United States multicenter study of factors predicting the persistence of GH deficiency during the transition period between childhood and adulthood

Abstract

Background: Many patients with childhood-onset growth hormone (GH) deficiency do not fulfill diagnostic criteria for GH deficiency (GHD) after attainment of adult height and may not require long-term GH treatment. Patients with history of idiopathic GHD (IGHD) pose the greatest management dilemma, as data regarding factors predictive of persistent GHD in this group are lacking.

Objectives: The objective of this study was to assess potential predictors of persistent GHD in a US patient cohort during transition from childhood to adulthood, particularly in patients with history of IGHD.

Methods: We studied 73 US patients with history of childhood-onset GHD screened at 21 US pediatric endocrine centers for a randomized clinical trial of GH replacement after attainment of adult height. The cohort comprised 42 boys/men and 31 girls/women aged14-22 years, who had received ≥1 year of GH treatment and had completed linear growth. The main outcome measures were sensitivity, specificity, positive and negative predictive values (PPV, NPV) of clinical and hormonal factors for persistent GHD (defined a priori in this study as peak GH < 5 μg/L).

Results: For the cohort as a whole, the best predictors of persistent GHD (100% PPV) were history of organic hypothalamic-pituitary disorder or ≥2 additional pituitary hormone deficiencies (PHD). Best predictors of persistent GHD in patients with childhood history of IGHD were standard deviation scores (SDS) for serum insulin-like growth factor binding protein-3 (IGFBP-3) below -2.0, and for insulin-like growth factor-I (IGF-I) below -5.3 (measured ≥6 weeks after completion of GH treatment; PPV 100% for both), and age <4 years at original diagnosis (PPV 89%). IGF-I above -1.6 SDS had 100% NPV.

Conclusions: US patients with an organic cause of childhood-onset GHD or ≥2 additional PHDs may not require GH stimulation testing to reconfirm GHD after completion of childhood treatment. In contrast, patients with idiopathic childhood-onset GHD almost invariably require retesting, as GHD persists in only a minority (those who were very young at initial diagnosis and those who have subnormal IGFBP-3 or extremely low IGF-I after completion of childhood treatment). Subnormal posttreatment IGF-I (<-2.0 SDS) lacked predictive power for persistent GHD, whereas IGF-I > -1.6 SDS was 100% predictive of GH sufficiency.

Figure 1

Participant flow diagram. Participant flow diagram of patients included in this study, showing numbers with history of organic vs. idiopathic GH deficiency, multiple pituitary hormone deficiencies vs. isolated GH deficiency, and numbers with confirmed GH deficiency on retesting. F = female; GH = growth hormone; M = male; N = number of patients; PH = pituitary hormone.

Figure 2

Peak GH response according to number of additional pituitary hormone deficiencies (PHD). Horizontal lines represent the median values of the peak stimulated GH concentrations for patients with 0, 1, 2, 3, and 4 additional PHDs; *p < 0.0001 for comparison of medians for the group with isolated GH deficiency (no additional PHDs) vs. all others. See “Results” for listing of stimulation tests used. ^#To avoid compressing the vertical axis, 1 GH value of 57 μg/L (idiopathic patient) is not shown. GH = growth hormone; N = total number of patients in each category (organic vs. idiopathic); n = number of patients in each subgroup.

Figure 3

a. Relationship between IGF-I SDS and peak GH response. Dashed vertical line represents the IGF-I threshold of -5.3 SDS, which provides 100% PPV for the diagnosis of persistent GH deficiency in patients with IGHD. IGF-I = insulin-like growth factor –I; b. Relationship between IGFBP-3 SDS and peak GH response. Dashed vertical line represents the IGFBP-3 threshold of -2.0 SDS, which provides 100% PPV for the diagnosis of persistent GH deficiency in patients with IGHD. IGFBP-3 = insulin-like growth factor binding protein 3; c. Relationship between body mass index and peak GH response. For the idiopathic group, Spearman r = -0.39, p = 0.003. There was no significant correlation for the organic group. Notes and abbreviations: See “Results” for listing of stimulation tests used; ^#One outlier idiopathic patient with a peak GH of 57 μg/L was excluded to avoid undue influence on the correlation and compressing the vertical axis. GH = growth hormone; IGHD = idiopathic GH deficiency; N = total number of patients in each category (organic vs. idiopathic); PPV = positive predictive value; SDS = standard deviation score.

Figure 4

Age at original childhood diagnosis of GH deficiency. Distribution of age at original diagnosis for patients who retested as GH deficient (left) vs. those who retested as GH sufficient (right). Within the persistently GH-deficient group, patients with history of idiopathic GH deficiency were significantly younger at diagnosis than those with history of organic GH deficiency (Table 1). Horizontal lines represent mean ages at initial diagnosis for patients with history of IGHD. GH = growth hormone; N = total number of patients in each category (organic vs. idiopathic).