Cassava Sciences, Hit or Miss? $SAVA $AVXL

Do not trade this blog as 4 billion years of evolution is against me and possibly even God laughs at it, He has a great sense of humor!

Cassava’s target is the good old Amyloid plaque, the only difference between the other amyloid palque targeting drugs and Sumifilam is that instead removal of the plaque the sumifilam blocks production of amyloid plaque in its misshaped state and therefore the whole cascade of other biomarkers are controled, especially those direcetly involved in inflammation and released in neuronal death. The Unified Alzheimer’s Theory claimes that there is a hypothetical event causing homeostsis disturbance, in response to which the plague is produced and finally the inflammation does is irreversable damage. This is the short vesion of it. As knowledge of biochemistry of amyloid plaque grows so do opportunities to disrupt the process. This slide expalained how the discovery was made.

Slide 5 from CTAD 2020 Presentation

First the target protein gets discovered, later the physiological process in which it was involved i.e. processes in which it binds with other proteins or catalyses some processes. All within some cascade of reactions leading to produce some final physiological effect. This opens an opportunity to either bind to this protein to lower its impact or block it altogether, that is to remove it, or supplement it with the drug to increase its impact. In point 3 of the slide 5, a small molecule by trial and error has been found to bind to altered version of the protein taking part in the cascade of events leading to deposits of plaque. Restoration to healthy state of A-Beta-42 lowers the neuroinflammation by removal of misshapen ABeta-42 otherwise acting on TLR4 toll-like receptor. Homeostasis is restored at least in this cascade.

The tau tangles are accumulating in the interstitial space between the cells. The best indicator of deposits of ABeta plaque (Amyloid) inside the cells is the P-Tau181. See this reference. https://pubmed.ncbi.nlm.nih.gov/29626426/
This very protein is considered a potential target for biomarker to detect levels of Abeta plaque and Tau tangles to screen patients for signs of early ABeta deposits indicating onset of Alzheimer’s, that at least is the intension of the authors of the above mentioned paper. Yet, as the reference acknowledges the screening has little efficacy in finding the patients with MCI (Mild Cognitive Imparaiment) but its power lies in predicting whether patients develops deposits of ABeta and Tau better than APOE4 allele and age. This is inline with the observation that 30% of diseased Alzheimer’s patients carry no sign of Abeta plaque deposits at autopsy. The significance of this is that $AVXL and The Unified Alzheimer’s Theory seem onto a better path to tackle the CNS diseases, as the cause to be addressed seems to be up stream from Abeta (Amyloid) and Tau physiology. This is due to the old story that murine model first develope cognitive decline 2 months before onset of the plaque. The lifespan of a mouse is about 11 months. This corresponds to 70 to 80 human years. 6 months corresponds to 42-48 years in humans but 4 months to 28-32 years. Of course these are not absolute numbers but they are presented here to give you insight into the timeline of events. See post https://piotrpeterblog.com/2018/08/10/the-winning-combination-for-anavex-2-73-and-is-it-all-about-inflammation/

The 10% drop in P-Tau181 indicates that the Sumifilam works as designed and lowers the biomarker for Abeta and Tau. Slide 19 confirms that Abeta-42 (the healthy form) has increased. I would like to turn your attention to slides 22 & 23, because they present the evidence that the neural death has been lowered substantially by the drug.

Antibodies that neutralize HMGB1 confer protection against damage and tissue injury during arthritiscolitisischemiasepsis, endotoxemia, and systemic lupus erythematosus.

This quotation comes from Wikipedia page on HMGB1. The protein is in positive feedback loop of inflammation. As neurons are dying, macrophages that digests them relase this companound which works like a signal to other receptors in the inflammation pathways. Reducing its levels in the cerebral fluid lowers the destructive effects of neuroinflammation.

The Brain Blood Barrier gets “rebuild” too as the levels of the most ubiquitous protein in blood which indicate the health of BBB gets lowered.

Sumifilam shows all the presented biomarkers going its way, on biomarkers this drug should have cured the Alzheimer’s already. All the ducks are in row for a big show of secondary measures of cognition. Before we get into this, let us remember that on basis of 57 weeks the greatest progress done by Blarcamesine was during first 5 weeks of dosing. After that the progress among 6 Supper Responders was a steady paced improvement. Here, the dose was provided for 28 days that is 4 weeks. Dr. McFarlane once said that Blarcamesine had 110% (1.1) Effect Size when run against placebo. I calculated that the effect size can go from just ~35% to ~300%, it all depends on the performance of the placebo arm and the subgroup of the patients taken for analysis. Patients in AD Phase2b/3 can be anywhere in between those two limiting cases. Of course, the above analysis pertains to patients in the first 57 weeks of AD Phase 2a. Now let us see the performance of Sumifilam on cognition.

The Effect Size for Sumifilam is large enough to assure an approval as it is above ES=.28 of Donezepli though it is a bit spotty with the 50mg cohort having 37% Effect Size vs. 23% 100mg cohort. I guess this is the best data Sumifilam could muster which overlaps with the worst case for the ES for the entire High Concentration Cohort (including the worst performers) in ANAVEX2-73 Phase 2a vs syntethic placebo. If Dr McFarlane’s remark that Blacamesine has exhibited 110% Effect Size in Australian trials is true then $AVXL still comes ahead. The PDD Phase 2 had 33% of patients to SIGMAR1 mutated variant, the reminder was SIGMAR1 Wlid Type. That was even larger incidence of the mutated version than in the general poplulation which is about 16-20%. I wonder whether the same composition will be retained in AD Phase 2b/3 or will the results be adjusted for the genetic composition of the cohort. https://piotrpeterblog.com/2020/11/10/preview-of-possible-effect-size-in-alzheimers-phase-2b-3-avxl-blarcamesine-anavex2-73/

  • My impression is that Sumifilam did not that much improved the lot of the patients through removal of the plaque what it did just by stopping the inflammation associated with the process. It is rather the proof of the destructive force of neurinflammation than the theory of plaque removal as a way to help Alzheimer’s patients.
  • The fact that 98% of trial population have improved indicates that there is no genetic aspect to Sumifilam efficacy unlike in Blarcamesine’s case. They work in two different ways and could be combined, hopefully. Blarcamesine can tackle the 80-84% of patients carrying the SIGMAR1 Wild Type gene.
  • There is almost instantaneous effect (4 weeks) just as it was in the case of Blarcamesine (5 weeks). Bodes well for both drugs.
  • Blarcamesine’s trials by Dr Missling design never looked into biomarkers (at least the same way as Sumifilam) since Dr Missling looked to control cost and concentrate on final outcomes that is cognitive measures. The consequence of this is that we do not know the impact of Blarcamesine on inflammation besides the global cognition impromement. One can be infered from the other, yet we lack numbers to make a direct comparison.
  • The reach of Blarcamesine is wider as there are multiple indication it can be used. Sumifilam actually is limited to the patients with Amyloid plaque, the telltale sign of Alzheimer’s disease, as it at least is now accepted to be.
  • Blarcamesine Rett Syndrome trials point to earier approval than Sumifilam’s but do not discount a partnership to speed up Sumifilam approval or $SAVA’s buyout.
  • $SAVA is just a variation of the theme of Amyloid removal and as such might get quicker acceptance than $AVXL with much more novel approach and still unknow MOA.
  • On the other hand unknown MOA makes $AVXL have a solid moat. Rather it is not the MOA per se but the unknown way ANAVEX2-73 interacts with SIGMAR1 receptor.
  • The cost of measuring 11 biomarkers must have been excessive and it follows the pattern of scientists to be in the driving seat. Seeking validation through biomarkers seems to be the approach taken by many Amyloid Theory trials as the biomarkers “proved” the theory valid while the cognitive tests were disappointing. Ultimately, what you “sell” is the cognitive improvement.
  • $SAVA is looking toward ~$75 million offering, compare it with the financing conducted by Dr Missling.

Leave a Reply