Behçet’s disease (BD), first described by a Turkish dermatologist, is an inflammatory disorder of uncertain origin characterized by an underlying vasculitis which is highly prevalent along the Silk Road [
Crohn’s disease (CD) is a chronic transmural inflammatory disorder that can affect the entire GI tract from the mouth to the anus [
Distinguishing between intestinal BD and CD is always a tough problem as they mimic each other clinically, endoscopically, and radiologically. Both the two conditions have younger age of onset, nonspecific GI symptoms, similar endoscopic features, and overlapping extraintestinal manifestations. Furthermore, lesions of both diseases could be located in any part of the alimentary tract, with the ileocecal region involved mostly [
A retrospective study of a single medical center was designed. The medical records of inpatients with intestinal BD and CD treated from February 1, 2004, to September 15, 2016, in Ruijin Hospital, Shanghai, China, were reviewed. All enrolled patients had undergone endoscopy as well as CT enterography (CTE) or MR enterography (MRE) in our hospital. Moreover, at least one-year follow-up after diagnosis was required for these patients. All enrolled patients had undergone T-SPOT (−) to exclude existing TB infection.
All enrolled BD patients had gastrointestinal involvement. A diagnosis of BD was based on the criteria suggested by the BD Research Committee of Japan in 1987 [
A diagnosis of CD was based on morphological (radiological, endoscopic, or surgical findings) and pathological criteria suggesting focal, asymmetrical, transmural, or granulomatous features [
Demographic, clinical, endoscopic, and radiologic data were collected at the time of initial diagnosis when the patients were naïve to immunomodulators’ therapy.
The demographic data included patients’ sex, age of onset, and BMI. Clinical manifestations were duration of symptoms, abdominal pain, diarrhea, abdominal distension, nausea, hematochezia, abdominal mass, fever, and weight loss. Extraintestinal manifestations included perianal disease, oral ulcers, genital ulcers, skin lesions, ocular lesions, musculoskeletal lesions, vascular abnormality, and neurologic lesions.
Double-balloon enteroscopy (DBE) was performed if the lesion was confined to the small intestine, while colonoscopy was performed for patients with only colonic involvement. Endoscopic features included multiple-site lesions, lesions confined to the ileocecal region, deformity of ileocecal valve, longitudinal ulcers, round and oval ulcers, punch-out ulcers, ulcers with discrete margin, ulcer
All of our enrolled patients had undergone CTE or MRE at least once which was evaluated by an experienced radiologist independently. Radiologic characteristics mainly included involvement of ≤3 segments, thickening bowel wall, asymmetrical pattern of involvement, intraluminal pseudopolyp formation, target sign, circumintestinal exudation, stricture with proximal dilation, comb sign, fistula, abscess, phlegmon, and ascites.
SPSS 19.0 was used for the data analyses and screening for potential valuable parameters for differential diagnosis between CD and BD. Continuous variables were expressed as mean ± SD, and a comparison was performed by using the Student
The medical records of 42 BD patients and 97 CD patients were reviewed. The demographic characteristics and clinical manifestations of intestinal BD and CD are summarized in Table
Demographic characteristics and clinical manifestations of intestinal BD and CD.
Features | Intestinal BD ( |
CD ( |
|
Score |
---|---|---|---|---|
Sex (male/female) | 20/22 | 54/43 | 0.382 | N/A |
Age of onset | 39.3 ± 17.3 | 36.9 ± 16.4 | 0.384 | N/A |
BMI (kg/m2) | 21.63 ± 3.2 | 21.08 ± 4.3 | 0.394 | N/A |
Duration of symptoms (months) | 25.2 ± 23.8 | 25.0 ± 24.3 | 0.969 | N/A |
Abdominal pain | 27 (64.3) | 75 (77.3) | 0.110 | N/A |
Diarrhea | 13 (31.0) | 65 (67.0) | <0.001 | −1 |
Abdominal distension | 11 (26.2) | 32 (33.0) | 0.426 | N/A |
Nausea | 4 (9.5) | 14 (14.4) | 0.429 | N/A |
Hematochezia | 11 (26.2) | 25 (25.8) | 0.959 | N/A |
Abdominal mass | 2 (4.8) | 6 (6.2) | 1.000 | N/A |
Fever | 14 (33.3) | 14 (14.4) | 0.011 | 1 |
Weight loss | 16 (38.1) | 51 (52.6) | 0.117 | N/A |
Perianal disease | 5 (11.9) | 42 (43.3) | <0.001 | −1 |
Oral ulcers | 42 (100.0) | 32 (33.0) | <0.001 | 1 |
Genital ulcers | 29 (69.0) | 14 (14.4) | <0.001 | 1 |
Skin lesions | 23 (54.8) | 8 (8.2) | <0.001 | 1 |
Ocular lesions | 3 (7.1) | 2 (2.1) | 0.326 | N/A |
Musculoskeletal lesions | 15 (35.7) | 8 (8.2) | <0.001 | 1 |
Vascular abnormality | 5 (11.9) | 7 (7.2) | 0.565 | N/A |
Neurologic lesions | 2 (4.8) | 1 (1.0) | 0.217 | N/A |
The endoscopic features of intestinal BD and CD are listed in Table
Endoscopic features of intestinal BD and CD.
Features | Intestinal BD ( |
CD ( |
|
Score |
---|---|---|---|---|
Multiple-site lesions | 22 (52.4) | 87 (89.7) | <0.001 | −1 |
Lesions confined to the ileocecal region | 38 (90.5) | 20 (20.6) | <0.001 | 1 |
Deformity of ileocecal valve | 8 (19.0) | 19 (19.6) | 0.941 | N/A |
Longitudinal ulcers | 5 (11.9) | 85 (87.6) | <0.001 | −1 |
Round or oval ulcers | 33 (78.6) | 29 (29.9) | <0.001 | 1 |
Punch-out ulcers | 25 (59.6) | 13 (13.4) | <0.001 | 1 |
Ulcers with discrete margin | 33 (78.6) | 59 (60.8) | 0.042 | 1 |
Ulcer |
22 (52.4) | 32 (33.0) | 0.031 | 1 |
Cobblestone appearance | 5 (11.9) | 31 (32.0) | 0.013 | −1 |
Aphthous lesions | 22 (52.4) | 45 (46.4) | 0.516 | N/A |
Stricture of bowel | 3 (7.1) | 23 (23.7) | 0.021 | −1 |
Anorectal involvement | 3 (7.1) | 21 (21.6) | 0.038 | −1 |
Endoscopic and CTE features in intestinal BD. (a, b) Colonoscopy revealed oval punch-out ulcer with discrete margin in the ileocecal region. (c, d) CTE showed the same patient with thickening bowel wall and circumintestinal exudation in the terminal ileum.
Endoscopic and CTE features in CD: (a) longitudinal ulcer that stretched across several folds under endoscopy; (b) bowel obstruction in the ascending colon due to chronic inflammation; (c) perianal fistula with multiple internal openings; (d) stretching and densifying of distal mesenteric artery the so-called comb sign in the ileum; (e) internal bowel-bowel fistula the so-called petal sign; (f) asymmetrical thickening of the bowel wall with target sign.
For radiologic findings shown in Table
Radiologic findings of intestinal BD and CD.
Features | Intestinal BD ( |
CD ( |
|
Score |
---|---|---|---|---|
Involvement |
32 (76.2) | 37 (38.1) | <0.001 | 1 |
Thickening bowel wall | 37 (88.1) | 89 (91.8) | 0.717 | N/A |
Asymmetrical pattern of involvement | 6 (14.3) | 40 (41.2) | <0.001 | −1 |
Intraluminal pseudopolyp formation | 8 (19.0) | 58 (59.8) | <0.001 | −1 |
Target sign | 2 (4.8) | 21 (21.6) | <0.001 | −1 |
Circumintestinal exudation | 31 (73.8) | 74 (76.3) | 0.755 | N/A |
Stricture with proximal dilation | 4 (9.5) | 24 (24.7) | 0.040 | −1 |
Comb sign | 7 (16.7) | 73 (75.3) | <0.001 | −1 |
Fistula | 1 (2.4) | 37 (38.1) | <0.001 | −1 |
Abscess | 3 (7.1) | 12 (12.3) | 0.539 | N/A |
Phlegmon | 1 (2.4) | 8 (8.2) | 0.360 | N/A |
Ascites | 0 (0.0) | 6 (6.2) | 0.233 | N/A |
The total score was calculated by pooling all of the valuable differential parameters together. A differentiating diagnostic model was established, with high sensitivity (90.5%), specificity (93.8%), accuracy (92.8%), PPV (86.4%), and NPV (95.8%). A ROC curve was plotted. Based on the Youden index, a diagnostic point of 0.5 was obtained (
ROC curve of the differentiation model (area under the ROC curve is 0.981).
BD is a systematic disease mainly characterized by oral, genital, ocular, and skin lesions. Sometimes, BD patients can present with GI ulcers, the so-called intestinal BD. CD is a chronic inflammatory GI disease which may also have extraintestinal manifestations mimicking BD greatly [
Among various demographic and clinical parameters, our study showed that diarrhea, fever, and perianal disease were most useful in differentiating intestinal BD from CD. Among them, fever favored a diagnosis of intestinal BD, whereas diarrhea and perianal disease favored a diagnosis of CD. These had further proved that intestinal BD and CD had overlapping clinical manifestations. For several systematic symptoms listed in the diagnostic criteria of BD, we found that only four of them, including oral, genital, skin, and musculoskeletal lesions, aided the differential diagnosis. Thus, it could be inferred that systematic symptoms of BD could mimic extraintestinal manifestations of CD. Consequently, we could not make a differential diagnosis between the two conditions only according to the symptoms. Our findings were similar to those of Li et al. [
Endoscopy is the first choice for clinicians to detect bowel lesions and evaluate therapeutic response. In our study, we found that the ulcers’ distribution of the two diseases was different. CD patients tended to have multiple-site involvement, whereas lesions of intestinal BD were more likely to be confined to the ileocecal region. Moreover, the morphology of the lesions was different from each other. Ulcers of intestinal BD were always round or oval in shape, punch-out in feature, >2 cm in size, and with a discrete margin. On the other hand, ulcers of CD were mostly longitudinal in shape. Cobblestone appearance, stricture of bowel, and anorectal involvement in CD patients may help distinguishing them from patients with intestinal BD. Our study has proved that although quite similar, some distinctions do exist between intestinal BD and CD in endoscopy, which is helpful to the differential diagnosis. These findings are in good agreement with those of Lee et al. [
CTE and MRE are emerging noninvasive technology for the diagnosis and evaluation of small-bowel diseases. Compared with endoscopy, they have a better diagnostic efficacy of both bowel wall lesions and extraenteric manifestations [
Although a series of differentiating parameters had been proposed, none of them enjoyed high sensitivity and specificity at the same time. As a result, differentiating between these two conditions through a single parameter is really difficult. Thus, we graded each parameter and established a diagnostic model that combined all of the valuable parameters together. Through later statistical analysis, we proved that our model had high diagnostic efficacy, with high sensitivity (90.5%), specificity (93.8%), accuracy (92.8%), PPV (86.4%), and NPV (95.8%). Based on the Youden index, we acquired a cutoff value of 0.5 and the area under the curve was 0.981. We believe that our differentiation model is more integrated and serve better, helping clinical practitioners to solve this problem.
This study has some limitations. First, as its retrospective nature and limited number of patients, the level of evidence is limited. We expect more prospective studies and multicenter collaboration being carried out regarding this field. Second, we did not include other ulcerous bowel disease into the analysis. Thus, this differentiating model could be used only if other diseases had been excluded, which may hamper its application. Third, all of the enrolled patients were inpatients in our hospital, which may bring selection bias.
In conclusion, intestinal BD and CD have overlapping characteristics making it hard to distinguish from each other. However, some parameters including clinical manifestations, endoscopic features, and radiologic characteristics are valuable in differentiating these two conditions. The established differentiating model that collated different parameters together yields high diagnostic efficacy, which will be very helpful in clinical practice.
All authors declare that there is no conflict of interests regarding the publication of this paper.
Tianyu Zhang and Liwen Hong contributed equally to this article.
This study was supported by the National Natural Science Foundation of China (nos. 81670503 and 81602558) and Shanghai Committee of Science and Technology Foundation (nos. 16411950408 and 15ZR1426400).