Growth hormone (GH) replacement therapy in GH deficient adults: Predictors of one-year metabolic and clinical response

Abstract

Objective This study investigated whether baseline status could predict the responsiveness to one-year growth hormone (GH) replacement therapy in adult GH deficient (GHD) patients. Design A total of 380 European patients with adult onset GHD due to non-functioning pituitary adenoma that had been enrolled in Pfizer International Metabolic Database (KIMS), and that had completed one year of GH replacement therapy within KIMS, were studied. Results The mean initial dose of GH was 0.22 (SEM 0.01) mg/day and after one year, the mean dose was 0.36 (0.01) mg/day. The mean insulin-like growth factor-I (IGF-I) SD score increased from −1.75 (0.08) at baseline to 0.47 (0.05) after one year. Quality of life (QoL)-Assessment of GHD in Adults (QoL-AGHDA), waist circumference, waist:hip ratio, and serum lipid pattern improved. Women received a higher dose of GH than men after one year, and demonstrated similar treatment response. In multiple stepwise forward regression analyses, the one-year changes in QoL-AGHDA score, waist:hip ratio, and serum low density lipoprotein-cholesterol (LDL-C) level correlated inversely with the baseline values of the same variable. In addition, the change after one year in QoL-AGHDA score correlated inversely with duration of hypopituitarism and baseline serum high density lipoprotein-cholesterol (HDL-C) level, and the change in waist:hip ratio correlated inversely, although more weakly, with baseline serum HDL-C level and UK citizenship and positively with baseline waist circumference and the initial GH dose. The change in serum LDL-C level additionally correlated inversely with the mean GH dose and duration of hypopituitarism and positively with UK citizenship. Conclusions Baseline status could, with moderate strength, predict the responsiveness in the same variable whereas it could not, or only weakly, predict the response in other variables. Therefore, when the decision to start GH replacement is undertaken, as many outcome variables as possible should be evaluated in order to adequately evaluate the likelihood of clinical benefit. Finally, women have a similar response to GH replacement as men when individualised GH dosing schedules are employed and should therefore be selected for GH therapy to a similar extent.