Endocrine and ultrasonographic assessment of the ovarian reserve and their predictive potential for IVF outcome

Abstract

Objective: The assessment of basal cycle day 3 FSH, LH and E2 levels (usually referred to as ovarian reserve) has been demonstrated to predict response to controlled ovarian hyperstimulation and chances of conception in IVF. Recent studies have also shown that such predictability can be enhanced by ultrasonographic examination of basal ovarian follicle count. The objective of this study was to investigate the correlations between ovarian reserve, basal ovarian volume and antral follicle count, the main parameters reflective of ovarian response, and pregnancy outcome in IVF. Design: Prospective cohort analysis. Materials and Methods: A total of 135 consecutive women undergoing IVF (with or without ICSI) between January and December 2002 were enrolled independently of their age and the etiology of infertility. Pretreatment transvaginal ultrasound measurements and basal levels of FSH, LH and E2 on basal cycle day 3 were correlated with the parameters of subsequent ovarian response (peak E2, numbers of mature oocytes, number of ampoules of gonadotropins used) and cycle outcome (clinical pregnancy). Ovarian response outcomes were analyzed by ordinary least squares regression. Pregnancy outcome was analyzed by logistic regression. Univariate analyses were made for each predictor; a multivariable analysis was also made to adjust for the other predictors. Results: Peak E2 demonstrated a negative correlation with basal cycle day 3 FSH (r = −0.22, p = 0.01) and a positive correlation with both basal cycle day 3 LH and basal antral follicle count after correcting for other variables (r = 0.20, p = 0.004 and r = 0.27, p = 0.05, respectively). After adjustment for other variables, only age and basal antral follicle count correlated with the total number of ampoules of gonadotropins used (r = 0.43, p < 0.001 and r = −0.37, p < 0.001, respectively) with age being the strongest predictor. The number of mature oocytes obtained at retrieval showed a positive correlation with basal cycle day 3 LH and basal antral follicle count (r = 0.32, p < 0.001 and r = 0.29, p = 0.01, respectively), and a negative correlation with basal cycle day 3 FSH and E2 (r = 0.–30, p < 0.001 and r = −0.16, p < 0.001, respectively). Only age was a significant predictor of clinical pregnancy, but when adjusted for all other variables it was no longer statistically significant. Conclusion: Basal cycle day 3 assessment of ovarian reserve through hormonal evaluation of FSH, LH and E2, and determination of the basal antral follicle count, were equally predictive of response to controlled ovarian hyperstimulation using gonadotropins in IVF therapy. Basal ovarian volume did not correlate with such response.