Ischemic Gastroduodenitis: Navigating the Challenges of Enteral Nutrition Resumption

ABSTRACT

Ischemic gastritis (IG) and ischemic duodenitis (ID) are rare due to the extensive anastomotic vascular supply arising from the celiac trunk. Profound splanchnic hypoperfusion is required to induce a clinically significant ischemic insult. Etiologies typically include systemic hypotension, vasculitis, or thromboembolism. Patients with IG/ID often face high morbidity and mortality given the severity of ischemia required to induce extensive mucosal damage, leaving clinicians to balance prolonged bowel rest and early enteral feeding. We present the case of an 18-year-old man with spastic quadriplegic cerebral palsy and developmental delay who developed severe IG/ID in the setting of severe hemorrhagic shock. He presented to our institution with hemodynamic instability, profound anemia, and signs of upper gastrointestinal bleeding. Endoscopic evaluation found the patient to have severe IG/ID necessitating prolonged bowel rest and total parenteral nutrition (TPN). After evidence of mucosal recovery on surveillance endoscopy, he was gradually transitioned from TPN to enteral nutrition and successfully discharged to a long-term care facility. This report describes the successful reintroduction of enteral feeding in a patient with severe IG/ID after prolonged bowel rest. The use of surveillance endoscopy to guide the gradual reintroduction of enteral feeding, alongside careful TPN weaning, allowed for the successful discharge of a patient with IG/ID, a condition associated with poor outcomes.

INTRODUCTION

Ischemic gastritis (IG) and ischemic duodenitis (ID) are rare given the robust anastomotic vascular supply arising from the celiac artery trunk to the upper gastrointestinal tract, with ischemic injury typically arising only during profound splanchnic hypoperfusion. Systemic hypotension, vasculitis, and disseminated thromboembolism can result in such severe splanchnic hypoperfusion that gastric and duodenal ischemia may develop. It has been documented that during states of shock, total gastric perfusion decreases by approximately 90%, with portions of the mucosa receiving no perfusion. Prolonged ischemia results in mucosal breakdown, ulceration, bleeding and, in severe cases, transmural necrosis.¹

Given the extent of mucosal injury and the severity of the underlying etiologies, patients with IG/ID have a poor prognosis, with a 6-month mortality ranging from 24% to 29%.^2,3 Clinicians managing critically ill patients with IG/ID often face challenges in determining the appropriate timing for reintroducing enteral feeding, so as to avoid exacerbating mucosal damage through overdistension of the stomach and increased acid secretion. Total parenteral nutrition (TPN) is associated with inherent risks including infection, cholestatic liver injury, thrombosis, and serious metabolic derangements,⁴ while fasting is associated with nutritional compromise and thus impaired healing. We present a case of severe IG/ID in which surveillance endoscopy was used to guide the optimal timing for resumption of enteral nutrition, offering a proposed framework for clinicians managing this challenging condition.

CASE REPORT

An 18-year-old man with a medical history of spastic quadriplegic cerebral palsy, recurrent pneumonia, unspecified seizure disorder with long-term dependency on gastrostomy tube and tracheostomy presented via ambulance after his family noted increasing lethargy compared with baseline, associated with increasing abdominal distension. Of note, 2 days before admission, the patient was discharged from an outside facility for a humeral fracture that was treated nonoperatively using pain management with a prescription for oral oxycodone. Unfortunately, records from the outside hospital could not be obtained, thus limiting the clinical context.

On presentation, the patient immediately received naloxone for suspected opioid overdose given the recent addition of oxycodone to his home medication regimen. Physical examination demonstrated a pinpoint left pupil with an initial blood pressure of 56/29 mm Hg and pulse oximetry at 70%. Initial laboratory test results indicated a profound anemia with a hemoglobin of 6.1 g/dL. The patient received 4 L of the lactated Ringer solution, 2 units of packed red blood cells and was empirically started on intravenous (IV) vancomycin and piperacillin-tazobactam. Additional laboratory test results demonstrated a platelet count of 44 K/mm³ and an international normalized ratio of 3.39.

Abdominal and pelvic contrast-enhanced computed tomography (CT) scan (Figure 1) obtained in the emergency department demonstrated a massively distended stomach with retained gastric contents, fecalization within the upper gastrointestinal (GI) tract, pneumobilia, and right lower lung lobe infiltrates. The pylorus was noted to be widely patent, although there appeared to be a reduction in caliber of the fourth portion of the duodenum.

Figure 1.

Transverse sections from the abdominal and pelvic intravenous contrast-enhanced computed tomography obtained in the emergency department. (1) Distended stomach with retained gastric contents. (2) Fecalization of the duodenum. (3) Intrahepatic pneumobilia. (4) Patent pylorus. (5) Decreased caliber of the terminal duodenum. (6) Gastrostomy tube bumper nonadherent with gastric mucosa.

The patient was admitted to the intensive care unit for further management and started on vasopressor support with norepinephrine and vasopressin. Aspiration of the gastrostomy tube revealed maroon-colored gastric contents. The patient did not experience any hematemesis, coffee-ground emesis, hematochezia, or melena. At this time, the patient was started on IV pantoprazole 40 mg twice daily, and the Gastroenterology service was consulted given concerns for an upper GI bleed. Physical examination revealed a distended abdomen, hypoactive bowel sounds and the absence of a succussion splash. The patient's gastrostomy tube was freely movable, with the site noted to be clean, dry, and intact.

The patient underwent an esophagogastroduodenoscopy (EGD) on hospital day 1 (HD 1), which revealed diffuse ulceration and friable mucosa involving approximately 90% of the gastric mucosa with small areas of intervening normal mucosa within the cardia and antrum (Figure 2). No active hemorrhage was seen, although dark heme was noted throughout the stomach. There was no evidence of gastric outlet obstruction. The duodenum demonstrated diffuse ulcerations with friable mucosa in the duodenal bulb and second portion of the duodenum. There was no evidence of duodenal obstruction. Histopathology from gastric mucosal biopsies demonstrated necroinflammatory debris with admixed luminal contents as well as acute erosive gastritis on a background of reactive gastropathy in the antrum and corpus. The patient was continued on IV pantoprazole 40 mg twice daily and was made strictly nil per os to allow for bowel rest.

Figure 2.

Images from the index esophagogastroduodenoscopy on hospital day 1. (A) Normal mucosa surrounded by denuding mucosal lining. (B) Extensive heme deposition on gastric mucosa. (C) Percutaneous endoscopic gastrostomy tube bumper without evidence of buried bumper syndrome or focal ulceration. (D) Mucosal ulceration and heme deposition in duodenum.

Despite hemodynamic stabilization and successful vasopressor weaning by HD 2, enteral feeding was deferred given the severity of mucosal injury observed on the index endoscopy. TPN was started on HD 7 after the placement of a peripherally inserted central catheter (which underwent an unforeseen 2-day delay) to allow for continued bowel rest. Trophic feeding was attempted on HD 8, resulting in the patient developing tachycardia and lactic acidosis, necessitating readmission to the intensive care unit given concerns for acute mesenteric ischemia. On HD 9, the patient was transferred back to the intermediate care unit on TPN after resolution of the aforementioned abnormalities and an unremarkable abdominal CT angiography.

Repeat EGD was performed on HD 14 (Figure 3). During the follow-up examination, several superficial ulcers (Forrest III) > 2 cm in size were noted in the gastric fundus, body and antrum, with a single large ulcer consistent with Forrest IIc. There was no active bleeding noted within the stomach. The duodenum revealed multiple clean-based (Forrest III) ulcers within the duodenal bulb, with evidence of scarring that decreased in severity more distally in the duodenum. Overall, the findings were much improved from the EGD performed on HD 1, while remaining on pantoprazole 40 mg IV twice daily. Trophic feeding via the gastrostomy tube was reinitiated on HD 14, and the volume of trophic feeds was increased gradually over the next 4 days. On HD 18, the patient was able to attain more than 60% of his nutritional requirements with enteral feeding and was weaned from TPN. The patient's laboratory studies did not reveal any evidence of refeeding syndrome after the reintroduction of enteral nutrition.

Figure 3.

Images from esophagogastroduodenoscopy on hospital day 14. (A) Superficial ulcers in gastric fundus. (B) Unremarkable mucosa within the duodenum.

The patient remained in the hospital for a total of 35 days, with the latter part of his hospital course focused on the management of a pulmonary infection due to Stenotrophomonas maltophilia and Pseudomonas aeruginosa. He was discharged to a long-term care facility to complete a seven-day course of tigecycline and cefepime.

DISCUSSION

The development of IG and ID is rare as the tissue involved receives rich collateral blood supply via branches of the celiac trunk. Consequently, clinically meaningful injury generally occurs only in the setting of severe splanchnic hypoperfusion. Given the severity of shock required to induce IG/ID, patients who develop the condition have poor outcomes, with 6-month mortality ranging between 24% and 29% and 1-year mortality of up to 41%.^3,5 Though the data are limited, 90-day mortality for patients requiring TPN in the setting of IG has been found to be as high as 50%.² It is within this context that one of the principal challenges with respect to the management of IG/ID remains determining the optimal timing for the resumption of enteral nutrition.

Prolonged bowel rest, while necessary to prevent exacerbation of mucosal injury through overdistension and increased acid secretion, carries its own significant risks. TPN is associated with infectious complications, cholestatic liver injury, thrombosis, and metabolic derangements,⁴ while continued fasting impairs mucosal healing and overall recovery. This tension is particularly pronounced in patients with severe or extensive mucosal injury, in whom the window for safe enteral refeeding is difficult to define in the absence of established guidelines.

EGD remains the preferred modality by which to evaluate the extent of ischemic damage to the stomach and duodenum.^1,3 Furthermore, endoscopy allows for surveillance of healing, with direct visualization of mucosal surfaces during subsequent endoscopic procedures. We propose that surveillance endoscopy should serve as the primary tool for guiding the decision to reintroduce enteral nutrition, with timing individualized according to the severity of injury observed at the index endoscopy, the patient's hemodynamic trajectory, and response to ongoing medical therapy. In patients with superficial erosions or shallow ulcers, granulation tissue can be expected within approximately 72 hours,⁶ suggesting that early repeat endoscopy at around 3 days may be appropriate. Conversely, patients with deep or extensive ulcerations, as in the present case, may require a longer interval of approximately 10–14 days before sufficient mucosal recovery can be assessed.⁷ IV proton pump inhibitor therapy and successful vasopressor weaning should be considered prerequisites before surveillance endoscopy is pursued. We propose that Forrest III ulcers with granulation tissue and no active bleeding, in the context of hemodynamic stability, may provide a reasonable basis for reintroducing low-volume trophic enteral nutrition, with gradual escalation according to clinical tolerance.

In this case, the severity of shock on admission strongly implicated splanchnic hypoperfusion as the underlying etiology of IG/ID, with CT angiography confirming the absence of vascular abnormalities. Despite hemodynamic stabilization and vasopressor weaning by HD 2, the extent of mucosal injury on index endoscopy precluded early enteral feeding, and an attempt at trophic feeds on HD 8 was complicated by hemodynamic instability. Surveillance endoscopy on HD 14 demonstrated meaningful mucosal recovery, with Forrest III ulcers and no active bleeding, providing the clinical basis for the cautious reintroduction of trophic feeds. Enteral nutrition was gradually advanced over the following 4 days, with TPN successfully weaned by HD 18 and no evidence of refeeding syndrome.

This case contributes to the limited literature on IG/ID by describing the timeline, challenges, and methodology by which enteral feeding was successfully restarted, offering practical guidance for clinicians who encounter this condition. The use of surveillance endoscopy to individualize the timing of enteral refeeding, rather than relying on clinical parameters alone, represents a reproducible and physiologically grounded approach. More data are needed before evidence-based guidelines relating to optimal timing of enteral nutrition in IG/ID can be established.

We present the case of a young patient with a complex medical history who presented with signs of upper GI bleeding and shock. He was found to have severe IG/ID, requiring strict bowel rest with TPN before the gradual reintroduction of enteral nutrition and eventual discharge. This case highlights the importance of surveillance endoscopy in cases of IG/ID, as well as the necessity of a cautious approach to managing nutrition to support an uncomplicated recovery.

DISCLOSURES

Author contributions: M. Stasiewicz: contributed to study conception and design, interpreted the data, drafted the manuscript, approved the final version and agrees to be accountable for all aspects of the work. Y. Naga: contributed to data acquisition, critically revised the manuscript, and approved the final version. GV Ohning: contributed to data acquisition, critically revised the manuscript, informed management recommendations in the manuscript and approved the final version. J. Aponte-Pieras: contributed to study conception and design, critically revised the manuscript, informed management recommendations, and approved the final version. M. Stasiewicz is the article guarantor.

Financial disclosure: None to report.

Previous presentation: Presented as a posted at the American College of Gastroenterology Annual Scientific Meeting, October 27, 2025; Phoenix, AZ.

Informed consent was obtained for this case report from next of kin.

ABBREVIATIONS:

CT

computed tomography

EGD

esophagogastroduodenoscopy

GI

gastrointestinal

HD

hospital day

ID

ischemic duodenitis

IG

ischemic gastritis

IV

intravenous

TPN

total parenteral nutrition