V Degros1, C Cortet-Rudelli, B Soudan, D Dewailly. 1. Service d'Endocrinologie et de Diabetologie, Clinique Marc Linquette, 6, rue du Professeur Laguesse, 59037 Lille cedex, France. ddewailly@chru-lille.fr
Abstract
OBJECTIVE: The effectiveness of biological investigations aiming at discriminating isolated hypogonadotropic hypogonadism (IHH) from constitutional delayed puberty (CDP) in male patients is still controversial. We revisited the diagnostic power of the basal serum testosterone level, the Triptorelin test and the human chorionic gonadotropin (hCG) test in a cohort of 33 boys with delayed puberty. DESIGN: Boys were aged 14.2 to 26.2 Years at referral. A 5-Year-long clinical follow-up after the initial study allowed confirmation of the diagnosis. At the end of the follow-up period, IHH was found in 13 patients while the other 20 had normal spontaneous pubertal development (CDP). RESULTS: At referral, a basal morning testosterone level >1.7 nmol/l was observed in 55% of patients with CDP exclusively (predictive positive value (PPV)=100%; predictive negative value (PNV)=59%). For CDP, the PPV of the LH peak 3 h after Triptorelin was 100% by setting the upper threshold at 14 IU/l and the PNV was 72%. However, no lower threshold could discriminate IHH from CDP in the remaining patients with an LH peak 3 h after Triptorelin <14 IU/l. In CDP patients, the PPV of the serum testosterone increment after hCG stimulation (deltaT/hCG) was 100% for values >9 nmol/l (PNV=72%). In IHH patients, the PPV of deltaT/hCG was 100% for values <3 nmol/l (PNV=82%). Only 29% of the studied population had a deltaT/hCG between these lower and upper thresholds and therefore could not have been classified initially. CONCLUSIONS: (i) Dynamic testing for the diagnosis of delayed puberty is useful only when the basal testosterone level is lower than 1.7 nmol/l; (ii) in that case, the hCG test has better discriminating power than the Triptorelin test and appears as the best cost-effective investigation. It prevents useless and expensive investigations in about one-half of CDP patients with a basal morning testosterone level lower than 1.7 nmol/l.
OBJECTIVE: The effectiveness of biological investigations aiming at discriminating isolated hypogonadotropic hypogonadism (IHH) from constitutional delayed puberty (CDP) in male patients is still controversial. We revisited the diagnostic power of the basal serum testosterone level, the Triptorelin test and the human chorionic gonadotropin (hCG) test in a cohort of 33 boys with delayed puberty. DESIGN:Boys were aged 14.2 to 26.2 Years at referral. A 5-Year-long clinical follow-up after the initial study allowed confirmation of the diagnosis. At the end of the follow-up period, IHH was found in 13 patients while the other 20 had normal spontaneous pubertal development (CDP). RESULTS: At referral, a basal morning testosterone level >1.7 nmol/l was observed in 55% of patients with CDP exclusively (predictive positive value (PPV)=100%; predictive negative value (PNV)=59%). For CDP, the PPV of the LH peak 3 h after Triptorelin was 100% by setting the upper threshold at 14 IU/l and the PNV was 72%. However, no lower threshold could discriminate IHH from CDP in the remaining patients with an LH peak 3 h after Triptorelin <14 IU/l. In CDPpatients, the PPV of the serum testosterone increment after hCG stimulation (deltaT/hCG) was 100% for values >9 nmol/l (PNV=72%). In IHH patients, the PPV of deltaT/hCG was 100% for values <3 nmol/l (PNV=82%). Only 29% of the studied population had a deltaT/hCG between these lower and upper thresholds and therefore could not have been classified initially. CONCLUSIONS: (i) Dynamic testing for the diagnosis of delayed puberty is useful only when the basal testosterone level is lower than 1.7 nmol/l; (ii) in that case, the hCG test has better discriminating power than the Triptorelin test and appears as the best cost-effective investigation. It prevents useless and expensive investigations in about one-half of CDPpatients with a basal morning testosterone level lower than 1.7 nmol/l.