Yes .. It will always be trial specific in the sense of the...

Yes .. It will always be trial specific in the sense of the treatment being applied.

The treatment is attempting to outperform the control and so there needs to be some "events" to occur in the control and/or treatment groups for this to be quantified. This can happen slowly or quickly depending on how fast the survival curve of the treatment group separates itself away from the survival curve of the control group. There also needs to be enough participants and events for these differences to have some statistical significance. For any given trial, LC and her researchers would probably be the only ones who would have an inkling about how well the treatment will work in advance and subsequent approximate timelines to obtain an approrpiate number of events to bring the trial to a close. Although Im assuming you may be able to base this off the control group as they are most likely to have more events within a potentially estimable timeline.