NfL
GFAP
Optical coherence tomography (OCT)
Immune profiling and PRLs
Eye-movement biomarkers

Neurofilament-light chain (NfL) and glial fibrillary acidic protein (GFAP) are among the most-studied biomarkers of disease activity and progression in multiple sclerosis, but numerous other fluid and imaging biomarkers are currently being investigated. New advances in biomarker research, including the use of a combination of biomarkers, were among the highlights at the recent Americas Committee for Treatment and Research in MS (ACTRIMS) Forum, held February 27-March 1 in West Palm Beach, Florida. The following is a summary of some key developments.

NfL
Abdelhak and colleagues conducted a systematic review of NfL studies in MS comprising 24 randomized controlled trials and 51 non-RCTs (Abdelhak et al. ACTRIMS Forum 2025;P021). Overall, >85% of studies reported that a higher baseline sNfL level was associated with higher relapse risk over the next two years; the estimated risk was 2.5-fold higher with a baseline sNfL >60 pg/mL versus <30 pg/mL (Kuhle et al. Neurology 2019;92:e1007-e1015).

About 75% of studies have shown a modest association between elevated sNfL and EDSS progression, with one study reporting a hazard ratio of 1.4 (Manouchehrinia et al. Ann Clin Transl Neurol 2020;7:139-143). A limited number of studies have shown an association between sNfL and other disability measures, including the Timed 25-Foot Walk and the 9-Hole Peg Test.

Of particular interest was an analysis of the EPIC (n=609) and Swiss MS cohorts (n=1290) (Abdelhak et al. JAMA Neurol 2023;80:1317-1325). An elevated sNfL Z-score predicted relapse-associated worsening (RAW) at one year, and progression independent of relapse activity (PIRA) at 1-2 years, illustrating that fluid biomarkers can provide insights on the ongoing pathological processes that underlie disability worsening.

GFAP
A subsequent analysis of the EPIC cohort (n=621) has also identified astrocyte involvement prior to PIRA events (Ning et al. ACTRIMS Forum 2025;P063). PIRA was defined as 6-month confirmed progression on EDSS, T25FW or 9HPT in the absence of relapses. GFAP Z-scores were elevated at a median 12 months prior to a PIRA event. A Z-score >1 (either for GFAP for sNfL) was associated with a higher risk of subsequent PIRA (HR 1.29 for both) within the next 13 months. Interestingly, neither the GFAP Z-score nor the sNfL Z-score was elevated at the time of the visit documenting a PIRA event.

A separate 3T MRI study examined GFAP as a marker of deep-grey matter damage in RMS and PMS patients (N=97) (Stolting et al. ACTRIMS Forum 2025;P244). Elevated GFAP was associated with decreased volumes in the striatum, basal ganglia, caudate, putamen and pallidum suggestive of neurodegeneration and atrophy. Other findings indicated glial cell activation in deep GM structures as well as microstructural changes and neuronal loss in the hippocampus and ventral diencephalon.

Optical coherence tomography (OCT)
OCT measures the thickness of the retinal nerve fibre layer (NFL) and the macular ganglion cell and inner plexiform layer (GCIPL) as biomarkers of neuroaxonal damage. The value of OCT in predicting progression independent of relapse activity (PIRA) was evaluated in a retrospective analysis of 221 relapse-free RRMS patients at centres in Berlin and Munich (Braginets et al. ACTRIMS Forum 2025;P054). PIRA was defined as sustained disability accrual in any of the measures used (EDSS, T25FW, 9HPT, SDMT, low-contrast Letter Acuity). In the Berlin cohort (mean age 39 years), 28% of patients experienced PIRA at five-year follow-up. RNFL thickness was significantly associated with PIRA (HR 0.72); the association between GCIPL thickness and PIRA was not significant. In the Munich cohort (mean age 38 years), 30% of patients had a PIRA event. However, there was no association between RNFL or GCIPL thickness and PIRA in this group.

A separate study examined the utility of combining OCT with sNFL in 248 patients (mean age 54.6 years) with progressive MS (Lin et al. ACTRIMS Forum 2025;P037). EDSS worsening was defined as >1.5-point change if EDSS <5.5, and >1.0-point change if EDSS >5.5. Overall, 22.6% had EDSS worsening after a median follow-up of 4.6 years. There was an increased risk of disability worsening in patients with either lower RNFL Z-scores (HR 0.82), GCIPL Z-scores (HR 0.79), or higher sNfL Z-scores (HR 1.43). The risk of disability was higher in patients with abnormal GCIPL and sNfL (HR 3.28). The rate of EDSS change was 0.14 units/year with abnormal sNfL, 0.21 units/year with abnormal GCIPL and 0.19 units/year with abnormal RNFL. The annual change in EDSS was higher in patients with abnormal sNfL + abnormal GCIPL (0.26 units/year).

Immune profiling and PRLs
A new prospective study investigated whether blood immune cell profiling can provide an indication of CNS compartmentalized inflammation associated with paramagnetic rim lesions (PRL) (Thebault et al. ACTRIMS Forum 2025;S1.1). CSF and blood samples were collected from 21 untreated patients (mean age 38 years); 90% of patients had >1 PRL on 7T MRI. Blood biomarkers (sNfL, sGFAP) were not correlated with PRL volume. Among the 450 immune cell parameters that were investigated, there were 22 CSF immune cell subsets and 10 blood cell subsets that were associated with PRL volume. Immune cell features indicative of compartmentalized inflammation included higher numbers of NK cells and antibody-secreting B cells in CSF, and elevated CD8 CD20dim in CSF and blood. Regulatory CD4 T cell levels (higher in CSF, lower in blood) were also associated with PRL volume. Three components combined (NK cells, antibody-secreting B cells, Treg CSF/blood ratio) were moderately predictive of PRL volume.

Eye-movement biomarkers
A Canadian group investigated if subtle changes in eye movements are predictive of physical and/or cognitive progression in MS patients (Giacomini et al. ACTRIMS Forum 2025;P058). The interim analysis included 66 patients followed for up to 18 months. Eye movements were captured using a novel technology administered with an iPad. Clinical assessments evaluated clinically meaningful change, defined as a 1-point increase in EDSS score, a 4-point decrease in SDMT score, a 20% increase in T25FW time, or a 20% decrease in Brief Visuospatial Memory Test-Revised (BVMT-R) score. Overall, 50% of patients had a clinically meaningful change on at least one measure. Changes in 10 eye-movement measures (e.g. fixation, pro- and anti-saccade, smooth pursuit) were significantly different in patients with physical or cognitive disability progression. These metrics were able to distinguish patients with versus without motor or cognitive progression with 85% accuracy (sensitivity 82%, specificity 88%). A prospective 24-month Canadian study (ELIOS) is planned to evaluate the usefulness of eye-movement biomarkers for monitoring disability progression and therapeutic response in patients receiving ofatumumab (Devonshire et al. ACTRIMS Forum 2025;P059).