Discordance of insulin-like growth factor-1 results and interpretation on four different platforms.

ABSTRACT

OBJECTIVE: Inconsistent Insulin-like Growth Factor 1 (IGF-1) measurements among different platforms have been observed. In this study, we compared the IGF-1 assay on four different platforms. METHODS: A total of 110 serum specimens were analyzed in this comparison study. IGF-1 was measured on the three different chemiluminescent automated immunoassay of Siemens Immulite 2000 XPi, DiaSorin Liaison XL, IDS iSYS and LC-MS/MS method. Results were compared with Weighted Deming regression. Bias was evaluated using the Bland-Altman method. RESULTS: Weighted Deming regression analysis showed approximately 36 % negative variation on Immulite, compared to Liaison (Immulite = 0.64 * DiaSorin + 2.95, r2 = 0.95); 8 % negative variation on iSYS, compared to Liaison (iSYS = 0.92 * DiaSorin + 0.51, r2 = 0.97); 17 % negative variation on LC-MS/MS, compared to Liaison (LC-MS/MS = 0.83 * DiaSorin-11.23, r2 = 0.93); 34 % positive variation on LC-MS/MS compared to Immulite (LC-MS/MS = 1.34 * Immulite-21.97, r2 = 0.96); 81 % positive variation on IDS iSYS compared to Immulite (IDS iSYS = 1.81 * Immulite-117.65, r2 = 0.83). The Bland-Altman plot showed a significant negative variation of Immulite versus DiaSorin and positive variation of IDS iSYS versus Immulite. Overall agreement between different platforms was poor, which reflected systematic difference. The variation between platforms increased as IGF-1 values increased. CONCLUSIONS: There are wide variations between different platforms for IGF-1 measurement. The lack of standardization in IGF-1 measurement creates a challenge for clinicians to monitor IGF-1 and treat patients with pituitary disorders, when switching from one platform to another. The potential impact of the variations in IGF-1 measurement between different platforms should be taken into consideration when managing patients.