Your conclusion is only one of many speculative possibilities. Instead, it could be that the original OS (second) endpoint (now the fourth endpoint) is statistically significant (or very positively trending), but for a more granular filtered look at this efficacy impact, the new primary endpoint then goes a step further and filters out cross-over patients by instead using an external arm composed of highly matched standard of care patients with nGBM from “comparable” “ contemporaneous” trials. In this way, efficacy is easier to evaluate for more pragmatic concerns like QALY calculations.