Reversing Imatinib's Immunosuppressive Effects by Modulating Type I IFN Signaling

Multimodal therapies are considered by many to be the future for cancer treatment. However, combining therapies that target activated oncogenes and those that target immune signaling pathways is fraught with complications. Many immune and oncogene signaling pathways overlap. Thus, treatment of tumor autonomous oncogene-activated signaling pathways may negatively affect the immunogenicity of the tumor cells themselves and/or the activity of immune cells. The type I interferons (IFN), IFNα and IFNβ, have direct and indirect roles in antitumor immunity, and many new and exciting IFN-directed therapeutic opportunities are being explored (1). Activated oncoprotein kinases like KIT can activate IFN receptor (IFNAR)–stimulated pathways, including STAT1. As described in this issue (2), Liu and colleagues hypothesized that activated KIT oncogene expression in gastrointestinal stromal tumors (GIST) contributes to tumor cell type I IFN signaling and that KIT inhibition with the FDA-approved drug imatinib would reduce tumor immunity, thereby compromising therapy.

Activation of KIT is the most common driver alteration in GIST and has been successfully modeled in Kit^V558D/+ mice (3). This model has been useful for studying small-molecule kinase inhibition of a driver oncoprotein in an in vivo, immunoproficient setting. Dendritic cell GIST infiltration is reduced by imatinib treatment in this model, reducing antitumor immunity (4). GIST has an immune microenvironment profile suggesting patients could benefit from checkpoint blockade therapy, but a recent clinical trial in which ipilimumab was combined with imatinib showed low activity and no clear synergy (5). The work of Liu and colleagues in the Kit^V558D/+ mouse model and human GIST may partially explain this dilemma.

Liu and colleagues found that imatinib treatment reduces the expression of genes required for optimal response to type I IFN treatment including the IFN receptor genes Ifnar1 and Ifnar2, as well as Stat1. Indeed, activated KIT is required for optimal type I IFN response because it activates STAT1 directly. The outcome of imatinib treatment is reduced MHC class I expression in the Kit^V558D/+ mouse model of GIST and reduced human leukocyte antigen class I expression in human GIST cells. In Kit^V558D/+ mice, it is shown that lack of type I IFN signaling enhances tumor growth, and this is associated with reduced MHC class I expression and CD8⁺ T-cell infiltration. Importantly, the authors find that cGAS/STING activation using a small-molecule agonist that raises type I IFN transcription, when given in a course prior to imatinib treatment, leads to more effective therapy. Workers in the field will be motivated to determine the optimal forms and schedule for combining immunotherapy with KIT inhibition to achieve a durable and complete remission. As Liu and colleagues point out, use of cGAS/STING agonists, IFNα, and other approaches are all options at present. This goal is critical given that KIT inhibition for GIST typically culminates in resistance and rarely produces a cure. An appropriate way to combine kinase inhibition and immunotherapy in GIST could be paradigmatic for the other many solid tumors driven by activated kinase oncogenes.

D.A. Largaespada reports other support from NeoClone Biotechnology, Inc., Luminary Therapeutics, Inc., Recombinetics, Inc., Luminary Therapeutics, Inc., Styx Biotechnologies, Inc., and Immusoft, Inc. and grants from Genentech, Inc. outside the submitted work.