GUit, 1983. 24, 106-113

Split ileostomy and ileocolostomy for Crohn's disease ofthe colon and ulcerative colitis: a 20 year survey

P H HARPER. S C TRUELOVE, E C G LEE,M G W KETTLEWELL. AND D P JEWELL

Froti Joh/ii Radcliffe Hospital, O.xforrd

SUMMARY The clinical course of 140 patients who have had a split ileostomy for ulcerative colitisor colonic Crohn's disease over a 20 year period is reported. In 37 patients with ulcerative colitisthere was no sustained improvement. In the 102 patients with Crohn's disease there was animmediate clinical improvement in 95, which was sustained in 65. Thirty patients havesubsequently required a proctocolectomy for persistent inflammation, and 28 are stilldefunctioned. Bowel continuity was restored after 61 split ileostomies and in 44 patients intestinalcontinuity remains intact at the present time (mean follow up since closure = 62.5 months, range0-231 months). It is concluded that a split ileostomy is a safe conservative operation producing atleast temporary improvement in severely ill and malnourished patients with Crohn's colitis, andthat if a subsequent resection becomes necessary it may be less extensive than was thoughtapplicable at the initial operation. In 27 patients a resection has not been required.

Isolation of the large bowel by means of an endileostomy was first tried by Brown' in the treatmentof ulcerative conditions of the colon, but the resultswere disappointing. Truelove2 showed that cortico-steroid enemas improved rectal inflammation inulcerative colitis and this led to the suggestion thatan external diversionary operation would not onlydefunction the colon, but would also allow adminis-tration of topical corticosteroids to the whole colonin patients with acute exacerbations. Preliminaryresults of this combined treatment showed that the26 patients with ulcerative colitis who had failed torespond to medical treatment had no improvementafter diversion.3 Five patients with Crohn's colitiswho were included in the study, however, showed amarked improvement in general health whichwarranted further assessment of this method oftreatment. This paper reports our long-termexperience with the technique, and although ouraim was to use it only for patients with Crohn'scolitis, in some patients the diagnosis was foundeventually to be ulcerative colitis.The original aims of the operation were: (1) To

facilitate limited resection. In patients with asegment of severe colonic disease in association with

Address for correspondence: P H Harper. Nuffield Department of Surgerv.John Radcliffe Hospital. Oxford.

Received for publication 17 Mai 1982

a mild diffuse colitis it was hoped that a period offaecal diversion might allow less inflamed bowel toheal so that strictures or fistulae could be excisedwithout a major resection or a proctocolectomybeing undertaken. (2) To facilitate major resection.In severely ill patients it was envisaged that animprovement in the poor general condition wouldfollow faecal diversion, and thus would makesubsequent major surgery safer. (3) To allowhealing of inflamed bowel. Patients in this categoryusually had severe symptoms in spite of medicaltreatment. In most cases the symptoms weresufficient to warrant a proctocolectomy, althoughthe disease, while diffuse, did not appear to be verysevere.For a number of reasons these aims have

subsequently become modified. The effective use oftotal parenteral nutrition has helped to prevent andtreat severe malnutrition, and has also allowed a'medical defunctioning' of the bowel for shortperiods, but where it failed to improve a patient'sgeneral condition a surgical defunctioning has beenperformed. It has also become apparent that,despite improvement in the disease, completehealing of the inflamed bowel rarely occurs.Furthermore, in some patients the technique hasbeen modified to combine a primary limitedresection of severe lesions with defunctioning of lessseverely inflamed bowel.

106

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

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.24.2.106 on 1 February 1983. D

ownloaded from

Crohn's disease of the colon and ulcerative colitis

OPERATIONThe faecal stream is diverted and the colondefunctioned by the operation of split ileostomy,operative details of which have been previouslydescribed.4 The operation has usually beenperformed under intravenous steroid cover. Theterminal ileum is divided about 5-10 cm from theileocaecal valve and a Brooke ileostomy5 isfashioned in the right iliac fossa. The distal ileum isbrought to the skin surface as a mucous fistula in theright hypochondrium. This operation, withoutresection of bowel, is referred to as a 'simple split' inthe following text.

Postoperatively, oral steroid treatment isdecreased gradually and if possible stopped aftertwo or three months. Daily topical steroid infusions(prednisolone 20 mg, Predsol Retention Enema,Glaxo) through the mucous fistula are continued forseveral months or a year. If the disease becomesquiescent or asymptomatic the colon is leftdefunctioned for about 18 months. The patient isthen reassessed to determine the extent and severityof residual disease. If investigation shows noevidence of active disease the continuity of thebowel is restored. Sometimes a segment of the colonmay show continued inflammation and this may

require a limited resection at the time of closure. Ifthe disease remains active and widespread a procto-colectomy is performed.

Methods

The current study has been undertaken to analysethe long-term results of a split ileostomy. Onehundred and forty patients have been treated. Thepreliminary report3 described 31 patients, but onlyeight are identifiable from their case histories inorder to be included in this study. In addition, afurther 109 cases have been reviewed making a totalof 117. A full follow-up was possible in 88 patients.Seventy-eight were seen in the outpatient clinic, sixpatients were reviewed by completion of a postalquestionnaire, and full details of four patients whohave moved from this region were forwarded bytheir present doctors. Nine patients have diedduring the 20 year period and 20 were reviewed untilthey were lost to follow-up.The diagnosis of each patient was reassessed using

the combination of clinical, radiological, and histo-pathological criteria described by Lockhart-Mummery and Morson.6 7 Over the period ofreview the diagnosis of several patients has beenaltered, especially as operative specimens becameavailable for histopathological study.A detailed record was made of the course of the

disease and treatment for each patient. The date of

onset of symptoms and subsequent major episodesof the illness such as relapses and operations wererecorded as 'events'. At each 'event', the clinicalfindings, laboratory investigations, histopatho-logical, and radiological reports were reviewed, andthe extent of the disease was assessed. Medical andsurgical treatment and their complications wererecorded in detail. The state of the patient's healthat the time of follow up was recorded.

Results

TYPES OF INFLAMMATORY BOWEI DISEASEPatients with ulcerative colitisThirty-seven patients with ulcerative colitis havebeen treated with a split ileostomy. Twenty-sixpatients were previously reported,3 but subse-quently the diagnosis of three was changed toCrohn's colitis. The identification of these 23patients is not possible from the earlier report. Since1965 a further 14 patients have been treated by asplit ileostomy but only four were known to haveulcerative colitis at the time of operation. In theremainder the final diagnosis of ulcerative colitis wasmade only after histological examination of resectedspecimens. The disease did not settle in 10 and theyhave required a proctocolectomy after a mean timeof 13.5 months' defunction (range 1-37 months).Two patients responded well to faecal diversion butrelapsed soon after closure of the split ileostomy andrequired proctocolectomies. The remaining twopatients had subtotal colectomies with ileorectalanastomoses. Both have recurrent disease needingmedical treatment. There were no deaths in thisgroup of 14 patients.One patient had a good clinical response to split

ileostomy for supposed Crohn's disease, but activecolonic disease persisted. She had a subtotalcolectomy and ileorectal anastomosis but hascontinued to have relapses of rectal disease needingmedical treatment. This is the only patient in thisseries in whom the histopathology of the operativespecimen is not conclusive of either ulcerative colitisor Crohn's disease and is thus classified asindeterminate colitis. This patient is, of course, notincluded in the Crohn's group.

Patients with Crohn's diseaseOne hundred and two patients have had splitileostomies for Crohn's colitis. Sixty-two patientswere women and 40 men with a mean age of onset ofdisease of 26.6 years (range 4-74 years). Onehundred and fourteen split ileostomies have beenperformed in these patients; 10 were defunctionedtwice and one a third time. The mean time fromonset of disease to the first split ileostomy was 57-3

1()7

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

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.24.2.106 on 1 February 1983. D

ownloaded from

Harper, Truelove, Lee, Kettlewell, and Jewell

months (range 0-309 months), and the meanduration of disease from its onset to the review was127 months (range 5-479 months). Fifty-eight casespresented primarily in Oxford and 44 patients werereferred from other doctors in the United Kingdomand abroad.At the time of the split ileostomy the majority of

patients had colonic or ileocolic disease, but five haddisease confined to the terminal ileum (Table 1).Eighty-five elective split ileostomies wereperformed; 69 for the failure of medical treatment,five without prior medical treatment, and 11 forperianal disease without major intestinalinflammation. The laparotomy findings duringemergency surgery are shown in Table 2. Faecaldiversion was used during 29 emergency operations,to defunction inflamed bowel, and also as a safe-guard when an anastomosis was at risk ofdehiscence. The combination of a split ileostomyand a bowel resection was used more frequently inemergency operations than during electiveprocedures (Table 3).

IMMEDIATE EFFECT OF SPLIT ILEOSTOMY INPATIENTS WITH CROHN'S DISEASEAfter faecal diversion, with or without an associatedresection, 95 patients (93%) showed an immediateimprovement of general health and symptoms. Theimprovement was sustained in 65 patients (64%) butwas partial and unsustained in 30 (Table 4), ofwhom 16 had a proctocolectomy after a mean periodof 16.4 months (range 4-54 months). Seven patientsdid not respond and all these were brought to aproctocolectomy within a year (mean 5.6 months).Of the 37 patients who had a poor or absent

response to split ileostomy the symptoms continuedin 23 patients and relapsed in 14 after diversion.Relapse was defined as the appearance of theclinical features of Crohn's disease after a symptom-free interval in patients with known disease,provided that other non-related causes of thesymptoms had been excluded.8 The cumulativerelapse rates (Fig. 1) of disease in defunctionedbowel and after closure of the ileostomy were

Table 1 Anatomical distribution of Crohn'.s disease at timeofsplit ileostomy

Numnber ofoperations

Small bowel 5llcocolic 22Colonic 54Colonic and small bowel 32No information ITotal 114

Table 2 Laparatomyfindings in 29patients treated byemergency operation

Inflammatory mass 14Acute colonic dilatation 4Obstruction 6Perforation 7Severely inflamed bowel 2Anastomotic dehiscence 3

calculated by the method described by Armitage.9

LONG-TERM RESULTS OF SPLIT ILEOSTOMY INPATIENTS WITH CROHN'S DISEASEThe sequence of operations performed is seen inFig. 2. Thirty patients (29%) have required aproctocolectomy; in 25 without closure of their splitileostomy. There was no significant improvement inhaemoglobin, white cell count, sedimentation rate,or serum albumin measurements, and little clinicalimprovement (Table 5) between the time of theinitial split ileostomy and proctocolectomy in thesepatients.

Twenty-eight patients have not been closed. Fivedied before closure of the split ileostomy, and onepatient has been lost to follow-up. Seven patientsare to have bowel continuity restored within thenext few months. Fifteen patients are not to havetheir ileostomies closed. In two of these analsphincter function is uncertain because of previouslysevere rectal and perianal disease, and five arereceiving medical treatment for active colonicdisease. The remaining eight patients prefer toremain defunctioned, rather than risk a relapse afterreconnection.

Sixty-one closures of split ileostomies have beenperformed after a mean period of external faecaldiversion of 20-1 months (range 1-61 months). Theclosure operation usually consisted of a simplereanastomosis, but in 13 patients a limited resectionwas necessary for residual disease. Of the 61patients, 44 remain closed at the present time (meanfollow up since closure, 62-5 months, range 0-231months), and 17 have been resplit or progressed to

Table 3 Split ileostomy and associated resections

Elective Etnergencyv Total

,Simple, 67 8 75with small bowcl resection 7 3 1()with right hemicolectomy 3 7 1()with subtotal colectomy 7 11 IXwith left hemicolectomy and

small bowel resection I - ITotal 85 29 114

108

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

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.24.2.106 on 1 February 1983. D

ownloaded from

Crohn's disease of the colon and ulcerative colitis

Table 4 Immediate response to split ileostomy and;eventual outcome in 102 patients with Crohn's colitis

Procto- Closurecol- of Defunc-

Total ectomy ileostomy tioned

No response 7 7 - -Partial response 30 16 6 8Good response 65 7 38 20

proctocolectomy (Fig. 2). Twenty-seven (44%) ofthe patients who were closed were simply split andreanastomosed without resection, and 13 have notneeded any further surgical treatment. Only twopatients have required a resection at bothoperations.During the period of colonic defunction there was

obviously a significant clinical improvement inpatients whose ileostomies were closed (Table 6). Inthese patients there was a significant reduction in theneed for steroid treatment by the time of ileostomyclosure in contrast with those patients who requireda proctocolectomy (Table 7). Although the colonicinflammation in the defunctioned colon mostlysettled, histological examination of endoscopicbiopsies shows evidence of residual inflammation.The colon becomes contracted and relativelyshortened during the period of defunction butdilates satisfactorily after reconnection.

EFFECTS OF SPLIT ILEOSTOMY FOR SPECIFICINDICATIONS IN PATIENTS WITH CROHN'S COLITISElective indicationsForty-four (52%) of the elective split ileostomieswere closed, 19 remain open and 22 came toproctocolectomy.

60

50 F

40 1

Fig. 1 The cumulative relapse rateduring colonic defunction (n= 14) andafter closure ofsplit ileostomy (n=61).

30

20

10

Proctocolectomy Defunctioned

234- "''

102 --24

Closure of44- *- ---_ - _- Isplit

/5524----- [jlit] . 3

11\

Closure of

split/6

1

Fig. 2 The sequence ofoperations.

Twenty-nine patients had severe perianal diseaseat the time of the operation. In 11 patients thediversion was indicated for perianal disease aloneand in the remainder for a combination of severeintestinal inflammation and perianal disease. Animprovement was recorded in the perianal inflam-mation of 19 patients, but only nine had theirileostomies closed and eight came to procto-colectomy. A full description of these patients willbe reported separately.

0- Relapse after closure of split ileostomy

01 Relapse in defunctioned bowel

6 1 2 1 8 24 30 36 42 48 54 60

TIME ( months)

109

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

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.24.2.106 on 1 February 1983. D

ownloaded from

Harper, Truelosve, Lee, Kettlewell, and Jewell

Table 5 Incidence of s!s'mptomns and signts at time of splitileostotmvi tantd suibsequenit pro(ctoc.olectoptmn wtit/louit priorclosure of ileostotms (n=25)

Split Pro( to-ileosto)ttil olect*oplx!

Constitutional 19 1(1Acute abdomen 3 2Diarrhoea/rectal discharge 16 12Pain 211 13Ulceraitise eolitis pictuire 4 2Periainall disea.se 7 12Extratintestinall disea.se 4

Six patients who presented with severe colonicCrohn's disease in childhood required large doses ofsteroids, and in one case azathioprine, to control thedisease. Steroid side effects and growth retardationwere prominent, and the patients showed no signs ofpuberty. There was a reluctance on the part of theparents and clinicians to perform a proctocolectomyin such young patients especially if the colitis wasthought to be because of Crohn's disease and so thedefunctioning technique was used. The operationswere performed between the ages of 12 and 16years. One patient died soon after his split ileostomyas the result of peritonitis from an extending fistula.Two later required a proctocolectomy and onepatient is still defunctioned. Two patients have hadtheir ileostomies closed and are fit and well. All thesurviving patients passed through puberty satis-factorily when the corticosteroid therapy wasdiscontinued or reduced.

Acute indicationsSeventeen (59%) of the split ileostomies performedacutely were closed, nine remain open, and threepatients have had a proctocolectomy. After theoperation there was an improvement in the generalhealth of 14 patients with acute inflammatorymasses. Diversion was combined with a resection ofthe mass in eight of these patients. In six the masswas not resected and it gradually resolved and

Table 6 Incidence ofsymptoms and signs at time ofsplitileostomv and subsequent closure (n=61)

Split Closure ofileostomv ileostoms'

Constitutional 38 2Acute abdomen 21 2Diarrhoea/rectal discharge 41 3Pain 38 1Ulcerative colitis picture 3 -

Perianal disease 17 1Extraintestinal disease 12 3

Table 7 Patients receivting steroid treatment at titmle of splitileostomn, closutre of'ileostomv. anid pro(toc(olec(tnov

Steroid treatmnent

Split ileostomv* 38 23and closure 9 52

Split ikcostomyt 19 6and proctocolcctomv 14 11

X2=27 13. P<'()1.()1:2X= 1.4.3. NS.

eventually disappeared, so that in one patient asimple reanastomosis was possible. Eight patientswere closed, two required a proctocolectomy, threeare still defunctioned, and one has died.

MORTALITY AND MORBIDITYThree patients (3%) died in hospital after a splitileostomy. One woman (aged 27 years) died of amassive pulmonary embolus two days post-operatively, and a boy (aged 13 years) developedobstruction and peritonitis after the operation.Another young man (aged 21 years) developedseveral intra-abdominal abscesses after a splitileostomy. He was seriously ill with multiple entero-cutaneous fistulae for over a year and wasmaintained on total parenteral nutrition until hisdeath. Six other patients died during the period ofreview. Two of these were unrelated to the disease(extradural haematoma and chronic respiratorydisease). Two patients died after proctocolectomy,one from overwhelming sepsis and a pulmonaryembolus, and the other from multiple pulmonaryemboli. A further two patients died after dischargefrom hospital. A woman (aged 40 years) had acardiac arrest while being anaesthetised elsewherefor the incision and drainage of an ischiorectalabscess and a man (aged 46 years) was admitted tohospital with shock and pneumonia and died almostimmediately.The morbidity recorded in the hospital notes of

the patients reviewed is shown in Table 8. Splitileostomy was uncomplicated in nearly two-thirds ofthe cases. The ileostomy stomas caused considerableproblems and were refashioned on 16 occasions.Pulmonary emboli were diagnosed in four patientsand were implicated in three of the deaths. Twelvepatients became obstructed after the split ileostomy;eight of these resolved on conservative manage-ment, and four needed operative correction. One ofthese patients was obstructed from a small bowelstricture of Crohn's disease that should, inretrospect, have been resected at the initial

11(

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

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.24.2.106 on 1 February 1983. D

ownloaded from

Crohn's disease of the colon and ulcerative colitis

Table 8 Morbidity ofsplit ileostomv for Crohn's disease

C'omplications No.

Total operations 114Uncomplicated 68Spout problems 16Peritonitis 4Wound and abdominal infections 12Wound dehiscence 3Fistulac 5Obstruction 12Pulmonary embolus 4Other 12

operation. The other patients were obstructed byadhesions and intra-abdominal herniation of bowel.

Discussion

Surgery has an undisputed place in the managementof inflammatory bowel disease. In Crohn's diseasethere is a continuing debate about the particularoperation which is best suited to a specific indicationand site of disease, because the disease is at presentincurable, tends to relapse, and often requiresrepeated operations. There is, therefore, a need topreserve intestine if possible. In ulcerative colitis thesituation is more clearly defined, mainly because thedisease is curable at the expense of a procto-colectomy.

In ulcerative colitis our results confirm the earlierstudies.3 it) Faecal diversion does not cause

sustained clinical improvement in the disease or ingeneral health and most patients soon progress tomajor excisional surgery. Since 1965 split ileostomyhas been used intentionally in only four young

patients with known ulcerative colitis in an attemptto spare them a proctocolectomy. At the time ofdiversion the other 10 patients had diagnoses ofCrohn's colitis or of an indeterminate colitis. Thesepatients were diverted mostly over 10 years ago andreflect the occasional difficulty in making a definitepreoperative diagnosis in some cases. Today,however, with colonoscopy and biopsy, doublecontrast radiology, and greater clinical experiencethis uncertainty is greatly reduced.

In contrast the effectiveness of a split ileostomy inCrohn's colitis is striking. The diversion usuallyresults in an immediate improvement in generalhealth. It is this respite in the course of the diseasethat makes the technique so valuable, so that time isobtained to improve the patient's condition andallow any resection which is necessary. In a majorityof patients the improvement followed simple splitileostomy, although in some it occurred afterdiversion combined with a resection. The

improvement was confirmed in several smallseries,11-4 after the publication of the preliminarystudy. Subsequently, enthusiasm for the procedurediminished, because of the incidence of extensionand relapses of disease in the isolated colons, thepoor rate of restoration of bowel continuity, and thefrequent progression of perianal disease.

Recently there has been renewed interest inconservative surgical techniques for the treatment ofCrohn's disease. Zelas and Jagelmanr'5 havereported 79 patients with Crohn's colitis in whom aloop ileostomy was used to divert the faecal stream.They emphasised the immediate generalimprovement in these patients and recommendedthe procedure as the initial management ofdebilitated patients before definitive surgery. Theloop ileostomy is quick and easy to construct andreplace but it does not ensure complete colonicdefunction. We have seen several complications ofthis operation, such as ileostomy prolapse anddifficulty in fitting the bag, and prefer to continuewith our own technique of complete faecaldiversion.The favourable response to external faecal

diversion is not universal. In seven of our patientsthere was no response at all and they soon requiredproctocolectomy (mean 5.6 months). In 30 others itwas partial and not sustained, and over half of theseeventually required a roctocolectomy (mean 16.4months). Oberhelman6 reported an improvementin all his 31 cases, Zelas and Jagelman'5 in 72 of 79,and Burman et al13 in 28 of 29. Other groups did notdo as well; Howell Jones et all obtained only atemporary remission in 19 of 21 patients, and onlytwo patients had a sustained remission.The role of faecal diversion in emergency surgery

is limited. Bowel frequently has to be resected(Table 3) - for example, for perforation or toxicdilatation. The defunctioning of inflammatorymasses, however, has resulted in the reduction ofinflammation and facilitated subsequent surgery. Inacute operations, therefore, faecal diversion can beused to conserve mild or moderately inflamed bowelnot needing resection, to isolate inflammatorymasses, and also as a safeguard when ananastomosis is at risk of dehiscence.There were five patients in the series with

inflammation confined to the small bowel, and theywere diverted because it was complicated by thepresence of an inflammatory mass or perianaldisease. Uncomplicated small bowel disease shouldbe treated by resection and anastomosis.There is good corroborative evidence to confirm

the improvements in health that we have seen.Generally, symptoms disappear promptly, weightincreases,'3 1' 1 and abdominal masses resolve.'5

111

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

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.24.2.106 on 1 February 1983. D

ownloaded from

Harper, Truelove, Lee, Kettlewell, and Jewell

Mean haemoglobin, sedimentation rate, serumalbumin, and seromucoid levels improve.1'3 Theneed for corticosteroid treatment is much reducedbut the radiological improvements lag behind theclinical. 'lWe have restored bowel continuity on 61

occasions. At the present time 44 patients havebowel continuity intact, with the prospect of sevenmore in the near future. Kivel et al"' test eachpatient before closure with a faecal challengeintroduced into the mucous fistula, but we have onlyrecently started systematically assessing the value ofthis. Zelas and Jagelman'5 who have used a loopileostomy specifically as a holding procedure toimprove their patients before resection, haverestored bowel continuity in combination with aresection in 52 of 79 patients. In three cases theyhave closed the ileostomy without resection but allthree have relapsed. We have closed 27 without anyresection and 13 have needed no further surgery fortheir Crohn's disease. Most other series have beenable to restore continuity in less than a third ofpatients13 14 1618 and many of these patients havethen relapsed.One of the major criticisms of external faecal

diversion is that the disease in the defunctionedbowel may progress and extend. If disease hascontinued after diversion then further excisionalsurgery is indicated. If relapses occur after an initial,good response to split ileostomy, medicalmanagement is often sufficient to control symptoms.Relapses of disease in defunctioned bowel areusually not an indication for urgent surgery. It isinteresting that relapses occur more frequently inpatients once bowel continuity has been restoredand this suggests a possible role of the faecal streamin the perpetuation of the inflammation.

Eleven patients had repeated split ileostomies.Their initial diversions were closed after animprovement in the colonic inflammation and agood clinical response. When recurrence occurredthe patients often preferred a repeat split ileostomyespecially if the only alternative was a procto-colectomy.

Despite the overall improvement in perianaldisease in response to split ileostomy, the ileostomyclosure rate has been poor. Split ileostomy was usedin 11 patients for perianal disease alone and itsadvantages over a sigmoid colostomy are that it iseasy to manage, permits administration of topicalcorticosteroids through the mucous fistula andensures complete faecal diversion. Although therewas a relief from sepsis and discomfort in 22 of 23patients with perianal disease in Zelas andJagelman's's patients, other authors have found thatthe disease progresses. Burman et al13 report the

healing of fissures but only one of six fistulae in anohealed and four other patients developed themwhile defunctioned. The response of perianaldisease to faecal diversion is variable, but in patientsincapacitated by their disease it is worth a trial, ifonly to facilitate a later proctocolectomy.There are reports of tumours developing in

bypassed Crohn's disease of the small bowel.'9 Twotumours have been reported occurring in defunc-tioned rectums,20' and two other at the site ofileocolovesical fistulae'8 also in defunctioned bowel.We have not seen any tumours in our patients, butthey will be periodically re-examined to exclude anoccult tumour.

Conclusions

While the aetiology of Crohn's disease in unknown,treatment can only be empirical. The advantage of asplit ileostomy is that it is a conservative operationwhich reduces the extent of resected bowel to aminimum. A split ileostomy is of little benefit inulcerative colitis. On the other hand, in Crohn'scolitis it is a relatively simple procedure with anacceptably low mortality and morbidity, which canbe performed in a sick patient in preference to anextensive resection, with an immediateimprovement in the general condition of mostpatients. It may improve the safety of subsequentoperations, limit the extent of subsequentresectional surgery, and in some cases remove theneed for resection entirely.Using the defunctioning technique for Crohn's

colitis has resulted in sustained remission ofsymptoms for considerable lengths of time. In asubstantial number of patients who would haveotherwise been treated by proctocolectomy forsymptoms which could not be controlled medically,it has been possible to restore intestinal continuity.Although defunctioning does not cure the disease,many of these patients have subsequently remainedwell enough not to need further surgery.

After our extensive experience with faecaldiversion we feel that it is indicated in patients witha definite diagnosis of Crohn's colitis in thefollowing circumstances: electively: for the failure ofmedical treatment, especially in debilitated patients;for symptomatic perianal disease not responding tosimple local measures; in the young to preservebowel and to attempt to prevent or delay aproctocolectomy. Acutely: to defunction inflamedcolon not needing resection; to defunctioninflammatory masses; to safeguard an anastomosisat risk of dehiscence.

Finally, the observation that the diversion offaeces in Crohn's colitis often results in a marked

112

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

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.24.2.106 on 1 February 1983. D

ownloaded from

Crohn's disease of the colon and ulcerative colitis 113

symptomatic improvement, and that local lesionsmay improve substantially, poses the question ofwhat part the faecal stream plays in the pathogenesisof Crohn's lesions.

Several other surgeons have been involved irn thetreatment of these patients - namely, ProfessorHarold Ellis, Mr Charles Webster, Mr Milo Keynes,and Mr Brian Dowling, and it is a pleasure toacknowledge our indebtedness to them.PHH is supported by the Wellcome Trust with a

Surgical Fellowship.

References

1 Brown JY. The value of complete physiological rest ofthe large bowel in the treatment of certain ulcerativeand obstructive lesions of this organ. Surg Gynec Obstet1913; 16: 610-3.

2 Truelove SC. Treatment of ulcerative colitis with localhydrocortisone hemisuccinate sodium: a report on acontrolled therapeutic trial. Br Med J 1958; 2: 1072-7.

3 Truelove SC, Ellis H, Webster CU. Place of adouble-barrelled ileostomy in ulcerative colitis andCrohn's disease of the colon: a preliminary report. BrMedJ 1965; 1: 150-3.

4 Lee E. Split ileostomy in the treatment of Crohn'sdisease of the colon. Ann R Coll Surg Engl 1975; 56:94-102.

5 Brooke BN. The management of an ileostomyincluding its complications. Lancet 1952; 2: 102-4.

6 Lockhart-Mummery HE, Morson BC. Crohn's disease(regional enteritis) of the large intestine and itsdistinction from ulcerative colitis. Gut 1960; 1: 87-105.

7 Lockhart-Mummery HE, Morson BC. Crohn's diseaseof the large intestine. Gut 1964; 5: 493-509.

8 Lee ECG, Papaioannou N. Recurrences following

surgery for Crohn's disease. Clin Gastroenterol 1980; 9:419-38.

9 Armitage P. Statistical methods in medical research.Oxford: Blackwell Scientific Publications. 1971.

10 Kivel RM, Taylor KB. Oberhelman H. Response tobypass ileostomy in ulcerative colitis and Crohn'sdisease of the colon. Lancet 1967; 2: 632-6.

11 Jones JH, Lennard-Jones JE, Lockhart-Mummery HE.Experience in the treatment of Crohn's disease of thelarge intestine. Gut 1966; 7: 448-52.

12 Oberhelman HA, Kohatsu S. Taylor KB. Kivel RM.Diverting ileostomy in the surgical management ofCrohn's disease of the colon. Am J Surg 1968; 115:231-40.

13 Burman JH, Thompson H, Cooke WT, AlexanderWilliams J. The effects of diversion of intestinalcontents on the progress of Crohn's disease of the largebowel. Gut 1971; 12: 11-15.

14 Mcllrath DC. Diverting ileostomy or colostomy in themanagement of Crohn's disease of the colon. Arch Surg1971; 103: 308-10.

15 Zelas P, Jagelman DG. Loop ileostomy in the manage-ment of Crohn's colitis in the debilitated patient. AnnSurg 1980; 191: 164-8.

16 Oberhelman HA. The effect of intestinal diversion byileostomy on Crohn's disease of the colon. In:Weterman IT, Pena AS, Booth CC, eds. Themanagement of Crohn's disease. Amsterdam-Oxford:Excerpta Medica, 1976: 216-9.

17 Weterman IT, Pena AS. The place of split ileostomy inthe treatment of Crohn's disease. In: Weterman IT,Pena AS, Booth CC, eds. The management of Crohn'sdisease. Amsterdam-Oxford: Excerpta Medica, 1976:224-32.

18 Aufses AH, Kreel I. Ileostomy for granulomatousileocolitis. Ann Surg 1971; 173: 91-6.

19 Lightdale CJ, Sternberg SS, Posner G, Sherlock P.Carcinoma complicating Crohn's disease. Am J Med1975; 59: 262-8.

20 Block GE. Surgical management of Crohn's colitis. NEngl J Med 1980; 302: 1068-70.

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

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.24.2.106 on 1 February 1983. D

ownloaded from