SPECIAL REPORT

The following summarizes some of the key data on biomarkers from the American Academy of Neurology annual meeting (AAN), San Diego, CA, April 5-9, 2025.

Neurofilament-light chain (NfL)
Combining sNfL and GFAP
Kappa free light chain

Fluid biomarkers are emerging as important diagnostic and prognostic tools in multiple sclerosis and in the evaluation of treatment response to disease-modifying therapies. The use of biomarkers in clinical practice was among the key topics addressed at the American Academy of Neurology annual meeting, held April 5-9, 2025, in San Diego, California.

Neurofilament-light chain (NfL)
The prognostic value of baseline NfL levels in serum was evaluated in the cohort of patients enrolled in the ASCLEPIOS I/II trials of ofatumumab (Ziemssen et al. Front Immunol 2022:13:852563). The median baseline sNfL level was 9.3 pg/mL; patients were stratified as high (above the median) versus low (below the median). A higher baseline sNfL was prognostic of new T2 lesions in year 1 and 2, and greater whole-brain and thalamic volume loss. The subsequent subgroup analysis (n=1678) looked at the prognostic value of baseline sNfL according to prior treatment status and prior treatments (Bittner et al. AAN 2025;P.009). There was a two-fold higher annualized rate of new T2 lesions with high versus low baseline sNfL in the previously-treated group (rate ratio 2.25), and in the treatment-naïve group (RR 2.16). The annualized rate of new T2 lesions for high vs. low sNfL was also prognostic in patients previously treated with interferon-beta (RR 2.05), glatiramer acetate (RR 2.28), dimethyl fumarate (RR 4.41), natalizumab (RR 3.41), and other DMTs (RR 4.71), but not for patients who had received fingolimod (RR 0.98).

The utility of sNfL assessments in practice is being evaluated in the NeofiLos pilot study of 83 office-based neurologists (Akgun et al. AAN 2025;P5.015). sNfL levels are measured at baseline and every three months thereafter. Elevated sNfL (Z-score >1.5) was detected in 16.1% of samples at study entry and in 21.7% at month 12. In most cases (73.8%), sNfL was not associated with clinical disease activity. In patients with clinical activity, sNfL was elevated in >50% of cases. A majority of neurologists stated that an elevated sNfL provided an initial hint that a treatment switch may be needed (64.8% at baseline, 78.3% at 12 months). Over 90% of neurologists said that sNfL was useful for evaluating treatment response. A majority of neurologists said that sNfL value was important to treatment decision-making at month 6 (51.0%), month 9 (54.3%) and month 12 (55.5%). However, few neurologists actually switched treatments in patients with elevated sNfL (3.8% at 3 months, 6.0% at 6 months, 1.8% at 12 months). The results indicate that at present, clinicians rely more on clinical or radiological assessments when determining the need to change DMTs.

Combining sNfL and GFAP
NfL primarily reflects neuroaxonal injury from neuroinflammation whereas glial fibrillary acidic protein (GFAP) is a marker of the astrocyte response to neurodegeneration. As such, serial measures of these and other biomarkers may provide a more complete picture of the complex pathophysiology of MS. A study at the University of Basel examined the trajectory of sNfL and sGFAP at baseline and in the 12 months following DMT initiation in a cohort of 106 MS patients (Uher et al. AAN 2025;P6.012). Mean age was 3.6 years; median EDSS score was 2.0; and mean follow-up was 4.8 years. sNfL Z-scores conformed to four different trajectories. In cluster 1 (23.6% of patients), elevated sNfL Z-scores were normalized with treatment. In cluster 2 (9.4%), elevated sNfL Z-scores decreased but remained elevated. In cluster 3 (41.5%), sNfL Z-scores were low and remained low. In cluster 4 (25.5%), initially low sNfL Z-scores increased slightly during treatment. Elevated sNfL Z-scores at 12 months (34.9% of the cohort) were predictive of a shorter time to relapse (hazard ratio 1.67). In contrast, sGFAP Z-scores were initially low and were unchanged after treatment initiation in 63.2% of cases. sGFAP Z-scores increased in 33.9%, which was associated with a non-significant higher risk of EDSS worsening.

Kappa free light chain
Kappa free light chain (KFLC), a marker of intrathecal B cell activity, has been incorporated into the new McDonald diagnostic criteria. Recent studies have also reported it is predictive of early disease activity in radiologically isolated syndrome and MS (Levraut et al. Neurol Neuroimmunol Neuroinflamm 2023;10:e200156. Rosenstein et al. J Neurol 2023;270:4800-4811).

An Austrian group found that in patients with CIS/early MS (mean age 33 years), those with a baseline KFLC index >100 had a two-fold higher risk of a second clinical attack in the next year compared to individuals with a low KFLC index. KFLC index was calculated as [CSF KFLC/serum KFLC]/albumin quotient) (Berek et al. Neurol Neuroimmunol Neuroinflamm 2021;8:e1005). The median time to a second attack was shorter in patients with a higher versus lower KFLC index (11 vs. 36 months).

The group presented new data on the utility of the KFLC index in predicting long-term disease activity (Hegen et al. AAN 2025;P6.004). Patients were followed for a median of 113 months. Overall, 72% had a relapse and 47% demonstrated worsening disability during the observation period. The baseline KFLC index predicted the time to relapse (hazard ratio 1.45), and disability worsening (HR 1.86). In patients with a high vs. low KFLC index (>100 vs. <100) at baseline, the chance of remaining relapse-free was decreased 40%; the probability of remaining free of disability worsening was 27% lower.