Mechanisms of cancer immunotherapy-associated thrombosis

Cancer immunotherapy is one of the most important advances in cancer treatment in decades, and has rapidly moved to front-line therapy for many cancers. The mechanism of cancer immunotherapy is to disable normal immunoregulatory pathways through administration of Immune Checkpoint Inhibitors (ICI), which are monoclonal antibodies directed toward key immune regulatory proteins including PD-1, PD-L1 and CTLA-4. Disabling these pathways enhances anti-tumor immunity. However, since these responses are not tumor-specific, ICIs are associated with a variety of immune-related adverse events (irAEs). We and others have recently reported a high incidence of thrombosis, which may exceed 20%, in patients treated with ICI; given the increasing use of ICIs in cancer treatment and the frequency of cancer diagnoses, it is clear that ICI-Cancer Associated Thrombosis (ICI-CAT) has become a major clinical problem and that better understanding of this disorder is urgently needed. However, there is little information available concerning mechanisms of ICI-CAT, and there are no published studies addressing this issue. We hypothesize that ICI-CAT is an irAE resulting from ICI-induced cellular activation and prothrombotic activity in the setting of underlying tumor-associated inflammation. Our murine model demonstrates that ICI-CAT requires the presence of an underlying tumor, with markedly increased expression of tumor cell tissue factor (TF) occurring after ICI treatment. Our model also supports a role for neutrophil extracellular traps (NETs) and platelet activation in ICI-CAT; platelet activation is also suggested in patients treated with ICI by our demonstration of increased levels of circulating platelet-neutrophil aggregates. On a cellular level, anti-PD-1 antibodies stimulate neutrophil NET release and prothrombotic activity, and may also enhance platelet activation in the presence of subthreshold thrombin concentrations. In this application, we propose to advance our understanding of ICI-CAT using both cellular and animal models, and to extend these studies to clinical samples from patients before and after initiating treatment with ICI. In Aim 1, we will determine the effect of different ICI and ICI combinations on the development of thrombi in tumor-bearing mice, and assess our model using other tumor types and mouse strains. We will examine thrombus size and composition, and identify critical cell types involved in thrombus formation by depleting mice of T cells, neutrophils, monocytes or platelets. We will also further define the cellular mechanisms involved in expression of prothrombotic activity using isolated leukocytes from normal human donors, mixed leukocyte populations, and endothelial cells, incubated with cytokines and ICI. In Aim 2, we will extend these studies by measuring mechanistic markers of inflammation and vascular activation in patient plasma before and after initiating ICI, and by comparing procoagulant gene expression in myeloid cells from patients treated with ICI who did, or did not develop thrombosis. Taken together, these studies should provide important new information leading to better understanding of mechanisms, and potentially new approaches for prevention and treatment of ICI-CAT.