Journal of Clinical Pathology, 1978, 31, 567-577

Overlap in the spectrum of non-specific inflammatorybowel disease 'colitis indeterminate'ASHLEY B. PRICE

From Northwick Park Hospital and CRC, Harrow, Middlesex (work carried out at St. Mark's Hospital,City Road, London EC1)

SUMMARY It is stated that 10-20O,' of cases of non-specific inflammatory bowel disease cannot beclassified. Thirty such cases, designated colitis indeterminate at the time of colectomy, were identifiedfrom the pathology files of St. Mark's Hospital. The histopathological features of the surgicalspecimens and any available biopsy specimens were studied. In nearly all the cases urgent surgery

had been required and the features of incipient or established fulminating disease were present.The pathology of these cases of Crohn's disease and ulcerative colitis overlapped, and differentiatingfeatures were scant or unreliable. Accepted criteria of Crohn's disease namely, fissuring ulceration,transmural inflammation, and a maintained goblet-cell population-were found in cases sub-sequently proved to be ulcerative colitis. Disease activity greatly affected the evaluation ofmorphological features. Many of the difficulties were resolved when biopsy material obtainedduring a quiescent phase was examined. These specimens gave a dynamic perspective of thedisease process, often more valuable than the static, non-specific picture of acute disease seen inthe surgical specimens. Cases of colitis indeterminate form a small but distinctive group in thespectrum of inflammatory bowel disease which is characterized by a common pattern of pathologythat presents a diagnostic dilemma.

The differential diagnosis between Crohn's diseaseand ulcerative colitis can be made in most cases onthe cumulative clinical, pathological, and radiologicalevidence (Lockhart-Mummery and Morson, 1960;Lennard-Jones et al., 1968; Marshak and Lindner,1975). Lockhart-Mummery and Morson (1964)and others (Farmer et al., 1968; Glotzer et al.,1970) have described the many distinguishinghistopathological features. Recently, the diagnosticvalue of these features has been tested by clusteranalysis and observer reliability studies in order toselect those that were most dependable (Hywel-Joneset al., 1973; Cook and Dixon, 1973). However,owing to the variable occurrence of even the mostreliable features the diagnostic difficulty stillremains (Schachter et al., 1970). Thus in most seriesof patients with non-specific inflammatory largebowel disease a confident diagnosis cannot bemade in 10-15% of cases (Hawk and Turnbull,1966; Glotzer et al., 1970; Kent et al., 1970).The difficulties in diagnosis are due to overlapping

Received for publication 20 October 1977

in the pathological features of Crohn's disease andulcerative colitis, which may even be oppositeends of a spectrum of one disease (Lewin and Swales,1966; Margulis et al., 1971). But few, apart fromLewin and Swales (1966) and Glotzer et al. (1970),have studied the pathology of these indeterminatecases in any detail. This paper reports a study oftheir histopathology which aimed firstly, to investi-gate the reason for the diagnostic difficulty, and,secondly, from clinical correlations, to try to placethe group within the overall perspective of non-specific inflammatory bowel disease.

Material and methods

At St. Mark's Hospital between 1960 and 1973there were 330 colectomies or proctocolectomiesfor ulcerative colitis and Crohn's disease. All thegross specimens had been photographed. Routineparaffin-embedded 5-,Lm haemotoxylin and eosin-stained sections from up to eight standard anatomicalsites together with additional blocks from sites ofspecial interest were available in all cases. Thehistopathological diagnoses had been made on the

567

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from

568

criteria of Lockhart-Mummery and Morson (1960),Cook and Dixon (1973), and Schachter and Kirsner(1975). This pool of material formed the backgroundfor the study.The cases that had caused diagnostic difficulty

were identified from the hospital's pathology filesunder three headings: (1) colitis unclassified,'not further categorised'; (2) colitis unclassified,'probably/possibly Crohn's disease'; and (3) colitisunclassified, 'probably/possibly ulcerative colitis'.A case had been indexed as colitis unclassified whenno confident histopathological diagnosis was madefrom the surgical specimen. When some diagnosticattributes were present but were insufficient for aconfident diagnosis, or atypical features were presentthat prevented a confident diagnosis, the categorycolitis unclassified had been qualified by the ap-propriate suffix, probably/possibly Crohn's diseaseor probably/possibly ulcerative colitis. Thirty suchcases were found: 13 colectomies and 17proctocolec-tomies. The pathological diagnosis on the surgicalspecimen had been made without reference to anyavailable biopsy material.From here onwards in this paper I substitute

the term 'colitis indeterminate' in preference tothat of 'colitis unclassified' used in the hospitalpathology files, and, for simplicity, the subdivisionsare considered as one in Table 1.The surgical specimens from 30 confidently

indexed cases of colonic Crohn's disease and from 30confidently indexed cases of ulcerative colitis fromthe same period were selected for comparison.It was soon evident that the indeterminate groupwere almost all examples of acute severe disease,and two additional control groups were therefore

Ashley B. Price

selected. They comprised confidently indexed casesthat had undergone urgent surgery, this beingdefined as surgery performed within 24 hours ofthe decision to operate (Ritchie, 1974). Thirtyexamples of ulcerative colitis were chosen, butonly two such examples of Crohn's colitis could beidentified from the records. The mean number ofsections examined per case was 11.Without knowledge of the diagnostic category

each case was assessed for a limited number ofattributes with positive discriminating value forCrohn's disease or ulcerative colitis. The attributeswere derived from the work of Lockhart-Mummeryand Morson (1960, 1964), Lennard-Jones (1971),and Cook and Dixon (1973). They are shown inTable 1.

After this the available preoperative and post-operative material was examined and classifiedas either 'diagnostic' or 'not diagnostic'. Ulcerativecolitis was diagnosed when the accepted collectiveappearances of mucin depletion, diffuse mucosalinflammation, and glandular irregularity or atrophywere seen. Crohn's disease was diagnosed when agranuloma was noted or when the following com-bination of features was seen: focal mucosalinflammation, regular glands with preserved epith-elial mucin, aggregates of lymphocytes adjacent tocrypts, submucosal inflammation (Price and Morson,1975). Crohn's disease was also diagnosed if anormal rectal mucosa was found at follow-upexamination in a case with a longstanding and typicalclinical history.

There were preoperative biopsies in 23 cases,which had been carried out at a time varying fromone day to as long as six years before surgery.

Table. 1 Incidence of selected discriminating attributes in each diagnostic category

Colitis indeterminate Ulcerative colitis Crohn's disease

Urgent surgery Elective surgery Urgent surgery Elective surgery

Number of casesAttributes suggestive of Crohn's disease:

Discontinuous disease orcontinuous disease but unevendistributionGranulomasTypical 'Crohn's' fissuresTransmural inflammationNormal goblet cells with a regularglandular pattern and focal mucosalinflammation

Attributes suggestive of ulcerative colitis:Continuous disease involving rectumGoblet-cell depletion with glandularirregularity ± diffuse mucosalinflammation

Attributes suggestive of acute disease idilatation:

Myocytolysis and capillaryengorgementAcute V-shaped clefts

30

214

30 30

2

42820

205

2

2222

1410

30

27

231230272

2825

3028

21 (13 toxicmegacolon)18

33

22 (10 toxicmegacolon)6

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from

Overlap in the spectrum ofnon-specific inflammatory bowel disease

From 10 of these 23 cases follow-up material wasavailable. The follow-up period ranged from fourmonths to nine years.The clinical details were studied along with the

radiological findings, and a final clinicopathologicalcorrelation was made.

Results

CLINICAL ACTIVITY AND OCCURRENCE OFACUTE DILATATIONIt was obvious early in the study that. the activityof the disease process at the time of major surgeryaffected the subsequent pathology. Of the 30 cases,27 (90%) indexed as colitis indeterminate hadundergone urgent surgery. The figure contrastswith the overall hospital incidence of 30% amongthe confidently indexed cases of Crohn's diseaseand ulcerative colitis seen during the same period.Thirteen cases in the indeterminate group hadradiologically proved toxic megacolon as had 10cases among the confidently indexed cases ofulcerative colitis (Table 1). The urgent surgery in thetwo indexed cases of Crohn's disease was forlocalised perforation not for incipient dilatation.

PATHOLOGYTwo additional criteria were recorded (Table 1)that reflected the incidence of established or incipientcolonic dilatation. These were myocytolysis withtransmural capillary engorgement and multipleV-shaped clefts or fissures (Roth et al., 1959). These

5 10 15 20 25 30

(cm)

and the other features presented in the Table arenow described in more detail.

Gross pathologyFifty-eight of the confidently indexed cases ofulcerative colitis and 29 of the 32 confidentlyindexed cases of Crohn's disease had characteristicmacroscopic appearances. In the indeterminategroup 28 cases had a continuous pattern of disease,and in 23 of them this amounted to a total colitis.In 14 the distribution was uneven, regardless of itsextent, and conformed to one of two patternsboth of which would normally bias towards adiagnosis of Crohn's disease.The first pattern was that of relative rectal

sparing, sufficient to suggest segmental disease(Fig 1). Of the seven cases with this pattern,three were considered to be 'probably ulcerativecolitis' on cumulative histological evidence. Inonly two cases was a diagnosis of Crohn's diseasethought to be more likely. On histological examina-tion the rectal mucosa was mildly abnormal in allseven. In the three judged to be 'probably ulcerativecolitis' glandular irregularity was noted, but in theremainder the inflammation was mild and thefeatures were equivocal.The second pattern, that of intermittent ulceration

(Fig. 2), was also seen in seven cases. This superficialresemblance to skip-like lesions was due to thevariable intensity of the disease in different parts ofthe colon. The histology of the mucosa betweenthe ulcers was of mild mucosal inflammation in

35 40 45 50 55 60 65In a a

Fig. 1 A total proctocolectomyin a case indexed colitis in-determinate-probably/possiblyulcerative colitis. Relative rectalsparing is seen which, with thenormal ascending colon, givesthe impression of a segmentaldisease pattern. There is severeulceration and dilatation of thetransverse colon.

3

569

*

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from

j_ 2 Fig. 2 A total colectomy ina case indexed colitis in-determinate-probablylpossibly ulcerative colitis.The variable intensity ofdisease in a severe totalcolitis with its patchy pattern

P> | a g v e l 1 e U of ulceration leaves areasthat may, on superficialinspection, be mistaken forskip-like lesions.

0, SW..Za Fig. 3 A typical fissure in

Crohn's disease. The tract4, is lined by inflammatory

cells with healthy muscleon each side. Accompany-ing mucosal ukceration is a

a ~~~~~~~~minorfeature. (Haema-2KtI ~~~toxylin and eosin x 10)

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from

Overlap in the spectrum ofnon-specific inflammatory bowel disease

Fig. 4 Multiple acute clefts with myocytolysis and capillary engorgement. These 'fissures' are

characteristic offulminant colitis, in this instance from a case indexed colitis indeterminate,probably ulcerative colitis. The diagnosis was confirmed by the histology of the subsequentproctectomy. (H and E x 15)

Fig. 5 Severe myocytolysis and capillary engorgement, histological markers of incipient or establisheddilatation (H and E x 55)

571

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from

Ashley B. Price

all instances. Three of these seven cases showed areasof glandular irregularity that were more like ulcer-ative colitis. The cumulative evidence favouredCrohn's disease as the probable diagnosis in twoothers of this group, and in two of the seven notentative diagnosis could be offered.

HistopathologyGranulomas Granulomas are considered to bean absolute distinction between Crohn's disease andulcerative colitis. They were present in 25 of the 32confidently indexed cases of Crohn's disease andwere absent in all the other categories of cases.

Fissuring ulceration Two qualitatively differentforms of fissuring ulceration could be distinguished.Characteristic fissuring of Crohn's disease was seenin 14 of the 32 confidently indexed cases. Thefissures were mostly single and narrow, had aprominent lining of inflammatory cells, and extendeddown into muscle (Fig. 3). They were not usuallyfound in areas of extensive mucosal denudation.The muscle nearby, apart from the presence of anaggregated pattern of lymphocytes, was intact.They were also seen in four cases in which the otherhistological criteria were so atypical (Table 1)

that a diagnosis of Crohn's disease was not justified.A second form of fissuring ulceration was present

in 18 of the 30 cases indexed as colitis indeterminate.These were squat, V-shaped clefts, sparsely lined byinflammatory cells, often multiple, and usually inareas of extensive ulceration. Accompanying myo-cytolysis and capillary engorgement were invariablypresent in the adjacent bowel wall (Figs. 4, 5). Theywere seen in six patients who had undergoneurgent surgery but in whom the cumulative attributesallowed a confident diagnosis of ulcerative colitisto be made.

It was not always possible to separate the twoforms of fissure from each other or from deepundermining ulceration. The latter was a prominentfeature in all cases of ulcerative colitis that hadrequired urgent surgery (Fig. 6). Only unequivocalexamples of each form were recorded.

Lymphoid aggregates and transmural inflammationTwo cases among those indexed as colitis inde-terminate showed a transmural pattern of aggregatedgroups of lymphocytes. Even in these two it waspoorly developed. It is a feature accepted as highlysignificant in the diagnosis of Crohn's disease(McGover and Goulston, 1968) and was present

Fig. 6 Florid undermining ulceration from a case confidently indexed ulcerative colitis. The muscle layer hasnot been breached, myocytolysis and the vascular element are absent (H and E x 10)

572

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from

Overlap in the spectrum ofnon-specific inflammatory bowel disease

Fig. 7 An acute fissure with an adjacent surviving mucosal island. Note the sur-prising lack of inflammation in the mucosa with such a severe disease process present.This case was indexed colitis indeterminate probably Crohn's disease. (H and E x 22)

in 26 of the 32 control cases of Crohn's colitis.It was absent from all indexed cases -of ulcerativecolitis. In addition, non-aggregated, full-thickness,transmural inflammation with a mixed cellularpattern was seen in all the confidently indexed casesof Crohn's disease. It was also present in 28 of the 30cases indexed as colitis indeterminate, and in 20 ofthe control cases of ulcerative colitis that hadrequired urgent surgery (Table 1). In all the casesindexed ulcerative colitis (urgent surgery) andcolitis indeterminate the transmural inflammationwas related to regions of severe ulceration.

Mucosal glandular pattern and goblet-cell populationA regular glandular pattern with a preservedgoblet-cell population and a focal distribution ofmucosal inflammation was the dominant featurein 29 of the 32 confidently indexed cases of Crohn's

disease. Mucin depletion, an irregular or atrophicmucosa, and diffuse mucosal inflammation pre-dominated in 53 of the 60 confidently indexed casesof ulcerative colitis. In the colitis indeterminategroup these features were obscured or modifiedby extensive ulceration and stripping of the mucosa.However, in islands of surviving mucosa theinflammation was surprisingly mild, the epitheliumregular, and the goblet-cell population well main-tained (Fig. 7)-features that favour a diagnosisof Crohn's disease but seen here in cases that oncumulative evidence had been judged as 'inde-terminate', which included four of the six casessubsequently proved to be ulcerative colitis.

Biopsy dataA confident diagnosis was made in seven casesfrom the indeterminate group of 23 with preopera-

573

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from

574

tive biopsy reports available. In three cases thefindings in a series of biopsies wele diagnosticof ulcerative colitis. At the time of surgery the acutedisease obscured any diagnostic pathology. Fourcases could be classified as Crohn's disease beforesurgery. In two the findings in repeated rectalbiopsies were normal, and in the other two, granu-lomas were present. Again, at surgery acute changespredominated. Furthermore, no granulomas werefound.

Follow-up material permitted a confident diag-nosis in an additional eight cases. Three wereclassified as ulcerative colitis and all showed acharacteristic follicular proctitis in the rectal stump.The other five cases were indexed Crohn's disease.In two, the findings in multiple follow-up rectalbiopsies remained normal, and in a third case theyshowed the characteristic inflammatory pattern ofCrohn's disease. In the remaining two, granulomaswere found in follow-up biopsies, one of whichwas four years after colectomy.A final confident diagnosis was therefore made in

15 of the 23 cases with biopsy material available.When this final diagnosis was compared with thetentative diagnosis at the time of major surgery-for example, colitis unclassified, probably/possiblyulcerative colitis or Crohn's disease-a changewas necessary in only two cases.

CLINICAL CORRELATIONSThe preoperative clinical diagnosis on the 30 in-determinate cases are shown in Table 2. In two, nodiagnoses were made, and in 12, an element ofdoubtwas expressed by the clinician. Initially, these werealso correlated with the tentative diagnoses in thepathology index. There was clinicopathologicaldisagreement in six of the 30 cases. At the end of thestudy there were 11 cases that carried a firm clinicalopinion and a final confident pathological diagnosis.Clinicopathological conflict was present in two.

Table 2 Clinical diagnoses in cases indexed colitisindeterminate2 Equivocal-no diagnosis

12 Crohn's disease (7 clinical doubt)

16 Ulcerative colitis (5 clinical doubt)

Discussion

I have applied the term 'colitis indeterminate' to asmall group of cases in which there had beendifficulty in distinguishing Crohn's disease fromulcerative colitis in the excised specimen. Theyform from 10 to 15% of the cases in many series ofnon-specific inflammatory disease (Hawk and Turn-bull, 1966; Glotzer et al., 1970; Kent et al., 1970).

Ashley B. Price

The results presented here suggest that most of themconform to a recognisable pathological and clinicalpattern. In twenty-seven out of the 30 cases thepatient had undergone urgent surgery and thepathological features were those associated withacute severe colitis, with or without accompanyingtoxic megacolon. The pathology of Crohn's colitisin the acute phase and that of ulcerative colitishave much in common (Roth et al., 1959; Morsonand Dawson, 1972). This was borne out here andwas the cause of the diagnostic dilemma. Certaingross attributes of discriminating value in thechronic state are common to both conditions in theincipient or established fulminant phase. Relativerectal sparing, for example, is an accepted featureof the cumulative evidence for a diagnosis ofCrohn's disease (Schachter and Kirsner, 1975). Inthis study it was seen in some indeterminate casessubsequently proved to be ulcerative colitis (Fig. 1).These unusual macroscopic appearances wereprobably accentuated by the healing effect of thetopical and systemic corticosteroids given in allthe cases.A second misleading macroscopic appearance was

of intermittent ulceration caused by the variableintensity of the disease in the acute episode (Fig. 2).Such a pattern suggests a diagnosis of Crohn'sdisease but was a feature of some cases proved byfollow-up biopsies to be ulcerative colitis. Carefulhistological examination of these atypical macro-scopic sites will avoid major errors in classification.Although no diagnostic pathology was seen minorchanges suggesting previous inflammation werepresent in those cases eventually proven to beulcerative colitis.

Certain accepted microscopical discriminatingattributes also require careful interpretation infulminant disease. A regular glandular pattern with awell-maintained goblet-cell population in diseasedareas (Fig. 7) is normally a criterion for the diag-nosis of Crohn's disease. Yet this was a markedfeature in up to 60% of the indeterminate casesin which there was no other evidence for a diagnosisof Crohn's disease. Furthermore, some were subse-quently proved to be ulcerative colitis. The suddenextensive but irregular stripping of the mucosathat occurs in the fulminant episode seems toleave intact less damaged mucosal islands. Thiswould account for the intermittent pattern ofulceration seen macroscopically (Fig. 2) and theminimal inflammation of the mucosa covering many'inflammatory pseudo-polyps' (Price, 1978). Theseare recognised as the hallmarks of an episode ofprevious severe disease.

Fissuring ulceration is another accepted diagnosticcriterion of Crohn's disease (Lennard-Jones, 1971)

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from

Overlap in the spectrum of non-specific inflammatory bowel disease

but in some of the cases presented here it was afeature of developing fulminant ulcerative colitis.However, the fissuring ulceration in the acuteepisodes of both Crohn's disease and ulcerativecolitis had certain qualitative differences from thatof the typical chronic case of Crohn's disease.In the latter they were usually single serpiginoustracts lined by inflammatory cells and penetratingthrough intact muscle (Fig. 3). The acute fissuresthatcharacterised the indeterminate cases seen in thisstudy occurred mostly in areas of severe mucosalulceration. They were multiple, short, V-shapedclefts sparsely lined by inflammatory cells. Theadjacent muscle was always diseased, showingmyocytolysis and intense capillary congestion(Figs. 4, 5), these two features being forerunners tothe onset of dilatation. It is important to distinguishthe two forms of fissuring. Only one is an attributeof Crohn's disease. Unfortunately, in the four casesthat had the characteristic fissures of Crohn'sdisease but in which all the other features wereatypical no biopsy material was available on whichto make a final diagnosis. Although it seems safeto use the characteristic fissure as an unequivocal

distinguishing feature of Crohn's disease fromulcerative colitis, in this series the proof was stillincomplete.The acute fissure seems to be the end result of

deep undermining ulceration, for a complete rangewas seen from one to the other (Figs. 3, 4, 6, 8).It is particularly difficult to differentiate acutefissuring from deep undermining ulceration and thefissure of chronic Crohn's disease when the acutevariety occurs singly rather than in multiple groups(Fig. 8).The absence of granulomas and well-formed

lymphoid aggregates can be explained by caseselection. Granulomas certainly occur in the acuteforms of Crohn's disease (Glass and Baker, 1976;Fisher et al., 1976; Anslem et al., 1973), but fromthe literature the incidence appears to be less thanthat in chronic disease-25% as opposed to 60%.

It is well documented that transmural inflam-mation, an attribute usually associated with Crohn'sdisease, may be seen in severe ulcerative colitis(Lumb et al., 1955; Roth et al., 1959). This wasconfirmed in this investigation. As a diagnosticsign it must be evaluated with reference to the

Fig. 8 An early stage in the formation of the acute fissure. A penetrating ulcer is seen encroaching onthe muscle in which there is some myocytolysis and numerous dilated vascular channels. The adjacentmucosa once again shows little evidence of the overall severity of disease. From a case indexed colitisindeterminate, probably/possibly ulcerative colitis. (H and E x 15)

S75

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from

576 Ashley B. Price

Resolving Increasina tivityphase of d

A Typical|/' Crohn's disease

Colitis indeterm'at3evre ocute colitis

TypicalI s--ulcerat~~~~~~~~lcraed colitis

Fig. 9 Diagrammatic representation of the spectrumofpathology in non-specific inflammatory bowel disease.In the small percentage of cases pursuing the courserepresented by the thin lines, separation of Crohn'sdisease from ulcerative colitis begins to become difficultat point B (incipient dilatation) and may be impossibleat point A (toxic megacolon). Biopsies at times repre-sented by either end of the chart often solve the diagnosticdilemma.

disease activity. The data in Table 1 suggest thatit is a good descriminating attribute in non-urgentcases but a non-specific feature in acute disease.

Focal or widespread myocytolysis plus intensemural and mucosal capillary engorgement wereconfined to the acute cases and were good indicatorsof incipient fulminant disease. The dissolution anddisintegration of the muscle plays a constant rolein the development of dilatation (Lumb et al., 1955;Roth et al., 1959) and is common to toxic megacolonof any cause. Its precise aetiology, however, isunknown. Drug therapy and damage to the neuralplexus have been cited as precipitant causes (Nor-land and Kirsner, 1959).From the review of the biopsy material in 23 of

the cases a final confident diagnosis was made in 15.This emphasises the importance of establishing asequential record of pathological changes in everycase of non-specific inflammatory bowel disease.The surgical specimen represents only one instantin time in these two chronic diseases-a time in thecase of a severe exacerbation when a confidentdiagnosis may not be available.

In those cases where a final histopathologicaldiagnosis was possible agreement with the tentativepathological diagnosis was noted in 85% and with aconfident clinical diagnosis in 80 %. Justificationfor the use of the temporary terminology 'colitisindeterminate', probably/possibly ulcerative colitisor Crohn's disease, is that it avoids permanenterrors in classification and is intellectually honest.Furthermore, it stimulates the clinician to seek forother confirmatory evidence yet acts as a therapeuticguideline.

Finally, the position of the cases labelled colitis

indeterminate in the spectrum of inflammatory boweldisease is illustrated (Fig. 9). They formed a smallbut distinctive group in which the diagnosticdifficulty was due to a common pattern. They wereclinically acute, severe cases that had undergoneurgent surgery. Macroscopically there was usually atotal colitis, sometimes with rectal sparing or anintermittent ulcerated pattern. Histologically thedominant pattern was of severe ulceration withnon-specific transmural inflammation. In all casesthere were foci of myocytolysis. Acute fissuringwas prominent. The islands of surviving mucosashowed only minimal inflammation, were intenselycongested, and had maintained a regular glandularpattern with preservation of mucin. Colonic dila-tation was a variable finding. Most of these featureswould suggest a diagnosis of Crohn's disease in thechronic phase. Outside the acute phase of the diseasereliable diagnostic attributes may again be found.

I thank Dr B. C. Morson, Mr H. E. Lockhart-Mummery, Dr J. Ritchie, Mr Norman Mackie,the Medical Illustration Departtnent at NorthwickPark Hospital, and Mrs S. McHale for their helpin preparing this manuscript.

References

Anslem, K., Rosenblum, S. A., and de Marco, V. (1973).Toxic megacolon in granulomatous colitis. HenryFord Hospital Journal, 21, 69-74.

Cook, M. G., and Dixon, M. F. (1973). An analysis of thereliability of detection and diagnostic value ofvarious pathological features in Crohn's disease andulcerative colitis. Gut, 14, 255-262.

Farmer, R. G., Hawk, W. A., and Turnbull, R. B. Jr.(1968). Regional enteritis of colon: a clinical andpathologic comparison with ulcerative colitis. AmericanJournal of Digestive Diseases, 13, 501-514.

Fisher, J., Mantz, F., and Calkins, W. G. (1976). Colonicperforation in Crohn's disease. Gastroenterology,71, 835-838.

Glass, R. E., and Baker, W. N. W. (1976). Role of thegranuloma in recurrent Crohn's disease. Gut, 17, 75-77.

Glotzer, D. J., Gardner, R. C., Goldman, H., Hinrichs,H. R., Rosen, H., and Zetzel, L. (1970). Comparativefeatures and course of ulcerative colitis and granulo-matous colitis. New England Journal of Medicine,282, 582-587.

Hawk, W. A., and Turnbull, R. B., Jr (1966). Primaryulcerative disease of the colon. Gastroenterology, 51,802-805.

Hywell-Jones, J., Lennard-Jones, J. E., Morson, B. C.,Chapman, M., Sackin, M. J., Sneath, P. H. A.,Spicer, C. C., and Card, W. I. (1973). Numericaltaxonomy and discriminant analysis applied tonon-specific colitis. Quarterly Journal of Medicine,42, 715-732.

Kent, T. H., Ammon, R. K., and Denbesten, L. (1970).

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from

Overlap in the spectrum of non-specific inflammatory bowel disease 577

Differentiation of ulcerative colitis and regionalenteritis of colon. Archives of Pathology, 89, 20-29.

Lennard-Jones, J. E. (1971). Definition and diagnosis.(Skandia International Symposia, 5) edited by A.Engel and T. Larsson., pp. 105-112.In RegionalEnteritis(Crohn's Disease), Nordiska Bokhandelns Forlag,Stockholm.

Lennard-Jones, J. E., Lockhart-Mummery, H. E., andMorson, B. C. (1968). Clinical and pathologicaldifferentiation of Crohn's disease and proctocolitis.Gastroenterology, 54, 1162-1170.

Lewin, K., and Swales, J. D. (1966) Granulomatouscolitis and atypical ulcerative colitis. Gastroenterology,50, 211-223.

Lockhart-Mummery, H. E., and Morson, B. C. (1960).Crohn's disease (regional enteritis) of the large in-testine and its distinction from ulcerative colitis.Gut, 1, 87-105.

Lockhart-Mummery, H. E., and Morson, B. C. (1964).Crohn's disease of the large intestine. Gut, 5, 493-509.

Lumb, G., Protheroe, R. H. B., and Ramsay, G. S.(1955). Ulcerative colitis with dilatation of the colon.British Journal ofSurgery, 43,182-188.

McGovern, V. J., and Goulston, S. J. M. (1968). Crohn'sdisease of the colon. Gut, 9, 164-176.

Margulis, A. R., Goldberg, H. I., Lawson, T. L., Mont-gomery, C. K., Rambo, N., Noonan, C. D., andAmberg, J. R. (1971).Theoverlappingspectrumofulcer-ative and granulomatous colitis: A roentgenographic-pathologic study. American Journal ofRoentgenologyRadium Therapy and Nuclear Medicine, 113, 325-334.

Marshak, R. H., and Lindner, A. E. (1975). Radiologicdiagnosis of chronic ulcerative colitis and Crohn'sdisease of the colon. In Inflammatory Bowel Disease,

Edited by J. B. Kirsner, and R. G. Shorter, pp. 241-276.Lea and Febiger, Philadelphia.

Morson, B. C., and Dawson, I. M. P., Eds. (1972).Gastro-Intestinal Pathology, p. 467. Blackwell, Oxford.

Norland, C. C., and Kirsner, J. B. (1969). Toxic dilatationof colon (toxic megacolon): etiology, treatment andprognosis in 42 patients. Medicine, 48, 229-250.

Price, A. B. (1978). Inflammatory and lymphoid polypsof the large intestine. In Pathogenesis of ColorectalCancer. (Major Problems in Pathology), edited byB. C. Morson., W. B. Saunders Company, Phila-delphia (In press).

Price, A. B., and Morson, B. C. (1975). Inflammatorybowel disease. The surgical pathology of Crohn'sdisease and ulcerative colitis. Human Pathology,6, 7-29.

Ritchie, J. K. (1974). Results of surgery for inflammatorybowel disease: a further survey of one hospital region.British Medical Journal, 1, 264-268.

Roth, J. L. A., Valdes-Dapena, A., Stein, G. N., andBockus, H. L. (1959). Toxic megacolon in ulcerativecolitis. Gastroenterology, 37, 239-255.

Schachter, H., Goldstein, M. J., Rappaport, H., FennessyM. B., and Kirsner, J. B. (1970). Ulcerative and'granulomatous' colitis-validity of differential diag-nostic criteria. Annals ofInternal Medicine, 72, 841-851.

Schachter, H., and Kirsner, J. B. (1975). Definitions ofinflammatory bowel disease of unknown etiology.Gastroenterology, 68, 591-600.

Requests for reprints to: Dr A B. Price, HistopathologyDepartment, Northwick Park Hospital, Watford Road,Harrow, Middlesex HAI 3UJ.

on June 23, 2022 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.31.6.567 on 1 June 1978. D

ownloaded from