Infective Endocarditis
Paul C. Cremer
I. INTRODUCTION
A. Epidemiology. The overall incidence of infective endocarditis is 3 to 10 per 100,000 patient-years, and in the United States, approximately 15,000 new cases are diagnosed every year. In developing countries, infective endocarditis is still most often associated with rheumatic heart disease, and patients are more likely to have a subacute or chronic presentation. Despite the decline of rheumatic heart disease in the developed world, the incidence of infective endocarditis has not decreased. Instead, the predominant risk factors have changed. In the developed world, an acute presentation is more common, and patients are more likely to be elderly with degenerative valve disease, prosthetic valves, chronic indwelling devices, or other predisposing medical conditions.
B. Morbidity and mortality. In the past few decades, outcomes have improved dramatically in other cardiovascular diseases, but endocarditis still has significant morbidity and mortality. Despite advances in diagnostic and therapeutic procedures, in-hospital mortality is 15% to 20%, and 1-year mortality is approximately 40%. In part, this persistently high morbidity and mortality is related to the shift in the epidemiology of infective endocarditis. More cases are now attributable to highly destructive microorganisms in older patients with prior cardiac disease and other comorbidities.
C. Heart valve team. Because of this changing risk profile and the diverse nature of presentation, the diagnosis and treatment of endocarditis remain a challenge. A high index of suspicion is required because delayed or missed diagnoses can lead to complications related to invasive and metastatic disease. Therefore, a heart valve team approach including cardiology, infectious disease, and cardiothoracic surgery is recommended to ensure expeditious diagnosis and treatment. In patients with destructive disease who may require urgent or emergent surgery, care should be provided in a center with immediate access to cardiac surgery. Occasionally, metastatic disease may also require the expertise of neurology, interventional neurology, neurosurgery, and abdominal or vascular surgery.
II. ETIOLOGY
A. Classification. Given the variable presentations of infective endocarditis, a systematic approach to the disease is essential. Sir William Osler famously distinguished subacute from acute bacterial endocarditis in 1885. Acute infective endocarditis leads to valve destruction and metastatic complications in days to weeks, whereas subacute disease progresses over weeks to months. This distinction remains a cornerstone in any evaluation of a patient with suspected endocarditis. However, in the current era of cardiac surgery and devices, frequent health care contact, and resistant microorganisms, patients can be further classified according to the structures involved, mode of acquisition, and culpable pathogen.
B. Structures involved. Patients are typically classified as having native valve, prosthetic valve, or cardiac device-related endocarditis.
1. Native valve endocarditis is further classified into left- and right-sided disease. The most common predisposing valve lesion for left-sided endocarditis is degenerative mitral valve disease with mitral regurgitation, which has been observed in 40% to 45% of cases. In the developed world, mitral valve prolapse with regurgitation is the most frequent abnormality, but in elderly patients, mitral annular calcification can also serve as a nidus for endocarditis. In 25% to 30% of patients with native left-sided endocarditis, aortic regurgitation is the predisposing lesion, and a bicuspid aortic valve is also a risk factor.
2. Right-sided endocarditis typically involves the tricuspid valve. The pulmonic valve is rarely infected. The most common risk factor for right-sided endocarditis is intravenous drug use. Indwelling lines and cardiac devices are also common risk factors as line-related infection can spread to involve the tricuspid valve. A ventricular septal defect is a risk for endocarditis, often involving the septal leaflet of the tricuspid valve. Other congenital heart defects that increase the risk for endocarditis include a patent ductus arteriosus, coarctation of the aorta, and tetralogy of Fallot. Conversely, secundum atrial septal defects are not associated with an increased risk of infective endocarditis.
3. When compared to the general population, patients with prosthetic valves are 50 times more likely to develop infective endocarditis. Two-thirds of all cases occur within the first year after surgery, and the incidence of infective endocarditis is similar in bioprosthetic and mechanical valves. Given the preponderance of infection within the first year, prosthetic valve endocarditis is categorized as early-onset (within 2 months of surgery), intermediate-onset (2 months to 1 year after surgery), and late-onset (greater than 1 year after surgery) disease. Early-onset prosthetic valve endocarditis is most likely related to surgery or another health care encounter. Intermediate-onset disease is less likely to be related to the index surgery, but health care-related infection is still the most common cause. Risks for early- and intermediate-onset endocarditis include prolonged surgery, reoperation, and sternal wound infections. As intermediate prosthetic valve endocarditis approaches 1 year, infection is more likely to be community acquired, and late-onset disease more closely resembles native valve endocarditis in microbiology. Data on infective endocarditis after transcatheter valve replacement are emerging, but the incidence and epidemiology appear similar to surgical valve replacement, with most cases occurring within the first year after the procedure.
4. Given the rising number of patients with cardiac devices (permanent pacemakers and implantable defibrillators) and the increased risk of infection with an intravenous lead, the incidence of endocarditis is increasing in these patients. Of note, local device infection should be distinguished from device-related endocarditis. A local device infection is restricted to the pocket. In device-related endocarditis, infection can spread along the lead to the endocardium and the electrode tip. The primary cause of cardiac device-related infection is contamination by local bacteriologic flora at the time of device implantation, although hematogenous seeding is also common. In general, risks for device-related endocarditis include diabetes mellitus, heart failure, generator replacement, and renal dysfunction.
C. Mode of acquisition. Infective endocarditis is community acquired, health care associated, or related to intravenous drug use. Health care-associated infection is further divided into nosocomial and non-nosocomial acquisition. The latter refers to infection acquired after a recent hospitalization, at a nursing home or long-term care facility, at a hemodialysis center, or at an infusion clinic. The etiologies of infective endocarditis in specific communities can vary widely, and the spectrum of disease is primarily determined by the rates of intravenous drug use and health care-associated infection. Overall, community acquisition is most common (˜70%), followed by nosocomial (˜15%) and non-nosocomial (˜10%) infection. In North America, health care-associated endocarditis is more common (35% to 40%) and is primarily related
to non-nosocomial infection; chronic intravenous access is present in approximately 25% of cases of infective endocarditis. In intravenous drug users, concomitant human immunodeficiency virus (HIV) infection is common and increases the risk for infective endocarditis. Finally, with community-acquired disease, the source of infection is usually not identified but may be attributed to a dental procedure or an infected skin lesion.
to non-nosocomial infection; chronic intravenous access is present in approximately 25% of cases of infective endocarditis. In intravenous drug users, concomitant human immunodeficiency virus (HIV) infection is common and increases the risk for infective endocarditis. Finally, with community-acquired disease, the source of infection is usually not identified but may be attributed to a dental procedure or an infected skin lesion.
D. Microbiology
1. When the mode of acquisition and infected cardiac structures are established, the most likely microorganisms can be determined (Table 9.1). Overall, in the developed world, staphylococci (˜40%) are now a more common cause of infective endocarditis than streptococci (˜30%). In fact, Staphylococcus aureus is not only the most common microorganism in intravenous drug users, but also the most frequent pathogen in native valve nonintravenous drug users as well as in patients with prosthetic valves and cardiac devices. Among community-acquired cases of infective endocarditis, staphylococcal infection is most frequent (30% to 35%) and is closely followed by oral streptococci (20% to 25%). In health care-associated infection, staphylococcal infection predominates and is responsible for approximately 70% of nosocomial and 65% of non-nosocomial infections.
The coagulase-negative S. epidermidis is a frequent cause of prosthetic valve and cardiac device-related endocarditis.
The coagulase-negative S. epidermidis is a frequent cause of prosthetic valve and cardiac device-related endocarditis.
TABLE 9.1 Frequency of Infective Endocarditis by Most Common Microorganisms in Patients with Native Valves, Intravenous Drug Use, Prosthetic Valves, and Cardiac Devices | ||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||
2. Enterococcal species cause approximately 10% of cases of infective endocarditis, typically related to Enterococcus faecalis (85% to 90%). The remaining cases are mostly because of E. faecium, with a few cases attributed to E. durans. Enterococci typically reside in the gastrointestinal and genitourinary tracts, and endocarditis can be nosocomial, related to manipulation of the colon, urethra, or bladder.
3. Culture-negative endocarditis represents another 10% of cases and is defined by the lack of identification of a causative organism after three blood cultures.
4. The HACEK organisms (Haemophilus, Aggregatibacter actinomycetemcomitans, Cardiobacterium hominis, Eikenella corrodens, and Kingella species) are now a rare cause of endocarditis (˜2%), especially in North America. Other Gram-negative bacteria (Enterobacteriaceae, Acinetobacter species, Pseudomonas aeruginosa) infrequently cause endocarditis, and risk factors include intravenous drug use, end-stage liver disease, central venous catheters, and older age. P. aeruginosa can be highly destructive and poorly responsive to antibiotics.
5. Coxiella burnetii, the causative agent of Q fever, can cause endocarditis; herding cattle, sheep, or goats is a risk factor. Bartonella henselae (cat scratch disease) and B. quintana (trench fever) can also cause endocarditis, most often in homeless or alcoholic patients. Most cases of endocarditis related to C. burnetii or Bartonella species have been diagnosed in Europe.
6. Fungi are another rare cause of endocarditis (˜2%). Candida albicans causes approximately 25% of cases of fungal endocarditis, other Candida species are responsible for another 25%, and Aspergillus species also cause 25% of cases. Fungal endocarditis is most common in patients with prosthetic heart valves. Other risk factors include intravenous drug use and an immunocompromised condition. Often, Candida endocarditis is nosocomial.
III. PATHOPHYSIOLOGY
A. Vegetations. The primary manifestation of infective endocarditis is a vegetation on an endothelial defect. Normal valve endothelium is resistant to colonization by circulating bacteria. Endothelial damage commonly occurs from inflammation as in rheumatic heart disease, degenerative valve changes in the elderly, and indwelling catheters. These damaged areas often have platelets and fibrin that can promote the growth and adherence of microorganisms. Typical endocarditis pathogens can adhere to damaged valves and promote local procoagulant activity. Bacterial colonies thus attract further fibrin and inflammatory cells and nurture vegetations.
B. Foreign material. Microorganisms can also adhere to foreign material. In cardiac device-related infective endocarditis, vegetations can be found from the subclavian vein to the superior vena cava, on the electrode lead, the tricuspid valve, and endocardium of the right atrium and ventricle. In prosthetic valve endocarditis, the sewing ring is often infected, and bioprosthetic cusps seem more resistant to disintegration by bacterial enzymes compared to native valve cusps.
C. Invasive disease. Tissue disintegration that involves the valve annulus and spreads to extravascular areas is referred to as invasive disease. Typically, invasive disease develops in stages beginning with cellulitis, then forming an abscess and abscess cavity, and finally resulting in a pseudoaneurysm. Specific consequences of invasive disease include internal fistulas, perforations, and heart block. In general, enzymatic degradation by invasive bacteria prefers connective tissue and fat, and muscle is often preserved. Also, the invasive nature of microorganisms is different, and this distinction relates to acute and subacute presentations of infective endocarditis. Enterococcal infections are often minimally invasive, and fungal infections can have large vegetations with minimal invasion or tissue destruction. Oral streptococci can develop invasive complications, but usually over weeks to months. Conversely, S.
aureus can dissolve annular tissue and cause dehiscence of a prosthesis within days to weeks. The coagulase-negative S. lugdunensis can also have a rapidly destructive clinical course.
aureus can dissolve annular tissue and cause dehiscence of a prosthesis within days to weeks. The coagulase-negative S. lugdunensis can also have a rapidly destructive clinical course.
D. Anatomic considerations
1. Vegetations on native valves typically form on leaflets, but can also occur on chordae. In general, vegetations form on the low-pressure side of the valve where blood flow strikes the valve surface. Regurgitation results from a loss of leaflet integrity or chordal rupture. In aortic valve endocarditis, a secondary (“kissing”) lesion commonly forms on the anterior mitral valve leaflet and less commonly develops on the aortic wall.
2. The magnitude of invasive disease is defined by the percentage of annulus involved, extent beyond the annulus, and the depth and size of cavities. Invasive disease is more common on the left side of the heart and is more frequent with the aortic valve compared to the mitral valve. In native valve endocarditis, extra-aortic invasion is often localized. The site of subcommissural invasion may be small, effectively hiding an extra-aortic focus of infection. Invasive disease is more common in prosthetic valve endocarditis, and extra-aortic invasion is frequently circumferential. However, cellulitis and abscesses can occur anywhere along the annulus. Heart block is most often associated with aortic valve endocarditis. Typically, bacterial invasion extends from the posterior aortic root into the right atrium and triangle of Koch, resulting in destruction of the atrioventricular node and bundle of His. In mitral valve endocarditis, anterior annular invasion leads to destruction of the aortomitral curtain, and posterior invasion enters the atrioventricular groove and separates the atrium from the ventricle.
IV. CLINICAL MANIFESTATIONS
Infective endocarditis can have a variable presentation depending on the pathogen and preexisting cardiac disease. Nonetheless, the hallmark manifestations of infective endocarditis are fever and a heart murmur.
A. Fever. Most patients (˜80% to 90%) will present with a fever and often have chills, poor appetite, or weight loss. Notable exceptions include patients that are pretreated with antibiotics, elderly or immunocompromised patients, and patients with cardiac device-related endocarditis. These patients can have atypical presentation, and a high index of suspicion is essential. In addition, infective endocarditis should be considered in any patient with fever and one of the following: known cardiac lesion, recent procedure associated with bacteremia, congestive heart failure, new stroke or unexplained embolic event, or a peripheral abscess without a known cause. In addition, endocarditis should be suspected in any patient with a catheter-related bloodstream infection and persistently positive blood cultures after catheter removal.
B. Heart murmur. The majority of patients have a heart murmur (˜85%). Many patients have preexisting heart murmurs related to valvular stenosis or regurgitation, valve replacement, congenital heart disease, or previous endocarditis.
C. Classic signs. Three-quarters of patients in the developed world present acutely within 30 days of infection. Embolic or vasculitic skin lesions are uncommon in these patients. In addition, patients with health care-associated and cardiac device-related endocarditis are unlikely to have classic features. However, these features are more common in the developing world where oral streptococcal infection still predominates. Examples include subungual hemorrhages, Roth spots (oval retinal hemorrhages), Osler nodes (subcutaneous nodules in the pulp of the digits), and Janeway lesions (hemorrhagic macular nontender lesions on the palms and soles of the feet).
D. Modified Duke criteria (Table 9.2). The modified Duke criteria are the most sensitive and specific criteria available and are divided into definite, probable, and rejected diagnostic groups. These criteria have been well validated in numerous studies in diverse populations including children, elderly, patients with prosthetic valves, and injection
drug users. However, although still clinically useful, the modified Duke criteria may be less sensitive in these populations.
drug users. However, although still clinically useful, the modified Duke criteria may be less sensitive in these populations.
TABLE 9.2 Modified Duke Criteria for the Diagnosis of Infective Endocarditis | |||||||
|---|---|---|---|---|---|---|---|
|
1. A definite pathologic diagnosis requires one of two criteria. For a definite clinical diagnosis, two major, one major and three minor, or five minor criteria are needed.
2. The possible diagnostic group includes one major and one minor or three minor criteria.
3. With a rejected diagnosis, an alternative diagnosis is evident. A diagnosis is also rejected if clinical manifestations resolve within a few days of antibiotics, or if there is no pathologic evidence of infective endocarditis at surgery or autopsy.
E. Complications. Patients with infective endocarditis can have complications related to valve damage, invasive disease, metastatic disease or as a result of treatment.
1. The most common complication of endocarditis is heart failure (30% to 40%), and the most common cause is aortic or mitral regurgitation. Heart failure is less common in right-sided endocarditis.
2. Stroke occurs in 15% to 20% of patients with infective endocarditis. The most common mechanism is ischemic stroke from a septic embolism. Clinically apparent embolic strokes most often involve the middle cerebral artery. Hemorrhagic transformation of an ischemic stroke can also occur and is often devastating. Other mechanisms of stroke include septic erosion of an artery without aneurysm and rupture of a mycotic aneurysm. Cerebral mycotic aneurysms tend to occur at branch points in the distal vessels.
3. Emboli besides stroke affect another 20% of patients. Septic pulmonary emboli or abscesses in right-sided or cardiac device-related endocarditis can present as pleuritic chest pain, cough, or hemoptysis. In addition to the brain, left-sided endocarditis frequently embolizes to the spleen. Vegetations can also embolize to coronary arteries, kidneys, and bowel. In general, embolism is more common with larger vegetations, and the risk of embolism is greatest in the first few days of antibiotic therapy. Fungi in particular may have especially large vegetations that embolize and cause major limb artery occlusion.
4. Periannular complications can frequently present with heart failure or heart block and are associated with S. aureus endocarditis. In patients with infective endocarditis, especially involving the aortic valve, the electrocardiogram should be followed closely for PR prolongation and complete heart block. The development of complete heart block in a patient with significant aortic regurgitation can result in rapid hemodynamic decompensation, but it is unclear if patients with a prolonged PR interval benefit from a prophylactic temporary pacemaker.
5. Acute renal failure is a frequent complication (˜30%) in patients with infective endocarditis. Common causes include hemodynamic impairment, antibiotic toxicity, contrast nephropathy, renal infarction, and immune complex and vasculitic glomerulonephritis.
V. LABORATORY TESTING
A. Blood cultures. In patients with fever for more than 48 hours who have risk factors for endocarditis or newly diagnosed left-sided valvular regurgitation, at least two sets of blood cultures should be drawn. Risk factors include preexisting valvular heart disease, prosthetic valves, certain congenital heart diseases, immunodeficient conditions, injection drug users, or patients with a history of infective endocarditis. In stable patients with a subacute or chronic presentation, three sets of blood cultures should be drawn at peripheral sites at least 6 hours apart before antibiotics are given. Similarly, if a stable patient has suspected subacute or chronic endocarditis and is already on antibiotics, therapy should be stopped and blood cultures repeated. Antibiotic therapy may need to be discontinued for 7 to 10 days until blood cultures return positive. In fact, “culture-negative” endocarditis is often a result of the use of antibiotics before blood cultures are obtained. Furthermore, in endocarditis, bacteremia is constant. Therefore, blood cultures can be drawn when a patient is afebrile, and positive results from only one set of several blood cultures may be a contaminant. Finally, with continuous monitoring of blood cultures and nonculture-based diagnostics, routine incubation for more than 7 days is no longer necessary.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
