Replies to post #252971 on Anavex Life Sciences Corp (AVXL)

[Investor2014]

It may be possible that the baseline cognitive score being on the better side means more good days for placebo patients than otherwise because they think or feel they are getting A2-73 when they are not.

For the actual genetic and fluid based biomarkers I doubt they would predisposition patients differently in terms of possible response to placebo, but they would influence outcomes once that patient actually receives A2-73 during OLE just as is expected if on A2-73 during the trial.

[bas2020]

Simply put, the placebo effect is psychological, whereas the gene variant effect is biological, whether it be beneficial or detrimental. Thus, the placebo effect is temporary, lasting from a few days to several weeks; whereas the gene variant effect is sustained. Placebo effects are independent of any gene variant effects, meaning they may or may not coexist with a given patient.

Many suspect that patients with moderately detrimental gene variants may experience improvements in therapeutic response at higher dosages, up to the individual's MTD.

[frrol]

The company will analyze the data from our trials by running OLS regressions on the data. The dependent variable will be simply the efficacy measures. The independent variables will be everything: dosage (0 for placebo), biomarkers (which will be binary: yes or no), gender, age, baseline measures, donepezil dosage, etc, etc. The correlation coefficients that are produced will show what factors were 'relevant', and to what statistical confidence. The point being that such analysis will reveal the validity (relevance) of the biomarkers, independently of anything else, including whether the participant was on placebo or not.

(They will also be analyzing covariance between independent variables. Eg, to check the relationship between changes in GABA and glutamate, and changes between efficacy measures like MMSE and ADSL.)

So whatever you're concerned about, with a large n trial that's been properly placebo controlled and blinded, whatever effects our drug, the placebo, the biomarkers, the age, the baseline MMSE, etc has on any efficacy outcomes, it will be statistically rendered for analysis, and will hopefully be clear and significant. If it isn't clear and significant, it won't be because of using placebos and biomarkers together. If that is what you were concerned with, no need to be.

[Steady_T]

Because an individual has an identified biomarker doesn't really say anything about what placebo response that individual will have. Identifying subjects with a particular biomarker in both treatment and placebo arms allows for a comparison of the two groups to determine if there is a difference. If there is a difference only then is the biomarker validated.

You have identified the reason for a placebo control group. It is unknown what placebo response a group of people will have to the treatment protocol in advance. To find that out a placebo arm of a study is set up.

A placebo control group does more than control for the psychological aspects of a trial. It also controls for many other unknown variables in the trial. For example, may be the disease of interest in the trial responds to the dye used in a scan. Or maybe there is some sort of selection bias in the population selection. Maybe just being in a trial reduces stress enough to cause a response.

I remember a particular rat behavior study from school where that experiment was duplicated in England and the researchers got very different results to overcrowding. After a great deal of testing and research it was discovered that the newspaper used for bedding in the US differed from the newspaper used in England. The English newspaper material was made with a different process and had an estrogen mimicking hormone residue in it that the US paper did not. That Estrogen mimicking hormone caused the difference in rat behavior.

It is that sort of unknown, off the wall, sort of thing that placebo group help account for along with the more obvious things.