Clinicopathological features of nonsteroidal antiinflammatory drug-induced small intestinal strictures

Clinicopathological features of nonsteroidal antiinflammatory drug-induced small intestinal strictures

GASTROENTEROLOGY CASE 1988:94:1079-4 REPORT Clinicopathological Features of Nonsteroidal Antiinflammatory Drug-Induced Small Intestinal Strictures...

858KB Sizes 1 Downloads 39 Views





Clinicopathological Features of Nonsteroidal Antiinflammatory Drug-Induced Small Intestinal Strictures I. BJARNASON, A. B. PRICE, G. ZANELLI, P. SMETHURST, M. BURKE, J. M. GUMPEL, and A. J, LEVI Section of Gastroenterology and Division of Radioisotopes, Medical Research Centre, Clinical Research Centre, Departments of Gastroenterology, Radiology, Rheumatology, and Histopathology, Northwick Park Hospital, Harrow, Middlesex, United Kingdom

The clinicopathological features in 4 patients are described where nonsteroidal antiinflammatory drugs appear to have caused small intestinal strictures. Two patients had rheumatoid arthritis and 2 patients had osteoarthritis; all had received nonsteroidal antiinflammatory drugs for 1.5-30 yr. Three patients had an initial illness characterized by diarrhea, profound weight loss, and hypoalbuminemia. Intestinal radiology at this stage ranged from subtle changes to those of Crohn’s disease. Three patients underwent surgery. At operation the bland external appearance contrasted with the striking mucosal appearances of multiple concentric, circumferential, diaphragmatic strictures that caused luminal stenosis to a pinhole. These septa were due to submucosal fibrosis and the histopathological picture appears to represent a new nosological entity.


ecent studies have shown that 70% of patients on long-term nonsteroidal antiinflammatory drugs (NSAIDs) develop small intestinal inflammation (1). The pathogenesis of the inflammation is incompletely understood but thought to involve several interacting factors. Increased intestinal permeability is evident within hours of NSAID ingestion (2-4) and it is suggested that this exposes the mucosa to luminal macromolecules and toxins. In conjunction with the effects of NSAIDs on chemotaxis and neutrophil function (5), this may make way for bacterial invasion and hence inflammation, It is surprising, especially when compared with Crohn’s disease, and almost paradoxical that despite the wide use of NSAIDs and the high prevalence of NSAID-induced small intestinal inflammation, only a few patients have required surgical intervention for intestinal strictures (6-9). However, in a recent study 3 cases were identified by radiology where NSAIDs

were believed to have caused mucosal ulceration and strictures (1). However, tissue for pathological confirmation in these patients was not available. We now report on 4 patients in whom such a correlation between clinical features and pathology has been possible. These cases demonstrate the difficulty of diagnosis as well as characteristic histopathology that may be specific for NSAID-induced small intestinal disease. Case 1 A 40-yr-old woman was referred for a second opinion on the management of her rheumatoid arthritis. Clinical details are shown in Table 1. She had previously received a wide variety of NSAIDs and gold injections. One month later she was admitted for research studies (1). She had clearly been unwell for at least 4 wk, with weight loss (4 kg), fever, and diarrhea up to six times a day. Laboratory findings were unchanged except for increasing severity of the hypoalbuminemia (17 g/L). Indium 111 leukocyte studies were undertaken. Scintigrams showed the characteristic right iliac fossa abnormality of patients with NSAID-induced small intestinal inflammation (1) and fecal excretion was 4.64% (normal cl%). Simultaneous studies with %r-labeled proteins over a IO-day period (10) indicated a severe protein-losing enteropathy with a mean daily plasma protein clearance into the intestine of 318 ml/day (normal <25 ml/day). A barium follow-through study showed a stricture in the distal jejunum, and several strictures close together in the distal ileum. The terminal ileum was irregular and ulcerated, consistent with Crohn’s disease. Abdominal computed tomography scans showed mesenteric lymph node en-

Abbreviation used in this paper: NSAID, nonsteroidal antiinflammatory drug. 0 1988 by the American Gastroenterological Association 0918~5085/88/$3.50

April 1988



I. Clinical Age


Diagnosis and duration of disease





Rheumatoid (30 Yr)








Rheumatoid arthritis (15 yr) Osteoarthrosis (7 yr)

ESR, erythrocyte

Drug treatment


(2 yr)


Azapropazone Aspirin Paracetamol Penicillamine

1.2 g/day (3 2.4 g/day (5 3.0 g/day (3 250 mg/day


20-90 mgiday (1.5 yr)


30 gkg (3 yr)

Naproxen Sodium cromoglycate Theophylline

1.0 g/day (6 yr) 1.0 g/day (4 yr) 750 mg/day (4 yr)

rate. ’ Maximum


Abnormal hematologic and biochemical findings

and duration of treatment


yr) yr) yr) (3 yr)

Anemia of chronic disease Hemoglobin 9.2 g/d1 (normal, >12 gldl) White cells 19 x 103/yl (normal, cl1 x 103/p1) ESR 50 mm/h (normal, <8 mm/h) Serum albumin 27 g/L (normal, >35 g/L) Iron deficiency anemia Hemoglobin 9.9 g/d1 Serum albumin 25 g/L

Serum albumin

26 g/L

dose 30 mg.

strated a normal terminal ileum and a a-cm tight stricture in the midileum but no apparent holdup of barium. The patient developed subacute intestinal obstruction and underwent elective intestinal resection. A fresh 50-cm specimen of small bowel was received in the laboratory and fixed by inflation with formalin. Multiple short constrictions were apparent externally, and on opening the bowel, these corresponded to mucosal diaphragms encircling the bowel, reducing the lumen to 1 mm in some areas (Figure l]. There were 19 such strictures. Figure 2 shows their histologic features.

largement. Gastroscopy, barium enema, and colonoscopy with biopsy were all normal. The patient continued to lose weight with intermittent fever but no pathogens were cultured from blood, urine, or feces. A [14C]glycocholic acid breath test and a dual (75SeHCAT, 58Co vitamin Blz) isotopic ileal function test were all normal (1). Nine weeks later she had an explorative laparotomy. The serosal appearance of the small bowel was normal, but on palpation there was felt to be several strictured areas, none of which were causing critical obstruction. The findings were thought to be compatible with mild early Crohn’s disease. A mesenteric node biopsy showed a benign reactive pattern. Because of a flare-up of her arthritis she received sulfasalazine (Salazopyrin; Pharmacia Ltd., Milton Keynes, U.K. ), I g b.i.d. (other drugs being unchanged) for 18 mo with marked improvement of joint symptoms (11). A repeat “‘In leukocyte study after 6 mo on sulfasalazine showed improvement of the intestinal inflammation, and the fecal excretion had dropped to 2.01%. A small bowel barium eneTa (enteroclysis) at this time demon-



Case 2 A 70-yr-old woman was admitted to the hospital for investigation because of fatigue, weight loss (20 kg in 4 mo), and colicky abdominal pain. Two months previously she had been treated at another hospital for an acute diarrhea1 illness with fever, vomiting, and hypoalbuminemia (26 g/L.), but no firm diagnosis was reached despite a full gastrointestinal investigation. Clinical details are shown in Table 1. A plain abdominal x-ray revealed

B 1. A. The excised ileum of a patient (case 1) showing the small intestine divided into segments by multiple “diaphragms.” The bowel was prepared by inflating with formalin and cutting out multiple windows to demonstrate the compartments. B. This close-up demonstrates the “diaphragmatic” nature of each mucosal fold.


Figure 2. The diaphragms comprise only mucosa and submucosa. There is fibrosis of the submucosa at the apex with slight distortion of the overlying architecture. A normal plica circularis is immediately adjacent (H/E, x15.5).

distended small intestinal loops suggesting obstruction. The results of a small bowel enema examination (enteroclysis) are shown in Figure 3. A barium meal, endoscopic retrograde cholangiopancreatography, barium enema, and abdominal computed tomography scans were all normal. No pathogens were isolated from blood, urine, or feces. Her clinical condition worsened over the next 3 mo and albumin fell to 18 g/L. An explorative laparotomy was undertaken. The serosal aspect of the small bowel was normal but palpation of a [email protected] length some 70 cm distal to the ligament of Treitz suggested multiple luminal narrowings. Ileotomy revealed multiple mucosal diaphragms. After several stricturoplasties the extent and number of these (at least 30) was appreciated and 70 cm of small bowel was resected. Some 4 wk later a second 30-cm length of small bowel, identical to the first, was resected because of recurrent obstructive symptoms. Some affected bowel was still left in situ. Macroscopically and microscopically, the strictures in both resections were identical to those seen in case 1. Treatment with NSAIDs was discontinued and 14 mo later the patient is well. A d-day fecal excretion of ‘llIn was 1.56% and she has a mild protein-losing enteropathy (38 ml/day).


3. At the time of study this small (enteroclysis) from case 2 was thought retrospect, a number of strictures are tween arrows] as exaggerated mucosal spond to the “diaphragmatic” strictures 1. 2, and 4.

bowel enema to be normal. In evident (in befolds and correseen in Figures

April 1988



Her symptoms subsided over the course of 4 mo and she resumed her asthma medication, took paracetamol for the arthritic pain, and was given sulfasalazine (1 g t.i.d) for her bowels. Four months after discharge from the hospital she had made a full clinical recovery. However, “‘In leukocyte scintigrams were still abnormal, with a 4-day fecal excretion of 3%, and she had a protein-losing enteropathy (200 ml/day) despite normal serum albumin levels (38 g/L). A small bowel enema 1 yr later revealed that the thickened folds had resolved completely, but in the middle third of the small intestine there were now six focal strictures identical to those shown in Figure 3, each 2-10 mm in length. These reduced the lumen down to
Case 3 A 64-yr-old woman was admitted to the hospital with intermittent small intestinal obstruction of 3-mo duration. Clinical details are shown in Table 1. Indium 111 leukocyte scintigrams were abnormal as was fecal “‘In excretion (1.9%). A [14C]glycocholic acid breath test and stool cultures were normal. A small bowel enema showed dilatation of the small intestine but the site and nature of obstruction was not demonstrated. At laparotomy a 15-cm segment of distal ileum was resected containing three strictures that were identical to those in cases 1 and 2 (Figure 4).

Case 4 A 65-yr-old woman was admitted with a diarrhea1 illness of short standing with a stool frequency of 20 times a day. Clinical details are shown in Table 1. Over the next 4 mo her illness continued and was characterized by severe and resistant hypoalbuminemia (albumin, 11-18 g/L). Despite treatment with elemental diet (12), corticosteroids, and >46 U of purified plasma proteins she lost 20 kg in weight. Hemoglobin, white blood cells, and erythrocyte sedimentation rate remained normal throughout. Barium enema and colonoscopy with biopsy were normal. Barium follow-through showed that the distal jejunum and the whole of the ileum were diffusely abnormal, with thickened folds consistent with “nonspecific” malabsorption associated with hypoalbuminemia. An “‘In leukocyte study showed the characteristic right iliac fossa abnormality of NSAIDs and fecal excretion was increased to 4.01%. At the peak of her illness (3 mo after admission] she underwent laparotomy. No gross intestinal pathology was noted. A full-thickness biopsy specimen was taken from the ileum. Microscopy showed intact, though slightly irregular, villi. The lamina propria was expanded by inflammatory cells, mainly by plasma cells, histiocytes, and eosinophils. The muscle layer was uninvolved. A mesenteric lymph node showed reactive follicular hyperplasia. No more precise a diagnosis could be made than nonspecific ileitis.

Discussion Iatrogenic or drug-induced small intestinal disease has been thought to be rare, manifesting itself as malabsorption, perforation, or hemorrhage, or any combination thereof (7,13-15). Obstruction has only rarely been described as a presenting symptom, although small intestinal strictures have been ascribed to ingestion of slow-release potassium preparations (14-16). The pathogenesis in these cases is believed to involve local mucosal irritation and ulceration caused by tablets lodging against the intestinal mucosa. The high potassium concentration causes a vascular spasm that leads to ischemic necrosis with subsequent fibrotic repair and hence stenosis of the involved bowel (17,18).‘0ther drugs have been implicated less frequently. Although the precise mechanism of stricture formation almost certainly differs between the various drugs, the final histologic picture of broad-based strictures with submucosal fibrosis and overlying ulceration is sufficiently characteristic to be able to suggest a druginduced etiology. As multiple drugs have often been taken together in documented cases, there is often doubt about the precise causative agent. The final “proof” is at best circumstantial and always retrospective. The strictures in our patients differ markedly from those described above. Macroscopically, they were thin concentric diaphragmatic septalike structures producing variable luminal occlusion from little more than accentuated plica circularis to a pinhole. These lesions have, to date, only been found in patients taking NSAIDs and may represent an early stage of what we think to be a final common histopathological pathway of NSAID-induced intestinal strictures (19, and unpublished results] that eventually evolves into the characteristic drug-induced stricture, as was commonly described for potassium. The diaphragm lesion does pose a particular problem for diagnosis, which is illustrated in our pa-




tients. Because of the thin profile of the diaphragm, the radiologic findings were quite subtle except in the first case, where the findings were consistent with Crohn’s disease. Consequently, the number of strictures was greatly underestimated and it seems likely that they would be completely missed on follow-through examinations. For similar reasons, the bowel appearances at laparotomy are unimpressive and the abnormality is easily missed unless specifically sought. These findings suggest that NSAID-induced small intestinal strictures are more common than appreciated (19) and raise the question whether the previous reported association of Crohn’s disease with rheumatoid arthritis may in fact represent NSAIDinduced damage (20). In any case, it is clear that NSAID ingestion should now be considered in the differential diagnosis of small bowel disease, with and without radiologically demonstrable lesions. The microscopic features of the diaphragmatic lesions, together with the macroscopic appearances, appear to represent a new nosologic entity. The precise mechanism by which NSAIDs cause these appearances is unknown. The high prevalence of small intestinal inflammation due to NSAIDs measured by *“In leukocyte scintigrams and fecal collections, and the apparent rarity of the strictures, suggest that they are not simply the natural consequence or the end result of the inflammation. There are two features in our patients that may be relevant to the pathogenesis of the strictures. First, one patient (case 2) was clearly exceeding the maximum recommended dose of piroxicam some threefold to fourfold, which is approaching the dose used in the experimental animal to produce perforations. Whether the others had done so in the past we cannot assess, but it is of interest that previously reported cases of NSAID-induced ulceration or stricturing all involved larger doses of NSAIDs than currently recommended (6,8,9). Second, and perhaps of greater interest, is the history in 3 patients of a severe illness resembling, for lack of a better term, “gastroenteritis” with moderate to severe hypoalbuminemia. Histopathological assessment during this illness in 1 patient showed nonspecific inflammation. As the patient recovered and subsequently has apparently developed NSAID-induced strictures, we suggest that it is this intercurrent illness that is important in the pathogenesis of the diaphragmaticlike intestinal strictures. The nature of this “gastroenteritic” episode is uncertain. It is possible that the small-intestinal inflammation caused by NSAIDs predisposes these individuals to an infectious agent or, alternatively and not mutually exclsive, it may be that the diaphragmatic lesions represent a faulty reparative process after a particularly severe gastroenteritis in patients on NSAIDs.

References Bjarnason I, Zanelli G, Smith T, et al. Nonsteroidal antiinflammatory drug-induced intestinal inflammation in humans. Gastroenterology 1987;93:480-9. 2. Bjarnason I, Williams P, So A, et al. Intestinal permeability and inflammation in rheumatoid arthritis; effects of nonsteroidal anti-inflammatory drugs. Lancet 1984;ii:1171-4. 3. Bjarnason I, Williams P, Smethurst P, Peters TJ, Levi AJ. The effects of non-steroidal anti-inflammatory drugs and prostaglandins on the permeability of the human small intestine. Gut 1986;27:1292-7. 4. Jenkins RT, Rooney PJ, Jones DB, Bienenstock J, Goodacre RL. Increased intestinal permeability in patients with rheumatoid arthritis: a side-effect of oral nonsteroidal anti-inflammatory drug therapy? Br J Rheumatol 1987;26:103-7. 5. Bjarnason I, Zanelli G. Smith T, et al. The pathogenesis of non-steroidal anti-inflammatory drug induced small intestinal inflammation in man. Stand J Rheumatol (in press]. 6. Sturges HF, Krone CL. Ulceration and stricture of the jejunum in a patient on long-term indomethacin therapy. Am J Gastroenterol 1973:59:162-g. 7. Longstreith GF, Newcomer AD. Drug induced malabsorption. Mayo Clin Proc 1975;50:284-93. 8. Neoptolemos JP. Locke TJ. Recurrent small bowel obstruction associated with phenylbutazone. Br J Surg 1983;70:244-5. 9. Madhok R, MacKenzie JA, Lee FD. et al. Small bowel ulceration in patients receiving non-steroidal anti-inflammatory drugs for rheumatoid arthritis. Q J Med 1986;225:53-8. 10. Van Tongeren JHM, Reichert WJ. Demonstration of proteinlosing enteropathy: the quantitative estimation of gastrointestinal protein loss using 5’Cr-labelled plasma proteins. Clin Chim Acta 1966;14:42-8. 11. Pullar T, Hunter JA, Cape11 AJ. Sulphasalazine in rheumatoid arthritis: a double blind comparison of sulphasalazine with placebo and sodium aurothiomalate. Br Med J 1983;287: 1

1102-4. 12. O’Morain

C, Segal AW, Levi AJ. Elemental diets in the treatment of acute Crohn’s disease. Br Med J 1980;281:1173-

5. 13. Langman MJS, Morgan L, Worrall A. Use of anti-inflammatory

drugs by patients admitted with small or large bowel perforations and haemorrhage. Br Med J 1985;290:347-9. 14. Davies DR, Brightmore T. Idiopathic and drug-induced ulceration of the small intestine. Br J Surg 1970:57:134-9. or 15. Wayte DM, Helwig EB. Small-bowel ulceration-iatrogenic multifactoral origin? Am J Clin Path01 1968;49:26-40. 16. Lawrason FD, Alpert E, Mohr FL, McMahon FG. Ulcerativeobstructive lesions of the small intestine. JAMA 1965;191: 641-4. 17. Boley SJ, Schults L, Krieger H, et al. Experimental evaluation of thiazides and potassium as a cause of small-bowel ulcer. JAMA 1965;192:763-8. 18. Riddell RH. The gastrointestinal tract. In: Riddell RH, ed. Pathology of drug-induced and toxic disease. New York: Churchill Livingstone, 1982:515-605. 19. Lang J. Price AB, Levi AJ, Burke M, Gumpel JM, Bjarnason I. Diaphragm disease: the pathology of nonsteroidal antiinflammatory drug-induced disease of the small intestine. J Clin Path01 1988 (in press). 20. Ansell BM, Wigley RAD. Arthritic manifestations in regional enteritis. Ann Rheum Dis 1964;43:64-72.

Received July 14, 1987. Accepted October 20. 1987. Address requests for reprints to: Ingvar Bjarnason, M.D., Section of Gastroenterology, Division of Clinical Sciences, MRC Clinical Research Centre, Watford Road, Harrow, Middlesex HA1 3UJ, United Kingdom.