Abstract
Background
Liver abscess is a serious disease traditionally managed by open drainage. The advances in interventional radiology over the last two decades have allowed a change in approach to this condition. We have reviewed our experience in managing liver abscess over the last 7 years.
Methods
Details of all patients admitted with liver abscess between 1995 and 2002 were prospectively entered onto our database. A review was performed to document the use of imaging and drainage techniques. Aetiology, morbidity, mortality and duration of hospital stay were recorded.
Results
Forty-two patients (median age 53 [22–85] years; M:F 18:24) were admitted with liver abscess (multiple abscess 20); 19 cases were of portal tract origin, 16 cases were of biliary tract origin and 7 cases were spontaneous. Forty-one patients were managed non-operatively, all received antibiotics (cephalo-sporins 76%, metronidazole 88%, quinolones 33%). Diagnosis was made on ultrasound scan (22) or CT (20). Five patients were managed with antibiotics alone. Fifteen patients were managed initially with percutaneous aspiration and five subsequently required percutaneous drainage. Twenty-one patients had primary percutaneous drainage, nine requiring a further procedure (aspiration 3, drainage 6). One patient underwent hepatic resection. Median hospital stay was 16 (6–35) days. There was one death, but no procedure-related morbidity.
Discussion
Non-operative management of solitary and multiple liver abscesses is safe and effective.
Keywords: Pyogenic, liver, abscess, antibiotics, aspiration, drainage
Introduction
Pyogenic hepatic abscess is a rare but potentially lethal condition, with a reported incidence of 20 per 100000 hospital admissions in a western population 1. Historically the commonest cause has been acute appendicitis 2, but in the era of modern surgery and antibiotics, this cause has declined drastically in importance. Other causes of ascending portal pyaemia have become increasingly important; in recent series the biliary tract has been the commonest source of sepsis, due to gallstones, bile duct strictures, endoscopic interventions and operation 1,3. The increasing medical use of immunosuppression also provides a fertile environment for the development of this condition 4.
The management of pyogenic hepatic abscess has changed greatly over the last three decades. Traditionally liver abscess has been regarded as a high-morbidity disease, routinely managed by open drainage, with mortality rates between 9% and 80% 2,4,5,6,7,8,9. Advances in interventional radiology over the last three decades have allowed a change in approach to this condition. Ultrasound imaging and computed tomography (CT) have made early diagnosis possible and allow accurate guidance of percutaneous aspiration or drainage 3,10,11. In combination with modern antibiotics, these percutaneous techniques now form the mainstay of treatment for liver abscess 1,3,12,13,14- We have reviewed our experience in managing liver abscess over the last 7 years, to document the current aetiology of this disease and the results of routine percutaneous aspiration and drainage.
Methods
Details of all patients admitted with liver abscess between January 1995 and August 2002 were prospectively entered onto our database. A review was performed to document the use of imaging and drainage techniques during this period. Aetiology, treatment morbidity, mortality and duration of hospital stay were also recorded.
Patients with parasitic disease or tuberculosis as the causative organism were excluded. Similarly patients with intrahepatic extension of gallbladder empyema were not included. Subphrenic and subhepatic abscesses were excluded unless shown to be contiguous with an intrahepatic abscess. The final exclusion was for patients with infected central necrosis of primary or secondary hepatic tumours.
Aspiration and drainage of abscesses were always performed under local anaesthesia, with either ultrasound or CT guidance. Aspiration was performed with an 18-gauge needle. Drainage was achieved with a variety of different 8–12 Fr all purpose drainage catheters placed by the Seldinger technique.
Results
Patients
Forty-two eligible patients were admitted with pyogenic liver abscess during the study period. Full details were available on all patients. The median age was 53 years (range 22–85) and there were 24 women. Twenty patients had multiple abscesses.
Aetiology
The causes of pyogenic liver abscess could be divided into three broad categories: portal tract sepsis, biliary sepsis and spontaneous. Nineteen cases were due to portal tract sepsis and ascending portal pyaemia, diverticular disease (12 patients) being the single most common cause of pyogenic liver abscess in this series (Table 1).
Table 1. Primaryof pyogenic liver abscess in 42 patients.
| Primary aetiology | Number of cases |
|---|---|
| Portal tract sepsis | |
| Diverticulitis | 12 |
| Crohn's disease | 2 |
| Appendix abscess | 1 |
| Ischaemic colitis | 1 |
| Acute pancreatitis | 1 |
| G1 tract cancer | 2 |
| Total | 19 |
| Biliary tract sepsis | |
| Malignant biliary stricture | 6 |
| Benign biliary stricture | 1 |
| Gallstones | 6 |
| Post liver transplant | 2 |
| Blunt abdominal trauma | 1 |
| Total | 16 |
| Spontaneous | 7 |
Sixteen patients had ascending biliary sepsis, most of which (10 patients) were due to benign causes (Table 1). Eleven of these patients had undergone previous biliary tract surgery (Table 2). Seven patients had also had an ERCP, four with stent placement, and one further patient had a percutaneous biliary stent placed.
Table 2. Surgical(n = 11) and radiological (n = 8) interventions performed in 16 patients before the development of hepatic abscess of biliary tract origin.
| Operation or procedure | Number of patients |
|---|---|
| Liver transplant | 2 |
| Biliary reconstruction | 3 |
| Whipple's procedure | 1 |
| Laparotomy for liver trauma | 1 |
| Open cholecystectomy | 1 |
| Laparoscopic cholecystectomy | 2 |
| Bile duct exploration | 1 |
| ERCP (+biliary stent) | 7(4) |
| PTC and stent | 1 |
Seven patients had no readily identifiable cause of pyogenic hepatic abscess and were termed spontaneous. These patients were investigated with ultrasound scan, magnetic resonance (MR) cholangiography (3 cases), liver MR imaging (3 cases) and barium enema (5 cases) to exclude the common underlying biliary, hepatic and colonic causes of liver abscess; one patient also had a liver biopsy to exclude underlying malignancy.
Overall 19 patients from the three groups were immunocompromised or had underlying malignancy (Table 3), while four of these patients had both.
Table 3. Secondary factors compromising immune function and predisposing to the development of pyogenic hapatic abscess in 19 patients.
| Secondary factors | Number of cases |
|---|---|
| Malignancy | |
| Cholangiocarcinoma | 2 |
| Pancreatic carcinoma | 3 |
| Neuroendocrine tumour | 1 |
| Gastric carcinoma | 1 |
| Caecal carcinoma | 1 |
| Lymphoma | 1 |
| Pharmocological immunosuppression | |
| Chemotherapy | 2 |
| Post liver transplant | 2 |
| Autoimmune disease | 2 |
| Inflammatory bowel disease | 2 |
| Asthma | 1 |
| Other | |
| Diabetes mellitus | 4 |
| Hereditary immunodeficiency syndrome | 1 |
Investigation and imaging
Initial diagnosis was made on ultrasound scan (22) or CT (20). There was one hepatic abscess that was not identified on ultrasound, giving a sensitivity of 96%. CT was 100% sensitive for hepatic abscess in this series.
Microbiology
Positive blood or pus cultures were obtained in 31 patients (Table 4). Coliforms were the most common isolate in liver abscesses of both biliary and portal origin. Gram-positive cocci accounted for most spontaneous liver abscesses. The portal group were more likely to have a mixed growth on culture (47%) than either the biliary (25%) or spontaneous (0%) groups.
Table 4. Results of blood and pus cultures in 41 patients with pyogenic hepatic abscess.
| Organisms | Biliary (n = l6) | Portal (n = l9) | Spontaneous (n = 7) |
|---|---|---|---|
| No growth | 4 | 6 | 1 |
| Streptococci | 5 | 4 | 3 |
| Staphylococci | 3 | 1 | 2 |
| Pseudomonas | 1 | 0 | 0 |
| Klebsiella | 0 | 3 | 0 |
| Coliforms | 6 | 12 | 1 |
| Bacteroides | 0 | 3 | 0 |
| Proteus | 1 | 1 | 0 |
| Mixed growth | 4 | 9 | 0 |
Treatment
Forty-one patients were managed non-operatively. All received antibiotics, as follows: cephalosporins 76%, metronidazole 88%, ciprofloxacin 33%, imipenem 12%, Tazocin 10%, amoxycillin 12%, gentamicin 12%. Five patients were managed with antibiotics alone; two had multiple small abscesses, two had a small solitary abscess and one had a large abscess involving most of segments 2 and 3. Patients were maintained on intravenous antibiotics until clinical signs of bacteraemia resolved; oral antibiotics were then given for a further 4–12 weeks.
Fifteen patients were managed initially with percutaneous aspiration and five subsequently required percutaneous drainage. Twenty-one patients had primary percutaneous drainage, nine requiring a further procedure (aspiration 3, drainage 6). Four drains fell out, one patient subsequently required repeat drainage and one underwent percutaneous aspiration.
One patient underwent operation for hepatic abscess. This was a 48-year old woman who presented at another hospital with cholangitis, right portal vein thrombosis and acute renal failure. She underwent open bile duct exploration and insertion of a T-tube, but was transferred to this unit 4 days later with multi-organ failure, multiple hepatic abscesses and a bile leak. Further CT scans revealed infarction of the right lobe of the liver. She was taken to theatre and underwent an emergency right hemihepatectomy. She made a good recovery and was discharged after 35 days.
Hospital stay
Median hospital stay was 16 (6–35) days.
Morbidity
One patient developed Clostridium difficile colitis after a month on broad-spectrum antibiotics. She responded to oral metronidazole and made a complete recovery.
There were no serious complications directly related to either placement of the drain or its subsequent removal. Although one patient had a 0.62 mmol/L (4 g/dL) fall in haemoglobin after drain insertion, there were no cases of post-procedure haemorrhage sufficient to require blood transfusion.
One patient had a biliary fistula after drain insertion, but in this case the abscess was an infected intrahepatic biloma, which had developed following blunt abdominal trauma with extensive hepatic parenchymal disruption and intrahepatic bile duct injury. Thus this was an anticipated outcome of the procedure and has not been classified as a complication. The fistula resolved after endoscopic biliary stenting.
Mortality
There was one death, in a 68-year old man with hepatic abscess secondary to ischaemic colitis who was on chemotherapy for non-Hodgkin's lymphoma. He presented with generalised peritonitis and underwent emergency laparotomy and right hemicolectomy for bowel infarction. He remained septic and further investigation revealed a hepatic abscess, which was drained percutaneously; he had established multi-organ failure and died 4 days after operation.
Discussion
Cases of pyogenic hepatic abscess seen in a large hepatobiliary tertiary referral centre include many complicated cases. Patients have often had prior operation or ERCP (Table 2) and may have several co-morbid conditions, underlying malignancy or be pharmacologically immunocompromised (Table 3). These results are comparable to those from other large reported series 1,3,6,12 and demonstrate that despite complex clinical scenarios a high level of success can be achieved using an aggressive policy of percutaneous drainage and aspiration, combined with high-dose broad-spectrum intravenous antibiotics.
The decision as to which treatment modality is most appropriate is based on clinical and radiological grounds. A stable patient with few signs of sepsis or an incidental finding of liver abscess can be managed safely with a trial of antibiotics alone, clinical signs of septicaemia with multi-organ dysfunction make aspiration or drainage mandatory. In multiple abscesses it is usually only the large or dominant collection which is drained, smaller abscesses are aspirated or treated with antibiotics alone. The decision as to whether to leave a drain in the abscess is made by the radiologist during the procedure; successful aspiration with collapse of the cavity makes drain placement unnecessary. Where aspiration appears complete drainage is not routinely performed since there is often discomfort and morbidity associated with an indwelling hepatic catheter. Very thick pus often cannot be satisfactorily aspirated or drained, under these circumstances aspirate is sent for culture and the patient is treated with appropriate antibiotics. The patient must be closely monitored; clinical deterioration or radiological evidence of progressive liquefaction triggers a repeat attempt at aspiration or drainage.
The high level of prior hepatobiliary surgical and endoscopic interventions in the biliary group reflects the expansion of surgical activity in this field. Any biliary tract instrumentation increases the risk of bacterial contamination of the biliary tree, with the consequent potential for ascending sepsis and abscess formation. Although most series indicate that biliary infection is now the leading cause of pyogenic hepatic abscess in western populations 1,3,4,6,12, it is interesting to observe the large number of portal tract causes seen in this series. Diverticular disease is an established cause of ascending portal pyaemia, yet we observed a higher incidence of this than most other series 1,12. This may be in part due to our policy of performing detailed colonic investigations in patients with unexplained hepatic abscess. It remains important to try and elucidate the cause of hepatic abscess in patients initially labelled as spontaneous, in order to exclude an otherwise treatable cause that might lead to recurrent disease. We advocate performing MR cholangiography in patients with biliary symptoms, an obstructive liver function test picture or ultrasound evidence of cholelithiasis or bile duct dilatation. In patients without these findings, we routinely perform an MRI of the liver to identify any hepatic parenchymal abnormality and an outpatient barium enema to exclude a colonic source of portal pyaemia, after resolution of the acute illness. Three of the remaining cases of spontaneous hepatic abscess in this report were in patients who were immunocompromised, leaving only four patients (10%) with truly unexplained hepatic abscess. A high level of immunocompromise or underlying malignancy was observed in this series (46%). Current management of portal tract and biliary sepsis with antibiotics plus endoscopic or surgical intervention makes progression to hepatic abscess a rare event in most cases. However, patients with advanced malignancy or impaired immune function remain at particular risk of liver abscess 4.
Two cases of liver abscess were seen in patients who had previously received an orthotopic liver transplant. In each case this was the presenting feature of an ischaemic bile duct injury due to late onset hepatic artery thrombosis. In patients with liver abscess and a past history of liver transplant, one must always consider underlying arterial occlusion as a possible cause of biliary tract stricture and subsequent abscess.
The absence of any serious procedure-related morbidity highlights the effectiveness of the non-operative approach in this group of complex, critically ill patients. The contrast with the results of series advocating a policy of frequent open drainage is stark 1,4. The one patient in our series who underwent operation presented with a clinical scenario that made operation unavoidable. Portal vein thrombosis is a recognised complication of severe cholangitis, but in this case it occurred in conjunction with a pre-existing iatrogenic injury to the right hepatic artery, which combined with the biliary sepsis to produce multiple abscesses in the necrotic right hepatic lobe. Right hemihepatectomy was the only possible approach. Two other patients developed right portal vein thrombosis, both following portal tract sepsis, but neither had compromise of the hepatic arterial supply. Both also developed multiple abscesses; one resolved on antibiotics alone, while the other required percutaneous drainage. Although the CT scan appearance is dramatic, isolated lobar portal vein thrombosis should not alter management of the underlying abscess. The only death in this series was in a patient who was in extremis on intensive care, following operation for ischaemic colitis before presentation of the abscess.
Portal and biliary sources of sepsis are equally important causes of liver abscess in the UK. Prior biliary tract operation or underlying immunocompromise are implicated in the development of most pyogenic hepatic abscesses, while truly spontaneous hepatic abscess is rare. Non-operative management of solitary and multiple liver abscesses is safe and effective.
References
- 1.Huang C-J, Pitt HA, Lipsett PA, et al. Pyogenic hepatic abscess. Ann Surg. 1996;223:600–9. doi: 10.1097/00000658-199605000-00016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Oschner A, DeBakey M, Murray S. Pyogenic abscess of the liver. Am J Surg. 1938;40:292–353. [Google Scholar]
- 3.Giorgio A, Tarantino L, Mariniello N, et al. Pyogenic liver abscess: 13 years of experience in percutaneous needle aspiration with US guidance. Radiology. 1995;195:122–4. doi: 10.1148/radiology.195.1.7892451. [DOI] [PubMed] [Google Scholar]
- 4.Branum GD, Tyson GS, Branum MA, Meyers WC. Hepatic abscess. Ann Surg. 1990;212:655–62. doi: 10.1097/00000658-199012000-00002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Barbour GL, Juniper K., Jr A clinical comparison of amoebic and pyogenic hepatic abscess of the liver in sixty six patients. Arm Intern Med. 1972;77:629. doi: 10.1016/0002-9343(72)90175-1. [DOI] [PubMed] [Google Scholar]
- 6.Farges O, Leese T, Bismuth H. Pyogenic liver abscess: an improvement in prognosis. Br J Surg. 1988;75:862–5. doi: 10.1002/bjs.1800750910. [DOI] [PubMed] [Google Scholar]
- 7.Rubin RH, Swartz MN, Malt R. Hepatic abscess: changes in clinical, bacteriologic and therapeutic aspects. Am J Surg. 1974;57:601–10. doi: 10.1016/0002-9343(74)90012-6. [DOI] [PubMed] [Google Scholar]
- 8.Pitt HA, Zuidema GD. Factors influencing mortality in the treatment of pyogenic hepatic abscess. Surg Gynecol Obstet. 1975;140:228–34. [PubMed] [Google Scholar]
- 9.Miedema BW, Dineen P. The diagnosis and treatment of pyogenic liver abscesses. Ann Surg. 1984;200:328–35. doi: 10.1097/00000658-198409000-00010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Berger LA, Osborne DR. Treatment of pyogenic abscesses by percutaneous drainage. Lancet. 1982;1:132–4. doi: 10.1016/s0140-6736(82)90381-6. [DOI] [PubMed] [Google Scholar]
- 11.Gerzof SG, Johnson WC, Robbins AH, Nabseth DC. Intrahepatic pyogenic abscesses: treatment by percutaneous drainage. Am J Surg. 1985;149:487–94. doi: 10.1016/s0002-9610(85)80045-3. [DOI] [PubMed] [Google Scholar]
- 12.Rintoul R, O'Riordain MG, Laurenson IF, Crosbie JL, Allan PL, Garden OJ. Changing management of pyogenic liver abscess. Br J Surg. 1996;83:1215–18. [PubMed] [Google Scholar]
- 13.Ch Yu S, Hg Lo R, Kan PS, Metreweli C. Pyogenic liver abscess: treatment with needle aspiration. Clin Radiol. 1997;52:912–16. doi: 10.1016/s0009-9260(97)80223-1. [DOI] [PubMed] [Google Scholar]
- 14.Rajak CL, Gupta S, Jain S, Chawla Y, Gulati M, Suri S. Percutaneous treatment of liver abscesses: needle aspiration versus catheter drainage. Am J Roentgenol. 1998;179:1035–9. doi: 10.2214/ajr.170.4.9530055. [DOI] [PubMed] [Google Scholar]
