What are the likely IVF/ICSI outcomes if there is a discrepancy between serum AMH and FSH levels? A multicenter retrospective study.

Objective: The principal outcome was to assess the ovarian response to controlled hyperstimulation during in vitro fertilization (IVF) with or without micro-injection (ICSI) in patients whom ovarian reserve testing revealed a discrepancy between the serum levels of FSH and AMH. The secondary outcome was to determine whether AMH and FSH profiles could predict the IVF/ICSI response.

Study Design: This was a multicenter, retrospective study analysing all controlled ovarian hyperstimulation cycles with attempted fresh embryo transfer(s) carried out during IVF/ICSI treatment and in which the AMH level had been assayed between January 01, 2008 and December 31, 2011. This enabled us to form 2 control groups (NOR, normal ovarian reserve: normal AMH and FSH and DOR, diminished ovarian reserve: diminished AMH, increased FSH) and 2 study groups (DAMH: diminished AMH, normal FSH and NAMH: normal AMH, increased FSH). The principal assessment criterion was quantitative ovarian response to stimulation defined by the mean number of oocytes punctured, the secondary assessment criterion the qualitative response to stimulation defined by the pregnancy rate per cycle.

Results: We were able to analyse 1803 stimulation cycles. The mean number of oocytes punctured was significantly reduced in the DAMH and DOR groups compared to the NAMH and NOR groups (5.2±3.9 and 4.1±3.3 vs. 11.5±7 and 9.5±5.6, respectively [P<0.01]). The pregnancy rate per initiated cycle was significantly reduced in the DAMH and DOR groups compared to the NAMH and NOR groups (20% and 24% vs. 32 and 35%, respectively [P<0.01]). Live birth rates did not differ between the groups however. Multivariate analysis with logistic regression revealed that AMH, FSH and age independently had an effect on the number of oocytes punctured, although the effect exerted by AMH seemed to be preponderant (OR: 2.75: 95%CI [2.39-3.19]). AMH appeared to be the sole factor independently predictive of pregnancy per cycle.

Conclusion: The serum AMH level appears to provide an additional item of discriminatory information, which should not be overlooked. Ovarian reserve work-up should include routine AMH assay.