Two-dimensional largest ovarian area to predict ovarian response in in vitro fertilization cycle

Objective: To evaluate the usefulness of two dimensional largest ovarian area (LOA) as a predictive parameter of oocyte yield and pregnancy in in vitro fertilization (IVF) cycle. Methods: Ninety-two stimulated IVF cycles from sixty-eight infertile women were included. On the day 3 of cycles, antral follicle count (AFC) and LOA were measured by ultrasonography before starting gonadotropins. LOA was calculated by ellipsoid formula (D1×D2×0.8) using two perpendicular diameters in the largest cross-sectional view of each ovary and then summed. Results: Both LOA and AFC had a positive correlation with the number of retrieved oocytes. The best cut-off value of LOA was 6.14 cm2 for predicting poor ovarian response (area under curve [AUC], 0.783; 95% confidence interval [CI], 0.684 to 0.862), 7.70 cm2 for predicting cycle cancellation (AUC, 0.794; 95% CI, 0.697 to 0.871) and 8.36 cm2 for predicting pregnancy (AUC, 0.641; 95% CI, 0.534 to 0.739). For AFC, the best cut-off value was 9 for poor ovarian response (AUC, 0.877; 95% CI, 0.792 to 0.936), 6 for predicting cycle cancellation (AUC, 0.942; 95% CI, 0.872 to 0.980) and 9 for pregnancy (AUC, 0.683; 95% CI, 0.578 to 0.776). The predictability for pregnancy was similar between LOA and AFC. Conclusion: LOA is easy to measure and a good marker for predicting ovarian response and IVF pregnancy.