Antral follicle count as a predictor of ovarian responsiveness in women with endometriomas or with a history of surgery for endometriomas

Objective To evaluate the accuracy of antral follicular count (AFC) in predicting ovarian responsiveness in ovaries with endometriomas or with a past history of surgical excision of endometriomas. Design Retrospective review. Setting Academic hospital. Patient(s) Eighty-three women for a total of 166 gonads. Intervention(s) None. Main Outcome Measure(s) Total number of developing follicles. Result(s) The ovaries were characterized as four groups: [1] unoperated gonads without endometriomas (n = 42, control group), [2] unoperated gonads with endometriomas (n = 46), [3] operated gonads without endometriomas (n = 55), and [4] operated gonads with endometriomas (n = 23). The analyses subsequently considered all ovaries with endometriomas (groups 2 + 4, n = 69) and all operated ovaries (groups 3 + 4, n = 78). The capacity of AFC to predict low response (≤2 follicles) or hyperresponsiveness (≥7 follicles) was evaluated using receiver operating characteristic curves. We used a linear regression model to calculate the adjusted B coefficients. The adjusted B coefficients in unaffected ovaries, in all ovaries with endometriomas, and in all operated ovaries were 0.55 (95% confidence interval [CI], 0.07-1.03), 0.76 (95% CI, 0.54-0.98), and 0.51 (95% CI, 0.26-0.76), respectively. The area under the curve (AUC) for the prediction of low response was 0.83 (95% CI, 0.68-0.99), 0.83 (95% CI, 0.73-0.93), and 0.74 (95% CI, 0.63-0.85), respectively. The AUC for the prediction of hyperresponse was 0.84 (95% CI, 0.70-0.97), 0.74 (95% CI, 0.63-0.85), and 0.77 (0.60-0.94), respectively. Conclusion(s) The accuracy of AFC for predicting ovarian response is similar in unaffected ovaries, ovaries with endometriomas and ovaries with a history of surgery for endometriomas.