Cancer immunotherapy is a form of cancer treatment that educates the immune system to recognize and attack cancer cells. In particular, two clinically approved antibodies known as immune checkpoint inhibitors (CPIs), which block cytotoxic T lymphocyte antigen-4 (CTLA-4) and programmed death-1 (PD-1) have provided improved survival benefits for patients across a range of cancer types. However, by activating the immune system, we also observe a broad range of immune-mediated side-effects. There are currently no biomarkers to identify which patients may be susceptible to the development of CPI-induced immunotoxicity and the mechanism by which they occur remains unclear due to the rarity of certain irAEs, their variable timing, and an inability to access the irAE-affected tissue site. In this study, we aimed to better define potential clinical and demographic information alongside the dysregulation of longitudinal serum proteins, such as cytokine responses, to assess their ability to predict the development of CPI-diabetes in CPI-treated stage IV melanoma patients. We identified that certain cytokines appeared to be upregulated preceding initiation of CPI treatment in patients with CPI-diabetes (including IL-27, EOTAXIN, SFAS/TNFRSF6), whereas others appeared to be significantly modulated during the treatment course (including IL1A and MIP1B/CCL4). By comparing serum concentrations of circulating factors at baseline, early during treatment and near to initiation of CPI-induced diabetes, we identified potential drivers of CPI-associated diabetes, which could represent biomarkers for risk or potential therapeutic targets to inhibit irAEs. We will next assess the mechanism for how these changes in the serum promote differential immune responses that could contribute to CPI-related diabetes. This study has allowed us to identify a potential predictive cytokine profile that may assist in determining cancer patients that are at higher risk of developing CPI-induced diabetes.