OBJECTIVE: To evaluate the ability of a prediction model to identify the individual starting dose of FSH for ovulation induction using a step-down regimen. DESIGN: Retrospective analysis of clinical data in an academic fertility unit. Fifty-six normogonadotropic anovulatory infertile patients who failed to ovulate or conceive with clomiphene citrate were included. They were treated with exogenous gonadotropins with a flexible starting dose for ovulation induction using a step-down regimen. The clinically applied starting dose of exogenous gonadotropins was compared with the calculated response-dose using a previously published prediction model. RESULTS: Patients were arbitrarily divided into three groups according to the day of the first decrease in gonadotropin dose: (a) early step-down (day 3 or earlier); (b) standard step-down (day 4 or later); (c) no step-down. These groups had average starting doses of 28.5 IU (group a) and 13 IU (group b) above the calculated response-dose, and 43 IU (group c) under the calculated response-dose. A significant correlation between day of first step-down and the difference between clinically applied and calculated response-dose was observed (P<0.0001, F-test for ANOVA). CONCLUSIONS: The patient group with the best step-down profile for ovulation induction exhibited the closest match between the clinically applied and calculated starting dose of gonadotropins. Therefore, this study provides support for the concept that the individual effective FSH starting dose for gonadotropin induction of ovulation in anovulatory infertile patients can be predicted on the basis of initial screening characteristics, such as body mass index, clomiphene resistance or failure, free IGF-I and FSH. This may result in more effective patient treatment protocols, reduced complication rates and health-economic benefits.
EJ van Santbrink, MJ Eijkemans, NS Macklon and BC Fauser
H Bili, J Laven, B Imani, MJ Eijkemans and BC Fauser
OBJECTIVE: To assess the effect of age on clinical, endocrine and sonographic features associated with polycystic ovary syndrome (PCOS) in normogonadotrophic anovulatory infertile women of reproductive years. DESIGN: Cross-sectional study. METHODS: Four hundred and seventy-two oligo-amenorrhoeic infertile patients, presenting with normal FSH and oestradiol concentrations, aged 17-42 years underwent a standardised initial evaluation including: cycle history, body mass index, waist-to-hip ratio and transvaginal ultrasound scanning of ovaries. Fasting blood samples were obtained for extensive endocrine evaluation. Cycle duration, serum levels of gonadotrophins, androgens, oestradiol, insulin, glucose, inhibin B as well as mean number of follicles, ovarian volume and ovarian stroma echogenicity were assessed. RESULTS: Older women had significantly lower LH and androgen and inhibin B serum levels. Similarly, older women presented with a reduced number of ovarian follicles. Age was inversely correlated with cycle duration (r=-0.112, P=0.02), LH (r=-0.154, P=0.001), testosterone (r=-0.194, P=0.001), androstenedione (r=-0.170, P=0.001), dehydroepiandrosterone (r=-0.157, P=0.001), insulin (r=-0.126, P=0.02), inhibin B (r=-0.118, P=0.03) serum levels and mean follicle number (r=-0.100, P=0.03). A positive correlation was observed between age and glucose to insulin ratio (r=0.138, P=0.009). CONCLUSIONS: Advanced age in normogonadotrophic anovulatory infertile women is associated with lower LH and androgen levels and with a decreased number of ovarian follicles. Although during reproductive years observed differences are relatively small, these age-related changes may affect the observed incidence of PCOS.