"Co-Flare" stimulation in the poor responder patient: predictive value of the flare response.

PURPOSE: In this study we review our experience with a day 2 start, "Co-Flare" protocol analyzing the flare response as a predictor of outcome in patients with a history of a poor response. METHODS: This study was conducted at a university based IVF Center. A total of 564 patient cycles over a 2.5-year time frame in patients that had either elevated FSH levels or a previous poor response to conventional leuprolide acetate down-regulated stimulation were retrospectively analyzed. These patients were treated with our co-flare protocol (1.0 mg of leuprolide acetate administered on day 2 and decreased on day 5 to 0.5 mg; gonadotropins initiated on day 3). We analyzed the initial flare response and the outcome for these patients. RESULTS: A total of 564 patients attempted to initiate an IVF cycle. The patients mean age was 39.4+/-3.7 years. Of the 450 cycles that started, the outcomes were as follows: 24% (108/450) cancellation, 20.4% (92/450) clinical pregnancy per initiated cycle, and a 14.0% delivery rate per initiated cycle. Patients with an estradiol flare that doubled were less likely to have cycles cancelled (13.6% vs. 35.6%, P < 0.01), but no differences were noted in pregnancy outcome if the cycle was not cancelled once the patient made it to retrieval. CONCLUSIONS: We have demonstrated an overall 14.0% delivery rate per initiated cycle in these "poor prognosis" patients. While the initial flare response (as indicated by a doubling of the estradiol by the second day of stimulation) was indicative of a better stimulation, no difference in pregnancy outcome was seen if the patient underwent retrieval.