LIVER ABSCESS (Part Two)

PYOGENIC LIVER ABSCESS (Cont.)

Treatment

• Pyogenic hepatic abscess is almost always fatal in patients in whom the disease is not diagnosed and treated appropriately.
  
• As a result, traditional treatment recommendations have included surgical drainage and a prolonged course of antibiotics.
  

Antibiotics

• Most patients with pyogenic hepatic abscesses present with fever, and more than half of these patients have positive blood cultures at the time of presentation.
  
• Therefore, early institution of antibiotic therapy is important.
  
• The antibiotic regimen chosen should be based on knowledge of the spectrum of organisms likely to be isolated from pyogenic hepatic abscesses.
  
• The organisms cultured most frequently include
  
• (1) gram-negative aerobes,
  
• (2) streptococcal species, and
  
• (3) anaerobes.
  
• In a series reported by Huang et al., gram-negative aerobes were found in the majority of patients.
• Klebsiella and streptococcal species were found in 50 to 60% of patients.
  
• E. coli, staphylococci, and Pseudomonas species were isolated in 30% of patients.
  
• Anaerobes of Bacteroides and Clostridium species were each cultured in 10% of patients.
  
• In more recent years, fungal species were also isolated in one fourth of the patients with pyogenic abscesses.

Systemic antibiotics should be initiated early, and the regimen chosen should be based on knowledge of the spectrum of organisms likely to be isolated.

• The underlying cause influences the spectrum; anaerobes are more likely with a colonic source, and enterococcal infection is more likely with a biliary source.
  
• As mentioned earlier, streptococci and Pseudomonas species as well as multiply resistant bacteria and fungi have been isolated more often in patients with indwelling biliary stents.
  
• Many of these patients have received broad-spectrum antibiotics for recurrent episodes of cholangitis.

Until specific bacteria have been cultured and sensitivities are known, broad antibacterial coverage is indicated.

• More recently, broad-spectrum penicillins such as imipenem, piperacillin-tazobactam, ticarcillin-clavulanate, and ampicillin-sulbactam have been recommended to provide coverage for the gram-negative organisms, enterococci, and anaerobes frequently found in pyogenic abscesses.
  
• Many of these patients are already at increased risk of developing renal failure because of advanced age, sepsis, jaundice, or pre-existing renal disease.
  
• Of course, other alternatives are required in patients with a significant penicillin allergy.

The exact length of antibiotic therapy should be individualized on the basis of the number of abscesses, the clinical response, and the toxicity of the antibiotic regimen that is required.

• Patients with multiple small miliary abscesses should receive at least 4 to 6 weeks of antibiotic therapy.
  
• However, a shorter antibiotic course may suffice for patients with a small solitary abscess that has been adequately drained.
  
• Initially, antibiotics should be administered parenterally.
  
• Once sepsis has cleared with no more fever and no leukocytosis, appropriate oral agents may be used in patients who require prolonged treatment.
  
• Ciprofloxacin has the advantage in this situation of providing a broad spectrum as well as excellent biliary excretion.
  
• However, oral antibiotics specific for the organisms isolated and those that achieve adequate blood and liver levels should be chosen in preference to an inappropriate antibiotic with excellent bile levels.

Closed Aspiration

• Closed needle aspiration and intracavitary antibiotics have been used successfully in many patients with amebic liver abscesses.
  
• However, until recently, most experts did not advocate closed aspiration of pyogenic hepatic abscesses.
  
• In contradistinction to the generally recommended surgical approach, reports by Civardi and associates and Balasegaram suggested that selected cases of pyogenic hepatic abscess can be managed with closed aspiration and systemic antibiotics.
  
• In the report from the Johns Hopkins Hospital, only 10 of 153 patients (7%) underwent aspiration, and 1 of these 10 patients (10%) died.
  
• These reports confirmed that closed aspiration and antibiotics can be successful in carefully selected patients including those who are young, who are otherwise healthy, and who have no other source of intra-abdominal infection.
  
• These very patients, however, tolerate percutaneous drainage or a laparotomy without difficulty.
  
• Theoretically, drainage in these patients should provide more rapid resolution of the abscess.
  
• Thus, aspiration may be attempted in carefully selected patients with a solitary abscess, but if sepsis persists, prompt drainage should be performed.

Percutaneous Drainage

• Percutaneous drainage of intra-abdominal and liver abscesses was introduced in the late 1970s.
  
• By the mid 1980s, the safety and efficacy of percutaneous abscess drainage in selected patients had been established.
  
• Several more recent reports documented that a significant proportion of patients can be managed with a combination of systemic antibiotics and percutaneous drainage.
  
• In the analysis from the Johns Hopkins Hospital, 45% of patients managed between 1973 and 1993 initially underwent percutaneous drainage.
  
• In earlier reports, percutaneous drainage was reserved for patients with solitary abscesses.
  
• However, some patients with multiloculated abscesses can be managed with aggressive percutaneous techniques that include disruption of loculations and placement of large-bore sump catheters.

In patients with indwelling biliary stents who develop an abscess, treatment options include systemic antibiotics and (1) change of the stent or (2) stent change and percutaneous drainage.

• If the abscess clearly communicates with the biliary tree on cholangiography, stent change may be adequate.
  
• However, if no communication exists, if the tract is small, or if the patient's condition remains septic, abscess drainage is indicated.
  
• Percutaneous and surgical drainage should not be considered competitive but, rather, complementary techniques.
  
• Most solitary and some carefully selected macroscopic multiple abscesses are amenable to percutaneous abscess drainage.
  
• In addition, multiple microscopic abscesses resulting from biliary tract obstruction may be managed by percutaneous biliary drainage.

Extraperitoneal Surgical Drainage

Before the introduction of antibiotics, the "extraserous" approach to drainage of hepatic abscesses was recommended because of fear of contamination of the pleural or peritoneal cavities with the transpleural or transperitoneal approaches.

• Entrance into a liver abscess through an area of adhesion between the liver and the parietal peritoneum was once the safest route because of the extremely high mortality associated with contamination of the pleural or peritoneal cavities.
  
• The exact route of extraperitoneal drainage was dictated by the position of the abscess. Anterior abscesses were approached through a subcostal (Clairmont) incision.
  
• Posteriorly situated liver abscesses in the right lobe were approached using a posterior retroperitoneal incision through the bed of the twelfth rib.

Although most hepatic abscesses can be reached through these incisions, abscesses high on the dome of the right lobe may be far from the anterior or the posterior approach.

• Because a short route for drainage is also an important principle, transpleural approaches to these high-lying lesions have been described.
  
• With this approach, the pleural space can be protected either by dissecting the pleura away from the diaphragm or by incising the pleural and suturing the diaphragm to the incision before opening into the abscess.
  
• These extraperitoneal or transpleural approaches are best suited for patients with a solitary abscess in whom the cause is known and has been treated adequately.
  
• Absolute certainty of these two conditions, however, is a rarity.
  
• The obvious disadvantages of these techniques, therefore, are that
  
• they do not allow for adequate exploration of the entire liver and
  
• they do not permit recognition of an inadequately treated intra-abdominal source of infection.

Transperitoneal Surgical Drainage

In the 1950s and 1960s, the ability to diagnose liver abscesses was limited, and many patients received antibiotics alone or no treatment for fever of unknown origin. Most of these patients had end-stage malignant disease or died before the establishment of a diagnosis. Approximately one half of the patients underwent surgical exploration; most underwent direct abscess drainage, and the remainder had the biliary tree decompressed.

• In more recent years, approximately 90% of patients managed at the Johns Hopkins Hospital underwent some form of abscess decompression .
  
• As in previous years, approximately one third of patients had surgical exploration and underwent abscess drainage.
  
• Although 10% of patients also had biliary tree decompression, most of these procedures were performed nonoperatively.

Transperitoneal drainage provides the advantages that

(1) the entire liver can be exposed,

(2) the best drainage site can be determined,

(3) multiple abscesses can be located with the assistance of intraoperative ultrasound,

(4) the entire abdomen can be explored for the primary source of infection, and

(5) if indicated, operative cholangiography, duct exploration, and drainage can be performed.

Failure to appreciate and treat an underlying source of intraabdominal infection or to drain an obstructed biliary tree adequately certainly contributed to the high mortality in an earlier era.

• However, with the widespread availability of abdominal and pelvic CT, the likelihood of missing an intra-abdominal source of infection is low.
  
• At present, surgical drainage is usually reserved for patients in whom percutaneous drainage has failed, those who require surgical management of an underlying problem, and selected patients with multiple macroscopic abscesses, patients receiving steroids, and patients with ascites.

Before a liver abscess is drained, meticulous exploration of the entire abdomen is mandatory.

• Patients with acute suppurative cholangitis with multiple abscesses of biliary origin also require common duct exploration.
  
• Choledochotomy with insertion of a T-tube is necessary, because cholecystostomy without choledochotomy does not provide adequate biliary drainage in many of these patients.
  
• However, the extent to which the biliary tree should be explored and the decision whether to risk increasing intrabiliary pressure with choledochoscopy or a completion cholangiogram must be individualized on the basis of the biliary disease and the stability of the patient during the surgical exploration.
  
• With the availability of percutaneous transhepatic drainage, some surgeons may consider delaying surgery if adequate percutaneous drainage can be achieved.

Once the abdomen has been explored and the primary focus of infection has been managed appropriately, the liver should be palpated carefully.

• In most cases, preoperative CT or ultrasound has been performed and aids the surgeon greatly in the localization of abscesses.
  
• If the abscess is not obvious during the surgical procedure, however, intraoperative ultrasound may aid in localization.
  
• Needle aspiration should be performed, and pus should be placed directly into anaerobic culture media, which is kept in the operating room, or it should be immediately transported to the laboratory for anaerobic and aerobic cultures.
  
• The aspirated material also should undergo Gram staining and should be examined for amebae.

After carefully protecting the adjacent peritoneal cavity with laparotomy pads, and with adequate suction available, the surgeon should drain the cavity dependently by creating a tract through the liver with his or her finger.

• If the abscess is multiloculated, some attempt should be made to open into the various pockets.
  
• This principle must be tempered, however, by the amount of hemorrhage that may ensue.
  
• A biopsy specimen should be taken from the abscess wall to rule out a necrotic, infected tumor and to search further for amebic trophozoites.
  
• In addition, biopsy of an area of the liver that appears grossly normal should be performed to rule out multiple microscopic abscesses, which may require prolonged systemic antibiotic therapy.
  
• Multiple drains, including Penrose and soft sump or rubber tubes, are placed into the cavity and adjacent perihepatic spaces and are brought out through separate stab incisions.
• The presence of drainage tubes allows for irrigation and sinograms to assess progress in closure of the space.

In patients with multiple microscopic abscesses, no formal liver drainage is required, but a careful search for a biliary or intra-abdominal source should be undertaken.

• Catheterization of the umbilical vein for antibiotic administration also may be considered.
  
• Occasionally, débridement of necrotic hepatic tissue is necessary to control infection.
  
• Infrequently, the only method to control multiple intrahepatic abscesses is formal hepatic resection.
  
• This form of treatment generally is reserved for patients with long-standing biliary obstruction confined to one hepatic lobe, in whom multiple previous attempts to control intrahepatic infection by biliary drainage have been unsuccessful.
  
• Hepatic resection may be the only answer for this selected group of patients.
  
• The largest experience with hepatic resection for pyogenic abscess comes from Balasegaram, who reported 3 deaths among 21 patients.
  
• Life-threatening hemobilia after drainage of an abscess may be another unusual indication for hepatic resection.

Results of Treatment

The mortality rates for patients with pyogenic hepatic abscess reported between 1950 and 2000 varied from as low as 24% to as high as 88%. This large variation can be explained, however, by differences among the various series in the relative percentages of multiple abscesses, abscesses of biliary origin, abscesses secondary to generalized septicemia, and anaerobic abscesses.

Number

As mentioned earlier, the number of abscesses is directly related to the outcome.

• In the 4-decade analysis from the Johns Hopkins Hospital, the overall mortality decreased significantly (P < .05) from 65% in the period from 1952 to 1972 to 31% in the period from 1973 to 1993.
• This change was largely the result of a significant (P < .05) reduction in the mortality observed in multiple abscesses (88% versus 45%).
• However, the mortality also decreased for solitary abscesses (31% versus 19%).
  
• In both periods, the number of abscesses was one of several factors that predicted outcome.

Etiology

Before 1973, mortality was extremely high in patients with malignant or benign biliary obstruction, as well as in patients with a portal vein route of entry of infection.

• Since 1973, mortality decreased in almost all categories.
  
• These decreases reached statistical significance (P < .05) in the malignant and benign biliary categories, with the most dramatic change in the latter subgroup.
• Only the hepatic artery route had an increase in mortality, but the number of patients in the group was small.

Bacteriology

• In the early report from the Johns Hopkins Hospital, mortality in patients with purely aerobic abscesses (78%) was significantly higher (P < .05) than in patients with purely anaerobic abscesses (28%).
• However, in the more recent report, mortality in patients with purely aerobic abscesses had decreased significantly (P < .05) to 32%, and the difference observed earlier between aerobic and anaerobic abscesses was no longer significant.
• However, the emergence of secondary fungal infection was associated with a significantly increased (P < .02) mortality rate.

Treatment Method

• Before 1973, treatment with antibiotics alone or no treatment was uniformly fatal.
  
• After 1973, no patients received antibiotics alone, whereas the universal mortality rate in patients without treatment did not change.
  
• As biliary drainage changed from operative to nonoperative, mortality decreased.
  
• However, because of small numbers of patients, this reduction did not reach statistical significance.
  
• Mortality in patients who had surgical drainage, in association with systemic antibiotics, was 21% before 1973 and 14% afterward.
  
• By comparison, mortality in patients who had percutaneous drainage after 1973 was 26% (P = .19).
  
• However, direct comparison of mortality for surgery and percutaneous drainage should not be made, because some patients with end-stage malignant diseases were managed with percutaneous drainage in preference to further surgery.

Risk Factors

Multiple abscesses, an associated malignant disease, jaundice, hypoalbuminemia, leukocytosis, bacteremia, and a significant complication were associated with increased mortality.

• In more recent years, septic shock and fungal infection also correlated with increased mortality.
  
• An analysis from Taiwan also attempted to define risk factors for a poor outcome in patients with pyogenic hepatic abscesses.
  
• In a univariate analysis of 72 patients, Lee and colleagues found several of the same parameters as in the Johns Hopkins report, such as multiple abscesses, jaundice, hypoalbuminemia, and leukocytosis.