To the Editor Duan et al1 conducted a well-designed meta-analysis comparing the efficacy and safety of programmed cell death 1 (PD-1) and programmed cell death ligand 1 (PD-L1) inhibitors for treating cancer. Most analyses in their study focused on overall pooled outcomes for all cancer types. Subgroup analyses for tumor types could only be found in the OS Comparison: Frequentist Approach section in the Results. In other parts of the Results section, the subgroup of tumor type was only used to examine the source of heterogeneity and verify the stability of overall pooled outcomes in sensitivity analysis. Additionally, the authors did not mention their results concerning tumor types in the Discussion or Conclusions sections. In the Conclusions, the statement “anti–PD-1 exhibited better survival outcomes than anti–PD-L1 in patients with solid tumors…with comparable safety profiles”1 easily misled readers to believe that anti–PD-1 was better than anti–PD-L1 for all types of solid tumors. In fact, tumor type did make a difference in the selection of anti–PD-1 or anti–PD-L1 according to the authors’ analyses. For example, in the comparison of overall survival (OS) through the frequentist approach, the overall pooled hazard ratio indicated that anti–PD-1 provided better OS than anti–PD-L1, while we noticed that anti–PD-1 and anti–PD-L1 showed no significant difference in urothelial cancer and renal cell carcinoma. Although the authors explained that failure to reach statistical significance in these 2 tumor types was because of insufficient statistical power with only 1 mirror group available for each indication, we thought it was unreasonable because numbers of participants in both the urothelial cancer and renal cell carcinoma groups were not small and even larger than other mirror groups. Furthermore, in the Safety Analysis section, the authors reported that overall safety profiles of anti–PD-1 and anti–PD-L1 were comparable for any adverse events, but they never stated or emphasized that anti–PD-1 was significantly more toxic than anti–PD-L1 in gastric carcinoma for any adverse events (any grades: risk ratio, 2.42; 95% CI, 1.74-3.38; grades 3-5: risk ratio, 7.28; 95% CI, 3.09-17.11).