Clin Chem Lab Med. 2026 Aug 11. doi: 10.1515/cclm-2026-0844. Online ahead of print.
ABSTRACT
OBJECTIVES: Serum M-protein analysis is central to the assessment of monoclonal gammopathies. Quantitation and isotyping are commonly performed by SPE or CZE and IFE or IT, respectively. While these assays are widely used, they are characterized by limited sensitivity and subjective interpretation. We evaluated the inter-site reproducibility of an automated MALDI-TOF mass spectrometry (MS) platform and its concordance with electrophoretic methods.
METHODS: We prepared 120 blinded, serially diluted serum specimens from 17 patients with confirmed monoclonal proteins, of which 118 were eligible for quantitative analyses. Five academic laboratories analyzed identical samples using routine SPE/CZE, IFE/IT, Optilite® total immunoglobulin measurements and MALDI-TOF MS (the EXENT® System). We assessed mass precision, quantitative and qualitative agreement, inter-site reproducibility, and CV-based variability.
RESULTS: Mass precision was excellent with >99 % of results within 4 m/z of the five-site mean. Agreement with IFE/IT was 81.7 %. Bland-Altman analysis resulted in mean biases ranging from -1.14 g/L to 4.61 g/L with isotype-dependent differences observed. Global comparison of SPE/CZE to MS resulted in a median Passing-Bablok slope of 1.004 (IQR: 0.055). Inter-site reproducibility was improved for MS over SPE/CZE with median slopes of 1.029 (IQR: 0.083) and 0.997 (IQR: 0.113), respectively. Results with CV ≤15 % were more frequent for MS than by SPE/CZE (67.8% vs. 46.9 %), particularly at MP concentrations <10 g/L.
CONCLUSIONS: Automated MALDI-TOF MS demonstrated strong inter-site reproducibility and close concordance with electrophoretic methods. Residual variability was attributable to upstream immunoglobulin quantification, IgM-related effects and known limitations of electrophoretic integration. These findings support MS as a robust and harmonizable platform for M-protein analysis in multi-center settings.
PMID:42573501 | DOI:10.1515/cclm-2026-0844