surgery amoebic colitis - gut · collar-stud abscesses. it was surprising howlittle disability the...

Gut, 1970,11, 941-946

Surgery in amoebic colitis

D. STEIN AND SIMMY BANKFrom the Departments of Surgery and Medicine and Gastro-intestinal Clinic, Groote Schuur Hospitaland University of Cape Town, Observatory, Cape Town, South Africa

SUMMARY The feasibility of surgery in amoebic colitis is presented by illustrative casehistories in two patients. Subtotal colectomy and ileostomy were carried out for fulminatingcolitis in the one and exteriorization, excision, and subsequent re-anastomosis of a localizedperforated area of bowel in the other.

It is suggested that the place of surgery in amoebic colitis should be reappraised, particularlywith the advent of modern medical and surgical methods now available. The indications forcontemplating surgery in abdominal amoebiasis are outlined and a workable approachis presented.

The majority of patients with severe amoebiccolitis respond rapidly and effectively to con-servative medical measures. Surgical interventionis usually reserved for free colonic perforation orthe drainage of intraperitoneal abscesses (Cope,1920; Evans, 1925; Ochsner and Debakey, 1942;Wilmot, 1950; Barker, 1958; Wilmot, 1962;Pelaez,Villazon, and Zaraboso, 1966). Apart fromthese two indications, surgery has usually beenregarded as meddlesome and, in fact, contra-indicated in amoebic colitis. The teaching thatthe colon resembles 'wet blotting paper' and istherefore hazardous to handle has resulted in adefeatist attitude towards operative procedures.Although isolated reports of colectomy anddrainage of abscesses have been published, mosttextbooks disregard surgery completely (Theron,1947; Barker, 1958).The purpose of this article is to demonstrate

that surgery is feasible in amoebic colitis, todiscuss the surgical indications in colonic amoe-

biasis, and to provide reasons for the desirabilityof surgical intervention in some cases of ful-minating amoebic colitis.

Received for publication 25 May 1970.

Case Reports

CASE 1The patient, an African male of 36 years, wasapparently quite well until one month before hisadmission to hospital on 4 July 1968 when hedeveloped pain in the right upper quadrantradiating to the right loin and back. Four daysbefore admission, he developed diarrhoea withblood and mucus in the stools associated withgeneralized cramp-like abdominal pains. He wasgiven a 'potent enema' by two of his friends tohelp clear his bowel, after which he becameincreasingly ill; the enema was later establishedto consist of soap and water only. There was nopast history of bowel upset and he had not beenin contact with anyone with diarrhoea.Examination on admission revealed him to be

an ill patient with mild dehydration. The pulse was120, temperature 99°F, and blood pressure120/90. There was diminished air entry and dull-ness at the base of the right lung and diffuseronchi scattered throughout both lung fields. Theabdomen was slightly distended but no masseswere palpable. 'Punch' tenderness was present

on March 19, 2020 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.11.11.941 on 1 N

ovember 1970. D

ownloaded from

http://gut.bmj.com/

942 D. Stein and Simmy Bank

over the liver and bowel sounds were present.Rectal examination was difficult due to markedtenderness. The other systems were normal. Thehaemoglobin was 9.5 g %, ESR 11 mm/hour, andwhite cell count 3,500 mm3. Examination of thestools showed blood and mucus but no amoebaewere visible on microscopy. Despite the negativestool examination, the diagnosis of amoebiccolitis was favoured but typhoid, bacillarydysentery, or subphrenic abscess were alsoconsidered. He was treated with penicillin andstreptomycin and intravenous fluids while beingfurther investigated. On 6 July 1968 he suddenlydeveloped generalized abdominal pain and hiscondition deteriorated rapidly. Examination atthat time revealed gross abdominal distension andpossible free fluid in the abdomen. Generalizedtenderness was present with guarding but therewas no true rigidity. A straight radiographshowed a dilated transverse colon and fluid levelsin the small bowel (Fig. 1). Proctoscopy at thatstage revealed a complete slough of the rectalmucosa. There was no normal mucosa to be seenand the slough extended to the mucocutaneousjunction. Examination of the slough again failedto reveal any amoebae on microscopy. Theproctoscopic findings appeared to substantiateamoebic colitis which had been aggravated by anenema given just before his admission. It wassuspected that the slough involved the wholecolon and that he had impending peritonitis.Laparotomy was performed the same evening

through a left paramedian incision. A largeamount of turbid free peritoneal fluid was found.The whole colon was wrapped up in the omentum.

Fig. 1 Erect radiograph of the abdomen incase 1 showing a dilated large and small bowel infulminating amoebic colitis.

The descending colon was stuck to the posteriorabdominal wall. Several necrotic patches werepresent in the rectum and sigmoid colon butthere was no free perforation. The colon wasopened at the hepatic flexure and the mucosa inthe area had sloughed and this lesionhad extendedmore proximally, but to a lesser extent. Theupper rectum, sigmoid, descending and transversecolon were dissected free and excised. Therectal stump was loosely closed with catgutsutures and replaced in the peritoneal cavity.Because of bleeding from the inflamed, rawretroperitoneal area the colon was resected asfar as the hepatic flexure only. The ascendingcolon was brought out through a stab woundin the right upper quadrant as a colostomy. Themucosa was necrotic at this area but the serosawas still intact. Several drains were inserted intothe peritoneal cavity and the wound closed. Onopening the resected transverse and the des-cending colon the whole mucosa was seen to besloughed, no viable islands of mucosal tissuebeing present. The histology report showed 'aspecimen consisting of large bowel measuringapproximately 44 cm long. The peritoneal surfaceis thickened and partly covered by fibrinousexudate. The proximal portion of the bowel isdilated, measuring 11 cm in diameter, whilst thedistal portion is 7 cm in diameter. The mucosa isdiffusely covered by a necrotic membranous typeof lesion but very occasional areas of normaltissue irtervene suggesting that this is severeamoebic dysentery. The histology confirms asevere florid amoebic dysentery with vast numbersof amoebae in the tissue.'

Emetine hydrochloride, 1 g (65 mg) daily,chloroquine diphosphate, 250 mg intramuscu-larly, and intravenous terramycin was started onreceiving the histological report. Despite thistherapy, the patient had a stormy postoperativecourse and on 11 July 1968 it was noted that themucosa at the colostomy site (where one smallisland of viable mucosa had previously beenpresent) had now completely sloughed and it wasdecided to remove the remaining colon. Theprevious laparotomy wound was opened and thisrevealed necrotic patches ofcaecum and ascendingcolon as well as four gangrenous patches of theterminal ileum in its distal 18 inches. Thecolonandinvolved ileum were excised and an ileostomy wasfashioned. After the operation the patientrequired intensive supportive measures, includingcorrection of the acidosis, isoprenaline infusions,central venous pressure monitoring, and electro-lyte replacement therapy. He discharged largequantities of pus from the rectal stump anddrainage wounds.The histology report of the second operation

showed 'a specimen consisting of proximal partof the ascending colon, caecum, appendix, anddistal ileum. The colon measured 20 cm and theileum measured 30 cm. The surface has a haemor-rhagic appearance and in areas appears to be



j.com/


ovember 1970. D

ownloaded from

http://gut.bmj.com/

943 Surgery in amoebic colitis

covered by fibrinous exudate. The appendixappears necrotic. The wall in areas appearsthickened while in others it is haemorrhagic andviable. In the lumen there is a slough whichinvolves the ileum proximally but at the distalline of excision it is absent. The whole of thecaecum contains the slough. Cut section of theappendix shows necrotic material within itscavity. A severe amoebic infection of the colon,appendix, and distal ileum is present. The des-truction of the bowel wall in areas has reachedthe serosa while in other areas has advanced asfar as the muscularis. In no section examinedhistologically is the mucosa seen. The amoebaeare extremely plentiful in all the sections. Inaddition, there is a local peritonitis.'The ileostomy functioned satisfactorily from

the outset despite the formation of a small boweland rectal stump fistula. These fistulae closedspontaneously after several days of antibiotictherapy. The patient improved slowly andon 15 August 1968 a large liver abscess wasdrained under a local anaesthetic through theeleventh intercostal space in the posterioraxillary line. An underwater drain was inserted.Microscopy of the pus obtained revealed thepresence of active amoebae despite a full courseof emetine. Radio-opaque dye was inserted viathe tube into the abscess cavity confirming thepresence of the abscess within the liver substance.A repeat course of emetine hydrochloride, 1 graindaily, was commenced but the patient continued todischarge a large amount of pus from the liverdrain. On 25 August 1968, metronidazole, 800 mgfour times daily, was commenced and from thenon the patient made rapid strides towards

Fig. 2 Barium enema one month after subtotalcolectomy in case 1 showing the rectum to be a

fibrotic tube with ulceration. The fistulous tract onthe left of the photograph resolved but the rectumremained unchanged on subsequent examinations.

recovery. The pus from the abscess cavitybecame less in quantity and the drain was ulti-mately removed. The rectal stump had dischargedlarge quantities of pus after operation and wasirrigated throughout his stay in hospital with asolution of kanamycin. Six weeks after theoperation he no longer had any dischargealthough sigmoidoscopic examination of therectal stump revealed a tube of granulationtissue with no evidence of any residual normalmucosa. Biopsy of the rectal stump at this stageshowed granulation tissue only. There was noevidence of any active amoebiasis. Barium enemawas performed (Fig. 2) and this revealed a dis-torted tube of granulation tissue with severalcollar-stud abscesses. It was surprising how littledisability the patient had from his grosslyabnormal rectal remnant.The patient was discharged from hospital and

was fitted with an ileostomy bag on 28 October1968. He was subsequently seen in the OutpatientDepartment one year after his discharge fromhospital and was perfectly fit and well. He had noproblem from the rectal stump and he was copingperfectly satisfactorily with the ileostomy. Thefistulae on the abdominal wall had all closed.

CommentThere are several points of importance in thiscase. (1) Despite the florid amoebic colitis,examination of the stools and rectal slough failedto demonstrate any active amoebae. (2) While it ispossible that early anti-amoebic therapy at thetime of admission may have averted the necessityof surgery, the finding ofan impending perforationat the rectosigmoid junction as well as severalother patches of necrosis of the rectum andsigmoid colon makes this unlikely. Colectomywould appear to have been the only procedurethat could have saved this man's life. (3) Despiteanti-amoebic therapy, the ascending colon andterminal ileum had already sloughed and thisrequired removal with a secondary procedure.Examination of the pathological specimen sug-gested that recovery would have been impossiblehad the colon remained in situ. Microscopicallythere was hardly any evidence of viable mucosaltissue and gangrenous patches extended throughto the serosa in several areas. (4) Marked involve-ment of the terminal ileum was present, a findingwhich is not uncommon in the more severe casesat necropsy. (5) The drainage of a liver abscesssix weeks after treatment with a full course ofemetine hydrochloride and chloroquine diphos-phate revealed the presence of active amoebicorganisms. A further course of emetine wasineffective but there was a dramatic response tometronidazole. (6) The fibrosis and contractureof the rectal stump, still present many monthsafter colectomy, suggested that complete re-generation of the colonic mucosa would havebeen virtually impossible. It is likely that the



j.com/


ovember 1970. D

ownloaded from

http://gut.bmj.com/


patient would have been left with a severe post-amoebic colitis.

CASE 2

A coloured female aged 23 years had diarrhoea(six stools per day) with blood and mucus forthree weeks before admission. For the last eightdays she had also had a stabbing pain in theright upper quadrant with bouts of nausea andvomiting. She had lost a great deal of weight.She was seven months pregnant. On examinationshe appeared ill. Her temperature was 1000F butshe was not dehydrated or jaundiced. The pulserate was 100 per minute and there was a systolicmurmur at the cardiac apex. The blood pressurewas 110/60. The chest and central nervous systemwas normal. Palpation of the abdomen revealeda uterus that was 30 weeks in size and the foetuswas lying as a breech. The liver was not enlargedbut there was a tender, well defined mass in theright upper quadrant just lateral to the fundus ofthe uterus. The urine contained a trace of proteinand trace of bilirubin, the haemoglobin was12 g %, the white cell count 20,500 mm3. and theESR 80 mm/hour. Sigmoidoscopically the rectalmucosa was inflamed and much blood and mucuswere seen. A small ulcer was visualized and a swabrevealed red blood cells, pus cells, and deadvegetative amoebae. In view of the pregnancy,emetine was not given but the patient received800 mg metronidazole three times daily, rolitetra-cycline (Reverin), 275 mg intravenously, andchloroquine diphosphate, 500 mg twice daily.The patient appeared to be responding to

treatment and the temperature gradually sub-sided. The pulse rate subsided to 80 per minute,but 48 hours after admission, she suddenlydeveloped severe generalized abdominal pain.Examination at that stage revealed guarding andrigidity of the whole abdomen and the mass wasno longer palpable.

Within six hours of the acute episode, laparo-tomy was performed. Exposure was obtained bya right subcostal incision (in view of the largeuterus) and this revealed a large, free perforationin the ascending colon just below the hepaticflexure. The hole had been plugged by largesheets of necrotic tissue and there was soiling ofthe peritoneal cavity. The regional colon wasoedematous and particularly friable around theperforation. The perforated bowel was exterior-ized by bringing it out through the wound andretained in position with a colostomy rod. In-spection of the lumen of the exteriorized bowelrevealed large necrotic ulcers with normal inter-vening mucosa. Biopsies were taken. Severaldrains were inserted into the peritoneal cavity.Histology showed necrotic tissue and exudatewith an extreme degree of autolysis, correspondingto that usually associated with amoebic colitis.While this appearance was suggestive of amoe-

biasis, it was not complete confirmation in theabsence of amoebae which had not been demon-strated (the patient had already had 48 hours ofanti-amoebic therapy). Another specimen ofbowel showed non-specific chronic inflammatoryinfiltrate. A small vessel in the submucosa con-tained a recent thrombus.The patient made a satisfactory postoperative

recovery and emetine, 1 grain daily, and chloro-quine, 300 mg twice daily, were given intra-muscularly. As soon as the ileus had subsided,metronidazole, 800 mg three times daily, wascommenced orally. Thirty-six hours after theoperation the patient delivered spontaneously ofa 30-week breech baby which subsequently diedfrom pulmonary complications. During therecovery phase from the first operation, sig-moidoscopy through the exteriorized colon wasperformed and regeneration of the ulcers wasvisualized.The patient maintained a steady postoperative

recovery. The stools were particularly loose andcontrol of the colostomy was difficult, but at nostage was there any digestion of the skin. Shewas ultimately discharged from hospital in a goodcondition and was re-admitted six weeks laterwhen the exteriorized colon was resected and anend-to-end anastomosis was performed. It wasinteresting to note that the mucosa had regener-ated completely but where there had been parti-cularly large ulcers, there was loss of mucosalfolds.

CommentExtension of a suspected colonic perforationoccurred while the patient was receiving anti-amoebic therapy and appeared to be showing agratifying clinical response. Exteriorization of theperforated area of the colon was undertaken andthis was followed by resection of the exteriorizedsegment and end-to-end anastomosis.

Discussion

Although amoebic colitis is a mild disease inmost patients, and readily cured by anti-amoebicmeasures, it occasionally presents as a severe oreven fulminating illness. Mortality figures arenot available for acute amoebic colitis, butWilmot (1950) estimates it at about 3%. Whenthe frequency of amoebiasis in South Africa andelsewhere in the world is considered, thisestimated mortality for amoebic colitis is quitesubstantial and second probably only to deathsdue to amoebic liver abscess. Furthermore, mostdeaths occur in the severe or fulminating varietyof the disease so that this extreme form probablyaccounts for nearly all of the 3% of deaths. Theincidence of fulminating colitis in all patients with



j.com/


ovember 1970. D

ownloaded from

http://gut.bmj.com/

945 Surgery in amoebic colitis

amoebic colitis is also not known. It can beassumed that the frequency is low, as manypatients with mild colitis are treated on an out-patient basis, and a further group probablyresponds fairly adequately to simple antibioticswithout diagnostic proof of the amoebic aetiologyof the bowel upset.A review of the causes of death in these severe

cases is fundamental to our suggested policy oftreatment. In a necropsy series reported by Kean,Gilmore, and van Stone (1956) the most commonfinding was peritonitis and there were 304%perforations in 148 fatal cases. This must indicatethat the patients either entered the hospital in amoribund condition with free or impendingperforation or that medical treatment was in-effective in controlling the disease before death.Despite this finding the authors fail to mentioncolectomy, or indeed, abdominal surgery in theirsection on treatment.

This fulminating form of the disease is usuallycharacterized by severe bloody diarrhoea, oftenwith mucosal sloughs, extreme toxaemia withhigh pyrexia, tachycardia, leucocytosis andsedimentation rate, or alternatively with a lackof homeostatic mechanisms with toxaemia, onlya slightly elevated temperature, leucocyte count,and sedimentation rate. In either case, the mostimportant diagnostic and prognostic signs arefound on abdominal examination. Abdominaldistension and/or tenderness should alert one tothe potential severity and possibly fatality of theoutcome. These signs may be present on admissionor develop during the course ofmedical treatment.Abdominal distension may be due to (1) para-

lytic ileus caused either by toxaemia, peritonitis,or electrolyte imbalance, or (2) toxic megacolonas found in ulcerative colitis. Localized or gener-alized abdominal tenderness suggests that theinflammatory process has traversed the muscularlayer and provides a pointer to subserosal exten-sion and potential or actual perforation of thebowel. An abdominal radiograph may revealeither subdiaphragmatic air (which is rare), smalland large bowel distension with or without air-fluid levels, or dilatation of the large bowel alone,the latter being characteristic of toxic megacolon.

It is our contention that surgical interventionmay become necessary at any stage in this formof the disease, and the results of the two patientsreported above suggest that this is a feasibleproposition with modern medical care and thewide variety of anti-amoebic measures nowavailable. Our policy at present is to instituteintensive medical measures consisting of naso-gastric suction, intravenous therapy, correctionof acidosis, hypokalaemia, and other electrolyteimbalance and, in addition, the administration ofanti-amoebic treatment and antibiotics. Frequentre-assessment of the patient's clinical status ismandatory with special attention to the abdominalsigns. Although each individual case should bejudged on its merits, increasing distension, and/or

tenderness despite electrolyte correcXion andanti-amoebic treatment, warrants the seriousconsideration of surgical intervention. A work-able approach might well be that (1) medicalmeasures should be continued if the patient'sgeneral and abdominal signs improve progress-ively or do not deteriorate. (2) Emergencysurgery should be seriously considered if thegeneral and local or radiological signs get worseafter 24 or 48 hours despite electrolyte correctionand anti-amoebic measures, or if acute episodesoccur, ie, sudden, increasing distension or acuteabdominal pain with increasing tenderness. Con-tinued 'watching' of the patient under thesecircumstances may result in a situation where thepatient is too ill for surgery to be undertaken.(3) Where the physical signs remain staticand profuse diarrhoea continues, with difficultyin maintaining electrolyte balance despite in-tensive therapy after five or more days, and noother cause for diarrhoea can be found, surgeryshould be contemplated. Amoebae can often nolonger be found in the stools but mucosalsloughing continues.

SURGERY IN ABDOMINAL AMOEBIASISWhile there are few indications for surgery inamoebic colitis, the following conditions may beadvanced for consideration.

(1) A liver abscess which fails to respond afterrepeated needle aspiration and a combination ofemetine, chloroquine, and metronidazole. Thismay require formal laparotomy and drainage(Rivers, Heibner, and Powell, 1955).

(2) Perforation of the bowel with localizedabscess formation not responding to medicalmeasures will require open drainage, and

(3) free perforation of the colon with intra-abdominal soiling. There would appear to be noplace for conservatism under these circumstances(Wilmot, 1962). Emergency laparotomy shouldbe undertaken and if the disease appears to belocalized to one segment of the bowel, then theperforated segment should be exteriorized orexcised and a proximal colostomy performed(Theron, 1947). After recovery, re-anastomosisshould be carried out.

(4) Fulminating amoebic colitis (as described)failing to respond or becoming worse on intensivemedical therapy. If laparotomy shows multipleareas of sloughing, gangrene, or impendingperforation, subtotal colectomy and colostomywith retention of the rectum should be carried out.Although Palaez et al (1966) mentioned this formof treatment, they state that the patient is usuallytoo ill for surgery. Clearly one should operatebefore the patient's condition makes surgery ahopeless proposition. If after treatment, therectum returns to normal, an ileorectal anasto-mosis may be carried out.

(5) In cases of severe postamoebic colitis, thatis, those patients in whom medical measures



j.com/


ovember 1970. D

ownloaded from

http://gut.bmj.com/


have succeeded in eradicating the amoebae butthe patient is left with an 'ulcerative colitis-like'state with anaemia, hypoalbuminaemia, and araised sedimentation rate. If the clinical courseis progressively downhill, the same surgicalindications as for severe ulcerative colitis shouldapply (Powell and Wilmot, 1966).

We should like to thank Dr I. N. Marks for helpfulcriticism and advice and Professor J. H. Louwfor permission to publish.

This study was supported in part by the Councilfor Scientific and Industrial Research and theMedical Research Council of South Africa.

References

Barker, E. M. (1958). Colonic perforation in amoebiasis S. Afr.med. J., 32, 634-638.

Cope, Z. (1920). Surgical aspects of dysentery. Lancet, 1, 579-585.Evans, T. C. (1925). The surgical aspects of amoebiasis. Trans.

roy. Soc. trop. Med. Hyg., 19, 282-311.Kean, B. H., Gilmore, H. R., Jr., and van Stone, W. W. (1956).

Fatal amoebiasis: Report of 148 fatal cases from ArmedForces. Institute of Pathology. Ann. intern. Med., 44,831-843.

Ochsner, A., and Debakey, M. (1942). Surgical amoebiasis.Int. Clin., 1, 68-99.

Pelaez, M., Villaz6n, A., and Zaraboso, R. S. (1966). Amoebicperforation of the colon. Dis. Colon. Rect., 9, 356-362.

Powell, S. J., and Wilmot, A. J. (1966). Ulcerative post-dysentericcolitis. Gat, 7, 438-443.

Rives, J. D., Heibner, W. C., and Powell, J. L. (1955). Thesurgical complications of amebiasis of the colon (exclusiveof liver abscess). Sarg. Clin. N. Amer., 35, 1421-1426.

Theron, P. (1947). Surgical aspects of amoebiasis. Brit. med. J.,2,123-126.

Wilmot, A. J. (1950). Clinical manifestations of amoebiasis in theBantu. D.M. Thesis, University of Oxford.

Wilmot, A. J. (1962). Clinical Amoebiasis, p. 39. Blackwell,London.



j.com/


ovember 1970. D

ownloaded from

http://gut.bmj.com/

surgery amoebic colitis - gut · collar-stud abscesses. it was surprising howlittle disability the...

Documents