Chapter 37. Surgical Treatment of Aortic Valve Endocarditis

Infective endocarditis is a disease in which a microorganism colonizes a focus in the heart, producing fever, heart murmur, splenomegaly, embolic manifestations, and bacteremia or fungemia. Early diagnosis of this condition is extremely important because it almost invariably leads to devastating complications and death if not treated with antibiotics, combined or not with surgery. Infective endocarditis is defined as community-acquired or health care–associated infection. The latter can be nosocomial or non–nosocomial health care associated.

Predisposing factors for infective endocarditis are cardiac abnormalities that disrupt the endocardium by means of a jet injury, as well as the presence of blood-borne microorganisms that colonize these abnormal surfaces. Among patients with aortic valve endocarditis, congenitally bicuspid aortic valve is the most common predisposing lesion.1 Other congenital abnormalities of the aortic valve, degenerative calcific aortic stenosis, aortic insufficiency secondary to connective tissue disorders, and rheumatic aortic valve disease, are also predisposing lesions for infection. Depending on the virulence of the offending microorganism, normal aortic valves can also be affected. Patients with prosthetic heart valves have a constant risk of developing infective endocarditis.

It is difficult to determine the incidence and prevalence of native aortic valve endocarditis in the general population because this disease is continuously changing.2 The annual incidence of infective endocarditis is estimated to range from 1.7 to 7.0 episodes per 100,000 person-years in North America.3–5

Patients with prosthetic aortic valves are reported to have an incidence of infective endocarditis of 0.2 to 1.4 episodes per 100 patient-years, which varies with the type of aortic valve.6–12 Approximately 1.4% of patients undergoing aortic valve replacement develop prosthetic valve endocarditis during the first postoperative year.13

The incidence of nosocomial endocarditis is increasing because more patients undergo invasive procedures. Infective endocarditis in hemodialysis patients is relatively infrequent, but it is associated with high mortality.14 Dental extractions have been demonstrated to produce bacteremia. Dental flossing can produce bacteremia in periodontally healthy and periodontally diseased individuals at a rate comparable with that caused by some dental treatments, for which antibiotic prophylaxis is usually given to prevent endocarditis.15 Endoscopic procedures may also produce bacteremia. Intravenous drug users are particularly susceptible to infective endocarditis, which often occurs in structurally normal heart valves. (Please see Prophylaxis of Infective Endocarditis elsewhere in this chapter.)

In 1928, Grant and colleagues16 theorized that platelet-fibrin thrombi on the heart valve served as a nidus for bacteria adherence. In 1963, Angrist and Oka17 introduced the term nonbacterial thrombotic endocarditis to describe sterile vegetations on a heart valve and provided experimental animal evidence supporting the role of these vegetations in the pathogenesis of endocarditis. Experimental inoculation in animals with preexisting nonbacterial thrombotic endocarditis produced by mechanical abrasion of the endothelial covering of heart valves causes a prompt leukocytic infiltration of the thrombi.18 As the microorganism multiplies, more leukocytes and thrombotic material accumulate in the area and a verrucous vegetation begins to form.

Depending on the virulence of the microorganism and the resistance of the host, the aortic valve can be destroyed and the infection may spread into the annulus and surrounding structures with abscess formation. The abscess may rupture into the pericardial or a cardiac cavity.

Infective endocarditis of the aortic valve not only causes destruction of the aortic cusps, paravalvular abscess, and cardiac fistulas but also can cause coronary and systemic embolization of vegetations.19 Cerebral infarction, either ischemic because of arterial occlusion, or hemorrhagic, because of rupture of the mycotic aneurysm, is common in these patients.20,21 Mycotic aneurysms, infarcts, and abscesses of other organs such as spleen, liver, kidneys, and limbs are also common.18 Aortic valve endocarditis with a large vegetation that prolapses into the left ventricle and comes in contact with the anterior leaflet of the mitral valve can cause secondary involvement of this valve.22,23

Infection of a mechanical heart valve is usually located in its sewing ring.24,25 Infection of a porcine or pericardial valve may involve the cusps, the sewing ring, or both.26,27 Infection in aortic valve homografts and pulmonary autografts resembles that of the native aortic valve: It begins in the aortic cusps and destroys them, causing aortic insufficiency, but it may also extend into surrounding structures.28 Endocarditis after aortic root replacement with mechanical valves frequently causes dehiscence of the valve from the aortic annulus with consequent false aneurysm.29

The microbiology of infective endocarditis of the aortic valve depends on whether the valve is native or prosthetic, and whether the infection is hospital or community acquired. Staphylococcus aureus and Streptococcus viridans are the most common microorganisms responsible for native aortic valve endocarditis.30,31Staphylococcus aureus is extremely virulent and able to cause infection in patients with normal aortic valves. Streptococcus viridans is not as virulent and causes infection that often follows a protracted course. Staphylococcus epidermidis and various other streptococci can also cause endocarditis. Coagulase-negative staphylococci have emerged as an important cause of native valve endocarditis in both community and health care settings.32

Endocarditis caused by gram-negative bacteria is uncommon, but it is often resistant to antibiotic therapy and may cause serious complications. Haemophilus, Actinobacillus, Cardiobacterium, Eikenella, and Kingella (the HACEK group) are gram-negative bacilli grouped together because of their characteristic fastidiousness requiring a prolonged incubation period before growth. Endocarditis resulting from the HACEK group is also uncommon. Fungal endocarditis is rare but extremely serious. Candida albicans and Aspergillus fumigatus are the usual agents.

The microbiology of prosthetic aortic valve endocarditis is somewhat different from that of the native valve.25,33 Prosthetic valve endocarditis has been arbitrarily classified as early when it occurs within the first 2 months after surgery and late when it occurs after 2 months.34 However, it is possible that many cases of prosthetic valve endocarditis that occur during the first year after surgery are acquired at the time of implantation of the artificial heart valve.35,36 This may be particularly true when the infection is caused by coagulase-negative staphylococci and the HACEK group of bacteria. Early prosthetic valve endocarditis is caused by contamination of the valve at the time of implantation by perioperative bacteremia.13Staphylococcus epidermidis, S. aureus, and Enterococcus faecalis are among the more common microorganisms responsible for early prosthetic valve endocarditis.13,33,36 The sources of late prosthetic valve endocarditis are more difficult to determine. Bacteremia is probably the principal cause of late endocarditis. Although streptococci and staphylococci are commonly encountered in these patients, a myriad of microorganisms can cause late prosthetic valve endocarditis.33,36

Nosocomial infections are often caused by S. aureus or other staphylococci.

In a small proportion of cases of aortic valve endocarditis, no microorganism can be cultured from either the blood or surgical specimens.33,36 This is called "culture-negative endocarditis," but it is important to rule out fastidious microorganisms and every effort should be made to identify them.

It is helpful to classify infective endocarditis as acute and subacute because there are major differences between these two clinical presentations. Subacute endocarditis is often caused by less virulent microorganisms such as S. viridans. When this organism affects a diseased aortic valve, the clinical course is protracted and antibiotics alone cure most cases. On the other hand, acute endocarditis is frequently caused by a virulent microorganism such as S. aureus and may affect a normal aortic valve. The clinical course is acute, and antibiotics alone often fail to cure the infection.

The onset of subacute endocarditis in most cases is subtle, with low-grade fever and malaise. Patients think they have the "flu" and are often treated with oral antibiotics for a week to 10 days with improvement of symptoms. However, in most cases the symptoms recur a few days after stopping antibiotics. In the majority of cases no predisposing factor is identified. An aortic valve murmur is present in nearly all patients because they have preexisting aortic valve disease. Splenomegaly is common. Clubbing of the fingers and toes may develop in longstanding cases. Skin and mucous membrane signs occur late in this form of endocarditis. Petechiae appear on any part of the body. Small areas of hemorrhage may be seen in the ocular fundi. Hemorrhages in the nail beds usually have a linear distribution near the distal end, hence the name splinter hemorrhages. Osler nodes are acute, tender, barely palpable nodular lesions in the pulp of the fingers and toes. Bacteria have been cultured from these lesions. Embolization of large vegetation fragments may cause dramatic clinical events such as acute myocardial infarction (AMI), stroke, or splenic or hepatic infarcts. Any other organ also may be involved. Destruction of the aortic cusps causes aortic insufficiency and heart failure. The blood pathology is not distinctive in subacute endocarditis. Anemia without reticulocytosis develops in patients untreated for more than a few weeks. The leukocyte count is moderately elevated. Blood cultures frequently identify the offending microorganism.

The clinical course of acute endocarditis is often fulminating. A preexisting source of bacteremia may be identified. This form of endocarditis can present with all the symptoms and signs described under subacute endocarditis, but they are more acute and patients are often sicker with overwhelming signs of sepsis. Early metastatic infections are common. Two physical signs are seen only in acute endocarditis: the Janeway lesion (a painless red-blue hemorrhagic lesion a few millimeters in diameter found in the palms of the hands and the soles of the feet) and the Roth spot (an oval pale area surrounded by hemorrhage near the optic disc). Acute endocarditis is common in patients with no preexisting aortic valve disease. Early cardiac decompensation caused by aortic insufficiency is common. Paravalvular abscess is also common, and depending on the location of the abscess, the electrocardiogram may show an increased PR interval or heart block. The blood picture is one of acute sepsis. Blood culture often isolates the infecting agent.

Prosthetic valve endocarditis may present as acute or subacute endocarditis.

Doppler echocardiography is very useful in the diagnosis and management of infective endocarditis.37–40 Transesophageal echocardiography is usually better than transthoracic echocardiography and multiplane is better than monoplane for the diagnosis of endocarditis. Echocardiography can detect vegetations as small as 1 or 2 mm in size, but it is more reliable in native than prosthetic valve endocarditis. It is more useful for tissue than for mechanical valves because of the acoustic shadowing of ball, disc, or leaflet motion of mechanical heart valves. Echocardiography is also highly sensitive for detecting paravalvular abscess and cardiac fistulas.40,41

Clinical investigators from Duke University proposed certain criteria for confirming or rejecting the diagnosis of infective endocarditis.42 These criteria have been confirmed by other investigators and their limitations addressed by others.43–45 A modified version of the Duke criteria has been proposed45 and it is shown in Table 37-1.

Heart catheterization and coronary angiography increase the risk of embolization in patients with aortic valve vegetations and should be avoided. Newer computed tomography (CT) imaging techniques to diagnose coronary artery disease are useful in these patients.

The investigators of the International Collaboration on Endocarditis–Prospective Cohort Study (ICE-PCS), an international and multicenter database on patients with confirmed endocarditis maintained at the Duke Clinical Research Institute recently reported on the clinical presentation, etiology, and outcome of infective endocarditis with definite diagnosis according to Duke criteria (see Table 37-1).31 The ICE-PCS report included a cohort of 2781 patients whose median age was 57.9 years. The endocarditis was native in 72%, prosthetic in 21% and pacemaker/ICD related in 7%. Approximately one-fourth of the patients had history of recent health care exposure. The mitral valve was infected in 41.1% of the patients and the aortic valve in 37.6%. Staphylococcus aureus was the offending microorganism in 31.2% of all cases. The diagnosis of stroke was made in 16.9% and other emboli were diagnosed in 22.6%. Congestive heart failure (CHF) developed 32.2%. Paravalvular abscess occurred in 14% of the patients. Surgical treatment was common for the entire cohort (48.2%), and overall in-hospital mortality was 17.7%.

An appropriate antibiotic is the most important aspect of the management of patients with infective endocarditis.30,31,36 Antibiotic therapy should be started soon after obtaining several blood cultures. The initial choice of antibiotics is based on clinical circumstances and the suspected source of infection. Patients who had recent dental work should receive antibiotics to counteract bacteria from the oral cavity; those who had recent urinary or colonic procedures should be treated with antibiotics that are effective against gram-negative bacteria.

Intravenous drug users are usually infected with S. aureus or coagulase-negative staphylococci, and antibiotics should be chosen accordingly. Once the microorganism is identified by blood cultures and its sensitivity to specific antibiotics is known, antibiotic therapy is adjusted accordingly. A combination of two or three antibiotics that potentiate each other is often needed in the treatment of endocarditis caused by virulent microorganisms. Intravenous antibiotic therapy is continued for 6 weeks.

It is difficult to eradicate infection caused by virulent microorganisms with antibiotics alone because these microorganisms often destroy the native aortic valve very rapidly and cause aortic insufficiency and CHF. These infections are usually caused by S. aureus, Pseudomonas aeruginosa, Serratia marcescens, or fungi.

Surveillance blood cultures are performed in 48 hours to monitor the efficacy of antibiotic therapy. The patient must be watched closely for signs of CHF, coronary and systemic embolization, and persistent infection. Daily electrocardiograms and frequent echocardiograms are performed during the first 2 weeks of treatment. With any evidence of increasing aortic insufficiency, enlarging vegetations, recurrent embolism, paravalvular abscess, or persistent infection, surgery should be immediately performed. It is important to operate on patients before they develop intractable heart failure, cardiogenic or septic shock, or extensive aortic root abscesses. Patients with vegetations larger than 10 mm present a clinical problem because they are more likely to develop serious complications and early surgery is justifiable.37–39

Anticoagulation is not effective in preventing embolization of vegetations in native and biologic valves and is associated with an increased risk of neurologic complications.46,47

Surgical treatment should be considered in patients with signs of CHF, acute valve dysfunction, paravalvular abscess or cardiac fistulas, recurrent systemic embolization when aortic valve vegetations are present, and persistent sepsis despite adequate antibiotic therapy for more than 4 to 5 days.

Patients with neurologic deficits should have CT or magnetic resonance imaging performed to determine if the cerebrovascular accident is ischemic or hemorrhagic. Ischemic damage is far more common than hemorrhagic damage, but both are associated with increased mortality and morbidity.46–49 Mycotic aneurysms should be treated before valve surgery. Aortic valve replacement should be postponed for 2 weeks after an ischemic stroke and 4 weeks after a hemorrhagic stroke if possible.49

Patients who need surgery are often very sick and may be in CHF. For this reason and because they often require complex and long surgical procedures, myocardial protection is of utmost importance. Another important aspect of surgery for endocarditis is avoidance of contamination of the surgical field, instruments, drapes, and gloves with vegetations and pus. Instruments used to extirpate contaminated areas in the heart should be discarded before reconstruction of the ventricle and aortic root begins. In addition, local drapes, suction equipment, and surgical gloves should all be changed.

When the infection is limited to the cusps of the native aortic valve or a bioprosthetic valve, complete removal of the valve and implantation of a biologic or mechanical valve usually resolves the problem. There is no evidence that bioprostheses are better than mechanical valves in patients with active infective endocarditis.50 Some investigators believe that aortic valve homograft is ideal for patients with active endocarditis,51–53 but the fact is that it can become infected like other valves and there is no evidence that the risk of persistent or recurrent infection is different from other valves.54,55 Some surgeons favor the pulmonary autograft, particularly in young patients.56

If the aortic annulus is involved in the infective process, resection of the necrotic or inflamed area is needed before a prosthetic valve can be implanted. The defect created by the resection should be patched before a prosthetic valve is implanted. We prefer to use fresh autologous pericardium to patch small defects (1 or 2 cm wide) in the aortic root and left ventricular outflow tract (LVOT), and glutaraldehyde-fixed bovine pericardium for larger defects.57,58 Some surgeons also use Dacron fabric to reconstruct the aortic root.59,60 Here again, aortic valve homograft is believed to be ideal for reconstruction of the aortic root and LVOT.52,61–63 The mitral valve of the aortic valve homograft can be used to patch defects in the LVOT by correctly orienting the homograft. However, an aortic valve homograft is by no means a substitute for radical resection of all infected tissues, because persistent infection can occur with this biologic valve.64,65 The pulmonary autograft has also been used in cases of extensive destruction of the aortic root,66 but again, this valve is no substitute for radical resection of all infected tissues.

Surgery for aortic root abscess and/or cardiac fistulas is challenging. The most important aspect in the surgical treatment of these patients is radical resection of all infected tissues.25,55,57,58 We believe that the type of valve implanted is less important than complete extirpation of all infected and edematous tissues.58 These patients frequently require replacement of the entire aortic root and reconstruction of the surrounding structures that are also involved by the abscess. These operations must be individualized because the pathology of aortic root abscess is variable. Extensive resection and reconstruction may be needed.58,67–69 Thus, patching of the interventricular septum, dome of the left atrium, intervalvular fibrous body, right atrium, and pulmonary artery may be necessary, as well as repair of the left and/or right coronary arteries. The aortic root is often replaced with a valved conduit.

Aortic root abscess extending into the intervalvular fibrous body or into a prosthetic mitral valve is particularly difficult to treat.67–69 In these cases, the resection and reconstruction can be performed through the aortic root and dome of the atrium.67–69 When an aortic valve homograft is used for this type of reconstruction, the anterior leaflet of the mitral valve of the homograft can be used as patch material for the new fibrous body between the aortic and mitral valves. Actually, aortic and mitral valve homografts in a single bloc of tissue have been used to treat this condition.70

Postoperative complications are common after surgery for active infective endocarditis. Septic patients may have severe coagulopathy and may bleed excessively after cardiopulmonary bypass. Antifibrinolytic agents such as tranexaminic acid or aminocaproic acid should be used in addition to platelets, cryoprecipitate, and fresh-frozen plasma in patients with coagulopathy. The administration of recombinant Factor VII may also be necessary after correction of thrombocytopenia, fibrinogen level, and thromboplastin and prothrombin times. Radical resection of aortic root abscess may cause heart block, for which a permanent pacemaker will be needed postoperatively. Depending on the patient's clinical condition before surgery, multiorgan failure may develop postoperatively. Neurologic deterioration may occur in patients with preexisting cerebral emboli. Pulmonary, splenic, hepatic, and other metastatic abscesses seldom require surgical treatment. Large metastatic abscesses may have to be drained, and in the case of the spleen, splenectomy should be performed because of the risk of rupture.71

Clinical Results

The prognosis of aortic valve endocarditis depends largely on when the disease is diagnosed, on the offending microorganism, and how promptly it is treated.30,31,72 Patients with prosthetic aortic valve endocarditis have a more serious prognosis than patients with native aortic valve endocarditis,31,36 and nosocomial infections are associated with higher mortality than community-acquired infections.73,74 The results of surgery for infective endocarditis have improved significantly since the introduction of antibiotics and surgery, but in-hospital mortality remains high at approximately 18%.31 Prosthetic valve endocarditis, increasing age, pulmonary edema, S. aureus infection, coagulase-negative staphylococcal infection, mitral valve vegetation, and paravalvular abscess were associated with an increased risk of in-hospital mortality at the ICE-PCS report on 2781 patients.31 Approximately one-half of those patients required surgery as part of their treatment.31

The operative mortality for patients with infection limited to the cusps of the aortic valve is largely dependent on the patients' clinical presentation at the time of surgery, age, and comorbidities. Most reports indicate that the operative mortality is under 10%.36,55,75 The operative mortality is higher for prosthetic valve endocarditis and ranges from 20 to 30%.25,26,31,76 Similarly, surgery for aortic root abscess is associated with higher operative mortality.58–63

We reviewed our 25-year experience with surgery for infective endocarditis in 383 consecutive patients.36 There were 226 patients with native and 117 with prosthetic valve endocarditis. The overall operative mortality was 12%. Preoperative shock, prosthetic valve endocarditis, paravalvular abscess and S. aureus were independent predictors of operative mortality. The 15-year survival for patients with native valve endocarditis was 59% and for patients with prosthetic valve endocarditis was 25% (p < .01). The freedom from recurrent infective endocarditis at 15 years was 86%, and similar for patients with native and prosthetic valve endocarditis. In most of these patients, a different microorganism caused the second episode of endocarditis.

The American Heart Association guidelines for prophylaxis of infective endocarditis were updated in 2007.77

The major changes in the updated recommendations include the following: (1) The Committee concluded that only an extremely small number of cases of infective endocarditis might be prevented by antibiotic prophylaxis for dental procedures even if such prophylactic therapy were 100% effective. (2) Infective endocarditis prophylaxis for dental procedures is reasonable only for patients with underlying cardiac conditions associated with the highest risk of adverse outcome from infective endocarditis. (3) For patients with these underlying cardiac conditions, prophylaxis is reasonable for all dental procedures that involve manipulation of gingival tissue or the periapical region of teeth or perforation of the oral mucosa. (4) Prophylaxis is not recommended based solely on an increased lifetime risk of acquisition of infective endocarditis. (5) Administration of antibiotics solely to prevent endocarditis is not recommended for patients who undergo a genitourinary or gastrointestinal tract procedure. These changes are intended to define more clearly when infective endocarditis prophylaxis is or is not recommended and provide more uniform and consistent global recommendations.

These new recommendations represent a dramatic shift with regard to which patients should receive antibiotic prophylaxis for prevention of infective endocarditis. However, they are "guidelines" and physicians caring for patients with organic heart valve disease should advise them according to the perceived risk of endocarditis and potential benefit of antibiotic prophylaxis.