Evaluation of immunoassays for the measurement of insulin-like growth factor-I and procollagen type III peptide, indirect biomarkers of recombinant human growth hormone misuse in sport

Insulin-like growth factor-I (IGF-I) and procollagen type III peptide (P-III-P) have been proposed as indirect biomarkers for the detection of the misuse of recombinant human growth hormone in sport. An extended intra- and inter-laboratory validation of commercially available immunoassays was carried out. For total IGF-I, two radioimmunoassay (RIA) kits (IGF-I/RIA1, Nichols Institute Diagnostics and IGF-I/RIA2, Mediagnost) and one enzyme-linked immunosorbent assay (ELISA) (R&D) were evaluated. For P-III-P, two RIA kits (P-III-P/RIA3, Cis-bioInternational and P-III-P/RIA4, Orion Diagnostica) were studied. The intra-laboratory precision and accuracy values for all IGF-I assays were better than 15%. The IGF-I/ELISA showed the lowest limit of quantification (LOQ) and its calibration curve covered the range of concentrations found in human serum samples. Higher agreement between laboratory results was obtained for IGF-I/ELISA and IGF-I/RIA1. Low inter-technique correlation was obtained for the three assays; the only comparable results were obtained between IGF-I/ELISA and IGF-I/RIA1. For P-III-P, intra-laboratory precision and accuracy values better than 15% were obtained for both assays in almost all cases. The calibration curve for P-III-P/RIA4 covered the range of concentrations of serum samples, while 30% of the values for P-III-P/RIA3 were below the calibration sample with the lowest concentration. Inter-laboratory correlation was also higher for P-III-P/RIA4. In summary, ELISA and RIA4 were the most suitable assays for measurement of IGF-I and P-III-P, respectively, in serum samples. However, the validation studies carried out show the need for harmonization of immunoassay parameters to improve the reproducibility and comparability of results between different laboratories and in different studies.