80 INFECTIVE ENDOCARDITIS
Harrison’s Manual of Medicine
When classified according to temporal evolution, acute endocarditis is a hectically febrile illness associated with damage to cardiac structures, seeding of extracardiac sites, and progression to death within weeks if left untreated. Subacute endocarditis is an indolent illness associated with slow or no destruction of cardiac structures and rare metastatic infection; it is gradually progressive unless complicated by an embolic event or a ruptured mycotic aneurysm. Other bases for the classification of endocarditis include site of infection, microbiologic cause, or predisposing risk factors (such as injection drug use).
A small number of bacterial species cause the majority of endocarditis cases (see Table 126-1, HPIM-15, p. 810). The causative microorganisms vary somewhat among the major clinical types of endocarditis, depending in part on which portal of entry into the bloodstream is used and where the infection is acquired (i.e., nosocomially or in the community). In native valve endocarditis, the viridans streptococci, staphylococci, and HACEK organisms (Haemophilus spp., Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella corrodens, Kingella kingae) enter through the skin and upper respiratory tract, Streptococcus bovis enters via the GI tract, enterococci tend to enter through the genitourinary tract, and nosocomially acquired Staphylococcus aureus usually enters via an infected IV catheter. Prosthetic valve endocarditis (PVE) arising within 2 months of valve replacement is due to a number of organisms associated with postoperative bacteremia. Coagulase-negative staphylococcal PVE arising within 12 months of valve replacement is usually nosocomially acquired. The microbiologic etiology of PVE arising >12 months after valve replacement is similar to that of native valve endocarditis. Transvenous pacemaker or implanted defibrillator-associated endocarditis is usually nosocomial and due to S. aureus or coagulase-negative staphylococci. Injection drug users (IDUs) with endocarditis usually have S. aureus (often methicillin-resistant) isolated, but a number of organisms can cause left-sided endocarditis in IDUs. Polymicrobial endocarditis occurs more often in IDUs than in pts who do not inject drugs. In the 5–15% of pts with “culture-negative” endocarditis, the negative cultures are attributable to prior antibiotic exposure in up to one-half of cases. The remainder of “culture-negative” cases are due to fastidious organisms, including the HACEK group, pyridoxal-requiring streptococci (Abiotrophia spp.), Bartonella spp., Coxiella burnetii, and Brucella spp.
The clinical syndrome of endocarditis is highly variable and spans a continuum between acute and subacute presentations. b-Hemolytic streptococci, S. aureus, pneumococci, Staphylococcus lugdunensis, and enterococci typically cause acute endocarditis. Viridans streptococci, coagulase-negative staphylococci, agents of “culture-negative” endocarditis, and occasionally enterococci and S. aureus typically cause subacute endocarditis.
The clinical features of endocarditis are nonspecific (Table 80-1), but these symptoms in a febrile pt with predisposing factors, bloodstream infection due to organisms that frequently cause endocarditis, unexplained arterial emboli, or progressive cardiac valvular incompetence should heighten the suspicion of endocarditis. Fever may be higher in acute than in subacute endocarditis and may be blunted or absent entirely in pts with cardiac or renal failure, in the severely debilitated, and in the elderly. Cardiac murmurs are ultimately detected in 85% of pts with acute endocarditis of a native valve. Congestive heart failure develops in 30–40% of pts as a result of valvular dysfunction, myocarditis, or intracardiac fistula formation. Extension of infection into the valve ring structure results in perivalvular abscesses, which in turn may extend to cause fistulae, pericarditis, or heart block. The classic peripheral manifestations are related to the duration of infection and, because of early diagnosis and treatment, have become less common. Back pain and myalgias are common presenting features that remit promptly with treatment. Renal dysfunction usually results from immune complex deposition and glomerulonephritis; embolic renal infarcts rarely cause renal insufficiency. Almost 50% of IDUs with endocarditis have infection limited to the tricuspid valve; these pts frequently present with pulmonary symptoms, such as cough and chest pain, and have nodular pulmonary infiltrates detectable on CXR.
Table 80-1 Clinical and Laboratory Features of Infective Endocarditis
Arterial emboli may be present in up to 50% of pts; vegetations that are >10 mm in diameter on echocardiogram and mitral vegetations are most likely to embolize. Emboli can travel to almost any organ or site. The risk of embolization is highest in the first week of antibiotic therapy, decreasing from 13 events/100,000 pt-days during the first week to 1.2 events/100,000 pt-days after the third week. Neurologic symptoms (most often due to embolic phenomena) occur in up to 40% of endocarditis pts; purulent meningitis, brain abscesses, ruptured mycotic aneurysm, seizures, and encephalopathy are also observed in some pts.
The Duke Criteria constitute a highly sensitive and specific diagnostic schema developed on the basis of clinical, laboratory, and echocardiographic findings (Table 80-2). Documentation of two major criteria, of one minor criterion and three major criteria, or of five minor criteria allows a clinical diagnosis of definite endocarditis. Three blood cultures separated by at least 1 h but not by >24 h should be obtained. If the initial blood cultures are negative after 48–72 h, two or three additional sets, including a lysis-centrifugation culture, should be obtained and the laboratory alerted to pursue fastidious organisms by prolonging incubation time and performing special subcultures. Serology for Brucella, Bartonella, Legionella, and C. burnetii can be useful in identifying fastidious organisms.
Table 80-2 The Duke Criteria for the Clinical Diagnosis of Infective Endocarditis
Echocardiography should be strongly considered for pts with a clinical diagnosis of endocarditis and for pts in whom endocarditis is strongly suspected. Transesophageal echocardiography (TEE) is >90% sensitive for detection of vegetations, whereas transthoracic echocardiography (TTE) is only 65% sensitive. TTE is not adequate for evaluating prosthetic valves or intracardiac complications. In pts with a low (<5%) pretest probability of having endocarditis, a negative TTE is sufficient to exclude endocarditis. In pts with a 5–50% pretest probability of having endocarditis, initial evaluation with TEE is cost-effective.
Empirical antibiotics should be reserved for pts with deteriorating hemodynamics. The regimens recommended for treatment of PVE (except for staphylococcal infections) are similar to those used to treat native valve infection (Table 80-3). Effective treatment for some intracardiac and CNS complications of endocarditis requires valve replacement, although most of the clinical indications for surgical treatment of endocarditis are not absolute (Table 80-4). Surgery should not be delayed when valvular dysfunction, progressive congestive heart failure, or uncontrolled or perivalvular infection is present. Surgery should be delayed for 2–3 weeks after a nonhemorrhagic embolic stroke and for 4 weeks after a hemorrhagic embolic stroke. Splenic abscesses develop 3–5% of the time and should be drained or eliminated by splenectomy. Mycotic aneurysms develop in 2–15% of pts; half of these cases involve the cerebral arteries. Although some will resolve, these aneurysms should be watched closely for signs of enlargement or leakage indicating the need for excision.
Table 80-3 Antibiotic Treatment for Infective Endocarditis Caused by Common Organismsa
Table 80-4 Indications for Cardiac Surgical Intervention in Patients with Endocarditis
The benefits of endocarditis prophylaxis are not established and in fact may be modest: only 50% of pts with native valve endocarditis know they have a valve lesion predisposing to endocarditis, most cases do not follow an identified procedure during which inoculation occurs, and 35% of cases are caused by organisms not targeted by prophylaxis. Nevertheless, the American Heart Association has identified procedures that may precipitate high-risk bacteremia (Table 80-5), cardiac lesions for which prophylaxis is advised (Table 80-6), and recommended regimens for prophylaxis (Table 80-7).
Table 80-5 Procedures for which Endocarditis Prophylaxis Is Advised in Patients at High or Moderate Risk for Endocarditisa
Table 80-6 Cardiac Lesions for which Endocarditis Prophylaxis Is Advised
Table 80-7 Antibiotic Regimens for Prophylaxis of Endocarditis in Adults at Moderate or High Riska
For a more detailed discussion, see Karchmer AW: Infective Endocarditis, Chap. 126, p. 809, in HPIM-15.