OR18-05 Interleukin 21 Drives The Development Of Autoimmune Diabetes Mellitus During Immune Checkpoint Inhibitor Cancer Therapy

Disclosure: J.G. Ortega: None. L.N. Scott: None. S. Wang: None. S. Zhang: None. H. Chen: None. N. Huang: None. M. Lechner: None. Autoimmune diabetes mellitus is a rare but life-threatening side effect of immune checkpoint inhibitor (ICI) cancer therapy. Checkpoint inhibitors increase immune activation by blocking natural regulatory molecules on T cells, including programmed death protein (PD1) and cytotoxic T lymphocyte antigen (CTLA4). As a result of increased immune activation, up to 60% of patients may develop autoimmunity in healthy tissues. Endocrine tissues are among the most often affected by these immune related adverse events (IRAEs). ICI-associated diabetes mellitus (ICI-T1DM) results from autoimmune destruction of insulin-producing pancreatic beta-cells. Patients with ICI-T1DM frequently present in diabetic ketoacidosis and require life-long insulin therapy. Unfortunately, the cause of many IRAEs remains poorly understood. We recently developed a mouse model of IRAEs in which anti-PD1 + anti-CTLA4 antibody treatment of autoimmunity-prone non-obese diabetic mice leads to multi-organ autoimmune infiltrates, including acceleration of insulinitis and DM. ICI-treated mice had increased CD4(+) T follicular (Tfh; PD1(+)ICOS(+)CXCR5(+)) and T peripheral helper (Tph; PD1(+)ICOS(+)CXCR5(-)) cells compared to isotype-treated controls. Such Tfh and Tph cells secrete interleukin (IL)-21 and can promote cytotoxic CD8(+) T cell function in autoimmunity. Indeed, CD8(+) T cells expressing Granzyme B, interferon gamma (IFNγ), and chemokine receptor CXCR6 ("CD8(+) autoimmune mediators") were expanded in ICI-treated mice. Furthermore, in vitro, IL21 induced the autoimmune mediator phenotype in CD8(+) T cells, with upregulation of IFNγ, cytotoxic and trafficking molecules. Using mice with genetic deletion of T cell receptor alpha (NOD.TCRa(-/-)), we showed that CD4(+) and CD8(+) T cells were required for the development of IRAEs in ICI-treated mice. In addition, adoptive transfer of CD4(+) and CD8(+) T cells into immunodeficient NOD.SCID mice confirmed that these populations were sufficient to cause DM during ICI treatment. Thus, we predicted that IL21 from Tfh and Tph cells contributes to the development of pancreas IRAEs by activating CD8(+) autoimmune mediator T cells. Indeed, in our model, mice with genetic blockade of IL21 signaling (IL21R(-/-)) were protected from developing diabetes. After 4 weeks of ICI treatment, only one of 11 (9.1%) IL21R(-/-) mice developed DM with onset at day 28, compared to 80% of ICI-treated wild-type (WT) mice, with a median DM-free time of 10 days from start of ICI therapy (p<0.0001, Log Rank test); no isotype-treated WT or IL21R(-/-) mice developed diabetes. Histologic analyses of pancreatic tissue from ICI-treated mice also showed a significantly lower insulinitis index in IL21R(-/-) compared to WT mice (mean 0.19±0.06 vs. 0.7±0.14, p<0.01). In summary, we identify IL21(+) CD4(+) Tfh/Tph cells as drivers of pancreas IRAEs and a potential therapeutic target to prevent ICI-T1DM in cancer patients. Presentation: Saturday, June 17, 2023