Abstract
Tumor-associated antigens (TAA) are important targets for cancer vaccines. However, TAA-based vaccines have not yet achieved their full potential in clinical trials. In contrast, immune checkpoint blockade (ICB) has emerged as an effective therapy, leading to durable responses in selected patients with cancer. To date, few generalizable associations between TAAs and ICB benefit have been reported, with most studies focusing on melanoma, which has the highest mutation rate in cancer. In this study, we developed a TAA burden (TAB) algorithm based on known and putative TAAs and investigated the association of TAB with ICB benefit. Analysis of the IMvigor210 patient cohort of urothelial carcinoma treated with anti-PDL1 revealed that high tumor mutation burden weakened the association of TAB with ICB benefit. Furthermore, TAB correlated with ICB efficacy in tumors characterized by negative PDL1 staining on immune cells; however, high levels of PDL1 staining on immune cells were linked to T-cell exhaustion. Validation across independent clinical datasets-including urothelial carcinoma cohorts treated with anti-PD1/PDL1 agents and neoadjuvant anti-PD1 trials for head and neck cancers-corroborated the finding that TAB correlates with ICB benefit in tumors with low T-cell exhaustion. Pan-cancer analyses revealed that in most cancer entities, tumors with higher T-cell exhaustion exhibited lower TAB levels, implying possible immunoediting of TAAs in tumors with established antitumor immunity. Our study challenges the prevailing notion of a lack of association between TAAs and ICB response. It also underscores the need for future investigations into the immunogenicity of TAAs and TAA-based vaccine strategies in tumors with low levels of T-cell exhaustion.