Department of Endocrinology, Institution of Internal Medicine, The Sahlgrenska Academy, Sahlgrenska University Hospital, Gothenburg University, Gothenburg, Sweden.
Eur J Endocrinol. 2010 May;162(5):869-78. doi: 10.1530/EJE-09-0973. Epub 2010 Feb 9.
Clinical response to GH therapy in GH-deficient (GHD) adults varies widely. Good predictors of treatment response are lacking. The aim of the study was to develop mathematical models to predict changes in serum IGF1 and body composition (BC) in response to GH therapy in GHD adults.
One hundred and sixty-seven GHD patients (103 men, median age 50 years) were studied before and after 12 months of GH treatment. GH dose was tailored according to serum IGF1 concentrations. Good responders (GR) and poor responders (PR) to GH therapy were defined as patients with a response >60th and <40th percentile respectively, for changes in serum IGF1 levels (adjusted for GH cumulative dose) and in BC (lean body mass (LBM) and body fat determined using dual-energy X-ray absorptiometry). A logistic regression model was used to predict the probability of being a GR or PR.
In the IGF1 prediction model, men (odds ratio (OR) 5.62: 95% confidence interval 2.59-12.18) and patients with higher insulin levels (OR 1.06: 1.00-1.12) were more likely to be GR. The accuracy of the prediction model was 70%. In the BC model, men (OR 10.72: 1.36-84.18) and GHD patients with lower LBM (OR 0.82: 0.73-0.92) and greater height (OR 1.23: 1.08-1.40) at baseline were more likely to be GR. The accuracy of the prediction model was 80%.
Accurate mathematical models to predict GH responsiveness in GHD adults were developed using gender, body height, baseline LBM, and serum insulin levels as the major clinical predictors.
生长激素缺乏症（GHD）成人对 GH 治疗的临床反应差异很大。缺乏良好的治疗反应预测指标。本研究旨在建立数学模型，预测 GHD 成人 GH 治疗后血清 IGF1 和身体成分（BC）的变化。
研究了 167 例 GHD 患者（103 例男性，中位年龄 50 岁），在接受 GH 治疗 12 个月前后进行了研究。根据血清 IGF1 浓度调整 GH 剂量。将 GH 治疗后血清 IGF1 水平（校正 GH 累积剂量）和 BC（使用双能 X 线吸收法测定的瘦体重（LBM）和体脂）变化>60%和<40%的患者定义为良好反应者（GR）和不良反应者（PR）。使用逻辑回归模型预测成为 GR 或 PR 的概率。
在 IGF1 预测模型中，男性（比值比（OR）5.62：95%置信区间 2.59-12.18）和胰岛素水平较高的患者（OR 1.06：1.00-1.12）更有可能成为 GR。预测模型的准确性为 70%。在 BC 模型中，男性（OR 10.72：1.36-84.18）和基线 LBM 较低（OR 0.82：0.73-0.92）和身高较高（OR 1.23：1.08-1.40）的 GHD 患者更有可能成为 GR。预测模型的准确性为 80%。
使用性别、身高、基线 LBM 和血清胰岛素水平作为主要临床预测指标，建立了预测 GHD 成人 GH 反应性的准确数学模型。
