I tried this once before and made every error possible. This time I think I have it right. I repeated KISS (keep it simple stupid) to myself as I figured. My thesis is: The Yin study first-treated cohort results, the most comparable to the HEAT trial control arm per the CLSN CMO, and the results of the T-dox P1 trial, show that is the worst case scenario T-dox PFS median time will best the 1.33 P3 trial requirement but is far more likely to double or triple the control arm median time, and therefore January is unlikely to be a binary event and the CLSN share price is in little danger of collapse from a results failure.
In Yin PFS was not calculated. Only DR (distant recurrence) and LR (local recurrence) were calculated. LR was about 22% for both first and prior treated cohorts, with a 4.6 month median LR time. Conservatively only about 12% of these LRs will be below the PFS of the first-treated in Yin and therefore 38% of these PFS events will come from DRs. The DRs in Yin first-treated reach 38% at about 8 months per the chart on page 1919 of Yin. Therefore the 50% median PFS event in Yin first-treated came at about 8 months. If the HEAT control arm mimics this then its PFS median event will also come at 8 months. In the P1 T-dox trial there was only one LR out of 21 complete ablations or 4.8% LRs. Let's say this 5% is matched by the treatment arm LRs in the P3 HEAT trial and, to be conservative, that all 5% are below the T-dox arm median PFS. Therefore 45% of T-dox arm below the median will come from DRs. The DR rate in Yin is 45% at about 11 months so if T-dox had no effect on DRs and only matched Yin (based on ablation only) then the T-dox P3 median would come at 11 months. 11/8 = 1.375, which passes the HEAT trial requirement of 1.33 treatment over control. HOWEVER, the P1 trial also showed that T-dox doubled the TTF (time to treatment failure), a measure similar to PFS, in the very sick P1 patients and therefore must also be very effective against DRs. So it is likely that the T-dox P3 arm will double or triple the control arm's PFS time in its much less sick P3 patients. Therefore failure in P3 is very unlikely.