3) Instead of looking at PD-L1, we could look at immune gene signatures.
CD8 T cells are associated to response with ICI but not with BCG.
Likely because different immunotherapies are likely to need different biomarkers
How would that be translated into the vaccine trials?