Method comparison and clinical classification agreement between Freelite and N-Latex serum-free light chain assays for clinical laboratory implementation

Abstract

Background: Measurement of serum free light chains (SFLC) is an essential component in the diagnosis and monitoring of plasma cell dyscrasias. Two analytical platforms are widely used in clinical laboratories: the Freelite assay, which employs polyclonal antibodies with turbidimetric detection, and the N-Latex assay, which uses monoclonal antibodies with nephelometric measurement. Because these assays differ in antibody design, calibration, and analytical detection principles, numerical concentrations and κ/λ ratios may
not be directly interchangeable across platforms. This study aimed to evaluate the analytical agreement and clinical classification concordance between the Freelite and N-Latex assays to assess the feasibility of implementing the N-Latex method in routine laboratory practice.

Materials and methods: A cross-sectional method-comparison study was performed using 33 paired archived residual serum samples obtained from routine clinical testing. Free κ and λ concentrations were measured using both Freelite and N-Latex assays on the same specimens. Analytical agreement was assessed using the Wilcoxon signed-rank test, Spearman correlation, Passing–Bablok regression, and Bland-Altman analysis. κ/λ ratios were interpreted according to assay-specific reference intervals, and categorical agreement was evaluated using Cohen’s kappa coefficient.

Results: Freelite produced significantly higher κ concentrations than N-Latex (median 43.0 vs 38.9 mg/L, p=0.0002), whereas λ concentrations did not differ significantly. Strong correlations were observed for both κ (ρ=0.897) and λ (ρ=0.951); however, correlation alone was not sufficient to demonstrate quantitative interchangeability. Passing-Bablok regression demonstrated proportional bias for κ measurements, with Freelite yielding higher values at elevated concentrations and the 95% confidence interval of the slope not including 1.0. Bland-Altman analysis revealed wide limits of agreement for κ values at extreme concentrations, indicating clinically relevant numerical differences between methods. When κ/λ ratios were interpreted categorically, using assay-specific reference intervals, concordant classification was observed in 30 of 33 samples (91%), corresponding to substantial agreement (κ=0.79).

Conclusions: The Freelite and N-Latex assays demonstrate strong correlation but exhibit systematic quantitative differences, particularly for κ measurements at high concentrations. Although κ/λ ratio interpretation is largely concordant, the two assays should not be considered directly interchangeable for absolute numerical interpretation or longitudinal monitoring. Implementation of the N-Latex assay is feasible in routine clinical laboratories provided that assay-specific reference intervals, appropriate dilution protocols, and antigen-excess precautions are applied, and clinicians are informed about expected differences when transitioning between
analytical platforms.