simple diagnosis causes · idiopathic inflammatory bowel disease fromtheothergroups. conclusion-at...

8JClin Pathol 1997;50:580-585

Simple objective criteria for diagnosis of causes ofacute diarrhoea on rectal biopsy

D Jenkins, A Goodall, B B Scott

AbstractAim-To identify simple, objective, accu-rate histological criteria for distinguish-ing acute infective-type colitis, chronicidiopathic inflammatory bowel disease,and irritable bowel syndrome on rectalbiopsy in patients with acute onset diar-rhoea at first presentation, one to 10 weeksafter onset.Methods-Cell counts and measurementsofmucosal architecture were made on ini-tial rectal biopsies from 18 patients withacute infective-type colitis, 17 patientswith first acute presentation of chronicidiopathic inflammatory bowel disease,and 23 patients with irritable bowel syn-drome. The data were analysed by ANOVAand discriminant analysis.Results-Lamina propria cells weremainly in the upper third in irritablebowel syndrome patients. Increasedlamina propria cellularity, mainly in themiddle third, and numbers of cryptintraepithelial neutrophils distinguishedacute infective-type colitis from irritablebowel syndrome in 93% of cases. Chronicidiopathic inflammatory bowel diseasediffered from irritable bowel syndromeand acute infective-type colitis in a de-creased number of crypts and alteredcrypt architecture. Chronic idiopathicinflammatory bowel disease showedhigher lamina propria cellularity, espe-cially in the basal third, with an increasednumber of lamina propria neutrophils.On discriminant analysis, crypt numbersdistinguished 86% of the cases of chronicidiopathic inflammatory bowel diseasefrom the other groups.Conclusion-At one week or more fromonset, acute infective-type colitis is char-acterised by a superficial increase inlamina propria cellularity, with only aslight increase in the number of poly-morphs. At this stage, chronic idiopathicinflammatory bowel disease is character-ised by a transmucosal increase in cellu-larity together with crypt loss andarchitectural abnormality. Thus,measurement of mucosal architectureestablishes simple, accurate, objective cri-teria for routine biopsy diagnosis ofchronic idiopathic inflammatory boweldisease from acute infective-type colitisand irritable bowel syndrome at initialpresentation, one to 10 weeks after onset.( Clin Pathol 1997;50:580-585)

Keywords: chronic idiopathic inflammatory boweldisease; infective colitis; biopsy diagnosis

The differential diagnosis of the causes of acutediarrhoea includes infective-type colitis,chronic idiopathic inflammatory bowel disease,and irritable bowel syndrome. Usually, thecolorectal mucosa in irritable bowel syndromeis regarded as histologically normal for inflam-mation and architecture. Rectal biopsy is usedroutinely to detect clinically important inflam-mation and distinguish acute infective-typecolitis from chronic idiopathic inflammatorybowel disease.' The histological features of dif-ferent types of infective colitis and of antibioticassociated colitis have been described.2`' Manyof the features used to diagnose acute infective-type colitis and chronic idiopathic inflamma-tory bowel disease in routine diagnosticpractice are not reproducible or accurate.5'6Specific histological criteria have not beenidentified for acute infective-type colitis,79although abnormal mucosal architecture andsome inflammatory features have been foundto distinguish chronic idiopathic inflammatorybowel disease from acute infective-type colitisand normality in some studies.5'-3

Previously, we reported the accuracy ofquantitative histology in separating chronic idio-pathic inflammatory bowel disease from irrita-ble bowel syndrome on rectal biopsy using theincreased cellularity in the basal third of thelamina propria, surface irregularity, and sur-face epithelial flattening.'2 The present studywas performed to obtain additional, objectivehistological criteria for routine use in thedifferential diagnosis of acute diarrhoea, in-cluding the problem of identifying acuteinfective-type colitis. Counts and measure-ments were made on rectal biopsy specimensfrom groups of patients with acute infective-type colitis, with acute presentation of chronicidiopathic inflammatory bowel disease con-firmed by subsequent follow-up, and with irri-table bowel syndrome. Statistically significantdifferences between the groups were investi-gated further using discriminant analysis toidentify diagnostic criteria.

MethodsThe study was a retrospective study of patientspresenting in routine practice to the Gastroen-terology Clinic at Lincoln County DistrictGeneral Hospital.

Division of Pathology,Queen's MedicalCentre, Nottingham,UKD Jenkins

Department ofMedicine, CountyHospital, Lincoln, UKA GoodallB B Scott

Correspondence to:Dr Jenkins, Division ofPathology, Queen's MedicalCentre, NottinghamNG7 2UH, UK.

Accepted for publication8 April 1997

580

on March 25, 2021 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.50.7.580 on 1 July 1997. Dow

nloaded from

http://jcp.bmj.com/

Diagnosis of causes of acute diarrhoea on rectal biopsy

Table 1 Measurements used in the analyses

Abbreviations Measurement

CIE Crypt intraepithelial neutrophils per10 crypts

CLUM Crypt lumen neutrophils per 10crypt lumens

LPP Lamina propria neutrophils/mmmucosal length

CELLEN Total lamina propria cells/mmmucosal length

HCDENS 1, 2, 3 Total cell densities in superficial,middle, and basal bands of laminapropria, respectively

HCDRAT Total cell density in superficial,horizontal band of laminapropria/density in basal band

SLR Measured surface length/smoothedsurface length

MHMEAN Mean of mucosal heightmeasurements

CRYPTS Number of crypts/mm surface lengthCLMEAN Mean of crypt length measurementsICDMEAN Mean of intercrypt distance

measurements

PATIENTS

Irritable bowel syndromeTwenty three patients with acute presentationof diarrhoea. There were nine men and 14women, median age 31 years (range 20-73),considered to have irritable bowel syndrome bystandard criteria and no relevant organicdisease at presentation or disclosed on followup for at least five years.

Acute presentation offirst onset of chronicidiopathic inflammatory bowel diseaseSeventeen patients with a total history up to thebiopsy of less than 10 weeks (median fiveweeks). There were nine men and eightwomen, median age 38 years (range 18-75).All had subsequent follow up of at least threeyears. Six patients had a final diagnosis ofulcerative colitis (four men, two women) withtypical sigmoidoscopic appearances and dis-ease above the sigmoid colon. Four hadCrohn's colitis (two men, two women). Thefinal diagnosis was based on characteristic his-tology and typical radiological appearances. Sixhad distal proctosigmoiditis (two men, fourwomen) with sigmoidoscopically continuousdisease. In three patients the disease extendedinto the sigmoid colon but not into thedescending colon. One man had chronicidiopathic inflammatory bowel disease of inde-terminate type with no firm diagnosis ofCrohn's disease or ulcerative colitis.

Acute infective-type colitisEighteen patients with acute diarrhoea. Therewere 10 men and eight women, median age 51years (range 21-80). All had a stool culture:five were positive for salmonella, four forcampylobacter, four for Clostridium difficile, andone for Yersinia spp. Four had antibiotic associ-ated diarrhoea, which was self-limiting and waspresumed to be infective, although C difficileculture was not performed. The median dura-tion of disease was two weeks (range one toeight).

PROCEDUREFormalin fixed rectal biopsy specimens takenduring routine management were embedded

into paraplast and cut at 1 gm thicknessperpendicular to the mucosal surface. Sectionswere stained with haematoxylin and eosin.Measurements and cell counts were carried outon the stained sections using a MOPPETimage analyser. All processing, measurements,and counts were performed without knowledgeof the diagnostic group to minimise experi-mental and observer bias. The method hasbeen described in detail previously.'2 Themeasurements are shown in table 1. None ofthe biopsies showed granulomas or giant cellsand these, therefore, were not studied furtheras diagnostic criteria.

STATISTICAL ANALYSISThe measurements were log transformed whennecessary to normalise the distribution. TheMann-Whitney U test was used to compare ageand duration of disease. All other analyses werecarried out on the Amdahl computer at theUniversity College (London) Computer Cen-tre using package BMDP (University ofCalifornia, 1985). Comparison by ANOVAand discriminant analysis were performed.Discriminant analysis (BMDP7D) computes alinear classification for each of the selectedgroups in a stepwise manner; at each step thevariable that adds most to the separation of thegroups is entered into the function. A "jack-knife" procedure was then used. This proce-dure simulates a separate dataset by removingone case, using all the remaining cases to pro-duce a discriminant function, and then classify-ing the excluded case with this function. Thewhole process is repeated for each case in theset.

ResultsTable 2 shows the results of selected measure-ments for each diagnostic group with statisti-cally significant differences between thegroups.

ACUTE INFECTIVE-TYPE COLITIS VERSUSIRRITABLE BOWEL SYNDROMEIn acute infective-type colitis there was in-creased lamina propria cellularity comparedwith irritable bowel syndrome. In all cases ofirritable bowel syndrome cell density wasgreatest in the superficial third of the laminapropria compared with the middle or basalthirds. This pattern usually is considerednormal. In acute infective-type colitis theincreased lamina propria cellularity was in themiddle third or the basal third. The increasewas usually less than twofold. Importantly, in95% of the cases of acute infective-type colitisthe cell density was greater in the uppertwo-thirds of the lamina propria compared tothe basal third (fig 1).

Neutrophil counts were increased in thecrypt epithelium and lamina propria in acuteinfective-type colitis compared with irritablebowel syndrome. Some increase in the numberof crypt intraepithelial neutrophils (> 10 percrypt lumen over 10 crypt lumens) was seen in39% of acute infective-type colitis comparedwith 4% of irritable bowel syndrome, and smallnumbers of intraluminal neutrophils (> 10 per

581



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

52enkins,Goodall, Scott

10

E

inCD+JQ0

0-

8

6

4

2I-

0

3 (DX._CoQ0

a

CL2 X

._

E

=

Fa)

u)

1 L:._UC;._

vControls INF IBD

acute

Figure 1 Bar chart showing median and 95% confidencelimits of measurements of superficiallbasal lamina propriacell density (thick lines) (HCDRAT) and crypt counts permillimetre surface (thin lines) (CRYPTS). Controls,irritable bowel disease patients; INF, acute infective-typecolitis; IBD, chronic idiopathic inflammatory bowel disease.

crypt lumen over 10 crypt lumens) werepresent in 17% of acute infective-type colitisbut not in irritable bowel syndrome. Surfaceintraepithelial neutrophils were increased inonly a few acute infective-type colitis cases.

In discriminant analysis, total lamina propriacellularity was the most important variable dis-tinguishing acute infective-type colitis fromirritable bowel syndrome, with crypt intraepi-thelial neutrophils also making a contribution.On jack-knife testing a discriminant functionderived from these correctly classified 93% ofcases.

CHRONIC IDIOPATHIC INFLAMMATORY BOWEL

DISEASE VERSUS ACUTE INFECTIVE-TYPE COLITIS

AND IRRITABLE BOWEL SYNDROMEIn chronic idiopathic inflammatory boweldisease lamina propria cellularity was increasedtwo- to threefold compared with irritable bowelsyndrome and was up to twofold greater thanin acute infective-type colitis. The increase wasmost marked in the basal third compared withboth irritable bowel syndrome and acuteinfective-type colitis (fig 1).There was a reduction in crypt numbers per

millimetre ofsurface epithelium in chronic idio-pathic bowel disease, with significant differ-ences in intercrypt distance and crypt volume.

The measures of crypt and mucosal architec-tural change correlated highly. In discriminantanalysis, crypt counts alone classified 86% ofcases of acute infective-type colitis and chronicidiopathic inflammatory bowel disease cor-rectly (fig 1).Although lamina propria counts of up to

100/mm were found frequently (83%) in acuteinfective-type colitis, very high counts oflamina propria polymorphs (> 1000/mm),crypt intraepithelial neutrophils (> 10 percrypt over 10 crypts) and crypt lumenneutrophils (> 10 per crypt lumen over 10crypt lumens) were infrequent (6%, 12%, and7%, respectively) but unique features ofchronic idiopathic inflammatory bowel disease.

DiscussionThe study defined objective, histological fea-tures that distinguished accurately acuteinfective-type colitis, chronic idiopathic inflam-matory bowel disease, and irritable bowel syn-drome in colorectal biopsies taken at first pres-entation from patients with acute onsetdiarrhoea. The cases studied represent routine,British district hospital practice and reflect theusual time of presentation after onset of symp-toms of patients with acute infective-type coli-tis (median two weeks), and chronic idiopathicinflammatory bowel disease (median fiveweeks). In this type of referral practice very fewpatients are seen in the first week of acute diar-rhoea. Other studies of patients during the veryearly stage of acute colitis have suggested thatthe architectural changes described here inacute onset chronic idiopathic bowel diseasemay not be present during the first week.'4 15

Also, more intense acute inflammation witherosion and ulceration has been described inacute infective-type colitis during the very earlyphase.SA The findings of this study, therefore,apply to biopsy diagnosis of colorectal biopsiestaken for the investigation of suspected chronicidiopathic inflammatory bowel disease of acuteonset in routine practice at one to 10 weeksafter onset. The purpose of this study was notto separate Crohn's disease from ulcerativecolitis, as clinical management at this stage isdetermined mainly by the need to separatechronic idiopathic inflammatory bowel diseasefrom acute infective-type colitis, and fromfunctional causes of diarrhoea (irritable bowelsyndrome). The distinction of Crohn's disease

Table 2 Derived means and 95% confidence limits (CL) for measurements in each diagnostic group

Controls.(IBS) (n=23) Acute infective-type colitis (AIC) (n=18) Chronic idiopathic inflammatory bowel disease (n=1 7)

Variable 95% CL Mean 96% CL Mean p value (v IBS) 95% CL Mean p value (v IBS) p value (v AIC)

CIE 0-2 1 3-14 7 <0.001 10-59 25 <0.001CLUM 0-1 0 0-14 1 1-5 3 <0.01LPP 8-14 11 31-93 53 <0.001 86-269 160 <0.001 <0.05CELLEN 1255-1490 1368 1885-2622 2223 <0.001 2675-3999 3271 <0.001 <0.01HCDENS1 6695-8915 7727 7459-9984 8630 6900-9484 8089HCDENS2 4587-6287 5370 7502-9568 8472 <0.001 7834-9913 8812 <0.001HCDENS3 3656-4820 4198 5369-7702 6427 <0.001 7661-10150 8818 <0.001 <0.05HCDRAT 1.73-2.01 1.87 1.16-1.72 1.44 <0.05 0.82-1.009 0.96 <0.001 <0.01SLR 1.04-1.06 1.05 1.05-1.11 1.08 1.08-1.16 1.12 <0.01MHMEAN 369-427 398 436-528 482 <0.01 517-628 572 <0.001CRYPTS 5.8-7.5 6.6 7.3-8.3 7.8 4.1-5.3 4.7 <0.001 <0.001CLMEAN 333-378 356 385-474 430 <0.05 451-569 497 <0.001ICDMEAN 28-38 32 24-33 29 39-58 48 <0.001 <0.001

IBS, irritable bowel syndrome.

n

582



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


V.~~~~~~~~~~~~

Figure 2 Irritable bowel syndrome showing the superficial distribution of cells in the upperthird of the lamina propria. Note also the "test-tube rack" arrangement of crypts.

Figure 3 Acute infective-type colitis showing a twofold increase in cells in the mid-zone ofthe lamina propria. This produces the appearance of a "mild superficial increase in chronicinflammation".

Figure 4 Chronic idiopathic inflammatory bowel disease. Crohn 's disease showingincreased cellularity in the basal third of the lamina propria. This produces the appearanceof a "moderate or severe transmucosal increase in chronic inflammation".

from ulcerative colitis becomes important inany subsequent surgical management. Themeasurements used to distinguish chronic idio-pathic inflammatory bowel disease from acuteinfective-type colitis have been shown previ-ously to identify both Crohn's disease andulcerative colitis.'2

The biopsy appearances of irritable bowelsyndrome usually are considered "normal".Irritable bowel syndrome was used in this studyas the "control" group because it comprises themain group of individuals who can presentwith acute diarrhoea and initially are suspectedof having the diseases studied but turn out notto have acute colitis or any other organicdisease. The main objective differences identi-fied by discriminant analysis between acuteinfective-type colitis and normal/irritablebowel syndrome are increased lamina propriacellularity and numbers of neutrophils in acuteinfective-type colitis. In normal/irritable bowelsyndrome the maximal lamina propria cellular-ity was beneath the surface epithelium in theupper third (figs 1 and 2). This distribution wasaltered in acute infective-type colitis by a mild(up to twofold) increase, affecting mainly themiddle third giving a "mild superficial increasein lamina propria cellularity" in 95% of cases(figs 1 and 3). The main differences identifiedbetween chronic idiopathic inflammatorybowel disease and acute infective-type colitis ornormal/irritable bowel syndrome were bothfeatures that are typical of chronic idiopathicinflammatory bowel disease: reduced cryptnumbers (which correlate with architecturalirregularity) and a two- or threefold increase incellularity in the basal third of the lamina pro-pria, produc ing a "moderate or severetransmucosal increase in lamina propria cellu-larity" (figs 1 and 4).Numbers of neutrophils provided some

additional discrimination between acuteinfective-type colitis and irritable bowel syn-drome. There were very few in irritable bowelsyndrome/normal, with an upper 95% confi-dence limit of fewer than two neutrophils per10 crypts. Neutrophils were increased in acuteinfective-type colitis and in chronic idiopathicinflammatory bowel disease. However, veryhigh lamina propria neutrophil counts, severecryptitis, and crypt abscesses (as definedabove) were found only (and infrequently) inchronic idiopathic inflammatory bowel disease.The relatively low neutrophil counts in acuteinfective-type colitis in this study may reflectthe late presentation (one to eight weeks afteronset) in these patients. Biopsies taken duringthe first few days of a severe acute infective epi-sode may show much heavier neutrophilinfiltration.2 4 8

Generally, there is good agreement betweenthe findings of this study and studies ofconventional biopsy diagnosis.5-9 1-6 Both ap-proaches suggest that distorted crypt architec-ture, crypt atrophy, and mixed lamina propriainflammation are important distinguishing fea-tures of chronic idiopathic inflammatory boweldisease. Previously, we have identified theimportance of the diffuse transmucosal in-crease in lamina propria cellularity for distin-guishing chronic idiopathic inflammatorybowel disease from normal/irritable bowel syn-drome cases. Other studies have drawn atten-tion to different patterns of increased basal cel-lularity in diagnosing chronic idiopathicinflammatory bowel disease.57 9 1-13 These in-clude basal lymphoid aggregates, basal giant

583



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

54enkins, Goodall, Scott

cells, and basal plasmacytosis. None of thebiopsies in this study showed granulomas orgiant cells. Although both are reported as spe-cific diagnostic features in other series, theirabsence in this series confirms their poor diag-nostic sensitivity.We agree with the conclusions of Surawicz

and Belic7 and Allison et al9 that normal cryptarchitecture is an important feature distin-guishing acute infective-type colitis fromchronic idiopathic inflammatory bowel disease,even at acute, first onset. Our study did not findany evidence in the British population ofdistorted crypt architecture in acute infective-type colitis comparable to that reported inIndian patients with shigellosis or any increasein crypt architectural abnormality with timeduring the period (one to eight weeks)studied.'7 18 Although the extremes of theranges of architectural measurements approxi-mated between chronic idiopathic inflamma-tory bowel disease and other groups, there wasminimal overlap. Our data suggest that mostpatients who subsequently are proven to havechronic idiopathic inflammatory bowel diseasehave abnormal crypt architecture within a fewweeks of initial clinical onset. Comparisonbetween acute chronic idiopathic inflammatorybowel disease at first onset and its more chronicform showed that the decrease in cryptnumbers was similar to that reported previ-ously in longstanding chronic idiopathic in-flammatory bowel disease with no correlationbetween the duration of disease and cryptresponse.The series agrees with other studies' 9 11

that it is possible to diagnose acute infective-type colitis by histology beyond the first fourdays as proposed by Nostrant et al and suggeststhat there is a characteristic histological patternin the late/resolving phase of acute infective-type colitis (one to eight weeks).8 This latephase diagnostic pattern comprises a mildincrease in lamina propria cellularity confinedmainly to the upper two-thirds and slightincreases in neutrophil numbers in cryptepithelium and lamina propria. The late phasepattern is that seen routinely in UK practiceand may explain the appearance reported bysome pathologists as "non-specific proctitis orcolitis".5 It is possible that many patientslabelled as having "non-specific inflammation"have resolving acute infective-type colitis.Understanding the time sequence of the evolu-tion of acute infective-type colitis and chronicidiopathic inflammatory bowel disease to-gether with the use ofbetter criteria to separateirritable bowel syndrome, acute infective-typecolitis, and chronic idiopathic inflammatorybowel disease should reduce the need for thisunsatisfactory "non-specific" category.The objective changes described can be used

to refine conventional histopathological diag-nosis. Crypt architectural abnormality can bedetected by conventional microscopy as changefrom the straight, close-packed, parallel, "test-tube rack" arrangement of crypts in irritablebowel syndrome or acute infective-type colitisto the fewer, irregular crypts of chronicidiopathic inflammatory bowel disease (figs

2-4). Often, subjective detection of such cryptabnormalities is not easy.6 Crypt countsprovide an easy accurate alternative and can bemade directly using a microscope micrometer.A reduction in numbers of crypt openings tofive or fewer per millimetre of surface epithe-lium measured along one 3 mm of surface epi-thelium in a single section was pathognomonicof chronic idiopathic inflammatory boweldisease (fig 1).The pattern of lamina propria cellularity (fig

1) usually can be detected easily in conven-tional examination. In normal/irritable bowelsyndrome the cells are mostly in the upperthird (fig 2). In acute infective-type colitis usu-ally there is a mild superficial increase, involvingthe middle third (fig 3), and in chronicidiopathic inflammatory bowel disease there iseither a moderate or severe transmucosalincrease (fig 4) in active disease or involvementof the basal part by lymphoid aggregates, giantcells, or plasmacytosis.Polymorph numbers are useful in distin-

guishing normal/irritable bowel syndromefrom active inflammation but need to be inter-preted carefully in attempting to separate acuteinfective-type colitis and chronic idiopathicinflammatory bowel disease. There is a wideoverlap and any diagnostic differences dependon the timing of the biopsy in relation to pres-ence and onset of clinical activity.

Precise measurement has refined diagnosticcriteria for distinguishing acute infective-typecolitis, chronic idiopathic inflammatory boweldisease, and irritable bowel syndrome at one to10 weeks after clinical onset. The preciselydefined changes in lamina propria cellularity inthe late phase of acute infective-type colitis andin chronic idiopathic inflammatory boweldisease, and the reduced crypt numbers andaltered architecture in chronic idiopathic in-flammatory bowel disease can be used to makeaccurate conventional biopsy diagnoses ofpatients presenting with acute diarrhoea.

We are grateful to the National Association for Colitis andCrohn's Disease for financial support. We also thank Mr P MStephenson for technical help.

1 Surawicz CM. Diagnosing colitis: biopsy is best. Gastroenter-ology 1987;92:538-40.

2 Price AB, Jewkes J, Sanderson PJ. Acute diarrhoea, Campy-lobacter colitis and the role of rectal biopsy. J Clin Pathol1979;32:990-7.

3 Price AB, Day DW. Pseudomembranous and infective coli-tis. In: Anthony PP, MacSween RN M, eds. Recent advancesin histopathology 11. Edinburgh: Churchill-Livingstone,1981:99-117.

4 Talbot IC, Price AB. Biopsy pathology in colorectal disease.London: Chapman & Hall Medical, 1987.

5 Theodossi A, Spiegelhalter DJ, Jass J, Firth J, Dixon M,Leader M, et al. Observer variation and discriminatoryvalue of biopsy features in inflammatory bowel disease. Gut1994;35:961-8.

6 Cooke MG, Dixon MF. An analysis of the reliability ofdetection and diagnostic value of various pathological fea-tures in Crohn's disease and ulcerative colitis. Gut 1973;14:255-62.

7 Surawicz CM, Belic L. Rectal biopsy helps to distinguishacute self-limited colitis from idiopathic inflammatorybowel disease. Gastroenterology 1984;86:104-13.

8 Nostrant TT, Kumar NB, Appelman HD. Histopathologydifferentiates acute self-limited colitis from ulcerative coli-tis. Gastroenterology 1987;92:318-28.

9 Allison MC, Hamilton-Dutoit SJ, Dhillon AP, Pounder RE.The value of rectal biopsy in distinguishing self-limitedcolitis from early inflammatory bowel disease. QuarJ Med1987;65:985-95.

584



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


10 Thompson EM, Price AB, Altman DG, Sowter C, Slavin G.Quantitation in inflammatory bowel disease using compu-terised interactive image analysis. Clin Pathol 1985;38:631-8.

11 Surawicz CM, Haggitt RC, Husseman M, McFarland LV.Mucosal biopsy diagnosis of colitis: acute self-limited coli-tis and idiopathic inflammatory bowel disease. Gastroenter-ology 1994;107:755-63.

12 Jenkins D, Goodall A, Drew K, Scott BB. What is colitis?Statistical approach to distinguishing clinically importantinflammatory bowel change in rectal biopsy specimens. .7

Clin Pathol 1988;41:82-9.13 Seldenrijk CA, Morson BC, Meuwissen SGM, Schipper

NW, Lindeman J, Meijer CJLM. Histopathological evalua-tion of colonic mucosal biopsy specimens in chronicinflammatory bowel disease: diagnostic implications. Gut1991;32:1514-20.

14 Schumacher G. First attack of inflammatory bowel diseaseand infectious colitis. Scand .7 Gastroenterol 1993;28(suppl198):1-24.

15 Therkildsen MH, Jensen BN, Tegibjaerg PS, RasmussenSN. The final outcome of patients presenting with theirfirst episode of acute diarrhoea and an inflamed rectalmucosa with preserved crypt architecture. A clinicopatho-logic study. ScandJ7 Gastroenterol 1989;24: 158-64

16 Kumar NB, Nostrant TT, Appleman HD. The histopatho-logic spectrum of acute self-limited colitis (acuteinfectious-type colitis). Am _7 Surg Pathol 1982;6:523-9.

17 Anand BS, Malhotra V, Battacharya SK, Datta P, Datta D,Sen D, et al. Rectal histology in acute bacillary dysentery.Gastroenterology 1986;90:654-60.

18 Mathan MM, Mathan VI. Morphology of rectal mucosa ofpatients with shigellosis. Rev Infect Dis 1991;13(suppl4):S314-18.

585



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

simple diagnosis causes · idiopathic inflammatory bowel disease fromtheothergroups. conclusion-at...

Documents