Long‐term Phase 1/2 intraspinal stem cell transplantation outcomes in ALS

Abstract

Abstract Objective: Intraspinal human spinal cord‐derived neural stem cell (HSSC) transplantation is a potential therapy for amyotrophic lateral sclerosis (ALS); however, previous trials lack controls. This post hoc analysis compared ambulatory limb‐onset ALS participants in Phase 1 and 2 (Ph1/2) open‐label intraspinal HSSC transplantation studies up to 3 years after transplant to matched participants in Pooled Resource Open‐Access ALS Clinical Trials (PRO‐ACT) and ceftriaxone datasets to provide required analyses to inform future clinical trial designs. Methods: Survival, ALSFRS‐R, and a composite statistic (ALS/SURV) combining survival and ALS Functional Rating Scale revised (ALSFRS‐R) functional status were assessed for matched participant subsets: PRO‐ACT n = 1108, Ph1/2 n = 21 and ceftriaxone n = 177, Ph1/2 n = 20. Results: Survival did not differ significantly between cohorts: Ph1/2 median survival 4.7 years, 95% CI (1.2, ∞) versus PRO‐ACT 2.3 years (1.9, 2.5), P = 1.0; Ph1/2 3.0 years (1.2, 5.6) versus ceftriaxone 2.3 years (1.8, 2.8), P = 0.88. Mean ALSFRS‐R at 24 months significantly differed between Ph1/2 and both comparison cohorts (Ph1/2 30.1 ± 8.6 vs. PRO‐ACT 24.0 ± 10.2, P = 0.048; Ph1/2 30.7 ± 8.8 vs. ceftriaxone 19.2 ± 9.5, P = 0.0023). Using ALS/SURV, median PRO‐ACT and ceftriaxone participants died by 24 months, whereas median Ph1/2 participant ALSFRS‐Rs were 23 (P = 0.0038) and 19 (P = 0.14) in PRO‐ACT and ceftriaxone comparisons at 24 months, respectively, supporting improved functional outcomes in the Ph1/2 study. Interpretation Comparison of Ph1/2 studies to historical datasets revealed significantly improved survival and function using ALS/SURV versus PRO‐ACT controls. While results are encouraging, comparison against historical populations demonstrate limitations in noncontrolled studies. These findings support continued evaluation of HSSC transplantation in ALS, support the benefit of control populations, and enable necessary power calculations to design a randomized, sham surgery‐controlled efficacy study.