Precious Metals

[gfp927z]

Axsome - >>> A Last-Minute Look At AXS-05 Revealed That Axsome Therapeutics May Have A Better Risk/Reward Profile In Its Major Depressive Disorder Pipeline


Seeking Alpha

Dec. 9, 2019

WS BioPharma Wolf



https://seekingalpha.com/article/4311382-last-minute-look-axsminus-05-revealed-axsome-therapeutics-may-better-risk-reward-profile-in


Previous scientific literatures did not show consistent conclusions about the effectiveness of one of AXS-05's active ingredient dextromethorphan on MDD.

The recent failure of SAGE-217 actually provides some insights which favor AXS-05.

AXS-05's phase III (GEMINI) result is coming very soon. One less opponent, less competition, and larger market potential reshape AXS-05's risk-reward profile.

DM is a very common OTC antitussive medicine for treating cold and cough, sold under the form of syrup, tablet, and spray. Being an OTC medicine, it has a good safety profile with a cheap price. DM has multiple mechanisms of action, including non-selective serotonin reuptake inhibitor and a sigma-1 receptor agonist. DM and its major metabolite, dextrorphan (DX) also act as an NMDA receptor antagonist. Altogether, it produces additional neuropsychiatric effects which are sedative, dissociative, and stimulant. In 2010, it was approved by FDA to treat pseudobulbar effect (uncontrolled crying/ laughing).

Axsome Therapeutics (NASDAQ:AXSM) is currently testing DM in combination with bupropion, an approved medicine for MDD in the Phase III GEMINI trial. Its Phase II trial ASCEND study showed an almost perfect result that I am not going to re-state here.

Why previously I was not convinced about AXS-05

Given that ASCEND result was so clear and indisputable, people would naturally ask why I was still not convinced at first. Well, it's because research literature does not generate clean evidence: most pre-clinical and animal study proposed DM's strong anti-depressive potential but scant human trials may tell us otherwise. Let me elucidate.

First, preclinical studies did find that DM exerts antidepressant-like effects in mice using the forced swim test and tail suspension test, two of the most validated and widely used animal models for predicting antidepressant efficacy.

Further, based on its mechanisms of action, DM would naturally be considered to have anti-depressive effect. In various in-vitro studies, DX was observed to have stronger binding affinity to NMDA receptors comparing to ketamine, indicating a potential rapid-acting anti-depressive effect. (DM's binding affinity to NMDA receptors is more controversial).

The sigma-1 agonist effect has been reported to demonstrate rapid antidepressant effects in several studies as well, but again, majority of the studies are pre-clinical and human trials do not generate strong evidence. For example, the sigma-1 ligand igmesine showed antidepressant efficacy under double-blind conditions in a UK population, but not in Czechs or Poles. So being a sigma-1 agonist, DM serves potential on this pathway but not 100% convinced.

One of the most negative evidence comes from a 2019 research literature published by a group of researchers in Cleveland Clinic Akron General and Northeast Ohio Medical University. The small study included 40 patients, about half the size of ASCEND trial. Being fully-aware of DM's fast metabolism, the study used DM pairing with various cytochrome P450 (CYP) 2D6 inhibitors like bupropion, fluoxetine, or paroxetine to extend its plasma concentration and half-life. The population selection is shown in Fig.1.

Figure 1. The patient population selection. Note DXM is the same as DM. (Source: Internet)

The study result is shown in Fig. 2, where the DM did not show a statistically significant improvement in depressive symptoms compared to control. In fact, DM may show a negative correlation to symptom improvement, as it prolongs the time to improvement from nurse documentation (Fig. 2).

Figure 2. Summary of primary efficacy outcome. (Source: Internet)

It is clear that this study had a very different design comparing to ASCEND or GEMINI. For example, the dosing on DM is not consistent across treatment group, with most common dose being 30mg. Also, neither treatment nor control group use bupropion consistently. One may argue that the 30mg DM dose is too low, but the study also found higher instances of perceptual disturbances and delusions in the DM group compared with the control group, potentially indicating that DM was taking some effect. People would also argue that the study's primary endpoint, the time to improvement is very different from remission rate. But intuitively, these two metrics should be positively correlated.

In a different study, however, the conclusion is opposite. A single patient with treatment-resistant MDD was treated with 600 mg/day of bupropion and a 300 mg oral loading dose of DM, followed by 60 mg po bid after an additional dose-finding period. In this single case, DM exhibited a ketamine-like rapid-acting antidepressant effect, thought to be mediated by mTOR and 5HTT inhibition, resulting in AMPA receptor trafficking, and dendritogenesis, spinogenesis, synaptogenesis, and increased neuronal survival. Of course, this study only observed one single patient thus the result may not easily be generalized to entire MDD population.

In conclusion, based on research literatures, DM can be a potentially strong anti-depressant, but it may need more consistent physiological evidence.

What did SAGE-217's failure tell us

On Dec. 5th, Sage Therapeutics (NASDAQ:SAGE) shocked the market with its Phase III MDD trial candidate SAGE-217 failed its primary endpoint. The result was completely unexpected. After carefully examining the data, we think SAGE-217 is an okay MDD drug, but the relative weak efficacy and questionable trial design cause its MOUNTAIN trial failure.

1. Sage's Phase II data is not all clean. In its Phase II data, under treatment group, the HAM-D change from baseline actually worsened after day 15 (Fig. 3). Further, its day 35 and day 42 readout shows insignificant difference. These negative signs were not seen in ASCEND trial, where AXS-05 treated patients experienced continuous improvement all the way through week 6 (Fig. 5). This potentially indicates preferable efficacy of AXS-05 over Sage-217.

Figure 3. Sage-217 Phase II primary outcome. (Source: Internet)

2. Both Sage-217's Phase II and Phase III trials used placebo as control. But AXS-05's Phase II actually used bupropion as control. Being a known antidepressant, bupropion would naturally augment the improvement in control cohort, but ASCEND still managed to produce a significant result. Now, in the Phase III GEMINI study, with placebo as control back in place, one should reasonably expect GEMINI result would turn out better than ASCEND.

Another way to establish a hypothetical direct comparison is to use the conversion between MADRS and HAM-D scores. In this research literature, it was stated that "An absolute HAM-D improvement of 10, 20, 25 points corresponded to a MADRS improvement of 12, 26, and 34". Now, if we interpolate ASCEND's treatment cohort of 17.2 points improvement in MADRS score into HAM-D score, it would yield an improvement of (17.2-12)/14*10+10 = 13.7 points, which is much better than Sage-217's day 15 outcome. Of course, such a comparison is not too fair. To illustrate, Sage-217 Phase II trial result also contains MADRS improvement, and in its placebo cohort, patients in average improved 17.2 MADRS points at week 6 (Fig. 4), which is the same as AXS-05 treatment effect.

Figure 4. Sage-217 different endpoints and results. (Source: Internet)

3. Sage chose day 15 as its primary endpoint purely based on the maximum separation between treatment and control in its Phase II data. I think this decision may be a bad idea if trial designer's goal was solely on maximizing the chance of success. In its Phase II trial, there were no readings between day 8 and day 15 and people were aware that starting from some time close to day 15 (could be day 11 or day 13), Sage-217 started to lose efficacy. In fact, it is day 6 that has a much higher chance to succeed because multiple readings are taken before and after that time point, and those "surrounding" data all show significant differences. That is why MOUNTAIN failed on 15 days, but statistical significance on HAM-D scale was actually achieved at days 3, 8, and 12. Of course, trial designers need to balance between duration of response (directly affecting market potential) and chance of success, but simply choosing day 15 as its primary outcome still seems lack of further support (or no other better option).

Axsome, on the other hand, chose a different approach. Although its phase II ASCEND study showed largest separation between treatment group and control at week 3 (see Fig. 5), it does not limit itself to it. The GEMINI actually targets at week 6. This is quite a bold decision since ASCEND trial only showed a mediocre separation (P = 0.013), leaving a very small buffer. But given the evidence that AXS-05 continuously benefits patients towards week 6, we think it exists higher probability that AXS-05 may be successful.

Figure 5. The primary outcome of Phase II ASCEND trial. (Source: Internet).

4. Sage stated that approximately 9% of patients in the SAGE-217 30 mg group had no measurable drug plasma concentration, consistent with non-compliance in taking SAGE-217. This may due to its critical flaw: its slow onset effect. For MDD patients, keeping a scheduled dose is not easy. If it is slow onset, it will make them harder to follow the schedule. AXS-05, on the other hand, due to its NMDA agonist feature (just like ketamine), may potentially display a rapid-onset effect (although not consistently showed in previous human trials).

Further, given such a surprising number of patients missed the dosing, it remains possible that some of those patients did actually take the medicine. They may just have a very fast SAGE-217 metabolism rate. For Axsome investors, this is less of a concern because AXS-05's Phase I result clearly revealed that bupropion increases DM's plasma concentration by over two orders of magnitudes (Fig. 6).

Figure 6. AXS-05 phase I result showing BUP greatly increases DM's plasma concentration as well as half-life. (Source: Internet)

5. Lastly, the SAGE-217's failure re-assured MDD's huge market potential: $4B market cap was wiped out in a single day simply because SAGE-217 failed its MDD phase III trial. Its failure also opens up a large portion of market share that would otherwise be taken by SAGE-217.

According to the National Institute of Mental Health, 16 million U.S. adults suffered from at least one episode of depression each year and 10.3 million adults experienced at least one depressive episode with severe impairment. The highest age range that experienced depressive disorders was 18-25 years. The US anti-depressive market is about $5B currently. Previous market consensus was that Sage-217 and Fabre Karamer's Travivo would be two major market players taking a significant amount of market share. Now, with Sage-217 experiencing a major failure and Axsome Therapeutics currently sits only at a 1.57B market cap, we think it would bring additional $1.25B evaluation (assume AXSM's current market cap already takes about $750 evaluation in AXS-05 pipeline) if GEMINI succeeds.

What now?

I would like to re-iterate that the failure of Sage-217 does not increase the probability of AXS-05's GEMINI success. In fact, the chance of a positive GEMINI trial should not be impacted by Sage's failure at all. All the above reasoning just demonstrated that GEMINI may not experience some of the drawbacks MOUNTAIN trial had. Of course, GEMINI may have its own flaws, such as a very bold week 6 primary outcome. Further, no matter how clear ASCEND's result was, it only enrolled a small number of 80 patients. When the trial size was increased by 4 times, it really put AXS-05's therapeutics efficacy into test. All in all, previous failures in various MDD trials draw to a same conclusion that MDD is indeed a difficult-to-treat disease. Given all these evidences, AXS-05, in my judgement, represents a stronger MDD candidate over Sage-217, nothing more, nothing less. For me, I just cannot ignore the market opportunity released by Sage-217's failure, so I took an initial position in AXSM, fully aware of the risks associated with it.

Also, there are 3 late phase trials revealing top-line data towards the next few weeks. As a result, Axsome's stock price will not be purely determined by GEMINI trial outcome. Note that this article is solely for the purpose of discussing AXS-05 in MDD, so other Axsome's leading candidates, such as AXS-07, AXS-12 will not be covered in this article.

Keep in mind that, by the end of Q3, Axsome had $43M cash in hand. With a recent $80M offering and a quarterly burn rate of around $20M, it probably has $100M cash balance right now which could go for another year. With so many Phase 3 top lines due, it may offer further dilutions for continuous development/commercialization purposes, probably around $150M if any of its late-stage trial outcome being highly positive.

Author: Weiwei Wang

Disclosure: I am/we are long AXSM. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

<<<