A mass in the liverBMJ 2013; 346 doi: http://dx.doi.org/10.1136/bmj.f2036 (Published 03 April 2013) Cite this as: BMJ 2013;346:f2036
- Alex Mentzer, core medical trainee,
- Michael Jacobs, consultant
- Correspondence to: A Mentzer
A 38 year old woman presented with a three day history of fever, generalised myalgia and arthralgia, and the sensation of difficulty when breathing in. She had poorly localised discomfort in the left diaphragmatic region on deep inspiration. Her stools had been loose over the preceding three days but she had no other symptoms. Three days before presentation she had returned from a three week trip to Western Australia and 18 months earlier she had travelled to Vietnam.
On examination she appeared well but was febrile at 38.0°C. She was tachypnoeic at 22 breaths/min, but all other observations were normal. Examination of all other systems was unremarkable. Blood tests showed severe neutrophilia (18.9×109/L; reference range 2.0-8.1), a raised C reactive protein (354 mg/L; reference value <10; 1 mg/L=9.52 nmol/L), and normal renal and liver biochemistry. Chest radiography and urinalysis were normal.
After two days of empirical treatment with intravenous co-amoxiclav (1.2 g three times daily) and oral doxycycline (200 mg once daily), her condition had not changed. She continued to have fevers up to 39.5°C, and her inflammatory markers remained high. Blood cultures taken at presentation were negative.
Computed tomography of the chest, abdomen, and pelvis with intravenous contrast showed a rim enhancing abscess in the left lobe of the liver and a filling defect in the portal vein consistent with portal vein thrombosis.
1 What is the differential diagnosis?
2 Which further investigations or procedures would you undertake?
3 What treatment would you start?
4 What are the potential complications of this condition?
1 What is the differential diagnosis?
The main differential diagnosis of a large mass in the liver is a pyogenic or an amoebic liver abscess. A hepatic cyst complicated by haemorrhage or secondary infection could also present in a similar clinical way.
The clinical presentation and computed tomography findings (figure⇓) are consistent with a pyogenic or amoebic liver abscess, although a hepatic cyst that has undergone spontaneous haemorrhage or secondary infection could present in a similar way. Simple hepatic cysts are highly prevalent in the general population, but subsequent infective or haemorrhagic complications are rare. Complicated cysts lack contrast enhancement on computed tomography, making a cyst less likely in this case.
A pyogenic abscess results from adherence and accumulation of bacteria within the liver after inadequate clearance of the pathogen by the immune system. Such abscesses can be classified according to the pathogen’s route of access—through the biliary system, hepatic artery, portal vein; through direct extension from another focus of infection, such as cholecystitis; or after trauma. Several organisms are often involved in pyogenic abscesses. These organisms vary and depend on the source. The worldwide incidence is about 11 cases per million people per year, and the most common underlying disease process is biliary cholelithiasis. The most common pathogens are enteric Gram negative bacteria.1 2
The causative organism in amoebic liver abscess is Entamoeba histolytica. This protozoan spreads through the faecal-oral route as cysts in contaminated food and water. This accounts for the high frequency of infection and disease in areas of the world with poor sanitation. Cysts undergo excystation in the colonic lumen to form trophozoites. These multiply by simple binary fission, and in turn form new cysts that are excreted in the stool to complete the lifecycle. Mature trophozoites may migrate to and colonise the colon wall, triggering epithelial cell apoptosis and subsequent dysentery. Trophozoites less commonly cause invasive disease via the portal venous system, mainly resulting in amoebic liver abscesses.3 4 These abscesses comprise a centre of acellular proteinaceous necrotic debris surrounded by a rim of trophozoites that actively invade the peripheral liver tissue. The total parasite burden is small. Trophozoites subvert neutrophil killing to increase cell death as they invade the surrounding tissue.4 5 The precise steps involved in initial liver invasion and abscess formation are poorly understood.
Estimates of the incidence of amoebic liver abscesses in endemic areas in the developing world range from 30 to 200 cases per million population per year. Incidence in developed countries is much lower. It is estimated that just over three in every 1000 returning travellers presenting with a fever have an amoebic liver abscess.6 Men are four to 13 times more likely to develop an amoebic abscess than women. Because risk seems to increase in women after the menopause, hormonal differences may contribute to the relative protection in women.3 An HLA-DR3 haplotype was associated with an increased risk in a Mexican cohort, but this finding has not been replicated in other ethnic groups.7 HIV disease has, in general, not been associated with an increased risk of invasive amoebic disease, with the exception of some recent studies in Asia.4 8 9
The table⇓ summarises the clinical features associated with amoebic liver abscess in the largest case series to date. Comparative studies have shown that the two causes of liver abscess are clinically indistinguishable, and outside endemic areas the patient’s travel history is usually the key feature that suggests the possibility of amoebic abscess.10 11 12 In countries in the developing world where amoebiasis remains endemic, the diagnosis of either condition depends on a high level of clinical suspicion and prompt implementation of appropriate investigations. Older age (>50 years) and diabetes increase the probability of a pyogenic abscess. The most common clinical features in both types of liver abscess are fever and abdominal pain (table), but these are not present in all patients. Other potentially confusing features such as cough and dyspnoea are well recognised and, importantly, jaundice is an uncommon feature. Interestingly, only 10-35% of patients with amoebic liver abscess report diarrhoea.
There are two cohorts of patients with amoebic liver abscess: those who present acutely (within five to seven days of symptom onset, as with our patient) and those who present later (often more than two weeks after the onset of symptoms). Around two thirds of patients present acutely, and these are typically younger with high fever and abdominal tenderness. In contrast, those who present later are often older with low grade undulant fevers and relatively non-tender abdomens, but often with palpable hepatomegaly.13 The presentation of most patients with pyogenic liver abscess is similar to that of the acute presentation of amoebic liver abscess.2 Right upper quadrant tenderness on examination, particularly focal or punch tenderness over the liver, may be a clue to diagnosis of an abscess of either cause.
Ninety per cent of abscesses occur in the right lobe of the liver, irrespective of the underlying infection, reflecting the distribution of portal venous drainage.14 15 The clinical diagnosis of a left sided abscess may be particularly challenging because patients may be asymptomatic for a long time.16 In non-endemic areas, 95% of cases of amoebic liver abscess present within five months of travel to an area of high prevalence.4 There are isolated reports of amoebic liver abscesses presenting many years after apparent exposure,10 but, as in our case, it was not possible to prove the timing of acquisition in these cases, and sporadic transmission outside of high prevalence areas could not be excluded.
Simple blood investigations also cannot help distinguish between amoebic and pyogenic liver abscesses. Leucocytosis with predominant neutrophilia and a mild increase in serum aminotransferases are common in both conditions. Serum alkaline phosphatase may also be mildly raised, and this is more common in patients with amoebic abscesses.10 11 12 Imaging of the liver is essential if there is any clinical suspicion of a hepatic abscess. Studies have repeatedly shown little difference between the sensitivity of ultrasound and computed tomography in the diagnosis of hepatic abscess, so the modality used usually depends on availability.3
2 Which further investigations or procedures would you undertake?
Aside from abdominal imaging, amoebic serology should be performed. Urgent therapeutic aspiration of the abscess should be undertaken owing to its location in the left lobe of the liver and risk of rupture into the pericardium.
There is general consensus that all but the smallest pyogenic abscesses should be drained or aspirated to ensure resolution and minimise the risk of fatal sequelae. Aspiration may also be used for accurate diagnosis of a pyogenic abscess. Gram stain and culture of the pus may be the only clue to the causative organisms because peripheral blood cultures are positive only 50% of the time.1 Some studies have suggested that an uncomplicated pyogenic abscess smaller than 3 cm in diameter may be managed with antibiotics alone.17
Unlike pyogenic liver abscesses, most uncomplicated amoebic liver abscesses can be treated with drugs alone and do not require aspiration or drainage.18 The diagnosis of amoebic liver abscess is usually confirmed by amoebic serology, and rapid availability of results may prevent unnecessary invasive intervention. However, both in the developed and developing world, serology is generally performed in specialist reference laboratories, so there may be a considerable delay. For this reason, diagnostic aspiration of liver abscesses is often undertaken pending a serological result and may also have therapeutic consequences. The aspirate from an amoebic abscess is classically described as “anchovy paste,” which refers to its pinkish tinge.1 Most importantly, trophozoites are not seen in amoebic liver abscess aspirates so their absence does not weigh against the diagnosis. Clues to an amoebic liver abscess are absent or scanty white cells on microscopy and a negative Gram stain. Lectin antigen and polymerase chain reaction detection of E histolytica from aspirates for early diagnosis is under development, and initial results are encouraging.19 20
A variety of techniques have been developed to detect amoebic antibodies in serum.21 Enzyme linked immunosorbent assays (ELISAs) are most widely used and have good test performance characteristics for extraintestinal amoebic disease, including amoebic liver abscesses. Ninety five per cent of these cases will have a positive amoebic antibody ELISA. This test is less sensitive for amoebic colitis, with only 70% of cases having positive amoebic serology by ELISA.1 The test is positive in 5-10% of people with asymptomatic carriage of E histolytica. Therefore, a positive result does not prove the diagnosis of amoebic disease, especially in areas with a high prevalence of Entamoeba infection, and the result has to be interpreted in the clinical context. The diagnosis of amoebic liver abscess will be strongly supported if the patient has symptoms and signs suggestive of this condition and a positive ELISA.
Some countries in the developing world without resources for ELISA still depend on the indirect haemagglutination assay because of its relative ease of completion, widespread utility, and adequate performance statistics. This test is less sensitive than ELISA, however, and increases the risk of false negative results. Immunofluorescence antibody testing is available in some reference laboratories and is more sensitive and specific than ELISA. The risk of a false negative serological result is greatest early in invasive disease, so repeat testing one week later is useful in patients who present acutely and in whom the diagnosis is strongly suspected but serology is negative. The cellulose acetate precipitin assay is not widely used and is less sensitive than ELISA and immunofluorescence,22 23 but a positive result is highly specific for invasive amoebic diseases. A positive CAP assay result is therefore useful corroborative evidence in addition to other serological tests and it may become more widely used in reference laboratories in the developed world.
After successful treatment, ELISA and immunofluorescence tests will be negative within six to 36 months.1 21 Immunofluorescence antibody titres seem to fall more rapidly, so this test can help distinguish previous treated infection from repeat infection with invasive disease,24 25 although this may still be difficult if reinfection occurs early. The cellulose acetate precipitin assay almost always becomes negative within six months of successful treatment of invasive amoebic disease, so a positive result after this time strongly suggests reinfection.
There has been interest in developing sensitive assays for gastrointestinal infection with E histolytica to support the diagnosis of invasive disease in suspected cases. Amoebic culture, antigen detection, and polymerase chain reaction on stool specimens do not play a role in diagnosis of amoebic liver abscess because they are often negative as a result of the low gastrointestinal amoebic load and a positive result can be seen in asymptomatic carriers of E histolytica.26
In confirmed cases of amoebic liver abscess, therapeutic ultrasound guided aspiration or drainage is indicated in certain situations. Aspiration should be performed urgently if imaging suggests that rupture into an adjacent organ or body space is an imminent risk. This is based on an abscess diameter greater than 5 cm or an estimated volume of 300 ml or more.3 Therapeutic aspiration is also widely used in patients with left sided abscesses because of the danger of rupture into the pericardium (see below).27 Percutaneous or surgical drainage is essential in the management of abscesses that have already ruptured into peritoneal, pleural, or pericardial spaces.27 28 29 Failure of medical treatment after five to seven days also mandates therapeutic aspiration. Age older than 55 years may also predict the need for aspiration.30
3 What treatment would you start?
In addition to urgent therapeutic aspiration, broad spectrum antibiotics covering organisms that cause pyogenic liver abscesses should be started in accordance with local antibiotic policy. High dose oral metronidazole (800 mg three times daily) should also be given. If the diagnosis of amoebic liver abscess is confirmed by amoebic serology, metronidazole should be continued alone for 10 days, followed by a 10 day course of an intestinal amoebicide, such as paromomycin or diloxanide furoate.
Given the difficulty in differentiating pyogenic and amoebic abscesses clinically in any healthcare setting, it is common practice to provide antimicrobial cover for both conditions if an abscess is suspected or identified on imaging. Broad spectrum antibiotics should be used for pyogenic abscesses in accordance with local policy. The choice of antibiotic should be revised when the causative organism(s) is identified and sensitivity profiles obtained from abscess aspirates and blood cultures. Parenteral antibiotics are usually given for two to three weeks, followed by a further two to four weeks of oral equivalents.31
Empirical treatment for amoebic liver abscess is warranted in these cases pending serological confirmation. The mainstay of treatment is high dose oral metronidazole (or tinidazole), a systemic amoebicide. Metronidazole is curative in most cases without the need for aspiration or drainage.16 Metronidazole produces rapid clinical improvement, and resistance to nitroimidazoles in E histolytica has rarely been described.32 Therefore, if the fever does not abate within 96 hours of treatment initiation, an alternative diagnosis should be sought or the diagnosis of complicated amoebic liver abscess considered. Before trials showed the efficacy of metronidazole in the 1960s and 1970s, treatment relied on toxic agents such as dehydroemetine. This agent is now reserved for people who cannot tolerate nitroimidazoles.3
After 10 days of treatment with a systemic amoebicide, another agent that acts within the gut lumen to eradicate carriage of E histolytica must be given to reduce the risk of recurrence. Several luminal amoebicides have been used, but no controlled trials have compared their effectiveness. The most commonly recommended agents are the oral aminoglycoside, paromomycin, and diloxanide furoate, both of which are given for 10 days and have an efficacy of over 80% in eradication of gut carriage.33 Paromomycin is favoured by some experts3 because it has greater potency in vitro,34 and there is a suggestion that it is slightly more effective than diloxanide furoate.33 This is supported by the observation that paromomycin is also more effective in the treatment of symptomatic intestinal amoebiasis.34 35 However, diloxanide furoate is still widely recommended, including in the BNF,36 although its availability has been limited in recent years.
4 What are the potential complications of this condition?
Feared complications of amoebic liver abscess include pericardial rupture and cerebral amoebiasis, which have high mortality rates.
Amoebic liver abscess can be fatal. Death usually occurs because of extension of the abscess into other organs or body compartments. Many patients with uncomplicated abscesses have a reactive pleural effusion. However, rupture of the abscess through the diaphragm into the pleural space or lung is more common than rupture into the peritoneal space. Transdiaphragmatic rupture may manifest as chest or abdominal pain, or a cough that is sometimes productive of anchovy sauce pus. Microscopy of sputum or pleural aspirate may sometimes be positive for trophozoites. Patients with pleural complications may develop respiratory distress and shock but usually respond well to metronidazole and drainage with appropriate supportive care.28 37 A much less common but feared complication is rupture into the pericardium, which causes death in 30-60% of cases. Patients may present with early symptoms and signs of pericarditis, followed by acute cardiac failure or cardiac tamponade.16 28 Peritoneal rupture is heralded by acute onset of abdominal pain and development of peritonitis with associated shock. Laparotomy is usually indicated in such cases, in addition to antibiotics.
Cerebral amoebiasis is rare (less than one in 1000 cases of amoebic liver abscess worldwide) but often lethal. It is almost always associated with amoebic liver abscess and is caused by haematogenous spread of trophozoites to the brain.38 Patients with an amoebic liver abscess who have a reduced level of consciousness, seizures, or focal neurology should undergo urgent cross sectional imaging of the brain. Successful treatment usually involves surgical debulking alongside metronidazole.
Bacterial superinfection of an amoebic liver abscess is rare, occurring in 2.5% of 577 patients diagnosed with a hepatic abscess in a case series from Pakistan.11 The diagnosis should, however, be suspected in patients with positive blood cultures or positive cultures of abscess aspirates and positive amoebic serology. Such patients should respond to the locally guided empirical antibiotic regimen discussed above, with a switch to a tailored regimen based on organism culture and sensitivities. The possibility of iatrogenic superinfection after aspiration or drainage should always be considered.
Classic autopsy studies show an association of portal and hepatic vein thromboses with large amoebic abscesses, and early case reports described cases of severe portal hypertension associated with large abscesses and portal vein thromboses.39 40 The increased use of cross sectional imaging techniques has led to an increased number of reports of thrombotic complications associated with amoebic liver abscess, especially affecting the hepatic veins and inferior vena cava. Thrombus formation in the inferior vena cava, which can lead to obstruction, may be a consequence of direct extension of inflammation and compression of the vessel. These cases have been treated successfully with metronidazole, and therapeutic aspiration may be a rational adjunct. Treatment also includes anticoagulation.41 It is not clear how common this subclinical complication is or whether it influences clinical outcomes.
The complications associated with pyogenic liver abscesses mainly relate to sepsis, septic shock, and its associated features of acute kidney injury and liver failure.42 Rupture is less common and is usually peritoneal. Portal vein thromboses have been reported, as have pulmonary emboli, which—as with amoebic liver abscesses—are probably partly caused by direct venous compression.31 41 43
Immediate ultrasound guided aspiration of the abscess was performed and 20 ml of pus was aspirated. Microscopy of the pus showed scanty white blood cells; no organisms were identified on Gram stain. Amoebic liver abscess was considered, and high dose oral metronidazole (800 mg three times daily) was started. Culture of the aspirated pus was sterile. The diagnosis of amoebic liver abscess was confirmed by positive serology, with an immunofluorescence antibody titre of 1:640 and positive cellulose acetate precipitin. The patient was discharged to complete a 10 day course of oral metronidazole, followed by a 10 day course of oral paromomycin (500 mg three times daily).
As treatment for her portal vein thrombosis, she was given warfarin to achieve an international normalised ratio of 2.0-3.0 for three months. At follow-up, she had recovered completely. Repeat imaging at the end of warfarin treatment showed resolution of the portal vein thrombosis and the liver abscess.
Cite this as: BMJ 2013;346:f2036
We thank John Bligh for his contribution to the serology section of this article.
Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent obtained.