Identifying strategies to improve the efficacy of immune checkpoint blockade (ICB) remains a major clinical need. Here, we show that therapeutically targeting the COX2/PGE2/EP2-4 pathway with widely used nonsteroidal and steroidal anti-inflammatory drugs synergized with ICB in mouse cancer models. We exploited a bilateral surgery model to distinguish responders from nonresponders shortly after treatment and identified acute IFNγ-driven transcriptional remodeling in responder mice, which was also associated with patient benefit to ICB. Monotherapy with COX2 inhibitors or EP2-4 PGE2 receptor antagonists rapidly induced this response program and, in combination with ICB, increased the intratumoral accumulation of effector T cells. Treatment of patient-derived tumor fragments from multiple cancer types revealed a similar shift in the tumor inflammatory environment to favor T-cell activation. Our findings establish the COX2/PGE2/EP2-4 axis as an independent immune checkpoint and a readily translatable strategy to rapidly switch the tumor inflammatory profile from cold to hot. Significance: Through performing in-depth profiling of mice and human tumors, this study identifies mechanisms by which anti-inflammatory drugs rapidly alter the tumor immune landscape to enhance tumor immunogenicity and responses to immune checkpoint inhibitors. See related commentary by Melero et al., p. 2372. This article is highlighted in the In This Issue feature, p. 2355
Acute myeloid leukemia (AML) is a hematological malignancy caused by clonal expansion of myeloid progenitor cells. Most patients with AML respond to chemotherapy, but relapses often occur and infer a very poor prognosis. Thirty to thirty-five percent of AMLs carry a four base pair insertion in the nucleophosmin 1 gene (NPM1) with a C-terminal alternative reading frame of 11 amino acids. We previously identified various neopeptides from the alternative reading frame of mutant NPM1 (dNPM1) on primary AML and isolated an HLA-A*02:01-restricted T-cell receptor (TCR) that enables human T-cells to kill AML cells upon retroviral gene transfer. Here, we isolated T-cells recognizing the dNPM1 peptide AVEEVSLRK presented in HLA-A*11:01. The TCR cloned from a T-cell clone recognizing HLA-A*11:01+ primary AML cells conferred in vitro recognition and lysis of AML upon transfer to CD8 cells, but failed to induce an anti-tumor effect in immunodeficient NSG mice engrafted with dNPM1 OCI-AML3 cells. In conclusion, our data show that AVEEVSLRK is a dNPM1 neoantigen on HLA-A*11:01+ primary AMLs. CD8 cells transduced with an HLA-A*11:01-restricted TCR for dNPM1 were reactive against AML in vitro. The absence of reactivity in a preclinical mouse model requires further preclinical testing to predict the potential efficacy of this TCR in clinical development.
Identifying strategies to improve the efficacy of immune checkpoint blockade (ICB) remains a major clinical need. Here, we show that therapeutically targeting the COX2/PGE2/EP2-4 pathway with widely used nonsteroidal and steroidal anti-inflammatory drugs synergized with ICB in mouse cancer models. We exploited a bilateral surgery model to distinguish responders from nonresponders shortly after treatment and identified acute IFNγ-driven transcriptional remodeling in responder mice, which was also associated with patient benefit to ICB. Monotherapy with COX2 inhibitors or EP2-4 PGE2 receptor antagonists rapidly induced this response program and, in combination with ICB, increased the intratumoral accumulation of effector T cells. Treatment of patient-derived tumor fragments from multiple cancer types revealed a similar shift in the tumor inflammatory environment to favor T-cell activation. Our findings establish the COX2/PGE2/EP2-4 axis as an independent immune checkpoint and a readily translatable strategy to rapidly switch the tumor inflammatory profile from cold to hot.Significance:Through performing in-depth profiling of mice and human tumors, this study identifies mechanisms by which anti-inflammatory drugs rapidly alter the tumor immune landscape to enhance tumor immunogenicity and responses to immune checkpoint inhibitors.See related commentary by Melero et al., p. 2372.This article is highlighted in the In This Issue feature, p. 2355