Evaluation of Readministration of Immune Checkpoint Inhibitors after Immune-Related Adverse Events in Patients with Cancer

Importance: Although immune checkpoint inhibitors (ICIs), such as anti-PD-1 (programmed cell death 1) or anti-PD-L1 (programmed cell death 1 ligand 1), have proved effective in treating many cancers, patients receiving ICIs may experience immune-related adverse events (irAEs). Little evidence exists on the safety of resuming these treatments after an irAE. Objective: To investigate the safety of a rechallenge with anti-PD-1 or anti-PD-L1 immunotherapies after an irAE. Design, Setting, and Participants: This cohort study of the safety of an ICI rechallenge involved consecutive adult patients (n = 93) who were referred to the ImmunoTOX assessment board at the Gustave Roussy cancer center in Villejuif, France, between August 1, 2015, and December 31, 2017. Data were analyzed from May 28 to November 25, 2018. Main Outcomes and Measures: Incidence of a second irAE in patients who had a readministration of an anti-PD-1 or anti-PD-L1 inhibitor after an initial grade 2 or higher irAE. Characteristics of the patients and the irAEs were reviewed, and the primary end point was the rate of occurrence of second irAEs. Results: A total of 93 patients were included, among whom 48 (52%) were female, and the median (range) age was 62.5 (33-85) years. The main cancer types or tumor sites were melanoma (31 [33%]), lung (15 [16%]), colorectal (8 [9%]), and lymphoma (8 [9%]). For the initial irAE, 43 grade 2 events (46%), 36 grade 3 events (39%), and 14 grade 4 events (15%) were found, presenting primarily as hepatitis (17 [18%]), skin toxic effect (14 [15%]), pneumonitis (13 [14%]), colitis (11 [12%]), or arthralgia (7 [7.5%]). Forty patients (43%) were rechallenged with the same anti-PD-1 or anti-PD-L1 agent. The rechallenged and non-rechallenged groups did not differ in terms of median (range) age (61 [34-84] years vs 63 [33-85] years; P =.37), time to initial irAE (5 [1-40] treatment cycles vs 3 [1-22] treatment cycles; P =.32), irAE severity (grade 2: 18 [47.5%] vs 27 [51%]; grades 3-4: 22 [52.5%] vs 26 [49%]; P =.70), or steroid use (17 [42.5%] vs 32 [60%]; P =.09). With a median follow-up period of 14 months, the same irAE or a different irAE occurred in 22 patients (55%). Shorter time to the initial irAE was linked to the occurrence of a second irAE (9 vs 15 weeks; P =.04). The second irAEs were not found to be more severe than the first. Conclusions and Relevance: The risk-reward ratio for an anti-PD-1 or anti-PD-L1 rechallenge appears to be acceptable, although these patients require close monitoring; further investigation into rechallenge conditions through a prospective clinical trial is needed..