Lung abscess

Table I.

Neoplasm	Primary lung – ~80% of cases SCCA. Small cell lung CA does not cavitate.Lymphoma – multiple lesions, upper lobe predominance.Metastatic CA – esp. head and neck SCCA.
Vasculitis	Wegener’s – single or multiple lesions, cANCA present in >90%.Rheumatoid arthritis -multiple, lower lobes, systemic disease evident, may coexist with malignancy.
Pulmonary infarct	~15% of PEs infarct; cavitation occurs in ~5% of infarcts. Bacterial superinfection rare.
Structural lung abnormalities with air-fluid levels-blebs, cystic bronchiectasis, pulmonary sequestration, bullae	usually distinguished by CT
Sarcoid	multiple in ~90%, median diameter 2 cm

Hematogenous spread of staphylococcal infections from conditions like septic phlebitis or tricuspid valve endocarditis in injection drug users can result in septic embolization to multiple sites, with predilection for the lungs.

Oropharyngeal infection causing septic internal jugular vein thrombophlebitis and Fusobacterium necrophorum bacteremia produces multifocal septic pulmonary infarcts (Lemierre’s syndrome). Progression to frank cavitation is less commonly seen now than in the pre-antibiotic era.

The recognition that chronically-ill patients with extensive health care system exposure develop oral colonization with aerobic gram-negative rods and Staphylococcus aureus is well-documented and reflected in the treatment guidelines for health-care associated pneumonia. Virulent pathogens such as Pseudomonas aeruginosa, Klebsiella pneumoniae and Staphylococcus aureus may cause cavitary pneumonia, and, in severe cases, pulmonary gangrene.

Patients with cell-mediated immune defects are susceptible to infection with fungi and intracellular bacterial pathogens. Granulocytopenic patients are more prone to infection with aerobic gram-negative rods and Staphylococcus aureus.

The likelihood of opportunistic infection increases with the intensity of endogenous or exogenous immunosuppression. Many opportunistic pathogens also cause infection in immunocompetent adults; conversely, conventional pathogens may cause more severe or unusual disease manifestations in patients with altered immunity.

Nocardia asteroides is an aerobic gram-positive actinomycete with a predilection for lung, CNS, and cutaneous involvement. A third of cases of nocardiosis occurs in immunocompromised patients. Approximately 40% of cases of nocardiosis are confined to the lung.

Rhodococcus equi is a gram-positive acid fast coccobacillus that has been associated with the human immunodeficiency virus (HIV) epidemic. It produces isolated pulmonary involvement in ~80% of cases. Cavitary disease is more frequently seen with Rhodococcus than with mycobacterial or nocardia infections in HIV-infected patients.

Invasive pulmonary aspergillosis has been most frequently described in heavily immunosuppressed populations, such as hematopoietic or solid organ transplant recipients. Invasive disease is well-described, however, in chronic obstructive pulmonary disease (COPD) patients on high doses or protracted courses of corticosteroids.

Reactivation tuberculosis leading to cavitation occurs in elderly patients as a consequence of waning immunologic surveillance. Immunosuppressed states such as HIV infection, post-transplantation, tumor necrosis factor (TNF)-alpha therapy, silicosis, dialysis, and head and neck cancer are associated with the highest risk for reactivation disease in patients with latent tuberculous infection (LTBI).

The nontuberculous mycobacteria Mycobacterium avium-intracellulare (MAC) and M. kansasii also cause cavitary lung disease. Illness classically presents with slowly-progressive cavitary disease in older men with underlying COPD. Cavitary disease is less commonly seen in HIV infected patients.

Histoplasma capsulatum is endemic to the St. Lawrence, Ohio, and Mississippi River Valleys. Underlying COPD predisposes to chronic pulmonary histoplasmosis, in which the host inflammatory response mediates a chronic fibrosing process. Resulting upper lobe cavitary disease closely mimics tuberculosis; dual infection or coexisting lung cancer are well-described.

Coccidioides immitis is endemic to the southwest United States. Cavity formation is a late sequelae of primary infection. Patients may develop thin-walled cavities, typically peripheral and solitary, and often asymptomatic. Alternatively, chronic fibrocavitary disease, more frequent in diabetic patients, produces progressive systemic symptoms, infiltrates with thick walled cavitation, and often hilar adenopathy.

Less common fungal pathogens associated with cavitary lung disease include blastomycosis, cryptococcosis, and sporotrichosis. Cavitary disease caused by Pneumocystis jiroveci is well-described in HIV infected patients and typically features thin-walled lesions.

Microaerophilic streptococci (Streptococcus anginosus group, formerly S. milleri) are a subgroup of viridans streptococci with a unique propensity for causing abscess formation. Monomicrobial lung abscesses have been described. Metastatic infection should be considered in bacteremic cases.

Type 3 Streptococcus pneumoniae has been reported as a sole abscess pathogen, but its role is unclear as coexisting anaerobic infection may have been contributory.

Actinomyces israelii is an anaerobic actinomycete most commonly associated with cervicofacial and abdominal infection. Isolated thoracic actinomycosis is uncommon but notoriously mimics lung cancer.

Other bacteria reported as pathogens in lung abscesses include Pasturella multocida, Haemophilus influenzae, Legionella, and Burkholderia cepacia.

Burkholderia pseudomallei, a gram negative intracellular bacterial pathogen, is the cause of melioidosis. This environmental saprophyte, endemic in Southeast Asia and the northern region of Australia, is transmitted by inoculation and presents with diverse clinical manifestations including lung abscess. The incubation period varies widely from days to years.

Klebsiella pneumoniae infection has been reported as a lung and liver abscess pathogen in Taiwan. Published experience is limited and the true prevalence of Klebsiella and its role as a sole pathogen in lung abscess is incompletely defined.

Entamoeba histolytica lung abscesses most commonly result from extension of a liver abscess into the lung and are therefore generally seen in the right lower or middle lobe.

Paragonimus westerman is a lung fluke endemic to Asia. Infection is acquired through ingestion of undercooked fresh water shellfish. Late infection may be associated with peripheral lung cavities, rarely with air-fluid levels, consequent to inflammatory lung changes around adult flukes. Lesions are typically small, peripheral, and found in the lower lung fields.

Echinococcus granulosa, the dog tapeworm, the etiologic agent of cystic echinococcal disease, is widely distributed throughout the world and classically produces a thin-walled cyst on CT that may demonstrate a crescent sign. Coexisting liver involvement is seen in ~20% of patients.

Physical examination may support a diagnosis of anaerobic lung abscess. Associated findings include fever, poor dentition with gingival disease, digital clubbing, and putrid sputum in ~50% of patients. Pulmonary exam abnormalities may include amphoric breath sounds, and signs of consolidation or effusion.

Laboratory testing is directed at establishing a microbiologic diagnosis, particularly in cases not clearly consistent with primary lung abscess.

Blood cultures are routinely sterile in anaerobic lung infection. Routine sputum cultures are of limited utility in primary lung abscess as they are routinely contaminated with aerobic upper airway flora. A putrid odor to sputum is characteristic of anaerobic infection. Sputum cultures should be obtained when health-care associated pathogens are suspected or in post-influenza settings when Staphylococcus aureus superinfection is a concern.

Blood cultures may be positive in Lemierre’s syndrome for Fusobacterium necrophorum and in endovascular infections for Staphylococcus aureus.

Tuberculosis should always be excluded in any atypical case of lung abscess. All patients should be placed in respiratory isolation and sputum for acid fast smear and culture obtained.

Blood testing for galactomannan, a component of the Aspergillus cell wall, should be obtained if invasive aspergillosis is a concern. Testing for beta-glucan (Fungitell), a fungal cell wall component with utility in diagnosing aspergillosis and Pneumocystis, may also be obtained if available. Antibody and antigen testing for histoplasmosis and coccidioidomycosis is available and should be sent in suspect cases; antigen testing for cryptococcus is also available. Laboratory lag times prevent these tests from informing initial decision-making in atypical cases but results may be useful in the event of a desultory clinical response.

Antinuclear cytoplasmic antibodies should be sent if vasculitis is a consideration. As noted, cavitary rheumatoid disease generally occurs in the setting of clinically-apparent disease. Cavitation of a pulmonary infarct is an unusual initial presentation for thromboembolic disease; D-dimer elevations would be expected in systemic inflammatory disease and would not be useful.

Lung abscess is generally a radiographic diagnosis. The typical location for a lung abscess is in a lung segment where gravity drainage carries aspirated material, i.e., the apical segment of the right lower lobe or posterior segment of the right upper lobe. Most aspirations occur in supine or recumbent patients. The right lung is favored because of the larger size and less acute angulation of the right mainstem bronchus. Primary lung abscesses are most often unilateral. Unusual (i.e. anterior) abscess location raises concern for an underlying neoplasm.

If an abscess is detected on chest radiograph, chest computed tomography (CT) provides enhanced assessment for an underlying endobronchial lesion, hilar adenopathy which is not usual in primary lung abscess, and additional underlying parenchymal lung disease or small abscesses. Chest CT also helps differentiate peripheral lung abscess from empyema with an air-fluid level indicative of bronchopleural fistula. In addition, CT chest may detect unsuspected lung abscess when performed in patients with non-resolving pneumonias.

Initial management necessitates determining whether a cavitary lung lesion is likely to represent a primary lung abscess. If the clinical presentation supports this diagnosis, therapy active against oral anaerobes is initiated. Commonly isolated pathogens include peptostreptococci, Prevotella spp, non-fragilis Bacteroides, and Fusobacteria spp. Parenteral clindamycin has supplanted penicillin as the standard of care, as several oral anaerobes (Fusobacterium, Prevotella sp, non-fragilis Bacteroides sp) now produce penicillinase.

Limited clinical data exists for other regimens, but based on in-vitro susceptibility of the usual pathogens, a number of other regimens may be used. These include combination beta-lactamase inhibitors (e.g., ampicillin-sulbactam, amoxicillin-clavulanate, piperacillin-tazobactam), carbapenems, second-generation cephalosporins (e.g., cefoxitin, cefotetan), and quinolones with good anaerobic activity (e.g. moxifloxacin). Concern has been raised that the anti-tuberculous activity of newer quinolones may lead to partial responses and diagnostic delays. As this question is unresolved, it may be preferable to select another drug regimen if tuberculosis is a serious consideration.

Metronidazole as monotherapy is not recommended for primary lung abscess since it is associated with a 50% failure rate due to poor activity against peptostreptococci; however it has been used effectively in combination with penicillin. Tetracyclines have variable anaerobic activity and poor penetrance through the abscess membranes, and are not recommended.

If health care-associated pathogens are likely, coverage for gram-negatives and Staphylococcus aureus should be initiated.

Patients should be treated for 4-6 weeks, ensuring sustained abatement of fever, resolution of putrid sputum, and improvement of radiographic findings.

Initial management should include:

• Ensuring hemodynamic stability and adequate oxygenation;
• Initiation of antibiotics;
• If tuberculosis is a consideration, initiate respiratory isolation and sputum collection;
• If the patient is an alcoholic, thiamine repletion and withdrawal protocol.

Early bronchoscopy should be considered in patients with secondary lung abscess if CT suggests a neoplasm. Bronchoscopy may also be performed to obtain specimen for microbiologic studies if sputum cultures are non-diagnostic, particularly if unusual bacterial, mycobacterial, fungal or other opportunistic pathogens are suspected.

Medical management failure occurs in 10-15% of cases. Early engagement of appropriate consulting services in such cases is prudent. Clinical factors associated with treatment failure include:

• Abscesses >6 centimeters (cm);
• Severe hemoptysis (rare);
• Resistant flora, e. g., Pseudomonas;
• Presence of obstructing neoplasm.

Management of malignant post-obstructive abscesses often requires a multidisciplinary approach including medical oncology, radiation oncology and thoracic surgery consultations. Bronchoscopic lavage and drainage of abscesses is not recommended because of the potential for large volume pus spillage into other lung segments causing acute respiratory distress syndrome (ARDS) or mechanical asphyxiation.

There has been limited published experience with CT-guided percutaneous drainage of large or poorly responding pyogenic abscesses. Endoscopic placement of pigtail catheter for drainage under bronchoscopic visualization has also been reported. Expertise in these potential treatment modalities is institution specific. For patients failing medical therapy in whom less invasive modalities are not available, surgical options should be considered. These includes chest tube drainage, surgical resection via lobectomy (central or large abscesses) or video-assisted thoracoscopy.

If chest tube drainage is performed, it is usually continued for 10-14 days. Potential complications include empyema, bronchopleural fistula, and bleeding.

Indications for surgical resection include hemoptysis, prolonged sepsis, bronchopleural fistula, abscess rupture with empyema formation, failure of medical therapy (i.e., 4-6 weeks of antibiotics), cavity larger than 6 cm, and suspicion for malignancy.

Patient monitoring should include daily assessment of fever trend, oxygenation, and any lung exam abnormalities. Watch for signs of alcohol withdrawal and adverse reactions to antibiotics (e.g., rash, diarrhea, drug fever.)

Laboratory monitoring should include periodic complete blood count (CBC), to assess for resolution of infection-related leucocytosis and anemia. The frequency of monitoring is determined by the clinical evolution of the patient. Initially daily monitoring may be necessary. Once the patient is improving every 2-3 days testing is adequate.

Baseline renal and hepatic function should be assessed in anticipation of protracted antibiotic therapy in the event that symptoms or signs of drug toxicity subsequently appear. Baseline renal function will also inform the need for dosage adjustment of agents excreted by the kidneys.

Initial improvement in patient symptoms and fever should appear in the first 2-3 days of therapy, with defervescence in 7-10 days. If there is no clear improvement by 10-14 days further evaluation is necessary. In addition to the previously outlined unfavorable prognostic features, diagnostic considerations include:

• Drug fever-particularly if the clinical appearance and fever curve are discordant;
• Inaccurate microbiologic diagnosis;
• Non-infectious etiology of cavitary lung disease or superinfection of non-infectious condition.

In such cases, results of serologic testing for unusual pathogens (e.g., endemic mycoses in patients with travel history) should be reviewed and further epidemiologic history should be explored for clues suggesting an unusual infectious exposure. Diagnostic bronchoscopy for biopsy, culture (fungal, mycobacterial, bacterial), and exclusion of radiographically-occult cancer or occluding foreign body should be performed on poor responders.

The optimal frequency of serial chest imaging in patients progressing favorably has not been established. Initial worsening of chest imaging during the first week of therapy is described. Repeat imaging while in-hospital will be generally reserved for patients in whom the clinical response is uncertain or poor.

Weekly CBC should be monitored to assess therapy response and monitor for drug toxicity. Weekly renal function testing is prudent in patients with underlying renal insufficiency. Chest x-rays should be obtained weekly to biweekly while the patient is under treatment. As noted there are no strict guidelines regarding the total duration of therapy. Many experts treat to radiographic resolution or stable small residual scarring. Abscess cavities and surrounding infiltrates often require weeks to months to resolve, typically lagging behind clinical improvement.

Patients receiving parenteral therapy for the duration of treatment will require peripherally inserted central catheter (PICC) placement. Skilled nursing facility placement may be necessary if there are logistic issues related to the complexity of the treatment regimen or if the patient requires physical and occupational therapy for debility.

Patients with primary lung abscesses responding well to initial therapy may be transitioned to oral therapy with clindamycin, amoxicillin-clavulanic acid, or moxifloxacin to complete therapy. Therapy regimens for hospital associated and opportunistic pathogens require individualization.

Primary lung abscesses have a favorable prognosis; approximately 90% of treated patients are cured. Underlying altered immunity, neoplastic bronchial obstruction, and hospital associated infection confer a much worse prognosis with fatal outcomes in 2/3 or more of patients reported.

Kuhajda, I, Zarogoulidis, K, Tsirgogianni, K, Tsavlis, D, Kioumis, C, Tsakiridis, K, Mpakas, A, Zarogoulidis, P, Zissimopoulos, A, Baloukas, D, Kuhajda, D. “Lung abscess-etiology, diagnostic and treatment options”. Ann Transl Med. vol. 3. 2015. pp. 183

Wali, SO. “An update on the drainage of pyogenic lung abscesses”. Ann Thorac Med. vol. 7. 2012. pp. 3-7.

Mandell, LA, Wunderink, RG, Anzueto, A. “Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults”. Clin Infect Dis. vol. 44. 2007. pp. S27-72.