Neoantigen architectures define immunogenicity and drive immune evasion of tumors with heterogenous neoantigen expression.

Background: Intratumoral heterogeneity (ITH) and subclonal antigen expression blunt antitumor immunity and are associated with poor responses to immune-checkpoint blockade immunotherapy (ICB) in patients with cancer. The underlying mechanisms however thus far remained elusive, preventing the design of novel treatment approaches for patients with high ITH tumors.

Methods: We developed a mouse model of lung adenocarcinoma with defined expression of different neoantigens (NeoAg), enabling us to analyze how these impact antitumor T-cell immunity and to study underlying mechanisms. Data from a large cancer patient cohort was used to study whether NeoAg architecture characteristics found to define tumor immunogenicity in our mouse models are linked to ICB responses in patients with cancer.

Results: We demonstrate that concurrent expression and clonality define NeoAg architectures which determine the immunogenicity of individual NeoAg and drive immune evasion of tumors with heterogenous NeoAg expression. Mechanistically, we identified concerted interplays between concurrent T-cell responses induced by cross-presenting dendritic cells (cDC1) mirroring the tumor NeoAg architecture during T-cell priming in the lymph node. Depending on the characteristics and clonality of respective NeoAg, this interplay mutually benefited concurrent T-cell responses or led to competition between T-cell responses to different NeoAg. In tumors with heterogenous NeoAg expression, NeoAg architecture-induced suppression of T-cell responses against branches of the tumor drove immune evasion and caused resistance to ICB. Therapeutic RNA-based vaccination targeting immune-suppressed T-cell responses synergized with ICB to enable control of tumors with subclonal NeoAg expression. A pan-cancer clinical data analysis indicated that competition and synergy between T-cell responses define responsiveness to ICB in patients with cancer.

Conclusions: NeoAg architectures modulate the immunogenicity of NeoAg and tumors by dictating the interplay between concurrent T-cell responses mediated by cDC1. Impaired induction of T-cell responses supports immune evasion in tumors with heterogenous NeoAg expression but is amenable to NeoAg architecture-informed vaccination, which in combination with ICB portrays a promising treatment approach for patients with tumors exhibiting high ITH.