what could go wrong?, page-91

You've obviously done some home work Southoz...

Re the extra 25

"We have been encouraged by the initial statistical review and modeling of the available blinded data, as it suggests that patients as a group are continuing on treatment for longer than anticipated before their disease progresses"

The answer as to why they recruited 25 additional patients could be quite simple, and it clearly was not a planned event. As above, they'd worked out from their modeling that patients were progressing at a slower rate and remaining on treatment for longer than anticipated- however they had no way of knowing if one arm was progressing at a slower rate than the other( since the trial was still randomized/ double blinded). Therefore they've gone ahead and recruited additional patients to help maintain the power of the study, in-case BOTH arms were progressing at a slower rate...

Re Different patients population

Yes you're correct, there will be some variance in patients between different international recruitment sites. In studies comparing Irontecann vs FOLFIRI, there's also been a variance from different ethnic groups in terms of response rate, and reported toxicities...