Issuu on Google+

ORIGINAL E n d o c r i n e

ARTICLE R e s e a r c h

Ethinyl Estradiol-Cyproterone Acetate Versus Low-Dose Pioglitazone-Flutamide-Metformin for Adolescent Girls with Androgen Excess: Divergent Effects on CD163, TWEAK Receptor, ANGPTL4, and LEPTIN Expression in Subcutaneous Adipose Tissue Marta Díaz, Matilde R. Chaco´n, Abel Lo´pez-Bermejo, Elsa Maymo´-Masip, Cristina Salvador, Joan Vendrell, Francis de Zegher, and Lourdes Iba´n˜ez Endocrinology Unit (M.D., L.I.), and Department of Obstetrics and Gynecology (C.S.), Hospital Sant Joan de De´u, University of Barcelona, 08950 Esplugues, Barcelona, Spain; Hospital Universitari de Tarragona Joan XXIII (M.R.C., E.M.-M., J.V.),Institut d’Investigacio´ Sanita`ria Pere Virgili, University Rovira i Virgili, 43007 Tarragona, Spain; Centro de Investigacio´n Biome´dica en Red de Diabetes y Enfermedades Metabo´licas (M.D., M.R.C., E.M.-M., J.V., L.I.), Instituto de Salud Carlos III, 28029 Madrid, Spain; Department of Pediatrics (A.L.-B.), Dr. Josep Trueta Hospital, 17007 Girona; Girona Institute for Biomedical Research (A.L.-B.), 17007 Girona, Spain; and Pediatric Endocrinology (F.d.Z.), University Hospital Gasthuisberg, 3000 Leuven, Belgium

Objective: The aim was to compare the effects of a traditional therapy (an oral estroprogestagen) to those of a novel treatment (a low-dose combination of generics) in adolescent girls with androgen excess. Study Design and Methods: In an open-label trial over 1 yr, 34 adolescents (age, 16 yr; body mass index, 23 kg/m2) with hyperinsulinemic androgen excess and without pregnancy risk were randomized to receive daily ethinyl estradiol-cyproterone acetate (EE-CA; Diane 35 Diario) or a lowdose combination of pioglitazone 7.5 mg/d, flutamide 62.5 mg/d, and metformin 850 mg/d (PioFluMet). Markers of androgen excess, C-reactive protein, high molecular weight adiponectin, lipids, carotid intima media thickness, body composition (absorptiometry), abdominal fat partitioning (magnetic resonance imaging), and gene expression in longitudinal biopsies of sc adipose tissue at the abdominal level (RT-PCR) were assessed at baseline and after 1 yr. Results: EE-CA and low-dose PioFluMet reduced androgen excess comparably, but had divergent effects on C-reactive protein, high molecular weight adiponectin, lipids, carotid intima media thickness, lean mass, abdominal and visceral fat, and on the expression of CD163, leptin, TNF-like weak inducer of apoptosis receptor, and angiopoietin-like protein 4, respectively, related to macrophage activation, fat accretion, inflammation, and lipoprotein metabolism in adipose tissue. All these divergences pointed to a healthier condition on low-dose PioFluMet. Conclusion: EE-CA and PioFluMet are similarly effective in reversing androgen excess over 1 yr, but low-dose PioFluMet is superior in reversing inflammatory, metabolic, and cardiovascular anomalies that are often associated with androgen excess. (J Clin Endocrinol Metab 97: 3630 –3638, 2012)

ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2012 by The Endocrine Society doi: 10.1210/jc.2012-1754 Received March 22, 2012. Accepted June 25, 2012. First Published Online July 12, 2012


Abbreviations: AMPK, AMP kinase; ANGPTL4, angiopoietin-like protein 4; BMI, body mass index; cIMT, carotid intima media thickness; CRP, C-reactive protein; Ct, threshold cycle; CV, coefficient of variation; EE-CA, ethinyl estradiol-cyproterone acetate; HMW, high molecular weight; HSD11␤1, 11␤-hydroxysteroid dehydrogenase type 1; LDL, low-density lipoprotein; LEP, leptin; LPL, lipoprotein lipase; MRI, magnetic resonance imaging; NF-␬B, nuclear factor-␬B; PCOS, polycystic ovary syndrome; PioFluMet, pioglitazone-flutamidemetformin; PPAR␥, peroxisome proliferator-activated receptor ␥.

J Clin Endocrinol Metab, October 2012, 97(10):3630 –3638

J Clin Endocrinol Metab, October 2012, 97(10):3630 –3638

n adolescent girls with ovarian androgen excess, there are essentially two therapeutic strategies to revert the signs and symptoms associated with hyperandrogenism, such as hirsutism, acne, and irregular menses. The classic approach is an ovary-silencing therapy with oral contraceptives containing different combinations of estroprogestagens; among them, ethinyl estradiol-cyproterone acetate (EE-CA) is commonly used for its purported antiandrogenic properties (1, 2). An alternative approach is to correct the underpinning metabolic abnormalities— frequently including hyperinsulinemic insulin resistance—thereby normalizing ovarian function and attenuating the risk for subsequent diabetes and cardiovascular disease (3, 4). In the latter approach, insulin-sensitizing agents such as metformin, have been used alone (3), or in a low-dose combination with flutamide, or with both flutamide and pioglitazone (PioFluMet); such combinations were recently shown to have beneficial effects on body composition, on markers of low-grade inflammation, and on circulating adipokines in adolescents and young women with ovarian androgen excess (5, 6). Adipose tissue dysfunction in hyperandrogenic girls— who combine enlarged adipocytes, hypoadiponectinemia, and decreased lipoprotein lipase (LPL) activity—may be a key factor in the pathogenesis of insulin resistance (7). Gene expression studies in sc and visceral adipose tissue of women with androgen excess point to a dysregulation of functional pathways involved in insulin signaling and resistance, adipogenesis, lipid metabolism, inflammation, immune function, and oxidative stress (8 –11). So far, these studies have included only a few adolescents and/or nonobese patients (12) and have not addressed the potential effects of pharmacological intervention. In a first head-to-head study over 6 months, EE-CA and low-dose PioFluMet were found to have comparable effects on hyperandrogenemia in girls with androgen excess, but divergent effects on cardiovascular health markers, adipokines, visceral fat, and low-grade inflammation—all these divergences favoring treatment with PioFluMet (13). Here, we report the results over 12 months together with a first comparison of gene expression patterns in consecutive biopsies of sc adipose tissue. We hypothesized that the two treatments would also have divergent effects on gene expression in adipose tissue, and accordingly, we assessed the longitudinal changes in gene expression within subgroups (0 –12 months) and the differences in 0to 12-month changes between subgroups. For the purpose of the study, we elected to study genes encoding transcription factors and other proteins involved in glucose and lipid metabolism and in inflammatory pathways.



Subjects and Methods Study population and design The study population consisted of 34 girls with hyperinsulinemic androgen excess [age, 16 yr; range, 14 –17 yr; body mass index (BMI), 23 kg/m2; range, 19 –29 kg/m2]; all of them were at least 2 yr beyond menarche. The girls were recruited among patients consecutively seen in the Adolescent Endocrinology Unit of Sant Joan University Hospital, Barcelona, Spain, between January and November 2010. The recruitment procedure has been previously reported in detail (13). As described (13), the inclusion criteria were: 1) hyperinsulinemia, defined as fasting insulin above 15 ␮U/ml, and/or a peak insulin above 150 ␮U/ml, and/or mean insulin greater than 84 ␮U/ml on a 2-h oral glucose tolerance test (14); 2) clinical and biochemical androgen excess, as defined by the presence of hirsutism (score above 8 in the Ferriman and Gallwey scale), amenorrhea (absence of menses for more than 3 months), or oligomenorrhea (menstrual cycles of ⬎45 d), and high serum levels of androstenedione and/or testosterone in the follicular phase of the cycle (d 3–7) or after 2 months of amenorrhea. Exclusion criteria were: pregnancy risk at study start or on follow-up, evidence of anemia, thyroid dysfunction, bleeding disorder, Cushing syndrome, hyperprolactinemia, glucose intolerance, diabetes mellitus, late-onset adrenal hyperplasia due to 21-hydroxylase deficiency (17-hydroxyprogesterone levels ⱖ200 ng/dl in the follicular phase of the cycle or after 2 months of amenorrhea), abnormal electrolytes or abnormal liver or kidney function, and use of medication affecting gonadal or adrenal function, or carbohydrate or lipid metabolism. Randomization was performed with the SealedEnvelope program (Sealed Envelope Ltd., London, UK) (http://www., using random permuted blocks with strata for age and BMI (13). The girls were randomly assigned to receive daily, at dinner time, either Diane 35 Diario (BayerSchering, Madrid, Spain; 35 ␮g ethinyl estradiol plus 2 mg cyproterone acetate for 21 of 28 d and placebo for 7 of 28 d) or PioFluMet in low-dose [7.5 mg pioglitazone (one half of a 15-mg tablet of Actos; Takeda, Madrid, Spain); 62.5 mg flutamide (one quarter of a 250-mg tablet of Flutamida; Cinfa, Navarra, Spain), and 850 mg metformin (one full 850-mg tablet of Metformina; Sandoz, Barcelona, Spain)], as reported (13). Clinical and endocrine-metabolic variables, carotid intima media thickness (cIMT), body composition, and abdominal fat partitioning were all assessed at study start (0 months) and after 12 months. Adipose tissue biopsies were performed at baseline and after 12 months of treatment. The study was registered as ISRCTN45546616 and was conducted in Barcelona, without support from the pharmaceutical industry, after approval by the Institutional Review Board of Sant Joan de De´u University Hospital, after written consent by the parents, and after assent by each of the study girls. None of the 12-month results have previously been reported.

Clinical and endocrine-metabolic assessment The same investigator (L.I.) measured weight and height (Harpenden stadiometer), derived BMI, and scored hirsutism (Ferriman-Gallwey) and acne (Leeds grading scale). Baseline assessments were performed in the follicular phase (d 3–7) or after 2 months of amenorrhea. Fasting blood glucose,


Díaz et al.

Gene Expression in Adipose after EE-CA or PioFluMet

serum insulin, lipid profile, SHBG androgens, C-reactive protein (CRP), and high molecular weight (HMW) adiponectin were assessed together with blood count and a screening of liver and kidney function. Serum glucose, immunoreactive insulin, SHBG, and testosterone and serum lipids were assayed as described (13). CRP was measured with a highly sensitive method (Architect c8000; Abbott, Wiesbaden, Germany) with intra- and interassay coefficients of variation (CV) of less than 2% and a detection limit of 0.1 mg/liter. HMW adiponectin was assessed with an ELISA kit (13); the intra- and interassay CV were 2.4 and 8%, respectively. Serum samples were stored at ⫺20 C until assay.

Carotid intima media thickness Longitudinal ultrasound scans of the carotid arteries were obtained by the same investigator (blinded to treatment allocation), with high-resolution equipment (Acuson Sequoia 512 SHA; Medisales, Los Alamitos, CA) (13). The mean values obtained on the left and right sides were averaged; the intraobserver CV was less than 10%.

Body composition and abdominal fat partitioning Body composition was assessed by dual-energy x-ray absorptiometry with a Lunar Prodigy and Lunar software (version 3.4/ 3.5; GE Lunar Corp., Madison, WI); absolute whole body fat and lean mass (kilograms), as well as fat content in the abdominal region were assessed (13, 15). The sc vs. visceral partitioning of abdominal fat was assessed by magnetic resonance imaging (MRI) using a multiple-slice MRI 1.5 Tesla scan (Signa LX Echo Speed Plus Excite; General Electric, Milwaukee, WI); MRI was also used to assess intrahepatic lipid content (13, 15). All scans were performed by the same operator, and all images were analyzed by the same radiologist (both blinded to treatment allocation).

J Clin Endocrinol Metab, October 2012, 97(10):3630 –3638

Between 0.2 and 1 ␮g of total RNA was reverse-transcribed to cDNA using TaqMan Reverse Transcription reagents (Applied Biosystems, Foster City, CA) and subsequently diluted with nuclease-free water (Sigma, St. Louis, MO) to 20 ng/␮l cDNA. The gene amplifications were assessed using specific pairs of primers; the assay IDs for each predesigned probe are listed in Table 1. Real-time quantitative PCR was performed with duplicates on a 7900HT Fast Real-Time PCR System using commercial Taqman Assays (Applied Biosystems). SDS software 2.3 and RQ Manager 1.2 (Applied Biosystems) were used to analyze the results with the comparative threshold cycle (Ct) method (2⌬⌬Ct) (16). Ct values for each sample were normalized with an optimal reference gene (GAPDH; Hs03929097), after testing three additional housekeeping genes: ␤-actin, cyclophilin, and 18s RNA. GAPDH was selected based on its stable expression in all samples and conditions analyzed. All data were expressed as an n-fold difference relative to a calibrator (mixing T ⫽ 0 and T ⫽ 12 from one sample of each treatment arm). The relative expression results of the selected genes are reported in Table 1.

Statistical analyses Statistical analyses were performed with SPSS 12.0 (SPSS, Chicago, IL). For uniformity, results are expressed as mean ⫾ SEM. Results with non-Gaussian distributions were mathematically transformed to improve symmetry before statistical analysis. The effects of each treatment over time were assessed by using unpaired Student’s t test. Two-way ANOVA followed by Bonferroni post hoc comparison test (Prism 5 software; GraphPad Software, Inc., San Diego, CA) was used to assess the simultaneous interaction of treatment and time on gene expression changes. Associations between continuous variables were sought by correlation analyses. P ⬍ 0.05 was considered statistically significant.

Adipose tissue biopsy Longitudinal adipose tissue gene expression studies were performed in 31 of 34 girls (n ⫽ 14 in the Diane subgroup, and n ⫽ 17 in the PioFluMet subgroup). In one girl, no biopsy was performed because she had a history of hypertrophic scars; two other girls consented in having the baseline biopsy but declined to undergo a second biopsy at 12 months. Subcutaneous adipose tissue samples were obtained through a 5- to 7-mm cold scalpel incision about 1 cm below the umbilicus, under local anesthesia (2% mepivacaine). An anesthetic cream (Emla; AstraZeneca, Madrid, Spain) was applied in the periumbilical area in the previous 5–7 h to further minimize the discomfort of the procedure. The samples were immediately frozen in liquid nitrogen and stored until analysis.

Quantitative RT-PCR Between 10 and 50 mg of frozen sc white adipose tissue was homogenized using Polytron benchtop homogenizer (Kinematica Ag, Littau-Luzern, Switzerland) for RNA extraction. Total RNA was isolated using Tripure reagent (Roche Diagnostics, Indianapolis, IN) according to the manufacturer’s instructions. The quantity and quality of isolated RNA were determined with a Nanodrop ND-1000 spectrophotometer (Nanodrop Technologies, Wilmington, DE) and an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA), respectively.

Results Endocrinology, body composition, and abdominal fat partitioning EE-CA and PioFluMet had similarly normalizing effects on clinical and biochemical markers of androgen excess. Divergent effects were observed on fasting insulin, total and low-density lipoprotein (LDL)-cholesterol, CRP, HMW adiponectin, cIMT, lean mass, and abdominal and visceral fat (Table 2 and Fig. 1), all in favor of PioFluMet. Liver and renal function tests remained within normal limits (at 12 months, for EE-CA and PioFluMet subgroups: alanine aminotransferase, 14 ⫾ 1 and 15 ⫾ 2 IU/liter; aspartate aminotransferase, 16 ⫾ 1 and 17 ⫾ 1 IU/liter; creatinine, 0.68 ⫾ 0.01 and 0.67 ⫾ 0.01 mg/dl). Gene expression analysis: comparisons within subgroups Gene expression analysis at baseline was comparable between subgroups for all assessed genes (Table 1).

J Clin Endocrinol Metab, October 2012, 97(10):3630 –3638


TABLE 1. Results of the gene expression analysis in sc adipose tissue of adolescent girls with androgen excess who were randomized to receive EE-CA (n ⫽ 14) or low-dose PioFluMet (n ⫽ 17) for 12 monthsa Gene symbol

Assay ID

Gene denomination


Hs00174877 Hs00171993


Hs01016661 Hs00211522


Hs00160173 Hs99999174 Hs00218766




Hs00193510 Hs01547870 Hs00167385 Hs00605917 Hs00167155 Hs00918082


Hs00975988 Hs00168966 Hs01057145 Hs02340011 Hs00234592


Hs00174097 Hs00606407


Hs01009006 Hs00188012 Hs00173425 Hs00201385 Hs01037704 Hs00299515 Hs00609791 Hs00386101


Hs00985641 Hs00992664 Hs00231882



Leptin Tumor necrosis factor receptor superfamily, member 12 Macrophage-associated antigen FIAF; fasting-induced adipose factor/ angiopoietin-like protein 4 Perilipin; lipid droplet-associated protein GRP78; glucose-regulated protein 78 kDa Acyl-CoA-synthetase short-chain family member 2 Tumor necrosis factor superfamily, member 12 HSL (hormone-sensitive lipase) 11␤-hydroxysteroid dehydrogenase type 1 ACC1 (acetyl-coenzyme carboxylase-1) APM1 (adiponectin) PAI1 (plasminogen-activator-inhibitor-1) Nucleotide-binding domain, leucine-rich containing family, pyrine domain-containing-3 Serine/threonine kinase 11 GLUT 4 (glucose transporter 4) Fat mass and obesity-associated gene Glycerol kinase Peroxisome proliferator-activated receptor ␥ Interleukin 1-␤ Peroxisome proliferator-activated receptor ␦ Sirtuin 1 Fatty acid synthase Lipoprotein lipase Diacylglycerol acyltransferase-1 Pyruvate dehydrogenase kinase 4 Lipin 1 Fatty acid binding protein 4 Patatin-like phospholipase domaincontaining protein 2 gene Interleukin-6 Carnitine palmitoyltransferase 1 Peroxisome proliferator-activated receptor ␣ AMPK (AMP-activated protein kinase)







Treatment interaction P value

1.21 ⫾ 0.3 1.70 ⫾ 0.3

2.32 ⫾ 0.6 6.98 ⫾ 2.1b

2.04 ⫾ 0.5 2.91 ⫾ 0.7

0.82 ⫾ 0.2b,d 3.22 ⫾ 0.8

0.008 0.028

0.94 ⫾ 0.3 1.34 ⫾ 0.5

1.41 ⫾ 0.3 3.87 ⫾ 1.2

0.93 ⫾ 0.2 1.96 ⫾ 0.5

0.74 ⫾ 0.1d 1.56 ⫾ 0.4

0.031 0.038

1.24 ⫾ 0.2 1.28 ⫾ 0.2 1.73 ⫾ 0.4

1.92 ⫾ 0.5 2.53 ⫾ 0.3c 3.20 ⫾ 0.9

1.61 ⫾ 0.2 1.57 ⫾ 0.2 2.24 ⫾ 0.4

1.11 ⫾ 0.3 1.98 ⫾ 0.2 1.94 ⫾ 0.5

0.074 0.074 0.114

0.52 ⫾ 1.0

1.24 ⫾ 0.4

0.80 ⫾ 0.2

0.77 ⫾ 0.1


0.98 ⫾ 0.2 0.96 ⫾ 0.2 1.13 ⫾ 0.3 1.89 ⫾ 0.4 0.71 ⫾ 0.1 1.46 ⫾ 0.4

2.23 ⫾ 1.0 1.16 ⫾ 0.4 1.64 ⫾ 0.4 2.22 ⫾ 0.5 1.24 ⫾ 0.4 1.69 ⫾ 0.4

1.01 ⫾ 0.2 0.81 ⫾ 0.1 1.68 ⫾ 0.4 2.15 ⫾ 0.4 0.85 ⫾ 0.3 1.07 ⫾ 0.2

0.85 ⫾ 0.3 0.40 ⫾ 0.1b 1.42 ⫾ 0.3 1.64 ⫾ 0.4 0.85 ⫾ 0.2 1.93 ⫾ 0.4

0.149 0.186 0.299 0.341 0.358 0.379

1.10 ⫾ 0.2 0.88 ⫾ 0.2 1.08 ⫾ 0.1 0.71 ⫾ 0.2 1.90 ⫾ 0.6

4.20 ⫾ 1.1c 1.93 ⫾ 0.5 1.33 ⫾ 0.2 0.94 ⫾ 0.4 0.84 ⫾ 0.1

2.30 ⫾ 0.6 1.93 ⫾ 0.4 1.25 ⫾ 0.2 0.60 ⫾ 0.1 1.35 ⫾ 0.2

4.07 ⫾ 1.0 2.36 ⫾ 0.6 1.25 ⫾ 0.1 1.10 ⫾ 0.2 0.66 ⫾ 0.1b

0.415 0.513 0.533 0.554 0.568

1.27 ⫾ 0.4 1.60 ⫾ 0.3

0.99 ⫾ 0.3 1.20 ⫾ 0.1

1.75 ⫾ 0.5 1.52 ⫾ 0.3

1.95 ⫾ 0.5 1.41 ⫾ 0.3

0.581 0.622

1.42 ⫾ 0.3 1.59 ⫾ 0.6 1.73 ⫾ 0.4 1.16 ⫾ 0.2 1.38 ⫾ 0.3 1.78 ⫾ 0.4 1.98 ⫾ 0.6 1.44 ⫾ 0.4

1.68 ⫾ 0.3 1.75 ⫾ 0.5 1.47 ⫾ 0.3 1.30 ⫾ 0.3 1.73 ⫾ 0.6 1.33 ⫾ 0.3 1.09 ⫾ 0.2 1.66 ⫾ 0.5

1.30 ⫾ 0.2 1.82 ⫾ 0.4 1.48 ⫾ 0.2 1.30 ⫾ 0.2 1.08 ⫾ 0.2 1.86 ⫾ 0.2 1.56 ⫾ 0.3 1.25 ⫾ 0.3

1.34 ⫾ 0.2 1.54 ⫾ 0.5 1.00 ⫾ 0.2 1.29 ⫾ 0.2 1.21 ⫾ 0.3 1.24 ⫾ 0.2 0.85 ⫾ 0.2 1.34 ⫾ 0.4

0.632 0.669 0.725 0.750 0.764 0.765 0.785 0.874

1.67 ⫾ 0.6 2.60 ⫾ 0.7 0.89 ⫾ 0.2

1.33 ⫾ 0.6 2.87 ⫾ 0.8 1.06 ⫾ 0.2

1.32 ⫾ 0.7 2.84 ⫾ 0.7 0.86 ⫾ 0.1

0.85 ⫾ 0.1 3.01 ⫾ 0.7 1.04 ⫾ 0.1

0.911 0.952 0.960

1.33 ⫾ 0.2

1.26 ⫾ 0.1

1.11 ⫾ 0.1

1.03 ⫾ 0.1


Values are expressed as mean ⫾ SEM. a

No significant differences between randomized subgroups at start.


P ⬍ 0.05; c P ⱕ 0.01 within subgroups for 0 –12 months.


P ⬍ 0.05 between subgroups for 0 –12 months.

The longitudinal changes within subgroups (0 –12 months) disclosed an up-regulation of 78 kDa glucoseregulated protein precursor (HSPA5), of serine-threonine kinase 11 (STK11), and of TNF-like weak inducer of apoptosis (TWEAK) receptor [also named fibroblast growth factor-inducible 14 (Fn14)] mRNA expression in EE-CAtreated girls (Table 1 and Fig. 2A), and a down-regulation in the mRNA expression of peroxisome proliferator-activated receptor ␥ (PPAR␥), 11␤-hydroxysteroid dehydrogenase type 1 (HSD11␤1), and leptin (LEP) in the PioFluMet-treated girls (Table 1 and Fig. 2B) (all P ⬍ 0.05 to P ⬍ 0.01 vs. baseline). In the latter subgroup, STK11 expression tended to increase after 12 months.

Gene expression analysis: comparisons between subgroups The comparison of gene expression profiles between treatments after 12 months disclosed an up-regulation of CD163, TWEAK receptor, LEP, and of angiopoietin-like protein 4 (ANGPTL4) in EE-CA-treated girls (P ⬍ 0.05 to P ⬍ 0.01 vs. PioFluMet-treated girls; Table 1 and Fig. 2C). Correlations The up-regulation in CD163 expression in EE-CAtreated girls was associated with the increase in visceral fat (r ⫽ 0.59; P ⬍ 0.005) and the decrease in circulating HMW adiponectin (r ⫽ 0.41; P ⫽ 0.04), whereas the up-


Díaz et al.

Gene Expression in Adipose after EE-CA or PioFluMet

J Clin Endocrinol Metab, October 2012, 97(10):3630 –3638

TABLE 2. Data from adolescent girls with androgen excess who were randomized to receive EE-CA (n ⫽ 17) or low-dose PioFluMet (n ⫽ 17) for 12 months EE-CA

Age (yr) BMI (kg/m2) Hirsutism score Acne score Endogenously regular cycles SHBG (nmol/liter) Testosterone (ng/dl) Fasting insulin (␮U/ml)* LDL-cholesterol (mg/dl)* HDL-cholesterol (mg/dl)* Triglycerides (mg/dl)* CRP (mg/liter)* HMW adiponectin (mg/liter)* cIMT (mm) Absorptiometry BMD (g/cm2) Lean mass (kg) Fat mass (kg) Abdominal fat (kg) Lean mass minus fat mass MRI Abdominal fat (cm2) Visceral fat (cm2) Intrahepatic lipid content (%)

Low-dose PioFluMet

Baselinea 15.9 ⫾ 0.3 23.0 ⫾ 0.8 13.5 ⫾ 0.9 2.2 ⫾ 0.2 35% 23 ⫾ 3 58 ⫾ 7 9.7 ⫾ 1.1 81 ⫾ 3 50 ⫾ 2 89 ⫾ 20 0.9 ⫾ 0.2 15 ⫾ 3 0.35 ⫾ 0.01

12 months

⌬ 0 –12 months

23.9 ⫾ 0.6 9.1 ⫾ 0.6d 1.1 ⫾ 0.1d 0%c ˆ 162 ⫾ 8d 30 ⫾ 3d 5.3 ⫾ 1.1b 105 ⫾ 6c 64 ⫾ 4d 132 ⫾ 22b 2.6 ⫾ 0.4c 7 ⫾ 1c 0.37 ⫾ 0.01c

1.13 ⫾ 0.03 35.2 ⫾ 0.7 19.9 ⫾ 1.9 5.3 ⫾ 0.4 15.2 ⫾ 0.7 163 ⫾ 17 32 ⫾ 2 16.9 ⫾ 1.6

⌬ 0 –12 months

0.8 ⫾ 0.2 ⫺4.5 ⫾ 0.7 ⫺1.1 ⫾ 0.2 ⫺35% 139 ⫾ 8 ⫺27 ⫾ 6 ⫺4.4 ⫾ 1.6 24 ⫾ 6 14 ⫾ 2 43 ⫾ 17 1.7 ⫾ 0.4 ⫺8 ⫾ 2 0.02 ⫾ 0.01

Baselinea 12 months 16.5 ⫾ 0.3 22.6 ⫾ 0.6 23.4 ⫾ 0.5 14.0 ⫾ 0.9 8.4 ⫾ 0.4d 2.3 ⫾ 0.2 1.1 ⫾ 0.04d 41% 94%c 28 ⫾ 3 36 ⫾ 3c 63 ⫾ 7 40 ⫾ 3c 10.2 ⫾ 1.9 4.3 ⫾ 0.9c 80 ⫾ 5 82 ⫾ 5 56 ⫾ 3 57 ⫾ 3 66 ⫾ 11 71 ⫾ 6 1.0 ⫾ 0.2 0.4 ⫾ 0.1c 11 ⫾ 2 12 ⫾ 1 0.35 ⫾ 0.01 0.34 ⫾ 0.01

0.5 ⫾ 0.3 ⫺5.7 ⫾ 0.8 ⫺1.2 ⫾ 0.2 53%g 8 ⫾ 2g ⫺22 ⫾ 6 ⫺5.9 ⫾ 1.8e 2 ⫾ 4f 1 ⫾ 2g 5⫾9 ⫺0.7 ⫾ 0.2g 1 ⫾ 2f ⫺0.01 ⫾ 0.01f

1.14 ⫾ 0.02 34.0 ⫾ 0.8c 22.0 ⫾ 1.5b 5.9 ⫾ 0.4c 12.0 ⫾ 0.5c

0.01 ⫾ 0.02 ⫺1.2 ⫾ 0.3 2.1 ⫾ 0.7 0.6 ⫾ 0.2 ⫺3.0 ⫾ 0.1

1.17 ⫾ 0.02 37.4 ⫾ 0.7 18.7 ⫾ 1.2 5.3 ⫾ 0.3 18.7 ⫾ 0.5

1.16 ⫾ 0.02 38.0 ⫾ 0.6 19.3 ⫾ 1.3 5.2 ⫾ 0.3 18.7 ⫾ 0.5

⫺0.01 ⫾ 0.02 0.6 ⫾ 0.3g 0.6 ⫾ 0.6 ⫺0.1 ⫾ 0.2e ⫺0.0 ⫾ 0.1f

196 ⫾ 19c 37 ⫾ 2b 17.1 ⫾ 1.4

33 ⫾ 9 5⫾2 0.2 ⫾ 1.5

164 ⫾ 16 36 ⫾ 3 15.1 ⫾ 1.5

165 ⫾ 17 31 ⫾ 2b 11.6 ⫾ 1.3c

1 ⫾ 9e ⫺5 ⫾ 2f ⫺3.5 ⫾ 1.1

Values are expressed as mean ⫾ SEM. HDL, High-density lipoprotein; BMD, bone mineral density. ˆ 100% of these girls had menses with an invariable interval of 28 d, occurring during the intake of placebo pills for 7 of 28 d. * Indicative values in healthy adolescents (n ⫽ 18; age, 16.9 ⫾ 0.3 yr; BMI, 21.1 ⫾ 0.6 kg/m2): fasting insulin, 3.0 ⫾ 0.3 ␮U/ml; LDL-cholesterol, 81 ⫾ 5 mg/dl; HDL-cholesterol, 54 ⫾ 2 mg/dl; triglycerides, 54 ⫾ 4 mg/dl; CRP, 0.7 ⫾ 0.2 mg/liter; HMW adiponectin, 13 ⫾ 2 mg/liter. a

No significant differences between randomized subgroups at start.


P ⬍ 0.05; c P ⱕ 0.01; d P ⱕ 0.001, within subgroups for 0- to 12-month change (⌬).


P ⬍ 0.05; f P ⱕ 0.01; g P ⱕ 0.001 between subgroups for 0- to 12-month change (⌬).

regulation in TWEAK receptor was related to the increase in CRP (r ⫽ 0.49; P ⫽ 0.01) after 12 months (Supplemental Fig. 1).

Discussion Main findings The present study indicates that EE-CA and PioFluMet are similarly effective in the reversal of clinical and biochemical hyperandrogenism over at least 12 months but have divergent effects on markers of metabolic health and on the adipose-tissue expression of genes such as CD163, TWEAK receptor, LEP, and ANGPTL4, respectively related to macrophage activation, inflammation, fat accretion, and lipoprotein metabolism. Of note, the divergences observed in cIMT—an established cardiovascular risk marker—after 6 months were maintained after 12 months. All divergences pointed to a healthier condition in adolescents receiving PioFluMet.

CD163, macrophage activation, and inflammation CD163 is a member of the scavenger receptor cysteinerich superfamily, is almost exclusively expressed on monocytes and macrophages, and is closely related to macrophage activation (17). Typically, this marker is associated with the M2 macrophage subpopulation involved in the down-regulatory phase of the inflammatory process. Recently, increased CD163 expression was found in crownlike structures within sc adipose tissue of obese individuals, reflecting macrophage infiltration and inflammation and suggesting a compensatory activation of the immunosuppressive pathways in states of low-grade inflammation (18). The increase in CD163 expression was correlated with visceral fat deposition (independent of adiposity) and associated with increased levels of proinflammatory cytokines whose expression is regulated by the nuclear factor-␬B (NF-␬B) stress pathway, which may impair insulin signaling and result in insulin resistance and ectopic lipid accumulation (18, 19). The CD163 up-reg-

Δ CRP (mg/L)



-4 -8 -12

2 1 0 -1



1 0 -1 -2


0,02 0


30 20 10 0 -10



4 0 -4 -8


EE-CA PioFluMet




<0.001 Δ LDL-Chol (mg/dL)


Δ Visc Fat (cm )


Δ cIMT (mm)

Δ Lean Mass (kg)

Δ HMW-adipo (mg/L)

J Clin Endocrinol Metab, October 2012, 97(10):3630 –3638

EE-CA PioFluMet

EE-CA PioFluMet

FIG. 1. Changes over 12 months in HMW-adiponectin (HMW-adipo), CRP, LDL-cholesterol, lean mass (by absorptiometry), cIMT, and visceral fat area (Visc Fat, by MRI) in adolescent girls with androgen excess treated with EE-CA (n ⫽ 17; dark gray) or low-dose PioFluMet (n ⫽ 17; light gray). Results are shown as mean and SEM.

ulation in the EE-CA-treated girls in parallel with the increase in the mRNA levels of proinflammatory genes such as TWEAK receptor or LEP supports the notion that prolonged estroprogestagen intake may enhance the state of low-grade inflammation characteristic of ovarian androgen excess, and as a result, the CD163 compensatory activation process mentioned above may be operative. TWEAK receptor and CD163 interaction: effects of metformin and adiponectin CD163 may also bind TWEAK—a proinflammatory cytokine in adipocytes—in pathological conditions, serving as an alternate receptor for TWEAK in cells lacking its receptor (20). The proinflammatory actions and increased expression of both TWEAK and its receptor (Fn14) in adipocytes have been shown to be mediated by NF-␬B (21, 22) in humans. In girls receiving EE-CA, the combined increase in CD163 and TWEAK receptor expression was closely related to the changes in visceral fat, the increase in CRP—a marker of low-grade inflammation—and the decrease in HMW adiponectin after 12 months. These data suggest that estroprogestagens, when given in supraphysiological doses, may deregulate the CD163-TWEAK system, resulting in a stimulation of the NF-␬B stress pathway, which in turn may enhance the already existing metabolic and body composition abnormalities. In addition, adiponectin and metformin have been shown recently to down-regulate cellular CD163 in vitro, probably through AMP kinase (AMPK) activation, resulting in a suppression of inflammation and in a modulation of cytokine signaling pathways (23, 24). This mechanism may also have influenced the relative difference in CD163 expression after 12 months in the two subgroups, with op-


posing influences of metformin (accounting for CD163 down-regulation in the PioFluMet girls) and decreasing levels of HMW adiponectin (resulting in CD163 up-regulation in EE-CA girls). Notwithstanding, the changes in adipose tissue AMPK activity in vitro, as described in diabetic men treated with metformin (25), are not necessarily paralleled by changes in gene expression in vivo. To our knowledge, no studies have been performed in polycystic ovary syndrome (PCOS) women, assessing the effects of metformin on CD163 and/or AMPK expression/activity.

ANGPTL4 and lipids The up-regulated ANGPTL4 expression in EE-CA-treated girls may have contributed to the rise of circulating LDL-cholesterol and triglyceride levels observed in these patients after 12 months. Indeed, ANGPTL4 is preferentially expressed in adipose tissue, wherein it serves as a regulator of lipolysis; in peripheral tissues, ANGPTL4 regulates fatty acid uptake through inhibition of LPL by promoting conversion of active LPL dimers into inactive monomers (26, 27). Genetic knockout of ANGPTL4 in mice protects against development of atherosclerosis and suppresses the ability of macrophages to become foam cells in vitro (28); in contrast, ANGPTL4 overexpression results in hypertriglyceridemia and peripheral insulin resistance (26). In diabetic mice, metformin down-regulates ANGPTL4 expression in liver and in adipose tissue (28), whereas systemic PPAR␥ stimulation facilitates adipose tissue ANGPTL4 secretion, resulting in angiogenesis and endothelial cell growth necessary for adipose tissue expansion (29). It is tempting to speculate that in the PioFluMet-treated girls, the joint effects of metformin and low-dose pioglitazone may have accounted for a relatively lower expression of ANGPTL4, supporting the notion that low-dose pioglitazone acts as inhibitor of CDK5-mediated phosphorylation of PPAR␥ rather than as a PPAR␥ activator (30). Interestingly, in the PioFluMet girls, PPAR␥ was significantly down-regulated after 12 months, suggesting that this combination may indeed tone down adipocyte differentiation activity. LEP, adiposity, and inflammation EE-CA-treated girls also showed increased LEP mRNA levels after 12 months, as compared with the PioFluMet girls where a significant down-regulation of LEP was ob-


Díaz et al.

Gene Expression in Adipose after EE-CA or PioFluMet

J Clin Endocrinol Metab, October 2012, 97(10):3630 –3638

FIG. 2. Results of gene expression analyses in sc adipose tissue of adolescent girls with androgen excess who were randomized to receive EE-CA (n ⫽ 14) or low-dose PioFluMet (n ⫽ 17) for 12 months. A and B, Histograms show mean and SEM results for 0- to 12-month changes in gene expression for EE-CA-treated (A) and PioFluMet-treated (B) girls. C, The between-treatment differences in 0- to 12-month changes in gene expression. Significant within- and between-treatment differences in gene expression are marked with asterisks.

served after 12 months. Previous studies have revealed that LEP expression in hyperandrogenic women is more related to adiposity than to androgen excess (10, 31); however, most of these studies have been performed in obese women and have failed to clarify the relationship between LEP expression and the androgen excess phenotype (32, 33). It is well known that leptin may alter insulin action in isolated adipocytes and that it promotes glucose and fatty acid oxidation and lipolysis (34). These observations from in vitro and in vivo studies suggest that leptin promotes energy dissipation and decreases lipid deposition in adipose tissue. LEP overexpression in EE-CA-treated girls may reflect the increase in visceral fat and might be a feedback mechanism to compensate for this unfavorable pattern. Finally, the increase in LEP expression could be part of the inflammatory response, reflecting an unfavorable inflammatory profile in the adipose tissue of these girls (35). Longitudinal gene expression changes within subgroups The potential implications of the gene expression changes occurring only within subgroups deserve further

consideration. For example, increased expression of HSD11␤1 has been reported in insulin-resistant states, including in obesity and androgen excess (9, 36). In hyperandrogenic women, the increase in HSD11␤1 expression has also been related to increased upper body fat distribution, even in the absence of obesity (37). The down-regulation in HSD11␤1 expression after PioFluMet coincided with a decrease in fasting insulin and in abdominal and visceral fat, and accordingly, seems to fit well with this concept. In adipose tissue from both treatment subgroups, STK11 (formerly called LKB1) was among the genes whose expression was most up-regulated. STK11 is a primary upstream kinase of AMPK, a necessary element in cell metabolism that is required for maintaining energy homeostasis. Our ex vivo results align thus well with the recent in vitro evidence indicating that androgens downregulate STK11 expression in adipocytes and that such down-regulation can be reversed by an androgen-receptor blocker (flutamide) or by an estrogen-receptor ␣ agonist (37). Because STK11 seems to influence the efficacy of

J Clin Endocrinol Metab, October 2012, 97(10):3630 –3638

metformin therapy in hyperinsulinemic girls with androgen excess (38), low-dose flutamide could henceforth be viewed as a “metformin sensitizer” in the adipose tissue of such girls. Finally, the increase in the expression of HSPA5—a resident chaperone within the endoplasmic reticulum—may reflect enhanced endoplasmic reticulum stress in adipocytes after EE-CA intervention; indeed, HSPA5 up-regulation has been linked to apoptosis and ␤-cell dysfunction in both animal and human models (39, 40). Strengths and weaknesses The present study appears to be the first to assess gene expression in sc adipose tissue of nonobese PCOS adolescents. The longitudinal design allowed us to detect divergent effects of EE-CA and low-dose PioFluMet on gene regulation in adipose tissue and to disclose a parallelism between changes in phenotype and in adipose-tissue gene expression. The main limitation of the study is the scarce amount of adipose tissue obtained with a minimal scalpel incision; this limitation precluded the assessment of adipose tissue morphology and also impeded the study of protein expression. Conclusion In conclusion, EE-CA and PioFluMet were similarly effective in reverting hyperandrogenism in nonobese PCOS adolescents but had divergent effects on the cardiometabolic phenotype, including on cIMT, paralleled by differences in the expression of genes related to macrophage activation, inflammation, lipoprotein metabolism, and fat accretion, all the divergences being in favor of low-dose PioFluMet. Further exploration of low-dose PioFluMet seems warranted, not only to strengthen its safety side over the longer term and in larger cohorts, but also to verify further whether the efficacy of this low-dose combination holds enough potential to allow it to become a first-choice therapy for the majority of young adolescents with androgen excess, namely those who are not at pregnancy risk and who are nowadays nevertheless exposed to supraphysiological doses of estroprogestagens.

Acknowledgments We thank Luis del Río, M.D., and Silvana di Gregorio, M.D., from Centro Me´dico CETIR, Barcelona, Spain, for performing the absorptiometry and MRI measurements. We are grateful to Giorgia Sebastiani for performing the carotid ultrasound assessments.


Address all correspondence and requests for reprints to: Lourdes Iba´n˜ez, M.D., Ph.D., Hospital Sant Joan de De´u, University of Barcelona, Passeig de Sant Joan de De´u, 2, 08950 Esplugues, Barcelona, Spain. E-mail: This study was supported by grants from the Ministerio de Ciencia e Innovacio´n, Instituto de Salud Carlos III, Madrid, Spain (PI09/90444, PI11/00049, and PI11/00085). M.R.C. is supported by a fellowship from the Fondo de Investigacio´n Sanitaria (FIS) “Miguel Servet” CP 06/00119. F.d.Z. is a Clinical Investigator supported by the Clinical Research Council of the Leuven University Hospitals, and by the University of Leuven (OT-04-35). A.L.-B. is a Clinical Investigator of the I3 Fund for Scientific Research (Ministry of Science and Innovation, Spain). Trial registration: ISRCTN45546616. Access at http:// Disclosure Summary: M.D., M.R.C., A.L.-B., E.M.-M., C.S., J.V., F.d.Z., and L.I. have nothing to disclose.

References 1. Pfeifer SM, Kives S 2009 Polycystic ovary syndrome in the adolescent. Obstet Gynecol Clin North Am 36:129 –152 2. Franks S 2008 Polycystic ovary syndrome in adolescents. Int J Obes (Lond) 32:1035–1041 3. Dunaif A 2008 Drug insight: insulin-sensitizing drugs in the treatment of polycystic ovary syndrome—a reappraisal. Nat Clin Pract Endocrinol Metab 4:272–283 4. Wild RA, Carmina E, Diamanti-Kandarakis E, Dokras A, EscobarMorreale HF, Futterweit W, Lobo R, Norman RJ, Talbott E, Dumesic DA 2010 Assessment of cardiovascular risk and prevention of cardiovascular disease in women with the polycystic ovary syndrome: a consensus statement by the Androgen Excess and Polycystic Ovary Syndrome (AE-PCOS) Society. J Clin Endocrinol Metab 95:2038 –2049 5. Iba´n˜ez L, de Zegher F 2006 Low-dose flutamide-metformin therapy for hyperinsulinemic hyperandrogenism in non-obese adolescents and women. Hum Reprod Update 12:243–252 6. Iba´n˜ez L, Lo´pez-Bermejo A, del Rio L, Enríquez G, Valls C, de Zegher F 2007 Combined low-dose pioglitazone, flutamide, and metformin for women with androgen excess. J Clin Endocrinol Metab 92:1710 –1714 7. Mannerås-Holm L, Leonhardt H, Kullberg J, Jennische E, Ode´n A, Holm G, Hellstro¨m M, Lo¨nn L, Olivecrona G, Stener-Victorin E, Lo¨nn M 2011 Adipose tissue has aberrant morphology and function in PCOS: enlarged adipocytes and low serum adiponectin, but not circulating sex steroids, are strongly associated with insulin resistance. J Clin Endocrinol Metab 96:E304 –E311 8. Corto´n M, Botella-Carretero JI, Benguría A, Villuendas G, Zaballos A, San Milla´n JL, Escobar-Morreale HF, Peral B 2007 Differential gene expression profile in omental adipose tissue in women with polycystic ovary syndrome. J Clin Endocrinol Metab 92:328 –337 9. Svendsen PF, Madsbad S, Nilas L, Paulsen SK, Pedersen SB 2009 Expression of 11␤-hydroxysteroid dehydrogenase 1 and 2 in subcutaneous adipose tissue of lean and obese women with and without polycystic ovary syndrome. Int J Obes (Lond) 33:1249 –1256 10. Lecke SB, Mattei F, Morsch DM, Spritzer PM 2011 Abdominal subcutaneous fat gene expression and circulating levels of leptin and adiponectin in polycystic ovary syndrome. Fertil Steril 95:2044 – 2049 11. Lindholm A, Blomquist C, Bixo M, Dahlbom I, Hansson T, Sundstro¨m Poromaa I, Bure´n J 2011 No difference in markers of adipose tissue inflammation between overweight women with polycystic ovary syndrome and weight-matched controls. Hum Reprod 26: 1478 –1485


Díaz et al.

Gene Expression in Adipose after EE-CA or PioFluMet

12. Chazenbalk G, Chen YH, Heneidi S, Lee JM, Pall M, Chen YD, Azziz R 2012 Abnormal expression of genes involved in inflammation, lipid metabolism, and Wnt signaling in the adipose tissue of polycystic ovary syndrome. J Clin Endocrinol Metab 97:E765–E770 13. Iba´n˜ez L, Diaz M, Sebastiani G, Sa´nchez-Infantes D, Salvador C, Lopez-Bermejo A, de Zegher F 2011 Treatment of androgen excess in adolescent girls: ethinylestradiol-cyproteroneacetate versus lowdose pioglitazone-flutamide-metformin. J Clin Endocrinol Metab 96:3361–3366 14. Ibáñez L, Potau N, Zampolli M, Prat N, Virdis R, Vicens-Calvet E, Carrascosa A 1996 Hyperinsulinemia in postpubertal girls with a history of premature pubarche and functional ovarian hyperandrogenism. J Clin Endocrinol Metab 81:1237–1243 15. Iba´n˜ez L, Lo´pez-Bermejo A, Díaz M, Marcos MV, de Zegher F 2011 Early metformin therapy (age 8 –12 yr) in girls with precocious pubarche to reduce hirsutism, androgen excess, and oligomenorrhea in adolescence. J Clin Endocrinol Metab 96:E1262–E1267 16. Schmittgen TD, Livak KJ 2008 Analyzing real-time PCR data by the comparative CT method. Nat Protoc 3:1101–1108 17. Møller HJ 2012 Soluble CD163. Scand J Clin Lab Invest 72:1–13 18. Leˆ KA, Mahurkar S, Alderete TL, Hasson RE, Adam TC, Kim JS, Beale E, Xie C, Greenberg AS, Allayee H, Goran MI 2011 Subcutaneous adipose tissue macrophage infiltration is associated with hepatic and visceral fat deposition, hyperinsulinemia, and stimulation of NF-␬B stress pathway. Diabetes 60:2802–2809 19. Farb MG, Bigornia S, Mott M, Tanriverdi K, Morin KM, Freedman JE, Joseph L, Hess DT, Apovian CM, Vita JA, Gokce N 2011 Reduced adipose tissue inflammation represents an intermediate cardiometabolic phenotype in obesity. J Am Coll Cardiol 58:232–237 20. Bover LC, Cardo´-Vila M, Kuniyasu A, Sun J, Rangel R, Takeya M, Aggarwal BB, Arap W, Pasqualini R 2007 A previously unrecognized protein-protein interaction between TWEAK and CD163: potential biological implications. J Immunol 178:8183– 8194 21. Winkles JA 2008 The TWEAK-Fn14 cytokine-receptor axis: discovery, biology and therapeutic targeting. Nat Rev Drug Discov 7:411– 425 22. Vendrell J, Maymo´-Masip E, Tinahones F, García-Espan˜a A, Megia A, Caubet E, García-Fuentes E, Chaco´n MR 2010 Tumor necrosislike weak inducer of apoptosis as a proinflammatory cytokine in human adipocyte cells: up-regulation in severe obesity is mediated by inflammation but not hypoxia. J Clin Endocrinol Metab 95: 2983–2992 23. Sporrer D, Weber M, Wanninger J, Weigert J, Neumeier M, Sto¨gbauer F, Lieberer E, Bala M, Kopp A, Scha¨ffler A, Buechler C 2009 Adiponectin downregulates CD163 whose cellular and soluble forms are elevated in obesity. Eur J Clin Invest 39:671– 679 24. Sag D, Carling D, Stout RD, Suttles J 2008 Adenosine 5⬘-monophosphate-activated protein kinase promotes macrophage polarization to an anti-inflammatory functional phenotype. J Immunol 181: 8633– 8641 25. Boyle JG, Logan PJ, Jones GC, Small M, Sattar N, Connell JM, Cleland SJ, Salt IP 2011 AMP-activated protein kinase is activated in adipose tissue of individuals with type 2 diabetes treated with metformin: a randomised glycaemia-controlled crossover study. Diabetologia 54:1799 –1809 26. Lichtenstein L, Berbe´e JF, van Dijk SJ, van Dijk KW, Bensadoun A, Kema IP, Voshol PJ, Mu¨ller M, Rensen PC, Kersten S 2007 Angptl4 upregulates cholesterol synthesis in liver via inhibition of LPL- and

27. 28.













J Clin Endocrinol Metab, October 2012, 97(10):3630 –3638

HL-dependent hepatic cholesterol uptake. Arterioscler Thromb Vasc Biol 27:2420 –2427 Mattijssen F, Kersten S 2012 Regulation of triglyceride metabolism by angiopoietin-like proteins. Biochim Biophys Acta 1821:782–789 Adachi H, Fujiwara Y, Kondo T, Nishikawa T, Ogawa R, Matsumura T, Ishii N, Nagai R, Miyata K, Tabata M, Motoshima H, Furukawa N, Tsuruzoe K, Kawashima J, Takeya M, Yamashita S, Koh GY, Nagy A, Suda T, Oike Y, Araki E 2009 Angptl 4 deficiency improves lipid metabolism, suppresses foam cell formation and protects against atherosclerosis. Biochem Biophys Res Commun 379: 806 – 811 Mizutani N, Ozaki N, Seino Y, Fukami A, Sakamoto E, Fukuyama T, Sugimura Y, Nagasaki H, Arima H, Oiso Y 2012 Reduction of insulin signaling upregulates angiopoietin-like protein 4 through elevated free fatty acids in diabetic mice. Exp Clin Endocrinol Diabetes 120:139 –144 Choi JH, Banks AS, Estall JL, Kajimura S, Bostro¨m P, Laznik D, Ruas JL, Chalmers MJ, Kamenecka TM, Blu¨her M, Griffin PR, Spiegelman BM 2010 Anti-diabetic drugs inhibit obesity-linked phosphorylation of PPAR␥ by Cdk5. Nature 466:451– 456 Carmina E, Chu MC, Moran C, Tortoriello D, Vardhana P, Tena G, Preciado R, Lobo R 2008 Subcutaneous and omental fat expression of adiponectin and leptin in women with polycystic ovary syndrome. Fertil Steril 89:642– 648 You T, Yang R, Lyles MF, Gong D, Nicklas BJ 2005 Abdominal adipose tissue cytokine gene expression: relationship to obesity and metabolic risk factors. Am J Physiol Endocrinol Metab 288:E741– E747 Skopkova´ M, Penesova´ A, Sell H, Ra´dikova´ Z, Vlcek M, Imrich R, Koska J, Ukropec J, Eckel J, Klimes I, Gasperíkova´ D 2007 Protein array reveals differentially expressed proteins in subcutaneous adipose tissue in obesity. Obesity (Silver Spring) 15:2396 –2406 Ceddia RB, William Jr WN, Lima FB, Curi R 1998 Leptin inhibits insulin-stimulated incorporation of glucose into lipids and stimulates glucose decarboxylation in isolated rat adipocytes. J Endocrinol 158:R7–R9 Kirchgessner TG, Uysal KT, Wiesbrock SM, Marino MW, Hotamisligil GS 1997 Tumor necrosis factor-␣ contributes to obesityrelated hyperleptinemia by regulating leptin release from adipocytes. J Clin Invest 100:2777–2782 Baudrand R, Domínguez JM, Carvajal CA, Riquelme A, Campino C, Macchiavello S, Bozinovic M, Morales M, Pizarro M, Solis N, Escalona A, Boza C, Arrese M, Fardella CE 2011 Overexpression of hepatic 5␣-reductase and 11␤-hydroxysteroid dehydrogenase type 1 in visceral adipose tissue is associated with hyperinsulinemia in morbidly obese patients. Metabolism 60:1775–1780 McInnes KJ, Brown KA, Hunger NI, Simpson ER 30 August 2011 Regulation of LKB1 expression by sex hormones in adipocytes. Int J Obes (Lond) doi: 10.1038/ijo.2011.172 Lo´pez-Bermejo A, Díaz M, Mora´n E, de Zegher F, Iba´n˜ez L 2010 A single nucleotide polymorphism in STK11 influences insulin sensitivity and metformin efficacy in hyperinsulinemic girls with androgen excess. Diabetes Care 33:1544 –1548 Laybutt DR, Preston AM, Akerfeldt MC, Kench JG, Busch AK, Biankin AV, Biden TJ 2007 Endoplasmic reticulum stress contributes to ␤ cell apoptosis in type 2 diabetes. Diabetologia 50:752–763 Song XJ, Yang CY, Liu B, Wei Q, Korkor MT, Liu JY, Yang P 2011 Atorvastatin inhibits myocardial cell apoptosis in a rat model with post-myocardial infarction heart failure by downregulating ER stress response. Int J Med Sci 8:564 –572

Journal of clinical endocrinology and metabolism 2012 97 (10) 3630