Agreement of LC-MS assays for IGF-1 traceable to NIST and WHO standards permits harmonization of reference intervals between laboratories

Abstract

Objectives: In this study, we aimed to determine the feasibility of transferring IGF-1 reference intervals between two liquid chromatography-mass spectrometry assays with distinct assay formats and calibration traceability.

Design and methods: To adopt a reference interval (RI) for our new assay we have conducted RI transference and verification studies according to the CLSI EP28-A3c and EP9c guidelines. Specifically, the analytical agreement between the assays was evaluated using the linear model and the appropriateness of the linear model for RI transference was assessed using Deming regression, correlation coefficients, Q-Q plot, difference plot and studentized residues for the LC-MS/MS against DiaSorin LiaisonXL IGF-1 immunoassay and the liquid chromatography-high resolution mass spectrometry (LC-MS/HRMS) IGF-1 assay. Both Diasorin immunoassay and LC-MS/HRMS assays are traceable to WHO, 02/254.

Results: Our study showed a strong correlation (R² > 0.93) and agreement (slope = 1.006, negligible intercept) between LC-MS/MS and LC-MS/HRMS regardless of their traceability and all statistical criteria were met per CLSI guidelines. Conversely, while the LC-MS/MS and Diasorin immunoassay results showed a strong correlation (R² > 0.97, slope = 1.055), they failed to meet all statistical criteria for RI transference due to the bias (-44.91) and non-normal distribution of the residues. The RI verification study showed that 90% of the local LC-MS results fell within the RIs transferred from the reference LC-MS method, thus meeting CLSI EP28-A3c guidelines and permitting the transference of the reference LC-MS RIs.

Conclusions: Taken together, this study provides data to suggest excellent agreement between assays traceable to distinct reference standards for IGF-1.

Keywords: IGF-1; Immunoassay interferences; LC-MS/MS; Reference interval transference; Traceability.