Safety and efficacy of anti-programmed death 1 antibodies in patients with cancer and pre-existing autoimmune or inflammatory disease

Objective: Patients with autoimmune or inflammatory disease (AID) are susceptible to immune-related adverse events (irAEs) when treated with immune check-point inhibitors (ICIs). We decided to analyse the safety and effectiveness of anti-PD-1 antibodies in AID patients and look for an association between the presence of pre-existing AID and the clinical outcome. Methods: In a prospective study of the REISAMIC registry of grade >= 2 irAEs occurring in ICI-treated patients, we studied the associations between pre-existing AID on one hand and irAE-free survival, overall survival and best objective response rate on the other. Results: We identified 45 patients with 53 AIDs in REISAMIC. The cancer diagnoses included melanoma (n = 36), non-small-cell lung cancer (n = 6) and others (n = 3). The most frequent pre-existing AIDs were vitiligo (n=17), psoriasis (n=12), thyroiditis (n=7), Sjogren syndrome (n = 4) and rheumatoid arthritis (n=2). Twenty patients (44.4%) presented with at least one irAE: eleven of these were associated with a pre-existing AID ('AID flare'). Treatment with anti-PD-1 antibodies was maintained in 15 of the 20 patients with an irAE. The IrAE-free survival time was significantly shorter in AID patients (median: 5.4 months) than in AID-free patients (median: 13 months, p = 2.1 x 10(-4)). The AID and AID-free groups did not differ significantly with regard to the overall survival time and objective response rate (p = 0.38 and 0.098, respectively). Conclusion: In patients treated with anti-PD-1 antibody, pre-existing AID was associated with a significantly increased risk of irAEs. Our results indicate that cancer treatments with anti-PD-1 antibodies are just as effective in AID patients as they are in AID-free patients. (C) 2017 Elsevier Ltd. All rights reserved.