Constitutional delayed puberty in males and hypogonadotropic hypogonadism: a reliable and cost-effective approach to differential diagnosis.

Forty-six male teenagers 13-19 years old with delayed puberty (DP) underwent gonadotropin releasing hormone (GnRH) and human chorionic gonadotropin (HCG) stimulation as part of their work-up. All were followed to age 18 and beyond. Thirty-seven had constitutional delayed puberty (CDP). Nine had hypogonadotropic hypogonadism (HH). At referral 34 youngsters with CDP were properly diagnosed when the lower limit for the luteinizing hormone (LH) response to GnRH (Factrel 0.1 mg i.v.) was set at 12 IU/l. Three boys with CDP failed to reach that level and were not assigned appropriately. All nine patients with HH had basal serum testosterone (T) < 50 ng/dl when first seen and LH responses to GnRH stimulation < 8.0 IU/l. In the late 1970s, five subjects with DP were given HCG 3,000 IU (two patients daily for 5 days; three on 3 alternate days). Serum T was measured before the first, and 48 hours after the last injection (day 7). With recognition of the long biological half-life of injected HCG and receptor downregulation by daily doses, the protocol was changed. In the early 1980s, the dose of HCG was randomized to either 500 IU or 1,000 IU given on 3 alternate days. T was measured before the first injection (basal), 48 hours later (day 3) and 48 hours after the third injection (day 7). At referral 35 patients with CDP, including one GnRH failure, met the criterion for a positive response to HCG stimulation based on their own reactions (T > 170 ng/dl on day 3; > 200 on day 7). Eleven patients with DP failed the test. Nine had HH and two had CDP. The nine patients with HH included the two given daily injections and the three given HCG 3,000 IU on 3 alternate days. Of the two with CDP, one, an obese boy with a normal GnRH test, only received 500 IU HCG (5.6 IU/kg), which may have been inadequate. The other failed both tests. Of the 35 responders, 17 (group 1) were given HCG 500 IU and 18 (group 2) were given 1,000 IU i.m. on 3 alternate days. Seven boys in group 1 and 12 in group 2 had serum T determined on day 3, and all 35 had T measured on day 7. There was no significant difference between the basal T levels in the two groups or in their responses to HCG, and the results were pooled. The combined data in the patients with CDP were then compared with those of the nine patients with HH, recognizing that the result on day 7 in the two patients given daily injections may reflect receptor down-regulation. Setting the limit for a normal T response at > 170 ng/dl 48 hours after a single injection of HCG and > 200 ng/dl after the third injection assigned all the patients with CDP appropriately. The data on the patients with HH are less complete, with only three T values 48 hours after the first injection However, even after three injections of HCG on alternate days, only one of nine patients with HH approached the lowest level achieved by the patients with CDP after a single injection. The data confirm that a basal serum T 50 ng/dl is evidence of the onset of puberty. In those with serum T < 50 ng/ dl our data suggest that a single injection of HCG 15 IU/kg, with serum T determined 48 hours later, is more discriminatory and offers the most reliable, easy to perform, least painful, and by far the most cost effective test to differentiate CDP from HH.