Beyond the method change in clinical practice: evaluation of insulin-like growth factor I assay.

Abstract

Insulin-like growth factor I (IGF-I) is the preferred biomarker for diagnosing and monitoring growth-related disorders but its serum quantification presents several difficulties since different IGF-I assays still leads to different IGF-I concentrations, especially when results are either above or below the normal range. We conducted a prospective study between November and December 2020 at a tertiary University Hospital with 212 serum samples to determine the analytical performance of the IGF-I assay on the Cobas e411 (Roche Diagnostics) and compare it with that of the Immulite 2000XPi (Siemens). In this work, we report for the first time the existence of discrepancies between IGF-I levels measured by Immulite 2000XPi and Cobas e411. Deming regression model provided a slope of 1.570 (95% CI: 1.395-1.745) and an intercept of-58.591 (95% CI:-89.151 to-28.030), with R2=0.967 and average bias of +53.061 with overestimation of IGF-I. It was found that Cobas e411 provides abnormally high IGF-I concentrations, but further studies are required to elucidate the cause of the discrepancies. Our data can alert clinicians and laboratory professionals of this situation and avoid misinterpretation of increased IGF-I levels as a therapeutic failure rather than as a problem associated with this method change.