Senti, major correction for post 191145. I think I got it right this time.
If there were 22 met+ patients at 18 months and one is asking how many are expected to be alive at 24 months, would it not be appropriate to look at curve b (publication). The survival for met+ at 18 months is about 81% (268 of 331 patients) and going down to 66.7% or 220.7 patients at 24 months (curve b and Table 2). So from 18 to 24 months we have a reduction of 14.3% of 331 or 47.3 patients.
Now a 47.3 patients loss from 268 to 220.7 is 17.6% not 14.3% (the % loss from 331) and we will end up at 24 months with 220.7 patients (268-47.3 The calculation that gets us there would be
66.7%X100/81%=82.3% (the survival %). 268X82.3%=220.6 (pretty close to 220.7)
Now applying this to the 22 patients going from 18-24 months we do the parallel calculation namely: 66.7%X100/81%=82.3% (survival %) 82.3%X22=18.1 patients left at 24 months from the previous 22.
Utilizing 21, 24, 27, and 33 months as the midpoints of the 6 months time spans 18-24, 24-30 and 30-36, the calculation method above and recalculating the survivor number for every increasing 6 month time frame, I estimated that of the 149 patients processed before 36 months (3/14), about 25 met+ and 9 met- will makr it to 36 months. With the previous 44 that would be a total of only 78 patients. If enough of them survive to 5 years that would be enough for approval especially if most of them are met+ then the approval for that category seems very strong. Also I hope that the last contingent of 108 will surprise us and will do better than my calculations indicate.
I will now proceed to detail the calculations and estimates in a post to follow.
