In the United States, amebic liver abscess has become a rare disease that is found almost exclusively in travelers and immigrants. In 1994, the last year in which incidence data were collected, there were only 2983 cases of amebiasis; the percentage of cases complicated by abscess is unknown, but it is probably well under 10%, roughly 1 case per million persons per year.¹ Worldwide, contamination of food and drinking water has maintained E. histolytica infection second only to malaria as a cause of death from parasitic disease. The epidemiology of this disease has been greatly informed by the appreciation that a closely related nonpathogenic species, Entamoeba dispar, colonizes between 5% and 25% of persons.^2,3 E. dispar has no apparent propensity for invasive disease, even among patients with acquired immunodeficiency syndrome (AIDS). In industrialized countries, most asymptomatic individuals with Entamoeba in their stool are colonized with E. dispar, whereas in highly endemic regions, asymptomatic infection with E. histolytica may exceed the rate of E. dispar colonization.^4,5,6–11 In addition, other Entamoeba species have been identified; these include Entamoeba moshkovskii, which causes noninvasive diarrhea, and Entamoeba bangladeshi, which is of unknown virulence. Because these species are indistinguishable by light microscopy, the presence of Entamoeba in the stool is not sufficient to establish the cause of a liver abscess. Men and women experience equivalent rates of E. histolytica infection, but adult men are at a 10-fold higher risk for invasive disease (colitis or extracolonic disease) than other populations.¹² The reason for this disparity remains uncertain, although gender differences in complement-mediated killing of E. histolytica and cytokine responses to amebic liver infection have recently been proposed.^13,14 In addition, testosterone has been shown to be a risk factor for the development of amebic liver abscesses in experimental animals.^15,16

Infection with E. histolytica results from ingestion of cysts in fecally contaminated food or water. Excystation occurs in the intestinal lumen, and trophozoites migrate to the colon, where they adhere by means of a lectin that specifically binds galactose N-acetyl-D-galactosamine (the Gal/GalNAc lectin) on colonic epithelium and multiply by binary fission.^19,20 At high densities, trophozoite encystation is initiated, and newly formed cysts are released into the stool to complete the life cycle. Most individuals experience asymptomatic infection, but approximately 10% develop symptomatic colitis when invasion of the colonic mucosa occurs. Spread to the liver via the portal system occurs in less than 1% of cases.² A number of virulence factors have been implicated in the development of amebic liver abscess, including small proteins (amoebapores) that punch holes in lipid bilayers of target cells, cysteine proteases, and the Gal/GalNAc lectin.^21–27 E. histolytica induces apoptosis in hepatocytes and neutrophils, forming large, nonpurulent, “anchovy paste” abscesses that grow inexorably without treatment.^28,29 The mechanism by which E. histolytica induces apoptosis is unknown, but its importance is underscored by the resistance of caspase-3-deficient mice to amebic liver abscess formation.^30,31 Recently, an E. histolytica surface protein that participates in apoptotic corpse phagocytosis, the phagosome-associated transmembrane kinase (PATMK), was shown to be required for intestinal amebiasis in a murine infection model; however, PATMK was found to not contribute to the development of experimental amebic liver abscess.³² By contrast, amoebapore expression was found to be required for the formation of liver abscesses but not amebic colitis.²⁵ These findings suggest that the contribution of particular amebic virulence factors to pathogenesis may be tissue specific, and many of these factors have been recently reviewed.³³

Predisposition to liver disease is influenced by host genetics. In studies of Mexican mestizo children and adults, individuals with the major histocompatibility complex haplotype HLA-DR3 have an increased frequency of amebic liver abscess compared with healthy populations of the same socioeconomic background.^34,35 Host factors thought to be critical in the containment and clearance of invasive amebae include complement, neutrophils, interferon-γ (IFN-γ), nitric oxide, and adaptive immune responses (see Chapter 274).^{13,14,36–38,39,40}

Although E. histolytica is the only species of Entamoeba known to cause invasive infections, certain strains may be more proficient at causing liver disease than others. In surveys of clinical E. histolytica isolates, genetically distinct strains recovered from amebic liver abscesses were infrequently found to cause asymptomatic intestinal colonization or intestinal disease.^75,76 More recently, Ali and colleagues⁷⁷ analyzed paired isolates from patients with concurrent amebic liver abscesses and intestinal infection. In all cases, the liver and intestinal amebae had discordant genotypes, suggesting that either amebae undergo genetic reorganization during invasion or that only a subset of strains is capable of metastasizing to the liver.

Patients with amebic liver abscess typically present with fever and a dull, aching pain localizing to the right upper quadrant, but jaundice is rare. Only 15% to 35% of patients present with concurrent gastrointestinal symptoms, including nausea, vomiting, abdominal cramping, or diarrhea (Table 77-3).^{18,42–45,46–71,96,97–103} Symptoms are acute (<2 weeks’ duration) in about two thirds of cases but can develop months to years after travel to an endemic area.^2,3 The presentation is indistinguishable from pyogenic liver abscess on clinical grounds, and a careful search for epidemiologic risk factors is of paramount importance. Corticosteroid use and male sex are well-established risk factors for invasive amebic disease.^12,96

Diagnosis

The diagnosis of liver abscess should be suspected in all patients with fever, leukocytosis, and a space-occupying liver lesion. Because of the nonspecific symptoms on presentation, the initial clinical impression is frequently wrong and may include cholangitis, pneumonia, hepatic malignancy, intra-abdominal catastrophe, or pneumonia.⁴⁶ Before the widespread use of noninvasive imaging, liver abscess was among the most frequently identifiable causes of fever of unknown origin. Leukocytosis is present in most patients; recent studies^{65,66,69–72} report leukocytosis in 68% to 88% of all patients, with mean white blood cell counts of 15,000 to 17,000/mm³. An elevated alkaline phosphatase concentration is the most frequently abnormal liver function test, occurring in about two thirds of patients with liver abscess, but a normal value does not exclude the diagnosis. Abnormalities of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and bilirubin are generally small, although they may be more pronounced in some patients with biliary disease. Albumin concentration and prothrombin time tend to be normal or nearly so. Chest radiographs are abnormal about half of the time but are of no real value in making the diagnosis. Therefore, although laboratory studies may suggest liver abnormalities, they are of no use in making the diagnosis of pyogenic liver abscess, and a high index of suspicion is required if the diagnosis is to be made in a timely fashion. Laboratory abnormalities may be of prognostic significance, most likely as markers of comorbidities. A multivariate analysis of risk factors found that a hemoglobin concentration of less than 10 g/dL and a blood urea nitrogen (BUN) concentration greater than 28 mg/dL were independent predictors of mortality in patients found to have pyogenic liver abscess (odds ratios of 13 and 14, respectively).⁶¹

Once the diagnosis is suspected, radiographic imaging studies are essential to diagnose pyogenic liver abscess.⁶² Ultrasonography and computed tomography (CT) scanning have proved particularly useful for demonstration and drainage of abscesses. Ultrasonography is the study of choice in patients with suspected biliary disease and in those who must avoid intravenous contrast or radiation exposure; it has a sensitivity of 70% to 90%. Contrast-enhanced CT scanning has improved sensitivity (≈95%) and is superior for guiding complex drainage procedures. Intravenous contrast is required for optimal imaging in two thirds of patients.¹⁰⁴ Magnetic resonance imaging (MRI) studies are seldom required, but they may be better at distinguishing abscesses from noninfectious liver lesions such as neoplasia. Fine-needle aspiration is the definitive diagnostic procedure, and MRI is a cumbersome tool for guiding drainage procedures.

In patients with pyogenic liver abscess, blood cultures are positive about half of the time. It is imperative that multiple sets of anaerobic and aerobic cultures be obtained because these are frequently the only cultures obtained before antibiotic administration, and in about 7% of cases they are the only positive culture data obtained.^50,57,69 The diagnosis ultimately rests on obtaining purulent material from an aspiration of the abscess cavity, usually under radiographic guidance. Failure to obtain the expected pus should prompt a reevaluation of the differential diagnosis, considering liver cyst, malignancy, and amebic liver abscess. Purulent material should always be sent for Gram stain, which may provide the only clue to a mixed infection in patients heavily treated with antibiotics. The importance of prompt delivery of anaerobic specimens under proper conditions cannot be stressed enough.

Patients with amebic liver abscess classically present with a single large abscess in the right lobe of the liver, but the abscess can be anywhere and multiple lesions are not infrequent. Moreover, there is near-complete overlap in the symptomatology of the two diseases (see Table 77-3). A recent study from Spain showed that age 45 or younger, presence of diarrhea, and a solitary right lobar abscess favored amebic etiology.¹⁰⁵ Distinguishing amebic from pyogenic liver abscess is not possible with the use of clinical criteria alone; however, a presumptive diagnosis of amebic liver abscess can be made in a patient with positive serology and a space-occupying liver lesion on CT or ultrasonography. Amebic serology (antiamebic antibodies) has a sensitivity of about 95% and is highly specific for E. histolytica infection.

Although the test can be negative in a minority of acute presentations (symptom duration <2 weeks), a repeat serology determination is usually positive.^106–109 It should be remembered that a positive serology result only confirms present or prior E. histolytica infection and cannot distinguish colitis from extraintestinal disease. Antigen detection represents a complementary diagnostic method. A commercially available ELISA assay that detects the Gal/GalNAc lectin has been shown to be positive in more than 95% of serum samples and about 40% of stool or abscess specimens in patients with amebic liver abscess.¹¹⁰ Several other U.S. Food and Drug Administration–approved ELISA antigen detection assays are also available, but some may cross-react with E. dispar.^111,112 Microscopic examination of the stool for cysts is of little value because E. histolytica cannot be distinguished from E. dispar. Examination of aspirated liver abscess pus is also not recommended, given that trophozoites are identified in only 11% to 25% of cases.^110,113 Nucleic acid amplification methods are promising diagnostic tools for invasive amebiasis and amebic liver abscess, but their use is currently limited to research laboratories.^{112,114–116} Some investigators have argued that in areas of low endemicity, suspected amebic liver abscess should be aspirated to exclude pyogenic liver abscess. Certainly, if there is no response to initial therapy, the abscess should be aspirated to confirm the diagnosis and to exclude pyogenic liver abscess. Bacterial superinfection of amebic liver abscess has been described in about 1% to 5% of cases, frequently as a complication of drainage procedures.¹¹⁷

Echinococcal cysts of the liver, which typically do not cause fever or tenderness, are distinguishable from amebic and pyogenic abscesses by CT but require separate consideration for aspiration (see Chapter 291).

Amebic liver abscess can almost always be treated with medical therapy alone. Metronidazole (750 mg three times daily) is typically given for 7 to 10 days.^2,3 An alternative, tinidazole (2 g daily for 3 days), has been used extensively in Europe and the developing world and has been approved for use in the United States for the treatment of amebiasis.¹¹⁸ Other nitroimidazoles with extended half-lives that are efficacious but not currently available in the United States include secnidazole and ornidazole. Decreased fever and abdominal pain are usually observed within 3 to 5 days after initiation of therapy. Patients frequently remain colonized with E. histolytica despite nitroimidazole treatment and should be treated with paromomycin, a nonabsorbable aminoglycoside with E. histolytica activity to eliminate this condition. It is generally accepted that uncomplicated amebic liver abscess does not require drainage. Percutaneous image-guided aspiration has replaced surgical drainage and should be employed if there is no response to appropriate therapy or the diagnosis is uncertain, to exclude pyogenic liver abscess and bacterial superinfection. Drainage should also be considered for large lesions at risk for rupture, particularly left-sided abscesses that can rupture into the pericardium. Percutaneous drainage has not been demonstrated to shorten hospital stay or to hasten clinical improvement, with the exception of one randomized controlled trial that found a salutary effect in patients with large abscesses (>300 mL).^97,119–124

Pyogenic Liver Abscess

Unlike amebic liver abscess, pyogenic liver abscesses usually require drainage in addition to antibiotic therapy. Surgical drainage was traditionally the treatment of choice and, in the preantibiotic era, the only hope for cure. As early as 1953, McFadzean and associates¹²⁵ reported the use of percutaneous drainage with antibiotic therapy to treat 14 patients with liver abscess. After ultrasonography and CT became widely available, multiple studies confirmed this approach and made image-guided drainage procedures the preferred primary therapy, although some still advocate surgical drainage (Fig. 77-1). Surgical intervention should be considered if percutaneous drainage fails or management of concurrent intra-abdominal disease is required, and also for some patients with multiple large or loculated abscesses.^126,127 Percutaneous catheter drainage is successful in 69% to 90% of cases; the procedure is generally well tolerated and usually can be performed at the time of radiographic diagnosis.^{56,57,58,60,128} Aspirated material should be sent for Gram stain and cultured for aerobic and anaerobic bacteria. Meticulous handling of the specimen and rapid transportation to a qualified laboratory are essential for efficient recovery of anaerobes. Histopathologic biopsy specimens should also be obtained if possible. Depending on host and epidemiologic factors, microbiologic evaluation for fungi, mycobacteria, and E. histolytica should be considered. The catheter is usually left in place until drainage becomes minimal, typically 5 to 7 days. Percutaneous aspiration without catheter placement has received recent attention, with several studies reporting success rates between 58% and 88% for solitary abscesses 5 cm or smaller in diameter, similar to those observed with catheter placement.^{57,58,60,129–131} Success rates of 94% to 98% were reported in two studies in which percutaneous aspiration without catheter placement was followed by close clinical monitoring and serial ultrasound studies as indicated. Both investigations found that reaspiration was required in about 50% of cases, and a minority of patients required three or more aspiration procedures.^132,133 Therefore, catheter placement may not be required in some patients, but prospective randomized control trials are necessary to clarify the role of aspiration alone versus catheter placement in the management of hepatic abscess.