Anavex Life Sciences Corp (AVXL)

[stockaesthetics]

This board, Stocktwits, YMB, and other boards have been littered with misinformation and poor analysis of the efficacy data from the Phase 2A trial, so I figured it was time for a clinical trial bootcamp (along with citations to show that these statements are facts and not just opinion).

First let's start off discussing the actual title of the clinical trial on clinicaltrials.gov: "Phase 2a Dose Finding, PK/PD and 12 Month Exploratory Efficacy Study of ANAVEX2-73 in Patients With Alzheimer's Disease (ANAVEX)." (citation: https://www.clinicaltrials.gov/ct2/show/NCT02244541) Exploratory efficacy, say that 10x to yourself to really instill the words "Exploratory efficacy" into our brains.

Let's also scroll down to the endpoints in the study. A quick primer on clinical endpoints is nicely detailed on its wikipedia site (citation: https://en.wikipedia.org/wiki/Clinical_endpoint). An important quote: "The primary endpoint of a clinical trial is the endpoint for which subjects are randomized and for which the trial is powered. Secondary endpoints are endpoints that are analyzed post hoc, for which the trial may not be powered nor randomized." This is a critical distinction.

Also, powered is an important word here, meaning statistical power. From another site: "Power analysis can be used to calculate the minimum sample size required so that one can be reasonably likely to detect an effect of a given size." Meaning you need to calculate the appropriate number of patients to include in the study to reject the null hypothesis (what we are testing). Again, remember the study is designed with statistical power to test the primary endpoint, and not necessarily powered to make conclusions from the secondary endpoints.

So why is that important for the discussion of the Phase 2a trial data? If you go back to the clinicaltrials.gov site and scroll down to the endpoints, you will see that the primary endpoint of the study (the endpoint for which the study is powered) is: "To determine maximum tolerated dose of Anavex2-73. [ Time Frame: 36 Days ] [ Designated as safety issue: No ]."

Several of the secondary endpoints were designed to test for EXPLORATORY efficacy (our new favorite words), such as MMSE, CBB, ADSC-ADL, among others. Again, these are post hoc analyses, and also there is a small number of n in the study since it was powered for dose-ranging and not for efficacy.

Many Phase 2a studies incorporate exploratory efficacy measurements, and this is important, as it gives the company preliminary information that may be useful when they are working with the FDA to determine how best to measure efficacy in larger, pivotal (Phase 2b, Phase 3, Phase 2/3) studies that are actually powered to test efficacy, almost always with a much larger subset of patients.

Can we really make conclusive determinations about the efficacy data here? No. You should be wary of anyone who tells you "this data is terrible!" or "this data is amazing, we have a cure!" I can tell you, as a clinical scientist whose job it is to discuss data from all phases of clinical trials to physicians in the field, I would most likely get fired if I made conclusive efficacy determinations based on data that is most likely not powered for statistical significance. I would have to specifically say something like "this is exploratory data, and although it shows a trend, we cannot make definitive conclusions from it."

Also, we have to remember that not only was the n very small, but since this trial was designed to figure out the correct dose, all of the patients from different doses were POOLED, and then assessed for efficacy. So say hypothetically, people from Group A received 1mg, people from Group B received 5mg, and people from Group C received 10mg, and the dose ranging study showed that 5mg was the most effective dose, the efficacy data was still analyzed by pooling all the group As, group Bs, and group Cs together. Imagine that all the Group As didn't respond because the dose was too low, and all the Group Cs were irritable and didn't feel well because of the high dose. Do you really think we are going to get the most perfect efficacy results? No, but we may get some useful information, which we did, where we showed that there were positive trends in certain cognitive and biologic measures that persisted over a period of approximately 31 weeks. Again, POOLED EXPLORATORY data that may not be powered to test for statistical significance. Not clean, but still useful to detect trends.

We figure out that the Group Bs do well with 5 mg, and we now have some info on the persistence of effect (a trend), so now we are going to design a larger trial (and even a pivotal registration study) that is powered for efficacy, and where efficacy is the primary endpoint of the study. This data is what we can use to make conclusive determinations about whether or not the drug is effective.

In conclusion, anyone who is telling you that the data is "bad" either is unaware of how clinical trials are designed (i.e. they are analyzing exploratory data from secondary endpoints in a Phase 2a trial as if they are efficacy data from primary endpoints of a larger Phase 2b/3 trial, how unfair is that?), or they are bashing and know they can instill fear in people who are investing in biotech but may not have the knowledge of how clinical trials work.

I will leave you with the PR we received in July that might make more sense now that we have the information above:

Dr. Norman Relkin, MD, PhD, an Alzheimer clinical trialist and an advisor to Anavex, commented: “To interpret the ANAVEX 2-73 results presented at the 2016 AAIC meeting, it is important to keep in mind the stated goals of this first in Alzheimer’s patients study. This was a Phase 2a study, primarily designed to determine which ANAVEX 2-73 dosages are safe to administer to mild to moderate stage Alzheimer’s patients. The study was successful in establishing the maximum tolerated dose and in revealing the range of ANAVEX 2-73 doses that are well-tolerated by Alzheimer patients. It also provided encouraging evidence that previously reported positive trends in certain cognitive and biologic measures persisted over a period of approximately 31 weeks. However, this analysis was based on pooled data from a relatively small number of subjects receiving a variety of doses. It is therefore unlikely that these findings reflect the full potential ANAVEX 2-73 in treating Alzheimer’s disease. It is unreasonable to draw conclusions about any limits to the long-term efficacy of ANAVEX 2-73 based on the interim Phase 2a findings, especially since no statistically significant decline from baseline was reported, which is impressive. Detailed pre-planned analysis of the pharmacodynamic results is in progress, which is one of the key factors of relevance for regulatory agencies and which will also determine the optimal dose for future studies.”