Association Between Rheumatic Autoantibodies and Immune-Related Adverse Events

Background: Side effects of immune checkpoint inhibitors (ICIs), called immune-related adverse events (irAEs), closely resemble primary autoimmune or rheumatic diseases. We aimed to understand the clinical utility of rheumatic autoantibodies (rhAbs) for diagnosing irAEs. Patients and Methods: Patients without pre-existing autoimmune disease (pAID) who had cancer treated with ICI(s) treatment from 1/1/2011 to 12/21/2020 and a rhAb checked were retrospectively identified. Logistic regression assessed associations between autoantibodies and irAEs, cancer outcome, and survival. Specificity, sensitivity, and positive/negative predictive values (PPV, NPV) were estimated for key rhAbs and ICIarthritis. Kaplan-Meier analyzed objective response rate (ORR) and overall survival (OS). Results: A total of 2662 patients were treated with >= 1 ICIs. One hundred and thirty-five without pAID had >= 1 rhAb tested. Of which 70/135(52%) were female; median age at cancer diagnosis was 62 years with most common cancers: melanoma (23%) or non-small cell lung cancer (21%), 96/135 (75%) were anti-PD1/PDL1 treated. Eighty had a rhAb ordered before ICI, 96 after ICI, and 12 before and after. Eighty-two (61%) experienced an irAE, 33 (24%) with rheumatic-irAE. Pre-ICI RF showed significant association with rheumatic-irAEs (OR = 25, 95% CI, 1.52-410.86, P =.024). Pre- and post-ICI RF yielded high specificity for ICI-arthritis (93% and 78%), as did pre- and post-ICI CCP (100% and 91%). PreICI RF carried 93% NPV and pre-ICI CCP had 89% PPV for ICI-arthritis. No variables were significantly correlated with ORR. Any-type irAE, rheumatic-irAE and ICI-arthritis were all associated with better OS (P =.000, P =.028, P =.019). Conclusions: Pre-ICI RF was associated with higher odds of rheumatic-irAEs. IrAEs had better OS; therefore, clinical contextualization for rhAbs is critical to prevent unnecessary withholding of lifesaving ICI for fear of irAEs.