RE: Estimates of MoS for r/r HR MDS with Bex & AZA12 Aug 2025 08:29
If we also factor in the average TP-53 mutation burden in HR MDS, is around 20%, we have ~double the amount of TP-53 patients than would be expected in a stratified for TP-53, double blinded, randomised, phase 3 pivotal trial. This obviously lowers the overall groups median overall survival, given the stark difference in survival between the 2 groups.
I asked ChatGPT5 to factor this in as if it were an average sized TP-53, relative to average prevalence, see below:
The higher the TP53 fraction, the more the overall median is pulled down toward the poor TP53 median (9.3 mo).
For a TP53 prevalence of ~15–25% (typical for many HR‑MDS cohorts), the combined median tends toward the mid‑20s to ~30 months, consistent with our earlier simple weighted approximation but now computed exactly from a Weibull mixture.
The exact combined median is sensitive to the Weibull shape assumptions (how hazard changes over time). I included shape sensitivity to show that uncertainty.
I don’t think I need to show how we are producing data that is highly statistically significant over what we would expect using standard of care as the comparator.
I can see why Juho, at Capital Markets Day 2024, said that he would be so interested to see what the FDA says as the r/r group survival data matures.
Basically, we are potentially heading towards survival rates eclipsing those seen in the front line under standard of care, in the relapsed / refractory population. That is a massive advance in the treatment of HR MDS, and if I can calculate this, I am quite sure the big pharma BD teams are anticipating the same scenarios.
The FDA End of phase 2 meeting gets closer everyday.
Good luck all…