Immune checkpoint blockade, as a breakthrough in cancer therapy with anti-PD-L1, anti-PD-1, and anti- CTLA4, has demonstrated impressive therapeutic effects in multiple clinical trials. However, only a small minority of patients respond to such therapies [1]. Higher tumor infiltrating lymphocytes (TILs) and PD-L1 expression in tumors have been found to correlate with better prognoses [2, 3]. However, due to a lack of proper tumor models, it is difficult to determine the exact contributions of PD- L1 expression or TILs to checkpoint blockade therapies. In a recent study published in Cancer Cell, Tang H. et al. performed an extensive investigation to address these issues, and provided strong evidence for designing therapeutic strategies to extend the benefits of checkpoint blockade for more patients [4].