Prediction of response to growth hormone treatment in pre-pubertal children with growth hormone deficiency

Abstract

GH therapy response varies substantially among patients. Several models were developed to predict the efficacy of GH therapy in children. To evaluate the accuracy of a growth prediction model using data from an Italian pediatric GH deficiency (GHD) cohort (GeNeSIS, Growth Prediction Sub-study). Open-label, multicenter study in 22 Italian pre-pubertal GH treatment- naïve patients with GHD (8 female, 14 male, 0.5 to 12.2 yr), 18 isolated GHD, 4 multiple pituitary hormone deficiency given recombinat human GH therapy (0.025-0.035 mg/kg/day) for 12 months. Growth prediction was performed, after 3 months of treatment, using baseline data [bone age (BA) and IGF-I], a urinary marker of bone turnover [deoxypyridinoline crosslinks (DPD)] at 4 weeks, and height velocity (HV) at 3 months. Results were expressed as 1st-yr HV using the following equation: 1-yr HV (cm) = 3.543 - (2.337 × BA) - (0.010 × IGF-I) + (0.100 × DPD) + (0.299 × 3-month HV). Predictions were compared to the 1st-yr HV and accuracy was calculated as percentage of the difference between mean calculated HV and the real 1st-yr HV. For females predicted HV was 12.98 ± 4.82 cm/yr and actually was 13.05 ± 3.91 cm/yr after the 1st year; for males predicted HV was 13.95 ± 5.39 cm/yr and actually was 12.93 ± 5.02 cm/yr. In this paediatric Italian cohort with GHD, a growth prediction model seems to be a valid tool to assess 1st-yr response to GH treatment in Italian children.