Examining the Evidence
Q Which drug is best for infertile PCOS patients—clomiphene or metformin?
<huc>A</huc> Metformin wins hands-down in nonobese women. Although the 2 drugs induced ovulation at roughly the same rate, metformin was associated with a higher pregnancy rate, a lower abortion rate, and a higher positive trend for live births.
While metaanalysis suggests metformin improves ovulatory frequency in women with PCOS, until now the question of whether it helps achieve and maintain pregnancy has been explored only in small trials. The superiority of metformin for primary treatment of PCOS-related anovulatory infertility over standard-of-care clomiphene citrate was a matter of speculation (partly because metformin therapy was often reserved for “clomiphene failure”).
Palomba et al are to be commended for their study design (double-dummy, double-blind, randomized controlled trial) and choice of pregnancy as the primary outcome—a quantum leap forward for clinical trials involving PCOS. They studied 100 nonobese (BMI <30) women with PCOS and primary infertility (male and tubal factor excluded) and randomized them in equal groups to metformin (850 mg bid) or clomiphene citrate (150 mg/day for 5 days per treatment cycle). Both groups were monitored with ultrasound for follicular development, ovulation, and pregnancy during 6 months of treatment. Ovulation occurred without human chorionic gonadotropin trigger, with no inseminations.
Twice the pregnancy rate
The pregnancy rate after 6 months was significantly higher in the metformin group (69%) than in the clomiphene group (34%), and the abortion rate was significantly lower with metformin (10% versus 38% for clomiphene). There also was a trend toward a better live birth rate with metformin (84% versus 56% with clomiphene).
Intriguingly, ovulation and fecundity rates improved progressively with metformin and were highest during the sixth month of treatment, whereas an opposite trend was noted with clomiphene.
Flaws may limit credibility
Several imperfections mark this trial. Although it was billed as double-dummy, the dummy used for both clomiphene and metformin was described as “polyvitamin tablets similar in appearance to metformin and/or CC.” A true dummy is identical in appearance to the medication; any suggestion that a medication is inactive will lead to unblinding, potentially biasing the results.
Another problem: 10% of metformin patients and 6% of clomiphene patients were excluded from the analyses, in some cases for vague reasons (eg, significant weight loss). An intention-to-treat analysis including all randomized patients would have been more appropriate, although pregnancy rates would have been lower.
Finally, this comparatively large sample size is not nearly large enough to detect a significant difference in the ultimate pregnancy goal: a live birth.
Metformin best in nonobese women
This study reinforces the use of metformin as first-line therapy for PCOS in nonobese women with anovulatory infertility. It is too soon to extrapolate results to an obese PCOS population, which is more characteristic of the United States.