neutropenic enterocolitis due to clostridium septicum infectionneutropenic enterocolitis due to...

J Clin Pathol 1984;37:335-343

Neutropenic enterocolitis due to Clostridium septicuminfectionASHLEY KING,* ANITA RAMPLING,t DGD WIGHT,* RE WARRENt

Departments of *Histopathology and tMedical Microbiology, Addenbrooke's Hospital, Cambridge

SUMMARY Three cases of neutropenic enterocolitis are described. This unusual condition occurs

almost exclusively in neutropenic patients and has a fulminating course which is almost invariablyfatal without surgical intervention. The lesion is centred on the caecum and hitherto itspathogenesis has been unclear, partly because most studies have been performed on materialobtained at necropsy. In these three patients, all of whom were treated surgically, Clostridiumsepticum was identified by specific immunofluorescence in the bowel wall of the resected speci-mens. Two patients also had C septicum septicaemia. Various forms of mucosal damage can beidentified which predispose towards invasion of the bowel wall by this organism. These cases

provide further confirmation of a primary role for C septicum in the pathogenesis of neutropenicenterocolitis.

Several types of gastrointestinal lesion have beendescribed in patients with leukaemia.' These includeleukaemic infiltration, mucosal haemorrhage, fungalinfection, and atypical epithelial changes directlyattributable to cytotoxic drugs.2 More recently, nec-rotising enterocolitis, associated with localised infec-tion or bacteraemia, or both, has been increasinglyrecognised as a particular complication ofleukaemia.3The name necrotising enterocolitis is confusing

and probably includes three separate and distinctentities: pseudomembranous colitis, ischaemic col-itis, and neutropenic enterocolitis.4 Pseudomem-branous colitis has a characteristic morphology andis linked with antibiotic treatment, and Clostridiumdifficile or its toxin or both is present in the stool.Acute ischaemic colitis is usually associated withulceration, extensive mucosal and submucosalhaemorrhage, and fibrin thrombi in vessels. Thethird group variously known as the ileocaecal syn-drome, typhlitis, or, most commonly, neutropenicenterocolitis57 is a distinct clinicopathological syn-drome of unknown aetiology in which patients oftenpresent with abdominal pain and fever resemblingappendicitis. The lesion, often diagnosed only atnecropsy, is usually centred on the caecum, whichshows pronounced thickening of its wall with intense

Accepted for publication 16 November 1983

oedema and necrosis and variable numbers ofmicroorganisms. Although the mortality is high,there have been recent reports of successful treat-ment by surgery-'0 and even antibiotics alone."The pathogenesis of neutropenic enterocolitis is

not entirely clear. The link with neutropenia seemsstrong since the condition is also encountered inneutropenia attributable to other causes-forexample, cyclic neutropenia'2-` and drug inducedagranulocytosis.'6

Although there are few detailed descriptions ofthe pathology, most authors have seen organismswithin the lesions. The organisms cultured fromblood or bowel have, however, been varied andoften mixed, and, as most descriptions refer topostmortem material, it is not always clear whetherthe organisms have played a primary role or aremerely secondary invaders. There have been threerecent reports of four cases of neutropenic col-itis'2-'4 in which C septicum septicaemia was associ-ated with characteristic ileocaecal disease. In threeof these cases Gram positive rods were seen in thebowel wall. Rifkin postulated that this organismmight be crucial to the development of neutropenicenterocolitis.'3 The many other organisms found byother authors may then be secondary invaders oreven represent postmortem growth.C septicum is not a common cause of non-

traumatic sepsis in man. Most of the recent reportsof C septicum infection, however, have been in

335

on May 7, 2021 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.37.3.335 on 1 March 1984. D

ownloaded from

http://jcp.bmj.com/

336

patients with leukaemia or colonic cancer.'7-'9Although none of these authors gives a detaileddescription of the pathological findings, mostauthors describe haemorrhage and necrosis of thecaecum or distal ileum or "caecitis."

In this report we describe three leukaemicpatients with classic neutropenic enterocolitis, eachof whom was subjected to bowel resection. In allthree patients material for histological examinationwas available early in the course of the patient'sillness and there was therefore little or no opportun-ity for secondary bacterial invasion. In all threecases C septicum bacilli alone were shown in thebowel wall by specific immunofluorescence.

Material and methods

C septicum fluorescein isothiocyanate labelled "0"antiserum was used, which is a mixture of three dif-ferent rabbit antisera raised against three strains:groups I and II described by Moussa20 and a thirdserologically distinct strain isolated in Tasmania(donated by Wellcome Research Laboratories).C difficile fluorescein isothiocyanate labelled "0"

antiserum was raised in rabbits against an 18 h cul-ture of a toxigenic strain of C difficile (CLN 34854/82), which was isolated from a patient at Adden-brooke's Hospital. Gammaglobulin containing theactive antibody was conjugated to fluoresceinisothiocyanate at Wellcome Research Laboratories.Work at Wellcome Research Laboratories (Femie,personal communication) and in our own laboratoryhas shown that all strains of C difficile so far testedhave a common "O" antigen which will stain withthis fluorescent antiserum.

Tissues were processed in paraffin wax and sec-tions were cut at 5 ,um. Dewaxed sections wererinsed in phosphate buffered saline and stained for30 min at 37°C in antiserum diluted 1/5 with phos-phate buffered saline, rinsed, and mounted in phos-phate buffered glycerol. They were examined with aLeitz Orthoplan microscope using a KP490 exciterand K5/5 barrier filter.The sinrplified alcohol shock method, which

selects for clostridial spores,2' was used for thedetection of C septicum from gut contents. A 1/10dilution of the specimen was added to an equal vol-ume of absolute ethanol and left at room tempera-ture for 1 h. Serial 10 fold dilutions were made inbrain-heart infusion broth, plated on blood agar,and incubated for 48 h at 37°C in anaerobic jars.Cultures were identified by sugar fermentation testsand gas-liquid chromatography.

King, Rampling, Wight, Warren

Case reports

CASE 1A 67 year old woman presented with acute myeloidleukaemia, which was treated with four courses ofDAT (daunorubicin, an anthracycline in combina-tion with cytosine arabinoside, and 6 thioguanine)over the next three months. During this period shehad several episodes of neutropenia, and at the endof the second course of DAT she became febrile andintravenous vancomycin and ceftazidime werestarted. Griping abdominal pain and diarrhoeadeveloped and C diffiile colitis was suspected (andsubsequently confirmed when both the toxin and theorganism were found in the stool). Oral vancomycinwas added to the regimen. Ceftazidime was with-drawn when a group D streptococcus was subse-quently grown from the blood. Diarrhoea andabdominal pain persisted for two weeks despite theoral vancomycin.Bone marrow remission of the leukaemia was

achieved at the end of the second month of treat-ment. Two months later she was readmitted to hos-pital, three weeks after a maintenance course ofcytotoxic treatment. She complained of anorexia forfive days and a one day history of abdominal pain,vomiting, fever, and constipation. She was febrile,pale, and sweaty with generalised guarding, rigidity,and rebound tenderness of the abdomen. Bowelsounds were diminished. The white blood cell countwas 2-2 x 109/l, with neutrophils less than 0-5 x19/Il.Her condition deteriorated over the next few

hours. A recurrence of C difficile infection compli-cated by colonic performation was suspected,although she had not received any antibiotics duringthe previous month. At laparotomy free bloodstained fluid was found in the peritoneal cavity,apparently emanating from an abnormal ileocaecalregion, which was excised. Perioperatively she wasgiven intravenous fluids, cefotaxime, gentamicin,and metronidazole, but she remained hypotensiveand anuric and died the next day. Blood taken onadmission to hospital subsequently yielded a pureculture of C septicum. C difficile was not found in thebowel contents.

PathologyThe resected bowel consisted of 20 cm of terminalileum together with the right colon. On both sides ofthe ileocaecal valve the bowel mucosa wasoedematous with bulging mucosal folds and focalhaemorrhage. Several small ulcers (about 0-5 cmdiameter) were also present in the terminal ileum.Sections showed that each of the ileal ulcers wasrelated to a mucosal leukaemic deposit associated



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Neutropenic enterocolitis due to Clostridium septicum infection

Fig. 1 Case 1. Leukaemic cells in the mucosa of theterminal ileum. Haematoxylin and eosin x 800.

with areas of coagulative necrosis of both mucosaand submucosa (Fig. 1). The muscularis propriashowed patchy necrosis, mainly of the inner coat. Nomature polymorphonuclear leukocytes were seen,but leukaemic cells were rather sparsely scatteredthrough the oedematous submucosa. Submucosalvessels were dilated and some contained thrombiadjacent to the areas of necrosis. Many Gram posi-

0

tive bacilli were seen especially in the areas of nec-rosis, and these were all shown to be C septicum byspecific immunofluorescence. No C difficile organ-isms were found. There were no leukaemic depositsin the caecum, but there were a few foci of mucosalnecrosis and submucosal oedema was much morepronounced than in the ileum.

CASE 2A 20 year old man presented with acute myeloidleukaemia and treatment with DAT was started.Two days after completing the second course he wasreadmitted to hospital with slight fever, abdominalpain, and diarrhoea. He was tender to palpation inthe right iliac fossa. The white blood cell count was0-4 x 109/l with no neutrophils, and the plateletcount was also reduced ( 19 x 109/1). C difficile infec-tion was suspected, although he had not receivedantibiotics during the preceding three weeks, and acourse of oral vancomycin was started.

During the next 24 h his condition deterioratedand guarding developed in the right iliac fossa.Intravenous gentamicin, cefotaxime, and met-ronidazole were started before diagnosticlaparotomy. At operation the caecum and ascendingcolon were found to be grossly thickened andhaemorrhagic, and there was a perforation in thelateral wall of the colon. A right hemicolectomy wasperformed.

5 (cm) 10

Fig. 2 Case 2. Resected caecum showing central ulceration plus gross thickening ofthe wall and mucosaloedema.

E:z:::m.

337



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

338

His condition improved, but a blood culture takenon admission yielded clostridium species on the firstpostoperative day and intravenous vancomycin wasadded to the antibiotic regimen. The clostridiumwas identified as C septicum and this organism wasalso cultured from the gut contents, but neither Cdifficile nor its toxin was detected. The patient madean uneventful recovery.

PathologyThe surgical specimen showed a thickened purplecaecum with an oedematous mucosa covered by agrey slough (Fig. 2). Sections showed strikinghaemorrhagic oedema of the submucosa with patchynecrosis of the mucosa and submucosa. There wasno gas formation. The muscularis propria was alsonecrotic in some places, and at one point involve-ment of the full thickness of the wall had led toperforation. The intact mucosa was strikinglyabnormal, the epithelium of both surface and glandshaving been totally replaced by atypical immaturecells, which were elongated and flattened on the sur-face with foci of surface epithelial loss. Many glandlumina contained necrotic desquamated cells. Therewas some evidence of regenerative activity at thebases of the crypts (Fig. 3). In the foci of mucosalnecrosis there were large numbers of organisms both


in the ulcers and in the adjacent submucosa. Mostorganisms were Gram positive bacilli, but somewere short and -others filamentous; Gram negativeforms were also present (Fig. 4). Despite this vari-able appearance all the organisms were positivelyidentified as C septicum by specific immunofluores-cence (Fig. 5). None stained with the C diffrile anti-serum.

CASE 3A 17 year old boy in his first relapse of acute lym-phoblastic leukaemia has already been describedbecause he had what was believed to be an atypicalC difficile colitis.22 Remission induction treatmentwith DAT was followed by the development ofneutropenia and fever. This was treated byintravenous cefotaxime and gentamicin for five days,followed by intravenous ceftazidine and vancomy-cin. On the fifth day of the second course he com-plained of colicky-central abdominal pain and overthe next 8 h tenderness and guarding developed inthe right iliac fossa. This was accompanied byabdominal distension, scanty bowel sounds, andslight icterus. Although he had vomited severaltimes, there was no diarrhoea.At laparotomy the caecum and ascending colon

were found to be thickened and oedematous, and so

id1 4jll Fig. 3 Case 2. Completereplacement ofthe normal cryptand surface epithelium by atypical

*lreN;|!&immature cells, attributable tocytotoxic drugs. Haematoxylin and

U eosin x 200.



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


'p . a right hemicolectomy was performed. C difficilecytotoxin and organisms were found subsequently incaecal contents. Blood cultures were negative, but

3 9 * the patient was already on intravenous vancomycinat the time of operation. Although his condition ini-tially improved and C difficile disappeared from his

, stools after treatment with oral vancomycin, it reap-* ' * , * / peared three and a half weeks later, when he had a

*/ mild attack of diarrhoea.

Fig. 4 Case 2. Gram stain shows the pleomorphicmorphology of the C septicum organisms. Filamentousforms are intermingled with short citron forms. Most of theorganisms are Gram positive but a few are negative. Gramx 800.

Fig. 5 Case 2. With fluorescein isothiocyanate labelledantisera to C septicum all the organisms fluoresce, includingthe filamentous forms. x 800.

PathologyThe resected colon showed a strikingly oedematousmucosa with bulging mucosal folds and focalhaemorrhage but no obvious ulceration. The mesen-

teric vessels were patent. Sections showed intensesubmucosal oedema and necrosis without gas forma-tion, which was most pronounced in the caecum butalso present in the ileal stump. There was patchynecrosis of the muscularis propria, particularly of itsinner layer, but the mucosa appeared intact andnormal. It was only after careful search and multiplesections that a single focus of mucosal necrosis (02cm) was discovered (Fig. 6). Multiple Gram positivebacilli could be seen streaming into the submucosafrom this point. All the organisms were shown byspecific immunofluorescence to be Csepticum. No Cdifficile was found.

Discussion

C septicum was long thought to be important inhuman pathology, but in recent years it has suffereda relative eclipse.23 It is responsible for 1020% of

cases of gas gangrene, which is now a rare diseaseboth in civilian life and under war conditions.24 Inveterinary published work C septicum has been cal-led the malignant oedema bacillus because of thecharacteristic findings in animal infections. Theaffected tissues are infiltrated with large quantitiesof gelatinous exudate with little or no gas.25 Theorganism is common in the intestinal tract of herbi-vores and is associated with braxy in sheep, a condi-tion with histological similarities to neutropenicenterocolitis. Before vaccines became availablebraxy caused high losses in Norway, Scotland, andIceland.26 The disease follows ingestion of frozengrass which fills the rumen and chills the anatomi-cally adjacent abomasum (fourth stomach). C sep-ticum invades the wall of the abomasum, producingenormous thickening of the wall due to inflammat-ory oedema, in which there are large numbers ofbacilli and white cells. Mucosal haemorrhages arealso seen.26The classic descriptions of neutropenic

enterocolitis and C septicum infection are similarand correspond closely to the findings in our three

-"--

P% 0 -

/o

339



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


*, i

*.4 4 1$+1.a%

f iA?

8'

N Fig. 6 Case 3. The mucosawas normal apart from this onetiny area ofmucosal necrosisfrom which Gram positivebacilli were seen streaming intothe submucosa. Intensesubmucosal oedema andnecrosis ofthe inner musclecoat are also apparent.Haematoxylin and eosin.Original magnification x 15.

.1

8'

8.;

*1.A'

4 W

I

f

.. I

/

-f

e

patients. In each of the cases extensive submucosaloedema with necrosis and haemorrhage without gasformation were the main findings. Myonecrosis inthe absence of vasculitis was also seen. Organismswere plentiful and always most numerous near tothe surface, where the mucosa was invariably necro-tic, suggesting that infection was from the lumen ofthe bowel. All of our patients had C septicum in thegut wall as shown by immunofluorescence. C sep-ticum was cultured from the blood of two of ourpatients, while the third was already receiving sys-temic antibiotics effective against clostridia, as aremany such neutropenic patients.

All the patients presented with abdominal painand fever and, as in other cases associated with C

septicum septicaemia, symptoms developed rapidlyover 24-48 h. Vomiting and change in bowel habitwere inconsistent findings.There is some disagreement as to how commonly

C septicum is found in the normal human intestine.Although it is often stated that it is a member of thenormal flora,23 recent studies suggest that it israre.27 28 Gut contents were cultured for C septicumonly in case 2, but there seems no doubt that thiswas the portal of entry in our patients. It is not clearwhy the ileocaecal region is so often the prime targetof the infection. Bacterial proliferation at this site isusual and is probably affected by local factors suchas pH and the redox potential, the osmotic environ-ment, and bile salt concentration.'8 By analogy with

340

i-ir"tl>!: W:.

*4

I

it:



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


Possible pathogenetic mechanisms in neutropenicenterocolitis

C septicum and mucosal damage due to:1 Leukaemic deposits (case 1)2 Cytotoxic drugs (case 2)3 Ulceration of tumours4 Neutropenic ulceration5 Haemorrhage due to thrombocytopenia6 ?Clostridium difficile (case 3)

C difficile infection, it is likely that Csepticum colon-isation is affected by antibiotics, and it is importantto note that this organism is much more resistant tothird generation cephalosporins and thienamycinthan earlier cephalosporins or penicillins.29

It seems probable that, as in braxy, predisposingphysiological and structural damage to the mucosalbarrier is necessary to initiate invasive C septicuminfection (Table). Also the relative rarity of C per-fringens bacteraemia in neutropenic patients sug-gests that they may have a specific susceptibility to Csepticum. Our patients showed three possiblemechanisms. In case 2 there was extensive epithelialatypia present in the colonic mucosa, which, as inthe cases described by Slavin,2 was associated withpatchy necrosis of epithelial cells leading to attenua-tion and interruption of the surface layer in places.These changes are attributable to cytosinearabinoside, which was received by all of ourpatients. Slavin et a12 do not indicate whether theepithelial lesions have any predisposition for theileocaecal region, but caecal necrosis has beenencountered with other cytotoxic regimens.303'

In case 1 the lesions were clearly spatially relatedto ulcerated mucosal leukaemic deposits, anotherobvious portal of entry. This same mechanism, sur-face ulceration of tumours, probably accounts forthe strong association between C septicum and can-cers of the caecum.''-'9 Another possible mechan-ism may be mucosal damage resulting from haemor-rhage due to thrombocytopenia.' I

None of these mechanisms seems able to explaininfection in patients with non-neoplasticagranulocytosis. Neutropenia, however, is associ-ated with ulceration of the oral mucous membrane,'2and similar lesions occur in the caecum.'2

Case 3 seems the most problematic. This case was

initially reported as an atypical response to Cdifficile in a patient without neutrophils.22 C difficileand its toxin were undoubtedly present in the stool,but we have conclusively shown by specificimmunofluorescence that the organisms present inthe lesions were not C difficile but were in fact Csepticum.C difficile is now considered to be the main

organism responsible for pseudomembranous

enterocolitis,"33 3 which is usually associated withalteration to the normal gut bacterial flora caused byantibiotics.34 Morphologically, in non-neutropenicpatients, the changes are distinctive.35 The colonicmucosa is covered by raised yellowish plaques orpseudomembranes, which have no special predilec-tion for the caecum. Microscopically, the plaquesare composed of foci of mucosal necrosis covered bya volcano-like eruption of neutrophils, fibrin, andepithelial debris. True ulceration or involvement ofthe submucosa are not normally seen. Bacteria arenot found in the lesions.

Since 1975 the hamster has been widely adoptedas an experimental model of pseudomembranousenterocolitis since it has been shown that a haemor-rhagic caecitis can be reproduced by C difficile or itstoxin.3637 Although this anatomical setting suggestsa close parallel with neutropenic enterocolitis, mor-phologically the lesions are dissimilar. Microscopi-cally the main changes are seen in the mucosa, whichappears hyperplastic with diffuse haemorrhage.3'-39In the most advanced cases there is extensiveepithelial necrosis with the formation ofpseudomembranes. Organisms have not been foundwithin the tissues. Furthermore, it has been conclu-sively shown that the lesions can be reproduced byintracaecal injection of the toxin alone.36 Here toothe principal histological change is epithelial nec-rosis.

Case 3 showed none of the usual features ofpseudomembranous enterocolitis, and we remainignorant of the usual morphological consequences ofinfection by C difficile in neutropenia. It remains tobe shown that C difficile has the power to causetissue damage in the bowel wall by invasion. It isprobable that this organism was responsible for thetrivial mucosal damage in case 3, allowing invasionby the more aggressive C septicum. The .attractivehypothesis that neutropenic enterocolitis might havebeen directly a manifestation of C difficile infectionalone2240 is thus no longer tenable in this particularcase, and a synergic process entailing two clostridialspecies seems more likely.

Other clostridia are well recognised as causes ofgut necrosis in animals, and C perfringens type Ccauses enteritis necroticans in man. The infection ishistologically different from the cases we describe.As well as tending to occur in the upper gastrointes-tinal tract, the findings are dominated by haemor-rhagic necrosis and submucosal gas bubbles ratherthan oedema, as well as systemic effects such as dis-seminated intravascular coagulation.

Previous reports of neutropenic enterocolitis haveemphasised that multiple organisms are oftenfound.34 It is interesting that many of theseorganisms-for example, pseudomonas, aeromonas,

341



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

342

and aspergillus-may be associated with vascularinvasion and might thus be responsible for mucosaldamage predisposing to clostridial invasion.There are other explanations for the mixed bac-

terial flora seen on Gram stained slides. Firstly, Csepticum is a highly pleomorphic organism whichmay range in size from long filaments to short citronand coccobacillary forms. Furthermore, as in oldcultures, many of the organisms in lesions appearGram negative, giving an impression that more thanone organism is present. Secondly, as Blenkinsoppand Dupont" indicate, invasion of extensively nec-rotic gut wall by bowel flora may mask a primarilyclostridial disease. This could also mask the respons-ible organism in postmortem bacteriology.7 Since allour cases were surgical excisions there had been lit-tle or no opportunity for overgrowth by other bac-terial species.Our experience with these three patients suggests

that it may be valuable to screen neutropenicpatients with abdominal symptoms for gut colonisa-tion by C sepiicum as well as C difficile. The alcoholshock method using non-selective nutrient bloodagar is an ideal and simple method. This may beparticularly valuable in patients receiving antibioticsthat prevent detection of bacteraemia as the sporeswill not be affected by the antibiotic. No model forclostridial infection in neutropenic animals has beendevised. In neutropenic patients many species ofclostridia may be found in the blood as componentsof polymicrobial bacteraemia. By contrast C sep-ticum is usually isolated in pure growth, suggesting aprimary pathogenic role. Other clostridia may alsohave this potential. The role of C difficile in neut-ropenic enterocolitis remains controversial. Wehave recovered two isolates as components ofpolymicrobial bacteraemia which could suggesteither a potential for tissue invasion or a polymicro-bial bacteraemia associated with colonic ulceration(Rampling et al. Unpublished observations). Ineither case it is likely that oral vancomycin andparenteral antibiotics effective against anaerobeswill be appropriate treatment, as they would also bein C septicum infection.

In conclusion, the lesions of neutropenicenterocolitis have many of the features of the classicdescriptions of infection by C septicum. Others havefound the organism in the blood-we have found itin the bowel wall. This observation, made on threecases occurring in one unit in nine months, providesfurther support for Rifkin's suggestion'3 that thisorganism may play an important primary role inmany more cases of neutropenic colitis than hithertoappreciated. The epidemiology of C septicum thusdeserves further study, and active immunisation ofpatients at risk may well be justified in the future.


We are most grateful to Dr David Fernie of Well-come Research Laboratories, who kindly arrangedfor the preparation of the fluorescent antibody to Cdifficile and who donated the C septicum fluorescentantibody for this investigation. We are also indebtedto Mr AJ Newton, who performed theimmunocytochemical methods, and to Mr Christ-opher Burton for the photographs.

References

Prolla JC, Kirsner JB. The gastrointestinal lesions and complica-tions of the leukaemias. Ann Intern Med 1964;61: 1084-103.

2 Slavin RE, Dias MA, Saral R. Cytosine arabinoside induced gas-trointestinal toxic alterations in sequential chemotherapeuticprotocols. A clinicopathologic study of 33 patients. Cancer1978;42: 1747-59.

Steinberg D, Gold J, Brodin A. Necrotizing enterocolitis inleukaemia. Arch Intern Med 1973;131:538-44.

4 Dosik GM, Luna M, Valdivieso M, et al. Necrotizing colitis inpatients with cancer. Am J Med 1979;67:646-56.

Wagner ML, Rosenberg HS, Fernbach DJ, Singleton EB. Typh-litis: A complication of leukaemia in childhood. AJR1970;109:341-50.

6 Sherman NJ, Woolley MM. The ileocaecal syndrome in acutechildhood leukaemia. Arch Surg 1973;107:39-42.

Moir DH, Bale PM. Necropsy findings in childhood leukaemia,emphasizing neutropenic enterocolitis and cerebralcalcification. Pathology 1976;8:247-58.

8 Kies MS, Luedke DW, Boyd JF, McCue MJ. Neutropenicenterocolitis. Two case reports of long-term survival followingsurgery. Cancer 1979;43:730-4.

Varki AP, Armitage JO, Feagler JR. Typhlitis in acuteleukaemia. Successful treatment by early surgical intervention.Cancer 1979;43: 695-7.

'0 Ikard RW. Neutropenic typhlitis in adults. Arch Surg198 1; 116:943-5.

Shaked A, Shinar E, Freund H. Neutropenic typhlitis: a plea forconservation. Dis Colon Rectum 1983;26:351-2.

12 Bignold LP, Harvey HPB. Necrotizing enterocolitis associatedwith invasion by Clostridium septicum complicating cyclicneutropenia. Aust NZ J Med 1979;9:426-9.

Rifkin GD. Neutropenic enterocolitis and Clostridium septicuminfection in patients with agranulocytosis. Arch Intern Med1980; 140:834-5.

4 Hopkins DG, Kushner JP. Clostridial species in the pathogenesisof necrotizing enterocolitis in patients with neutropenia. Am JHematol 1983;14:289-94.

Wright DG, Dale DC, Fauci AS;, Wolff SM. Human cyclic neut-ropenia. Clinical review and long-term follow-up of patients.Medicine (Balt) 1981;60:1-13.

Braye SG, Copplestone JA, Gartell PC. Neutropenicenterocolitis during mianserin-induced agranulocytosis. BrMed J 1982;285:1117.

'' Alpem RJ, Dowell VR. Clostridium septicum infections andmalignancy. JAMA 1969;209:385-8.

18 Koransky JR, Stargel MD, Dowell VR. Clostridium septicumbacteraemia. Its clinical significance.Am J Med 1979;66: 63-6.

'9 Katlic MR, Derkac WM, Coleman WS. Clostridium septicuminfection and malignancy. Ann Surg 1981; 193:361-4.

20 Moussa RS. Antigenic formulae for Clostridium septicum andClostridium chauvei. J Pathol Bacteriol 1959;77:341-50.

21 Borriello SP, Honour P. Simplified procedure for the routineisolation of Clostridium difficile from faeces. J Clin Pathol1981;34: 1124-7.

22 Rampling A, Warren RE, Berry PJ, Swirsky D, Hoggarth C,Bevan PC. Atypical Clostridium difficile colitis in neutropenic



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Neutropenic enterocolitis due to Clostridium septicum infectionpatients. Lancet 1982;ii: 162-3.

23 MacLennan JD. The histotoxic clostridial infections of man. Bac-teriol Rev 1962;26:236-76.

24 Finegold SM. Anaerobic bacteria in human disease. New York:Academic Press, 1977:388-9.

2 Topley WWC, Wilson GS. Gas gangrene; Anaerobic infectionsof animals. The principles of bacteriology and immunity. Lon-don: Edward Arnold and Co, 1936:1403-8.

26 Gaiger SH. Anaerobic infections of animals. J Comp Pathol1924;37: 163-91.

27 Drasar BS, Goddard P, Heaton S. Clostridia isolated fromfaeces. J Med Microbiol 1976;9:63-71.

28 Finegold SM, Atterbery HR, Sutter VL. Effect of diet on humanfecal flora: comparison of Japanese and American diets. Am JClin Nutr 1974;27:1456-69.

29 Gabay EL, Rolfe RD, Finegold SM. Susceptibility of Clostridiumsepticum to 23 antimicrobial agents. Antimicrob AgentsChemother 1981;20:852-3.

30 Matolo NM, Garfinkle SE, Wolfman EF. Intestinal necrosis andperforation in patients receiving immunosuppressive drugs.Am J Surg 1976; 132:753-4.

3' Kingry RL, Hobson RW, Muir RW. Caecal necrosis and perfora-tion with systemic chemotherapy. Am Surg 1973;39: 129-33.

32 Geelhoed GW, Kane MA, Dale DC, Wells SA. Colon ulceration-and perforation in cyclic neutropenia. J Pediatr Surg1973;8:379-82.

33 Tedesco FJ. Pseudomembranous colitis: Pathogenesis andtherapy. Med Clin North Am 1982;66:655-64.

I O'Connor TW. Pseudomembranous enterocolitis: A historicaland clinical view. Dis Colon Rectum 1981;24:445-8.

3S Bartlett JG. Antibiotic-associated pseudomembranous colitis.Rev Infect Dis 1979; 1: 530-9.

36 Libby JM, Jortner BS, Wilkins TD. Effects of the two toxins ofClostridium difficile in antibiotic-associated cecitis in hamsters.Infect Immun 1982;36:882-9.

37 Lusk RH, Fekety R, Silva J, Browne RA, Ringler DH, AbramsGD. Clindamycin-induced enterocolitis in hamsters. J InfectDis 1978;137:464-75.

38 Humphrey CD, Lushbaugh WB, Condon CW, Pittman JC, Pitt-man FE. Light and electron microscopic studies of antibioticassociated colitis in the hamster. Gut 1979;20:6-15.

31 Price AB, Larson HE, Crow J. Morphology of experimentalantibiotic-associated enterocolitis in the hamster: a model forhuman pseudomembranous colitis and antibiotic associateddiarrhoea. Gut 1979;20:467-75.

40 Conzigou P, Reiffers J, Leveque AM, Amouretti M, Broustet A,Berand C. Les enterocolites aigues necrosantes desagranulocytoses sout des formes majeures de colites pseudo-membraneuses. Gastroenterol Clin Biol 1982;6:48-51.

Blenkinsopp WK, Dupont PA. Bacteria in necrotisingenterocolitis. Lancet 1977;ii:617.

Requests for reprints to: Dr DGD Wight, Department ofHistopathology. Addenbrooke's Hospital, Cambridge CB22QQ. England.

343



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

neutropenic enterocolitis due to clostridium septicum infectionneutropenic enterocolitis due to...

Documents