Overestimation of free light chain antigen excess rate.

BACKGROUND: Free light chains (FLC) are useful biomarkers for diagnosis and follow-up of plasma cell disorders. FLC quantification is encumbered by non-linearity and antigen excess (>4-fold difference between results obtained at the 2000- and 100-fold dilution). METHODS: FLC concentration was measured with Freelite® reagents on the BNII, using 100- and 2000-fold dilutions in 3645 samples. Samples displaying antigen excess were re-measured at the 2000- and/or 400-fold dilution. Carryover was evaluated by tracing samples to cuvettes and by measuring a normal sample in cuvettes that previously contained samples with a high FLC concentration. RESULTS: Antigen excess occurred in 0.93% of samples for κ and in 0.55% of samples for λ. In 81.5% of the cases, it could not be confirmed by a re-analysis of a 2000-fold and/or a 400-fold diluted sample. Real antigen excess was documented in 0.25% and 0.03% of the samples for κ and λ FLC, respectively. In the low concentration range (2000-fold dilution), imprecision was high. False antigen excess was reduced by batch analysis, introducing cleaning and rinsing procedures and using the 400-fold dilution. No antigen excess was detected in samples with normal FLC concentrations. CONCLUSION: Falsely high results occur by imprecision in the low concentration range and/or by carryover in cuvettes.

Effect of sample dilution on serum free light chain concentration by immunonephelometric assay.

BACKGROUND: Free light chains (FLC) are useful biomarkers in the assessment of plasma cell disorders. Concerns have been raised about some technical aspects of the assay. This report examined the occurrence of dilution anomalies and/or antigen excess. METHODS: FLC were determined on a BNII instrument at at-least two dilutions (100- and 2000-fold) in 2088 consecutive samples from 865 different patients. In part of them, the 400-fold dilution was also available. RESULTS: Higher than 2-fold differences and inconsistencies ["<" or ">" result at one dilution not consistent with the result obtained at another dilution] between the 100- and 2000-fold dilutions were found in 12.7% of patients for κ FLC and in 3.1% of patients for λ FLC. More than 4-fold differences between results obtained at the 2000-fold and the 100-fold dilutions were observed in 5.4% of patients for κ FLC and in 1.2% of patients for λ FLC. CONCLUSIONS: The FLC assay on BNII suffers from sample dilution anomalies and/or failure of antigen excess detection in a substantial fraction of patients. Laboratory professionals and clinicians should be alert to such analytical difficulties.

Diagnostic accuracy of the FIDIS multiplex fluorescent microsphere immunodetection system for anti-extractable nuclear antigen (ENA) antibodies in connective tissue diseases.

BACKGROUND: Anti-extractable nuclear antigen antibodies (ENA) are markers of connective tissue diseases (CTDs). METHODS: We compared FIDIS reagents in the multiplex fluorescent microsphere immunodetection system to INNO-LIA and immunodiffusion for 174 antinuclear antibody-positive patients, 102 with well-defined CTDs and 72 disease controls. RESULTS: No significant differences were found in sensitivity or specificity between FIDIS and immunodiffusion, or between FIDIS and INNO-LIA for all anti-ENA in all CTD patients; nor were any differences found for individual anti-ENAs within distinct CTDs. The FIDIS sensitivity was 41% (anti-SSA) and 17% (anti-SSB) in lupus erythematosus (LE) or primary Sjögren's syndrome; 5% (anti-ribosome and anti-Sm) in LE; 17% (anti-RNP) in LE or mixed CTD; 21% (anti-Scl70) in systemic sclerosis; and 61% (anti-centromere) in limited systemic sclerosis. The specificity reached 88%-100%. Receiver operating characteristic curve areas did not differ between FIDIS and INNO-LIA. Agreement ranged from 91% (anti-SSB) to 99% (anti-Jo1) between FIDIS and INNO-LIA, and from 95% (anti-Scl70) to 100% (anti-Sm) between FIDIS and immunodiffusion. Samples scored positive with all techniques in 83% (anti-centromere), 70% (anti-RNP), 67% (anti-Jo1), 60% (anti-SSA), 40% (anti-SSB), 33% (anti-ribosome), 25% (anti-Sm) and 13% (anti-Scl70) of cases. CONCLUSIONS: The diagnostic performance of FIDIS anti-ENA reagents is comparable to immunodiffusion and INNO-LIA.