The diagnostic value of amh levels in the prediction of poor ovarian response in assisted reproductive technique cycles

Poor ovarian reserve reduces the success of the outcome in women undergoing controlled ovarian hyperstimulation. Our aim is to determine the predictive value of AMH concentrations on ovarian response and pregnancy rates after a cycle of controlled ovarian hyperstimulation in patients with poor ovarian responseOur study included 98 primarily infertile patients who applied to our infertility clinic and initiated IVF treatment between October-February 2009. The study was approved by our university ethics council. Patients were identified as poor responders if only 3 or less oocytes were collected after controlled ovarian hyperstimulation, or if the cycle was cancelled due to insufficient ovarian response following treatment. Patients were then divided into subgroups in accordance with their response, and the occurrence of clinical pregnancy and cycle cancellation. Hormonal, sonographic and demographic data were then compared between groupsOur study concluded that, patients demonstrating poor response to ovarian stimulation had lower AMH levels, decreased AFC (antral follicle count), were older and hence had higher bFSH levels. Logistic regression revealed that the best parameter for predicting poor ovarian response was AMH. ROC analysis used to identify poor ovarian response resulted in an AUC value of 0.89 (%95 CI: 0.82-0.97). The predictive threshold value was appointed at 0.94ng/ml with sensitivity of % 84 and specificity of % 85. In patients with confirmed clinical pregnancies, AMH and AFC were deemed significantly high and both of these parameters were also regarded as the best predictors of clinical pregnancy with AUC values of 0.75 (%95 CI: 0.63-0.85) and 0.72 (%95 CI: 0.60-0.80) respectively. AMH and AFC were significantly lower in cancelled cycles; although the one and only effective predictor of cycle cancellation was b FSH.