Background: Immune checkpoint inhibitors (ICIs) are potent new cancer therapies but can cause serious immune-related adverse events. Radiation therapy (RT) also induces systemic immunologic effects, and data on the interaction and safety of combining ICIs and RT are limited. Methods: In this retrospective cohort study using a large medical claims database from 2010 to 2017, we ascertained the risk of immunosuppressive steroid therapy as well as the risk of hospitalization within 180 days of treatment with an ICI in patients with diagnoses of malignant melanoma or lung cancer. Patients were stratified by use of RT within 30 days before and after ICI therapy. ICIs included pembrolizumab, nivolumab, and ipilimumab, while immunosuppressive agents included oral prednisone and intravenous methylprednisolone. Results: 2020 patients (218 with RT, 1802 without RT) met inclusion criteria for prednisone analysis, while 3519 patients (361 with RT, 3158 without RT) met inclusion criteria for all other analyses. On univariable analysis, RT was not associated with need for prednisone or methylprednisolone (RR 1.2, 95%CI 0.8-1.9 and RR 1.2, 95%CI 0.7-2.1 respectively). When assessing hospitalization, RT was significantly associated with hospitalization following ICI therapy for certain cancer/drug combinations (RR 1.4, 95%CI 1.2-1.6, p<0.001 for lung cancer/PD-1 inhibitors, RR 2.0, 95%CI 1.0-3.5, p=0.03 for melanoma/ipilimumab). Conclusions: In patients treated with ICIs, receiving RT was not associated with a higher risk of requiring immunosuppressive steroid therapy as compared to not receiving RT. However, in those with ICIs, RT was associated with a higher risk of hospitalization as compared to no RT, though this may be a result of underlying differences in patent severity (more severe disease may require ICI and RT).

^aNo values were significantly different at p<0.05. ^bThe patient populations for prednisone and methylprednisolone are not identical because the separate analyses required distinct inclusion criteria.