Rationale and Hypothesis We have previously reported that theca cells (TC) responses to hCG in women with PCOS represent a spectrum where some exhibit exaggerated increases of 17OHP while in others 17OHP responses resemble those of normal women (Maas KH et al, JCEM, 2015). The basis for this differential responsiveness is not clear. Earlier studies reported that 17OHP responses to gonadotropin stimulation were heterogeneous among PCOS women, which was attributed to the degree of hyperinsulinemia (Pasquali R et al, JCEM, 2007). However, assessment of the ovary was omitted in the analysis. In preliminary studies, we have found that in women with PCOS, insulin sensitivity was strongly correlated with insulin sensitivity index as assessed by the method of Matsuda and DeFronzo (Diabetes Care, 1999). However, the study lacked sufficient numbers. Further analysis of insulin sensitivity with respect to hCG stimulated theca cell responses is warranted.

We have also examined 17OHP responses to hCG in relationship to antral follicle count and anti-Müllerian hormone (AMH) in PCOS and normal women. In PCOS women, as expected, serum AMH correlated with antral follicle count. However, TC responses in PCOS were inversely related to AMH (Maas KH et al, JCEM, 2015). These novel observations suggested that in PCOS AMH production may reflect redistribution of the follicle population. In human ovaries maximal immunodetection of AMH is observed in small (< 4 mm) antral follicles followed by a rapid and progressive decline until an absence of the protein by 8 mm (Weenen C et al, Mol Hum Reprod, 2004). This consideration raises the issue of whether normal AMH levels represent more advanced follicle growth in some PCOS women compared with that of others with elevated AMH levels. An increased stage of follicle development would be accompanied by increased TC hyperplasia and may account for greater 17OHP responses to hCG stimulation. A comparison of TC responses to hCG with ovarian morphology has not be done in women with PCOS.

Based on these findings, we hypothesize that in PCOS, heterogeneous TC responses to hCG reflect differences in morphometric development of the follicle population. In addition, the positive correlation between insulin sensitivity and TC responses to hCG suggest an effect of hyperinsulinemia. We propose to investigate the relationship between theca cell responses to hCG, follicle morphology, and insulin sensitivity before and following treatment with an insulin lowering drug, metformin.

50-100 PCOS and 50-100 normal women matched for age (<37 yr) and BMI (<35) will be studied. PCOS and normal women in the mid-follicular phase (day 5-7) of the menstrual cycle will receive an iv injection of hCG as previously described (see Experimental Design). Basal and stimulated blood samples will be assessed for 17OHP, androgens, gonadotropins and AMH. Detailed 3-D ultrasound imaging will be done prior to hCG stimulation (see Experimental Design). Ovarian morphology will include volume, antral follicle number, and the size of each follicle visualized.

Screening 100 women with PCOS and 100 normal women will be studied.

PCOS will be defined according to NIH criteria:

• <8 spontaneous menses per year
• hyperandrogenism clinically (Ferriman-Gallway score≥8 or total testosterone>0.5ng/ml).
• Absence of CAH, hyperprolactinemia, thyroid disorder, Cushings syndrome, hypothalamic anovulation.

Subjects will have a screening visit to establish eligibility and to sign the informed consent. They will have a history and physical examination performed which will include a urine pregnancy test, a transvaginal ultrasound, thyroid studies, a fasting chemistry panel, and a hemoglobin.

Experimental Design Phase 1 (Pretreatment)

1. In all subjects images of the both ovaries will be obtained using 3D technology and the number, size, and spatial arrangement of ovarian follicles will be noted for both ovaries in each subject.
2. Normal subjects will be studied during the mid-follicular phase of the cycle (Days 5-9) while PCOS women will be anovulatory and will not have a cycle day parameter.
3. On study day one, recombinant-hCG (r-hCG) will be administered intravenously at a dose of 25 micrograms.
4. Blood samples will be obtained at 1 hour before (T = -1), immediately prior (T = 0), and 24 hours after (T = +24) iv hCG injection.
5. Sera will be assayed at all time points for steroid and peptide hormones including 17OHP, androstenedione, testosterone, estradiol, progesterone, LH, FSH, and AMH.
6. All subjects will undergo an oral glucose tolerance test. Each will be placed on a 300-gram carbohydrate diet for three days before the study. On the day of testing each subject will ingest 75 gm of a glucose solution and blood samples for glucose and insulin will be obtained at 0, 15, 30, 60, 120 and 180 minutes after the glucose load.

Phase 2 (Posttreatment)

1. Each PCOS subject will be administered metformin, 1500 mg/day for 3 months. All subjects will be encouraged to maintain same dietary and exercise activity throughout the course of study.
2. After 3 months of treatment, all tests as described in Phase 1 will be repeated.

Inclusion Criteria:

• <8 spontaneous menses per year • hyperandrogenism clinically (Ferriman-Gallway score≥8 or total testosterone>0.5ng/ml).

Exclusion Criteria:

• CAH, hyperprolactinemia, thyroid disorder, Cushings syndrome, hypothalamic anovulation