Editors' ChoiceCancer Immunotherapy

Interfer-ing with immunotherapy-induced autoimmunity

See allHide authors and affiliations

Science Translational Medicine  09 Jan 2019:
Vol. 11, Issue 474, eaaw0533
DOI: 10.1126/scitranslmed.aaw0533


Inhibition of type I IFN prevents autoimmune toxicity while preserving antitumor efficacy of combined cancer vaccine and adoptive cell therapy.

Cancer immunotherapy—includings immune checkpoint inhibition (ICI), adoptive cell therapy (ACT), and cancer vaccines—is revolutionizing cancer treatment. Although ACT is successful in some hematological malignancies, it has limited effects in solid tumors. Therapeutic cancer vaccines have not achieved much clinical success so far. Additionally, immunotherapy can be accompanied by widespread and potentially fatal autoimmune toxicities. Intense investigation is under way to enhance the efficacy of immunotherapy and to minimize autoimmune toxicity.

Walsh and colleagues combined tumor antigen–specific T cell ACT with a viral cancer vaccine to efficiently treat established melanoma in mice, suggesting that the combination of vaccination and ACT, but not individual therapy, could be effective in solid tumors. As expected, their combination therapy induced autoimmunity. The location of the autoimmune response was dependent on where the tumor antigen, often a self-antigen, is expressed, confirming that T cell recognition of self-antigen in tumor leads to autoimmunity. Importantly, they discovered that type I interferon (IFN) induced by the cancer vaccine was coupled to the autoimmune response. Blockade of type I IFN signaling, either by administration of neutralizing antibody or by genetic ablation of type I IFN signaling in the mice, prevented autoimmunity. Moreover, utilization of an alternative viral vector with reduced type I IFN induction also inhibited the autoimmune response without compromising antitumor efficacy. Mechanistically, type I IFN increased major histocompatibility complex I (MHC I) expression on normal tissues, which permitted the autoimmune response. In contrast, their combination therapy induced MHC I expression on tumor cells independent of type I IFN.

This study identifies a potential target to reduce cancer immunotherapy–associated autoimmune toxicity in animal models of combined ACT and cancer vaccine therapy. Several key questions remain. Does the pathway apply to other modes of cancer immunotherapy with significant autoimmune toxicity, such as ICI therapy? More importantly, whether this insight can be translated into human clinical application awaits further investigation. In addition, type I IFN is also known to enhance memory T cell differentiation. Thus, it remains to be determined whether type I IFN blockade may inadvertently increase cancer relapse in the long run.

Highlighted Article

View Abstract

Navigate This Article