OBJECTIVE: To study the predictive value of initial low response (LR) in IVF/intracytoplasmic sperm injection (ICSI). DESIGN: Retrospective analysis. SETTING: Two Finnish fertility centers. PATIENT(S): A total of 3,846 IVF/ICSI cycles performed from 1994 to 2002. INTERVENTION(S): Consecutive cycles in the same subject were identified. The study groups consisted of subjects who had three treatment cycles and at least one LR cycle (n = 80). MAIN OUTCOME MEASURE(S): Pregnancy rate (PR), total gonadotropin dose, and embryo quality. RESULT(S): Only 2.5% (2/80) of subjects had a LR in all three consecutive cycles. In 43 women an initial LR was followed by >/=1 normal response (NR) cycle, and in 35 women an initial NR was followed by >/=1 LR cycle. The PR/cycle was similarly low in women with an initial LR and an initial NR (10.1% vs. 16.2%). An increase in gonadotropin dose resulted in a higher number of oocytes in women with an initial LR (from 2.1 +/- 0.9 to 6.7 +/- 2.7) but the PR/cycle remained low, compared to the overall mean PR (27.2%). In cycles in which top quality embryos were transferred, subjects with an initial LR had a lower PR than women with an initial NR (17.8% vs. 41.2%). CONCLUSION(S): An initial LR is a predictor of poor outcome in subsequent cycles, even if ovarian response is improved by increasing the gonadotropin dose or a top quality embryo is replaced.
OBJECTIVE: To study the predictive value of initial low response (LR) in IVF/intracytoplasmic sperm injection (ICSI). DESIGN: Retrospective analysis. SETTING: Two Finnish fertility centers. PATIENT(S): A total of 3,846 IVF/ICSI cycles performed from 1994 to 2002. INTERVENTION(S): Consecutive cycles in the same subject were identified. The study groups consisted of subjects who had three treatment cycles and at least one LR cycle (n = 80). MAIN OUTCOME MEASURE(S): Pregnancy rate (PR), total gonadotropin dose, and embryo quality. RESULT(S): Only 2.5% (2/80) of subjects had a LR in all three consecutive cycles. In 43 women an initial LR was followed by >/=1 normal response (NR) cycle, and in 35 women an initial NR was followed by >/=1 LR cycle. The PR/cycle was similarly low in women with an initial LR and an initial NR (10.1% vs. 16.2%). An increase in gonadotropin dose resulted in a higher number of oocytes in women with an initial LR (from 2.1 +/- 0.9 to 6.7 +/- 2.7) but the PR/cycle remained low, compared to the overall mean PR (27.2%). In cycles in which top quality embryos were transferred, subjects with an initial LR had a lower PR than women with an initial NR (17.8% vs. 41.2%). CONCLUSION(S): An initial LR is a predictor of poor outcome in subsequent cycles, even if ovarian response is improved by increasing the gonadotropin dose or a top quality embryo is replaced.