Ileocecal involvement in intestinal Behçet's disease and Crohn's disease: comparison of clinicopathological and immunophenotypic features

Intestinal Behçet's disease (BD) predominantly affects the ileocecal region and is currently diagnosed based on endoscopic features and clinical manifestations. It is difficult to distinguish between intestinal BD and Crohn's disease (CD) due to similar patient populations, gastrointestinal involvement, extraintestinal manifestations, and long‐term recurrent course. In this study we aimed to compare the clinicopathological and immunophenotypic features of intestinal BD to CD.

endoscopic finding is a large, deep, volcano-shaped ulcer with histological features of vasculitis and non-specific inflammation. 2,4,5rrently, diagnosis of intestinal BD is made based on clinical manifestations associated with systemic BD.7][8] Cheon et al developed a diagnostic model based on endoscopic features and extraintestinal manifestations of patients with intestinal BD in 2009, offering the possibility of an early diagnosis. 9Notably, none of these diagnostic criteria are established based on a reliable histological basis.
Intestinal BD is difficult to differentiate from Crohn's disease (CD) due to their similarities in patient populations, region of GI involvement, extraintestinal manifestations, and a long-term relapsing-remitting disease course. 1,10There is also an overlap in the management between the two diseases; however, vascular-related disease and the potential recurrence of ulcers at the anastomosis in intestinal BD require aggressive treatment and close monitoring. 11,12The features of longitudinal ulcers, cobblestone appearance, perianal lesions, and non-caseating granulomas can be used to distinguish CD from intestinal BD. 9,13,14 While the absence of these typical features, along with limited detection of granulomas and vasculitis in biopsy specimens, makes the differential diagnosis difficult, especially in patients with ileocecal ulcer. 1,2,10Detailed comparative studies on the histopathological features of ileocecal ulcer biopsies between intestinal BD and CD have not yet been performed.In this study we aimed to evaluate the clinicopathological and immunophenotypic features of patients with intestinal BD and CD.

| Patients and diagnostic criteria
Medical records including pathological findings of 29 cases with intestinal BD and 120 cases with CD diagnosed at the Department of Gastroenterology, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine (Hangzhou, Zhejiang Province, China) between January 2015 and December 2021 were retrospectively reviewed.The diagnosis of intestinal BD was made based on endoscopic features and clinical manifestations of the patients according to the criteria developed by Cheon et al. 9 "Typical" ulcers were defined as colonoscopic findings of less than five ulcers that were oval in shape, deep, with discrete borders, and located in the ileocecal area.While those did not show all characteristics were considered atypical. 9Only patients met the classification criteria for definite and probable intestinal BD were included in the analysis.Definite intestinal BD was diagnosed based on typical intestinal ulceration at the ileocecal region and systemic BD.And probable intestinal BD was diagnosed based on typical intestinal ulceration at the ileocecal region with oral ulcer only, or atypical intestinal ulcer with systemic BD.CD was diagnosed based on the 3rd European consensus on the diagnosis and management of CD according to patients' clinical, radiological, endoscopic, and histopathological findings. 15The inclusion criteria for intestinal BD and CD were as follows: (a) presence of endoscopically confirmed ileocecal ulcer(s); (b) biopsy specimens of the ulcers obtained from the ileocecal region; and (c) with sufficient baseline clinical and follow-up data for analysis.Those with comorbid tumors, intestinal tuberculosis (TB), lack of biopsy specimens from the ileocecal region, or insufficient baseline or follow-up information were excluded.Patients suspected to have intestinal TB was excluded by tuberculin skin test, tissue TB polymerase chain reaction, Myobacterium tuberculosis culture, tissue acid-fast bacilli staining, and even empirical antituberculosis treatment. 10l cases were followed up for at least 1 year, and follow-up information was obtained from their medical records or telephone followup.The study was approved by the Institutional Ethics Committee of Sir Run Run Shaw Hospital, Zhejiang University School of Medicine.
Written informed consent was waived due to the retrospective study design.

| Clinicopathological evaluation and definitions
Demographics, clinical symptoms, intestinal complications, previous history of surgery, endoscopic findings, laboratory test results, pathological features of the biopsy specimens, and follow-up data of the patients were reviewed and collected.
Clinical manifestations included abdominal pain, chronic diarrhea, melena or bloody stool, fever, weight loss, perianal disease, and extraintestinal manifestations (such as oral ulcer, genital ulcer, skin lesions), and a previous history of surgery.Perianal lesions included anal fistula, perianal abscess, and anal fissure.Recurrent oral ulcers were defined as three or more episodes of oral ulcers occurred within 12 consecutive months. 3Intestinal complications included stricture, perforation, obstruction, adhesion, and intestinal fistula detected by radiological examination, endoscopy or during surgery.Erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and hemoglobin (Hb) levels were collected from the initial test results during the corresponding hospitalization. 16Missing values were defined when no relevant laboratory tests were performed.A previous history of surgery was referred to intestinal resection due to intestinal complications or ineffective treatment of the disease (excluding those for perianal lesions). 10The 5-year cumulative probabilities of surgery were examined.The outcome event was evaluated from the GI disease onset to surgery using data obtained from the medical records and follow-up information.
Patients without information on the date of a specific surgery or lost to follow-up were excluded from the analysis.Histopathological features included vasculitis, crypt irregularity, cryptitis, crypt abscess, transmucosal inflammation, basal plasmacytosis, granulomas, pseudopyloric gland metaplasia, villous atrophy, hemorrhage, excessive neutrophils, chronic inflammatory infiltrate in mucosal lamina propria, and prominent lymphocytic infiltration in ulcer tissue, and neutrophilic exudate on the ulcer surface were carefully evaluated in all cases by pathologists.
The histological definitions of crypt irregularity, cryptitis, crypt abscess, basal plasmacytosis, granulomas, and villous atrophy were derived from the European consensus. 14,17Transmucosal inflammation was characterized by increased lymphocytes and plasma cells in the lamina propria of the mucosa.Vasculitis was histopathologically defined as degeneration or regeneration of the endothelial cells, thickening of the blood vessel wall, and infiltration of inflammatory cells, especially lymphocytes, in the vascular wall, after exclusion of the capillaries or small vessels immediately adjacent to the ulcer. 5Neutrophils and chronic inflammatory infiltrates in the lamina propria of the mucosa were classified into four grades according to the Geboes score: 0, no increase; 1, mild but unequivocal increase; 2, moderate increase; and 3, marked increase. 18Excessive neutrophils were defined as moderately and markedly increased neutrophils in the lamina propria.Plasma cells and other mononuclear cells were assessed separately.Dense lymphocytic infiltration was defined as infiltration of a predominantly large number of lymphocytes of the ulcer tissues, excluding a mixed infiltration of chronic inflammatory cells, such as lymphocytes, plasma cells, and histiocytes.Neutrophilic exudate was frequently observed on the BD ulcer surface in intestinal BD.Thickness of neutrophilic exudate was defined as the maximum thickness of neutrophil accumulation above the intact ulcer tissues and the maximum thickness of neutrophil exudate tissue for the few fragmented ulcer tissues.If no ulcerated tissue was biopsied, a missing value was defined.The sections were scanned using a digital section scanner (KF-PRO-020; KFBIO, Yuyao, Zhejiang Province, China), and five different circular areas (200 μm in diameter) with the most prominent infiltration and the highest degree of positivity were selected to count the numbers of FOXP3 + , MPO + , and CD3 + , CD20 + cells.The average value was considered for statistical analysis.Lymphoid follicles were excluded from the analysis.

| Statistical analysis
Categorical variables were expressed as numbers and percentages, and were compared between groups using the χ 2 test or Fisher's exact test.Statistical analysis was performed using a nonparametric test (Mann-Whitney U-test) for rank variables.Normal distribution of continuous variables was determined by using the Kolmogorov-Smirnov test.Continuous variables with a normal distribution were expressed as mean ± standard deviation, and were compared by using the Student's t-test, while those with non-normal distribution were expressed as median and interquartile range (IQR) and were compared by using the nonparametric analysis (Mann-Whitney U-test).The receiver operating characteristic (ROC) curve analysis was used to evaluate the discriminative ability and identify the optimal cut-off value of neutrophil exudate thickness on the ulcer surface.The area under the ROC curve (AUROC) was used to identify a cut-off value by maximizing the combination of sensitivity and specificity.The cumulative probabilities of surgery were assessed using the Kaplan-Meier method, and the curves were compared using the log-rank test.P < 0.05 was considered statistically significant.Statistical analysis was performed using the SPSS Statistics version 25.0 (IBM, Armonk, NY, USA).

| Clinical characteristics of patients with intestinal BD and CD
Clinical characteristics and endoscopic findings of 29 intestinal BD patients and 120 CD patients are summarized in Table 1.Among the 29 patients with intestinal BD, 20 were male and nine were female, with a median age at presentation of enteropathy of 25.0 years (IQR 11.5-33.0years).While among the CD patients, 85 were male and 35 were female, with a median age of 25.0 years (IQR 19.0-33.0years) at disease onset.Age at disease onset and gender were comparable between the two groups (P = 0.531 and 0.843, respectively).Compared with CD, abdominal pain was more common in patients with intestinal BD (96.6% vs 79.2%, P = 0.027), whereas chronic diarrhea was less frequent (24.1% vs 48.3%, P = 0.018).While melena or bloody stool, fever, and weight loss did not differ between the two groups.ESR was significantly higher in patients with intestinal BD than those in CD (48.6 ± 38.4 mm/h vs 19.0 ± 20.2 mm/h, P = 0.009), while the CRP level and the prevalence of anemia were not statistically different between the two groups.

| Histopathological features of ileocecal ulcers in BD and CD patients
Histopathological features of the ileocecal biopsy specimens were compared between the intestinal BD and CD groups (Table 2).Crypt irregularity was identified in 92.9% of patients with intestinal BD and all the CD patients (P = 0.036).The differences between intestinal BD and CD in cryptitis, transmucosal inflammation, and pseudopyloric gland metaplasia were not statistically significant.Additionally, crypt abscesses (14.3% vs 35.9%,P = 0.027) and basal plasmacytosis (3.6% vs 22.6%, P = 0.021) were significantly less frequent in intestinal BD than in CD.Small intestinal mucosa was examined in 10 of the 29 BD patients and 34 of the 120 CD patients, and there was a statistically significant difference in villous atrophy between the two groups (80.0%[8/10] vs 100% [34/34], P = 0.048).We also detected granulomas in 41.9% of biopsy specimens of the CD patients but none in those with intestinal BD (P < 0.001).Typical vasculitis was not found in either group.Excessive neutrophils in the mucosal lamina propria were significantly more frequent in intestinal BD (46.4% vs 12.8%, P < 0.001).Plasma cells and other mononuclear cells were evaluated separately.Plasma cells were moderately and markedly increased in most cases of CD but mildly increased in those with intestinal BD (P < 0.001), whereas other mononuclear cells did not differ between the two groups (P = 0.126).We observed that neutrophilic exudate on the ulcer surface (89.7% vs 51.1%, P < 0.001; Figure 1A,B) and dense infiltration of lymphocytes in ulcer tissues (93.1% vs 8.9%, P < 0.001; Figure 1C,D) were more common in intestinal BD than in CD.We measured neutrophilic exudate thickness and found that it was significantly thicker in intestinal BD patients than in CD patients (430.00 ± 291.24 μm vs 124.11 ± 163.81 μm, P < 0.001; Figure 1E).
T A B L E 1 Clinical information and endoscopic performance of patients with intestinal Behçet's disease (BD) and Crohn's disease (CD).F I G U R E 1 Histopathological findings of ileocecal biopsy specimens of (A,C) intestinal Behçet's disease and (B,D) Crohn's disease, as well as (E,F) the marker of neutrophilic exudate thickness and its receiver operating characteristic curve.HE stain.Magnification: A,B, Â160; C-E, Â200.

| Differentiating capacity of neutrophilic exudate thickness
Neutrophilic exudate thickness exhibited a relatively strong discriminatory ability between intestinal BD and CD patients.The AUROC was 0.820 (95% confidence interval [CI] 0.726-0.913)for the marker of the neutrophilic exudate thickness (Figure 1F).When using a cut-off value of 145 μm, neutrophilic exudate thickness showed a sensitivity of 89.7% and a specificity of 64.4%.Meanwhile, there were more intestinal BD cases with neutrophilic exudate thickness ≥145 μm on the ulcer surface than those with CD (89.7% vs 35.6%, P < 0.001).

| Outcome
The cumulative probability of surgery after the onset of GI involvement differed significantly between patients with intestinal BD and CD (Figure 3).At 1, 3, and 5 years after the onset GI involvement, the cumulative rates of surgery for intestinal BD were 6.9%, 30.0%, and 54.1%, respectively, and for CD patients the rates were 6.0%, 12.5%, and 25.4%, respectively (P = 0.029).

| DISCUSSION
It is challenging to make a differential diagnosis between intestinal BD and CD at early presentation because these two diseases share many similarities. 1,10,19Studies on the comparison of pathological characteristics of GI biopsy specimens between BD and CD are scarce, and intestinal biopsies have not been systematically analyzed, especially in patients with ileocecal ulcers.In a study on upper GI biopsy specimens from 50 CD patients and 34 BD patients, Akemoto et al found that in H. pylori-negative patients, focal gastric neutrophilic infiltration and wide duodenal gastric foveolar metaplasia were important in distinguishing CD from BD. 20 The histology of esophageal tissues typically reveals non-specific inflammation rather than the typical features of vasculitis. 21Ciscato et al described a case of BD that mimicked CD as the pathological findings of GI biopsies exhibited cryptitis, crypt abscess, and ruptured crypt-associated granulomas. 22The presence of vasculitis suggests a possibility of intestinal BD; however, it is difficult to be detected in biopsied specimens. 1,21,23Therefore, superficial colonoscopic biopsies may lead to misdiagnosis of BD by pathologists in the absence of typical features.
Our current study demonstrated some significant differences in histological characteristics of ileocecal ulcer biopsy specimens between intestinal BD and CD, which is the most involved region in BD.We found that excessive neutrophils in the the lamina propria of the mucosa, relatively thick neutrophilic exudate on the ulcer surface, and prominently lymphocytic infiltration in the ulcer tissues were significantly more common in intestinal BD than in CD.The neutrophilic exudate thickness was significantly higher in patients with intestinal BD than in those with CD.The sensitivity and specificity of neutrophilic exudate thickness in distinguishing BD and CD were 89.7% and 64.4%, respectively, when the cut-off value of neutrophilic exudate thickness was 145 μm (AUROC 0.820).However, the value of neutrophil exudate thickness in biopsy tissue is relatively limited due to the limited availability of biopsy tissue, which would reflect the full picture of neutrophil exudate on the ulcer surface.We hope to analyze this histological variable in surgical specimens to find the best and most accurate threshold for distinguishing between these two diseases.
Plasma cells were mostly moderately and markedly increased in CD but only slightly increased in intestinal BD.Additionally, we showed that intestinal BD and CD shared some similar histopathological features, such as cryptitis, transmucosal inflammation, and pseudopyloric gland metaplasia.No vasculitis was identified in either group, once again highlighting the difficulty in diagnosing intestinal BD on small mucosal biopsy specimens.It is controversial regarding the presence of granulomas in intestinal BD, and a few studies have reported the detection of granulomas in BD patients. 23,24Our study found no typical granulomas in BD biopsy specimens that could be used for differential diagnosis.
9][30] FOXP3 + CD4 + T cells were increased in mucosal lymphoid tissues (mucosal lamina propria and mesenteric lymph nodes) in patients with active CD, 29 and those with BD had significantly higher CD4 + CD25 + FOXP3 + regulatory T cell counts than the healthy controls. 30However, no comparative immunophenotypic study, including CD3, CD20, and FOXP3, between intestinal BD and CD has been conducted.In our study, immunohistochemical results revealed that intestinal BD had significantly higher CD3 + T cells and CD20 + B cells than CD in the lamina propria and ulcer tissue.
FOXP3 + T cells were significantly higher in BD than in CD, but the ratio of Foxp3 + /CD3 + T cells did not differ.One study showed that level of MPO, a component of neutrophil extracellular traps, was elevated in active BD patients and might contribute to the procoagulant F I G U R E 3 Five-year cumulative probabilities of surgery after gastrointestinal (GI) disease onset between intestinal Behçet's disease (BD) and Crohn's disease (CD).
state. 31Our histological findings demonstrated that excessive neutrophils in the mucosal lamina propria were significantly more frequent in intestinal BD than in CD.Immunohistochemical results showed that there was an increasing trend in MPO + cells in intestinal BD than in CD, despite the absence of a statistically significant difference.These morphological and immunohistochemical findings suggest that lymphocyte and neutrophil infiltrates are more common in intestinal BD, whereas plasmacytosis is more common in CD, possibly indicating differences in the major inflammatory cells involved in the pathogenesis.
We also analyzed the differences in clinical and endoscopic features between patients with intestinal BD and CD.We observed that abdominal pain was more common in intestinal BD than in CD, while chronic diarrhea was more common in CD.Liu et al observed that ESR and CRP levels were not statistically different between 29 active CD and 17 BD patients. 16However, we revealed that intestinal BD patients had a higher average level of ESR.Additionally, our patients patients with GI involvement. 32Consequently, we adopted this diagnostic model to identify intestinal BD patients.We hope our findings may help the early diagnosis of intestinal BD and its differentiation from CD.In clinical practice, it might not be possible to fully distinguish between the two diseases on the basis of histological features alone.A comprehensive clinicopathological analysis of the two diseases is therefore of particular importance.We hope that our histological findings will serve as one of the bases for differential diagnosis.Unfortunately, the lack of some histological variables in the biopsy specimens in our study made it impossible to construct a clinicopathological model.More cases should be included to comprehensively evaluate and to construct a more complete clinicopathological differential model.
There were some limitations to our study.We conducted a retrospective study with a relatively small sample size, which might have led to selection bias.In addition, some of the laboratory test results and morphological variables were missing, which might have affected the statistical results to some extent.Therefore, multicenter collaboration and further validation studies are needed.
In conclusion, intestinal BD and CD share similarities in clinical manifestations, relative examination, and pathological features.In addition to the typical clinical and endoscopic findings, intestinal BD can show some histological and immunophenotypic features that may be important in differentiating it from CD. Understanding these features and focusing on biopsies of the ulcers and adjacent mucosa may help guide the differential workup of intestinal BD and CD.
Variables expressed as the number of positive cells per circular area (200 μm in diameter).Abbreviations: IQR, interquartile range; MPO, myeloperoxidase; SD, standard deviation.*P < 0.05.F I G U R E 2 Immunohistochemistry of inflammatory cell infiltration in ulcerated tissues of (A-C) intestinal Behçet's disease and (D-F) Crohn's disease.Magnification, Â400.28.75-36.00],P = 0.002), and CD20 + B cells (15.00 [IQR 7.50-35.50]vs 8.50 [IQR 4.00-12.25],P = 0.008) per area (200 μm diameter) in the mucosal lamina propria than CD, while the FOXP3 + /CD3 + T cell ratio was not statistically different.The mean number of MPO + cells within the lamina propria of the mucosa had an increasing trend in intestinal BD than in CD, although the difference was not statistically significant.
with intestinal BD tended to have a higher cumulative probability of surgical treatment in the first 5 years of GI involvement onset than CD.More studies are needed to validate the discrepancy.Establishing efficient diagnostic criteria for intestinal BD remains challenging, and diagnostic criteria based on BD-related clinical manifestations often lead to delayed diagnosis.A diagnostic model based on clinical manifestations, particularly recurrent oral ulcers, and endoscopy offers the possibility of early diagnosis of intestinal BD. 8,9 Zou et al verified the reliability of a diagnostic model in a group of Chinese BD Histological features of ileocecal ulcer in intestinal Behçet's disease (BD) and Crohn's disease (CD).