ICI-Related Esophagitis in Cancer Remains Rare, Physicians Say

Retrospective data show a low prevalence of esophagitis among immune checkpoint inhibitor-related adverse event in patient with cancer.

The rate of immune-checkpoint-inhibitor (ICI) related esophagitis in patients with gastrointestinal symptoms undergoing esophagogastroduodenoscopy (EGD) remains low and rarely leads to treatment discontinuation, according to a recent study published in the Journal of The National Comprehensive Cancer Network.

To treat these rare cases, proton pump inhibitors (PPIs), sucralfate, histamine 2 receptor (H2) blockers, and steroids are all used in clinical practice, but the results of the study have led the investigator to recommend PPIs over the other therapies in certain patients.

“Because the prevalence of ICI-related esophagitis is low, we recommend initial empirical therapy with PPIs in the absence of alarming symptoms. If symptoms persist, endoscopy could be pursued along with additional therapies, such as steroids. Although evidence supporting this proposed management of ICI-related esophagitis is lacking, it is routinely used for esophagitis of other etiologies with reasonable success,” wrote the study authors.

ICI’s have helped to improve survival across multiple cancer types and are associated with a nonspecific immune response. This nonspecific response can result in toxic effects known as immune-related adverse events (irAEs). These irAEs commonly affect organs and organ systems such as the gastrointestinal tract, skin, liver, lungs, and the endocrine system. Gastrointestinal irAEs tend to present as lower gastrointestinal (GI) manifestations including colitis and diarrhea. These GI AEs tend to be more common with CTLA-4 blockade than with PD-1 and PD-L1 blockade. The average rate of grade 3/4 diarrhea and colitis in patients receiving CTLA-4 inhibitors was 18% and 5% respectively.

T retrospective analysis looked at 657 patients receiving ICI-based therapy and underwent EGD at the University of Texas MD Anderson Cancer Center to determine the rate of esophagitis in this patient population. Data was collected between June 2011 and January 2020. Patients with preexisting gastroesophageal reflux disease (GERD), eosinophilic esophagitis, esophageal cancer, or know esophageal infection were excluded.

Of the 657 included in the analysis, 21 had confirmed esophagitis. Of the 636 that were excluded, 509 had no evidence of esophageal abnormality, 26 had radiation damage, 27 had esophageal cancer, 23 had esophagogastric surgery, 24 had a history of GERD, and 25 had a positive infectious workup at the time of diagnosis.

The median age of included patients was 64, 62% were male and 62% were white. Comorbidities included hypertension (29%), hyperlipidemia (29%), types 2 diabetes (24%), hypothyroidism (24%), chronic kidney disease (14%), ischemic heart disease (10%), and COPD (10%). Current cancer types included non-small cell lung cancer (29%), gastrointestinal (19%), melanoma (19%), genitourinary (14%), head and neck (10%), hematologic (5%), and other (5%). Ninety-five percent of patients had stage IV disease and 71% received anti-PD-1/PD-L1 therapy. Five percent received anti-CTLA-4 therapy and 24% received a combination of the 2.

Nineteen percent of patients were given ICI with chemotherapy known to cause esophagitis and 33% were given ICIs concurrently with any chemotherapy. Other irAEs include colitis (24%), skin (19%), joint (10%), endocrine system (5%), hematologic system (5%), and liver (5%).

The median duration of ICI-based therapy was 105 days and the median number of infusions at the time of esophagitis onset was 3. The median time to onset was 126 days and the median duration of symptoms was 30 days. Common clinical esophagitis presentations included nausea/vomiting (67%), dysphagia (29%), hematemesis (195), melena (14%), dyspepsia (10%), and reflux (5%). The majority of patients, 81%, had a peak grade of 1-2.

Nearly half, 48%, of patients were hospitalized for esophagitis with a median duration of stay of 7 days. Forty-three percent of patients stopped ICI use due to gastrointestinal toxicity and 38% of patients died. Ten percent had recurrence of esophagitis symptoms.

“The median time from initiation of ICI use to onset of esophagitis in our study was approximately 4 months, which is similar to the time of onset described by other authors. This is a longer time to onset than is generally observed for other irAEs. For example, lower gastrointestinal irAEs such as colitis tend to have an onset of 6 to 8 weeks, in contrast with upper gastrointestinal irAEs, which tend to have an onset of 4 to 9 months,” wrote the study authors. “However, retrospective analyses have shown that colitis associated with PD-1/PD-L1 blockade has a later onset than does colitis associated with CTLA-4 blockade. Delayed onset of upper gastrointestinal toxic effects may be explained by their association with anti–PD-1/PD-L1 therapy. Confirming this requires further evaluation.”


Panneerselvan K, Amin R, Wei D, et al. Clinicopathologic features, treatment response, and outcomes of immune checkpoint inhibitor–related esophagitis. J Natl Compr Canc Netw. 2021;19 (8):896-904. doi: https://doi.org/10.6004/jnccn.2020.7675