Immunotherapy-induced coeliac disease in curative lung cancer

Abstract

The advent of immunotherapy has revolutionised the treatment of metastatic lung cancer and it has recently been established as the standard of care in the radical treatment of lung cancer. However, immune-related adverse events (IrAEs) frequently occur in patients treated with immunotherapy, and rare IrAEs continue to be identified. We report a case of immunotherapy-induced coeliac disease due to adjuvant durvalumab post-chemoradiotherapy in a patient receiving curative treatment for lung cancer. The patient had raised anti-tissue transglutaminase IgA and histological findings consistent with coeliac disease. This is the first published case report of probable immunotherapy-induced coeliac disease both with the immunotherapy drug durvalumab and in the curative lung cancer setting.

Background

The PACIFIC study led to the approval of the immunotherapy drug durvalumab in the adjuvant setting for patients with unresectable, stage III non-small cell lung cancer (NSCLC) with no disease progression after two or more cycles of platinum-based chemoradiotherapy,¹ and whose tumours express programmed death-ligand 1 (PD-L1) in >1% of cells.² PD-L1 is an immune checkpoint protein which has been hypothesised to aid neoplastic lesions to evade the host’s immune system. Durvalumab is a human monoclonal IgG1 antibody which binds to PD-L1 on tumour cells, preventing its interaction with PD-1 on cytotoxic T-cells. This blocks the inhibitory PD-1–PD-L1 interaction, leading to increased cytotoxic T-cell activity, and cytokine production,¹ with the aim of causing tumour necrosis. However, activation of T-cells can lead to immune-related adverse events (IrAEs) which require prompt recognition and treatment.

Case presentation

A 68-year-old Caucasian woman with stage III NSCLC received adjuvant durvalumab after completion of concurrent chemoradiotherapy. Her PD-L1 status was 90%–100%. After the fourth cycle of durvalumab, she presented with a 1 week history of diarrhoea with 3–5 bowel motions per day (Common Terminology Criteria for Adverse Event (CTCAE) grade 2 diarrhoea). She was admitted to hospital and started on prednisolone 50 mg (1 mg/kg). After 3 days of steroids, she had a flexible sigmoidoscopy which appeared macroscopically normal. Biopsies showed a non-specific increase in intraepithelial lymphocytes. The diarrhoea resolved, and she was discharged on a 6-week weaning regime of prednisolone for presumed immunotherapy-induced colitis. Durvalumab therapy was paused.

Two months later, the patient re-presented with a 5-day history of diarrhoea with >10 bowel motions per day, including nocturnal diarrhoea (CTCAE grade 3). She had not restarted durvalumab in this time, but on re-admission the prednisolone was restarted. Her abdomen was soft and non-tender. Abdominal X-ray was normal. A full diarrhoea screen was performed, including urea and electrolytes, thyroid function tests, coeliac disease screen (anti-tissue transglutaminase (anti-tTG) IgA), stool cultures, 7-α-cholestenone, serology for blood-borne viruses, cytomegalovirus and Epstein-Barr virus. Positive findings were an elevated serum anti-tTG IgA level (10.2 unit/mL, reference range 0.1–5 unit/mL). She was also deficient in both folic acid and vitamin D. She had not had a coeliac screen performed prior to starting durvalumab and she did not have any stored serological samples which could be used to test whether the anti-tTG IgA was raised prior to starting the drug.

An oesophagogastroduodenoscopy was performed. Macroscopically, there was scalloping of the duodenum (figure 1). Biopsies taken from the first and second part of the duodenum showed small bowel mucosa with marked villous blunting, chronic inflammation of the lamina propria and increased intraepithelial lymphocytes (figure 2), consistent with Marsh 3 criteria,³ confirming the diagnosis of coeliac disease. She was homozygous for human leukocyte antigen (HLA) DQ2 on HLA typing.

Figure 1.

Endoscopic imaging of D2 showing scalloping of the duodenum.

Figure 2.

Histological image of D2 showing villous blunting, increased intraepithelial lymphocytes and chronic inflammation of the lamina propria (magnification 10×). These appearances are consistent with the diagnosis of coeliac enteropathy.

Treatment

The patient was started on a lifelong gluten-free diet while in hospital. As she had been started on a weaning course of oral prednisolone (50 mg daily, reduced by 10 mg every 7 days) with the intention of improving her symptoms rapidly, this was continued until the course was complete.

Outcome and follow-up

Her diarrhoea immediately improved on the steroids and gluten-free diet. A decision was made to restart immunotherapy once the steroids were complete. She did not report recurrence of her symptoms following an additional seven cycles of durvalumab therapy while on a gluten-free diet. The anti-tTG IgA normalised after 3 months.

Discussion

Coeliac disease is an immune-mediated disease where a loss of immune tolerance to ingested gluten peptides leads to small intestinal inflammation, tissue damage and the formation of antibodies to deamidated gluten peptides.⁴ Although approximately 1% of the population have a positive serological test for coeliac disease, the majority of these people are minimally symptomatic and are undiagnosed. It is hypothesised that coeliac disease develops in people with a genetic predisposition (95% of people with coeliac disease carry the HLA-DQ2 and 5% the HLA-DQ8 haplotype) who are exposed to an additional trigger, for example, a viral illness.⁵ The diagnosis is most reliably made by positive serology (raised anti-tTG IgA) and small intestinal histological appearances. These changes may result in small bowel malabsorption and consequent weight loss, fatigue, diarrhoea, osteopenia and anaemia due to iron, vitamin B12 and folate deficiencies. In most patients it is possible to control the symptoms and reduce the risks of malabsorption by completely removing gluten from the patient’s diet.

Immune checkpoint inhibitors (ICIs) are novel antibodies designed to bind to specific immune checkpoint proteins on T-cells, or on malignant cells. Binding of checkpoint proteins to their natural ligand normally results in downregulation of the immune response. When ICIs bind to checkpoint proteins they prevent interaction with the respective ligand, thereby removing the down-regulatory signal. This results in T-cell activation with subsequent elimination of neoplastic cells in the context of cancer immunotherapy. The following two immune checkpoint targets in the T-cell regulation pathway have been identified: PD-1/PD-L1 and CTLA-4. Durvalumab binds to PD-L1 to prevent it from binding to and activating PD-1.

Gastrointestinal toxicity is one of the most frequent IrAEs.⁶ In the durvalumab arm of the PACIFIC trial (n=476), 18.3% of patients had diarrhoea and 0.6% had severe (grades 3–4) diarrhoea. All cases were diagnosed as immunotherapy-induced colitis;¹ however, it was not clear from the trial if all these patients had endoscopic investigations to confirm the diagnosis. The detection of rare toxicities requires a large number of patients to be treated with immunotherapy, therefore rarer toxicities are often identified post licensing, when the number of patients receiving these drugs increases exponentially. There are three published cases of patients receiving immunotherapy developing coeliac disease after the commencement of the drug^7–9 but the exact incidence of this IrAE is presently unknown. These cases occurred in patients prescribed either PD-1 inhibitors or CTLA-4 inhibitors, and symptoms started as early as cycle 1, and as late as cycle 7.

This case report highlights a rare IrAE of coeliac disease presenting during treatment with durvalumab. Durvalumab may have unmasked previously undiagnosed coeliac disease and as serological samples from before the durvalumab was started were not available, it is not possible to exclude this. However, there was no history suggestive of malabsorption and serological samples over the previous 6 years had not shown any history of iron-deficiency anaemia. Alternatively, the patient may have developed coeliac disease de novo, as a direct consequence of immunotherapy treatment. The exact underlying mechanism for the development of coeliac disease secondary to immunotherapy is unknown. A mouse model of the pathogenesis of coeliac disease demonstrates the interaction of several codependent factors. Upregulation of interleukin (IL)-15 in the lamina propria of the small intestinal epithelium induces the adaptive immune response to dietary gluten. HLA-DQ8 enhances the production of interferon-gamma required to cause tissue damage. However, the presence of IL-15 from the epithethial compartment is required before there is sufficient activation and expansion of cytotoxic T-cells to lead to CD8⁺ T-cell-dependent killing of epithelial cells and villous atrophy.¹⁰ As described above, ICIs block the actions of regulatory T-cell populations, enhancing the inflammatory cascade. Although this is a highly desirable therapeutic goal in treating malignancies, it is likely to allow unchecked intestinal inflammation throughout the gastrointestinal tract. Patients with untreated coeliac disease have both mucosal damage and high serum concentrations of CTLA-4.¹¹ Gentile et al⁹ hypothesise that CTLA-4 may have an immunomodulatory effect on T cells, enabling drugs, such as ipilimumab which blocks CTLA-4, to possibly upregulate the T-cell response. Durvalumab selectively blocks the interaction of PD-L1 with PD-1 and does not play a role in the CTLA-4 pathway; however, the mechanism of activation of inflammation may be similar, increasing the likelihood of patients taking durvalumab developing autoimmune diseases such as coeliac disease.

Coeliac disease is one of many causes of villous atrophy or enteropathy—other causes include infections such as Giardiasis and tuberculosis, autoimmune conditions and intestinal lymphoma. A retrospective analysis by Fazal et al¹² examined the cases of 40 patients who presented with diarrhoea following treatment with ICIs which act on the PD-1–PD-L1 pathway. 17.5% of patients had macroscopic evidence of inflammation in the duodenum and 71% of those had biopsy-proven villous atrophy. 86% of patients had evidence of chronic inflammation or increased intraepithelial lymphocytes on histology. Although this is not specific for any particular cause of inflammation, it does show that following treatment with ICIs, inflammatory changes in the upper gastrointestinal tract are common. All patients in their series also had a colonoscopy and 65% of patients also had inflammation present in the colon. Fazal et al did not comment on anti-tTG IgA serology, but it seems that immune checkpoint inhibition can lead to an enteropathy in its own right.

Immunotherapy-induced colitis usually occurs after the first few cycles of treatment and is most effectively managed when immunosuppressive treatment is initiated early. If gastrointestinal IrAEs are unrecognised or treatment is delayed, the consequences can be significant, including death.¹³ Guidelines produced by the American Society of Clinical Oncology¹⁴ and guidelines endorsed by the British Society of Gastroenterology¹⁵ give advice on the management of new diarrhoea in patients on immunotherapy.

This is the first reported case of coeliac disease arising following immunotherapy treatment in both the adjuvant lung cancer setting, and with durvalumab. A new diagnosis of coeliac disease will have long-term consequences for this patient who may have been cured with chemoradiotherapy alone. However, the majority of patients with stage III NSCLC have disease progression despite chemoradiotherapy with curative intent and therefore, as shown in the PACIFIC trial, the addition of immunotherapy in improving survival is significant. This case highlights immunotherapy-induced coeliac disease and other enteropathies as potential causes of severe diarrhoea. Although diarrhoea due to coeliac disease in the context of immunotherapy is an exceptional IrAE, symptoms can resolve quickly once a diagnosis is made and a gluten-free diet commenced. We would recommend a wide differential diagnosis is considered when patients present with symptoms of classic IrAEs, but with investigation findings that are inconsistent with immunotherapy-induced colitis. If no evidence of colonic inflammation is demonstrated, an oesophagogastroduodenoscopy is recommended to examine for enteropathy.

Learning points.

• Patients prescribed immunotherapy who develop diarrhoea of Common Terminology Criteria for Adverse Event grade 2 or greater should have a full diarrhoeal screen performed, including coeliac serology.

• Steroid therapy may mask the underlying cause of immunotherapy-induced diarrhoea.

• It can be safe to restart immunotherapy in patients with immunotherapy-induced coeliac disease after appropriate management of coeliac disease (lifelong gluten-free diet).

Ethics statements

Patient consent for publication

Obtained.