P012 Patients with cancer and preexisting autoimmune and inflammatory diseases treated by anti-programmed death 1 (PD-1) antibodies at gustave roussy cancer center

Abstract

Introduction Immune checkpoint inhibitors anti PD-1 are monoclonal antibodies used in cancers.1 PD-1 ligand/PD-1 pathway antagonism between cancer cells and antitumoral cytotoxic CD8 T cells increase antitumoral immunity but is involved in autoimmune diseases (AID). Objectives Evaluate tolerance and efficacy of anti PD-1 in AID patients. Methods Patients had been included in REISAMIC registry (Registry of Severe Adverse Events of Immunomodulating Monoclonal Antibodies in Oncology) between June 1st, 2014, and December 31st, 2017. Patients were treated with anti-PD-1. Exclusion criteria were malignant hematologic disease, second advanced cancer or chronic viral infection. AID subtypes were defined for sensitivity analyses: diseases of clinical interest (DCI), prognostic interest (DPI) and vitiligo. One patient with pre-existing AID were matched on age, sex and cancer type with 3 patients without preexisting AID (controls). Analyses were adjusted for OMS status, corticosteroid, cerebral metastasis, LDH, albuminemia and neutrophil/lymphocyte ratio >3. Endpoints were grade ≥2 irAE free survival and overall survival (OS). Objective response rates (ORR) were described. Results 641 patients were included in REISAMIC. Among them, 69 patients were excluded. 572 patients had been included: 63 with AID and 509 controls. Among AID patients, we observed 38% DCI, 38% DPI and 24% vitiligo. DCI were Sjögren syndrome (n=4), rheumatoid arthritis (n=4), polymyalgia rheumatica/giant cell arteritis (n=2) and others. DPI were psoriasis (n=11), thyroiditis (n=8) and others. We matched 55 AID patients with 165 controls. Cancer type, TNM or AJCC grade, age and sex were similar between AID patients and controls. Overall survival (HR 0,67, IC95% [0,42-1,08], p=0,098) and irAE free survival (HR 1,27, IC95% [0,84-1,90], p=0,25) were not different between AID patients and controls. After vitiligo exclusion in sensitivity analyses, irAE free survival was shorter for AID patients than controls (HR 1,69, IC05% [1- 2,87], p=0,05) and more for DCI patients only (HR 2,66, IC95% [1,01-7], p=0,049). ORR were 35,5% and 28,5% for AID patients and controls respectively. Conclusions Preexisting AID should not be a contraindication for anti PD-1. We suspect that mechanisms involved in AID are important in irAE mechanism and antitumoral response. CD4 T cells and Th17 cells should be study in these situations.