Verification study of free light chains assays on reagent-optimized analysers

Our aim was to compare analytical specifications of two assays (monoclonal vs. polyclonal) for free light chains (FLCs) quantification optimized for two different analytical platforms, nephelometer ProSpec (Siemens, Erlangen, Germany) and turbidimetric analyser Optilite (The Binding Site, Birmingham, UK).

The evaluation included verification of the precision, repeatability and reproducibility, estimation of accuracy and method comparison study with 37 serum samples of haematological patients. Kappa and lambda FLC were measured in each sample by both methods and kappa/lambda ratio was calculated.

Results show satisfactory precision of both methods with coefficients of variation for ProSpec of CV_wr = 2.20% and CV_br = 3.44%, and for Optilite CV_wr = 2.82% and CV_br = 4.15%. Estimated bias for FLC lambda was higher on the ProSpec analyser, but bias for FLC kappa was higher on the Optilite analyser. Correlation coefficients were 0.98; P < 0.001 for FLC kappa and 0.97; P < 0.001 for FLC lambda. Considering normal/pathological FLC ratio moderate agreement within assays was detected (κ = 0.621). When the results were categorized according to criteria for progressive disease, 4/37 (0.10) cases were differently classified. Lambda FLC values by Optilite in three samples with monoclonal FLC lambda were more than twelve times higher than by ProSpec. A 25% difference in FLC ratio was detected in 16/37 (0.43) and 50% difference in 13/37 (0.35) patients.

All manufacturers' precision claims could not be achieved in the verification study. The comparison of results to biological variations data showed that coefficients of variations are acceptable for both assays. The assays should not be used interchangeably in haematological patients.