- Top of page
- Case Reports
Scalloping of the duodenal mucosal folds is an endoscopic finding of small bowel mucosal pathology that is generally due to villous atrophy. Though it can be seen in many disease processes, it is most commonly associated with celiac disease. We report three patients with scalloping of duodenal folds and histologic confirmation of villous atrophy due to Crohn's disease. All patients had negative celiac serologies and two had positive markers for Crohn's disease (anti–Saccharomyces cerevisiae antibodies). Patients had either ileitis or ileocolitis in addition to duodenal abnormalities. These cases illustrate that scalloping can occur in the duodenum in Crohn's disease.
- Top of page
- Case Reports
Scalloping of duodenal mucosal folds on endoscopic examination is most often a sign of villous atrophy secondary to celiac disease. 1,8 The three cases presented here demonstrate that this endoscopic finding is not specific for celiac disease. Each of our three patients had, in addition to scalloping on upper endoscopy, either ileitis or ileocolitis making a unifying diagnosis of Crohn's disease more likely. Pathologic examination of duodenal biopsies revealed mild villous atrophy, patchy inflammation, and granulomas. These findings are consistent with Crohn's disease. Two patients had positive serological markers (ASCA) supportive of the diagnosis of Crohn's disease. 9
The patients' clinical course and response to medication were also consistent with Crohn's disease. None of the patients had positive celiac serologies. Though patients with celiac disease may improve on steroids and immunosuppressive agents, the long-term follow-up of our patients showed histologic improvement without a gluten-free diet. These clinical scenarios and the follow-up biopsies make a diagnosis of celiac disease unlikely.
A recent paper identified scalloping as being one of the least sensitive markers for celiac disease (6%), but nevertheless highly specific (100%). 10 Most often, one sees loss or reduction of duodenal folds, scalloping of the Kerkring folds, and a mosaic duodenal mucosal pattern in celiac disease. The sensitivity of these findings is 50%, but they are highly specific (99%) with positive and negative predictive values of 60% and 99%, respectively. 11 Endoscopic markers of duodenal villous atrophy are not specific for celiac disease. They may suggest duodenal involvement with Crohn's disease in addition to tropical sprue, HIV enteropathy, Giardiasis, small intestinal opportunistic infections, and amyloidosis. 3
Crohn's disease may affect any part of the gastrointestinal tract. Early estimates of a 5% incidence of gastroduodenal involvement have been replaced by higher rates of up to 83%. 12 Better diagnostic procedures including upper endoscopy with biopsy have led to greater detection of proximal Crohn's disease. A retrospective histologic study by Wright and Riddell found that more than 75% of patients with Crohn's disease of the distal gastrointestinal tract have upper gastrointestinal abnormalities on biopsy. 6 Crohn's disease of the esophagus, stomach, and duodenum can occur in 5%–30% of patients with ileocolonic disease. 13
Patients with proximal disease may present with symptoms of anorexia, epigastric pain, dyspepsia, diarrhea, fever, weight loss, or a combination of these symptoms. 14 Upper gastrointestinal involvement with Crohn's disease is most commonly found in the antrum, duodenal bulb, and distal duodenum. Duodenal biopsy findings include villous atrophy (23%), focal inflammation (40%), intraepithelial lymphocytosis (22%), gastric metaplasia (18%), erosions (3%), and granulomas (1%). 15 Granulomatous inflammation, long considered to be the diagnostic histopathologic hallmark of gastroduodenal Crohn's disease, is actually only present in 6%–15% of biopsy specimens. 16 Changes classically associated with celiac disease, such as villous atrophy and intraepithelial lymphocytosis, are more common. 17
In Crohn's disease, endoscopic evaluation of the esophagus, stomach, and duodenum may show a variety of findings. The esophagus may have erosions, ulcers, or nodularity. 18 The antrum and duodenum may have granular mucosa with patchy erythema, friability, nodularity, or ulceration. The duodenum may have irregular, thickened folds or scalloped mucosa. 19 Stricture formation may be seen in the upper tract in severe cases.
Pathologic examination of the duodenal biopsies of these three patients showed an increase in chronic inflammatory cells, granulomas, villous atrophy, and crypt hypertrophy. Except for granulomas, each of these individual histologic findings is typically seen in celiac disease. In Crohn's duodenitis, however, changes are patchy, varying in intensity or alternating between normal and abnormal within a single biopsy or between separate biopsy pieces. Because the demonstration of patchiness is so crucial for arriving at the correct diagnosis, multiple biopsy specimens need to be taken.
Several case reports have described patients with celiac disease in addition to Crohn's disease or ulcerative colitis. 20–22 The diagnoses are based on a combination of serologies for celiac disease, histopathology, and response to therapy with either a gluten-free diet or steroids and 5-ASA compounds. The coexistence of celiac disease with other autoimmune disorders is commonly described, but the diagnosis of celiac disease and inflammatory bowel disease is infrequent. 22
In our three cases, serologic markers for both diseases proved to be useful. Tissue transglutaminase antibodies based on the ELISA assay are highly sensitive and specific for celiac disease. 23 In addition, immunoglobulin A autoantibodies to endomysium are close to 100% sensitive and specific for celiac disease. 24 Unfortunately, the disappearance of endomysial antibodies in treated celiac disease patients does not indicate histologic recovery. 25 It has been suggested as a tool for monitoring dietary compliance. In our three patients, this was not at issue.
Serological markers are helpful diagnostic tools to differentiate Crohn's disease from ulcerative colitis. Anti-Sacchromyces cerevisiae antibody has up to 60% sensitivity and 91% specificity for Crohn's disease. 26 The positive and negative predictive values are 88% and 68%, respectively. For perinuclear antineutrophil cytoplasmic antibodies (pANCA), the sensitivity, specificity, positive and negative predictive values are 50%, 95%, 69%, and 89%, respectively. 26 The specificity for these markers is high, but their low sensitivity makes them less useful. In combination with clinical scenarios and histopathologic confirmation, there is a greater diagnostic value. A retrospective study evaluated 37 patients with celiac disease by history (chronic diarrhea) and serology (87% EMA positive, 87% tTG positive) and found that 16 of the 37 patients were ASCA positive. 27 The presence of ASCA in the serum of patients with celiac disease can occur.
The type of inflammatory infiltrate in Crohn's disease is different from that in celiac disease. The concentration of chronic inflammatory cells are at the base of crypts rather than along the surface. The admixture of histiocytes, sometimes aggregating into ill-defined granulomas, and aggregates of neutrophils, sometimes associated with acute cryptitis, are features in favor of Crohn's disease. The study of Wright and Riddell suggests that intraepithelial lymphocytosis does not help to distinguish between Crohn's and celiac disease. Focal acute inflammation of the duodenal mucosa, in the absence of gastric Helicobacter pylori infection, was highly specific (92% specificity) for Crohn's disease, but was not found in our three cases. 6 Because H. pylori is a significant cause of gastritis, it must be excluded before focal acute inflammation can be attributed to Crohn's disease.
Scalloping of duodenal mucosal folds is most commonly seen as a result of celiac disease. These cases, however, illustrate that proximal Crohn's disease may manifest with this endoscopic appearance as well. Consequently, it is useful to correlate the finding of scalloping with clinical, histologic, and serologic data in making a specific diagnosis.