Jan 8, 2020
Towards improved characterization of immune-related adverse events in the setting of pre-existing autoimmune disease.
This JCO Podcast provides observations and commentary on the JCO article “Immune Checkpoint Inhibitor Therapy in Patients with Preexisting Inflammatory Bowel Disease”, by Abu-Sbeih et al. My name is Katy Tsai, and I am Assistant Professor of Medicine in the Division of Hematology/Oncology at the University of California, San Francisco. My oncologic specialty is the treatment of advanced melanoma and non-melanoma skin cancers.
Immune checkpoint inhibitors, referred to as ICIs in this podcast, have transformed the landscape of treatment options in oncology. While ICIs were first approved for the treatment of advanced melanoma in 2011, since that time, ICIs have shown activity in a variety of other histologies. Anti-PD-1 or anti-PD-L1, with or without anti-CTLA-4, are now approved for the treatment of lung cancer, head and neck squamous cell carcinoma, renal cell carcinoma, and many others. While ICIs can result in durable responses, their continued use can be limited by the development of immune-related adverse eventsimmune-related adverse events. Because these events are believed to be autoimmune in nature, there are intuitive safety concerns about ICI use in patients with known autoimmune disease. Can patients with pre-existing autoimmune disease be safely treated with ICI? Are these patients more likely to experience immune-related adverse events related to their autoimmune disease? Is this risk increased if their autoimmune disease is severe, with a history of required immunosuppression? These are all important questions faced by healthcare providers, questions which are not well studied due to the exclusion of patients with known autoimmune disease from pivotal ICI clinical trials.
To address these questions, Abu-Sbeih and colleagues chose to focus on ICI use in patients with pre-existing inflammatory bowel disease. This is a clinically relevant population of interest, given the relatively high incidence of immune-related diarrhea and colitis with ICI use; these immune-related adverse events occur in almost half of patients receiving combination anti-CTLA-4 and anti-PD-1 or PD-L1, about one-third of patients receiving anti-CTLA-4, and less frequently in patients receiving anti-PD-1 or PD-L1 alone. Abu-Sbeih and colleagues conducted a retrospective, multicenter study in which 102 patients – half with Crohn’s disease, half with ulcerative colitis – received ICI between 2010 and 2019. 17 patients received anti-CTLA-4, and 85 received anti-PD-1 or anti-PD-L1. This is a notable cohort of patients, as previous meta-analyses have reported on the safety of ICI use in much smaller numbers of patients with inflammatory bowel disease, and with less detailed clinical characterization of their inflammatory bowel disease. Univariate and multivariate logistic regression were used to assess the risk of gastrointestinal, or GI, adverse events, and was compared to a control population of 11,377 ICI-treated patients without inflammatory bowel disease, from the same participating institutions.
An important observation made by the authors was that the rate of GI immune-related adverse events in the inflammatory bowel disease cohort was significantly higher at 41%, compared to 11% in the non-inflammatory bowel disease cohort. Univariate analysis identified anti-CTLA-4 (given as monotherapy or in combination) as a risk factor for GI immune-related adverse events compared to anti-PD-1/L1 therapy but showed only a tendency toward significance in multivariate analysis, as did inflammatory bowel disease involving the colon. Although none of the collected clinical variables reached significance in multivariate analysis, the authors’ analysis of outcomes in the inflammatory bowel disease cohort is illuminating. Of the 41 inflammatory bowel disease patients who developed diarrhea, 51% had peak grade 3 or 4 diarrhea, 76% received glucocorticoids, and 29% required additional immunosuppression with infliximab or vedolizumab. 4% developed colonic perforation, with half of those patients requiring surgical intervention. Also of note, patients who were identified as having active inflammatory bowel disease within 3 months of ICI start had higher grade diarrhea compared to patients with inactive inflammatory bowel disease. Endoscopy data were also available for 48 inflammatory bowel disease patients. Interestingly, of the 41 patients who were noted to have normal or mild inflammatory findings, 18 (43%) developed any GI immune-related adverse events, 5 (12%) of which were grades 3-4. In the 7 patients with moderate/severe inflammatory findings, 5 (71%) had GI immune-related adverse events of any grade, 2 (29%) of which were grades 3-4.
Despite the higher rate of GI immune-related adverse events and associated complications in the inflammatory bowel disease cohort compared to the non-inflammatory bowel disease cohort, there were no fatalities. Additionally, 48% of patients in the inflammatory bowel disease cohort were identified as having complete response, partial response, or stable disease to ICI therapy, a clinical benefit rate similar to those reported in ICI clinical trials. It seems, then, that the benefits of ICI therapy in this population may well outweigh the risks, particularly for patients with inflammatory bowel disease who may have no viable alternative therapy available for their malignancy.
Overall, despite the inherent limitations of retrospective analysis, this work represents the largest study to date investigating the risk of GI immune-related adverse events in patients with cancer and comorbid inflammatory bowel disease who were treated with ICIs. It provides evidence for increased incidence and severity of GI immune-related adverse events, and complications thereof, in a well-annotated cohort of patients with inflammatory bowel disease. As these patients continued to receive clinical benefit from their ICI therapy, these findings can help better inform pre-ICI treatment counseling in patients with preexisting inflammatory bowel disease, and better prepare them for expected risks of treatment. At the same time, this work also raises a number of questions for further investigation. Is anti-CTLA-4 truly safe to use in this population, or was this limited by the small number of patients receiving anti-CTLA-4 in this study? Should ICI therapy be delayed to allow for optimal treatment of active inflammatory bowel disease, to decrease the risk of severe GI immune-related adverse events? What additional guidance can be given to providers regarding when to initiate additional immunosuppressive therapy, and should this be driven by endoscopic findings? These questions and more can and should be investigated in future larger-scale prospective studies. For the time being, based on the data presented, I would certainly be more cautious about giving a more aggressive immunotherapy regimen – that is, combination anti-CTLA-4 and anti-PD-1/L-1 – in a patient with highly symptomatic inflammatory bowel disease. Also, while numbers in this study were small, it seems that endoscopic evaluation prior to ICI start could be helpful in risk stratifying for severe GI IRAE. While inflammatory findings on endoscopy may not necessarily be an absolute contraindication to use of ICI, it seems these results would certainly be helpful in shared decision-making with the patient regarding their risk of developing GI IRAE and weighing other potential treatment options. Finally, and most importantly, this data demonstrates that some patients with inflammatory bowel disease who receive ICI for their malignancy can derive clinical benefit with manageable toxicities, suggesting that the mere presence of comorbid inflammatory bowel disease should not be a blanket exclusion criterion for ICI clinical trials.
This concludes this JCO Podcast. Thank you for listening.