Case 4 -

Non-Steroidal Anti-Inflammatory Drug-Related Injury Laura W. Lamps, M.D. Audrey J. Lazenby, M.D. Joel K. Greenson, M.D.

Clinical Presentation: This 54 year old Jewish physician underwent screening colonoscopy for polyps. Case 4 - Figure 1 - Low magnification of ileal biopsy shows superficial mucosal ulceration, mild villous blunting, and prominent lymphoid aggregates. Prominent lymphoid aggregates are normally present within ileal mucosa, but may confound the histologic evaluation of inflammatory processes. Case 4 - Figure 2 - Higher magnification shows villous edema and mixed inflammatory infiltrate in lamina propria, features that may be seen in the context of NSAID-related injury. Endoscopic Findings: He was found to have ulcers in the terminal ileum. Diagnosis : Non-steroidal anti-inflammatory drug-related injury General Comments: It has been well recognized that NSAIDs cause "peptic" ulcers in the duodenum, albeit at a lower rate than in the stomach. These ulcers are usually bland with "nonspecific" histology. This duodenal injury has been well recognized because of the ease of examination of the duodenum with the upper endoscope. The fate of the distal small bowel in patients taking NSAIDs has been less well researched because of the relative inaccessibility of the jejunum and ileum to endoscopic examination (except for the distal terminal ileum.) However, in the past decade or two, studies focusing on the small bowel have shown that NSAID damage to the distal small bowel is not uncommon. A recent endoscopic study of 1,900 consecutive colonoscopies in which the terminal ileum was entered showed that 2% of the patients had lesions of the terminal ileum. These lesions ranged from erythema to erosions to ulcers. On follow up, 85% of the patients with these small bowel lesions were found to be taking NSAIDs up to one week before the endoscopic examination. Most of these patients had been endoscoped for a personal or family history of colorectal dysplasia and were asymptomatic. These patients had no history of symptoms of any primary small bowel disease and none developed small bowel disease on follow up for up to 60 months. The most common NSAID used in these patients was enteric coated aspirin, which was taken at low dose (235 mg/d) for health maintenance reasons. When a greater length of the small bowel is studied, even more evidence of damage is seen in the distal small bowel of NSAID users. One careful study by Allison, et al, from the UK, looked for evidence of NSAID damage to the upper GI tract at post mortem. A single investigator examined the small bowel and stomach of all autopsies over a 22 month period. A drug history of NSAID usage was studied for each patient by reviewing hospital records as well as personal physician records. Of the 713 patients included in the study, 249 had some type of NSAID prescribed during the six months before death. Nonspecific ulcers were found in the jejunum or ileum of 21 (8.4%) of the NSAID group, while ulcers were seen in only 3 patients (0.6%) in the control group. Looking at just the long-term users of NSAIDS, the prevalence of ulcers was up to 13.5%. The pathology of the lesions in the NSAID usage group ranged from single to multiple tiny punched out ulcers to confluent areas of deep ulceration and stricture formation. With even more sensitive detection methods, subclinical damage to the distal small bowel can be seen in a majority of patients taking NSAIDs even at a low dose. Bjarnson and colleagues have used radiolabeled RBCs and leukocytes followed by scintigrams and have found 60% to 70% of patients taking NSAIDs to have accumulations of RBCs and leukocytes in the distal small bowel, indicating increased small bowel inflammation. These same investigators have also found increased fecal excretion of labeled RBCs and leukocytes as well as increased intestinal permeability in patients taking NSAIDs. The pathology of the small bowel in most of the patients in the Bjarnson studies has not been investigated. However, in one of the Bjarnson studies, 18 patients with the highest fecal excretion of radionuclide had barium studies of the distal small bowel. Three of these patients had terminal ileal ulcers and two of these had strictures as well. Thus, much of the damage detected by these very sensitive radionuclide scans is likely very small or superficial erosions. Nonetheless, these tiny lesions spread over a long length of distal small bowel may contribute to clinically significant blood loss. Other studies have shown that up to 50% of NSAIDs users, with iron deficiency anemia, have no detectable lesions on gastroscopy or colonoscopy, suggesting the blood loss comes from the small bowel. While most of the NSAID intestinal lesions are nonexciting, bland ulcers, there is one distinctive and striking small bowel lesion associated with NSAID use. A small percentage of patients taking NSAIDs develop "diaphragm disease" in which concentric luminal protrusions of fibrotic mucosa and submucosa occlude the lumen. These occlusions range from thin, delicate structures that appear to be only exaggerations of the plicae circulares to thicker, more rigid structures. These diaphragms can lead to a markedly narrowed lumen, leaving only a few millimeter opening in the center of the bowel. Most of diaphragm disease cases have been associated with slow release NSAIDs and most are in the terminal ileum. However, "diaphragm disease" has more recently been described in the jejunum and proximal colon as well. While the diaphragms have a very dramatic gross appearance, by histology they consist only of bland fibrosis, demonstrating a column of fibrotic tissue extending from the mucosa into deeper regions. Pathogenesis: The pathogenesis is uncertain, but it is hypothesized that NSAIDs have a direct effect on enterocytes causing damage to mitochondria and other subcellular changes that ultimately translate into increased intestinal permeability. This increase in permeability makes enterocytes more susceptible to injury by other forces such as ingested foods, bacteria, and particularly bile acids. The fibrotic changes of diaphragm disease may be a further reaction to tissue injury by the abovementioned factors. Natural History and Treatment: Most patients' symptoms resolve with cessation of NSAID administration. However, cessation of NSAIDs may be impractical in many patients. Therapeutic trials of drugs that may prevent NSAID damage (nabumetone, sulindac) or actually treat it (sulfasalazine, metronidazole) are ongoing. Differential diagnosis: Since the ulcers seen in NSAID usage have no distinctive pathologic features, the main problem is in attributing the ulcers to some other disease process, most notably Crohn's disease. Crohn's induced ulcers are more likely to show signs of chronic injury, including increased chronic inflammation, scarring, or pyloric metaplasia, and, if one is lucky, a granuloma. In reality, however, it may not be possible to reliably distinguish aphthous ulcers of Crohn's disease from NSAID ulcers Thus clinical history and clinical context may be of great import in arriving at the correct diagnosis. In asymptomatic patients, being endoscoped for cancer screening, with only the findings of small ulcers in the terminal ileum, the most likely diagnosis is NSAID damage. Other causes of small erosions or ulcers in the terminal ileum include trauma (bones or toothpicks,) radiation injury, vasculitis, or potassium related injury. References: Allison MC, et. al. Gastrointestinal damage associated with the use of nonsteroidal antiinflammatory drugs. N Engl J Med 327(11):749-754, 1992. Bjarnason I, et. al. Nonsteroidal antiinflammatory drug-induced intestinal inflammation in humans. Gastroenterol 93:480-489, 1987. Fellows IW, et. al. Non-steroidal anti-inflammatory drug-induced jejunal and colonic diaphragm disease: a report of two cases. Gut 33(10):1424-1426, 1992. Gargot D, et. al. Nonsteroidal anti-inflammatory drug-induced colonic strictures: two cases and literature review. Am J Gastroenterol 90(11):2035-2038, 1995. Kwo PY, et. al. Nonsteroidal anti-inflammatory drug-induced enteropathy: case discussion and review of the literature. Mayo Clin Proc 70:55-61, 1995. Lang J, et. al. Diaphragm disease: pathology of disease of the small intestine induced by non-steroidal anti-inflammatory drugs. J Clin Pathol 41(5):516-526, 1988. Lee, FD. Drug-related pathological lesions of the intestinal tract. Histopathol 25:303-308, 1994 Morris AJ, et. al. Enteroscopic diagnosis of small bowel ulceration in patients receiving non-steroidal anti-inflammatory drugs. Lancet 337:520, 1991. Bjarnason I, Hayllar J, Macpherson AJ, Russell AS. Side effects of nonsteroidal anti-inflammatory drugs on the small and large intestine in humans. Gastroenterol 104:1832-47, 1993. Cappell MS, Simon T. Colonic toxicity of administered medications and chemicals. Am J Gastroenterol 88:1684-1699, 1993. Faucheron JL. Toxicity of non-steroidal anti-inflammatory drugs in the large bowel. Eur J Gastroenterol Hepatol 11:389-92, 1999. Puspok A, Kiener H-P, Overhuber B. Clinical, Endoscopic, and Histologic Spectrum of non-steroidal anti-inflammatory drug-induced lesions in the colon. Dis Colon Rectum 43:685-91, 2000.