Turning Setbacks into Strategies: Lessons from RELYVIRO’s journey in ALS
In this article and the next we look at the recent developments for new ALS treatments. To begin with we focus on what many had hoped would be a breakthrough: the story of RELYVIRO (AMX0035). Its development was marked by successes and setbacks, and ultimately resulted in failure with the product being withdrawn from the market by its manufacturer Amylyx earlier this year. Our purpose here is to learn from this in order to inform future drug development in rare diseases. In the second article to follow, we will look at the late-stage ALS pipeline and see how sponsors are responding to these challenges and the promise they hold for improving the lives of ALS patients.
ALS: a devastating disease
ALS is a progressive neurodegenerative disease that affects motor neurons in the brain and spinal cord, leading to muscle weakness, atrophy, and eventual paralysis. Most patients succumb to respiratory failure within 2–5 years. Treatments are limited with RILUZOLE, a glutamate antagonist, and EDARAVONE, a free radical scavenger, offering modest benefits in slightly slowing disease progression. QALSODY (tofersen) an oligonucleotide antisense therapy is also approved but only for patients with SOD1 gene mutations, representing a small proportion of the overall ALS population.
The causes of ALS are still unclear, although a range of interrelated mechanisms are understood to contribute, including:
- Excitotoxicity caused by excessive glutamate overstimulation of neurons
- Oxidative stress driven by an imbalance in reactive oxygen species
- Dysfunctional mitochondria
- Protein aggregation, particularly of TDP-43
- Chronic neuroinflammation mediated by activated microglia and astrocytes
- Genetic mutations e.g., in C9orf72, SOD1, and TARDBP
- Environmental triggers e.g., exposure to toxins, physical trauma
The range of contributing factors means that treatment is likely to require a multifaceted approach.
Learnings from AMX0035
AMX0035, a fixed-dose combination of sodium phenylbutyrate and taurursodiol, was designed to target mitochondrial dysfunction and endoplasmic reticulum stress—two key pathways implicated in ALS.
Amylyx sought its IND in April 2017 and several years later in September 2022, secured FDA approval based on a single Phase 2 study. Shortly after obtaining its IND, Amylyx obtained Orphan Drug Status for AMX0035, and continued to engage routinely with FDA over the course of development. Amylyx also applied for both Fast Track Designation and Breakthrough Status but was unsuccessful. FDA’s rationale for turning down these applications was that the planned Phase 2 study did not enable a clear benefit to be demonstrated over the currently approved therapies.
Despite this, FDA granted AMX0035 Priority Review in December 21 following the completion of its Phase 2 CENTAUR study. In March 22, the FDA’s Peripheral and Central Nervous System Drugs Advisory Committee (PCNSDAC) voted against approval of AMX0035 by 6-4 votes. However, in an unusual move, the committee met again in September of that year following a submission of additional analysis and data from Amylyx. Upon reviewing this information, the PCNSDAC voted in favour of approval by 7-2. What then led to this change of heart by the FDA?
Phase 2 CENTAUR study
The pivotal Phase 2 CENTAUR trial enrolled 137 ALS patients and evaluated AMX0035 over 24 weeks. The primary endpoint was the change in ALS Functional Rating Scale-Revised (ALSFRS-R) scores. The study showed a statistically significant improvement in the primary end point of 2.32 pts (p=0.034) relative to placebo, with positive improvements (but not statistically significant) on some secondary endpoints.
From its March 22 meeting, FDA expressed several criticisms: the data at 24 weeks showed “modest results on primary endpoint with limited support from any secondary endpoints.” and “no survival benefit at 24 weeks”. FDA expressed concern over the quality of the data and analysis including the appropriateness of the efficacy analysis i.e., whether the actual decline of patients was linear over time, as well as study conduct issues such as the potential for unblinding and randomisation implementation problems.
Data from the open label extension demonstrated an improvement in the composite time to survival of 4.8 months (p=0.045), however FDA felt this was not persuasive given a range of factors that complicated the interpretation of this data including limited sample size and baseline imbalances. (1)
Following this, Amylyx submitted additional analysis including individual responder analysis, post-hoc survival sensitivity analysis using external natural history data and a statistical model, as well as data for Cerebral Spinal Fluid (CSF) biomarkers from its PEGASUS Alzheimer’s Disease trial. These data and analyses combined with the ongoing Phase 3 PHOENIX trial, the relatively safe profile of the drug, and the high level of unmet need are all likely to have contributed to the FDA choosing to exercise more flexibility in its decision making, ultimately leading to its approval. (2, 3)
An important point that should not be overlooked is the role played by the patient organisations. For instance, the ALS Association had been extremely active throughout not only through the FDA review but also the development of AMX0035. It is likely that their engaging and passionate campaigns contributed to the case for more flexibility by the FDA.
Decision to withdraw from the market: failure of Phase 3 PHOENIX trial
Despite a successful approval by the FDA, the Phase 3 PHEONIX trial did not go as expected. The global 48-week study aimed to confirm AMX0035’s efficacy and safety in 664 ALS patients, significantly larger than the CENTAUR trial. The primary endpoint was the change in ALSFRS-R scores from baseline, with secondary endpoints including survival and respiratory function. Disappointingly, the results did not support those from the Phase 2 study:
- Primary Endpoint: The trial failed to show a statistically significant difference in ALSFRS-R scores compared to placebo (p=0.667)
- Secondary Endpoints: No significant improvements were observed in survival, respiratory function, or quality of life measures
Furthermore, in the analysis of the patient sub-set that met the CENTAUR trial patient criteria, no significant difference was observed relative to placebo.
Despite the disappointing efficacy results, AMX0035 was found to be well-tolerated, with no new safety concerns. (4)
Regulators and payers have frequently been cautious when approving and agreeing access to drugs that have limited data e.g., from small sample Phase 2 studies. It is possible that these results may have reinforced their perspective on this matter, something that should be borne in mind by drug developers following a similar regulatory/launch pathway.
Why did AMX0035 fail in Phase 3?
We are still waiting to see the full data from the PHOENIX trial, and so questions around the underpinning science of AMX0035 remain to be answered. However, it is clear that the Phase 3 study did not reflect what was observed in the previous smaller patient sample. It’s likely that the concerns originally raised by the FDA in its March 22 meeting regarding weaknesses in the data and analysis had merit and should have had greater weight in the final approval decision.
A related factor could be the heterogeneity of the ALS population which manifests in different progression rates and phenotypes across patients. A larger, more diverse trial population such as the PHOENIX study may well have masked the treatment effect observed in Phase 2. However, as we saw earlier, the patient subgroup from the PHOENIX study that matched the CENTAUR sample showed no significant improvement, raising questions on the validity of the Phase 2 data.
What can rare disease drug developers learn from the AMX0035 journey?
- Invest well in developing a robust study design, execution of the study and statistical analysis. This is easier to say than do, however it is essential if you are looking to gain approval from phase 2 studies with small samples of patients
- Engage early and frequently throughout the development journey with regulators. Listen to their concerns and accommodate their needs where possible and practical. Take advantage of opportunities to respond to their issues raised with additional data and analysis
- Build supporting evidence using real world and natural history data, taking care to ensure the best match with your clinical data
- Align with the needs of patient associations on common goals and collaborate where appropriate on research programs
- Assess the therapeutic relevance of your drug to specific patient subgroups to address patient heterogeneity
Conclusion
The withdrawal of AMX0035 underscores the complexities of ALS drug development and the importance of robust confirmatory data. For the biotech and pharma sectors, there are a rich set of lessons to be learned. In the next article we will see how sponsors are continuing to respond to the challenges in ALS.
Contact
If you are interested in discussing any of the issues above for your company/drug development program, please contact me through my email address dniven@nivenbiopharma.com . Feel free to also visit my website at www.nivenbiopharma.com for more information. I have no conflicts of interest in the production of this article.
References
- AMX0035 (presentation), FDA Peripheral and Central Nervous System Drugs Advisory Committee March 30 2022
- FDA Advisory Committee Votes to Recommend AMX0035 for Treatment of ALS, Marco Meglio, Neurology Live, Sep 2022
- FDA Briefing Document, Peripheral and Central Nervous System Drugs Advisory Committee (PCNS) Meeting September 7, 2022, CDER, FDA
- Amylyx Pharmaceuticals Announces Topline Results From Global Phase 3 PHOENIX Trial of AMX0035 in ALS, Amlyx Press Release, March 2024
