Constructing an Insulin-Like Growth Factor-based Prediction Model

Serum insulin-like growth factor-I (IGF-I) measurements have been shown to correlate well with growth hormone action and effect, and recent data show that serum IGF-I may be related to safety and efficacy of growth hormone (GH) treatment in patients. Some studies indicate that high IGF-I levels are associated with increased cancer risk, and low IGF-I levels are associated with increased risk for cardiovascular disease. Studies in children also show that the serum IGF-I level is correlated with the change in height score achieved (that is, the higher the IGF-I level, the greater the gain in height). Pediatric endocrinologists have therefore begun to use serum IGF-I levels, in addition to growth rate and weight gain, to adjust the GH dose in treated children. Although monitoring of serum IGF-I levels is becoming standard of care in patients begin treated with GH, there are few guidelines regarding the actual logistics of adjusting GH dose. As serum IGF-I level has been linked to both safety and efficacy of GH treatment, the ideal practice would be to maintain serum IGF-I levels within a certain target range. The overall goal of our study is to construct a mathematical model which predicts the change in GH dose necessary to achieve a desired change in IGF-I level. Hypotheses to be tested by our study include the following: IGF-I measurement has a role in optimization of GH therapy; GH dose change to achieve IGF-I changes are predictable; and gender and puberty affect the relationship between dose change and target IGF-I changes.