Checkpoint inhibitor (CPI) therapy has greatly improved the treatment of cancers that had previously been considered intractable. As a result of the intended biologic activity of these drugs, which enable activation of T cells that can cause tumor destruction, new autoimmune adverse events have occurred. Major targets of these adverse events have been endocrine tissues including thyroid and beta cells in the pancreas. Thyroiditis is frequent and autoimmune diabetes has emerged as a serious adverse event often requiring intensive care. The reasons why some individuals develop these autoimmune events and why some are protected are not known but this information may lead to ways of preventing the occurrence of these adverse events. The overall objective if this proposal is to understand the molecular and cellular immunologic basis for autoimmune diabetes and thyroid disease and to test whether they can be prevented with agents that are specific in their actions. The hypothesis that we wish to test is that in individuals who develop endocrinopathies CPI therapy induces pathologic T cells, loss of B cell tolerance, and dysfunctional regulatory T cells. Our preliminary data has identified phenotypic differences in autoantigen reactive effector T cells and Tregs in those who do and do not develop endocrinopathies. In addition, our analysis of autoreactive B cells suggests that CPI therapy affects peripheral B cell tolerance checkpoints and results in an increased frequency of autoreactive mature nave B cells. We plan to analyze, using Seq-Well single cells in patients who are followed prospectively from before treatment to when they present with autoimmune endocrinopathies. In the subgroup of individuals who are HLA-A2 (~40%) we will study thyroid and diabetes reactive CD8+ T cells with cellular libraries and CyTOF and determine whether the T cell receptors that are found on the antigen reactive cells can be detected prior to treatment and in which subpopulation. We will also determine whether CPIs affect the number and function of Tregs and address specifically whether the CPI causes the Tregs to produce pathologic cytokines. We will use established techniques to determine whether the CPIs induce a failure of peripheral B cell tolerance and identify the relationship between changes in B cells and Tregs. We will be collecting data on the autoreactive T and B cell repertoire that we will correlate with clinical responses to the primary tumors. Finally, our studies of the immunologic mechanisms that underlie these adverse events suggest ways in which they may be prevented, which we will test in a murine model of anti-PD-L1 induced diabetes in NOD mice. We will test whether B cell depletion or enhancement of Tregs, either by low doses of IL-2 or by infusion of diabetes antigen specific Tregs can prevent diabetes onset. These studies therefore, will elucidate the mechanisms of these serious adverse events, identify individuals who are at greatest risk for these events, and test whether therapies that do not interfere with the anti-tumor effects of the CPIs can be used to prevent them.
This proposal will identify the mechanisms of autoimmune endocrinopathies that occur following checkpoint inhibitor therapy in patients with cancers. The studies may determine which individuals are at greatest risk for these events and suggest therapies that may prevent their occurrence.
