AAIC 2021 Updates & Anavex’s 29 July 2021 PR
Annovis
First let’s talk about Annovis (ANVS) data. Presented on 28 July 2021, ANVS failed to impress with lackluster outcomes and even worst slide-ology (as we call it). As a reminder, this trial was for Alzheimer’s disease and Parkinson’s disease, and was administered via IV. Before we dive into it, let’s acknowledge that their baseline MMSE score for the AD cohort was 24.5 - 25.4. This means the vast majority of their patients are MCI or very early-stage Alzheimer’s disease. Keep in mind that Anavex’s AD 2a MMSE baseline was 21 at the beginning of the trial, which indicates Anavex’s cohort was significantly more degraded than Annovis’. This is an important distinction because most trials see better results in less-impaired patients. If ANVS is disease-modifying, their scores should have blown Anavex out of the water on this fact alone.
ANVS Summary: ANVS grasped at straws with 7 different cognitive tests out of all 7, only ADAS-COG 11 dosed vs. baseline produced statistically significant results (with a 4.4 point improvement from baseline/30%). They failed to produce statistically significant results in ADAS-COG 11 for dosed vs. placebo (dosed improved 3.3 points/22% vs. placebo, indicating placebo managed to improve 8% on their own in the same amount of time – 25 days). Their PD patients were tested using UPDRS which did show trends of improvement, none of which was meaningful. Considering ANVS’s high MMSE baseline, vast slew of cognitive tests, and less intensive ADAS-COG 11 measure, (as opposed to ADAS-COG 14), ANVS should have been able to provide dramatically better data for both Alzheimer’s disease and Parkinson’s disease. Their ADAS-COG 11 outcome must be looked at skeptically as they failed to produce significant results against placebo. Safety profile was great. More time and data is needed to reevaluate at a later time.
ANVS Breakdown: ANVS tested 4x separate ADAS-COG measures (ADAS-COG 3, 6, 11, and 14) with ADAS-COG 11 being their primary cognitive endpoint. Intriguingly, ANVS was only able to garner statistically significant results against baseline, but NOT for placebo for ADAS-COG 11. Furthermore, all other ADAS-COG tests failed to reach statistical significance. Slide 8 demonstrates how ADAS-COG 11 is relatively less intensive than ADAS-COG 14, which is likely why the company chose it to begin with, in order to try to inch out statistical significance.
UPDRS data for the PD cohort was abysmal:
UPDRS Part 1: Placebo improved 14% compared to 23% dosed
UPDRS Part 2: Placebo improved 26% compared to 27% dosed (nearly identical)
UPDRS Part 3: Placebo declined 1% compared to 2% improvement dosed
UPDRS Part 4: Placebo declined 11% compared to 7% decline dosed
UPDRS Total: Placebo improved 1% compared to 3% improvement dosed
To put the Parkinsons’s UPDRS data into perspective, ANVS trialed for 25 days. Meanwhile, over 14 weeks Anavex’s UPDRS total score was: placebo declined 3.53 points and dosed improved -10.98% for a total placebo/dosed differentiation of -14.51 points, an 18.9% improvement. A -7.1 point improvement is considered clinically meaningful. Anavex surpassed that by over double, ANVS didn’t even come close.
Furthermore, ANVS failed to produce statistically significant or clinically meaningful data in the WAIS coding test, MMSE, and CDR sum of boxes. There is some promising correlation however, which may be proven out in a longer trial.
The last item of note for ANVS was their ability to reap statistical significance in inflammatory markers in AD and PD patients. This is overall a positive but doesn’t necessarily mean anything if their cognitive outcomes can’t follow up with meaningful data.
Cassava
Cassava’s (SAVA) 9 month AAIC data is actually quite favorable to the company. The first 50 patients (data presented) had a mean MMSE baseline of 22.6 – somewhat less impaired than Anavex’s AD 2a, but relatively comparable.
SAVA Summary: Overall the cognitive results for SAVA are quite favorable. At 6 months they were able to produce a -1.6 point ADAS COG 11 improvement (10%) to baseline and a -3.0 point ADAS COG 11 improvement (19%) at 9 months. What’s more, at 9 months 66% of patients improved over baseline. To top off their good cognitive results, the dose is given orally, the trial appears to be safe, and they had some titillating biomarker data.
SAVA Breakdown: As mentioned earlier, SAVA was able to produce very nice ADAS-COG 11 improvements at week 24 (6 mo) and week 36 (9 mo). The improvement to that point cannot be understated; however, as a word of caution, it is important to note that many companies do well in early phases of their AD trials with a drop off typically occurring sometime at or before week 40 – 60. One of the reasons this knowledge is so important is when we account for Anavex’s long term AD 2a data. Let’s compare them below.
ALL POINT VALUES ARE ADAS-COG/CORRELATED VALUES
SAVA 24 week (6 mo): -1.6 point
SAVA 36 week (9 mo): -3.0 point
BASELINE MMSE was 22.6
AVXL 57 week (14.25 mo): -3.4 point
AVXL 70 week (17.5 mo): -5.1 point
BASLEINE MMSE was 21
Additionally, it should be noted that the relatively large N Anavex PDD trial yielded significant results in Episodic Memory outcomes, which have a 70% correlation to ADAS-COG. The 29 July 2021 PR makes sure to calculate those correlations for us.
AVXL PDD 14 week (3.5 mo): -1.9 point in the 50mg cohort (8% mean improvement over baseline).
Not to mention the massive 4.0 difference between dosed and placebo at week 14 in the PDD trial.
Back to SAVA, the company was able to reduction in a number of cerebral spinal fluid biomarkers.
AB42: low in AD patients (increased 84% at 6 mo)
Total tau: marker of neurodegeneration (decreased 38% at 6 mo)
P-tau181: marker of disrupted tau function (decreased 18% at 6 mo)
Neurogranin: synaptic loss/degeneration (decreased 72% at 6 mo)
Neurofilament Light Chain: axonal loss/degeneration (decreased 55% at 6 mo)
YKL-40: marker of neuroinflammation (decreased 44% at 6 mo)
sTREM2: microglial-induced neuroinflammation: (decreased 65% at 6 mo)
HMGB1: pathogenic “danger” molecule: (decreased 53% at 6 mo)
Finally, for their behaviors endpoint (NPI): “at baseline, 34% of patients had no neuropsychiatric symptoms on the NPI scale. At month 9, >50% of patients had no symptoms on NPI”. So the company was able to reduce NPI to 0 for about 16% of affected patients.
Summation: Annovis and Cassava are both showcasing phase 2 AD trials. Annovis has a lot of ground to make up for in their phase 3 trial. Current Annovis data lacks teeth and presents a number of holes towards its potential. Additionally, Annovis is administered via IV, and their starting MMSE was considerably higher than both AVXL and SAVA – which should have produced a much better outcome. Cassava data is quite compelling. Longer term data with a placebo control and blinding is needed to corroborate these results, but if I had to choose between an investment in Anavex, Cassava, and Annovis, I would invest capital in that order. While Cassava’s data looks as good as Anavex’s in many ways, we must remember that Anavex’s data has held over a long time horizon. Anavex has learned a lot from the AD 2a trial and has positioned themselves very well to have an even greater effect in the ongoing 2b/3 with less impaired patients, biomarker/genomic indicators, and dosing understanding. The 2b/3 cognitive data will likely be significantly greater than their 2a – and we didn’t even mention their ADCS-ADL daily living scores which were extraordinarily powerful. Furthermore, unlike its competition, Anavex has extremely significant PDD data, and is likely nearing approval and commercialization in Rett Syndrome.
Anavex 29 July 2021 Press Release
Anavex surprises shareholders by announcing plans for a phase 3 trial to prevent Alzheimer’s disease. The company has been making comments about this potential for months now with a very public proclamation days ago on a Yahoo Finance interview with the CEO. It is extremely interesting this trial is already being planned, as most shareholders would probably have assumed it would have happened post-AD therapy approval. It is fantastic news the company is sparing no time (and newly acquired cash) to expand the pipeline even further. I look forward to seeing more preclinical preventative data later this year (having already seen signs of prevention in clinical trials, and other preclinical trials). As a final note and mentioned earlier, the company has provided the ADAS-COG correlations for not only the Alzheimer’s 2a trial, but also the recent PDD cognitive data, and this is to my knowledge the first time the PDD calculations have been readily provided to investors.
