FILLING DEFECTS OF THE LIVER
Pyogenic Hepatic Abscess
Echinococcal Cyst (Hydatid Disease)
Neoplasm and Other Causes of Hepatic Filling Defects
Filling defects of the liver are defined in this chapter as lesions of sufficient size to be demonstrated by imaging studies, such as standard roentgenograms, computed tomography (CT), radioisotope scanning, or hepatic ultrasonography. Generally, lesions that are 0.5 to 1.0 cm or larger can be documented by newer CT scans with intravenous contrast. Differential diagnosis includes infection, hematoma, neoplasm, and cysts (Table 14-1). Granulomatous processes such as tuberculosis and fungal infection, although capable of involving hepatic parenchyma, are unusual causes of filling defects as defined above.
Table 14-1. Amebic versus pyogenic hepatic abscess
Radiographic assessment of the liver is performed for reasons that include (a) hepatomegaly or right upper quadrant (RUQ) tenderness on physical examination, (b) elevation of hepatic enzymes, most notably alkaline phosphatase, (c) jaundice, (d) fever of undetermined origin or other constitutional syndromes, and (e) assessment of metastases in patients with selected known primary malignancies. Clinical presentation is highly variable, depending on the underlying condition. As an example, uncomplicated hepatic cysts may remain asymptomatic and be documented as an incidental finding when roentgenographic studies are performed for other reasons. Alternatively, pyogenic or amebic abscess may present with high-grade constitutional abnormalities including shock. Many patients have subtle complaints of RUQ discomfort and low-grade constitutional abnormalities. Severity and acuteness of presentation may provide useful information about the cause of the process.
Pyogenic Hepatic Abscess
Hepatic abscess has an incidence of at least 0.016%, is primarily a disease of the elderly, and is felt to be underreported. Recent investigations comparing this disease during two 20-year periods suggest that the incidence is increasing from 13 to 20 cases per 100,000 hospital admissions. Contributory factors include (numbers in parentheses represent percentage of macroscopic abscesses) the following: biliary tract infections (33%), direct extension from contiguous infected foci (25%), bacteremia (10%), blunt trauma to the RUQ (e.g., steering wheel injuries) (15%), and pylephlebitis (e.g., complications of perforated appendix) (6%). Approximately 10% are cryptogenic; this group is becoming more common. Patients with pyogenic hepatic abscess diagnosed within the past 20 years are more likely to have underlying malignancy (typically of the biliary tract or pancreas).
Clinical presentation varies with the predisposing factors. Most cases present subacutely with low-grade constitutional complaints and RUQ discomfort. Fever is the most commonly noted abnormality and is seen in more than 90% of cases. Hepatomegaly is documented in only 50% of patients, and jaundice occurs in approximately 25% of patients. Up to 20% of patients present with pulmonary complaints that include right-sided pleural effusions, elevation of the right hemidiaphragm, or rales on physical examination. Pneumonia may be the initial consideration. Routine laboratory data are often noncontributory. Anemia and leukocytosis are seen in 60% and 70% of cases, respectively. Elevations of the alkaline phosphatase are noted in 75% of patients; hyperbilirubinemia is seen in only 20% to 25%.
Noninvasive imaging techniques are generally employed to assess the patient further. Both RUQ ultrasonography and CT (often with IV contrast) have excellent published results and should be employed preferentially to radioisotope scanning. Neither of the preferred modalities is associated with significant risk for false-negative results. Ultrasound has a reported sensitivity for the diagnosis of pyogenic hepatic abscess of 85% to 95% and has the advantage of providing excellent visualization of the biliary tract (diseases of which may contribute to liver abscess). Additionally, ultrasound may be better at distinguishing solid structures from cysts. CT is more than 95% sensitive for the diagnosis of hepatic abscess and has advantages that include the ability to evaluate selected portions of the liver and discriminate among other abdominal structures, and usefulness in obese or distended patients. Additionally, patients with multiple pyogenic abscesses may be better diagnosed by CT than by ultrasound. In the absence of pylephlebitis or biliary tract disease, random assessment of the gastrointestinal organs is generally not indicated.
Percutaneous drainage of the filling defect at the time of initial evaluation allows simultaneous diagnosis and therapy. Materials obtained can be sent for Gram’s stain, aerobic or anaerobic culture, parasitology, and histopathology. When a filling defect is clinically suspected, percutaneous drainage often can obviate the need for surgical exploration. CT or ultrasonography can then be used to follow clinical response to therapy and ensure resolution of the process.
The bacteriology of pyogenic hepatic abscess is often complex and includes combinations of “gut flora.” The role of anaerobes, including Bacteroides fragilis, is well defined. These often coexist with enteric gram-negative bacilli, such as Escherichia coli or Klebsiella species. Trends during the past two decades suggest an increase in hepatic abscesses associated with fungi, streptococci, and Pseudomonas aeruginosa. In adults, hepatic abscesses associated with Actinomycetes as a single pathogen have recently been reviewed. Most cases presented in typical fashion and had unknown primary sources. Recent reports have also documented the likelihood of metastatic dissemination of Klebsiella infection (often causing endophthalmitis) from pyogenic abscess. This has been especially noted in diabetics.
Therapy of pyogenic hepatic abscess consists of long-term parenteral antimicrobials plus drainage. When drainage cannot be performed for technical or medical reasons, certain cases may be managed medically with careful follow-up; however, this should not be considered the standard of care.
Percutaneous catheter drainage is recommended as the drainage procedure of choice. Percutaneous aspiration is generally considered inferior, but comparative studies are needed. Open surgical procedures should be performed when (a) multiple or septated lesions are identified that cannot be drained percutaneously, (b) surgery is indicated to manage the initiating cause of the abscess (e.g., biliary stones), and (c) the patient has failed to respond to percutaneous drainage. Surgical extirpation should also be considered when hydatid disease is in the differential diagnosis. The optimal length of antimicrobial treatment is unknown. Most authorities recommend at least 4 weeks of therapy when drainage has been accomplished, and up to 8 weeks when drainage could not be performed. Agents that include quinolones, metronidazole, and trimethoprim/sulfamethoxazole have excellent bioavailability following oral administration and may be considered in selected cases.
Amebiasis afflicts up to 10% of the population, is associated with 50 million cases of invasive disease, and may account for as many as 100,000 deaths. Although two species of Entamoeba have recently been identified that are morphologically identical, it is only E. histolytica that causes invasive disease. Amebic abscess is the most common complication of intestinal amebiasis and can occur many years after exposure in endemic areas. It should be suspected as a cause of intrahepatic filling defects in patients with such an epidemiologic history, and it generally involves the right hepatic lobe. Although many features overlap with those of pyogenic abscess, selected clues summarized in Table 14-1 may occasionally help with differentiation. Geographic areas of risk are Mexico, Central America, and Southeast Asia. Disease is also associated with homosexuality and residence in institutions for the mentally retarded.
Up to 66% of cases present with the acute onset of fever (often with rigors) and abdominal pain with a duration of fewer than 10 days. The others often have complaints for up to 12 weeks. Persons with acute presentations often have high-grade fevers (above 102°F), and abdominal pain is often located in the RUQ. Selected cases may present only as an acute febrile syndrome. Patients with subacute or chronic presentations have lower fevers and more benign presentations. In both groups, pulmonary complaints such as pleurisy and cough may be noted, and the initial consideration may be pneumonia. Only about one third of patients give a history of active diarrhea at any time before clinical presentation. Rarely, initial presentation will be that of rupture—either into the peritoneum, pleural space, or pericardium. Intraabdominal rupture is more likely when abscesses of the left lobe are present.
For patients with acute presentations, laboratory data typically demonstrate leukocytosis with bandemia. WBC counts may be normal in those with more chronic illness. Both groups may become anemic. Approximately 85% of patients have elevations of alkaline phosphatase. This is the most reliable chemical test, and levels are likely to be higher in persons with chronic disease. Hyperbilirubinemia is not commonly noted. Indirect hemagglutination testing is a reliable serologic test for hepatic amebiasis, yielding positive results in at least 85% of patients with extraintestinal disease, and titers often exceed 1:256. It should be performed in all patients with space-occupying lesions when a specific alternative diagnosis cannot be made. Results of serologic testing may be negative early in the course of disease, and a negative result does not preclude this illness in acute cases. However, elevations are almost always noted by 2 weeks into disease and may remain high for many years following successful therapy.
The diagnosis can be suspected by hepatic imaging, but a pathognomonic picture is not seen. Hepatic ultrasound or CT provides highly sensitive results and can be repeated during and after therapy to gauge response. Most authors now feel that the former, because of ease of administration and lower expense, is the study of choice for both diagnosis and follow-up. Satisfactory treatment of amebic abscess can be anticipated to result in return to an echographically normal liver. However, lesion size may actually increase during early therapy, and thus some authorities do not recommend routine radiographic studies if the patient is clinically responding. The time to normalization is 2 to 20 months. Generally, there is no value to gallium or sulfur colloid scans.
Up to 50% of cases demonstrate multiple space-occupying lesions. The most common sites remain the superior or posterior-superior portion of the right lobe. Results of hepatic ultrasonography usually demonstrate round or oval hypoechoic lesions, often containing debris. Fifty percent may appear as cystic lesions, and many that present acutely may appear as either solid or heterogeneous masses. Thus, the ultrasonic presentation, especially in acute cases, is variable.
The average size of abscesses is 7 to 10 cm. Lesions larger than 10 cm can be seen in at least 33% of cases. In acute cases with clinical presentations of fewer than 5 days, however, initial ultrasonographic findings may be falsely negative. The diagnosis should continue to be suspected by the presentation of fever, RUQ discomfort, and elevated alkaline phosphatase in a person with an appropriate epidemiologic history.
The role of percutaneous lesion aspiration for either diagnosis or treatment remains controversial. Surgical intervention is rarely indicated for the uncomplicated case but may be lifesaving if rupture has occurred. Although some recent data suggest that patients with amebic abscess respond more rapidly when aspiration accompanies medication, many patients in whom the diagnosis is secure can be managed without it. Aspiration is indicated if (a) alternative diagnoses are being considered (excluding hydatid disease), (b) lesions are larger than 10 cm, (c) patients fail to respond to standard therapy within approximately 72 hours, or (c) left-sided lesions are at risk for rupture. Material obtained at aspiration should be evaluated for the presence of organisms. Typical fluid does not smell of food and is brownish (“anchovy paste”), often devoid of organisms, and typically without polymorphonuclear leukocytes. Standard cultures (aerobic and anaerobic) and histopathology should be carried out in all cases.
Medical therapy for amebic abscess consists of oral or IV administration of 750 mg of metronidazole three times daily for 5 to 10 days. A regimen of 2.4 g daily by mouth for several days has also been successfully employed. Up to 90% of patients respond clinically within 72 hours. Aspiration of the hepatic lesion should be performed if clinical response is not seen by this time. More toxic regimens that are uncommonly indicated consist of 1 to 1.5 mg of dehydroemetine per kilogram of body weight daily for 5 days plus either diloxanide furoate or paromomycin.
Echinococcal Cyst (Hydatid Disease)
Echinococcal disease is endemic in areas of southern Europe, the Middle East, and Australia. It represents the larval stage of the canine tapeworm Echinococcus granulosus. Cases reported in the United States are primarily imported. Human disease is generally asymptomatic and identified during routine roentgenography for other reasons. Sixty percent to 75% of all cysts occur in the liver. Between 20% and 40% of persons may harbor multiple cysts.
Clinical presentation is often chronic and generally consists of an enlarging space-occupying lesion in the absence of significant constitutional symptomatology. Symptoms are typically vague abdominal pain and pressure. At least one third of patients remain asymptomatic. Complaints are generally not noted until cysts reach 7 cm or more in size; average cyst size at surgery is approximately 11 cm. Hepatic enlargement may be noted, often in association with a palpable mass. Leakage and frank rupture are dreaded complications and can occur spontaneously, after trauma, or at the time of surgical removal. Material is “allergenic,” and leakage may be associated with anaphylaxis.
Diagnosis should be suspected in patients with a history of travel to endemic areas. Usual laboratory data are not useful. About 60% of patients will have an elevation of alkaline phosphatase, and 33% may demonstrate eosinophilia. An indirect hemagglutination titer of more than 1:64 is seen in about 50% of cases. CT, ultrasonography, or routine abdominal films often demonstrate single or multiple cystic areas, generally with partial calcification. The presence of intracystic septations associated with partial calcification of the cyst is considered pathognomonic.
Therapy consists of surgical extirpation under carefully controlled conditions to avoid spillage of cyst contents. Percutaneous aspiration is contraindicated because of risk for spillage. Medications are not useful. For the management of cysts when complete excision is not feasible, authorities recommend instillation of cetrimide plus removal of the inner embryonic membranes of the cyst.
Neoplasm and Other Causes of Hepatic Filling Defects
Benign and malignant neoplasms (both primary and metastatic) often involve the liver. The most common primary neoplasms are hepatocellular carcinoma and cholangiocarcinoma. Primary cancers commonly associated with hepatic metastases include those of the lung, breast, colon, rectum, stomach, and pancreas, and melanoma. One investigation demonstrated that 39% of adults dying with solid tumors had hepatic metastases.
Clinical presentation varies and may overlap that seen with infectious etiologies. Ascites, however, may be seen and is distinctly unusual for infectious causes of hepatic filling defects. Alternatively, patients with pyogenic abscess are more likely to have at least three of the following: leukocytosis, fever, risk factors for pyogenic abscess, shorter clinical course, and normal hepatic size.
CT is considered the best screening test for hepatic metastases. Sensitivity is above 90%, but specificity is low. Lesions larger than 1 cm can be described in this fashion. Hepatic ultrasonography is generally considered the next most important test and (like CT) can distinguish solid from cystic lesions. Up to 90% of these can be discovered with this technique. CT, often with the use of IV contrast, can detect lesions larger than 0.5 cm. Lesions that include cavernous hemangioma, focal fatty deposits, and nodular hyperplasia can often be diagnosed by a characteristic CT scan plus radionucleotide imaging. Both ultrasonography and CT often can discriminate simple cysts and other lesions. Although one or another study can provide clues to etiology, the ultimate diagnosis generally rests on tissue and fluid sampling.
Hepatic filling defects are associated with many causes. The clinical presentation is variable, and diagnosis is usually identified by noninvasive radiographic assessment plus sampling of the lesion. Percutaneous evaluation of lesion contents is generally acceptable and well tolerated. However, a consideration for hydatid disease should contraindicate this approach. (R.B.B.)
Barnes PF, et al. A comparison of amebic and pyogenic abscess of the liver. Medicine (Baltimore) 1987;66:472–483.
The authors compare the clinical characteristics of 96 patients with amebic abscess with those of 48 patients with pyogenic abscess. Features that included Latin American birthplace, younger age, abdominal pain with RUQ tenderness, symptoms for fewer than 14 days, and selected hepatic enzyme determinations helped predict amebic abscess. Ultrasonographic features of round or oval shape and “hypoechoic appearance with fine, homogeneous, low-level echoes at high gain” were also statistically more common with amebic abscess.
Cady B. Natural history of primary and secondary tumors of the liver. Semin Oncol 1983;10:127–134.
This review provides an excellent overview of the likely causes of hepatic neoplasms and their clinical presentation.
Halvorsen RA Jr, Thompson WM. Imaging primary and metastatic cancer of the liver. Semin Oncol 1991;18:111–122.
The authors provide an approach to the imaging of hepatic neoplasms. Lesions are placed in several groups according to the presence of characteristic changes by imaging modalities. Unfortunately, most are insufficiently differentiated by imaging alone and require tissue diagnosis.
Huang CJ, et al. Pyogenic hepatic abscess. Changing trends over 42 years. Ann Surg 1996; 223:600–607.
The authors assess 233 cases of pyogenic hepatic abscess treated during a 40-year period. They divide patients into two periods: before and after 1972. Differences in bacteriology and treatment were noted. Those cases seen more recently were more likely to be associated with biliary or pancreatic malignancy. Overall mortality decreased from 65% to approximately 30%.
Klotz SA, Penn RL. Clinical differentiation of abscess from neoplasm in newly diagnosed space-occupying lesions of the liver. South Med J 1987;80:1537–1541.
The authors recognized the difficulty in differentiating malignant from pyogenic intrahepatic lesions by imaging techniques alone and looked critically at clinical presentations. Patients with pyogenic abscesses had shorter prodromes, risk factors for abscess, fever, leukocytosis, and normal hepatic size. Presence of at least three of these correctly predicted all abscesses.
Miyamoto MI, Fang FC. Pyogenic liver abscess involving Actinomyces: case report and review. Clin Infect Dis 1993;16:303–309.
The authors report a case of hepatic abscess associated with Actinomyces and review the literature on 35 other cases. Most presented in a subacute fashion and were diagnosed by standard, noninvasive tests. Nonsurgical drainage appears appropriate, and antimicrobial therapy with penicillin or tetracycline resulted in satisfactory outcomes.
Ravdin JI. Amebiasis. Clin Infect Dis 1995;20:1453–1466.
The author provides an excellent overview of recent developments in amebiasis. Biology, epidemiology, diagnosis, management, and prevention are explored. Much practical information, obviously provided by a hands-on clinician, is included in this article.
Rubin RH, Swartz MN, Malt R. Hepatic abscess: changes in clinical, bacteriologic, and therapeutic aspects. Am J Med 1974;57:601–610.
This classic review of the topic summarizes the Massachusetts General Hospital experience with 50 patients during a 12-year period. Most pyogenic abscesses were macroscopic and associated with biliary tract abnormalities. Combinations of enteric flora were most commonly noted. Mortality was greatest in the elderly. Successful treatment required a combination of prolonged antimicrobials and drainage.
Schaefer JW, Khan Y. Echinococcosis (hydatid disease): lessons from experience with 59 patients. Rev Infect Dis 1991;13:243–247.
This article reports the experience from Saudi Arabia with almost 60 patients who had hydatid disease. Currently, diagnosis can be made by ultrasound or CT presentation. Serologic diagnosis is specific but insensitive. Surgery is the management strategy of choice and should be reserved for symptomatic cases. Outcomes were uniformly satisfactory when surgery was performed electively, but morbidity and mortality were encountered when rupture had occurred.
Seeto RK, Rockey DC. Pyogenic liver abscess. Changes in etiology, management, and outcome. Medicine 1996;75:99–113.
The authors provide an excellent review of the subject, summarizing information about epidemiology, bacteriology, and management. The central role of CT and ultrasound in both the diagnosis and management of pyogenic hepatic abscess is stressed. The authors feel that drainage rather than aspiration of abscess is the therapy of choice. This therapeutic regimen has cure rates of up to 75%—and up to 90% within the past 5 years.
Thompson JE Jr, Forlenza S, Verma R. Amebic liver abscess: a therapeutic approach. Rev Infect Dis 1985;7:171–179.
Metronidazole remains the therapy of choice for most amebic abscesses. Failure is occasionally noted and can be anticipated when clinical response is not noted within 72 hours.