If you look at the Sept VBL slides that wcopeland linked, it clearly says N = 29 for that particular trial. That slide has the sample size for all the trials, including the year, median age, median survival, etc. It's all on that slide. 8 trials that they summarized in one slide.