Predicting ALS Survival Using Combined ALSFRS-R Slope and NfL: Insights from the ALS/ MND Natural History Consortium's Data and Biofluid Collection Study

Document Type

Conference Proceeding

Publication Date

10-6-2025

Publication Title

Muscle Nerve

Abstract

Introduction: The Amyotrophic Lateral Sclerosis Functional Rating Scale—Revised (ALSFRS-R) is the gold standard for measuring disease progression. Interest has grown in developing fluid progression biomarkers to assess risk and prognosis in ALS. We present findings on the effectiveness of exploratory fluid biomarkers as indicators of ALS disease progression. Methods: In this study we analyze data from the ALS/MND Natural History Consortium to evaluate three proposed blood biomarkers in disease progression models. Blood was collected from people living with ALS and analyzed for neurofilament light chain (NfL), plasma phosphorylated tau-181 (pTau181), and cardiac troponin T (cTnT). We used blood biomarker values measured at a single time point, along with the initial ALSFRS-R score and the slope between the first and last observed scores, adjusted for the interval between assessments. Analyses accounted for the delay from diagnosis to ALSFRS-R and blood sample collection. Associations between biomarkers and disease progression were studied using a linear regression model comparing ALSFRS-R slope to biomarker measurements. To study the relationship with overall survival, Cox proportional hazards models were fitted using biomarkers and the ALSFRS-R derived variables. The predictive performance of the models was assessed through the Integrated Brier Score (IBS), dynamic AUC, and C-index. Results: Data from 123 participants were available for analysis. NfL had a significant negative correlation with ALSFRS-R slope (r = −0.45, p < 0.001). NfL was also significantly associated with overall survival (HR=1.02, p = 0.015), accounting for the first ALSFRS-R measurement and its slope. Other biomarkers were not significantly associated with survival. The predictive capability of a model containing only the first observed ALSFRS-R score (C-index: 0.79) improved with the addition of NfL (C-index: 0.89). A model with the ALSFRS-R slope (C-index: 0.88) also showed improved predictive metrics after adding NfL (C-index: 0.90). Predictive metrics for these models were high when including NfL, showing the added value of a single NfL value at any point during disease course in predicting survival. Discussion: A single NfL value improved prediction of survival when compared with ALSFRS-R metrics alone. pTau181 and cTnT, while potentially useful as markers of phenotype, did not contribute significantly to predictions of overall survival.

Volume

72

First Page

S97

Find It @ Sladen

Share

COinS