Pinning down infective endocarditis cases
The diagnositic criteria are diverse, and many cases are far from obvious. An infectious disease specialist explicates the workup and initial therapy.
Despite all the recent technologic advances in cardiology, there continue to be 10,000-15,000 cases of infective endocarditis (IE) annually in the United States. Originally, the infection was diagnosed by pathologic findings on the infected heart valves, but this criterion excluded many cases that could have been treated before the disease had advanced to that point. Better diagnostic techniques and greater awareness of the subtle signs and symptoms have changed the face of IE.
A changing at-risk population While the incidence of IE has been stable over the past 40 years, the risk of contracting it is changing. Classically, individuals with predisposing heart lesions, such as those of rheumatic heart disease, were thought to be at highest risk and to make up the greatest number of IE patients.
Today, the disease strikes mostly those over the age of 60, three quarters of whom have a predisposing heart lesion, which is usually clinically silent until IE occurs. In addition, permanent vascular access, e.g., for chronic hemodialysis or administration of chemotherapy, places patients at increased risk. Other predisposing heart conditions, such as prosthetic heart valves and previous endocarditis, continue to be important but in a smaller number of overall cases. IV drug abusers comprise the bulk of younger IE patients.
Update on the classic symptoms
The classic IE stigmata, namely recurrent bacteremia, fevers, immunologic phenomena, peripheral emboli, and active valvulitis manifested as changing murmurs and cardiac dysfunction, as a whole are specific for IE. Fever and a heart murmur are found in up to 80% of patients presenting with IE.
However, many cases are not so obvious. In order to understand the diverse clinical criteria, it is useful to understand the pathophysiology of IE. The heart valves are an immunologically privileged site and do not have their own dedicated microvasculature.
Consequently, an infective organism that enters this milieu enjoys the benefits of a nutritious environment, leading to the formation of valvular vegetations, without too much threat from the host immune response. Nonetheless, the immune system becomes alerted to the infection by the constant, low-level bacteremia that results from the regular seeding of the circulation by actively replicating organisms.
The host response is systemic, with fevers and peripheral signs of immune activation. The resulting immune complexes are deposited wherever there is a distal bed of microcirculation—the kidney (glomerulonephritis with active urinary sediment), retina (Roth’s spots visualized on funduscopic exam), or skin (painful distal subcutaneous nodules known as Osler’s nodes), or in the circulation (positive rheumatoid factor [RF]). Splinter hemorrhages may also be seen in the nail beds, but these findings are nonspecific and seen in various conditions.
There will also be peripheral clues from the active replication of infective organisms that have traveled to distal points. Septic emboli are seen on lung CT as well as elsewhere in the peripheral circulation. These emboli can cause tissue ischemia as well as infarction in the brain, spleen, kidney, bone, liver, and other sites. Additional vascular destruction can occur in the form of mycotic aneurysms.
These are not sterile, but rather contain actively replicating organisms that can be cultured using routine microbiologic methods (although this is not usually done). Some patients present with endocarditis only after peripheral emboli or mycotic aneurysms have developed, so any deep pyogenic abscess or infected vessel of unclear etiology demands consideration of IE.
Diagnosis is complicated by a need to avoid overdiagnosis and long courses of IV antibiotics. The modified Duke criteria (Table 1) have evolved to capture the greatest number of IE cases without being overly sensitive. The criteria categorize culture results, physical findings, and historical information into major and minor criteria, which must be present in sufficient numbers and specific groupings to confirm the diagnosis.
History should evaluate for any potential risk factors as well as other signs of systemic illness (such as night sweats or weight loss). Cardiac examination is essential, and a thorough physical examination should look for clues that point to minor modified Duke criteria.
Laboratory studies may be diagnostically helpful. Changes consistent with a chronic inflammatory process, although nonspecific, provide additional supportive evidence. Anemia of chronic disease, leukocytosis, elevated inflammatory markers, active urinary sediment, and elevated RF are possible findings. Blood cultures are of central importance, and three separate sets (two cultures taken more than 12 hours apart) should be obtained to maximize the chances of diagnosis.
Doppler echocardiography now plays a central role. Noninvasive transthoracic imaging can often provide direct visualization of valvular vegetation and is a good first step in suspected IE. The more invasive transesophageal echocardiography is preferred when imaging quality is diminished by physical habitus or prosthetic valves, for example. When considering such imaging, you may also want to consult a cardiologist and an infectious disease specialist.
Use of the modified Duke criteria can help clinicians make a diagnosis of IE (Table 2). If IE is suspected, most authorities agree that hospital admission is indicated, as the inflammation and tissue destruction of IE have been known to progress rapidly, with development of significant cardiac sequelae. Any sign of regurgitation murmur is of obvious concern, suggesting valve destruction and the consideration for valve replacement surgery.
Direct myocardial tissue destruction can develop into a local abscess. Ventricular dysfunction may result, with symptoms of frank heart failure. Also possible are conduction defects due to local destruction of conduction fibers, with development of arrhythmias as well as atrioventricular heart block.
Electrocardiograms and telemetry monitoring are necessary to follow for such complications. Although myocardial ischemia may occur from local damage, the clinical utility of cardiac enzymes has not been thoroughly studied.
Identifying the causative pathogen
Many organisms can cause IE. Pyogenic organisms, such as Staphylococcus aureus, are leading causes. Gram-positive organisms are responsible for the bulk of cases. Clinical confusion can occur when coagulase-negative staphylococci are present because, while they are a frequent contaminant, they have also been significant players in recent large studies. Gram-negative bacteria and other atypical organisms are less frequent causes. Anaerobes are extremely unusual.
Frequently implicated in cases of IE, S. aureus also can cause a significant amount of local and systemic damage. Most experts recommend that IE be considered in any patient who has documented S. aureus bacteremia. Obtain screening echocardiography unless the source of infection is known or there is no clinical indication of IE.
Some microorganisms are infrequent causes of IE but can have devastating effects. Candida species can cause extensive myocardial damage and require surgical consultation for emergent valve surgery. HACEK organisms (Hemophilus aphrophilus, Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella corrodens, Kingella kingae), historically referred to as “fastidious organisms,” are now routinely detected by modern, automated blood culture methods within five days. Coxiella burnetii is another unusual cause of IE, but it has been found to be such a common cause of “culture-negative IE” that whenever this organism is demonstrated in a blood culture or implicated by a positive serology, IE should always be considered.
Empiric therapy in patients with suspected IE should be guided by the usual microbiology of the disease. Good gram-positive coverage with a beta-lactam antibiotic is the recommended course. Because of the large number of cases caused by S. aureus and the rising frequency of methicillin-resistant S. aureus, it is reasonable to provide therapeutic coverage with vancomycin as well until additional culture data become available. Synergistic dosing of gentamicin is also routinely encouraged.
There are few data to support the use of quinolones, and they cannot be recommended as an initial choice. Microbiology results will guide long-term therapy, which should be decided after consultation with an infectious disease specialist.
Duration of antibiotic therapy will depend on the type of microorganism involved as well as the extent of the infection. In almost all situations, therapy will be IV antibiotics exclusively, as tissue concentrations from oral antibiotics are too low to effectively eradicate infections within the heart valves. Once the diagnosis has been established and the patient’s clinical course has stabilized, the remainder of therapy may occur in the outpatient setting, as long as there is access to an outpatient IV antibiotic provider and close follow-up with an experienced clinician to monitor for any adverse effects of long-term antibiotic use (most commonly leukopenia or renal insufficiency).
Dr. Spak is a clinical research fellow in the Department of Allergy and Infectious Diseases, University of Washington, Seattle.
- Li JS, Sexton DJ, Mick N, et al. Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis. Clin Infect Dis. 2000;30:633-638.
- Infective endocarditis: diagnosis, antimicrobial therapy, and management of complications. Circulation. 2005;111:e394-e434.
- Mylonakis E, Calderwood SB. Infective endocarditis in adults. N Engl J Med. 2001;345:1318-1330.
- Fowler VG Jr, Miro JM, Hoen B, et al. Staphylococcus aureus endocarditis: a consequence of medical progress. JAMA. 2005;293:3012-3021.