IGF-I ICMA and RIA assays: methodological spects and its implications in the diagnosis of growth hormone deficiency (GHD) in childhood

MG BALLERINI; HM DOMENÉ; A MARTINEZ; M MORINI; P SCAGLIA; A KESELMAN; RA REY; L KARABATAS; MC GUIDA; V PIPMAN; SV BENGOLEA; H CASSINELLI; I BERGADA; JJ HEINRICH; HG JASPER; MG ROPELATO

Congreso; The XXI Annual Meeting of the Latin American Society for Pediatric Endocrinology (SLEP); 2010

The diagnosis of GH-IGF-I axis disorders is partially dependent on accurate measurements of IGF-I. Objectives: To evaluate: 1-Analytical performance (ISO-15189) of ICMA-IGF-I, 2-The prevalence of low IGF-I levels in SS by ICMA compared to an in-house extractive RIA. IGF-I concentration was measured in 193 normal children, 17 GHD and 76 idiopathic-SS, 5.0-17.5yr by ICMA (IMMULITE2000, Siemens) and RIA. Log-transformed IGF-I results were compared using Passing-Bablok analysis. IGF-I<-2.0SDS were considered low for both methods. Results: ICMA analytical performance was acceptable: linearity (r=0.986), total CVs(<5.4%), bias (<8.7%), total error (19.4%) and sigma(>3.0). ICMA results were positively biased compared to RIA (y=1.567x-41, r=0.92) and showed higher divergences at puberty. In normal children, ICMA-IGF-I increased with age and puberty without sexual dimorphism. Low IGF-I values were observed in 12 versus 10/17GHD and in 25 versus 11/76 ISS by RIA and ICMA respectively, being the between-method agreement 83%(GHD) and 44%(ISS). Patients with discordant ICMA- and RIA-IGF-I results (n=16, 2/12GHD and 14/25 ISS), showed higher IGFBP3-SDS(p=0.005) and ALS-SDS(p=0.007) than those with low IGF-I results by both assays. In conclusion, some interference of IGFs binding proteins may not be completely nullified in ICMA and could explain its lower diagnostic accuracy of low IGF-I levels in short-statured children.