CTAD 2024

Slide 6: Cellular stress described earlier likely lends to protein aggregates and clutter, which repeats in circular fashion. This eventually leads to lysosomal dysfunction. Lysosomes are needed during autophagic processes, clearing away waste in the brain. With dampened lysosomes, proteins begin clumping - causing compensatory functions to work harder and harder to combat the resultant inflammation. These normally cell-saving processes eventually become neurotoxic and cause degeneration. Anavex proposes the drug works by activating SIGMAR1 which latches unto/fuses with lysosomes, which negate lysosomal dysfunction - thus allowing autophagic function and proteostasis to continue in the brain. There are a myriad of other extremely beneficial pathways also enhanced by SIGMAR1 described briefly in the slide 7 section.

An independent paper exploring SIGMAR1 relationship to autophagy was published in Oct. This paper showed that SIGMAR1 enhances ATG8-family genes responsible for autophagy. Cells lacking SIGMAR1 showed lower ATG8 expression and worse autophagic flux. 

Slide 7: Anavex suggests that while Blarcamesine could be used as an independent monotherapy, the drug would likely work well in concert with monoclonal antibodies like Donanemab (Kisunla) or Lecanemab (Leqembi). This is because the monoclonals and Blarcamesine use different pathways to target similar effects, as if tackling the disease in a pincer movement. The monoclonals fight amyloid and tau via downstream mechanisms, while Blarcamesine enhances autophagy, calcium regulation, chromatin remodeling, cholesterol, and metabolic functions to provide neuroprotection more at the root cause of the disease (upstream). In concert, the drugs remove and prevent neurotoxicity while repairing and regenerating brain cells - which even without pathological conditions slowly continue during aging.

Slide 8: In conclusion, the mechanism of Blarcamesine is via upstream activation of autophagic function, which counters neurodegeneration and markers of aging.

Slides 10 & 11: Anavex explains that SIGMAR1 gene mutations exist within Alzheimer's populations, which impair protein trafficking. These gene mutations are found in approximately 30% of the population, whereas approximately 70% of people have fully functional SIGMAR1 genes. All patients reap benefit from SIGMAR1 activation, but fully functional patients (wild type / WT) have greater benefit. 

According to an analysis we conducted in 2023, approximately 84% of the global population have SIGMAR1 WT. An important distinction made in this slide deck is that Anavex considers patients to have SIGMAR1 WT if patients are homozygous dominant (both alleles are WT). Our analysis included all individuals with even one WT allele. Patients with one WT allele and one mutation allele likely respond worse than a patient with two WT alleles because in the heterozygous case the WT allele would be carrying function for the mutated alleles dampened function. As mentioned earlier, all patients benefit; however, we will adjust our SIGMAR1 WT expectation to be in-line with Anavex from this moment forward to be consistent (70% vs 30%). We highlight some interesting theoretical population numbers below built on this.

• ~70% of population have WT as defined by homozygous WT alleles

• ~30% of population have mutation as defined by homozygous mutation alleles or heterozygous combination

• ~84% of people globally have at least one WT allele

  • • Of the 30% mutation population ~47% have a WT and mutation allele, and 53% have two mutation alleles.

    • To put these percentages into action, the 2024 Alzheimer’s Disease Facts and Figures Report published by the Alzheimer’s Association increased the number of Americans with Alzheimer’s to 6,900,000.

      • • WT patients: 4,830,000 (homozygous)

        • Mutation carrying patients: 2,070,000

          • • Of the 2,070,000:

              • • Homozygous mutation: 1,097,100

                • Heterozygous WT + mutation: 972,900

These numbers do NOT include MCI due to Alzheimer’s. We use 5,000,000 for this figure (Alzheimer’s Association suggests 5,000,000 to 7,000,000) but sources vary widely.

Slides 13, 14, 15: Anavex showcases the three-arm trial design. They highlight that a sole cognitive measure (i.e. ADAS-COG) can serve as the primary endpoint for a trial according to existing EMA guidelines and draft FDA guidelines. There were no significant baseline outliers in patient populations per arm at the start of the trial.

Anavex for the first time reveals the intent-to-treat (ITT) scores for the ADAS-COG13 (cognitive) and CDR-SB (global) endpoints. There are two types of data analyses used in clinical trials, ITT and per-protocol (PP). ITT includes all patients as initially assigned regardless of adherence to protocol (like missing doses or early withdrawal). ITT reflects real-world effectiveness as it highlights what a global population on Blarcamesine with normal day-to-day hinderances might look like. In this case, Anavex defined ITT as any patients that had at least one dose and one office visit. Per-protocol analyzes only participants who strictly followed study protocol without significant division. This is usually looked at as ideal condition as the company excludes data from patients that deviate from instruction. Both types are desired by regulators, but ITT is shown significantly more weight.

Slide 16: Anavex reveals clinically meaningful scores with p-values under 0.025 in both the ITT ADAS-COG13 measure, and SIGMAR1 WT only ADAS-COG13 measures. As expected, SIGMAR1 WT patients improved more than patients with mutation genes, but all patients (pooled) improved to levels deemed to be clinically meaningful by researchers. This is extremely valuable data and is the best cognitive data I have seen in any pivotal trial with 36.3% slowing in ITT and 49.8% slowing in SIGMAR1 WT patients. The significance of this data is amplified by the fact that Anavex only screened out <10% of patients during enrollment, thus representing real-world global Alzheimer's populations. Monoclonal trials needed significant screening to produce ideal populations for their trials - representing only a sliver of real-world applicability.

Slide 17: Similarly, CDR-SB scores in the ITT and SIGMAR1 WT only populations improved significantly with p-values under 0.025. Lecanemab and Donanemab had improved CDR-SB scores in their trials by 0.45 and 0.6 points respectively. This showcases that Blarcamesine improved CDR-SB global scores in-line with standard of care.

In both ADAS-COG13 and CDR-SB, Anavex accomplished these scores in only 48 weeks as opposed to 18 months in Lecanemab and Donanemab. The therapeutic envelope is expected to increase as time goes on, so Anavex could reasonably expect their scores to be even larger in a comparable time period. We showcase the internally designed graphic below to illustrate Anavex's ADAS-COG13 data and comparability to monoclonal antibodies.

Slides 18, 19 & 20: Anavex summarizes that the drug is safe with minimal side effects, most of which are transient and mild. These side effects are able to be diminished by adjusting titration and dosing at night. There is no ARIA associated with orally available Blarcamesine, as opposed to monoclonal antibodies. Separately, the pre-specified analysis proves Anavex's mechanism of action via SIGMAR1 activation which is highly beneficial in a very large patient population. 

In addition to ADAS-COG13, CDR-SB, and CGI-I improvements (CGI-I not in this deck), brain atrophy in the whole brain, grey matter, and lateral ventricles reduced significantly (neurodegeneration, N). Amyloid 42/40 ratio also improved significantly in the CSF (amyloid, A), and tau trended favorably (tau, T). Neurofilament light (NfL) and ADCS-ADL (function) also trended favorably.

Importantly, slide 20 reveals that the long-awaited 2b/3 Alzheimer's paper has been submitted to the journal.

Conclusion: Considering what is known about Anavex's statistical analysis plan (SAP), the importance of the ITT populations in the tethered ADAS-COG13 and CDR-SB measures reaching statistical significance at values under 0.025 (each) and garnering scores higher than (ADAS-COG13) or equal to (CDR-SB) standard of care cannot be understated. This bolsters confidence that the EMA in Europe will approve Anavex - a much needed step for European clinicians and Alzheimer's patients. Additionally, if FDA guidance is published as was written in draft, Anavex will be able to file in the United States as well. We anticipate the FDA's stated draft desire to allow for one cognitive endpoint to carry an early-Alzheimer's filing (assuming strong response) will very likely be maintained in finished guidelines. Anavex's 2b/3 trial results address all requirements currently listed in aforementioned draft guidance.

Pre-specified SIGMAR1 WT analysis reveals large-population therapeutic benefit well-exceeding clinically meaningful status, and simultaneously serves to affirm the compounds mechanism of action. While many pre-specified analyses for other compounds usually limit market share, Blarcamesine is expected to benefit all patients (as demonstrated by pooled data) regardless of SIGMAR1 type. Beyond affirming mechanism of action, SIGMAR1 WT identification will serve clinicians when prescribing the drug, as it allows clinicians the ability to predict a significant treatment response (possibly as high or higher than 49.8% cognitive slowing along with atrophy reduction, amyloid improvement, and more).

The drug is safe, has no associated ARIA, is orally available, addresses all aspects of the A/T/N (amyloid, tau, neurodegeneration), and sees large clinically meaningful improvement in real-world, global patient populations.

Anavex still expects to file for wide-spread central authorization at the EMA by end of year, which would mean the drug would go up at the 11-14 Nov or 9-12 Dec CHMP meetings. Other near term catalysts include the 2b/3 Alzheimer's peer review publication and open label extension data (safety, ADAS-COG13, and ADCS-ADL). 

In our informed opinion, Blarcamesine approval in the European Union is highly likely forthcoming. It will be interesting to see Anavex build out its sales force, distribution, education materials, etc., partner with big pharma to commercialize, or be bought out under favorable circumstances. 

Bottom-line Up Front (BLUF): On 30 Oct 2024, Cognition Therapeutics further expanded on their CTAD presentation (Update 1) in an informative webcast. It was pleasing to hear that the company and I came to similar conclusions regarding the p-tau217 subgroup analysis. Due to thoroughness of our previous analysis, there wasn't much new information. Highlights of the webcast can be found below - many of which reiterate our earlier findings.

1. Lower levels of p-tau217 - which is an accurate measure of amyloid and tau burden - typically correspond to earlier disease stages (MCI or mild Alzheimer's). Higher amyloid and tau (pathology) more often than not indicate more advanced clinical stage. However, due to the extreme heterogenous nature of the disease there are plenty of patients throughout the Alzheimer's continuum (moderate and even some advanced stage) that manifest significant cognitive loss with lower levels of p-tau217. These facets indicate that the drug is theoretically clinical-stage agnostic as long as the actual pathological load isn't too high.

2. To conduct this analysis, the company used the median p-tau217 measure (1.0pg/ml) and created two simple groups. One of the groups was patients with 1.0pg/ml or more, and the other was under 1.0pg/ml. In the lower group, the mean was 0.66pg/ml and represented 45% of the dosed patient population. According to Cognition Therapeutics' CMO, the team that measured these samples have assessed thousands of patients, and the SHINE p-tau217 median (1.0pg/ml) was within the standard measure for mild-to-moderate patients. This means that Cognition Therapeutics could theoretically replicate these results in about 50% of all early-to-moderate patients in the global Alzheimer's community. Note: 0.66pg/ml is 3-5x greater than cognitively unimpaired individuals. 

3. Relatedly, due to the simplistic nature of p-tau217 blood test diagnostics it would be easy for the company to screen patients and reasonably expect therapeutic benefit based on their p-tau217 levels. We postulate that this could lead to prophylactic/preventative pathways.

4. Patients in the enrolling START trial feature Lecanemab (Leqembi) and Donanemab (Kisunla) patients. This will be one of the first large trials to feature these types of patients in their trial. It will be interesting to see how CT1812 interacts with patients on monoclonal antibodies.

5. By the end of the presentation we had two suggestions. One of the suggestions was to investigate SIGMAR2 (TMEM97) mutations - if any - and see if patients with mutation respond better or worse to treatment. The second suggestion was to devise a 'time saved' metric. Because the drug shows practical halting of disease progression at 6 months, it would be interesting to see how much time this saves patients in earlier stages, reducing need for caregiver, extending independence, and possibly increasing overall lifespan.

Conclusion: Overall the webcast was definitely worth a listen, especially for new investors. SHINE demonstrated itself as an excellent exploratory pivot point and the path towards a much better optimized phase 3 trial has been revealed. By screening out ~50% of patients with non-optimal p-tau217, reducing the high-dose arm from 300mg to 200mg, and conducting a longer trial, the company will likely be able to demonstrate extremely significant clinical outcomes addressing a large patient population in multiple clinical stages (mild-to-moderate). We look forward to continue following CT1812's story and will be ready to digest outcomes of forthcoming phase 2 LBD topline data.

Bottom-line Up Front (BLUF): On 29 Oct 2024, Cognition Therapeutics (CGTX) released prespecified subgroup analysis regarding the p-tau217 plasma biomarker. This biomarker is elevated in Alzheimer’s patients, accurately correlates to amyloid/tau burden, and is associated with greater cognitive decline. While not exclusively true, it can be expected that earlier stage patients have lower p-tau217 over more advanced clinical stage patients. The company found that patients with lower levels of p-tau217 at trial commencement fared extremely well compared to patients with higher p-tau217 in multiple clinical assessments; including ADAS-COG11 (cognition), MMSE (a measure of cognitive impairment), ADCS-ADL (function), and ADCS-CGIC (global). This subgroup was not insignificant, and represented 45% of the total dosed population. These measures all trended positively and beat out the total pooled population (pre-subgroup analysis) but still did not reach statistical significance. In our opinion, the statistical significance miss is practically irrelevant for this phase 2 study and the results presented are tremendously compelling. We have three primary takeaways:

1. These results emphasize the probability that an earlier-stage Alzheimer’s trial (like START) would likely have more robust outcomes. The primary mechanism of action for CT1812 is to reduce and prevent oligomer binding to synapses. Patients in earlier stages (typically corresponding to lower p-tau217) would likely benefit more from oligomer displacement because they are at an earlier stage of tau pathology. Preventing or displacing oligomers could delay or halt downstream cascades that would normally lead to phosphorylation and tangle formation, thus slowing disease progression. In contrast, patients with higher p-tau217 at treatment commencement would have more advanced tau pathology, making it more difficult for the drug to fight through inflammation and existing tau damage. Evidence for this theory can be seen in graphic 1 below. The first chart shows the primary analysis (placebo vs. total pooled population), the second chart shows low p-tau217 vs. comparable placebo, and the third chart shows high p-tau217 vs. comparable placebo. Attention should be paid to the second and third charts in particular, as we can visualize that higher p-tau217 patients performed worse than placebo. Ultimately, these charts tell us that had the higher p-tau217 patients been removed or excluded from the trial entirely, the primary analysis would have likely reached statistical significance. We noted that the low p-tau217 slope and overall result looks similar to Donanemab trial data at a similar timeframe. You can view those results here.

2. As pointed out by Cognition Therapeutics, p-tau217 can now be used as a predictive biomarker of response to CT1812 treatment. The company could reasonably expect that Alzheimer’s patients with low p-tau217 would respond exceedingly well to therapy. In this moderately-sized subpopulation, patients could possibly completely or near-completely halt cognitive and functional decline for 6 or more months. After 6 months, decline would marginally continue as indicated by published slopes. At this point, we theorize decline rate would be significantly slower than placebo patients, and CT1812-dosed patients could expect to see a tremendous gap in decline compared to placebo patients. All in all, this indicates that patients with low p-tau217 measures would fare significantly better than untreated patients and likely significantly increase a patient's time in early stages where daily life free of caregiver can be achieved. Today’s presentation significantly advances future packages with regulators including the FDA and EMA.

3. A better optimized phase 3 Alzheimer's trial can be planned and enrolled utilizing targeted p-tau217 thresholds. Cognition Therapeutics plans to pursue this method of screening in this upcoming phase 3 trial.

Conclusion: CT1812 featuring unique SIGMAR2 modulation dramatically improved 45% of all dosed patients. Cognitive scores for these patients as assessed by ADAS-COG11 showed a relative halt to disease progression at 6 months - an improvement over comparable placebo patients of 2.7 points. This is a clinically meaningful improvement. If Cognition Therapeutics could screen for low p-tau217 patients in their upcoming phase 3 trial and was able to replicate these results over a longer timeframe (ideally 52 - 76 weeks), the company would have better cognitive data than Donanemab (Kisunla) with easier administration, better safety, and a greater patient population/market share. While we focus heavily on earlier-stage patients, it is notable that the drug maintains efficacy against moderate-stage patients as well - a feat that most cannot achieve. We look forward to listening in to the company webcast tomorrow (30 Oct), and we will include our commentary on that webcast in the evening. A link to the webcast can be found at the bottom of their 29 Oct press release.