Abstract Study question In patients with polycystic ovarian syndrome (PCOS), are serum androgen levels (total and free testosterone) associated with the risk of endometrial hyperplasia or cancer? Summary answer Patients with PCOS diagnosed with endometrial hyperplasia or cancer had statistically significantly lower serum levels of androgens compared to PCOS patients with a normal endometrium. What is known already PCOS affects as many as 12% of women of reproductive age. Women with PCOS are almost three times as likely to develop endometrial cancer as women without PCOS, and endometrial hyperplasia precedes cancer in 36% of cases. Known risk factors of endometrial hyperplasia include anovulation, obesity, and resultant unopposed estrogen stimulation of the endometrium. A hallmark of PCOS is hyperandrogenism, however the effect of androgens on the endometrium is unclear. Study design, size, duration This was a cross-sectional, retrospective study of patients aged 18-40, being evaluated for oligomenorrhea and/or PCOS in the Reproductive Endocrinology Clinic at Los Angeles County Hospital, California, USA. All patients underwent an endometrial biopsy and serum measurement of free and total testosterone, estradiol, and hemoglobin A1c. Patients diagnosed with PCOS and with an endometrial biopsy diagnosis were included in the study. Patients on any hormonal therapy were excluded. Participants/materials, setting, methods In all patients included in the study, serum androgen levels (free and total testosterone) were measured by mass spectrometry. All endometrial biopsies were analyzed by an experienced histopathology, and all diagnoses of endometrial hyperplasia or cancer were confirmed by histologic studies. Main results and the role of chance Using the Rotterdam criteria, a total of 232 patients were diagnosed with PCOS during the study period. Of these patients, 161 patients (70%) had a normal endometrium, 19 (8%) were diagnosed with endometrial hyperplasia (EH) without atypia, 38 (16%) were diagnosed with EH with atypia, and 14 (6%) were diagnosed with endometrial cancer. Patients with a normal endometrium had a mean total testosterone of 60.9 ng/dL and mean free testosterone of 9.3 ng/dL. Relatively, patients with EH without atypia, EH with atypia, and endometrial cancer had lower mean total testosterone levels of 50.1 ng/dL, 45.5 ng/dL, and 30.1 ng/dL, respectively (p-values 0.04, 0.02, 0.001, respectively), as well as lower mean free testosterone levels of 8.2 pg/mL, 6.7 pg/mL, and 5.5 pg/mL, respectively (p-values 0.04, 0.03, and 0.01, respectively). There was no significant difference in age, parity, BMI, estradiol levels, or hemoglobin A1c between the four groups. Limitations, reasons for caution Limitations of our study include its retrospective nature. Furthermore, hormone levels were only measured in the serum, and not in the endometrial tissue directly, which would give the most insight into their actions on the endometrium. Wider implications of the findings Our findings suggest that androgens have a protective effect on the endometrium against endometrial hyperplasia and cancer. Further studies are needed to evaluate the therapeutic potential role of androgens in the prevention or treatment of these conditions. Trial registration number not applicable
Read full abstract