BACKGROUND: There is scanty information analysing the predictive value of a poor response, in terms of cancellation of the IVF cycle because of poor follicular development, as a predictor of ovarian response in a subsequent treatment cycle. This study, where logistic regression analysis was used, was undertaken to investigate the relative power of the woman's age, basal FSH, and previous cycle cancellation both as single and combined predictors of ovarian response in an IVF program where pituitary desensitization is routinely used. METHODS: One hundred and twenty-nine consecutive patients having their first cycle of IVF/ICSI treatment cancelled because of poor follicular response and undergoing a second attempt within 6 months after the failed treatment cycle were initially selected (group 1). Group 2 comprised 129 patients undergoing the first cycle of IVF/ICSI treatment and who were randomly selected from our assisted reproductive treatment program matching by BMI and indication for IVF/ICSI to those in group 1. RESULTS: Cancellation rate was significantly higher but ovarian response significantly lower in group 1 as compared with group 2. As indicated by the AUC(ROC) determined with ROC analysis, such a poor outcome in patients having a previous IVF/ICSI cycle cancelled due to poor response was observed whatever the level of basal FSH. In a logistic regression analysis and according to the odds ratio values, the predictive capacity of a previous poor response was 9 and 7.6 times higher than the predictive capacity of age and basal FSH, respectively. Any two or all three variables studied did not improve the predictive value of previous cycle cancellation alone. CONCLUSIONS: The history of an IVF/ICSI cancelled cycle due to poor follicular response in a standard stimulation protocol is a better predictor of cancellation in subsequent treatment cycles than age or FSH. The poor ovarian response associated with previous cycle cancellation occurs whatever the level of basal FSH.
BACKGROUND: There is scanty information analysing the predictive value of a poor response, in terms of cancellation of the IVF cycle because of poor follicular development, as a predictor of ovarian response in a subsequent treatment cycle. This study, where logistic regression analysis was used, was undertaken to investigate the relative power of the woman's age, basal FSH, and previous cycle cancellation both as single and combined predictors of ovarian response in an IVF program where pituitary desensitization is routinely used. METHODS: One hundred and twenty-nine consecutive patients having their first cycle of IVF/ICSI treatment cancelled because of poor follicular response and undergoing a second attempt within 6 months after the failed treatment cycle were initially selected (group 1). Group 2 comprised 129 patients undergoing the first cycle of IVF/ICSI treatment and who were randomly selected from our assisted reproductive treatment program matching by BMI and indication for IVF/ICSI to those in group 1. RESULTS: Cancellation rate was significantly higher but ovarian response significantly lower in group 1 as compared with group 2. As indicated by the AUC(ROC) determined with ROC analysis, such a poor outcome in patients having a previous IVF/ICSI cycle cancelled due to poor response was observed whatever the level of basal FSH. In a logistic regression analysis and according to the odds ratio values, the predictive capacity of a previous poor response was 9 and 7.6 times higher than the predictive capacity of age and basal FSH, respectively. Any two or all three variables studied did not improve the predictive value of previous cycle cancellation alone. CONCLUSIONS: The history of an IVF/ICSI cancelled cycle due to poor follicular response in a standard stimulation protocol is a better predictor of cancellation in subsequent treatment cycles than age or FSH. The poor ovarian response associated with previous cycle cancellation occurs whatever the level of basal FSH.