For diagnosis and discussion, see the following page.

Diagnosis: Septic pulmonary emboli.

DISCUSSION

Septic pulmonary embolism (SPE) results when fragments of thrombus containing bacteria or fungi travel to the pulmonary circulation and lodge in segmental and subsegmental pulmonary arteries. These septic thrombi are large enough to occlude the vessel and result in septic infarction. SPE complicates a minority of underlying infectious processes, such as tricuspid endocarditis and septic thrombophlebitis of the extremities. Other less common sources of septic emboli include osteomyelitis (1); deep pelvic infections; infected transvenous pacemakers, catheters, or shunts; and head and neck infections. Lemierre's syndrome (postanginal sepsis), in which periodontal abscesses extend into the parapharyngeal space and cause septic thrombosis of the internal jugular vein and resultant SPE, is well described. Mastoiditis may have a similar course (1). Occasionally, a bland pulmonary infarction may be complicated by secondary bacterial or fungal infection.

Risk factors for SPE include intravenous drug abuse (IVDA), which is the most common underlying cause, as well as congenital heart disease, immunosuppression in the presence of systemic infection, skin infections, and indwelling venous catheters. Although IVDA has long been associated with tricuspid endocarditis, one group studying patients with group A streptococcal sepsis found that only a minority of IVDA patients with documented SPE had echocardiographically evident endocarditis (2). The investigators argue that SPE in the context of IVDA is most commonly acquired as the result of suppurative thrombophlebitis of the upper extremities, often in the absence of endocarditis. HIV infection is also known to be a specific risk factor for SPE, with or without an underlying history of IVDA (3).

Organisms associated with SPE include, most commonly, coagulase-positive Staphylococcus aureus and group A Streptococcus; in primary oropharyngeal infections, Bacteroides and Fusobacterium are common (1). Patients with IVDA-related SPE may harbor polymicrobial emboli, including anaerobic and gram-negative organisms. Blood cultures may be falsely negative, particularly early in the disease process (3, 4).

The clinical presentation of SPE may be indolent or fulminant, depending upon the organism and the underlying disease. IVDA patients commonly experience an indolent course. Some patients have fever, productive cough, and hemoptysis (1). Fever of unknown origin is also a common presentation.

Because infective endocarditis may not be present or detectable, a high level of clinical suspicion--with or without positive blood cultures--has previously been necessary to make the diagnosis of SPE. The classic chest radiograph findings of SPE in the appropriate clinical context have also been extremely useful in confirming the diagnosis. These findings include multiple pulmonary nodules of variable size (5 mm to 2 cm) and variable degrees of cavitation, located predominantly in the periphery and bases (5). Associated wedge-shaped subpleural densities, hilar or mediastinal lymphadenopathy, and ill-defined infiltrates and effusions may also be present. However, one study found that fewer than half of chest radiographs in patients with known SPE actually demonstrated pulmonary nodules. Most radiographs revealed only nonspecific, ill-defined pulmonary infiltrates and blunting of the costophrenic angles, consistent with small pleural effusions (5).

CT significantly increases both sensitivity and specificity in the diagnosis of SPE. In one study, CT demonstrated pulmonary nodules in 83% of patients with known SPE (4). Although pulmonary nodules themselves are nonspecific, the use of CT allows for much better evaluation of cavitation, which is a frequent finding in SPE. Huang et al described cavitation in 67% of their SPE patients (5). Even more specific for SPE is the presence of a nodule with an associated feeding vessel, a finding seen occasionally in metastatic disease, but commonly in SPE--in 67% of cases by Kuhlman et al (4). The presence of both a feeding vessel and cavitation is highly specific for SPE (5).

Further CT findings of SPE that have been described include air bronchograms within the parenchymal nodules in 28% of cases (4), wedge-shaped subpleural densities (Hampton's humps) in 73% of cases (5), mediastinal or hilar lymphadenopathy in 27% of cases (1), and axillary lymphadenopathy in 80% of cases (5). Empyema, a frequent complication of SPE, is not readily seen on plain radiographs. It is secondary to rupture of subpleural lesions into the pleural space and was identified on CT scans in 39% of cases in one study (4).

A final advantage of CT over chest radiography is the option of administering intravenous contrast, which results in peripheral enhancement of these parenchymal nodules as well as easy delineation of lymphadenopathy relative to vascular structures in the hila and mediastinum. Contrast also allows better identification of feeding vessels, which indicate the hematogenous origin of these nodules.

In the appropriate clinical context, CT can add significant sensitivity and specificity in the diagnosis of SPE.

1. Fraser RS, Pare PD. Diagnosis of Diseases of the Chest, vol 3. Philadelphia: WB Saunders Co, 1999.
2. Bernaldo de Quiros JC, Moreno S, Cercenado E, Diaz D, Berenguer J, Miralles P, Catalan P, Bouza E. Group A streptococcal bacteremia. A 10-year prospective study. Medicine (Baltimore) 1997;76:238-248.
3. Kuhlman JE, Fishman EK, Hruban RH, Knowles M, Zerhouni EA, Siegelman SS. Diseases of the chest in AIDS: CT diagnosis. Radiographics 1989;9:827-857.
4. Kuhlman JE, Fishman EK, Teigen C. Pulmonary septic emboli: diagnosis with CT. Radiology 1990;174:211-213.
5. Huang RM, Naidich DP, Lubat E, Schinella R, Garay SM, McCauley DI. Septic pulmonary emboli: CT-radiographic correlation. AJR Am J Roentgenol 1989;153:41-45.