Abstract and Introduction
Objectives: Checkpoint inhibitor (CPI)–associated colitis can limit therapy and has resemblance to inflammatory bowel disease (IBD). Studies exploring mechanistic similarities between these colitides are limited, yet therapeutic targets for either disorder could emerge from shared pathophysiology.
Methods: The morphology and inflammatory content of colonic biopsy specimens from anti–CTLA-4 and anti–PD-1/PD-L1 antibody-treated patients with CPI colitis were compared with initial biopsy specimens from patients with IBD. Predictors of the need for infliximab were sought in CPI patients.
Results: Biopsy specimens from CPI patients showed significantly lower chronicity scores and similar activity scores compared with patients with IBD. Anti–CTLA-4 and IBD groups showed equivalent CD8, CD4, PD-1, and PD-L1 expression, while FoxP3 scores were lower and CD68 scores were higher in anti–CTLA-4 compared with IBD biopsy specimens. Anti–PD-1/PD-L1 group had lower scores for CD8, CD4, and PD-1 and equivalent scores for FoxP3, PD-L1, and CD68 compared with IBD. Anti–CTLA-4 biopsy specimens had higher scores for CD8, PD-1, PD-L1, and CD68 than anti–PD-1/PD-L1 biopsy specimens. CD8/FoxP3 ratios and CD68 scores were higher among CPI patients requiring infliximab therapy for colitis compared with those responding to steroids.
Conclusions: The proinflammatory immune phenotype of anti–CTLA-4–associated colitis has significant overlap with IBD. CD8/FoxP3 ratios may predict therapeutic response in CPI-associated colitis.
Immunotherapy has become a paradigm in cancer treatment that potentiates native host immune responses to tumor by blocking inhibitory immune checkpoints. Of available agents in the checkpoint inhibitor (CPI) class of drugs, anti–CTLA-4 (cytotoxic T lymphocyte antigen 4) and anti–PD-1/PD-L1 (programed cell death 1 receptor/programed cell death ligand 1) monoclonal antibodies are the current leaders for treating patients with locally advanced and metastatic cancers. CTLA-4, a protein expressed on activated cytotoxic T cells, blocks the costimulation required to amplify T-cell responses, thereby diminishing immune reactivity and facilitating the development of immunologic tolerance.[2,3] Anti–CTLA-4 antibodies remove this inhibition, leading to an increased amplitude of early stage T-cell activation.[1,4] PD-1 receptor (expressed on the surface of CD4+ and CD8+ T cells, natural killer cells, B cells, macrophages, and dendritic cells) and its ligands, PD-L1 and PD-L2 (expressed on immune cells and also on epithelial cells, endothelial cells, and fibroblasts), act more peripherally, inhibiting signaling pathways downstream of antigen-specific T-cell receptors to reduce the expression of cytokines and transcription factors associated with effector T-cell function in tissues.[5,6] Therefore, antibodies against PD-1 and its ligands diminish immune inhibition in the periphery, allowing for heightened antitumor activity.[1,4]
Since their introduction in 2003, the indications for CPI use have continued to broaden beyond the original indications of melanoma and non–small cell lung cancer, as clinical responses in the post- and now neoadjuvant setting are under evaluation in many cancer types in over 100 clinical trials of monotherapy and combination therapy.[8–11]
The development of clinically debilitating immune-related adverse events (irAEs) due to the generalized loss of T-cell inhibition set off by these drugs has become a frequent and sometimes treatment-limiting side effect of CPI therapy, with reported incidence rates ranging from 3% to 68% depending on the site of involvement. Commonly affected sites include the luminal gastrointestinal tract, lungs, skin, endocrine organs, and liver,[4,13–21] all of which are affected by an autoimmune-like inflammatory injury mediated by T-cell hyperactivation against self-antigens.
irAE-related colitis occurs in up to 20% of CPI-treated patients and can threaten the ability to continue therapy, often requiring steroids and anti–tumor necrosis factor α (TNF-α) therapy to treat and carrying a risk of perforation.[22,23] The histopathology of initial-onset CPI-associated colitis has been described in case reports and small series.[24–27] To our knowledge, only one histologic study has sought to compare CPI-associated colitis (anti–CTLA-4) with idiopathic inflammatory bowel disease (IBD), the autoimmune disease this irAE most closely mimics, and data comparing mechanistic and immune cellular content between these colitides are limited.[27,28] In this study, the morphology and immune phenotype of CPI-associated colitis are contrasted with that of IBD to examine whether and how closely these two conditions overlap. A secondary goal is to consider whether any mechanistic derivations can be implied from this comparison that may shed light onto either condition.