Recent IPO(Jan 7th) Amylyx pharmaceuticals has developed a drug(AMX0035) for the treatment of ALS. They believe it is the first to both slow decline and lengthen the life of ALS patients. We believe the drug is inert and will fail its upcoming Phase 3 Phoenix trial. The company has made the extraordinary claim that they believe data from the PHOENIX trial will not be required for the FDA to make a determination on the approval of AMX0035 for the treatment of ALS based on their discussions with the FDA. This comes despite a long history of failure in ALS drug clinical trials(https://ncbi.nlm.nih.gov/pmc/articles/PMC5360725/… ) and an extremely weak Phase 2 clinical trial(CENTAUR). The study contained only 137 patients (48 on placebo and 87 on AMX0035) We will deep dive into it here as all of the value of the company hinges on the trial results.
The main claims to fame of the study are the longer survival time of AMX0035 patients vs placebo(25 months vs 18.5) and the slower rate of decline in ALFRS-R score(a 48 point scale to measure 12 aspects of physical function in ALS patients, a lower score indicates more decline). The study barely met a key level of significance(p=0.05) for both measures(p-value for the rate of ALFRS-R decline=0.03, 0.023 for survival). The key problems, however, arise with the comparison of the results with existing literature and previous clinical studies for ALS.
The baseline characteristics of the patients are notable as they are needed for comparison to other studies. Average age =57.5(approx same for placebo and drug). Average time since symptom onset=13.6 for placebo, 13.5 for drug). Thus, the median survival time for placebo patients from the time since symptom onset is 32.1(18.5+13.6). Similarly, for AMX0035 it's 38.5(25+13.5). This is impressive in isolation. However, there are a large number of studies showing much longer survival times even before Riluzole. Riluzole(FDA approved:1995) is generally accepted as the only drug that meaningfully extends the life of ALS patients, typically by a few months. It was taken by many patients in the study. I will summarise the important studies but a review can be found here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3515205/.
1978 study(https://tinyurl.com/yckukx47 n=668, age at onset=56, survival from onset=36 months), 1985(https://tinyurl.com/4msdkvw8 n=318,age=56.6,survival=36 months), 1989 (https://tinyurl.com/4w466vm6, n=194, age=59, survival=36), 1993 (https://tinyurl.com/ycknepha, n=708, age=58.5,survival=38). (Riluzole now active)1995(https://tinyurl.com/bdedev4m, n=831, age=55.7, survival=35), 1998( https://tinyurl.com/mrytn2au, n=62, age=57,survival=38.4), 2002(https://tinyurl.com/52eh5kbj, n=841, age=56, survival=43), 2004(https://tinyurl.com/2p96sx8c, n=793, age=5, survival=34.8). 2008(https://tinyurl.com/47fsu9um, n=47, age=58, survival=42.8).
Given these studies, many of which have much larger sample sizes, we find it extremely unlikely that $AMLX will be able to prove statistically significant survival benefits given their median 38.5 months for survival. The other key claim of the 6-month study is the lower rate of decline in the ALSFRS-R score for AMX0035(-1.24 score change per month) vs placebo(-1.66 score change per month). In isolation, this is impressive. Compared to previous trials and literature, it is not.
A phase 3 trial testing minocycline vs placebo for efficacy in P3 trials https://tinyurl.com/446dnjhn(We are more interested in placebo for all these examples) used 206 placebo patients(158 more than AMLX), age=57.5, months since symptom onset=18(4.5 months more than $AMLX phase 2). This showed a -1.04 score change per month for placebo. Better than both the placebo and drug for AMLX P2 trial despite lasting 3 months longer and having patients with longer time since symptom diagnosis. An NP001 P2 trial for ALS https://tinyurl.com/2p96zhsc had placebo n=42, age=53.7, time since symptom onset=17 months and showed a placebo -0.88 change in score per month. Again, better than either placebo or the AMX0035 drug in the CENTAUR trial. A perampanel P2 trial for ALS https://tinyurl.com/mr2hvv3n had placebo n=22, age=62.6, 48 weeks long(11 months), 15.3 months since symptom onset, and a -9 total change for a per month change of -0.8 per month. Again, better than CENTAUR. A P2 Tirasemtiv P2 Trial for ALS https://tinyurl.com/37vtnxzc had placebo n=210, age=56.8, time since symptom onset=26.7 months and a -2.4 points change in 12 weeks for a -0.8 change per month. Finally, a Reldesemtiv P2 ALS Trial https://tinyurl.com/muxw3vcj had a placebo n=115, age=59.6, time since symptom onset=22.1 months, and a -3.5 change over 3 months for a -1.16 change per month. About equivalent to the average change for AMX0035(Thought still better).
We believe that all of these studies point to the drug being inert and that the trial will not meet the level of significance needed for the larger sample size planned in the planned phase 3 trial called PHOENIX n=600. The PDUFA date has been accelerated to June 29th. Even with positive feedback from the FDA, successful trial results will be needed. Projected sales for Riluzole for 2021 are $220M. Thus, even if AMX0035 were already approved, Amylyx would still be trading at 5.5x Riluzole sales.
AMX0035 is also planned for the treatment of Alzheimer's. This is another disease with a huge graveyard of failed drugs. If the phase 3 PHOENIX trial fails, expect 70%+ downside from the current $1.2B valuation given they are currently relying on grant revenue(650k in 2020) and experience a FY20 net loss of 42MM.
I do not hold a position with the issuer such as employment, directorship, or consultancy.
I and/or others I advise do not hold a material investment in the issuer's securities.
Continued cash burn from operations with only grant revenue.
Negative results from phase 3 trial topline results showing AMX0035 is inert.
Increasing short selling from institutional investors(ALS/Alzheimer's candidates tend to be heavily shorted given their history of failure.)