Antimicrobial therapy of prosthetic valve endocarditis

INTRODUCTION — 

Issues related to the antimicrobial therapy of prosthetic valve infective endocarditis (PVE) will be reviewed here; the content reflects American and European guidelines [1,2].

An overview of the management of infective endocarditis (IE) in adults is presented separately. (See "Overview of management of infective endocarditis in adults".)

General issues related to echocardiography are discussed separately. (See "Role of echocardiography in infective endocarditis".)

Issues related to clinical manifestations and diagnosis of PVE, complications of PVE, and indications for surgery are discussed separately. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis" and "Complications and outcome of infective endocarditis" and "Prosthetic valve endocarditis: Surgical management".)

Issues related to management of native valve endocarditis (NVE) are discussed separately. (See "Antimicrobial therapy of left-sided native valve endocarditis" and "Surgery for left-sided native valve infective endocarditis".)

Issues related to management of cardiac device infections are discussed separately. (See "Infections involving cardiac implantable electronic devices: Epidemiology, microbiology, clinical manifestations, and diagnosis" and "Infections involving cardiac implantable electronic devices: Treatment and prevention".)

Issues related to management of mycotic aneurysm and brain abscess are discussed separately. (See "Overview of infected (mycotic) arterial aneurysm" and "Treatment and prognosis of bacterial brain abscess".)

GENERAL CONSIDERATIONS — 

Management of prosthetic heart valve infection can be challenging; optimal treatment requires [1]:

●Identification of the causative microorganism and selection of an effective bactericidal antimicrobial regimen

●Monitoring for complications of prosthetic valve infective endocarditis (PVE) (see "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis" and "Complications and outcome of infective endocarditis" and "Overview of management of infective endocarditis in adults", section on 'Echo monitoring during therapy')

●Assessing for indications for surgical intervention (eg, when infection has extended beyond the valve to contiguous cardiac tissue, resulting in abscess formation or valve dysfunction) (see "Prosthetic valve endocarditis: Surgical management")

●Management in conjunction with an infectious disease consultant

The approach to antibiotic therapy is the same for PVE involving surgically implanted valves and transcatheter implanted aortic valves (TAVI).

Site of care — The initial management of PVE should occur in the hospital, preferably in an institution where cardiac surgery is available. This is important because treatment of PVE with antimicrobial therapy alone often fails, particularly when infection occurs within the initial six months after valve surgery, is caused by a virulent organism, or is complicated by paravalvular extension or by valve dysfunction.

Patients should remain hospitalized until fever resolves and it is clear that surgery is not needed (or, for patients who undergo surgery that they are stable postoperatively). After initial treatment in the hospital (with resolution of fever and bacteremia, imaging assessment to exclude paravalvular infection, and evaluation regarding the potential need for surgical intervention), consideration may be given to completion of antibiotic therapy in the setting of closely supervised outpatient management. (See 'Completing therapy' below.)

Empiric therapy — After blood cultures have been obtained (three sets from separate venipuncture sites, ideally 30 to 60 minutes apart), empiric broad-spectrum antibiotic therapy with activity against gram-positive organisms (including methicillin-resistant staphylococci and enterococci) and aerobic gram-negative organisms (including Pseudomonas) should be initiated. A regimen including vancomycin (with activity against enterococci and methicillin-resistant staphylococci) and either cefepime or piperacillin-tazobactam (the latter has antipseudomonal activity) would provide such coverage. In the context of healthcare-associated infection with concern for multidrug-resistant gram-negative bacilli or enterococcal infection, gentamicin could be added after weighing the associated risk of nephrotoxicity. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis", section on 'Microbiology'.)

Assessing clinical response to initial therapy

●Defervescence – Most patients with PVE become afebrile three to five days after initiation of appropriate antimicrobial therapy. Patients with Staphylococcus aureus PVE may respond more slowly, remaining febrile for five to seven days after initiation of therapy. Patients with methicillin-resistant S. aureus (MRSA) infection may experience persistent bacteremia during this period.

●Surveillance blood cultures – The initial microbiologic response to therapy should be assessed by obtaining repeat blood cultures 48 hours after antibiotics are begun; thereafter, we obtain two sets of blood cultures every 24 to 48 hours until the bloodstream infection has cleared [1].

●Serial examination for complications – After initiating antimicrobial therapy, careful serial physical examinations and periodic electrocardiograms (ECGs) should be performed to evaluate for signs of heart failure, emboli, or other complications. This is particularly important if there is persistent bacteremia and/or fever for greater than five days; in patients with PVE, this may indicate invasive paravalvular infection or an extracardiac site of undrained infection. Patients who develop new complications while on appropriate antimicrobial therapy (such as new emboli, heart failure, conduction disturbances, heart block, or other complications) should have a repeat echocardiogram to assess for worsening valve dysfunction, cardiac abscess, or fistula. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis".)

Tailoring therapy — Antibiotic therapy should be tailored to culture and susceptibility data once available. (See 'Specific pathogens' below.)

Duration of therapy

●Clinical approach – The optimal duration of therapy for PVE is uncertain. In general, we agree with treatment guidelines issued by the American Heart Association (AHA) and the European Society for Cardiology (ESC), which recommend that PVE be treated with an agent(s) that is bactericidal against the isolated microorganism for at least six weeks after the last positive blood culture [1,2].

This approach is based on observational studies and expert opinion guided by our understanding of the pathogenesis of vegetation formation and pathogen susceptibility. In general, a minimum duration of six weeks is felt to be important given that organisms deep within vegetations or within biofilm are metabolically relatively inactive (and therefore are less responsive to antibiotics) than actively proliferating organisms near the surface. (See "Pathogenesis of vegetation formation in infective endocarditis" and "Prosthetic joint infection: Epidemiology, microbiology, clinical manifestations, and diagnosis", section on 'Biofilm'.)

●Role of shortened duration of parenteral therapy – We favor completion of treatment for PVE using a parenteral regimen, given high morbidity and mortality.

While some have suggested that a shortened duration of parenteral therapy may be sufficient in some cases (as an example, the 2023 ESC guidelines suggest consideration of oral step-down therapy for PVE due to oral streptococci, Streptococcus gallolyticus, Enterococcus faecalis, S. aureus, and coagulase negative staphylococci [2]), in our view the weight of the available evidence has been insufficient to change practice. Completion of parenteral therapy is particularly important for patients with virulent or relatively resistant pathogens, patients with less typical causes of PVE, patients with secondary cardiac or extracardiac complications, and in patients with prolonged duration of infection prior to initiation of therapy.

Effective therapy using a shortened duration of parenteral therapy followed by oral therapy was suggested in a study that included patients with PVE [3]; however, this study had a small sample size and with some biases in patient selection. As examples, many patients had streptococcal IE (approximately 50 percent), and no patients had infective endocarditis (IE) due to MRSA; furthermore, surgical intervention (which has a positive impact on source control) was performed in more than one-third of patients.

Completing therapy

●Outpatient parenteral therapy – Patients may complete intravenous therapy as outpatients once infection has been controlled and they are hemodynamically stable. The patient (or a caregiver) must be capable of managing the technical aspects of intravenous therapy. Such patients require careful monitoring and must have ready access to full medical care should complications occur [1,4]. (See "Outpatient parenteral antimicrobial therapy".)

●Monitoring

•Laboratory and echocardiographic monitoring – While on antimicrobial therapy, patients should be monitored for antimicrobial toxicity. Weekly laboratory monitoring (complete blood count, chemistries, liver, and kidney function tests) and periodic C-reactive protein should be performed. Serial audiograms and therapeutic drug monitoring are warranted for patients receiving long-term aminoglycosides [1].

Issues related to echocardiographic monitoring during therapy are discussed separately. (See "Overview of management of infective endocarditis in adults", section on 'Echo monitoring during therapy'.)

•Clinical monitoring – Patients should be monitored for development of complications related to IE, including embolic events, cardiac conduction disturbances (especially with aortic valve infection in view of the proximity of the valve to the conduction system), and heart failure (see "Complications and outcome of infective endocarditis"). Development of complications should prompt evaluation for cardiac surgery. (See "Prosthetic valve endocarditis: Surgical management".)

•Indications for repeat evaluation – Patients should be counseled regarding the need for immediate evaluation in the setting of new fever, chills, or other signs of systemic toxicity or symptoms suggesting congestive heart failure or neurologic complications. The assessment should include a thorough clinical evaluation, imaging as indicated by symptoms, and repeat blood cultures [1,2]. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis".)

After achieving sterile blood cultures, if breakthrough bacteremia develops while on appropriate antibiotic therapy, the antimicrobial susceptibility of the breakthrough isolate must be reassessed.

Role of surgery — Indications for surgery in patients with PVE are discussed separately. (See "Prosthetic valve endocarditis: Surgical management".)

Follow-up — Issues related to follow-up after completion of antibiotic therapy for IE are discussed separately. (See "Overview of management of infective endocarditis in adults", section on 'Follow-up'.)

Relapse

●Evaluation – Patients with relapse of PVE should undergo repeat diagnostic evaluation, including cardiac imaging for evaluation for extension of infection into paravalvular tissues. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis".)

In addition, patients should also be evaluated for extracardiac sites of infection that may have been inadequately treated.

●Management

•Presence of paravalvular extension – Patients with paravalvular extension of infection warrant surgery, followed by another course of antimicrobial therapy. (See "Prosthetic valve endocarditis: Surgical management".)

•Absence of paravalvular extension – In the absence of paravalvular extension of infection, patients with relapse of PVE following completion of appropriate antimicrobial therapy should receive a repeat course of antibiotics as described in the following sections. In addition, consideration should be given to the role of surgery. Bacterial isolates obtained in the context of relapse should be retested carefully for complete antibiotic susceptibility profiles.

SPECIFIC PATHOGENS — 

Antimicrobial therapy should be adjusted based on culture results, as described in the following sections. Selection of antimicrobial therapy for prosthetic valve infective endocarditis (PVE) is largely based upon clinical experience; in general, the antimicrobial regimen used to treat a specific pathogen causing PVE is based on the treatment used for that organism when it causes native valve endocarditis (NVE); staphylococci are an exception. (See 'Staphylococci' below and "Antimicrobial therapy of left-sided native valve endocarditis".)

If cultures remain negative, therapy for culture-negative PVE should be administered. (See 'Culture negative PVE' below.)

Staphylococci

General principles

●Antimicrobial therapy – The primary consideration in choosing antimicrobial therapy for treatment of staphylococcal PVE hinges upon whether the organism is sensitive to methicillin and other beta-lactam antibiotics (table 1) [5,6].

Combination antimicrobial treatment is advised for treatment of staphylococcal PVE. We agree with the American Heart Association (AHA), the European Society for Cardiology (ESC), and the International Society of Antimicrobial Chemotherapy (ISAC), each of which favors a three-drug regimen consisting of a primary antistaphylococcal agent, gentamicin, and rifampin (if the organism is susceptible to these agents) [1,2,7]. The antistaphylococcal agent and rifampin should be administered for at least six weeks; gentamicin should be administered for the initial two weeks of treatment to prevent the emergence of rifampin resistance [5].

This antibiotic approach is supported by data from animal models of prosthetic device infection and retrospective clinical series [5,8-11] as well as expert opinion. In one retrospective study including 61 patients with staphylococcal PVE who underwent surgery, valves from patients treated preoperatively with combination therapy were 5.9 times more likely to be culture-negative than valves from patients treated with monotherapy; all six patients treated with a triple-drug regimen (including rifampin) had negative valve cultures at surgery [8].

Inclusion of an aminoglycoside and/or rifampin in combination therapy has been questioned by some authors [12-14]; this is discussed further below. (See 'Addition of an aminoglycoside' below and 'Addition of rifampin' below.)

●Role of surgery – In addition to antimicrobial therapy, treatment of staphylococcal PVE may require surgical intervention. In the past, PVE caused by S. aureus was itself an indication for early surgery; however, the risk to benefit ratio of surgical intervention in this setting is more nuanced than previously recognized, and some patients with S. aureus PVE may be successfully treated with medical therapy alone [15-17]. Issues related to surgery for PVE are discussed further separately. (See "Prosthetic valve endocarditis: Surgical management", section on 'Difficult-to-treat pathogens'.)

Primary antistaphylococcal agent

●Methicillin-susceptible isolates

•Clinical approach – For patients with PVE due to methicillin-susceptible S. aureus (MSSA), we favor initial treatment with a semisynthetic penicillinase-resistant penicillin (eg, nafcillin, oxacillin, cloxacillin, or flucloxacillin) to achieve maximum efficacy in the setting of high inoculum infection. After 4 to 10 days of treatment, with eradication of bacteremia, switching to cefazolin for ease of administration and transition to outpatient therapy is reasonable.

•Patient with beta-lactam allergy – For patients with beta-lactam hypersensitivity, allergy consultation for desensitization should be sought to facilitate use of a beta-lactam anti-staphylococcal agent.

If beta-lactam desensitization is not feasible, a regimen for methicillin-resistant S. aureus (MRSA) may be used. We agree with the ESC which favors daptomycin over vancomycin as the primary antistaphylococcal agent (table 1), given lower risk of nephrotoxicity in the setting of concomitant gentamicin [2]. Daptomycin efficacy against MSSA is generally comparable to vancomycin; in addition, the emergence of daptomycin non-susceptibility during therapy appears to be less frequent among MSSA than MRSA isolates [18-22].

●Methicillin-resistant isolates – For isolates resistant to methicillin, vancomycin is the antistaphylococcal agent of choice; emergence of daptomycin non-susceptibility during therapy is more likely to occur among MRSA than MSSA isolates [2,18,20-22].

In the combination regimen, daptomycin may be substituted for vancomycin (if the isolate remains daptomycin susceptible) in the following circumstances:

•Patients with infection due to an MRSA isolate with vancomycin MIC >1.0 mcg/mL, a level which requires toxic vancomycin doses to achieve an optimal vancomycin antibacterial effect [23-25]. (See "Vancomycin: Parenteral dosing, monitoring, and adverse effects in adults".)

•Patients with infection due to a methicillin-heteroresistant S. aureus wherein there is a sub-population with vancomycin MIC >2 mcg/mL.

•Patients with high risk of nephrotoxicity precluding coadministration of vancomycin and gentamicin. If gentamicin is to be avoided entirely, daptomycin should be given in combination with ceftaroline and rifampin.

For patients treated with vancomycin who develop breakthrough bacteremia or have microbiologic failure (including isolation of the causative staphylococcus from an explanted valve or vegetation), the recovered isolate should be retested for susceptibility to vancomycin, daptomycin, and rifampin; resistance to both daptomycin and rifampin has emerged during treatment with vancomycin [5,18,19,26].

●Isolates with reduced vancomycin susceptibility – For treatment of PVE due to staphylococcal isolates with reduced vancomycin susceptibility (minimum inhibitory concentration >1 mcg/mL by broth microdilution or 1.5 mcg/mL by E-test), or for treatment of PVE with inadequate response to vancomycin therapy, the optimal approach is uncertain and clinical experience is limited.

In such cases, we favor combination therapy with daptomycin (8 to 10 mg/kg once daily) plus a beta-lactam (ceftaroline, oxacillin, or nafcillin), which may result in synergy and efficacy even against daptomycin non-susceptible isolates. The ESC suggests combination therapy with daptomycin and either a beta-lactam (cloxacillin or ceftaroline) or fosfomycin (not available for parenteral use in the United States) [2].

Other approaches include monotherapy with daptomycin, ceftaroline, telavancin, or linezolid [27-31].

●Duration of therapy – The duration of treatment with the primary antistaphylococcal agent (and rifampin) is at least six weeks from the time blood cultures become negative. If cultures of an explanted valve or paravalvular abscess are positive, antibiotic therapy should be continued for at least six weeks from the date of surgery. Some clinicians continue antibiotic therapy beyond six weeks for patients with persistent bacteremia or slow resolution of fever (suggesting a delayed response to therapy) or for patients treated with a nonstandard regimen.

Addition of an aminoglycoside

●Clinical approach – An aminoglycoside (usually gentamicin) is used to enhance the antistaphylococcal activity of the primary antistaphylococcal agent, by synergy and/or direct antimicrobial activity [32]. This enhanced killing of staphylococci reduces the likelihood of emergence of a rifampin-resistant subpopulation, and thus preserves rifampin effectiveness throughout treatment.

•Gentamicin-susceptible isolates – Gentamicin is coadministered with a primary antistaphylococcal agent and rifampin; dosing is summarized in the table (table 1).

•Gentamicin-resistant isolates – If the organism is resistant to gentamicin, an alternative aminoglycoside (to which the isolate is susceptible) should be used.

•Aminoglycoside-resistant isolates – If the organism is resistant to all aminoglycosides, we favor treatment as follows:

-MSSA: Daptomycin, rifampin, and a beta-lactam (oxacillin, cloxacillin, or flucloxacillin)

-MRSA: Daptomycin, rifampin, and either ceftaroline (600 mg IV every 8 hours) or fosfomycin (8-12 g/day IV in four divided doses, where available)

An alternative approach consists of treatment with a primary staphylococcal agent in conjunction with rifampin and a fluoroquinolone to which the strain is highly susceptible; this approach is based on data from animal models of endocarditis [33-35].

Use of other agents, such as trimethoprim-sulfamethoxazole and linezolid as an alternative to an aminoglycoside has been described in case reports [35]. Treatment with these and other non-standard regimens requires careful monitoring and should be managed in consultation with infectious disease guidance.

●Rationale – Addition of an aminoglycoside (as a third agent to a regimen including an antistaphylococcal agent plus rifampin) reduces the likelihood of emergence of rifampin resistance.

Intrinsic mutation of staphylococci to rifampin resistance emerges at a relatively high frequency.

In any high inoculum of staphylococci there is likely a rare organism that has mutated spontaneously to rifampin-resistant. This mutant can become the dominant population if exposed to the selective pressure of rifampin. In vitro studies and clinical reports have detected emergence of rifampin resistance when a high-inoculum of staphylococci are exposed to rifampin in the presence of a single, modestly effective antistaphylococcal agent [32,33]. Emergence of rifampin resistance during therapy has been reported in patients treated for S. aureus infections with vancomycin plus rifampin [5,26]. In settings of high staphylococcal inoculum foreign body infections, the goal of adding gentamicin is to enhance the eradication of both rifampin-sensitive and rifampin-resistant organisms by its synergistic interaction with the antistaphylococcal agent and its direct antistaphylococcal activity. This prevents the emergence of a dominant rifampin-resistant staphylococcal population during treatment.

In a small randomized trial including 32 patients with methicillin-resistant coagulase-negative staphylococcal PVE, rifampin resistance emerged in 7 of 19 patients (37 percent) treated with vancomycin plus rifampin but in none of 13 patients treated with vancomycin, gentamicin, and rifampin (p = 0.025) [5].

Addition of rifampin

●Clinical approach – The approach to treatment of staphylococcal PVE consists of initial antibiotic therapy with two antimicrobial agents other than rifampin (eg, a primary anti-staphylococcal agent and gentamicin, if feasible based on antimicrobial susceptibility and kidney function) for three to five days (with clearance of bacteremia) prior to initiation of rifampin. If the susceptibility of the isolate (or other clinical considerations) precludes treatment with two anti-staphylococcal antimicrobials, treatment with a single effective agent should be administered for five days (with clearance of bacteremia) before beginning rifampin.

To avoid selection of a progressively rifampin-resistant staphylococcal population, reduction in the overall staphylococcal inoculum is important. Delaying the initiation of rifampin until the bacterial density at the site of infection is significantly reduced diminishes the likelihood of selecting spontaneous resistant mutants.

If breakthrough bacteremia or microbiologic failure occurs in patients on a rifampin-containing regimen, rifampin susceptibility must be reassessed [5]. If the isolate has become resistant to rifampin, rifampin should be discontinued.

●Rationale – Rifampin is unique in its ability to kill stationary growth-phase staphylococci embedded in biofilm that is adherent to infected foreign material; this is based on in vitro data, evidence from animal models, and clinical observations [5,10,11,33-36].

Use of rifampin is supported by the following studies:

•In a post-hoc analysis of a prospective cohort study including 964 patients with S. aureus bacteremia (including 374 patients with deep-seated infection, of whom 121 had endocarditis), outcomes were compared among those receiving combination antibiotic therapy (most frequently including rifampin) and those not receiving combination therapy [37]. There was no difference in 30-day or 90-day mortality; however, lower mortality was observed in the subgroup with implanted foreign bodies treated with a combination regimen (30-day mortality hazard ratio [HR] 0.53 [95% CI 0.26-1.07] and 90-day mortality HR 0.55, 95% CI 0.33-0.91). In addition, late complications of S. aureus bacteremia, including relapse, were observed less frequently among recipients of combination therapy.

•In a randomized trial including more than 750 patients with S. aureus bacteremia treated with an antistaphylococcal regimen and randomly assigned to addition of rifampin or placebo, a small reduction in clinical and microbiologic recurrence of infection was observed in rifampin- recipient patients with deep infections where source control was not affected [38].

However, retrospective studies [12,14] and a systematic literature review [13] have questioned the benefit of combination antibiotic regimens for treatment of staphylococcal PVE:

•One retrospective study included 180 PVE cases, of which 114 were caused by S. aureus and 66 by coagulase negative staphylococci; methicillin resistance was noted in 17 and 39 patients in these groups, respectively [12]. Overall, in-hospital and one-year mortality rates were 23 and 35 percent, respectively. In comparing the outcome of 101 patients treated with a rifampin-containing regimen (median rifampin duration 33 days [interquartile range 12.5 to 41.2]) to 79 patients not receiving rifampin, the in-hospital and one-year mortality and relapse rates in the two groups were not significantly different. Details of the primary antistaphylococcal therapy were not provided for either study arm. The treatment groups appeared comparable; however, the study was retrospective and thus limited by undetected selection bias.

•Another retrospective study included of 373 patients with staphylococcal PVE (due to either S. aureus [228 patients] or coagulase-negative species [145 patients]), patients were treated with an anti-staphylococcal antibiotic plus and adjunctive agent (275 patients received rifampin; 225 patients received gentamicin) [14]. Based on an unadjusted univariate analysis, no association was observed between recurrent PVE or mortality within one year and use of rifampin (adjusted hazard ratio [HR] 0.77, 95% CI 0.48–1.24) or gentamicin (adjusted HR 1.11, 95% CI 0.71–1.74).

•A literature review stated that the ESC and AHA recommendation for a combination antimicrobial regimen including rifampin was not based on sufficient clinical data, but largely on expert opinion combined with animal model studies [13]. In view of the potential toxicities associated with combination therapy, the authors suggested that the addition of gentamicin and rifampin to an antistaphylococcal antibiotic for treatment of PVE be used primarily within the context of clinical trials.

Given the high mortality rates for staphylococcal PVE (25 percent in-hospital and at least 35 percent at one year), we continue to favor the combination regimens recommended by the AHA and ESC [1,2]. The potential benefits of combination therapy (including rifampin and gentamicin) support this approach, particularly in patients treated medically (in whom infected biofilm may pose a unique challenge). Where there is specific concern regarding aminoglycoside use and nephrotoxicity, the daptomycin-beta-lactam-rifampin regimens are better tolerated alternatives. (See 'Aminoglycoside resistance' below.)

Streptococci

Viridans streptococci and S. bovis/S. equinus complex — For treatment of streptococcal PVE in the United States, the penicillin susceptibility of viridans group streptococci and Streptococcus bovis/Streptococcus equinus complex (includes S. gallolyticus subsp gallolyticus [the major species associated with endocarditis and colon cancer (table 2)]) are defined in the AHA treatment guidelines [1]; ESC guidelines use different minimum inhibitory concentration (MIC) susceptibility parameters (table 3 and table 4) [2]. In the United States, most streptococci that cause endocarditis are highly penicillin susceptible (MIC ≤0.12 mcg/mL); occasional strains are relatively resistant to penicillin (MIC >0.12 mcg/mL and <0.5 mcg/mL), and rare strains are fully resistant to penicillin (MIC ≥0.5 mcg/mL) [39].

Penicillin-susceptible strains

●Clinical approach – For treatment of streptococcal PVE due to penicillin-susceptible strains, we agree with the AHA which recommends a beta-lactam antibiotic (eg, penicillin, ampicillin, or ceftriaxone) for six weeks, with the option to add an aminoglycoside (eg, gentamicin) for the initial two weeks [1]. The addition of an aminoglycoside is relatively contraindicated in the setting of kidney function impairment and cranial nerve VIII dysfunction [5]. Even with the addition of an aminoglycoside, the beta-lactam regimen should be continued for six weeks. An alternative regimen for beta-lactam intolerant patients consists of vancomycin monotherapy for six weeks. Antibiotic dosing is summarized in the table (table 3).

Laboratory testing for high-level gentamicin resistance should be performed prior to initiation of combination therapy, if possible. However, the optimal approach to screening streptococci for high-level aminoglycoside resistance is uncertain, and there is no Clinical and Laboratory Standards Institute (CLSI) guidance. Bacterial growth on brain-heart infusion agar plates (which are commercially available for screening enterococci) in the face of high aminoglycoside concentrations is suggestive of high-level resistance and loss of bactericidal synergy with combination therapy.

●Rationale for aminoglycoside use – Combination beta-lactam-aminoglycoside therapy is used to achieve synergistic killing of streptococci [40]. This approach is based on in vitro studies, in vivo studies using experimental endocarditis models, clinical series, and clinical experience [6,41].

However, there are no studies demonstrating the superiority of combination therapy over penicillin monotherapy. In addition, studies have demonstrated that various genetic pathways mediate high-level resistance to streptomycin (and less frequently to gentamicin) in a small percentage of streptococci that cause endocarditis; in these isolates, the presence of high-level aminoglycoside resistance abrogates the bactericidal synergy of combination therapy [42-45].

Relatively penicillin-resistant strains — The preferred regimen for streptococcal PVE due to relatively penicillin-resistant strains (defined by the AHA as MIC >0.12 mcg/mL and <0.5 mcg/mL) consists of combination therapy with a beta-lactam antibiotic (eg, penicillin, ampicillin, or ceftriaxone) for six weeks and an aminoglycoside (eg, gentamicin) for at least two weeks. Antibiotic dosing is summarized in the table (table 4).

Prior to initiation of combination therapy, laboratory testing for high-level gentamicin resistance should be performed, if possible. (See 'Penicillin-susceptible strains' above.)

The optimal approach to administration of gentamicin (dosing and duration) is uncertain. To reduce the risk of kidney injury, we favor administration of gentamicin as a single daily dose (rather than in three divided doses) for a minimum of two weeks [2]. We favor this approach based on limited data from patients with NVE with these organisms and the increased risk of aminoglycoside nephrotoxicity and cranial nerve VIII toxicity with longer courses of therapy.

If aminoglycoside therapy is relatively contraindicated (eg, in the setting of kidney function impairment or cranial nerve VIII dysfunction), or the strain is gentamicin resistant, monotherapy with a beta-lactam antibiotic may be given [5]. An alternative regimen consists of vancomycin monotherapy for six weeks. Daptomycin should not be substituted for vancomycin. (See 'Penicillin allergy' below.)

Fully penicillin-resistant strains — For patients with PVE due to fully penicillin-resistant streptococci (defined by the AHA as MIC ≥0.5 mcg/mL), management consists of combination therapy with a beta-lactam antibiotic (eg, penicillin, ampicillin, or ceftriaxone) and, in the absence of high-level resistance, an aminoglycoside (eg, gentamicin), both for six weeks. We favor administration of gentamicin as a single daily dose (rather than in three divided doses). Antibiotic dosing is summarized in the table (table 4).

If the patient is intolerant of penicillin and ceftriaxone, or if gentamicin cannot be used, vancomycin monotherapy should be used; patients with NVE caused by streptococci that are very highly resistant to penicillin have been cured with vancomycin alone [46]. In general, we do not favor combination treatment with vancomycin plus gentamicin (despite synergistic bactericidal activity against penicillin resistant streptococci), given increased risk of nephrotoxicity with this approach.

Penicillin allergy — For patients with immediate-type reactions (urticaria or anaphylaxis) to penicillin, vancomycin is an acceptable alternative regimen. For patients with non-immediate type reactions, cefotaxime or ceftriaxone may be used. (See "Allergy evaluation for immediate penicillin allergy: Skin test-based diagnostic strategies and cross-reactivity with other beta-lactam antibiotics".)

Daptomycin should not be used routinely as an alternative to vancomycin for treatment of endocarditis caused by non-speciated viridans streptococci; emergence of stable high-level resistance to daptomycin has been noted in at least 25 percent of the Streptococcus mitis group isolates (S. mitis, Streptococcus oralis, and Streptococcus sanguinis) with exposure to daptomycin in vitro and in experimental endocarditis models. Resistance has also emerged in Streptococcus parasanguinis, Abiotrophia spp, and Granulicatella adiacens [46-48]. In vitro and endocarditis model studies suggest that the likelihood of emergence of daptomycin resistance in streptococci may be diminished when daptomycin is combined with gentamicin or ceftriaxone, although this approach would not be recommended therapy [49,50].

Other streptococci

●Streptococcus pneumoniae – For treatment of PVE due to penicillin-susceptible Streptococcus pneumoniae (in the absence of meningitis), the regimen consists of penicillin or ceftriaxone for six weeks, with the option to add an aminoglycoside (eg, gentamicin) for the initial two weeks (as for penicillin-susceptible viridans streptococci) (table 3). (See 'Penicillin-susceptible strains' above.)

For treatment of PVE due to S. pneumoniae with relative or full penicillin resistance (MIC ≥0.12 mcg/mL) in the absence of meningitis, the regimens in the table may be used (table 3). In such cases an aminoglycoside may be added, but is not required simply because of the elevated MIC.

For treatment of PVE due to penicillin-resistant pneumococci complicated by meningitis, the regimen must be adjusted to ensure adequate antimicrobial penetration into the cerebrospinal fluid (with meningeal doses of ceftriaxone, vancomycin, or both, contingent on susceptibility). (See "Initial therapy and prognosis of community-acquired bacterial meningitis in adults".)

●Streptococci belonging to serogroups A, B, C, F, and G

•Serogroup A – For treatment of PVE due to Group A Streptococcus, the regimen consists of penicillin or ceftriaxone for six weeks; gentamicin is not used for serogroup A because these strains are highly susceptible to penicillin.

•Serogroups B, C, F, and G – For treatment of PVE due to streptococci belonging to serogroups B, C, F, and G, the regimen consists of penicillin or ceftriaxone for six weeks, with the option to add an aminoglycoside (eg, gentamicin) for the initial two weeks (as for penicillin-susceptible viridans streptococci) (table 3). (See 'Penicillin-susceptible strains' above.)

Streptococci-like organisms — Streptococci-like organisms include Abiotrophia defectiva, Granulicatella spp, and Gemella spp. The approach to treatment of PVE caused by these organisms is the same as the approach for fully penicillin-resistant strains. (See 'Fully penicillin-resistant strains' above.)

A regimen of high-dose penicillin or ampicillin plus gentamicin may be used initially; confirming the antimicrobial susceptibility of the isolate is important in tailoring therapy. Among Granulicatella adiacens, 40 percent are penicillin susceptible but penicillin susceptibility does not predict ceftriaxone susceptibility; ceftriaxone must be tested specifically [51]. In contrast, for Abiotrophia defective, ceftriaxone susceptibility is typical but does not predict penicillin susceptibility. Gentamicin should be administered for at least the first two weeks.

Enterococci — Among the enterococci species, Enterococcus faecalis is the most common human pathogen and the species that most commonly causes endocarditis. Enterococcus faecium is a far less frequent cause of endocarditis but is the most antimicrobial resistant species. Other species are rare causes of endocarditis and their antibiotic susceptibility is less predictable. Speciation will facilitate optimal planning of therapy. Given increasing antibiotic resistance among enterococci, strains causing endocarditis must be tested carefully to select a synergistic regimen [52,53].

In general, bactericidal activity against enterococci (which is needed for effective treatment of endocarditis) can be achieved via synergistic combination of two beta-lactam agents or combination of an aminoglycoside with a cell wall agent; these are discussed further below. (See 'Susceptible strains' below.)

Susceptible strains

●General approach – Bactericidal synergistic activity against enterococci can be achieved via combination of ampicillin and ceftriaxone or combination of an aminoglycoside with a cell wall agent. We favor the ampicillin plus ceftriaxone combination regimen, particularly in patients with pre-existing kidney dysfunction or in whom the risk of aminoglycoside nephrotoxicity or oto/vestibular toxicity is a concern [54-56].

•Beta-lactam combination regimen – Bactericidal synergy against E. faecalis (but not E. faecium) can be achieved by combining ampicillin plus ceftriaxone (table 5) [57,58].

Exposure of E. faecalis to high concentrations of ampicillin plus ceftriaxone results in bactericidal synergy through the saturation of penicillin binding proteins 2, 3, 4, and 5; this synergy is achieved irrespective of high-level resistance to aminoglycosides.

In retrospective studies, the efficacy of this regimen has been comparable to the aminoglycoside combination regimen and has been associated with significantly less frequent nephrotoxicity [54-56,59,60]. Data supporting use of a beta-lactam combination regimen are discussed further separately. (See "Antimicrobial therapy of left-sided native valve endocarditis", section on 'Susceptible strains' and "Treatment of enterococcal infections".)

•Aminoglycoside combination regimen – Bactericidal activity against enterococci can be achieved with the synergistic interaction of a cell wall active agent (penicillin, ampicillin, or vancomycin) and an aminoglycoside (gentamicin or streptomycin) [1,2]. Gentamicin is administered as a single daily dose for two weeks when combined with penicillin or ampicillin (table 5); it is administered in three divided doses for six weeks if used in combination with vancomycin (table 6) [2].

-Synergy and resistance testing – The cell wall agent is bacteriostatic against enterococci but promotes entry of the aminoglycoside into the cell, where it is bactericidal [59]. To achieve synergistic activity, the organism (E. faecalis or E. faecium) must be susceptible to the cell wall active agent at achievable serum concentrations and must be susceptible to gentamicin (at 500 mcg/mL) or streptomycin (at 1000 mcg/mL in broth or at 2000 mcg/mL on agar culture).

High-level resistance to gentamicin and streptomycin is mediated by two independent genes; therefore, organisms need to be tested for high-level resistance to each of these drugs to fully establish potential for synergism. Growth in the presence of the aminoglycoside at the above concentrations indicates high-level resistance and precludes a synergistic bactericidal effect when that aminoglycoside is combined with a cell wall active agent. In addition, gentamicin resistance at the above concentration indicates that synergy cannot be achieved with netilmicin, tobramycin (which is never synergistic even when there is no high-level gentamicin resistance), amikacin, or kanamycin.

-Mitigating nephrotoxicity – To mitigate potential aminoglycoside nephrotoxicity, gentamicin may be administered for two weeks [61,62].

This approach is supported by a study of 93 patients with enterococcal endocarditis including 27 cases of PVE; clinical cure was achieved in 78 percent of cases; among cured patients, a cell wall active agent (primarily ampicillin) was given for a median of 42 days and the synergistic aminoglycoside was given for a median of 15 days [61]. In another study including 84 patients with enterococcal endocarditis, the outcome of patients treated with ampicillin for six weeks and gentamicin for 14 days was comparable to that of patients treated ampicillin for six weeks with plus gentamicin for a median of 28 days (range 18 to 42 days) [62].

There is scant data for use of brief courses of gentamicin combined with vancomycin for treatment of enterococcal endocarditis. Accordingly, an abbreviated duration gentamicin is not recommended when vancomycin is the cell wall active agent.

Use of an aminoglycoside combination regimen is based on in vitro studies, animal models, and clinical series [59,63]. (See "Treatment of enterococcal infections".)

●Penicillin allergy – For patients with immediate-type hypersensitivity to beta-lactams, options include:

•Desensitization to penicillin, to allow treatment with ampicillin plus ceftriaxone (or a beta-lactam plus an aminoglycoside); this is preferred, if feasible (table 5).

•Treatment with vancomycin plus gentamicin for six weeks, if kidney function allows (table 6). There is very little published data to support a shorter course of gentamicin when combined with vancomycin.

Patients with a less definitive history of beta-lactam allergy should be evaluated by an allergist to clarify the history and facilitate use of the preferred beta-lactam based therapies if feasible.

Aminoglycoside resistance

●PVE due to Enterococcus faecalis

•Gentamicin resistance – For PVE caused by E. faecalis strains susceptible to penicillin, vancomycin, and streptomycin but with high-level resistance to gentamicin, we favor treatment with a beta-lactam combination regimen (ceftriaxone plus ampicillin for six weeks) (table 7). (See 'Susceptible strains' above.)

For strains that do not exhibit high-level resistance to streptomycin, the combination of penicillin or ampicillin plus streptomycin will provide bactericidal synergy and may be used as an alternative regimen, provided that the patient does not have pre-existing kidney dysfunction, cranial nerve VIII dysfunction, or visual impairment. There is a risk of irreversible cranial nerve VIII toxicity with four to six weeks treatment. Dosing is summarized in the table (table 7).

•Resistance to gentamicin and streptomycin – If the organism exhibits high-level resistance to both streptomycin and gentamicin, no available aminoglycoside will provide bactericidal synergy; thus, an aminoglycoside should not be administered. In such cases, the beta-lactam combination regimen should be used (table 7).

In the setting of beta lactam intolerance, an allergy consultation should be pursued for desensitization to allow use of the beta-lactam combination regimen (table 7) or use of daptomycin in combination with a beta-lactam (table 8) [54-56].

Prolonged treatment with a cell wall active agent alone (eg, ampicillin for 8 to 12 weeks) may be effective in some cases, but the risk of relapse is very high [60].

●PVE due to Enterococcus faecium – E. faecium is a relatively infrequent cause of PVE; when this species does cause infection, it is typically resistant to multiple antibiotics, and optimal treatment is not certain. (See 'Resistance to penicillins, aminoglycosides, and vancomycin' below.)

Penicillin resistance — Intrinsic high-level penicillin or ampicillin resistance in enterococci is usually due to alterations in penicillin-binding proteins. For patients with PVE due to enterococcal strains with intrinsic high-level penicillin resistance, we favor combination therapy with vancomycin for six weeks plus gentamicin for six weeks (duration as permitted by kidney function) [2,64]. Dosing is summarized in the table (table 6). To reduce the risk of nephrotoxicity, treatment with daptomycin plus ceftaroline or ampicillin are potential alternative regimens (table 8) [59,64].

Occasionally, E. faecalis may be resistant to penicillin and ampicillin by virtue of beta-lactamase production; if this is suspected, confirmation by testing for beta-lactamase can be requested. In such cases, vancomycin or ampicillin-sulbactam or amoxicillin-clavulanate could be used as the cell wall active agent in a combination regimen with an aminoglycoside, contingent on absence of high-level aminoglycoside resistance.

Resistance to penicillins, aminoglycosides, and vancomycin — For PVE caused by vancomycin-resistant E. faecium (VRE) organisms (which often are also resistant to penicillin and ampicillin and highly resistant to gentamicin and streptomycin), the optimal approach is uncertain (table 8). A full evaluation of the isolate's resistance profile is required for selection of therapy. Occasionally, a standard regimen is available. Infectious disease consultation is advised.

●Role of daptomycin – In general, daptomycin is active against enterococci, including VRE [39]. Clinical experience with daptomycin treatment of enterococcal PVE is limited, raising important concerns regarding reporting bias [65].

Daptomycin combination therapy is preferable over monotherapy. Daptomycin should be administered in high doses (10 to 12 mg/kg IV once daily) and combined with either ampicillin, ceftaroline, ertapenem, or fosfomycin (parenteral fosfomycin not available in the United States) [2]. Combining daptomycin with one of these agents may enhance the bactericidal activity of daptomycin (even against daptomycin-resistant enterococci); however, this effect is difficult to predict [64,66-68].

●Role of linezolid – Linezolid is often active against E. faecium and E. faecalis, but its effectiveness in the treatment of PVE caused by VRE is not fully established [1,69].

Quinupristin-dalfopristin is active against E. faecium but causes unacceptable infusion-related toxicity that usually precludes routine use.

Patients with PVE caused by highly resistant enterococci should be managed with the assistance of infectious disease consultation [59]. In addition, consultation regarding surgical intervention during suppressive bacteriostatic therapy is warranted. Surgery may also be warranted for patients not responsive to antimicrobial therapy. (See "Prosthetic valve endocarditis: Surgical management".)

HACEK organisms — HACEK organisms are fastidious gram-negative bacilli:

●Haemophilus species (including H. parainfluenzae, the most likely cause of endocarditis among HACEK organisms)

●Aggregatibacter species [including Aggregatibacter actinomycetemcomitans (previously called Actinobacillus actinomycetemcomitans), Aggregatibacter aphrophilus (previously called Haemophilus aphrophilus), and Aggregatibacter paraphrophilus (previously called Haemophilus aphrophilus)]

●Cardiobacterium species (including Cardiobacterium hominis)

●Eikenella corrodens

●Kingella species (including Kingella kingae)

These organisms are highly susceptible to third generation cephalosporins; some are ampicillin-resistant due to the production of beta-lactamase. Patients with HACEK PVE in the absence of valvular dysfunction can often be cured with antibiotics alone [70].

For treatment of HACEK PVE, treatment with ceftriaxone (or a comparable third- or fourth-generation cephalosporin, such as cefotaxime or cefepime, respectively) for six weeks is preferred (table 9) [1,2].

An alternative regimen in the ESC guidelines consists of ampicillin for six weeks in combination with gentamicin for the initial two weeks [2]. This regimen should be used only if growth in vitro is adequate for reliable beta-lactam susceptibility testing, confirmation of ampicillin susceptibility, and absence of beta-lactamase production. In addition, the potential for aminoglycoside nephrotoxicity should be considered carefully.

For patients unable to tolerate either of the above regimens, ciprofloxacin or another fluoroquinolone may be used [1,2].

Other gram-negative organisms — Treatment of PVE caused by non-HACEK gram-negative bacilli should be based upon the susceptibility of the causative organism.

Combination antimicrobial therapy with a beta-lactam (penicillin, cephalosporin, or carbapenem) and an aminoglycoside (and in some cases, with the addition of a fluoroquinolone or trimethoprim-sulfamethoxazole) for six weeks is reasonable [1,2].

Maximal antimicrobial doses of (eg, rather than low doses used for synergy) should be used, in an attempt to expose a potentially high bacterial inoculum deep within vegetations to high antimicrobial concentrations. The rationale for combination therapy is to minimize the likelihood of antibiotic resistance in the setting of a potentially large bacterial burden within the protected environment of a vegetation.

Surgery to excise the infected valve is often required in PVE due to gram-negative bacilli, especially that caused by Pseudomonas aeruginosa or when infection involves the left-sided heart valves. (See "Prosthetic valve endocarditis: Surgical management".)

Corynebacteria (diphtheroids) — Sporadic cases of endocarditis due to Corynebacterium species have been reported. These organisms should not be dismissed as contaminants when isolated in blood cultures from a patient with a prosthetic valve. In addition, antibiotic resistance to beta-lactam agents has become common in Corynebacteria spp, particularly among C. striatum and C. jeikeium, two of the more common species causing endocarditis [71]. Susceptibility to vancomycin remains nearly universal: however, rapid emergence of resistance to daptomycin has been reported.

Pending susceptibility test results, vancomycin is the primary agent for treatment [72-74]. Earlier studies indicated that for organisms susceptible to gentamicin (MIC <4.0 mcg/mL), penicillin plus gentamicin resulted in synergistic bactericidal activity [30]. However, this has not been confirmed in more recent isolates with increased penicillin resistance. Thus, penicillin combined with gentamicin may be used for treatment only if the causative isolate is confirmed to be susceptible to both of these antibiotics.

Antibiotic treatment should be administered for six weeks. Corynebacterial PVE is a highly morbid infection; many patients will develop indications for surgical intervention [72-74]. (See "Prosthetic valve endocarditis: Surgical management".)

Fungi — Treatment of fungal endocarditis consists of antifungal therapy and generally valve replacement.

Issues related to Candida endocarditis are discussed separately. (See "Candida endocarditis and suppurative thrombophlebitis".)

Issues related to endocarditis caused by Aspergillus are discussed separately. (See "Treatment and prevention of invasive aspergillosis", section on 'Antifungal therapy' and "Treatment and prevention of invasive aspergillosis", section on 'Role of surgery' and "Treatment and prevention of invasive aspergillosis".)

CULTURE NEGATIVE PVE

●Etiology and clinical approach – Culture-negative infective endocarditis (IE) is defined as endocarditis without etiology following inoculation of three blood samples in a standard blood culture system (eg, negative cultures after seven days) [1].

Cultures are negative in IE for three major reasons:

•Administration of antimicrobial agents prior to obtaining blood cultures

•Inadequate microbiologic techniques

•Infection with fastidious bacteria or nonbacterial pathogens

In the absence of prior antibiotic exposure, the most common causes of culture-negative IE are fastidious organisms (eg, fastidious bacteria, zoonotic agents, and fungi); in the setting of prior antibiotic exposure, the most common causes are Streptococcus spp, staphylococci, and enterococci.

Several pathogens warrant special consideration in culture-negative PVE:

•Coxiella burnetii, the causative agent of Q fever, is a relatively common cause of culture-negative endocarditis and has a unique predisposition for prosthetic valves; the frequency varies in different geographic locations. Further information regarding Q fever can be found on the United States Centers for Disease Control and Prevention website [75,76]. (See "Chronic Q fever, including endocarditis".)

•Fungal endocarditis should be considered, especially in patients with a complex perioperative course or a known, even somewhat remote, prior episode of candidemia. (See 'Fungi' above.)

•Rarely, culture-negative PVE due to atypical mycobacteria has been described. Reports have described valve contamination during manufacture. Mycobacterium chimaera has caused culture-negative PVE, although this organism can be recovered from blood when cultures are placed in special media. This infection has occurred as a consequence of intraoperative wound contamination by aerosol from contaminated heater-cooler machines used during cardiopulmonary bypass. These cases can have an indolent, markedly delayed onset (eg, years after valve replacement surgery) and can present with multiple foci of infection [77]. (See "Overview of nontuberculous mycobacterial infections", section on 'M. chimaera associated with cardiac surgery'.)

Many other pathogens can present as culture-negative PVE. Diagnostic evaluations as described in other sections should be pursued. (See "Prosthetic valve endocarditis: Epidemiology, clinical manifestations, and diagnosis", section on 'Culture-negative endocarditis' and "Blood culture-negative endocarditis: Epidemiology, microbiology, and diagnosis".)

●Management – Empiric treatment of patients with culture-negative PVE should be pursued in consultation with an infectious disease specialist. We are in agreement with the American Heart Association (AHA), which favors the following initial approach while awaiting results from an expanded diagnostic evaluation [1]:

•Onset ≤1 year after surgery – Patients with culture-negative PVE with onset ≤1 year after surgery should receive antimicrobial therapy for coverage of infection due to staphylococci, enterococci, and aerobic gram-negative bacilli. Broad coverage can be achieved using an empiric regimen with vancomycin as well as cefepime or piperacillin-tazobactam. With an acute septic onset, gentamicin can be added to further enhance coverage, with care to avoid nephrotoxicity. Addition of rifampin should be delayed at least three to five days after start of therapy.

•Onset >1 year after surgery – Patients with culture-negative PVE with onset >1 year after surgery should receive antimicrobial therapy for coverage of infection due to staphylococci, viridans group streptococci, enterococci, and Haemophilus, Aggregatibacter, Cardiobacterium, Eikenella, Kingella (HACEK) organisms. Empiric therapy could be initiated with vancomycin and ceftriaxone; this regimen provides initial empiric coverage but is not adequate treatment for enterococcal infection.

Patients with nosocomial or hospital-associated culture-negative PVE should be treated empirically as outlined above for patients with onset ≤1 year after valve implantation.

If additional diagnostic evaluations identify the etiologic agent, antibiotic therapy should be targeted accordingly. If negative blood cultures may be attributable to prior antimicrobial exposure, consideration should be given to the prior antimicrobial's spectrum of activity and the patient's relevant epidemiology to select empiric therapy.

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: Outpatient parenteral antimicrobial therapy".)

SUMMARY AND RECOMMENDATIONS

●General considerations − Management of prosthetic valve endocarditis (PVE) requires bactericidal antimicrobial therapy, monitoring for complications, and assessment of indications for surgical intervention. Specific recommendations are based primarily upon clinical experience and in vitro studies, rather than on clinical trial data. (See 'General considerations' above.)

●Empiric therapy – After blood cultures have been obtained (three sets), empiric broad-spectrum antibiotic therapy should be initiated. We suggest an empiric regimen active against gram-positive and gram-negative bacteria (Grade 2C). Vancomycin (with activity against enterococci and methicillin-resistant staphylococci), and either cefepime or piperacillin-tazobactam would provide such coverage. In the context of healthcare-associated infection with concern for multidrug-resistant gram-negative bacilli or enterococcal infection, gentamicin could be added after weighing the associated risk of nephrotoxicity. (See 'Empiric therapy' above.)

●Targeted therapy − Therapy for PVE should be targeted to the organism isolated from blood cultures, along with the in vitro susceptibility results. The approach is summarized in the tables and is discussed in the sections above:

•Staphylococci (table 1) − For treatment of staphylococcal PVE, we suggest combination therapy with a three-drug regimen (including a primary antistaphylococcal agent, gentamicin, and rifampin), rather than one- or two-drug regimens (Grade 2C), given high mortality rates for staphylococcal PVE; the potential for adverse events must be monitored carefully. (See 'Staphylococci' above.)

We begin treatment with the antistaphylococcal agent plus gentamicin. After three to five days of treatment, with eradication of bacteremia, rifampin can be added.

-Methicillin-susceptible S. aureus (MSSA) – For patients with PVE due to MSSA, we favor initial treatment with a semisynthetic penicillinase-resistant penicillin (eg, nafcillin, oxacillin, cloxacillin, or flucloxacillin). Changing the anti-staphylococcal penicillin to cefazolin for ease of administration and transition to outpatient therapy is reasonable after control of infection.

For patients with PVE due to MSSA in the setting of beta-lactam hypersensitivity, allergy consultation for desensitization should be sought to facilitate use of a beta-lactam anti-staphylococcal agent. If this is not feasible, we favor daptomycin as the primary antistaphylococcal agent (rather than vancomycin), given lower risk of nephrotoxicity in the setting of concomitant gentamicin. (See 'Primary antistaphylococcal agent' above and 'Addition of an aminoglycoside' above.)

-Methicillin-resistant S. aureus (MRSA) – For patients with PVE due to MRSA, we favor vancomycin as the primary antistaphylococcal agent; emergence of daptomycin non-susceptibility is more likely to occur among MRSA than MSSA isolates. (See 'Primary antistaphylococcal agent' above and 'Addition of an aminoglycoside' above.)

•Viridans streptococci and Streptococcus bovis/Streptococcus equinus complex (table 3 and table 4) – For patients with streptococcal PVE due to strains that are relatively resistant to penicillin (defined by American Heart Association [AHA] as minimum inhibitory concentration [MIC] >0.12 and <0.5 mcg/mL; defined by European Society of Cardiology [ESC] as MIC ≥0.25 mcg/mL) or fully resistant to penicillin (defined by the AHA as MIC ≥0.5 mcg/mL), the preferred regimen consists of combination therapy with a beta-lactam antibiotic (penicillin, ampicillin, or ceftriaxone) for six weeks, plus gentamicin.

Our approach to the duration of gentamicin depends on the degree of penicillin resistance. For relatively penicillin-resistant strains, we administer gentamicin for at least two weeks, to minimize the risk of nephrotoxicity and oto/vestibular toxicity. For fully penicillin-resistant strains, we administer gentamicin for six weeks. (See 'Viridans streptococci and S. bovis/S. equinus complex' above.)

•Enterococci (table 5 and table 6 and table 7 and table 8) (See 'Enterococci' above.)

-For treatment of PVE due to susceptible strains of Enterococcus faecalis, we suggest combination therapy with ampicillin plus ceftriaxone since it avoids the toxicity of aminoglycosides (Grade 2C).

-For patients with PVE due to susceptible strains of E. faecalis who are treated with the combination regimen of penicillin or ampicillin plus gentamicin, we favor administration of gentamicin for two weeks as a single daily dose, to minimize the risk of nephrotoxicity and oto/vestibular toxicity. (See 'Enterococci' above.)

•Other pathogens – Suggested regimens are provided for the following pathogens in the relevant sections and tables above:

-Streptococcus pneumoniae and streptococcal groups A, B, C, F, and G (see 'Other streptococci' above)

-Streptococci-like organisms (Abiotrophia defectiva, Granulicatella spp, and Gemella spp) (see 'Streptococci-like organisms' above)

-HACEK (Haemophilus, Aggregatibacter, Cardiobacterium, Eikenella, Kingella) organisms (table 9) (see 'HACEK organisms' above)

-Other gram-negative organisms (see 'Other gram-negative organisms' above)

-Corynebacteria (diphtheroids) (see 'Corynebacteria (diphtheroids)' above)

-Fungi (see 'Fungi' above)

-Culture-negative endocarditis (CNE) (see 'Culture negative PVE' above)

●Treatment duration − We suggest that patients with PVE be treated with parenteral antibiotics for at least six weeks rather than for shorter durations (Grade 2C). This approach is based upon our understanding of the pathogenesis of vegetation formation and pathogen susceptibility, rather than on an evidence-based approach based on outcome data. Some data suggest that using oral antimicrobials to complete the late phase of treatment may be effective and allow shortening the duration of intravenous therapy in some cases. However, thus far, we judge that the weight of this evidence has been insufficient to change antibiotic treatment of PVE. (See 'Duration of therapy' above.)

●Monitoring during treatment − Patients with PVE require careful regular clinical follow-up including serial physical examinations and obtaining follow-up blood cultures to document clearance of bacteremia. If breakthrough bacteremia develops while on appropriate antibiotic therapy, the antimicrobial susceptibility of the breakthrough isolate should be reassessed. Patients should be monitored for development of complications including embolic events and heart failure. Development of complications should prompt evaluation for cardiac surgery. (See 'Assessing clinical response to initial therapy' above and 'Completing therapy' above.)