Infective Endocarditis - Clinical Tree

Definition

Infective endocarditis includes conditions in which structures of the heart, most frequently the valves, harbor an infective process that leads to valvar dysfunction, localized or generalized sepsis, or sites for embolism ( Box 15-1 ). The term infective endocarditis (IE) includes acute, subacute, and chronic processes; infection of bacterial, viral, rickettsial, or fungal etiology; and involvement of either native or prosthetic valves. As such, the term has no implication as to duration of the processes, infecting agent, or site of infection and thus supplants previously used terms such as “subacute bacterial endocarditis.”

Box 15-1

Definitions

Endocarditis: Exudative and proliferative inflammatory alterations of the endocardium, characterized by vegetations on the endocardial surface or within the endocardium. It may occur as a primary disorder (infective endocarditis) or as a complication of or in association with another disease (e.g., lupus erythematosus, rheumatic heart disease).
Infective endocarditis: Invasion and multiplication of microorganisms on the endocardial surface, within the endocardium, within the myocardium, or on prosthetic materials within and around cardiac structures. It includes conditions in which structures of the heart, most frequently the valves, harbor an infective process that leads to valvar dysfunction, localized or generalized sepsis, or sites for embolism. The term covers:
- ▪
  Acute, subacute, and chronic processes
- ▪
  Infection of bacterial, viral, rickettsial, or fungal etiology
- ▪
  Involvement of either native or prosthetic valves (or other prosthetic material)
Acute endocarditis: A severe form of infective endocarditis caused by virulent pyogenic microorganisms such as hemolytic streptococci or staphylococci. It can become life threatening within days.
Subacute endocarditis: A form of infective endocarditis that develops subtly over a period of weeks to several months. It may produce symptoms for months before heart valve damage or emboli make the diagnosis clear. It is usually caused by Streptococcus viridans or Streptococcus fecalis .
Active endocarditis: A surgical term indicating an operation carried out in the presence of obvious local cardiac infection manifested by inflammation, active vegetations, abscesses, burrowing sinuses, or fistulae. If such an operation is carried out while a patient is being treated with antibiotics for active infection, or has been treated within 2 weeks of operation, the disease is considered active.
Healed endocarditis: A surgical term indicating an operation carried out in the absence of obvious local cardiac infection and inflammation, generally following treatment and supposed eradication of microorganisms. It is characterized by lack of local inflammation; vegetations may be present but are generally endothelialized, and abscesses have resulted in well-defined and stable cavities, including sinuses and fistulae.
Native valve endocarditis: Infectious endocarditis involving a patient's own (native) heart valve.
Prosthetic valve endocarditis: Infectious endocarditis involving a surgically implanted prosthetic heart valve. Prosthetic valve endocarditis and its abbreviation PVE are familiar, standard, and historical designations for infective endocarditis on any heart valve substitute, but “prosthetic” is at times a misnomer (e.g., infection of a pulmonary autograft). A more appropriate term is replacement device endocarditis , but in this chapter the historical term is maintained for the sake of familiarity.

Historical Note

In 1806, Corvisart described mitral valve vegetations found at autopsy in a 39-year-old man. In their 1824 book on heart disease, Bertin and Bouillaud discuss induration and vegetations on the valves of patients dying with endocarditis. The term endocarditis was introduced by Bouillard in 1841 when he described clinical and pathologic features of the disease. Earlier, Morgagni, Lancisi, and Sandifort had described hearts with probable endocardial vegetations.

Weinstein and Brusch reported that in 1886, Wyssokowitch and Orth designed an experimental model for endocarditis in which aortic valve cusps of animals were traumatized and the animals subsequently injected with bacterial suspensions from patients with endocarditis. The animals developed murmurs, embolic complications, and valve lesions at autopsy.

At the 1885 Gulstonian Lecture, Osler described the classic features of endocarditis. By 1909, he had refined his understanding of the pathologic anatomy (“proliferative vegetations”) and introduced the clinical finding of changing murmurs. Along with Horder, Osler emphasized the role of blood cultures for diagnosis.

Successful treatment of endocarditis lagged behind pathologic and clinical descriptions. In the 1940s, the era of sulfa drugs, cure was achieved in about 5% of patients. By 1950, however, principles of antibiotic therapy had been established, including high-dose penicillin, long duration of treatment, and antibiotic suppression. It was also recognized at that time that delay in treatment, heart failure, advanced age, and preexisting rheumatic valvulitis were adverse prognostic factors. In 1940, Tauroff and Vessell successfully ligated a patent ductus arteriosus in treating a 2-year-old female with endocarditis. In 1957, Bahnson and colleagues reported occurrence of staphylococcal infection on silk sutures used for great vessel and intracardiac repairs.

Direct surgical treatment of IE began in 1961, when Kay and colleagues reported successful treatment of Candida endocarditis of the tricuspid valve. The native valve was débrided and an accompanying ventricular septal defect (VSD) closed. The first report of replacing a cardiac valve for native IE was published in 1965 by Wallace and colleagues. Their patient was a 45-year-old man who had severe aortic regurgitation with heavy Klebsiella vegetations on each cusp. He was treated intensively with antibiotics over 3 weeks, but there was resistant active infection and heart failure. Valve replacement was successful, eradicating the infection and restoring satisfactory hemodynamics.

In 1972, Merendino's group from the University of Washington reported the collective results of cardiac operations for endocarditis in 139 patients, 24 of whom were treated by mitral valve repair, mechanical prosthetic mitral replacement, and replacement of the aortic valve with β-propiolactone–sterilized allografts; 17 of 24 patients survived. The report emphasized continuing sepsis and heart failure as indications for operation, compared native (“primary”) to prosthetic (“secondary”) valve endocarditis, and differentiated between active and healed lesions.

Most of the modern concepts of surgical treatment of IE were articulated by the late 1970s and reviewed in publications by Stinson and by Richardson and colleagues.

Pathogenesis And Morphology

Pathogenesis

The most common site of cardiac involvement is on the line of closure of a valve surface, typically on the atrial side of atrioventricular valves and on the ventricular surface of semilunar valves. Once bacteria become attached to the surface, the vegetation matures through bacterial proliferation and fibrin deposition. The preponderance of bacteria below the surface of the vegetation provides protection from phagocytes and high antibiotic concentration.

Several predisposing factors have been identified that may contribute to or be responsible for development of IE. Despite Osler's initial observations, valvulitis of rheumatic fever was often confused histologically with IE. However, unlike rheumatic valvulitis, IE is not characterized in its early stages by global neovascularization or global inflammation of valve cusps. In most cases, the valvar endocardial surface must be altered to allow deposition of fibrin and platelets and subsequent attachment of bacteria. This injury may result from preexisting valvar lesions such as rheumatic valvulitis, anular or valvar calcification, or catheter trauma. Hemodynamic factors may contribute, such as the jet effect of blood flow through a patent ductus arteriosus or restrictive VSD, mitral valve prolapse, or bicuspid aortic valve. Bacteremia must occur in bacterial-based IE. Frequent transient bacteremias are found in 60% to 80% of normal individuals, but the number of organisms is small, and without the presence of one or more of the above factors, infection and vegetations do not result.

Several lines of evidence have demonstrated the importance of a compromised or altered immune system in the pathogenesis of IE. Histopathologic analysis of kidney tissue in patients with IE may reveal diffuse proliferative glomerulonephritis, with evidence of deposition of immunoglobulin (Ig)G and IgM. Circulating immune complexes (CICs) may be found in the glomerular basement membrane, retina, and peripheral lesions (Roth spots and Janeway lesions). Various manifestations of complement activation have been found in IE. Hooper and colleagues identified CICs in patients with prosthetic valve endocarditis (PVE). Kauffmann and colleagues found a positive correlation between CIC levels and duration of illness, and several investigators have noted a decline in CIC levels with successful treatment.

Staphylococcus aureus binds to porcine valvar endothelial cells by a mechanism that is specific and receptor mediated. This characteristic most likely represents a specific physiochemical interaction between microbial adhesins and a host-cell receptor that involves fibronectin lipoteichoic acid. IE involving a previously normal valve is often caused by S. aureus .

In several studies, one quarter of IE cases occurred on normal valves. Likely organisms are those that have increased adhesion molecules noted on dextran polymerization, namely Staphylococcus, Streptococcus viridans, and Enterococcus . Common risk factors are presence of overwhelming sepsis, resuscitation from shock, use of long-term indwelling catheters, intravenous (IV) drug abuse, and fungemia associated with prolonged antibiotic therapy. Only about 5% of patients with catheter sepsis are found to have IE, and this is generally due to staphylococcal organisms. Iatrogenic IE most often occurs in patients undergoing chronic hemodialysis who have frequent staphylococcal bacteremias and also may have sclerotic aortic or mitral valves.

It has long been acknowledged that usual bacterial vegetations of IE can result from seeding of a platelet-thrombin nidus after mechanical trauma. Garrison and Freedman produced endocarditis in rabbits by inserting a catheter into the endocardium and injecting bacterial isolates from humans with endocarditis. This model remains the principal experimental paradigm for the mechanical basis of IE.

Perhaps the most persuasive hypothesis for the pathogenesis of IE has been put forward by Rodbard. Basically, high-velocity jets of blood from a high-pressure source form at an orifice and enter a low-pressure sink. Venturi currents deposit bacteria immediately beyond the orifice to form vena contracta and result in mechanical erosion and deposition of platelets and thrombin ( Fig. 15-1 ). These mechanical conditions exist beyond stenotic valves, on the pulmonary artery opposite a patent ductus arteriosus ( Fig. 15-2 ), and on the left atrial aspect of a regurgitant mitral valve. Because most IE lesions of the aortic valve begin on the ventricular aspect, however, this suggests a role for valvar regurgitation in the pathogenesis of these lesions, and the Venturi effect would also apply ( Table 15-1 ).

Figure 15-1, Flow through a permeable tube. High-pressure source drives fluid through an orifice into a low-pressure sink. Curved arrows leaving the stream entering wall in upstream segment represent normal perfusion of lining layer. Velocity is maximal and perfusing pressure is low immediately beyond orifice, where momentum of stream converges stream lines to form a vena contracta. Low pressure in this segment results in reduced perfusion and may cause retrograde flow from deeper layers of vessel into the flowing stream. It is at the vena contracta that bacteria and other formed elements in blood accumulate.

Figure 15-2, Representation of infective endocarditis at a patent ductus arteriosus. Vegetations are deposited on pulmonary artery wall opposite a high-velocity jet through an open ductus.

Table 15-1

Loci of Infective Endocarditic Lesions

Modified from Rodbard ; in Weinstein and Schlesinger.

Condition	High-Pressure Source	Orifice	Low-Pressure Sink	Location of Lesions	Satellite Lesions
Coarctation of aorta	Central aorta	Coarctation	Distal aorta	Downstream wall of aorta	Lateral wall peripheral to stenotic lesion
Patent ductus arteriosus	Aorta	Ductus	Pulmonary artery	Pulmonary artery	Pulmonary artery Tricuspid leaflets
Arteriovenous fistula	Artery	Fistula	Vein	Communications and veins
Ventricular septal defect	Left ventricle	Defect	Right ventricle	Right ventricular surface of defect	Pulmonary artery
Aortic regurgitation	Aorta	Closed aortic cusps	Left ventricle	Ventricular surface of aortic valves	Mitral chordae
Mitral regurgitation	Left ventricle	Closed mitral leaflets	Left atrium	Atrial surface of mitral valves	Atrium
Pulmonary regurgitation	Pulmonary artery	Closed pulmonary cusps	Right ventricle	Ventricular surface Pulmonary cusps
Tricuspid regurgitation	Right ventricle	Closed tricuspid leaflets	Right atrium	Atrial surface Tricuspid leaflets

The abnormal endocardial and endothelial surfaces and increased turbulence imparted by numerous cardiac malformations create a substrate that is vulnerable to infection. The risk of infection associated with various cardiac abnormalities is reflected in the American Heart Association recommendations for prophylactic antibiotics in patients with such malformations.

Morphology

Usual sites for IE found at operation in patients with native valve endocarditis (NVE) not related to IV drug use are shown in Fig. 15-3 . Vegetations and erosive cavities are on the ventricular aspect of the aortic valve cusps and at the base of the atrial aspect of the mitral valve leaflets, often resulting in separation or discontinuity at the ventriculoarterial or atrioventricular junction. Less often, discrete perforations caused by isolated vegetations are located on the aortic cusps themselves. Occasionally, “drop lesions” from the aortic valve occur on the anterior mitral leaflet or the tensor apparatus of the mitral valve.

Figure 15-3, Usual sites of native valve infective endocarditis. A, Aortic valve with vegetation on noncoronary cusp and partial destruction of left coronary cusp. B, Aortic valve with vegetation between noncoronary and right coronary cusps extending as an anular abscess. C, Leaflet vegetation and ring abscess of posterior medial aspect of mitral valve. D, Mitral valve with drop lesion of anterior leaflet. Key: LCA, Left coronary artery; RCA, right coronary artery.

In the vast majority of non–drug-related cases of NVE, valve deposits are left-sided. In IV drug–related IE, the tricuspid valve is involved in about half the cases and aortic or mitral valves in the remainder. Perianular pseudoaneurysms (abscesses) are more frequently associated with aortic than with mitral vena contracta endocarditis and occur in about one third of cases studied by transesophageal echocardiography (TEE); S. aureus is the predominant organism. Clinically, presence of pericarditis, rapid progression of symptoms, and a high degree of atrioventricular block are associated with perianular abscess.

In PVE, all or most vegetations are on the ventricular aspect of the prosthesis. Only a small area of sewing-ring detachment may be apparent and may appear sterile. At operation, therefore, a thorough search must be made beneath the prosthesis for the bulk of the pathologic process. As noted in some series, there may be important detachment without apparent vegetations; this may occur in either the presence or absence of positive blood cultures. This represents PVE and requires replacement of the prosthesis.

Clinical Features And Diagnostic Criteria

Infective endocarditis is present when defined in accordance with New York Heart Association (NYHA) criteria : positive blood cultures associated with either new or changing murmurs or embolic phenomena; or new or changing murmurs in a patient with a congenital cardiac anomaly or prior valve damage, associated with either embolic phenomena or sustained fever, anemia, and splenomegaly. Most authorities accept a modification of that definition to include progressive heart failure in the presence of positive blood cultures. In an effort to improve specificity and sensitivity for the diagnosis of IE, investigators at Duke University have proposed criteria modeled after the Jones criteria for diagnosing rheumatic fever ( Box 15-2 ). These criteria include major and minor signs and symptoms, echocardiographic findings, possible iatrogenic and nosocomial factors (e.g., indwelling catheters), and history of IV drug abuse.

Box 15-2

Diagnostic Criteria for Infective Endocarditis

Major Criteria

Positive Blood Culture

Typical microorganisms for infective endocarditis (IE) from two separate blood cultures (viridans streptococci, ^a

a Including nutritional variant strains.

Streptococcus bovis , HACEK group, or community-acquired Staphylococcus aureus or enterococci) in absence of a primary focus, or
Persistently positive blood culture, defined as recovery of a microorganism consistent with IE from:
- Blood cultures drawn more than 12 hours apart, or
- All of three or majority of four or more separate blood cultures, with first and last drawn at least 1 hour apart

Evidence of Endocardial Involvement

Positive echocardiogram for infective endocarditis:
- Oscillating intracardiac mass on valve or supporting structures, or in path of regurgitant jets, or on implanted material, in absence of an alternative anatomic explanation, or
- Abscess, or
- New partial dehiscence of prosthetic valve or new valvar regurgitation (increase or change in preexisting murmur not sufficient for diagnosis)

Minor Criteria

Predisposition

Predisposing heart condition or IV drug use
Fever ≥38°C

Vascular Phenomena

Major arterial emboli, septic pulmonary infarcts, mycotic aneurysm, intracranial hemorrhage, conjunctival hemorrhages, Janeway lesions

Immunologic Phenomena

Glomerulonephritis, Osler nodes, Roth spots, rheumatoid factor

Microbiological Evidence

Positive blood culture but not meeting major criterion, ^b

b Excluding single positive cultures for coagulase-negative staphylococci and organisms that do not cause endocarditis.

or serologic evidence of active infection with organism consistent with infective endocarditis

Echocardiogram

Consistent with infective endocarditis but not meeting major criterion

Key: HACEK, Haemophilus spp., Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella spp., and Kingella kingae; IV, intravenous.

The most common clinical manifestation of IE is fever, which is present in 95% to 100% of patients. This applies to both NVE and PVE. The fever may be low grade or spiking and generally follows peaks of bacteremia by about 2 hours. Patients at risk for IE who develop unexplained fever for more than 48 hours should have two or more sets of blood cultures drawn from different sites. Because of the importance of proper identification of the offending organism, administration of antibiotics should be delayed until blood cultures have been obtained.

Positive blood cultures are obtained in about 95% of patients even when right-sided endocarditis, endocarditis caused by fungus, endocarditis in addicts, and endocarditis caused by fastidious organisms are included.

In IE cases confirmed by echocardiography, autopsy, or operation, positive blood cultures are obtained in 95% of cases with two blood specimens and in 98% with four specimens. However, occurrence of negative-culture endocarditis rises to about 10% in most surgical series, and PVE predominates in this group of patients.

Culture-negative IE is more likely with intracellular or fastidious organisms, or with previous antibiotic therapy. A history of antibiotic therapy or serologic evidence of Mycoplasma or Chlamydia species is likely responsible for negative-culture IE. Other causes include Candida, Aspergillus, and fastidious slow-growing organisms such as Q-fever (Coxiella burnetii) and Bartonella organisms.

A heart murmur is found in about 85% to 95% of IE patients. Changing murmurs occur much less frequently (≈15% to 20%). Although the concept of changing murmurs has been classically associated with IE since Osler's Gulstonian Lecture in 1885, clinically changing murmurs are not often appreciated. Ten percent of IE patients lack murmurs, particularly those with tricuspid involvement. Infection involving the aortic valve and root is often characterized by a relatively short diastolic murmur. There may be a murmur only in early systole or midsystole, although murmurs are often obscured by tachycardia. The murmur of mitral regurgitation caused by IE is similar to that of other mitral regurgitation murmurs and may exhibit a typical radiation posteriorly when the anterior leaflet is perforated. There also may be signs of mitral stenosis secondary to obstruction by large vegetations, in which case the murmur may be diastolic.

Pulse pressure may be normal or even narrowed in aortic regurgitation if IE is acute or important heart failure exists. Narrowed pulse pressure is caused by high left ventricular end-diastolic pressure with low cardiac output. There is frequently a gallop rhythm.

Infection of the mitral valve and its supporting structures is considered to be less frequent than aortic valve endocarditis, but may be more indolent in its course. The mitral valve is most commonly involved (≈40% of cases), followed by the aortic valve in about 36%, when S. aureus is the infecting organism. Right-sided endocarditis is almost uniquely related to the tricuspid valve. Right-sided IE accounts for about 10% of IE cases and usually occurs in the setting of IV drug abuse. Affected patients tend to be younger and have fewer comorbidities and less structural heart disease than patients with left-sided IE. Right-sided lesions are associated with fever, but often a cardiac murmur is initially absent.

Anemia frequently occurs in IE patients and has multifactorial causes, but primarily it results from marrow suppression secondary to chronic disease. Arthritis and arthralgias are infrequently seen today, generally because endocarditis is diagnosed earlier. Myalgias are common and may be associated with bacteremia or occasionally may result from myocardial microabscesses, generally occurring in staphylococcal bacteremia.

Severity of heart failure in a hospitalized IE patient is not appropriately classified by NYHA functional class. In general, heart failure can be termed mild when only small doses of diuretic or digitalis are necessary; moderate when large doses of diuretic, afterload-reducing medications and bed rest are necessary; and severe when cardiogenic shock is present and inotropic agents are needed.

Embolization is the presenting manifestation in about 10% to 15% of patients with left-sided IE, and about half of patients with IE have evidence of embolic phenomena on physical exam or via diagnostic imaging. These emboli seem about evenly distributed between cerebral and peripheral sites. Classic peripheral signs of endocarditis—Osler nodes, Janeway lesions, Roth spots, petechiae, and clubbing—are late manifestations and are infrequently seen in a surgical practice. The one exception may be Janeway lesions, and when noted, the infecting organism is almost always Staphylococcus .

In practice, diagnosis of IE is primarily based on two tests: blood cultures and echocardiography. The diagnosis is made most often by the presence of positive blood cultures and a cardiac lesion characterized by new stenosis, new regurgitation, or echocardiographic evidence for a vegetation. There are many more individuals with sepsis and positive blood cultures but without cardiac manifestations, and they are not considered to have IE.

An excellent microbiology laboratory is essential for accurate and prompt diagnosis. With current techniques, ability to culture fastidious organisms (e.g., Q-fever [Coxiella burnetii], Mycoplasma ) is high. Thus, diagnosis of IE involving cardiac valves or congenital malformations is made on a constellation of findings, usually fever, positive blood cultures, and hemodynamic derangement within the cardiac structures that is best assessed and followed by echocardiography.

Echocardiography has become a standard modality for diagnosis and continuing observation of patients with IE. In all situations, TEE has greater sensitivity and specificity than transthoracic echocardiography (TTE). Specificity for TEE is approximately 90% and sensitivity 95%. For TTE, accuracy ranges from 40% to 80%. These figures apply equally to NVE and PVE. Perivalvar or perianular cavities associated with prosthetic valves are more easily delineated with TEE, and in the view of some authorities, these represent pseudoaneurysms rather than abscess cavities in most patients. Box 15-3 summarizes the echocardiographic and clinical findings in IE that indicate the potential need for operation.

Box 15-3

Echocardiographic and Clinical Features Suggesting Potential Need for Surgical Intervention in Patients with Infective Endocarditis

Vegetation

Persistent vegetation after systemic embolization
Anterior mitral valve leaflet vegetation, particularly with size ≥10 mm
One or more embolic events during first 2 weeks of antimicrobial therapy ^a

a Surgery may be required because of risk of embolization.
Two or more embolic events during or after antimicrobial therapy ^a
Increase in vegetation size after 4 weeks of antimicrobial therapy ^b

b Surgery may be required because of heart failure or failure of medical therapy.

Valvar Dysfunction

Acute aortic or mitral regurgitation with signs of ventricular dilatation ^b
Heart failure unresponsive to medical therapy ^b
Valve perforation or rupture ^b

Perivalvar Extension

Valvar dehiscence, rupture, or fistula ^b
New heart block ^b
Large abscess or extension of abscess despite appropriate antimicrobial therapy ^b

An algorithm for diagnosis of IE begins with fever, requires positive blood cultures, includes some sign or symptom referable to the heart, and receives anatomic corroboration with TEE. From TEE, vegetation size, mobility, and position can be documented; degree of stenosis and regurgitation associated with the valve lesion can be assessed; and in PVE the presence of prosthetic leakage or perianular cavities can be determined.

The decision to use cardiac catheterization and angiography has varied over the years and continues to be made on a case-by-case basis. In few cases have vegetations been disturbed by intraarterial or venous catheters, and a reasonable policy is that if coronary artery disease or coronary embolization is suspected, coronary angiography should be undertaken before operation.

Neurologic abnormalities may have occurred in as many as 25% to 30% of IE patients at initial presentation. They are protean in nature and include stroke, transient ischemic attack, toxic encephalopathy, meningitis, brain abscess, loss of vision, seizures, headache, backache, and acute mononeuropathy. Funduscopic examination (Roth spots or flame hemorrhages), cerebrospinal fluid examination, and computed tomography (CT) scanning or magnetic resonance imaging (MRI) of the brain are performed as indicated, and results may alter timing of surgical therapy (see following discussion). MRI may be particularly effective in detecting cerebral embolic lesions in IE patients without neurologic symptoms.

Embolic events are reported frequently in IE patients (24%-67%). Prevalence is probably higher in IV drug–related endocarditis and perhaps slightly lower in PVE. The brain is the most frequently identified site of emboli.

Natural History

Epidemiology

The incidence of IE in the United States is reported to be 1.7 to as high as 11.6 episodes per 100,000 persons per year and is highest in the oldest age group, with variance due in great part to varying diagnostic and reporting criteria. Approximately 10,000 to 20,000 new cases of IE are diagnosed each year in the United States, accounting for about 1 in 1000 hospital admissions. The IE profile has changed over the past several decades; in the current era, IE is more frequently associated with invasive medical procedures and old age and less frequently associated with rheumatic heart disease and poor dentition. In the elderly, it is more often linked to a prosthetic valve and bacteria from the gastrointestinal tract. Terpenning and colleagues found that indwelling catheters were implicated as the source of bacteremia in half the cases of nosocomial IE.

Prosthetic heart valves represent a strong risk factor for IE. In a large study of adult patients, occurrence of PVE was 4.1% at 48 months after valve replacement in 1465 consecutive hospital survivors, and 64% of these patients died. Occurrence of PVE after valve replacement appears to be much lower in the current era, with overall risk about 1% to 5% in the first year after valve replacement and about 1% per year thereafter. Other studies report an annual incidence of PVE ranging from 0.12% to 0.4% per patient-year. The hazard for developing PVE is greater in patients operated on for NVE than in those undergoing valve replacement for other reasons ( Fig. 15-4 ). In addition to NVE as a risk factor for PVE, placement of a mechanical prosthesis (vs. a tissue valve), black race, male gender, and longer cardiopulmonary bypass time are incremental risk factors for subsequent development of PVE. Finally, prosthetic valve reoperation is a greater hazard for development of PVE than primary valve replacement ( Fig. 15-5 ).

Figure 15-4, Hazard function for prosthetic valve endocarditis (PVE) in patients with and without previous native valve endocarditis.

Figure 15-5, Hazard functions for prosthetic valve endocarditis (PVE) after original valve replacement and after reoperation. Both have early peaking and constant phases.

It is difficult to identify a precise causative factor for PVE. However, intraoperative surface contamination, introduction of contaminated blood or blood substitute, bacterial colonization of a member of the surgical team, bacterial aerosolization in ventilators, nasal colonization of the patient, and preexisting urosepsis have all been implicated. In contrast to prosthetic heart valves, implanted anuloplasty rings and indwelling pacemaker leads are infrequently disposed to IE.

Among the pediatric population, IE most often occurs in patients with VSDs and valvar aortic stenosis. Incidence is low (14.5 per 10,000 person-years) but 35 times that of the normal population. Incidence of IE in patients with VSD is reduced by about 50% after surgical closure. Postoperative IE in children most often occurs after aortic valvotomy, valve replacement, or use of a right ventricular–pulmonary artery conduit. In both pediatric and adult populations, mitral valve prolapse has emerged as a frequent preexisting malformation in the spectrum of IE (29% of patients in McKinsey and colleagues’ series ).

Causative Organisms

Approximately 80% of endocarditis cases are caused by streptococcal or staphylococcal species. S. aureus predominates as the infecting organism in the majority of hospital-acquired and drug-related cases of IE. S. aureus involves the mitral valve more than the aortic valve and results in higher occurrence of embolism compared with other organisms. Streptococcal IE accounts for about 30% of IE cases, and viridans streptococci are the most common causative organisms. Enterococci are the third leading cause of IE, implicated in about 10% of cases. Enterococcal IE typically occurs in elderly males with multiple comorbidities, results less commonly in embolic events, and disproportionately affects the aortic valve. Gram-negative bacilli account for about 5% of cases. Distribution of infecting organisms in the usual NVE population, however, may vary among institutions and temporally.

In PVE occurring within 2 months of operation, Staphylococcus epidermidis is the major offending organism. Late-onset PVE has the same general spectrum of causative organisms as NVE. Enterococcal PVE (usually caused by Enterococcus faecalis or Enterococcus faecium ) is usually associated with manipulation of the gastrointestinal or genitourinary tract or with malignancy.

Complications

The most frequent cardiac complication of NVE is heart failure, primarily caused by valvar regurgitation. However, NVE occasionally results in mitral or tricuspid valve stenosis and infrequently in aortic valve stenosis.

Perianular leakage and abscess and occasionally stenosis are the major causes of heart failure in PVE, with perianular extension occurring in over 50% of cases. It occurs in 10% to 40% of NVE cases and is more common with aortic valve involvement. This dangerous complication can lead to abscess formation, pseudoaneurysms, and aortocavitary fistula formation (which can develop from any aortic sinus). Myocardial abscesses (most commonly related to S. aureus infection) complicate IE in 20% to 40% of cases and are particularly common with PVE and aortic valve involvement. Patients with a bicuspid aortic valve and IE carry a high risk of abscess formation. Development of conduction abnormalities should prompt further TEE evaluation for abscess formation or extension. If left untreated, abscess cavities may progress to fistula formation and intracardiac shunting from myocardial perforation . Once these complications develop, mortality may exceed 40% despite surgical intervention. If the rare patient with a small abscess cavity (≤1 cm), clinically controlled infection, and multiple comorbidities is treated medically, close follow-up with serial echocardiography during prolonged antibiotic therapy is mandatory. Pericarditis typically occurs in association with anular abscess or myocardial perforation.

Renal complications of IE take at least four forms: prerenal failure secondary to low cardiac output, microabscess formation caused by septic emboli, glomerular dysfunction resulting from circulating immune complexes, and renal failure caused by antibiotic toxicity.

Embolic events are common in IE patients ( Fig. 15-6 ), with a reported prevalence of 43% in NVE, 67% in IV drug–associated IE, and 25% in PVE. Ting and colleagues report a 19% occurrence of splenic emboli in their series of patients with IV drug–related endocarditis. Occurrence is about 5% in the usual type of left-sided NVE. Metastatic infection of viscera is typically caused by Staphylococcus . Other classic peripheral manifestations of IE (e.g., petechiae, Osler nodes, splinter hemorrhages) are infrequently seen now, probably because of earlier intervention in the disease process. Multiple coronary emboli may result in myocardial infarction and ventricular dysfunction; the highest risk of embolic complications occurs with S. aureus, Candida, and HACEK ( Haemophilus, Actinobacillus, Cardiobacterium, Eikenella, and Kingella ) species.

Figure 15-6, Number of embolic events by site of embolization in a series of 365 patients with 131 (34%) embolic events (some patients had more than one site of embolization).

About half the embolic complications of IE are associated with neurologic manifestations , and one fourth to one third of patients with NVE or PVE at some time have neurologic complications, 90% of which are related to emboli in the distribution of the middle cerebral artery. Presence of S. aureus increases risk of neurologic complications. Stroke is the most common neurologic event. Vegetations are seen on echocardiography in about 40% of patients with neurologic complications and in about 30% of patients without neurologic sequelae. Some characteristics of vegetations visualized on TTE or TEE (e.g., density, mobility) may not be helpful in defining emboli risk. Cerebral embolism generally occurs before the start of antibiotic therapy, with risk of stroke falling rapidly after initiation of effective antibiotics. A European multicenter study estimated the prevalence of acute ischemic stroke at 12% (CL 10%-14%) on hospital admission, but only 3.7% (2.7%-4.9%) after start of appropriate antibiotic therapy. Data from the International Collaboration on Endocarditis indicated a stroke incidence of 4.8 per 1000 patient-days during the first week of antibiotic therapy, which decreased to 1.7 per 1000 patient-days in the second week and fell further thereafter. After instituting antibiotic therapy, however, both absolute vegetation size and observed increase in vegetation size are associated with increased embolic risk ( Fig. 15-7 ). Large vegetations are commonly caused by the HACEK group of organisms and fungi.

Figure 15-7, Risk of embolic events according to vegetation size in patients followed 1 to 942 days (mean, 151 days) after initiation of antibiotic treatment for left-sided infective endocarditis (217 episodes in 211 patients; multicenter study, April 1996 to June 2000). Key: CI, Confidence interval; RR, relative risk.

The most devastating neurologic complication is intracerebral hemorrhage, which complicates about 5% of IE cases and carries a mortality exceeding 50%. The pathophysiology may involve septic arteritis with erosion of the vessel wall during uncontrolled infection, hemorrhage following cerebral infarction, or rupture of a mycotic aneurysm. For patients on chronic anticoagulation who develop neurologic symptoms, anticoagulation should be discontinued (or maintained at low therapeutic levels in patients with a mechanical prosthetic valve) until intracranial hemorrhage can be excluded by CT scanning or MRI.

Medical treatment alone (vs. operative) increases risk of embolism. However, delaying repair of the cardiac lesion is usually advised in the presence of central nervous system (CNS) complications. Morbidity is less when repair is done in the presence of cerebral infarction versus cerebral hemorrhage (see Indications for Operation ). Evidence indicates that cerebral mycotic emboli will regress after extirpation of the valvar septic lesion, so early cardiac operation in patients with mycotic aneurysm and cerebral abscess may be advisable.

Mortality

Mortality associated with IE is reported at 15% to 20% during the initial hospitalization and 20% to 30% during the first year. Early mortality is similar between NVE and PVE and between mitral and aortic valve IE. Gram-negative bacillus and fungal IE carry a mortality exceeding 50%. Development of heart failure, intracardiac abscess, embolism, a large mobile vegetation, hemodynamic instability, altered mental status, immunocompromise, and advanced age have also been identified as risk factors for mortality.

Therapy

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here