RESUMEN
Background: Transforming growth factor beta (TGFß) is well-recognized as an immunosuppressive player in the tumor microenvironment but also has a significant impact on cancer cell phenotypes. Loss of TGFß signaling impairs DNA repair competency, which is described by a transcriptomic score, ßAlt. Cancers with high ßAlt have more genomic damage and are more responsive to genotoxic therapy. The growing appreciation that cancer DNA repair deficits are important determinants of immune response prompted us to investigate the association of ßAlt with response to immune checkpoint blockade (ICB). We predicted that high ßAlt tumors would be infiltrated with lymphocytes because of DNA damage burden and hence responsive to ICB. Methods: We analyzed public transcriptomic data from clinical trials and preclinical models using transcriptomic signatures of TGFß targets, DNA repair genes, tumor educated immune cells and interferon. A high ßAlt, immune poor mammary tumor derived transplant model resistant to programmed death ligand 1 (PD-L1) antibodies was studied using multispectral flow cytometry to interrogate the immune system. Results: Metastatic bladder patients in IMvigor 210 who responded to ICB had significantly increased ßAlt scores and experienced significantly longer overall survival compared to those with low ßAlt scores (hazard ratio 0.62, P=0.011) . Unexpectedly, 75% of high ßAlt cancers were immune poor as defined by low expression of tumor educated immune cell and interferon signatures. The association of high ßAlt with immune poor cancer was also evident in TCGA and preclinical cancer models. We used a high ßAlt, immune poor cancer to test therapeutic strategies to overcome its inherent anti-PD-L1 resistance. Combination treatment with radiation and TGFß inhibition were necessary for lymphocytic infiltration and activated NK cells were required for ICB response. Bioinformatic analysis identified high ßAlt, immune poor B16 and CT26 preclinical models and paired biopsies of cancer patients that also demonstrated NK cell activation upon response to ICB. Conclusions: Our studies support ßAlt as a biomarker that predicts response to ICB albeit in immune poor cancers, which has implications for the development of therapeutic strategies to increase the number of cancer patients who will benefit from immunotherapy.