STA and MG

[biopearl123]

Something you might find helpful is to take a look at MF registries around the world in areas that overlap with Impact centers with patient entry times after 2019. The Moffett data seemed like such an outlier but only looked at 27 patients with a very wide confidence interval and lots of censored patients, but as I mentioned previously I drew a sharp breath when I saw it, the implications being obvious as we looked at it from 20 thousand different angles. It is probably reasonable to think that many US and European centers will also show longer mOS as well, but as we mentioned DIPSS class and inclusion of access or lack of, to other newer JAKi drugs pre or post Impact (ex US) could affect ultimate OS as well as changes in Rux usage as noted by the Moffett team. In any case the non US and non European MF registries have catalogued large numbers of patients that may not be subjected to as much of the "era drift" we are trying to understand. It might be worth it (if you have time on your hands) to try to duplicate DIPSS class, time on Rux to failure, time after failure etc to death etc and in that way get a better sense for what the real BAT is likely to show in other geographical areas. Entry criteria is very strict for Impact less so for some of the registries. Also the points raised about the back loading of patients I think will mean many more later enrollees will still be alive at the time of interim meaning we have to depend more on the early enrollees to reach the trigger. As you look at the K-M curves there is a steep drop off of early deaths followed by a much gentler slope of later deaths so deaths are contributed more slowly the more distant the entry. I know everyone is a little sick of this but it is interesting.