Yeah there is something behind all the patients living longer.
Of the 960 actual glioblastoma patients they screened, they kicked out 250 for progressing. When you read a 5 year survival rate for glioblastoma, that doesn't exclude patients who progress early. Not to mention all the other patients that didn't pass the screening, many for various reasons that would have had the "not living longer".
It's a classic selection bias issue that will end the same way as CLDX, with the control arm greatly outperforming expectations.