Managing IEC-associated enterocolitis following CAR-T therapy in multiple myeloma

Abstract

We appreciate the thoughtful letter by Somay and colleagues regarding potential etiologies of enterocolitis following immune effector cell (IEC) therapy in multiple myeloma (MM) [ 1 ]. As we observed in our original manuscript, IEC-associated enterocolitis typically manifests with non-bloody diarrhea in the weeks to months following chimeric antigen receptor T-cell (CAR-T) therapy in MM [ 2 ]. Endoscopic biopsy findings generally mimic those of graft-versus-host disease following allogeneic transplantation, including crypt dropout and increased apoptosis. In our original case series from the United States (US) MM Immunotherapy Consortium, we had identified 14 cases of IEC-associated enterocolitis following ciltacabtagene autoleucel (cilta-cel, n = 13) or idecabtagene vicleucel (ide-cel, n = 1). Another multicenter group has subsequently identified an additional 12 cases of IEC-associated enterocolitis following cilta-cel, all with similar clinical presentations and biopsy findings [ 3 ]. We again highlight the importance of ruling out infectious and neoplastic etiologies of diarrhea before making a diagnosis of IEC-associated enterocolitis. For example, colitis caused by cytomegalovirus infection can occur even in the absence of frank viremia. With regard to neoplastic etiologies, we had intentionally excluded 3 patients from our original series because they clearly appeared to have been driven by a lymphoproliferative process based on validated T-cell receptor (TCR) clonality assays; one such case has since been described separately [ 4 ]. In their letter, Somay and colleagues provide an excellent workflow for the evaluation of lymphoproliferative processes in this setting [ 1 ]. They suggest the routine use of anti-camelid antibodies (appropriate for cilta-cel but not ide-cel) for CAR detection and also recommend TCR clonality testing. We wholeheartedly agree with these steps as best practices for diagnostic workup. Additionally, as demonstrated in a recently published report where enteric T-cell lymphoma following cilta-cel was likely caused by insertional mutagenesis [ 5 ], integration-site analyses can be helpful to elucidate underlying causes if CAR T cells are identified. We acknowledge that such testing is not commercially available at the vast majority of cancer centers, even in academic settings. This includes in our original case series, where TCR clonality testing and integration-site analyses could not be done in most cases due to the lack of Clinical Laboratory Improvement Amendments (CLIA)-grade assays or research protocols to allow tissue transfer between centers. In such cases, we recommend close collaboration with CAR-T product manufacturers to investigate potential etiologies using the methods described above. As we noted in our initial report, reactive and neoplastic T-cell processes likely exist along a spectrum in this setting. Somay and colleagues suggest that all cases of IEC-associated enterocolitis may involve some degree of neoplastic T-cell growth [ 1 ]. This assertion merits further dedicated investigation but is unverifiable at the present time. Even in the absence of a pathologic lymphoproliferative process, post-infusion CAR T-cell populations have been demonstrated to possess oligoclonal properties with selection of particular clones at various timepoints [ 6 ]. In our case series, 42% of patients ( n = 6) had resolution or improvement in their symptoms with corticosteroids and biological agents. Anecdotally, we have had additional patients diagnosed with IEC-associated enterocolitis since our initial publication whose symptoms have resolved with brief courses of corticosteroid therapy alone. These facts again suggest a pathological spectrum in the presentation of IEC-associated enterocolitis, whereby oligoclonal T-cell clones may represent a true malignant entity in some cases (as highlighted by Somay and colleagues) [ 1 ] but not in other cases. When shifting from pathologic diagnoses to clinical decision-making, we would also recommend caution with the suggestion that a lymphotoxic agent (in particular, cyclosporine) should be used empirically for frontline therapy in clinical practice. Cyclosporine has little activity in T-cell lymphoma, although it may be effective for certain indolent T-cell lymphoproliferative disorders of the gastrointestinal tract [ 7 , 8 ]. Cyclosporine is associated with significant toxicities (e.g., infections and nephrotoxicity) as well as its own risk of second primary neoplasms. More importantly to this setting, cyclosporine inhibits T cells much more potently than biologic agents such as infliximab or vedolizumab. As such, systemic cyclosporine may interfere with CAR-T efficacy. In refractory cases where patients continue to have persistent post-CAR-T symptoms despite corticosteroids and biological agents, it is reasonable to consider frank T-cell suppression with drugs like cyclosporine or cyclophosphamide [ 4 , 9 ]. Conversely, other small molecules such as ruxolitinib or dasatinib may be considered given their inhibitory effects on CAR T cells in certain settings [ 10 , 11 ]. Because IEC-associated enterocolitis clearly spans a wide spectrum of presentations, we believe that a personalized and stepwise approach to management is essential. Ultimately, given that infliximab or vedolizumab may lead to symptom resolution in many cases without any known detrimental impact on T-cell expansion [ 2 ], we generally favor the use of these biologic agents first if a brief course of corticosteroids is ineffective. A staged treatment approach (summarized in Fig. 1 , including suggested dosing details for these agents) may provide the best balance of safety and efficacy for managing IEC-associated enterocolitis. The term IEC-associated enterocolitis remains the best descriptor of this syndrome for the time being, particularly given that some cases improve with drugs used to manage autoimmune conditions (or immune-mediated colitis following checkpoint inhibitor therapy in solid oncology) [ 12 , 13 , 14 ] rather than requiring lymphoma treatment regimens. Further translational investigations to better understand this disease process through the lens of clonality, using methods similar to those eloquently described by Somay and colleagues [ 1 ], are active ongoing areas of research for the US MM Immunotherapy Consortium as well. Fig. 1: Suggested approach for IEC-associated enterocolitis. Dosing considerations: for biologic agents, we suggest infliximab 5 mg/kg or vedolizumab 300 mg in line with SITC guidelines [ 12 ]. The optimal choice between the two biological agents is unclear in immune-mediated colitis following checkpoint inhibitor therapy [ 14 ]. However, vedolizumab may be more helpful for cases with predominantly upper GI symptoms or in infliximab-refractory cases [ 13 ]. For cyclosporine, we suggest an initial total daily dose of 5 mg/kg with subsequent pharmacokinetic monitoring [ 4 ]. For cyclophosphamide, we suggest a single dose of 2000 mg/m 2 as done for the treatment of parkinsonism in this setting [ 9 ]. CAR-T chimeric antigen receptor T-cell, CsA cyclosporine A, GI gastroenterology/gastrointestinal, FDA United States Food and Drug Administration, IEC immune effector cell, mg milligrams, mg/kg milligrams per kilogram, mg/m 2 milligrams per square meter of body surface area, SITC Society for Immunotherapy of Cancer. Full size image

Topics & Concepts