Article Text
Abstract
As immune checkpoint inhibitors (ICIs) are increasingly used, clinicians are more frequently encountering the side effects of these therapies. ICIs have been implicated in numerous adverse effects against healthy tissues. We present a case of a patient who developed treatment refractory checkpoint inhibitor colitis. Following colonoscopy, it was discovered that this patient had cytomegalovirus (CMV) coinfection. This case report highlights the importance of undertaking an appropriate assessment, including endoscopic and histologic investigation, of patients with presumed ICI colitis. Accurately diagnosing a superimposed CMV colitis changes clinical management and can improve patient outcomes.
- gastroenterology
- oncology
- hepatitis and other GI infections
Statistics from Altmetric.com
Background
Immune checkpoint inhibitors (ICIs) such as ipilimumab, nivolumab and pembrolizumab are monoclonal antibodies that target pathways of T-cell activation and function.1 By removing inhibition of T-cell function, these monoclonal antibodies promote T-cell-mediated suppression of tumours.1 ICIs are now approved for use in more than 14 different malignancies.1 While these drugs have a therapeutic benefit, they may also upregulate T-cell activity against healthy tissues resulting in adverse effects. ICIs have been implicated in colitis, hepatitis, pneumonitis, myocarditis, dermatitis, endocrine dysfunction and neurotoxicity.1
Case presentation
A 66-year-old man with melanoma of the scalp, type 2 diabetes, tobacco abuse and checkpoint inhibitor colitis treated with prednisone 40 mg two times per day presented for colonoscopy to evaluate refractory diarrhoea. The patient was diagnosed with primary cutaneous melanoma of the scalp (T3aN2aM0, AJCC8 stage IIIB) in August 2017. Wide local excision in September 2017 showed negative margins. Sentinel lymph node biopsy from the left posterior neck revealed a 3 mm focus of metastatic melanoma. Lymph node dissection of the left posterior neck was performed in November 2017 and sequencing for BRAF mutation was negative. In late November 2017, the patient was started on adjuvant nivolumab. No cytomegalovirus (CMV) diagnostic tests were done prior to the initiation of nivolumab. In April 2018, the patient developed non-bloody, watery diarrhoea. In the first week, he had 9–10 bowel movements a day and lost 10 pounds. He tested negative for enteric bacterial infections and Clostridium difficile colitis and was treated for presumed checkpoint inhibitor colitis with prednisone 10 mg two times per day for 5 days with improvement. As his steroids were tapered, he developed worsening diarrhoea. Prednisone was increased to 40 mg two times per day and his symptoms resolved. After 2 weeks of high-dose steroids, he was again tapered with the addition of budesonide. As his steroids were tapered, he developed breakthrough diarrhoea. In March 2019, he underwent colonoscopy for evaluation. Colonoscopy revealed mild erythema and localised punctate ulcerations (figure 1). Biopsies were obtained and revealed an ‘owl’s eye’ inclusion body, diagnostic of CMV colitis (figure 2). His CMV DNA PCR was positive at 631 IU/mL. The patient was referred to infectious disease and was treated with valganciclovir 900 mg two times per day for 1 month. Following 1 month of treatment, his CMV DNA PCR was undetectable. The patient had complete resolution of his gastrointestinal symptoms and was successfully weaned off steroids.
Outcome and follow-up
Following treatment with valganciclovir, the patient’s diarrhoea resolved. His CMV PCR following treatment was undetectable. At 2-month follow-up, he reported that his bowel movements had returned to normal consistency and frequency.
Discussion
Colitis is a well-known side effect of ICIs and occurs in 25% of patients treated with the cytotoxic T-lymphocyte associated protein 4 (CTLA-4) blocking antibodies (ipilimumab/Yervoy) and 5% of patients treated with the programmed cell death protein 1 (PD-1) antibodies (nivolumab/Opodivo and pembrolizumab/Keytruda).1 These effects are dose dependent. A recent chart review of 139 case reports between January 2016 and April 2018 found that 14 out of 139 patients treated with ICIs developed colitis.2
The clinical symptoms of ICI colitis include watery bowel movements, abdominal cramping, flatulence and occasional blood/mucus in the stool.3 In rare cases, ICI colitis can lead to colonic perforation and death.3 The onset of colitis typically occurs 6–7 weeks after starting ipilimumab4 and 6–18 weeks after starting nivolumab or pembrolizumab.5 It is best to recognise ICI colitis early as acute inflammation can lead to chronic inflammation within 1 month of the onset of symptoms.3 ICI colitis should be in the differential of any patient presenting with symptoms of colitis with recent ICI exposure. The first step in management is to rule out infectious causes of diarrhoea such as enteric bacteria and C. difficile.3 Once infection is ruled out, patients with ICI colitis can be treated with corticosteroids.3 Guideline-based recommendations suggest 0.5–1 mg/kg prednisone for mild diarrhoea, and 1–2 mg/kg prednisone for severe diarrhoea, with tapers lasting over at least 4–6 weeks.6 If symptoms do not improve after 3–5 days or recur after treatment, treatment with anti-tumor necrosis factor (TNF) agents and other biologics can be considered.7 Patients with severe colitis refractory to steroids and biologics may require colectomy.
This report highlights treatment refractory ICI colitis in a patient with CMV colitis infection. In this case, it is unclear if the patient’s clinical symptoms are due to ICI colitis (which did initially respond to high-dose steroids, although lower than recommended doses for immune-mediated colitis) or superimposed CMV colitis (which did respond to valganciclovir). Concomitant CMV colitis has been seen in numerous prior reports of patients with inflammatory bowel disease (IBD).8 The observation of CMV coinfection in both IBD and ICI colitis suggests parallels between these two disease entities. Research has shown patients with acute severe colitis and CMV coinfection are less responsive to treatment with corticosteroids than non-infected patients.8 One proposed mechanism for corticosteroid resistance is that T-cell inhibition by immunosuppression may result in reactivation of CMV.8 The review of the literature reveals two prior case reports of CMV colitis detected in a patient with therapy refractory ICI colitis.9 10 A retrospective cohort study performed in Germany found that CMV infection in patients with treatment refractory ICI colitis may be common.11 Over a 10-year period, 370 patients with metastatic melanoma were treated with ICIs at the University Hospital of Essen. Forty one of the 370 patients treated with ICIs developed ICI colitis. Of the 41 patients with colitis, 5 were refractory to traditional immunosuppressive therapy (no improvement within 2 weeks of immunosuppressive therapy). All five patients with treatment refractory ICI colitis underwent colonic biopsy with four out of five biopsies positive for CMV. This study demonstrates CMV reactivation or infection may be a clinically relevant contributor to treatment refractory colitis in patients with presumed ICI colitis.
Our case report highlights the importance of a detailed diagnostic evaluation for patients with presumed ICI colitis. In cases of presumed ICI colitis, an appropriate assessment, including endoscopic and histologic investigation, is needed to exclude other differential diagnoses. In this case, endoscopic evaluation revealed a superimposed CMV colitis. Following treatment with valganciclovir, the patient’s gastrointestinal symptoms resolved.
Learning points
Immune checkpoint inhibitor (ICI) colitis and cytomegalovirus (CMV) colitis share many common symptoms, making the diagnosis difficult to distinguish clinically.
In patients with presumed ICI colitis, it is important to undertake an appropriate assessment, including endoscopic and histologic investigation, to determine the correct diagnosis.
Accurately diagnosing a superimposed CMV colitis changes clinical management and can improve patient outcomes.
Footnotes
Contributors KBH: primary author of the manuscript. PF: edited the manuscript. BBB: edited the manuscript and is the article guarantor.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent for publication Obtained.
Provenance and peer review Not commissioned; externally peer reviewed.