The prospective REDUCE-GAD study reports that higher serum neurofilament-light chain (NfL) levels are associated with the presence of contrast-enhancing (CE) lesions, which may enable clinicians to forego contrast agents for follow-up MRIs (Schaefer et al. Eur J Neurol 2023;30:2393-2400).
Gadolinium-based contrast agents are routinely used in suspected MS since the co-occurrence of Gd+ lesions and non-enhancing lesions establishes the criterion of dissemination in time outlined in the McDonald criteria (Thompson et al. Lancet Neurol 2018;17:162-173). However, gadolinium is not advised for routine follow-up imaging due to concerns about cumulative toxicity, according to the most recent MAGNIMS-CMSC-NAIMS recommendations (Wattjes et al. Lancet Neurol 2021;20:653-670).
REDUCE-GAD enrolled 102 patients (mean age 36 years); 55.9% had CE at baseline. Blood samples were obtained within five days of the MRI. Samples were analysed for five biomarkers: NfL, glial fibrillary acidic protein (GFAP), tau protein, ubiquitin-carboxyl-terminal-hydrolase (UCH-L1), and matrix metalloproteinase (MMP-9). (UCH-L1 is a marker of neuronal cell body injury.)
NfL was most robustly correlated with CE vs. no CE. The NfL Z-score was more predictive of CE (adjusted odds ratio 1.521). NfL values >9.1 pg/mL indicated the presence of radiological disease activity. The authors noted that this was similar to the cut-off of 9.3 pg/mL that was predictive of brain lesions and neuronal loss in a post-hoc analysis of the ASCLEPIOS trials of ofatumumab (Ziemssen et al. Front Immunol 2022;13:852563).
The most useful cut-off value for NfL and CE was 14.1 pg/mL (sensitivity 49.1%, specificity 82.2%, positive predictive value 77.8%, negative predictive value 56.0%). Values >59.2 pg/mL were only seen in conjunction with CE. sNfL level was also correlated with the number and largest diameter of CE lesions.
GFAP was correlated with the number and size of CE lesions but not with CE positivity; results were influenced by patient age and prior steroid use. MMP-9 was correlated with clinical relapses but not CE; the authors noted that MMP-9 levels in CSF might be more sensitive. There was no correlation of CE with serum tau protein, UCH-L1 and S100B, suggesting that markers of neurodegeneration are less useful in younger patients.
The authors concluded that a biomarker-guided approach may enable clinicians to identify patients at higher risk of CE lesions while avoiding repeated exposure to gadolinium during long-term monitoring.