Ho Khek Yu MD, FAMS, FRACP, Albert Low FRCR, MMed,¹ Priyanthi Kumarasinghe MBBS, FRCPA,² Jacqueline Hwang FRCPA, FAMS ²
Department of Medicine, National University Hospital, Singapore
¹ Department of Diagnostic Radiology, SGH
² Department of Pathology, SGH

* Presented at the SGH Clinio-pathological Conference on 5 March 2005.

PROTOCOL

Mr A is a 57-year-old Malay patient with a known history of ischaemic heart disease, benign prostatic hypertrophy, and iritis. He had been on specialist follow-up for the past 5 years and was reportedly also treated medically for what he was told to be tuberculosis of the bowel. His surgical history included an appendicectomy in 1977 and a laparotomy for intestinal obstruction in 1985, the cause of which was not available to us. He has no significant family and medical history.

Prior to his admission to the Singapore General Hospital (SGH), Mr A presented to another hospital for a primary complaint of per rectal (PR) bleeding on 21 March 2004. A flexible endoscopy, done till 90cm from the anal verge, revealed an irregular and suspicious stenotic lesion. His carcinoembryonic antigen (CEA) level was normal at 1.6ng/ml. A computer tomogram (CT) scan performed on 24 May 2004 showed no evidence of colonic carcinoma or distant metastasis. However, there was mural thickening and stricturing at the ileocolic region, which appeared to have been stable since the last CT scan performed on 3 October 2002. Review of an enteroclysis study done on 29 December 2003 showed correlative findings suggestive of extrinsic indentation at the medial aspect of the terminal ileum. The histopathological report of the biopsy was that of a hyperplastic polyp. A repeat colonoscopy performed at the same hospital on 15 July 2004 again showed suspicious mucosa with nodules at the previous ileocolic anastomosis. A repeat biopsy was reported as that of chronic inflammation with hypertrophic glands. No malignancy was seen.

Subsequently, Mr A presented to SGH with a further episode of PR bleeding. He was otherwise asymptomatic. He had no loss of weight or appetite. He did not have any abdominal pain, was able to move his bowels regularly but had mild obstructive symptoms. Physical examination was essentially normal. He was well, alert, and afebrile. His vital signs were stable. Heart sounds were normal and air-entry to lungs was good. His abdomen was soft, non-tender, non-distended and there was no palpable mass. PR examination and proctoscopy were unremarkable.

Laboratory investigations were largely unremarkable. His haemoglobin was 11.7g/dl, total white cell count 6.8´ 10(9)/L and the CEA was at 1.3ng/ml. Other haematological and biochemical parameters were well within normal limits. His chest X-ray was also normal. A repeat colonoscopy done on 30 July 2004 demonstrated inflammation, stricture and ulceration at the ileocolic anastomosis, which was biopsied. On 5 August 2004, he underwent a right hemicolectomy and adhesioloysis. Intra-operative findings were that of an anastomotic stricture with a thickened mass in the mesentery next to it. There were dense adhesions but the rest of the colon was grossly normal. Post-operatively, he was well and was eventually discharged on 12 August 2004, the seventh post-operative day.

Ho Khek Yu

In summary, Mr A is a 57-year-old Malay man who presented with per rectal (PR) bleeding and mild obstructive symptoms. He had significant past medical history of ischaemic heart disease. He was also reportedly suffering from iritis and had been medically treated for what he was told to be tuberculosis of the bowel; the details of which were not available to us. Mr A had also undergone 2 surgeries in the past: an appendicetomy and a laparotomy for intestinal obstruction. The operative and histologic findings for both the surgeries were not known to us.

Based on the above, I would like to speculate on a possible unifying diagnosis of Crohn’s disease for the following reasons. He had a long history of intestinal problems and was on specialist follow-up for at least 5 years. Although not distinctly indicative, his history was suggestive of a long-standing gastrointestinal disorder, such as Crohn’s disease. Mr A had several conditions which were also consistent with the clinical manifestations of Crohn’s disease. These included ileitis that might have mimicked the appendicitis; ileitis/ileal stricture that might have caused the intestinal obstruction; and ileocolitis that might have been mistreated as intestinal tuberculosis. In addition, Mr A had iritis, a condition prevalent in 5% of patients with inflammatory bowel disease.

Prior to his admission to SGH, Mr A presented to another hospital for a primary complaint of PR bleeding. A flexible endoscopy showed an irregular and suspicious stenotic lesion at 90cm from the anal verge. The histopathological report of a biopsy which had been taken showed the presence of a hyperplastic polyp only. A repeat colonoscopy performed at the same hospital 6 months later revealed nodular mucosa at the previous ileocolic anastomosis, which on histopathological examination again demonstrated chronic inflammation with hypertrophic glands, but no evidence of malignancy. Mr A’s carcinoembryonic antigen level was normal. A subsequent computed tomography (CT) scan showed the only abnormality of a mural thickening and stricturing at the ileocolic region, which had apparently already been noted during a CT scan performed 15 months previously. The lesion appeared to have remained stable since the last scan. An enteroclysis showed correlative findings suggestive extrinsic indentation as opposed to intrinsic growth at the medial aspect of the terminal ileum. All the above findings suggested the unlikelihood of malignancy as the cause of his ileocolonic lesion.

When Mr A subsequently presented to SGH with a further episode of PR bleeding, he was well, and afebrile. Despite having mild obstructive symptoms, he was not experiencing any abdominal pain and his bowels were regular. Mr A also did not show any loss of weight or appetite. His vital signs were all stable. Abdominal palpation detected no abnormality. In addition, PR examination and proctoscopy yielded no remarkable finding. Laboratory investigations were largely unremarkable. Mr A’s haemoglobin level and total white cell count were well within normal limits. His chest X-ray was normal. The above findings were not supportive of infection, such as tuberculosis, Salmonella, or Yersinia, as a cause of his presentation.

Colonoscopy was repeated at SGH. This time, features of active inflammation including stricture and ulceration were detected at the ileocolic anastomosis. On this basis and the persistence of his symptoms, Mr A underwent a right hemicolectomy and adhesioloysis. Intra-operative findings were that of an anastomotic stricture with a thickened mass in the mesentery next to it. Dense adhesions were found there but the rest of the colon was grossly normal.

My main differential diagnosis is Crohn’s disease with acute exacerbation for those reasons I have mentioned. My differential diagnoses include neoplastic lesions, such as colon/ileal adenocarcinoma and lymphoma; and infective lesions, such as gut tuberculosis; and bacterial ileocolitis, such as Salmonella and Yersinia. All these differential diagnoses could give rise to those operative findings as described above. It may be noted that long-standing Crohn’s disease is known to predispose to cancer of both small intestine and colon. In such cases, the cancer usually arises in areas of chronic disease.

Albert Low

CT abdomen and pelvis from 3 Oct 2002 and 24 May 2004 were reviewed (Fig. 1). Both showed concentric mural thickening at the neo-ileocolic junction and neo-terminal ileum, of homogeneous attenuation with absence of mural stratification, associated with luminal narrowing. The extent and morphology of mural thickening was stable between the scans. In addition, there was proliferation of perienteric fat of increased attenuation with perienteric inflammatory strands. This resulted in abnormally wide separation of the small bowel. There was no clearly tumourous mesenteric mass. There were several stable small volume mesenteric lymph nodes, which were non-specific.

Fig. 1. CT scan on 24 May 2004 showed concentric mural thickening of the neo-terminal ileum of homogeneous attenuation (arrow) associated with luminal narrowing (Fig. 1a). There had been no interval change since the CT scan on 3 Oct 2002 (Fig. 1b). CT scan on 24 May 2004 in an adjacent section shows proliferation of perienteric fat of with perienteric inflammatory strands (long arrow) (Fig. 1c). This resulted in abnormally wide separation (arrowheads) of the neoterminal ileum from an adjacent loop of normal ileum medially (star), compared with the closely apposed loops of normal small bowel in the left side. There was clearly no tumourous mesenteric mass. A few small volume mesenteric lymph nodes (short arrow) were present. The findings had been stable since the CT scan on 3 Oct 2002 (Fig. 1d).

In between the CT scans, a barium enema (BE) study and a small bowel enema (SBE) were performed on 21 Oct 2002 and 29 Dec 2003, respectively (Figs. 2 and 3). Both showed a neo-terminal ileal stricture. The BE showed deep linear "rose-thorn" ulcers, while the subsequent SBE showed asymmetric straightening at the mesenteric border with completely effaced mucosal folds on this side, possibly with a longitudinal ulcer. More proximally, nodular mucosal fold thickening was present. Ileal disease appeared far more pronounced than colonic disease. The widely-separated small bowel in the right flank correlated with the CT findings.

Fig. 2. Double contrast barium enema on 21 Oct 2002 showed neoterminal ileum stricture delineated by caecal-ileal reflux, with deep linear "rose-thorn" ulcers (arrow and arrowheads) (Figs. 2a and 2b).

Fig. 3. Small bowel enema on 29 Dec 2003 showed neoterminal ileum stricture and asymmetric straightening at the mesenteric border with completely effaced mucosal folds on this side, possibly with a longitudinal ulcer (Fig. 3a). More proximally, nodular mucosal fold thickening is present (Fig. 3b). The widely separated small bowel (double-headed arrows) correlates the CT images in Figs. 1c and 1d.

The findings of the fluoroscopic studies (BE and SBE) were characteristic of Crohn’s disease. Tuberculosis is unlikely to show asymmetrically severe disease at the mesenteric border, and longitudinal ulcers are not a feature. The abnormal separation of the small bowel was shown to be due to fibrofatty mesenteric proliferation, or creeping fat of the mesentery on CT. This finding is almost exclusive to Crohn’s disease due the characteristic transmural extent of disease. The stability of the CT findings over 18 months virtually rules out malignancy and implies chronicity of disease. Further, the homogeneous attenuation of the mural thickening is a known indicator of transmural fibrosis in long-standing Crohn’s disease.

Priyanthi Kumarasinghe

The specimen consisted of an adherent loop of small intestine 47cm long, anastomosed to a length of large intestine 12cm long. Small intestine showed patchy cobblestone areas and focal ulceration proximal to the anastomotic site. There was flattening of the mucosal folds with thickening and fibrosis of the wall up to 1.5cm involving the muscularis propria at the area of the anastomosis. Large intestine showed oedematous mucosal folds and patchy ulceration. No diverticulae were noted. The appendix was not present.

Sections showed focal, patchy ulceration with intervening normal mucosa (Fig. 4). Abnormal areas showed characteristic transmural inflammation (Fig. 5). There were deep fissuring ulcers and abscesses (Fig. 6). Areas, especially those deeper to ulceration showed marked sub-mucosal oedema, fibrosis, lymphangiectasia, and transmural inflammation with prominent lymphoid follicles with the characteristic "string-of-beads" appearance involving all layers of the intestinal wall (Fig. 7). Abnormal mucosa showed evidence of chronicity featuring glandular distortion, crypt shortening, mixed acute and chronic inflammation with basal plasmacytosis (Fig. 8), focal cryptitis and crypt abscesses (Fig. 9) and focal pyloric metaplasia (Fig. 10). Mucosa also showed regenerative and hyperplastic features (Fig. 11). There were enlarged lymph nodes. No granulomata were found in any of the sections examined.

Fig. 4. Abrupt ulceration with intervening normal mucosa (haematoxylin and eosin ´ 100).

Fig. 5. Ulcerated areas with underlying characteristic transmural inflammation (haematoxylin and eosin ´ 100), low power view of the ulcer with deeper wall, inset (haematoxylin and eosin ´ 40).

Fig. 6. Deep fissuring ulcers and abscesses (haematoxylin and eosin ´ 100).

Fig. 7. Hyperplastic lymphoid follicles in the submucosa and muscularis propria (haematoxylin and eosin ´ 200).

Fig. 8. Crypt distortion with mixed acute and chronic inflammation in the ileal mucosa (haematoxylin and eosin ´ 100).

Fig. 9. Cryptitis and crypt distortion (haematoxylin and eosin ´ 400).

Fig. 10. Focal pyloric metaplasia (haematoxylin and eosin ´ 200, inset ´ 400).

Fig. 11. Regenerative and hyperplastic mucosa (haematoxylin and eosin ´ 200).

Previous endoscopic biopsies had shown isolated, non-specific features including chronic inflammation, repair and regenerative features and hyperplastic mucosa at various points of the evolution of the disease. None had shown granulomatous inflammation.

The features of the right hemicolectomy specimen were consistent with Crohn’s disease. The characteristic gross and microsocpic features as demonstrated (patchy and transmural active chronic inflammation with regenerative and metaplastic mucosal changes) enabled a definitive pathological diagnosis at colectomy. A definitive diagnosis of Crohn’s disease is more difficult on endoscopic biopsies as characteristic transmural involvement cannot be demonstrated. Endoscopic biopsies include only the mucosa and at most, muscularis mucosa and superficial submucosa. The diagnosis of Crohn’s disease therefore involves observation of a constellation of mucosal abnormalities of different sites, including the endoscopically normal and abnormal mucosa and good clinico-pathological correlation.

A remarkable feature in this patient was the absence of the granulomatous component. However granulomata have been observed in approximately 35% of endoscopic biopsies and 50 to 60% of resected specimens of Crohn’s disease.¹ Presence of granulomata enhances the diagnostic strength of Crohn’s disease. At the same time, the presence of granulomatous inflammation brings in all other causes, most importantly mycobacterial infection in the tropics, to the differential diagnosis. It is of paramount importance to exclude the possibility of such infections before a definitive diagnosis is made.

The pathological differential diagnosis of Crohns’ diseases on mucosal biopsies also includes ulcerative colitis, medication-induced injury and right-sided diverticular disease prevalent among Asians, particularly the Chinese.^2-4

From the pathologist’s point of view, the approach to the diagnosis of Crohn’s disease in the tropics should be based on excluding the possibility of infections and also mimics common or even unique in the local context.

References

1. Robert ME. Inflammatory disorders of the small intestine. In: Odze RD, Goldblum JR, Crawford JM, editors. Surgical Pathology of the GI tract, Liver, Biliary Tract, and Pancreas. Philadelphia: Saunders, 2004;198.
2. Chan CC, Lo KK, Chung EC, Lo SS, Hon TY. Colonic diverticulosis in Hong Kong: distribution pattern and clinical significance. Clin Radiol 1998; 53:842-4.
3. Wong SK, Ho YH, Leong AP, Seow-Choen F. Clinical behavior of complicated right-sided and left-sided diverticulosis. Dis Colon Rectum 1997; 40:344-8.
4. Lee YS. Diverticular disease of the large bowel in Singapore. An autopsy survey. Dis Colon Rectum 1986; 29:330-5.