Version July 2025
INTRODUCTION —
Prosthetic valve endocarditis (PVE) refers to infection of surgically implanted cardiac valves or transcatheter implanted aortic valves (TAVI) or other non-CIED intracardiac foreign bodies [1-4]. PVE accounts for 20 to 30 percent of endocarditis cases reported from tertiary medical centers and 4 percent of endocarditis cases recorded in a United States national database between 2003 and 2017 [4-6]. Patients with PVE experience a high in-hospital or 30-day mortality rate (20 percent for surgically implanted valves and 36 percent for TAVI) [4-6].
The pathogenesis, epidemiology, microbiology, pathology, clinical manifestations, and diagnosis of PVE will be reviewed here.
Issues related to management of PVE (including antimicrobial therapy and surgery) are discussed separately, as are issues related to prevention. (See "Antimicrobial therapy of prosthetic valve endocarditis" and "Prosthetic valve endocarditis: Surgical management" and "Prevention of endocarditis: Antibiotic prophylaxis and other measures".)
General issues related to TAVI are discussed separately. (See "High-gradient aortic stenosis in adults: Indications for valve replacement", section on 'Choice of surgical or transcatheter intervention' and "Transcatheter aortic valve implantation: Periprocedural and postprocedural management" and "Transcatheter aortic valve implantation: Complications".)
SURGICAL VALVE IMPLANTATION
Timing and mechanisms — The timing of PVE involving surgically implanted valves reflects different pathogenic mechanisms; this, in turn, influences the epidemiology, microbiology, pathology, and clinical manifestations of these infections [7]. (See 'Microbiology' below.)
●Early infection – Cases occurring during the initial two months after surgery are largely nosocomial or health care-associated; in such cases, microorganisms reach the prosthetic valves via direct intraoperative contamination or via hematogenous spread during the days and weeks after surgery.
Cases occurring 2 to 12 months after surgery are a blend of health-care associated infections (delayed-onset) and community-acquired infections. (See 'Microbiology' below.)
Early after surgery, the valve sewing ring, valve annulus, and anchoring sutures are not yet covered with endothelium, but are coated with host proteins (such as fibronectin and fibrinogen) to which organisms can adhere. Therefore, adherent organisms have direct access to the prosthesis-annulus interface and to perivalvular tissue along suture pathways. Consequently, perivalvular infection, including abscesses and prosthesis dehiscence, are common with early infection of both mechanical and bioprosthetic valves [8].
●Late infection – The pathogenesis of late PVE (>12 months postoperatively) has been postulated to resemble that of native valve endocarditis (NVE). (See "Pathogenesis of vegetation formation in infective endocarditis".)
As the sewing ring, sutures, and adjacent tissues become endothelialized and as alterations occur in the surface and flow characteristics of bioprosthetic valve leaflets, platelet-fibrin microthrombi are deposited. These microthrombi serve as hospitable surfaces to which organisms adhere.
The pathogens associated with late PVE tend to be bacteremic isolates able to survive serum bactericidal activity and adhere to these microthrombi. They are similar to those causing NVE. (See "Native valve endocarditis: Epidemiology, risk factors, and microbiology", section on 'Microbiology'.)
With time after surgery, the perivalvular tissues are somewhat protected from infection by overlying endothelium. Therefore, unless the infecting organism is Staphylococcus aureus or another highly virulent or invasive pathogen, infection is less likely to extend into the perivalvular tissues. Accordingly, late-onset infections are less often complicated by perivalvular abscess and valve dehiscence and are more commonly restricted to the sewing ring or the bioprosthetic leaflet.
Epidemiology
●PVE incidence – PVE occurs in 1 to 6 percent of cases, with an incidence of 0.3 to 1.2 percent per patient-year (PY) [1,4,9,10]. In multinational series (primarily from tertiary medical centers), PVE accounts for 20 to 30 percent of endocarditis cases [4,5]. In patients who have undergone surgical valve replacement, PVE occurs with equal frequency at aortic and mitral sites.
Among patients who underwent surgical aortic valve replacement (SAVR) in Denmark between 1996 and 2015, the cumulative incidence of PVE was 6.0 per 1000-PY; this is lower than the incidence of infective endocarditis (IE) in individuals with a prior history of endocarditis (16.1 out of 1000 PY), but higher than the incidence of IE in matched controls who did not undergo valve replacement (hazard ratio [HR] 19.1, 95% CI 15.0-24.4) [7].
●Bioprosthetic valve PVE more common than mechanical valve PVE – PVE occurs more frequently on bioprosthetic valves than mechanical valves. In data from the Swedish National Patient Registry for patients age 50 to 69 years old who underwent aortic valve replacement between 1997 and 2012, the rate of PVE was higher among recipients of bioprosthetic valves (8.6 percent; followed for a mean of 5.0 years) than among recipients of mechanical valves (7.3 percent; followed for a mean of 8.8 years) [11]. The Swedish data also show a greater incidence of PVE among recipients of bioprosthetic valves than mechanical valves at multiple time points (1 year, 2 to 5 years, and 6 to 10 years) (table 1) [12,13].
Similarly, in an American observational study including more than 38,000 patients ≥65 years of age with prosthetic valves implanted between 1991 and 1999, the cumulative risk of endocarditis at 12-year follow-up was higher among those with bioprosthetic valves than those with mechanical valves (2.2 versus 1.4 percent; adjusted HR 1.60, 95% CI 1.31-1.95) [14].
●Health care-associated infection – In addition to typical nosocomial early onset PVE, health care-associated PVE resulting from infection acquired in outpatient health care settings or in the context of ongoing invasive care occurs with significant frequency.
In one cohort study including more than 550 patients with PVE, health care-associated infection (non-nosocomial) was defined as PVE diagnosed within 48 hours of admission in a patient with extensive nonhospital health care contact (defined as follows) [4]:
•Intravenous (IV) therapy, wound care, or specialized nursing care at home or IV chemotherapy within the prior 30 days
•Residence in a nursing home or other long-term care facility
•Hospitalization in an acute care hospital for two or more days within the prior 90 days
•Attendance at a hospital or hemodialysis clinic within the prior 30 days
Health care-associated PVE was observed in 37 percent of cases; of these, 70 percent were nosocomial and 30 percent were acquired in an outpatient context. Approximately 70 percent were diagnosed within the first year after valve implantation; the majority occurred within the first 60 days. S. aureus was the most common pathogen (identified in 34 percent of cases).
Nosocomial bacteremia is associated with significant risk for seeding prosthetic valves. In one study including 115 prosthetic valve recipients with nosocomial bacteremia judged not to be the sentinel event of endocarditis, PVE due to the bacteremic organism developed in 16 percent of cases between 7 and 170 days thereafter (median interval 28 days) [15]. In another study including 37 patients with prosthetic valves who developed postoperative candidemia (without evidence of endocarditis), candida PVE was diagnosed 26 to 690 days later in 11 percent of patients [16].
●Risk factors – Other risk factors for PVE include risk factors for NVE; these are discussed separately. (See "Native valve endocarditis: Epidemiology, risk factors, and microbiology".)
Microbiology
●Correlation with time since implantation – The microbiology of PVE involving surgically implanted valves is related in part to the time of onset after implantation (table 2) [1,4,17-25]:
•During the initial two months of implantation, the most frequently encountered pathogens are S. aureus and coagulase-negative staphylococci (CoNS); next in frequency are gram-negative bacilli, Candida species and diphtheroids (Corynebacterium). This spectrum of organisms reflects the typical nosocomial origin of these infections.
•Between 2 and 12 months after implantation, the most frequently encountered pathogens are CoNS, S. aureus, and streptococci, followed by enterococci. In general, cases occurring during this time interval are a blend of delayed-onset nosocomial and healthcare-associated or community-acquired infections (non-healthcare associated).
•Beyond 12 months after implantation, the most frequently encountered pathogens are streptococci and S. aureus, followed by CoNS and enterococci. Late PVE, like NVE, usually results from transient bacteremia occurring among ambulatory patients. (See "Native valve endocarditis: Epidemiology, risk factors, and microbiology", section on 'Microbiology'.)
•Culture-negative PVE occurs in all time intervals after surgery. (See 'Culture-negative endocarditis' below.)
•Sporadic cases of PVE due to a variety of other bacteria, fungi [26], Mycoplasma hominis [27], and nontuberculous mycobacteria (some related to intraoperative exposure to aerosolized organisms due to heater-cooler contamination by Mycobacterium chimaera; others related to contamination in the manufacture of bioprostheses) have been reported at various intervals after valve replacement. One case of infective endocarditis due to enterovirus has been described [28].
●CoNS antibiotic susceptibility – The species and antibiotic susceptibility of CoNS isolated from patients with PVE correlate with time of onset after valve implantation, which likely reflects the selective pressure of antibiotic exposure. CoNS causing PVE during the initial year after surgery are almost exclusively Staphylococcus epidermidis, most of which are methicillin resistant. In contrast, in the absence of recent antibiotic selective pressure, those CoNS causing PVE more than one year after valve surgery are commonly non-epidermidis species and most are methicillin susceptible [25,29]. (See "Infection due to coagulase-negative staphylococci: Epidemiology, microbiology, and pathogenesis" and "Infection due to coagulase-negative staphylococci: Treatment".)
Clinical manifestations — The clinical features of PVE are driven by the virulence of the infecting organism and whether any of the following are present: intracardiac complications, systemic emboli, and hematogenous infection involving extracardiac sites. Fever is the most common symptom; other symptoms include chills, anorexia and weight loss, malaise, headache, myalgias, arthralgias, night sweats, abdominal pain, dyspnea, cough, and pleuritic pain. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis", section on 'Symptoms and signs'.)
●Intracardiac complications – In patients with PVE, the frequency of valvular dysfunction, heart failure, perivalvular abscess, and new electrocardiographic conduction disturbances is higher than in patients with NVE [4,5,30]. Cardiac imaging may demonstrate evidence of abscess, fistulae, or aneurysm formation [31,32].
Perivalvular invasion, commonly with associated dehiscence of the prosthesis and perivalvular regurgitation, occurs in approximately 11 to 46 percent of cases, and frank extension into tissue causing myocardial abscess occurs in 15 to 29 percent of cases [4,5]. Aortic valve PVE may extend through the annulus to cause pericarditis or, more commonly, into the membranous portion of the interventricular septum where it disrupts the conduction system, resulting in heart block [5,33-35]. Large vegetations may prevent closure of the prosthesis, producing incompetence, or encroach upon the valve orifice leading to functional stenosis.
In PVE involving bioprosthetic valves, invasive disease is the predominant pathology during the initial year after valve surgery; thereafter, leaflet infection is predominant. In one study including 85 patients, annular and myocardial invasion was observed in 45 percent of cases and was noted more frequently among infections occurring in the first year after surgery than in cases presenting later (59 versus 25 percent) [30]. Similarly, in another series, invasive disease was more common in patients with early compared with later bioprosthetic PVE (79 versus 31 percent) [36].
●Systemic emboli – The incidence of clinically overt arterial emboli is 30 to 40 percent; central nervous system complications (primarily embolic infarcts or hemorrhages), occur in 11 to 40 percent of cases [4,5,37-39].
●Infection involving extracardiac sites – Clinical manifestations of complications, including those potentially arising through bacteremic seeding of extracardiac sites, warrant independent diagnostic evaluation, concurrent with evaluation for PVE. (See "Complications and outcome of infective endocarditis".)
In addition, patients with early PVE may present with postoperative manifestations that are more prominent than features of endocarditis.
TRANSCATHETER AORTIC VALVE IMPLANTATION (TAVI)
Timing — The timing and features of TAVI-PVE likely occur due to a pathogenesis similar to prosthetic valve endocarditis (PVE) after surgical valve implantation; earlier infections reflect nosocomial or health care-associated bacteremia in close temporal proximity to valve implantation, while later infections reflect to community occurrence of bacteremia. (See 'Timing and mechanisms' above.)
Epidemiology
●Incidence – The incidence of TAVI-PVE ranges from 0.3 to 1.2 per 100 patient years [40-44].
The incidence is highest during the initial year after valve placement and decreases over time (table 3) [42,43]. In an international registry including 579 patients with TAVI-PVE, the median time from implantation to onset of TAVI-PVE symptoms was 171 days (interquartile range 53 to 421 days) [31].
●Risk of TAVI-PVE versus SAVR-PVE – Data are conflicting regarding risk of PVE in the setting of TAVI compared with surgical aortic valve replacement (SAVR):
•Some data suggest the risk of PVE is comparable. In a meta-analysis of four randomized controlled trials comparing TAVI and SAVR, the incidence of PVE was similar at one year (representing early PVE) as well as at a mean follow-up of 2 and 3.4 years (representing late PVE) [45]. Among patients with an intermediate cardiac surgical risk, there was a trend toward increased risk of PVE among those who underwent TAVI (2.3 versus 1.2 percent; odds ratio 1.92, 95% CI 0.99-3.77). In addition, there was no difference in incidence of TAVI-PVE in recipients of a self-expanding valve (SEV) or a balloon expandable valve (BEV) compared to that in SAVR patients [45]. Other studies also have found rates of TAVI-PVE after implantation of a BEV or a SEV similar to rates of PVE after SAVR [46,47].
•Other data suggest lower risk of PVE associated with TAVI compared with SAVR. In a retrospective cohort study (without propensity score matching), the incidence of PVE was 4.81 (95% CI 4.61-5.03) per 1000 patient years (PY) among SAVR patients compared with 3.57 (95% CI 3.00-4.21) per 1000 PY among TAVI patients; in multivariable analysis, SAVR was an independent predictor of infective endocarditis [47]. Similarly, a review of data from three randomized trials reported a higher cumulative incidence of PVE at five years after SAVR versus TAVI (1.58 versus 1.01 percent) [48].
●Risk factors – The risk factors associated with TAVI-PVE have been evaluated using multivariable analysis [13,17,41,43,47,49,50]. Patient factors associated with TAVI-PVE include male sex, chronic renal disease (creatinine clearance <30 mL/min/1.75m2), pulmonary disease, cirrhosis, and endocarditis within the prior year. Procedure factors include postoperative aortic regurgitation (moderate to severe), need for cardiac implantable electronic device (CIED) placement, complications (such as cardiac arrest, major bleeding, and sepsis), low valve placement, and transapical access. In addition, concerns have been raised about non-standardized preprocedure antibiotic prophylaxis and variability in the sterility of the insertion setting (catheterization laboratory versus operating room).
Microbiology — In general, the microbiology of TAVI-PVE is similar to that of PVE after surgical valve replacement (see 'Microbiology' above); however, one important difference is the increased frequency of enterococci causing TAVI-PVE (table 2).
In one study including 91 patients with TAVI-PVE, S. aureus and enterococci accounted for the majority of TAVI-PVE cases within 30 days of implantation (35 and 34 percent, respectively) [51]. It is uncertain whether the higher frequency of enterococci causing TAVI-PVE reflects use of femoral vascular access, preoperative antibiotic prophylaxis that does not cover enterococci, the increased likelihood of a genitourinary portal of entry in older adults, or other factors.
Clinical manifestations — In general, the clinical manifestations of TAVI-PVE are similar to those of PVE after surgical valve replacement. (See 'Clinical manifestations' above.)
In one study including 578 patients with TAVI-PVE, fever was noted in 77 percent, CNS symptoms and other systemic emboli in 18 and 13 percent, respectively, new aortic regurgitation in 10 percent, and mitral valve regurgitation in 12 percent [52]. Rates of heart failure on admission range from 22 to 60 percent [17,18,52]. TAVI-PVE tends to occur in older patients; therefore, it may present with nonspecific symptoms (anorexia, weight loss, fatigue), and fever may be blunted.
Vegetations on the valve leaflets, the stent, or both are seen on echocardiography in 68 percent of cases [17,44]. Vegetations have been observed more commonly on stents of self-expanding valves (SEV) and on the leaflets of balloon expandable valves (BEV) [17].
In a study including 579 patients with TAVI-PVE, perivalvular extension of infection detected with various imaging techniques or at surgery was observed in 18 percent of cases [31]. Extension occurred with similar frequency among patients with SEV and BEV. Perivalvular abscesses were the most common lesion (observed in 87 patients), followed by pseudoaneurysms, fistulae, and a combination of lesions (7, 7, and 4 patients, respectively). In 65 patients, perivalvular extension was associated with infection of the TAVI itself, in 10 patients with mitral valve infection, in 25 patients with infection involving multiple sites, and in 5 patients with right-sided infection. TAVI dysfunction (primarily regurgitation) and large vegetations were more frequent in patients with perivalvular extension.
EVALUATION
When to suspect PVE — The diagnosis of prosthetic valve endocarditis (PVE) should be suspected in patients with history of valve replacement who present with signs and symptoms of endocarditis. Fever is the most common symptom; other symptoms include chills, anorexia and weight loss, malaise, headache, myalgias, arthralgias, night sweats, abdominal pain, dyspnea, cough, and pleuritic pain. (See 'Clinical manifestations' above.)
In the absence of these signs and symptoms, PVE should be suspected in the following circumstances:
●Bacteremia with an organism commonly causing PVE (see 'Microbiology' above)
●Persistent, unexplained bacteremia with an organism not commonly associated with PVE
●Persistent symptoms (such as fever, chills, anorexia, weight loss) in the absence of bacteremia
●New prosthetic valve dysfunction, particularly regurgitation, even in the absence of other signs of infection
●Systemic emboli
●Electrocardiogram (EKG) with new conduction abnormality (first degree or complete atrioventricular block)
Blood cultures
Collection and interpretation
●Collection – At least three sets of blood cultures should be obtained, ideally from separate venipuncture sites, prior to initiation of antibiotic therapy [53,54].
The 2023 Duke-ISCVID criteria schema no longer require that blood cultures be drawn from unique venipunctures (although this is highly desirable), nor obtained over a specified period of time [55].
In patients with untreated endocarditis, bacteremia is continuous, thus blood cultures will be positive irrespective of their temporal relationship to fever.
●Yield and interpretation
•In the absence of antecedent antibiotic therapy, blood cultures are positive in ≥90 percent patients with PVE. If all or most blood cultures drawn over a period of hours to days in a patient with a prosthetic valve are positive, PVE is highly probable.
•For blood cultures with isolates commonly considered contaminants (such as coagulase-negative staphylococci, Corynebacterium, or Cutibacterium, a high rate of blood culture positivity helps to distinguish infecting pathogens from contaminants. In addition, molecular evidence that a sporadically isolated organism represents a single clone may help to distinguish infecting pathogens from contaminants [56].
Issues related to blood cultures for diagnosis of endocarditis are discussed further separately. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis", section on 'Blood cultures'.)
Culture-negative endocarditis — Culture-negative endocarditis refers to endocarditis with no definitive microbiologic etiology following inoculation of at least three independently obtained blood samples into a standard blood-culture system with five days of incubation. (See "Blood culture-negative endocarditis: Epidemiology, microbiology, and diagnosis".)
●Microbiology – In the absence of antibiotic administration prior to blood culture collection, it is unusual for patients with PVE due to common bacterial causes of endocarditis to have persistently negative blood cultures.
Culture-negative PVE can occur when infection is caused by fastidious organisms such as Legionella species, Bartonella species, Coxiella burnetii, Mycoplasma hominis, mycobacteria, and fungi.
Rare cases of blood culture-negative PVE (and other cardiac surgery related focal infections) have been caused by M. chimaera, a non-tuberculosis mycobacterium. These infections have been linked to the intraoperative exposure of surgical wounds to aerosolized organisms from contaminated heater-cooler units used during cardiac surgery [57-59]. Onset of these infections after cardiac surgery has often been very delayed and indolent [59]. (See "Overview of nontuberculous mycobacterial infections", section on 'M. chimaera associated with cardiac surgery'.)
●Diagnosis – Tools for diagnosis of culture-negative endocarditis are discussed separately. (See "Blood culture-negative endocarditis: Epidemiology, microbiology, and diagnosis".)
In addition, fluorescence in situ hybridization (FISH) combined with 16S rRNA-gene polymerase chain reaction (PCR) and sequencing (FISHseq) is an emerging technique that can be applied to a surgically excised prosthetic valve. The technique may enable organism identification at the species level, as well as provide histopathologic information regarding organism location [60]. This technique also underscores the diagnostic utility of combining molecular detection of specific organisms in tissue specimens with the identification of consistent organisms on histopathology. While not widely available, this is an innovative approach for challenging diagnoses.
Cardiac imaging — Our suggested approach to cardiac imaging with suspected PVE is outlined in the algorithm (algorithm 1). Supporting evidence for this approach is described below.
Echocardiography as primary modality — Transthoracic echocardiography (TTE) and transesophageal echocardiography (TEE) are the primary tools for diagnosis of PVE; both are recommended for assessment of patients with suspected PVE [1,2,61].
TTE (often the initial study) provides information regarding cardiac function and in some patients, anatomic evidence of endocarditis. TEE is the echocardiographic approach of choice [1,61]. Among patients with PVE, TEE is significantly more sensitive for the detection of anatomic evidence of infection than TTE (86 to 92 percent versus 17 to 52 percent, respectively), particularly for assessing a mitral valve prosthesis or perivalvular complications [62-70]. Thus, in the absence of contraindications, TEE should be performed for patients with suspected PVE [1].
If PVE is suspected in spite of negative echocardiography, repeat echocardiography (5 to 7 days later) may establish a diagnosis of PVE in some cases [1,64,71]. In one study among patients with suspected PVE, repeat TEE resulted in a shift from possible to definite endocarditis in 42 percent of cases; there was no diagnostic benefit beyond a third TEE [72].
For diagnosis of transcatheter implanted aortic valves (TAVI)-PVE, the sensitivity of echocardiography (including TEE) is limited due to artifacts and shadowing caused by the large metal stents anchoring the valve. In one study including 578 patients with TAVI-IE, echocardiography failed to detect evidence of endocarditis in 15 percent of cases [52]. Accordingly, alternative imaging modalities may be needed for the assessment of suspected TAVI-PVE. (See 'Emerging imaging modalities' below.)
Intracardiac echocardiography (ICE) may provide diagnostic information in selected patients with suspected PVE involving surgically implanted valves or TAVI. For patients with possible endocarditis, ICE may allow reclassification to definite endocarditis [73].
Issues related to echocardiography for diagnosis of endocarditis are discussed further separately. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis".)
If the diagnosis of PVE remains uncertain after TEE, other imaging technologies should be considered (if available), as discussed below.
Emerging imaging modalities — The most useful additional cardiac imaging tools for diagnosis of PVE include electrocardiogram-gated cardiac computed tomography angiogram (CTA) and fluorine-18 fluorodeoxyglucose positron emission tomography computed tomography (18F-FDG PET/CT) [74]. CTA or 18F-FDG PET/CT may be useful in patients for whom the diagnosis of PVE remains uncertain following TEE, in circumstances where TEE is not feasible, and when additional information regarding perivalvular extension of infection is sought. Findings based on these tools are now included as major diagnostic criteria in the 2023 Duke-ISCVID and 2023 ESC diagnostic criteria [1,55].
Decisions regarding use of these tools depend on local availability and expertise; they are often guided by the diagnostic or management questions that persist after traditional echocardiographic imaging has been performed. There are multiple subtle aspects of modality selection, imaging technique, and image interpretation; optimal use of these tools requires careful consultation between managing physicians and imagers [32].
●Cardiac CTA – Cardiac CTA may be helpful for cases in which definitive evidence of PVE and its complications cannot be fully assessed with TEE, or in planning a surgical strategy for patients with extra-valvular complications. In general, TEE is superior to CTA for detection of vegetations (especially small ones) or valve perforations, whereas CTA is comparable or superior to TEE for detection of perivalvular infection, abscess, or pseudoaneurysm [75-78]. CTA may also provide satisfactory coronary imaging for patients at intermediate risk of coronary artery disease who require surgical intervention [77,78].
●18F-FDG PET/CT – 18F-FDG PET/CT is useful for diagnosis of selected cases of PVE in which echocardiography is not diagnostic, or when clarification of perivalvular extension or possible extension of infection to an ascending aorta graft is needed.
With this modality, uptake of positron-labeled glucose by inflammatory leukocytes allows anatomic localization of infection. The sensitivity likely diminishes as inflammation resolves with administration of antimicrobial therapy; in addition, images must be reviewed carefully to distinguish between PVE and non-infectious post-surgical inflammation, particularly in patients who are within two to three months after valve implantation or when BioGlue has been used in the prior surgery [79].
18F-FDG PET/CT includes whole body imaging; therefore, extracardiac sites of infection also may be detected.
Studies comparing echocardiography, 18F-FDG PET/CT, and CTA include:
●In a 2015 study, more than 90 patients with suspected endocarditis involving prosthetic valve(s) and/or other intracardiac devices were evaluated with TEE, 18F-FDG PET/CT, and CTA [69]. Patients were initially classified according to the 2000 modified Duke criteria using TEE. Addition of 18F-FDG PET/CT to the modified Duke criteria as an additional major criterion was associated with increased diagnostic sensitivity for PVE (from 52 to 91 percent) with little loss in specificity (from 95 to 90 percent), and with an increase in the negative predictive value (from 60 to 88 percent). In a subgroup of patients who underwent both 18F-FDG PET/CT and CTA, the classification was improved further by CTA detection of abscesses, pseudoaneurysms, and fistulas.
●In a 2020 study, more than 180 patients with suspected PVE were classified diagnostically according to the 2000 modified Duke criteria and then were reclassified with 18F-FDG PET/CT findings included as an additional major criterion [70]. Inclusion of the 18F-FDG PET/CT results improved the sensitivity (42 to 91 percent), positive predictive value (74 to 86 percent), and negative predictive value (65 to 93 percent) over the modified Duke criteria. Of the 62 episodes initially classified as possible PVE, 18F-FDG PET/CT findings allowed reclassification to definite PVE in 76 percent of cases and increased conclusive diagnoses (definite or rejected) from 67 to 92 percent.
●A prospective study examined the impact of FDG PET/CT on the diagnosis and management of patients with suspected IE, including 70 patients with suspected PVE [80]. FDG PET-CT led to modification of the classification by Duke criteria in 24 percent of patients with PVE, mostly due to perivalvular uptake. Patient management was modified in 21 percent of cases, resulting in a change in antibiotic therapy in 15 of 70 patients and change in cardiac surgery management in 4 of 70 patients. Those who benefitted the most from FDG PET-CT were those with noncontributory baseline echocardiography or initial classification of possible IE.
●For patients with suspected TAVI-PVE, use of 18F-FDG PET/CT and/or CTA may be especially beneficial given diminished sensitivity of echocardiography in this context. In one study including 22 patients with possible TAVI-PVE (based on echocardiography), additional imaging with 18F-FDG PET/CT and CTA confirmed or excluded TAVI-PVE in 10 cases but 12 cases remained classified as possible PVE, requiring a clinical decision regarding treatment [81]. In another report including 16 patients with suspected TAVI-PVE, multimodality imaging increased the sensitivity of the modified Duke criteria from 50 to 100 percent [82].
Thus far, other imaging modalities (including intracardiac echocardiography, cardiac MRI, and SPECT/CT) are less well studied.
Additional evaluation for PVE source or complications — Additional diagnostic evaluation for patients with suspected or known PVE includes electrocardiography (ECG), chest radiography, and potentially further radiographic imaging tailored to clinical manifestations.
●ECG – Baseline ECG should be performed as part of the initial evaluation for all patients with suspected infective endocarditis, with subsequent telemetry monitoring or serial electrocardiograms. The presence of heart block or conduction delay (which may manifest initially as a prolonged PR interval) may provide an important clue to extension of infection into the valve annulus and adjacent septum that should prompt further evaluation. In addition, the presence of findings consistent with ischemia or infarction may suggest the presence of emboli to the coronary circulation. (See "ECG tutorial: Basic principles of ECG analysis".)
●Chest radiography – Chest radiography is warranted to evaluate for an infiltrate, congestive heart failure, and potential alternative causes of fever and systemic symptoms. Findings suggestive of septic pulmonary emboli may be seen in patients with a fistula from left sided endocarditis into the right side of the heart, patients with tricuspid valve PVE, patients with concurrent infection of a cardiac implantable electronic device (CIED), or infection of a valved pulmonary artery conduit in patients with corrected congenital heart disease.
●Computed tomography (CT) – CT of the torso (chest, abdomen, and pelvis) is not routinely done but may be used to evaluate for systemic emboli with splenic or renal infarcts) or sites of metastatic infection (vertebral osteomyelitis/discitis, psoas abscess, or other sites of infection) that may warrant localized drainage [2,83]. The decision to pursue this imaging should be guided by a careful history and clinical assessment.
●Additional radiographic imaging – Additional radiographic imaging to evaluate for complications of PVE should be tailored to findings on history and physical examination [2]. As an example, patients with back pain should be evaluated for vertebral osteomyelitis with magnetic resonance imaging (MRI).
•Role of brain MRI – Patients with headache, neurologic deficits, or meningeal signs should be evaluated with brain MRI/magnetic resonance angiogram for neurologic complications (including mycotic aneurysms or central nervous system bleeding). (See "Overview of infected (mycotic) arterial aneurysm".)
Routine brain imaging with CT or MRI is not standard of care in the absence of focal neurologic signs or symptoms. However, in patients with endocarditis, brain MRI commonly detects lesions considered related to endocarditis in patients with and without clinical neurologic findings [2,84-87]. The presence of such lesions can be diagnostically important. For example, in one prospective study including 130 patients with suspected endocarditis, of whom only 16 had acute neurologic symptoms, brain MRI (with angiography) detected findings related to endocarditis in all patients with neurologic symptoms and 79 percent of patients without clinical neurologic abnormalities [84]. Among those with no neurologic symptoms, small ischemic lesions and microhemorrhages (in 46 and 67 patients, respectively) were the most common abnormalities. The MRI findings led to an upgraded diagnostic classification to definite or possible endocarditis in 32 percent of patients previously classified as non-definite endocarditis, and to a modification of planned therapy in 18 percent of cases. The impact of the MRI findings on clinical outcome was not assessed in this study. Larger studies are needed to clarify the role of routine brain MRI in clinical practice.
Cerebral microhemorrhages are seen commonly on brain MRI in settings other than endocarditis. They do not satisfy a minor diagnostic criteria as emboli, nor are they associated with intracerebral hemorrhage [1,84].
ESTABLISHING A DIAGNOSIS OF PVE —
The diagnosis of prosthetic valve endocarditis (PVE) is established based on clinical manifestations, blood cultures (or other microbiologic data), and cardiac imaging. The diagnosis of PVE requires (1) identification of the infecting pathogen by culture, serologic testing, or molecular testing and (2) cardiac imaging to identify a valvular vegetation, paravalvular abscess, or other structural complication of infection.
The accepted criteria for diagnosis of IE are the 2023 Duke-International Society for Cardiovascular Infectious Disease (ISCVID) criteria [55]. These criteria are summarized in the tables (table 4 and table 5).
The 2023 European Society for Cardiology (ESC) Guidelines has a schema for the diagnosis of IE [1], which is similar to that of the 2023 Duke-ISCVID criteria.
2023 Duke-ISCVID criteria — The 2023 Duke-International Society for Cardiovascular Infectious Disease (ISCVID) criteria (table 4 and table 5) [55] were updated from the 2000 modified Duke criteria.
These criteria use combination of major and minor criteria to yield clinical diagnoses of definite, possible, or rejected endocarditis:
●Definite endocarditis – Pathologic diagnosis or clinical diagnosis with one of the following: 2 major criteria, 1 major plus 3 minor criteria, or 5 minor criteria.
●Possible endocarditis – One of the following: 1 major plus 1 minor criteria, or 3 minor criteria.
●Reject endocarditis – Any of the following: firm alternate diagnosis, lack of recurrence despite antibiotic therapy for less than four days, or no pathologic evidence of endocarditis with antibiotic therapy for less than four days.
The major changes in the 2023 Duke criteria (from the 2000 Duke criteria) include:
●The pathological criteria for definite endocarditis has been expanded (beyond molecular or culture evidence of an organism typically causing endocarditis from intracardiac tissue or a systemic embolus) to include evidence of such an organism from an extracted cardiac implantable electronic device (CIED).
●The major microbiologic criterion for positive blood cultures yielding organisms that typically cause endocarditis has been simplified to at least two positive blood cultures; it no longer requires that these be drawn over a specified time interval from unique venipunctures. In addition, the list of typical organisms now includes Staphylococcus lugdunensis and Enterococcus faecalis. For organisms that rarely or occasionally cause endocarditis, at least three unique positive blood cultures are required.
●The major microbiological criteria now list organisms considered typical causes of endocarditis involving intracardiac prosthetic material: coagulase negative staphylococci, Corynebacterium striatum, Corynebacterium jeikeium, Serratia marcescens, Pseudomonas aeruginosa, Cutibacterium acnes, nontuberculous mycobacteria, and Candida spp.
●New major microbiological criteria have been added:
•Positive polymerase chain reaction (PCR) or (other nucleic acid-based technique) for Coxiella burnetii, Bartonella spp, or Tropheryma whipplei from blood
•IFA for detection of IgG antibodies to B. henselae or B. quintana with titer ≥1:800
●The major imaging criteria have been expanded to include new technologies: cardiac computed tomography (CT) and fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) with abnormal activity involving with a prosthetic valve or cardiac device, more than three months after cardiac surgery. These are equivalent to diagnostic echocardiographic findings.
●New minor clinical criteria have been added:
•Predispositions: Prior history of endocarditis, transcatheter prosthetic valve placement or valve repair, or endovascular CIED.
•Vascular phenomenon: Splenic or cerebral abscess.
•Microbiological: Culture or PCR (or other nucleic acid-based test) for an organism consistent with IE from valve tissue, systemic embolus, or a sterile non-cardiac body site.
•Imaging: Abnormal activity on 18F-FDG PET/CT (within three months of valve implantation) involving a valve, ascending aortic graft (with concomitant evidence of valve involvement), intracardiac device leads, or other prosthetic material.
Validation of Duke criteria — Validation studies of the Duke criteria and ESC criteria are discussed separately. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis", section on 'Validation of Duke criteria'.)
Clinical judgement is paramount in application of diagnostic criteria in clinical practice. The 2023 Duke-ISCVID and 2023 ESC diagnostic criteria provide improved sensitivity for the diagnosis of PVE over previous criteria; however, each loses specificity (incorrectly diagnosing definite endocarditis, ie, false positive cases) in certain situations. Furthermore, these criteria are not 100 percent sensitive for diagnosis of endocarditis.
DIFFERENTIAL DIAGNOSIS —
Noninfective, calcific, vegetative-like lesions with inflammatory infiltrates may develop on the leaflets of bioprosthetic valves with age; these can be misdiagnosed as infection. In one study including 88 patients who underwent removal of bioprosthetic valve tissue (21 for suspected endocarditis and 67 for noninfective dysfunction), PVE was characterized histologically by surface vegetations with microorganisms and neutrophil-rich inflammatory infiltrates; in the non-infected group, valve pathology was characterized by calcific inflammatory infiltrates consisting mainly of macrophages and lymphocytes [88]. This study emphasizes the benefit of combining molecular diagnostics with histopathology in assessing valve tissue in the diagnosis of culture-negative endocarditis, especially PVE. (See 'Culture-negative endocarditis' 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".)
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 5th to 6th 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 10th to 12th 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 email 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 topic (see "Patient education: Endocarditis (The Basics)")
SUMMARY AND RECOMMENDATIONS
●Forms of PVE – Prosthetic valve endocarditis (PVE) refers to infection of infection of surgically implanted cardiac valves or transcatheter implanted aortic valves (TAVI) or other non-CIED intracardiac foreign bodies. (See 'Surgical valve implantation' above and 'Transcatheter aortic valve implantation (TAVI)' above.)
●Surgical valve implantation
•Timing and microbiology – The microbiology of PVE typically reflects the time since implantation (table 2) (see 'Timing and mechanisms' above and 'Microbiology' above):
-In early PVE (<2 months after surgery), microorganisms reach the prosthetic valve via direct intraoperative contamination or via hematogenous spread during the initial days and weeks after surgery; perivalvular abscess is common.
-Cases occurring 2 to 12 months after surgery are a blend of delayed-onset nosocomial infections (reflecting infection at the time of surgery or during the surgical admission) and community-acquired infections.
-In late PVE (>12 months after surgery), the pathogens tend to be similar to those that cause native valve endocarditis (NVE).
•Epidemiology – PVE represents 20 percent of all cases of endocarditis; it occurs in 1 to 6 percent of patients with valve prostheses, with an incidence of 0.3 to 1.2 percent per patient-year; the risk is greatest during the initial year after implantation. PVE occurs with slightly greater frequency on bioprosthetic than mechanical valves (table 3). (See 'Epidemiology' above.)
•Clinical manifestations – Patients with PVE present with symptoms and signs similar to those encountered in NVE; however, many patients with PVE present with nonspecific symptoms. Clinical manifestations of PVE include fever, chills, anorexia, and weight loss. The frequency of new or changing murmurs, heart failure, and new electrocardiographic conduction disturbances in patients with PVE is higher than in patients with NVE. (See 'Clinical manifestations' above.)
●Transcatheter aortic valve implantation – TAVI-PVE is an increasingly frequent form of PVE. Cases primarily present in the initial year after valve replacement. The microbiology of TAVI-PVE is similar to that seen with PVE involving surgically implanted valves, with the exception of an increased frequency of enterococcal infections. The incidence of and clinical manifestations of TAVI-PVE are similar to that of PVE involving surgically implanted aortic bioprosthetic valves. (See 'Transcatheter aortic valve implantation (TAVI)' above.)
●Evaluation
•Clinical suspicion – The diagnosis of PVE should be suspected in patients with history of valve replacement who present with any of the following (see 'When to suspect PVE' above):
-Bacteremia with an organism commonly causing PVE
-Persistent, unexplained bacteremia with an organism uncommonly associated with PVE
-Persistent nonspecific symptoms (fever, chills, anorexia, weight loss) in the absence of bacteremia
-New prosthetic valve dysfunction, particularly regurgitation, even in the absence of other signs of infection
-Systemic emboli
-EKG with new conduction abnormality (first degree or complete atrioventricular block)
•Blood cultures – At least three sets of blood cultures should be obtained, ideally from separate venipuncture sites, prior to initiation of antibiotic therapy. For patients with signs of clinical instability, initiation of empiric antimicrobial therapy (after blood cultures have been obtained) is indicated. (See 'Blood cultures' above.)
•Cardiac imaging
-Echocardiography as primary modality – Echocardiography should be performed in all patients with suspected PVE (algorithm 1). In general, transthoracic echocardiography (TTE) is the first imaging test for patients with suspected PVE; however, transesophageal echocardiography (TEE) has higher sensitivity than TTE and should be pursued in all patients with suspected PVE in the absence of contraindications. In patients with TAVI-PVE, the sensitivity of echocardiography is limited due to artifacts and shadowing caused by the metal stents anchoring the valve. (See 'Echocardiography as primary modality' above.)
-Emerging imaging modalities – Additional cardiac imaging tools, including 18F-fluorodeoxyglucose positron emission computed tomography (18F-FDG PET/CT) and electrocardiogram-gated cardiac computed angiography (CTA), are useful in patients for whom the diagnosis of PVE (particularly TAVI-PVE) remains uncertain following echocardiography. With 18F-FDG PET/CT, uptake of positron-labeled glucose by inflammatory leukocytes allows anatomic localization of infection. CTA is useful when TEE is contraindicated; compared with TEE, CTA is less sensitive for detecting vegetations, but more sensitive for detecting perivalvular abnormalities. Findings from these imaging techniques satisfy a major diagnostic criteria in the 2023 Duke diagnostic schema. (See 'Emerging imaging modalities' above.)
•Additional evaluation – Additional evaluation for patients with suspected PVE includes electrocardiography, chest radiography, and other radiographic imaging tailored to clinical manifestations. (See 'Additional evaluation for PVE source or complications' above.)
●Establishing a diagnosis – The diagnosis of PVE is established based on clinical manifestations, blood cultures (or other microbiologic data), and cardiac imaging (usually echocardiography). The accepted criteria for diagnosis of infective endocarditis are the 2023 Duke-International Society for Cardiovascular Infectious Disease (ISCVID) criteria. (See 'Establishing a diagnosis of PVE' above.)
