They collected 40 samples and therefore had a bigger data set, except they had to filter the 40-patient data set "post-hoc" to get the ones with the MAP scores supporting their hypothesis - and for that they could only use 25.
Just like your patent point (which was valid), this might seem like a petty point, but saying they only took 25 patient serum levels when they took 40, especially since you said "they wouldn't arbitrarily leave some out - which one's would they chose to leave out", is in my opinion incorrect as they made a conscious decision to leave an additional 15 data points out aince they didn't fit into their new/reversed hypothesis. ---------------
Fair point. I think you've persuaded me I was misusing the number 25 as a filter. There are no petty points - I'm interested in what is true any error in reasoning you find helps me. And I hope vice versa. Sometimes I've got to think a bit to understand the significance of the error though.
There is still only 11 data points on the graphic though.
Here I quote Silviu from page 67 of the ODAC transcript talking about slide CC-32 (which is the one with the 11 data points on it). (BTW: Its otherwise the same as in PA'154 (the patent application) except for there is an R-squared = 0.38 and p=0.04 at ODAC and a Spearman r =0.76 p=0.0086 in Figure 7 of PA'154. )
"Shown in the next slide is the relationship between in vitro inhibition of IL-2 receptor expression for lot and reduction in activated CD4 T cells in vivo after 28 days of treatment with remestemcel-L in the phase 3 trial in patients who received a single product lot" (my underline)
Now given the purpose of this slide is to show as the heading on it says "Remestemcel-L Bioactivity in Vivo in Study 001:Inhibition of IL2Ralpha Expression in Vitro is Associated With In Vivo Reduction of Activated CD4+ T Cell.
I don't think Silviu is holding back any data points he has - they are all going to be on there (we can hold in suspension Bazsa's suggestion that there may be more than one patient per data point - so each of the 11 might be an average - (with 2 patients behind one data point perhaps 3 behind another perhaps all adding up to say 40 at the max (but its less than that though because you can't include subjects who got multiple lots - this is still only single lots (that's written on the slide) I can deal with that later multiples per dot suggestion of Bazsa's later I think).
Do you agree that its reasonable to assume that at the time of ODAC that was all the data MSB had represented on the slide - that in this case they would and should use all the data they had and not hold any back that was relevant to the subject of the slide (ie Study 001, in vivo and in vitro association) - only exclude patients that got more than one lot or for whom the blood data (28 days apart - activated T cells) was unavailable or something - is that reasonable in your view?
(20min delay)