RE:RE:RE:Excel spreadsheet display of new data from MDA
Hi DJDawg, I tend to look at data presented in MD&A's alone - not suggesting those are more reliable, just to keep things manageable for myself. From these, I take whatever raw numbers they include (i.e. # of treated patients, evaluable patients, CRs, IRs, etc.) and I keep track of how they develop from MD&A to MD&A. I typically exclude the Phase Ib results (assuming they won't be presented to FDA again) and I believe I have a decent understanding of how the results for the undertreated cohort differ from the results for the optimized cohort (the MD&A of 28 April 2022 provides the key).
Yesterday's data format deviates from the previous ones, but still provides enough raw numbers (first table and the swimmers' plot) to continue to track the development of CRs and IRs, with results as given in my previous post .
New, for me, is the insight that apparently not all patients receive all assessments. Many seem not to be further evaluated after some point (which may be after 90, 180, 270 or 360 days), for whatever reason (e.g. trial protocol, patient's desiring to discontinue). It's acknowledgement of this (by the team) that imo accounts for the reduction in evaluable patients numbers. It cannot be the result of the undertreated cohort having simply been omitted. First of all, there are now ~31 patients long enough in the trial to have made it to their 450-days assessment. Deducting the 12 undertreated from those would leave 19 optimized patients evaluable at 450 days, but the plot only shows 12. Second, among the undertreated there are two who made it CR all the way through 450 days, and one ended IR at 450 days (that is, if the data presented in MD&A 28/4/2022 is correct). These are included in the swimmers' plot (at least, that's what my tracking results suggest).
Now, how response numbers get translated into response rates is an entirely different story again. In the early stages they divided response numbers by number of patients treated, resulting in artificially low response rates. Then they changed to using evaluable patients, resulting in more accurate rates. And now they changed again, and have started using actually evaluated patients, resulting in much higher (C&I) response rates. What's correct and what's not? I don't dare say. As I mentioned, I presume there is protocol to follow when achievement of trial objectives is calculated (the second table in that respect is more important than the first one), and I expect the biostatistics/regulatory experts they have hired to know what they are doing.
Meanwhile, I'll just continue tracking any raw numbers they throw at us:).