IGF-I IGMA and RIA assays: methodological aspects and its implications in the diagnosis of growth hormone deficiency in childhood

ROPELATO M; BALLERINI M, DOMENE H, MORINI M, MARTINEZ A, SCAGLIA P, KARABATAS L, HEINRICH J, BERGADA I, JASPER H.

Praga

Congreso; 49th Annual Meeting Europ. Soc. Peditr. Endocrinol; 2010

The diagnosis of GH axis disorders is partially dependent on accurate measurements of IGF-I. Hence, IGF-I measurements must be done in a reliable way and must allow uniform interpretation. Objectives: To evaluate the analytical performance of a commercially available IGF-I ICMA and its diagnostic efficiency in patients with GHD compared to those of a longstanding in-house RIA. Analytical performance of IGF-I ICMA (IMMULITE-2000, Siemens) was assessed within the framework of standard ISO 15189. Linearity, precision, bias, total error measurement (TE) and six sigma were calculated. ICMA  and RIA (JCEM 2000;85:4168) were compared using Passing- Bablock regression analysis. IGF-I was measured by both assays in 193 normal children (104 girls) and in 16 severe GHD patients (4 girls) (peak GH<7 ng/ml), age range 5 to 17.5 years. Log-transformed IGF-I values were used to establish reference ranges for each method. The diagnostic efficiency (DEf) of IGF-I SDS for GHD diagnosis was evaluated by ROC analysis. Results: ICMA analytical performance was found acceptable: linearity (r=0.9860, y=1.007x-12.62), total CVs (<5.4%), bias (<8.7 %), TE (<19.4%) and sigma (>3.0). ICMA results were proportional and positively biased compared to RIA (y=1.567X-41, r: 0.92) showing higher divergences at concentrations >250 ng/ml. IGF-I ICMA values increased with age and Tanner stage (maximum at Tanner IV) without sexual dimorphism. IGF-I ICMA SDS cut-off was -1.88 [sensitivity (S) 0.96 (95% CI: 0.93-0.99); specificity (Sp): 0.77 (0.50- 0.93) DEf 0.91 (0.80-1.03)]. Using RIA, IGFI SDS cut-off was -1.52 [S 0.93 (0.89-0.96), Sp 0.88 (0.63-0.99), DEf 0.92 (0.81-1.02)]. Considering values in SDS, there was a 94% of concordance for both methods in GHD diagnosis. In conclusion, higher values of IGF-I ICMA compared to RIA suggest that some interference of IGF-I-binding proteins remnants occur. The use of SDS cut-off for both methods allowed a comparable diagnostic efficiency in GHD children. Our findings stress the need for IGF-I local control values according to assay employed, age, and Tanner stage