Résumé
Background Amyotrophic lateral sclerosis (ALS) is a life-threatening disease characterised by progressive loss of motor neurons with few therapeutic options. The MIROCALS study tested the hypothesis that low-dose interleukin-2 (IL-2LD) improves survival and function in ALS. Methods In this randomised, double-blind, placebo-controlled trial, male and female riluzole-naive participants, with either a possible, laboratory-supported probable, probable, or definite ALS diagnosis (revised El Escorial criteria), aged 18-76 years, with symptom duration of 24 months or fewer, and slow vital capacity of 70% or more, underwent a riluzole-only 12-18 week run-in period before randomisation in a 1:1 ratio to either 2 million international units (MIU) IL-2LD or placebo by subcutaneous injection daily for 5 days every 28 days over 18 months. The primary endpoint was survival at 640 days (21 months). Secondary outcomes included safety, ALS Functional Rating Scale-Revised (ALSFRS-R) score, and biomarker measurements including regulatory T-cells (Tregs), cerebrospinal fluid (CSF)phosphorylated-neurofilament heavy-chain (CSF-pNFH), and plasma and CSF-chemokine ligand 2 (CCL2). The primary endpoint analysis used unadjusted log-rank and Cox's model adjusted analyses using pre-defined prognostic covariates to control for the disease and treatment response heterogeneity. The study was 80% powered to detect a two-fold decrease in the risk of death by the log-rank test in the intention-to-treat (ITT) population, including all randomly allocated participants. MIROCALS is registered with ClinicalTrials.gov (NCT03039673) and is complete. Findings From June 19, 2017, to Oct 16, 2019, 304 participants were screened, of whom 220 (72%) met all criteria for random allocation after the 12-to-18-week run-in period on riluzole. 136 (62%) of participants were male and 84 participants (38%) were female. 25 (11%) of the 220 randomly allocated participants were defined as having possible ALS under El Escorial criteria. At the cutoff date there was no loss to follow-up, and all 220 patients who were randomly allocated were documented as either deceased (90 [41%]) or alive (130 [59%]), so all participants were included in the ITT and safety populations. The primary endpoint unadjusted analysis showed a non-significant 19% decrease in risk of death with IL-2LD (hazard ratio 0<middle dot>81 [95% CI 0<middle dot>54-1<middle dot>22], p=0<middle dot>33), failing to demonstrate the expected two-fold decrease in risk of death. The analysis of the primary endpoint adjusted on prognostic covariates, all measured at time of random allocation, showed a significant decrease of the risk of death with IL-2LD (0<middle dot>32 [0<middle dot>14-0<middle dot>73], p=0<middle dot>007), with a significant treatment by CSF-pNFH interaction (1<middle dot>0003 [1<middle dot>0001-1<middle dot>0005], p=0<middle dot>001). IL-2LD was safe, and significantly increased Tregs and decreased plasma-CCL2 at all timepoints. Stratification on CSF-pNFH levels measured at random allocation showed that IL-2LDwas associatedwith a significant 48% decrease in risk of death (0<middle dot>52 [0<middle dot>30-0<middle dot>89], p=0<middle dot>016) in the 70% of the population with low (750-3700 pg/mL) CSF-pNFH levels, while in the 21% with high levels (>3700 pg/mL), there was no significant difference (1<middle dot>37 [0<middle dot>68-2<middle dot>75], p=0<middle dot>38). Interpretation With this treatment schedule, IL-2LD resulted in a non-significant reduction in mortality in the primary unadjusted analysis.
However, the difference between the results of unadjusted and adjusted analyses of the primary endpoint emphasises the importance of controlling for disease heterogeneity in ALS randomised controlled trials. The decrease in risk of death achieved by IL-2LD therapy in the trial population with low CSF-pNFH levels requires further investigation of the potential benefit of this therapy in ALS. Copyright (c) 2025 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY 4.0 license.