Synthetic Odds Ratios in Phase 2a and Phase 2b/3 in Alzheimer’s Trial for Blarcamesine $AVXL

This Blog Is Only For Educational Purposes. Do Not Trade On This Blog Alone. Do Your Due Deligence or Consult Professional.

Lately, I have been posting videos but there ia limited capability of videos to disseminate graphics or plots. I decided to do a post and video later.

The Anavex’s data release on the results of Phase 2b/3 included Odds Ratio versus Placebo arm for cognition (MMSE) and Activities of Daily Living (ADCS-ADL). Again, we assume that the Low and Medium Concentration cohorts did not exhibited any therapeutic effect or efficacy so that we can assume them to be a synthetic placebo arm in Phase2a. Let’s see the marked up illustrations/presentation slides.

The criteria for cognitive improvement will be the zero line. If we want to arrive at Odds Ratio for cognition improvement we count the improved cases on Low and Medium concentrations columns, namely population of 2. The population of those improved with the High Concentration can be taken to be either 5 or 4, one case on borderline. this gives tentative Odds Ratio of 2.5 or 2 for lower value when excluding the border case. If we compare it to the press release on Phase 2b/3 Odds Ratio of 1.84 we get very close value.

Now, let’s turn our attention to the ADCS-ADL. I bet it will get more interesting since we have never anylyzed daily living results.

The values for ADL are within round parenthesis and the corresponding to patient MMSE (cognition scores) are within square brackets. First thing to notice here is that the border criteria for performance in ADCS-ADL is set at minus 2 points. Notice that all the 4 cases of synthetic placebo belonging to the Low and Medium Cohorts are almost exclusively negative on cognition. The Odds Ratio for ADCS-ADL seems to be 4 to 6 or 1.5., or 150% improvement. If we move the border criteria for improvement up to zero line the Odds Ratio becomes 2 to 4 or 2, 200%. Compare this with 2.67 (267%) Odds ratio in the Phase 2b/3. On very superficial level Blarcamesine in Phase 2b/3 has repeated the performance of Phase 2a. Notice that there is high correlation between general improvement in cognition and improvement in daily living scores.

Another aspect of the data from Phase 2b/3 was that the cutoff criteria was +3.5 ADCS-ADL points, not zero or -2 points. Just a remainder that positive is the direction of the increased ability to carry on the daily living. This suggests much stronger response in Phase 2b/3.

If indeed the Low and Medium Concentration cohort or rather only the Medium Concentration Cohort corresponds to the 30mg dose then one should expect very low performance in this cohort versus the higher dose 50mg cohort. This being the case the 50mg cohort would have higher concentration of better MMSE and ADCS-ADL scores. I touched on this issue in the video below.

The reason for combining the cohorts is a mystery. However, the theory of Lane Simonian published in Seeking Alpha article that the combining of cohorts was done to support achieving statistical significance does not seen to be a valid as the calculations to arrive at it do not involve directly the n of the dosed cohort. Since the Standard Deviation of the Placebo arm is set and its Standard Error as well it leaves only the mean of the dosed cohort to be given a consideration. If the mean of dosed cohort is outside (to the right of) the value of 1.96 Standard Error plus Mean of Placebo then we have statistical significance. Including the 30mg cohort with the expectation of lower efficacy should rather move to the left the value of the mean for combined cohort versus the 50mg cohort mean which could be more to the right. That is only the case , if our expectations are correct. In other words, at this point of my knowledge how p-values are computed I think that Lane’s idea is false.

link to the article : https://seekingalpha.com/article/4592603-biogen-and-eisai-cassava-sciences-and-anavex-clinical-trial-results-in-need-of-revisions?source=content_type%3Areact%7Csection%3Asummary%7Csection_asset%3Aall_analysis%7Cfirst_level_url%3Asymbol%7Cbutton%3ATitle%7Clock_status%3ANo%7Cline%3A1

In conclusion, the results from the Phase 2b/3 as they are posted to the public give a picture of performance compatible with the results of the Phase 2a. Yet, the combining of cohorts on the face of facts known at this point of Phase 2a dosing seems odd at the least.

See more videoes on my YouTube channel @TheLomLom9 or Energy and Alzheimer’s