Op0270 targeted Therapies for Severe Immune-Related Adverse Events of Immune Checkpoint Inhibitors Are Not Associated With a Worse Prognosis

Abstract

BackgroundIn about 10% of patients, the immune response to immune checkpoint inhibitors (ICI) exceeds the anti-tumor response and leads to autoimmune complications (immune-related Adverse Events, irAEs), which can sometimes be severe and require the use of targeted therapies. Two studies recently reported a deleterious impact of targeted therapies on the survival of patients with irAEs.ObjectivesTo compare overall survival in patients treated or not with a biological or targeted synthetic DMARD (b/tsDMARD) for an hospitalized irAE.MethodsAll adults included in the French national health database who initiated an ICI between 2014 and 2019, for any type of cancer, were retrospectively analyzed. Follow-up was analyzed between 2013 to 2020, to have a one-year look-back and one year look-ahead period. The occurrence of an irAE during ICI treatment or in the 12 months after its last administration was defined by the combination of (1) hospitalization for a cause evoking an irAE of any nature, except endocrinopathy, and (2) either the discontinuation of the ICI, or the initiation of corticosteroids, or of a conventional DMARD or a b/tsDMARD, or a new recognition as a long-term disabling condition (LTD) for an autoimmune or inflammatory disease. Patients with irAE treated with b/tsDMARD were matched (1:3) at the time they initiated b/tsDMARD to those who did not initiate b/tsDMARD, using a dynamic propensity score calculated every 30 days. The propensity score included gender, age, type of cancer, type of ICI, time from cancer to initiation of ICI and from initiation of ICI to irAE, use of corticosteroids, hospitalization in intensive care unit, type of irAE, number of LTDs, Charlson’s index and FDep social deprivation index. Overall survival was compared between the two groups using a Cox model. A sensibility analysis restricted to gastrointestinal and rheumatic irAEs was performed.Results71,723 patients (men: 66.0%, median age: 66 years) initiating an ICI were analyzed. An hospitalized irAE occurred in 7883 patients (11.0%). irAE occurred at a median time of 72 days after ICI initiation, and the 4 more frequent were gastrointestinal (inflammatory colitis, 4.7%), cardiologic (2.0%), rheumatic (1.8%), and pulmonary (1.7%). Median patient follow-up after irAE was 356 days. Mortality after irAE was 4.8% at 1 month and 11.3% at 3 months.After matching, 325 patients were treated for an irAE with a b/tsMDARD (Infliximab (57.7%), Rituximab (25.5%), Tocilizumab (4.2%), Vedolizumab (3.3%), others (9.3%)), including 257 and 40 for a gastrointestinal or rheumatic irAE, respectively. They were compared to 975 patients who were not treated with a b/tsDMARD. The median time from ICI initiation to irAE was well balanced by matching (93 versus 92 days, respectively). ICI was discontinued in 25.4% of b/tsDMARD treated patients patients within 3 months after the irAE, compared to 20.9% in the group not treated with a b/tsDMARD.Mortality after irAEs was 6.8% and 7.6% at 3 months, respectively, and 16.0% and 17.9% at 6 months, respectively. Overall survival whatever the type of irAE did not significantly differ in patients treated with or without a b/tsDMARD (HR=0.87, CI=[0.74, 1.03]), or in patients treated for a rheumatic irAE (HR=0.80, CI=[0.51, 1.25]). Overall survival was significantly improved in gastrointestinal irAE in patients treated with a b/tsDMARD compared to patients not treated with a b/tsDMARD (HR=0.80, CI=[0.66, 0.97]).ConclusionIn one of the largest study to date in terms of number of hospitalized irAEs and number of b/tsDMARD-treated patients, targeted therapies were not associated with a worse prognosis and significantly improved overall survival in patients with induced colitis, the most frequent and one of the most severe irAEs.Figure 1.Overall survival in days in patients treated with a b/tsDMARD (blue) or not (red)AcknowledgementsThis analysis was planned in the setting of the PRAISE study, which received an unrestricted grant from BMS. BMS had no access to data, results and interpretation of the present analysis.Disclosure of InterestsNathanaël Sedmak: None declared, Eden SEBBAG: None declared, Pierre Tran Ba Loc: None declared, Thibaut Fabacher: None declared, Thibault Bahougne: None declared, Cécile Arnold: None declared, Guillaume Desjeux: None declared, Hervé Servy: None declared, Nicolas Meyer: None declared, Erik Sauleau: None declared, Raphaèle Seror: None declared, Jacques-Eric Gottenberg Consultant of: Abbvie, BMS, Gilead, Galapagos, Lilly, Janssen, Roche Chugai, Pfizer, Sanofi, UCB, Grant/research support from: Abbvie, BMS, Pfizer.