Pathogenic autoantibodies against aquaporin 4 (AQP4) in neuromyelitis optica spectrum disorder (NMOSD) cause central nervous system injury, with subsequent release of astroglial and neuronal proteins such as glial fibrillary acidic protein (GFAP), neurofilament light chain (NfL), ubiquitin carboxyl-terminal hydrolase L1 (UCH-L1) and Tau into the circulation. N-MOmentum is a randomized, placebo-controlled, double-masked trial of inebilizumab, a B-cell-depleting monoclonal antibody (NCT02200770).
Investigate relationships of NfL, UCH-L1, Tau and serum (s)GFAP to disease activity and Expanded Disability Status Scale (EDSS) disability in N-MOmentum trial participants with either AQP4-immunoglobulin G (IgG) seropositive or seronegative NMOSD.
Serum biomarkers NfL, UCH-L1, Tau and sGFAP were measured using the single molecular array (SIMOA; Quanterix) in 1260 serial and attack-related samples from N-MOmentum participants (n=215) and healthy controls (HC; n=25).
At baseline, biomarkers were elevated in subsets of patients with NMOSD (NfL, 16%; UCH-L1, 6%; Tau, 12%; sGFAP, 29%); NfL and UCH-L1 levels correlated with sGFAP (r=0.53 [p<0.001] and 0.18 [p=0.007]). Baseline elevations were significantly associated with increased attack risk (NfL, hazard ratio [HR] 2.5, p=0.01; UCH-L1, HR 2.8, p=0.039; Tau, HR 2.6, p=0.01; sGFAP, HR 3.03, p<0.001). After controlling for baseline sGFAP in Cox regressions, the other markers were not independently associated with attack risk (all HR <2; p>0.05). In the total cohort, a greater proportion of patients had an attack with placebo than inebilizumab (39% vs 12%). All biomarker levels increased after attacks and median-fold increases from baseline (95% confidence interval) trended higher with placebo than inebilizumab, reaching significance with sGFAP (NfL, 1.49 [0.93–3.37] vs 1.30 [0.84–2.14], p=0.4; UCH-L1, 6.70 [1.59–52.4] vs 1.85 [0.89–23], p=0.12; Tau, 2.19 [0.96–9.46] vs 1.09 [0.40–3.7], p=0.23; sGFAP, 20.2 [4.4–98] vs 1.11 [0.75–24.6], p=0.037). Following attacks, NfL correlated with EDSS score at attack assessments (R=0.55; p<0.001); other biomarkers did not correlate with EDSS score after controlling for NfL levels.
In NMOSD, serum NfL, UCH-L1 and Tau levels were higher than in HC; increased baseline sGFAP levels were associated with greater attack risk. Although sGFAP levels showed the greatest increase following attacks, NfL correlated with attack-related disability.