At present, polycystic ovary syndrome (PCOS) is diagnosed ultrasonographically by an ovarian volume exceeding 10 mL and/or 12 or more follicles measuring 2 to 9 mm in each ovary. Counting follicles may, however, be difficult and unreliable, and it remains uncertain that a count greater than 12 is specific for polycystic ovaries. This study sought to determine whether assaying serum for anti-Müllerian hormone (AMH) can replace the antral follicle count in diagnosing PCOS. AMH was estimated using a second-generation immunoassay in 73 patients diagnosed as having PCOS using the Rotterdam system and in 96 control subjects.
Mean serum AMH levels were 3-fold higher in patients with PCOS than in control subjects. There was a similar difference in the number of follicles per ovary, and these variables correlated significantly with one another in both patients and control subjects. Levels of AMH correlated positively with serum testosterone in both groups, but AMH levels did not correlate significantly with follicle-stimulating hormone levels in the PCOS group. A cutoff value of 60 pmol/L was 92% specific and 67% sensitive for PCOS. This figure is slightly higher than the 90th percentile of control subjects. Both AMH levels and follicle counts, but not age, differed in patients presenting with amenorrhea, oligomenorrhea, or regular cycles. In patients with PCOS with abnormal cycles, AMH levels varied significantly depending on whether or not hyperandrogenism was present, but no such differences were found for follicle counts.
If reliable ultrasound studies are not available, the serum AMH can point to the presence of polycystic ovaries in women with hyperandrogenism and/or oligoanovulation. Conceivably, the level of AMH could indicate the extent of ovarian dysfunction in patients with PCOS and predict the response to attempted induction of ovulation.