Central laboratory reassessment of IGF-I, IGF-binding protein-3, and GH serum concentrations measured at local treatment centers in growth-impaired children: implications for the agreement between outpatient screening and the results of somatotropic axis functional testing.

BACKGROUND Childhood GH deficiency, suspected in the presence of decreased height velocity and short stature, is usually characterized by low IGF-I and IGF-binding protein-3 (IGFBP-3) serum concentrations and is conventionally confirmed by diminished GH peak responses to pharmacological stimuli. OBJECTIVE We evaluated the agreement between different IGF-I (IGFBP-3) assays in predicting GH deficiency and tested whether variability between growth factor screening and pharmacological testing could be diminished by reassessment of growth factor and GH peak concentrations in a single laboratory. DESIGN Using the Tuebingen IGF-I (IGFBP-3) RIA, 317 (321) sera from children evaluated for growth disorders in 19 centers were reanalyzed. In 103 children with insulin hypoglycemia and arginine tests, we evaluated how the association between the outcome of growth factor screening and functional testing would change if different assays were employed. RESULTS Locally measured IGF-I correlated better than IGFBP-3 with the results of the central laboratory (Tuebingen) assay (slope of the regression curve 1.05; 95% confidence interval (95% CI) 1.01-1.1 versus 1.18; 95% CI 1.09-1.3). Agreement between local and central laboratory assays in predicting GH deficiency was better for IGF-I than for IGFBP-3 assays (kappa =0.59 versus kappa =0.47). The poor agreement between growth factor screening and GH pharmacological testing was not improved when hormone concentrations were remeasured in the central laboratory (kappa local=-0.0031, central=0.12). CONCLUSIONS In children with impaired growth, growth factor screening reflects different aspects of GH insufficiency than does functional testing. Agreement between these approaches is poor and could not be improved by reduction of assay-related variability.