Can highly sensitive antimüllerian hormone testing predict failed response to ovarian stimulation?

Objective: To determine whether a newer commercially available antimüllerian hormone (AMH) enzyme-linked immunosorbent assay (picoAMH ELISA, AnshLabs) with a lower threshold of detection is predictive of successful ovarian stimulation in a population of women with diminished ovarian reserve (DOR).

Design: Retrospective case-control study.

Setting: University-based IVF program.

Patient(s): Cases were patients whose first IVF cycle was cancelled for lack of ovarian response (<3 follicles; n = 24). Controls were patients with DOR (early follicular FSH of ≥10 IU/L), whose first cycle resulted in aspiration of at least 3 oocytes (n = 24).

Intervention(s): Frozen serum samples collected during routine clinical care between 2008 and 2012 before starting IVF were analyzed for AMH using the picoAMH ELISA.

Main Outcome Measure(s): Serum AMH levels in patients who successfully reached oocyte retrieval compared with patients with a failed controlled ovarian hyperstimulation (COH) cycle. Receiver operator curve analysis was used to identify a predictive threshold AMH value.

Result(s): No demographic differences were found between groups. The successful group had a higher antral follicle count (8.5 vs. 6) and higher AMH levels (847 vs. 406 pg/mL). The AMH level correlated with the antral follicle count (R = 0.61). The AMH level of >500 pg/mL had 83.3% sensitivity and 70.8% specificity to detect patients who proceeded to successful oocyte retrieval. Below AMH levels of 100 pg/mL, no patients achieved oocyte retrieval.

Conclusion(s): Due to a lower threshold of detection, picoAMH may be able to predict successful ovarian stimulation among women with DOR using a threshold of 500 pg/mL, with good sensitivity and specificity.