The effects of metformin and rosiglitazone, alone and in combination, on the ovary and endometrium in polycystic ovary syndrome

Abstract

To examine the effects of metformin and rosiglitazone, alone and in combination, on endometrial histology and ovarian steroid production. Randomized open-label study of metformin and rosiglitazone in 16 women with polycystic ovary syndrome (PCOS) performed at a single academic health center. The study consisted of a 6-week baseline observation period, a 3-month treatment period of single-agent therapy (rosiglitazone or metformin), and then a 3-month period of combined therapy. Abnormal endometrial histology was found in 3 subjects at baseline, including 1 case of adenocarcinoma of the endometrium in an asymptomatic subject, who was excluded from further study. The 2 other abnormal cases (simple hyperplasia) resolved with treatment. Three months of single-agent therapy showed a benefit of rosiglitazone (n = 9) over metformin (n = 6) in terms of reducing circulating unbound testosterone levels (-11.8; 95% CI: -21.7 to -2.0 ng/dL) and 2-hour glucose (-42.0; 95% CI: -76.2 to -7.8 mg/dL), 2-hour insulin (-150.4; 95% CI: -272.7 to -28.1 microU/mL) as well as a significant decrease in integrated levels of glucose and insulin by area under the curve analysis, all obtained from oral glucose tolerance testing. Daily urinary progestin-to-estrogen ratios improved on rosiglitazone compared to metformin therapy (0.08; 95% CI: 0.02 to 0.14). Ovulatory rates tended to improve on both single-agent and combined treatments (30/90 cycles, 33%), compared to baseline ovulatory rate (2/15, 13%). Despite 6 months of therapy alone or in combination, 5 women displayed no evidence of biochemical ovulation by urinary or serum progestin measurements. This study provides preliminary evidence that insulin-sensitizing drugs may have beneficial effects on the endometrium, although the exact mechanism beyond improving ovulatory function is still unknown. In addition, we suggest that rosiglitazone may be more beneficial than metformin therapy on raised insulin and androgen levels in an obese PCOS population. Combined therapy did not demonstrate significant benefit above and beyond single-agent therapy.