Results from the Global Phase 3 PHOENIX Trial Evaluating Sodium Phenylbutyrate and Taurursodiol in ALS
Source: View publication →
The phase 3 PHOENIX trial failed to confirm the efficacy of AMX0035 in amyotrophic lateral sclerosis, missing all primary and secondary endpoints and leading to the drug's voluntary market withdrawal.
Key Findings
Study Design
Study Limitations
Clinical Significance
The negative results of the PHOENIX trial represent a major clinical disappointment, definitively establishing that AMX0035 (Relyvrio/Albrioza) does not alter ALS disease progression. Consequently, the drug was voluntarily withdrawn from the market in early 2024. The trial underscores the critical necessity of conducting large, adequately powered phase 3 confirmatory trials for rare disease therapeutics that receive accelerated or conditional approvals based on phase 2 data.
Historical Context
AMX0035 received full FDA approval in September 2022 following intense patient advocacy and a highly unusual, reconvened FDA advisory committee meeting that narrowly supported the drug based on the 137-patient phase 2 CENTAUR trial (which showed a 25% slower decline in ALSFRS-R and a post-hoc survival benefit). Because the data were viewed by many regulators as borderline, Amylyx pledged to voluntarily withdraw the drug if the larger phase 3 PHOENIX trial failed. When PHOENIX missed all endpoints, the manufacturer honored this commitment, establishing a landmark precedent for post-market regulatory accountability and the withdrawal of definitively ineffective neurodegenerative treatments.
Guided Discussion
High-yield insights from every perspective
What is the proposed mechanism of action of sodium phenylbutyrate and taurursodiol (AMX0035) in ALS, and how does this relate to the underlying cellular pathophysiology of the disease?
Key Response
AMX0035 was designed to simultaneously target endoplasmic reticulum (ER) stress through sodium phenylbutyrate and mitochondrial dysfunction through taurursodiol. Understanding this mechanism helps connect basic cellular biology of motor neuron degeneration to targeted therapeutic strategies.
How should you counsel a patient with ALS who is currently taking AMX0035 (Relyvrio) regarding the results of the PHOENIX trial and its subsequent market withdrawal?
Key Response
Residents must navigate difficult conversations about stopping therapies. Counseling should include explaining the lack of efficacy demonstrated in the Phase 3 trial, assisting with safe discontinuation, and pivoting the focus to other evidence-based disease-modifying therapies (like riluzole) and robust multidisciplinary symptom management.
The FDA originally approved AMX0035 based on the Phase 2 CENTAUR trial. As a neuromuscular fellow, how do you reconcile the conflicting efficacy results between the CENTAUR and PHOENIX trials, and what does this teach us about ALS trial design?
Key Response
Differences likely stem from patient heterogeneity; CENTAUR enrolled faster-progressing patients, while PHOENIX had broader inclusion criteria. This highlights the statistical risks of regulatory approval based on a single Phase 2 study and the critical need for robust, adequately powered Phase 3 data in highly variable neurodegenerative diseases.
The withdrawal of Relyvrio removes a heavily discussed treatment option for ALS. How does this outcome reshape our approach to shared decision-making regarding early access or accelerated approval drugs in fatal neurodegenerative diseases?
Key Response
Attendings must balance patient hope for novel therapies with the reality of incomplete efficacy data and potential financial or physical toxicity. This case serves as a paradigm for discussing the uncertainty of accelerated approvals, emphasizing the importance of managing expectations and prioritizing quality of life.
Scholarly Review
Critical appraisal through the lens of expert reviewers and guideline development
How did the statistical powering, patient heterogeneity, and primary endpoint selection (ALSFRS-R slope) in the PHOENIX trial impact its ability to detect a drug effect compared to the earlier CENTAUR trial?
Key Response
ALSFRS-R trajectories are highly variable and non-linear. Evaluating whether the mixed models for repeated measures (MMRM) or joint rank analyses adequately accounted for slower disease progression and missing data in the broader PHOENIX cohort is critical for designing and powering future ALS clinical trials.
In reviewing the PHOENIX trial manuscript, how would you critically evaluate the risk of unblinding due to the known gastrointestinal side effect profile of the treatment, and could this have influenced the subjective components of the ALSFRS-R?
Key Response
A critical reviewer would flag whether GI side effects disproportionately unblinded the treatment arm. While unblinding typically biases results toward positive outcomes, evaluating this ensures the negative result wasn't confounded by dropout imbalances or reporting bias, confirming that the failure to meet endpoints was a true lack of biological efficacy.
Given the definitive negative results of the PHOENIX trial and subsequent market withdrawal, what specific updates must be made to ALS management guidelines regarding AMX0035, and what level of evidence supports this change?
Key Response
Previous consensus guidelines had tentatively included AMX0035 based on Phase 2 data. The committee must now issue a strong recommendation against its use, citing Level A evidence (a well-powered Phase 3 RCT showing lack of efficacy), formally removing it from algorithms and refocusing guidelines on riluzole, edaravone, and symptomatic care.
Clinical Landscape
Noteworthy Related Trials
Riluzole ALS Trial
Tested
Riluzole 100 mg daily
Population
Patients with probable or definite ALS
Comparator
Placebo
Endpoint
Survival without tracheostomy
Edaravone Phase 3 Trial (MCI186-19)
Tested
Edaravone 60 mg IV
Population
Patients with early-stage ALS and preserved respiratory function
Comparator
Placebo
Endpoint
Change in ALSFRS-R score at 24 weeks
CENTAUR Trial
Tested
Sodium phenylbutyrate and taurursodiol (AMX0035)
Population
Adults with definite ALS within 18 months of symptom onset
Comparator
Placebo
Endpoint
Rate of decline in the ALS Functional Rating Scale-Revised (ALSFRS-R) score
Tailored to your role
Want this tailored to you?
Add your specialty or training stage to get role-specific takeaways and more questions.
Personalize this analysis