Surgery for left-sided native valve infective endocarditis

INTRODUCTION — Since the 1960s, valve replacement and valve repair have become common procedures for management of selected cases of infective endocarditis (IE). A review found that a mean of 32 percent of IE patients in published studies underwent valve surgery [1], but the percentage of patients undergoing valve surgery approaches nearly 50 percent in contemporary studies of predominantly left-sided IE [2]. The indications for, efficacy of, and considerations for surgery in the setting of left-sided native valve endocarditis will be reviewed here. The evaluation and medical treatment of native valve endocarditis, management of right-sided endocarditis, and the role of surgery in prosthetic valve endocarditis are discussed separately. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis" and "Antimicrobial therapy of left-sided native valve endocarditis" and "Right-sided native valve infective endocarditis" and "Surgery for prosthetic valve endocarditis".)

OVERVIEW — Surgical consultation is warranted for all patients with IE, since indications for surgery may develop or clinical status may worsen quickly. Determination of whether an indication for early surgical intervention in IE is present is a critical aspect in the management of IE patients. In general, surgery is warranted in the setting of complications for which antibiotic treatment alone is unlikely to cure the condition or prevent further complications or death. For patients with IE with an indication for surgical intervention, an individualized risk-benefit analysis is performed by a multispecialty heart valve team to determine whether, when, and what surgical procedure should be performed to improve patient outcome.

SURGICAL REFERRAL

Indications for surgical consultation — Prompt surgical consultation is warranted in all patients with IE, as indications for surgical intervention may appear or progress rapidly. Patients with IE are best managed by a multispecialty heart valve team that includes an infectious disease specialist, cardiologist, and cardiac surgeon, with input from neurology, neurosurgery, radiology, and other specialists based on the individual case [3-7]. Multidisciplinary treatment of IE has been found to reduce mortality in IE by improving the use of surgery based on indications and consideration of operative risk and optimal timing [6,8].

Surgical indications — Determination of whether a patient with IE requires early surgical treatment depends upon many clinical and prognostic factors [3]. Early surgery in IE has been broadly defined as surgery performed before completion of a full course of antibiotics [9]. The following recommendations are similar to those in guidelines established by United States and European cardiology societies and the American Association for Thoracic Surgery [3,5,10,11].

Referral for early surgery is indicated in patients with left-sided native valve IE in the following clinical settings. The evidence supporting these recommendations is discussed below. (See 'Evidence for surgical indications' below.)

●For patients with IE-associated valve dysfunction (usually aortic or mitral regurgitation) causing symptoms or signs of heart failure (HF), we recommend referral for early valve surgery. (See 'Heart failure' below.)

●For patients with paravalvular extension of infection with development of annular or aortic abscess, destructive penetrating lesion (eg, fistula), and/or heart block, we suggest referral for early valve surgery. (See 'Paravalvular extension' below.)

●For patients with infection due to a difficult-to-treat pathogen, we suggest referral for early valve surgery. Difficult-to-treat pathogens include fungi and multidrug-resistant organisms. We do not consider Staphylococcus aureus IE alone an indication for early surgery. (See 'Difficult-to-treat pathogens' below.)

●For patients with persistent infection (manifested as persistent bacteremia or fever lasting more than five days after initiation of appropriate antibiotic therapy, provided other sites of infection and causes of fever have been excluded), we suggest referral for early valve surgery.

●Early surgery (within the first week of antibiotic therapy) may reduce the risk of embolism in patients with large vegetations (≥10 mm) but criteria for referral for surgery in this setting are controversial, and guideline recommendations have varied [3,5,10,12]. For patients with large vegetations, we perform an individualized risk-benefit assessment comparing early surgery with expectant management based upon multiple factors including the diameter and volume of the vegetation, change in size of the vegetation on appropriate antibiotic therapy, the infecting pathogen, history of prior systemic embolization, likelihood that the patient will soon require valve surgery (eg, due to severe valve dysfunction), and patient age and life expectancy (which impacts prosthetic valve choice and exposure to long-term risks of prosthetic valve replacement). (See 'Vegetation characteristics and risk of embolization' below.)

Evidence for surgical indications — Recommendations for surgical management of IE are based largely on observational studies, since only one small randomized trial has been performed of surgery for IE as discussed below [3] (see 'Vegetation characteristics and risk of embolization' below). Since observational studies are subject to bias (including survivor and selection bias) and confounding, studies have used various methods in an attempt to more accurately assess the effect of surgery on outcomes. Observational studies that have adjusted for bias and confounding have produced mixed results. Among studies of patients with IE that have adjusted for survivor bias, some have found harm or no significant benefit associated with surgery compared with medical therapy [13-15], while others have found an association between valve surgery and reduced mortality (in-hospital, at one year, or at five years), particularly for patients with IE complications [16-19]. The differing results may be due to a difference in analytic methods, as well as differences in patient populations, particularly since studies were not limited to patients with commonly accepted indications for surgery.

Surgery is recommended for selected patients with IE who meet the previously mentioned criteria, and, in the absence of large randomized trials, the most relevant evidence on efficacy comes from studies that included propensity scoring and/or subset analysis of patients with high-risk features considered indications for surgery. Evidence of an association between surgery and reduced mortality [3] has been found in patients with complicated IE, including HF [16,18], intracardiac abscess or fistula [17,18], native valve S. aureus IE [17], or systemic embolization [17,18].

Further studies are needed to identify patients with IE that will benefit from valve surgery.

Heart failure — The survival benefit of surgery in IE appears to be greatest among patients with severe valve dysfunction causing HF. Ideally, surgery should be undertaken as soon as signs and symptoms of HF appear and before hemodynamic instability occurs.

Prior to valve replacement or repair for management of endocarditis, approximately 90 percent of deaths among patients with IE were attributable to HF. Moderate to severe HF due to IE confers a high risk for mortality in the absence of surgery. In such cases, medical therapy alone is associated with approximately 75 percent mortality; with surgery, the mortality rate is ≤25 percent [20-23]. Among patients with IE who undergo surgery, HF is the indication in 65 to 75 percent of patients [24].

Surgery should be undertaken as soon as signs and symptoms of HF appear, before extreme or refractory hemodynamic deterioration occurs. This recommendation is supported by a prospective multicenter study of 4075 patients with IE, including 1359 with HF; surgery was performed in 62 percent of patients with HF [16]. Surgery was associated with lower in-hospital mortality versus no surgery in patients with New York Heart Association (NYHA) functional class I or II HF and a greater reduction in mortality rate with surgery was seen in patients with NYHA class III or IV HF. Both in-hospital and one-year mortality rates were significantly lower among patients with HF who underwent surgery during the initial hospitalization than patients who did not (21 versus 45 percent and 29 versus 58 percent, respectively).

Nearly all patients with severe valvular regurgitation due to endocarditis will have signs and symptoms of HF due to acute volume overload of the nondilated left ventricle. IE in patients with preexisting compensated valvular regurgitation may lead to decompensation due to worsening valve damage and further regurgitation [25].

HF caused by severe valvular regurgitation is an indication for early surgical intervention. This is particularly important in the treatment of acute severe aortic insufficiency when medical therapy is ineffective; in this setting, intraaortic balloon pump use is ineffective and can be harmful. Early surgical consultation in these patients is critical, as clinical deterioration can be very rapid and the window of surgical opportunity for patients with acute valvular regurgitation due to IE can be very short. If there is an unavoidable delay in surgery, temporary stabilization with medical therapy is attempted in an intensive care setting.

Even in the absence of overt HF, individuals with severe regurgitation (based on quantitative echo parameters) and hemodynamic compromise (such as tachycardia or evidence of elevated pressure within the pulmonary circulation) may be candidates for early surgery. In the rare patient with IE and severe regurgitation in whom there are no clinical or echocardiographic signs of cardiac decompensation, initial non-surgical management may be reasonable with the timing of intervention than based on standard indications for chronic native valve regurgitation. The role of surgery in patients with severe valve stenosis or regurgitation and a large mobile vegetation is discussed below. (See 'Vegetation characteristics and risk of embolization' below.)

Noncardiac factors may exaggerate manifestations of HF; these include fever, anemia, sepsis, and renal insufficiency. In addition, patients with antecedent left ventricular dysfunction may present with signs or symptoms of HF that are out of proportion to the degree of valvular dysfunction caused by IE. Such patients require careful evaluation and management with diuretics and vasodilators, as well as close surveillance of the response to therapy. In some cases of acute HF in IE, HF symptoms may improve or resolve with medical therapy [26]. (See "Overview of the management of heart failure with reduced ejection fraction in adults".)

Complicated infection — Complicated infection refers to extension of infection beyond the valve leaflets, which can manifest as paravalvular abscess, fistula formation, septal defects, heart block, leaflet perforation or other destructive penetrating lesions, or sinus of Valsalva aneurysm. Endocarditis caused by a difficult-to-treat pathogen may also be considered complicated.

Complicated infection should be suspected in patients with persistent fever and/or persistently positive blood cultures after five to seven days of appropriate antibiotic therapy. In such cases, meticulous clinical evaluation for metastatic infection, drug fever, and secondary (often nosocomial) infection is needed. Patients should also undergo transesophageal echocardiography to evaluate for the presence of paravalvular infection (as transthoracic echocardiography is not sufficient to identified complicated infection). Issues related to complicated IE are discussed further separately. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis" and "Complications and outcome of infective endocarditis".)

Paravalvular extension — We recommend early valve surgery for patients with paravalvular extension of infection with development of annular or aortic abscess, destructive penetrating lesion (eg, fistula), and/or heart block. Extension of infection beyond the valve leaflets can manifest as paravalvular abscess, fistula formation, heart block, leaflet perforation, or sinus of Valsalva aneurysm. Paravalvular extension occurs in 10 to 20 percent of patients with native valve endocarditis [27]. Evidence of invasive infection may be incidentally found during diagnostic imaging, particularly by transesophageal echocardiography, or during investigation for new atrioventricular conduction abnormalities or persistent infection. (See "Complications and outcome of infective endocarditis", section on 'Cardiac complications' and "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis".)

Surgery is warranted for patients with paravalvular extension, which is often fatal and typically requires complex repair procedures by a surgeon experienced in such repairs; the likelihood of cure with medical therapy alone in such cases is low [28]. Indeed, any delay in surgical therapy in these cases can lead to a considerably more complex and potentially hazardous operative repair as the infection continues. In one study including 20 patients with paravalvular abscess managed medically, the mortality rate was 40 percent; after six months, an additional 40 percent of patients had at least one complication including recurrent sepsis, HF, or renal failure [29]. In another series including 41 patients with IE and annular abscess, survival among patients treated with medical therapy alone was <25 percent [30]. The mortality for surgical patients with annular abscesses ranged from 19 to 43 percent in various series [31-33] and in one series including 233 patients, the survival rate at 27 months was 59 percent [29]. Independent risk factors for mortality included age, staphylococcal infection, and abscess with fistula formation.

Difficult-to-treat pathogens — We suggest referral for early surgery in IE cases due to difficult-to-treat pathogens, include fungi and multidrug-resistant organisms (eg, Pseudomonas aeruginosa and vancomycin-resistant Enterococcus). Endocarditis caused by a difficult-to-treat pathogen may also be considered complicated [3,34]. Pathogens that are difficult to cure with medical therapy alone include:

●Fungal infection:

•The optimal approach for management of Candida endocarditis is a combination of valve replacement and a long course of antifungal therapy; this approach is based on case reports, case series, cohort studies, a meta-analysis, and clinical experience [35]. There are conflicting data about the mortality rate associated with a combined medical and surgical approach compared with antifungal therapy alone; medical therapy of Candida endocarditis has occasionally been curative. This issue is discussed further separately. (See "Candida endocarditis and suppurative thrombophlebitis".)

•Combined medical therapy and valve replacement are essential for management of Aspergillus endocarditis in attempts to prevent embolic complications and valvular decompensation; this approach is based on case series [36,37]. In one review including 61 patients, survival was 20 percent among those who received combined surgical and medical therapy; none survived among those who received medical therapy alone [38]. (See "Treatment and prevention of invasive aspergillosis".)

●Multidrug-resistant organisms:

•Pseudomonas aeruginosa – Cardiac valve surgery has been recommended as a cornerstone of treatment of IE due to P. aeruginosa based on case series up to 1990 [39,40]. A subsequent case series in 2009 suggested surgery is warranted if blood cultures remain positive for seven days on appropriate antimicrobial therapy, or if Pseudomonas bacteremia recurs after completion of a six-week course of combination antimicrobial therapy [41]. (See "Pseudomonas aeruginosa bacteremia and endocarditis".)

•Vancomycin-resistant Enterococcus – Most cases of IE due to vancomycin-resistant Enterococcus are nosocomial, and surgical management in such patients is complicated, as they often have serious comorbid conditions that would preclude surgical intervention [42-44]. One retrospective review including 50 cases noted that surgical intervention trended toward improved survival, but was not statistically significant [43].

We do not consider S. aureus IE, particularly methicillin-sensitive pathogens, an absolute indication for early surgery and instead assess for other indications for early surgery. While some favor early surgery (during the initial hospitalization before completion of a full therapeutic course of antibiotics) in patients with left-sided IE caused by S. aureus [10,11], other literature suggests that early surgery in IE patients may be associated with a higher risk of mortality in the early postoperative period [45]. Patients with S. aureus IE commonly have higher operative risk due to comorbid medical conditions such as renal failure and sepsis [2]. No randomized trials have demonstrated better survival or outcome for early surgery in S. aureus IE compared with medical therapy. In an observational study of 203 patients with S. aureus IE, an analysis adjusted for treatment selection bias found no significant survival benefit from early valve surgery (death hazard rate 0.71, 95% CI 0.34-1.49) [46]. Routine use of early surgery for S. aureus IE without other IE complication may not improve long-term survival due to adverse host factors.

Persistent infection — Limited data are available on the impact of valve surgery in patients with native valve IE with persistent infection (manifested as persistent bacteremia or fever lasting more than seven days after initiation of appropriate antibiotic therapy, provided other sites of infection and causes of fever have been excluded). Persistent bacteremia has been independently associated with higher six-month mortality in IE, and surgery was found to improve survival after propensity adjustment [9]. A rationale for proceeding with early surgery in patients with persistent infection is that paravalvular abscess may be missed by transesophageal echocardiography, particularly when the abscess is located adjacent to mitral annular calcification. (See "Role of echocardiography in infective endocarditis".)

Vegetation characteristics and risk of embolization — Among patients with IE, risk factors for embolization include vegetation size >10 mm [47], vegetation mobility, vegetation location on the anterior mitral leaflet, prior embolization, and infection with S. aureus, Streptococcus bovis, or fungus [7]. The risk of embolization in the setting of native valve endocarditis is 13 to 44 percent, and in many cases embolization occurs before a diagnosis of IE has been established [48-50]. The risk of embolization tends to decline rapidly after initiation of effective antimicrobial therapy, and serious embolic events are uncommon a week after such therapy is begun. In a multicenter study of 384 patients with definite IE (75 percent with native valve endocarditis), S. aureus and S. bovis were predictors of total embolic events, while vegetation length >10 mm and severe vegetation mobility were significant predictors of embolic events after initiation of antibiotic therapy [51]. (See "Complications and outcome of infective endocarditis", section on 'Metastatic infection' and "Role of echocardiography in infective endocarditis", section on 'Prognostic markers'.)

A 2018 meta-analysis found that patients with IE and vegetation size >10 mm had a significantly higher risk of embolic events and in-hospital mortality [47]. However, the available data have not established that surgical management improves outcomes in patients with large vegetations without other indication for valve surgery (such as heart failure or uncontrolled infection).

●An observational study of 1006 patients with left-sided IE (584 with vegetation size >10 mm and 422 with vegetation size ≤10 mm) confirmed that the risk of embolic events (44.3 versus 28.4) and six-month mortality (25.1 versus 19.4 percent) was higher among patients with large vegetations >10 mm compared with patients with smaller vegetations [52]. After propensity adjustment, higher mortality was observed in patients with large vegetations managed medically (hazard ratio [HR] 1.86, 95% CI 1.48 to 2.34) but not in those managed surgically (HR 1.01, 95% CI 0.69 to 1.49) compared with patients with smaller vegetations. Among patients with large vegetations managed medically, 52 percent had a documented reason for not undergoing surgery; the reasons included 31 percent with a poor prognosis, 21 percent had a stroke, 21 percent with good prognosis without surgery and in 19 percent the surgeon declined to operate. Thus, these data do not establish a causal relationship between surgical management and improved outcomes in patients with large vegetations as the only reason for surgery.

●A later observational study of 726 patients with left sided IE (420 with vegetation size >10 mm and 306 with vegetation size ≤10 mm) found that although the group with large vegetations had a higher in hospital mortality rate (31.7 versus 24.8 percent), vegetation size was not an independent predictor of mortality [53]. Among patients with large vegetation without heart failure or uncontrolled infection, mortality rate was similar with and without surgery (18.6 versus 11.6 percent).

Increasing vegetation size during antimicrobial therapy has been associated with a less favorable prognosis [54], although surgery is not routinely indicated for patients who are otherwise responding well to medical therapy.

Limited data are available on the impact of the timing of surgery in patients with a large vegetation. A small, randomized study of patients with native left-sided IE with severe valve regurgitation or stenosis and vegetation diameter >10 mm compared surgery within 48 hours of randomization versus conventional timing for surgery (69 percent during the initial hospitalization) [55]. Surgery reduced the number of systemic embolic events without a difference in mortality. Yet observational studies have shown that the frequency of embolic events is highest at the time of presentation for IE and declines to a low daily rate within the first week of appropriate antibiotic treatment [55].

PREOPERATIVE ASSESSMENT — For patients with IE with indications for valve surgery, preoperative assessment includes clinically-guided imaging in addition to echocardiography to assess for IE complications or other comorbid condition that may adversely affect outcome and an individualized assessment of surgical risks and benefits.

Preoperative imaging — We do not recommend routine preoperative brain imaging in the absence of neurologic symptoms or signs because of the high incidence of asymptomatic brain emboli in IE [56], although some experts have suggested preoperative neurologic imaging for all patients with IE [7]. For patients with symptoms of stroke or transient ischemic attack, a thorough neurologic evaluation including brain imaging and evaluation of prognosis is indicated to guide management. Computed tomography (CT) angiography and/or conventional angiography should be performed in patients with IE with a cerebral hemorrhage to evaluate for possible ruptured infectious aneurysm. (See "Initial assessment and management of acute stroke" and "Initial evaluation and management of transient ischemic attack and minor ischemic stroke" and "Spontaneous intracerebral hemorrhage: Pathogenesis, clinical features, and diagnosis", section on 'Evaluation and diagnosis'.)

Routine coronary angiography is not required prior to valve surgery for IE but may be performed in hemodynamically stable patients with risk factors for coronary artery disease or patients with symptoms or signs of myocardial ischemia. Cardiac catheterization in patients with IE does come with unique considerations. For example, catheter manipulation in the setting of aortic valve vegetations does carry an embolic risk. Although not yet well studied, coronary CT angiography may be a viable preoperative imaging option in patients for whom cardiac catheterization is not possible [57]. (See "Clinical use of coronary computed tomographic angiography".)

The use of noncardiac imaging, such as abdominal CT scan with contrast or positron emission tomographic-CT scan, to look for other sites of infection should be guided by the specific clinical situation and not routinely performed.

Operative risk/benefit assessment — To assess the operative risk of mortality and morbidity associated with valve surgery, we recommend using the online Society of Thoracic Surgeons’ (STS) Risk Calculator (acsdriskcalc.research.sts.org) which now includes a variable for treated or active IE. Using the large STS database, a risk model specific to surgery for IE was developed and validated [58]. Major factors predictive of operative mortality included surgical urgency (especially emergency status), hemodynamic instability or cardiogenic shock, renal failure, and surgery before completion of antibiotic therapy. In addition to predicting operative mortality, this STS-IE risk score has been found to predict six-month mortality in IE, likely because it captures many adverse conditions, such as renal failure [2]. In another study, similar variables were found to be associated with higher operative mortality, such as age, surgical urgency, S. aureus IE, and comorbid host factors [59].

The risk of operative mortality is as high as 25 percent among patients with end-stage kidney disease on hemodialysis [60]. In a retrospective study including more than 11,000 patients on hemodialysis with IE, 11 percent underwent valve replacement [61]. Risk factors for mortality included older age, diabetes mellitus as the cause of chronic kidney disease, surgery during the index hospitalization, dysrhythmia, and infection with S. aureus.

There is an interaction between the presence of surgical indication, operative risk, and surgical intervention that is associated with outcome. Patients with a surgical indication and low operative risk who undergo surgery have a relatively low mortality rate at six months (approximately 10 percent); in contrast, patients who have surgical indications, high operative risk, and no surgery performed have an estimated six-month mortality of over 50 percent [2]. In a similar study in France, IE patients who had a surgical indication and surgery performed had a nominally but not significantly lower mortality rate than those with a surgical indication but no surgery performed due to high operative risk [26].

A substantial proportion of patients with IE and indications for surgery do not undergo surgery, largely because their operative risk is deemed too high for a reasonable expectation of benefit. Although more than half of patients with left-sided IE will develop a surgical indication during their index hospitalization, nearly one in four patients with an indication will not undergo surgery due to operative risk [2]. In one cohort including 803 patients with IE and indications for surgery, only 661 patients actually underwent surgery [2]. Patients with indications for surgery who did not undergo surgery more commonly had an underlying poor prognosis (34 percent), hemodynamic instability (20 percent), stroke (23 percent), and sepsis (21 percent). Specific clinical characteristics independently associated with not proceeding with surgery despite presence of an indication include liver disease, stroke before surgery, and S. aureus IE (likely related to adverse host factors and comorbid illness in these patients) [2].

TIMING OF SURGERY

Approach to timing surgery — The optimal timing for surgery in native valve endocarditis remains controversial in many cases, as the timing depends on numerous clinical considerations. Timing must balance the urgency of indications for surgery against the risk factors or contraindications for surgery.

For patients with native valve endocarditis with indications for early surgery (see 'Surgical indications' above), we favor proceeding with surgery within a few days of development of the surgical indication [7]. In most cases of complicated IE, there is no advantage to delaying intervention [3]. As an example, in patients with severe HF due to endocarditis-associated valve dysfunction, emergency or urgent surgery is warranted unless surgery is deemed contraindicated (eg, large ischemic stroke with coma or large intracranial hemorrhage that precludes heparin anticoagulation) or futile (eg, multiple or severe stroke in patient with comorbidities) [7].

In general, early surgery is warranted for patients with the indications discussed above, irrespective of the duration of preoperative antibiotic therapy. The duration of intravenous antibiotic therapy is typically four to six weeks, a wide interval to select an appropriate time for surgery. (See 'Evidence for timing of surgery' below.)

Surgery performed less than seven days after receiving intravenous antibiotic treatment for S. aureus or enterococcal left-sided IE has been associated with positive valve culture from cardiac surgery [62]. However, surgery, when indicated, should not be delayed for this reason, as positive valve culture may impact duration of antibiotic treatment after surgery but not clinical outcome or even relapse rate [63].

United States guidelines for the management of IE suggest that early cardiac surgery occur during ongoing antibiotic therapy [3]. European guidelines provide more specific recommendations for timing of surgery based on the IE complication; emergency surgery (within 24 hours of diagnosis) is recommended for left sided IE complicated by severe valve dysfunction and refractory heart failure, while urgent surgery (after three to five days) is recommended for less severe heart failure, locally uncontrolled infection, or to prevent embolic complications from large vegetations (≥10 mm) [64].

No delay in valve surgery is required for the following conditions if indications for surgery are present [7]:

●Silent microembolism.

●Transient ischemic attack.

●Cerebral abscess, if asymptomatic and without significant neurologic impairment.

●Cerebral microhemorrhages detected by brain magnetic resonance imaging as T2* hypodense regions <10 mm in size, if asymptomatic [65].

●Ischemic stroke without hemorrhagic conversion and without severe neurological impairment or decreased level of consciousness. (See 'Timing following stroke' below.)

However, in certain clinical situations, delayed or deferred surgery is appropriate:

●Low life expectancy and/or multi-organ failure.

●Stroke or cerebral abscess with severe neurological impairment or decreased level of consciousness or coma (suggested surgical delay of four weeks or more). (See 'Timing following stroke' below.)

●Intracranial hemorrhage (suggested surgical delay of four weeks or more). (See 'Timing following stroke' below.)

●Ruptured mycotic aneurysm.

For patients with IE and intracerebral hemorrhage, vascular imaging should be performed to determine the cause. For patients with IE and intracerebral hemorrhage, the timing of surgery ranges from immediate to a four-week delay depending upon the size of the hemorrhage and the urgency of valve surgery [7]. In this scenario, a multidisciplinary approach involving neurology and neurosurgery is critical to make an informed decision on the timing of surgery.

Evidence for timing of surgery

General timing — In general, patients with IE complication who have an indication for surgery and acceptable operative risk should proceed to surgery as soon as possible, because additional medical therapy may not significantly improve outcome after surgery. A meta-analysis of 21 observational studies of surgical treatment of IE found that surgery performed at seven days or less from admission had the lowest risk of all-cause mortality, but no adjustment for operative risk or urgency was performed [66]. However, another multicenter study with adjustment for operative risk and surgical propensity found no significant association between early surgery within seven days of diagnosis or transfer and six-month survival in IE compared with later surgery [67].

Timing following stroke — The optimum timing of valve surgery in patients with IE and stroke is controversial because only limited, observational data are available to guide management [3]. Traditional practices of delaying surgery for a six-week period after a stroke are unsupported by the available clinical data. Concerns include the potential risk of neurological deterioration due to either hemorrhagic transformation with anticoagulation or worsened cerebral ischemia during cardiopulmonary bypass. Although early studies suggested a risk of neurological deterioration with surgery early after ischemic stroke secondary to IE, later studies, including those with propensity-matched analysis to adjust for selection bias, have not found a significantly increased risk of neurologic deterioration or mortality with early surgery in this setting [7]. For patients with IE with ischemic stroke without severe neurological impairment or decreased level of consciousness/coma, early surgery enables better survival with low risk of neurologic worsening.

Anticoagulation associated with cardiopulmonary bypass does not necessarily increase the risk for conversion of a nonhemorrhagic cerebral infarct into a hemorrhagic lesion [7,34,48,68-74]. In one study including 60 patients with IE who underwent neurologic examinations and screening brain magnetic resonance imaging (regardless of neurological symptoms), no neurologic deterioration was observed after surgery among patients with a symptomatic cerebrovascular complication preoperatively [48]. Similarly, a study including 214 patients who underwent cardiac surgery (including 65 of patients with embolic stroke) demonstrated no difference in perioperative mortality between patients with or without embolic stroke [69]. In a study including 198 patients with IE and ischemic stroke who underwent surgery (30 percent underwent surgery ≤7 days after stroke and 70 percent underwent surgery >7 days after stroke), there was no apparent survival benefit in delaying surgery when indicated; one-year mortality was comparable between the groups (27 versus 19 percent) [68].

For patients with ischemic stroke with severe neurologic deficits or diminished consciousness and for patients with hemorrhagic stroke, the risk of postoperative neurological deterioration and mortality is greater, so an individualized approach to timing of surgery is indicated. The timing of cardiac surgery for IE with these neurologic complications is challenging and debatable. As noted above, we favor delaying surgery by four weeks or more in patients with IE with ischemic stroke with severe neurologic deficit or altered consciousness; and we favor delaying valve surgery for ≥4 weeks after intracranial hemorrhage. A rationale for delaying surgery for at least four weeks after major ischemic stroke is to reduce the risk of hemorrhagic transformation in this high-risk population. Thus, the 2015 American Heart Association (AHA) IE guidelines contain recommendations similar to the 2017 AHA/American College of Cardiology focused update on valvular heart disease, which notes that it is reasonable to delay valve surgery for at least four weeks after major ischemic stroke or intracranial hemorrhage if the patient is hemodynamically stable [3,75]. In an observational, multicenter study of patients with hemorrhagic stroke, mortality was higher among patients undergoing surgery within the first four weeks of hemorrhagic stroke compared with patients who underwent later surgery (75 versus 40 percent) [76]. (See "Spontaneous intracerebral hemorrhage: Pathogenesis, clinical features, and diagnosis".)

CHOICE OF PROCEDURE

Approach to choice of procedure — In the setting of active infection (particularly involving the mitral valve), valve repair is preferable to valve replacement when feasible, particularly given the small risk of prosthetic valve infection following valve replacement. Valve repair may be possible when a leaflet perforation occurs in the absence of extensive leaflet destruction or annular involvement.

Valve repair may be particularly desirable in young intravenous drug users in whom compliance with anticoagulation required for mechanical prosthetic heart valves is often poor and use of a bioprosthetic heart valve is associated with a high rate of structural deterioration over the long term. (See "Mechanical prosthetic valve thrombosis or obstruction: Clinical manifestations and diagnosis".)

Choice of prosthetic valve is based upon standard clinical criteria and patient preference. There is no clear evidence demonstrating the superiority of a particular valve type in preventing reinfection [77]. Life expectancy is an important factor in choosing a prosthetic valve. As an example, for patients on chronic hemodialysis with IE, the two-year mortality after valve replacement surgery is 65 percent; as a result, the durability of a bioprosthetic valve may be adequate in these patients [61]. (See "Choice of prosthetic heart valve for surgical aortic or mitral valve replacement".)

Management of patients with prosthetic valves and prosthetic valve complications are discussed separately. (See "Overview of the management of patients with prosthetic heart valves" and "Mechanical prosthetic valve thrombosis or obstruction: Clinical manifestations and diagnosis" and "Antithrombotic therapy for mechanical heart valves".)

Evidence on choice of valve procedure — Limited data are available to guide the choice of valve procedure. Data on mitral valve repair for active IE are limited [78-80]. There is evidence to suggest that early surgery before extensive tissue destruction may facilitate mitral valve repair [81]. In one series including 37 patients who underwent mitral valve repair in the setting of active IE, 31 underwent prosthetic annuloplasty [79]. The 10-year survival rate was 80 percent; no repeat mitral valve operations were required in 90 percent of cases. Recurrence of IE was observed in one patient.

Valve replacement surgery (irrespective of whether with a bioprosthetic or mechanical valve) for IE is associated with increased risk of recurrent IE [82]. In patients with IE who have valve replacement rather than repair, host factors (eg, older age and other comorbid conditions that may reduce longer survival) have been found to be associated with implantation of a bioprosthetic rather than mechanical prosthesis [83]. These adverse host factors likely influence long-term survival more than the type of prosthetic valve.

ANTIBIOTIC THERAPY — The approach to selection of antimicrobial therapy is discussed further separately. (See "Antimicrobial therapy of left-sided native valve endocarditis".)

In the setting of initially positive blood cultures, the first day of antibiotic therapy (for determining the duration of treatment) is counted as the first day of negative blood cultures or the day of surgery if operative tissue cultures are positive (whichever is later) [3,5]. Antimicrobial therapy should be continued for at least two weeks postoperatively, even if that extends the originally planned antibiotic course.

In the setting of initially positive blood cultures and negative operative tissue cultures, the approach to duration of antimicrobial therapy is uncertain. If operative tissue cultures are negative, it may be reasonable to count the first day of negative blood cultures as the first day of antibiotic therapy (for determining the duration of treatment) including any preoperative days of negative blood cultures [3,5].

In the setting of negative blood cultures and negative operative cultures, the optimal duration of therapy is uncertain. We favor 14 days of therapy; the duration should be counted from the first day of antibiotic administration [3]. This approach is based on a retrospective review of more than 350 patients who underwent surgery for treatment of IE in which relapse of IE was unrelated to duration of antibiotic therapy following surgery [84].

OUTCOME — Studies evaluating the outcome of surgery for IE have largely focused on in-hospital or operative (30-day) survival. There have been relatively few studies of the long-term outcome for patients undergoing surgery for native valve endocarditis. The outcome of patients with native valve endocarditis is highly dependent on the preoperative presence of key factors such as stroke, congestive heart failure, or uncontrolled bacteremia. As an example, in one retrospective study of 100 patients with native mitral valve endocarditis, five-year survival was 87 percent for patients who had none of the preceding factors but only 34 percent for those who had all three [85].

Several observational studies have found reduced longer-term survival in IE patients after initial hospitalization. One study included 475 patients who underwent surgery for isolated native mitral valve endocarditis; the five-year survival was 59.6 percent [86]. In a study including 881 patients who underwent surgery for native valve endocarditis between 1997 and 2007, survival rates at 30 days and five years were approximately 95 and 66 percent, respectively [87].

In a nationwide, population-based study in Taiwan, IE survivors (n = 10,116) had higher rates of ischemic stroke, hemorrhagic stroke, myocardial infarction, readmission for HF, and sudden and all-cause death than matched controls without IE [82].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Cardiac valve disease" and "Society guideline links: Treatment and prevention of infective endocarditis".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Endocarditis (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Approach to identifying candidates for early surgical treatment – Determination of whether and when a patient with infective endocarditis (IE) requires early surgical treatment (ie, before completion of a full course of antibiotics) requires an individualized multispecialty approach including involvement of experts in infectious disease, cardiology, imaging, and cardiothoracic surgery. The first step is to identify patients with one or more indications for early surgical referral. This is followed by preoperative evaluation including an individualized assessment of surgical risks. A substantial proportion of patients with IE and indications for surgery do not undergo surgery, largely because their operative risk is deemed too high. (See 'Surgical referral' above and 'Preoperative assessment' above.)

●Timing of surgery – Timing of early surgery in patients with indications is based upon balancing the urgency of indications for surgery against risk factors or contraindications for surgery, generally irrespective of the duration of preoperative antibiotic therapy (see 'Timing of surgery' above).

●Indications for referral for early valve surgery – Referral for early valve surgery is indicated in patients with left-sided native valve IE and one or more of the following features:

•Heart failure (HF) symptoms – For patients with IE-associated valve dysfunction (usually aortic or mitral regurgitation) causing symptoms or signs of HF, we recommend referral for early valve surgery (Grade 1B) (see 'Heart failure' above). The benefits of surgery appear to be greatest among patients with severe valve dysfunction causing HF. Ideally, surgery should be undertaken as soon as signs and symptoms of HF appear and before hemodynamic instability occurs.

•Complicated infection (see 'Complicated infection' above):

-For patients with paravalvular extension of infection with development of annular or aortic abscess, destructive penetrating lesion (eg, fistula), and/or heart block, we recommend referral for early valve surgery (Grade 1C). (See 'Paravalvular extension' above.)

-For patients with infection due to a difficult-to-treat pathogen, we suggest referral for early valve surgery. Difficult-to-treat pathogens include fungi and multidrug-resistant organisms (eg, Pseudomonas aeruginosa and vancomycin-resistant Enterococcus) (Grade 2C). We do not consider S. aureus IE an indication for early surgery. (See 'Difficult-to-treat pathogens' above.)

•Persistent infection – We suggest referral for early valve surgery for patients with persistent infection (defined as persistent bacteremia or fever lasting >7 days after initiation of appropriate antibiotic therapy, provided other sites of infection and causes of fever have been excluded) (Grade 2C). (See 'Persistent infection' above.)

•Large vegetations – Early surgery (within the first week of antibiotic therapy) may reduce the risk of embolism in patients with large vegetations (>10 mm) but criteria for referral for surgery in this setting are controversial. For patients with large vegetations, we perform an individualized risk-benefit assessment comparing early surgery with expectant management based upon multiple factors including the diameter and volume of the vegetation, change in size of the vegetation on appropriate antibiotic therapy, the infecting pathogen, history of prior systemic embolization, likelihood that the patient will soon require valve surgery for another indication (eg, due to severe valve dysfunction), and patient age and life expectancy (which impacts prosthetic valve choice and exposure to long-term risks of prosthetic valve replacement). (See 'Vegetation characteristics and risk of embolization' above.)

●Timing following stroke – No delay in valve surgery is required for silent microembolism, transient ischemic attack, cerebral abscess, or ischemic stroke with no hemorrhagic conversion and without severe neurological impairment or decreased level of consciousness. For patients with major ischemic stroke (ie, with severe neurologic deficit or altered consciousness) or intracranial hemorrhage, we typically delay surgery for ≥4 weeks. (See 'Timing following stroke' above.)

●Choice of valve procedure – In the setting of active infection, valve repair is preferable to valve replacement when feasible; valve replacement is associated with a small risk of infection of prosthetic materials. Valve repair may be possible when a leaflet perforation occurs in the absence of extensive leaflet destruction or annular involvement. (See 'Approach to choice of procedure' above.)

ACKNOWLEDGMENTS

The UpToDate editorial staff acknowledges Gabriel S Aldea, MD, Catherine Otto, MD, Stephen B Calderwood, MD, and Rakesh M Suri, MD, DPhil, who contributed to earlier versions of this topic review.

The UpToDate editorial staff also acknowledges William Gaasch, MD (deceased), who contributed to earlier versions of this topic.