The association of pre-existing autoimmune disease and immune-related adverse events secondary to immune checkpoint inhibition therapy in a UK multicenter cohort.

Abstract

2522 Background: Pre-existing autoimmune disease (AID) potentially increases the propensity for the development of immune related adverse events (irAE) in response to oncological immune checkpoint inhibitors (ICIs) is biologically plausible and clinically observed. However, due to consistent clinical trial exclusion of those with pre-existing AID, the impact on the frequency and severity of irAEs is uncertain. Here we analyse this relationship in a large, real-world, UK multi-centre cohort. Methods: A retrospective analysis of 2049 patients treated with ICIs over a two year period was undertaken across 12 National Health Service centres by the UK National Oncology Trainees Collaborative for Healthcare Research (NOTCH). Patients received ICIs as standard of care for malignant melanoma, non-small cell lung cancer and renal cell carcinoma. The presence of pre-existing AIDs was assessed and classified as either autoantibody driven or autoinflammatory then correlated with clinically significant irAEs (i.e. ≥grade 2 or all-grade endocrinopathies). Statistical analyses included T-test, Mann-Whitney and Chi-squared. For overall survival (OS) Kaplan-Meier and log-rank tests were utilised. Results: Pre-existing AID were present in 13% (n = 257) of the overall cohort. Pre-existing endocrinopathies (30%; n = 76) were most common followed by rheumatological AIDs (18%; n = 46). In the pre-existing AID cohort there was a female predominance (48% vs 39%; p = 0.006) but no difference in smoking history (p = 0.074) or ethnicity (p = 0.12). There was no difference in ICI treatment between those with and without pre-existing AID (p = 0.2800). IrAEs occurred in 45% (n = 117) patients with pre-existing AID vs 33% (n = 583) without (p£0.001). The median time to onset of irAEs was similar. IrAEs with an increased incidence in the pre-existing AID cohort were colitis (p = < 0.001), arthralgia (p = 0.008) and dermatological irAEs (p = 0.014). There was no difference in the incidence of irAEs in patients with autoantibody driven vs autoinflammatory pre-existing AID (44.0 % vs 44.8%, p = 0.905). In the overall cohort, those with pre-existing AIDs had a median OS of 20.4 months (95% CI: 19.4-21.7) vs 14.1 months (95% CI: 12.8-16.3) in those without pre-existing AID (p = 0.004). Conclusions: This large multi-centre ICI-treated cohort demonstrates that pre-existing AID is a predisposing factor for the development of irAEs, however the incidence is lower than previously quoted. The pathological basis of pre-existing AID did not differentially affect irAE manifestation. Patients with pre-existing AID had improved OS compared to those without which has not been observed in previously reported studies. ICI treatment should be considered in those with pre-existing AID but further studies are needed to determine how best to optimise outcomes whilst mitigating the impact of irAEs.