Page 1


BLUK077-Stein

April 14, 2007

15:29

Pulmonary Embolism

i


BLUK077-Stein

April 14, 2007

15:29

Pulmonary Embolism Second Edition by

Paul D. Stein,

MD

Director of Research Education St. Joseph Mercy-Oakland Pontiac, Michigan, USA Professor, Full Time Affiliate Department of Medicine Wayne State University School of Medicine Detroit, Michigan, USA Adjunct Professor of Medical Physics Oakland University Rochester, Michigan, USA

iii


BLUK077-Stein

April 14, 2007

15:29

 C 1996 by Williams & Wilkins, Maryland  C 2007 by Paul D. Stein

Published by Blackwell Publishing Blackwell Futura is an imprint of Blackwell Publishing Blackwell Publishing, Inc., 350 Main Street, Malden, Massachusetts 02148-5020, USA Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK Blackwell Science Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia All rights reserved. No part of this publication may be reproduced in any form or by any electronic or mechanical means, including information storage and retrieval systems, without permission in writing from the publisher, except by a reviewer who may quote brief passages in a review. First published 1996 Second edition 2007 1

2007

ISBN: 978-1-4051-3807-9 Library of Congress Cataloging-in-Publication Data Stein, Paul D. Pulmonary embolism / by Paul D. Stein. – 2nd ed. p. ; cm. Includes bibliographical references and index. ISBN 978-1-4051-3807-9 (alk. paper) 1. Pulmonary embolism. I. Title. [DNLM: 1. Pulmonary Embolism. WG 420 s819p

2007]

RC776.P85s74 2007 616.2 49–dc22 2007005029 A catalogue record for this title is available from the British Library Commissioning Editors: Steve Korn and Gina Almond Development Editor: Beckie Brand Editorial Assistant: Victoria Pittman Set in 9/12 Minion and Frutiger by Aptara Inc., New Delhi, India Printed and bound in Singapore by Markono Print Media Pte. Ltd. Cost for publication of this book was supported in part by unrestricted grants from Diatide, Inc., Londonderry, New Hampshire and Dupont Pharmaceuticals Co., Wilmington, Delaware. For further information on Blackwell Publishing, visit our website: www.Blackwellmedicine.com The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards. Blackwell Publishing makes no representation, express or implied, that the drug dosages in this book are correct. Readers must therefore always check that any product mentioned in this publication is used in accordance with the prescribing information prepared by the manufacturers. The author and the publishers do not accept responsibility or legal liability for any errors in the text or for the misuse or misapplication of material in this book.

iv


BLUK077-Stein

April 14, 2007

15:29

Contents

Preface, ix Part I Prevalence, risks, and prognosis of pulmonary embolism and deep venous thrombosis

1 Pulmonary embolism and deep venous thrombosis at autopsy, 3 2 Incidence of pulmonary embolism and deep venous thrombosis in hospitalized patients, 16 3 Case fatality rate and population mortality rate from pulmonary embolism and deep venous thrombosis, 19 4 Prognosis in acute pulmonary embolism based on right ventricular enlargement, prognostic models, and biochemical markers, 24 5 Changing risks of untreated deep venous thrombosis and acute pulmonary embolism, 31 6 Resolution of pulmonary embolism, 35 7 Upper extremity deep venous thrombosis, 37 8 Thromboembolic disease involving the superior vena cava and brachiocephalic veins, 41 9 Venous thromboembolic disease in the four seasons, 44 10 Regional differences in the United States of rates of diagnosis of pulmonary embolism and deep venous thrombosis and mortality from pulmonary embolism, 47 11 Venous thromboembolism in the elderly, 52

12 Pulmonary thromboembolism in infants and children, 66 13 Venous thromboembolism in men and women, 68 14 Comparison of the diagnostic process in black and white patients, 72 15 Pulmonary thromboembolism in Asians/Pacific Islanders, 76 16 Pulmonary thromboembolism in American Indians and Alaskan Natives, 83 17 Venous thromboembolism in patients with cancer, 85 18 Venous thromboembolism in patients with heart disease, 93 19 Venous thromboembolism in patients with ischemic and hemorrhagic stroke, 98 20 Pulmonary embolism and deep venous thrombosis in hospitalized adults with chronic obstructive pulmonary disease, 101 21 Pulmonary embolism and deep venous thrombosis in hospitalized patients with asthma, 107 22 Deep venous thrombosis and pulmonary embolism in hospitalized patients with sickle cell disease, 109 23 Venous thromboembolism in pregnancy, 113 24 Air travel as a risk for pulmonary embolism and deep venous thrombosis, 119 25 Estrogen-containing oral contraceptives and venous thromboembolism, 122

v


BLUK077-Stein

April 14, 2007

15:29

vi

26 Obesity as a risk factor in venous thromboembolism, 125 27 Hypercoagulable syndrome, 128 Part II Diagnosis of deep venous thrombosis

28 Deep venous thrombosis of the lower extremities: clinical evaluation, 139 29 Clinical model for assessment of deep venous thrombosis, 144 30 Clinical probability score plus single negative ultrasound for exclusion of deep venous thrombosis, 147 31 D-dimer for the exclusion of acute deep venous thrombosis, 149 32 D-dimer combined with clinical probability assessment for exclusion of acute deep venous thrombosis, 158 33 D-dimer and single negative compression ultrasound for exclusion of deep venous thrombosis, 160

Contents

42 The history and physical examination in all patients irrespective of prior cardiopulmonary disease, 192 43 Clinical characteristics of patients with acute pulmonary embolism stratified according to their presenting syndromes, 197 44 Clinical assessment in the critically ill, 203 45 The electrocardiogram, 206 46 The plain chest radiograph, 216 47 Arterial blood gases and the alveolar–arterial oxygen difference in acute pulmonary embolism, 221 48 Fever in acute pulmonary embolism, 229 49 Leukocytosis in acute pulmonary embolism, 232 50 Alveolar dead-space in the diagnosis of pulmonary embolism, 234 51 Neural network computer-assisted diagnosis, 236

34 Contrast venography, 161

52 Empirical assessment and clinical models for diagnosis of acute pulmonary embolism, 239

35 Compression ultrasound for the diagnosis of deep venous thrombosis, 164

53 D-dimer for the exclusion of acute pulmonary embolism, 243

36 Impedance plethysmography and fibrinogen uptake tests for diagnosis of deep venous thrombosis, 168

54 D-dimer combined with clinical probability for exclusion of acute pulmonary embolism, 250

37 Computed tomography for diagnosis of deep venous thrombosis, 171

55 D-dimer in combination with amino-terminal pro-B-type natriuretic peptide for exclusion of acute pulmonary embolism, 253

38 Magnetic resonance angiography for diagnosis of deep venous thrombosis, 175 39 P-selectin and microparticles to predict deep venous thrombosis, 179

56 Low tissue plasminogen activator plasma levels and low plasminogen activator inhibitor-1 levels as an aid in exclusion of acute pulmonary embolism, 254

Part III Diagnosis of acute pulmonary embolism

57 Echocardiogram in the diagnosis and prognosis of acute pulmonary embolism, 255

40 Clinical characteristics of patients with no prior cardiopulmonary disease, 183

58 Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism, 260

41 Relation of right-sided pressures to clinical characteristics of patients with no prior cardiopulmonary disease, 190

59 Techniques of perfusion and ventilation imaging, 262


BLUK077-Stein

April 14, 2007

15:29

vii

Contents

60 Ventilation–perfusion lung scan criteria for interpretation prior to the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED), 267 61 Observations from PIOPED: ventilation–perfusion lung scans alone and in combination with clinical assessment, 271 62 Ventilation–perfusion lung scans in patients with a normal chest radiograph, patients with no prior cardiopulmonary disease, patients with any prior cardiopulmonary disease, and patients with chronic obstructive pulmonary disease, 278 63 Perfusion lung scans alone in acute pulmonary embolism, 280 64 Probability interpretation of ventilation–perfusion lung scans in relation to largest pulmonary arterial branches in which pulmonary embolism is observed, 282 65 Revised criteria for evaluation of lung scans recommended by nuclear physicians in PIOPED, 284

72 Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries, 318 73 Quantification of pulmonary emboli by conventional and CT angiography, 319 74 Complications of pulmonary angiography, 321 75 Contrast-enhanced spiral CT for the diagnosis of acute pulmonary embolism before the Prospective Investigation of Pulmonary Embolism Diagnosis, 325 76 Methods of PIOPED II, 340 77 Multidetector spiral CT of the chest for acute pulmonary embolism: results of the PIOPED II trial, 348 78 Outcome studies of pulmonary embolism versus accuracy, 355 79 Contrast-induced nephropathy, 357 80 Radiation exposure and risk, 359 81 Magnetic resonance angiography for the diagnosis of acute pulmonary embolism, 364

66 Criteria for very low probability interpretation of ventilation–perfusion lung scans, 288

82 Serial noninvasive leg tests in patients with suspected pulmonary embolism, 371

67 Probability assessment based on the number of mismatched segmental equivalent perfusion defects or number of mismatched vascular defects, 294

83 Predictive value of diagnostic approaches to venous thromboembolism, 373

68 Probability assessment based on the number of mismatched vascular defects and stratification according to prior cardiopulmonary disease, 298 69 The addition of clinical assessment to stratification according to prior cardiopulmonary disease further optimizes the interpretation of ventilation–perfusion lung scans, 304 70 Single photon emission computed tomographic perfusion lung scan, 310 71 Standard and augmented techniques in pulmonary angiography, 311

84 Diagnostic approaches to acute pulmonary embolism, 376 Part IV Prevention and treatment of deep venous thrombosis and pulmonary embolism

85 New and old anticoagulants, 389 86 Prevention of deep venous thrombosis and pulmonary embolism, 405 87 Treatment of deep venous thrombosis and acute pulmonary embolism, 414 88 Withholding treatment of patients with acute pulmonary embolism who have a high risk of bleeding provided and negative serial noninvasive leg tests, 422


BLUK077-Stein

April 14, 2007

15:29

viii

89 Thrombolytic therapy in acute pulmonary embolism, 425 90 Thrombolytic therapy for deep venous thrombosis, 437 91 Inferior vena cava filters: trends in use, complications, indications, and use of retrievable filters, 444 92 Catheter-tip embolectomy in the management of acute massive pulmonary embolism, 454

Contents

93 Pulmonary embolectomy, 459 94 Chronic thromboembolic pulmonary hypertension and pulmonary thromboendarterectomy, 464 Index, 467


BLUK077-Stein

April 14, 2007

15:29

Preface

Since the first edition of Pulmonary Embolism was published in 1996, major strides have been made in many aspects of pulmonary embolism and its immediate cause, deep venous thrombosis. The purpose of this second edition is to provide detailed, in depth information on pulmonary embolism in a format that is readily usable by practicing physicians, and at the same time, provide fully referenced data that can serve as a resource for physicians with a deeper interest in the field. Many associates contributed to the investigations upon which much of the information in this book is based. The PIOPED I investigators supplied data for an impeccable database used for much of what we know about the accuracy of clinical assessment as well as ventilation–perfusion lung scans. The PIOPED II investigators supplied an equally impeccable database upon which much of what we know about the accuracy of multidetector CT angiography is based. This database is just beginning to be examined for useful information on a variety of additional subjects. The PIOPED III investigators are just starting to evaluate magnetic resonance angiography, and hope to acquire an equally impressive database. Grants from the National Institutes of Health, National Heart Lung and Blood Institute (NHLBI) made all of this possible. Guidance by representatives of the NHLBI contributed to the success of the study. Many of the investigators who participated in PIOPED I continued through PIOPED II and into PIOPED III. Some even participated in the Urokinase-Pulmonary Embolism Trial, published in 1973, which contributed a huge amount of information about pulmonary embolism, beyond that which was learned about thrombolytic therapy. Investigators with preferences for diverse approaches to the diagnosis and management of pulmonary embolism subverted their personal interests to collabo-

rate on scientific levels. This resulted in collaborations with only one goal: advancement of the field. The investigators, including physicians, statisticians, nursecoordinators, and technicians, often contributed long and hard hours with little reward. I am thankful for their efforts, and for the deep friendships with many that have resulted from these collaborations. Another database that was used extensively for epidemiological information related to pulmonary embolism and deep venous thrombosis was The National Hospital Discharge Summary. This database is available to the general public. Many details about its correct use, however, required consultation with representatives of the National Center for Health Statistics, and this assistance was graciously given. Several bright young men have worked with me over the years. Jerald W. Henry, MD, worked for several years on obtaining data from PIOPED I before going to medical school. He is now a practicing radiologist. More recently, Kalpesh C. Patel, MBBS, and Neeraj K. Kalra, MD, assisted. They are now finishing subspecialty training. Fadi Kayali, MD, did dedicated and superior work. He has been accepted into fellowship training. Afzal Beemath, MD, is a brilliant former fellow. He not only contributed importantly to several investigations, but also helped in a major way in the completion of this book. Nikunj Patel, MD, also worked long hours in assisting in the preparation of this book, as did his brother Hiren who assisted for a few months. Fadi Matta, MD, and Abdo Yaekoub, both of whom recently started with me, have done a sensational job. Finally, thanks to Steven Korn, formerly of Futura Publishing Company and Beckie Brand of Blackwell Publishing Company for encouraging me to complete this labor.

ix


BLUK077-Stein

March 12, 2007

I

14:45

PART I

Prevalence, risks, and prognosis of pulmonary embolism and deep venous thrombosis

1


BLUK077-Stein

March 12, 2007

1

14:45

CHAPTER 1

Pulmonary embolism and deep venous thrombosis at autopsy

Prevalence of pulmonary embolism at autopsy The prevalence of pulmonary embolism (PE) at autopsy varies according to the age and morbidity of the population studied. Dalen and Alpert in 1975 estimated that 15% of deaths in acute general hospitals and 25% of deaths in nursing homes or chronic hospitals were due to PE [1]. In more recent years, with more extensive use of antithrombotic prophylaxis, PE at autopsy was shown with similar prevalences among patients who died in acute care hospitals (24%) and patients who died in chronic care hospitals (22%) [2]. Outpatients, however, had a lower prevalence of PE at autopsy (5%) [2]. The prevalence of PE at autopsy of patients in general hospitals and in entire communities, with one exception, ranged from 9 to 28% and has not changed in over 60 years [2–21] (Table 1.1). One study, however, reported gross PE in 55% of patients at autopsy [10]. On average, PE at autopsy occurred in 7031 of 55,090 patients (13%) (Table 1.1, Figure 1.1).

Small PE at autopsy In an autopsy study that employed postmortem pulmonary arteriography as well as gross dissection and microscopic examination, gross dissection showed PE in 34 of 225 (15%) of autopsied patients [4]. Among these, PE was limited to muscular pulmonary artery branches (0.1–1 mm diameter) in 26 of 34 patients (76%) and PE was in elastic pulmonary artery branches (>1 mm diameter) in 8 of 34 patients (24%) [4]. Microscopic examination showed PE in pulmonary arterioles in 13 of 34 patients (38%) with grossly visible PE. The smallest PE that have been identified in living patients were with wedge pulmonary arteriography, which showed PE in 1–2-mm-diameter pulmonary artery branches [24] (see Chapter 71). Fibrous bands, webs, and intimal fibrosis have been interpreted as the final state of organization of PE and these have been reported by some to indicate old PE at autopsy [7]. Meticulous dissection and microscopic examination for minute and barely visible fragments showed traces of fresh or old PE at autopsy in 52% and 64% of patients [7, 8].

Large or fatal PE at autopsy Large or fatal PE in patients at autopsy in general hospitals or communities from 1939 to 2000 occurred in 2264 of 54,364 patients (4%) (range 0.3–24%) [2, 3, 8–11, 13–19, 21, 22] (Table 1.1, Figure 1.1). In most studies, the prevalence of large or fatal PE ranged from 3 to 10%. In elderly institutionalized patients, the rate of fatal PE at autopsy was within that range, 18 of 234 (8%) [23]. Data on institutionalized patients are not included in Table 1.1. A sudden increase in the rate of PE at autopsy was observed in London in 1940 due to cramped conditions in air raid shelters [22]. These rates also are not included in Table 1.1.

Unsuspected PE at autopsy Pulmonary embolism was unsuspected or undiagnosed antemortem in 3268 of 3876 patients in general hospitals or communities who had PE at autopsy (84%) (range 80–93%) [3, 5, 8, 11, 12, 16, 18] (Table 1.2, Figure 1.2). Remarkably, even in patients with large or fatal PE at autopsy, the majority, 1902 of 2448 (78%), were unsuspected or undiagnosed antemortem [2, 11, 12, 14–16, 18, 19, 25] (Table 1.2, Figure 1.2). In our experience, PE at autopsy caused death in 5%, contributed to death in 0.5%, and was incidental in 9.2% of 404 autopsies, and the distribution, according

3


BLUK077-Stein

March 12, 2007

14:45

4

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 1.1 Prevalence of pulmonary embolism at autopsy in general hospitals and communities. Any PE/No autopsies (%)

Fatal or large PE/No autopsies (%)

Study years

First author, year [Ref] Simpson, 1940 [22]

4/242 (2)

1939

606/4391 (14)

1945–1954

Coon, 1959 [3]

34/225 (15)

1960–1961

Smith, 1964 [4]

118/981 (12)

1956–1960

Uhland, 1964 [5]

17/61 (28)*

1951–1959

Freiman, 1965 [6]

55/263 (21)†

1964–1965

Morrell, 1968 [7]

567/4600 (12)

202/4600 (4)

1964–1974

Coon, 1976 [8]

319/1350 (24)

1976

Schwarz, 1976 [9]

280/508 (55)‡

92/508 (18)

1969–1970

Havig, 1977 [10]

216/1455 (15)

54/1455 (4)

1973–1974

Goldhaber, 1982 [11]

389/2398 (16)

1966–1976

Dismuke, 1984 [12]

105/1133 (9)

1966–1970

Dismuke, 1986 [13]

53/1124 (5)

1971–1975

””

43/1128 (4)

1976–1980

””

44/1276 (3)

1980–1984

313/2388 (13)

239/2388 (10)

1979–1983

Rubenstein, 1988 [14] Sandler, 1989 [15]

1934/21,529 (9)

67/21,529 (0.3)

1960–1984

Karwinski, 1989 [16] Linblad, 1991 [17]

161/766 (21)

68/766 (9)

1957

250/1117 (22)

93/1117 (8)

1964

””

346/1412 (25)

83/1412 (6)

1975

””

260/994 (26)

93/994 (9)

1987

59/404 (15)

20/404 (5)

1985–1986

Stein, 1995 [18]

””

92/2427 (4)

1985–1989

Morgenthaler, 1995 [19]

288/3334 (9)§

1966–1974

Mandelli, 1997 [20]

182/1144 (16)§

1989–1994

431/2356 (18)

178/2356 (8)

1987

Nordstrom, 1998 [2]

525/3764 (14)

221/3764 (6)

1980–2000

Pheby, 2002 [21]

””

* An additional 22/61 (36%) showed traces of residual pulmonary embolism (PE), fibrous bands, or webs. †

An additional 31% had had fibrous bands or intimal fibrosis indicative of old PE. An additional 72 of 508 (14%) were visible only by microscopy. § Massive and submassive PE.

PE, DVT at autopsy (%)

50 45 40 35 30 25 20 15 10 5 0

43

13 4

Any PE

Large or fatal PE

Any DVT

Figure 1.1 Prevalence of pulmonary embolism (PE) and deep venous thrombosis (DVT) at autopsy.


BLUK077-Stein

March 12, 2007

CHAPTER 1

14:45

5

PE and DVT autopsy

Table 1.2 Unsuspected pulmonary embolism at autopsy. Any unsuspected or

Unsuspected or

undiagnosed PE

undiagnosed minor or

Unsuspected or undiagnosed fatal or large PE

[unsuspected PE/total

small PE [unsuspected

[unsuspected large PE/total

PE (%)]

small PE/total PE (%)]

PE (%)]

Study years

First author, year [Ref]

563/606 (93)

1945–1954

Coon, 1959 [3]

91/107 (85)

1955–1960

Uhland, 1964 [5]

514/567 (91)

1964– 1974

Coon, 1976 [8]

199/217 (92)

161/162 (99)

38/54 (70)

1973–1974

Goldhaber, 1982 [11]

310/389 (80)

219/244 (90)

91/145 (63)

1966–1976

Dismuke, 1984 [12]

30/44 (68)

1980–1984

Rubenstein, 1988 [14]

186/195 (95)

1979–1983

Sandler, 1989 [15]

1619/1934 (84)

436/484 (90)

1183/1450 (82)

1960–1984

Karwinski, 1989 [16] Stein, 1995 [18]

52/59 (88)

36/37 (97)

14/20 (70)

1985–1986

47/92 (51)

1985–1989

Morgenthaler, 1995 [19]

189/279 (68)

1987

Nordstrom, 1998 [2]

124/169 (73)

1995–2002

Attems, 2004* [25]

* All patients ≥70 years old. PE, pulmonary embolism.

to whether diagnosed and treated, suspected but not diagnosed or treated, or unsuspected is shown in Table 1.3 [18]. Many patients with unsuspected large or fatal PE had advanced associated disease [18]. Patients who suffer sudden and unexplained catastrophic events in the hospital are a group in whom the diagnosis might be suspected more frequently if physicians maintain a high index of suspicion [18].

Rate and sequence of organization of thromboemboli

Unsuspected PE/total PE (%)

A thrombus contains extensive regions of masses of agglutinated platelets [26]. Platelets are deposited first,

Figure 1.2 Prevalence of unsuspected pulmonary embolism (PE) at autopsy.

90

followed by leukocytes, followed after a variable period of time by fibrin with trapped red cells and a few scattered leukocytes [26]. The rate of organization of thromboemboli has been assessed in rabbits [27, 28]. The following results were shown [27, 28]: 8 minutes. Thrombus covered by an eosinophilic rim of platelets. Small amounts of fibrin were interspersed among the platelets at the edge of the thrombus [28]. 3 days. Thrombi contained masses of red cells, fibrin, platelets, and white cells together with a number of macrophages. Parts of the surface not in contact with the vessel wall were covered by flattened cells and in places these were buttressed by a layer of elongated cells beneath. Platelets were particularly

84 78

75 60 45 30 15 0

Patients with any unsuspected PE

Patients with unsuspected large or fatal PE


BLUK077-Stein

March 12, 2007

14:45

6

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 1.3 Autopsy patients ≥18 years old (n = 404). PE caused death (%)

PE contributed to death (%)

PE incidental (%)

PE total (%)

Diagnosed and treated

3 (0.7)

0 (0)

1 (0.2)

4(1.0)

Suspected but not

3 (0.7)

0 (0)

0 (0)

3 (0.7)

Unsuspected

14 (3.5)

2 (0.5)

36 (8.9)

52(12.9)

Total

20 (5.0)

2 (0.5)

37 (9.2)

59 (14.6)

diagnosed or treated

Modified from Stein and Henry [18] and reproduced with permission. PE, pulmonary embolism.

Table 1.4 Deep venous thromboses; autopsies with full limb dissection. DVT n/N (%)

Site (number of thrombi)

95/324 (29)

Site (number of patients)

First author, year [Ref]

Thighs or pelvis 7

Rossle, 1937 [29]

Thighs and Calves 38 Calves only 50 100/165 (61)

Thighs 22

Neumann, 1938 [30]

Calves 87 Ankle 17 Foot 71 88/200 (44)

Thighs only 3

Hunter, 1945 [31]

Thighs and Calves 28 Calves only 57 35/130 (27) 32/100 (32)*

Raeburn, 1951 [32] Thighs only 18

McLachin, 1962 [33]

Thighs and Calves 10 Calves only 4 149/253 (59)

Thighs only 24

Gibbs, 1957 [34]

Thighs and Calves 39 Calves only 86 13/27 (48)

IVC 1

Thighs only 1

Pelvic 1

Thighs and Calves 7

Thigh 23

Calves only 5

Stein, 1967 [35]

Calves 35† 540/1350 (40)

Pelvic 41‡

Schwarz, 1976 [9]

Thigh 21 Calves 74 161/261 (62)

IVC 8 Pelvic 31 Thigh 129 Calves 128 Foot 87

* Males >40 years old. †

Calf 11 microscopic thrombi in addition. Sample of 37 patients. DVT, deep venous thrombosis; n, number of patients with DVT; N, number of patients necropsied. ‡

Havig, 1977 [10]


BLUK077-Stein

March 12, 2007

CHAPTER 1

14:45

PE and DVT autopsy

7

Figure 1.3 Extensive antemortem thrombus located in popliteal and calf veins. Previously unpublished figure from Stein and Evans [35].

prevalent near the thrombus–vessel wall junction. Mononuclear cells were prominent [27]. 5 and 7 days. Beginnings of vascularization were apparent. Capillaries were within the thrombus mass and in cellular areas of attachment to the intima. The central area of the thrombus showed mainly debris [27]. 7 days. Occluding thrombi had retracted in places and were covered by flattened cells, and showed one or more firm cellular attachments to the intima. Macrophages were conspicuous and contained lipid, fibrin, and cellular debris together with fibroblastic cells [27]. 14 days. Thrombi consisted of cellular masses containing small clumps of fibrin and variable amounts of fat and fibrous tissue [27]. 20 days. Some thrombi appeared as polypoid masses protruding into the lumen and containing variable amounts of fat, fibrous, and elastic tissue, and on occasion calcium, while others showed lipid within foamy cells and a fibrous tissue cap containing fibroblasts, collagen, and elastic tissue [27]. 30 days. Thromboemboli were converted to eccentric fibrofatty thickenings of the intima [27].

Deep venous thrombosis at autopsy Data on patients who had complete dissection of the lower extremities at autopsy are from prior decades,

and before the general use of antithrombotic prophylaxis [10, 29–36]. Among patients at autopsy who had full limb dissection, 1213 of 2810 (43%) showed deep venous thrombosis (DVT) [10, 29–36] (Table 1.4, Figures 1.1 and 1.3). Among 161 patients with DVT at autopsy, 7 patients had thrombi in the common iliac vein and 22 had thrombi in the external iliac vein. Each of these patients also showed DVT in the femoral vein [10]. The external iliac vein showed thrombi in 12 of 161 patients (7%) without femoral vein involvement. In 4 of these patients, the calf veins showed DVT, but not the femoral veins [10]. Deep venous thrombosis affected the veins of the calves more frequently than the veins of the thighs, and both were more frequently affected than the veins of the pelvis. The distribution of 601 thrombi found in 311 patients who had dissection of the pelvic, thigh, and calf veins was 54% in the veins of the calves, 32% in the veins of the thighs, 12% in the pelvic veins, and 1% in the inferior vena cava [10, 30, 35, 36] (Figure 1.4). The distribution of 563 thrombi among 261 necropsied patients who had dissection of the veins of the foot as well as the veins of the calf, thigh, and pelvis was 28% in the veins of the foot, 38% in the calf, 27% in the thigh, 6% in the pelvic veins, and 1% in the inferior vena cava (IVC) [10, 30] (Figure 1.5). Among 282 necropsied patients who had complete dissection of the veins of the thighs and veins of the calves, the thrombi were located only in the veins of the calves in 54% of patients [31, 33–35] (Figure 1.6).


BLUK077-Stein

March 12, 2007

14:45

60

54

Thrombi (%)

50 40 32 30 20 12 10 0

1 IVC

Pelvis

Calf

Thigh

40

Figure 1.4 Distribution of deep venous thrombosis among patients at autopsy in whom pelvic, thigh, and calf veins were dissected.

38

35 Thrombi (%)

30

28

27

25 20 15 10 6 5 1 0 IVC

Pelvis

Thigh

Calf

Foot

Figure 1.5 Distribution of deep venous thrombosis among patients at autopsy in whom veins of the foot as well as pelvic, thigh, and calf veins were dissected.

Patients (%)

60

54

50 40 30

30 20

16

10 0 Thigh only

Thighs and calves

Calves only

Figure 1.6 Percentage of patients at autopsy with deep venous thrombosis who had involvement of veins of thigh only, veins of thighs and calf veins, and veins of calf only.

Figure 1.7 Normal postmortem venogram of calf (lateral projection) showing anterior tibial (AT), posterior tibial (PT), and peroneal (Pe) veins. The deep veins are paired. (Reproduced from Stein and Evans [35], with permission.)

8


BLUK077-Stein

March 12, 2007

CHAPTER 1

14:45

PE and DVT autopsy

9

Both the veins of the thighs and calves were affected in 30% of patients. Only the veins of the thighs showed DVT in 16% of patients. Bilateral DVT was observed in 81 of 96 patients (84%) with extensive DVT at autopsy and in 26 of 65 (40%) of patients with minor DVT at autopsy [10]. Postmortem venography illustrates the extent and location of DVT at autopsy in unselected patients [35]. For comparison, normal postmortem venograms of the calf and thighs are shown (Figures 1.7 and 1.8). Postmortem venograms of DVT involving the veins of the thighs are shown in Figures 1.9 and 1.10. Figure 1.8 Normal postmortem venogram of the thighs (anteroposterior projection) showing the femoral (F), deep femoral (DF), greater saphenous (GS), and popliteal (P) veins. Valve pockets are shown. (Reproduced from Stein and Evans [35], with permission.)

Figure 1.9 Postmortem venogram of the veins of both thighs. Extensive thrombosis of the femoral, deep femoral, and popliteal veins was found by dissection of the left thigh. The venogram of the left thigh shows absence of filling of the popliteal and deep femoral veins and only a faint outline of the femoral vein (F). The left greater saphenous vein is dilated and joined by numerous collateral vessels. The veins of the right thigh were normal. (Reproduced from Stein and Evans [35], with permission.)

Figure 1.10 Postmortem venogram of right thigh. The femoral vein has not filled with contrast material because of a completely occluding thrombus. The greater saphenous (GS) vein is distended. Collateral vessels formed at the site of an occluding thrombus in the greater saphenous vein (arrow). (Reproduced from Stein and Evans [35], with permission.)


BLUK077-Stein

March 12, 2007

14:45

10

PART I

Prevalence, risks, and prognosis of PE and DVT

Figure 1.11 Organized thrombus in anterior tibial vein (same patient as Figure 1.12). This thrombus is older than the thrombus in the femoral vein, and there is no phlebitis here. Hematoxylin and eosin ×40. (Previously unpublished figure from Stein and Evans [35].)

Forward thrombosis versus retrograde thrombosis

Thrombophlebitis and phlebothrombosis

In every case that we examined in which the veins of the thigh and the calf showed DVT in continuity, the thrombi in the calf were older than those in the thigh [35] (Figures 1.11 and 1.12). This supports the concept that forward thrombosis is more common than retrograde thrombosis.

The terms “thrombophlebitis” and “phlebothrombosis” in prior years were used to distinguish between DVT associated with inflammation (thrombophlebitis) and DVT not associated with inflammation (phlebopthrombosis). These are outdated terms. Histological investigations have not supported a distinction between the clinical diagnoses of thrombophlebitis and phlebothrombosis. Thrombosis of the veins of the lower extremities usually occurs without inflammation [35] (Figures 1.11 and 1.14–1.16). Inflammation of the walls of the veins, when it occurs (Figure 1.12), is usually secondary to the thrombosis [35]. No clear evidence indicates that inflammation

Collateral veins around occlusions Clinically unsuspected DVT at autopsy was often extensive, causing collateral circulation around occlusions and dilatation of collateral veins [35] (Figures 1.10 and 1.13).

Figure 1.12 Thrombus attached to femoral vein (same patient as Figure 1.11). Lymphocytic infiltrate is shown throughout the wall of the vein. The patient had signs and symptoms of deep venous thrombosis. Hematoxylin and eosin ×13. (Previously unpublished figure from Stein and Evans [35].)


BLUK077-Stein

March 12, 2007

CHAPTER 1

14:45

PE and DVT autopsy

Figure 1.13 Postmortem venogram of the thighs. There is definite radiographic evidence of occlusion of the femoral vein between points 1 and 2. There is no filling of the femoral vein (F) between these points. Dilated and tortuous collaterals pass around the site of occlusion. No thrombus was found on dissection of the veins of the thigh of this patient, presumably because dissection was carried out along the collateral vessels in this area rather than the femoral vein. (This apparent femoral vein occlusion was not included among the positive cases reported in Stein and Evans [35].)

Figure 1.14 Recent thrombus attached in vein of soleal plexus. Hematoxylin and eosin ×16. (Previously unpublished figure from Stein and Evans [35].)

11

of the veins prevents embolization, or that embolization is more frequent in those patients with thrombi not associated with venous inflammation. The distinction between “thrombophlebitis” and “phlebothrombosis” is of no clinical consequence [35]. A thrombus can induce inflammation in the underlying wall of the vein, and this inflammation in some patients is extensive enough to produce pain, tenderness, swelling, and fever compatible with the clinical diagnosis of thrombophlebitis [36]. However, the underlying pathogenic mechanism is primary thrombosis and not primary phlebitis [36]. The following historical background explains the evolution of these outdated diagnostic terms. John Hunter, after studying infected venesections in human beings and in horses, attributed the thrombosis to phlebitis [37]. Virchow, however, observed that the cellular reaction in the wall of the vein usually does not occur until after the thrombus has been laid down [38]. Welch [39], in studying DVT in patients with infectious diseases such as typhoid fever, found an inflammatory lesion beneath the endothelium in which he could not demonstrate any organisms. He termed this “toxic endophlebitis” and attributed some instances of DVT to inflammation of the veins. Subsequently, patients were described who had clinical evidence of thrombosed leg veins and also had clinical signs of inflammation (warmth, redness, tenderness). A diagnosis of thrombophlebitis was made. In view of Welch’s observations, it was concluded that the primary event was inflammation of the wall of the vein. In contrast, asymptomatic patients were later described who had thrombosis of


BLUK077-Stein

12

March 12, 2007

14:45

PART I

Prevalence, risks, and prognosis of PE and DVT

Figure 1.15 Fresh unattached thrombus in fomoral vein. Lines of Zahn distinguish this from postmortem clot. Hematoxylin and eosin ×4. (Previously unpublished figure from Stein and Evans [35].)

Figure 1.16 Photomicrograph showing thrombus originating in valve pocket of a posterior tibial vein. The well-organized fibrous point of attachment is capped by a fresh red cell, platelet, and fibrin clot. There is no inflammation of the vein. Hematoxylin and eosin ×4. (Previously unpublished figure from Stein and Evans [35].)

Figure 1.17 Thrombus attached to valve pocket in femoral vein and propagating along the vein. Venous valve is shown (arrow). Hematoxylin and eosin ×10. (Previously unpublished figure from Stein and Evans [35].)


BLUK077-Stein

March 12, 2007

CHAPTER 1

14:45

PE and DVT autopsy

13

Figure 1.18 Section of left posterior tibial vein. An antemortem thrombus, 0.2 cm in largest dimension, is located within a valve pocket. (Previously unpublished figure from Stein and Evans [35].)

the lower extremities that resulted in PE [40]. These patients, because of the lack of leg signs, were said to have phlebothrombosis. Although there are situations in which phlebitis is primary and thrombosis is secondary (such as mechanical and chemical injury) [36], these are rarely compared with the incidence of thrombosis without inflammation [31, 36]. In patients with DVT at autopsy, fresh components of the thrombus as well as older components were shown, indicating that the thrombosis was continuing [35] (Figure 1.16). None of the patients were diagnosed antemortem as having DVT. A patient with clinical signs and symptoms of DVT showed lymphocytic infiltration in the media of the veins (Figure 1.12). The inflammation occurred not only at the sites of attachment of the thrombus, but also where the thrombus was apposed to the endothelium without being attached, suggesting that the thrombus induced the inflammation.

Valve pockets as site of origin of DVT The valve pockets were a frequent site of origin of thrombi (Figures 1.16–1.18). Thrombi located in valve pockets consisted of organized fibrous points of attachment capped by fresh fibrin and red cell clot [35] (Figure 1.16). Dilated veins and enlarged valve pockets were frequently seen (Figure 1.19). There was no correlation of either of these abnormalities with the presence of thrombosis [35].

Figure 1.19 Postmortem venogram showing dilated valve pocket in femoral (F) vein of left thigh (arrow). The deep femoral vein (DF) is also shown. (Reproduced from Stein and Evans [35], with permission.)


BLUK077-Stein

March 12, 2007

14:45

14

PART I

References

18 Stein PD, Henry JW. Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest 1995; 108: 978–981. 19 Morgenthaler TI, Ryu JH. Clinical characteristics of fatal pulmonary embolism in a referral hospital. Mayo Clin Proc 1995; 70: 417–424. 20 Mandelli V, Schmid C, Zogno C, Morpurgo M. “False negatives” and “false positives” in acute pulmonary embolism: a clinical–postmortem comparison. Cardiologia 1997; 42: 205–210. 21 Pheby DF, Codling BW. Pulmonary embolism at autopsy in a normal population: implications for air travel fatalities. Aviat Space Environ Med 2002; 73: 1208–1214. 22 Simpson K. Shelter deaths from pulmonary embolism. Lancet 1940; 2: 744. 23 Gross JS, Neufeld RR, Libow LS, Gerber I, Rodstein M. Autopsy study of the elderly institutionalized patient. Review of 234 autopsies. Arch Intern Med 1988; 148: 173–176. 24 Stein PD. Wedge arteriography for the identification of pulmonary emboli in small vessels. Am Heart J 1971; 82: 618–623. 25 Attems J, Arbes S, Bohm G, Bohmer F, Lintner F. The clinical diagnostic accuracy rate regarding the immediate cause of death in a hospitalized geriatric population; an autopsy study of 1594 patients. Wien Med Wochenschr 2004; 154; 159–162. 26 Poole JC, French JE, Cliff WJ. The early stages of thrombosis. J Clin Pathol 1963; 16: 523–528. 27 Still WJ. An electron microscopic study of the organization of experimental thromboemboli in the rabbit. Lab Invest 1966; 15: 1492–1507. 28 Thomas DP, Gurewich V, Ashford TP. Platelet adherence to thromboemboli in relation to the pathogenesis and treatment of pulmonary embolism. N Engl J Med 1966; 274: 953–956. 29 Rossle R. Uber die Bedeutung und die entstehung der wadenvenenthrombosen. Virchow Arch Path Anat 1937; 300: 180–189. 30 Neumann R. Ursprungszentren und entwicklungsformen der bein-thrombose. Virchow Arch Path Anat 1938; 301: 708–735. 31 Hunter WC, Krygier JJ, Kennedy JC, Sneeden VD. Etiology and prevention of thrombosis of the deep leg veins: a study of 400 cases. Surgery 1945; 17: 178–190. 32 Raeburn C. The natural history of venous thrombosis. BMJ 1951; 2: 517–520. 33 McLachlin J, Richards T, Paterson JC. An evaluation of clinical signs in the diagnosis of venous thrombosis. Arch Surg 1962; 85: 738–744. 34 Gibbs NM. Venous thrombosis of the lower limbs with particular reference to bed-rest. Br J Surg 1957; 45: 209– 236.

1 Dalen JE, Alpert JS. Natural history of pulmonary embolism. Prog Cardiovas Dis 1975; 17: 259–270. 2 Nordstrom M, Lindblad B. Autopsy-verified venous thromboembolism within a defined urban population— the city of Malmo, Sweden. Acta Path Microbiol Immunol Scand 1998; 106: 378–384. 3 Coon WW, Coller FA. Clinicopathologic correlation in thromboembolism. Surg Gynecol Obstet 1959; 109: 259– 269. 4 Smith GT, Dammin GJ, Dexter L. Postmortem arteriographic studies of the human lung in pulmonary embolization. JAMA 1964; 188: 143–151. 5 Uhland H, Goldberg LM. Pulmonary embolism: a commonly missed clinical entity. Dis Chest 1964; 45: 533–536. 6 Freiman DG, Suyemoto J, Wessler S. Frequency of pulmonary thromboembolism in man. N Engl J Med 1965; 272: 1278–1280. 7 Morrell MT, Dunnill MS. The post-mortem incidence of pulmonary embolism in a hospital population. Br J Surg 1968; 55: 347–352. 8 Coon WW. The spectrum of pulmonary embolism: twenty years later. Arch Surg 1976; 111: 398–402. 9 Schwarz N, Feigl W, Neuwirth E, Holzner JH. Venous thromboses and pulmonary emboli in autopsy material. Wien Klin Wochenschr 1976; 88: 423–428. 10 Havig O. Deep venous thrombosis and pulmonary embolism. Chapters 2, 4: Pulmonary thromboembolism. Acta Chir Scand 1977; 478(suppl): 4–11, 24–37. 11 Goldhaber SZ, Hennekens CH, Evans DA, Newton EC, Godleski JJ. Factors associated with correct antemortem diagnosis of major pulmonary embolism. Am J Med 1982; 73: 822–826. 12 Dismuke SE, VanderZwaag R. Accuracy and epidemiological implications of the death certificate diagnosis of pulmonary embolism. J Chronic Dis 1984; 37: 67–73. 13 Dismuke SE, Wagner EH . Pulmonary embolism as a cause of death. The changing mortality in hospitalized patients. JAMA 1986; 255: 2039–2042. 14 Rubenstein I, Murray D, Hoffstein V. Fatal pulmonary emboli in hospitalized patients—an autopsy study. Arch Int Med 1988; 148: 1425–1426. 15 Sandler DA, Martin JF. Autopsy proven pulmonary embolism in hospital patients: are we detecting enough deep vein thrombosis? J R Soc Med 1989; 82: 203–205. 16 Karwinski B, Svendsen E. Comparison of clinical and post-mortem diagnosis of pulmonary embolism. J Clin Pathol 1989; 42: 135–139. 17 Lindblad B, Sternby NH, Bergqvist D. Incidence of venous thromboembolism verified by necropsy over 30 years. BMJ 1991; 302: 709–711.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

March 12, 2007

CHAPTER 1

14:45

PE and DVT autopsy

35 Stein PD, Evans H. An autopsy study of leg vein thrombosis. Circulation 1967; 35: 671–681. 36 Allen EV, Barker NW, Hines EA, Jr. Peripheral Vascular Diseases. WB Saunders, Philadelphia, 1962: 559– 569. 37 Hunter J. Observation on the inflammation of the internal coats of veins. Trans Soc Imp Med Chir Knowl 1793; 1: 18. Quoted by Stein and Evans in Reference [35].

15

38 Virchow R. Cellular Pathology as Based Upon Physiological and Pathological Histology. J. & A. Churchill, Ltd., London, 1860: 197–203. Quoted from Gibbs NM in Reference [34]. 39 Welch WH. Thrombosis. In: Allbutt TC, ed. A System of Medicine, Vol. 6. Macmillan, New York, 1899: 180. 40 Homans J. Thrombosis of the deep veins of the lower leg causing pulmonary embolism. New Engl J Med 1934; 211: 993.


BLUK077-Stein

2

March 12, 2007

14:48

CHAPTER 2

Incidence of pulmonary embolism and deep venous thrombosis in hospitalized patients

Introduction In 1999, 140,000 patients were discharged from nonfederal short-stay hospitals in the United States with a diagnosis of pulmonary embolism (PE) and 370,000 patients were discharged with deep venous thrombosis (DVT) [1]. In 2001, the number of patients diagnosed with PE increased to 156,000, and the number discharged with DVT remained the same [2]. Either PE or DVT was diagnosed in 466,000 hospitalized patients [2]. Throughout the entire United States, from 1979 to 2001, the number of patients discharged from short-stay nonfederal hospitals with PE was 2,741,000, with DVT 6,475,000, and with either venous thromboembolism (VTE) (either PE or DVT) 8,575,000 [3]. During this 23-year period, the average populationbased rate of diagnosis of PE in hospitalized patients was 47/100,000 population, the population-based rate of diagnosis of DVT was 112/100,000 population, and for VTE it was 148/100,000 population [2].

Incidence of PE in hospitals The National Hospital Discharge Survey showed the prevalence of PE in patients ≥20 years of age averaged over a 21-year period of study from 1979 to 1999 of 0.40% [4] (Figure 2.1). During this period there were 612,000,000 short-stay nonfederal hospital admissions throughout the United States [4]. These data were based entirely on discharge codes. The results were comparable to the incidence of PE in hospitalized patients as was shown by much smaller but more rigidly defined retrospective evaluations (0.27–0.4%) [5–8]. (National Hospital Discharge Survey. Available at: http://www.cdc.gov/nchs/about/major/hdasd/nhds .htm.)

16

The incidence of PE in hospitalized patients did not change over 21 years [4] (Figure 2.2). The incidence of PE in hospitalized patients was nearly the same in men and women (relative risk of men to women 1.11) [4] (Figure 2.3). The incidence of PE in hospitalized patients was the same in white and black patients (relative risk of white patients to black patients 1.00) [4] (Figure 2.4). The prevalence of PE in a general hospital, based on clinical diagnoses, many of which were confirmed at autopsy, in an era prior to pulmonary angiography or ventilation–perfusion scans was 0.2% [9]. The prevalence of acute PE in patients in a clinic of digestive surgery, diagnosed by pulmonary angiography, high probability V-Q scans or autopsy was 0.3% [10]. Using comparable criteria, we found the same prevalence (0.3%) [8]. The inclusion of patients estimated to have PE based on non-high-probability interpretations of the ventilation–perfusion lung scans and the inclusion of patients with clinically undiagnosed PE at autopsy caused the estimated prevalence of PE to be higher, 1.0% [5]. There are, in addition, patients with silent PE, the frequency of which is undetermined. On average, PE occurs in 13% of patients at autopsy, among whom the diagnosis was unsuspected antemortem in 84% (Chapter 1).

Incidence of DVT in hospitals Based on results of the National Hospital Discharge Survey, the prevalence of DVT in patients ≥20 years of age averaged over the 21-year period of study was 0.93% [4] (Figure 2.1). Venous thromboembolic disease (VTE), defined as PE and/or DVT, occurred in 1.24% [4] (Figure 2.1). The incidence of DVT in


BLUK077-Stein

March 12, 2007

CHAPTER 2

14:48

17

PE and DVT in hospitalized patients

0.5 0.40

PE

VTE

2

VTE

1.5 DVT 1 0.5

PE

0 99

97

95

93

91

89

87

85

83

81

VTE, DVT, PE in hospitalized adults (%)

DVT

1.8 1.6

Male

1.4

DVT

1.2 Female

1 0.8

Male

0.6

PE

0.4

Female

0.2 99

97

95

93

91

89

87

85

83

81

PE and DVT in hospitalized adults according to sex (%)

Year

Year

2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2

White patients Black patients White patients Black patients

99

97

95

93

Year

91

89

87

85

83

81

79

Figure 2.4 Incidences of pulmonary embolism (PE) and deep venous thrombosis (DVT) in hospitalized adults from 1979 to 1999 according to race. The incidences in black and white patients were the same for PE and nearly the same for DVT. (Reprinted from Stein et al. [4], with permission from Elsevier.)

0.93

79

Figure 2.3 Incidences of pulmonary embolism (PE) and deep venous thrombosis (DVT) in hospitalized adults from 1979 to 1999 according to sex. The incidences in men and women were nearly the same. (Reprinted from Stein et al. [4], with permission from Elsevier.)

1.24

1

79

Figure 2.2 Incidences of pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) in hospitalized adults from 1979 to 1999. The incidence of DVT increased (slope = 0.028%/year, r = 0.92, P < 0.0005). The incidence of PE did not change. The incidence of VTE increased in parallel to the incidence of DVT. (Reprinted from Stein et al. [4], with permission from Elsevier.)

1.5

0

PE and DVT in hospitalized adults according to race (%)

Figure 2.1 Prevalence of pulmonary embolism (PE), deep venous thrombosis (DVT) and either PE or DVT, venous thromboembolism (VTE) in hospitalized adults (â&#x2030;Ľ20 years). (Data from Stein et al. [4].)

PE and DVT in hospitalized adults (%)

2


BLUK077-Stein

March 12, 2007

14:48

18

PART I

hospitalized patients increased from 239,000 of 30,771,000 (0.8%) in 1979 to 363,000 of 28,504,000 (1.3%) in 1999 and the incidence of VTE increased in parallel (Figure 2.2). The incidence of DVT in hospitalized patients was nearly the same in men and women (relative risk of men to women 1.05) [4] (Figure 2.3). The incidence of DVT in hospitalized patients was nearly the same for white and black patients (relative risk 1.05) [4] (Figure 2.4). The increasing incidence of DVT in hospitalized patients from 1979 to 1999 may represent an increasing availability and use of venous ultrasound during much of that period [1]. Early diagnosis and treatment of DVT may have prevented a parallel increase in the incidence of PE in hospitalized patients. Whether the trend toward an increasing incidence of DVT in hospitalized patients will continue is uncertain, particularly in view of outpatient treatment of DVT, which was introduced in 1996 and 1997 [11–13]. The reported incidence of DVT in hospitals, 0.1– 3.17%, ranged more widely than the incidence of PE [14–17]. In Asian hospitals the prevalence of DVT was lower [18, 19], but VTE has been reported to be uncommon among Asians [20, 21] (see Chapter 15).

7 Proctor MC, Greenfield LJ. Pulmonary embolism: diagnosis, incidence and implications. Cardiovasc Surg 1997; 5: 77–81. 8 Stein PD, Patel KC, Kalra NJ et al. Estimated incidence of acute pulmonary embolism in a community/teaching general hospital. Chest 2002; 121: 802–805. 9 Hermann RE, Davis JH, Holden WD. Pulmonary embolism. A clinical and pathologic study with emphasis on the effect of prophylactic therapy with anticoagulants. Am J Surg 1961; 102: 19–28. 10 Huber O, Bounameaux H, Borst F et al. Postoperative pulmonary embolism after hospital discharge: an underestimated risk. Arch Surg 1992; 127: 310–313. 11 The Columbus Investigators. Low-molecular-weight heparin in the treatment of patients with venous thromboembolism. N Engl J Med 1997; 337: 657–662. 12 Levine M, Gent M, Hirsh J et al. A comparison of low-molecular-weight heparin administered primarily at home with unfractionated heparin administered in the hospital for proximal deep-vein thrombosis. N Engl J Med 1996; 334: 677–681. 13 Koopman MMW, Prandoni P, Piovella F et al. Treatment of venous thrombosis with intravenous unfractionated heparin administered in the hospital as compared with subcutaneous low-molecular-weight heparin administered at home. N Engl J Med 1996; 334: 682–687. 14 Klatsky AL, Armstrong MA, Poggi J. Risk of pulmonary embolism and/or deep venous thrombosis in AsianAmericans. Am J Cardiol 2000; 85: 1334–1337. 15 Igbinovia A, Malik GM, Grillo IA et al. Deep venous thrombosis in Assir region of Saudi Arabia. Case–control study. Angiology 195; 46: 1107–1113. 16 Schuurman B, den Heijer M, Nijs AM. Thrombosis prophylaxis in hospitalized medical patients: does prophylaxis in all patients make sense? Neth J Med 2000; 56: 171–176. 17 Stein PD, Patel KC, Kalra NK et al. Deep venous thrombosis in a general hospital. Chest 2002; 122: 960–962. 18 Liam CK, Ng SC. A Review of patients with deep vein thrombosis diagnosed at university hospital, Kuala Lumpur. Ann Acad Med Singapore 1990; 19: 837–840. 19 Kueh YK, Wang TL, Teo CP et al. Acute deep vein thrombosis in hospital practice. Ann Acad Med Singapore 1992; 21: 345–348. 20 Stein PD, Kayali F, Olson RE, Milford CE. Pulmonary thromboembolism in Asian/Pacific Islanders in the United States: analysis of data from the National Hospital Discharge Survey and the United States Bureau of the Census. Am J Med 2004; 116: 435–442. 21 White RH, Zhou H, Romano PS. Incidence of idiopathic deep venous thrombosis and secondary thromboembolism among ethnic groups in California. Ann Intern Med 1998; 128: 737–740.

References 1 Stein PD, Hull RD, Ghali WA et al. Tracking the uptake of evidence: two decades of hospital practice trends for diagnosing deep venous thrombosis and pulmonary embolism. Arch Intern Med 2003; 163: 1213–1219. 2 Unpublished data from Stein PD, Kayali F, Olson RE. Regional differences in rates of diagnosis and mortality of pulmonary thomboembolism. Am J Cardiol 2004; 93: 1194–1197. 3 Stein PD, Kayali F, Olson RE. Regional differences in rates of diagnosis and mortality of pulmonary thomboembolism. Am J Cardiol 2004; 93: 1194–1197. 4 Stein PD, Beemath A, Olson RE. Trends in the incidence of pulmonary embolism and deep venous thrombosis in hospitalized patients. Am J Cardiol 2005; 95: 1525– 1526. 5 Stein PD, Henry JW. Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest 1995; 108: 978–981. 6 Stein PD, Huang H, Afzal A et al. Incidence of acute pulmonary embolism in a general hospital: relation to age, sex and race. Chest 1999; 116: 909–913.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

March 28, 2007

3

15:2

CHAPTER 3

Case fatality rate and population mortality rate from pulmonary embolism and deep venous thrombosis

Overview

Untreated patients

Among all patients with pulmonary embolism (PE) throughout the United States, irrespective of treatment or the severity of the PE, the estimated case fatality rate (death from PE/100 patients with PE) in 1998 was 7.7% [1]. The case fatality rate from untreated clinically apparent PE, obtained before anticoagulant therapy was universally used, was 26â&#x20AC;&#x201C;37% [2, 3]. In recent years, however, the case fatality rate from mild untreated PE, based on a small number of patients was 5% [4]. The case fatality rate from acute PE in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) was 2.5% [5]. This rate applies to patients who were well enough to be recruited into an investigation of the diagnostic validity of ventilationâ&#x20AC;&#x201C; perfusion lung scans [5]. Most fatalities from PE occur within the first 2.5 hours after the diagnosis is made [6]. Such patients could not have been included in PIOPED. Case fatality rates from PE in trials of treatment with low-molecular-weight heparin were 0.6â&#x20AC;&#x201C; 1.0% [7, 8]. In such trials patients with massive PE requiring thrombolytic therapy, and patients at risk of bleeding, among others, were excluded. The most important factor affecting mortality is shock due to right ventricular failure secondary to PE [9]. In patients with over 50% occlusion of the pulmonary circulation who were in shock, the case fatality rate was 32%, whereas in those with over 50% occlusion of the pulmonary circulation who were not in shock, the case fatality rate was 6% [9].

Regarding untreated patients with acute PE in decades before diagnostic imaging tests were available, Barritt and Jordan reported a 26% mortality from the initial PE [3]. Some of these patients perhaps died from recurrent PE [3]. The diagnosis was made on the basis of clinical features that included evidence of right ventricular failure, pulmonary infarction, or both. Clinical features of pulmonary infarction included pleuritic pain, hemoptysis, pleural friction rub, loss of resonance at the lung base, rales, and the chest radiograph. Features that they relied upon for the detection of acute right ventricular failure were faintness, chest pain, fall of blood pressure, rise of jugular venous pressure, and the electrocardiogram. In 1961, Hermann and associates calculated a 37% case fatality rate from the initial PE [2] (Figure 3.1). The diagnosis was based on clinical features, and autopsy among those who died. The treatment of these patients was not reported, although data were collected between 1943 and 1957 and anticoagulant therapy was not in general used before 1947 [2]. Hermann and associates also reported a 36% frequency of fatal recurrent PE. The total estimated frequency of death that included the original PE and recurrent PE was 73% (Figure 3.1). There was, in addition, a 21% frequency of nonfatal recurrent PE among untreated patients with clinically diagnosed PE (Figure 3.2). Presumably, PE was severe among these patients with apparent clinical features.

19


BLUK077-Stein

March 28, 2007

15:2

20

Untreated fatal PE Initial and recurrent 73%

80 60

Initial 37%

Treated patients in the modern era Case fatality rates (PE deaths/100 PE) have been reported in regional surveys [10], regional estimates [11], registries [12], and prospective investigations [5]

Non-fatal recurrent PE (%)

Untreated PE 40 30

21% 20

5%

10 0

Clinically apparent initial PE

Mild initial PE

Figure 3.2 Nonfatal recurrent pulmonary embolism (PE) among untreated patients diagnosed on the basis of clinical findings (clinically apparent PE) and among patients with mild PE diagnosed by objective tests. (Data from Hermann et al. [2] and Stein and Henry [4].)

1997

We evaluated the 20 patients who received no treatment for PE during the first 3 months of follow-up of PIOPED [4]. Only 1 of these patients (5%) died of PE [4] (Figure 3.1). These 20 patients from PIOPED are described in Chapter 5.

1995

Figure 3.1 Fatal initial pulmonary embolism (PE) and fatal initial and recurrent PE among untreated patients diagnosed on the basis of clinical findings (clinically apparent PE) and among patients with mild PE diagnosed by objective tests. (Data from Hermann et al. [2] and Stein and Henry [4].)

1993

Mild PE

1991

Clinically apparent PE

1989

Clinically apparent PE

0 1987

0

4

1985

Initial & recurrent 5%

20

8

1983

40

12

1981

Mortality (%)

100

Prevalence, risks, and prognosis of PE and DVT

1979

Fatal PE/all PE (%)

PART I

Year Figure 3.3 Estimated case fatality rate for pulmonary embolism (PE) from 1979 to 1998. (Reprinted from Stein et al. [1], with permission from Elsevier.)

where the number of deaths was in the hundreds, and in elderly patients [13] where the number of deaths was several thousand. We calculated case fatality rates of PE from a database with 194,000 PE deaths, based on the entire population of the United States from 1979 to 1998 [1]. The estimated case fatality rate from PE increased from 6.7% in 1979 to 10.5% in 1989 (Figure 3.3). From 1989 to 1998 the estimated case fatality rate decreased to 7.7 PE deaths/100 PE. The estimated case fatality rate in the Minneapolis-St. Paul metropolitan area in 1995 ranged from approximately 2 to 6% depending on age [11]. As in our investigation, these are estimated rates [11]. The case fatality rate in short-stay hospitals in metropolitan Worcester in 1985â&#x20AC;&#x201C;1986 (12%) was somewhat higher than what we calculated during those years [10]. The estimated case fatality rate from PE increased with age (Figure 3.4). The relation of case fatality to age was described by an exponential function. The higher case fatality rate with age is concordant with regional investigations [10, 11] and studies in individual hospitals [2, 14]. The estimated case fatality rate from PE over the 20year interval of observation was higher among African Americans than Caucasians (Figure 3.5). The rate ratio of African Americans to Caucasians was 1.43. Others observed a higher case fatality rate among elderly African Americans than elderly Caucasians [13]. The estimated case fatality rate was comparable in men and women (Figure 3.6) as had been shown in regional studies [10, 11]. Among the patients â&#x2030;Ľ65 years of age, we showed no difference in the case fatality rate between men and women. In an investigation of Medicare patients aged 65â&#x20AC;&#x201C;74 years, the case fatality


BLUK077-Stein

March 28, 2007

CHAPTER 3

15:2

21

Case fatality rate and population mortality rate

20

Figure 3.4 Estimated case fatality rates for pulmonary embolism (PE) according to decades of age. The estimated case fatality rates are the average of yearly values over a 20-year period. (Reprinted from Stein et al. [1], with permission from Elsevier.)

Fatal PE/all PE (%)

17.4 15 10.9 10

8.2 3.6

5

6.9

5.5

0 25−34 35−44

45−54 55−64 65−74 75−84

>85

Age group (years)

rate in 1991 of 10.0% in white men and 9.4% in white women [13] was comparable to our estimated case fatality rates of 7.3 and 9.2% in white men and women aged 65 years or older the same year. The case fatality rate in black females aged 65 years or older in 1991 was virtually identical in both studies (11.4 and 11.1%), but we calculated a higher rate in black males (20.9% versus 13.5%).

Massive PE: hypotensive patients Among patients with massive PE, defined as a systolic arterial pressure <90 mm Hg, death from PE within 90 days occurred in 35 of 108 (32%) [15] (Figure 3.7). Among patients with systolic arterial pressure ≥90 mm Hg, death from PE within 90 days occurred in 119 of 2284 (5%).

Treated patients with PE compared with treated patients with deep venous thrombosis (DVT) Pooled data among treated patients with PE and treated patients with DVT showed higher death rates from recurrent PE among the PE patients [16]. Among 2429 patients with DVT treated 5 days to 3 months with anticoagulants, death from PE occurred in 0.3% (Figure 3.8). Among 949 patients with PE treated 5 days to 3 months, death from recurrent PE occurred in 1.4%. The death rates from PE excluded deaths within the first 5 days of diagnosis [16]. Patients treated with thrombolytic agents or inferior vena cava filters were also excluded.

Population mortality rate from PE The number of patients who died from PE in 1998 based on death certificates was 24,947 [17]. With a

Caucasians 1997

1995

1993

1991

1989

1987

1985

1983

1981

Male

8

Female

4 0

1997

1995

1993

1991

1989

1987

1985

1983

1981

Figure 3.5 Case fatality rate from pulmonary embolism (PE) among African Americans and Caucasians. The estimated case fatality rate from PE was higher among African Americans than Caucasians over the 20-year period of observation (P = 0.014). (Reprinted from Stein et al. [1], with permission from Elsevier.)

12

1979

Year

Fatal PE/all PE (%)

African American

20 16 12 8 4 0 1979

Fatal PE/all PE (%)

4.7

Year Figure 3.6 Case fatality rate according to sex. The estimated case fatality rate among males and females was comparable. (Reprinted from Stein et al. [1], with permission from Elsevier.)


BLUK077-Stein

March 28, 2007

15:2

Mortality from PE (%)

22

PART I

40 35 30 25 20 15 10 5 0

32%

5%

PE BP < 90 mm Hg

PE BP > 90 mm Hg

Figure 3.7 Death rates from pulmonary embolism (PE) within 90 days among patients with massive PE, defined as a systolic arterial pressure (BP) <90 mm Hg, and patients with nonmassive PE (systolic arterial pressure ≥90 mm Hg). (Data are from Kucher et al. [15].)

population of 270,299,000 in 1998, this would amount to 9 PE deaths/100,000 population. Assuming that death certificates are only 26.7% accurate for the diagnosis of fatal PE [18], the estimated number of deaths from PE in 1998 may be 93,000 and the death rate may be 34 PE deaths/100,000 population. Over the last two decades the population mortality rate from PE (deaths from PE/100,000 population) decreased [17, 19]. This could be a consequence of a declining incidence of PE (diagnoses of PE/100,000 population) or a declining case fatality rate from PE (deaths from PE/100 cases of PE) or the combination. For the interval 1979–1989, our data showed no decline in the case fatality rate (Figure 3.3), which suggests that

1.4%

Fatal PE (%)

1.5

1

0.5

n = 949 0.3%

n = 2429 0 DVT

PE

Figure 3.8 Pooled data among patients with pulmonary embolism (PE) and patients with deep venous thrombosis (DVT) treated 5 days to 3 months with anticoagulants. The death rate from recurrent PE was higher in patients with PE than the death rate from PE in patients with DVT. (Data are from Douketis et al. [16].)

Prevalence, risks, and prognosis of PE and DVT

the declining population mortality rate from PE was largely due to a declining incidence of PE. From 1979 to 1999, the incidence of diagnosis has, in fact, decreased [20]. Improved prevention of PE may have been a key factor. From 1989 to 1998, the declining population mortality rate appears to be largely due to a declining case fatality rate. The rate of diagnosis of PE no longer declined during this interval [20], but the population mortality rate continued to decline [17]. The estimated case fatality rate also declined during this period (Figure 3.3). The declining population mortality rate during this period, therefore, reflected a decreased case fatality rate. Earlier diagnosis and improved management rather than improved prevention would account for the decreased population mortality rate during this period. The trends in estimated case fatality that we showed were concordant with trends from the Minneapolis-St. Paul metropolitan area [21]. Both databases showed either an increasing or unchanging case fatality rate from 1979 to the mid- or late 1980s [21] followed by a declining case fatality rate.

References 1 Stein PD, Kayali F, Olson RE. Estimated case fatality rate of pulmonary embolism, 1979–1998. Am J Cardiol 2004; 93: 1197–1199. 2 Hermann RE, Davis JH, Holden WD. Pulmonary embolism: a clinical and pathologic study with emphasis on the effect of prophylactic therapy with anticoagulants. Am J Surg 1961; 102: 19–28. 3 Barritt DW, Jordan SC. Anticoagulant drugs in the treatment of pulmonary embolism: a controlled trial. Lancet 1960; 1: 1309–1312. 4 Stein PD, Henry JW, Relyea B. Untreated patients with pulmonary embolism: outcome, clinical and laboratory assessment. Chest 1995; 107: 931–935. 5 Carson JL, Kelley MA, Duff A et al. The clinical course of pulmonary embolism. New Engl J Med 1992; 326: 1240– 1245. 6 Stein PD, Henry JW. Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest 1995; 108: 978–981. 7 The Columbus Investigators. Low-molecular-weight heparin in the treatment of patients with venous thromboembolism. N Engl J Med 1997; 337: 657–662. 8 Simonneau G, Sors H, Charbonnier B et al. A comparison of low-molecular-weight heparin with unfractionated heparin for acute pulmonary embolism. The THESSE Study Group. N Engl J Med 1997; 337: 663–669.


BLUK077-Stein

March 28, 2007

CHAPTER 3

15:2

Case fatality rate and population mortality rate

9 Alpert JS, Smith R, Carlson J et al. Mortality in patients treated for pulmonary embolism. JAMA 1976; 236: 1477– 1480. 10 Anderson FA, Wheeler HB, Goldberg RJ et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med 1991; 151: 933–938. 11 Janke RM, MCGovern PG, Folsom AR. Mortality, hospital discharges, and case fatality for pulmonary embolism in the twin cities: 1980–1995. J Clin Epidemiol 2000; 53: 103–109. 12 Goldhaber SZ, Visani L, De Rosa M, for ICOPER. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet 1999; 353: 1386–1389. 13 Siddique RM, Siddique MI, Conners AF, Jr, Rimm AA. Thirty-day case-fatality rates for pulmonary embolism in the elderly. Arch Intern Med 1996; 156: 2343–2347. 14 Byrne JJ. Phlebitis: a study of 748 cases at the Boston City Hospital. New Engl J Med 1955; 253: 579–586. 15 Kucher N, Rossi E, De Rosa M, Goldhaber SZ. Massive pulmonary embolism. Circulation 2006; 113: 577–582.

23

16 Douketis JD, Kearon C, Bates S, Duku EK, Ginsberg JS. Risk of fatal pulmonary embolism in patients with treated venous thromboembolism. JAMA 1998; 279: 458– 462. 17 Horlander KT, Mannino DM, Leeper KV. Pulmonary embolism mortality in the United States, 1979–1998: an analysis using multiple-cause mortality data. Arch Intern Med 2003; 163: 1711–1717. 18 Attems J, Arbes S, Bohm G, Bohmer F, Lintner F. The clinical diagnostic accuracy rate regarding the immediate cause of death in a hospitalized geriatric population; an autopsy study of 1594 patients. Wien Med Wochenschr 2004; 154: 159–162. 19 Lilienfeld DE. Decreasing mortality from pulmonary embolism in the United States, 1979–1996. Int J Epidemiol 2000; 29: 465–469. 20 Stein PD, Hull RD, Patel KC et al. Venous thromboembolic disease: comparison of the diagnostic process in men and women. Arch Intern Med 2003; 163: 1689–1694. 21 Lilienfeld DE, Godbold JH, Burke GL, Sprafka JM, Pham DL, Baxter J. Hospitalization and case fatality for pulmonary embolism in the twin cities: 1979–1984. Am Heart J 1990; 120: 392–395.


BLUK077-Stein

4

April 6, 2007

15:6

CHAPTER 4

Prognosis in acute pulmonary embolism based on right ventricular enlargement, prognostic models, and biochemical markers

Right ventricular enlargement and prognostic models The prognosis in acute pulmonary embolism (PE) depends on the severity of the PE and the prior cardiorespiratory condition of the patient. The size of the PE has been assessed by the pulmonary angiogram, CT angiogram (Chapter 73), and perfusion lung scans [1]. Indices of severity include shock, pulmonary hypertension, right ventricular dilatation, and Pa O2 . Elderly patients had a worse prognosis than young patients (Chapter 3). Literature review of the prognostic value of echocardiographically assessed right ventricular dysfunction showed that short-term PE-related deaths occurred in 4–14% more patients with right ventricular dysfunction than in those without right ventricular dysfunction [2]. Among normotensive patients, the short-term mortality was 4 and 5% higher among those with right ventricular dysfunction [2]. Others, in a review, found a short-term mortality of 5% in normotensive patients with ventricular dysfunction [3]. Among patients with PE in whom the right ventricular to left ventricular short-axis ratio (RV/LV) was measured by CT, the positive predictive value of death from PE in 3 months was 10% in 69 patients with RV/LV > 1 and it was zero among 51 patients with RV/LV ≤ 1 [4]. However, among patients with PE in PIOPED II who were treated only with anticoagulants and/or an inferior vena cava filter, who were not hypotensive, in shock, critically ill, on ventilatory support, did not have a myocardial infarction within the previous month, and did not have an episode of ven-

24

tricular tachycardia or ventricular fibrillation within the previous 24 hours, in-hospital outcome was the same in those with and those without an enlarged right ventricle (Stein PD, Beemath A, Matta F, et al., unpublished data from PIOPED II). The in-hospital mortality from PE on in these patients with an RV/LV short axis dimension ratio >1 measured on CT angiograms was 0 of 76 (0%) versus 1 of 79 (1.3%) in those with an RV/LV dimension ratio ≤1. The in-hospital all cause mortality in those with an RV/LV dimension ratio >1 was 2 of 76 (2.6%) versus 2 of 79 (2.5%) in those with an RV/LV dimension ratio ≤1. The case fatality rate in hypotensive patients and in patients with right ventricular enlargement or dysfunction is described in Chapter 89. Correlation of pulmonary artery mean pressure with angiographic severity was low (r = 0.38) as was the correlation of Pa O2 (r = −0.34) [5]. Prognostic models based on weighting of several indices of severity have been described [6–8]. The most recent consists of 11 routinely available predictors of 30-day all-cause mortality [6]. These were age, male sex, cancer heart failure, chronic lung disease, pulse ≥110/min, respiratory rate ≥30/min, temperature <36◦ C, altered mental status (disorientation, lethargy, stupor, or coma), and oxygen saturation <90 mm Hg [6]. These were assigned point scores as shown in Table 4.1. Validation in a subsequent investigation showed a 30-day all-cause mortality of 0, 1.0, 3.1, 10.4, and 24.4% in the respective five classes of risk, which did not differ from the original derivation values [9]. Several laboratory tests were also independently associated with 30-day all-cause mortality [9].


BLUK077-Stein

April 6, 2007

CHAPTER 4

15:6

Prognosis in acute pulmonary embolism

Table 4.1 Predictors of 30-day mortality. Predictor

Point Score

Age

1 point/yr

Male

10

Cancer

30

Heart failure

10

Chronic lung disease

10

Pulse >110/min

20

Systolic blood pressure <100 mm HG

30

Respiratory rate ≥30/min

20

Temperature <36◦ C

20

Disorientation, lethargy, stupor, or coma

60

Oxygen saturation <90 mm Hg

20

Risk

Point Score

very low

≤65

low intermed

66–85 86–105

high

106–125

very high

>125

Table base on Aujesky et al. [9].

These included hemoglobin <12 g/dL, white blood cell count <4000 or >12,000/mm3 , platelets <100,000/mm3 , sodium <130 or >150 meq/L, blood urea nitrogen ≥30 mg/dL, arterial pH 7.25, and Pa CO2 <25 or >55 mm Hg [9]. The more complex model that included laboratory findings showed 30-day mortalities that were similar to mortalities with the less complex model [9].

Cardiac troponins Cardiac troponins are regulatory proteins of the thin actin filaments of the cardiac muscle [10]. They control the calcium-mediated interaction of actin and myosin [11]. Troponin T and troponin I are highly sensitive and specific markers of myocardial injury [10]. The release of cardiac troponin from the myocyte to the blood can be due to reversible or irreversible cell damage [11]. Ischemia without coronary stenosis (demand ischemia), resulting from a mismatch between myocardial oxygen supply and demand, may occur [11]. This may cause increased membrane permeability and release of smaller troponin fragments into the systemic circulation [11]. It has been thought for many years that some of the electrocardiographic changes in acute PE reflect myocardial ischemia [12–14]. Myocardial

25

infarction has been shown at autopsy of patients who died of PE and had normal coronary arteries [12, 13, 15, 16]. In fact, investigations in dogs [17] and in pigs [18] with experimentally induced PE showed that blood flow increased in both the right and left coronary arteries. Coronary blood flow increased concordantly with increasing pulmonary artery pressure and decreasing PaO2 [17, 18]. The troponin complex consists of three subunits: troponin T, which binds to tropomyosin and facilitates contraction; troponin I, which binds to actin and inhibits actin–myosin interactions; and troponin C, which binds to calcium ions [11]. The amino acid sequences of the skeletal and cardiac isoforms of cardiac troponin T and troponin I are sufficiently dissimilar and therefore detectable by monoclonal antibodybased assays [11]. Troponin C is not used clinically because both cardiac and smooth muscle share troponin C isoforms. In-hospital, all-cause mortality among patients with acute PE who had an elevated troponin I ranged from 14 to 36% and in patients with a normal troponin I, in-hospital mortality ranged from 2 to 7% [19–26] (Table 4.1). Pooled data among patients with an elevated troponin I showed an in-hospital mortality 24 of 109 (22%) versus 17 of 291 (6%) among patients with a normal troponin I. Patients with a markedly elevated troponin I level above 1.5 ng/mL had a higher mortality from PE (22% mortality) than those with a modest elevation of 0.07–1.5 ng/mL (10% mortality) [20]. Among unselected PE patients with an elevated troponin T, in-hospital, all-cause mortality was 13 of 57 (23%) versus 3 of 105 (3%) [20, 24] among patients with a normal troponin T (Table 4.2).

Troponin levels in combination with right ventricular dysfunction or right ventricular dilatation Acute right ventricular dilatation or dysfunction has been shown to indicate a poor prognosis by several investigators. This is reviewed in Chapter 57. Patients with right ventricular dysfunction in combination with an elevated troponin I [23] or troponin T [27] showed high rates of in-hospital mortality compared with those in whom the right ventricle was normal and troponin level was not elevated (Table 4.3). In-hospital mortality, however, was the same in patients with


BLUK077-Stein

April 6, 2007

15:6

26

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 4.2 Mortality and troponin levels in patients with acute pulmonary embolism. Troponin I

Cutoff

Abnormal, mortality

Normal, mortality

value

Cause of

Follow-up

First author [Ref]

n/ N (%)

n/ N (%)

(ng/mL)

death

duration

Yalamanchili [19]

8/24 (33)

All cause

In-hospital

Unselected

7/43 (16)*

9/123 (7) 1/63 (2)*

≥2.0

Konstantinides [20]

≥0.07

All cause

In-hospital

Unselected

Kucher [21]

4/28 (14)

1/63 (2)

≥0.06

All cause

In-hospital

Unselected

La Vecchia [22]

5/14 (36)

1/42 (2)

>0.6

All cause

In-hospital

Unselected

Scridon [23]

23/73 (32)*

5/68 (7)*

>0.1

All cause

30 days

Unselected

Troponin T

Selection

Cutoff

Abnormal, mortality

Normal, mortality

value

Cause of

Follow-up

First author [Ref]

n/ N (%)

n/ N (%)

(ng/mL)

death

duration

Konstantinides [20]

5/39 (13)*

2/67 (3)*

≥0.04

All cause

In-hospital

Unselected

Giannitsis [24]

8/18 (44)

1/38 (3)

≥0.1

All cause

In-hospital

Unselected

Selection

Pruszczyk [25]

8/32 (25)

0/32 (0)

>0.01

All cause

In-hospital

Normotensive

Pruszczyk [26]

6/24 (25)

1/22 (5)

>0.01

All cause

In-hospital

RV Dilatation

* Estimated from authors’ graphs.

elevations of troponin, irrespective of right ventricular function.

Myoglobin Myoglobin is a low-molecular-heme protein that is found in both cardiac and skeletal muscle and, therefore, is not cardiac specific [28, 29]. It is among the earliest markers released into the circulation [30], and may be detected as early as 2 hours after the onset of myocardial necrosis [28]. Serum myoglobin after myocardial infarction increases even before a detectable rise of cardiac troponin levels occurs [29]. Myoglobin measurement for the diagnosis of the acute coronary syndrome is most efficient when measured within 6 hours after the onset of myocardial infarc-

tion [28]. Literature on myoglobin levels in pulmonary embolism is sparse. Among patients with PE who had right ventricular distention and an elevated myoglobin level, in-hospital, all-cause mortality was 7 of 21 (33%) versus 0 of 25 (0%) among such patients who had a normal myoglobin level [16] (Table 4.4).

Natriuretic peptides The natriuretic peptides are useful diagnostic and prognostic biomarkers for patients with congestive heart failure [31]. In contrast to atrial natriuretic peptide that originates mainly from atrial tissue, brain natriuiretic peptide (BNP) is produced to a large degree from ventricular myocytes [31]. The principal stimulus for BNP synthesis and secretion is cardiomyocyte

Table 4.3 Mortality and right ventricular dysfunction with elevated troponin levels in patients with acute pulmonary embolism. RV dysfunction

RV normal and

and elevated

troponin not

troponins

elevated

[mortality

[mortality

First author [Ref]

n/N (%)]

n/N (%)]

Scridon [23]

17/45 (38)

Binder [27] ∗

3/16 (19)*

Estimated from authors’ graphs.

Troponin type

Cutoff value

Cause of

Follow-up

(ng/mL)

death

duration

— (5)

Trop I

>0.1

All cause

30 days

0/53 (0)

Trop T

≥0.04

All cause

In-hospital


BLUK077-Stein

April 6, 2007

CHAPTER 4

15:6

27

Prognosis in acute pulmonary embolism

Table 4.4 Mortality and myoglobin levels in patients with acute pulmonary embolism. Myoglobin High, mortality

Normal, mortality

Cutoff value

Cause of

Follow-up

First author [Ref]

n/N (%)

n/N (%)

(ng/mL)

death

duration

Selection

Pruszczyk [20]

7/21 (33)

0/25 (0)

>0.01

All cause

In-hospital

RV dilatation

stretch [31]. Brain natriuretic peptide is a 32-amino acid peptide hormone first isolated from porcine brain tissue [31]. In plasma, the intact 108 amino acid prohormone (proBNP), the biologically active 32-amino acid BNP, and the remaining part of the prohormone, N-terminal (NT)-proBNP, which is 76 amino acids, can be measured by immunoassay [31]. Prohormones in normal ventricular myocytes are not stored to a significant amount. Therefore, it takes several hours for the plasma natruiretic levels to increase significantly after the onset of stretch [32]. This process includes BNP messenger ribonucleic acid synthesis, prohormone synthesis, and prohormone release into the circulation. Elevations in BNP [33] and NT-proBNP [34] are associated with right ventricular dysfunction in acute PE. Natriuretic peptide levels are also increased in patients with right ventricular pressure overload due to causes other than PE, including primary pulmonary hypertension, chronic thromboembolic hypertension, congenital heart disease, and chronic lung disease [35–38].

Brain natriuretic peptides in patients with PE, when low, predict a benign clinical outcome, with few inhospital deaths from PE [31]. There were no inhospital deaths among 70 patients with PE who had a low NT-proBNP level [28, 39] and 2 in-hospital deaths among 66 patients with PE who had a low BNP level [40, 41] (Table 4.5). Published cutoff values for NT-proBNP [27, 34, 39] and for BNP [40–42] vary (Table 4.4). Because BNP release into the circulation may take several hours after the onset of myocardial injury, a second measurement should be obtained 6– 12 hours after an initially negative test in a PE patient with a symptom duration <6 hours [31].

Serum uric acid Uric acid is the final product of purine nucleotide degradation. Tissue hypoxia depletes adenosine triphosphate (ATP) and activates the purine nucleotide degradation pathway to uric acid, resulting in urate overproduction [43, 44]. Serum uric acid level is therefore determined by the imbalance between

Table 4.5 Mortality, brain natriuretic peptide, and N-terminal pro-brain natriuretic peptide levels in patients with acute pulmonary embolism. Natriuretic peptide

Peptide and

First author

High, mortality

Normal, mortality

cutoff value

Cause of

Follow-up

[Ref]

n/N (%)

n/N (%)

(pg/mL)

death

duration

Selection

Binder [27]

7/67 (10)*

0/57 (0)

NT-proBNP

All cause

In-hospital

Unselected

All cause

In-hospital

Unselected

Pulmonary

In-hospital

Unselected

In-hospital

Unselected

3 months

Unselected

≥1000 Pruszczyk [39]

15/66 (23)

0/13 (0)

NT-proBNP 88–334†

Kruger [40]

1/17 (6)

1/25 (4)

BNP >90

embolism Kucher [41]

4/32 (13)

1/41 (2)

BNP ≥90

Pulmonary embolism

ten Wolde [42]

4/36 (11)

1/74 (1)

BNP >21.7 pmol/L

Pulmonary embolism

Estimated from authors’ graphs. NT-proBNP cutoff level age and sex dependent. NT-proBNP, N-terminal pro-brain natriuretic peptide; BNP, brain natriuretic peptide. †


BLUK077-Stein

April 6, 2007

15:6

28

PART I

production and excretion and is increased in hypoxic states such as left ventricular failure, cyanotic heart disease, and obstructive pulmonary disease [45–50]. Among 71 patients with acute PE, serum uric acid was elevated (mean ± SD 6.2 ± 2.3 mg/dL) compared with 62 age- and sex-matched controls (4.5 ± 0.9 mg/dL) [51]. Serum uric acid was higher in the 27 patients who died from PE during hospitalization than in the remaining survivors (8.3 ± 2.2 versus 6.5 ± 2.2 mg/dL). After treatment including thrombolysis and pulmonary embolectomy, serum uric acid level significantly decreased in patients with PE from 6.7 ± 2.0 to 5.8 ± 1.9 mg/dL.

10 Ammann P, Pfisterer M, Fehr T, Rickli H. Raised cardiac troponins: causes extend beyond acute coronary syndromes. BMJ 2004; 328: 1028–1029. 11 Ammann P, Maggiorini M, Bertel O et al. Troponin as a risk factor for mortality in critically ill patients without acute coronary syndromes. J Am Coll Cardiol 2003; 41: 2004–2009. 12 Horn H, Dack S, Friedberg CK. Cardiac sequelae of embolism of the pulmonary artery. Arch Int Med 1939; 64: 296. 13 Dack S, Master AM, Horn H, Grishman A, Field LE. Acute coronary insufficiency due to pulmonary embolism. Am J Med 1949; 7: 464. 14 Weber DM, Phillips JH, Jr. A re-evaluation of electrocardiographic changes accompanying acute pulmonary embolism. Am J Med Sci 1966; 251: 381–398. 15 Currens J, Barnes AR. The heart in pulmonary embolism. Arch Int Med 1943; 71: 325. 16 Coma-Canella I, Gamallo C, Martinez Onsurbe P, LopezSendon J. Acute right ventricular infarction secondary to massive pulmonary embolism. Eur Heart J 1988; 9: 534– 540. 17 Stein PD, Alshabkhoun S, Hatem C et al. Coronary artery blood flow in acute pulmonary embolism. Am J Cardiol 1968; 21: 32–37. 18 Stein PD, Alshabkhoun S, Hawkins HF, Hyland JW, Jarrett CE. Right coronary blood flow in acute pulmonary embolism. Am Heart J 1969; 77: 356–362. 19 Yalamanchili K, Sukhija R, Aronow WS, Sinha N, Fleisher AG, Lehrman SG. Prevalence of increased cardiac troponin I levels in patients with and without acute pulmonary embolism and relation of increased cardiac troponin I levels with in-hospital mortality in patients with acute pulmonary embolism. Am J Cardiol 2004; 93: 263– 264. 20 Konstantinides S, Geibel A, Olschewski M et al. Importance of cardiac troponins I and T in risk stratification of patients with acute pulmonary embolism. Circulation 2002; 106: 1263–1268. 21 Kucher N, Wallmann D, Carone A, Windecker S, Meier B, Hess OM. Incremental prognostic value of troponin I and echocardiography in patients with acute pulmonary embolism. Eur Heart J 2003; 24: 1651– 1656. 22 La Vecchia L, Ottani F, Favero L et al. Increased cardiac troponin I on admission predicts in-hospital mortality in acute pulmonary embolism. Heart 2004; 90: 633– 637. 23 Scridon T, Scridon C, Skali H, Alvarez A, Goldhaber SZ, Solomon SD. Prognostic significance of troponin elevation and right ventricular enlargement in acute pulmonary embolism. Am J Cardiol 2005; 96: 303– 305.

References 1 The Urokinase Pulmonary Embolism Trial: A National Cooperative Study. Perfusion lung scanning. Circulation 1973; 47(2, suppl): II46–II50. 2 ten Wolde M, Sohne M, Quak E, Mac Gillavry MR, Buller HR. Prognostic value of echocardiographically assessed right ventricular dysfunction in patients with pulmonary embolism. Arch Intern Med 2004; 164: 1685–1689. 3 Gibson NS, Sohne M, Buller HR. Prognostic value of echocardiography and spiral computed tomography in patients with pulmonary embolism. Curr Opin Pulmon Med 2005; 11: 380–384. 4 van der Meer RW, Pattynama PM, van Strijen MJ et al. Right ventricular dysfunction and pulmonary obstruction index at helical CT: prediction of clinical outcome during 3-month follow-up in patients with acute pulmonary embolism. Radiology 2005; 235: 798–803. 5 The Urokinase Pulmonary Embolism Trial: A National Cooperative Study. Interrelationships of pulmonary angiograms, lung scans, hemodynamic measurements, and fibrinolytic findigs. Circulation 1973; 47(2, suppl): II73– II80. 6 Aujesky D, Obrosky DS, Stone RA et al. Derivation and validation of a prognostic model for pulmonary embolism. Am J Respir Crit Care Med 2005; 172: 1041–1046. 7 Wicki J, Perrier A, Perneger TV, Bounameaux H, Junod AF, Predicting adverse outcome in patients with acute pulmonary embolism: a risk score. Thromb Haemost 2000; 84: 548–552. 8 Nendaz MR, Bandelier P, Aujesky D et al. Validation of a risk score identifying patients with acute pulmonary embolism, who are at low risk of clinical adverse outcome. Thromb Haemost 2004; 91: 1232–1236. 9 Aujesky D, Roy PM, Le Manach CP et al. Validation of a model to predict adverse outcomes in patients with pulmonary embolism. Eur Heart J 2006; 27: 476–481.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

April 6, 2007

CHAPTER 4

15:6

Prognosis in acute pulmonary embolism

24 Giannitsis E, Muller-Bardorff M, Kurowski V et al. Independent prognostic value of cardiac troponin T in patients with confirmed pulmonary embolism. Circulation 2000; 102: 211–217. 25 Pruszczyk P, Bochowicz A, Torbicki A et al. Cardiac troponin T monitoring identifies high-risk group of normotensive patients with acute pulmonary embolism. Chest 2003; 123: 1947–1952. 26 Pruszczyk P, Bochowicz A, Kostrubiec M et al. Myoglobin stratifies short-term risk in acute major pulmonary embolism. Clin Chim Acta 2003; 338: 53– 56. 27 Binder L, Pieske B, Olschewski M et al. N-terminal pro-brain natriuretic peptide or troponin testing followed by echocardiography for risk stratification of acute pulmonary embolism. Circulation 2005; 112: 1573– 1579. 28 Braunwald E, Antman EM, Beasley JW, et al. ACC/AHA guidelines for the management of patients with unstable angina and non-ST-segment elevation myocardial infarction: a report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines (Committee on the Management of Patients With Unstable Angina). J Am Coll Cardiol, 2000: 36: 970– 1062. 29 de Lemos JA, Morrow DA, Gibson M et al. The prognostic value of serum myoglobin in patients with nonST-segment elevation acute coronary syndromes. Results from the TIMI 11B and TACTICS-TIMI 1B Studies. J Am Coll Cardiol 2002; 40: 238–244. 30 Mair J, Wagner I, Jakob G et al. Different time courses of cardiac contractile proteins after acute myocardial infarction. Clin Chim Acta 1994; 231: 47–60. 31 Kucher N, Goldhaber SZ. Cardiac biomarkers for risk stratification of patients with acute pulmonary embolism. Circulation 2003; 108: 2191–2194. 32 Hama N, Itoh H, Shirakami G et al. Rapid ventricular induction of brain natriuretic peptide gene expression in experimental acute myocardial infarction. Circulation 1995; 92: 1558–1564. 33 Tulevski II, Hirsch A, Sanson BJ et al. Increased brain natriuretic peptide as a marker for right ventricular dysfunction in acute pulmonary embolism. Thromb Haemost 2001; 86: 1193–1196. 34 Kucher N, Printzen G, Doernhoefer T, Windecker S, Meier B, Hess OM. Low pro-brain natriuretic peptide levels predict benign clinical outcome in acute pulmonary embolism. Circulation 2003; 107: 1576–1578. 35 Nagaya N, Nishikimi T, Okano Y et al. Plasma brain natriuretic peptide levels increase in proportion to the extent of right ventricular dysfunction in pulmonary hypertension. J Am Coll Cardiol 1998; 31: 202– 208.

29

36 Nagaya N, Nishikimi T, Uematsu M et al. Plasma brain natriuretic peptide as a prognostic indicator in patients with primary pulmonary hypertension. Circulation 2000; 102: 865–870. 37 Bando M, Ishii Y, Sugiyama Y, Kitamura S. Elevated plasma brain natriuretic peptide levels in chronic respiratory failure with cor pulmonale. Respir Med 1999; 93: 507–514. 38 Tulevski II, Groenink M, van Der Wall EE et al. Increased brain and atrial natriuretic peptides in patients with chronic right ventricular pressure overload: correlation between plasma neurohormones and right ventricular dysfunction. Heart 2001; 86: 27–30. 39 Pruszczyk P, Kostrubiec M, Bochowicz A et al. Nterminal pro-brain natriuretic peptide in patients with acute pulmonary embolism. Eur Respir J 2003; 22: 649– 653. 40 Kruger S, Graf J, Merx MW et al. Brain natriuretic peptide predicts right heart failure in patients with acute pulmonary embolism. Am Heart J 2004; 147: 60–65. 41 Kucher N, Printzen G, Goldhaber SZ. Prognostic role of brain natriuretic peptide in acute pulmonary embolism. Circulation 2003; 107: 2545–2547. 42 ten Wolde M, Tulevski II, Mulder JW et al. Brain natriuretic peptide as a predictor of adverse outcome in patients with pulmonary embolism. Circulation 2003; 107: 2082–4208. 43 Fox AC, Reed GE, Meilman H, Silk BB. Release of nucleosides from canine and human hearts as an index of prior ischemia. Am J Cardiol 1979; 43: 52–58. 44 Mentzer RM, Jr, Rubio R, Berne RM. Release of adenosine by hypoxic canine lung tissue and its possible role in pulmonary circulation. Am J Physiol 1975; 229: 1625– 1631. 45 Leyva F, Anker S, Swan JW, Godsland IF, Wingrove CS, Chua TP et al. Serum uric acid as an index of impaired oxidative metabolism in chronic heart failure. Eur Heart J 1997; 18: 858–865. 46 Anker SD, Leyva F, Poole-Wilson PA, Kox WJ, Stevenson JC, Coats AJ. Relation between serum uric acid and lower limb blood flow in patients with chronic heart failure. Heart 1997; 78: 39–43. 47 Hayabuchi Y, Matsuoka S, Akita H, Kuroda Y. Hyperuricaemia in cyanotic congenital heart disease. Eur J Pediatr 1993; 152: 873–876. 48 Hasday JD, Grum CM. Nocturnal increase of urinary uric acid:creatinine ratio: a biochemical correlate of sleepassociated hypoxemia. Am Rev Respir Dis 1987; 135: 534– 538. 49 Braghiroli A, Sacco C, Erbetta M, Ruga V, Donner CF. Overnight urinary uric acid:creatinine ratio for detection of sleep hypoxemia: validation study in chronic


BLUK077-Stein

April 6, 2007

15:6

30

PART I

obstructive pulmonary disease and obstructive sleep apnea before and after treatment with nasal continuous positive airway pressure. Am Rev Respir Dis 1993; 148: 173– 178. 50 Elsayed NM, Nakashima JM, Postlethwait EM. Measurement of uric acid as a marker of oxygen tension

in the lung. Arch Biochem Biophys 1993; 302: 228– 232. 51 Shimizu Y, Nagaya N, Satoh T et al. Serum uric acid level increases in proportion to the severity of pulmonary thromboembolism. Circulation 2002; 66: 571– 575.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

April 6, 2007

5

15:7

CHAPTER 5

Changing risks of untreated deep venous thrombosis and acute pulmonary embolism

Introduction The frequency of fatal pulmonary embolism (PE) in patients with untreated deep venous thrombosis (DVT) has diminished as diagnostic tests have made it possible to diagnose mild DVT [1]. Prior to the use of venography and of sensitive noninvasive tests for the early detection of DVT, the risk of fatal PE in untreated patients with clinically apparent DVT was 37% [2] (Figure 5.1). In pooled data of patients with un-

100

60 Clinical DVT (%)

Fatal PE (%)

80

treated DVT identified by radioactive fibrinogen scintiscans, most of which was distal and subclinical, fatal PE occurred in 5% [3]. It is apparent that the risk of fatal PE was greater among patients with more severe DVT. The percentage of patients with acute PE who have clinically detectable DVT has diminished as physicians have developed the ability to diagnose subtle PE [2]. Among patients who died from acute PE, 53% had clinically identified DVT [4] (Figure 5.2). In an investigation of patients with massive or submassive angiographically diagnosed acute PE (the Urokinase Pulmonary Embolism Trial), 34% of patients had clinically identifiable DVT [5]. Among patients with mild as well as severe acute PE (PIOPED), only 15% of patients had clinically apparent DVT [6] and 47% had signs of DVT in PIOPED II [7].

40

20

60 53 34 20 15 0

0 Clinical DVT

Subclinical DVT

Figure 5.1 Frequency of fatal pulmonary embolism (PE) in untreated patients with clinically apparent deep venous thrombosis (DVT), and patients, most of whom had subclinical DVT diagnosed by radioactive fibrinogen scintiscans. (Data are from Byrne [2] and Collins et al. [3]. Reprinted from Stein [1] with permission.)

47

40

Fatal PE

Massive Massive or Massive or or mild PE mild PE submassive (PIOPED) (PIOPED II) PE

Figure 5.2 Frequency of clinically apparent deep venous thrombosis (DVT) among patients with fatal acute pulmonary embolism (PE), massive or submassive pulmonary embolism, or massive or mild pulmonary embolism. (Data are based on Byrne and Oâ&#x20AC;&#x2122;Neil [4], the Urokinase Pulmonary Embolism Trial [5], Stein [6] and Stein et al. [7].)

31


April 6, 2007

15:7

32

In the present era of early diagnosis of acute PE, the risk of fatal recurrent PE as well as the risk of death from the initial PE in untreated patients is lower than in patients with severe PE reported in prior years [1]. Among untreated patients with acute PE diagnosed on the basis of clinical features that included evidence of right ventricular failure, pulmonary infarction, or both, Barritt and Jordan reported a 26% mortality from the initial PE, although some of these patients, perhaps died from recurrent PE [8] (see Chapter 3, untreated patients). In 1961, Hermann and associates [9] calculated a 37% mortality from the initial PE and a 36% frequency of fatal recurrent PE (see Chapter 3, untreated patients).

Untreated patients in present era of early diagnosis In an investigation of the clinical course of acute PE, Carson and associates observed that in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED), 24 patients escaped treatment in the hospital [10]. However, 4 patients were begun on anticoagulant therapy during the first month after discharge. We evaluated the 20 patients who received no treatment for PE during the first 3 months of follow-up [11]. Only 1 of these patients died of PE [10, 11]. In the era of early diagnosis by ventilation–perfusion lungs scans and pulmonary angiography, mortality from the initial PE and from recurrent PE among patients with untreated mild PE was 1 of 20 (5%) (Figure 5.3) [11]. The circumstances involving no therapy in these 20 patients from PIOPED are as follows: 19 had pulmonary angiograms interpreted as showing no PE by the local radiologist, but the interpretation of no PE was reversed in 18 following reevaluation by the central panel of angiogram readers [11]. The diagnosis of no PE was reversed in 1 patient because PE was found at autopsy 6 days after the pulmonary angiogram. One patient had no pulmonary angiogram; the patient died of unrelated causes 4 days after a ventilation– perfusion lung scan and autopsy showed small peripheral PE. The untreated patient who died was a 33-year-old woman with underlying primary pulmonary hypertension with right ventricular failure [11]. Organized

PART I

Prevalence, risks, and prognosis of PE and DVT

100

Fatal initial and recurrent PE (%)

BLUK077-Stein

80

60

40

20

0 Clinically apparent PE

Mild PE

Figure 5.3 Fatal initial and fatal recurrent pulmonary embolism (PE) among untreated patients. Comparison is made between patients in whom the diagnosis was clinically apparent, and presumably pulmonary embolism was severe and patients in whom pulmonary embolism was mild. (Data are from Hermann et al. [9], and Stein and Henry [11]. Reprinted from Stein [1] with permission.

and fresh pulmonary emboli were shown at autopsy 6 days after a pulmonary angiogram that failed to show PE. Whether this death resulted from the original PE or recurrent PE is uncertain. In regard to the course of untreated mild PE over 1 year, there were no instances of fatal recurrent PE [11]. This assumes that the cause of death in the only patient who died was the original PE. Fatal recurrent PE, therefore, was 0 of 19 (0%) among untreated survivors of mild PE during months 4–12 of observation [11]. One patient died of aspiration pneumonia following hysterectomy for endometrial carcinoma 4 days after an intermediate probability ventilation–perfusion lung scan was obtained [11]. Multiple small thromboemboli in peripheral branches were observed at autopsy. These thromboemboli did not contribute to death. Based on clinical assessment, recurrent PE was thought to have occurred. The frequency of nonfatal recurrent PE among survivors of the first PE was 1 of 19 (5.3%).


BLUK077-Stein

April 6, 2007

CHAPTER 5

15:7

100 90 80 Fatal initial PE and fatal and nonfatal recurrent PE (%)

33

Changing risks of untreated DVT and acute PE

70 60 50 40 30 20 10 0 Clinical PE

Mild PE

Figure 5.4 Fatal initial pulmonary embolism (PE), fatal recurrent PE, and nonfatal recurrent PE among patients with clinically apparent severe PE and patients with mild PE. (Data are from Hermann et al. [8] and Stein and Henry [10]).

The frequency of fatal initial PE, fatal recurrent PE, and nonfatal recurrent PE over the course of 1 year among untreated patients with acute PE was 2 of 20 (10%) (Figure 5.4). Untreated patients with PE had mild PE as suggested by the following data. Among patients who had measurements of the PaO2 while breathing room air, the PaO2 was lower in untreated patients compared to treated patients (39 ± 16 versus 55 ± 31 mm Hg) (P < 0.001) [11]. The pulmonary artery mean pressure did not show a statistically significant difference between untreated and treated patients (23 ± 13 mm Hg versus 24 ± 10 mm Hg). Ventilation–perfusion (V–Q) lung scans were interpreted as high probability in a smaller percent of untreated patients with PE than treated patients, 0 of 20 (0%) versus 160 of 376 (43%) [11]. Low probability, nearly normal, or normal ventilation–perfusion scans were more frequent among untreated patients 10 of 20 (50%) versus 58 of 376 (15%). Ventilation– perfusion lung scans among untreated patients more

often showed no mismatched segmental equivalent perfusion defects than among treated patients, 14 of 20 (70%) versus 122 of 376 (32%) [11]. All untreated patients, 20 of 20 (100%) showed fewer than 3 mismatched segmental equivalent perfusion defects compared with 227 of 376 (60%) among treated patients. Pulmonary angiograms at the time of PIOPED entry were obtained in 19 of the untreated patients [11]. Thromboemboli involved only segmental pulmonary arteries or smaller branches in 16 of 19 (84%) of untreated patients compared with 132 of 362 (36%) treated patients. Thromboemboli were not observed on the angiogram of 1 untreated patient, but were shown at autopsy 6 days later. The frequency of fatal initial and fatal recurrent PE in untreated patients with mild PE (5.0%) is strikingly lower than the mortality from untreated PE reported in past decades among patients who presumably had severe PE [8, 9]. This lower mortality appears to relate to the milder severity of PE in these untreated patients. The mortality of untreated patients with mild PE is comparable to the mortality from fatal PE in untreated patients with subtle DVT, approximately 5% [3].

Rates of recurrent PE in untreated patients with PE based on calculations in patients with suspected PE and negative serial noninvasive leg tests The estimated frequency of PE during the 3-month follow-up of untreated patients with nonmassive PE and serial noninvasive leg tests was between 3 and 9%. The estimated frequency of fatal PE was 1% [12] (see Chapter 88).

References 1 Stein PD. Changing patterns of risk of untreated thromboembolic disease. Semin Respir Crit Care Med 1996; 17: 3–6. 2 Byrne JJ. Phlebitis: a study of 748 cases at the Boston City Hospital. New Engl J Med 1955; 253: 579–586. 3 Collins R, Scrimgeour A, Yusuf S, Peto R. Reduction in fatal pulmonary embolism and venous thrombosis by perioperative administration of subcutaneous heparin. Overview of results of randomized trials in general, orthopedic, and urologic surgery. New Engl J Med 1988; 318: 1162–1173.


BLUK077-Stein

April 6, 2007

15:7

34

4 Byrne JJ, O’Neil EE. Fatal pulmonary emboli. A study of 130 autopsy-proven fatal emboli. Am J Surg 1952; 83: 47–49. 5 A National Cooperative Study. Clinical and electrocardiographic observations. The Urokinase Pulmonary Embolism Trial. Circulation 1973; 47/48(suppl II): II-60– II-65. 6 Stein PD. Unpublished data from the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). 7 Stein PD, Beemath A, Matta F, et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med 2007 (In press). 8 Barritt DW, Jordan SC. Anticoagulant drugs in the treatment of pulmonary embolism: a controlled trial. Lancet 1960; 1: 1309–1312.

PART I

Prevalence, risks, and prognosis of PE and DVT

9 Hermann RE, Davis JH, Holden WD. Pulmonary embolism: a clinical and pathologic study with emphasis on the effect of prophylactic therapy with anticoagulants. Am J Surg 1961; 102: 19–28. 10 Carson JL, Kelley MA, Duff A et al. The clinical course of pulmonary embolism. New Engl J Med 1992; 326: 1240– 1245. 11 Stein PD, Henry JW, Relyea B. Untreated patients with pulmonary embolism: outcome, clinical and laboratory assessment. Chest 1995; 107: 931–935. 12 Stein PD, Hull RD, Raskob GE. Withholding treatment in patients with acute pulmonary embolism who have a high risk of bleeding and negative serial noninvasive leg tests. Am J Med 2000; 109: 301– 306.


BLUK077-Stein

March 28, 2007

6

15:5

CHAPTER 6

Resolution of pulmonary embolism

Pulmonary emboli resolve because of natural thrombolytic processes [1, 2]. The rate of resolution of perfusion defects, calculated as a percent of the pretreatment defect among 70 patients treated with anticoagulants in the Urokinase Pulmonary Embolism Trial [1], is shown in Figure 6.1. After 24 hours, there was only a 7% mean resolution of the pretreatment perfusion scan defect [1] (Figure 6.1). By 2 days, there was 16% mean resolution. The mean resolution progressively increased to 75% by 3 months, and thereafter, increased only slightly. Among patients with no prior cardiopulmonary disease, ≥90% resolution was shown at 1 year in 29 of 32 (91%) [1] (Figure 6.2).

However, among patients who had prior cardiopulmonary disease, ≥90% resolution was shown at 1 year in only 13 of 18 (72%). Others showed complete clearing of the perfusion scan in 7 of 10 (70%) patients with no prior cardiopulmonary disease [3], and in 22 of 33 (67%), many of who had prior cardiopulmonary disease [4]. In the Urokinase Pulmonary Embolism Trial, the proportion of patients with ≥90% resolution of the perfusion defect was similar in those treated with anticoagulants and those treated with Urokinase in both patients with prior cardiopulmonary disease and patients with no prior cardiopulmonary disease [1] (Figure 6.3).

Figure 6.2 Proportion of patients with <10% residual perfusion defects after 1 year of treatment with anticoagulants according to whether they had cardiopulmonary disease (CPD) or no cardiopulmonary disease. (Data from the Urokinase Pulmonary Embolism Trial [1].)

Mean resolution (%)

60

77

53 42

40

32 16

20

21

7 0 24hrs day 2

Percent of patients with <10% residual perfusion defects

Figure 6.1 Mean resolution of perfusion defects shown as a percentage (%) of pretreatment defects in relation to time of treatment. Hrs = hours, mos = months. (Data from the Urokinase Pulmonary Embolism Trial [1].)

77

75

80

day 3

100 80

day 5

day 7 day 14 3 mos 6 mos 12 mos

91 72

60 40 20 0 CPD (n=18)

No CPD (n=32)

35


BLUK077-Stein

March 28, 2007

15:5

Percent of patients with <10% residual perfusion defects

36

PART I

100 80

AC 72

UK 77

AC 91

Prevalence, risks, and prognosis of PE and DVT

UK 88

60 40 20 0

CPD (n=18)

CPD (n=22)

No CPD No CPD (n=32) (n=37)

Using newer scintigraphic techniques than used in the Urokinase Pulmonary Embolism Trial, and somewhat different methods of calculating the improvement of perfusion, de Groot showed a mean rate of improvement of the perfusion scan of 4% in 2–4 days [5]. Importantly, a high probability interpretation of the lung scan remained high probability in 77 of 79 patients (97%) during the 2–4 day period. The V–Q scan became nondiagnostic in 1 of 79 (1%) and it became normal in 1 of 79 (1%) [5]. Resolution of acute PE based on pulmonary angiograms obtained 1 to 7 days after the initial angiogram in 6 patients showed no change in 1 and minimal resolution with persistent marked abnormalities in 5 [2]. Ten to 21 days after the initial pulmonary angiogram, among 10 patients, no resolution was shown in 1, minimal resolution in 2, moderate resolution with persistent minimal abnormalities in 5, and complete resolution (normal angiogram) in 2. After 34 days, among 2 patients, 1 showed minimal resolution and 1 showed complete resolution. The pulmonary diffusing capacity of carbon monoxide (DLCO) after 1 year was only 72% of predicted among 21 patients with PE and no prior cardiopulmonary disease, who were treated with anticoagulants [6]. Others also observed a continuing impairment of DLCO in 10 patients treated with anticoagulants, although the perfusion lung scans tended to return to normal [3]. Among 11 patients such pa-

Figure 6.3 Percent of patients with <10% residual perfusion defects after treatment with anticoagulants (AC) or Urokinase (UK) according to whether they had cardiopulmonary disease (CPD) or no cardiopulmonary disease. (Data from The Urokinase Pulmonary Embolism Trial [1].)

tients followed a mean of 7.4 years, mean pulmonary artery pressure (22 mm Hg) was somewhat elevated and pulmonary vascular resistance at rest was also elevated (171 dyne –sec/cm−5 ) [7].

References 1 The Urokinase Pulmonary Embolism Trial. A national cooperative study. Perfusion lung scanning. Circulation 1973; 47(2 suppl): II46–II50. 2 Dalen JE, Banas JS, Jr, Brooks HL, Evans GL, Paraskos JA, Dexter L. Resolution rate of acute pulmonary embolism in man. N Engl J Med 1969; 280: 1194–1199. 3 Wimalaratna HS, Farrell J, Lee HY. Measurement of diffusing capacity in pulmonary embolism. Respir Med 1989; 83: 481–485. 4 Paraskos JA, Adelstein SJ, Smith RE et al. Late prognosis of acute pulmonary embolism. N Engl J Med 1973; 289: 55–58. 5 de Groot MR, Oostdijk AH, Engelage AH, van Marwijk Kooy M, Buller HR. Changes in perfusion scintigraphy in the first days of heparin therapy in patients with acute pulmonary embolism. Eur J Nucl Med 2000; 27: 1481– 1486. 6 Sharma GV, Burleson VA, Sasahara AA. Effect of thrombolytic therapy on pulmonary–capillary blood volume in patients with pulmonary embolism. N Engl J Med 1980; 303: 842–845. 7 Sharma GV, Folland ED, McIntyre KM, Sasahara AA. Long-term benefit of thrombolytic therapy in patients with pulmonary embolism. Vasc Med 2000; 5: 91–95.


BLUK077-Stein

March 12, 2007

7

15:14

CHAPTER 7

Upper extremity deep venous thrombosis

Prior to 1967, thrombosis of the upper extremities constituted less than 2% of cases of deep venous thrombosis (DVT) [1]. Since the 1970s, there has been an increased recognition of upper extremity DVT [2–6]. The incidence of upper extremity DVT in a community teaching general hospital was reviewed during the 2-year period, July 1998 through June 2000 [7]. The incidence of upper extremity DVT in adults (≥20 years) was 64 of 34,567 hospital admissions (0.19%). The incidence of upper extremity DVT that we observed among hospitalized patients was the same as reported by Kroger and associates (0.2%) [3]. Upper extremity DVT in adults was accompanied by proximal DVT of the lower extremity in 2 of 64 patients [7] (see Figure 7.1). Upper extremity DVT involved the subclavian vein in 48 patients (75%) and the axillary vein in 25 patients (39%) [7]. The internal jugular vein was included among patients with upper extremity DVT, and it involved 29 patients (45%) [7]. Nine patients (14%) had involvement only of a deep distal vein (brachial

Figure 7.1 Patient with upper extremity deep venous thrombosis (DVT). (Courtesy of Syed Mustafa, MD, St. Joseph Mercy-Oakland Hospital, Pontiac Michigan.)

veins in 6 patients, ulnar vein alone in 1 patient, radial vein alone in 1 patient, and both radial and ulnar veins in 1 patient). In 7 patients, the upper extremity DVT was shown by venography to extend proximally to the brachiocephalic vein. Among these, 2 extended to the superior vena cava. In addition to these patients with DVT of the upper extremity, 16 patients had involvement only of the superficial veins of the upper extremity. These patients are not included in the various computations. All the patients with upper extremity DVT received therapy with anticoagulants [7]. None developed pulmonary embolism (PE). Cancer was diagnosed in 30 of 65 (46%). Upper extremity DVT in the past had been considered to be benign and self-limited [8]. More recent data suggest that this is not the case [8–11]. Systematic review of the literature in 1991 showed a 9% incidence of symptomatic PE in patients with upper extremity DVT, half of which was diagnosed by objective tests [12]. Review in 1991 by others showed a 7% incidence of symptomatic PE in patients with upper extremity DVT, none of which was confirmed by objective tests [13]. More recently, symptomatic PE diagnosed by ventilation–perfusion lung scans was observed in 7% of patients with upper extremity DVT [8]. Most PE (94%) in patients with upper extremity DVT occurred in untreated patients [9]. Fatal PE in patients with upper extremity DVT has been reported [10]. Routine ventilation/perfusion lung scans in patients with upper extremity DVT were high probability for PE in 13% [13]. The increased incidence of upper extremity DVT has been attributed to the use of central venous catheters and transvenous pacemakers [2]. As the use of central venous lines and pacemaker wires has increased, their role in the etiology of upper extremity DVT has become prominent [3, 14–17]. The incidence of central venous catheter-related DVT assessed by venography has been reported in 27–66% of patients [18]. Most of

37


BLUK077-Stein

March 12, 2007

15:14

38

PART I

the thrombi in these cases were asymptomatic. The reported incidence of symptomatic catheter-related DVT in adults ranged from 0.3 to 28.3% [18]. In the series we reported, central venous access lines on the side of the upper extremity DVT were inserted in 39 of 65 (60%) [7]. Nearly the same percentage (55%) of upper extremity DVT was associated with central venous access lines in the experience of others [19]. Others reported the use of indwelling catheters in 28–33% of patients with upper extremity DVT [3, 9, 14]. Six patients in our series had arteriovenous shunts on the side of the upper extremity DVT [7]. There were 3 additional patients who had thrombosis only of the arteriovenous shunt without involvement of contiguous veins. These patients were not included among the patients we reported. Nineteen patients (29%) had no apparent cause, although all patients developed upper extremity DVT in the hospital and had received intravenous infusions of medications. Three of 39 patients (8%) who had central venous access lines had been on antithrombotic prophylaxis with lowdose warfarin (1–2 mg/day) prior to developing upper extremity DVT. Among 30 patients with venous access lines and upper extremity DVT in whom there was data, the lines had been inserted 3–14 days in 90%. Swelling of the arm was the most frequent sign, and was present in all patients with upper extremity DVT. Pain was present in 26 of 65 (40%) patients. Some discomfort due to the swelling was present in all. Four of 65 (6%) had erythema over the affected site. One patient with internal jugular vein thrombosis had swelling of the neck. He also had thrombosis of the superior vena cava. Campbell and associates, among 25 patients with upper extremity DVT, observed swelling in 96%, pain in 76%, discoloration in 52%, prominent veins in 52%, and a palpable cord in 8% [2].

We observed an association of malignancy with upper extremity DVT in 29 of 65 (45%) [7]. Among these patients, 23 also had a central line. An association of upper extremity DVT with malignancy is well established [20] and was reported in 64% by Baarslag et al. [19]. A hypercoagulable state may also be associated with proximal upper vein thrombosis [21], but our patients generally were not evaluated for this. During the 2-year period of this investigation, 10 of 34,567 patients (0.03%) had thrombosis of the superior vena cava or brachiocephalic vein, unaccompanied by DVT of the subclavian or axillary veins or more distal veins of the upper extremity and unaccompanied by proximal DVT of the lower extremity [22] (Figure 7.2). During the same period, proximal DVT of the lower extremity was diagnosed in 271 of 34,567 patients (0.78%) [23] (Figure 7.2). Deep venous thrombosis of the upper extremity, sometimes accompanied by thrombosis of the superior vena cava or brachiocephalic vein, but unaccompanied by proximal DVT of the lower extremity, was observed in 62 of 34,567 adult patients (0.18%) [7]. Among all patients with symptomatic DVT, thrombosis of the superior vena cava or brachiocephalic vein alone constituted 10 of 343 (3%) [7, 22, 23] (Figure 7.3). Upper extremity DVT, unaccompanied by proximal lower extremity DVT but sometimes accompanied by thrombosis of the superior vena cava or brachiocephalic vein, constituted 62 of 343 (18%), and proximal lower extremity DVT constituted 271 of 343 (79%). More recently, a prospective registry of 324 patients with central venous catheter-associated upper extremity DVT and 268 patients with noncentral venous catheter-associated upper extremity DVT was published [24]. An indwelling central venous catheter was

Prevalence, risks, and prognosis of PE and DVT

DVT in hosp pts (%)

1 0.8 0.6

0.78%

0.4 0.18%

0.2

0.03%

0 Lower extremity

Upper extremity, SVC, or brachioceph

SVC or brachioceph

Figure 7.2 Prevalence in hospitalized (hosp) patients (pts) of proximal lower extremity deep venous thrombosis (DVT), upper extremity, superior vena cava (SVC) or brachiocephalic (brachioceph) DVT unaccompanied by lower extremity DVT, and isolated SVC or brachiocephalic DVT. (Data from Mustafa et al. [7], Otten et al. [22], and Stein et al. [23].)


BLUK077-Stein

March 12, 2007

CHAPTER 7

15:14

Upper extremity deep venous thrombosis

SVC or brachiocephalic 3%

Upper extremity 18%

Lower extremity 79%

Figure 7.3 Proportion of patients with proximal deep venous thrombosis (DVT) of lower extremities, upper extremity DVT, unaccompanied by proximal lower extremity DVT but sometimes accompanied by thrombosis of the superior vena cava or brachiocephalic vein. (Data from Mustafa et al. [7], Otten et al. [22], and Stein et al. [23].)

the strongest independent predictor of upper extremity DVT. Only 20% of patients with upper extremity DVT who did not have a contraindication to anticoagulation were receiving anticoagulant prophylaxis at the time of diagnosis of upper extremity DVT [24]. Frequent risk factors in patients with central venous catheter-associated upper extremity DVT and in patients with noncentral venous catheter-associated upper extremity DVT were cancer, personal history, or family history of venous thromboembolism (VTE). Surgery or immobilization within 30 days, and ongoing chemotherapy were frequent risk factors in patients with central venous catheter-associated upper extremity DVT. Ongoing chemotherapy was a frequent risk factor in patients with noncentral venous catheterassociated upper extremity DVT [24]. Among 324 patients with upper extremity DVT, 25% had upper extremity swelling, 11% had upper extremity discomfort, and 4% had erythema [24]. Pulmonary embolism at the time of diagnosis of upper extremity DVT was confirmed in 0.9%.

References 1 Coon WW, Willis PW, III. Thrombosis of axillary and subclavian veins. Arch Surg 1967; 94: 657–663. 2 Campbell CB, Chandler JG, Tegtmeyer CJ, Bernstein EF. Axillary, subclavian, and brachiocephalic vein obstruction. Surgery 1977; 82: 816–826.

39

3 Kroger K, Schelo C, Gocke C, Rudofsky G. Colour Doppler sonographic diagnosis of upper limb venous thromboses. Clin Sci 1998; 94: 657–661. 4 Huber P, Hauptli W, Schmitt HE, Widmer LK. [Axillary and subclavian vein thrombosis and its sequelae]. Internist (Berl) 1987; 28: 336–343. 5 Theis W, Zaus M, Klefhaber M et al. Primare und sekundare Schultergurtelvenenthrombose; eine Analyse von 227 Patienten [Abstract]. VASA 1994; l43(suppl): 167– 169. 6 Layher T, Heinrich F. Retrospektive Betrachtung von Arm- bzw-Schultervenenthrombosen am Krankenhaus Bruchsal im Zeitraum von 1973 und 1993 [Abstract]. VASA 1994; 43(suppl): 169. 7 Mustafa S, Stein PD, Patel KC, Otten TR, Holmes R, Silbergleit A. Upper extremity deep venous thrombosis. Chest 2003; 123: 1953–1956. 8 Hingorani A, Ascher E, Lorenson E et al. Upper extremity deep venous thrombosis and its impact on morbidity and mortality rates in a hospital-based population. J Vasc Surg 1997; 26: 853–860. 9 Horattas MC, Wright DJ, Fenton AH et al. Changing concepts of deep venous thrombosis of the upper extremityreport of a series and review of the literature. Surgery 1988; 104: 561–567. 10 Monreal M, Raventos A, Lerma R et al. Pulmonary embolism in patients with upper extremity DVT associated to venous central lines—a prospective study. Thromb Haemost 1994; 72: 548–550. 11 Prandoni P, Polistena P, Bernardi E et al. Upperextremity deep vein thrombosis. Risk factors, diagnosis, and complications. Arch Intern Med 1997; 157: 57–62. 12 Becker DM, Philbrick JT, Walker FB, IV. Axillary and subclavian venous thrombosis. Prognosis and treatment. Arch Intern Med 1991; 151: 1934–1943. 13 Monreal M, Lafoz E, Ruiz J, Valls R, Alastrue A. Upperextremity deep venous thrombosis and pulmonary embolism. A prospective study. Chest 1991; 99: 280– 283. 14 Timsit JF, Farkas JC, Boyer JM et al. Central vein catheterrelated thrombosis in intensive care patients: incidence, risks factors, and relationship with catheter-related sepsis. Chest 1998; 114: 207–213. 15 Haire WD, Lieberman RP, Edney J et al. Hickman catheter-induced thoracic vein thrombosis. Frequency and long-term sequelae in patients receiving high-dose chemotherapy and marrow transplantation. Cancer 1990; 66: 900–908. 16 Ryan JA, Jr, Abel RM, Abbott WM et al. Catheter complications in total parenteral nutrition. A prospective study of 200 consecutive patients. N Engl J Med 1974; 290: 757–761.


BLUK077-Stein

March 12, 2007

15:14

40

PART I

17 Dollery CM, Sullivan ID, Bauraind O et al. Thrombosis and embolism in long-term central venous access for parenteral nutrition. Lancet 1994; 344: 1043–1045. 18 Verso M, Agnelli G, Venous thromboembolism associated with long-term use of central venous catheters in cancer patients. J Clin Oncol 2003; 21: 3665–3675. 19 Baarslag HJ, van Beek EJ, Koopman MM, Reekers JA. Prospective study of color duplex ultrasonography compared with contrast venography in patients suspected of having deep venous thrombosis of the upper extremities. Ann Intern Med 2002; 136: 865–872. 20 Prandoni P, Bernardi E. Upper extremity deep vein thrombosis. Curr Opin Pulm Med 1999; 5: 222–226.

21 Martinelli I, Cattaneo M, Panzeri D, Taioli E, Mannucci PM. Risk factors for deep venous thrombosis of the upper extremities. Ann Intern Med 1997; 126: 707–711. 22 Otten TR, Stein PD, Patel KC, Mustafa S, Silbergleit A. Thromboembolic disease involving the superior vena cava and brachiocephalic veins. Chest 2003; 123: 809– 812. 23 Stein PD, Patel KC, Kalra NK et al. Deep venous thrombosis in a general hospital. Chest 2002; 122: 960–962. 24 Joffe HV, Kucher N, Tapson VF, Goldhaber SZ; for the deep vein thrombosis (DVT) FREE Steering Committee. Upper-extremity deep vein thrombosis: a prospective registry of 592 patients. Circulation 2004; 110: 1605–1611.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

March 12, 2007

8

15:15

CHAPTER 8

Thromboembolic disease involving the superior vena cava and brachiocephalic veins

Isolated venous thromboembolic disease of the brachiocephalic veins and superior vena cava (SVC) was observed in hospitalized adults ≥20 years old in 10 of 34,567 (0.03%) [1] (see Chapter 7). It was accompanied by deep venous thrombosis (DVT) at other sites in an additional 13 patients (0.04%). Previous data are limited to 2 case reports [2, 3]. Thrombosis of the SVC is an uncommon cause of the SVC syndrome [4–6]. Rarely, pulmonary embolism (PE) with the SVC syndrome has been observed at autopsy [7, 8]. It was suggested, however, that SVC thrombosis may pose a significant risk for PE [9]. Brachiocephalic vein thrombosis was observed in 22 patients and SVC thrombosis was observed in 6 patients [1]. The number of patients with brachiocephalic vein thrombosis alone or in combination with SVC thrombosis or with thrombosis of the subclavian or axillary veins is shown in Table 8.1 [1]. The diagnosis was made by contrast venography in 21 patients, and by contrast enhanced spiral CT in 2 patients. All 6 patients with SVC thrombosis showed incomplete obstruction of the SVC. Among the 22 patients with brachiocephalic vein involvement, 13 showed total occlusion or occlusion sufficient to produce collateral veins.

Table 8.2 Predisposing factors among patients with brachiocephalic or superior vena cava thrombosis (n = 23).

Table 8.1 Vessels showing thrombosis (n = 23). SVC only

Two of 23 (8.7%) patients had nonfatal PE, both of which were diagnosed by high-probability ventilation– perfusion lung scans [1]. One patient had brachiocephalic vein thrombosis and, in addition, had bilateral thrombosis of the axillary, subclavian, and internal jugular veins. The PE occurred before the venous thrombosis was diagnosed and treated. The other patient had brachiocephalic vein thrombosis with no other upper body DVT, but in addition had proximal lower extremity DVT. The PE occurred while on treatment for lower extremity DVT. This was the only patient with coincident DVT of the lower extremities. To our knowledge, only one case of PE from brachiocephalic thrombosis has been reported previously [10]. Predisposing factors for patients with thrombosis of the SVC or brachiocephalic vein are shown in Table 8.2 [1]. Among patients who had central venous access lines, 10 had Groshong catheters, 4 had Infusaports, and 1 had a PICC line (peripherally inserted central catheter) (all from Bard Access Systems, Salt Lake City, Utah). An association of use of central venous access lines with thrombosis has been described [11–15]. Three patients with brachiocephalic vein thrombosis had neither cancer nor central venous lines, but had

1

Number

Percent

SVC + brachiocephalic

1

Cancer and central venous access line

14

61

SVC + brachiocephalic + subclavian or axillary

4

Cancer alone

3

13

Brachiocephalic only

9*

Central venous access line alone

1

4

Brachiocephalic + subclavian or axillary

8

External mechanical compression;

1

4

4

17

not cancer * One of these patients also had DVT of a lower extremity.

Cause not apparent

SVC, superior vena cava. Reprinted with permission from Otten et al. [1].

Reprinted with permission from Otten et al. [1].

41


BLUK077-Stein

March 12, 2007

15:15

42

PART I

Table 8.3 Signs and symptoms, SVC and/or brachicephalic vein thrombosis.

2 Kwong T, Leonidas JC, Ilowite NT. Asymptomatic superior vena cava thrombosis and pulmonary embolism in an adolescent with SLE and antiphospholipid antibodies. Clin Exp Rheum 1994; 12: 215–217. 3 Goldstein MF, Nestico P, Olshan AR et al. Superior vena cava thrombosis and pulmonary embolus: association with right atrial mural thrombus. Arch Intern Med 1982; 142: 1726–1728. 4 Gucalp R, Dutcher J. Oncologic emergencies. In: Braunwald E, Fauci AS, Kasper DL et al., eds. Harrison’s Principles of Internal Medicine, 15th edn. McGraw-Hill, New York, 2001: 642–650. 5 Abner A. Approach to the patient who presents with superior vena cava obstruction. Chest 1993; 103: 394S–397S. 6 Salsali M, Cliffton EE. Superior vena caval obstruction with lung cancer. Ann Thorac Surg 1968; 6: 437–442. 7 Maddox A-M, Valdivieso M, Lukeman J et al. Superior vena cava obstruction in small cell bronchogenic carcinoma. Clinical parameters and survival. Cancer 1983; 52: 2165–2172. 8 Ryan JA, Abel RM, Abbott WM et al. Catheter complications in total parenteral nutrition. A prospective study of 200 consecutive patients. N Engl J Med 1974; 290: 757– 761. 9 Adelstein DJ, Hines JD, Carter SG et al. Thromboembolic events in patients with malignant superior vena cava syndrome and the role of anticoagulation. Cancer 1988; 62: 2258–2262. 10 Black MD, French GJ, Rasuli P et al. Upper extremity deep venous thrombosis. Underdiagnosed and potentially lethal. Chest 1993; 103: 1887–1890. 11 Prandoni P, Polistena P, Bernardi E et al. Upper-extremity deep vein thrombosis. Risk factors, diagnosis, and complications. Arch Intern Med 1997; 157: 57–62. 12 Torosian MH, Meranze S, McLean G et al. Central venous access with occlusive superior central venous thrombosis. Ann Surg 1986; 203: 30–33. 13 Haire WD, Lieberman RP, Edney J et al. Hickman catheter- induced thoracic vein thrombosis. Frequency and long-term sequelae in patients receiving high-dose chemotherapy and marrow transplantation. Cancer 1990; 66: 900–908. 14 Gore JM, Matsumoto AH, Layden JJ et al. Superior vena cava syndrome. Its association with indwelling ballooning-tipped pulmonary artery catheters. Arch Intern Med 1984; 144: 506–508. 15 Dollery CM, Sullivan ID, Bauraind O. Thrombosis and embolism in long-term central venous access for parenteral nutrition. Lancet 1994; 344: 1043–1045. 16 Knudson GJ, Wiedmeyer DA, Erickson SJ et al. Color Doppler sonographic imaging in the assessment of upperextremity deep venous thrombosis. Am J Roentgenol 1990; 154: 399–403.

Number (percent) All patients

Brachiocephalic vein

(n = 23)

only* (n = 9)

Arm edema

18 (78)

7 (78)

Pain/discomfort

15 (65)

5 (56)

Head/neck edema

10 (43)

3 (33)

Distended veins

7 (30)

2 (22)

Erythema

5 (22)

2 (22)

2 (9)

0 (0)

1 (4)

1 (11)

CNS symptoms None (incidental finding)

* Patients with accompanying thrombosis of axillary, subclavian, or internal jugular vein were excluded. CNS, central nervous system. Reprinted with permission from Otten et al. [1].

arteriovenous shunts for renal dialysis [1]. There was no evidence of thrombosis of the shunts. One of these patients had subclavian vein thrombosis in addition to thrombosis of the brachiocephalic vein. The other 2 patients had no evidence of thrombosis at other sites. The patient who had SVC thrombosis alone had no clear cause for the thrombosis. She was taking estrogen replacement therapy following menopause. Signs and symptoms observed in all patients with SVC thrombosis and/or brachiocephalic vein thrombosis and in 9 patients who showed involvement only of a brachiocephalic vein are shown in Table 8.3 [1]. In the latter group, the thrombosis caused partial occlusion in 6 of 9 (67%). Three of those with partial occlusion showed collateral vessels. Contrast venography is the most conclusive test [16–20]. Helical computed tomographic phlebography [21], magnetic resonance angiography [19, 20], and gadolinium-enhanced magnetic resonance venography [22] may be useful. Color duplex Doppler ultrasonography cannot image the SVC and proximal segment of the brachiocephalic veins [16, 23].

References 1 Otten TR, Stein PD, Patel KC, Mustafa S, Silbergleit A. Thromboembolic disease involving the superior vena cava and brachiocephalic veins. Chest 2003; 123: 809–812.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

March 12, 2007

CHAPTER 8

15:15

VTE involving SVC and brachiocephalic veins

17 Gooding GA, Hightower DR, Moore EH et al. Obstruction of the superior vena cava or subclavian veins: sonographic diagnosis. Radiology 1986; 159: 663–665. 18 Schwartz EE, Goodman LR, Haskin ME. Role of CT scanning in the superior vena cava syndrome. Am J Clin Oncol 1986; 9: 71–78. 19 Hartnell GG, Hughes LA, Finn JP et al. Magnetic resonance angiography of the central chest veins. A new gold standard? Chest 1995; 107: 1053–1057. 20 Finn JP, Zisk JHS, Edelman RR et al. Central venous occlusion: MR angiography. Radiology 1993; 187: 245–251.

43

21 Qanadli SD, EL Hajjam M, Bruckert F et al. Helical CT phlebography of the superior vena cava: diagnosis and evaluation of venous obstruction. Am J Roentgenol 1999; 172: 1327–1333. 22 Kroencke TJ, Taupitz M, Arnold R et al. Threedimensional gadolinium-enhanced magnetic resonance venography in suspected thrombo-occlusive disease of the central chest veins. Chest 2001; 120: 1570–1576. 23 Falk RL, Smith DF. Thrombosis of upper extremity thoracic inlet veins: diagnosis with duplex Doppler sonography. Am J Roentgenol 1987; 149: 677–682.


BLUK077-Stein

March 28, 2007

9

15:5

CHAPTER 9

Venous thromboembolic disease in the four seasons

Introduction

26

30

11.0

11.8

12.1

10

12.0

20

O ct −D ec

ep −S ly Ju

un e Ap r− J

Ja

n−

M ar

0

44

There is no meaningful seasonal variation of mortality from PE [13]. From 1980 to 1998, quarterly mortality rates throughout the entire United States, based on data from the United States National Center for Health Statistics, ranged from 0.93 to 1.05 PE deaths/quarter/100,000 population [13] (Table 9.1). Quarterly mortality rates from PE in four regions of the United States (Northeast, South, Midwest, and West) were evaluated separately [13] (Table 9.2). All were within 1 PE death/quarter/100,000 population [13]. Small differences were statistically significant due to the large number of patients evaluated (184,201 death certificates indicating PE as the cause of death). Recognizing that the death certificate diagnosis of fatal or large PE is accurate in only 32–35% of patients [14, 15] we assumed that whatever inaccuracy exists in the death certificates was constant throughout the seasons [13]. Differing observations had been reported previously on seasonal differences of mortality from acute PE. Several investigators reported peak mortality rates

.2

.7 27

28

.0 28

40 Rate per 100,000

.1

An absence of seasonal variation was shown in all regions of the United States, including the Southern region where winters are mild, and the Northeastern and Midwestern regions where seasons are sharply defined [1] (Figure 9.1). These observations were based on 21 years of data from the National Hospital Discharge Survey [1]. The results were based on data in 2,457,000 hospitalized patients with pulmonary embolism (PE) and 5,767,000 hospitalized patients with deep venous thrombosis (DVT). The data apply to patients with broad differences of ethnic, social, and racial backgrounds and to regions with wide differences in climate. This absence of seasonal variation was concordant with results of others who found no seasonal variations in PE [2] or DVT [3–5]. Seasonal variability of thromboembolic disease has been suggested to be present since 1939 [6]. Some showed a peak incidence in winter months for PE [2–8], fatal PE [7], DVT [8, 9], and thromboembolic disease [10]. Others showed a decreased incidence of PE in the winter, or a peak in spring and autumn [11] or peaks in both summer and winter [12].

Mortality from acute PE according to season

DVT PE

Figure 9.1 Rates of diagnosis in hospitalized patients/100,000 population for pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) according to quarter of year. Data are averaged from 1979 to 1999. (Reprinted from Stein et al. [1], with permission from Elsevier.)


BLUK077-Stein

March 28, 2007

CHAPTER 9

15:5

45

Venous thromboembolic disease in the four seasons

Table 9.1 Average quarterly rate of deaths from pulmonary embolism: 1980–1998 (PE deaths/100,000 population). Jan–Mar

Apr–Jun

Jul–Sep

Oct–Dec

All seasons

Northeast

1.11

0.98

0.97

1.03

1.02

Midwest

1.14

1.00

0.99

1.06

1.05

South

1.22

1.06

1.09

1.16

1.13

West

0.65

0.59

0.60

0.65

0.62

All regions

1.03

0.91

0.91

0.98

0.96

Differences between regions: South > Northeast, Midwest, West (P < 0.05); Northeast, Midwest > West (P < 0.05). Differences between seasons: Jan–Mar > Apr–Jun, Jul–Sep, Oct–Dec (P < 0.05); Oct–Dec > Apr–Jun, Jul–Sep (P < 0.05). PE, pulmonary embolism. Reprinted with permission from Stein et al. [13].

in the first quarter of the year [16–20], sometimes with overlap in the last quarter [21] and sometimes with second peaks in the third quarter [16, 17]. Others reported peak mortality rates in the second quarter [22–24], sometimes with second peaks in the third and fourth quarter [22], or fourth quarter alone [23]. Some reported peaks only in the third and fourth quarter [25]. Some reported more frequent fatal PE during “fine weather phases” of the year and “at the beginning of fine weather” [26]. Some reported no quarterly varia-

tion [27, 28]. Many of these investigations were based on observations in less than 200 patients [16–18, 20, 21]. The largest investigation included less than 1500 patients [26]. The absence of meaningful seasonal variation of the rate of diagnosis in hospitalized patients [1] and mortality from PE [13] based on data from several thousands of patients indicate that PE is not affected by the season, contrary to reports based on smaller investigations.

Table 9.2 Regions of United States defined according to states and district of Columbia. West

Midwest

South

Northeast

Alaska

Illinois

Delaware

Maine

Arizona

Indiana

Maryland

New Hampshire

California

Iowa

District of Columbia

Vermont

Colorado

Kansas

Virginia

Massachusetts

Hawaii

Michigan

West Virginia

Connecticut

Idaho

Minnesota

North Carolina

Rhode Island

Montana

Missouri

South Carolina

New York

Nevada

Nebraska

Georgia

New Jersey

New Mexico

North Dakota

Florida

Pennsylvania

Oregon

Ohio

Kentucky

Utah

South Dakota

Tennessee

Washington

Wisconsin

Wyoming

Alabama Mississippi Arkansas Louisiana Oklahoma Texas

Reprinted from Stein et al. [1], with permission from Elsevier.


BLUK077-Stein

March 28, 2007

15:5

46

PART I

References

15 Attems J, Arbes S, Bohm G, Bohmer F, Lintner F. The clinical diagnostic accuracy rate regarding the immediate cause of death in a hospitalized geriatric population; an autopsy study of 1594 patients. Wien Med Wochenschr 2004; 154: 159–162. 16 Chau KY, Yuen ST, Wong MP. Seasonal variation in the necropsy incidence of pulmonary thromboembolism in Hong Kong. J Clin Pathol 1995; 48: 578–579. 17 Colantonio D, Casale R, Natali G, Pisqualetti P. Seasonal periodicity in fatal pulmonary thromboembolism. Lancet 1990; 335: 56–57. 18 Gallerani M, Manfredini R, Ricci L et al. Sudden death from pulmonary thromboembolism: chronobiological aspects. Eur Heart J 1992; 13: 661–665. 19 Mobius C, Gunther U, Klinker L, Putzke HP. [Meteoropathologic effects on the development of fatal lung embolism]. Z Gesamte Hyg 1989; 35: 391–392. 20 Manfredini R, Gallerani M, Salmi R, Zamboni P, Fersini C. Fatal pulmonary embolism in hospitalized patients: evidence for a winter peak. J Int Med Res 1994; 22: 85– 89. 21 Wroblewski BM, Siney PD, White R. Fatal pulmonary embolism after total hip arthroplasty. Seasonal variation. Clin Orthop Relat Res 1992; 276: 222–224. 22 Green J, Edwards C. Seasonal variation in the necropsy incidence of massive pulmonary embolism. J Clin Pathol 1994; 47: 58–60. 23 Hackl H. [Environmental effects and pulmonary embolism]. Dtsch Med J 1968; 19: 475–477. 24 Montes Santiago J, Rey Garcia G, Mediero Dominguez A. [Seasonal changes in morbimortality caused by pulmonary thromboembolism in Galicia]. An Med Interna 2003; 20: 457–460. 25 Steiner I, Matejek T. [Pulmonary embolism–temporal aspects]. Cesk Patol 2003; 39: 185–188. 26 Putzke HP, Mobius C, Gunther U, Bargenda M, Dobberphul J. [The incidence of fatal lung emboli with special reference to the underlying disease and the effect of weather]. Z Gesamte Inn Med 1989; 44: 106–110. 27 Coon WW, Coller FA. Some epidemiologic considerations of thromboembolism. Surg Gynecol Obstet 1959; 109: 487–501. 28 Golin V, Sprovieri SR, Bedrikow R, Salles MJ. Pulmonary thromboembolism: retrospective study of necropsies performed over 24 years in a university hospital in Brazil. Sao Paulo Med J 2002; 120: 105–108.

1 Stein PD, Kayali F, Olson RE. Analysis of occurrence of venous thromboembolic disease in the four seasons. Am J Cardiol 2004; 93: 511–513. 2 Galle C, Wautrecht JC, Motte S et al. The role of season in the incidence of deep vein thrombosis. J Mal Vasc 1998; 23: 99–101. 3 Bounameaux H, Hicklin L, Desmarais S. Seasonal variation in deep vein thrombosis. BMJ 1996; 312: 284–285. 4 Luthi H, Gruber UF. Is there a seasonal fluctuation in the appearance of deep venous thrombosis? Anasth Intensivther Notfallmed 1982; 17(3): 158–160. 5 National Hospital Discharge Survey Multi-year Data File 1979–1999. CD-ROM Series 13, No. 19A. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics, Hyattsville, MD, reissued March 2001. http://www.cdc.gov/nchs/about/major/hdasd/nhds.htm. 6 Oschner A, DeBakey M. Thrombophlebitis and phlebothrombosis. South Surg 1939; 8: 269–290. 7 Lawrence JC, Xabregas A, Gray L, Ham JM. Seasonal variation in the incidence of deep vein thrombosis. Br J Surg 1977; 64: 777–780. 8 Boulay F, Berthier F, Schoukroun G, Raybaut C, Gendreike Y, Blaive B. Seasonal variations in hospital admission for deep vein thrombosis and pulmonary embolism: analysis of discharge data. BMJ 2001; 323: 601–602. 9 Ferrari E, Baudouy M, Cerboni P et al. Clinical epidemiology of venous thromboembolic disease. Results of a French multicentre registry. Eur Heart J 1997; 18: 685– 691. 10 Green J, Edwards C. Seasonal variation in the necropsy incidence of massive pulmonary embolism. J Clin Pathol 1994; 47: 58–60. 11 Bilora F, Manfredini R, Petrobelli F, Vettore G, Boccioletti V, Pomerri F. Chronobiology of non fatal pulmonary thromboembolism. Panminerva Med 2001; 43: 7–10. 12 Coon WW. The spectrum of pulmonary embolism: twenty years later. Arch Surg 1976; 111: 398–402. 13 Stein PD, Kayali F, Beemath A et al. Mortality from acute pulmonary embolism according to season. Chest 2005; 128: 3156–3158. 14 Dismuke SE, VanderZwaag R. Accuracy and epidemiological implications of the death certificate diagnosis of pulmonary embolism. J Chronic Dis 1984; 37: 67–73.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

April 6, 2007

10

15:8

CHAPTER 10

Regional differences in the United States of rates of diagnosis of pulmonary embolism and deep venous thrombosis and mortality from pulmonary embolism

Regional rates of diagnosis of pulmonary embolism and deep venous thrombosis The Western region of the United States from 1979 to 2001 showed lower rates of diagnosis of deep venous thrombosis (DVT) and venous thromboembolism (VTE) in hospitalized patients than any other region [1] (Table 10.1, Figure 10.1). The rates of diagnosis of DVT and VTE were lower in the Western region than in other regions from 1979 to 1989 and remained lower from 1990 to 2001 (Table 10.1). Rate ratios of the rates of the diagnosis of DVT, PE, and VTE comparing the Western region to other regions ranged from 0.65 to 0.87 [1] (Table 10.2). Rates of diagnosis were based on data from The National Hospital Discharge Survey [2]. Population estimates were from the United States Bureau of the Census [3]. Regions of the United States were defined by the National Hospital Discharge Survey (see Table 10.2, Chapter 9). In Caucasians, from 1979 to 2001, the rates of diagnosis of DVT and VTE were lower in the Western region than all other regions and the rate of diagnosis of PE was lower in the West than other regions except the Midwest [1]. In African Americans, the rates of diagnosis of PE, DVT, and VTE were lower in the West than in the Midwest (Table 10.3). In both men and women, the rates of diagnosis of DVT and VTE were lower in the West than any other region [1] (Table 10.4). Within each region, the rates of diagnosis of DVT and VTE in men were lower than in women [1] (Table 10.4).

In patients ≥65 years, rates of diagnosis of DVT and VTE were lower in the Western region than other regions, but there was only a trend toward a lower rate of PE in the Western region [1]. In the Western region, rates of DVT in men and women ≥65 years were comparable, although in the Midwestern and Southern regions, rates of DVT were higher in women ≥65 years than in men ≥65 years. Caucasians ≥65 years of age and <65 years of age, from 1979 to 2001, had lower rates of diagnosis of DVT and VTE in the Western region than in other regions (P < 0.01 to P < 0.001). African Americans ≥65 years showed no regional differences in rates of diagnosis. Younger African Americans (aged <65 years) showed lower rates of DVT, PE, and VTE in the West than the Midwest [1]. Relatively low rates of diagnosis of PE and DVT were observed on the Pacific and Atlantic coasts of the United States from 1986 to 1989 in patients ≥65 years, based on a sample of Medicare enrollees [4].

Regional mortality rates from PE The PE mortality rate from 1979 to 1998 was lower in the Western region than any other region [1] (Figure 10.2). It was lower in the Western region in both men and women and in African Americans and Caucasians (all P < 0.001). Rate ratios for rates of mortality comparing the Western region to other regions ranged from 0.55 to 0.60. In patients ≥65 years and patients <65 years, the mortality rates were lower in the Western region in both Caucasians and African

47


BLUK077-Stein

April 6, 2007

15:8

48

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 10.1 Rate of diagnosis of deep venous thrombosis, pulmonary embolism, and venous thromboembolism according to region. Year

West

Midwest

South

Northeast

Deep venous thrombosis/100,000/yr 1979–1989

88

116

101

114

1990–2001

78

131

120

138

1979–2001

83

124

112

127

Pulmonary embolism/100,000/yr 1979–1989

48

59

51

55

1990–2001

34

46

41

50

1979–2001

40

52

46

53

Venous thromboembolism/100,000/yr 1979–1989

128

164

145

160

1990–2001

102

163

150

173

1979–2001

113

163

147

167

16 7

VTE DVT

53

40

50

46

100

52

83

112

124

150

127

11 3

200

PE

or

th

ea

st

So ut h N

M

id

w

es

es t

t

0 W

DX/100,000/yr

14 7

16 3

Deep venous thrombosis: 1979–1989: West < Northeast, Midwest (P < 0.001); West < South (P < 0.01) 1990–2001: West < Northeast, Midwest, and South (all P < 0.001) 1979–2001: West < Northeast, Midwest, and South (all P < 0.001) 1979–1989: South < Northeast, Midwest (P < 0.01) 1990–2001: South < Northeast (P < 0.05) 1979–2001: South < Northeast, Midwest (P < 0.01). Pulmonary embolism: 1990–2001: West < Northeast, Midwest (P < 0.001); West < South (P < 0.05) 1979–2001: West < Northeast, Midwest (P < 0.001) 1990–2001: South < Northeast (P < 0.05). Venous thromboembolism: 1979–1989: West < Northeast, Midwest (P < 0.001); West < South (P < 0.05) 1990–2001: West < Northeast, Midwest, and South (all P < 0.001) 1979–2001: West < Northeast, Midwest, and South (all P < 0.001) 1979–1989: South < Northeast, Midwest (P < 0.05) 1990–2001: South < Northeast (P < 0.01) 1979–2001: South < Northeast, Midwest (P < 0.01). Comparisons with probabilities of P > 0.05 were excluded. Reprinted from Stein et al. [1], with permission from Elsevier.

Figure 10.1 Rates of diagnosis (Dx)/100,000 population/year of pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) according to region from 1979 to 2001. (Reprinted from Stein et al. [1], with permission from Elsevier.)


BLUK077-Stein

April 6, 2007

CHAPTER 10

15:8

Americans (P < 0.01 to P < 0.001). The mortality rate in women was higher than in men within all regions (P = 0.04 to P = 0.004). The rate ratios of mortality of women to men within each region ranged from 1.09 to 1.16. Death from PE was obtained from the United States Bureau of the Census records. The Western region, from 1990 to 2001, had a higher percentage of Asian Americans-Pacific Islanders (8.2%) than the Northeastern (3.2%), Midwestern (1.6%), or Southern (1.7%) regions [1]. Racial categories for Asian-Pacific Islanders did not exist in census data before 1990 [5]. Data were insufficient to make regional comparisons with Asian-Pacific Islanders. Men ≥65 years constituted 9.5% of the population in the Western region and 10.6–11.5% in other regions. Women ≥65 years constituted 12.1% of the population in the Western region and 13.8–15.4% in other regions. The rate ratio of women to men aged ≥65 years (1.3) was the same in all regions.

Table 10.2 Rate ratios of rates of diagnosis of deep venous thrombosis, pulmonary embolism, and venous thromboembolism according to region. Year

49

Regional differences of VTE in the United States

West/Northeast

West/Midwest

West/South

Deep venous thrombosis/100,000/yr 1979–1989

0.77

0.76

0.87

1990–2001

0.57

0.60

0.65

1979–2001

0.65

0.66

0.74

Pulmonary embolism/100,000/yr 1979–1989

0.87

0.82

0.94

1990–2001

0.67

0.74

0.82

1979–2001

0.76

0.77

0.87

Venous thromboembolism/100,000/yr 1979–1989

0.80

0.78

0.88

1990–2001

0.59

0.63

0.68

1979–2001

0.68

0.69

0.77

Reprinted from Stein et al. [1], with permission from Elsevier.

Table 10.3 Rates of diagnosis of deep venous thrombosis, pulmonary embolism, and venous thromboembolism according to region and race (1979–2001). Race

West

Midwest

South

Northeast

Deep venous thrombosis/100,000/yr White patients

68

95

109

122

Black patients

75

113

96

96

Pulmonary embolism/100,000/yr White patients

32

39

44

52

Black patients

38

50

39

40

Venous thromboembolism/100,000/yr White patients

92

125

143

162

Black patients

106

150

128

126

Deep venous thrombosis: Black patients: West < Midwest (P < 0.001) White patients: West < Midwest, South, Northeast (P < 0.001); Midwest < Northeast (P < 0.001); Midwest < South (P < 0.01); South < Northeast (P < 0.01). Pulmonary embolism: Black patients: West < Midwest (P < 0.05) White patients: West < Northeast (P < 0.001); West < South (P < 0.01); Midwest < Northeast (P < 0.01); South < Northeast (P < 0.05). Venous thromboembolism: Black patients: West < Midwest (P < 0.001); South, Northeast < Midwest (P < 0.05) White patients: West < Midwest, South, Northeast (P < 0.001); Midwest < Northeast (P < 0.001); Midwest < South (P < 0.01); South < Northeast (P < 0.01). Comparisons with probabilities of P > 0.05 were excluded. Reprinted from Stein et al. [1], with permission from Elsevier.


BLUK077-Stein

April 6, 2007

15:8

50

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 10.4 Rates of diagnosis of deep venous thrombosis, pulmonary embolism, and venous thromboembolism according to region and sex (1979–2001). Sex

West

Midwest

South

Northeast

Deep venous thrombosis/100,000/yr Men

76

107

90

114

Women

90

143

134

141

Pulmonary embolism/100,000/yr Men

37

48

41

48

Women

43

56

50

57

Venous thromboembolism/100,000/yr Men

104

143

121

148

Women

123

185

174

185

Deep venous thrombosis: Men: West < Midwest, Northeast (P < 0.001); West < South (P < 0.05); South < Northeast (P < 0.001); South < Midwest (P < 0.01) Women: West < Midwest, South, Northeast (P < 0.001). Pulmonary embolism: Men: West < Midwest, Northeast (P < 0.05) Women: West < Midwest, Northeast (P < 0.001); West < South (P < 0.05). Venous thromboembolism: Men: West < Midwest, Northeast (P < 0.001); West < South (P < 0.01); South < Northeast, Midwest (P < 0.001) Women: West < Midwest, South, Northeast (P < 0.001). Comparisons with probabilities of P > 0.05 were excluded. Reprinted from Stein et al. [1], with permission from Elsevier.

PE deaths/100,000/yr

Kniffin and associates, in a population of patients ≥65 years of age, showed lower age adjusted rates of diagnosis of PE in women than men and a tendency toward higher rates of DVT in women than men [4]. We showed comparable rates of diagnosis of PE in men and women ≥65 years of age in each of the regions and comparable rates of DVT in the Western and

10 8 6 4

4.2

4.5

4.1

Midwest

South

Northeast

2.5

2 0 West

Figure 10.2 Mortality rates from PE (deaths from PE/100,000 population/year) from 1979 to 1998. The mortality rate in the Western region was lower than the mortality rate in all other regions (P < 0.001). (Reprinted from Stein et al. [1], with permission from Elsevier.)

Northeastern regions, but higher rates of DVT in elderly women than elderly men in the Midwestern and Southern regions [1]. Throughout the United States for patients of all ages, the rate of diagnosis of DVT was higher in women than men [6], but in elderly patients the rates of DVT and PE were comparable in men and women [7]. A somewhat younger population would have contributed to the lower rates of DVT and VTE and the lower mortality rate in the Western region [7]. However, such lower rates were observed in patients ≥65 years as well. A higher percentage of Asian Americans and/or Pacific Islanders in the Western region than in other regions would also have contributed to the lower rates of diagnosis and lower mortality rate in the Western region, because the incidences of PE and of DVT are lower in Asian Americans than in African Americans or Caucasians [8–10]. However, lower rates of diagnosis of DVT, PE, and VTE were shown in Caucasians in the Western region and lower mortality rates from PE were shown in Caucasians and African Americans in the Western region. The observed difference


BLUK077-Stein

April 6, 2007

CHAPTER 10

15:8

Regional differences of VTE in the United States

in regional rates of diagnosis of DVT and VTE are unlikely to be related to differences in climate. We observed no seasonal variation in the rate of diagnosis of DVT, PE, or VTE in any of the regions, including the Southern region, where winters are mild, and the Northeastern and Midwestern regions, where seasons are sharply defined [5] (see Chapter 9). Lilienfeld and Godbold, based on data from 1980 to 1984, showed lower mortality rates from PE in the Pacific and Mountain regions than other parts of the country [11].

References 1 Stein PD, Kayali F, Olson RE. Regional differences in rates of diagnosis and mortality of pulmonary thomboembolism. Am J Cardiol 2004; 93: 1194–1197. 2 National Hospital Discharge Survey 1979–2001 Multiyear Public-use data file documentation. US Department of Health and Human Services, Public Health Service, National Center for Health Statistics. http://www.cdc.gov/ nchs/about/major/hdasd/nhds.htm. 3 Bureau of the Census, Department Of Commerce, United States Department of Health and Human Services (US DHHS) Centers for Disease Control and Prevention (CDC), CDC WONDER On-line Database. http:// wonder.cdc.gov/census.shtml.

51

4 Kniffin WD, Jr, Baron JA, Barrett J, Birkmeyer JD, Anderson FA, Jr. The epidemiology of diagnosed pulmonary embolism and deep venous thrombosis in the elderly. Arch Intern Med 1994; 154: 861–866. 5 Stein PD, Kayali F, Olson RE. Analysis of venous thromboembolic disease in the four seasons. Am J Cardiol 2004; 93: 511–513. 6 Stein PD, Hull RD, Patel KC et al. Venous thromboembolic disease: comparison of the diagnostic process in men and women. Arch Intern Med 2003; 163: 1689–1694. 7 Stein PD, Hull RD, Kayali F, Ghali WA, Alshab AK, Olson RE. Venous thromboembolism according to age: the impact of an aging population. Arch Intern Med 2004; 164: 2260–2265. 8 Klatsky AL, Armstrong MA, Poggi J. Risk of pulmonary embolism and/or deep venous thrombosis in AsianAmericans. Am J Cardiol 2000; 85: 1334–1337. 9 White RH, Zhou H, Romano PS. Incidence of idiopathic deep venous thrombosis and secondary thromboembolism among ethnic groups in California. Ann Intern Med 1998; 128: 737–740. 10 Stein PD, Kayali F, Olson RE, Milford, CE. Pulmonary thromboembolism in Asian-Pacific Islanders in the United States: analysis of data from the National Hospital Discharge Survey and the United States Bureau of the Census. Am J Med 2004; 116: 435–442. 11 Lilienfeld DE, Godbold JH. Geographic distribution of pulmonary embolism mortality rates in the United States, 1980 to 1984. Am Heart J 1992; 124: 1068–1072.


BLUK077-Stein

March 28, 2007

11

15:8

CHAPTER 11

Venous thromboembolism in the elderly patients than in younger patients (20–69 years). The 21-year trends for the diagnosis of DVT according to age are shown in Figure 11.1b [1]. The elderly population showed the greatest increase in the 1990s. The diagnosis of pulmonary embolism (PE) in patients 70 years or older was 6.2 than the rate in younger patients (Figure 11.2a) [1]. Contrary to DVT, the rate of diagnosis of PE decreased from 370 PE/100,000 population in 1979 to 254 PE/100,000 population in 1990 (Figure 11.2a) and then remained constant.

Rates of diagnosis and trends in the diagnosis of deep venous thrombosis and pulmonary embolism in the elderly Deep venous thrombosis (DVT) in elderly patients (70 years or older) increased 44% from 454 DVT/100,000 population in 1990 to 655 DVT/100,000 population in 1999 (Figure 11.1a) [1]. Deep venous thrombosis was diagnosed 4.7 times more frequently in elderly (a) DVT DVT/100,000 population

800

Age >70

600 400 Age 20−69

200 0

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Year

(b) DVT: age distribution

DVT/100,000 population

1000 800

70−89

600 400

60−69 50−59 40−49 20−39

200 0 1999

1997

1995

1993

52

1991

1989

1987

1985

1983

1981

1979

Year

Figure 11.1 (a) Among elderly patients (70 years or older), trends over 21 years in the rate of diagnosis of deep venous thrombosis (DVT) was constant from 1979 to 1990 and increased from 1990 to 1999. In younger patients (20–69 years), there was a slight but significant decline in the rate of diagnosis of DVT between 1979 and 1990. The rate then increased somewhat between 1990 and 1999. (b) From 1979 to 1990, the rate of diagnosis of DVT was constant in patients aged 60–69 and 70–89 years. During this time interval, the rate decreased in younger age groups. From 1990 to 1999, the rate of diagnosis of DVT increased in all age groups. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)


BLUK077-Stein

March 28, 2007

CHAPTER 11

15:8

53

Venous thromboembolism in the elderly (a) PE

200 100

Age 20−69

0 1999

1997

1995

1993

1991

1989

1987

1985

1983

PE/100,000 population

500

PE: age distribution

400 300

70−89

200 60−69 50−59 40−49 20−39

100 0 1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

Year

DVT/100,000 population/yr

Smaller changes were observed in patients aged 20– 69 years from 67 PE/100,000 population in 1979 to 30 PE/100,000 population in 1990 (Figure 11.2b). The rate increased somewhat between 1990 and 1999. The rates of diagnosis of DVT or PE in elderly men and women, and elderly black and white patients were comparable [1].

Figure 11.3 Deep venous thrombosis (DVT)/100,000 population/year, diagnosed at hospital discharge, shown according to age for the year 1999. (Data from Stein et al. [1, 2].)

1981

Year

(b)

1979

Figure 11.2 (a) Among elderly patients (70 years or older), trends over 21 years in the rate of diagnosis of pulmonary embolism (PE) decreased from 1979 to 1990 and then remained constant from 1990 to 1999. In younger patients there was a slight but significant decline in the rate of diagnosis of PE between 1979 and 1990. The rate then increased somewhat between 1990 and 1999. (b) Trends over 21 years in the rate of diagnosis of PE in patients as shown by age group. In all age groups, the rate of diagnosis of PE significantly decreased from 1979 to 1990. From 1990 to 1999, the rate of diagnosis of PE remained constant in all age groups except age 20–39, which showed a slight increase. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)

Age >70

300

1979

PE/100,000 population

400

Deep venous thrombosis, based on hospital discharges, was diagnosed in 700 patients/100,000 population/year aged 70–89 years, 300/100,000 population/year aged 60–69 years, and lower proportions of the population of younger people [1, 2] (Figure 11.3). Comparing the rate of DVT at each decade of age with the rate at age 20–29, the rate ratio increased 700

700 600 500 400

300

300 200

200 100 0

<5

10

30

60

100

0−14 15−19 20−29 30−39 40−49 50−59 60−69 70−89

Age groups (years)


BLUK077-Stein

March 28, 2007

15:8

54

PART I

40

DVT 1980−1999

Prevalence, risks, and prognosis of PE and DVT

PE 1980−1999

40

.7

11

1

Rate ratio

Rate ratio

.7

20

1

6.

2.

6

10

7 7

90−99

80−89

70−79

60−69

250

50−59

300

300

40−49

exponentially up to age 89 (Figure 11.4) [1]. In patients aged 70–79 years and 80–89 years, the rate ratios for DVT were 12.7 and 17.7, respectively. Pulmonary embolism, based on hospital discharges, was diagnosed in 300 patients/100,000 population/year aged 70–89 years, 100/100,000 population/year aged 60–69 years, and lower proportions of the population of younger people [1, 2] (Figure 11.5). Comparing the rate of PE at each decade of age with the rate at age 20–29, the rate ratio increased exponentially up to age 89 (Figure 11.6) [1]. In patients aged 70–79 years and 80–89 years, the rate ratios for

30−39

Figure 11.4 Rate ratios for the rate of diagnosis of deep venous thrombosis (DVT), comparing each decade of age with the rate at age 20–29 years. Rate ratios were averaged over 21 years. Between ages 20–29 and 80–89 years, the rate ratios for the diagnosis of DVT increased exponentially. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)

0 20−29

90−99

80−89

70−79

60−69

50−59

40−49

30−39

20−29

Age groups (years)

PE/100,000 population/yr

1

1.

7

3.

1. 0

Age groups (years) Figure 11.6 Rate ratios for the rate of diagnosis of pulmonary embolism (PE), comparing each decade of age with the rate at age 20–29 years. Rate ratios were averaged over 21 years. Between ages 20–29 and 80–89 years, the rate ratios for the diagnosis of PE increased exponentially. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)

PE were 20.6 and 27.9, respectively. There was no step change of rate of diagnosis of PE or DVT at any age. The recommended approach to the diagnosis of PE is applicable to elderly patients [3]. An abundance of literature documents that the risk of venous thromboembolism increases with age [1, 4–19].

Case fatality rate From 1989 to 1998, the estimated case fatality rate from PE in the United States was 7.7 PE deaths/100 cases of PE [20]. The estimated case fatality rate from PE is strongly age-dependent, and increased exponentially with age from 3.6% in patients aged 25–34 years to 17.4% in patients aged >85 years [20] (see Chapter 3).

Antithrombotic prophylaxis and age

200 150

100

100

80

50 0

.3

.9 .6

12 8. 4.

10

20

.7

20

27

.9

17

17

30

27

30

<1 0−14

4

10

20

20

15−19 20−29 30−39 40−49 50−59 60−69 70−89

Age groups (years)

Figure 11.5 Pulmonary embolism (PE)/100,000 population/ year, diagnosed at hospital discharge, shown according to age for the year 1999. (Data from Stein et al. [1, 2].)

Recommendations for antithrombotic prophylaxis in patients undergoing surgical procedures are partially based on the age of the patient [21]. Patients less than 40 years have been considered at low-risk for venous thromboembolism for specific in-hospital surgical groups, such as general surgical patients, and age as a risk factor has become more important at 40 years or older with regard to thromboprophylaxis [21]. This led to the general belief that age as a risk factor has a


BLUK077-Stein

March 28, 2007

15:8

Venous thromboembolism in the elderly

55

break point at the age of 40 years. For patients aged 30–39 years, there is almost a 2-fold increase in the risk of DVT or PE compared with younger patients [1]. The concept that the elderly are at the greatest need for thromboprophylaxis is emphasized by these data which show an 18–28-fold increase in the risk of DVT or PE for patients ≥70 years compared with those 20–29 years of age [1].

the PE in 44% [25]. Comparable percentages of patients in the younger age groups were immobilized or underwent surgery before the PE. Malignancy was more frequent among patients ≥70 years (26%) than among the patients <40 years old (2%), but patients 40–69 years had malignancy nearly as frequently as did patients ≥70 years old (24%). Estrogen use was infrequent among female patients ≥70 years old. Its use among female patients <40 years of age preceded PE in 35% and childbirth in women <40 years of age preceded PE in 25% [25].

CHAPTER 11

Diagnosis of acute PE The diagnosis of PE among elderly patients has been thought to be particularly difficult because the expected signs and symptoms may be absent or ignored [22–24]. This did not seem to be the case in the experience of the Prospective Investigation of the Pulmonary Embolism Diagnosis (PIOPED I) [25] and, in general, not in the experience of PIOPED II [26]. In PIOPED II, however, dyspnea and tachypnea were less frequent in patients ≥70 years old than in patients <40 years old [26]. The typical signs and symptoms known to occur among younger patients were common among elderly patients [25, 26]. In the absence of these signs and symptoms, unexplained radiographic abnormalities were important diagnostic clues [25]. When the diagnosis of PE is uncertain, computed tomographic (CT) angiography can be performed safely in elderly patients, providing renal function is adequate [3]. Renal failure was a problem among elderly patients who underwent conventional angiography [25].

Predisposing factors according to age Among 72 patients ≥70 years old in PIOPED I, 67% were immobilized before the PE, and surgery preceded

Syndromes of PE according to age The usual syndromes of PE, among all patients, irrespective of prior cardiopulmonary disease, characterized by (1) hemoptysis or pleuritic pain, (2) isolated dyspnea, or (3) circulatory collapse were observed in PIOPED I among elderly patients [25]. However, 11% of patients ≥70 years of age in PIOPED I and 15% in PIOPED II did not show these syndromes [25, 26]. In PIOPED I they were identified on the basis of unexpected radiographic abnormalities, which may have been accompanied by tachypnea or a history of thrombophlebitis [25]. Unexplained radiographic abnormalities may be an important clue to the diagnosis of PE, particularly among elderly patients in whom the expected signs and symptoms are absent, as has been previously observed [22]. Among all patients with PE in PIOPED II, the syndrome of hemoptysis or pleuritic pain was less frequent in patients ≥70 years than in patients <40 years (P < 0.025) and isolated dyspnea was more frequent (P < 0.05) (Table 11.1). In PIOPED II, among patients with no prior cardiopulmonary disease, the prevalence of the various syndromes was similar among all age groups (Table 11.2).

Table 11.1 Syndromes of acute PE according to age: all patients with pulmonary embolism: PIOPED II. Syndromes Hypotension, LOC

≥70 yr (n = 55) [n (%)]

40–69 yr (n = 106) [n (%)]

<40 yr (n = 31) [n (%)]

2 (4)

10 (9)

3 (10)

Hemoptysis or pleuritic pain

15 (27)

43 (41)

16 (52)

Isolated dyspnea

30 (55)

39 (37)

10 (32)

8 (15)

14 (13)

2 (6)

No syndrome

The syndrome of hemoptysis or pleuritic pain was less frequent in patients ≥70 years than in patients <40 years (P <0.025) and isolated dyspnea was more frequent (P <0.05). PE, pulmonary embolism; LOC, loss of consciousness. Data from Stein et al. [26].


BLUK077-Stein

March 28, 2007

15:8

56

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 11.2 Syndromes of acute PE according to age: patients with no prior cardiopulmonary disease: PIOPED II. ≥70 yr (n = 35) [n (%)]

Syndromes Hypotension, LOC

40–69 yr (n = 76) [n (%)]

<40 yr (n = 22) [n (%)]

2 (6)

7 (9)

2 (9)

Hemoptysis or pleuritic pain

11 (31)

30 (39)

10 (45)

Isolated dyspnea

17 (49)

29 (38)

8 (36)

6 (17)

11 (14)

2 (9)

No syndrome

Differences comparing age groups not significant. PE, pulmonary embolism; LOC, loss of consciousness. Data from Stein et al. [26].

Symptoms according to age Dyspnea was the most frequent symptom in all patients with PE, occurring in 78 and 75% of ≥70 years old in PIOPED I and PIOPED II [25, 26] (Tables 11.3) and in 66% of patients ≥70 years old with no prior cardiopulmonary disease (Table 11.4). Pleuritic pain occurred in 51 and 33% of all patients with PE in PIOPED I and PIOPED II. Pleuritic pain occurred more frequently than hemoptysis in all age groups. In gen-

eral, all symptoms occurred with equal frequency in the different age groups, but there were occasional exceptions, as noted in Tables 11.3–11.4

Signs according to age Tachypnea (respiratory rate ≥20/min) was the most frequent sign in all age groups with PE [25, 26] (Table 11.5) and in those with no prior cardiopulmonary disease (Table 11.6). All signs, occurred with

Table 11.3 Symptoms in all patients with acute pulmonary embolism according to age: PIOPED I and PIOPED II. ≥70 yr

Dyspnea

<40 yr

40–69 yr

PIOPED I

PIOPED II

PIOPED I

PIOPED II

PIOPED I

(n = 72)

(n = 53–55)

(n = 144)

(n = 100–106)

(n = 44)

PIOPED II (n = 30–31)

(%)

(%)

(%)

(%)

(%)

(%) 87

78

75

78

79

82

Dyspnea (rest or exertion)

75

79

87

Dyspnea (at rest)

60

59

71

Dyspnea (exertion only)

13

19

13

Orthopnea (>2-pillow)

31

40

32

51

33

58

53

70

53

Cough

35

44

42

46

45

35

Purulent

11

10

6

Clear

9

11

6

Nonproductive

20

21

23

Leg swelling

35

26

33

50

14

33

Leg pain

31

28

26

45

20

48

Palpitation

13

15

9

Wheezing

10

25

12

32

16

35

Angina-like pain

10

13

13

18

7

23

8

4

4

6

32*

10

Pleuritic pain

Hemoptysis

*P <0.01, ≥70 years vs. <40 years; P <0.001, 40–69 years vs. <40 years among patients in PIOPED I. Data from Stein et al. [25, 26].


BLUK077-Stein

March 28, 2007

15:8

Table 11.4 Symptoms in patients with PE and no preexisting cardiac or pulmonary disease according to age: PIOPED II.

Syndromes

≥70 yr

40–69 yr

<40 yr

(n = 30–35)

(n = 72–76)

(n = 21–22)

(%)

(%)

(%)

Dyspnea Dyspnea (rest or exertion)

66

75

82

Dyspnea (at rest)

49

55

68

Dyspnea (exertion only)

14

19

9

Orthopnea (≥2-pillow)

23

33

18

Pleuritic pain

35

46

45

Chest pain (not pleuritic)

18

20

23

Cough

29

36

36

Hemoptysis

3

5

9

Purulent

6

5

5

Clear Nonproductive Wheezing

6

5

5

11

21

27

9

24

27

Calf swelling

26

49*

24

Thigh swelling

11

5

9

0

0

5

Calf pain

26

47

41

Thigh pain

11

24

19

0

3

5

Thigh swelling, no calf swelling

Thigh pain, no calf pain

*P = 0.025, age ≥70 years vs. age 40–69 years, P = 0.048, age 40–69 years vs. age <40 years. All other differences between age groups not significant. Data from Stein et al. [26].

Table 11.5 Signs in all patients with acute pulmonary embolism according to age: PIOPED I and PIOPED II. ≥70 yr

<40 yr

40–69 yr

PIOPED I

PIOPED II

PIOPED I

PIOPED II

PIOPED I

PIOPED II

(n = 72)

(n = 52–55)

(n = 144)

(n = 101–106)

(n = 44)

(n = 29–31)

(%)

(%)

(%)

(%)

(%)

(%)

Tachypnea (>20/min)

74

51

69

58

82

60

Rales

65

26

61

17

41

26

Tachycardia (>100/min)

29

21

26

24

32

43

Increased P2

15

7

20

19

34

18

Deep venous thrombosis

15

47

17

49

9

42

Diaphoresis

8

2

10

7

18

0

Wheezes

8

4

10

2

5

6

Temperature >38.5◦ C

7

0

5

2

14

3

Third heart sound

7

6

5

Right ventricular lift

4

7

0

Jugular venous distention

19

12

10

Rhonchi

6

6

0

Decreased breath sounds

29

19

13

Pleural friction rub

6

2

5

1

0

0

Homans’ sign

4

2

2

Cyanosis

3

0

3

1

2

0

Differences between age groups in both PIOPED I and PIOPED II were not significant. Data from Stein et al. [25, 26].

57


BLUK077-Stein

March 28, 2007

15:8

58

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 11.6 Signs in patients with PE and no preexisting cardiac or pulmonary disease according to age: PIOPED II. ≥70 yr

40–69 yr

<40 yr

(n = 30–35)

(n = 71–76)

(n = 20–22)

(%)

(%)

(%)

Tachypnea (≥20/min)

47

59

52

Tachycardia (>100/min)

23

21

38

Diaphoresis

3

3

0

Cyanosis

0

0

0

Signs General

Temperature >38.5◦ C (>101.3◦ F)

0

3

5

Cardiac examination (any)

23

22

14

Increased P2*

12

15

19

0

6

0

Jugular venous distension

20

14

5

Lung examination (any)

43

22

27

Rales (crackles)

Right ventricular lift†

26

14

19

Wheezes

3

1

0

Rhonchi

0

3

0

23

14

18

0

0

0

Decreased breath sounds Pleural friction rub DVT Calf or thigh

51

48

36

Calf only

34

31

32

Calf and thigh

14

16

5

3

1

0

Thigh only

All differences comparing age groups not significant. * Data in 26 patients ≥70 years, 61 patients 40–69 years, 16 patients ≤40 years. † Data in 30 patients ≥70 years, 64 patients 40–69 years, 16 patients <40 years. P2, pulmonary component of second heart sound; DVT, deep venous thrombosis. Data from Stein et al. [26].

similar frequency among all age groups (Tables 11.5 and 11.6).

Chest radiograph The chest radiograph among all patients with PE in PIOPED I was normal in 4% ≥70 years old [25]. Atelectasis or pulmonary parenchymal abnormalities were the most frequent radiographic abnormalities among all age groups. All radiographic abnormalities occurred with a comparable frequency among all age groups (Table 11. 7).

Combinations of symptoms and signs and radiographic abnormalities Even among patients ≥70 years old, a combination of nonspecific symptoms and signs that typically occur

with PE was present in the great majority of patients with PE [25, 26]. Dyspnea or tachypnea among all patients with PE occurred in 92% of patients ≥70 years of age in PIOPED I, but in only 77% in PIOPED II (Table 11.8) and in 70% with no prior cardiopulmonary disease in PIOPED II (Table 11.9). Among all patients with PE, dyspnea or tachypnea or pleuritic pain occurred in 94% of patients ≥70 years old in PIOPED I and in 87% in PIOPED II [25, 26]. If signs of DVT were added, 94% of patients ≥70 years in PIOPED I and 96% in PIOPED II had 1 or more of these findings (Tables 11.8). Combinations of signs and symptoms occurred with similar frequency in all age groups, except dyspnea or tachypnea among patients with PE and no prior cardiopulmonary disease were less frequent in patients 70 years than in patients <40 years [26] (Tables 11.8 and 11.9).


BLUK077-Stein

March 28, 2007

CHAPTER 11

15:8

59

Venous thromboembolism in the elderly

Table 11.7 Chest radiograph in all patients with acute pulmonary embolism according to age: PIOPED I. ≥70 yr

40–69 yr

<40 yr

(n = 72) (%)

(n = 144) (%)

(n = 44) (%)

4

8

Atelectasis or pulmonary parenchymal abnormality

Normal

71

69

14 64

Pleural effusion

57

46

45

Pleural-based opacity

42

34

43

Prominent central pulmonary artery

29

20

11

Elevated diaphragm

28

27

18

Cardiomegaly

22

17

14

Decreased pulmonary vascularity

19

22

20

Pulmonary edema

13

12

7

Westermark’s sign

7

8

0

Differences among age groups were not significant. Westermark’s sign = prominent central pulmonary artery and decreased pulmonary vascularity. Reprinted from Stein et al. [25], with permission from the American College of Cardiology Foundation.

The electrocardiogram according to age Nonspecific ST segment or T wave changes were the most frequent electrocardiographic abnormalities in all patients with PE, either or both occurring in 56% of patients ≥70 years old and with nearly the same

frequency in younger patients [25]. With the exception of left anterior hemiblock (left axis deviation) among patients ≥70 years of age, other electrocardiographic abnormalities occurred in 12% or fewer patients in all age groups in PIOPED I. No differences in the frequency of occurrence of any ECG abnormalities

Table 11.8 Combinations of signs and symptoms in all patients with acute pulmonary embolism according to age: PIOPED I and PIOPED II. ≥70 yr

<40 yr

40–69 yr

PIOPED I

PIOPED II

PIOPED I

PIOPED II

PIOPED I

PIOPED II

(n = 72)

(n = 52)

(n = 144)

(n = 104)

(n = 44)

(n = 30)

(%)

(%)

(%)

(%)

(%)

(%)

Dyspnea or tachypnea

92

77

90

88

95

Dyspnea or tachypnea or hemoptysis

92

91

98

93 —

Dyspnea or tachypnea or pleuritic pain*

94

87

98

93

100

Dyspnea or tachypnea or signs of deep

92

91

98

97

94

96

99

96

100

100

97

98

100

99

100

venous thrombosis Dyspnea or tachypnea or pleuritic pain or

100

signs of deep venous thrombosis* Dyspnea or tachypnea or radiographic atelectasis or parenchymal abnormality Dyspnea or tachypnea or pleuritic pain or radiographic atelectasis or parenchymal abnormality* *The addition of hemoptysis did not improve the sensitivity of the combination for the detection of pulmonary embolism. Tachypnea = respiratory rate ≥20/min. Differences among age groups in PIOPED I and PIOPED II were not significant. Data from Stein et al. [25, 26].


BLUK077-Stein

March 28, 2007

15:8

60

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 11.9 Combinations of signs and symptoms in patients with acute pulmonary embolism and no prior cardiopulmonary disease according to age: PIOPED II. ≥70 yr (n = 33) (%)

40–69 yr (n = 75) (%)

<40 yr (n = 21) (%)

Dyspnea or tachypnea (≥20/min)

70*

88

90

Dyspnea or tachypnea (≥20/min) or

85

93

95

97

97

100

pleuritic pain Dyspnea or tachypnea (≥20/min) or pleuritic pain or signs of DVT Tachypnea = respiratory rate ≥20/min. P = 0.025, age ≥70 years vs. age <40 years. Other differences among age groups were not significant. Data from Stein et al. [26].

were apparent between patients ≥70 years of age and younger patients, although incomplete right bundle branch block was less frequent among patients 40–69 years than among patients less than 40 years of age (Table 11.10).

Blood gases according to age The partial pressure of oxygen in arterial blood (PaO2 ) was lower among patients ≥70 years than among those <40 years of age based on data in PIOPED I among all patients with PE [25]. Among patients 40–69 years, the PaO2 was lower than in patients <40 years, but not significantly lower than in patients ≥70 years of age [25]. The PaO2 among patients with PE ≥70 years of age, 40–69 years of age, and <40 years of age was 61 ± 12,

67 ± 15, and 75 ± 18 mm Hg, respectively (mean ± standard deviation). In PIOPED II, the PaO2 in all patients with PE who were 70 years old was 91 ± 72 mm Hg and did not differ significantly from values in patients 40–69 years (83 ± 31 mm Hg) (unpublished data from [26]). In those with PE and no prior cardiopulmonary disease, the PaO2 in patients 70 years old was 88 ± 45 mm Hg and it also did not differ significantly from values in patients 40–69 years (84 ± 27 mm Hg) (unpublished data from [26]). The alveolar–arterial oxygen difference (gradient) among patients with PE ≥70 years of age was 47 ± 14 mm Hg, which was higher than among patients 40–69 years old (40 ± 17 mm Hg) and it was higher than in patients <40 years old (31 ± 17 mm Hg). The alveolar–arterial oxygen difference in normal adults increases with age [27–30].

Table 11.10 Electrocardiographic findings in all patients with acute pulmonary embolism according to age: PIOPED I. ≥70 yr (n = 72) (%)

40–69 yr (n = 113) (%)

<40 yr (n = 36) (%)

Normal

21

27

22

ST segment or T wave changes

56

51

56

Left axis deviation

18

11

8

Left ventricular hypertrophy

12

7

11

Acute myocardial infarction pattern

12

4

6

Low voltage QRS

9

5

0

Complete right bundle branch block

7

4

3

Right ventricular hypertrophy

4

3

3

Right axis deviation

2

3

8

P-pulmonale

2

2

0

Incomplete right bundle branch block

2

0

11*

*P < 0.01, age 40–69 years vs. age <40 years. Reprinted from Stein et al. [25], with permission from the American College of Cardiology Foundation.


BLUK077-Stein

March 28, 2007

CHAPTER 11

15:8

61

Venous thromboembolism in the elderly

Clinical assessment according to age The accuracy of clinical assessment was comparable among patients in all age groups [25]. When physicians were 80–100% confident that PE was present on the basis of clinical judgment and laboratory tests with the exception of ventilation–perfusion scans, they were correct in 90% of 10 patients ≥70 years old. When they believed that there was less than 20% likelihood of PE, it was present in 19% of 69 patients. In most patients physicians were uncertain of the diagnosis, believing that there was a 20–79% chance of PE.

Ventilation–perfusion lung scan according to age Among patients ≥70 years of age the positive predictive value of a high-probability interpretation of the ventilation–perfusion lung scan (94%) was higher than in patients <40 years of age (69%) [25] (Table 11.11). The positive predictive value of other interpretations was comparable in all age groups. The sensitivity of ventilation–perfusion lung scans indicating a high probability of PE among patients ≥70 years of age (47%) did not differ significantly from the sensitivity of such scans among younger age groups [25].

CT angiography according to age Among 773 patients with an adequate CT pulmonary angiography and 737 patients with an adequate CT pulmonary angiography/CT venography, the sensitivity and specificity for PE for age groups 18–59, 60–79, and 80–99 years did not differ to a statistically significant extent [31]. Multidetector CT pulmonary an-

giography and CT pulmonary angiography/CT venography may be used with various diagnostic strategies in adults of all ages.

Acute hemorrhage/infarction syndrome in the elderly The syndrome of hemoptysis or pleuritic pain, in addition to signs and symptoms, was investigated in detail in elderly patients [32]. The electrocardiogram was normal in 62% of such patients [32]. If abnormal, the most frequent abnormalities were nonspecific ST segment or T wave changes (38%). The chest radiograph showed atelectasis or a pulmonary parenchymal abnormality in 82% of elderly patients with the hemoptysis/pleuritic pain syndrome. The central pulmonary artery dilated in 29% of such elderly patients. A normal chest radiograph was uncommon, occurring in only 6% of elderly patients. The ventilation–perfusion lung scan was interpreted as high probability for PE in 41% of elderly patients with the hemoptysis/pleuritic pain syndrome. Elderly patients with the hemoptysis/pleuritic pain syndrome had a higher pulmonary artery mean pressure (25 ± 9 versus 17 ± 7 mm Hg) and lower PaO2 (64 ± 10 versus 81 ± 14 mm Hg) than patients <40 years of age, and elderly patients tended to have more mismatched segmental perfusion defects on the ventilation–perfusion lung scan than patients <40 years of age [32].

Use of diagnostic tests in the elderly Diagnostic approaches to DVT and PE have changed markedly over the past two decades in temporal

Table 11.11 Results of ventilation–perfusion lung scans in all patients with acute pulmonary embolism according to age: PIOPED I. ≥70 yr V–Q scan probability High Intermediate Low Near normal/normal

PE/n

≤40 yr

40–69 yr (%)

PE/n

(%)

PE/n

(%)

34/36

(94)*

60/68

(88)

11/16

(69)

27/100

(27)

59/199

(30)

22/52

(42)

10/71

(14)

24/172

(14)

8/57

(14)

1/8

(13)

1/55

(2)

3/68

(4)

*P ≤ 0.025 ≤70 years vs ≤40, all other differences among age groups were not significant. n, number of patients with the scan result shown in column; PE, pulmonary embolism. Reprinted from Stein et al. [25], with permission from the American College of Cardiology Foundation.


BLUK077-Stein

March 28, 2007

15:8

62

Venograms/100,000 population

PART I

Prevalence, risks, and prognosis of PE and DVT

200 150 Age >70

100 Age 20−69

50 0

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Year

harmony with the evolving literature [3, 33]. Changes in the use of diagnostic tests in the elderly parallel the changes in the general population. The use of diagnostic tests in elderly black and white patients was comparable [1]. During the 1980s and early 1990s, the use of contrast venography of the lower extremities was strikingly higher in elderly patients than in younger patients

Figure 11.7 Trends over 21 years in the use of contrast venography of the lower extremities in elderly patients compared with younger patients. The use of contrast venography peaked in 1986 and began to decline sharply in 1989 and thereafter in patients aged 70 or older. The use of contrast venography was higher in elderly patients than in younger patients. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)

(Figure 11.7) [1]. The use of contrast venography sharply declined as the use of ultrasonography increased. Doppler ultrasonography has supplanted ascending contrast venography as the preferred diagnostic approach for DVT [33]. Between 1989 and 1999, the elderly population utilized 5.7 times far more venous ultrasound tests of the lower extremities than the younger population (20–69 years) (Figure 11.8a) [1].

(a) Ultrasound/100,000 population

Ultrasound 300

Age >70 200 100

Age 20−69 0 1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Year

(b) Ultrasound 1989−1999

30

25.3

Rate ratio

20.4

20

14.6 8.4

10 1.4

2.5

30−39

40−49

4.8

0 90−99

80−89

70−79

60−69

50−59

20−29

Age groups (years)

Figure 11.8 (a) The use of Doppler ultrasonography in the elderly markedly increased after 1982 and stabilized in the 1990s. (b) Rate ratios comparing the rate of use of venous ultrasound at age 20–29 to older decades for the interval of 1989 to 1999. Between ages 20–29 and age 80–89, the rate ratios increased exponentially with age. The 95% confidence intervals were too narrow to illustrate. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)


March 28, 2007

63

Venous thromboembolism in the elderly

80−89

400

70−79

300 200

60−69 50−59

100

40−49 20−39

0 1999

1997

1995

1993

1991

1989

1987

1985

Year

427 patients 40–69 years, and in 0.7% of 135 patients <40 years of age. Renal failure, either major or minor, was the most frequent complication of conventional pulmonary angiography among elderly patients [25]. It occurred in 3% of 200 patients ≥70 years of age, compared with 0.7% of 562 patients ≤69 years of age. Among the 10 patients who developed renal failure, 3 required dialysis. “Minor” complications of renal failure were important complications, although dialysis was not required. Patients with these complications showed either an elevation of the serum creatinine from previously normal levels to ≥2.1 mg/100 mL (range 2.1–3.5 mg/100 mL) or an increase in a previously abnormal serum creatinine level ≥2 mg/100 mL.

40 30 Age >70

20

AGE 20−69

10 0

9 −9 97 19 6 −9 94 19 3 −9 91 19 0 −9 88 19 7 −8 85 19 4 −8 82 19 1 −8 79

19

The rate ratio of use of ultrasound, comparing each decade of age with the rate at age 20–29, based on average values from 1989 to 1999, was the highest in the elderly (Figure 11.8b). The use of lung scans over 21 years of observation was highest in elderly patients (Figure 11.9a) [1]. In all age groups, the use of lung scans has decreased since mid-1980s (Figure 11.9a). The utility of ventilation– perfusion lung scans among patients ≥70 years old was comparable with that in younger patients [25]. The positive predictive value of all probabilities of ventilation–perfusion lung scans using original PIOPED criteria [34] were comparable in all age groups (Table 11.11) [25]. However, a higher proportion of patients ≥70 years of age had nondiagnostic (intermediate or low probability) V–Q scans, (80%), than patients ≤40 years of age (56%) (Table 11.11). Among patients ≥70 years of age with ventilation–perfusion lung scans indicating a high probability of PE, 94% had PE (Table 11.11). The rate of use of pulmonary angiograms over a 21-year period of observation was higher in elderly patients than in younger patients (Figure 11.10) [1]. Both for elderly patients and for younger patients the use of pulmonary angiograms increased from 1979 to 1999. Pulmonary angiography was not more hazardous among the elderly, although renal failure was a more frequent sequela among patients ≥70 years of age than among younger patients [25]. Major complications occurred in 1.0% of 200 patients ≥70 years, in 1.2% of

1983

1981

1979

Figure 11.9 Trends over 21 years in the use of radioisotopic lung scans in elderly patients (70 years or older) showed a higher utilization compared with younger patients. The use of lung scans in elderly patients began to increase sharply in 1983, peaked in 1986, and showed a progressive decline in 1987. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)

500

Angios/100,000 population

CHAPTER 11

15:8

Lung scans/100,000 population

BLUK077-Stein

Year Figure 11.10 Trends over 21 years in the use of pulmonary angiograms in elderly patients compared with younger patients. The rate of use of pulmonary angiograms was higher in elderly patients than in younger patients. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)


BLUK077-Stein

March 28, 2007

15:8

64

PART I

References 1 Stein PD, Hull RD, Kayali F, Ghali WA, Alshab AK, Olson RE. Venous thromboembolism according to age: the impact of an aging population. Arch Intern Med 2004; 164: 2260–2265. 2 Stein PD, Kayali F, Olson RE. Incidence of venous thromboembolism in infants and children: data from the National Hospital Discharge Survey. J Pediatr 2004; 145: 563–565. 3 Stein PD, Woodard PK, Weg JG et al. Diagnostic pathways in acute pulmonary embolism: recommendations of the PIOPED II Investigators. Am J Med 2006; 119: 1048–1055. 4 Anderson FA, Jr, Wheeler HB, Goldberg RJ et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT study. Arch Intern Med 1991; 151: 933–938. 5 Silverstein MD, Heit JA, Mohr DN, Petterson TM, O’Fallon WM, Melton LJ, III. Trends in the incidence of deep vein thrombosis and pulmonary embolism. A 25year population based study. Arch Intern Med 1998; 158: 585–593. 6 Gillum RF. Pulmonary embolism and thrombophlebitis in the United States, 1970–1985. Am Heart J 1987; 114: 1262–1264. 7 Giuntini C, Ricco GD, Marini C, Mellilo E, Palla A. Pulmonary embolism: epidemiology. Chest 1995; 107 (suppl): 3S–9S. 8 Nordstrom M, Linblad B, Bergqvist D, Kjellstrom T. A prospective study of the incidence of deep-vein thrombosis within a defined urban population. J Int Med Res 1992; 232: 155–160. 9 Coon WW, Willis PW, III, Keller JB. Venous thromboembolism and other venous disease in the Tecumseh Community Health Study. Circulation 1973; 48: 839–846. 10 Kniffin WD, Jr, Baron JA, Barrett J, Birkmeyer JD, Anderson FA, Jr. The epidemiology of diagnosed pulmonary embolism and deep venous thrombosis in the elderly. Arch Intern Med 1994; 154: 861–866. 11 Coon WW, Coller FA. Some epidemiologic considerations of thromboembolism. Surg Gynecol Obstet 1959; 109: 487–501. 12 Goldhaber SZ, Visani L, De Rosa M. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet 1999; 353: 1386–1389. 13 Ferrari E, Baudouy M, Cerboni P et al. Clinical epidemiology of venous thromboembolic disease. Results of a French Multicentre Registry. Eur Heart J 1997; 18: 685– 691. 14 Hansson P-O, Welin L, Tibblin G, Eriksson H. Deep vein thrombosis and pulmonary embolism in the general pop-

15

16

17 18

19

20

21 22

23 24 25

26

27

28

29

30

31

Prevalence, risks, and prognosis of PE and DVT

ulation. The study of men born in 1913. Arch Intern Med 1997; 157: 1665–1670. Hume M, Sevitt S, Thomas DP. Venous Thrombosis and Pulmonary Embolism. A Commonwealth Fund Book. Harvard University Press, Cambridge, MA, 1970. Nicolaides AN, Irving D. Clinical factors and the risk of deep venous thrombosis. Thromboembolism. In: Nicolaides AN, ed. Etiology, Advances in Prevention and Management. University Park Press, Baltimore, 1975: 199–204. Stein PD, Patel KC, Kalra NK et al. Deep venous thrombosis in a general hospital. Chest 2002; 122: 960–962. Stein PD, Huang H-L, Afzal A, Noor H. Incidence of acute pulmonary embolism in a general hospital: relation to age, sex, and race. Chest 1999; 116: 909–913. Stein PD, Patel KC, Kalra NK et al. Estimated incidence of acute pulmonary embolism in a community/teaching general hospital. Chest 2002; 121: 802–805. Stein PD, Kayali F, Olson RE. Estimated case fatality rate from pulmonary embolism, 1979–1998. Am J Cardiol 2004; 93: 1197–1199. Geerts WH, Heit JA, Clagett GP et al. Prevention of venous thromboembolism. Chest 2001; 119(suppl): 132S–175S. Taubman LB, Silverstone FA. Autopsy proven pulmonary embolism among the institutionalized elderly. J Am Geriatr Soc 1986: 34: 752–756. Morrell MT. The incidence of pulmonary embolism in the elderly. Geriatrics 1970; 25: 138–153. Busby W, Bayer A, Pathy J. Pulmonary embolism in the elderly. Age Ageing 1988; 17: 205–209. Stein PD, Gottschalk A, Saltzman HA, Terrin ML. Diagnosis of acute pulmonary embolism in the elderly. J Am Coll Cardiol 1991; 18: 1452–1457. Stein PD, Beemath A, Matta F et al. Clinical Characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med 2007 (In press). Mellemgaard K. The alveolar–arterial oxygen difference: its size and components in normal man. Acta Physiol Scand 1966; 67: 10–20. Harris EA, Kenyon AM, Nisbet HD, Seelye ER, Whitlock RML. The normal alveolar–arterial oxygen-tension gradient in man. Clin Sci (Colch) 1974; 46: 89–104. Filley GF, Gregoire F, Wright GW. Alveolar and arterial oxygen tensions and the significance of the alveolar– arterial oxygen tension difference in normal men. J Clin Invest 1954; 33: 517–529. Kanber GJ, King FW, Eshchar YR, Sharp JT. The alveolar– arterial oxygen gradient in young and elderly men during air and oxygen breathing. Am Rev Respir Dis 1968; 97: 376–381. Stein PD, Beemath A, Quinn DA et al. Usefulness of multidetector spiral computed tomography according to age and sex for diagnosis of acute pulmonary embolism. Am J Cardiol (In press).


BLUK077-Stein

March 28, 2007

15:8

Venous thromboembolism in the elderly

65

32 Stein PD, Henry JW. Acute pulmonary embolism presenting as pulmonary hemorrhage/infarction syndrome in the elderly. Am J Geriatr Cardiol 1998; 7: 36– 42. 33 Stein PD, Hull RD, Ghali WA et al. Tracking the uptake of evidence: two decades of hospital practice trends for diag-

nosing deep vein thrombosis and pulmonary embolism. Arch Intern Med 2003; 163: 1213–1219. 34 The PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759.

CHAPTER 11


BLUK077-Stein

March 28, 2007

12

15:10

CHAPTER 12

Pulmonary thromboembolism in infants and children

Twenty-three years of data were analyzed in children discharged from hospitals in the United States were based on the database of the National Hospital Discharge Survey (NHDS) [1]. Venous thromboembolic disease was coded at discharge from short-stay hospitals with sufficient frequency to indicate that it should be considered in the diagnosis of children with appropriate clinical findings [1]. Infants, teenage girls, and black children had the highest rates of diagnosis [1]. From 1979 to 2001, pulmonary embolism (PE) was diagnosed at discharge from short-stay non-Federal hospitals throughout the United States in 13,000 infants and children ≤17 years of age, deep venous thrombosis (DVT) in 64,000, and venous thromboembolism (VTE) in 75,000 [1]. Rates of diagnosis were 0.9 PE/100,000 children/year, 4.2 DVT/100,000 children/year, and 4.9 VTE/100,000 children/year (Figure 12.1) [1]. The rates of diagnosis of DVT and of VTE did not change from the triennial periods 1979–1982 to 1999–2001. Rates of diagnosis of PE, DVT, and VTE were higher in infants than children aged 2–14 years (Figures 12.2– 12.4) [1]. The rates were also higher in teenagers aged 15–17 years than in children aged 2–14 years, but the rates in teenagers were comparable to the rates in infants.

In teenagers aged 15–17 years, the rate of diagnosis of DVT was 2.1 times higher in girls than boys. Teenage girls with DVT had an associated pregnancy in 27%. The rate of DVT in nonpregnant teenage girls was 10 DVT/100,000 teenage girls/year, and the rate of pregnancy-associated DVT was 109 DVT/100,000 teenage girls/year. The rate of DVT in nonpregnant teenage girls did not differ significantly from the rate for teenage boys. The rate of diagnosis of PE in black children, 1.6/100,000/year, was 2.4 times higher than in white children, 0.7/100,000/year. The rate of DVT in black children, 5.7/100,000/year, was 1.7 times the rate in white children, 3.3/100,000/year. The rate of VTE in black children was 1.8 times the rate in white children. A double-peaked curve was shown for DVT [1, 2] and PE [1] with the highest rates of diagnosis in infants less than 1-year-old and a second peak in teenagers. Teenage girls had twice the rates of DVT and VTE as teenage boys, although in younger children the frequencies were comparable. Bernstein and associates, among adolescents, observed PE in twice as many girls as boys [3]. Pregnancy-related DVT accounted for the difference in rates between teenage boys and teenage girls.

6 5 4 3 2 1 0

4.9 4.2

0.9

3 2.2

1

0.4 0−1

PE

2.0

2

0

DVT

VTE

Figure 12.1 Thromboembolic disease in children ≤17 years of age. (Data from Stein et al. [1].)

66

PE/100,000/yr

PE, DVT, VTE/100,000 children/yr

4

2−14

15−17

Age (years) Figure 12.2 Pulmonary embolism (PE) in children. (Data from Stein et al. [1].)


March 28, 2007

CHAPTER 12

DVT/100,000/yr

10

15:10

67

Pulmonary thromboembolism in infants and children 9.9 8.7

8 6 4 2.1

2

12 VTE/100,000/yr

BLUK077-Stein

11.4

10.5

10 8 6 4

2.4

2 0

0 0−1

2−14

15−17

0−1

Age (years) Figure 12.3 Deep venous thrombosis (DVT) in children. (Data from Stein et al [1].)

Abortion and/or contraceptives were shown to be risk factors in 75% of female adolescents who had PE [3]. Teenage users of oral contraceptives did not appear to be at an increased risk of VTE compared to older users [4]. Indwelling catheter use was the most common predisposing factor for PE or DVT in children and adolescents, followed by surgery and trauma [5]. Neonatal thrombosis, with the exception of spontaneous renal vein thrombosis was almost always associated with indwelling catheters (86 of 97 cases, 89%) [6]. Lower extremity DVT in children, when unrelated to venous catheterization or surgery, appeared to be related to local infection of the involved extremity, trauma or immobilization [7]. One or more coagulopathies were reported in 26 of 40 (65%) children with venous thrombosis who had an evaluation for a deficiency of protein C, protein S, or antithrombin III and assessment for a lupus anticoagulant [8]. In the Canadian registry, 9% of children with VTE had a coagulopathy, but the majority did not receive coagulation testing [2, 9]. Among patients with heart disease at autopsy, the prevalence of PE as a cause of death was particularly high in children <10 years old [10]. Among medical patients with heart disease <10 years old, PE as a cause of death was found at autopsy in 14.5% and in postoperative children with heart disease at autopsy PE caused death in 8.0% [10]. Prior to 1962, under 50 cases of PE in children had been reported in the world literature [11]. Only 36 children with DVT and 10 with PE were identified among all Scottish hospital inpatients from 1968 to 1971 [12]. From 1975 to 1991, 308 children with DVT or PE were reported in the English and French literature [5]. A

2−14 Age (years)

15−17

Figure 12.4 Venous thromboembolism (VTE) in children. (Data from Stein et al. [1].)

Canadian registry of 15 hospitals from July 1990 to December 1992 identified 137 children aged 1 month to 17 years old who had DVT [2].

References 1 Stein PD, Kayali F, Olson RE. Incidence of venous thromboembolism in infants and children: data from the National Hospital Discharge Survey. J Pediatr 2004; 145: 563–565. 2 Andrew M, David M, Adams M et al. Venous thromboembolic complications (VTE) in children: first analyses of the Canadian Registry of VTE. Blood 1994; 83: 1251–1257. 3 Bernstein D, Coupey S, Schonberg SK. Pulmonary embolism in adolescents. Am J Dis Child 1986; 140: 667–671. 4 Royal College of General Practitioners’ Oral Contraception Study. Oral contraceptives, venous thrombosis, and varicose veins. J R Coll Gen Pract 1978; 28: 393–399. 5 David M, Andrew M. Venous thromboembolic complications in children. J Pediatr 1993; 123: 337–346. 6 Schmidt B, Andrew M. Neonatal thrombosis: report of a prospective Canadian and international registry. Pediatrics 1995; 96: 939–943. 7 Wise RC, Todd JK. Spontaneous, lower-extremity venous thrombosis in children. Am J Dis Child 1973; 126: 766– 769. 8 Nuss R, Hays T, Manco-Johnson M. Childhood thrombosis. Pediatrics 1995; 96: 291–294. 9 Manco-Johnson MJ. Disorders of hemostasis in childhood: risk factors for venous thromboembolism. Thromb Haemost 1997; 78: 710–714. 10 Pulido T, Aranda A, Zevallos MA et al. Pulmonary embolism as a cause of death in patients with heart disease. An autopsy study. Chest 2006; 129: 1282–1287. 11 Emery JL. Pulmonary embolism in children. Arch Dis Child 1962; 37: 591–595. 12 Jones DR, Macintyre IM. Venous thromboembolism in infancy and childhood. Arch Dis Child 1975; 50: 153–155.


BLUK077-Stein

March 12, 2007

13

15:31

CHAPTER 13

Venous thromboembolism in men and women The Tecumseh Community Health Study showed 4.5 PE/10,000 women/year compared with 1.75 PE/10,000 men/year [4]. The largest investigation was based on data from the National Hospital Discharge Survey [7]. From 1979 to 1999, 2,448,000 patients were discharged

(a)

(b)

80 Women

60 40 Men

20

80

Age-adjusted PE/100,000

PE/100,000 population

Pulmonary embolism (PE) has been reported to occur more frequently in women than in men due to estrogen use, childbearing, and a higher frequency of deep venous thrombosis (DVT) [1â&#x20AC;&#x201C;5]. A postmortem study showed PE in 11% of women and 7% of men [6].

Women

60

40

Men

20

(d) Women

160

120 Men

80

40

Men

100

50 1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

Figure 13.1 (a) Unadjusted rates of diagnosis of pulmonary embolism (PE) per 100,000 population over a 21-year period. The unadjusted rate of diagnosis was higher in women than men. (b) Age-adjusted rates of diagnosis of PE per 100,000 population over a 21-year period. The age-adjusted rates of diagnosis were comparable in men and women. (c) Unadjusted rates of diagnosis of deep venous thrombosis (DVT) per 100,000

Women

150

1979

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Year

Age-adjusted DVT/100,000

DVT/100,000 population

1999

1997

1995

1993

1991

Year

(c)

68

1989

1987

1985

1983

1981

1979

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Year

Year

population over a 21-year period. The unadjusted rate of diagnosis of DVT was higher in women than in men. (d) Age-adjusted rates of diagnosis of DVT per 100,000 population over a 21-year period. The age-adjusted rate of diagnosis was higher in women. (Reproduced from Stein et al. [7], with permission from American Medical Association. All rights reserved.)


March 12, 2007

1999

1997

1995

1993

1991

1989

(b) 60 45

Women

30

Men

15 0 1999

1997

1995

1993

1991

1989

1987

1985

Figure 13.2 Rates of use per 100,000 population of ventilation–perfusion (V–Q) lung scans. The rates were higher among women than in men. (Reproduced from Stein et al. [7], with permission from American Medical Association. All rights reserved.)

1987

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Year

Year

1983

Men

0

Men

0 1985

20

Women

1981

Women

15

1983

40

30

1981

60

45

1979

80

(a)

1979

from short-stay non-Federal hospitals with PE and 5,691,000 were discharged with DVT. Either PE or DVT was listed as the discharge diagnosis in 7,589,000 patients [7]. Given that PE and DVT are conditions that increase in incidence with age, age-adjusted rates of diagnosis as well as crude (unadjusted) rates were calculated [7]. In 1999, 139,000 patients were discharged from nonFederal short-stay hospitals with a diagnosis of PE. The rate of diagnosis not adjusted for age was 42 PE/100,000 men/year and 60 PE/100,000 women/year (Figure 13.1a) [7]. The age-adjusted rate of diagnosis of PE/100,000 population/year in men and women was comparable (Figure 13.1b) [7]. This was concordant with some prior investigations of PE [8–11], but PE has also been reported more frequently in men [12– 15] and older men [16]. Both the unadjusted rate of diagnosis of DVT/ 100,000 population/year and the age-adjusted rates of diagnosis of DVT/100,000 population/year were higher in women (Figures 13.1c and 13.1d) [7]. This is concordant with most prior literature [8, 16, 17], but the prior literature is not uniform. Objectively diagnosed DVT had also been reported to be more frequent in men [9]. In 1999, the unadjusted rate of diagnosis was 115 DVT/100,000 men/year and 154 DVT/100,000 women/year [7]. In 1999, 369,000 patients were discharged from non-Federal short-stay hospitals with a diagnosis of DVT. The rates of use/100,000 population/year of ventilation–perfusion lung scans (Figure 13.2) and

V−Q scans/100,000 population

69

Venous thromboembolism in men and women

Venograms/100,000 population

CHAPTER 13

15:31

Ultrasonography/100,000 population

BLUK077-Stein

Year Figure 13.3 (a) Rates of venography per 100,000 population over a 21-year period. The rates of use among men and women were comparable. (b) Rates of venous ultrasonography of the lower extremities per 100,000 population over a 21-year period. The rate of use was higher among women. (Reproduced from Stein et al. [7], with permission from American Medical Association. All rights reserved.)

venous ultrasound examinations of the lower extremities were higher among women (Figure 13.3b) [7]. In 1999, the rate of use of ventilation–perfusion lung scans was 18/100,000 men/year and 24/100,000 women/year. During the same year the rate of use of venous ultrasound examinations of the lower extremities was 31/100,000 men/year and 38/100,000 women/year. A more frequent use of venous ultrasonography for the diagnosis of DVT had been observed previously in women [18]. The rates of use/100,000 population/year of contrast venography among men and women were comparable (Figure 13.3a). The durations of hospitalization among patients with a primary discharge diagnosis of PE and of DVT were comparable among men and women (Figure 13.4a) [7].


BLUK077-Stein

March 12, 2007

15:31

70

PART I

(a)

Table 13.1 Signs and symptoms of acute pulmonary embolism in women and men.

Prevalence, risks, and prognosis of PE and DVT

PE hospitalization days

16 12 Women

8 Men

4 0 1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Year

DVT hospitalization days

8

Women

Men

4

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

0

(n = 132)

percent

percent

80

78

Pleuritic pain

60

57

Cough

41

40

Leg swelling

24

36†

Leg pain

23

30

Hemoptysis

10

21‡

Rales (crackles)

60

57

2

7*

* P < 0.05. † P < 0.04. ‡ P < 0.02, women vs. men. Data are from Quinn et al. [19].

16 12

Men

(n = 119)

Dyspnea

Pleural friction rub

(b)

Women

Year Figure 13.4 (a) Duration of hospitalization for men and women with a primary discharge diagnosis of pulmonary embolism (PE). Duration was comparable among genders. (b) Duration of hospitalization for men and women with a primary discharge diagnosis of deep venous thrombosis (DVT). The duration of hospitalization was comparable among genders. (Reproduced from Stein et al. [7], with permission from American Medical Association. All rights reserved.)

Regarding the signs and symptoms of acute PE, women recruited in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) less frequently had hemoptysis (10% versus 21%), leg swelling (24% versus 36%), and pleural friction rub (2% versus 7%) (Table 13.1). Except for minor differences of heart rate, the prevalence of other signs and symptoms of acute PE did not differ significantly between women and men [19]. The sensitivity and specificity of the ventilation– perfusion scan, using original PIOPED criteria [20], were the same in women and men. The positive predic-

tive values of a high-probability ventilation–perfusion scan were similar between women (86%) and men (90%) and the negative predictive values of normal or nearly normal ventilation–perfusion scans were also similar, 93% in women and 88% in men [19]. There was no difference between women and men during 1-year follow-up in the recurrence of PE [19].

References 1 Palevsky HI. Pulmonary hypertension and thromboembolic disease in women. Cardiovasc Clin 1989; 19: 267– 283. 2 Bernstein D, Goupey S, Schonberg SK. Pulmonary embolism in adolescents. AJDC 1986; 140: 667–671. 3 Coon W. Epidemiology of venous thromboembolism. Ann Surg 1977; 186: 149–164. 4 Coon WW, Willis PW, III, Keller JB. Venous thromboembolism and other venous disease in the Tecumseh Community Health Study. Circulation 1973; 48: 839–846. 5 Breckenridge RT, Ratnoff OD. Pulmonary embolism and unexpected death in supposedly normal persons. N Engl J Med 1964; 270: 298–299. 6 Karwinski B, Svendsen E. Comparison of clinical and postmortem diagnosis of pulmonary embolism. J Clin Pathol 1989; 42: 135–139. 7 Stein PD, Hull RD, Patel KC et al. Venous thromboembolic disease: comparison of the diagnostic process in men and women. Arch Intern Med 2003; 163: 1689–1694. 8 Ferrari E, Baudouy M, Cerboni P et al. Clinical epidemiology of venous thromboembolic disease. Results of a


BLUK077-Stein

March 12, 2007

CHAPTER 13

9

10

11

12

13

14

15:31

Venous thromboembolism in men and women

French multicentre registry. Eur Heart J 1997; 18: 685– 691. Anderson FA, Jr, Wheeler HB, Goldberg RJ et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med 1991; 151: 933–938. Stein PD, Huang H-L, Afzal A, Noor H. Incidence of acute pulmonary embolism in a general hospital: relation to age, sex, and race. Chest 1999; 116: 909–913. Stein PD, Patel KC, Kalra NK et al. Estimated incidence of acute pulmonary embolism in a community/teaching general hospital. Chest 2002; 121: 802–805. Lilienfeld DE, Godbold JH, Burke GL, Sprafka JM, Pham DL, Baxter J. Hospitalization and case fatality for pulmonary embolism in the twin cities: 1979–1984. Am Heart J 1990; 120: 392–395. Janke RM, McGovern PG, Folsom AR. Mortality, hospital discharges, and case fatality for pulmonary embolism in the twin cities 1980–1995. J Clin Epidemiol 2000; 53: 103– 109. Silverstein MD, Heit JA, Mohr DN et al. Trends in the incidence of deep vein thrombosis and pulmonary

15

16

17 18

19

20

71

embolism. A 25-year population-based study. Arch Intern Med 1998; 158: 585–593. Giuntini C, Di Ricco G, Marini C, Melillo E, Palla A. Pulmonary embolism: epidemiology. Chest 1995; 107(suppl): 3S–9S. Kniffin WD, Jr, Baron JA, Barrett J, Birkmeyer JD, Anderson FA, Jr. The epidemiology of diagnosed pulmonary embolism and deep venous thrombosis in the elderly. Arch Intern Med 1994; 154: 861–866. Stein PD, Patel KC, Kalra NK et al. Deep venous thrombosis in a general hospital. Chest 2002; 122: 960–962. Beebe HG, Scissons RP, Salles-Cunha SX et al. Gender bias in use of venous ultrasonography for diagnosis of deep vein thrombosis. J Vasc Surg 1995; 22: 538– 542. Quinn DA, Thompson BT, Terrin ML et al. A prospective investigation of pulmonary embolism in women and men. JAMA 1992; 268: 1689–1696. A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759.


BLUK077-Stein

April 14, 2007

14

14:48

CHAPTER 14

Comparison of the diagnostic process in black and white patients

Introduction The death rate among patients with pulmonary embolism (PE) in the United States is higher among black patients than among white patients [1]. Recognizing that there is a racial disparity in the use of some health services [2â&#x20AC;&#x201C;10], one could speculate that such a disparity could contribute to the higher death rate among black patients with PE. However, nothing was found to suggest that diagnostic tests were being withheld in black patients [11]. Also, there was no evidence of a failure to reach a diagnosis in black patients with thromboembolic disease [11]. Caution has been recommended in the use of race as a variable [12]. It is necessary to account for distinctions between race and socioeconomic status [12]. An emphasis on ethnic groups rather than on race implies an appreciation of cultural and behavioral attitudes, beliefs, lifestyle patterns, diet, environmental living conditions, and other factors [13]. Disparities in health care may be due to a lack of appropriate health care messages, limited access to care and services, lack of trust of health care providers, and racial bias among medical care providers [12]. Data on trends over 21 years and relative differences among races in the prevalence of PE and DVT (deep venous thrombosis) and the use of diagnostic tests were obtained from the National Hospital Discharge Survey from 1979 to 1999 [11]. During these 21 years, 2,448,000 patients were discharged from short-stay non-Federal hospitals with PE, 5,691,000 patients were discharged with DVT and 7,589,000 patients were discharged with either PE or DVT. The rate of diagnosis of PE/100,000 population/year not adjusted for age, was comparable among black and white patients (Figure 14.1a) [11]. In 1999, the rate of diagnosis among black patients was 41 PE/100,000

72

population/year and among white patients it was 42 PE/100,000 population/year. Despite some year-toyear fluctuation, the rate of diagnosis of PE among black patients did not change appreciably over the 21year period studied. White patients, however, showed a decreasing rate of PE in the population over the period of survey [11]. Adjustment for age caused the rate of diagnosis of PE among black patients to separate from white patients after the mid-1980s (Figure 14.1b) [11]. The age-adjusted rates indicate the hospitalization rates for black and white patients assuming identical age distributions for both populations. The age-adjusted rate of diagnosis of PE/100,000 population/year was higher in black patients than in white patients (Figure 14.1b). From 1979 to 1992, there was a decline in the rate of diagnosis both among black patients and among white patients. From 1992 to 1999, the rate of diagnosis was constant in black patients, but the rate of diagnosis increased somewhat in white patients. The rate of diagnosis of DVT/100,000 population/ year, not adjusted for age, was comparable among black and white patients (Figure 14.1c) [11]. In 1999, the rate of diagnosis among black patients was 110 DVT/100,000 population/year and among white patients it was 115 DVT/100,000 population/year. The rate of diagnosis increased over the 21-year period of survey in black patients. Among white patients, the rate of diagnosis remained unchanged during the period of survey. The age-adjusted rate of diagnosis of DVT/100,000 population/year was higher in black patients than in white patients (Figure 14.1d). From 1979 to 1992, the age-adjusted rate of diagnosis was constant in black patients, whereas white patients showed a gradual decrease in the rate. From 1992 to 1999, the rate of diagnosis increased sharply in black patients; the rate increased more gradually in white patients.


BLUK077-Stein

April 14, 2007

CHAPTER 14

14:48

73

VTE in black and white patients

60

Black patients

40 White patients

20

Black patients

60 40

White patients

20 0

Year

150

Black patients

100 White patients

50 0

200

Black patients

150 100 White patients

50 0

1999

1997

1995

60 45

Black patients

30 15

White patients

0 1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

V−Q scans/100,000 population

diagnosis over a 21-year period of deep venous thrombosis (DVT) per 100,000 population according to race. The rate of diagnosis was comparable among black and white patients. (d) Age-adjusted rates of diagnosis of DVT per 100,000 population according to race. The rate was higher in black patients than in white patients. (Reproduced from Stein et al. [11], with permission from American Medical Association. All rights reserved.)

1979

The use of pulmonary angiograms/100,000 population/year was low among both black and white patients. The frequency of use of angiograms among black patients was too low to calculate rates. However, inspection of the data showed no suggestion of a disparity of use between black and white patients. Rates of use of ventilation–perfusion lung scans among black and white patients from 1979 to 1999 were comparable (Figure 14.2). The use of ventilation– perfusion lung scans in both groups increased sharply between 1979 and 1986. The use then declined in black as well as white patients.

1993

Use of diagnostic tests for PE and deep vein thrombosis

1991

Year

Year Figure 14.1 (a) Unadjusted (crude) rates of diagnosis over a 21-year period of pulmonary embolism (PE) per 100,000 population according to race. The rates of diagnosis were comparable among black and white patients. (b) Ageadjusted rates of diagnosis of PE per 100,000 population according to race. There was no change in the rate of diagnosis among black patients over the 21-year period. White patients, however, showed a decreasing rate of PE over the period of survey. (c) Unadjusted (crude) rates of

1989

1987

1985

1983

1981

1979

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Age-adjusted DVT/100,000

(d)

(c) DVT/100,000 population

1999

1997

1995

1993

1991

Year

1989

1987

1985

1983

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

1981

0

80

1979

PE/100,000 population

80

Age-adjusted PE/100,000

(b)

(a)

Year Figure 14.2 Rates of use per 100,000 population of ventilation–perfusion (V–Q) lung scans among black and white patients. The rates were comparable. (Reproduced from Stein et al. [11], with permission from American Medical Association. All rights reserved.)


BLUK077-Stein

April 14, 2007

14:48

74

PART I

Prevalence, risks, and prognosis of PE and DVT

Venography/100,000 population

(a) 50

25

White patients Black patients

0 1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Year

60 Black patients 40 White patients

20

0 1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Ultrasonography/100,000 population

(b)

Year

Figure 14.3 (a) Rates of use per 100,000 population of contrast venography among black and white patients. The rates were comparable. (b) Rates of use per 100,000 population of venous ultrasound of the lower extremities among black and white patients. The rates were comparable. (Reproduced from Stein et al. [11], with permission from American Medical Association. All rights reserved.)

(a)

PE, days

18

12 Black patients White patients 1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

6

Year

(b)

DVT, days

15 10 Black patients 5

White patients

0 1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Year

Figure 14.4 (a) Duration of hospitalization for black and white patients with a primary discharge diagnosis of pulmonary embolism (PE). The durations were comparable among races. (b) Duration of hospitalization for black and white patients with a primary discharge diagnosis of deep venous thrombosis (DVT). The durations were comparable among races. (Reproduced from Stein et al. [11], with permission from American Medical Association. All rights reserved.)


BLUK077-Stein

April 14, 2007

CHAPTER 14

14:48

VTE in black and white patients

The rates of use of contrast venography among black and white patients from 1979 to 1999 were comparable (Figure 14.3a) [11]. The rates of use of venograms in black and white patients peaked in the late 1980s. The rates of use of venous ultrasound of the lower extremities among black and white patients from 1979 to 1999 were comparable (Figure 14.3b). Between 1988 and 1999, the ratio of use of ultrasounds in black to white patients ranged from 1.2 to 1.8. The rate of use of venous ultrasounds increased in both until 1991.

Use of medical facilities for treatment of PE and deep vein thrombosis The duration of hospitalization for black and white patients with a primary discharge diagnosis of PE was comparable (Figure 14.4a) [11]. The duration of hospitalization decreased over time in both groups. The duration of hospitalization for black and white patients with a primary discharge diagnosis of DVT was also comparable (Figure 14.4b). As with PE, the duration of hospitalization decreased over time in both. This reflects the use of protocols for a rapid attainment of therapeutic levels of heparin, a shorter duration of therapy with heparin, and use of low-molecular heparin [14]. These data were obtained entirely from hospitalized patients [11]. Information related to care after hospitalization was not available. Some have reported a shorter duration of anticoagulant therapy in black patients [15].

References 1 Lilienfeld DE. Decreasing mortality from pulmonary embolism in the United States, 1979–1996. Int J Epidemiol 2000; 29: 465–469. 2 Freeman HP, Payne R. Racial injustice in health care. N Engl J Med 2000; 342: 1045–1047.

75

3 Schneider EC, Zaslavsky AM, Epstein AM. Racial disparities in the quality of care for enrollees in Medicare managed care. JAMA 2002; 287: 1288–1294. 4 Ayanian JZ, Udvarhelyi S, Gatsonis CA et al. Racial differences in the use of revascularization procedures after coronary angiography. JAMA 1993; 269: 2642–2646. 5 Gornick ME, Eggers PW, Reilly TW et al. Effects of race and income on mortality and use of services among Medicare beneficiaries. N Engl J Med 1996; 335: 791–799. 6 Roetzheim RG, Pal N, Tennant C, et al. Effect of health insurance and race on early detection of cancer. J Natl Cancer Inst 1999; 91: 1409–1415. 7 Kasiske BL, Neylan JF, III, Riogio RR et al. The effect of race on access and outcome in transplantation. N Engl J Med 1991; 324: 302–307. 8 Bach PB, Cramer LD, Warren JL, Begg CB. Racial differences in the treatment of early stage of lung cancer. N Engl J Med 1999; 341: 1198–1205. 9 Brawley OW, Freeman HP. Race and outcomes: is this the end of the beginning for minority health research? J Natl Cancer Inst 1999; 91: 1908–1909. 10 Roach M, III, Cirrincione C, Budman D et al. Race and survival from breast cancer based on Cancer and Leukemia Group B Trial 8541. Cancer J Sci Am 1997; 3: 107–112. 11 Stein PD, Hull RD, Patel KC et al. Venous thromboembolic disease: comparison of the diagnostic process in blacks and whites. Arch Intern Med 2003; 163: 1843–1848. 12 Thomas SB. The color line: race matters in the elimination of health disparities. Am J Public Health 2001; 91: 1046– 1048. 13 Haynes MA, Smedley BD (eds.) The Unequal Burden of Cancer: An Assessment of NIH Research and Programs for Ethnic Minorities and the Medically Underserved. National Academy Press, Washington, DC, 1999: 19. 14 Hyers TM, Agnelli G, Hull RD et al. Antithrombotic therapy for venous thromboembolic disease. Chest 2001; 119(suppl): 176S–193S. 15 Ganz DA, Glynn RJ, Mogun H et al. Adherence to guidelines for oral anticoagulation after venous thrombosis and pulmonary embolism. J Gen Intern Med 2000; 15: 776– 781.


BLUK077-Stein

March 12, 2007

15

15:39

CHAPTER 15

Pulmonary thromboembolism in Asians/Pacific Islanders

Introduction Most investigations of pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) in countries in Asia suggest that the rate of diagnosis is lower in Asians [1–13] than in persons from North America or Europe [14–18], although some did not describe differences [19–22]. Data from the National Hospital Discharge Survey combined with data from the Bureau of the Census show that in the United States the rate of diagnosis of DVT, PE, and VTE; the incidence in hospitalized patients; and mortality rate from PE were all lower in Asians/Pacific Islanders than in whites and African Americans [23].

Unadjusted population-based rates of diagnosis From 1990 to 1999, rates of diagnosis of PE in Asian/Pacific/Islanders were 5/100,000 population/ year, compared with 34/100,000/year in whites and 39/100,000/year African Americans [23] (Table 15.1, Figure 15.1). Regarding DVT, rates of diagnosis in Asian/Pacific Islanders were 20/100,000 population/year compared with 98/100,000/year in whites and 105/100,000/ year in African Americans. Rates for VTE were 23/ 100,000 population/year in Asian/Pacific Islanders, 122/100,000/year in whites and 134/100,000/year in African Americans [23]. Others too, reported VTE among Asian Americans to be lower than in whites and African Americans in two investigations [12, 13]. Klatsky et al. [12] studied the rate of a primary diagnosis of VTE among hospitalized patients in California from 1978 to 1985, and reported rates of 2/100,000 population/year in Asian Americans, 21/100,000 population/year in whites, and 22/100,000 population/year in African Americans.

76

Klatsky et al. [12] reported even lower rates of diagnosis of venous thromboembolic disease among Asian Americans, whites, and African Americans in California than we observed throughout the United States [23]. The likely reason is that they required a primary diagnosis of PE or DVT, whereas we included all patients with PE or DVT [23]. White et al. [13] investigated idiopathic DVT in hospitalized patients in California between 1991 and 1994, and reported rates of 6/100,000 population/year in Asian Americans, 23/100,000 population/year in whites, and 29/100,000 population/year in African Americans. White et al. [13] also reported lower rates of DVT among Asian Americans, whites, and African Americans in California, which would seem to relate to the study requirement of idiopathic DVT and the exclusion of patients with cancer and those with temporary risk factors such as surgery and trauma. We included all patients with DVT [23].

Age-adjusted population-based rates of diagnosis Age-adjusted rates of diagnosis of PE were also lower in Asians/Pacific Islanders than in African Americans and whites (Table 15.1, Figure 15.1). Considering VTE (DVT and/or PE) [23], age-adjusted rates for the 10year period were also lower among Asians/Pacific Islanders than in African Americans and whites [23] (Table 15.1, Figure 15.1). Among Asians/Pacific Islanders, the age-adjusted rates of DVT were similar in men and women (19/100,000/year versus 24/100,000/year), as were the rates for PE (5/100,000/year versus 8/100,000/year), and VTE (22/100,000/year versus 30/100,000/year) [23].

Incidence in hospitals There was a disproportionately low rate of hospitalizations among Asians/Pacific Islanders (5000


March 12, 2007

CHAPTER 15

15:39

77

Pulmonary thromboembolism in Asians

Figure 15.1 Age-adjusted rates (diagnoses/100,000 population/year) of pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) in hospitalized patients from 1990 to 1999. (Reprinted from Stein et al. [23], with permission from Elsevier.)

Age-adjusted rates/100,000/year

BLUK077-Stein

200 138

130 150 107

104

100

26 50

VTE

22 36

40

7

0 Whites

hospitalizations/100,000 population/year) compared with whites (9000 hospitalizations/100,000 population/year) and African Americans (11,000 hospitalizations/100,000 population/year) [23]. The incidence of DVT among hospitalized Asians/Pacific Islanders aged ≥20 years (0.4/100 hospitalizations) was lower than among African Americans (1.1/100 hospitalizations) and whites (1.1/100 hospitalizations) (Table 15.2) [23]. In patients ≥20 years, the incidence of PE in hospitalized patients was also lower among Asians/Pacific Islanders (0.1/100 hospitalizations) than among whites (0.4/100 hospitalizations) and African Americans (0.4/100 hospitalizations) [23]. Similarly, in patients aged ≥20 years, the incidence of VTE in hospitalized patients was lower in Asians/Pacific Islanders (0.5/100 hospitalizations) than in whites (1.4/100 hospitalizations) and African Americans (1.4/100 hospitalizations) [23].

Population mortality rates Age-adjusted mortality from 1990 to 1998 for PE was lower in “others” (1.0 deaths/100,000 population/year) than in whites (3.4 deaths/100,000 population/year) and African Americans (6.9 deaths/100,000 population/year) (Table 15.3) [23]. Mortality from PE among Asians/Pacific Islanders was included in mortality rates described in “others.” Lower rates of fatal PE in Asians were reported at autopsy in Singapore [1] but not in Chinese in Hong Kong [21]. The low rate of diagnosis was related partly to the low rate of all hospitalizations in Asians/Pacific Islanders. The lower incidence of thromboembolic disease in hospitalized patients was independent of effects of a disproportionately low rate of all-cause hospitalization

African Americans

DVT PE

Asian/Pacific Islanders

among Asians/Pacific Islanders. The lower mortality among Asians/Pacific Islanders was also independent of the lower rate of all-cause hospitalizations. A lower rate of VTE in Asians than whites has been observed in several Asian ethnicities, as well as in many different clinical settings, including autopsy, where possible ethnic customs related to outpatient therapy would hardly have affected the results (Table 15.4). In the study by Klatsky et al. on Asians in northern California, 42.2% were Chinese, 11.9% were Japanese, 29.5% were Filipinos, 5.0% were South Asians, and 11.2% were other Asians [13]. Lower rates of thromboembolic disease have been shown in investigations in China [1, 5, 8–11], Japan [2], Thailand [3, 4], and Malaysia [6, 7]. Some of these investigations included patients from other Asian regions, including India [1, 5, 6]. A lower rate of PE was found at autopsy among Indians from Vellore, South India, than among patients in Boston and Los Angeles [2]. A lower prevalence of genetically induced abnormalities predisposing to VTE, such as factor V Leiden, has been speculated to contribute to the lower incidence of VTE in Asians [24]. Factor V Leiden, an abnormal factor V protein that is relatively resistant to degradation by protein C [25], is the most common genetic mutation predisposing to VTE [25, 26]. Factor V Leiden has been found in 4–5% of whites in North America and Europe [27, 28], 0.9–1.2% of African Americans, and in only 0–0.5% of Asians [27–29]. Differences in coagulation factors may also contribute to the low rate of VTE in Asians. Plasma fibrinogen levels are lower in Japanese than in whites, irrespective of whether they live in a rural or urban area of Japan or whether they are Japanese Americans [30]. Blood levels of factor VIIc and factor VIIIc were


78 18,000

5,000

127,000

736,000

embolism

21,000

440,000

2,641,000

embolism

P < 0.0005 Asian-Pacific Islanders vs. Caucasians and African American. Reprinted from Stein et al. [23], with permission from Elsevier.

* 10 year sum of census estimates.

Asian-Pacific Islander

344,000

2,128,000

Caucasian

African American

thrombosis

thrombo-

Race

Venous

venous

Pulmonary

Deep

Number of patients (1990–1999)

92,179,599

328,136,068

2,170,606,118

Population*

39 5†

20†

34

embolism

Pulmonary

105

98

thrombosis

venous

Deep

23†

134

122

embolism

thrombo-

Venous

(Diagnoses/100,000 population/year)

Crude rate

22†

107

104

thrombosis

venous

Deep

7†

40

36

embolism

Pulmonary

26†

138

130

embolism

thrombo-

Venous

(Diagnoses/100,000 population/year)

Age-adjusted rate

March 12, 2007

Table 15.1 Deep venous thrombosis, pulmonary embolism, and venous thromboembolism according to race (1990–1999).

BLUK077-Stein

15:39


BLUK077-Stein

March 12, 2007

CHAPTER 15

15:39

79

Pulmonary thromboembolism in Asians

Table 15.2 Deep venous thrombosis and venous thromboembolic disease in hospitalized patients according to race and age groups (1990–1999). Number of patients

Diagnoses/100

(1990–1999)

hospitalizations

Deep

Venous

Age

venous

thrombo-

All discharges

Deep venous

Venous thrombo-

(years)

thrombosis

embolism

(1990–1999)

thrombosis

embolism

Caucasians 0–19

25,000

29,000

20,222,000

0.1

0.1

20–49

419,000

503,000

67,389,000

0.6

0.7

50–69

705,000

869,000

48,630,000

1.5

1.8

≥70

979,000

1,240,000

67,112,000

1.5

1.8

2,128,000

2,641,000

203,352,000

1.0

1.3

All ages African Americans 0–19 20–49

9,000

12,000

6,620,000

0.1

0.2

130,000

166,000

16,553,000

0.8

1.0

50–69

104,000

133,000

7,683,000

1.4

1.7

≥70

101,000

129,000

6,429,000

1.6

2.0

All ages

344,000

440,000

37,285,000

0.9

1.2

Asian-Pacific Islanders 0–19

—*

—*

20–49

6,000

8,000

592,000

—*

—*

2,110,000

0.3

0.4

50–69

7,000

8,000

892,000

0.8

0.9

≥70

4,000

5,000

936,000

0.4

0.5

18,000†

21,000

4,530,000

0.4

0.5

All ages * Insufficient data. †

Represents the sum of all age groups before rounding. Reprinted from Stein et al. [23], with permission from Elsevier.

lower in rural and urban Japanese than in whites and Japanese Americans [30]. Some differences of coagulation factors between Asians and whites are attributable to environmental factors, especially diet and smoking, as well as genetic differences [26, 30]. In some venographic studies of patients recovering from total hip replacement, the incidence of asymp-

tomatic DVT in Asians was reported to be comparable with rates observed in the general population in North America [19, 20]. This may suggest that Asians have a more efficient inactivation of coagulation by activated protein C or more fibrinolytic activity [24], thereby lowering the rate of PE and the rate of symptomatic DVT.

Table 15.3 Mortality from pulmonary embolism (1990–1998). Crude mortality rate Death count Race

(1990–1998)

Population*

Age-adjusted mortality rate

(deaths/100,000

deaths/100,000

population/year)

(population/year)

Caucasian

66,879

1,946,786,446

3.4

3.4

African American

14,033

293,437,621

4.8

6.9

Other

584

100,803,416

0.6

1.0

* Sum of census estimates 1990–1998. Reprinted from Stein et al. [23], with permission from Elsevier.


Clinical/laboratory Clinical/laboratory

Major gynecological surgery Hospitalized patients

Deep venous thrombosis

Deep venous thrombosis

Venogram

80 Hospital Hospital idiopathic

Primary venous thromboembolism

Deep venous thrombosis

Discharge codes

Discharge codes and chart review

Clinical/laboratory autopsy

Autopsy

Venogram

Venogram

Venogram

Venogram

Fibrinogen uptake test

Fibrinogen uptake test/venogram

Clinical/laboratory

Reprinted from Stein et al. [23], with permission from Elsevier.

* Diagnosis of deep venous thrombosis was made by venous ultrasound or venography only [16].

Autopsy Hospital

Femoral fracture

Deep venous thrombosis

Fatal pulmonary embolism

Femoral fracture

Deep venous thrombosis

Venous thromboembolism

Total hip replacement

Gynecological surgery

Deep venous thrombosis Total knee replacement

Stroke

Deep venous thrombosis

Fibrinogen uptake test

General surgery

Deep venous thrombosis

Deep venous thrombosis

Deep venous thrombosis

Fibrinogen uptake test

General surgery General surgery

Deep venous thrombosis

Deep venous thrombosis

Laboratory

Pregnancy Pregnancy

Deep venous thrombosis

Venous thromboembolism

Fibrinogen uptake test

Fibrinogen uptake test

Total hip replacement

Autopsy

Autopsy

Hysterectomy

Autopsy

Pulmonary embolism

Deep venous thrombosis

Autopsy

Pulmonary embolism

Method

6/100,000/ year

2/100,000/year

0.013

3.6

53

50

77

64

2.6

17

2.6

12

0.02

0.2

0.04

0.08

3.8

1.7

4

1.4

0.16

Asia (%)

13â&#x20AC;&#x201C;19/100,000/year

21â&#x20AC;&#x201C;22/100,000/year

0.27

4.2

48

48

64

54

16

55

25

25

25

0.1

0.1

0.8

35

16

54

14.0, 14.7

14

European (%)

North American/

California

California

China

China

China

Malaysia

Malaysia

Malaysia

China

China

China

Malaysia

Malaysia

China

Malaysia

Singapore

Thailand

Thailand

Thailand

Japan

Singapore

location

Asian

13

12

11, 18

21

15, 20

15, 19

15, 19

15, 19

10, 15

9, 15

8, 15

7, 15

15

17, 22

6, 17

5, 16*

4

4

3, 15

2

1, 14

References

March 12, 2007

Deep venous thrombosis

Group

Venous thromboembolism

Rate of diagnosis

Table 15.4 Comparison of rates of venous thromboembolism, pulmonary embolism, and deep venous thrombosis in Asians versus North Americans/Europeans.

BLUK077-Stein

15:39


BLUK077-Stein

March 12, 2007

15:39

Pulmonary thromboembolism in Asians

81

Asians appear to be more sensitive to warfarin than whites [31]. The target range of the International Normalized Ratio (INR) for patients with nonvalvular atrial fibrillation, 1.5–2.1 in Japanese [32], and for mechanical prosthetic heart valves, 1.5–2.5 in Japanese [33], is lower than the target range of 2.0–3.0 for most indications in patients from North America and Europe [34, 35]. Clinical problems with toxicity and dosage adjustment of warfarin have been found in patients with a genetic polymorphism to CYP2C9 [36, 37], but this polymorphism does not appear to be more frequent in Asians than whites, nor does it explain the observed greater sensitivity to warfarin [38]. Differences in the activity of CYP2C9 between Asians and whites may relate to differences in the pharmacokenetics of warfarin [38].

bolism in the Chinese in Hong Kong. Int J Cardiol 1988; 20: 373–380. Klatsky AL, Armstrong MA, Poggi J. Risk of pulmonary embolism and/or deep venous thrombosis in AsianAmericans. Am J Cardiol 2000; 85: 1334–1337. White RH, Zhou H, Romano PS. Incidence of idiopathic deep venous thrombosis and secondary thromboembolism among ethnic groups in California. Ann Intern Med 1998; 128: 737–740. Coon WW, Coller FA. Some epidemiologic considerations of thromboembolism. Surg Gynecol Obstet 1959; 109: 487–501. Geerts WW, Heit JA, Clagett GP et al. Prevention of venous thromboembolism. Chest 2001; 119(suppl): 132S– 175S. Stein PD, Patel KC, Kalra NK et al. Deep venous thrombosis in a general hospital. Chest 2002; 122: 960–962. Andersen BS, Steffensen FH, Sorensen HT, Nielsen GL, Olsen J. The cumulative incidence of venous thromboembolism during pregnancy and puerperium—an 11 year Danish population-based study of 63,300 pregnancies. Acta Obstet Gynecol Scand 1998; 77: 170–173. Stein PD, Patel KC, Kalra NK et al. Estimated incidence of acute pulmonary embolism in a community/teaching general hospital. Chest 2002; 121: 802–805. Dhillon KS, Askander A, Doraisamy S. Postoperative deep-vein thrombosis in Asian patients is not a rarity: a prospective study of 88 patients with no prophylaxis. J Bone Joint Surg 1996; 78B: 427–430. Mok CK, Hoaglund FT, Rogoff SM, Chow SP, Ma A, Yau AC. The incidence of deep vein thrombosis in Hong Kong Chinese after hip surgery for fracture of the proximal femur. Br J Surg 1979; 66: 640–642. Dickens P, Knight BH, Ip P, Fung WS. Fatal pulmonary embolism: a comparative study of autopsy incidence in Hong Kong and Cardiff, Wales. Forensic Sci Int 1997; 90: 171–174. Chan LY, Tam WH, Lau TK. Venous thromboembolism in pregnant Chinese women. Obstet Gynecol 2001; 98: 471–475. Stein PD, Kayali F, Olson RE, Milford, CE. Pulmonary thromboembolism in Asian-Pacific Islanders in the United States: analysis of data from the National Hospital Discharge Survey and the United States Bureau of the Census. Am J Med 2004; 116: 435–442. White RH. The epidemiology of venous thromboembolism. Circulation 2003; 107: I4–I8. Svensson PJ, Dahlback B. Resistance to activated protein C as a basis for venous thrombosis. N Engl J Med 1994; 330: 517–522. Franco RF, Reitsma PH. Genetic risk factors of venous thrombosis. Hum Genet 2001; 109: 369–384.

CHAPTER 15

12

13

14

15

16 17

References 1 Hwang WS. The rarity of pulmonary thromboembolism in Asians. Singapore Med J 1968; 9: 276–279. 2 Hirst AE, Gore I, Tanaka K, Samuel I, Krishtmukti I. Myocardial infarction and pulmonary embolism. Arch Pathol 1965; 80: 365–370. 3 Atichartakarn V, Pathepchotiwong K, Keorochana S, Eurvilaichit C. Deep vein thrombosis after hip surgery among Thai. Arch Intern Med 1988; 148: 1349–1353. 4 Chumnijarakij T, Poshyachinda V. Postoperative thrombosis in Thai women. Lancet 1975; 1: 1357–1358. 5 Kueh YK, Wang TL, Teo CP, Tan YO. Acute deep vein thrombosis in hospital practice. Ann Acad Med Singapore 1992; 21: 345–348. 6 Liam CK, Ng SC. A review of patients with deep vein thrombosis diagnosed at university hospital, Kuala Lumpur. Ann Acad Med 1990; 19: 837–840. 7 Cunningham IGE, Yong NK. The incidence of postoperative deep vein thrombosis in Malaysia. Br J Surg 1974; 61: 482–483. 8 Nandi P, Wong KP, Wei WI, Ngan H, Ong GB. Incidence of postoperative deep vein thrombosis in Hong Kong Chinese. Br J Surg 1980; 67: 251–253. 9 Tso SC. Deep vein thrombosis after strokes in Chinese. Aust N Z J Med 1980; 10: 513–514. 10 Tso SC, Wong V, Chan V, Chan TK, Ma HK, Todd D. Deep vein thrombosis and changes in coagulation and fibrinolysis after gynaecological operations in Chinese: the effect of oral contraceptives and malignant disease. Br J Haematol 1980; 46: 603–612. 11 Woo KS, Tse LK, Tse CY, Metreweli C, Vallance-Owen J. The prevalence and pattern of pulmonary thromboem-

18

19

20

21

22

23

24 25

26


BLUK077-Stein

March 12, 2007

15:39

82

PART I

27 Ridker PM, Miletich JP, Hennekens CH, Buring JE. Ethnic distribution of factor V Leiden in 4047 men and women. Implications for venous thromboembolism screening. JAMA 1997; 277: 1305–1307. 28 Rees DC, Cox M, Clegg JB. World distribution of factor V Leiden. Lancet 1995; 346: 1133–1134. 29 Gregg JP, Yamane AJ, Grody WW. Prevalence of the factor V-Leiden mutation in four distinct American ethnic populations. Am J Med Genet 1997; 73: 334–336. 30 Iso H, Folsom AR, Wu KK et al. Hemostatic variables in Japanese and Caucasian men. Plasma fibrinogen, factor VIIc, factor VIIIc, and von Willebrand factor and their relations to cardiovascular disease risk factors. Am J Epidemiol 1989; 130: 925–934. 31 Takahashi H, Echizen H. Pharmacogenetics of CYP2C9 and interindividual variability in anticoagulant response to warfarin. Pharmacogenomics J 2003; 3: 202–214. 32 Yamaguchi T. Optimal intensity of warfarin therapy for secondary prevention of stroke in patients with nonvalvular atrial fibrillation: a multicenter, prospective, randomized trial. Stroke 2000; 31: 817–821.

33 Matsuyama K, Matsumoto M, Sugita T et al. Anticoagulant therapy in Japanese patients with mechanical mitral valves. Circulation 2002; 66: 668–670. 34 Hirsh J, Dalen JE, Anderson DR et al. Managing oral anticoagulant therapy. Chest 2001; 119(suppl): 3S–7S. 35 Stein PD, Alpert JS, Bussey HI, Dalen JE, Turpie AG. Antithrombotic therapy in patients with mechanical and biological prosthetic heart valves. Chest 2001; 119(suppl): 220S–227S. 36 Tabrizi AR, Zehnbauer BA, Borecki IB, McGrath SD, Buchman TG, Freeman BD. The frequency and effects of cytochrome P450 (CYP) 2C9 polymorphisms in patients receiving warfarin. J Am Coll Surg 2002; 194: 267–273. 37 Aithal GP, Day CP, Kesteven PJL, Daly AK. Association of polymorphisms in the cytochrome P450 CYP2C9 with warfarin dose requirement and risk of bleeding complications. Lancet 1999; 353: 717–719. 38 Takahashi H, Wilkinson GR, Caraco Y et al. Population differences in S-warfarin metabolism between CYP2C9 genotype-matched Caucasian and Japanese. Clin Pharmacol Ther 2003; 73: 253–263.

Prevalence, risks, and prognosis of PE and DVT


January 5, 2007

CHAPTER 16

Pulmonary thromboembolism in American Indians and Alaskan Natives 155 131 150

71

100 126

107

50

55

VTE DVT

A Afr m ic er a ic n an A s A me la ri sk ca an n N Ind at ia iv n es s

ca

si

an

s

0

au

Rates of diagnosis of deep venous thrombosis (DVT), pulmonary embolism (PE), and venous thromboembolism (VTE) are lower in American Indians and Alaskan Natives than rates in African Americans or Caucasians [1, 2]. From 1996 to 2001, the rate of diagnosis of VTE (PE and/or DVT) in American Indians/Alaskan Natives, based on combined data from the National Hospital Discharge Survey and the Indian Health Service, was 71/100,000/year, compared with 155/100,000/year in African Americans and 131/100,000 in Caucasians (Figure 16.1) [1]. Rates of diagnosis of DVT according to race are shown in Figure 16.1 [1]. The number of PEs in American Indians/Alaskan Natives was too low to give an accurate estimate of the rate of diagnosis. Only 1 patient with PE was hospitalized in Indian Health Service hospitals between 1996 and 2001. During this interval, an estimated 420,000 patients were hospitalized [1]. The rate of diagnosis of VTE among patients discharged from Indian Health Service hospital care from 1980 to 1996 was reported as 33/100,000/year in American Indians/Alaskan Natives [2]. The relatively low incidence of VTE in American Indians/Alaskan Natives would seem to be due to as yet undetermined genetic factors. A lower prevalence of Factor V Leiden in American Indians/Alaskan Natives populations (1.25%) compared with Caucasians (5.3%) perhaps contributes to the lower incidence of VTE in American Indians/Alaskan Natives [3]. The concept that racial groups can differ genetically and the differences can have medical importance has recently been discussed [4]. The possibility that American Indians/Alaskan Natives have different diets or lifestyles

DX/100,000/yr

16

18:14

C

BLUK077-Stein

Figure 16.1 Rates of diagnosis of deep venous thrombosis (DVT) and venous thromboembolic disease (VTE) in American Indians/Alaskan Natives, Caucasians, and African Americans from 1996 to 2001 based on combined data from the National Hospital Discharge Survey and the Indian Health Service. Rates of diagnosis of DVT and of VTE were lower in Indians/Alaskan Natives than Caucasians or African Americans (all differences P < 0.001). The rates of diagnosis in Caucasians and African Americans were comparable. (Reproduced from Stein et al. [1], with permission from American Medical Association. All rights reserved.)

that affect the rate of diagnosis of VTE cannot be excluded [5].

References 1 Stein PD, Kayali F, Olson RE, Milford, CE. Pulmonary thromboembolism in American Indians and Alaskan Natives. Arch Intern Med 2004; 164: 1804â&#x20AC;&#x201C;1806. 2 Hooper WC, Holman RC, Heit, JA, Cobb N. Venous thromboembolism hospitalizations among American Indians and Alaska Natives. Thrombosis Res 2003; 108: 273â&#x20AC;&#x201C; 278.

83


BLUK077-Stein

January 5, 2007

18:14

84

3 Ridker PM, Miletich JP, Hennekens CH, Buring JE. Ethnic distribution of Factor V Leiden in 4047 men and women. Implications for venous thromboembolism screening. JAMA 1997; 277: 1305– 1307.

PART I

Prevalence, risks, and prognosis of PE and DVT

4 Burchard EG, Ziv E, Coyle N et al. The importance of race and ethnic background in biomedical research and clinical practice. N Engl J Med 2003; 348: 1170–1175. 5 Cooper RS, Kaufman JS, Ward R. Race and genomics. N Engl J Med 2003; 348: 1166–1170.


BLUK077-Stein

April 14, 2007

17

14:54

CHAPTER 17

Venous thromboembolism in patients with cancer

Introduction Venous thromboembolism (VTE) is one of the most common complications in patients with cancer [1, 2]. The association of venous thrombosis with gastric carcinoma was described in 1865 [3]. A fourfold

increased risk of VTE has been found among patients with malignant neoplasm [4, 5]. Among patients insured through the Medicare program (patients aged 65 years or older), 0.6% of patients admitted with malignancy also had VTE [5]. A higher rate was reported by others [6]. Among patients hospitalized with solid

Table 17.1 Venous thromboembolism, pulmonary embolism, and deep venous thrombosis in patients with cancer as well as no cancer. All ages 1979â&#x20AC;&#x201C;1999 Diagnoses/ 100 hospitalizations

Number of patients Cancers Pancreas Brain Myeloproliferative, other lymphatic/ hematopoeitic Stomach Lymphoma, lymphosarcoma, reticulosarcoma Uterus Trachea, bronchus, and lung Esophagus Prostate Rectum, rectosigmoid junction, anus Kidney Colon Ovary Liver, gallbladder, intra- and extrahepatic ducts Leukemia Breast (female) Cervix Bladder Lip, oral cavity, pharynx Average incidences No cancer

VTE

PE

DVT

All discharges

VTE

PE

DVT

51,000 27,000 15,000

14,000 8,000 *

41,000 22,000 13,000

1,176,000 772,000 521,000

4.3 3.5 2.9

1.2 1.0 *

3.5 2.8 2.5

24,000 79,000

7,000 20,000

20,000 63,000

887,000 3,182,000

2.7 2.5

0.7 0.6

2.3 2.0

26,000 170,000 12,000 95,000 30,000

6,000 56,000 *

1,180,000 8,120,000 603,000 4,643,000 1,457,000

2.2 2.1 2.0 2.0 2.1

0.5 0.6 *

30,000 11,000

21,000 129,000 8,000 72,000 21,000

0.6 0.7

1.8 1.6 1.3 1.6 1.4

19,000 69,000 31,000 13,000

5,000 24,000 8,000 6,000

15,000 49,000 27,000 8,000

939,000 3,614,000 1,669,000 703,000

2.0 1.9 1.9 1.8

0.5 0.6 0.5 0.9

1.6 1.4 1.6 1.1

45,000 82,000 14,000 21,000 <5,000

10,000 22,000 *

2,655,000 4,932,000 875,000 2,011,000 849,000

1.7 1.7 1.6 1.0 <0.6

0.4 0.4 *

5,000 *

38,000 64,000 12,000 17,000 *

0.3 *

1.4 1.3 1.4 0.8 *

823,000 6,854,000

232,000 2,212,000

640,000 5,124,000

40,788,000 662,309,000

2.0 1.0

0.6 0.3

1.6 0.8

* Insufficient data. VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis. Reprinted from Stein et al. [15], with permission from Elsevier.

85


BLUK077-Stein

April 14, 2007

14:54

PART I

tumors, 7.8% also had VTE and 1.1% had PE (pulmonary embolism) [7]. The risk of VTE varies according to the type of cancer [5â&#x20AC;&#x201C;8]. Carcinoma of the pancreas has been associated with the greatest risk of VTE [9]. High rates of VTE have been reported among patients with solid neoplasms originating at several sites including the lung, ovary, brain, pancreas, stomach, kidney, colon [5â&#x20AC;&#x201C;7]. Patients with lymphoma, leukemia, and myeloproliferative syndromes also often have associated VTE [5, 6]. However, cancers originating at some sites, including head and neck and bladder, did not have a high incidence of associated VTE [5, 6]. A review of cancer patients who underwent surgery showed that they had twice the risk of postoperative deep venous thrombosis (DVT) as noncancer patients who underwent similar procedures [10].

Several mechanisms may be involved in the pathogenesis of thromboembolic events in patients with cancer [11]. These include (1) tumor cell procoagulants and/or cytokines, (2) tumor associated inflammatory cell procoagulants and/or cytokines, and (3) mediators of platelet adhesion or aggregation generated by tumor cells and/or tumor-associated inflammatory cells [11]. Stasis and endothelial damage may also be involved in the pathogenesis of thromboembolic events in patients with cancer [11]. The extent of cancer influences the risk of VTE [12]. Chemotherapy and radiation increase the risk of VTE [10, 12]. Risk factors may interact [10]. The risk of VTE increases with age [13] and age is also associated with malignancy. Among cancer patients undergoing surgery, advanced age, debility, prolonged and difficult surgery, and a lengthy and

(a)

4 3.5 3 2.5 2 1.5 1 0.5 0

Prevalence, risks, and prognosis of PE and DVT

Cancer patients

Noncancer patients 99

97

95

93

91

89

87

85

83

81

79

VTE in hospitalized cancer and noncancer patients (%)

86

(b)

2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Cancer patients

Noncancer patients 99

97

95

93

91

89

87

85

83

81

79

PE in hospitalized cancer and noncancer patients (%)

Year

(c)

4 Cancer patients

3.5 3 2.5 2 1.5 1

Noncancer patients

0.5 0

99

97

95

93

Year

91

89

87

85

83

81

79

DVT in hospitalized cancer and noncancer patients (%)

Year

Figure 17.1 (a) Incidence of venous thromboembolism (VTE) in patients hospitalized with cancer and those without cancer. From 1989 to 1999, there was a prominent increase in the incidence of VTE in patients discharged with cancer. The incidence of VTE in patients without cancer also increased but the slope was lower. (b) Incidence of pulmonary embolism (PE) in patients hospitalized with cancer and those without cancer. From 1989 to 1999, the incidence of PE in patients discharged with cancer increased. The incidence of PE in patients without cancer increased a little. (c) Incidence of deep venous thrombosis (DVT) in patients hospitalized with cancer and those without cancer. From 1989 to 1999, there was a prominent increase in the incidence of DVT in patients discharged with cancer. The incidence of DVT in patients without cancer also increased but the slope was lower. (Reprinted from Stein et al. [15], with permission from Elsevier.)


BLUK077-Stein

April 14, 2007

CHAPTER 17

14:54

87

Venous thromboembolism in cancer patients

twice the incidence of VTE in hospitalized patients who did not have any of these malignancies, 6,854,000 of 662,309,000 (1.0%) [15]. The highest incidence of VTE was in patients with carcinoma of the pancreas (4.3%) (Table 17.1). The lowest incidence of VTE was

complicated postoperative course add to the risk of DVT [14]. Among patients hospitalized from 1979 to 1999, with any of the malignancies listed in Table 17.1, 827,000 of 40,787,000 (2.0%) had VTE [15]. This was

(a)

4 3.5 3 2.5 2 1.5 1 0.5

Pancreas Brain Myeloprol Stomach Lymphoma Uterus Lung Esophagus Prostate Rectal Kidney Colon Ovary Liver Leukemia Breast Cervix Bladder

Relative risk of VTE in cancer patients

4.5

4.5 4 3.5 3 2.5 2 1.5 1 0.5 Pancreas Brain Myeloprol Stomach Lymphoma Uterus Lung Esophagus Prostate Rectal Kidney Colon Ovary Liver Leukemia Breast Cervix Bladder

Relative risk of PE in cancer patients

(b)

(c) 5 4 3 2 1 Cervix

Bladder

Breast

Liver

Leukemia

Colon

Ovary

Rectal

Kidney

Prostate

Lung

Esophagus

Uterus

Stomach

Lymphoma

Myeloprol

Brain

0 Pancreas

Relative risk of DVT in cancer patients

Figure 17.2 (a) Relative risks of venous thromboembolism (VTE) in patients hospitalized with cancer compared to those without cancer. The relative risk of VTE ranged from 1.02 to 4.34. (b) Relative risks of pulmonary embolism (PE) in patients hospitalized with cancer compared to those without cancer. The relative risk of PE ranged from 0.77 to 3.66. (c) Relative risks of deep venous thrombosis (DVT) in patients hospitalized with cancer compared to those without cancer. The relative risk of DVT ranged from 1.07 to 4.65. (Reprinted from Stein et al. [15], with permission from Elsevier.)


BLUK077-Stein

April 14, 2007

14:54

88

PART I

in patients with carcinoma of the bladder (1.0%) (Table 17.1). The incidence of VTE in patients with carcinoma of the lip, oral cavity or pharynx was lower, but data were insufficient to calculate the incidence. Pulmonary embolism among patients hospitalized with carcinoma of the pancreas occurred in 1.2% and in patients with carcinoma of the bladder in 0.3% [15] (Table 17.1). Patients hospitalized with malignancies, on average, had twice the incidence of PE as in those who did not have malignancies (0.6% versus 0.3%) (Table 17.1).

The incidence of DVT associated with patients hospitalized with various neoplasms are shown in Table 17.1. As in patients with VTE and PE, patients hospitalized with cancer had twice the incidence of DVT as those who did not have cancer (1.6% versus 0.8%) (Table 17.1). The incidence of VTE in hospitalized patients with malignant neoplasms began to sharply increase in 1989 [15] (Figure 17.1a). The incidence of VTE in hospitalized patients who did not have cancer also showed an increasing incidence beginning in 1989, but

Prevalence, risks, and prognosis of PE and DVT

Table 17.2 Venous thromboembolism, pulmonary embolism, and deep venous thrombosis in patients with cancer as well as no cancer. Age group 40â&#x20AC;&#x201C;59, 1979â&#x20AC;&#x201C;1999 Diagnoses/ Number of patients Cancers

VTE

PE

100 hospitalizations DVT

All discharges

VTE

PE

DVT

Pancreas

10,000

*

10,000

217,000

4.6

*

4.6

Brain

11,000 *

*

9,000 *

206,000

5.3 *

*

4.4 *

Myeloproliferative, other

*

49,000

*

lymphatic/ hematopoeitic Stomach Lymphoma, lymphosarcoma,

5,000

*

5,000

171,000

2.9

*

2.9

18,000

5,000

16,000

738,000

2.4

0.6

2.2

reticulosarcoma 6,000

*

5,000

318,000

1.9

*

1.6

52,000 *

13,000 *

43,000 *

2,057,000 157,000

2.5 *

0.6 *

2.1 *

Prostate

5,000

*

*

332,000

1.5

*

*

Rectum, rectosigmoid

5,000

*

*

291,000

1.7

*

*

Uterus Trachea, bronchus, and lung Esophagus

junction, anus *

*

*

232,000

*

*

*

Colon

11,000

*

8,000

612,000

1.8

*

1.3

Ovary

9,000 *

*

8,000 *

572,000

1.6 *

*

1.4 *

8,000

5,000

7,000

358,000

2.2

1.4

2.0

26,000

6,000 *

21,000

1,734,000

1.5

1.2

347,000

*

5,000 *

247,000

1.4 *

0.3 *

*

*

293,000

Kidney

Liver, gallbladder, intra- and

*

149,000

*

extrahepatic ducts Leukemia Breast (female) Cervix Bladder Lip, oral cavity, pharynx Average incidences No cancer

5,000 * *

*

1.4 *

*

*

*

171,000

29,000

137,000

9,080,000

1.9

0.3

1.5

1,693,000

504,000

1,295,000

131,893,000

1.3

0.4

1.0

* Insufficient data. VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis. Reprinted from Stein et al. [15], with permission from Elsevier.


BLUK077-Stein

April 14, 2007

CHAPTER 17

14:54

89

Venous thromboembolism in cancer patients

the increasing incidence was not as high. The incidence of PE among patients with cancer increased in the 1990s (Figure 17.1b), but the increasing incidence was not as high as with DVT (Figure 17.1c) or VTE (Figure 17.1a). The relative risk of VTE, PE, and DVT among hospitalized patients with various malignancies, compared to patients who did not have any of these malignancies, is shown in Figures 17.2a–17.2c [15]. The highest relative risk of VTE, 4.3, was among hospitalized patients with carcinoma of the pancreas. The lowest relative risk, 1.0, was among patients with carcinoma

of the bladder. The relative risks of PE and DVT paralleled the relative risks of VTE (Figures 17.2b and 17.2c). Among patients aged 40–59 years who were hospitalized with cancer, average incidence of VTE and DVT were 46 to 50% higher than in patients who did not have these types of cancer but the incidence of PE was not higher [15] (Table 17.2). Among patients aged 60–79 years who were hospitalized with cancer, average incidence of VTE and DVT were only 29 to 42% higher than in patients who did not have these cancers (Table 17.3).

Table 17.3 Venous thromboembolism, pulmonary embolism and deep venous thrombosis in patients with cancer as well as no cancer. Age group 60–79, 1979–1999 Diagnoses/ Number of patients Cancers

VTE

PE

100 hospitalizations DVT

All discharges

VTE

PE

DVT

Pancreas

37,000

724,000

5.1

9,000

245,000

4.9

1.7 *

3.9

12,000

12,000 *

28,000

Brain

9,000

*

8,000

263,000

3.4

*

3.0

Myeloproliferative, other

3.7

lymphatic/ hematopoeitic Stomach

13,000

*

11,000

494,000

2.6

*

2.2

Lymphoma, lymphosarcoma,

43,000

12,000

33,000

1,432,000

3.0

0.8

2.3

16,000

5,000

13,000

697,000

2.3

0.7

1.9

105,000

76,000 *

5,159,000

2.0

7,000

38,000 *

359,000

1.9

0.7 *

1.5 *

Prostate

66,000

21,000

51,000

2,992,000

2.2

0.7

1.7

Rectum, rectosigmoid

19,000

7,000

13,000

846,000

2.2

0.8

1.5

reticulosarcoma Uterus Trachea, bronchus, and lung Esophagus

junction, anus Kidney

12,000

*

9,000

507,000

2.4

*

1.8

Colon

44,000

32,000

2,087,000

2.1

17,000

15,000

842,000

2.0

0.8 *

1.5

Ovary

16,000 *

6,000

*

4,000

383,000

1.6

*

1.0

Liver, gallbladder, intra- and

1.8

extrahepatic ducts Leukemia

25,000

5,000

21,000

1,060,000

2.4

0.5

2.0

Breast (female)

42,000

13,000 *

32,000

2,300,000

1.8

1.4

4,000

263,000

1.9

0.6 *

*

11,000 *

1,217,000

*

425,000

1.2 *

*

0.9 *

Cervix Bladder Lip, oral cavity, pharynx Average incidences No cancer

5,000 14,000 *

*

1.5

492,000

129,000

370,000

22,033,704

2.2

0.6

1.7

3,018,000

1,038,000

2,151,000

174,352,000

1.7

0.6

1.2

* Insufficient data. VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis. Reprinted from Stein et al. [15], with permission from Elsevier.


BLUK077-Stein

April 14, 2007

14:54

90

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 17.4 Venous thromboembolism, pulmonary embolism, and deep venous thrombosis according to sex in patients with cancer as well as no cancer. Incidence (%) VTE

PE

DVT

Cancers

Male

Female

Male

Female

Male

Female

Pancreas

4.4

4.3 3.0

1.3 *

3.5

3.9

1.2 *

3.5

Brain

3.4

2.2

Myeloproliferative, other lymphatic/hematopoeitic

2.8

3.1

*

*

2.3

2.6

Stomach

3.0

2.1

*

*

2.4

1.8

Lymphoma, lymphosarcoma, reticulosarcoma

2.4

2.6

0.8

0.5

1.8

2.2

Trachea, bronchus, and lung

2.1 2.2

2.1 *

0.7 *

0.7 *

1.5

Esophagus

1.6

1.7 *

Rectum, rectosigmoid junction, anus

1.9

2.2 2.2

0.7 *

1.5

1.9

0.7 *

1.4

Renal

1.5

1.7

Colon

2.0 2.6

0.7 *

0.7

Liver/gallbladder/extrahepatic duct

1.7 *

1.2 *

1.5

Leukemia

1.3

2.2

*

Bladder

1.1 *

*

0.5 *

Lip, oral cavity, pharynx

1.1 *

*

*

0.8 *

Average incidences

2.1

2.1

0.3

0.4

1.6

1.6

No cancer

1.1

1.0

0.4

0.3

0.8

0.8

1.4

1.0

1.5 1.9 * *

* Insufficient data. VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis. Reprinted from Stein et al. [15], with permission from Elsevier.

On average, the incidences of VTE, PE, and DVT in patients with cancer were similar in men and women, but differences were shown with various malignancies [15] (Table 17.4). The incidence of VTE, PE, and DVT according to race, on average, were similar in African Americans and whites (Table 17.5). Differences were observed with various malignancies. The rates we reported [15] were the same order of magnitude as reported by Levitan et al. [6]. In patients with carcinoma of pancreas, the relative risk of VTE was 4.3 times that of noncancer patients. The rates of VTE, PE, and DVT showed an increase starting in the late 1980s. The incidences of DVT and PE increase exponentially with age [13]. However, the relative risk of VTE among patients aged 40–59 years was paradoxically higher than in older patients. This reflected a disproportionately higher incidence of VTE in the older noncancer patients. Cancer age did not greatly increase the incidence of VTE in older patients.

The cancers that we evaluated were shown by others to have a higher incidence of VTE than in patients without cancer [5–7]. The extent of malignancy, invasion, metastasis, chemotherapy, radiotherapy, and surgery, which may increase the incidence of associated VTE, was not available on discharge codes.

Pulmonary embolism as a cause of death in patients who died with cancer Among 506 autopsies of patients in whom 96% had “some sort of neoplasm,” 35 (7%) died from PE [16]. Numerous other investigations describe the association of PE with cancer, but do not describe the frequency of fatal PE in patients who died with cancer. Among patients who died from PE, 22.9% had cancer [17]. Among patients with cancer who died from 1980 to 1998, the estimated average frequency of PE as the cause of death, after adjustment for the


BLUK077-Stein

April 14, 2007

CHAPTER 17

14:54

91

Venous thromboembolism in cancer patients

Table 17.5 Venous thromboembolism, pulmonary embolism, and deep venous thrombosis according to race in patients with cancer as well as no cancer. Incidence (%) VTE

PE

DVT

Cancers

White

Black

White

Black

White

Black 2.9

Pancreas

4.4

4.2

1.2

1.7

3.5

Brain

3.8

5.1

1.0

3.2

3.2

2.0

Myeloproliferative, other lymphatic/hematopoeitic

3.0

3.7

0.7

0.4

2.6

3.3

Stomach

2.9

2.6

0.9

0.6

2.4

2.1

Lymphoma, lymphosarcoma, reticulosarcoma

2.5 2.2

0.6 *

0.5 *

2.1

Uterus

2.3 *

2.1 *

Trachea, bronchus, and lung

2.1

Esophagus

2.2

1.8 *

0.7 *

0.7 *

1.9 1.6

1.3 *

1.5

Prostate

2.2

1.6

0.6

0.6

1.7

1.3

Rectum, rectosigmoid junction, anus

1.9

3.3

0.7

1.2

1.3

2.1

Kidney

1.9

1.7

0.5

0.5

1.5

1.3

Colon

1.8

2.3

0.6

0.7

1.3

1.7

Ovary

1.9 1.8

0.5 *

0.5 *

1.6

Liver, gallbladder, intra- and extrahepatic ducts

2.0 *

1.6 *

1.2

Leukemia

1.8

1.8

0.4

0.6

1.6

1.2

Breast (female)

1.7 1.8

0.4 *

0.4 *

1.3

Cervix

1.7 *

1.2 *

Bladder

*

*

*

*

*

*

0.8 *

*

Lip, oral cavity, pharynx

1.0 *

Average incidences

2.2

1.8

0.5

0.6

1.7

1.3

No cancer

1.1

0.9

0.4

0.3

0.8

0.7

1.6

*

* Insufficient data. VTE, venous thromboembolism; PE, pulmonary embolism; DVT, deep venous thrombosis. Reprinted from Stein et al. [15], with permission from Elsevier.

myeloproliferative disease and lowest in patients who died with cancer of the liver, gallbladder, intra- and extrahepatic ducts. The rate of fatal PE among patients who died with cancer decreased from 1980 to 1998 (Figure 17.3). This

0.6 0.5 0.4 0.3 0.2 0.1 0 98

96

94

92

Year

90

88

86

84

82

80

Figure 17.3 Frequency of fatal pulmonary embolism (PE) among patients who died with cancer from 1980 to 1998. The frequency declined over the 19-year period of study. (Reprinted from Stein et al. [18], with permission from Elsevier.)

Deaths from PE in patients who died with cancer (%)

inaccuracy of death certificates, was between 0.60 and 1.05% [18]. However, these values appear low compared to the prevalence of large or fatal PE in all patients at autopsy, about 4% (Chapter 1). The incidence of fatal PE was highest among those who died with


BLUK077-Stein

April 14, 2007

14:54

92

PART I

presumably reflects an improvement in achieving an early diagnosis and more vigorous prophylaxis over these 19 years.

8 Thodiyil PA, Kakkar AK. Variation in relative risk of venous thromboembolism in different cancers. Thomb Haemost 2002; 87: 1076–1077. 9 Rickles FR, Edwards RL. Activation of blood coagulation in cancer: Trousseau’s syndrome revisited. Blood 1983; 62: 14–31. 10 Bergqvist D. Venous thromboembolism and cancer: prevention of VTE. Thromb Res 2001; 102: V209–V213. 11 Rickles FR, Levine M, Edwards RL. Hemostatic alterations in cancer patients. Cancer Metast Rev 1992; 11: 237–248. 12 Lee AYY, Levine MN. Venous thromboembolism and cancer: risks and outcomes. Circulation 2003; 107: I-17–I-21. 13 Stein PD, Hull RD, Kayali F, Ghali WA, Alshab AK, Olson RE. Venous thromboembolism according to age: the impact of an aging population. Arch Intern Med 2004; 164: 2260–2265. 14 Gallus AS. Prevention of post-operative deep leg vein thrombosis in patients with cancer. Thromb Haemost 1997; 78: 126–132. 15 Stein PD, Beemath A, Meyers FA, Skaf E, Sanchez J, Olson RE. Incidence of venous thromboembolism in patients hospitalized with cancer. Am J Med 2006; 119: 60–68. 16 Ambrus JL, Ambrus CM, Mink IB, Pickren JW. Causes of death in cancer patients. J Med 1975; 6: 61–64. 17 Horlander KT, Mannino DM, Leeper KV. Pulmonary embolism mortality in the United States, 1979–1998: an analysis using multiple-cause mortality data. Arch Intern Med 2003; 163: 1711–1717. 18 Stein PD, Beemath A, Meyers FA, Kayali F, Skaf E, Olson RE. Pulmonary embolism as a cause of death in patients who died with cancer. Am J Med 2006; 119: 163–165.

References 1 Arklel YS. Thrombosis and cancer. Semin Oncol 2000; 27: 362–374. 2 Donati MB. Cancer and thrombosis. Haemostatis 1994; 24: 128–131. 3 Trousseau A. Phlegmassia alba dolens. In: Clinique Medicale de l’Hotel-Dieu de Paris, Vol. 3. New Sydenham Society, London, 1865: 94. Quoted by Rickles and Edwards in Reference [9]. 4 Heit JA, Silverstein MD, Mohr DN, Petterson TM, O’Fallon WM, Melton LJ, III. Risk factors for deep vein thrombosis and pulmonary embolism. A populationbased case–control study. Arch Intern Med 2000; 160: 809– 815. 5 Blom JW, Doggen CJM, Osanto S, Rosendaal FR. Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA 2005; 293: 715–722. 6 Levitan N, Dowlati A, Remick SC et al. Rates of initial and recurrent thromboembolic diseases among patients with malignancy versus those without malignancy. Risk analysis using Medicare claims data. Medicine 1999; 78: 285–291. 7 Sallah S, Wan JY, Nguyen NP. Venous thrombosis in patients with solid tumors: determination of frequency characteristics. Thromb Haemost 2002; 87: 575–579.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

March 28, 2007

18

15:12

CHAPTER 18

Venous thromboembolism in patients with heart disease

Patients with heart failure Heart failure (HF) is considered a major risk factor for venous thromboembolism (VTE), defined as pulmonary embolism (PE) and/or deep venous thrombosis (DVT) [1–6]. In patients with advanced HF, thrombophlebitis of the pelvic or leg veins is a common complication and this may also be a source of PE [7]. Based on data from the National Hospital Discharge Survey (NHDS) [8], among hospitalized patients with HF, PE was diagnosed in 0.73%, DVT in 1.03%, and

VTE in 1.63% [9] (Table 18.1). Among hospitalized patients who were not diagnosed with HF, PE was diagnosed in 0.34%, DVT was diagnosed in 0.85%, and VTE in 1.11%. The prevalence of PE in HF was similar to findings reported in smaller investigations that used defined criteria for HF [10, 11]. Others showed that HF patients with lower ejection fractions had a higher risk of thromboembolic events [10, 12]. The reported frequency of PE in patients with HF has ranged widely from 0.9 to 39% of patients [3, 7, 10, 12–15]. At autopsy the incidence of PE in patients with HF ranged

Table 18.1 Prevalence and relative risk of pulmonary embolism, deep venous thrombosis, and venous thromboembolism in hospitalized patients according to age groups. n/N (%)* Age groups (yrs)

HF

No HF

HF vs. No HF(95% CI)

<40

13,000/1,129,000 (1.15)

333,000/339,063,000 (0.10)

11.72 (11.52–11.93)

40–59

65,000/6,410,000 (1.01)

655,000/164,010,000 (0.40)

2.54 (2.52–2.56)

60–79

236,000/28,745,000 (0.82)

1,229,000/205,140,000 (0.60)

1.37 (1.36–1.38)

>80

157,000/22,590,000 (0.69)

404,000/74,491,000 (0.54)

1.28 (1.27–1.29)

All ages

431,000/58,873,000 (0.73)

2,662,000/782,704,000 (0.34)

2.15 (2.15–2.16)

<40

19,000/1,129,000 (1.68)

1,045,000/339,063,000 (0.31)

5.46 (5.38–5.54)

40–59

85,000/6,410,000 (1.33)

1,762,000/164,010,000 (1.07)

1.23 (1.23–1.24)

60–79

317,000/28,745,000 (1.10)

2,833,000/205,140,000 (1.38)

0.80 (0.80–0.80)

>80

281,000/22,590,000 (1.24)

948,000/74,491,000 (1.27)

0.98 (0.97–0.98)

All ages

607,000/58,873,000 (1.03)

6,683,000/782,704,000 (0.85)

1.21 (1.20–1.21)

<40

30,000/1,129,000 (2.66)

1,304,000/339,063,000 (0.38)

6.91 (9.83–6.99)

40–59

139,000/6,410,000 (2.17)

2,227,000/164,010,000 (1.36)

1.60 (1.59–1.61)

60–79

509,000/28,745,000 (1.77)

3,749,000/205,140,000 (1.83)

0.97 (0.97–0.97)

>80

405,000/22,590,000 (1.79)

1,256,000/74,491,000 (1.69)

1.06 (1.06–1.07)

All ages

960,000/58,873,000 (1.63)

8,660,000/782,704,000 (1.11)

1.47 (1.47–1.48)

Pulmonary embolism

Deep venous thrombosis

Venous thromboembolism

* 95% confidence intervals are all ≤0.04%. HF, heart failure; yrs, years; CI, confidence interval. Reproduced from Beemath et al. [9], with permission from Elsevier.

93


BLUK077-Stein

March 28, 2007

15:12

94

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 18.2 Prevalence and relative risks of pulmonary embolism and deep venous thrombosis in hospitalized patients according to sex and race. Pulmonary embolism Prevalence (%)* HF

No HF

Deep venous thrombosis Prevalence (%)*

Relative risk HF vs. No HF

HF

No HF

Relative risk HF vs. No HF

Sex Male

0.69

0.38

1.80 (1.79–1.80)

0.89

0.90

0.98 (0.98–0.99)

Female

0.77

0.31

2.45 (2.44–2.46)

1.14

0.82

1.39 (1.39–1.40)

Race White

0.75

0.36

2.10 (2.09–2.11)

1.03

0.89

1.15 (1.15–1.15)

African American

0.85

0.30

2.82 (2.79–2.84)

1.13

0.79

1.44 (1.43–1.45)

* 95% confidence intervals are all ≤0.02%. HF, heart failure. Reproduced from Beemath et al. [9], with permission from Elsevier.

DVT and PE in patients hospitalized with heart failure (%)

from 28 to 48% [14, 16, 17]. The reported frequency of DVT in patients with HF also ranged widely from 10 to 59% [2, 3, 5]. The relative risk of PE in patients with HF was highest in patients <40 years of age (relative risk 11.72) and the relative risk for DVT was 5.46 [9] (Table 18.1). In patients aged 40–59 years, the relative risk for PE was 2.54 and for DVT 1.23. With older age groups the relative risks for both PE and DVT decreased (Table 18.1). With increasing age and its accompanying risk factors for PE and DVT, other risk factors balance or outweigh the risk of HF alone. In older patients therefore, the higher relative risk for PE or DVT is thereby reduced or eliminated. The high relative risk of PE and DVT in younger adults is readily explained, but the somewhat higher rates of PE and DVT in younger adults are not explained. This observation is not in accordance with

previous findings of an increased risk of PE and DVT with age [18]. The relative risk of PE was higher in patients with HF compared to those patients with no HF among both women and men and African Americans and whites [9] (Table 18.2). The relative risk of DVT was higher in patients with HF compared with no HF among women but not in men. Both African Americans and whites with HF had higher relative risks for DVT compared with those who did not have HF (Table 18.2). The rate of PE among patients hospitalized with HF decreased from 1.38% in 1979–1981 to 0.61% in 2000– 2003 [9] (Figure 18.1). The rate of diagnosis of DVT in patients hospitalized with HF decreased from 0.89% in 1979–1981 to 0.71% in 1988–1990 and then increased to 1.35% in 2000–2003 (Figure 18.1).

1.6 1.4

DVT

1.2 1 0.8 0.6

PE

0.4 '79−81 '82−84 '85−87 '88−90 '91−93 '94−96 '97−99 '00−03 Years

Figure 18.1 Rates of pulmonary embolism (PE) and deep venous thrombosis (DVT) in hospitalized patients with heart failure (HF). Data were averaged over 3-year periods from 1979 to 1999 and over the 4-year period 2000–2003. The rate of PE decreased from 1979–1981 to 2000–2003. The rate of DVT decreased from 1979–1981 to 1988–1990. The rate of DVT then increased from 1988–1990 to 2000–2003. (Reproduced from Beemath et al. [9], with permission from Elsevier.)


March 28, 2007

CHAPTER 18

15:12

95

VTE in patients with heart failure

Figure 18.2 Frequency of fatal pulmonary embolism (PE) among adults who died with heart failure (HF) from 1980 to 1998. The frequency declined over the 19-year period of study. (Reproduced from Beemath et al. [27], with permission from Elsevier.)

Deaths from PE in adults who died with heart failure (%)

BLUK077-Stein

6 5 4 3 2 1 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 Year

Figure 18.3 Frequency of fatal pulmonary embolism (PE) among adults who died with heart failure (HF) among the different age groups. (Reproduced from Beemath et al. [23], with permission from Elsevier.)

Deaths from PE in adults who died with heart failure (%)

Heart failure appears to be a stronger risk factor in woman. Dries et al. [10] reported a significant difference between the sexes in the distribution of thromboembolic events in patients with HF. Among women with HF, 24% had PE whereas among men with HF 14% had PE [10]. African Americans with HF also had a somewhat higher risk for PE and DVT than whites. In 2004, it was recommended that patients hospitalized with HF should receive antithrombotic prophylaxis [19]. Roberts et al. reported PE as the cause of death in 12 of 139 (9%) autopsied patients who died with HF [15]. Goldhaber et al. reported PE as the cause of death at autopsy in 13 of 41 (32%) patients with HF [16]. In other studies, PE was found in the wide range of 0.4–50% of autopsied patients who died with HF, but whether PE caused or contributed to these deaths was not stated [11, 14, 20–22]. We investigated the rate of death from PE in patients who died with HF from 1980 to 1998 based on data from death certificates, as listed by the United States Bureau of the Census [23].

Among adults with HF who died over a 19-year period of study, PE was the listed cause of death in 20,387 of 755,807 (2.7%) [23]. Assuming a sensitivity of 26.7% for the death certificate diagnosis of fatal PE [24], the frequency of fatal PE among patients who died with HF over the 19-year period of study would be 10.1%. The frequency of death from PE in patients who died with HF decreased from 5.0% in 1980 to 1.6% in 1998 (Figure 18.2). These rates were not adjusted for the 26.7% sensitivity of PE diagnoses on death certificates. We assume that whatever inaccuracy exists in the death certificates was constant throughout the 19-year period of study. The trend of a decreasing rate of death from PE in adults who died with HF, therefore, would appear to be accurate. The decreasing rate of fatal PE among patients who died with HF during the 19-year period of study presumably reflects an improvement in achieving an early diagnosis and more vigorous prophylaxis. The risk of PE in patients with HF is twice the risk of PE observed in patients who do not have

12 10 8 6 4 2 0

20−34

35−44 45−54

55−64

65−74

Age group (years)

75−84

85−99


BLUK077-Stein

March 28, 2007

15:12

96

HF [9]. Antithrombotic prophylaxis is recommended in such patients [19]. The rate of fatal PE in adults who died with HF decreased with age with the highest rate in the age group 20–34 years (10.6%) (Figure 18.3). The reason for this is not clear.

PART I

9

10

Patients with any heart disease Among 1032 autopsies of patients with heart disease, 231 patients (24.4%) had PE and 100 (9.7%) had massive PE, considered to be the cause of death [25]. Strikingly, the prevalence of PE as a cause of death was particularly high in young patients (<10 years old) with heart disease [25]. Among 59 patients who died from PE and in whom the cardiac pathology was described, 34 (58%) had right ventricular dilatation, and 8 (14%) biventricular dilatation [25]. An antemortem suspicion of PE was raised in only 18% of patients with heart disease who died from PE [25].

References 1 Shively BK. Deep venous thrombosis prophylaxis in patients with heart disease. Curr Cardiol Rep 2001; 3(1): 56–62. 2 Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000; 160(22): 3415–3420. 3 Anderson FA, Jr, Wheeler HB, Goldberg RJ et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med 1991; 151(5): 933–938. 4 Isnard R, Komajda M. Thromboembolism in heart failure, old ideas and new challenges. Eur J Heart Fail 2001; 3(3): 265–269. 5 Cogo A, Bernardi E, Prandoni P et al. Acquired risk factors for deep-vein thrombosis in symptomatic outpatients. Arch Intern Med 1994; 154(2): 164–168. 6 Jafri SM, Ozawa T, Mammen E, Levine TB, Johnson C, Goldstein S. Platelet function, thrombin and fibrinolytic activity in patients with heart failure. Eur Heart J 1993; 14(2): 205–212. 7 Segal JP, Harvey WP, Gurel T. Diagnosis and treatment of primary myocardial disease. Circulation 1965; 32: 837– 844. 8 US Department of Health and Human Services, Public Health Service, National Center for Health Statistics

11

12

13

14 15

16

17 18

19

20

21

22

Prevalence, risks, and prognosis of PE and DVT

National Hospital Discharge Survey 1979–1999 Multiyear Public-Use Data File Documentation. Available at: http://www.cdc.gov/nchs/about/major/hdasd/nhds.htm. Last accessed Jan/ 20/, 2006. Beemath A, Stein PD, Skaf E, Al Sibae MR et al. Risk of venous thromboembolism in patients hospitalized with heart failure. Am J Cardiol. 2006; 98(6): 793–795. Dries DL, Rosenberg YD, Waclawiw MA, Domanski MJ. Ejection fraction and risk of thromboembolic events in patients with systolic dysfunction and sinus rhythm: evidence for gender differences in the studies of left ventricular dysfunction trials. J Am Coll Cardiol 1997; 29(5): 1074–1080. Al-Khadra AS, Salem DN, Rand WM, Udelson JE, Smith JJ, Konstam MA. Warfarin anticoagulation and survival: a cohort analysis from the Studies of Left Ventricular Dysfunction. J Am Coll Cardiol 1998; 31(4): 749–753. Kyrle PA, Korninger C, Gossinger H et al. Prevention of arterial and pulmonary embolism by oral anticoagulants in patients with dilated cardiomyopathy. Thromb Haemost 1985; 54(2): 521–523. Dunkman WB, Johnson GR, Carson PE, Bhat G, Farrell L, Cohn JN, for The V-HeFT VA Cooperative Studies Group. Incidence of thromboembolic events in congestive heart failure. Circulation 1993; 87(6 suppl): VI94– VI101. Kinsey D, White P. Fever in congestive heart failure. Arch Intern Med 1940; 65: 163–170. Roberts WC, Siegel RJ, McManus BM. Idiopathic dilated cardiomyopathy: analysis of 152 necropsy patients. Am J Cardiol 1987; 60: 1340–1355. Goldhaber SZ, Savage DD, Garrison RJ et al. Risk factors for pulmonary embolism. The Framingham Study. Am J Med 1983; 74(6): 1023–1028. Greenstein J. Thrombosis and pulmonary embolism. South African Med J 1945; 19: 350–377. Stein PD, Hull RD, Kayali F, Ghali WA, Alshab AK, Olson RE. Venous thromboembolism according to age: the impact of an aging population. Arch Intern Med 2004; 164(20): 2260–2265. Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126(3 suppl): 338S–400S. Harvey WP, Finch CA. Dicumarol prophylaxis of thromboembolic disease in congestive heart failure. N Engl J Med 1950; 242: 208–211. Anderson GM, Hull E. The effect of dicumarol upon the mortality and incidence of thromboembolic complications in congestive heart failure. Am Heart J 1950; 39: 697–702. Spodick DH, Littmann D. Idiopathic myocardial hypertrophy. Am J Cardiol 1958; 1: 610–623.


BLUK077-Stein

March 28, 2007

CHAPTER 18

15:12

VTE in patients with heart failure

23 Beemath A, Skaf E, Stein PD. Pulmonary embolism as a cause of death in patients who died with heart failure. Am J Cardiol 2006; 98: 1073–1075. 24 Attems J, Arbes S, Bohm G, Bohmer F, Lintner F. The clinical diagnostic accuracy rate regarding the immediate cause of death in a hospitalized geriatric population; an

97

autopsy study of 1594 patients. Wien Med Wochenschr 2004; 154: 159–162. 25 Pulido T, Aranda A, Zevallos MA et al. Pulmonary embolism as a cause of death in patients with heart disease. An autopsy study. Chest 2006; 129: 1282– 1287.


BLUK077-Stein

March 10, 2007

19

14:23

CHAPTER 19

Venous thromboembolism in patients with ischemic and hemorrhagic stroke

Patients with stroke are at particular risk of developing deep venous thrombosis (DVT) and pulmonary embolism (PE) because of limb paralysis, prolonged bed rest, and increased prothrombotic activity [1]. Deep venous thrombosis in the paralyzed legs of patients with stroke was reported as early as 1810 by Ferriar and again by Lobstein in 1833 [2]. Among 14,109,000 patients hospitalized with ischemic stroke, PE occurred in 72,000 (0.51%), DVT occurred in 104,000 (0.74%), and VTE (venous thromboembolism) occurred in 165,000 (1.17%) [3] (Figure 19.1). Among 1,606,000 patients hospitalized with hemorrhagic stroke, rates were higher. Pulmonary embolism occurred in 11,000 (0.68%), DVT occurred in 22,000 (1.37%), and VTE occurred in 31,000 (1.93%)

[3] (Figure 19.1). The rate of VTE in hospitalized patients with ischemic stroke and with hemorrhagic stroke did not change significantly over the 25-year period of observation [3] (Figure 19.2). The higher rate of PE, DVT, and VTE among patients with hemorrhagic stroke compared with patients with ischemic stroke may represent more frequent use of antithrombotic prophylaxis in patients with ischemic stroke although treatment was not known [3]. Since 1980, with the use of modern diagnostic techniques and general awareness of the importance of antithrombotic prophylaxis, only a few previous case series have been reported, the largest of which had 607 patients [4â&#x20AC;&#x201C;10]. In most, antithrombotic prophylaxis was either not given or not described [4â&#x20AC;&#x201C;7, 9].

2 1.93

1.8 1.6 1.4 PE, DVT or VTE 1.2 (%) 1 0.8 0.6 0.4 0.2 0

1.37

0.68 0.51

1.17 0.74 Hemorrhagic stroke Ischemic stroke

PE DVT VTE Figure 19.1 Rates of pulmonary embolism (PE), deep venous thrombosis (DVT), and venous thromboembolism (VTE) in hospitalized patients with ischemic and

98

hemorrhagic stroke. Data were averaged from 1979 to 2003. (Reprinted from Skaf et al. [3], with permission from Elsevier.)


March 10, 2007

99

VTE in patients with stroke

Figure 19.2 Trends in the rates of venous thromboembolism (VTE) in hospitalized patients with ischemic and hemorrhagic stroke. Data were pooled in 5-year intervals. The rate of VTE in patients with ischemic stroke and in patients with hemorrhagic stroke did not change over the 25-year period of observation. (Reprinted from Skaf et al. [3], with permission from Elsevier.)

2.8 2.4

Hemorrhagic stroke

2 1.6 1.2 Ischemic stroke 0.8 1979−84

The prevalence of PE in these patients with stroke ranged from 0 to 4%. The prevalence of PE was 0.8% among 245 patients who received antithrombotic prophylaxis [8]. The International Stroke Trial Collaboration Group showed that among patients with ischemic stroke who received low-dose heparin (5000 IU twice daily), the rate of PE within 14 days was 33 of 4860 (0.7%) and among those who received medium-dose heparin (12,500 IU twice daily), the rate of PE within 14 days was 20 of 4856 (0.4%) [11]. Some received aspirin 300 mg/day in addition. Among all patients treated with aspirin, heparin, both, or neither, the rate of PE ranged from 0.5 to 0.8% [11]. Most investigations of antithrombotic therapy for stroke [11] or to prevent DVT among stroke patients [10, 12] excluded patients with hemorrhagic stroke. Clinically apparent DVT was reported in 1.7–5.0% of patients with stroke [4–8]. Subclinical DVT occurred in 28–73% of patients with stroke, usually in the paralyzed limb [10, 12, 13]. A high proportion of patients with DVT also have subclinical PE [14]. Although deaths within a few days of stroke are usually due to the direct consequence of brain damage, those occurring over the following weeks are mainly due to potentially preventable problems including PE [4]. In the experience of some, PE is the leading cause of death during the 2–4 weeks after onset of stroke [15, 16], yet PE is one of the preventable causes of death after stroke [16–18]. Prior to the general use of antithrombotic prophylaxis (1941–1952), 26% of immediate survivors of stroke who subsequently died, died of PE [19]. More recently, among patients with stroke who died and had autopsies, PE was the cause of death

1985−89

1990−94

1995−99

2000−03

Year

in 8–16% [15, 17, 18, 20]. Others, who described the rate of PE as the cause of death in patients who died with stroke, based on death certificates and autopsy between 1961 and 1984, reported rates of 1.3–5.9% [16, 21–25]. In 1997, the International Stroke Trial Collaborative Group showed PE as a cause of death in 75 of 1781 (4.2%) of patients with ischemic stroke who died [11]. Among patients with ischemic stroke who died from 1980 to 1998, PE was the listed cause of death on death certificates in 11,101 of 2,000,963 (0.55%) [26]. Adjusted rates of fatal PE in stroke, based on an assumed sensitivity for fatal PE on death certificates of 26.7– 37.2%, were 1.5–2.1% [26]. Death rates from PE among patients who died with ischemic stroke decreased from 1980 to 1998 (Figure 19.3) [26]. The uncorrected death rate from PE

Fatal PE among stroke deaths (%)

CHAPTER 19

14:23

VTE in hospitalized patients with ischemic or hemorrhagic stroke (%)

BLUK077-Stein

0.7 0.6 0.5 0.4 '80 '82 '84 '86 '88 '90 '92 '94 '96 '98 Year

Figure 19.3 Proportion of deaths from pulmonary embolism (PE) in patients who died with ischemic stroke. The proportion of deaths from PE with ischemic stroke decreased over the 19-year period of observation. (Reprinted from Skaf et al. [26], with permission from Elsevier.)


BLUK077-Stein

March 10, 2007

14:23

100

PART I

among patients with ischemic stroke, 0.55%, was similar to the average uncorrected death certificate rate from PE in the general population (0.45%) [27]. The decreasing proportion of deaths from PE among patients with ischemic stroke who died suggests an increased and effective use of antithrombotic prophylaxis in these patients.

13 Warlow C, Ogston D, Douglas AS. Deep venous thrombosis of the legs after strokes. BMJ 1976; 1: 1178– 1181. 14 Moser KM, Fedullo PF, LittleJohn JK, Crawford R. Frequent asymptomatic pulmonary embolism in patients with deep venous thrombosis. JAMA 1994; 271: 223– 225. 15 Viitanen M, Winblad B, Asplund K. Autopsy-verified causes of death after stroke. Acta Med Scand 1987; 222: 401–408. 16 Silver FL, Norris JW, Lewis AJ, Hachinski VC. Early mortality following stroke: a prospective review. Stroke 1984; 15: 492–496. 17 Brown M, Glassenberg M. Mortality factors in patients with acute stroke. JAMA 1973; 224: 1493–1495. 18 Bounds JV, Wiebers DO, Whisnant JP, Okazaki H. Mechanisms and timing of deaths from cerebral infarction. Stroke 1981; 12: 474–477. 19 Marquardsen J. The natural history of acute cerebrovascular disease: a retrospective study of 769 patients. Acta Neurol Scand 1969; 45: 9–88. 20 Ulbrich J. Woran sterben die apoplektiker. Ther Umsch 1981; 38: 703–708. 21 Marshall J, Kaeser AC. Survival after non-haemorrhagic cerebrovascular accidents. A prospective study. BMJ 1961; 2: 73–77. 22 Baker RN, Schwartz WS, Ramseyer JC. Prognosis among survivors of ischemic stroke. Neurology 1968; 18: 933– 941. 23 Marquardsen J. The natural history of acute cerebrovascular disease. Acta Neurol Scand 1969; 45: 131–137. 24 Matsumoto N, Whisnant JP, Kurland LT, Okazaki H. Natural history of stroke in Rochester, Minnesota, 1955 through 1969: an extension of a previous study, 1945 through 1954. Stroke 1973; 4: 20–29. 25 Miah K, von Arbin M, Britton M, de Faire U, Helmers C, Maasing R. Prognosis in acute stroke with special reference to some cardiac factors. J Chronic Dis 1983; 36(3): 279–288. 26 Skaf E, Stein PD, Beemath A, Sanchez J, Olson RE. Fatal pulmonary embolism and stroke. Am J Cardiol 2006; 97: 1776–1777. 27 Horlander KT, Mannino DM, Leeper KV. Pulmonary embolism mortality in the United States, 1979–1998. An analysis using multiple-cause mortality data. Arch Intern Med 2003; 163: 1711–1717.

References 1 Harvey RL. Prevention of venous thromboembolism after stroke. Topics Stroke Rehab 2003; 10: 61–69. 2 Lobstein JF. Traite d’ Anatomie Pathologique, Vol 2. Levranle FG, Paris, 1833: 610. Quoted by Warlow et al. in Reference [13]. 3 Skaf E, Stein PD, Beemath A, Sanchez J, Bustamante MA, Olson RE. Venous thromboembolism in patients with ischemic and hemorrhagic stroke. Am J Cardiol 2005; 96: 1731–1733. 4 Davenport RJ, Dennis MS, Wellwood I, Warlow CP. Complications after acute stroke. Stroke 1996; 27: 415–420. 5 Dromerick A, Reding M. Medical and neurological complications during inpatient stroke rehabilitation. Stroke 1994; 25: 358–361. 6 Dobkin BH. Neuromedical complications in stroke patients transferred for rehabilitation before and after diagnostic related groups. J Neurol Rehab 1987; 1: 3–7. 7 McClatchie G. Survey of the rehabilitation outcomes of stroke. Med J Aust 1980; 1: 649–651. 8 Kalra L, Yu G, Wilson K, Roots P. Medical complications during stroke rehabilitation. Stroke 1995; 26: 990–994. 9 Subbararao BJ, Smith J. Pulmonary embolism during stroke rehabilitation. Illinois Med J 1984; 165: 328–332. 10 Turpie AGG, Levine MN, Hirsh J et al. Double-blind randomized trial of org 10172 low-molecular-weight heparinoid in prevention of deep-vein thrombosis in thrombotic stroke. Lancet 1987; 1: 523–527. 11 International Stroke Trial Collaborative Group. The International Stroke Trial (IST): a randomized trial of aspirin, subcutaneous heparin, both, or neither among 19 435 patients with acute ischaemic stroke. Lancet 1997; 349: 1569–1581. 12 McCarthy ST, Turner JJ, Robertson D, Hawkey CJ. Lowdose heparin as a prophylaxis against deep-vein thrombosis after acute stroke. Lancet 1977; 2: 800–801.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

March 28, 2007

20

15:14

CHAPTER 20

Pulmonary embolism and deep venous thrombosis in hospitalized adults with chronic obstructive pulmonary disease

When patients with chronic obstructive pulmonary disease (COPD) are routinely evaluated for pulmonary embolism (PE), the rate of diagnosis has been reported as 9 of 31 (29%) and 49 of 197 (25%) [1, 2]. Rates of diagnosis of PE among hospitalized patients with COPD who were not routinely evaluated for PE were much lower than found among patients with COPD who underwent routine testing for PE [3]. Over the 25-year period of evaluation (1979–2003), 58,392,000 adults ≥20 years of age were hospitalized with COPD [3]. Among these patients, PE was diagnosed in 0.65%, DVT (deep venous thrombosis) in 1.08%, and VTE (venous thromboembolism) (PE and/or DVT) in 1.58% [3] (Tables 20.1 and 20.2). Among 789,403,000 adults hospitalized with illnesses other than COPD, PE was diagnosed in 0.34%, DVT in 0.83%, and VTE in 1.09% [3] (Tables 20.1 and 20.2). The relative risk for PE in adults hospitalized with COPD compared with those who did not have COPD was 1.92 [3] (Table 20.1, Figure 20.1) [3]. Among hos-

pitalized patients aged 20–39 years with COPD, the relative risk for PE compared with hospitalized adults the same age who did not have COPD was 5.34 [3]. Among patients aged 40–59 years, the relative risk for PE decreased to 2.02. Among patients aged 60–79 years and 80–99 years, the relative risk for PE was 1.23 and 1.41, respectively. The relative risk for DVT among hospitalized adults with COPD compared with those who did not have COPD was 1.30 and as with patients with PE, the relative risk was high in younger patients (Table 20.2) [3]. In young adults other risk factors in combination with COPD are uncommon, so the contribution of COPD to the risk of PE and DVT becomes more apparent than in older patients. The relative risk for PE and DVT in all patients increases exponentially with age [4]. Congestive heart failure, cancer, and stroke have been shown to be risk factors for VTE [5–7]. The prevalence of these risks, as we showed in this study and as others have shown previously [6, 7], is age dependent. The prevalence of congestive heart failure, cancer, and

Table 20.1 Rates of pulmonary embolism and relative risk in hospitalized patients with COPD. n/N(%) Age group (yrs)

PE, COPD

PE, No COPD

Relative risk

20–39

5,500/846,000 (0.65)

313,000/257,457,000 (0.12)

5.34

40–59

67,000/9,700,000 (0.69)

653,000/190,519,000 (0.34)

2.02

60–79

226,000/35,057,000 (0.64)

1,238,000/236,328,000 (0.52)

1.23

80–99

82,000/12,789,000 (0.64)

479,000/105,099,000 (0.46)

1.41

All adults ≥20

381,000/58,392,000 (0.65)

2,684,000/789,403,000 (0.34)

1.92

PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease; yrs, years. Reprinted from Stein et al. [3], with permission.

101


BLUK077-Stein

March 28, 2007

15:14

102

PART I

Prevalence, risks, and prognosis of PE and DVT

Table 20.2 Rates of deep vein thrombosis and relative risk in hospitalized patients with COPD. n/N(%) Age group (yrs)

DVT, COPD

DVT, No COPD

Relative risk

20–39

8,000/846,000 (0.95)

952,000/257,457,000 (0.37)

2.56

40–59

98,000/9,700,000 (1.01)

1,749,000/190,519,000 (0.92)

1.10

60–79

389,000/35,057,000 (1.11)

2,761,000/236,328,000 (1.17)

0.95

80–99

138,000/12,789,000 (1.08)

1,092,000/105,099,000 (1.04)

1.04

All adults ≥20

632,000/58,392,000 (1.08)

6,554,000/789,403,000 (0.83)

1.30

DVT, deep venous thrombosis; COPD, chronic obstructive pulmonary disease; yrs, years. Reprinted from Stein et al. [3], with permission.

stroke increased with age among patients who did not have COPD [3]. The relative risk of PE in patients with COPD, therefore, decreased with age because of such competing risk factors. In younger patients with COPD, however, comorbid conditions were uncommon in the control population, so the contribution of COPD to the risk of VTE was evident. Symptomatic DVT was found in 3 of 196 (1.5%) of patients hospitalized in a respiratory intensive care unit with an exacerbation of COPD [8]. Subclinical DVT was diagnosed in an additional 18 of 196 (9.2%) [8]. The prevalence of subclinical DVT in other studies of patients with an exacerbation of COPD, each with fewer than 60 patients, ranged from 0 to 45% [9–12]. The prevalence of PE in hospitalized patients throughout the United States with COPD (0.65%),

Table 20.3 Predisposing factors in patients with COPD and suspected acute PE.

6 Relative risk for PE with COPD

compared with the rate of diagnosis of PE in hospitalized patients with COPD, all of whom underwent diagnostic tests for PE, 25–29% [1, 2] indicates that PE in patients with COPD is generally underdiagnosed. The clinical features in 108 patients in PIOPED I who had COPD and were suspected of having PE were evaluated [13]. In patients with COPD, wheezing was less frequent and crackles were more frequent in patients with PE than in those with COPD who did not have PE. The predisposing factors, symptoms, other signs, chest radiographic findings, blood gases, and alveolar–arterial oxygen differences (gradients) did not differ to a statistically significant extent in patients with COPD who had PE compared with patients with COPD who did not have PE [13] (Tables 20.3–20.6,

PE,* No. (%)

No PE, No. (%)

(n = 21)

(n = 87)

Immobilization

9 (43)

36 (41)

Surgery

7 (33)

21 (24)

2

Thrombophlebitis, ever

3 (14)

5 (6)

1 ----------------------------------------------------------

Malignancy

2 (10)

12 (14)

Trauma, lower

1 (5)

7 (8)

Stroke

1 (5)

4 (5)

Estrogen

0 (0)

2 (2)

5 4 3

0

20−39

40−59

60−79

80−99

Age (years) Figure 20.1 Relative risk for pulmonary embolism (PE) in patients with chronic obstructive pulmonary disease (COPD) compared to patients the same decade of age hospitalized without COPD. (Reprinted from Stein et al. [3], with permission.)

extremities

All differences between PE and No PE were not significant. * Some patients had more than one predisposing factor. PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease. Reprinted with permission from Lesser et al. [13].


BLUK077-Stein

March 28, 2007

CHAPTER 20

15:14

103

PE and DVT in patients with COPD

Table 20.4 Symptoms in patients with COPD and suspected acute PE.

Table 20.6 Chest radiographs: patients with COPD and suspected acute PE.

PE, No. (%)

No PE, No. (%)

PE, No. (%)

No PE, No. (%)

(n = 21)

(n = 87)

(n = 21)

(n = 87)

Dyspnea

19 (90)

80 (92)

Atelectasis

12 (57)

36 (41)

Cough

13 (62)

48 (55)

Effusion

10 (48)

31 (34)

Pleuritic pain

9 (43)

32 (37)

Volume loss

10 (48)

31 (34)

Leg swelling

9 (43)

23 (26)

Infiltrate

9 (43)

30 (33)

Wheezing

8 (38)

35 (40)

Elevated hemidiaphragm

6 (29)

18 (20)

Leg pain

4 (19)

13 (15)

Oligemia

4 (19)

16 (18)

Hemoptysis

4 (19)

7 (8)

Palpations

2 (10)

16 (18)

Angina-like pain

0 (0)

8 (9)

Figures 20.2 and 20.3). Changes of the alveolar–arterial oxygen gradients from prior values to values at the time of the suspected PE were no greater in those with PE than in those in whom PE was excluded [13] (Figure 20.4). Physicians, when confident of a low-probability clinical assessment, were usually correct in excluding

Table 20.5 Signs of acute PE in patients with COPD and suspected acute PE. PE, No. (%)

No PE, No. (%)

(n = 21)

(n = 87)

Crackles

17 (81)

46 (53)

Tachypnea (≥20/min)

15 (71)

71 (82)

Tachycardia (>100/min)

7 (33)

31 (36)

Wheezes

2 (10)

34 (39)

Deep vein thrombosis

2 (10)

9 (10)

Third heart sound

2 (10)

7 (8)

Diaphoresis

1 (5)

10 (11)

Temperature >38.5◦ C

1 (5)

7 (8)

Cyanosis

1 (5)

3 (3)

Increased pulmonary

0 (0)

11 (13)

component of second

Table 20.7 V–Q findings in 108 patients with COPD and suspected acute PE. V–Q scan probability

PE/total (%)

High

5/5 (100)

Intermediate

14/65 (22)

Low

2/33 (6)

Near normal/normal

0/5 (0)

Total

21/108 (19)

V–Q, ventilation perfusion; PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease. Reprinted with permission from Lesser et al. [13].

75

PaCO2 (mm Hg)

All differences between PE and No PE were not significant. PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease. Reprinted with permission from Lesser et al. [13].

All differences between PE and No PE were not significant. PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease. Reprinted with permission from Lesser et al. [13].

NS

60 45 30 15 0 PE (n = 14)

heart sound Pleural friction rub

0 (0)

4 (5)

Homan’s sign

0 (0)

1 (1)

PE, pulmonary embolism; COPD, chronic obstructive pulmonary disease. Reprinted with permission from Lesser et al. [13].

NO PE (n = 57)

Figure 20.2 Partial pressure of carbon dioxide in arterial blood (PaCO2 ) while breathing room air in patients with pulmonary embolism (PE) and in patients in whom PE was excluded (No PE). The difference was not statistically significant (NS). (Reprinted with permission from Lesser et al. [13].)


BLUK077-Stein

March 28, 2007

15:14

A-a gradient (mm Hg)

104

PART I

225 200 175 150 125 100 75 50 25 0

Prevalence, risks, and prognosis of PE and DVT

NS

PE (n = 21)

NO PE (n = 83)

Figure 20.3 Alveolar–arterial (A-a) oxygen gradient among patients with pulmonary embolism (PE) and patients in whom PE was excluded (No PE). The difference was not statistically significant (NS). (Reprinted with permission from Lesser et al. [13].)

PE [correct in 29 of 30 (97%)] but physicians most often were uncertain of the diagnosis (67 patients) and rarely were confident of a high probability of PE (correct in 3 of 3) [13]. In patients with COPD, V–Q scans sometimes give diagnostic information, but less often than in those with no cardiopulmonary disease or cardiopulmonary disease exclusive of COPD [13, 14] (Table 20.7). Presumably, contrast enhanced CT is more likely to make or exclude the diagnosis of PE. The majority of ventilation–perfusion scans in patients with COPD, 60%, were interpreted as intermediate (indeterminate) probability for PE [13]. The cause of the perfusion defect, at least in patients with emphysema due to alpha-1

250

PE (n = 10)

250

A-a gradient (mm Hg)

A-a gradient (mm Hg)

150 100 50 0

antitrypsin deficiency, is destruction of the distal pulmonary arterial branches and capillary bed [15]. This is apparent on pulmonary angiograms and wedge angiograms of such patients, compared with the perfused capillary network of normal patients [15, 16] (Figures 20.5 and 20.6) (see normally perfused capillary network illustrated by wedge angiogram in Chapter 71).

NO PE (n = 41)

P < .01

NS 200

Figure 20.5 Pulmonary wedge arteriogram in a patient with emphysema associated with alpha-1 antitrypsin deficiency. There is diminished arborization of small pulmonary artery branches. (Reproduced with permission from Stein et al. [15].)

200 150 100 50 0

Pre Current

Pre

Current

Figure 20.4 Left: Alveolar–arterial (A-a) oxygen gradient among patients with pulmonary embolism (PE) who had both prior assessments and assessments at the time of the PE (current). Right: Prior and current values of A-a gradient among patients in whom PE was excluded. The A-a gradient increased in both, but the difference was not significant comparing those with PE and those who did not have PE. (Reprinted with permission from Lesser et al. [13].)


BLUK077-Stein

March 28, 2007

CHAPTER 20

15:14

105

PE and DVT in patients with COPD

(a)

(b)

Figure 20.6 (a) Pulmonary arteriogram in a patient with emphysema associated with alpha-1 antitrypsin deficiency. A diminished number of branches is shown in arteries of the lower lung zones. Capillary hypoperfusion in the lower zones shown on this film was confirmed in subsequent films. (Reproduced with permission from Stein et al. [15].) (b) Pulmonary wedge arteriogram in the same patient with emphysema associated with alpha-1 antitrypsin deficiency as shown in Figure 20.6a. There is a prominent reduction in the number of vessels branching from this artery. (Reproduced with permission from Stein et al. [15].)

References 1 Mispeleare D, Glerant JC, Audebert M et al. Pulmonary embolism and sibilant types of chronic obstructive pulmonary disease decompensations. Rev Mal Respir 2002; 19: 415–423. 2 Tillie-Leblond I, Marquette CH, Perez T et al. Pulmonary embolism in patients with unexplained exacerbation of chronic obstructive pulmonary disease: prevalence and risk factors. Ann Intern Med 2006; 144: 390– 396. 3 Stein PD, Beemath A, Meyers FA, Olson RE. Pulmonary embolism and deep venous thrombosis in hospitalized adults with chronic obstructive pulmonary disease. J Cardiovasc Med 2007 (in press). 4 Stein PD, Hull RD, Kayali F et al. Venous thromboembolism according to age: the impact of an aging population. Arch Intern Med 2004; 164: 2260– 2265. 5 Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism—the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 338s–400s. 6 Stein PD, Beemath A, Meyers FA, Skaf E, Sanchez J, Olson RE. Incidence of venous thromboembolism in patients hospitalized with cancer. Am J Med 2006; 119(1): 60– 68. 7 Skaf E, Stein PD, Beemath A, Sanchez J, Bustamante MA, Olson RE. Venous thromboembolism in patients with ischemic and hemorrhagic stroke. Am J Cardiol 2005; 96(12): 1731–1733. 8 Schonhofer B, Kohler D. Prevalence of deep-vein thrombosis of the leg in patients with acute exacerbation of chronic obstructive pulmonary disease. Respiration 1998; 65: 173–177. 9 Erelel M, Cuhadaroglu C, Ece T, Arseven O. The frequency of deep venous thrombosis and pulmonary embolus in acute exacerbation of chronic obstructive pulmonary disease. Respir Med 2002; 96: 515–518. 10 Pek WY, Johan A, Stan S et al. Deep vein thrombosis in patients admitted for exacerbation of chronic obstructive pulmonary disease. Singapore Med J 2001; 42: 308–311. 11 Prescott SM, Richards KL, Tikoff G et al. Venous thromboembolism in decompensated chronic obstructive pulmonary disease. A prospective study. Am Rev Respir Dis 1981; 123: 32–36. 12 Winter JH, Buckler PW, Bautista AP et al. Frequency of venous thrombosis in patients with an exacerbation of chronic obstructive lung disease. Thorax 1983; 38: 605– 608. 13 Lesser BA, Leeper KV, Stein PD et al. The diagnosis of acute pulmonary embolism in patients with chronic


BLUK077-Stein

March 28, 2007

15:14

106

PART I

obstructive pulmonary disease. Chest 1992; 102: 17– 22. 14 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603.

15 Stein PD, Leu JD, Welch MH, Guenter CA. Pathophysiology of the pulmonary circulation in emphysema associated with alpha antitrypsin deficiency. Circulation 1971; 43: 227–239. 16 Stein PD. Wedge arteriography for the identification of pulmonary emboli in small vessels. Am Heart J 1971; 82: 618–623.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

March 10, 2007

21

14:27

CHAPTER 21

Pulmonary embolism and deep venous thrombosis in hospitalized patients with asthma

Asthma is not included among the recognized risk factors for pulmonary embolism (PE) or deep venous thrombosis (DVT) [1–3]. The association of PE with asthma had been described only in case reports [4, 5]. However, among 18,041,000 patients ≥20 years of age who were hospitalized with asthma from 1979 to 2003, based on data from the National Hospital Discharge Survey (NHDS), PE was diagnosed in 93,000 (0.51%) and DVT was diagnosed in 169,000 (0.93%). Venous thromboembolism (VTE) (PE and/or DVT) occurred in 242,000 (0.97%) (Stein et al., unpublished data from the National Hospital Discharge Survey). On average, the risk for PE in patients hospitalized with asthma was higher than the risk for PE in patients who did not have asthma (relative risk 1.21) (Stein et al., unpublished data from the National Hospital Discharge Survey). The relative risk was age-dependent (Figure 21.1). Among patients aged 20–39 years, the relative risk for PE was 2.58. Among patients aged

40–59 years, the relative risk for PE decreased to 1.31. The relative risk in older patients ranged from 0.86 to 1.21 (Stein et al., unpublished data from the National Hospital Discharge Survey) (Figure 21.1). The risk for DVT in patients hospitalized with asthma, on average, was not higher than the risk for DVT in patients who did not have asthma (relative risk 0.94) (Stein et al., unpublished data from the National Hospital Discharge Survey). However, the relative risk was higher in younger patients aged 20–39 years (relative risk 1.36) (Figure 21.2). These observations indicate that young adults hospitalized with asthma have a high risk for PE and DVT in comparison to hospitalized patients the same age who do not have asthma. With increasing age and its accompanying risk factors for PE and DVT, other risk factors balance or outweigh the risk of asthma alone. In older patients, therefore, a higher relative risk for VTE in patients with asthma is thereby eliminated.

2

2

1 ----------------------------------------------------------

0

20−39

40−59

60−79

80−99

Age (years) Figure 21.1 Relative risk for pulmonary embolism (PE) in hospitalized patients with asthma compared to patients who did not have asthma shown in relation to age. The relative risk was age-dependent.

Relative risk for DVT with asthma

Relative risk for PE with asthma

3

1 ----------------------------------------------------------

0

20−39

40−59

60−79

80−99

Age (years) Figure 21.2 Relative risk for deep venous thrombosis (DVT) in hospitalized patients with asthma compared to patients who did not have asthma shown in relation to age. The relative risk was higher in younger patients.

107


BLUK077-Stein

March 10, 2007

14:27

108

PART I

References

3 Cohen AT. Venous thromboembolic disease management of the nonsurgical moderate- and high-risk patient. Semin Hematol 2000; 37: 19–22. 4 Labay Matias MV, Hervas Palazon J, Puges Bassols E, Perez Ferron J. [Status asthmaticus, adult respiratory distress syndrome and pulmonary vascular thrombosis]. An Esp Pediatr 1985; 22: 169–170. 5 Divac A, Djordjevic V, Jovanovic D et al. Recurrent pulmonary embolism in a patient with asthma. Respiration 2004; 71: 428.

1 Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 338S–400S. 2 Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000; 160: 3415– 3420.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

March 10, 2007

22

14:33

CHAPTER 22

Deep venous thrombosis and pulmonary embolism in hospitalized patients with sickle cell disease

Nearly every component of hemostasis, including platelet function and the procoagulant, anticoagulant, and fibrinolytic states is altered in sickle cell disease (SCD) [1]. This has led to the view that SCD is a hypercoagulable state [1, 2]. Interactions among sickle erythrocytes, other blood cells, and the various matrix proteins that line the endothelium contribute to the multifactorial vaso-occlusive process [1]. Prior to our evaluation of the prevalence of deep venous thrombosis (DVT) in patients with SCD [3], DVT had been described only in case reports [4, 5]. Deep venous thrombosis is not included among the usual complications of SCD [6]. In past years, there may have been a reluctance to fully evaluate patients with SCD who had suspected DVT, because highosmolarity contrast material used for venograms had a potential for inducing sickling [7]. With the ready availability of venous ultrasound after 1991 [8], a definitive noninvasive diagnosis of DVT became possible in these patients. Among 1,804,000 patients hospitalized with SCD from 1979 to 2003, a discharge diagnosis of DVT was made in 11,000 (0.61%) [3]. From 1991 to 2003, when venous ultrasound was in general use, a somewhat higher percentage of patients hospitalized with SCD had DVT, 7800 of 1,107,000 (0.70%). Acute pulmonary embolism (PE) in patients with SCD is a more contentious issue than acute DVT. As would be expected with a hypercoagulable state, pulmonary thromboembolism occurs in SCD [9–11]. Its frequency, however, is undetermined, largely because of difficulties in distinguishing it from thrombosis in situ [10–13]. Pulmonary embolism has been documented in case reports of patients with SCD [14, 15],

and in 25% [16] and 60% [11] of autopsied patients with SCD. The distinction between PE and thrombosis in situ is particularly important in patients with SCD because both perhaps could be the cause of the acute chest syndrome (pulmonary infiltrates, fever, chest pain, and respiratory symptoms) in these patients. The acute chest syndrome is a leading cause of death in patients with SCD [17–19]. Even though pulmonary infarction occurs in 16% of patients with the acute chest syndrome [20], PE generally is not considered a cause [7]. The likely cause of pulmonary infarction in patients with SCD has been suggested to be thrombosis in situ [10, 12, 21]. Others suggested that engorge ment of the small pulmonary vessels, perhaps due to vasospasm and sickling, can produce pulmonary infarction in the absence of thrombosis of these vessels [22–24]. Whether acute PE is an initiating cause of the acute chest syndrome or whether recurrent PE leads to pulmonary hypertension in SCD has been an unsolved issue [9]. Imaging tests for PE were not obtained in a large study of causes of the acute chest syndrome in SCD [20]. Generally considered causes of the acute chest syndrome include infection, fat embolism, bone marrow embolism, fluid overload, hypoxemia, and vascular obstruction due to sickling and endothelial adherence of erythrocytes [7, 25]. Treatment of the acute chest syndrome includes antibiotics, oxygen, fluid management, respiratory therapy, pain management, bronchodilator therapy, and exchange transfusion [20]. Anticoagulant therapy is not included [20] and its role in the treatment of the acute chest syndrome is uncertain [7] or unconvincing [1, 26, 27]. Low doses of oral anticoagulants (average INR = 1.6) have been speculated to be a relevant form of therapy

109


BLUK077-Stein

March 10, 2007

14:33

PART I

because low doses can normalize the hypercoaguable state in patients with SCD [13]. This presumes that PE is not an etiologic factor. However, once pulmonary hypertension occurs following the acute chest syndrome, PE contributes substantially to further morbidity and mortality [12]. Over the 25-year period from 1979 to 2003, a discharge diagnosis of PE was made in 9000 of 1,804,000 patients hospitalized with SCD. (0.50%) [3]. During this same period, among hospitalized African Americans who did not have SCD, 322,000 of 97,463,000 (0.33%) had a discharge diagnosis of PE. Approximately, 7000 of 9000 (78%) patients with SCD who had a discharge diagnosis of PE were <40 years of age. In patients <40 years old, 7000 of 1,581,000 (0.44%) with SCD had a discharge diagnosis of PE, whereas only 59,000 of 48,611,000 (0.12%) of African Americans without SCD had a discharge diagnosis of PE (Figure 22.1). The prevalence of DVT was similar in patients <40 years old with SCD, 7000 of 1,581,000 (0.44%) and in African Americans who did not have SCD, 193,000 of 48,611,000 (0.40%). The relatively high prevalence of apparent PE in patients with SCD, and the comparable prevalences of DVT in both groups are consistent with the possibility of thrombosis in situ. Thrombosis in situ, however, may not be the exclusive cause of pulmonary vascular occlusion in patients with SCD. If the prevalence of PE in SCD patients aged <40 years were the same as in non-SCD African Americans the same age, then 2000 of 1,581,000 (0.13%) of patients with SCD would be

expected to have PE. This accounts for 29% of the 7000 patients with SCD who had a discharge diagnosis of PE. Thrombosis in situ is limited to muscular pulmonary artery branches (<1 mm diameter) and arterioles [12] and cannot be identified by ordinary pulmonary angiography or CT pulmonary angiography. The smallest PE that have been identified in living patients were in 1–2-mm-diameter pulmonary artery branches [28]. These were shown with wedge pulmonary arteriography [28]. Based on pulmonary angiography, PE occurs in main, lobar, or segmental pulmonary arteries in 96% of patients [29]. In view of the fact that thrombosis in situ occurs only in muscular pulmonary arteries or arterioles, the demonstration of thrombi in elastic vessels (>1 mm diameter) is diagnostic of PE. To diagnose PE, an imaging study is required to show an intraluminal filling defect in an elastic artery (>1 mm). Standard pulmonary angiography and CT pulmonary angiography are the commonly used methods for imaging intraluminal filling defects. However, the risk of injection of radio-opaque contrast material requires assessment in view of evidence that ionic contrast material may induce sickling [30]. Whether nonionic contrast material carries the same risk is uncertain. The package insert of nonionic contrast material cautions against its use in patients with SCD [31]. Magnetic resonance angiography has been used in a few patients with PE and shows promise [32–34]. The package insert for gadopentetate dimeglumine, a magnetic resonance contrast agent, warns that deoxygenated sickle erythrocytes align perpendicular to a magnetic field in vitro [35]. This may result in vasoocclusive complications in vivo. Gadopentetate dimeglumine has not been studied in patients with SCD. A ventilation–perfusion lung scan would not contribute to the differentiation between thrombosis in situ and pulmonary thromboembolism. In patients with the acute chest syndrome, if DVT were present it would be reasonable to assume that venous thromboembolism occurred. Clinically apparent DVT, however, is present in only 11% of patients with acute PE [36]. A venous compression ultrasound would be useful if it were positive, but it lacks sensitivity for DVT in patients with suspected PE who do not have clinical evidence of DVT, being positive in only 29% [37]. Pulmonary thromboembolism is not rare in patients with SCD [3]. This raises the issue of whether

Hospitalized patients <age 40 with PE, DVT (%)

110

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

0.44

0.44 0.40

SCD

0.12 Af-Amer No SCD

PE

SCD AfAmer No SCD

DVT

Figure 22.1 Hospitalized patients younger than 40 years of age with sickle cell disease (SCD) who had a discharge diagnosis of pulmonary embolism (PE) or deep venous thrombosis (DVT) (light bars) compared with African Americans (Af-Amer) the same age who did not have SCD (dark bars). (Reprinted from Stein et al. [3], with permission from Elsevier.)

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

March 10, 2007

CHAPTER 22

14:33

Patients with sickle cell disease

PE might be an etiologic factor in patients with SCD who have the acute chest syndrome or pulmonary hypertension. In such patients, the risks of an imaging procedure need to be balanced with the advantage of proper diagnosis and treatment.

References 1 Ataga KI, Orringer EP. Hypercoagulability in sickle cell disease: a curious paradox. Am J Med 2003; 115: 721–728. 2 Francis RB, Jr. Platelets, coagulation, and fibrinolysis in sickle cell disease: their possible role in vascular occlusion. Blood Coagul Fibrinolysis 1991 2: 341–353. 3 Stein PD, Beemath A, Meyers FA, Skaf E, Olson RE. Deep venous thrombosis and pulmonary embolism in hospitalized patients with sickle cell disease. Am J Med 2006; 119: 897 e7–11. 4 Brion L, Dupont M, Fondu P, Rutsaert J. Sickle cell anemia and venous thrombosis. Acta Paediatr Belg 1978; 31: 241– 244. 5 Koren A, Zalman L, Levin C et al. Venous thromboembolism, Factor V Leiden, and methylenetetrahydrofolate reductase in a sickle cell anemia patient. Pediatr Hematol Oncol 1999; 16: 469–472. 6 Wang WC, Lukens JN. Sickle cell anemia and other sickling syndromes, In: Lee GR, Paraskevas S, Foerster J, Greer JP et al., eds. Wintrobe’s Clinical Hematology, 10th edn. Williams & Wilkins, Baltimore, 1999: 1346–1357. 7 Kirkpalrick MB, Haynes J. Sickle cell disease and the pulmonary circulation. Semin Respir Crit Care Med 1994; 15: 473–481. 8 Stein PD, Hull RD, Ghali WA et al. Tracking the uptake of evidence: two decades of hospital practice trends for diagnosing deep vein thrombosis and pulmonary embolism. Arch Intern Med 2003; 163: 1213–1219. 9 Serjeant GR. Sickle Cell Disease, 2nd edn. Oxford University Press, Oxford, UK, 1992: 184–187. 10 Moser KM., Shea JG. The relationship between pulmonary infarction, cor pulmonale and the sickle states. Am J Med 1957; 22: 561–579. 11 Oppenheimer EH, Esterly JR. Pulmonary changes in sickle cell disease. Am Rev Respir Dis 1971; 103: 858–859. 12 Adedeji MO, Cespedes J, Allen K, Subramony C, Hughson MD. Pulmonary thrombotic arteriopathy in patients with sickle cell disease. Arch Pathol Lab Med 2001; 125: 1436– 1441. 13 Walker B K, Ballas SK, Burka ER. The diagnosis of pulmonary thromboembolism in sickle cell disease. Am J Hematol 1979; 7: 219–232. 14 Maggi JC, Nussbaum E. Massive pulmonary infarction in sickle cell anemia. Pediatr Emerg Care 1987; 3: 30–32.

111

15 Rubler S, Fleischer RA. Sickle cell states and cardiomyopathy. Sudden death due to pulmonary thrombosis and infarction. Am J Cardiol 1967; 19: 867–873. 16 Haupt HM, Moore GW, Bauer TW, Hutchins GM. The lung in sickle cell disease. Chest 1982; 81: 332–337. 17 Vichinsky E. Comprehensive care in sickle cell disease: its impact on morbidity and mortality. Semin Hematol 1991; 28: 220–226. 18 Castro O, Brambilla DJ, Thorington B et al. The acute chest syndrome in sickle cell disease: incidence and risk factors: the Cooperative Study of Sickle Cell Disease. Blood 1994; 84: 643–649. 19 Platt OS, Brambilla DJ, Rosse WF et al. Mortality in sickle cell disease: life expectancy and risk factors for early death. N Engl J Med 1994; 330: 1639–1644. 20 Vichinsky EP, Neumayr LD, Earles AN et al., for National Acute Chest Syndrome Study Group. Causes and outcomes of the acute chest syndrome in sickle cell disease. N Engl J Med 2000; 342: 1855–1865. 21 Durant JR, Cortes FM. Occlusive pulmonary vascular disease associated with hemoglobin SC disease. Am Heart J 1966; 71: 100–106. 22 Athanasou NA, Hatton C, McGee JO, Weatherall DJ. Vascular occlusion and infarction in sickle cell crisis and the sickle chest syndrome. J Clin Pathol 1985; 38: 659– 664. 23 Francis RB, Johnson CS. Vascular occlusive in sickle cell disease: current concepts and unanswered questions. Blood 1991; 77: 1405–1414. 24 Gladwin MT, Rodgers GP. Pathogenesis and treatment of acute chest syndrome of sickle-cell anaemia. Lancet 2000; 355: 1476–1478. 25 Minter KR, Gladwin MT. Pulmonary complications of sickle cell anemia. A need for increased recognition, treatment, and research. Am J Respir Crit Care Med 2001; 164: 2016–2019. 26 Collins FS, Orringer EP. Pulmonary hypertension and cor pulmonale in the sickle hemoglobinopathies. Am J Med 1982; 73: 814–821. 27 Barrett-Connor E. Pneumonia and pulmonary infarction in sickle cell anemia. JAMA 1973; 224: 997–1000. 28 Stein PD. Wedge arteriography for the identification of pulmonary emboli in small vessels. Am Heart J 1971; 82: 618–623. 29 Stein PD, Henry JW. Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries and relation to probability interpretation of ventilation/perfusion lung scans. Chest 1997; 111: 1246– 1248. 30 Richards D, Nulsen FE. Angiographic media and the sickling phenomenon. Surg Forum 1971; 22: 403–404. 31 Optiray (Ioversol Injection). Mallinckrodt Inc., St. Louis, Missouri, September 2000.


BLUK077-Stein

March 10, 2007

14:33

112

PART I

32 Meaney JFM, Weg JG, Chenevert TL et al. Diagnosis of pulmonary embolism with magnetic resonance angiography. N Engl J Med 1997; 336: 1422–1427. 33 Oudkerk M, van Beek EJ, Wielopolski P et al. Comparison of contrast-enhanced magnetic resonance angiography and conventional pulmonary angiography for the diagnosis of pulmonary embolism: a prospective study. Lancet 2002; 359: 1643–1647. 34 Gupta A, Frazer CK, Ferguson JM et al. Acute pulmonary embolism: diagnosis with MR angiography. Radiology 1999; 210: 353–359.

35 Magnevist Injection (Brand of Gadopentetate Dimeglumine). Berlix Laboratories, Wayne, New Jersey, May 2000. 36 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 37 Turkstra F, Kuijer PM, van Beek EJ, Brandjes DP, ten Cate JW, Buller HR. Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med 1997; 126: 775–781.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

April 6, 2007

23

16:40

CHAPTER 23

Venous thromboembolism in pregnancy

Introduction

Trends in DVT according to age

Pulmonary venous thromboembolism in developed countries is one of the leading causes of maternal death [1–6]. Limited survey data from the United States, Denmark, and the United Kingdom have yielded conflicting findings as to whether venous thromboembolism associated with pregnancy is increasing or declining [3, 7–10]. We used the database of the National Hospital Discharge Survey, available at: http:// www.cdc.gov/nchs/about/major/hdasd/nhds.htm, to assess trends in venous thromboembolism during pregnancy [11].

The rate of pregnancy-associated DVT was higher among women aged 35–44 years than in younger women (Figure 23.2). On average, 91% of maternal patients were 15– 34 years of age (Figure 23.3). The percentage of deliveries in women 35–44 years of age increased over 21 years from 4.5 to 13.1% and the percentage of deliveries in younger women decreased proportionately (Figure 23.3).

Pregnancy-associated DVT according to race

Twenty-one-year trends in rate of diagnosis of deep venous thrombosis

Figure 23.1 Triennial rates of deep venous thrombosis (DVT) in women aged 15–44 years. (Reprinted from Stein et al. [11], with permission from Elsevier.)

Deep venous thrombosis/ 100,000/yr

Pregnancy-associated deep venous thrombosis (DVT) was diagnosed in 93,000 of 80,798,000 women (0.12%) from 1979 to 1999 [11]. The rate of pregnancyassociated DVT (vaginal delivery and cesarean section) increased from 1982–1984 to 1997–1999 (Figure 23.1). The rate of nonpregnancy-associated DVT decreased from 1979–1981 to 1991–1993. Thereafter, the rate of nonpregnancy-associated DVT remained constant.

The rate of pregnancy-associated DVT among black women was higher than among white women during 1979–1988 and 1989–1999 (Figure 23.4). The rate of pregnancy-associated DVT in both black women and white women increased from 1979–1988 to 1989– 1999. Higher rates of pregnancy-associated DVT in black women compared to white women have previously been observed [6, 12]. Higher rates of fatal pulmonary embolism (PE) have also been reported among black maternity patients [9].

200 Pregnancy-associated 150 100 Nonpregnancy-associated

50

0 1979−1981 1982−1984 1985−1987 1988−1990 1991−1993 1994−1996 1997−1999 Years

113


BLUK077-Stein

April 6, 2007

16:40

114

Deep venous thrombosis/ 100,000/yr

PART I

Prevalence, risks, and prognosis of PE and DVT

350 300 250 200

35−44

150 100 50

25−34 15−24

Figure 23.2 Rates of pregnancy-associated deep venous thrombosis (DVT) according to decades of age, 1979–1988 to 1989–1999. (Data from Stein et al. [11].)

0 1979−1988

1989−1999 Years

The percentage of deliveries in black and white women remained constant over the 21-year period of observation (Figure 23.5). The rate of nonpregnancy-associated DVT was higher in black women than in white women (Figure 23.6). The rates of nonpregnancy-associated DVT decreased from 1979–1988 to 1989–1999 in both black and white women (Figure 23.6). The observation of a higher rate of DVT in black women in the nonmaternal population is compatible with our previous observation of an increased rate of DVT in age-matched blacks compared to whites [13, 14].

Pregnancy-associated DVT according to mode of delivery The rate of diagnosis of DVT following cesarean section and following vaginal delivery increased from 1979–1988 to 1989–1999 (Figure 23.7).

All deliveries (%)

100

The percentage of all cesarean section deliveries among women aged 15–34 years increased from 15.9% in 1979 to 22.6% in 1987. Thereafter, the rate decreased to 17.0% in 1999 (Figure 23.8). Among women aged 35–44 years, the percentage of all cesarean section deliveries increased linearly from 1.1% in 1979 to 3.9% in 1999 (Figure 23.8). Among all women who had cesarean sections, the proportion performed in women aged 35–44 years increased from 6.2% in 1979 to 18.7% in 1999 (Figure 23.9). Concordantly, the proportion of cesarean sections in women aged 15–34 years decreased.

Twenty-one-year trends in the rate of diagnosis of PE The rate of pregnancy-associated PE was lower than the rate of nonpregnancy-associated PE (Figure 23.10). However, the rate of pregnancy-associated DVT was higher than the rate of nonpregnancy-associated DVT.

15−34

50

35−44 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Years

Figure 23.3 Proportion of deliveries according to age. Over the 21-year period of observation, deliveries in women 35–44 years of age increased. (Data from Stein et al. [11].)


April 6, 2007

CHAPTER 23

16:40

115

Venous thromboembolism in pregnancy

Figure 23.4 Rates of deep venous thrombosis (DVT) between 1979–1988 and 1989–1999 among black and white pregnancy-associated women. (Data from Stein et al. [11].)

Deep venous thrombosis/ 100,000/yr

BLUK077-Stein

200 144

150

118 100

100 61 50 0 White

Black

1979−1988

White

Black

1989−1999

All deliveries (%)

100 White

50

Figure 23.5 The proportion of deliveries in black and white women. The proportions were constant over the 21-year period of observation. (Data from Stein et al. [11].)

Figure 23.7 Rates of DVT following cesarean section and vaginal delivery during 1979–1988 and 1989–1999. (Data from Stein et al. [11].)

0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999

Deep venous thrombosis/ 100,000/yr

Years

200 150 100

70

55 50

57 30

0

White

Black

1979−1988 Deep venous thrombosis/ 100,000/yr

Figure 23.6 Rates of nonpregnancy-associated deep venous thrombosis (DVT) in black and white women. (Data from Stein et al. [11].)

Black

White

Black

1989−1999

120

104

80

63 47

40

29

0 Vaginal delivery

C-section

1979−1988

Vaginal delivery

C-section

1989−1999


BLUK077-Stein

April 6, 2007

16:40

Cesarean section (%)

116

PART I

Prevalence, risks, and prognosis of PE and DVT

30 15 −34

20

10 35 −44 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Years

The ratio of pregnancy-associated DVT to pregnancyassociated PE, 12.8, was higher than the ratio of nonpregnancy-associated DVT to nonpregnancyassociated PE, 2.9 (Figure 23.10). The reason for this difference is unknown and could reflect a difference of the natural history of DVT in pregnancy. It also could reflect a reluctance to expose pregnant women to ionizing radiation associated with imaging for PE resulting in a decreased frequency of diagnosis of PE. Our findings of an increasing rate of pregnancyassociated DVT in the United States [11] is in harmony with that observed in Denmark [7] where a similar upward trend was noted. The upward trend is concordant with reported findings in the United Kingdom for venous thromboembolism [3]. However, some reported a relatively constant rate of antepartum DVT and PE and a decreasing rate of postpartum PE [10]. Our observations support the impression held for many years that pregnancy predisposes to thromboembolism. The rate of pregnancy-associated DVT was twice the rate of nonpregnancy-associated DVT in women the same age [11]. A sixfold increase in the rate of thromboembolism during pregnancy and the

Figure 23.8 Cesarean sections, shown as a percentage of all deliveries according to age groups from 1979 to 1999. (Reprinted from Stein et al. [11], with permission from Elsevier.)

puerperium in comparison to nonpregnant women has been reported by others [15]. The reason for the diverging trends for the rates of diagnosis of pregnancy-associated DVT and nonpregnancy-associated DVT is not apparent. Analysis of covariance showed that age, race, and the percentage of deliveries by cesarean section did not explain the increasing rates of pregnancy-associated DVT over time. Although the percentage of deliveries in women aged 35–44 years increased over the 21-year period of observation, the increased risk of pregnancyassociated DVT was not limited to this age group. In fact, the greatest increase in rates of pregnancyassociated DVT was in younger women. Even though the rate of pregnancy-associated DVT was higher in black women than in white women, the percentage of deliveries in black women remained constant over the 21-year period of observation. Therefore, possible changes of racial characteristics of the maternal population would not explain the increasing rate of pregnancy-associated DVT. The proportion of deliveries by cesarean section decreased during most of the years that the rate of pregnancy-associated DVT was increasing (from 1987 to 1999). Therefore, the

Cesarean section (%)

100

15 −34

50

35 −44 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999

Years

Figure 23.9 Percentage of cesarean sections according to age from 1979 to 1999. (Data from Stein et al. [11].)


BLUK077-Stein

April 6, 2007

CHAPTER 23

16:40

117

Venous thromboembolism in pregnancy

150

Rate per 100,000/yr

11 5 100

49

50

17

9 Figure 23.10 Pregnancy-associated and nonpregnancy-associated deep venous thrombosis (DVT) and pulmonary embolism (PE) from 1979 to 1999. (Data From Stein et al. [11].)

0

percentage of deliveries by cesarean section could not explain the increasing rate of pregnancy-associated DVT. Increasing pregnancy-associated comorbidity due to acute and chronic medical conditions and comorbidity, which previously precluded conception may be co-factors, but this could not be assessed. Regarding nonpregnancy-associated DVT, a decreasing rate from 1973–1975 to 1988–1990 in the general population of women (pregnancy-associated and nonpregnancy-associated) was reported in women aged 15–44 years [16]. We observed the same trend during this time period in women of all ages, but from 1991 to 1999 the rate increased [14]. Our observations support the need for continued vigilance in the prevention of pregnancy-associated DVT. Further understanding is needed of the factors that contribute to this trend for an increasing rate of a potentially lethal condition in young healthy women.

References 1 Atrash HK, Koonin LM, Lawson HW, Franks AL, Smith JC. Maternal mortality in the United States, 1979–1986. Obstet Gynecol 1990; 76: 1055–1060. 2 Koonin LM, MacKay AP, Berg CJ, Atrash HK, Smith JC. Pregnancy-related mortality surveillance-United States, 1987–1990. MMWR 1997; 46: 17–36. 3 Department of Health, Welsh Office, Scottish Office Department of Health, Department of Health and Social Services, Northern Ireland. Why Mothers Die. Report

DVT

PE

Pregnancyassociated

4

5

6

7

8

9

10

11

12

DVT

PE

Nonpregnancyassociated

on Confidential Enquiries into Maternal Deaths in the United Kingdom, 1994–1996. 1998; Chapt 2. Hogberg U, Innala E, Sandstrom A. Maternal mortality in Sweden, 1980–1988. Obstet Gynecol 1994; 84: 240– 244. Berg CJ, Atrash HK, Koonin LM, Tucker M. Pregnancyrelated mortality in the United States, 1987–1990. Obstet Gynecol 1996; 88: 161–167. Rochat RW, Koonin LM, Atrash HK, Jewett JF. Maternal mortality in the United States: report from the Maternal Mortality Collaborative. Obstet Gynecol 1988; 72: 91– 97. Andersen BS, Steffensen FH, Sorensen HT, Nielsen GL, Olsen J. The cumulative incidence of venous thromboembolism during pregnancy and puerperium—an 11 year Danish population-based study of 63,000 pregnancies. Acta Obstet Gynecol Scand 1998; 77: 170–173. Macklon NS, Greer IA. Venous thromboembolic disease in obstetrics and gynecology: the Scottish experience. Scot Med J 1996; 41: 83–86. Franks AL, Atrash HK, Lawson HW, Colberg KS. Obstetrical pulmonary embolism mortality, United States, 1970–85. Am J Pub Health 1990; 80: 720–722. Heit JA, Kobbervig CE, James AH, Petterson TM, Bailey KR, Melton LJ, 3rd. Trends in the incidence of venous thromboembolism during pregnancy or postpartum: a 30-year population-based study. Ann Intern Med 2005; 143: 697–706. Stein PD, Hull RD, Kayali F et al. Venous thromboembolism in pregnancy: 21 year trends. Am J Med 2004; 117: 121–125. Buehler JW, Kaunitz AM, Hogue CJR, Hughes JM, Smith JC, Rochat RW. Maternal mortality in women aged 35 years or older: United States. JAMA 1986; 255: 53–57.


BLUK077-Stein

April 6, 2007

16:40

118

PART I

13 Stein PD, Hull RD, Patel KC et al. Venous thromboembolic disease: comparison of the diagnostic process in blacks and whites. Arch Intern Med 2003; 163: 1843–1848. 14 Stein PD, Hull RD, Patel KC, Olson RE, Ghali WA, Meyers FA. Venous thromboembolic disease: comparison of the diagnostic process in men and women. Arch Intern Med 2003; 163: 1689–1694.

15 Anonymous. Oral contraception and thrombo-embolic disease. J R Coll Gen Pract 1967; 13: 267–279. 16 Silverstein MD, Heit JA, Mohr DN, Petterson TM, O’Fallon WM, Melton, LJ, III. Trends in the incidence of deep vein thrombosis and pulmonary embolism. A 25-year population-based study. Arch Intern Med 1998; 158: 585–593.

Prevalence, risks, and prognosis of PE and DVT


BLUK077-Stein

March 12, 2007

24

15:45

CHAPTER 24

Air travel as a risk for pulmonary embolism and deep venous thrombosis

Introduction Air travel was suggested to be a risk for deep venous thrombosis (DVT) and pulmonary embolism (PE) in 1954, presumably induced by stasis [1]. However, the entire gamut of risk factors for venous thromboembolism has been suggested to participate in the occurrence of venous thromboembolic disease (VTE) following air travel [2]. The possibility of VTE after travel is not unique to air travel, having been reported with various modes of transportation [1, 3, 4]. Prolonged periods in cramped quarters, irrespective of travel, can lead to PE [5]. The term “economy class syndrome” was introduced in 1988 [6], but has since been replaced with “flight-related DVT” in recognition of the fact that all travelers are at risk, irrespective of the class of travel [7]. Systematic review indicates that there is a strong and significant association between prolonged air travel and PE and DVT [8]. Those with preexisting risk factors for VTE were most vulnerable [8, 9]. Flight duration between 3 and 18 hours appears to be a risk for DVT [9–13] and fatal PE [14]. Among patients who died of PE during flight, 10 of 11 died during travel lasting 12–18 hours [14]. Among 6.58 million passengers who arrived at Sydney International Airport following travel of ≥9 hours duration, 17 passengers had acute PE upon arrival (2.6 PE/million travelers) [15] (Figure 24.1). Among passengers who arrived at Madrid-Barajas Airport, 15 of 9.07 million passengers (1.65/million passengers) who traveled ≥8 hours had acute PE on arrival [16]. Only 1 of 3.93 million passengers who traveled 6–8 hours (0.25 per million) had acute PE on arrival and 0 of 28.04 million passengers who traveled ≤6 hours had acute PE on arrival [16] (Figure 24.1).

In a prospective investigation of travelers who traveled ≥10 hours, 4 of 878 (0.5%) developed PE and 5 of 878 (0.6%) developed DVT [17]. In terms of distance traveled, among those who traveled >6200 miles, 4.8 passengers per million passenger arrivals required transfer to a hospital by a French emergency medical team for acute PE [18]. Among those who traveled 4650–6199 miles, 2.7 passengers per million passenger arrivals had acute PE. The proportion with acute PE decreased to 0.4 passengers per million passenger arrivals with 3100–4649 miles traveled, 0.1 acute PE passengers per million passenger arrivals with 1550–3099 miles traveled, and no acute PE with <1549 miles traveled [18]. Recent travel has shown a positive association with DVT and PE in some [19–21], but not all investigations [22, 23]. Among inpatients, 41 of 168 (24.4%) with DVT or PE reported recent travel >4 hours by plane compared with 12 of 160 (7.5%) with other illnesses [19]. Other patients with DVT reported long distance travel in 62 of 494 (12.6%), whereas only 31 of 494 (6.3%) patients who did not have DVT reported long distance travel [20]. Still others showed that 20 of 185 patients with DVT (10.8%) traveled >3 hours compared with 31 of 383 (8.1%) who did not have DVT [21]. Regarding studies that showed no association of VTE with travel, 9 of 186 (5%) patients with DVT traveled ≥3 hours by train, car, or boat within the past 4 weeks, whereas even a somewhat higher proportion with no DVT, 42 of 602 (7%) also traveled >3 hours [22]. Similarly, 14 of 198 (7.1%) patients with DVT traveled >3 hours during the past 4 weeks, whereas the same proportion, 44 of 615 (7.2%) with no DVT also traveled >3 hours during the past 4 weeks [23]. The positive reports suggest a contribution of travel to the risk of PE and DVT, but the apparently negative

119


BLUK077-Stein

March 12, 2007

15:45

PE/million passenger arrivals

120

PART I

Prevalence, risks, and prognosis of PE and DVT

3 2.6 2.5 2 1.7 1.5 1 0.5 0

0.3 0 <6

≥8 6−8 Flight duration (hours)

studies failed to exclude a moderate effect because controls may have been identified, in part, by a history of recent travel [24].

Trials of prophylactic agents Elastic stockings appear to be effective in preventing flight-related DVT. Pooled data of randomized trials showed asymptomatic DVT, identified by screening with ultrasound, in 1 of 868 passengers (0.1%) who received elastic stockings and 38 of 883 passengers (4.3%) who received no prophylaxis [25–29]. Low-molecular-weight heparin, in small numbers of patients, also appeared to prevent DVT. No DVT was found in 84 passengers who received enoxaparin 1 mg/kg 2–4 hours before travel, whereas 4 of 82 (4.9%) untreated passengers developed DVT [30]. Aspirin, in small numbers of patients, was ineffective in preventing flight-related DVT. Among patients who received aspirin 400 mg/day for 3 days, starting 12 hours before travel, DVT (usually asymptomatic) developed in 3 of 84 (3.6%) versus 4 of 82 (4.9%) in controls [30].

Recommendations Recommendations of the American College of Chest Physicians for travel ≥6 hours are [31]. General measures (Clear risk/benefit, strong recommendation):

1 Avoid constrictive clothing around the lower extremities or waist. 2 Avoid dehydration.

≥9

Figure 24.1 Prevalence of pulmonary embolism (PE) per million passenger arrivals according to duration of flight. (Data from Hertzberg et al. [15] and Perez-Rodriguez et al. [16].)

3 Stretch calf muscles frequently. Specific measures:

1 Below knee graduated compression stockings providing 15–30 mm Hg at the ankle (Unclear risk/benefit, weak recommendation), or 2 Single prophylactic dose of low-molecularweight heparin injected prior to departure (Unclear risk/benefit, weak recommendation), 3 It is recommended that aspirin not be used (Clear risk/benefit, strong recommendation).

References 1 Homans J. Thrombosis of the deep leg veins due to prolonged sitting. N Engl J Med 1954; 250: 148–149. 2 Giangrande PL. Air travel and thrombosis. Br J Haematol 2002; 117: 509–512. 3 Symington IS, Stack BH. Pulmonary thromboembolism after travel. Br J Dis Chest 1977; 71: 138–140. 4 Tardy B, Page Y, Zeni F et al. Phlebitis following travel. Presse Med 1993; 22: 811–814. 5 Simpson K. Shelter deaths from pulmonary embolism. Lancet 1940; 2: 744. 6 Cruickshank JM, Gorlin R, Jennett B. Air travel and thrombotic episodes: the economy class syndrome. Lancet 1988; 2: 497–498. 7 Collins J. Thromboembolic disease related to air travel: what you need to know. Semin Roentgen 2005; 40: 1–2. 8 Ansari MT, Cheung BM, Qing Huang J, Eklof B, Karlberg JP. Traveler’s thrombosis: a systematic review. J Travel Med 2005; 12: 142–154. 9 Arfvidsson B. Risk factors for venous thromboembolism following prolonged air travel: a “prospective” study. Cardiovasc Surg 2001; 9: 158–159.


BLUK077-Stein

March 12, 2007

CHAPTER 24

15:45

121

Air travel as a risk for PE and DVT

10 Paganin F, Laurent Y, Gauzere BA, Blanc P, Roblin X. Pulmonary embolism on non-stop flights between France and Reunion Island. Lancet 1996; 347: 1195–1196. 11 Ribier G, Zizka V, Cysique J, Donatien Y, Glaudon G, Ramialison C. Venous thromboembolic complications following air travel. Retrospective study of 40 cases recorded in Martinique. Rev Med Interne 1997; 18: 601– 604. 12 Eklof B, Kistner RL, Masuda EM, Sonntag BV, Wong HP. Venous thromboembolism in association with prolonged air travel. Dermatol Surg 1996; 22: 637–641. 13 Mercer A, Brown JD. Venous thromboembolism associated with air travel: a report of 33 patients. Aviat Space Environ Med 1998; 69: 154–157. 14 Sarvesvaran R. Sudden natural deaths associated with commercial air travel. Med Sci Law 1986; 26: 35–38. 15 Hertzberg SR, Roy S, Hollis G, Brieger D, Chan A, Walsh W. Acute symptomatic pulmonary embolism associated with long haul air travel to Sydney. Vasc Med 2003; 8: 21–23. 16 Perez-Rodriguez E, Jimenez D, Diaz G et al. Incidence of air travel-related pulmonary embolism at the Madrid-Barajas airport. Arch Intern Med 2003; 163: 2766– 2770. 17 Hughes RJ, Hopkins RJ, Hill S et al. Frequency of venous thromboembolism in low to moderate risk long distance air travellers: the New Zealand Air Traveller’s Thrombosis (NZATT) study. Lancet 2003; 362: 2039–2044. 18 Lapostolle F, Surget V, Borron SW et al. Severe pulmonary embolism associated with air travel. N Engl J Med 2001; 345: 779–783. 19 Ferrari E, Chevallier T, Chapelier A, Baudouy M. Travel as a risk factor for venous thromboembolic disease: a casecontrol study. Chest 1999; 115: 440–444. 20 Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000; 160: 3415–3420. 21 Arya R, Barnes JA, Hossain U, Patel RK, Cohen AT. Longhaul flights and deep vein thrombosis: a significant risk

22

23

24 25

26

27

28

29

30

31

only when additional factors are also present. Br J Haematol 2002; 116: 653–654. Kraaijenhagen RA, Haverkamp D, Koopman MM, Prandoni P, Piovella F, Buller HR. Travel and risk of venous thrombosis. Lancet 2000; 356: 1492–1493. Ten Wolde M, Kraaijenhagen RA, Schiereck J et al. Travel and the risk of symptomatic venous thromboembolism. Thromb Haemost 2003; 89: 499–505. Gallus AS, Goghlan DC. Travel and venous thrombosis. Curr Opin Pulm Med 2002; 8: 372–388. Belcaro G, Geroulakos G, Nicolaides AN, Myers KA, Winford M. Venous thromboembolism from air travel: the LONFLIT study. Angiology 2001; 52: 369–374. Scurr JH, Machin SJ, Bailey-King S, Mackie IJ, McDonald S, Smith PD. Frequency and prevention of symptomless deep-vein thrombosis in long-haul flights: a randomised trial. Lancet 2001; 357: 1485–1489. Belcaro G, Cesarone MR, Shah SS et al. Prevention of edema, flight microangiopathy and venous thrombosis in long flights with elastic stockings. A randomized trial: The LONFLIT 4 Concorde Edema-SSL Study. Angiology 2002; 53: 635–645. Cesarone MR, Belcaro G, Nicolaides AN et al. The LONFLIT4-Concorde–Sigvaris Traveno stockings in long flights (EcoTraS) study: a randomized trial. Angiology 2003; 54: 1–9. Cesarone MR, Belcaro G, Errichi BM et al. The LONFLIT4–Concorde deep venous thrombosis and edema study: prevention with travel stockings. Angiology 2003; 54: 143–154. Cesarone MR, Belcaro G, Nicolaides AN et al. Venous thrombosis from air travel: the LONFLIT3 study– prevention with aspirin vs low-molecular-weight heparin (LMWH) in high-risk subjects: a randomized trial. Angiology 2002; 53: 1–6. Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism: the seventh ACCP conference on antithrombotic and thrombolytic therapy. Chest 2004; 126: 338S–400S.


BLUK077-Stein

April 14, 2007

25

14:57

CHAPTER 25

Estrogen-containing oral contraceptives and venous thromboembolism

Risks of venous thromboembolic disease in users of estrogen-containing oral contraceptives compared with nonusers Most preparations of estrogen-containing oral contraceptives consist of a mixture of estrogen and progesterone [1]. The US Food and Drug Administration in 1980 recommended the use of the lowest possible dose of estrogen for birth control [2]. In the past it had been thought that different types of progestins did not affect the risk of deep venous thrombosis (DVT) [3], but more recent data indicate that some third generation progestins have double the risk of second generation progestins [4]. In most countries, medications containing 50 μg of ethinyl estradiol or its equivalent are no longer available except as a shortterm postcoital medication [1]. Modern-day oral contraceptives contain 20–35 μg of ethinyl estradiol [5, 6]. Although the relative risk of VTE is higher among users of oral estrogen-containing contraceptives than nonusers [5, 7], the absolute risk is low [8]. Review by Vandenbroucke et al. [8] showed relative risks for VTE that, in general, ranged from 2.5 to 6.1 [9–12] although one study showed a relative risk of 11 [13]. A review by Lewis [5] of second and third generation estrogen-containing oral contraceptives, showed odds ratios that ranged from 0.8 to 2.3 [5, 9, 14–19]. Importantly, an absolute baseline risk <1/10,000 patients/ year increased to only 3 to 4/10,000 patients/year during the time oral contraceptives are being used [8]. Among 234,218 users of estrogen-containing oral contraceptives from 1980 to 1986, at doses that ranged from <50 μg/day to >50 μg/day, inpatient diagnoses of PE (pulmonary embolism) or DVT were

122

made in only 142 women (5.8/10,000 oral contraceptive users/year) [20]. Others, in women taking <50 μg/day of estrogen-containing contraceptives, reported VTE in 4.2/10,000 users/year [20], 4.7/10,000 [6] and 2.5/10,000 users/year [21]. Pooled data showed VTE in 127/417,915 patient-years (3.0/10,000 contraceptive users/year) [6, 20, 21]. Users of a patch designed to deliver 20 μg of ethinyl estradiol/day and 150 μg of norelgestromin/day over a period of 1 week [22] showed VTE in 4.1/10,000 users/year [21] and 5.3/10,000 users/year [6]. With earlier generation estrogen-containing oral contraceptives evaluated in 1961–1969, the incidence of DVT among users was higher than reported in recent years, 9 of 4965 (18/10,000), but higher rates were also reported among nonusers, 8 of 4933 (16/10,000) [23].

Relative risk in relation to dose The relative risk for VTE in women using oral contraceptives containing 50 μg of estrogen, compared with users of oral contraceptives that contained <50 μg was 1.5 [20]. The relative risk for VTE in women using oral contraceptives containing >50 μg of estrogen, compared with users of oral contraceptives that contained <50 μg was 1.7 [20]. No difference in the risk of VTE was found with various levels of low doses of 20, 30, 40, and 50 μg/day [3]. With doses of estrogen of 50 μg/day, the rate of VTE was 7.0/10,000 contraceptive users/year and with >50 μg/day, the rate of VTE was 10.0/10,000 oral contraceptive users/year [20] (Figure 25.1). Some, however, found no appreciable difference in the relative risk of VTE in relation to low or higher estrogen doses [11].


April 14, 2007

CHAPTER 25

14:57

VTE/10,000 estrogen users/yr

10.0 10 9 8 7 6 5 4 3 2 1 0

123

Estrogen-containing oral contraceptives and VTE

7.0

3.0

Relative risk of VTE (estrogen users vs. nonusers)

BLUK077-Stein

4 3.5 3 2.5 2 1.5 1 0.5 0

3.7

1.9 1.0

<20

1.2

21−25

26−30

>30

Body mass index (kg/m2) <50

50 Estrogen (μg/day)

>50

Figure 25.1 Venous thromboembolism (VTE) per 10,000 estrogen using patients per year according to daily dose. (Data from Gerstman et al. [20], Jick et al. [6] and Drug Safety Final Report [21]. Data with estrogen <50 μg/day were pooled.)

Relative risk according to duration of use Reports of risk of VTE in relation to the duration of use of oral contraceptives are inconsistent. Some showed relative risks increased as the duration of use of estrogen-containing oral contraceptives increased [13]. The relative risks were 0.7 in women who used oral contraceptives <1 year, 1.4 for those who used oral contraceptives 1–4 years and 1.8 in those who used it ≥5 years [13]. Others showed quite the opposite effect with a decreasing relative risk with duration of use [3]. The relative risk for DVT or PE was 5.1 with use <1 year, 2.5 with use 1–5 years, and 2.1 with use >5 years [3]. Finally, some showed the risk to be unaffected by the duration of use [11].

Estrogen-containing oral contraceptives in combination with smoking Users of oral contraceptives in Europe who smoked ≥10 cigarettes/day had odds ratios of 2.59 compared with users of oral contraceptives who did not smoke [11]. Users of oral contraceptives in developing countries who smoked ≥10 cigarettes/day had odds ratios of 1.22 compared with users of oral contraceptives who did not smoke [11]. Users of oral contraceptives who smoked <10 cigarettes/day did not have an increased risk of VTE compared to users of oral contraceptives who did not smoke [11].

Figure 25.2 Relative risk of venous thromboembolism (VTE) comparing estrogen users to nonusers shown in relation to body mass index (BMI). (Data from Lidegaard et al. [3].)

Estrogen-containing oral contraceptives and obesity The odds ratio for VTE comparing users of oral estrogen-containing oral contraceptives with matched nonusers of various body mass indexes (BMIs), showed that the combination of obesity with oral contraceptives carried a higher relative risk [3] (Figure 25.2). The World Health Organization reported higher odds ratios with oral contraceptives among women with BMIs >25 kg/m2 [11].

Estrogen-containing oral contraceptives and postoperative VTE The possibility of an increased risk of postoperative thromboembolism with oral contraceptive use was raised by Vessey et al. in 1970 [24]. The risk of postoperative PE appears to have increased in women who use oral contraceptives, even when the oral contraceptives have a low estrogen content [2]. In PIOPED, if PE was suspected, and if the women who underwent surgery used oral contraceptives, 50% had PE. If PE was suspected, and if the women who underwent surgery did not use oral contraceptives, 12% had PE [2].

References 1 Carter CJ. Epidemiology of venous thromboembolism. In: Hull RD & Pineo GF, eds. Disorders of Thrombosis. W. B. Saunders Co., Philadelphia, PA, 1996: 159–174. 2 Quinn DA, Thompson BT, Terrin ML et al. A prospective investigation of pulmonary embolism in women and men. JAMA 1992; 268: 1689–1696.


BLUK077-Stein

April 14, 2007

14:57

124

3 Lidegaard O, Edstrom B, Kreiner S. Oral contraceptives and venous thromboembolism. A case–control study. Contraception 1998; 57: 291–301. 4 Tanis BC, Rosendaal FR. Venous and arterial thrombosis during oral contraceptive use: risks and risk factors. Semin Cardiovasc Med 2003; 3: 69–83. 5 Lewis MA. The epidemiology of oral contraceptive use: a critical review of the studies on oral contraceptives and the health of young women. Am J Obstet Gynecol 1998; 179: 1086–1097. 6 Jick SS, Kaye JA, Russmann S, Jick H. Risk of nonfatal venous thromboembolism in women using a contraceptive transdermal patch and oral contraceptives containing norgestimate and 35 microg of ethinyl estradiol. Contraception 2006; 73: 223–228. 7 Realini JP, Goldzieher JW. Oral contraceptives and cardiovascular disease: a critique of the epidemiologic studies. Am J Obstet Gynecol 1985; 152: 729–798. 8 Vandenbroucke JP, Rosing J, Bloemenkamp KW et al. Oral contraceptives and the risk of venous thrombosis. N Engl J Med 2001; 344: 1527–1535. 9 Jick H, Jick SS, Gurewich V, Myers MW, Vasilakis C. Risk of idiopathic cardiovascular death and nonfatal venous thromboembolism in women using oral contraceptives with differing progestagen components. Lancet 1995; 346: 1589–1593. 10 Vessey M, Mant D, Smith A, Yeates D. Oral contraceptives and venous thromboembolism: findings in a large prospective study. BMJ (Clin Res Ed) 1986; 292: 526. 11 World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Venous thromboembolic disease and combined oral contraceptives: results of international multicentre case– control study. Lancet 1995; 346: 1575–1582. 12 Lewis MA, Heinemann LA, MacRae KD, Bruppacher R, Spitzer WO, The increased risk of venous thromboembolism and the use of third generation progestagens: role of bias in observational research. Transitional Research Group on Oral Contraceptives and the Health of Young Women. Contraception, 1996; 54: 5–13. [Erratum, Contraception 1996; 54: 121.] 13 Helmrich SP, Rosenberg L, Kaufman DW, Strom B, Shapiro S. Venous thromboembolism in relation to oral contraceptive use. Obstet Gynecol 1987; 69: 91–95. 14 World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contracep-

PART I

15

16

17

18

19

20

21

22

23

24

Prevalence, risks, and prognosis of PE and DVT

tion. Effect of different progestagens in low oestrogen oral contraceptives on venous thromboembolic disease. Lancet 1995; 346: 1582–1588. Spitzer WO, Lewis MA, Heinemann LA, Thorogood M, MacRae KD, for Transnational Research Group on Oral Contraceptives and the Health of Young Women. Third generation oral contraceptives and risk of venous thromboembolic disorders: an international case–control study. BMJ 1996; 312: 83–88. Bloemenkamp KW, Rosendaal FR, Helmerhorst FM, Buller HR, Vandenbroucke JP. Enhancement by factor V Leiden mutation of risk of deep-vein thrombosis associated with oral contraceptives containing a thirdgeneration progestagen. Lancet 1995; 346: 1593–1596. Farmer RDT, Preston TD. The risk of venous thromboembolism associated with low oestrogen oral contraceptives. J Obstet Gynaecol 1995; 1: 13–20. Farmer RD, Lawrenson RA, Thompson CR, Kennedy JG, Hambleton IR. Population-based study of risk of venous thromboembolism associated with various oral contraceptives. Lancet 1997; 349: 83–88. Herings RMC, de Boer A, Urquhart J, Leufkens HGM. Non-causal explanations for the increased risk of venous thromboembolism among users of third generation oral contraceptives [Abstract]. Pharmacoepidemiol Drug Saf 1996; 5(suppl 1): S88. Gerstman BB, Piper JM, Tomita DK, Ferguson WJ, Stadel BV, Lundin FE. Oral contraceptive estrogen dose and the risk of deep venous thromboembolic disease. Am J Epidemiol 1991; 133: 32–37. i3 Drug Safety Final Report. The risk of venous thromboembolism, myocardial infarction, and ischemic stroke among women using the transdermal contraceptive system compared to women using norgestimate-containing oral contraceptives with 35 μg of ethinyl estradiol. Prepared for Johnson and Johnson PRD, August 3, 2006. Abrams LS, Skee D, Natarajan J, Wong FA. Pharmacokinetic overview of Ortho EvraTM /EvraTM . Fertility Sterility 2002; 77(Suppl 2): S3–S12. Fuertes-de la Haba A, Curet JO, Pelegrina I, Bangdiwala I. Thrombophlebitis among oral and nonoral contraceptive users. Obstet Gynecol 1971; 38: 259–263. Vessey M, Doll R, Fairbain A, Glober G. Postoperative thromboembolism and the use of oral contraceptives. BMJ 1970; 3: 123–126.


BLUK077-Stein

March 11, 2007

26

7:7

CHAPTER 26

Obesity as a risk factor in venous thromboembolism

Obesity was shown to be a risk factor for pulmonary embolism (PE) and deep venous thrombosis (DVT) [1]. Among hospitalized patients diagnosed with obesity, based on data from the National Hospital Discharge Survey (NHDS) [2], 91,000 of 12,015,000 (0.8%) had PE [1] (Figure 26.1). Among hospitalized patients who were not diagnosed with obesity, PE was diagnosed in 2,366,000 of 691,000,000 (0.3%). Deep venous thrombosis was diagnosed in 243,000 of 12,015,000 (2.0%) patients diagnosed with obesity, and in 5,524,000 of 691,000,000 (0.8%) who were not diagnosed with obesity. The relative risk of PE, comparing obese patients with non-obese patients, was 2.21 and for DVT it was 2.50 [1]. The relative risks for PE and DVT were agedependent (Table 26.1). Obesity had the greatest impact on patients <40 years of age, in whom the relative risk for PE in obese patients was 5.19 and the relative risk for DVT was 5.20 (Table 26.1) [1]. Obese females had a greater relative risk for DVT than obese males, 2.75 versus 2.02 [1]. The prevalence of both PE and DVT in hospitalized obese females was

PE, DVT (%)

2.5

DVT

2 2.0 1.5 PE

1

DVT PE

0.8

0.5 0

0.8

0.3 Non-obese

Obese

Figure 26.1 Pulmonary embolism (PE) and deep venous thrombosis (DVT) in hospitalized patients from 1979 to 1999 showing the prevalence in obese and non-obese patients. (Data based on Stein et al. [1].)

higher than in obese males (Figure 26.2). In females <40 years of age, the relative risk for DVT comparing obese to non-obese patients was 6.10. In males <40 years of age, the relative risk for DVT was 3.71. The proportion of hospitalized patients diagnosed with obesity was within a narrow range (1.4–2.4%) over the 21-year period of observation from 1979 to 1999, indicating consistency in the diagnostic process

Table 26.1 Relative risks of pulmonary embolism and deep venous thrombosis according to age among obese and non-obese patients. Obese vs. non-obese Pulmonary embolism Age groups

Relative risk

(95% CI)

Deep venous thrombosis Relative risk

(95% CI)

<40

5.19

(5.11–5.28)

5.20

(5.15–5.25)

40–49

1.94

(1.91–1.97)

2.13

(2.11–2.15)

50–59

1.25

(1.23–1.27)

1.67

(1.65–1.68)

60–69

1.42

(1.40–1.44)

1.88

(1.87–1.90) (1.87–1.91)

70–79

2.07

(2.04–2.10)

1.89

>80

3.15

(3.08–3.22)

2.16

(2.12–2.20)

All ages

2.21

(2.20–2.23)

2.50

(2.49–2.51)

Reprinted from Stein et al. [1], with permission from Elsevier.

125


BLUK077-Stein

March 11, 2007

7:7

126

PART I

DVT

PE, DVT (%)

2.5 DVT

2 1.5 1 0.5 0

PE

PE

2.2

1.7

0.6 Obese male

0.8 Obese females

Figure 26.2 Pulmonary embolism (PE) and deep venous thrombosis (DVT) in hospitalized patients from 1979 to 1999 showing the prevalence in obese men and obese women. (Data based on Stein et al. [1].)

[1]. Previous investigators used several indices of obesity including a body mass index (BMI) >35 kg/m2 as well as BMI 30–35 kg/m2 [3], >30 kg/m2 [4, 5], BMI ≥29 kg/m2 [6], weight >20% of median recommended weight for height [3], and waist circumference ≥100 cm [7]. In 28–33% of patients reported by Anderson and associates, the physicians’ assessment was accepted [8, 9]. It is likely that all patients diagnosed with obesity in the NHDS database were in fact obese, irrespective of the criteria used. However, some obese patients may not have had a listed discharge diagnosis of obesity, and they would have been included in the non-obese group. This would have tended to reduce the relative risk of obesity in venous thromboembolism (VTE). The relative risks for PE and DVT were similar to relative risks reported in smaller investigations that used defined criteria for obesity. In women with a BMI >30 kg/m2 , the relative risk for DVT, compared to non-obese women was 2.4 [10]. In women with a BMI ≥29 kg/m2 , the relative risk for PE was 2.9 [6]. In the Framingham Study, obesity was a risk factor only in women [4]. Coon and Coller also showed that obesity was a risk factor only in women. Oral contraceptives in obese women increased the relative risk of DVT to 9.8 [10]. Men with a waist circumference ≥100 cm had a relative risk of 3.9 for VTE compared to men with smaller waists [7]. Among men and women together, the risk ratio for DVT, comparing obese to non-obese patients, was 2.39 [11]. Some found that obesity was not an independent risk factor for VTE [12]. Obesity has been suggested to be a risk factor for fatal PE since 1927 [13]. Investigations that reported an increased risk due to obesity have been criticized because they failed to control for hospital confinement

Prevalence, risks, and prognosis of PE and DVT

or other risk factors [12]. High proportions of patients with venous thromboembolic disease have been found to be obese [8, 9], but the importance of the association is diminished because of the high proportion of obesity in the general population [14]. Some investigations showed an increased risk ratio for DVT or PE, in women [4, 6, 10, 15], but data in men were less compelling. One investigation showed obesity to be a risk factor in men [7] and two did not [4, 15]. Some found no evidence that obesity was an independent risk factor in men or women [12]. Case series of morbidly obese patients (>100 pounds overweight or twice ideal weight) who underwent gastric bypass surgery, showed only a small incidence of postoperative VTE [5, 16]. Enoxaparin was shown to be effective for thromboprophylaxis in morbidly obese patients following bariatric surgery [17]. With various dosing regimens among 544 patients, PE occurred in 0.7% and none developed DVT [17]. All PE occurred after the cessation of enoxaparin, 7 days to 1 month after operation.

References 1 Stein PD, Beemath A, Olson RE. Obesity as a risk factor in venous thromboembolism. Am J Med 2005; 118: 978–980. 2 US Department of Health and Human Services, Public Health Service, National Center for Health Statistics National Hospital Discharge Survey 1979–1999 Multiyear Public-Use Data File Documentation. Available at: http://www.cdc.gov/nchs/about/major/hdasd/nhds.htm. 3 Farmer RD, Lawrenson RA, Todd JC et al. A comparison of the risks of venous thromboembolic disease in association with different combined oral contraceptives. Br J Clin Pharmacol 2000; 49: 580–590. 4 Goldhaber SZ, Savage DD, Garrison RJ et al. Risk factors for pulmonary embolism. The Framingham Study. Am J Med 1983; 74: 1023–1028. 5 Printen KJ, Miller EV, Mason EE, Barnes RW. Venous thromboembolism in the morbidly obese. Surg Gynecol Obstet 1978; 147: 63–64. 6 Goldhaber SZ, Grodstein F, Stampfer MJ et al. A prospective study of risk factors for pulmonary embolism in women. JAMA 1997; 277: 642–645. 7 Hansson PO, Eriksson H, Welin L, Svardsudd K, Wilhelmsen L. Smoking and abdominal obesity: risk factors for venous thromboembolism among middle-aged men: “the study of men born in 1913.” Arch Intern Med 1999; 159: 1886–1890. 8 Anderson FA, Jr, Wheeler HB, Goldberg RJ et al. A population-based perspective of the hospital incidence


BLUK077-Stein

March 11, 2007

CHAPTER 26

9

10

11

12

7:7

Obesity as a risk factor in VTE

and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med 1991; 151: 933–938. Anderson FA, Jr, Wheeler HB, Goldberg RJ, Hosmer DW, Forcier A. The prevalence of risk factors for venous thromboembolism among hospital patients. Arch Intern Med 1992; 152: 1660–1664. Abdollahi M, Cushman M, Rosendaal FR. Obesity: risk of venous thrombosis and the interaction with coagulation factor levels and oral contraceptive use. Thromb Haemost 2003; 89: 493–498. Samama MM. An epidemiologic study of risk factors for deep vein thrombosis in medical outpatients: the Sirius study. Arch Intern Med 2000; 160: 3415–3420. Heit JA, Silverstein MD, Mohr DN et al. The epidemiology of venous thromboembolism in the community. Thromb Haemost 2001; 86: 452–463.

127

13 Snell AM. The relation of obesity to fatal postoperative pulmonary embolism. Arch Surg 1927; 15: 237– 244. 14 Hedley AA, Ogden CL, Johnson CL et al. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999–2002. JAMA 2004; 291: 2847–2850. 15 Coon WW, Coller FA. Some epidemiologic considerations of thromboembolism. Surg Gynecol Obstet 1959; 109: 487–501. 16 Kerstein MD, McSwain NE, Jr, O’Connell RC, Webb WR, Brennan LA. Obesity: is it really a risk factor in thrombophlebitis? South Med J 1987; 80: 1236–1238. 17 Hamad GG, Choban PS. Enoxaparin for thromboprophylaxis in morbidly obese patients undergoing bariatric surgery: findings of prophylaxis against VTE outcomes in bariatric surgery patients receiving enoxaparin (PROBE) study. Obesity Surg 2005; 15: 1368–1374.


BLUK077-Stein

April 6, 2007

27

15:23

CHAPTER 27

Hypercoagulable syndrome

Inherited thrombophilia Patients with inherited thrombophilia (Table 27.1) tend to have clinical episodes of venous thromboembolism (VTE) before 45 years of age, and episodes tend to be recurrent [1].

Antithrombin III deficiency Antithrombin III is a naturally occurring anticoagulant that inactivates a number of enzymes in the coagulation cascade (factors IIa, IXa, Xa, XIa, and XIIa) (Figure 27.1). A diagram of the coagulation cascade is shown in Chapter 85, Figure 85.1. Antithrombin III deficiency is inherited as an autosomal dominant trait with heterozygotes having an increased risk of VTE [2]. There are two types of antithrombin III deficiency [3]. In type I deficiency, there is a reduction of functional antithrombin and in type II deficiency, there is an abnormal molecule [3]. Heterozygosity for antithrombin III deficiency is found in about 4% of families with inherited thrombophilia, in 1% of patients with a first episode of deep venous thrombosis (DVT), and

in 0.02% of healthy individuals [4] (Table 27.2). The prevalence of antithrombin III deficiency among patients with thrombosis (predominantly venous thrombosis) ranges from 0.5 to 8% [5]. Deep venous thrombosis, with or without pulmonary embolism was present in 90% of patients with antithrombin III deficiency [1]. Patients with antithrombin III deficiency have a 8â&#x20AC;&#x201C;10 times greater risk of developing thrombosis than individuals with normal coagulation [1, 6]. The antithrombinâ&#x20AC;&#x201C;heparin cofactor assay using a factor Xa and a thrombin inhibition assay are laboratory screening tests for this disorder. Since antithrombin III is a cofactor for heparin, heparin will not be effective in patients with antithrombin III deficiency [7]. In fact, heparin resistance may be an indication of antithrombin III deficiency [7].

Protein C deficiency Protein C deficiency is inherited as an autosomal dominant disorder and heterozygosity is a significant risk factor for VTE [8]. Two types of protein C deficiency have been reported [9]. Type I deficiency is a quantitative deficiency with decreased amounts protein C

Table 27.1 Inherited and acquired thrombophilic factors. Inherited thrombophilic factors Antithrombin III deficiency

Antiphospholipid syndrome

Protein C deficiency

Heparin induced thrombocytopenia

Protein S deficiency

Dysfibrinogenemia

Factor V Leiden

Myeloproliferative disorders

Prothrombin 20210A mutation

Malignancy

Elevated levels of factor VIII Elevated levels of factor XI Heparin cofactor II deficiency Dysfibrinogenemia Decreased levels of plasminogen Decreased levels of plasminogen activator Hyperhomocystenemia Sticky platelet syndrome

128

Acquired thrombophilic factors


BLUK077-Stein

April 6, 2007

CHAPTER 27

15:23

129

Hypercoagulable syndrome in VTE

XIIa

Tissue factor (TF) VIIa TF/VIIa complex

XIa

XI IX X

AT III IXa AT III

VIIIa Figure 27.1 Simplified coagulation cascade showing sites of action of antithrombin III (AT III). Deficiency of AT III results in failure to naturally inhibit the action of factors XIIa, XIa, IXa, Xa, and IIa.

activity due to reduced synthesis. Type II deficiency is a qualitative defect due to a defective protein C molecule [9]. Heterozygosity for protein C deficiency is found in about 6% of families with inherited thrombophilia, in 3% of patients with a first episode of unexplained DVT, and in 0.3% of healthy individuals [4] (Table 27.2). The prevalence of protein C deficiency among patients with thrombosis ranges from 1.5 to 11.5% [5]. Deep venous thrombosis, with or without pulmonary embolism was present in 88% of patients with protein C deficiency [1]. Patients with Protein C deficiency have a 4â&#x20AC;&#x201C;10 times greater risk of thrombosis as compared with control groups with normal coagulation [1, 6, 10]. The best screening tests for deficiencies of protein C are functional assays that detect both quantitative and qualitative defects of protein C. Immunologic assays detect only quantitative deficiencies of protein C [7]. Coagulation assays for protein C can give falsely low values if the factor V Leiden mutation is present, and as a result, the presence of factor V Leiden mutation should be assessed prior to application of coagulation assays for protein C. Short-term management of thrombosis among patients with protein C deficiency should be with heparin or low-molecular-weight heparin. A vitamin K antagonist such as warfarin should be considered for long-term treatment [7].

Xa IIa II prothrombin Va thrombin I fibrinogen

Ia fibrin

inadequate amount of normally functional protein S present in both the free and bound forms. Type II deficiency is a defective protein S molecule. Type III protein S deficiency is characterized by a low amount of free protein S, but an overall normal amount of total protein S. The large majority of patients with protein S deficiency have a type I deficiency, the prevalence of which is 6% in families with inherited thrombophilia and 1â&#x20AC;&#x201C;2% of patients with first time unexplained DVT [4] (Table 27.2). The prevalence of protein S deficiency among patients with thrombosis ranges from 1.5 to 13.2% [5]. Among patients with protein S deficiency, 74% develop DVT and 72% develop superficial thrombophlebitis [11]. Others reported DVT, with or without PE, to be present in 100% of patients with protein S deficiency [1]. Patients with protein S deficiency have an 8â&#x20AC;&#x201C;10 times higher risk of thrombosis compared with individuals with normal coagulation [1, 6]. As with protein C deficiency, the screening tests for protein S deficiency are functional assays that are most reliable when factor V Leiden mutation is ruled out. It is important to measure free protein S, since some patients with hereditary protein S deficiency may have normal or borderline total protein S levels. Treatment for protein S deficiency is the same as in protein C deficiency.

Protein S deficiency Protein S deficiency is inherited as an autosomal dominant disorder, with heterozygotes having an increased risk of VTE when compared with their unaffected family members [11]. There are three classifications of protein S deficiency. Type I deficiency results from an

Activated protein C resistance associated with factor V mutation (factor V Leiden) The Leiden mutation of factor V is the most common genetic abnormality associated with VTE. It is found in


130 —

6 — — — — — — —

Protein S deficiency

Factor V Leiden

Prothrombin 20210A mutation

Elevated factor VIII levels

Elevated factor XI levels

Heparin cofactor II deficiency

Dysfibrinogenemia

Malignancy

25

1–2

Prevalence of

10

25

6

20

1.5–13.2

1.5–11.5

0.5–8

unexplained VTE (%)

disorders in Pts with

* Prevalence of disorder among Caucasians in general population. Pts, patients; Gen’l, general; DVT, deep venous thrombosis; VTE, venous thromboembolism.

Antiphospholipid syndrome

Acquired thrombophilic factors

Hyperhomocystenemia

6

3

4

1

DVT (%)

Protein C deficiency

thrombophilia (%)

Antithrombin III deficiency

Inherited thrombophilic factors

Thrombophilic disorders

Patients with first episode of

family members with

5–10

2* 11*

— 5*

0.3

0.02

population (%)

disorders in Gen’l

Prevalence of

15–50

29–55

10

36

6

57

74–100

88

90

disorder (%)

VTE with

Frequency of

2

11

2.5

2.2

5–6

2.8

2–8

8–10

4–10

8–10

VTE

risk For

Relative

[38, 39]

[33–35]

[7, 10, 13, 23]

[22]

[21]

[10, 19]

[10, 13, 16]

[3, 10, 13]

[1, 10, 13]

[1, 4, 6, 11]

[1, 4, 6, 10]

[1, 4, 6]

References

April 6, 2007

Incidence of disorder among

Table 27.2 Incidence of various thrombophilic disorders and associated venous thromboembolism.

BLUK077-Stein

15:23


BLUK077-Stein

April 6, 2007

CHAPTER 27

15:23

131

Hypercoagulable syndrome in VTE

Tissue factor (TF)

XIIa

VIIa TF/VIIa complex

XI

XIa IX

IXa

X VIIIa

Figure 27.2 Simplified coagulation cascade showing the involvement of factor V Leiden in the coagulation pathway. Activated factor V Leiden is resistant to inhibition by activated protein C.

about 5% of Caucasians, but is extremely rare in people of African and Asian descent [7, 12, 13]. Factor V Leiden is an autosomal dominant condition in which the mutated factor V is resistant to inactivation by activated protein C (Figure 27.2). About 4â&#x20AC;&#x201C;7% of the general population is heterozygous for factor V Leiden and about 0.06â&#x20AC;&#x201C;0.25% of the population is homozygous for factor V Leiden [1] (Table 27.2). The prevalence of factor V Leiden among patients with unexplained venous thrombosis is 20% [13]. Deep venous thrombosis, with or without pulmonary embolism, was present in 57% of patients with factor V Leiden mutations [1]. Patients with factor V Leiden mutation presented with a 2- to 8-fold increase risk of thrombosis compared with individuals with normal coagulation [1, 10]. The relative risk of thrombosis for carriers was shown to have increased 7-fold for heterozygotes and 80-fold for homozygotes among patients <70 years of age with no malignancy [14]. One way to test for the factor V Leiden mutation is to measure the plasma clotting time in the absence and presence of activated protein C [3]. Newer versions of this test can be performed when patients are receiving anticoagulants [15]. The most direct test for factor V Leiden is DNA testing (polymerase chain reactionbased test) [3].

Prothrombin 20210A mutation The prothrombin 20210A allele is due to a glycine-toarginine transition in position 20210 of the prothrombin gene [3]. The abnormal gene causes increased concentrations of prothrombin. It is thought that the increased amount of circulating prothrombin upreg-

Xa Va Leiden II prothrombin

V Leiden Activated protein C

IIa thrombin

I fibrinogen

Ia fibrin

ulates the coagulation cascade [3]. The prevalence of prothrombin 20210A mutation among Caucasians in the general population is 2% [13]. About 6.2% of patients with the prothrombin 20210A gene mutation had venous thrombosis [3, 10, 13]. The diagnosis of prothrombin 20210A is based entirely on DNA analysis of the genes since it is the only genetic coagulation defect that cannot be reliably diagnosed by functional or immunological tests [3].

Elevated factor VIII levels The prevalence of elevated factor VIII levels among Caucasians in the general population is 11% [13]. Patients with high levels of factor VIII (>1500 IU/L) had a 5-fold increase risk for thrombosis when compared to patients with lower levels of factor VIII (<1000 IU/L) [16]. Among patients presenting with an initial episode of unexplained DVT, 25% had factor VIII levels above 1500 IU/L [10, 13]. For every 100 IU/L rise in levels of factor VIII, the risk for a single episode of DVT increased 10% and the risk for recurrent DVT increased 24% [17]. Factor VIII may be increased during an acute illness [18]. A C-reactive protein level should therefore be measured to determine if the increase of factor VIII is likely to be an acute phase reactant [18]. Pregnancy and use of oral contraceptives may also raise levels of factor VIII [7].

Elevated factor XI levels Increased levels of factor XI, a component of the intrinsic pathway, has been implicated as a risk factor for VTE [19]. A 120.8% increase in levels of factor XI


BLUK077-Stein

April 6, 2007

15:23

132

PART I

was associated with a 2.2-fold increased risk of venous thrombosis [10, 19].

Disorders of plasmin generation

Heparin cofactor II deficiency Heparin cofactor II, also known as heparin cofactor A or dermatan sulfate cofactor, inhibits thrombin by forming a heparin cofactor II-thrombin complex [20]. The physiologic function of the molecule is unclear, but it may serve as a natural antithrombotic agent in the presence of dermatan sulfate. Inherited deficiency of heparin cofactor II is rare. It is inherited as an autosomal dominant trait. Limited studies have shown that heterozygosity for heparin cofactor II deficiency is not a likely risk for thrombosis without other concomitant risk factors [21]. Some, however, have reported thrombotic episodes in 36% of individuals with the deficiency [21].

Inherited dysfibrinogenemias A number of abnormal fibrinogens are associated with the hypercoagulable syndrome of dysfibrinogenemias. Ten percent of patients with dysfibrinogenemias develop venous thrombosis with arterial thrombosis being rare [22]. Hereditary dysfibrinogenemias is an autosomal dominant trait except for a few cases that appear to be recessive traits [22]. The thrombin time (TT) is the most sensitive screening test for dysfibrinogenemias. The thrombin time may be prolonged or shortened and reptilase time is also prolonged. Patients with recurrent thrombotic events may require long-term anticoagulation with Coumadin or subcutaneous heparin [20].

Prevalence, risks, and prognosis of PE and DVT

Dysplasminogenemia, decreased levels of plasminogen, decreased synthesis or release of tissue plasminogen activator, and increased levels of plasminogen activator inhibitor are associated with impaired fibrinolysis (Figure 27.3). Routine screening is thought not to be cost-effective and is not indicated [20]. Long-term treatment is with warfarin or low-molecular-weight heparin [7].

Hyperhomocystenemia Elevated plasma homocysteine levels constitute a risk factor for venous as well as arterial thrombosis [23]. Some suggest that hyperhomocystenemia contributes to venous and arterial thrombosis by its toxic effect on the vascular endothelium and on the clotting cascade [23]. Hyperhomocystenemia occurs in 5â&#x20AC;&#x201C;10% of the general population [7] and in 10% of patients with VTE [13], with a relative risk of venous thromboembolism of 2.5 [10, 23]. Hyperhomocystenemia can be diagnosed by measuring fasting homocysteine plasma levels [15]. Hyperhomocysteinemia can also be diagnosed by fluorescence polarization immunoassays [24]. Thrombosis is treated in a standard fashion in addition to vitamin B6, B12, and folic acid supplementation [7].

Sticky platelet syndrome Sticky platelet syndrome is inherited in an autosomal dominant trait that results in platelets that are hyperaggregable to epinephrine and/or adenosine

Thrombin Fibrinogen

Fibrin

Plasmin

Plasminogen

PA inhibitor

tPA

Fibrin degradation products Figure 27.3 Fibrinolysis pathway. Dysplasminogenemia, decreased levels of plasminogen, decreased synthesis or release of tissue plasminogen activator (tPA), and increased levels of plasminogen activator inhibitor (PAI) contribute to the maintenance of fibrin, thereby promoting a hypercoagulable state.


BLUK077-Stein

April 6, 2007

CHAPTER 27

15:23

133

Hypercoagulable syndrome in VTE

phosphate [7]. Both arterial, and to a lesser-degree venous thrombi may occur in patients with sticky platelet syndrome [25]. Treatment is with either 81 mg or 325 mg aspirin daily [25]. If aspirin fails, an adenosine diphosphate receptor agonist such as clopidrogel, may be used [25].

Acquired thrombophilia The antiphospholipid syndrome Antiphospholipid antibodies are associated with both arterial and venous thrombosis [26]. The most commonly detected subgroups of antiphospholipid antibodies are lupus anticoagulant antibodies, anticardiolipin antibodies, and anti-B2 -glycoprotein I antibodies [27]. Increased expression and activity of tissue factor is mediated by antibodies against phospholipidbinding proteins present in endothelial cells or other cells. The antigen–antibody reaction releases tissue factor which initiates the coagulation cascade leading to the prothrombotic state [27–29]. The exact mechanism by which these antibodies induce the transduction signal to produce tissue factor are not yet clarified [28]. Primary antiphospholipid syndrome occurs in patients without clinical evidence of an autoimmune disease, whereas secondary antiphospholipid syndrome occurs in association with autoimmune diseases. Systemic lupus erythematosus (SLE) is the most common autoimmune disorder associated with the antiphospholipd syndrome [27]. Among patients with SLE who had the antiphospholipid antibodies, anticardiolipin antibodies were present in 12–30% [30, 31] and lupus anticoagulant antibodies were present in 15–34% [31, 32]. Deep venous thrombosis, the most common manifestation of the antiphospolipid syndrome, occurs in 29–55% of patients with the syndrome, and about half of these patients have pulmonary emboli [33–35]. The estimated relative risk for a first episode of venous thromboembolism in patients with a thrombophilic defect as compared to healthy individuals is 11 [6]. Arterial thrombosis is less common than venous thrombosis in patients with antiphospholipid syndrome [33– 35]. Since there is no definitive association between specific clinical manifestations and particular subgroups of antiphospholipid antibodies, and patients may be negative according to one test and positive for another,

multiple tests for antiphospholipid antibodies are recommended [27].

Heparin-induced thrombocytopenia In patients with heparin-induced thrombocytopenia, platelet activation occurs due to the binding of heparin to platelet factor 4. This forms a platelet– heparin complex that causes endothelial cell injury and activates tissue factor and the coagulation cascade [20, 36]. Thrombosis, either venous or arterial, occurs in about 20% of patients with heparininduced thrombocytopenia [20]. Platelet transfusion in patients with heparin-induced thrombocytopenia is contraindicated because in such patients intravascular platelet aggregation also occurs, thus further contributing to the thrombosis [20]. Management involves cessation of heparin and switching to alternative anticoagulant options.

Acquired dysfibrinogenemia Acquired dysfibrinogenemia occurs most often in patients with severe liver disease [22]. The impairment of the fibrinogen is due to a structural defect caused by an increased carbohydrate content impairing the polymerization of the fibrin, depending on the degree of abnormality of the fibrinogen molecule [22]. Screening tests and treatment are the same as with acquired dysfibrinogenemia.

Myeloproliferative disorders Polycythemia vera, essential thrombocytopenia, and agnogenic myeloid metaplasia are associated with thrombocytosis and an increase in whole blood viscosity [20], thereby contributing to the hypercoagulable state. These complications may result because of altered interactions between platelets, white blood cells, or endothelial cells, due to either altered receptor expression, receptor–ligand interactions, or signaling events. Thrombosis may occur in arteries or veins [37].

Malignancy Cancer is the second most common cause of hypercoagulability, accounting for 10–20% of spontaneous DVTS [7]. Approximately 15% of patients with cancer


BLUK077-Stein

April 6, 2007

15:23

134

have clinical thombosis and up to 50% have thrombosis on autopsy [38]. The mechanism by which tumor cells contribute to thrombosis is not clear. It has been shown that tumor cells interact with thrombin and plasmin and directly influence thrombus formation [20]. Patients with malignancy have a 2-fold increase of developing venous thromboembolism [39].

Diagnostic approach to patients suspected of thrombophilia “Strongly” thrombophilic patients are those patients who sustained their first VTE event before 50 years of age, have a history of recurrent thrombosis, or have a first-degree relative with documented VTE events occurring before 50 years of age [40]. If one or more of these features are present, a complete evaluation for hereditary thrombophilia is appropriate [40]. Testing should be extended to their first-degree family members as well. Because most of the tests are not reliable during anticoagulation, it is preferable to postpone laboratory testing until after discontinuation of treatment [40].

References 1 Martinelli I, Mannucci PM, De Stefano V et al. Different risks of thrombosis in four coagulation defects associated with inherited thrombophilia: a study of 150 families. Blood 1998; 92: 2353–2358. 2 Thaler E, Lechner K. Antithrombin III deficiency and thromboembolism. Clin Haematol 1981; 10: 369–390. 3 Bertina RM. Factor V Leiden and other coagulation factor mutations affecting thrombotic risk. Clin Chem 1997; 43: 1678–1683. 4 Lane DA, Mannucci PM, Bauer KA et al. Inherited thrombophilia: Part 1. Thromb Haemost 1996; 76: 651–662. 5 Mateo J, Oliver A, Borrell M, Sala N, Fontcuberta J. Laboratory evaluation and clinical characteristics of 2,132 consecutive unselected patients with venous thromboembolism—results of the Spanish Multicentric Study on Thrombophilia (EMET-Study). Thromb Haemost 1997; 77: 444–451. 6 Weitz JI, Middeldorp S, Geerts W, Heit JA. Thrombophilia and new anticoagulant drugs. Hematology (Am Soc Hematol Educ Program) 2004: 424–438. 7 Thomas RH. Hypercoagulability syndromes. Arch Intern Med 2001; 161: 2433–2439. 8 Allaart CF, Poort SR, Rosendaal FR, Reitsma PH, Bertina RM, Briet E. Increased risk of venous thrombosis in carri-

PART I

9

10

11

12

13 14

15 16

17

18

19

20 21

22 23

24

Prevalence, risks, and prognosis of PE and DVT

ers of hereditary protein C deficiency defect. Lancet 1993; 341: 134–138. Reitsma PH, Bernardi F, Doig RG et al. Protein C deficiency: a database of mutations, 1995 update. On behalf of the Subcommittee on Plasma Coagulation Inhibitors of the Scientific and Standardization Committee of the ISTH. Thromb Haemost 1995; 73: 876–889. Kamphuisen PW, Rosendaal FR. Thrombophilia screening: a matter of debate. Neth J Med 2004; 62: 180– 187. Engesser L, Broekmans AW, Briet E, Brommer EJ, Bertina RM. Hereditary protein S deficiency: clinical manifestations. Ann Intern Med 1987; 106: 677–682. Bounameaux H. Factor V Leiden paradox: risk of deepvein thrombosis but not of pulmonary embolism. Lancet 2000; 356: 182–183. Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet 1999; 353: 1167–1173. Rosendaal FR, Koster T, Vandenbroucke JP, Reitsma PH. High risk of thrombosis in patients homozygous for factor V Leiden (activated protein C resistance). Blood 1995; 85: 1504–1508. Federman DG, Kirsner RS. An update on hypercoagulable disorders. Arch Intern Med 2001; 161: 1051–1056. Koster T, Blann AD, Briet E, Vandenbroucke JP, Rosendaal FR. Role of clotting factor VIII in effect of von Willebrand factor on occurrence of deep-vein thrombosis. Lancet 1995; 345: 152–155. Kraaijenhagen RA, in’t Anker PS, Koopman MM et al. High plasma concentration of factor VIIIc is a major risk factor for venous thromboembolism. Thromb Haemost 2000; 83: 5–9. Cumming AM, Shiach CR. The investigation and management of inherited thrombophilia. Clin Lab Haematol 1999; 21: 77–92. Meijers JC, Tekelenburg WL, Bouma BN, Bertina RM, Rosendaal FR. High levels of coagulation factor XI as a risk factor for venous thrombosis. N Engl J Med 2000; 342: 696–701. Nachman RL, Silverstein R. Hypercoagulable states. Ann Intern Med 1993; 119: 819–827. Adcock DM, Jensen R, Johns CS, Macy PA. Coagulation Handbook. Esoterix Coagulation, Aurora, Colorado, 2002: 25–27. Brick W, Burgess R, Faguet GB. Dysfibrinogenemia. WebMD. www.webmd.com. Last accessed June 26, 2006. den Heijer M, Koster T, Blom HJ et al. Hyperhomocysteinemia as a risk factor for deep-vein thrombosis. N Engl J Med 1996; 334: 759–762. Tripodi A, Negri B, Bertina RM, Mannucci PM. Screening for the FV:Q506 mutation–evaluation of thirteen plasmabased methods for their diagnostic efficacy in comparison with DNA analysis. Thromb Haemost 1997; 77: 436–439.


BLUK077-Stein

April 6, 2007

CHAPTER 27

15:23

Hypercoagulable syndrome in VTE

25 Mammen EF. Sticky platelet syndrome. Semin Thromb Hemost 1999; 25: 361–365. 26 Greaves M. Antiphospholipid antibodies and thrombosis. Lancet 1999; 353: 1348–1353. 27 Levine JS, Branch DW, Rauch J. The antiphospholipid syndrome. N Engl J Med 2002; 346: 752–763. 28 Amengual O, Atsumi T, Khamashta MA. Tissue factor in antiphospholipid syndrome: shifting the focus from coagulation to endothelium. Rheumatology (Oxford) 2003; 42: 1029–1031. 29 Ames PR. Antiphospholipid antibodies, thrombosis and atherosclerosis in systemic lupus erythematosus: a unifying ‘membrane stress syndrome’ hypothesis. Lupus 1994; 3: 371–377. 30 Merkel PA, Chang Y, Pierangeli SS, Convery K, Harris EN, Polisson RP. The prevalence and clinical associations of anticardiolipin antibodies in a large inception cohort of patients with connective tissue diseases. Am J Med 1996; 101: 576–583. 31 Cervera R, Khamashta MA, Font J et al. Systemic lupus erythematosus: clinical and immunologic patterns of disease expression in a cohort of 1,000 patients. The European Working Party on Systemic Lupus Erythematosus. Medicine (Baltimore) 1993; 72: 113–124. 32 Love PE, Santoro SA. Antiphospholipid antibodies: anticardiolipin and the lupus anticoagulant in systemic lupus erythematosus (SLE) and in non-SLE disorders. Prevalence and clinical significance. Ann Intern Med 1990; 112: 682–698.

135

33 Asherson RA, Khamashta MA, Ordi-Ros J et al. The “primary” antiphospholipid syndrome: major clinical and serological features. Medicine (Baltimore) 1989; 68: 366– 374. 34 Alarcon-Segovia D, Perez-Vazquez ME, Villa AR, Drenkard C, Cabiedes J. Preliminary classification criteria for the antiphospholipid syndrome within systemic lupus erythematosus. Semin Arthritis Rheum 1992; 21: 275– 286. 35 Vianna JL, Khamashta MA, Ordi-Ros J et al. Comparison of the primary and secondary antiphospholipid syndrome: a European Multicenter Study of 114 patients. Am J Med 1994; 96: 3–9. 36 Arepally GM, Mayer IMM. Antibodies from patients with heparin-induced thrombocytopenia stimulate monocytic cells to express tissue factor and secrete interleukin8. Blood 2001; 98: 1252–1254. 37 Kessler CM. Propensity for hemorrhage and thrombosis in chronic myeloproliferative disorders. Semin Hematol 2004; 41: 10–14. 38 Luzzatto G, Schafer AI. The prethrombotic state in cancer. Semin Oncol 1990; 17: 147–159. 39 Stein PD, Beemath A, Meyers FA, Skaf E, Sanchez J, Olson RE. Incidence of venous thromboembolism in patients hospitalized with cancer. Am J Med 2006; 119: 60– 68. 40 Bauer KA. The thrombophilias: well-defined risk factors with uncertain therapeutic implications. Ann Intern Med 2001; 135: 367–373.


BLUK077-Stein

March 12, 2007

II

15:49

PART II

Diagnosis of deep venous thrombosis

137


BLUK077-Stein

March 12, 2007

28

15:49

CHAPTER 28

Deep venous thrombosis of the lower extremities: clinical evaluation

Physical examination remains the initial diagnostic modality that calls attention to the potential diagnosis and upon which physicians must rely. Only 13 of 32 patients (41%) with autopsy proven deep venous thrombosis (DVT) who died of trauma or burns were diagnosed antemortem, and 63 of 118 patients (53%) with autopsy proven DVT who died of pulmonary embolism (PE) had antemortem clinical signs of DVT [1, 2]. Among 37 legs of patients screened by 125 I fibrinogen scans, the clinical signs showed a similar sensitivity (49%) [3]. The specificity of clinical signs among 16 legs evaluated by venography was also low (56%), but these data are biased because patients were selected due to clinical findings [4]. In 12 extremities of patients with DVT shown by dissection at autopsy, the sensitivity of ankle asymmetry ≥1.27 cm was 83%, Homans’ sign was 8%, and local tenderness was 41% [5]. The specificity among 18 extremities of patients in whom DVT was excluded by dissection at autopsy for both ankle asymmetry and Homans’ sign was 94% and the specificity of local tenderness was 89% [5]. Homans’ sign is active and/or passive dorsiflexion of the foot associated with any of the following: (1) pain, (2) incomplete dorsiflexion (with equal pressure applied) to prevent pain, or (3) flexion of the knee to release tension in the posterior muscles with dorsiflexion [6]. A Homans’ sign was present in 44% of patients with DVT of the lower leg, and in 60% of patients with femoral venous thrombosis [7]. The elicitation of pain with inflation of a blood pressure cuff around the calf to 60–150 mm Hg has been recommended as a test for DVT [8]. This test, however, was not shown to be more helpful than the assessment of direct tenderness or leg circumference [7].

Calf asymmetry indicates a need for noninvasive diagnostic tests of the lower extremities to determine whether DVT is present [9]. Asymmetry of the circumference of the ankle, calf, or thigh ≥1 cm has been shown in 90% of patients with DVT, but such asymmetry was also shown in 92% of patients with suspected DVT in whom the diagnosis was excluded [10]. A difference ≥3 cm of calf circumference, however, was associated with a high likelihood of having DVT [11, 12]. A difference of circumference of the calves ≥1 cm, measured 10 cm below the tibial tuberosity, is abnormal [9] and was defined as a quantitative sign. As the difference of calf circumference increased from ≥1 cm to ≥4 cm, the sensitivity decreased from 43% or 53% to 4% or 10% and the specificity increased from 57–98% (Table 28.1, Figure 28.1) [13]. The sensitivity and specificity of signs of DVT in two different diagnostic categories of patients was assessed, the results of which gave an envelope of values [13]. The first diagnostic category was 350 patients from the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED I) [14] with PE proven by pulmonary angiography. Deep venous thrombosis was assumed to be present based on the 83% and 91% incidence of DVT at autopsy of patients with PE [2, 15]. This category is advantageous for the investigation of the sensitivity of signs of DVT because more than 99% of these patients were identified on the basis of respiratory complaints associated with PE [16]. Although DVT was presumably present, the patients were not selected for evaluation because of signs of DVT in this diagnostic category. The second diagnostic category was 30 patients with suspected PE in PIOPED I in whom DVT was

139


BLUK077-Stein

March 12, 2007

15:49

140

PART II

Diagnosis of DVT

Table 28.1 Calf asymmetry with deep venous thrombosis. Positive leg test

PE +, Assumed DVT

Negative leg test

Circum diff (cm)

sensitivity [N/n (%)]

sensitivity [N/n (%)]

specificity [N1 /n1 (%)]

≥1

16/30 (53)

149/350 (43)

27/47 (57)

≥2

8/30 (27)

59/350 (17)

41/47 (87)

≥3

4/30 (13)

23/350 (7)

45/47 (96)

≥4

3/30 (10)

13/350 (4)

46/47 (98)

N, number of patients with sign; n, number of patients with deep venous thrombosis; N1 , number of patients with no sign; n1 , number of patients with no deep venous thrombosis; Circum diff, circumference difference of calves. Stein et al. [13], and based on unpublished data from PIOPED I.

diagnosed by objective leg tests of the lower extremities. The advantage of this diagnostic category is that the diagnosis of DVT was made with confidence. The disadvantage is that many patients presumably had leg tests obtained because of clinical manifestations suggestive of DVT. This bias would increase the apparent sensitivity of clinical findings. However, among the entire group of patients in whom leg tests were performed (30 with DVT and 47 with no DVT) 48 of 77 (62%) had no qualitative signs of DVT, and 29 of 77 (38%) had no qualitative signs or measured asymmetry. Therefore, a significant number of patients were referred for objective leg tests only because they had a suspicion of PE. The individual qualitative signs of DVT (edema, erythema, calf tenderness, palpable cord, Homans’ sign) showed a sensitivity of 47% or less, irrespective of whether DVT was diagnosed by objective tests of the

lower extremities or whether it was assumed to be present in patients with PE (Table 28.2) [13]. Edema was the most sensitive sign. Unilateral edema was not statistically significantly more sensitive or specific than bilateral edema. All signs showed a specificity of 83% or higher (Table 28.2). The specificities of signs did not differ to a statistically significant extent. The presence of any sign (edema, erythema, calf tenderness, palpable cord, or Homans’ sign) or measured asymmetry of the calves increased the sensitivity for detection of DVT above the sensitivity of a physical sign alone or asymmetry alone (Table 28.3). The specificity varied inversely with the sensitivity [13]. The combination of a sign on physical examination (edema, erythema, calf tenderness, palpable cord, or Homans’ sign) plus ipsilateral asymmetry was associated with a sensitivity of 33% or lower, but the specificity was 87% or higher (Table 28.4). Among the

100

Sensitivity (%)

80

60

Leg test +

40

PE +

20 0 0

20

40

60

False positives (%)

80

100

Figure 28.1 Relation of sensitivity of measured calf asymmetry to the frequency of false-positive findings. The upper curve (leg test +) shows data of patients in whom objective tests of the lower extremities showed deep venous thrombosis (DVT) to be present or absent. The lower curve (PE +) shows data of patients in whom DVT was estimated to be present because of pulmonary embolism (PE). The numbers indicate the measured difference of calf circumference. (Data from Stein et al. [13], based on unpublished data from PIOPED I.)


BLUK077-Stein

March 12, 2007

CHAPTER 28

15:49

141

Clinical evaluation of DVT

Table 28.2 Signs of deep venous thrombosis. Positive leg test

PE +, Assumed DVT

Negative leg test

Sign

sensitivity [N/n (%)]

sensitivity [N/n (%)]

specificity [ N1 /n1 (%)]

Edema, any

14/30 (47)

79/350 (23)

40/47 (85)

Edema, unilat

30/350 (9)

43/47 (91)

Edema, bilat

49/350 (14)

44/47 (94)

21/350 (6)

43/47 (91)

Erythema

5/30 (17)

Calf tender

5/30 (17)

47/350 (13)

39/47 (83)

Palp cord

0/30 (0)

8/350 (2)

46/47 (98)

Homan’s sig

1/30 (3)

9/350 (3)

44/47 (94)

Unilat, unilateral; bilat, bilateral; N, number of patients with sign; n, number of patients with deep venous thrombosis; N1 , number of patients with no sign; n1 , number of patients with no deep venous thrombosis. Stein et al. [13], and based on unpublished data from PIOPED I.

Table 28.3 Signs of deep venous thrombosis and/or calf asymmetry. Sign +/or circum

Positive leg test

PE +, Assumed DVT

Negative leg test

diff (cm)

sensitivity [N/n (%)]

sensitivity [N/n (%)]

specificity [N1 /n1 (%)]

Any sign +/or ≥1 cm

24/30 (80)

192/350 (55)

23/47 (49)

Any sign +/or ≥2 cm

20/30 (67)

131/350 (37)

33/47 (70)

Any sign +/or ≥3 cm

19/30 (63)

111/350 (32)

35/47 (74)

Any sign +/or ≥4 cm

18/30 (60)

104/350 (30)

35/47 (74)

N, number of patients with sign; n, number of patients with deep venous thrombosis; N1 , number of patients with no sign; n1 , number of patients with no deep venous thrombosis; circum diff, circumference difference of calves. Stein et al. [13], and based on unpublished data from PIOPED I.

Table 28.4 Ipsilateral signs of deep venous thrombosis and calf asymmetry. Sign and circum

Positive leg test

PE +, Assumed DVT

Negative leg test

diff (cm)

sensitivity [N/n (%)]

sensitivity [N/n (%)]

specificity [N1 /n1 (%)]

Any sign and ≥1

10/30 (33)

52/350 (15)

41/47 (87)

Any sign and ≥2

5/30 (17)

27/350 (8)

43/47 (91)

Any sign and ≥3

3/30 (10)

11/350 (3)

45/47 (96)

Any sign and ≥4

2/30 (7)

8/350 (2)

46/47 (98)

N, number of patients with sign; n, number of patients with deep venous thrombosis; N1 , number of patients with no sign; n1 , number of patients with no deep venous thrombosis; circum diff, circumference difference of calves. Stein et al. [13], and based on unpublished data from PIOPED I.


BLUK077-Stein

March 12, 2007

15:49

142

PART II

Diagnosis of DVT

Figure 28.2 Relation of sensitivity of combinations of qualitative signs and measured calf asymmetry to the frequency of false-positive findings. The upper curve (leg test +) shows data of patients in whom objective tests of the lower extremities showed deep venous thrombosis (DVT). The lower curve (PE +) shows data of patients in whom DVT was estimated to be present because of pulmonary embolism (PE). The numbers indicate the measured difference of calf circumference. Sign +/or asym indicates patients who had a qualitative sign and/or asymmetry. Sign + ipsi asym indicates patients who had a qualitative sign plus ipsilateral asymmetry. (Data from Stein et al. [13], based on unpublished data from PIOPED I.)

3–10% of patients who had one or more qualitative signs plus ≥3 cm calf asymmetry, the specificity for DVT was 96% [13]. Among patients in whom DVT was diagnosed by objective leg tests, signs of DVT or measured asymmetry showed a higher sensitivity than in patients in whom DVT was estimated to be present because of PE (Figure 28.2) [13]. If either a qualitative sign or measured asymmetry was present, the presence of one or both was more sensitive than a qualitative sign plus measured asymmetry. The former, however, showed more false-positive values than patients with qualitative signs and ipsilateral measured calf swelling. As the measured difference of calve circumferences increased from ≥1 cm to ≥4 cm in combination with qualitative signs, the sensitivity diminished, and the presence of false-positive values also diminished. Among patients with DVT diagnosed by objective leg tests, 2 of 30 (7%) had leg pain, but no qualitative signs or asymmetry ≥1 cm. An additional 1 of 30 (3%) reported swelling in the leg or foot, but had no leg pain, qualitative signs, or asymmetry ≥1 cm. The sensitivity for the detection of DVT of one or more symptoms, qualitative signs, or asymmetry ≥1 cm was 27 of 30 (90%). Among patients with DVT estimated to be present on the basis of PE, the addition of symptoms of tenderness or swelling to the qualitative or quantitative signs increased the sensitivity for the detection of DVT to 229 of 350 (65%) [13]. The addition of symptoms of tenderness or swelling to qualitative signs or measured asymmetry ≥1 cm

decreased the specificity of one or more of any of these findings to 18 of 47 (38%) [13]. These data show, therefore, that combinations of qualitative signs and measured asymmetry of the calves do not reliably identify patients with DVT, nor does the absence of such signs exclude DVT. However, in patients who have ≥3 cm asymmetry of the calves, the finding is 96% specific [13]. Meta-analysis identified covariates that provided diagnostic accuracy for DVT. Only malignancy and previous DVT were useful for ruling in DVT and recent immobilization, difference in calf diameter, and recent surgery were of borderline value [17]. The positive likelihood ratios [18] ranged from 1.76 to 2.71 indicating that none provided high certainty for a diagnosis [17]. Only absence of calf swelling and absence of a difference in calf diameter were somewhat (borderline) useful for ruling out DVT [17]. The negative likelihood ratios were 0.67 and 0.57 [17].

References 1 Sevitt S, Gallagher N. Venous thrombosis and pulmonary embolism: a clinico-pathological study in injured and burned patients. Br J Surg 1961; 48: 475–489. 2 Byrne JJ, O’Neil EE. Fatal pulmonary emboli. A study of 130 autopsy-proven fatal emboli. Am J Surg 1952; 83: 47–49. 3 Milne RM, Gunn AA, Griffiths JMT, Ruckley CX. Postoperative deep venous thrombosis: a comparison of diagnostic techniques. Lancet 1971; 2: 445–447. 4 Johnson WC. Evaluation of newer techniques for the diagnosis of venous thrombosis. J Surg Res 1974; 16: 473–481.


BLUK077-Stein

March 12, 2007

CHAPTER 28

15:49

Clinical evaluation of DVT

5 McLachlin J, Richards T, Paterson JC. An evaluation of clinical signs in the diagnosis of venous thrombosis. Arch Surg 1962; 85: 738–744. 6 Homans J. Disease of the veins. N Engl J Med 1944; 231: 51–60. 7 DeWeese JA, Rogoff SM. Phlebographic patterns of acute deep venous thrombosis of the leg. Surgery 1963; 53: 99– 108. 8 Lowenberg RI. Early diagnosis of phlebothrombosis with aid of a new clinical test. JAMA 1954; 155: 1566–1570. 9 Stein PD, Henry JW, Gopalakrishnan D, Relyea B. Asymmetry of the calves in the assessment of patients with suspected acute pulmonary embolism. Chest 1995; 107: 936–939. 10 Cranley JJ, Canos AJ, Sull WJ. The diagnosis of deep venous thrombosis: fallibility of clinical symptoms and signs. Arch Surg 1976; 111: 34–36. 11 Lambie JM, Mahaffy RG, Barber DC, Karmody AM, Scott MM, Matheson NA. Diagnostic accuracy in venous thrombosis. BMJ 1970; 2: 142–143. 12 Nypaver TJ, Shepard AD, Kiell CS, McPharlin M, Fenn N, Ernst CB. Outpatient duplex scanning for deep vein thrombosis: parameters predictive of a negative study result. J Vasc Surg 1993; 18: 821–826.

143

13 Stein PD, Henry JW, Relyea B. Sensitivity and specificity of physical examination of the lower extremity in evaluation of deep venous thrombosis. Unpublished data from PIOPED I. 14 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 15 Short DS. A survey of pulmonary embolism in a general hospital. BMJ 1952; 1: 790–796. 16 Stein PD, Saltzman HA, Weg JG. Clinical characteristics of patients with acute pulmonary embolism. Am J Cardiol 1991; 68: 1723–1724. 17 Goodacre SG, Sutton AJ, Sampson FC. Meta-analysis: the value of clinical assessment in the diagnosis of deep venous thrombosis. Ann Intern Med 2005; 143: 129– 139. 18 Jaeschke R, Guyatt GH, Sackett DL, for The EvidenceBased Medicine Working Group. Users’ guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? JAMA 1994; 271: 703– 707.


BLUK077-Stein

March 11, 2007

29

7:5

CHAPTER 29

Clinical model for assessment of deep venous thrombosis

Scoring systems for deep venous thrombosis Recognizing the difficulty in making a clinical diagnosis of acute deep venous thrombosis (DVT), a clinical model was developed which permits an assessment of the clinical probability [1] (Table 29.1). An earlier model with more features had been developed by the same group [2]. Although it gave good results, it was more complex and seems to have been abandoned. A clinical model for assessment of probability of DVT may be particularly useful when combined with another test, such as D-dimer. This will be discussed in Chapter 32.

Table 29.1 Clinical model for assessment of probability of deep venous thrombosis. Feature Active cancer (treatment ongoing or within

Score 1

previous 6 months or palliative) Paralysis, paresis, or recent immobilization of lower

1

extremities Recently bedridden >3 days or major surgery within

1

4 weeks Localized tenderness along the deep venous system

1

Entire leg swollen

1

Calf swelling >3 cm compared with the

1

asymptomatic leg (measured 10 cm below tibial tuberosity) Pitting edema (greater in the symptomatic leg) Collateral superficial veins (non-varicose) Alternative diagnosis as likely or greater than that

1 1 −2

of deep vein thrombosis High probability ≥3; moderate probability 1–2; low probability ≤0. In patients with symptoms in both legs, the more symptomatic leg is used. Based on data from Wells et al. [1].

144

Positive predictive values of the Wells test in patients with suspected DVT are shown in Table 29.2. Among patients with a low probability Wells test, DVT was shown in 3–13% [1–7] (Table 29.2). Among patients with a high probability assessment by the Wells test, DVT was present in 38–75%. Review of the Wells test in a meta-analysis of 22 studies showed that 11% of patients with a low probability Wells score had DVT [8], A low Wells score markedly reduced the probability of DVT (negative likelihood ratio 0.25 [8]. In this meta-analysis, 56% of patients with a high probability Wells score had DVT. A high Wells score markedly increased the probability of DVT (positive likelihood ratio 5.2) [8]. Risk stratification was more accurate for proximal DVT than for distal DVT [8]. Empirical assessment showed that in patients with suspected DVT who had a low probability clinical assessment, DVT was present in 1–13% [3, 5, 9] (Table 29.2, Figure 29.1). With a high probability clinical assessment, DVT was present in 63–100% [3, 5, 9]. In articles reviewed for a meta-analysis, 52% of patients with a high probability empirical assessment had DVT [8]. Meta-analysis showed that empirical assessment gave similar likelihood ratios as the Wells score, but there were a limited number of studies, and the confidence intervals were wide [8]. In patients with suspected DVT who had a low probability clinical assessment, DVT was present in 8% [8]. The Wells scoring system has been criticized because DVT is not entirely excluded in patients with a low score [6]. Resident physicians calculated a probability score for DVT that disagreed with senior staff in 30 of 165 patients (18%) [10]. Awareness of the subjective aspect of some parts of the Wells’ score was recommended before its implementation in clinical practice [10]. Putting this in perspective, the real value of the Wells score (and any clinical decision aid) is its ability to complement, rather than displace, physicians’


BLUK077-Stein

March 11, 2007

CHAPTER 29

7:5

145

Clinical model for assessment of DVT

Table 29.2 Positive predictive value of empirical assessment and clinical model for probability of deep venous thrombosis.

Method

First author [Ref]

Low probability

Intermediate probability

High probability

[DVT/ N (%)]

[DVT/ N (%)]

[DVT/ N (%)]

Empirical

Cornuz [5]

11/86 (13)

30/127 (24)

41/65 (63)

Empirical

Perrier [9]

3/29 (1)

56/291 (19)

52/54 (96)

Empirical

Miron [3]

1/78 (1)

30/166 (18)

26/26 (100)

15/193 (8)

116/584 (20)

119/145 (82)

Wells [2]

16/301 (5)

47/143 (33)

72/85 (85) 34/46 (74)

Average Wells expanded Wells

Miron [3]

4/126 (3)

19/98 (19)

Wells

Wells [4]

5/50 (10)

14/71 (20)

22/29 (76)

Wells

Cornuz [5]

14/121 (12)

36/121 (30)

32/48 (67)

Wells

Odega [6]

61/507 (12)

53/321 (17)

175/467 (38)

Wells

Kraaijn’n [7]

71/896 (8)

133/508 (26)

208/322 (65)

Wells

Wells [1]

10/39 (3)

32/193 (17)

53/71 (75)

165/1739 (9)

287/1312 (22)

524/983 (53)

Average

Figure 29.1 Pooled data showing prevalence of deep venous thrombosis (DVT) among patients with low, intermediate, and high probability (prob) assessments based upon empirical (empir) evaluation and upon the Wells scoring system. Data based on values shown in Table 29.2.

Deep vein thrombosis (%)

100

Empir 82%

75 Wells 53%

50

25

Empir Wells 8% 9%

Empir Wells 20% 22%

0 Low prob

empirical assessment [11]. The Wells score will classify the patient into a probability range that is probably correct, but it is up to the physician to apply clinical judgment to fine-tune this estimate of disease probability [11].

4

5

References 1 Wells PS, Anderson DR, Bormanis J et al. Value of assessment of pretest probability of deep-vein thrombosis in clinical management. Lancet 1997; 350: 1795–1798. 2 Wells PS, Hirsh J, Anderson DR et al. Accuracy of clinical assessment of deep-vein thrombosis. Lancet 1995; 345: 1326–1330. 3 Miron M-J, Perrier A, Bounameaux H. Clinical assessment of suspected deep vein thrombosis: comparison be-

6

7

Intermediate prob

High prob

tween a score and empirical assessment. J Intern Med 2000; 247: 249–254. Wells PS, Anderson DR, Bormanis J et al. Application of a diagnostic clinical model for the management of hospitalized patients with suspected deep-vein thrombosis. Thromb Haemost 1999; 81: 493–497. Cornuz J, Ghali WA, Hayoz D et al. Clinical prediction of deep venous thrombosis using two risk assessment methods in combination with rapid quantitative D-dimer testing. Am J Med 2002; 112: 198–203. Oudega R, Hoes AW, Moons KGM. The Wells rule does not adequately rule out deep venous thrombosis in primary care patients. Ann Intern Med 2005; 143: 100– 107. Kraaijenhagen RA, Piovella F, Bernardi E et al. Simplification of the diagnostic management of suspected deep vein thrombosis. Arch Intern Med 2002; 162: 907–911.


BLUK077-Stein

March 11, 2007

7:5

146

8 Goodacre SG, Sutton AJ, Sampson FC. Meta-analysis: the value of clinical assessment in the diagnosis of deep venous thrombosis. Ann Intern Med 2005; 143: 129–139. 9 Perrier A, Desmarais S, Miron M-J et al. Non-invasive diagnosis of venous thromboembolism in outpatients. Lancet 1999; 353: 190–195.

PART II

Diagnosis of DVT

10 Bigaroni A, Perrier A, Bounameaux H. Is clinical probability assessment of deep vein thrombosis by a score really standardized? Thromb Haemost 2000; 83: 788–789. 11 Douketis JD. Use of clinical prediction score in patients with suspected deep venous thrombosis: two steps forward, one step back? Ann Intern Med 2005; 143: 140–142.


BLUK077-Stein

March 11, 2007

30

7:6

CHAPTER 30

Clinical probability score plus single negative ultrasound for exclusion of deep venous thrombosis

Ultrasonography A clinical model that combines a pretest probability by a scoring system with compression ultrasonography is safe and feasible, and reduces the need for serial ultrasound testing [1]. The Wells point score system for assessment of probability of deep venous thrombosis (DVT) is shown in Chapter 29 [1]. A low probability score for DVT (score ≤0) in combination with a negative compression ultrasound of the lower extremities had a negative predictive value of 97.8–99.7% [1–4]

and an “unlikely” probability (score <2) had a negative predictive value of 98.3–98.9% [5, 6] (Table 30.1). A moderate clinical probability of DVT in combination with a negative ultrasound had a negative predictive value of 96.6–97.0% [1, 2] (Table 30.1). However, if the clinical probability by the point score system was high probability for DVT, a negative venous ultrasound could not be relied upon to exclude DVT. The negative predictive value of a high-probability clinical score and negative ultrasound, in small numbers of patients ranged from 71 to 73% [1, 2] (Table 30.1).

Table 30.1 Negative predictive value in patients with suspected deep vein thrombosis and negative ultrasound according to clinical assessment scores. First author [Ref]

Population

Negative predictive value (%)

Ultrasound negative and low probability or “unlikely” clinical score (Wells score)* Wells score ≤0 (low probability) Wells [2]

Hospitalized

45/46 (97.8)

Wells [1]

Outpatients

317/318 (99.7)

Kraaijenhagen [4]

Outpatients

821/834 (98.4)

Tick [3]

Outpatients

245/250 (98.0)

Wells score <2 (unlikely) Wells [5]*

Outpatients

264/267 (98.9)

Perrier [6]†

Outpatients

234/238 (98.3)

Wells [2]

Hospitalized

58/60 (96.7)

Wells [1]

Outpatients

161/166 (97.0)

Wells [2]

Hospitalized

5/7 (71.4)

Wells [1]

Outpatients

11/15 (73.3)

Wells score 1–2 (moderate probability)

Wells score ≥3 (high probability)

*Unlikely = Wells score <2. † Low or moderate probability. Low probability = Wells score ≤0; moderate probability = Wells score 1–2; high probability = Wells score ≥3.

147


BLUK077-Stein

March 11, 2007

7:6

148

PART II

Diagnosis of DVT

References 1 Wells PS, Anderson DR, Bormanis J et al. Value of assessment of pretest probability of deep-vein thrombosis in clinical management. Lancet 1997; 350: 1795–1798. 2 Wells P, Anderson D, Bormanis J et al. Application of a diagnostic clinical model for the management of hospitalized patients with suspected deep-vein thrombosis. Thromb Haemost 1999; 81: 493–497. 3 Tick LW, Ton E, van Voorthuizen T et al. Practical diagnostic management of patients with clinically suspected deep vein thrombosis by clinical probability test, compression

ultrasonography and D-dimer test. Am J Med 2002; 113: 630–635. 4 Kraaijenhagen RA, Piovella F, Bernardi E et al. Simplification of the diagnostic management of suspected deep vein thrombosis. Arch Intern Med 2002; 162: 907–911. 5 Wells PS, Anderson DR, Rodger M et al. Evaluation of Ddimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003; 349: 1227–1235. 6 Perrier A, Desmarais S, Miron MJ et al. Non-invasive diagnosis of venous thromboembolism in outpatients. Lancet 1999; 353: 190–195.


BLUK077-Stein

March 12, 2007

31

15:51

CHAPTER 31

D-dimer for the exclusion of acute deep venous thrombosis

The D-dimer assays in patients with suspected deep venous thrombosis (DVT) differ in sensitivity, specificity, and likelihood ratios from values in patients with suspected pulmonary embolism (PE) [1]. Values of sensitivity of the enzyme-linked immunosorbent assay (ELISA) and quantitative rapid ELISA assays are significantly superior to those for the quantitative latex, semiquantitative latex, and whole blood agglutination assays in patients with suspected DVT [1] (Figure 31.1). The quantitative rapid ELISA assay is more convenient than the conventional ELISA and it provides a high certainty for a negative diagnosis of DVT in patients with a low or intermediate clinical probability (but not high clinical probability) (Chapter 32).

Negative likelihood ratios of the quantitative rapid ELISA assay were 0.09, 0.10, and 0.08 in Tier 1, Tier 2, and Tier 3 studies (Tables 31.1 and 31.2). Tier 1 studies (20 investigations) compared an ELISA assay and at least one other D-dimer assay for exclusion of DVT [2–21] (Table 31.3). Tier 2 included the Tier 1 studies and 29 additional studies [22–50] that met all inclusion criteria (Table 31.4). Tier 3 combined 21 methodologically weaker studies [51–71] with the 49 Tier-2 studies. Negative likelihood ratios <0.1 generate large and often conclusive changes from pretest to posttest probability [72] providing high certainty for excluding DVT. As will be shown, combining a negative rapid ELISA with a low or moderate clinical probability for

Table 31.1 Deep vein thrombosis. Sensitivity

Specificity

Positive likelihood

Negative likelihood

(95% CI)

(95% CI)

ratio (95% CI)

ratio (95% CI)

ELISA

0.96 (0.91–1.00)

0.38 (0.28–0.48)

1.55 (1.32–1.81)

0.12 (0.04–0.33)

Quant rapid ELISA

0.96 (0.90–1.00)

0.44 (0.32–0.55)

1.70 (1.39–2.09)

0.09 (0.02–0.41)

Semiquant rapid ELISA

0.89 (0.81–0.98)

0.39 (0.28–0.50)

1.47 (1.21–1.78)

0.27 (0.12–0.60)

Qual rapid ELISA

0.93 (0.84–1.00)

0.47 (0.30–0.63)

1.75 (1.28–2.39)

0.15 (0.04–0.56)

Quant latex

0.85 (0.74–0.95)

0.66 (0.55–0.78)

2.49 (1.77–3.51)

0.24 (0.12–0.45)

Semiquant latex

0.78 (0.67–0.89)

0.66 (0.56–0.76)

2.30 (1.69–3.13)

0.33 (0.21–0.54)

Whole blood

0.87 (0.68–1.00)

0.83 (0.65–1.00)

4.97 (1.84–13.42)

0.16 (0.04–0.65)

ELISA

0.95 (0.91–0.99)

0.40 (0.32–0.49)

1.60 (1.39–1.83)

0.12 (0.05–0.29)

Quant rapid ELISA

0.96 (0.90–1.00)

0.44 (0.34–0.54)

1.71 (1.43–2.05)

0.10 (0.03–0.36)

Semiquant rapid ELISA

0.90 (0.83–0.98)

0.39 (0.29–0.50)

1.48 (1.24–1.78)

0.25 (0.12–0.55)

Qual rapid ELISA

0.93 (0.87–0.99)

0.46 (0.35–0.57)

1.73 (1.40–2.13)

0.15 (0.07–0.37)

Quant latex

0.86 (0.78–0.94)

0.61 (0.51–0.71)

2.20 (1.70–2.84)

0.23 (0.13–0.41)

Semiquant latex

0.79 (0.69–0.88)

0.66 (0.57–0.75)

2.33 (1.75–3.11)

0.32 (0.20–0.51)

Whole blood

0.86 (0.80–0.93)

0.67 (0.61–0.73)

2.62 (2.17–3.16)

0.20 (0.13–0.32)

Test Tier 1 analysis

Tier 2 analysis

Quant, quantitative; semiquant, semiquantitative; qual, qualitative. Modified and reprinted with permission from Stein et al. [1].

149


BLUK077-Stein

March 12, 2007

15:51

150

Diagnosis of DVT

Proportion

Proportion

PART II

Sensitivity

Specificity

Figure 31.1 Boxplots of sensitivity and specificity among the D-dimer assays for patients with suspected deep venous thrombosis (DVT). (Reprinted with permission from Stein et al. [1].)

DVT essentially rules out DVT (Chapter 32). However, a negative D-dimer does not reliably exclude DVT if the clinical probability is high (Chapter 32). Sensitivity and specificity for DVT according to various assays using a cutoff level of 500 ng/mL are shown

in Figure 31.1. In patients with DVT, the least variability for sensitivity was seen with the ELISA, qualitative rapid ELISA, and quantitative rapid ELISA assays. Limited data for sensitivity, specificity, and likelihood ratios were available for cutoff values of 250 ng/mL and


BLUK077-Stein

March 12, 2007

CHAPTER 31

15:51

151

D-dimer for exclusion of acute DVT

Table 31.2 Deep vein thrombosis.

Test

Sensitivity

Specificity

Positive likelihood

Negative likelihood

(95% CI)

(95% CI)

ratio (95% CI)

ratio (95% CI)

Cutoff 500 ng/mL – Tier 3 analysis (all data) ELISA

0.94 (0.89–0.98)

0.43 (0.36–0.50)

1.65 (1.46–1.87)

0.15 (0.07–0.30)

Quant rapid ELISA

0.97 (0.92–1.00)

0.42 (0.32–0.52)

1.67 (1.42–1.97)

0.08 (0.02–0.38)

Semiquant rapid ELISA

0.91 (0.85–0.98)

0.43 (0.34–0.52)

1.60 (1.37–1.88)

0.21 (0.10–0.42)

Qual rapid ELISA

0.93 (0.87–0.99)

0.53 (0.43–0.64)

1.99 (1.60–2.48)

0.13 (0.06–0.32)

Quant latex

0.88 (0.80–0.95)

0.59 (0.49–0.69)

2.14 (1.68–2.73)

0.21 (0.12–0.38)

Semiquant latex

0.78 (0.69–0.87)

0.70 (0.62–0.78)

2.60 (1.95–3.46)

0.31 (0.21–0.47)

Whole blood

0.82 (0.76–0.89)

0.70 (0.64–0.76)

2.77 (2.27–3.38)

0.25 (0.18–0.36)

Cutoff 250 ng/mL – All studies meeting inclusion criteria ELISA

NA

NA

NA

NA

Quant rapid ELISA

0.98

0.39

1.58

0.07

Semiquant rapid ELISA

0.92

0.56

2.10

0.14

Qual rapid ELISA

NA

NA

NA

NA

Quant Latex

0.91

0.53

1.96

0.16

Semiquant latex

0.91

0.47

1.73

0.19

Whole blood

0.88

0.66

2.57

0.18

Cutoff 1000 ng/ml – All studies meeting inclusion criteria ELISA

0.90

0.72

3.21

0.14

Quant rapid ELISA

0.93

0.58

2.23

0.13

Semiquant rapid ELISA

0.94

0.57

2.19

0.11

Qual rapid ELISA

NA

NA

NA

NA

Quant latex

0.81

0.70

2.73

0.27

Semiquant latex

0.73

0.78

3.35

0.34

Whole blood

0.88

0.66

2.57

0.18

Quant, quantitative; semiquant, semiquantitative; qual, qualitative. Modified and reprinted with permission from Stein et al. [1].

1000 ng/mL for some but not all of the D-dimer assays (Table 31.2). A negative D-dimer has highest negative predictive values in populations with a low prevalence of DVT. The negative predictive value of the quantitative rapid ELISA for exclusion of acute DVT was estimated based on a sensitivity of 96% and specificity of 44% (Tier 1 and Tier 2 studies), knowing the prevalence of DVT in the population tested. The prevalence of DVT in the included investigations of D-dimer ranged from 20 to 78% (average 36%) [1]. If the prevalence of DVT in the population tested were 36%, the negative predictive value of a negative quantitative rapid ELISA would be 95%. If the prevalence of DVT were 20%, the neg-

ative predictive value would be 98%. If the prevalence of DVT were 78%, the negative predictive value would be only 75% (Figure 31.2). A negative D-dimer may be useful for the exclusion of recurrent DVT as well as the exclusion of a suspected first episode of DVT [73]. The reliability of a negative D-dimer for the exclusion of DVT is highest in populations with a low prevalence of DVT and limited when the prevalence is high. Non-ELISA assays should not be used as stand-alone tests [74] given their inferior sensitivity and negative likelihood ratio values. However, in patients with a low clinical probability, non-ELISA assays provide a reasonable certainty for ruling out DVT [1].


BLUK077-Stein

March 12, 2007

15:51

Table 31.3 Patients with suspected DVT.

Study

Year

D-dimer type

Cutoff analyzed (ng/mL)

No. Pts enrolled

Patients with DVT (%)

1989

ELISA,

<500

56

37.5

<500

100

45.0

<500

97

40.2

<500

116

29.3

<500

32

78.1

Quant rapid ELISA,

<500

100

40.0

semiquant rapid ELISA,

<250 <500

114

42.1

ELISA,

<250

96

37.5

semiquant latex

<500

ELISA,

<500

171

43.9

Quant rapid ELISA,

<500

81

51.9

semiquant rapid ELISA,

<250

Tier 1 analysis Bounameaux [3]

semiquant latex Elias [6]

1990

ELISA, semiquant latex

Speiser [18]

1990

Quant rapid ELISA, qual latex

Boneu [2]

1991

ELISA, semiquant latex

Chang-Liem [4]

1991

Quant rapid ELISA, quant latex

Dale [5]

1994

quant latex Hansson [10]

1994

ELISA, quant latex, semiquant latex

Tengborn [19] Elias [7]

1994 1996

Quant rapid ELISA, semiquant rapid ELISA, qual rapid ELISA, semiquant latex Legnani [12]

1997

qual rapid ELISA, ELISA, quant latex, semiquant latex Mayer [15]

1997

Semiquant rapid ELISA, whole blood

<500

108

30.6

Escoffre-Barbe [8]

1998

ELISA,

<500

464

59.5

<250

180

33.3

99

39.4

97

39.2

177

20.3

99

50.5

quant latex Lindahl [14]

1998

Semiquant rapid ELISA, semiquant latex

Legnani [13]

˚ Wahlander [21]

1999

1999

Quant rapid ELISA,

<700

ELISA,

<155

quant latex

<70

Quant rapid ELISA,

≤ 500 ≤ 1000

semiquant rapid ELISA,

<500

Qual rapid ELISA, ELISA, semiquant latex Sadouk [16] van der Graaf [20]

2000 2000

Quant rapid ELISA,

<500

quant latex

<250

Quant rapid ELISA,

<500

semiquant rapid ELISA,

<1000

qual rapid ELISA,

<250

ELISA, quant latex, semiquant latex, whole blood (Continued )

152


BLUK077-Stein

March 12, 2007

15:51

Table 31.3 Patients with suspected DVT (Continued ).

Study

Year

D-dimer type

Cutoff analyzed (ng/mL)

No. Pts enrolled

Patients with DVT (%)

Funfsinn [9]

2001

Quant rapid ELISA,

<500

106

44.3

108

25.0, 20.4, 26.9

113

43.4

ELISA, quant latex Shitrit [17]

Larsen [11]

2001

2002

Quant rapid ELISA,

<1000

quant latex,

<500

semiquant latex

<250

Quant rapid ELISA,

<500

semiquant rapid ELISA, quant latex

DVT, deep venous thrombosis; semiquant, semiquantitative; quant, quantitative; qual, qualitative. Modified and reprinted with permission from Stein et al. [1]. Table 31.4 Tier 2 analysis (Tier 2 includes Tier 1 and the following investigations).

Study

Year

D-dimer type

Cutoff analyzed (ng/mL)

No. Pts enrolled

Patients with DVT (%)

Van Bergen [47]

1989

ELISA

<250

239

24.7

de Boer [28]

1991

Qual latex

<200

33

63.6

Kroneman [39]

1991

Quant rapid ELISA

<268

239

24.7

Ibrahim [34]

1992

Semiquant latex

<250

85

45.9

Jossang [37]

1992

Semiquant latex

<500

69

53.6

Heijboer [33]

1992

ELISA

<300

474

12.2

Pini [45]

1993

Semiquant latex

<200

425

45.9

Wells [49]

1995

Whole blood

214

24.8

D’Angelo [27]

1996

Quant rapid ELISA

<500

103

21.4

<1000 Gavaud [30]

1996

ELISA

<370

80

32.5

Borg [23]

1997

ELISA

<500

76

42.1

Gauzzaloca [31]

1997

Qual rapid ELISA

<500

68

52.9

Jacq [35]

1997

Whole blood

50

48.0

Knecht [38]

1997

Quant latex

<500

154

38.3

Leroyer [42]

1997

Qual rapid ELISA,

<500

448

59.2

ELISA Khaira [37]

1998

Semiquant rapid ELISA

<500

79

36.7

Wijns [50]

1998

Quant rapid ELISA,

<1000

74

43.2

qual rapid ELISA,

<500

ELISA Aschwanden [22]

1999

Whole blood

343

24.2

Caliezi [26]

1999

ELISA

<500

106

44.3

Legnani [40]

1999

Quant latex

<230

92

38.0

Lennox [41]

1999

Whole blood

200

23.0

Lindahl [43]

1999

Quant latex

<700

236

43.2

Wells [48]

1999

Whole blood

150

26.0

Bradley [24]

2000

Semiquant rapid ELISA

<300

138

31.9

Farrell [29]

2000

Whole blood

198

8.1

Permpikul [44]

2000

Whole blood

65

66.2

Bucek [25]

2001

Quant latex,

<500

100

35.0

Harper [32]

2001

<250

235

21.7, 21.3

1739

24.4

whole blood Quant latex, whole blood ten Wolde [42]

2002

Whole blood

DVT, deep venous thrombosis; qual, qualitative; quant, quantitative; semiquant, semiquantitative. Modified and reprinted with permission from Stein et al. [1].

153


BLUK077-Stein

March 12, 2007

15:51

154

PART II

Diagnosis of DVT

Negative predictive value (%)

100 98

95

80 75

60 40 20 0 20

36 Prevalence of DVT (%)

References 1 Stein PD, Hull RD, Patel KC et al. D-dimer for the exclusion of acute venous thrombosis and pulmonary embolism: a systematic review. Ann Intern Med 2004; 140: 589–602. 2 Boneu B, Bes G, Pelzer H, Si´e P, Boccalon H. D-dimers, thrombin antithrombin III complexes and prothrombin fragments 1 + 2: diagnostic value in clinically suspected deep vein thrombosis. Thromb Haemost 1991; 65: 28–31. 3 Bounameaux H, Schneider PA, Reber G, de Moerloose P, Krahenbuhl B. Measurement of plasma D-dimer for diagnosis of deep venous thrombosis. Am J Clin Pathol 1989; 91: 82–85. 4 Chang-Liem GS, Lustermans FA, van Wersch JW. Comparison of the appropriateness of the latex and Elisa plasma D-dimer determination for the diagnosis of deep venous thrombosis. Haemostasis 1991; 21: 106–110. 5 Dale S, Gogstad GO, Brosstad F et al. Comparison of three D-dimer assays for the diagnosis of DVT: ELISA, latex and immunofiltration assay (NycoCard D-dimer). Thromb Haemost 1994; 71: 270–274. 6 Elias A, Aillaud F, Roul C et al. Assessment of D-dimer measurement by ELISA or latex methods in deep vein thrombosis diagnosed by ultrasonic duplex scanning. Fibrinolysis 1990; 4: 237–240. 7 Elias A, Aptel I, Huc B et al. D-dimer test and diagnosis of deep vein thrombosis: a comparative study of 7 assays. Thromb Haemost 1996; 76: 518–522. 8 Escoffre-Barbe M, Oger E, Leroyer C et al. Evaluation of a new rapid D-dimer assay for clinically suspected deep venous thrombosis (Liatest D-dimer). Am J Clin Pathol 1998; 109: 748–753. 9 Funfsinn N, Caliezi C, Biasiutti FD et al. Rapid D-dimer testing and pre-test clinical probability in the exclusion of deep venous thrombosis in symptomatic outpatients. Blood Coagul Fibrinolysis 2001; 12: 165–170.

78

Figure 31.2 Negative predictive value for deep venous thrombosis (DVT) of quantitative rapid ELISA shown in relation to prevalence of DVT in population studied. (Data from Stein et al. [1].)

10 Hansson PO, Eriksson H, Eriksson E, Jagenburg R, Lukes P, Risberg B. Can laboratory testing improve screening strategies for deep vein thrombosis at an emergency unit? J Intern Med 1994; 235: 143–151. 11 Larsen TB, Stoffersen E, Christensen CS, Laursen B. Validity of D-dimer tests in the diagnosis of deep vein thrombosis: a prospective comparative study of three quantitative assays. J Intern Med 2002; 252: 36–40. 12 Legnani C, Pancani C, Palareti G et al. Comparison of new rapid methods for D-dimer measurement to exclude deep vein thrombosis in symptomatic outpatients. Blood Coagul Fibrinolysis 1997; 8: 296–302. 13 Legnani C, Pancani C, Palareti G, Guazzaloca G, Coccheri S. Contribution of a new, rapid, quantitative and automated method for D-dimer measurement to exclude deep vein thrombosis in symptomatic outpatients. Blood Coagul Fibrinolysis 1999; 10: 69–74. 14 Lindahl TL, Lundahl TH, Ranby M, Fransson SG. Clinical evaluation of a diagnostic strategy for deep venous thrombosis with exclusion by low plasma levels of fibrin degradation product D-dimer. Scand J Clin Lab Invest 1998; 58: 307–316. 15 Mayer W, Hirschwehr R, Hippmann G, Odpadlik H, Bayer P, Partsch H. Whole-blood immunoassay (SimpliRED) versus plasma immunoassay (Nyco-Card) for the diagnosis of clinically suspected deep vein thrombosis. VASA 1997; 26: 97–101. 16 Sadouk M, Desmarais S, Patenaude JV, Lepage R. Comparison of diagnostic performance of three new fast Ddimer assays in the exclusion of deep vein thrombosis. Clin Chem 2000; 46: 286–287. 17 Shitrit D, Heyd J, Raveh D, Rudensky B. Diagnostic value of the D-dimer test in deep vein thrombosis: improved results by a new assay method and by using discriminate levels. Thromb Res 2001; 102: 125–131. 18 Speiser W, Mallek R, Koppensteiner R et al. D-dimer and TAT measurement in patients with deep venous thrombosis: utility in diagnosis and judgement of anticoagulant


BLUK077-Stein

March 12, 2007

CHAPTER 31

19

20

21

22

23

24

25

26

27

28

29

30

15:51

155

D-dimer for exclusion of acute DVT

treatment effectiveness. Thromb Haemost 1990; 64: 196– 201. Tengborn L, Palmblad S, Wojciechowski J, Peterson LE, Stigendal L. D-dimer and thrombin/antithrombin III complex—diagnostic tools in deep venous thrombosis? Haemostasis 1994; 24: 344–350. van der Graaf F, van den Borne H, van der Kolk M, de Wild PJ, Janssen GW, van Uum SH. Exclusion of deep venous thrombosis with D-dimer testing—comparison of 13 D-dimer methods in 99 outpatients suspected of deep venous thrombosis using venography as reference standard. Thromb Haemost 2000; 83: 191–198. W˚ahlander K, Tengborn L, Hellstr¨om M et al. Comparison of various D-dimer tests for the diagnosis of deep venous thrombosis. Blood Coagul Fibrinolysis 1999; 10: 121–126. Aschwanden M, Labs KH, Jeanneret C, Gehrig A, Jaeger KA. The value of rapid D-dimer testing combined with structured clinical evaluation for the diagnosis of deep vein thrombosis. J Vasc Surg 1999; 30: 929–935. Borg JY, Levesque H, Cailleux N, Franc C, Hellot MF, Courtois H. Rapid quantitative D-dimer assay and clinical evaluation for the diagnosis of clinically suspected deep vein thrombosis [Letter]. Thromb Haemost 1997; 77: 602– 603. Bradley M, Bladon J, Barker H. D-dimer assay for deep vein thrombosis: its role with colour Doppler sonography. Clin Radiol 2000; 55: 525–527. Bucek RA, Quehenberger P, Feliks I, Handler S, Reiter M, Minar E. Results of a new rapid D-dimer assay (cardiac D-dimer) in the diagnosis of deep vein thrombosis. Thromb Res 2001; 103: 17–23. Caliezi C, Funfsinn N, Mauron T et al. Performance of a new fibrin monomer assay to exclude deep vein thrombosis in symptomatic outpatients. Thromb Haemost 1999; 81: 50–53. D’Angelo A, D’Alessandro G, Tomassini L, Pittet JL, Dupuy G, Crippa L. Evaluation of a new rapid quantitative D-dimer assay in patients with clinically suspected deep vein thrombosis. Thromb Haemost 1996; 75: 412– 416. de Boer WA, de Haan MA, Huisman JW, Klaassen CH. Ddimer latex assay as screening method in suspected deep venous thrombosis of the leg. A clinical study and review of the literature. Neth J Med 1991; 38: 65–69. Farrell S, Hayes T, Shaw M. A negative SimpliRED Ddimer assay result does not exclude the diagnosis of deep vein thrombosis or pulmonary embolus in emergency department patients. Ann Emerg Med 2000; 35: 121– 125. Gavaud C, Ninet J, Ville D et al. [Diagnosis of venous thrombosis and/or pulmonary embolism by determination of D-dimers using ELISA. Review based on a study

31

32

33

34

35

36

37

38

39

40

41

42

of 80 consecutive patients hospitalized in an emergency unit]. J Mal Vasc 1996; 21: 22–30. Guazzaloca G, Palareti G, Legnani C et al. [Deep venous thrombosis. Evaluation of a non-invasive diagnostic procedure based on compression ultrasonography and measurement of D-dimer plasma levels]. Minerva Cardioangiol 1997; 45: 259–266. Harper P, Marson C, Grimmer A, Monahan K, Humm G, Baker B. The rapid whole blood agglutination D-dimer assay has poor sensitivity for use as an exclusion test in suspected deep vein thrombosis. N Z Med J 2001; 114: 61–64. Heijboer H, Ginsberg JS, Buller HR, Lensing AW, Colly LP, Wouter ten Cate J. The use of the D-dimer test in combination with non-invasive testing versus serial non-invasive testing alone for the diagnosis of deep-vein thrombosis. Thromb Haemost 1992; 67: 510–513. Ibrahim KM, O’Neill AI, Parkin DJ. D-dimer (latex) assay in the diagnosis of deep venous thrombosis [Letter]. Med J Aust 1992; 157: 574–575. Jacq F, Heron E, Rance A, Cesarini ML, Emmerich J, Fiessinger JN. [Evaluation of a test for rapid detection of D-dimers for the exclusion of the diagnosis of venous thrombosis]. Presse Med 1997; 26: 1132–1134. Jossang B, Runde I. [Diagnostic value of C-reactive protein and D-dimer in deep venous thrombosis]. Tidsskr Nor Laegeforen 1992; 112: 1153–1155. Khaira HS, Mann J. Plasma D-dimer measurement in patients with suspected DVT—a means of avoiding unnecessary venography. Eur J Vasc Endovasc Surg 1998; 15: 235–238. Knecht MF, Heinrich F. Clinical evaluation of an immunoturbidimetric D-dimer assay in the diagnostic procedure of deep vein thrombosis and pulmonary embolism. Thromb Res 1997; 88: 413–417. Kroneman H, Van Bergen PF, Knot EA, Jonker JJ, de Maat MP. Diagnostic value of D-dimer for deep venous thrombosis in outpatients. Haemostasis 1991; 21: 286–292. Legnani C, Pancani C, Palareti G, Guazzaloca G, Coccheri S. Performance of a new, fast D-dimer test (IL Test Ddimer) for the management of outpatients with suspected deep vein thrombosis in emergency situations. Fibrinolysis Proteolysis 1999; 13: 139–141. Lennox AF, Delis KT, Serunkuma S, Zarka ZA, Daskalopoulou SE, Nicolaides AN. Combination of a clinical risk assessment score and rapid whole blood D-dimer testing in the diagnosis of deep vein thrombosis in symptomatic patients. J Vasc Surg 1999; 30: 794–803. Leroyer C, Escoffre M, Le Moigne E et al. Diagnostic value of a new sensitive membrane based technique for instantaneous D-dimer evaluation in patients with clinically suspected deep venous thrombosis. Thromb Haemost 1997; 77: 637–640.


BLUK077-Stein

March 12, 2007

15:51

156

43 Lindahl TL, Lundahl TH, Fransson SG. Evaluation of an automated microlatex D-dimer assay (Tina-quant on Hitachi 911 analyser) in symptomatic outpatients with suspected DVT [Letter]. Thromb Haemost 1999; 82: 1772– 1773. 44 Permpikul C, Chantavatcharakorn K, Bouranasompop C, Chaiyasoot W. Whole blood agglutination D-dimer test for the diagnosis of deep vein thrombosis. J Med Assoc Thai 2000; 83: 732–736. 45 Pini M, Quintavalla R, Pattacini C et al. Combined use of strain-gauge plethysmography and latex D-dimer test in clinically suspected deep venous thrombosis. Fibrinolysis 1993; 7: 391–396. 46 ten Wolde M, Kraaijenhagen RA, Prins MH, Buller HR. The clinical usefulness of D-dimer testing in cancer patients with suspected deep venous thrombosis. Arch Intern Med 2002; 162: 1880–1884. 47 van Bergen PF, Knot EA, Jonker JJ, de Boer AC, de Maat MP. Is quantitative determination of fibrin(ogen) degradation products and thrombin–antithrombin III complexes useful to diagnose deep venous thrombosis in outpatients? Thromb Haemost 1989; 62: 1043–1045. 48 Wells PS, Anderson DR, Bormanis J et al. Application of a diagnostic clinical model for the management of hospitalized patients with suspected deep-vein thrombosis. Thromb Haemost 1999; 81: 493–497. 49 Wells PS, Brill-Edwards P, Stevens P et al. A novel and rapid whole-blood assay for D-dimer in patients with clinically suspected deep vein thrombosis. Circulation 1995; 91: 2184–2187. 50 Wijns W, Daoud N, Droeshout I et al. Evaluation of two D-dimer assays in the diagnosis of venous thromboembolism. Acta Clin Belg 1998; 53: 270–274. 51 Bouman CS, Ypma ST, Sybesma JP. Comparison of the efficacy of Ddimer, fibrin degradation products and prothrombin fragment 1–2 in clinically suspected deep venous thrombosis. Thromb Res 1995; 77: 225–234. 52 Brenner B, Pery M, Lanir N et al. Application of a bedside whole blood D-dimer assay in the diagnosis of deep vein thrombosis. Blood Coagul Fibrinolysis 1995; 6: 219–222. 53 Carter CJ, Doyle DL, Dawson N, Fowler S, Devine DV. Investigations into the clinical utility of latex D-dimer in the diagnosis of deep venous thrombosis. Thromb Haemost 1993; 69: 8–11, 24. 54 Carter CJ, Serrano K, Breen DJ, Buckley AR, Devine DV. Rapid fibrin D-dimer tests for deep venous thrombosis: factors affecting diagnostic utility. J Emerg Med 1999; 17: 605–610. 55 Chapman CS, Akhtar N, Campbell S, Miles K, O’Connor J, Mitchell VE. The use of D-dimer assay by enzyme immunoassay and latex agglutination techniques in the diagnosis of deep vein thrombosis. Clin Lab Haematol 1990; 12: 37–42.

PART II

Diagnosis of DVT

56 Heaton DC, Billings JD, Hickton CM. Assessment of Ddimer assays for the diagnosis of deep vein thrombosis. J Lab Clin Med 1987; 110: 588–591. 57 Janssen MC, Verbruggen BW, ter Hark PJ, Novakova IR. Factor VIIA determination compared to D-dimer in diagnosis of deep venous thrombosis. Thromb Res 1997; 86: 423–426. 58 Janssen MC, Heebels AE, de Metz M et al. Reliability of five rapid D-dimer assays compared to ELISA in the exclusion of deep venous thrombosis. Thromb Haemost 1997; 77: 262–266. 59 Killick SB, Pentek PG, Mercieca JE, Clarke MF, Bevan DH. Comparison of immunofiltration assay of plasma Ddimer with diagnostic imaging in deep vein thrombosis. Br J Haematol 1997; 96: 846–849. 60 Kozman H, Flemmer MC, Rahnama M. Deep venous thrombosis: prediction by D-dimer? South Med J 1997; 90: 907–910. 61 LaCapra S, Arkel YS, Ku DH, Gibson D, Lake C, Lam X. The use of thrombus precursor protein, D-dimer, prothrombin fragment 1.2, and thrombin antithrombin in the exclusion of proximal deep vein thrombosis and pulmonary embolism. Blood Coagul Fibrinolysis 2000; 11: 371–377. 62 Mauron T, Baumgartner I, Z’Brun A et al. SimpliRED D-dimer assay: comparability of capillary and citrated venous whole blood, between-assay variability, and performance of the test for exclusion of deep vein thrombosis in symptomatic outpatients [Letter]. Thromb Haemost 1998; 79: 1217–1219. 63 Ott P, Astrup L, Jensen RH, Nyeland B, Pedersen B. Assessment of D-dimer in plasma: diagnostic value in suspected deep venous thrombosis of the leg. Acta Med Scand 1988; 224: 263–267. 64 Pannocchia A, Chiappino I, Valpreda S et al. DVT exclusion in symptomatic outpatients: use of a rapid D-dimer immunofiltration assay and comparison with three enzyme immuno assays. Fibrinolysis 1995; 10: 117–119. 65 Rowbotham BJ, Carroll P, Whitaker AN et al. Measurement of crosslinked fibrin derivatives—use in the diagnosis of venous thrombosis. Thromb Haemost 1987; 57: 59–61. 66 Sassa H, Sone T, Tsuboi H, Kondo J, Yabashi T. Diagnostic significance of thrombin-antithrombin III complex (TAT) and D-dimer in patients with deep venous thrombosis. Jpn Circ J 1996; 60: 201–206. 67 Scarano L, Bernardi E, Prandoni P et al. Accuracy of two newly described D-dimer tests in patients with suspected deep venous thrombosis. Thromb Res 1997; 86: 93–99. 68 Siragusa S, Terulla V, Pirrelli S et al. A rapid D-dimer assay in patients presenting at the emergency room with suspected acute venous thrombosis: accuracy and relation to clinical variables. Haematologica 2001; 86: 856–861.


BLUK077-Stein

March 12, 2007

CHAPTER 31

15:51

D-dimer for exclusion of acute DVT

69 Trujillo-Santos AJ, Garcia de Lucas MD, Rios-Tamayo R, Jimenez-Puete A, Garcia-Sanchez JE. [Clinical and analytic diagnostic evaluation of deep venous thrombosis of the lower limbs.] Med Clin (Barc) 2000; 114: 46–49. 70 Wildberger JE, Vorwerk D, Kilbinger M, Lentner A, Wienert V, G¨unther RW. [The diagnosis of deep venous thromboses of the leg using a new rapid test (SimpliRED)]. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 1997; 167: 79–82. 71 Wildberger JE, Vorwerk D, Kilbinger M et al. Bedside testing (SimpliRED) in the diagnosis of deep vein thrombosis. Evaluation of 250 patients. Invest Radiol. 1998; 33: 232–235.

157

72 Jaeschke R, Guyatt GH, Sackett DL, for The EvidenceBased Medicine Working Group. Users’ guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? JAMA 1994; 271: 703–707. 73 Rathbun SW, Whitsett TL, Raskob GE. Negative Ddimer result to exclude recurrent deep venous thrombosis: a management trial. Ann Intern Med 2004; 141: 839– 845. 74 Lee AY, Julian JA, Levine MN et al. Clinical utility of a rapid whole-blood D-dimer assay in patients with cancer who present with suspected acute deep venous thrombosis. Ann Intern Med 1999; 131: 417–423.


BLUK077-Stein

March 11, 2007

32

7:13

CHAPTER 32

D-dimer combined with clinical probability assessment for exclusion of acute deep venous thrombosis

Introduction If a D-dimer assay is negative, deep venous thrombosis (DVT) will be ruled out in a high proportion of patients, providing clinical probability assessment also indicates that the likelihood of DVT is low [1, 2]. How-

ever, calculations based on negative likelihood ratios, according to Bayes theorem, show that DVT would often be present in patients with a discordantly high clinical assessment in combination with a negative Ddimer [2]. This holds true for all D-dimer assays [1] (Figure 32.1a). A high clinical probability, therefore,

DVT 100 90 80 70 60 50 40 30 20 10 0

Whole blood

Quant rapid ELISA

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 10

D-Dimer posttest probability (%)

(a)

Pretest probability (%)

DVT

12 10 8

Whole blood 6 4 Quant rapid ELISA

2

Pretest probability (%)

158

40

35

30

25

20

15

10

0 5

D-Dimer posttest probability (%)

(b)

Figure 32.1 (a) Relation of posttest probability of deep venous thrombosis (DVT) to pretest probability of DVT in patients with a negative D-dimer assay. (Data from Stein et al. [1].) (b) Enlargement of Figure 32.1a.


BLUK077-Stein

March 11, 2007

CHAPTER 32

7:13

D-dimer and clinical probability to exclude DVT

Table 32.1 Negative D-dimer and clinical assessment for DVT by Wells score. Proportion with First author [Ref]

D-dimer

DVT [n/N (%)]

Clinical assessment low probability (score ≤0) Likhanipour [6]

Rapid ELISA

0/60 (0)

Bates [7]

Quantitative latex

0/193 (0)

Kraaijenhagen [3]

Whole blood

10/561 (1.8)

Kearon [4]

Whole blood

1/177 (0.6)

Janes [5]

Whole blood

1/98 (1.0)

Clinical assessment unlikely (score <2) Wells [8]

Whole blood

Rapid ELISA

blood agglutination [3–5] and in none tested by rapid ELISA or quantitative latex [6, 7] (Table 32.1). With an unlikely clinical probability and negative whole blood agglutination D-dimer, DVT was present in 0.9% [8] (Table 32.1). With a moderate probability Wells score, DVT was found in 1.1–3.0% with a negative rapid ELISA or quantitative latex assay [6, 7] and DVT occurred in 5.8% with a negative whole blood agglutination assay [4] (Table 32.1). Results were highly variable with a high probability score and negative D-dimer assay [4, 7] (Table 32.1), but sparse data showed that DVT was not reliably excluded [4] (Table 32.1).

2/218 (0.9)

Clinical assessment moderate probability (score 1–2) Likhanipour [6]

159

2/67 (3.0)

Bates [7]

Quantitative latex

1/90 (1.1)

Kearon [4]

Whole blood

7/120 (5.8)

Clinical assessment high probability (score ≥3) Bates [7]

Quantitative latex

0/20 (0)

Kearon [4]

Whole blood

33/41 (80.5)

Likhanipour [6]

Rapid ELISA

0/19 (0)

DVT, deep venous thrombosis.

indicates a need for further testing even though the Ddimer assay is negative. Differences according to assay are apparent when clinical assessment indicates approximately a 20–30% likelihood of DVT, as may occur with an intermediate probability Wells score. This is illustrated in Figure 32.1b, which is an enlargement of Figure 32.1a. For example, if the pretest probability assessment is 25% and a rapid ELISA is negative, approximately 3% of patients would have DVT, whereas with a negative whole blood assay, approximately 5% would have DVT [1, 2] (Figure 32.1b). Among patients with a negative D-dimer assay and a low-probability clinical assessment, DVT on average was shown in 1.4% of patients tested by whole

References 1 Stein PD, Hull RD, Patel KC et al. D-dimer for the exclusion of acute venous thrombosis and pulmonary embolism: a systematic review. Ann Intern Med 2004; 140: 589–602. 2 Sox HC. Commentary. Ann Intern Med 2004; 140: 602. 3 Kraaijenhagen RA, Piovella F, Bernardi E et al. Simplification of the diagnostic management of suspected deep vein thrombosis. Arch Intern Med 2002; 162: 907–911. 4 Kearon C, Ginsberg JS, Douketis J et al. Management of suspected deep venous thrombosis in outpatients by using clinical assessment and D-dimer testing. Ann Intern Med 2001; 135: 108–111. 5 Janes S, Ashford N. Use of a simplified clinical scoring system and D-dimer testing can reduce the requirement for radiology in the exclusion of deep vein thrombosis by over 20%. Br J Haematol 2001; 112: 1079–1082. 6 Likhanipour K, Wolfson AB, Walker H et al. Combining clinical risk with D-dimer testing to rule out deep vein thrombosis. J Emerg Med 2004; 27: 233–239. 7 Bates SM, Kearon C, Crowther M et al. A diagnostic strategy involving a quantitative latex D-dimer assay reliably excludes deep venous thrombosis. Ann Intern Med 2003; 138: 787–794. 8 Wells PS, Anderson DR, Rodger M et al. Evaluation of Ddimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003; 349: 1227–1235.


BLUK077-Stein

March 11, 2007

33

7:14

CHAPTER 33

D-dimer and single negative compression ultrasound for exclusion of deep venous thrombosis

D-dimer alone (Chapter 31), combined with clinical assessment (Chapter 32), or combined with a single compression ultrasound examination of the lower extremities (sometimes with clinical assessment) have been used in various management algorithms. Bernardi and associates, among all patients with suspected deep venous thrombosis (DVT), showed DVT in 0.2% on 3-month follow-up after a normal rapid ELISA D-dimer and single negative venous ultrasound [1] (Table 33.1). Tick and associates, in patients with a high or moderate clinical assessment, negative whole blood agglutination test, and negative ultrasound showed no DVT on follow-up [2]. Oudega and associates, among patients with a lowprobability Wells clinical assessment for DVT, negative rapid ELISA, and single negative compression ultrasound showed DVT in 2.3% on 3-month follow-up [3]. The average rate of DVT on 3-month follow-up after a single negative compression ultrasound and negative

D-dimer was 6 of 1268 (0.5%) [1–3]. Clearly, however, the studies were heterogeneous with differences in the pretest probability of DVT and differences in the D-dimer assay used.

References 1 Bernardi E, Prandoni P, Lensing A et al. D-dimer testing as an adjunct to ultrasonography in patients with clinically suspected deep vein thrombosis: prospective cohort study. BMJ 1998; 17: 1037–1040. 2 Tick LW, Ton E, van Voorthuizen T et al. Practical diagnostic management of patients with clinically suspected deep vein thrombosis by clinical probability test, compression ultrasonography and D-dimer test. Am J Med 2002; 113: 630–635. 3 Oudega R, Hoes AW, Moons KGM. The Wells rule does not adequately rule out deep venous thrombosis in primary care patients. Ann Intern Med 2005; 143: 100–107.

Table 33.1 Negative D-dimer, single compression ultrasound, and clinical evaluation for DVT. Proportion with DVT: First author [Ref]

Clinical

D-dimer

3-month outcome (%)

Bernardi [1]

All suspected DVT

Rapid ELISA

1/598 (0.2)

Tick [2]

Moderate/high (≥1) Wells

Whole blood agglutination

0/148 (0)

Oudega [3]

Low (≤0) Wells

Rapid ELISA

5/222 (2.3)

DVT, deep venous thrombosis.

160


BLUK077-Stein

April 6, 2007

34

15:24

CHAPTER 34

Contrast venography

Introduction Contrast venography continues to be the diagnostic reference standard for deep venous thrombosis (DVT) [1], but by 1999 it was rarely performed [2] and by 2003 fewer than 5000 hospitalized patients in the United States had a contrast venogram (Stein unpublished) (Figure 34.1). For the evaluation of patients with suspected DVT, contrast venography has been replaced by compression ultrasound [2] and perhaps venous phase computed tomography (CT) when used in combination with contrast enhanced CT pulmonary angiography [3].

Method of contrast venography

125 100

US

75 50 25

Venograms

0 2003

2001

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Number of patients (×103)

The procedure for contrast venography recommended in PIOPED II [4] was modified from Rabinov and Paulen [5]. The patient is placed in a semi-upright position (30–45◦ ). The patient stands on a box or rotating

Year Figure 34.1 Number of hospitalized patients in United States from 1979 to 2003 who had venous ultrasonography or contrast venography. By 2003, fewer than 5000 hospitalized patients in the United States had contrast venography. (Modified and reproduced from Stein et al. [2], with permission from American Medical Association. All rights reserved.)

platform with the other foot so that the injected leg is non-weight-bearing and completely relaxed. A 19–21gauge scalp vein needle or a plastic cannula is placed in a superficial vein of the foot, preferably in the superficial dorsal metatarsal vein of the great toe. A tourniquet is placed at the ankle. Low- or iso-osmolar nonionic contrast material is recommended at a concentration of 40–50% (200–250 mg/mL). The injection is done by hand using a 60 cc syringe. Power injectors are not recommended. The injection is performed under fluoroscopic monitoring at a rate of 0.5–1.0 cc per second. The amount of contrast material necessary for complete opacification of the veins is determined by direct fluoroscopic observation. Usually, 125–150 mL is required for each leg. The injection site should be observed fluoroscopically and/or visually in order that any extravasation can be recognized at once. Each segment of the venous system from the foot to the groin is recorded on spot films or digitally and/or on 14 × 17 films with an overhead tube at 40 inches target-film distance and Bucky film tray in various projections (at least two orthogonal views) for the calf veins and popliteal veins, and a single AP film for the thigh veins, pelvic veins, and IVC, so that optimal display of the venous pathways is accomplished without overlapping. Gentle palpation can be helpful to achieve complete opacification in local areas of imperfect filling. For the best radiographic detail, a small focal-spot tube (0.6 mm) and an automatic phototimer are recommended. After an initial bolus of approximately 50 cc of contrast material, 15–25 mL boluses of contrast material are injected prior to each overhead film. The examiner manually compresses the ipsilateral common femoral vein and the fluoroscopic table is turned into the horizontal or slight Trendelenburg position. The imaging field is positioned to cover the ipsilateral femoral vein, iliac vein, and the lower inferior vena. The ipsilateral leg is elevated as much as possible by the examiner and manual pressure on the femoral vein is released. A single AP image is obtained.

161


BLUK077-Stein

April 6, 2007

15:24

162

Diagnostic criteria Criteria for diagnosis of acute DVT by contrast venography that were agreed upon by PIOPED II investigators [4] were (1) a complete filling defect, i.e., failure to opacify the entire lumen due to a central filling defect (the vessel may enlarge compared to the opposite vein) and (2) a partial filling defect surrounded by contrast material. Criteria for chronic DVT are large collateral veins, webs, and synechiae (signs of recanalization).

Side-effects with low-osmolar, nonionic contrast material Severe discomfort among 282 patients who received high concentration standard contrast material, sodium and meglumine diatrizoates (Renografin 60) occurred in 9% and moderate discomfort occurred in 14% [6]. Postvenography thrombosis occurred in 9% [6]. New subclinical DVT was shown in 20 of 61 (33%) initially normal patients following venograms, based on substantially increased uptake of I-125 labeled fibrinogen after the procedure [7]. In 4 of 61 patients (7%), there was evidence of PE or DVT by ventilation–perfusion lung scan or repeat venography [7]. With Renografin 45, among 212 patients, severe pain was reduced to 2%, moderate pain reduced to 12%, and postvenography thrombosis reduced to 4% [6]. With nonionic lowosmolar contrast material (iopamidol), among 200 patients, severe and moderate pain was eliminated and postvenography thrombosis decreased to 2% or less [6]. All patients had mild pain with venography [6]. In another investigation with nonionic low-osmolar contrast material (iohexol), among 463 patients, local pain occurred on 12% and persisted for 1 week in 4% of patients [8]. Postvenography superficial phlebitis occurred in 0.6% and serious bronchospasm occurred in 0.4%. Altogether, minor side-effects including local pain, nausea, dizziness, skin reactions, vomiting, delayed edema, and superficial phlebitis occurred in 22% [8].

Adequacy of visualization with clinical venograms Clinical venograms showed adequate visualization of the veins of the thighs in 108 of 111 (97%) and adequate visualization of the veins of the calves in 109

PART II

Diagnosis of DVT

of 111 (98%) [6]. Some indicate, however, that filling is inadequate in as many as 5% of patients, and there is disagreement about the presence of thrombi in approximately 10% of patients [9].

Correlation of contrast venography with I-125-fibrinogen uptake Among patients with normal contrast venography of the calves, I-125-fibrinogen uptake showed DVT in 8 of 71 (11%) in one investigation [6] and 10 of 41 (24%) in another [8].

Correlation of postmortem contrast venography with dissection Postmortem contrast venography in 47 limbs compared with careful extensive dissection showed 87– 100% correlation with thrombi in the femoral vein, deep femoral vein, popliteal vein, and greater saphenous vein [10]. However, postmortem venography of the veins of the calves was an unreliable guide to the presence of thrombosis because of apparent filling defects and absence of vessels [10]. Postmortem venography had the advantage over clinical contrast venography of being able to inject an unlimited volume of contrast material. The potential problem of false positive filling defects caused by postmortem clots was rare. In the veins of the thighs, intraluminal filling defects, absence of vessels, dilated veins, enlarged valve pockets, and collateral circulation were readily shown.

References 1 Redman HC. Deep venous thrombosis: is contrast venography still the diagnostic “gold standard”? Radiology 1988; 168: 277–278. 2 Stein PD, Hull RD, Ghali WA et al. Tracking the uptake of evidence: two decades of hospital practice trends for diagnosing deep venous thrombosis and pulmonary embolism. Arch Intern Med 2003; 163: 1213–1219. 3 Cham MD, Yankelevitz DF, Shaham D et al., for The Pulmonary Angiography-Indirect CT Venography Cooperative Group. Deep venous thrombosis: detection by using indirect CT venography. Radiology 2000; 216: 744–751.


BLUK077-Stein

April 6, 2007

CHAPTER 34

15:24

Contrast venography

4 Gottschalk A, Stein PD, Goodman LR, Sostman HD. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173–182. 5 Rabinov. K, Paulen S. Venography of the lower extremities. In: Abrams HL, ed. Abrams Angiography, 3rd edn. Little Brown & Co., Boston, Massachusetts, 1983: 1885– 1887. 6 Bettmann MA, Robbins A, Braun SD, Wetzner S, Dunnick NR, Finkelstein J. Contrast venography of the leg: diagnostic efficacy, tolerance, and complication rates with ionic and nonionic contrast media. Radiology 1987; 165: 113–116.

163

7 Albrechtsson U, Olsson CG. Thrombotic side-effects of lower-limb phlebography. Lancet 1976; 1: 723–724. 8 Lensing AW, Prandoni P, Buller HR, Casara D, Cogo A, ten Cate JW. Lower extremity venography with iohexol: results and complications. Radiology 1990; 177: 503– 505. 9 Baldt MM, Zontsich T, Stumpflen A et al. Deep venous thrombosis of the lower extremity: efficacy of spiral CT venography compared with conventional venography in diagnosis. Radiology 1996; 200: 423–428. 10 Stein PD, Evans H. An autopsy study of leg vein thrombosis. Circulation 1967; 35: 671–681.


BLUK077-Stein

April 6, 2007

35

15:28

CHAPTER 35

Compression ultrasound for the diagnosis of deep venous thrombosis

Definitions Conventional gray-scale real-time ultrasonography, as the name indicates, gives a gray-scale real-time (Bmode) two-dimensional image of the vein. The image is obtained by real-time computation of the reflected signals from an array of ultrasound sources. The reflections are due to boundaries between adjacent structures with different acoustic properties [1]. Duplex ultrasonography is the combined use of conventional gray-scale real-time ultrasonography with pulsed Doppler technology. The combination permits simultaneous imaging of anatomic structures and audible or visual characterization of venous and arterial blood flow [1]. Color duplex ultrasonography gives Doppler shift information from moving erythrocytes. Color coding shows both velocity and direction of flow. This is superimposed in the real-time gray-scale image [1].

Equipment preference As long as compression is used, a clear advantage of one system over another has not been shown in prospective clinical trials [1]. Even so, nonocclusive thrombi may be more easily documented with color flow imaging, and studies of obese patients are generally more easily achieved with this technique [1].

Diagnostic criteria for acute DVT Noncompressibility of the vein is the most reliable sign of acute deep venous thrombosis (DVT) [1]. In Prospective Investigation of Pulmonary Embolism

164

Diagnosis II (PIOPED II), color duplex imaging of the common femoral, superficial femoral, popliteal, and proximal greater saphenous veins was used [2]. Acute DVT was diagnosed if there was noncompressibility of the vein in combination with at least one of the following: an enlarged vein, a hypoechoic lumen, or absence of collaterals [2]. Usually, an enlarged vein is found in combination with noncompressibility (T. Wakefield, personal communication, 2006). Some use, as a secondary diagnostic criterion, an echogenic thrombus within the vein lumen [1]. This may be false positive, however, and acute thrombi may be anechoic [1]. Loss of flow phasicity, response to Valsalva, and augmentation have also been cited as secondary diagnostic criteria [1]. An abbreviated imaging study of only the common femoral veins and popliteal veins but not the superficial femoral veins has been used by some to conserve time and resources. However, among 269 patients with DVT identified by compression ultrasound, the DVT was isolated to the superficial femoral vein in 60 patients (22%) [3]. The remaining patients showed DVT in the common femoral or popliteal veins or both [3]. Abbreviated imaging studies, therefore, may fail to identify DVT in a significant number of patients.

Accuracy of compression ultrasound Symptomatic DVT In patients with clinically suspected symptomatic acute DVT, compression ultrasound of the proximal deep veins is highly sensitive and specific. In prospective investigations of symptomatic patients with suspected DVT, using contrast venography as a reference


BLUK077-Stein

April 6, 2007

CHAPTER 35

15:28

165

Compression ultrasound for diagnosis of DVT

Symptomatic High risk asymptomatic

100 97 97

94

96 96

97

Figure 35.1 Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of compression ultrasound in symptomatic patients with deep venous thrombosis (DVT) and in high-risk asymptomatic patients with DVT. Data are pooled data described in text.

Percent

80 60

68

63

40 20 0 Sensitivity

standard with blinded readers, sensitivity with realtime B-mode ultrasound [4–6], duplex ultrasound [7, 8], and color duplex ultrasound [9, 10], the respective sensitivities were 192 of 207 (93%), 120 of 126 (95%), and 45 of 47 (96%). Specificities were 293 of 299 (98%), 111 of 120 (93%), and 124 of 125 (99%). All these patients had ultrasonic examination of the common femoral vein, superficial femoral, and popliteal vein. On average, using any of these three methods with compression ultrasound (recognizing heterogeneity of the methods, so these averages are estimates), sensitivity was 94%, specificity 97%, positive predictive value 97%, and negative predictive value 96% (Figure 35.1).

Asymptomatic DVT In asymptomatic patients with a high risk for DVT, sensitivity 62 of 99 (63%) and positive predictive value 62 of 91 (68%) of compression ultrasound were lower than in symptomatic patients (Figure 35.1) [11–17]. Similar values to those in symptomatic patients, however, were shown with specificity 1024 of 1060 (97%) and negative predictive value 1001 of 1038 (96%).

Calf veins The sensitivity of compression ultrasound for isolated DVT of the calf veins, in a limited number of symptomatic patients, was 92 of 112 (82%) [6, 10, 18]. For this reason, serial studies to examine for extension to proximal veins are recommended [1]. In the event that serial studies cannot be obtained, contrast venography has been recommended [1].

Specificity

PPV

NPV

The sensitivity of compression ultrasound for isolated DVT of the calf veins, in a small number of asymptomatic patients, was low, 36 of 58 (62%) [14, 17, 19].

Compression ultrasound in patients with acute pulmonary embolism Among 149 patients with acute pulmonary embolism (PE) diagnosed by a high-probability lung scan or positive pulmonary angiogram, bilateral B-mode grayscale compression ultrasonography of the proximal leg veins showed DVT in 43 of 149 (29%) [20]. Others showed DVT by B-mode compression ultrasound in 70 of 162 patients with PE (43%) diagnosed by a high-probability ventilation–perfusion lung scan and in 35 of 60 (58%) diagnosed by a high-probability ventilation–perfusion lung scan in combination with a high clinical probability [21]. Autopsies have shown that 80% or more of patients with PE have thrombi that originate in the lower extremities [22–25]. Even so, only 11% of patients with acute PE had clinically apparent DVT in PIOPED [26] and 47% in PIOPED II [27]. Therefore, the limited sensitivity of compression ultrasound for the detection of DVT in patients with acute PE may partially reflect the fact that the patients are asymptomatic from DVT. Among 178 patients in whom PE was excluded, compression ultrasound showed DVT in 5 of 178 (3%) [20]. Others showed a lower prevalence of coincidental DVT, 2 of 334 (0.6%) [21].


BLUK077-Stein

April 6, 2007

15:28

166

Outcome studies of untreated patients with suspected DVT and negative compression ultrasound: conversion to positive after 1 week Among outpatients with suspected DVT who had negative initial real-time B-mode ultrasonography and were untreated, 2% showed DVT on days 2–8 [28– 31]. Detailed results were 7 of 342 (2%) [28], 12 of 1302 (0.9%) [29], 0 of 118 (0%) [30], and 6 of 413 (1.5%) [31]. One of the 12 patients evaluated by Cogo and associates died of massive PE during the first week after a negative ultrasound [29].

Outcome of patients with suspected DVT and negative ultrasound after repeat testing Patients with suspected DVT were followed off treatment for 3 months after repeat testing with ultrasound in ≤8 days of the first negative test [28–31]. Venous thromboembolic events occurred in 2 of 335 (0.6%) [28], 8 of 1290 (0.6%) [29], and 6 of 378 (1.6%) [30]. Among 118 outpatients with suspected DVT, 1.3% of the negative ultrasound patients developed PE or DVT in 3 months [31]. Results in inpatients were quite different, however. Among 56 inpatients with suspected DVT, 10% of the ultrasound negative patients developed thromboembolic complications within 3 months [31]. It has been shown that compression ultrasonography repeated only once at 5–7 days, and if negative, permits a safe withholding of anticoagulants in patients with suspected DVT [28].

Outcome studies of compression ultrasound and clinical score See Chapter 30.

Outcome studies of compression ultrasound and D-dimer See Chapter 33.

Comparison of compression ultrasound and CT venous phase imaging for DVT There was 95.5% concordance between CT venous phase imaging and compression ultrasound [32].

PART II

Diagnosis of DVT

Kappa statistic was 0.809 (“almost perfect”). Among patients with PE by the reference test the proportion with a positive CT venogram and positive venous ultrasound was nearly the same and showed no statistically significant difference (32).

References 1 Tapson VF, Carroll BA, Davidson BL et al., for the American Thoracic Society. The diagnostic approach to acute venous thromboembolism. Clinical practice guideline. Am J Respir Crit Care Med 1999; 160: 1043–1066. 2 Gottschalk A, Stein PD, Goodman LR, Sostman HD. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173– 182. 3 Maki DD, Kumar N, Nguyen B, Langer JE, Miller WT, Jr, Gefter WB. Distribution of thrombi in acute lower extremity deep venous thrombosis: implications for sonography and CT and MR venography. Am J Roentgenol 2000; 175: 1299–1301. 4 Cronan JJ, Dorfman GS, Scola FH, Schepps B, Alexander J. Deep venous thrombosis: US assessment using vein compression. Radiology 1987; 162: 191–194. 5 Pedersen OM, Aslaksen A, Vik-Mo H, Bassoe AM. Compression ultrasonography in hospitalized patients with suspected deep venous thrombosis. Arch Intern Med 1991; 151: 2217–2220. 6 Lensing AW, Prandoni P, Brandjes D et al. Detection of deep-vein thrombosis by real-time B-mode ultrasonography. N Engl J Med 1989; 320: 342–345. 7 O’Leary DH, Kane RA, Chase BM. A prospective study of the efficacy of B-scan sonography in the detection of deep venous thrombosis in the lower extremities. J Clin Ultrasound 1988; 16: 1–8. 8 Mitchell DC, Grasty MS, Stebbings WS et al. Comparison of duplex ultrasonography and venography in the diagnosis of deep venous thrombosis. Br J Surg 1991; 78: 611–613. 9 Lewis BD, James EM, Welch TJ, Joyce JW, Hallett JW, Weaver AL. Diagnosis of acute deep venous thrombosis of the lower extremities: prospective evaluation of color Doppler flow imaging versus venography. Radiology 1994; 192: 651–655. 10 Rose SC, Zwiebel WJ, Nelson BD et al. Symptomatic lower extremity deep venous thrombosis: accuracy, limitations, and role of color duplex flow imaging in diagnosis. Radiology 1990; 175: 639–644. 11 Froehlich JA, Dorfman GS, Cronan JJ, Urbanek PJ, Herndon JH, Aaron RK. Compression ultrasonography for the detection of deep venous thrombosis in patients who have a fracture of the hip. A prospective study. J Bone Joint Surg Am 1989; 71: 249–256.


BLUK077-Stein

April 6, 2007

CHAPTER 35

15:28

Compression ultrasound for diagnosis of DVT

12 Barnes CL, Nelson CL, Nix ML, McCowan TC, Lavender RC, Barnes RW. Duplex scanning versus venography as a screening examination in total hip arthroplasty patients. Clin Orthop Relat Res 1991; 271: 180–189. 13 Woolson ST, Pottorff G. Venous ultrasonography in the detection of proximal vein thrombosis after total knee arthroplasty. Clin Orthop Relat Res 1991; 273: 131–135. 14 Elliott CG, Suchyta M, Rose SC et al. Duplex ultrasonography for the detection of deep vein thrombi after total hip or knee arthroplasty. Angiology 1993; 44: 26–33. 15 Davidson BL, Elliott CG, Lensing AW, for the RD Heparin Arthroplasty Group. Low accuracy of color Doppler ultrasound in the detection of proximal leg vein thrombosis in asymptomatic high-risk patients. Ann Intern Med 1992; 117: 735–738. 16 Mattos MA, Londrey GL, Leutz DW et al. Color-flow duplex scanning for the surveillance and diagnosis of acute deep venous thrombosis. J Vasc Surg 1992; 15: 366–375; discussion 375–376. 17 Lensing AW, Doris CI, McGrath FP et al. A comparison of compression ultrasound with color Doppler ultrasound for the diagnosis of symptomless postoperative deep vein thrombosis. Arch Intern Med 1997; 157: 765–768. 18 Habscheid W, Hohmann M, Wilhelm T, Epping J. Realtime ultrasound in the diagnosis of acute deep venous thrombosis of the lower extremity. Angiology 1990; 41: 599–608. 19 Rose SC, Zwiebel WJ, Murdock LE et al. Insensitivity of color Doppler flow imaging for detection of acute calf deep venous thrombosis in asymptomatic postoperative patients. J Vasc Interv Radiol 1993; 4: 111–117. 20 Turkstra F, Kuijer PM, van Beek EJ, Brandjes DP, ten Cate JW, Buller HR. Diagnostic utility of ultrasonography of leg veins in patients suspected of having pulmonary embolism. Ann Intern Med 1997; 126: 775–781. 21 Wells PS, Ginsberg JS, Anderson DR et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998; 129: 997– 1005. 22 Sevitt S, Gallagher N. Venous thrombosis and pulmonary embolism: a clinico-pathological study in injured and burned patients. Br J Surg 1961; 48: 475–489.

167

23 Cohn R, Walsh J. The incidence and anatomical site of origin of pulmonary emboli. Stanford Med Bull 1946; 4: 97–99. 24 Short DS. A survey of pulmonary embolism in a general hospital. BMJ 1952; 1: 790–796. 25 Byrne JJ, O’Neil EE. Fatal pulmonary emboli. A study of 130 autopsy-proven fatal emboli. Am J Surg 1952; 83: 47–49. 26 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 27 Stein PD, Beemath A, Matta F et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med 2007 (In press). 28 Birdwell BG, Raskob GE, Whitsett TL et al. The clinical validity of normal compression ultrasonography in outpatients suspected of having deep venous thrombosis. Ann Intern Med 1998; 128: 1–7. 29 Cogo A, Lensing AW, Koopman MM et al. Compression ultrasonography for diagnostic management of patients with clinically suspected deep vein thrombosis: prospective cohort study. BMJ 1998; 316: 17–20. 30 Heijboer H, Buller HR, Lensing AW, Turpie AG, Colly LP, ten Cate JW. A comparison of real-time compression ultrasonography with impedance plethysmography for the diagnosis of deep-vein thrombosis in symptomatic outpatients. N Engl J Med 1993; 329: 1365– 1369. 31 Sluzewski M, Koopman MM, Schuur KH, van Vroonhoven TJ, Ruijs JH. Influence of negative ultrasound findings on the management of in- and outpatients with suspected deep-vein thrombosis. Eur J Radiol 1991; 13: 174–177. 32 Goodman LG, Stein PD, Beemath A, Sostman HD, Wakefield TW, Woodard PK, Yankelevitz DF. Equivalence of venous phase CT venograms and compression ultrasound in the evaluation of patients with suspected acute pulmonary embolism: data from the Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II). (Submitted for publication).


BLUK077-Stein

March 12, 2007

36

18:9

CHAPTER 36

Impedance plethysmography and fibrinogen uptake tests for diagnosis of deep venous thrombosis

Impedance plethysmography Impedance plethysmography (IPG) has been largely replaced by compression ultrasonography for the diagnosis of deep venous thrombosis (DVT). Data from the National Hospital Discharge Survey show that from 1990 through 2003 fewer than 5000 impedance plethysmograms were coded at hospital discharge of all patients in the United States (Stein PD, unpublished). Impedance plethysmography has the advantage of being less expensive than other diagnostic tests for DVT; it is portable and it requires less technician training [1]. Even so, it is operator dependent, and requires trained technicians and adherence to a validated protocol [2]. Impedance plethysmography is sensitive for proximal DVT, but not distal DVT. Only 17â&#x20AC;&#x201C;25% of limbs with isolated calf DVT showed an abnormality with IPG [3, 4]. Among patients with DVT in the proximal veins shown by venography, 86% (30 of 35) had a positive impedance plethysmogram [5]. Others showed that impedance plethysmography was indicative of DVT in 15 of 16 (94%) of legs that showed positive venograms of both the proximal and distal leg veins [4]. Hull and associates showed that 124 of 133 limbs with DVT (93%) were identified by IPG and 386 of 397 limbs without DVT (97%) showed a negative IPG [3]. One study of IPG, however, showed a sensitivity of only 26 of 40 (65%) and specificity of 79 of 85 (93%) for proximal DVT [6]. This may have reflected referral of patients with less severe symptoms and smaller, less occlusive thrombi [6]. On average, based on 10 investigations, IPG for proximal DVT showed a sensitivity

168

of 650 of 738 (88%) and specificity of 330 of 354 (93%) [7]. Outcome study has shown that only 9 of 365 (2.5%) of patients with suspected DVT and negative serial impedance plethysmograms over a period of 8 days had symptoms of PE (pulmonary embolism) or DVT in 6 months [8]. Although IPG detects proximal DVT in approximately 88% of patients, noninvasive testing for DVT in patients with documented PE is positive in only 43â&#x20AC;&#x201C; 57% of the patients [5, 9]. This may reflect the following possibilities: (1) The residual DVT may be nonobstructive or confined to the calf and remain undetected. (2) The deep venous thrombi causing the pulmonary emboli may embolize entirely from the deep veins of the thighs, leaving no residual thrombosis to detect. (3) The source of the pulmonary emboli may be from sites other than the lower extremities (for example, deep pelvic veins, renal veins, and subclavian veins). Comparison of IPG with compression ultrasound has shown the latter to be more accurate [10]. Among symptomatic patients with clinically suspected DVT, the sensitivity of IPG was 100 of 130 (77%) versus 117 of 130 (90%) with compression ultrasound. The specificities were 339 of 365 (93%) with IPG and 358 of 365 (98%) with compression ultrasound.

Outcome of patients with suspected DVT who had a single negative impedance plethysmogram Only 1 of 284 patients with suspected DVT (0.4%) and a negative impedance plethysmogram developed DVT on follow-up [11]. None developed PE.


BLUK077-Stein

March 12, 2007

CHAPTER 36

18:9

Impedance plethysmography and fibrinogen uptake tests

Outcome of patients with suspected DVT who had negative serial impedance plethysmograms Among 139 pregnant women with suspected DVT and negative serial impedance plethysmograms obtained over 14 days, none developed DVT or PE within a 3month follow-up period [12]. Among 289 patients with suspected DVT who had negative serial impedance plethysmograms obtained over 10 days, none developed PE within a 6-month follow-up period after all serial tests were completed [13]. One patient, however, may have had a minor PE during the 10-day period of serial testing [13]. One patient (0.3%) had DVT 9 weeks after the initial evaluation.

7

8

9

10

Fibrinogen uptake test Fibrinogen leg scanning, also called the fibrinogen uptake test, had been used extensively to detect subclinical DVT [14]. The test is no longer available because of concerns about the potential for viral transmission with this human blood product [15]. Geerts et al. [15] indicated that the fibrinogen uptake test lacks both specificity and sensitivity for the detection of DVT [16–21] and is poorly correlated with major thromboembolic events [22].

References 1 Tapson VF, Carroll BA, Davidson BL et al., for the American Thoracic Society. The diagnostic approach to acute venous thromboembolism. Clinical practice guideline. Am J Respir Crit Care Med 1999; 160: 1043–1066. 2 Taylor DW, Hull R, Sackett DL, Hirsh J. Simplification of the sequential impedance plethysmograph technique without loss of accuracy. Thromb Res 1980; 17: 561–565. 3 Hull R, van Aken WG, Hirsh J et al. Impedance plethysmography using the occlusive cuff technique in the diagnosis of venous thrombosis. Circulation 1976; 53: 696– 700. 4 Moser KM, LeMoine JR. Is embolic risk conditioned by location of deep venous thrombosis? Ann Intern Med 1981; 94: 439–444. 5 Hull RD, Hirsh J, Carter CJ et al. Pulmonary angiography, ventilation lung scanning, and venography for clinically suspected pulmonary embolism with abnormal perfusion lung scan. Ann Intern Med 1983; 98: 891–899. 6 Ginsberg JS, Wells PS, Hirsh J et al. Reevaluation of the sensitivity of impedance plethysmography for the

11

12

13

14 15

16

17

18

19

169

detection of proximal deep vein thrombosis. Arch Intern Med 1994; 154: 1930–1933. Kraaijenhagen RA, Lensing AW, Wallis JW, van Beek EJ, ten Cate JW, Buller HR. Diagnostic management of venous thromboembolism. Baillieres Clin Haematol 1998; 11: 541–586. Heijboer H, Buller HR, Lensing AW, Turpie AG, Colly LP, ten Cate JW. A comparison of real-time compression ultrasonography with impedance plethysmography for the diagnosis of deep-vein thrombosis in symptomatic outpatients. N Engl J Med 1993; 329: 1365–1369. Hull RD, Hirsh J, Carter CJ et al. Diagnostic value of ventilation–perfusion lung scanning in patients with suspected pulmonary embolism. Chest 1985; 88: 819–828. Wells PS, Hirsh J, Anderson DR et al. Comparison of the accuracy of impedance plethysmography and compression ultrasonography in outpatients with clinically suspected deep vein thrombosis. A two centre paireddesign prospective trial. Thromb Haemost 1995; 74: 1423– 1427. Wheeler HB, Anderson FA, Jr, Cardullo PA, Patwardhan NA, Jian-Ming L, Cutler BS. Suspected deep vein thrombosis. Management by impedance plethysmography. Arch Surg 1982; 117: 1206–1209. Hull RD, Rascob GE, Carter CJ. Serial impedance plethysmography in pregnant patients with clinically suspected deep-vein thrombosis. Clinical validity of negative findings. Ann Intern Med 1990; 112: 663–667. Huisman MV, Buller HR, ten Cate JW, Vreeken J. Serial impedance plethysmography for suspected deep venous thrombosis in outpatients: The Amsterdam General Practitioner Study. N Engl J Med 1986; 314: 823–828. Kakkar VV. The diagnosis of deep vein thrombosis using the 125 I fibrinogen test. Arch Surg 1972; 104: 152–159. Geerts WH, Pineo GF, Heit JA et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 338S–400S. Moskovitz PA, Ellenberg SS, Feffer HL et al. Low-dose heparin for prevention of venous thromboembolism in total hip arthroplasty and surgical repair of hip fractures. J Bone Joint Surg Am 1978; 60: 1065–1070. Fauno P, Suomalainen O, Bergqvist D et al. The use of fibrinogen uptake test in screening for deep vein thrombosis in patients with hip fracture. Thromb Res 1990; 60: 185–190. Lensing AW, Hirsh J. 125 I-fibrinogen leg scanning: reassessment of its role for the diagnosis of venous thrombosis in post-operative patients. Thromb Haemost 1993; 69: 2–7. Agnelli G, Radicchia S, Nenci GG. Diagnosis of deep vein thrombosis in asymptomatic high-risk patients. Haemostasis 1995; 25: 40–48.


BLUK077-Stein

March 12, 2007

18:9

170

20 Bergqvist D, Burmark US, Flordal PA et al. Low molecular weight heparin started before surgery as prophylaxis against deep vein thrombosis: 2500 versus 5000 XaI units in 2070 patients. Br J Surg 1995; 82: 496–501. 21 Nurmohamed MT, Verhaeghe R, Haas S, et al. A comparative trial of a low molecular weight heparin (enoxaparin) versus standard heparin for the; prophylaxis of postope-

PART II

Diagnosis of DVT

rative deep vein thrombosis in general surgery. Am J Surg 1995 169: 567–571. 22 Flordal PA, Bergqvist D, Ljungstrom KG et al. Clinical relevance of the fibrinogen uptake test in patients undergoing elective general abdominal surgery: relation to major thromboembolism and mortality. Fragmin Multicentre Study Group. Thromb Res 1995; 80: 491–497.


BLUK077-Stein

April 14, 2007

37

14:54

CHAPTER 37

Computed tomography for diagnosis of deep venous thrombosis

Ascending venography Spiral CT ascending venography compared with conventional contrast venography In an evaluation of spiral CT venography (CTV) compared with conventional contrast venography combined with duplex ultrasonography and follow-up, 52 patients with suspected deep venous thrombosis (DVT) were evaluated [1]. Sensitivity was 100%. Regarding results according to limb and site, sensitivity in calf veins was 36 of 36, thigh veins 32 of 32, and pelvic veins 13 of 13. Specificities in limbs were calf veins 64 of 68 (94%), thigh veins 68 of 70 (97%), and thigh veins 88 of 90 (98%). False-positive results with CTV may have been caused by false-negative results of conventional venography [1]. Spiral CTV more clearly demonstrated extension of DVT into the pelvic veins and inferior vena cava than conventional contrast venography [1]. In another investigation of 102 lower extremities compared with ascending venography and color duplex ultrasonography, sensitivity was 100% and specificity was 96% [2]. The quality of CTV was superior to ascending venography in all veins [2]. Methods of ascending CTV The method of ascending CTV used by Baldt and associates is as follows: A 22-gauge cannula was inserted in a dorsal vein of both feet and a tourniquet was placed around both ankles [1]. Forty milliliters of iohexol diluted with 200 mL of saline was injected via a Y adapter simultaneously into both legs at 4 mL/s using a power injector. This achieved a flow of 2 mL/s for both legs. After a 35-second delay, a 100-cm section was imaged from the ankle to the inferior vena cava. Immediately

after the study, 100 mL of saline was injected into both cannulas to flush contrast material from the veins. A single slice scanner was used. From the data set, 200 images with a nominal thickness of 5 mm were reconstructed.

Complications Complications of ascending contrast venography are described in Chapter 34. Complications relate to local and systemic effects of contrast material, and presumably the complications rate are the same with CTV and ascending contrast venography.

Venous phase CTV Validity of CT venous phase imaging in patients with suspected acute pulmonary embolism On average, CTV of the veins of the calf, thigh, and pelvis in patients with suspected PE (pulmonary embolism) showed a sensitivity of 34 of 37 (92%) and specificity of 71 of 82 (87%) using ascending venography as a reference standard (Table 37.1) [3, 4]. Venous phase imaging limited to the veins of the thighs showed an average sensitivity of 83 of 85 (98%) and average specificity of 399 of 405 (99%) using compression ultrasonography as a reference standard (Table 37.1) [5â&#x20AC;&#x201C;7]. With an average prevalence of DVT of 17% in these patients, positive predictive value was 83 of 89 (93%) and negative predictive value was 399 of 401 (99.5%) (Figure 37.1) [5â&#x20AC;&#x201C;7]. Venous phase imaging of the veins of the thighs also was sensitive and specific based on Doppler sonography as the reference standard [8] (Table 37.1). Technical quality of CTV, among 541 patients, was excellent in 47%, good in 29%, fair in 17%, and poor in 6% [7]. 171


BLUK077-Stein

April 14, 2007

14:54

172

PART II

Diagnosis of DVT

Table 37.1 Sensitivity and specificity of spiral CT venous phase imaging in patients with suspected pulmonary embolism. n/N (%) First author (Ref)

Number slice CT

Sites evaluated

Reference test

Sensitivity

Coche [3]

2

Mid-calf, thigh, pelvis

Venogram

16/17 (94)

Ghaye [4]

1

Calf, thigh, pelvis

Venogram

18/20 (90)

17/26 (65)

Ghaye [4]

1

Calf, thigh, pelvis

Compression ultrasound

82/84 (98)

81/93 (87)

Garg [5]

1

Thigh

Compression ultrasound

Loud [6]

1

Thigh

Compression ultrasound

63/65 (97)

243/243 (100)

Cham [7]

1, 2, 4

Thigh

Compression ultrasound

15/15 (100)

93/97 (96)

Begemann [8]

4

Thigh

Doppler sonography

11/11 (100)

28/29 (97)

Methods The methods for CT venous phase imaging of the veins of the pelvis and lower extremities used in the Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II) were as follows [9]. An injection of 135 mL of low osmolar nonionic contrast material (300– 320 mg iodine/mL) was made into an arm vein at 4 mL/s. Scans of the iliac, femoral, and popliteal veins were obtained 3 minutes after the beginning of the injection of contrast material for chest CT. Scans commenced at the iliac crest and continued in a helical mode to the level of the knees. Veins of the calves were not studied because CT venous phase imaging usually fails to opacify the calf veins. Diagnostic criteria for DVT The diagnostic criteria for acute DVT based on venous phase imaging in the PIOPED II [9] were: Occlusive = complete filling defect, i.e., failure to opacify the entire lumen due to a central filling defect (the vessel may enlarge compared to the opposite vein); Nonocclusive = partial filling defect surrounded by contrast material.

5/5 (100)

Specificity 54/56 (96)

63/65 (97)

The criteria for chronic DVT are not well worked out in the CT literature, but are well established in the venous compression ultrasound literature. The following criteria were designated as DVT of unknown age: a Complete filling defect, but the vein is smaller than its peers; b Contrast flowing through a thickened, often smaller vein; c Calcification in vein; d Secondary sign, i.e., increased collateral vessels.

Increased sensitivity for venous thromboembolism using venous phase imaging in patients with suspected acute PE In patients with suspected acute PE, the number of patients with only a positive contrast-enhanced chest CT, positive chest CT and positive CTV, and only a positive CTV is shown in Table 37.2 [3, 5–8, 10–15]. Images were obtained with 1-slice, 2-slice, 4-slice, 8-slice, and 16-slice CT (Table 37.2). Pooled data showed that 647 patients with suspected acute PE had a positive chest CT and an additional 121 patients (19%) had only a

100 99.5

95 Percent

90

93

92

85

87

80 75 70 SENS

SPEC

PPV

NPV

Figure 37.1 Sensitivity (sens), specificity (spec), positive predictive value (PPV), and negative predictive value (NPV) of venous phase imaging (CTV) of veins of thighs using conventional contrast venography as a reference standard. Data are pooled data described in text.


BLUK077-Stein

April 14, 2007

CHAPTER 37

14:54

173

Computed tomography for diagnosis of DVT

Table 37.2 Distribution of positive (POS) findings on contrast enhanced chest CT and venous phase imaging (CTV) in patients with suspected pulmonary embolism. POS CT

POS CT chest

POS CT chest

First author (Ref)

Number slice CT

chest only

and CTV

± POS CTV

POS CTV only

POS CTV only/VTE (%)*

Richman [10]

4

31

42

88

15

15/88 (17)

Coche [3]

2

9

13

25

3

3/25 (12)

Loud [6]

1

27

58

116

31

31/116(27)

Garg [5]

1

7

5

12

0

0/12 (0)

Cham [7]

1, 2, 4

62

29

107

16

16/107(15)

Revel [11]

4

33

21

59

5

5/59 (8)

Nicolas [12]

1

18

29

52

5

5/52 (10)

Begemann [8]

4

7

13

31

11

11/31 (35)

Stein [13]

4, 8, 16

71

79

164

14

14/183 (8)

Ghaye [14]

16

15

38

73

20

20/73 (27)

Johnson [15]

4, 8

30

10

41

1

1/41 (2)

CT = computed tomography, POS = positive, Neg = negative, CTV = CT venogram, Anglo = angiogram, VTE = venous thromboembolic disease.

positive CTV [3, 5–8, 10–15] (Figure 37.2). Among patients in whom single slice CT was used for the diagnosis of suspected PE, 36 of 180 (20%) patients with VTE were diagnosed on the basis of a positive CTV [5, 6, 12]. Among patients with suspected acute PE who were shown to have venous thromboembolism by multidetector contrast-enhanced spiral CT, the diagnosis was made by a positive CTV in 66 of 456 (14%) [8, 10, 11, 13–15]. In PIOPED II, CTV’s were obtained using continuous helical imaging with 7.5 mm collimation and 7.5 mm reconstruction (14). Subsequently, a random sample of the original data set was examined after every two of three 7.5 mm images were deleted. The refor-

matted cases, therefore, consisted of 7.5 mm images with a 15 mm gap (14). Per patient agreement with continuous images for DVT was 89%. Use of 5 mm axial images, skipping 15 mm, should yield similar results (14). With 5 mm, skip 15 mm images, plus automated dose modulation, the calculated radiation dose dropped from 9.1 to 3 mSv. In PIOPED II, thrombi isolated to the inferior vena cava and iliac vessels occurred in only 3% of patients (13). Therefore, scanning from the acetabulum to the tibial plateau, rather than from the iliac crest, would further reduce the radiation to 2.0 mSv (16), Continuous helical CTVs and compression ultrasound showed similar results (17) (See Chapter 35).

100 84

VTE (%)

80 60 40 16

20 Figure 37.2 Patients with suspected pulmonary embolism in whom venous thromboembolism (VTE) was diagnosed by multidetector CT.

0 Positive Chest CT

Positive CTV Only


BLUK077-Stein

April 14, 2007

14:54

174

References 1 Baldt MM, Zontsich T, Stumpflen A et al. Deep venous thrombosis of the lower extremity: efficacy of spiral CT venography compared with conventional venography in diagnosis. Radiology 1996; 200: 423–428. 2 Zontsich T, Turetschek K, Baldt M. CT-phlebography. A new method for the diagnosis of venous thrombosis of the upper and lower extremities. Radiologe 1998; 38: 586–590. 3 Coche EE, Hamoir XL, Hammer FD, Hainaut P, Goffette PP. Using dual-detector helical CT angiography to detect deep venous thrombosis in patients with suspicion of pulmonary embolism: diagnostic value and additional findings. AJR Am J Roentgenol 2001; 176: 1035–1039. 4 Ghaye B, Szapiro D, Willems V, Dondelinger RF. Combined CT venography of the lower limbs and spiral CT angiography of pulmonary arteries in acute pulmonary embolism: preliminary results of a prospective study. JBRBTR 2000; 83: 271–278. 5 Garg K, Kemp JL, Wojcik D et al. Thromboembolic disease: comparison of combined CT pulmonary angiography and venography with bilateral leg sonography in 70 patients. Am J Roentgenol 2000; 175: 997–1001. 6 Loud PA, Katz DS, Bruce DA, Klippenstein DL, Grossman ZD. Deep venous thrombosis with suspected pulmonary embolism: detection with combined CT venography and pulmonary angiography. Radiology 2001; 219: 498–502. 7 Cham MD, Yankelevitz DF, Shaham D et al., for the Pulmonary Angiography-Indirect CT Venography Cooperative Group. Deep venous thrombosis: detection by using indirect CT venography. Radiology 2000; 216: 744–751. 8 Begemann PG, Bonacker M, Kemper J et al. Evaluation of the deep venous system in patients with suspected pulmonary embolism with multi-detector CT: a prospective study in comparison to Doppler sonography. J Comput Assist Tomogr 2003; 27: 399–409. 9 Gottschalk A, Stein PD, Goodman LR, Sostman HD. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173–182.

PART II

Diagnosis of DVT

10 Richman PB, Wood J, Kasper DM et al. Contribution of indirect computed tomography venography to computed tomography angiography of the chest for the diagnosis of thromboembolic disease in two United States emergency departments. J Thromb Haemost 2003; 1: 652– 657. 11 Revel MP, Petrover D, Hernigou A, Lefort C, Meyer G, Frija G. Diagnosing pulmonary embolism with fourdetector row helical CT: prospective evaluation of 216 outpatients and inpatients. Radiology 2005; 234: 265– 273. 12 Nicolas M, Debelle L, Laurent V et al. Incremental lower extremity CT venography, a simplified approach for the diagnosis of phlebitis in patients with pulmonary embolism. J Radiol 2001; 82: 251–256. 13 Stein PD, Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 2006; 354: 2317–2327. 14 Ghaye B, Nchimi A, Noukoua C, Dondelinger RF. Does multi-detector row CT pulmonary angiography reduce the incremental value of indirect CT venography compared with single-detector row CT pulmonary angiography? Radiology 2006; 240: 256–262. 15 Johnson JC, Brown MD, McCullough N, Smith S. CT lower extremity venography in suspected pulmonary embolism in the ED. Emerg Radiol 2006; 12: 160–163. 16 Goodman LR, Stein PD, Beemath A, Sostman HD, Wakefield TW, Woodard PK, Yankelevitz D. CT venography: continuous helical images versus reformatted discontinuous images using PIOPED II data. Am J Roentgen 2007 (In press). 17 Goodman LG, Stein PD, Beemath A, Sostman HD, Wakefield TW, Woodard PK, Yankelevitz DF. Equivalence of venous phase CT venograms and compression ultrasound in the evaluation of patients with suspected acute pulmonary embolism: data from the Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II). (Submitted for publication).


BLUK077-Stein

March 28, 2007

38

19:12

CHAPTER 38

Magnetic resonance angiography for diagnosis of deep venous thrombosis

Introduction Magnetic resonance angiography (MRA) has advantages compared to contrast-enhanced spiral CT, which include no ionizing radiation, no iodinated contrast material, and ability to perform repeated examinations safely [1]. Disadvantages are that MRA is expensive and cumbersome [1]. At present, MRA of the veins of the lower extremity is a second-line diagnostic tool because of higher cost, technical limitations, limited availability, and logistical constraints [2]. Magnetic resonance imaging was available, however, in 59% of non-Federal short-stay general and special hospitals in the United States in 2003 [3].

MR venography without contrast material Deep venous thrombosis (DVT) can be readily detected on MRI sequences without requiring a contrast medium. Results with spin-echo-black blood imaging [4], two-dimensional gradient-echo, time-of-flight MR venograms [5â&#x20AC;&#x201C;9], or direct thrombus imaging on spin echo with fat saturation [10] have reported high sensitivity and specificity (87â&#x20AC;&#x201C;100%) for detection of thrombosis of the pelvic and lower extremity veins. A disadvantage is the time required to obtain multiple axial images, and small superficial and perforating veins are often not delineated [6].

Gd-MR venography of the lower extremities Compared to two-dimensional time-of-flight techniques (which do not employ contrast material) three-

dimensional gadolinium-enhanced peripheral venography has a much shorter acquisition time, images look sharper, and more veins can be seen [11]. The use of low-dose contrast permits repeated measurements [11]. The pedal injection of diluted paramagnetic contrast material in combination with a dedicated vascular coil showed images comparable to conventional contrast venograms [7]. Sensitivity for venous segments, compared with contrast venography, was 13 of 13 (100%) and specificity was 88 of 90 (98%).

Delayed-enhanced MR venography of the veins of the thighs Fraser and associates [8] following an antecubital injection of gadopentetate dimeglumine showed that in the femoral vein, Gd-MRV (Gd-MR venography)was sensitive in 20 of 20 (100%) and specific in 31 of 32 (97%). In the iliac vein, Gd-MRV was sensitive in 7 of 7 (100%) and specific in 39 of 39 (100%). A normal Gd-MRA venogram following an intravenous injection of Gd is shown in Figure 38.1. Good visualization of the veins of the lower extremities was obtained with this technique immediately after imaging the pulmonary arteries (Figures 38.2 and 38.3) using a single intravenous injection of a gadoliniumcontaining contrast agent. No additional gadolinium was necessary following Gd-MRA pulmonary angiography. The gadolinium injected for the pulmonary MR angiogram was used for venous phase imaging of the thigh veins. Veins of the calf in general are not clearly shown.

175


BLUK077-Stein

March 28, 2007

19:12

176

(a)

PART II

(b)

Diagnosis of DVT

Femoral vein

(c)

Saphenous vein

Superficial femoral vein (d)

Popliteal vein Figure 38.1 (a) Subtracted maximum intensity projection images of the lower extremity venous system. (b) Contiguous reconstructed transverse images through the data set without subtraction. These transverse images show both veins (arrows) and arteries. (c) Contiguous reconstructed transverse images through the data set without subtraction. These transverse images show both

(a)

veins (arrows) and arteries. (d) Contiguous reconstructed transverse images through the data set without subtraction. These transverse images show both veins (arrows) and arteries. (Courtesy of Thomas Chenevert, PhD, Department of Radiology, University of Michigan, Ann Arbor, MI.)

Thrombus

Figure 38.2 Deep vein thrombosis shown with antecubital injection of gadolinium: comparison of techniques. (a) 2D interleaved fast spoiled gradient recalled echo with fat suppression 512 Ă&#x2014; 192 with one average; 28 locations 5 mm thick in 0.5 minutes. (b) 2D sequential time of flight

(b)

256 Ă&#x2014; 129 with two averages; 31 locations 5 mm thick in 6.3 minutes. (Courtesy of Pamela K. Woodard, MD, Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, MO.)


BLUK077-Stein

March 28, 2007

CHAPTER 38

19:12

177

Magnetic resonance angiography for DVT

Vein Thrombus

Figure 38.3 Reformatted 2D fast spoiled gradient recalled echo showing deep venous thrombosis following an antecubital injection with gadolinium. (Courtesy of Pamela K. Woodard, MD, Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, MO.)

Gd-MRV: diagnostic criteria for acute DVT The diagnostic criteria for acute DVT on Gd-MRV are: Occlusive = complete filling defect, i.e., failure to opacify the entire lumen due to a central filling defect (the vessel may enlarge compared to the opposite vein); Nonocclusive = partial filling defect surrounded by opacification.

Comprehensive examination The use of MRV of the veins of the lower extremities in combination with Gd-MRA of the pulmonary arteries creates a comprehensive study for thromboembolism comparable to the combination of contrast-enhanced pulmonary spiral CT of the pulmonary arteries in combination with venous phase CT of the veins of the lower extremities (CTV) [12, 13]. A pilot study of Gd-MRA in 20 patients indicated that such a comprehensive study was feasible and appeared to be a reliable and rapid diagnostic approach [14].

Gd-MRV of the upper extremities and central veins Patients with suspected thrombois, stenosis, occlusion, or compression of the axillary vein, subclavian vein, brachiocephalic vein, or superior or inferior vena cava were imaged with three-dimensional Gd-MRA [15]. The Gd-MRA was sensitive in 4 of 4 (100%) with

axillary or subclavian vein thrombosis [15]. Absence of disease was correctly identified in 13 of 13 (100%). A high number of patients, however, are not able to undergo MRA in this setting [16].

Complications Severe reactions to magneto-pharmaceuticals are rare, 1 in 350,000â&#x20AC;&#x201C;450,000 [17]. In the 2 reported cases of anaphylactoid reactions, the patients had a history of reactive airway disease [17]. Any adverse reaction to gadolinium contrast material has been reported in 0.06% [18] but perhaps as many as 1% may have mild allergic reactions (hives) according to the package insert [19]. In June 2006, the US Food and Drug Administration (FDA) issued an alert indicating that nephrogenic systemic fibrosis or nephrogenic fibrosing dermopathy (NSF/NFD) had been reported in patients with advanced renal failure who received high doses of gadolinium-containing contrast agent [20]. In December 2006, the FDA reported NSF/NFD following gadolilnium-containing contrast agents in patients with only moderately severe renal disease [21]. This is discussed in Chapter 81.

References 1 Sostman HD. MRA for diagnosis of venous thromboembolism. Q J Nucl Med 2001; 45: 311â&#x20AC;&#x201C;323. 2 Kanne JP, Lalani TA. Role of computed tomography and magnetic resonance imaging for deep venous thrombosis


BLUK077-Stein

March 28, 2007

19:12

178

3

4

5

6

7

8

9

10

11

12

and pulmonary embolism. Circulation 2004; 109: 115– 121. AHA Hospital Statistics, 2005 Edition. Health Forum LLC (an affiliate of American Hospital Association), Chicago, 2005: 159. Catalano C, Pavone P, Laghi A et al. Role of MR venography in the evaluation of deep venous thrombosis. Acta Radiol 1997; 38: 907–912. Spritzer CE, Norconk JJ, Jr, Sostman HD, Coleman RE. Detection of deep venous thrombosis by magnetic resonance imaging. Chest 1993; 104: 54–60. Ruehm SG, Zimny K, Debatin JF. Direct contrastenhanced 3D MR venography. Eur Radiol 2001; 11: 102– 112. Ruehm SG, Wiesner W, Debatin JF. Pelvic and lower extremity veins: contrast-enhanced three-dimensional MR venography with a dedicated vascular coil-initial experience. Radiology 2000; 215: 421–427. Fraser DG, Moody AR, Davidson IR, Martel AL, Morgan PS. Deep venous thrombosis: diagnosis by using venous enhanced subtracted peak arterial MR venography versus conventional venography. Radiology 2003; 226: 812–820. Gottschalk A, Stein PD, Goodman LR et al. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173–182. Botner RM, Buecker A, Wiethoff AJ et al. In vivo magnetic resonance imaging of coronary thrombosis using a fibrin binding molecular magnetic resonance contrast agent. Circulation 2004; 110: 1463–1466. Li W, David V, Kaplan R, Edelman RR. Three-dimensional low dose gadolinium-enhanced peripheral MR venography. J Magn Reson Imaging 1998; 8: 630–633. Loud PA, Katz DS, Klippenstein DL, Shah RD, Grossman ZD. Combined CT venography and pulmonary angiography in suspected thromboembolic disease: diagnostic accuracy for deep venous evaluation. AJR 2001; 174: 61– 65.

PART II

Diagnosis of DVT

13 Cham MD, Yankelevitz DF, Shaham D et al., for the Pulmonary CT Angiography-Indirect CT Venography Cooperative Group. Deep venous thrombosis: detection by using indirect CT venography. Radiology 2000; 216: 744– 751. 14 Obernosterer A, Aschauer M, Portugaller H, Koppel H, Lipp RW. Three-dimensional gadolinium-enhanced magnetic resonance angiography used as a “one-stop shop” imaging procedure for venous thromboembolism: a pilot study. Angiology 2005; 56: 423–430. 15 Thornton MJ, Ryan R, Varghese JC, Farrell MA, Lucey B, Lee MJ. A three-dimensional gadolinium-enhanced MR venography technique for imaging central veins. AJR Am J Roentgenol 1999; 173: 999–1003. 16 Baarslag HJ, Van Beek EJ, Reekers JA. Magnetic resonance venography in consecutive patients with suspected deep vein thrombosis of the upper extremity: initial experience. Acta Radiol 2004; 45: 38–43. 17 Carr JJ. Magnetic resonance contrast agents for neuroimaging. Safety issues. Neuroimaging Clin N Am 1994; 4: 43–54. 18 Cochran ST, Bornyea K, Sayre JW. Trends in adverse events after IV administration of contrast media. AJR 2001; 176: 1385–1388. 19 Package Insert: Magnevist. Berlex Laboratories, Wayne, NJ. Revised May 2000. 20 U.S. Food and Drug Administration, Healthcare professional sheet. Gadolinium-containing contrast agents for magnetic resonance imaging (MRI) (marketed as Omniscan, OptiMARK, Magnevist, ProHance and MultiHance). http://www.fda.gov/cder/drug/InfoSheets/HCP/ gccaHCP.htm Accessed 23 June 2006. 21 US Food and Drug Administration Public Health Advisory. Update on magnetic resonance imaging (MRI) contrast agents containing gadolinium and nephrogenic fibrosing dermopathy. http://www.fda.gov/cder/drug/ advisory/gadolinium agents 20061222.htm.


BLUK077-Stein

March 11, 2007

39

8:17

CHAPTER 39

P-selectin and microparticles to predict deep venous thrombosis

A preliminary investigation of the predictive value of a combination of D-dimer, P-selectin, and microparticles showed that plasma markers have potential for diagnostic evaluation, but sensitivity and specificity need to be improved for clinical applicability [1]. Pselectin is present in the alpha granules of platelets and the Weibel-Palade bodies of endothelial cells [1]. P-selectin is the first upregulated glycoprotein on activated endothelial cells and platelets and has procoagulant properties [1]. Microparticles are fragments of cell membranes shed from platelets [2], leucocytes [3], and endothelial cells [4]. They are <1 μm in size and circulate within the vascular system [1]. Hematopoietic cells may also be a significant source of microparticles that participate in thrombosis via P-selectindependent mechanisms [5]. Microparticles are rich in tissue factor [6], phosphatidylserine [7], and the receptor for P-selectin, PSGL-1 [8]. Through interactions between P-selectin and its receptor, PSGL-1, microparticles are elaborated and then taken up into thrombi [1]. The presence of these microparticles has an important role in thrombus generation and its subsequent amplification [9, 10]. Combinations of P-selectin, microparticles, and Ddimer were evaluated among 22 patients with deep venous thrombosis (DVT) and 21 symptomatic patients in whom DVT was excluded [1]. Threshold values were soluble P-selectin 0.68 ng/mg of total protein, total microparticles 125% compared to controls, and D-dimer 3 mg/L by latex turbidimetric assay. Sensitivity of P-selectin alone was 68%, total microparticles alone 50%, and D-dimer 64%. Respective specificities were 81, 67, and 76%. Combined application of these variables gave a sensitivity of 73% and specificity of 81%.

References 1 Rectenwald JE, Myers DD, Jr, Hawley AE et al. D-dimer, P-selectin, and microparticles: novel markers to predict deep venous thrombosis. A pilot study. Thrombo Haemost 2005; 94: 1312–1317. 2 Gilbert GE, Sims PJ, Wiedmer T, Furie B, Furie BC, Shattil SJ. Platelet-derived microparticles express high affinity receptors for factor VIII. J Biol Chem 1991; 266: 17261– 17268. 3 Mesri M, Altieri DC. Endothelial cell activation by leukocyte microparticles. J Immunol 1998; 161: 4382–4387. 4 Sabatier F, Roux V, Anfosso F, Camoin L, Sampol J, Dignat-George F. Interaction of endothelial microparticles with monocytic cells in vitro induces tissue factordependent procoagulant activity. Blood 2002; 99: 3962– 3970. 5 Chou J, Mackman N, Merrill-Skoloff G, Pedersen B, Furie BC, Furie B. Hematopoietic cell-derived microparticle tissue factor contributes to fibrin formation during thrombus propagation. Blood 2004; 104: 3190–3197. 6 Falati S, Liu Q, Gross P et al. Accumulation of tissue factor into developing thrombi in vivo is dependent upon microparticle P-selectin glycoprotein ligand 1 and platelet P-selectin. J Exp Med 2003; 197: 1585–1598. 7 Martinez MC, Tesse A, Zobairi F, Andriantsitohaina R. Shed membrane microparticles from circulating and vascular cells in regulating vascular function. Am J Physiol Heart Circ Physiol 2005; 288: H1004–H1009. 8 Furie B, Furie BC. Role of platelet P-selectin and microparticle PSGL-1 in thrombus formation. Trends Mol Med 2004; 10: 171–178. 9 Myers DD, Hawley AE, Farris DM et al. P-selectin and leukocyte microparticles are associated with venous thrombogenesis. J Vasc Surg 2003; 38: 1075–1089. 10 Myers D, Jr, Farris D, Hawley A et al. Selectins influence thrombosis in a mouse model of experimental deep venous thrombosis. J Surg Res 2002; 108: 212–221.

179


BLUK077-Stein

April 6, 2007

16:0

III Diagnosis of acute PART III

pulmonary embolism

181


BLUK077-Stein

April 6, 2007

40

16:0

CHAPTER 40

Clinical characteristics of patients with no prior cardiopulmonary disease

Introduction The clinical manifestations of pulmonary embolism (PE) are recognized to be nonspecific. When considered together, frequent signs and symptoms, and combinations of signs and symptoms, particularly in patients with predisposing factors for PE, can strongly suggest the need for further diagnostic evaluation [1, 2]. The most extensive databases on the clinical characteristics of acute PE are from the Urokinase Pulmonary Embolism Trial (UPET) [3], PIOPED I (Prospective Investigation of Pulmonary Embolism Diagnosis I) [4], and more recently, PIOPED II [5]. It is clear, however, that the clinical characteristics reported among patients in PIOPED I, as in PIOPED II and other investigations of patients enrolled in clinical trials, are characteristics of patients whose findings were sufficient to alert physicians to the diagnosis and characteristics of patients who met the inclusion criteria. Patients who were too ill to participate, or who died suddenly, or were unidentified because the clinical findings were mild or atypical would not be included. Patients in whom PE was suspected but excluded would also be likely to have clinical findings similar to patients with PE, because such clinical findings are recognized by physicians as those associated with PE, and for this reason patients were evaluated. In spite of these constraints, recognition of the clinical characteristics is vital in the initial identification of patients who may have PE. Critical to identification of patients with PE is a high level of suspicion, particularly in patients with risk factors for PE, which would then trigger further evaluation. In order to characterize the diagnostic features of the history and physical examination in patients with acute PE, patients with no preexisting cardiac or pulmonary disease were evaluated. In such patients, the features

of acute PE become clear and manifestations related to coexistent disease are excluded. In PIOPED I [4] there were 117 patients with PE and no prior cardiopulmonary disease. All were diagnosed by pulmonary angiography or autopsy. PE was excluded in 248 such patients. It was excluded by a normal pulmonary angiogram in 176 patients and by the absence of adverse events over the course of 1 year in 72 patients who did not receive anticoagulant therapy [6]. In PIOPED II, there were 133 patients with PE and no prior cardiopulmonary disease and PE was excluded in 366 such patients. The diagnosis and exclusion of PE was on the basis of a composite set of reference tests [7] (Chapter 76). Patients in PIOPED I were considered to have no preexisting cardiac or pulmonary disease if they had no history or evidence of valvular heart disease, coronary artery disease, myocardial infarction, “other heart disease,” or if prior to this acute episode, they had not suffered left- or right-sided congestive heart failure. Such patients had no history of asthma, chronic obstructive pulmonary disease, interstitial lung disease, “other lung disease,” and no recognized acute pneumonia or acute respiratory distress syndrome at the time of the evaluation for suspected PE. These patients also had no previous history of PE. In PIOPED II, the definitions of “no prior cardiopulmonary disease” differed somewhat because the information collected in the two investigations was not exactly the same. In PIOPED II, patients were defined as having no prior cardiopulmonary disease if they had no current asthma or pneumonia, no history of chronic bronchitis, emphysema, chronic obstructive pulmonary disease, lung cancer, or prior PE, and no current or history of rightor left-sided heart failure. Patients in PIOPED I with PE and no prior cardiopulmonary disease were 54 ± 17 years of age 183


BLUK077-Stein

April 6, 2007

16:0

184

(mean ± standard deviation) and 49% were males [4]. Patients in PIOPED II with PE were 56 ± 16 years of age. In PIOPED II, 38% were males [5].

Predisposing factors Immobilization, irrespective of the cause, was shown to be the most prevalent predisposing factor. Data from PIOPED I [4] indicate, as others have shown [8], that immobilization of even 1 or 2 days may predispose to PE. Among 117 patients with PE who had no preexisting cardiac or pulmonary disease, immobilization usually due to surgery occurred in 56% of the patients (Table 40.1) [4]. Among the 66 patients in whom immobilization was a predisposing factor, 43 (65%) were immobilized for ≤2 weeks. Immobilization of 2 days or less preceded PE in 8 of 117 (7%). One or more predisposing factors were present in 96 of 117 (82%) patients with PE. In PIOPED II, immobilization was not as frequent a predisposing factor as in PIOPED I (Table 40.1) [4, 5]. Recent travel ≥4 hours was reported in 14% of the patients with PE. Almost half of the patients had smoked at one time. One or more risk

PART III

Diagnosis of acute PE

factors were reported in 92% of the patients with PE in PIOPED II. Subsequent to PIOPED II, we evaluated the incidence of venous thromboembolism in patients with the nephrotic syndrome using data from the National Hospital Discharge Survey (Kayali F, Najjar R, Stein PD. Unpublished data). The rate of diagnosis of deep venous thrombosis (DVT) in patients with the nephrotic syndrome was 1.5%. The relative risk, comparing with patients who did not have the nephrotic syndrome was 1.67. Among patients 18–39 years old, the rate of DVT with the nephrotic syndrome was 2.8% (relative risk 7.0). There were too few patients with renal vein thrombosis to calculate its rate of occurrence. The data show that the nephrotic syndrome is a risk factor for DVT. Pulmonary embolism, if it occurs, is likely to be due to DVT and not renal vein thrombosis, because of the infrequent occurrence of the latter.

Syndromes of acute PE Consideration of the diagnosis of PE in terms of its presenting syndrome may focus one’s thinking about the diverse manifestations of PE. The syndrome of

Table 40.1 Predisposing factors in patients with pulmonary embolism and no previous cardiac or pulmonary disease. n (%) Risk factors

PIOPED I ( N = 117)

PIOPED II ( N = 131–133)

Immobilization

66 (56)*

27 (20)†

Surgery (≤3 mo)

63 (54)

30 (23)‡

Malignancy (excluding lung cancer)*

27 (23)

29 (22)

Thrombophlebitis, ever

16 (14)

11 (8)

12 (10)

14 (11)

Travel ≥4 h in last month

19 (14)

Trauma (≤3 mo) Lower extremities Other

4 (3)

Smoke (ever)

55 (42)

<1 pack/day

29 (22)

1–2 packs/day

20 (15)

≥2 packs/day

0 (0)

Central venous instrumentation (≤3 mo)

12 (9)

Stroke, paresis, or paralysis

8 (7)

Postpartum (≤3 mo)

5 (4)

4 (3)

*Actively treated in last 3 months. † Within last month. ‡ Among 30 patients with surgery as a risk factor, the prevalence of heart, abdominal, pelvic, hip/knee – open, hip/knee – replacement, and neurosurgery ranged from 3 to 5.  Includes pelvis. Data from Stein et al. [4, 5].


BLUK077-Stein

April 6, 2007

CHAPTER 40

16:0

185

Patients with no prior cardiopulmonary disease

Table 40.2 Syndromes of acute pulmonary embolism. n (%) Syndrome

PIOPED I ( N = 117)

PIOPED II ( N = 133)

Pleuritic pain or hemoptysis

76 (65)

55 (41)

Uncomplicated dyspnea

26 (22)

48 (36)

Circulatory collapse*

9 (8)

11 (8)

Different Presentation

6 (5)

19 (14)

*Systolic blood pressure <80 mm Hg or loss of consciousness. Data from Stein et al. [4, 5].

pleuritic pain or hemoptysis, in the absence of circulatory collapse, was the most frequent mode of presentation occurring in 65% of the patients with PE and no prior cardiopulmonary disease in PIOPED I and in 41% in PIOPED II (Table 40.2) [4, 5]. Circulatory collapse (systolic blood pressure <80 mm Hg or loss of consciousness) was an uncommon mode of presentation in these investigations occurring in 8% of the patients with PE and no prior cardiopulmonary disease in PIOPED I and PIOPED II. A syndrome of dyspnea in the absence of hemoptysis or pleuritic pain or circulatory collapse occurred in 22% in PIOPED I and in 36% in PIOPED II. Many of the remaining patients in PIOPED II were identified on the basis of testing in patients with deep venous thrombosis (DVT), or other findings in the absence of any of these syndromes. The syndrome of pleuritic pain or hemoptysis may be more frequent in PIOPED I and PIOPED II than the other syndromes of PE because such patients usually have less extensive PE than patients with other syndromes of PE [1]. Such patients, therefore, may have been more likely to be recruited.

Symptoms of acute PE Among patients with PE and no prior cardiac or pulmonary disease, dyspnea was the most common symptom, occurring in 73% in PIOPED I and 73% in PIOPED II (Table 40.3) [4, 5]. Among patients with PE and dyspnea, 21 of 97 (22%) had dyspnea only on exertion, whereas 73 of 97 (75%) had dyspnea at rest [5]. In a few patients, the information was not known (Table 40.3). Orthopnea was shown to be a symptom of PE. Orthopnea occurred in 37 of 73 (51%) of the patients with PE who had dyspnea at rest and it occurred in 7 of 21 (33%) with dyspnea only on exertion (information not known in 2 patients with orthopnea) [5]. The rate of onset of dyspnea was within seconds

in 46% of patients, within minutes in 26%, and within hours in 11% (5). In some patients (17%), the rate of onset was within days (5). In both PIOPED I and PIOPED II, pleuritic chest pain was more frequent in patients with PE and no prior cardiopulmonary disease than was hemoptysis (Table 40.3). Cough was common among patients with PE and no prior cardiopulmonary disease in PIOPED I and PIOPED II, occurring in 37 and 34% [4, 5]. It was usually nonproductive (Table 40.3). In those in whom the cough was productive, similar proportions of patients had clear, purulent, or bloody sputum (Table 40.3). In small number of patients, blood-streaked sputum (4 patients) was more frequent than all blood (1 patient), and the amount of blood in all instances was <1 teaspoonful (Table 40.3). Calf or thigh pain occurred in 44% of the patients with PE and no prior cardiopulmonary disease in PIOPED II (Table 40.3) Thigh pain alone was uncommon, 3% [5]. The relation of the sensitivity of individual symptoms (percent of patients with PE who had the particular symptoms) to the percent of false positives among patients in PIOPED I who did not have PE is shown in Figure 40.1.

Signs of acute PE Tachypnea (respiratory rate â&#x2030;Ľ 20/min) was the most common sign of acute PE among patients with no prior cardiac or pulmonary disease, occurring in 70% in PIOPED I and 54% in PIOPED II (Table 40.4) [4, 5]. Tachycardia (heart rate > 100/min) occurred in 30 and 24% of the patients with PE. One of the signs of right atrial, right ventricular, or pulmonary artery pressure elevation occurred in 21% of the patients (Table 40.4).


BLUK077-Stein

April 6, 2007

16:0

Table 40.3 Symptoms in patients with pulmonary embolism and no preexisting cardiac or pulmonary disease. PE n (%) Symptoms

PIOPED I ( N = 117)

PIOPED II ( N = 127–133)

85 (73)

97 (73)

Dyspnea Dyspnea (rest or exertion) Dyspnea (at rest)

73 (55)

Dyspnea (exertion only)

21 (16)

Orthopnea (≥2-pillow)

37 (28)

Pleuritic pain Chest pain (not pleuritic) Cough Hemoptysis

77 (66)

58 (44)

5 (4)

25 (19)

43 (37)

45 (34)

15 (13)

7 (5)*

Purulent

7 (5)

Clear

7 (5)

Nonproductive

26 (20)

Wheezing

10 (9)

Palpitations

12(10)

27 (21)

Calf swelling

51 (40)

Thigh swelling

10 (8)

Thigh swelling, no calf swelling

1 (1) 30 (26)†

Calf or thigh pain

56 (44)

Calf pain only

30 (23)

Calf and thigh pain

22 (17)

Thigh pain only

4 (3)

*Hemoptysis; patients with PE: 2 = slightly pinkish, 4 = blood-streaked, 1 = all blood (<1 teaspoonful). † “Leg pain.” PE, pulmonary embolism. Data from Stein et al. [4, 5].

100

80

Sensitivity (%)

1 2 60

40

1. Dyspnea 2. Pleuritic pain 3. Cough 4. Leg swelling 5. Leg pain 6. Hemoptysis 7. Palpitations 8. Wheezing 9. Anginal pain

3 4 5

20 6 9

8

7

0 0

20

40 60 False positives (%)

Figure 40.1 Sensitivity and frequency of false-positive symptoms in patients with suspected acute pulmonary embolism (PE) who had no prior cardiopulmonary disease

80

(CPD) (n = 365). The numbers refer to the individual symptoms shown in the insert. (Data from Stein et al. [4].)

186

100


BLUK077-Stein

April 6, 2007

CHAPTER 40

16:0

187

Patients with no prior cardiopulmonary disease

Table 40.4 Signs in patients with pulmonary disease and no preexisting cardiac or pulmonary disease. PE n (%) Signs

PIOPED I ( N = 117)

PIOPED II ( N = 128–132)

General Tachypnea (≥20/min)

82 (70)

71 (54)

Tachycardia (>100/min)

35 (30)

32 (24)

Diaphoresis

13 (11)

3 (2)

Cyanosis

1 (1)

0 (0)

Temperature >38.5◦ C (>101.3◦ F)

8 (7)

1 (1)

Cardiac examination (any) Increased P2 Third heart sound

28 (21) 27 (23)

Fourth heart sound

28 (24)

Right ventricular lift

5 (4)

4 (4)∗∗

Jugular venous distension

18 (14)

Lung examination (any) Rales (crackles) Wheezes

38 (29) 60 (51)

23 (18)

6 (5)

2 (2)

Rhonchi

2 (2)

Decreased breath sounds Pleural friction rub

15 (15)∗

3 (3)

22 (17) 3 (3)

0 (0)

13 (11)

62 (47)*

DVT Calf or thigh Calf only

42 (32)

Calf and thigh

18(14)

Thigh only Homans’ sign

2 (2) 5 (4)

Number of patients with PE who had one or more signs of DVT: edema = 55, erythema = 5, tenderness = 32, palpable cord = 2. PE, pulmonary embolism; P2, pulmonary component of second sound; DVT, deep venous thrombosis. Data from Stein et al. [4, 5]. *Data in 103 pts. **Data in 110 pts.

Examination of the lungs showed an abnormality in 29% of the patients with PE and no prior cardiopulmonary disease (Table 40.4). Rales (crackles) and decreased breath sounds were the most frequently detected abnormalities. Wheezes and rhonchi were heard occasionally. A pleural friction rub was rare in PIOPED I and was not observed in PIOPED II. In PIOPED I, most patients with PE who had rales (88%) had pulmonary parenchymal abnormalities, atelectasis, or a pleural effusion on the chest radiograph. Signs of DVT were observed in the calf or thigh in 47% of the patients with PE and no prior cardiopulmonary disease (Table 40.4). Both the thigh and calf showed signs of DVT in 14%, but the thigh alone uncommonly showed signs of DVT in 2%.

The relation of the sensitivity of individual signs (percent of patients with PE who had the particular signs) to the percent of false positives among patients in PIOPED I who did not have PE is shown in Figure 40.2.

Combinations of signs and symptoms Combinations of clinical characteristics in patients with PE and no prior cardiopulmonary disease are shown in Table 40.5. Dyspnea or tachypnea was shown in 90 and 84% of such patients in PIOPED I and PIOPED II [4, 5]. Dyspnea or tachypnea or pleuritic pain was shown in 97 and 92%. One of these findings or signs of DVT was found in 98% of such patients in


BLUK077-Stein

April 6, 2007

16:0

188

PART III

Diagnosis of acute PE

100

80 1. RR ≥ 20/min 2. Rales 3. HR > 100/min 4. S4 5. Increased P2 6. DVT 7. Diaphoresis 8. TEMP > 38.5 9. Wheezes 10. Homans' 11. RV Lift 12. Pleural rub 13. S3 14. Cyanosis

Sensitivity (%)

1 60 2 40 3 4

5

20

6

7 9

0

8

0

20

40 60 False positives (%)

80

100

Figure 40.2 Sensitivity and frequency of false-positive signs in patients with suspected acute pulmonary embolism (PE) who had no prior cardiopulmonary disease

(CPD) (n = 365). The numbers refer to the individual signs shown in the insert. (Data from Stein et al. [4].)

PIOPED II. Conversely, in the absence of dyspnea or tachypnea, pleuritic pain, or signs of DVT, PE was infrequently diagnosed [4, 5]. Comparable observations among patients with no prior cardiopulmonary disease were also made in the

UPET [3]. Among 215 patients, either dyspnea or tachypnea (respiratory rate ≥20/min) occurred in 96% of the patients with no preexisting cardiac or pulmonary disease in UPET [3]. Either dyspnea, tachypnea, or signs of DVT occurred in 99% [3].

Table 40.5 Combinations of clinical characteristics in patients with pulmonary embolism and no prior cardiopulmonary disease. PE n (%) Signs

PIOPED I ( N = 117)

PIOPED II ( N = 131)

Dyspnea or tachypnea (≥20/min)

105 (90)

110 (84)

Dyspnea or tachypnea (≥20/min) or pleuritic pain

113 (97)

120 (92)

Dyspnea or tachypnea (≥20/min) or pleuritic pain

128 (98)

or signs of DVT Dyspnea or tachypnea (≥20/min) or signs of DVT

107 (91)

Dyspnea or tachypnea (≥20/min) or pleuritic pain

115 (98)

or radiographic evidence of atelectasis or a parenchymal abnormality PE, pulmonary embolism. Data from Stein et al. [4, 5].


BLUK077-Stein

April 6, 2007

CHAPTER 40

16:0

189

Patients with no prior cardiopulmonary disease

Table 40.6 Clinical assessment in patients with no preexisting cardiac or pulmonary disease. PE/n (%) Pulmonary embolism

Pulmonary embolism

Clinical assessment

(assessment by fellow)

(assessment by staff)

High probability

26/41 (63)

15/17 (88)

Intermediate probability

47/99 (47)

74/200 (37)

8/74 (11)

14/103 (14)

Low probability

Differences between assessments by fellows and staff were not significant. PE, pulmonary embolism. Data are from Stein et al. [4].

Clinical assessment Physicians, on the basis of clinical assessment, were more consistent in their ability to exclude PE than they were in their ability to make the diagnosis of PE (Table 40.6). In the majority of patients (46–63%), however, physicians were suspicious of the diagnosis of PE, but uncertain in their assessment. Senior staff, when they made a high-probability assessment, were correct in 88% and fellows were correct in 63% (Table 40.6) [4]. The clinical assessment of the likelihood of acute PE was based upon all available noninvasive data, not any specific predetermined criteria. In conclusion, symptoms may be mild and traditionally important symptoms may be absent. There may be dyspnea only on exertion. Orthopnea may be a symptom of PE.

References 1 Stein PD, Willis PW, III, DeMets DL. History and physical examination in acute pulmonary embolism in patients without pre-existing cardiac or pulmonary disease. Am J Cardiol 1981; 47: 218–223.

2 Wenger NK, Stein PD, Willis PW, III. Massive acute pulmonary embolism: the deceivingly nonspecific manifestations. JAMA 1972; 220: 843–844. 3 A National Cooperative Study. Clinical and electrocardiographic observations (Chapter X of The Urokinase Pulmonary Embolism Trial). Circulation 1973; 47(suppl II): 60–65. 4 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 5 Stein PD, Beemath A, Matta F et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med, 2007 (In press). 6 The PIOPED Investigators. Value of the ventilation/ perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 7 Stein PD. Fowler SE, Goodman LR et al., for the PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis II (PIOPED II). N Engl J Med 2006; 354: 2317–2327. 8 Coon WW, Coller FA. Some epidemiologic considerations of thromboembolism. Surg Gynecol Obstet 1959; 109: 487– 501.


BLUK077-Stein

March 12, 2007

41

16:18

CHAPTER 41

Relation of right-sided pressures to clinical characteristics of patients with no prior cardiopulmonary disease

The pulmonary artery mean pressure in patients with pulmonary embolism (PE) and no prior cardiopulmonary disease in Prospective Investigating Pulmonary Embolism Diagnosis I (PIOPED I) did not exceed the upper limits of normal (20 mm Hg) in 58% of the 111 randomized patients (Figure 41.1) (P.D. Stein and B. Relyea unpublished data from PIOPED I, 1990). Pulmonary artery mean pressure was 30 mm Hg or less in 85% of these patients. Right atrial mean pressure was less than the upper limits of normal (8 mm Hg) in 75% of the 109 patients with PE and no prior cardiopulmonary disease (Figure 41.2). Right atrial mean pressure was rarely elevated in patients in whom the pulmonary artery mean pressure was normal. Even among patients in whom pulmonary artery mean pressure exceeded 20 mm Hg, there was a poor correlation with right atrial mean pressure (r = 0.11). Right-sided pressures were not routinely obtained in PIOPED II. The relation of signs and symptoms to pulmonary artery mean pressure, right atrial mean pressure,

and the partial pressure of oxygen in arterial blood (Pa O2 ) while breathing room air is shown in Table 41.1. The pulmonary artery mean pressure was lowest in patients with pleuritic chest pain and highest in those with a third heart sound, right ventricular lift, diaphoresis, and palpitations. A few statistically significant differences were shown at the P < 0.01 level. The pulmonary artery mean pressure was higher in patients with a third heart sound than in patients with the following signs and symptoms who did not have a third heart sound: pleural friction rub, accentuated pulmonary component of the second sound, tachycardia, tachypnea, wheezing, hemoptysis, cough, pleuritic pain, and dyspnea. Patients with pleuritic pain had a lower pulmonary artery mean pressure than patients with palpitations and patients with tachypnea who did not have pleuritic pain. There was a poor correlation between pulmonary artery mean pressure and the Pa O2 (Figure 41.3). 100

25

80

n = 111 20

60 15

40 10

20

5

0

4

8

12

16

20

24

28

32

36

40

48

52

PA mean pressure (mm Hg)

190

0

Cumulative frequency (%)

Number of patients

30

Figure 41.1 Distribution and cumulative frequency of pulmonary artery mean pressure among randomized patients with pulmonary embolism and no prior cardiopulmonary disease (P.D. Stein and B. Relyea, unpublished data from PIOPED I, 1990).


BLUK077-Stein

March 12, 2007

16:18

100

Figure 41.2 Distribution and cumulative frequency of right atrial mean pressure among randomized patients with pulmonary embolism and no prior cardiopulmonary disease (P.D. Stein and B. Relyea, unpublished data from PIOPED I, 1990).

30

n = 109

25

80

60

20 15

40

10 20 5

Cumulative frequency (%)

Number of patients

35

0

0 2

4

6

8

10 12 14 16 18

20 22 24 26

RA mean pressure (mm Hg)

Table 41.1 Relation between signs, symptoms, right-sided pressures, and Pa O2 . Signs/symptoms

PA mean (mm Hg)

RA mean (mm Hg)

Pa O2 (mm Hg)

Dyspnea

23 ± 9 (n = 121)

6 ±5 (n = 114)

69 ± 14 (n = 98)

Pleuritic pain

21 ± 8 (n = 115)

5 ± 4 (n = 114)

73 ± 14 (n = 97)

Cough

24 ± 9 (n = 68)

6 ± 5 (n = 63)

71 ± 13 (n = 54)

Hemoptysis

22 ± 8 (n = 24)

5 ± 5 (n = 24)

74 ± 14 (n = 23)

Palpitations

27 ± 10 (n = 18)

6 ± 6 (n = 15)

64 ± 15 (n = 15)

Wheezing

21 ± 7 (n = 16)

4 ± 3 (n = 14)

70 ± 15 (n = 13)

Angina-like pain

25 ± 14 (n = 9)

3 ± 3 (n = 6)

62 ± 9 (n = 8)

Tachypnea (≥20/min)

22 ± 9 (n = 118)

6 ± 5 (n = 112)

69 ± 15 (n = 95)

Rales (crackles)

22 ± 8 (n = 81)

6 ± 4 (n = 78)

68 ± 14 (n = 68)

Tachycardia (>100/min)

24 ± 9 (n = 59)

7 ± 5 (n = 53)

69 ± 17 (n = 42)

Increased P2

23 ± 9 (n = 42)

6 ± 5 (n = 40)

67 ± 14 (n = 33)

Diaphoresis

26 ± 11 (n = 16)

8 ± 7 (n = 14)

61 ± 15 (n = 12)

Right ventricular lift

28 ± 12 (n = 3)

8 ± 9 (n = 3)

67 ± 23 (n = 4)

Friction rub

21 ± 8 (n = 6)

9 ± 6 (n = 6)

67 ± 4 (n = 5)

Third heart sound

39 ± 2 (n = 3)

12 ± 8 (n = 4)

71 ± 9 (n = 3)

Values are mean ± standard deviation. PA, pulmonary artery; RA, right atrium; RV, right ventricle; Pa O2 , partial pressure of oxygen in arterial blood in patients breathing room air. P.D. Stein and B. Relyea, unpublished data from PIOPED I, 1990).

Figure 41.3 Relation of the partial pressure of oxygen in arterial blood (Pa O2 ) while breathing room air to the pulmonary artery (PA) mean pressure in patients with no prior cardiopulmonary disease. Both randomized and referred patients are included (P.D. Stein and B. Relyea, unpublished data from PIOPED I, 1990).

191


BLUK077-Stein

March 28, 2007

42

15:25

CHAPTER 42

The history and physical examination in all patients irrespective of prior cardiopulmonary disease

Introduction In this chapter, the clinical characteristics of all patients with acute pulmonary embolism (PE) are described. Evaluation of the clinical characteristics of acute PE in patients with no prior cardiopulmonary disease eliminates patients with PE whose clinical findings may be due to underlying disease and was described in Chapter 40. Data from Prospective Investigation of Pulmonary Embolism Diagnosis I (PIOPED I) include patients who were randomized for pulmonary angiography as described in the primary PIOPED report [1] and patients who were referred for angiography and not described in the original PIOPED report. The patients with acute PE in PIOPED I were 58 ± 17 years of age (mean ± standard deviation) and 54% were males [2]. In PIOPED II, patients with PE were 57 ± 17 years of age and 42% were males [3].

Predisposing factors Predisposing factors are shown in Table 42.1. In PIOPED II, 94% had one or more risk factors. Immobilization was the most frequent predisposing factor, and surgery was the usual cause of immobilization. In PIOPED I and PIOPED II, 19 and 10% of the patients, respectively, had a history of ever having had deep venous thrombosis (DVT). Six percent and 4% had prior PE in PIOPED I and PIOPED II (Table 42.1).

Syndromes of acute PE The syndrome of Hemoptysis or pleuritic pain, in the absence of circulatory collapse, and the syndrome of

192

uncomplicated dyspnea were frequent modes of presentation in patients with acute PE in PIOPED II (Table 42.2) [3]. Circulatory collapse (systolic blood pressure <80 mm Hg or loss of consciousness) was an uncommon mode of presentation, possibly because such patients were not recruited.

Symptoms of acute PE Dyspnea was the most frequent symptom, occurring in 78 and 79% of patients with PE in PIOPED I and PIOPED II (Table 42.3) [2, 3]. Dyspnea was usually present at rest (61% of the patients with PE), but in some (16% with PE) it was present only with exertion (Table 42.3) [3]. Orthopnea was reported in 36% of the patients with PE [3]. The onset of dyspnea was within seconds in 41%, minutes in 26%, hours in 14%, and over a period of days in 19% [3] Pleuritic chest pain occurred four to eight times more frequently than hemoptysis in patients with PE. Pure bloody sputum was observed in only 2 patients, and it was <1 teaspoonful. Cough was common and occurred in 43% of the patients with PE in both PIOPED I and PIOPED II (Table 42.3).

Signs of acute PE Tachypnea (respiratory rate ≥20/min) was the most common sign of acute PE and was observed in 73% of the patients with PE in PIOPED I and 57% in PIOPED II (Table 42.4) [2, 3]. Rales and decreased breath sounds were the most frequent findings on examination of the lungs, but rhonchi and wheezes were also observed. Rarely, signs of DVT were found only in the


BLUK077-Stein

March 28, 2007

CHAPTER 42

15:25

193

All patients irrespective of prior cardiopulmonary disease

Table 42.1 Predisposing factors in all patients with pulmonary embolism. n (%) PIOPED I (N = 383)

Risk factors Immobilization

206 (54)

Travel ≥4 h in last month Surgery (≤3 mo)

PIOPED II (N = 185 – 192)* 48 (25)‡ 23 (12)

160 (42)

Coronary heart disease

76 (20)

Myocardial infarction

48 (13)

Heart failure

45 (12)

Collagen vascular disease

15 (4)

41 (21)§

10 (5)

Malignancy†

69 (18)

Stroke, paresis, or paralysis

37 (10)

37 (19)¶ 7 (4)

Thrombophlebitis, ever

71 (19)

19 (10)

Prior pulmonary embolism

23 (6)

7 (4)

47 (12)

16 (8)||

27 (7)

22 (12)

Trauma (≤3 mo) Lower extremities Other

5 (3)

Asthma Pneumonia (current)

27 (7)

5 (3)

Chronic obstructive pulmonary disease

37 (10)

10 (5)

Emphysema

7 (4)

Interstitial lung disease

6 (2)

Lung cancer

5 (3)

Estrogen

22 (6)

Males, therapeutic

1 (1)

Smoke (ever)

90 (47)

<1 pack/day

43 (22)

1–2 packs/day

37 (19)

>2 packs/day

1 (1)

Central venous instrumentation (≤3 mo) Postpartum (≤3 mo)

22 (12) 9 (2)

Sepsis (current)

0 (0)

* Nominal value. If value was not reported, it was assumed to be absent. †

Actively treated in last 3 months. Within last month. § Among 41 patients with surgery as a risk factor, abdominal surgery was in 9, and heart, pelvic, hip/knee – open, hip/knee – replacement, and neurosurgery ranged from 3 to 5. ¶ Excluding lung cancer. || Includes pelvis. Data from PIOPED I (unpublished) and Stein et al.[3]. ‡

Table 42.2 Syndromes of acute pulmonary embolism. Syndrome

n (%) in PIOPED II (N = 192)

Hemoptysis or pleuritic pain

84 (44)

Uncomplicated dyspnea Circulatory collapse*

70 (36) 15 (8)

* Systolic blood pressure <80 mm Hg or loss of consciousness. Data from Stein et al. [3].


BLUK077-Stein

March 28, 2007

15:25

194

PART III

Diagnosis of acute PE

Table 42.3 Symptoms in all patients with pulmonary embolism. PE n (%) Symptoms

PIOPED I (N = 383)

PIOPED II (N = 184 – 192)

Dyspnea Dyspnea (rest or exertion)

299 (78)

Dyspnea (at rest) Dyspnea (exertion only)

31 (16)

Orthopnea (≥2 pillows) Pleuritic pain Chest pain (not pleuritic) Cough Hemoptysis

151 (79) 117 (61) 69 (36)

225 (59)

89 (47)

23 (6)

33 (17)

166 (43)

82 (43)

62 (16)

11 (6)*

Purulent

18 (10)

Clear

18 (10)

Nonproductive Wheezing

40 (21) 54 (14)

Palpitations

48 (13)

Calf swelling

118 (31)†

58 (31) 54 (29)‡

Calf or thigh swelling

72 (39)

Calf and thigh swelling

15 (8)

Thigh swelling, no calf swelling Calf pain

3 (2) 103 (27)

43 (23)‡

Calf or thigh pain

78 (42)

Calf and thigh pain

30 (16)

Thigh pain, no calf pain

5 (3)

* Hemoptysis; patients with PE: 3 = slightly pinkish; 6 = blood-streaked; 2 = all blood (<1 teaspoonful). † “Leg.” ‡ Calf only. PE, pulmonary embolism. Data from Stein et al. [2, 3].

thighs. Tachycardia (heart rate >100/min) was present in only 30 and 26% of the patients with PE in PIOPED I and PIOPED II (Table 42.4). In the Urokinase Pulmonary Embolism Trial, among all patients with PE irrespective of the presence or absence of prior cardiopulmonary disease, syncope, accentuated pulmonary component of the second, third, or fourth heart sound (S3 or S4 ), and cyanosis were more frequent among patients with massive PE than submassive PE [4]. Pleuritic pain was less frequent among patients with massive PE than submassive PE.

Combinations of signs and symptoms Among all patients with acute PE , dyspnea or tachypnea (respiratory rate ≥20/min) was present in 91 and

86% of the patients in PIOPED I and PIOPED II (Table 42.5) [2, 3]. Dyspnea or tachypnea or pleuritic pain was present in 97 and 92% [2, 3]. Dyspnea or tachypnea or pleuritic pain or signs of DVT were observed in 97 in both PIOPED I and PIOPED II [2, 3] In PIOPED I, all patients with acute PE had dyspnea or tachypnea or pleuritic pain or unexplained radiographic evidence of atelectasis or a parenchymal abnormality or an unexplained low Pa O2 [2].

Dyspnea, hypoxia, and normal chest radiograph Dyspnea and a low partial pressure of oxygen in arterial blood (Pa O2 ≤70 mm Hg) in a patient with a normal chest radiograph suggested the diagnosis of PE and was seen in 9 of 17 (53%) with PE in PIOPED I and


BLUK077-Stein

March 28, 2007

CHAPTER 42

15:25

195

All patients irrespective of prior cardiopulmonary disease

Table 42.4 Signs in all patients with pulmonary embolism. PE n (%) PIOPED I (N = 383)

PIOPED II (N = 184 – 191)

Tachypnea (≥20/min)

278 (73)

108 (57)

Tachycardia (>100/min)

115 (30)

49 (26)

Signs General

Diaphoresis

38 (10)

8 (4)

Cyanosis

13 (3)

1 (1)

Temperature >38.5◦ C (>101.3◦ F)

27 (7)

3 (2)

Cardiac examination (any)

42 (22)

Increased P2

87 (23)

Third heart sound

21 (5)

22 (15)

Right ventricular lift

8 (5)

Jugular venous distension

25 (13)

Lung examination (any) Rales (crackles) Wheezes

70 (37) 210 (55)

40 (21)

44 (11)

6 (3)

Rhonchi

9 (5)

Decreased breath sounds Pleural friction rub

40 (21) 17 (4)

2 (1)

56 (15)

90 (47)

DVT Calf or thigh Calf only

65 (34)

Calf and thigh

23 (12)

Thigh only Homans’ sign

2 (1) 10 (3)

Number of patients with PE who had one or more signs of DVT: edema = 79, erythema = 7, tenderness = 36, palpable cord = 2. PE, pulmonary embolism; P2, pulmonary component of second sound; DVT, deep venous thrombosis. Data from Stein et al. [2, 3].

Table 42.5 Combinations of clinical characteristics in patients with pulmonary embolism. n (%) Signs

PIOPED I (N = 383)

PIOPED II (N = 188)

Dyspnea or tachypnea (≥20/min)

347 (91)

162 (86)

Dyspnea or tachypnea (≥20/min) or pleuritic pain

371 (97)

173 (92)

Dyspnea or tachypnea (≥20/min) or pleuritic pain or

373 (97)

182 (97)

signs of DVT Dyspnea or tachypnea (≥20/min) or pleuritic pain or signs of DVT or radiographic evidence of atelectasis or a parenchymal abnormality Data from Stein et al. [2, 3].

381 (99)


BLUK077-Stein

March 28, 2007

15:25

196

in 22 of 93 (24%) with no PE [5]. This combination draws attention to the likelihood of PE and suggests the necessity for further diagnostic studies. Neither this combination nor other combinations, however, were diagnostic. The signs and symptoms in patients with PE were not specific. The frequency of individual signs and symptoms in patients with PE in most instances did not differ statistically from signs and symptoms in patients in whom PE was suspected, but excluded. This is expected, because patients with suspected PE were identified for further diagnostic studies on the basis of signs and symptoms suggestive of PE. The nonspecific signs and symptoms as well as the nonspecific but important findings on the chest radiograph and electrocardiogram, particularly in a patient with risk factors for PE, and a necessary high level of awareness of the possibility of PE, may lead to further diagnostic tests and a correct diagnosis.

PART III

Diagnosis of acute PE

References 1 The PIOPED Investigators. Value of the ventilation/ perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 2 Stein PD, Saltzman HA, Weg JG. Clinical characteristics of patients with acute pulmonary embolism. Am J Cardiol 1991; 68: 1723–1724. 3 Stein PD, Beemath A, Matta F et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med, 2007 (In press). 4 The Urokinase Pulmonary Embolism Trial: A National Cooperative Study. Associated laboratory and clinical findings. Circulation 1973; 47/48(suppl II): II81– II85. 5 Stein PD, Alavi A, Gottschalk A et al. Usefulness of noninvasive diagnostic tools for diagnosis of acute pulmonary embolism in patients with a normal chest radiograph. Am J Cardiol 1991; 67: 1117–1120.


BLUK077-Stein

March 28, 2007

43

15:25

CHAPTER 43

Clinical characteristics of patients with acute pulmonary embolism stratified according to their presenting syndromes

Introduction Many of the findings in the various syndromes of pulmonary embolism (PE) can be understood in terms of the severity of PE as it increases from mild with the pulmonary infarction syndrome to moderate with the isolated dyspnea syndrome to severe with circulatory collapse. Patients with submassive PE more commonly have the pulmonary infarction syndrome than patients with massive PE [1]. Patients with pulmonary infarction were shown to have less severe PE than patients with isolated dyspnea based on an objective pulmonary angiography scoring system [2]. Patients with circulatory collapse had the most severe PE based on the angiographic score, but the score was not statistically significantly higher than in patients with isolated dyspnea [2]. The prevalence of various clinical and laboratory characteristics of patients with the syndrome of pulmonary infarction, isolated dyspnea, or circulatory collapse may give clues to the diagnosis or suggest characteristics which may reduce the likelihood of inadvertently discarding the diagnosis of PE. Among patients with PE and no prior cardiopulmonary disease who survived long enough to undergo diagnostic evaluation, 65% in PIOPED I (Prospective Investigation of Pulmonary Embolism Diagnosis I) and 41% in PIOPED II presented with the syndrome of pulmonary hemorrhage or infarction characterized by pleuritic pain or hemoptysis [3, 4]. The syndrome of isolated dyspnea in the absence of circulatory collapse, pleuritic pain, or hemoptysis occurred in 22 and 36% in PIOPED I and PIOPED II [3, 4]. Circulatory collapse among patients who survived long enough for diagnostic evaluation was observed in only 8% with

PE and no prior cardiopulmonary disease in both PIOPED I and PIOPED II. However, approximately one third of patients with acute PE die within 2.5 hours [5]. The diagnosis of PE is occasionally made in asymptomatic patients with deep venous thrombosis (DVT) [3, 4] or, particularly in elderly patients, the diagnosis is suggested on the basis of unexplained radiographic abnormalities in asymptomatic patients [6]. The mean ages of patients in PIOPED I with pulmonary infarction syndrome, isolated dyspnea, and circulatory collapse were 51 ± 17 (mean ± SD), 55 ± 17, and 66 ± 18 years, respectively, which were not statistically significantly different [7] (Table 43.1).

Predisposing factors There were no statistically significant differences of predisposing factors between the three groups of presenting syndromes in PIOPED I [7] (Table 43.2). Absence of predisposing factors was observed in 13 and 15% of patients with pulmonary infarction or isolated dyspnea syndrome, respectively. Two of 5 patients with circulatory collapse had no predisposing factors. By definition none of these patients had prior cardiac or pulmonary disease, which, in some instances, may predispose them to PE.

Symptoms There were no statistically significant differences of symptoms between the groups of patients with PE and no prior cardiopulmonary disease in PIOPED I, excluding differences due to the defining characteristics [7] (Table 43.3). In patients with PE and no prior

197


BLUK077-Stein

March 28, 2007

15:25

Table 43.1 Age of patients in PIOPED I with pulmonary embolism and no prior cardiopulmonary disease. n (%)

Age (years) <21

Pulmonary infarction syndrome

Isolated dyspnea syndrome

(N = 119)

(N = 31)

1 (1)

1 (3)

21–30

20 (17)

3 (10)

31–40

15 (13)

3 (10)

41–50

16 (13)

4 (13)

51–60

27 (23)

6 (19)

61–70

27 (23)

7 (23)

71–80

9 (8)

7 (23)*

81–90

4 (3)

0 (0)

Patients with circulatory collapse ranged between age 31 and 90 years. * P < 0.02, pulmonary embolism vs. isolated dyspnea. Reprinted with permission from Stein and Henry [7]. Table 43.2 Predisposing factors for patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%) Pulmonary infarction syndrome

Isolated dyspnea syndrome

Predisposing factor

(N = 119)

(N = 31)

Immobilization

63 (53)

19 (61)

Surgery

61 (51)

14 (45)

Malignancy

23 (19)

4 (13)

Thrombophlebitis, ever

15 (13)

8 (26)

Trauma, lower extremity

22 (18)

2 (6)

Estrogen

11 (9)

4 (13)

Stroke

8 (7)

3 (10)

Postpartum ≤3 mo

5 (4)

2 (6)

None of the above

18 (15)

4 (13)

Only 5 patients had circulatory collapse. Two had no apparent predisposing factor. By definition, none of these patients had prior cardiopulmonary disease, which in some instances may predispose to PE. Reprinted with permission from Stein and Henry [7]. Table 43.3 Symptoms in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%)

Symptoms Dyspnea Pleuritic pain Cough

Pulmonary infarction

Isolated dyspnea

Circulatory collapse

syndrome (N = 119)

syndrome (N = 31)

syndrome (N = 5)

31 (100)* 0 (0)*

2 (40)

11 (35)

2 (40)

86 (72) 115 (97)* 52 (44)

0 (0)

Leg swelling

33 (28)

12 (39)

2 (40)

Leg pain

30 (25)

10 (32) 0 (0)*

2 (40)

5 (16)

0 (0)

Hemoptysis

25 (21)*

Palpitations

11 (9)

Wheezing

10 (8)

5 (16)

1 (20)

5 (4)

2 (6)

0 (0)

Angina-like pain

* Presence or absence of symptom defines syndrome. Reprinted with permission from Stein and Henry [7].

198

0 (0)


BLUK077-Stein

March 28, 2007

CHAPTER 43

15:25

199

Clinical characteristics according to presenting syndromes

cardiopulmonary disease in PIOPED I who had circulatory collapse, dyspnea was present in 2 of 5 (40%) [7], and in PIOPED II dyspnea was present in 9 of 11 (82%) [4]. Among all patients with PE and circulatory collapse in PIOPED II, dyspnea was present in 13 of 15 (87%) [4].

Signs Among patients with no prior cardiopulmonary disease in PIOPED I, rales were more prevalent in the pulmonary infarction group, 67 of 119 (56%), than in the isolated dyspnea group, 6 of 31 (19%) (P < 0.001) (Table 43.4). This was not, however, observed in PIOPED II in which 20% of the patients with the hemoptysis/pleuritic pain syndrome had rales compared with 26% in the isolated dyspnea syndrome [4]. Tachycardia in PIOPED I was more prevalent in patients with PE who had isolated dyspnea, 14 of 31 (45%), than in the pulmonary infarction group, 28 of 119 (24%) (P < 0.02). The frequency of tachycardia, however, was similar in both the groups in PIOPED II, 32 and 20%, respectively.

Combinations of signs and symptoms Among patients with pulmonary infarction syndrome and no prior cardiopulmonary disease in PIOPED I,

102 of 119 (86%) had dyspnea or tachypnea [7]. In PIOPED II, 41 of 51 (80%) had dyspnea or tachypnea [4]. Among patients with PE and circulatory collapse in PIOPED I and PIOPED II, dyspnea or tachypnea was not always present [4, 7]. Even dyspnea or tachypnea or pleuritic pain was not always present, but patients with circulatory collapse had one of these signs or symptoms or signs of DVT [4, 7].

Clinical assessment Physicians were confident of the diagnosis of PE (80– 100% clinical likelihood) in only 14–23% of the patients irrespective of the presenting syndrome. Physicians thought PE was probably absent (0–19% clinical likelihood) in 10% or fewer patients irrespective of the group.

Electrocardiographic manifestations A normal electrocardiogram (ECG) was more prevalent among patients with pulmonary infarction syndrome, 45 of 97 (46%), than among patients with isolated dyspnea syndrome, 2 of 21 (10%) (P < 0.01) (Table 43.5). Right bundle branch block was more

Table 43.4 Signs in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%)

Signs

Pulmonary infarction

Isolated dyspnea

Circulatory collapse

syndrome (N = 119)

syndrome (N = 31)

syndrome (N = 5)

Tachypnea (≥20/min)

84 (71)

20 (65)

2 (40)

Rales (crackles)

67 (56)*

6 (19)

1 (20)

Tachycardia (>100/min)

28 (24)†

14 (45)

3 (60)

Increased pulmonary component of sec-

32 (27)

10 (32)

0 (0)

ond sound Deep venous thrombosis

11 (9)

3 (10)

1 (20)

Diaphoresis

8 (7)

5 (16)

1 (20)

Temperature >38.5◦ C (>101.3◦ F)

7 (6)

2 (6)

0 (0)

Wheezes

8 (7)

2 (6)

0 (0)

Homans’ sign

2 (2)

2 (6)

1 (20)

Pleural friction rub

6 (5)

0 (0)

0 (0)

Third heart sound

2 (2)

2 (6)

0 (0)

Cyanosis

2 (2)

1 (3)

0 (0)

* P < 0.001, pulmonary infarction vs. isolated dyspnea. † P < 0.02, pulmonary infarction vs. isolated dyspnea. Reprinted with permission from Stein and Henry [7].


BLUK077-Stein

March 28, 2007

15:25

200

PART III

Diagnosis of acute PE

Table 43.5 Electrocardiographic manifestations in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%) Pulmonary infarction

Isolated dyspnea

Circulatory collapse

Manifestations

syndrome (N = 97)

syndrome (N = 21)

syndrome (N = 5)

Normal electrocardiogram

45 (46)*

2 (10)

1 (20)

ST-segment or T-wave changes

39 (40)

12 (57)

3 (60)

Left-axis deviation

13 (13)

3 (14)

0 (0)

Complete right bundle branch block

3 (3)

2 (10)

2 (40)†

Left ventricular hypertrophy

9 (9)

0 (0)

0 (0)

Incomplete right bundle branch block

4 (4)

2 (10)

0 (0)

Acute myocardial infarction pattern

2 (2)

1 (5)

0 (0)

Low-voltage QRS

3 (3)

1 (5)

1 (20)

P pulmonale

1 (1)

1 (5)

1 (20)

Right-axis deviation

1 (1)

0 (0)

0 (0)

Right ventricular hypertrophy

1 (1)

1 (1)

0 (0)

Some patients had more than one abnormality. * P < 0.01, pulmonary infarction vs. isolated dyspnea. † P < 0.001, pulmonary infarction vs. circulatory collapse. Reprinted with permission from Stein and Henry [7].

prevalent among patients with circulatory collapse than among patients with pulmonary infarction syndrome, 2 of 5 (40%) versus 3 of 97 (3%) (P < 0.001).

Plain chest radiograph The pulmonary infarction group tended to have a higher prevalence of atelectasis or pulmonary parenchymal abnormalities, 89 of 119 (75%), than the isolated dyspnea group, 16 of 31 (52%) (P < 0.02), and it had a higher prevalence than the circulatory collapse group, 1 of 5 (20%) (P < 0.01) (Table 43.6). The pulmonary infarction group also had a higher prevalence of pleural effusion, 67 of 119 (56%), than the isolated dyspnea group, 8 of 31 (26%) (P < 0.01), and it tended to have a higher prevalence than the circulatory collapse group, 0 of 5 (0%) (P < 0.02) (Table 43.6). The circulatory collapse group tended to have a higher prevalence of cardiomegaly, 2 of 5 (40%), than the pulmonary infarction group, 10 of 119 (8%) (P < 0.02) (Table 43.6).

creased and the Pa O2 decreased. The pulmonary infarction group had a higher mean Pa O2 (72.6 ± 14.2 mm Hg) than the isolated dyspnea group (63.2 ± 12.9 mm Hg) (P < 0.01) (Table 43.7). A Pa O2 greater than 80 mm Hg was observed in 27 of 99 (27%) patients with pulmonary infarction and 2 of 19 (11%) with isolated dyspnea. The pulmonary infarction group had a lower pulmonary artery mean pressure than the isolated dyspnea group (20.1 ± 8.1 mm Hg versus 24.5 ± 9.3 mm Hg (P < 0.02).

Ventilation–perfusion lung scan A high-probability ventilation–perfusion (V–Q) lung scan was more prevalent among the isolated dyspnea group than the pulmonary infarction group, 20 of 31 (65%) versus 38 of 119 (32%) (P < 0.001) (Table 43.8). Intermediate probability V–Q lung scans were most prevalent among patients with pulmonary infarction syndrome.

Pulmonary artery pressure, Pa O2 , Pa CO2 , and pH

Pulmonary infarction syndrome compared with isolated dyspnea syndrome

With increasing severity of PE, from pulmonary infarction to isolated dyspnea to circulatory collapse, trends suggest that the pulmonary artery mean pressure in-

Rales are more prevalent among patients with pulmonary infarction and tend to be less prevalent among patients with isolated dyspnea or circulatory collapse


BLUK077-Stein

March 28, 2007

CHAPTER 43

15:25

201

Clinical characteristics according to presenting syndromes

Table 43.6 Plain chest radiograph findings in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%) Pulmonary infarction

Isolated dyspnea

Circulatory collapse

Findings

syndrome (N = 119)

syndrome (N = 31)

syndrome (N = 5)

Atelectasis or pulmonary

89 (75)*

16 (52)

1 (20)†

Pleural effusion

67 (56)‡

8 (26)

0 (0)§

Pleural-based opacity

43 (36)

7 (23)

0 (0)

Elevated diaphragm

31 (26)

6 (19)

1 (20)

Decreased pulmonary vasculature

24 (20)

8 (26)

0 (0)

Prominent central pulmonary artery

17 (14)

5 (16)

0 (0)

Cardiomegaly

10 (8)

5 (16)

2 (40)§

7 (6)

2 (6)

0 (0)

Pulmonary edema

3 (3)

1 (3)

0 (0)

Normal chest X-ray

17 (14)

8 (26)

2 (40)

parenchymal abnormality

Westermark’s sign¶

* P < 0.02, pulmonary infarction vs. isolated dyspnea. † P < 0.01, pulmonary infarction vs. circulatory collapse. ‡ P < 0.01, pulmonary infarction vs. isolated dyspnea. § P < 0.02, pulmonary infarction vs. circulatory collapse. ¶ Prominent central pulmonary artery and decreased pulmonary vascularity. Reprinted with permission from Stein and Henry [7].

(Table 43.4). We previously showed that rales are associated with patients who have radiographic evidence of a parenchymal abnormality [3]. A normal ECG is frequently seen in patients with the pulmonary infarction syndrome, but a normal ECG is uncommon in patients with isolated dyspnea (Table 43.5). A Pa O2 higher than 80 mm Hg is not uncommon in patients with the pulmonary infarction syndrome,

but such levels are uncommon in patients with isolated dyspnea (Table 43.7). Abnormalities on the chest radiograph, although more common among patients with pulmonary infarction, are often observed in patients with isolated dyspnea (Table 43.6). A high-probability interpretation of the V–Q scan occurs in a minority of patients with the pulmonary infarction syndrome, but it was found

Table 43.7 Pulmonary artery mean pressure, Pa O2 , Pa CO2 , and pH (on room air) in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. Pulmonary infarction

Isolated dyspnea

Circulatory collapse

syndrome

syndrome

syndrome

Mean PA pressure (mm Hg)

20.1 ± 8.1* (n = 115)

24.5 ± 9.3 (n = 30)

25.8 ± 5.9 (n = 5)

Mean Pa O2 (mm Hg)

72.6 ± 14.2† (n = 99)

63.2 ± 12.9 (n = 19)

56.5 ± 2.7‡ (n = 4)

Mean Pa CO2 (mm Hg)

34.6 ± 5.2 (n = 99)

34.8 ± 4.3 (n = 19)

34.5 ± 4.7 (n = 4)

Mean pH

7.45 ± 0.04* (n = 99)

7.42 ± 0.08 (n = 19)

7.41 ± 0.07‡ (n = 4)

Values are mean ± standard deviation. * P < 0.02, pulmonary infarction vs. isolated dyspnea. † P < 0.01, pulmonary infarction vs. isolated dyspnea. ‡ P < 0.05, pulmonary infarction vs. circulatory collapse. PA, pulmonary artery; Pa O2 , partial pressure of oxygen in arterial blood; Pa CO2 , partial pressure of carbon dioxide in arterial blood. Reprinted with permission from Stein and Henry [7].


BLUK077-Stein

March 28, 2007

15:25

202

PART III

Diagnosis of acute PE

Table 43.8 Ventilation–perfusion lung scan in patients with pulmonary embolism and no prior cardiopulmonary disease in PIOPED I. n (%) Pulmonary infarction

Isolated dyspnea

Circulatory collapse

syndrome (N = 119)

syndrome (N = 31)

syndrome (N = 5)

38 (32)* 63 (53)*

20 (65)

4 (80)†

Intermediate

4 (13)

0 (0)†

Low

15 (13)

4 (13)

1 (20)

3 (3)

3 (10)

0 (0)

High

Near normal/normal

* P < 0.001, pulmonary infarction vs. isolated dyspnea. † P < 0.05, pulmonary infarction vs. circulatory collapse. Reprinted with permission from Stein and Henry [7].

in the majority of patients with isolated dyspnea (Table 43.8).

reduce the likelihood of inadvertently discarding the diagnosis.

References Circulatory collapse syndrome Some patients with circulatory collapse from PE, even though they have the most severe PE, are not short of breath. Based on pooled data from PIOPED I and PIOPED II, dyspnea was absent in 5 of 16 (31%) of patients with circulatory collapse who had PE and no prior cardiopulmonary disease [4, 7]. Among patients in shock in the Urokinase Pulmonary Embolism Trial, 6 of 21 (29%) did not have dyspnea, and among those with syncope, 2 of 19 (11%) did not have dyspnea [2]. Even frequent combinations of signs and symptoms may be absent in these patients with severe PE. Combined data from patients with PE and no prior cardiopulmonary disease in PIOPED I and PIOPED II showed that dyspnea and tachypnea were absent in 3 of 16 ( 19%) and the same number had an absence of dyspnea, tachypnea, and pleuritic pain [4, 7]. Patients with PE and circulatory collapse may have a normal radiograph (Table 43.6) and they may have a low-probability interpretation of the V–Q scan (Table 43.8). Awareness that some typical characteristics of PE may be absent, even with severe PE, may

1 A National Cooperative Study: the Urokinase Pulmonary Embolism Trial. Associated clinical and laboratory findings. Circulation 1973; 47(suppl II): II-51–II-85. 2 Stein PD, Willis PW, III, DeMets DL. History and physical examination in acute pulmonary embolism in patients without preexisting cardiac or pulmonary disease. Am. J Cardiol 1981; 47: 218–223. 3 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 4 Stein PD, Beemath A, Matta F et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med, 2007 (In press). 5 Stein PD, Henry JW. Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest 1995; 108: 978–981. 6 Stein PD, Gottschalk A, Saltzman HA, Terrin ML. Diagnosis of acute pulmonary embolism in the elderly. J Am Coll Cardiol 1991; 18: 1452–1457. 7 Stein PD, Henry JW. Clinical characteristics of patients with acute pulmonary embolism stratified according to their presenting syndromes. Chest 1997; 112: 974–979.


March 11, 2007

44

7:53

CHAPTER 44

Clinical assessment in the critically ill

Introduction Autopsy studies have consistently shown that pulmonary embolism (PE) is frequently not considered antemortem in patients who died from PE (see Chapter 1). It is apparent, therefore, that the clinical diagnosis of PE in critically ill patients is difficult. The clinical characteristics in several categories of critically ill patients were evaluated in order to identify features that may alert physicians to the diagnosis [1].

Ventilatory support, intensive care units, hypoxemic not on ventilatory support The clinical diagnosis of acute PE in an intensive care unit is generally considered to be exceedingly difficult [2]. A clinical assessment of the likelihood of PE was based on bedside evaluation, the plain chest radiograph, electrocardiogram, and ordinary laboratory tests [3]. Physicians were uncertain of whether PE was present in 67–78% of critically ill patients [1]. In noncritically ill patients, physicians were uncertain whether PE was present in 67%. Among patients on ventilatory support, patients in intensive care units not on ventilatory support, and hypoxemic patients not on ventilatory support, when physicians assessed an 80–100% likelihood of PE, 75–88% of patients in fact had PE [1]. This was comparable to the positive predictive value in noncritically ill patients (77%). There were few critically ill patients in whom physicians were confident enough to make a high-likelihood clinical assessment. A high-likelihood clinical assessment was made in only 9% of the critically ill patients. However, the proportion of patients with a high-likelihood clinical assessment in noncritically ill patients was comparable, 8%. If physicians assessed a 0–19% clinical likelihood of PE in these critically ill patients, only 0– 5% had PE. The number of critically ill patients with a

low-likelihood clinical assessment was also small. The frequency of PE in noncritically ill patients with a lowlikelihood clinical assessment was 15%. After making a clinical assessment, physicians in Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) were uncertain whether PE was present in 63–78% of patients with no prior cardiopulmonary disease [4], any prior cardiopulmonary disease (P.D. Stein and J.W. Henry, unpublished data from PIOPED, 1990), chronic obstructive pulmonary disease (COPD) [5], patients in intensive care units not on ventilatory support [1], patients on ventilatory support [1], and hypoxemic patients not on ventilatory support [1] (Figure 44.1). However, physicians were more often uncertain of whether PE was present in making a clinical assessment among patients in an intensive care unit not on ventilatory support than among patients with no prior cardiopulmonary disease. The percent of patients with an uncertain clinical assessment in other categories did not differ from each other to a statistically significant extent.

80

Patients (%)

BLUK077-Stein

60 40 20 0 NO CPD (n = 320)

ANY CPD (n = 478)

COPD (n = 100)

ICU NO VENT (n = 85)

VENT (n = 46)

Figure 44.1 Percent of patients in whom clinical assessment was uncertain regarding the likelihood of pulmonary embolism. NO CPD, no prior cardiopulmonary disease; ANY CPD, any prior cardiopulmonary disease; COPD, chronic obstructive pulmonary disease; ICU NO VENT, intensive care unit not on ventilatory support; VENT, ventilatory support. NO CPD versus ICU NO VENT, P < 0.01. All other differences were not statistically significant.

203


BLUK077-Stein

March 11, 2007

7:53

Pulmonary embolism (%)

204

PART III

Table 44.1 Positive predictive value of lung scan interpretation and clinical assessment in patients on ventilatory support.

100 80 60 40 20 0 NO CPD (n = 17)

ANY CPD (n = 22)

COPD (n = 3)

ICU NO VENT (n = 8)

VENT (n = 4)

Figure 44.2 Frequency of pulmonary embolism among patients in whom a high-likelihood clinical assessment was made. Abbreviations are as in Figure 44.1. There were no statistically significant differences in the frequency of pulmonary embolism between the various groups.

40 Pulmonary embolism (%)

Diagnosis of acute PE

30 20 10 0 NO CPD (n = 103)

ANY CPD (n = 123)

COPD (n = 30)

ICU NO VENT (n = 11)

VENT (n = 11)

Figure 44.3 Frequency of pulmonary embolism among patients in whom the clinical assessment indicated a low likelihood of pulmonary embolism. Abbreviations are as in Figure 44.1. Pulmonary embolism was more frequent among patients with no cardiopulmonary disease (CPD) than with any CPD (P < 0.05). Other differences were not statistically significant.

V–Q scan

Total PE+/Test+ (%)

High

4/4 (100)

Intermediate

6/24 (25)

Low

2/17 (12)

Near normal/normal

0/1 (0)

Total

12/46 (26)

PE, pulmonary embolism. Reprinted with permission from Henry et al. [1].

Among patients with no prior cardiopulmonary disease, any prior cardiopulmonary disease, COPD, patients in intensive care units not on ventilatory support, patients on ventilatory support, and hypoxemic patients not on ventilatory support, if clinical assessment indicated a high likelihood of PE, it was present in 75% or more of the patients in each of these categories [1, 4–6] (Figure 44.2). Few patients, however, had a high-likelihood clinical assessment among those with COPD, in the intensive care unit not on ventilatory support, and patients on ventilatory support. Among patients in whom physicians made a clinical assessment of PE as being low likelihood, PE was present in 0–14% of the patients in the various categories (Figure 44.3).

Ventilation–perfusion lung scans Most patients on ventilatory support, 41 of 46 (89%), showed nondiagnostic (intermediate or low probability) ventilation–perfusion lung scans (Table 44.1, Figure 44.4).

Patients (%)

80 60 40 20

References

0 NO CPD (n = 365)

ANY CPD (n = 526)

COPD (n = 98)

ICU NO VENT (n = 89)

VENT (n = 46)

Figure 44.4 Percent of patients with intermediate probability ventilation–perfusion (V–Q) lung scans. Abbreviations are as in Figure 44.1. Fewer patients with no prior CPD had intermediate probability V–Q scans than any other category (P < 0.01–0.001). Patients with COPD had intermediate probability V–Q scans more frequently than any other group (P < 0.01–0.001) except in patients on ventilatory support.

1 Henry JW, Stein PD, Gottschalk A, Relyea B, Leeper KV, Jr. Scintigraphic lung scans and clinical assessment in critically ill patients with suspected acute pulmonary embolism. Chest 1996; 109: 462–466. 2 Cowen JC, Kelley MA. An organized approach to detecting pulmonary embolism in the critically ill. J Crit Illn 1994; 9: 551–567. 3 The PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary


BLUK077-Stein

March 11, 2007

CHAPTER 44

7:53

Clinical assessment in critically ill

Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753– 2759. 4 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no pre-existing

205

cardiac or pulmonary disease. Chest 1991; 100: 598– 603. 5 Lesser BA, Leeper KV, Stein PD et al. The diagnosis of pulmonary embolism in patients with chronic obstructive pulmonary disease. Chest 1992; 102: 17–22.


BLUK077-Stein

March 12, 2007

45

16:20

CHAPTER 45

The electrocardiogram

In 1935, McGinn and White described the electrocardiograms of 7 patients with acute cor pulmonale secondary to pulmonary embolism (PE) [1]. They described the presence of a Q wave and late inversion of the T wave in lead III and a low origin of the T wave with a gradual staircase ascent of the ST segment in lead II. They also noted an S wave in lead I. Their description of the S1 Q3 T3 complex became a traditional electrocardiographic manifestation of acute cor pulmonale. This observation was expanded by Murnaghan and associates [2] in 1943, with the report of 10 additional patients with acute cor pulmonale and no prior heart disease. An S1 Q3 T3 pattern was again observed, sometimes associated with right-axis deviation. Some of the patients showed a depressed ST segment; others showed ST-segment elevation [2]. However, even in the presence of preexistent cardiac disease, these electrocardiographic indications of acute cor pulmonale were present in only about one-third of the patients. Based upon studies of patients and of dogs, Love and associates, in 1938, concluded that the typical electrocardiogram of PE was ST-segment depression and T-wave inversion [3]. Transient right bundle branch block in PE was reported in 1939 [4]. Wood [5], and subsequently others [6–9], reported sharp T-wave inversion in the precordial leads facing the right ventricle. Clockwise rotation was described by Wilson and associates in 1947 [10] and was confirmed subsequently [9]. A shift of the transitional zone to the left and evidence of incomplete right bundle branch block was found helpful in distinguishing the prominent Q waves of PE from those of myocardial infarction [7–11]. P pulmonale in acute PE was described by Katz in 1946 [12]. By 1940, Sokolow and associates stated that no single electrocardiographic abnormality is consistently present in PE [13]. In a study of 50 patients, they found that a large percentage had either normal electrocardiograms or nonspecific changes [13]. The distinctive pattern described by McGinn and White occurred only in 10% of the patients [1].

206

The prevalence of various electrocardiographic abnormalities in 90 patients with massive or submassive PE and in 89 patients with mild to massive PE, all of whom had no prior cardiopulmonary disease, is shown in Table 45.1 [14, 15]. Definitions of various electrocardiographic abnormalities are shown in Table 45.2 [14]. The electrocardiographic abnormalities observed in patients with no prior cardiopulmonary disease who had submassive or massive PE were comparable to the electrocardiographic changes observed in patients with no prior cardiopulmonary disease in whom the severity of the PE ranged from mild to severe (Table 45.1) [15]. Definitions of electrocardiographic changes are shown in Table 45.2 [16–21]. A normal electrocardiogram was shown in only 6% of the patients with massive acute PE , whereas in those with submassive acute PE , it was normal in 23%. Although this would appear to suggest a trend, the difference was not statistically significant. In patients with mild to massive PE, a normal electrocardiogram was shown in 30% [15]. Cutforth and Oram reported 50 patients with PE, 49 of whom had patent coronary arteries and 24% of whom had normal electrocardiograms [19]. None of the patients who had massive or submassive PE with no prior cardiopulmonary disease had atrial flutter or atrial fibrillation [14], and only 5% of the patients with no prior cardiopulmonary disease in whom the severity of the PE ranged from mild to severe had atrial fibrillation or atrial flutter [15]. None had second- or third-degree atrioventricular block [14, 15]. Only 1 patient had first-degree atrioventricular block [14, 15]. Among 23 patients with acute PE and no heart disease reported by Szucs and associates, only 4% (1 of 23) had atrial flutter or atrial fibrillation [22]. This observation led Dalen to conclude that atrial flutter or atrial fibrillation in patients with PE is nearly always limited to individuals with prior heart disease [23]. Right-axis deviation in PE has been observed for many years [1, 11]. Electrocardiographic


BLUK077-Stein

March 12, 2007

CHAPTER 45

16:20

207

Electrocardiogram

Table 45.1 Electrocardiographic manifestations: patients without prior cardiac or pulmonary disease. Pulmonary embolism (%) Electrocardiogram

Mild to massive (n = 89)*

Submassive to massive (n = 90)†

Normal electrocardiogram

30

13

Atrial flutter

1

0

Atrial fibrillation

4

0

Atrial premature contractions

4

2

Ventricular premature contractions

4

3

2

6

Rhythm disturbances

P wave P pulmonale QRS abnormalities 2

7

Left-axis deviation

Right-axis deviation

13

7

Clockwise rotation (V5 )

7

S1 S2 S3

7

S1 Q3 T3

12

Incomplete right bundle branch block

4

6

Complete right bundle branch block

6

9

Right ventricular hypertrophy

2

6

Pseudoinfarction

3

11

Low voltage (frontal plane)

3

6

ST segment and T wave Nonspecific T wave

42

ST-segment depression

26

ST segment elevation

16

Nonspecific ST segment or T wave

49

Some patients had more than one abnormality. ∗ Data from Stein et al. [15]. † Data from Stein et al. [14].

manifestations of acute cor pulmonale (S1 Q3 T3 , complete right bundle branch block, P pulmonale, or right-axis deviation) were less common than ST-segment or T-wave changes (Table 45.1) [14]. One or more of these abnormalities occurred in 26% of the patients with submassive or massive acute PE not associated with cardiac or pulmonary disease and occurred in 32% of such patients who suffered massive PE. Cutforth and Oram reported that 14% of the patients with PE had right bundle branch block and 28% had an S1 Q3 T3 [19]. A pseudoinfarction pattern was seen in 11% of the patients with submassive or massive PE who had no prior cardiopulmonary disease [14]. A pseudoinfarction pattern, however, was seen in only 3% of the patients with no prior cardiopulmonary disease who had PE that ranged in severity from mild to massive [15]. The problem of distinguishing the prominent Q waves

known to occur with PE from those of myocardial infarction has been recognized for many years [7, 9]. The electrocardiogram may simulate an inferior infarction with Q waves and T-wave inversion in leads II, III, and aVF [7, 24] or anteroseptal infarction characterized by QS or QR waves in V1 and T-wave inversion in the right precordial leads [19, 20]. The development of Q waves and extensive T-wave inversion in the anterior and lateral leads was also observed in 1 patient following PE [25]. Leftward shifts of the frontal plane axis in PE are frequent [26, 27]. Karlen and Wolff noted a shift of the frontal plane axis toward the horizontal position in 4 patients with PE [24]. An example of a patient with a previously normal frontal plane QRS axis that showed a shift to −25◦ following an acute PE and subsequently reverted to normal is shown in Figure 45.1 [26]. In a review of 115 patients with acute PE, a leftward shift


BLUK077-Stein

March 12, 2007

16:20

208

PART III

Diagnosis of acute PE

Table 45.2 Definitions of the various electrocardiographic abnormalities. Abnormalities of the P wave Right atrial enlargement. The P wave was considered to be suggestive of right atrial enlargement if it was 0.25 mV (2.5 mm) in the extremity leads or over 0.15 mV in lead V1 [16, 17]. Abnormalities of the QRS complex Right-axis deviation. Right-axis deviation was defined as a mean frontal plane QRS electrical axis greater than 90◦ [17]. Left-axis deviation. This was defined as a mean frontal plane QRS axis equal to or leftward of −30◦ . The A QRS was measured by the simplified method of Grant [18]. Many patients with marked left-axis deviation undoubtedly had left anterior fascicular block, but this was not indicated as a separate interpretation. S1 S2 S3 pattern. This was defined as the occurrence of S waves in leads I, II, and III of at least 0.15 mV (1.5 mm) amplitude in each lead. Clockwise rotation. This was defined as a shift in the transition zone (R = S) in the precordial leads. Usually a shift to the V4 position or further leftward is considered as clockwise rotation [19, 20]. For the purpose of this study, a shift in the transition zone to V5 was analyzed to avoid problems of interpretation that may occur with minor errors of precordial electrode position. Right bundle branch block [19, 20]. Incomplete right bundle branch block was diagnosed when the QRS duration was 0.10–0.11 s and when the terminal QRS forces were directed rightward and anteriorly causing the production of S waves in leads I and R waves in lead V1 . Complete right bundle branch block. This was considered to be present in those patients with a QRS duration of 0.12 s or greater and with terminal QRS forces directed as previously described. S1 Q3 T3 pattern [1, 19, 20]. The presence of S waves in lead I and Q waves in lead III, each of amplitude exceeding 0.15 mV (1.5 mm), associated with inversion of the T wave in lead III constituted the S1 Q3 T3 pattern. Right ventricular hypertrophy [17, 21]. Right ventricular hypertrophy was considered to be present if the R wave in lead V1 exceeded 0.5 mV (5 mm) or the R/S ratio in lead V1 was greater than 1. In this study, right-axis deviation was not required to make this diagnosis nor was the presence of right-axis deviation alone considered adequate for the diagnosis of right ventricular hypertrophy. Low-voltage QRS complexes. The QRS complexes were considered to be low voltage if the greatest overall deflection of the QRS was 0.5 mV (5 mm) or less in all of the limb leads [17]. Pseudoinfarction pattern. This was defined as ST-segment or T-wave abnormalities associated with prominent Q waves, which simulated a recent myocardial infarction. ST-segment depression. This was considered present if the ST segment was depressed 0.05 mV (0.5 mm) or greater in any lead except aVR. No distinction was made between junctional or ischemic ST-segment depression. Changes were considered primary in the absence of complete bundle branch block or ventricular hypertrophy. ST-segment elevation. This was recorded as present when found in any lead except aVR if the ST segment was elevated at least 0.1 mV (1 mm) in the absence of ST depression in any other lead. T-wave inversion. This was considered to be present if it occurred in any lead except aVL, III, aVR, or V1 . Changes were considered primary in the absence of complete bundle branch block or ventricular hypertrophy. Reprinted from Stein et al. [14], with permission from Elsevier.

of the frontal plane QRS axis was observed more frequently than a rightward shift, and left-axis deviation was present more often than right-axis deviation [27]. In patients with submassive or massive PE and no prior cardiopulmonary disease, left-axis deviation was recorded as frequently as right-axis deviation (Table 45.1) [14]. Among patients with PE who had no prior cardiopulmonary disease in PIOPED, left-axis deviation was more frequent than right-axis deviation (Table 45.1) [15]. Emphysema is also a cause of left-axis deviation [28]. Since the electrocardiogram of emphy-

sema is usually easily recognized, the term pseudo leftaxis deviation has been termed for such individuals to distinguish the electrocardiogram from that of left ventricular conduction abnormalities. Abnormal conduction through the emphysematous lungs has been postulated as a possible cause of the left-axis deviation in these patients [28]. Low-voltage frontal plane QRS complexes in PE was first observed by Stein and associates [14]. It was observed in 6% of the patients with massive or submassive PE who did not have prior cardiopulmonary


BLUK077-Stein

March 12, 2007

CHAPTER 45

16:20

209

Electrocardiogram

(a)

(b)

Figure 45.1 (a) Electrocardiogram 1 week prior to acute pulmonary embolism. Maximal frontal plane QRS axis, +70◦ , is within normal limits. (b) Electrocardiogram on the day of hospitalization for acute pulmonary embolism. The frontal plane QRS axis has shifted leftward to −25◦ . Nonspecific ST-segment and T-wave changes are also present. (c) Electrocardiogram 5 days after transvenous embolectomy shows return of frontal plane QRS axis toward control value. Frontal plane QRS axis is now +15◦ . (Reprinted from Stein and Bruce [26], with permission from Elsevier.)

(c)


BLUK077-Stein

March 12, 2007

16:20

210

PART III

Diagnosis of acute PE

Figure 45.2 (a) Electrocardiogram on the day of diagnostic pulmonary angiogram showing low-voltage QRS complexes in frontal plane. (b) Electrocardiogram 13 days later. Amplitude of QRS complexes is normal. (Reprinted from Stein et al. Bruce [14], with permission from Elsevier.)

disease (Table 45.1) [14]. Among patients in Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) with mild to massive PE who had no prior cardiopulmonary disease, 3% had low-voltage QRS complexes [15]. An electrocardiogram showing lowvoltage frontal plane QRS complexes that subsequently resolved is shown in Figure 45.2. Right ventricular endocardial potential is reduced in massive PE, possibly reflecting acute myocardial stress [29]. In 1938, Love and associates considered changes of the ST segment and T wave to be â&#x20AC;&#x153;the significant electrocardiographic changes notedâ&#x20AC;? in acute PE [3]. Abnormalities of the ST segment and T wave are by far the most frequent electrocardiographic manifestation of PE [14, 15]. Nonspecific T-wave changes occurred in 42% of the patients with massive or submassive embolization [14]. A similar number of patients (41%) showed either depression or elevation of the ST seg-

ment [14]. Nonspecific ST-segment or T-wave changes were observed in 49% of the patients who had no prior cardiopulmonary disease and in whom the severity of PE ranged from mild to severe [15]. Inversion of the T wave was also the most common electrocardiographic abnormality reported by Cutforth and Oram (46%) [19]. They observed depression of the ST segment in 18% of the patients with PE [19]. As early as 1935, the electrocardiographic abnormalities in acute PE were known to be transient and in some instances lasted only 48 hours [1]. Among patients with submassive or massive PE and no prior cardiopulmonary disease, inversion of the T waves was the most persistent electrocardiographic abnormality [14]. Abnormalities of the QRS complex appeared to be more transient [15]. The duration of electrocardiographic abnormalities following acute PE is shown in Table 45.3 [14]. (The


BLUK077-Stein

March 12, 2007

CHAPTER 45

16:20

211

Electrocardiogram

Table 45.3 Disappearance of electrocardiographic abnormalities following acute massive or submassive pulmonary embolism. Days after diagnostic pulmonary angiography Electrocardiographic abnormality*

5 or 6

14

P-wave abnormality* P pulmonale

80% (4 of 5)

80% (4 of 5)

Right-axis deviation

60% (3 of 5)

60% (3 of 5)

Left-axis deviation

50% (3 of 6)

50% (3 of 6)

Clockwise rotation (V5 )

67% (4 of 6)

67% (4 of 6)

Incomplete right bundle branch block

80% (4 of 5)

100% (5 of 5)

QRS abnormalities

Complete right bundle branch block

75% (6 of 8)

75% (6 of 8)

Right ventricular hypertrophy

60% (3 of 5)

80% (4 of 5)

S 1 S2 S3

67% (4 of 6)

80% (4 of 5)

S 1 Q3 T3

100% (11 of 11)

100%(11 of 11)

67% (6 of 9)

67% (6 of 9)

0% (0 of 4)

67% (2 of 3)

Pseudoinfarction Low voltage (frontal plane) Primary RST-segment and T-wave abnormalities RST-segment depression (not reciprocal)

48% (11 of 23)

74% (17 of 23)

RST-segment elevation (not reciprocal)

31% (4 of 13)

14% (2 of 14)†

T-wave inversion

22% (8 of 37)

49% (18 of 37)

*Some patients had more than one abnormality. † Recurrence noted in 2 patients. Reprinted from Stein et al. [14], with permission from Elsevier.

day of diagnostic arteriography and baseline electrocardiogram is considered as day 0.) The acute PE, by history, may have occurred as many as 5 days previously. Abnormalities of repolarization tended to persist. Inversion of the T wave disappeared in only 22% of the patients 5 or 6 days after the PE was diagnosed by angiography, although it resolved in 49% by 2 weeks [14]. Depression of the ST segment tended to resolve somewhat faster. In contradistinction to repolarization, many of the abnormalities of depolarization disappeared faster. In well over half of the patients with pseudoinfarction, S1 S2 S3 , S1 Q3 T3 , right ventricular hypertrophy, or right bundle branch block, these abnormalities were no longer apparent on the electrocardiograms 5 or 6 days after the angiographic diagnosis. Examples of the occurrence and disappearance of abnormal left-axis deviation (left anterior fascicular block), low-voltage QRS complexes in the frontal plane, pseudoinfarction, and apparent acute ischemic event are shown in Figures 45.1–45.4, respectively. Electrocardiographic abnormalities of acute massive or submassive pulmonary embolization were frequently related to the severity of the embolization as

indicated by lung scans or pulmonary arteriography [14]. Patients with ST-segment abnormalities, T-wave inversion, pseudoinfarction patterns, S1 Q3 T3 patterns, incomplete right bundle branch block, right-axis deviation, right ventricular hypertrophy, or ventricular premature beats had statistically significantly larger perfusion defects on the lung scan or larger defects on the pulmonary arteriogram than those with normal electrocardiograms [14]. Elevation of the pulmonary arterial mean pressure or right ventricular end-diastolic pressure was often associated with electrocardiographic abnormalities. In patients with right-axis deviation, incomplete right bundle branch block, S1 S2 S3 pattern, pseudoinfarction pattern, ST-segment depression, or T-wave inversion, either the pulmonary arterial mean pressure or right ventricular end-diastolic pressure or both were significantly higher than in patients who had normal electrocardiograms [14]. The partial pressure of oxygen in arterial blood (Pa O2 ) in patients with various electrocardiographic abnormalities, with the exception of patients with right-axis deviation, did not differ significantly from


BLUK077-Stein

March 12, 2007

16:20

212

PART III

Diagnosis of acute PE

Figure 45.3 (a) Electrocardiogram on the day of pulmonary angiogram showing Q waves in II, II, aVF, and V1 simulating recent inferior and anteroseptal myocardial infarction. Pulmonary artery pressure was 63/24 mm Hg and Pa O2 was 63 mm Hg. (b) Electrocardiogram next day shows no significant Q waves. Inversion of T waves persists. Pulmonary artery pressure decreased to 32/12 mm Hg and Pa O2 increased to 89 mm Hg. (Reprinted from Stein et al. [14], with permission from Elsevier.)

the depressed levels found in patients who had normal electrocardiograms [14]. Even so, a contribution of hypoxemia to the electrocardiographic changes cannot be excluded. Neither right ventricular work nor the total pulmonary resistance of patients with abnormal electrocardiograms differed significantly from values measured in patients with normal electrocardiograms. The product of heart rate and pulmonary arterial peak pressure also showed no difference between groups. Peak brachial arterial pressure was 90 mm Hg or higher in all but 3 patients (96%). The magnitude of the T-wave inversion was not related to any of the measured hemodynamic abnormalities or to the severity of PE [14]. Patients in whom the

T wave was inverted, 0.05â&#x20AC;&#x201C;0.1 mV, had hemodynamic, angiographic, and perfusion abnormalities comparable to patients in whom the T waves were inverted more than 0.3 mV in the frontal plane. The electrocardiographic changes in PE have been postulated to be related to several etiologic mechanisms. Since some of the electrocardiographic abnormalities simulate myocardial ischemia, and even infarction, it has been suggested that myocardial underoxygenation is the cause [30]. The right ventricle would be especially embarrassed by myocardial ischemia because of the load placed upon it by obstruction of the pulmonary circulation. The main cardiac effects of PE are (1) strain upon the right ventricle and


BLUK077-Stein

March 12, 2007

CHAPTER 45

16:20

Electrocardiogram

213

Figure 45.4 (a) 56-year-old man with acute pulmonary embolism shown by CT angiography. Coronary angiogram on same day was normal. Electrocardiogram simulates acute ischemic event showing prominent ST-segment elevation in leads V1 and V2 . (b) Electrocardiogram next day no longer shows ST-segment elevation. (Courtesy of Elias Skaf, MD.)

(2) myocardial ischemia [30]. The latter has been postulated to be due to shock, hypoxemia, or vagal reflexes from the lung to the coronary arteries [30]. Myocardial necrosis, usually involving the left ventricle, has been observed at autopsies of patients who died of PE and had normal coronary arteries [30, 31]. Because studies of left coronary flow [32] and right coronary blood flow [33] following experimentally induced acute PE in dogs and pigs showed an increased coronary flow (Figure 45.5), rather than a reduction, reflex spasm of the coronary arteries appears unlikely. Shock also appears to be an unlikely cause, because electrocardiographic changes frequently occur in the absence of shock [14]. An effect of hypoxemia on the myocardium cannot be excluded. Such an effect, however, would be secondary, since the partial pressure of oxygen in arterial blood (Pa O2 ) in patients with abnormal electrocardiograms, in general, did not differ from that of patients with normal electrocardiograms [14]. The possibility also exists that some patients may have had asymptomatic coronary heart disease. In the

presence of acute PE and hypoxemia, such individuals would not be able to increase coronary flow adequately to compensate for the underoxygenation of arterial blood. Love and associates concluded that dilatation of the right ventricle produced the electrocardiographic changes in PE [3]. Mechanical obstruction of the pulmonary artery produced electrocardiographic changes similar to those produced by PE. The electrocardiographic abnormalities were preceded by visible and marked dilatation of the right ventricle [3]. Section of the vagi or cervical sympathetic cords had no effect on the production of the electrocardiographic changes [3]. Wood also concluded that the essential factor in the production of electrocardiographic abnormalities in PE was right ventricular stress [5]. Electrocardiographic changes appeared to depend on the size of the PE [5]. Occlusion of the pulmonary artery of human beings during thoracic surgery produced some of the electrocardiographic changes observed in PE, thereby supporting this contention [34].


BLUK077-Stein

March 12, 2007

16:20

214

PART III

Diagnosis of acute PE

Figure 45.5 Simultaneous mean left circumflex coronary artery blood flow (LCBF), femoral artery mean pressure (FAM), pulmonary artery mean pressure (PAM), right atrial mean pressure (RAM), and electrocardiogram (ECG) in a dog during the administration of clot pulmonary embolism (PE). The LCBF was measured with an electromagnetic flow transducer placed around the artery. The LCBF increased from a control value of 88 to 210 cm3 /min 1 minute after PE. The PAM increased to 72 mm Hg 25 seconds after PE and decreased to 53 mm Hg 1 minute after PE. There was a transient reduction of FAM and a transient elevation of RAM. (Reprinted from Stein et al. [32], with permission from Elsevier.)

Occasional patients with a Pa O2 as low as 51 mm Hg and mean pulmonary arterial pressures as high as 31 mm Hg had normal electrocardiograms [14]. Even so, most of the patients with electrocardiographic abnormalities had more severe evidence of PE as indicated by the lung scan or pulmonary angiogram than patients with normal electrocardiograms. The frequent association of electrocardiographic abnormalities with an elevated pulmonary arterial mean pressure or elevated right ventricular end-diastolic pressure supports the hypothesis that ventricular dilatation represents the major cause of the electrocardiographic changes. Arterial hypoxemia did not correlate with the electrocardiographic changes, although the Pa O2 was low in many patients and may have contributed to the observed abnormalities [14].

References 1 McGinn S, White PD. Acute cor pulmonale resulting from pulmonary embolism. JAMA 1935; 104: 1473–1480. 2 Murnaghan D, McGinn S, White PD. Pulmonary embolism with and without acute cor pulmonale, with special reference to the electrocardiogram. Am Heart J 1943; 25: 573–597. 3 Love WS, Jr, Brugler GW, Winslow N. Electrocardiographic studies in clinical and experimental pulmonary embolization. Ann Intern Med 1938; 11: 2109–2123.

4 Durant TM, Ginsburg IW, Roesler H. Transient bundle branch block and other electrocardiographic changes in pulmonary embolism. Am Heart J 1939; 17: 423–430. 5 Wood P. Pulmonary embolism: diagnosis by chest lead electrocardiography. Br Heart J 1941; 3: 21–29. 6 Myers GB, Klein HA, Stoffer BE. The electrocardiographic diagnosis of right ventricular hypertrophy. Am Heart J 1948; 35: 1–40. 7 Phillips E, Levine HD. A critical evaluation of extremity and precordial electrocardiography in acute cor pulmonale. Am Heart J 1950; 39: 205–216. 8 Wood P. Electrocardiographic appearances in acute and chronic pulmonary heart disease. Br Heart J 1948; 10: 87. 9 Eliaser M, Jr, Giansiracusa F. The electrocardiographic diagnosis of acute cor pulmonale. Am Heart J 1952; 43: 533–545. 10 Wilson FN, Rosenbaum FF, Johnston FD. Interpretation of the ventricular complex of the electrocardiogram. Adv Intern Med 1947; 2: 1–63. 11 Kuo PT, Vander Veer JB. Electrocardiographic changes in pulmonary embolism with special reference to an early and transient shift of the electrical axis of the heart. Am Heart J 1950; 40: 825–838. 12 Katz LN. Electrocardiography, 2nd edn. Lea & Febiger, Philadelphia, 1946: 401–406. 13 Sokolow M, Katz LN, Muscovitz AN. The electrocardiogram in pulmonary embolism. Am Heart J 1940; 19: 166– 184. 14 Stein PD, Dalen JE, McIntyre KM, Sasahara AA, Wenger NK, Willis PW, III. The electrocardiogram in acute


BLUK077-Stein

March 12, 2007

CHAPTER 45

15

16

17 18 19 20

21

22

23

24

16:20

Electrocardiogram

pulmonary embolism. Prog Cardiovasc Dis 1975; 17: 247–257. Stein PD, Terrin ML, Hales CA et al.. Clinical, laboratory, roentgenographic and electrocardiographic findings in patients with acute pulmonary embolism and no pre-existing cardiac or pulmonary disease. Chest 1991; 100: 598–603. Thomas P, Dejong D. The P wave in the electrocardiogram in the diagnosis of heart disease. Br Heart J 1954; 16: 241– 254. Friedberg CK. Diseases of the Heart, 3rd edn. Saunders, Philadelphia, 1966: 38, 171, 188. Grant RP. Clinical Electrocardiography. McGraw-Hill, New York, 1957: 13. Cutforth RH, Oram S. The electrocardiogram in pulmonary embolism. Br Heart J 1958; 20: 41–60. Weber DM, Phillips JH, Jr. A re-evaluation of electrocardiographic changes accompanying acute pulmonary embolism. Am J Med Sci 1966; 251: 381–398. Milnor WR. Electrocardiogram and vectorcardiogram in right ventricular hypertrophy and right bundle branch block. Circulation 1957; 16: 348–367. Szucs MM, Jr, Brooks HL, Grossman W et al. Diagnostic sensitivity of laboratory findings in acute pulmonary embolism. Ann Intern Med 1971; 74: 161–166. Dalen JE. Diagnosis of acute pulmonary embolism. In: Dalen JE, ed. Pulmonary Embolism., Medcom, New York, 1972: 28–39. Karlen WS, Wolff L. The vectorcardiogram in pulmonary embolism. II. Am Heart J 1956; 51: 839–860.

215

25 Romhilt D, Susilavorn B, Chou T. Unusual electrocardiographic manifestation of pulmonary embolism. Am Heart J 1970; 80: 237–241. 26 Stein PD, Bruce TA. Left axis deviation as an electrocardiographic manifestation of acute pulmonary embolism. J Electrocardiol 1971; 4: 67–69. 27 Lynch RE, Stein PD, Bruce TA. Leftward shift of frontal plane QRS axis as a frequent manifestation of acute pulmonary embolism. Chest 1972; 61: 443–446. 28 Pryor R, Blount SG, Jr. The clinical significance of true left axis deviation. Left intraventricular blocks. Am Heart J 1966; 72: 391–413. 29 Chatterjee K, Sutton GC, Miller GAH. Right ventricular endocardial potential in acute massive pulmonary embolism. Br Heart J 1972; 34: 271–273. 30 Horn H, Dack S, Friedberg CK. Cardiac sequelae of embolism of the pulmonary artery. Arch Intern Med 1939; 64: 296–321. 31 Dack S, Master AM, Horn H et al. Acute coronary insufficiency due to pulmonary embolism. Am J Med 1949; 7: 464–477. 32 Stein PD, Alshabkhoun S, Hatem C et al. Coronary artery blood flow in acute pulmonary embolism. Am J Cardiol 1968; 21: 32–37. 33 Stein, PD, Alshabkhoun S, Hawkins HF et al. Right coronary blood flow in acute pulmonary embolism. Am Heart J 1969; 77: 356–362. 34 Semisch CW, III, Merves L. Electrocardiographic studies on artificially produced pulmonary artery occlusion in human beings. Arch Intern Med 1942; 69: 417–428.


BLUK077-Stein

March 12, 2007

46

16:22

CHAPTER 46

The plain chest radiograph

Introduction Abnormalities on the plain chest radiograph of patients with pulmonary embolism (PE) were described in the late 1930s and early 1940s [1–3]. The radiographic signs of acute PE include oligemia, pleonemia of the obstructed lung, dilated hilar arteries, elevated hemidiaphragm, dilated pulmonary artery trunk, pleural effusion, atelectasis, and infiltrate [4, 5]. Occasionally, cavity formation may occur with pulmonary infarction [5]. Dilatation of the azygos vein and superior vena cava has been described [4]. Pulmonary edema has also been described [4]. It was thought by some to reflect underlying heart disease [5], although pulmonary edema has been observed in patients with no prior cardiopulmonary disease [6]. Clearly, the plain chest radiograph cannot be considered a definitive examination [7]. Importantly,

however, the chest radiograph may provide an additional data point in the evaluation of patients with suspected acute PE and may assist in the exclusion of disease processes, such as pneumothorax or rib fracture, which may simulate PE. The frequency of abnormalities on the plain chest radiograph in 383 patients in the Prospective Investigation of Pulmonary Embolism Diagnosis I (PIOPED I) with acute PE, the severity of which ranged from mild to massive, is shown in Table 46.1 [8]. Many of these patients had prior cardiopulmonary disease. The frequency of abnormalities on the plain chest radiograph in 128 patients in the Urokinase Pulmonary Embolism Trial who had submassive to massive PE is also shown in Table 46.1 [9]. Approximately one-third of these patients had prior cardiopulmonary disease [10]. Radiographic findings among patients with acute PE and no prior cardiopulmonary disease are shown

Table 46.1 Plain chest radiograph in patients with acute pulmonary embolism, irrespective of prior cardiopulmonary disease. Mild to massive

Submassive to massive

PE (%) (n = 383)*

PE (%) (n = 128)†

Atelectasis or pulmonary parenchymal abnormality

69

Atelectasis

20

Consolidation

41

Pleural effusion

47

28

Pleural-based opacity

34

Elevated hemidiaphragm

28

41

Decreased pulmonary vascularity

20

15

Prominent central pulmonary artery

20

23

Cardiomegaly

16

Westermark’s sign‡

6

Pulmonary edema

10

Pulmonary venous hypertension pattern

3

Pulmonary arterial hypertension pattern

3

*Data are from Stein et al. [8] based on data from PIOPED. † Data are from the Urokinase Pulmonary Embolism Trial [9]. ‡ Prominent central pulmonary artery and decreased pulmonary vascularity.

216


BLUK077-Stein

March 12, 2007

CHAPTER 46

16:22

217

Plain chest radiograph

Table 46.2 Plain chest radiograph in patients with acute pulmonary embolism and no prior cardiopulmonary disease. Mild to massive

Submassive to massive

PE (%) (n = 117)*

PE (%) (n = 169)†

Atelectasis or pulmonary parenchymal abnormality

68

Atelectasis

28

Consolidation

39

Pleural effusion

48‡

30

Pleural-based opacity

35

— 46

Elevated hemidiaphragm

24

Decreased pulmonary vascularity

21

22

Prominent central pulmonary artery

15

21

Cardiomegaly

12

7

Westermark’s sign§ Pulmonary edema

4

Pulmonary venous hypertension pattern

4

Pulmonary arterial hypertension pattern

2

*Data are from Stein et al. [6]. † Data are from Stein et al. [11]. ‡ Among patients with a pleural effusion, 86% had only blunting of the costophrenic angle. None had a pleural effusion that occupied more than one-third of a hemithorax. § Prominent central pulmonary artery and decreased pulmonary vascularity. Reproduced from Stein et al. [11], with permission from S. Karger AG, Basel.

no prior cardiopulmonary disease, the plain chest radiograph was normal in 16% [6]. Among 169 patients with submassive or massive acute PE and no prior cardiopulmonary disease, one or more of the following signs (elevated hemidiaphragm, consolidation, pleural effusion, or atelectasis) occurred in 67% [11]. Parenchymal signs occurred with comparable frequency among patients with submassive PE (71%) and massive PE (65%). Either consolidation, atelectasis, or pleural effusion occurred in 62% with submassive PE and 50% with massive PE [11]. Atelectasis or a pulmonary parenchymal

in Table 46.2. The sensitivity of these findings in relation to the frequency of false-positive radiographic signs among 364 patients with no prior cardiopulmonary disease in whom PE was suspected is shown in Figure 46.1 [6]. Among 169 patients with submassive or massive acute PE and no prior cardiopulmonary disease, a normal chest radiograph was observed in 24% [11]. The frequency of a normal plain chest radiograph was similar among patients with submassive PE (25%) and massive PE (24%). Among 117 patients with PE that ranged in severity from mild to massive and who had

Figure 46.1 Sensitivity and frequency of false-positive values of various abnormalities on the plain chest radiograph. Data are from Stein et al. [6] in 364 patients with suspected acute pulmonary embolism and no prior cardiopulmonary disease. The numbers refer to the individual radiographic abnormalities listed in the insert.

Sensitivity (%)

100 80 1

60

1. Atelectasis/parenchymal abnormality 2. Pleural effusion 3. Pleural-based opacity 4. Elevated diaphragm 5. Decreased pulmonary vasculature 6. Prominent central PA 7. Cardiomegaly , 8. Westermark s sign 9. Pulmonary edema

2

40

3 4

5 6

20

7 9

0 0

8

20

40

60

False positives (%)

80

100


BLUK077-Stein

March 12, 2007

16:22

218

abnormality were the most frequent abnormalities on the plain chest radiographs in patients with no prior cardiopulmonary disease who had mild to massive PE (Table 46.2) [6]. Among these patients, a pleural effusion occurred in 48% [6]. Only blunting of the costophrenic angle was shown in 86% of the patients with PE who had a pleural effusion. No patient with PE and no prior cardiopulmonary disease had a pleural effusion that occupied more than one-third of a hemithorax [6]. Among patients with no prior cardiopulmonary disease, one or more vascular signs (focal oligemia, distension of the proximal pulmonary artery, pulmonary artery hypertension pattern, or pulmonary venous hypertension pattern) occurred in 37%. Vascular signs on the plain chest radiograph were more frequent among patients with massive PE than with submassive PE (43% versus 28%) [11]. Either focal oligemia or distension of the proximal portion of the pulmonary artery or both occurred more frequently among patients with massive PE than with submassive PE (42% versus 26%) [11]. Pulmonary edema occurred in 4% of the patients with PE and no prior cardiopulmonary disease [6]. Alveolar pulmonary edema occurred in only 1% and interstitial edema occurred in 3%. Among 25 patients with massive acute PE that involved 50% or more of the pulmonary arterial tree, 96% of whom had no prior cardiopulmonary disease, Kerr and associates identified zones of focal oligemia in all patients [12]. Other findings were hyperemic zones (40%), pulmonary infarction (56%), hilar enlargement (56%), elevated hemidiaphragm (32%), and enlarged pulmonary trunk (12%). Among 41 patients (the number with prior cardiopulmonary disease was unstated), Moses and associates observed that the most common signs on the plain chest radiograph were infiltrate (54%), pleural effusion (51%), and atelectasis (27%) [13]. An elevated hemidiaphragm was observed only in 17%. Focal oligemia was uncommon (2%) as was a pattern of pulmonary hypertension (5%). A normal plain chest radiograph was uncommon (7%). Among 126 patients who had PE (the number with prior cardiopulmonary disease was unstated), Laur [14] reported an elevated hemidiaphragm in 74% of severe cases and 50% in mild cases of PE. Dilatation of the hilar vessels was shown in 77% of the severe cases and 48% of the mild cases of PE. Dilatation of pul-

PART III

Diagnosis of acute PE

monary artery was noted in 57% of severe cases and 22% of mild cases of PE. Among 38 patients with PE, 5 of whom had heart failure, Wiener and associates [15] observed an elevated hemidiaphragm in 32% and changes suggestive of pleuritis in 45%. A negative plain chest radiograph was observed in 21%. Among 29 patients with PE, Follath and associates [16] reported dilatation of the right pulmonary artery in 59% and dilatation of the left pulmonary artery in 45%. Dilatation of the main pulmonary artery was observed in fewer, 27%. Infiltrates occurred in 45% and oligemia in 17%.

Radiographic abnormalities related to pulmonary artery mean pressure Patients with no prior cardiopulmonary disease and PE, who had normal plain chest radiographs, had the lowest pulmonary artery mean pressures (Table 46.3) [17]. Patients with normal plain chest radiographs had pulmonary artery mean pressures lower than in patients with an elevated diaphragm, pleural-based opacity, decreased pulmonary vascularity, prominent central pulmonary artery, Westermarkâ&#x20AC;&#x2122;s sign, or cardiomegaly. Patients with normal chest radiographs also had pulmonary artery mean pressures lower than in patients with atelectasis or patients with a pleural effusion. The highest pulmonary artery mean pressures were in patients with no prior cardiopulmonary disease who had PE that ranged in severity from mild to massive with a prominent central pulmonary artery or cardiomegaly (Table 46.3) [17]. Patients with a prominent central pulmonary artery had pulmonary artery mean pressures higher than in patients with atelectasis or a pulmonary parenchymal abnormality or with a pleural effusion. Patients with cardiomegaly also had pulmonary artery mean pressures higher than in patients with atelectasis or a pulmonary parenchymal abnormality or with a pleural effusion. Patients with cardiomegaly also had pulmonary artery mean pressures higher than in patients with a pleural-based opacity, elevated diaphragm, or decreased pulmonary vascularity. Patients with cardiomegaly and a dilated central pulmonary artery had pulmonary artery mean pressures higher than in


BLUK077-Stein

March 12, 2007

CHAPTER 46

16:22

219

Plain chest radiograph

Table 46.3 Pulmonary artery mean pressure and chest radiographic abnormalities among 123 patients with acute pulmonary embolism and no prior cardiopulmonary disease. Number (%)

PA mean pressure (mm Hg)*

Normal

10 (8)

16 ± 5

Atelectasis or pulmonary parenchymal abnormality

94 (76)

21 ± 9

Pleural effusion

71 (58)

21 ± 10

Pleural-based opacity

45 (37)

23 ± 10

Elevated diaphragm

33 (27)

23 ± 8

Decreased pulmonary vascularity

26 (21)

24 ± 11

Prominent central pulmonary artery

20 (16)

30 ± 14

Cardiomegaly

15 (12)

34 ± 13

Westermark’s sign†

7 (6)

31 ± 11

Pulmonary edema

5 (4)

25 ± 10

*Values are mean ± standard deviation. † Prominent central pulmonary artery and decreased pulmonary vascularity. PA, pulmonary artery. Reprinted from Stein et al. [17], with permission from Elsevier.

patients with a prominent central pulmonary artery not accompanied by cardiomegaly. Pulmonary infarction or pulmonary hemorrhage (defined as the abrupt onset of pleuritic pain with or without hemoptysis and infiltrate on the chest radiograph) was uncommon when emboli obstructed the central pulmonary artery [18]. However, pulmonary infarction was frequent when distal arteries were occluded [18]. Presumably,

those with central pulmonary artery occlusion had higher pulmonary artery pressures. Radiographic abnormalities, related to the partial pressure of oxygen in arterial blood (Pa O2 ) and the alveolar–arterial oxygen gradient, of patients with PE that ranged in severity from mild to massive and who had no prior cardiopulmonary disease are shown in Table 46.4 [17].

Table 46.4 Partial pressure of oxygen in arterial blood alveolar–arterial oxygen gradient and chest radiographic abnormalities among 93 patients with acute pulmonary embolism and no prior cardiopulmonary disease. Number (%)

Pa O2 (mm Hg)*

A–a gradient (mm Hg)*

Normal

10 (11)

75 ± 14

30 ± 16

Atelectasis or pulmonary parenchymal abnormality

71 (76)

70 ± 13

36 ± 15

Pleural effusion

53 (57)

72 ± 15

35 ± 17

Pleural-based opacity

31 (33)

70 ± 15

37 ± 14

Elevated diaphragm

25 (27)

67 ± 12

39 ± 14

Decreased pulmonary vascularity

16 (17)

69 ± 15

39 ± 17

Prominent central pulmonary artery

11 (12)

61 ± 11

44 ± 16

Cardiomegaly

9 (10)

61 ± 16

48 ± 22

Westermark’s sign†

4 (4)

60 ± 14

41 ± 18

Pulmonary edema

2 (2)

55 ± 4

58 ± 2

*Values are mean ± standard deviation. † Prominent central pulmonary artery and decreased pulmonary vascularity. Pa O2 , partial pressure of oxygen in arterial blood; A–a = alveolar–arterial. Reprinted from Stein et al. [17], with permission from Elsevier.


BLUK077-Stein

March 12, 2007

16:22

220

References 1 Westermark N. On the Roentgen diagnosis of lung embolism. Acta Radiol 1938; 19: 357–372. 2 Hampton AO, Castleman B. Correlation of postmortem chest teleroentgenograms with autopsy findings: with special reference to pulmonary embolism and infarction. Am J Roentgenol 1940; 43: 305–326. 3 Fleischner F, Hampton AO, Castleman B. Linear shadows in the lung (interlobar pleuritis, atelectasis, and healed infarction). Am J Roentgenol 1941; 46: 610– 618. 4 Fleischner FG. Pulmonary embolism. Clin Radiol 1962; 13: 169–182. 5 Kelly MJ, Elliot LP. The radiologic evaluation of the patient with suspected pulmonary thromboembolic disease. Med Clin North Am 1974; 59: 3–36. 6 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 7 Greenspan RH, Ravin CE, Polansky SM, McLoud TC. Accuracy of the chest radiograph in diagnosis of pulmonary embolism. Invest Radiol 1982; 17: 539– 543. 8 Stein PD, Saltzman HA, Weg JG. Clinical characteristics of patients with acute pulmonary embolism. Am J Cardiol 1991; 68: 1723–1724. 9 National Cooperative Study: the Urokinase Pulmonary Embolism Trial: associated clinical and laboratory findings. Circulation 1973; 47/48(suppl II): 81–85.

PART III

Diagnosis of acute PE

10 National Cooperative Study. The Urokinase Pulmonary Embolism Trial: clinical and electrocardiographic observations. Circulation 1973; 47/48(suppl II): 60–65. 11 Stein PD, Willis PW, III, DeMets DL, Greenspan RH. Plain chest roentgenogram in patients with acute pulmonary embolism and no pre-existing cardiac or pulmonary disease. Am J Noninvas Cardiol 1987; 1: 171–176. 12 Kerr IH, Simon G, Sutton GC. The value of the plain radiograph in acute massive pulmonary embolism. Br J Radiol 1971; 44: 751–757. 13 Moses DC, Silver TM, Bookstein JJ. The complementary roles of chest radiography, lung scanning, and selective pulmonary angiography in the diagnosis of pulmonary embolism. Circulation 1974; 49: 179–188. 14 Laur A. Roentgen diagnosis of pulmonary embolism and its differential from myocardial infarction. Am J Roentgenol 1963; 90: 632–637. 15 Wiener SN, Edelstein J, Charms BL. Observations on pulmonary embolism and the pulmonary angiogram. Am J Roentgenol 1966; 98: 859–873. 16 Follath F, Burkart F, Fridrich R. On evaluation of the thoracic picture in acute pulmonary embolism. Schweiz Med Wochenschr 1968; 98: 1589–1592. 17 Stein PD, Athanasoulis C, Greenspan RH, Henry JW. Relation of plain chest radiographic findings to pulmonary arterial pressure and arterial blood oxygen levels in patients with acute pulmonary embolism. Am J Cardiol 1992; 69: 394–396. 18 Dalen JE, Haffajee CI, Alpert JS, Howe JP, III, Ockene IS, Paraskos JA. Pulmonary embolism, pulmonary hemorrhage, and pulmonary infarction. N Engl J Med 1977; 296: 1431–1435.


BLUK077-Stein

March 11, 2007

47

8:32

CHAPTER 47

Arterial blood gases and the alveolarâ&#x20AC;&#x201C;arterial oxygen difference in acute pulmonary embolism

The partial pressure of oxygen in arterial blood (Pa O2 ) when low in patients with suspected acute pulmonary embolism (PE) has been shown to be a helpful adjunct in the diagnostic assessment [1]. It is recognized, however, that patients with acute PE may have a normal Pa O2 [2â&#x20AC;&#x201C;4]. Among 132 patients with acute PE and

no prior cardiopulmonary disease who had measurements of the Pa O2 while breathing room air, 24% had a Pa O2 of 80 mm Hg or higher (Figure 47.1) (P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990). These data were obtained from patients in the randomized arm of the Prospective Investigation of Pulmonary Embolism Diagnosis I (PIOPED I) and in the referred arm [5]. In PIOPED II, the Pa O2 while breathing room

Figure 47.1 Distribution of partial pressure of oxygen in arterial blood (Pa O2 ) and cumulative frequency while breathing room air among 132 patients with angiographically proven pulmonary embolism and no preexisting cardiac or pulmonary disease. The number of

patients with indicated values of the Pa O2 is shown on the left and the cumulative percent of patients with the indicated values of the Pa O2 is shown of the right. (Data from P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990.)

Arterial blood gases

221


BLUK077-Stein

March 11, 2007

8:32

222

PART III

Diagnosis of acute PE

Table 47.1 Arterial blood gases and alveolar–arterial oxygen difference while breathing room air (PIOPED II). n (%) Probability PE

Pa O2 (mm Hg) ≤49

Probability PE

PE no Prior

No PE no prior

vs. No PE (No

PE all patients

No PE all patients

vs. No PE

CPD ( N = 48)

CPD ( N = 88)

Prior CPD)

( N = 74)

( N = 186)

(all patients)

1 (2)

2 (2)

NS

4 (5)

17 (9)

50–59

6 (13)

12 (14)

NS

12 (16)

32 (17)

NS

60-69

15 (31)

14 (16)

≤0.05

20 (27)

35 (19)

NS

8 (17)

13 (15)

NS

14 (19)

32 (17)

NS

18 (38)

47 (53)

NS

24 (32)

70 (38)

NS

70–79 ≥80 Pa CO2 (mm Hg) ≤35 36–39 ≥40

NS

30 (63)

39 (44)

≤0.05

42 (57)

65 (35)

≤0.01

12 (25)

17 (19)

NS

18 (24)

39 (21)

NS

6 (13)

32 (36)

≤0.01

14 (19)

82 (44)

≤0.001

≤0.05

pH (units) <7.35

0 (0)

7 (8)

0 (0)

13 (7)

≤0.025

7.35–7.45

29 (60)

60 (68)

NS

41 (55)

131 (70)

≤0.025

>7.45

19 (40)

21 (24)

NS

33 (45)

42 (23)

≤0.001

A–a O2 difference (mm Hg) ≤20

17 (35)

44 (50)

NS

24 (32)

70 (38)

NS

21–30

4 (8)

10 (11)

NS

5 (7)

32 (17)

≤0.05

31–40

11 (23)

13 (15)

NS

18 (24)

30 (16)

NS

41–50

9 (19)

13 (15)

NS

14 (19)

32 (17)

NS

51–60

5 (10)

6 (7)

NS

10 (14)

17 (9)

NS

≥61

2 (4)

2 (2)

NS

3 (4)

5 (3)

NS

CPD, cardiopulmonary disease; PE, pulmonary embolism; NS, not significant. Reprinted from Stein et al. [6], with permission.

air was measured in only 48 patients with PE and no prior cardiopulmonary disease [6] (Table 47.1). The Pa O2 was 80 mm Hg or higher in 38% of these patients [6]. Even among patients with submassive or massive acute PE, 12% had a Pa O2 of 80 mm Hg or higher [2]. The distribution of the Pa O2 among 88 randomized patients with PE who had no prior cardiopulmonary disease was comparable to the distribution of the Pa O2 among 202 such patients in whom PE was suspected, but excluded [3]. Among 277 patients with acute PE, 53% of whom had prior cardiopulmonary disease, the Pa O2 was 80 mm Hg or higher in 19% [7]. Clearly, acute PE cannot be excluded on the basis of a normal Pa O2 . Typically, however, among patients with acute PE, the Pa O2 is low. The distribution of the partial pressure of carbon dioxide in arterial blood (Pa CO2 ) among 132 patients with acute PE and no prior cardiopulmonary disease who had measurements while breathing room air is shown in Figure 47.2. The Pa CO2 was less than

35 mm Hg in 49% of these patients, and it was greater than 45 mm Hg in 2% of the patients (P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990). Among 48 patients in PIOPED II with PE and no prior cardiopulmonary disease, the Pa CO2 was ≤35 mm Hg in 65% and ≥40 mm Hg in 13% (6). The pH of arterial blood was greater than 7.44 in 53% of such patients in PIOPED I and it was less than 7.38 in 4% of patients (Figure 47.3) (P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990). In PIOPED II, more of the patients with PE and no prior cardiopulmonary disease had an arterial blood pH less than 7.35 and 40% had a pH greater than 7.45 (6).

Alveolar–arterial oxygen difference A normal alveolar–arterial oxygen difference (alveolar–arterial oxygen gradient) does not exclude acute PE [3, 6, 7, 8]. The distribution of values of


BLUK077-Stein

March 11, 2007

8:32

100

30

90 25

80

60 50

15

Cumulative frequency (%)

Number of patients

70 20

40 10

30 20

5 10 0

0 25

28

31

34

37

40

43

47

PaCO2 (mm Hg) Figure 47.2 Distribution of partial pressure of carbon dioxide in arterial blood (Pa CO2 ) and cumulative frequency while breathing room air among 132 patients with

angiographically proven pulmonary embolism and no prior cardiopulmonary disease. (Data from P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990.)

100

60

50 80

Number of patients

60 30 40 20

20 10

0

0 7.17

7.21

7.25

7.29

7.33

7.37

7.41

7.45

7.49

7.53

7.59

pH Figure 47.3 Distribution of pH of arterial blood and cumulative frequency while breathing room air among 132 patients with angiographically proven PE and no prior

cardiopulmonary disease. (Data from P.D. Stein and B. Relyea, unpublished data from PIOPED, 1990.)

223

Cumulative frequency (%)

40


BLUK077-Stein

March 11, 2007

8:32

224

PART III

35

PE (n = 88)

30 Percent of patients

Diagnosis of acute PE

No PE (n = 202)

25 20 15 10 5 0 <10

10−20 21−30 31−40 41−50 51−60 >61 Alveolar-arterial oxygen gradient (mm Hg)

the alveolar–arterial oxygen differences in patients with acute PE and in patients in whom acute PE was excluded is shown in Figure 47.4 [3]. No value of the alveolar–arterial oxygen difference was diagnostic of PE and no value excluded PE. Subsequent to the publication of our observations on the alveolar–arterial oxygen difference in patients with no prior cardiopulmonary disease, there has been rekindling of interest in the ability of a normal alveolar– arterial oxygen difference to exclude PE. McFarlane and Imperiale indicated that only 1.9% of the patients with a normal alveolar–arterial oxygen difference and no history of prior PE or prior DVT (deep venous thrombosis) had PE [9]. They concluded, therefore, that a normal alveolar–arterial oxygen difference among patients with no prior PE or prior DVT makes the diagnosis of PE so unlikely that further diagnostic evaluation may be unnecessary in this subgroup. Because the alveolar–arterial oxygen difference seemed to be more useful in the experience of these investigators than what our data showed, we evaluated further the value of the alveolar–arterial oxygen difference among patients in PIOPED [7]. We used more than one definition of a normal alveolar–arterial oxygen difference. We stratified according to categories of patients in whom a normal alveolar–arterial oxygen difference is thought by some to assist in excluding the diagnosis of PE. We used an expanded database from PIOPED that included patients randomized for obligatory pulmonary angiography and patients referred for pulmonary angiography by their physicians.

Figure 47.4 Distribution of alveolar–arterial oxygen gradient among 88 patients with angiographically proven pulmonary embolism (PE) and no preexisting cardiac or pulmonary disease (solid bars) and 202 such patients in whom PE was excluded (striped bars). (Reprinted with permission from Stein et al. [3].)

The alveolar–arterial oxygen difference was calculated according to the following equation [10]: Alveolar–arterial oxygen difference = 150 − 1.25Pa CO2 − Pa O2 where Pa CO2 is the partial pressure of carbon dioxide in arterial blood (mm Hg) and Pa O2 is the partial pressure of oxygen in arterial blood (mm Hg). Measurements of arterial blood gases were obtained with the patient breathing room air. All measurements were obtained within 24 hours prior to the diagnostic pulmonary angiogram. This may have been as long as 48 hours after the onset of symptoms, although usually it was less than 36 hours after the onset of symptoms [5]. The normal alveolar–arterial oxygen difference increases with age. Most patients have an alveolar– arterial oxygen difference ≤20 mm Hg [11–14]. Normal age-related values obtained from data reported by Mellemgaard [11] are as follows: Patients 15–19 years of age had an alveolar–arterial oxygen difference ≤14 mm Hg; patients 20–29 years of age had an alveolar–arterial oxygen difference ≤20 mm Hg; and patients ≥30 years of age (with one exception) had an alveolar–arterial oxygen difference ≤27 mm Hg [11]. These values correspond closely with normal agerelated values of the alveolar–arterial oxygen difference reported by Harris and associates [12] and Kanber and associates [14]. Others reported somewhat lower normal values of the alveolar–arterial oxygen difference,


BLUK077-Stein

March 11, 2007

CHAPTER 47

8:32

225

Blood gases and alveolar–arterial oxygen difference

which did not exceed 20 mm Hg in subjects 23–60 years of age [13]. Normal values of the alveolar–arterial oxygen difference were also approximated by the equation age/4 + 4 [9, 15]. This equation is restrictive, however. It would have excluded 28% of the normal patients reported by Mellemgaard [11], 20% reported by Kanber and associates [14], 27% reported by Harris and associates [12], and 16% reported by Filley and associates [13]. Averaging these data, the age-corrected equation would incorrectly define as abnormal 19% of the patients [7]. Irrespective of how a normal alveolar–arterial oxygen difference was defined, comparable percentages of patients with normal alveolar–arterial oxygen differences had PE or no PE [7]. This applied to all patients with PE including patients with prior cardiopulmonary disease. It applied to patients stratified according to the absence of prior cardiopulmonary disease and to patients stratified according to the absence of prior PE or prior DVT.

All patients Among 280 patients with acute PE in PIOPED I, including patients with prior cardiopulmonary disease as well as patients with prior PE and/or prior DVT, 12% had an alveolar–arterial oxygen difference ≤20 mm Hg versus 16% among patients who did not have PE (Table 47.2) [7]. Using the same method in PIOPED II, among 74 patients with PE the alveolar-arterial oxygen difference was ≤20 mmHg in 32% (6). Among

patients with a normal alveolar–arterial oxygen difference based on reported age-related values, 20% with PE had a normal alveolar–arterial oxygen difference versus 25% among patients who did not have PE. If a normal alveolar–arterial oxygen difference was defined on the basis of the equation age/4 + 4, then 8% had a normal alveolar–arterial oxygen difference versus 12% who did not have PE. None of these comparisons of patients with PE to those without PE showed statistically significant differences.

Patients with no prior cardiopulmonary disease Among 130 patients with no prior cardiopulmonary disease in PIOPED I, 14% with an alveolar–arterial oxygen difference ≤ 20 mm Hg had acute PE versus 20% among patients who did not have PE (Table 47.2) [7]. Using the same method in PIOPED II, among 48 patients with PE and no prior cardiopulmonary disease, the alveolar-arterial oxygen difference was ≤ 20 mmHg in 35% (6). Among patients with a normal alveolar–arterial oxygen based on reported age-related values, 23% with PE had a normal alveolar–arterial oxygen difference versus 27% with no PE. Using the age/4 + 4 equation to define normal values of the alveolar–arterial oxygen difference, 10% with PE had normal values of the alveolar–arterial oxygen difference versus 12% among patients who did not have PE. None of these comparisons of patients with PE to those without PE showed statistically significant differences.

Table 47.2 Frequency of normal A–a gradients among patients with pulmonary embolism and patients with no pulmonary embolism. (PIOPED I). n/N (%) A–a gradient

A–a gradient

A–a gradient normal =

normal ≤20 mm Hg

normal ≤ Age/4 +

Age-related data

(A–a ≤ N1)

4 (A–a ≤ N1)

(A–a ≤ N1)

PE

No PE

PE

No PE

PE

No PE

All patients

33/280 (12)

82/499 (16)

23/280 (8)

59/499 (12)

57/280 (20)

123/499 (25)

No prior CPD

18/130 (14)

40/200 (20)

13/130 (10)

24/200 (12)

30/130 (23)

54/200 (27)

No prior PE or DVT

21/190 (11)

55/365 (15)

15/190 (8)

40/365 (11)

38/190 (20)

86/365 (24)

All differences (PE vs. no PE) not significant. CPD, cardiopulmonary disease; DVT, deep vein thrombosis; Nl, normal; Pts, patients. Reprinted with permission from Stein et al. [7].


BLUK077-Stein

March 11, 2007

8:32

226

PART III

Among 190 patients with no prior PE or prior DVT, 11% with acute PE had an alveolar–arterial oxygen difference ≤ 20 mm Hg versus 15% who did not have PE (Table 47.2) [7]. Among patients with a normal alveolar–arterial oxygen difference based on reported age-related values, 20% had PE and 24% did not have PE. Based on the equation for age-corrected values of the alveolar–arterial oxygen difference, 8% with PE had a normal alveolar–arterial oxygen difference versus 11% with no PE. None of these comparisons of patients with PE to those without PE showed statistically significant differences.

Relation of alveolar–arterial oxygen difference to the Pa O2

A–a gradient (mm Hg)

The alveolar–arterial oxygen difference showed a strong inverse linear correlation with the Pa O2 among patients with PE and no prior cardiopulmonary disease (Figure 47.5) [7]. The alveolar–arterial oxygen difference also showed a strong inverse linear correlation with the Pa O2 and patients with PE who had prior cardiopulmonary disease (Figure 47.6). This indicates that the value of both tests for screening for acute PE is similar. Neither values of the alveolar–arterial oxygen difference nor values of the Pa O2 can be used to exclude PE.

100 90 80 70 60 50 40 30 20 10 0 −10

No CPD

A−a gradient (mm Hg)

Patients with no prior PE or prior DVT

Diagnosis of acute PE

PaO2 (mm Hg) Figure 47.5 Alveolar–arterial (A–a) oxygen difference shown as a function of the partial pressure of oxygen in arterial blood (Pa O2 ) among patients with no prior cardiopulmonary disease (No CPD). Correlation coefficient r = −0.917 (P < 0.005). (Reprinted with permission from Stein et al. [7].)

Relation of alveolar–arterial oxygen difference to pulmonary pressure and to the ventilation–perfusion lung scan Patients with less severe PE were more likely to have a normal alveolar–arterial oxygen difference than those with more severe PE. The alveolar–arterial oxygen difference showed a loose linear correlation with the pulmonary artery mean pressure among patients with PE and no prior cardiopulmonary disease (r = 0.32, P < .005) and

CPD

20

40

60 80 PaO2 (mm Hg)

100

120

Figure 47.6 Alveolar–arterial (A–a) oxygen difference shown as a function of the partial pressure of oxygen in arterial blood (Pa O2 ) among patients with any prior cardiopulmonary disease (CPD). Correlation coefficient r = −0.842 (P < 0.005). (Reprinted with permission from Stein et al. [7].)


BLUK077-Stein

March 11, 2007

CHAPTER 47

8:32

227

Blood gases and alveolar–arterial oxygen difference

Table 47.3 Negative predictive value of blood gases and alveolar–arterial oxygen gradient, alone or in combination: PIOPED I. No PE/Test Neg (%) No prior CPD

Prior CPD

All patients

Pa CO2 ≥35

105/171 (61)

142/211 (67)

247/382 (65)

Pa O2 ≥80

55/87 (63)

62/84 (74)

117/171 (68)

A–a ≤20

40/58 (69)

38/50 (76)

78/108 (72)

Pa CO2 ≥35 and Pa O2 ≥80

30/49 (61)

28/36 (78)

58/85 (68)

Pa CO2 ≥35 and A–a ≤20

30/46 (65)

28/33 (85)

58/79 (73)

Pa O2 ≥80 and A–a ≤20

36/54 (67)

34/45 (76)

70/99 (71)

Pa CO2 ≥35, Pa O2 ≥80, and A–a ≤20

26/42 (62)

24/28 (86)

50/70 (71)

Blood gases and alveolar–arterial (A–a) oxygen differences are in mm Hg. PE, pulmonary embolism; CPD, cardiopulmonary disease. NEG, negative. Reprinted with permission from Stein et al. [17].

among patients with PE who had prior cardiopulmonary disease (r = 0.34, P < .005) [7]. The alveolar–arterial oxygen difference showed a loose linear correlation with the number of mismatched perfusion defects on the ventilation– perfusion lung scan among patients with PE and no prior cardiopulmonary disease (r = 0.40, P < .005) [7]. The alveolar–arterial oxygen difference also showed a loose linear correlation with the number of mismatched perfusion defects on the ventilation– perfusion lung scan among patients with PE who had prior cardiopulmonary disease (r = 0.39, P < .005).

Blood gases in combination with the alveolar-arterial oxygen difference It has been suggested that if both the Pa CO2 and the alveolar–arterial oxygen difference are normal, this can be used as evidence against the presence of PE [16]. This is based on the observation among patients with essentially no prior cardiac or pulmonary disease who had PE that 93% had hypoxemia or hypocapnia and 98% had an increased alveolar–arterial oxygen difference or hypocapnia [16]. We tested this among patients in PIOPED to determine if normal values of one or more blood gases, alone or in combination, could be used to exclude the diagnosis of acute PE [17]. Whether PE is absent if blood gases are normal is shown by the negative predictive value. Based on data

from PIOPED, we were not able to exclude PE based on an absence of abnormal values of the Pa O2 , Pa CO2 , or alveolar–arterial oxygen difference [17]. Among patients with no prior cardiopulmonary disease, PE was absent in 69% or fewer with a Pa O2 ≥80 mm Hg, Pa CO2 ≥35 mm Hg, and/or alveolar–arterial oxygen difference ≤20 mm Hg (Table 47.3). Even with values of the Pa O2 ≥80 mm Hg, Pa CO2 ≥35 mm Hg, and alveolar–arterial oxygen difference ≤20 mm Hg, 62% did not have PE. Conversely, in patients who did not have prior cardiopulmonary disease, PE was present in 38% when the Pa O2 and Pa CO2 were not low and the alveolar–arterial oxygen difference was normal. Among patients with prior cardiopulmonary disease, the highest negative predictive value was shown with the combination Pa O2 ≥80 mm Hg, Pa CO2 ≥35 mm Hg, and alveolar–arterial oxygen difference ≤20 mm Hg (Table 47.3) [17]. With this combination, PE was absent in 86% of patients who had prior cardiopulmonary disease. Conversely, in patients with prior cardiopulmonary disease, PE was present in 14% when the Pa O2 and Pa CO2 were not low and the alveolar–arterial oxygen difference was normal. We observed that 88% of the patients with PE and no prior cardiopulmonary disease had either a low Pa CO2 or a high alveolar–arterial oxygen difference [17]. We also observed that 97% of the patients with prior cardiopulmonary disease who had PE had a low Pa CO2 or a high alveolar–arterial oxygen difference. Pulmonary embolism could not be excluded on the basis of such


BLUK077-Stein

March 11, 2007

8:32

228

blood gases [17]. No combinations of blood gases were identified that reliably excluded pulmonary embolism. Our data indicate that the Pa O2 and the alveolar– arterial oxygen difference may be included among laboratory and clinical findings that typically are abnormal in patients with PE, but may be normal in some patients with PE. Normal values of an alveolar–arterial oxygen difference do not exclude the diagnosis of acute PE.

References 1 Szucs MM, Brooks HL, Grossman W et al. Diagnostic sensitivity of laboratory findings in acute pulmonary embolism. Ann Intern Med 1971; 74: 161–166. 2 The National Cooperative Study: The Urokinase Pulmonary Embolism Trial. Circulation 1973; 47/48(suppl II): II-18–II-21 and II-81–II-85. 3 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 4 Menzoian JO, Williams LF. Is pulmonary angiography essential for the diagnosis of acute pulmonary embolism? Am J Surg 1979; 137: 543–548. 5 A collaborative study by the PIOPED investigators. Value of the ventilation–perfusion scan in acute pulmonary embolism: Results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 6 Stein PD, Beemath A, Weg JG et al. Clinical characteristics of patients with acute pulmonary embolism: data from PIOPED II. Am J Med, submitted.

PART III

Diagnosis of acute PE

7 Stein PD, Goldhaber SZ, Henry JW. Alveolar–arterial oxygen gradient in the assessment of acute pulmonary embolism. Chest 1995; 107: 139–143. 8 Overton DT, Bocka JJ. The alveolar–arterial oxygen gradient in patients with documented pulmonary embolism. Arch Int Med 1988; 148: 1617–1619. 9 McFarlane MJ, Imperiale TF. Use of the alveolar–arterial oxygen gradient in the diagnosis of pulmonary embolism. Am J Med 1994; 96: 57–62. 10 Guenter CA. Respiratory function of the lungs and blood. In: Guenter CA, Welch MH, eds. Pulmonary Medicine, 2nd edn. Lippincott, Philadelphia, 1982: 168. 11 Mellemgaard K. The alveolar–arterial oxygen difference: its size and components in normal man. Acta Physiol Scand 1966; 67: 10–20. 12 Harris EA, Kenyon AM, Nisbet HD, Seelye ER, Whitlock RML. The normal alveolar–arterial oxygen-tension gradient in man. Clin Sci Mol Med 1974; 46: 89– 104. 13 Filley GF, Gregoire F, Wright GW. Alveolar and arterial oxygen tensions and the significance of the alveolar– arterial oxygen tension difference in normal men. J Clin Invest 1954; 33: 517–529. 14 Kanber GJ, King FW, Eshchar YR, Sharp JT. The alveolar– arterial oxygen gradient in young and elderly men during air and oxygen breathing. Am Rev Respir Dis 1968; 97: 376–381. 15 Skorodin MS. Respiratory diseases and A–a gradient measurement. JAMA 1984; 252: 1344. 16 Cvitanic O, Marino PL. Improved use of arterial blood gas analysis in suspected pulmonary embolism. Chest 1989; 95: 48–51. 17 Stein PD, Goldhaber SZ, Henry JW, Miller AC. Arterial blood gas analysis in the assessment of suspected acute pulmonary embolism. Chest 1996; 109: 78–81.


BLUK077-Stein

March 28, 2007

48

15:26

CHAPTER 48

Fever in acute pulmonary embolism

Patients (%)

Low-grade fever based on data from Prospective Investigation of Pulmonary Embolism Diagnosis I (PIOPED I), is not uncommon in pulmonary embolism (PE) and high fever, although rare, may occur [1]. Low-grade fever need not be accompanied by pulmonary hemorrhage/infarction [1]. Temperature was measured on the day of recruitment into PIOPED I [1]. Symptoms of PE began within 24 hours of study entry [2]. Fever was defined as temperature of 100.0◦ F (37.8◦ C) or higher [3, 4]. Although the normal temperature in human beings has been said to be 98.6◦ F (37◦ C) based on Wunderlich’s original observations over 120 years ago, the maximum oral temperature in normal individuals at 4 p.m. is 99.9◦ F (37.7◦ C) [3, 4]. Among all patients, fever [temperature ≥ 100.0◦ F (37.8◦ C)] was present in 95 of 363 (26%) patients with PE. Other definite or possible causes of fever were present in 46 of these patients [1]. Among those with pulmonary hemorrhage/infarction, fever was present in 80 of 308 (26%). Among those who did not have pulmonary hemorrhage/infarction, fever was present in a similar percentage, 15 of 55 (27%). Among patients with PE and no other definite or possible cause of fever, temperature was normal in

100 90 80 70 60 50 40 30 20 10 0

268 of 311 (86%) (Figure 48.1) [1]. Among those with fever, the fever was usually low grade. Temperature of 101◦ F (38.3◦ C) or higher occurred only in 19 of 311 (6%) patients and only 5 of 311 (1.6%) had a temperature of 102◦ F (38.9◦ C) or higher. Some conditions causing fever may not have been listed in the discharge summaries or computer listing of discharge diagnoses. The frequency of patients that we reported with fever caused by PE, therefore, may have been higher than the actual value. The distribution of temperatures according to whether patients had pulmonary hemorrhage/infarction syndrome versus uncomplicated PE or circulatory collapse is shown in Figure 48.2 [1]. Signs or symptoms of deep venous thrombosis (DVT) were present in 24 of 43 patients (56%) with PE who had fever. Among patients with a temperature of 102◦ F (38.9◦ C) or higher, 3 of 5 (60%) had clinical findings compatible with DVT. Each of 4 patients with PE and otherwise unexplained fever who did not have pulmonary hemorrhage or infarction had clinical findings compatible with DVT. In PIOPED I, among patients with PE and no prior cardiopulmonary disease, a temperature higher than 101.3◦ F (38.5◦ C) was reported in 7% [5], but no attempt was made to eliminate patients with other causes

86

8 <100.0

100.0− 100.9

5 101.0− 101.9

1 102.0− 102.9

Temperature (°F)

<1

<1

103.0− 103.9

>103.9

Figure 48.1 Temperature distributions among patients with pulmonary embolism and no other definite or possible cause of fever. (Data from Stein et al. [1].)

229


BLUK077-Stein

March 28, 2007

15:26

230

PART III

90

85

91

80 Patients (%)

Diagnosis of acute PE

Hemorrhage/infarction syndrome

70

Uncomplicated PE or circulatory collapse syndrome

60 50 40 30 20

8 7

10

5

2

0 <100.0

100.0− 100.9

101.0− 101.9

1

0

102.0− 102.9

<1 0

<1 0

103.0− 103.9

>103.9

Temperature (°F)

of fever. Among patients with no prior cardiopulmonary disease who had the pulmonary infarction syndrome, temperature higher than 101.3◦ F (38.5◦ C) was reported in 6% [6] and with the syndrome of isolated dyspnea, 16% were reported with such a temperature elevation [6]. Again, no attempt was made to eliminate patients with other causes of fever. In the Urokinase Pulmonary Embolism Trial and the Urokinase-Streptokinase Embolism Trial, among patients with no prior cardiopulmonary disease, a temperature higher than 99.9◦ F (37.5◦ C) was observed in 50% [7]. The prevalence of fever was the same among patients with massive and with submassive PE. At a tertiary care hospital, among all patients with PE diagnosed by pulmonary angiography between 1980 and 1984, 14% had a temperature higher than 100◦ F (37.8◦ C) [8]. Other causes of fever were not excluded in any of these case series. Murray and associates attempted to clarify whether fever in PE was due to the PE or accompanying disease [9]. Among patients with angiographically diagnosed PE and no other apparent cause of fever, a temperature of 100.4◦ F (38◦ C) or higher was observed in 20 of 31 (64%) and 2 of 31 (6%) had a temperature of 103.1◦ F (39.5◦ C) or higher [9]. Pulmonary infarction was not more frequent among patients with fever [9]. Some reported that unless concomitant pulmonary infection is present, the temperature generally is 101◦ F (38.3◦ C) or less [10, 11]. Higher temperatures were generally observed in patients with extensive pulmonary infarction or in whom secondary pneumonitis developed distal to the embolus [11].

Figure 48.2 The distribution of temperatures according to whether patients had pulmonary hemorrhage/infarction syndrome versus uncomplicated pulmonary embolism (PE) or circulatory collapse. (Data from Stein et al. [1].)

Some have suggested that low-grade fever, especially after an operation, is occasionally the only symptom of PE [12]. Occasional case reports describe PE as a cause of fever of undetermined origin [13, 14].

References 1 Stein PD, Afzal A, Henry JW, Villarreal CG. Fever in acute pulmonary embolism. Chest 2000; 117: 39–42. 2 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 3 Mackowiak PA, Wasserman SS, Levine MM. A critical appraisal of 98.6◦ F, the upper limit of the normal body temperature, and other legacies of Carl Reinhold August Wunderlich. JAMA 1992; 268: 1578–1580. 4 Gelfand JA, Dinarello CA. Alterations in body temperature. In: Fauci AS, Braunwald E, Isselbacher KJ et al., eds. Harrison’s Principles of Internal Medicine, 14th edn. McGraw-Hill Inc., New York, 1998: 84–90. 5 Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. 6 Stein PD, Henry JW. Clinical characteristics of patients with acute pulmonary embolism stratified according to their presenting syndromes. Chest 1997; 112: 974– 979. 7 Stein PD, Willis PW, III, DeMets DL. History and physical examination in acute pulmonary embolism in patients


BLUK077-Stein

March 28, 2007

CHAPTER 48

15:26

Fever in acute PE

without preexisting cardiac or pulmonary disease. Am J Cardiol 1981; 47: 218–223. 8 Leeper KV, Jr, Popovich JP, Jr, Adams DA, Stein PD. The clinical manifestations of acute pulmonary embolism: HFH experience. A five year review. Henry Ford Hosp Med J 1988; 36: 29–34. 9 Murray HW, Ellis GC, Blumenthal DS, Sos TA. Fever and pulmonary thromboembolism. Am J Med 1979; 67: 232– 235. 10 Sasahara AA, Cannilla JE, Morse RL et al. Clinical and physiologic studies in pulmonary thromboembolism. Am J Cardiol 1967; 20: 10–20.

231

11 Sharma GV, Sasahara AA, McIntyre KM. Pulmonary embolism: the great imitator. Dis Mon 1976; 22: 4–38. 12 Hodgson CH. Pulmonary embolism and infarction. Dis Chest 1965; 47: 577–588. 13 Aburahma AF, Saiedy S. Deep vein thrombosis as a probable cause of fever of unknown origin. W V Med J 1997; 93: 368–370. 14 Stallman JS, Aisen PS, Aisen ML. Pulmonary embolism presenting as fever in spinal cord injury patients: report of two cases and review of the literature. J Am Paraplegia Soc 1993; 16: 157–159.


BLUK077-Stein

April 6, 2007

49

16:2

CHAPTER 49

Leukocytosis in acute pulmonary embolism syndrome) had a WBC count of >10,000/mm3 in 32 of 183 (17%) [1]. Among patients who did not have the pulmonary hemorrhage/infarction syndrome, 20 of 83 (24%) had a WBC count of >10,000/mm3 . Among patients with the pulmonary hemorrhage/infarction syndrome, the frequency of an elevated WBC count (16–17%) was the same irrespective of whether the pulmonary hemorrhage/infarction syndrome was characterized by atelectasis/infiltrate on the chest radiograph, hemoptysis, or pleuritic chest pain. The percent of neutrophils in patients with PE and a WBC count of >10,000/mm3 and in those with a WBC count of ≤10,000/mm3 is shown in Table 49.1 [1]. Older clinicians found the WBC count useful in distinguishing between pneumonia and pulmonary infarction [2]. Pulmonary infarction was considered to be more likely than pneumonia if the WBC count was <15,000/mm3 [2]. Sasahara showed that among patients with PE, not necessarily with the pulmonary hemorrhage/infarction syndrome, the WBC count was in the range of 15,000–25,000/mm3 in only 1 of 20 (5%) [2]. Others evaluated leukocytosis with pulmonary infarction. Among 200 patients with

Leukocytosis is uncommon in acute pulmonary embolism (PE) [1]. The frequency of an elevated white blood cell (WBC) count was not greater in patients with the pulmonary hemorrhage/infarction syndrome than in patients with acute PE who did not have pulmonary hemorrhage/infarction [1]. One reason that the prevalence of leukocytosis in acute PE has been unclear, in spite of an extensive number of investigations of PE over a period of several years, may relate to the fact that acute PE is usually associated with other conditions, which themselves may cause leukocytosis. Based on a detailed review of records from a tertiary care hospital, among all patients with PE, the WBC count was ≤10,000/mm3 in 214 of 386 (55%) and >10,000/mm3 in 172 of 386 (45%) [1]. Among patients in whom other definite causes for leukocytosis were eliminated, 214 of 324 (66%) had a WBC count of ≤10,000/mm3 and 110 of 324 (34%) had a WBC count of >10,000/mm3 [1]. The distribution of the WBC count in patients with PE and no other cause for WBC count elevation is shown in Figure 49.1. Among patients with no other cause for a WBC count elevation, those with pulmonary hemorrhage/infarction syndrome (hemoptysis/pleuritic pain 80

80

PE (%)

60 40 20

7

8

3

1

1

0

0 <10.1

10.1− 12.0

12.1− 14.0

14.1− 16.0

16.1− 18.0

WBC × 103/mm3

232

18.1− 20.0

>20

Figure 49.1 The distribution of white blood cell (WBC) count in patients with pulmonary embolism (PE) and no other cause for WBC count elevation. (Reproduced with minor modifications from Afzal et al. [1].)


BLUK077-Stein

April 6, 2007

CHAPTER 49

16:2

233

Leukocytosis in acute PE

Table 49.1 Distribution of percent neutrophils in patients with PE according to the presence or absence of leukocytosis. No. of patients (%) Neutrophils (%)

WBC >10,000/mm3 ( N = 27)

40–50

WBC ≤10,000/mm3 ( N = 81) 10 (12.3)

61–70

5 (18.5)

24 (29.6)

71–80

14 (51.8)

20 (24.6)

81–90

6 (22.2)

13 (16.0)

>90

1 (3.7)

2 (2.4)

WBC = white blood cells. Reproduced with minor modifications from Afzal et al. [1].

pulmonary hemorrhage/infarction shown at autopsy, the WBC count was usually between 10,000 and 15,000/mm3 [3]. In patients with pulmonary infarction accompanied by pneumonia or patients with large pulmonary infarctions, the WBC count was reported to be “around” 20,000/mm3 . In two patients a WBC count of 28,000/mm3 and a count of 48,000/mm3 were observed. Among patients with pulmonary infarction at autopsy, Miller and Berry observed a WBC count of >10,000/mm3 in 79 of 104 (76%) [4]. The WBC count was 15,000–20,000/mm3 in 34 of 104 (33%). It was >20,000/mm3 in 17 of 104 (16%).

References 1 Afzal A, Noor HA, Gill SA, Brawner C, Stein, PD. Leukocytosis in acute pulmonary embolism. Chest 1999; 115: 1329–1332. 2 Sasahara AA, Stein M. PulmonaryEmbolic Disease. Grune & Straton, New York, 1965: 258–259. 3 Gsell O. Der hamorrhagische lungeninfarct und seine komplikationen (Infarktpleuritis, infarktpnoumonie, infarktkaverne usw). Deut Med Wchschr 1935; 61:1317– 1360. 4 Miller R, Berry JB. Pulmonary infarction: a frequently missed diagnosis. Am J Med Sci 1951; 222: 197–206.


BLUK077-Stein

March 28, 2007

50

15:30

CHAPTER 50

Alveolar dead-space in the diagnosis of pulmonary embolism

Alveolar dead-space volume occurs in areas of the lung that are ventilated, but not perfused and that contain a very low partial pressure of carbon dioxide (PCO2 ) [1]. Exhaled dead-space volume dilutes the total amount of CO2 in exhaled breaths relative to the arterial partial pressure of CO2 (Pa CO2 ). Therefore, the alveolar dead-space volume can be estimated by simultaneously measuring CO2 in exhaled breaths and Pa CO2 [1]. Physiological dead-space fraction = VDphys /VT = (Pa CO2 − PE CO2 )/Pa CO2 where PE CO2 is the partial pressure of mixed expired CO2 . VDphys = physiological dead-space and VT = tidal volume. Because the physiological dead-space fraction (VDphys /VT ) measures the fraction of each tidal volume that is wasted on alveolar and airway dead-space, the airway dead-space fraction was subtracted from the physiological dead-space fraction. This yielded the alveolar dead-space fraction (VDalv ). Multiplication by 100 gives the fraction as a percentage. The airway deadspace volume was calculated from geometric analysis of the CO2 volume curve [2, 3]. Normal alveolar deadspace was defined as ≤20% [1]. Some calculate “late dead-space fraction” [3]. This represents the dead-space fraction of CO2 at the end of a hypothetical expiration with a volume equal to 15% of predicted total lung capacity [3]. Volumetric capnograms were obtained at the bedside using a commercially available volumetric-based capnography machine, which can simultaneously record time-based capnograms, expired CO2 content, breath volume, and flow data. Measurement of the alveolar dead-space volume with volumetric capnography and blood gas analysis can be completed in less than 5 minutes [2]. The most timeconsuming portion of the measurement was the col-

234

lection and processing of arterial blood for blood gas analysis [2]. The end-tidal to arterial PCO2 gradient as well as physiological dead-space is often increased in obstructive lung disease and interstitial lung disease [3]. These techniques, therefore, suffer an inherent limited ability to differentiate between pulmonary embolism (PE) and other pulmonary disease [3]. Efforts to increase the specificity of measurement include measurements after a deep expiration, which reduces the arterialend expiratory gradient in obstructive lung disease, but not in PE [3, 4]. Others suggested that measurements could be made in combination with spirometry [3, 5]. Physiological dead-space fraction, alveolar deadspace fraction, and late dead-space fraction are shown in Table 50.1. The sensitivities and specificities varied widely [1, 3, 5–7]. On the other hand, alveolar dead-space fraction [1, 6, 7] or a change in alveolar dead-space fraction [1] when used in combination with D-dimer had a high sensitivity, but it was not specific (Table 50.2). The combination has a high negative predictive value (Table 50.2). Alveolar dead-space fraction, therefore, in the opinion of the authors, may be useful for exclusion of PE when used in combination with a negative D-dimer [1]. The alveolar dead-space fraction showed a correlation with systolic pulmonary artery pressure and an approximate correlation with the percent perfusion defect [2]. Among patients in whom bedside evaluation showed any two of the following—alveolar deadspace fraction ≤15%, D-dimer normal, or Wells score “unlikely”, PE was safely excluded in a higher proportion of patients than by only an “unlikely” Wells score plus a normal D-dimer [8]. If any two of these assessments was normal or “unlikely,” PE was excluded in 34% of the patients in whom the diagnosis was suspected, and venous thromboembolism (VTE) on a 3month follow-up occurred in only 2.4% [8]. If clinical


BLUK077-Stein

March 28, 2007

CHAPTER 50

15:30

235

Alveolar dead-space in diagnosis of PE

Table 50.1 Physiological dead-space fraction, alveolar dead-space fraction, and late dead-space fraction in pulmonary embolism. n/N (%) First author (Ref.)

Test

Cutoff value

Sensitivity

Specificity

Burki [5]

VDphys /VT

>0.4

16/16 (100)

16/29 (55)

Eriksson [3]

VDphys /VT

>0.4

8/9 (89)

22/29 (76)

Eriksson [3]

fDlate

>0.2

6/9 (67)

Kline [6]

VDalv /VT

0.2

23/26 (88)

Johanning [7]

VDalv /VT

>0.45

Kline [1]

VDalv /VT *

>0.2

4/7 (57)* 43/64 (67%)

29/29 (100) 135/144 (94) 14/14 (100) 241/316 (76)

*VDalv /VT calculated as VDphys /VT – airway dead-space fraction. VDphys , physiological dead-space; VT , tidal volume; fDlate , late dead-space fraction; VDalv /VT , alveolar dead-space fraction, = (Pa CO2 – end-tidal CO2 )/Pa CO2 ). Table 50.2 Dead-space in combination with D-dimer in pulmonary embolism. Negative First author (Ref.)

Test

Index value

D-dimer

Sensitivity

Kline [6]

VDalv /VT

≥0.2

Semiquant latex agglut

26/26 (100)*

Johanning [7]

VDalv /VT

>0.45

Semiquant latex agglut

Kline [1]

VDalv /VT‡

>0.2

Whole blood agglut

4/5 (80%) 63/64 (98)§

Specificity

95/144 (66)† 16/16 (100%) 163/316 (52)†

predictive value 95/95 (100) 16/17 (94) 163/164 (99)

*Both tests positive. † Both tests negative. ‡ PE positive required a dead-space level of >0.45 or a change in dead-space of >0.10 from control values before symptoms in combination with a positive D-dimer. PE negative required a dead-space level of <0.45 and a change in dead-space of <0.10 in combination with a negative D-dimer. § Either test positive. Semiquant, semiquantitative; agglut, agglutination.

assessment plus D-dimer alone were used, PE was excluded in 18% [8].

References 1 Kline JA, Israel EG, Michelson EA, O’Neil BJ, Plewa MC, Portelli DC. Diagnostic accuracy of a bedside D-dimer assay and alveolar dead-space measurement for rapid exclusion of pulmonary embolism: a multicenter study. JAMA 2001; 285: 761–768. 2 Kline JA, Kubin AK, Patel MM et al. Alveolar dead space as a predictor of severity of pulmonary embolism. Acad Emer Med 2000; 7: 611–617. 3 Eriksson L, Wollmer P, Olsson CG et al. Diagnosis of pulmonary embolism based upon alveolar dead space analysis. Chest 1989; 96: 357–362. 4 Hatle L, Rokseth R. The arterial to end-expiratory carbon dioxide tension gradient in acute pulmonary embolism and other cardiopulmonary diseases. Chest 1974; 66: 352– 357.

5 Burki NK. The dead space to tidal volume ratio in the diagnosis of pulmonary embolism. Am Rev Respir Dis 1986; 133: 679–685. 6 Kline JA, Meek S, Boudrow D, Warner D, Colucciello S. Use of the alveolar dead space fraction (Vd/Vt) and plasma D-dimers to exclude acute pulmonary embolism in ambulatory patients. Acad Emerg Med 1997; 4: 856– 863. 7 Johanning JM, Veverka TJ, Bays RA, Tong GK, Schmiege SK. Evaluation of suspected pulmonary embolism utilizing end-tidal CO2 and D-dimer. Am J Surg 1999; 178: 98–102. 8 Rodger MA, Bredeson CN, Jones G et al. The bedside investigation of pulmonary embolism diagnosis study: a double-blind randomized controlled trial comparing combinations of 3 bedside tests vs ventilation– perfusion scan for the initial investigation of suspected pulmonary embolism. Arch Intern Med 2006; 166: 181– 187.


BLUK077-Stein

March 12, 2007

51

18:30

CHAPTER 51

Neural network computer-assisted diagnosis

We attempted to determine if objective clinical characteristics that can be readily identified by physicians without special knowledge of pulmonary embolism (PE) could be evaluated by computer to achieve a correct estimate of the likelihood of PE [1]. To accomplish this, we used a neural network [1]. A neural network is an artificial intelligence paradigm that integrates multiple clinical variables to arrive at a correct diagnosis. It has been used for clinical problems such as myocardial infarction and low back disorders [2, 3]. Neural network paradigms also have

been applied to the evaluation of ventilation–perfusion lung scans, with good results [4, 5]. Neural network paradigms have been reported that perform computerized interpretations of scintillation data on thallium scans of the heart [6]. Artificial neural network models were developed by presenting clinical data to the neural network in the form of numeric input variables [1]. Fifty characteristics of the history, physical examination, chest radiograph, electrocardiogram, and arterial blood gases were presented to the computer algorithm (Table 51.1).

Table 51.1 Variables presented to the neural network. Alveolar–arterial gradient Pa O2 ∗

Complaints of dyspnea

Atrial flutter

Hemoptysis (≤1 mo) Stroke, ever

Hypoxia (Pa O2 < 60 mm Hg) QRS axis >120 and <−30 degrees

Postpartum (≤3 mo) Cough (≤3 days)

ST-segment and/or T-wave changes

Pa CO2 ∗∗

Recent surgery (≤3 mo) Lower extremity trauma (≤3 mo)

Cyanosis Prominent pulmonary artery (radiograph)

Angina-like pain (≤3 day) Palpitations (≤3 day)

Pulmonary edema (radiograph) Right ventricular hypertrophy (ECG)

History of chronic obstructive pulmonary disease A–a gradient > twice age-corrected History of immobilization (≤3 mo) T-wave changes alone Incomplete right bundle branch block

Asthma, ever Estrogen use (at onset of symptoms) Pleuritic chest pain (≤3 days) Wheezing Pleural friction rub

Loss of consciousness (≤1 wk) Calf tenderness Thrombophlebitis (ever) Leg or foot swelling, not joints

Cardiomegaly (radiograph) ST-segment depression alone P pulmonale Respiratory alkalosis (Pa CO2 ≤ 30 mm Hg)

Any leg pain, not arthritic

Palpable venous cord

Leg erythema Homans’ sign Low-voltage QRS complexes

Left ventricular hypertrophy (ECG) Atrial fibrillation Diaphoresis

Premature ventricular contractions

Premature atrial contractions Malignancy (≤3 mo) Complete right bundle branch block

Pa O2 = partial pressure of oxygen in arterial blood. Pa CO2 = partial pressure of carbon dioxide in arterial blood.

∗∗

236


BLUK077-Stein

March 12, 2007

CHAPTER 51

18:30

237

Neural network computer-assisted diagnosis

Data were obtained from patients recruited in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). The interpretations of the ventilation–perfusion scans were not included in the data presented to the neural network. Variables were coded as either binary variables (present or absent) or continuous variables. The percentage of patients with PE was determined for deciles of risk estimated on the basis of clinical assessment by physicians in PIOPED [1]. This was compared with deciles of risk estimated by neural network clinical assessment (Figure 51.1). In general, there was good agreement between the likelihood of PE estimated by physicians in PIOPED and estimated by the neural network. The areas under the receiver operating curves (ROC) [7] relating sensitivity to the frequency of false-positive values were calculated for clinical assessment estimated by the neural network and by intuitive clinical assessment The areas under the ROC did not differ significantly, indicating that the neural networks performed as well in achieving a clinical diagnosis as physicians

knowledgeable and experienced in the diagnosis of acute PE. Data that were of particular value to the neural network in making or rejecting the diagnosis of PE were identified [1]. Among these were a high alveolar– arterial gradient, a low partial pressure of oxygen in arterial blood (Pa O2 ) on room air, atrial flutter, marked right-axis or left-axis deviation, ST-segment depression, T-wave inversion, and surgery within the last 3 months. It is important to keep in mind that all of the patients who were recruited in PIOPED were identified on the basis of well-known clinical characteristics suggestive of acute PE. Consequently, the well-known features of acute PE were present in patients in whom the diagnosis was made and also in patients in whom the diagnosis was excluded [8, 9]. Typical characteristics of PE were not useful to the neural network in distinguishing between patients with PE and those without PE. Even though similar signs and symptoms were present in patients with and without PE, experienced physicians in PIOPED were often able to correctly

100 90

Neural (n =1213)

80

PIOPED (n =1213)

70

PE (%)

60 50 40 30 20 10

91 –1 00

81 –9 0

71 –8 0

61 –7 0

51 –6 0

41 –5 0

31 –4 0

21 –3 0

11 –2 0

0– 10

0

Predicted PE (%) Figure 51.1 Percentage of patients with pulmonary embolism (PE) for deciles of the predicted risk based on PIOPED clinical assessment and on neural network clinical assessment. (Reprinted with permission from Patil et al. [1].)


BLUK077-Stein

March 12, 2007

18:30

238

estimate the likelihood of PE, based on undefined clinical judgment [10]. This suggests that there were details of the clinical characteristics that were not recorded in the PIOPED database. It also suggests that there may be recorded combinations of observations that were weighted by physicians in reaching a correct clinical assessment.

PART III

6

7

References 1 Patil S, Henry JW, Rubenfire M, Stein PD. Neural network in the clinical diagnosis of acute pulmonary embolism. Chest 1993; 104: 1685–1689. 2 Baxt WG. Use of an artificial neural network for the diagnosis of myocardial infarction. Ann Intern Med 1991; 115: 843–848. 3 Bounds DG, Lloyd PJ. A comparison of neural network and other pattern recognition approaches to the diagnosis of low back disorders. Neural Netw 1990; 3: 583–591. 4 Scott JA, Palmer EL. Neural network analysis of ventilation–perfusion lung scans. Radiology 1993; 186: 661–664. 5 Tourassi GD, Floyd CE, Sostman HD, Coleman RE. Ar-

8

9

10

Diagnosis of acute PE

tificial neural network for diagnosis of acute pulmonary embolism: effect of case and observer selection. Radiology 1995; 194: 889–893. Fujita H, Katafuchi T, Uehara T, Nishimura T. Application of artificial neural network to computer-aided diagnosis of coronary artery disease in myocardial SPECT bull’s-eye images. J Nucl Med 1992; 33(2): 272–276. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982; 143: 29–36. Stein PD, Terrin ML, Hales CA et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest 1991; 100: 598–603. Stein PD, Saltzman HA, Weg JG. Clinical characteristics of patients with acute pulmonary embolism. Am J Cardiol 1991; 68: 1723–1724. A Collaborative Study by the PIOPED Investigators. Value of the ventilation–perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759.


BLUK077-Stein

April 6, 2007

52

16:5

CHAPTER 52

Empirical assessment and clinical models for diagnosis of acute pulmonary embolism

The most prominent standardized clinical scores based on objective or mostly objective clinical findings are the Wells (or “Canadian”) extended clinical model [1–3], the Wells simplified model [3–5], and the original “Geneva” clinical model [5, 6]. More recently, a revised Geneva clinical model was developed [7]. A Pisa model has also been developed [8]. In addition, many use an empirical assessment [3, 6, 9]. Differences in the Wells simplified model [4] and the Wells extended model [2] relate to whether leg swelling is measured (extended model) or appears swollen (simplified model) and whether pulmonary embolism (PE) is as likely or more likely than an alternative diagnosis (extended model) or just more likely (simplified model). The Wells scoring system is shown in Table 52.1. A Wells score of <2 points indicates a

low probability of PE, 2–6 points indicates a moderate probability, and >6 points indicates a high probability of PE [1, 4]. A Wells score of ≤4 points also has been used to indicate that PE is “unlikely” and >4 points indicates that PE is likely [4, 10]. The Wells extended and simplified models are not strictly objective. Judgment is required to evaluate the final variable “if pulmonary embolism is as likely or more likely than an alternative diagnosis.” To make this judgment, physicians may use the history, physical examination, chest radiograph, electrocardiogram, and whatever blood tests are considered necessary to diagnose PE [1]. Scoring for the original Geneva system is shown in Table 52.2. The original Geneva scoring system, although objective, requires an arterial blood gas analysis. A revised Geneva score has been developed based

Table 52.1 Clinical scoring system for suspected pulmonary embolism (Wells). Clinical feature*

Wells extended model*

Wells simplified model†

Score

Signs/symptoms of DVT

Measured leg swelling plus pain

Leg swelling plus pain with

3.0

with palpation in deep vein region

palpation in deep vein region

Heart rate >100/min

Yes

Yes

1.5

Immobilization: bedrest

Yes

Yes

1.5

Prior PE or DVT

Yes

Yes

1.5

Hemoptysis

Yes

Yes

1.0

Malignancy: treated within

Yes

Yes

1.0

PE as likely or more likely

PE more likely

3.0

(± bathroom privileges) ≥3 consecutive days or surgery in last 4 wk

6 mo or on palliative care Alternative diagnosis

Low probability of PE: score <2; moderate probability: score 2–6; high probability: score >6; unlikely probability: score ≤4† ; likely probability: score >4. ∗Wells et al. [1]. † Wells et al. [4]. PE, pulmonary embolsim; DVT, deep venous thrombosis.

239


BLUK077-Stein

April 6, 2007

16:5

240

PART III

Diagnosis of acute PE

Table 52.2 Original Geneva scoring system for suspected pulmonary embolism.

Table 52.3 Revised Geneva scoring system for suspected pulmonary embolism.

Feature

Feature

Score

Score

Age >65 yr

1

60–79

1

Previous PE or DVT

3

≥80

2

Surgery (under general anesthesia) or fracture of

2

Age (yr)

Previous PE or DVT

2

Recent surgery

3

Heart rate > 100/min

1

lower limbs (≤1 mo) Solid or hematologic malignant condition,

2

currently active or considered cured <1 yr

Pa CO2

Unilateral lower limb pain

3

<4.8 kPa (<36 mm Hg)

2

Hemoptysis

2

4.8–5.19 kPa (36–39 mm Hg)

1

Heart rate

Pa O2 <6.5 kPa (<49 mm Hg)

4

6.5–7.99 kPa (49–60 mm Hg)

3

8–9.49 kPa (61–71 mm Hg)

2

9.5–10.99 kPa (71–82 mm Hg)

1 1

Elevation of hemidiaphragm

1

3

≥95 beats/min

5

Pain on lower limb deep venous palpation and

4

unilateral edema Low probability of PE: score 0–3; intermediate probability: score 4–10; high probability: score ≥11. PE, pulmonary embolsim; DVT, deep venous thrombosis. From Le Gal et al. [7], with permission.

Chest X-ray Plate-like atelectasis

75–94 beats/min

Low probability of PE: score 0–4; moderate probability: score 5–8; high probability: score ≥9. Wicki et al. [6]. PE, pulmonary embolsim; DVT, deep venous thrombosis. Pa O2 = partial pressure of oxygen in arterial blood. Pa CO2 = partial pressure of carbon dioxide in arterial blood.

entirely on clinical variables and independent of arterial blood gases (Table 52.3) [7]. A “Pisa” clinical model showed excellent results (Table 52.4), but is more difficult to apply [8]. The Wells extended clinical model and simplified clinical model are applicable to inpatients as well as

outpatients [1, 2]. The original Geneva model and the revised Geneva model are applicable to patients seen on an emergency service [6, 7]. The Wells extended and simplified scores [2, 4], and the original and revised Geneva scores give points for recent surgery [6, 7]. The simplified Wells prediction rule and the original Geneva prediction rule performed poorly in patients referred from surgical wards [11]. Results with these prediction rules are shown in Table 52.4. Comparisons between the simplified Wells system and the original Geneva system in the same

80

0 Clin Low

Clin Moderate

Geneva revised

Geneva original

Empirical Geneva revised

Wells simplified Geneva original

10

Empirical Wells extended

20

Geneva revised

30

Wells simplified

40

Geneva original

50

Wells extended Wells simplified

60

Empirical Wells extended

Positive predictive value (%)

70

Clin High

Figure 52.1 Positive predictive value for pulmonary embolism of low, moderate, and high probability clinical assessments using empirical judgment, Wells extended model, Wells simplified model, original Geneva model, and revised Geneva model. Data based on values shown in Table 52.4.


BLUK077-Stein

April 6, 2007

CHAPTER 52

16:5

241

Empirical assessment and clinical models

Table 52.4 Positive predictive value of empirical clinical assessment and clinical prediction scores for pulmonary embolism. PE / N (%) Method/first author (Ref.)

Low probability

Moderate probability

High probability

Unlikely

Likely

9/178 (5%)

27/69 (39%)

Empirical Sanson [3]

80/278 (29)

35/77 (46)

PIOPED [9]

21/228 (9)

11/58 (19)

170/569 (30)

61/90 (68)

Wicki [6]

33/368 (9)

174/523 (33)

62/94 (66)

Average

65/654 (10)

424/1390 (31)

158/261 (61)

25/734 (3)

Wells extended Wells [1]

112/403 (28)

80/102 (78)

Sanson [3]

17/60 (28)

54/138 (39)

18/39 (46)

Wells [2]

7/357 (2)

45/168 (27)

28/43 (65)

Average

49/1151 (4)

211/709 (30)

126/184 (68)

2/99 (2)

24/128 (19)

10/20 (50)

41/147 (28)

78/259 (30)

3/8 (38)

Wells simplified Wells [4] Sanson [3] Chagnon [5]

19/162 (12)

42/104 (40)

10/11 (91)

Average

62/408 (15)

144/491 (29)

23/39 (59)

Wicki [6]

49/486 (10)

166/437 (38)

51/63 (81)

Chagnon [5]

20/152 (13)

43/113 (38)

8/12 (67)

Average

69/638 (11)

209/550 (38)

59/75 (79)

18/229 (8)

132/463 (29)

42/57 (74)

19/432 (4)

62/283 (22)

306/313 (98)

Geneva original

Geneva revised Le Gal [7] Pisa Miniati [8] PE, pulmonary embolsim.

patients were made by Chagnon and associates [5] and are also shown in Table 52.4. Receiver operating characteristic curve analysis showed no difference between results with these prediction rules [5]. The positive predictive value of a low-probability clinical assessment, whether empirical or by any of the prediction rules (Wells extended, Wells simplified, Geneva original, Geneva revised, or Pisa), on average, was 4–15% (Table 52.4, Figure 52.1). The positive predictive value of a moderate- or intermediateprobability clinical assessment was 22–38%. Highprobability clinical assessment showed PE in 59–98%.

2

3

4

References 1 Wells PS, Anderson DR, Rodger M et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pul-

5

monary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med 2001; 135: 98–107. Wells PS, Ginsberg JS, Anderson DR et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 1998; 129: 997– 1005. Sanson BJ, Lijmer JG, Mac Gillavry MR, Turkstra F, Prins MH, Buller HR, for the ANTELOPE-Study Group. Comparison of a clinical probability estimate and two clinical models in patients with suspected pulmonary embolism. Thromb Haemost 2000; 83: 199–203. Wells PS, Anderson DR, Rodger M et al. Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer. Thromb Haemost 2000; 83: 416– 420. Chagnon I, Bounameaux H, Aujesky D et al. Comparison of two clinical prediction rules and implicit assessment


BLUK077-Stein

April 6, 2007

16:5

242

among patients with suspected pulmonary embolism. Am J Med 2002; 113: 269–275. 6 Wicki J, Pernerger TV, Junod AF et al. Assessing clinical probability of pulmonary embolism in the emergency ward: a simple score. Arch Intern Med 2001; 161: 92–97. 7 Le Gal G, Righini M, Roy P-M, et al. Prediction of pulmonary embolism in the emergency department: the revised Geneva score. Ann Intern Med 2006; 144: 165– 171. 8 Miniati M, Monti S, Bottai M. A structured clinical model for predicting the probability of pulmonary embolism. Am J Med 2003; 114: 173–179.

PART III

Diagnosis of acute PE

9 A Collaborative Study by the PIOPED Investigators. Value of the ventilation–perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 10 van Belle A, Buller HR, Huisman MV et al; Christopher Study Investigators. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography. JAMA 2006; 295: 172–179. 11 Ollenberger GP, Worsley DF. Effect of patient location on the performance of clinical models to predict pulmonary embolism. Thromb Res 2006; 118: 685–690.


BLUK077-Stein

March 28, 2007

53

15:31

CHAPTER 53

D-dimer for the exclusion of acute pulmonary embolism

Tests for D-dimer have been available since the 1980s for use in the exclusion of venous thromboembolic disease [1]. In spite of this extensive literature, the role of D-dimer in the diagnosis of deep vein thrombosis (DVT) or pulmonary embolism (PE) was unclear, in part due to the multiple D-dimer assays that are available, the role of central laboratory testing versus bedside testing, and concerns about variability of the assays [1]. A systematic review of the literature showed that in ranking the D-dimer assays according to the sensitivity values and likelihood of increasing certainty for ruling out PE, the values for sensitivity for the ELISA (enzyme-linked immunosorbent assay) and quantitative rapid ELISA assays were significantly superior to those for the semiquantitative latex and whole blood agglutination assays [2]. The quantitative rapid ELISA assay is more convenient than the conventional ELISA, and provides a high level of certainty for a negative diagnosis of PE as well as DVT [2]. In this chapter, D-dimer as a stand-alone test for the exclusion of acute PE will be discussed. D-dimer combined with clinical probability assessment for the exclusion of PE will be discussed in Chapter 54. Chapter 31 discussed D-dimer as a stand-alone test for the exclusion of DVT. D-dimer for the exclusion of DVT in combination with clinical probability was reviewed in Chapter 32. A brief description of various D-dimer tests is shown in Table 53.1. Tier 1 (11 studies) compared an ELISA assay and at least one other D-dimer assay [3â&#x20AC;&#x201C;13]. Tier 2 studies included the Tier 1 studies and 20 other studies that met all inclusion criteria [6, 14â&#x20AC;&#x201C;33]. There were, therefore, 31 investigations that met all inclusion criteria, and these comprised Tier 2 (Table 53.2). Tier 3 combined 10 methodologically weaker studies [34â&#x20AC;&#x201C;43] with the 31 Tier 2 studies. Sensitivity, specificity, and likelihood ratios differed according to whether D-dimer was being used to exclude PE or DVT. Non-ELISA assays should not be used as stand-alone tests for PE [44] given their in-

ferior sensitivity and negative likelihood ratio values. However, in patients with a low clinical probability of PE, non-ELISA assays provide a reasonable certainty for ruling it out [2]. In patients with PE, the highest sensitivity was seen among the ELISA assays using a cutoff level of 500 ng/mL (Figure 53.1) [2]. The variability of the sensitivity values is generally less for the ELISA assays as a group (Figure 53.1), probably reflecting assay reproducibility given the very large database in this systematic review [2]. The ELISA and quantitative rapid ELISA had a negative likelihood ratio of 0.13 in Tier 1 studies (Table 53.3), 0.07 in Tier 2 studies (Table 53.3), and 0.05 in Tier 3 studies (Table 53.4) [2]. Likelihood ratios of <0.1 generate large and often conclusive changes from pretest to posttest probability [45] providing high certainty for excluding PE. Combining a negative rapid ELISA with a low or moderate clinical probability for PE essentially rules out PE (Chapter 31). However, a negative D-dimer does not reliably exclude PE if the clinical probability is high (Chapter 31). Furthermore, a negative D-dimer has the highest negative predictive values in populations with a low prevalence of PE. It is possible to estimate the negative predictive value of the quantitative rapid ELISA for exclusion of acute PE from the sensitivity (97%) and specificity (41%) (based on Tier 2 studies) knowing the prevalence of PE in the population tested. The prevalence of PE in the included investigations of D-dimer ranged from 8 to 62% (average 25%) [2]. If the prevalence of PE in the population to be tested were 25%, the negative predictive value of a negative quantitative rapid ELISA would be 98% [2] (Figure 53.2). If the prevalence of PE were 8%, the negative predictive value would be 99%. If the prevalence of PE were 62%, the negative predictive value would be only 89%. Limited data for sensitivity, specificity, and likelihood ratios were available for cutoff values of 250 and

243


BLUK077-Stein

March 28, 2007

15:31

244

PART III

Diagnosis of acute PE

Table 53.1 Brief description of D-dimer tests. Type of test

Description

Time to results

Enzyme-linked

The conventional ELISA is considered a reference standard

Approximately 8 h; may

immunosorbent assay

for determination of D-dimer concentration. The

not be available the

(ELISA)

conventional ELISA is not practical for use in the

same day

diagnosis of venous thromboembolism in individual patients, because it is laboratory intensive. Availability in most centers is limited Quantitative rapid ELISA

The quantitative rapid ELISA test uses an antibody to

Approximately 35 min

D-dimer that is fluorescent labeled. An automated immunoanalyzer provides a numerical result Semiquantitative rapid ELISA

The semiquantitative rapid ELISA employs an antibody to

Less than 10 min

D-dimer that is tagged to a color-producing agent. Plasma D-dimer concentration is quantified from the degree of color intensity produced

Qualitative rapid ELISA

The qualitative rapid ELISA method involves the presence

10 min

of D-dimer detected by an anti-D-dimer monoclonal antibody coupled to alkaline phosphatase. Activity is revealed by the addition of a substrate that causes a color change. The intensity of the color is read visually by comparison with a positive and negative control Quantitative latex

Quantitative latex agglutination assays use monoclonal

agglutination

antibodies to D-dimer that are coated onto latex

7â&#x20AC;&#x201C;15 min

particles. In the presence of D-dimer, the particles aggregate to form larger aggregates and light scattering decreases. Quantification of the D-dimer concentration is done with an analyzer that detects agglutination and precipitation turbidimetrically Semiquantitative latex agglutination

Semiquantitative latex agglutination assays rely on the

3â&#x20AC;&#x201C;4 min

use of monoclonal antibodies to D-dimer that are coated onto latex particles. Macroscopic agglutinations are seen when elevated D-dimer levels are present in the plasma

Whole blood agglutination

The most frequently studied whole blood assay uses a

2 min

freshly collected drop of capillary or venous whole blood, which is mixed with a conjugate of monoclonal antibody to D-dimer. The latter is linked to a monoclonal antibody to human red blood cells. Visible agglutination of red cells takes place in the presence of elevated D-dimer levels. Interpreter experience is important due to difficulty in discriminating between weak positive and normal results

1000 ng/mL for some but not all of the D-dimer assays (Table 53.4) [2]. The clinical utility of the D-dimer assays is limited by the nonspecificity of a positive result due to factors such as inflammation, trauma, and surgery to name some. The clinical utility differs among patient populations. The proportion of inpatients in whom PE

can be excluded by a normal D-dimer is lower than in outpatients, because inpatients more often have unrelated disorders that cause a positive D-dimer test [46]. Similarly, the proportion of normal D-dimer tests in patients with a history of prior DVT or PE is lower than in patients without previous venous thromboembolic disease [47].


BLUK077-Stein

March 28, 2007

CHAPTER 53

15:31

245

D-dimer for exclusion of acute PE

Table 53.2 Patients with clinically suspected pulmonary embolism.

Study

Cutoff

No. of

Patients

analyzed

patients

with

enrolled

PE (%)

Year

D-dimer type

(ng/mL)

1991

Quant rapid ELISA, qual

<500

100

25.0

Tier 1 analysis Leitha [8]

latex Ginsberg [6]

1993

ELISA, semiquant latex

<500

221

19.5

Lenzhofer [9]

1993

ELISA, semiquant latex

<250

118

14.4

Flores [5]

1995

Quant rapid ELISA, quant

<500

85

36.5

<500, <1000

183

10.4, 10.4,

latex Veitl [12]

1996

ELISA, semiquant rapid ELISA, quant latex,

9.8, 10.4,

semiquant latex,

9.3

whole blood Duet [4]

1998

ELISA, quant latex

<500

85

18.8

Meyer [10]

1998

Qual rapid ELISA, quant

<500

142

35.9, 42.3

latex Oger [11]

1998

ELISA, quant latex

<500

386

37.8

Heit [7]

1999

Semiquant rapid ELISA,

≤ 500, < 250,

105

31.4, 31.4

<500

131

22.9, 25.2

<500

287

31.4

≤ 250

ELISA, quant latex, semiquant latex Wallis [13]

2000

Quant rapid ELISA, whole blood

de Monye [3]

2002

Quant rapid ELISA, quant latex

Tier 2 analysis (Tier 2 includes Tier 1 and the following patients) Bounameaux [15]

1989

ELISA

<500

99

9.1

Bounameaux [16]

1990

ELISA

<500

67

29.9

Bounameaux [17]

1992

ELISA

<500

170

32.4

Demers [21]

1992

ELISA

<500

156

15.4

Pappas [28]

1993

Semiquant latex

<250

20

35.0

Ginsberg [24]

1995

Whole blood

86

18.6

Reber [31]

1995

ELISA, semiquant rapid

<500

301

30.9

de Moerloose [20]

1996

Quant rapid ELISA

<500

195

23.6

Perrier [29]

1996

ELISA

<500

308

35.4

Bonnin [14]

1997

Quant rapid ELISA

<500

83

22.9

Knecht [26]

1997

Quant latex

<500

154

10.4

Perrier [30]

1997

ELISA

<500

671

29.2

Egermayer [22]

1998

Whole blood

499

8.0

Ginsberg [25]

1998

Whole blood

1177

16.7

de Groot [19]

1999

Whole blood

245

24.9

Kruip [32]

1999

Quant rapid ELISA

<500

137

24.1

ELISA

Kutinsky [33]

1999

Semiquant latex

<250, <500

98

30.6

Bova [18]

2000

Semiquant rapid ELISA

<300

140

32.1

Farrell [23]

2000

Whole blood

198

20.7

MacGillavry [27]

2001

Whole blood

403

31.0

PE, pulmonary embolism; Quant, quantitative; qual, qualitative; semiquant, semiquantitative. Reprinted with minor modifications with permission from Stein et al. [2].


Negative predictive value (%) 1.0

99

8 E Q L ra ua ISA p i lit d at E i Q LI ve ra uan SA* pi tit Se d E ativ LI e m ra iqu SA pi an d EL tita IS tiv A e Q la uan te tit Se x as ativ m s e la iqu ay te an x as tita sa tiv y e W ho l e as b sa loo y d

E Q L ra ua ISA p i lit d at E i Q LI ve ra uan SA* pi tit Se d E ativ LI e m ra iqu SA pi an d EL tita IS tiv A e Q la uan te tit Se x as ativ m s e la iqu ay te an x as tita sa tiv y e W ho l e as b sa loo y d

Proportion

March 28, 2007

E Q L ra ua ISA pi lit d at E i Q LI ve ra uan SA* pi tit Se d E ativ LI e m ra iqu SA pi an d EL tita IS tiv A e Q la uan te tit x Se as ativ m s e la iqu ay te an x t as ita sa tiv y e W ho l as e b sa loo y d

E Q L ra ua ISA pi lit d at E i Q LI ve ra uan SA* pi tit Se d E ativ LI e m ra iqu SA pi an d EL tita IS tiv A e Q la uan te tit x Se as ativ m s e la iqu ay te an x t as ita sa tiv y e W ho l as e b sa loo y d

Proportion

BLUK077-Stein

15:31

Tier 1

0.95 0.7

0.6

0.85 0.5

0.4

0.75 0.3

Tier 2

0.6

80

25

Prevalence of PE (%)

246 0.8

0.9 0.7

0.8 0.6

0.7 0.5

0.4

0.5 0.3

Sensitivity Specificity

Figure 53.1 Boxplots of the finding for sensitivity and specificity among the D-dimer assays for patients with suspected pulmonary embolism. (Reprinted with permission from Stein PD et al. [2].)

100 98

89

60

40

20

0

62

Figure 53.2 Negative predictive value of D-dimer quantitative rapid ELISA (enzyme-linked immunosorbent assay) for pulmonary embolism (PE) in relation to prevalence of PE in population tested. (Data from Stein et al. [2].)


BLUK077-Stein

March 28, 2007

15:31

Table 53.3 Pulmonary embolism. Sensitivity

Specificity

Positive likelihood

Negative likelihood

Test

(95% CI)

(95% CI)

ratio (95% CI)

ratio (95% CI)

ELISA

0.95 (0.85–1.00)

0.44 (0.34–0.54)

1.68 (1.44–1.95)

0.13 (0.03–0.58)

Quant rapid ELISA

0.95 (0.83–1.00)

0.39 (0.28–0.51)

1.56 (1.32–1.83)

0.13 (0.02–0.84)

Tier 1 analysis

Semiquant rapid ELISA

0.93 (0.79–1.00)

0.36 (0.23–0.50)

1.45 (1.20–1.76)

0.20 (0.04–0.96)

Qual rapid ELISA

0.93 (0.74–1.00)

0.68 (0.50–0.87)

2.92 (1.77–4.79)

0.11 (0.01–0.93)

Quant latex

0.89 (0.81–0.98)

0.45 (0.36–0.53)

1.62 (1.43–1.84)

0.24 (0.13–0.45)

Semiquant latex

0.92 (0.79–1.00)

0.45 (0.31–0.59)

1.68 (1.35–2.09)

0.17 (0.04–0.78)

Whole blood

0.78 (0.64–0.92)

0.74 (0.60–0.88)

2.93 (1.89–4.52)

0.31 (0.18–0.51)

Tier 2 analysis ELISA

0.96 (0.88–1.00)

0.51 (0.44–0.59)

1.97 (1.72–2.26)

0.08 (0.01–0.43)

Quant rapid ELISA

0.97 (0.87–1.00)

0.41 (0.30–0.51)

1.64 (1.40–1.91)

0.07 (0.00–1.55)

Semiquant rapid ELISA

0.93 (0.79–1.00)

0.40 (0.27–0.54)

1.55 (1.25–1.93)

0.18 (0.04–0.94)

Qual rapid ELISA

0.91 (0.68–1.00)

0.70 (0.47–0.93)

3.01 (1.52–5.96)

0.13 (0.01–1.28)

Quant latex

0.89 (0.80–0.99)

0.47 (0.38–0.57)

1.69 (1.44–1.99)

0.23 (0.11–0.48)

Semiquant latex

0.80 (0.65–0.94)

0.56 (0.42–0.70)

1.81 (1.35–2.42)

0.36 (0.20–0.67)

Whole blood

0.83 (0.74–0.92)

0.64 (0.55–0.73)

2.32 (1.87–2.88)

0.27 (0.17–0.42)

Quant, quantitative; qual, qualitative; semiquant, semiquantitative. Reprinted with minor modifications with permission from Stein et al. [2]. Table 53.4 Pulmonary embolism.

Test

Sensitivity

Specificity

Positive likelihood

Negative likelihood

(95% CI)

(95% CI)

ratio (95% CI)

ratio (95% CI)

Tier 3 analysis: Cutoff 500 ng/mL (all data) ELISA

0.95 (0.88–1.00)

0.45 (0.38–0.53)

1.74 (1.55–1.96)

0.11 (0.03–0.39)

Quant rapid ELISA

0.98 (0.88–1.00)

0.40 (0.29–0.50)

1.62 (1.38–1.91)

0.05 (0.00–4.15)

Semiquant rapid ELISA

0.94 (0.81–1.00)

0.39 (0.26–0.52)

1.55 (1.24–1.92)

0.15 (0.02–1.13)

Qual rapid ELISA

0.92 (0.71–1.00)

0.68 (0.46–0.90)

2.92 (1.52–5.61)

0.11 (0.01–1.45)

Quant latex

0.90 (0.81–1.00)

0.46 (0.37–0.56)

1.68 (1.42–1.98)

0.21 (0.09–0.49)

Semiquant latex

0.86 (0.74–0.97)

0.51 (0.39–0.62)

1.73 (1.39–2.16)

0.29 (0.14–0.58)

Whole blood

0.82 (0.74–0.91)

0.63 (0.54–0.71)

2.21 (1.81–2.70)

0.28(0.18–0.43)

0.07

Tier 3 analysis: Cutoff 250 ng/mL (all studies meeting inclusion criteria). ELISA

0.96

0.55

2.13

Quant rapid ELISA

NA

NA

NA

NA

Semiquant rapid ELISA

NA

NA

NA

NA

Qual rapid ELISA

NA

NA

NA

NA

Quant latex

0.94

0.44

1.69

0.14

Semiquant latex

0.90

0.63

2.39

0.17

Whole blood

0.84

0.62

2.22

0.26

NA

Tier 3 analysis: Cutoff 1000 ng/mL (all studies meeting inclusion criteria) ELISA

NA

NA

NA

Quant rapid ELISA

NA

NA

NA

NA

Semiquant rapid ELISA

0.74

0.75

2.96

0.35

Qual rapid ELISA

NA

NA

NA

NA

Quant latex

NA

NA

NA

NA

Semiquant latex

NA

NA

NA

NA

Whole blood

0.84

0.62

2.22

0.26

Quant, quantitative; qual, qualitative; semiquant, semiquantitative. Reprinted with minor modifications with permission from Stein et al. [2].

247


BLUK077-Stein

March 28, 2007

15:31

248

References 1 Becker DM, Philbrick JT, Bachhuber TL, Humphries JE. D-dimer testing and acute venous thromboembolism. A shortcut to accurate diagnosis? Arch Intern Med 1996; 156: 939–946. 2 Stein PD, Hull RD, Patel KC et al. D-dimer for the exclusion of acute venous thrombosis and pulmonary embolism: a systematic review. Ann Intern Med 2004; 140: 589–602. 3 de Monye W, Sanson BJ, Buller HR, Pattynama PMT, Huisman MV. The performance of two rapid quantitative D-dimer assays in 287 patients with clinically suspected pulmonary embolism. Thromb Res 2002; 107: 283–286. 4 Duet M, Benelhadj S, Kedra W et al. A new quantitative D-dimer assay appropriate in emergency: reliability of the assay for pulmonary embolism exclusion diagnosis. Thromb Res 1998; 91: 1–5. 5 Flores J, Lancha C, Perez Rodriguez E, Garcia Avello A, Bollo E, Garvia Frade LJ. Efficacy of D-dimer and total fibrin degradation products evaluation in suspected pulmonary embolism. Respiration 1995; 62: 258–262. 6 Ginsberg JS, Brill-Edwards PA, Demers C, Donovan D, Panju A. D-dimer in patients with clinically suspected pulmonary embolism. Chest 1993; 104: 1679–1684. 7 Heit JA, Minor TA, Andrews JC, Larson DR, Li H, Nichols WL. Determinants of plasma fibrin D-dimer sensitivity for acute pulmonary embolism as defined by pulmonary angiography. Arch Pathol Lab Med 1999; 123: 235–240. 8 Leitha T, Speiser W, Dudczak R. Pulmonary embolism: efficacy of D-dimer and thrombin–antithrombin III complex determinations as screening tests before lung scanning. Chest 1991; 100: 1536–1541. 9 Lenzhofer R, Wimmer F, Haydl H et al. Prospective study of determining the value of D-dimer in diagnosing pulmonary embolism. Wien Klin Wochenschr 1993; 105: 492– 496. 10 Meyer G, Fischer AM, Collignon MA et al. Diagnostic value of two rapid and individual D-dimer assays in patients with clinically suspected pulmonary embolism: comparison with microplate enzyme linked immunosorbent assay. Blood Coagul Fibrinolysis 1998; 9: 603–608. 11 Oger E, Leroyer C, Bressollette L, et al. Evaluation of a new, rapid, and quantitative D-dimer test in patients with suspected pulmonary embolism. Am J Respir Crit Care Med 1998; 158: 65–70. 12 Veitl M, Hamwi A, Kurtaran A, Virgolini I, Vukovich T. Comparison of four rapid D-dimer tests for diagnosis of pulmonary embolism. Thromb Res 1996; 82: 399–407. 13 Wallis JW, Kruip M, de Jongh-Leuvenink J, B¨uller HR. A comparative study of two rapid D-dimer tests for the exclusion of pulmonary embolism in symptomatic patients [letter]. Thromb Haemost 2000; 84: 925.

PART III

Diagnosis of acute PE

14 Bonnin F, Hadjikostova H, Jebrak G et al. Complementarity of lung scintigraphy and D-dimer test in pulmonary embolism. Eur J Nucl Med 1997; 24: 444–447. 15 Bounameaux H, Slosman D, de Moerloose P, Reber G. Laboratory diagnosis of pulmonary embolism: value of increased levels of plasma D-dimer and thrombin– antithrombin III complexes. Biomed Pharmacother 1989; 43: 385–388. 16 Bounameaux H, Schneider PA, Slosman D, de Moerloose P, Reber G. Plasma D-dimer in suspected pulmonary embolism: a comparison with pulmonary angiography and ventilation–perfusion scintigraphy. Blood Coagul Fibrinolysis 1990; 1: 577–579. 17 Bounameaux H. Significance of the determination of Ddimers in venous thromboembolic disease. J Mal Vasc 1992; 17(suppl B): 91–92. 18 Bova C, Greco F, Ferrari A, Serafini O et al. The usefulness of the association of clinical probability, rapid plasma measurement of D-dimer, compression echography of the lower limbs and echocardiography in the diagnosis of acute pulmonary embolism. Ital Heart J 2000; 1: 116– 121. 19 de Groot MR, van Marwijk Kooy M, Pouwels JG, Engelage AH, Kuipers BF, Buller HR. The use of a rapid D-dimer blood test in the diagnostic work-up for pulmonary embolism: a management study. Thromb Haemost 1999; 82: 1588–1592. 20 de Moerloose P, Desmarais S, Bounameaux H et al. Contribution of a new, rapid, individual and quantitative automated D-dimer ELISA to exclude pulmonary embolism. Thromb Haemost 1996; 75: 11–13. 21 Demers C, Ginsberg JS, Johnston M, Brill-Edwards P, Panju A. D-dimer and thrombin–antithrombin III complexes in patients with clinically suspected pulmonary embolism. Thromb Haemost 1992; 67: 408– 412. 22 Egermayer P, Town GI, Turner JG, Heaton DC, Mee AL, Beard ME. Usefulness of D-dimer, blood gas, and respiratory rate measurements for excluding pulmonary embolism. Thorax 1998; 53: 830–834. 23 Farrell S, Hayes T, Shaw M. A negative SimpliRED Ddimer assay result does not exclude the diagnosis of deep vein thrombosis or pulmonary embolus in emergency department patients. Ann Emerg Med 2000; 35: 121– 125. 24 Ginsberg JS, Wells PS, Brill-Edwards P et al. Application of a novel and rapid whole blood assay for D-dimer in patients with clinically suspected pulmonary embolism. Thromb Haemost 1995; 73: 35–38. 25 Ginsberg JS, Wells PS, Kearon C et al. Sensitivity and specificity of a rapid whole-blood assay for D-dimer in the diagnosis of pulmonary embolism. Ann Intern Med 1998; 129: 1006–1011.


BLUK077-Stein

March 28, 2007

CHAPTER 53

15:31

D-dimer for exclusion of acute PE

26 Knecht MF, Heinrich F. Clinical evaluation of an immunoturbidimetric D-dimer assay in the diagnostic procedure of deep vein thrombosis and pulmonary embolism. Thromb Res 1997; 88: 413–417. 27 MacGillavry MR, Lijmer JG, Sanson BJ, Buller HR, Brandjes DP. Diagnostic accuracy of triage tests to exclude pulmonary embolism. Thromb Haemost 2001; 85: 995– 998. 28 Pappas AA, Dalrymple G, Harrison K et al. The application of a rapid D-dimer test in suspected pulmonary embolus. Arch Pathol Lab Med 1993; 117: 977– 980. 29 Perrier A, Bounameaux H, Morabia A et al. Diagnosis of pulmonary embolism by a decision analysis-based strategy including clinical probability, D-dimer levels, and ultrasonography: a management study. Arch Intern Med 1996; 156: 531–536. 30 Perrier A, Desmarais S, Goehring C et al. D-dimer testing for suspected pulmonary embolism in outpatients. Am J Respir Crit Care Med 1997; 156: 492–496. 31 Reber G, Vissac AM, de Moerloose P, Bounameaux H, Amiral J. A new, semi-quantitative and individual ELISA for rapid measurement of plasma D-dimer in patients suspected of pulmonary embolism. Blood Coagul Fibrinolysis 1995; 6: 460–463. 32 Kruip MJHA, Slob MJ, Buller HR, van der Heul C. A prospective clinical study on the feasibility and safety of a combined noninvasive/pulmonary angiography strategy in patients with suspected pulmonary embolism: an interim analysis. Clin Appl Thromb Haemost 1999; 5(4): 222–225. 33 Kutinsky I, Blakley S, Roche V. Normal D-dimer levels in patients with pulmonary embolism. Arch Intern Med 1999; 159: 1569–1572. 34 Goldhaber SZ, Vaughan DE, Tumeh SS, Loscalzo J. Utility of cross-linked fibrin degradation products in the diagnosis of pulmonary embolism. Am Heart J 1988; 116: 505–508. 35 Goldhaber SZ, Simons GR, Elliott CG et al. Quantitative plasma D-dimer levels among patients undergoing pulmonary angiography for suspected pulmonary embolism. JAMA 1993; 270: 2819–2822. 36 Gosselin RC, Owings JT, Utter GH, Jacoby RC, Larkin EC. A new method for measuring D-dimer using immunoturbidometry: a study of 255 patients with suspected pulmonary embolism and deep vein thrombosis. Blood Coagul Fibrinolysis 2000; 11: 715–721.

249

37 Harrison KA, Haire WD, Pappas AA et al. Plasma Ddimer: a useful tool for evaluating suspected pulmonary embolus. J Nucl Med 1993; 34: 896–898. 38 Johanning JM, Veverka TJ, Bays RA, Tong GK, Schmiege SK. Evaluation of suspected pulmonary embolism utilizing end-tidal CO2 and D-dimer. Am J Surg 1999; 178: 98–102. 39 Kline JA, Israel EG, Michelson EA, O’Neil BJ, Plewa MC, Portelli DC. Diagnostic accuracy of a bedside D-dimer assay and alveolar dead-space measurement for rapid exclusion of pulmonary embolism: a multicenter study. JAMA 2001; 285: 761–768. 40 LaCapra S, Arkel YS, Ku DH, Gibson D, Lake C, Lam X. The use of thrombus precursor protein, D-dimer, prothrombin fragment 1.2, and thrombin antithrombin in the exclusion of proximal deep vein thrombosis and pulmonary embolism. Blood Coagul Fibrinolysis 2000; 11: 371–377. 41 Lichey J, Reschofski I, Dissmann T, Priesnitz M, Hoffmann M, Lode H. Fibrin degradation product D-dimer in the diagnosis of pulmonary embolism. Klin Wochenschr 1991; 69: 522–526. 42 Quinn RJ, Nour R, Butler SP et al. Pulmonary embolism in patients with intermediate probability lung scans: diagnosis with Doppler venous US and D-dimer measurement. Radiology 1994; 190: 509–511. 43 Quinn DA, Fogel RB, Smith CD et al. D-dimers in the diagnosis of pulmonary embolism. Am J Respir Crit Care Med 1999; 159: 1445–1449. 44 Jaeschke R, Guyatt GH, Sackett DL, for the EvidenceBased Medicine Working Group. Users’ guides to the medical literature. III: How to use an article about a diagnostic test. B: What are the results and will they help me in caring for my patients? JAMA 1994; 271: 703– 707. 45 Lee AY, Julian JA, Levine MN et al. Clinical utility of a rapid whole-blood D-dimer assay in patients with cancer who present with suspected acute deep venous thrombosis. Ann Intern Med 1999; 131: 417–423. 46 Schrecengost JE, LeGallo RD, Boyd JC et al. Comparison of diagnostic accuracies in outpatients and hospitalized patients of D-dimer testing for the evaluation of suspected pulmonary embolism. Clin Chem 2003; 49: 1483–1490. 47 Le Gal G, Righini M, Roy PM et al. Value of D-dimer testing for the exclusion of pulmonary embolism in patients with previous venous thromboembolism. Arch Intern Med 2006; 166: 176–180.


BLUK077-Stein

April 6, 2007

54

16:8

CHAPTER 54

D-dimer combined with clinical probability for exclusion of acute pulmonary embolism

Combining a negative rapid ELISA (enzyme-linked immunosorbent assay) with a low or moderate clinical probability for pulmonary embolism (PE) essentially rules out the diagnosis of PE [1]. The non-ELISA assays when combined with a low clinical probability

for PE (but not a moderate probability) also provide a reasonable certainty of ruling out PE [1]. Figure 54.1 shows the relation of posttest probability to pretest probability using negative likelihood ratios according to Bayesian analysis [2]. Posttest probabilities with a

(a)

100 80 Whole blood

60 40 20

Quantitative rapid ELISA 95

85

75

65

55

45

35

25

15

0 5

D-dimer posttest probability (%)

120

Pretest probability (%)

20 18 16 14 12

Whole blood

10 8 6 4

Quantitative rapid ELISA

2

Pretest probability (%)

250

40

35

30

25

20

15

10

0 5

D-dimer posttest probability (%)

(b)

Figure 54.1 (a) Relation of posttest probability to pretest probability. (Data from Stein et al. [1].) (b) Enlargement of part (a).


BLUK077-Stein

April 6, 2007

CHAPTER 54

16:8

251

D-dimer and clinical probability for exclusion of PE

Table 54.1 Negative D-dimer with clinical evaluation to exclude PE: 3-month or 6-month outcome. PE/N (%)

First author

Wells score

Wells score

Wells score

Wells score

Wells score

(Ref.)

D-dimer

<2 (low)

≤ 4 (unlikely)

2–6 (moderate)

>4 (likely)

>6 (high)

Wells [3]

Whole blood

2/73 (2.7)

2/118 (1.7)

2/69 (2.9)

3/29(10.3)

1/5 (20)

Kearon [4]

Whole blood

1 or 2 /373

Christopher

Rapid ELISA

4/1026(0.4)†

Rapid ELISA

Geneva score

(0.3–0.5)* Group [5] Perrier [6]

0–8 (low or moderate) 0/220 (0) *6 mo outcome (1 PE and 1 “VTE,” 2 additional DVT on testing). † In addition, 1/1026 (0.1%) had DVT on follow-up. PE, pulmonary embolism; DVT, deep venous thrombosis; VTE, venous thromboembolsim; ELISA, enzyme-linked immunosorbent assay.

negative D-dimer assessed with the whole blood agglutination assay are shown in comparison with the quantitative rapid ELISA (Figure 54.1). If the pretest clinical probability for PE is high, for example, 70% or higher, the probability of PE, even with a negative D-dimer, irrespective of whether measured with whole blood agglutination or quantitative ELISA, is greater than 20%. Clearly, a high-clinical probability, irrespective of a negative D-dimer, indicates a need for further testing. The advantage of a negative quantitative rapid ELISA compared to a negative whole blood agglutination assay becomes apparent with a low or moderate clinical probability [1]. An enlargement (Figure 54.1b) shows that when the clinical probability is 4–15%, which is typical of a low-probability clinical assessment, the posttest probability with D-dimer testing by whole blood agglutination may be as high as 6%, but with the rapid ELISA it is 2%. With a 29–38% clinical probability, which is typical of a moderate-probability assessment, the posttest probability with whole blood agglutination may be as high as 16%, but with the rapid ELISA it is approximately 7% (Figure 54.1b). The prevalence of PE with levels of clinical assessment is shown in Chapter 52 “Empirical assessment and clinical models for diagnosis of acute pulmonary embolism.” The quantitative rapid ELISA method per-

mits a reasonably safe exclusion of PE with a lowor moderate-probability clinical assessment, but the whole blood agglutination may give uncertain results with a moderate-probability clinical assessment. Tests of whole blood D-dimer in combination with clinical assessment in small numbers of patients showed that PE occurred in 2.7% of the patients with a low-probability Wells score (score <2) on 3-month follow-up and PE occurred in 0.3–1.7% with an “unlikely” probability (score ≤ 4) [3, 4] (Table 54.1). Results were better in larger investigations using the rapid ELISA [4, 5]. With the D-dimer rapid ELISA in combination with an unlikely probability by Wells criteria, PE on 3-month follow-up occurred in 0.4% and an additional 0.1% had deep venous thrombosis (DVT) [5]. With the D-dimer rapid ELISA in combination with a low or moderate probability by original Geneva criteria, none had PE on 3-month follow-up [6]. The criteria for the Wells clinical assessment and the Geneva clinical assessment are described in Chapter 52 “Empirical assessment and clinical models for diagnosis of acute pulmonary embolism.” Among patients with suspected acute PE, a negative rapid ELISA in combination with an unlikely probability clinical assessment by Wells criteria or either a low or moderate clinical assessment by Geneva criteria occurred in 31–32% of the patients [5, 6] (Figure 54.2).


BLUK077-Stein

April 6, 2007

16:8

252

PART III

Diagnosis of acute PE

References

PE excluded Further tests necessary

31−32%

68−69%

Figure 54.2 Proportion of patients with suspected pulmonary embolism (PE) in whom the diagnosis of PE was excluded by a negative rapid ELISA D-dimer plus low or moderate /original Geneva score or unlikely clinical probability by the simplified Wells score. (Data from Christopher Study Investigators [5] and Perrier et al. [6].)

1 Stein PD, Hull RD, Patel KC et al. D-dimer for the exclusion of deep venous thrombosis and acute pulmonary embolism: a systematic review. Ann Intern Med 2004; 140: 589–602. 2 Sox HC. Commentary. Ann Intern Med 2004; 140: 602. 3 Wells PS, Anderson DR, Rodger M et al. Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer. Thromb Haemost 2000; 83: 416– 420. 4 Kearon C, Ginsberg JS, Douketis J et al. An evaluation of D-dimer in the diagnosis of pulmonary embolism: a randomized trial. Ann Intern Med 2006; 144: 812–821. 5 van Belle A, Buller HR, Huisman MV et al; Christopher Study Investigators. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing and computed tomography. JAMA 2006; 295: 172–179. 6 Perrier A, Roy P-M, Sanchez O et al. Multidetectorrow computed tomography in suspected pulmonary embolism. N Engl J Med 2005; 352: 1760–1768.


BLUK077-Stein

April 14, 2007

55

14:58

CHAPTER 55

D-dimer in combination with amino-terminal pro-B-type natriuretic peptide for exclusion of acute pulmonary embolism

Fragments of pro-B-type natriuretic peptide (pro BNP), B-type natriuretic peptide (BNP), and aminoterminal B-type natriuretic peptide (NT-proBNP) are released in response to myocardial stretch and volume overload [1]. Elevations of natriuretic peptide have been described in patients with pulmonary embolism (PE) [2–5], particularly in those with right ventricular strain [2]. On the basis of indicating right ventricular overload, both BNP and NT-proBNP have been used to predict adverse outcomes in patients with PE [3– 5]. It was speculated that a low value of NT-proBNP in combination with a low value of D-dimer would strengthen the validity of an excluded diagnosis of PE over exclusion of PE on the basis of a low value of D-dimer alone [6]. A preliminary investigation calls attention to a potential advantage of using these two tests in combination [6]. Results, however, are not yet clinically applicable. D-dimer, measured by a quantitative latex agglutination test, and NT-proBNP were assessed in only 19 patients with dyspnea who were found to have PE and in 199 patients with dyspnea in whom PE was not necessarily suspected [6]. The patients in the control group most commonly had a pulmonary infection, exacerbation of chronic obstructive pulmonary disease, or asthma [6]. Among 199 patients with no PE, specificity for a low D-dimer alone was 72% and for a low D-dimer or low NT-proBNP specificity was 85% [6].

However, the negative predictive values were similar. The negative predictive value of D-dimer alone was 143 of 144 (99%) and of NT-proBNP was 115 of 119 (97%) [6]. The negative predictive value of the combination of a low D-dimer and low NT-proBNP was 169 of 173 (98%).

References 1 Cowie MR, Struthers AD, Wood DA et al. Value of natriuretic peptides in assessment of patients with possible new heart failure in primary care. Lancet 1997; 350: 1349–1353. 2 Pruszczyk P, Kostrubiec M, Bochowicz A et al. N-terminal pro-brain natriuretic peptide in patients with acute pulmonary embolism. Eur Respir J 2003; 22: 649–653. 3 Kucher N, Printzen G, Doernhoefer T, Windecker S, Meier B, Hess OM. Low pro-brain natriuretic peptide levels predict benign clinical outcome in acute pulmonary embolism. Circulation 2003; 107: 1576–1578. 4 Kucher N, Printzen G, Goldhaber SZ. Prognostic role of brain natriuretic peptide in acute pulmonary embolism. Circulation 2003; 107: 2545–2547. 5 ten Wolde M, Tulevski, II, Mulder JW et al. Brain natriuretic peptide as a predictor of adverse outcome in patients with pulmonary embolism. Circulation 2003; 107: 2082– 2084. 6 Melanson SE, Laposata M, Camargo CA, Jr, et al. Combination of D-dimer and amino-terminal pro-B-type natriuretic peptide testing for the evaluation of dyspneic patients with and without acute pulmonary embolism. Arch Pathol Lab Med 2006; 130: 1326–1329.

253


BLUK077-Stein

April 6, 2007

56

16:10

CHAPTER 56

Low tissue plasminogen activator plasma levels and low plasminogen activator inhibitor-1 levels as an aid in exclusion of acute pulmonary embolism

D-dimer is one of the last products derived from the thrombotic/fibrinolytic process. Its levels rise when thrombin activates fibrinogen, fibrin is formed, and factor XIIIa stabilizes it [1]. Then tissue plasminogen activator (tPA) activates plasminogen to plasmin and plasmin splits fibrin. One of the fibrin degradation products is D-dimer [1, 2]. This fibrinolytic process is regulated by plasminogen activator inhibitor type-1 (PAI-1) inhibiting tPA in the circulation and alpha2antiplasmin inhibiting free plasmin [3]. Fibrinolytic system activation means that tPA must be released into the circulation [4]. This conclusion led to the investigation of tPA levels and PAI-1 levels in patients with suspected acute pulmonary embolism (PE) [4]. Plasma levels were measured using an enzyme-linked immunosorbent assay (ELISA). Flores and associates evaluated tPA levels and PAI-1 levels in 27 patients with acute PE and 39 patients with suspected PE in whom the diagnosis was excluded [4]. A cutoff level of 8.5 ng/mL tPA was used. In 13 of 39 (33%) in whom PE was absent, the tPA level was lower than 8.5 ng/mL. Sensitivity was 27 of 27 (100%) and negative predictive value was 13 of 13 (100%).

254

Levels of PAI-1 were investigated using a cutoff level of 15 ng/mL [4]. In 9 of 39 (23%) in whom PE was absent, the PAI-1 level was lower than 15 ng/mL [4]. Sensitivity was 27 of 27 (100%) and negative predictive value was 9 of 9 (100%). This preliminary investigation suggests that tPA levels and PAI-1 levels below a specified cutoff value may potentially be useful for the exclusion of the diagnosis of acute PE.

References 1 Garcia Frade LJ, Sureda A. Degradation products of fibrinogen/fibrin: methods of determination and clinical utility. Rev Iberoam Thromb Hemost 1991; 4: 89–98. 2 Lijnen HR, Collen D. Mechanisms of physiological fibrinolysis. Baillieres Clin Haematol 1995; 8: 277–290. 3 Booth NA, Bennett B. Fibrinolysis and thrombosis. Baillieres Clin Haematol 1994; 7: 559–572. 4 Flores J, Garcia-Avello A, Flores VM, Navarro JL, Canseco F, Perez-Rodriguez E. Tissue plasminogen activator plasma levels as a potential diagnostic aid in acute pulmonary embolism. Arch Pathol Lab Med 2003; 127: 310– 315.


BLUK077-Stein

March 28, 2007

57

15:45

CHAPTER 57

Echocardiogram in the diagnosis and prognosis of acute pulmonary embolism

Visualization of emboli within the right atrium, right ventricle, or pulmonary artery

Right ventricular enlargement or dysfunction: all patients or normotensive patients

Echocardiography has a low sensitivity for visualization of emboli within the right atrium, right ventricle, or pulmonary artery. Among any patient with pulmonary embolism (PE), pooled data showed emboli within the right ventricle or right atrium in 19 of 195 (10%) (Table 57.1) [1–3]. Thrombi in the right side of the heart or pulmonary artery were shown in 24 of 165 (15%) (Table 57.1, Figure 57.1) [1, 2]. Among patients with massive PE or unstable patients, the sensitivity of echocardiography for thrombi in the right atrium or ventricle ranged from 7 to 18% (Table 57.2) [4–6]. Patients with extremely mobile right heart thrombi that resembled a worm or snake had a high thrombusrelated mortality rate, 20 of 48 (42%) [7]. Among these, fatal PE occurred in 13 (27%), fatal paradoxical PE occurred in 1 (2%), and perioperative death from thrombectomy occurred in 6 (13%). Patients with more or less immobile right heart thrombi had a lower mortality. Among 57 patients, none had fatal PE and 2 (4%) died perioperatively at thrombectomy [7].

Echocardiographic evidence of right ventricular enlargement or dysfunction has been used for prognostic purposes and to guide therapy (see Chapter 89 on thrombolytic therapy). In-hospital mortality from PE among unselected patients and among normotensive patients with right ventricular dysfunction ranged from 5 to 13% compared with 0–1% among those with normal right ventricular function (Table 57.3) [8–10]. Pooled data showed an in-hospital mortality from PE in 23 of 222 (10%) with right ventricular dysfunction compared with 1 of 383 (0.3%) among those with normal right ventricular function [8–10]. Data based on right ventricular-to-left ventricular dimension ratios derived from CT angiography in normotensive, not critically ill patients, however, showed no increased in-hospital mortality in those with right ventricular enlargement (Stein PE, Beemath A, Matta F et al. Unpublished data from PIOPED II) (see Chapter 4). The cumulative 1-year mortality rate from acute PE was higher in patients with significant right ventricular hypokinesis [11].

Table 57.1 Intracardiac or pulmonary artery thrombi on echocardiography: any patient with pulmonary embolism. n/N (%) First author (Ref.)

RA or RV thrombi only sensitivit/

Cheriex [1]

10/60 (17)

10/60 (17)

Kasper [2]

4/105 (4)

14/105 (13)

Franzoni [3]

5/30 (17)

RA, RV, or PA thrombi sensitivity

5/30 (17)

RA, right atrium; RV, right ventricle; PA, pulmonary artery.

255


BLUK077-Stein

March 28, 2007

15:45

256

PART III

100

Diagnosis of acute PE

87

90 Sensitivity (%)

80 64

70 60 50 40 30

15

20 10

n = 393

n = 191

Echo RV Enlarged or dysfunction

Echo or US

n = 195

0

RA, RV, or PA thrombi

Figure 57.1 Sensitivity of echocardiographic findings for pulmonary embolism. Data based on information from Tables 57.1, 57.4, and 57.6. RA, right atrium; RV, right ventricle; PA, pulmonary artery; echo, echocardiogram; US, ultrasound.

Table 57.2 Thrombi within the right atrium or right ventricle on echocardiography: massive pulmonary embolism or unstable patient. First author (Ref.)

n/N (%): Sensitivity

Casazza [4]

23/130 (18)

Castillo-Fenoy [5]

4/34 (12)

Chapoutot [6]

12/170 (7)

Table 57.3 In-hospital mortality from pulmonary embolism according to the presence of right ventricular dysfunction. n/N (%) First Author (Ref.)

PE mortality: RV dysfunction

PE mortality: no RV dysfunction

Ribeiro [8]

9/70 (13)

Grifoni [9]

3/65 (5)

0/56 (0) 0/97 (0)

Kasper [10]

11/87 (13)

2/230 (1)

PE, pulmonary embolism; RV, right ventricle.

Table 57.4 All patients suspected with pulmonary embolism: echocardiographic right ventricular enlargement or dysfunction. n/N (%) First author (Ref.)

Sensitivity

Specificity

Positive predictive value

Negative predictive value

Grifoni [14]

32/63 (51)

Mansencal [15]

97/173 (56)

Perrier [16]

12/18 (67)

30/32 (94)

12/14 (86)

30/36 (83)

Miniati [17]

24/43 (56)

60/67 (90)

24/31 (77)

60/79 [76]

Cheriex [1]

46/46 (100)

Nazeyrollas [18]

29/31 (94)

32/39 (82)

29/36 (81)

32/34 (94)

Rudoni [19]

13/24 (54)

46/47 (98)

13/14 (93)

46/55 (84)


BLUK077-Stein

March 28, 2007

CHAPTER 57

15:45

257

Echocardiogram

Table 57.5 Echocardiograhic indices related to tricuspid regurgitation.

First author (Ref.)

TR velocity

TR velocity

TR velocity

TR velocity

TR color

> 2.5 m/s

≥ 2.7 m/s

≥ 3.0 m/s

moderate-severe

coded (1–4)

Grifoni [14] Mansencal [15] Perrier [16]

X

Miniati [17]

X

Cheriex [1]

X

Nazeyrollas [18]

X

X

Rudoni [19]

X

TR, tricuspid regurgitation.

Thirty-day all-cause mortality was 108 of 681 (16%) in PE patients with right ventricular dysfunction compared with 71 of 785 (9%) in PE patients with normal right ventricular function [12, 13]. Regarding the diagnosis of PE in all patients, the sensitivity of right ventricular enlargement or dysfunction ranged from 51 to 100% (Table 57.4) [1, 14–19]. On average, the sensitivity of right ventricular enlargement or dysfunction was 253 of 393 (64%), but it is recognized that the data are heterogeneous. The criteria used for right ventricular enlargement or dysfunction varied (Tables 57.5–57.7).

Echocardiogram plus venous ultrasound: all patients The sensitivity of echocardiographic findings of right ventricular enlargement or dysfunction combined with venous compression ultrasound in all patients

with suspected PE, on average, was 222 of 254 (87%) (Table 57.8, Figure 57.1) [15, 16]. If any two of clinical probability high, echocardiogram, and ultrasound were used, the sensitivity was 56 of 63 (89%) (Table 57.8, Figure 57.1) [14].

Right ventricular enlargement or dysfunction: massive PE or unstable patients The sensitivity of right ventricular enlargement or dysfunction for the diagnosis of PE in patients with massive PE or unstable patients, combining data from three small case series, was 33 of 33 (100%) (Table 57.9) [1, 15, 19]. Specificity and positive predictive value were not stated. Considering high-probability clinical assessment, echocardiogram, and ultrasound, if any 2 of 3 were positive, the sensitivity was 56 of 63 (89%) (Table 57.8) [14].

Table 57.6 Echocardiograhic indices related to RV dysfunction.

First author (Ref.)

RVEDD

RVEDD

>27 mm

>30 mm

Grifoni [14]

RVEDD high

RV/LV

RV/LV

RV/LV area ≥ 0.6 &

EDD > 0.5

EDD > 1.0

RV free wall < 7.0 mm

X

X

Mansencal [15]

X

Perrier [16] Miniati [17] Cheriex [1]

X X

X X

Nazeyrollas [18] Rudoni[19]

RV hypokinetic

X X

RV, right ventricle; LV, left ventricle; RVEDD, right ventricular end diastolic diameter.


BLUK077-Stein

March 28, 2007

15:45

Table 57.7 Echocardiograhic miscellaneous indices.

First author (Ref.)

IVS paradoxical

IVS systolic

Pulmonary acceleration

RV/RA gradient

IVC collapse

bulge

flattening

< 90 m/s

> 30 mm Hg

index < 40%

X

X

X

Grifoni [14] Mansencal [15]

X

Perrier [16] Miniati [17] Cheriex [1]

X

X

Nazeyrollas [18] Rudoni [19]

X

* Grifoni used any one criteria from Tables 57.5â&#x20AC;&#x201C;57.7. Mansencal required both criteria from Tables 57.5 and 57.6. Perrier required both criteria from Tables 57.5 and 57.6. Miniati and Rudoni used any two criteria from Tables 57.5â&#x20AC;&#x201C;57.7. IVS, intraventricular septum; IVC, inferior vena cava; RV, right ventricle; RA, right atrium. Table 57.8 All patients suspected with pulmonary embolism. n/N (%) Positive

Negative

First author (Ref.)

Basis for diagnosis

Sensitivity

Specificity

predictive value

predictive value

Grifoni [14]

Clin/echo/US*

56/63 (89)

40/54 (74)

56/70 (80)

40/47 (85)

Mansencal [15]

Echo or US

154/173 (89)

Perrier [16]

Echo or US

12/18 (67)

30/32 (94)

12/14 (86)

30/36 (83)

* Any 2 of 3. US, ultrasound.

Table 57.9 Massive pulmonary embolism or unstable patient: echocardiographic right ventricular enlargement or dysfunction. First author (Ref.)

Basis for Dx

n/N (%): sensitivity

Grifoni [14]

Clin/echo/US*

33/34 (97)

Mansencal [15]

Echo

17/17 (100)

Cheriex [1]

Echo

11/11 (100)

Rudoni [19]

Echo

5/5 (100)

* Any 2 of 3. US, ultrasound.

100 100

Sensitivity (%)

90 80 70 60 50 40 30 20 10 0

12

n = 334 RA or RV thrombi

n = 33 Echo RV dysfunction or enlargement

258

Figure 57.2 Sensitivity of echocardiogram in patients with massive pulmonary embolism (PE) or unstable from PE. Right atrial (RA) or right ventricular (RV) thrombi were shown in 12 patients and echocardiographic RV dysfunction or enlargement was shown in all patients. Data based on information from Tables 57.2 and 57.7.


BLUK077-Stein

March 28, 2007

CHAPTER 57

15:45

259

Echocardiogram

References 1 Cheriex EC, Sreeram N, Eussen YF, Pieters FA, Wellens HJ. Cross sectional Doppler echocardiography as the initial technique for the diagnosis of acute pulmonary embolism. Br Heart J 1994; 72: 52–57. 2 Kasper W, Meinertz T, Henkel B et al. Echocardiographic findings in patients with proved pulmonary embolism. Am Heart J 1986; 112: 1284–1290. 3 Franzoni P, Cuccia C, Zappa C, Volpini M, Gei P, Visioli O. Thromboembolus migrating into the right heart in pulmonary embolism. Echocardiographic and clinicotherapeutic aspects in 7 cases and review of the literature. G Ital Cardiol 1989; 19: 7–16. 4 Casazza F, Bongarzoni A, Centonze F, Morpurgo M. Prevalence and prognostic significance of right-sided cardiac mobile thrombi in acute massive pulmonary embolism. Am J Cardiol 1997; 79: 1433–1435. 5 Castillo-Fenoy A, Scheuble C, Benacerraf A, Brau J, Cereze P. Echographical symptomatology of acute pulmonary embolism. Arch Mal Coeur Vaiss 1980; 73: 971–978. 6 Chapoutot L, Metz D, Canivet E et al. Mobile thrombus of the right heart and pulmonary embolism: diagnostic and therapeutic problems. Apropos of 12 cases. Arch Mal Coeur Vaiss 1993; 86: 1039–1045. 7 European Working Group on Echocardiography. The European Cooperative Study on the clinical significance of right heart thrombi.. Eur Heart J 1989; 10: 1046– 1059. 8 Ribeiro A, Lindmarker P, Juhlin-Dannfelt A, Johnsson H, Jorfeldt L. Echocardiography Doppler in pulmonary embolism: right ventricular dysfunction as a predictor of mortality rate. Am Heart J 1997; 134: 479–487. 9 Grifoni S, Olivotto I, Cecchini P et al. Short-term clinical outcome of patients with acute pulmonary embolism, normal blood pressure, and echocardiographic right ventricular dysfunction. Circulation 2000; 101: 2817– 2822. 10 Kasper W, Konstantinides S, Geibel A, Tiede N, Krause T, Just H. Prognostic significance of right ventricular after-

11 12

13

14

15

16

17

18

19

load stress detected by echocardiography in patients with clinically suspected pulmonary embolism. Heart 1997; 77: 346–349. Ribeiro A. The role of echocardiography Doppler in pulmonary embolism. Echocardiography 1998; 15: 769–778. Kucher N, Rossi E, De Rosa M, Goldhaber SZ. Prognostic role of echocardiography among patients with acute pulmonary embolism and a sysytolic arterial pressure of 90 mm Hg or higher. Arch Intern Med 2005; 165: 1777–1781. Schoepf UJ, Kucher N, Kipfmueller F, Quiroz R, Costello P, Goldhaber SZ. Right ventricular enlargement on chest computed tomography: a predictor of early death in acute pulmonary embolism. Circulation 2004; 110(20): 3276– 3280. Grifoni S, Olivotto I, Cecchini P et al. Utility of an integrated clinical, echocardiographic, and venous ultrasonographic approach for triage of patients with suspected pulmonary embolism. Am J Cardiol 1998; 82: 1230– 1235. Mansencal N, Redheuil A, Joseph T et al. Use of transthoracic echocardiography combined with venous ultrasonography in patients with pulmonary embolism. Int J Cardiol 2004; 96: 59–63. Perrier A, Tamm C, Unger PF, Lerch R, Sztajzel J. Diagnostic accuracy of Doppler-echocardiography in unselected patients with suspected pulmonary embolism. Int J Cardiol 1998; 65: 101–109. Miniati M, Monti S, Pratali L et al. Value of transthoracic echocardiography in the diagnosis of pulmonary embolism: results of a prospective study in unselected patients. Am J Med 2001; 110: 528–535. Nazeyrollas P, Metz D, Jolly D et al. Use of transthoracic Doppler echocardiography combined with clinical and electrocardiographic data to predict acute pulmonary embolism. Eur Heart J 1996; 17: 779–786. Rudoni RR, Jackson RE, Godfrey GW, Bonfiglio AX, Hussey ME, Hauser AM. Use of two-dimensional echocardiography for the diagnosis of pulmonary embolus. J Emerg Med 1998; 16: 5–8.


BLUK077-Stein

March 11, 2007

58

10:5

CHAPTER 58

Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism

In 2004, CT (computed tomography) scanners, mostly single slice, were available in a larger proportion of nonfederal short-stay general and special hospitals in the United States than were radioisotope facilities (88% versus 61%) (Table 58.1) [1]. If one assumes that hospitals with <25 beds did not have CT, then 95% of the short-stay hospitals with ≥25 beds had CT. Multislice CT scanners were available in 39% of the short-stay hospitals [1]. Diagnostic ultrasound was available in 86% and magnetic resonance imaging was available in 59% of the short-stay hospitals. The distribution of sizes of short-stay nonfederal hospitals in the United States in 2004 was 7% with 6–24 beds, 20% with 25–49 beds, 21% with 50–99 beds, 23% with 100–199 beds, and 29% with ≥200 beds [1]. The proportion of hospitals with various imaging facilities differs in European countries. In the Netherlands in 1997, among 122 hospitals, facilities for ventilation–perfusion (V–Q) scans were available in about 57% of the hospitals, spiral CT in 61%, and pulmonary angiography in 89% [2]. In Austria in 1998,

Table 58.1 Proportion of short-stay nonfederal hospitals in the United States with computed tomography (CT), ventilation–perfusion (V–Q) lung scans, magnetic resonance imaging (MRI), and ultrasound. Percent of hospitals CT

88

Multislice CT

39

MRI

60

V–Q

61

Ultrasound

86

Data from American Hospital Association Hospital Statistics [1].

260

among 127 hospitals, ultrasound was available in 97%, V–Q scans in 19%, spiral CT in 54%, and pulmonary angiography in 59% [3]. In the United Kingdom, according to a survey published in 1999, 66% of 327 hospitals had facilities for V–Q scans, 44% had facilities for spiral CT, and 46% had facilities for pulmonary angiography [4]. In the United Kingdom, V–Q scans were frequently the only imaging test other than chest radiography [4]. It has been suggested that spiral CT replace V–Q scintigraphy as the primary modality for screening patients with suspected acute pulmonary embolism (PE) [5], particularly among hospitalized patients [6]. It has also been proposed that spiral CT be included in the workup of patients with nondiagnostic V–Q lung scans [6]. Among patients discharged from hospitals with a diagnosis of PE in 2003, 47% of the diagnostic tests for PE were with CT scans (Figure 58.1). In 2003, only 17% of the tests in patients discharged with PE were with V–Q scans, 24% were pulmonary angiograms, and venous ultrasound of the lower extremities was obtained in 13%. These data were assessed using the database of the National Hospital Discharge Survey from 1979 through 2001 [7] and updated through 2003 (unpublished). Whether venous ultrasound in patients discharged with PE served to identify deep venous thrombosis as a surrogate diagnosis for PE or whether there was an interest in diagnosing deep venous thrombosis in patients already diagnosed with PE is undetermined. We assumed that all imaging studies we reported were used for the diagnosis of PE in these patients [7]. This assumption is stronger for CT, V–Q scans, and pulmonary angiography. Tests in patients with suspected PE in whom the diagnosis was excluded were


BLUK077-Stein

March 11, 2007

CHAPTER 58

10:5

261

Trends in use of diagnostic imaging

Tests in patients with PE (%)

100

80 V−Q SCANS CT

60

40 ANGIOS 20 US 0 2003

2001

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

Year Figure 58.1 Relative use of diagnostic imaging tests in patients hospitalized with pulmonary embolism (PE) from 1979 through 2003. V–Q, ventilation–perfusion, US,

ultrasound, ANGIOS, pulmonary angiograms. (Modified and reprinted from Stein et al. [7], with permission from Elsevier.)

not included in these trends because there would be less certainty about whether CT was used to exclude PE. It is possible that pulmonary angiography was coded more consistently than other diagnostic tests.

3 Schibany N, Fleischmann D, Thallinger C et al. Equipment availability and diagnostic strategies for suspected pulmonary embolism in Austria. Eur Radiol 2001; 11: 2287– 2294. 4 Burkill, GJC, Bell JRG, Padley SPG. Survey on the use of pulmonary scintigraphy, spiral CT and conventional pulmonary angiography for suspected pulmonary embolism in the British Isles. Clin Radiol 1999; 54: 807–810. 5 Goodman LR, Lipchik RJ. Diagnosis of acute pulmonary embolism: time for a new approach. Radiology 1996; 199: 25–27. 6 Goodman LR, Lipchik RJ, Kuzo RS. Acute pulmonary embolism: the role of computed tomographic imaging. J Thor Imaging 1997; 12: 83–86. 7 Stein PD, Kayali F, Olson RE. Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism. Am J Cardiol 2004; 93: 1316–1317.

References 1 AHA Hospital Statistics, 2006 Edition. Health Forum LLC (An Affiliate of American Hospital Association), Chicago, 2006. 2 Hagen PJ, van Strijen MJ, Kieft GJ, Prins MH, Postmus PE. Availability of diagnostic facilities in the Netherlands for patients with suspected pulmonary embolism. ANTELOPE Study Group, Advances in New Technologies Evaluating the Localisation of Pulmonary Embolism. Neth J Med 2000; 57: 142–149.


BLUK077-Stein

March 11, 2007

59

10:14

CHAPTER 59

Techniques of perfusion and ventilation imaging

Introduction With the advent and ready availability of contrastenhanced spiral CT, the use of ventilation–perfusion (V–Q) lung scans for the diagnosis of acute pulmonary embolism (PE) has decreased. By 2001, the use of CT surpassed the use of V–Q scans [1], and unpublished updated information show that divergent trends have continued through 2003 (Chapter 58). Even so, V–Q scans are recommended for the diagnosis of acute PE in patients with severe allergy to iodine containing contrast material, patients with renal failure, women of reproductive age in whom radiation is a particular concern, and sometimes in patients in whom the CT angiogram gives an uncertain reading [2].

Perfusion imaging In perfusion lung imaging, the amount of microembolization of radiolabeled particles is proportional to the pulmonary arterial blood flow to that region [3]. Macroaggregated human albumin labeled with 99m Tc is used. The aggregated albumin preparation typically contains 90% of particles within the 10–90 μm range. It is estimated that there are over 600 million pulmonary arterioles small enough to trap the administered particles [4]. Because approximately 200,000– 500,000 particles are routinely injected for an individual study, the effect is physiologically insignificant. The arterioles are only temporarily occluded because the material degrades into smaller particles, the biologic half-time being about 6–8 hours [5]. The fragments are then phagocytized by cells of the reticuloendothelial system. The minimum toxic dose for albumin is 20 mg/kg [6]. The usual imaging dose is about 0.14 μg/kg albumin. At least 60,000 particles should be injected for an adequate study. Otherwise, an uneven distribution of

262

radioactivity in the vascular bed is likely [7]. Because the usual 2–4 mCi dose from most kits is distributed among approximately 500,000 particles, low-particle injections are rarely a problem. Perfusion imaging is begun by an intravenous injection of macroaggregates with the patient in the supine position to permit an even distribution of flow from the base of the lung to the apex, although there is a ventral to posterior gradient [8]. After injection, if a single-head camera is used, imaging is best performed with the patient erect, which allows maximum expansion of the lungs, especially at the lung bases. With a dual-headed camera, the patient is usually imaged in the supine position, which allows simultaneous anterior and posterior views and other 180◦ combination simultaneous views to be made. Anterior, posterior, both laterals, and both posterior and anterior obliques have been advocated [9–11]. Segmental delineation of the basal segments of both lower lobes is best seen in posterior oblique projections [11]. Posterior oblique views separate the two lungs, thereby avoiding “shine through” artifacts sometimes present in the lateral images. Oblique images contributed to the definition and clarification of the abnormalities, and resulted in improved lesion localization in 73% of patients [12].

Ventilation imaging Current methods for performing a ventilation study utilize a xenon gas, either 133 Xe or 127 Xe, radioactive krypton gas, 81m Kr, or a radioactive aerosol, usually the radioaerosol of 99m Tc diethylenetriamine pentaacetic acid (DTPA) or 99m Tc pyrophosphate (PYP). In addition, an ultra-fine dry dispersion of carbon “soot” that can be labeled with 99m Tc has been used, particularly in Europe and Australia [13]. Some authorities believe the ventilation study with 133 Xe can be effectively performed following the


BLUK077-Stein

March 11, 2007

CHAPTER 59

10:14

Techniques of perfusion and ventilation imaging

perfusion portion of the ventilation–perfusion study [14]. Carrying out the ventilation scan study after the perfusion scan would be convenient, because the ventilation scan need be performed only on those patients whose perfusion scans are indeterminate [15]. The experience of others, however, indicates that scattered irradiation from the previously administered perfusion tracer (99m Tc) significantly decreases the accuracy of the washout phase of the ventilation scan [8]. 127 Xe has photon energies higher than that of 99m Tc. As a consequence, 127 Xe ventilation studies can be performed after the perfusion scan. The uncertain availability of this tracer and its considerable cost have limited its general application [8].

Xenon ventilation studies The preperfusion 133 Xe ventilation lung scan has three phases: the single-breath image, the equilibrium phase, and the washout phase. If a single-head camera is used, the single-breath image is obtained with the patient initially positioned with his back to a wide field of view gamma camera (ideally erect, but often supine). A bolus of approximately 15–20 mCi of 133 Xe is injected into the mouthpiece of the spirometer system at a time when the patient begins a maximal inspiration. The patient holds his breath for the next 15–20 seconds and a single-breath image is obtained. Following this, equilibrium phase imaging is obtained with the patient breathing in a closed spirometer for 4–5 minutes. During this interval, a posterior and both right and left posterior oblique images are obtained. If there are any areas of the lungs that xenon cannot reach, this will be demonstrated on these views. If a dual-headed camera is used, simultaneous anterior and posterior images are obtained, and the oblique views are not needed. It is important to employ a tight-sealing facemask or mouthpiece at this point to ensure that complete equilibrium is reached [8]. The washout phase is begun at the end of this time by readjusting the intake valves of the spirometer system to permit the patient to inhale ambient air and exhale into a shielded charcoal trap that absorbs the expired radioactive gas. In general, the washout phase should last at least 5 minutes. With a single-head camera, images are taken at 30–60 second intervals during the course of the washout. Posterior oblique views approximately halfway through the washout process enhance the ability to detect and lo-

263

cate xenon retention in the anterior–posterior plane. These are not necessary if a dual-headed camera is employed. The equilibrium image has an activity distribution that corresponds to the aerated lung. Images made during washout should show a rapid clearance of activity from the lungs, usually within 90 seconds. Often, there is slight retention in the upper zones compared to the lower zones. In contrast to the single-breath image, which displays an abnormality as a deficit in radioactivity, abnormal washout images show 133 Xe “hot spots” that characterize the uneven distribution of ventilation. In this situation, abnormally retained 133 Xe activity (“air trapping”) is superimposed on a background of decreasing activity from normally ventilated lung regions. A sufficiently long rebreathing equilibrium time (at least 4 minutes) is imperative to maximize the usefulness of the washout phase by permitting the radiotracer to enter abnormal lung zones by collateral airdrift [8]. The washout phase of the 133 Xe study is the most likely phase to show ventilation abnormalities [16]. The major disadvantage of the xenon technique is that images of the lung are required in a preselected view, typically in the posterior projection, with some ancillary anteroposterior information added by oblique views. The simultaneous anterior and posterior views obtained with a dualheaded camera largely avoid this problem. Additional information at times can be added by obtaining a “second-look” single-breath xenon study in a different view. The typical analogue display of a Xe ventilation study for pulmonary embolism relies on a compromise intensity setting. The intensity is usually determined by experience, such that the early phases of the washout examination may be slightly intense relative to the later phases. The intensity of the later phases, at the time when the “unevenness” of abnormal xenon washout is best detected, will have an intensity set to optimize the abnormal appearance. In practice, however, it is sometimes difficult to adjust the washout intensity correctly, and other variables such as chest wall thickness, lung thickness (i.e., volume), and the depth and frequency of respiration may all affect the washout phase of the ventilation study. To avoid these problems, serial computerized washout images (e.g., 30–45 second intervals) may be modified by normalizing each image in the series. In this way, regions with


BLUK077-Stein

March 11, 2007

10:14

264

delayed clearance become more prominent, particularly in the later phases of the washout. There is no absolute normal range of xenon clearance times because of the normal variability in the rate and depth of respiration between subjects. The normal lungs are homogeneous in horizontal appearance, although there may be changes from apex to base. Inhomogeneity of xenon lung clearance at any time during the washout phase indicates an abnormality.

Krypton ventilation studies 81m

Kr is a relatively insoluble inert gas with a 13second half-life. It decays to 81 Kr by isomeric transition, emitting 190 keV gamma rays (65%) and internal conversion electrons. The gas is obtained by eluting a generator containing its parent 81 Rb, which has a 4.7-hour half-life. The short half-life of the 81 Rb generator is a disadvantage, since the generator contains only enough activity to be used for clinical studies on the day it is delivered. This markedly increases cost, because a new generator is required for each day’s work. Inhalation of 81m Kr results in images that demonstrate the distribution of regional tidal ventilation. Because of the ultrashort half-life of 13 seconds, by the time collateral air drift occurs, the 81m Kr has decayed away. There is, therefore, no opportunity to obtain washout images. The inability to obtain washout images is balanced by the fact that ventilation images can be obtained in multiple projections. This facilitates direct comparison with perfusion scans for the evaluation of ventilation–perfusion mismatches. 81m Kr images provide a sensitive means for studying regional ventilation [17, 18]. If the minute ventilation becomes greater than normal, 81m Kr regional count rates become more dependent on lung volume than on ventilation. This rarely, if ever, occurs in adults. Thus, 81m Kr images tend to underestimate ventilation in regions with high airflow because of potential partial dependence on the regional lung volume. Furthermore, because 81m Kr has a higher photon energy than 99m Tc, a low energy collimator cannot be used without getting some septal penetration. This decreases the resolution of the ventilation scan. If a medium energy collimator is used to avoid septal penetration, the collimator resolution is inherently less than a low energy collimator, and resolution of both the perfusion and ventilation scans decreases. This is also true for 127 Xe.

PART III

Diagnosis of acute PE

Radioaerosol inhalation studies Radioaerosols also provide a means for investigating regional ventilation and are commercially available with the introduction of small, easily contained, and efficient aerosol nebulizers. Radioaerosols are small particles, rather than gases. The radioaerosols deposit in the lung by impaction in central airways, sedimentation in more distal airways, and random contact with alveolar walls during diffusion in the alveoli [19]. The commercially available nebulizers produce submicron size aerosols that penetrate relatively evenly to the lung periphery, whereas larger (3–5 μm) radioaerosols, which were used in the past, demonstrated substantial degrees of central airway deposition, resulting in poor images of peripheral ventilation patterns. Currently, the radioaerosol most often employed is 99m Tc-diethylenetriamine pentaacetate (DTPA). DTPA is a low-molecular-weight solute that is able to cross the respiratory epithelium and be removed from the lungs rapidly through the bloodstream. Aerosols are typically used as a preperfusion ventilation scan with counts about one-fifth the counts of the perfusion scan so they do not “bleed through” on the perfusions images. Typically, if an eight-view perfusion scan is to be obtained, the aerosol ventilation images will be obtained in the same projections allowing precise comparison of ventilation and perfusion. However, good results have been shown with obtaining perfusion lung scans before ventilation scans when the ventilation scans were obtained with DPTA [20]. Performing the perfusion scan prior to the ventilation scan permitted the ventilation study to be tailored for optimal positioning to determine if mismatched defects were present. In addition, ventilation scans were unnecessary in patients in whom the perfusion scans were normal. Aerosolized 99m Tc-pyrophosphate has also been used and has the advantage of a longer lung residence time, which is especially important in smokers [21]. To yield uniform apex-to-base deposition, the inhalation of radioaerosol can be performed with the patient in the supine position. However, with a singlehead camera, if postperfusion aerosol images are obtained, and the perfusion defects are basilar, the erect position is often used. A nebulizer that generates submicron size particles should be filled with about 30 mCi of 99m Tc DTPA in 2–3 mL of saline, and the nebulizer


BLUK077-Stein

March 11, 2007

CHAPTER 59

10:14

Techniques of perfusion and ventilation imaging

should be aerated at a flow rate of 8–10 L/minute [21]. After approximately 3 minutes of breathing, about 750 mCi of 99m Tc aerosol will have been deposited in the lungs, yielding 100 K count images in approximately 2 minutes with a standard wide-field gamma camera with an all-purpose low-energy collimator. If postperfusion aerosol scanning is done, the aerosol dose in the nebulizer is increased to as much as 45 mCi of 99m Tc-pyrophosphate or DTPA and the inhalation of aerosol continues until the count rate obtained for the perfusion scan is doubled [21]. There is no readily apparent difference between the various agents used for ventilation scans from a diagnostic standpoint [22]. Cost, patient logistics, and referral patterns determine which ventilation agent is best suited for a specific institution [8]. Regarding the distribution of ventilation, in upright normal subjects, the lung bases ventilate better than the apices. In the lateral decubitus position, the lower lung ventilates better per unit volume than the upper lung. When the subject is supine, there is little apex-tobase gradient, but there is still an anterior–posterior gradient, with the most dependent regions receiving more ventilation than the uppermost regions.

Single photon emission computed tomographic perfusion lung scan Single photon emission computed tomography (SPECT) is discussed in Chapter 70.

References 1 Stein PD, Kayali F, Olson RE. Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism. Am J Cardiol 2004; 93: 1316–1317. 2 Stein PD, Woodard PK, Weg JG et al. Diagnostic pathways in acute pulmonary embolism: recommendations of the PIOPED II Investigators. Am J Med 2006; 119: 1048– 1055. 3 Stein PD, Gottschalk A. Critical review of ventilation/ perfusion lung scans in acute pulmonary embolism. Prog Cardiovasc Dis 1994; 37: 13–24. 4 Miller WS. The structure of the lungs. J Morphol 1893; 8: 165. Quoted by Dalen JE, Haynes FW, Hoppin FG, Jr, Evans GL, Bhardwaj P, Dexter L. Cardiovascular responses to experimental pulmonary embolism. Am J Cardiol 1967; 20: 3–9.

265

5 Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. 6 Taplin GV, MacDonald NS. Radiochemistry of macroaggregated albumin and newer lung scanning agents. Semin Nucl Med 1971; 1: 132–152. 7 Heck LL, Duley JW. Statistical considerations in lung imaging with 99 m-Tc albumin particles. Radiology 1974; 113: 675–679. 8 Gottschalk A, Alderson PO, Sostman HD. Nuclear medicine techniques and applications. In: Murray JF & Nadel JA, eds. Textbook of Respiratory Medicine, 2nd edn. WB Saunders, Philadelphia, 1994: 682–710. 9 Mack JF, Wellman HN, Saenger EL, Friedman BI. Clinical experience with oblique views in pulmonary perfusion camera-imaging in normal and pathological anatomy. Radiology 1969; 92: 897–902. 10 Mack JF, Wellman HN, Saenger EL. Oblique pulmonary scintiphotography in the analysis of perfusion abnormalities due to embolism. J Nucl Med 1969; 10: 420. 11 Caride VJ, Puri S, Slavin JD, Lange RC, Gottschalk A. The usefulness of posterior oblique views in perfusion lung imaging. Radiology 1976; 121: 669–671. 12 Nielsen PE, Kirchner PT, Gerber FH. Oblique views in lung perfusion scanning: clinical utility and limitations. J Nucl Med 1977; 18: 967–972. 13 Fawdry RM, Gruenewald SM. Initial experience with technegas: a new ventilation agent. Australas Radiol 1988; 32: 232–238. 14 Kipper MS. Alazraki N. The feasibility of performing 133 Xe ventilation imaging following the perfusion study. Radiology 1982; 144: 581–586. 15 Stein PD, Terrin ML, Gottschalk A, Alavi A, Henry JW. Value of ventilation/perfusion scans versus perfusion scans alone in acute pulmonary embolism. Am J Cardiol 1992; 69: 1239–1241. 16 Alderson PO, Biello DR, Khan AR, Barth KH, McKnight RC, Siegel BA. Comparison of 133 Xe single-breath and washout imaging in the scintigraphic diagnosis of pulmonary embolism. Radiology 1980; 137: 481– 486. 17 Alderson PO, Line BR. Scintigraphic evaluation of regional pulmonary ventilation. Semin Nucl Med 1980; 10: 218–242. 18 Miller TR, Biello DR, Lee JI et al. Ventilation imaging with Kr-81 m: A comparison with Xe-133. Eur J Nucl Med 1981; 6: 11–16. 19 Stuart BO. Deposition of inhaled aerosols. Arch Intern Med 1973; 131: 60–73. 20 Freitas JE, Sarosi MG, Nagle CC, Yeomans ME, Freitas AE, Juni JE. Modified PIOPED criteria used in clinical practice. J Nucl Med 1995; 36: 1573–1578.


BLUK077-Stein

March 11, 2007

10:14

266

21 Krasnow AZ, Isitman AT, Collier BD et al. Diagnostic applications of radioaerosols in nuclear medicine. In: Freeman LM, ed. Nuclear Medicine Annual 1993. Raven Press, New York, 1993: 123â&#x20AC;&#x201C;193.

PART III

Diagnosis of acute PE

22 Alderson PO, Biello DR, Gottschalk A et al. Tc-99 mDTPA aerosol and radioactive gases compared as adjuncts to perfusion scintigraphy in patients with suspected pulmonary embolism. Radiology 1984; 153: 515â&#x20AC;&#x201C;521.


BLUK077-Stein

April 14, 2007

60

15:0

CHAPTER 60

Ventilation–perfusion lung scan criteria for interpretation prior to the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED)

Introduction Scintigraphic studies of normal lungs demonstrate homogeneous patterns of matched ventilation and perfusion [1]. Nearly all pulmonary diseases, including neoplasms, infections, and chronic obstructive pulmonary disease, can produce decreased pulmonary blood flow to affected lung zones [2]. Pulmonary consolidation, bronchial obstructive disease, and restrictive disease as well as vascular occlusion disease may cause abnormalities on the perfusion lung scan as well as the ventilation lung scan. Vascular occlusion may result from neoplastic vascular compression and pulmonary vasculitis as well as pulmonary embolism (PE). Other processes occurring in the pulmonary arterial lumen can cause a ventilation–perfusion mismatch, including embolism of material other than thrombus (fat embolism and foreign material injected by intravenous drug abusers), in situ thrombosis (sickle cell disease), and pulmonary artery tumors [3–7]. Ventilatory defects may accompany the perfusion defect due to bronchial constriction caused from substances released from emboli or by reductions in alveolar levels of carbon dioxide, a potent airway dilator. This response, however, is usually transient and uncommon in scintigraphic studies [8–10]. Pulmonary infection, hemorrhage, infarction, atelectasis, or other conditions associated with alveolar collapse and/or leakage of fluids into the alveolar space cause perfusion and ventilation abnormalities. The degree of impaired ventilation and perfusion can vary, but a matched ventilation and perfusion defect typically is present [11].

Emphysema due to alpha1-antitrypsin deficiency and presumably due to other etiologies results in destructive loss of alveoli and the capillary bed, thereby affecting perfusion as well as ventilation [12]. The pulmonary arterioles also constrict with hypoxia [12]. Destruction of alveoli and the capillary bed as well as arteriolar constriction tend to cause matched ventilation and perfusion abnormalities [1]. Restrictive disease, particularly if caused by chronic fibrosis, usually results in a greater impairment of perfusion than ventilation [1]. A generalized abnormality on the chest radiograph such as diffuse pulmonary edema or diffuse reticulonodular disease may not cause the perfusion lung scan to be abnormal. Among 55 patients with diffuse radiographic opacities, 73% had normal or near-normal perfusion images [13]. A ventilation–perfusion scan in such patients, therefore, may assist in excluding PE, often guiding management decisions in these patients.

Diagnostic criteria for PE The scintigraphic hallmarks of PE are large, wedgeshaped, pleural-based perfusion defects in areas that ventilate normally and are radiographically clear [14]. Wagner and associates [15] and DeNardo and associates [16] suggested combined ventilation–perfusion lung imaging as a way to improve the specificity of diagnosing pulmonary emboli by scintigraphy. McNeil and coworkers [17] highlighted the findings of numerous investigators. They pointed out that in regions where the chest radiograph is normal, abnormalities in the

267


BLUK077-Stein

April 14, 2007

15:0

268

PART III

Diagnosis of acute PE

Table 60.1 Criteria for high probability in the interpretation of ventilation–perfusion scans for pulmonary embolism. Sullivan and Gottschalk [18] Biello [19]

Biello et al. [20]

Based on McNeil [21]

Hull et al. [22]

≥2 large V–Q mismatches

≥1 large V–Q mismatches

Multiple V–Q mismatches

≥1 segmental V–Q

with Nl X-ray

with Nl X-ray

mismatch

largest being greater than or equal to segmental

or

or

≥2 moderate V–Q

≥2 moderate V–Q

mismatches with Nl X-ray

mismatches with Nl X-ray

or

or

≥1 Q > X-ray

≥1 Q > X-ray

Mod: 25–90% Seg

Mod: 25–75% Seg

Large: >90% Seg

Large: >75% Seg

V–Q mismatch = Nl V in

V–Q mismatch = Nl V in

region of Q defect

region of Q defect

V–Q match = abnormal V in

V–Q mismatch = Nl or mildly reduced V with Nl X-ray in

V–Q match = abnormal V in

region of Q defect

region of Q defect

region of Q defect

Nl, normal; V–Q, ventilation–perfusion; X-ray, radiographic abnormality in region of perfusion abnormality; seg, segment. Reprinted from Stein and Gottschalk [18], with permission from Elsevier.

Table 60.2 Criteria for intermediate probability in the interpretation of ventilation–perfusion scans for pulmonary embolism. Sullivan and Gottschalk [18] Biello [19]

Biello et al. [20]

Based on McNeil [21]

Hull et al. [22]

≥1 Q = X-ray

≥1 Q = X-ray

1 segmental V–Q mismatch

≥1 subsegmental V–Q

or

or

or

or

1 large V–Q mismatch

1 moderate V–Q mismatch

≥1 Q = X-ray*

≥1 segmental V–Q

mismatch

with Nl X-ray

match

or

or

or

Severe COPD with Q

Severe COPD with Q defects

Mixed V–Q mismatched and

defects

matched defects or 1 V–Q mismatch ≥lobar size or Multiple subsegmental V–Q mismatches

* Our interpretation of data. Nl, normal; V–Q, ventilation–perfusion; X-ray, radiographic abnormality in region of perfusion abnormality. Reprinted from Stein and Gottschalk [18], with permission from Elsevier.


BLUK077-Stein

April 14, 2007

CHAPTER 60

15:0

269

V–Q scan diagnostic criteria before PIOPED

Table 60.3 Criteria for low probability in the interpretation of ventilation–perfusion scans for pulmonary embolism. Sullivan and Gottschalk [18] Biello [19]

Biello et al. [20]

Based on McNeil [21]

Hull et al. [22]

Small V–Q mismatches

Small V–Q mismatches

1 subsegmental V–Q

≥1 subsegmental V–Q

mismatch or

or

or

V–Q matches with Nl

V–Q matches with Nl

≥1 V–Q match

X-ray

match

X-ray

or

or

Q  X-ray

Q  X-ray

Small: <25% seg

Small: <25% seg

*Our interpretation of data. Nl, normal; V–Q, ventilation–perfusion; X-ray, radiographic abnormality in region of perfusion abnormality; seg, segment. Reprinted from Stein and Gottschalk [18], with permission from Elsevier.

perfusion scan that are matched by zones of abnormal ventilation are less likely to represent pulmonary emboli than the mismatched abnormalities (i.e., reduced perfusion with normal ventilation). Many studies have led to a gradual evolution of diagnostic criteria for scintigraphy. This was reviewed by Stein and Gottschalk [18]. Tables 60.1–60.3 list diagnostic criteria for the interpretation of ventilation–perfusion lung scans that were employed by various investigators or seem to be suggested by their observations [19–22]. The positive predictive values for PE with high, intermediate, low, and normal interpretations of ventilation–perfusion scans according to various proposed criteria were comparable (Table 60.4). When the chest radiograph is called “normal” this refers to the radiographic appearance in the region of the ventilation or perfusion defect and need not mean

Table 60.4 Probability of pulmonary embolism obtained with different criteria for the interpretation of ventilation–perfusion scans for pulmonary embolism. V–Q

Biello

Hull

probability

Biello [19]

et al. [20]

et al. [22]

High

90%

87–92%

86%

Intermediate

30%

20–33%

Low

10%

0–8%

Normal

0%

0%

21–40% 13% 0%

V–Q, ventilation–perfusion. Reprinted from Stein and Gottschalk [18], with permission from Elsevier.

the entire radiograph. Similarly, a chest radiographic defect indicates a radiographic opacity in the region related to the ventilation or perfusion defect. A lung zone is one-third of a lung divided craniocaudally (i.e., upper, middle, and lower zones). Definitions of a mismatched defect, matched defect, large segmental defect, and moderate size segmental defect varied among investigators (Table 60.1). McNeil showed likelihood ratios for several ventilation–perfusion scan patterns [21]. Based upon her data, others developed criteria for the interpretation of ventilation–perfusion scans [23, 24]. These criteria differ somewhat from each other. Neither of these developed criteria showed areas under receiver operating characteristic curves that differed significantly from criteria suggested by Biello and associates [19, 20, 23, 24]. A common cause of ventilation–perfusion mismatched defects not due to acute PE is persistent mismatched defects from prior PE. Rates and extent of resolution of perfusion lung scans are described in Chapter 6.

References 1 Gottschalk A, Alderson PO, Sostman HD. Nuclear medicine techniques and applications. In: Murray JF and Nadel JA, eds. Textbook of Respiratory Medicine, 2nd edn., WB Saunders Co, Philadelphia, 1994: Chap. 25. 2 Secker-Walker RH, Siegel BA. The use of nuclear medicine in the diagnosis of lung disease. Radiol Clin North Am 1973; 11: 215–241.


BLUK077-Stein

April 14, 2007

15:0

270

3 McNeil BJ. A diagnostic strategy using ventilation– perfusion studies in patients suspect for pulmonary embolism. J Nucl Med 1976; 17: 613–616. 4 Smith RR, Hutchins GM. Pulmonary fecal embolization complicating the Budd–Chiari syndrome. N Engl J Med 1978; 298: 1069–1070. 5 Sostman HD, Brown M, Toole A, Bobrow S, Gottschalk A. Perfusion scan in pulmonary vascular/lymphangitic carcinomatosis: the segmental contour pattern. Am J Radiol 1981; 137: 1072–1074. 6 Myerson PJ, Myerson DA, Katz R, Lawson JP. Gallium imaging in pulmonary artery sarcoma mimicking pulmonary embolism. Case report. J Nucl Med 1976; 17: 893–895. 7 Velchik MG, Tobin M, McCarthy K. Nonthromboembolic causes of high-probability lung scans. Am J Physiol Imaging 1989; 4: 32–38. 8 Thomas D, Stein M, Tanabe G, Rege V, Wessler S. Mechanism of bronchoconstriction produced by thromboemboli in dogs. Am J Physiol 1964; 206: 1207–1212. 9 Kessler RM, McNeil BJ. Impaired ventilation in a patient with angiographically demonstrated pulmonary emboli. Radiology 1975; 114: 111–112. 10 Epstein J, Taylor A, Alazraki N, Coel M. Acute pulmonary embolus associated with transient ventilatory defect case report. J Nucl Med 1975; 16: 1017–1020. 11 Finley TN, Tooley WH, Swenson EW, Gardner RE, Clements JA. Pulmonary surface tension in experimental atelectasis. Am Rev Respir Dis 1964; 89: 372–378. 12 Stein PD, Leu JD, Welch MH, Guenter CA. Pathophysiology of the pulmonary circulation in emphysema associated with alpha1 antitrypsin deficiency. Circulation 1971; 43: 227–239. 13 Newman GE, Sullivan DC, Gottschalk A, Putman CE. Scintigraphic perfusion patterns in patients with diffuse lung disease. Radiology 1982; 143: 227–231.

PART III

Diagnosis of acute PE

14 Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. 15 Wagner HN, Jr, Lopez-Majano V, Langan JK, Joshi RC. Radioactive xenon in the differential diagnosis of pulmonary embolism. Radiology 1968; 91: 1168–1174. 16 DeNardo GL, Goodwin DA, Ravasini R, Dietrich PA. The ventilatory lung scan in the diagnosis of pulmonary embolism. N Engl J Med 1970; 282: 1334–1336. 17 McNeil BJ, Holman BL, Adelstein SJ. The scintigraphic definition of pulmonary embolism. JAMA 1974; 227: 753–756. 18 Stein PD, Gottschalk A. Critical review of ventilation– perfusion lung scans in acute pulmonary embolism. Prog Cardiovasc Dis 1994; 37: 13–24. 19 Biello DR. Radiological (scintigraphic) evaluation of patients with suspected pulmonary thromboembolism. JAMA 1987; 257: 3257–3259. 20 Biello DR, Mattar AG, McKnight RC, Siegel BA. Ventilation–perfusion studies in suspected pulmonary embolism. Am J Radiol 1979; 133: 1033–1037. 21 McNeil BJ. Ventilation–perfusion studies and the diagnosis of pulmonary embolism: concise communication. J Nucl Med 1980; 21: 319–323. 22 Hull RD, Hirsh J, Carter CJ et al. Pulmonary angiography, ventilation lung scanning, and venography for clinically suspected pulmonary embolism with abnormal perfusion lung scan. Ann Intern Med 1983; 98: 891–899. 23 Webber MM, Gomes AS, Roe D, LaFontaine RL, Hawkins RA. Comparison of Biello, McNeil, and PIOPED criteria for the diagnosis of pulmonary emboli on lung scans. Am J Roentgenol 1990; 154: 975–981. 24 Sullivan DC, Coleman RE, Mills SR, Ravin CE, Hedlund LW. Lung scan interpretation: effect of different observers and different criteria. Radiology 1983; 149: 803– 807.


BLUK077-Stein

March 11, 2007

61

10:38

CHAPTER 61

Observations from PIOPED: ventilation–perfusion lung scans alone and in combination with clinical assessment

Introduction The Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) was a national collaborative study designed to determine the sensitivity and specificity of ventilation–perfusion lung scans in patients with suspected acute pulmonary embolism (PE) [1]. All patients in whom attending physicians suspected acute PE were asked to consent to undergo obligatory angiography if their ventilation–perfusion lung scan was abnormal. Among those who consented for participation in PIOPED, a random sample was selected for investigation. PE was diagnosed or excluded by pulmonary angiography or at autopsy. Patients with a normal ventilation–perfusion scan were not required to undergo a pulmonary angiogram. PE was excluded in such patients by the absence of PE during 1 year of follow-up while not receiving anticoagulant therapy. In addition to the randomized patients described in the primary PIOPED report [1], an arm of PIOPED included patients who were referred for pulmonary angiography by attending physicians. These nonrandomized patients were not included in the analysis of sensitivity and specificity reported in PIOPED [1], and these patients are not included in the data described in this chapter unless specifically stated. Useful information, however, was obtained from data in these patients. A number of secondary investigations based on PIOPED data included both arms of PIOPED. Ventilation studies were performed, preferably in the upright position with 5.6 × 108 − 11.1 × 108 Bq of xenon-133. They started with a 100,000-count, posterior view, first-breath image, and then posterior equi-

librium (wash-in) images for two consecutive 120second periods. Then, perfusion scans were obtained with 1.5 × 108 Bq of 99m Tc macroaggregated albumin that contained 100,000–700,000 particles. Particles were injected into an antecubital vein. The perfusion images consisted of anterior, posterior, both posterior oblique, and both anterior oblique views, with 750,000 counts per image for each. For the lateral view with the best perfusion, 500,000 counts per image were collected; the other lateral view was obtained for the same length of time. Two nuclear medicine readers, not from the center that performed the scan and blinded to all clinical and laboratory data, independently interpreted the lung scans with chest roentgenograms. The criteria for the interpretation of ventilation– perfusion scans used in PIOPED were developed on the basis of 2 or more mismatched segmental equivalent defects being indicative of a high probability of PE (Table 61.1). A mismatched segmental equivalent defect is 1 large mismatched segmental defect or 2 moderate size segmental defects. A large segmental defect is >75% of a segment. A moderate size segmental defect is ≥25% of a segment, but ≤75% of a segment (Figures 61.1 and 61.2). These criteria assume that a mismatched moderate size segmental equivalent defect is of less diagnostic value than a mismatched large segmental defect [2]. One of the difficulties with this system of grading is the interpretation of the size of the mismatched defect. The distinguishability of large versus moderate size segmental defects requires skill and judgment. Experienced readers of radionuclide lung scans often underestimate the size of segmental defects [3].

271


BLUK077-Stein

March 11, 2007

10:38

272

PART III

Diagnosis of acute PE

Table 61.1 PIOPED central scan interpretation categories and criteria. High probability ≥2 Large (>75% of a segment) segmental perfusion defects without corresponding ventilation or roentgenographic abnormalities or substantially larger than either matching ventilation or chest roentgenogram abnormalities. ≥2 Moderate segmental (25 and 75% of a segment) perfusion defects without matching ventilation or chest roentgenogram abnormalities and 1 large mismatched segmental defect. ≥4 Moderate segmental perfusion defects without ventilation or chest roentgenogram abnormalities. Intermediate probability (indeterminate) Not falling into normal, very low, low-, or high-probability categories. Borderline high or borderline low. Difficult to categorize as low or high. Low probability Nonsegmental perfusion defects (e.g., very small effusion causing blunting of the costophrenic angle, cardiomegaly, enlarged aorta, hila, and mediastinum, and elevated diaphragm). Single moderate mismatched segmental perfusion defect with normal chest roentgenogram. Any perfusion defect with a substantially larger chest roentgenogram abnormality. Large or moderate segmental perfusion defects involving no more than four segments in one lung and no more than three segments in one lung region with matching ventilation defects either equal to or larger in size, and chest roentgenogram either normal or with abnormalities substantially smaller than perfusion defects. >3 Small segmental perfusion defects (<25% of a segment) with a normal chest roentgenogram. Very low probability ≤ 3 Small segmental perfusion defects with a normal chest roentgenogram. Normal No perfusion defects present. Perfusion outlines exactly the shape of the lungs as seen on the chest roentgenogram (hilar and aortic impressions may be seen, chest roentgenogram and/or ventilation study may be abnormal). PIOPED, Prospective Investigation of Pulmonary Embolism Diagnosis. From PIOPED [1], with permission.

The PIOPED criteria for interpretation of ventilation–perfusion lung scans defined a “very low probability” ventilation–perfusion scan, but the results of PIOPED reported “nearly normal/normal” interpretations. The latter interpretation includes nearly normal ventilation–perfusion lung scans and entirely normal ventilation–perfusion lung scans. A nearly normal ventilation–perfusion lung scan category included readings of very low probability by one central scan reader and low probability by the other, very low probability by both, very low probability by one, and normal by the other. An entirely normal ventilation–perfusion lung scan was one that was agreed upon by both central scan readers as being normal (Figure 61.3). Agreement among scan readers was excellent in PIOPED for high-probability (95%), very low probability (92%), and normal (94%) scan categories [1]. For intermediate-probability (indeterminate) and

low-probability scan categories, the readers agreed less frequently (75 and 70%, respectively). The sensitivity of a high-probability ventilation– perfusion lung scan interpretation among 251 patients in PIOPED with angiographically diagnosed PE was 41% [1]. Sensitivity was defined as the percentage of patients with PE who had a high-probability ventilation–perfusion lung scan. The specificity of a high-probability ventilation–perfusion lung scan among 480 patients with negative pulmonary angiograms was 97% [1]. Specificity was defined as the percentage of patients with pulmonary angiograms free of signs of PE who had non-high-probability interpretations of the ventilation–perfusion lung scans. The frequency of PE among patients in whom the diagnosis was made or excluded entirely by pulmonary angiography differed somewhat from the frequency in patients in whom outcome analysis as well as pulmonary angiography were employed


BLUK077-Stein

March 11, 2007

10:38

(a)

Figure 61.1 (a) (Left) Chest radiograph, right lower zone, showing a mild parenchymal abnormality; (right) perfusion scan, anterior view, showing a perfusion defect (arrow) that is much larger than the radiographic abnormality. (b) (Left) Perfusion scan, posterior view, showing the same large perfusion defect seen in the anterior view (arrow); (right) ventilation scan, posterior view, equilibrium phase, showing nearly normal ventilation. The washout phase was normal. The ventilation–perfusion lung scan was interpreted as high probability for pulmonary embolism. A pulmonary angiogram showed a large embolism in the right lower zone. The perfusion scan was obtained with 99m Tc macroaggregated albumin. The ventilation scan was obtained with xenon-133. (Courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, MI.)

(b)

Perfusion

Anterior Ventilation

Perfusion

Posterior (a)

Figure 61.2 (a) (Left) Chest radiograph showing mild linear atelectasis at left base; (right) perfusion scan, anterior view, showing a massive perfusion defect (arrow) that is much larger than the radiographic abnormality. (b) (Left) Perfusion scan, posterior view, showing the same large perfusion defect (arrow) seen in the anterior view in Figure 61.2(a); (right) ventilation scan, posterior view, equilibrium phase, showing nearly normal ventilation. The ventilation–perfusion lung scan was interpreted as high probability for pulmonary embolism. A pulmonary angiogram showed pulmonary embolism. Technique of ventilation–perfusion scans was as in Figure 61.1. (Courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, MI.)

L Perfusion

(b)

L

L Perfusion

273

Ventilation


BLUK077-Stein

March 11, 2007

10:38

274

PART III

(a)

Anterior

Diagnosis of acute PE

Posterior

R

R

(b)

RPO

(Table 61.2). A larger number of patients with low probability or nearly normal interpretations of the ventilation–perfusion scans were included in the database when outcome analysis was employed in addition to pulmonary angiography. The following results refer to patients in whom outcome analysis in combination with pulmonary angiography was employed, although both sets of data are shown in Table 61.2. Among the general population of patients suspected of PE, a high-probability ventilation–perfusion scan using original PIOPED criteria was indicative of PE in 87% (Table 61.2) [1]. Among patients in whom the ventilation–perfusion scan probability interpretation was intermediate, PE was present in 30%. Therefore, the intermediate or indeterminate interpretation was uninformative. Among patients in whom a low-probability interpretation of the ventilation–perfusion lung scan was made, PE was present in 14%. Some physicians believe that this percentage of patients with PE who have a low-probability ventilation–perfusion scan is too high to adequately exclude PE [4]. Therefore, patients

LPO

Figure 61.3 Normal perfusion lung scan in a patient with borderline cardiomegaly and a prominent aorta. (a) (Left) Anterior view; (right) posterior view. (b) (Left) Right posterior oblique (RPO) view; (right) left posterior oblique (LPO) view. A pulmonary angiogram was negative for pulmonary embolism. The perfusion scan was obtained with Tc-99m macroaggregated albumin. (Courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, MI.)

Table 61.2 Ventilation–perfusion lung scan results: all randomized patients using original pioped criteria. #PE/#PTS (%) V–Q

Pulmonary

scan probability

angiography and

Diagnostic quality pulmonary

category

outcome analysis*

angiography only

High probability

103/118 (87)

102/116 (88)

Intermediate

104/345 (30)

105/322 (33)

40/296 (14)

39/238 (16)

5/128 (4)

5/55 (9)

probability Low probability Near normal/ normal Total

252/887 (28)

251/731 (34)

*Pulmonary embolism was diagnosed by pulmonary angiography in 248 patients and by autopsy in 4 patients. Pulmonary embolism was excluded by pulmonary angiography in 465 patients. Pulmonary embolism was excluded by outcome assessment (the absence of adverse events during 1-year follow-up while not receiving antibiotics) in 170 patients. PE: pulmonary embolism; PTS, patients; V–Q, ventilation– perfusion. Data are from PIOPED [1].


BLUK077-Stein

March 11, 2007

CHAPTER 61

10:38

275

V–Q scan results from PIOPED

with low-probability ventilation–perfusion scans require further diagnostic studies [4–6]. A nearly normal or normal ventilation–perfusion lung scan by original PIOPED criteria showed PE in only 4% of the patients (Table 61.2). Conversely, such an interpretation excluded PE in 96% of the patients. Among patients with nearly normal ventilation– perfusion lung scans included in both arms of PIOPED (patients who volunteered for random assignment to obligatory angiography and patients referred for angiography), the frequency of PE was 8 of 165 (5%) [7]. PE was diagnosed more frequently among patients with nearly normal ventilation–perfusion lung scans who underwent pulmonary angiography, 8 of 75 (11%), than in those in whom PE was diagnosed on the basis of an adverse outcome while receiving no anticoagulant therapy, 0 of 90 (0%) [7]. At first glance, this may suggest that recurrent PE was not apparent in patients with PE of such mild severity that the ventilation–perfusion scan was nearly normal. There is no evidence that this was the case, however [7]. In patients with nearly normal ventilation–perfusion scans who were assessed by outcome events in comparison to the patients assessed by angiography, the ventilation–perfusion scan showed fewer mismatched segmental perfusion defects, very low probability, or normal ventilation–perfusion interpretations by one of the two ventilation–perfusion readers as well as a generally lower clinical assessment [7]. The observed lower frequency of PE in patients evaluated by outcome events in comparison to those who underwent angiography, therefore, can be attributed to a lower likelihood of PE. A normal ventilation–perfusion scan entirely excluded PE in the experience of PIOPED and others [1, 8–10]. There were 21 patients in PIOPED who had ventilation–perfusion scans read as normal by both the central readers. Three underwent angiography and none showed thromboemboli. None of the remaining 18 patients received anticoagulants and none had clinically evident PE on follow-up. Sporadic case reports suggest that PE may occur in the presence of a normal perfusion lung scan. If these reports are correct, such cases are extremely rare. The major theoretical reasons for PE associated with a normal perfusion scan are central, nonobstructing, nonlateralized PE or minimal defects on the perfusion scan that are not appreciated [11]. Patients with a single, small, or partially occluding embolus may not show

a perfusion defect because of limitations of perfusion scanning. A history of prior PE diminished the positive predictive value of a ventilation–perfusion lung scan for acute PE. Among 19 patients with histories of prior PE and a high-probability ventilation–perfusion scan, the positive predictive value of the high-probability ventilation–perfusion scan was only 74% compared with 91% for those without a history of PE. This difference of positive predictive values reflected a loss of specificity in the high-probability ventilation– perfusion scan diagnosis for patients with histories of PE. The specificity of the ventilation–perfusion scan was only 88% among patients with a history of PE whereas it was 98% among patients with no prior PE [1]. In PIOPED, only 13% of the patients had highprobability ventilation–perfusion scans. Intermediateprobability ventilation–perfusion scan readings occurred in 39% of the patients (Figure 61.4). Lowprobability ventilation–perfusion scans occurred in 34% and nearly normal ventilation–perfusion scans in 14%. Among patients with PE, 41% showed a highprobability ventilation–perfusion scan, but some had low probability or even nearly normal ventilation– perfusion scans (Figure 61.5).

Clinical assessment in combination with lung scans One of the important contributions of PIOPED was the demonstration of the value of clinical assessment in combination with ventilation–perfusion lung scan. Clinical assessment used in combination with the findings of ventilation–perfusion lung scans strengthened the diagnostic value of the ventilation–perfusion scan [1]. In PIOPED, if both the independent clinical assessment and findings by ventilation–perfusion lung scans were high probability for PE, the diagnosis was correct in 97% of the patients (Table 61.3) [1]. If both the independent clinical assessment and findings by ventilation–perfusion lung scans were low probability, PE was present in only 4% of the patients. Unfortunately, these concordant diagnostic combinations were uncommon, occurring in only 28% of the patients with clinically suspected PE. Either clinical uncertainty or uncertainty regarding the ventilation–perfusion lung scan findings (i.e., intermediate ventilation–perfusion scan pattern) was present in 72% of the patients.


BLUK077-Stein

March 11, 2007

Near NL/NL 14%

Low 34%

High 41%

10:38

High 13%

Intermed 39%

Figure 61.4 Distribution of interpretations of ventilation–perfusion lung scans in all patients with suspected pulmonary embolism randomized for investigation in PIOPED. Intermed, intermediate; Near NL /NL, near normal/normal. (Data from PIOPED [1].)

Near normal 16%

Low 16%

Intermed 41%

Figure 61.5 Distribution of interpretations of ventilation–perfusion lung scans in patients with proven pulmonary embolism randomized for investigation in PIOPED. Intermed, intermediate; Near NL /NL, near normal/normal. (Data from PIOPED [1].)

Table 61.3 Clinical assessment and ventilation–perfusion scan probability in PIOPED. #PE/#PTS (%) V–Q scan probability category

Highly likely

High probability

28/29 (97)

70/80 (88)

5/9 (56)

Intermediate probability

27/41 (66)

66/236 (28)

11/68 (16)

6/15 (40)

30/191 (16)

4/90 (4)

0/5 (0)

4/62 (6)

1/61 (2)

Low probability Near normal/normal Total

61/90 (68)

Uncertain

170/569 (30)

Unlikely

21/228 (9)

Highly likely = 80–100% likelihood of PE based on clinical assessment; uncertain = 20–79% likelihood of PE based on clinical assessment; unlikely = 0–19% likelihood of PE based on clinical assessment. Pulmonary embolism was diagnosed by pulmonary angiography in 248 patients and by autopsy in 4 patients. Pulmonary embolism was excluded by pulmonary angiography in 465 patients. Pulmonary embolism was excluded by outcome assessment (the absence of adverse events during 1-year follow-up while not receiving antibiotics) in 170 patients. PE, pulmonary embolism; PTS, patients; V–Q, ventilation–perfusion. Modified from PIOPED [1].

276


BLUK077-Stein

March 11, 2007

CHAPTER 61

10:38

V–Q scan results from PIOPED

The probability of PE with various ventilation– perfusion scan probabilities combined with various concordant and discordant clinical suspicions is shown in Table 61.3.

References 1 A Collaborative Study by the PIOPED Investigators: Value of the ventilation/perfusion scan in acute pulmonary embolism: Results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 2 Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. 3 Morrell NW, Nijran KS, Jones BE, Biggs T, Seed WA. The underestimation of segmental defect size in radionuclide lung scanning. J Nucl Med 1993; 34: 370–374. 4 Hull RD, Raskob GE. Low-probability lung scan findings: a need for change. Ann Intern Med 1991; 114: 142–143. 5 Stein PD, Hull RD, Saltzman HA, Pineo G. Strategy for diagnosis of patients with suspected acute pulmonary embolism. Chest 1993; 103: 1553–1559.

277

6 Stein PD, Hull RD, Pineo G. Strategy that includes serial and noninvasive leg tests for diagnosis of thromboembolic disease in patients with suspected acute pulmonary embolism: estimated percentage of patients, based on data from PIOPED, in whom a noninvasive diagnosis or exclusion of thromboembolic disease might be safely made. Arch Intern Med 1995; 155: 2101–2104. 7 Henry JW, Stein PD, Gottschalk A, Raskob G. Pulmonary embolism among patients with a nearly normal ventilation/perfusion lung scan. Chest 1996; 110: 395– 398. 8 Stein PD. Low-dose heparin for prevention of pulmonary embolism and significance of normal lung scan. ACCP Bull 1982; 21: 12–14. 9 Hull RD, Raskob GE, Coates G, Panju AA. Clinical validity of a normal perfusion lung scan in patients with suspected pulmonary embolism. Chest 1990; 97: 23–26. 10 Kipper MS, Moser KM, Kortman KE, Ashburn WL. Long term follow-up of patients with suspected pulmonary embolism and a normal lung scan. Chest 1982; 82: 411– 415. 11 Stein PD, Gottschalk A. Critical review of ventilation/ perfusion lung scans in acute pulmonary embolism. Prog Cardiovasc Dis 1994; 37: 13–24.


BLUK077-Stein

April 6, 2007

62

16:14

CHAPTER 62

Ventilation–perfusion lung scans in patients with a normal chest radiograph, patients with no prior cardiopulmonary disease, patients with any prior cardiopulmonary disease, and patients with chronic obstructive pulmonary disease

Data from the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) included 133 patients with a normal chest radiograph, among whom only 20 had pulmonary embolism (PE) [1] (Table 62.1, Figure 62.1). Although the ventilation–perfusion (V–Q) lung scan was diagnostic in a higher proportion of patients with a normal chest radiograph than with more complex chest findings, still only 52% were diagnostic (high probability or nearly normal/normal) [1]. Others, however, showed diagnostic V–Q scans (high probability or normal) in 201 of 220 patients (91%) with a normal chest radiograph [2]. Based on data from PIOPED and using the original criteria for the evaluation of ventilation–perfusion scans from PIOPED, as the complexity of associated

cardiopulmonary disease increased, the proportion of patients with diagnostic interpretations decreased [1, 3, 4]. In patients with no prior cardiopulmonary disease, 37% of patients showed diagnostic V–Q scans and in patients with any cardiopulmonary disease, 23% showed diagnostic interpretations [3] (Table 62.1, Figure 62.1). With even more complex disease (chronic obstructive pulmonary disease, COPD), only 10% showed diagnostic V–Q scans [4] (Table 62.1, Figure 62.1). Among each of these categories of patients, the positive predictive value of high, low, and nearly normal ventilation–perfusion lung scans was similar [1, 3, 4] (Table 62.1). The sensitivities of high probability ventilation– perfusion scans and specificities of a nearly normal/

Table 62.1 V–Q scan findings in patients with various degrees of complexity of associated cardiopulmonary disease. V–Q scan

Normal chest

No prior

Any prior

COPD‡

probability

radiograph* [PE/n (%)]

CPD† [PE/n (%)]

CPD† [PE/n (%)]

[PE/n (%)]

High

6/9 (67)

5/54 (93)

55/66 (83)

4/17 (24)

47/119 (39)

58/227 (26)

14/65 (22)

Low

8/47 (17)

17/113 (15)

25/182 (14)

2/33 (6)

Near normal/NL

2/60 (3)

3/79 (4)

2/51 (4)

0/5 (0)

Intermediate

5/5 (100)

*Stein et al. [1], † Stein et al. [3], ‡ Lesser et al. [4]. V–Q, ventilation–perfusion; PE, pulmonary embolism; CPD, cardiopulmonary disease; COPD, chronic obstructive pulmonary disease; NL, normal.

278


BLUK077-Stein

April 6, 2007

CHAPTER 62

16:14

Normal chest X-ray

Low 35%

279

V–Q scan and normal chest radiograph

Any CPD

No prior CPD

Near NL/NL Low 31% 45%

COPD

Near NL/NL 22%

Near NL/NL 22%

High Low 13% 31%

Low 35%

Near NL/NL High 10% 10%

High 15% Intermed High 13% 7%

Intermed 33%

Intermed 43%

Intermed 60%

Figure 62.1 Distribution of interpretations of ventilation–perfusion scans among patients with a normal chest radiograph, no prior cardiopulmonary disease (CPD), patients with any prior cardiopulmonary disease, and patients with chronic obstructive pulmonary disease (COPD). As the complexity of disease increased, the

proportion of patients with diagnostic ventilation– perfusion scans (high probability or nearly normal/normal) decreased from 52 to 10%. Some of the distributions do not add to exactly 100% because percentages were rounded off. NL, normal; Intermed, intermediate. (Data from Stein et al. [1, 3] and Lesser et al. [4].)

Table 62.2 Sensitivity and specificity of V–Q scans in patients with a normal chest radiograph, patients with no prior cardiopulmonary disease, patients with any prior

cardiopulmonary disease, and patients with chronic obstructive pulmonary disease.

Normal chest radiograph*

Sensitivity high prob

Specificity normal or near normal V–Q

V–Q high/PE + (%)

V–Q normal or near NL/PE − (%)

9/20 (45)

60/113 (53)

No prior cardiopulmonary disease†

54/72 (75)

79/293 (27)

Any prior cardiopulmonary disease†

66/140(47)

51/386 (13)

5/21 (24)

5/87 (6)

COPD‡

*Stein et al. [1], † Stein et al. [3], ‡ Lesser et al. [4]. V–Q, ventilation–perfusion; PE, pulmonary embolism; prob, probability; COPD, chronic obstructive pulmonary disease.

normal V–Q scan are shown in patients with a normal chest radiograph, patients with no prior cardiopulmonary disease, patients with any prior cardiopulmonary disease, and patients with COPD [1, 3, 4]. With increasing complexity of pulmonary disease, in general, the sensitivity of a high-probability V–Q scan decreased and the specificity of a nearly normal/normal V–Q scan also decreased (Table 62.2).

References 1 Stein PD, Alavi A, Gottschalk A et al. Usefulness of noninvasive diagnostic tools for diagnosis of acute pulmonary

embolism in patients with a normal chest radiograph. Am J Cardiol 1991; 67: 1117–1120. 2 Forbes KP, Reid JH, Murchison JT. Do preliminary chest X-ray findings define the optimum role of pulmonary scintigraphy in suspected pulmonary embolism? Clin Radiol 2001; 56: 397–400. 3 Stein PD, Coleman RE, Gottschalk A, Saltzman HA, Terrin ML, Weg JG. Diagnostic utility of ventilation/perfusion lung scans in acute pulmonary embolism is not diminished by pre-existing cardiac or pulmonary disease. Chest 1991; 100: 604–606. 4 Lesser BA, Leeper KV, Stein PD et al. The diagnosis of pulmonary embolism in patients with chronic obstructive pulmonary disease. Chest 1992; 102: 17–22.


BLUK077-Stein

April 14, 2007

63

15:1

CHAPTER 63

Perfusion lung scans alone in acute pulmonary embolism

Ventilation lung scans are obtained in combination with perfusion lung scans with the thought that ventilation would be abnormal in areas of pneumonia or local hypoventilation, but would be normal in pulmonary embolism (PE) [1]. However, among 98 patients in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED I), the positive predictive values of high probability, intermediate probability, low probability, and near normal/normal interpretations of perfusion scans alone were comparable with the positive predictive values of ventilation– perfusion scans [2] (Table 63.1). A high probability perfusion scan had no less positive predictive value for acute PE than a high probability ventilation– perfusion scan [2]. Similarly, a low probability or near normal/normal perfusion scan excluded PE with no less validity than a low probability or near normal/normal ventilation–perfusion scan [2]. Original criteria for ventilation–perfusion scan interpretation from the PIOPED I were used [3]. Perfusion defects, in the absence of a ventilation scan, were interpreted on the basis of regional findings on the plain chest radiograph. Table 63.1 Positive predictive values of perfusion scans alone compared to ventilation–perfusion scans. Perfusion scan

V–Q scan

Scan interpretation

PE/n(%)

PE/n(%)

High

14/15 (93)

15/16 (94)

Intermediate

14/38 (37)

9/25 (36)

0/12 (0)

5/25 (20)

Low Near normal/normal

1/2 (50)

0/1 (0)

All differences between perfusions scan alone and ventilation–perfusion (V–Q) scans were not significant. PE, pulmonary embolism; n, number of patients; V–Q, ventilation–perfusion. Reprinted and modified from Stein et al. [2], with permission from Elsevier.

280

The positive predictive value of high probability interpretations of perfusion scans alone was 93%, which did not differ significantly from that of high probability interpretations of ventilation–perfusion scans 94% (Table 63.1) [2]. The positive predictive value for PE among patients with low probability interpretations of perfusion lung scans alone, 0%, did not differ significantly from that of low probability interpretations of ventilation– perfusion scans, 20% [2] (Table 63.1). Few patients had nearly normal/normal ventilation–perfusion or perfusion scans. Comparisons in this category, therefore, were not meaningful. Fewer patients who had perfusion scans alone had low probability interpretations than did those with ventilation–perfusion scans (23% versus 38%) and a trend suggested that more patients with perfusion scans alone had intermediate probability interpretations than did those with ventilation–perfusion scans [2] (Table 63.2). The sensitivity of high probability perfusion scans alone, 48%, was comparable with that of high probability ventilation–perfusion scans, 52%. The specificity of

Table 63.2 Distribution of lung scan interpretations in 98 patients with perfusion scans alone and ventilation– perfusion scans. Perfusion scan

V–Q scan

Scan interpretation

(%)

(%)

High

17

18

Intermediate

49

35

Low

23*

38

Near normal/normal

10

9

*P < 0.05. V–Q, ventilation–perfusion. Reprinted and modified from Stein et al. [2], with permission from Elsevier.


BLUK077-Stein

April 14, 2007

CHAPTER 63

15:1

281

Perfusion lung scans alone in acute PE

high probability perfusion scans alone, 97%, was identical to that of high probability ventilation–perfusion scans [2]. Investigators in the PISA-PED study used only perfusion lung scans; they did not use ventilation lung scans [4]. A perfusion scan compatible with PE was defined as a scan with a single or multiple wedge-shaped perfusion defects of any size. Diversion of blood away from an underperfused area usually resulted in associated overperfused single or multiple wedge-shaped areas. A perfusion scan was considered negative for PE if single or multiple perfusion defects did not exhibit a wedge shape. Positive perfusion scans had a positive predictive value of 95%. A negative perfusion scan had a negative predictive value of 81%. Combining clinical assessment with the perfusion scan showed good results when clinical assessment and the perfusion scan reading were concordant. Only the minority of patients (21%) had discordant clinical and perfusion scan assessments, and these patients required pulmonary angiography.These results at first glance appear better than the results of PIOPED. In the PISA-PED study, however, 34% of the patients had normal perfusion

scans and 61% had normal or nearly normal perfusion scans, whereas in PIOPED, only 2% had normal scintiscans and 14% had normal or nearly normal perfusion scans. The two study populations, therefore, were dissimilar.

References 1 Wagner HN, Jr, Lopez-Majano V, Langan JK, Joshi RC. Radioactive xenon in the differential diagnosis of pulmonary embolism. Radiology 1968; 91: 1168–1174. 2 Stein PD, Terrin ML, Gottschalk A, Alavi A, Henry JW. Value of ventilation/perfusion scans versus perfusion scans alone in acute pulmonary embolism. Am J Cardiol 1992; 69: 1239–1241. 3 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism—results of the prospective investigation of pulmonary embolism diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 4 Invasive and noninvasive diagnosis of pulmonary embolism. Preliminary results of the Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis (PISAPED). Chest 1995; 107: 33S–38S.


BLUK077-Stein

March 28, 2007

64

15:49

CHAPTER 64

Probability interpretation of ventilation–perfusion lung scans in relation to largest pulmonary arterial branches in which pulmonary embolism is observed

In the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED I) patients, with high probability interpretations of V–Q scans were likely to have PE (pulmonary embolism) in main or lobar pulmonary arteries, whereas patients with low probability V–Q scans had a higher prevalence of PE limited to subsegmental pulmonary arterial branches [1]. The prevalence of PE according to the largest pulmonary artery involved is shown for patients with high, intermediate, and low probability interpretations of V–Q scans (Table 64.1). Eighty-five percent of patients with a high probability interpretation of a V–Q scan had PE in main or lobar pulmonary arteries (Table 64.1) [1]. Conversely, those with low probability interpretations had PE in main or lobar pulmonary arteries in only 31%. Patients with a low probability interpretation of the V–Q lung scan had a higher prevalence of PE lim-

ited to the subsegmental branches, 17%, than patients with an intermediate probability interpretation, 6%, or patients with a high probability interpretation, 1%. These data are further analyzed in patients according to the presence or absence of prior cardiopulmonary disease (Table 64.2) [1]. Patients with low probability interpretations of the V–Q scan and no prior cardiopulmonary disease had PE limited to the subsegmental pulmonary arteries in 30% whereas patients with low probability V–Q scans who had prior cardiopulmonary disease had PE limited to subsegmental pulmonary arteries in 8%. Oser and associates showed PE limited to subsegmental pulmonary arteries or smaller in 23 of 76 (30%) patients with angiographically diagnosed PE [2]. Goodman and associates showed PE limited to subsegmental pulmonary arteries in 4 of 11

Table 64.1 Largest pulmonary artery with pulmonary embolism in relation to the ventilation–perfusion lung scan. Interpretation

Main or lobar [n/N(%)]

High prob V–Q

126/148 (85)

21/148 (14)

1/148 (1)

71/160 (44)

80/160 (50)

9/160 (6)

18/59 (31)

31/59 (53)

217/375 (58)

136/375 (36)

Intermediate prob V–Q Low prob V–Q All patients

Segmental [n/N(%)]

Subsegmental [n/N(%)]

10/59 (17)*† 22/375 (6)

*P < 0.01 low V–Q vs. intermediate V–Q. † P < 0.001 low V–Q vs. high V–Q. High prob V–Q, high probability interpretation of ventilation–perfusion (V–Q) lung scan; intermediate prob V–Q, intermediate probability interpretation of V–Q; low prob V–Q, low probability interpretation of V–Q. Reprinted with permission from Stein et al. [1].

282


BLUK077-Stein

March 28, 2007

CHAPTER 64

15:49

283

Probability interpretation according to largest pulmonary arter involved

Table 64.2 Largest pulmonary artery with pulmonary embolism according to ventilation–perfusion scan among patients stratified for prior cardiopulmonary disease. Interpretation

Main or lobar [n/N(%)]

Segmental [n/N(%)]

Subsegmental [n/N(%)]

No CPD

56/68 (82)

12/68 (18)

0/68 (0)

CPD

70/80 (88)

9/80 (11)

1/80 (1)

No CPD

29/73 (40)

40/73 (55)

4/73 (5)

CPD

40/83 (48)

39/83 (47)

4/83 (5)

7/23 (30)

9/23 (39)

7/23 (30)

11/36 (31)

22/36 (61)

3/36 (8)*

High prob V–Q

Intermediate prob V–Q

Low prob V–Q No CPD CPD

*P < 0.05 low probability V–Q with no CPD vs. low probability V–Q with CPD. High prob V–Q, high probability interpretation of ventilation–perfusion (V–Q) lung scan; intermediate prob V–Q, intermediate probability interpretation of V–Q; low prob V–Q, low probability interpretation of V–Q; CPD, cardiopulmonary disease. Reprinted with permission from Stein et al. [1].

(36%) patients with angiographically diagnosed PE [3]. Quinn and associates reported PE limited to subsegmental pulmonary arteries in 2 of 20 (10%) patients with PE [4]. The prevalence that we observed of PE limited to subsegmental pulmonary arteries (6%) was lower than reported by Oser and associates (30%) and by Goodman and associates (36%), but was comparable to the frequency reported by Quinn and associates (10%) [2–4]. Goodman and associates also reported a higher prevalence of low probability interpretations of the V–Q scan among patients with PE, 5 of 11 (45%) than observed in PIOPED, 59 of 375 (16%) [3]. The patients in PIOPED, therefore, may have had more severe PE than the patients reported by Goodman and associates.

References 1 Stein PD, Henry JW. Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries and relation to probability interpretation of ventilation/ perfusion lung scans. Chest 1997; 111: 1246–1248. 2 Oser RF, Zuckerman DA, Gutierrez FR, Brink JA. Anatomic distribution of pulmonary emboli at pulmonary angiography: implications for cross-sectional imaging. Radiology 1996; 199: 31–35. 3 Goodman LR, Curtin JJ, Mewissen MW et al. Detection of pulmonary embolism in patients with unresolved clinical and scintigraphic diagnosis: helical CT versus angiography. Am J Roentgenol 1995; 164: 1369–1374. 4 Quinn MF, Lundell CJ, Klotz TA et al. Reliability of selective pulmonary arteriography in the diagnosis of pulmonary embolism. Am J Roentgenol 1987; 149: 469–471.


BLUK077-Stein

March 11, 2007

65

11:1

CHAPTER 65

Revised criteria for evaluation of lung scans recommended by nuclear physicians in PIOPED

After the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) was concluded, nuclear physicians in PIOPED reevaluated the original criteria for interpretation of ventilation–perfusion lung scans [1]. They correlated probability estimates of pulmonary embolism (PE) based on the ventilation– perfusion lung scans with the actual frequency of PE as determined by pulmonary angiography. The nuclear physicians in PIOPED determined that most of the original PIOPED criteria appropriately categorized ventilation–perfusion scans [1]. However, it was recommended that three criteria should be reconsidered (Table 65.1):

1 Two large segmental mismatches may not be the optimum threshold for high probability, and in some cases should be considered for intermediate probability. However, due to the small number of cases with this finding, no definite, statistically founded recommendation could be made. 2 A single moderate mismatched perfusion defect is appropriately categorized as intermediate, rather than as low probability. 3 Extensive matched ventilation–perfusion abnormalities are appropriate for low probability, provided that the chest radiograph is clear. Single matched defects may be better categorized as intermediate probability. Due to the small number of cases with this finding, no definite, statistically founded recommendation could be made. It was suggested that the revised criteria resulting from these adjustments should now be used for the interpretation of ventilation–perfusion scans [1]. Analyses focused upon individual PIOPED criteria. In many instances, combined patterns were excluded. Combined patterns involving mismatched

284

defects were considered in the analysis of criteria for “high probability,” since mismatched defects and high probability diagnoses took priority over other patterns in the PIOPED criteria. For example, a patient with three segmental mismatches (which meets criteria for high probability) and matched ventilation–perfusion defects (which meet the criteria for low probability) should have been assigned to the high probability category. Therefore, all patients with mismatched perfusion defects, including those with combined patterns, were considered in the analysis of high probability [1]. When evaluating low probability criteria, since these did not take precedence in scan categorization, it was necessary to isolate analyses to those patients who did not have scan findings that would place them in a higher category [1]. Correlations of scintigraphic patterns that fulfilled individual PIOPED criteria were made with pulmonary angiograms read as definitely positive or definitely negative for PE [1]. There were 731 patients in PIOPED in the randomized obligatory pulmonary angiography group who had definitive angiographic results [2]. Correlations of ventilation–perfusion scans with pulmonary angiograms were made in 393 of those patients (54%) who satisfied pertinent individual PIOPED criteria on their ventilation–perfusion lung scans [1]. Analyses of the PIOPED criteria for the evaluation of ventilation–perfusion scans arbitrarily defined the following ranges:

1 High probability = 80–100% likelihood of PE. 2 Intermediate probability = 20–79% likelihood of PE. 3 Low probability = 0–19% likelihood of PE. Two large mismatched perfusion defects did not provide a reliable interpretation of high probability,


mismatch

V–Q are borderline

285

mismatches

mismatches

normal perfusion in lung

≥1 small Q with Nl X-ray

Nl X-ray

very low probability ≤3 small Q with

or

>3 small Q with Nl X-ray

or

elevated diaphragm

aorta, hilum, mediastinum, and

effusion, cardiomegaly, enlarged

Perfusion defects due to pleural

elevated diaphragm or

or

mediastinum, and

aorta, hilum,

probability

cardiomegaly, enlarged

region

borderline intermediate to high

pleural effusion,

lung and ≤3 segment in 1 lung

mismatches are

Perfusion defects due to

or

 Q involving ≤4 segments in 1

V–Q matches with Nl X-ray or X-ray

or

Q  X-ray

X-ray and some areas of

or ≥2 V–Q matches with Nl

or

≥1 Q  X-ray defect

PIOPED revised (1)

X-ray (2 large V–Q

1 matched V–Q with Nl

or

mismatches

1–3 moderate V–Q

or

ray

1 moderate V–Q mismatch with Nl X-

1 large ±1 moderate V–Q mismatch

PIOPED (2)

PIOPED revised (1)

Low probability

Nl, normal; V–Q, ventilation–perfusion; X-ray, radiographic abnormality in region of perfusion abnormality; mod, moderate; seg, segment. Moderate = 25–75% segment; large ≥75% segment; V–Q mismatch = Nl V and Nl X-ray or Q > V or X-ray in region of Q defect. Data from PIOPED [2] and Gottschalk et al. [1].

≥4 moderate V–Q

≥4 moderate V–Q

mismatches

mismatches or

moderate V–Q

moderate V–Q

or

2 or 3 moderate V–Q

1 large V–Q and ≥2

1 large V–Q and ≥2 mismatches

or

or

or

high probability)

±1 moderate V–Q

1 large V–Q mismatch

mismatches (2 large

≥2 large V–Q

≥2 large V–Q

PIOPED (2)

Intermediate probability

March 11, 2007

mismatches

PIOPED revised (1)

PIOPED (2)

High probability

Table 65.1 Original PIOPED (Prospective Investigation of Pulmonary Embolism Diagnosis) criteria for ventilation–perfusion lung scans compared with proposed revised criteria.

BLUK077-Stein

11:1


BLUK077-Stein

March 11, 2007

11:1

286

PART III

Diagnosis of acute PE

Figure 65.1 Left: Perfusion scan, posterior view, showing solitary perfusion defect (arrow), in the posterior basal segment of the right lower lobe. Right: Ventilation scan, posterior view, initial breath showing no ventilation defect. Equilibrium and washout phases also showed no ventilation defect. According to original PIOPED criteria

this single mismatched moderate size perfusion defect should have been interpreted as low probability for pulmonary embolism (PE), but according to the revised PIOPED criteria this ventilation–perfusion scan should be interpreted as intermediate probability. Pulmonary embolism was shown on the pulmonary angiogram.

whereas 2.5 mismatched segmental equivalent perfusion defects provided a more accurate categorization [1] (Table 65.2). A mismatched segmental equivalent perfusion defect is 1 large (>75% of a segment) mismatched perfusion defect or 2 moderate size (25–75% of a segment) perfusion defects. The nuclear physicians

in PIOPED could not definitely recommend changing the threshold for high probability, however, because the number of patients with this pattern was small. Ten of 28 patients (36%) with a single moderate mismatched perfusion defect had PE [1]. It was clear that this was not a valid criterion for low probability. Scans with this finding should be considered intermediate probability for PE (Figure 65.1). Patients with a single matched perfusion appeared to have a higher likelihood of PE than patients with 2 or more matched perfusion defects [1]. A single matched defect was associated with PE in 26% of patients (Table 65.3). Two or more matched perfusion defects were associated with PE in 14% of patients. A single matched perfusion defect, therefore, should be considered for intermediate probability.

Table 65.2 Frequency of pulmonary embolism in patients with various patterns of mismatched perfusion defects. Number of Type of mismatch

patients

PE (%)

1 segmental equivalent

33

52

1 large defect

24

46

2 moderate defects

9

67

1.5 segmental equivalents

18

72

1 large + 1 moderate defect

11

73

3 moderate defects

7

71

2 segmental equivalents

7

71

2 large defects

5

80

1 large + 2 moderate defects

1

0

4 moderate defects

1

100

10

100

2 large + 1 moderate defect

8

100

1 large + 3 moderate defects

1

100

5 moderate defects

1

100

2.5 segmental equivalents

Defect = mismatched perfusion defect. PE, pulmonary embolism. Reprinted by permission of the Society of Nuclear Medicine from Gottschalk et al. [1].

Table 65.3 Patients with only matched defects on the ventilation–perfusion scan. Number of points

PE (%)

Single matched defect (any size)

23

26

Multiple matched defects (any size)

66

14

All matched defects

89

17

PE, pulmonary embolism. Data from Gottschalk et al. [1].


BLUK077-Stein

March 11, 2007

CHAPTER 65

11:1

Revised criteria for evaluation of V–Q scans

References 1 Gottschalk A, Sostman HD, Coleman RE et al. Ventilation– perfusion scintigraphy in the PIOPED study. Part II. Evaluation of the scintigraphic criteria and interpretations. J Nucl Med 1993; 34: 1119–1126.

287

2 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism—results of the prospective investigation of pulmonary embolism diagnosis (PIOPED). JAMA 1990; 263: 2753–2759.


BLUK077-Stein

March 12, 2007

66

16:44

CHAPTER 66

Criteria for very low probability interpretation of ventilation– perfusion lung scans

Introduction The criteria for the interpretation of low probability ventilation–perfusion (V–Q) lung scans in patients with suspected acute pulmonary embolism (PE) that were used in the collaborative study of Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED I) [1] have been modified since its conclusion [2]. In PIOPED I, using the original PIOPED criteria, 14% of patients with ventilation–perfusion scans interpreted as low probability had PE [1]. The PIOPED I criteria for low probability included [1]:

1 Nonsegmental perfusion defects ≤ radiographic abnormality. 2 Perfusion defects ≤ ventilation defects with normal chest radiograph. 3 Perfusion defects < radiographic defect. 4 ≤3 small perfusion defects with a normal chest radiograph. 5 A single moderate size mismatched perfusion defect with a normal chest radiograph. The Nuclear Medicine Working Group of PIOPED recommended that the following modifications be made in regard to low probability interpretations [2]:

1 A single moderate mismatched perfusion defect should be categorized as intermediate rather than as low probability. 2 Multiple and relatively extensive matched ventilation–perfusion abnormalities are appropriate for low probability, provided that the chest radiograph is clear. 3 Single matched defects may be better categorized as intermediate probability, although this cannot be definitely validated statistically. The modifications of the PIOPED I criteria for low probability were made on the assumption that patients

288

with low probability interpretations of ventilation– perfusion scans should have a positive predictive value of PE <20% [2]. These revised PIOPED criteria were found to be more accurate than the original PIOPED criteria [3]. The modified PIOPED I criteria, however, were tested prospectively in a population of patients that differed significantly from the population of patients investigated in PIOPED I [4]. In the population of patients investigated, 36% had normal pulmonary scintiscans, in comparison to 2% in PIOPED and the prevalence of PE was also lower, 10% versus 28% in PIOPED. Using the modified criteria, patients with a low probability interpretation had a 6% frequency of PE [4], whereas in PIOPED such patients had a 14% frequency of PE A “very low probability” interpretation with a positive predictive value less than 10% would be more useful than a “low probability” interpretation [5]. To identify such “very low probability” criteria of this, we evaluated the individual characteristics and combinations of characteristics of the low probability ventilation–perfusion lung scan in PIOPED I [6]. In some instances, we maximized the useful PIOPED database by evaluation of individual lungs or lung zones rather than individual patients [6, 7].

Nonsegmental perfusion abnormalities Nonsegmental perfusion abnormalities that were associated with a low positive predictive value for PE were enlargement of the heart, hila, or mediastinum, or elevated hemidiaphragm (Figure 66.1). Lungs with such nonsegmental perfusion abnormalities had a positive predictive value for PE of 8% [6] (Table 66.1, Figure 66.2).


BLUK077-Stein

March 12, 2007

CHAPTER 66

16:44

289

Very low probability interpretation of V–Q scans

Perfusion

(a) Figure 66.1 (a) Plain chest radiograph showing mild cardiomegaly and prominence of hila. (b) Perfusion lung scan, anterior view, showing nonsegmental perfusion abnormalities. The scan shows prominence of the hila, more clearly shown on the right (arrow) and cardiomegaly.

(b) These types of nonsegmental perfusion abnormalities were associated with a very low positive predictive value for pulmonary embolism (PE). (Images courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, Michigan.)

Table 66.1 Positive predictive value of individual criteria used for low probability assessment of ventilation–perfusion lung scans. PE/total (%) Lungs with only one type of perfusion defect Costophrenic angle effusion

4/14 (29)

Nonsegmental abnormality

8/103 (8)

Perfusion defect < X-ray 1 zone

2/24 (8)

2 or 3 zones

1/16 (6)

All zones

3/40 (8)

Matched V–Q (X-ray normal) 1 zone

4/34 (12)

2 or 3 zones

1/30 (3)

All zones

5/64 (8)

Lungs with two types of perfusion defects Costophrenic angle effusion and nonsegmental abnormality

1/10 (10)

Costophrenic angle effusion and matched V–Q (X-ray normal)

1/8 (13)

Nonsegmental abnormality and perfusion defect < X-ray

3/34 (9)

Nonsegmental abnormality and matched V–Q (X-ray normal)

4/29 (14)

Combinations of two perfusion defects were excluded from the table if the combination was observed in only 4 or fewer lungs. Costophrenic angle effusion = Pleural effusion with obliteration of the costophrenic angle with the perfusion defect ≤ X-ray abnormality. Nonsegmental abnormality = Nonsegmental perfusion abnormality including enlargement of the hilum, mediastinum, or heart or elevated diaphragm with the perfusion defect ≤ X-ray abnormality. Perfusion defect < X-ray = Parenchymal abnormality on the chest radiograph with the perfusion defect < radiographic abnormality. Matched V–Q (X-ray normal) = Matched ventilation–perfusion defect with normal chest radiograph and perfusion defect ≤ ventilation defect. PE, pulmonary embolism. Reprinted by permission of the Society of Nuclear Medicine from Stein et al. [6].


BLUK077-Stein

March 12, 2007

16:44

290

PART III

Perfusion

(a)

(b)

Perfusion defect smaller than radiographic defect A perfusion defect smaller than the radiographic defect, regardless of the findings on the ventilation scan, showed PE in 8% of patients [6] (Table 66.1).

Matched ventilation–perfusion abnormality in 2 or 3 zones of a single lung Matched ventilation–perfusion abnormalities, in the presence of a regionally normal chest radiograph, when occurring as the only type of perfusion defect in 2 or 3 zones of a single lung, showed a pos-

L (a)

Perfusion

L (b)

Diagnosis of acute PE

Figure 66.2 (a) Plain chest radiograph showing linear atelectasis in the left lower zone. (b) Perfusion lung scan, anterior view, showing small perfusion defect (arrow) in the region of the atelectasis. Perfusion defects smaller than associated parenchymal abnormalities on the chest radiograph, when occurring as the only type of perfusion defect, showed a positive predictive value for pulmonary embolism (PE) of 8%. The pulmonary angiogram in these patients showed no PE. (Images courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, Michigan.)

itive predictive value for PE of 3% [6] (Figure 66.3, Table 66.1). A matched ventilation–perfusion defect occurring as the only type of perfusion defect in 1 zone of a single lung showed a trend toward a higher positive predictive value for PE than matched ventilation–perfusion defects in 2 zones or 3 zones in a single lung. A matched ventilation–perfusion defect in only 1 zone was not considered a criterion for a very low probability interpretation [6]. The positive predictive value for PE with matched ventilation–perfusion defects in 1 zone was 4 of 34 (12%), in 2 zones it was 1 of 20 (5%), and in 3 zones it was 0 of 10 (0%) [6] (Table 66.1).

Ventilation

Figure 66.3 (a) Perfusion lung scan, posterior view, showing a perfusion defect (arrow) in the left lower zone. (b) Ventilation scan, posterior view, initial breath, showing impaired ventilation in the left lower zone. A matched ventilation–perfusion defect occurring as the only type of perfusion defect in 1 zone of a single lung showed a trend toward a higher positive predictive value for pulmonary embolism (PE) than matched perfusion defects in 2 zones or 3 zones in a single lung. This patient showed no PE on the pulmonary angiogram. (Images courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, Michigan.)


BLUK077-Stein

March 12, 2007

CHAPTER 66

16:44

291

Very low probability interpretation of V–Q scans

Figure 66.4 Posterior view of a perfusion lung scan. Both the round lesion in the left apex and the retrocardiac lesion in the base of the left lung were small segmental perfusion defects (<25% of a segment). The patient showed no PE on pulmonary angiography. (Reprinted by permission of the Society of Nuclear Medicine from Stein et al. [8].

One to three small perfusion defects The positive predictive value for PE of lung scans with 1–3 small (<25% of a segment) segmental perfusion defects with a regionally normal chest radiograph was 1–3% depending on the group analyzed [8] (Figure 66.4). Evaluation of this criterion was independent of findings on the ventilation scan; a regionally normal chest radiograph was required. This low positive predictive value for PE is comparable to the positive predictive value reported by Stein [9]. Among patients in whom PE was diagnosed or excluded by pulmonary angiography or follow-up, 1 of 68 (1%) showed PE [8]. Among patients with >3 small segmental perfusion defects, 3 of 27 (11%) showed PE [8].

Triple matched defects in the upper or middle lung zones A matched ventilation–perfusion defect with associated matching chest radiographic opacity is defined as a triple matched defect. The triple match can be caused by PE creating a pulmonary “infarction” (usually pulmonary hemorrhage). Worsley and associates, showed that PE was present more frequently with triple matches in the lower zones of the lung compared with the upper or middle zones [10] (Table 66.2). We showed PE with triple matched defects was infrequent in the upper or middle lung zones, occurring in 1 of 27 (4%) (Figure 66.5, Table 66.2) [11]. In the lower zone,

Table 66.2 Positive predictive value of triple matched defects according to lung zone. Gottschalk

Worsley

[PE/No. (%)]

[PE/No. (%)]

Upper zone

0/13 (0)

Middle zone

1/14 (7)

4/36 (11)

Lower zone

13/57 (23)

61/187 (33)

Total

14/84 (17)

71/275 (26)

6/52 (12)

PE, pulmonary embolism; No., number. Reprinted by permission of the Society of Nuclear Medicine from Gottschalk et al. [11].

however, with a triple match, PE occurred in 13 of 57 (23%) [11].

Stripe sign The stripe sign on perfusion images consists of a stripe of perfused lung tissue between a perfusion defect and the adjacent pleural surface [10]. In PIOPED, perfusion defects showing the stripe sign had a positive predictive value of PE of only 6 of 85 (7%) [12].

Pleural effusion A pleural effusion with blunting of the costophrenic angle in combination with a nonsegmental perfusion defect (enlargement of the heart, hila, or mediastinum, or elevated hemidiaphragm) showed a


BLUK077-Stein

March 12, 2007

16:44

292

PART III

Diagnosis of acute PE

Figure 66.5 (a) Chest radiograph, lateral view, showing a parenchymal abnormality in the posterior portion of the base of the right lung (arrow). (b) Perfusion scan, lateral view of right lung, showing perfusion defect in posterior inferior portion (arrow). (c) Ventilation scan, posterior view, equilibrium phase, showing absent ventilation in base of right lung. The finding of a matched ventilation–perfusion defect with associated matching

chest radiographic opacity (the triple match) in a lower lung zone has an intermediate/indeterminate probability with a positive predictive value for acute pulmonary embolism (PE) of 23%. This patient had PE. (Images courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, Michigan.)

positive predictive value for PE of 10% [2] (Table 66.1). A small pleural effusion alone was associated with a higher positive predictive value for PE (29%). We prospectively evaluated the database of PIOPED II to test these criteria for very low probability and the additional criterion of a large pleural effusion [13]. Some of the criteria, however, were defined somewhat differently in PIOPED II. The diagnostic reference standards for PE were CT angiogram, CT venogram, and standard pulmonary digital subtraction angiography (DSA) [13]. Criteria for a very low probability of PE tested prospectively from data in PIOPED II were:

6 Stripe sign (consists of a stripe of perfused lung tissue between a perfusion defect and the adjacent pleural surface, best seen on a tangential view). 7 Pleural effusion ≥1/3 of the pleural cavity with no other perfusion defect in either lung.

1 Nonsegmental perfusion abnormalities: these were enlargement of the heart or hilum, elevated hemidiaphragm, costophrenic angle effusion, or linear atelectasis with no other perfusion defect in either lung. 2 Perfusion defect smaller than corresponding radiographic lesion. 3 ≥2 matched V–Q defects with regionally normal chest radiograph and some areas of normal perfusion elsewhere in the lungs. 4 1–3 small segmental perfusion defects (<25% of a segment). 5 Solitary triple matched defect (defined as a matched V–Q defect with associated matching chest radiographic opacity) in the mid or upper lung zone confined to a single segment.

A very low probability consensus interpretation of the V–Q scan was made in 460 of 824 patients with suspected PE (56%). Among these patients, the CT angiogram was obtained and of diagnostic quality in 422 patients. In these patients, PE was shown in 38 of 422 (9.0%) [13]. Each of the component V–Q criteria for very low probability showed a positive predictive value <10% except 1–3 small segmental perfusion defects (Table 66.3). However, some categories had too few patients for a meaningful assessment. The positive predictive value of a very low probability interpretation was 15 of 185 (8.1%). If the criterion of 1–3 small perfusion defects were eliminated, the positive predictive value of a very low probability interpretation would be 12 of 161 (7.5%). The data from PIOPED II confirmed that the very low probability V–Q interpretation has a positive predictive value for PE of <10% [13]. This places the low probability interpretation into a useful category for the exclusion of acute PE, keeping in mind that a very low probability interpretation of the V–Q scan was made in 56% of patients with suspected PE [13]. All of the component V–Q features of a very low probability interpretation showed a low (<10%) positive


BLUK077-Stein

March 12, 2007

CHAPTER 66

16:44

293

Very low probability interpretation of V–Q scans

Table 66.3 Positive predictive value of criteria for very low probability interpretation. Criterion

PE/patients (%)

All very low probability*

15/185 (8.1)

Nonsegmental perfusion abnormalities

10/120 (8.3)

Perfusion defect smaller than corresponding radiographic lesion

0/9 (0)

≥2 matched V–Q defects

5/57 (8.8)

1–3 small segmental perfusion defects

8/64 (12.5)

A solitary triple matched defect

0/3 (0)

Stripe sign

1/25 (4.0)

Pleural effusion ≥1/3 of the pleural cavity with no other perfusion defect in either lung

0/5 (0)

*Includes only patients in whom central readers identified the criterion on which a very low probability interpretation was based, and in whom central reader 1 and central reader 2 agreed on the presence of the criterion. PE, pulmonary embolism. Reproduced with permission from Gottschalk et al. [13].

predictive value except 1–3 small segmental perfusion defects [13]. Many criteria, however, require further evaluation because of sparse data.

References 1 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. 2 Gottschalk A, Sostman HD, Coleman RE et al. Ventilation–perfusion scintigraphy in the PIOPED study. Part II. Evaluation of the scintigraphic criteria and interpretations. J Nucl Med 1993; 34: 1119–1126. 3 Sostman HD, Coleman RE, DeLong DM, Newman GE, Paine S. Evaluation of revised criteria for ventilation– perfusion scintigraphy in patients with suspected pulmonary embolism. Radiology 1994; 193: 103–107. 4 Freitas JE, Sarosi MG, Nagle CC, Yeomans ME, Freitas AE, Juni JE. Modified PIOPED criteria used in clinical practice. J Nucl Med 1995; 36: 1573–1578. 5 Hull RD, Raskob GE. Low-probability lung scan findings: a need for change. Ann Intern Med 1991; 114: 142–143. 6 Stein PD, Relyea B, Gottschalk A. Evaluation of individual criteria for low probability interpretation of ventilationperfusion lung scans. J Nucl Med 1996; 37: 577–581.

7 Stein PD, Gottschalk A. Review of criteria appropriate for a very low probability of pulmonary embolism on ventilation–perfusion lung scans: a position paper. Radiographics 2000; 20: 99–105. 8 Stein PD, Henry JW, Gottschalk A. Small perfusion defects in suspected pulmonary embolism. J Nucl Med 1996; 37: 1313–1316. 9 Silberstein E, Worsley DF, Alavi A, Elgazzar A. The clinical significance of the very low probability (PIOPED) lung scan pattern [abstract]. J Nucl Med 1995; 36: 113. 10 Worsley DF, Kim CK, Alavi A, Palevsky HI. Detailed analysis of patients with matched ventilation–perfusion defects and chest radiographic opacities. J Nucl Med 1993; 34: 1851–1853. 11 Gottschalk A, Stein PD, Henry JW, Relyea B. Matched ventilation, perfusion and chest radiographic abnormalities in acute pulmonary embolism. J Nucl Med 1996; 37: 1636–1638. 12 Sostman HD, Gottschalk A. Prospective validation of the stripe sign in ventilation–perfusion scintigraphy. Radiology 1992: 184: 455–459. 13 Gottschalk A, Stein PD, Sostman HD et al. Very low probability interpretation of ventilation-perfusion lung scans in combination with low probability clinical assessment reliably excludes pulmonary embolism: Data from PIOPED II (Submitted for Publication).


BLUK077-Stein

March 11, 2007

67

11:27

CHAPTER 67

Probability assessment based on the number of mismatched segmental equivalent perfusion defects or number of mismatched vascular defects

Introduction The diagnostic value of ventilation–perfusion lung scans was strengthened by creating a table of positive predictive values and specificities based on the observed number of mismatched segmental equivalent perfusion defects [1] (Table 67.1, Figure 67.1). This table was more informative than “high,” “intermediate,” or “low” probability clinical assessments, and permitted the nuclear scan reader to assign specific risks of pulmonary embolism (PE) to individual patients. In addition, stratification of patients according to the presence or absence of prior cardiopulmonary disease enhanced the ventilation–perfusion scan assessment of the probability of PE among patients in both of these clinical categories [1]. One segmental equivalent was defined as 1 large segment or 2 moderate size segments [2]. For example, 1.5 mismatched segmental equivalents can be obtained with either 1 large and 1 moderate mismatched segmental perfusion defect or 3 moderate mismatched segmental perfusion defects. With increasing numbers of mismatched segmental equivalent perfusion defects, there was a continuum of gradually increasing specificities and positive predictive values, and decreasing sensitivities [1] (Table 67.1, Figure 67.1). No discrete criterion for “high probability” can be clearly identified from the data. Rather than define an arbitrary number of mismatched segmental equivalent defects as “high probability” for all patients, use of Table 67.1 enables readers of ventilation–perfusion scans to assign a positive predictive value and specificity

294

to their individual patients according to prior disease status.

Number of mismatched segmental equivalent defects and stratification according to prior cardiopulmonary disease Stratification of patients according to the presence or absence of prior cardiopulmonary disease enhanced the ventilation–perfusion scan assessment of PE among both clinical categories of patients [1]. Among patients with no prior cardiopulmonary disease ≥0.5 mismatched segmental equivalent perfusion defects was associated with a positive predictive value for PE of 80% (Figure 67.2, Table 67.1). With ≥1.5 mismatched segmental equivalents, the positive predictive value for PE was 89% in patients with no prior cardiopulmonary disease. Among patients with prior cardiopulmonary disease, more mismatched segmental equivalent perfusion defects were required to give a particular positive predictive value and specificity than among patients with no prior cardiopulmonary disease. Among patients with prior cardiopulmonary disease, ≥2 mismatched segmental equivalent perfusion were required for an 80% positive predictive value of PE, and ≥4.5 mismatched segmental equivalents were required for a 90% positive predictive value of PE [1]. Further stratification of patients with prior cardiopulmonary disease into those with prior cardiac disease (exclusive of prior pulmonary disease) and


BLUK077-Stein

March 11, 2007

CHAPTER 67

11:27

295

Mismatched segmental perfusion defects or mismatched vascular defects

Table 67.1 Positive predictive value of cumulative number of mismatched segmental equivalent perfusion defects among patients with no prior cardiopulmonary disease and patients with prior cardiopulmonary disease. #PE/#PTS (%) Segmental

Patients with no prior

Patients with prior

All patients

equivalents

cardiopulmonary disease (n = 421)

cardiopulmonary disease (n = 629)

(n = 1050)

≥0.0

173/421 (41)

205/629 (33)*

378/1050 (36)

≥0.5

123/154 (80)

130/192 (68)†

253/346 (73)

≥1.0

102/118 (86)

113/155 (73)*

215/273 (79)

≥1.5

91/102 (89)

99/128 (77)†

190/230 (83)

≥2.0

79/87 (91)

91/114 (80)‡

170/201 (85)

≥2.5

72/80 (90)

87/105 (83)

159/185 (86)

≥3.0

65/73 (89)

81/97 (84)

146/170 (86)

≥3.5

60/67 (90)

77/88 (88)

137/155 (88)

≥4.0

57/63 (90)

74/84 (88)

131/147 (89)

≥4.5

50/53 (94)

70/78 (90)

120/131 (92)

≥5.0

49/52 (94)

65/72 (90)

114/124 (92)

≥5.5

47/50 (94)

61/66 (92)

108/116 (93)

≥6.0

42/44 (95)

59/64 (92)

101/108 (94)

≥6.5

40/42 (95)

56/61 (92)

96/103 (93)

≥7.0

38/40 (95)

51/56 (91)

89/96 (93)

≥7.5

34/36 (94)

43/47 (91)

77/83 (93)

*P < 0.01. † P < 0.02. ‡ P < 0.05: No prior cardiopulmonary disease vs. Prior cardiopulmonary disease. These probabilities were calculated with chi-square and are higher than reported in Stein et al. [1] because Yates’ correction, which was unnecessarily conservative, was used in the previous analysis. Modified and reprinted with permission from Stein et al. [1].

Pulmonary embolism (%)

100

Figure 67.1 Positive predictive value for pulmonary embolism (PE) according to the number of mismatched segmental equivalent perfusion defects on the ventilation–perfusion lung scan. Bars represent standard error.

90

80

70

60

≥0

≥1

≥2 ≥3 ≥4 ≥5 Cumulative mismatched segmental equivalents

≥6


BLUK077-Stein

March 11, 2007

11:27

296

PART III

Diagnosis of acute PE

Figure 67.2 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched segmental equivalent perfusion defects. Broken line indicates patients with no prior cardiopulmonary disease (NO CPD). Unbroken line indicates patients with any prior cardiopulmonary disease (CPD). Significant differences occurred with ≥0.5 (P < 0.02), ≥1.0 segmental equivalents (P < 0.01), ≥1.5 (P < 0.02), and ≥2.0 (P < 0.05) segmental equivalents. These probabilities were calculated by chi-square and are higher than reported by Stein et al. [1] because Yates’ correction was unnecessarily conservative, was used in the previous calculations. (Reprinted with permission from Stein et al. [1].)

those with prior pulmonary disease (exclusive of prior cardiac disease) did not strengthen the positive predictive value of either group. No significant differences were shown between patients with prior cardiac disease and patients with prior lung disease (Figure 67.3). As shown in Table 67.1, some patients with PE had no mismatched large or moderate size segmental defects. The patients who had PE and no mismatched segmental lesions had either small perfusion defects, mediastinal abnormalities, pleural effusions, a variety of parenchymal radiographic abnormalities, ventilation– perfusion matched defects or any combination of

these. None had an entirely normal ventilation– perfusion scan. Sensitivity and specificity according to the number of mismatched segmental equivalent perfusion defects among patients stratified according to the presence or absence of prior cardiopulmonary disease is shown in Table 67.2. With stratification, among patients with no prior cardiopulmonary disease the presence of 1.0 or more mismatched segmental equivalent perfusion defects was more sensitive for PE than the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) criteria for high probability (≥2

Figure 67.3 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched segmental equivalent perfusion defects among patients with no prior cardiopulmonary disease (NO CPD), any prior cardiac disease (CARD DIS), and any prior pulmonary disease (PULM DIS). There were no significant differences between patients with prior cardiac disease and patients with prior pulmonary disease. (Reprinted with permission from Stein et al. [1].)


BLUK077-Stein

March 11, 2007

CHAPTER 67

11:27

297

Mismatched segmental perfusion defects or mismatched vascular defects

Table 67.2 Number of mismatched segmental equivalent defects, sensitivity, and specificity among patients with no prior cardiopulmonary disease and patients with prior cardiopulmonary disease. Patients with no prior

Patients with prior

cardiopulmonary

cardiopulmonary

All patients

disease (n = 421)

disease (n = 629)

(n = 1050)

Segmental equivalents

SENS (%)

SPEC (%)

SENS (%)

SPEC (%)

SENS (%)

SPEC (%)

≥ 0.0

100

0

100

0

100

0

≥ 0.5

71

88

63

85

67

86

≥ 1.0

59

94

55

90

57

91

≥ 1.5

53

96

48

93

50

94

≥ 2.0

46

97

44

95

45

95

≥ 2.5

42

97

42

96

42

96

≥ 3.0

38

97

40

96

39

96

≥ 3.5

35

97

38

97

36

97

≥ 4.0

33

98

36

98

35

98

≥ 4.5

29

99

34

98

32

98

≥ 5.0

28

99

32

98

30

99

≥ 5.5

27

99

30

99

29

99

≥ 6.0

24

99

29

99

27

99

≥ 6.5

23

99

27

99

25

99

≥ 7.0

22

99

25

99

24

99

≥ 7.5

20

99

21

99

20

99

SENS, Sensitivity; SPEC, Specificity. Reprinted with permission from Stein et al. [1].

segmental equivalent defects) applied to unstratified patients (59% versus 41%) (P < 0.001) [1, 3]. This increased sensitivity was associated with a comparable positive predictive value (86% versus 87%) and specificity (94% versus 97%) [1, 3].

References 1 Stein PD, Gottschalk A, Henry JW, Shivkumar K. Stratification of patients according to prior cardiopulmonary disease and probability assessment based on the num-

ber of mismatched segmental equivalent perfusion defects: approaches to strengthen the diagnostic value of ventilation/perfusion lung scans in acute pulmonary embolism. Chest 1993; 104: 1461–1467. 2 Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. 3 A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759.


BLUK077-Stein

March 11, 2007

68

11:28

CHAPTER 68

Probability assessment based on the number of mismatched vascular defects and stratification according to prior cardiopulmonary disease

The diagnosis of pulmonary embolism (PE) on the basis of the number of mismatched segmental equivalents assumes that a mismatched moderate size segmental perfusion defect is of less diagnostic value than a mismatched large size segmental perfusion defect [1]. This assumption has not been fully tested. There is considerable skill and judgment required in distinguishing a moderate size mismatched segmental perfusion defect (25â&#x20AC;&#x201C;75% of a segment) from a large size mismatched segmental perfusion defect (>75% of a segment). Experienced readers of radionuclide lung scans often underestimate the size of segmental defects [2]. Ventilationâ&#x20AC;&#x201C;perfusion lung scans can be assessed without the necessity of distinguishing large mismatched segmental defects from moderate size mismatched seg-

mental defects [3]. This makes the interpretation of ventilationâ&#x20AC;&#x201C;perfusion lung scans easier and more objective. The positive predictive value of PE relative to the number of mismatched large segmental perfusion defects was similar to the positive predictive value of PE relative to the number of mismatched moderate size segmental defects [3]. The positive predictive value of PE based upon the cumulative number of mismatched large segmental defects, irrespective of the number of mismatched moderate size segmental defects, is shown in Figure 68.1 and Table 68.1. The positive predictive value of PE based upon the cumulative number of mismatched moderate size segmental perfusion defects, irrespective of the number

Figure 68.1 Positive predictive value of pulmonary embolism (PE) according to the cumulative number of large segmental (SEG) perfusion defects (irrespective of the number of moderate size segmental defects) (unbroken line) and according to the number of moderate size segmental perfusion defects (irrespective of the number of large segmental defects) (broken line). (Reprinted with permission from Stein et al. [3].)

298


BLUK077-Stein

March 11, 2007

CHAPTER 68

11:28

299

Interpretation of V–Q scans according to prior cardiopulmonary disease

Table 68.1 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched large segmental perfusion defects and moderate size segmental perfusion defects. Mismatched large segmental defects

Mismatched moderate size segmental

(n = 1064)

defects (n = 1064)

Cumulative number of defects

#PE/#PTS

%PE

#PE/#PTS

%PE

≥0

383/1064

36

383/1064

36

≥1

197/251

78

205/277

74

≥2

161/191

84

119/141

84

≥3

135/153

88

89/98

91

≥4

119/133

89

64/69

93

≥5

108/117

92

48/52

92

≥6

92/99

93

40/43

93

≥7

78/82

95

36/38

95

≥8

62/66

94

35/37

95

PE, pulmonary embolism; PTS, patients. Modified and reprinted with permission from Stein et al. [3].

of mismatched large segmental perfusion defects, is also shown in Figure 68.1 and Table 68.1. Based on the similarity of the positive predictive value of large mismatched perfusion defects and moderate size mismatched perfusion defects, we defined a mismatched vascular perfusion defect as a mismatched large or moderate segmental perfusion defect. The number of mismatched vascular perfusion defects, therefore, was the number of large and/or moderate size mismatched perfusion defects [3]. The size of the mismatched defect, providing it was ≥25% of a segment, was of no consequence. The positive predictive value of the cumulative number of mismatched vascular defects (mismatched large and/or

Figure 68.2 Positive predictive value of pulmonary embolism (PE) based on the cumulative number of mismatched vascular defects (the number of mismatched large and/or moderate size segmental defects). (Reprinted with permission from Stein et al. [3].).

moderate size segments) is shown in Table 68.2 and Figure 68.2. The sensitivity and specificity of various numbers of mismatched vascular defects are shown in Table 68.3. The cumulative number of mismatched segmental equivalents perfusion defects and the cumulative number of vascular defects were of comparable diagnostic value, as indicated by similar values of the maximum likelihood estimates of the areas under the receiver operating characteristics (ROC) curves (Table 68.4, Figure 68.3). Stratification of patients according to the presence or absence of prior cardiopulmonary disease enhanced the positive predictive value of the number of


BLUK077-Stein

March 11, 2007

11:28

300

PART III

Diagnosis of acute PE

Table 68.2 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched vascular perfusion defects (large and/or moderate size segmental perfusion defects). Cumulative number

All patients

of defects

(n = 1064)

Percent

(n = 421)

No prior CPD Percent

(n = 629)

Any prior CPD Percent

≥0

383/1064

36

173/421

41

205/629

33∗

≥1

255/350

73

123/154

80

130/190

68†

≥2

200/244

82

94/106

89

105/136

77‡

≥3

170/201

85

79/87

91

90/112

80‡

≥4

149/172

87

67/75

89

81/96

84

≥5

130/144

90

55/60

92

75/84

89

≥6

116/124

94

49/52

94

67/72

93

≥7

103/110

94

43/45

96

60/65

92

≥8

93/100

93

41/43

95

52/57

91

P < 0.01. P < 0.02. ‡ P < 0.05: Prior cardiopulmonary disease vs. No prior cardiopulmonary disease. These probabilities are higher than reported by Stein et al. [3] because the chi-square probabilities previously calculated used the unnecessarily conservative Yates’ correction. CPD, cardiopulmonary disease. Modified and reprinted with permission from Stein et al. [3]. †

mismatched vascular defects in each category (Table 68.2, Figure 68.4). Among patients with no prior cardiopulmonary disease, ≥1 vascular defect indicated a positive predictive value for PE of 80% and ≥2 mismatched vascular defects indicated a positive predictive value of 89% (Figure 68.5). Among patients with

prior cardiopulmonary disease, ≥1 mismatched vascular perfusion defect indicated a positive predictive value of 68% and ≥2 mismatched vascular perfusion defects indicated a positive predictive value of 77%. Tables 68.2 and 68.3 can be used to assess a positive predictive value and specificity for individual patients

Table 68.3 Sensitivity and specificity in relation to the number of mismatched vascular perfusion defects large segmental and/or moderate size segmental defects). All patients

Cumulative number of defects

SENS (%)

SPEC (%)

No prior CPD SENS (%)

SPEC (%)

Prior CPD SENS (%)

SPEC (%)

≥0

100

0

100

0

100

0

≥1

67

86

71

88

63

86

≥2

52

94

54

95

51

93

≥3

44

95

46

97

44

95

≥4

39

97

39

97

40

96

≥5

34

98

32

98

37

98

≥6

30

99

28

99

33

99

≥7

27

99

25

99

29

99

≥8

24

99

24

99

25

99

All differences between NO prior cardiopulmonary disease and Prior cardiopulmonary disease were not significant. SENS, sensitivity; SPEC, specificity; CPD, cardiopulmonary disease. Reprinted with permission from Stein et al. [3].


BLUK077-Stein

March 11, 2007

CHAPTER 68

11:28

301

Interpretation of V–Q scans according to prior cardiopulmonary disease

Table 68.4 Maximum likelihood estimates of the areas under receiver operating characteristic curves based on the number of mismatched vascular perfusion defects and the number of mismatched segmental equivalent defects. Mismatched vascular

Mismatched segmental

defects (area)

equivalents (area)

All patients (n = 1064)

0.8512

0.8530

Patients with no prior

0.8875*

0.8905†

0.8280

0.8215

cardiopulmonary disease (n = 421) Patients with any prior cardiopulmonary disease (n = 629) ∗

P = 0.051 – NO prior cardiopulmonary disease vs. Any prior cardiopulmonary disease – mismatched vascular defects. P = 0.026 – NO prior cardiopulmonary disease vs. Any prior cardiopulmonary disease – mismatched segmental equivalents. Differences between mismatched vascular defects and mismatched segmental equivalent defects were not significant. Reprinted with permission from Stein et al. [3]. †

Figure 68.3 Receiver operating characteristic curves based on the cumulative number of mismatched segmental equivalent perfusion defects and on the cumulative number of mismatched vascular perfusion defects. The curves are superimposed.

Figure 68.4 Positive predictive values of pulmonary embolism (PE) based on the cumulative number of mismatched vascular defects among patients with no prior cardiopulmonary disease (NO CPD) and among patients with any prior cardiopulmonary disease (CPD). Differences between patients with CPD and NO CPD occurred with ≥1 mismatched vascular defect (P < 0.02), ≥2 defects (P < 0.05), and ≥3 defects (P < 0.05). Probabilities are based on chi-square without Yates’ correction. (Reprinted with permission from Stein et al. [3].)


BLUK077-Stein

March 11, 2007

11:28

302

PART III

Perfusion

Posterior

Ventilation

Posterior

L LAT

RPO

Perfusion

Diagnosis of acute PE

Perfusion

Figure 68.5 Top left: Perfusion scan, posterior view, showing suggestion of perfusion defects, seen better in left lateral view (below). Top right: Ventilation scan, posterior view, washout phase, showing no abnormality. Bottom left: Perfusion scan in left lateral view showing moderate size perfusion defects (arrows) in left anterior segment of left upper lobe and anterior basal segment of left lower lobe. Bottom right: Perfusion scan, right posterior oblique (RPO) view, showing a nonsegmental perfusion abnormality associated with atelectasis (arrow). Among patients with no prior cardiopulmonary disease, 2 or more mismatched vascular defects indicated a positive predictive value for pulmonary embolism (PE) of 89%. Pulmonary embolism was shown on the pulmonary angiogram in this patient. (Images courtesy of Alexander Gottschalk, MD, Department of Radiology, Michigan State University, East Lansing, Michigan.)

Figure 68.6 Positive predictive value of pulmonary embolism (PE) according to the cumulative number of mismatched vascular defects among patients with no prior cardiopulmonary disease (NO CPD), prior pulmonary disease (PULM DIS), and prior cardiac disease (CARD DIS). The positive predictive value of PE among patients with prior cardiac disease and those with prior pulmonary disease was similar. (Reprinted with permission from Stein et al. [3].)


BLUK077-Stein

March 11, 2007

CHAPTER 68

11:28

Interpretation of V–Q scans according to prior cardiopulmonary disease

stratified according to prior cardiopulmonary disease and according to the number of mismatched vascular perfusion defects. Patients with prior cardiac disease and those with prior pulmonary disease showed a comparable positive predictive value (Figure 68.6). There was, therefore, no advantage in stratification according to the specific prior cardiac or pulmonary disease. The number of mismatched vascular defects is as powerful for the assessment of ventilation–perfusion scans as the number of mismatched segmental equivalents. The number of mismatched vascular defects is easier to interpret, and permits a more objective evaluation [3].

303

References 1 Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. 2 Morrell NW, Nijran KS, Jones BE, Biggs T, Seed WA. The underestimation of segmental defect size in radionuclide lung scanning. J Nucl Med 1993; 34: 370– 374. 3 Stein PD, Henry JW, Gottschalk A. Mismatched vascular defects: an easy alternative to mismatched segmental equivalent defects for the interpretation of ventilation/perfusion lung scans in pulmonary embolism. Chest 1993; 104: 1468–1471.


BLUK077-Stein

March 12, 2007

69

16:48

CHAPTER 69

The addition of clinical assessment to stratification according to prior cardiopulmonary disease further optimizes the interpretation of ventilation–perfusion lung scans

Introduction The diagnostic value of ventilation–perfusion lung scans was enhanced by showing the positive predictive value for pulmonary embolism (PE) relative to the cumulative number of mismatched segmental equivalent perfusion defects [1] (Chapter 67) where one segmental equivalent is one large segmental defect (>75% of a segment) or two moderate size segmental defects (25–75% of a segment) [2]. The positive predictive value for PE, relative to the cumulative number of mismatched vascular perfusion defects (large and/or moderate size segmental defects) was also shown [3] (Chapter 68). Interpretation of ventilation–perfusion scans on the basis of mismatched vascular defects eliminated the necessity of estimating whether a segmental defect was large or moderate in size. This facilitated the interpretation of ventilation–perfusion lung scans. The diagnostic value of ventilation–perfusion lung scans was further enhanced by combining prior clinical assessment in patients stratified according to the presence of prior cardiopulmonary disease [4]. Clinical assessment was based on intuitive judgment. By combining prior clinical assessment with stratification of patients according to the presence or absence of prior cardiopulmonary disease, a family of curves was derived which allowed an accurate assessment of the positive predictive value of PE based upon the number of mismatched segmental equivalent defects or upon the number of mismatched vascular defects [4]. The families of curves were comparable, irrespec-

304

tive of whether assessment was based upon segmental equivalent defects or vascular defects [4]. The primary advantage of the use of vascular equivalent defects, as indicated previously [3] (Chapter 68), is the ability to assess ventilation–perfusion lung scans without differentiating large segmental defects from moderate size segmental defects.

Patients with no prior cardiopulmonary disease Patients with no prior cardiopulmonary disease, irrespective of whether they were evaluated on the basis of mismatched segmental equivalent perfusion defects (Figure 69.1) or mismatched vascular perfusion defects (Figure 69.2), showed comparable families of curves. A high likelihood clinical assessment, in combination with ≥0.5 mismatched segmental equivalent perfusion defects or ≥1 mismatched vascular perfusion defect resulted in a positive predictive value for PE of 100% (Tables 69.1 and 69.2) [4]. Among patients with no prior cardiopulmonary disease and an intermediate likelihood clinical assessment, a positive predictive value for PE ≥80% was shown with ≥1 mismatched segmental equivalent perfusion defect and ≥2 mismatched vascular perfusion defects. To achieve ≥90% positive predictive value for PE, ≥4 mismatched segmental equivalent perfusion defects or ≥5 mismatched vascular perfusion defects were required (Tables 69.1 and 69.2) [4]. Among patients with no prior cardiopulmonary disease who had a low likelihood clinical assessment, ≥1


BLUK077-Stein

March 12, 2007

CHAPTER 69

16:48

Clinical assessment and prior cardiopulmonary disease for V–Q interpretation

305

Figure 69.1 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched segmental equivalent perfusion defects among patients with no prior cardiopulmonary disease (NO CPD). Patients were categorized according to a high likelihood clinical assessment (CLIN HIGH), intermediate likelihood clinical assessment (CLIN INTER), or low likelihood clinical assessment (CLIN LOW). Only few patients had a low likelihood clinical assessment and ≥1 mismatched segmental perfusion defects. There was, therefore, fluctuation of the CLIN LOW curve. (Reprinted with permission from Stein et al. [4].)

mismatched segmental equivalent perfusion defects or ≥2 mismatched vascular perfusion defects resulted in a positive predictive value for PE of ≥80%. With a larger number of mismatched defects there was inconsistency of the data due to small numbers of patients (Tables 69.1 and 69.2) [4].

Patients with prior cardiopulmonary disease Among patients with a high probability clinical assessment and prior cardiopulmonary disease, ≥0.5 mis-

Figure 69.2 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched vascular perfusion defects among patients with no prior cardiopulmonary disease (NO CPD). Abbreviations as in Figure 69.1. Only few patients had a low likelihood clinical assessment and ≥1 mismatched vascular defects. There was, therefore, fluctuation of the CLIN LOW curve. (Reprinted with permission from Stein et al. [4].)

matched segmental equivalent perfusion defects and ≥1 mismatched vascular perfusion defects resulted in a positive predictive value for PE ≥88% (Figures 69.3 and 69.4, Tables 69.3 and 69.4) [4]. Among patients with an intermediate probability clinical assessment, ≥1.5 mismatched segmental equivalent defects and ≥3 mismatched vascular defects indicated a positive predictive value ≥80%. To achieve ≥90% positive predictive value for PE, ≥3.5 mismatched segmental equivalent defects or ≥5 mismatched vascular defects were required (Figures 69.3 and 69.4, Tables 69.3 and 69.4) [4].


BLUK077-Stein

March 12, 2007

16:48

Table 69.1 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched segmental equivalent defects and clinical

assessment among patients with no prior cardiopulmonary disease (n = 324).

#PE/#PTS (%)

Mismatched segmental defects

High clinical likelihood

Intermed clin likelihood

Low clinical likelihood

≥0

27/31 (87)∗

107/251 (43)

16/42 (38)†

≥0.5

19/19 (100)‡

79/101 (78)

9/15 (60)§

≥1

18/18 (100)

64/78 (82)

7/8 (88)

≥1.5

16/16 (100)

58/67 (87)

6/7 (86)

≥2

15/15 (100)

51/57 (89)

4/5 (80)

≥2.5

14/14 (100)

46/52 (88)

3/4 (75)

≥3

14/14 (100)

41/47 (87)

2/3 (67)

≥3.5

13/13 (100)

38/43 (88)

1/2 (50)

≥4

13/13 (100)

35/39 (90)

1/2 (50)

≥4.5

11/11 (100)

31/32 (97)

1/2 (50)

≥5

11/11 (100)

30/31 (97)

1/2 (50)

≥5.5

10/10 (100)

29/30 (97)

1/2 (50)

≥6

8/8 (100)

26/27 (96)

1/1 (100)

≥6.5

8/8 (100)

26/27 (96)

1/1 (100)

≥7

8/8 (100)

24/25 (96)

1/1 (100)

≥7.5

7/7 (100)

21/22 (95)

1/1 (100)

Probabilities are higher than reported by Stein et al. [4] because chi-square probabilities previously were calculated using Yates’ correction, which was unnecessarily conservative. ∗ P < 0.001. † P < 0.001. ‡ P < 0.05: High vs. intermediate clinical likelihood. § P < 0.01: High vs. low clinical likelihood. PE, pulmonary embolism; PTS, patients; Intermed, intermediate; Clin, clinical. Modified and reprinted with permission from Stein et al. [4].

Table 69.2 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched vascular defects and clinical assessment

among patients with no prior cardiopulmonary disease (n = 324).

#PE/#PTS (%)

Mismatched vascular defects

High clinical likelihood

Intermediate clinical likelihood

Low clinical likelihood

≥0

27/31 (87)∗

107/251 (43)

16/42 (38)†

≥1

19/19 (100)‡

79/101 (78)

9/15 (60)§

≥2

16/16 (100)

60/70 (86)

7/8 (88)

≥3

16/16 (100)

50/56 (89)

4/5 (80)

≥4

15/15 (100)

42/48 (88)

2/3 (67)

≥5

11/11 (100)

36/39 (92)

1/2 (50)

≥6

10/10 (100)

31/32 (97)

1/2 (50)

≥7

8/8 (100)

28/29 (97)

1/2 (50)

≥8

8/8 (100)

26/27 (96)

1/1 (100)

Probabilities are higher than reported by Stein et al. [4] because chi-square probabilities previously were calculated using Yates’ correction, which was unnecessarily conservative. ∗ P < 0.001. † P < 0.001. ‡ P < 0.05: High vs. intermediate. § P < 0.01: High vs. low. PE, pulmonary embolism; PTS, patients; CI, 95% confidence interval. Modified and reprinted with permission from Stein et al. [4].

306


BLUK077-Stein

March 12, 2007

CHAPTER 69

16:48

Clinical assessment and prior cardiopulmonary disease for V–Q interpretation

307

Figure 69.3 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched segmental equivalent defects among patients with prior cardiopulmonary disease (ANY CPD). Abbreviations as in Figure 69.1. Only few patients had a low likelihood clinical assessment and ≥1 mismatched segmental perfusion defects. There was, therefore, fluctuation of the CLIN LOW curve. (Reprinted with permission from Stein et al. [4].)

Among patients with a low probability clinical assessment, and prior cardiopulmonary disease, the positive predictive value for PE did not exceed 33% (Figures 69.2 and 69.4, Tables 69.2 and 69.4). Few patients, however, had ≥1 mismatched segmental equivalent or ≥2 mismatched vascular defects [4]. Presentation of the data as continuous numbers of mismatched defects provided a positive predictive value for PE based upon the ventilation–perfusion scan characteristics of any particular patient. This flexibility was not possible with the criteria used for ventilation– perfusion scan interpretation data as presented in the

Figure 69.4 Positive predictive value of pulmonary embolism (PE) relative to the cumulative number of mismatched vascular defects among patients with prior cardiopulmonary disease (ANY CPD). Abbreviations as in Figure 69.1. Only few patients had a low likelihood clinical assessment and ≥1 mismatched vascular defects. There was, therefore, fluctuation of the CLIN LOW curve. (Reprinted with permission from Stein et al. [4].)

Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED), which employed fixed numbers of defects for various probabilities of PE [5]. Clinical assessment was important [4]. Among patients with no prior cardiopulmonary disease, a high likelihood clinical assessment was indicative of PE in >85% of patients, irrespective of the findings on the ventilation–perfusion lung scan. Also, a sufficiently large numbers of mismatched segmental equivalent perfusion defects or mismatched vascular perfusion defects negated a low or intermediate likelihood clinical impression.


BLUK077-Stein

March 12, 2007

16:48

Table 69.3 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched segmental equivalent perfusion defects and

#PE/#PTS (%)

Mismatched segmental defects ≥0

clinical assessment among patients with prior cardiopulmonary disease (n = 569).

High clinical likelihood ∗

24/30 (80)

Intermed clin likelihood

Low clinical likelihood

144/396 (36)

† †

13/143 (9)‡ 3/15 (20)‡ 3/11 (27)‡

≥0.5

17/19 (89)

99/139 (71)

≥1.0

16/18 (89)

86/115 (75)†

≥1.5

15/17 (88)

78/98 (80)

1/6 (17)

≥2.0

13/14 (93)

72/88 (82)

1/6 (17)

≥2.5

13/14 (93)

68/80 (85)

1/5 (20)

≥3.0

13/14 (93)

63/73 (86)

1/5 (20)

≥3.5

12/13 (92)

61/67 (91)

1/4 (25)

≥4.0

11/12 (92)

59/65 (91)

1/3 (33)

≥4.5

11/12 (92)

55/60 (92)

1/3 (33)

≥5.0

10/11 (91)

51/54 (94)

1/3 (33)

≥5.5

10/10 (100)

47/50 (94)

1/3 (33)

≥6.0

10/10 (100)

45/48 (94)

1/3 (33)

≥6.5

9/9 (100)

44/47 (94)

1/3 (33)

≥7.0

8/8 (100)

41/44 (93)

0/2 (0)

≥7.5

7/7 (100)

39/41 (95)

0/2 (0)

Probabilities are higher than reported by Stein et al. [4] because chi-square probabilities previously were calculated using Yates’ correction, which was unnecessarily conservative. ∗ P < 0.001: High vs. intermediate. † P < 0.001: Intermediate vs. low. ‡ P < 0.001: High vs. low. PE, pulmonary embolism; PTS, patients; Intermed, intermediate; Clin, clinical. Modified and reprinted with permission from Stein et al. [4].

Table 69.4 Positive predictive value of pulmonary embolism in relation to the cumulative number of mismatched vascular perfusion defects and clinical

#PE/#PTS (%)

Mismatched vascular defects ≥0

assessment among patients with prior cardiopulmonary disease (n = 569).

High clinical likelihood 24/30 (80)

Intermed clin likelihood 144/396 (36)

† †

Low clinical likelihood 13/143 (9)‡ 3/15 (20)‡

≥1

17/19 (89)

99/139 (71)

≥2

16/18 (89)

81/103 (79)§

2/7 (29)¶

≥3

13/14 (93)

71/86 (83)

1/6 (17)

≥4

13/14 (93)

63/72 (88)

1/5 (20)

≥5

12/13 (92)

59/64 (92)

1/4 (25)

≥6

11/11 (100)

52/55 (95)

1/3 (33)

≥7

10/10 (100)

46/49 (94)

1/3 (33)

≥8

7/7 (100)

42/45 (93)

0/2 (0)

Probabilities are higher than reported by Stein et al. [4] because chi-square probabilities previously were calculated using Yates’ correction, which was unnecessarily conservative. ∗ P < 0.001: High vs. intermediate. † P < 0.001. ‡ P < 0.001. § P < 0.01: Intermediate vs. low. ¶ P < 0.01: High vs. low. PE, pulmonary embolism; PTS, patients; Intermed, intermediate; Clin, clinical. Modified and reprinted with permission from Stein et al. [4].

308


BLUK077-Stein

March 12, 2007

CHAPTER 69

16:48

Clinical assessment and prior cardiopulmonary disease for V–Q interpretation

Among patients with prior cardiopulmonary disease, larger numbers of mismatched perfusion defects were required to negate an intermediate likelihood clinical assessment than among patients with no prior cardiopulmonary disease [4]. A low likelihood clinical assessment among patients with prior cardiopulmonary disease indicated a low positive predictive value for PE, irrespective of the number of mismatched defects, although there were few patients in this category. Prior clinical assessment in PIOPED I was made by physicians who were experienced and interested in PE. Readily obtained objective clinical information can be obtained by any physician or physician’s assistant, however, and processed with neural network computer intelligence [6]. This gives a clinical assessment which is as accurate as that of physicians experienced with PE (Chapter 52).

2

3

4

5

References 1 Stein PD, Gottschalk A, Henry JW, Shivkumar K. Stratification of patients according to prior cardiopulmonary disease and probability assessment based on the num-

6

309

ber of mismatched segmental equivalent perfusion defects: approaches to strengthen the diagnostic value of ventilation/perfusion lung scans in acute pulmonary embolism. Chest 1993; 104: 461–467. Neumann RD, Sostman HD, Gottschalk A. Current status of ventilation–perfusion imaging. Semin Nucl Med 1980; 10: 198–217. Stein PD, Henry JW, Gottschalk A. Mismatched vascular defects: an easy alternative to mismatched segmental equivalent defects for the interpretation of ventilation/perfusion lung scans in pulmonary embolism. Chest 1993; 104: 1468– 1471. Stein PD, Henry JW, Gottschalk A. The addition of clinical assessment to stratification according to prior cardiopulmonary disease further optimizes the interpretation of ventilation/perfusion lung scans in pulmonary embolism. Chest 1993; 104: 1472–1476. A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. Patil S, Henry JW, Rubenfire M, Stein PD. Neural network in the clinical diagnosis of acute pulmonary embolism. Chest 1993; 104: 1685–1689.


BLUK077-Stein

March 28, 2007

70

15:50

CHAPTER 70

Single photon emission computed tomographic perfusion lung scan

Single photon emission computed tomography (SPECT) is a well established and readily available imaging method that is widely used in modern nuclear medicine diagnostics [1]. Provided that a doubleor triple-head camera is used, the acquisition time of a SPECT scan is comparable to that of planar imaging [2]. Radiologists who have used SPECT for the diagnosis of acute pulmonary embolism (PE) suggest that ventilation–perfusion imaging is significantly improved with SPECT [3]. In particular, specificity is thought to be higher with SPECT perfusion lung scans [2, 4] or ventilation–perfusion lung scans [3] compared with planar lung scans. Fewer SPECT perfusion lung scans are interpreted as nondiagnostic compared with planar perfusion lung scans [2]. There are sparse data comparing SPECT perfusion lung scans or SPECT ventilation–perfusion lung scans with planar lung scans using an independent reference standard. A nearly independent composite reference standard was used by Collart et al. [2] (Table 70.1). A SPECT perfusion lung scan may clarify the presence of an apparent stripe sign [5]. A stripe sign is the appearance of normally perfused tissue between the perfusion defect and the pleural surface suggesting a nonpleural-based opacity [6, 7]. Pulmonary embolism is thought to produce a pleural-based defect. Nonpleural-based lesions with a stripe sign are thought to be less likely due to PE [6, 7].

References 1 Reinartz P, Nowak B, Weiss C, Buell U. Acute pulmonary embolism: thin-collimation spiral CT versus planar ventilation–perfusion scintigraphy. Radiology 2004; 232: 621. 2 Collart JP, Roelants V, Vanpee D et al. Is a lung perfusion scan obtained by using single photon emission computed tomography able to improve the radionuclide diagnosis of pulmonary embolism? Nucl Med Commun 2002; 23: 1107– 1113. 3 Lemb M, Pohlabeln H. Pulmonary thromboembolism: a retrospective study on the examination of 991 patients by ventilation/perfusion SPECT using Technegas. Nuklearmedizin 2001; 40: 179–186. 4 Vanninen E, Tenhunen-Eskelinen M, Mussalo H et al. Are three-dimensional surface-shaded SPET images better than planar and coronal SPET images in the assessment of regional pulmonary perfusion? Nucl Med Commun 1997; 18: 423–430. 5 Pace WM, Goris ML. Pulmonary SPECT imaging and the stripe sign. J Nucl Med 1998; 39: 721–723. 6 Sostman HD, Gottschalk A. The stripe sign: a new sign for diagnosis of nonembolic defects on pulmonary perfusion scintigraphy. Radiology 1982; 142: 737–741. 7 Sostman HD, Gottschalk A. Prospective validation of the stripe sign in ventilation–perfusion scintigraphy. Radiology 1992; 184: 455–459.

Table 70.1 SPECT and planar perfusion lung scans in pulmonary embolism. First author

Planar sensitivity

SPECT sensitivity

Planar specificity

SPECT specificity

[Ref]

[n/N (%)]

[n/N (%)]

[n/N (%)]

[n/N (%)]

Collart [2]

12/15 (80)

12/15 (80)

43/55 (78)

53/55 (96)

SPECT, single photon emission computed tomography.

310


BLUK077-Stein

March 11, 2007

71

12:43

CHAPTER 71

Standard and augmented techniques in pulmonary angiography

Introduction Pulmonary angiography is generally accepted as the definitive diagnostic test for pulmonary embolism (PE) (Figure 71.1). Pulmonary angiography has undergone evolutionary development, both in technique and in the identification of angiographic criteria diagnostic of PE. The validity of pulmonary angiography is dependent upon technique. It is usually technically easier to diagnose PE than to confidently exclude it by pulmonary angiography.

rized the angiographic signs into those of major artery occlusion and those of peripheral vessel occlusion [4]. Obstruction of major arteries seemed to be associated with a reduced number of peripheral branches, areas of increased transradiency (oligemia), slow passage

Historical development of pulmonary angiography The first attempt to investigate the appearance of the pulmonary vasculature in the presence of PE was made by Jesser and deTakats in 1941 [1]. They described a reduction of the size of the pulmonary arterial tree and blocked main vessels after embolization of the pulmonary artery of dogs with a barium sulfate suspension. The absence of filling of the pulmonary artery beyond the site of occlusion was also observed by Liberson and Liberson [2] in 1942, in pulmonary angiograms taken after the injection into rabbits of lead fillings mixed with paraffin. The intraluminal filling defect in PE was first described by Lochhead, Roberts, and Dotter in 1952 in dogs embolized with autologous blood clots [3]. They also described incomplete occlusion of vessels, reduced rate of flow through the region of the clot, diminished opacification of distal vessels, and dilation of the pulmonary artery proximal to the embolus. Chrispin, Goodwin, and Steiner, in 1963, based on some right atrial and some pulmonary arterial injections in 7 patients with clinical evidence of PE, catego-

Figure 71.1 Pulmonary angiogram showing intraluminal filling defects (arrows) in left pulmonary artery and its branches.

311


BLUK077-Stein

March 11, 2007

12:43

312

PART III

Diagnosis of acute PE

of contrast material into the capillaries (asymmetrical filling), and failure of the pulmonary veins to fill. Obstruction of smaller vessels was associated with slow disappearance of contrast material from the arteries (prolonged arterial phase), peripheral pruning, and tortuous vessels. Williams and associates in 1963, in addition to previous observations, reported narrowing and irregularity of opacified vessels, which they attributed to circumferential organization of thromboemboli [5]. Smith, Dammin, and Dexter, utilizing postmortem arteriography, showed reduced opacification of small vessels in the periphery of the lung when small vessels were embolized [6]. Sasahara and associates mentioned plaque-like filling defects among the angiographic abnormalities noted in patients with clinical evidence of PE [7]. They noted difficulty of interpretation of pulmonary angiograms in patients with mitral stenosis or chronic lung disease. Dalen et al. [8] obtained selective pulmonary angiograms in dogs embolized with autologous clots (some with radiopaque autologous clots) and observed cutoffs, areas of oligemia, and filling defects to be typical angiographic abnormalities due to pulmonary embolization. Stein and associates divided angiographic signs into those of morphologic significance and those of physiologic significance [9]. Morphologic signs were intraluminal filling defects, abrupt cutoff of an artery, and localized pruning (lack of fine branching). Physiologic signs were oligemia (areas of increased translucency), asymmetrical filling, prolongation of the arterial phase, and bilateral lower-zone filling delay. Absence of fine branching of the pulmonary vessels was subsequently shown to be nonspecific, and may reflect pulmonary vascular destruction such as occurs in emphysema [10, 11]. Abrupt cutoff of vessels may occur with tumor. It became apparent, therefore, that the only abnormality on the pulmonary angiogram that was diagnostic of PE was visualization of the intraluminal filling defect. Other angiographic abnormalities, in a patient with suspected PE who has no prior cardiopulmonary disease, may call attention to suspicious regions.

niques. Pulmonary digital subtraction angiography (DSA) was used in PIOPED II [13–15]. In PIOPED I, techniques of pulmonary angiography were similar to PIOPED II, although cut films were used as was ionic contrast material. The femoral-vein Seldinger technique with a multiple side-holed 4 French size to 8 French size pigtail catheter was used [16]. Small amounts of contrast material (5–8 mL) were injected by hand to check the patency of the inferior vena cava. The catheter was directed into the proximal portion of the pulmonary artery of the lung with the greatest abnormality on the ventilation–perfusion lung scan. The left lung was imaged in the 50◦ right anterior oblique (RAO) and 40◦ left anterior oblique (LAO) projection. The right lung was imaged in the 30◦ RAO and 40◦ LAO projections. Field size was the maximum magnification that allowed visualization of the entire lung in both views. The highest matrix (1024 × 1024 when possible) was used to allow imaging at 6 frames/sec. Low or iso-osmolar 60% contrast material was injected at 20–35 mL/sec for a total of 25–50 mL [12]. If emboli were not identified, superselective injections were made in areas of suspected abnormality. Magnified and coned images were obtained in the same projections [17]. The amount of contrast material for superselective injections was 5–30 mL injected at a rate of 5–20 mL/sec. If no emboli were found in the first lung, or if bilateral angiography in the clinical center was routine, identical techniques were used for the second lung.

Angiographic technique

Criteria for interpretation of acute PE were a partially occlusive filling defect within an arterial branch or a completely occlusive filling defect that leads to termination of the column of contrast agent in a meniscus that outlines the trailing edge of the embolus [21]. A

The techniques used in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED II) [12] may serve as a guide for most recent accepted tech-

Techniques that augment pulmonary DSA Techniques that augment DSA might be employed to reduce the volume of contrast material that is needed and enhance the visualization of small pulmonary emboli in distal branches of the pulmonary artery. Such techniques include cineangiography [18], balloonocclusion cineangiography [14, 19], and wedge arteriography [20] (Figures 71.2–71.4). These techniques are not routinely used, however.

Interpretation of pulmonary DSA


BLUK077-Stein

March 11, 2007

CHAPTER 71

12:43

Pulmonary angiography

313

Figure 71.2 Normal pulmonary wedge arteriogram. Vessels show narrow gradual tapering and numerous fine branches. A background blush of capillary filling and veins draining the segment are shown. Catheter diameter was 2.3 mm. (Reprinted from Stein [20], with permission from Elsevier.)

nonfilling vessel, or hypovascular area were not considered diagnostic, nor were webs, stenoses, or other vessel abnormalities. Criteria for chronic PE are termination of a vessel without a meniscus, mural irregularity suggestive of recanalization, webs, or synechiae [22], as well as focal stenosis and focal calcification.

Quality of pulmonary angiography Digital subtraction angiography was evaluated during the injection of contrast material through a flow

directed balloon catheter [21]. Selective or subselective pulmonary angiograms were obtained during the injection of 20â&#x20AC;&#x201C;30 mL of iodinated contrast material at 10â&#x20AC;&#x201C;15 mL/sec. Among 211 patients, image quality was excellent in 61.1%, adequate in 37.4%, and poor in 1.4%. The catheter employed was analogous to a Swan-Ganz catheter. The average amount of contrast material was 65 mL. The average duration of the examination was 28 minutes. In PIOPED I, 12 of 1111 (1%) of angiograms were not completed, usually because of complications [23]. Among completed angiograms, 35 of 1099 (3%) were


BLUK077-Stein

March 11, 2007

12:43

314

PART III

Diagnosis of acute PE

Figure 71.3 Wedge arteriogram showing a completely obstructed artery just distal to the catheter (arrow). The diameter of the occluded artery is approximately 2.0 mm. Contrast material has been forced in a retrograde direction due to the obstruction, and is seen along the side of the catheter (arrow). (Reprinted from Stein [20], with permission from Elsevier.)

nondiagnostic. This percentage of nondiagnostic angiograms is similar to the 5% of poor angiograms reported by Dalen and associates following an injection in the main pulmonary artery [24]. The angiographic interpretations that PIOPED I reported, however, were consensus readings, adjudicated if necessary by a panel of readers. Individual angiographers in a clinical environment may not achieve the accuracy of these consensus readings.

Outcome in patients with normal pulmonary angiograms Among the 681 patients in PIOPED I, with negative pulmonary angiograms, the diagnosis of no PE was reversed by the Outcome Classification Committee in 4 (1%) [16]. Each of these patients died within 6 days

of the pulmonary angiogram, and PE was found in each at autopsy. The validity of positive pulmonary angiograms could not be assessed. Pulmonary embolism occurred within 1 year in 6 of 380 (1.6%) patients with suspected PE and normal pulmonary angiograms who received no anticoagulants [25].

Reader agreement Overall agreement on all three categories of interpretation (both readers agreed PE was present, PE absent, or PE uncertain) was 81% [23]. There was closer agreement on the presence of PE than on the absence of PE. Both agreed PE was present or both agreed that PE could not be diagnosed with certainty in 92%. Both readers agreed that PE was absent or both agreed that PE could not be excluded with certainty in 82%.


BLUK077-Stein

March 11, 2007

CHAPTER 71

12:43

315

Pulmonary angiography

Figure 71.4 Wedge arteriogram showing intraluminal filling defects in arteries 1.5–2.0 mm in diameter (arrows). (Reprinted from Stein [20], with permission from Elsevier)

The quality of the angiograms had a greater impact upon the agreement on negativity than on positivity [23]. Agreement on positivity with good, fair, and poor quality angiograms was 93, 90, and 98%, respectively. Agreement on negativity was 88, 77, and 54%, respectively [23].

Pulmonary angiograms showed PE in main or lobar pulmonary arteries as well as smaller orders of branches in 217 of 375 (58%) patients with PE. Among these patients, angiographic readers showed an average co-positivity of 98% [26] (Table 71.1). Average co-positivity was the average

Table 71.1 Pulmonary emboli in main or lobar arteries. Reader 1 Reader 2

PE+

PE−

PE?

Total

PE+

207

2

0

209

PE−

3

0

1

4

PE?

3

1

0

4

213

3

1

217

Total

Data are numbers of patients. Average co-positivity = 1/2 (207/213 + 207/209) × 100 = 98%. PE−, pulmonary embolism absent; PE+, pulmonary embolism present; and PE?, pulmonary embolism uncertain. Reproduced from Stein et al. [26], with permission.


BLUK077-Stein

March 11, 2007

12:43

316

PART III

Diagnosis of acute PE

Table 71.2 Largest pulmonary emboli in segmental arteries. Reader 1 Reader 2

PE+

PE−

PE?

Total

PE+

110

6

3

119

PE−

11

0

1

12

PE?

5

0

0

5

126

6

4

136

Total

Data are numbers of patients. Average co-positivity = 1/2 (110/126 + 110/119) × 100 = 90%. PE−, pulmonary embolism absent; PE+, pulmonary embolism present; and PE?, pulmonary embolism uncertain. Reproduced from Stein et al. [26], with permission.

In PIOPED II, the unweighted kappa statistic [27] for DSA was 0.66 [28].

of the agreement on positivity of reader 1 with reader 2 and reader 2 with reader 1, expressed as percent. Pulmonary angiograms in which the largest order of arteries with PE were segmental branches were shown in 136 of 375 (36%) patients with PE. These patients may have had PE in subsegmental branches as well. Among these patients, average copositivity was 90% [26] (Table 71.2). Angiograms of only 22 of 375 (6%) showed PE limited to subsegmental pulmonary arteries. Readers of these angiograms showed an average co-positivity of 66% (Table 71.3). The angiographers agreed with themselves in 64 of 72 (89%) [26]. (See Chapter 72 for discussion of prevalence in proximal and peripheral pulmonary arteries)

Digital subtraction pulmonary angiography: peripheral injection Peripheral injection of an average of 150 mL of contrast material (range 120–220 mL) in patients with suspected acute PE was compared with conventional pulmonary angiography [29]. The peripheral DSA was satisfactory in 41 of 54 patients (76%). Among patients in whom the conventional pulmonary angiogram showed PE, 26 of 37 (70%) showed PE on peripheral DSA. If the PE was large (Miller index >11) the peripheral DSA showed PE in 16 of 20 (80%). Among patients in whom conventional pulmonary angiography was negative, peripheral DSA was negative in 7 of 15 (47%).

Table 71.3 Largest pulmonary emboli in subsegmental arteries. Reader 1 Reader 2

PE+

PE−

PE?

Total 16

PE+

9

7

0

PE−

2

0

3

5

PE?

1

0

0

1

12

7

3

22

Total

Data are numbers of patients. Average co-positivity = 1/2 (9/12 + 9/16) × 100 = 66%. PE−, pulmonary embolism absent; PE+, pulmonary embolism present; and PE?, pulmonary embolism uncertain. Reproduced from Stein et al. [26], with permission.


BLUK077-Stein

March 11, 2007

CHAPTER 71

12:43

317

Pulmonary angiography

References 1 Jesser JH, de Takats G. Visualization of the pulmonary artery during its embolic obstruction. Arch Surg 1941; 42: 1034–1041. 2 Liberson F, Liberson IR. The use of Diodrast in determining the location and extent of pulmonary embolism: an experimental study. Am J Roetgenol 1942; 48: 352–355. 3 Lochhead RP, Roberts DJ, Jr, Dotter CT. Pulmonary embolism, experimental and angiocardiographic study. Am J Roentgenol 1952; 68: 627–633. 4 Chrispin AR, Goodwin JF, Steiner RE. The radiology of obliterative pulmonary hypertension and thromboembolism. Br J Radiol 1963; 36: 705–714. 5 Williams JR, Wilcox C, Andrews GJ, Burns RR. Angiography in pulmonary embolism. JAMA 1963; 184: 473–476. 6 Smith GT, Dammin GJ, Dexter L. Postmortem arteriographic studies of the human lung in pulmonary embolization. JAMA 1964; 188: 143–151. 7 Sasahara AA, Stein M, Simon M, Littmann D. Pulmonary angiography in the diagnosis of thromboembolic disease. N Engl J Med 1964; 270: 1075–1081. 8 Dalen JE, Mathur VS, Evans H et al. Pulmonary angiography in experimental pulmonary embolism. Am Heart J 1966; 72: 509–520. 9 Stein PD, O’Connor JF, Dalen JE et al. The angiographic diagnosis of acute pulmonary embolism: evaluation of criteria. Am Heart J 1967; 73: 730–741. 10 Stein PD, Leu JD, Welch MH, Guenter CA. Pathophysiology of the pulmonary circulation in emphysema associated with alpha1 antitrypsin deficiency. Circulation 1971; 43: 227–239. 11 Jacobson G, Turner AF, Balchum OJ et al. Pulmonary arteriovenus shunts in emphysema demonstrated by wedge arteriography. Am J Roentgenol 1965; 93: 868–878. 12 Gottschalk A, Stein PD, Goodman LR, Sostman HD. Overview of Prospective Investigation of Pulmonary Embolism Diagnosis II. Semin Nucl Med 2002; 32: 173–182. 13 Goodman PC, Brant-Zawadzki M. Digital subtraction pulmonary angiography. Am J Roentgenol 1982; 139: 305– 309. 14 Ferris EJ, Holder JC, Lim WN et al. Angiography of pulmonary emboli: digital studies and balloon occlusion cineangiography. Am J Roentgenology 1984; 142: 369–373. 15 van Rooij W-JJ, den Heeten GJ, Sluzewski M. Pulmonary embolism: diagnosis in 211 patients with use of selective

16

17 18

19

20

21

22

23

24

25

26

27 28 29

pulmonary digital subtraction angiography with a flowdirected catheter. Radiology 1995; 195: 793–797. A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. Bookstein JJ. Segmental arteriography in pulmonary embolism. Radiology 1969; 93: 1007–1012. Meister SG, Brooks HL, Szucs MM, Banas JS, Jr, Dexter L, Dalen JE. Pulmonary cineangiography in acute pulmonary embolism. Am Heart J 1972; 84: 33–37. Wilson JE, III, Bynum LJ. An improved pulmonary angiographic technique using a balloon-tipped catheter. Am Rev Respir Dis 1976; 114: 1137–1144. Stein PD. Wedge arteriography for the identification of pulmonary emboli in small vessels. Am Heart J 1971; 82: 618–623. Dotter CT. Acquired Abnormalities of the Pulmonary Arteries in Abrams HL, 3rd edn. Little, Brown, Boston, MA, 1983: 743–761. Nicod P, Peterson K, Levine M et al. Pulmonary angiography in severe chronic pulmonary hypertension. Ann Intern Med 1987; 107: 565–568. Stein PD, Athanasoulis C, Alavi A et al. Complications and validity of pulmonary angiography in acute pulmonary embolism. Circulation 1992; 85: 462–468. Dalen JE, Brooks HL, Johnson LW, Meister SG, Szucs MM, Jr, Dexter L. Pulmonary angiography in acute pulmonary embolism: indications, techniques and results in 367 patients. Am Heart J 1971; 81: 175–185. Henry JW, Relyea B, Stein PD. Continuing risk of thromboemboli among patients with normal pulmonary angiograms. Chest 1995; 107: 1375–1378. Stein PD, Henry JW, Gottschalk A. Reassessment of pulmonary angiography for the diagnosis of pulmonary embolism: relation of interpreter agreement to the order of the involved pulmonary arterial branch. Radiology 1999; 210: 689–691. Landis RJ, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 159–174. Stein PD, Goodman LR, Sostman HD. Response to letter. N Engl J Med 2006; 355: 955–956. Musset D, Rosso J, Petitpretz P et al. Acute pulmonary embolism: diagnostic value of digital subtraction angiography. Radiology 1988; 166: 455–459.


BLUK077-Stein

March 28, 2007

72

15:53

CHAPTER 72

Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries

The prevalence of pulmonary embolism (PE) limited to subsegmental pulmonary arteries was evaluated with the pulmonary angiograms obtained in the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) [1]. Among all patients with PE, 22 of 375 (6%) had PE limited to subsegmental branches of the pulmonary artery, 136 of 375 (36%) showed PE in segmental as pulmonary arterial branches and 217 of 375 (58%) showed PE in main or lobar pulmonary arteries [1] (Figure 72.1). Therefore, 94% of patients with PE shown by pulmonary angiography had the PE in segmental branches or larger. The prevalence that we observed of PE limited to subsegmental pulmonary arteries (6%) was lower than reported by Oser and associates, 23 of 76 (30%) and by Goodman and associates, 4 of 11 (36%), but was comparable to the frequency reported by Quinn and associates, 2 of 20

(10%) [2–4]. Goodman and associates also reported a higher prevalence of low probability interpretations of the V–Q scan among patients with PE, 5 of 11 (45%) than observed in PIOPED, 59 of 375 (16%) [2]. The patients in PIOPED, therefore, may have had more severe PE, than the patients reported by Goodman and associates. In patients with intermediate probability V–Q scans, the PE was limited to subsegmental pulmonary arteries in 9 of 160 (6%) [1]. In patients with low probability V–Q scans, the PE was limited to subsegmental pulmonary arteries in 10 of 59 (17%) [1]. Among patients with no prior cardiopulmonary disease who had low probability V–Q scans and PE, the prevalence of PE limited to subsegmental pulmonary arteries was higher, 7 of 23 (30%) [1].

Largest artery showing PE (%)

References

70 60 50 40 30 20 10 0

58 36

6 Main or lobar

Segmental

Subsegmental

Figure 72.1 Largest pulmonary artery showing pulmonary embolism (PE) among 375 patients with PE on pulmonary angiogram. (Data from Stein and Henry [1].)

318

1 Stein PD, Henry JW. Prevalence of acute pulmonary embolism in central and subsegmental pulmonary arteries and relation to probability interpretation of ventilation/perfusion lung scans. Chest 1997; 111: 1246–1248. 2 Goodman LR, Curtin JJ, Mewissen MW et al. Detection of pulmonary embolism in patients with unresolved clinical and scintigraphic diagnosis: helical CT versus angiography. Am J Roentgenol 1995; 164: 1369–1374. 3 Oser RF, Zuckerman DA, Gutierrez FR, Brink JA. Anatomic distribution of pulmonary emboli at pulmonary angiography: implications for cross-sectional imaging. Radiology 1996; 199: 31–35. 4 Quinn MF, Lundell CJ, Klotz TA et al. Reliability of selective pulmonary arteriography in the diagnosis of pulmonary embolism. AJR 1987; 149: 469–471.


BLUK077-Stein

March 11, 2007

73

12:50

CHAPTER 73

Quantification of pulmonary emboli by conventional and CT angiography

CT angiography index Qanadli and associates described an index for quantifying the severity of pulmonary emboli (PE) by CT angiography [1]. The index is defined as the sum of the products of N × D, where N is the value of the proximal clot site equal to the number of segmental branches arising distally, and D is the degree of obstruction defined as 1 for any amount of partial obstruction (<100%) and 2 for total obstruction. For example, the presence of a PE in a segmental artery is scored as 1 point [2]. This value is multiplied by a factor D, defined as 1 for a partially occlusive PE and 2 for a completely occlusive PE [2]. A PE in a lower lobe artery that gives rise to 5 segmental pulmonary arteries would be scored as N = 5. The extent of the lower lobe artery occlusion would be graded as if each of the segmental arteries from it had the same degree of obstruction as the parent lobar artery. If the lower lobe artery was partially occluded, the score would be 5 × 1 = 5. If the lower lobe artery was totally occluded, the score would be 5 × 2 = 10. Each lung is regarded as having 10 segmental branches: 3 to the upper lobe, 2 to the middle lobe and lingual, and 5 to the lower lobe [2]. Accordingly, the maximum obstruction score for each lung is 20 and for each patient is 40 [2]. The score is converted into a percentage [2].

Prognosis based on CT angiography PE index Preliminary observations with the PE index described by Qanadli and associates [1] showed that a cutoff value of 60% was useful for prediction of survival from acute PE [2]. Among patients with a PE index ≥60%, only 1 of 6 (17%) survived [2]. Among patients with a PE

index <60%, 52 of 53 (98%) survived. The surviving patient with a PE index >60% received thrombolytic therapy. Another patient with a PE index of 92.5% died, even with thrombolytic therapy [2].

Angiographic index: Walsh index The “Walsh” angiographic index is an objective index based on PE shown by conventional pulmonary angiography [3]. The index reflects the number of segmental vessels potentially or actually underperfused because of PE in segmental or larger vessels. The maximum value of the index is 18, each unit representing the equivalent of total or partial occlusion of a segmental pulmonary arterial branch. Each definite total occlusion of a pulmonary vessel receives a score as shown in Table 73.1. Each vessel showing a definite partial occlusion (intraluminal filling defect) receives a score as shown in Table 73.2. A single filling defect that extends into >1 anatomical location receives a score for each location up to but not exceeding the maximum designated for each region. Table 73.1 Score for vessel showing total occlusion. Score for each total Vessel involved

occlusion

Main pulmonary artery

9

Intermediate artery

6

Lobar artery Upper

3

Middle (or lingular)

2

Lower

4

Segmental

1

Data from Walsh et al. [3].

319


BLUK077-Stein

March 11, 2007

12:50

320

PART III

Table 73.2 Score for vessels showing a definite partial occlusion (definite intraluminal filling defect). Score for a filling Anatomical location

defect

Central region (pulmonary

3

Diagnosis of acute PE

scale: normal = 0, mild = 1, moderate = 2, or severe = 3 [3]. The objective angiographic score (Walsh index) and the subjective index showed a close correlation (r = 0.86) [3]. Subjective scores were the means of the readings by three radiologists with an interest and experience in PE.

trunk, main pulmonary arteries, intermediate vessel) Lobar or lingular

2

Segmental

1

Data from Walsh et al. [3].

Abnormalities of a single upper lobe region may receive a score of ≤3, middle lobe or lingual ≤2, and lower lobe ≤4. Obstructions in central anatomical regions (primary trunk, main pulmonary arteries, and intermediate vessels) receive scores according to the vessel involved. If the total score of the lung in question is >4 without considering filling defects in the central region, the central filling defects are ignored. All filling defects in a single central region, whether single or multiple, receive a score of 3, if they are considered. If a single vessel contains both a filling defect and an obstruction, only the obstruction is scored. The sum total of scores for all abnormalities in one lung must not exceed 9. Computing score: Score 1 for each definite segmental obstruction or filling defect. Score each obstruction or filling defect in each lobar (or lingular) region and add to it the scores for segmental abnormalities up to the maximum for the lobe or region in question. Score obstructions in each central region according to the listed criteria and add to it the scores for lobar abnormalities up to a maximum of 9 for each lung. If the total score for a lung is less than 5 at this point, increase the score by 3 if any central region filling defects are present.

Subjective evaluation Subjective evaluation of the angiographic severity of PE was assessed on the basis of a four-point subjective

Miller index The Miller index is composed of an objective score for arterial obstruction and a subjectively determined score for reduction of peripheral perfusion [4]. The right pulmonary artery is assigned 9 segmental arteries: 3 in the right upper lobe, 2 in the right middle lobe, and 4 in the right lower lobe. The left lung is assigned seven segmental arteries: 2 in the left upper lobe, 2 to the lingual, and 3 in the left lower lobe. Partial or complete occlusion of a segmental artery receives a point score of 1. Proximal PE are scored equal to the number of segmental arteries arising distally according to the anatomic subdivisions. The maximal score for obstructions is 16. Reduction of peripheral perfusion is scored by dividing each lung into upper, middle, and lower zones and using a four-point scale: normal perfusion = 0, moderate reduction = 1, severe reduction = 2, and absent perfusion = 3. Maximal score of reduced perfusion in both lungs is 18. The maximal Miller index for both lungs is 34.

References 1 Qanadli SD, El Hajjam M, Vieillard-Baron A et al. New CT index to quantify arterial obstruction in pulmonary embolism: comparison with angiographic index and echocardiography. AJR Am J Roentgenol 2001; 176: 1415– 1420. 2 Wu AS, Pezzullo JA, Cronan JJ, Hou DD, Mayo-Smith WW. CT pulmonary angiography: quantification of pulmonary embolus as a predictor of patient outcome—initial experience. Radiology 2004; 230: 831–835. 3 Walsh PN, Greenspan RH, Simon M et al. An angiographic severity index for pulmonary embolism. Circulation 1973; 47/48(suppl II): 101–108. 4 Miller GA, Sutton GC, Kerr IH, Gibson RV, Honey M. Comparison of streptokinase and heparin in treatment of isolated acute massive pulmonary embolism. BMJ 1971; 2: 681–684.


BLUK077-Stein

March 11, 2007

74

13:30

CHAPTER 74

Complications of pulmonary angiography

Introduction The proportion of patients with pulmonary embolism (PE) who are diagnosed by conventional pulmonary angiography was relatively constant at about 10–20% from 1979 to 2003 [1]. Conventional pulmonary angiography still is the diagnostic reference standard for PE. However, because of the invasive nature of pulmonary angiography, imaging techniques that are relatively noninvasive are generally recommended [2]. Complications of pulmonary angiography have been reported by several groups of investigators. Mills and associates reported major complications (including death) in 43 of 1350 (3%) [3]. Three deaths occurred (0.2%), and each was in a patient with cor pulmonale, elevated right ventricular end-diastolic pressure, and elevated pulmonary artery pressure. Nonfatal major complications among these 1350 patients included cardiac perforation in 14 (1%), major arrhythmias in 11 (1%), successfully treated cardiac arrest in 6 (0.4%), myocardial injury in 6 (0.4%), and significant contrast reaction in 4 (0.3%). In the Urokinase Pulmonary Embolism Trial, among 310 patients with PE, 6 of 310 (2%) patients had nonfatal major complications, exclusive of local complications [4]. There were no deaths. Major arrhythmias occurred in 5 and myocardial perforation in 1 patient. Dalen and associates reported significant complications in 13 of 367 (4%) who underwent angiography because of suspected PE [5]. There was 1 death (0.3%). Hull and associates reported 2 reactions to contrast media, but no deaths among 104 patients [6]. Marsh and associates reported 1 death and 1 nonfatal perforation of the right ventricle among 106 patients [7]. Moses and associates reported major complications of pulmonary angiography in 3 of 298 (1%) patients who had suspected PE [8]. There were 2 deaths and

1 major arrhythmia. Stein and associates reported nonlocal complications in 12 of 122 (10%) patients with suspected PE or other pulmonary disorders [9]. Two remote deaths occurred, which probably were not a direct result of pulmonary arteriography. Ranniger reported nonlocal complications in 8 of 241 (3%) [10]. There were no deaths. Novelline and associates reported no major complications or deaths from pulmonary angiography in 302 patients and only minor arrhythmias in 3 (1%) [11]. Complications of pulmonary angiography were reported in 1111 patients who underwent angiography in the Prospective Investigation of Pulmonary Embolism Diagnosis I (PIOPED I) [12]. Major complications were defined as life-threatening, or did not respond promptly to pharmaceutical therapy, or required intensive or prolonged treatment within the hospital. Patients who required cardiopulmonary resuscitation, endotracheal intubation, dialysis, or blood transfusion were defined as having major complications. Only one complication (the most severe complication) was listed for any patient. Nonmajor or minor complications were defined as complications that spontaneously regressed with no apparent residual damage even though prolonged monitoring may have been required. Such complications, in some instances, were important and potentially dangerous events and may have prolonged the hospital stay. The techniques of pulmonary angiography that were employed are relevant. Pulmonary angiograms were performed through a femoral vein using the Seldinger technique with 6–8 French catheters (outside diameter 2–2.7 mm) [13]. Small amounts of contrast material (5–8 mL) were injected by hand to check the patency of the inferior vena cava by fluoroscopy. Seventy-six percent iodinated (ionic) contrast material was used.

321


BLUK077-Stein

March 11, 2007

13:30

322

PART III

Table 74.1 Major complications of angiography among 1111 patients with suspected acute pulmonary embolism. Complication

#PTS (%)

Death

5 (0.5)

Respiratory distress (CPR or intubation)

4 (0.4)

Renal failure (dialysis)

3 (0.3)

Hematoma (transfusion 2 units)

2 (0.1)

Total

14 (1.3)

Only one complication is listed for each patient. CPR, cardiopulmonary resuscitation; PTS, patients. Modified and reproduced from Stein et al. [12], with permission.

Major complications Major complications, including death, occurred in 1.3% of the patients (Table 74.1) [12]. Most of the patients who died or suffered major nonfatal cardiopulmonary complications were in critical condition with severely compromised cardiopulmonary function prior to pulmonary angiography. The severity of the clinical condition usually was not due to massive PE. Death occurred in 0.5% [12]. Although 5 deaths were attributed to pulmonary angiography, the cause of death in 3 patients may not have been due to the

Diagnosis of acute PE

catheterization procedure or pulmonary angiography. At least one of the patients was practically moribund and 4 others had severe underlying cardiac or respiratory disease. The frequency of death (0.5%) was similar to the pooled value reported by others, 9 of 3074 (0.3%) [4â&#x20AC;&#x201C;12]. The circumstances associated with respiratory decompensation in 0.4% of the patients who required endotracheal intubation or cardiopulmonary resuscitation are outlined in Table 74.2. Hematoma of the groin with bleeding severe enough to require two units of blood was observed in 0.2% of the patients. One of the two patients who had such bleeding was receiving anticoagulant therapy. Three patients (0.3%) following angiography developed renal dysfunction, presumably acute tubular necrosis, which required dialysis. Serum creatinine in one patient increased from 1.2 mg/100 mL to 6.0 mg/100 mL and in one patient it increased from 2.4 mg/100 mL to 6.0 mg/100 mL. One patient developed a serum creatinine of 2.7 mg/100 mL, which was associated with pulmonary edema due to fluid overload. The patient required endotracheal intubation. In addition to the patients with major renal complications, 10 patients (0.9%) developed an elevation of the serum creatinine and were treated with fluid balance, diuretics, and in one patient, dopamine. These

Table 74.2 Nonmajor or minor complications of angiography among 1111 patients with suspected acute pulmonary embolism. Complication Respiratory distress (prompt response to drugs) Renal dysfunction (responded to drug therapy and fluid balance)

#PTS (%) 4 (0.4) 10 (0.9)

Angina (monitored in coronary care unit)

2 (0.2)

Hypotension (prompt response to drugs/fluids)

2 (0.2)

Pulmonary congestion (prompt response to drugs) Urticaria, itching, or periorbital edema

4 (0.4) 16 (1.4)

Hematoma (not transfused)

9 (0.8)

Arrhythmia (spontaneous conversion or prompt response to drugs)

6 (0.5)

Subintimal contrast Stain

4 (0.4)

Narcotic overdose (treated with naloxone)

1 (0.1)

Nausea and vomiting

1 (0.1)

Right bundle branch block

1 (0.1)

Total Only one complication is listed for each patient. PTS, patients. Modified and reproduced from Stein et al. [12], with permission.

60 (5.4)


BLUK077-Stein

March 11, 2007

CHAPTER 74

13:30

Complications of pulmonary angiography

patients did not require dialysis and the complications, therefore, were defined as minor. Among these patients, 5 with a previously normal serum creatinine developed levels that ranged from 2.1 to 5.9 mg/100 mL. Five patients with a previously abnormal serum creatinine which ranged from 1.5 to 2.7 mg/100 mL developed a further increase of ≥2 mg/100 mL. Renal insufficiency induced by contrast material was a problem, particularly among elderly patients [12, 14]. All pulmonary angiograms were performed with a standard ionic, high-osmolar contrast agent [13]. Whether the frequency of renal insufficiency would have been lower following the use of a lowosmolar contrast agent is undetermined. There is no evidence that the incidence of pronounced renal toxicity is different following the intravascular use of highversus low-osmolar contrast agents [15–17]. Also, the potential clinical advantages in regard to cardiac function have not yet been clearly demonstrated [18, 19].

Nonmajor or minor complications Among the patients with less severe or minor complications, many of the complications raised serious concern, some were potentially life-threatening and some required prolonged monitoring. Minor complications, therefore, were not necessarily trivial (Table 74.2). Two patients had bronchospasm following the injection of contrast material that was severe enough to discontinue the procedure. One required subcutaneous epinephrine. Two patients developed angina during the catheterization and were transferred to the coronary care unit for monitoring. Two patients (in addition to 1 patient whose most prominent problem was angina) became hypotensive.

Relation of complications to clinical characteristics Major complications occurred more frequently among patients sent for angiography from the medical intensive care unit than patients from elsewhere, 5 of 122 (4%) versus 9 of 989 (1%) [12]. The severity of the underlying condition rendered these patients vulnerable to the procedure. Such critically ill patients often present the most vexing diagnostic dilemmas which require precise information for their optimal management, but the risk of angiography is higher in these

323

patients. Clinical judgment is important in this situation. Minor complications occurred with a similar frequency among patients sent from the medical intensive care unit and from elsewhere, 4 of 122 (3%) versus 56 of 989 (6%). The frequency of complications was not related to the presence or absence of PE. Only 4 of 11 patients with major complications, in whom a diagnosis was established, had PE. Major complications occurred in 1% of the patients with PE and in 1% of the patients who did not have PE. Minor complications occurred in 6% of the patients with PE and in 5% of the patients with no PE. Complications, in general, were not related to age, although renal complications occurred more often in elderly patients [12, 14]. Patients with either major or less severe renal dysfunction, in comparison to patients with no renal dysfunction, were 74 ± 13 years versus 57 ± 17 years (mean ± standard deviation). Complications were unrelated to gender. The frequency of complications was not related to pulmonary artery mean pressure [12]. Pulmonary artery mean pressure was measured in patients with major, minor, and no complications and it was 22 ± 14, 19 ± 9, and 23 ± 11 mm Hg, respectively. Patients with major or minor complications had right atrial mean pressures that were no higher than in patients with no complications. The volume of contrast material injected was not significantly larger in patients with major or minor complications than in patients with no complications, 106 ± 73, 185 ± 84, and 181 ± 61 mL, respectively [12]. Even among patients who suffered renal dysfunction, either major or minor (n = 13), the volume of contrast material injected was not significantly greater than among those who had no complications 207 ± 53 mL versus 181 ± 61 mL. The potential complication of myocardial perforation was specifically monitored and none was observed. The absence of myocardial perforation reflects the use of angiographic catheters with a “pigtail” shaped tip. Such catheters also contributed to the reduced incidence of major cardiac arrhythmias occurring during the passage of the catheter through the right ventricle. The risk of pulmonary hypertension has been emphasized by previous investigators [3], but procedural modifications diminished the risk in PIOPED. In the past, most deaths and serious complications were


BLUK077-Stein

March 11, 2007

13:30

324

reported in patients with pulmonary hypertension and an elevated right ventricular end-diastolic pressure [3]. In this setting, an injection of contrast material into the main pulmonary artery may result in death. In PIOPED, the protocol called for selective contrast injections into the right or left pulmonary artery, rather than the main pulmonary artery. Further, in patients with elevated pulmonary artery pressures, the protocol permitted a reduction of the flow rate and the total amount of contrast injected. Such precautions, however, cannot be viewed as a guarantee for the prevention of serious or fatal complications in patients with pulmonary hypertension. Marsh and associates [7] reported death in a patient with severe pulmonary hypertension after a hand injection of 10 mL of contrast material.

PART III

9

10

11

12

13

References 1 Stein PD, Kayali F, Olson RE. Trends in the use of diagnostic imaging in patients hospitalized with acute pulmonary embolism. Am J Cardiol 2004; 93: 1316–1317. 2 Stein PD, Woodard PK,Weg JG et al. Diagnostic pathways in acute pulmonary embolism: recommendations of the PIOPED II Investigators. Am J Med, 2006; 119: 1048– 1055. 3 Mills SR, Jackson DC, Older RA, Heaston DK, Moore AV. The incidence, etiologies, and avoidance of complications of pulmonary angiography in a large series. Radiology 1980; 136: 295–299. 4 The Urokinase Pulmonary Embolism Trial: A National Cooperative Study. Circulation 1973; 47(suppl II): 38–45. 5 Dalen JE, Brooks HL, Johnson LW, Meister SG, Szucs MM, Jr, Dexter L. Pulmonary angiography in acute pulmonary embolism: indications, techniques, and results in 367 patients. Am Heart J 1971; 81: 175–185. 6 Hull RD, Hirsh J, Carter CJ et al. Pulmonary angiography, ventilation lung scanning, and venography for clinically suspected pulmonary embolism with abnormal perfusion lung scan. Ann Intern Med 1983; 98: 891–899. 7 Marsh JD, Glynn M, Torman HA. Pulmonary angiography: application in a new spectrum of patients. Am J Med 1983; 75: 763–770. 8 Moses DC, Silver TM, Bookstein JJ. The complementary roles of chest radiography, lung scanning, and selective

14

15

16

17

18

19

Diagnosis of acute PE

pulmonary angiography in the diagnosis of pulmonary embolism. Circulation 1974; 49: 179–188. Stein MA, Winter J, Grollman JH, Jr. The value of the pulmonary-artery-seeking catheter in percutaneous selective pulmonary arteriography. Radiology 1975; 144: 299–304. Ranniger K. Pulmonary arteriography: a simple method for demonstration of clinically significant pulmonary emboli. Am J Roentgenol 1969; 106: 558–562. Novelline RA, Baltarowich OH, Athanasoulis CA, Waltman AC, Greenfield AJ, McKusick KA. The clinical course of patients with suspected pulmonary embolism and a negative pulmonary arteriogram. Radiology 1978; 126: 561–567. Stein, PD, Athanasoulis C, Alavi A et al. Complications and validity of pulmonary angiography in acute pulmonary embolism. Circulation 1992; 85: 462–468. A Collaborative Study by the PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 1990; 263: 2753–2759. Stein PD, Henry JW. Age related complications of pulmonary angiography for acute pulmonary embolism. Am J Geriatr Cardiol 1993; 2: 13–22. Schwab SJ, Hlatky MA, Pieper KS et al. Contrast nephrotoxicity: a randomized controlled trial of a nonionic and an ionic radiographic contrast agent. N Engl J Med 1989; 320: 149–153. Taliercio CP, Vlietstra RE, Ilstrup DM et al. A randomized comparison of the nephrotoxicity of iopamidol and diatrizoate in high risk patients undergoing cardiac angiography. J Am Coll Cardiol 1991; 17: 384– 390. Donadio C, Tramonti G, Giordani R et al. Glomerular and tubular effects of ionic and nonionic contrast media (diatrizoate and iopamidol). Contr Nephrol 1988; 68: 212– 219. Hirshfeld JW, Jr, Wieland J, Davis CA et al. Hemodynamic and electrocardiographic effects of ioversol during cardiac angiography. Comparison with iopamidol and diatrizoate. Invest Radiol 1989; 24: 138–144. Bettmann MA, Higgins CB. Comparison of an ionic with a nonionic contrast agent for cardiac angiography. Results of a multicenter trial. Invest Radiol 1985; 20: S70– S74.


BLUK077-Stein

April 6, 2007

75

17:4

CHAPTER 75

Contrast-enhanced spiral CT for the diagnosis of acute pulmonary embolism before the Prospective Investigation of Pulmonary Embolism Diagnosis

Introduction Case reports and short case series of patients with pulmonary embolism (PE) diagnosed by “conventional” computed tomography (CT), but not necessarily contrast-enhanced CT, have been published since 1978 [1–8]. The first evaluation of contrast-enhanced spiral CT compared with pulmonary angiography was published in 1992 [9]. Since then, dozens of original studies, reviews, commentaries, and technical notes related to contrast-enhanced spiral CT for the diagnosis of acute PE have been published. By 2001, a higher proportion of tests in patients with PE were obtained with spiral CT than with ventilation– perfusion (V–Q) lung scans [10] (Figures 75.1 and 75.2). In spite of extensive literature, the accuracy of spiral CT for the diagnosis of acute PE was not established definitively prior to the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED II) (Chapter 77). This lack of clarity is due, in part, to the difficulty in obtaining prospective comparisons with an established reference standard using the criteria of an investigation free of design-related bias [11]. There have been several systematic reviews of the topic [12–15] and other extensive reviews [16–21]. A European Collaborative Trial of the diagnostic performance of spiral CT (ESTIPEP Trial) was done between 1996 and 1997, but the results have been reported only as an abstract [22]. Divergent results were reported, depending on whether the spiral CT was interpreted by local or expert readers, and whether the reference standard was pulmonary angiography

alone or in combination with ventilation–perfusion lung scans [22]. Review of investigations of diagnostic accuracy of CT was based on inclusion criteria recommended by Lijmer and associates for avoidance of bias in studies of diagnostic tests [11]. Tier 1 was defined as those that met the following requirements: (1) the diagnosis of PE was made on the basis of objective tests; (2) patients were studied consecutively; (3) the study was performed prospectively; (4) the CT was read without knowledge of the results of the reference test; (5) all patients studied were suspected of having PE; (6) the study included patients with and without PE; (7) the decision to perform the reference diagnostic test was made independently of the result of the spiral CT; (8) descriptions of the CT methods were sufficiently detailed to permit replication; (9) patients with a broad spectrum of demographics were investigated; (10) patients with all severities of PE were evaluated, with no restriction on the results of preliminary diagnostic tests. Investigations that met most, but not all of these criteria were defined as Tier 2. Investigations that were retrospective were excluded from this evaluation of diagnostic accuracy.

Diagnostic accuracy, Tier 1 studies One Tier 1 investigation of diagnostic accuracy was with a 2-slice scanner [23]; the others used single detector scanners [9, 24–29] (Table 75.1). Pooled data from Tier 1 studies showed a sensitivity of 76% and specificity of 89% [23–26, 28, 29] (Table 75.2). Two

325


BLUK077-Stein

326

April 6, 2007

17:4

PART III

Diagnosis of acute PE

(a)

(b)

Figure 75.1 (a) Axial plane with 16-slice contrast-enhanced spiral CT showing bilateral pulmonary emboli. (Courtesy of Brian Sabb, DO, Department of Radiology, St. Joseph Mercy-Oakland Hospital, Pontiac, Michigan.) (b) Same patient as Figure 75.1a. Coronal plane.


BLUK077-Stein

April 6, 2007

CHAPTER 75

17:4

Contrast-enhanced spiral CT before PIOPED II

(a)

(b)

Figure 75.2 (a) Coronal plane with 16-slice contrast-enhanced CT showing bilateral pulmonary emboli (PE) with massive PE in the right pulmonary artery. (Courtesy of Brian Sabb, DO, Department of Radiology, St. Joseph Mercy-Oakland Hospital, Pontiac, Michigan.) (b) Same patient as Figure 75.2a. Axial plane.

327


328

1 1 1 2 1 1 1 1

Drucker [24]

Nilsson [25]

Perrier [26]

Qanadli [23]

Remy-Jardin [9]

Remy-Jardin [27]

Ruiz [28]

Van Strijen [29]

5

3

3 or 5

5

5

3

3

5

Collimation (mm)

Reference

V–Q or Angio

Angio

Angio

Angio

Angio

V–Q or Angio

Angio

Angio

standard

† Specificity not stated. PE, pulmonary embolism; Angio, angiography; V–Q, ventilation–perfusion.

* Main, lobar or segmental pulmonary artery.

Detector #

First author [Ref]

Sensitivity

88/128 (69)

21/23 (91)

56/62 (90)

51/74 (69)

30/33 (91)

9/15 (60)

[n/N (%)]

Any PE

92/109 (84)

31/38 (82)

89/95 (94)

88/98 (90)

55/57 (96)

26/32 (81)

[n/N (%)]

Specificity

80/93 (86)

39/43 (91)

18/18 (100)

[n/N (%)]

Sensitivity

25/32 (78)

23/24 (96)

[n/N (%)]

Specificity

Proximal PE*

6/28 (21)†

[n/N (%)]

Sensitivity

Distal PE

April 6, 2007

Table 75.1 Tier 1: sensitivity and specificity of individual investigations.

BLUK077-Stein

17:4


329 97/143 (68) 137/154 (89)

Single-slice CT, 3 or

proximal

5 mm collimation

collimation

collimation

Single-slice CT; 5 mm

199/273 (73)

Single-slice CT 102/130 (78)

255/335 (76)

Single and 2-slice CT

Single-slice CT; 3 mm

[n/N (%)]

Sensitivity

Group

117/141 (83)

174/193 (90)

291/334 (87)

380/429 (89)

[n/N (%)]

Specificity

97/121 (80)

102/121 (84)

199/244 (82)

255/304 (84)

value [n/N (%)]

Positive predictive

117/163 (72)

174/202 (86)

291/365 (80)

380/460 (83)

value [n/N (%)]

Negative predictive

3.98

7.97

5.66

6.66

likelihood ratio

Positive

0.38

0.24

0.31

0.27

likelihood ratio

Negative

29]

[9, 27,

[24, 29]

28]

[25, 26,

28, 29]

[24â&#x20AC;&#x201C;26,

28, 29]

[23â&#x20AC;&#x201C;26,

Refs

April 6, 2007

Table 75.2 Tier 1 Investigations: pooled data.

BLUK077-Stein

17:4


BLUK077-Stein

April 6, 2007

17:4

330

investigations used pulmonary angiography and ventilation–perfusion lung scans as the reference standard [26, 29]. All others used pulmonary angiography exclusively [22–25, 28]. When investigators reported results according to two observers [24, 28] the results of observer 1 were used.

Single-slice CT using 3 mm and 5 mm collimation Pooled data of all investigations with single-slice CT showed a sensitivity of 73% [24–26, 28, 29] (Table 75.2). Those with 3 mm collimation showed a sensitivity of 78% [25, 26, 28] (Table 75.2). Specificity was 90%. Positive predictive value with single-slice CT using 3 mm collimation was 84% and negative predictive value was 86% (Table 75.2). Likelihood ratios for a positive test and for a negative test were 7.97 and 0.24, respectively (Table 75.2) [positive likelihood ratio = sensitivity/(1–specificity) and negative likelihood ratio = (1–sensitivity)/specificity]. Data with single-slice CT using 5 mm collimation showed a sensitivity of 68% and specificity of 83% [24, 29] (Table 75.2).

Single-slice CT, proximal PE Sensitivity with single-slice CT using 3 or 5 mm collimation for identifying patients with PE in the main, lobar, or segmental pulmonary artery branches was 89% [9, 27, 29] (Table 75.2). In the main PE, sensitivity with 3 mm collimation was 20 of 20 (100%) and specificity was 101 of 101 (100%) [28]. In the lobar arteries, sensitivity was 61 of 67 (91%) and specificity was 277 of 291 (95%) [28].

Single-slice CT, distal PE Data on distal PE are limited to very few patients (Table 75.1). The number of subsegmental pulmonary artery branches with visible PE increases with decreasing section thickness [30]. More subsegmental arteries would be visible with 2 mm collimation and 0.75 sec/revolution than 3 mm collimation and 1.0 sec/revolution [31].

Tier 2 investigations Individual Tier 2 investigations are shown in Table 75.3, and pooled data of sensitivities, specificities, positive and negative predictive values of Tier 2 investigations are shown in Table 75.4 [32–46]. Single-slice CT

PART III

Diagnosis of acute PE

in Tier 2 investigations showed higher values than in Tier 1 investigations [32, 33, 35–43, 45] (Tables 75.3 and 75.4).

4-Slice CT Only two investigations of diagnostic accuracy prior to PIOPED II evaluated multidetector CT, and both were with 4-slice CT [34, 46]. Pooled data from these studies showed a sensitivity of 45 of 46 (98%) and specificity 132 of 141 (94%) [34, 46] (Table 75.3). With 4-slice CT, visualization of the pulmonary arteries in the middle and peripheral lung zones has been shown to be significantly higher than with single-slice CT, although visualization of the central lung zone was ranked equally with single-slice CT and 4-slice CT [47]. Similarly, 4-slice CT with thin-collimation (1.25 mm) showed improved small pulmonary artery visualization compared with single-slice CT [48].

CT in combination with clinical assessment Single-slice CT The apparent objectivity of CT may lull physicians into a sense of complacency [49]. In the diagnostic process, the probability of PE should be considered in combination with the results of CT [49]. A positive CT in patients with a low probability clinical impression is often falsely positive, and a negative CT in combination with a high probability clinical impression is often falsely negative [49]. Using likelihood ratios calculated from the pooled data, the probability of PE after a positive or negative single-slice CT was estimated in patients stratified according to clinical assessment. Calculations of Tier 1 data showed that the posttest probability of PE would be <84% after a positive single-slice CT in patients with a clinical probability of PE <40% (Figure 75.3). In patients with a negative single-slice CT, if the clinical probability of PE was >70%, there would be more than a 36% posttest probability of PE (Figure 75.3). 4-Slice CT Pooled data of only two investigations CT showed that in patients with a positive 4-slice CT and discordantly low clinical probability of <30%, the posttest probability of PE is <90% (Figure 75.4). Conversely, with a negative 4-slice CT, if the clinical probability of PE is discordantly high, >90%, the calculated posttest probability of PE is >17% (Figure 75.4).


331

1 3

1

1 4 1 1 1 1 1 1 1 1 1 1 and 2 1 4

Blachere [32]

Blum [33]

Coche [34]

Garg [35]‡

Goodman [36]

Kim [37]

Mayo [38]

Otmani [39]

Pruszczyk [40]

Sostman [41]

Steiner [42]

Van Rossum [43]

van Strijen [44]

Velm