Surgery for prosthetic valve endocarditis

INTRODUCTION — Infection of a prosthetic heart valve can be difficult to diagnose and manage. Optimal treatment of prosthetic valve endocarditis (PVE) requires:

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

●Surgical intervention (in selected cases), especially 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. The antimicrobial treatment of PVE and the role of surgery in native valve endocarditis are discussed separately. (See "Antimicrobial therapy of prosthetic valve endocarditis" and "Surgery for left-sided native valve infective endocarditis" and "Right-sided native valve infective endocarditis", section on 'Surgical assessment'.)

GENERAL CONCEPTS — Cardiac surgery is an important component of therapy for many patients with PVE. Among patients with PVE, 40 to 50 percent are treated surgically; surgery is likely warranted but not performed in another 15 to 20 percent due to patient refusal or because surgery is contraindicated by complications of PVE or comorbid conditions [1,2]. Paravalvular invasion is common in PVE (occurring in over 30 percent in one international study [3]), especially when infection arises within 12 months of surgery or involves an aortic prosthesis [4]. Among patients who are diagnosed prior to involvement of paravalvular tissue, antibiotic therapy alone may be effective.

The diagnosis and evaluation of PVE are discussed separately. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis".)

INDICATIONS FOR SURGERY — Indications for cardiac surgery in patients with PVE have been developed based on complications related to intracardiac pathology and inability to eradicate infection. Some of the indications for surgical treatment are more compelling than others and all require careful individualized risk-benefit analysis.

We agree with the following recommendations for early valve surgery for patients with PVE from the 2015 American Heart Association (AHA) scientific statement on infective endocarditis (IE) [5].

Early surgery (during the initial hospitalization for antimicrobial therapy) is recommended for patients with PVE with one or more of the following:

●Signs or symptoms of heart failure (HF) resulting from valve dehiscence, intracardiac fistula, or severe prosthetic valve dysfunction.

●IE complicated by heart block, annular or aortic abscess, or destructive penetrating lesion.

●PVE caused by fungi or other highly resistant organism.

●Persistent bacteremia despite appropriate antibiotic therapy for five to seven days and exclusion of other sites of infection.

As noted in the 2015 AHA scientific statement [5], early surgery may be helpful in some patients in the following clinical setting, but evidence is limited. In fact, in the largest study to date, early surgery for conditions other than valve dysfunction and perivalvular abscess was not associated with improved hospital or one-year survival [3] (see 'Evidence' below). Thus, for patients in the following clinical settings, we suggest basing a decision on whether or not to proceed with surgery upon weighing individualized estimates of benefit and risk:

●Patients with PVE who have recurrent emboli despite appropriate antibiotic treatment

●Patients with relapsing PVE

●Patients with PVE with mobile vegetations >10 mm in size

Similar recommendations are included in the 2015 European Society of Cardiology IE guidelines (table 1) [6] and the 2016 American Association for Thoracic Surgery guidelines [7]. Recommendations for surgery for PVE require continued reassessment as new data emerge.

EVIDENCE — 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. These limitations are only partially addressed by adjusted analyses. Studies have identified risk factors for mortality among patients with PVE and provide limited data on associations between treatment type (surgery versus only medical therapy) among high-risk groups [5]. Limited evidence suggests that patients with PVE with new (or severe) mitral or aortic regurgitation, perivascular invasion, prosthetic valve dehiscence, or heart failure are most likely to benefit from surgery [1,3,5,6,8,9].

A 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 [10]. Valve reoperation was associated with a lower risk of 30-day mortality, greater survival at mean 22-month follow-up, and a similar rate of PVE recurrence compared to medical therapy alone. However, the study lacked information on how many patients with indications for surgery were treated medically due to high operative risk, and there was no adjustment for risk factors or for time from onset of therapy to surgical intervention. The largest study in this meta-analysis contributed 39.4 percent of the surgically treated population and 43.9 percent of the medically treated group [3]. Importantly, the unadjusted analysis of survival in this study suggested that surgery provided end of hospitalization and one-year survival benefits; however, when the data were carefully adjusted for potential biases, the early and late survival benefits of surgery were not confirmed. These data, as shown below, call to question the simple conclusion of the meta-analysis and suggest that the decision to intervene surgically in the treatment of PVE is nuanced and requires clinical judgement.

The largest published study of PVE, an international multicenter prospective study, examined 1025 patients with PVE treated between January 2000 and December 2006 [3]. Of these patients, 490 (48 percent) underwent valve surgery during the initial hospitalization for treatment of PVE. Early surgery was associated with reduced in-hospital and one-year mortality in an unadjusted comparison with medical therapy. However, when outcome was adjusted 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), there was no reduction in in-hospital or one-year mortality among surgically treated patients. HF was the strongest predictor of in-hospital and one-year mortality. In a post hoc subgroup analysis not adjusted for survivor bias, surgery was significantly 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). However, surgery was not beneficial 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, particularly when surgery is to address less immediately threatening intracardiac complications [11].

Similar findings were observed in studies with only in-hospital follow-up [2,12,13]. 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 [12]. 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 medically treated patients 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 score 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 [2]. The three latter variables are considered indications for surgical treatment, while S. aureus is often accompanied by paravalvular invasion that is an indication for surgical treatment. (See 'Microorganisms usually requiring surgery' below.)

Antibiotic therapy alone is often 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 (more than 12 months after prosthesis implantation) and infection by viridans streptococci, HACEK, or enterococci (that can be treated with bactericidal therapy) [14-16].

These observations illustrate substantial mortality risks among patients with PVE with or without surgical therapy, which may reflect the age and burden of comorbidities in this population as well as the severity of complications from endocarditis [3,8,11,14-16]. The studies also indicate that identifying patients with PVE who will benefit from cardiac surgery during active infection is more complex than previously recognized. Recommended indications for cardiac surgery are not absolute and should be weighed carefully in decision making, together with the patient’s overall clinical condition and estimated surgical risk. Survival benefit from surgery may be limited to those patients with the most severe and demanding complications of infection: paravalvular invasion (eg, abscess) and HF from valve dysfunction or dehiscence [1,3,5,6,8,9].

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 [3,13,17]. Among patients with moderate to severe HF and PVE, few survive beyond six months if treated with antibiotics alone, whereas 44 to 64 percent survive with timely surgical intervention [15,18-21]. Surgery must be performed before HF becomes intractable. There is no evidence that delaying surgery in this setting improves outcome [22-24]. In fact, the operative mortality of these patients is proportional to the severity of hemodynamic disability at the time of surgery [22,25]. Thus, hemodynamic status should weigh more heavily than the duration of antibiotic therapy in determining the timing of surgical intervention.

Paravalvular invasion — Prosthetic valve dehiscence is also an indication of paravalvular invasion. An unstable, hypermobile prosthesis due to dehiscence of anchoring sutures causing HF and/or severe valve regurgitation is a surgical emergency, requiring urgent intervention, because the valve is likely to become increasingly unstable with acute severe valve regurgitation. The management of lesser degrees of dehiscence is very nuanced. First, it must be clear that the dehiscence is new and related to PVE rather than a long-standing residual of the original surgery. If minimal dehiscence is related to PVE, the interplay of mortality related to surgery versus that related to inadequate treatment of PVE must be assessed. In some patients with minimal dehiscence, serial monitoring during antibiotic therapy rather than early surgery may be preferred.

Identification of evidence of paravalvular invasion (including use of transesophageal echocardiography) is discussed separately. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis" and "Complications and outcome of infective endocarditis".)

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 lesion) experience high mortality rates and are rarely curable with medical treatment alone. Complex reconstructive procedures have been associated with survival rates of 80 percent [26,27]. Patients with these complications are among those most likely to benefit from cardiac surgery. For the most severe of these intracardiac complications, there is generally no benefit from delay of surgery. However, as noted above for prosthetic valve dehiscence, surgery might be appropriately delayed for some echocardiographic changes thought to indicate a paravalvular abscess. For patients with small defects, apparently controlled infection, and minimal valve dysfunction, delay may be appropriate. This may be particularly relevant in the setting of a large embolic stroke or an intracerebral hemorrhage wherein early surgery could risk significant exacerbation of neurologic injury or death. (See "Surgery for left-sided native valve infective endocarditis", section on 'Timing following stroke'.)

Microorganisms usually requiring surgery — Surgery is generally advised for PVE caused by S. aureus when accompanied by intracardiac complications and also for fungi, gram-negative (non-HACEK) microorganisms (particularly Pseudomonas aeruginosa and multidrug-resistant gram-negative bacilli), and multidrug-resistant enterococci.

S. aureus PVE is associated with significant mortality [28-33]. Although earlier guidelines included S. aureus PVE as an indication for early surgery, the indications are based upon presence of specific complications of PVE rather than infection with S. aureus. This approach is supported by a 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 [8]. While subject to the limitations of a retrospective review of prospectively collected data and not from a randomized trial, these observations suggest that the decision for surgical therapy of S. aureus PVE is more complex than previously thought. Surgical intervention is likely to be beneficial in those patients with major intracardiac complications and marked hemodynamic dysfunction [1,9,18,31,32]. Selected patients with S. aureus PVE can be effectively treated with antibiotics alone. In a study of 55 patients with S. aureus PVE, patients who were cured without surgical intervention were characterized by age less than 50 years old, American Society of Anesthesia class III status, and the absence of cardiac, central nervous system, or systemic complications [34]. These findings call into question the general recommendation for surgery in patients with staphylococcal PVE. The role of surgery may be more appropriate for those patients demonstrating overt uncontrolled infection or other intracardiac complications requiring surgical intervention. 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.

Although carefully adjusted analyses to reduce confounding bias are not available to assess surgical treatment of fungal PVE, early surgical intervention is considered by most experts to be a standard element of treatment for this entity [35]. Among 15 patients with fungal PVE treated with antifungal agents plus surgery, 10 (67 percent) survived with an average follow-up of 4.5 years [36]. One review of 17 patients suggested that survival rates for Candida PVE were comparable with and without surgery (46 versus 50 percent) [37]. However, only patients with uncomplicated PVE survived in this review, and many remained on long-term suppressive oral therapy. Patients who have fungal PVE often have significant comorbidities (such as end-stage kidney disease on hemodialysis or immunosuppressed states from chemotherapy or HIV, for example). Timing and duration of medical therapy and surgical intervention are frequently complex.

Embolization — The role of surgery in patients with PVE and recurrent emboli or large mobile vegetation(s) has not been clearly established. The frequency of embolic complications is higher in patients with native valve endocarditis (NVE) who have vegetations exceeding 10 mm in diameter and hypermobile compared with those with smaller, less mobile vegetations. Comparable data are not available for patients with PVE. However, the overall rate of embolic complications is similar for patients with PVE and NVE and, in both conditions, emboli decrease with effective therapy [38]. While prevention of emboli that cause irreversible end-organ damage (eg, central nervous system and myocardial infarction) is a laudable goal, it has not been established that surgical intervention achieves this aim in patients with PVE and large vegetations in the absence of other intracardiac complications. The risk of emboli should be weighed with other conditions that might benefit from surgical intervention. In this setting, surgery should be weighed carefully.

Relapse after optimal medical therapy — PVE relapses after appropriate antimicrobial therapy may reflect unrecognized paravalvular infection [4,18] and thus merit careful re-evaluation, including imaging to assess for new evidence of paravalvular complications (see "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis"). The benefit of surgery versus additional enhanced antibiotic therapy must be weighed carefully.

ANTIBIOTIC THERAPY FOLLOWING SURGERY — Following valve replacement for active bacterial endocarditis, the Task Force on Infective Endocarditis of the European Society of Cardiology recommends another full course (six weeks) of antimicrobial treatment if the intraoperative valve culture is positive [39]. We agree with this approach for culture-positive settings and when an acute necrotizing process is found at surgery. If the culture is negative and pathology does not reveal acute necrotizing infection, we suggest completion of at least the full treatment course (counting the duration of preoperative antibiotics). We favor longer courses of postoperative antibiotic therapy when the new prosthetic valve has been placed at a site where there was overt paravalvular infection.

ANTICOAGULATION — Among patients with prosthetic valve infective endocarditis, the potential benefit of preventing 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".)

COMPLICATIONS AND OUTCOMES — Complications following surgery for PVE reflect the risks of redo surgical valve replacement in the setting of infection frequently associated with paravalvular tissue invasion. Major complications include severe coagulopathy, bleeding or tamponade requiring re-exploration 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 [6]. In the hands of experienced cardiac surgeons, operative mortality rates for patients with paravalvular infection due to PVE, treated with valve replacement and surgical reconstruction of paravalvular tissue, range from 10 to 30 percent; in contrast, the projected mortality of PVE complicated by extensive paravalvular invasion, extensive valve dehiscence, or marked valve dysfunction and progressive HF would approach 100 percent without surgery [1,26,27,40-43]. These 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 [27,40-42].

Among patients with PVE, factors associated with increased mortality during initial hospitalization and at one year following infection include increased age, health care-associated infection, Staphylococcus aureus infection, persistent bacteremia, vegetation, intracardiac abscess, valve dehiscence or severe regurgitation, HF, and stroke [2-4,13,18,19,44,45]. Five-year survival rates ranging from 54 to 82 percent have been reported for patients undergoing surgery for PVE [22,27,40,42,46,47].

The rate of recrudescent PVE after surgery is 6 to 15 percent; repeat surgery is required for recurrent PVE or for dysfunction of the newly implanted prosthesis in 18 to 26 percent [4,22,26,27,40,43]. While the frequencies of these complications are not insignificant, they are relatively small compared with the anticipated mortality with antibiotic therapy alone.

TRANSCATHETER HEART VALVES — Endocarditis has been described in patients with transcatheter aortic or pulmonic heart valves (THV), as discussed separately. The mortality risk associated with THV-PVE is high. For transcatheter aortic valves, thus far, data suggest that considerations similar to those operative in the treatment of surgically placed aortic valve PVE may be reasonable for patients with transcatheter aortic valve replacement-associated PVE, including careful risk/benefit analysis. Management of transcatheter pulmonic valve endocarditis is discussed separately. (See "Transcatheter pulmonary valve implantation", section on 'Endocarditis' and "Transcatheter aortic valve implantation: Complications".)

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

●Treatment of prosthetic valve endocarditis (PVE) is more difficult than treatment of native valve endocarditis and may require surgical replacement of the prostheses in addition to antibiotic therapy. (See 'Introduction' above.)

●Paravalvular abscess is common in PVE, especially when infection arises within 12 months of surgery or involves an aortic prosthesis. (See 'General concepts' above.)

●Indications for cardiac surgery in patients with PVE have been developed based on complications related to intracardiac pathology and inability to eradicate infection. Decisions on whether to proceed with surgical treatment of PVE require careful individualized risk-benefit analysis.

●Early surgery is recommended for patients with PVE with one or more of the following:

•Signs or symptoms of heart failure (HF) resulting from valve dehiscence, intracardiac fistula, or severe prosthetic valve dysfunction.

•IE complicated by heart block, annular or aortic abscess, or destructive penetrating lesion.

•Infection with a highly resistant organism (including fungi).

•Persistent bacteremia despite appropriate antibiotic therapy for five to seven days and exclusion of other sites of infection.

●Early surgery may be helpful in some patients in the following clinical setting, but evidence is limited. Thus for patients in the following clinical settings, we suggest basing a decision on whether or not to proceed with surgery upon weighing individualized estimates of benefit and risk:

•Patients with PVE who have recurrent emboli despite appropriate antibiotic treatment

•Patients with relapsing PVE

•Patients with PVE with mobile vegetations >10 mm

●Observational data indicate that the outcome of PVE in patients who experience severe HF due to prosthesis dysfunction or who have evidence of valve dehiscence, paravalvular abscess, or fistula formation is improved with surgical treatment. There is no evidence that delaying surgery in the setting of deteriorating cardiac function improves outcome.

●Surgery is generally advised for PVE caused by Staphylococcus aureus when accompanied by intracardiac complications and also for fungi, gram-negative (non-HACEK) microorganisms (particularly Pseudomonas aeruginosa and multidrug-resistant gram-negative bacilli), and multidrug-resistant enterococci. (See 'Microorganisms usually requiring surgery' above.)

●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. The projected mortality for patients with PVE with severe paravalvular invasion or HF due to prosthetic valve dysfunction approaches 100 percent without surgery. The rate of recurrent PVE after surgery is 6 to 15 percent. (See 'Complications and outcomes' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff would like to thank Dr. Gabriel S. Aldea for his contributions as Section Editor to prior versions of this topic review.