To evaluate high efficacy early treatment (HEET) versus low efficacy early treatment (LEET) paradigms on serum neurofilament light chain (sNfL) and serum glial fibrillary acidic protein (sGFAP) outcomes in a real-world MS cohort.

There are limited studies on the applicability of sNfL and sGFAP to inform the HEET or LEET approach.

Patients enrolled in the CLIMB study treated with HEET (fingolimod, natalizumab, ocrelizumab, rituximab) or LEET (dimethyl fumarate, glatiramer acetate, interferons, teriflunomide) within five years of symptoms onset were divided into two cohorts. Cohort A [HEET (n=99), LEET (n=89)] had sNfL and sGFAP 3-0 years before and 0.5-3 years after treatment. Cohort B [HEET (n=56), LEET (n=158)] had two samples after treatment start (0-4 and 1-5 years). A linear mixed-effects model adjusted for age and sex was used to estimate the difference in the change with time in log-transformed sNFL/sGFAP comparing the HEET and LEET groups.

In both cohorts, patients in the HEET group were older at baseline (Cohort A: mean age 35.9±9.1 vs. 39.4±9.9, p=0.01; female 70.7% vs. 70.8%, p=1.0; EDSS 1.5±1.5 vs. 1.2±1.2, p=0.1; sNfL 1±0.9 vs. 0.98±0.8, p=0.85; sGFAP 3.1±0.53 vs. 3.1±0.45, p=0.62), (Cohort B: age 35.1±9.3 vs. 39.2±10.6, p= 0.01; female 69.6% vs. 66.5%, p=0.6; EDSS 1.8±1.4 vs. 1.1±1.0, p<0.01; sNfL 0.75±0.7 vs. 0.93±0.6, p=0.09, sGFAP 2.97±0.5 vs. 3.2±0.6, p=0.01). In Cohort A, the adjusted linear mixed-effects model showed more pronounced decrease of sNFL and sGFAP in the HEET group (-0.082, p=0.076; and -0.035, p=0.12, respectively). In cohort B, sNFL and sGFAP trended to increase more in the HEET group (0.075, p=0.050, and 0.034, p=0.17, respectively). 

In cohort A, with pretreatment levels, sNFL and sGFAP trended to decrease more rapidly in HEET compared to LEET. Thus, pretreatment biomarkers followed longitudinally may inform prognostication and monitoring of the HEET or LEET approach; additional randomized studies are necessary.