Combined Effect of the PGR +331C > T, CYP17A1 -34A > G and CYP19A1 1531G > A Polymorphisms on the Risk of Developing Endometriosis

► polymorphisms ► estrogens ► endometriosis ► biomarkers

Purpose To evaluate the magnitude of the association of the polymorphisms of the genes PGR, CYP17A1 and CYP19A1 in the development of endometriosis.

This is a retrospective case-control study involving 161 women with endometriosis (cases) and 179 controls . The polymorphisms were genotyped by real-time polymerase chain reaction usin g the TaqMan system. The association of the polymorphisms with endometriosis was evaluated using the multivariate logistic regression.

The endometriosis patients were signi ficantly younger than the controls (36.0 /C6 7.3 versus 38.0 /C6 8.5 respectively, p ¼ 0.023), and they had a lower body mass index (26.3 /C6 4.8 versus 27.9 /C6 5.7 respectively, p ¼ 0.006), higher average duration of the menstrual flow (7.4 /C6 4.9 versus 6.1 /C6 4.4 days respectively, p ¼ 0.03), and lower average time intervals between menstrual periods (25.2 /C6 9.6 versus 27.5 /C6 11.1 days respectively, p ¼ 0.05). A higher prevalence of symptoms of dysmenorrhea, dyspareunia, chronic pelvic pain, infertility and intestinal or urinary changes was observed in the case group when compared with the control group. The received August 23, 2016 accepted March 23, 2017 published online June 14, 2017 DOI https://doi.org/ 10.1055/s-0037-1604097. ISSN 0100-7203. Copyright © 2017 by Thieme Revinter Publicações Ltda, Rio de Janeiro, Brazil THIEME Original Article 273

Endometriosis is a benign gynecological estrogen-depen- dent disease characterized by the presence of endometrial tissue out of the uterine cavity, affecting nearly 10% of women of reproductive age. Symptoms may include dysme- norrhea, dyspareunia, chronic pelvic pain and infertility. 1 The pathogenesis and the molecular mechanisms that are involved in the development of endometriosis are not yet clear, and hereditary susceptibility is an area of growing investigation for the identification of genetic polymorphisms that may lead to an increased risk of developing the disease. 2,3 Estrogen performs a fundamental role in endometriosis, which predominantly occurs in women of reproductive age who have high estrogen production. 4,5 An increase in en- zyme expression is responsible for the estrogen synthesis and reduction of progesterone receptor (PGR) expression interval between the onset of symptoms and the de finitive diagnosis of endometriosis was 5.2 /C6 6.9 years. When comparing both groups, signi ficant differences were not observed in the allelic and genotypic frequencies of the polymorphisms PGR þ331C > T, CYP17A1 -34A > G and CYP19A1 1531G > A,e v e nw h e nc o n s i d e r i n gt h es y m p - toms, classi fication and stage of the endometriosis. The combined genotype PGR þ331TT/CYP17A1 -34AA /CYP19A11531AA is positively associated with endometriosis (odds ratio [OR] ¼ 1.72; 95% con fidence interval [95%CI] ¼ 1.09–2.72).

The combined analysis of the polymorphisms PGR-CYP17A1-CYP19A1 suggests a gene-gene interaction in the susceptibility to endometriosis. These results may contribute to the identi fication of biomarkers for the diagnosis and/or prognosis of the disease and of possible molecular targets for individualized treatments. Resumo Objetivo Avaliar a magnitude de associação de polimor fismos nos genes PGR, CYP17A1 e CYP19A1 no desenvolvimento da endometriose. Métodos Este é um estudo retrospectivo do tipo caso-controle, envolvendo 161 mulheres com endometriose (casos) e 179 controles. Os polimor fismos foram ge- notipados pela reação em cadeia da polimerase em tempo real utilizando o sistema TaqMan. A associação dos polimor fismos estudados com a endometriose foi avaliada pela regressão logística multivariada. Resultados As pacientes com endometriose eram signi fic a t i v a m e n t em a i sj o v e n sd o que os controles (36,0 /C6 7,3 versus 38,0 /C6 8,5, respectivamente, p ¼ 0,023), apre- sentaram um índice de massa corporal menor (26,3 /C6 4,8 versus 27,9 /C6 5,7, respec- tivamente, p ¼ 0,006), maior tempo médio de duração do fluxo menstrual (7,4 /C6 4,9 versus 6,1 /C6 4,4 dias, respectivamente, p ¼ 0,03) e menor tempo médio do intervalo entre as menstruações (25,2 /C6 9,6 versus 27,5 /C6 11,1 dias, respectivamente, p ¼ 0,05). Uma maior prevalência dos sintomas de dismenorreia, dispareunia, dor pélvica crônica, infertilidade, alterações i ntestinais e urinárias foi observada no grupo casos comparado ao grupo controle. O tempo médio entre o início dos sintomas e o diagnóstico de finitivo de endometriose foi de 5,2 /C6 6,9 anos. Comparando os dois grupos, não foram observadas diferenças signi fic a t i v a sn a sf r e q u ê n c i a sa l é l i c a se genotípicas dos polimor fismos PGR þ331C > T, CYP17A1 -34A > G e CYP19A1 1531G > A, e nem considerando os sintomas, a classi ficação e o estadiamento da endo- metriose. O genótipo combinado PGR þ331TT/CYP17A1 -34AA /CYP19A11531AA está associado positivamente com a endometriose (razão de possibilidades [RP] ¼ 1,72; intervalo de con fiança de 95% [IC95%] ¼ 1,09 –2,72). Conclusões A análise combinada dos polimor fismos PGR-CYP17A1-CYP19A1 sugere uma interação gene-gene na susceptibilidade à endometriose. Estes resultados podem contribuir para a identi ficação de biomarcadores para o diagnóstico e/ou prognóstico da doença, assim como de possíveis alvos moleculares para um tratamento individualizado. Palavras-chave ► polimor fismos ► estrógenos ► endometriose ► biomarcadores Rev Bras Ginecol Obstet Vol. 39 No. 6/2017 Combined Effect of Gene Polymorphisms on the Risk of Developing Endometriosis Cardoso et al.274 observed in samples of endometrial injuries, but these phenomena are not found in controls. 6,7 As endometriosis is an estrogen-dependent disease, genetic polymorphisms involved in the biosynthesis and regulation of estrogens could be considered possible biomarkers for its diagnosis and/or prognosis. Cytochrome P450 17A1 (CYP17A1) is involved in the initial stages of estrogen synthesis, convert- ing pregnenolone into 17 α-hydroxypregnenolone and, sub- sequently, into dehydroepiandrosterone. Furthermore, cytochrome P450 19A1 (CYP19A1) acts in the final stage by converting androstenedione into estrone, and testoster- one into estradiol. 4,8 T h ee n z y m eC Y P 1 7 A 1 ,a l s ok n o w na s1 7 α-hydroxylase, is encoded by the gene with the same name, located in chromosome 10q24.3. 8 The single nucleotide polymorphism (SNP) CYP17A1 -34A > G is located in the 5 ′ untranslated region (UTR) of the CYP17A1 gene, and it causes a significant increase in the expression of 17 α-hydroxylase.8 A different gene product, the aromatase enzyme, is encoded by the gene CYP19A1, which is located in chromosome 15q21. The SNP CYP19A1 1531G > A is found in the 3 ′UTR of this gene, and it causes a signi ficant change in the levels of circulating estradiol. 8 Progesterone is also involved in the pathogenesis of endometriosis, as it is a strong antagonist of estrogen, and thus plays an essential role in the regulation of endometrial cell proliferation. 6 The PGR gene, located in chromosome 11q22 – q23, is responsible for encoding both progesterone receptor isoforms (PR-A and PR-B) by tran- scribing from alternative promoters. 9 The SNP PGR þ331C/T, located in the gene promoter region, creates an additional TATA box, and causes higher transcription of PR-B. 9 As the SNPs PGR þ331C > T, CYP17A1 -34A > G and CYP19A1 1531G > A are located near potential elements that regulate their respective genes, interfering with the levels of expression of the corresponding proteins, it be- comes relevant to evaluate the in fluence of these SNPs in the development of endometriosis. To date, 12 studies have evaluated the association of the SNPs PGR þ331C > T, CYP17A1 -34A > G and CYP19A1 1531G > A in the develop- ment of endometriosis in different populations. However, the results of the analyses are controversial. 10–21 In this context, the objective of this study was to evaluate the magnitude of the association of the SNPs PGR þ331C > T, CYP17A1 -34A > G and CYP19A1 1531G > A with the devel- opment of endometriosis in women treated in two public

hospitals in Brazil.

Study Design This was a retrospective, case-control study approved by the Human Research Ethics Committees of two of our institu- tions (under protocols number 414/11 and 1.244.29 respec - tively), both located in the city of Rio de Janeiro, Brazil. All participating patients ( n ¼ 340) provided written informed consent, and visited one of the two institutions between March 2011 and October 2015. The research was conducted in accordance with the Declaration of Helsinki, which was revised in 2008. Patients with a surgical diagnosis (after laparoscopy or laparotomy) of endometriosis with histologi- cal confirmation of the disease, as well as those diagnosed by magnetic resonance imaging (MRI), were considered the case group ( n ¼ 161). The control group ( n ¼ 179) consisted of women with a negative diagnosis of endometriosis, after laparoscopy or laparotomy for tubal ligation ( n ¼ 48) or for the treatment of benign diseases, such as myoma ( n ¼ 52), ovarian cysts ( n ¼ 30), hydrosalpinx ( n ¼ 5), or for other reasons ( n ¼ 44). Women with any history or diagnosis of cancer or adenomyosis were excluded. 2 The stage of the endometriosis was determined according to the revised American Fertility Society classi fication, which divides the disease into four stages: I (minimum), II (mild), III (moderate) and IV (severe). 22 Regarding the classi fication of the endometriosis, we considered the proposal of Nisolle and Donnez:23 superficial endometriosis (SUP), ovarian endome- trioma (OMA), and deep in filtrative endometriosis (DIE). Superficial endometriosis and ovarian endometrioma may be found in association with deep endometriosis, 24 and the cases in which this association was observed were consid- ered DIE. The body mass index (BMI) was calculated as the weight (kg) divided by the height squared (m 2). Only severe and incapaci- tating symptoms of pain were included. Women who failed to conceive after one year of regular, contraceptive-free inter- course were considered infertile. Cyclical intestinal or urinary symptoms were defined as bowel and/or urinary pain and/or bleeding coinciding with menstrual periods. Genotyping Genomic DNA was extracted from the peripheral blood sample using a genomic DNA extraction kit (Genomic DNA Extraction, Real Biotech Corporation, Banqiao City, Taiwan), according to the manufacturer ’ s instructions. Genotyping ofPGR þ331C > T (rs10895068),CYP17A1 -34A > G (rs743572) and CYP19A1 1531G > A (rs10046) SNPs was performed by real-time polymerase chain reaction (PCR) using the TaqMan system. Oligonucleotides and probes speci fict o each SNP were obtained from Applied Biosystems: rs10895068 (C_27858738_10), rs743572 (C_2852784_30) and rs10046 (C_8234731_30). For all SNPs, PCRs were per- formed with 30 ng of template DNA, 1 /C2 TaqMan Universal Master Mix (Applied Biosystems, Foster City, CA, US), and with each primer and probe assay at 1 /C2 dilution, and H2O to 8 μL. The PCR conditions were: 95°C for 10 minutes, followed by 40 cycles of denaturation at 92°C for 15 seconds, and annealing at 60°C for 1 minute. Allele-detection was performed on a 7500 Real-Time System (Applied Biosystems, Foster City, CA, US), and the genotypes were then determined directly. Statistical Analyses The continuous variables were expressed as the mean /C6 standard deviation (SD), and the differences between means were evaluated using the Student ’ s t-test. The cate- gorical data were expressed as percentages, and evaluated by the Chi-square ( χ2) test or Fisher ’ s exact test, when applicable. For each SNP, the Hardy-Weinberg equilibrium (HWE) was calculated, and the allelic and genotypic Rev Bras Ginecol Obstet Vol. 39 No. 6/2017 Combined Effect of Gene Polymorphisms on the Risk of Developing Endometriosis Cardoso et al. 275 Table 1 Demographic and clinical characteristics of the study population ( N ¼ 340) Variable Controls (N ¼ 179) Cases (N ¼ 161) p/C3 Age (years) n (%) 18–29 30 (16.8) 29 (18.0) 0.011 30–39 60 (33.5) 79 (49.1) /C21 40 86 (48.0) 48 (29.8) No information 3 (1.7) 5 (3.1) Marital status Married/partner 96 (53.6) 111 (68.9) 0.046 Single 53 (29.6) 38 (23.7) Divorced/Widow 5 (2.8) 1 (0.6) No information 25 (14.0) 11 (6.8) Level of Schooling Elementary education 38 (21.3) 24 (14.9) < 0.001 High school 89 (49.7) 60 (37.3) Higher education 26 (14.5) 67 (41.6) No information 26 (14.5) 10 (6.2) BMI 40 13 (7.3) 8 (5.0) Infertility Primary 19 (10.6) 53 (32.9) < 0.001 Secondary 3 (1.7) 16 (9.9) None /C3/C3 133 (74.3) 58 (36.1) No attempt 20 (11.2) 32 (19.9) No information 4 (2.2) 2 (1.2) Parity 0 19 (10.6) 53 (32.8) < 0.001 1 23 (12.8) 37 (23.0) 2 58 (32.5) 27 (16.8) 3o rm o r e 5 5( 3 0 . 7 ) 1 0( 6 . 2 ) No attempt 20 (11.2) 32 (20.0) No information 4 (2.2) 2 (1.2) Symptoms /C3/C3/C3 Dysmenorrhea 29 (16.1) 77 (47.5) < 0.001 Chronic pelvic pain 70 (38.9) 124 (76.5) Dyspareunia 50 (27.8) 102 (63.0) Cyclical urinary complaints /C3/C3/C3/C3 9 (6.4) 41 (27.5) Cyclical intestinal complaints /C3/C3/C3/C3 8 (6.0) 74 (49.7) Abbreviation: BMI, body mass index. Notes:/C3 p-value obtained by Pearson ’sC h i - s q u a r e d(χ2) test. /C3/C3 Number of fertile women. /C3/C3/C3 The same woman can have more than one symptom. /C3/C3/C3/C3 Pain or bleeding during the menstrual period. Rev Bras Ginecol Obstet Vol. 39 No. 6/2017 Combined Effect of Gene Polymorphisms on the Risk of Developing Endometriosis Cardoso et al.276 distributions were compared between cases and controls by the χ2 test or Fisher ’ s exact test. To evaluate the association between the SNPs and the development of endometriosis, as well as the presence of symptoms, the classi fication and staging of the disease were used to estimate the odds ratios (ORs) and their respective 95% con fidence intervals (95% CIs), with adjustment for possible confounding factors, using a multivariate logistic regression. Values of p < 0.05 were considered statistically signi ficant. All analyses were performed using the Statistical Package for the Social Sciences (SPSS, IBM Corp., Armonk, NY, US) software, version 20.0.

The endometriosis cases were diagnosed through laparosco- py ( n ¼ 90, 55.9%), laparotomy ( n ¼ 27, 16.8%) or MRI (n ¼ 44, 27.3%), and 87 (54%) were classi fied as DIE. Of the 117 surgery cases, 37 (31.6%) presented stages I-II diseases, and 80 (68.4%) presented stages III-IV. The most common locations of the endometriotic lesions were in the ovary (31%), followed by the intestine (19%), and the uterosacral ligaments (15%). The average age of the endometriosis patients at the time of diagnosis was 31.5 /C6 7.5, and the average time for disease diagnosis, which was the time since the beginning of the symptoms until the de finitive diagnosis, was 5.2 /C6 6.9 years. In ►Table 1 , the demographic and clinical data from the study population are described. The patients with endometri- osis were significantly younger than the controls (36.0 /C6 7.3 versus 38.0 /C6 8.5 respectively, p ¼ 0.023), and presented a higher education level (41.6% versus 14.5%), as well as a lower BMI (26.3 /C6 4.8 versus 27.9 /C6 5.7 respectively, p ¼ 0.006). The number of infertile women (primary or secondary) was significantly higher in the case group (42.9%) than in the control group (12.3%), and nearly 63% of the women from the control group had 2 or more children. The patients with endometriosis presented a higher prevalence ( p < 0.001) of symptoms, including dysmenorrhea, dyspareunia, chronic pelvic pain, and urinary and intestinal changes. Considering the characteristics of the menstrual cycle, a signi ficant differ- ence was detected between cases and controls in relation to the average duration of the menstrual flow (7.4 /C6 4.9 versus 6.1 /C6 4.4 days respectively,p ¼ 0.03) and the average interval between menstrual periods (25.2 /C6 9.6 versus 27.5 /C6 11.1 days respectively, p ¼ 0.05). When comparing the allelic and genotypic frequencies of the SNPs PGR þ331C > T, CYP17A1 -34A > G and CYP19A1 G > A between the cases and controls ( ►Fig. 1 ), no signi fi- cant differences were detected, even when considering the staging and endometriosis classi fication (data not shown). The allelic distribution of the SNPs PGR þ331C > T, CYP17A1 -34A > G and CYP19A1 G > A in relation to the absence or presence of symptoms (dysmenorrhea, pelvic pain, dyspar- eunia, infertility, and urinary and intestinal problems, for example) in the endometriosis patients is summarized in ►Fig. 2 . Considering the symptoms of the disease, no significant differences were found. A combined analysis of the three studied SNPs ( PGR þ331C > T, CYP17A1 -34A > G and CYP19A1 G > A), com- pared between the endometriosis cases and controls, was performed to investigate whether the presence of more than 1 SNP would increase the risk of developing the disease ( ►Table 2 ). It has been observed that relative to the com- bined wild-type genotype ( PGR þ331CC/CYP17A1 -34AA / CYP19A11531GG), the combined genotype PGR þ331TT/ CYP17A1 -34AA /CYP19A11531AA is associated with an in- creased risk of developing endometriosis. A combined anal- ysis of the PGR þ331C > T, CYP17A1 -34A > G and CYP19A1 G > A genotypes was also performed in relation to the absence or presence of symptoms (dysmenorrhea, pelvic pain, dyspareunia, infertility, and urinary and intestinal problems, for example) in the endometriosis patients. How- ever, no significant differences were found (data not shown). In ►Table 3 , we describe the variant allele frequencies of the SNPs PGR þ331 T , CYP17A1 -34 G and CYP19A1A in Fig. 1 Allelic and genotypic frequencies of the polymorphisms PGR þ331C > T, CYP17A1 -34A > G and CYP19A1 G > A in the study population. Rev Bras Ginecol Obstet Vol. 39 No. 6/2017 Combined Effect of Gene Polymorphisms on the Risk of Developing Endometriosis Cardoso et al. 277 different populations. The allele frequency PGR þ331 T varied between 2% and 10% and 5% and 10% in the cases and controls respectively, based on 3 studies that have evaluated this SNP, including the present one. The SNP rs743572 has been evaluated in 7 studies, in addition to the present study, and the frequency of the CYP17A1 -34 G allele varied between 35% and 58% in the cases and between 31% and 63% in the controls. In addition to our study, 4 other studies also evaluated the SNP rs10046, and the allele frequency of CYP19A1 1531A varied between 35% and 58% and between 41% and 60% in the cases and controls respectively.

Endometriosis is a complex multifactorial gynecological disease caused by the combination of hormonal, genetic and environmental factors, as well as immunological pro- cesses. Estrogen and progesterone are essential in the regu- lation of endometrial tissue growth, and, as such, they may play a central role in the pathogenesis of endometriosis. 5,6 In this study, a positive association between the combined genotype PGR þ331TT/CYP17A1 -34AA/CYP19A1 1531AA and the development of endometriosis was observed. Supporting our results, neither Lamp et al, 18 in Estonia, nor Trabert et al, 19 in the United States, could find an association with only the PGR þ331T allele. However, van Kaam et al,15 in the Netherlands, observed a protective effect (OR ¼ 0.22; 95%CI ¼ 0.06– 0.77) for the development of en- dometriosis. In agreement with our findings, 5 studies from China,13 Japan,10 Turkey,17 Italy16 and the United States 19 failed to observe an association between the SNP CYP17A1 -34G and endometriosis. However, 2 studies in the Chinese population11,12 found a positive association between endo- metriosis and the CYP17A1 -34A allele ( p ¼ 0.046 and p ¼ 0.009). With relation to the SNP CYP19A1 1531G > A, none of the studies found an association with endometriosis, Fig. 2 Minor allelic frequencies of the polymorphisms studied among symptomatic and asymptomatic endometriosis patients. Rev Bras Ginecol Obstet Vol. 39 No. 6/2017 Combined Effect of Gene Polymorphisms on the Risk of Developing Endometriosis Cardoso et al.278 Table 2 Combined genotype frequencies of the polymorphisms PGR þ331C > T, CYP17 -34A > G and CYP19 1531G > A between controls and cases, and their association with the risk of endometriosis Genotypes Controls N( % ) Cases N( % ) p /C3 OR (95%CI) /C3/C3 PGR þ331C > T, CYP17 -34A > G and CYP19 1531G > A WT / WT / WT 23 (13.5) 11 (7.3) 1 /C3/C3/C3 WT / WT / VAR 29 (17.1) 28 (18.6) 0.20 1.82 (0.73 –4.57) WT / VAR / VAR 70 (41.2) 60 (40.0) 0.17 1.35 (0.88 –2.05) VAR / WT / WT 2 (1.2) 4 (2.7) 0.18 1.57 (0.82 –3.00) VAR / VAR / WT 7 (4.1) 10 (6.7) 0.20 1.23 (0.90 –1.67) VAR / WT / VAR 1 (0.6) 7 (4.7) 0.02 1.72 (1.09 –2.72) WT / VAR / WT 25 (14.7) 22 (14.7) 0.15 1.13 (0.96 –1.34) VAR / VAR / VAR 13 (7.6) 8 (5.3) 0.87 1.02 (0.85 –1.21) Abbreviations: 95%CI, 95% con fidence interval; OR, odds ratio. Notes: WT/WT/WT, CC/AA/GG ;W T / W T / V A R ,CC/AA/GA or CC/AA/AA ; WT/VAR/VAR, CC/GG/AA or CC/AG/GA or CC/AG/AA or CC/GG/GA;V A R / W T / W T ,CT/ AA/GG or TT/AA/GG ; VAR/VAR/WT, CT/AG/GG or CT/GG/GG or TT/AG/GG or TT/GG/GG ;V A R / W T / V A R ,CT/AA/GA or CT/AA/AA or TT/AA/GA or TT/AA/AA ; WT/VAR/WT, CC/GG/GG or CC/AG/GG; VAR/VAR/VAR, TT/GG/AA ; /C3 p-value obtained through the Chi-squared test (Pearson P-value) or Fisher ’se x a c t test. /C3/C3 Adjusted by age and BMI. /C3/C3/C3 Reference group. Table 3 Frequency of the studied polymorphisms among different populations (endometriosis patients and controls) Polymorphism Population N /C3 Frequency among the controls (%) Frequency among the cases (%) Association with endometriosis

PGR þ331 C > T (rs10895068) PGR þ331 T Estonia 349 8.8 6.7 No association Lamp et al 18 United States 823 4.8 5.1 No association Trabert et al 19 Netherlands 165 9.7 2.3 Protective (T allele) van Kaam et al 15 Brazil 179 6.6 10.1 No association Present study CYP17A1 -34 A > G (rs743572) CYP17A1 -34 G Turkey 93 30.8 57.6 No association Bozdag et al 17 China 247 60.2 50.8 Risk (A allele) Hsieh et al 11 China 227 62.9 50.8 Risk (A allele) Hsieh et al 12 China 510 54.3 58.2 No association Juo et al 13 Japan 317 39.8 42.9 No association Kado et al 10 United States 823 39.3 34.8 No association Trabert et al 19 Italy 190 37.2 42.3 No association Vietri et al 16 Brazil 180 42.7 44.4 No association Present study CYP19A1 1531 G > A (rs10046) CYP19A1 1531 A Korea 412 55.9 55.8 No association Hur et al 14 Estonia 349 59.8 55.0 No association Lamp et al 18 China 371 56.4 57.5 No association Wang et al 20 China 202 41.0 35.3 No association Yang et al 21 Brazil 180 39.4 40.6 No association Present study Abbreviations: CYP17A1, cytochrome P450 17A1; CYP19A1, cytochrome P450 19A1; PGR, progesterone receptor. Note: /C3 N total number of individuals included in the study (cases þ controls). Rev Bras Ginecol Obstet Vol. 39 No. 6/2017 Combined Effect of Gene Polymorphisms on the Risk of Developing Endometriosis Cardoso et al. 279 which is similar to our results. 14,18,20,21 To date, no study has investigated the combined effect of the SNPs PGR þ331TT/ CYP17A1 -34AA/CYP19A1 1531AA on the development of endometriosis. The strength of the present study is that it is the first study performed in the Brazilian population that evaluated the SNPs PGR þ331C > T, CYP17A1 -34A > G and CYP19A1 1531G > A in terms of endometriosis development, while consider- ing the symptoms of the disease. All control patients were surgically evaluated to con firm a negative diagnosis of endometriosis; 27% of them had previously undergone ster- ilization, and 63% had already had 2 or more children. A

of this study was that the controls also included women with other gynecological diseases, providing lower risk estimations. Furthermore, considering the endometri- osis patients, 27% were diagnosed by MRI. However, this has been a specific and accurate method for the detection of deep endometriosis. 25–28 In conclusion, the combined analysis of the polymor- phisms PGR-CYP17A1-CYP19A1 suggests a gene-gene inter- action in the susceptibility to endometriosis. The results may contribute to the identi fication of genetic biomarkers that are able to help in disease diagnosis and/or prognosis, as well as in the identi fication of possible molecular targets for individualized treatments. Conflicts of Interest The authors have no con flicts of interest to disclose. Acknowledgment The authors would like to thank Lucas Rafael Lopes and Caroline Passos from Centro Universitário Estadual da Zona Oeste, Rio de Janeiro, Brazil, for their technical assistance. This study was supported by the Brazilian agencies Conselho Nacional de Desenvolvimento Cientí- fico e Tecnológico (CNPq), Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) and Fundação Ary Frauzino - Oncobiologia.

1 Acién P, Velasco I. Endometriosis: a disease that remains enig- matic. ISRN Obstet Gynecol 2013;2013:242149 2 Perini JA, Cardoso JV, Berardo PT, et al. Role of vascular endothelial growth factor polymorphisms (-2578C > A, -460 T > C, -1154G > A, þ405G > Ca n d þ936C > T) in endometriosis: a case-con- trol study with Brazilians. BMC Womens Health 2014;14:117 3 Zondervan KT, Rahmioglu N, Morris AP, et al. Beyond endome- triosis genome-wide association study: from genomics to phe- nomics to the patient. Semin Reprod Med 2016;34(04):242 – 254 4 Bell DW, Brannigan BW, Matsuo K, et al. Increased prevalence of EGFR-mutant lung cancer in women and in East Asian popula- tions: analysis of estrogen-related polymorphisms. Clin Cancer Res 2008;14(13):4079 – 4084 5 Barcelos ID, Donabella FC, Ribas CP, et al. Down-regulation of the CYP19A1 gene in cumulus cells of infertile women with endome- triosis. Reprod Biomed Online 2015;30(05):532 – 541 6 Bedaiwy MA, Dahoud W, Skomorovska-Prokvolit Y, et al. Abun- dance and Localization of Progesterone Receptor Isoforms in Endometrium in Women With and Without Endometriosis and in Peritoneal and Ovarian Endometriotic Implants. Reprod Sci 2015;22(09):1153– 1161 7 Lafay Pillet MC, Schneider A, Borghese B, et al. Deep in filtrating endometriosis is associated with markedly lower body mass index: a 476 case-control study. Hum Reprod 2012;27(01): 265– 272 8 Ghisari M, Eiberg H, Long M, Bonefeld-Jørgensen EC. Polymorph- isms in phase I and phase II genes and breast cancer risk and relations to persistent organic pollutant exposure: a case-control study in Inuit women. Environ Health 2014;13(01):19 9 De Vivo I, Huggins GS, Hankinson SE, et al. A functional poly- morphism in the promoter of the progesterone receptor gene associated with endometrial cancer risk. Proc Natl Acad Sci U S A 2002;99(19):12263– 12268 10 Kado N, Kitawaki J, Obayashi H, et al. Association of the CYP17 gene and CYP19 gene polymorphisms with risk of endometriosis in Japanese women. Hum Reprod 2002;17(04):897 – 902 11 Hsieh YY, Chang CC, Tsai FJ, Lin CC, Tsai CH. Cytochrome P450c17alpha 5 ′-untranslated region /C3 T/C polymorphism in en- dometriosis. J Genet 2004;83(02):189 – 192 12 Hsieh YY, Chang CC, Tsai FJ, Lin CC, Tsai CH. Estrogen receptor alpha dinucleotide repeat and cytochrome P450c17alpha gene polymorphisms are associated with susceptibility to endome- triosis. Fertil Steril 2005;83(03):567 – 572 13 Juo SH, Wang TN, Lee JN, Wu MT, Long CY, Tsai EM. CYP17, CYP1A1 and COMT polymorphisms and the risk of adenomyosis and endometriosis in Taiwanese women. Hum Reprod 2006;21(06): 1498– 1502 14 Hur SE, Lee S, Lee JY, Moon HS, Kim HL, Chung HW. Polymorph- isms and haplotypes of the gene encoding the estrogen-metabo- lizing CYP19 gene in Korean women: no association with advanced-stage endometriosis. J Hum Genet 2007;52(09): 703– 711 15 van Kaam KJ, Romano A, Schouten JP, Dunselman GA, Groothuis PG. Progesterone receptor polymorphism þ331G/A is associated with a decreased risk of deep in filtrating endometriosis. Hum Reprod 2007;22(01):129 – 135 16 Vietri MT, Ciof fi M, Sessa M, et al. CYP17 and CYP19 gene polymorphisms in women affected with endometriosis. Fertil Steril 2009;92(05):1532 – 1535 17 Bozdag G, Alp A, Saribas Z, Tuncer S, Aksu T, Gurgan T. CYP17 and CYP2C19 gene polymorphisms in patients with endometriosis. Reprod Biomed Online 2010;20(02):286 – 290 18 Lamp M, Peters M, Reinmaa E, et al. Polymorphisms in ESR1, ESR2 and HSD17B1 genes are associated with fertility status in en- dometriosis. Gynecol Endocrinol 2011;27(06):425 – 433 19 Trabert B, Schwartz SM, Peters U, et al. Genetic variation in the sex hormone metabolic pathway and endometriosis risk: an evaluation of candidate genes. Fertil Steril 2011;96(06): 1401 – 1406.e3 20 Wang L, Lu X, Wang D, et al. CYP19 gene variant confers suscept- ibility to endometriosis-associated infertility in Chinese women. Exp Mol Med 2014;46:e103 21 Yang X, Chen SQ, Liu M. [Association of the CYP19 gene poly- morphism with genetic susceptibility to endometriosis]. Zhon- ghua Yi Xue Yi Chuan Xue Za Zhi 2010;27(06):692 – 696 22 Revised American Society for Reproductive Medicine classi fica- tion of endometriosis: 1996. Fertil Steril 1997;67(05):817 – 821 23 Nisolle M, Donnez J. Peritoneal endometriosis, ovarian endome- triosis, and adenomyotic nodules of the rectovaginal septum are three different entities. Fertil Steril 1997;68(04):585 – 596 24 Chapron C, Santulli P, de Ziegler D, et al. Ovarian endometrioma: severe pelvic pain is associated with deeply in filtrating endome- triosis. Hum Reprod 2012;27(03):702 – 711 Rev Bras Ginecol Obstet Vol. 39 No. 6/2017 Combined Effect of Gene Polymorphisms on the Risk of Developing Endometriosis Cardoso et al.280 25 Bianek-Bodzak A, Szurowska E, Sawicki S, Liro M. The importance and perspective of magnetic resonance imaging in the evaluation of endometriosis. BioMed Res Int 2013;2013:436589 26 Ma K, Majumder K, Clayton R, Rajashanker B, Ed-Osagie E. Correlation between magnetic resonance imaging results and findings at surgery for cases of severe endometriosis. J Minim Invasive Gynecol 2015;22(6S):S55 27 Barcellos MB, Lasmar B, Lasmar R. Agreement between the preoperative findings and the operative diagnosis in patients with deep endometriosis. Arch Gynecol Obstet 2016;293(04): 845– 850 28 Ito TE, Abi Khalil ED, Taffel M, Moawad GN. Magnetic resonance imaging correlation to intraoperative findings of deeply in filtra- tive endometriosis. Fertil Steril 2017;107(02):e11 – e12 Rev Bras Ginecol Obstet Vol. 39 No. 6/2017 Combined Effect of Gene Polymorphisms on the Risk of Developing Endometriosis Cardoso et al. 281