<<

Clinical and pathological correlations in inflammatory bowel disease

Reaffirmation of histologic criteria for diagnosis

A decade ago, we1 and others2 emphasized the importance of correlation between clinical and pathological features in inflammatory bowel disease Richard G. Farmer, M.D. (IBD). These studies demonstrated that most cases of IBD could be differentiated into Crohn's disease Department of or ulcerative , although a small number had indecisive features, both clinically and histologi- William A. Hawk, M.D. cally. There were 16 (11%) such patients in our Department of Pathology study.1 Further, despite much emphasis on the non- caseating sarcoid-type for the diagnosis Rupert B. Turnbull, Jr., M.D. of Crohn's disease, our study documented this find-

Associate Emeritus Consultant, Department ing in only 55% of those with Crohn's disease of Colon and Rectal Surgery confirmed by other histologic criteria.1 That the problem of exact histologic differentiation of ulcer- ative colitis and Crohn's disease persists has been recently reaffirmed by Kirsner3 who stated "typical pathological findings notwithstanding, the features in an individual case may be indecisive." He also noted that " are identified in approxi- mately 50% of patients." Because of our continued interest in this subject, and to reevaluate our experience in the histologic confirmation of the diagnosis of IBD, we undertook a study of 366 consecutive surgically resected spec- imens from patients with known IBD clinically. There were 121 patients with a diagnosis of ul- cerative colitis and 245 with Crohn's disease. Path- ological features are listed in the Table. 35

Downloaded from www.ccjm.org on October 1, 2021. For personal use only. All other uses require permission. 36 Cleveland Clinic Quarterly Vol. 46, No. 1

Table. Pathological features of submucosa only is an important differ- and Crohn's disease ential point in ulcerative colitis as con-

Ulcerative coli- ^ . , .. trasted with Crohn's disease. Crohn s disease tis Noncaseating sarcoid-type granu-

Pathological fea- No. of No. of lomas were found in only 49% of the tures cases % cases % cases of Crohn's disease. Particular at- Transmural inflam- 34 28 240 98 tention was focused on finding granu- mation lomas in the resected specimens, and Serositis 33 27 235 96 Granulomas 0 0 122 49 emphasis was placed on the clinical cor- Lymphedema 4 3 215 88 relation as well. This was done also in Fissuring 0 0 186 76 Crypt abscesses 115 95 137 56 our earlier study1 in which we differen- Mucosal dysplasia 56 46 5 2 tiated the clinical features of granu- lomatous and nongranulomatous forms Thus, highly characteristic pathological of Crohn's disease. However, subsequent features correlated with the character- experience has not shown this to be a istic clinical features of the diseases. In reliable clinical differential point. ulcerative colitis, the most characteristic Among the 122 patients with granu- histologic criterion was finding a crypt lomas in this series, 46 cases were from abscess in 95% of cases. Correlating with surgically resected specimens from pa- this finding were the mucosal changes tients with Crohn's disease who had un- grossly, the relative lack of transmural dergone more than one resection. Of involvement, and the complete absence these 46 cases, granulomas were found of granulomas. Histologically, these fea- in 26 cases, in specimens from both sur- tures correlated with clinical features gical procedures; 14 with granulomas in characteristic of ulcerative colitis. the first surgical specimen, but not in subsequent resected specimens; and six For patients with Crohn's disease, with granulomas in the second specimen transmural with serositis not having had granulomas in the first was the most common pathological fea- specimen. There was no notable clinical ture in 96% of the cases. Lymphedema correlation with the presence or absence and fissuring were the next most com- of granulomas as a histologic character- mon pathological features in 88% and istic of Crohn's disease. Based on timing 76% of the cases respectively. These of the resected specimen, it would ap- findings correlated well with the char- pear that a granuloma has a "life his- acteristic features of "cobblestone mu- tory" of about 3 months based on timing cosa" and the propensity for develop- of elected resections in this group of ment of fistulae. Although crypt ab- patients with Crohn's disease. scesses were found in 56% of the cases with Crohn's disease, they were often in During this period, 12 patients were areas of mucosa with relatively normal found to have carcinomas, one involving mucosa adjacent. This is in striking con- the , eight in the colon, trast to the mucosa of ulcerative colitis, and three in the . Nine of these which is diffusely and uniformly in- patients had ulcerative colitis; in seven volved. Therefore, although the crypt of these the carcinoma was in the prox- abscess is a nonspecific finding, its pres- imal colon, and in two instances the ence in association with diffuse and uni- carcinoma was in the rectum. Two of form involvement of the mucosa and the patients had Crohn's disease, with a

Downloaded from www.ccjm.org on October 1, 2021. For personal use only. All other uses require permission. Summer 1979 Inflammatory bowel disease 37 carcinoma being found once in the small and in prospective studies to define clin- intestine and once in the rectum. One ical criteria and correlate them with patient had ulcerative proctosigmoiditis mucosal dysplasia and subsequent de- and a rectal carcinoma subsequently de- velopment of carcinoma. This study veloped. For the patients with ulcerative does reaffirm that well-defined histo- colitis, the finding of the carcinomas logic criteria are present in most cases correlated well with the presence of mu- of ulcerative colitis and in Crohn's dis- cosal dysplasia, as emphasized by Mor- ease, and that by careful clinical and son.4 Awareness of the potential impor- pathological observation, differentia- tance of mucosal dysplasia in the long- tion between the two diseases can be term surveillance of patients with ulcer- made in more than 90% of cases. ative colitis is being increasingly recog- nized. ' References In summary, this study reaffirms the 1. Farmer RG, Hawk WA, Turnbull RB Jr: ability to differentiate the clinical and Regional of the colon; a clinical and histologic features of Crohn's disease pathologic comparison with ulcerative colitis. Am J Dig Dis 13: 501-514, 1968. and ulcerative colitis in the vast major- 2. Lennard-Jones JE, Lockhart-Mummery HE, ity of cases. There are highly character- Morson BC: Clinical and pathological differ- istic histologic features in each disease, entiation of Crohn's disease and proctocolitis. and these correlate well with clinical Gastroenterology 54: 1162-1170, 1968. features. Our studies have not indicated 3. Kirsner JB: Problems in the differentiation of ulcerative colitis and Crohn's disease of the prognostic features among the histologic colon; the need for repeated diagnostic eval- characteristics, either as to recurrent dis- uation. Gastroenterology 68: 187-191, 1975. ease or severity of disease. Particularly 4. Morson BC: Rectal and colonic biopsy in disappointing as a histologic parameter inflammatory bowel disease. Am J Gastroen- of significance clinically, is the granu- terol 67: 417-426, 1977. loma in Crohn's disease. Mucosal dys- 5. Fuson J, Sullivan BH, Farmer RG, et al: Endoscopic cancer surveillance in chronic ul- plasia does appear to be a clinically and cerative colitis. Presented at annual meeting histologically important prognostic fea- of the American Society for Gastrointestinal ture in patients with long-standing ul- Endoscopy, Las Vegas, Nevada, 1978. cerative colitis. More work in this area 6. Nugent FW, Haggitt RC, Colcher H, et al: clearly needs to be done, both in defi- Malignant potential of chronic ulcerative col- itis; preliminary report. Gastroenterology 76: nition and documentation of dysplasia, 1-5, 1979.

Downloaded from www.ccjm.org on October 1, 2021. For personal use only. All other uses require permission.