Targeting IDH2^R140Q and other neoantigens in acute myeloid leukemia

• Neoantigens arising from recurrent AML mutations are immunogenic and represent novel immunotherapeutic targets.

• IDH2^R140Q-specific T cells can selectively eliminate neoantigen/HLA-matched AML in vitro and in vivo.

For patients with high-risk or relapsed/refractory acute myeloid leukemia (AML), allogeneic stem cell transplantation (allo-HSCT) and the graft-versus-leukemia effect mediated by donor T cells, offer the best chance of long-term remission. However, the concurrent transfer of alloreactive T cells can lead to graft-versus-host disease that is associated with transplant-related morbidity and mortality. Furthermore, ∼60% of patients will ultimately relapse after allo-HSCT, thus, underscoring the need for novel therapeutic strategies that are safe and effective. In this study, we explored the feasibility of immunotherapeutically targeting neoantigens, which arise from recurrent nonsynonymous mutations in AML and thus represent attractive targets because they are exclusively present on the tumor. Focusing on 14 recurrent driver mutations across 8 genes found in AML, we investigated their immunogenicity in 23 individuals with diverse HLA profiles. We demonstrate the immunogenicity of AML neoantigens, with 17 of 23 (74%) reactive donors screened mounting a response. The most immunodominant neoantigens were IDH2^R140Q (n = 11 of 17 responders), IDH1^R132H (n = 7 of 17), and FLT3^D835Y (n = 6 of 17). In-depth studies of IDH2^R140Q-specific T cells revealed the presence of reactive CD4⁺ and CD8⁺ T cells capable of recognizing distinct mutant-specific epitopes restricted to different HLA alleles. These neo–T cells could selectively recognize and kill HLA-matched AML targets endogenously expressing IDH2^R140Q both in vitro and in vivo. Overall, our findings support the clinical translation of neoantigen–specific T cells to treat relapsed/refractory AML.