⁎4479 Is routine mucosal biopsy of value in patients with diarrhoea and normal colonoscopy in an open access setting?

*4478 ENDOSCOPIC PREDICTION OF EARLY POSTSURGICAL RECURRENCE IN PATIENTS WITH CROHN’S DISEASE. Toshifumi Ashida, Atsuo Maemoto, Takanori Fujiki, Tohru Kohno, Akitoshi Kakisaka, Masaki Taruishi, Mikihiro Fujiya, Fumika Orii, Kaori Fujiya, Jiro Watari, Yusuke Saitoh, Yutaka Kohgo, Asahikawa Med Coll, Asahikawa, Japan. Background/Aim: High relapse rate of re-stenosis at the site of intestinal anastomosis in postsurgical patients with Crohn’s disease is reported from several study groups. These studies revealed that appearance of endoscopic changes at the site of anastomosis frequently preceded to production of typical symptoms. However, it is still unclear when endoscopic observation should be performed, or how we can detect risky patients of earlier recurrence of anastomotic re-stenosis. To find out the predictive markers for earlier recurrence, we have prospectively observed the sequential changes of endoscopic features appeared at the site of anastomosis, from immediately after surgery to recurrence of typical stenotic lesions. Patients/Methods: Twenty-nine patients with Crohn’s disease who underwent intestinal/colonic resection in Asahikawa Medical College Hospital from 1990 to 1999 were subjected in this study. All the patients had ileocolonic or colo-colonic anastomosis, which were accessible by colonoscopy. Endoscopic observation and combined endoscopic retrograde ileography (ERIG) were performed at 1,6, and 12 months after surgery. One year after surgery, these observations were repeated at once a year. Results: At 1 month after surgery, 24.9% (7/29) patients already had small aphthous ulcer(s) at the site of anastomosis. These lesions were not disappeared, then incidence of relapsing these lesions were increased ( 6 months; 48.3% (14/29) 12 months; 65.5% (19/29)). Stenotic lesions due to multiple or longitudinal ulceration of anastomosis were detected from 2 years after surgery, in the patients of ulcer(+) at 12 months (31.6% (6/19)). However, no patients developed stenotic lesions within 5 years after surgery among 10 patients of ulcer (-) at 12 months. Significant statistical correlation in Logrank test was detected between ulcer (+) at 12 months and development of stenosis, or requirement of re-operation. Other factors such as disease duration, administration of 5-ASA, nocternal nutritional supprement with elemental diet, or methods of anastomosis did not correlate with postsurgical relapse rate in our patients. Conclusion: Developement of aphthous or small ulcer(s) within a year after surgery at the site of anastomosis is a risk factor of earlier re-stenosis or re-operation. Endoscopic evaluation of anastomotic areas at 12 months after surgery is necessary in management for postsurgical CD patients. *4479 IS ROUTINE MUCOSAL BIOPSY OF VALUE IN PATIENTS WITH DIARRHOEA AND NORMAL COLONOSCOPY IN AN OPEN ACCESS SETTING? Ian F. Yusoff, Neville E. Hoffman, Donald G. Ormonde, Sir Charles Gairdner Hosp, Perth, WA, Australia. BACKGROUND: Routine mucosal biopsy in patients undergoing colonoscopy for diarrhoea, in whom macroscopic examination is normal, remains controversial and practice varies widely without clear guidelines. Reported rates of clinically significant microscopic abnormalities vary from 2-27%.It is unclear if ileal biopsy adds anything to colonic biopsy alone. OBJECTIVES: We sought to evaluate the diagnostic yield of colonic and ileal mucosal biopsy in patients undergoing colonoscopy for diarrhoea in whom the macroscopic examination was normal. METHODS: We retrospectively reviewed all colonoscopies performed over a nine year period in a tertiary referral centre with an open access endoscopy service. Cases were selected where the sole indication for colonoscopy was diarrhoea, the musosa was macroscopically normal (other than diverticulosis) and biopsies were performed. Cases were excluded if the examination was inadequate. The histopathology reports of the selected cases were then reviewed. RESULTS: 362 cases were identified. Colonoscopy and biopsy was normal in 260 patients.Ileal biopsies were performed (in addition to colonic biopsies) in 158 cases, none of which revealed clinically significant abnormalities. Clinically significant histological findings were present in 18 cases (5%). Findings included collagenous colitis (5 cases), lymphocytic colitis (1 case), possible lymphocytic colitis (1 case), possible collagenous colitis (1 case), inflammatory bowel disease (2 cases), melanosis coli (2 cases) and significant eosinophil mucosal infiltration (6 cases). 28 patients (8%) had minor histological abnormalities with no specific diagnostic features. The diagnostic yield was highest in patients above 60 years old, where 10% had clinically significant histological abnormalities. All patients with collagenous colitis were female and only 1 was less than 60 years old. CONCLUSIONS: When colonoscopy is normal in patients with diarrhoea, routine colonic biopsy identifies significant pathology in 5% of cases. The diagnostic yield is highest in patients over 60 years old. Routine ileal biopsy is unhelpful.

*4480 PREDICTION OF THE CLINICAL COURSE OF BEHCET’S COLITIS ACCORDING TO MACROSCOPIC-TYPE CLASSIFICATION BY COLONOSCOPY. Joo Sung Kim, Hae Moon, Il Ju Choi, Hyun Chae Jung, In Sung Song, Chung Yong Kim, Dept of Internal Medicine, Soul National Univ, Seoul, South Korea; Dept of Internal Medicine, Seoul National Univ, Seoul, South Korea; Dept of Internal Medicine, Seoul, South Korea. Background : Behcet’s colitis has denoted predominant gastrointestinal symptoms and colonic ulcerative lesions documented by objective measures in patients with Behcet’s disease. The lesions of Behcet’s colitis showed aphthoid or punched-out ulceration, but the macroscopic types of lesions have not been defined. To predict the clinical outcome of patients with Behcet’s colitis according to colonoscopic findings, we classified the characteristic colonic ulcers and evaluated the efficacy of medical treatment and calculated the operation rates and the recurrence rates according to the macroscopic types. Methods :We retrospectively reviewed the medical records and colonoscopic photographs of 50 patients with Behcet’s colitis. The colonic lesions were examined by colonoscopy in patients with Behcet’s disease and gastrointestinal symptoms. Colonoscopic findings were categorized according to three types: volcano-, geographic- and aphthous-type. Volcano-type ulcers were defined as well-demarcated deep-penetrating ulcers with nodular margins or converging folds. Geographic-type ulcers were defined as being shallow various-shaped ulcers with sharp edges, and aphthous-type ulcers as small,round or ovoid, punchedout shallow ulcers. The efficacy of medical treatment was assessed by follow-up colonoscopy or double-contrast barium enema 4 to 8 weeks after treatment and recurrence of the lesions was evaluated during the follow-up periods. Cumulative operation rates were obtained by the KaplanMeier method. Results : Macroscopic types of colonic ulcers revealed 25 out of 50 (50%) patients with volcano-type, 11 (22%)with geographic-type, and 14 (28%) with aphthoustype lesions. Complete remission rates by medical treatment and operation rates in volcano-type ulcerations were 6 of 25 (24%) and 13 of 25 (52%), geographic-type 8 of 11 (73%) and 1 of 11 (9%), and aphthous-type 9 of 14 (64%) and 2 of 14 (14%), respectively. The recurrence rates in volcano-type ulcerations were 9 of 19 (47%), geographic-type 1 of 9 (11%) and aphthous-type 1 of 11 (9%). Conclusion : Volcano-type ulcerations in Behcet’s colitis showed less favorable response to medical treatment and more episodes of operation and recurrence than geographic- and aphthous-type ulcerations. *4481 THE COLON SINGLE-STRIPE SIGN: DIAGNOSTIC IMPLICATIONS FOR ISCHEMIC COLITIS. Gary R. Zuckerman, Raphael B. Merriman, Chandra Prakash, Ray E. Clouse, Washington Univ Sch of Medicine, St. Louis, MO. A unique endoscopic appearance of a single linear ulcer has been noted at colonoscopy in a group of patients with acute abdominal symptoms, often in the clinical setting of an ischemic event. Twelve patients (8F/4M, mean 75.2 yr) with endoscopic evidence of only a single ulcerated band running along the longitudinal axis of the colon (colon single-stripe sign, CSSS) were further studied. The CSSS was >8 cm in length and isolated to a segment of the left colon. To better understand the etiology of this lesion, the clinical characteristics and course of CSSS were compared with 26 cases of ischemic colitis with large circumferential ulcers (15F/11M, mean 63.7 yr) and 58 consecutive patients with other forms of non-ischemic colitis including IBD (25 cases), infectious colitis (13), radiation colitis (4), amyloid (1), nonspecific colitis (15)(36F/22M, mean 48.1 yr). RESULTS: Both CSSS and ischemic colitis groups were older than the non-ischemic colitis group. Unequivocal evidence of a preceding ischemic event was noted with comparable frequency (p=0.2) in the CSSS and ischemic colitis groups, significantly different from the non-ischemic colitis group (p<0.01, see table). A modified stripe configuration (longitudinal stripe extending from the distal and/or proximal end of a circumferential ulceration) was seen in 15% of the ischemic colitis group. None of the CSSS patients required surgical intervention while 5 (19%) from the ischemic colitis group underwent surgical exploration (p=0.04). Eight patients (31%) in the ischemic colitis group died; there were 2 deaths (3.4%) in the non-ischemic colitis group and none in the CSSS group (p<0.01). CONCLUSIONS: 1. The colon single-stripe sign appears to represent an endoscopic appearance for ischemic colitis and characterizes a milder form in the clinical spectrum of ischemic colitis. 2. This endoscopic sign may reflect colonic ischemic injury along a vascular watershed territory.

Ischemic event Abdominal pain Gross bleeding Days to discharge (mean)

CSSS

Ischemic colitis

Non-ischemic colitis

10/12 5/12 9/12 2.9

15/26 18/26 17/26 9.5

4/58* 30/58 21/58* 5.2

*p<0.05 across groups

AB146

GASTROINTESTINAL ENDOSCOPY

VOLUME 51, NO. 4, PART 2, 2000