Hepatic abscess associated with subclinical ulcerative colitis

Hepatic abscesses are caused by a variety of intra-abdominal inflammatory conditions, and usually result as direct extension of bacteria or of portal pyemia. A case of hepatic abscess developing prior to the acute onset of ulcerative colitis without other intra-abdominal pathology is presented Can J Gastroenterol 1989;3( 3 ): 123-125


J IVER ABSCESS AS A COMPLICATION OF
L egional enteritis or ulcerative colitis is exceedingly uncommon, especially in the latter.Presented here is a case in which the development of a liver abscess imminently preceded the first clinical sign of ulcerative colitis.The authors propose chat chis is a rare example of a pyogenic liver abscess associated with ulcerative colitis possibly ansmg from portal vein bacrerem1a.
Hore/ Dieu Hospital , Kingscon.Onwno CASE PRESENTATION A 45 -year-old man presented co a Kingston emergency department in July, 1985 w1th a two week history of spiking temperature associated with chills, rigors, lethargy and an 8 kg weight loss.Prior co the onset of this acute illness th e patient had been completely healthy.There were no systemic or gastrointestinal symptoms to suggest active inflammatory bowel disease Physical examination revealed an illlooking man with a low grade temperature of 3 7. S 0 C. The abdomen was slightly distended and tender to palpation in the right upper quadrant, but no mass was present The liver was no t palp able.Laboratory investigations revealed the following: leucocyte count 16,900/L (normal 4000 to 10,000) with 75% polymorphonuclear leucocytes and 0% bands; hemoglobin concentration 101 g/L (normal 130 co 180); e rythrocyte sedimenca-t1 o n rate 120 mm/h (normal 1 to 7); aspartate aminotransferase (AST) 39 iu/L (normal l to 21); alanine ami notransferase (ALT) 69 iu/ L (normal 1 to 30); alkaline phosphatase 259 1u/L ( normal 34 to 108); and a total serum bilirubin of 10 µmoVL (normal 2 to 18).Cultures of sputum, urine and b lood were all negative for pathogens.
An ultrasound scan of the abdomen showed a hyperechoic mass in the posterior segment of the right lobe of the liver.A computer assisted tomography, (CAT) scan confirmed multiple small hypodense lesions m the right lobe of the liver suggestive of multiple liver abscesses (Figure l ).No air was identified in the mass by plain films or CAT scan.
The patient underwent an exploratory laparocomy.Careful examination of the intra-abdominal organs did not reveal any pathology of the appendix or the biliary tract and, aside from a few sigmoid diverticula, die findings were confined to the liver.There was no pericolonic ad hesions or mesenteric thickening to suggest recent o r remote episodes of diverticulitis.A large right lobe liver abscess was identified.incised and drained.Numerous smaller abscesses were found to be in contin uity with the large abscess cavity.The abscess cavity contai ned viscous yellow-whi te, foul smell ing pus.Stains and cultu res of the abscess fluid for bacteria (aerobes and anaerobes).fungi and acid-fast bacilli were all negative.Serology for Entamoeba histolytica, using the ELLSA method, was nonreactive.In spite of the negative bacteriological cultures the odour of the abscess was indicative of an anaerobic pyogenic abscess.lmproper culture or isolation techniques were presumed to be the reason for fai lure to isolate anaerobic bacteria.
The patient's postoperative cou rse was slow but uneventful.He was treated with intravenous metronidazole 500 mg every 6 h, netilmycin 80 mg every 8 hand penicillin 2 million units every 4 h.The drains were removed and the patient was discharged home two weeks after surgery, on oral merronidazole 500 mg 124 every 6 hand ccphalexin 500 mg every 6 h for two weeks.The patient presented again two weeks after discharge with crampy.abdominal pain and bloody diarrhea.There was no history of travel o utside O ntario.A preliminary diagnosisofantibiotic induced colitis was made, however, no clinical improvement occurred within two weeks of discontinuing the antibiotics.Multiple stool cultures for bacteria, ova and parasites were all negative for pathogens.Specific cultures fo r enterohcmorrhagic Escherichia coli were not obtained.Assays for Clostridtum difficle cytotoxin were also neganvc.
Flexible sigmoiJoscopy revealed a moderately inflamed rectum and sigmoid colon with mucosa!hyperemia.granularity and friabiliry.Histology of a rectal biopsy showed acute colitis with crypt abscesses consisten t with a diagnosis of ulcerative colitis.At this point, the patient was treated with oral prednisone 40 mg/day and su lfasalazine 2 g/day.Symptoms promptly resolved over a period of one week.
Once the acute phase of illness had subsided the remainder of the gastrointestinal tract was examined radiologically.An upper gastro in testinal series with small bowel follow through was normal.
A double contrast barium enema showed only mild diverticulosis The patient was subsequently followed as an outpatient.He was weaned off the prednisone within six wccb and then maintained on sulfasalazinc 2 g/day with good symptom control.A fo llow-up s1gmoidoscopy six weeks later revealed mild crythema and friability confined to the distal 20 cm of the colon.Computed tomography (CT) scan o f the abdomen showed no evidence of residual hepatic abscess.Complete blood count, erythrocyte sedimentation rate, AST, ALT and alkaline phosphatase were normal.One year postoperatively the patie nt is completely well on maintenance sulfasalazine.Sigmoidoscopy and biopsy at this rime were normal.

DISCUSSION
In developed countries the maJoTlty ofliver abscesses arc pyogenic in origin Only a minority arc related to Enwmoeba hiswlynca.The pathogenesis is usuall y related to biliary sepsis or portal vein hac-tcrem1a.However.hepatic artery bacteria and sepsis adjacent to the li ver can also be causative ( I ,2).
The incidence of portal vein bacteremia in patients with regio nal enterins or ulcerative colitis is unknown.Studies of patients with ulcerative colitis have suggested an increased incidence of portal bactercmia.However, the significance of this has not yet been determined (1,4).In spite of this possible increased incidence of portal bactcremia.the literature contains very few reports of liver abscess associated with regional enteritis or ulcerative colitis ( 5-8).ln fact, many authors do not even cite these two conditions when discussing the possible etiology of hepatic abscess.Lansbu rg ct al (6) reported only one case of hepatic abscess among 1333 patients with ulcerative colitis and this was in a patient with very severe disease Thus, evidence to date suggests that liver abscess as a complication of inflammatory bowel disease is a rare phenomenon.
The development of a liver abscess in patients with inflammatory howel disease can be due to suppurativc pylephlcbitis (9, 10), a fistulization into the biliary tree ( 11) or direct extension of bacteria from an adjacent subhcpatic absce$.The latter two occur only with C rohn\ disease while the former ran occu1 with ulcerauve coin is ( > ).
The presenr patient is unusual m that the hepatic abscess preceded the first clinical sign of ulcerative colitis.Thor ough review of any anteccndcnr history of similar large bowrl symptoms elicited one episode lasting two days of watery nonbloody diarrhea sevt'n years previously which resolved without diagnosis or treatment.This may represent an early episode of 'suhdm1,al' mflammatory bowel disease hut with no otht•r manifestations for seven years this seems less likdy It 1s 11nposs1hlc to provt' the exact euology of the present patient\ liver abscess.Howcwr.the tinly other operative pathology found was divertic-ulos1s.At nn time had the patient had signs or symptoms to suggest diwrt1cu-lit1s and the double contrast barium enema showed the d1verticular disease to he quite mild Thus, m VIC\\' of the more active disease process of ulcerati\'e colitis, i1 1s felt that the d1vert1culos1s was unrdated to the hepatic .ihsccs~ The fact that this p,Hient's abscess ap• peared to he sterile is not nn uncommon finding ma review of pyogen1c hepatic absccsst'S by McDonald and colle.igue,( 12).pathogens could not be isolated in at least r;, of cases.The pathogen in these cases was presumed to have been an anaerobe Sabbaj ( 13) suggested that only 2 5''~ of anaerobes arc recovered from hepatic abscesses with the most commonly isolated bt•mg a Gram-ros1uve cocci This poor microbiologic isolation of anaerobic bacteria 1s often due to improper handling of samples and poor culturing techniques Specific series suggest a high frequency of isolation of anaerobes when samples arc handled adequately and stringent anaerobic culture techniques ;ire 1mpkmented ( 14-16) Amoebic hepanc absces~ must ,1lways be excluded and serology ts nn excellent discriminator in distinguishing bt•twt'en amoebic and pyogt'111c abscesses The sens1t1v1ty of serology ts greater than 9','\, and ewn h1ghL'r with invasive amocbiasb ( I .16).In the present patient the serology was nonreactive and Encamoeba /rn- Submit your manuscripts at http://www.hindawi.com Correspondence and reprinrs: Dr Dale C. Mercer.Ass1sranr Pro/e.1sor.Deparcmenr of Surgery, Queen's Ur11versity.Hore/ D1eu Hospical Kingswn, Onrario Kil 502 Receit•ed for pablicarion January 3, 1989.Accepced March 8. 1989 C.~N J GASTROENTEROL VOL 3 No 3 )L 'IE 1989