Version May 2024
INTRODUCTION — Prosthetic valve endocarditis (PVE) refers to infection of one or more prosthetic heart valves [1-4]. The timing of the infection after surgical valve replacement reflects different pathogenic mechanisms that, in turn, influence the clinical presentation.
The pathogenesis, epidemiology, microbiology, pathology, clinical manifestations, and diagnosis of PVE (both surgical aortic valve replacement [SAVR] and transcatheter aortic valve implantation [TAVI]) 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 "Surgery for prosthetic valve endocarditis" and "Prevention of endocarditis: Antibiotic prophylaxis and other measures".)
General issues related to TAVI are discussed separately. (See "Indications for valve replacement for high gradient aortic stenosis in adults", section on 'Choice of surgical or transcatheter intervention' and "Transcatheter aortic valve implantation: Periprocedural and postprocedural management" and "Transcatheter aortic valve implantation: Complications".)
DEFINITIONS AND PATHOGENESIS — The timing of PVE reflects different pathogenic mechanisms that, in turn, influence the epidemiology, microbiology, pathology, and clinical manifestations of the infection [5-9]. (See 'Microbiology' below.)
Infection engrafted on a transcatheter aortic valve replacement occurs with greatest frequency during the initial year after placement, but studies have not correlated time of onset with pathology or microbiology.
Early infection — In early PVE (during the initial two months after surgery), microorganisms reach the prosthetic valve via direct intraoperative contamination or via hematogenous spread during the initial days and weeks after surgery. In general, cases occurring 2 to 12 months after surgery are a blend of delayed-onset nosocomial (infection reflecting infection at the time of surgery or during the surgical admission) and community-acquired infections. (See 'Microbiology' below and 'Health care-associated infection' below.)
Early after valve implantation, the valve sewing ring, cardiac annulus, and anchoring sutures have not yet become covered with endothelium; therefore, organisms have direct access to the prosthesis-annulus interface and to perivalvular tissue along suture pathways. These structures are coated with host proteins, such as fibronectin and fibrinogen, to which organisms can adhere. Perivalvular abscesses are particularly common with prosthetic valves because the annulus is commonly the primary site of infection involving both mechanical and bioprosthetic valves, especially in early PVE [10].
Late infection — The pathogenesis of late PVE (>12 months postoperatively) has been postulated to resemble native valve endocarditis (NVE). (See "Pathogenesis of vegetation formation in infective endocarditis".)
As the sewing ring, sutures, and adjacent tissues become endothelialized following valve replacement, alterations in the surface and flow characteristics of valve leaflets may facilitate deposition of microthrombi (comprised of platelets and fibrin) on the bioprosthetic leaflets and the anchoring stent of bioprosthetic and mechanical valves. These microthrombi serve as hospitable surfaces for organisms to adhere. The pathogens associated with late PVE tend to be bacteremic isolates able to survive serum bactericidal activity and adhere to these microthrombi, and are similar to those inducing 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, the perivalvular tissues are less likely to be affected in late PVE. 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
General principles — PVE represents 20 percent of all cases of endocarditis; it occurs in 1 to 6 percent of patients with valve prostheses [11], with an incidence of 0.3 to 1.2 percent per patient-year [4,11,12]. In general, PVE in patients who have undergone surgical valve replacement occurs with equal frequency at aortic and mitral sites [6,7,13-16].
The epidemiology of aortic valve PVE depends on whether the valve replacement was a surgical aortic valve replacement (SAVR) or a transcatheter aortic valve implantation (TAVI). (See "Indications for valve replacement for high gradient aortic stenosis in adults", section on 'Choice of surgical or transcatheter intervention'.)
Risk factors for endocarditis are discussed further separately. (See "Native valve endocarditis: Epidemiology, risk factors, and microbiology".)
Surgical replacement — Patients who undergo SAVR are at risk of developing PVE. In data from Danish national registries between 1996 and 2015, the cumulative incidence of endocarditis among patients following SAVR was 6.0 per 1000-patient years (PY). This is less than the incidence of infective endocarditis in individuals with a prior history of endocarditis (16.1 out of 1000 PY), but greater than the incidence of native valve endocarditis in non-valve replacement matched controls (hazard ratio [HR] 19.1, 95% CI 15.0-24.4) [9].
In early studies, the frequency of PVE during the initial postoperative year among patients with a mechanical valve is similar to or greater than the frequency of PVE among patients with a bioprosthetic valve; with increasing time after valve implantation, PVE may be slightly more common among patients with a bioprosthetic valve [5,6,13]. In several early studies, the cumulative percentage of patients with a bioprosthetic valve who developed PVE ranged from 1 to 3 percent during the initial postoperative year and 3 to 6 percent by five years after valve replacement [5,7,8,13,14,17].
Subsequent studies have demonstrated that the risk of infection is higher for bioprosthetic valves than mechanical valves. In three randomized trials including more than 1400 patients with 8 to 20 years of follow-up, rates of PVE were nominally (but not significantly) higher in patients receiving bioprosthetic versus mechanical valves [18-20].
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) [21]. In addition, the Swedish data are notable for a greater incidence of PVE among recipients of bioprosthetic than mechanical valves at multiple time points (1 year, 2 to 5 years, and 6 to 10 years) (table 1) [22,23].
Similarly, in an American observational study including more than 38,000 patients ≥65 years of age with prosthetic valves implanted from 1991 to 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) at 12-year follow-up (adjusted HR 1.60, 95% CI 1.31-1.95) [24].
Transcatheter aortic valve implantation — Factors that may confer risk for TAVI-PVE include extensive implanted foreign material, residual perivalvular regurgitation following implantation, and mucosal injury associated with device placement.
The median time from implantation to onset of PVE symptoms in an international registry including 250 TAVI-PVE episodes was 5.3 months (interquartile range 1.5 to 13.4 months) [25]. In an expansion of that registry to include 579 patients, the median time to onset was 171 days (interquartile range 53 to 421 days) [26].
In data from prospective clinical trials and national registries (in which propensity scoring or multivariable analysis was used to overcome differences in baseline patient characteristics), the incidence rates of TAVI-PVE are similar to the incidence rates of SAVR-PVE with a bioprosthetic valve [27-30]. Rates of PVE for TAVI and SAVR are highest during the initial year after placement and decrease over time (table 2) [29,30].
In a meta-analysis including 10 randomized trials comparing TAVI and SAVR, the overall incidence of PVE was similar in patients who underwent TAVI or SAVR at one year (representing early PVE) as well as at a mean follow-up of 2 and 3.4 years (representing late PVE) [31]; among patients with intermediate 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). There was no difference in incidence of PVE among patients who underwent TAVI with a self-expanding valve (SEV), TAVI with a balloon expendable valve (BEV), or SAVR. Similarly, in a registry including more than 6200 patients who underwent TAVI, the cumulative incidence of PVE at one year after implantation for SEV and BEV was 0.95 and 1.25 percent, respectively [32].
The incidence of TAVI-PVE and SAVR-PVE is similar among older patients with severe aortic stenosis and intermediate operative risk. In a propensity-matched comparison of more than 1000 such patients who underwent TAVI with a BEV or SAVR, the rates of PVE at one year were 0.8 and 0.7 percent, respectively [33]. Similarly, in a trial including more than 280 patients at lower surgical risk managed with a self-expanding bioprosthesis by TAVI or a bioprosthesis by SAVR, the cumulative five-year incidences of PVE was 6.2 and 4.4 percent, respectively [34].
In contrast to the aforementioned findings, a large retrospective cohort study using linked databases (without propensity score matching) found an overall infective endocarditis incidence of 4.81 (95%CI 4.61-5.03) per 1000 PY among patients undergoing SAVR compared with 3.57 (95% CI 3.00-4.21) per 1000 PY among patients undergoing TAVI. In a multivariable analysis, SAVR was an independent predictor of infective endocarditis. The patients who underwent SAVR had a longer follow-up period (53.9 versus 24.5 months). More than one-third of patients who underwent SAVR underwent concomitant coronary artery bypass grafting and 7.7 percent had more than one valve intervention [35]. In addition, in a review of data from three randomized trials, a higher cumulative incidence of PVE at five years was observed among patients who underwent SAVR versus TAVI [36]. These studies illustrate the challenges of comparing outcomes among patient groups, including the effect of differences in baseline characteristics, surgical procedure, and duration of follow-up.
A number of studies have evaluated risk factors for TAVI-PVE using multivariable analysis [23,25,28,30,35,37,38]. 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 postprocedure aortic regurgitation (moderate to severe), need for cardiac electrical device placement, hospital complications (such as cardiac arrest, major bleeding, and sepsis), low valve placement, and transapical access. In addition, concerns have been raised about nonstandardized preprocedure antibiotic prophylaxis and variable sterility in implant facilities.
Health care-associated infection — In addition to the typical nosocomial nature of early onset PVE, some cases of PVE (including TAVI-PVE) are health care-associated (in that they result from infection acquired in outpatient health care settings or in the context of ongoing invasive care).
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 therapy, wound care, or specialized nursing care at home or intravenous 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 infection 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.
In patients with prosthetic valves, nosocomial bacteremia is associated with significant risk for seeding the prosthesis. 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) [39]. Similarly, in another study including 37 patients with prosthetic valves and postoperative candidemia without evidence of endocarditis, fungal endocarditis developed in 11 percent of cases, between 26 and 690 days later [40]. The patients who developed candida PVE had persistent fungemia (mean 8.1 days) without evidence of endocarditis during the month after cardiac surgery.
MICROBIOLOGY — The microbiology of PVE involving surgically implanted valves is relatively predictable, depending on the time since implantation (table 3) [4,25,41-46]:
●During the initial two months of implantation, the most frequently encountered pathogens were S. aureus and coagulase-negative staphylococci (CoNS); next in frequency were gram-negative bacilli and Candida species. This spectrum of organisms reflects the typical nosocomial origin of these infections.
●Between 2 and 12 months after implantation, the most frequently encountered pathogens were coagulase-negative staphylococci, S. aureus, and streptococci, followed by enterococci. In general, cases occurring 2 to 12 months after surgery are a blend of delayed-onset nosocomial and community-acquired infections.
●Beyond 12 months after implantation, the most frequently encountered pathogens were streptococci and S. aureus, followed by coagulase-negative staphylococci and enterococci. In general, the range of pathogens is similar to that of native valve endocarditis (NVE) in patients who are not injection drug users. This is because 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 [47], Mycoplasma hominis [48], 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 [49].
The above time-related descriptions of PVE microbiology are illustrated by observations of PVE due to CoNS. Among patients with PVE due to CoNS during the initial year after surgery, the cause is almost exclusively Staphylococcus epidermidis, of which most are methicillin resistant (reflecting the likely nosocomial origin); in contrast, among patients with PVE due to CoNS more than one year after surgery, almost half are non-epidermidis species, and most are methicillin susceptible (reflecting the likely community origin) [13,50]. (See "Infection due to coagulase-negative staphylococci: Epidemiology, microbiology, and pathogenesis" and "Infection due to coagulase-negative staphylococci: Treatment".)
Among 780 episodes of PVE (TAVI-PVE cases are included as bioprosthetic valve infections) in a Swedish registry between 2008 and 2020, the microbiology was similar to that noted above and in the table (table 3). Coagulase-negative staphylococci, Corynebacterium, and Candida were more common in year 1 (compared with later-onset cases) and alpha-hemolytic streptococci and S. bovis were more frequent after year 1. S. aureus more commonly infected mechanical valves; streptococci, enterococci, and coagulase-negative staphylococci more commonly infected bioprosthetic valves [51].
In reports of very early onset of TAVI-PVE (onset within 30 days), S. aureus and enterococci accounted for the majority of cases (35 and 34 percent, respectively) [52]. Generally, the microbiology of TAVI-PVE is similar to that of surgical PVE; however, one difference is an increased frequency of enterococci causing TAVI-PVE (table 3). It is uncertain whether this may reflect use of femoral vascular access, preoperative antibiotic prophylaxis that does not cover enterococci, increased likelihood of a genitourinary portal of entry in older adults, or other factors.
PATHOLOGY — The intracardiac pathology of infection involving surgically placed valves, particularly when PVE presents during the early months after surgery or when it is caused by invasive organisms, shapes the requirement for therapy. Perivalvular invasion, commonly with associated dehiscence of the prosthesis and perivalvular regurgitant flow, occurs in approximately 40 percent of patients and frank extension into tissue causing myocardial abscess is seen in 15 percent [53,54]. In one series, invasion of perivalvular tissue was noted in more than 80 percent of cases [55].
In some patients, infection of an aortic valve prosthesis extends 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 various degrees of heart block [56-58]. Large vegetations may prevent closure of the prosthesis producing incompetence or encroach upon the valve orifice causing functional stenosis.
Bioprosthetic valve endocarditis also is associated with invasive infection. Annular and myocardial invasion was noted in 38 of 85 patients (45 percent) in one study and was more frequent among bioprosthetic PVE occurring in the first year after valve replacement than in cases presenting later (59 versus 25 percent) [59]. Similarly, in another series, invasive disease was more common in patients with early compared with later bioprosthetic PVE (79 versus 31 percent) [60].
The histologic features that characterize PVE in bioprosthetic valves are not well defined. As bioprosthetic valves degenerate, they may form noninfective, calcific, vegetative-like lesions with inflammatory infiltrates, thus potentially causing a noninfectious process that can be misdiagnosed as PVE. In one study of the histopathology of bioprosthetic valve tissue removed at surgery from 88 patients (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 [61].
Given relatively limited autopsy and surgical reports, the pathology of TAVI-PVE has been largely inferred from transesophageal echocardiography and other cardiac imaging. In a registry including more than 240 patients with TAVI-PVE, echocardiography demonstrated vegetations and perivalvular complications in 68 and 18 percent of cases, respectively [25]. Invasive complications appeared similar with infection of self-expanding valves (SEV) and balloon expandable valves (BEV) [32]; vegetations were attached to the stent in 18 percent and/or on leaflets in 48 percent of cases, respectively. Vegetations were more common on stents of SEV compared with BEV (26 and 11 percent, respectively), and vegetations were found more commonly on leaflets of BEV compared with SEV (59 versus 36 percent, respectively) [25]. In addition, mitral valve infection (including leaflet perforation and regurgitation) has been observed in 20 percent of patients with TAVI-PVE [25].
In an expansion of the above multinational registry to include 579 patients, 105 patients (18 percent) were reported to have perivalvular extension of infection, found with various imaging techniques or at surgery [26]. Extension occurred with similar frequency among patients with SEV and BEV. Perivalvular abscesses were the most common lesion (noted in 87 patients), followed by pseudoaneurysms, fistulae, and a combination of lesions in 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 larger vegetations were more frequent in patients with perivalvular extension.
CLINICAL MANIFESTATIONS — Patients with PVE present with symptoms and signs similar to those encountered in native valve endocarditis (NVE); however, many patients with PVE present with nonspecific symptoms. For example, fever, chills, anorexia, and weight loss, particularly if these are persistent or otherwise unexplained, may be manifestations of PVE. Other symptoms include malaise, headache, myalgias, arthralgias, night sweats, abdominal pain, dyspnea, cough, and pleuritic pain.
In patients with PVE, the frequency of new or changing murmurs, heart failure, and new electrocardiographic conduction disturbances is higher than in patients with NVE [4,59,62]. In addition, patients with early PVE may present with postoperative manifestations that are more prominent than features of endocarditis. Supportive signs of infective endocarditis include cutaneous manifestations such as petechiae or splinter hemorrhages. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis", section on 'Symptoms and signs'.)
Complications of PVE include intracardiac structural damage and central nervous system events, septic emboli, metastatic infection, and systemic immune reactions. Clinical manifestations reflecting these complications may be present at the time of initial presentation and/or may develop subsequently. The incidence of clinically overt arterial emboli is 40 percent; central nervous system complications, primarily embolic infarcts or hemorrhages, occur in 20 to 40 percent of cases [55,63-65]. Clinical manifestations of a complication warrant independent diagnostic evaluation, concurrent with evaluation for PVE. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis", section on 'Complications as initial presentation' and "Complications and outcome of infective endocarditis".)
Manifestations of invasive infection include valvular dysfunction, persistent fever ≥9 days despite appropriate antibiotic therapy, new electrocardiographic conduction disturbances, and cardiac imaging evidence of abscess, fistulae, or aneurysm formation [26,62,66,67]. In one study including more than 100 patients with PVE after surgical valve replacement, these findings were observed in 64 percent of patients; they occurred more frequently in the setting of aortic valve involvement and in the first year after valve replacement [62].
On admission of patients with TAVI-PVE, fever is 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 [68]. Rates of heart failure on admission range from 22 to 60 percent [25,41,68]. TAVI-PVE, which tends to occur in older patients, may present with nonspecific symptoms (anorexia, weight loss, and fatigue) and fever may be blunted. Echocardiography, other imaging, and surgery have demonstrated that invasive infection, including abscesses, fistulae, and pseudoaneurysms occur in 18 percent of patients with TAVI-PVE [26]. (See 'Pathology' above.)
DIAGNOSIS
Overview of diagnostic approach — The diagnosis of PVE should be considered in patients with history of valve replacement who present with any of the following:
●Bacteremia with an organism commonly causing PVE (see 'Microbiology' above)
●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
●Unexplained stroke or systemic emboli
The diagnosis of PVE is established based on clinical manifestations, blood cultures (or other microbiologic data), and echocardiography. The accepted criteria for diagnosis of infective endocarditis are the 2023 Duke-International Society for Cardiovascular Infectious Disease (ISCVID) criteria [69]. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis", section on '2023 Duke-ISCVID criteria'.)
However, the sensitivity of these criteria for diagnosis of PVE is lower than their sensitivity for native valve endocarditis (NVE). Therefore, in the setting of persistent clinical suspicion for PVE but "possible" or "rejected" endocarditis based on the modified Duke criteria or when TTE and TEE are not diagnostic, additional cardiac imaging should be pursued if feasible.
To address these limitations, particularly in the setting of suspected PVE, abnormalities in multislice electrocardiogram-gated cardiac computed tomography (with or without angiography) and nuclear imaging (18F-fluorodeoxyglucose positron emission computed tomography [18F-FDG PET/CT]) and radiolabeled leukocyte single photon emission computed tomography (SPECT/CT) have been included as major criteria in the 2015 European Society of Cardiology criteria for the diagnosis of endocarditis [2]. (See 'Diagnostic (2023 Duke-ISCVID) criteria' below and 'Cardiac imaging' below.)
At least three sets of blood cultures should be obtained from separate venipuncture sites prior to initiation of antibiotic therapy. For patients who are clinically stable, antimicrobial therapy may be deferred while awaiting the results of blood cultures and other diagnostic tests. For patients with signs of clinical instability or sepsis, initiation of empiric antimicrobial therapy (after blood cultures have been obtained) is appropriate. Follow-up blood cultures should be obtained 48 to 72 hours after antimicrobial therapy is begun and repeated every 48 to 72 hours until clearance of bacteremia is documented. (See 'Blood cultures' below and "Detection of bacteremia: Blood cultures and other diagnostic tests" and "Antimicrobial therapy of prosthetic valve endocarditis".)
Echocardiography should be performed in all patients with suspected PVE (algorithm 1) [1,70-72]. Transthoracic echocardiography (TTE) is often the initial study; however, transesophageal echocardiography (TEE) has higher sensitivity than TTE for both the diagnosis of PVE and the detection of perivalvular extension of infection. Accordingly, in the absence of contraindications when TTE findings are nondiagnostic, TEE should be pursued for diagnostic purposes as well as to assess for perivalvular infection. If initial TEE is negative or indeterminate and clinical suspicion for PVE persists, repeat TEE should be pursued (5 to 7 days later). (See 'Echocardiography' below.)
When the diagnosis of PVE remains uncertain following TEE and/or in circumstances where TEE is not feasible, electrocardiogram-gated cardiac computed angiography (CTA), 18F-FDG PET/CT, and SPECT/CT may be useful imaging tools. (See 'Additional imaging tools' below.)
Additional diagnostic evaluation for patients with suspected or known PVE includes electrocardiography (ECG), chest radiography, and additional radiographic imaging tailored to clinical manifestations. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis".)
●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 (which should prompt further evaluation with echocardiography as discussed below). 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.
●Computed tomography (CT) – CT of the torso (chest, abdomen, and pelvis) is not routinely done but may be used to evaluate for sites of metastatic infection (such as splenic infarct, renal infarcts, psoas abscess, or other sites of infection) that may warrant localized drainage [1,73]. The decision to pursue this imaging should be guided by a careful history and clinical assessment, because of potential contrast associated nephrotoxicity.
●Additional radiographic imaging to evaluate for complications of PVE should be tailored to findings on history and physical examination [1]. As examples, patients with back pain should be evaluated for vertebral osteomyelitis with magnetic resonance imaging (MRI), and patients with headache, neurologic deficits, or meningeal signs should be evaluated with head MRI/magnetic resonance angiogram for neurologic complications (including intracranial mycotic aneurysm 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 both patients with and without clinical neurologic findings [1,74-77]. The presence of such lesions can be diagnostically important. For example, in one prospective study, brain MRI (with angiography) in 130 patients with suspected endocarditis, of whom only 16 had acute neurologic symptoms, detected findings related to endocarditis in all patients with neurologic symptoms and 79 percent of patients without clinical neurologic 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 [74].
Diagnostic (2023 Duke-ISCVID) criteria — The accepted criteria for diagnosis of infective endocarditis are the 2023 Duke-International Society for Cardiovascular Infectious Disease (ISCVID) criteria [69]. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis", section on '2023 Duke-ISCVID criteria'.)
However, the sensitivity of the modified Duke criteria for diagnosis of PVE is lower than the sensitivity for diagnosis of NVE, given limitations of echocardiography, including TEE. In two studies of pathologically confirmed cases of PVE involving surgically implanted valves, assessed by the Duke criteria, 76 percent were categorized as definite PVE and 24 percent as possible PVE [78,79].
For this reason, we agree with the diagnostic criteria issued by the 2015 European Society of Cardiology guidelines, which include intracardiac abnormalities detected by multislice electrocardiogram-gated cardiac computed tomography (with or without angiography) and nuclear imaging with 18F-FDG PET/CT or SPECT/CT as major diagnostic criteria [2]. These additional imaging techniques should be used for diagnostic clarification in the setting of persistent clinical suspicion of PVE but only "possible" or "rejected" endocarditis based on the Duke criteria or when therapeutic decisions require additional information regarding intracardiac pathology. (See 'Additional imaging tools' below.)
Diagnostic tools
Blood cultures
General principles — At least three sets of blood cultures should be obtained from separate venipuncture sites prior to initiation of antibiotic therapy [80,81]. In the absence of prior antibiotic therapy, blood cultures will be positive in ≥90 percent patients with PVE. Because bacteremia is continuous, blood cultures will be positive regardless of whether or not they are obtained in proximity to the fever. When 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.
The duration of documented bacteremia is particularly important when the isolate is an organism that is commonly considered a contaminant, such as coagulase-negative staphylococci, Corynebacterium, or Cutibacterium. A high rate of blood culture positivity, or molecular evidence that a sporadically isolated organism represents a single clone, helps to distinguish infecting pathogens from contaminants [82].
Blood cultures for diagnosis of endocarditis are discussed further separately. (See "Clinical manifestations and evaluation of adults with suspected left-sided native valve endocarditis".)
Culture-negative endocarditis — Culture-negative endocarditis is defined as endocarditis with no definitive microbiologic etiology following inoculation of at least three independently obtained blood samples into a standard blood-culture system after five days of incubation and subculturing.
If antibiotics have not been administered prior to obtaining blood cultures, it is unusual to have persistently negative blood cultures in patients with PVE caused by the commonly encountered bacteria causing endocarditis. Nevertheless, 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. (See "Endocarditis caused by Bartonella".)
Rare cases of blood culture-negative PVE (and other cardiac surgery related focal infections) have been caused by M. chimaera, a non-tuberculosis mycobacterium. Contamination of heater-cooler units used during cardiac surgery has resulted in intraoperative exposure of patients to aerosolized organisms [83-85]. Onset of these infections after cardiac surgery has often been very delayed and indolent [85]. (See "Overview of nontuberculous mycobacterial infections", section on 'M. chimaera associated with cardiac surgery'.)
Detection of these and other unusual fastidious pathogens causing PVE relies upon the same evaluation used to assess culture-negative NVE. Next-generation sequencing of cell-free pathogen deoxyribonucleic acid (DNA) from plasma is an emerging technology that may facilitate rapid detection of fastidious pathogens causing infective endocarditis or organisms where growth in blood cultures has been inhibited by recently administered antibiotics [86]. 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 infected prosthetic valve where the causative pathogen is unclear or the significance of a microbe recovered from a blood culture (eg, a common skin colonizer such as S. epidermidis) is uncertain. The technique potentially provides both organism identification at the species level as well as histopathologic information regarding organism location [87]. While not widely available, this is an innovative approach to challenging diagnoses.
Cardiac imaging
Echocardiography — Echocardiography is an important tool for diagnosis of PVE. TTE is often the initial study and provides useful information regarding cardiac function. However, TEE is the study of choice for detection of vegetations as well as other complications including abscess, fistula, leaflet perforation, pseudoaneurysm, and paraprosthetic leak [88]; therefore, TEE should always be performed in patients with suspected PVE (in the absence of contraindications).
Following surgical valve replacement, the sensitivity of TEE is greater than that of TTE (86 to 92 percent versus 17 to 52 percent, respectively), particularly for assessing a mitral valve prosthesis or perivalvular complications [89-97]. In one study of 114 endocarditis episodes (34 PVE, 80 NVE) who underwent TTE and TEE, the Duke criteria classification results were concordant in 63 cases (55 percent). TEE results led to a reclassification by Duke criteria as definite in 22 cases and possible in 3 cases. Reclassification occurred more frequently among patients with endocarditis involving prosthetic valves versus native valves (34 versus 11 percent), including 10 patients reclassified as definite PVE [95].
The negative predictive value of the modified Duke criteria, including a complete echocardiographic evaluation, in a patient with PVE may be as low as 60 to 65 percent [96,97]. If PVE is strongly suspected in the setting of negative echocardiography, repeat echocardiography (5 to 7 days later) may establish a diagnosis of PVE in some cases [91,98]. 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 [99]. If the diagnosis of PVE remains uncertain, other imaging technologies, if available, should be considered. (See 'Additional imaging tools' below.)
Similarly in the diagnosis of TAVI -PVE, the sensitivity of echocardiography (including TEE) is limited due to artifacts and shadowing caused by the large metal stents anchoring the valve. Among 578 patients with TAVI-IE, echocardiography failed to detect evidence of endocarditis in 87 patients (15 percent) [68]. Accordingly, alternative imaging modalities may be needed for the assessment of suspected TAVI-PVE. (See 'Additional imaging tools' below.)
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".)
Additional imaging tools — The most useful alternative cardiac imaging tools for diagnosis of PVE include CTA and 18F-FDG PET/CT [2]; thus far, other imaging modalities (including intracardiac echocardiography, cardiac MRI, and SPECT/CT) are less well studied.
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. The use of such imaging tools depends in large part on their local availability and the expertise of local staff and is often guided by the diagnostic or management questions that remain after traditional echocardiographic imaging has been performed. There are multiple subtle aspects of modality selection, images obtained, image interpretation, and even preparation of patients for optimal imaging that suggest optimal application of imaging strategies requires expertise in the imagers and prestudy consultation between managing physicians and imagers [67].
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 [100-103]. CTA may also provide satisfactory coronary imaging for patients at intermediate risk of coronary artery disease who require surgical intervention [102,103].
18F-FDG PET/CT is useful for diagnosis of select cases of PVE in which echocardiography is not diagnostic or when clarification of perivalvular extension is needed. With this modality, uptake of positron-labeled glucose by inflammatory leukocytes allows anatomic localization of infection. The sensitivity of this modality 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 [104]:
●In one study including more than 90 patients (with indwelling prosthetic valve[s] and/or other intracardiac devices) with suspected endocarditis who were evaluated with TEE, 18F-FDG PET/CT, and CTA, patients were initially classified according to the modified Duke criteria [96]. 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 another study, more than 180 patients with suspected PVE were classified diagnostically according to the modified Duke criteria and then were reclassified with 18F-FDG PET/CT findings included as an additional major criterion [97]. 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) of the modified Duke criteria for diagnosis of PVE. 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 [105]. 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 [106]. 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 [107].
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
●Prosthetic valve endocarditis (PVE) refers to infection of one or more prosthetic heart valves. In early PVE (during the initial two 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 (infection reflecting infection at the time of surgery or during the surgical admission) and community-acquired infections. In late PVE (>12 months postoperatively), the pathogens tend to be bacteremic isolates similar to those inducing native valve endocarditis (NVE). The microbiology of PVE typically reflects the time since implantation (table 3). (See 'Definitions and pathogenesis' above and 'Microbiology' above.)
●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 2). (See 'Epidemiology' above.)
●Transcatheter aortic valve implantation PVE (TAVI-PVE) is a new and 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 TAVI-PVE is similar to that of PVE involving surgically implanted aortic bioprosthetic valves. (See 'Transcatheter aortic valve implantation' above.)
●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.)
●The diagnosis of PVE should be suspected in patients with history of valve replacement who present with any of the following (see 'Overview of diagnostic approach' 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
•Unexpected stroke or systemic emboli
●The diagnosis of PVE is established based on clinical manifestations, blood cultures (or other microbiologic data), and echocardiography. The accepted criteria for diagnosis of infective endocarditis are the 2023 Duke-International Society for Cardiovascular Infectious Disease (ISCVID) criteria. However, the sensitivity of these criteria for diagnosis of PVE is lower than their sensitivity for NVE; therefore, in the setting of persistent clinical suspicion for PVE but 'possible' or 'rejected' endocarditis based on the modified Duke criteria, additional cardiac imaging should be pursued if feasible. (See 'Overview of diagnostic approach' above and 'Diagnostic (2023 Duke-ISCVID) criteria' above and 'Additional imaging tools' above.)
●At least three sets of blood cultures should be obtained from separate venipuncture sites prior to initiation of antibiotic therapy. For patients who are clinically stable, antimicrobial therapy may be deferred while awaiting the results of blood cultures and other diagnostic tests. For patients with signs of clinical instability, initiation of empiric antimicrobial therapy (after blood cultures have been obtained) is appropriate. (See 'Overview of diagnostic approach' above.)
●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 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' above.)
●Additional cardiac imaging tools include 18F-fluorodeoxyglucose positron emission computed tomography (18F-FDG PET/CT) and electrocardiogram-gated cardiac computed angiography (CTA); these tools may be 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 for detection of perivalvular infection, abscess, or pseudoaneurysm. (See 'Additional imaging tools' above.)
●Additional evaluation for patients with suspected PVE includes electrocardiography, chest radiography, and other radiographic imaging tailored to clinical manifestations. (See 'Overview of diagnostic approach' above.)
آیا می خواهید مدیلیب را به صفحه اصلی خود اضافه کنید؟
