OBJECTIVE: To assess the reliability of growth hormone (GH) secretion tests using provocative agents in women of different ovarian status. DESIGN: Comparison of GH secretion in response to clonidine (Catapressan; Boehringer Ingelheim, Reims, France) and growth hormone-releasing factor (GH-RF). SETTING:University Hospital, Hôpital Tenon, Paris, France. PATIENTS: Women categorized as "normal" (n = 6) or "poor" responder (n = 7) to ovarian stimulation with gonadotropins, depending on the follicular development attained at previous IVF attempts. INTERVENTIONS:Clonidine (0.300 mg administered orally) once and GH-RF (1 micrograms/kg IV) repeated twice. The tests were performed in random order in each individual on following cycles. MAIN OUTCOME MEASURES: Basal and peak GH values, area under the curve (AUC). RESULTS: Poor responder patients show significantly higher basal levels of FSH, GH, and insulin-like growth factor 1; FSH and basal GH levels are positively correlated. Peak GH levels and AUC are not significantly different in both categories of patients, whether GH-releasing factor or clonidine are used as provocative agents. True positive rate is 56.4% at the cutoff value of 7 micrograms/L, with no significant difference between the patients of the normally or poorly responding groups. At the 10-micrograms/L cutoff level, the true positive rate is almost half in the poor responder group (19.0% versus 38.9%), but the difference is not significant. CONCLUSIONS: These results raise concern about using the GH secretory response to a single clonidine administration as a predictive test of the therapeutic benefit that could be obtained by co-stimulating the somatotropic axis during a treatment with gonadotropins in poor responder patients, especially when their FSH basal levels are elevated.
RCT Entities:
OBJECTIVE: To assess the reliability of growth hormone (GH) secretion tests using provocative agents in women of different ovarian status. DESIGN: Comparison of GH secretion in response to clonidine (Catapressan; Boehringer Ingelheim, Reims, France) and growth hormone-releasing factor (GH-RF). SETTING: University Hospital, Hôpital Tenon, Paris, France. PATIENTS: Women categorized as "normal" (n = 6) or "poor" responder (n = 7) to ovarian stimulation with gonadotropins, depending on the follicular development attained at previous IVF attempts. INTERVENTIONS:Clonidine (0.300 mg administered orally) once and GH-RF (1 micrograms/kg IV) repeated twice. The tests were performed in random order in each individual on following cycles. MAIN OUTCOME MEASURES: Basal and peak GH values, area under the curve (AUC). RESULTS: Poor responder patients show significantly higher basal levels of FSH, GH, and insulin-like growth factor 1; FSH and basal GH levels are positively correlated. Peak GH levels and AUC are not significantly different in both categories of patients, whether GH-releasing factor or clonidine are used as provocative agents. True positive rate is 56.4% at the cutoff value of 7 micrograms/L, with no significant difference between the patients of the normally or poorly responding groups. At the 10-micrograms/L cutoff level, the true positive rate is almost half in the poor responder group (19.0% versus 38.9%), but the difference is not significant. CONCLUSIONS: These results raise concern about using the GH secretory response to a single clonidine administration as a predictive test of the therapeutic benefit that could be obtained by co-stimulating the somatotropic axis during a treatment with gonadotropins in poor responder patients, especially when their FSH basal levels are elevated.