Detection Rates of Non-Cavitary Epithelioid Cell Granuloma by Gastrointestinal Biopsy in Patients with Treatment-Naïve Crohn’s Disease

Crohn’s disease (CD) is a chronic, progressive inflammatory bowel disease that may cause irreversible bowel injury or disfunction [1]. The diagnosis of CD at earlier stages of disease is important as early intervention and treatment can prevent disease progression and improve long-term prognosis [1‒3].

The diagnosis of CD is based on physical examination, endoscopic evaluation, imaging, and histological examination [4]. The diagnosis of CD is relatively straightforward in patients with advanced stages of disease as characteristic findings such as longitudinal ulcers and cobblestone appearances are frequently observed on endoscopic examination of the gastrointestinal (GI) tract; however, the diagnosis is often clinically challenging in cases without longitudinal ulcers or cobblestone appearances. In cases with earlier stages of CD which lack longitudinal ulcers or cobblestone appearances, histopathologic evidence of non-cavitary epithelioid cell granuloma is considered one of the most important findings for a definitive diagnosis of CD. The detection of non-cavitary epithelioid cell granuloma is also important for differentiating CD from other forms of enterocolitis including intestinal tuberculosis [5, 6]. Furthermore, a recent study reported a trend toward a more aggressive disease phenotype and poorer prognosis in granuloma-positive cases of CD [7]. Accordingly, the detection of granuloma at the time of initial diagnosis is important for the accurate prediction of prognosis.

In Japan, the diagnosis of CD is made according to the official Japanese diagnostic criteria for CD [4] (Table 1). A unique feature of these diagnostic criteria is that CD can be diagnosed according to the presence of granuloma on histological examination of GI biopsies in combination with other secondary findings, even in the absence of advanced lesions such as longitudinal ulcers or cobblestone appearances. A recent study reported that 176 (30.4%) out of the 579 cases of newly diagnosed CD in Japan were diagnosed based on the detection of non-cavitary epithelioid cell granuloma on histological examination of GI biopsies [8].

Although several previous studies have evaluated the rate of granuloma detection on histological examination of GI biopsies in CD according to the number of cases [9‒14], few reports have examined the rate of granuloma detection per number of biopsies [5]. Further, the effects of using serial or step sections during histological examination of GI biopsy specimens on the rate of granuloma detection remain unknown.

The present study therefore evaluated the rate of granuloma detection on histological examination of GI biopsies in treatment-naïve CD and determined the effectiveness of histological examination of step sections of biopsy specimens when assessing the presence of granuloma. We also describe the characteristics of cases in which the detection of granuloma was essential for a diagnosis of CD.

This single-center retrospective cohort study comprised all consecutive patients newly diagnosed with CD at our hospital between April 2017 and March 2023. The diagnosis of CD was made based on physical examination, endoscopic findings, surgical findings, and histopathological findings according to the official Japanese diagnostic criteria for CD [4] (Table 1). During endoscopic examinations, biopsy specimens were taken from lesions including aphthae, erosions, ulcers, and areas of mucosa with cobblestone appearance. One biopsy specimen was obtained from each of these endoscopic findings. The models of biopsy forceps used in this study were Radial Jaw 4 (Boston Scientific Corporation, USA) for the upper GI and EndoJaw (Olympus Medical Systems Corp., JAPAN) for the lower GI, respectively. Initially, one section stained with hematoxylin and eosin (HE) per biopsy was examined to assess for the presence of non-cavitary epithelioid cell granuloma. A non-cavitary epithelioid cell granuloma was defined as a collection of epithelioid cells with or without accompanying multinucleated giant cells and without caseating necrosis or foreign material. In cases where granulomas were not detected, step sections (10 slides at 30 μm intervals) were additionally prepared and examined to assess for the presence of granuloma. All biopsies were reviewed by two experienced GI pathologists. This study was approved by the Institutional Ethics Committee (approval number: 2021-02-037). Informed consent was obtained from all patients.

The presence or absence of primary and secondary findings of CD according to the Japanese diagnostic criteria for CD was determined by retrospective analysis of medical records. The proportion of patients with each finding was then calculated. The proportions of patients with diagnostic patterns 1, 2, or 3 according to the Japanese diagnostic criteria for CD were also calculated. In diagnosis pattern 2, the detection of non-cavitary epithelioid cell granuloma is essential for a diagnosis of CD. We therefore also investigated the clinical features of patients diagnosed with pattern 2 CD.

The positive rate of granuloma in the total number of CD cases was calculated. The comparison of the patient characteristics between the granuloma-positive cases and the granuloma-negative cases was performed. Furthermore, the frequency of granuloma detection in the total number of endoscopic biopsies was examined. Regarding the granuloma detection rate in the number of endoscopic biopsies, the improvement of the granuloma detection frequency after the step sectioning examination was evaluated.

Statistical analyses were performed using JMP Pro 13.2.1 (SAS Institute Inc., Cary, NC, USA). Categorical variables were presented as numbers and percentages, and continuous variables were presented as median values and ranges. Data between the two groups were compared using the χ² test for categorical variables and the Mann-Whitney U test for numerical variables. p values <0.05 were considered statistically significant.

A total of 30 patients with newly diagnosed CD were included in the present study. The baseline characteristics of included patients are summarized in Table 2. Study participants comprised 20 males (66.7%) and 10 females (33.3%). The median age at diagnosis was 24 years. All patients underwent both esophagogastroduodenoscopy (EGD) and ileocolonoscopy, with GI biopsies performed in 29 patients. Small bowel examinations were performed using several modalities on a case-by-case basis. Regarding disease location, 13.3%, 16.7%, and 70.0% of cases had ileitis, colitis, and ileocolitis, respectively. Regarding disease behavior, 83.4%, 13.3%, and 3.3% of cases had inflammatory type, structuring type, and penetrating type disease, respectively. Fifteen of the 30 cases (50%) had perianal disease. Two cases (6.7%) underwent intestinal resection for small bowel strictures and were diagnosed with CD based on histological examination of resected specimens.

The proportions of study participants with primary and secondary findings of CD according to the Japanese diagnosis criteria for CD are shown in Table 3. Regarding primary findings, longitudinal ulcers, cobblestone appearance, and non-cavitary epithelioid cell granuloma were present in 27 cases (90.0%), 12 cases (40.0%), and 19 cases (63.3%), respectively. In terms of secondary findings, extensive irregular to round ulcers or aphthae, characteristic anorectal lesions, and characteristic gastric and duodenal lesions were present in 16 cases (53.3%), 15 cases (50.0%), and 23 cases (76.7%), respectively. Twenty-seven of the 30 cases (90.0%) were diagnosed with CD according to diagnosis pattern 1, and 3 cases (10.0%) were diagnosed according to diagnosis pattern 2. Diagnosis pattern 3 was not observed in the present study.

The present study comprised three cases diagnosed with CD according to diagnostic pattern 2 in which the detection of non-cavitary epithelioid granuloma is necessary for a definitive diagnosis of CD. The clinical features of each case were as follows.

A 15-year-old male was diagnosed with ileocolonic CD (Fig. 1). EGD revealed small, round-shaped ulcers in the esophagus and mucosal redness and small erosions in the stomach and duodenum. Ileocolonoscopy revealed multiple small ulcers in the ileum and edema and erosions in the rectum. Ten biopsy specimens were taken from the ileum and colon. Granuloma was detected in an ileal specimen only after histological examination of step sections. CD was diagnosed according to diagnostic pattern 2.

A 19-year-old male was diagnosed with ileocolonic CD with perianal disease (Fig. 2). EGD revealed multiple aphthoid erosions in the stomach and duodenum. Ileocolonoscopy revealed multiple aphthoid erosions in the ileum, colon, and rectum. Eight GI biopsy specimens were taken from the ileum and colon. Granulomas were detected on histological examination of a biopsy specimen from the ascending colon only after examination of step sections. CD was diagnosed according to diagnostic pattern 2.

A 15-year-old male was diagnosed with colonic CD (Fig. 3). Ileocolonoscopy revealed multiple erosions and irregular-shaped ulcers in the colon. Twelve biopsy specimens were taken from the colon. Granulomas were detected on histological examination of a biopsy specimen from the ascending colon. CD was diagnosed according to diagnostic pattern 2.

The proportion of patients with non-cavitary epithelioid cell granuloma among patients newly diagnosed with CD was 63.3% (19 out of 30 cases; Table 3). Among these 19 cases, granuloma was detected on histological examination of biopsy specimens in 17 cases and of surgical specimens in 2 cases. There was no significant difference in the patient characteristics between granuloma-positive cases and granuloma-negative cases (online suppl. Table 1; for all online suppl. material, see https://doi.org/10.1159/000533479).

Among all patients with a new diagnosis of CD, GI biopsies were performed in 29 cases. Accordingly, the rate of granuloma detection by GI biopsy per case was 58.6% (17 out of 29 cases). The anatomic locations and the endoscopic images where the granulomas were detected are shown in Table 4. The locations were scattered among the stomach, various parts of the large intestine, and the ileum, and no consistent trend was observed. Regarding endoscopic images, of the total 33 granulomas detected, 17 (51.5%) were detected from longitudinal ulcers, 12 (36.4%) from aphthae or erosions, and 4 (12.1%) from irregular- or round-shaped ulcers.

The rates of granuloma detection according to the number of biopsies taken are summarized in Table 5. EGD with biopsy was performed in 16 cases, with a total of 63 biopsies taken. The rate of granuloma detection on histological examination of upper GI tract biopsies was 0% on initial examination and 3.2% following examination of step sections. Ileocolonoscopy with biopsy was performed in 29 cases, with a total of 221 biopsies taken. The rate of granuloma detection on histological examination of lower GI tract biopsies was 7.7% on initial examination and 14.0% following examination of step sections. Regarding the lower GI tract, the rate of granuloma detection was significantly improved after histological examination of step sections (7.7 vs. 14.0%; p < 0.05). A total of 284 GI biopsies were taken in 29 cases. The total number of biopsy specimens found to contain granuloma on histological examination was 17 (6.0%) on initial examination and 33 (11.6%) following histological examination of step sections. The rate of granuloma detection was significantly improved following histological examination of step sections compared with the initial examination (p < 0.05).

The present study evaluated the rate of non-cavitary epithelioid cell granuloma detection by histological examination of GI biopsies in patients with treatment-naïve CD. The findings of this study are noteworthy for the following two respects. First, this is a rare study that calculated the rate of granuloma detection according to the number of biopsies taken. Second, this is the first study to compare the rate of granuloma detection by conventional histological examination and examination of step sections of GI biopsies, with a significantly higher rate of granuloma detection following examination of step sections. Third, the analysis of endoscopic images of granuloma-detected sites showed a relatively high detection rate from aphthae and erosions, indicating their importance as biopsy target sites.

We used the official Japanese diagnostic criteria for CD. Using these diagnostic criteria, three diagnosis patterns can be used to make a definitive diagnosis of CD, as shown in Table 1. Hisabe et al. [8] studied the proportion of patients with each diagnostic pattern among Japanese patients with a new diagnosis of CD and found that 87.4% of patients were diagnosed according to pattern 1, 30.4% according to pattern 2, and 7.1% according to pattern 3. In our study, 90.0% of cases were diagnosed according to pattern 1. This finding corroborates the results of Hisabe et al. The proportion of patients with longitudinal ulcers was 81.7% in the study by Hisabe et al. and 90.0% in the present study. These results indicate that longitudinal ulcers may be the most frequent characteristic finding when diagnosing treatment-naïve CD.

On the other hand, the proportion of cases diagnosed according to pattern 2 were 30.4% in the study by Hisabe et al. and 10.0% in the present study. Notably, both longitudinal ulcers and cobblestone appearance were absent in cases diagnosed according to pattern 2, and the detection of a non-cavitary epithelioid cell granuloma was required for the diagnosis of CD. It is also worth noting that most cases diagnosed according to pattern 2 had early stages of disease. These findings indicate the improvement of granuloma detection rates in GI biopsies is important to ensuring an accurate and definitive diagnosis of CD can be made at an early stage of disease. The findings of the present study also demonstrate the clinical characteristics of the 3 cases in which the detection of granuloma was essential for a diagnosis of CD. All 3 cases were considered to have early-stage disease, which predominantly consisted of aphthous erosions and small ulcers in the absence of longitudinal ulcers or cobblestone appearances. As CD is a chronic progressive disease, diagnosis at an early stage of disease stages is particularly important for improving the natural history and long-term prognosis of CD. In cases with early-stage CD with aphthous erosions and small ulcers, it is important to take multiple biopsies from the GI tract and perform histological examinations using multiple sections of biopsy specimens to ensure the detection of granuloma.

Several previous studies have evaluated the rate of granuloma detection on histological examination of GI biopsies from patients with CD. We summarized the main results of five studies in which the methods of biopsy and pathological examination were precisely described (Table 6) [5, 10, 11, 14, 15]. In these reports, the rate of granuloma detection following histological examination of GI biopsies ranged from 34% to 67%. Among these five studies, three studies reported that the rate of granuloma detection was 60% or greater [5, 10, 11]. It should be noted that multiple biopsies and histological examination of serial sections were intentionally performed in these three studies. A similar detection rate to these three reports was observed in the present study, highlighting the importance of multiple biopsies and additional histological examination of serial or step sections.

The number of GI biopsies that should be obtained for the detection of granuloma during the initial diagnosis of CD is currently unclear. A limited number of previous studies have examined the rate of granuloma detection according to the number of biopsies taken. Ye et al. evaluated the rate of granuloma detection in 164 biopsies in a study of 52 patients with CD. In this study, an average of six biopsies were taken from each affected lesion of the lower GI tract, and histological examination of serial sections was performed to assess for the presence of granuloma, with a reported granuloma detection rate per biopsy of 39.6% (65 granulomas/164 biopsies). It should be emphasized that this report used a considerably more detailed series of examinations than conventionally used in clinical practice. The benefit of performing histological examinations using serial or step sections in detecting granuloma in GI biopsy specimens is also unclear. In the present study, we compared the rate of granuloma detection between the conventional histological examinations and examinations using step sections. In the present study, the rate of granuloma detection according to the number of biopsies taken was 0% from the upper GI tract and 7.7% from the lower GI tract before histological examination of step sections. However, the detection rates improved to 3.2% and 14.0% after histological examination of step sections of biopsies from the upper and lower GI tracts, respectively. This finding indicates that the rate of granuloma detection was significantly improved after histological examination of step sections. The findings of Ye et al. report and the present study highlight the importance of performing multiple GI biopsies and pathological examinations of serial or step sections when assessing the presence of granuloma.

The endoscopic images where the granulomas were detected are another important issue. It should be noted that granulomas were frequently detected from longitudinal ulcers. Here, we also emphasize that granulomas were detected at a relatively high frequency from aphthae and erosions. The results suggest that intentional biopsy from aphthae and erosions is highly significant for the detection of granulomas, especially in the diagnosis of early CD that lacks ulcerative lesions.

The limitations of the present study include the small number of cases from a single center and the variability in the methods used to obtain GI biopsies given its retrospective study design. Prospective studies using uniform methods are required to determine the appropriate number of biopsies required and clarify the appropriate site of biopsy when assessing for the presence of granuloma.

In conclusion, the present study evaluated the rate of granuloma detection from histological examination of GI biopsies during the initial diagnosis of treatment-naïve CD. The rate of granuloma detection per case was approximately 60% in the present study. The rate of granuloma detection per number of biopsies performed was only 6.0% before histological examination of step sections; however, assessing for the presence of granuloma using histological examination of step sections significantly improved the detection rate to 11.6%. These results indicate that performing multiple GI biopsies and assessing for granuloma using histological examination of serial or step sections is necessary when making an initial diagnosis of CD.

All procedures followed have been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. The study design was observational. The study was approved by the Ethics Committee of the Institutional Review Board of Tohoku Medical and Pharmaceutical University School of Medicine. Approved IRB number is 2021-02-037. The study was performed in accordance with the ethics guidelines in Japan. For adult participants, written informed consent was obtained from the patient for publication of the details of their medical case and any accompanying images. For underage participants, written informed consent was obtained from the parent/legal guardian of the patient for publication of the details of their medical case and any accompanying images. Patients who expressed unwillingness to participate in this study were excluded.

The authors declare no conflicts of interest.

The authors received no financial support for the research.

Katsuya Endo established the study design and was involved in designing the original study protocol, collecting data, and writing the manuscript. Yoko Kawakami, Yuki Yoshino, Shiho Kondo, Daisuke Fukushi, and Kazuhiro Murakami participated in data collection. Atsuko Takasu, Takayuki Kogure, Morihisa Hirota, and Kennichi Satoh contributed in discussing and critically reviewing the manuscript.