Clinical applications of anti-mullerian hormone measurement in reproductive medicine

Abstract

Anti-Mullerian hormone (AMH) is a hormone which is exclusively produced in ovarian granulosa cells in adult women and is measurable in the blood circulation. It has been introduced as a marker of ovarian reserve and function. Serum AMH has good correlation with antral follicle count (AFC). In the past years, new commercial assays for AMH have been evolving. We compared the Gen II ELISA assay with two older ELISA kits as well as three newer assay methods including two on automated platforms. Good correlations were demonstrated among them, and yet differences in calibration did exist. It is reasonable to consider moving onto the new automated assays for better convenience and assay precision as compared with the manual ELISA methods.

Recognising that serum AMH level changes with age of women, we established a set of year-by-year age-specific reference ranges of serum AMH levels in healthy Chinese women measured by the automated Access AMH assay to guide interpretation of AMH levels in individual women in clinical practice.

We carried out a retrospective study which confirmed that serum AMH concentration was raised in women with World Health Organisation group 2 anovulatory disorder (including polycystic ovary syndrome) and decreased to very low levels in women with premature ovarian insufficiency, while being unchanged in hypogonadotrophic hypogonadism and hyperprolactinaemia compared with ovulatory control subjects. Serum AMH had good discriminatory performance between hypogonadotrophic hypogonadism and polycystic ovary syndrome. We also validated for the first time the use of age-specific multiples of the median (MoM) AMH derived from our local reference ranges instead of a “one for all ages” cut-off for the diagnosis of polycystic ovary syndrome. Our results suggested that an MoM AMH of 1.5 or above was a promising substitute for antral follicle count in the diagnosis of polycystic ovary syndrome.

We showed for the first time that serum AMH level was significantly higher in subjects with a live birth in the first cycle or cumulatively after three cycles of ovarian stimulation and intrauterine insemination treatment compared to those failing treatment. Serum AMH level had only modest predictive performance on pregnancy or ovarian over-response. In in vitro fertilisation treatment, our study confirmed that AMH and AFC only had modest performance in predicting occurrence of cumulative live birth and may not give much additional value on top of the women’s age. We further compared gonadotrophin dosing based on either serum AMH level or AFC in a randomised controlled trial, and demonstrated that both algorithms resulted in no significant difference in the proportion achieving desired ovarian response. We also studied cases where AMH and AFC were discordant, which revealed that those having higher AMH within the same AFC quartile had higher oocyte yield and cumulative live birth rate, and the ovarian responsiveness was intermediate between that when both are concordant on either end. Hence, both AMH and AFC can be utilised for individualisation of stimulation regimen; when the AMH and AFC fall into discordant categories, it would be sensible to adopt an intermediate gonadotrophin dose.