2023-07-12 09:45:03 ET
Summary
- Cassava Sciences' results for mild Alzheimer's disease are better than placebo at 12 months but little better beyond this. Thus, there is no compelling reason to invest in the stock.
- Simufilam may help maintain cognition in those with mild cognitive impairment over long periods of time, but this has not been definitively determined yet.
- To gain FDA approval, Simufilam likely has to perform at least as well as acetylcholinesterase inhibitors such as Aricept.
Cassava Sciences ( SAVA ) recently announced results from its Cognition Maintenance Study for Alzheimer’s disease. In this study, some of those who completed the twelve month open label extension trial on Simufilam continued on the drug while others were switched to placebo for six months. The company presented two sets of data: one for those who it said had mild to moderate Alzheimer’s disease and one for those who it said had mild Alzheimer’s disease. However, most of the first group had mild Alzheimer’s disease whereas most of the second group had mild cognitive impairment. Once this adjustment is made, Cassava Sciences’ results for mild Alzheimer’s disease are almost the same as placebo between 12 and 18 months. Given this, there is no compelling reason to invest in the stock.
Cassava Sciences ADAS-Cog scores are once again for earlier stages of cognitive impairment than its MMSE scores ( previous article ). Simufilam does appear to modestly slow the progression of mild Alzheimer's disease for about a year, but its beneficial impact stops after that.
Before looking at the specific numbers from the Cognition Maintenance Study, however, it is necessary to discuss defining cognitive status based on MMSE (Mini-Mental State Examination) scores, and the conversion of ADAS-Cog11 (Alzheimer’s Disease Assessment Scale-Cognitive) scores to MMSE scores. MMSE ranges for various forms of cognitive impairment vary by a couple of points either way depending upon the source. Moreover, some scientists and some companies, such as Cassava Sciences, extend the definition of mild Alzheimer’s disease to include some of those with mild cognitive impairment. However doing so ignores the fact that those with mild Alzheimer’s disease have considerably greater cognitive deficits than those with mild cognitive impairment ( comparisons ). The following are relatively standard MMSE ranges for those with cognitive issues:
MMSE Score
- 24-27 Mild Cognitive Impairment
- 20-24 Mild Alzheimer’s Disease
- 10-20 Moderate Alzheimer’s Disease
Likewise there is no agreed upon conversion rate between ADAS-Cog11 scores and MMSE scores. The conversion tables are based on different clinical trials in which both MMSE and ADAS-Cog scores were used so there will be some minor differences between them. However even the most liberal of these instruments comes nowhere close to matching the MMSE scores and ADAS-Cog scores provided by Cassava Sciences. The co
Here are the specific mean baseline numbers which Cassava Sciences gave for each group followed by more accurate MMSE scores "derived" from the actual ADAS-Cog scores (lower MMSE scores and higher ADAS-Cog scores reflect greater cognitive impairment):
Mild Alzheimer’s Disease Simufilam
- MMSE 18.6
- ADAS-Cog11 19.3
- Adjusted MMSE 23.5
Mild Alzheimer's Disease Placebo
- MMSE 18.1
- ADAS-Cog 11 21.9
- Adjusted MMSE 22.5
Mild Cognitive Impairment Simufilam
- MMSE 24.1
- ADAS-Cog11 11
- Adjusted MMSE 27
Mild Cognitive Impairment Placebo
- MMSE 24
- ADAS-Cog11 11.2
- Adjusted MMSE 27
Cassava Sciences never provided the number of patients who truly had moderate Alzheimer’s disease from its open label extension trial nor from its Cognition Maintenance Study. Given the ADAS-Cog mean baseline number in the Simufilam group in the CMS, the number of patients who had moderate Alzheimer’s disease at this point was likely quite small. What we do know is that in the open label extension trial these moderate patients declined modestly less than the expected placebo decline for this group ( 4.4 points versus approximately 6 points ).
Simufilam seemed to perform only slightly better than a true placebo in mild Alzheimer's disease between 12 and 18 months, as shown by the following chart (Cassava Sciences did not provide any numbers for the "mild to moderate" group but their decline is easy to calculate based on the data they did provide for the "mild" group and for the full set of patients).
Simufilam (mean baseline of 19.3 baseline):
- 2.4 point decline between 12 and 18 months
Placebo (mean baseline of 21.9):
- 2.4 point decline between 12 and 18 months
True Placebo Decline (mean baseline of 20.7)
- Approximate 3.3 point decline between 12 and 18 months ( table three )
Simufilam does provide an advantage over placebo for mild Alzheimer's disease up to a year ( .5 point decline versus a 3.5 point decline ). The almost inescapable conclusion is that Simufilam provides a temporary benefit for those with mild Alzheimer’s disease but that it is not a disease modifying treatment.
The least cognitively impaired group provides the greatest challenge in interpretation. Part of this is due to the fact that ADAS-Cog may not be the most sensitive measure for those with mild cognitive impairment ( debate ). Nevertheless the improvement in this group appears to be impressive. In the open label extension trial, those with mild cognitive impairment improved from 15 to 12.6 points on the ADAS-COG11 scale. The original starting mean for those who continued on to the Cognition Maintenance Study is not known, but the mean baseline at the beginning of the Cognition Maintenance Study was 11 for those in the drug group and 11.2 in the placebo group. Those on Simufilam improved by another .6 points between 12 and 18 months whereas those on placebo declined by .6 points. Other types of intervention can lead to temporary improvements in cognition in mild cognitive impairment and not all of those with mild cognitive impairment even those with substantial amounts of amyloid go on to develop Alzheimer’s disease. In addition because of the small number of patients (40), the length of the trial, and the relatively small difference between improvement in Simufilam patients and the relatively small decline for those on placebo between 12 months and 18 months, it is impossible to say how long or if Simufilam can delay the progression to Alzheimer’s disease.
Cassava Sciences has been very creative in its presentation of data. The initial downward movement of the stock following the release of this data appears to be the correct response. Long-term use in those with mild cognitive impairment (although better evidence is needed for its efficacy in this group), short-term use in those with mild Alzheimer’s disease, and no use in those with moderate to severe Alzheimer’s disease may be the appropriate future course of action.
Cassava Sciences’ likely has another year before it presents its phase 3 clinical trial results (from two trials). The outcome of the Department of Justice’s investigation and a court case against Cassava Sciences for alleged manipulation of data could still have a significant impact on the stock’s value between now and then. The CMS data, though, may have been the last chance to significantly move Cassava Sciences stock in an upward direction and it did not do so. I don’t see anyone buying out the company nor any acceleration of the process toward FDA approval (or non-approval) based on the CMS results. My prediction is that Cassava Sciences will remain around 20 points until the release of phase 3 clinical trial results. Those who want to hold on to the stock through the phase 3 clinical trials have to hope that data is at least as good as it is for acetylcholinesterases like Aricept. I don’t expect it to be better; indeed is appears to be about the same ( table 3 ). Therefore, without much conviction, I continue to give the stock a hold recommendation.
For further details see:
Cassava Sciences' Simufilam For Alzheimer's Disease: Not As Good As Advertised