USE OF INSULIN-LIKE GROWTH FACTOR-BINDING PROTEIN-2 (IGFBP-2), IGFBP-3, AND IGF-I FOR ASSESSING GROWTH-HORMONE STATUS IN SHORT CHILDREN

Insulin-like growth factor-binding protein-2 and -3 (IGFBP-2 and -3) are members of a family of proteins that are present in extracellular fluids and bind IGF-I and -II. IGFBP-2 is regulated differently from IGF-I and IGFBP-3, because its serum concentrations are elevated in some adults with GH deficiency (GHD), whereas IGF-I and IGFBP-3 concentrations are usually decreased. The purposes of this study were to define the normal range of IGFBP-2 concentrations in children, to determine its efficacy in the diagnosis of GHD, and to compare the diagnostic value of measurements of the serum GH response to provocative testing with results of measurements of IGFBP-2, IGFBP-3, and IGF-I. Mean serum IGFBP-2 concentrations ranged from 263 +/- 101 ng/mL (mean +/- SD) during infancy to 136 +/- 38 ng/mL in normal 15- to 18-yr-olds (P < 0.001), whereas IGFBP-3 increased from 1211 +/- 384 to 2781 +/- 382 ng/mL in the same age groups. Thirty-nine of 49 children with GHD and low IGF-I values (serum GH response, less than or equal to 1 ng/mL after 2 provocative tests) had serum IGFBP-2 concentrations that were greater than 2 SD above their corresponding age-adjusted means. In contrast all 49 of these children had IGFBP-3 values that were below normal for age. Because serum IGFBP-2 concentrations are regulated by GH directly and not through IGF-I, the IGFBP-2 to IGF-I ratio was used to determine whether it improved diagnostic accuracy. Fifty of 57 GH-deficient children had IGFBP-2/IGF-I ratios that were greater than 2 SD above the mean. This included 48 of 49 children with low IGF-I and 2 of 8 children with normal IGF-I. Fifty-three of the 57 children with GHD had decreased IGFBP-3 values. Among 23 children with idiopathic short stature (ISS) who had normal responses to GH stimulation testing (serum GH, >10 ng/mL), had low IGF-I values. Of the 7, all had an increased IGFBP-2/IGF-I ratio and a low IGFBP-3 level. Of the remaining 16 children with normal IGF-I, 13 had a normal IGFBP-2/IGF-I ratio and normal IGFBP-3 values. Three had low IGFBP-3 and an increased IGFBP-2/IGF-I ratio. In 76% of the 80 short-statured patients studied, there was concordance among serum GH responses to provocative tests, IGF-I, IGFBP-2/IGF-I ratio, and IGFBP-3. Of the remaining 19 patients, 11 had concordance among the values of IGF-I, IGFBP-2/IGF-I ratios, and IGFBP-3, but these results were discordant with the serum GH response. We propose that the GH test results in these 11 patients fail to provide an accurate assessment of the status of GH secretion or action. We conclude that there is no completely reliable test for diagnosing or excluding GH deficiency in all short children. When used alone, none of the tests we have assessed has superior diagnostic specificity or sensitivity. Measurement of GH-responsive peptides, such as IGF-I, IGFBP-2, and IGFBP-3, can add insights even when the results do not agree with the GH responses to provocative stimuli. Interpretation of the tests together improves the reliability of diagnostic assessment.