Decidual Endovascular Trophoblast Invasion in Women with Polycystic Ovary Syndrome: An Experimental Case-Control Study

ORIGINAL E n d o c r i n e

ARTICLE R e s e a r c h

Decidual Endovascular Trophoblast Invasion in Women with Polycystic Ovary Syndrome: An Experimental Case-Control Study Stefano Palomba, Tiziana Russo, Angela Falbo, Annalisa Di Cello, Giuseppina Amendola, Rosa Mazza, Achille Tolino, Fulvio Zullo, Luigi Tucci, and Giovanni Battista La Sala Obstetrics and Gynecology Unit (S.P., T.R., A.F., G.B.L.S.), Department of Obstetrics, Gynecology, and Pediatrics, Azienda Ospedaliera Arcispedale Santa Maria Nuova, Istituto di Ricovero e Cura a Carattere Scientifico, University of Modena and Reggio Emilia, 42123 Reggio Emilia, Italy; Department of Obstetrics and Gynecology (A.D.C., G.A., R.M., F.Z.), University “Magna Graecia” of Catanzaro, and Unit of Pathology (L.T.), “Pugliese-Ciaccio” Hospital, 88100 Catanzaro, Italy; Department of Obstetrics and Gynecology (A.T.), University “Federico II” of Naples, 80131 Naples, Italy Context: Previous experimental and clinical data suggest impaired decidual trophoblast invasion in patients with polycystic ovarian syndrome (PCOS). Objective: The objective of the study was to test the hypothesis that decidual endovascular trophoblast invasion in pregnant patients with PCOS is impaired and to clarify the potential mechanisms involved. Design: This was an experimental case-control study. Setting: The study was conducted at the academic Departments of Obstetrics and Gynecology and the Unit of Pathology (Italy). Patients: Forty-five pregnant subjects screened from a wide population of women waiting for legal pregnancy termination were included in the final analysis. Specifically, 15 pregnant patients with PCOS were enrolled as cases and another 30 age- and body mass index (BMI)-matched healthy pregnant women without any feature of PCOS were enrolled as the controls. Intervention: Interventions included the collection of trophoblastic and decidual tissue at the 12th week of gestation. Main Outcome Measures: Clinical, ultrasonographic, and biochemical data as well as the histological analysis of decidual endovascular trophoblast invasion. Results: The rate of implantation site vessels with endovascular trophoblast invasion (ratio between total number of implantation site vessels and total number of vessels with endovascular trophoblast invasion) and the extent of endovascular trophoblast invasion (proportion between immunoreactive areas to cytokeratin 7 and to CD34) were significantly lower in patients with PCOS compared with healthy non-PCOS controls. Endovascular trophoblast invasion data were significantly and indirectly related to the markers of insulin resistance and testosterone concentrations in PCOS patients. Conclusions: Pregnant patients with PCOS patients have impaired decidual trophoblast invasion. Further studies are needed to evaluate the exact mechanisms through which insulin resistance and hyperandrogenemia exert this effect. (J Clin Endocrinol Metab 97: 2441–2449, 2012)

ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2012 by The Endocrine Society doi: 10.1210/jc.2012-1100 Received January 12, 2012. Accepted March 19, 2012. First Published Online April 25, 2012

Abbreviations: A, Androstenedione; AUC, area under curve; AUCglucose, AUC for glucose; AUCinsulin, AUC for insulin; BMI, body mass index; CV, coefficient of variation; DHEAS, dehydroepiandrosterone sulfate; DM, diabetes mellitus; FAI, free androgen index; GIR, fasting glucose-to-insulin ratio; GLUT4, glucose transporter 4; HOMA, homeostasis model of assessment; HOXA 10, homeobox A10; IGFBP-1, IGF-binding protein-1; OGTT, oral glucose tolerance test; P, progesterone; PCOS, polycystic ovarian syndrome; PI, pulsatility index; RI, resistance index; T, testosterone; WHR, waist to hip ratio.

J Clin Endocrinol Metab, July 2012, 97(7):2441–2449

jcem.endojournals.org

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 11 January 2014. at 03:49 For personal use only. No other uses without permission. . All rights reserved.

2441

2442

Palomba et al.

Decidual Trophoblast Invasion in PCOS

olycystic ovary syndrome (PCOS) is one of the most common endocrine disorders in women of childbearing age. The impacts of the syndrome on human reproduction range from anovulatory infertility to detrimental effects on oocyte (1, 2) and embryo (3) quality, endometrial receptivity (4), and pregnancy outcomes (5). The spectrum of pregnancy complications associated with PCOS seems to be related to defects in placentation and impaired decidual trophoblast invasion (6). In particular, the failure of trophoblastic invasion into the musculoelastic coat of the spiral arteries has been shown to result in incomplete vascular transformation and persistently increased impedance of the uterine arteries (7). In fact, one of the main macroscopic alterations in trophoblast invasion observed in the placenta of patients with complicated pregnancies, which is known to be closely related to adverse pregnancy outcomes, is the anatomic failure of the formation of competent uterine spiral arteries in the decidua (7). Defects in decidual trophoblast invasion have been indirectly evaluated in patients with PCOS using noninvasive methods. Several proteins thought to play an important role in the pathophysiology of this defect include glycodelin and IGF-binding protein-1 (IGFBP-1), which are regulated by insulin and produced by the epithelium and stroma, respectively. The secretion of these proteins may play an important role in endometrial receptivity and has been shown to be reduced in PCOS (8). Any impairment in decidual trophoblast invasion induces a high-resistance state in the uteroplacental circulation. In fact, a direct relationship has been demonstrated between blood flow resistance as measured by Doppler ultrasonography and trophoblastic invasion of the decidual vessels on histological stains (9). In particular, pregnancies with a low-resistance uterine artery flow pattern during the first trimester of pregnancy are associated with a more extensive decidual trophoblast invasion pattern than pregnancies with high-resistance uterine artery flow patterns (9). Furthermore, a recent prospective case-control study showed that uterine artery Doppler indices are more commonly altered in pregnant patients with PCOS than in controls, suggesting that PCOS patients are characterized by a defect in trophoblast invasion (10). This hypothesis was also supported by the higher predictive value of the uterine artery Doppler indices for abnormal pregnancy and perinatal outcomes in the PCOS population (10). To date, no specific and direct data evaluating decidual trophoblast invasion in PCOS are available in the literature. Based on these considerations, the goals of the current experimental case-control study were to confirm that pregnant patients with PCOS have abnormalities in de-

P

J Clin Endocrinol Metab, July 2012, 97(7):2441–2449

cidual endovascular trophoblast invasion and to clarify the potential mechanisms involved.

Materials and Methods Ethics The procedures used in the present study were in accordance with the Helsinki Declaration on human experimentation guidelines. The study was approved by the Institutional Review Board of the Department of Obstetrics and Gynecology of the “Pugliese-Ciaccio” Hospital, Catanzaro (Italy). The purpose of the protocol was carefully explained to all women before they were entered into the study, and the written consent to participate to the experimental study was obtained from all participants. In any case, incentives for the patients were not provided.

Subjects Between December 2010 and March 2011, a group of women, waiting for legal pregnancy termination for nonmedical reasons, as regulated by the Italian law, were consecutively screened at the outpatient pregnancy termination service of the Department of Obstetrics and Gynecology of Catanzaro. In particular, we selected pregnant patients with PCOS who were eligible to be enrolled as cases. Other healthy pregnant women matched with cases for age and body mass index (BMI) who were without signs or symptoms of PCOS (as detailed in the text below) were enrolled as controls. To improve the statistical efficiency of the matching procedure, multiple controls were used. In particular, each case was matched with two controls (one to two). Controls were defined as age and BMI matched with the cases when the differences between them were less than 2 yr and 1 kg/m2 for age and BMI, respectively. PCOS was confirmed at the study enrollment visit according to the criteria specified by the European Society of Human Reproduction and Embryology/American Society of Reproductive Medicine (11). Healthy women were included as controls if they had regular menstrual cycles (26 –32 d in length) before pregnancy, no signs of clinical hyperandrogenism (Ferriman-Gallwey score less than 8) (12), serum androgens levels within normal ranges [according to our reference range for total testosterone (T), androstenedione (A), and dehydroepiandrosterone sulfate (DHEAS); ⬍ 81 ng/dl, ⬍ 430 mg/dl, and ⬍ 3.3 ng/ml for T, DHEAS, and A, respectively, coefficient of variations (CV) ranging between 5 and 10%], and no polycystic ovary morphology on ultrasound (13). Exclusion criteria were age younger than 18 yr or older than 35 yr, severe obesity (defined as a BMI ⬎ 35 kg/m2), multiple gestations, blighted ovum, the absence of embryo heart activity at the moment of legal termination, premalignancies or malignancies, medical conditions or other concurrent medical illnesses, metabolic abnormalities [including diabetes mellitus (DM), impaired glucose tolerance, and impaired fasting glucose], cigarette smoking, drug/alcohol use, uterine malformations, noncompliance with the study protocol, and current or previous (within the last 3 months) use of any hormonal and/or antidiabetic drugs and/or ovulation inductors and/or multivitamins and/or folic acid. Patients who lacked histological evidence of sites of endovascular trophoblast invasion were also excluded from the final analysis (see below).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 11 January 2014. at 03:49 For personal use only. No other uses without permission. . All rights reserved.

J Clin Endocrinol Metab, July 2012, 97(7):2441–2449

Protocol At the time of pregnancy termination, each subject underwent a clinical evaluation and an ultrasonographic assessment, and a venous blood sample was taken for biochemical analysis. All legal pregnancy terminations were performed at the 12th week of gestation using a hysterosuction technique in the Hospital “Pugliese-Ciaccio” (Catanzaro, Italy). At the time of pregnancy termination, histological samples were taken, processed and stored for evaluation of trophoblast invasion as detailed below.

Clinical assessments Clinical assessments consisted of anthropometric measurements [height, weight, BMI, and waist to hip ratio (WHR)], Ferriman-Gallwey score (12) calculation, and heart rate and blood pressure assessments. During the same visit, a semiquantitative questionnaire to evaluate physical activity and job and daily activities was completed by each woman (14); simultaneously, the daily diet was assessed with two well-validated questionnaires (15, 16). Socioeconomic status, work status, education level, ethnicity, and associated medical conditions were carefully assessed for each woman (17). Finally, women were asked to complete a questionnaire on their family history of preeclampsia, DM, and complicated pregnancies (17).

Biochemical evaluations A venous blood drawn was taken from each subject to assess hormonal and glucose metabolism patterns, blood count, liver and renal function, and serum glycodelin and IGFBP-1 levels. The blood samples were obtained in the morning between 0800 and 0900 h after a 12-h period of overnight fasting and resting in bed. A complete set of hormonal assays consisting of prolactin, thyroid-stimulating hormone, progesterone (P), 17-hydroxy-P, T, A, DHEAS, and SHBG was performed. The free androgen index (FAI) [T (nanomoles per liter)/SHBG (nanomoles per liter) ⫻ 100]) was calculated in each subject. All plasma hormone concentrations were measured by specific RIA, and SHBG levels were measured using an immunoradiometric assay, whereas the serum concentrations of glycodelin and IGFBP-1 were assayed by a two-site RIA. Glucose and insulin concentrations were also assayed while fasting and after a 2-h oral glucose tolerance test (OGTT). For each patient, the fasting glucose-to-insulin ratio (GIR) (mg/10⫺4 U), the homeostasis model of assessment (HOMA) [fasting glucose (millimoles per liter) ⫻ fasting insulin (picomoles per liter)/22.5] and the ratio of the areas under curve (AUC) for glucose and insulin (AUCglucose to AUCinsulin ratio) were calculated. Overall, intra- and interassay CV were less than 10%.

Ultrasound and Doppler assessments The sonographic study was performed in each subject by the same experienced operator (T.R.), who was formally blinded to the group assignment and any clinical detail. Ultrasonographies were performed by using an ultrasonic scanner equipped with a 7.5-MHz vaginal probe. In each case, the embryo heart beats were recorded to confirm pregnancy viability, and the crown-rump length of each embryo was measured to date the pregnancy. In addition, a careful evaluation of blood flow through the uterine artery was performed as previously detailed (10), and the blood flow impedance of

jcem.endojournals.org

2443

both sides was automatically calculated by the ultrasonographic instrument using at least three clear and similar consecutive waveforms. This information was expressed as the pulsatility index (PI) and the resistance index (RI). For each patient, the mean PI and RI values of the right and left artery were used. In addition, the morphology (presence or absence of a notch, i.e. a definitive upward change in velocity after the initial deceleration slope of the primary wave) of the waveforms was also recorded for each side (10). Subjects were defined as having an abnormal uterine flow at uterine artery in presence of a uni-/bilateral notch and/or a PI value higher than the 95th percentile of reference range for our population at 12th week of gestational age (10). Because all scans were performed by the same operator, only the intraoperator variability for each measurement was calculated. In particular, the intraoperator CV was calculated comparing a set of measurements in 20 women and repeating them in the same conditions after 10 min. The intraoperator CV were 8 and 0% for PI values and notch detection, respectively.

Tissue collection and histological evaluation In no case was any abortifacient drug administered (i.e. prostaglandins and/or mifepristone). In both cases and controls, trophoblast and decidual tissue was collected by the same operators (A.D.C. and G.A.) at pregnancy termination using the suctionaspiration technique. All collected tissues were immediately processed separating chorionic villous from placental tissue and then the decidua from trophoblastic tissue with repeated washes with Hanks’ balanced salts solution 1. The tissue samples were fixed in 4% paraformaldehyde for at least 24 h and subsequently embedded in paraffin. Serial 4-␮m sections were cut and placed in bioboxes. Histological sections were reviewed by an experienced pathologist (L.T.), who was blinded to the group assignment and any clinical detail. At the time of the histological and immunohistochemical evaluations, all paraffin-embedded sections were deparaffinized in xylene and rehydrated through a series of graded alcohol solutions. To identify the implantation site for each sample, staining with hematoxylin-eosin was performed (Fig. 1A). To confirm the presence of an implantation site and to identify endovascular trophoblast invasion, the positive slides on hematoxylin-eosin staining were processed for immunohistochemical analysis. In addition, Masson trichromic staining was performed with the goal of evaluating the exact number of implantation site vessels and vessels with endovascular trophoblast invasion (Fig. 1B). For the immunohistochemical evaluation, the deparaffinized and rehydrated sections were processed as detailed in the following steps: 1) pretreatment (heat induced antigen retrieval with a 750 W microwave oven) in three sequential steps of 4 min each using a citrate buffer [citrate acid monohydrated 10 mM (pH 6.0)]; 2) incubation for 20 min at room temperature with 0.3% hydrogen peroxide in methanol to inhibit the endogenous peroxidases; 3) incubation for 25 min at room temperature with nonimmune horse serum (Dako, Milan, Italy) at dilutions of 1:20 in PBS/BSA (1%) to prevent nonspecific immunostaining; 4) incubation overnight at 4 C with the primary antibodies, i.e. IgG monoclonal mouse antihuman cytokeratin 7 and CD34 antibodies (Dako) at a dilution of 1:500; 5) biotinylated goat antimouse IgG (Beckton-Dickinson, Milan, Italy) at a dilution of

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 11 January 2014. at 03:49 For personal use only. No other uses without permission. . All rights reserved.

2444

Palomba et al.

Decidual Trophoblast Invasion in PCOS

J Clin Endocrinol Metab, July 2012, 97(7):2441–2449

2A) and CD34 (Fig. 2B) (18). In particular, the immunoreactive areas to cytokeratin 7 specifically indicated the vessels with endovascular trophoblast invasion, whereas the immunoreactive areas to CD34 specifically represented all vessels that were in the section (18). The five top slides for each immune-histochemical staining were considered for this analysis. To quantify the immunoreactive areas, an advanced image analysis software (Image-Pro Plus 6.0; Media Cybernetics, Silver FIG. 1. Histological images of trophoblast invasion by hematoxylin-eosin (A) or Masson Spring, MD) was used by a single operator trichromic (B) staining. Asterisk, Trophoblast and implantation site; bold arrow, decidual (R.M.). The images were acquired directly by tissue; arrow, spiral arteries. optic microscopy with ⫻200 enlargement using an advanced digital camera and, after op1:200 was used as secondary antiserum; 6) incubation with avitimization of the optic quality by modifying the brightness and the din-biotin-peroxidase complex (Vector Laboratories, Burlincontrast, were digitally and successively processed. Their extension game, CA) was used to reveal the antigens; 7) nickel-sulfate enwas automatically calculated and expressed as pixels, and the avhancement with diaminobenzidine chromogen in the presence of erage of pixels was calculated for each tissue sample. The propor0.03% H2O2 was used to enzymatically reveal the antigens; and tion (percentage of pixels) between the immunoreactive areas to 8) counterstaining with hematoxylin-eosin. cytokeratin 7 and to CD34 was used as a marker of the extension The five best sections were selected by the pathologist of endovascular trophoblast invasion. (L.T.), and 10 fields per section were analyzed. In each field, The intraoperator CV for the histological evaluations were the presence or absence of implantation site fragments was less than 10%. The CV was evaluated in a prestudy phase at 7-d assessed. Specifically, the implantation sites were identified by intervals comparing a set of measurements in 50 samples from the presence of interstitial extravillous trophoblast surroundvolunteers and repeating them in the same conditions after a ing decidual vessels. All samples (and/or patients) without week. identified implantation sites were excluded from the final analysis. All implantation site decidual vessels were examined and the presence or absence of endovascular trophoblast invasion was determined (9). The total number of implantation site vessels and vessels with endovascular trophoblast invasion was noted for each subject. The percentage of implantation site vessels with endovascular trophoblast invasion was then calculated. The extension of endovascular trophoblast invasion was evaluated calculating the immunoreactive areas to specific antigens, i.e. cytokeratin 7 (Fig.

Statistical analysis Categorical variables were compared using the Pearson’s ␹2 test; a Fisher’s exact test was used for the frequency tables when more than 20% of the expected values were less than five. The normal distribution of continuous variables data were evaluated with the Kolmogrov-Smirnov test. Thus, our data were expressed as the mean ⫾ SD and analyzed using an unpaired Student’s t test. The strength of the association between endovascular trophoblast invasion, expressed both as a percentage of implantation site vessels showing signs of endovascular trophoblast invasion and endovascular trophoblast invasion extension (percentage of pixels), and the patient’s clinical, biochemical, and ultrasonographic characteristics were measured by correlation analysis and expressed using Pearson’s correlation coefficient (r). Statistical significance was set at P ⬍ 0.05, whereas a statistical trend was arbitrarily established for P values between 0.05 and 0.07. We used the Statistical Package for Social Sciences (SPSS 14.0.1; SPSS Inc., Chicago, IL) for all statistical analyses.

Results FIG. 2. Immunohistochemical evaluation of positive cells to cytokeratin 7 antigen in cases (A) and controls (B) or CD34⫹ antigen in cases (C) and controls (D). Brown, Immunoreactive cells to cytokeratin 7 (A and B) and to CD34⫹ (C and D); blue, nonimmunoreactive cells.

One hundred twenty-five women waiting for legal pregnancy termination were consecutively screened, and a total of 57 pregnant women were included in the study. Five and

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 11 January 2014. at 03:49 For personal use only. No other uses without permission. . All rights reserved.

J Clin Endocrinol Metab, July 2012, 97(7):2441–2449

jcem.endojournals.org

2445

TABLE 1. Clinical and biochemical data in cases (PCOS group) and controls (control group) Group Age (yr) BMI (kg/m2) WHR Ferriman-Gallwey score HR (beats/min) SBP (mm Hg) DBP (mm Hg) T (ng/ml) A (ng/ml) DHEAS (ng/ml) SHBG (nmol/liter) FAI (%) Fasting glucose (mg/dl) Fasting insulin (␮U/ml) GIR (mg/10⫺4 U) HOMA OGTT AUCglucose (mg/dl per 120 min) AUCinsulin (␮U/ml per 120 min) AUCglucose to AUCinsulin ratio IGFBP-1 (ng/ml) Glycodelin (ng/ml)

PCOS (n. 15) 28.4 ⫾ 6.2 26.2 ⫾ 7.3 0.85 ⫾ 0.12 10.9 ⫾ 3.6 75.0 ⫾ 15.3 123.0 ⫾ 12.6 80.5 ⫾ 10.3 1.9 ⫾ 0.3 3.6 ⫾ 1.2 2563.9 ⫾ 267.2 22.9 ⫾ 9.1 10.9 ⫾ 4.6 82.0 ⫾ 9.4 11.29 ⫾ 4.3 4.5 ⫾ 2.1 16.6 ⫾ 3.2

Control (n. 30) 28.3 ⫾ 5.8 25.6 ⫾ 6.9 0.74 ⫾ 0.11 3.1 ⫾ 2.2 76.0 ⫾ 14.8 120.5 ⫾ 14.2 80.8 ⫾ 9.6 1.1 ⫾ 0.5 1.9 ⫾ 1.1 1792.5 ⫾ 198.9 39.9 ⫾ 10.2 4.6 ⫾ 2.1 78.1 ⫾ 10.3 5.81 ⫾ 3.6 4.8 ⫾ 1.9 14.2 ⫾ 2.0

P 0.958 0.194 0.002