Characterization of Spontaneous Colitis in Cotton-Top Tamarins (Saguinus oedipus) and Its Response to Sulfasalazine

Transcription

1 GASTROENTEROLOGY 1985;88:13-9 Characterization of Spontaneous Colitis in Cotton-Top Tamarins (Saguinus oedipus) and Its Response to Sulfasalazine JAMES 1. MADARA, DANIEL K. PODOLSKY, NORVAL W. KING, PRABHA T K. SEHGAL, RONDA MOORE, and HARLAND S. WINTER New England Regional Primate Center, Department of Pathology, Brigham and Women's Hospital, and Departments of Medicine and Pediatrics, Harvard Medical School, Massachusetts General Hospital and Childrens' Hospital, Boston, Massachusetts Chronic colitis in the cotton-top tamarin (CTT) has been characterized by obtaining distal colonic biopsy specimens, hematocrits, serum albumins, and stools for bacteriologic and parasitic examination in nondebilitated living CTTs. The species specificity of the histologic features of colitis observed in the CTT was assessed by obtaining distal colonic biopsy specimens from 10 animals of other primate species for histologic examination. Histologic evidence of active colitis was found in 50% of adult CTTs but was absent in all non-ctt species studied. Fortytwo stool samples obtained from 18 CTTs yielded only one isolate (Campylobacter). In addition to active colitis, CTT rectal mucosa also often had subtle irregularities in mucosal structure that were not present in nonrelated primate species and might represent chronic colitis. Metaplasia was not observed. The therapeutic effects of oral sulfasalazine (50 mg/kg' day) on CTT colitis were assessed in a randomized 10-wk placebo controlled crossover study. This study demonstrated significant improvement in disease activity as judged histologically (p < 0.05) and significant increases in animal Received February 7, Accepted July 5, Address requests for reprints to: James L. Madara, M.D., Department of Pathology, Brigham and Women's Hospital, 75 Francis Street, Boston, Massachusetts This work was supported in part by grants AM27972, RR , AM1392, AM3014, AM31070, and T32-AM07477 from the National Institutes of Health and by a grant from the National Foundation for Ileitis and Colitis to Dr. Podolsky. Dr. Madara is supported in part by the American Gastroenterological AssociationIRoss Research Scholar Award. The authors thank Michael O'Connell and Dennis Walsh for their excellent technical assistance in the performance of these studies by the American Gastroenterological Association /85/$3.30 weight (p < 0.01) and serum albumin (p < 0.01) during sulfasalazine therapy when compared with saline control. Sulfasalazine therapy can ameliorate the effects of this disease and offers promise in maintaining experimental colonies of this endangered species for future studies. Past animal models of colitis have produced selflimited disease that lacks the structural and clinical features of chronic colitis (1-3). Reports of spontaneous colitis in primates generally have described sporadic occurrence of an acute fulminant illness (4,5). In the absence of signs of chronicity, the relevance of these available putative animal models to human idiopathic chronic colitis remains uncertain. More recently, a chronic wasting syndrome has been reported in tamarins (6,7) and an autopsy study of the nonhuman primate, the cotton-top tamarin (CTT, Saguinus oedipus), has suggested that this wasting syndrome is attributable to chronic colitis (8). Furthermore, CTTs that die with this form of colitis have an extraordinarily high incidence of colonic carcinoma (9). These preliminary reports suggest a similarity between spontaneously occurring CTT colitis and idiopathic chronic ulcerative colitis of humans. However, little information is available concerning the histologic and clinical features of this disorder in living nondebilitated CTTs. For example, the temporal course of disease activity in this potential model is unknown. and the response of this disease to antiinflammatory agents used in the treatment of chronic colitis in humans has not been objectively assessed. The development of the therapeutic approaches that would limit the Abbreviations used in this paper: CTT, cotton-top tamarin.

2 14 MADARA ET AL. GASTROENTEROLOGY Vol. 88, No.1, Part 1 severity of colitis in the CTT would be particularly important in preserving the existing colonies of this endangered species. We performed studies directed toward a more complete characterization of CTT colitis. Mucosa was examined in distal colonic biopsy specimens from a population of CTTs to assess the presence, distribution, and structural characterization of colitis in living animals. The presence of structurally defined active colitis was correlated with potential clinical and laboratory parameters of disease activity. Multiple stool samples were examined for ova, parasites, and bacterial pathogens. We also evaluated the species specificity of the colonic mucosal structural alterations in the CTT by defining the histologic appearance of the distal colonic mucosa in five other primate species. Lastly, the effects of oral sulfasalazine therapy on the clinical, laboratory, and histologic features of CTT colitis were assessed in a prospective, placebo-controlled, blind crossover study. Materials and Methods General The animal population used for these studies included 18 CTTs (S. oedipus) and 2 monkeys from each of the following species: rhesus (Macaca mulatta), cynomolgus (Macaca fascicularis), squirrel (Saimiri sciureus), owl (Aotus trivirgatus), and common marmoset (Callithrix jacchus). These latter species were assessed to ensure that the structural lesions seen in CTTs were not commonly seen in primate species housed at our institution. The last species is closely related to CTT. All animals were housed at the New England Regional Primate Center. Cotton-top tamarins received commercial marmoset diet (ZuPreen, Hill's Pet Products Inc., Topeka, Kan.) supplemented by insects and fruits. The age of animals born in captivity ranged from 1 to 5 yr whereas feral animals were residents in the colony for >3 yr; both sexes were included in these studies. After general anesthesia with intramuscular ketamine, multiple rectal mucosal biopsy specimens were obtained with a suction biopsy apparatus at a site 3-5 cm proximal to the pectinate line. Morphologic Techniques Tissue samples for morphologic examination were carefully oriented on Gelfoam (Upjohn, Kalamazoo, Mich.) and immersed in 10% phosphate-buffered formalin. After coding and paraffin embedment, well-oriented 6-J-tm sections were obtained and stained with hematoxylin and eosin. Histologic features were graded in a blind fashion by one individual (J.L.M.). In preliminary studies, biopsy specimens were examined as a group to define the range of structural abnormalities seen within the population of CTTs. Based on this experience, a grading system was developed to quantitate histologic parameters of both acute and chronic inflammation in CTTs. Details of the histologic grading used in these studies are outlined in Table 1. Intraobserver reproducibility of the grading system was assessed by reevaluation of eight randomly selected biopsy specimens 1 wk after initial review. This group of specimens contained a broad range of active and chronic lesions. Design of Studies In preliminary studies, two to three specimens from 6 CTTs were evaluated by light microscopy to assess the prevalence of active colitis and to define the distribution of activity within the rectum (Le., focal or diffuse). The specificity of the histologic findings in CTTs was assessed by comparison with histologic features of distal colonic mucosal specimens from other primate species housed at the New England Regional Primate Center. Colonic suction biopsy specimens were obtained from 2 animals in each of five additional primate species including both New World and Old World species (see above). The range of histologic features of CTT colitis was further explored by review of postmortem histologic slides of the colon and small intestine from 10 CTTs that had previously died of colonic adenocarcinomas complicating colitis and were selected on the basis of the absence of autolytic mucosal alterations. To examine further the relevance of CTT colitis to human idiopathic colitis, the histologic and clinical response to sulfasalazine was studied in 12 CTTs in a 10-wk prospective, randomized, blinded crossover trial. The animals, randomly divided into two groups, received either a suspension of sulfasalazine (50 mg/kg' day administered as two equal doses) or placebo (saline) for 5 wk at which time treatment groups crossed over and were followed for a second 5-wk study period. Multiple distal colonic biopsy specimens were obtained at the beginning, middle (after the initial 5-wk treatment period), and end of the 10-wk study period for histologic evaluation as described above. In addition, serial venous blood samples were obtained at the same intervals for determination of complete blood count and serum albumin levels performed in a commerciallaboratory (Vetpath, Teterboro, N\J.). Serial stool samples were also obtained at the same intervals and examined for the presence of bacterial pathogens (including Salmonella, Shigella, Yersinia, and Campylobacter species) as well as for ova and parasites in the clinical bacteriology laboratories of the tvwfsachusetts General Hospital. Body weights were recorderci on a weekly basis. Results Histologic Characterization of Acute Cotton Top Tamarin Colitis Initial efforts were directed toward defining the histologic features of colitis in CTTs. Blind comparison of the histologic findings in two mucosal biopsy specimens obtained from each of 6 CTTs showed a nearly uniform intraanimal concordance. The interpretations of the degree of inflammatory

3 January 1985 COLITIS IN TAMARINS 15 Table 1. Histologic Evaluation of Activity and Chronicity in Rectal Biopsy Specimens From Cotton-Top Tamarins Grade Activity Normal (0) Mild (1) Moderate (2) Definition Absence of single intraepithelial polymorphonuclear leukocytes (PNMs) and absence of clusters of PMNs in the lamina propria Focal a ggregates of lamina propria PMNs + the presence of i solated intraepithelial PMNs in three or fewer crypts per biopsy specimen Small clusters of intraepithelial PMNs or isolated intraepithelial PMNs in more than three crypts per biopsy specimen Severe (3) Crypt abscesses Structural alterations that may relate to chronicity Normal (0) No structural features of chronicity Mild (1) Equivocal irregularity of crypts or of surface with the presence of single branched crypts Moderate (2) Definite irregularity of crypts or surface with at least two foci of crypt branching Severe (3) Definite irregularity of crypts or surface with multifocal branching and focal loss of crypts activity in each of two specimens were in agreement for 5 animals, and in the 6th animal one biopsy specimen was interpreted as displaying mild, and the other, moderate activity. These data indicate that inflammatory activity of the rectum, when present, was diffuse and relatively uniform in nature. Acute inflammatory activity was present in 4 of 6 (67%) animals in this initial group and 5 of 12 (42%) in a second group of animals studied subsequently as part of the therapeutic trial (see below). Because duplicate rectal biopsy specimens were always taken and were highly consistent in showing the same degree of activity, it is unlikely that histologic activity in the rectum was missed due to sampling error. Histologic features of activity and their grading are summarized in Table 1. Active disease was most often characterized by the presence of neutrophils invading the epithelium of the crypts with concurrent presence of neutrophils in the lamina propria. In addition, the inflammatory infiltrate was invariably accompanied by an increase in the density of lymphocytes and plasma cells in the lamina propria. In specimens containing no active disease, the lamina propria alternately showed increased lymphocyte and plasma cell density or a density of these mononuclear cells similar to that seen in other primate species without colitis (Le., normal). The presence of intraepithelial neutrophils unassociated with mucous granule discharge by nearby goblet cells was occasionally noted. Although this feature was not observed in the majority of actively inflamed crypts, it was noted focally, particularly in CTTs in which a large volume of mucosa was available for study (e.g., necropsy specimens). However, it should be stressed that neutrophil invasion into the epithelium was usually accompanied by loss of goblet cell mucus and attenuation of remaining epithelial cells. Histologic Features of Chronic Colitis The presence of histologic features of chronic colitis in CTTs was also evaluated according to criteria outlined in Table 1. We emphasize that this grading system is based on the range of mucosal structure seen in CTT colitis which, as covered in detail in the Discussion, is far milder than the range of structural lesions seen in chronic ulcerative colitis in humans. Features of chronicity in most instances (grade 1) were very mild and consisted of slight surface irregularity and curling of crypts (Figure 1). Indeed this low-grade chronicity might be termed "equivocal" chronicity. Architectural changes of chronicity were most easily evaluated in 10 animals lacking concurrent activity: 4 had biopsies showing moderate (definite) chronic changes and none in this group of 12 animals were classified as showing severe chronic changes. Metaplasia, a common feature of chronic colitis in humans (10-12), was not observed. In addition to the mucosal biopsy specimens obtained from living CTTs, histologic features were also evaluated retrospectively in colonic mucosa obtained from 10 animals that died from colonic carcinoma superimposed on colitis. The entire colons of these animals were studied, and many animals had regions of severe acute activity. However, in regions containing either mild or no activity, features of moderate to severe chronicity (Figure lb) were often diffusely present (50%). However, again metaplasia was not found. When it was possible to be evaluated, the distal small bowel was normal in these animals. Although it is not the focus of the present study, it is interesting to note that no instances of epithelial dysplasia were noted in this small group of autopsied animals. In order to determine the significance of the histologic features of inflammation described above and in Table 1, the colonic mucosa from a variety of additional primate species was examined (Figure 2). Among the five non-ctt species examined, 1 of 2 squirrel monkeys had elongated colonic crypts with numerous mitoses and mucous depletion of crypt goblet cells but had no intraepithelial or lamina propria polymorphonuclear leukocytes-findings consistent with a regenerative response to a recent colonic epithelial insult. The remaining squirrel monkey, as well as the owl, rhesus, and cynomolgus monkeys had mucosal architecture that appeared

4 16 MADARA ET AL. GASTROENTEROLOGY Vol. 88, No.1, Part 1 Figure 1. Light micrographs of rectal mucosa from 4 different cotton-top tamarins (Saguinus oedipus). A. Rectal mucosa which appears to have only equivocal, if any, structural abnormality consistent with chronic injury. Although the surface is slightly irregular and perhaps focally villiform (arrowhead) and an occasional crypt is irregular (double arrowhead), these putative alterations are extremely subtle. X200. B. Rectal mucosa with "definite," grade 3 atrophy. Alterations include branching (1) and irregular (2) crypts. Although similar changes were present in other animals, these changes are representative of the most marked degree of atrophy observed in this study. Metaplasia was never found. x 200. C. Mild (grade 1) active inflammation with single intraepithelial neutrophils (arrowheads). x 310. D. Severe (grade 3) active inflammation with focal crypt abscesses (asterisks) accompanied by regenerative change and crypt goblet cell mucin depletion. X310. structurally indistinguishable from normal human colon (Figure 2). These observations suggest that the structural features of chronicity that were noted in mucosal biopsy specimens from CTTs do not simply represent a baseline difference between human and nonhuman primates. Interestingly, however, rectal mucosa from both common marmosets (a species closely related to the CTT) showed either slight surface or crypt irregularity-features consistent with mild (i.e., equivocal) chronicity in our histologic grading system (Figure 2). Sulfasalazine Study Of the 12 animals entered in the therapeutic trial, 10 completed the study; 2 animals were removed at the discretion of the clinical veterinarian due to severity of clinical disease activity with loss of -20% of body weight. Although both animals were subsequently noted to be receiving placebo at the time of their removal from the study, data from these animals were not included in subsequent analysis. Of the 20 5-wk animal treatment periods evaluated (10 animals each having one placebo and one sulfasalazine treatment period), nine periods were not associated with histologic evidence of disease activity at either the beginning or the end of a study period. The remaining 11 periods were thus analyzed to determine the relative efficacy of drug versus placebo in reducing or preventing the development of active colitis. As shown in Table 2, of the

5 January 1985 COLITIS IN TAMARINS 17 Figure 2. Light micrographs of rectal mucosa of the (A) common marmoset (Callithrjx jaeehus), (B) cynomolgus (Macuea faseieularis), (e) rhesus (Maeaea mullata), and (D) owl (Aotus trivirgatus) monkey, With the exception of the common mari)1oset-a species closely related to the cotton-top tamarin-the mucosa of all species have parallel straight crypts and a regular surface. However, the rectal mucosa of the common marmoset may exhibit irregular crypts and surface, thus giving it the appearance of mild atrophy. In all nonhuman primate species studied, an occasional crypt or small group of crypts can "herniate" through the muscularis mucosa (D, arrowheads). This latter finding, which often occurs in association with submucosal lymphoid patches, may thus be considered to be normal. All micrographs X animals that increased their grade of activity during the treatment period, all were receiving placebo. These 4 animals had no activity at the beginning of the study period and 2 developed grade 1 activity, whereas 1. each developed grade 2 and 3 activity. In contrast, 6 of the 7 animals demonstrating improvement in disease activity during a treatment period were receiving sulfasalazine, and only 1 was receiving placebo (Table 2). The placebo animal improved from grade 1 to grade 0 activity. Of the CTTs that improved with sulfasalazine therapy, 3 improved by two to three grades and 3 improved by one grade. Analyzing these data by x 2 analysis with the Yates' correction revealed that sulfasalazine was significantly (p < 0.05) more efficacious than placebo in ameliorating histologic disease activity. The clinical and laboratory parameters of disease activity derived from the entire group studied are summarized in Table 3. When data from all study animals were examined regardless of the presence or absence of histologic disease activity, no significant changes in hematocrit, white blood cell count, or serum albumin were observed after either placebo or sulfasalazine treatment. However, it should be noted that weight gain was significantly greater in sulfasalazine-treated animals than in those receiving placebo (p < 0.05 and p < 0.01 for groups 1 and 2, respectively). When data from animals demonstrating histologic signs of disease activity at some time during the study period were analyzed separately, significant increases in serum albumin (p < 0.01) and gains in weight (p < 0.01) were found after Table 2. Effect of Sulfasqlazine on Colitis Activity in Cotton-Top Tamarins Having or Developing Activity During Treatment Period Outcome Activity improved a Activity worsened a p < Sulfasalazine 6 o Treatment Placebo 1 4

6 18 MADARA ET AL. GASTROENTEROLOGY Vol. 88, No, 1, Part 1 Table 3. Clinical and Laboratory Parameters in the Cotton-Top Tamarin: Evaluation of a Crossover Therapeutic Trial Parameters Group Week of Hct WBC Albumin ~ Weight No." study Rx (%) (x 1 O ~ 3 ) (g/dl) (%) 1 0 Sulfasalazine 38.1 ± ± ± 0.3 ~ 5 Placebo 38.4 ± ± ± ± 0.48 b ~ ± ± ± ± 0.42 e 2 0 Placebo 47.0 ± ± ± 0.5 t 5 Sulfasalazine 38.0 ± ± ± ± 2.10 d t ± ± ± ± 1.48 e Rx, treatment; Hct, hematocrit; WBC, leukocytes. " Cotton-top tamarin randomly assigned to receive either sulfasalazine (50 mg/kg. day) or saline placebo for 5 wk; animals were then changed to the alternate treatment for an additiona,l 5 wk. band c are significantly different, p < d and e are significantly different, p < sulfasalazine but not after placebo (Table 4). In contrast to its effect on active colitis, sulfasalazine therapy had no influence on the presence of histologically defined chronic colitis. No parasites were identified in any CTT during the course of this study; 1 animal was noted to have a culture demonstrating Campylobacter ssp jejuni. This animal was 1 of 2 animals, noted above, that were removed from the study. No other enteric pathogens were recovered. Discussion The presence of colonic adenocarcinoma in association with colitis has been previously described in autopsy studies of CTTs (9). Although these findings have an apparent association with a clinical syndrome of wasting in the CTT, the histologic and dinical features of CTT colitis remain incompletely defined (7-9). Our initial review of colonic mucosal biopsy specimens from living CTTs demonstrated the frequent presence of acute and chronic colitis. These findings were not simply a reflection of differences in human versus nonhuman primate mucosal architecture in that similar active lesions were not present in a wide range of other nonhuman primates. We also demonstrate that the activity of CTT colitis is lessened by sulfasalazine as assessed by morphologic, clinical, and laboratory criteria. The activity of CTT colitis within the distal colon could be adequately assessed by a single rectal biopsy because duplicate or triplicate rectal biopsies showed highly comparable grades of activity. This finding suggests that variations in histologically defined activity reflect real differences in rectal disease activity and are not due to sampling error. The finding of a wide variation in the presence and degree of histologic activity in serial mucosal specimens of the same animal suggests that the course of CTT colitis is marked by spontaneous flares in disease activity, e.g., 1 animal while on placebo was noted to have no activity on initial study whereas subsequent biopsy specimens showed moderate and severe degrees of activity. The mucosal damage resulting from repetitive flares presumably results in the distortion of normal mucosal architecture-an appearance commonly termed atrophy which is also the hallmark of chronic colitis in humans (10). Although structural changes consistent with chronic colitis were frequently found in the CTT rectal mucosa, they were not uniformly present and tended to be of relatively mild degree. It is interesting to note that, in contrast to other primates, the common marmoset, which is closely related to the CTT, demonstrated histologic features consistent with mild rectal mucosal atrophy. The possible presence of clinically significant colitis has not been well studied in the common marmoset. However, the presence of more severe structural features of chronicity was observed only in CTT colonic mucosa, although even these were not as severe as those often seen in human chronic colitis. For example, meta- Table 4. Effect of Sulfasalazine on Clinical and Laboratory Parameters in Cotton-Top Tamarins With Active Colitis Parameters Sample Hct WBC Albumin ~ Weight (n = 7) (%) (x 1 O ~ 3 ) (g/dj) (%) Pre-Rx 35.6 ± ± ± 0.5 Post-Rx 39.9 ± ± ± ± 8.4 NS NS P < 0.01 P < 0.01 Hct, hematocrit; WBC, leukocytes; Rx, treatment; NS, not significant.

7 January 1985 COLITIS IN TAMARINS 19 plasia was not found in any CTT. It is possible that the relative pal.j.city of histologic features of severe chronic injury in these animals may reflect a more limited duration of active lesions in the CTT in comparison to human chronic colitis. In addition, the relatively short life spans of these animals may limit the extent of the development of signs of chronicity. Alternatively, lack of severe mucosal atrophy may reflect the infrequency with which severe architectural damage accompanies the acute disease process in the CTT. For example, mucosal erosions and ulcerations are uncommon in CTTs even in the presence of multifocal crypt abscesses, but such destructive lesions are routinely found during acute flares of chronic ulcerative colitis in humans. Although epithelial erosions were not commonly found in tissue sections of CTT, it is probable that occasionally such lesions do occur in these animals insofar as CTTs with the wasting syndrome may demonstrate occult blood in their stools (unpublished observations). Although 1 CTT studied had Campylobacter species isolated from stool samples, no other enteric pathogens or intestinal parasites were found in the multiple stool samples that were obtaineq from each animal. In addition, no parasites were found in mucosal biopsy specimens. Although the stil.j.ctural features of chronic ulcerative colitis and infectious (or acute self-limited) colitis in humans overlap (11), infectious colitis may show combinations of structural features that are less commonly found in chronic ulcerative colitis (10), including intense neutrophilic infiltration of the mucosal surface with preservation of glandular architecture (10,12,13) and relative absence of chronic inflammation in the lamina propria (14). These characteristic features of infectious colitis were not observed in CTT colitis. Although the nature of the structural lesion observed in CTTs, like that found in chronic ulcerative colitis, is nonspecific, a recent study in humans showed that the appearance of distorted crypt architecture such as that seen in Figure lb in CTTs is highly predictive of idiopathic inflammatory bowel disease and is not seen in self-limited colitis in humans (15). However, in the absence of an identifiable etiologic factor or a disease-specific marker, it is difficult to assess the relevance of any potential animal model to the human disorder. The response of CTT colitis with activity to sulfasalazine is not only similar to that observed in controlled studies of idiopathic chronic colitis with activity in humans (16-18), but currently offers hope of controlling disease severity in the CTT. These animals are now an endangered species and enlargement of existing colonies will depend on the ability to breed relatively healthy pairs of CTT in captivity. Adult CTTs currently die with alarming frequency, often of complications of colitis. Indeed one recent study reported that 65% of CTTs died within the first year after arrival and death rates in subsequent years ranged from 16% to 38% (9). We hope that judicious use of sulfasalazine therapy may help preserve this potentially useful model for subsequent study. References 1. Kirsner JB. Experimental "colitis" with particular reference to hypersensitivity reactions in the colon. Gastroenterology 1961;40: Anver MR, Cohen BJ. Animal model: ulcerative colitis induced in guinea pigs with degraded carrageenins. Am J Pathol 1976;84: Watt S, Marcus R. Experimental ulcerative disease of the colon in animals. Gastroenterology 1973;14: Stout C, Snyder RL. Ulcerative colitis-like lesion in Siamang gibbons. Gastroenterology 1969;57: Scott GBD, Keyman IF. Ulcerative colitis in apes: a comparison with the human disease. J PathoI1975;115: King GJ. An investigation into "Wasting Marmoset Syndrome" at Jersey Zoo. Extract from the Jersey Wildlife Preservation Trust Thirteenth Annual Report, Richter CB, Tankersley W, Webb A. Chronic recurrent colitis: a wasting syndrome in marmosets and tamarins (abstr). Twenty-ninth Annual Session of the American Association for Laboratory Animal Science, New York, Chalifoux LV, Bronson RT, Escajadillo A, McKenna S. An analysis of the association of gastroenteric lesions with chronic wasting syndrome of marmosets. Vet PathoI1982:19(Suppl 7): Chalifoux LV, Bronson RT. Colonic adenocarcinoma associated with chronic colitis in cotton-top marmosets, Saguinus oedipus. Gastroenterology 1981;80: Goldman H, Antonioli DA. Mucosal biopsy of the rectum, colon and distal ileum. Hum Pathol 1982;13: Yardley JH, Donowitz M. Colo-rectal biopsy in inflammatory bowel disease in the gastrointestinal tract. In: Yardley JH, Morson BC, Abell MR, eds. The gastrointestinal tract. Baltimore: Williams & Wilkins, 1977: Kumar KB, Nostraut TT, Appleman HD. The histopathologic spectrum of acute self-limited colitis (acute infectious-type colitis). Am J Surg Pathol 1982;6: Price AB, Jewkes J, Sanderson PJ. Acute diarrhea: campylobacter colitis and the role of rectal biopsy. J Clin PClthol 1979;32: Morson Be. Rectal biopsy in inflammatory bowel disease. N Engl J Med 1972;287: Surawicz CM, Belic L. Rectal biopsy helps to distinguish acute self-limited colitis from idiopathic inflammatory bowel disease. Gastroenterology 1984;86: Baron JH, Conell AM, Lennard-Jones JE, Jones FA. Sulphasalazine and salicylazosulphadimidine in ulcerative colitis. Lancet 1962;i:l Dick AP, Grayson MJ, Carpenter RG, Petrie A. Controlled trial of sulphasalazine in treatment of ulcerative colitis. Gut 1964; 5: Lennard-Jones JE, Longmore AI. Newell AC, Wilson CWF, Jones FA. An assessment of prednisone salazoprin and topical hydrocortisone used as outpatient treatment for ulcerative colitis. Gut 1960;1: