[Value of anti-Müllerian hormone in predicting pregnant outcomes of polycystic ovary syndrome patients undergone assisted reproductive technology]

Objective: Explore the value of anti-Müllerian hormone (AMH) in predicting pregnant outcomes of polycystic ovary syndrome (PCOS) patients undergone assisted reproductive technology. Methods: The study totally recruited 1 697 patients who underwent the first in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI) cycle in Sun Yat-sen Memorial Hospital from the January 2014 to December 2015. The patients were divided into two groups based on the age<35 (n=758) and ≥35 years old (n=939) , compare the basic data and pregnant outcomes of controlled ovarian hyerstimulation. Spearman correlation method was conducted to analyze the relations between AMH and clinical outcomes. The logistic regression method and partial correlation analysis were used to judge the main factors which determine pregnancy outcomes by controlled the confounding factors. The receiver operating characteristic curve (ROC) was used to evaluate the predictive sensitivity and specificity of AMH. Results: In the group of PCOS patient younger than 35 years, AMH were correlated with the number of antral follicles (r=0.388) and retrieved oocytes (r=0.235) . When the effect of total dosage and starting dosage of gonadotropin were controlled, AMH was still significantly associated with the number of retrieved oocytes (P<0.05) . AMH had no predictive value for the clinical pregnancy of PCOS patient younger than 35 years (area under ROC curve=0.481, P=0.768) . In the group of PCOS patient≥35 years old, AMH were correlated with the number of antral follicles (r=0.450) , retrieved oocytes (r=0.399) , available embryo (r=0.336) and high quality embryo (r=0.235) . When the effect of total dosage and starting dosage of gonadotropin were controlled, the correlations were still significant between those indexes (all P<0.05) . AMH had no predictive value for the clinical pregnancy of PCOS patient ≥35 years old (area under ROC curve=0.535, P=0.560) . However, the clinical pregnancy rate of the group of PCOS patient ≥35 years old was slightly higher than the control group (P=0.062) . Conclusions: AMH has no predictive value for the pregnancy outcome of PCOS patient. The pregnancy rate of PCOS patient ≥35 years old is slightly higher than the younger group, because the PCOS patient may have better ovarian reserve.