Objective: Type 1 diabetes mellitus (T1DM) is a rare, irreversible immune-related adverse event reported in patients receiving treatment with immune checkpoint inhibitors (ICI). However, clinical risk factors for ICI-induced T1DM (ICI-T1DM) and its impact on survival in patients remain unknown.
Research design and methods: We used Optum's Clinformatics Data Mart database for assessment of the incidence and characteristics of T1DM in a large de-identified cohort of patients treated with ICI between 2017 and 2020. We applied Fine-Gray and cause-specific hazard models to study associations between patient/treatment characteristics and ICI-T1DM and applied the Cox model with ICI-T1DM as a time-varying covariate to assess the impact of ICI-T1DM on survival.
Results: ICI-T1DM was observed in 261 of 30,337 (0.86%) patients. Dual use of antibodies to cytotoxic T lymphocyte antigen 4 (CTLA-4) and programmed cell death 1 (PD-1) or programmed cell death ligand 1 (PD-L1) was associated with increasing risk of ICI-T1DM (hazard ratio [HR] 1.62; 95% CI 1.15-2.26) vs. anti-PD-L1 or anti-PD-1 alone. Younger age (HR 1.19 for every 5-year decrease; 95% CI 1.13-1.25) and preexisting non-T1DM diabetes (HR 4.48; 95% CI 3.45-5.83) were also associated with higher risk of ICI-T1DM. Conversely, prior use of immunosuppressive medications (HR 0.57; 95% CI 0.34-0.95) was associated with lower incidence of ICI-T1DM, but part of its protective effect may be due to the increased mortality rate. Development of ICI-T1DM does not seem to significantly impact patient survival.
Conclusions: The risk of ICI-T1DM is associated with the type of ICI therapy, patient age, and preexisting non-T1DM diabetes. These data may help guide risk assessment and screening practices for patients during ICI therapy.
© 2022 by the American Diabetes Association.