Prosthetic valve endocarditis: Surgical management

INTRODUCTION — 

Prosthetic valve endocarditis (PVE) occurs in 1 to 6 percent of patients with valve prostheses, with an incidence of 0.3 to 1.2 percent per patient-year, with similar frequencies of PVE after surgical or transcatheter valve replacement. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis".)

Infection of a prosthetic heart valve can be difficult to diagnose and manage. Optimal treatment of PVE requires [1-4]:

●Identification of the causative microorganism.

●Selection of a bactericidal antimicrobial regimen of proven efficacy.

●A clear understanding of the intracardiac pathology and attendant complications of PVE.

●Surgical intervention in selected cases, such as when infection has extended beyond the valve to contiguous cardiac tissue or resulted in major valvular dysfunction.

The surgical management of PVE will be reviewed here.

Related topics are discussed separately, including:

●(See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis".)

●(See "Antimicrobial therapy of prosthetic valve endocarditis".)

FREQUENCY OF SURGERY FOR PVE — 

Most patients with PVE have an indication for cardiac surgery and the risk of mortality remains high [3,5].

For surgical valve PVE — Most of the data on PVE are for surgically implanted valves.

In large series of patients with infective endocarditis (IE) reported by national and international study groups (many based at tertiary care referral centers and thus subject to referral bias), PVE composed 20 to 30 percent of IE cases; in 18 to 50 percent of PVE cases, management included surgery [6-8]. As an example, among 26,111 patients with PVE in the 2003 to 2017 United States National Inpatient Sample database, 18 percent underwent valve replacement surgery [9]. In an international study of patients with PVE admitted or referred to tertiary medical centers, treatment of 50 percent included cardiac surgery [10].

In 20 to 35 percent of patients with PVE, surgery is warranted (based on standard indications) but not performed for a number of reasons, including patient refusal, death before surgery, or contraindications (related to PVE complications or comorbid conditions) [3,7,8,10,11]. Preoperative assessment frequently indicates a high risk of surgical mortality and postsurgical complications with associated disability [12]. Thus, patients with PVE are at high risk for morbidity and mortality with or without surgical management. (See 'For surgical valve PVE' below.)

Surgery for PVE is largely directed at eradicating infection unresponsive to antibiotic therapy due to highly resistant and/or virulent pathogens involving foreign material, debriding paravalvular invasion, and addressing prosthetic valve dysfunction and resulting heart failure (HF) with valve replacement. Paravalvular invasion occurs commonly in PVE involving surgically implanted valves (eg, occurring in over 30 percent of patients in one international study [13] and in 60 percent of patients in a study from a major referral center [14]). The frequency of invasive disease is higher when infection arises within 12 months of valve surgery, involves an aortic prosthesis, or is caused by an invasive organism such as Staphylococcus aureus [15].

For transcatheter heart valve PVE — Endocarditis is increasingly encountered in patients with transcatheter aortic or pulmonic heart valve replacement. The frequency of PVE involving transcatheter aortic valve implantation (TAVI-PVE) appears at least comparable with the frequency of PVE in patients with surgical aortic valve replacement (SAVR-PVE).

In-hospital and one-year mortality with TAVI-PVE are high. On imaging, 18 to 20 percent of patients with TAVI-PVE have findings of paravalvular invasive infection and 70 percent have indications for surgical intervention similar to those guiding the treatment of IE in patients with SAVR-PVE [16-18]. However, given the age range and preexisting comorbidities in the population receiving TAVI, only 20 percent of patients with TAVI-PVE undergo cardiac surgery.

CLINICAL CARE SETTING — 

Decisions to intervene surgically and optimal timing are nuanced; optimally, such decisions are made by a multidisciplinary heart valve team with infectious disease expertise in endocarditis, in a center with reconstructive cardiac surgical experience and robust perioperative support services [2,3]. Early referral to such centers likely improves outcomes [3,7,14].

SURGICAL INDICATIONS — 

Indications for cardiac surgery in patients with PVE have been developed to address complications related to intracardiac pathology and inability to eradicate infection in the absence of surgical debridement of necrotic and infected tissue. Some of the indications for surgical treatment are more compelling than others, and all require careful, individualized risk-benefit analysis.

The following recommendations for early valve surgery for patients with PVE are in broad agreement with the 2014 American Heart Association (AHA) IE scientific statement, the 2016 American Association of Thoracic Surgery consensus guidelines for surgical treatment of endocarditis, the 2020 American College of Cardiology/AHA valve guidelines, and the 2023 European Society of Cardiology guidelines for endocarditis [1-4,19].

Referral for early surgery is indicated in patients with PVE 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) or valve dehiscence causing symptoms or signs of HF, we recommend referral for early valve surgery rather than nonsurgical management. (See 'Heart failure' below and 'Prosthetic valve dehiscence' below.)

●For patients with paravalvular extension of infection with development of annular or aortic abscess, destructive penetrating lesion (eg, fistula), unstable hypermobile prosthetic valve dehiscence causing regurgitation, and/or heart block, we recommend referral for early valve surgery rather than nonsurgical management. Paravalvular infection is more common with PVE compared with native valve endocarditis and typically requires transesophageal echocardiography (TEE) or other advanced imaging modalities for diagnosis. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis", section on 'Cardiac imaging' and 'Prosthetic valve dehiscence' below and 'Paravalvular extension' below.)

●For patients with PVE due to a difficult-to-treat pathogen, we suggest referral for early valve surgery. Difficult-to-treat pathogens include multidrug-resistant organisms (enterococci [particularly Enterococcus faecium] and gram-negative non-HACEK organisms [particularly Pseudomonas aeruginosa]). We do not regard PVE due to staphylococcal or fungal organisms to be an independent indication for early surgery in the absence of other indications. (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 valve surgery.

It has not been established whether surgery to remove large vegetations (in the absence of other indications for surgical intervention) provides a survival benefit. (See 'Vegetation characteristics and risk of embolization' below.)

Antibiotic therapy alone can be successful in patients with PVE who have no evidence of HF, significant prosthetic valve dysfunction, or paravalvular infection and who are infected by less virulent organisms. These patients are characterized by later onset of infection (commonly more than 12 months after prosthesis implantation) and infection by viridans streptococci, HACEK, or enterococci (that can be treated with bactericidal therapy) [20-22]. In a study of patients managed medically (56 percent of whom had an indication for surgery), relapse occurred in 5 percent, and one-year mortality was 24 percent (16 percent in those without an indication for surgery) [23].

EVIDENCE FOR SURGICAL INDICATIONS — 

Recommendations for valve surgery in patients with PVE are based upon observational studies that are subject to bias (including treatment selection and survivor bias) and confounding [13]. The available evidence suggests that patients with PVE with new (or severe) mitral or aortic regurgitation, paravalvular invasion, prosthetic valve dehiscence, or HF are most likely to benefit from surgery [1,11,13,14,24,25].

Heart failure — Observational data suggest that the outcome of PVE in patients who experience HF due to prosthetic valve dysfunction is improved with valve surgery, although these observations are subject to bias [13,14,26,27]. Among PVE patients with moderate to severe HF, few survive beyond six months if treated with antibiotics alone, whereas 44 to 64 percent survive with timely surgical intervention [21,28-31]. Surgery must be performed before HF becomes refractory. There is no evidence that delaying surgery to allow further antibiotic therapy in this setting improves outcome [3,4,14]. In fact, the operative mortality in these patients increases with the severity of hemodynamic compromise at the time of surgery. Thus, hemodynamic status should weigh more heavily than the duration of antibiotic therapy in determining the timing of surgical intervention.

Emergency surgery (within 24 hours) is warranted for patients with invasive infection and valvular dysfunction who have refractory pulmonary edema or cardiogenic shock, and urgent surgery (within three to five days) is warranted for patients with HF due to acute severe prosthetic aortic or mitral valve dysfunction and/or echocardiographic evidence of hemodynamic compromise (eg, left ventricular systolic dysfunction or pulmonary hypertension) [3]. In patients with milder HF who stabilize with medical therapy, surgery can be delayed until later in antibiotic therapy when the usual indications for surgery can be applied [3].

Complicated infection

Paravalvular extension — Patients with PVE complicated by paravalvular invasion (evidenced by development of complete atrioventricular [AV] block, annular or aortic abscess, aortic aneurysm or pseudoaneurysm, fistula formation, or other destructive penetrating lesions diagnosed by echocardiography or computed tomography [CT]) experience high mortality rates and are rarely cured with medical treatment alone [7,10]. Relative to patients with native valve endocarditis (NVE), these paravalvular complications occur at higher rates among patients with PVE, including paravalvular abscess (34 versus 13 percent), intracardiac fistula (4.6 versus 2.1 percent), and pseudoaneurysm (10.9 versus 5.2 percent) [7]. In contrast, valve perforation or rupture is more common with NVE compared with PVE (22.9 versus 3.6 percent) [7]. Patients with these complications are among those most likely to benefit from cardiac surgery [7,14].

On the other hand, surgery might be appropriately delayed for limited echocardiographic or ¹⁸F-fluorodeoxyglucose-positron emission tomography (FDG-PET)/CT findings thought to reflect limited paravalvular infection that appears to be controlled. Identification of paravalvular invasion by TEE and other imaging technologies (cardiac CT and FDG-PET/CT) is discussed separately. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis".)

PVE with onset ≤6 months after valve replacement is frequently associated with paravalvular extension of infection and/or more virulent and resistant organisms [3]; patients with PVE early after valve replacement should be followed carefully for these factors.

Prosthetic valve dehiscence — Prosthetic valve dehiscence is defined as a gap between the prosthetic valve sewing ring and the native heart tissue. A small area of dehiscence can be due to suture disruption from a mechanical cause, such as implantation technique or mitral annular calcium, presenting as paravalvular regurgitation, often early after valve implantation. However, new valve dehiscence in the setting of PVE is most often a manifestation of paravalvular infection and abscess.

An unstable, hypermobile prosthesis due to dehiscence of anchoring sutures causing HF and/or valve regurgitation requires urgent intervention, because the valve is likely to become increasingly unstable with acute severe valve regurgitation.

Difficult-to-treat pathogens

●Multidrug-resistant organisms – For patients with PVE caused by multidrug-resistant organisms (enterococci [particularly E. faecium] and gram-negative non-HACEK organisms [particularly P. aeruginosa]), we suggest early surgical intervention (ideally after clearance of bacteremia, if feasible) even in the absence of other indications. The reasons for this approach include limited therapeutic options and the need for use of treatment regimens with uncertain efficacy (which may be associated with significant risk of toxicity).

For patients with PVE caused by staphylococcal or fungal infection, we do not regard infection due to these pathogens as an independent indication for early surgery.

•S. aureus – Surgery for PVE due to S. aureus is warranted in the context of cardiac indications for surgery (see 'Surgical indications' above); we do not consider S. aureus IE alone an indication for early surgery.

This approach is supported by a retrospective study of 168 patients with definite S. aureus PVE. After adjusting for selection bias (propensity for surgery), prognostic metrics, and survivor bias (time of surgery), surgical treatment conveyed no benefit for in-hospital or one-year mortality [24]. These observations suggest that the decision for surgical therapy in S. aureus PVE is more complex than previously thought.

Given the very high mortality rates in patients with PVE with indications for surgical intervention who do not undergo surgery, surgical intervention is likely to be beneficial in patients with S. aureus PVE with clinically significant intracardiac complications and hemodynamic dysfunction [11,14,25,28,32,33]. However, for select patients with S. aureus PVE (such as those who are younger, have few comorbidities, and PVE without paravalvular invasion or major stroke), management with antibiotics alone can be effective [34]. When attempting medical therapy in patients with S. aureus PVE, careful follow-up is required to monitor for development of intracardiac complications requiring surgical intervention.

•Fungal infection – Early surgical intervention is considered by most experts to be a standard component of treatment for fungal PVE; however, carefully adjusted analyses to reduce confounding bias are not available to evaluate this approach [7]. In view of the available data, benefit from surgery is most likely in the setting of cardiac indications for surgery. (See 'Surgical indications' above.)

In one review including 70 cases of Candida spp endocarditis, including 32 cases of PVE, outcome was not improved by surgical intervention nor by whether primary antifungal therapy consisted of a lipid formulation of amphotericin B or a high-dose echinocandin regimen [35]. In a retrospective study including 46 patients with Candida PVE (primarily caused by C. parapsilosis and C. albicans) surgical intervention was not associated with improved survival at six months, although a survival benefit was associated with liposomal amphotericin B therapy [36].

In these studies, patients who undergo surgery are typically those with intracardiac complications, HF, and uncontrolled infection; in contrast, those who are treated medically often have either uncomplicated infection or contraindications to surgery. Accordingly, selection bias likely compromises these analyses of outcome.

Persistent infection — Limited data are available on the impact of valve surgery in patients with PVE 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 [37]. A rationale for proceeding with early surgery in patients with persistent infection is that paravalvular abscess may be missed by TEE, particularly when the abscess is located adjacent to mitral annular calcification. (See "Role of echocardiography in infective endocarditis".)

Vegetation characteristics and risk of embolization — It has not been established that surgery to remove large vegetations in patients with PVE (in the absence of other indications for surgery) is beneficial. While prevention of emboli that can cause irreversible end-organ damage (eg central nervous system and myocardial infarction) is a laudable goal, the mortality risk of surgery itself, which is potentially increased in patients with PVE, should be considered carefully. (See 'Effect of surgery on mortality' below.)

The approach to patients with PVE and emboli or large mobile vegetations differs from that for patients with NVE. Among patients with NVE, the frequency of embolic complications is higher in patients with S. aureus infection or hypermobile vegetations exceeding 10 mm in diameter compared with those with smaller, less mobile vegetations [38-40]. In addition, large vegetations are associated with increased frequency of coincident intracardiac complications of endocarditis. Comparable data are not available for patients with PVE. However, the overall rate of embolic complications is similar for patients with PVE and NVE. In both conditions, the embolization rate decreases rapidly over the initial two weeks of effective antimicrobial therapy and continues to decline thereafter [41].

A randomized trial of early surgery for patients with left-sided NVE complicated by large vegetations plus other indications for valve surgery demonstrated decreased embolic events in those treated with early surgery compared with conventional management, but there was no survival benefit with early surgery [42]. The lack of a survival benefit may have been confounded by frequent late surgery for many patients in the conventional treatment group.

The following observational studies suggesting benefit to early surgery among patients with vegetations ≥10 mm included some patients with PVE; however, most patients had NVE.

●Retrospective studies (with variable statistical adjustments to control for biases) in patients with left-sided IE have found decreased in-hospital and six-month mortality among patients with vegetations ≥10 mm and intracardiac complications who undergo early surgery compared with those treated medically, although it is unclear whether the survival benefit is related to prevention of embolic events [38,40,43].

●Retrospective studies of patients with left-sided NVE with vegetations ≥10 mm and no prior embolic or intracardiac complications have found an association between early surgical intervention and improved survival [38] and absence of endocarditis-related complications at six months compared with similar patients treated medically [44].

Effect of surgery on mortality

For surgical valve PVE — Most of the data on outcomes for surgical management of PVE are for patients with PVE affecting surgically implanted valves.

●Meta-analysis – The impact of surgical intervention for PVE on mortality is uncertain, despite the results of a large meta-analysis which suggested a survival benefit from surgical intervention for PVE [45]. The meta-analysis of 32 observational studies of PVE compared 1320 patients who underwent surgical intervention (valve reoperation) with 1316 patients who were treated with medical therapy alone [45]. In unadjusted analyses, valve reoperation was associated with a lower risk of 30-day mortality (25 versus 34 percent; risk ratio [RR] 0.73, 95% CI 0.64-0.82), greater survival at a mean of 22-month follow-up (69 versus 58 percent; RR 1.27, 95% CI 1.05-1.53), and a similar rate of PVE recurrence compared with medical therapy alone. However, the studies included in the analysis often lacked information on how many patients with indications for surgery were treated medically due to high operative risk, and the studies were often not adjusted for risk factors and for time from onset of therapy to surgical intervention.

●Largest study in the meta-analysis – Of note, the largest study included in this meta-analysis (contributing approximately 40 percent of the patients [45]) did not confirm early or late survival benefits from surgery after careful adjustment for potential biases [13]. This international multicenter prospective study examined 1025 patients with PVE treated between January 2000 and December 2006 [13]. Of these patients, 490 (48 percent) underwent valve surgery during the initial hospitalization for treatment of PVE.

•No overall mortality benefit – In an unadjusted analysis, surgery during the initial hospitalization was associated with reduced in-hospital (22 versus 27 percent; hazard ratio [HR] 0.68, 95% CI 0.53-0.87) and one-year mortality (27 versus 37 percent; HR 0.68, 95% CI 0.55-0.85). However, when outcomes were analyzed with adjustment for the propensity to operate (treatment selection bias) and for time of surgical intervention (survivor bias caused by attributing to medical therapy early deaths in patients who may have had surgical indications), surgical treatment was not associated with a reduction in in-hospital or one-year mortality. HF was the strongest predictor of in-hospital and one-year mortality.

•Benefit based upon indication for surgery – In post hoc subgroup analyses not adjusted for survivor bias, surgery was associated with decreased in-hospital and one-year mortality in the subpopulation with the strongest indications for surgery (new mitral or aortic valve regurgitation, paravalvular abscess or fistula, prosthetic valve dehiscence or severe regurgitation, or HF) [13,46]. However, surgery was not associated with decreased mortality when the indications for intervention were less strong (mitral or aortic valve vegetations, persistent bacteremia). These data suggest that redo cardiac surgery in patients with PVE may carry increased intrinsic mortality risk, which, in turn, offsets the potential benefits of surgery in some patients, particularly when surgery is to address less immediately threatening intracardiac complications.

●Single institution study of early surgery – In a retrospective single institution study, the outcome of 404 patients with PVE treated surgically was compared with that of 119 patients with PVE who received only medical treatment [14]. The indications for surgery were those generally recommended in guidelines. Importantly, the study institution has recognized expertise in cardiac surgery and many in the study population were referred for surgical consideration. After careful analysis to control biases common in retrospective studies of outcome, including a substudy of propensity score-matched cohorts, surgical intervention was associated with a significantly lower risk of death compared with medical treatment (HR 0.32, 95% CI 0.22-0.48). Surgery was undertaken early with the decision to operate made a median of four days (interquartile range [IQR] 2 to 6 days) after admission, and the additional time to surgery was two days (IQR 1 to 5 days).

●Studies with only in-hospital follow-up – Similar findings were observed in studies with only in-hospital follow-up [10,26,47].

•An international multicenter prospective study of 355 patients with PVE noted the following variables independently associated with surgical therapy: intracardiac abscess, HF, younger age, coagulase-negative staphylococci, and S. aureus [47]. Unadjusted in-hospital mortality rates were similar for surgical and medical treatment (25 and 23 percent, respectively). A propensity analysis to compare surgically treated patients with those treated medically demonstrated a nonsignificant trend toward reduced in-hospital mortality with surgical treatment (odds ratio [OR] 0.51, 95% CI 0.23-1.36). A similar trend was noted among patients with high propensity scores for cardiac surgery.

•In a later study using this international database, in-hospital mortality among 556 patients with PVE was significantly associated with age ≥65 years, health care-associated infection, stroke, S. aureus infection, persistent bacteremia, HF, and intracardiac abscess [10]. The three latter variables are considered indications for surgical treatment, and S. aureus infection is often accompanied by paravalvular invasion, which is an indication for surgical treatment.

For transcatheter heart valve PVE — Most of the data on transcatheter heart valve endocarditis (THV-PVE) come from studies on PVE after transcatheter aortic valve implantation (TAVI). Studies of TAVI-IE have not found surgical therapy to convey a survival benefit compared with medical treatment [48,49]. This may relate to the significantly high operative risk associated with age and preoperative comorbidity or selection bias. With increasing use of TAVI in younger patients with few comorbidities, some patients with TAVI-IE may benefit from surgical intervention. A study reported similar crude postoperative mortality in surgically treated PVE in patients with TAVI-PVE and SAVR-PVE (13.6 versus 10.8 percent, respectively) and in a multivariable analysis, prior TAVI was not significantly associated with mortality [50].

TIMING OF SURGERY

●General considerations – The optimal timing for surgery for PVE remains controversial, as it depends on numerous clinical considerations. Timing of surgery must balance the urgency of indications for surgery against the risk factors or contraindications for surgery. For most patients with PVE with indications for early surgery (see 'Surgical indications' above), we favor proceeding with surgery within a few days of development of the surgical indication unless surgery is contraindicated or considered futile. In most cases of complicated PVE, there is no advantage to delaying intervention.

●Timing after stroke – Physiologic changes and anticoagulation required during cardiac surgery can exacerbate prior ischemic strokes or intracerebral hemorrhage. The risk of neurologic worsening relates to the nature of the lesion (ischemic versus hemorrhagic), the severity of the neurologic insult, and the time elapsed since the lesion occurred. That risk must be weighed against immediacy of the surgical indication.

The recommendations of the American Heart Association, European Society of Cardiology, and the American Association for Cardiac Surgery for the timing of surgical intervention after a neurologic event are similar [1-4,23]:

•Transient ischemic attacks and asymptomatic "microhemorrhages" noted on T2 magnetic resonance images (which are not in fact hemorrhagic lesions) are not contraindications to cardiac surgery.

•Following an ischemic infarct without coma, major neurologic impairment, or hemorrhagic conversion, early surgical treatment (within seven days of admission) to address major surgical indications is associated with better neurologic outcomes and survival and thus should not be delayed.

•Following cerebral infarcts with coma, major neurologic impairment, or significant intracerebral hemorrhage, cardiac surgery should be delayed for at least three to four weeks and then reevaluated, unless the indication for surgery is an immediate risk of death.

We agree with the above approach; ideally, these complex decisions should be made by a multidisciplinary endocarditis team including a neurologist.

ANTIMICROBIAL THERAPY

●Selection – Issues related to antibiotic selection are discussed separately. (See "Antimicrobial therapy of prosthetic valve endocarditis".)

●Duration

•We treat with a six-week course of antimicrobial therapy from the date of valve replacement in the following circumstances:

-Patients with positive intraoperative valve culture and/or pathology demonstrating acute necrotizing infection

-Patients who underwent valve placement at a site of overt paravalvular infection

•For patients with negative intraoperative culture and lack of necrotizing infection who undergo surgery late in the planned course of antimicrobial therapy, we count the duration of preoperative antibiotics toward the total treatment course.

For fungal PVE, prolonged suppressive therapy to prevent relapse is warranted. (See "Candida endocarditis and suppurative thrombophlebitis", section on 'Chronic suppressive therapy'.)

For certain organisms (eg, Coxiella burnetii [51], Brucella spp, Tropheryma whipplei), prolonged regimens are warranted, independent of surgical intervention. (See "Brucellosis: Treatment and prevention", section on 'Endocarditis' and "Whipple's disease", section on 'Treatment'.)

ANTICOAGULATION — 

Among patients with prosthetic valve infective endocarditis, the potential benefit of preventing valve obstruction due to thrombus formation and embolization with anticoagulation must be weighed against the increased risk of intracerebral hemorrhage. This issue is discussed in detail separately. (See "Antithrombotic therapy in patients with infective endocarditis".)

RELAPSE — 

PVE that relapses after appropriate antimicrobial therapy may reflect paravalvular infection or a previously unrecognized extracardiac focus of infection [15,28]. Patients who relapse should be evaluated carefully, including imaging to assess for new evidence of paravalvular complications. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis".)

The antimicrobial susceptibility must be evaluated for emergence of resistance. The benefit of surgery (particularly as related to intracardiac complications) relative to retreatment with antibiotic therapy must be weighed carefully.

COMPLICATIONS AND OUTCOMES

Surgical complications — Complications following surgery for PVE reflect the risks of redo surgical valve replacement in the setting of infection frequently associated with paravalvular tissue invasion. Preoperative concurrent comorbidities may also be exacerbated. Major complications include severe coagulopathy, bleeding or tamponade requiring reexploration of the chest, acute renal failure, low cardiac output syndrome, complete AV block, respiratory failure requiring prolonged ventilatory support, stroke, sepsis, recrudescent PVE, and prosthetic valve dysfunction [3,12,19,52]. While the frequencies of these complications are not insignificant, they are relatively small compared with the anticipated mortality when antibiotic therapy alone is pursued in the face of indications for surgical intervention.

At institutions with experienced cardiac surgeons and perioperative support services, operative mortality rates for patients with PVE treated with valve replacement and surgical reconstruction of paravalvular tissue range from 10 to 30 percent [12,14,52]. In contrast, the projected mortality of PVE complicated by extensive paravalvular invasion, extensive valve dehiscence, or marked valve dysfunction and progressive HF would exceed 50 percent without surgery [7,11,14,53-58]. Data suggest that surgical intervention for markedly complicated PVE should be performed in centers with extensive experience when possible. Complex reconstruction of the aortic or mitral valve apparatus and the supporting structures is often required to achieve an optimal outcome of PVE [14].

Repeat episodes of endocarditis occur in 6 to 15 percent of surgically treated patients with PVE; repeat surgery on the newly implanted prosthesis may be indicated when paravalvular infection or valve dysfunction is present [15,53-55,58,59]. The rates of recrudescent PVE after surgery or valve dysfunction (either of which might require repeat surgery) are not insignificant, but they are relatively small compared with the survival benefits from surgery when indicated and the anticipated mortality when indicated surgery is not performed [12,52].

Surgical outcomes — The effect of surgery on mortality is discussed above. (See 'Effect of surgery on mortality' above.)

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: Treatment and prevention of infective endocarditis" and "Society guideline links: Cardiac valve disease".)

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

●Frequency of surgery for PVE (See 'Frequency of surgery for PVE' above.)

•For surgical valve PVE – Most patients with prosthetic valve endocarditis (PVE) have an indication for cardiac surgery, and the risk of mortality remains high. The response to medical treatment of PVE is often more limited than for native valve endocarditis (NVE), and surgical replacement of the prosthesis is often required in conjunction with antibiotic therapy. Paravalvular infection and abscess is common in PVE, especially when infection is caused by more aggressive pathogens (staphylococci), arises within 12 months of surgery, or involves an aortic prosthesis.

•For transcatheter valve PVE – PVE following transcatheter aortic valve implantation (TAVI-PVE) occurs with a frequency similar to that seen for surgically implanted bioprosthetic aortic valves. More than half of patients with TAVI-PVE present with indications for surgical intervention; however, only a small fraction undergo explantation and surgical aortic valve replacement (SAVR) because of high estimated surgical mortality risk in this older population with comorbidities.

●Clinical care setting – Decisions on whether to proceed with surgical treatment of PVE require careful individualized risk-benefit analysis and are optimally made by a multidisciplinary heart valve team with infectious disease expertise in endocarditis in a center with significant reconstructive cardiac surgical experience and robust perioperative support services. (See 'Clinical care setting' above.)

●Indications for referral for early valve surgery – Referral for valve surgery during the initial hospitalization for PVE is indicated in the following circumstances (See 'Surgical indications' above.)

•High-risk complications – For patients with PVE with one or more high-risk complications, we recommend early surgery rather medical intervention alone (Grade 1C). We define a high-risk complication of PVE as:

-Heart failure (HF) resulting from prosthetic valve dehiscence, intracardiac fistula or prosthetic valve dysfunction. (See 'Heart failure' above.)

-Paravalvular complications such as heart block, annular or aortic abscess, or destructive penetrating lesion.

or

-Unstable hypermobile prosthetic valve dehiscence causing regurgitation. (See 'Paravalvular extension' above and 'Prosthetic valve dehiscence' above.)

The rationale for this approach is that observational studies indicate high mortality risk in patients with a high-risk complication and suggest that surgery may improve survival. (See 'Heart failure' above.)

•Multidrug-resistant organisms – For patients with PVE due to multidrug-resistant organisms (including enterococci [particularly Enterococcus faecium] and gram-negative non-HACEK organisms [particularly Pseudomonas aeruginosa]), we suggest early surgery rather than medical management alone (Grade 2C). The reasons for this approach include limited therapeutic options and the need for use of treatment regimens with uncertain efficacy (which may be associated with significant risk of toxicity). (See 'Difficult-to-treat pathogens' above.)

In contrast, for patients with PVE caused by staphylococcal or fungal infection, we do not regard infection due to these pathogens as an independent indication for early surgery.

•Persistent infection – 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 valve surgery rather than medical management alone (Grade 2C). (See 'Persistent infection' above.)

•Large vegetations – For patients with left-sided PVE with large vegetations (≥10 mm) at risk for initial or recurrent systemic emboli, the decision on whether to proceed with surgery is based upon individualized estimates of benefit and risk. (See 'Vegetation characteristics and risk of embolization' above.)

●Timing of surgery – For most patients with PVE with indications for surgery (see 'Surgical indications' above) there is no advantage of delaying surgery; we favor proceeding with surgery within a few days of development of the surgical indication unless surgery is contraindicated or considered futile. (See 'Timing of surgery' above.)

The timing of surgery for PVE in patients who have had a stroke or intracerebral hemorrhage must consider the potential exacerbation of the neurologic injury by the cardiac surgery. Neurology consultation is advised.

●Complications and outcomes – In the hands of experienced cardiac surgeons, operative mortality rates for patients with PVE treated with valve replacement and surgical reconstruction of paravalvular tissue range from 10 to 30 percent. Without surgery, the projected mortality for patients with PVE with severe paravalvular infection or HF due to prosthetic valve dysfunction exceeds 50 percent. (See 'Complications and outcomes' above.)

ACKNOWLEDGMENT — 

The UpToDate editorial staff thank Gabriel S Aldea, MD, who contributed to earlier versions of this topic review.