Review Article

Drug-induced Small Bowel Injury

Shadi Hamdeh; Dejan Micic; Stephen Hanauer

Aliment Pharmacol Ther. 2021;54(44512):1370-1388.

Abstract and Introduction

Abstract

Background: Drug-induced gastrointestinal injury has been increasingly reported, but its exact incidence is not known. The small and large intestines represent the most affected sites of injury, accounting for 20%-40% of all gastrointestinal side effects.

Aim: To provide an updated literature review detailing medications linked to the development of small bowel injury.

Methods: We conducted a literature search on PubMed from its inception to May 1, 2021. We included English-language original studies, meta-analyses, systematic reviews, review articles and case reports.

Results: Drug-induced enteropathy can range from asymptomatic histological changes resulting in a subtle, self-limited disease to a chronic inflammatory condition mimicking inflammatory bowel disease, or bowel perforation. Endoscopy can demonstrate erythema, mucosal friability, oedema, erosions, ulcers or strictures in severe cases. Histology may include mucosal erosions and ulcerations, focal active enteritis, villous atrophy, epithelial apoptosis or necrotising enteritis. A well-established association has been found with the use of nonsteroidal anti-inflammatory drugs, immunosuppressants, chemotherapeutic agents, antibiotics, immunotherapies, etanercept and olmesartan. Possible associations have been reported with other biologic agents, medications used for glycemic control, antihypertensives, cholinesterase inhibitors, potassium and iron supplements, with conflicting data regarding contraceptives/hormonal therapy and isotretinoin.

Conclusion: Physicians should be aware of the manifestations of drug-induced enteropathy as early recognition can lead to prompt discontinuation of the offending therapy and, therefore, a reduced risk of future complications.

Introduction

Drug-induced gastrointestinal (GI) injury has been increasingly encountered with the ever-growing armamentarium of pharmaceutical agents and availability of over-the-counter drugs. Each year in the United States, 4.02 billion prescriptions are dispensed.^[1] According to the Food and Drug Administration (FDA) annual report from 1995, all medications are associated with adverse events with 10% of these adverse events related to the GI tract, making the GI tract one of the most affected targets for adverse events.^[2] Based on a survey of hospital in-patients, this number can be even higher in the in-patient setting.^[3] Drug-induced GI injury can affect any part of the GI tract including the mouth,^[3] esophagus,^[3,4] stomach,^[3,5] small intestine^[6,7] and colon.^[8] The small and large intestine are of the most common sites for drug-related adverse events, accounting for 20%-40% of all GI side effects,^[9] however, the exact incidence of drug induced small bowel (SB) injury is not fully known. Drug-induced SB injury can be focal and limited to the SB or extensive with simultaneous injury to the colon.

Drug-induced colitis has been recently reviewed by Hamdeh et al^[8] and is beyond the scope of this review. In our review, we aim to provide an updated literature review detailing medications linked to the development of SB injury. We will describe the pathophysiology of drug-induced enteropathy (DIE), and current approaches for diagnosis and treatment. In general, medication discontinuation leads to clinical and endoscopic improvement but in more severe cases, immunosuppressive therapies or surgery may be warranted.

1 2 3 4 5 6 7

Next Section

Table 1. Medications associated with drug-induced small bowel injury
Drug class	Subclass	Summary of evidence
NSAIDs	Non-selective NSAIDs	Strong association is well-established
	Aspirin	Strong evidence even with low dose ASA
	Enteric coated ASA	Association with SB injury is well-established
	COX-2	Association with SB is well-established, but potentially decreased compared to non-selective NSAIDs
Biologics	Anti-TNF	Evidence is well-established based on several retrospective and prospective studies. Risk is the highest with ETN use
Biologics	IL-17 inhibitors	Well-established association based on evidence from clinical trials (including prospective and RCTs) and real world (case reports)
Immunosuppressive agents	Mycophenolate mofetil (MMF)	Well-established association based on several studies. Switching to EC-MMF to reduce the risk of SB toxicity demonstrated conflicting results
	Calcineurin inhibitors	Well-established association. Single case report of tacrolimus-induced intestinal angioedema
	Chemotherapeutic agents	Strong evidence linking several chemotherapeutic agents to SB injury including 5-FU, irinotecan and docetaxel. Linalidomide use was associated with bile salt malabsorption. Other cytotoxic agents including cisplatin, cytosine arabinoside, doxorubicin, vincristine, mercaptopurine and dasatinib linked to diarrhoea
	Checkpoint inhibitors (CPI)	Outside of an established colitis, growing strong evidence shows an association between CPIs and enteritis mimicking celiac disease on endoscopy. The risk is the highest with ipilimumab
	VEGF inhibitors	GI perforation is well-documented with the use of this class based on several studies and meta-analyses
	EGFR inhibitors	Evidence based on several case reports and one meta-analysis of 18 clinical trials
	HER-2 inhibitors	Association with diarrhoea based on evidence from a large clinical trial
	Tyrosine Kinase inhibitors (TKI)	Association with SB perforation was reported in several cases in the literature. Also diarrhoea is common based on several studies with a recent meta-analysis of 43 studies
	Rituximab	Increased risk of bowel perforation based on several studies and pooled analysis. This led to a warning about this potential adverse event
	Phosphoinositide-3 kinase (PI-3K) inhibitors	Increased risk of SB injury was shown in a retrospective analysis
Blood pressure medications	Olmesartan	Olmesartan associated enteropathy (OAE) is a well-established AE associated with olmesartan use based on large cohort studies beside numerous case reports. This association was also seen with other ARBs
Blood pressure medications	Calcium channel blockers (CCB)	Risk of bowel obstruction associated with nifedipine use based on a large population based study, but overall risk is low
Antibiotics		Very strong association between antibiotic use and diarrhoea, also risk of de novo IBD with prior antibiotic use
Diabetes medications	Acarbose	Diarrhoea is common with acarbose use. Few case reports of pneumatosis intestinalis
	Dipeptidyl peptidase-4 (DPP-4) inhibitors	Association with IBD based on large population-based study
	Metformin	Strong association with diarrhoea based on several studies and real-world data
Hormonal therapy	Oral contraceptives (OCP)	Increased risk of CD development and relapse
Hormonal therapy	Hormonal therapy post menopausal	Conflicting results compared to OCP use with potential for a protective effect
Isotretinoin		Possible association was initially suggested after reviewing the FEARs. Discredited in subsequent studies
Supplements	Potassium supplements	Well-reported association in the literature. The risk is decreased with the use of the slow released tablets
Supplements	Iron supplements	A strong evidence based on several published studies including a meta-analysis of 43 studies
Cholinesterase inhibitors (ChEI)		38 cases of peptic ulcers with 30 reported case of GI bleeding with donepezil use. However, a population based study failed to show an association
Miscellaneous	Nicorandil	Sporadic case reports of SB ulcers
	Alendronate	Four case reports of de-novo or exacerbation of celiac disease
	Clozapine	One case report of necrotising enterocolitis
	Gold compounds	Enterocolitis was reported in a literature review of 29 cases
	Clofazimine	Sporadic cases

Abbreviations: AE, adverse events; ASA, aspirin; COX, cyclooxygenase; EGFR, epidermal growth factor receptor; ETN, etanerecept; HER-2, human epidermal growth factor receptor 2; IL, interleukin; NSAIDS, non-steroidal anti-inflammatory drug; OCPs, oral contraceptives; RCT, randomised controlled trial; SB, small bowel, TNF, tumour necrosis factor; VEGF, vascular endothelial growth factor.

Table 2. Key differences between small bowel Crohn's disease and NSAID-induced enteropathy
	NSAID-induced enteropathy	Crohn's disease
Clinical presentation	Symptoms range from asymptomatic to abdominal pain, diarrhoea, nausea, vomiting, meleana, heamatochezia or obscure GI bleed secondary to ulcerations. Complications including obstruction (diaphragm disease) or rarely perforation. Not associated with penetrating disease	Symptoms range from asymptomatic to abdominal pain, nausea, vomiting, changes in bowel habits based on the location of disease, GI bleeding, weight loss, fever and anorexia. Complications including obstruction and fistula formation
Extra-intestinal manifestations	No clear association with extraintestinal manifestations; however NSAIDs are commonly used for pre-existing joint pain	Aphthous stomatitis, uveitis, scleritis, episcleritis, peripheral and axial arthropathy, skin rash including erythema nodosum and pyoderma gangrenosum
Endoscopy	Mucosal breaks (erosions or ulcers), denuded mucosa, reddened folds, petechiae or red spots. In the setting of frank ulcerations, red blood can be seen. With chronic use, a short segment stricture (diaphragm disease) is often seen in the distal small bowel and is pathognomonic for NSAID enteropathy	Focal erythema, edema, erosions, ulcers in a discontinuous pattern. Transmural inflammation associated with strictures and fistulae can be seen endoscopically. The most common location is terminal ileum and proximal colon with less common findings in proximal small bowel or gastro-duodenal
Histology	Mucosal erosions and ulcerations, focal active enteritis, villous atrophy mimicking celiac disease, epithelial apoptosis or necrotising enteritis/enterocolitis. Findings are limited to the mucosa and submucosa	Both mucosal and transmural accumulation of inflammatory cells including lymphocytes, plasma cells and neutrophils with active disease. Cryptitis, crypt abscess formation and crypt architectural distortion are also commonly seen suggestive of chronicity and granulomas are identified in 20%-30%
Radiology	Isolated ulcers, mucosal thickening and hyperenhancement. Strictures can be isolated or multiple. Diaphragm disease can be missed on cross-sectional imaging and small bowel follow-up due to the short nature of the stricture	Focal ulcers, mucosal thickening and hyperenhancement in short or long segments. Stricture or fistualae are common
Treatment	NSAID discontinuation Chronic changes may require endoscopic interventions (dilation) or surgery	Steroids, immunomodulators, biologic agents and surgery
Prognosis	Very good if NSAIDs are discontinued early and permanently discontinued	High risk of disease progression based on underlying risk factors including age, gender and extent of disease, cigarette smoking, and transmural complications at diagnosis