Immune related adverse events in patients with breast cancer and autoimmune disease treated with immunotherapy.

1103 Background: Pembrolizumab is approved for early and advanced triple negative breast cancer (TNBC), and atezolizumab was approved for advanced TNBC. Toxicity of immunotherapy (IO) and immune related adverse events (irAEs) in patients with breast cancer with an underlying autoimmune disease are not well understood as autoimmune disease is often an exclusion criterion in immunotherapy trials. Understanding real-world IO toxicity and irAEs in patients with breast cancer and autoimmune disease may inform clinical decision making. Methods: We studied IO toxicity and irAEs in patients with early and metastatic breast cancer, and underlying autoimmune disease vs. those without autoimmune disease (at IO start) who received IO at an academic institution. A retrospective review was conducted to identify IO toxicity and irAEs (CTCAE v 5.0). Cohorts were compared with Pearson’s chi-squared test (categorical variables) and Wilcoxon rank-sum test (continuous variables), with p<0.05 for statistical significance. Results: Cohorts had 23 patients with breast cancer and ≥1 autoimmune disease (AD) prior to receipt of IO (thyroid: 48%, rheumatologic: 26%, gastrointestinal: 13%, dermatologic: 9%), and 106 patients with breast cancer but without autoimmune disease (non-AD). Median age at IO start was 63 (interquartile range [IQR] 54-69) in AD vs. 50 in non-AD (IQR 39-59). Stage I-III/IV distribution was 39%/61% for AD and 67%/33% in non-AD. IO was mainly pembrolizumab (AD: 87%; non-AD: 85%) vs. atezolizumab. Table depicts characteristics of IO toxicity. A significantly higher overall rate of irAEs was seen in AD (96%) vs. non-AD (59%) [p<0.001]. Significantly greater immune related transaminitis was observed in AD (52%) vs. non-AD (24%) [p=0.006]. Immune related hypothyroidism (AD: 30% vs. non-AD: 15%, p=0.08), nephritis (AD: 9% vs. non-AD 2%, p=0.08), and hyperthyroidism (AD: 9% vs. non-AD: 3%, p=0.19) had numerically higher rates in AD, but these findings did not achieve statistical significance. Flare of the baseline AD with IO was observed in 6/23 (26%) patients, and 50% of flares occurred within 1 month of starting IO (range 1-5 months). Management of irAEs was similar in AD vs. non-AD, and rates of full resolution of irAEs were similar (AD: 55%; non-AD: 61%, p=0.65). Late onset irAEs (> 3 months after IO discontinuation) were rare (AD: 12%; non-AD: 4%, p=0.21). Conclusions: A higher overall rate of irAEs was observed in patients with underlying autoimmune disease, including greater immune related transaminitis and flare of the underlying autoimmune disease, suggesting the need for close monitoring of irAEs in this population. [Table: see text]