@Owl vs Fox, nice find. Another nerve-racking and/or exciting...

@Owl vs Fox, nice find. Another nerve-racking and/or exciting few months ahead of us. Note the statement that "The study has completed enrollment and final data is expected in late 2021".

The additional 9 patients (to reach the final sample of 36 patients) from the 27 patients included in the interim results are presumably randomised on 1:1 allocation (i.e. 1:1 treatment:control). We also know that there are 24 PFS events according to the 21 April 2021 announcement. The obvious missing part is what the split of those patients with PFS events is across the existing and additionally enrolled patients.

I created some dummy data to simulate some survival curves to try to replicate the interim survival curves but it's almost impossible because I don't have the precise time-to-event values! I produced some very similar survival curves to the interim results but with different p values. The hazard ratios (HR) are based on Cox proportional hazard regression which represent risk over a time period. I thought I could get away with calculating relative risk (RR) as we have the observed number of PFS events but it wouldn't be appropriate because it's different from the formal measure in the Her-Vaxx study.

There's also the overall survival (OS) which is showing a lower HR compared to PFS. In the case that the final PFS result is not statistically significant due to more PFS event favouring the treatment arm, we should remember that there is still the OS result that could still be statistically significant which is in fact the primary outcome measure for the study. We shall find out soon enough.