Abstract
PURPOSE: Our purpose was to determine prospectively, using receiver-operating characteristic (ROC) analysis, whether the ovarian reserve test with hMG could improve the predictive value of a woman's age and basal levels of follicle stimulating hormone (FSH), E2, and inhibin or any combination of them regarding ovarian response and pregnancy rate in IVF treatment following pituitary desensitization. METHODS: The hMG test was performed within 3 months of IVF treatment in 80 women undergoing the first cycle of IVF and consisted of 2 ampoules of hMG daily for 5 days starting on cycle days 2 to 3. Hormone and ultrasound evaluation was performed on cycle days 2 to 3 and 7 to 8. RESULTS: The mean age and basal FSH levels were significantly higher in the canceled (n = 28) than in the control (n = 52) group, whereas the basal inhibin level was significantly higher in the latter. Regarding ovarian response, the combination FSH plus inhibin had the better diagnostic accuracy (predictive value of 70%) among basal variables. When post-hMG parameters (alone or in combination) were analyzed, E2 alone, with a 77% diagnostic accuracy, emerged as the best predictive variable of cancellation in IVF cycles. When ROC analysis was used, the area under the ROC curve for E2 post-hMG (diagnostic accuracy of 84.5%) was significantly higher than that for the estimates based on the combination of basal FSH and inhibin (diagnostic accuracy of 71.3%). However, woman's age was the only variable independently associated with pregnancy rate. CONCLUSIONS: The predictive power of the hMG test of ovarian reserve is better than that of age and basal hormone values (FSH and inhibin) and it is based mainly on the E2 response to hMG treatment. However, given that age is the only predictor of pregnancy and considering the cost and discomfort of the hMG test, the usefulness, if any, of the test in predicting IVF performance in the daily clinical setting remains to be established.