Rectal Histology in Acute Bacillary Dysentery

Transcription

1 GASTROENTEROLOGY 1986;90: Rectal Histology in Acute Bacillary Dysentery B. S. ANAND, V. MALHOTRA, S. K. BHATTACHARYA, P. DATTA, D. DATTA, D. SEN, M. K. BHATTACHARYA, P. P. MUKHERJEE, and S. C. PAL Department of Gastroenterology, G. B. Pant Hospital, New Delhi; and National Institute of Cholera and Enteric Diseases, Calcutta, India' A recent epidemic of acute Shigella dysentery in West Bengal (India) provided us with an opportunity to examine the rectal mucosal abnormalities seen in this condition. One hundred two patients were investigated using sigmoidoscopy, rectal biopsy, and rectal swab for culture. Pure culture of Shigella was obtained in 37 cases, and the rectal biopsy specimens from these patients were assessed in detail. The mean (±SD) duration of illness was 47.8 ± 27.4 h (range h), and most patients (31 of 37,84%) had diarrhea with blood and mucus in the stools. Significant findings at histology were as follows. (a) Cellular infiltrate was predominantly round cell or mixed round cell and neutrophilic in the majority of patients (27, 73%). (b) Disorganization of crypts was seen in as many as 31 patients (84%); in most subjects the distorted architecture was mild, but in a few the defect was severe with crypt branching and dilatation. (c) In the majority of patients the inflammatory process extended to the muscularis mucosae and submucosa; edema with or without increased cellular infiltrate was seen in the muscularis mucosae in 92% and in the submucosa in 80%. (d) There was no difference in the rectal histology of patients with a short history of disease «48 h) compared with those with a longer history, except for goblet cell depletion which was more in those with diarrhea for more than 48 h. (e) The mucosal abnormalities in patients with watery diarrhea were, in general, milder than in those with dysentery, although the difference was statistically not significant; 2 of 6 patients with watery diarrhea had severe colitis. (f) The mucosal abnormalities were more severe in patients with Shigella dysenteriae infection compared with Shigella flexneri. Received February 28, Accepted September 17, Address requests for reprints to: Dr. B.S. Anand, Department of Gastroenterology, G.B. Pant Hospital, New Delhi , India by the American Gastroenterological Association /86/$3.50 Shigella is a gram-negative, nonmotile, non-lactosefermenting organism that causes acute diarrhea with or without the presence of blood in the stools. There are four species (Shigella dysenteriae, Shigella flexneri, Shigella boydii, and Shigella sonnei) and many serotypes (1). In the United States and other developed countries S. sonnei and S. flexneri predominate, whereas S. dysenteriae is seen primarily in the developing countries. Shigella infection can occur both sporadically and in an epidemic form (2). Clinically, the disease produced by this organism covers a wide sprectrum, from asymptomatic or subclinical infection, to a mild watery diarrhea, to fatal hemorrhagic dysentery, which is caused typically by S. dysenteriae type 1 (Shiga bacillus) (2). Since 1964 the clinical division of the National Institute of Cholera and Enteric Diseases has been maintaining a continuous surveillance of Shigella infection in patients admitted with acute diarrhea at the Infectious Disease Hospital (Calcutta, West Bengal, India). The isolation rate of Shigella species over all these years varied from 1.7% to 4.5%, and S. dysenteriae had not been encountered for 10 yr (3). However, an epidemic of bacillary dysentery due predominantly to S. dysenteriae type 1 began sweeping through West Bengal from the beginning of February 1984, and by June, 78,000 cases had been reported with 2200 deaths (4). This provided us with an opportunity to examine, among other aspects, the rectal mucosal abnormalities produced by this organism. The histologic appearances in Shigella colitis have not been well described and the present study documents in detail the mucosal abnormalities seen in this condition. Materials and Methods Control Group Rectal biopsy specimens from 30 patients with irritable bowel syndrome were included in the control

2 March 1986 RECTAL HISTOLOGY IN BACILLARY DYSENTERY 655 group. Diagnosis of irritable bowel syndrome was made on clinical grounds and on the exclusion of organic pathology by investigation. Investigations carried out were as follows: routine blood tests, repeated (x 3) stool examinations for ova and cysts, and sigmoidoscopy. At sigmoidoscopy, stool and rectal biopsy specimens were obtained and immediately cultured for Entamoeba histolytica in Robinson's culture medium as described earlier (5). Part of the biopsy specimen was used for histopathological examination. In addition, barium examination and fiberoptic endoscopy were performed if considered necessary. These subjects were part of our control group in a study on E. histolytica cyst passers (6) and were under observation for more than a year; none developed any organic pathology during this period. Patient Group One hundred two patients with acute diarrhea presenting to the Infectious Diseases Hospital (Calcutta) were included in the study. After recording the clinical history and physical findings, the patients underwent sigmoidoscopy with a rigid instrument. No bowel preparation was used before the procedure. At sigmoidoscopy a rectal swab for bacterial culture and a biopsy specimen for histopathology were obtained. Culture The rectal swabs obtained at the time of sigmoidoscopy were analyzed for various bacterial enteropathogens using standard techniques (7-9). Characteristic Shigella-like colonies on MacConkey agar, desoxycholate citrate agar, Salmonella-Shigella agar (small, colorless, and discrete), and xylose-lysine desoxycholate agar (small, pink, and discrete) were biochemically identified using API 20 Enterobacteriaceae strips (API system, Montalieu Vercieu, France). Biochemically confirmed Shigella isolates were finally confirmed by slide agglutination with Shigella group polyvalent and monospecific antisera obtained from Wellcome Reagents Limited (U.K.). Rectal Biopsy Histology The biopsy specimens were fixed in 10% formol saline. Serial sections (5 /Lm thick) were obtained and stained with hematoxylin and eosin. The slides were examined blind by two examiners who were unaware of the culture report and clinical details of the patients. For the evaluation of the biopsy specimens, we selected the same histopathological features as described by Surawicz and Belic (10). The criteria adopted for assessing the various histologic features were as follows: crypt architecture was considered normal if the crypts were arranged in a regular parallel order; derangement of this order or presence of branched crypts was categorized as distorted crypt architecture. Crypt atrophy was considered if the crypts were widely spaced or when the area between the crypts and surface epithelium or between the crypts and the muscularis mucosae was increased. A subjective analysis of the relative numbers of neutrophils and round cells within the lamina propria in comparison with the control group was made. The results were categorized into (a) predominant increase of neutrophils, (b) predominant increase of round cells, and (c) an equal increase in both neutrophils and round cells. Mucosal erosions were considered only if neutrophils were present at the site of damage; absence of neutrophils suggested traumatic artifact. We also assessed some additional morphologic features not included by Surawicz and Belic (10): presence of thrombus within capillaries, cellular infiltrate and edema of the muscularis mucosae, and inflammation of the submucosa. On the basis of the above findings, a final subjective assessment of the rectal biopsy specimens was made and the results were categorized into two groups: (a) mild colitis and (b) moderate/severe colitis. Mild colitis was diagnosed when the surface epithelium was flattened and showed focal erosions but without any surface exudate; there was an increase in the cellular infiltrate in the lamina propria, in the absence of any crypt abscesses (mild crypt atrophy was occasionally seen); the muscularis mucosae and submucosa were essentially normal or showed only mild edema and increase in cellular infiltrate. Moderate/severe colitis was diagnosed in the presence of diffuse mucosal damage with surface exudate, crypt atrophy, and crypt abscesses, and dense infiltration of the lamina propria by round cells and neutrophils. Results Control Group There were 24 men and 6 women with a mean (±SD) age of 35.4 ± 10 yr (range yr) in this group. Rectal histology. The surface epithelium was composed of tall columnar cells with a regular basal arrangement of nuclei; in 2 subjects focal areas of flattening were noted. Occasional neutrophils were seen in the surface epithelium; the number varied from 0 to 1 per high power field (HPF, x 40) in 27 subjects (90%) and from 0 to 2 in the remaining 3 subjects. The crypts were normal in all subjects; there was no distortion, atrophy, focal dilatation, crypt abscess, or neutrophilic infiltration of the crypt epithelium in any instance. In the lamina propria, the usual cell type was the round cell. These cells occurred more often in the superficial zone of the mucosa, and in this region their number varied from 20 to 35 in individual crypts. By contrast, neutrophils were rare, numbering only one or zero per HPF. Other features such as capillary thrombi, goblet cell depletion, granuloma formation, giant cells, and histiocytic collection were not seen in any subject. The muscularis mucosae was normal in all subjects. The submucosa was available for examination in only 5 subjects and was normal in all.

3 656 ANAND ET AL. GASTROENTEROLOGY Vol. 90, No.3 Table 1. Histologic Features of 37 Patients With Shigella Dysentery and Their Relation to the Duration of Illness and the Infecting OrganisrnG Duration of illness Infecting organism :0::48 h >48 h SDl SF2 Histologic criteria No. (n = 26) (n = 11) P (n =22) (n = 8) P Surface epithelium Flattened 33 (89) 22 (85) 11 (100) NS 20 (91) 7 (87.5) NS Erosions 32 (86) 22 (86) 11 (100) NS 21 (95) 6 (75) NS Villus Neutrophils within 36 (97) 25 (96) 11 (100) NS 22 (100) 8 (100) NS Exudate 13 (35) 8 (31) 5 (45) NS 6 (27) 2 (25) NS Crypts Distorted 31 (84) 21 (81) 10 (91) NS 19 (86) 6 (75) NS Atrophy 32 (86) 22 (85) 10 (91) NS 21 (95) 6 (75) NS Focal dilatation 32 (86) 18 (69) 8 (73) NS 18 (82) 6 (75) NS Crypt abscess 8 (22) 5 (19) 3 (27) NS 3 (14) 2 (25) NS Neutrophils within epithelium 26 (70) 17 (65) 9 (82) NS 16 (73) 5 (62.5) NS Lamina propria Increased round cells 21 (57) 10 (38) 6 (55.5) NS 13 (59) 5 (62.5) NS Increased neutrophils 10 (27) 5 (19) 5 (45) NS 6 (27) 2 (25) NS Both neutrophils and round 6 (16) 6 (23) 0 NS 3 (14) 1 (12.5) NS cells Superficial zone 28 (76) 20 (77) 8 (73) NS 16 (73) 7 (81.5) NS Capillary thrombi 15 (40.5) 8 (31) 7 (64) NS 10 (45) 3 (37.5) NS Goblet cell depletion 25 (68) 14 (54) 11 (100) < (77) 2 (25) <0.01 Epithelioid granulomas, giant cells, and histiocytic collection Muscularis mucosae Edema 34 (92) 23 (88) 11 (100) NS 22 (100) 7 (87.5) NS Increased cells 23 (62) 15 (58) 8 (73) NS 19 (86) 4 (50) <0.05 Submucosa Edema ± increased cells 28 (80) 19 (76) 9 (90) NS 16 (80) 5 (62.5) NS NS, not significant; SD1, S. dysenteriae type 1; SF2, S. flexneri type 2." Note that the submucosa was not included in the biopsy specimen in 2 patients, one with a short disease history and the other with disease for more than 48 h; both specimens were from patients with S. dysenteriae type 1 infection. Numbers in parentheses are percentages. Patient Group Shigella organism was cultured in 39 of the 102 patients studied. Two of the 39 patients showed a mixed culture of Shigella and Vibrio parahaemolyticus and were excluded from the study. Therefore, for the purpose of analysis of rectal biopsy specimens, only 37 patients were included. These cases involved S. dysenteriae type 1 (n = 22), type 2 (n = 1), and type 4 (n = 1); S. flexneri type 1 (n = 1), type 2 (n = 8), type 4 (n = 1), and type 6 (n = 1); and S. boydii type 12 (n = 2). No case of S. sonnei was isolated. Clinical profile. There were 34 men and 3 women in the patient group. The mean (±SD) age was 27.7 ± 11.6 yr (range yr). The mean (±SD) duration of illness was 47.8 ± 27.4 h (range h). Most of the patients were passing stools that contained blood and mucus (31 of 37, 84%). Watery diarrhea was present in only 6 patients (16%). For the purpose of analysis the patients were divided into two groups: those with diarrhea for <48 h (26 patients, 70%) and those with a longer history (11 patients, 30%). Histologic analysis. The results of analysis of the various histologic features are shown in Table 1. Abnormal surface epithelium was observed in the majority of subjects: 89% had a flattened epithelium involving a portion of or the entire biopsy specimen, 86% had mucosal erosions, and infiltration by polymorphonuclear cells was seen in all but 1 patient (97%). Exudate containing neutrophils over the surface epithelium was seen in 13 of 37 patients (35%); in 2 of the 13 a pseudomembrane was present (Figure 1). A villus surface epithelium was not seen in any patient. Abnormality of crypt architecture was seen in the majority of subjects: 84% had distorted crypts, 86% had crypt atrophy, and 86% had focal crypt dilatation (Figures 2 and 3). In most patients, the crypt distortion was mild and was the result of mucosal edema and increase in the cellular infiltrate. Presence of crypt abscesses was uncom-

4 March 1986 RECTAL HISTOLOGY IN BACILLARY DYSENTERY 657 :> ~.. ' (.,~,.;"., - j', C.;.....: ~ '.~ -i Figure 1. Rectal biopsy specimen showing severe mucosal inflammation with damage to the surface epithelium and pseudomembrane formation (H&E, x50). mon (8 of 37 patients, 22%), with half of the patients showing only a single crypt abscess. No specific distribution of the crypt abscesses or "string-ofpearls" appearances were noted. However, infiltration of the crypt epithelium by neutrophils was seen frequently (25 of 37 patients, 68%). An increase in the cellular infiltrate (round cells and neutrophils) in the lamina propria was observed in every patient. The predominant cell type most often was round cells (21 of 37 patients, 57%) or both round cells and neutrophils (16%), whereas a predominant increase of neutrophils was seen in only 27% of the patients (Figure 4). An increase in the cellular infiltrate in the superficial zone compared with the rest of the mucosa was observed in the majority of patients (28 of 37,76%). Thrombi within the capillaries was seen in 41% of the patients. In most patients the inflammatory process extended into the muscularis mucosae and submucosa (Figure 5). Edema with or without increased cellular infiltrate was seen in the muscularis mucosae of 34 of 37 patients (92%) and in the submucosa of 28 of 35 patients (80%). (In 2 patients the submucosa was not included in the biopsy specimen.) On the basis of subjective assess- Figure 2. Focal area of cuboidal surface epithelium with dilatation and branching of the crypts in a rectal biopsy specimen (H&E, X100). Figure 3. Rectal biopsy specimen showing flattened and ulcerated surface epithelium with regenerating glands in the superficial zone. Note dilated crypts in the deeper area (H&E, x 100).

5 658 ANAND ET AL. GASTROENTEROLOGY Vol. 90, No. 3 Figure 4. Increased cellular infiltration of the lamina propria in this rectal biopsy specimen. Most of the cells are lymphocytes.and plasma cells (H&E, x200). Figure 5. Diffuse inflamm'ltory infiltrate of the muscularis mucosae and submucosa of a rectal biopsy specimen (H&E, x100). ment, 14 patients (38%) had mild abnormalities and the remaining 23 patients (62%) showed moderate to severe colitis. Comparison of histologic features and clinical profi.le. Only 6 patients (16%) had watery diarrhea compared with 31 patients (84%) with dysentery. The types of infecting organism in the group with watery diarrhea [So dysenteriae type 1 (n = 2). S. flexneri type 2 (n = 2). S. flexneri type 1 (n = 1), and S. boydii type 12 (n = 1)] were no different from those in the group with blood and mucus in the stools. Four of the 6 patients with watery diarrhea (67%) had mild histologic abnormalities compared with 10 of the 31 patients with dysentery (32%); the difference was statistically not significant L~ = 2.53) (Table 2). Surprisingly, 2 of the 6 patients with watery diarrhea showed significant mucosal abnormalities. One of these was a 15-yr-old boy who presented 5 days (130 h) after the onset of diarrhea that had remained watery throughout the period of illness. Sigmoidoscopy showed a moderate degree of diffuse hyperemia, and stool culture revealed S. dysenteriae type 1 organism. Rectal biopsy showed severe colitis with mucosal erosions, purulent exudate, and crypt abscesses. The rectal biopsy appearances of patients with a short disease history «48 h) were compared with those with a longer history (>48 h) (Table 1). There was a remarkable degree of similarity in the two groups, and no significant difference was observed with respect to any 'of the histologic features except one. The one exception was goblet cell depletion, which was seen in all 11 patients with diarrhea for more than 48 h as compared with 54% (14 of 26) in those with a short history (.t = 7.51, p < 0.01). Capillary thrombi were seen more frequently in p&tients with a longer history (64% vs. 31 %), but this difference was statistically not significant. It is interesting to note that infiltration of the lamina propria by neutrophils was seen in a proportionately larger number of patients with diarrhea for more than 48 h compared with those with a shorter history (45% vs. 19%), although the difference was statistically not significant. Comparison of histologic features with the type of infecting organism. In view of the small numbers of patients with Various Shigella species, Table 2. Comparison of the Severity of Mucosal Damage With the Duration of Illness, Character of Stools, and Type of Infecting Organism Q Moderate/severe Parameter Mild colitis colitis Duration of illness ~48 h 11 (42) 15 (58) > 48 h 3 (27) 8 (73) Character of slools Blood and mucus 10 (32) 21 (68) Watery 4 (67) 2 (33) Infecting organism S. dysenteriae Type Type Type S. flexneri Type Type Type Type 6 0 S. boydii type o Patients with S. dysenteriae type 1 had more severe rectal mucosal abnormalities compared with S. flexneri type 2 (X 2 = 5.56, P < 0.02). When the two Shigella species were assessed as a group, S. dysenteriae resulted in more severe mucosal abnormalities compared with S. flexneri (X 2 = 4.82, P < 0.05). Numbers in Pilrentheses are percentages.

6 March 1986 RECTAL HISTOLOGY IN BACILLARY DYSENTERY 659 comparison was only possible between S. dysenteriae type 1 (n = 22) and S. flexneri type 2 (n = 8). Subjective assessment of the severity of mucosal damage revealed a significantly larger number of moderate to severe abnormalities with S. dysenteriae type 1 compared with S. flexneri type 2 (73% vs. 25%,.I = 5.56, P < 0.02). Qualitatively, the histologic damage caused by the two species of Shigella was very similar, and a statistical difference was observed with respect to only two features. Goblet cell depletion (77% vs. 25%,.I = 6.9, P < 0.01) and cellular infiltration of the muscularis mucosae (86% vs. 50%,.I = 4.33, P < 0.05) occurred significantly more often in biopsy specimens from patients with S. dysenteriae type 1 compared with S. flexneri type 2. Patients in the S. dysenteriae group as a whole had more severe mucosal damage compared with the S. flexneri group; moderate to severe abnormalities were seen in 18 of 24 patients with S. dysenteriae (75%) compared with only 4 of 11 patients with S. flexneri infection (36%) (.I = 4.82, p < 0.05) (Table 2). Discussion Rectal mucosal appearances in acute shigellosis have not been well described. Most reports on rectal histology in acute infective diarrheas include only a few patients with Shigella and contain a description common to a number of different infecting organisms. In this prospective study, we present a detailed assessment of the rectal histology of 37 patients with acute Shigella infection. Biopsy specimens of Shigella patients were compared with those obtained from our control group, which consisted of patients with irritable bowel syndrome. Parasitic infections were carefully excluded in this group; for E. histolytica, in addition to stool examination by the concentration technique, culture of rectal biopsy and stools obtained through a sigmoidoscope was also carried out. Moreover, all these subjects had been under our observation for a sufficiently long time for us to be reasonably certain that they did not suffer from any organic pathology. In most patients with Shigella infection the histologic abnormalities involved the entire biopsy specimen, extending from the surface epithelium to the submucosa. Derangement of crypt architecture was seen in 31 of 37 patients (84%). In some, the crypt distortion was so severe as to mimic the abnormality seen in inflammatory bowel disease (Figure 2). This finding is in contrast to that reported by other workers. For example, Surawicz and Belic (10) did not find distorted crypt architecture in any of their patients with acute self-limited colitis. Similar findings were reported by Kumar et al. in their patients with acute infective colitis (11). Another frequent finding was the presence of crypt atrophy, seen in 86% of the patients (Figure 3). We believe that edema and increased cellular infiltrate in the lamina propria contributed to a large extent to the disorganization of the crypts seen in our patients. Moreover, the presence of crypt atrophy added to the general picture of crypt disarray. Another interesting observation was the type of cellular infiltrate seen in our patients. Other workers report a predominant neutrophilic response in the lamina propria in acute infective colitis (12,13). In contrast, the majority (27, 73%) of our patients showed either a predominant round cell or a combined round cell and neutrophilic infiltrate. Neutrophils tended to aggregate in the superficial zone of the mucosa, and all except 1 patient showed polymorphs infiltrating the surface epithelium-with one-third of the patients having a purulent mucosal exudate. Goblet cell depletion was present in the majority (25 of 37 patients, 68%). This finding differs from some earlier reports where mucus depletion of goblet cells was not considered a feature of acute infective colitis (10,12). When analyzed according to the severity of mucosal damage, mucus depletion was more frequent in patients with a severe disease than in those with mild abnormalities, the difference being statistically significant (82% vs. 36%,.I = 10.42, P < 0.01). The inflammatory process extended beyond the mucosa to involve the muscularis mucosae and submucosa in 92% and 80% of the patients, respectively (Figure 5). This finding again has not been emphasized in previous reports. Assessment of the mucosal abnormalities according to the duration of illness showed a remarkable degree of similarity between those with a short illness and those with a longer disease history. The only significant difference was observed with respect to goblet cell depletion, whch was more extensive (p < 0.01) in patients with disease for more than 48 h. To assess whether or not this effect was due to the severity of colitis, a comparison was made between patients with only mild inflammation in the two groups. Although the numbers were small, a definite trend could be discerned; only 2 of 11 patients with short illness (18%) had goblet cell depletion compared with all 3 patients who had diarrhea for more than 48 h; the difference was statistically significant (.I = 10.64, p < 0.01). Thus both the severity of colitis and the duration of illness were associated with mucus depletion of goblet cells. Most of the patients (31 of 37, 84%) were passing blood and mucus in the stools; only 6 patients (16%) had watery diarrhea. Four of these patients had mild colitis, whereas the remaining 2 patients showed moderate to severe abnormalities, despite the fact

7 660 ANAND ET A1. GASTROENTEROLOGY Vol. 90, No.3 that they were not passing any blood in their stools. This observation raises two issues. First, significant colitis may be present in patients with acute watery diarrhea and, therefore, all such patients should undergo sigmoidoscopy and rectal biopsy. The second issue concerns the pathophysiology of this defect. Animal experiments indicate that Shigella organisms have a dual pathogenic effect on the gastrointestinal tract. According to these studies, the pathogens release an enterotoxin that induces the secretion of water and electrolytes from the proximal jejunum in the absence of any bacterial invasion of the mucosa, resulting in watery diarrhea (14,15). The same organisms also produce the dysentery syndrome when they invade the colonic epithelial cells and initiate the inflammatory colitis (16,17). These preliminary observations have only been made in animals but, if extrapolated to humans, would suggest that the character of stools depends upon the pathogenic mechanism (enterotoxic or enteroinvasive) predominating in a particular patient. The present study offers some support for this hypothesis: 68% of the patients with blood and mucus in stools had moderate to severe colitis compared with 33% of the patients with watery diarrhea (the difference, however, was not significant). Shigella dysenteriae type 1 (Shiga bacillus) produced more severe rectal mucosal abnormalities than S. flexneri type 2. This finding is in keeping with the clinical observation that S. dysenteriae type 1 is the most virulent of the Shigella species (2). There is much debate as to whether it is possible to differentiate acute infective colitis from chronic inflammatory bowel disease on the basis of rectal mucosal appearances. In a recent study, Surawicz and Belic identified seven histologic features that in their opinion were highly discriminant of inflammatory bowel disease and helped to differentiate it from acute infectious colitis (10). These were villus surface epithelium, distorted crypts, crypt atrophy, increased numbers of both round cells and neutrophils in lamina propria, epithelioid granulomas, giant cells, and lymphoid aggregates. Of these, a villus surface epithelium, epithelioid granulomas, and giant cells were not seen in any of our patients and, therefore, if present, may favor the diagnosis of inflammatory bowel disease. However, these features were relatively uncommon in the study of Surawicz and Belic, being seen in only 27%, 25%, and 19% of patients, respectively. By contrast, the more frequent abnormalities of crypt disorganization (48%) and increased numbers of both round cells and neutrophils (37.5%) observed by these workers in idiopathic inflammatory bowel disease but not in acute self-limited colitis were present in our patients also. We therefore believe that in some instances it would be extremely difficult to differentiate acute infective colitis from inflammatory bowel disease, and only the clinical course and subsequent biopsy specimens will be of diagnostic help. References 1. Bagshawe AP, LaBrooy JT. Bacillary dysentery. In: Shearman DJC, Finlayson NDC. eds. Diseases of the gastrointestinal tract and liver. New York: Churchill Livingstone. 1982: Nalin DR. Shigellosis. In: Strickland GT, ed. Hunter's tropical medicine. 6th ed. Philadelphia: WB Saunders. 1984: Sen D, Saha MR, Niyogi SK. et al. Etiological studies on hospital in patients with acute diarrhoea in Calcutta. Trans R Soc Trop Med Hyg 1983;77: Pal Sc. Epidemic bacillary dysentery in West Bengal. India (lett). Lancet 1984;i: Sargeaunt PG. Baveja UK, Nanda R, Anand BS. Influence of geographical factors in the distribution of pathogenic zymodemes of Entamoeba histolytica: identification of zymodeme XIV in India. Trans R Soc Trop Med Hyg 1984;76: Nanda R, Baveja U. Anand BS. Entamoeba histolytica cyst passers: clinical features and outcome in untreated subjects. Lancet 1984;ii: Butzler JP. Skirrow ME. Campylobacter enteritis. Clin GastroenteroI1979;8: Edwards PRo Ewing WH. Identification of Enterobacteriaceae. 3rd ed. Minneapolis: Burgess Publishing, Sakazaki R. Tamura K. Prescott LM, Benzie Z. Sanual SC, Sinha R. Bacteriological examination of diarrhoeal stools in Calcutta. Ind J Med Res 1971;59: Surawicz CM. Belic 1. Rectal biopsy helps to distinguish acute self-limited colitis from idiopathic inflammatory bowel disease. Gastroenterology 1984;86: Kumar NB, Nostrant TT. Appelman HD. The histopathologic spectrum of acute self limited colitis (acute infectious-type colitis). Am J Surg PathoI1982;6: Dickinson RJ. Gilmour HM. McClelland DBL. Rectal biopsy in patients presenting to an infectious disease unit with diarrhoeal disease. Gut 1979;20: Day DW. Mandai BK. Morson Be. The rectal biopsy appearances in Salmonella colitis. Histopathology 1978;2: Keusch GT. Shigella infections. Clin Gastroenterol 1979; 8: Keusch GT. Donohue-Rolfe A. Jacewicz M. Shigella toxin(s): description and role in diarrhoea and dysentery. Pharmacol Ther 1982;15: Gemski P Jr. Formal SB. Shigellosis: an invasive infection of the gastrointestinal tract. In: Schlessinger D, ed. Microbiology. Washington. D.C.: American Society for Microbiology. 1975: Kinsay MD. Formal SB. Dammin GJ, Giannella RA. Fluid and electrolyte transport in rhesus monkeys challenged intracaecally with Shigella flexneri 2 a. Infect Immun 1976;14: