Management of infliximab refractory immune checkpoint inhibitor gastrointestinal toxicity: a multicenter case series
Catriona Harvey,Kazi J Nahar,Janet McKeown,Serigne N Lo,Sheima Farag,Nadia Yousaf,Kate Young,Liselotte Tas,Aafke Meerveld-Eggink,Christian Blank,Austin Thomas,Jennifer McQuade,Bastian Schilling,Douglas B Johnson,Roberto Martín Huertas,Ana Arance,Joanna Lee,Lisa Zimmer,Georgina V Long,Matteo S Carlino,Yinghong Wang,Alexander Maxwell Menzies
DOI: https://doi.org/10.1136/jitc-2023-008232
IF: 12.469
2024-01-01
Journal for ImmunoTherapy of Cancer
Abstract:Background Immune checkpoint inhibitor (ICI) gastrointestinal toxicity (gastritis, enteritis, colitis) is a major cause of morbidity and treatment-related death. Guidelines agree steroid-refractory cases warrant infliximab, however best management of infliximab-refractory ICI gastrointestinal toxicity (IRIGItox) is unknown. Methods We conducted an international multicenter retrospective case series. IRIGItox was defined as failure of symptom resolution ≤grade 1 (Common Terminology Criteria for Adverse Events V.5.0) following ≥2 infliximab doses or failure of symptom resolution ≤grade 2 after one dose. Data were extracted regarding demographics, steroid use, response to treatment, and survival outcomes. Toxicity was graded at symptom onset and time of infliximab failure. Efficacy of infliximab refractory therapy was assessed by symptom resolution, time to resolution and steroid wean duration. Survival outcomes were examined based on immunosuppressive therapy received. Results 78 patients were identified: median age 60 years; 56% men; majority melanoma (N=70, 90%); 60 (77%) received anti-cytotoxic T-lymphocyte-associated protein 4 alone or in combination with anti-programmed cell death protein-1 and most had colitis (N=74, 95%). 106 post-infliximab treatments were given: 31 calcineurin inhibitors (CNIs); 27 antimetabolites (mycophenolate, azathioprine); 16 non-systemic immunomodulatory agents (eg, mesalazine or budesonide); 15 vedolizumab; 5 other biologics (anti-interleukin-12/23, 16, Janus kinase inhibitors) and 7 interventional procedures (including colectomy); 5 did not receive post-infliximab therapy. Symptom resolution was achieved in most (N=23/31, 74%) patients treated with CNIs; 12/27 (44%) with antimetabolites; 7/16 (44%) with non-systemic immunomodulation, 8/15 (53%) with vedolizumab and 5/7 (71%) with interventional procedures. No non-vedolizumab biologics resulted in toxicity resolution. CNIs had the shortest time to symptom resolution (12 days) and steroid wean (43 days); however, were associated with poorer event-free survival (6.3 months) and overall survival (26.8 months) than other agents. Conversely, vedolizumab had the longest time to toxicity resolution and steroid wean, 66 and 124 days, but most favorable survival data: EFS 24.5 months; median OS not reached. Six death occurred (three due to IRIGItox or management of toxicity; three with persisting IRIGItox and progressive disease). Conclusions IRIGItox causes major morbidity and mortality. Management is heterogeneous. CNIs appear most likely to result in toxicity resolution in the shortest time period, however, are associated with poorer oncological outcomes in contrast to vedolizumab.
oncology,immunology