Discussion

Immune-checkpoints inhibitors have revolutionized the treatment of metastatic malignancies; however, they can trigger various organ-specific irAEs, such as nausea and diarrhea, which can limit their use even with evidence of regression of the underlying malignancy. One third of patients treated with ipilimumab, an anti-CTLA-4 antibody, develop diarrhea and 16% of patients will go on to develop severe colitis, which can lead to perforations (0.5%) and/or colectomy [4, 11]. Nivolumab, an anti-PD-1 antibody, causes diarrhea in 8–19% of patients, of whom only 1% experience grade 3 or 4 diarrhea [5, 12, 13]. Patients treated with a combination of ipilimumab and nivolumab have a 44% chance of developing diarrhea, with grade 3 diarrhea accounting for 20% of all cases [12]. Typically, the onset of diarrhea occurs 6 weeks after initiation of treatment, but can be delayed up to 6 months after the last dose of ICI [13]. Patients can also experience other irAEs separately or concomitantly, such as thyroiditis, myositis, and hepatitis, which suggests a systemic auto-immune like reaction to ICIs.

While colitis is the most common cause of diarrhea in the ICI-treated patient, alternative etiologies of diarrhea must also be considered. There has been only one published case reporting ICI-induced celiac disease due to ipilimumab [14]. Our case report highlights a rare irAE, celiac disease, associated with ICI treatment. It is unclear whether the patient had previously undiagnosed celiac disease or whether ICIs triggered her enteritis, but the patient was asymptomatic prior to initiation of ICI. Given that she also exhibited other well-characterized concomitant irAEs such as pancreatitis and hypothyroidism, we suspect that her celiac disease was triggered by ICIs. The initiation of ipilimumab in particular seemed to trigger her symptoms, which is in concordance with the literature that has shown the strong immunogenic effects of ipilimumab compared to other ICIs. The patient’s celiac diagnosis was coincidental; given that her diarrhea was of a lower grade, she was treated symptomatically with loperamide. Upper endoscopy was mainly performed for her nausea and vomiting.

The pathogenesis of celiac disease is due to gluten-mediated activation of intestinal CD4+ T cells in the lamina propria. Gliadin peptides from gluten are converted by tissue transglutaminase (TTG) to a form that increases their affinity for HLA-DQ2 and HLA-DQ8 molecules and results in enhanced antigen presentation. Antigen presenting cells activate CD4+ T-helper cells in the lamina propria, causing expansion of cells that produce antibodies to gliadin and TTG [15]. The histologic hallmarks of celiac disease on small bowel biopsies are intraepithelial lymphocytosis, lymphoplasmacytic inflammation of the lamina propria, villous atrophy, and crypt hyperplasia [16].

While the pathogenesis of immune-mediated colitis is not well understood, CTLA-4 blockade leads to T-cell activation that increases secretion of CD4 T-helper cell cytokines and cytolytic CD8 T-cell tissue infiltration [17]. In contrast to celiac disease, ICI-induced colitis usually presents with an array of histologic findings. Usually biopsies demonstrate features of acute active colitis such as intraepithelial neutrophilic infiltrates or crypt abscesses, increased mononuclear cells in the lamina propria [18]. Both celiac disease and ICI colitis show increased apoptotic cells in crypts. Interestingly a subset of ICI colitis patients may demonstrate intraepithelial or basal lymphocytes, excess plasma cells in the lamina propria, and lymphocytic cryptitis on colonic biopsy, which is more consistent with findings in chronic colitis [19].

As histologic features may overlap, clinical features are crucial for differentiation of distinct diseases. This patient tested positive for tissue transglutaminase antibodies; the serum ELISA anti-TTG test has 93% sensitivity and 98% specificity for a celiac diagnosis [15]. Typically celiac disease is diagnosed by biopsy when there are greater than forty lymphocytes per hundred epithelial nuclei in the villi. Thus, although the patient’s histology showed less than thirty lymphocytes per hundred epithelial nuclei, a diagnosis of early celiac disease can be made in conjunction with a positive serology. Gastroenterologists must consider histologic and endoscopic features, clinical symptoms, and laboratory findings such as celiac serology and genetic testing to arrive at the correct diagnosis.

Our patient was able to continue treatment with ICIs with dietary modifications, suggesting correct diagnosis is critical for optimal patient outcome. As highlighted by this case, active inflammation can affect the small bowel and/or upper GI tract alone. For a patient with symptomatic diarrhea, evaluation shouldn’t be limited to colonoscopy alone, and biopsies should be done to look for microscopic evidence of inflammation even if the mucosa of the GI appears normal.

Both early recognition and initiation of the appropriate treatment of irAEs are crucial to relieve symptoms, avoid complications, and when indicated enable continued ICI therapy. While enterocolitis is by far the most common cause of diarrhea, we here report a case of celiac disease induced by ICI therapy. In a patient with symptomatic diarrhea following initiation of ICIs, infectious pathology should be ruled out followed by initiation of systemic corticosteroids. In patients with unusual features and/or failure to respond to steroid treatment, consideration should be given to a full endoscopic work up including a colonoscopy and exam of the terminal ileum and an upper endoscopy with biopsies in order to discern the underlying etiology. This case of ICI-induced celiac disease demonstrates that multidisciplinary collaboration among oncologists, gastroenterologists, and pathologists is crucial for correct diagnosis and treatment.