A French multicenter analytical evaluation of the automated Lumipulse G sNfL blood assay (Fujirebio®) and its comparison to four other immunoassays for serum neurofilament light chain assessment in clinical settings.

OBJECTIVES

Measurement of serum neurofilament light chain (sNfL) protein is becoming a key biomarker for many neurological diseases. Several immunoassays have been developed to meet these clinical needs, revealing significant differences in terms of variability and results. Here, we propose a French multicenter comparison of 5 sNfL assays.

METHODS

6 replicates of 3 pools with low (10 pg/mL), medium (30 pg/mL) and high (100 pg/mL) sNfL values and one replicate of 12 samples with growing sNfL values were analyzed by six independent French clinical laboratories. The analytical performances of the sNfL blood assay (Fujirebio®) on Lumipulse G were first evaluated then compared to four other immunoassays: NF-light V2 (Quanterix®) on SiMOA HD-X, Human NF-L (Biotechne®) on Ella, R-Plex Human Neurofilament L (MSD®) on Sector 2400; manual ELISA test using Uman Diagnostic/Quanterix®.

RESULTS

Inter-center comparison of the Lumipulse blood assay revealed limited but significant differences in the mean sNfL values across low, medium, and high pools between each city (p < 0.001) and between the two different batches used. Coefficients of variation of pools ranged from 2.0 to 16.9 %. Z-score of sNfL results of the 12 samples ranged from -1.70 to +1.71. Inter-technique comparison showed a systematic difference of sNfL values, with a overestimation of MSD and Ella over other tests. Nonetheless, results were all significantly correlated (p < 0.001).

CONCLUSION

The automated Lumipulse assay produced comparable sNfL values across laboratories; but further adjustments are needed to harmonize sNfL results. Biologists and physicians should be aware of the variability in results between different immunoassay suppliers.