Amylyx Pharmaceuticals' Relyvrio (sodium phenylbutyrate and taurursodiol) may offer a more substantial survival advantage in patients with amyotrophic lateral sclerosis (ALS) than indicated by initial trial results, according to a new analysis leveraging the Pooled Resource Open-Access ALS Clinical Trials (PRO-ACT) database. The post-hoc study, published in medRxiv, compared survival outcomes of patients treated with Relyvrio in the CENTAUR trial against a propensity-matched, treatment-naïve external control group from PRO-ACT, revealing a median overall survival benefit of 10.39 months.
The CENTAUR trial, a randomized, placebo-controlled study with an open-label extension (OLE), initially demonstrated a 4.8-month longer median overall survival for patients receiving Relyvrio compared to placebo (HR, 0.64; 95% CI, 0.42–1.00; p = 0.048). However, the OLE phase introduced a significant crossover effect, with 71% of patients initially on placebo switching to Relyvrio, potentially underestimating the drug's true impact on survival.
Addressing Crossover with External Controls
To mitigate the impact of crossover, researchers conducted a post-hoc survival analysis comparing Relyvrio-randomized participants from CENTAUR with a matched external control cohort from the PRO-ACT database. The PRO-ACT database, the largest ALS clinical trials database, incorporates anonymized longitudinal data from 11,675 participants in 29 Phase 2 and 3 ALS clinical trials.
The analysis involved 89 participants in the CENTAUR PB and TURSO group, 85 of whom had a match in the PRO‐ACT external control group on 1:1 propensity score matching using a caliper width of 0.4. Propensity score matching adjusted for key baseline covariates known to influence ALS prognosis, including time since symptom onset, pre-baseline ALSFRS-R slope, vital capacity, and age at study baseline.
Significant Survival Extension Observed
Results showed a median overall survival of 23.54 months in the CENTAUR Relyvrio group compared to 13.15 months in the PRO-ACT external control group, representing a 10.39-month difference. The hazard of death was 52% lower in the Relyvrio group (HR, 0.48; 95% CI, 0.31–0.72; p = 0.00048).
Sensitivity analyses using different caliper widths, the full PRO-ACT analysis set without propensity score matching, and the full set with propensity score inverse probability treatment weighting (IPTW) all yielded consistent results.
Methodological Rigor and Limitations
The researchers emphasized the importance of careful methodology when using external controls, including matching eligibility criteria and known prognostic factors. Propensity score matching was employed to minimize bias due to imbalance among nonrandomized cohorts or other trial variables.
The authors acknowledge limitations, including the potential for missed death events in clinical trial participants due to loss to follow-up and the lack of data on respiratory support or gastrostomy tube placement in the PRO-ACT database. Variations in clinical practice over time may have also contributed to confounding.
Implications and Future Directions
Despite these limitations, the study suggests a potentially greater survival benefit with Relyvrio in ALS than previously observed in the CENTAUR trial's intent-to-treat analysis. The ongoing global Phase 3 PHOENIX trial (NCT05021536) will further evaluate Relyvrio's effect on survival and other outcomes in a larger ALS population over a longer duration.
"Analyses using external controls may provide additional context for survival outcomes in future trials in ALS," the authors concluded. "Adherence to robust methodology, particularly selection of optimally matched external controls, is pivotal when performing such analyses."
