Page 1


Complicated Pancreatitis Thesis, Utrecht University, The Netherlands Š 2015 Olaf Bakker ISBN/EAN 978-94-6233-020-7 Design: wenz iD, Wendy Schoneveld Printed by: Gildeprint Drukkerijen B.V., Enschede The research described in this thesis was financially supported by: A grant from The Netherlands Organization for Health Research and Development, the ZonMw Healthcare Efficiency Research Program (170992902), and by a grant from Nutricia (08/KR/AB/002). Printing of this thesis was financially supported by: Chirurgisch Fonds Heelkunde UMC Utrecht, Krijnen Medical Innovations, Angiocare, Vascutek Nederland


C ompl i c ate d Panc re atiti s G e c ompli c e erd e Pan cre atitis (met een samenvatting in het Nederlands)

Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof.dr. G.J. van der Zwaan, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op donderdag 9 juli 2015 des avonds te 6.00 uur door Olaf Jan Bakker geboren op 4 september 1978 te Bergeyk


Promotoren: Prof.dr. H.G. Gooszen Prof.dr. F.L. Moll


Voor Janine en mijn ouders


C ON T E N T S

CHAPTER 1

General introduction Current Gastroenterology Reports 2009

9

PART I EARLY IDENTIFICATION OF COMPLICATED PANCREATITIS CHAPTER 2

Blood urea nitrogen in the early assessment of acute pancreatitis: An international validation study Archives of Internal Medicine 2011

21

CHAPTER 3

Extrapancreatic necrosis without pancreatic parenchymal necrosis: A separate entity in necrotizing pancreatitis? Gut 2013

39

CHAPTER 4

The impact of organ failure on mortality in necrotizing pancreatitis Submitted

57

PART II PREVENTING COMPLICATED PANCREATITIS CHAPTER 5

Timing of enteral nutrition in acute pancreatitis: meta-analysis of individuals using a single-arm of randomized trials Pancreatology 2014

75

CHAPTER 6

Early versus on-demand nasoenteric tube feeding in acute pancreatitis New England Journal of Medicine 2014

93

CHAPTER 7

Timing of cholecystectomy after mild biliary pancreatitis British Journal of Surgery 2011

127


PART III TREATMENT OF COMPLICATED PANCREATITIS CHAPTER 8

A step-up approach or open necrosectomy for necrotizing pancreatitis New England Journal of Medicine 2010

147

CHAPTER 9

Endoscopic transgastric vs surgical necrosectomy for infected necrotizing pancreatitis: a randomized trial JAMA 2012

167

CHAPTER 10 Endoscopic transpapillary stenting or conservative treatment for pancreatic fistulas in necrotizing pancreatitis: multicenter series and literature review Annals of Surgery 2011

187

CHAPTER 11 A conservative and minimally invasive approach to necrotizing pancreatitis improves outcome Gastroenterology 2011

205

CHAPTER 12 Treatment of necrotizing pancreatitis Clinical Gastroenterology & Hepatology 2012

227

PART IV SUMMARY AND FUTURE PERSPECTIVES CHAPTER 13 Summary British Journal of Surgery 2014

225

CHAPTER 14 Future perspectives Nature Reviews Gastroenterology & Hepatology 2014

267

APPENDIX Dutch Summary (Nederlandse Samenvatting) Review committee Acknowledgements (Dankwoord) List of publications Curriculum vitae

282 292 293 298 304


CHAPTER 1 General introduction Current Gastroenterology Reports. 2009 Apr;11(2):104-10

Adapted from: Prevention, detection, and management of infected necrosis in severe acute pancreatitis Olaf J Bakker, Hjalmar C van Santvoort, Marc G Besselink, Erwin van der Harst, H Sijbrand Hofker, Hein G Gooszen Dutch Pancreatitis Study Group


CHAPTER 1

BAC KG ROU N D Acute pancreatitis is an inflammatory disorder of the pancreas, a long flattened gland located deep in the abdomen. The pancreas has two major functions: secretion of digestive enzymes by pancreatic acinar cells through the pancreatic duct into the duodenum and secretion of hormones from the islets of Langerhans into the blood stream. Acute pancreatitis is caused by obstruction of the pancreatic duct or by a direct toxic effect on acinar cells.1,2 Duct obstruction or toxic agents (such as alcohol) induce pathophysiologic processes in the pancreatic acinar cell which eventually cause the premature activation of the inactive enzyme trypsinogen to active trypsin. Active trypsin within the acinar cell instead of within the pancreatic duct leads to autodigestive injury. Autodigestion of the pancreas stimulates an inflammatory response within the pancreatic parenchyma such as infiltration of neutrophils and macrophages, and release of cytokines. This inflammatory response is extremely diverse and varies from minor abdominal complaints followed by swift recovery on one side to climaxing multiple organ failure and almost inevitable death within days on the other.3,4 Inci d ence Acute pancreatitis is actually the most common gastrointestinal disease requiring acute hospital admission.5-6 In the United States in 2009, more than 270.000 patients were discharged with the diagnosis acute pancreatitis. These admissions were related to over 2500 in-hospital deaths and more than 2,5 billion US dollars of healthcare costs. Compared with the year 2000, a 30 percent increase was seen. The exact mechanism causing this increase is unclear, but it is most likely related to an increase in age, obesity, diabetes and gallstone disease.8 In the Netherlands, an average hospital admits around 30 to 50 patients with acute pancreatitis each year.7 D utch Pancre atitis Stu dy Group Although, it has become a common disease, annually per hospital the number of patients with acute pancreatitis who develop complications such as infected pancreatic necrosis or multiple organ failure is very low. Therefore, it is challenging for physicians to become familiar with the disease and promptly recognize and treat complications if they arise. As a result, clinical research in acute pancreatitis mainly consisted of single-center retrospective patient series instead of robust multicenter randomized trials. This changed considerably when the Dutch Pancreatitis Study Group was founded (www.pancreatitis. nl). In 2002 professor Hein Gooszen together with Marc Besselink toke the initiative to organize a nationwide study group. Since its foundation in 2002, the Dutch Pancreatitis Study Group has finished several multicenter randomised trials, published numerous papers in peer review journals, contributed to a revised classification of the disease, 10


GENERAL INTRODUCTION

developed and published a new international guideline on treatment of acute pancreatitis and presented prizewinning research during many national and international meetings. This thesis is the result of 7 years of intensive clinical research from our Study Group. The research aimed to increase the understanding but foremost improve the clinical outcome of patients who are suddenly struck by an attack of acute pancreatitis. O utline of thesis ‘C ompli c ate d p ancre atitis’ Most patients with acute pancreatitis recover uneventfully and are discharged after a few days.9 In around 20 percent of patients, however, the disease course is complicated by major infections, such as infected pancreatic necrosis, or other complications such as organ failure.10-12 Infection of pancreatic necrosis is associated with a mortality of around 15 percent.13 Organ failure in acute pancreatitis is associated with a mortality of up to 30 percent.12 In general, the type of complication that may develop is closely related to the time from symptom onset.10 It has been postulated that acute pancreatitis runs a so-called biphasic clinical course.14 The first or acute phase (approximately the first 2 weeks after symptom onset) is characterized by the systemic inflammatory response syndrome (SIRS), which progresses to actual organ failure in some patients.4 The second or late phase (approximately after 2 weeks) is mostly complicated by major infections such as infection of pancreatic necrosis. These major infections may also lead to organ failure. Inability to adequately treat the infection and reverse existing organ failure may lead to weeks of ongoing sepsis with high mortality rates. During these different phases of the disease, specific complications may be managed differently at different time points. For example, during the first week, pancreatic necrosis is never managed with pancreatic necrosectomy, but after week 3 a necrosectomy may very well be a lifesaving procedure.15 The research presented in PART I EARLY IDENTIFICATION OF COMPLICATED PANCREATITIS was aimed to early identify and accurately evaluate patients with complicated acute pancreatitis. Fortunately, the majority of patients with acute pancreatitis recovers uneventfully and patients who develop complications are relatively rare. However, those patients who require urgent treatment in an intensive care unit setting, in most cases need to be transferred from the general medical ward within the first 24 to 48 hours from hospitalization.16 Therefore, the first 24 to 48 hours from hospitalization is an important phase in the natural course of acute pancreatitis. In this phase, patients who need transfer to the intensive care unit or require other interventions such as aggressive fluid resuscitation or endoscopic retrograde cholangiopancreatography (ERCP) with sphincterotomy, must be identified as soon as possible.

11


CHAPTER 1

This lead to the following research question (see also Table 1): 1. What is the accuracy of serum blood urea nitrogen as early predictor of complicated pancreatitis? (chapter 2) Around weeks 3 and 4 after symptom onset, patients with pancreatitis complicated by pancreatic necrosis or necrosis of the tissue surrounding the pancreas, may develop infection of these necrotic collections.10 The extent and location of necrotic pancreatic or extrapancreatic tissue varies considerably during the disease course and between patients. As a result, through the years all kinds of names and definitions have been given to the different necrotic collections.17 While these vulnerable patients need to be treated optimally, these different names and definitions do not improve the interdisciplinary communication among treating physicians and neither does it improve the comparability of clinical research. It is imperative to uniformly identify and classify these patients. The following research question was formulated: 2. What is difference in clinical outcome between patients with pancreatic parenchymal necrosis and patients with extrapancreatic necrosis without necrosis of the pancreatic parenchyma? (chapter 3) The extent of pancreatic and extrapancreatic necrosis varies among patients. However, the timing, extent and duration of organ failure in acute pancreatitis varies even more. Earlier studies report on the impact of transient organ failure (lasting less than 48 hours) versus persistent organ failure (lasting more than 48 hours) and the impact of organ failure with and without the presence of infection of pancreatic necrosis.4,12 However, detailed data regarding the timing, type and combination (respiratory, cardiovascular or renal failure), or duration of organ failure are not available. These data are needed to clinically evaluate the prognosis of patients with pancreatitis and organ failure and to assist clinical decision-making regarding continuation of treatment. The following research question was formulated: 3. What is the impact of organ failure on mortality in necrotizing pancreatitis? (chapter 4) In the first hours to days after symptom onset, acute pancreatitis may progress from a local inflammatory process within the pancreas to a systemic inflammatory condition with multiple organ failure. This indicates that there is a small window of opportunity for potential therapeutic interventions to interact. In the second part of this thesis, PART II PREVENTING COMPLICATED PANCREATITIS, our attempts at preventing complications in acute pancreatitis are described. 12


GENERAL INTRODUCTION

Table 1. Summary of research questions addressed in this thesis Chapter 2.

What is the accuracy of serum blood urea nitrogen as early predictor of complicated pancreatitis?

3.

What is difference in clinical outcome between patients with pancreatic parenchymal necrosis and patients with extrapancreatic necrosis without necrosis of the pancreatic parenchyma?

4.

What is the impact of organ failure on mortality in necrotizing pancreatitis?

5.

Based on individual patient data from randomized trials, does early enteral tube feeding improve outcome in patients with acute pancreatitis?

6.

Does early enteral tube feeding compared with an oral diet after 72 hours with tube feeding only if necessary, reduce the number of infections or death in patients with pancreatitis who are at risk for complications?

7.

What is the optimal timing of cholecystectomy following an attack of gallstone pancreatitis?

8.

Does a ‘step-up approach’, with the use of percutaneous drainage and minimally invasive necrosectomy if necessary, reduce the number of complications or death compared with primary open necrosectomy in patients with necrotizing pancreatitis?

9.

Does endoscopic transgastric necrosectomy compared with primary necrosectomy reduce the proinflammatory response and clinical outcome in patients with necrotizing pancreatitis?

10.

In patients with pancreatic fistulas following pancreatic necrosectomy, does endoscopic transpapillary stenting compared with conservative treatment reduce the time to fistula closure?

11.

What is the outcome of conservative and minimally invasive approaches in patients with necrotizing pancreatitis?

12.

What is the current state-of-the art approach to patients with necrotizing pancreatitis?

Major infections (infected pancreatic necrosis, pneumonia and bacteremia), have a large impact on outcome in acute pancreatitis.10,11 These infections are thought to be mediated by bacterial translocation from the gut provoked by disturbed intestinal motility, bacterial overgrowth and increased mucosal permeability.18-23 Because of this close association with the gut and the fact that pancreatitis is caused by premature activation of digestive enzymes, the best way to feed patients with pancreatitis has puzzled ‘pancreatologists’ for years. Recent insights show that enteral tube feeding is believed to stimulate intestinal motility – thus reducing bacterial overgrowth - and may increase splanchnic blood flow which helps to preserve the integrity of the gut mucosa.24-27 Non-randomized studies suggest that very early enteral tube feeding started within 24 to 48 hours of admission, compared with a start after 48 hours, improves outcome even further.28-31 The following research questions were formulated: 4. Based on individual patient data from randomized trials, does early enteral tube feeding improve outcome in patients with acute pancreatitis? (chapter 5) 5. Does early enteral tube feeding compared with an oral diet after 72 hours (with tube feeding only if necessary), reduce the number of infections or death in patients with pancreatitis who are at risk for complications? (chapter 6) 13


CHAPTER 1

Patients who suffered an attack of gallstone pancreatitis are at risk for a second attack with potential complicated pancreatitis as long as the gallstones have not been removed. The following research question was formulated: 6. What is the optimal timing of cholecystectomy following an attack of gallstone pancreatitis? (chapter 7) The traditional approach to the treatment of necrotizing pancreatitis complicated by infection of necrotic tissue is open necrosectomy to completely remove the infected necrotic tissue.32 This invasive approach is associated with high rates of complications and death and with a risk of long-term pancreatic insufficiency. As an alternative to open necrosectomy, less invasive techniques, including percutaneous drainage, endoscopic (transgastric) drainage, and minimally invasive retroperitoneal necrosectomy, are increasingly being used.15 These less invasive approaches may reduce complications by minimizing surgical trauma (i.e., tissue damage and a systemic inflammatory response) in already critically ill patients.33-35 In the final part of this thesis, PART III TREATMENT OF COMPLICATED PANCREATITIS, the research is presented that answers the following questions: 7. Does a ‘step-up approach’, with the use of percutaneous drainage and minimally invasive necrosectomy if necessary, reduce the number of complications or death compared with primary open necrosectomy in patients with necrotizing pancreatitis? (chapter 8) 8. Does endoscopic transgastric necrosectomy compared with primary necrosectomy reduce the proinflammatory response and clinical outcome in patients with necrotizing pancreatitis? (chapter 9) 9. In patients with pancreatic fistulas following pancreatic necrosectomy, does endoscopic transpapillary stenting compared with conservative treatment reduce the time to fistula closure? (chapter 10) 10. What is the outcome of conservative and minimally invasive approaches in patients with necrotizing pancreatitis? (chapter 11) 11. What is the current state-of-the art approach to patients with necrotizing pancreatitis? (chapter 12)

14


GENERAL INTRODUCTION

REFERENCES 1. Whitcomb DC. Clinical practice. Acute pancreatitis. N Engl J Med 2006;354(20):214250. 2. Lankisch PG, Apte M, Banks PA. Acute pancreatitis. Lancet. 2015 Jan 20. pii: S01406736(14)60649-8. 3. Di Fabio F, Abu Hilal M, Johnson CD. Acute pancreatitis: mild, severe or potentially fatal. Pancreatology. 2011;11(4):373-5 4. Johnson CD, Abu-Hilal M. Persistent organ failure during the first week as a marker of fatal outcome in acute pancreatitis. Gut. 2004 Sep;53(9):1340-4. 5. Peery AF, Dellon ES, Lund J, Crockett SD, McGowan CE, Bulsiewicz WJ et al. Burden of gastrointestinal disease in the United States: 2012 update. Gastroenterology 2012; 143(5):1179-1187. 6. Fagenholz PJ, Castillo CF, Harris NS, Pelletier AJ, Camargo CA, Jr. Increasing United States hospital admissions for acute pancreatitis, 1988-2003. Ann Epidemiol 2007;17(7):491-7. 7. Spanier B, Bruno MJ, Dijkgraaf MG. Incidence and mortality of acute and chronic pancreatitis in the Netherlands: a nationwide record-linked cohort study for the years 1995-2005. World J Gastroenterol 2013;19(20):3018-26.

11. Wu BU, Johannes RS, Kurtz S, Banks PA. The impact of hospital-acquired infection on outcome in acute pancreatitis. Gastroenterology 2008;135(3):816-20. 12. Petrov MS, Shanbhag S, Chakraborty M, Phillips AR, Windsor JA. Organ failure and infection of pancreatic necrosis as determinants of mortality in patients with acute pancreatitis. Gastroenterology. 2010 Sep;139(3):813-20. 13. van Santvoort HC, Bakker OJ, Bollen TL, Besselink MG, Ahmed AU, Schrijver AM et al. A conservative and minimally invasive approach to necrotizing pancreatitis improves outcome. Gastroenterology 2011; 141(4):1254-1263. 14. Banks PA, Bollen TL, Dervenis C, Gooszen HG, Johnson CD, Sarr MG et al. Classification of acute pancreatitis--2012: revision of the Atlanta classification and definitions by international consensus. Gut 2013; 62(1):102-111. 15. Raraty MG, Halloran CM, Dodd S, Ghaneh P, Connor S, Evans J, et al. Minimal access retroperitoneal pancreatic necrosectomy: improvement in morbidity and mortality with a less invasive approach. Ann Surg. 2010 May;251(5):787-93. 16. Wu BU, Banks PA. Clinical management of patients with acute pancreatitis. Gastroenterology. 2013 Jun;144(6):1272-81.

8. Yadav D, Lowenfels AB. Trends in the epidemiology of the first attack of acute pancreatitis: a systematic review. Pancreas 2006;33(4):323-30.

17. Bollen TL, van Santvoort HC, Besselink MG, van Leeuwen MS, Horvath KD, Freeny PC et al. The Atlanta Classification of acute pancreatitis revisited. Br J Surg 2008; 95(1):6-21.

9. Singh VK, Bollen TL, Wu BU, Repas K, Maurer R, Yu S et al. An assessment of the severity of interstitial pancreatitis. Clin Gastroenterol Hepatol 2011;9(12):1098-103.

18. Ammori BJ, Leeder PC, King RF, Barclay GR, Martin IG, Larvin M et al. Early increase in intestinal permeability in patients with severe acute pancreatitis: correlation with endotoxemia, organ failure, and mortality. J Gastrointest Surg 1999;3(3):252-62.

10. Besselink MG, van Santvoort HC, Boermeester MA, Nieuwenhuijs VB, van Goor H, Dejong CH et al. Timing and impact of infections in acute pancreatitis. Br J Surg 2009;96(3):267-73.

19. Besselink MG, van Santvoort HC, Renooij W, de Smet MB, Boermeester MA, Fischer K et al. Intestinal barrier dysfunction in a randomized trial of a specific probiotic composition in acute pancreatitis. Ann Surg 2009;250(5):712-9.

15


CHAPTER 1

20. Fritz S, Hackert T, Hartwig W, Rossmanith F, Strobel O, Schneider L et al. Bacterial translocation and infected pancreatic necrosis in acute necrotizing pancreatitis derives from small bowel rather than from colon. Am J Surg 2010;200(1):111-7. 21. Nieuwenhuijs VB, Verheem A, van Duijvenbode-Beumer H, Visser MR, Verhoef J, Gooszen HG et al. The role of interdigestive small bowel motility in the regulation of gut microflora, bacterial overgrowth, and bacterial translocation in rats. Ann Surg 1998;228(2):188-93. 22. Rahman SH, Ammori BJ, Holmfield J, Larvin M, McMahon MJ. Intestinal hypoperfusion contributes to gut barrier failure in severe acute pancreatitis. J Gastrointest Surg 2003;7(1):26-35; discussion -6. 23. Van Felius ID, Akkermans LM, Bosscha K, Verheem A, Harmsen W, Visser MR et al. Interdigestive small bowel motility and duodenal bacterial overgrowth in experimental acute pancreatitis. Neurogastroenterol Motil 2003;15(3):267-76. 24. Marik PE. What is the best way to feed patients with pancreatitis? Curr Opin Crit Care 2009;15(2):131-8. 25. McClave SA, Heyland DK. The physiologic response and associated clinical benefits from provision of early enteral nutrition. Nutr Clin Pract 2009;24(3):305-15. 26. Ziegler TR. Parenteral nutrition in the critically ill patient. N Engl J Med. 2009 Sep 10;361(11):1088-97. 27. Marik PE, Zaloga GP. Early enteral nutrition in acutely ill patients: a systematic review. Crit Care Med 2001;29(12):2264-70. 28. Li JY, Yu T, Chen GC, Yuan YH, Zhong W, Zhao LN et al. Enteral nutrition within 48 hours of admission improves clinical outcomes of acute pancreatitis by reducing complications: a meta-analysis. PLoS One 2013;8(6):e64926.

16

29. Petrov MS, Pylypchuk RD, Uchugina AF. A systematic review on the timing of artificial nutrition in acute pancreatitis. Br J Nutr 2009;101(6):787-93. 30. Sun JK, Mu XW, Li WQ, Tong ZH, Li J, Zheng SY. Effects of early enteral nutrition on immune function of severe acute pancreatitis patients. World J Gastroenterol 2013;19(6):917-22. 31. Wereszczynska-Siemiatkowska U, SwidnickaSiergiejko A, Siemiatkowski A, Dabrowski A. Early enteral nutrition is superior to delayed enteral nutrition for the prevention of infected necrosis and mortality in acute pancreatitis. Pancreas 2013;42(4):640-6. 32. Beger HG, Buchler M, Bittner R, Oettinger W, Block S, Nevalainen T. Necrosectomy and postoperative local lavage in patients with necrotizing pancreatitis: results of a prospective clinical trial. World J Surg 1988; 12(2):255-262. 33. Carter CR, McKay CJ, Imrie CW. Percutaneous necrosectomy and sinus tract endoscopy in the management of infected pancreatic necrosis: an initial experience. Ann Surg 2000; 232(2):175-180. 34. Horvath K, Freeny P, Escallon J, Heagerty P, Comstock B, Glickerman DJ et al. Safety and efficacy of video-assisted retroperitoneal debridement for infected pancreatic collections: a multicenter, prospective, single-arm phase 2 study. Arch Surg 2010; 145(9):817-825. 35. van Santvoort HC, Besselink MG, Horvath KD, Sinanan MN, Bollen TL, van RB et al. Videoscopic assisted retroperitoneal debridement in infected necrotizing pancreatitis. HPB (Oxford) 2007; 9(2):156159.


17


PART I EARLY IDENTIFICATION OF COMPLICATED PANCREATITIS

CHAPTER 2 Blood urea nitrogen in the early assessment of acute pancreatitis: An international validation study Archives of Internal Medicine 2011 CHAPTER 3 Extrapancreatic necrosis without pancreatic parenchymal necrosis: A separate entity in necrotizing pancreatitis? Gut 2013 CHAPTER 4 The impact of organ failure on mortality in necrotizing pancreatitis Submitted


PART I EARLY IDENTIFICATION OF COMPLICATED PANCREATITIS


CHAPTER 2 Blood urea nitrogen (BUN) in the early assessment of acute pancreatitis. An international validation study Archives of Internal Medicine 2011

Bechien U Wu, Olaf J Bakker, Georgios I Papachristou, Marc G Besselink, Kathryn Repas, Hjalmar C van Santvoort, Venkata Muddana, Vikesh K Singh, David C Whitcomb, Hein G Gooszen, Peter A Banks


PART I CHAPTER 2

A B S T R AC T B ackg round Objective assessment of acute pancreatitis (AP) is critical to help guide resuscitation efforts. Objectives: 1) Validate serial blood urea nitrogen (BUN) measurement for early prediction of mortality 2) Identify potential thresholds for early changes in BUN to serve as potential targets for resuscitation. Metho ds Secondary analysis of three prospective AP cohort studies (Brigham and Women’s Hospital (BWH): June 2005 – May 2009, Dutch Pancreatitis Study Group (DPSG): March 2004 -March 2007 and University of Pittsburgh Medical Center (UPMC): June 2003 – September 2007). Meta-analysis and stratified multivariate logistic regression adjusted for age-, gender- and creatinine were calculated to determine risk of mortality associated with elevated BUN at admission and rise in BUN at 24 hours. Accuracy of BUN determined by area-under the receiver operator curve (AUC) and compared to APACHE II. A BUN based assessment algorithm was derived on BWH data and validated on the DPSG and UPMC cohorts. Resu lts 1,043 AP cases were included in analysis. In pooled analysis, a BUN≥20 mg/dL was associated with OR 4.6 [2.5, 8.3] for mortality. Any rise in BUN at 24 hours was associated with OR 4.3 [2.3, 7.9] for death. Accuracy of serial BUN measurement (AUC 0.82-0.91) was comparable to APACHE II (AUC 0.72-0.92) in each of the cohorts. A BUN-based assessment algorithm was able to identify patients at increased risk for mortality during the initial period of resuscitation. C onclusi ons We have confirmed the accuracy of BUN for early prediction of mortality in acute pancreatitis and developed an algorithm that may assist physicians in their early resuscitation efforts.

22


BLOOD UREA NITROGEN IN ACUTE PANCREATITIS

I N T RODU C T ION Acute pancreatitis is a common cause for hospitalization worldwide. Each year, there are over 200,000 hospitalizations for acute pancreatitis in the United States at a direct cost exceeding $2 billion.1, 2 Although the majority of patients experience only mild, self-limited disease, up to 20% experience a more severe form of illness that can be lifethreatening.3-6 Most patients that ultimately require treatment in an intensive care setting are transferred from the general medical ward during the first 24-48 hours of hospitalization.7 Therefore, management during the initial 24 hours of hospitalization is an important phase of treatment for patients with acute pancreatitis. Since the publication of the Ranson criteria nearly 40 years ago,8 an objective approach to the early assessment of acute pancreatitis has been the focus of a great deal of clinical investigation. Although numerous clinical scoring systems9-12 and candidate biomarkers13, 14 have been proposed, none of these approaches has become incorporated into routine clinical practice. Despite its limitations, the APACHE II score remains the assessment scale most often recommended for evaluation of patients with acute pancreatitis.3, 5, 15 Practice guidelines in acute pancreatitis also call for vigorous fluid resuscitation in the early treatment of acute pancreatitis.16 To evaluate the effectiveness of initial resuscitation efforts, physicians rely on physical examination findings and results of laboratory tests. In this regard, surrogate markers of intravascular volume status such as hematocrit or blood urea nitrogen (BUN) can provide valuable feedback regarding a patient’s response to initial resuscitation efforts. To help determine which routine laboratory test might be most useful in the early assessment of patients with acute pancreatitis, we recently conducted a retrospective cohort study from 69 U.S hospitals. In that study, we determined that serial measurement of BUN was more accurate than hematocrit for the early prediction of mortality in patients with acute pancreatitis.17 The purpose of this study was to develop an objective approach to early clinical assessment of acute pancreatitis based on serial BUN measurement. Our first aim was to validate early changes in BUN as an early risk factor for mortality using prospectively collected data from the U.S and Netherlands. Our secondary aim was to then develop an algorithm based on early changes in BUN to help guide clinicians in their early resuscitation efforts.

M E T HOD S Pati ent p opu l ati on and d ata c ol l e c ti on We collected data from three concurrently conducted prospective cohort studies in acute pancreatitis. All centers used the same uniform criteria for diagnosis of acute pancreatitis. Specifically, patients were required to satisfy at least two of the following three criteria 23


PART I CHAPTER 2

to confirm the diagnosis: 1) typical epigastric abdominal pain, 2) elevation in amylase/ lipase ≼3 times upper limit of normal and/or 3) confirmatory findings on cross-sectional abdominal imaging. In addition, each of the studies calculated the APACHE II score based on the worst (most extreme) laboratory and vital sign measurement obtained during the initial 24 hours of hospitalization. All patients were managed with initial fast (nil per os) at the time of hospitalization. Specific resuscitation parameters including frequency of laboratory monitoring were determined by individual treating clinicians according to standards of practice within each institution. Within each study cohort, patients were enrolled at the time of admission and followed until hospital discharge. Ethical approval was obtained from the appropriate Institutional Review Board for each of the individual cohort studies as well as from the Partners Healthcare Institutional Review Board for the present combined study. A further description of the study cohorts is provided below: The Markers of Severity in Acute Pancreatitis (MOSAP) study was conducted at Brigham and Women’s Hospital (BWH) from June 2005 - May 2009 for the purpose of evaluation of prognostic markers in acute pancreatitis. Details of the data collection methods used in this study have been previously published.18 All patients directly admitted or transferred to BWH with a diagnosis of acute pancreatitis during the study period were eligible for participation. Patients were identified prospectively based on lipase elevation and/or admission diagnosis codes. Demographic data was obtained through patient interviews. Specific clinical parameters that were prospectively collected included vital signs, laboratory tests and fluid resuscitation parameters. Medical records were obtained for all transferred patients admitted during the study period. The Dutch Pancreatitis Study Group (DPSG) cohort consisted of data collected from 15 Dutch hospitals from March 2004 - March 2007. The hospitals comprised 8 Dutch university medical centers and 7 major teaching hospitals. The ethics review board of each participating hospital approved the protocol of the cohort study. The study cohort consisted of patients with a first attack of acute pancreatitis potentially eligible for participation in a prospective randomized controlled trial on use of probiotics in severe acute pancreatitis (PROPATRIA;19 registration no. ISRCTN38327949). The cohort for the present study was restricted to patients randomized to the placebo arm of the PROPATRIA trial and to patients who could not be randomized due to exclusion criteria e.g. predicted mild disease.20 Patients not randomized in the context of the PROPATRIA trial continued to have prospective data collected regarding clinical parameters and laboratory tests for the duration of their hospitalization. The Severity of Acute Pancreatitis Study (SAPS) was conducted at 2 tertiary care hospitals of the University of Pittsburgh Medical Center (UPMC) from June 2003 - September 2007 for the purpose of evaluation of early prognostic markers in acute pancreatitis. All patients directly admitted or transferred to UPMC with a diagnosis of acute 24


BLOOD UREA NITROGEN IN ACUTE PANCREATITIS

pancreatitis during the study period were eligible for participation. Patients had demographic, radiographic and laboratory data prospectively collected. Details of the collection methods used in the SAPS study have been reported previously.21 To evaluate the role of serial BUN measurement as an early prognostic marker, we limited our analysis to patients that had at least 2 BUN measurements collected within 24 hours of hospitalization. Specifically, patients were required to have BUN collected at admission and a repeat value at 24 hours (Âą6 hours) for study inclusion. For purposes of uniform data presentation, units of BUN are presented as mg/dL.

S TAT I ST IC A L A NA LYSI S We conducted 2 sets of analyses: 1. Evaluation of serial BUN measurement for the early prediction of mortality in acute pancreatitis and 2. Development of an approach to early assessment based on serial BUN measurement. 1. Evaluation of BUN for early prediction of mortality We conducted a meta-analysis using data from each of the aforementioned studies to determine the impact of an elevated BUN at admission or a rise in BUN during the initial 24 hours of hospitalization on risk of in-hospital death. A threshold of 20 mg/dL was used for admission BUN based on previously published findings.17 To evaluate the effect of a rise in BUN during the initial 24 hours of hospitalization, changes in BUN were dichotomized as an increase (≼0 mg/dL) vs. decrease from admission value. Odds ratios and 95% confidence limits were calculated for each parameter for individual studies and then pooled using a random-effects model. The I2-test was used to evaluate for heterogeneity between studies (Comprehensive Meta-analysis Version 2, Biostat, Englewood NJ). We then compared mean admission BUN values and mean change in BUN at 24 hours according to survival status (fatal cases versus survivors, Wilcoxon rank sum test) within study cohorts. The Cochrane-Armitage test of trend was used to evaluate the relationship between extent of elevation in admission BUN as well as extent of rise in BUN during the initial 24 hours of hospitalization with subsequent in-hospital mortality. We used multivariate logistic regression to obtain effect estimates adjusted for age-, gender and serum creatinine for admission BUN and change in BUN at 24 hours. We tested for interaction between change in BUN and change in serum creatinine. To evaluate accuracy of serial BUN measurement for prediction of mortality we plotted a receiver-operator characteristic (ROC) curve based on BUN at admission and change in BUN at 24 hours. Accuracy of this model was compared to a similar creatinine-based model and the APACHE II score by comparison of area-under the ROC curve (AUC) according to the methods described by DeLong et al.22 25


PART I CHAPTER 2

2. Development of a BUN-based approach to early assessment (initial 24 hours) We used recursive partitioning to develop an assessment algorithm that incorporates both the BUN at admission and a patientâ&#x20AC;&#x2122;s change in BUN after initial resuscitation for use in the evaluation of patients with acute pancreatitis. Specifically, we applied Classification and Regression Tree (CART, Salford Systems) analysis to identify specific thresholds for BUN at admission and change in BUN at 24 hours that were associated with either increased or reduced risk of in-hospital mortality in the BWH dataset. We used a supervised approach to CART analysis in which we pre-specified that model development should proceed according to a manner that reflected clinical practice such that admission BUN was forced entry as the first split variable followed thereafter by change in BUN at 24 hours. Based on a priori knowledge, we specified an initial threshold of 20 mg/dL for BUN at admission.17 All model development was performed exclusively using BWH data with 10-fold cross-validation and the Gini index as split criteria. Following model development, we applied the BUN assessment algorithm to the UPMC and DPSG cohorts for validation purposes. Unless otherwise noted, all statistical programming was performed in SAS statistical software version 9.2 (Cary, N.C.). All reported p-values are two-sided with alpha <0.05 threshold used for significance.

R E SU LT S Stu dy Popu l ati on Data from 1,407 potentially eligible cases of acute pancreatitis were available from the three study cohorts. Figure 1 summarizes the composition of the study cohorts after application of the study exclusionary criteria. Demographic and clinical characteristics of the cases included from each study are presented in Table 1a. There were a total of 521 AP cases included from the BWH cohort (median age 52 years, 51% women), 374 cases from the DPSG (median age 57 years, 44% women) and 148 cases from UPMC (median age 53 years, 46% women). The most frequent etiology of pancreatitis was biliary in all three cohorts. There were differences between study cohorts with respect to age, etiology of pancreatitis, proportion of hospital transfers and mortality. Demographic and clinical characteristics of the cases excluded based on lack of serial BUN measurement are presented in Table 1b. A significant proportion of excluded cases were hospital transfers (26% transfers in BWH, 60% transfers in UPMC). Excluded cases also tended to have mild severity of disease as measured by median APACHE II score.

26


BLOOD UREA NITROGEN IN ACUTE PANCREATITIS

Figure 1. Study population. Potentially eligible subjects were those originally included in each individual study cohort. Further eligibility for the present study was restricted to patients with serial measurement of blood urea nitrogen (BUN) during the initial 24 hours of hospitalization.

Table 1A. Demographic and clinical characteristics of the three study cohorts. BWH N=521

DPSG N=374

UPMC N=148

p-value*

52 (41, 62)

57 (41.0, 70.0)

53 (39, 67)

0.003

51%

44%

46%

0.14

Biliary

43%

57%

36%

<0.0001

Alcohol

19%

18%

14%

0.28

Post-ERCP

12%

20%

14%

<0.0001

Idiopathic

11%

5%

27%

<0.0001

12%

13%

43%

<0.0001

7 (4,11)

6 (4,10)

6 (3,10)

0.001

3.7%

7.8%

3.4%

0.04

Age, median IQR Women Etiology

Hospital transfer APACHE II Score, median IQR Mortality

* Ď&#x2021;2 for categorical or ANOVA for continuous variables where appropriate

Table 1B. Demographic and clinical characteristics of patients excluded from the present study due to lack of serial measurement of blood urea nitrogen (BUN). BWH N=121

DPSG N=205

UPMC N=40

52 (41, 62)

57 (41.0, 70.0)

53 (39, 67)

60%

51%

57%

Biliary

41%

62%

40%

Alcohol

15%

18%

20%

Post-ERCP

21%

15%

15%

Idiopathic

11%

5%

10%

Age, median IQR Women Etiology

Hospital transfer APACHE II Score, median IQR Mortality

26%

10%

60%

5 (3, 8)

6 (3,8)

5 (3,6)

2.5%

3.9%

5.0%

27


PART I CHAPTER 2

BUN for the e arly pre di c ti on of mor ta lity Admission BUN values were significantly higher among fatal cases compared to survivors in the BWH and DPSG cohorts (BWH: mean 38.9 mg/dL non-survivors vs 17.7 mg/dL survivors, Wilcoxon rank sum p<0.0001, DPSG: 27.6 mg/dL vs 18.7 mg/dL p<0.0001). There was a non-significant trend in a similar direction in the UPMC cohort (22.8 mg/ dL vs 17.6 mg/dL p=0.11). The Forest Plot for risk of hospital mortality associated with an elevated admission BUN (20 mg/dL) is presented in Figure 2a. In pooled analysis, an elevated BUN at admission was associated with an overall odds-ratio of 4.6 [95% CL 2.5, 8.3] for in-hospital mortality (I2<0.0001). The Forest Plot for the risk of hospital mortality associated with a rise (³0 mg/dL) in BUN during the initial 24 hours of hospitalization is presented in Figure 2b. In pooled analysis, a rise in BUN during the initial 24 hours of hospitalization was associated with an overall odds-ratio of 4.3 [2.3, 7.9] for mortality (I2<0.0001). As shown in Figure 3a, there was a significant trend towards increased mortality with higher levels of BUN at admission for the BWH and DPSG cohorts (BWH CochraneArmitage trend p<0.0001 and DPSG trend p<0.0001) but not in the UPMC study (UPMC trend p=0.14). Change in BUN at 24 hours differed among fatal cases vs. survivors in

Figure 2a. Risk of mortality associated with an elevated blood urea nitrogen level (≥20 mg/dL) at admission. Random-effects model was used for pooled analysis. BWH=Brigham and Women’s Hospital, DPSG=Dutch Pancreatitis Study Group, UPMC=University of Pittsburgh Medical Center

28


BLOOD UREA NITROGEN IN ACUTE PANCREATITIS

Figure 2b. Risk of Mortality associated with Rise in Blood Urea Nitrogen (â&#x2030;Ľ0 mg/dL) at 24 hours. Randomeffects model was used for pooled analysis. BWH=Brigham and Womenâ&#x20AC;&#x2122;s Hospital, DPSG=Dutch Pancreatitis Study Group, UPMC=University of Pittsburgh Medical Center

each of the 3 individual study cohorts (BWH: mean change in BUN +1.3 mg/dL among non-survivors vs. -2.5 mg/dL among survivors p=0.01, DPSG: +6.3 mg/dL vs. -0.5 mg/ dL p<0.0001, UPMC: +12.0 mg/dL vs. -1.8 mg/dL p=0.003). Figure 3b displays the relationship between extent of rise in BUN at 24 hours and subsequent risk for in-hospital mortality. A greater rise in BUN at 24 hours was associated with increased mortality in all the study cohorts (BWH Cochrane-Armitage trend p<0.0001, DPSG p<0.0001 and UPMC p=0.05). We constructed an age-, gender- and creatinine adjusted multivariate logistic regression model to investigate the independent effect of admission and 24-hour change in BUN on risk of mortality. Table 2 lists the adjusted effect estimates for each of the three study cohorts. Both admission BUN and change in BUN at 24-hours were independent predictors of mortality in the BWH and DPSG cohorts. This relationship did not hold for the UPMC dataset. There was evidence of significant positive interaction between change in BUN and change in serum creatinine (p=0.03). Table 3 presents the area under the receiver operator curve (AUC) of a two-variable BUN model (admission value and change at 24 hours) compared to AUC of the APACHE II score. In each of the study cohorts, the BUN model performed similarly to the APACHE 29


PART I CHAPTER 2

II score for prediction of mortality (BUN model AUC=0.82-0.91, APACHE II score AUC=0.72-0.92) and outperformed the APACHE II in the DPSG study cohort (BUN model AUC=0.82 vs. APACHE II AUC=0.72,χ2 p=0.02). A BU N - b a s e d ap pr o a c h t o e a r l y a s s e s s m e nt of a c u t e p a n c r e at i t i s : CART ana lysis We applied a CART analysis using data from the BWH cohort to develop a quantitative approach to patient assessment that incorporates serial BUN measurement during the initial 24 hours of hospitalization. We specified an initial threshold of ≥20 mg/dL for BUN at admission. Among patients with an elevated BUN at admission, a decrease in BUN of at least 5 mg/dL during the first 24 hours of hospitalization was associated with significantly reduced mortality in the BWH cohort (2.9% vs. 15%, χ2 p=0.01). By contrast, a rise of as little as 2 mg/dL or more was associated with increased mortality among patients with a normal BUN at admission (6.7% vs. 0.9% mortality, χ2 p=0.008). The algorithm performed similarly when applied to the DPSG and UPMC study cohorts (Figure 4). In each study cohort, patients with an elevated BUN at admission that did not decline by at least 5 mg/dL after 24 hours of hospitalization were at the highest risk

Figure 3a. In-hospital mortality according to blood urea nitrogen (BUN) level at admission. BWH= Brigham and Women’s Hospital Cohort, n=521 (Cochrane-Armitage trend p<0.0001) DPSG= Dutch Pancreatitis Study Group, n=374 (trend p<0.0001). UPMC= University of Pittsburgh Cohort, n=148 (trend p=0.14)

30

Figure 3b. In-hospital mortality according to extent of rise in blood urea nitrogen. (BUN) among patients whose BUN level rose during initial 24 hours of hospitalization. BWH= Brigham and Women’s Hospital Cohort, n=145 (Cochrane-Armitage trend p<0.0001). DPSG= Dutch Pancreatitis Study Group, n=172 (trend p=0.0002). UPMC= University of Pittsburgh Medical Center, n=55 (trend p=0.05)


BLOOD UREA NITROGEN IN ACUTE PANCREATITIS

for in-hospital mortality (15%-21%). By contrast, patients with an elevated BUN at admission that declined by 5 mg/dL or more after 24 hours had substantially reduced mortality (0-3.2%). Among patients with a normal BUN at admission, those with an increase in BUN of 2 mg/dL or more at 24 hours experienced significantly increased mortality (6-11%) compared to those whose BUN did not rise by this amount (0%-1.0%).

Table 2. Risk of in-hospital mortality associated with elevation in blood urea nitrogen (BUN) at admission and extent of rise in BUN at 24 hours adjusted for age-, gender- and creatinine (both at admission and change in 24 hours).

BWH

Admission BUN Odds-ratio (95% Confidence Limit)

Rise in BUN at 24 hours Odds-ratio (95% Confidence Limit)

2.0 (1.3, 3.0)

2.6 (1.5, 4.4)

DPSG

1.6 (1.1, 2.5)

1.8 (1.3, 2.5)

UPMC

1.0 (0.2, 4.5)

2.9 (0.5, 18.0)

Effect estimates provided are for every increase of 5 mg/dL in BUN at admission or at 24 hours respectively. BWH=Brigham and Women’s Hospital DPSG= Dutch Pancreatitis Study Group UPMC= University of Pittsburgh Medical Center Table 3. Accuracy of serial measurement of blood urea nitrogen (BUN) compared to the APACHE II score. BUN AUC (95% Confidence Limit)

APACHE II AUC (95% Confidence Limit)

Χ2 p-value

BWH

0.84 (0.70, 0.94)

0.89 (0.81, 0.98)

0.30

DPSG

0.82 (0.74, 0.90)

0.72 (0.63, 0.82)

0.02

UPMC

0.91 (0.83, 0.99)

0.92 (0.85, 0.98)

0.88

Accuracy determined by calculation of the area-under the receiver operator characteristic curve (AUC). Accuracy of BUN determined by logistic regression model including BUN value at admission and change in BUN at 24 hours. BWH= Brigham and Women’s Hospital DPSG= Dutch Pancreatitis Study Group UPMC= University of Pittsburgh Medical Center

31


PART I CHAPTER 2

Figure 4. Algorithm for serial BUN measurement in the early assessment of acute pancreatitis. Results based on CART analysis developed using BWH data with subsequent application to DPSG and UPMC study cohorts. BWH= Brigham and Womenâ&#x20AC;&#x2122;s Hospital (0.9% vs. 6.7%, p=0.008, 2.9% vs. 15%, p=0.01) DPSG= Dutch Pancreatitis Study Group (1.0% vs. 15%, p<0.0001, 3.2% vs. 20%, p=0.03) UPMC=University of Pittsburgh Medical Center (0% vs. 11%, p=0.004, 0% vs. 9.1%, p=0.21)

DI S C U S SION We have confirmed that both an elevated BUN at admission and an increase in BUN during the initial 24 hours of hospitalization are risk factors for mortality in acute pancreatitis. Using data from three concurrent prospective cohort studies conducted in the United States and Netherlands we have also demonstrated that serial BUN measurement provided accuracy comparable to the more complex APACHE II score for early prediction of in-hospital mortality. We have incorporated these findings into an assessment algorithm based on early changes in BUN that may provide useful information to physicians during their initial management of patients with acute pancreatitis. A great deal of research has focused on development of approaches to early riskstratification in acute pancreatitis. Various approaches have included clinical scoring systems such as the Ranson criteria,8 Modified Glasgow23 and APACHE II score10 among others9 including our recently developed simplified Bedside Index of Severity in Acute Pancreatitis (BISAP) score.12 Additional approaches have included use of inflammatory markers such as C-reactive protein14 or interleukins (IL)-6 and IL-8.24 Prior studies have also evaluated the potential role of specific routine laboratory tests in predicting outcome in acute pancreatitis.25, 26 However, none of these approaches has gained widespread acceptance in clinical practice. Because there is no specific pharmacologic treatment for 32


BLOOD UREA NITROGEN IN ACUTE PANCREATITIS

acute pancreatitis, an ideal prognostic marker should help guide physicians in their approach to accepted interventions such as fluid resuscitation. The potential importance of BUN in the early assessment of acute pancreatitis has been noted previously. A rise in BUN of â&#x2030;Ľ5 mg/dL at 48 hours was one of eleven criteria originally established as part of the Ranson score.8 The BUN has also been featured in several multi-factor clinical scoring systems in acute pancreatitis.9, 12, 25 However, many additional routine laboratory tests have also been suggested for early assessment in acute pancreatitis, most notably hemoconcentration (serum hematocrit).27, 28 Unfortunately, attempts to validate the role of hemoconcentration have failed to confirm the accuracy of this test as a prognostic marker in acute pancreatitis29-31 likely due to the heterogeneity of patient populations studied.31, 32 Early changes in BUN may reflect several important pathophysiologic processes in acute pancreatitis. In addition to intravascular volume depletion, a rise in BUN may be secondary to impairment in renal function or potentially concurrent upper gastrointestinal hemorrhage. Renal failure is a relatively common form of organ dysfunction among patients with acute pancreatitis.33 Nevertheless, an elevated BUN at admission or rise in BUN during the initial 24 hours of hospitalization was an independent risk factor for mortality after adjusting for serum creatinine. The incidence of clinically significant gastrointestinal hemorrhage (>500 mL blood loss) was prospectively tracked in the BWH cohort and found to be low (<2%) making gastrointestinal hemorrhage an unlikely explanation for the impact of increasing BUN on mortality. In a recent study of 69 U.S. hospitals, we determined that serial measurement of BUN was the most accurate routine laboratory test for prediction of mortality when compared to the serum hematocrit, creatinine, calcium, white blood cell count, and glucose.17 That study was limited by use of retrospectively collected administrative data. Since then, Faisst et al. have also determined that an elevated admission BUN was associated with prolonged stay in an intensive care unit among a selected group of patients with severe acute pancreatitis.33 The present multicenter study has now validated the role of serial BUN measurement for prediction of mortality across three independent cohort studies. Although the admission BUN was not a significant risk factor for mortality in the UPMC study cohort, we attribute this finding to the smaller sample size as evidenced by the wide confidence limits for this study cohort depicted in Figures 2a and 2b. This further emphasizes the importance of the multicenter nature of the present study. To extend the practical application of the current study findings, we developed and validated an approach to assessment of patients with acute pancreatitis based on early changes in BUN. Among patients with an elevated BUN at admission (>20 mg/dL), a decrease of at least 5 mg/dL at 24 hours was associated with reduced risk of in-hospital death. By contrast, among patients with a normal BUN at admission, even a slight rise in BUN (2 mg/dL or more) was associated with increased risk of mortality. 33


PART I CHAPTER 2

There were several potential limitations to the present study. First, there were differences with respect to patient demographics and clinical outcomes among the study cohorts. We therefore used a random-effects model in the meta-analysis and conducted a series of stratified analyses to evaluate the effect of early changes of BUN on risk of mortality. A second limitation was the exclusion of patients based on lack of serial BUN measurement. This is a potential source of selection bias as all management decisions including collection of laboratory tests were left to the discretion of each patientâ&#x20AC;&#x2122;s individual treating physician. To evaluate the extent to which this may have impacted our study findings, we have presented a summary of the demographic and clinical characteristics of the patients excluded from the present study (Table 1b). In general, patients excluded from the study had milder disease as evidenced by lower APACHE II scores during the initial 24 hours of hospitalization. There were also a high percentage of transferred patients excluded from the BWH and UPMC study due to an inability to obtain initial hospitalization records. Nevertheless, the mortality rates among the three cohorts included in the present study are consistent with other reports of acute pancreatitis.5, 34, 35 In summary, we have confirmed the accuracy of serial BUN measurement for the early prediction of mortality using data from three concurrently conducted prospective cohort studies of acute pancreatitis. An elevated BUN at admission or a subsequent rise in BUN level during the initial 24 hours of hospitalization was an independent risk factor for mortality. We have developed an algorithm based on early changes in BUN that may help clinicians evaluate a patientâ&#x20AC;&#x2122;s response to early resuscitation efforts in acute pancreatitis. Ack now l e dgements The authors would like to thank the National Pancreas Foundation (NPF), USA and Senter, an agency of the Dutch Ministry of Economic Affairs for their support of the original studies upon which this work is based. Specifically, funding from the NPF was used for data collection in the BWH study cohort and Senter funds were used to help support data collection in the DPSG study cohort.

34


BLOOD UREA NITROGEN IN ACUTE PANCREATITIS

REFERENCES 1. Fagenholz PJ, Castillo CF, Harris NS, Pelletier AJ, Camargo CA, Jr. Increasing United States hospital admissions for acute pancreatitis, 1988-2003. Ann Epidemiol. Jul 2007; 17(7):491-497. 2. Fagenholz PJ, Fernandez-del Castillo C, Harris NS, Pelletier AJ, Camargo CA, Jr. Direct medical costs of acute pancreatitis hospitalizations in the United States. Pancreas. Nov 2007;35(4):302-307. 3. Banks PA, Freeman ML. Practice guidelines in acute pancreatitis. Am J Gastroenterol. Oct 2006;101(10):2379-2400. 4. Whitcomb DC. Clinical practice. Acute pancreatitis. N Engl J Med. May 18 2006; 354(20):2142-2150. 5. Forsmark CE, Baillie J. AGA Institute technical review on acute pancreatitis. Gastroenterology. May 2007;132(5):20222044. 6. Swaroop VS, Chari ST, Clain JE. Severe acute pancreatitis. Jama. Jun 16 2004;291(23):28652868. 7. Harrison DA, D’Amico G, Singer M. Case mix, outcome, and activity for admissions to UK critical care units with severe acute pancreatitis: a secondary analysis of the ICNARC Case Mix Programme Database. Crit Care. 2007;11 Suppl 1:S1. 8. Ranson JH, Rifkind KM, Roses DF, Fink SD, Eng K, Localio SA. Objective early identification of severe acute pancreatitis. Am J Gastroenterol. Jun 1974;61(6):443-451.

of Ranson, Glasgow, APACHE-II, and APACHE-O criteria to predict severity in acute biliary pancreatitis. Int Surg. Jul-Sep 2001;86(3):158-161. 12. Wu BU, Johannes RS, Sun X, Tabak Y, Conwell DL, Banks PA. The early prediction of mortality in acute pancreatitis: a large population-based study. Gut. Dec 2008; 57(12):1698-1703. 13. Johnson CD, Lempinen M, Imrie CW, et al. Urinary trypsinogen activation peptide as a marker of severe acute pancreatitis. Br J Surg. Aug 2004;91(8):1027-1033. 14. Papachristou GI, Whitcomb DC. Inflammatory markers of disease severity in acute pancreatitis. Clin Lab Med. Mar 2005;25(1):17-37. 15. UK guidelines for the management of acute pancreatitis. Gut. May 2005;54 Suppl 3:iii1-9. 16. Gardner TB, Vege SS, Pearson RK, Chari ST. Fluid resuscitation in acute pancreatitis. Clin Gastroenterol Hepatol. Oct 2008;6(10):10701076. 17. Wu BU, Johannes RS, Sun X, Conwell DL, Banks PA. Early changes in blood urea nitrogen predict mortality in acute pancreatitis. Gastroenterology. Jul 2009; 137(1):129-135. 18. Singh VK, Wu BU, Bollen TL, et al. A prospective evaluation of the bedside index for severity in acute pancreatitis score in assessing mortality and intermediate markers of severity in acute pancreatitis. Am J Gastroenterol. Apr 2009;104(4):966-971.

9. Spitzer AL, Barcia AM, Schell MT, et al. Applying Ockham’s razor to pancreatitis prognostication: a four-variable predictive model. Ann Surg. Mar 2006;243(3):380-388.

19. Besselink MG, van Santvoort HC, Buskens E, et al. Probiotic prophylaxis in predicted severe acute pancreatitis: a randomised, double-blind, placebo-controlled trial. Lancet. Feb 23 2008;371(9613):651-659.

10. Larvin M, McMahon MJ. APACHE-II score for assessment and monitoring of acute pancreatitis. Lancet. Jul 22 1989;2(8656): 201-205.

20. Besselink MG, van Santvoort HC, Boermeester MA, et al. Timing and impact of infections in acute pancreatitis. Br J Surg. Mar 2009;96(3):267-273.

11. Osvaldt AB, Viero P, Borges da Costa MS, Wendt LR, Bersch VP, Rohde L. Evaluation

21. Papachristou GI, Muddana V, Yadav D, et al. Comparison of BISAP, Ranson’s, APACHE-

35


PART I CHAPTER 2

II, and CTSI scores in predicting organ failure, complications, and mortality in acute pancreatitis. Am J Gastroenterol. Feb 2010;105(2):435-441; quiz 442. 22. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. Sep 1988;44(3):837-845. 23. Corfield AP, Cooper MJ, Williamson RC, et al. Prediction of severity in acute pancreatitis: prospective comparison of three prognostic indices. Lancet. Aug 24 1985;2(8452): 403-407. 24. Pooran N, Indaram A, Singh P, Bank S. Cytokines (IL-6, IL-8, TNF): early and reliable predictors of severe acute pancreatitis. J Clin Gastroenterol. Sep 2003; 37(3):263-266. 25. Fan ST, Choi TK, Lai EC, Wong J. Prediction of severity of acute pancreatitis: an alternative approach. Gut. Nov 1989;30(11):1591-1595. 26. Muddana V, Whitcomb DC, Khalid A, Slivka A, Papachristou GI. Elevated serum creatinine as a marker of pancreatic necrosis in acute pancreatitis. Am J Gastroenterol. Jan 2009;104(1):164-170. 27. Baillargeon JD, Orav J, Ramagopal V, Tenner SM, Banks PA. Hemoconcentration as an early risk factor for necrotizing pancreatitis. Am J Gastroenterol. Nov 1998;93(11): 2130-2134. 28. Brown A, Orav J, Banks PA. Hemoconcen足 tration is an early marker for organ failure and necrotizing pancreatitis. Pancreas. May 2000;20(4):367-372. 29. Lankisch PG, Mahlke R, Blum T, et al. Hemoconcentration: an early marker of severe and/or necrotizing pancreatitis? A critical appraisal. Am J Gastroenterol. Jul 2001;96(7):2081-2085. 30. Remes-Troche JM, Duarte-Rojo A, Morales G, Robles-Diaz G. Hemoconcentration is a poor predictor of severity in acute pancreatitis. World J Gastroenterol. Nov 28 2005;11(44):7018-7023.

36

31. Gardner TB, Olenec CA, Chertoff JD, Mackenzie TA, Rob er ts on DJ. Hemoconcentration and pancreatic necrosis: further defining the relationship. Pancreas. Aug 2006;33(2):169-173. 32. Wu BU, Conwell DL, Singh VK, et al. Early Hemoconcentration Is Associated With Pancreatic Necrosis Only Among Transferred Patients. Pancreas. Feb 22 2010. 33. Faisst M, Wellner UF, Utzolino S, Hopt UT, Keck T. Elevated blood urea nitrogen is an independent risk factor of prolonged intensive care unit stay due to acute necrotizing pancreatitis. J Crit Care. Mar; 25(1):105-111. 34. Bruennler T, Hamer OW, Lang S, et al. Outcome in a large unselected series of patients with acute pancreatitis. Hepatogastroenterology. May-Jun 2009; 56(91-92):871-876. 35. Lowenfels AB, Maisonneuve P, Sullivan T. The changing character of acute pancreatitis: epidemiology, etiology, and prognosis. Curr Gastroenterol Rep. Apr 2009;11(2):97-103.


37


PART I EARLY IDENTIFICATION OF COMPLICATED PANCREATITIS


CHAPTER 3 Extrapancreatic necrosis without pancreatic parenchymal necrosis: A separate entity in necrotizing pancreatitis? Gut 2013

Olaf J Bakker, Hjalmar C van Santvoort, Marc G Besselink, Marja A Boermeester, Casper H van Eijck, Cornelis H Dejong, Harry van Goor, H Sijbrand Hofker, Usama Ahmed Ali, Hein G Gooszen, Thomas L Bollen for the Dutch Pancreatitis Study Group


PART I CHAPTER 3

A B S T R AC T Obje c tive In the revised Atlanta classification of acute pancreatitis, the term necrotizing pancreatitis also refers to patients with only extrapancreatic fat necrosis without pancreatic parenchymal necrosis (EXPN), as determined on contrast-enhanced computed tomography (CECT). Patients with EXPN are thought to have a better clinical outcome, although robust data are lacking. D esig n We performed a post-hoc analysis of a prospective multicenter database including 639 patients with necrotizing pancreatitis on CECT. All CECTs were reviewed by a single radiologist blinded for the clinical outcome. Patients with EXPN were compared with patients with pancreatic parenchymal necrosis (with or without extrapancreatic necrosis). Outcomes were persistent organ failure, need for intervention and mortality. A predefined subgroup analysis was performed on patients who developed infected necrosis. Resu lts A total of 315 patients with EXPN were compared with 324 patients with pancreatic parenchymal necrosis. Patients with EXPN less often suffered from complications: persistent organ failure (21% versus 45%; P < 0.001), persistent multiple-organ failure (15% versus 36%; P < 0.001), infected necrosis (16% versus 47%; P < 0.001), intervention (18% versus 57%; P < 0.001) and mortality (9% versus 20%; P < 0.001). When infection of extrapancreatic necrosis developed, outcomes between groups were equal (mortality with infected necrosis; EXPN 28% vs pancreatic necrosis 18%, P = 0.16). C onclusi on EXPN causes fewer complications than than pancreatic parenchymal necrosis. It should therefore be considered a separate entity in acute pancreatitis. Outcome in cases of infected necrosis is similar.

40


EXPN: A SEPARATE ENTITY IN NECROTIZING PANCREATITIS?

What is a lre ady k now n ab out this subje c t? • Patients with acute pancreatitis can develop extensive extrapancreatic necrosis without developing necrosis of the pancreatic parenchyma as determined on contrastenhanced computed tomography. What are the ne w f indings? • Patients with extrapancreatic necrosis only have a better prognosis than patients with pancreatic parenchymal necrosis. Outcome in case of infected necrosis is similar. How mig ht it imp ac t clini c a l pr ac ti ce in the fores e e abl e future? • The revised Atlanta classification of acute pancreatitis will distinguish necrotizing pancreatitis into extrapancreatic necrosis only and pancreatic parenchymal necrosis. • To accurately compare outcomes, future studies in necrotizing pancreatitis should report the number of patients with EXPN.

I N T RODU C T ION In a disease as complex as acute pancreatitis, correct terminology and clear definitions are essential in interdisciplinary communication among treating physicians as well as in reports of clinical research. In the 1992 Atlanta classification, necrotizing pancreatitis was defined as diffuse or focal area(s) of non-viable pancreatic parenchyma, typically associated with extrapancreatic fat necrosis or non-enhanced pancreatic parenchyma more than 3 cm in length or involving more than 30% of the area of the pancreas1. This definition did not include patients with extrapancreatic necrosis only (EXPN). Over the years a few small case series have described patients undergoing surgical removal of extensive extrapancreatic necrosis, whilst the pancreatic parenchyma appeared to be viable2-5. Since these first reports, no prospective study has compared the outcome of patients with EXPN with patients with pancreatic parenchymal necrosis (or in short: pancreatic necrosis) in a large consecutive cohort. The primary aim of this study was to investigate whether the rate of complications and mortality of patients with EXPN differs from patients with pancreatic necrosis with or without extrapancreatic necrosis. The second aim is to determine whether the rates of complications differ between patients with EXPN or pancreatic necrosis who develop infected necrosis.

41


PART I CHAPTER 3

M E T HOD S Pati ents This was a post-hoc analysis of a database of a prospective multicenter cohort study.6 Patients with acute pancreatitis were included between March 2004 and November 2008 in all 8 Dutch university medical centers and 13 large teaching hospitals of the Dutch Pancreatitis Study Group. During the study period, all patients admitted with acute pancreatitis were screened for eligibility for the Dutch PROPATRIA and PANTER trials7,8. Regardless of eligibility for the randomized trials, patients were asked for informed consent for registration in the prospective database on admission. Details on the general treatment protocol, data collection and definitions of outcomes (e.g. definition of primary infected necrosis or organ failure) of this cohort have been described previously6. Infected necrosis was defined as a positive culture of pancreatic or extrapancreatic necrosis obtained by means of fine-needle aspiration or from the first drainage procedure or first necrosectomy, or the presence of gas in the peripancreatic collection on contrast-enhanced CT (CECT). In the current study, new analyses were performed to specifically compare patients with pancreatic necrosis (with or without extrapancreatic necrosis) with patients with extrapancreatic necrosis only (EXPN) as determined on CECT. Patients with interstitial pancreatitis were excluded. The following baseline parameters were assessed: age, etiology, American Society of Anaesthesiologists (ASA) classification on admission, Acute Physiology, Age, and Chronic Health Evaluation (APACHE)-II score on admission, Imrie or Modified Glasgow score, C-reactive protein (CRP) during the first 48 hours of admission and number of transferred patients from other hospitals to one of the hospitals of the Dutch Pancreatitis Study Group. The following outcome parameters were assessed: organ failure, multiple-organ failure, development of sterile or infected necrosis, conservative treatment or intervention, and mortality. The study was conducted in accordance with the principles of the Declaration of Helsinki, and was investigator initiated and investigator driven. The ethics review board of each participating hospital approved the study. Patients or their legal representatives gave written informed consent. We adhered to the STROBE statement guidelines for reporting on observational cohort studies9. D ef initi ons of extr ap ancre ati c ne crosis and p ancre ati c ne crosis Pancreatic necrosis was defined as focal or diffuse non-enhancement of the pancreatic gland as determined on CECT. Extrapancreatic necrosis only without pancreatic necrosis (i.e. EXPN; Figure 1) was defined as extrapancreatic morphologic changes exceeding fat stranding with complete enhancement of the pancreatic parenchyma without signs of focal or diffuse non-enhancement able to be determined on the final CECT of 42


EXPN: A SEPARATE ENTITY IN NECROTIZING PANCREATITIS?

A

B

C

D

Figure 1. Extrapancreatic Necrosis (A-D): A and B: Axial and coronal reconstructed contrast-enhanced CT in a 37-year-old male with alcohol-induced pancreatitis 6 days after onset of symptoms. There is normal enhancement of the pancreatic parenchyma (white asterisks) with a poorly marginated peripancreatic collection (arrows) extending in the left anterior pararenal compartment.C and D: Axial and coronal reconstructed contrast-enhanced CT 2 months later, again showing a normal enhancement of the pancreas surrounded by a well encapsulated heterogeneous peripancreatic collection (arrows) with areas of nonliquid (fat) densities (arrowheads) in the left anterior pararenal compartment. During surgery large amount of necrotic material was debrided and pus was drained. Patient recovered uneventfully with preservation of normal pancreatic function.

hospitalization or before any intervention. Patients with both pancreatic and extrapancreatic necrosis were included in the pancreatic necrosis group (Figure 2). Patients with pancreatic necrosis without extrapancreatic necrosis, which is very rare, were also included in the group of pancreatic necrosis.

43


PART I CHAPTER 3

Figure 2. Acute pancreatic parenchymal necrosis: Axial contrast­-enhanced CT in a 45-year-old female with acute biliary pancreatitis. There is nonenhancement of the neck and part of the body of the pancreas (white asterisk) indicative for parenchymal necrosis with normal enhancement of the tail of the pancreas (black asterisk).

Eva lu ati on of imag ing CECT generally was performed approximately 7 days after admission. In patients with clinical deterioration or without clinical improvement, additional CECTs were performed. In patients participating in the PROPATRIA trial, a CECT after 7 to 10 days was part of the protocol. Radiologists in the participating centers assessed the CECT for the presence or absence of necrotizing pancreatitis. Based on their evaluation, patients were included in the prospective database. After closure of the database following completion of the PANTER trial, all CECTs performed during and after hospitalization were reviewed by a single experienced abdominal radiologist (TLB) who was unaware of the patients’ clinical background, possible interventions and the initial radiology report. Because extrapancreatic or parenchymal necrosis can be missed on CECTs performed early in the disease course, all CECTs performed in each patient were reviewed. The final decision on the presence or absence of EXPN or pancreatic necrosis was based on the final CECT of hospitalization or before any intervention. Statisti c a l ana lysis Continuous data are presented as median with interquartile range (IQR). Differences were tested with the Mann-Whitney U test. Proportions were compared by the χ2, the Fisher exact or the linear-by-linear association test, as appropriate. To assess whether EXPN is an independent predictor of clinical outcome, multivariable regression analysis was performed adjusting for potential confounders (e.g. prognostic variables on admission such as age). The following clinical outcomes were analysed: persistent organ failure, persistent multiple-organ failure, infected necrosis, the need for intervention and mortality. EXPN was entered into the model as the main factor. As co-variables, all prognostic variables that were potentially different between patients with EXPN and 44


EXPN: A SEPARATE ENTITY IN NECROTIZING PANCREATITIS?

pancreatic necrosis on admission in univariable analysis (P < 0.2) were included. Results are reported as adjusted odds ratios (OR) with 95%-confidence intervals (CI). A twosided P < 0.05 was considered statistically significant. Analyses were performed with SPSS version 15.0 (SPSS, Chicago, IL).

R E SU LT S Pati ents Of the 639 patients included in the prospective database, 315 patients (49%) had EXPN (determined on CECT) and 324 patients (51%) had signs of pancreatic necrosis (with or without extrapancreatic necrosis) (Figure 3). Only 4 out of 639 patients (0.6%) had pancreatic necrosis without any signs of peripancreatic necrosis and these patients were included in the pancreatic necrosis group. Patients with interstitial pancreatitis were excluded. Baseline characteristics are shown in Table 1. No differences were seen in age, sex, etiology, ASA classification and predicted severity based on APACHE-II scores and CRP levels. The median Imrie score at 48 hours after admission was higher in patients with pancreatic necrosis, 4 vs. 3 (P < 0.001). The percentage of transferred patients was also higher in patients with pancreatic necrosis, 36% vs. 13% (P < 0.001).

Figure 3. Study population

45


PART I CHAPTER 3

Table 1. Admission characteristics of patients with EXPN compared with patients with pancreatic parenchymal necrosis. All patients N = 639

Extrapancreatic necrosis only (EXPN) N = 315

Age – yr

58 (45 - 70)

58 (44 – 72)

58 (45 – 69)

0.24a

Male sex

398 (62%)

184 (58%)

214 (66%)

0.05b

Characteristic

Pancreatic necrosis

P-value

N = 324

Etiology

0.26c

Gallstones

304 (48%)

149 (47%)

155 (48%)

Alcohol abuse

150 (24%)

76 (24%)

74 (23%)

Other

63 (10%)

37 (12%)

26 (8%)

Unknown

122 (19%)

53 (17%)

69 (21%)

ASA# class on admission

0.79d

I (healthy status)

202 (32%)

97 (31%)

105 (32%)

II (mild systemic disease)

347 (54%)

174 (55%)

173 (53%)

III (severe systemic disease)

90 (14%)

44 (14%)

46 (14%)

APACHE-II† score on admission

8 (5 - 11)

7 (5 – 10)

8 (5 – 11)

0.98a

Imrie / modified Glasgow score after 48 hours

3 (2 - 5)

3 (2 – 4)

4 (2 – 5)

<0.001a

Highest CRP level in first 48 h

291 (216 - 382)

293 (216 – 278)

290 (215 – 388)

0.60a

156 (24%)

40 (13%)

116 (36%)

<0.001b

2 (4 – 7)

2 (1 – 4)

6 (3 – 10)

<0.001a

14 (7 – 46)

9 (5 – 21)

34 (9 – 71)

<0.001a

Predicted severity of pancreatitis

Transferred from other hospital No. of CECTs performed per patient* Timing of final CECT* (days after admission)

Continuous variables are median (interquartile range). #ASA stands for American Society of Anaesthesiologists. †APACHE stands for Acute Physiology Age and Chronic Health Evaluation. ‡CRP stands for C-reactive protein. *before discharge or any intervention. aMann-Whitney U test; bFisher’s Exact test; cPearson Chisquare; dLinear-by-linear association

Us e of CECT A median number of 2 (IQR 4 – 7) CECTs were performed in each patient (Table 1). Patients with EXPN were scanned less frequent than patients with pancreatic necrosis; a median number of 2 CECTs per patient (IQR 1 - 4) compared with a median of 6 CECTs per patient (IQR 3 - 10; P < 0.001). The overall median time between onset of symptoms and the final CECT of hospitalization or before any intervention was 14 days (IQR 7 – 46 days). Patients with EXPN had their final CECT earlier than patients with pancreatic necrosis; a median of 9 days (IQR 5 – 21 days) after admission compared with a median of 34 days (IQR 9 – 71 days; P < 0.001).

46


EXPN: A SEPARATE ENTITY IN NECROTIZING PANCREATITIS?

O utcomes The different outcomes of patients with pancreatic necrosis and EXPN are presented in Table 2. All outcomes occurred less often in patients with EXPN. More specifically, patients with EXPN had a significant lower risk of developing single or multiple-organ failure. Twice as many patients with pancreatic necrosis developed persistent organ failure during admission (21% versus 45%; P < 0.001). Persistent multiple-organ failure was also seen in twice as many patients with pancreatic necrosis (15% versus 36%; P < 0.001). Patients with EXPN had a lower risk of developing infected necrosis (16% versus 47%; P < 0.001). The need for intervention was lower in patients with EXPN (18% versus 57%; P < 0.001). Finally, mortality in patients with EXPN was significantly lower (9% versus 20%; P < 0.001).

Table 2. Outcome of patients with EXPN compared with patients with pancreatic parenchymal necrosis. Extrapancreatic necrosis only(EXPN) N = 315

Pancreatic necrosis N = 324

P-value#

77 (24%)

163 (50%)

<0.001

At any time during admission, persistent

66 (21%)

147 (45%)

<0.001

In the first week of admission

47 (15%)

93 (29%)

<0.001

56 (18%)

138 (43%)

<0.001

At any time during admission, persistent

46 (15%)

115 (36%)

<0.001

In the first week of admission

30 (10%)

64 (20%)

<0.001

Characteristic Organ failure At any time during admission

Multiple-organ failure At any time during admission

Sterile or infected necrosis Sterile necrosis

264 (84%)

173 (54%)

<0.001

Primary Infected necrosis

51 (16%)

151 (47%)

<0.001

Conservative treatment

258 (82%)

139 (43%)

<0.001

Any intervention*

57 (18%)

185 (57%)

<0.001

3 (1%)

29 (9%)

<0.001

Conservative treatment or intervention

Emergency laparotomy PCD ** (as first intervention)

37 (12%)

94 (29%)

<0.001

Necrosectomy¶

40 (13%)

129 (40%)

<0.001

29 (9%)

64 (20%)

<0.001

Mortality

#Fisher’s Exact test. * i.e. emergency laparotomy, PCD or necrosectomy. ** PCD stands for Percutaneous Catheter Drainage. ¶ i.e. with or without previous PCD or emergency laparotomy.

47


PART I CHAPTER 3

When we adjusted for potential confounding factors with multivariable regression, patients with EXPN still had better clinical outcomes. After adjusting for male sex, Imrie score and transferred patients, EXPN was independently associated with a lower risk of organ failure, adjusted OR 0.53 (CI 0.37 – 0.78; P < 0.001); multiple-organ failure, adjusted OR 0.48 (CI 0.32 – 0.72, P < 0.001); infected necrosis, adjusted OR 0.30 (CI 0.20 – 0.45, P < 0.001); any intervention, adjusted OR 0.25 (CI 0.17 – 0.38, P < 0.001), and mortality, adjusted OR 0.59 (CI 0.35 – 0.97, P = 0.04). Subg roup ana lysis of primar y infe c te d ne crosis The subgroup of 202 patients who developed infected necrosis was separately analysed. There were no significant differences between patients with infected EXPN and patients with infected pancreatic necrosis regarding all the aforementioned clinical outcomes (Table 3). The incidence of persistent organ failure (55% of EXPN versus 60% of pancreatic necrosis; P = 0.62) and persistent multiple-organ failure (41% of EXPN versus 46% of pancreatic necrosis; P = 0.63) was similar. Almost all patients with infected necrosis required an intervention (92% versus 94%; P = 0.74). The likelihood of recovery

Table 3. Outcome in subgroup of patients with primary infected extrapancreatic or pancreatic parenchymal necrosis. Infected extrapancreatic necrosis only (EXPN) N = 51

Infected pancreatic necrosis

At any time during admission

31 (61%)

101 (67%)

0.50

At any time during admission, persistent

28 (55%)

90 (60%)

0.62

Characteristic

P-value#

N = 151

Organ failure

Multiple-organ failure At any time during admission

26 (51%)

86 (57%)

0.52

At any time during admission, persistent

21 (41%)

70 (46%)

0.63

4 (8%)

9 (6%)

0.74

47 (92%)

142 (94%)

0.74

0 (0%)

4 (3%)

0.58

Conservative treatment or intervention Conservative treatment Any intervention* Emergency laparotomy PCD ** (as first intervention) PCD only*** Necrosectomy¶ Mortality

32 (63%)

84 (56%)

0.42

14 (28%)

34 (23%)

0.57

33 (65%)

108 (72%)

0.23

14 (28%)

27 (18%)

0.16

#Fisher’s Exact test. * i.e. emergency laparotomy, PCD or necrosectomy. ** PCD stands for Percutaneous Catheter Drainage. ***In subgroup of patients with PCD as first intervention. ¶ i.e. with or without previous PCD or emergency laparotomy.

48


EXPN: A SEPARATE ENTITY IN NECROTIZING PANCREATITIS?

following percutaneous catheter drainage without the need for additional necrosectomy was 28% versus 23% (P = 0.57). Finally, mortality did not differ significantly; 28% versus 18% (P = 0.16).

DI S C U S SION This study, evaluating the clinical outcome of a large prospective cohort of patients with necrotizing pancreatitis, shows that patients with EXPN (as determined on CECT) have a better prognosis than patients with pancreatic necrosis. However, in case of infected necrosis, rates of complication and mortality are similar. To our knowledge, this is the first report on clinical outcomes of EXPN in a large unselected cohort of patients with acute necrotizing pancreatitis. The presence of extrapancreatic necrosis as a separate entity in the absence of pancreatic necrosis was first recognised by Howard and Wagner in 19892. In a group of 13 patients, pancreatography after debridement of necrosis showed an intact pancreatic duct in most patients. Subsequently, Madry and Fromm reported the results of 40 patients operated for infected necrosis4. Patients were described as having necrotic retroperitoneal fat without overt pancreatic necrosis. Sakorafas and colleagues were the first to report a better outcome in patients with EXPN who underwent surgical debridement5. Operative and CT findings of 12 patients operated for necrotizing pancreatitis between 1983 and 1997 suggested that necrotizing pancreatitis did not always involve the pancreatic parenchyma. In contrast to the current study, this retrospective study only reported the outcome of patients who underwent surgical debridement for necrotizing pancreatitis and did not provide the clinical outcome of the total group of patients (with or without organ failure, sterile or infected, conservative treatment or after intervention). In their study, including 12 patients with EXPN, mortality was 8% which is comparable to the 9% mortality rate found in 315 patients with EXPN in this study. In both studies, the mortality rate in patients with pancreatic necrosis was 20%. Accurate comparison of mortality rates between the two studies, however, is hampered due to differences in patient characteristics, CECT assessment of EXPN, and study design (i.e. experienced single center study versus a multicenter study with academic and non-academic hospitals). In the present study, EXPN was diagnosed if CECT showed extrapancreatic fat with morphologic changes that exceeded fat stranding and if the pancreatic parenchyma did not show any signs of necrosis as determined on the last CECT of hospitalization or before any intervention. We cannot exclude, however, that small focal areas of parenchymal necrosis were overlooked on CT in patients with EXPN. Inversely, extrapancreatic morphologic changes could also consist of fluid only instead of fat necrosis. Previous studies have indeed questioned the accuracy of CECT for diagnosing EXPN10,11. 49


PART I CHAPTER 3

Small extrapancreatic collections, for example smaller than 2 cm in diameter, most likely contain fluid which can easily be absorbed. Large extrapancreatic effusions, for example more than 5 cm in diameter, are less easily absorbed12. Collections that do not dissolve early in the course of the disease, will most likely contain some degree of peripancreatic tissue necrosis and, hence, have a greater chance of being detected. If extrapancreatic changes have been overestimated and incorrectly scored as EXPN (while those patients actually had interstitial pancreatitis without necrosis), this could potentially question the validity of the differences found in this study. It is known that patients with interstitial pancreatitis have a better prognosis than patients with EXPN. A recent study from Boston compared the outcome of patients with interstitial pancreatitis with 8 patients with EXPN13. Patients with interstitial pancreatitis had a better outcome in terms of organ failure, length of hospital stay, need for interventions and mortality. This study, though, did not include patients with pancreatic necrosis. However, we have several reasons to believe that the presence of EXPN has not been overestimated in this study. First, several studies have actually demonstrated a good accuracy of CECT for EXPN when CECT findings were compared with the presence of fat necrosis at operation or autopsy14-16. Second, in patients with EXPN the median time between onset of symptoms and the last CECT before discharge was 9 days. Extrapancreatic collections that persist throughout the second week after onset of symptoms will in the majority of cases, contain some amount of fat necrosis17,18. Third, the APACHE-II scores on admission and CRP levels after 48 hours did not differ between patients with EXPN and patients with pancreatic necrosis. If patients with interstitial disease (with extrapancreatic fluid instead of extrapancreatic necrosis) would have been incorrectly scored as EXPN and included in the study, one could speculate that the median severity scores on admission in the EXPN group would have been lower than the median severity scores of patients with pancreatic necrosis. However, as patients with interstitial disease (based on CECT) were not included in the study, we did not investigate whether the severity scores in this group of patients is actually lower compared with patients with EXPN. In a previous multicenter study, we found a mortality of 0.8% in patients with interstitial pancreatitis (mortality in EXPN in this study was 9%).19 Although the Imrie scores did differ significantly after 48 hours, these scores reflect the ongoing inflammatory response and this is likely to be more severe in patients with parenchymal necrosis. The same difference in ongoing inflammatory response between groups was seen in patients with organ failure during the first week. The pathophysiologic explanation for the better outcome of patients with EXPN remains speculative. It is generally believed that trypsin activation within pancreatic acinar cells leads to autodigestion and local inflammation20,21. Following a cascade of intra-cellular events, the pancreatic acinar cells may become necrotic. In some patients, extensive local 50


EXPN: A SEPARATE ENTITY IN NECROTIZING PANCREATITIS?

inflammation causes a severe systemic inflammatory response syndrome that may lead to organ failure22,23. In EXPN, necrosis of the acinar cells does not seem to occur and the pancreatic parenchyma is preserved (at least on CECT). This might be indicative of a less severe local inflammatory response. Subsequently, a less severe local inflammatory response could explain the less severe systemic inflammatory response in patients with EXPN. Another potential explanation may be that in EXPN the amount of released inflammatory mediators or cytokines is lower. The concentration of inflammatory mediators might be higher in the pancreatic acinar cells than in the extrapancreatic fatty tissue. Parenchymal necrosis causes these mediators to be released into the systemic circulation. Conceivably, the higher the amount of mediators released, the more severe the systemic inflammatory response will be24. Lastly, it might be possible that ductal disruption, which may occur with parenchymal necrosis, causes more complications. Ductal disruption could facilitate bacterial invasion of pancreatic tissue. In this study, we found that patients with pancreatic necrosis significantly more often develop infection (47% versus 16%). In patients with necrotizing pancreatitis, pancreatic exocrine and endocrine insufficiencies, as well as pancreatic fistulas following interventions, are common long term complications8,25,26. In theory, EXPN will not cause these complications. None of the patients with EXPN in the study from Sakorafas developed endocrine or exocrine insufficiency during long term follow-up. In a previous study, pancreatic fistulas were shown to cause considerable morbidity.27 For this cohort, however, our prospective database did not include data on exocrine and endocrine insufficiency or pancreatic fistulas. This clinical study does not necessarily prove, in terms of biology, that EXPN is a separate entity. Based on the results of this study, however, EXPN should be recognized as a different clinical entity within necrotizing pancreatitis (compared with parenchymal necrosis) because it heralds several significant clinical implications. First, in case of EXPN as determined on CECT, during necrosectomy (e.g. in case of infected necrosis) surgeons should refrain from debridement of the pancreas to prevent iatrogenic pancreatic injury. This knowledge is crucial as during surgery the differentiation between necrotic pancreatic parenchyma and peripancreatic fat necrosis can be difficult. Second, to accurately compare reports of clinical studies on new interventions in necrotizing pancreatitis, future studies should report the number of patients with EXPN. Third, the presence of EXPN is a favourable prognosticator compared with pancreatic necrosis. The presence of EXPN should not change overall management but does provide the clinician with information for risk stratification. Fourth, our finding that infected EXPN has a similar poor outcome as infected pancreatic necrosis, underlines the importance of adequate monitoring and timely recognition and treatment of sepsis in all patients with necrotizing pancreatitis. 51


PART I CHAPTER 3

In summary, this study provides necessary data on the outcome of patients with EXPN. The prognosis of patients with non-infected EXPN has been shown to be significantly better than of patients with non-infected pancreatic necrosis and could be considered a different clinical entity in acute pancreatitis. In case of infection, however, prognosis is similar among patients with EXPN and parenchymal necrosis.

52


EXPN: A SEPARATE ENTITY IN NECROTIZING PANCREATITIS?

REFERENCES 1. Bradley EL, III. A clinically based classification system for acute pancreatitis. Summary of the International Symposium on Acute Pancreatitis, Atlanta, Ga, September 11 through 13, 1992. Arch Surg 1993; 128(5):586-590. 2. Howard JM, Wagner SM. Pancreatography after recovery from massive pancreatic necrosis. Ann Surg 1989; 209(1):31-35. 3. Howard JM. Delayed debridement and external drainage of massive pancreatic or peripancreatic necrosis. Surg Gynecol Obstet 1989; 168(1):25-29. 4. Madry S, Fromm D. Infected retroperitoneal fat necrosis associated with acute pancreatitis. J Am Coll Surg 1994; 178(3):277-282. 5. Sakorafas GH, Tsiotos GG, Sarr MG. Extrapancreatic necrotizing pancreatitis with viable pancreas: a previously underappreciated entity. J Am Coll Surg 1999; 188(6):643-648.

10. Balthazar EJ. Acute pancreatitis: assessment of severity with clinical and CT evaluation. Radiology 2002; 223(3):603-613. 11. Merkle EM, Gorich J. Imaging of acute pancreatitis. Eur Radiol 2002; 12(8):1979-1992. 12. Kloppel G. Acute pancreatitis. Semin Diagn Pathol 2004; 21(4):221-226. 13. Singh VK, Bollen TL, Wu BU et al. An assessment of the severity of interstitial pancreatitis. Clin Gastroenterol Hepatol 2011; 9(12):1098-1103. 14. Gambiez LP, Denimal FA, Porte HL, Saudemont A, Chambon JP, Quandalle PA. Retroperitoneal approach and endoscopic management of peripancreatic necrosis collections. Arch Surg 1998; 133(1):66-72. 15. Larvin M, Chalmers AG, McMahon MJ. Dynamic contrast enhanced computed tomography: a precise technique for identifying and localising pancreatic necrosis. BMJ 1990; 300(6737):1425-1428.

6. van Santvoort HC, Bakker OJ, Bollen TL et al. A conservative and minimally invasive approach to necrotizing pancreatitis improves outcome. Gastroenterology 2011; 141(4):1254-1263.

16. Vege SS, Fletcher JG, Talukdar R, Sarr MG. Peripancreatic collections in acute pancreatitis: correlation between computerized tomography and operative findings. World J Gastroenterol 2010; 16(34):4291-4296.

7. Besselink MG, van Santvoort HC, Buskens E et al. Probiotic prophylaxis in predicted severe acute pancreatitis: a randomised, double-blind, placebo-controlled trial. Lancet 2008; 371(9613):651-659.

17. Bollen TL, Singh VK, Maurer R et al. A comparative evaluation of radiologic and clinical scoring systems in the early prediction of severity in acute pancreatitis. Am J Gastroenterol 2012; 107(4):612-619.

8. van Santvoort HC, Besselink MG, Bakker OJ et al. A step-up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med 2010; 362(16):1491-1502.

18. Bradley EL, III. A fifteen year experience with open drainage for infected pancreatic necrosis. Surg Gynecol Obstet 1993; 177(3):215-222.

9. von EE, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007; 370(9596):1453-1457.

19. Besselink MG, van Santvoort HC, Boermeester MA et al. Timing and impact of infections in acute pancreatitis. Br J Surg 2009; 96(3):267-273. 20. Halangk W, Lerch MM. Early events in acute pancreatitis. Gastroenterol Clin North Am 2004; 33(4):717-731.

53


PART I CHAPTER 3

21. Pandol SJ, Saluja AK, Imrie CW, Banks PA. Acute pancreatitis: bench to the bedside. Gastroenterology 2007; 132(3):1127-1151. 22. Frossard JL, Steer ML, Pastor CM. Acute pancreatitis. Lancet 2008; 371(9607):143-152. 23. Whitcomb DC. Clinical practice. Acute pancreatitis. N Engl J Med 2006; 354(20):2142-2150. 24. Makhija R, Kingsnorth AN. Cytokine storm in acute pancreatitis. J Hepatobiliary Pancreat Surg 2002; 9(4):401-410. 25. Connor S, Alexakis N, Raraty MG et al. Early and late complications after pancreatic necrosectomy. Surgery 2005; 137(5):499-505. 26. Tsiotos GG, Luque-de LE, Sarr MG. Longterm outcome of necrotizing pancreatitis treated by necrosectomy. Br J Surg 1998; 85(12):1650-1653. 27. Bakker OJ, van Baal MC, van Santvoort HC et al. Endoscopic transpapillary stenting or conservative treatment for pancreatic fistulas in necrotizing pancreatitis: multicenter series and literature review. Ann Surg 2011; 253(5):961-967.

54


55


PART I EARLY IDENTIFICATION OF COMPLICATED PANCREATITIS


CHAPTER 4 The impact of organ failure on mortality in necrotizing pancreatitis Submitted

Olaf J Bakker, Nicolien J Schepers, Marc G Besselink, Usama Ahmed Ali, Thomas L Bollen, Hein G Gooszen, Hjalmar C van Santvoort and Marco J Bruno for the Dutch Pancreatitis Study Group


PART I CHAPTER 4

A B S T R AC T B ackg round & Aims Mortality in necrotizing pancreatitis is primarily caused by organ failure. This study aims to specify the impact of the characteristics of organ failure on mortality and the association with infected pancreatic necrosis. Metho ds We performed a post-hoc analysis of a prospective database of 639 patients with necrotizing pancreatitis from 21 hospitals. During the entire hospital stay, data on organ failure (i.e. respiratory, cardiovascular and renal failure) were registered daily. Resu lts In total, 240 of 639 patients with necrotizing pancreatitis (38%) developed organ failure. Overall, the longest episode of organ failure (any type or combination) lasted a median of 9 days (interquartile range 3 to 17) and started in the first week in 47% of patients (41% mortality), in 14% during the second week (43% mortality), and in 39% after the second week (32% mortality; P = .51). Mortality of persistent multiple organ failure lasting less than 1 week, 1 to 2 weeks, 2 to 3 weeks or longer than 3 weeks was 43%, 38%, 46% and 52%, respectively (P = .68). Simultaneous persistent failure of all 3 organ systems occurred in 52 patients with 63% mortality. In patients with infected pancreatic necrosis and organ failure compared with organ failure alone, we found no difference in mortality after 10 days of admission (28% vs. 30%, P = .4). Overall, patients with organ failure alone had a higher risk for mortality than patients with organ failure and infected pancreatic necrosis (44% vs. 29%, P < .001). C onclusi ons Although mortality is high in patients with necrotizing pancreatitis and persisting organ failure, half of patients may survive after several weeks of organ failure. Infected pancreatic necrosis does not increase mortality when compared with organ failure without infected pancreatic necrosis.

58


ORGAN FAILURE IN NECROTIZING PANCREATITIS

BAC KG ROU N D & A I M S Organ failure is the most important determinant for severity in acute pancreatitis and is associated with a mortality rate of around 30%.1-4 Approximately 45% of patients with necrotizing pancreatitis develop organ failure, suffer from extended intensive care unit (ICU) admissions and most often need to undergo multiple invasive interventions.5, 6 Next to high complication and mortality rates, organ failure in necrotizing pancreatitis leads to considerable healthcare utilization.7 Early organ failure is caused by a systemic inflammatory response syndrome (SIRS) which is mediated by the release of inflammatory chemokines and cytokines. Transient organ failure, i.e. lasting less than 48 hours, is associated with low mortality and low risk of complications.8 Persisting organ failure, i.e. lasting more than 48 hours, is associated with increased complication and mortality rates.9, 10 Organ failure may also develop at a later stage in the disease. Infected pancreatic necrosis may develop several weeks after disease onset and lead to sepsis and concomitant organ failure. However, infected pancreatic necrosis or other infections such as pneumonia may also occur early in the disease course and therefore it can be difficult to determine whether organ failure is caused by SIRS or sepsis.2, 11 Three organ systems are important in the assessment of organ failure in acute pancreatitis â&#x20AC;&#x201C; the cardiovascular, respiratory and renal system.2 Although, numerous studies have reported the general impact of organ failure, the impact of the specific characteristics of organ failure on mortality (such as timing and combination of failing organ systems), are unknown due to the relative low incidence of organ failure and lack of robust, prospective, detailed, homogeneous data. Previous studies have several limitations. First, data are limited to the first week of the disease.8-10, 12, 13 Second, no detailed daily assessment of organ function was performed, for example, only data on admission or only daily assessment during the first week was reported.9, 14-22 Third, various definitions for organ failure were used.12, 16, 18, 23 Fourth, studies were performed in a single center,10, 12-20, 22, 24 included less than 100 patients with organ failure,10, 12, 13, 15-22 and were performed in tertiary referral centers with more than 40% of patients transferred from other hospitals during different phases of the disease.19, 21 Finally, when comparing the impact of organ failure with the impact of infected pancreatic necrosis on mortality, the timing of organ failure in relation to infected pancreatic necrosis is critical. In previous studies, the exact timing of organ failure is not reported. These limitations seriously hamper the comparativeness and generalisability of the reported outcomes after organ failure and, prohibit the pooling of studies in meta-analyses. Improved knowledge about the details of organ failure in acute pancreatitis will help to evaluate the prognosis of patients who suffer from organ failure, especially when endof-life decisions need to be discussed. Furthermore, insight into the details of organ failure 59


PART I CHAPTER 4

may provide a base for new treatment strategies in severe acute pancreatitis. The aim of this study is to specify the impact of the different characteristics of organ failure on mortality and the association with infected pancreatic necrosis.

PAT I E N T S & M E T HOD S Pati ents & Stu dy D esig n The Dutch Pancreatitis Study Group performed a prospective multicenter cohort study in 21 Dutch hospitals of consecutive patients with necrotizing pancreatitis. The protocol of this study was described previously.6, 25 The present study is a detailed analysis of all patients with organ failure in this prospective database of 639 patients with necrotizing pancreatitis. In this study, we investigated type, onset and duration of organ failure in necrotizing pancreatitis. The association between these characteristics of organ failure and mortality was studied. Finally, the association between organ failure and infected pancreatic necrosis was analyzed. D ata C ol l e c ti on During the entire hospital stay, all data regarding organ failure were registered daily including type (i.e. respiratory, cardiovascular and renal failure), duration and sequence of organ failure. Furthermore, the following characteristics were recorded: age, sex, prognostic scores (APACHE-II, Modified Glasgow score, C-reactive protein) at hospital admission, computed tomography (CT) severity index, total hospital and ICU stay, and interventions. The presence of extrapancreatic necrosis and extent of parenchymal necrosis was based on the final CT of hospitalisation or before any intervention. D ef initi ons Organ failure was defined as failure of the respiratory, cardiovascular or renal system (for definitions see Box 1). Persistent organ failure was defined as organ failure that occurred on three consecutive days. Multiple organ failure was defined as organ failure of two or more organ systems. If patients suffered from intermittent episodes of organ failure, the longest continuous episode of failure of either the respiratory, cardiovascular or renal system was used to study the relation with mortality. Based on the duration of the longest episode, patients were defined as having persistent or transient organ failure. Subsequently, a distinction was made between single of multiple organ failure. For example, a patient who suffered from persistent single organ failure but only transient multiple organ failure, was included in the group of patients with persistent single organ failure. A patient who suffered from persistent single organ failure and persistent multiple organ failure was only included in the group of patients with persistent multiple organ failure. Patients with 60


ORGAN FAILURE IN NECROTIZING PANCREATITIS

Box 1. Definitions Organ failure Pulmonary

PaO2 <60 mmHg, despite FIO2 of 0.30, or need for mechanical ventilation

Circulatory

Circulatory systolic blood pressure <90 mmHg, despite adequate fluid resuscitation, or need for inotropic catecholamine support

Renal

Creatinin level >177 µmol/L after rehydration or need for hemofiltration or hemodialysis

Single organ failure

Failure of one organ system on the same day

Multiple organ failure

Failure of at least 2 organ systems on the same day

Transient organ failure

Organ failure that resolves within 48 hours

Persistent organ failure

Organ failure that lasts longer than 48 hours (occurs on 3 consecutive days)

Necrotizing pancreatitis

Pancreatic parenchymal necrosis with or without extrapancreatic necrosis and extrapancreatic necrosis alone

• Pancreatic necrosis

Focal non-enhancement of the pancreatic gland (CT severity index of >4)

• Extrapancreatic necrosis alone

Peripancreatic morphologic changes exceeding fat stranding without pancreatic parenchymal necrosis (CT severity index of 3 or 4)

Infected pancreatic necrosis

Positive culture of pancreatic or extrapancreatic necrotic tissue obtained by fine-needle aspiration or from the first intervention, or the presence of gas in the fluid collection on contrast-enhanced CT

fulminant pancreatitis who died within 48 hours after onset of organ failure and hence were not able to develop organ failure lasting for more than 48 hours (i.e. persistent organ failure), were included in the group of patients with persistent organ failure because of the obvious presence of severe pancreatitis. The prevalence of organ failure demonstrates the total number of organs failing per system per day divided by the number of patients alive; the incidence represents the number of new organs failing per system per day divided by the number of patients alive. Necrotizing pancreatitis was defined as either pancreatic parenchymal necrosis with or without extrapancreatic necrosis, or extrapancreatic necrosis alone.2 Statisti c a l Ana lysis Normal data are presented as mean ± standard deviation (SD) or as median with inter quartile ranges (IQR). Differences in continuous and ordinal variables were tested with the Mann-Whitney U test. Proportions were compared by using the Χ2 test. The association between organ failure, infected pancreatic necrosis and mortality was studied. A multivariable logistic regression analysis was performed to adjust for baseline variables with a P-value of < .2. 61


PART I CHAPTER 4

R E SU LT S Organ failure occurred in 240 of 639 patients (38%) with necrotizing pancreatitis. Table 1 presents the characteristics of patient with organ failure compared with patients without organ failure. Compared with patients without organ failure, patients with organ failure had a higher percentage of pancreatic parenchymal necrosis, a prolonged hospital and ICU stay, more often developed infected pancreatic necrosis and intervention was needed more often. Ty p e and C ombinati on of Organ Fai lure and Imp ac t on Mor ta lity Figure 1a and 1b represent the prevalence and incidence of respiratory, cardiovascular and renal failure during hospital admission in 240 patients with organ failure. The prevalence and incidence of organ failures peaked during the first week. In total, 221 patients (92%) experienced respiratory failure with 38% mortality, followed by cardiovascular failure in 197 patients (82%) with 41% mortality, and renal failure in 106 patients (44%) with 49% mortality. Table 2 presents mortality rates in different subgroups of patients according to type and combinations of organ failure. Transient organ failure occurred in only 21 of 240 patients with organ failure (9%) and was associated with 10% mortality. In 219 patients with persistent organ failure (91%), mortality was 38%. Of these 219 patients, 166 patients (76%) had multiple organ failure with 43% mortality. Failure of the respiratory and cardiovascular system occurred most often (88 patients), followed by failure of the respiratory and renal system (23 patients) with similar mortality rates of 35%. In total, 52 patients had simultaneous failure of all three systems with 63% mortality. Timing of Organ Failure and Impact on Mortality Figure 2 shows the timing of the longest persistent episode, categorized per organ system. The respiratory system was the first to fail and the longest to persist. The longest episode of persistent organ failure (any type) lasted a median of 9 days (interquartile range 3â&#x20AC;&#x201C;17 days). Overall, a median of 2 episodes of organ failures (interquartile range 2 to 3, range 1 to 12) occurred. Table 3 describes the association between the onset and duration of organ failure and mortality. The longest episode of persistent organ failure (any type or combination) started in the first week in 47% of patients with 41% mortality, in 14% during the second week with 43% mortality, and in 39% after the second week with 32% mortality (P = .51). Mortality in persistent multiple organ failure lasting less than 1 week, 1 to 2 weeks, 2 to 3 weeks or longer than 3 weeks was 43%, 38%, 46% and 52% respectively (P = .68). In other words, even after more than 3 weeks of organ failure, 48% of patients survived.

62


ORGAN FAILURE IN NECROTIZING PANCREATITIS

Table 1. Characteristics of patients with organ failure compared with patients without organ failure. Characteristic

All patients Organ failure No organ failure (n=639) (n=240; 38%) (n=399; 62%)

Age – year

58 (45-70)

61 (50-71)

56 (43-70)

.003A

398 (62)

158 (66)

240 (60)

.15B

8 (5-11)

9 (6-13)

7 (4-10)

< .001A

3 (2-5)

4 (3-5)

3 (2-4)

< .001A

291 (±131)

284 (±134)

295 (±127)

.24A

4 (4-8)

6 (4-10)

4 (4-6)

< .001A

324 (51)

163 (68)

161 (40)

< .001B

Male sex – no. (%)

P value

Predicted severity • APACHE II score* • Modified Glasgow score* • Highest CRP 48 hrs after admission

#

CT severity index Pancreatic parenchymal necrosis – no. (%) Extent of pancreatic necrosis • <50%

530 (83)

116 (71)

360 (90)

• >50%

109 (17)

70 (29)

39 (10)

Extrapancreatic necrosis alone – no. (%)

315 (49)

77 (32)

238 (60)

< .001 B

Hospital stay in days (interquartile range)

28 (14-66)

64 (31-106)

19 (12-38)

< .001A

ICU stays in days (interquartile range)

3 (0-22)

22 (9-42)

0 (0-0)

< .001A

Infected pancreatic necrosis

202 (32)

132 (55)

70 (18)

< .001B

177 (74)

65 (16)

< .001B

Any intervention (e.g. drainage, necrosectomy)

242 (38)

* <1% missing data, # 7% missing data. Mann-Whitney test, χ2 test, Linear-by-linear association test A

B

C

Table 2. Mortality in different subgroups of 240 patients with organ failure. Subgroups

Mortality in transient organ failure

Mortality in persistent organ failure

2 / 15 (13)

11 / 53 (21)

Single Organ Failure Any Organ System Cardiovascular

0 / 6 (0)

1 / 3 (33)

Respiratory

0 / 3 (0)

9 / 45 (20)

Renal

2 / 6 (30)

1 / 5 (20)

Any 2 or more Organ Systems

0 / 6 (0)

72 / 166 (43)

Cardiovascular and Respiratory

Multiple Organ Failure 0 / 6 (0)

31 / 88 (35)

Respiratory and Renal

-

8 / 23 (35)

Renal and Cardiovascular

-

0 / 3 (0)

All 3 Organ Systems

-

33 / 52 (63)

Note; All values in parentheses are percentages. Data represent the longest persistent episode of organ failure per system.

63


PART I CHAPTER 4

Figure 1. Prevalence of respiratory, cardiovascular and renal failure in 240 patients with organ failure Note; Graph presents the daily number of patients with organ failure per system divided by the patients alive.

Figure 1b. Incidence of pulmonary, cardiovascular and renal failure 240 patients with organ failure Note; Graph presents the daily number of patients with new onset organ failure per system divided by the patients alive.

64


ORGAN FAILURE IN NECROTIZING PANCREATITIS

Figure 2. Timing of failure per organ system in 240 patients Note; Data represent the longest persistent episode of failure per organ system. The solid horizontal lines represent the interquartile range of the start (upper line) and end day (lower line) of the longest persistent episode of organ failure. The vertical solid line represents the median day the organ failure started and ended. The shaded grey area is constructed between the median start and end day of organ failure and can be interpreted as the median timing of organ failure per system.

Subgroup Infected Pancreatic Necrosis and Impact on Mortality Infected pancreatic necrosis was present in 132 of 240 patients with organ failure (55%). Of these 132 patients, 50 patients (38%) experienced organ failure in the first week compared with 82 patients (62%) who developed organ failure after the first week. Of all 639 patients with necrotizing pancreatitis, 70 patients had infected pancreatic necrosis without organ failure (11%) with 4% mortality. In 132 patients with organ failure and infected pancreatic necrosis, the overall mortality was 29%. Surprisingly, patients with organ failure but without infected pancreatic necrosis had a higher risk for mortality (44% of 108 patients) than patients with organ failure and infected pancreatic necrosis (29% of 132 patients; relative risk 2.6; 95% confidence interval 1.7-4.1; P < .001). However, when only mortality occurring at least 10 days after admission was included, mortality in patients with infected pancreatic necrosis and organ failure was comparable to mortality in patients with organ failure without infected pancreatic necrosis (28% vs. 30%, respectively, relative risk 1,3; P= .4). These results did not change after adjusting for potential confounding factors such as age or comorbidity (data not given).

65


PART I CHAPTER 4

Table 3. Mortality related to onset and duration of persistent organ failure Mortality (%) Onset Organ Failure First week Second week

Single Organ Failure

Multiple Organ Failure

7 / 30 (23)

36 / 74 (49)

0 / 4 (0)

13 / 26 (50) 2 / 15 (13)

Third week

1 / 4 (25)

Fourth week

1 / 4 (25)

9 / 16 (56)

Fifth week and later

2 / 11 (18)

12 / 35 (34)

48 hours to one week

4 / 22 (18)

34 / 80 (43)

Duration organ failure One to two weeks

2 / 11 (18)

15 / 40 (38)

Two to three weeks

1 / 12 (8)

6 / 13 (46)

More than three weeks

4 / 8 (50)

17 / 33 (52)

Note; data represent the longest persistent episode of organ failure (any organ system).

DI S C U S SION This is the first study to provide detailed insight into the association between type and timing of organ failure and mortality in patients with necrotizing pancreatitis. In contrast to previous studies, we found no significant association between the onset of organ failure and mortality. Also, no significant association was found between the duration of organ failure and mortality. Finally, we found a higher mortality in patients with organ failure without infected pancreatic necrosis compared with patients with organ failure and infected pancreatic necrosis. This difference disappeared when only patients who died more than 10 days after admission were accounted for. These results are in contrast to previous studies and require careful interpretation. First of all, a previous meta-analysis suggested that the influence of organ failure and infection of pancreatic necrosis on mortality was similar and that presence of both increases mortality even further.4 This discrepancy may be explained by the following differences between study design and data collection. First, in our study all patients with necrotizing pancreatitis were prospectively included and data collection started on the day of admission. This complete registration of patients also includes patients who succumb very early in the disease course as a result of fulminant early organ failure. Sensitivity analyses showed that when only mortality occurring after 10 days of admission was considered, this difference in mortality between patients with and without infection of pancreatic necrosis was no longer present. Second, in the meta-analysis patient series from different countries during different time periods are included and all results are pooled. As registration of organ failure requires prolonged, extensive and accurate data 66


ORGAN FAILURE IN NECROTIZING PANCREATITIS

collection including start and finish, type and sequence, one may question the accuracy of the data when organ failure is only registered as a binominal variable, meaning organ failure was present of absent. For these reasons, we believe this is not only the largest but also the most accurate study to report a prospective multicenter cohort of patients with organ failure using the same definition and accurate data collection for all participating centers. At present, there are two classification systems defining the severity of pancreatitis.2, 3 The Revised Atlanta classification distinguishes mild, moderate and severe pancreatitis according to the presence of transient or persistent organ failure. The determinant-based classification contains an additional, category of patients with infected necrosis and organ failure, named critical pancreatitis. The clinical relevance for this category is based on studies suggesting that patients with infected pancreatic necrosis and organ failure are at excessive risk for mortality. However the included studies had serious limitations as mentioned previously. We found that mortality rates between the “severe” and “critical” group were similar when considering mortality beyond the early phase, and even lower in patients with organ failure and infected pancreatic necrosis when the entire disease course was taken into account. Therefore, our data do not support the relevance of a critical category next to mild, moderate and severe pancreatitis. Acute severe pancreatitis is thought to run a biphasic course.2, 26, 27 The first two weeks are characterized by the systemic inflammatory response syndrome (SIRS) that may cause early organ failure. This early phase is hypothesized to be followed by a state of compensatory anti-immune response syndrome in which patients are susceptible for infections.28, 29 From this hypothesis, mortality rates in severe pancreatitis are believed to have two peaks, i.e. following “early” organ failure and following “late” organ failure and subsequent sepsis. Surprisingly, our data do not support the presence of a biphasic model. No increase in the onset of organ failure, nor a second peak in mortality, was observed after the early phase of the disease course. The biphasic model might be predominantly based on experimental, immunological data that are not confirmed by parameters such as organ failure in clinical studies. Another possible explanation for these results is that improved intensive care unit treatment has influenced outcome and caused the second peak in mortality to disappear. This has been observed in the field of trauma surgery.30 A potential limitation of our study is that the underlying cause of organ failure is not specified. For example, when organ failure occurs after open necrosectomy for infected pancreatic necrosis, the cause may be obvious and is a result of abdominal sepsis. However, to determine whether organ failure is caused by severe SIRS or a potential major infection is more complex. The exact timing of infection of pancreatic necrosis or other infections such as pneumonia are unclear in most cases, especially when persisting organ failure is already present. Organ failure caused by SIRS without documented infection and organ failure caused by sepsis may overlap in time and are difficult to 67


PART I CHAPTER 4

distinguish.11 Another limitation is that no patients with organ failure but without necrotizing pancreatitis are included. As organ failure in interstitial pancreatitis is extremely rare, we expect that this limitation is of minor importance. In addition, patients that died before informed consent was given are not included. However, including these patients would likely only further strengthen the important role of organ failure by increasing mortality rates in patients with organ failure alone compared with patients with organ failure and infected pancreatic necrosis. Finally, the definitions of organ failure slightly differ from the definitions proposed in the revised Atlanta classification, which uses the modified Marshall scoring system.2 Although, proposed as a relative simple and easy to use scoring system, the modified Marshall score has two major disadvantages. First, patients with persisting organ failure are sometimes admitted for over 6 months in the hospital. It is cumbersome to calculate the modified Marshall score for 3 separate organ systems on a daily basis throughout the whole disease course. Especially, when retrospective cohorts describe the incidence of organ failure with the modified Marshall system one might question the accuracy of the reported organ failure rates. Second, endotracheal intubation for respiratory failure, the use of vasopressors for persistent hypotension after fluid resuscitation and renal replacement therapy for renal insufficiency are not incorporated in the definitions of the revised Atlanta classification. If patients receive these supportive therapies, this may result in normal arterial oxygen measurements and normal systolic blood pressures. Hence, the Marshall score will not reach the threshold for organ failure and these patients would be categorised as having moderate pancreatitis. The definitions used in our study are more stringent than the modified Marshall scoring system. This may also be the reason why transient organ failure was only observed in 21 of 639 patients (3%) with necrotizing pancreatitis. In conclusion, this study highlights the importance of organ failure in patients with necrotizing pancreatitis and challenges the concept that infected pancreatic necrosis has an equal impact on outcome compared with organ failure. We believe our findings of similar survival rates after several weeks of organ failure compared with organ failure of shorter duration, is a strong plea for intensive and prolonged treatment of patients with necrotizing pancreatitis.

68


ORGAN FAILURE IN NECROTIZING PANCREATITIS

REFERENCES 1. Guidelines for intensive care unit admission, discharge, and triage. Task Force of the American College of Critical Care Medicine, Society of Critical Care Medicine. Crit Care Med 1999;27:633-8. 2. Banks PA, Bollen TL, Dervenis C, et al. Classification of acute pancreatitis--2012: revision of the Atlanta classification and definitions by international consensus. Gut 2013;62:102-11. 3. Dellinger EP, Forsmark CE, Layer P, et al. Determinant-based classification of acute pancreatitis severity: an international multidisciplinary consultation. Ann Surg 2012;256:875-80. 4. Petrov MS, Shanbhag S, Chakraborty M, et al. Organ failure and infection of pancreatic necrosis as determinants of mortality in patients with acute pancreatitis. Gastroenterology 2010;139:813-20. 5. Banks PA, Freeman ML. Practice guidelines in acute pancreatitis. Am J Gastroenterol 2006;101:2379-400. 6. van Santvoort HC, Bakker OJ, Bollen TL, et al. A conservative and minimally invasive approach to necrotizing pancreatitis improves outcome. Gastroenterology 2011;141:1254-63. 7. Fagenholz PJ, Fernandez-del Castillo C, Harris NS, et al. Direct medical costs of acute pancreatitis hospitalizations in the United States. Pancreas 2007;35:302-7. 8. Johnson CD, Abu-Hilal M. Persistent organ failure during the first week as a marker of fatal outcome in acute pancreatitis. Gut 2004;53:1340-4. 9. Mofidi R, Duff MD, Wigmore SJ, et al. Association between early systemic inflammatory response, severity of multiorgan dysfunction and death in acute pancreatitis. Br J Surg 2006;93:738-44. 10. Thandassery RB, Yadav TD, Dutta U, et al. Dynamic nature of organ failure in severe acute pancreatitis: the impact of persistent

and deteriorating organ failure. HPB (Oxford) 2013;15:523-8. 11. Besselink MG, van Santvoort HC, Boermeester MA, et al. Timing and impact of infections in acute pancreatitis. Br J Surg 2009;96:267-73. 12. Buter A, Imrie CW, Carter CR, et al. Dynamic nature of early organ dysfunction determines outcome in acute pancreatitis. Br J Surg 2002;89:298-302. 13. Remes-Troche JM, Uscanga LF, Pelaez-Luna M, et al. When should we be concerned about pancreatic necrosis? Analysis from a single institution in Mexico City. World J Surg 2006;30:2227-33; discussion 2234-5. 14. Halonen KI, Pettila V, Leppaniemi AK, et al. Multiple organ dysfunction associated with severe acute pancreatitis. Crit Care Med 2002;30:1274-9. 15. Isenmann R, Rau B, Beger HG. Early severe acute pancreatitis: characteristics of a new subgroup. Pancreas 2001;22:274-8. 16. Le Mee J, Paye F, Sauvanet A, et al. Incidence and reversibility of organ failure in the course of sterile or infected necrotizing pancreatitis. Arch Surg 2001;136:1386-90. 17. Li XY, Wang XB, Liu XF, et al. Prevalence and risk factors of organ failure in patients with severe acute pancreatitis. World J Emerg Med 2010;1:201-4. 18. Lytras D, Manes K, Triantopoulou C, et al. Persistent early organ failure: defining the high-risk group of patients with severe acute pancreatitis? Pancreas 2008;36:249-54. 19. Perez A, Whang EE, Brooks DC, et al. Is severity of necrotizing pancreatitis increased in extended necrosis and infected necrosis? Pancreas 2002;25:229-33. 20. Sharma M, Banerjee D, Garg PK. Characterization of newer subgroups of fulminant and subfulminant pancreatitis associated with a high early mortality. Am J Gastroenterol 2007;102:2688-95.

69


PART I CHAPTER 4

21. Wig JD, Bharathy KG, Kochhar R, et al. Correlates of organ failure in severe acute pancreatitis. JOP 2009;10:271-5. 22. Zhu AJ, Shi JS, Sun XJ. Organ failure associated with severe acute pancreatitis. World J Gastroenterol 2003;9:2570-3. 23. Tenner S, Sica G, Hughes M, et al. Relationship of necrosis to organ failure in severe acute pancreatitis. Gastroenterology 1997;113:899-903. 24. Guo Q, Li A, Xia Q, et al. The role of organ failure and infection in necrotizing pancreatitis: a prospective study. Ann Surg 2014;259:1201-7. 25. Bakker OJ, van Santvoort H, Besselink MG, et al. Extrapancreatic necrosis without pancreatic parenchymal necrosis: a separate entity in necrotising pancreatitis? Gut 2013;62:1475-80. 26. Bone RC. Immunologic dissonance: a continuing evolution in our understanding of the systemic inflammatory response syndrome (SIRS) and the multiple organ dysfunction syndrome (MODS). Ann Intern Med 1996;125:680-7. 27. Moore FA, Sauaia A, Moore EE, et al. Postinjury multiple organ failure: a bimodal phenomenon. J Trauma 1996;40:501-10; discussion 510-2. 28. Cobb JP, O'Keefe GE. Injury research in the genomic era. Lancet 2004;363:2076-83. 29. Mayerle J, Dummer A, Sendler M, et al. Differential roles of inflammatory cells in pancreatitis. J Gastroenterol Hepatol 2012;27 Suppl 2:47-51. 30. Minei JP, Cuschieri J, Sperry J, et al. The changing pattern and implications of multiple organ failure after blunt injury with hemorrhagic shock. Crit Care Med 2012;40:1129-35.

70


71


PART II PREVENTING COMPLICATED PANCREATITIS

CHAPTER 5 Timing of enteral nutrition in acute pancreatitis: Meta-analysis of individuals using a single-arm of randomized trials Pancreatology 2014 CHAPTER 6 Early versus on-demand nasoenteric tube feeding in acute pancreatitis New England Journal of Medicine 2014 CHAPTER 7 Timing of cholecystectomy after mild biliary pancreatitis British Journal of Surgery 2011


PART II PREVENTING COMPLICATED PANCREATITIS


CHAPTER 5 Timing of enteral nutrition in acute pancreatitis: Meta-analysis of individuals using a single-arm of randomised trials Pancreatology 2014

Olaf J Bakker, Sandra van Brunschot, Antoni Farre, Colin D Johnson, Fotis Kalfarentzos, Brian E Louie, Attila OlĂĄh, Stephen J Oâ&#x20AC;&#x2122;Keefe, Maxim S Petrov, James J Powell, Marc G Besselink, Hjalmar C van Santvoort, Maroeska M Rovers, Hein G Gooszen


PART II CHAPTER 5

A B S T R AC T Intro du c ti on In acute pancreatitis, enteral nutrition (EN) reduces the rate of complications, such as infected pancreatic necrosis, organ failure, and mortality, as compared to parenteral nutrition (PN). Starting EN within 24 hours of admission might further reduce complications. Me tho ds A literature search for trials of EN in acute pancreatitis was performed. Authors of eligible trials were requested to provide the data of all patients in the EN-arm of their trials. A meta-analysis of individual patient data was performed. The cohort of patients with EN was divided into patients receiving EN within 24 hours or after 24 hours of admission. Multivariable logistic regression, adjusting for predicted disease severity and trial, was used to study the effect of timing of EN on a composite endpoint of infected pancreatic necrosis, organ failure, or mortality. Resu lts Observational data from 165 individuals from 8 randomised trials were obtained; 100 patients with EN within 24 hours and 65 patients with EN after 24 hours of admission. In the multivariable model, EN started within 24 hours of admission compared to EN started after 24 hours of admission, reduced the composite endpoint from 45% to 19% (adjusted odds ratio [OR] of 0.44; 95% confidence interval [CI] 0.20 â&#x20AC;&#x201C; 0.96). Within the composite endpoint, organ failure was reduced from 42% to 16% (adjusted OR 0.42; 95% CI 0.19 â&#x20AC;&#x201C; 0.94). C onclusi ons In this meta-analysis of observational data from individuals with acute pancreatitis, starting EN within 24 hours, compared starting after 24 hours after admission, was associated with a reduction in complications.

76


META-ANALYSIS OF TIMING OF ENTERAL NUTRITION

I N T RODU C T ION Acute pancreatitis is the most common gastro-intestinal reason for hospitalization in the United States with over 274.000 admissions annually.1 Inpatient costs are estimated to exceed 2.5 billion dollar per year. The mortality of patients with acute pancreatitis is determined by the presence of organ failure or infection of pancreatic necrosis.2,3 Organ failure and infected pancreatic necrosis are associated with a mortality of up to 30%. To improve outcome, early intervention strategies in patients with acute pancreatitis aim at reducing the duration of organ failure and the development of infected pancreatic necrosis. So far, several randomised trials, using pharmacotherapeutics, antibiotics, or probiotics, have failed to reduced the rate of either organ failure or infected pancreatic necrosis.4-6 However, enteral nutrition (EN) has shown that, when compared to parenteral nutrition (PN), it reduces the rate of infected pancreatic necrosis, organ failure and mortality.7 In acute pancreatitis, in the first hours after onset of symptoms, injury to the pancreatic gland initiates a local inflammatory response which causes the release of cytokines, chemokines, neutrophils and other inflammatory mediators.8,9 This evolving inflammatory response can cause a series of pathophysiological events in distant organs and especially the gut, including disturbed gastrointestinal motility, bacterial overgrowth, reduction of arterial blood flow, increased permeability of the gastrointestinal mucosal barrier and bacterial translocation.10-13 Increased bacterial translocation exacerbates the systemic inflammatory response and may cause distant infections such as infected pancreatic necrosis.14 So, the gut plays a pivotal role in the development of organ failure and infected necrosis in acute pancreatitis. As a potent stimulator of intestinal blood flow and gut motility, early intraluminal EN is hypothesized to preserve intestinal mucosa, prevent bacterial translocation, reduce the inflammatory response and potentially reduce the rate of organ failure and infected necrosis.15 In a meta-analysis of 15 randomised trials in critically ill patients, but without patients with acute pancreatitis, EN started within 24 hours of admission reduced the rate of infections.16 To investigate whether the hypothesized benefit of starting EN very early in acute pancreatitis is supported by evidence from clinical trials, we performed a meta-analysis of individual patient data from the EN-arm of randomised trials.

M E T HOD S S e arch str ateg y and s el e c ti on of tri a ls A systematic literature search was performed for randomised trials investigating a very early start of EN. However, the results of this literature search showed that no randomised 77


PART II CHAPTER 5

trials have been performed that specifically investigated a very early start of EN compared to a delayed start of EN. Therefore, a second literature search was performed that was directed at randomised trials with early EN in one arm of the study in adults with acute pancreatitis. To investigate the effect of timing of EN based on data from randomised trials with an EN-arm, we aimed to aggregate the individual data of all patients who were included in the EN-arm of the trials. The following inclusion criteria were used: consecutive patients with acute pancreatitis, use of a validated classification system or generally accepted parameter to predict severity on admission, and initiation of EN according to a prespecified protocol. We searched in Pubmed, Embase, Cochrane, CINAHL and Web of knowledge with the Highly Sensitive Search Strategy17 and the search term ‘acute pancreatitis’ (date of literature search 31st of January 2011). In addition, Haynes’ clinical query (broad, sensitive search) in combination with ‘acute pancreatitis’ was used to check if all trials were retrieved with the first strategy.18 Titles and abstracts were screened for eligibility and reference lists of reviews, meta-analyses and all included studies were cross-checked manually to search for additional eligible papers. Tri a l elig ibi lity and qu a lity ass essment The quality of included studies was assessed for the presence of allocation concealment, reporting of incomplete data, selective outcome reporting and intention-to-treat analysis.19 Two authors (OJB and SvB) independently assessed the eligibility of the selected full text publications. The primary investigators of all selected trials were asked for the raw data of their trials. In case primary investigators did not reply they were contacted at regular intervals or attempts were made to reach them by telephone. The data obtained were checked for consistency, integrity of randomisation and follow-up. All trials obtained informed consent and ethics approval. O utcome vari abl es The primary outcome was a composite of mortality, infected pancreatic necrosis, or organ failure. This composite endpoint was chosen as infected pancreatic necrosis and organ failure are the major causes of mortality in acute pancreatitis.2,3 Secondary outcomes were the individual components of the primary outcome. Presp e cif i e d subg roups Prespecified subgroup analyses were performed for patients with predicted severe pancreatitis (regardless of the predictive score used) and for patients with necrotizing pancreatitis. These subgroups were prespecified as most complications occur in patients with predicted severe pancreatitis and in patients with necrotizing pancreatitis.20

78


META-ANALYSIS OF TIMING OF ENTERAL NUTRITION

Statisti c a l ana lysis To assess whether timing of EN is associated with the incidence of the different endpoints, we performed a one stage meta-analysis using logistic regression analysis. In this model, the independent variables were: EN within or after 24 hours after admission; the potential confounders (age, gender, etiology, presence of necrosis, and predicted severity based on Acute Physiology And Chronic Health Evaluation-II [APACHE-II], Imrie or modified Glasgow score, Ranson score, or CRP). A dummy variable was added to adjust for in between variance and to identify each study within the regression analysis.19 To increase statistical efficiency and to reduce bias, we developed the covariate ‘predicted severe pancreatitis’. Patients were classified as having ‘predicted severe pancreatitis’ if these patients either had an APACHE-II ≥ 8, an Imrie score ≥ 3, a Ranson score ≥ 3, or a CRP ≥ 150 mg/L. This covariate was available for all patients and was used in the model instead of the separate severity predictors. Risk differences (RD) and adjusted odds ratios (ORs) with corresponding 95% confidence intervals (CIs) and the number-needed-to-treat (NNT) were calculated for the composite primary endpoint and for the secondary endpoints. Sensitivity analyses were planned to assess whether results would differ if data from unavailable trials would have been included in the meta-analysis.19

R E SU LT S Out of 20.320 abstracts retrieved from 5 different databases, we identified 13 trials that met the eligibility criteria (Figure 1).21-33 Data from 6 trials23,24,30,32,33 could not be retrieved because data were no longer available (n = 3) or no contact was established with the principal investigators despite repeated attempts (n = 3). After personal communication with one of the authors, data from 29 consecutive patients with acute pancreatitis and EN, who were considered for inclusion in another randomised trial that was pending publication at time of our literature search, was added to the dataset.34 Out of 13 potentially eligible trials, including 325 patients treated with EN, we obtained the data from 165 patients (51%). From one trial, which included additional patients during a second phase of recruitment to reach the estimated sample size, only data from the first phase of recruitment were available.21 One trial compared EN with conventional therapy instead of comparing EN to PN.31 Conventional therapy consisted of a nil-by-mouth regimen until patients were able to tolerate oral nutrition. Although, no patients within the conventional therapy group received EN within 24 hours, they were included in the study. These patients were included for two reasons. First, the numbers of fasting days were carefully registered prior to start of EN or oral diet. Second, they did not receive PN and hence were not 79


PART II CHAPTER 5

prone to its potential complications. Therefore these patients were included in the cohort of patients with start of EN after 24 hours. Table 1 shows the main characteristics of the 8 included studies. Six of 8 trials included patients with predicted severe pancreatitis based on APACHE-II, Imrie score, or CRP22,25-27,29,31; one trial included patients who did not clinically improve after 48 hours after admission and registered Ranson score as predicted severity parameter21; one trial included patients with predicted severe pancreatitis and inability to tolerate oral fluids27, and one trial included consecutive patients with acute pancreatitis without signs of severe or critical disease on admission34. All trials used a protocol for the timing of start of EN which differed from within 6 hours of diagnosis to after 48 hours of admission. The methodological quality of included trials was generally high, although none of the studies were blinded for intervention (as this was not possible due to the nature of interventions) and none of the trials used a blinded endpoint assessment. Allocation concealment was adequate. All patients randomised completed the study and data for the composite endpoint were complete. Individual patient data showed that despite protocol, timing of EN differed between patients within trials and ranged from start on day of admission to start on the 10th day after admission. One hundred patients (61%) were given EN within 24 hours of admission and 65 patients (39%) received EN more than 24 hours after admission.

Figure 1. Selection of studies for individual patient data meta-analysis on timing of enteral nutrition. NOTE *After personal communication with the author, data from one other trial were included. This trial was published after the literature search for this study.

80


No clinical improvement after 48 hrs APACHE ≥ 6, Imrie ≥ 3, CRP ≥ 150 APACHE ≥ 6

2000 28*

2002 20#

2005 10

2006 11

2012 29

USA

Hungary 2002 41

2003 8

UK

UK

Canada

Spain

New Zealand

Powell31

Abou-Assi21

Olah29

Gupta25

Louie27

Casas22

Petrov34

Route

Type

Outcomes

< 6 hrs of diagnosis

Nasojejunal polymeric

< 24 hrs of admission Nasojejunal elemental

> 48 hrs of admission Nasojejunal elemental

< 24 hrs vs after Nasojejunal polymeric clinical improvement

Measurements of systemic inflammation, complications

Complications

Duration of hospitalization, complications

Modulation of markers of the inflammatory response, complications

< 48 hrs of admission Nasojejunal semielemental Complications

Timing

EN protocol

< 24 hrs of admission Nasogastric semielemental Complications

< 48 hrs of admission Nasojejunal semielemental Acute inflammatory response, complications

NOTE. EN denotes enteral nutrition. * In the original study, one patient was excluded after randomisation.31 According to the intention-to-treat principle this patient has been included in the present study. # Original study recruited patients in 2 phases.21 For this meta-analysis only data from the first phase of the study were available.

Consecutive patients with acute pancreatitis

APACHE ≥ 8, CRP ≥ 150, Balthazar D or E

Ranson ≥ 3, intolerance for oral < 24 hrs of enrolment Nasojejunal semielemental Measures of nutrition and fluids after a maximum of 96 inflammation, complications hrs

APACHE ≥ 7, Imrie ≥ 3

APACHE ≥ 8, Imrie ≥ 3, CRP ≥ 120, Balthazar D or E

1997 18

Greece

Kalfarentzos

26

Patients Participants with EN

Year

Country

Study

Table 1. Characteristics of the 8 included trials in the meta-analysis.

META-ANALYSIS OF TIMING OF ENTERAL NUTRITION

81


PART II CHAPTER 5

Baseline characteristics are given in Table 2. As definitions and parameters used for predicted severe pancreatitis differed between trials, no particular score or parameter was available for all patients (see Table 2). The overall parameter ‘predicted severe pancreatitis’ based on APACHE-II score ≥ 8, Imrie score ≥ 3, Ranson score ≥ 3, or CRP > 150 mg/L was available for all patients. Although, 6 out of 8 trials aimed to include patients with predicted severe pancreatitis, individual data showed that only 95 of 165 patients (58%) were deemed to have predicted severe pancreatitis (as defined in the methods section). In the cohort of patients who received EN after 24 hours, a significantly greater proportion was defined as having predicted severe pancreatitis (74% compared to 47%; P < 0.001). The proportion of patients with necrotizing pancreatitis did not differ between groups. The modified Glasgow score differed significantly between both groups (median 2 versus 4; P = 0.01) but was only available for 47% of patients. Our one-stage meta-analysis, using multivariable logistic regression analysis and adjusting for predicted severity and trial, showed that EN within 24 hours after admission was associated with a lower incidence of the primary composite endpoint (Table 3). Nineteen percent of patients with EN within 24 hours suffered from infected pancreatic necrosis, organ failure, or mortality compared to 45% of patients with EN after 24 hours. Table 2. Baseline characteristics of patients in the 8 trials EN within 24 hrs EN after 24 hrs (n = 100) (n = 65) Age Women

55 (45 – 70)

0.20

1 (1%)

37 (37%)

22 (34%)

0.74

0 (0%)

0.06

0 (0%)

< 0.001

0 (0%)

Alcohol

44 (44%)

23 (35%)

Biliary

35 (35%)

25 (39%)

Hypertriglyceridemia

0 (0%)

3 (5%)

Medication

1 (1%)

4 (6%)

Other

4 (4%)

4 (6%)

Idiopathic

Missing Data (n = 165)

53 (42 – 66)

Etiology

Predicted severe pancreatitis*

P value

16 (16%)

6 (9%)

47 (47%)

48 (74%)

APACHE-II Score

8 (6 – 12)

9 (7 – 13)

0.17

72 (44%)

CRP on admission

53 (18 – 129)

110 (34 – 229)

0.05

47 (28%)

CRP within 48 hours of admission

186 (77 – 285)

149 (58 – 286)

0.66

118 (72%)

Imrie or modified Glasgow score

2 (1 – 4)

4 (2 – 5)

0.01

78 (47%)

40 (40%)

29 (45%)

0.40

17 (10%)

Necrotizing pancreatitis

Continuous variables are median (interquartile range). EN denotes enteral nutrition. *Predicted severe pancreatitis defined as APACHE-II score ≥ 8, Imrie score ≥ 3, Ranson score ≥ 3, or CRP > 150 mg/L (on admission or within 48 hours of admission). APACHE-II = Acute Physiology and Chronic Health Evaluation II; CRP = C-reactive protein

82


META-ANALYSIS OF TIMING OF ENTERAL NUTRITION

Table 3. Outcome of patients with enteral nutrition within and after 24 hours of admission.

All patients I nfected pancreatic necrosis, Organ Failure, or Mortality Infected pancreatic necrosis Organ Failure Mortality Predicted severe pancreatitis* I nfected pancreatic necrosis, Organ Failure, or Mortality

EN within 24 hrs

EN after 24 hrs

Adjusted OR (95% CI)

NNT

(n = 100)

(n = 65)

19 (19%)

29 (45%)

0.44 (0.20 – 0.96)

4

7 (7%)

9 (14%)

0.65 (0.21 – 1.99)

15

16 (16%)

27 (42%)

0.42 (0.19 – 0.94)

4

3 (3%)

8 (12%)

0.38 (0.09 – 1.56)

11

(n = 47)

(n = 48)

18 (38%)

26 (54%)

0.55 (0.24 – 1.26)

6

Infected pancreatic necrosis

7 (15%)

9 (19%)

0.66 (0.21 – 1.99)

26

Organ Failure

15 (32%)

24 (50%)

0.51 (0.22 – 1.20)

6

0.46 (0.11 – 1.93)

12

0.50 (0.18 – 1.43)

4

Mortality Necrotizing pancreatitis I nfected pancreatic necrosis, Organ Failure, or Mortality

3 (6%)

7 (15%)

(n = 40)

(n = 29)

14 (35%)

17 (59%)

Infected pancreatic necrosis

7 (18%)

9 (31%)

0.56 (0.17 – 1.81)

7

Organ Failure

11 (28%)

15 (52%)

0.48 (0.16 – 1.42)

4

3 (8%)

5 (17%)

0.67 (0.13 – 3.53)

10

Mortality

EN denotes enteral nutrition. *Predicted severe pancreatitis defined as APACHE-II score ≥ 8, Imrie score ≥ 3, Ranson score ≥ 3, or CRP > 150 mg/L. CI = Confidence Interval; NNT = Number Needed to Treat; OR = Odds Ratio

This resulted in an adjusted OR of 0.44 (95% CI 0.20 – 0.96) and a NNT of 4 patients. Within the composite endpoint, a significant reduction was seen in the rate of organ failure following EN within 24 hours (16% compared to 42%; adjusted OR 0.42; 95% CI 0.19 – 0.94, NNT 4). Results for the prespecified subgroups are given in Table 3. In the subgroups of patients with predicted severe pancreatitis and in patients with pancreatic necrosis results were consistently better if EN was started within 24 hours. However, none of these differences reached statistical significance. In the subgroup of patients with predicted severe pancreatitis, the composite endpoint was reduced from 54% to 38% (adjusted OR 0.55, 95% CI 0.24 – 1.26). We were unable to perform the planned sensitivity analyses as from the publications of the unavailable trials the essential data on start of EN could not be extracted.

83


PART II CHAPTER 5

DI S C U S SION This meta-analysis of individual patient data from a single-arm of randomised trials showed that start of EN within 24 hours of admission was associated with a reduction in the composite endpoint of mortality, infected pancreatic necrosis, and organ failure. In a multivariable logistic regression analysis adjusting for a significant difference in predicted severity at baseline, the NNT was 4, i.e. 4 patients should be treated with EN to prevent one endpoint. Within the composite endpoint, it also reduced the onset of organ failure compared to EN after 24 hours of admission. This study is the first individual patient data meta-analysis of randomised trials in acute pancreatitis. A collaboration between physicians from 7 countries resulted in a dataset of patients from 8 trials. Individual patient data meta-analyses have certain advantages over meta-analyses of the literature such as the ability to perform additional analyses, adequate subgroup analyses and multivariable regression analyses.19 From the previous published data of the trials included in this meta-analysis, the timing of EN in individual patients cannot be extracted. The individual data on the exact start of EN in each patient enabled us to perform a comparison between a start of EN within and after 24 hours of hospital admission. Second, the individual data enabled the construction of a multivariable logistic regression model. In this model, we were able to adjust for the potential confounding effect of the baseline characteristic â&#x20AC;&#x2DC;predicted severe pancreatitisâ&#x20AC;&#x2122;. In addition, by adding a dummy variable for each trial, we adjusted for potential hidden confounders and heterogeneity between trials. Third, no large scale randomised trials on timing of EN in acute pancreatitis have been performed. So, for a meaningful comparison, it is necessary to combine data from several studies. At this moment, this observational meta-analysis of single-arm data from randomised trials provides the most accurate conclusions based on the available literature and provides a basis for future guideline formulations. Fourth, the individual data enabled us to perform subgroup analyses. In the subgroup analyses of patients with predicted severe pancreatitis and in patients with necrotizing pancreatitis, the effect of EN within 24 hours seemed consistently in favour of EN within 24 hours of admission although statistical significance was not reached. This lack of statistical significance may be due to a type II error or a false negative result. A larger sample of patients with predicted severe pancreatitis might show a significant reduction. Unfortunately in acute pancreatitis such large studies are seldom performed. Last, the NNT for the primary endpoint was 4. If this beneficial effect of EN turns out to be only half as strong as was found in this study, it will still be an important clinical improvement that can be achieved with a relatively simple, safe and inexpensive treatment strategy. This meta-analysis also has several weaknesses. The aim of the included randomised trials was to compare EN to PN instead of investigating timing of EN. Therefore this 84


META-ANALYSIS OF TIMING OF ENTERAL NUTRITION

study is a secondary analysis of single-arm data from randomised trials and should therefore be considered as a meta-analysis of prospective observational studies (instead of a pooled comparison of two randomized interventions). Second, the composite primary outcome of this study was not the primary outcome in the included trials. One could argue that data collection on secondary endpoints might be less accurate. The endpoints used in this study are, however, well-known and established complications of acute pancreatitis.2,3 Even more, mortality, infected pancreatic necrosis, and organ failure, are major clinical events which do not require extensive additional investigations to be detected during standard care. Hence, we consider the risk of missing endpoints to be low. Third, it is known that combined outcomes can make a treatment seem more, or less, effective than it really is.35 We therefore performed sensitivity analyses using different composite outcomes, which all showed similar results. By using such a composite endpoint we were able to aggregate our data and generalise results. This was critical to accomplish our aim of establishing whether early EN is effective in patients with pancreatitis. Fourth, inclusion criteria differed between the trials. Different cut-offs were used to define predicted severe pancreatitis. A combination of available APACHE-II scores, Imrie score, Ranson scores, and CRPs resulted in an overall binary â&#x20AC;&#x2DC;predicted severe pancreatitisâ&#x20AC;&#x2122; parameter. This parameter was available for all patients. Previous studies have also used a combination of scores or parameters to define predicted severe pancreatitis.4,26,29 One trial included patients after a period of fasting of at least 48 hours after admission.21 As a result, patients from that trial could only be included in the cohort of EN after 24 hours. Exclusion of patients from that trial did not change results (data not shown). Fifth, of 13 potentially eligible trials data from only 7 trials were collected (see Figure 1). Would results differ if other trials would have been included? We believe results would not be different as the unavailable trials used comparable inclusion criteria and EN regimens. Most importantly, outcomes across the unavailable trials are comparable to the trials included in this study.23,24,28,30,32,33 Unfortunately the planned sensitivity analyses to statistically confirm lack of selection bias could not be performed as the start of EN is not available in the published articles. The results found in this study confirm findings from earlier experimental and clinical studies on the role of EN and its influence on the gut barrier in acute pancreatitis.36 Intestinal mucosal integrity is compromised in acute pancreatitis. A clinical study in 58 patients with acute pancreatitis showed an early increase in intestinal permeability.11 Experimental studies have shown that, compared with TPN, EN has the ability to maintain gut barrier function.37,38 In a small retrospective study of 17 selected patients with acute pancreatitis in the intensive care unit, early initiation of EN was associated with a low mortality rate.39 A larger retrospective study of 197 patients with predicted severe pancreatitis showed that EN started within the first 48 hours after admission might lead to a reduction in mortality, infected pancreatic necrosis, respiratory failure and 85


PART II CHAPTER 5

intensive care admission.40 A meta-analysis of randomised controlled trials of EN versus PN investigated the effect of timing of EN on clinical outcomes in patients with acute pancreatitis.41 This was based on the start of nutrition that was described in the manuscripts as opposed to the individual patient data as was used in this study. The outcomes in trials that planned to start EN within 24 hours and within 48 hours after admission were compared to the outcome in trials that planned to start EN after 24 hours and after 48 hours. Results suggested that EN administered within 48 hours might be more effective. A recent randomised trial compared EN within 48 hours after admission to EN after 8 days of fasting in 60 patients with acute pancreatitis.42 The primary aim was to investigate immune function. The authors concluded that early EN moderates the excessive immune response. Despite the relative small sample size differences in clinical outcomes were found. Multiple organ dysfunction syndrome, systemic inflammatory response syndrome, pancreatic infection and duration of intensive care stay were significantly reduced in the early EN group. ­This study also confirms results found in studies on early EN in critically ill patients in general. In these patients, early EN is recommended as standard treatment.43,44 This is the result of substantial evidence confirming the benefits of very early initiation of EN in critically ill patients.45 For example, in a meta-analysis of 15 randomised trials including critically ill patients but no pancreatitis patients, very early EN was associated with a lower incidence of infections (relative risk reduction, 0.45 95% CI 0.30 - 0.66).16 A recent large cohort study in intensive care patients found that EN within 48 hours of the start of mechanical ventilation was associated with reduced mortality.46 This finding is remarkable as all included patients were on vasopressors and required ventilatory support for more than 2 days. Patients with mild pancreatitis in general do not need tube feeding.47,48 Oral feeding is feasible in these patients though it may lead to pain relapse in a considerable proportion of patients.49,50 Patients with severe pancreatitis need nutritional support. To predict on admission whether patients will develop severe pancreatitis, predicted severity scoring systems are usually used. Six out of 8 trials in this study included patients with predicted severe pancreatitis. However, the subgroup of patients with predicted severe pancreatitis in this study consisted of only 95 of 165 (56%) patients. This is explained by the fact that the definition we used for predicted severe pancreatitis is more stringent than the definition used in some of the trials. More specifically, in this study predicted severe pancreatitis was based on an APACHE-II score ≥ 8, while in 2 trials an APACHE-II ≥ 625,26 was used and in one trial an APACHE-II ≥ 7 31 was used. In addition, one trial did not use a predictive scoring system but included patients that showed no clinical improvement after 48 hours 21 and one trial used Ranson ≥ 3 or intolerance for oral fluids27. Also, most trials used 48 to 96 hours after admission to stratify patients into predicted mild or predicted severe pancreatitis. In this study, patients were stratified 86


META-ANALYSIS OF TIMING OF ENTERAL NUTRITION

within 24 hours into predicted mild or predicted severe based on the individual data on admission. For these reasons, the subgroup of patients with predicted severe pancreatitis in this study is relatively small. On the other hand, when we consider the non-negligible percentage of complications such as organ failure and infected necrosis that occurred in the group of patients that we stratified as having predicted mild pancreatitis, we can conclude that our stratification has been conservative. In conclusion, this study found that initiation of EN within 24 hours after hospital admission in patients with acute pancreatitis was associated with a large reduction in the onset of complications, especially organ failure. Early EN in only 4 patients is needed to prevent complications in one. However, as this was a secondary analysis of single-arm data from a selection of randomised trials and only a subgroup of patients were defined as having predicted severe pancreatitis, further studies are warranted to investigate the optimal timing of EN administration in patients with acute pancreatitis. Author contributi ons OJB and MMR and several other members of the group designed the study. OJB requested and received the data from the different trials. SvB performed on-site additional data collection. OJB and MMB designed the analysis plan and performed the statistical analysis. OJB drafted the first and subsequent versions of the manuscript. OJB had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors read, corrected and approved the final manuscript.

87


PART II CHAPTER 5

REFERENCES 1. Peery AF, Dellon ES, Lund J, et al. Burden of gastrointestinal disease in the United States: 2012 update. Gastroenterology 2012; 143:1179-1187. 2. van Santvoort HC, Bakker OJ, Bollen TL, et al. A conservative and minimally invasive approach to necrotizing pancreatitis improves outcome. Gastroenterology 2011; 141:1254-1263. 3. Petrov MS, Shanbhag S, Chakraborty M, et al. Organ failure and infection of pancreatic necrosis as determinants of mortality in patients with acute pancreatitis. Gastroenterology 2010;139:813-820. 4. Besselink MG, van Santvoort HC, Buskens E, et al. Probiotic prophylaxis in predicted severe acute pancreatitis: a randomised, double-blind, placebo-controlled trial. Lancet 2008;371:651-659. 5. Johnson CD, Kingsnorth AN, Imrie CW, et al. Double blind, randomised, placebo controlled study of a platelet activating factor antagonist, lexipafant, in the treatment and prevention of organ failure in predicted severe acute pancreatitis. Gut 2001;48:62-69. 6. Wittau M, Mayer B, Scheele J, et al. Systematic review and meta-analysis of antibiotic prophylaxis in severe acute pancreatitis. Scand J Gastroenterol 2011;46:261-270. 7. Petrov MS, van Santvoort HC, Besselink MG, et al. Enteral nutrition and the risk of mortality and infectious complications in patients with severe acute pancreatitis: a meta-analysis of randomized trials. Arch Surg 2008;143:1111-1117. 8. Makhija R, Kingsnorth AN. Cytokine storm in acute pancreatitis. J Hepatobiliary Pancreat Surg 2002;9:401-410. 9. Pandol SJ, Saluja AK, Imrie CW, et al. Acute pancreatitis: bench to the bedside. Gastroenterology 2007;132:1127-1151. 10. Alverdy JC, Chang EB. The re-emerging role of the intestinal microflora in critical illness and inflammation: why the gut hypothesis of

88

sepsis syndrome will not go away. J Leukoc Biol 2008;83:461-466. 11. Ammori BJ, Leeder PC, King RF, et al. Early increase in intestinal permeability in patients with severe acute pancreatitis: correlation with endotoxemia, organ failure, and mortality. J Gastrointest Surg 1999;3:252-262. 12. Rahman SH, Ammori BJ, Holmfield J, et al. Intestinal hypoperfusion contributes to gut barrier failure in severe acute pancreatitis. J Gastrointest Surg 2003;7:26-35. 13. Van Felius I, Akkermans LM, Bosscha K, et al. Interdigestive small bowel motility and duodenal bacterial overgrowth in experimental acute pancreatitis. Neurogastroenterol Motil 2003;15:267-276. 14. Fritz S, Hackert T, Hartwig W, et al. Bacterial translocation and infected pancreatic necrosis in acute necrotizing pancreatitis derives from small bowel rather than from colon. Am J Surg 2010;200:111-117. 15. McClave SA. Drivers of oxidative stress in acute pancreatitis: the role of nutrition therapy. JPEN J Parenter Enteral Nutr 2012;36:24-35. 16. Marik PE, Zaloga GP. Early enteral nutrition in acutely ill patients: a systematic review. Crit Care Med 2001;29:2264-2270. 17. Robinson KA, Dickersin K. Development of a highly sensitive search strategy for the retrieval of reports of controlled trials using PubMed. Int J Epidemiol 2002;31:150-153. 18. Montori VM, Wilczynski NL, Morgan D, et al. Optimal search strategies for retrieving systematic reviews from Medline: analytical survey. BMJ 2005;330:68. 19. Riley RD, Lambert PC, Abo-Zaid G. Metaanalysis of individual participant data: rationale, conduct, and reporting. BMJ 2010;340:c221. 20. Banks PA, Freeman ML. Practice guidelines in acute pancreatitis. Am J Gastroenterol 2006;101:2379-2400.


META-ANALYSIS OF TIMING OF ENTERAL NUTRITION

21. Abou-Assi S, Craig K, O'Keefe SJ. Hypocaloric jejunal feeding is better than total parenteral nutrition in acute pancreatitis: results of a randomized comparative study. Am J Gastroenterol 2002;97:2255-2262. 22. Casas M, Mora J, Fort E, et al. [Total enteral nutrition vs. total parenteral nutrition in patients with severe acute pancreatitis]. Rev Esp Enferm Dig 2007;99:264-269. 23. Doley RP, Yadav TD, Wig JD, et al. Enteral nutrition in severe acute pancreatitis. JOP 2009;10:157-162. 24. Eckerwall GE, Axelsson JB, Andersson RG. Early nasogastric feeding in predicted severe acute pancreatitis: A clinical, randomized study. Ann Surg 2006;244:959-965. 25. Gupta R, Patel K, Calder PC, et al. A randomised clinical trial to assess the effect of total enteral and total parenteral nutritional support on metabolic, inflammatory and oxidative markers in patients with predicted severe acute pancreatitis (APACHE II > or =6). Pancreatology 2003;3:406-413. 26. Kalfarentzos F, Kehagias J, Mead N, et al. Enteral nutrition is superior to parenteral nutrition in severe acute pancreatitis: results of a randomized prospective trial. Br J Surg 1997;84:1665-1669. 27. Louie BE, Noseworthy T, Hailey D, et al. 2004 MacLean-Mueller prize enteral or parenteral nutrition for severe pancreatitis: a randomized controlled trial and health technology assessment. Can J Surg 2005;48:298-306. 28. McClave SA, Greene LM, Snider HL, et al. Comparison of the safety of early enteral vs parenteral nutrition in mild acute pancreatitis. JPEN J Parenter Enteral Nutr 1997;21:14-20. 29. Olah A, Pardavi G, Belagyi T, et al. Early nasojejunal feeding in acute pancreatitis is associated with a lower complication rate. Nutrition 2002;18:259-262.

30. Petrov MS, Kukosh MV, Emelyanov NV. A randomized controlled trial of enteral versus parenteral feeding in patients with predicted severe acute pancreatitis shows a significant reduction in mortality and in infected pancreatic complications with total enteral nutrition. Dig Surg 2006;23:336-344. 31. Powell JJ, Murchison JT, Fearon KC, et al. Randomized controlled trial of the effect of early enteral nutrition on markers of the inflammatory response in predicted severe acute pancreatitis. Br J Surg 2000;87:13751381. 32. Windsor AC, Kanwar S, Li AG, Barnes et al. Compared with parenteral nutrition, enteral feeding attenuates the acute phase response and improves disease severity in acute pancreatitis. Gut 1998;42:431-435. 33. Wu XM, Ji KQ, Wang HY, et al. Total enteral nutrition in prevention of pancreatic necrotic infection in severe acute pancreatitis. Pancreas 2010;39:248-251. 34. Petrov MS, McIlroy K, Grayson L, et al. Early nasogastric tube feeding versus nil per os in mild to moderate acute pancreatitis: A randomized controlled trial. Clin Nutr 2012. 35. Tomlinson G, Detsky AS. Composite end points in randomized trials: there is no free lunch. JAMA 2010; 303:267-268. 36. Marik PE. What is the best way to feed patients with pancreatitis? Curr Opin Crit Care 2009;15:131-138. 37. Kang W, Kudsk KA. Is there evidence that the gut contributes to mucosal immunity in humans? JPEN J Parenter Enteral Nutr 2007;31:246-258. 38. Zou XP, Chen M, Wei W, et al. Effects of enteral immunonutrition on the maintenance of gut barrier function and immune function in pigs with severe acute pancreatitis. JPEN J Parenter Enteral Nutr 2010;34:554-566. 39. Hegazi R, Raina A, Graham T, et al. Early jejunal feeding initiation and clinical outcomes in patients with severe acute pancreatitis. JPEN J Parenter Enteral Nutr 2011;35:91-96.

89


PART II CHAPTER 5

40. Wereszczynska-Siemiatkowska U, SwidnickaSiergiejko A, Siemiatkowski A, et al. Early enteral nutrition is superior to delayed enteral nutrition for the prevention of infected necrosis and mortality in acute pancreatitis. Pancreas 2013;42:640-646. 41. Petrov MS, Pylypchuk RD, Uchugina AF. A systematic review on the timing of artificial nutrition in acute pancreatitis. Br J Nutr 2009;101:787-793. 42. Sun JK, Mu XW, Li WQ, et al. Effects of early enteral nutrition on immune function of severe acute pancreatitis patients. World J Gastroenterol 2013;19:917-922. 43. Kreymann KG, Berger MM, Deutz NE, et al. ESPEN Guidelines on Enteral Nutrition: Intensive care. Clin Nutr 2006;25:210-223. 44. McClave SA, Martindale RG, Vanek VW, et al. Guidelines for the Provision and Assessment of Nutrition Support Therapy in the Adult Critically Ill Patient: Society of Critical Care Medicine (SCCM) and American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.). JPEN J Parenter Enteral Nutr 2009;33:277-316. 45. McClave SA, Heyland DK. The physiologic response and associated clinical benefits from provision of early enteral nutrition. Nutr Clin Pract 2009;24:305-315. 46. Khalid I, Doshi P, DiGiovine B. Early enteral nutrition and outcomes of critically ill patients treated with vasopressors and mechanical ventilation. Am J Crit Care 2010;19:261-268. 47. UK guidelines for the management of acute pancreatitis. Gut 2005;54 Suppl 3:iii1-iii9. 48. Forsmark CE, Baillie J. AGA Institute technical review on acute pancreatitis. Gastroenterology 2007;132:2022-2044. 49. Eckerwall GE, Tingstedt BB, Bergenzaun PE, et al. Immediate oral feeding in patients with mild acute pancreatitis is safe and may accelerate recovery--a randomized clinical study. Clin Nutr 2007;26:758-763.

90

50. Petrov MS, van Santvoort HC, Besselink MG, et al. Oral refeeding after onset of acute pancreatitis: a review of literature. Am J Gastroenterol 2007; 102: 2079


91


PART II PREVENTING COMPLICATED PANCREATITIS


CHAPTER 6 Early versus on-demand nasoenteric tube feeding in acute pancreatitis New England Journal of Medicine 2014

Olaf J Bakker, Sandra van Brunschot, Hjalmar C van Santvoort, Marc G Besselink, Thomas L Bollen, Marja A Boermeester, Cornelis H Dejong, Harry van Goor, Koop Bosscha, Usama Ahmed Ali, Stefan Bouwense, Wilhelmina M van Grevenstein, Joos Heisterkamp, Alexander P Houdijk, Jeroen M Jansen, Thom M Karsten, Eric R Manusama, Vincent B Nieuwenhuijs, Alexander F Schaapherder, George P van der Schelling, Matthijs P Schwartz, BW Marcel Spanier, Adriaan Tan, Juda Vecht, Bas L Weusten, Ben J Witteman, Louis M Akkermans, Marco J Bruno, Marcel G Dijkgraaf, Bert van Ramshorst and Hein G Gooszen for the Dutch Pancreatitis Study Group


PART II CHAPTER 6

ABSTRACT B ackg round Early enteral feeding through a nasoenteric feeding tube is often used in patients with severe acute pancreatitis to prevent gut-derived infections, but evidence to support this strategy is limited. We conducted a multicenter, randomized trial comparing early nasoenteric tube feeding with an oral diet at 72 hours after presentation to the emergency department in patients with acute pancreatitis. Metho ds We enrolled patients with acute pancreatitis who were at high risk for complications on the basis of an Acute Physiology and Chronic Health Evaluation II score of 8 or higher (on a scale of 0 to 71, with higher scores indicating more severe disease), an Imrie or modified Glasgow score of 3 or higher (on a scale of 0 to 8, with higher scores indicating more severe disease), or a serum C-reactive protein level of more than 150 mg per liter. Patients were randomly assigned to nasoenteric tube feeding within 24 hours after randomization (early group) or to an oral diet initiated 72 hours after presentation (ondemand group), with tube feeding provided if the oral diet was not tolerated. The primary end point was a composite of major infection (infected pancreatic necrosis, bacteremia, or pneumonia) or death during 6 months of follow-up. Resu lts A total of 208 patients were enrolled at 19 Dutch hospitals. The primary end point occurred in 30 of 101 patients (30%) in the early group and in 28 of 104 (27%) in the on-demand group (risk ratio, 1.07; 95% confidence interval, 0.79 to 1.44; P= 0.76). There were no significant differences between the early group and the on demand group in the rate of major infection (25% and 26%, respectively; P= 0.87) or death (11% and 7%, respectively; P= 0.33). In the on-demand group, 72 patients (69%) tolerated an oral diet and did not require tube feeding. C onclusi ons This trial did not show the superiority of early nasoenteric tube feeding, as compared with an oral diet after 72 hours, in reducing the rate of infection or death in patients with acute pancreatitis at high risk for complications. (Funded by the Netherlands Organization for Health Research and Development and others; PYTHON Current Controlled Trials number, ISRCTN18170985.)

94


TRIAL OF NUTRITION IN ACUTE PANCREATITIS

I N T RODU C T ION Acute pancreatitis is the most common gastrointestinal disease leading to hospital admission, and its incidence continues to rise.1-4 Most patients with acute pancreatitis recover uneventfully and are discharged after a few days.5,6 In 20% of patients, the disease course is complicated by major infection, such as infected pancreatic necrosis, which is associated with a mortality of 15%.7-11 A meta-analysis of eight randomized trials involving 348 patients showed that nasoenteric tube feeding, as compared with total parenteral nutrition, reduced the rate of infections and mortality among patients with severe pancreatitis.12 These infections are thought to be mediated by bacterial translocation from the gut, provoked by disturbed intestinal motility, bacterial overgrowth, and increased mucosal permeability.13-18 Nasoenteric tube feeding is believed to stimulate intestinal motility - thus reducing bacterial overgrowth - and may increase splanchnic blood flow, which helps to preserve the integrity of the gut mucosa.19,20 Total parenteral nutrition lacks the trophic effect of enteric feeding and is associated with central venous catheterâ&#x20AC;&#x201C;related infections as well as metabolic complications.21 A meta-analysis of randomized trials involving acutely ill patients admitted to the hospital for indications other than pancreatitis showed a 22% reduction in the rate of major infection when nasoenteric tube feeding was started very early (â&#x2030;¤36 hours after admission or surgery) as compared with a later start.22 Similarly, nonrandomized studies of acute pancreatitis have shown that nasoenteric tube feeding started within 48 hours after admission, as compared with a start after 48 hours, significantly reduced the rate of major infection and in some studies even reduced mortality.23-26 On the basis of these potential benefits, American and European nutritional societies recommend routine early nasoenteric tube feeding in all patients with severe pancreatitis.27-29 Guidelines from gastroenterologic and pancreatic societies, however, state that, regardless of disease severity, tube feeding is indicated when patients are not able to tolerate an oral diet for up to 7 days.30,31 Unfortunately, it takes 3 to 4 days after admission to make this assessment,32 and by that time the window of opportunity for effective prevention of infection with early tube feeding has passed.7 To address this problem in the management of acute pancreatitis, we compared the effects of early nasoenteric tube feeding with those of an oral diet started at 72 hours, with a switch to nasoenteric tube feeding only in the case of insufficient oral intake.

95


PART II CHAPTER 6

M E T HOD S Stu dy p ar ti cip ants The protocol of the Pancreatitis, Very Early Compared with Selective Delayed Start of Enteral Feeding (PYTHON) trial has been published previously33 and is available with the full text of this article at NEJM.org. The study was conducted according to the protocol. Adults with a first episode of acute pancreatitis who were at high risk for complications (i.e., patients predicted to have severe pancreatitis) were eligible to undergo randomization. Patients were considered to be at high risk for complications if, within 24 hours after presentation to the emergency department, the Acute Physiology and Chronic Health Evaluation (APACHE) II34 score was 8 or higher (on a scale of 0 to 71, with higher scores indicating more severe disease), if the Imrie or modified Glasgow score35 was 3 or higher (on a scale of 0 to 8, with higher scores indicating more severe disease), or if the serum C-reactive protein level was more than 150 mg per liter.36 These assessments predict the development of complications during the course of the disease. Pancreatitis was diagnosed if at least two of the three following features were present: typical abdominal pain, a serum lipase or amylase level that was more than 3 times the upper limit of the normal range, or characteristic findings on cross-sectional imaging of the abdomen. The exclusion criteria are given in the Supplementary Appendix, available at NEJM.org. Stu dy d esig n and oversig ht The PYTHON trial was a multicenter, randomized, controlled superiority trial performed in six university medical centers and 13 large teaching hospitals of the Dutch Pancreatitis Study Group. Patients were randomly assigned in a 1:1 ratio either to nasoenteric tube feeding initiated within 24 hours after randomization (the early group) or to an oral diet starting at 72 hours (the on-demand group). Randomization was performed centrally by the study coordinator with the use of a Web-based system that used permuted-block randomization with a concealed, varying block size. Randomization was stratified according to treatment center and a dichotomized APACHE II score (<13 vs. â&#x2030;Ľ13); the latter stratification factor was used because patients with an APACHE II score of 13 or higher are at increased risk for major infection. All the patients or their legal representatives provided written informed consent. The study protocol was approved by the institutional review board of the University Medical Center Utrecht and by all the participating centers. All the authors vouch for the veracity and completeness of the data and data analyses. The sponsors were not involved in the design or conduct of the study or in the preparation of the manuscript or the decision to submit it for publication.

96


TRIAL OF NUTRITION IN ACUTE PANCREATITIS

Stu dy pro ce dures Patients underwent randomization within 24 hours after presentation to the emergency department. Those assigned to early nasoenteric feeding received a nasojejunal feeding tube as soon as possible but not later than 24 hours after randomization. Feeding tubes were placed endoscopically or radiologically, according to local practice. Nasoenteric feeding was administered as Nutrison Protein Plus (Nutricia). After tube placement, feeding was started at 20 ml per hour during the first 24 hours and was gradually increased (see the Supplementary Appendix). In the two study groups, full nutrition was defined as an energy target of 25 kcal per kilogram of body weight per day for patients in the intensive care unit (ICU) and 30 kcal per kilogram per day for patients in the ward.28,37,38 Patients assigned to an oral diet did not receive nutrition by any means other than that provided by standard intravenous f luids during the first 72 hours after presentation to the emergency department. Exceptions were made for patients who requested oral food during this 72-hour period. At 72 hours, all the patients in the on-demand group were given an oral diet. If an oral diet was not tolerated, it was offered again after 24 hours. If an oral diet still was not tolerated after 96 hours from the time of presentation, nasoenteric feeding was started after the placement of a nasojejunal tube, and the same procedure was followed as in the early group. E nd p oints The primary end point was a composite of major infection or death within 6 months after randomization. Major infection was defined as infected pancreatic necrosis, bacteremia, or pneumonia (for definitions, see Box S1 in the Supplementary Appendix). Predefined secondary end points included the development of necrotizing pancreatitis as diagnosed on the basis of computed tomography (CT) performed 5 to 7 days after admission (because pancreatic parenchymal necrosis may take up to 72 hours to develop) and the development of organ failure after randomization. D ata col l e c ti on and end-p oint ass essment Dieticians registered the caloric intake and calculated energy-intake targets during the first week after admission on the basis of actual body weight. All CT studies were interpreted by an author who is an experienced radiologist and who was unaware of the treatment assignments. An adjudication committee, consisting of four pancreatic surgeons and a gastroenterologist who were unaware of the treatment assignments, individually evaluated each patient for the occurrence of the primary end point before interim and final analyses. Disagreements with respect to major infection were resolved during a plenary consensus meeting.

97


PART II CHAPTER 6

Pati ent s afety An independent data and safety monitoring committee evaluated the progress of the trial and examined safety end points after the completion of follow-up in each consecutive group of 25 patients. Adverse events were listed and presented to the data and safety monitoring committee in an unblinded fashion. Statisti c a l ana lysis The expected incidence of the primary end point in the on-demand group was based on data from individual patients in the placebo group of a previous randomized trial.39 For the early group, data from randomized trials comparing nasoenteric with parenteral nutrition were used to estimate the incidence.33,40-42 The sample-size calculation was based on an expected reduction in the primary composite end point associated with early tube feeding from 40 to 22%.33 We estimated that a sample of 208 patients would provide the study with at least 80% power, at a twosided alpha level of 5% and assuming a 1% loss to follow-up. Analysis was based on the intention-to-treat method, with the exclusion only of patients for whom the adjudication committee, whose members were unaware of the treatment assignments, decided before any analysis that the diagnosis of acute pancreatitis was incorrect. Predefined subgroups included patients with an APACHE II score below 13 and those with a score of 13 or higher at randomization. Two post hoc subgroup analyses were performed: one for patients with the systemic inflammatory response syndrome (SIRS, as defined by the Consensus Conference criteria of the American College of Chest Physiciansâ&#x20AC;&#x201C;Society of Critical Care Medicine) at randomization, because such patients are at high risk for complications,43 and one for a low or high body-mass index (BMI; the weight in kilograms divided by the square of the height in meters), since the BMI differed significantly between the two treatment groups at baseline. An interim analysis of the primary end point was performed after 50% of the patients had completed 6 months of follow-up. The interim analysis was performed by an independent statistician, who was unaware of the treatment assignments, applying the Peto approach with symmetric stopping boundaries at a P value of less than 0.001.33,44 For the final analyses, a two-sided P value of less than 0.05 was considered to indicate statistical significance. P values were not adjusted for multiple testing.

RESULTS E nrol lment and r and omizati on From August 2008 through June 2012, a total of 867 patients were assessed for eligibility (Fig. S1 in the Supplementary Appendix). A total of 208 patients (24%) were enrolled 98


TRIAL OF NUTRITION IN ACUTE PANCREATITIS

and randomly assigned to early nasoenteric tube feeding (102 patients) or an oral diet with tube feeding if required (106). The adjudication committee excluded 3 patients who had undergone randomization and had been incorrectly diagnosed with acute pancreatitis (2 patients had gastric carcinoma and 1 had intestinal volvulus). A total of 101 patients in the early group and 104 in the on-demand group were included in the intention-totreat analysis. Baseline characteristics, presented in Table 1, were equally distributed between the groups except for the mean (±SD) BMI (29±5 in the early group vs. 27±5 in the on-demand group, P= 0.01). Details regarding the number of calories delivered during the first week after admission and the timing of feeding are shown in Figure 1, and in Table S1 in the Supplementary Appendix. As specified by the protocol, patients in the early group received feeding earlier than those in the on-demand group. Nasoenteric tube feeding in the early group was started a median of 8 hours after randomization and a median of 23 hours after presentation to the emergency department, as compared with initiation of an oral diet 64 hours after randomization and 72 hours after presentation in the on-demand group (P<0.001) (Table S1 in the Supplementary Appendix). A total of 5 of 104 patients (5%) assigned to on-demand feeding requested and received food within the first 72 hours after presentation.

Figure 1. Calories Delivered with the Use of Early versus On-Demand Naso enteric Tube Feeding. Each rectangle shows the mean value (horizontal line) and 95% confidence interval (top and bottom of the rectangle).

99


PART II CHAPTER 6

Table 1. Characteristics of the Patients at Baseline.* Characteristic

Early Tube Feeding (N = 101)

On-Demand Tube Feeding (N = 104)

Female sex - no. (%)

46 (46)

45 (43)

Age - yr

65±16

65±15

Gallstones

59 (58)

56 (54)

Alcohol abuse

14 (14)

23 (22)

Other

28 (28)

25 (24)

<25

20/99 (20)

33/103 (32)

25 to <35

69/99 (70)

67/103 (65)

≥35

10/99 (10)

3/103 (3)

Cause of pancreatitis - no. (%)

Body-mass index - no./total no. (%)†

Disease severity APACHE II score‡ Mean

11±4

11±5

≥13§

32 (32)

29 (28)

Imrie or modified Glasgow score ¶ Median Range

2

2

0–6

0–5

C-reactive protein - mg/liter Median

70

75

21–179

11–189

SIRS - no. (%)◆

63 (62)

70 (67)

Respiratory failure - no. (%)

30 (30)

27 (26)

6 (6)

5 (5)

Interquartile range

Multiple organ failure - no. (%)** Duration - hr Onset of symptoms to presentation at the emergency department Median Interquartile range

12

13

5–28

4–33

Presentation at the emergency department to randomization Median Interquartile range

13

11

5–19

4–19

* Plus–minus values are means ±SD. There were no significant between-group differences at baseline, except for body- mass index (P = 0.01). † The body-mass index is the weight in kilograms divided by the square of the height in meters. ‡ Scores on the Acute Physiology and Chronic Health Evaluation (APACHE) II range from 0 to 71, with higher scores indicating more severe disease.34 § Patients with an APACHE II score of 13 or higher constituted a predefined subgroup. ¶ Imrie or modified Glasgow35 scores range from 0 to 8, with higher scores indicating more severe disease. ◆ The systemic inflammatory response syndrome (SIRS) was diagnosed with the use of the Consensus Conference criteria of the American College of Chest Physicians–Society of Critical Care Medicine. ** Organ failure was defined as a modified Marshall score of 2 or more (on a scale of 0 to 12, with higher scores indicating more severe disease), as proposed in the revised Atlanta classification of acute pancreatitis.45 Multiple organ failure was defined as failure of two or more organs on the same day.

100


TRIAL OF NUTRITION IN ACUTE PANCREATITIS

O utcomes Primary End Point The primary composite end point of major infection or death occurred in 30 patients (30%) in the early group, as compared with 28 (27%) in the on-demand group (risk ratio, 1.07; 95% confidence interval [CI], 0.79 to 1.44; absolute risk difference, 3 percentage points; 95% CI, â&#x2C6;&#x2019;9 to 15; P= 0.76). Major infections occurred in 25% of the patients in the early group and in 26% of those in the on-demand group (P= 0.87) (Table 2). Mortality was 11% in the early group, as compared with 7% in the on-demand group (P= 0.33), and most of the deaths were related to persistent multiple organ failure (defined as failure of two or more organs on â&#x2030;Ľ3 consecutive days). Secondary End Points Necrotizing pancreatitis developed in 63% of the patients in the early group and in 62% of those in the on-demand group. A total of 18% of the patients in the early group and 19% of those in the on-demand group required ICU admission (Table 2). In the on-demand group, 32 patients (31%) required nasoenteric tube feeding; 72 patients (69%) tolerated an oral diet and did not require tube feeding (Table 3). In 9 of these 32 patients (28%), tube feeding was prompted by the use of mechanical ventilation. The on-demand tubefeeding strategy reduced the number of days to full tolerance of an oral diet (9 days with the early strategy vs. 6 days with the on-demand strategy, P = 0.001). Gastrointestinal events occurred frequently, but the frequency did not differ significantly between the groups. Attenuation of the acute inflammatory response was hypothesized to be part of the beneficial effect of early feeding. However, such an effect did not occur (Fig. S3 in the Supplementary Appendix). In a predefined subgroup analysis restricted to patients with an APACHE II score of 13 or higher at randomization, the occurrence of the primary end point did not differ significantly between the two treatment groups (Table S3 in the Supplementary Appendix). Post hoc subgroup analyses also did not show a significant betweengroup difference in the primary end point for patients with SIRS at randomization or those with a BMI of less than 25 or 35 or more (Table S3 in the Supplementary Appendix). No significant differences were observed in healthcare utilization except for the number of tube placements (145 tube placements in the early group vs. 57 in the on-demand group, P<0.001) (Table S4 in the Supplementary Appendix).

101


102 12 (12) 11 (11) 64 (63)

Pneumonia

Death - no. (%)

Necrotizing pancreatitis - no. (%)‡

12 (12)

Mechanical ventilation - no. (%)

10/67 (15) 7/67 (10) 4/67 (6)

Persistent single organ failure

Multiple organ failure

Persistent multiple organ failure

4/73 (5)

6/73 (8)

10/73 (14)

31/73 (42)

14 (13)

20 (19)

4±3

65 (62)

7 (7)

13 (12)

18 (17)

15 (14)

27 (26)

28 (27)

On-Demand Tube Feeding (N = 104)

1.05 (0.51–2.14)

1.14 (0.67–1.95)

1.05 (0.65–1.70)

0.92 (0.65–1.32)

0.93 (0.60–1.44)

0.95 (0.66–1.38)

1.06 (0.77–1.47)

1.27 (0.85–1.89)

0.97 (0.63–1.50)

0.98 (0.68–1.43)

0.74 (0.43–1.26)

0.97 (0.70–1.34)

1.07 (0.79–1.44)

Risk Ratio (95% CI)

1.00

0.77

1.00

0.73

0.84

0.86

0.29

0.76

0.33

1.00

1.00

0.28

0.87

0.76

P Value

* Plus-minus values are means ±SD. Risk ratios are for early tube feeding as compared with on-demand tube feeding. ICU denotes intensive care unit. † Patients may have had more than one type of infection. ‡ Necrotizing pancreatitis was defined as pancreatic parenchymal necrosis or extrapancreatic necrosis.45,46 In nine patients (9%) in the early group and seven (7%) in the on-demand group, no CT was performed. § Scores on the CT severity index range from 0 to 10, with higher scores indicating more extensive pancreatic or extra- pancreatic necrosis. ¶ New-onset organ failure was defined as organ failure that was not present at randomization. Persistent organ failure was defined as organ failure present on 3 or more consecutive days (>48 hours). Multiple organ failure was defined as failure of two or more organs on the same day.

26/67 (39)

Single organ failure

New-onset organ failure - no./total no. at risk (%)¶

18 (18)

ICU admission after randomization - no. (%)

4±2

17 (17)

Bacteremia

CT severity index§

9 (9)

25 (25)

30 (30)

Infected pancreatic necrosis

Infection - no. (%)†

Secondary end points

Primary composite end point: infection or death - no. (%)

Outcome

Early Tube Feeding (N = 101)

Table 2. Primary and Secondary End Points, According to the Intention-to-Treat Analysis.*

PART II CHAPTER 6


21 (21)

Diarrhea

29 (28)

11 (11)

0.81 (0.57–1.17)

0.96 (0.60–1.54)

0.82 (0.57–1.20)

0.91 (0.68–1.24)

0.66 (0.32–1.37)

0.90 (0.62–1.30)

0.97 (0.70–1.33)

0.95 (0.77–1.16)

Risk Ratio (95% CI)

0.26

1.00

0.12

0.31

0.66

0.001

0.28

0.61

0.76

0.68

P Value

* NA denotes not applicable. † Dislodging or obstruction of the nasogastric tube was noted in case-record forms by the attending physician or nurse. The denominator is the number of patients who had a feeding tube inserted. Two patients in the early group declined tube feeding. ‡ The need for a nasogastric tube to be inserted for decompression or the need for parenteral nutrition was indicated by the attending physician. § Full tolerance of an oral diet was defined as tolerance of the oral diet without receipt of any other type of nasoenteric or parenteral nutrition. ¶ Gastrointestinal events were assessed during each day of the hospital stay. " Data are for suspected aspiration as noted by a physician or nurse in the case-record forms. ** Ileus was diagnosed by the attending physician and noted in the case-record forms.

10 (10)

Ileus**

0

4 (4)

26 (25)

19 (19)

Vomiting

Aspiration

5–10 37 (36)

6–12

6

10 (10)

5 (5) 9

23 (22)

4/32 (12)

11/99 (11) 19 (19)

14/32 (44)

32 (31)

On-Demand Tube Feeding (N = 104)

38/99 (38)

NA

Early Tube Feeding (N = 101)

32 (32)

Nausea

Gastrointestinal event - no. (%)¶

Interquartile range

Median

Days from admission to full tolerance of oral diet§

Need for parenteral nutrition - no. (%)‡

Need for insertion of nasogastric tube for decompression - no. (%)‡

Obstruction of nasoenteric feeding tube - no./total no. (%)†

Dislodging of nasoenteric feeding tube - no./total no. (%)†

Need for nasoenteric feeding tube - no. (%)

Nutrition variable

Outcome

Table 3. Nutrition Tolerance and Gastrointestinal Events.*

TRIAL OF NUTRITION IN ACUTE PANCREATITIS

103


PART II CHAPTER 6

DISCUSSION This multicenter, randomized trial involving patients with acute pancreatitis who were at high risk for complications did not show that an early start of nasoenteric tube feeding was superior to the introduction of an oral diet after 72 hours, with tube feeding only if required, in reducing the composite end point of major infection or death. With the oral diet and on-demand tubefeeding strategy, only approximately one third of patients required a nasojejunal feeding tube. The absolute between-group difference in the primary end point was 3 percentage points, with the 95% confidence interval ranging from 9 percentage points lower to 15 percentage points higher. These findings do not support clinical guidelines recommending the early start of nasoenteric tube feeding in all patients with severe acute pancreatitis in order to reduce the risks of infection and death. However, this trial was not powered to exclude a substantial benefit of early feeding. The results of our trial differ from those of previous trials and observational studies.12,23-26 Previous trials showed an improved outcome after early nasoenteric tube feeding as compared with total parenteral nutrition. This may be explained in part by complications associated with providing total parenteral nutrition, such as catheter-related infections.21 The negative outcome of our study, as compared with the outcomes in these previous trials, is not explained by differences in the timing of early tube feeding or the severity of pancreatitis in the study participants. The timing of early nasoenteric tube feeding in our study was similar to the timing in the previous studies. In addition, we used similar criteria for enrolling patients at high risk for complications, and we observed similar rates of major infection and death. Previous observational studies investigating the initiation of nasoenteric tube feeding within 48 hours after admission, as compared with initiation more than 48 hours after admission, cannot differentiate between cause and effect (i.e., less severely ill patients may have been fed earlier). This is in line with a recently revived debate on the presumed benefit of early enteral feeding in critically ill patients in general. Early enteral feeding is recommended in most current ICU guidelines.38,47 However, the methodologic quality of the trials that form the basis for these general ICU recommendations has been criticized.48 Thus, for critically ill patients in general and for those with acute pancreatitis specifically, large, high-quality, randomized, controlled trials that show an improved outcome with early enteral feeding are lacking.49 There are several possible explanations for the negative result of our study. First, early enteral feeding may not be as effective as we anticipated. Our hypothesis was that the trophic effect of early enteral feeding would stabilize the integrity of the gut mucosa, reducing inflammation and improving the outcome. Early enteral feeding was not associated with a reduction in any of the variables indicating inflammation (Fig. S3 in 104


TRIAL OF NUTRITION IN ACUTE PANCREATITIS

the Supplementary Appendix). We did not evaluate gut permeability and bacterial translocation on the basis of the serum intestinal fatty acidâ&#x20AC;&#x201C;binding protein level or endotoxin exposure.17,50 Therefore, we cannot determine whether gut permeability was influenced by early feeding in a subset of our patients. Increased gut permeability and bacterial translocation may be restricted to patients with acute pancreatitis who have multiple organ failure,14 a subgroup that accounted for only a small fraction of the patients in this trial. However, a study of acute pancreatitis in which the rates of multiple organ failure and death were similar to the rates in our study did show an increase in gut permeability and endotoxin exposure in most patients with severe acute pancreatitis.51 Another possibility is that tube feeding in the early group in our trial should have been started even earlier. In a trial involving a small number of patients at one center, it would be possible to start nasogastric tube feeding some hours earlier by using a feeding tube that could be placed at the bedside. In daily practice, however, we believe that an earlier start of tube feeding would not be feasible. Starting an oral diet later in the on-demand group in order to increase the difference in timing between the two study groups would not be ethical because it would put patients at risk for malnutrition. A third explanation for the negative result may be that the study was too small to detect a difference between the two groups. To our knowledge, this is the largest trial of nutrition in patients with acute pancreatitis that has been performed so far, but the wide confidence interval for the primary end point may indicate that an even larger trial is needed. Fourth, the widely accepted scoring systems for prediction of severity in acute pancreatitis are only moderately accurate.52 In early-intervention studies in acute pancreatitis, it is therefore unavoidable that mild or moderate disease will develop in a proportion of patients who were classified at presentation as having severe pancreatitis. Nevertheless, at randomization, approximately one third of our patients had organ failure and two thirds had SIRS. Organ failure is one of the determinants of severe pancreatitis, and SIRS is increasingly recognized as an early indicator of severe pancreatitis.30,45 A feeding tube frequently causes discomfort, excessive gagging, or esophagitis and is often dislodged or becomes obstructed, which necessitates the replacement of the feeding tube.53,54 If tube feeding were restricted to patients who could not tolerate an oral diet, this would result in substantial avoidance of discomfort and costs. In conclusion, our trial did not show the hypothesized benefit of early nasoenteric tube feeding in patients with acute pancreatitis who were at high risk for complications. The observation that the clinical outcomes of early tube feeding were similar to those of an oral diet initiated at 72 hours, with tube feeding only if required, challenges the concept of the gut mucosapreserving effect of early enteral feeding during acute pancreatitis.

105


PART II CHAPTER 6

Supported by grants from the Netherlands Organization for Health Research and Development, the ZonMw Healthcare Efficiency Research Program (170992902), and Nutricia (08/KR/AB/002). Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. We thank Vera Zeguers for assistance as a study research nurse, all the medical and nursing staff at the participating centers for assistance in the enrollment and care of the patients in this study, and the patients and their families for contributions to the study. APPENDIX The authorsâ&#x20AC;&#x2122; full names and academic degrees are as follows: Olaf J. Bakker, M.D., Sandra van Brunschot, M.D., Hjalmar C. van Santvoort, M.D., Ph.D., Marc G. Besselink, M.D., Ph.D., Thomas L. Bollen, M.D., Marja A. Boermeester, M.D., Ph.D., Cornelis H. Dejong, M.D., Ph.D., Harry van Goor, M.D., Ph.D., Koop Bosscha, M.D., Ph.D., Usama Ahmed Ali, M.D., Stefan Bouwense, M.D., Wilhelmina M. van Grevenstein, M.D., Ph.D., Joos Heisterkamp, M.D., Ph.D., Alexander P. Houdijk, M.D., Ph.D., Jeroen M. Jansen, M.D., Ph.D., Thom M. Karsten, M.D., Ph.D., Eric R. Manusama, M.D., Ph.D., Vincent B. Nieuwenhuijs, M.D., Ph.D., Alexander F. Schaapherder, M.D., Ph.D., George P. van der Schelling, M.D., Ph.D., Matthijs P. Schwartz, M.D., Ph.D., B.W. Marcel Spanier, M.D., Ph.D., Adriaan Tan, M.D., Ph.D., Juda Vecht, M.D., Ph.D., Bas L. Weusten, M.D., Ph.D., Ben J. Witteman, M.D., Ph.D., Louis M. Akkermans, M.D., Ph.D., Marco J. Bruno, M.D., Ph.D., Marcel G. Dijkgraaf, Ph.D., Bert van Ramshorst, M.D., Ph.D., and Hein G. Gooszen, M.D., Ph.D. The authorsâ&#x20AC;&#x2122; affiliations are as follows: the Department of Surgery, University Medical Center Utrecht, Utrecht (O.J.B., H.C.S., U.A.A., W.M.G., L.M.A.), the Departments of Gastroenterology and Hepatology (S.B.) and Surgery (M.G.B., M.A.B.) and the Clinical Research Unit (M.G.D.), Amsterdam Medical Center, and the Department of Gastroenterology, Onze Lieve Vrouwe Gasthuis (J.M.J.), Amsterdam, the Departments of Radiology (T.L.B.), Gastroenterology (B.L.W.), and Surgery (B.R.), St. Antonius Hospital Nieuwegein, Nieuwegein, the Department of Surgery, Maastricht University Medical Center and NUTRIM School for Nutrition, Toxicology, and Metabolism, Maastricht (C.H.D.), the Departments of Surgery (H.G., S.B.) and Operation Room/ Evidence-Based Surgery (H.G.G.), Radboud University Medical Center, and the Department of Gastroenterology, Canisius Wilhelmina Hospital (A.T.), Nijmegen, the Department of Surgery, Jeroen Bosch Hospital, Hertogenbosch (K.B.), the Department of Surgery, St. Elisabeth Hospital, Tilburg (J.H.), the Department of Surgery, Medical Center Alkmaar, Alkmaar (A.P.H.), the Department of Surgery, Reinier de Graaf Gasthuis, Delft (T.M.K.), the Department of Surgery, Medical Center Leeuwarden, Leeuwarden (E.R.M.), the Department of Surgery, University Medical Center Groningen, Groningen (V.B.N.), the Department of Surgery, Leiden University Medical Center, Leiden (A.F.S.), 106


TRIAL OF NUTRITION IN ACUTE PANCREATITIS

the Department of Surgery, Amphia Hospital, Breda (G.P.S.), the Department of Gastroenterology, Meander Medical Center, Amersfoort (M.P.S.), the Department of Gastroenterology, Rijnstate Hospital, Arnhem (B.W.M.S.), the Department of Gastroenterology, Isala Clinics, Zwolle (J.V.), the Department of Gastroenterology, Hospital Gelderse Vallei Ede, Ede (B.J.W.), and the Department of Gastroenterology and Hepatology, Erasmus University Medical Center, Rotterdam (M.J.B.) â&#x20AC;&#x201D; all in the Netherlands.

107


PART II CHAPTER 6

REFERENCES 1. Peery AF, Dellon ES, Lund J, et al. Burden of gastrointestinal disease in the United States: 2012 up d at e . G a s t ro e nt e rol o g y 2012;143:1179-87. 2. Fagenholz PJ, Castillo CF, Harris NS, Pelletier AJ, Camargo CA Jr. Increasing United States hospital admissions for acute pancreatitis, 1988-2003. Ann Epidemiol 2007;17:491-7. 3. Spanier B, Bruno MJ, Dijkgraaf MG. Incidence and mortality of acute and chronic pancreatitis in the Netherlands: a nationwide record-linked cohort study for the years 1995-2005. World J Gastroenterology 2013;19:3018-26. 4. Yadav D, Lowenfels AB. Trends in the epidemiology of the first attack of acute pancreatitis: a systematic review. Pancreas 2006;33:323-30. [Erratum, Pancreas 2007; 34:174.] 5. Singh VK, Bollen TL, Wu BU, et al. An assessment of the severity of interstitial pancreatitis. Clin Gastroenterol Hepatol 2011;9:1098-103. 6. Whitcomb DC. Acute pancreatitis. N Engl J Med 2006;354:2142-50. 7. Besselink MG, van Santvoort HC, Boermeester MA, et al. Timing and impact of infections in acute pancreatitis. Br J Surg 2009;96:267-73. 8. Wu BU, Johannes RS, Kurtz S, Banks PA. The impact of hospital-acquired infec- tion on outcome in acute pancreatitis. Gastroenterology 2008;135:816-20. 9. Rodriguez JR, Razo AO, Targarona J, et al. Debridement and closed packing for sterile or infected necrotizing pancreati- tis: insights into indications and outcomes in 167 patients. Ann Surg 2008;247:294-9. 10. van Santvoort HC, Bakker OJ, Bollen TL, et al. A conservative and minimally invasive approach to necrotizing pancreati- tis improves outcome. Gastroenterology 2011;141:1254-63.

108

11. van Santvoort HC, Besselink MG, Bakker OJ, et al. A step-up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med 2010;362:1491-502. 12. Al-Omran M, Albalawi ZH, Tashkandi MF, Al-Ansary LA. Enteral versus parenteral nutrition for acute pancreatitis. Cochrane Database Syst Rev 2010;1: CD002837. 13. Ammori BJ, Leeder PC, King RF, et al. Early increase in intestinal permeability in patients with severe acute pancreatitis: correlation with endotoxemia, organ failure, and mortality. J Gastrointest Surg 1999;3:252-62. 14. Besselink MG, van Santvoort HC, Renooij W, et al. Intestinal barrier dysfunction in a randomized trial of a specific probiotic composition in acute pancreatitis. Ann Surg 2009;250:712-9. 15. Fritz S, Hackert T, Hartwig W, et al. Bacterial translocation and infected pancreatic necrosis in acute necrotizing pancreatitis derives from small bowel rather than from colon. Am J Surg 2010;200: 111-7. 16. Nieuwenhuijs VB, Verheem A, van Duijvenbode-Beumer H, et al. The role of interdigestive small bowel motility in the regulation of gut microflora, bacterial overgrowth, and bacterial translocation in rats. Ann Surg 1998;228:188-93. 17. Rahman SH, Ammori BJ, Holmfield J, Larvin M, McMahon MJ. Intestinal hypoperfusion contributes to gut barrier failure in severe acute pancreatitis. J Gastrointest Surg 2003;7:26-35. 18. Van Felius ID, Akkermans LM, Boss- cha K, et al. Interdigestive small bowel motility and duodenal bacterial overgrowth in experimental acute pancreatitis. Neurogastroenterol Motil 2003;15:267- 76. 19. Marik PE. What is the best way to feed patients with pancreatitis? Curr Opin Crit Care 2009;15:131-8. 20. McClave SA, Heyland DK. The physiologic response and associated clinical benefits from provision of early enteral nutrition. Nutr Clin Pract 2009;24:305-15.


TRIAL OF NUTRITION IN ACUTE PANCREATITIS

21. Ziegler TR. Parenteral nutrition in the critically ill patient. N Engl J Med 2009; 361:1088-97. 22. Marik PE, Zaloga GP. Early enteral nutrition in acutely ill patients: a systematic review. Crit Care Med 2001;29:2264-70. [Erratum, Crit Care Med 2002;30:725.] 23. Li JY, Yu T, Chen GC, et al. Enteral nutrition within 48 hours of admission improves clinical outcomes of acute pancreatitis by reducing complications: a metaanalysis. PLoS One 2013;8(6):e64926. 24. Petrov MS, Pylypchuk RD, Uchugina AF. A systematic review on the timing of artificial nutrition in acute pancreatitis. Br J Nutr 2009;101:787-93. 25. Sun JK, Mu XW, Li WQ, Tong ZH, Li J, Zheng SY. Effects of early enteral nutrition on immune function of severe acute pancreatitis patients. World J Gastroenterol 2013;19:917-22. 26. Wereszczynska-Siemiatkowska U, SwidnickaSiergiejko A, Siemiatkowski A, Dabrowski A. Early enteral nutrition is superior to delayed enteral nutrition for the prevention of infected necrosis and mortality in acute pancreatitis. Pancreas 2013;42:640-6. 27. McClave SA, Martindale RG, Vanek VW, et al. Guidelines for the provision and assessment of nutrition support therapy in the adult critically ill patient: Society of Critical Care Medicine (SCCM) and American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.). JPEN J Parenter Enteral Nutr 2009;33:277-316. 28. Meier R, Ockenga J, Pertkiewicz M, et al. ESPEN guidelines on enteral nutrition: pancreas. Clin Nutr 2006;25:275-84. 29. Mirtallo JM, Forbes A, McClave SA, Jensen GL, Waitzberg DL, Davies AR. Inter- national consensus guidelines for nutrition therapy in pancreatitis. JPEN J Parenter Enteral Nutr 2012;36:284-91. 30. IAP/APA evidence-based guidelines for the management of acute pancreatitis. Pancreatology 2013;13:Suppl 2:e1-e15.

31. Tenner S, Baillie J, DeWitt J, Vege SS. American College of Gastroenterology guideline: management of acute pancreatitis. Am J Gastroenterol 2013;108:1400-15. 32. Banks PA, Freeman MLy. Practice guidelines in acute pancreatitis. Am J Gastroenterol 2006;101:2379-400. 33. Bakker OJ, van Santvoort HC, van Brunschot S, et al. Pancreatitis, very early compared with normal start of enteral feeding (PYTHON trial): design and ratio- nale of a randomised controlled multicenter trial. Trials 2011;12:73. 34. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818-29. 35. Corfield AP, Cooper MJ, Williamson RC, et al. Prediction of severity in acute pancreatitis: prospective comparison of three prognostic indices. Lancet 1985;2: 403-7. 36. Neoptolemos JP, Kemppainen EA, Mayer JM, et al. Early prediction of sever- ity in acute pancreatitis by urinary trypsinogen activation peptide: a multicentre study. Lancet 2000;355:1955-60. 37. Dickerson RN, Vehe KL, Mullen JL, Feurer ID. Resting energy expenditure in patients with pancreatitis. Crit Care Med 1991;19:484-90. 38. Kreymann KG, Berger MM, Deutz NE, et al. ESPEN guidelines on enteral nutrition: intensive care. Clin Nutr 2006;25: 210-23. 39. Besselink MG, van Santvoort HC, Buskens E, et al. Probiotic prophylaxis in predicted severe acute pancreatitis: a randomised, double-blind, placebocontrolled trial. Lancet 2008;371:651-9. [Erratum, Lancet 2008;371:1246.] 40. Gupta R, Patel K, Calder PC, Yaqoob P, Primrose JN, Johnson CD. A randomised clinical trial to assess the effect of total enteral and total parenteral nutritional support on metabolic, inflammatory and oxidative markers in patients with pre- dicted severe acute pancreatitis (APACHE II > or =6). Pancreatology 2003;3:406-13.

109


PART II CHAPTER 6

41. Oláh A, Pardavi G, Belágyi T, Nagy A, Issekutz A, Mohamed GE. Early nasojejunal feeding in acute pancreatitis is associated with a lower complication rate. Nutrition 2002;18:259-62. 42. Petrov MS, Kukosh MV, Emelyanov NV. A randomized controlled trial of enteral versus parenteral feeding in patients with predicted severe acute pancreatitis shows a significant reduction in mortality and in infected pancreatic complications with total enteral nutrition. Dig Surg 2006; 23:336-44. 43. Singh VK, Wu BU, Bollen TL, et al. Early systemic inflammatory response syndrome is associated with severe acute pancreatitis. Clin Gastroenterol Hepatol 2009;7:1247-51. 44. Chan AW, Tetzlaff JM, Gøtzsche PC, et al. SPIRIT 2013 explanation and elaboration: guidance for protocols of clinical trials. BMJ 2013;346:e7586. 45. Banks PA, Bollen TL, Dervenis C, et al. Classification of acute pancreatitis — 2012: revision of the Atlanta classification and definitions by international consensus. Gut 2013;62:102-11. 46. Bakker OJ, van Santvoort H, Besselink MG, et al. Extrapancreatic necrosis without pancreatic parenchymal necrosis: a separate entity in necrotising pancreatitis? Gut 2013;62:1475-80. 47. Heyland DK, Dhaliwal R, Drover JW, Gramlich L, Dodek P. Canadian clinical practice guidelines for nutrition support in mechanically ventilated, critically ill adult patients. JPEN J Parenter Enteral Nutr 2003;27:355-73. 48. Koretz RL, Lipman TO. The presence and effect of bias in trials of early enteral nutrition in critical care. Clin Nutr 2014; 33:240-5. 49. Casaer MP, Van den Berghe G. Nutri- tion in the acute phase of critical illness. N Engl J Med 2014;370:1227-36. 50. Pan L, Wang X, Li W, Li N, Li J. The intestinal fatty acid binding protein diag- nosing gut dysfunction in acute pancreatitis: a pilot study. Pancreas 2010;39:633-8.

110

51. Rahman SH, Ammori BJ, Larvin M, McMahon MJ. Increased nitric oxide excretion in patients with severe acute pancreatitis: evidence of an endotoxin mediated inflammatory response? Gut 2003; 52:270-4. 52. Mounzer R, Langmead CJ, Wu BU, et al. Comparison of existing clinical scoring systems to predict persistent organ failure in patients with acute pancreatitis. Gastroenterology 2012;142:1476-82. 53. Jones BJ. Enteral feeding: techniques of administration. Gut 1986;27:Suppl 1: 47-50. 54. Marik PE. Aspiration pneumonitis and aspiration pneumonia. N Engl J Med 2001;344:665-71.


SUPPLEMENTARY APPENDIX Early nasoenteric versus on-demand tube feeding in acute pancreatitis

Olaf J Bakker, Sandra van Brunschot, Hjalmar C van Santvoort, Marc G Besselink, Thomas L Bollen, Marja A Boermeester, Cornelis H Dejong, Harry van Goor, Koop Bosscha, Usama Ahmed Ali, Stefan Bouwense, Wilhelmina M van Grevenstein, Joos Heisterkamp, Alexander P Houdijk, Jeroen M Jansen, Thom M Karsten, Eric R Manusama, Vincent B Nieuwenhuijs, Alexander F Schaapherder, George P van der Schelling, Matthijs P Schwartz, BW Marcel Spanier, Adriaan Tan, Juda Vecht, Bas L Weusten, Ben J Witteman, Louis M Akkermans, Marco J Bruno, Marcel G Dijkgraaf, Bert van Ramshorst and Hein G Gooszen for the Dutch Pancreatitis Study Group 111


PART II CHAPTER 6

A DDI T IONA L I N F OR M AT ION - Exclusion criteria - Feeding regimen and general supportive treatment - End points - Data collection and end point assessment - Patient safety - Statistical analysis - Interim analysis - IBox S1. Definitions of infections - Figure S1. Trial Enrollment, Randomization and Follow-up - Figure S2. Amount of calories delivered (% of energy target) - Table S1. Time to start of feeding - Table S2. Baseline characteristics of patients â&#x20AC;&#x201C; additional information - Table S3. Results of subgroup analyses for primary end point - Figure S3. APACHE II score, C-reactive protein and SIRS during first week - Table S4. Healthcare utilization - Figure S4. Abdominal pain during the first week after admission - Table S5. Adverse events E xclusi on criteri a Excluded were patients with recurrent acute or chronic pancreatitis, pancreatitis due to endoscopic retrograde cholangiopancreaticography or malignancy, patients with enteral or parenteral nutrition at home, pregnant patients, patients assessed (or transferred from other hospitals) more than 24 hours after presentation to the emergency department, or patients presenting to the emergency department more than 96 hours after symptom onset. Fe e ding reg i men Early Nasoenteric Feeding If assigned to early nasoenteric feeding, patients received a nasojejunal feeding tube as soon as possible but not later than 24 hours following randomization. Feeding tubes were placed endoscopically or radiologically according to local practice. After tube placement nasoenteric feeding was started at 20 ml per hour during the first 24 hours following randomization. After 24 hours the volume of nutrition was increased to 45 ml per hour, after 48 hours to 65 ml per hour and after 72 hours to full nutrition depending on patientâ&#x20AC;&#x2122;s actual body weight. Nasoenteric feeding was administered as Nutrison protein enriched (Nutricia, Zoetermeer, The Netherlands). Per 100 ml this provided 125 kcal, 6.3 g protein, 4.9 g fat and 14.2 g carbohydrate. Standard amounts of minerals, vitamins and trace elements were included. The solutions osmolarity was 290 mOsmol/l and the osmolality 112


SUPPLEMENTARY APPENDIX PYTHON TRIAL

was 355 mOsmol/kg H20. For both study groups, full nutrition was defined as an energy target of 25 kcal/kg/day for patients in the intensive care unit and 30 kcal/kg/day for patients in the ward.28, 37 At 3 and 7 days after admission a dietician assessed nutritional status and nutritional requirements and made adjustments accordingly. The position of the feeding tube was not checked routinely but only in case of suspected dislodging. A plain abdominal radiograph was performed in case of nausea or vomiting, lowered consciousness in non-intubated patients, or gastric residual volumes exceeding 250 ml per 6 hours in patients with a nasogastric decompression tube. In patients with symptoms of delayed intestinal passage, the nutritional dose was decreased by 50% and gradually increased again the next day or stopped completely if symptoms progressed to a distended abdomen with an abdominal X-ray that showed dilated small and large bowel. When a patient was judged to be able to tolerate an oral diet, tube feeding was gradually decreased and replaced by an oral diet. If pain relapsed during start of an oral diet, tube feeding was restarted. On-Demand Tube Feeding In the on demand tube feeding group, patients were kept â&#x20AC;&#x2DC;nil by mouthâ&#x20AC;&#x2122; without any artificial nutrition during the first 72 hours after presentation to the emergency ward. Exceptions were made for patients who requested oral food within these 72 hours (exceptions were infrequent with only 5% requesting earlier feeding). These patients were offered liquids or solid food on request. At 72 hours all patients in the on demand group were offered an oral diet. If this was not tolerated, it was offered again after 24 hours. If it still was not tolerated after 96 hours from admission, nasoenteric feeding was started after placement of a nasojejunal tube and the same procedure was followed for nasoenteric tube feeding as in the early feeding group. If patients were transferred to the intensive care unit (ICU), a nasogastric or nasojejunal tube was placed and feeding was started with the same regimen as in the early feeding group. G ener a l supp or tive tre atment Fluid resuscitation was commenced at the emergency department and continued throughout the first days. The amount of fluid administered was titrated using vital signs and serum markers. Antibiotics were administered based on culture results and were not given as prophylaxis in patients with necrotizing pancreatitis. A contrast enhanced abdominal CT was performed 5 to 7 days after admission in all patients to evaluate the presence of local complications. In patients with infected pancreatic necrosis, interventions were generally postponed until the acute necrotic collections had progressed to walled-off necrosis. Patients were discharged from the hospital if this was deemed feasible by the treating physician. In general, patients were discharged after resolution of major abdominal pain, 113


PART II CHAPTER 6

normalization of inflammatory parameters such as serum C-reactive protein and when oral intake reached energy targets. E nd p oints The primary end point was a composite of major infections or death within 6 months from randomization. Major infections were defined as infected pancreatic necrosis, bacteremia, or pneumonia (for definitions see Box 1 in the supplementary appendix). Predefined secondary end points included the development of necrotizing pancreatitis as diagnosed on CT, and new onset organ failure. APACHE II scores, C-reactive protein levels and the presence of the Systemic Inflammatory Response Syndrome (SIRS) measured daily from admission up to day 7, and healthcare utilization. Secondary end points also included gastrointestinal symptoms and tolerance of early nasoenteric or oral feeding, such as the number of patients requiring requiring a nasoenteric feeding tube in the on demand tube feeding group. Full tolerance of an oral diet was defined as consuming an oral diet that reaches energy targets without clinically relevant gastrointestinal symptoms as judged by the treating physician. D ata col l e c ti on and end p oint ass essment A case-record form was filled in by local dieticians who registered the caloric intake and calculated caloric energy targets during the first week after admission based on actual body weight. For patients with nasoenteric tube feeding the caloric intake in kcal was deducted from the amount of feeding. For patients with an oral diet, the dieticians registered the type and quantity of hospital food tolerated and calculated the amount of kcal provided per patient. Routine follow-up visits took place 3 and 6 months after discharge. Data collection was performed by local physicians using a paper case-record form. The trial nurse, who was not involved in patient care, repeatedly motivated physicians to fill in all forms and collected the forms . A special case-record form was filled in by local dieticians who registered the caloric intake and calculated caloric energy targets during the first week after admission. For patients with nasoenteric tube feeding the caloric intake in kcal was deducted from the amount of Nutrison administered. For patients with an oral diet, the dieticians registered the type and quantity of hospital food offered and tolerated and subsequently calculated the amount of kcal provided per patient. Possible reasons for discontinuation of nasoenteric tube or oral feeding such as vomiting were noted and accounted for when calculating the caloric intake.

114


SUPPLEMENTARY APPENDIX PYTHON TRIAL

Pati ent s afety To optimize patient safety an independent data and safety monitoring committee (DSMC) evaluated the progress of the trial and examined safety end points following every 25 patients who completed follow-up. All involved physicians were repetitively asked to report any potential adverse events. These adverse events were listed and presented to the DSMC in an unblinded fashion. The DSMC discussed the implications of the data presented. In addition, all deceased patients were extensively evaluated by the DSMC for cause of death and possible intervention related serious adverse events. The outcome of the meeting of the DSMC was discussed with the trial steering committee and was reported to the responsible IRB. All adverse events were reported to the Dutch Central Committee on Research involving Human Subjects and the IRB. Statisti c a l ana lysis The expected incidence of the primary end point in the on demand tube feeding strategy was based on the individual patient data from the placebo arm of a previous randomized trial using the same inclusion criteria as the current study. To estimate the expected incidence of the primary end point in early nasoenteric feeding group, results from patients who received early nasoenteric tube feeding in randomized trials comparing nasoenteric to parenteral nutrition were used. The sample-size calculation was based on an expected reduction in the primary composite end point from 40 to 22%. With at least 80% power, a two-sided alpha level of 5% and 1% loss to follow-up, the sample size was set at 208 patients. Analysis was based on intention-to-treat, only excluding patients in whom the adjudication committee, blinded for treatment allocation and prior to any analysis, decided that the diagnosis acute pancreatitis was incorrect. Variables are summarized as frequencies and percentages, means with standard deviations or 95% confidence intervals, or medians and interquartile ranges, as appropriate. Results are presented as risk ratios with corresponding 95% confidence intervals. Dichotomous data were compared with the use of Fisherâ&#x20AC;&#x2122;s exact test, continuous data with the MannWhitney U test, and categorical data with the linear-by-linear association test. The variables used as stratification factors were not included in the analyses. Predefined subgroups were patients with an APACHE II below 13 or an APACHE II of 13 or higher at randomization. Post hoc subgroup analyses were performed for patients with SIRS at randomization as recent insights show a high risk of complications in these patients, and patients with a body mass index below 25 or above 35 because of differences at baseline (Table 1 manuscript). An interim-analysis was performed on the primary end point after 50% of patients had completed 6 months follow-up. The interim-analysis was performed by an independent statistician blinded for treatment allocation applying the Peto approach with symmetric stopping boundaries at P<0.001. 115


PART II CHAPTER 6

For final analyses, a two-sided P value of less than 0.05 was considered to indicate statistical significance. P values were not adjusted for multiple testing. Interim ana lysis Planned interim analysis was performed after all 104 patients had completed 6 months follow-up. At that time, 142 patients were randomized. Although P values for stopping were not reached, interim results showed a trend towards increased mortality in the early nasoenteric feeding group. The DSMC unanimously recommended the steering committee to pause recruitment to allow for additional safety analyses. These analyses showed a higher rate of baseline organ failure in the early nasoenteric feeding group. After a joint meeting the steering committee and DSMC decided to resume recruitment anticipating the difference in baseline organ failure to disappear.

Box S1. Definitions of infections used for primary end point.

116

Infection

Definition

Infected pancreatic necrosis

Positive culture of pancreatic or extrapancreatic necrotic tissue obtained with fine-needle aspiration or from the first drainage procedure or operation, or the presence of gas in the fluid collection on contrast-enhanced CT.

Bacteremia

Positive blood culture. For non-pathogens (e.g. coagulase negative staphylococci) at least 2 samples had to be positive.

Pneumonia

Coughing or dyspnoea, radiography with infiltrative abnormalities, raised inflammatory variables and positive sputum culture. For intubated patients a positive endotracheal culture was mandatory.


SUPPLEMENTARY APPENDIX PYTHON TRIAL

Figure S1. Trial Enrollment, Randomization and Follow-up.

117


PART II CHAPTER 6

Figure S2. Amount of calories delivered (% of energy target)

Table S1. Time to start of Early Nasoenteric Feeding or On-Demand Tube Feeding Time to start â&#x20AC;&#x201C; hrs* From randomization

On-Demand Tube Feeding (N = 104)

P value

8 (2-20)

64 (50-69)

<0.001

From presentation to the emergency ward

23 (18-27)

72 (69-79)

<0.001

From onset of symptoms

41 (27-54)

91 (77-117)

<0.001

* Values are medians and interquartile ranges.

118

Early Nasoenteric Tube Feeding (N = 101)


SUPPLEMENTARY APPENDIX PYTHON TRIAL

Table S2. Baseline Characteristics of the Patients – additional information^ Early Nasoenteric Tube Feeding (N = 101)

On-Demand Tube Feeding (N = 104)

Cardiovascular disease

62 (62)

51 (49)

Pulmonary disease

14 (14)

16 (15)

Characteristic Coexisting condition – no. (%)

Chronic renal insufficiency

3 (3)

0 (0)

Diabetes

17 (17)

20 (19)

Other systemic disease

26 (26)

25 (24)

I: healthy status

24 (24)

26 (25)

II: mild systemic disease

59 (58)

55 (53)

III: severe systemic disease

16 (16)

22 (21)

2 (2)

1 (1)

22±12

21±12

8 (8)

9 (9)

ASA class on admission – no. (%)‡

IV: severe systemic disease constant threat to life Disease severity Blood Urea Nitrogen – mg/dL Renal failure – no. (%) Cardiovascular failure – no. (%)

2 (2)

1 (1)

Single-organ failure – no. (%)††

34 (34)

31 (30)

^ Plus-minus values represent means±SD. ‡ ASA denotes American Society of Anesthesiologists. †† Organ failure was defined as a modified Marshall score of 2 or more as proposed in the revised Atlanta classification of acute pancreatitis.

Table S3. Results of subgroup analyses for primary end point* Early Nasoenteric Tube Feeding

On-Demand Tube Feeding

Risk Ratio (95% CI)

P value

14 (44)

14 (48)

0.92 (0.57 – 1.49)

0.80

SIRS at randomization (N = 133)

23 (37)

25 (36)

1.02 (0.70 – 1.48)

1.00

BMI below 25 or above 35 (N = 66)

10 (33)

6 (17)

1.56 (0.94 – 2.60)

0.15

Subgroup – No. (%) Predefined APACHE II 13 or higher (N = 61) Post hoc

*Data are number and percentages. Tests for interaction were not significant (P>0.05).

119


PART II CHAPTER 6

A

B

C

Figure S3. Inflammatory response during the first week after admission as depicted by daily APACHE II scores (A), serum C-reactive protein levels (B) and presence of SIRS (C). For Figures A and B, the boxes represent 95% confidence intervals and the horizontal lines within the boxes represent the means. For Figure C the height of the boxes represents the number of patients.

120


SUPPLEMENTARY APPENDIX PYTHON TRIAL

Table S4. Healthcare utilization Outcome Days in hospital# Days in ICU##

Early Nasoenteric Tube Feeding (N = 101)

On-Demand Tube Feeding (N = 104)

P value^

15 (10 – 22)

14 (9 – 25)

0.98

5 (3 – 14)

6 (3 – 23)

0.70

Interventions - Total no. (range per patient) Nasojejunal tube placements‡

145 (0 – 6)

57 (0 – 11)

<0.001

ERCP*

37 (0 – 3)

50 (0 – 6)

0.76

Percutaneous catheter drainage**

24 (0 – 13)

46 (0 – 15)

0.13

8 (0 – 7)

6 (0 – 2)

0.43

Endoscopic transgastric drainage or necrosectomy** Surgical necrosectomy**

3 (0 – 1)

7 (0 – 2)

0.49

Other interventions†

39 (0 – 3)

49 (0 – 3)

0.58

X-ray (chest or plain abdominal)

449 (0 – 39)

592 (0 – 69)

0.84

CT

232 (0 – 16)

283 (0 – 19)

0.38

Radiology - Total no. (range per patient)

17 (0 – 2)

16 (0 – 2)

0.65

Microbiology cultures

MRCP or MRI

2 (0 – 6)

3 (0 – 7)

0.49

Visits to outpatient clinic

2 (1 – 6)

2 (1 – 4)

0.39

Visits to general practitioner

2 (1 – 4)

2 (1 – 4)

0.30

226 (0 – 40)

299 (0 – 93)

0.85

Visits to physiotherapist – Total no. (range per patient)

Continuous data are median and interquartile ranges (IQR) or total number per study group and range per patient when indicated. ICU denotes intensive care unit. ERCP denotes Endoscopic Retrograde Cholangiopancreaticography. CT denotes Computed Tomography. MRCP denotes Magnetic Resonance Cholangiopancreaticography. MRI denotes Magnetic Resonance Imaging. # Including readmissions. ## For patients admitted in the ICU for at least 24 hours. ‡ Endoscopically or radiologically and including replacements after dislodging or obstruction. * Endoscopic retrograde cholangiopancreaticography. ** For necrotizing pancreatitis. † For example cholecystectomy for biliary pancreatitis. ^ P value calculated with Mann-Whitney U test.

121


PART II CHAPTER 6

Figure S4. Abdominal pain during the first week after admission. Abdominal pain measured daily during the first week on a scale from 0 to 10, with 10 meaning the most severe pain imaginable, i.e. the Numeric Rating Scale.

Table S5. Adverse Events other than primary and secondary endpoints* Adverse Events

Early Nasoenteric Tube Feeding (N = 101)

On-Demand Tube Feeding (N = 104)

2

1

Gastro-intestinal Bowel ischemia

122

Abdominal compartment syndrome

0

1

Peritonitis

0

1

Perforation of a hollow viscus

1

2

Entero-cutaneous fistula

1

0

Volvulus

1

0

Acute cholecystitis or cholangitis

8

7

Bile duct injury

1

0

Bleeding after ERCP#

1

0

Dislocated percutaneous transhepatic biliary catheter

0

1

Dislocated percutaneous gallbladder catheter

0

1

Suspected gastrointestinal blood loss

1

7

Colitis

1

0

Ascites

2

0


SUPPLEMENTARY APPENDIX PYTHON TRIAL

Table S5. Continued Adverse Events Clostridium infection

Early Nasoenteric Tube Feeding (N = 101)

On-Demand Tube Feeding (N = 104)

0

2

0

2

Cardiovascular Acute Coronary Syndrome Atrial Fibrillation

3

11

Myocardial Infarction

2

3

Congestive heart failure

10

5

Pericarditis

2

0

Portal thrombosis

0

2

Malignant hypertension

2

2

Postoperative bleeding

0

1

2

0

Pulmonary Pulmonary embolus Exacerbation of chronic obstructive pulmonary disease

1

0

Pleural fluid requiring percutaneous catheter drainage

1

1

0

1

Neurologic Brain herniation Delerium

3

5

Headache

1

0

Peroneal nerve palsy

0

1

Critical illness polyneuropathy

1

3

Urinary tract Urinary Tract Infection

10

9

Pyelonephritis

1

0

Hematuria

0

1

Other Gout

1

0

Infection of epidural catheter

1

1

Surgical site infection

1

0

Hyperkalemia

2

0

Hypokalemia

3

3

Hyponatremia

2

0

Hyperglycemia

1

4

Exocrine pancreatic insufficiency

0

2

* Adverse events as noted in case record forms by attending physicians and reported to the Data and Safety Monitoring Board. These adverse events were not predefined in the study protocol.

123


PART II CHAPTER 6

Memb ers of the Tri a l Ste ering C ommitte e OJ Bakker, HC van Santvoort, MGH Besselink, U Ahmed Ali, LMA Akkermans, University Medical Center Utrecht; MA Boermeester, Academic Medical Center Amsterdam; CHC Dejong, Maastricht University Medical Center; H van Goor, S van Brunschot, Radboud University Medical Center; AFM Schaapherder, Leiden University Medical Center; VB Nieuwenhuijs, University Medical Center Groningen: MA Brink, Meander Medical Center Amersfoort; B van Ramshorst, BLAM Weusten, TL Bollen, St. Antonius Hospital Nieuwegein; BJM Witteman, Hospital Gelderse Vallei Ede; HG Gooszen (chairman), Radboud University Medical Center. Memb ers of the D ata and S afety Monitoring B o ard E Buskens, epidemiologist, University Medical Center Groningen (chairman); PMNYH Go, surgeon, St. Antonius Hospital Nieuwegein; DA Legemate, surgeon and clinical epidemiologist, Academic Medical Center Amsterdam; EM Mathus-Vliegen, Gastroenterologist, Academic Medical Center Amsterdam; PD Siersema, gastroenterologist, University Medical Center Utrecht. Ind ep end ent statisti ci ans GF Borm and R Donders, statisticians, Radboud University Medical Center, Nijmegen. Memb ers of the Adju di c ati on C ommitte e MA Boermeester, Academic Medical Center Amsterdam; CHC Dejong, Maastricht University Medical Center; H van Goor, Radboud University Medical Center Nijmegen, B van Ramshorst, St. Antonius Hospital Nieuwegein; BJM Witteman, Hospital Gelderse Vallei Ede. Authors' contributi ons OJB, HCvS, MGB, MAB, CHD, HvG, BvR, BJW, UAA, HGG and several other members of the study group participated in the design of the study. OJB, SvB and TLB collected the data. OJB performed the statistical analyses apart from the primary end point which was performed by the independent statisticians. OJB drafted the first and subsequent versions of the manuscript. All authors read and approved the final manuscript. HGG supervised the current study. The authors vouch for the accuracy and completeness of the data and analyses.

124


125


PART II PREVENTING COMPLICATED PANCREATITIS


CHAPTER 7 Timing of cholecystectomy after mild biliary pancreatitis British Journal of Surgery 2011

Olaf J Bakker, Hjalmar C van Santvoort, Julia C Hagenaars, Marc G Besselink, Thomas L Bollen, Hein G Gooszen and Alexander F Schaapherder for the Dutch Pancreatitis Study Group


PART II CHAPTER 7

A B S T R AC T B ackg round The aim of the study was to evaluate recurrent biliary events as a consequence of delay in cholecystectomy following mild biliary pancreatitis. Metho ds Between 2004 and 2007, 731 patients with acute pancreatitis were registered prospectively in 15 Dutch hospitals. Of 308 patients with mild biliary pancreatitis, 267 were candidates for cholecystectomy. Recurrent biliary events requiring admission before and after cholecystectomy, and after endoscopic sphincterotomy (ES), were evaluated. Resu lts Eighteen patients underwent cholecystectomy during the initial admission, leaving 249 potential candidates for cholecystectomy after discharge. Cholecystectomy was performed after a median of 6 weeks in 188 patients (75.5 per cent). Before cholecystectomy, 34 patients (13.7 per cent) were readmitted for biliary events, including 24 with recurrent biliary pancreatitis. ES was performed in 108 patients during the initial admission. Eight (7.4 per cent) of these patients suffered from biliary events after ES and before cholecystectomy, compared with 26 (18.4 per cent) of 141 patients who did not have ES (risk ratio 0.51, 95 per cent confidence interval 0.27 to 0.94; P = 0.015). Following cholecystectomy, 8 (3.9 per cent) of 206 patients developed biliary events after a median of 31 weeks. Only 142 (53.2 per cent) of 267 patients were treated in accordance with the Dutch guideline, which recommends cholecystectomy or ES during the index admission or within 3 weeks thereafter. C onclusi on A delay in cholecystectomy after mild biliary pancreatitis carries a substantial risk of recurrent biliary events. ES reduces the risk of recurrent pancreatitis but not of other biliary events.

128


TIMING OF CHOLECYSTECTOMY IN BILIARY PANCREATITIS

I N T RODU C T ION Small gallstones and sludge are the most common cause of acute pancreatitis in the Western world.1,2 Following an episode of biliary pancreatitis, patients may develop recurrent biliary events, such as recurrence of biliary pancreatitis, acute cholecystitis, cholangitis or biliary colic. To prevent these, international treatment guidelines advise cholecystectomy or endoscopic sphincterotomy (ES) following an episode of acute biliary pancreatitis.3–5 Failure to provide definitive treatment may potentially put a patient at risk of a lethal episode of acute pancreatitis.6 The appropriate timing of cholecystectomy in acute pancreatitis depends on the clinical course. After severe pancreatitis, with local complications such as pancreatic necrosis and systemic complications including respiratory failure, it is advisable to delay cholecystectomy until such complications have resolved.7 After mild pancreatitis, without local and systemic complications, the optimal timing of cholecystectomy remains unclear. The aim is to minimize the risk of recurrent biliary events, but published guidelines differ with regard to the timing of cholecystectomy: during the initial admission or within 2, 3 or 4 weeks after discharge.3–5,8,9 ES early in the course of mild biliary pancreatitis is indicated only for clearance of bile duct stones in the presence of obstructive jaundice or suspected cholangitis.3–5,8 After recovery from pancreatitis, ES can be performed as an alternative to cholecystectomy in patients unfit for surgery, such as those with significant co-morbidity.3–5 In patients with mild biliary pancreatitis who are fit for surgery, but have already undergone urgent ES for obstructive jaundice or suspected cholangitis, the use of cholecystectomy to prevent recurrent biliary events is debated. Some authors argue that, as ES is highly effective in preventing recurrent pancreatitis, cholecystectomy is not necessary in these patients.10,11 Others draw a parallel with the situation in symptomatic gallstone disease, where a Cochrane review has clearly shown that cholecystectomy should always be performed after ES.12 The primary aim of this study was to estimate the risk of recurrent biliary events occurring in the time between discharge after mild biliary pancreatitis and admission for cholecystectomy in a prospective multicentre cohort. Secondary aims were to evaluate the influence of ES on the recurrence of biliary events and to determine the incidence of recurrent biliary events after cholecystectomy.

M E T HOD S This was a retrospective observational study of patients included in a prospective database. The study population consisted of adult patients with mild biliary pancreatitis. They were identified from a prospective multicentre cohort with a primary attack of acute 129


PART II CHAPTER 7

pancreatitis, and were included between March 2004 and March 2007. Patients were potentially eligible for a randomized trial on probiotic prophylaxis and were prospectively registered in one of 15 hospitals of the Dutch Pancreatitis Study Group.13,14 The study was reported in accordance with the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines.15 All patients gave written informed consent to participate and the ethics review board of each participating hospital approved the study. In general, patients were treated with adequate fluid resuscitation and analgesics. They underwent abdominal ultrasonography and routine laboratory investigations during the first 3 days of admission and thereafter if necessary. Computed tomography (CT) was performed if there was no clinical improvement within a week of admission and thereafter on clinical indication. Patients were discharged from hospital once oral diet was tolerated, inflammatory parameters had normalized, and after sufficient clinical improvement as judged by the treating physician. Tre atment of bi li ar y p ancre atitis The responsible physician decided if and when a cholecystectomy was necessary and whether the patient was fit for surgery. The study protocol did not provide guidelines on when to perform cholecystectomy, although the Dutch acute pancreatitis guideline published in 2005 advises cholecystectomy or ES in patients with mild biliary pancreatitis during admission or within 3 weeks thereafter.9 If possible, cholecystectomy was performed laparoscopically. Intraoperative cholangiography was not carried out routinely. ES was also performed at the discretion of the physician. The majority of endoscopic retrograde cholangiopancreaticography (ERCP) procedures were performed during the first 24â&#x20AC;&#x201C;72 h of admission. Indications for urgent ES have been described previously, and included obstructive jaundice or suspected cholangitis and predicted severe biliary pancreatitis16. D ef initi on of p ancre atitis For this study, all prospective laboratory data and imaging results obtained during the first 72 h of admission were reviewed by two investigators. Acute biliary pancreatitis was defined as:16 gallstones and/or sludge on ultrasonography or CT; a dilated common bile duct (CBD) on ultrasonography or CT (larger than 8 mm in diameter for patients aged 75 years or less; more than 10 mm in diameter for those aged over 75 years); or two of the following three laboratory abnormalities: serum bilirubin level greater than 40 Âľmol/l, alanine aminotransferase (ALT) level greater than 100 units/l with ALT level greater than the aspartate aminotransferase level, and alkaline phosphatase level greater than 195 units/l with a g-glutamyltransferase level greater than 45 units/l. Other causes of pancreatitis or indications for chronic pancreatitis had to be absent.

130


TIMING OF CHOLECYSTECTOMY IN BILIARY PANCREATITIS

Stu dy outcomes Mild pancreatitis was defined as pancreatitis without local complications such as necrosis or organ failure.17 Criteria for predicted severe pancreatitis were an Acute Physiology And Chronic Health Evaluation (APACHE) II score of at least 8, or an Imrie score of 3 or more, or a C-reactive protein level greater than 50 mg/l.18–20 Biliary events included recurrent biliary pancreatitis, acute cholecystitis, cholangitis and symptomatic cholelithiasis requiring admission to hospital. Recurrent biliary pancreatitis was diagnosed when at least two of the following three criteria were present: typical epigastric abdominal pain, increase in serum amylase or lipase level to more than three times the upper limit of normal, or confirmatory findings on cross-sectional abdominal imaging. Cholecystitis was defined as signs of peritonitis in the right upper abdomen with a positive Murphy’s sign in combination with fever and a raised white blood cell count.21 Symptomatic cholelithiasis requiring admission was defined as the presence of biliary colic or other complaints caused by the presence of gallstones in the CBD.22 Cholangitis was defined as acute abdominal pain, serum bilirubin level greater than 40 µmol/l and/or a dilated CBD on ultrasonography or CT and temperature greater than 38.5°C.16 D ata col l e c ti on Clinical data were collected prospectively at the time of patient inclusion. All CT images were reviewed by one experienced radiologist to assess the presence of peripancreatic or pancreatic necrosis and to determine the CT severity index.23 Follow-up was completed in January 2009. Postdischarge data concerning readmissions for gallstone complications were obtained from the medical records during site visits at the end of follow-up. If a patient was transferred to another hospital, the transferring hospital was contacted for information on possible readmission during follow-up. Statisti c a l ana lysis The number and timing of cholecystectomies, and biliary events before cholecystectomy, in patients with mild biliary pancreatitis were analysed. A further analysis was carried out with patients stratified according to whether or not they had undergone ES during the initial admission. In addition, ES during the initial admission was incorporated into a multivariable logistic regression model, with serum bilirubin and serum ALT levels during the first 48 h of admission as co-variables. The treatment strategies for biliary events, the rate of laparoscopic cholecystectomy, the number of readmissions after cholecystectomy and adherence to the Dutch guideline were assessed. Continuous data are presented as median (interquartile range), and risk ratios and odds ratios with 95 per cent confidence intervals. Between-group differences were analysed

131


PART II CHAPTER 7

using the Mann–Whitney U test for continuous data, and the Fisher’s exact test for categorical data. Two-sided P < 0.050 was considered statistically significant. Analyses were performed with SPSS® version 15.0 (SPSS, Chicago, Illinois, USA).

R E SU LT S Some 731 patients with a primary episode of acute pancreatitis were enrolled in the study; 203 patients with severe pancreatitis were excluded, leaving 528 patients with mild pancreatitis. A biliary aetiology was established in 308 patients (58.3 per cent), of whom a further 41 patients were excluded (Figure 1). Demographic and clinical characteristics of the 267 candidates for cholecystectomy, either during admission or after discharge, are shown in Table 1. The minimum follow-up was 22 months (median 39 (28–48) months). Timing of chol e c yste c tomy Of 267 patients with mild biliary pancreatitis eligible for cholecystectomy, 206 (77.2 per cent) patients underwent the procedure. Eighteen patients (6.7 per cent) had cholecystectomy during the initial admission for mild biliary pancreatitis, leaving 249 candidates for cholecystectomy after discharge. Cholecystectomy was performed after a median of 6 (4–13) weeks in 188 (75.5 per cent) of these 249 patients. Sixty-one (22.8 per cent) of 267 patients did not have a cholecystectomy during follow-up: 28 patients who had and 33 who did not have ES (Figure 1). Re c urrent bi li ar y e vents Biliary events requiring readmission before cholecystectomy are summarized in Table 2. Among the 18 patients who underwent cholecystectomy during the initial admission, no biliary events were documented in the short interval between recovery from pancreatitis and cholecystectomy. Thirty-four (13.7 per cent) of 249 candidates for cholecystectomy after discharge were readmitted owing to biliary events. Twenty-four patients (9.6 per cent) had an attack of recurrent biliary pancreatitis, two (0.8 per cent) were readmitted for acute cholecystitis and eight (3.2 per cent) with symptomatic cholelithiasis. No patient developed cholangitis during the study. The median time between discharge and readmission was 31 (11–76) days. Sixteen readmissions (6.4 per cent) occurred within 4 days of discharge (Table 2). Rol e of end os copi c sphinc terotomy ES was performed after admission in 108 (43.4 per cent) of 249 patients who were candidates for cholecystectomy after discharge (Figure 1). Stones or sludge were found 132


TIMING OF CHOLECYSTECTOMY IN BILIARY PANCREATITIS

in the CBD during ERCP in 51 (47.2 per cent) of these patients. A stent was placed in the CBD in three of the patients who had early ES. Reasons for biliary stents were incomplete removal of CBD stones in two patients and contraindication to ES because of coagulopathy in one. None of the patients underwent ES as an elective procedure after discharge for prevention of recurrent biliary events. Baseline characteristics of patients who had ES and those who did not were comparable, except for levels of liver-related serum enzymes which were higher in the ES group during the first 48 h of admission (Table 3).

Figure 1. Study population

133


PART II CHAPTER 7

The rate of interval cholecystectomy did not differ between patients with and without ES: 80 (74.1 per cent) of 108 versus 108 (76.6 per cent) of 141 respectively (P = 0.658). The median time to cholecystectomy was 6 (4–12) weeks in patients who had ES, compared with 7 (4–14) weeks in those who did not (P = 0.168). Patients who had ES developed fewer recurrent biliary events before cholecystectomy, primarily owing to a reduced risk of biliary pancreatitis. Two patients developed recurrent pancreatitis before cholecystectomy in the ES group compared with 22 in the group without ES: risk ratio 0.18 (0.05 to 0.68; P < 0.001) (Table 4). The potential protective effect of ES on the development of biliary events was explored in a multivariable logistic regression model including serum bilirubin and serum ALT levels during the first 48 h of admission as co-variables. In the multivariable model, ES significantly reduced the

Table 1. Baseline characteristics of 267 patients who were candidates for cholecystectomy during admission or after discharge No. of patients* Age (years)†

59 (41–72)

Women

160 (59.9)

ASA grade I (healthy)

140 (52.4)

II (mild systemic disease)

104 (39.0)

III (severe systemic disease)

23 (8.6)

Predicted severity of pancreatitis APACHE-II†‡

6 (3–9)

Imrie/modified Glasgow score†

2 (1–3)

C-reactive protein (mg/l)†§ CT performed CT severity index†

123 (37–232) 142 (53.2) 3 (2–4)

Highest liver enzymes during first 48 h of admission† Bilirubin (µmol/l)

38 (20–65)

AST (units/l)

176 (69–314)

ALT (units/l)

237 (99–438)

AP (units/l)

153 (101–242)

GGT (units/l)

309 (122–513)

Length of hospital stay (days)† Follow-up (months)†

9 (5–14) 39 (28–48)

*With percentages in parentheses unless indicated otherwise; †values are median (interquartile range). ‡Highest score on day of admission. §Highest value during first 48 h. ASA, American Society of Anesthesiologists; APACHE, Acute Physiology And Chronic Health Evaluation; CT, computed tomography; AST, aspartate aminotransferase; ALT, alanine aminotransferase; AP, alkaline phosphatase; GGT, g-glutamyltransferase.

134


TIMING OF CHOLECYSTECTOMY IN BILIARY PANCREATITIS

Table 2. Recurrent biliary events in 249 candidates for cholecystectomy after discharge. No. of patients* No. of recurrent biliary events‡

34 (13.7)

Recurrent biliary pancreatitis

24 (9.6)

Acute cholecystitis

2 (0.8)

Symptomatic cholelithiasis

8 (3.2)

Cholangitis

0 (0)

Time of onset of recurrent biliary events after discharge Within 2 weeks

11 (4.4)

Within 4 weeks

16 (6.4)

Within 6 weeks

21 (8.4)

Within 10 weeks

26 (10.4)

Time after discharge (days)†

31 (11–76)

*With percentages in parentheses unless indicated otherwise; †values are median (interquartile range). ‡Readmissions for recurrent biliary events before cholecystectomy or before end of follow-up without cholecystectomy.

risk of recurrent biliary events (odds ratio 0.38, 0.16 to 0.91; P = 0.029). Univariable analysis was carried out to explore whether patients at risk of recurrent biliary events after ES could be identified. Predicted severity of pancreatitis, high liver enzyme levels and the presence of stones or sludge during ES were tested, but none was associated with recurrent biliary events (data not shown). C lini c a l cours e and tre atment of bi li ar y e vents The median duration of readmission for the 34 patients with recurrent biliary events was 8 (4–12) days. No patient died during readmission or thereafter, and none developed necrotizing pancreatitis or organ failure. Of 24 patients who developed recurrent pancreatitis, 17 eventually underwent cholecystectomy. Three patients underwent ES during readmission and cholecystectomy 3 months thereafter. One patient underwent ES alone, and in three patients no ES or cholecystectomy was performed. The eight patients who were readmitted with symptomatic cholelithiasis were treated with cholecystectomy. The two patients readmitted with cholecystitis underwent cholecystectomy. C h ol e c yste c tomy There were 18 open and 188 laparoscopic cholecystectomies, of which 15 (8.0 per cent) were converted to an open procedure.

135


PART II CHAPTER 7

Bi li ar y e vents after chol e c yste c tomy Eight (3.9 per cent) of 206 patients were readmitted with recurrent biliary events after a median of 31 (8–44) weeks following cholecystectomy. Three patients developed recurrent biliary pancreatitis, four were readmitted for symptomatic choledocholithiasis and one had cholangitis. One of the 18 patients who had cholecystectomy during the initial admission developed recurrent pancreatitis within a month of the procedure. Furthermore, 26 (8.4 per cent) of a total of 308 patients with mild biliary pancreatitis had a cholecystectomy before the study, of whom three were readmitted with symptomatic cholelithiasis.

Table 3. Baseline characteristics of patients who did or did not undergo endoscopic sphincterotomy ES (n = 108)

No ES (n = 141)

57 (41–73)

59 (41–72)

0.834¶

70 (64.8)

79 (56.0)

0.192

I (healthy)

58 (53.7)

74 (52.5)

0.898

II (mild systemic disease)

44 (40.7)

53 (37.6)

0.694

6 (5.6)

14 (9.9)

0.341

APACHE-II*†

6 (3–10)

6 (4–9)

0.504¶

Imrie/modified Glasgow score*

2 (1–3)

2 (1–3)

0.847¶

141 (55–244)

124 (29–231)

0.422¶

61 (56.5)

72 (51.1)

0.442

3 (2–4)

3 (2–4)

0.645¶

54 (31–91)

29 (16–52)

< 0.001¶

AST (units/l)

222 (151–336)

129 (49–289)

< 0.001¶

ALT (units/l)

319 (159–495)

190 (70–391)

< 0.001¶

AP (units/l)

203 (121–298)

135 (88–209)

< 0.001¶

GGT (units/l)

388 (233–594)

247 (100–450)

< 0.001¶

9 (6–16)

9 (5–13)

0.207¶

39 (28–48)

39 (31–48)

0.287¶

Age (years)* Women ASA grade

III (severe systemic disease) Predicted severity of pancreatitis

C-reactive protein (mg/l)*‡ CT performed CT severity index* Highest liver enzymes levels during first 48 h of admission* Bilirubin (µmol/l)

Length of hospital stay (days)* Total length of follow-up (months)*

Values in parentheses are percentages unless indicated otherwise; *values are median (interquartile range). †Highest score on day of admission. ‡Highest value during first 48 h. ES, endoscopic sphincterotomy; ASA, American Society of Anesthesiologists; APACHE, Acute Physiology And Chronic Health Evaluation; CT, computed tomography; AST, aspartate aminotransferase; ALT, alanine aminotransferase; AP, alkaline phosphatase; GGT, g-glutamyltransferase. § Fisher’s exact test, except ¶Mann–Whitney U test.

136


TIMING OF CHOLECYSTECTOMY IN BILIARY PANCREATITIS

Ad herence to D utch gui d eline Some 142 (53.2 per cent) of 267 patients were treated in accordance with the Dutch guideline, which recommends cholecystectomy or ES during admission or within 3 weeks thereafter. Four of these patients (2.8 per cent) did not have planned cholecystectomy within 3 weeks after discharge, but during readmission for a biliary complication. Excluding these four patients, 138 patients (51.7 per cent) were treated according to the Dutch guideline. Adherence to the guideline differed between hospitals, especially the proportion of patients who had urgent ES. However, no significant differences in baseline characteristics were found between patients who were treated in accordance with the guideline and those who were not. Type of hospital (university versus non-university) did not influence guideline adherence (data not shown).

Table 4. Recurrent biliary events in patients who did or did not undergo endoscopic sphincterotomy before cholecystectomy ES* (n = 108)

No ES* (n = 141)

Risk ratio†

8 (7.4)

26 (18.4)

0.51 (0.27, 0.94)

0.015

Recurrent pancreatitis

2 (1.9)

22 (15.6)

0.18 (0.05, 0.68)

< 0.001

Cholecystitis

2 (1.9)

0 (0)

2.33 (2.02, 2.69)

0.187

Symptomatic cholelithiasis

4 (3.7)

4 (2.8)

1.16 (0.57, 2.35)

0.730

Biliary events‡

Values parentheses are *percentages and †95 per cent confidence intervals. ES, endoscopic sphincterotomy; ‡Readmissions for biliary events before cholecystectomy or before end of follow-up without cholecystectomy. § Fisher’s exact test

DI S C U S SION The main finding of this multicentre study on the timing of cholecystectomy in mild biliary pancreatitis was that 13.7 per cent of patients were readmitted for biliary events when cholecystectomy was performed at a median of 6 weeks after discharge. This included an almost 10 per cent risk of recurrent pancreatitis. If ES was performed during the initial admission, the risk of biliary events was significantly reduced but not eliminated. Furthermore, there was a low risk of recurrent biliary events after cholecystectomy. Finally, this study showed that compliance with the Dutch guideline was poor. Surgeons who advocate cholecystectomy 6 weeks after discharge argue that very early cholecystectomy is associated with a more difficult dissection, potentially leading to more conversions and more complications, such as bile duct injuries.24 They accept the risk of 137


PART II CHAPTER 7

recurrent biliary events, arguing that these can usually be treated by simple cholecystectomy. It has also been suggested that patients should be given time to recover fully from an episode of acute pancreatitis. On the other hand, surgeons advocating cholecystectomy during the initial admission argue that recurrent biliary pancreatitis may be severe and potentially lethal.6 There is evidence to suggest that early cholecystectomy might even be associated with a less difficult dissection than delayed cholecystectomy.25 Unfortunately, no prospective comparative studies have been published to settle this debate. However, a recent investigation from Boston suggested that an interval of 2 weeks between discharge and cholecystectomy might be too long.26 In the present study, 11 patients (4.4 per cent) were readmitted within 2 weeks after discharge. Several, mostly retrospective, studies have suggested that there is a substantial risk of recurrent biliary events after discharge from hospital following an episode of mild biliary pancreatitis and before interval cholecystectomy.27â&#x20AC;&#x201C;29 None reported on the period after cholecystectomy. The present study has shown that there is a small risk of recurrent biliary events after cholecystectomy. These events are most likely caused by retained CBD stones. Another possible explanation for recurrent biliary pancreatitis occurring months to years after cholecystectomy is formation of new stones in the bile duct.30 It is of interest that ES reduced the risk of recurrent pancreatitis in this study. However, most patients with mild biliary pancreatitis do not need urgent ES as this is indicated only in the event of suspected cholangitis or retained CBD stones. Nonetheless, in patients who undergo urgent ES, the question has been raised whether cholecystectomy should be performed early or even at all. In a meta-analysis of five randomized trials in patients without pancreatitis, biliary pain, cholecystitis, cholangitis and even mortality were reduced in patients who underwent cholecystectomy after ES compared with a wait-andsee policy after ES.12 No randomized trial has compared cholecystectomy with a waitand-see policy in patients with mild biliary pancreatitis who have undergone ES. A single-centre study from Sweden showed that, of 109 patients with biliary pancreatitis who had ES without planned cholecystectomy, only one had recurrent pancreatitis.10 However, 45 patients developed other recurrent biliary events, of whom 20 later underwent cholecystectomy. In a study from Ohio, no significant difference in recurrent pancreatitis was found between 38 patients who had ES versus 83 who underwent cholecystectomy.11 In the present study, although ES significantly reduced the risk of recurrent pancreatitis, 7.4 per cent of patients still had recurrent biliary events. Early cholecystectomy may be indicated to prevent such biliary events, which are associated with patient discomfort, hospital admission and additional costs. As a result of the varying study outcomes, international guidelines differ on when to perform a cholecystectomy.3â&#x20AC;&#x201C;5 More recent guidelines tend to recommend early cholecystectomy, either during the hospital admission or within 2 weeks after discharge. 138


TIMING OF CHOLECYSTECTOMY IN BILIARY PANCREATITIS

Nonetheless, in general adherence to such guidelines is low. After the release of the UK guidelines for acute pancreatitis, several studies reported that early cholecystectomy was performed in only a minority of patients.31â&#x20AC;&#x201C;34 In studies from the USA and Australia, cholecystectomy was performed too late and only in a minority of patients.35,36 However, a recent audit from Scotland claimed that a majority of surgeons perform cholecystectomy in accordance with guideline recommendations.37 In the present study adherence was poor as only half of all patients were treated in accordance with the Dutch guideline. A substantial proportion of patients had no cholecystectomy during follow-up. In accordance with the intention-to-treat principle, these patients were not excluded from the analyses and calculations were based on all patients who were candidates for cholecystectomy. This study has some limitations. Although it was a non-randomized cohort study, the population consisted of consecutive prospectively registered patients with a primary attack of acute pancreatitis. The emphasis was on recruiting patients with predicted severe pancreatitis potentially eligible for a randomized trial on probiotic prophylaxis.14 As a consequence, the number of patients with predicted severe disease in the cohort may have been relatively high, as well as the median C-reactive protein values and the median CT severity index. Nevertheless, the results are unlikely to be influenced by this as all patients in the study fulfilled the criteria for mild biliary pancreatitis. Patient selection could have influenced the timing of cholecystectomy or the decision to perform ES. However, multivariable analysis adjusting for liver-related serum enzyme levels showed a significant reduction in recurrent biliary events following ES. This study has shown that a delay in cholecystectomy after mild biliary pancreatitis is associated with a substantial risk of recurrent biliary events. ES, if performed during the initial admission, might reduce the risk of recurrent pancreatitis but does not decrease the likelihood of other biliary events. There is a small risk of recurrent biliary events even after cholecystectomy. C ontributors In addition to the authors, the following clinicians participated in this study: B. van Ramshorst, B. L. Weusten and R. Timmer (St Antonius Hospital, Nieuwegein), L. M. Akkermans, G. A. Cirkel, V. Zeguers, A. Roeterdink, H. G. Rijnhart, M. P. Schwartz, M. S. van Leeuwen and B. U. Ridwan (University Medical Centre Utrecht, Utrecht), B. J. Witteman and P. M. Kruyt Gelderse (Vallei Hospital, Ede), C. J. van Laarhoven and T. A. Drixler (St Elisabeth Hospital, Tilburg), V. B. Nieuwenhuijs, R. J. Ploeg, H. S. Hofker, M. R. Kruijt Spanjer, H. T. Buitenhuis, S. U. van Vliet and S. Ramcharan (University Medical Centre Groningen, Groningen), H. van Goor, A. Nooteboom, J. B. Jansen, G. T. Bongaerts and H. C. Buscher (Radboud University Nijmegen Medical Centre, Nijmegen), M. A. Brink, M. Mundt, R. Frankhuisen and E. C. Consten (Meander Medical Centre, 139


PART II CHAPTER 7

Amerfoort), M. A. Boermeester, O. van Ruler, D. J. Gouma and M. J. Bruno (Academic Medical Centre, Amsterdam), C. H. Dejong and J. P. Rutten (Maastricht University Medical Centre, Maastricht), A. C. Tan, C. Rosman, L. Ootes and B. Houben (Canisius Wilhelmina Hospital, Nijmegen), A. Haasnoot (Leiden University Medical Centre, Leiden), C. H. van Eijck, J. B. C. van der Wal, G. van â&#x20AC;&#x2122;t Hof and E. J. Kuipers (Erasmus Medical Centre, Rotterdam), P. Wahab, E. J. Spillenaar Bilgen and P. van Embden (Rijnstate Hospital, Arnhem), F. J. Kubben, E. van der Harst, J. F. Lange, N. A. Wijffels and L. A. van Walraven (Maasstad Hospital, Rotterdam), and M. A. Cuesta and C. J. Mulder (Vrije Universiteit Medical Centre, Amsterdam).

140


TIMING OF CHOLECYSTECTOMY IN BILIARY PANCREATITIS

REFERENCES 1 Whitcomb DC. Clinical Practice. Acute Pancreatitis. N Engl J Med 2006; 354: 2142–2150. 2 Venneman NG, Buskens E, Besselink MG, Stads S, Go PM, Bosscha K et al. Small gallstones are associated with increased risk of acute pancreatitis: potential benefits of prophylactic cholecystectomy? Am J Gastroenterol 2005; 100: 2540–2550. 3 Working Party of the British Society of Gastroenterology; Association of Surgeons of Great Britain and Ireland; Pancreatic Society of Great Britain and Ireland; Association of Upper GI Surgeons of Great Britain and Ireland. UK guidelines for the management of acute pancreatitis. Gut 2005; 54(Suppl 3): iii1–iii9. 4 Banks PA, Freeman ML; Practice Parameters Committee of the American College of Gastroenterology. Practice guidelines in acute pancreatitis. Am J Gastroenterol 2006; 101: 2379–2400. 5 Uhl W, Warshaw A, Imrie C, Bassi C, McKay CJ, Lankisch PG et al. IAP guidelines for the surgical management of acute pancreatitis. Pancreatology 2002; 2: 565–573. 6 Gullo L, Migliori M, Pezzilli R, Oláh A, Farkas G, Levy P et al. An update on recurrent acute pancreatitis: data from five European countries. Am J Gastroenterol 2002; 97: 1959–1962. 7 Nealon WH, Bawduniak J, Walser EM. Appropriate timing of cholecystectomy in patients who present with moderate to severe gallstone-associated acute pancreatitis with peripancreatic fluid collections. Ann Surg 2004; 239: 741–749. 8 Toouli J, Brooke-Smith M, Bassi C, CarrLocke D, Telford J, Freeny P et al. Guidelines for the management of acute pancreatitis. J Gastroenterol Hepatol 2002; 17(Suppl): S15–S39. 9 Dutch Guideline Acute Pancreatitis. Update 2005 (in Dutch). http://www.internisten.nl/ upl o a d s / J N / G M / J NG Mf S a 1 F 5 - P U _

eRJvR5xQ/3-nivrichtlijnacutepancreatitisupd ate.2005website.pdf [accessed 9 March 2011]. 10 Gislason H, Vetrhus M, Horn A, Hoem D, Söndenaa K, Søreide O et al. Endoscopic sphincterotomy in acute gallstone pancreatitis: a prospective study of the late outcome. Eur J Surg 2001; 167: 204–208. 11 Kaw M, Al-Antably Y, Kaw P. Management of gallstone pancreatitis: cholecystectomy or ERCP and endoscopic sphincterotomy. Gastrointest Endosc 2002; 56: 61–65. 12 <JCIT>McAlister VC, Davenport E, Renouf E. Cholecystectomy deferral in patients with endoscopic sphincterotomy. Cochrane Database Syst Rev 2007; (4)CD006233. 13 Besselink MG, van Santvoort HC, Boermeester MA, Nieuwenhuijs VB, van Goor H, Dejong CH et al. Timing and impact of infections in acute pancreatitis. Br J Surg 2009; 96: 267–273. 14 Besselink MG, van Santvoort HC, Buskens E, Boermeester MA, van Goor H, Timmerman HM et al. Probiotic prophylaxis in patients with predicted severe acute pancreatitis: a randomised, double-blind, placebo-controlled trial. Lancet 2008; 371: 651–659. 15 von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Lancet 2007; 370: 1453–1457. 16 van Santvoort HC, Besselink MG, de Vries AC, Boermeester MA, Fischer K, Bollen TL et al. Early endoscopic retrograde cholangiopancreatography in predicted severe acute biliary pancreatitis: a prospective multicenter study. Ann Surg 2009; 250: 68–75. 17 Bradley EL III. A clinically based classification system for acute pancreatitis. Summary of the International Symposium on Acute Pancreatitis, Atlanta, Ga, September 11 through 13, 1992. Arch Surg 1993; 128: 586–590.

141


PART II CHAPTER 7

18 Knaus WA, Zimmerman JE, Wagner DP, Draper EA, Lawrence DE. APACHE – acute physiology and chronic health evaluation: a physiologically based classification system. Crit Care Med 1981; 9: 591–597. 19 Blamey SL, Imrie CW, O’Neill J, Gilmour WH, Carter DC. Prognostic factors in acute pancreatitis. Gut 1984, 25: 1340–1346. 20 Werner J, Hartwig W, Uhl W, Müller C, Büchler MW. Useful markers for predicting severity and monitoring progression of acute pancreatitis. Pancreatology 2003; 3: 115–127. 21 Hirota M, Takada T, Kawarada Y, Nimura Y, Miura F, Hirata K et al. Diagnostic criteria and severity assessment of acute cholecystitis: Tokyo Guidelines. J Hepatobiliary Pancreat Surg 2007; 14: 78–82. 22 Rome Group for the Epidemiology and Prevention of Cholelithiasis (GREPCO). Radiologic appearance of gallstones and its relationship with biliary symptoms and awareness of having gallstones. Observations during epidemiological studies. Dig Dis Sci 1987; 32: 349–353. 23 Balthazar EJ, Robinson DL, Megibow AJ, Ranson JH. Acute pancreatitis: value of CT in establishing prognosis. Radiology 1990; 174: 331–336. 24 Ranson JH. The timing of biliary surgery in acute pancreatitis. Ann Surg 1979; 189: 654–663. 25 Sinha R. Early laparoscopic cholecystectomy in acute biliary pancreatitis: the optimal choice? HPB (Oxford) 2008; 10: 332–335. 26 Ito K, Ito H, Whang EE. Timing of cholecystectomy for biliary pancreatitis: do the data support current guidelines? J Gastrointest Surg 2008; 12: 2164–2170. 27 Nebiker CA, Frey DM, Hamel CT, Oertli D, Kettelhack C. Early versus delayed cholecystectomy in patients with biliary acute pancreatitis. Surgery 2009; 145: 260–264. 28 Cameron DR, Goodman AJ. Delayed cholecystectomy for gallstone pancreatitis:

142

re-admissions and outcomes. Ann R Coll Surg Engl 2004; 86: 358–362. 29 McCullough LK, Sutherland FR, Preshaw R, Kim S. Gallstone pancreatitis: does discharge and readmission for cholecystectomy affect outcome? HPB (Oxford) 2003; 5: 96–99. 30 Gloor B, Stahel PF, Müller CA, Worni M, Büchler MW, Uhl W. Incidence and management of biliary pancreatitis in cholecystectomized patients. Results of a 7-year study. J Gastrointest Surg 2003;7: 372–377. 31 Senapati PS, Bhattarcharya D, Harinath G, Ammori BJ. A survey of the timing and approach to the surgical management of cholelithiasis in patients with acute biliary pancreatitis and acute cholecystitis in the UK. Ann R Coll Surg Engl 2003; 85: 306–312. 32 Barnard J, Siriwardena AK. Variations in implementation of current national guidelines for the treatment of acute pancreatitis: implications for acute surgical service provision. Ann R Coll Surg Engl 2002; 84: 79–81. 33 Toh SK, Phillips S, Johnson CD. A prospective audit against national standards of the presentation and management of acute pancreatitis in the South of England. Gut 2000; 46: 239–243. 34 Sargen K, Kingsnorth AN. Management of gallstone pancreatitis: effects of deviation from clinical guidelines. JOP 2001; 2: 317– 322. 35 Chiang DT, Thompson G. Management of acute gallstone pancreatitis: so the story continues. ANZ J Surg 2008; 78: 52–54. 36 Nguyen GC, Boudreau H, Jagannath SB. Hospital volume as a predictor for undergoing cholecystectomy after admission for acute biliary pancreatitis. Pancreas 2010; 39: e42–e47. 37 Campbell EJ, Montgomery DA, Mackay CJ. A national survey of current surgical treatment of acute gallstone disease. Surg Laparosc Endosc Percutan Tech 2008; 18: 242–247.


143


PART III TREATMENT OF COMPLICATED PANCREATITIS

CHAPTER 8 A step-up approach or open necrosectomy for necrotizing pancreatitis New England Journal of Medicine 2010 CHAPTER 9 Endoscopic transgastric vs surgical necrosectomy for infected necrotizing pancreatitis: A randomized trial JAMA 2012 CHAPTER 10 Endoscopic transpapillary stenting or conservative treatment for pancreatic fistulas in necrotizing pancreatitis: Multicenter series and literature review Annals of Surgery 2011 CHAPTER 11 A conservative and minimally invasive approach to necrotizing pancreatitis improves outcome Gastroenterology 2011 CHAPTER 12 Treatment of necrotizing pancreatitis Clinical Gastroenterology & Hepatology 2012


PART III TREATMENT OF COMPLICATED PANCREATITIS


CHAPTER 8 A step-up approach or open necrosectomy for necrotizing pancreatitis New England Journal of Medicine 2010

Hjalmar C van Santvoort, Marc G Besselink, Olaf J Bakker, H Sijbrand Hofker, Marja A Boermeester, Cornelis H Dejong, Harry van Goor, Alexander F Schaapherder, Casper H van Eijck, Thomas L Bollen, Bert van Ramshorst, Vincent B Nieuwenhuijs, Robin Timmer, Johan S LamĂŠris, Flip M Kruyt, Eric R Manusama, Erwin van der Harst, George P van der Schelling, Tom Karsten, Eric J Hesselink, Cornelis J van Laarhoven, Camiel Rosman, Koop Bosscha, Ralph J de Wit, Alexander P Houdijk, Maarten S van Leeuwen, Erik Buskens and Hein G Gooszen, for the Dutch Pancreatitis Study Group


PART III CHAPTER 8

A B S TAC T B ackg round Necrotizing pancreatitis with infected necrotic tissue is associated with a high rate of complications and death. Standard treatment is open necrosectomy. The outcome may be improved by a minimally invasive step-up approach. Metho ds In this multicenter study, we randomly assigned 88 patients with necrotizing pancreatitis and suspected or confirmed infected necrotic tissue to undergo primary open necrosectomy or a step-up approach to treatment. The step-up approach consisted of percutaneous drainage followed, if necessary, by minimally invasive retroperitoneal necrosectomy. The primary end point was a composite of major complications (newonset multiple-organ failure or multiple systemic complications, perforation of a visceral organ or enterocutaneous fistula, or bleeding) or death. Resu lts The primary end point occurred in 31 of 45 patients (69%) assigned to open necrosectomy and in 17 of 43 patients (40%) assigned to the step-up approach (risk ratio with the stepup approach, 0.57; 95% confidence interval, 0.38 to 0.87; P=0.006). Of the patients assigned to the step-up approach, 35% were treated with percutaneous drainage only. New-onset multiple-organ failure occurred less often in patients assigned to the step-up approach than in those assigned to open necrosectomy (12% vs. 40%, P=0.002). The rate of death did not differ significantly between groups (19% vs. 16%, P=0.70). Patients assigned to the step-up approach had a lower rate of incisional hernias (7% vs. 24%, P=0.03) and new-onset diabetes (16% vs. 38%, P=0.02). C onclusi ons A minimally invasive step-up approach, as compared with open necrosectomy, reduced the rate of the composite end point of major complications or death among patients with necrotizing pancreatitis and infected necrotic tissue. (Current Controlled Trials number, IRCTN13975868.)

148


STEP-UP APPROACH VS. NECROSECTOMY FOR NECROTIZING PANCREATITIS

I N T RODU C T ION Acute pancreatitis is the third most common gastrointestinal disorder requiring hospitalization in the United States, with annual costs exceeding $2 billion.1,2 Necrotizing pancreatitis, which is associated with an 8 to 39% rate of death, develops in approximately 20% of patients.3 The major cause of death, next to early organ failure, is secondary infection of pancreatic or peripancreatic necrotic tissue, leading to sepsis and multiple organ failure.4 Secondary infection of necrotic tissue in patients with necrotizing pancreatitis is virtually always an indication for intervention.3,5-7 The traditional approach to the treatment of necrotizing pancreatitis with secondary infection of necrotic tissue is open necrosectomy to completely remove the infected necrotic tissue.8,9 This invasive approach is associated with high rates of complications (34 to 95%) and death (11 to 39%) and with a risk of longterm pancreatic insufficiency.10-16 As an alternative to open necrosectomy, less invasive techniques, including percutaneous drainage,17,18 endoscopic (transgastric) drainage,19 and minimally invasive retroperitoneal necrosectomy, are increasingly being used.14,20-22 These techniques can be performed in a so-called step-up approach.23 As compared with open necrosectomy, the step-up approach aims at control of the source of infection, rather than complete removal of the infected necrotic tissue. The first step is percutaneous or endoscopic drainage of the collection of infected fluid to mitigate sepsis; this step may postpone or even obviate surgical necrosectomy.17-19 If drainage does not lead to clinical improvement, the next step is minimally invasive retroperitoneal necrosectomy.14,20-22 The step-up approach may reduce the rates of complications and death by minimizing surgical trauma (i.e., tissue damage and a systemic proinflammatory response) in already critically ill patients.14,21 It remains uncertain which intervention in these patients is optimal in terms of clinical outcomes, health care resource utilization, and costs. We performed a nationwide randomized trial called Minimally Invasive Step Up Approach versus Maximal Necrosectomy in Patients with Acute Necrotising Pancreatitis (PANTER).

M E T HOD S Stu dy D esig n The design and rationale of the PANTER study have been described previously.24 Adults with acute pancreatitis and signs of pancreatic necrosis, peripancreatic necrosis, or both, as detected on contrast-enhanced computed tomography (CT), were enrolled in 7 university medical centers and 12 large teaching hospitals of the Dutch Pancreatitis Study Group. Patients with confirmed or suspected infected pancreatic or peripancreatic necrosis were eligible for randomization once a decision to perform a surgical intervention 149


PART III CHAPTER 8

had been made and percutaneous or endoscopic drainage of the fluid collection was deemed possible. Infected necrotic tissue was defined as a positive culture of pancreatic or peripancreatic necrotic tissue obtained by means of fine-needle aspiration or from the first drainage procedure or operation, or the presence of gas in the fluid collection on contrast-enhanced CT. Suspected infected necrosis was defined as persistent sepsis or progressive clinical deterioration despite maximal support in the intensive care unit (ICU), without documentation of infected necrosis. The exclusion criteria were a flare-up of chronic pancreatitis, previous exploratory laparotomy during the current episode of pancreatitis, previous drainage or surgery for confirmed or suspected infected necrosis, pancreatitis caused by abdominal surgery, and an acute intraabdominal event (e.g., perforation of a visceral organ, bleeding, or the abdominal compartment syndrome). Patients were randomly assigned to either primary open necrosectomy or the minimally invasive step-up approach. Randomization was performed centrally by the study coordinator. Permuted-block randomization was used with a concealed block size of four. Randomization was stratified according to the treatment center and the access route that could be used for drainage (i.e., a retroperitoneal route or only a transabdominal or endoscopic transgastric route). Stu dy O versig ht All patients or their legal representatives provided written informed consent before randomization. This investigator-initiated study was conducted in accordance with the principles of the Declaration of Helsinki. The institutional review board of each participating hospital approved the protocol. Q u a lity C ontrol The indication for intervention and the optimal timing of intervention in necrotizing pancreatitis are frequently subject to discussion.25 Therefore, an expert panel consisting of eight gastrointestinal surgeons, one gastroenterologist, and three radiologists was formed. Whenever infected necrosis was suspected or there was any other indication for intervention in a patient, the expert panel received a case description, including CT images, on a standardized form by e-mail. Within 24 hours, the members of the expert panel individually assessed the indication for intervention and the pa- tientâ&#x20AC;&#x2122;s eligibility for randomization. Whenever possible, the randomization and intervention were postponed until approximately 4 weeks after the onset of disease.5,6,26,27 All interventions were performed by gastrointestinal surgeons who were experienced in pancreatic surgery and by experienced interventional radiologists and endoscopists. Whenever necessary, the most experienced study clinicians visited the participating centers to assist with interventions.

150


STEP-UP APPROACH VS. NECROSECTOMY FOR NECROTIZING PANCREATITIS

O p en Ne c ros e c tomy The open necrosectomy, originally described by Beger et al.,8 consisted of a laparotomy through a bilateral subcostal incision. After blunt removal of all necrotic tissue, two large-bore drains for postoperative lavage were inserted, and the abdomen was closed. Minima l ly Invasive Step-up Appro ach The first step was percutaneous or endoscopic transgastric drainage. The preferred route was through the left retroperitoneum, thereby facilitating minimally invasive retroperitoneal necrosectomy at a later stage, if necessary. If there was no clinical improvement (according to prespecified criteria24) after 72 hours and if the position of the drain (or drains) was inadequate or other fluid collections could be drained, a second drainage procedure was performed. If this was not possible, or if there was no clinical improvement after an additional 72 hours, the second step, video- assisted retroperitoneal dĂŠbridement (VARD) with postoperative lavage,21,22 was performed. (Details on the stepup approach and postoperative management in both groups are included in the Supplementary Appendix, available with the full text of this article at NEJM.org.) E nd Points and D ata C ol l e c ti on The predefined primary end point was a composite of major complications (i.e., newonset multiple organ failure or systemic complications, enterocutaneous fistula or perforation of a visceral organ requiring intervention, or intraabdominal bleeding requiring intervention) (Table 1) or death during admission or during the 3 months after discharge. The individual components of the primary end point were analyzed as secondary end points. Secondary end points also included other complications (Table 1), health care resource utilization, and total direct medical costs and indirect costs from admission until 6 months after discharge (details are available in the Supplementary Appendix). Follow-up visits took place 3 and 6 months after discharge. Data collection was performed by local physicians using Internet-based case-record forms. An independent auditor who was unaware of the treatment assignments checked all completed case-record forms against on-site source data. Discrepancies detected by the auditor were resolved on the basis of a consensus by two investigators who were unaware of the study-group assignments and were not involved in patient care. All CT scans were prospectively evaluated by one experienced radiologist who was unaware of the treatment assignments and outcomes. A blinded outcome assessment was performed by an adjudication committee consisting of eight experienced gastrointestinal surgeons who independently reviewed all data regarding complications. Disagreements were resolved during a plenary consensus meeting with concealment of the treatment assignments. 151


PART III CHAPTER 8

Table 1. Definitions of the Primary and Secondary End Points Major Complications New onset multi organ failure or systemic complications: new onset failure (i.e. not present at any time in the 24 hours before first intervention) of 2 or more organs or occurrence of 2 or more systemic complications at the same moment in time Organ failure* - P  ulmonary failure: PaO2 below 60 mm Hg despite FIO2 of 30% or need for mechanical ventilation - C  irculatory failure: circulatory systolic blood pressure below 90 mm Hg despite adequate fluid resuscitation or need for inotropic catecholamine support - R  enal failure: creatinine level over 177 μmol/L after rehydration or need for hemofiltration or hemodialysis Systemic complications* - D  isseminated intravascular coagulation: platelets <100 × 109/L - S evere metabolic disturbance: calcium level <1.87 mmol/L - G  astrointestinal bleeding: more than 500 ml/24 hours Enterocutaneous fistula: secretion of fecal material from a percutaneous drain or drainage canal after removal of drains or from a surgical wound; either from small or large bowel, confirmed with imaging or during surgery† Perforation of a visceral organ requiring intervention: either surgical, radiological or endoscopic† Intra-abdominal bleeding requiring intervention: either surgical, radiological or endoscopic Other Complications Pancreatic fistula: output via a percutaneous drain or drainage canal after removal of drains or from a surgical wound of any measurable volume of fluid with an amylase content greater than 3 times the serum amylase activity‡ New onset diabetes: the need for insulin or oral antidiabetic drugs at 6 months after discharge which was not present before onset of pancreatitis Use of pancreatic enzymes: the use of oral pancreatic enzyme supplementation to treat clinical symptoms of steatorrhea at 6 months after discharge which was not present before onset of pancreatitis Incisional hernia: full-thickness discontinuity in abdominal wall with bulging of abdominal contents with or without obstruction at 6 months after discharge from hospital∫ * Adapted from the 1992 Atlanta classification for acute pancreatitis.29 † Prior to any analysis, the adjudication committee decided to combine the endpoints enterocutaneous fistula and perforation of a visceral organ, because one is often caused by the other and may coexist in the same patient. ‡ Adapted from the International Study Group on Pancreatic Fistula Definition (ISGPF) criteria for postoperative pancreatic fistula.30 ∫ The original study protocol25 stated “incisional hernia requiring intervention”. Prior to any analysis, the adjudication committee decided to report incisional hernias without intervention because surgical reconstruction of the abdominal wall is usually not performed within 6 months after recovery of necrotizing pancreatitis.

152


STEP-UP APPROACH VS. NECROSECTOMY FOR NECROTIZING PANCREATITIS

Statisti c a l Ana lysis We calculated that we would need to enroll 88 patients24 in order to detect a 64% relative reduction in the rate of the composite primary end point with the step-up approach (from 45% to 16%), with a power of 80% and a two-sided alpha level of 0.05. The large risk reduction with the step-up approach was expected on the basis of results from a Dutch nationwide retrospective multicenter study30 and other previous studies.17,31 Moreover, a larger sample was not thought to be feasible because necrotizing pancreatitis with secondary infection is uncommon. All analyses were performed according to the intention-to-treat principle. The occurrences of the primary and secondary end points were compared between the treatment groups. Results are presented as risk ratios with corresponding 95% confidence intervals. Differences in other outcomes were assessed with the use of the Mannâ&#x20AC;&#x201C;Whitney U test. Predefined subgroup analyses were performed for the presence or absence of organ failure at randomization and the timing of intervention (â&#x2030;¤28 days or >28 days after the onset of symptoms). A formal test of interaction in a logisticregression model was used to assess whether treatment effects differed significantly between the subgroups. No interim analysis was performed. As a precautionary measure, an independent biostatistician who was unaware of the study-group assignments performed sequential monitoring32 of the major complications and deaths reported during the trial (details are available in the Supplementary Appendix). All reported P values are two-sided and have not been adjusted for multiple testing.

R E SU LT S Stu dy Par ti cip ants Between November 3, 2005, and October 29, 2008, a total of 378 patients with acute pancreatitis who had signs of pancreatic necrosis, peripancreatic necrosis, or both were enrolled in the study. A total of 88 patients were randomly assigned to a treatment group (Figure 1). Baseline characteristics of the treatment groups were similar (Table 2). Primar y O p en Ne cros e c tomy Of the 45 patients assigned to primary open necrosectomy, 44 underwent a primary laparotomy. In one patient, who had previously undergone esophagectomy, it was decided after randomization that laparotomy would potentially compromise the gastric conduit. Therefore, primary VARD without previous percutaneous drainage was performed. Patients underwent a median of 1 open necrosectomy (range, 1 to 7). Nineteen patients (42%) required one or more additional laparotomies for additional necrosectomy because 153


PART III CHAPTER 8

of ongoing sepsis (in eight patients), complications (in five patients) or both (in six patients). Fifteen patients (33%) required additional percutaneous drainage after laparotomy. Minima l ly Invasive Step-up Appro ach Forty of 43 patients assigned to the step-up approach (93%) underwent retroperitoneal percutaneous drainage; 1 patient (2%) underwent transabdominal percutaneous drainage and 2 patients (5%) underwent endoscopic transgastric drainage. After the first 72 hours of observation, 19 patients (44%) underwent a second drainage procedure. Details of the drainage procedures are available in the Supplementary Appendix. Fifteen patients (35%) survived after percutaneous or endoscopic drainage only, without the need for necrosectomy. The condition of two patients with progressive multiple organ failure was too unstable for surgery, and they subsequently died. The remaining 26 patients (60%) underwent necrosectomy a median of 10 days (range, 1 to 52) after percutaneous

Figure 1. Enrollment, Randomization, and Follow-up of the Study Patients. VARD denotes video-assisted retroperitoneal dĂŠbridement

154


STEP-UP APPROACH VS. NECROSECTOMY FOR NECROTIZING PANCREATITIS

drainage. A VARD procedure was performed in 24 of the patients, and the other 2 patients underwent primary laparotomy according to the protocol because there was no retroperitoneal access route. A median of 1 VARD procedure (range, 0 to 3) was performed in each patient. In one patient, VARD was intraoperatively converted to laparotomy because it was not possible to reach the pancreatic necrosis through the retroperitoneum. Fourteen patients (33%) required one or more additional operations for further necrosectomy (five patients), complications (seven patients), or both (two patients). Seven of the 26 patients who underwent necrosectomy (27%) required percutaneous drainage afterward. C lini c a l E nd Points The primary and secondary end points are listed in Table 3. The composite primary end point of major complications or death occurred in 31 of 45 patients after primary open necrosectomy (69%) and in 17 of 43 patients after the step-up approach (40%) (risk ratio with the step-up approach, 0.57; 95% confidence interval [CI], 0.38 to 0.87; P= 0.006). All major complications tended to occur more frequently after primary open necrosectomy than after the step-up approach, although the difference was significant only for the composite end point of new-onset multiple organ failure or multiple systemic complications (P= 0.001). This difference was mainly driven by the occurrence of organ failure (Table 3). The rate of death between the two study groups did not differ significantly (P = 0.70) (Table 3). A total of 15 patients in the study died (17%): 8 patients in the step-up group (19%) and 7 patients in the open-necrosectomy group (16%). The causes of death were multiple organ failure in seven patients in the step-up group and six patients in the opennecrosectomy group, postoperative bleeding in one patient in the step-up group and no patients in the open-necrosectomy group, and respiratory failure due to pneumonia in no patients in the step-up group and one patient in the open-necrosectomy group. At the 6-month follow-up, patients who had undergone primary open necrosectomy, as compared with patients who had been treated with the step-up approach, had a higher rate of incisional hernias (24% vs. 7%, P = 0.03), new-onset diabetes (38% vs. 16%, P= 0.02), and use of pancreatic enzymes (33% vs. 7%, P= 0.002). He a lth C are Res ource Uti lizati on and C osts Utilization of health care resources for operations (i.e., necrosectomies and reinterventions for complications) was lower in the group of patients who were treated with the step-up approach than in the group of patients who underwent primary open necrosectomy (P= 0.004) (Table 3).

155


PART III CHAPTER 8

Table 2. Baseline Characteristics of the Patients.*

Characteristic

Minimally invasive step-up approach (N= 43)

Age – yr

57.6 (±2.1)

57.4 (±2.0)

0.94

31 (72)

33 (73)

0.89

Male sex – no. (%)

Primary open P Value necrosectomy (N= 45)

Cause of pancreatitis – no. (%) Gallstones Alcohol abuse Other

0.98 26 (61)

29 (64)

3 (7)

5 (11)

14 (33)

11 (24)

Co-morbidity – no. (%) 19 (44)

21 (47)

0.82

Pulmonary disease

Cardiovascular disease

4 (9)

4 (9)

0.95

Chronic renal insufficiency

3 (7)

2 (4)

0.61

Diabetes

5 (12)

4 (9)

0.67

ASA class on admission – no. (%)†

0.99

I (healthy status)

11 (26)

11 (24)

II (mild systemic disease)

19 (44)

20 (44)

III (severe systemic disease)

13 (30)

14 (31)

28 (20 – 55)

27 (22 – 39)

0.12

8 (4 – 10)

8 (4 – 10)

0.95

less than 30%

17 (40)

19 (42)

30% - 50%

14 (33)

10 (22)

Body-mass index on admission – kg/m2 ‡ Computed tomography∫ CT-severity index¶ Extent of pancreatic necrosis – no. (%)

greater than 50%

0.52

12 (28)

16 (36)

Necrosis extending > 5cm down the paracolic gutters – no. (%)

24 (56)

27 (60)

0.69

Retroperitoneal access route to collection possible – no. (%)

40 (93)

40 (89)

0.50

SIRS – no. (%)**

42 (98)

45 (100)

0.49

Admitted on ICU – no. (%)

23 (54)

21 (47)

0.52

Disease severity║

156

Admitted on ICU anytime before randomization – no. (%)

28 (65)

29 (64)

0.95

Single organ failure – no. (%)

21 (49)

22 (49)

0.99

Multiple organ failure – no. (%)

15 (35)

13 (29)

0.55

Positive blood culture within last 7 days – no. (%)

14 (33)

15 (33)

0.94

Positive blood culture anytime before randomization – no. (%)

22 (51)

25 (56)

0.68


STEP-UP APPROACH VS. NECROSECTOMY FOR NECROTIZING PANCREATITIS

Table 2. Continued

Characteristic

APACHE II score‡‡

Minimally invasive step-up approach (N= 43) 14.6 (±6.1)

Primary open P Value necrosectomy (N= 45)

15.0 (±5.3)

0.75

APACHE II score ≥ 20 – no. (%)

10 (23)

9 (20)

0.71

MODS∫∫

2 (0 – 9)

1 (0 – 10)

0.71

3 (0 – 11)

2 (0 – 12)

0.39

C-reactive protein – mg/L

213.6 (±106)

215.9 (±111)

0.93

White blood cell count – x 109/liter

17.6 (±10.6)

15.9 (±6.3)

0.38

Time since onset of symptoms – days

30 (11 – 71)

29 (12 – 155)

0.86

37 (86)

38 (84)

0.83

23 (54)

23 (51)

SOFA score¶¶

Antibiotic treatment anytime before randomization – no. (%) Nutritional support anytime before randomization – no. (%) Enteral feeding only Parenteral feeding only Enteral feeding and parenteral feeding Oral diet

0.92 3 (7)

4 (9)

12 (28)

11 (24)

5 (12)

7 (16)

Tertiary referrals – no. (%)

21 (49)

23 (51)

0.83

Infected necrosis – no. (%)

39 (91)

42 (93)

0.65

* Continuous data are means (±SD), or medians (range). † American Society of Anesthesiologists (ASA). ‡ The body-mass index is the weight in kilograms divided by the square of the height in meters. ∫ This information was derived from the CT performed just prior to randomization. ¶ CT severity index can range from 0 to 10, with higher scores indicating more extensive pancreatic necrosis and peripancreatic fluid collections. ║ Based on maximum values during 24 hours prior to randomization unless stated otherwise. ** Systemic inflammatory response syndrome (SIRS), as defined by the consensus-conference criteria of the American College of Chest Physicians-Society of Critical Care Medicine. ‡‡ Acute Physiology and Chronic Health Evaluation (APACHE) II scores can range from 0 to 71, with higher scores indicating more severe disease. ∫∫ Multiple organ dysfunction score (MODS) can range from 0 to 24, with higher scores indicating more severe organ dysfunction. ¶¶ Sequential organ failure assessment (SOFA) scores can range from 0 to 24, with higher scores indicating more severe organ dysfunction.

157


158 7 (16) 3 (7)

New onset diabetes∫

Use of pancreatic enzymes∫

8 (18) 1 (2)

6 (13)

6 (14)

2

≥3

0 (0) 31 (69)

17 (40) 19 (44)

0

1

Necrosectomies (laparotomy or VARD) – no. (%)

Healthcare utilization

17 (38)

3 (7)

15 (33)

11 (24)

12 (28)

Pancreatic fistula

17 (38)

7 (16)

10 (22)

Incisional hernia∫

Other morbidity – no. (%)

6 (14) 8 (19)

Enterocutaneous fistula or perforation of a visceral organ requiring intervention

Death – no. (%)

1 (2) 10 (22)

0 (0) 7 (16)

18 (40)

19 (42)

31 (69)

Primary open necrosectomy (N= 45)

5 (12)

5 (12)

17 (40)

Minimally invasive step-up approach (N= 43)

Intra-abdominal bleeding requiring intervention

Multiple systemic complications

Multiple organ failure

New onset multiple organ failure or systemic complications‡

Major morbidity – no. (%)

Secondary endpoints

Major complications or death (combined) – no. (%)†

Primary endpoint

Table 3. Primary and secondary endpoints*

0.21 (0.07 – 0.67)

0.43 (0.20 – 0.94)

0.29 (0.09 – 0.95)

0.74 (0.40 – 1.36)

1.20 (0.48 – 3.01)

0.63 (0.25 – 1.58)

0.73 (0.31 – 1.75)

0.28 (0.11 – 0.67)

0.57 (0.38 – 0.87)

Risk Ratio (95%-CI)

<0.001

0.002

0.02

0.03

0.33

0.70

0.32

0.48

0.001

0.006

P Value

PART III CHAPTER 8


50 (1-287)

Days in hospital

60 (1-247)

11 (0-111)

0.41 (0.19 – 0.88)

0.53

0.26

0.01

<0.001

0.004

* Continuous data are medians (range). † Multiple events in the same patient were considered as one endpoint. ‡ Only for patients without multiple organ failure or multiple systemic complications at any time in the 24 hours before first intervention. ∫ Assessed at 6 months after discharge from the index admission (readmission within 10 days was considered the same admission). ¶ This included necrosectomies (laparotomies or VARD procedures) and additional operations to treat complications (e.g. relaparotomy for abdominal bleeding) during index admission. ║ This included primary drainage procedures as part of the minimally invasive step-up approach and additional drainages procedures after necrosectomy in both treatment groups during index admission. ** Only for patients not admitted to ICU at any time in the 24 hours before first intervention.

9 (0-281)

18 (40)

32 (0-6)

82 (1-7) 7 (16)

91 (1-7)

53 (0-6)

Days in ICU

New ICU admission anytime after first intervention – no. (%)**

Total no. of drainage procedures (range per patient)

Total no. of operations (range per patient) ¶

Table 3. Continued

STEP-UP APPROACH VS. NECROSECTOMY FOR NECROTIZING PANCREATITIS

159


PART III CHAPTER 8

After primary open necrosectomy, 40% of patients required a new ICU admission, as compared with 16% of patients who had been treated with the step-up approach (P= 0.01). The mean total of direct medical costs and in- direct costs per patient during admission and at the 6-month follow-up was €78,775 ($116,016) for the step-up approach and €89,614 ($131,979) for open necrosectomy, for a mean absolute difference of €10,839 ($15,963) per patient. Thus, the step-up approach reduced costs by 12% (details of costs are available in the Table in the Supplementary Appendix). Pre d ef ine d Subg roup Ana lys es Treatment effects with respect to the primary end point were similar across the subgroups on the basis of organ failure at the time of randomization and the timing of intervention (≤28 days or >28 days after the onset of symptoms). None of the tests for interaction were significant (P>0.05).

DI S C U S SION This study showed that the minimally invasive step-up approach, as compared with primary open necrosectomy, reduced the rate of the composite end point of major complications or death, as well as long-term complications, health care resource utilization, and total costs, among patients who had necrotizing pancreatitis and confirmed or suspected secondary infection. With the step-up approach, more than one third of patients were successfully treated with percutaneous drainage and did not require major abdominal surgery. There are several possible explanations for the favorable outcome of the step-up approach. First, as we postulated when designing the study,24 infected necrosis may be similar to an abscess because both contain infected fluid (pus) under pressure. Although a true abscess is more easily resolved with percutaneous drainage because it is composed entirely of liquid, simple drainage may also be sufficient to treat infected necrotic tissue. After the infected fluid is drained, the pancreatic necrosis can be left in situ, an approach that is similar to the treatment of necrotizing pancreatitis without infection. This hypothesis apparently holds true, since 35% of our patients who were treated with the step-up approach did not require necrosectomy. Second, it has been suggested that minimally invasive techniques provoke less surgical trauma (i.e., tissue injury and a proinflammatory response) in patients who are already severely ill.14,20,21 This hypothesis is supported by the substantial reduction in the incidence of new-onset multiple organ failure in our step-up group. Third, in the attempt to completely debride necrosis, viable pancreatic parenchyma may be unintentionally removed. This could explain why, at the 6-month follow-up, significantly more patients 160


STEP-UP APPROACH VS. NECROSECTOMY FOR NECROTIZING PANCREATITIS

who underwent primary open necrosectomy had new-onset diabetes or were taking pancreatic enzymes. For pragmatic reasons, we defined pancreatic insufficiency on the basis of the use of pancreatic-enzyme supplements to treat clinical symptoms of pancreatic insufficiency instead of objective analyses of exocrine insufficiency (e.g., the fecal elastase test). It is possible that some of these patients did not have exocrine insufficiency, although the rate of pancreatic-enzyme supplementation in the open necrosectomy group is consistent with data on exocrine insufficiency after open necrosectomy.15 Our findings are consistent with observations from several retrospective studies. It has been suggested previously that percutaneous drainage can be performed in almost every patient who has necrotizing pancreatitis with infection and obviates the need for necrosectomy in approximately half the patients.17,18,33 Several authors have reported promising results of minimally invasive necrosectomy,14,20,22 including endoscopic procedures.19,34-36 Most studies, however, included only a small number of patients and may have unintentionally selected patients who were less ill than the patients treated with open necrosectomy or were better candidates for minimally invasive techniques. In contrast, the current study was randomized and included a relatively large number of patients, with a high incidence of confirmed infected necrotic tissue and organ failure at the time of intervention. The benefit of the step-up approach in terms of preventing major abdominal surgery and associated complications, such as multiple organ failure requiring ICU admission, is of obvious importance. The reduction in long-term complications, including new-onset diabetes and incisional hernias, is also clinically relevant. Diabetes due to necrotizing pancreatitis is known to worsen over time.15 Moreover, secondary complications from diabetes have a considerable effect on the quality of life and potentially on life expectancy. Incisional hernias often cause disabling discomfort and pain, carry a risk of small-bowel strangulation, and frequently require surgical intervention.37 Aside from these clinical implications, the estimated economic benefit from reduced health care resource utilization and costs may be substantial. Approximately 233,000 patients are admitted with a new diagnosis of acute pancreatitis in the United States each year,38 and necrotizing pancreatitis with secondary infection develops in about 5% of these patients.3,28 On the basis of these numbers, the step-up approach may reduce annual costs in the United States by $185 million. The nationwide multicenter setting of our study and the applicability of the minimally invasive techniques provide support for the generalizability of its results. Percutaneous catheter drainage is a relatively easy and well-established radiologic procedure. VARD is considered a fairly straightforward procedure that can be performed by any gastrointestinal surgeon with basic laparo scopic skills and experience in pancreatic necrosectomy.21,22 Our study specifically compared two treatment strategies and does not provide a direct comparison of open necrosectomy with minimally invasive retroperitoneal necrosectomy. Although there are theoretical advantages of a minimally invasive approach, we have not 161


PART III CHAPTER 8

proved that VARD is superior to open necrosectomy in patients in whom percutaneous drainage has failed. This study was not designed or powered to demonstrate a difference in the rate of death between the two treatment strategies. A study showing a clinically relevant difference in mortality would require thousands of patients and is not likely to be performed. Our results indicate that the preferred treatment strategy for patients with necrotizing pancreatitis and secondary infection, from both a clinical and an economic point of view, is a minimally invasive step-up approach consisting of percutaneous drainage followed, if necessary, by minimally invasive retroperitoneal necrosectomy. Ack now l e dgemens We thank the study research nurses, Anneke Roeterdink and Vera Zeguers, for their tremendous work, all medical and nursing staff in the participating centers for their assistance in enrollment and care of patients in this study, the patients and their families for their contributions to the study, and Ale Algra and Marco Bruno for critically reviewing an earlier version of the manuscript.

162


STEP-UP APPROACH VS. NECROSECTOMY FOR NECROTIZING PANCREATITIS

REFERENCES 1. Shaheen NJ, Hansen RA, Morgan DR, et al. The burden of gastrointestinal and liver diseases, 2006. Am J Gastroenterol 2006;101:2128-38. 2. Fagenholz PJ, Fernández-del Castillo C, Harris NS, Pelletier AJ, Camargo CA Jr. Direct medical costs of acute pancreatitis hospitalizations in the United States. Pancreas 2007;35:302-7. 3. Banks PA, Freeman ML. Practice guide- lines in acute pancreatitis. Am J Gastroenterol 2006;101:2379-400. 4. Whitcomb DC. Acute pancreatitis. 5. Uhl W, Warshaw A, Imrie C, et al. IAP guidelines for the surgical management of acute pancreatitis. Pancreatology 2002;2: 56573. 6. Nathens AB, Curtis JR, Beale RJ, et al. Management of the critically ill patient with severe acute pancreatitis. Crit Care Med 2004;32:2524-36. 7. Forsmark CE, Baillie J, AGA Institute Clinical Practice and Economics Commit- tee, AGA Institute Governing Board. AGA Institute technical review on acute pancreatitis. Gastroenterology 2007;132:2022- 44. 8. Beger HG, Büchler M, Bittner R, Oet- tinger W, Block S, Nevalainen T. Necrosectomy and postoperative local lavage in patients with necrotizing pancreatitis: re- sults of a prospective clinical trial. World J Surg 1988;12:255-62. 9. Traverso LW, Kozarek RA. Pancreatic necrosectomy: definitions and technique. J Gastrointest Surg 2005;9:436-9. 10. Rau B, Bothe A, Beger HG. Surgical treatment of necrotizing pancreatitis by necrosectomy and closed lavage: changing patient characteristics and outcome in a 19year, single-center series. Surgery 2005; 138:28-39. 11. Büchler MW, Gloor B, Müller CA, Friess H, Seiler CA, Uhl W. Acute necrotizing

pancreatitis: treatment strategy ac- cording to the status of infection. Ann Surg 2000;232:619-26. 12. Rodriguez JR, Razo AO, Targarona J, et al. Debridement and closed packing for sterile or infected necrotizing pancreatit- is: insights into indications and outcomes in 167 patients. Ann Surg 2008;247:294-9. 13. Ashley SW, Perez A, Pierce EA, et al. Necrotizing pancreatitis: contemporary analysis of 99 consecutive cases. Ann Surg 2001;234:572-9. 14. Connor S, Alexakis N, Raraty MG, et al. Necrotizing pancreatitis: contemporary analysis of 99 consecutive cases. Ann Surg 2001;234:572-9. 15. Tsiotos GG, Luque-de León E, Sarr MG. Long-term outcome of necrotizing pancreatitis treated by necrosectomy. Br J Surg 1998;85:1650-3. 16. Howard TJ, Patel JB, Zyromski N, et al. Declining morbidity and mortality rates in the surgical management of pancreatic necrosis. J Gastrointest Surg 2007; 11:43-9. 17. Freeny PC, Hauptmann E, Althaus SJ, Traverso LW, Sinanan M. Percutaneous CTguided catheter drainage of infected acute necrotizing pancreatitis: techniques and results. AJR Am J Roentgenol 1998; 170:96975. 18. Baril NB, Ralls PW, Wren SM, et al. Does an infected peripancreatic fluid collection or abscess mandate operation? Ann Surg 2000;231:361-7. 19. Papachristou GI, Takahashi N, Cha- hal P, Sarr MG, Baron TH. Peroral endoscopic drainage/ debridement of walled-off pancreatic necrosis. Ann Surg 2007;245: 943-51. 20. Carter CR, McKay CJ, Imrie CW. Percutaneous necrosectomy and sinus tract endoscopy in the management of infected pancreatic necrosis: an initial experience. Ann Surg 2000;232:175-80. 21. van Santvoort HC, Besselink MG, Horvath KD, et al. Videoscopic assisted retroperitoneal

163


PART III CHAPTER 8

debridement in infected necrotizing pancreatitis. HPB (Oxford) 2007;9:156-9. 22. Horvath KD, Kao LS, Ali A, Wherry KL, Pellegrini CA, Sinanan MN. Laparoscopic assisted percutaneous drainage of infected pancreatic necrosis. Surg Endosc 2001;15:67782. 23. Windsor JA. Minimally invasive pancreatic necrosectomy. Br J Surg 2007;94: 132-3. 24. Besselink MG, van Santvoort HC, Nieuwenhuijs VB, et al. Minimally invasive ‘step-up approach’ versus maximal necrosectomy in patients with acute necrotising pancreatitis (PANTER trial): design and rationale of a randomised controlled multicenter trial [ISRCTN13975868]. BMC Surg 2006;6:6. 25. Connor S, Raraty MG, Neoptolemos JP, et al. Does infected pancreatic necrosis require immediate or emergency debridridement? Pancreas 2006;33:128-34. 26. Fernández-del Castillo C, Rattner DW, Makary MA, Mostafavi A, McGrath D, Warshaw AL. Debridement and closed packing for the treatment of necrotizing pancreatitis. Ann Surg 1998;228:676-84. 27. Besselink MG, Verwer TJ, Schoenmaeckers EJ, et al. Timing of surgical intervention in necrotizing pancreatitis. Arch Surg 2007;142:1194-201. 28. Bradley EL III. A clinically based classification system for acute pancreatitis: summary of the International Symposium on Acute Pancreatitis, Atlanta, GA, September 11–13, 1992. Arch Surg 1993;128:586-90. 29. Bassi C, Dervenis C, Butturini G, et al. Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery 2005;138:8-13. 30. Besselink MG, de Bruijn MT, Rutten JP, Boermeester MA, Hofker HS, Gooszen HG. Surgical intervention in patients with necrotizing pancreatitis. Br J Surg 2006; 93:593-9. 31. van Santvoort HC, Besselink MG, Bol- len TL, Buskens E, van Ramshorst B, Goo- szen

164

HG. Case-matched comparison of the retroperitoneal approach with laparotomy for necrotizing pancreatitis. World J Surg 2007;31:1635-42. 32. Bolland K, Whitehead J. Formal approaches to safety monitoring of clinical trials in lifethreatening conditions. Stat Med 2000;19:2899-917. 33. Besselink MG, van Santvoort HC, Schaapherder AF, van Ramshorst B, van Goor H, Gooszen HG. Feasibility of minimally invasive approaches in patients with infected necrotizing pancreatitis. Br J Surg 2007;94:604-8. 34. Charnley RM, Lochan R, Gray H, O’Sullivan CB, Scott J, Oppong KE. Endoscopic necrosectomy as primary therapy in the management of infected pancreatic necrosis. Endoscopy 2006;38:925-8. 35. Seifert H, Biermer M, Schmitt W, et al. Transluminal endoscopic necrosectomy after acute pancreatitis: a multicentre study with long-term follow-up (the GEPARD Study). Gut 2009;58:1260-6. 36. Gardner TB, Chahal P, Papachristou GI, et al. A comparison of direct endoscopic necrosectomy with transmural endoscopic drainage for the treatment of walled-off pancreatic necrosis. Gastrointest Endosc 2009;69:1085-94. 37. Luijendijk RW, Hop WC, van den Tol MP, et al. A comparison of suture repair with mesh repair for incisional hernia. N Engl J Med 2000;343:392-8. 38. DeFrances CJ, Lucas CA, Buie VC, Golosinskiy A. National Hospital Discharge Survey. Natl Health Stat Report 2008;5:1-20.


165


PART III TREATMENT OF COMPLICATED PANCREATITIS


CHAPTER 9 Endoscopic transgastric vs. surgical necrosectomy for infected necrotizing pancreatitis: A randomized trial JAMA 2012

Olaf J Bakker, Hjalmar C van Santvoort, Sandra van Brunschot, Ronald B Geskus, Marc G Besselink, Thomas L Bollen, Casper H van Eijck, Paul Fockens, Eric J Hazebroek, Rian M Nijmeijer, Jan-Werner Poley, Bert van Ramshorst, Frank P Vleggaar, Marja A Boermeester, Hein G Gooszen, Bas L Weusten and Robin Timmer for the Dutch Pancreatitis Study Group


PART III CHAPTER 9

A B ST R AC T C ontext Most patients with infected necrotizing pancreatitis require necrosectomy. Surgical necrosectomy induces a proinflammatory response and is associated with a high complication rate. Endoscopic transgastric necrosectomy, a form of natural orifice transluminal endoscopic surgery (NOTES), may reduce the proinflammatory response and reduce complications. Obje c tive To compare the pro-inflammatory response and clinical outcome of endoscopic transgastric and surgical necrosectomy. D esig n, S e tting and Pati ents Randomized controlled assessor-blinded clinical trial in 3 academic hospitals and one regional teaching hospital in The Netherlands between August 20th 2008 and March 3rd 2010. Patients had signs of infected necrotizing pancreatitis and an indication for intervention. Inter venti ons Random allocation to endoscopic transgastric or surgical necrosectomy. Endoscopic necrosectomy consisted of transgastric puncture, balloon dilatation, retroperitoneal drainage and necrosectomy. Surgical necrosectomy consisted of video-assisted retroperitoneal debridement or, if not feasible, laparotomy. Main O utc ome Me asure The primary endpoint was the post-procedural proinflammatory response as measured by serum interleukine-6 (IL-6) levels. Secondary clinical endpoints included a predefined composite endpoint of major complications (new-onset multiple organ failure, intraabdominal bleeding, enterocutaneous fistula or pancreatic fistula) or death. Resu lts We randomized 22 patients; two patients did not undergo necrosectomy following percutaneous catheter drainage and could not be analyzed for the primary endpoint. Endoscopic transgastric necrosectomy reduced the post-procedural IL-6 levels compared to surgical necrosectomy (P = 0.004). The composite clinical endpoint occurred less often after endoscopic necrosectomy (20% vs 80%; risk difference [RD] 0.60, 95% confidence interval [CI] 0.16 to 0.80; P = 0.03). Endoscopic necrosectomy did not cause new-onset

168


ENDOSCOPIC TRANSGASTRIC VS SURGICAL NECROSECTOMY

multiple organ failure (0% vs. 50%, RD 0.50, CI 0.12 - 0.76; P = 0.04) and reduced the number of pancreatic fistulas (10% vs. 70%, RD 0.60, CI 0.17 - 0.81; P = 0.02). C onclusi on In patients with infected necrotizing pancreatitis, endoscopic necrosectomy reduced the proinflammatory response as well as the composite clinical endpoint compared to surgical necrosectomy. Trial registration The PENGUIN trial is registered, number ISRCTN07091918.

I N T RODU C T ION Acute pancreatitis is a common and potentially lethal disorder.1 In the United States alone over 50.000 patients are admitted with acute pancreatitis each year.2 One of the most dreaded complications in these patients is infected necrotizing pancreatitis that leads to sepsis and is often followed by multiple-organ failure.3 In these patients interventions are necessary to debride the infected necrosis but the interventions themselves cause substantial morbidity.4-6 The treatment of infected necrotizing pancreatitis has undergone fundamental changes in recent years. Whenever possible, intervention is postponed until the collections with necrosis are demarcated.7,8 Demarcation facilitates necrosectomy and reduces complications related to the drainage and debridement procedures.9 A recent randomized trial demonstrated that a step-up approach of percutaneous catheter drainage with subsequent minimally invasive surgical necrosectomy, is superior to primary open necrosectomy.10 The step-up approach reduced new-onset multiple-organ failure and long term complications such as diabetes and need for pancreatic enzyme supplementation. Approximately one-third of patients required necrosectomy after percutaneous catheter drainage. Current guidelines recommend surgical necrosectomy, 4-6 however, minimally invasive and open surgical necrosectomy for infected necrosis has a 55 up to 81% complication rate.10,11 A less invasive approach to necrosectomy is natural orifice transluminal endoscopic surgery (NOTES).12 NOTES combines endoscopic and surgical techniques.13 Endoscopic transgastric necrosectomy is a new NOTES technique that involves drainage and direct retroperitoneal endoscopic necrosectomy through the gastric wall using endoscopic ultrasound (EUS) guidance.14 Endoscopic necrosectomy is performed under conscious sedation without the need for general anaesthesia and potentially reduces the

169


PART III CHAPTER 9

proinflammatory response and risk of procedure-related complications such as multiple organ failure in these already ill patients.15-17 NOTES has not yet been compared to surgery in a randomized clinical trial in any disease. Here, we report the results of the first randomized clinical trial with NOTES called Pancreatitis, Endoscopic Transgastric versus Primary Necrosectomy in Patients With Infected Necrosis (PENGUIN).

M E T HOD S Pati ents Adult patients needing necrosectomy for suspected or confirmed infected necrotizing pancreatitis who could undergo both endoscopic and surgical necrosectomy, based on computed tomography (CT) imaging, were eligible for randomization. Infected necrosis was defined as a positive culture of pancreatic or peripancreatic necrosis obtained from fine needle aspiration or the first drainage procedure or operation, or the presence of gas in the collection on contrast-enhanced CT. Suspected infected pancreatic necrosis was defined as persisting sepsis or progressive clinical deterioration despite maximal support on the intensive care unit without documented infected necrosis. The trial was performed in 3 university medical centers and a large non-university teaching hospital of the Dutch Pancreatitis Study Group. Exclusion criteria were previous surgical or endoscopic necrosectomy, previous exploratory laparotomy, pancreatitis as a consequence of abdominal surgery, a flare up of chronic pancreatitis, abdominal compartment syndrome, perforation of a visceral organ or bleeding as indication for intervention. All patients or their legal representatives gave written informed consent before randomization. This study was investigator-initiated and was done in accordance with the principles of the Declaration of Helsinki. The institutional review board of each participating hospital approved the protocol. Q u a lity control As previously described, an expert panel of 8 gastrointestinal surgeons, 3 gastroenterologists and 3 radiologists evaluated all candidates prior to randomization.10 The panelâ&#x20AC;&#x2122;s advice on the decision to intervene and possibility to randomize into either arm was discussed with the treating physicians who made the final decision to intervene. Whenever possible, intervention was postponed to at least one month after the onset of disease.7,8 All interventions were performed by gastrointestinal surgeons experienced in pancreatic surgery and experienced endoscopists.

170


ENDOSCOPIC TRANSGASTRIC VS SURGICAL NECROSECTOMY

E nd os copi c tr ansgastri c ne cros e c tomy All procedures were carried out under conscious sedation using midazolam or propofol and pethidine. Linear-array endoscopic ultrasound (EUS) was used to visualize the extent of the necrosis and to determine the optimal puncture site.18,19 Under EUS guidance, the collection was punctured using a 19 Gauge needle (EUSN-19-T, Wilson-Cook, WinstonSalem, N.C., USA). After withdrawal of the stylette, the content of the collection was aspirated to confirm the correction position. Through the 19 Gauge needle, a 0.035 inch (0.89 mm) Jagwire guidewire (Boston Scientific, Boston, MA, USA) was advanced under fluoroscopic guidance. Using electrocautery, the outer sheath of a 10 French (WilsonCook) or a 6 French (Endo-technik, Solingen, Germany) cystogastrotome was advanced into the stomach wall followed by balloon dilatation of the tract up to 8 mm. Thereafter, 2 or more double-pigtail plastic stents (7 French, 5 or 7 cm) and a 6 French nasocystic catheter were placed and the necrotic collection was irrigated with one litre of normal saline per 24 hours. During the subsequent days, the site of access was dilated up to 15 or 18 mm using a dilatation balloon. A forward viewing endoscope was advanced in the cavity and the necrotic tissue was evacuated with a basket, net or polypectomy snare. At the end of each procedure, multiple 7 French double-pigtail plastic stents were placed. This procedure was repeated until the majority of the necrotic material was removed. Surg i c a l ne cros e c tomy Patients underwent video-assisted retroperitoneal debridement (VARD).10 This included debridement of all loosely adherent necrosis through a 5 cm flank incision, using a previously placed retroperitoneal percutaneous drain and videoscopic assistance for guidance. If VARD was not possible because there was no safe retroperitoneal access route, open necrosectomy through laparotomy using a bilateral subcostal incision was performed.20 Continuous postoperative lavage was performed using two large bore drains. Ad diti ona l inter venti ons Additional necrosectomies after surgical necrosectomy were only performed if there was no clinical improvement. Acute complications such as bleeding were treated according to the treating physiciansâ&#x20AC;&#x2122; preference. All patients received intravenous antibiotics (imipenem / cilastatin, meropenem or piperacillin / tazobactam depending on center), which were adjusted according to culture results or stopped if there was clinical improvement. O utcome me asures The primary endpoint was the proinflammatory response following endoscopic or surgical necrosectomy as measured by the serum level of the proinflammatory cytokine interleukine-6 (IL-6). Secondary endpoints included a composite clinical endpoint of 171


PART III CHAPTER 9

death or major complications (see Table 1 for definitions of clinical endpoints). Major complications comprised of new-onset multiple organ failure, intra-abdominal bleeding requiring intervention, enterocutaneous fistula or perforation of a visceral organ requiring intervention and pancreatic fistula. Other secondary endpoints included long term complications such as new-onset diabetes, use of pancreatic enzymes or persisting fluid collections measured at 6 months after discharge. Follow-up visits took place at 3 and 6 months after discharge. Data collection was performed by the study coordinator and the trial nurse at each study site. One experienced radiologist blinded for treatment allocation evaluated all CT scans for the presence and extent of necrosis prior to randomization and persisting fluid collections at 6 months after discharge. An adjudication committee consisting of 5 gastrointestinal surgeons and 2 gastroenterologists independently reviewed all clinical endpoints and performed a blinded outcome assessment.

Table 1. Definitions of clinical endpoints. Organ failure* - Pulmonary failure: PaO2 below 60 mm Hg despite FIO2 of 30% or need for mechanical ventilation - C  irculatory failure: circulatory systolic blood pressure below 90 mm Hg despite adequate fluid resuscitation or need for inotropic catecholamine support - Renal failure: creatinine level over 177 μmol/L after rehydration or need for hemofiltration or hemodialysis. Multiple-organ failure: Failure of 2 or more organs at the same moment in time. Major complications New-onset multiple-organ failure: multiple-organ failure not present in the 24 hours before randomization. Intra-abdominal bleeding requiring intervention: either surgical, radiological or endoscopic. Enterocutaneous fistula or perforation of a visceral organ requiring intervention: secretion of fecal material from a percutaneous drain or drainage canal after removal of drains or from a surgical wound; either from small or large bowel, confirmed with imaging or during surgery; requiring either surgical, radiological or endoscopic intervention. Pancreatic fistula: output via a percutaneous or nasocystic drain or drainage canal after removal of percutaneous drains or from a surgical wound of any measurable volume of fluid with an amylase content greater than 3 times the serum amylase activity‡. Long term complications New-onset diabetes: the need for insulin or oral antidiabetic drugs at 6 months after discharge which was not present before onset of pancreatitis. Use of pancreatic enzymes: the use of oral pancreatic enzyme supplementation to treat clinical symptoms of steatorrhea at 6 months after discharge which was not present before onset of pancreatitis. Persisting fluid collections: the presence of pancreatic or peripancreatic fluid collections on CT scan at 6 months after discharge. * Adapted from the 1992 Atlanta classification for acute pancreatitis. ‡Adapted from the International Study Group on Pancreatic Fistula Definition (ISGPF) criteria for postoperative pancreatic fistula.

172


ENDOSCOPIC TRANSGASTRIC VS SURGICAL NECROSECTOMY

D eterminati on of IL-6 l e vels IL-6 was used as a marker for the overall inflammatory state.21,22 In both groups, blood samples were drawn just before the start of the first intervention and 2 hours, 5 hours, 24 hours and 7 days thereafter. In the endoscopic group, the first procedure was the starting point of blood sampling; the first sample was drawn before gastric puncture, balloon dilatation and drainage of the collection. This first step in treatment was anticipated to cause the greatest pro-inflammatory response because manipulation of an infected collection under pressure could cause bacteremia and subsequent clinical deterioration. EDTA blood samples were drawn, centrifuged at 4ºC at 3000 rounds per minute for 10 minutes immediately and plasma aliquots were stored in -80ºC. Serum levels of IL-6 were determined in all blood samples using a multiplex suspension bead array system according to the manufacturer’s protocol (Bio-Rad Laboratories, Hercules, CA, USA). Data analysis was performed using the Bio-Plex 100 system and Bio-Plex Manager software version 4.1 (Bio-Rad). The lower limit of detection for IL-6 was 0.2 pg/ml. All analyses were done blinded to treatment allocation. Statisti c a l ana lysis When we designed the study, endoscopic necrosectomy was performed in only four specialized hospitals in The Netherlands and only a few small case series had been published on endoscopic necrosectomy. Furthermore, infected necrotizing pancreatitis is a relatively uncommon condition. For these reasons a primary endpoint was chosen with a relatively small necessary sample size. Endoscopy was hypothesized to reduce the post-procedural proinflammatory response, as measured by serum IL-6, as compared to surgical necrosectomy. Calculation of the sample size was based on differences demonstrated in randomized trials of open versus laparoscopic abdominal surgery.23,24 We calculated that 10 necrosectomies per group would be needed to detect a reduction of 45%, with 30% within group standard deviation, with 80% power and a two-sided alpha level of 0.05. As predefined in the study protocol, patients excluded from the study for any reason (e.g., withdrawal of consent) before the first endoscopic or surgical intervention was performed, were replaced by new patients. Reason for this is that only after intervention the blood samples for the primary endpoint could be drawn. Randomization was performed centrally by the study coordinator using a computergenerated permuted block sequence with a concealed block size of 4. Trends in serum IL-6 levels in both treatment groups were assessed using a linear random effects model. We assumed a linear trend for the first 24 hours and a separate value after one week. Results are presented as fitted trend curves with 95% confidence intervals. Overall difference between both treatment groups was tested with the likelihood ratio test, whereas a Wald test was used for the differences at the different observation times. Analyses were performed according to the intention-to-treat principle. A post hoc 173


PART III CHAPTER 9

multivariable logistic regression was performed to investigate the confounding effect of the two variables showing strongest asymmetry at baseline. Continuous data are presented as medians with interquartile ranges. Differences between the treatment groups were quantified via risk differences with 95% confidence intervals for dichotomous outcomes. The Mann-Whitney U test was used for all continuous data. For the longitudinal analyses on IL-6, missing data were considered to be missing at random (6 out of 100 samples). For the other analyses, individuals with missing data were excluded, hence data was considered to be missing completely at random. Data on long term complications was not avaiable in the non-surviving patients (5 out of 20 patients). No interim analysis was performed. All reported P values are two-sided and have not been adjusted for multiple testing. P values lower than 0.05 were considered statistically significant. Analyses were performed using the R statistical program (version 2.13) and SPSS (version 15.0).

R E SU LT S Pati ents Between August 20th 2008 and March 3rd 2010 a total of 34 patients with signs of infected necrotizing pancreatitis were assessed for eligibility and 22 patients underwent randomization (Figure 1). Two patients underwent percutaneous catheter drainage after randomization. Subsequently, these patients clinically improved so that necrosectomy was no longer needed. As predefined in the study protocol, these patients were excluded and laboratory measurements for the primary endpoint were not performed. Baseline characteristics of the two groups of 10 patients were comparable (Table 2). Nineteen of 20 patients (95%) had infected necrosis as proven by cultures at the first intervention. The median time from onset of disease to randomization was around 48 and 59 days (P = 0.91). In the acute phase of the disease before randomization, 40% of all patients suffered from organ failure and 30% suffered from multiple organ failure. E nd os copy All 10 patients assigned to the endoscopic arm underwent endoscopic transgastric necrosectomy. Nine patients were sedated using midazolam and fentanyl, and in one patient propofol was used. The median number of procedures was 3 (interquartile range, 2 to 6). In one patient, necrosectomy was performed during the first procedure. In 2 patients, an additional VARD procedure was performed using a left-sided retroperitoneal percutaneous catheter as guidance following respectively 7 and 5 endoscopic necrosectomies. These 2 patients were analysed in the endoscopic arm.

174


ENDOSCOPIC TRANSGASTRIC VS SURGICAL NECROSECTOMY

Table 2. Baseline Characteristics of the Patients.* Endoscopic Transgastric Necrosectomy (N= 10)

Characteristic Serum IL-6 levels prior to intervention - pg/ml Age – yr Male sex – no. (%)

Surgical necrosectomy (N= 10)

49.7 (16.5 – 103.4)42.4 (15.9 – 138.9)

P-value

0.82

62 (58 – 70)

64 (46 – 72)

0.97

6 (60)

8 (80)

0.63

Cause of pancreatitis – no. (%)

0.82

Biliary

6 (60)

7 (70)

Alcohol abuse

2 (20)

2 (20)

Other

2 (20)

1 (10)

I (healthy status)

1 (10)

1 (10)

II (mild systemic disease)

9 (90)

8 (80)

III (severe systemic disease)

0 (0)

1 (10)

29 (26 – 35)

27 (23 – 37)

0.51

9 (90)

7 (70)

0.58

10 (6 – 14)

11 (7 – 14)

0.76

ASA class on admission – no. (%)†

0.59

Body-mass index on admission – kg/m ‡ 2

Disease severity at randomization║ SIRS – no. (%)** APACHE II score‡‡ C-reactive protein – mg/L

141 (113 – 196)

232 (140 – 275)

0.29

White blood cell count – x 109/liter

11.8 (9.4 – 18.6)

17.9 (9.8 – 19.7)

0.45

Organ failure – no. (%)

2 (20)

3 (30)

1.00

Multiple organ failure – no. (%)

2 (20)

1 (10)

1.00

Admitted on ICU – no. (%)

2 (20)

3 (30)

1.00

Organ failure – no. (%)

4 (40)

4 (40)

1.00

Multiple organ failure – no. (%)

2 (20)

4 (40)

0.63

Positive blood culture – no. (%)

3 (30)

2 (20)

1.00

Admitted on ICU – no. (%)

5 (50)

4 (40)

1.00

8 (4 – 10)

8 (4 – 10)

1.00

Disease severity anytime before randomization

Computed tomography∫ CT severity index¶ - median (range) Extent of pancreatic parenchymal necrosis – no. (%)

1.00

less than 30%

3 (30)

3 (30)

30% - 50%

2 (20)

2 (20)

greater than 50%

4 (40)

4 (40)

Peripancreatic necrosis only – no. (%) Time since onset of symptoms – days Previous percutaneous catheter drainage

1 (10)

1 (10)

1.00

48 (36 – 74)

59 (29 – 69)

0.91

4 (40)

8 (80)

0.17

175


PART III CHAPTER 9

Table 2. Continued Nutritional support anytime before randomization – no. (%)

0.36

Enteral nutrition

6 (60)

5 (50)

Parenteral nutrition

1 (10)

0 (0)

Enteral and parenteral nutrition

1 (10)

4 (40)

Oral diet

2 (20)

1 (10)

Tertiary referrals – no. (%)

8 (80)

9 (90)

1.00

Infected necrosis – no. (%)

10 (100)

9 (90)

1.00

Continuous data are medians and interquartile ranges. † American Society of Anesthesiologists (ASA). ‡ The body-mass index is the weight in kilograms divided by the square of the height in meters. ║Based on maximum values during 24 hours prior to randomization unless stated otherwise. ** Systemic inflammatory response syndrome (SIRS), as defined by the consensus-conference criteria of the American College of Chest Physicians-Society of Critical Care Medicine. ‡‡ Acute Physiology and Chronic Health Evaluation (APACHE) II scores can range from 0 to 71, with higher scores indicating more severe disease. ∫ This information was derived from the CT performed just prior to randomization. ¶ CT severity index can range from 0 to 10, with higher scores indicating more extensive pancreatic necrosis and peripancreatic fluid collections.

Figure 1. A, Cross-sectional view depicting an enlarged, partially necrotic pancreas with a peripancreatic collection containing fluid and necrosis. The preferred access route for video-assisted debridement is within the left retroperitoneal space to reach the necrotic collection between the left kidney and descending colon. A laparoscope is inserted, and long grasping forceps are used to debride the necrosis. B, The access route for natural orifice transluminal endoscopic surgery is through the posterior wall of the stomach. The necrotic collection most often bulges into the stomach facilitating endoscopic transgastric necrosectomy. After balloon dilatation of the puncture site in the stomach wall, the endoscope is introduced into the retroperitoneal space and loose necrotic material is removed.

176


ENDOSCOPIC TRANSGASTRIC VS SURGICAL NECROSECTOMY

Surger y Of all 12 patients in the surgical necrosectomy group, 6 underwent VARD, 4 underwent laparotomy and 2 patients were treated with percutaneous drainage only. All operations were performed under general anaesthesia. The median number of necrosectomies was 1 (interquartile range, 1 to 2). In one patient with necrosis extending down the left and right retrocolic space, 4 necrosectomies were necessary. One patient underwent 3 relaparotomies following necrosectomy; 1 for a gastric perforation and 2 for irrigation of a contaminated abdomen. Another patient underwent an additional laparotomy for an enterocutaneous fistula caused by perforation of the large intestine following VARD. O utcomes With respect to the primary endpoint, IL-6 levels increased after surgical necrosectomy, whereas IL-6 levels decreased after endoscopy. Figure 2 presents the results of IL-6 levels following intervention. Since IL-6 levels had a skewed distribution, a logarithmic transformation was used when fitting the model. Results are presented as fitted trend curves with 95% confidence intervals on the logarithmic scale. The corresponding original IL-6 values are shown on the axis. The overall test for difference between arms was significant (P = 0.004). The largest difference between the two groups was seen at 24 hours after intervention (P = 0.005). The composite clinical endpoint of death and major complications was also reduced in the patients in the endoscopy group (20% vs 80%, risk difference [RD] 0.60; 95% confidence interval [CI] 0.16 to 0.80; P = 0.04; Table 3). New-onset multiple organ failure did not occur after endoscopic transgastric necrosectomy (0% vs. 50%, RD 0.50; CI 0.12 to 0.76; P = 0.04). Fewer patients in the endoscopic group developed pancreatic fistulas (10% vs. 70%, RD 0.60; CI 0.17 to 0.81; P = 0.02). At 6 months after discharge, patients assigned to endoscopic necrosectomy less often used pancreatic enzymes (0% vs. 50%, RD 0.50; CI 0.07 to 0.81; P = 0.04). In total, 5 of 20 patients died (10% vs 40%, RD 0.30; CI -0.08 to 0.60; P = 0.30). All deaths were attributable to persistent multiple-organ failure. Following endoscopic necrosectomy, one patient died 8 days after randomization. Following surgical necrosectomy, 4 patients died respectively, 21, 29, 79 and 155 days after randomization. S ensitiv ity and Mu ltivari abl e Ana lysis All 22 randomized patients, including the two patients who were treated without necrosectomy, were included in a sensitivity analysis for clinical endpoints. One of the patients treated without necrosectomy used pancreatic enzymes at 6 months follow-up. No other clinical endpoints were seen in these patients. When including all 22 randomized patients, the composite clinical endpoint occurred in 2 of 10 patients (20%) after endoscopic necrosectomy and in 8 of 12 patients (67%) after surgical necrosectomy (risk 177


PART III CHAPTER 9

difference 0.47; 95% confidence interval 0.05 to 0.71; P = 0.04). Furthermore, endoscopic necrosectomy still reduced new-onset multiple-organ failure (0% vs. 42%; P = 0.04) and pancreatic fistulas (10% vs. 58%, P = 0.03). Although not statistically significant, the number of patients with previous percutanous drainage at baseline was higher in the endoscopy group than in the surgery group (40% vs. 80%) and the median CRP was higher in the surgery group (141 vs. 232). To investigate the effect on the clinical outcomes from asymmetries in patient and disease characteristics despite randomization, we performed multivariable logistic regression adjusting for percutaneous catheter drainage and CRP. After adjustment for these two asymmetric variables, endoscopic necrosectomy was still associated with a reduced risk for major complications or death (adjusted odds ratio 0.06, 95% confidence interval 0.01 â&#x20AC;&#x201C; 0.78; P = 0.03).

Figure 2. Study Flow Chart

178


1 (10)

36 (17 – 74)

5 (50)

-

0.4 (-0.002 - 0.68)

-

0.28 (-0.17 - 0.63)

0.50 (0.07 - 0.81)

0.28 (-0.17 - 0.63)

0.91

0.14

0.007

0.33

0.04

0.33

0.02

0.47

-

0.03

0.30

0.03

P-value

* Continuous data are medians and interquartile ranges. † Multiple events in the same patient were considered as one endpoint. ‡ Only for patients without (multiple) organ failure at any time in the 24 hours before first intervention. ∫ Surviving patients were assessed 6 months after discharge from the index admission (readmission within 10 days was considered the same admission). ‡‡ Persisting fluid collections as seen on CT scan. ** Surviving patients only.

45 (12 – 69)

1 (10)

New ICU admission anytime after randomization – no. (%)

Days in hospital after randomization**

3 (2 – 6)

No. of necrosectomies (endoscopic or surgical)

1 (1 – 2)

3 (50) (N = 10)

2 (22) (N = 10)

Healthcare utilization

‡‡

Persisting fluid collections

3 (50)

0 (0)

3 (50)

2 (22)

Use of pancreatic enzymes

0.60 (0.17 - 0.81)

7 (70) (N = 6)

1 (10)

-

(N = 9)

0 (0)

New-onset diabetes

Long term complications – no. (%)∫

Pancreatic fistula

0 (0)

 nterocutaneous fistula or perforation of a visceral organ requiring E intervention

0.50 (0.12 - 0.76)

0.30 (-0.08 - 0.60)

0.60 (0.16 - 0.80)

Risk Difference (95% CI)

0.20 (-0.11 - 0.51)

0 (0)

Intra-abdominal bleeding requiring intervention

5 (50)

4 (40)

8 (80)

Surgical necrosectomy (N = 10)

2 (20)

0 (0)

New-onset multiple organ failure‡

Major complications – no. (%)

2 (20)

Death – no. (%)

Endoscopic Transgastric Necrosectomy (N = 10)

Major complications or death† – no. (%)

Table 3. Clinical endpoints*

ENDOSCOPIC TRANSGASTRIC VS SURGICAL NECROSECTOMY

179


PART III CHAPTER 9

Figure 3. Serum Interleukin 6 Levels After Endoscopic Transgastric or Surgical Necrosectomy. Error bars indicate 95% confidence intervals. Data were available from all 10 participants in the endoscopic transgastric group at all times except at 1 week (8 of 10). Similarly, all data were available among the 10 participants in the surgery group except at 2 hours (8 of 10) and at 24 hours and 1 week (9 of 10).

DI S C U S SION The transition from open to laparoscopic surgery over the past 25 years greatly reduced surgical morbidity and mortality. NOTES has the potential for another quantum leap in improved surgical outcomes.12 In this first randomized trial comparing NOTES to current standard surgical therapy, endoscopic necrosectomy was associated with reduced proinflammatory response as opposed to increases in the inflammatory marker IL-6 that was seen with surgical necrosectomy. NOTES was also associated with a significant reduction in the composite clinical endpoint of major complications or death. Within the composite endpoint, NOTES necrosectomy was associated with fewer episodes of multiple-organ failure and pancreatic fistulas. The lower rate of complications after endoscopic transgastric relative to surgical necrosectomy in this study (20% versus 80%) is consistent with prior reports for these procedures .10,11,18,19,25-31 Previous, nonrandomized reports of outcomes for these interventions demonstrated encouraging results of endoscopic transgastric necrosectomy with complications rates of approximately 17%. Two recent European studies on minimally invasive and open surgical necrosectomy reported complications in 55 to 81% of patients.10,11 The very high complication rate following surgery in the current study is explainable by our preoperative protocol of percutaneous catheter drainage. In 80% of patients, surgery was performed in patients not responding to percutaneous catheter drainage and, therefore, patients were more ill than might have been observed in prior studies of pancreatic infected necrosis.7,10 Furthermore, 40% of patients in the surgical group required open necrosectomy that accounted for the high complication rate. Open necrosectomy is typically associated with complication rates as high as 81% 11 180


ENDOSCOPIC TRANSGASTRIC VS SURGICAL NECROSECTOMY

The reduction of organ failure with endoscopy is clinically relevant because organ failure is one of the major causes of long-term morbidity and death following acute pancreatitis.32,33 Patients with acute pancreatitis and either organ failure or infected necrosis have a mortality rate of up to 30%.34 When organ failure and infected necrosis coincide, outcome is even worse.34 The current study was not powered to show a difference in death rate. A larger clinical trial will be necessary to determine if NOTES necrosectomy can reduce mortality from infected pancreatic necrosis. The reduced proinflammatory response and prevention of new-onset multiple-organ failure after endoscopic necrosectomy might be explained by two factors. First, with the use of a natural orifice as access route to the retroperitoneal cavity, surgical dissection to reach the omental sac or retroperitoneum is no longer needed. Second, endoscopic necrosectomy is performed under conscious sedation and does not require general anaesthesia. This is an important difference as general anaesthesia is known to induce or prolong systemic inflammation in critically ill patients.35 Although, the results of our trial are consistent with earlier reports of endoscopic necrosectomy, there are important differences between PENGUIN and others trials that preceded it. Non randomized studies are prone to selection bias, favoring the results of the intervention. Since our groups were randomly allocated, they were balanced, facilitating a more definitive comparison of the interventions than is possible by non randomized studies. Because patients only received our intervention after failing more conservative means of treatment, the proportion of them with infected necrosis (95%) is much higher than the 34-54% reported in the two largest series to date.30,31 These other trials were of pancreatic necrosis that was not necessarily infected so that, on average, patients in those studies were less ill than those we treated. Another important difference between PENGUIN and prior studies was the use of measured inflammatory responses in addition to other outcomes. Most prior studies had radiological findings, such as resolution of the necrotic collection on CT, as the primary outcome measure.25,27-31 Radiological findings do not necessarily correlate to the actual state of disease for pancreatitis. This study has its limitations. The primary endpoint reflected the inflammatory response (i.e. IL-6) and was not a clinical endpoint. Second, the requirement for necrosectomy is relatively rare resulting in a small number of patients enrolled in PENGUIN. Nevertheless, despite the small numbers, the difference in clinical endpoints between groups was striking. A statistically significant difference in the composite clinical outcome was found even with the small number of patients we studied. The direction and magnitude of the clinical effect of endoscopic necrosectomy was similar to earlier reports of endoscopic necrosectomy.36 Third, the first procedure in the endoscopic arm was a drainage procedure. Transgastric necrosectomy was the second intervention. In the surgical arm, necrosectomy was the first procedure. This might have influenced the IL-6 measurements 181


PART III CHAPTER 9

following these procedures. However, new-onset multiple-organ failure was measured up to 6 months after discharge accounting for the cumulative effect of repeated necrosectomies. Although, a median of 3 endoscopic necrosectomies were performed per patient in the endoscopy group, these repeated interventions did not result in newonset multiple-organ failure in these patients. Patients undergoing NOTES necrosectomy required a relatively large number of procedures (median 3, IQR 3 â&#x20AC;&#x201C; 6) that might limit the utility of this new NOTES approach. However, a reduction in serious complications such as new-onset organ failure despite repeated endoscopic necrosectomies, may well justify NOTES necrosectomy. Current literature suggests that endoscopic necrosectomy is increasingly used worldwide.36 As occurred with the introduction of other new techniques such as laparoscopic cholecystectomy, widespread clinical implementation often preceded randomized trials.37,38 The results of the present trial are preliminary and larger, more definitive studies are needed before NOTES necrosectomy can be recommended for routine clinical practice. A nationwide multicenter trial has been started in the Netherlands to confirm these first favourable results (Current Controlled Trials number, ISRCTN09186711). We would like to stress that responsible implementation of this new technique can only be achieved with the use of readily accessible training programs. This is recognized by various organisations involved in the clinical application of NOTES.12,38,39 In this first randomized clinical trial on NOTES in patients with infected necrotizing pancreatitis, endoscopic transgastric necrosectomy reduced the pro-inflammatory response as well as the composite clinical endpoint, including new-onset multiple-organ failure, as compared to surgical necrosectomy. However, these early, promising results require confirmation from a larger clinical trial. Ack now l e dgements We thank Ms.Roeterdink, PENGUIN research nurse, department of Surgery, Utrecht University Medical Center, and all medical and nursing staff in the participating centers for their assistance in enrollment and care of patients in this study. Ms. Roeterdink received no compensation for her role in the study. Ste ering committe e MA Boermeester, OJ Bakker, MG Besselink, P Fockens, HG Gooszen, B van Ramshorst, HC van Santvoort, B Weusten, R Timmer (chair). E xp er t p anel MA Boermeester, TL Bollen, MJ Bruno, CH Dejong, CH van Eijck, P Fockens, H van Goor, HG Gooszen, HS Hofker, JS LamĂŠris, MS van Leeuwen, VB Nieuwenhuijs, AF Schaapherder, R Timmer. 182


ENDOSCOPIC TRANSGASTRIC VS SURGICAL NECROSECTOMY

Adju di c ati on committe e MG Besselink, MA Boermeester, HG Gooszen, B van Ramshorst, BL Weusten, HC van Santvoort, R Timmer. Authorsâ&#x20AC;&#x2122; contributi ons MGB, HCvS, HGG, BvR, RT, BLW and several other members of the study group participated in the design of the study. OJB and TLB collected the data. OJB, HCvS and SvB drew the blood samples. RMN performed the cytokine analyses. OJB, MAB and RBG designed the analysis plan and performed the statistical analysis. OJB drafted the first and subsequent versions of the manuscript. OJB had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors read, corrected and approved the final manuscript. RT, BLW and HGG supervised the current study. All authors vouch for the accuracy and completeness of the data and analyses. C ol l ab or ator s In addition to the authors, the following clinicians participated in this study. None of the collaborators received compensation for their role in the study. MJ Bruno MD PhD1, K Bosscha MD PhD2, CH Dejong MD PhD3, H van Goor MD PhD4, HS Hofker MD5, JS Lameris MD PhD6, MS van Leeuwen MD PhD7, VB Nieuwenhuijs MD PhD5, AF Schaapherder MD PhD8, MP Schwartz MD PhD9, GT Rijkers MD PhD10, RP Voermans MD PhD11. 1 Dept. of Gastroenterology & Hepatology, Erasmus MC, University Medical Center, Rotterdam; 2Dept. of Surgery, Jeroen Bosch hospital, â&#x20AC;&#x2DC;s Hertogenbosch; 3Dept. of Surgery, Maastricht University Medical Center, Maastricht; 4Dept. of Surgery, Radboud University Nijmegen Medical Centre; 5Dept. of Surgery, University Medical Center Groningen, Groningen; 6Dept. of Radiology, Amsterdam Medical Center, Amsterdam; 7Dept. of Radiology, University Medical Center Utrecht, Utrecht; 8Dept. of Surgery, Leiden University Medical Center, Leiden; 9Dept. of Gastroenterology, Meander hospital, Amersfoort; 10Dept. of Surgery, University Medical Center Utrecht, Utrecht; 11Dept. of Gastroenterology & Hepatology, Amsterdam Medical Center, Amsterdam.

183


PART III CHAPTER 9

REFERENCES 1. Fagenholz PJ, Castillo CF, Harris NS, Pelletier AJ, Camargo CA, Jr. Increasing United States hospital admissions for acute pancreatitis, 1988-2003. Ann Epidemiol. 2007;17:491-497. 2. Singla A, Simons J, Li Y et al. Admission volume determines outcome for patients with acute pancreatitis. Gastroenterology. 2009;137:1995-2001. 3. Frossard JL, Steer ML, Pastor CM. Acute pancreatitis. Lancet. 2008;371:143-152. 4. Forsmark CE, Baillie J. AGA Institute technical review on acute pancreatitis. Gastroenterology. 2007;132:2022-2044. 5. UK guidelines for the management of acute pancreatitis. Gut. 2005;54 Suppl 3:iii1-iii9. 6. Uhl W, Warshaw A, Imrie C et al. IAP Guidelines for the Surgical Management of Acute Pancreatitis. Pancreatology. 2002;2:565-573. 7. Rodriguez JR, Razo AO, Targarona J et al. Debridement and closed packing for sterile or infected necrotizing pancreatitis: insights into indications and outcomes in 167 patients. Ann Surg. 2008;247:294-299. 8. Besselink MG, Verwer TJ, Schoenmaeckers EJ et al. Timing of surgical intervention in necrotizing pancreatitis. Arch Surg. 2007;142:1194-1201. 9. Werner J, Hartwig W, Hackert T, Buchler MW. Surgery in the treatment of acute pancreatitis--open pancreatic necrosectomy. Scand J Surg. 2005;94:130-134. 10. van Santvoort HC, Besselink MG, Bakker OJ et al. A step-up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med. 2010;362:1491-1502. 11. Raraty MG, Halloran CM, Dodd S et al. Minimal access retroperitoneal pancreatic necrosectomy: improvement in morbidity and mortality with a less invasive approach. Ann Surg. 2010;251:787-793. 12. Rattner DW, Hawes R, Schwaitzberg S,

184

Kochman M, Swanstrom L. The Second SAGES/ASGE White Paper on natural orifice transluminal endoscopic surgery: 5 years of progress. Surg Endosc. 2011. 13. Baron TH, Thaggard WG, Morgan DE, Stanley RJ. Endoscopic therapy for organized pancreatic necrosis. Gastroenterology. 1996;111:755-764. 14. Seifert H, Wehrmann T, Schmitt T, Zeuzem S, Caspary WF. Retroperitoneal endoscopic debridement for infected peripancreatic necrosis. Lancet. 2000;356:653-655. 15. McGee MF, Schomisch SJ, Marks JM et al. Late phase TNF-alpha depression in natural orifice translumenal endoscopic surgery (NOTES) peritoneoscopy. Surgery. 2008;143:318-328. 16. Bone RC. Immunologic dissonance: a continuing evolution in our understanding of the systemic inflammatory response syndrome (SIRS) and the multiple organ dysfunction syndrome (MODS). Ann Intern Med. 1996;125:680-687. 17. Makhija R, Kingsnorth AN. Cytokine storm in acute pancreatitis. J Hepatobiliary Pancreat Surg. 2002;9:401-410. 18. Voermans RP, Veldkamp MC, Rauws EA, Bruno MJ, Fockens P. Endoscopic transmural debridement of symptomatic organized pancreatic necrosis (with videos). Gastrointest Endosc. 2007;66:909-916. 19. Schrover IM, Weusten BL, Besselink MG, Bollen TL, van RB, Timmer R. EUS-guided endoscopic transgastric necrosectomy in patients with infected necrosis in acute pancreatitis. Pancreatology. 2008;8:271-276. 20. Beger HG, Buchler M, Bittner R, Oettinger W, Block S, Nevalainen T. Necrosectomy and postoperative local lavage in patients with necrotizing pancreatitis: results of a prospective clinical trial. World J Surg. 1988;12:255-262. 21. Aruna D. Pradhan AD, Everett BM, Cook NR, Rifai N, Ridker PM. Effects of Initiating Insulin and Metformin on Glycemic Control and Inflammatory Biomarkers Among


ENDOSCOPIC TRANSGASTRIC VS SURGICAL NECROSECTOMY

Patients With Type 2 Diabetes. The LANCET Randomized Trial. JAMA.2009;302:11861194. 22. Ketoconazole for early treatment of acute lung injury and acute respiratory distress syndrome: a randomized controlled trial. The ARDS Network. JAMA. 2000;283:19952002. 23. Ordemann J, Jacobi CA, Schwenk W, Stosslein R, Muller JM. Cellular and humoral inflammatory response after laparoscopic and conventional colorectal resections. Surg Endosc. 2001;15:600-608. 24. Schwenk W, Jacobi C, Mansmann U, Bohm B, Muller JM. Inflammatory response after laparoscopic and conventional colorectal resections - results of a prospective randomized trial. Langenbecks Arch Surg. 2000;385:2-9. 25. Seewald S, Groth S, Omar S et al. Aggressive endoscopic therapy for pancreatic necrosis and pancreatic abscess: a new safe and effective treatment algorithm (videos). Gastrointest Endosc. 2005;62:92-100. 26. Charnley RM, Lochan R, Gray H, Oâ&#x20AC;&#x2122;Sullivan CB, Scott J, Oppong KE. Endoscopic necrosectomy as primary therapy in the management of infected pancreatic necrosis. Endoscopy. 2006;38:925-928. 27. Papachristou GI, Takahashi N, Chahal P, Sarr MG, Baron TH. Peroral endoscopic drainage/ debridement of walled-off pancreatic necrosis. Ann Surg. 2007;245:943-951. 28. Escourrou J, Shehab H, Buscail L et al. Peroral transgastric/transduodenal necrosectomy: success in the treatment of infected pancreatic necrosis. Ann Surg. 2008;248:1074-1080. 29. Mathew A, Biswas A, Meitz KP. Endoscopic necrosectomy as primary treatment for infected peripancreatic fluid collections (with video). Gastrointest Endosc. 2008;68:776-782.

long-term follow-up (the GEPARD Study). Gut. 2009;58:1260-1266. 31. Gardner TB, Coelho-Prabhu N, Gordon SR et al. Direct endoscopic necrosectomy for the treatment of walled-off pancreatic necrosis: results from a multicenter U.S. series. Gastrointest Endosc. 2011;73:718-726. 32. Johnson CD, bu-Hilal M. Persistent organ failure during the first week as a marker of fatal outcome in acute pancreatitis. Gut. 2004;53:1340-1344. 33. Buchler MW, Gloor B, Muller CA, Friess H, Seiler CA, Uhl W. Acute necrotizing pancreatitis: treatment strategy according to the status of infection. Ann Surg. 2000;232:619-626. 34. Petrov MS, Shanbhag S, Chakraborty M, Phillips AR, Windsor JA. Organ failure and infection of pancreatic necrosis as determinants of mortality in patients with acute pancreatitis. Gastroenterology. 2010;139:813-820. 35. Strom T, Martinussen T, Toft P. A protocol of no sedation for critically ill patients receiving mechanical ventilation: a randomised trial. Lancet. 2010;375:475-480. 36. Haghshenasskashani A, Laurence JM, Kwan V et al. Endoscopic necrosectomy of pancreatic necrosis: a systematic review. Surg Endosc. 2011. 37. Levine DW. An analysis of laparoscopic cholecystectomy. N Engl J Med. 1991;325:967968. 38. Barkun JS, Aronson JK, Feldman LS et al. Evaluation and stages of surgical innovations. Lancet. 2009;374:1089-1096. 39. Meining A, Feussner H, Swain P et al. Natural-orifice transluminal endoscopic surgery (NOTES) in Europe: summary of the working group reports of the Euro-NOTES meeting 2010. Endoscopy. 2011;43:140-143.

30. Seifert H, Biermer M, Schmitt W et al. Transluminal endoscopic necrosectomy after acute pancreatitis: a multicentre study with

185


PART III TREATMENT OF COMPLICATED PANCREATITIS


CHAPTER 10 Endoscopic transpapillary stenting or conservative treatment for pancreatic fistulas in necrotizing pancreatitis: Multicenter series and systematic review Annals of Surgery 2011

Olaf J Bakker, Mark C van Baal, Hjalmar C van Santvoort, Marc G Besselink, Jan-Werner Poley, Joos Heisterkamp, Thomas L Bollen, Hein G Gooszen, and Casper H van Eijck for the Dutch Pancreatitis Study Group


PART III CHAPTER 10

A B S T R AC T Obje c tive Endoscopic transpapillary stenting (ETS) of the pancreatic duct facilitates ductal outflow and may reduce time to pancreatic fistula closure. However, data on the feasibility of ETS in patients with necrotizing pancreatitis are scarce. B ackg round Pancreatic fistulas often occur after intervention in necrotizing pancreatitis and frequently close only after months of conservative treatment. Metho ds From a prospective cohort of patients with acute pancreatitis admitted in 15 hospitals (2004–2007), all patients who underwent ETS or conservative treatment for a pancreatic fistula were identified. Safety, feasibility, and outcome of ETS were evaluated. Furthermore, a literature review was performed for similar studies in necrotizing pancreatitis. Resu lts Of 731 patients with acute pancreatitis, 19 patients were treated with ETS and 16 patients were treated conservatively for a pancreatic fistula. Fistula closure was achieved in 16 of 19 patients (84%) in the ETS group and in 8 of 12 patients (75%) in the conservative group (P = 0.175). The median time to fistula closure after ETS was 71 days (interquartile range [IQR] 34–142) compared with 120 days (IQR 51–175 days) in the conservative group (P = 0.130). Complications were observed in 6 patients. A total of 10 studies reporting the results of 281 patients with stent placement for pancreatic fistulas were included in the literature review. Fistula closure was achieved in 200 patients (71%). Stent-related complications were reported in 9% of patients. C onclusi ons ETS seems a feasible and safe alternative to conservative treatment in patients with pancreatic fistulas after intervention for necrotizing pancreatitis.

188


STENTING FOR FISTULAS IN NECROTIZING PANCREATITIS

I N T RODU C T ION Infection of pancreatic necrosis occurs in around 30% of patients with necrotizing pancreatitis and is considered an indication for intervention1. Surgical necrosectomy and percutaneous catheter drainage are the most frequently used techniques to debride and drain the infected collections. The drains are generally kept in place until the spontaneous production decreases or the collected fluid becomes clear following postprocedural lavage. A subset of patients will, however, have continued spontaneous production of clear fluids from the drain, indicative of a pancreatico-cutaneous fistula.2-5 The estimated incidence of persisting pancreatic fistulas varies from 17 to 76% following intervention for necrotizing pancreatitis.6-9 Pancreatic fistulas are associated with considerable morbidity such as metabolic and nutritional disturbances, prolonged hospitalization and even with mortality.7 Pancreatic fistulas most often are treated conservatively, although the time for a pancreatic fistula to resolve spontaneously usually takes more than 3 months and in some cases even over a year.10 If conservative treatment fails, ultimately, a pancreaticojejunostomy may be indicated.6,11 Over the years, several groups of investigators have proposed endoscopic transpapillary stenting (ETS) as an alternative strategy for the management of pancreatic duct injuries.12-17 ETS decreases intraductal pressure which facilitates drainage of pancreatic secretions to the duodenum instead of through the fistula. Most series describe the results of ETS in patients with fistulas following pancreatic surgery for suspected malignancy or for pseudocysts complicating chronic pancreatitis. Performing an endoscopic intervention in a critically ill patient with necrotizing pancreatitis and concomitant papillary edema, pancreatic ductal obstruction or duodenal obstruction by pancreatic collections may be technically challenging and may potentially even worsen clinical outcome. However, only limited data are available on ETS and conservative treatment of fistulas in patients with necrotizing pancreatitis.13 We evaluated the feasibility and clinical outcome of ETS and conservative treatment in patients with pancreatic fistulas following intervention for necrotizing pancreatitis from a prospective multicenter database. In addition, we performed a literature review for similar studies reporting the result of ETS and conservative treatment in patients with acute necrotizing pancreatitis.

M E T HOD S D esig n This was a retrospective analysis of a prospective database of 731 patients with a primary episode of acute pancreatitis admitted to the 15 centers of the Dutch Pancreatitis Study Group between March 2004 and March 2007. The ethical review board of each 189


PART III CHAPTER 10

participating hospital approved the protocol for prospective data collection and all patients or their legal representatives gave written informed consent for inclusion in the database. The baseline characteristics and outcome of this cohort has previous been reported.18-20 Pati ent s el e c ti on All patients who underwent percutaneous drainage and/or surgical treatment for (suspected) infected necrotizing pancreatitis and subsequently developed a pancreatic fistula were included in the current study. Patients were identified by screening of endoscopic retrograde pancreaticography (ERP) reports, imaging reports and case notes of all patients in the database. No strict criteria existed for the indication for ETS or conservative treatment in the participating hospitals. Although, the attending physicians decided if and when ETS was attempted, the gastroenterologists in this study that decided to perform ETS have extensive endoscopic experience. No patient related factors, fistula production or hospital policies influenced these decisions. A pancreatico-cutaneous fistula was defined as output via an operatively or percutaneously placed catheter (or drainage canal after removal of drains or from a surgical wound) of any measurable volume with an amylase content greater than 3 times the serum amylase activity. A pancreatico-abdominal fistula was defined as the presence of ascites with amylase content greater than 3 times the serum amylase activity. These definitions were adapted from the International Study Group on Pancreatic Fistula criteria.2 Organ failure was defined as: Pulmonary failure: PaO2 below 60 mm Hg despite FIO2 of 30% or need for mechanical ventilation; Circulatory failure: circulatory systolic blood pressure below 90 mm Hg despite adequate fluid resuscitation or need for inotropic catecholamine support; Renal failure: creatinine level over 177 Îźmol/L after rehydration or need for hemofiltration or hemodialysis. Te chni qu e of E nd os copi c Tr ansp api l l ar y Stenting (ET S) Endoscopic retrograde cholangiopancreatography (ERCP) was performed to locate the site of oancreatic duct (PD) injury or disruption, as demonstrated by contrast leakage from the pancreatic duct (Figure 1). In presence of a obstruction or disruption of the PD, a guidewire was inserted and, if possible, a stent was placed. The preferred position was a bridging stent to bypass or cover the PD disruption, and thereby restore the flow of pancreatic secretions. In all patients, bridging of the PD disruption was attempted. If bridging of the leakage site was not possible, an internal stent was placed with the proximal tip of the stent in the collection and the distal tip through the papilla (Figure 2). Finally, if neither a bridging stent nor an internal stent could be placed, a short transpapillary stent was placed to reduce intraductal pressure.

190


STENTING FOR FISTULAS IN NECROTIZING PANCREATITIS

Figure 1. Endoscopic retrograde pancrea­ tography (ERP) of a patient with ductal leakage from the head of the pancreas. A percutaneous catheter drain is visible.

Figure 2. Computed Tomography (CT) of a patient that underwent surgical necrosectomy with postoperative lavage (2 large bore transabdominal drains visible). An internal stent is visible with the proximal tip of the stent in the collection and the distal tip through the papilla.

O utcome and d ata col l e c ti on Data on patient demographics and clinical outcome were available from the prospectively collected database. The outcomes of the study were procedure related complications during ETS, any deterioration in clinical condition following ETS or ERCP and successful fistula treatment following ETS or following conservative treatment. Successful fistula treatment was defined as total resolution of fistula output and absence of large fluid collections on follow-up computed tomography (CT) after removal of all drains without the need for additional interventions. L iter ature re v i e w A MEDLINE search was performed for similar studies reporting the results of endoscopic stenting or conservative treatment for pancreatic fistulas in patients with acute pancreatitis. Search terms were ‘pancreatic fistula’ OR ‘duct disruption’. Cross references were searched in the studies found. Only studies published in the English language were included. Studies reporting the results of surgery for pancreatic fistulas were excluded. All studies reporting the result of transgastric or transduodenal endoscopic treatment for pancreatic pseudocysts or pancreatic fluid collections instead of endoscopic treatment for pancreatic fistulas were also excluded. 191


PART III CHAPTER 10

Statisti cs Descriptive statistics were performed and data are presented as numbers with subsequent percentages. Non-normally distributed data are presented as median with interquartile range (IQR). All statistical analyses were performed using SPSS® for Windows® version 15.0 (SPSS Chicago Illinois, USA).

R E SU LT S Ac ute p ancre atitis cohor t Between March 2004 and March 2007, 731 patients with a first episode of acute pancreatitis were included. From 203 patients with severe pancreatitis, 129 patients (64%) suffered from organ failure and 98 patients (48%) developed infected necrosis. In 115 patients (57%) either percutaneous drainage, necrosectomy or both was performed because of (suspected) infected necrotizing pancreatitis. In 64 patients (56%) percutaneous drainage was performed, 43 of these patients underwent subsequent necrosectomy and 51 patients underwent a primary necrosectomy (without previous drainage). Pancre ati c f istu l a cohor t From these 115 patients, 35 patients (30%) with a pancreatic fistula after radiologic and/ or surgical intervention for necrotizing pancreatitis were identified. Nineteen patients were treated with ETS and 16 patients were treated conservatively (see Table 1 for baseline characteristics). An overview of all patients with ETS and conservative treatment are given in Tables 2 and 3. Univariate logistic regression analysis was performed to assess potential association of each of the variables with ETS or conservative treatment. The patients in the conservative group were older with a median age 61 years compared with 46 years in the ETS group (P = 0.028). One patient had a pancreatico-abdominal fistula; the remaining 34 patients had a pancreatico-cutaneous fistula. A PD injury or disruption was con- firmed by contrast leakage during ERP in 18 of 19 patients treated with ETS (95%). Concomitant obstruction of the PD was observed in 10 of 19 patients (53%). The median fistula output was approximately 150 (IQR 200–300) mL/day for the ETS group and 250 (IQR 75–338) mL/day for the conservative group. These 19 patients were treated in 6 of 15 participating hospitals. Of 19 patients with a pancreatic fistula treated with ETS, 12 patients (63%) had recovered from (multi)organ failure. Three patients were admitted to the intensive care unit (ICU) due to multiorgan failure at the time of ETS. From 16 patients with conservative treatment, 9 patients (56%) suffered from (multi)organ failure at any moment during admission. In 24 patients, ETS for a pancreatic fistula was attempted and succeeded in 19 patients (79%). The median duration of conservative treatment before ETS was 34 days (IQR 192


STENTING FOR FISTULAS IN NECROTIZING PANCREATITIS

18â&#x20AC;&#x201C;92). The remaining 5 patients, in whom ETS was not possible, were treated conservatively. ETS failed in these patients because cannulation of the PD was not possible due to papillary oedema (n = 4) and due to a complete stop of the PD (n = 1). In 13 patients (68%), a PD stent could be placed during the initial procedure (Table 4). In 4 patients (21%), a second procedure was required and 2 patients (11%) required 3 procedures.

Table 1. Characteristics of patients with ETS and conservative treatment for a pancreatic fistula after intervention for infected necrotizing pancreatitis. Patient characteristics Male gender

ETS (N=19)

Conservative (N=16)

P**

11 (37)

8 (50)

0,435

46 (32-61)

61 (52-70)

0,028

CT severity index

7 (4-9)

8 (6-10)

0,182

Persistent organ failure during admission#

15 (79)

9 (56)

0,156

Pancreatic parenchymal necrosis

14 (74)

15 (94)

0,147

Peripancreatic necrosis/collections only*

5 (26)

1 (6)

0,147

Infected necrosis

13 (68)

12 (75)

0,668

26 (11-67)

22 (12-35)

0,227

Surgical necrosectomy

11 (58)

14 (88)

Percutaneous catheter drainage

8 (42)

2 (13)

Octreotide therapy

5 (26)

5 (33)

0,656

Sphincterotomy

8 (42)

5 (31)

0,509

34 (18-92)

-

Pancreatico-cutaneous

15 (94)

16 (100)

Pancreatico-abdominal

1 (6)

0 (0)

Age (y)

Time from onset of symptoms to intervention for infected necrosis (days) Type of initial intervention

Time from intervention to ERP (days)

0,067

Type of fistula

1,000

Location of PD disruption

0,178

Head

2 (11)

4 (25)

Body

7 (37)

4 (25)

Tail

9 (47)

1 (6)

Normal pancreatic duct

1 (5)

0 (0)

Not identified

0 (0)

7 (44)

10 (53)

-

150 (200-300)

250 (75-338)

Pancreatic duct obstruction Fistula output (ml/day)

0,350

Data are n (%) or median (interquartile range). CT = computed tomography; ERP = endoscopic retrograde pancreatography; PD = pancreatic duct. ETS = endoscopic transpapillary stenting. *No pancreatic parenchymal necrosis. #Organ failure more than 48hrs. **Univariate logistic regression analysis was used to test for differences between groups.

193


194

F

F

42

48

50

69

32

45

75

13

14

15

16

17

18

19

M

F

M

M

M

M

F

F

M

M

M

M

M

6

6

8

8

8

8

4

6

4

10

10

4

6

10

10

4

6

4

8

CTSI

PCD

PCD

SN

SN

SN

SN

PCD

SN

SN

SN

PCD

PCD

PCD

SN

SN

PCD

SN

SN

PCD

None

Resp/Renal

Circ/Resp/Renal

Circ/Resp/Renal

Resp

Resp

None

Circ/Resp/Renal

None

Circ/Resp/Renal

Circ/Resp

Circ/Resp

Circ/Resp/Renal

Circ/Resp

Circ/Resp/Renal

Circ/Resp

Circ/Resp

None

Circ/Resp/Renal

Organ failure

Type of Prior to ETS intervention

No

No

No

No

No

No

No

No

No

No

No

No

Yes

No

Yes

Yes

No

No

No

Organ failure

No

No

No

No

No

No

No

No

No

No

No

No

Yes

No

Yes

Yes

No

No

No

ICU admission

At time of ETS

103

53

42

7

28

82

17

375

74

104

145

97

139

76

230

251

92

-

52

CRP

26

97

151

123

92

18

80

40

12

4

34

28

1

19

39

34

30

228

17

Time to ETS (days)

Bridging

Transpapillary

Transpapillary

Transpapillary

Transpapillary

Internal

Bridging

Transpapillary

Internal

Bridging

Bridging

Transpapillary

Internal

Transpapillary

Internal

Transpapillary

Internal

Transpapillary

Internal

Type of stent

31

133

104

63

71

32

7

28

186

151

179

91

-

38

365

-

36

36

95

Time to fistula closure (days)

Yes

No + PR

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Deceased

Yes

Yes

Deceased

Yes

Yes + ETD

Yes

None

Migration

Migration

None

None

None

None

None

occlusion

occlusion

Migration

None

None

None

None

None

Migration

None

None

Successful Complication fistula closure

F=female; M=male; PCD=percutaneous drainage; SN=surgical necrosectomy; ETD=endoscopic transgastric drainage; PR=pancreatic resection; CTSI = CT severity index; PD = pancreatic duct; ETS = endoscopic transpapillary stenting.

28

31

11

12

16

74

8

58

61

7

9

46

6

10

F

45

5

F

M

68

M

44

51

2

3

27

1

Gender

4

Age

Patient No.

Table 2. Overview of patients with necrotizing pancreatitis undergoing endoscopic transpapillary treatment for pancreatic fistulas.

PART III CHAPTER 10


47

61

70

51

39

70

52

53

61

72

63

61

54

53

81

75

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

F

F

M

M

M

M

M

M

F

F

M

M

F

F

F

F

Gender

10

10

5

4

8

6

8

8

10

8

10

10

8

6

6

8

CTSI

PCD

SN

SN

PCD

SN

SN

SN

SN

SN

SN

SN

SN

SN

SN

SN

SN

Type of intervention

None

Circ/Resp

None

None

Resp

None

Circ/Resp

Circ/Resp/Renal

Circ/Resp

Circ/Resp

Resp

None

Circ/Resp

None

Circ/Resp

None

No

Yes

No

No

Yes

No

Yes

Yes

Yes

Yes

Yes

No

Yes

No

Yes

No

No

No

No

No

No

No

No

No

No

No

No

Yes

Yes

Yes

Yes

Yes

Organ failure ICU admission ETS attempted

112

182

177

53

92

90

170

212

140

49

131

50

43

47

128

349

Time to fistula closure (days)

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

No + PR

Yes

No + ETD

No + PR

No + PR

Successful fistula closure

None

None

None

None

None

None

None

None

None

None

None

None

None

None

Deterioration*

None

Complication

F=female; M=male; PCD=percutaneous drainage; SN=surgical necrosectomy; ETD=endoscopic transgastric drainage; PR=pancreatic resection; CTSI = CT severity index; ETS = endoscopic transpapillary stenting. *Clinical deterioration (raise of inflammatory parameters) following unsuccessful stent placement.

Age

Patient No.

Table 3. Overview of patients with necrotizing pancreatitis undergoing conservative treatment for a pancreatic fistula

STENTING FOR FISTULAS IN NECROTIZING PANCREATITIS

195


PART III CHAPTER 10

Bridging of the PD disruption was achieved in only 4 of 19 patients (21%). In the remaining 15 patients, 6 patients received an internal stent and 9 patients a short transpapillary stent. C ompli c ati ons Complications occurred in 6 patients after stent placement. Migration of the stent occurred in 4 of 19 patients (21%) and clogging in 2 of19 patients (11%). In 3 of 4 patients with a migrated stent a new stent was placed. In the fourth patient, the fistula had resolved at the time the stent migration was discovered. The clogged stents were both exchanged during subsequent ERPs. In 1 patient in whom ETS was attempted but failed a clinical deterioration was observed with an increase in abdominal pain for a day and a transient rise in inflammatory markers. O utcome Fistula closure (ie, complete resolution of the fistula without need for other interventions) was achieved in 16 of 19 patients (84%) in the ETS group and in 8 of 12 patients (75%) in the conservative group (P = 0.175). In the ETS group, 1 patient required a pancreaticojejunostomy for fistula closure 118 days after ETS. Two patients died before fistula closure, 18 and 22 days after stent placement. These were 2 patients suffering from severe multiorgan failure that started before any intervention and persisted throughout the disease course. In the conservatively treated patients, 4 patients (25%) underwent an additional intervention to achieve fistula closure. Pancreaticojejunostomy was performed in 3 patients and 1 patient underwent endoscopic transgastric drainage of the collection. The median time to fistula closure after ETS was 71 days (IQR 34â&#x20AC;&#x201C;142) compared with 120 days (IQR 51â&#x20AC;&#x201C;175 days) in the conservative group (P = 0.130). The association between age, ETS, or conservative treatment and time to fistula resolution was explored using multivariate linear regression. Adjusting for the effect of age did not change Table 4. Characteristics of endoscopic stent placement (ETS) for pancreatic fistulas in patients with necrotizing pancreatitis. ETS characteristics

N = 19

Number of ETS attempts for stent placement (median)

1 (1-3)

Stent position Bridging

4 (21)

Transpapillary

9 (47)

Internal

6 (32)

Diameter stent (Fr)

7 (5-10)

Length stent (cm)

7 (3-15)

Data are n (%) or median (range). *PD = pancreatic duct

196


STENTING FOR FISTULAS IN NECROTIZING PANCREATITIS

outcome of fistula closure or time to fistula resolution.In the ETS group, the time to fistula closure was not associated with the number of days between intervention for necrotizing pancreatitis and ETS. Neither was it associated with the position of the endoprothesis (transpapillary, bridging, or internal) or the presence of a PD stenosis (data not shown). Fol l ow-up The presence of PFCs was investigated in all surviving patients that underwent imaging (US, CT, or MRI) during follow-up. Six months after initial fistula resolution, PFCs were seen at imaging in 6 of 15 patients (40%) after ETS and in 8 of 13 patients (62%) after conservative treatment (P = 0.449). After 1 year, 1 patient in the ETS group had died due to metastatic disease. One year and 2 years after fistula resolution, PFCs were noticed in 3 of 14 patients (21%) after ETS and in 7 of 13 patients (54%) after conservative treatment (P = 0.120). L iter ature re v i e w After reviewing 58 potential relevant manuscripts, 10 articles with a total of 281 patients (including the present series) with stent placement for pancreatic fistulas were selected.12-15,17,21-24 An overview is given in table 5. Only the results for patients with acute pancreatitis are given (as far as these results could be deducted from the articles). Twentyfive complications were reported in the studies, although the pancreatitis severity of most patients was not described in detail. Time to fistula closure ranged from a median 2 days to 4 months between studies and within the individual studies a wide range of time to closure was observed (similar to our findings). A direct comparison of results between studies is unfortunately seriously hampered due to lack of homogeneity. The study by Boerma and colleagues is most similar to the current study. This study, like our study, included only patients with necrotizing pancreatitis and a pancreatic fistula following necrosectomy. In this study, however, only patients with a pancreatic obstruction underwent endoscopic stenting. The overall reported rate of fistula closure after stent placement in the studies retrieved varied between 58 and 100 percent. In addition, the literature was searched for studies reporting the result of conservative treatment for pancreatic fistulas in patients with necrotizing pancreatitis. After an extensive MEDLINE and cross-reference search only a single study was identified.10 Sikora and colleagues reported on 156 patients with acute severe pancreatitis that underwent percutaneous drainage or surgical intervention. Out of the 81 surviving patients, 43 developed a pancreatic fistula (53%) and were managed conservatively. In 38 patients (88%) fistula closure was achieved after a median of 70 days (28 â&#x20AC;&#x201C; 424 days). However, from these 38 patients with spontaneous closure, 3 patients underwent cystogastrostomy (n=3) or cystojejunostomy (n=4) at a later stage for treatment of pseudocysts. 197


198

38

16

1998

Howard

26

2005

2005

2006

2006

2009

Halttunen

Varadarajulu

Brennan

Cicek

Bakker

360

35

97

50

43

281 (81)

19 (54)

14 (54)

30 (100)

92 (95)

43 (86)

43 (100)

11 (69)

13 (81)

7 (18)

9 (100)

No. of patients with endoscopic treatment¶

131 (45)

35 (100)

NS

7 (23)

44 (46)§

9 (18)

NS

8 (50)

16 (100)

10 (26)

2 (22)

No. of patients with acute necrotizing pancreatitis

TP, IN, B

TP, IN, B

TP

TP, IN, B

TP

TP, IN, B

IN, B

TP, B

TP

TP, B

200 (71)

16 (84)

12 (86)

21 (70)

52 (55)

36 (84)

25 (58)

10 (91)

13 (100)

7 (100)

8 (89)

2–122 (2-984)

71 (34–142)†

7 ** ‡

NS

NS

122 (16-984)

NS

9 (2-33)**

10 (2-64)

NS

2 (2-7)

26 reported complications in 281 patients (9%)

2 stent occlusions; 4 stent migrations; 1 (mild) clinical deterioration

2 stent migrations

1 stent occlusion

1 patient died with multiorgan failure 1 week after ETS; 3 stent occlusions; 2 patients with pus from stent

1 patient with fever; 1 patient with mild post-ERCP pancreatitis

4 patients with clinical deterioration

NS

1 stent occlusion

NS

1 stent occlusion; 1 infected pseudocyst

Position of PD Successful fistula Duration until Number and type of stent stent+ closure after stent fistula closure related complications placement¶ after stent placement* (days)

NS = not specified; ¶N (percentage); + TP = transpapillary stent; IN = internal stent; B = bridging stent; *Data are median (range); ** Data are mean (range); † Median (interquartile range); ‡ Patients were referred to surgery if drain production did not decrease significantly within 10 days after stent insertion.

Total

30

2002

Telford

16

2000

2001

Boerma

Costamagna

9

1997

Kozarek

Total No. of patients

Year

Study

Table 5. Systematic literature review on endoscopic treatment of pancreatic fistulas.

PART III CHAPTER 10


STENTING FOR FISTULAS IN NECROTIZING PANCREATITIS

DI S C U S SION This is the largest series on ETS and conservative treatment for pancreatic fistulas after surgical or radiological intervention in patients with necrotizing pancreatitis. Although ETS was performed selectively and not in a prospective manner, our results suggest that ETS is feasible, relatively safe and might form an alternative for conservative treatment in critically ill patients. Previous studies have described the results of ETS for pancreatic fistulas, as is presented in the literature review. Most studies report a high rate of successful fistula closure following endoscopic stenting. The results, however, are difficult to compare. The studies differ in several aspects. First of all, most studies also included patients with pancreatic fistulas following surgery for pancreatic cancer or chronic pancreatitis. Second, the definition of a pancreatic fistula varied between studies. Third, the timing and indication for endoscopic treatment differed or were not reported. Fourth, the patient characteristics of patients with acute pancreatitis (e.g. presence of necrosis or organ failure) are not given. Finally, the position of the stent (transpapillary, internal or bridging) was not reported in most studies. Kozarek et al. reported in 1997 that transpapillary stenting can effectively close pancreatic fistulas within a week.12 In 9 retrospectively identified patients with ductal injury following surgery for chronic and acute pancreatitis, 8 fistulas closed within a week. One stent occlusion and one stent migration were described. These promising results unfortunately have not been reproduced by other case series. Howard et al. described 7 patients with different types of pancreatic fistulas and treatment with transpapillary stenting.14 ETS was technically successful in all patients, although patients with complete ductal disruption were not deemed amenable for ETS and only patients with partial ductal disruption were included. No complications of ETS were reported. Telford et al. described results in 43 patients identified with pancreatic ductal disruption of different etiology.15 Four patients experienced a clinical deterioration after stent placement, although stent occlusion was not documented. The study by Boerma et al. is the only study, next to the present series, that solely included patients with necrotizing pancreatitis that had undergone surgical necrosectomy.13 Following ETS, after a median of 10 days, fistula production stopped in all 13 patients. All patients, however, that were treated with endoscopic stenting had a ductal obstruction and all stents were placed through the obstruction. This might have shortened overall time to fistula closure compared to the present series. In the current study, fistula closure was achieved in approximately 80% of patients. The median time to fistula closure, however, was more than 2 months after stent placement. Our results differ from some of the results reported. The observed difference may be explained by the fact that we only included patients that underwent intervention for suspected infected necrosis. In these patients, the ductal anatomy of the pancreas was severely disrupted as is reflected by the high rate of ductal obstruction, ductal 199


PART III CHAPTER 10

disconnection and the median CT severity index (CTSI). CTSI was 8 or higher in 50% of patients. A CTSI of 8 correlates with at least 30% lack of enhancement of pancreatic parenchyma (along with the presence of 2 or more peripancreatic collections).25 As a result of the disrupted anatomy, in only 4 patients a bridging stent could be positioned. The transpapillary or internal stent placed in the remaining 15 patients lowers the pancreatic intraductal pressure, but probably does not entirely reduce flow of pancreatic juices into the pancreatic fluid collections. As a consequence, complete closure of the fistula will still take considerable time. A bridging stent has been correlated with a more favourable outcome.15 In the current study, outcome for patients with a bridging stent was similar to results after transpapillary or an internal stentalthough patient numbers in the current study probably are too small to formally compare the outcome for the different stent positions. Despite many challenging factors most often encounteredin critically ill patients with necrotizing pancreatitis, no complications such as bleeding or perforation directly related to ETS were noted in the current study. Several other complications did occur, such as occlusion and migration of stents. Notably, these events did not result in clinical deterioration. In this study, we used the adapted definition for pancreatic fistula as proposed by the International Study Group on Pancreatic Fistula (ISGPF) that was designed for pancreatic resections.2 Fistulas following pancreatic resections represent failure of a healing or sealing of a pancreatic-enteric anastomosis. We defined a pancreatic fistula following intervention for necrotizing pancreatitis as output via a percutaneous drain (or drainage canal after removal of drains or from a surgical wound) of any measurable volume of fluid with an amylase content greater than 3 times the serum amylase activity. We feel this definition is more appropriate for patients with necrotizing pancreatitis. From the number of patients with a pancreatic fistula in the study cohort, the incidence of pancreatic fistulas after intervention for necrotizing pancreatitis cannot accurately be deducted. The physicians decided to perform ETS at their own discretion and the conservatively treated patients were identified from the ERP reports, imaging reports and casenotes. It is possible that some patients with a pancreatic fistula have gone unnoticed. This study has several shortcomings. It was a retrospective analysis and ETS was not performed according to a standardized protocol. Conservative treatment before ETS differed among centers. For example, 4 patients received octreotide and 8 patients underwent a biliary sphincterotomy during early ERCP when signs of an impacted stone or significant cholestasis were present. It is unlikely that these differences in conservative treatment strategies may have influenced outcome of ETS because in most patients conservative treatment was attempted for more than 3 weeks. More importantly, selection of patients for ETS and conservative treatment very likely influenced the time to fistula 200


STENTING FOR FISTULAS IN NECROTIZING PANCREATITIS

resolution. It is possible that some of the patients undergoing ETS would have had spontaneous closure of their fistula in the same time period with conservative treatment. On the other hand, it is also very much possible that some conservative treated patients would have had earlier resolution if they were treated with ETS. The overall median time to fistula closure did not differ significantly between groups. If ETS really shortens time to fistula closure can only be answered in a comparative, preferably randomized, design. Such a study is needed, but unfortunately difficult to perform. In conclusion, this study suggests that endoscopic transpapillary treatment in patients with pancreatic fistulas after intervention for infected necrotizing pancreatitis is a feasible and a safe alternative to conservative treatment

201


PART III CHAPTER 10

REFERENCES 1. Banks PA, Freeman ML. Practice guidelines in acute pancreatitis. Am J Gastroenterol 2006; 101:2379-2400. 2. Bassi C, Dervenis C, Butturini G et al. Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery 2005; 138:8-13. 3. Nealon WH, Walser E. Main pancreatic ductal anatomy can direct choice of modality for treating pancreatic pseudocysts (surgery versus percutaneous drainage). Ann Surg 2002; 235:751-758. 4. Nealon WH, Bhutani M, Riall TS et al. A unifying concept: pancreatic ductal anatomy both predicts and determines the major complications resulting from pancreatitis. J Am Coll Surg 2009; 208:790-799. 5. Werner J, Buchler MW. Management of pancreatic fistulas in acute pancreatitis. In: Beger H, Warshaw A, eds. The Pancreas, an Integrated Textbook of Basic Science, Medicine, and Surgery, 2nd. Wiley-Blackwell 2008. 356-361.

202

11. Alexakis N, Sutton R, Neoptolemos JP. Surgical treatment of pancreatic fistula. Dig Surg 2004; 21:262-274. 12. Kozarek RA, Ball TJ, Patterson DJ et al. Transpapillary stenting for pancreaticocutaneous fistulas. J Gastrointest Surg 1997; 1:357-361. 13. Boerma D, Rauws EA, van Gulik TM et al. Endoscopic stent placement for pancreaticocutaneous fistula after surgical drainage of the pancreas. Br J Surg 2000; 87:1506-1509. 14. Howard TJ, Stonerock CE, Sarkar J et al. Contemporary treatment strategies for external pancreatic fistulas. Surgery 1998; 124:627-632. 15. Telford JJ, Farrell JJ, Saltzman JR et al. Pancreatic stent placement for duct disruption. Gastrointest Endosc 2002; 56:18-24. 16. Deviere J, Bueso H, Baize M et al. Complete disruption of the main pancreatic duct: endoscopic management. Gastrointest Endosc 1995; 42:445-451.

6. Ho HS, Frey CF. Gastrointestinal and pancreatic complications associated with severe pancreatitis. Arch Surg 1995; 130:817822.

17. Varadarajulu S, Noone TC, Tutuian R et al. Predictors of outcome in pancreatic duct disruption managed by endoscopic transpapillary stent placement. Gastrointest Endosc 2005; 61:568-575.

7. Tsiotos GG, Smith CD, Sarr MG. Incidence and management of pancreatic and enteric fistulas after surgical management of severe necrotizing pancreatitis. Arch Surg 1995; 130:48-52.

18. van Santvoort HC, Besselink MG, de Vries AC et al. Early endoscopic retrograde cholangiopancreatography in predicted severe acute biliary pancreatitis: a prospective multicenter study. Ann Surg 2009; 250:68-75.

8. Fotoohi M, Dâ&#x20AC;&#x2122;Agostino HB, Wollman B et al. Persistent pancreatocutaneous fistula after percutaneous drainage of pancreatic fluid collections: role of cause and severity of pancreatitis. Radiology 1999; 213:573-578.

19. Besselink MG, van Santvoort HC, Boermeester MA et al. Timing and impact of infections in acute pancreatitis. Br J Surg 2009; 96:267-273.

9. Connor S, Ghaneh P, Raraty M et al. Minimally invasive retroperitoneal pancreatic necrosectomy. Dig Surg 2003; 20:270-277.

20. Besselink MG, van Santvoort HC, Buskens E et al. Probiotic prophylaxis in predicted severe acute pancreatitis: a randomised, double-blind, placebo-controlled trial. Lancet 2008; 371:651-659.

10. Sikora SS, Khare R, Srikanth G et al. External pancreatic fistula as a sequel to management of acute severe necrotizing pancreatitis. Dig Surg 2005; 22:446-451.

21. Costamagna G, Mutignani M, Ingrosso M et al. Endoscopic treatment of postsurgical external pancreatic fistulas. Endoscopy 2001; 33:317-322.


STENTING FOR FISTULAS IN NECROTIZING PANCREATITIS

22. Halttunen J, Weckman L, Kemppainen E et al. The endoscopic management of pancreatic fistulas. Surg Endosc 2005; 19:559-562. 23. Brennan PM, Stefaniak T, Palmer KR et al. Endoscopic transpapillary stenting of pancreatic duct disruption. Dig Surg 2006; 23:250-254. 24. Cicek B, Parlak E, Oguz D et al. Endoscopic treatment of pancreatic fistulas. Surg Endosc 2006; 20:1706-1712. 25. Balthazar EJ, Robinson DL, Megibow AJ et al. Acute pancreatitis: value of CT in establishing prognosis. Radiology 1990; 174:331-336.

203


PART III TREATMENT OF COMPLICATED PANCREATITIS


CHAPTER 11 A conservative and minimally invasive approach to necrotizing pancreatitis improves outcome Gastroenterology 2011

Hjalmar C van Santvoort, Olaf J Bakker, Thomas L Bollen, Marc G Besselink, Usama Ahmed Ali, A Marjolein Schrijver, Marja A Boermeester, Harry van Goor, Cornelis H Dejong, Casper H van Eijck, Bert van Ramshorst, Alexander F Schaapherder, Erwin van der Harst, Sijbrand Hofker, Vincent B Nieuwenhuijs, Menno A Brink, Philip M Kruyt, Eric R Manusama, George P van der Schelling, Tom Karsten, Eric J Hesselink, Cornelis J van Laarhoven, Camiel Rosman, Koop Bosscha, Ralph J de Wit, Alexander P Houdijk, Miguel A Cuesta, Peter J Wahab, and Hein G Gooszen, for the Dutch Pancreatitis Study Group


PART III CHAPTER 11

A B S T R AC T B ackg round & Aims In recent years, treatment of necrotizing pancreatitis has changed towards more conservative and less invasive strategies but prospective data are lacking. We describe treatment results in a prospective multicenter cohort covering the entire clinical spectrum of necrotizing patients. Metho ds From 2004 to 2008, 639 consecutive patients with necrotizing pancreatitis from 21 Dutch hospitals were prospectively included in this observational cohort study. Data on disease severity, interventions (i.e., radiological, endoscopical, surgical) and outcome were recorded. Resu lts Overall mortality was 15% (n=93). Organ failure occurred in 240/639 patients (38%), with 35% mortality. Treatment was conservative in 397/639 patients (62%) with 7% mortality. An intervention was performed in 242/639 patients (38%), with 27% mortality: this included early emergency laparotomy in 32/639 patients (5%), with 78% mortality. The longer the time between admission and intervention, the lower the risk of mortality: 0-14 days; 56%, 14-29 days; 26% and >29 days; 15%, P<0.001. 208/639 patients (33%) underwent intervention for infected necrosis, with 19% mortality. Catheter drainage was most often performed (63%) as first intervention, without additional necrosectomy in 35% of patients. Primary catheter drainage had fewer complications than primary necrosectomy (42% vs. 64%, P=0.003). Patients with pancreatic parenchymal necrosis (n=324), as compared to patients with peripancreatic necrosis alone (n=315), had a higher risk of organ failure (50% vs. 24%, P<0.001) and mortality (20% vs. 9%, P<0.001). C onclusi ons Two-thirds of patients with necrotizing pancreatitis can be treated without an intervention and have relatively low mortality. In patients with infected necrosis, delayed intervention and catheter drainage as first treatment improve outcome.

206


CONSERVATIVE APPROACH TO NECROTIZING PANCREATITIS

I N T RODU C T ION Acute pancreatitis is complicated by necrosis of the pancreatic parenchyma and/or the peripancreatic fat tissue in around 20% of patients.1,2 The clinical course of necrotizing pancreatitis can be divided into two phases. In the first phase (i.e., 1-2 weeks after onset of symptoms), a systemic inflammatory response syndrome (SIRS) occurs, which is often associated with multiple organ failure.3-5 Interventions in the first phase are relatively contra-indicated, although some patients require an emergency laparotomy for complications such as abdominal compartment syndrome or bowel ischemia.3 It has been suggested that around half the deaths from necrotizing pancreatitis are caused by multiple organ failure in the early phase.6,7 In the late phase of the disease (i.e., after 1-2 weeks), systemic inflammation often regresses and infected necrosis occurs in about 30% of patients with necrotizing pancreatitis.8,9 Without radiological, endoscopical or surgical intervention, infected necrosis ultimately leads to death in nearly every patient. Over the last years, treatment of necrotizing pancreatitis has changed considerably. First, the indication for intervention has shifted. Whereas historically most patients with sterile necrosis underwent necrosectomy, it is now accepted that sterile necrosis should largely be managed conservatively and that the main indication for intervention is infected necrosis.3,10-12 Second, the timing of intervention has changed. Necrosectomy was once performed at a very early stage,13 but now it is thought that intervention should be delayed to around 3-4 weeks after onset of disease.14-16 This timeframe allows for encapsulation of peripancreatic collections, which may improve conditions for intervention and thereby decrease the risk of complications such as bleeding and perforation. Third, as an alternative to the historical standard of primary open necrosectomy17, minimally invasive techniques, such as percutaneous catheter drainage and minimally invasive necrosectomy, are increasingly used.18-24 The outcome of patients with necrotizing pancreatitis after the recent implementation of the abovementioned changes is unknown. Although several cohort studies on necrotizing pancreatitis were published over the last decades9, these data are outdated and of limited clinical relevance for several reasons. First, these studies mostly did not include patients with peripancreatic necrosis only (i.e., without pancreatic parenchymal necrosis), whereas in the updated Atlanta classification peripancreatic necrosis is also included under necrotizing pancreatitis.25 Second, most of the reviewed studies were performed in a time when the abovementioned changes regarding intervention had not yet fully occurred. Third, most reviewed studies are small retrospective case-series. Although there are some larger studies, with sample sizes ranging from 281 to 392 patients, these studies cover long periods (range 12 to 19 years).26-28 Fourth, as most studies report only the subgroup of patients who underwent operative necrosectomy22,27,29,30, data on the outcome of conservative treatment of necrotizing pancreatitis and outcome 207


PART III CHAPTER 11

of emergency laparotomy are scarce. Fifth, although sterile necrosis should be managed conservatively, many of the reviewed studies report low proportions of infected necrosis (range 63 to 74%).22,29,30 Sixth, whereas percutaneous drainage alone may be sufficient treatment in a proportion of patients, the percentage of patients undergoing percutaneous drainage prior to necrosectomy is low (range 5 to 30%) or not even reported.12,26,29-31 Finally, the vast majority of data comes from highly experienced single centers.12,26,27,29-31 It is questionable whether these results can be extrapolated to daily practice in non-expert centers. Given all the above, new data from large prospective multicenter studies are needed to serve as a reference for the current outcome of necrotizing pancreatitis. The aim of this study was to evaluate outcome of conservative and interventional treatment in a prospective, nationwide cohort of 639 patients covering the entire clinical spectrum of necrotizing pancreatitis in recent years. We also analyzed outcome in disease subgroups based on of organ failure, pancreatic or peripancreatic necrosis and status of infection.

M E T HOD S Pati ents and stu dy d esig n We performed a prospective observational cohort study between March 2004 and November 2008 in all 8 Dutch university medical centers and 13 large teaching hospitals of the Dutch Pancreatitis Study Group. The study aims, outcomes, definitions and endpoints were predefined. During the study period, all patients admitted with acute pancreatitis were registered in a prospective database and were screened for eligibility for the PROPATRIA and PANTER randomized trials.24,32 Regardless of eligibility for the trials, patients were included in the current prospective study if they met the following inclusion criteria: signs of pancreatic necrosis and/or peripancreatic necrosis on contrastenhanced computed tomography (CECT).33 This is the first time this complete cohort is reported. The study was conducted in accordance with the principles of the Declaration of Helsinki. The ethics review board of each participating hospital approved the study. Patients or their legal representatives gave written informed consent. We adhered to the STROBE guidelines for reporting observational cohort studies.34 C ompute d tomog r aphy CECT was performed per protocol in the patients randomized in the PROPATRIA study 7-10 days after admission and on demand therefore and thereafter.32 In all other patients, CECT was performed by discretion of the treating physician, but generally only in case there was no clinical improvement after the initial 7-10 days of admission. Local radiologists judged the CECT for pancreatic necrosis and/or peripancreatic 208


CONSERVATIVE APPROACH TO NECROTIZING PANCREATITIS

necrosis, and based on their evaluation, patients were prospectively included. However, after the study was completed, all digitalized CECTs were reviewed by an experienced abdominal radiologist (TLB) who was unaware of the clinical background and treatment. He made the final decision on inclusion or exclusion based on the highest CT-severity index33 measured on all CECTs performed during admission. Median time between onset of symptoms and the final CECT before discharge was 14 days (interquartile range [IQR] 7-47). Tre atment proto col The following treatment was protocolized: following admission, patients received rigorous fluid resuscitation and full laboratory investigations were performed on the first three days and in the 24 hours prior to any intervention. Nasojejunal enteral feeding was initiated if an oral diet was not tolerated. Parenteral nutrition was only initiated when enteral nutrition was not tolerated or not sufficient. Half the patients randomized in the PROPATRIA trial received probiotic prophylaxis (n=105).32 Antibiotics were administered only in the presence of suspected or documented infection. Patients with (pending) organ failure were treated in the intensive care unit. In line with international guidelines, intervention was generally only performed in case of suspected or confirmed infection of pancreatic necrosis or peripancreatic necrosis alone.3,10,11 There was no routine fine needle aspiration of peripancreatic collections. The decision to intervene was mostly based on clinical grounds (i.e. deterioration). Whenever possible, intervention was postponed until approximately 4 weeks after the onset of disease. In case of proven or suspected infection of pancreatic necrosis within 4 weeks after onset of disease, antibiotics were administered and intervention was postponed as long as possible.14-16 Some patients underwent emergency laparotomy early in the course of disease. The indication was severe clinical deterioration suspected to be caused by abdominal compartment syndrome, bowel ischemia, or perforation of a visceral organ. Early 2006, we installed a multidisciplinary expert panel consisting of eight gastrointestinal surgeons, one gastroenterologist, and three radiologists to guide decisions on intervention. Whenever infected necrosis was suspected or there was any other indication for intervention, the expert panel received a case description, including CT images, on a standardized form by e-mail. Within 24 hours, the members of the expert panel individually advised on the indication for intervention and possible methods of interventions. The ultimate decision for intervention was made by the treating physician. Intervention could be primary necrosectomy, primary percutaneous catheter drainage with 12 to 14 French drains18 or endoscopic transluminal (gastric wall or duodenal wall) catheter drainage,21 with or without subsequent necrosectomy. Necrosectomy consisted of open necrosectomy by laparotomy and continuous postoperative lavage,17 or minimally invasive necrosectomy: video-assisted retroperitoneal debridement (VARD)20 or 209


PART III CHAPTER 11

endoscopic transluminal necrosectomy (ETN).21 A subgroup of patients included in the PANTER trial24 was randomly assigned to either primary open necrosectomy (n=45), or a protocolized step-up approach which included primary catheter drainage, followed, if necessary, by VARD (n=43). D ef initi ons Pancreatic necrosis was defined as a CT-severity index of greater than 4, which means focal nonenhancement of the pancreatic gland. Peripancreatic necrosis alone (i.e., without pancreatic parenchymal necrosis) was defined as a CT-severity index of 3 or 4, which means peripancreatic morphological changes exceeding fat stranding. In the most recent draft of the revised Atlanta Classification25, the definition of necrotizing pancreatitis includes both pancreatic parenchymal necrosis with or without peripancreatic necrosis, and peripancreatic necrosis alone.

Table 1. Definitions of complications. Complication

Definition

Infected necrosis

A positive culture of pancreatic necrosis or peripancreatic necrosis obtained by means of fine-needle aspiration or from the first drainage procedure or necrosectomy, or the presence of gas in the peripancreatic collection on contrast-enhanced CT

Organ failure

210

Pulmonary failure

PaO2 <60 mm Hg, despite FIO2 of 0.30, or need for mechanical ventilation

Circulatory failure

Circulatory systolic blood pressure <90 mm Hg, despite adequate fluid resuscitation, or need for inotropic catecholamine support

Renal failure

Creatinine level >177 Îźmol/liter after rehydration or new need for hemofiltration or hemodialysis

Multiple organ failure

Failure of at least two organ systems on the same day

Persistent organ failure

Presence of organ failure on at least 3 consecutive days (lasting more than 48 hours)

New-onset multiple organ failure

New-onset failure (i.e., not present at any time in the 24 hours before first intervention) of two or more organs

Enterocutaneous fistula

Secretion of fecal material from a percutaneous drain or drainage canal after removal of drains or from a surgical wound, either from small or large bowel; confirmed by imaging or during surgery

Perforation of visceral organ

Perforation requiring surgical, radiological, or endoscopic intervention

Intraabdominal bleeding

Bleeding requiring surgical, radiological, or endoscopic intervention


CONSERVATIVE APPROACH TO NECROTIZING PANCREATITIS

D ata col l e c ti on Data from patients randomized in the PROPATRIA and PANTER trial were collected as described previously.24,32 In addition, data from the other patients were entered prospectively into a database. This included data on patient demographics, laboratory investigations in the first three days of admission and 24 hours prior to intervention, clinical course, microbiology cultures, and interventions. If a patient had been referred from another center, all CECTs, clinical data, and laboratory investigations were retrieved. O utcomes The primary study outcome was mortality during index admission. Readmission within 10 days after discharge from index admission was considered a prolonged index admission. Secondary outcomes during admission were organ failure, infected necrosis, the number of drainage procedures, endoscopic and surgical necrosectomies, other operations, and complications (Table 1). Statisti c a l ana lysis Analyses were performed with SPSS version 15.0 (Chicago, IL). The KolmogorovSmirnov test was used to assess whether continuous data were normally distributed (P>0.05). Continuous data are presented as mean Âą standard deviation (SD) and in case of non-normal distributions as median with IQR. In case of missing data, patients were not included in the analyses for the specific parameter and this is reported. Differences were tested by the Mann-Whitney, Kruskal-Wallis, or Linear-by-linear association tests for ordinal variables. Proportions were compared by the chi-square test. When comparing different time points of intervention and methods of intervention (i.e., drainage or necrosectomy), in order to reduce potential selection bias, we used multivariate logistic regression to adjust for prognostic baseline variables. These variables were: age, CTseverity index, ASA-class, APACHE-II score, and organ failure. Results of the adjusted analyses are presented as odds ratios (OR) with corresponding 95%-confidence intervals (CI). A two-sided P<0.05 was considered statistically significant.

R E SU LT S During the 4.5 years study period, 639 patients with pancreatic necrosis or peripancreatic necrosis alone were included. No data were missing with regard to the primary and secondary study outcomes. Patient characteristics on admission and details from the CECTs prior to any intervention are given in Table 2. Overall mortality was 15% (93/639). Table 3 lists mortality for the several treatment and disease subgroups that will be consecutively presented. 211


PART III CHAPTER 11

Table 2. Characteristics of 639 patients with necrotizing pancreatitis Characteristic

All patients (n=639)

Conservative treatment Intervention* (n=397) (n=242)

Age – yr

58 (45-70)

57 (43-71)

59 (48-69)

0.41a

Male sex

398 (62%)

237 (60%)

161 (67%)

0.08b

Etiology

P value

0.41b

Gallstones

304 (48%)

Alcohol abuse

150 (24%)

192 (48%) 95 (24%)

112 (46%) 55 (23%)

Other

63 (10%)

42 (11%)

21 (9%)

Unknown

122 (19%)

68 (17%)

54 (22%)

ASA-class on admission

0.02c

I (healthy status)

202 (32%)

136 (34%)

66 (27%)

II (mild systemic disease)

347 (54%)

214 (54%)

133 (55%)

III (severe systemic disease)

90 (14%)

47 (12%)

43 (18%)

8 (5-11)

7 (5-10)

8 (5-11)

0.25a

324 (51%)

192 (48%)

132 (55%)

0.13b

Imrie/modified Glasgow score (48 hrs of admission)

3 (2-5)

3 (2-4)

4 (3-5)

<0.001a

Imrie/modified Glasgow score ≥ 3

413 (65%)

225 (57%)

188 (78%)

<0.001b

Highest CRP level in first 48 hrs of admission (mg/L)

291 (±130)

287 (±133)

294 (±129)

0.51a

513 (80%)

318 (80%)

195 (81%)

0.42b

44 (7%)

32 (8%)

12 (5%)

CT-severity index

4 (4-8)

4 (4-6)

8 (6-10)

<0.001a

Pancreatic necrosis

324 (51%)

139 (35%)

185 (76%)

<0.001b

Peripancreatic necrosis alone

315 (49%)

258 (65%)

57 (24%)

<0.001b

Predicted severity of pancreatitis APACHE-II score on admission APACHE-II score > 8 on admission

CRP >150 (mg/L) No CRP data available#

Extent of pancreatic necrosis

<0.001c

<30%

447 (70%)

329 (83%)

118 (49%)

30% to 50%

83 (13%)

25 (6%)

58 (24%)

>50%

109 (17%)

43 (11%)

66 (27%)

156 (24%)

39 (10%)

117 (48%)

Transferred from other hospital

<0.001b

Continuous variables are mean (±SD) or median (interquartile range). #CRP = C-reactive protein. It was not possible to measure CRP in 1 hospital. APACHE = Acute Physiology And Chronic Health Evaluation. ASA = American Society of Anaesthesiologists. aMann-Whitney. bChi-square. cLinear-by-linear association

212


CONSERVATIVE APPROACH TO NECROTIZING PANCREATITIS

C ons er vative tre atment Conservative treatment (i.e., without any radiological, endoscopical or surgical intervention) was performed in 397/639 patients (62%). Characteristics on admission for these patients are shown in Table 2. Overall mortality was 7% (28/397). The group of 397 patients treated conservatively included 63 patients (16%) with organ failure and 11 patients (3%) with infected necrosis. The latter 11 patients were only treated with intravenous antibiotics because of their extraordinary good clinical condition in the absence of sepsis or organ failure. Mortality in the patients with organ failure and conservative treatment was 37% (23/63) and mortality in the 11 patients with infected necrosis and conservative treatment was 0%.

Table 3. Mortality in the different subgroups of the 639 patients with necrotizing pancreatitis. Subgroups All patients with necrotizing pancreatitis

Mortality 93/639 (15%)

Pancreatic necrosis

64/324 (20%)

Peripancreatic necrosis alone

29/315 (9%)

No organ failure

8/399 (2%)

Organ failure At any time during admission

85/240 (35%)

At any time during admission, persistent

77/213 (36%)

In the first week of admission

57/140 (41%)

Multiple organ failure At any time during admission

79/194 (41%)

At any time during admission, persistent

66/161 (41%)

In the first week of admission

44/94 (47%)

Infected or sterile necrosis Primary infected necrosis

41/202 (20%)

Sterile necrosis

52/437 (12%)

Conservative treatment or intervention Conservative treatment

28/397 (7%)

Any intervention (i.e., emergency laparotomy, drainage, necrosectomy)

65/242 (27%)

Emergency laparotomy

25/32 (78%)

Any intervention for suspected or confirmed infected necrosis

40/208 (19%)

Catheter drainage as first intervention

26/130 (20%)

Necrosectomy as first intervention

14/78 (18%)

Necrosectomy (with or without previous drainage or emergency laparotomy)

41/169 (24%)

Any operation (i.e., necrosectomy or emergency laparotomy)

56/187 (30%)

213


PART III CHAPTER 11

Inter venti ona l tre atment An intervention was performed in 242/639 patients (38%). A flowchart of patients undergoing intervention is shown in Figure 1. As compared to patients treated conservatively, patients undergoing intervention had higher ASA-class, Imrie score, CT-severity index and extent of pancreatic necrosis on admission (Table 2). During admission, 177/242 patients (73%) developed organ failure and 189/242 patients (78%) developed infected necrosis. Mortality in patients undergoing intervention was 27% (65/242). Table 4 shows the baseline criteria for patients undergoing intervention, subdivided in 3 time periods. The longer the time between admission and intervention, the lower the risk of mortality: 0-14 days; 56% (25/45), 14-29 days; 26% (25/98) and >29 days; 15% (15/99), P<0.001. This did not change when adjusted for baseline prognostic factors (age, APACHE-II, CT-severity index, organ failure, ASA-class): adjusted OR 0.38, 95%-CI 0.24-0.59, P<0.001. Early emergenc y laparotomy An emergency laparotomy was necessary in 32/639 patients (5%), at a median of 5 days (IQR 2-14) after admission. These laparotomies were not performed for suspected or confirmed infected necrosis, but because an abdominal catastrophe for another reason. There were signs of abdominal compartment syndrome in 15 patients and bowel ischemia was observed during laparotomy in 11 patients. Of the 32 patients undergoing emergency laparotomy, 25 patients (78%) died. Inter ventions for infec ted necrosi s A radiological, endoscopical or surgical intervention for suspected or confirmed infected necrosis was performed in 208/639 patients (33%). The cultures of the first intervention were positive in 89% (185/208). Details on disease severity at the time of intervention are given in Table 5. Median time between onset of symptoms and intervention was 28 days (IQR 22-41). The first intervention was most often percutaneous (or endoscopic transluminal) catheter drainage (63%, n=130). Primary necrosectomy was performed in 78/208 patients (38%). Half of patients (n=104) undergoing intervention for infected necrosis suffered from one or more complications. New-onset organ failure occurred in 83/208 patients (40%), intra-abdominal bleeding in 33/208 patients (16%), and enterocutaneous fistula or perforation of a visceral organ in 35/208 patients (17%). The longer the time between admission and intervention, the lower the risk of complications: 0-14 days; 72% (13/18), 14-29 days; 57% (53/93), and >29 days; 39% (38/97), P=0.007. This did not change when adjusted for baseline variables: adjusted OR 0.46, 95%-CI 0.29-0.73, P=0.001.

214


CONSERVATIVE APPROACH TO NECROTIZING PANCREATITIS

Necrotizing pancreatitis N=639

Conservative treatment

Radiological, endoscopical or surgical intervention

397/639 (62%)

242/639 (38%)

Early emergency laparotomy

Intervention for suspected/ confirmed infected necrosis

32/639 (5%)

208/639 (33%)

Intervention for gastric outlet obstruction from sterile necrosis 2/639 (0.3%)

Catheter drainage

ETN

1/2 (50%)

1/2 (50%)

Primary catheter drainage

Primary necrosectomy

130/208 (63%)

78/208 (38%)

No additional necrosectomy

Additional necrosectomy

54/130 (42%)

76/130 (58%)

Laparotomy

VARD

ETN

68/78 (87%)

6/78 (8%)

4/78 (5%)

Laparotomy

VARD

ETN

25/76 (33%)

44/76 (58%)

7/76 (9%)

Figure 1. Flowchart on treatment strategies in 639 patients with necrotizing pancreatitis. 29

215


PART III CHAPTER 11

In the 130 patients undergoing catheter drainage as first intervention (percutaneous: n=113, endoscopic transluminal: n=17), a median of 1 drainage procedure (IQR 1-2) was performed per patient: 75/130 patients (58%) underwent 1 drainage procedure, 42/130 patients (32%) underwent 2 drainage procedures and 13/130 patients (10%) had more than 2 drainage procedures. Forty-five of the 130 patients (35%) who underwent primary drainage were successfully treated without any necrosectomy and left the hospital in good clinical condition. Nine of 130 patients (7%) treated with primary catheter drainage died without undergoing any other intervention. These patients were not considered candidates for surgery due to poor clinical condition and co-morbidity. Following primary catheter drainage, after a median of 10 days (IQR 5-22), 76/130 patients (58%) underwent additional necrosectomy (laparotomy: n=25, VARD: n=44, ETN: n=7). Mortality was 22% (17/76). There were no significant differences in mortality

Table 4. Baseline characteristics for patients who underwent early and late interventions Characteristic

Time from admission to intervention 0-14 days (n=45)

14-29 days (n=98)

More than 29 days (n=99)

P value

Age – yr

58 (±14)

58 (±13)

58 (±15)

0.98a

Male sex

29 (64%)

66 (67%)

66 (67%)

0.94b

Etiology

0.24b

Gallstones

14 (31%)

47 (48%)

51 (52%)

Alcohol abuse

15 (33%)

23 (24%)

17 (17%)

Other Unknown

3 (7%)

8 (8%)

10 (10%)

13 (29%)

20 (20%)

21 (21%)

ASA-class on admission

0.71c

I (healthy status)

11 (24%)

26 (27%)

29 (29%)

II (mild systemic disease)

26 (58%)

55 (56%)

52 (53%)

III (severe systemic disease)

8 (18%)

17 (17%)

18 (18%)

CT-severity index

6 (5-10)

8 (4-8)

8 (6-10)

0.76a

Pancreatic necrosis

35 (78%)

70 (71%)

80 (81%)

0.29a

Peripancreatic necrosis alone

10 (22%)

28 (29%)

19 (19%)

0.29a

Extent of pancreatic necrosis

0.87c

<30%

24 (53%)

49 (50%)

30% to 50%

7 (16%)

30 (31%)

21 (21%)

>50%

14 (31%)

23 (25%)

29 (29%)

APACHE-II at time of intervention

14 (10-21)

13 (9-19)

13 (8-17)

0.14a

Organ failure at time of intervention

12 (27%)

46 (47%)

27 (27%)

0.006b

Kruskal-Wallis. Chi-square. Linear-by-linear association.

a

216

45 (46%)

b

c


CONSERVATIVE APPROACH TO NECROTIZING PANCREATITIS

Table 5. Characteristics of 208 patients undergoing an intervention for infected necrosis. Characteristic APACHE-II score* CRP level*

Intervention for infected necrosis (n=208) 13 (9-17) 200 (Âą109)

Admitted to the Intensive Care Unit At the time of intervention

75 (36%)

Anytime before intervention

140 (67%)

Organ failure At the time of intervention

76 (37%)

Anytime before intervention

113 (54%)

Multiple organ failure At the time of intervention

46 (22%)

Anytime before intervention

84 (40%)

Infected necrosis confirmed by intra-operative culture

185 (89%)

Continuous variables are mean (ÂąSD) or median (interquartile range). * Based on the maximal values 24 hours before intervention, CRP not available in 46 patients (n=46). APACHE = Acute Physiology And Chronic Health Evaluation

or the number of patients with complications for the different techniques of necrosectomy (data not shown). Seventy-eight patients underwent necrosectomy as first intervention (laparotomy: n=68, VARD: n=6, or ETN n=4). Mortality was 18% (14/78) and this did not differ significantly for the different techniques of necrosectomy (data not shown). The number of patients with complications was significantly greater in patients undergoing laparotomy, as compared to VARD and ETN: 71% (48/68) vs. 33% (2/6) vs. 0% (0/4) respectively, P=0.004. Fewer complications occurred in patients undergoing catheter drainage as first intervention than in patients undergoing primary necrosectomy: 42% (54/130) vs. 64% (50/78), respectively, P=0.003, adjusted OR 0.37, 95%-CI 0.19-0.67, P=0.001. This difference was mainly driven by a lower incidence of new onset multiple organ failure after primary drainage: 17% (22/130) vs. 31% (24/78), P=0.02. There was no difference in mortality: 20% (26/130) vs. 18% (14/78), P=0.71. Dis e as e subg roups O rgan failure Organ failure during admission occurred in 240/639 patients (38%). Mortality occurred almost exclusively in patients with organ failure, as compared to patients without organ failure: 35% (85/240) vs. 2% (8/399), P<0.001. 217


PART III CHAPTER 11

The vast majority of patients with organ failure suffered from persistent (>48 hours) organ failure: 213/240 (89%). Mortality for patients with transient organ failure and persistent organ failure did, however, not differ significantly: 30% (8/27) vs. 36% (77/213), P=0.51. Multiple organ failure occurred in 194/639 patients (30%), and persistent multiple organ failure in 161/639 patients (25%). Mortality for transient and persistent multiple organ failure was 40% (13/33) and 41% (66/161), respectively, P=0.87. More than half of the cases (140/240, 58%) of organ failure occurred within the first week of admission. Mortality was higher in patients with organ failure in the first week of admission as compared to patients with organ failure after the first week of admission: 41% (57/140) vs. 28% (28/100), P=0.04. Multiple organ failure in the first week was associated with 47% (44/94) mortality. Ty pe of necrosi s On CECT, 324/639 patients (51%) had signs of pancreatic parenchymal necrosis and 315/639 patients (49%) patients had signs of peripancreatic necrosis alone. Organ failure occurred more often in patients with pancreatic necrosis as compared to patients with peripancreatic necrosis alone: 50% (163/324) vs. 24% (77/315), P<0.001. Mortality was also greater in patients with pancreatic necrosis than in patients with peripancreatic necrosis alone: 20% (64/324) vs. 9% (29/315), P<0.001. Infec ted necrosi s Primary infection of necrosis was diagnosed in 202/639 patients (32%), at a median of 26 days (IQR 18-38) after admission. The incidence of infected necrosis was higher in patients with pancreatic necrosis as compared to patients with peripancreatic necrosis alone: 47% (151/324) vs. 16% (51/315), P<0.001.

DI S C U S SION This is the largest prospective cohort of patients with necrotizing pancreatitis who underwent either conservative treatment or an intervention reported thus far. Compared to most of the earlier studies, this study included a large number of patients in a relatively short study period, it was conducted in a nationwide multicenter setting and covered the entire clinical spectrum of necrotizing pancreatitis. We demonstrated that treatment can be conservative in about two-thirds of patients with necrotizing pancreatitis. In case of infected necrosis, one-third of patients can be successfully treated with percutaneous (or endoscopic transluminal) catheter drainage, and do not need any form of necrosectomy. If catheter drainage is unsuccessful, minimally invasive retroperitoneal or endoscopic transluminal necrosectomy are safe and feasible techniques. 218


CONSERVATIVE APPROACH TO NECROTIZING PANCREATITIS

We confirmed that around half of the patients with necrotizing pancreatitis who die have sterile necrosis. Mortality in these patients is almost exclusively caused by multiple organ failure in the first week.6,7 There is currently no effective treatment to improve outcome in these patients. Intervention is generally contra-indicated.3 However, in 32 patients (5%) in the current study an early emergency laparotomy was performed. As to be expected, mortality was high (78%). Fifteen of these patients were suspected of an abdominal compartment syndrome. There are only a few other studies on emergency laparotomy for abdominal compartment syndrome early in the course of acute pancreatitis.35-38 These retrospective studies included 3 to 27 patients, with mortality varying from 30 to 75%. Interestingly, an 2007 international consensus conference suggests that catheter drainage of intra-abdominal fluid can be used as the initial step to decrease intra-abdominal pressure in patients with abdominal compartment syndrome in general.39 This strategy may improve outcome and is currently being evaluated in acute pancreatitis (ClinicalTrials.gov NCT00793715). Organ failure occurred in 38% of patients, with 35% mortality. This is in line with a recent meta-analysis of 14 cohorts of both interstitial and necrotizing pancreatitis, showing overall mortality after organ failure of 30%.40 Several studies have suggested that early persisting organ failure rather than transient organ failure drives mortality in acute pancreatitis.5,6,41 Interestingly, in the current study, the vast majority of patients with organ failure (89%) or multiple organ failure (83%) had persistent organ failure (i.e., lasting for more than 48 hours). Mortality in patients with organ failure in the first week was significantly higher than in patients with organ failure occurring after the first week (41% vs. 28%). This supports the theory that organ failure early in the course of acute pancreatitis, which is associated with systemic release of cytokines and SIRS, is a different clinical entity than organ failure as a result of sepsis from infected necrosis at a later stage. Over a third of patients with necrotizing pancreatitis in our study underwent an intervention, which was associated with 27% mortality. Several guidelines3,9,15 now advise to withhold intervention for several weeks after onset of symptoms, although there is only limited evidence for this from previous studies. A randomized study from 1997 suggested that delaying surgical intervention beyond the first 12 days, as compared to intervention as early as in the first 72 hours of admission, reduces mortality.13 Two retrospective studies14,16 and one systematic review16 have suggested that postponing surgery for approximately 4 weeks after admission further improves outcome. The current study yields further evidence in favour of delaying intervention whenever possible, as we found that morbidity and mortality decreased considerably if intervention was postponed. Notably, this is the first time that improved outcome for delayed intervention is reported in an analysis adjusted for potential confounders regarding patient characteristics and disease severity (i.e., age, ASA-class, organ failure, APACHE-II score, CT-severity index). 219


PART III CHAPTER 11

Infection of necrosis occurred in around 30% of patients with necrotizing pancreatitis. This shows that the incidence of infected necrosis has remained fairly constant over the last 20 years.9 Mortality in infected necrosis in our study was 20%. This seems to be lower than the mortality of around 30% for infected necrosis reported in reviews of the literature of the last two decades.9,40 However, our results are difficult to compare with the literature because most published reports present only selected subgroups of patients with infected necrosis (e.g., only those who underwent necrosectomy, excluding emergency laparotomy). Moreover, large, prospective, unselected cohort-studies on infected necrosis are rare.42 In the current cohort of 639 patients, 33% underwent an intervention for suspected or confirmed infected necrosis. Around a third of patients with infected necrosis were successfully managed with catheter drainage only. A median of 1 drainage procedure with normal sized drains (i.e., 12-14 French) was sufficient. In these patients, decompression of the infected collection with fluid/pus seems to be enough for full recovery, meaning that major abdominal surgery is not necessary. Patients who were treated with the step-up approach, using drainage as primary intervention and only followed by minimally invasive necrosectomy if needed, also had fewer complications than patients undergoing primary necrosectomy. These results are in line with our recent multicenter PANTER trial.24 We advise that all patients with suspected or confirmed infected necrosis are treated with percutaneous or endoscopic catheter drainage first. In the patients undergoing primary necrosectomy, minimally invasive necrosectomy was associated with fewer complications. There are only few other studies that directly compared minimally invasive necrosectomy with open necrosectomy.22,43 In a retrospective case-matched study in 30 patients, minimally invasive retroperitoneal necrosectomy was associated with a marked reduction in postoperative multiple organ failure.43 A recent retrospective monocenter study compared 137 patients who underwent minimally invasive retroperitoneal necrosectomy to 52 patients who underwent open necrosectomy.22 Complications and mortality were significantly lower after minimally invasive retroperitoneal necrosectomy than after open necrosectomy: 55% vs. 81% and 19% vs. 38%, respectively. These studies confirm the hypothesis that minimally invasive necrosectomy induces less surgical stress (i.e., a pro-inflammatory immune response) in these already critically ill patients. Next to minimally invasive surgical techniques, endoscopic necrosectomy through the stomach or duodenum is also gaining popularity.21,44,45 The two largest and most recent case-series have shown that ETN is safe, feasible and seems to compare favourable to the literature on surgical necrosectomy with complication rates of around 14% to 49% and mortality of 6% to 7.5%.44,45 In the current study, ETN was performed as primary intervention in 7 patients and was associated with no complications and no deaths. However, low patients numbers and possible selection bias may explain these favourable results. Results from the first randomized trial 220


CONSERVATIVE APPROACH TO NECROTIZING PANCREATITIS

comparing ETN with surgical necrosectomy are awaited (PENGUIN trial; ISRCTN07091918). Mortality in the subgroup of patients undergoing an intervention for suspected or confirmed infected necrosis in our study was 19%. This is comparable to the 24% mortality reported from Liverpool, UK,22 and the 15% mortality reported from Boston, both expert centers, in patients with infected necrosis.29 In our study, 5% of patients with infected necrosis were successfully managed without any intervention. This has been previously reported in case-reports and small case-series.46-49 Although our data confirm that a very small subset of patients with infected necrosis can be treated with antibiotics alone, infected necrosis should still be considered an indication for intervention in the overwhelming majority of patients. We recommend close monitoring of patients who have an extraordinarily good clinical condition that might justify conservative treatment. Patients with pancreatic parenchymal necrosis had significantly higher mortality than patients with peripancreatic necrosis alone: 20% vs. 9%. There is only one other study that previously compared these two subgroups, and it showed similar mortality rates as our study.50 Although the outcome of patients with parenchymal necrosis was substantially worse, patients with peripancreatic necrosis alone still had a considerable risk of organ failure (24%), infected necrosis (16%), and need for intervention (23%). The subgroup of patients in whom CECT shows normally enhancing pancreatic parenchyma but signs of peripancreatic necrosis should therefore also be monitored carefully. This study has potential shortcomings. First, during the study period, CECT was performed in patients admitted with acute pancreatitis by discretion of the treating physician (with the exception for the 296 patients in the PROPATRIA study who underwent CECT per protocol). Hence, there may be a small group of patients in whom CECT was not performed because either their clinical condition was exceptionally good or they were too ill to undergo CECT. These patients may have been missed for inclusion. Second, selection bias may have occurred for the comparative analyses on the different interventions and timing of intervention, as this study did not have a randomized design, even though our analyses adjusting for baseline prognostic factors did not confirm this. Third, we classified patients as having â&#x20AC;&#x2DC;peripancreatic necrosis aloneâ&#x20AC;&#x2122; if the highest CTseverity index33 measured during admission was 3 or 4. One can argue that this definition is not 100% accurate, especially as it is has been suggested that peripancreatic necrosis can not be reliably diagnosed by CECT.51,52 However, several studies demonstrated a good correlation between peripancreatic findings on CECT and the presence of fat necrosis at operation or autopsy.53-55 Moreover, the median time between onset of symptoms and the last CECT in our study was 14 days. If a patient demonstrates a heterogeneous peripancreatic collection on CECT (i.e., corresponding to a CT-severity index of 3-4) two weeks into the disease, this should be considered fat necrosis until proven otherwise.54 Notably, in all patients with signs of peripancreatic necrosis alone on CECT in whom 221


PART III CHAPTER 11

necrosectomy was performed, peripancreatic fat necrosis was found during operation. In conclusion, this large and prospective multicenter study demonstrated that about two-thirds of patients with necrotizing pancreatitis can be treated conservatively with relatively low mortality. Early emergency laparotomy and multiple organ failure, however, remain associated with high mortality. Patients with infected necrosis benefit from postponing intervention and minimally invasive techniques. A third of these patients can be treated with percutaneous (or endoscopic transluminal) catheter drainage only and do not require necrosectomy.

222


CONSERVATIVE APPROACH TO NECROTIZING PANCREATITIS

REFERENCES 1. Beger HG, Rau B, Mayer J, et al. Natural course of acute pancreatitis. World J Surg 1997;21:130-135. 2. Bradley EL, III. A clinically based classification system for acute pancreatitis. Summary of the International Symposium on Acute Pancreatitis, Atlanta, Ga, September 11 through 13, 1992. Arch Surg 1993; 128:586-590. 3. Nathens AB, Curtis JR, Beale RJ, et al. Management of the critically ill patient with severe acute pancreatitis. Crit Care Med 2004; 32:2524-2536. 4. Werner J, Feuerbach S, Uhl W, et al. Management of acute pancreatitis: from surgery to interventional intensive care. Gut 2005;54:426-436. 5. Mofidi R, Duff MD, Wigmore SJ, et al. Association between early systemic inflammatory response, severity of multiorgan dysfunction and death in acute pancreatitis. Br J Surg 2006;93:738-744. 6. Johnson CD, Abu-Hilal M. Persistent organ failure during the first week as a marker of fatal outcome in acute pancreatitis. Gut 2004; 53:1340-1344. 7. Blum T, Maisonneuve P, Lowenfels AB, et al. Fatal outcome in acute pancreatitis: its occurrence and early prediction. Pancreatology 2001;1:237-241. 8. Besselink MG, Van Santvoort HC, Boermeester MA, et al. Timing and impact of infections in acute pancreatitis. Br J Surg 2009;96:267-273. 9. Banks PA, Freeman ML. Practice guidelines in acute pancreatitis. Am J Gastroenterol 2006;101:2379-2400. 10. UK guidelines for the management of acute pancreatitis. Gut 2005;54 Suppl 3:iii1-iii9. 11. Forsmark CE, Baillie J. AGA Institute Technical Review on Acute Pancreatitis. Gastroenterology 2007;132:2022-2044.

12. Buchler MW, Gloor B, Muller CA, et al. Acute necrotizing pancreatitis: treatment strategy according to the status of infection. Ann Surg 2000;232:619-626. 13. Mier J, Luque-de Le贸n E, Castillo A, et al. Early versus late necrosectomy in severe necrotizing pancreatitis. Am J Surg 1997; 173:71-75. 14. Fernandez-del Castillo C, Rattner DW, Makary MA, et al. Debridement and closed packing for the treatment of necrotizing pancreatitis. Ann Surg 1998;228:676-684. 15. Uhl W, Warshaw A, Imrie C, et al. IAP Guidelines for the Surgical Management of Acute Pancreatitis. Pancreatology 2002; 2:565-573. 16. Besselink MG, Verwer TJ, Schoenmaeckers EJ, et al. Timing of surgical intervention in necrotizing pancreatitis. Arch Surg 2007; 142:1194-1201. 17. Beger HG, Buchler M, Bittner R, et al. Necrosectomy and postoperative local lavage in patients with necrotizing pancreatitis: results of a prospective clinical trial. World J Surg 1988;12:255-262. 18. Freeny PC, Hauptmann E, Althaus SJ, et al. Percutaneous CT-guided catheter drainage of infected acute necrotizing pancreatitis: techniques and results. Am J Roentgenol 1998;170:969-975. 19. Carter CR, McKay CJ, Imrie CW. Percutaneous necrosectomy and sinus tract endoscopy in the management of infected pancreatic necrosis: an initial experience. Ann Surg 2000;232:175-180. 20. van Santvoort HC, Besselink MG, Horvath KD, et al. Videoscopic assisted retroperitoneal debridement in infected necrotizing pancreatitis. HPB 2007;9:156-159. 21.

Baron TH, Thaggard, morgan DE, et al. Endoscopic therapy for organized pancreatic necrosis. Gastroenterology 1999;111:820-823.

22. Raraty MG, Halloran CM, Dodd S, et al. Minimal access retroperitoneal pancreatic necrosectomy: improvement in morbidity

223


PART III CHAPTER 11

and mortality with a less invasive approach. Ann Surg 2010;251:787-793. 23. Horvath K, Freeny P, Escallon J, et al. Safety and efficacy of video-assisted retroperitoneal debridement for infected pancreatic collections: a multicenter, prospective, single-arm phase 2 study. Arch Surg 2010; 145:817-25. 24. van Santvoort HC, Besselink MG, Bakker OJ, et al. A step-up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med 2010;362:1491-1502. 25. http://www.pancreasclub.com/resources/ AtlantaClassification.pdf. 26. Rau B, Bothe A, Beger HG. Surgical treatment of necrotizing pancreatitis by necrosectomy and closed lavage: changing patient characteristics and outcome in a 19year, single-center series. Surgery 2005; 138:28-39. 27. Farkas G, Marton J, Mandi Y, et al. Surgical management and complex treatment of infected pancreatic necrosis: 18-year expe足 rience at a single center. J Gastrointest Surg 2006;10:278-285. 28. Gotzinger P, Sautner T, Kriwanek S, et al. Surgical treatment for severe acute pancreatitis: extent and surgical control of necrosis determine outcome. World J Surg 2002;26:474-478. 29. Rodriguez JR, Razo AO, Targarona J, et al. Debridement and closed packing for sterile or infected necrotizing pancreatitis: insights into indications and outcomes in 167 patients. Ann Surg 2008;247:294-299. 30 Howard TJ, Patel JB, Zyromski N, et al. Declining morbidity and mortality rates in the surgical management of pancreatic necrosis. J Gastrointest Surg 2007;11:43-49. 31. Ashley SW, Perez A, Pierce EA, et al. Necrotizing pancreatitis: contemporary analysis of 99 consecutive cases. Ann Surg 2001;234:572-579. 32. Besselink MG, Van Santvoort HC, Buskens E, et al. Probiotic prophylaxis in predicted

224

severe acute pancreatitis: a randomised, double-blind, placebo-controlled trial. Lancet 2008;371:651-659. 33. Balthazar EJ, Robinson DL, Megibow AJ, et al. Acute pancreatitis: value of CT in establishing prognosis. Radiology 1990; 174:331-336. 34. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007;370:1453-1457. 35. Gecelter G, Fahoum B, Gardezi S, et al. Abdominal compartment syndrome in severe acute pancreatitis: an indication for a decompressing laparotomy? Dig Surg 2002;19:402-404. 36. De Waele JJ, Hoste E, Blot SI, et al. Intraabdominal hypertension in patients with severe acute pancreatitis. Crit Care 2005; 9:452-457. 37. Chen H, Li F, Sun JB, et al. Abdominal compartment syndrome in patients with severe acute pancreatitis in early stage. World J Gastroenterol 2008;14:3541-3548. 38. Mentula P, Hienonen P, Kemppainen E, Puolakkainen P, Leppaniemi A. Surgical decompression for abdominal compartment syndrome in severe acute pancreatitis. Arch Surg. 2010;145:764-769. 39. Cheatham ML, Malbrain ML, Kirkpatrick A, et al. Results from the International Conference of Experts on Intra-abdominal Hypertension and Abdominal Compartment Syndrome. II. Recommendations. Intensive Care Med 2007;33:951-962. 40. Petrov MS, Shanbhag S, Chakraborty M, et al. Organ Failure and infection of pancreatic necrosis as determinants of mortality in patients with acute pancreatitis. Gastroenterology 2010;139:813-820. 41. Buter A, Imrie CW, Carter CR, et al. Dynamic nature of early organ dysfunction determines outcome in acute pancreatitis. Br J Surg 2002;89:298-302.


CONSERVATIVE APPROACH TO NECROTIZING PANCREATITIS

42. Beger HG, Bittner R, Block S, et al. Bacterial contamination of pancreatic necrosis. A prospective clinical study. Gastroenterology 1986;91:433-438. 43. van Santvoort HC, Besselink MG, Bollen TL, et al. Case-matched comparison of the retroperitoneal approach with laparotomy for necrotizing pancreatitis. World J Surg 2007;31:1635-1642. 44. Gardner TB, Coelho-Prabhu N, Gordon SR et al. Direct endoscopic necrosectomy for the treatment of walled-off pancreatic necrosis: results from a multicenter U.S. series. Gastrointest Endosc 2011; epub ahead of print.

53. Larvin M, Chalmers AG, McMahon MJ. Dynamic contrast enhanced computed tomography: a precise technique for identifying and localising pancreatic necrosis. BMJ 1990;300:1425-1428. 54. Bradley EL, III. A fifteen year experience with open drainage for infected pancreatic necrosis. Surg Gynecol Obstet 1993;177:215222. 55. Gambiez LP, Denimal FA, Porte HL, et al. Retroperitoneal approach and endoscopic management of peripancreatic necrosis collections. Arch Surg 1998;133:66-72.

45. Seifert H, Biermer M, Schmitt W, et al. Transluminal endoscopic necrosectomy after acute pancreatitis: a multicentre study with long-term follow-up (the GEPARD study). Gut 2009;58:1260-1266 46. Dubner H, Steinberg W, Hill M, et al. Infected pancreatic necrosis and peripancreatic fluid collections: serendipitous response to antibiotics and medical therapy in three patients. Pancreas 1996;12:298-302. 47. Baril NB, Ralls PW, Wren SM, et al. Does an infected peripancreatic fluid collection or abscess mandate operation? Ann Surg 2000;231:361-367. 47. Adler DG, Chari ST, Dahl TJ, et al. Conservative management of infected necrosis complicating severe acute pancreatitis. Am J Gastroenterol 2003;98:98-103. 49. Runzi M, Niebel W, Goebell H, et al. Severe acute pancreatitis: nonsurgical treatment of infected necroses. Pancreas 2005;30:195-199. 50 Sakorafas GH, Tsiotos GG, Sarr MG. Extrapancreatic necrotizing pancreatitis with viable pancreas: a previously underappreciated entity. J Am Coll Surg 1999; 188:643-648. 51. Merkle EM, Gorich J. Imaging of acute pancreatitis. Eur Radiol 2002;12:1979-1992. 52. Balthazar EJ. Acute pancreatitis: assessment of severity with clinical and CT evaluation. Radiology 2002;223:603-613.

225


PART III TREATMENT OF COMPLICATED PANCREATITIS


CHAPTER 12 Treatment of necrotizing pancreatitis Clinical Gastroenterology & Hepatology 2012

Sandra van Brunschot, Olaf J Bakker, Marc G Besselink, Thomas L Bollen, Paul Fockens, Hein G Gooszen, Hjalmar C van Santvoort, for the Dutch Pancreatitis Study Group


PART III CHAPTER 12

A B S T R AC T Acute pancreatitis is a common and potentially lethal disease. It is associated with significant morbidity and consumes enormous healthcare resources. Over the last two decades, the treatment of acute pancreatitis has undergone fundamental changes based on new conceptual insights and evidence from clinical studies. The majority of patients with necrotizing pancreatitis have sterile necrosis, which can be successfully treated conservatively. Emphasis of conservative treatment is on supportive measures and prevention of infection of necrosis and other complications. Patients with infected necrosis generally need to undergo an intervention, which has shifted from primary open necrosectomy in an early disease stage to a step-up approach, starting with catheter drainage, if needed followed by minimally invasive surgical or endoscopic necrosectomy once peripancreatic collections have sufficiently demarcated. This review provides an overview of current standards for conservative and invasive treatment of necrotizing pancreatitis.

I N T RODU C T ION Epi d emi ol og y and di ag nosis Acute pancreatitis is an acute inflammation of the pancreas that in Western countries, is mainly caused by gallstones (40 to 50%) and alcohol abuse (10 to 40%). Other causes (20 to 30%) include medication, endoscopic retrograde cholangiopancreatography (ERCP), hypertriglyceridemia, hypercalcemia and surgery. In around 10% the etiology remains unknown.1,2 The pathophysiology of acute pancreatitis is generally considered as a premature or inappropriate activation of digestive enzymes within pancreatic acinar cells, causing autodigestion of the pancreas and surrounding tissues with subsequent local and systemic inflammation.3,4 The incidence of acute pancreatitis is increasing. In the United States, acute pancreatitis accounts for over 200.000 hospital admissions each year.4-6 In Europe, the incidence ranges from approximately 4 to 45 per 100.000 patients a year.2 Acute pancreatitis is associated with significant morbidity and enormous healthcare resources.5,7 Overall mortality in acute pancreatitis is approximately 5%.3 Diagnosis of acute pancreatitis requires at least two of the following three features: 1) abdominal pain, typically epigastric, 2) serum amylase or lipase â&#x2030;Ľ 3 times the upper limit of normal, and 3) characteristic findings of acute pancreatitis on contrast-enhanced computed tomography (CECT). In most cases, the clinical history and laboratory results accurately provide the diagnosis and no diagnostic imaging is required. A CECT in the 228


TREATMENT OF NECROTIZING PANCREATITIS

initial 3 to 4 days of acute pancreatitis might underestimate or miss the amount of necrosis.8-10 In general, a CECT is advised if a patient does not improve after the first week of treatment to evaluate the extent of local complications.9 In clinical practice, however, it is not uncommon for patients to undergo CT earlier than one week, especially in case of early complications. C lini c a l cours e Acute pancreatitis has a mild clinical course in around 80% of patients, in whom the disease resolves spontaneously within around a week.11 However, about 20% of patients develop severe acute pancreatitis, which is associated with mortality rates of 8 up to 39%.3 The 1992 Atlanta classification defined severe acute pancreatitis as the presence of organ failure or local complications such as pancreatic necrosis. Pancreatic necrosis occurs in around 15 to 20% of patients and is typically diagnosed as focal areas of non-enhancing pancreatic parenchyma on CECT (Figure 1).3 The Atlanta classification is currently under revision.12 In the revised classification the definition of â&#x20AC;&#x2DC;necrotizing pancreatitisâ&#x20AC;&#x2122; will not only include patients with pancreatic necrosis but also patients with extrapancreatic fat necrosis alone (i.e., with normal enhancing pancreatic parenchyma on CECT). Some studies have suggested that patients with extrapancreatic necrosis alone may have a better outcome than patients with pancreatic necrosis. However, extrapancreatic necrosis alone is clinically a more severe entity than acute edematous pancreatitis.13,14

Figure 1. Acute necrotizing pancreatitis: a 47-year-old man with necrotizing pancreatitis of biliary origin. Perfusion defect is observed at the neck of the pancreas (arrows), with remaining viable pancreatic tissue at the body and tail (asterisk). Note the presence of gallstones.

229


PART III CHAPTER 12

Several prognostic scoring systems are used to predict the severity of acute pancreatitis in the first days of admission: among them are APACHE-ll, (Modified) Glasgow (Imrie) score, Ranson score, procalcitonin, C-reactive protein and blood urea nitrogen (BUN).15 These prognostic scoring systems are mainly used for severity stratification in clinical studies, and one can argue about their importance in daily clinical practice. Theoretically, severe acute pancreatitis is divided in a biphasic clinical course. The first phase (i.e., up to 1 to 2 weeks after onset of symptoms) is characterized by a proinflammatory immune response. A systemic inflammatory response syndrome (SIRS) often occurs, which is frequently accompanied by failure of one or more organ systems.16-18 Organ failure develops in around 40% of patients with severe acute pancreatitis and is associated with a mortality rate of approximately 30%.14,19 More than half of the cases of organ failure occur in the first week of admission.14 Patients with persistent organ failure or multi-organ failure have a worse prognosis than patients with transient organ failure or single organ failure.3,16,20 It has been suggested that approximately half of the deaths from necrotizing pancreatitis are caused by multi-organ failure in the early phase.14,20,21 In the second phase of the disease (i.e., after 1 to 2 weeks from onset of symptoms) the pro-inflammatory immune response usually subsides. In this phase, the patient’s immune system is probably suppressed which renders patients more susceptible to infectious complications caused by bacterial translocation.22-24 The most severe infectious complication in necrotizing pancreatitis is infection of pancreatic or peripancreatic necrosis. The incidence of infected necrosis in patients with necrotizing pancreatitis has remained stable over the last decades (around 30%).14,25 The peak incidence of infected necrosis is between 2 and 4 weeks after onset of disease.26 Infected necrosis is typically suspected when there is persistent sepsis, new onset sepsis or progressive clinical deterioration (i.e., signs of sepsis) despite maximal support in the second phase of the disease, without another source of infection. A pathognomonic sign of infected necrosis are impacted peripancreatic or intrapancreatic gas bubbles in a collection on CECT (Figure 2), although this is present in only a minority of patients. In some patients, gas bubbles can also be explained by a fistulous communication between the collection and bowel. Which however, also means the collection is contaminated. A fine needle aspiration (FNA) for microbiological culture can be performed to diagnose infected necrosis. However, FNA might not always be necessary in patients with necrotizing pancreatitis and suspected infected necrosis. In addition, FNA is associated with a risk of false-negative results.27 Because “suspected infected necrosis” no longer represents an immediate indication for invasive treatment, a FNA culture result will not per se guide clinical decision making. Intervention is generally postponed to 3 to 4 weeks after onset of disease, and the need for intervention is primarily dictated by clinical deterioration and encapsulation of the infected collection rather than a positive microbiological culture obtained by FNA. A recent Dutch multicenter randomized 230


TREATMENT OF NECROTIZING PANCREATITIS

Figure 2. Infected necrosis: a 55-year-old woman with infected necrotizing pancreatitis. There is a large heterogeneous collection in the pancreatic and peripancreatic area (arrows point at the borders of the collection) with impacted gas bubbles (big arrowheads) and a gas-fluid level (small arrowheads), often a pathognomonic sign of infected necrosis.

controlled trial (RCT) demonstrated that a strategy of intervention in patients with clinical suspicion of infected necrosis, without the routine use of FNA, yielded definitive proof of infected necrosis (i.e., positive microbiological cultures from radiological drainage and operation) in over 90% of patients.28 Even though much has changed in the management of necrotizing pancreatitis over the last 20 years, mortality of infected necrosis remains as high as 12 to 39%.14,28-32

T R E AT M E N T I N T H E E A R LY P HA SE Initial treatment of acute pancreatitis is mainly conservative and focuses primarily on frequent monitoring of the clinical course, pain management, fluid-resuscitation and supportive measures for organ failure. Supp or tive me asures Patients admitted with acute pancreatitis should be closely monitored with adequate amounts of intravenous fluids and pain management. In case of hemodynamic, respiratory or renal insufficiency with a diuresis of <0.5ml/kg/hr despite adequate fluid resuscitation, or metabolic disorders, patients need to be managed in an intensive care unit. Aggressive fluid resuscitation is undertaken, especially in the setting of hemoconcentration which reflects intravascular volume depletion. Prevention or reversal of hemoconcentration 231


PART III CHAPTER 12

is the goal of volume resuscitation. Fluid balance should be maintained and closely monitored.4 The need for large amounts of fluid during the initial 24 hours is associated with poor outcome and therefore this group of patients should be watched carefully.33-36 Two retrospective cohort studies suggested that aggressive early fluid resuscitation (at least one third of the total 72-hour cumulative intravenous fluid volume given during the first 24 hours) is associated with decreased risk of SIRS and organ failure, a lower rate of admission to the intensive care unit, a reduced length of hospital stay, and reduced mortality.37,38 Most guidelines encourage targeting fluid resuscitation toward correcting hypotension, correcting hemoconcentration, and maintaining adequate urine output.3,9,37 However, a recent RCT from China demonstrated that aggressive, uncontrolled administration of intravenous fluids in the first days of acute pancreatitis can also be detrimental as it was related to a twofold increase in mortality.39 The type of fluid administered has been investigated in 2 studies. A cohort study of 434 patients showed no difference in outcome based on the type of fluid administered.38 Though, a recent RCT suggests that lactated Ringerâ&#x20AC;&#x2122;s solution reduces the systemic inflammation compared to fluid resuscitation with normal saline.40 Analgesia plays an important role in the treatment of acute pancreatitis. Parenteral analgesics are generally needed. There is no evidence to suggest an advantage of any particular type of medication. When abdominal pain is particularly severe, patientcontrolled analgesia is usually preferred. It is important to obtain measurements of bedside oxygen saturation frequently whenever narcotic agents are administered to relieve pain.3,9 When organ dysfunction or organ failure is present, supportive treatment should be provided in an appropriate critical care facility.41 Several medical treatment options (e.g., platelet activating factor antagonist (lexipafant), activated protein C) to prevent organ failure in the early phase have been investigated but none of them have been convincingly shown to be effective.42,43 Pre venti on of infe c ti on of ne crosis A recent prospective observational study of 731 patients with acute pancreatitis (28% with severe acute pancreatitis) showed that 25% of all patients developed one or more infections (i.e., pneumonia, bacteremia or infected necrosis). Mortality in patients with infection was 30%, whereas 80% of all deceased patients had an infection.26 In severe acute pancreatitis, disturbed gastrointestinal motility may lead to bacterial overgrowth and failure of the structural mucosal barrier which may lead to increased gut permeability.44-47 These events may result in bacteria which cross the gastrointestinal mucosal barrier and invade the systemic compartment, so called bacterial translocation.44,48 Bacterial translocation is thought to be the mechanism causing most infections in acute pancreatitis. Strategies aimed at preventing bacterial translocation and subsequent infections have therefore been widely studied in recent years: antibiotics, probiotics and enteral nutrition. 232


TREATMENT OF NECROTIZING PANCREATITIS

Antibiotics Several meta-analyses, including 15 randomized trials, have been published on systemic antibiotics aimed at preventing infectious complications in acute pancreatitis.49-51 Only 3 RCTs were double-blind placebo controlled.52-54 The design, methodological quality, and most importantly, outcome of the included studies vary widely.49 Most meta-analyses did not demonstrate a significant beneficial effect of antibiotic prophylaxis on infection of pancreatic necrosis and mortality.49-51 Although the discussion on antibiotic prophylaxis in acute pancreatitis continues, at the moment there is no convincing evidence in favor of routine antibiotic prophylaxis. If a beneficial effect actually exists, it will be difficult to perform a randomized study with sufficient statistical power to demonstrate this effect. Most international guidelines currently do not recommend routine antibiotic prophylaxis.3,55 Probiotic s Probiotics are non-pathogenic bacteria that, on delivery to the hostâ&#x20AC;&#x2122;s intestinal tract, are believed to prevent bacterial overgrowth, reinforce the mucosal barrier function and regulate the systemic immune system which may reduce bacterial translocation and subsequent infections. Probiotics have been shown to prevent infections in elective major abdominal surgery.48 Several studies on probiotics have also been performed in patients with acute pancreatitis. The first two double-blind, placebo controlled, randomized trials from the same study group included respectively 45 and 62 patients with predicted severe acute pancreatitis. The first trial showed a significant reduction of infected pancreatic necrosis in patients receiving probiotics. The second trial showed a lower, but not significant incidence of multi-organ failure, septic complications and mortality in the probiotics group.56,57 The third and largest double-blind, placebo controlled, trial Included 298 patients with predicted severe acute pancreatitis. This study did not show an effect of an enterally administered multispecies probiotic mixture on the incidence of infections. However, patients receiving probiotics had an increased mortality as compared to patients receiving placebo (16% versus 6%; p value 0.01).29 This negative effect was associated with an increase in non-occlusive mesenteric ischemia, which was predominantly seen in the patients with multi-organ failure, and has not yet been explained.58 There is currently a strong advice against the use of probiotics in patients with predicted severe acute pancreatitis. Enteral nutr ition Nutritional support has a fundamental role in the management of severe acute pancreatitis. Besides maintaining adequate caloric intake, nutritional support is important in prevention of infectious complications. Nutritional support can be achieved through parenteral and enteral feeding. Both strategies have been compared in several randomized 233


PART III CHAPTER 12

trials and meta-analyses. Results show that enteral feeding significantly reduce mortality, multi-organ failure, systemic infections and the need for operative intervention compared with parenteral feeding.59 Enteral nutrition can be administered through a nasogastric or nasojejunal tube. Two RCTs compared these two routes and did not show significant differences between recurrence or worsening of pain, hospital stay, complications, or mortality.60,61 These studies, therefore, suggested that the simpler, cheaper, and more easily used nasogastric feeding appears to be well tolerated and is as safe as nasojejunal feeding in patients with severe acute pancreatitis. A larger study to further test the safety of nasogastric feeding is currently underway in the US (SNAP trial, Clinicaltrials.gov, NCT00580749). Experimental and clinical research has shown that the phenomena of bacterial translocation already takes place within a few hours after onset of symptoms.22,62 This implies that there is only a very narrow therapeutic-window for preventing bacterial translocation and subsequent infections.26 Theoretically, enteral feeding should therefore be started as early as possible for a beneficial clinical effect. There is evidence in favor of this hypothesis in critically ill patients other than acute pancreatitis. In a meta-analysis of 15 randomized trials comparing early (within 36 hrs) versus delayed (after 36 hrs) start of enteral feeding on outcome of critically ill ICU patients. Early enteral nutrition significantly reduced the incidence of infections and length of hospital stay.63 In acute pancreatitis, there is only indirect evidence for an early start of enteral feeding. A metaanalysis comparing the effect of enteral versus parenteral nutrition with subgroups based on timing of start of nutrition showed that an early start of enteral feeding significantly reduced multi-organ failure, pancreatic infections and mortality.64 The first randomized trial specifically designed to compare early and selective delayed enteral feeding in predicted severe acute pancreatitis (PYTHON trial) is currently underway in the Netherlands (ISRCTN 18170985).65 ERCP for bi li ar y p ancre atitis Gallstones are the most common cause of acute pancreatitis in the Western world.2,9 In patients with biliary pancreatitis, decompression of the common bile duct and removal of gallstones or sludge by early endoscopic retrograde cholangiopancreatography (ERCP) with subsequent sphincterotomy may mitigate the pancreatic inflammation and reduce complications. Several RCTs have investigated the clinical effect of early ERCP in acute biliary pancreatitis.66-69 From the available evidence, two conclusions on the role of ERCP are generally drawn: 1) patients with biliary pancreatitis and concurrent cholangitis should undergo early ERCP and 2) in predicted non-severe biliary pancreatitis ERCP is not beneficial.3,9 However, the role of early ERCP in patients with predicted severe biliary pancreatitis remains controversial. Although the 2005 UK guidelines on acute pancreatitis recommend emergency ERCP in these patients41, two more recent US guidelines state that 234


TREATMENT OF NECROTIZING PANCREATITIS

the value of early ERCP in predicted severe biliary pancreatitis without cholangitis is yet undetermined.3,9 This is due to the fact that the published RCTs included only a small number of patients with predicted severe acute pancreatitis and, hence, were statistically underpowered to detect clinical effects in the group of most severe ill patients.66-69 A recent updated meta-analysis, showed no effect of ERCP on complications or mortality in all patients with predicted severe biliary pancreatitis. However, the pooled sample size was still small (n=126) and sphincterotomy was only performed in 53% of patients.70 A recent prospective observational study, including 153 patients with predicted severe biliary pancreatitis without cholangitis, showed no significant reduction of complications following ERCP in patients without radiological or biochemical signs of cholestasis. In the subgroup of patients with cholestasis, however, ERCP was significantly associated with fewer complications, including pancreatic necrosis.71 A future large and well-designed randomized trial should study the effect of ERCP in patients with predicted severe biliary pancreatitis without cholangitis, with a predefined subgroup analysis in patients with and without signs of cholestasis. E arly compli c ati ons re quiring inter venti on A rare but dramatic complication early in the course of severe acute pancreatitis is abdominal compartment syndrome (ACS).17 ACS is preceded by intra-abdominal hypertension (IAH) which is defined as an intra-abdominal pressure at or above 12 mmHg. ACS is diagnosed when the intra-abdominal pressure exceeds 20 mmHg and there are signs of new organ failure (e.g., respiratory, circulatory, renal).72 IAH generally occurs early and in some studies the incidence has been reported to be as high as 59 to 78% in patients with severe acute pancreatitis.73,74 The pathophysiology of IAH is directly related to the pancreatic inflammation, which may cause retroperitoneal edema, fluid collections, ascites and a paralytic ileus. IAH may also be partly iatrogenic, resulting from aggressive fluid resuscitation. IAH can also manifest in the later phase of acute pancreatitis, associated with local pancreatic complications.75 The incidence of ACS in severe acute pancreatitis has been reported up to 30% in some studies and is associated with extremely high mortality rates of 46 to 75%.73,74,76,77 ACS requires immediate measures such as sedation, analgesics, nasogastric decompression, fluid restriction, and diuretics, to lower the abdominal pressure. If these measures do not result in a rapid clinical improvement, invasive intervention is required. Percutaneous catheter decompression seems to be effective in resolving ACS in patients with intraperitoneal fluid, abscess, or blood, thereby avoiding the need for surgical decompression.78 This strategy may improve outcome and is currently evaluated in acute pancreatitis by a randomized trial (Clinicaltrials.gov NCT00793715). If percutaneous decompression does not immediately lower the intraabdominal pressure, surgical decompression laparotomy should be performed.73,77,78 In rare cases where decompressive laparotomy is necessary in the early phase of 235


PART III CHAPTER 12

necrotizing pancreatitis, it is advised not to open the retroperitoneum or to perform necrosectomy. At this stage, the necrosis is probably sterile which means a formal necrosectomy is not indicated and, conversely, may cause severe complications such as bleeding, perforation, infection of necrosis and death.79 Another uncommon but devastating complication requiring early intervention is bowel ischemia. The occurrence of non-occlusive mesenteric ischemia is well known in critically ill patients80 and several cases of non-occlusive mesenteric ischemia have been reported in acute pancreatitis.81 Although data on the incidence and outcome of bowel ischemia in acute pancreatitis are limited, the incidence seems to be low (approximately 4%). However, if present, mortality rates are approaching 100%.81

T R E AT M E N T I N T H E L AT E P HASE C ons er vative tre atment In around two third of patients with necrotizing pancreatitis, the pancreatic or peripancreatic necrosis remains sterile. These patients can develop walled-off pancreatic necrosis (WOPN) late in the disease. WOPN is characterized by a thickened wall between the necrosis and the adjacent viable tissue (fig 3). In accordance with international guidelines, patients with sterile necrosis can be successfully managed conservatively (i.e., without any form of radiologic, endoscopic or surgical intervention).3,17,55 An intervention for sterile peripancreatic collections with fluid and necrosis accommodates the risk of introducing infection of necrosis (55 to 59%).82,83 Iatrogenic infection of sterile necrosis requires additional interventions and considerably increases morbidity and mortality.82,84,85 Probably, the only exception are patients with persistent mechanical obstruction due to peripancreatic collections, in the absence of clinical signs of infection, causing ongoing nausea, vomiting, pain, anorexia and inability to resume oral intake. In this case, the decision for intervention will be solely based on clinical symptoms, supported by CT findings and should be delayed up to at least 4 to 6 weeks after onset of symptoms. This due to the fact that most collections will resolve spontaneously. In a recent prospective observational study of 639 patients with necrotizing pancreatitis, 62% of patients were treated conservatively. Mortality in these patients was 7%.14 Invasive tre atment Whereas historically many patients with sterile necrosis also underwent necrosectomy, it is now accepted that the main indication for intervention is infected necrosis.9,17,41,86 The timing of intervention has also changed. Necrosectomy was once performed at a very early stage79, whereas it is now believed that intervention should be delayed to 236


TREATMENT OF NECROTIZING PANCREATITIS

approximately 3 to 4 weeks after onset of disease.27,87,88 In order to postpone intervention, patients with signs of infected necrosis are initially treated with broad-spectrum antibiotics and maximal support. This allows for encapsulation and demarcation of peripancreatic collections, which may improve conditions for intervention and thereby theoretically decrease the risk of complications such as bleeding and perforation. However, in some patients this is not feasible and dramatic clinical deterioration will require earlier intervention. A recent study of 242 patients undergoing intervention for necrotizing pancreatitis showed, in a multivariable analysis adjusting for confounding factors, that patients with longer time between admission and intervention had lower mortality: 0 to 14 days, 56%; 14 to 29 days, 26%; and >29 days, 15% (P<.001).14 It should be noted that there are several reports of patients with infected necrosis who were in such a good clinical condition that allowed treatment with intravenous antibiotics without invasive intervention.14 However, the vast majority of patients with infected necrosis need to undergo radiological, endoscopic or surgical intervention at some point.

Figure 3. Walled-off necrosis: a 40-year-old man with necrotizing pancreatitis and walled-off necrosis. A completely encapsulated collection is observed in the pancreatic and peripancreatic area (arrows), with predominant fluid density interspersed with areas of fat density (arrowheads).

237


PART III CHAPTER 12

Primar y op en ne cros e c tomy The traditional approach to infected necrosis used to be primary open necrosectomy to completely remove the infected necrosis.25,89 This is an invasive approach associated with a high risk of complications (34 to 95%) and mortality (11 to 39%) and long-term pancreatic insufficiency.31,32,86,90-95 As an alternative to primary open necrosectomy, minimally invasive radiologic, surgical and endoscopic techniques for intervention have gained wide popularity. Minima l ly invasive appro aches Minimally invasive interventions include percutaneous catheter drainage (PCD)96, endoscopic transluminal drainage (ETD)97-102, endoscopic (transluminal) necrosectomy (ETN)98,99,103-112 and minimally invasive retroperitoneal surgical necrosectomy.28,31,92,113-117 Minimally invasive techniques are thought to induce less physiological stress as compared with open surgical necrosectomy. Reduced surgical stress might decrease the risk of complications in these often already severely ill patients. Percutaneou s catheter drainage Image-guided PCD (Figure 4) as primary treatment of infected necrosis was first described in 1998.96 The rationale of PCD is to treat infected necrosis as an abscess and drain the infected fluid (i.e., pus) under pressure, without removal of necrotic material. Successful drainage of the infected fluid will temporize sepsis and improve patientâ&#x20AC;&#x2122;s clinical condition. This may lead to a situation where the patient is capable of resorbing the necrotic material without the need for formal necrosectomy. PCD is feasible in >95%

Figure 4. PCD: a 55-year-old woman with infected necrotizing pancreatitis (same patient as in Figure 2). Axial CT (A) performed in right decubitus position for optimal retroperitoneal positioning of a 12F percutaneous drain (arrow) via the left flank. Successive follow-up CT (B) reveals reduction in size of infected pancreatic collection, with PCD centrally positioned via the left retroperitoneal route (between the descending colon [DC] and the right kidney [R]).

238


TREATMENT OF NECROTIZING PANCREATITIS

of patients with infected necrotizing pancreatitis, often via a left-sided retroperitoneal approach.28,118 In a recent systematic review of 11 studies with a total of 384 patients receiving PCD for necrotizing pancreatitis, more than half of the patients were successfully treated with PCD alone and thus did not undergo additional necrosectomy.119 This was confirmed by a recent prospective observational study. In 131 patients undergoing intervention for (suspected) infected necrosis, PCD was performed as the first intervention in 63% of patients, without the need for additional necrosectomy in 35% of patients. 14 If necrosectomy is still needed after PCD, PCD may have allowed for further encapsulation of the necrotic collections and improvement of the patientâ&#x20AC;&#x2122;s clinical condition. PCD thereby acts as a bridge to surgery. The preferred route for PCD is through the left retroperitoneum, so that the drain can be used as guidance for retroperitoneal surgical necrosectomy. Minimally invasive retroper itoneal necros ec tomy Several less invasive surgical techniques to perform necrosectomy have been described in recent years. The most commonly used techniques are sinus tract endoscopy31,114,120, laparoscopic trans-abdominal necrosectomy121,122 and video-assisted retroperitoneal debridement (VARD).115-117 Sinus tract endoscopy involves serial dilatation of a percutaneous catheter drain tract using fluoroscopic guidance in the operating room, with subsequent necrosectomy by jet irrigation and suction using a nephroscope or flexible endoscope. Residual solid necrotic tissue is evacuated using a variety of endoscopic instruments. Several retrospective studies reported a mean morbidity of 25 to 88% and mortality of 0 to 25%. A median of 3 to 4 (range 1 to 9) session per patient were necessary to remove all infected necrosis.92,114,120 VARD (Figure 5) can be considered a hybrid between sinus tract endoscopy and an open lumbar approach.123-125 Using a 5 cm subcostal incision, the previously placed percutaneous catheter drain is followed into the retroperitoneum to enter the necrotic collection. The first necrosis is removed under direct vision, followed by further debridement under videoscopic assistance.116,117 VARD is associated with a morbidity and mortality of respectively 24 to 54% and 0 to 8%.115,116,126 VARD has several advantages: it uses regular surgical equipment, it is a straightforward, semi-open procedure and mostly requires only 1 session per patient. Endos copic transluminal drainage and necros ec tomy As an alternative to radiologic and surgical techniques, endoscopic transluminal drainage (ETD) and endoscopic transluminal necrosectomy (ETN) are gaining popularity. 239


PART III CHAPTER 12

Endoscopic interventions are typically performed under conscious sedation without the need for general anaesthesia. This potentially reduces the inflammatory response and may further reduce complications such as new-onset multi-organ failure. First, the collection with infected necrosis is visualized by endoscopic ultrasound to determine the extent of necrosis and the optimal site of drainage. Next, the collection is punctured through the gastric or duodenal wall, followed by balloon dilatation of the tract. Two double pigtail stents and a nasocystic catheter are placed for continuously postoperative irrigation (Figure 6A and 7). Several retrospective cohort studies show promising results of ETD, with complication rates of 2 to 21% and mortality rates of 0 to 6%.97-102 In case of no improvement or deterioration after ETD, ETN can be performed to remove infected necrosis. The cystostomy tract is further dilated, the collection is entered by a forward viewing endoscope and necrosectomy is performed (figure 6B).

Figure 5. PCD and VARD. (A) Cross-sectional image and torso depicting a peripancreatic collection with fluid and necrosis. The preferred access route is through the left retroperitoneal space between the left kidney, dorsal spleen, and descending colon. A percutaneous drain is inserted in the collection to mitigate sepsis and postpone or even obviate necrosectomy. The area of detail is shown in (panel B). (C) A 5-cm subcostal incision is made, and the previously placed percutaneous drain is followed into the retroperitoneum to enter the necrotic collection. The first necrosis is removed under direct vision with a long grasping forceps. This is followed by further debridement under videoscopic assistance (D).

240


TREATMENT OF NECROTIZING PANCREATITIS

Figure 6. ETD and ETN. A large peripancreatic collection containing fluid and necrosis is shown. The preferred access route for endoscopic translumi- nal treatment is through the posterior wall of the stomach. The necrotic collection often bulges into the stomach, facilitating endoscopic transluminal treat- ment. (A) The collection is punc- tured through the gastric wall, followed by balloon dilatation of the tract. Two double-pigtail stents and a nasocystic catheter are placed for continuous postoperative irrigation. (B) The cystostomy tract is further dilated, the collection is entered by a forward viewing endoscope, and necrosectomy is performed.

At the end of the procedure, two double-pigtail stents and a nasocystic catheter are placed. If necessary, ETN can be repeated until the majority of necrotic material is removed.98,99,103-112 By avoiding any abdominal wall incision, typical complications related to surgical necrosectomy such as incisional hernias, pancreatic fistula and wound infection will probably be reduced with ETN. In a recent systematic review of 10 series on ETN in necrotizing pancreatitis overall mortality after ETN was 5% and the mean procedure related morbidity 27%. In 76% of 241


PART III CHAPTER 12

patients, complete resolution of the necrotic collection was achieved by endoscopic interventions alone. On average there were 4 (range 1 to 35) endoscopic sessions needed to achieve complete resolution.107 Although these results are promising, there is a risk of selection bias within these studies because the number of critically ill patients with infected necrosis included was relatively low. A recent pilot randomized controlled trial showed promising results. ETN significantly reduced the pro-inflammatory response (measured by serum interleukine-6 levels) as well as the composite clinical endpoint of major morbidity and mortality compared with surgical necrosectomy.127 The step-up approach The minimally invasive techniques can be applied in a so-called step-up approach.28,30,128 The first step is catheter drainage (i.e., radiologic percutaneous or endoscopic transluminal) of the collection with infected fluid and necrosis to mitigate sepsis and postpone or even obviate necrosectomy.99,119 If drainage does not lead to clinical improvement, the next step is minimally invasive necrosectomy performed either surgically or endoscopically.92,114,116,117 As compared with open necrosectomy, the step-up approach aims at control of the source of infection, rather than complete removal of the infected necrotic tissue. The step-up approach can be performed both surgically and endoscopically . The PANTER trial compared primary open necrosectomy with a surgical step-up approach in 88 patients with suspected or confirmed infected necrosis.28 The step-up approach, using PCD and followed, if necessary, by VARD, reduced the combined primary endpoint of death and major complications (i.e. new multi-organ failure, enterocutaneous fistula, perforation or bleeding) from 69% to 40%. Furthermore, at 6 months follow-up, patients assigned to the step-up approach had a significantly lower

Figure 7. ETD: a 63-year-old woman with infected necrotizing pancreatitis. Axial CT (A) and coronal reconstructed mean intensity projection (B) show an infected pancreatic collection (arrow) with an endoscopic pigtail drain (arrowheads) positioned inside the collection (L, liver; S, stomach).

242


TREATMENT OF NECROTIZING PANCREATITIS

rate of incisional hernias and new-onset diabetes. The step-up approach also reduced total costs with 12%. Finally, 35% of patients in the step-up approach group were treated with percutanous drainage alone, and did not need any form of surgery.28 These outcomes may further be improved by an endoscopic step-up approach which consists of ETD, followed, if necessary, by ETN. The Dutch Pancreatitis Study Group has recently started a nationwide randomized trial comparing the surgical step-up approach with the endoscopic step-up approach: TENSION (Trial registration: ISRCTN09186711).

SUM M A RY Necrotizing pancreatitis remains a complex and challenging disease, even though several major improvements have occurred in the management of the disease over the last two decades. In summary, the initial treatment of necrotizing pancreatitis should primarily focus on fluid resuscitation, pain management and supportive measures for organ failure. With regard to prevention of infection of necrosis, routine antibiotic or probiotic prophylaxis is not recommended. Enteral nutrition, compared with parenteral nutrition, appears to be effective in preventing infected necrosis, but the optimal timing of start of enteral feeding requires further study. In patients with biliary pancreatitis and absence of cholangitis, there is no evidence that early ERCP with spincterotomy is beneficial. However, in the subset of patients with predicted severe biliary pancreatitis and radiologic or biochemical signs of cholestasis, early ERCP and sphincterotomy may prevent further complications. Conservative treatment is successful in around two-third of patients. Unnecessary intervention for sterile necrosis accommodates the risk of introducing infection and subsequent complications. However, 30% of patients spontaneously develop infection of necrosis and need to undergo invasive intervention. Whenever clinically feasible, intervention is postponed until there is sufficient encapsulation and demarcation of the infected peripancreatic collections, generally 3 to 4 weeks after onset of symptoms. Primary open necrosectomy has been replaced by a minimally invasive step-up approach which lowers the risk of major morbidity. The initial step is drainage of infected peripancreatic collections which can be performed image-guided percutaneously or EUS guided endoscopic transluminally, depending on anatomic feasibility and local expertise. Catheter drainage is successful as definitive treatment in about 40% of patients. If catheter drainage does not lead to clinical improvement, the next step is minimally invasive drainguided retroperitoneal necrosectomy or endoscopic transluminal necrosectomy. A treatment algorithm for severe acute pancreatitis is given in Figure 8. Future studies should further elucidate the role of both minimally invasive surgical and endoscopic interventions in patients with necrotizing pancreatitis. 243


PART III CHAPTER 12

Figure 8. Treatment algorithm for severe acute pancreatitis.

244


TREATMENT OF NECROTIZING PANCREATITIS

REFERENCES 1. Neoptolemos JP, Kemppainen EA, Mayer JM, et al. Early prediction of severity in acute pancreatitis by urinary trypsinogen activation peptide: a multicentre study. Lancet 2000;355:1955-60. 2. Yadav D, Lowenfels AB. Trends in the epidemiology of the first attack of acute pancreatitis: a systematic review. Pancreas 2006;33:323-30. 3. Banks PA, Freeman ML. Practice guidelines in acute pancreatitis. Am J Gastroenterol 2006;101:2379-400. 4. Whitcomb DC. Clinical practice. Acute pancreatitis. N Engl J Med 2006;354:2142-50. 5. Fagenholz PJ, Fernandez-del Castillo C, Harris NS, et al. Direct medical costs of acute pancreatitis hospitalizations in the United States. Pancreas 2007;35:302-7. 6. Swaroop VS, Chari ST, Clain JE. Severe acute pancreatitis. JAMA 2004;291:2865-8. 7. Shaheen NJ, Hansen RA, Morgan DR, et al. The burden of gastrointestinal and liver diseases, 2006. Am J Gastroenterol 2006;101:2128-38. 8. Bollen TL, Singh VK, Maurer R, et al. A comparative evaluation of radiologic and clinical scoring systems in the early prediction of severity in acute pancreatitis. Am J Gastroenterol 2012;107:612-9. 9. Forsmark CE, Baillie J. AGA Institute technical review on acute pancreatitis. Gastroenterology 2007;132:2022-44. 10. Spanier BW, Nio Y, van der Hulst RW, et al. Practice and yield of early CT scan in acute pancreatitis: a Dutch Observational Multicenter Study. Pancreatology 2010; 10:222-8. 11. Beger HG, Rau B, Isenmann R. Natural history of necrotizing pancreatitis. Pancreatology 2003;3:93-101.

12. Revision of the ATLANTA classification of acute pancreatitis. 2008. (Accessed at http:// p anc rea s clu b. c om / w p - c onte nt / up loads/2011/11/AtlantaClassification.pdf.) 13. Sakorafas GH, Tsiotos GG, Sarr MG. Extrapancreatic necrotizing pancreatitis with viable pancreas: a previously underappreciated entity. J Am Coll Surg 1999;188:643-8. 14. van Santvoort HC, Bakker OJ, Bollen TL, et al. A conservative and minimally invasive approach to necrotizing pancreatitis improves outcome. Gastroenterology 2011; 141:1254-63. 15. Wu BU, Bakker OJ, Papachristou GI, et al. Blood urea nitrogen in the early assessment of acute pancreatitis: an international validation study. Arch Intern Med 2011; 171:669-76. 16. Mofidi R, Duff MD, Wigmore SJ, et al. Association between early systemic inflammatory response, severity of multiorgan dysfunction and death in acute pancreatitis. Br J Surg 2006;93:738-44. 17. Nathens AB, Curtis JR, Beale RJ, et al. Management of the critically ill patient with severe acute pancreatitis. Crit Care Med 2004;32:2524-36. 18. Werner J, Feuerbach S, Uhl W, et al. Management of acute pancreatitis: from surgery to interventional intensive care. Gut 2005;54:426-36. 19. Petrov MS, Shanbhag S, Chakraborty M, et al. Organ failure and infection of pancreatic necrosis as determinants of mortality in patients with acute pancreatitis. Gastroenterology 2010;139:813-20. 20. Johnson CD, Abu-Hilal M. Persistent organ failure during the first week as a marker of fatal outcome in acute pancreatitis. Gut 2004;53:1340-4. 21. Blum T, Maisonneuve P, Lowenfels AB, et al. Fatal outcome in acute pancreatitis: its occurrence and early prediction. Pancreatology 2001;1:237-41.

245


PART III CHAPTER 12

22. Ammori BJ, Leeder PC, King RF, et al. Early increase in intestinal permeability in patients with severe acute pancreatitis: correlation with endotoxemia, organ failure, and mortality. J Gastrointest Surg 1999; 3:252-62. 23. Deitch EA. The role of intestinal barrier failure and bacterial translocation in the development of systemic infection and multiple organ failure. Arch Surg 1990; 125:403-4. 24. Van Leeuwen PA, Boermeester MA, Houdijk AP, et al. Clinical significance of translocation. Gut 1994;35:S28-34. 25. Beger HG, Buchler M, Bittner R, et al. Necrosectomy and postoperative local lavage in patients with necrotizing pancreatitis: results of a prospective clinical trial. World J Surg 1988;12:255-62. 26. Besselink MG, van Santvoort HC, Boermeester MA, et al. Timing and impact of infections in acute pancreatitis. Br J Surg 2009;96:267-73.

33. de-Madaria E, Soler-Sala G, Sanchez-Paya J, et al. Influence of fluid therapy on the prognosis of acute pancreatitis: a prospective cohort study. Am J Gastroenterol 2011; 106:1843-50. 34. Eckerwall G, Olin H, Andersson B, et al. Fluid resuscitation and nutritional support during severe acute pancreatitis in the past: what have we learned and how can we do better? Clin Nutr 2006;25:497-504. 35. Kuwabara K, Matsuda S, Fushimi K, et al. Early crystalloid fluid volume management in acute pancreatitis: association with mortality and organ failure. Pancreatology 2011;11:351-61. 36. Wu BU. Editorial: fluid resuscitation in acute pancreatitis: striking the right balance. Am J Gastroenterol 2011;106:1851-2.

27. Besselink MG, Verwer TJ, Schoenmaeckers EJ, et al. Timing of surgical intervention in necrotizing pancreatitis. Arch Surg 2007;142:1194-201.

37. Gardner TB, Vege SS, Chari ST, et al. Faster rate of initial fluid resuscitation in severe acute pancreatitis diminishes in-hospital mortality. Pancreatology 2009;9:770-6.

28. van Santvoort HC, Besselink MG, Bakker OJ, et al. A step-up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med 2010;362:1491-502.

38. Warndorf MG, Kurtzman JT, Bartel MJ, et al. Early fluid resuscitation reduces morbidity among patients with acute pancreatitis. Clin Gastroenterol Hepatol 2011;9:705-9.

29. Besselink MG, van Santvoort HC, Buskens E, et al. Probiotic prophylaxis in predicted severe acute pancreatitis: a randomised, double-blind, placebo-controlled trial. Lancet 2008;371:651-9. 30. Besselink MG, van Santvoort HC, Nieuwenhuijs VB, et al. Minimally invasive 'step-up approach' versus maximal necrosectomy in patients with acute necrotising pancreatitis (PANTER trial): design and rationale of a randomised controlled multicenter trial [ISRCTN13975868]. BMC Surg 2006;6:6.

246

32. Rodriguez JR, Razo AO, Targarona J, et al. Debridement and closed packing for sterile or infected necrotizing pancreatitis: insights into indications and outcomes in 167 patients. Ann Surg 2008;247:294-9.

31. Raraty MG, Halloran CM, Dodd S, et al. Minimal access retroperitoneal pancreatic necrosectomy: improvement in morbidity and mortality with a less invasive approach. Ann Surg 2010;251:787-93.

39. Mao EQ, Fei J, Peng YB, et al. Rapid hemodilution is associated with increased sepsis and mortality among patients with severe acute pancreatitis. Chin Med J (Engl) 2010;123:1639-44. 40. Wu BU, Hwang JQ, Gardner TH, et al. Lactated Ringer's solution reduces systemic inflammation compared with saline in patients with acute pancreatitis. Clin Gastroenterol Hepatol 2011;9:710-7 e1. 41. UK guidelines for the management of acute pancreatitis. Gut 2005;54 Suppl 3:iii1-9. 42. Johnson CD, Kingsnorth AN, Imrie CW, et al. Double blind, randomised, placebo controlled study of a platelet activating


TREATMENT OF NECROTIZING PANCREATITIS

factor antagonist, lexipafant, in the treatment and prevention of organ failure in predicted severe acute pancreatitis. Gut 2001;48:62-9. 43. Pettila V, Kyhala L, Kylanpaa ML, et al. APCAP--activated protein C in acute pancreatitis: a double-blind randomized human pilot trial. Crit Care 2010;14:R139. 44. Dervenis C, Smailis D, Hatzitheoklitos E. Bacterial translocation and its prevention in acute pancreatitis. J Hepatobiliary Pancreat Surg 2003;10:415-8. 45. Nieuwenhuijs VB, van Dijk JE, Gooszen HG, et al. Obstructive jaundice, bacterial translocation and interdigestive small-bowel motility in rats. Digestion 2000;62:255-61. 46. Nieuwenhuijs VB, Verheem A, van Duijvenbode-Beumer H, et al. The role of interdigestive small bowel motility in the regulation of gut microflora, bacterial overgrowth, and bacterial translocation in rats. Ann Surg 1998;228:188-93. 47. Van Felius ID, Akkermans LM, Bosscha K, et al. Interdigestive small bowel motility and duodenal bacterial overgrowth in experimental acute pancreatitis. Neurogastroenterol Motil 2003;15:267-76. 48. van Santvoort HC, Besselink MG, Timmerman HM, et al. Probiotics in surgery. Surgery 2008;143:1-7. 49. de Vries AC, Besselink MG, Buskens E, et al. Randomized controlled trials of antibiotic prophylaxis in severe acute pancreatitis: relationship between methodological quality and outcome. Pancreatology 2007;7:531-8. 50. Villatoro E, Mulla M, Larvin M. Antibiotic therapy for prophylaxis against infection of pancreatic necrosis in acute pancreatitis. Cochrane Database Syst Rev 2010:CD002941. 51. Wittau M, Mayer B, Scheele J, et al. Systematic review and meta-analysis of antibiotic prophylaxis in severe acute pancreatitis. Scand J Gastroenterol 2011; 46:261-70. 52. Dellinger EP, Tellado JM, Soto NE, et al. Early antibiotic treatment for severe acute

necrotizing pancreatitis: a randomized, double-blind, placebo-controlled study. Ann Surg 2007;245:674-83. 53. Garcia-Barrasa A, Borobia FG, Pallares R, et al. A double-blind, placebo-controlled trial of ciprofloxacin prophylaxis in patients with acute necrotizing pancreatitis. J Gastrointest Surg 2009;13:768-74. 54. Isenmann R, Runzi M, Kron M, et al. Prophylactic antibiotic treatment in patients with predicted severe acute pancreatitis: a placebo-controlled, double-blind trial. Gastroenterology 2004;126:997-1004. 55. AGA Institute medical position statement on acute pancreatitis. Gastroenterology 2007;132:2019-21. 56. Olah A, Belagyi T, Issekutz A, et al. Randomized clinical trial of specific lactobacillus and fibre supplement to early enteral nutrition in patients with acute pancreatitis. Br J Surg 2002;89:1103-7. 57. Olah A, Belagyi T, Poto L, et al. Synbiotic control of inflammation and infection in severe acute pancreatitis: a prospective, randomized, double blind study. Hepatogastroenterology 2007;54:590-4. 58. Besselink MG, van Santvoort HC, Renooij W, et al. Intestinal barrier dysfunction in a randomized trial of a specific probiotic composition in acute pancreatitis. Ann Surg 2009;250:712-9. 59. Al-Omran M, Albalawi ZH, Tashkandi MF, et al. Enteral versus parenteral nutrition for acute pancreatitis. Cochrane Database Syst Rev 2010:CD002837. 60. Eatock FC, Chong P, Menezes N, et al. A randomized study of early nasogastric versus nasojejunal feeding in severe acute pancreatitis. Am J Gastroenterol 2005; 100:432-9. 61. Kumar A, Singh N, Prakash S, et al. Early enteral nutrition in severe acute pancreatitis: a prospective randomized controlled trial comparing nasojejunal and nasogastric routes. J Clin Gastroenterol 2006;40:431-4.

247


PART III CHAPTER 12

62. Rahman SH, Ammori BJ, Holmfield J, et al. Intestinal hypoperfusion contributes to gut barrier failure in severe acute pancreatitis. J Gastrointest Surg 2003;7:26-35; discussion -6. 63. Marik PE, Zaloga GP. Early enteral nutrition in acutely ill patients: a systematic review. Crit Care Med 2001;29:2264-70. 64. Petrov MS, Pylypchuk RD, Uchugina AF. A systematic review on the timing of artificial nutrition in acute pancreatitis. Br J Nutr 2009;101:787-93. 65. Bakker OJ, van Santvoort HC, van Brunschot S, et al. Pancreatitis, very early compared with normal start of enteral feeding (PYTHON trial): design and rationale of a randomised controlled multicenter trial. Trials 2011;12:73. 66. Fan ST, Lai EC, Mok FP, et al. Early treatment of acute biliary pancreatitis by endoscopic papillotomy. N Engl J Med 1993; 328:228-32. 67. Folsch UR, Nitsche R, Ludtke R, et al. Early ERCP and papillotomy compared with conservative treatment for acute biliary pancreatitis. The German Study Group on Acute Biliary Pancreatitis. N Engl J Med 1997;336:237-42. 68. Neoptolemos JP, Carr-Locke DL, London NJ, et al. Controlled trial of urgent endoscopic retrograde cholangiopancreato足 graphy and endoscopic sphincterotomy versus conservative treatment for acute pancreatitis due to gallstones. Lancet 1988; 2:979-83. 69. Oria A, Cimmino D, Ocampo C, et al. Early endoscopic intervention versus early conservative management in patients with acute gallstone pancreatitis and biliopancreatic obstruction: a randomized clinical trial. Ann Surg 2007;245:10-7. 70. Petrov MS, van Santvoort HC, Besselink MG, et al. Early endoscopic retrograde cholangiopancreatography versus conservative management in acute biliary pancreatitis without cholangitis: a meta-analysis of randomized trials. Ann Surg 2008;247:250-7.

248

71. van Santvoort HC, Besselink MG, de Vries AC, et al. Early endoscopic retrograde cholangiopancreatography in predicted severe acute biliary pancreatitis: a prospective multicenter study. Ann Surg 2009;250:68-75. 72. Malbrain ML, Cheatham ML, Kirkpatrick A, et al. Results from the International Conference of Experts on Intra-abdominal Hy p e r t e n s i on and Ab d om i n a l Compartment Syndrome. I. Definitions. Intensive Care Med 2006;32:1722-32. 73. Chen H, Li F, Sun JB, et al. Abdominal compartment syndrome in patients with severe acute pancreatitis in early stage. World J Gastroenterol 2008;14:3541-8. 74. De Waele JJ, Hoste E, Blot SI, et al. Intraabdominal hypertension in patients with severe acute pancreatitis. Crit Care 2005; 9:R452-7. 75. De Waele JJ, Leppaniemi AK. Intraabdominal hypertension in acute pancreatitis. World J Surg 2009;33:1128-33. 76. Gecelter G, Fahoum B, Gardezi S, et al. Abdominal compartment syndrome in severe acute pancreatitis: an indication for a decompressing laparotomy? Dig Surg 2002;19:402-4; discussion 4-5. 77. Mentula P, Hienonen P, Kemppainen E, et al. Surgical decompression for abdominal compartment syndrome in severe acute pancreatitis. Arch Surg 2010;145:764-9. 78. Cheatham ML, Malbrain ML, Kirkpatrick A, et al. Results from the International Conference of Experts on Intra-abdominal Hypertension and Abdominal Compartment Syndrome. II. Recommendations. Intensive Care Med 2007;33:951-62. 79. Mier J, Leon EL, Castillo A, et al. Early versus late necrosectomy in severe necrotizing pancreatitis. Am J Surg 1997;173:71-5. 80. Kolkman JJ, Mensink PB. Non-occlusive mesenteric ischaemia: a common disorder in gastroenterology and intensive care. Best Pract Res Clin Gastroenterol 2003;17:457-73.


TREATMENT OF NECROTIZING PANCREATITIS

81. Hirota M, Inoue K, Kimura Y, et al. Nonocclusive mesenteric ischemia and its associated intestinal gangrene in acute pancreatitis. Pancreatology 2003;3:316-22. 82. Walser EM, Nealon WH, Marroquin S, et al. Sterile fluid collections in acute pancreatitis: catheter drainage versus simple aspiration. Cardiovasc Intervent Radiol 2006;29:102-7.

92. Connor S, Alexakis N, Raraty MG, et al. Early and late complications after pancreatic necrosectomy. Surgery 2005;137:499-505. 93. Howard TJ, Patel JB, Zyromski N, et al. Declining morbidity and mortality rates in the surgical management of pancreatic necrosis. J Gastrointest Surg 2007;11:43-9.

83. Zerem E, Imamovic G, Omerovic S, et al. Randomized controlled trial on sterile fluid collections management in acute pancreatitis: should they be removed? Surg Endosc 2009;23:2770-7.

94. Rau B, Bothe A, Beger HG. Surgical treatment of necrotizing pancreatitis by necrosectomy and closed lavage: changing patient characteristics and outcome in a 19year, single-center series. Surgery 2005;138:28-39.

84. Besselink MG, van Santvoort HC, Bakker OJ, et al. Draining sterile fluid collections in acute pancreatitis? Primum non nocere! Surg Endosc 2011;25:331-2.

95. Tsiotos GG, Luque-de Leon E, Sarr MG. Long-term outcome of necrotizing pancreatitis treated by necrosectomy. Br J Surg 1998;85:1650-3.

85. van Santvoort HC, Besselink MG, Bakker OJ, et al. Endoscopic necrosectomy in necrotising pancreatitis: indication is the key. Gut 2010;59:1587.

96. Freeny PC, Hauptmann E, Althaus SJ, et al. Percutaneous CT-guided catheter drainage of infected acute necrotizing pancreatitis: techniques and results. AJR Am J Roentgenol 1998;170:969-75.

86. Buchler MW, Gloor B, Muller CA, et al. Acute necrotizing pancreatitis: treatment strategy according to the status of infection. Ann Surg 2000;232:619-26. 87. Fernandez-del Castillo C, Rattner DW, Makary MA, et al. Debridement and closed packing for the treatment of necrotizing pancreatitis. Ann Surg 1998;228:676-84. 88. Uhl W, Warshaw A, Imrie C, et al. IAP Guidelines for the Surgical Management of Acute Pancreatitis. Pancreatology 2002;2:565-73. 89. Traverso LW, Kozarek RA. Pancreatic necrosectomy: definitions and technique. J Gastrointest Surg 2005;9:436-9. 90. Ashley SW, Perez A, Pierce EA, et al. Necrotizing pancreatitis: contemporary analysis of 99 consecutive cases. Ann Surg 2001;234:572-9; discussion 9-80. 91. Babu BI, Sheen AJ, Lee SH, et al. Open pancreatic necrosectomy in the multidisciplinar y management of postinflammatory necrosis. Ann Surg 2010;251:783-6.

97. Baron TH, Thaggard WG, Morgan DE, et al. Endoscopic therapy for organized pancreatic necrosis. Gastroenterology 1996;111:755-64. 98. Gardner TB, Chahal P, Papachristou GI, et al. A comparison of direct endoscopic necrosectomy with transmural endoscopic drainage for the treatment of walled-off pancreatic necrosis. Gastrointest Endosc 2009;69:1085-94. 99. Papachristou GI, Takahashi N, Chahal P, et al. Peroral endoscopic drainage/ debridement of walled-off pancreatic necrosis. Ann Surg 2007;245:943-51. 100. Seewald S, Ang TL, Teng KC, et al. EUSguided drainage of pancreatic pseudocysts, abscesses and infected necrosis. Dig Endosc 2009;21 Suppl 1:S61-5. 101. Varadarajulu S, Phadnis MA, Christein JD, et al. Multiple transluminal gateway technique for EUS-guided drainage of symptomatic walled-off pancreatic necrosis. Gastrointest Endosc 2011;74:74-80.

249


PART III CHAPTER 12

102. Vitale GC, Davis BR, Vitale M, et al. Natural orifice translumenal endoscopic drainage for pancreatic abscesses. Surg Endosc 2009;23:140-6. 103. Charnley RM, Lochan R, Gray H, et al. Endoscopic necrosectomy as primary therapy in the management of infected pancreatic necrosis. Endoscopy 2006;38:925-8. 104. Coelho D, Ardengh JC, Eulalio JM, et al. Management of infected and sterile pancreatic necrosis by programmed endoscopic necrosectomy. Dig Dis 2008;26:364-9. 105. Escourrou J, Shehab H, Buscail L, et al. Peroral transgastric/transduodenal necrosectomy: success in the treatment of infected pancreatic necrosis. Ann Surg 2008;248:1074-80. 106. Gardner TB, Coelho-Prabhu N, Gordon SR, et al. Direct endoscopic necrosectomy for the treatment of walled-off pancreatic necrosis: results from a multicenter U.S. series. Gastrointest Endosc 2011;73:718-26. 107. Haghshenasskashani A, Laurence JM, Kwan V, et al. Endoscopic necrosectomy of pancreatic necrosis: a systematic review. Surg Endosc 2011;25:3724-30. 108. Mathew A, Biswas A, Meitz KP. Endoscopic necrosectomy as primary treatment for infected peripancreatic fluid collections (with video). Gastrointest Endosc 2008;68:776-82. 109. Schrover IM, Weusten BL, Besselink MG, et al. EUS-guided endoscopic transgastric necrosectomy in patients with infected necrosis in acute pancreatitis. Pancreatology 2008;8:271-6. 110. Seewald S, Groth S, Omar S, et al. Aggressive endoscopic therapy for pancreatic necrosis and pancreatic abscess: a new safe and effective treatment algorithm (videos). Gastrointest Endosc 2005;62:92-100. 111. Seifert H, Biermer M, Schmitt W, et al. Transluminal endoscopic necrosectomy after acute pancreatitis: a multicentre study with long-term follow-up (the GEPARD Study). Gut 2009;58:1260-6.

250

112. Voermans RP, Veldkamp MC, Rauws EA, et al. Endoscopic transmural debridement of symptomatic organized pancreatic necrosis (with videos). Gastrointest Endosc 2007;66:909-16. 113. Baril NB, Ralls PW, Wren SM, et al. Does an infected peripancreatic fluid collection or abscess mandate operation? Ann Surg 2000;231:361-7. 114. Carter CR, McKay CJ, Imrie CW. Percutaneous necrosectomy and sinus tract endoscopy in the management of infected pancreatic necrosis: an initial experience. Ann Surg 2000;232:175-80. 115. Horvath K, Freeny P, Escallon J, et al. Safety and efficacy of video-assisted retroperitoneal debridement for infected pancreatic collections: a multicenter, prospective, single-arm phase 2 study. Arch Surg 2010;145:817-25. 116. Horvath KD, Kao LS, Wherry KL, et al. A technique for laparoscopic-assisted percutaneous drainage of infected pancreatic necrosis and pancreatic abscess. Surg Endosc 2001;15:1221-5. 117. van Santvoort HC, Besselink MG, Horvath KD, et al. Videoscopic assisted retroperitoneal debridement in infected necrotizing pancreatitis. HPB (Oxford) 2007;9:156-9. 118. Besselink MG, de Bruijn MT, Rutten JP, et al. Surgical intervention in patients with necrotizing pancreatitis. Br J Surg 2006; 93:593-9. 119. van Baal MC, van Santvoort HC, Bollen TL, et al. Systematic review of percutaneous catheter drainage as primary treatment for necrotizing pancreatitis. Br J Surg 2011;98:18-27. 120. Connor S, Ghaneh P, Raraty M, et al. Minimally invasive retroperitoneal pancreatic necrosectomy. Dig Surg 2003;20:270-7. 121. Parekh D. Laparoscopic-assisted pancreatic necrosectomy: A new surgical option for treatment of severe necrotizing pancreatitis. Arch Surg 2006;141:895-902; discussion -3.


TREATMENT OF NECROTIZING PANCREATITIS

122. Zhu JF, Fan XH, Zhang XH. Laparoscopic treatment of severe acute pancreatitis. Surg Endosc 2001;15:146-8. 123. Castellanos G, Pinero A, Serrano A, et al. Infected pancreatic necrosis: translumbar approach and management with retrop eritoneos copy. Arch Surg 2002;137:1060-3; discussion 3. 124. Fagniez PL, Rotman N, Kracht M. Direct retroperitoneal approach to necrosis in severe acute pancreatitis. Br J Surg 1989;76:264-7. 125. Nakasaki H, Tajima T, Fujii K, et al. A surgical treatment of infected pancreatic necrosis: retroperitoneal laparotomy. Dig Surg 1999;16:506-11. 126. van Santvoort HC, Besselink MG, Bollen TL, et al. Case-matched comparison of the retroperitoneal approach with laparotomy for necrotizing pancreatitis. World J Surg 2007;31:1635-42. 127. Bakker OJ, van Santvoort HC, van Brunschot S, et al. Endoscopic transgastric vs surgical necrosectomy for infected necrotizing pancreatitis: a randomized trial. JAMA 2012;307:1053-61. 128. Windsor JA. Minimally invasive pancreatic necrosectomy. Br J Surg 2007;94:132-3.

251


PART IV SUMMARY AND FUTURE PERSPECTIVES

CHAPTER 13 Summary British Journal of Surgery 2014 CHAPTER 14 Future perspectives Nature Reviews Gastroenterology & Hepatology 2014


PART IV SUMMARY AND FUTURE PERSPECTIVES


CHAPTER 13 Summary British Journal of Surgery. 2014 Jan;101(1):e65-79

Adapted from: Staged multidisciplinary step-up management for necrotizing pancreatitis

David W da Costa, Djamila Boerma, Hjalmar C van Santvoort, Karen D Horvath, Jens Werner, Christopher R Carter, Thomas L Bollen, Hein G Gooszen, Marc G Besselink and Olaf J Bakker


PART IV CHAPTER 13

In this chapter, the contents and major conclusions from chapters 1 to 12 are briefly summarized. We have incorporated the individual study results in a pragmatic treatment algorithm. This algorithm is based on a multidisciplinary staged step-up management strategy and functions as an easy-to-use framework for physicians during the disease course. In the final section of this chapter this algorithm is presented. BACKGROUND Acute pancreatitis is an inflammatory disorder of the pancreas caused by obstruction of the pancreatic duct or by a direct toxic effect on pancreatic acinar cells. This inflammatory response is extremely diverse and varies from minor abdominal complaints followed by swift recovery on one side to climaxing multiple organ failure and death within days on the other.1,2 Acute pancreatitis is the most common gastrointestinal disease requiring acute hospital admission.3 However, each hospital admits only a few patients with pancreatitis complicated by infected pancreatic necrosis or multiple organ failure each year.4 As a result, clinical research in acute pancreatitis mainly consisted of single-center retrospective patient series instead of robust multicenter randomized trials. For this reason, the Dutch Pancreatitis Study Group was founded in 2002 and aimed to increase the understanding but foremost improve the clinical outcome of patients with complicated pancreatitis. This thesis is the result of 7 years of intensive clinical research from our Study Group. PART I EARLY IDENTIFICATION OF C OMPLICATED PANCREATITIS Patients with pancreatitis who will develop complications during the course of the disease need to be identified early for timely treatment. An international collaboration between Brigham and Womenâ&#x20AC;&#x2122;s hospital (Boston, USA), University of Pittsburgh Medical Center (Pittsburgh, USA) and the Dutch Pancreatitis Study Group resulted in an aggregated prospective dataset of 1043 patients with acute pancreatitis (chapter 2). Detailed data from these patients were available and pooled analysis showed that risk of mortality was strongly associated with elevated blood urea nitrogen (BUN) level at admission and rise in BUN level at 24 hours. A BUN level of 20 mg/dL or higher was associated with an odds ratio (OR) of 4.6 (95 percent confidence interval, 2.5-8.3) for mortality. Any rise in BUN level at 24 hours was associated with an OR of 4.3 (95 percent confidence interval, 2.3-7.9) for mortality. A BUN-based assessment algorithm identified patients at increased risk for mortality during the initial 24 hours of hospitalization and may assist physicians in their early resuscitation efforts. In a disease as complicated as acute pancreatitis, correct terminology and clear definitions are essential in interdisciplinary communication among treating physicians as well as in reports of clinical research (chapter 3). To improve the use of uniform terminology, a 256


SUMMARY

revision of the 1992 Atlanta classification was developed. In this new classification, necrotizing pancreatitis refers to patients with necrosis of the pancreatic parenchyma and to patients with only extrapancreatic fat necrosis without pancreatic parenchymal necrosis (EXPN). Patients with EXPN are thought to have a better clinical outcome, although robust data are lacking. From a multicenter prospective dataset, 315 patients with EXPN were compared with 324 patients with pancreatic parenchymal necrosis. Patients with EXPN less often suffered from complications including mortality (9 vs 20 percent, p<0.001). However, when infection of extrapancreatic necrosis developed, outcomes between groups were similar. So, overall EXPN causes fewer complications than pancreatic parenchymal necrosis and should therefore be considered a separate entity in acute pancreatitis. Future studies need to report the number of patients with pancreatic parenchymal necrosis and the number of patients with EXPN. In complicated pancreatitis the extent and duration of organ failure in acute pancreatitis varies considerably. Detailed data on timing, type and combination (respiratory, cardiovascular or renal failure), or duration of organ failure are not available in the literature. These data are needed to clinically estimate the prognosis of patients with pancreatitis and organ failure. The results of detailed analysis of 240 patients with necrotizing pancreatitis and organ failure are presented in chapter 4. In summary, mortality associated with organ failure is strongly depended on timing (for example, organ failure started in the first week, 41 percent mortality), duration (for example, multiple organ failure lasting less than 1 week, 43 percent mortality) and combination (for example, simultaneous persistent failure of all 3 organ systems has 63 percent mortality). In contrast to earlier reports, we found no difference in mortality occurring after 10 days of admission in patients with infected pancreatic necrosis and organ failure compared with organ failure alone (28 versus 30 percent, respectively, relative risk 1.3, P=0.4). Even more, half of patients survived after several weeks of organ failure. These findings may assist clinical decision-making regarding continuation of treatment and end-of-life decisions in patients with persistent multiple organ failure. PART II PREVENTING C OMPLICATED PANCREATITIS In the second part of this thesis, the research is presented that was aimed at preventing complications in acute pancreatitis. Major infections (infected pancreatic necrosis, pneumonia and bacteremia), have a large impact on outcome in acute pancreatitis.5,6 These infections are thought to be mediated by bacterial translocation from the gut provoked by disturbed intestinal motility, bacterial overgrowth and increased mucosal permeability.7-12 Recent insights show that enteral tube feeding is believed to stimulate intestinal motility â&#x20AC;&#x201C; thus reducing bacterial overgrowth - and may increase splanchnic blood flow which helps to preserve the integrity of the gut mucosa.13-20 257


PART IV CHAPTER 13

A meta-analysis including 165 individuals from 8 randomized trials was performed and using multivariable logistic regression, adjusting for predicted disease severity and trial, we studied the effect of timing of enteral nutrition on a composite endpoint of infected pancreatic necrosis, organ failure, or mortality (chapter 5). In the multivariable model, enteral nutrition started within 24 hours from admission compared with enteral nutrition started after 24 hours from admission, reduced the composite endpoint from 45 to 19 percent (adjusted OR 0.44; 95 percent CI 0.20 to 0.96). Organ failure was reduced from 42 to 16 percent (adjusted OR 0.42; 95 percent CI 0.19 to 0.94). In conclusion, these observational data from individuals with acute pancreatitis showed that early enteral nutrition was associated with a large reduction in complications. The meta-analysis as well as all other studies on early enteral nutrition have several important limitations. Therefore, we conducted a multicenter, randomized trial comparing early enteral tube feeding with an oral diet at 72 hours after presentation to the emergency department in patients with acute pancreatitis (PYTHON trial; chapter 6). Patients with acute pancreatitis who were at high risk for complications were randomly assigned to tube feeding within 24 hours after randomization (early group) or to an oral diet initiated 72 hours after presentation (on-demand group), with tube feeding provided if the oral diet was not tolerated. The primary end point was a composite of major infection (infected pancreatic necrosis, bacteremia, or pneumonia) or death during 6 months of follow-up. A total of 208 patients were enrolled and the primary end point occurred in 30 of 101 patients (30 percent) in the early group and in 28 of 104 (27 percent) in the on-demand group (risk ratio, 1.07; 95 percent CI, 0.79 to 1.44, P = 0.76). There were no significant differences between the early group and the on demand group in the rate of major infection (25% and 26%, respectively; P = 0.87) or death (11% and 7%, respectively; P = 0.33). In the on-demand group, 72 patients (69%) tolerated an oral diet and did not require tube feeding. In conclusion, the PYTHON trial did not show the superiority of early enteral tube feeding, as compared with an oral diet after 72 hours, in reducing the rate of infection or death in patients with acute pancreatitis at high risk for complications. The use of the on-demand strategy instead of routine use of tube feeding in all patients with acute pancreatitis and a high risk of complications, will substantially reduce patient discomfort and healthcare costs. In chapter 7, we studied the timing of cholecystectomy after an attack of mild pancreatitis as patients who have suffered from an attack of gallstone pancreatitis are at risk for a second attack as long as the gallstones have not been removed. The aim of the study was to evaluate recurrent gallstone-related events as a consequence of delay in cholecystectomy following mild gallstone pancreatitis. Cholecystectomy was performed after a median of 6 weeks in 188 patients (75,5 percent). Before cholecystectomy, 34 patients (13,7 per cent) 258


SUMMARY

were readmitted for gallstone-related events, including 24 with recurrent gallstone pancreatitis. Eight (7,4 percent) of 108 patients suffered from gallstone-related events after endoscopic sphincterotomy and before cholecystectomy, compared with 26 (18,4 percent) of 141 patients who did not have sphincterotomy (risk ratio 0,51, 95 percent CI 0,27 to 0,94; P = 0,02). Only 142 (53 percent) of 267 patients were treated in accordance with the Dutch guideline. As a result, we conclude that a delay in cholecystectomy after mild gallstone pancreatitis carries a substantial risk of recurrent gallstone-related events. Endoscopic sphincterotomy reduces the risk of recurrent pancreatitis but not of other biliary events. PART III TREATMENT OF C OMPLICATED PANCREATITIS In the third part of this thesis we present the research aimed at improving the outcome of patients with pancreatitis and complications such as infected pancreatic necrosis. In the PANTER trial, we have compared the traditional approach to the treatment of infected pancreatic necrosis, open necrosectomy, with the less invasive step-up approach (chapter 8). The step-up approach consisted of percutaneous drainage followed, if necessary, by minimally invasive retroperitoneal necrosectomy. In this multicenter study, we randomly assigned 88 patients. The primary end point occurred in 31 of 45 patients (69%) assigned to open necrosectomy and in 17 of 43 patients (40%) assigned to the step-up approach (risk ratio with the step-up approach, 0.57; 95 percent CI, 0.38 to 0.87; P = 0.006). Of the patients assigned to the step-up approach, 35 percent were treated with percutaneous drainage only. In conclusion, a minimally invasive step-up approach, as compared with open necrosectomy, reduced the rate of the composite end point of major complications or death among patients with infection of pancreatic necrosis. To investigate whether an even less invasive approach improves outcome even further, we performed the PENGUIN trial (chapter 9). In this pilot trial, we tested whether endoscopic transgastric necrosectomy, a form of natural orifice transluminal endoscopic surgery, may reduce the postprocedural proinflammatory response as measured with serum interleukin 6 (IL-6) levels following intervention for infected pancreatic necrosis. Endoscopic necrosectomy consisted of transgastric puncture, balloon dilatation, retroperitoneal drainage, and necrosectomy. Surgical necrosectomy consisted of videoassisted retroperitoneal debridement or, if not feasible, laparotomy. Endoscopic transgastric necrosectomy reduced the postprocedural IL-6 levels compared with surgical necrosectomy (P=.004). The secondary composite clinical end point occurred less often after endoscopic necrosectomy (20 vs 80 percent; risk difference, 0.60; 95 CI, 0.16 to 0.80; P=0.03). In conclusion, in patients with infected pancreatic necrosis, endoscopic necrosectomy reduced the proinflammatory response as well as the composite 259


PART IV CHAPTER 13

clinical end point compared with surgical necrosectomy. In patients with pancreatic fistulas following pancreatic necrosectomy, endoscopic transpapillary stenting may reduce the time to fistula closure. In chapter 10, we studied the safety, feasibility, and outcome of endoscopic transpapillary stenting in a prospective series of patients and in the literature. Our study showed a reduction in time to fistula closure following stenting. The literature review showed that there is a reasonable success rate (71 percent) of fistula closure following pancreatic stenting. Stent-related complications were found and reported in around 10 percent of patients. From these results we concluded that endoscopic transpapillary stenting is a feasible and safe alternative to conservative treatment in patients with pancreatic fistulas after intervention for necrotizing pancreatitis. In chapter 11, we report the data from 639 consecutive patients with necrotizing pancreatitis treated at 21 Dutch hospitals. The largest prospective series of patients with necrotizing pancreatitis till date. The main results were that the overall mortality rate was 15 percent, the overall intervention rate was 38 percent with 27 percent mortality, the longer the time between admission and intervention, the lower the mortality, and primary catheter drainage had fewer complications than primary necrosectomy. In conclusion, the majority of patients with necrotizing pancreatitis can be treated without an intervention and with low mortality. In patients with infected necrosis, delayed intervention and catheter drainage as first treatment improves outcome. The review in chapter 12 provides an overview of current standards for conservative and invasive treatment of necrotizing pancreatitis. Most of these new standards of treatment were based on data from studies from previous chapters.

260


SUMMARY

Research questions and main findings presented in this thesis. Chapter 2.

What is the accuracy of serum blood urea nitrogen (BUN) as early predictor of complicated pancreatitis? BUN is a simple, cheap and accurate early predictor of mortality in acute pancreatitis and we have developed an algorithm that may assist physicians in their early resuscitation efforts.

3.

What is difference in clinical outcome between patients with pancreatic parenchymal necrosis and patients with extrapancreatic necrosis without necrosis of the pancreatic parenchyma (EXPN)? EXPN causes fewer complications than pancreatic parenchymal necrosis. It should therefore be considered a separate entity in acute pancreatitis. Outcome in infected pancreatic necrosis or infected EXPN is similar.

4.

What is the impact of organ failure on mortality in necrotizing pancreatitis? Although mortality is high in patients with necrotizing pancreatitis and persisting organ failure, half of patients may survive after several weeks of organ failure.

5.

Based on individual patient data from randomized trials, does early enteral tube feeding improve outcome in patients with acute pancreatitis? In this meta-analysis of observational data from individuals with acute pancreatitis, starting EN within 24 hours after hospital admission, compared with after 24 hours, was associated with a reduction in complications and improved clinical outcome.

6.

Does early enteral tube feeding compared with an oral diet after 72 hours with tube feeding only if necessary, reduce the number of infections or death in patients with pancreatitis who are at risk for complications? The PYTHON trial did not demonstrate superiority of early nasoenteric tube feeding, compared with oral diet after 72 hours, in reducing infections or death in patients with acute pancreatitis at high risk for complications.

7.

What is the optimal timing of cholecystectomy following an attack of gallstone pancreatitis? A delay in cholecystectomy after mild biliary pancreatitis carries a substantial risk of recurrent biliary events. Cholecystectomy should performed during the primary admission for mild gallstone pancreatitis.

8.

Does a â&#x20AC;&#x2DC;step-up approachâ&#x20AC;&#x2122;, with the use of percutaneous drainage and minimally invasive necrosectomy if necessary, reduce the number of complications or death compared with primary open necrosectomy in patients with necrotizing pancreatitis? A minimally-invasive step-up approach, as compared with primary open necrosectomy, reduces the combination of major complications or death, as well as long term complications, healthcare utilization and total costs in patients with necrotizing pancreatitis.

9.

Does endoscopic transgastric necrosectomy compared with primary necrosectomy reduce the proinflammatory response and improve clinical outcome in patients with necrotizing pancreatitis? In patients with infected pancreatic necrosis, endoscopic necrosectomy reduced the proinflammatory response as well as the composite clinical end point compared with surgical necrosectomy.

261


PART IV CHAPTER 13

10.

In patients with pancreatic fistulas following pancreatic necrosectomy, does endoscopic transpapillary stenting compared with conservative treatment reduce the time to fistula closure? Endoscopic transpapillary stenting is a feasible and safe alternative to conservative treatment and may shorten time to fistula closure in patients with pancreatic fistulas after intervention for necrotizing pancreatitis.

11.

What is the outcome of conservative and minimally invasive approaches in patients with necrotizing pancreatitis? The majority of patients with necrotizing pancreatitis can be treated without an intervention and with low mortality. In patients with infected pancreatic necrosis, delayed intervention and catheter drainage as first treatment improves outcome.

12.

What is the current state-of-the art approach to patients with necrotizing pancreatitis? Intervention in necrotizing pancreatitis has shifted from early primary open necrosectomy to a step-up approach, starting with catheter drainage if needed, followed by minimally invasive necrosectomy or endoscopic necrosectomy once peripancreatic collections have sufficiently demarcated.

S TAG E D S T E P- U P M A NAG E M E N T In necrotizing pancreatitis, the type of complication that may develop is closely related to the time from symptom onset, and specific complications may be managed differently at different time points. Therefore, this paragraph including the treatment algorithm (Figure 1) briefly summarizes staged multidisciplinary ‘step-up’ strategies for necrotizing pancreatitis according to time from onset of symptoms. Frequent clinical evaluation of the patient’s condition remains paramount in the first 24 to 72 hours of the disease. Liberal goal-directed fluid resuscitation and early enteral nutrition should be provided. Urgent endoscopic retrograde cholangiopancreatography is indicated when cholangitis is suspected, but it is unclear whether this is appropriate in patients with predicted severe gallstone pancreatitis without cholangitis. Antibiotic prophylaxis does not prevent infection of necrosis and antibiotics are not indicated as part of initial management. Bacteriologically confirmed infections should receive targeted antibiotics. With the more conservative approach to necrotizing pancreatitis currently advocated, fine needle aspiration culture of pancreatic or extrapancreatic necrosis will less often lead to a change in management and is therefore indicated less frequently. Optimal treatment of infected pancreatic necrosis consists of a staged multidisciplinary ‘step-up’ approach. The initial step is drainage, either percutaneous or transluminal, followed by surgical or endoscopic transluminal debridement only if needed. Debridement is delayed until the acute necrotic collection has become ‘walled-off ’.

262


SUMMARY

Figure 1. Suggested treatment algorithm for necrotizing pancreatitis according to the time after onset of symptoms. CECT, contrast-enhanced computed tomography; SIRS, systemic inflammatory response syndrome; ERCP, endoscopic retrograde cholangiopancreatography; MRI, magnetic resonance imaging; FNA, fine-needle aspiration.

263


APPENDIX CHAPTER 13

REFERENCES 1. Whitcomb DC. Clinical practice. Acute pancreatitis. N Engl J Med 2006;354(20):214250. 2. Lankisch PG, Apte M, Banks PA. Acute pancreatitis. Lancet. 2015 Jan 20. pii: S01406736(14)60649-8. 3. Peery AF, Dellon ES, Lund J, Crockett SD, McGowan CE, Bulsiewicz WJ et al. Burden of gastrointestinal disease in the United States: 2012 update. Gastroenterology 2012; 143(5):1179-1187. 4. Spanier B, Bruno MJ, Dijkgraaf MG. Incidence and mortality of acute and chronic pancreatitis in the Netherlands: a nationwide record-linked cohort study for the years 1995-2005. World J Gastroenterol 2013;19(20):3018-26. 5. Besselink MG, van Santvoort HC, Boermeester MA, Nieuwenhuijs VB, van Goor H, Dejong CH et al. Timing and impact of infections in acute pancreatitis. Br J Surg 2009;96(3):267-73. 6. Wu BU, Johannes RS, Kurtz S, Banks PA. The impact of hospital-acquired infection on outcome in acute pancreatitis. Gastroenterology 2008;135(3):816-20. 7. Ammori BJ, Leeder PC, King RF, Barclay GR, Martin IG, Larvin M et al. Early increase in intestinal permeability in patients with severe acute pancreatitis: correlation with endotoxemia, organ failure, and mortality. J Gastrointest Surg 1999;3(3):252-62. 8. Besselink MG, van Santvoort HC, Renooij W, de Smet MB, Boermeester MA, Fischer K et al. Intestinal barrier dysfunction in a randomized trial of a specific probiotic composition in acute pancreatitis. Ann Surg 2009;250(5):712-9. 9. Fritz S, Hackert T, Hartwig W, Rossmanith F, Strobel O, Schneider L et al. Bacterial translocation and infected pancreatic necrosis in acute necrotizing pancreatitis derives from small bowel rather than from colon. Am J Surg 2010;200(1):111-7.

264

10. Nieuwenhuijs VB, Verheem A, van Duijvenbode-Beumer H, Visser MR, Verhoef J, Gooszen HG et al. The role of interdigestive small bowel motility in the regulation of gut microflora, bacterial overgrowth, and bacterial translocation in rats. Ann Surg 1998;228(2):188-93. 11. Rahman SH, Ammori BJ, Holmfield J, Larvin M, McMahon MJ. Intestinal hypoperfusion contributes to gut barrier failure in severe acute pancreatitis. J Gastrointest Surg 2003;7(1):26-35; discussion -6. 12. Van Felius ID, Akkermans LM, Bosscha K, Verheem A, Harmsen W, Visser MR et al. Interdigestive small bowel motility and duodenal bacterial overgrowth in experimental acute pancreatitis. Neurogastroenterol Motil 2003;15(3):267-76. 13. Marik PE. What is the best way to feed patients with pancreatitis? Curr Opin Crit Care 2009;15(2):131-8. 14. McClave SA, Heyland DK. The physiologic response and associated clinical benefits from provision of early enteral nutrition. Nutr Clin Pract 2009;24(3):305-15. 15. Ziegler TR. Parenteral nutrition in the critically ill patient. N Engl J Med. 2009 Sep 10;361(11):1088-97. 16. Marik PE, Zaloga GP. Early enteral nutrition in acutely ill patients: a systematic review. Crit Care Med 2001;29(12):2264-70. 17. Li JY, Yu T, Chen GC, Yuan YH, Zhong W, Zhao LN et al. Enteral nutrition within 48 hours of admission improves clinical outcomes of acute pancreatitis by reducing complications: a meta-analysis. PLoS One 2013;8(6):e64926. 18. Petrov MS, Pylypchuk RD, Uchugina AF. A systematic review on the timing of artificial nutrition in acute pancreatitis. Br J Nutr 2009;101(6):787-93. 19. Sun JK, Mu XW, Li WQ, Tong ZH, Li J, Zheng SY. Effects of early enteral nutrition on immune function of severe acute pancreatitis patients. World J Gastroenterol 2013;19(6):917-22.


SUMMARY

20. Wereszczynska-Siemiatkowska U, SwidnickaSiergiejko A, Siemiatkowski A, Dabrowski A. Early enteral nutrition is superior to delayed enteral nutrition for the prevention of infected necrosis and mortality in acute pancreatitis. Pancreas 2013;42(4):640-6.

265


PART IV SUMMARY AND FUTURE PERSPECTIVES


CHAPTER 14 Discussion and Future Perspectives Nature Reviews Gastroenterology & Hepatology. 2014 Aug;11(8):462-9

Adapted from: Treatment options for acute pancreatitis

Olaf J Bakker, Yama Issa, Hjalmar C van Santvoort, Marc G Besselink, Nicolien J Schepers, Marco J Bruno, Marja A Boermeester, and Hein G Gooszen for the Dutch Pancreatitis Study Group


PART IV CHAPTER 14

The research presented in this thesis has answered several topical questions in acute pancreatitis. Some of these answers have led to clear changes in management and hopefully will lead to improved outcomes for patients with pancreatitis. However, we need to understand that, in 2015, for a subgroup of patients with complicated pancreatitis the prognosis is still very poor. The Dutch Pancreatitis Study Group will continue its activities and several new and exciting trials are underway. In this chapter, we discuss the latest developments and potential research directions.

R E C E N T C HA N G E S I N C L A S SI F IC AT ION In 1992, a classification system for acute pancreatitis was developed during a consensus conference held in Atlanta, Georgia.1 ‘Atlanta 1992’ aimed to universally define the different local and systemic complications that can develop in acute pancreatitis. At the time, the classification was an important step forward, however, some of the definitions turned out to be confusing. Especially terms related to the content of peripancreatic collections, such as acute fluid collection, pseudocyst and even pancreatic necrosis, were interpreted differently.2-4 Systemic complications included in Atlanta 1992, such as a low serum calcium or low platelet count, were found to be general symptoms of critical illness instead of specific complications of pancreatitis. For these reasons, a clear need to revise the old Atlanta classification became apparent and an international web-based consultation was undertaken.5 The revised classification, ‘Atlanta 2012’, divides acute pancreatitis into 3 categories: mild, moderate and severe disease (Table 1). Box 1 presents the different local complications. The main argument put forward by the authors to make this distinction is to separate fluid collections containing only fluid from collections that contain a solid component (and thus contain an amount of necrosis). The amount of necrosis is related to specific complications that can occur such as infection of necrosis or pancreatic insufficiency and thereby determine patient outcome.6 The proposed distinction seems logical for stratification of patients for treatment and research purposes. However, the most used imaging technique in acute pancreatitis is contrast enhanced CT. And based on CT images it can sometimes be difficult, if not impossible, to distinct fluid from necrosis.7 MRI can clearly distinct fluid from solid components but in daily practice is not as easily performed as CT. This limitation of CT to some extent restricts the implementation of the new classification. Also, in the new classification the different types of local complications are defined according to the time after onset of disease. A cut-off value of around 4 weeks is proposed. For example, acute necrotic collections develop within 4 weeks and are defined as collections containing a variable amount of fluid or necrotic tissue. After 4 weeks, these collections start to become encapsulated by an enhancing wall of reactive tissue and are then named walled off necrosis. The timing 268


DISCUSSION AND FUTURE PERSPECTIVES

Box 1. Local complications in acute pancreatitis* Interstitial oedematous pancreatitis • A  cute peripancreatic fluid collection (peripancreatic fluid associated with interstitial oedematous pancreatitis with no associated extrapancreatic necrosis) • P  ancreatic pseudocyst (an encapsulated collection of fluid with a well-defined inflammatory wall usually outside the pancreas with minimal or no necrosis) Necrotizing pancreatitis • A  cute necrotic collection (a collection containing variable amounts of both fluid and necrosis associated with necrotizing pancreatitis; the necrosis can involve the pancreatic parenchyma or the extrapancreatic tissues) • W  alled-off necrosis (a mature, encapsulated collection of pancreatic or extrapancreatic necrosis that has developed a well-defined inflammatory wall) Infected pancreatic necrosis • A  cute necrotic collection • Walled-off necrosis Other local complications • Gastric outlet dysfunction • Splenic or portal vein thrombosis • Colonic necrosis *According to the 2012 Revised Atlanta Classification

of encapsulation, however, differs strongly between patients and can only be judged on CECT rather than according to time from symptom onset. It is advised to follow up on these collections during the disease course to eventually differentiate one from the other. The 2012 Revised Atlanta Classification repeatedly uses the term 'collection'. However, there is no definition of ‘collection’. It is used as a more broad term including several local complications. One may argue that there has to be a definition of ‘collection’. A collection in this context would contain fluid or necrosis (liquefaction of solid tissue through the process of tissue death) together in one place. The definition of pseudocysts seems incomplete as this need to contain amylase rich fluid and is lined not by epithelium but rather granulation tissue, probably a more correct term than "inflammatory wall" which is used in the revised Atlanta classification. In the context of severe necrotizing pancreatitis, ‘collection’ is used in the definition of walled-off necrosis. However, necrosis spreads in various degrees throughout the retroperitoneum and into the small and large mesentery as there are no anatomical barriers. If expansive, then the necrosis cannot be said to be co-located into a ‘collection’. An alternative description would be to divide pancreatic necrosis into pancreatic necrosis that is ‘limited’ or pancreatic necrosis that is ‘expansive’. Systemic complications are defined as persistent organ failure or an exacerbation of a pre-existing co-morbidity, such as coronary artery disease or chronic lung disease, precipitated by the acute pancreatitis attack. However, it can sometimes be very difficult to distinguish whether the acute pancreatitis attack (which may cause SIRS and SIRS may lead to organ failure) or the exacerbation of the pre-existing co-morbidity caused the 269


PART IV CHAPTER 14

development of organ failure. One may also argue that as organ failure is already included in the classification, the inclusion of exacerbation of a pre-existing co-morbidity in the classification is redundant. On the other hand, patients with pre-existing co-morbidity may also be admitted without SIRS, without local complications, without organ failure and still may be admitted for a prolonged time at the hospital; for example, pulmonary failure might occur chronic obstructive pulmonary disease. This scenario is now considered ‘moderate severe’ pancreatitis (that is, when a pre-existing comorbid condition is exacerbated by the pancreatitis attack). Persistent organ failure is emphasized as the defining feature of severe pancreatitis and is defined as organ failure persisting for more than 48 hours. The modified Marshall scoring system is used to define organ failure in one of the three organ systems.8 Although, proposed as a relative simple and easy to use scoring system, the modified Marshall score has 2 major disadvantages. First, patients with persisting organ failure are sometimes admitted for over 6 months in the hospital. It is cumbersome to calculate the modified Marshall score for 3 separate organ systems on a daily basis throughout the whole disease course. Especially, when retrospective cohorts describe the incidence of organ failure with the modified Marshall system one might question the accuracy of the reported organ failure rates.3 Second, endotracheal intubation for respiratory failure, the use of vasopressors for persistent hypotension after fluid resuscitation and renal replacement

Table 1. Classifications of Acute Pancreatitis Complications

1992 Atlanta 2012 Revised Atlanta

Determinant-based classification

Mild Severe Mild Moderate Severe Mild Moderate

Severe

Critical

No

Yes

No

Yes

Yes

No

Sterile

Infected

Infected

Transient organ failure

No

Yes

No

Yes

No

No

Yes

No

No

Persistent organ failure

No

Yes

No

No

Yes

No

No

Yes

Yes

Exacerbation of preexisting co-morbidity

NA

NA

No

Yes

Yes/No NA†

NA

NA

NA

Local complications* Systemic complications**

* In Atlanta 2012, several different local complications are distinguished (see also Box 1). ** In Atlanta 2012, systemic complications are defined as persistent organ failure or an exacerbation of a pre-existing co-morbidity, such as coronary artery disease or chronic lung disease, precipitated by the acute pancreatitis attack. Persistent organ failure is defined as organ failure persisting for more than 48 hours. Three organ systems are addressed: respiratory, cardiovascular and renal. Organ failure is defined as a score of 2 or more using the modified Marshall scoring system for respiratory, cardiovascular and renal organ systems. In the Determinant-based classification the SOFA score is used or when a certain threshold is breached.15 † NA denotes not applicable. ‡ For severe pancreatitis either infected pancreatic necrosis or persistent organ failure is mandatory

270


DISCUSSION AND FUTURE PERSPECTIVES

therapy for renal insufficiency are not incorporated in the definitions of â&#x20AC;&#x2DC;Atlanta 2012â&#x20AC;&#x2122;. If patients receive these supportive therapies, this may result in normal arterial oxygen measurements and normal systolic blood pressures. Hence, the Marshall score will not reach the threshold for organ failure and these patients would be categorised as having moderate pancreatitis. Despite these minor weaknesses, Atlanta 2012 is again a major step forward and is likely to improve communication between physicians and improve accuracy of studies that describe cohorts of pancreatitis patients. Parallel to the revision of the Atlanta classification, a second initiative was undertaken to likewise redefine acute pancreatitis and its complications.10 In contrast to Atlanta 2012, the determinant based classification proclaims to be based on determinants of severity and the epidemiological concept of causal inference. However, when comparing both classifications, a similar subdivision of local and systemic complications is found. Indeed, there are some differences. For example, organ failure is defined using the sepsis-related organ failure (SOFA) score instead of the modified Marshall score. Next to mild, moderate and severe, a fourth category is proposed: critical acute pancreatitis. This is characterized by the presence of infected necrosis and persistent organ failure. Actually, our study (chapter 4) showed that the presence of organ failure and infected pancreatic necrosis does not lead to a higher mortality than organ failure without infected pancreatic necrosis. Eventually, the existence of 2 parallel classifications systems seems conflicting. Widespread implementation of either one of the classifications in daily practice is likely to be limited by their co-existence. The near future will hopefully show which system is to be adopted by the pancreatic community.

N E W I N SIG H T S I N TO E A R LY M A NAG E M E N T E arly i d entif i c ati on of compli c ate d p ancre atitis In the absence of a specific drug therapy for patients with acute pancreatitis, all early management strategies are mainly supportive. As the clinical course varies widely, it is important to predict complications on admission or within 24 to 48 hours thereafter to initiate further management strategies such as early intensive care admission. A recent head-to-head comparison of the nine most frequently used systems showed a moderate overall efficacy of these scoring systems.11 Complex combinations of different scoring systems might improve efficacy but are virtually impossible to implement in daily practice. Perhaps novel approaches, for example rapid determination of biomarkers, might further improve accuracy. Recently, more simplistic ways to predict complications have been tested. Blood Urea Nitrogen (BUN) on admission and after 24 hours turned out to be an accurate, readily available, low cost and universal way to predict mortality (chapter 2).12 Creatinin and hematocrit levels are simple measures to predict the 271


PART IV CHAPTER 14

development of pancreatic necrosis.13 Another frequently used predictor with an increasing accuracy beyond 48 hours is C-reactive protein.14 However, the modified Glasgow score and C-reactive protein are calculated or measured after 48 hours after admission. And after 48 hours, scoring systems or single predictors actually reflect ongoing organ failure instead of predicting pending organ failure. The Systemic Inflammatory Response Syndrome (SIRS) can be used to monitor disease progression. A substantial number of patients with acute pancreatitis have SIRS on admission, but persistent SIRS after 24 to 48 hours after admission is strongly associated with the development of organ failure.15 The new International Association of Pancreatology and American Association of Pancreatology (IAP/APA) guideline advises repetitive monitoring of the presence of SIRS during the first days of admission.16

E A R LY SU P P ORT I V E M E A SU R E S F lui d ther apy Probably the most important component of early supportive measures in patients with acute pancreatitis is fluid suppletion. Third-space fluid loss can cause intravascular hypovolemia and might evoke organ failure. Liberal fluid suppletion to correct or preferably prevent intravascular hypovolemia is commonly accepted as the first step in supportive care. A practical approach for the amount of fluids to administer would be: 1) bolus infusion of 20 mL/kg of Lactated Ringerâ&#x20AC;&#x2122;s in the emergency department and 2) followed by a total infusion of approximately 2500 to 4000 ml in the first 24 hours after admission.16,17 Vital signs, physical examination and urine output should be repeatedly assessed every 6 hours. After 24 hours, fluid resuscitation can be tailored according to vital signs and urinary output. If any signs of organ dysfunction arise despite adequate resuscitative measures, intensive care admission is advised. Although, only a minority of all patients with acute pancreatitis develops organ failure during the first 72 hours after admission, it is crucial to recognise these patients early. Nutriti on Other supportive measures include nutrition, antibiotics and, in biliary pancreatitis, endoscopic retrograde pancreaticocholangiography (ERCP). Nutritional support is nowadays considered a therapeutic measure instead of merely a way to provide calories in a patient with severe pancreatitis. Our PYTHON trial showed that early enteral nutrition was not able to reduce the composite endpoint of infections or mortality (chapter 6).18

272


DISCUSSION AND FUTURE PERSPECTIVES

Antibi oti cs Prophylactic administration of intravenous antibiotics has been used with the aim to prevent infected necrosis and other infectious complications. Several randomised trials have shown that routine use of antibiotics does not prevent infected necrosis.19 Therefore, antibiotics in acute pancreatitis are indicated in case of proven infection or in case of a very strong clinical suspicion of cholangitis or infected necrosis.16 E nd os copi c retrog r ad e p ancre ati co chol ang i og r aphy Urgent ERCP with sphincterotomy is indicated in patients with biliary pancreatitis and concomitant cholangitis, whereas ERCP is not indicated in patients with mild biliary pancreatitis.16,20 The verdict whether ERCP with sphincterotomy should be performed routinely in patients with predicted severe biliary pancreatitis without cholangitis is still undecided. Current meta-analyses and guidelines show conflicting results.21 Although several randomised trials have been published on this matter, the optimal policy for patients with predicted severe biliary pancreatitis but without cholangitis cannot be extracted as these trials also included mild pancreatitis patients and patients with cholangitis. A multicenter randomised trial from our group including this specific group of patients has started recruitment (ISRCTN97372133, http://www.controlled-trials.com/ ISRCTN97372133/apec).

N OV E L T H E R A P I E S I N AC U T E PA N C R E AT I T I S Hemof i ltr ati on and p eritone a l di a lysis In severe pancreatitis, a local inflammatory process climaxes into a systemic inflammation with concomitant organ failure. Pro-inflammatory cytokines as mediators of inflammation are believed to play an important role in the exacerbation of the disease.22 A potential new therapeutic measure in acute pancreatitis is the elimination of these mediators with the use continuous veno-venous hemofiltration (CVVH) or peritoneal dialysis.23 Two small Chinese trials randomized patients to hemofiltration with or without peritoneal dialysis or to standard therapy without hemofiltration or peritoneal dialysis.24,25 Therapy was started daily during 4 to 6 hours from hospitalization. Reported results seem promising and the authors conclude that, compared to controls, following hemofiltration and peritoneal dialysis all cytokines dropped significantly sooner, the time to relief of abdominal pain was shortened, CT severity scores decreased, APACHE II scores decreased and the incidence of organ failure was decreased. Unfortunately, detailed insight into the early therapeutic strategy is lacking (for example, failed attempts of hemodialysis during the first days of severe pancreatitis when patients can be hemodynamically unstable). Also some other methodological weaknesses (for example, 273


PART IV CHAPTER 14

the analyses were not based on the intention-to-treat principle, no adverse events are reported, no information on end point adjudication is provided) hamper the validity of the results. So far, no convincing evidence has been produced to justify widespread application of these blood purification techniques. Hopefully in the near future, insights into the efficacy of this therapeutic measure will be broadened. C ontinu ous Reg i ona l Infusi on of a Prote as e In hibitor One of the possible mechanisms of the development of pancreatic necrosis is thought to be microvascular obstruction caused by local vasospasm and increased intravascular coagulability.26 Experimental studies have shown that intra-arterial protease inhibitors can inhibit the coagulation system, inhibit proteases such as trypsin, a pancreatic enzyme, and have other effects such as inhibition of the cytokine production.27 The use of protease inhibitors combined with antibiotics may prevent secondary infection of pancreatic necrosis by facilitating the delivery of antibiotics into the pancreatic tissue. Several studies have investigated the effect of protease inhibitors in acute pancreatitis. A study from Poland randomized 78 patients to either continuous regional arterial infusion with antibiotics or to standard treatment without proteases but with antibiotics.28 Although no differences were found in infectious complications, mortality following infusion of a protease inhibitor was significantly reduced. In line with the first positive results after hemofiltration in acute pancreatitis, also these first positive results are criticized due to methodological weaknesses. Potential baseline differences, no reporting of infected pancreatic necrosis and organ failure as outcomes, the long recruitment period (6 years) and high drop-out rates are several of these weaknesses. Despite convincing evidence, protease inhibitors have been recommended in the Japanese guidelines for treatment of acute pancreatitis for several years.29 A recent in-patient registration study including more than 1.000 Japanese hospitals, investigated mortality rates after infusion of protease inhibitors and used propensity-score matching to compare outcomes to controls.30 Unfortunately, continuous regional infusion of protease inhibitors was not able to reduce mortality. In contrast, hospital stay was prolonged and costs were higher with the use of protease inhibitors. Probably, these conflicting results provide enough clinical equipoise for a large clinical trial on the use of protease inhibitors. However, it is unlikely that such a trial will be feasible in the near future.

NEW INSIGHTS INTO MANAGEMENT OF INFECTED NECROSIS D el ay ing Inter venti ons The question when to intervene in patients with suspected or proven infected pancreatic necrosis, has puzzled pancreatologists for years. Very early surgical intervention (i.e. 274


DISCUSSION AND FUTURE PERSPECTIVES

within 72 hours after admission) is clearly harmful.31 In recent series describing results after intervention for necrotizing pancreatitis, intervention is performed after a median of approximately 4 weeks after onset of symptoms (either drainage or necrosectomy).32,33 Postponement of intervention theoretically provides an opportunity for acute necrotic collections to become walled-off and for the necrosis to liquefy. This facilitates necrosectomy and probably improves the outcome of invasive interventions such as percutaneous catheter drainage (PCD), endoscopic transluminal drainage or surgical or endoscopic necrosectomy. However, in severely ill patients, if intervention is deemed necessary, this should not be postponed just to wait for encapsulation. Step-up appro ach The PANTER trial (chapter 8) has shown that PCD as the first step in treatment of infected necrosis, followed by necrosectomy only when the patient does not improve, when compared to primary necrosectomy, reduces new-onset organ failure and prevents the need for necrosectomy in about a third of patients.34 This so-called step-up approach could be considered the standard approach to infected necrotizing pancreatitis. Whether the first step (i.e. drainage), should be performed endoscopically or percutaneously, is subject of ongoing research and probably also depends on local expertise. Endoscopic transluminal drainage has the theoretical advantage of preventing the development of an external pancreatic fistula as compared to PCD. Such fistulas prolong hospital stay, are cumbersome for patients and sometimes may be difficult to treat.35 Internal drainage of a pancreatic fistula may enhance recovery. Ne cros e c tomy : Minima l ly invasive surger y or end os copy? The traditional approach for necrosectomy is open surgical necrosectomy with complete removal of all infected pancreatic necrosis.36 Reported complication rates vary between institutions but are reported in up to 80% of patients.37 Minimally invasive techniques such as video-assisted retroperitoneal debridement (VARD) or minimal access retroperitoneal pancreatic necrosectomy (MARPN) might reduce complications.32;37-39 Endoscopic necrosectomy, using a transgastric or transduodenal approach, is theoretically even less invasive. It is a form of natural orifice transluminal endoscopic surgery (NOTES) and can be performed under conscious sedation without the need for general anaesthesia. A randomised pilot trial comparing surgical to endoscopic necrosectomy showed a reduced pro-inflammatory response following endoscopic necrosectomy (chapter 9).40 Interestingly, also a reduction in the composite endpoint of death and major complications was found. These promising but preliminary results require urgent validation. A multicenter randomised trial comparing endoscopic to surgical â&#x20AC;&#x2DC;step-up approachâ&#x20AC;&#x2122; in patients with infected necrotizing pancreatitis has recently completed recruitment (TENSION trial; ISRCTN09186711; http://www.controlled-trials.com/ 275


PART IV CHAPTER 14

ISRCTN09186711). Until results are available, the method and route of drainage and necrosectomy depend largely on local experience available in each institution. Interactive collaboration between surgeons, gastroenterologists, intensivists and radiologists in the treatment of these complex subgroup of patients with regard to the use of supportive measures and type and timing of interventions to manage infected pancreatic necrosis is crucial.

F U T U R E P E R SP E C T I V E S Despite the promising results of numerous experimental and small clinical studies in acute pancreatitis, eventually all adequately powered clinical trials in acute pancreatitis validating these early results turned out negative. Antibiotics, probiotics, antiinflammatory drugs such as lexipafant, and even early enteral nutrition have all been tested and did not improve outcome. 41-43 From these findings we need to conclude two things. First, all promising results from small studies require validation in large scale trials. Second, there is still an urgent need for new therapeutic agents. In a disease with increasing incidence and already over 270.000 admissions each year in the US, a strong collaborative platform is needed to rapidly test new therapeutic agents. Randomized trials remain the most powerful tool to evaluate the effectiveness of new agents while accounting for unmeasured confounders or selection bias. But large randomised trials are very complex, costly and require considerable time to recruit patients. A new tool for the pancreatic community may be the registry-based randomised trial.44 With the use of large scale low cost observational registries, new randomized trials may be performed with considerable less cost, in less time and in a real world population. Unfortunately, no large scale registries for patients with acute pancreatitis exist thus far. So, the first logical step would be to start enrolling patients in observational registries across multiple international sites. The researchers from CAPER (Collaborative Alliance for Pancreas Education and Research) from the United States have taken the initiative to construct such a registry. Once these registries are ongoing, they might provide a potent platform for the testing of new therapeutic agents. As acute pancreatitis will probably cause increasing mortality and increasing costs in future years, new therapeutic agents are very much needed.

276


DISCUSSION AND FUTURE PERSPECTIVES

REFERENCES 1. Bradley EL, III. A clinically based classification system for acute pancreatitis. Summary of the International Symposium on Acute Pancreatitis, Atlanta, Ga, September 11 through 13, 1992. Arch Surg 1993; 128(5):586-590. 2. Bollen TL, van Santvoort HC, Besselink MG, van Leeuwen MS, Horvath KD, Freeny PC et al. The Atlanta Classification of acute pancreatitis revisited. Br J Surg 2008; 95(1):621. 3. Besselink MG, van Santvoort HC, Bollen TL, van Leeuwen MS, Lameris JS, van der Jagt EJ et al. Describing computed tomography findings in acute necrotizing pancreatitis with the Atlanta classification: an interobserver agreement study. Pancreas 2006; 33(4):331-335. 4. Bollen TL. Imaging of acute pancreatitis: update of the revised Atlanta classification. Radiol Clin North Am 2012; 50(3):429-445. 5. Banks PA, Bollen TL, Dervenis C, Gooszen HG, Johnson CD, Sarr MG et al. Classification of acute pancreatitis--2012: revision of the Atlanta classification and definitions by international consensus. Gut 2013; 62(1):102-111. 6. Bollen TL, Singh VK, Maurer R, Repas K, van Es HW, Banks PA et al. Comparative evaluation of the modified CT severity index and CT severity index in assessing severity of acute pancreatitis. AJR Am J Roentgenol 2011; 197(2):386-392. 7. Bakker OJ, van Santvoort HC, Besselink MG, Boermeester MA, van EC, Dejong K et al. Extrapancreatic necrosis without pancreatic parenchymal necrosis: a separate entity in necrotising pancreatitis? Gut 2012. 8. Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med 1995; 23(10):1638-1652. 9. Petrov MS, Shanbhag S, Chakraborty M, Phillips AR, Windsor JA. Organ failure and

infection of pancreatic necrosis as determinants of mortality in patients with acute pancreatitis. Gastroenterology 2010; 139(3):813-820. 10. Dellinger EP, Forsmark CE, Layer P, Levy P, Maravi-Poma E, Petrov MS et al. Determinant-based classification of acute pancreatitis severity: an international multidisciplinary consultation. Ann Surg 2012; 256(6):875-880. 11. Mounzer R, Langmead CJ, Wu BU, Evans AC, Bishehsari F, Muddana V et al. Comparison of existing clinical scoring systems to predict persistent organ failure in patients with acute pancreatitis. Gastroenterology 2012; 142(7):1476-1482. 12. Wu BU, Bakker OJ, Papachristou GI, Besselink MG, Repas K, van Santvoort HC et al. Blood urea nitrogen in the early assessment of acute pancreatitis: an international validation study. Arch Intern Med 2011; 171(7):669-676. 13. Papachristou GI, Muddana V, Yadav D, Whitcomb DC. Increased serum creatinine is associated with pancreatic necrosis in acute pancreatitis. Am J Gastroenterol 2010; 105(6):1451-1452. 14. Neoptolemos JP, Kemppainen EA, Mayer JM, Fitzpatrick JM, Raraty MG, Slavin J et al. Early prediction of severity in acute pancreatitis by urinary trypsinogen activation peptide: a multicentre study. Lancet 2000; 355(9219):1955-1960. 15. Singh VK, Wu BU, Bollen TL, Repas K, Maurer R, Mortele KJ et al. Early systemic inflammatory response syndrome is associated with severe acute pancreatitis. Clin Gastroenterol Hepatol 2009; 7(11):1247-1251. 16. IAP/APA evidence-based guidelines for the management of acute pancreatitis. Pancreatology 2013; 13(4 Suppl 2):e1-15. 17. Wu BU, Hwang JQ, Gardner TH, Repas K, Delee R, Yu S et al. Lactated Ringer's solution reduces systemic inflammation compared with saline in patients with acute pancreatitis. Clin Gastroenterol Hepatol 2011; 9(8):710717.

277


CHAPTER

18. Bakker OJ, van Brunschot S, van Santvoort HC, Besselink MG, Bollen TL, Boermeester MA, et al. Early versus on-demand nasoenteric tube feeding in acute pancreatitis. N Engl J Med. 2014 Nov 20;371(21):1983-93. 19. Wittau M, Mayer B, Scheele J, Henne-Bruns D, Dellinger EP, Isenmann R. Systematic review and meta-analysis of antibiotic prophylaxis in severe acute pancreatitis. Scand J Gastroenterol 2011; 46(3):261-270. 20. Tse F, Yuan Y. Early routine endoscopic retrograde cholangiopancreatography strategy versus early conservative management strategy in acute gallstone pancreatitis. Cochrane Database Syst Rev 2012; 5:CD009779. 21. van Geenen EJ, van Santvoort HC, Besselink MG, van der Peet DL, van Erpecum KJ, Fockens P et al. Lack of Consensus on the Role of Endoscopic Retrograde Cholangiography in Acute Biliary Pancreatitis in Published Meta-Analyses and Guidelines: A Systematic Review. Pancreas 2013; 42(5):774-780. 22. Malmstrom ML, Hansen MB, Andersen AM, Ersboll AK, Nielsen OH, Jorgensen LN et al. Cytokines and organ failure in acute pancreatitis: inflammatory response in acute pancreatitis. Pancreas 2012; 41(2):271-277. 23. Schetz M. Non-renal indications for continuous renal replacement therapy. Kidney Int Suppl 1999;(72):S88-S94. 24. Yang C, Guanghua F, Wei Z, Zhong J, Penghui J, Xin F et al. Combination of hemofiltration and peritoneal dialysis in the treatment of severe acute pancreatitis. Pancreas 2010; 39(1):16-19. 25. Jiang HL, Xue WJ, Li DQ, Yin AP, Xin X, Li CM et al. Influence of continuous venovenous hemofiltration on the course of acute pancreatitis. World J Gastroenterol 2005; 11(31):4815-4821. 26. Inoue K, Hirota M, Beppu T, Ishiko T, Kimura Y, Maeda K et al. Angiographic features in acute pancreatitis: the severity of abdominal vessel ischemic change reflects the severity of acute pancreatitis. JOP 2003; 4(6):207-213.

278

27. Takeda K. Antiproteases in the treatment of acute necrotizing pancreatitis: continuous regional arterial infusion. JOP 2007; 8(4 Suppl):526-532. 28. Piascik M, Rydzewska G, Milewski J, Olszewski S, Furmanek M, Walecki J et al. The results of severe acute pancreatitis treatment with continuous regional arterial infusion of protease inhibitor and antibiotic: a randomized controlled study. Pancreas 2010; 39(6):863-867. 29. Takada T, Kawarada Y, Hirata K, Mayumi T, Yoshida M, Sekimoto M et al. JPN Guidelines for the management of acute pancreatitis: cutting-edge information. J Hepatobiliary Pancreat Surg 2006; 13(1):2-6. 30. Hamada T, Yasunaga H, Nakai Y, Isayama H, Horiguchi H, Matsuda S et al. Continuous regional arterial infusion for acute pancreatitis: a propensity score analysis using a nationwide administrative database. Crit Care 2013; 17(5):R214. 31. Mier J, Leon EL, Castillo A, Robledo F, Blanco R. Early versus late necrosectomy in severe necrotizing pancreatitis. Am J Surg 1997; 173(2):71-75. 32. Raraty MG, Halloran CM, Dodd S, Ghaneh P, Connor S, Evans J et al. Minimal access retroperitoneal pancreatic necrosectomy: improvement in morbidity and mortality with a less invasive approach. Ann Surg 2010; 251(5):787-793. 33. Rodriguez JR, Razo AO, Targarona J, Thayer SP, Rattner DW, Warshaw AL et al. Debridement and closed packing for sterile or infected necrotizing pancreatitis: insights into indications and outcomes in 167 patients. Ann Surg 2008; 247(2):294-299. 34. van Santvoort HC, Besselink MG, Bakker OJ, Hofker HS, Boermeester MA, Dejong CH et al. A step-up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med 2010; 362(16):1491-1502. 35. Bakker OJ, van Baal MC, van Santvoort HC, Besselink MG, Poley JW, Heisterkamp J et al. Endoscopic transpapillary stenting or conservative treatment for pancreatic fistulas


DISCUSSION AND FUTURE PERSPECTIVES

in necrotizing pancreatitis: multicenter series and literature review. Ann Surg 2011; 253(5):961-967. 36. Beger HG, Buchler M, Bittner R, Oettinger W, Block S, Nevalainen T. Necrosectomy and postoperative local lavage in patients with necrotizing pancreatitis: results of a prospective clinical trial. World J Surg 1988; 12(2):255-262.

43. Wittau M, Mayer B, Scheele J, Henne-Bruns D, Dellinger EP, Isenmann R. Systematic review and meta-analysis of antibiotic prophylaxis in severe acute pancreatitis. Scand J Gastroenterol 2011; 46(3):261-270. 44. Lauer MS, D'Agostino RB, Sr. The randomized registry trial--the next disruptive technology in clinical research? N Engl J Med 2013; 369(17):1579-1581.

37. Carter CR, McKay CJ, Imrie CW. Percutaneous necrosectomy and sinus tract endoscopy in the management of infected pancreatic necrosis: an initial experience. Ann Surg 2000; 232(2):175-180. 38. Horvath K, Freeny P, Escallon J, Heagerty P, Comstock B, Glickerman DJ et al. Safety and efficacy of video-assisted retroperitoneal debridement for infected pancreatic collections: a multicenter, prospective, single-arm phase 2 study. Arch Surg 2010; 145(9):817-825. 39. van Santvoort HC, Besselink MG, Horvath KD, Sinanan MN, Bollen TL, van RB et al. Videoscopic assisted retroperitoneal debridement in infected necrotizing pancreatitis. HPB (Oxford) 2007; 9(2):156159. 40. Bakker OJ, van Santvoort HC, van BS, Geskus RB, Besselink MG, Bollen TL et al. Endoscopic transgastric vs surgical necrosectomy for infected necrotizing pancreatitis: a randomized trial. JAMA 2012; 307(10):1053-1061. 41. Besselink MG, van Santvoort HC, Buskens E, Boermeester MA, van GH, Timmerman HM et al. Probiotic prophylaxis in predicted severe acute pancreatitis: a randomised, double-blind, placebo-controlled trial. Lancet 2008; 371(9613):651-659. 42. Johnson CD, Kingsnorth AN, Imrie CW, McMahon MJ, Neoptolemos JP, McKay C et al. Double blind, randomised, placebo controlled study of a platelet activating factor antagonist, lexipafant, in the treatment and prevention of organ failure in predicted severe acute pancreatitis. Gut 2001; 48(1):6269.

279


APPENDIX

Dutch Summary (Nederlandse Samenvatting) Review Committee Acknowledgements (Dankwoord) List of Publications Curriculum Vitae


APPENDIX

DU TC H SUM M A RY Ne d erl ands e s amenvatting Achterg rond Acute pancreatitis is een acute ontsteking van het pancreas veroorzaakt door verstopping van de ductus pancreaticus of door een direct toxisch effect op alvleeskliercellen. De mate van ontsteking is enorm divers en varieert van milde buikpijnklachten welke vlot herstellen tot escalerend orgaanfalen met praktisch onvermijdbaar overlijden.1,2 Acute pancreatitis is de meest voorkomende ziekte van het maag-darmstelsel waarvoor patiënten acuut worden opgenomen in het ziekenhuis.3 Echter, ieder jaar worden per ziekenhuis slechts enkele gevallen van acute pancreatitis, met complicaties zoals orgaanfalen of geïnfecteerde necrose, opgenomen.4 Als gevolg hiervan bestond klinisch onderzoek naar gecompliceerde acute pancreatitis voorheen vooral uit retrospectieve series van patiënten uit een enkel ziekenhuis in plaatse van robuuste gerandomiseerde multicentrische studies. Om deze redenen is de Pancreatitis Werkgroep Nederland in 2002 opgericht. De Werkgroep streeft naar meer inzicht in het ziektebeeld maar vooral naar een betere prognose van patiënten met gecompliceerde acute pancreatitis. Dit proefschrift bevat het resultaat van 7 jaar intensief klinisch onderzoek van onze Werkgroep. DEEL I VROEGE IDENTIFICATIE VAN GEC OMPLICEERDE PANCREATITIS Patiënten met acute pancreatitis bij wie een verhoogde kans bestaat op het ontwikkelen van complicaties moeten vroegtijdig worden geïdentificeerd om adequate behandeling en monitoring uit te voeren. Een internationale samenwerking tussen het Brigham and Women’s hospital (Boston, VS), University of Pittsburgh Medical Center (Pittsburgh, VS) en de Pancreatitis Werkgroep Nederland heeft geresulteerd in een prospectieve dataset van 1043 patiënten met acute pancreatitis (hoofdstuk 2). Gedetailleerde data van deze patiënten waren beschikbaar voor uitgebreide analyses van de geaggregeerde dataset. Hieruit kwam naar voren dat het risico op sterfte sterk geassocieerd was met een verhoogd serum ureum bij opname en een stijging van het serum ureum binnen 24 uur na opname. Een ureum van 20 mg/dL (7.14 mmol/L) of hoger was geassocieerd met een odds ratio (OR) van 4.6 (95 procent betrouwbaarheidsinterval 2.5 – 8.3) voor sterfte. Een gestegen ureum 24 uur na opname was geassocieerd met een OR van 4.3 (95 procent betrouwbaarheidsinterval 2.3 – 7.9) voor sterfte. Een algoritme gebaseerd op de hoogte van het ureum identificeert patiënten identificeren bij wie een verhoogde kans op sterfte bestaat en kan artsen helpen bij het instellen van vroegtijdige behandelingen. Acute pancreatitis is een complexe aandoening waarbij het van belang is om correcte terminologie en heldere definities te hanteren. Dit is met name van belang voor de 282


DUTCH SUMMARY

communicatie tussen artsen en voor het rapporteren van de uitkomsten van klinisch onderzoek (hoofdstuk 3). Om het gebruik van correcte terminologie te stimuleren is een revisie van de in 1992 opgestelde Atlanta classificatie ontwikkeld.5 In deze nieuwe classificatie worden patiënten met necrotiserende pancreatitis onderverdeeld in patiënten met necrose van het pancreasparenchym met of zonder necrose van het extrapancreatische vetweefsel, of patiënten met alleen necrose van het extrapancreatische vetweefsel zonder necrose van het pancreasparenchym (EXPN). Men veronderstelt dat patiënten met EXPN een betere prognose hebben maar robuuste data ontbreken om dit te onderbouwen. Uit een prospectieve multicentrische serie van patiënten met necrotiserende pancreatitis, zijn 315 patiënten met EXPN vergeleken met 324 patiënten met alleen necrose van het pancreasparenchym. Patiënten met EXPN ontwikkelden minder complicaties waaronder een lagere sterfte (9 versus 20 procent, p<0.001). Echter, wanneer infectie van extrapancreatische necrose aanwezig was, waren de uitkomsten tussen beide groepen gelijk. Concluderend zijn voor de gehele groep patiënten met EXPN de uitkomsten beter vergeleken met patiënten met pancreasnecrose. Om deze reden moet EXPN als een aparte entiteit worden gezien binnen het spectrum van necrotiserende pancreatitis. Toekomstige studies zouden het aantal patiënten met pancreasnecrose en het aantal patiënten met alleen EXPN moeten rapporteren. Bij patiënten met gecompliceerde pancreatitis varieert de uitgebreidheid en de duur van orgaanfalen aanzienlijk. Gedetailleerde data over de timing, type, duur en combinatie (respiratoir, cardiovasculair en renaal falen) van orgaanfalen zijn niet beschikbaar in de huidige literatuur. Deze data zijn hard nodig om de prognose van patiënten met acute pancreatitis en orgaanfalen te kunnen inschatten. Het resultaat van een gedetailleerde analyse van 240 patiënten met necrotiserende pancreatitis en orgaanfalen wordt gepresenteerd in hoofdstuk 4. Samengevat, de sterfte ten gevolge van orgaanfalen is sterk afhankelijk van de timing van multi-orgaanfalen (start in de eerste week van opname, 41 procent sterfte), duur van multi-orgaanfalen (korter dan 1 week, 43 procent sterfte en langer dan 3 weken 52 procent) en de combinatie (falen van alle 3 de orgaansystemen, 63 procent sterfte). In tegenstelling tot eerdere studies, hebben wij geen verschil gevonden in sterfte (na 10 dagen opname) tussen patiënten met orgaanfalen en met geïnfecteerde pancreasnecrose, en patiënten met orgaanfalen maar zonder infectie van pancreasnecrose (28 versus 30 procent sterfte, relatief risico 1.3, p=0.4). Uiteindelijk overleefde de helft van alle patiënten na een episode van gecompliceerde pancreatitis met enkele weken van orgaanfalen. Deze bevindingen kunnen de klinische beslisvorming omtrent patiënten met orgaanfalen en acute pancreatitis sturen. Zelfs eventuele besluiten over het wel of niet continueren van de behandeling kunnen worden gebaseerd op de gedetailleerde data uit hoofdstuk 4.

283


APPENDIX

DE E L I I HET VO ORKOMEN VAN GEC OMPLICEERDE PANCREATITIS In het tweede deel van dit proefschrift wordt het onderzoek gepresenteerd dat bedoelt is om complicaties bij patiënten met acute pancreatitis te voorkomen. Belangrijke infecties (infectie van pancreasnecrose, pneumonie, bacteriemie) beïnvloeden de prognose van patiënten met acute pancreatitis.6,7 In het algemeen wordt verondersteld dat deze infecties worden veroorzaakt door translocatie van bacteriën vanuit de darm, wat weer wordt veroorzaakt door verstoorde darmmotiliteit in combinatie met bacteriële overgroei in de darm en een toename van mucosale permeabiliteit.7-12 De laatste inzichten zijn dat enterale sondevoeding de intestinale motiliteit stimuleert - en daarmee bacteriële overgroei tegengaat – en de bloedtoevoer naar de darmen stimuleert – en daarmee de integriteit van de darmwand in stand houdt.13-20 Wij verrichtten een meta-analyse van 165 individuen uit 8 gerandomiseerde trials met gebruik van multivariabele logistische regressie, waarbij is gecorrigeerd voor voorspelde ernst van acute pancreatitis en voor de oorspronkelijk trial herkomst. Het effect van de start van enterale sondevoeding op een samengesteld eindpunt van infectie van pancreasnecrose, orgaanfalen en sterfte werd onderzocht (hoofdstuk 5). In het multivariabele model, reduceert enterale voeding gestart binnen 24 uur na opname vergeleken met een start na 24 uur het samengestelde eindpunt van 45 naar 19 procent (gecorrigeerde OR 0.44; 95 procent betrouwbaarheidsinterval 0.20 - 0.96). Orgaanfalen werd gereduceerd van 42 naar 16 procent (gecorrigeerde OR 0.42; 95 procent betrouwbaarheidsinterval 0.19 - 0.94). Concluderend, deze observationele data van individuen met acute pancreatitis tonen aan dat vroege enterale sondevoeding geassocieerd is met een sterke afname van het aantal complicaties. De bovenstaande meta-analyse evenals als eerdere studies hebben enkele sterke beperkingen. Om deze reden hebben wij een multicentrische gerandomiseerde trial uitgevoerd waarin vroege enterale sondevoeding is vergeleken met een normaal dieet gestart na 72 uur na presentatie op de Eerste Hulp in patiënten met acute pancreatitis en een verhoogde kans op complicaties (PYTHON trial, hoofdstuk 6). Patiënten werden gerandomiseerd tussen sondevoeding binnen 24 uur na randomisatie (vroege groep) of een normaal dieet na 72 uur na presentatie (on-demand groep) met sondevoeding alleen wanneer een normaal dieet niet werd getolereerd. Het primaire eindpunt was een samengesteld eindpunt van infecties (geïnfecteerde pancreasnecrose, pneumonie of bacteriemie) of sterfte binnen 6 maanden na randomisatie. In totaal werden 208 patiënten gerandomiseerd waarvan er 205 werden geïncludeerd in de analyse. Het primaire eindpunt trad op bij 30 van de 101 patiënten (30 procent) in de vroege groep en bij 28 van de 104 patiënten (27 procent) in de on-demand groep (risico ratio, 1.07; 95 procent betrouwbaarheidsinterval, 0.79 - 1.44, P = 0.76). Evenmin waren er verschillen tussen de groepen in infecties (25 en 26 procent, respectievelijk; p=0.87) of sterfte (11 en 7 284


DUTCH SUMMARY

procent, p=0.33). In de on-demand groep, tolereerden 72 patiënten (69 procent) een normaal dieet en was er dus geen reden voor sondevoeding. Concluderend, de PYTHON trial heeft niet aangetoond dat vroege sondevoeding superieur is in het reduceren van het aantal infecties of sterfte vergeleken met een normaal dieet gestart na 72 uur na presentatie bij patiënten met acute pancreatitis en een verhoogde kans op complicaties. Het gebruik van de on-demand strategie in plaats van routinematig gebruik van sondevoeding zorgt voor een significante reductie in patiënt ongemak en gezondheidszorgkosten. In hoofdstuk 7 hebben we de timing van cholecystectomie bij patiënten met een biliaire pancreatitis bestudeerd. Patiënten met een biliaire pancreatitis hebben kans op een 2de aanval zo lang de galblaas niet is verwijderd. Het doel van deze studie was om het aantal galsteen gerelateerde klachten te evalueren als gevolg van de wachttijd tot cholecystectomie. Bij 188 patiënten (76 procent) met een biliaire pancreatitis werd na mediaan 6 weken een cholecystectomie verricht. Voorafgaand aan cholecystectomie werden 34 patiënten heropgenomen voor galsteen gerelateerde klachten inclusief 24 gevallen van recidief acute pancreatitis. Acht (7 procent) patiënten van de 108 patiënten ontwikkelden galsteen gerelateerde klachten na een endoscopische papillotomie tijdens de eerste opname maar voor de cholecystectomie vergeleken met 26 patiënten (18 procent) van de 141 patiënten bij wie geen papillotomie werd verricht tijdens de eerste opname (risico ratio 0.51, 95 procent betrouwbaarheidsinterval 0.27 - 0.94; p=0.02). Slechts 142 (53 procent) van de 267 patiënten werden behandeld in overeenkomst met de Nederlandse richtlijn. Gebaseerd op de resultaten concludeerden wij dat een vertraging in cholecystectomie na milde biliaire pancreatitis een substantieel risico heeft op het krijgen van recidief galsteen gerelateerde klachten. Papillotomie verlaagt de kans op recidief pancreatitis maar niet op andere galsteen gerelateerde klachten. DEEL III BEHANDELING VAN GEC OMPLICEERDE PANCREATITIS In het derde deel van dit proefschrift presenteren we het onderzoek dat bedoelt is om de uitkomst van patiënten met gecompliceerde pancreatitis, zoals patiënten met geïnfecteerde pancreasnecrose, te verbeteren. In de PANTER trial, is de traditionele benadering van geïnfecteerde pancreasnecrose, open necrosectomie, vergeleken met de minimaal invasieve step-up benadering (hoofdstuk 8). De step-up benadering bestond uit percutane catheter drainage, indien nodig, gevolgd door minimaal invasieve retroperitoneale necrosectomie. In de multicentrische studie, werden 88 patiënten gerandomiseerd. Het primaire eindpunt trad op in 31 van de 45 patiënten (69 procent) in de open necrosectomie groep en in 17 van de 43 patiënten (40 procent) in de step-up groep (risico ratio met the step-up benadering, 0.57; 95 procent betrouwbaarheidsinterval, 0.38 - 0.87; p=0.006). Van de patiënten in de step-up groep werd 35 procent behandeld met alleen een percutane 285


APPENDIX

drain. Concluderend, een minimaal invasieve step-up benadering, in vergelijking met een open necrosectomie, reduceert het samengesteld eindpunt van complicaties en sterfte bij patiënten met infectie van pancreasnecrose. Om te onderzoek of een nog minder invasieve benadering voor een nog betere uitkomst zorgt, hebben we de PENGUIN trial uitgevoerd (hoofdstuk 9). In deze pilot trial hebben we onderzocht of endoscopische transgastrische necrosectomie, een vorm van endoscopische chirurgie door natuurlijke lichaamsopeningen, de postprocedurele proinflammatoire reactie kan verlagen. De proinflammatoire reactie is gemeten met serum Interleukine-6 (IL-6) bepalingen na interventie voor geïnfecteerde necrose. Endoscopische necrosectomie bestond uit transgastrische punctie, dilatatie door een ballon, retroperitoneale debridement en necrosectomie. Chirurgische necrosectomie bestond uit video-geassisteerde retroperitoneale necrosectomie of laparotomie. Endoscopische necrosectomie zorgde voor een verlaging van IL-6 in vergelijking met chirurgische necrosectomie (P=0.004). Het samengestelde secundaire klinische eindpunt trad minder vaak op in de endoscopische chirurgie groep (20 versus 80 procent; risico verschil, 0.60; 95 betrouwbaarheidsinterval, 0.16 - 0.80; p=0.03). Concluderend, bij patiënten met geïnfecteerde pancreasnecrose zorgt een endoscopische necrosectomie voor een verlaging van de postprocedurele proinflammatoire reaktie en voor het samengestelde klinische eindpunt in vergelijking met chirurgische necrosectomie. Bij patiënten met een pancreasfistel na een necrosectomie vanwege geïnfecteerde pancreasnecrose, kan een endoscopische transpapillaire stent zorgen voor een kortere tijd tot genezing van de pancreasfistel. In hoofdstuk 10 hebben we de veiligheid, haalbaarheid en uitkomst van endoscopische stentplaatsing bestudeerd in een prospectieve serie van patiënten en in de literatuur. Onze studie heeft aangetoond dat na stentplaatsing de tijd tot genezing van de pancreasfistel wordt verkort. De review van de literatuur toonde aan dat het percentage van het succesvol genezen van de pancreasfistel na stentplaatsing alleszins redelijk is (71 procent). Stent-gerelateerde complicaties werden gezien in ongeveer 10 procent van de patiënten in de literatuur. Uit deze resultaten concluderen wij dat endoscopische transpapillaire stentplaatsing een veilig en haalbaar alternatief is voor conservatieve therapie bij patiënten met een pancreasfistel na een necrosectomie vanwege geïnfecteerde pancreasnecrose. In hoofdstuk 11 rapporteren we de uitkomst van 639 opeenvolgende patiënten met necrotiserende pancreatitis behandeld in 21 Nederlandse ziekenhuizen. Dit is de grootste opeenvolgende serie van patiënten met necrotiserende pancreatitis. De belangrijkste bevinding was dat in de gehele groep de sterfte 15 procent was. Daarnaast dat 38 procent van alle patiënten een interventie behoefte en dat in deze groep de sterfte 27 procent was. Tevens vonden wij dat hoe langer de tijd tussen opname en interventie, hoe lager de sterfte. Tot slot, primaire catheter drainage (percutaan of endoscopisch) was geassocieerd met minder complicaties dan primaire necrosectomie. Concluderend, de meerderheid 286


DUTCH SUMMARY

van de patiënten met necrotiserende pancreatitis kan conservatief worden behandeld met een lage sterfte. Bij patiënten met geïnfecteerde pancreasnecrose, uitgestelde interventie en catheter drainage als primaire interventie zorgt voor een betere uitkomst. Het overzichtsartikel in hoofdstuk 12 beschrijft de huidige standaard behandeling van necrotiserende pancreatitis, zowel conservatief als invasief. Een groot deel van de veranderingen in behandeling van necrotiserende pancreatitis zijn gebaseerd op bevindingen uit eerdere hoofdstukken van dit proefschrift.

Onderzoeksvragen en belangrijkste bevindingen Hoofdstuk 2.

Wat is de accuratesse van serum ureum als vroege voorspeller van gecompliceerde pancreatitis? Ureum is een eenvoudige, goedkope en accurate vroege voorspeller van mortaliteit in acute pancreatitis en we hebben een algoritme opgesteld dat artsen kan helpen bij het tijdig starten van behandeling.

3.

Wat is het verschil in uitkomst tussen patiënten met pancreasnecrose en patiënten met alleen necrose van het extrapancreatische vetweefsel (EXPN)? EXPN veroorzaakt minder complicaties dan pancreasnecrose. Daarom moet EXPN worden beschouwd als een aparte entiteit binnen met spectrum van necrotiserende pancreatitis. Echter, de uitkomst in het geval van infectie van EXPN is gelijk aan infectie van pancreasnecrose.

4.

Wat is de invloed van orgaanfalen op sterfte in necrotiserende pancreatitis? Alhoewel de sterfte erg hoog is bij patiënten met necrotiserende pancreatitis en orgaanfalen, overleeft 50 procent van de patiënten, ook na weken van orgaanfalen.

5.

Gebaseerd op individuele patiënten data uit gerandomiseerde trials, zorgt vroege enterale sondevoeding voor een betere uitkomst in patiënten met acute pancreatitis? In deze observationele meta-analyse van individuele data van patiënten met acute pancreatitis, was het starten van sondevoeding binnen 24 uur na ziekenhuisopname, vergeleken met een start na 24 uur na opname, geassocieerd met een reductie in het aantal complicaties en een verbeterde klinische uitkomst.

6.

Zorgt vroege enterale sondevoeding, vergeleken met een normaal dieet na 72 uur na presentatie, voor een afname van het aantal infecties of mortaliteit bij patiënten met acute pancreatitis? Bij patiënten met acute pancreatitis en een hoge kans op complicaties, heeft de PYTHON trial de superioriteit van vroege sondevoeding vergeleken met een normaal dieet 72 uur na presentatie, niet kunnen aantonen voor wat betreft het reduceren van infecties of sterfte.

7.

Wat is de optimale timing van cholecystectomie na een aanval van milde biliaire pancreatitis? Een vertraging in de tijd tot cholecystectomie na milde biliaire pancreatitis zorgt voor een substantieel risico op recidief galsteen gerelateerde klachten. Cholecystectomie zou moeten worden verricht tijdens de initiële opname voor milde biliaire pancreatitis.

287


APPENDIX

8.

Zorgt een ‘step-up benadering’, met het gebruik van percutane catheter drainage en minimaal invasieve necrosectomie, voor een afname van het aantal complicaties of sterfte vergeleken met primaire open necrosectomie bij patiënten met infectie van pancreasnecrose? Een minimaal invasieve step-up benadering vergeleken met primaire open necrosectomie reduceert de combinatie van complicaties of sterfte, evenals lange termijn complicaties en gezondheidszorgkosten.

9.

Zorgt een endoscopische transgastrische necrosectomie vergeleken met een chirurgische necrosectomie voor een afname van de proinflammatoire response en een verbetering van klinische uitkomst bij patiënten met necrotiserende pancreatitis? In patiënten met geïnfecteerde pancreasnecrose zorgt endoscopische necrosectomie voor een reductie van de proinflammatoire response en voor een reductie van complicaties vergeleken met chirurgische necrosectomie.

10.

In patiënten met een pancreasfistel als gevolg van een necrosectomie van pancreasnecrose, zorgt een transpapillaire stent vergeleken met een conservatieve benadering voor een kortere tijd tot genezing van de pancreasfistel? Endoscopische transpapillaire stentplaatsing is een haalbaar en veilig alternatief naast conservatieve therapie en kan mogelijk de tijd tot fistelgenezing verkorten.

11.

Wat is de uitkomst van een conservatieve en minimaal invasieve benadering bij patiënten met necrotiserende pancreatitis? De meerderheid van de patiënten met necrotiserende pancreatitis kan worden behandeld zonder interventie en met een lage sterfte. Bij patiënten met geïnfecteerde pancreasnecrose zorgt een uitgestelde interventie en primaire catheter drainage voor een verbetering van de uitkomst.

12.

Wat is de huidige state-of-the-art benadering van patiënten met necrotiserende pancreatitis? In necrotiserende pancreatitis is interventie verschoven vroege primaire open necrosectomie naar een step-up benadering met catheter drainage en zo nodig gevolgd door een minimaal invasieve chirurgische of endoscopische necrosectomie zodra peripancreatische collecties zich hebben afgekapseld.

GEFASEERDE STEP-UP BEHANDELING In necrotiserende pancreatitis, is het type van complicatie dat ontstaat nauw gerelateerd aan de duur van de ziekte. Daarom worden specifieke complicaties verschillend behandeld op verschillende tijdsmomenten. In deze paragraaf wordt de behandeling van necrotiserende pancreatitis samengevat en een behandel algoritme gegeven. Beiden zijn gebaseerd op de tijd sinds het ontstaan van symptomen (figuur 1). In de eerste 24 tot 72 uur na ontstaan van de ziekte is frequente controle van de conditie van de patiënt cruciaal. Liberale vocht suppletie moet worden gestuurd op basis van laboratorium uitslagen zoals serum ureum en kreatinine. Sondevoeding is geïndiceerd indien na 96 uur een normaal dieet niet wordt verdragen. Bij een verdenking op cholangitis is een spoed endoscopische retrograde cholangiopancreaticografie 288


DUTCH SUMMARY

geïndiceerd, maar of dit nodig is in patiënten met een ernstige biliaire pancreatitis is onbekend. Antibiotica profylaxe kan infectie van pancreasnecrose niet voorkomen en antibiotica maken geen onderdeel uit van de initiële behandeling. Infecties bevestigd met een bacteriologisch kweek moeten worden bestreden met gerichte antibiotica. Vanwege de conservatieve strategie bij necrotiserende pancreatitis, zal een fijne naald aspiratie minder vaak lijden tot een verandering in het beleid en is daarom zelden geïndiceerd. Optimale behandeling van geïnfecteerde pancreasnecrose bestaat uit een gefaseerde multidisciplinaire ‘step-up’ strategie. De eerste stap is drainage, percutaan of transgastrisch, gevolgd door chirurgische of endoscopische transluminale necrosectomie indien nodig. Necrosectomie wordt uitgesteld totdat de necrotische collecties zijn afgekapseld.

Figuur 1. Suggestie voor een behandelalgoritme gebaseerd op de duur sinds ontstaan van ziekte. CECT, contrastenhanced computed tomografie; SIRS, systemische inflammatoire respons syndroom; ERCP, endoscopische retrograde cholangiopancreatografie; MRI, magnetic resonance imaging; FNA, fijne naald aspiratie.

289


APPENDIX

REFERENCES 1. Whitcomb DC. Clinical practice. Acute pancreatitis. N Engl J Med 2006;354(20):214250. 2. Lankisch PG, Apte M, Banks PA. Acute pancreatitis. Lancet. 2015 Jan 20. pii: S01406736(14)60649-8. 3. Peery AF, Dellon ES, Lund J, Crockett SD, McGowan CE, Bulsiewicz WJ et al. Burden of gastrointestinal disease in the United States: 2012 update. Gastroenterology 2012; 143(5):1179-1187. 4. Spanier B, Bruno MJ, Dijkgraaf MG. Incidence and mortality of acute and chronic pancreatitis in the Netherlands: a nationwide record-linked cohort study for the years 1995-2005. World J Gastroenterol 2013;19(20):3018-26. 5. Banks PA, Bollen TL, Dervenis C, Gooszen HG, Johnson CD, Sarr MG et al. Classification of acute pancreatitis--2012: revision of the Atlanta classification and definitions by international consensus. Gut 2013; 62(1):102-111. 6. Besselink MG, van Santvoort HC, Boermeester MA, Nieuwenhuijs VB, van Goor H, Dejong CH et al. Timing and impact of infections in acute pancreatitis. Br J Surg 2009;96(3):267-73. 7. Wu BU, Johannes RS, Kurtz S, Banks PA. The impact of hospital-acquired infection on outcome in acute pancreatitis. Gastroenterology 2008;135(3):816-20. 8. Ammori BJ, Leeder PC, King RF, Barclay GR, Martin IG, Larvin M et al. Early increase in intestinal permeability in patients with severe acute pancreatitis: correlation with endotoxemia, organ failure, and mortality. J Gastrointest Surg 1999;3(3):252-62. 9. Besselink MG, van Santvoort HC, Renooij W, de Smet MB, Boermeester MA, Fischer K et al. Intestinal barrier dysfunction in a randomized trial of a specific probiotic composition in acute pancreatitis. Ann Surg 2009;250(5):712-9.

290

10. Fritz S, Hackert T, Hartwig W, Rossmanith F, Strobel O, Schneider L et al. Bacterial translocation and infected pancreatic necrosis in acute necrotizing pancreatitis derives from small bowel rather than from colon. Am J Surg 2010;200(1):111-7. 11. Nieuwenhuijs VB, Verheem A, van Duijvenbode-Beumer H, Visser MR, Verhoef J, Gooszen HG et al. The role of interdigestive small bowel motility in the regulation of gut microflora, bacterial overgrowth, and bacterial translocation in rats. Ann Surg 1998;228(2):188-93. 12. Rahman SH, Ammori BJ, Holmfield J, Larvin M, McMahon MJ. Intestinal hypoperfusion contributes to gut barrier failure in severe acute pancreatitis. J Gastrointest Surg 2003;7(1):26-35; discussion -6. 13. Van Felius ID, Akkermans LM, Bosscha K, Verheem A, Harmsen W, Visser MR et al. Interdigestive small bowel motility and duodenal bacterial overgrowth in experimental acute pancreatitis. Neurogastroenterol Motil 2003;15(3):267-76. 14. Marik PE. What is the best way to feed patients with pancreatitis? Curr Opin Crit Care 2009;15(2):131-8. 15. McClave SA, Heyland DK. The physiologic response and associated clinical benefits from provision of early enteral nutrition. Nutr Clin Pract 2009;24(3):305-15. 16. Ziegler TR. Parenteral nutrition in the critically ill patient. N Engl J Med. 2009 Sep 10;361(11):1088-97. 17. Marik PE, Zaloga GP. Early enteral nutrition in acutely ill patients: a systematic review. Crit Care Med 2001;29(12):2264-70. 18. Li JY, Yu T, Chen GC, Yuan YH, Zhong W, Zhao LN et al. Enteral nutrition within 48 hours of admission improves clinical outcomes of acute pancreatitis by reducing complications: a meta-analysis. PLoS One 2013;8(6):e64926. 19. Petrov MS, Pylypchuk RD, Uchugina AF. A systematic review on the timing of artificial nutrition in acute pancreatitis. Br J Nutr 2009;101(6):787-93.


DUTCH SUMMARY

20. Sun JK, Mu XW, Li WQ, Tong ZH, Li J, Zheng SY. Effects of early enteral nutrition on immune function of severe acute pancreatitis patients. World J Gastroenterol 2013;19(6):917-22. 21. Wereszczynska-Siemiatkowska U, SwidnickaSiergiejko A, Siemiatkowski A, Dabrowski A. Early enteral nutrition is superior to delayed enteral nutrition for the prevention of infected necrosis and mortality in acute pancreatitis. Pancreas 2013;42(4):640-6.

291


APPENDIX

R E V I E W C OM M I T T E E Prof.dr. Y. van der Graaf Professor of Clinical Epidemiology Julius Center for Health Sciences and Primary Care University Medical Center Utrecht, The Netherlands Prof.dr. M.J.M. Bonten Professor of Molecular Epidemiology of Infectious Diseases Department of Medical Microbiology and Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, The Netherlands Prof.dr. P.D. Siersema Professor of Gastroenterology Department of Gastroenterology and Hepatology University Medical Center Utrecht, The Netherlands Prof.dr. M.A. Boermeester Professor of Surgery of Abdominal Infections Department of Surgery Amsterdam Medical Center, Amsterdam, The Netherlands Prof.dr. J.G. van der Hoeven Professor of Intensive Care Medicine Department of Intensive Care Medicine University Medical Centre St. Radboud, Nijmegen, The Netherlands

292


ACKNOWLEDGEMENTS

AC K N OW L E D G E M E N T S D an kwo ord Veel mensen hebben bijgedragen aan het onderzoek in dit proefschrift. Graag wil ik iedereen ontzettend bedanken. Het is een intensieve maar geweldig leerzame periode geweest. Het was vooral vele uren ploeteren maar de ultieme beloning doet alles vergeten. Klinisch onderzoek is een echte hobby geworden waar ik in de toekomst nog veel plezier aan hoop te beleven. Allereerst wil ik alle patiënten en hun familieleden bedanken voor deelname aan de verschillende trials. Het is ongelofelijk dapper om als een donder bij heldere hemel getroffen te worden door een onbekende scheurende buikpijn en dan ook nog toestemming te geven voor deelname aan een studie waar pas toekomstige patiënten misschien baat bij hebben en waarin ‘geloot’ gaat worden voor wat de behandeling precies gaat inhouden. Zonder uw moed geen vooruitgang in behandeling! Prof. Dr. H.G. Gooszen, beste Hein, u bent een unieke promotor, originele denker, een man van grote trials maar ook een clinicus pur sang. Ik heb u leren kennen tijdens een tumultueuze periode in uw carrière. Juist in deze periode heeft u er nooit over getwijfeld het onderzoek de rug toe te keren. Onder uw voorzitterschap heeft Pancreatitis Werkgroep deze periode doorstaan. Een prestatie van formaat. U geeft uw promovendi volledig de ruimte. Dit is uitzonderlijk in de academische wereld en heeft mijn eigen promotietraject vele malen leerzamer gemaakt. En zelfs nu in de laatste periode voor mijn promotie heb ik het gevoel dat ik u nog beter leer kennen. Bedankt voor alles. Prof. Dr. F.L. Moll, beste Frans, bij u promoveren is niet uniek, de promovendi staan zo ongeveer in de rij op het Domplein. Maar dat ik bij u mag promoveren vind ik wel heel bijzonder. U bent de pater familias van één van de mooiste vaatklinieken in Nederland, wereldberoemd vaatchirurg en alom gerespecteerd onderzoeker. Mijn overgang naar de vaatchirurgie voelt op deze manier helemaal goed. Ontzettend bedankt en ik kijk ernaar uit om nog jarenlang samen onderzoek doen. Geachte leden van de beoordelingscommissie, bedankt voor de interesse in mijn proefschrift. Dr. van Santvoort, Hjalmar! De energie die jij en Marc uitstraalden tijdens mijn eerste maanden in de pancreaskamer was overweldigend. Jullie (en al snel ‘wij’) waren niet te stoppen. Je hebt me in een rap tempo de basis van acute pancreatitis en het doen van 293


APPENDIX

wetenschappelijk onderzoek bijgebracht. In Nederland en daarbuiten leerde ik alle stakeholders kennen. Ik moest jouw PANTER op de rails houden en dat kon maar op een manier, 100% overgave. Het is toch de normaalste zaak van de wereld om op vrijdag na middernacht het datacentrum uit te wandelen. Of een expertmail versturen op oudejaarsavond. Maar de ervaring die ik tijdens PANTER heb opgedaan heb ik later tijdens PYTHON driedubbel en dwars nodig gehad! Uren kunnen we over onderzoek en chirurgie praten en ik vind het heerlijk. Hopelijk blijven we dat ook vooral doen ondanks alle drukte. Binnenkort gaat er voor jou ook weer een spannende nieuwe periode aanbreken. Ik wens je daarin en in de verdere toekomst het allerbeste. De chirurgische wereld gaat nog veel van je horen, vriend! Dr. Besselink, beste Marc, je hebt ongeëvenaarde kwaliteiten. Jouw organisatietalent, communicatieve vaardigheden en tomeloze energie krijgen meer voor mekaar in een paar jaar dan hele vakgroepen in een decennium. Tijdens jouw promotieperiode heb je de basis gelegd van de Pancreatitis Werkgroep en dus ook van dit proefschrift. Inmiddels ben je volwaardig HPB-chirurg, hard op weg naar een leerstoel en heb je zelfs de troepen bij elkaar gekregen om gezamenlijk pancreaskanker aan te vallen. Ik ben benieuwd wat je allemaal gaat ondernemen in de toekomst. Bedankt! Drs. van Brunschot, beste Sandra, de eerste vrouw met de ‘pancreatofoon’. Je kwam als geroepen na het vertrek van mijn geplande opvolger. Hierdoor werd jouw inwerkperiode wel heel kort maar je wist je razendsnel aan te passen. Wat heb ik het getroffen met zo’n opvolgster! Ondanks het kleine oponthoud gleed PYTHON dankzij jouw inspanningen heel soepeltjes over de eindstreep. En PENGUIN er achteraan. Daarnaast zette je jouw TENSION op de rit, kwam je in opleiding tot chirurg en ben je recent mama geworden. Je bent een ijzersterke vrouw en jij krijgt straks een schitterend proefschrift. Drs. Bollen, beste Thomas, je bent de radioloog die iedere pancreatoloog hoopt te kunnen bellen. Ik heb genoten van de telefoontjes, dag en nacht, wanneer we acuut beslissingen moesten nemen over pancreatitis patiënten. Binnen seconden had je de scans doorgrond. ‘Kan passen bij’ of ‘kan niet uitgesloten worden’ past niet in jouw radiologisch woordenboek! Je bent enorm gedreven, een creatieve geest en lieve pappa. Jouw artikelen verdienen een nietje en ik kijk ernaar uit. Dr. Ahmed Ali, beste Usama, ik geniet van jouw enthousiasme en aanstekelijke lach. Het chronische pancreatitis onderzoek staat dankzij jou op de kaart. No Access, statistiek, strijkijzertjes in Wenen, computer crashes of een roadtrip van Pittsburgh naar Columbus, bedankt voor alle overlegmomenten en de ondersteuning op alle fronten! Succes met jouw promotie en de opleiding in de mooiste regio. 294


ACKNOWLEDGEMENTS

Drs. Schepers, beste Nicolien, je bent een van de nieuwe sterren van het datacentrum. Dankzij jouw gedrevenheid, nuchterheid en aangename vrolijke persoonlijkheid dendert jouw APEC trein op volle vaart naar het eindstation. Ik vind het hartstikke leuk om je hierbij te helpen. Wanneer wordt er een APEC overleg in de sneeuw gepland? Ik neem mijn ABS mee! Mevrouw Zeguers, beste Vera, Ik ben je enorm dankbaar voor al het werk dat je in ‘onze’ PYTHON hebt gestoken. Je bent een van de fundamenten van de Werkgroep, weergaloos aan de telefoon met patiënten, artsen en verpleegkundigen, overleefde de storm en verhuisde en passant mee naar Nijmegen. Zonder jouw hulp had dit proefschrift er anders uit gezien. Super bedankt! Mevrouw Roeterdink, beste Anneke, wat heb ik bewondering voor jouw rol in het datacentrum. Met tomeloze inzet, flexibiliteit en discipline houd je alle ballen in de lucht. Tussen alle bedrijven door haalde je ook nog de berg aan PENGUIN data binnen uit alle centra. Ontzettend bedankt. Onderzoekers van de PWN, beste Mark, Stefan, Yama, Janneke, David, Bob en Xavier. Bedankt voor jullie hulp en heel veel succes met jullie eigen projecten. Keep it up! Drs. Lodewijkx, beste Piet, dank voor je hulp bij de economische evaluatie van PYTHON en veel succes met je verdere carrière. Prof. Dr. M.J. Bruno, beste Marco, met verve trad je in de voetsporen van professor Gooszen als nieuwe voorzitter van de Werkgroep. Ondanks je superdrukke agenda weet je altijd tijd vrij te maken voor acute pancreatitis onderzoek. Dat wordt zeer gewaardeerd! Ontzettend bedankt voor je adviezen en waardevolle suggesties bij de manuscripten. Ik hoop in de toekomst nog veel onderzoek met je te doen. Leden van de Pancreatitis Werkgroep Nederland, bedankt voor uw bijdrage aan de unieke projecten van onze Werkgroep. In het bijzonder wil ik Kees Dejong, Casper van Eijck, Marja Boermeester, Robin Timmer, Harry van Goor, Bert van Ramshorst, Vincent Nieuwenhuis, Ben Witteman, Sandro Schaapherder, Menno Brink, Jan-Werner Poley, Paul Fockens, Sijbrand Hofker, Joos Heisterkamp, Thijs Schwartz, Marcel Dijkgraaf en professor Akkermans bedanken voor de nimmer aflatende steun aan de Werkgroep. Leden van de Data & Safety Monitoring Committee van de PYTHON trial, beste Dink Legemate, Peter Go, Peter Siersema, Lisbeth Mathus-Vliegen en voorzitter Erik Buskens. Het is een intensieve maar zeer leerzame reis geweest. Ik ben u zeer dankbaar voor uw inzet en moedige besluitvorming. Een commissie om met trots op terug te kijken. 295


APPENDIX

Docters Wu and Papachristou, dear Bechien and George, new school pancreatologists. I really enjoyed doing research with you guys. I learned a lot from your original and intelligent projects. Hopefully we can do some more collaborative work in the future! Beste Koop Bosscha en de andere chirurgen in het Jeroen Bosch ziekenhuis, bij de start van mijn opleiding had ik alleen het artikel uit de inleiding van dit proefschrift gepubliceerd en was mijn beoogde opvolgster net gestopt……. Ik ben jullie heel veel dank verschuldigd voor het geduld, het vertrouwen en de steun om zowel opleiding als promotie tot een goed einde te mogen brengen. Vaatchirurgen in het Jeroen Bosch ziekenhuis, beste Olivier Koning, Ronald van den Haak, Bart Verhoeven, Jan-Willem Hinnen en natuurlijk Ruud van Loenhout. Jullie hebben me de weg gewezen in de wereld van de vaatchirurgie en hiervoor ben ik jullie zeer dankbaar. Ik heb geen moment spijt gehad van mijn keus en ik heb genoten van de dagen op OK 5. Jullie hebben een prachtige vaatchirurgische kliniek! Beste professor Menno Vriens, chirurgen en vaatchirurgen uit het UMC Utrecht, bedankt voor de fijne tijd, de mooie start in de vaatchirurgie en de ruimte die ik heb gekregen om de opleiding een aantal maanden te kunnen onderbreken om PYTHON af te ronden. Arts-assistenten in het Jeroen Bosch ziekenhuis en het UMC Utrecht, dank voor de leuke opleidingstijd en prettige samenwerking. Beste Maarten en Guus, huisgenoten, jaargenoten en vrinden, op vele momenten heeft dit proefschrift mij er niet gezelliger opgemaakt . Ik dank jullie voor de steun en vriendschap! Boys van de Jet Set, Michael, Temmie, Neel, Roelof, Erik, Roderik, Tjarda, Victor en John. Zo nu en dan net zoals toen, toppers! Mannen van jaarclub Temmink, de eerste en misschien wel laatste promotie binnen de club, daar moet op gedronken worden! Beste Lotti en Jeroen, Feline en Lucia, en natuurlijk Linda, dankzij jullie voelen we ons helemaal thuis in Houdringe. Bedankt. Lieve broers en zussen van Janine, jullie zijn een prachtige en kleurrijke familie. Bedankt voor de support.

296


ACKNOWLEDGEMENTS

Miepie, Lou en kleine Mink, jullie zijn een schitterend stel en ik geniet enorm van jullie. Lieve Died, je bent een geweldige moeder en schoonmoeder. Ik bewonder je kracht en waardeer je adviezen. Bedankt voor alle hulp en gezelligheid. Femke en Maarten en de kleine Schuurtjes, jullie hebben al die jaren meegeleefd en interesse getoond in het onderzoek. Maar jullie zijn vast net zo blij als ik dat het nu af is. Bedankt voor alles. En lieve grote zus, je bent mijn paranimf en dat is logisch. Bedankt! Lieve Pappa en Mamma, de afgelopen jaren waren best een beetje pittig. Dankzij jullie onvoorwaardelijke steun en liefde komt het allemaal goed. Jullie zijn nog altijd een voorbeeld. Ik ben ongelofelijk dankbaar. Het is nog steeds heerlijk om na een late dienst een nachtje in Den Bosch te blijven en te genieten van jullie warmte. Ik hou van jullie! Lieve kleine stoere Isa, jij kwam ongeveer halverwege dit proefschrift in ons leven. Wat ben ik trots op mijn kleine waaghals. Pappaâ&#x20AC;&#x2122;s boekje is af, ik ga met je schommelen! Lieve Tieme, wat ben jij een mooi vrolijk mannetje. Jij maakt van het leven een feestje. Ga je mee op straat voetballen? Lieve Janine, my sweetie, 2 kindjes, 3 verhuizingen, allebei in opleiding en de laptop mee op zoâ&#x20AC;&#x2122;n beetje alle vakanties de afgelopen 8 jaar. Toch blijf je me in alles steunen. Dat is bijzonder. Je bent mijn hele houvast en ik geniet van je iedere dag. Ik hou van je.

297


APPENDIX

L I S T OF P U B L IC AT ION S 1. Boersma D, Smulders DL, Bakker OJ, van den Haak RF, Verhoeven BA, Koning OH. Endovenous laser ablation of insufficient perforating veins: energy is key to success. Vascular.2015 May 12.[Epub ahead of print] 2. Bakker OJ, van Santvoort HC; Dutch Pancreatitis Study Group. Reply. Gastroenterology. 2015 Apr 30.[Epub ahead of print] 3. Hollemans RA, Bollen TL, van Brunschot S, Bakker OJ, Ahmed Ali U, van Goor H, Boermeester MA, Gooszen HG, Besselink MG, van Santvoort HC; Dutch Pancreatitis Study Group. Predicting Success of Catheter Drainage in Infected Necrotizing Pancreatitis. Ann Surg. 2015 Mar 13. 4. Bakker OJ, Besselink MG, Gooszen HG. Early versus on-demand tube feeding in pancreatitis. N Engl J Med. 2015 Feb 12;372(7):685. 5. van Grinsven J, Besselink MG, Bakker OJ, van Brunschot S, Boermeester MA, van Santvoort HC. Retroperitoneoscopic Approaches for Infected Necrotizing Pancreatitis. In: C.E. Forsmark and T.B. Gardner (eds.), Prediction and Management of Severe Acute Pancreatitis. Springer Media, New York, 2015, pp 189-195. 6. Bakker OJ, van Brunschot S, van Santvoort HC, Besselink MG, Bollen TL, Boermeester MA, Dejong CH, van Goor H, Bosscha K, Ahmed Ali U, Bouwense S, van Grevenstein WM, Heisterkamp J, Houdijk AP, Jansen JM, Karsten TM, Manusama ER, Nieuwenhuijs VB, Schaapherder AF, van der Schelling GP, Schwartz MP, Spanier BW, Tan A, Vecht J, Weusten BL, Witteman BJ, Akkermans LM, Bruno MJ, Dijkgraaf MG, van Ramshorst B, Gooszen HG; Dutch Pancreatitis Study Group. Early versus on-demand nasoenteric tube feeding in acute pancreatitis. N Engl J Med. 2014 Nov 20;371(21):1983-93. 7. Vennix S, Bakker OJ, Prins HA, Lips DJ. Re-interventions Following Laparoscopic Surgery for Colorectal Cancer: Data from 818 Individuals from the Dutch Surgical Colorectal Audit. J Laparoendosc Adv Surg Tech A. 2014 Nov;24(11):751-5. 8. Bakker OJ, van Brunschot S, Farre A, Johnson CD, Kalfarentzos F, Louie BE, Olรกh A, O'Keefe SJ, Petrov MS, Powell JJ, Besselink MG, van Santvoort HC, Rovers MM, Gooszen HG. Timing of enteral nutrition in acute pancreatitis: meta-analysis of individuals using a single-arm of randomised trials. Pancreatology. 2014 SepOct;14(5):340-6.

298


LIST OF PUBLICATIONS

9. Hollemans RA, van Brunschot S, Bakker OJ, Bollen TL, Timmer R, Besselink MG, van Santvoort HC; Dutch Pancreatitis Study Group. Minimally invasive intervention for infected necrosis in acute pancreatitis. Expert Rev Med Devices. 2014 Nov;11(6):637-48. 10. van Brunschot S, Schut AJ, Bouwense SA, Besselink MG, Bakker OJ, van Goor H, Hofker S, Gooszen HG, Boermeester MA, van Santvoort HC; Dutch Pancreatitis Study Group. Abdominal compartment syndrome in acute pancreatitis: a systematic review. Pancreas. 2014 Jul;43(5):665-74. 11. Issa Y, Bruno MJ, Bakker OJ, Besselink MG, Schepers NJ, van Santvoort HC, Gooszen HG, Boermeester MA. Treatment options for chronic pancreatitis. Nat Rev Gastroenterol Hepatol. 2014 Sep;11(9):556-64. 12. Bakker OJ, Issa Y, van Santvoort HC, Besselink MG, Schepers NJ, Bruno MJ, Boermeester MA, Gooszen HG. Treatment options for acute pancreatitis. Nat Rev Gastroenterol Hepatol. 2014 Aug;11(8):462-9. 13. van Brunschot S, Fockens P, Bakker OJ, Besselink MG, Voermans RP, Poley JW, Gooszen HG, Bruno M, van Santvoort HC. Endoscopic transluminal necrosectomy in necrotising pancreatitis: a systematic review. Surg Endosc. 2014 May;28(5):142538. 14. van Baal MC, Bollen TL, Bakker OJ, van Goor H, Boermeester MA, Dejong CH, Gooszen HG, van der Harst E, van Eijck CH, van Santvoort HC, Besselink MG; Dutch Pancreatitis Study Group. The role of routine fine-needle aspiration in the diagnosis of infected necrotizing pancreatitis. Surgery. 2014 Mar;155(3):442-8. 15. van Brunschot S, van Grinsven J, Voermans RP, Bakker OJ, Besselink MG, Boermeester MA, Bollen TL, Bosscha K, Bouwense SA, Bruno MJ, Cappendijk VC, Consten EC, Dejong CH, Dijkgraaf MG, van Eijck CH, Erkelens GW, van Goor H, Hadithi M, Haveman JW, Hofker SH, Jansen JJ, LamĂŠris JS, van Lienden KP, Manusama ER, Meijssen MA, Mulder CJ, Nieuwenhuis VB, Poley JW, de Ridder RJ, Rosman C, Schaapherder AF, Scheepers JJ, Schoon EJ, Seerden T, Spanier BW, Straathof JW, Timmer R, Venneman NG, Vleggaar FP, Witteman BJ, Gooszen HG, van Santvoort HC, Fockens P; Dutch Pancreatitis Study Group. Transluminal endoscopic step-up approach versus minimally invasive surgical step-up approach in patients with infected necrotising pancreatitis (TENSION trial): design and rationale of a randomised controlled multicenter trial [ISRCTN09186711]. BMC Gastroenterol. 2013 Nov 25;13:161. 299


APPENDIX

16. da Costa DW, Boerma D, van Santvoort HC, Horvath KD, Werner J, Carter CR, Bollen TL, Gooszen HG, Besselink MG, Bakker OJ. Staged multidisciplinary step-up management for necrotizing pancreatitis. Br J Surg. 2014 Jan;101(1):e65-79. 17. Schepers NJ, Besselink MG, van Santvoort HC, Bakker OJ, Bruno MJ; Dutch Pancreatitis Study Group. Early management of acute pancreatitis. Best Pract Res Clin Gastroenterol. 2013 Oct;27(5):727-43. 18. Vennix S, Abegg R, Bakker OJ, van den Boezem PB, Brokelman WJ, Sietses C, Bosscha K, Lips DJ, Prins HA. Surgical re-interventions following colorectal surgery: open versus laparoscopic management of anastomotic leakage. J Laparoendosc Adv Surg Tech A. 2013 Sep;23(9):739-44. 19. Bouwense SA, Besselink MG, Brunschot S, Bakker OJ, Santvoort HC, Schepers NJ, Boermeester MA, Bollen TL, Bosscha K, Brink MA, Bruno MJ, Consten EC, Dejong CH, Duijvendijk P, Eijck CH, Gerritsen JJ, Goor H, Heisterkamp J, Hingh IH, Kruyt PM, Molenaar IQ, Nieuwenhuijs VB, Rosman C, Schaapherder AF, Scheepers JJ, Spanier MB, Timmer R, Weusten BL, Witteman BJ, Ramshorst B, Gooszen HG, Boerma D. Pancreatitis of biliary origin, optimal timing of cholecystectomy (PONCHO trial): study protocol for a randomized controlled trial. Trials. 2012 Nov 26;13(1):225. 20. Bakker OJ. Systematic review and meta-analysis of safety of laparoscopic versus open appendicectomy for suspected appendicitis in pregnancy (Br J Surg 2012; 99: 1470-1478). Br J Surg. 2012 Nov;99(11):1478-9. 21. Bakker OJ, van Santvoort H, Besselink MG, Boermeester MA, van Eijck C, Dejong K, van Goor H, Hofker S, Ahmed Ali U, Gooszen HG, Bollen TL; for the Dutch Pancreatitis Study Group. Extrapancreatic necrosis without pancreatic parenchymal necrosis: a separate entity in necrotising pancreatitis? Gut. 2013 Oct;62(10):1475-80. 22. van Brunschot S, Bakker OJ, Besselink MG, Bollen TL, Fockens P, Gooszen HG, van Santvoort HC; Dutch Pancreatitis Study Group. Treatment of necrotizing pancreatitis. Clin Gastroenterol Hepatol. 2012 Nov;10(11):1190-201. 23. Bakker OJ. Should conservative treatment of appendicitis be first line? BMJ. 2012;344:e2546 24. van Baal MC, Besselink MG, Bakker OJ, van Santvoort HC, Schaapherder AF, Nieuwenhuijs VB, Gooszen HG, van Ramshorst B, Boerma D; for the Dutch Pancreatitis Study Group. Timing of Cholecystectomy After Mild Biliary Pancreatitis: A Systematic Review. Ann Surg. 2012 Mar 30. 300


LIST OF PUBLICATIONS

25. Bakker OJ, van Santvoort HC, van Brunschot S, Geskus RB, Besselink MG, Bollen TL, van Eijck CH, Fockens P, Hazebroek EJ, Nijmeijer RM, Poley JW, van Ramshorst B, Vleggaar FP, Boermeester MA, Gooszen HG, Weusten BL, Timmer R; Dutch Pancreatitis Study Group. Endoscopic transgastric vs surgical necrosectomy for infected necrotizing pancreatitis: a randomized trial. JAMA. 2012 Mar 14;307(10):1053-61. 26. van Santvoort HC, Bakker OJ, Bollen TL, Besselink MG, Ali UA, Schrijver AM, Boermeester MA, van Goor H, Dejong CH, van Eijck CH, van Ramshorst B, Schaapherder AF, van der Harst E, Hofker S, Nieuwenhuijs VB, Brink MA, Kruyt PM, Manusama ER, van der Schelling GP, Karsten T, Hesselink EJ, van Laarhoven CJ, Rosman C, Bosscha K, de Wit RJ, Houdijk AP, Cuesta MA, Wahab PJ, Gooszen HG; Dutch Pancreatitis Study Group. A conservative and minimally invasive approach to necrotizing pancreatitis improves outcome. Gastroenterology. 2011 Oct;141(4):1254-63 27. Bakker OJ, van Santvoort HC, Hagenaars JC, Besselink MG, Bollen TL, Gooszen HG, Schaapherder AF; for the Dutch Pancreatitis Study Group. Timing of cholecystectomy after mild biliary pancreatitis. Br J Surg. 2011 Oct;98(10):1446-54. 28. Wu BU, Bakker OJ, Papachristou GI, Besselink MG, Repas K, Van Santvoort HC, Muddana V, Singh VK, Whitcomb DC, Gooszen HG, Banks PA. Blood urea nitrogen in the early assessment of acute pancreatitis: an international validation study. Arch Intern Med. 2011;171(7):669-76. 29. Bakker OJ, van Baal MC, van Santvoort HC, Besselink MG, Poley JW, Heisterkamp J, Bollen TL, Gooszen HG, van Eijck CH, for the Dutch Pancreatitis Study Group. Endoscopic transpapillary stenting or conservative treatment for pancreatic fistulas in necrotizing pancreatitis: multicenter series and literature review. Ann Surg. 2011 May;253(5):961-7. 30. Bouwense SA, Bakker OJ, Van Santvoort HC, Boerma D, van RB, Gooszen HG, Besselink MG. Safety of Cholecystectomy in the First 48 hours After Admission for Gallstone Pancreatitis not yet Proven. Ann Surg. 2011;253(5):1053-4. 31. Bakker OJ, van Santvoort HC, van Brunschot S, Ahmed Ali U, Besselink MG, Boermeester MA, Bollen TL, Bosscha K, Brink MA, Dejong CH, van Geenen EJ, van Goor H, Heisterkamp J, Houdijk AP, Jansen JM, Karsten TM, Manusama ER, Nieuwenhuijs VB, van Ramshorst B, Schaapherder AF, van der Schelling GP, Spanier BM, Tan A, Vecht J, Weusten BL, Witteman BJ, Akkermans LM, Gooszen HG, Dutch Pancreatitis Study Group Pancreatitis, very early compared with normal start of

301


APPENDIX

enteral feeding (PYTHON trial): design and rationale of a randomised controlled multicenter trial. Trials. 2011 Mar 10;12(1):73. 32. Van Baal MC, van Santvoort HC, Bollen TL, Bakker OJ, Besselink MG, Gooszen HG, for the Dutch Pancreatitis Study Group. Systematic review of percutaneous catheter drainage as primary treatment for necrotizing pancreatitis. Br J Surg. 2011 Jan;98(1):18-27. 33. Besselink MG, Van Santvoort HC, Bakker OJ, Bollen TL, Gooszen HG. Draining sterile fluid collections in acute pancreatitis? Primum non nocere! Surg Endosc. 2011 Jan;25(1):331-2. 34. Van Santvoort HC, Bakker OJ, Besselink MG, Bollen TL, Fischer K, Nieuwenhuijs VB, Gooszen HG, Van Erpecum KJ. Prediction of common bile duct stones in the earliest stages of acute biliary pancreatitis. Endoscopy. 2011 Jan;43(1):8-13. 35. Van Santvoort, Besselink MG, Bakker OJ, Vleggaar FP, Timmer R, Weusten BL, Gooszen HG; on behalf of the Dutch Pancreatitis Study Group Endoscopic necrosectomy in necrotising pancreatitis: indication is the key. Gut. Nov;59(11):1587. 36. Van Doesburg IAJ, Besselink MGH, Bakker OJ, Van Santvoort HC, Gooszen HG, on behalf of the Dutch Pancreatitis Study Group. Antibiotics, Probiotics and Enteral Nutrition: Means to Prevent Infected Necrosis in AP. In:Â Tilg H, Mayerle J (eds): Clinical Update on Inflammatory Disorders of the Gastrointestinal Tract. Front Gastrointest Res. Basel, Karger, 2010, vol 26, pp 156â&#x20AC;&#x201C;164. 37. Bakker OJ, Go PM, Puylaert JB, Kazemier G, Heij HA; Werkgroup en klankbordgroup "Richtlijn acute appendicitis". Guideline on diagnosis and treatment of acute appendicitis: imaging prior to appendectomy is recommended. Ned Tijdschr Geneeskd. 2010;154:A303. Dutch. 38. Leeuwenburgh MM, Bakker OJ, Gorzeman MP, Bollen TL, Seldenrijk CA, Go PM. Fewer unnecessary appendectomies following ultrasonography and CT. Ned Tijdschr Geneeskd. 2010;154(17):A869. 39. Van Santvoort HC, Besselink MG, Bakker OJ, Boermeester MA, Dejong CH, Van Goor H, Schaapherder AFM, Van Eijck CH, Bollen TL, Van Ramshorst B, Nieuwenhuijs VB, Timmer R, Lameris JS, Kruyt PhM, Manusama E, Van der Harst E, Van der Schelling GP, Karsten T, Hesselink EJ, Van Laarhoven CJ, Rosman C, Bosscha K, De Wit RJ, Houdijk AP, Van Leeuwen MS, Buskens E, Gooszen HG for the Dutch Acute Pancreatitis Study Group. A step-up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med. 2010 Apr 22;362(16):1491-502. 302


LIST OF PUBLICATIONS

40. Bakker OJ, van Santvoort HC, Besselink MG, van der Harst E, Hofker HS, Gooszen HG; Dutch Pancreatitis Study Group. Prevention, detection, and management of infected necrosis in severe acute pancreatitis. Curr Gastroenterol Rep. 2009;11:104-10. 41. Cobben LP, Bakker OJ, Puylaert JB, Kingma LM, Blickman JG. Imaging of patients with clinically suspected appendicitis in the Netherlands: conclusions of a survey. Br J Radiol. 2009 Jun;82(978):482-5.

303


APPENDIX

C U R R IC U LUM V I TA E Olaf Jan Bakker was born on September 4th, 1978 in Bergeyk, The Netherlands. He grew up in ‘s-Hertogenbosch and graduated from secondary school at St.Jans Lyceum in 1997. After one year of biology, he started medical school at the University of Utrecht in 1999. During his years at the university he was a field hockey player for HC Den Bosch and sv Kampong and won the national championship as well as the European cup. As a medical student he was a member of the national guideline committee on acute appendicitis and worked for several months in a small rural hospital in Tanzania. After graduating from medical school in 2006 he started working at the department of surgery of St.Antonius hospital in Nieuwegein (dr. P.M.N.Y.H. Go). In October 2007 he joined the Dutch Pancreatitis Study Group headed by prof.dr. H.G. Gooszen and coordinated it’s activities for two years including the PANTER, PENGUIN and PYTHON trials. In 2009 he started his training in general surgery at Jeroen Bosch Hospital, ‘s-Hertogenbosch (dr. K. Bosscha) and in 2012 at the University Medical Center Utrecht (prof. dr. M.R. Vriens). In 2013 he started his training in vascular surgery at the University Medical Center Utrecht (prof.dr. F.L. Moll) and continued his training at Jeroen Bosch Hospital (dr. O.H.J. Koning). In the summer of 2015 he will work at the department of cardiothoracic surgery of St.Antonius hospital, Nieuwegein (dr. R.H. Heijmen) for 3 months after which he will start a fellowship in vascular and endovascular surgery at the department of vascular surgery of the University Medical Center Utrecht (dr. G.J. de Borst). Olaf Bakker is (co-) author of over 40 peer reviewed articles and book chapters. He has won several awards, including the Top Abstract Prize at the United European Gastroenterology Week in 2013, the National Pancreas Foundation Best Abstract Prize of the American Pancreatic Association (2013 and 2011) and the Axcan Sager Research Prize of the European Pancreatic Club in 2009. He lives in Bilthoven with his girlfriend Janine and their two children Isa and Tieme.

304


wenz iD - Olaf bakker  

Complicated Pancreatitis

wenz iD - Olaf bakker  

Complicated Pancreatitis

Advertisement