- Email: [email protected]
Endocarditis
CARDIOVASCULAR INFECTIONS
Endocarditis Sara E Cosgrove Adolf W Karchmer
Infectious endocarditis results from microbial infection of the endothelial surface of the heart. The characteristic lesion, a vegetation (Figure 1), is a mass of platelets and fibrin within which circulating micro-organisms have adhered and enmeshed.
1 Surgical specimen of a removed mitral valve showing vegetations (arrows).
Epidemiology and microbiology
Most cases of native-valve endocarditis are caused by viridans streptococci, Streptococcus bovis, Staphylococcus aureus or enterococci. In patients with community-acquired native-valve endocarditis who are not injecting drug-users, about one-third of cases are caused by streptococci, one-third by Staph. aureus and one-third by other organisms, including the HACEK group (Haemophilus spp., Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella spp. and Kingella kingae) of fastidious Gram-negative coccobacilli. When actively sought, Bartonella spp. have been found to cause 3% of cases, and Coxiella burnetii has caused a similar percentage of cases in selected areas. Both organisms are difficult to grow in blood cultures, and the diagnosis may require serological evaluation.2 The incidence of infectious endocarditis in injecting drug-users is 2–5%/year. Infection commonly involves the right heart valves and often occurs in previously normal valves. More than 50% of cases are caused by Staph. aureus. Nosocomial native-valve endocarditis is an emerging problem. Staph. aureus is the most common pathogen (55%), followed by enterococci (16%) and coagulase-negative staphylococci (10%). Central venous catheters are a major risk factor. The prevalence
Native-value endocarditis: in developed countries, the annual incidence of infectious endocarditis remained stable at 1.7–4.2/100,000 from 1950 to 1987, but recent data suggest that the incidence may be increasing; studies from Sweden and the USA report annual incidences of 5.9/100,000 and 11.6/100,000.1 The most likely explanation is the increased prevalence of predisposing conditions in an ageing population; 55–75% of patients with native-valve endocarditis have such conditions, including: • rheumatic valve disease • degenerative heart disease • mitral valve prolapse • congenital heart disease • hypertrophic cardiomyopathy • intravenous drug abuse • a prosthetic heart valve. The mean age of patients with infectious endocarditis has increased in parallel with a decline in rheumatic valve disease and an increase in degenerative valvulopathy as the substrate for infection. Most cases of infectious endocarditis in the over-60s are associated with degenerative heart disease. Mitral valve prolapse with mitral regurgitation is associated with only a 4–8-fold increased risk of infectious endocarditis compared with normal valves; however, because of its prevalence, it accounts for 7–30% of cases of nativevalve endocarditis. In young adults with infectious endocarditis, 20% of cases occur in those with congenital abnormalities, who now survive into adulthood as a result of corrective surgery. Bicuspid aortic valve is the most common congenital heart lesion and is emerging as an important predisposing cause of infectious endocarditis in the fourth and fifth decades of life.
Sara E Cosgrove is Assistant Professor of Medicine at Johns Hopkins University School of Medicine, Baltimore, USA. Conflict of interests: none declared. Adolf W Karchmer is Professor of Medicine at Harvard Medical School and Chief of the Division of Infectious Diseases at the Beth Israel Deaconess Medical Center, Boston, USA. Conflict of interests: none declared.
MEDICINE 33:4
2 Osler nodes on the hand and the pulp of the finger (arrows) in a young woman with rheumatic heart disease and aortic valve endocarditis caused by Staphylococcus aureus.
66
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
of infectious endocarditis following Staph. aureus catheter-related bacteraemia averages 6.1%, but was 23% in a study that systematically examined patients using transoesophageal echocardiography (TOE).3 Patients with Staph. aureus catheter-related bacteraemia should undergo TOE to exclude infectious endocarditis.
during the first year after valve surgery and almost 6% after 5 years. The risk is greatest during the first 6 months. Methicillin-resistant coagulase-negative staphylococci are responsible for most cases seen within the first year after valve placement. After 12 months, the microbiology of prosthetic-valve endocarditis resembles that of the native-valve form, though the incidence of coagulase-negative staphylococci is greater (10%).
Prosthetic-valve endocarditis: in 10–15% of patients with infectious endocarditis, infection develops on prosthetic valves. The cumulative prevalence of prosthetic-valve endocarditis is 3%
Marantic endocarditis (also known as thrombotic endocarditis) occurs in patients with malignancy, particularly adenocarcinoma.4 Platelets and fibrin form vegetations on heart valves as a consequence of a hypercoagulable state. Patients present with fever and embolization of the spleen, kidneys and CNS, but blood cultures are negative.
Modified Duke clinical criteria for diagnosis of infectious endocarditis Major criteria Positive blood culture • Two separate blood cultures yielding organisms that typically cause infectious endocarditis (viridans streptococci, Streptococcus bovis, HACEK,1 Staphylococcus aureus or community-acquired enterococci without a primary focus) • Persistently positive blood cultures (defined as positive blood cultures drawn more than 12 hours apart or all of three or a majority of four or more separate positive blood cultures, the first and last drawn at least 1 hour apart) • Single positive blood culture for Coxiella burnetii or antiphase I IgG antibody titre > 1:800 Evidence of endocardial involvement • Echocardiography positive for infectious endocarditis2 Oscillating intracardiac mass on a valve or supporting structure in the path of regurgitant jets or on implanted material in the absence of an alternative anatomical explanation Abscess New partial dehiscence of a prosthetic valve • New valvular regurgitation (change in pre-existing murmur is not adequate)
Clinical features Presentation of infectious endocarditis ranges from an acute syndrome of systemic toxicity and rapid progression with intracardiac and extracardiac complications, to an indolent illness that develops over weeks with low-grade fever and minimal cardiac dysfunction. Fever is the most common symptom (90% of patients), but may be muted or absent in the elderly, in those with chronic renal failure, and in patients who have received prior antibiotic therapy. Other nonspecific symptoms include chills, anorexia, weight loss, myalgia, cough and back pain. Heart murmurs are found in 85% of patients with native-valve endocarditis and most often represent the predisposing cardiac lesion. A new or changing murmur is seen in only 10–40% of patients; this may signal rapid valvular damage and impending heart failure. Peripheral signs – the classic peripheral Oslerian manifestations of infectious endocarditis (petechiae, splinter haemorrhages, Osler’s nodes – Figure 2, Janeway lesions, Roth spots) are not pathognomonic for infectious endocarditis and represent the sequelae of prolonged disease. They are now seen uncommonly, because most patients seek medical attention earlier in the course of the disease.
Minor criteria • Predisposing heart condition or intravenous drug use • Fever (≥ 38.0°C) • Vascular phenomenon (major arterial emboli, septic pulmonary infarcts, mycotic aneurysm, intracranial haemorrhage, conjunctival haemorrhages or Janeway lesions) • Immunological phenomenon (glomerulonephritis, Osler’s nodes, Roth spots or rheumatoid factor) • Microbiological evidence (positive blood culture, but less than major criterion,3 or serological evidence of active infection with an organism consistent with infectious endocarditis) Haemophilus spp., Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella corrodens and Kingella kingae 2 TOE recommended in patients with prosthetic valves or complicated endocarditis and who have possible endocarditis 3 Excludes a single positive culture for coagulase-negative staphylococci or organisms that do not cause infectious endocarditis 1
Source: Durack D T et al. Am J Med 1994; 96: 200–9 and Li J S et al. Clin Infect Dis 2000; 30: 633–8.
4 Echocardiogram of a mitral valve showing a large, round vegetation (arrow).
3
MEDICINE 33:4
67
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
Recommended antibiotic treatment in infectious endocarditis1 Infecting organism
Antibiotic
Dose2
Duration (weeks)
Comments
• Prosthetic-valve endocarditis is treated with 6 weeks’ benzylpenicillin (or vancomycin) and 2 weeks’ gentamicin (regimens 1 and 2) or 6 weeks’ gentamicin (regimen 3) • A 2-week regimen is appropriate when heart failure, aortic insufficiency and conduction abnormalities are absent, embolic disease is absent, the infection is on native valves, vegetations are < 5 mm on echocardiography, and there is a clinical response in 5 days • In regimen 1, a higher dose of penicillin (11.2 g daily) is recommended by most experts in the USA; in addition, ceftriaxone, 2 g i.v. or i.m. 24-hourly, can be substituted for benzylpenicillin • Regimen 3 is also preferred in patients infected with nutritionally variant streptococci (Abiotrophia spp.)
Viridans streptococcal and Streptococcus bovis native-valve endocarditis 1 Fully sensitive to penicillin (MIC ≤ 0.1 mg/litre)
Benzylpenicillin plus Gentamicin3 Ceftriaxone Benzylpenicillin
1.2–1.9 g i.v. 4-hourly
2
1 mg/kg i.v. 8–12-hourly 2 g i.v. or i.m. 24-hourly 1.2–1.9 g i.v. 4-hourly
2 4 4
2 Relatively resistant to penicillin (MIC > 0.1 to < 0.5 mg/litre)
Benzylpenicillin plus Gentamicin3
2.4 g i.v. 4-hourly
4
1 mg/kg i.v. 8–12-hourly
2
3 Penicillin (MIC ≥ 0.5 mg/litre)
Benzylpenicillin plus Gentamicin3
2.4–3 g i.v. 4 hourly
4–6
1 mg/kg i.v. 8–12-hourly
4–6
Vancomycin3
15 mg/kg i.v. over 100 minutes 12-hourly
4
Ampicillin or amoxicillin plus Gentamicin3
2 g i.v. 4-hourly
4–6
1 mg/kg i.v. 8–12 hourly
4–6
2 Highly resistant to gentamicin (MIC ≥ 500 mg/litre) but low-level resistance to streptomycin (MIC < 2000 mg/litre)
Ampicillin or amoxicillin plus Streptomycin3
2 g i.v. 4-hourly
4–6
1.5 mg/kg i.v. or i.m. 12-hourly
4–6
3 Penicillin allergy or resistance
Vancomycin3 plus Gentamicin3 Teicoplanin plus Gentamicin3
15 mg/kg i.v. over 100 minutes 12-hourly 1 mg/kg i.v. 8–12-hourly 10 mg/kg i.v. 24-hourly
4–6
1 mg/kg i.v. 8–12-hourly
4–6
4 Penicillin allergy in regimens 1 and 2 (see enterococcal disease for regimen 3)
Enterococcal native-valve endocarditis 1 Low-level resistance to gentamicin (MIC < 500 mg/litre)
Emboli – clinically apparent systemic emboli occur in up to 40% of patients with infectious endocarditis. Most occur before diagnosis, and the incidence decreases after the initiation of antimicrobial therapy. Splenic emboli may present with left shoulder or left upper quadrant pain, renal emboli with flank pain or haematuria, and emboli in the extremities with pain or ischaemia. Neurological complications occur in 25% of patients with endocarditis and are the first sign of endocarditis in up to oneMEDICINE 33:4
4–6 4–6
• Prosthetic-valve endocarditis is treated for 6 weeks using these regimens • Do not give cephalosporins in enterococcal infectious endocarditis • 4 weeks’ treatment is recommended in patients with native-valve endocarditis and a shorter history of illness (< 3 months) who respond promptly to treatment • In regimen 3, streptomycin should be substituted for gentamicin when the strain is highly resistant to gentamicin but not to streptomycin • Do not use aminoglycosides when there is high-level resistance to both gentamicin and streptomycin • The combination of vancomycin and an aminoglycoside carries an increased risk of renal toxicity
half of patients.5 Cerebral emboli occur in 15–20% of patients with infectious endocarditis and most commonly involve the middle cerebral artery. Transient ischaemic attacks, haemorrhages, meningitis, brain abscess and mycotic aneurysms can also occur. Development of focal neurological symptoms, persistent headache or confusion should be investigated by MRI and magnetic resonance angiography or angiography, and patients with meningeal signs should undergo lumbar puncture. 68
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
Infecting organism
Antibiotic
Dose2
Duration (weeks)
Comments
• Addition of gentamicin,3 1 mg/kg i.v. 8–12-hourly, can be considered for the first 3–5 days of therapy in regimen 1, in patients with normal renal function • If penicillin sensitive, benzylpenicillin can be used in place of flucloxacillin • Cefazolin, 2 g i.v. 8-hourly, can be used in place of flucloxacillin in patients with non-severe penicillin allergy (i.e. rash) • Right-sided Staphylococcus aureus infectious endocarditis can often be treated with flucloxacillin plus gentamicin for 2 weeks • Teicoplanin is not recommended in staphylococcal infectious endocarditis • A 2-week course of vancomycin and gentamicin is not recommended in right-sided Staph. aureus infectious endocarditis
Staphylococcal native-valve endocarditis 1 Methicillin sensitive
Flucloxacillin or nafcillin
2 g i.v. 4-hourly
4–6
2 Methicillin resistant or penicillin allergy
Vancomycin3
15 mg/kg i.v. over 100 minutes 12-hourly
4–6
Flucloxacillin or nafcillin plus Gentamicin3 plus Rifampin
2 g i.v. 4-hourly
6
1 mg/kg i.v. 8-hourly
2
300–600 mg p.o. 12-hourly
6
Vancomycin3 plus Gentamicin3 plus Rifampin
15 mg/kg i.v. over 100 minutes 12-hourly 1 mg/kg i.v. 8-hourly
6
300–600 mg p.o. 12-hourly
6
Ceftriaxone Ampicillin plus Gentamicin3
2 g i.v. or i.m. 24-hourly 2 g i.v. 4-hourly
4 4
1 mg/kg i.v. 8-hourly
4
Staphylococcal prosthetic-valve endocarditis 1 Methicillin sensitive
2 Methicillin resistant or penicillin allergy
2
• Early surgical intervention is indicated in complicated prosthetic-valve endocarditis caused by Staph. aureus • Coagulase-negative Staphylococcus isolates are often resistant or heteroresistant to methicillin; methicillin sensitivity must be documented assiduously • Cefazolin, 2 g i.v. 8-hourly, can be used in place of flucloxacillin in patients with non-severe penicillin allergy (i.e. rash) • If penicillin sensitive, benzylpenicillin can be used in place of flucloxacillin • Rifampin increases the dose of warfarin required for effective anticoagulation • Teicoplanin is not recommended in staphylococcal prosthetic valve endocarditis • Fluoroquinolones should be considered when there is aminoglycoside resistance
HACEK organisms • Test organism for β-lactamase production; do not use ampicillin plus gentamicin if β-lactamase is produced
Some of the recommendations for treatment of staphylococcal infectious endocarditis stated here differ from those of the Working Party of the British Society for Antimicrobial Chemotherapy, which recommend addition of gentamicin, rifampin or fusidic acid to vancomycin in methicillin-resistant native-valve staphylococcal endocarditis, and a choice of one of these three agents in addition to a penicillinase-resistant penicillin or vancomycin in staphylococcal prosthetic-valve endocarditis. 2 Doses are for normal renal function and should be adjusted in renal impairment. 3 Drug levels must be monitored; aim for the following: vancomycin – 1-hour post-infusion 30 mg/litre, trough 10–15 mg/litre; gentamicin – 1-hour post-infusion 3–5 mg/litre, trough ≤ 1 mg/litre 1
5
MEDICINE 33:4
69
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
Diagnosis and investigations
Indications for surgical intervention
The clinical, laboratory and echocardiographic features that define infectious endocarditis are organized in a diagnostic scheme termed the ‘modified Duke criteria’ (Figure 3).6,7 These criteria provide a sensitive and specific approach to diagnosis. The presence of two major criteria, one major and three minor criteria or five minor criteria defines definite infectious endocarditis; the presence of one major and one minor or three minor criteria defines possible infectious endocarditis. A diagnosis of infectious endocarditis is unlikely to be rejected erroneously using the modified Duke criteria, if febrile patients who are suspected to have the condition undergo careful clinical evaluation. Nevertheless, clinicians should remember that these criteria were developed primarily for research purposes; thus, patients with a febrile illness suspected to be endocarditis who do not meet the criteria for definite infectious endocarditis but in whom there is no alternative diagnosis should be considered to have possible infectious endocarditis and should be treated for it. Blood cultures yielding organisms that are common contaminants (coagulase-negative staphylococi and diphtheroids) or rare causes of endocarditis (Gram-negative bacilli) should be viewed with caution as evidence supporting a diagnosis of infectious endocarditis, unless they are persistently positive or positive with genetically identical organisms, and alternative sites of infection are excluded. Blood cultures – more than 95% of blood cultures are positive in infectious endocarditis. Failure to culture an organism from the blood is most likely to be a consequence of prior antimicrobial therapy, but may result from infection with a fastidious organism or from improper microbiological techniques. In patients with suspected infectious endocarditis, three sets of blood cultures (two bottles each) should be drawn, with 10 ml of blood in each bottle. When infectious endocarditis is likely, blood cultures should be incubated for up to 3 weeks, to improve detection of fastidious organisms. When a patient who is clinically stable without evidence of sepsis, progressive valvular dysfunction or heart failure is known to have received antibiotics before blood cultures are taken, empirical antimicrobial therapy should be delayed until cultures off antibiotics can be obtained. A delay in treatment of 2–4 days should not affect the clinical course in stable patients, and may allow detection of the causative organism. Two additional sets of blood cultures should be obtained on two consecutive days when the initial cultures are negative after 24–48 hours. Empirical antimicrobial therapy should be initiated immediately after obtaining blood cultures in patients with an acute presentation or when clinical findings suggest a need for urgent surgical intervention. In patients with infectious endocarditis caused by resistant or virulent organisms, or who remain febrile on therapy, repeat blood cultures are indicated to assess the persistence of bacteraemia. Echocardiography – ideally, echocardiography should be performed in cases of suspected infectious endocarditis. It facilitates diagnosis through visualization of vegetations (Figure 4), and management through detection of complications such as valvular and ventricular dysfunction, abscesses and prosthetic valve dehiscence. The specificity of transthoracic echocardiography for detecting vegetations is high (91–98%), but the sensitivity is only 45–60%; visualization of prosthetic valves, perivalvular abscesses, leaflet perforation and fistulas is suboptimal.8 Nevertheless, in patients
Surgery indicated for optimal outcome • Development of moderate-to-severe heart failure as a result of valve dysfunction • Partial dehiscence of a prosthetic valve • Persistent bacteraemia despite optimal antimicrobial therapy • Absence of effective bactericidal treatment • Fungal infectious endocarditis • Relapse of prosthetic-valve endocarditis after optimal antimicrobial treatment • Persistent unexplained fever (≥ 10 days) in culture-negative prosthetic-valve endocarditis • Staphylococcus aureus prosthetic-valve endocarditis
MEDICINE 33:4
Surgery may be required for optimal outcome (in order of strength of recommendation) • Perivalvular extension of infection (myocardial, septal or annulus abscess or intracardiac fistula) • Poorly responsive Staphylococcus aureus endocarditis involving the aortic or mitral valve • Infectious endocarditis caused by highly antibiotic-resistant organisms • Persistent unexplained fever (≥ 10 days) in culture-negative native-valve endocarditis • Relapse of native-valve endocarditis after optimal antimicrobial therapy • Large (> 10 mm diameter) hypermobile vegetations 6
in whom the likelihood of infectious endocarditis is relatively low, good-quality negative transthoracic echocardiography is adequate to exclude the condition. In patients at intermediate or high risk of infectious endocarditis or its complications, negative transthoracic echocardiography should be followed by TOE, or TOE should be the initial investigation (the latter strategy is more cost-effective).9 TOE is the procedure of choice for imaging the pulmonic valve and in patients with suspected prosthetic-valve endocarditis or intracardiac complications. The sensitivity of TOE for identifying vegetations is 90–94%, but a negative result does not exclude infectious endocarditis in patients in whom the level of suspicion of infectious endocarditis is intermediate to high. In this situation, further evaluation, including repeat TOE and consideration of empirical treatment, is appropriate. Other laboratory tests – patients with suspected blood culturenegative infectious endocarditis not related to prior antimicrobial therapy should undergo serological tests for infection with Brucella spp., Legionella spp., C. burnetii, Chlamydia spp. and Bartonella spp. Circulating immune complex titre, C-reactive protein, ESR and other nonspecific tests are often abnormal, but seldom aid diagnosis or management of infectious endocarditis and can usually be omitted.
Management A detailed history and examination, full blood count, serum creatinine, liver function tests, urinalysis, ECG and chest radiography should be undertaken on admission as baseline observations for 70
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
Recommendations for prophylaxis of endocarditis in adults Procedure
Standard prophylaxis
Penicillin allergy or penicillin given more than once within the last month Clindamycin, 600 mg p.o. 1 hour before procedure
• Dental extraction, scaling or periodontal surgery, tonsillectomy, adenoidectomy, oesophageal dilatation, surgery or sclerotherapy under local or no anaesthesia
Amoxicillin, 3 g p.o. 1 hour before procedure
• Above procedures under general anaesthesia
Amoxicillin, 1 g i.v. 15 minutes before procedure
Clindamycin, 300 mg i.v. 15 minutes before procedure, followed by 150 mg p.o. or i.v. 6 hours later
• Above procedures under general anaesthesia in patients with a prosthetic valve or cyanotic heart disease, or patients who have had previous infectious endocarditis1
Amoxicillin, 1 g i.v., and gentamicin, 120 mg i.v. or i.m., 15 minutes before procedure
Vancomycin, 1 g i.v. 100 minutes before procedure, followed by gentamicin, 120 mg i.v. 15 minutes before procedure or Teicoplanin, 400 mg i.v., and gentamicin 120 mg i.v., 15 minutes before procedure or Clindamycin, 300 mg i.v. 15 minutes before procedure, followed by 150 mg p.o. or i.v. 6 hours later
• Genitourinary surgery/instrumentation2 • Obstetric and gynaecological procedures or gastrointestinal procedures in patients with a prosthetic valve or previous infectious endocarditis
Amoxicillin, 1 g i.v., and gentamicin, 120 mg i.v., 15 minutes before procedure
Vancomycin/gentamicin or teicoplanin/ gentamicin regimens as above; clindamycin is not effective against enterococci and should not be used in these procedures
Current recommendations in the USA are prophylaxis with amoxicillin alone for the above procedures in both high-risk and moderate-risk patients. Oral prophylaxis is used in all cases except when the patient is unable to take oral medication. Clindamycin alone is used in patients with penicillin allergy. 2 A negative urine culture should be documented before the procedure. 1
Source: Littler W A et al. Lancet 1997; 350: 1100 and Simmons N A. J Antimicrob Chemother 1993; 31: 437–8.
7
enterococci. Although combination therapy is usually not essential to eradicate streptococci, it is required to eradicate enterococci in infectious endocarditis because enterococci are inhibited but not killed by penicillin, ampicillin and vancomycin. Enterococci with high-level resistance to gentamicin (MIC ≥ 500 mg/litre) should be tested for high-level resistance to streptomycin (MIC ≥ 2000 mg/litre). When this is absent, streptomycin should be used in lieu of gentamicin. Neither gentamicin nor streptomycin should be used when there is high-level resistance to both. Other aminoglycosides are ineffective against enterococci. Addition of gentamicin to a penicillinase-resistant penicillin (flucloxacillin, nafcillin) is not essential in the treatment of methicillin-sensitive staphylococcal native-valve endocarditis, but may accelerate control of infection and can be considered in patients with normal renal function. No clinical data indicate a benefit with gentamicin in the treatment of methicillin-resistant staphylococcal native-valve endocarditis. Also, addition of rifampin has not been shown to be beneficial in the treatment of staphylococcal native-valve endocarditis. In the treatment of staphylococcal prosthetic-valve endocarditis, US authorities recommend use of both gentamicin and rifampin
subsequent management. Prolonged PR interval or intraventricular conduction disturbances on ECG can indicate a valve ring abscess, particularly in aortic valve infectious endocarditis. New conduction abnormalities are highly specific for paravalvular infection, but are not sensitive.10 Antimicrobial therapy: recommendations for organism-specific antimicrobial therapy are shown in Figure 5.11,12 When empirical treatment is necessary, risk factors for certain infecting microorganisms and local bacterial resistance patterns should be considered. Antimicrobial sensitivities and the minimum inhibitory concentration (MIC) of selected antimicrobials should be obtained to determine appropriate therapy. Measurement of the minimum bactericidal concentration is no longer routinely recommended. Parenteral therapy is usually necessary to obtain high and sustained serum concentrations of bactericidal antimicrobial drugs. Patients who report penicillin allergy should be questioned carefully about the nature of the reaction, because alternative agents may be less effective. Addition of gentamicin to penicillin, ampicillin and vancomycin generally results in synergistic killing of streptococci and MEDICINE 33:4
71
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
in addition to a penicillinase-resistant penicillin or vancomycin to help protect against the emergence of rifampin resistance during treatment. When dealing with unusual organisms, or typical organisms with high levels of resistance, consultation with experts in infectious diseases and in microbiology is recommended. Adverse reactions to antibiotics are common in infectious endocarditis and should be sought by clinical and laboratory monitoring. • A full blood count should be obtained weekly in patients receiving a β-lactam antibiotic or vancomycin, to detect drug-induced cytopenias. • Liver function tests should be undertaken weekly in patients receiving flucloxacillin or nafcillin, to detect drug-induced hepatitis. • Serum creatinine levels should be determined at least weekly in patients receiving β-lactam antibiotics, vancomycin or aminoglycosides. • Serum concentrations of vancomycin and aminoglycosides should be measured, to facilitate appropriate dosing. Persistent fever beyond 7 days requires evaluation and may indicate failure of antimicrobial therapy or the presence of a myocardial abscess, focal extracardiac infection (splenic or renal abscess), emboli, hypersensitivity to an antimicrobial agent or a complication unrelated to infectious endocarditis.
heart disease except secundum atrial septal defect, hypertrophic cardiomyopathy) for the development of infectious endocarditis. Procedures requiring prophylaxis and specific regimens are shown in Figure 7.16,17
REFERENCES 1 Karchmer A W. Infective endocarditis. In: Braunwald E, ed. Heart disease: a textbook of cardiovascular medicine. 7th ed. London: Saunders, 2005: 1633–56. 2 Bayer A S, Bolger A F, Taubert K A et al. Diagnosis and management of infective endocarditis and its complications. Circulation 1998; 98: 2936–48. 3 Fowler V G Jr, Li J, Corey G R et al. Role of echocardiography in evaluation of patients with Staphylococcus aureus bacteremia: experience in 103 patients. J Am Coll Cardiol 1997; 30: 1072–8. 4 Gonzalez Quintela A, Candela H J, Vidal C et al. Non-bacterial thrombotic endocarditis in cancer patients. Acta Cardiol 1991; 46: 1–9. 5 Heiro M, Nikoskelainen J, Engblom E et al. Neurologic manifestations of infective endocarditis: a 17-year experience in a teaching hospital in Finland. Arch Intern Med 2000; 160: 2781–7. 6 Durack D T, Lukes A S, Bright D K. New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Am J Med 1994; 96: 200–9. 7 Li J S, Sexton D J, Mick N et al. Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis. Clin Infect Dis 2000; 30: 633–8. 8 Evangelista A, Gonzalez-Alujas M T. Echocardiography in infective endocarditis. Heart 2004; 90: 614–17. 9 Heidenreich P A, Masoudi F A, Maini B et al. Echocardiography in patients with suspected endocarditis: a cost-effectiveness analysis. Am J Med 1999; 107: 198–208. 10 Meine T J, Nettles R E, Anderson D J et al. Cardiac conduction abnormalities in endocarditis defined by the Duke criteria. Am Heart J 2001; 142: 280–5. 11 Elliott T S J, Foweraker J, Gould F K et al. Guidelines for the antibiotic treatment of endocarditis in adults: report of the Working Party of the British Society for Antimicrobial Chemotherapy. J Antimicrob Chemother 2004; 54: 971–81. 12 Wilson W R, Karchmer A W, Dajani A S et al. Antibiotic treatment of adults with infective endocarditis due to streptococci, enterococci, staphylococci, and HACEK microorganisms. JAMA 1995; 274: 1706–13. 13 Vikram H R, Buenconsejo J, Hastun R et al. Impact of valve surgery on 6-month mortality in adults with complicated left-sided native valve endocarditis. JAMA 2003; 290: 3207–14. 14 Olaison L, Pettersson G. Indications for surgical intervention in infective endocarditis. Cardiol Clin 2003; 21: 235–51. 15 Gillinov A M, Shah R V, Curtis W E et al. Valve replacement in patients with endocarditis and acute neurologic deficit. Ann Thorac Surg 1996; 61: 1125–9. 16 Littler W A, McGowan D A, Shanson D C. Changes in recommendations about amoxycillin prophylaxis for prevention of endocarditis. British Society for Antimicrobial Chemotherapy Endocarditis Working Party. Lancet 1997; 350: 1100. 17 Simmons N A. Recommendations for endocarditis prophylaxis. Endocarditis Working Party for Antimicrobial Chemotherapy. J Antimicrob Chemother 1993; 31: 437–8.
Surgical treatment: cardiac surgical intervention has an important role in the management of the intracardiac complications of infectious endocarditis and has led to a decrease in mortality, particularly when correcting anatomical complications that have caused congestive heart failure.13 Indications for cardiac surgery are listed in Figure 6.14 A cardiovascular surgeon should be consulted early in treatment, to enable surgery to be undertaken expediently, if needed. When valvular dysfunction results in heart failure or antimicrobial therapy fails, delaying surgery in favour of continued antimicrobial therapy leads to increased mortality. Recurrent infectious endocarditis involving prosthetic valves placed during active infectious endocarditis occurs in 2–3% of patients. This is lower than the mortality seen in similar patients treated medically. Timing of surgical intervention requires modification in patients who have suffered a recent neurological injury. If possible, surgery should be delayed for 2–3 weeks after an embolic infarct and for 1 month or more after intracerebral haemorrhage.15 In patients with negative valve cultures, preoperative plus postoperative antimicrobial therapy should at least equal a full course of recommended treatment. Patients with positive valve cultures, and most patients with prosthetic-valve endocarditis, should receive a full course of treatment after surgery.
Prophylaxis Although antimicrobial prophylaxis to prevent infectious endocarditis has not been proved effective, it remains the standard of care. Prophylaxis is recommended in patients with cardiac conditions who are considered to be at high risk (prosthetic valves, cyanotic congenital heart disease, previous infectious endocarditis) or at moderate risk (valvular heart disease including aortic or mitral regurgitation, aortic stenosis, mitral valve prolapse with mitral regurgitation, bicuspid aortic valve, non-cyanotic congenital
MEDICINE 33:4
72
© 2005 The Medicine Publishing Company Ltd
Endocarditis Sara E Cosgrove Adolf W Karchmer
Infectious endocarditis results from microbial infection of the endothelial surface of the heart. The characteristic lesion, a vegetation (Figure 1), is a mass of platelets and fibrin within which circulating micro-organisms have adhered and enmeshed.
1 Surgical specimen of a removed mitral valve showing vegetations (arrows).
Epidemiology and microbiology
Most cases of native-valve endocarditis are caused by viridans streptococci, Streptococcus bovis, Staphylococcus aureus or enterococci. In patients with community-acquired native-valve endocarditis who are not injecting drug-users, about one-third of cases are caused by streptococci, one-third by Staph. aureus and one-third by other organisms, including the HACEK group (Haemophilus spp., Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella spp. and Kingella kingae) of fastidious Gram-negative coccobacilli. When actively sought, Bartonella spp. have been found to cause 3% of cases, and Coxiella burnetii has caused a similar percentage of cases in selected areas. Both organisms are difficult to grow in blood cultures, and the diagnosis may require serological evaluation.2 The incidence of infectious endocarditis in injecting drug-users is 2–5%/year. Infection commonly involves the right heart valves and often occurs in previously normal valves. More than 50% of cases are caused by Staph. aureus. Nosocomial native-valve endocarditis is an emerging problem. Staph. aureus is the most common pathogen (55%), followed by enterococci (16%) and coagulase-negative staphylococci (10%). Central venous catheters are a major risk factor. The prevalence
Native-value endocarditis: in developed countries, the annual incidence of infectious endocarditis remained stable at 1.7–4.2/100,000 from 1950 to 1987, but recent data suggest that the incidence may be increasing; studies from Sweden and the USA report annual incidences of 5.9/100,000 and 11.6/100,000.1 The most likely explanation is the increased prevalence of predisposing conditions in an ageing population; 55–75% of patients with native-valve endocarditis have such conditions, including: • rheumatic valve disease • degenerative heart disease • mitral valve prolapse • congenital heart disease • hypertrophic cardiomyopathy • intravenous drug abuse • a prosthetic heart valve. The mean age of patients with infectious endocarditis has increased in parallel with a decline in rheumatic valve disease and an increase in degenerative valvulopathy as the substrate for infection. Most cases of infectious endocarditis in the over-60s are associated with degenerative heart disease. Mitral valve prolapse with mitral regurgitation is associated with only a 4–8-fold increased risk of infectious endocarditis compared with normal valves; however, because of its prevalence, it accounts for 7–30% of cases of nativevalve endocarditis. In young adults with infectious endocarditis, 20% of cases occur in those with congenital abnormalities, who now survive into adulthood as a result of corrective surgery. Bicuspid aortic valve is the most common congenital heart lesion and is emerging as an important predisposing cause of infectious endocarditis in the fourth and fifth decades of life.
Sara E Cosgrove is Assistant Professor of Medicine at Johns Hopkins University School of Medicine, Baltimore, USA. Conflict of interests: none declared. Adolf W Karchmer is Professor of Medicine at Harvard Medical School and Chief of the Division of Infectious Diseases at the Beth Israel Deaconess Medical Center, Boston, USA. Conflict of interests: none declared.
MEDICINE 33:4
2 Osler nodes on the hand and the pulp of the finger (arrows) in a young woman with rheumatic heart disease and aortic valve endocarditis caused by Staphylococcus aureus.
66
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
of infectious endocarditis following Staph. aureus catheter-related bacteraemia averages 6.1%, but was 23% in a study that systematically examined patients using transoesophageal echocardiography (TOE).3 Patients with Staph. aureus catheter-related bacteraemia should undergo TOE to exclude infectious endocarditis.
during the first year after valve surgery and almost 6% after 5 years. The risk is greatest during the first 6 months. Methicillin-resistant coagulase-negative staphylococci are responsible for most cases seen within the first year after valve placement. After 12 months, the microbiology of prosthetic-valve endocarditis resembles that of the native-valve form, though the incidence of coagulase-negative staphylococci is greater (10%).
Prosthetic-valve endocarditis: in 10–15% of patients with infectious endocarditis, infection develops on prosthetic valves. The cumulative prevalence of prosthetic-valve endocarditis is 3%
Marantic endocarditis (also known as thrombotic endocarditis) occurs in patients with malignancy, particularly adenocarcinoma.4 Platelets and fibrin form vegetations on heart valves as a consequence of a hypercoagulable state. Patients present with fever and embolization of the spleen, kidneys and CNS, but blood cultures are negative.
Modified Duke clinical criteria for diagnosis of infectious endocarditis Major criteria Positive blood culture • Two separate blood cultures yielding organisms that typically cause infectious endocarditis (viridans streptococci, Streptococcus bovis, HACEK,1 Staphylococcus aureus or community-acquired enterococci without a primary focus) • Persistently positive blood cultures (defined as positive blood cultures drawn more than 12 hours apart or all of three or a majority of four or more separate positive blood cultures, the first and last drawn at least 1 hour apart) • Single positive blood culture for Coxiella burnetii or antiphase I IgG antibody titre > 1:800 Evidence of endocardial involvement • Echocardiography positive for infectious endocarditis2 Oscillating intracardiac mass on a valve or supporting structure in the path of regurgitant jets or on implanted material in the absence of an alternative anatomical explanation Abscess New partial dehiscence of a prosthetic valve • New valvular regurgitation (change in pre-existing murmur is not adequate)
Clinical features Presentation of infectious endocarditis ranges from an acute syndrome of systemic toxicity and rapid progression with intracardiac and extracardiac complications, to an indolent illness that develops over weeks with low-grade fever and minimal cardiac dysfunction. Fever is the most common symptom (90% of patients), but may be muted or absent in the elderly, in those with chronic renal failure, and in patients who have received prior antibiotic therapy. Other nonspecific symptoms include chills, anorexia, weight loss, myalgia, cough and back pain. Heart murmurs are found in 85% of patients with native-valve endocarditis and most often represent the predisposing cardiac lesion. A new or changing murmur is seen in only 10–40% of patients; this may signal rapid valvular damage and impending heart failure. Peripheral signs – the classic peripheral Oslerian manifestations of infectious endocarditis (petechiae, splinter haemorrhages, Osler’s nodes – Figure 2, Janeway lesions, Roth spots) are not pathognomonic for infectious endocarditis and represent the sequelae of prolonged disease. They are now seen uncommonly, because most patients seek medical attention earlier in the course of the disease.
Minor criteria • Predisposing heart condition or intravenous drug use • Fever (≥ 38.0°C) • Vascular phenomenon (major arterial emboli, septic pulmonary infarcts, mycotic aneurysm, intracranial haemorrhage, conjunctival haemorrhages or Janeway lesions) • Immunological phenomenon (glomerulonephritis, Osler’s nodes, Roth spots or rheumatoid factor) • Microbiological evidence (positive blood culture, but less than major criterion,3 or serological evidence of active infection with an organism consistent with infectious endocarditis) Haemophilus spp., Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella corrodens and Kingella kingae 2 TOE recommended in patients with prosthetic valves or complicated endocarditis and who have possible endocarditis 3 Excludes a single positive culture for coagulase-negative staphylococci or organisms that do not cause infectious endocarditis 1
Source: Durack D T et al. Am J Med 1994; 96: 200–9 and Li J S et al. Clin Infect Dis 2000; 30: 633–8.
4 Echocardiogram of a mitral valve showing a large, round vegetation (arrow).
3
MEDICINE 33:4
67
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
Recommended antibiotic treatment in infectious endocarditis1 Infecting organism
Antibiotic
Dose2
Duration (weeks)
Comments
• Prosthetic-valve endocarditis is treated with 6 weeks’ benzylpenicillin (or vancomycin) and 2 weeks’ gentamicin (regimens 1 and 2) or 6 weeks’ gentamicin (regimen 3) • A 2-week regimen is appropriate when heart failure, aortic insufficiency and conduction abnormalities are absent, embolic disease is absent, the infection is on native valves, vegetations are < 5 mm on echocardiography, and there is a clinical response in 5 days • In regimen 1, a higher dose of penicillin (11.2 g daily) is recommended by most experts in the USA; in addition, ceftriaxone, 2 g i.v. or i.m. 24-hourly, can be substituted for benzylpenicillin • Regimen 3 is also preferred in patients infected with nutritionally variant streptococci (Abiotrophia spp.)
Viridans streptococcal and Streptococcus bovis native-valve endocarditis 1 Fully sensitive to penicillin (MIC ≤ 0.1 mg/litre)
Benzylpenicillin plus Gentamicin3 Ceftriaxone Benzylpenicillin
1.2–1.9 g i.v. 4-hourly
2
1 mg/kg i.v. 8–12-hourly 2 g i.v. or i.m. 24-hourly 1.2–1.9 g i.v. 4-hourly
2 4 4
2 Relatively resistant to penicillin (MIC > 0.1 to < 0.5 mg/litre)
Benzylpenicillin plus Gentamicin3
2.4 g i.v. 4-hourly
4
1 mg/kg i.v. 8–12-hourly
2
3 Penicillin (MIC ≥ 0.5 mg/litre)
Benzylpenicillin plus Gentamicin3
2.4–3 g i.v. 4 hourly
4–6
1 mg/kg i.v. 8–12-hourly
4–6
Vancomycin3
15 mg/kg i.v. over 100 minutes 12-hourly
4
Ampicillin or amoxicillin plus Gentamicin3
2 g i.v. 4-hourly
4–6
1 mg/kg i.v. 8–12 hourly
4–6
2 Highly resistant to gentamicin (MIC ≥ 500 mg/litre) but low-level resistance to streptomycin (MIC < 2000 mg/litre)
Ampicillin or amoxicillin plus Streptomycin3
2 g i.v. 4-hourly
4–6
1.5 mg/kg i.v. or i.m. 12-hourly
4–6
3 Penicillin allergy or resistance
Vancomycin3 plus Gentamicin3 Teicoplanin plus Gentamicin3
15 mg/kg i.v. over 100 minutes 12-hourly 1 mg/kg i.v. 8–12-hourly 10 mg/kg i.v. 24-hourly
4–6
1 mg/kg i.v. 8–12-hourly
4–6
4 Penicillin allergy in regimens 1 and 2 (see enterococcal disease for regimen 3)
Enterococcal native-valve endocarditis 1 Low-level resistance to gentamicin (MIC < 500 mg/litre)
Emboli – clinically apparent systemic emboli occur in up to 40% of patients with infectious endocarditis. Most occur before diagnosis, and the incidence decreases after the initiation of antimicrobial therapy. Splenic emboli may present with left shoulder or left upper quadrant pain, renal emboli with flank pain or haematuria, and emboli in the extremities with pain or ischaemia. Neurological complications occur in 25% of patients with endocarditis and are the first sign of endocarditis in up to oneMEDICINE 33:4
4–6 4–6
• Prosthetic-valve endocarditis is treated for 6 weeks using these regimens • Do not give cephalosporins in enterococcal infectious endocarditis • 4 weeks’ treatment is recommended in patients with native-valve endocarditis and a shorter history of illness (< 3 months) who respond promptly to treatment • In regimen 3, streptomycin should be substituted for gentamicin when the strain is highly resistant to gentamicin but not to streptomycin • Do not use aminoglycosides when there is high-level resistance to both gentamicin and streptomycin • The combination of vancomycin and an aminoglycoside carries an increased risk of renal toxicity
half of patients.5 Cerebral emboli occur in 15–20% of patients with infectious endocarditis and most commonly involve the middle cerebral artery. Transient ischaemic attacks, haemorrhages, meningitis, brain abscess and mycotic aneurysms can also occur. Development of focal neurological symptoms, persistent headache or confusion should be investigated by MRI and magnetic resonance angiography or angiography, and patients with meningeal signs should undergo lumbar puncture. 68
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
Infecting organism
Antibiotic
Dose2
Duration (weeks)
Comments
• Addition of gentamicin,3 1 mg/kg i.v. 8–12-hourly, can be considered for the first 3–5 days of therapy in regimen 1, in patients with normal renal function • If penicillin sensitive, benzylpenicillin can be used in place of flucloxacillin • Cefazolin, 2 g i.v. 8-hourly, can be used in place of flucloxacillin in patients with non-severe penicillin allergy (i.e. rash) • Right-sided Staphylococcus aureus infectious endocarditis can often be treated with flucloxacillin plus gentamicin for 2 weeks • Teicoplanin is not recommended in staphylococcal infectious endocarditis • A 2-week course of vancomycin and gentamicin is not recommended in right-sided Staph. aureus infectious endocarditis
Staphylococcal native-valve endocarditis 1 Methicillin sensitive
Flucloxacillin or nafcillin
2 g i.v. 4-hourly
4–6
2 Methicillin resistant or penicillin allergy
Vancomycin3
15 mg/kg i.v. over 100 minutes 12-hourly
4–6
Flucloxacillin or nafcillin plus Gentamicin3 plus Rifampin
2 g i.v. 4-hourly
6
1 mg/kg i.v. 8-hourly
2
300–600 mg p.o. 12-hourly
6
Vancomycin3 plus Gentamicin3 plus Rifampin
15 mg/kg i.v. over 100 minutes 12-hourly 1 mg/kg i.v. 8-hourly
6
300–600 mg p.o. 12-hourly
6
Ceftriaxone Ampicillin plus Gentamicin3
2 g i.v. or i.m. 24-hourly 2 g i.v. 4-hourly
4 4
1 mg/kg i.v. 8-hourly
4
Staphylococcal prosthetic-valve endocarditis 1 Methicillin sensitive
2 Methicillin resistant or penicillin allergy
2
• Early surgical intervention is indicated in complicated prosthetic-valve endocarditis caused by Staph. aureus • Coagulase-negative Staphylococcus isolates are often resistant or heteroresistant to methicillin; methicillin sensitivity must be documented assiduously • Cefazolin, 2 g i.v. 8-hourly, can be used in place of flucloxacillin in patients with non-severe penicillin allergy (i.e. rash) • If penicillin sensitive, benzylpenicillin can be used in place of flucloxacillin • Rifampin increases the dose of warfarin required for effective anticoagulation • Teicoplanin is not recommended in staphylococcal prosthetic valve endocarditis • Fluoroquinolones should be considered when there is aminoglycoside resistance
HACEK organisms • Test organism for β-lactamase production; do not use ampicillin plus gentamicin if β-lactamase is produced
Some of the recommendations for treatment of staphylococcal infectious endocarditis stated here differ from those of the Working Party of the British Society for Antimicrobial Chemotherapy, which recommend addition of gentamicin, rifampin or fusidic acid to vancomycin in methicillin-resistant native-valve staphylococcal endocarditis, and a choice of one of these three agents in addition to a penicillinase-resistant penicillin or vancomycin in staphylococcal prosthetic-valve endocarditis. 2 Doses are for normal renal function and should be adjusted in renal impairment. 3 Drug levels must be monitored; aim for the following: vancomycin – 1-hour post-infusion 30 mg/litre, trough 10–15 mg/litre; gentamicin – 1-hour post-infusion 3–5 mg/litre, trough ≤ 1 mg/litre 1
5
MEDICINE 33:4
69
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
Diagnosis and investigations
Indications for surgical intervention
The clinical, laboratory and echocardiographic features that define infectious endocarditis are organized in a diagnostic scheme termed the ‘modified Duke criteria’ (Figure 3).6,7 These criteria provide a sensitive and specific approach to diagnosis. The presence of two major criteria, one major and three minor criteria or five minor criteria defines definite infectious endocarditis; the presence of one major and one minor or three minor criteria defines possible infectious endocarditis. A diagnosis of infectious endocarditis is unlikely to be rejected erroneously using the modified Duke criteria, if febrile patients who are suspected to have the condition undergo careful clinical evaluation. Nevertheless, clinicians should remember that these criteria were developed primarily for research purposes; thus, patients with a febrile illness suspected to be endocarditis who do not meet the criteria for definite infectious endocarditis but in whom there is no alternative diagnosis should be considered to have possible infectious endocarditis and should be treated for it. Blood cultures yielding organisms that are common contaminants (coagulase-negative staphylococi and diphtheroids) or rare causes of endocarditis (Gram-negative bacilli) should be viewed with caution as evidence supporting a diagnosis of infectious endocarditis, unless they are persistently positive or positive with genetically identical organisms, and alternative sites of infection are excluded. Blood cultures – more than 95% of blood cultures are positive in infectious endocarditis. Failure to culture an organism from the blood is most likely to be a consequence of prior antimicrobial therapy, but may result from infection with a fastidious organism or from improper microbiological techniques. In patients with suspected infectious endocarditis, three sets of blood cultures (two bottles each) should be drawn, with 10 ml of blood in each bottle. When infectious endocarditis is likely, blood cultures should be incubated for up to 3 weeks, to improve detection of fastidious organisms. When a patient who is clinically stable without evidence of sepsis, progressive valvular dysfunction or heart failure is known to have received antibiotics before blood cultures are taken, empirical antimicrobial therapy should be delayed until cultures off antibiotics can be obtained. A delay in treatment of 2–4 days should not affect the clinical course in stable patients, and may allow detection of the causative organism. Two additional sets of blood cultures should be obtained on two consecutive days when the initial cultures are negative after 24–48 hours. Empirical antimicrobial therapy should be initiated immediately after obtaining blood cultures in patients with an acute presentation or when clinical findings suggest a need for urgent surgical intervention. In patients with infectious endocarditis caused by resistant or virulent organisms, or who remain febrile on therapy, repeat blood cultures are indicated to assess the persistence of bacteraemia. Echocardiography – ideally, echocardiography should be performed in cases of suspected infectious endocarditis. It facilitates diagnosis through visualization of vegetations (Figure 4), and management through detection of complications such as valvular and ventricular dysfunction, abscesses and prosthetic valve dehiscence. The specificity of transthoracic echocardiography for detecting vegetations is high (91–98%), but the sensitivity is only 45–60%; visualization of prosthetic valves, perivalvular abscesses, leaflet perforation and fistulas is suboptimal.8 Nevertheless, in patients
Surgery indicated for optimal outcome • Development of moderate-to-severe heart failure as a result of valve dysfunction • Partial dehiscence of a prosthetic valve • Persistent bacteraemia despite optimal antimicrobial therapy • Absence of effective bactericidal treatment • Fungal infectious endocarditis • Relapse of prosthetic-valve endocarditis after optimal antimicrobial treatment • Persistent unexplained fever (≥ 10 days) in culture-negative prosthetic-valve endocarditis • Staphylococcus aureus prosthetic-valve endocarditis
MEDICINE 33:4
Surgery may be required for optimal outcome (in order of strength of recommendation) • Perivalvular extension of infection (myocardial, septal or annulus abscess or intracardiac fistula) • Poorly responsive Staphylococcus aureus endocarditis involving the aortic or mitral valve • Infectious endocarditis caused by highly antibiotic-resistant organisms • Persistent unexplained fever (≥ 10 days) in culture-negative native-valve endocarditis • Relapse of native-valve endocarditis after optimal antimicrobial therapy • Large (> 10 mm diameter) hypermobile vegetations 6
in whom the likelihood of infectious endocarditis is relatively low, good-quality negative transthoracic echocardiography is adequate to exclude the condition. In patients at intermediate or high risk of infectious endocarditis or its complications, negative transthoracic echocardiography should be followed by TOE, or TOE should be the initial investigation (the latter strategy is more cost-effective).9 TOE is the procedure of choice for imaging the pulmonic valve and in patients with suspected prosthetic-valve endocarditis or intracardiac complications. The sensitivity of TOE for identifying vegetations is 90–94%, but a negative result does not exclude infectious endocarditis in patients in whom the level of suspicion of infectious endocarditis is intermediate to high. In this situation, further evaluation, including repeat TOE and consideration of empirical treatment, is appropriate. Other laboratory tests – patients with suspected blood culturenegative infectious endocarditis not related to prior antimicrobial therapy should undergo serological tests for infection with Brucella spp., Legionella spp., C. burnetii, Chlamydia spp. and Bartonella spp. Circulating immune complex titre, C-reactive protein, ESR and other nonspecific tests are often abnormal, but seldom aid diagnosis or management of infectious endocarditis and can usually be omitted.
Management A detailed history and examination, full blood count, serum creatinine, liver function tests, urinalysis, ECG and chest radiography should be undertaken on admission as baseline observations for 70
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
Recommendations for prophylaxis of endocarditis in adults Procedure
Standard prophylaxis
Penicillin allergy or penicillin given more than once within the last month Clindamycin, 600 mg p.o. 1 hour before procedure
• Dental extraction, scaling or periodontal surgery, tonsillectomy, adenoidectomy, oesophageal dilatation, surgery or sclerotherapy under local or no anaesthesia
Amoxicillin, 3 g p.o. 1 hour before procedure
• Above procedures under general anaesthesia
Amoxicillin, 1 g i.v. 15 minutes before procedure
Clindamycin, 300 mg i.v. 15 minutes before procedure, followed by 150 mg p.o. or i.v. 6 hours later
• Above procedures under general anaesthesia in patients with a prosthetic valve or cyanotic heart disease, or patients who have had previous infectious endocarditis1
Amoxicillin, 1 g i.v., and gentamicin, 120 mg i.v. or i.m., 15 minutes before procedure
Vancomycin, 1 g i.v. 100 minutes before procedure, followed by gentamicin, 120 mg i.v. 15 minutes before procedure or Teicoplanin, 400 mg i.v., and gentamicin 120 mg i.v., 15 minutes before procedure or Clindamycin, 300 mg i.v. 15 minutes before procedure, followed by 150 mg p.o. or i.v. 6 hours later
• Genitourinary surgery/instrumentation2 • Obstetric and gynaecological procedures or gastrointestinal procedures in patients with a prosthetic valve or previous infectious endocarditis
Amoxicillin, 1 g i.v., and gentamicin, 120 mg i.v., 15 minutes before procedure
Vancomycin/gentamicin or teicoplanin/ gentamicin regimens as above; clindamycin is not effective against enterococci and should not be used in these procedures
Current recommendations in the USA are prophylaxis with amoxicillin alone for the above procedures in both high-risk and moderate-risk patients. Oral prophylaxis is used in all cases except when the patient is unable to take oral medication. Clindamycin alone is used in patients with penicillin allergy. 2 A negative urine culture should be documented before the procedure. 1
Source: Littler W A et al. Lancet 1997; 350: 1100 and Simmons N A. J Antimicrob Chemother 1993; 31: 437–8.
7
enterococci. Although combination therapy is usually not essential to eradicate streptococci, it is required to eradicate enterococci in infectious endocarditis because enterococci are inhibited but not killed by penicillin, ampicillin and vancomycin. Enterococci with high-level resistance to gentamicin (MIC ≥ 500 mg/litre) should be tested for high-level resistance to streptomycin (MIC ≥ 2000 mg/litre). When this is absent, streptomycin should be used in lieu of gentamicin. Neither gentamicin nor streptomycin should be used when there is high-level resistance to both. Other aminoglycosides are ineffective against enterococci. Addition of gentamicin to a penicillinase-resistant penicillin (flucloxacillin, nafcillin) is not essential in the treatment of methicillin-sensitive staphylococcal native-valve endocarditis, but may accelerate control of infection and can be considered in patients with normal renal function. No clinical data indicate a benefit with gentamicin in the treatment of methicillin-resistant staphylococcal native-valve endocarditis. Also, addition of rifampin has not been shown to be beneficial in the treatment of staphylococcal native-valve endocarditis. In the treatment of staphylococcal prosthetic-valve endocarditis, US authorities recommend use of both gentamicin and rifampin
subsequent management. Prolonged PR interval or intraventricular conduction disturbances on ECG can indicate a valve ring abscess, particularly in aortic valve infectious endocarditis. New conduction abnormalities are highly specific for paravalvular infection, but are not sensitive.10 Antimicrobial therapy: recommendations for organism-specific antimicrobial therapy are shown in Figure 5.11,12 When empirical treatment is necessary, risk factors for certain infecting microorganisms and local bacterial resistance patterns should be considered. Antimicrobial sensitivities and the minimum inhibitory concentration (MIC) of selected antimicrobials should be obtained to determine appropriate therapy. Measurement of the minimum bactericidal concentration is no longer routinely recommended. Parenteral therapy is usually necessary to obtain high and sustained serum concentrations of bactericidal antimicrobial drugs. Patients who report penicillin allergy should be questioned carefully about the nature of the reaction, because alternative agents may be less effective. Addition of gentamicin to penicillin, ampicillin and vancomycin generally results in synergistic killing of streptococci and MEDICINE 33:4
71
© 2005 The Medicine Publishing Company Ltd
CARDIOVASCULAR INFECTIONS
in addition to a penicillinase-resistant penicillin or vancomycin to help protect against the emergence of rifampin resistance during treatment. When dealing with unusual organisms, or typical organisms with high levels of resistance, consultation with experts in infectious diseases and in microbiology is recommended. Adverse reactions to antibiotics are common in infectious endocarditis and should be sought by clinical and laboratory monitoring. • A full blood count should be obtained weekly in patients receiving a β-lactam antibiotic or vancomycin, to detect drug-induced cytopenias. • Liver function tests should be undertaken weekly in patients receiving flucloxacillin or nafcillin, to detect drug-induced hepatitis. • Serum creatinine levels should be determined at least weekly in patients receiving β-lactam antibiotics, vancomycin or aminoglycosides. • Serum concentrations of vancomycin and aminoglycosides should be measured, to facilitate appropriate dosing. Persistent fever beyond 7 days requires evaluation and may indicate failure of antimicrobial therapy or the presence of a myocardial abscess, focal extracardiac infection (splenic or renal abscess), emboli, hypersensitivity to an antimicrobial agent or a complication unrelated to infectious endocarditis.
heart disease except secundum atrial septal defect, hypertrophic cardiomyopathy) for the development of infectious endocarditis. Procedures requiring prophylaxis and specific regimens are shown in Figure 7.16,17
REFERENCES 1 Karchmer A W. Infective endocarditis. In: Braunwald E, ed. Heart disease: a textbook of cardiovascular medicine. 7th ed. London: Saunders, 2005: 1633–56. 2 Bayer A S, Bolger A F, Taubert K A et al. Diagnosis and management of infective endocarditis and its complications. Circulation 1998; 98: 2936–48. 3 Fowler V G Jr, Li J, Corey G R et al. Role of echocardiography in evaluation of patients with Staphylococcus aureus bacteremia: experience in 103 patients. J Am Coll Cardiol 1997; 30: 1072–8. 4 Gonzalez Quintela A, Candela H J, Vidal C et al. Non-bacterial thrombotic endocarditis in cancer patients. Acta Cardiol 1991; 46: 1–9. 5 Heiro M, Nikoskelainen J, Engblom E et al. Neurologic manifestations of infective endocarditis: a 17-year experience in a teaching hospital in Finland. Arch Intern Med 2000; 160: 2781–7. 6 Durack D T, Lukes A S, Bright D K. New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Am J Med 1994; 96: 200–9. 7 Li J S, Sexton D J, Mick N et al. Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis. Clin Infect Dis 2000; 30: 633–8. 8 Evangelista A, Gonzalez-Alujas M T. Echocardiography in infective endocarditis. Heart 2004; 90: 614–17. 9 Heidenreich P A, Masoudi F A, Maini B et al. Echocardiography in patients with suspected endocarditis: a cost-effectiveness analysis. Am J Med 1999; 107: 198–208. 10 Meine T J, Nettles R E, Anderson D J et al. Cardiac conduction abnormalities in endocarditis defined by the Duke criteria. Am Heart J 2001; 142: 280–5. 11 Elliott T S J, Foweraker J, Gould F K et al. Guidelines for the antibiotic treatment of endocarditis in adults: report of the Working Party of the British Society for Antimicrobial Chemotherapy. J Antimicrob Chemother 2004; 54: 971–81. 12 Wilson W R, Karchmer A W, Dajani A S et al. Antibiotic treatment of adults with infective endocarditis due to streptococci, enterococci, staphylococci, and HACEK microorganisms. JAMA 1995; 274: 1706–13. 13 Vikram H R, Buenconsejo J, Hastun R et al. Impact of valve surgery on 6-month mortality in adults with complicated left-sided native valve endocarditis. JAMA 2003; 290: 3207–14. 14 Olaison L, Pettersson G. Indications for surgical intervention in infective endocarditis. Cardiol Clin 2003; 21: 235–51. 15 Gillinov A M, Shah R V, Curtis W E et al. Valve replacement in patients with endocarditis and acute neurologic deficit. Ann Thorac Surg 1996; 61: 1125–9. 16 Littler W A, McGowan D A, Shanson D C. Changes in recommendations about amoxycillin prophylaxis for prevention of endocarditis. British Society for Antimicrobial Chemotherapy Endocarditis Working Party. Lancet 1997; 350: 1100. 17 Simmons N A. Recommendations for endocarditis prophylaxis. Endocarditis Working Party for Antimicrobial Chemotherapy. J Antimicrob Chemother 1993; 31: 437–8.
Surgical treatment: cardiac surgical intervention has an important role in the management of the intracardiac complications of infectious endocarditis and has led to a decrease in mortality, particularly when correcting anatomical complications that have caused congestive heart failure.13 Indications for cardiac surgery are listed in Figure 6.14 A cardiovascular surgeon should be consulted early in treatment, to enable surgery to be undertaken expediently, if needed. When valvular dysfunction results in heart failure or antimicrobial therapy fails, delaying surgery in favour of continued antimicrobial therapy leads to increased mortality. Recurrent infectious endocarditis involving prosthetic valves placed during active infectious endocarditis occurs in 2–3% of patients. This is lower than the mortality seen in similar patients treated medically. Timing of surgical intervention requires modification in patients who have suffered a recent neurological injury. If possible, surgery should be delayed for 2–3 weeks after an embolic infarct and for 1 month or more after intracerebral haemorrhage.15 In patients with negative valve cultures, preoperative plus postoperative antimicrobial therapy should at least equal a full course of recommended treatment. Patients with positive valve cultures, and most patients with prosthetic-valve endocarditis, should receive a full course of treatment after surgery.
Prophylaxis Although antimicrobial prophylaxis to prevent infectious endocarditis has not been proved effective, it remains the standard of care. Prophylaxis is recommended in patients with cardiac conditions who are considered to be at high risk (prosthetic valves, cyanotic congenital heart disease, previous infectious endocarditis) or at moderate risk (valvular heart disease including aortic or mitral regurgitation, aortic stenosis, mitral valve prolapse with mitral regurgitation, bicuspid aortic valve, non-cyanotic congenital
MEDICINE 33:4
72
© 2005 The Medicine Publishing Company Ltd