There is increasing evidence that fluorodeoxyglucose (FDG) positron emission tomography/computed tomography may be useful in infective endocarditis management, both for detection of septic emboli and for diagnosis of prosthetic valve endocarditis when echocardiography is inconclusive. Such evidences lead to inclusion of the FDG positron emission tomography/computed tomography result as a major criterion of prosthetic valve endocarditis in European Society of Cardiology guidelines, whereas the last American Heart Association recommendations on endocarditis state that more experience is needed to assess the diagnostic utility of FDG positron emission tomography/computed tomography. The present study shows that noninfected prosthetic heart valves often display a homogeneous FDG uptake. This pattern is present even years after valve implantation and should not be considered, per se, as a marker of prosthetic material infection. In addition, the intensity of the FDG uptake did not decrease according to time from valve surgery and seemed to be greater in patients with a history of vasculitis.
Characterization of 18F-Fluorodeoxyglucose Uptake Pattern in Noninfected Prosthetic Heart Valves
Circulation: Cardiovascular Imaging
Abstract
Background—
18F-Fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) has been recently acknowledged as a diagnostic tool for prosthetic valve endocarditis, but its specificity is limited by uptake on noninfected valves. The objective of this study was to outline the main features of FDG uptake on PET/CT in patients with noninfected prosthetic heart valve (PHV).
Methods and Results—
Our institution’s PET/CT database was reviewed to identify patients with PHV, excluding those suspected of infection or who had received antibiotic treatment. PET indication, valve location, and type (biological/mechanical) and time from implantation were collected for each patient. Images with and without attenuation correction were considered for interpretation. The pattern of FDG uptake (absent, homogeneous, or heterogeneous) was recorded. Fifty-four PHVs (51 patients) were identified, including 32 biological valves. Indications for PET were oncology (n=26), suspicion of prosthetic valve endocarditis subsequently excluded (n=17), and history of vasculitis (n=11). A periprosthetic FDG uptake was present in 47 (87%) and 30 (56%) PHVs with and without attenuation correction, respectively, and the pattern was homogeneous in all but 4 (7%) and 3 (6%) PHVs, respectively. On quantitative analysis, maximum standardized uptake values was greater in mechanical than in biological valves (4.0 [2.4–8.0] versus 3.3 [2.1–6.1]; P=0.01) and in patients with vasculitis than in those referred for other indications. The uptake intensity did not differ before and 3 months after valve replacement.
Conclusions—
Noninfected PHVs frequently display homogeneous FDG uptake, which remains steady over time. Caution is, therefore, needed when interpreting FDG PET/CT in suspected prosthetic valve endocarditis, with specific attention to uptake pattern.
Introduction
Prosthetic valve endocarditis (PVE) represents 10% to 30% of all cases of endocarditis and occurs in 1% to 6% of patients with valve prosthesis.1,2 It affects mechanical and bioprosthetic valves equally. PVE is associated with an in-hospital mortality rate of 20% to 40% and is even higher when the infecting agent is Staphylococcus aureus.3–6 Such a high mortality rate is at least in part due to a delay in treatment initiation because of the difficulty in diagnosing infection on prosthetic valves. The direct evidence of vegetation or perivalvular complications of infection—a major criterion in the Duke–Li diagnostic criteria7—relies on echocardiography, but the presence of composite material generating acoustic shadowing results in inconclusive results in a substantial proportion of patients8 and loss of diagnostic accuracy in PVE compared with native valve endocarditis, both for transthoracic and transesophageal echocardiography.6
See Editorial by Sambuceti et al
There is increasing evidence that FDG positron emission tomography (PET)/computed tomography (CT) may be useful in the management of infective endocarditis, both for detection of septic emboli9–13 and for diagnosis of prosthetic valve infection.14–16 A study showed that the sensitivity of the modified Duke criteria performed at admission was markedly increased when the presence of an abnormal FDG uptake was added as a major criterion.14 Such evidence lead to the inclusion of the FDG PET/CT results as a major criterion of PVE in recently published European Society of Cardiology guidelines.17 However, the latest American Heart Association guidelines on endocarditis state that more experience is needed to assess the diagnostic utility of FDG PET/CT.18 It is acknowledged that a mild FDG uptake in the perivalvular area may occur in the absence of infection and may, therefore, be considered as a normal pattern, in particular, early after valve surgery.17 In addition, the factors used to distinguish normal and abnormal patterns of FDG uptake are not standardized. To improve the diagnostic performance of FDG PET/CT, and particularly its specificity, it is critical to precisely describe FDG uptake in noninfected prosthetic heart valve (PHV).
We hypothesized that FDG uptake may be present in the perivalvular area in the absence of PVE. Secondary objectives were (1) to characterize the prevalence and pattern of FDG uptake in noninfected PHV, (2) to evaluate whether this uptake would disappear after a certain time interval after PHV implantation, and (3) to investigate the potential impact of patients’ medical history on this pattern.
In the present study, we included patients with a history of cardiac valve replacement who were referred for PET/CT for a purpose unrelated to PVE. This approach allowed the recruitment of a large array of patients. We then analyzed the pattern and intensity of FDG uptake in the perivalvular area. Demographic, clinical, and biological parameters likely to affect FDG uptake, such as the type of valve, the time interval after implantation, and underlying disease, were also analyzed.
Methods
Study Population
The flowchart for patient selection is presented in Figure 1. The PET/CT database of our institution was reviewed for 24 months to identify patients with PHVs. They were divided into 3 groups according to the clinical indication for PET/CT referral: oncology (initial staging of cancer or suspicion of relapse), history of large vessel vasculitis (patients referred for follow-up without immunosuppressive therapy or under stable background therapy), and suspicion of PVE subsequently rejected. In the latter group, the Duke score was applied after ≥3-month follow-up or at the time of cardiac surgery. For all other patients, definite or possible infection during a follow-up of ≥3 months after the PET/CT scan, regardless of the location, was a criterion of exclusion. Patients who received antibiotic therapy, whatever the indication, in the period from 2 weeks before until 6 weeks after PET/CT scan were also excluded from the analysis. Patients with intense FDG physiological myocardial uptake, preventing analysis of the valve plane, were also excluded to comply with clinical practice in the setting of suspicion of PVE. To this purpose, all PET/CT scans were graded for myocardial uptake in a 3-level scale: absence of uptake when the FDG uptake on cardiac walls was lower than that of blood pool, faint or heterogeneous when uptake was the same level as the blood pool or greater but with areas with no uptake, and intense when left ventricular uptake was intense and homogeneous. Only the latter group of patients was excluded from further analysis. When sequential PET/CT had been performed in the same patient, only the first scan after valve replacement was considered so that each patient could only be analyzed once. For each patient, the following data were collected: age, sex, leukocyte count, C-reactive protein value, diabetic status, as well as those pertaining to the PHV—date of implantation, location (aortic or mitral), and type (biological or mechanical). The Institutional Review Board of Bichat Hospital approved this study, and all subjects gave informed consent for review of their records.
18F-FDG PET/CT
Imaging Protocol
Patients were asked to fast for at least 6 hours before FDG injection. In diabetic patients, blood glucose plasma level had to be <12.0 mmol/L at the time of FDG injection. FDG was injected intravenously at a dose of 4 MBq/kg. PET/CT was performed 1 hour later on a PET/CT hybrid system (Discovery 690; General Electric Medical System, Buc, France). Imaging started with a nonenhanced, low-dose CT (120 kV, 80 mA) and was followed by whole-body PET acquisition in 3-dimensional mode, with an acquisition time of 3 minutes per bed position. PET images were reconstructed using 3-dimensional time-of-flight ordered-subsets expectation maximization with and without attenuation correction transaxial slices (256×256 matrix).
Interpretation Criteria
All cases were reviewed by an experienced nuclear medicine physician blinded to the type of cardiac valve and time interval from valve implantation. Image analysis was performed on ADW workstations (GE Medical Systems, Buc, France) and included both visual analysis and quantification of FDG uptake. First, the pattern of physiological myocardial FDG uptake was noted in 3 groups: absent, mild, or intense. Then the periprosthetic FDG uptake was qualitatively assessed on images that were corrected (AC) or not (NAC) for attenuation, on oblique views reoriented so that the plane of the slice coincided with the plane of the PHV, and quoted as (1) absent; (2) homogeneous when the uptake was diffuse around the PHV ring without focal enhancement; (3) heterogeneous when the uptake was either focal or diffuse with a focal enhancement (Figure 2). The perivalvular FDG uptake (perivalvular maximum standardized uptake value [SUV-V]) was measured on the same reoriented AC views to draw regions of interest in the plane of the valve. Planar circular regions of interest encompassing the perivalvular area were drawn on 3 adjacent slices, and the maximum standardized uptake values (SUVmax) of each slice were averaged as previously reported.14–16 For the purpose of normalization of FDG intensity and to provide measurements comparable with previous publications on this topic,14–16 the mean SUV in blood pool (SUV-B) was measured by manually positioning a 3-dimensional volume of interest in the right atrial cavity (avoiding atrial walls). The prosthetic valve-to-background ratio was calculated as SUV-V/SUV-B. Because in patients with a history of vasculitis, the FDG uptake in the aorta may have confounded the perivalvular assessment of PHV in aortic position, we also quantified FDG uptake in the ascending aorta. To that purpose, planar circular regions of interest encompassing the ascending aorta were drawn on adjacent slices, and the maximum standardized uptake value was referred to as SUV-aorta. The prosthetic aorta-to-background ratio was calculated as SUV-aorta/SUV-B.
Statistical Analysis
Statistical analysis was performed using the R software (version R.3.2.2; the R Foundation). Continuous variables were expressed by median and range. Categorical variables were expressed by their percentage. Comparisons across referral indication groups were performed using Fisher exact test for categorical variables and the Kruskal–Wallis test for continuous variables. Comparisons by valve type or valve position were performed using Fisher exact test for categorical variables and either the t test or Mann–Whitney test for continuous variables. The relationship between SUV-V (or SUV-V/SUV-B) and time interval after valve implantation was expressed using Pearson’s correlation coefficient. Comparisons of SUV-V and SUV-V/SUV-B between the 3 referral groups were performed using analysis of variance. Multiple linear regression analysis was used to investigate the relationships between SUV-V or SUV-V/SUV/B and variables related to the patients or the valves (valve type, valve position, and delay after implantation). A value of P<0.05 was considered significant.
The reproducibility of qualitative and quantitative assessment of perivalvular FDG uptake was checked by 2 experienced observers (Drs Mathieu and Rouzet) in patients included during the first 12 months of the study (n=36). The intraobserver and interobserver agreement of perivalvular uptake pattern (absent, homogeneous, or heterogeneous) was assessed by weighted Kappa coefficient, and the agreement of perivalvular FDG uptake intensity (SUV-V and SUV-V/SUV-B ratio) was expressed by Bland–Altman plot and compared using the intraclass correlation coefficient.
Results
Study Population
Among 585 patients with a PHV identified during the screening period, 51 met the inclusion criteria (Figure 1), with a total of 54 prosthetic valves. The selected patient characteristics are presented in Table 1. Out of 25 patients referred for oncological indication, most of them were investigated in the setting of initial staging of suspected (n=11) or proven (n=9) cancer. They had not had previous chemotherapy or radiotherapy when they underwent the PET/CT scan. Five additional patients were investigated for suspicion of cancer relapse. In this latter group, none had a history of radiotherapy involving the chest, and they received the last chemotherapy treatment at least 12 months before the PET/CT scan. After a follow-up of ≥3 months (median, 5 months; range, 3–12), none developed PVE. Out of 11 patients with a history of large vessel vasculitis, 3 presented with Takayasu disease, 3 with giant cell arteritis, 2 with Behcet disease, and 3 with other vasculitis. Among them, 5 were under stable background immunosuppressive therapy when they underwent the scan. After a follow-up of ≥3 months (median, 12 months; range, 3–12), none developed PVE. The main characteristics of the 15 patients referred for suspicion of PVE are presented in Table 2. Briefly, the suspicion was raised by either the occurrence of fever or an abnormal finding on echocardiography. The diagnosis of PVE had been ruled out by a diagnosis of rejected endocarditis as assessed by the Duke classification after follow-up in all patients. Three of them underwent valve replacement during the month after the PET scan, and the pathological analysis of the valve sample led to a conclusion of bioprosthesis degeneration in 2 patients and thrombus in 1 patient. One patient underwent heart transplant for advanced heart failure 3 months after the PET/CT scan without evidence of PVE at histological examination. In the 11 remaining patients, after a follow-up of >3 months (median, 6 months; range, 3 to 12), an alternative diagnosis was found in 4 patients, and no event occurred in 7 patients. No patients were lost to follow-up. It is noteworthy that most patients included in the study presented biological signs of inflammation with high values of leukocyte count and C-reactive protein, without significant difference according to referral indication (Table 1).
| Overall Population (n=51) | Oncology (n=25) | Suspected Infection (n=15) | Vasculitis (n=11) | P Value | |
|---|---|---|---|---|---|
| Age, y | 66 (25–85) | 70 (39–82) | 66 (35–85) | 55 (25–85) | 0.02 |
| Male sex, n (%) | 29 (57) | 19 (76) | 4 (27) | 6 (55) | 0.01 |
| Leukocyte count, 109/L | 7.27 (4.60–33.92) | 7.29 (4.60–33.92) | 6.29 (4.77–11.61) | 6.98 (5.76–9.60) | 0.8 |
| CRP value, mg/L | 10 (3–213) | 11.5 (3–213) | 10 (3–169) | 7 (3–117) | 0.4 |
| Glycemia,* mmol/L | 5.8 (1.8–11.3) | 6.0 (3.3–11.3) | 5.6 (1.8–9.1) | 5.6 (4.1–6.7) | 0.8 |
Continuous variables are expressed as median (range). CRP indicates C-reactive protein.
*
At the time of 18FDG injection.
| Valve Position/Type | Time Interval After Valve Replacement, months | Symptoms/Signs | Echocardiography | FU Duration, months | Event | Final Diagnosis |
|---|---|---|---|---|---|---|
| Aortic/Biol | 3 | Fever | Normal | 3 | None | No infection |
| Mitral/Biol | 1 | None | 3-mm mass on mitral annulus | 3 | None | Suture thread |
| Mitral/Mecha | 2 | Fever | Normal | 3 | None | No infection |
| Mitral/Biol | 1 | None | Paraprosthetic leakage | 3 | Heart transplantation | No infection |
| Mitral/Biol | 1 | None | Paraprosthetic mass | 1 | Mitral valve replacement | Thrombus |
| Mitral/Biol | 1 | None | Mass attached to the valve leaflet | 6 | Regression under anticoagulant treatment | Thrombus |
| Mitral/Mecha | 195 | Fever | Normal | 3 | None | No infection |
| Aorta/Biol (homograft after aortic endocarditis) | 206 | Dyspnea | Pseudoaneurysm of the aortic annulus | 1 | Aortic and mitral valve replacement | Homograft degeneration |
| Aorta/Biol | 40 | Fever | Normal | 6 | None | No infection |
| Aorta/Biol | 93 | Dyspnea | Aortic valve leakage | 1 | Aortic valve replacement | Bioprosthesis degeneration |
| Aorta/Biol | 47 | Suspicion of mycotic aneurysm | Normal | 12 | None | No infection |
| Aorta/Biol | 40 | None | Thickening of the prosthetic valve leaflets | 12 | None | Early bioprosthesis degeneration |
| Aorta/Biol | 7 | Inflammatory syndrome | Normal | 12 | None | No infection |
| Mitral/Mecha | 228 | Dyspnea | Paravalvular regurgitation | 6 | None | No infection |
| Aorta/Biol | 137 | Congestive heart failure | Mass attached to the mitral annulus | 12 | None | Mitral annular calcification |
The most recent valve surgery has been considered in the 2 patients who have 2 prosthetic valves. Biol indicates biological; FU, follow up; and Mecha, mechanical.
Valve characteristics are detailed in Table 3. The prosthetic valves were mostly biological (59%) and in aortic position (65%). The time interval between valve implantation and PET/CT ranged from 9 days to 25 years and was not statistically different between groups based on referral indication (Table 3). However, patients with biological valves were investigated earlier after implantation than those with mechanical valves (median, 3.6 versus 7.8 years; P=0.004). Fifteen patients were investigated <3 months after valve implantation, mainly with biological PHV (Table 4).
| All PHV (n=54) | Oncology (n=26) | Suspected PVE (n=17) | Vasculitis (n=11) | P Value | |
|---|---|---|---|---|---|
| Valve position, n (%) | |||||
| Aorta | 35 (65) | 17 (65) | 10 (59) | 8 (73) | 0.8 |
| Mitral | 19 (35) | 9 (35) | 7 (41) | 3 (27) | |
| Valve type, n (%) | |||||
| Biological | 32 (59) | 16 (62) | 13 (76) | 3 (27) | 0.03 |
| Mechanical | 22 (41) | 10 (38) | 4 (24) | 8 (73) | |
| Delay between valve implantation and imaging, days | 1957 (9–9044) | 2361 (9–9044) | 2039 (25–6952) | 786 (34–2335) | 0.1 |
Continuous variables are expressed as median (range). PVE indicates prosthetic valve endocarditis.
*
Three patients had 2 prosthetic valves (2 in the group suspected PVE and 1 in the group vasculitis), which were analyzed independently.
| All PHV (n=54) | Valve Type | Valve Position | |||||
|---|---|---|---|---|---|---|---|
| Biological (n=32) | Mechanical (n=22) | P Value | Aortic (n=35) | Mitral (n=19) | PValue | ||
| Time from implantation, n (%) | |||||||
| <3 mo | 15 (28) | 11 (34) | 4 (18) | 0.06 | 7 (20) | 8 (42) | 0.1 |
| >3 mo and <1 y | 4 (7) | 4 (13) | 0 (0) | 4 (11) | 0 (0) | ||
| >1 y | 35 (65) | 17 (53) | 18 (82) | 24 (69) | 11 (58) | ||
| Periprosthetic FDG uptake pattern | |||||||
| Absence of uptake, n (%) | |||||||
| AC | 7 (13) | 6 (19) | 1 (4.5) | 0.2 | 5 (14) | 2 (11) | 1 |
| NAC | 24 (44) | 18 (56) | 6 (27) | 0.03 | 17 (49) | 7 (37) | 0.1 |
| Homogeneous uptake, n (%) | |||||||
| AC | 43 (80) | 25 (78) | 18 (82) | 1 | 27 (77) | 16 (84) | 0.8 |
| NAC | 27 (50) | 13 (41) | 14 (64) | 0.2 | 16 (46) | 11 (58) | 0.6 |
| Heterogeneous uptake, n (%) | |||||||
| AC | 4 (7) | 1 (3) | 3 (13) | 0.3 | 3 (9) | 1 (5) | 1 |
| NAC | 3 (6) | 1 (3) | 2 (9) | 0.6 | 2 (6) | 1 (5) | 1 |
| Periprosthetic FDG uptake intensity | |||||||
| SUV-V | 3.5 (2.1–8.0) | 3.3 (2.1–6.1) | 4.0 (2.4–8.0) | 0.01 | 3.6 (2.1–8.0) | 3.5 (2.4–6.1) | 0.8 |
| SUV-B | 1.8 (1.0–2.8) | 1.8 (1.0–2.8) | 1.8 (1.0–2.7) | 0.9 | 1.8 (1.0–2.7) | 1.8 (1.0–2.8) | 0.8 |
| SUV-V/SUV-B | 2.0 (1.3–6.6) | 1.9 (1.3–5.1) | 2.1 (1.4–6.6) | 0.03 | 1.9 (1.3–6.6) | 2.0 (1.4–5.1) | 0.8 |
Continuous variables are expressed as median (range). AC indicates attenuation-corrected images; NAC, nonattenuation-corrected images; PHV, prosthetic heart valve; SUV-B, mean SUV in blood pool; and SUV-V, perivalvular SUVmax.
Periprosthetic FDG Uptake Pattern and Intensity
FDG uptake pattern and intensity are detailed in Tables 4 and 5. On visual analysis, FDG uptake was detectable in the periprosthetic area more frequently on AC images than on NAC images (87% versus 56%; P=0.0005) and with mechanical prosthesis than with biological ones (Table 4). The uptake was homogeneous in most cases (Table 4) both on AC and NAC images (Figure 2). Accordingly, in the overall population, heterogeneous uptake was present in only 4 PHV (7%) on AC and 3 (6%) on NAC images (Table 4). Among the 4 patients presenting with a heterogeneous uptake on AC images, 3 had a mechanical prosthetic valve in aortic position, and the remaining patient had a biological prosthesis implanted for <3 months (Figure 2C). Interobserver and intraobserver agreement of qualitative assessment of periprosthetic FDG uptake assessed by weighted Kappa coefficient were 0.89 (95% confidence interval, 0.75–1.0).
Quantitative analysis using either SUVmax alone (SUV-V) or corrected for bloodpool activity (SUV-V/SUV-B ratio) further evidenced the higher uptake of mechanical compared with biological valves (Table 4). Interobserver agreement of quantitative assessment of periprosthetic FDG uptake was 0.99 (95% CI, 0.99–1.0 for SUV-V and 0.98 (95% CI, 0.96–0.99) for SUV-V/SUV-B ratio. The mean difference between operators was 0.03 (95% limit of agreement, −0.36; 0.31) for SUV-V and 0.02 (95% limit of agreement, −0.43; 0.47) for SUV-V/SUV-B ratio. Intraobserver agreement of quantitative assessment of periprosthetic FDG uptake was 0.99 (95% CI, 0.99–1.0) for SUV-V and 0.97 (95% CI, 0.95–0.99) for SUV-V/SUV-B ratio. The mean difference between the 2 sets of measurements was 0.01 (95% limit of agreement, −0.17; 0.18) for SUV-V and −0.02 (95% limit of agreement, −0.31; 0.27) for SUV-V/SUV-B ratio.
A subgroup analysis showed that patients referred for large vessel vasculitis presented higher uptake values compared with those referred for oncology or suspicion of PVE (Table 5), with a median SUV-V of 4.7 versus 3.3 and 3.5 and median SUV-V/SUV-B of 2.7 versus 1.9 and 1.8, respectively, (P<0.001 for both SUV-V and SUV-V/SUV-B). Of note, SUVmax of the ascending aorta or SUV-aorta/SUV-B was not significantly greater in patients with vasculitis (Table 5). Although more patients in this subgroup were implanted with a mechanical prosthesis, the association remained significant by multivariate linear regression analysis (P=0.001 for both SUV-V and SUV-V/SUV-B). After exclusion of patients with vasculitis, FDG uptake was detectable in the periprosthetic area in 36 (84%) PHV with AC and in 22 (51%) PHV without AC in the remaining study population (in comparison to patients with vasculitis; P=0.3 and 0.09, respectively); the median SUV-V was 3.5 (2.1–5.7), and the median SUV-V/SUV-B ratio was 1.9 (1.3–3.8). No significant difference in SUV-V or SUV-V/SUV-B ratio was found according to sex (P=0.12), blood glucose level at the time of FDG injection (P=0.48), or diabetic status (P=0.08).
| Oncology (n=26) | Suspected PVE (n=17) | Vasculitis (n=11) | P Value | |
|---|---|---|---|---|
| Qualitative assessment of the perivalvular FDG uptake | ||||
| AC | ||||
| Absence of uptake, n (%) | 3 (11.5) | 4 (23.5) | 0 (0) | 0.5 |
| Homogeneous uptake, n (%) | 21 (81) | 12 (70.5) | 10 (91) | |
| Heterogeneous uptake, n (%) | 2 (7.5) | 1 (6) | 1 (9) | |
| NAC | ||||
| Absence of uptake, n (%) | 13 (50) | 9 (53) | 2 (18) | 0.4 |
| Homogeneous uptake, n (%) | 12 (46) | 7 (41) | 8 (73) | |
| Heterogeneous uptake, n (%) | 1 (4) | 1 (6) | 1 (9) | |
| Quantitative assessment of the perivalvular FDG uptake | ||||
| SUV-V | 3.3 (2.1–5.7) | 3.5 (2.1–4.7) | 4.7 (3.0–8.0) | <0.001 |
| SUV-B | 1.8 (1.0–2.7) | 1.9 (1.3–2.8) | 1.8 (1.5–2.2) | 0.5 |
| SUV-V/SUV-B | 1.9 (1.4–3.8) | 1.8 (1.3–2.3) | 3.1 (1.6–3.9) | <0.001 |
| FDG uptake in the ascending aorta | ||||
| SUV-aorta | 2.7 (2.0–3.6) | 2.9 (1.3–3.5) | 2.5 (2.1–3.7) | 0.4 |
| SUV-aorta/SUV-B | 1.4 (1.2–1.9) | 1.5 (1.2–1.9) | 1.5 (1.2–1.7) | 0.8 |
Continuous variables are expressed as median (range). AC indicates attenuation-corrected images; NAC, non attenuation-corrected images; PVE, prosthetic valve endocarditis; SUV-B, mean standardized uptake values in blood pool; and SUV-V, perivalvular maximum standardized uptake values.
Of note, both qualitative and quantitative analysis of perivalvular FDG uptake yielded similar results whatever the valve position (Table 4), suggesting that the close proximity of FDG physiological faint uptake by the myocardium and PHV in mitral position did not affect the evaluation of the pattern or the SUV of the valve.
Relationship Between FDG Uptake and Time From Implantation
On visual analysis, FDG uptake was detectable in the periprosthetic area more frequently in PHV implanted <3 months prior to the analysis (n=15) than in others, but the difference was not significant (respectively, 93% versus 85%; P=0.7 with attenuation correction; and 67% versus 51%, P=0.4 without attenuation correction). Likewise, there was no significant difference in SUV-V or in the SUV-V/SUV-B ratio between PHV implanted for <3 months and those that had been in place for longer (SUV-V, 3.5 [2.4–5.8] versus 3.5 [2.1–8.0]; P=0.5 and SUV-V/SUV-B ratio, 1.9 [1.4–3.9] versus 2.0 [1.3–6.6]; P=0.5). In addition, SUV-V was not correlated with time from implantation (R=−0.07; P=0.60; Figure 3). A subgroup analysis showed that SUV-V in biological valves was weakly and inversely correlated with time (R=−0.36; P=0.04), but this was not the case for mechanical valves (R=−0.19; P=0.40).
Discussion
The present study shows that noninfected PHV often display a homogeneous FDG uptake that is independent of the time interval between valve implantation and imaging. In addition, the intensity of the FDG uptake seems to be greater in patients with a history of vasculitis.
In the setting of PVE, FDG PET/CT proved to be helpful by providing evidence of periprosthetic hypermetabolism when echocardiography is inconclusive.14,15 It is generally considered that perivalvular uptake is increased during the few months after valve replacement. Indeed, most of the false-positive scans in our previous series occurred in patients with a prosthetic valve implanted <2 months before undergoing FDG PET/CT.15 For this reason, Saby et al14 did not include patients investigated during the first month after valve implantation, and European Society of Cardiology recommendations do not consider PET/CT results during the first 3 months after implantation.17 In the present study, the mean SUVmax was not statistically different whether prostheses were implanted before or after the 3-month threshold. Only a weak correlation between the time from valve replacement and the intensity of FDG uptake could be evidenced in the subgroup of patients with a biological prosthesis. Consequently, the presence of a perivalvular uptake should be interpreted with great caution even after the first 3 postoperative months.
The periprosthetic FDG uptake was present on visual analysis in most patients on AC images. This uptake was somewhat lower on NAC images, suggesting the occurrence of overcorrection artifacts19 related to the presence of composite components of PHV in some patients. Such artifacts have been well investigated in cardiac-implanted electronic devices,20,21 but much less in PHV. The greater values of SUVmax and valve-to-background ratio calculated in the presence of a mechanical valve also support the role of AC because more composite material is present. However, FDG uptake was also present in 56% of patients on noncorrected images. Based on data obtained in patients with a definite or rejected diagnosis of PVE, our group15 and others16 suggested cutoff values of SUVmax and uptake ratios designed to help discriminate between infected and noninfected PHV. In their series of patients suspected of PVE or cardiac implantable electronic devices infection, Pizzi et al16 report a specificity of 79% when SUVmax was ≥3.7 and 100% when SUVmax was ≥6.9. In the present series of patients free from infection, SUVmax ranged from 2 to 5 in most cases and ≤8 in the subgroup of patients with vascular disease. The comparison of the FDG uptake level with previous reports in patients with PVE14–16 shows that even if mean values are greater in patients with definite PVE, there is no threshold that allows reliable differentiation between infected and noninfected PHV because of a wide overlap between values. In the present study, a greater FDG uptake was evidenced in patients with a history of vasculitis but without evidence of active disease. The increased level of FDG uptake was possibly related to the presence of a baseline arterial inflammation in this specific subgroup of patients, and the discrimination between infection and mere inflammation should be even more careful in this subgroup of patients to avoid inappropriate redo surgery.
A homogenous uptake pattern present on visual analysis in almost all cases suggests that this feature was associated with the physiological scarring process around PHVs. Early after valvular implantation, migration and proliferation of inflammatory cells, myofibroblasts, and capillary endothelial cells are triggered by the host tissue reaction,22 which is likely to provoke an increased FDG uptake. Thereafter, the progression of the healing process leads to the formation of a fibrous cuff with increasing amounts of collagen and re-endothelialization of the sewing ring.22 The persistence of FDG uptake in the periprosthetic area, even on noncorrected images, is likely to account for the persistence of a remodeling process around the sewing ring. The variability of periprosthetic FDG uptake in the absence of infection may be related to the variability of the fibrous tissue ingrowth on the sewing ring. Endothelialization of the sewing ring is a physiological healing process that varies from one patient to another and may even consist in a marked fibroblastic proliferation with chronic inflammatory features, leading to obstructive pannus in the most severe cases.23 A similar pattern has been reported by Keidar et al24 in the setting of vascular prosthesis. In their study, the authors showed that FDG uptake could remain intense over time in synthetic prosthesis, whereas it decreased in native vein grafts. Because the sewing ring of PHV and vascular grafts are made of similar synthetic biomaterial, it is likely that a common reaction occurs in both settings.
With regards to the recently issued European Society of Cardiology guidelines on infective endocarditis,17 where the presence of abnormal activity around a PHV is considered as a major criterion of infective endocarditis, the interpretation of FDG PET should be performed carefully. In particular, the pattern of faint or homogeneous perivalvular uptake should not be considered abnormal, regardless of the delay from prosthesis implantation. The present findings also highlight the fact that diagnostic criteria should not be a substitute for the clinical judgment of the endocarditis team, as stated in the guidelines.
Limitations of the Study
Because of the retrospective nature of patients’ enrollment, they did not undergo the specific high-fat low-carbohydrate diet followed by 12 hours fasting intended to minimize FDG cardiac uptake. As a consequence, a substantial proportion presented an intense myocardial FDG uptake. Because they were not representative of the target population, we excluded those patients from analysis so that the results of the study may apply to patients referred for suspicion of PVE who undertook the appropriate diet. In the remaining study population, the spectrum of FDG cardiac uptake pattern was comparable to that of the population of our previous series with a low or mild uptake in most patients.15 Likewise, the mean values of SUVmax in patients without PVE in our previous series and in the present study population were in the same range. Hence, we can infer that the study population is representative of the general population undergoing FDG PET in the setting of suspicion of PVE as regards metabolic status. The inclusion criteria did not allow to formally exclude ongoing vasculitis that may have represented a confounding factor, leading to an overestimation of perivalvular FDG uptake in this subgroup. Although we ruled out any contribution of underlying aortic FDG uptake by determining that the signal in the ascending aorta was not significantly greater in patients with vasculitis compared with the rest of the study population, this finding should be further substantiated by dedicated studies.
Conclusions
Noninfected PHVs frequently display homogeneous FDG uptake in the perivalvular area on PET/CT. This pattern is present years after valve implantation and should not be considered, per se, as a marker of prosthetic material infection. In addition, such FDG uptake was slightly greater in mechanical versus biological prosthesis and in patients with vasculitis. Caution is, therefore, needed when interpreting FDG PET/CT in suspected PVE, and special attention should be paid to the uptake pattern. Finally, these findings highlight the need to integrate the results of FDG PET/CT in the clinical context.
CLINICAL PERSPECTIVE
References
1.
Vongpatanasin W, Hillis LD, Lange RA. Prosthetic heart valves. N Engl J Med. 1996;335:407–416. doi: 10.1056/NEJM199608083350607.
2.
Hoen B, Duval X. Clinical practice. Infective endocarditis. N Engl J Med. 2013;368:1425–1433. doi: 10.1056/NEJMcp1206782.
3.
Anguera I, Miro JM, San Roman JA, de Alarcon A, Anguita M, Almirante B, Evangelista A, Cabell CH, Vilacosta I, Ripoll T, Muñoz P, Navas E, Gonzalez-Juanatey C, Sarria C, Garcia-Bolao I, Fariñas MC, Rufi G, Miralles F, Pare C, Fowler VG, Mestres CA, de Lazzari E, Guma JR, del Río A, Corey GR; Aorto-Cavitary Fistula in Endocarditis Working Group. Periannular complications in infective endocarditis involving prosthetic aortic valves. Am J Cardiol. 2006;98:1261–1268. doi: 10.1016/j.amjcard.2006.05.066.
4.
Wang A, Athan E, Pappas PA, Fowler VG, Olaison L, Paré C, Almirante B, Muñoz P, Rizzi M, Naber C, Logar M, Tattevin P, Iarussi DL, Selton-Suty C, Jones SB, Casabé J, Morris A, Corey GR, Cabell CH; International Collaboration on Endocarditis-Prospective Cohort Study Investigators. Contemporary clinical profile and outcome of prosthetic valve endocarditis. JAMA. 2007;297:1354–1361. doi: 10.1001/jama.297.12.1354.
5.
Chirouze C, Alla F, Fowler VG, Sexton DJ, Corey GR, Chu VH, Wang A, Erpelding ML, Durante-Mangoni E, Fernández-Hidalgo N, Giannitsioti E, Hannan MM, Lejko-Zupanc T, Miró JM, Muñoz P, Murdoch DR, Tattevin P, Tribouilloy C, Hoen B; ICE Prospective Investigators. Impact of early valve surgery on outcome of Staphylococcus aureus prosthetic valve infective endocarditis: analysis in the International Collaboration of Endocarditis-Prospective Cohort Study. Clin Infect Dis. 2015;60:741–749. doi: 10.1093/cid/ciu871.
6.
Habib G, Thuny F, Avierinos JF. Prosthetic valve endocarditis: current approach and therapeutic options. Prog Cardiovasc Dis. 2008;50:274–281. doi: 10.1016/j.pcad.2007.10.007.
7.
Li JS, Sexton DJ, Mick N, Nettles R, Fowler VG, Ryan T, Bashore T, Corey GR. Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis. Clin Infect Dis. 2000;30:633–638. doi: 10.1086/313753.
8.
Hill EE, Herijgers P, Claus P, Vanderschueren S, Peetermans WE, Herregods MC. Abscess in infective endocarditis: the value of transesophageal echocardiography and outcome: a 5-year study. Am Heart J. 2007;154:923–928. doi: 10.1016/j.ahj.2007.06.028.
9.
Bonfiglioli R, Nanni C, Morigi JJ, Graziosi M, Trapani F, Bartoletti M, Tumietto F, Ambrosini V, Ferretti A, Rubello D, Rapezzi C, Viale PL, Fanti S. 18F-FDG PET/CT diagnosis of unexpected extracardiac septic embolisms in patients with suspected cardiac endocarditis. Eur J Nucl Med Mol Imaging. 2013;40:1190–1196. doi: 10.1007/s00259-013-2426-7.
10.
Kestler M, Muñoz P, Rodríguez-Créixems M, Rotger A, Jimenez-Requena F, Mari A, Orcajo J, Hernández L, Alonso JC, Bouza E; Group for the Management of Infectious Endocarditis (GAME). Role of (18)F-FDG PET in patients with infectious endocarditis. J Nucl Med. 2014;55:1093–1098. doi: 10.2967/jnumed.113.134981.
11.
Mikail N, Benali K, Ou P, Slama J, Hyafil F, Le Guludec D, Rouzet F. Detection of mycotic aneurysms of lower limbs by whole-body (18)F-FDG-PET. JACC Cardiovasc Imaging. 2015;8:859–862. doi: 10.1016/j.jcmg.2014.10.019.
12.
Vos FJ, Bleeker-Rovers CP, Sturm PD, Krabbe PF, van Dijk AP, Cuijpers ML, Adang EM, Wanten GJ, Kullberg BJ, Oyen WJ. 18F-FDG PET/CT for detection of metastatic infection in gram-positive bacteremia. J Nucl Med. 2010;51:1234–1240. doi: 10.2967/jnumed.109.072371.
13.
Asmar A, Ozcan C, Diederichsen AC, Thomassen A, Gill S. Clinical impact of 18F-FDG-PET/CT in the extra cardiac work-up of patients with infective endocarditis. Eur Heart J Cardiovasc Imaging. 2014;15:1013–1019. doi: 10.1093/ehjci/jeu054.
14.
Saby L, Laas O, Habib G, Cammilleri S, Mancini J, Tessonnier L, Casalta JP, Gouriet F, Riberi A, Avierinos JF, Collart F, Mundler O, Raoult D, Thuny F. Positron emission tomography/computed tomography for diagnosis of prosthetic valve endocarditis: increased valvular 18F-fluorodeoxyglucose uptake as a novel major criterion. J Am Coll Cardiol. 2013;61:2374–2382. doi: 10.1016/j.jacc.2013.01.092.
15.
Rouzet F, Chequer R, Benali K, Lepage L, Ghodbane W, Duval X, Iung B, Vahanian A, Le Guludec D, Hyafil F. Respective performance of 18F-FDG PET and radiolabeled leukocyte scintigraphy for the diagnosis of prosthetic valve endocarditis. J Nucl Med. 2014;55:1980–1985. doi: 10.2967/jnumed.114.141895.
16.
Pizzi MN, Roque A, Fernández-Hidalgo N, Cuéllar-Calabria H, Ferreira-González I, Gonzàlez-Alujas MT, Oristrell G, Gracia-Sánchez L, González JJ, Rodríguez-Palomares J, Galiñanes M, Maisterra-Santos O, Garcia-Dorado D, Castell-Conesa J, Almirante B, Aguadé-Bruix S, Tornos P. Improving the Diagnosis of infective endocarditis in prosthetic valves and intracardiac devices with 18F-fluordeoxyglucose positron emission tomography/computed tomography angiography: initial results at an infective endocarditis referral center. Circulation. 2015;132:1113–1126. doi: 10.1161/CIRCULATIONAHA.115.015316.
17.
Habib G, Lancellotti P, Antunes MJ, Bongiorni MG, Casalta JP, Del Zotti F, Dulgheru R, El Khoury G, Erba PA, Iung B, Miro JM, Mulder BJ, Plonska-Gosciniak E, Price S, Roos-Hesselink J, Snygg-Martin U, Thuny F, Tornos Mas P, Vilacosta I, Zamorano JL, Erol Ç, Nihoyannopoulos P, Aboyans V, Agewall S, Athanassopoulos G, Aytekin S, Benzer W, Bueno H, Broekhuizen L, Carerj S, Cosyns B, De Backer J, De Bonis M, Dimopoulos K, Donal E, Drexel H, Flachskampf FA, Hall R, Halvorsen S, Hoen B, Kirchhof P, Lainscak M, Leite-Moreira AF, Lip GY, Mestres CA, Piepoli MF, Punjabi PP, Rapezzi C, Rosenhek R, Siebens K, Tamargo J, Walker DM; Document Reviewers. 2015 ESC Guidelines for the management of infective endocarditis: The Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC). Endorsed by: European Association for Cardio-Thoracic Surgery (EACTS), the European Association of Nuclear Medicine (EANM). Eur Heart J. 2015;36:3075–3128. doi: 10.1093/eurheartj/ehv319.
18.
Baddour LM, Wilson WR, Bayer AS, Fowler VG, Tleyjeh IM, Rybak MJ, Barsic B, Lockhart PB, Gewitz MH, Levison ME, Bolger AF, Steckelberg JM, Baltimore RS, Fink AM, O’Gara P, Taubert KA; American Heart Association Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease of the Council on Cardiovascular Disease in the Young, Council on Clinical Cardiology, Council on Cardiovascular Surgery and Anesthesia, and Stroke Council. Infective Endocarditis in Adults: Diagnosis, Antimicrobial Therapy, and Management of Complications: A Scientific Statement for Healthcare Professionals From the American Heart Association. Circulation. 2015;132:1435–1486. doi: 10.1161/CIR.0000000000000296.
19.
Kinahan PE, Hasegawa BH, Beyer T. X-ray-based attenuation correction for positron emission tomography/computed tomography scanners. Semin Nucl Med. 2003;33:166–179. doi: 10.1053/snuc.2003.127307.
20.
DiFilippo FP, Brunken RC. Do implanted pacemaker leads and ICD leads cause metal-related artifact in cardiac PET/CT? J Nucl Med. 2005;46:436–443.
21.
Ghafarian P, Aghamiri SM, Ay MR, Rahmim A, Schindler TH, Ratib O, Zaidi H. Is metal artefact reduction mandatory in cardiac PET/CT imaging in the presence of pacemaker and implantable cardioverter defibrillator leads? Eur J Nucl Med Mol Imaging. 2011;38:252–262. doi: 10.1007/s00259-010-1635-6.
22.
Siddiqui RF, Abraham JR, Butany J. Bioprosthetic heart valves: modes of failure. Histopathology. 2009;55:135–144. doi: 10.1111/j.1365-2559.2008.03190.x.
23.
Vitale N, Renzulli A, Agozzino L, Pollice A, Tedesco N, de Luca Tupputi Schinosa L, Cotrufo M. Obstruction of mechanical mitral prostheses: analysis of pathologic findings. Ann Thorac Surg. 1997;63:1101–1106.
24.
Keidar Z, Pirmisashvili N, Leiderman M, Nitecki S, Israel O. 18F-FDG uptake in noninfected prosthetic vascular grafts: incidence, patterns, and changes over time. J Nucl Med. 2014;55:392–395. doi: 10.2967/jnumed.113.128173.
Information & Authors
Information
Published In
Copyright
© 2017 American Heart Association, Inc.
History
Received: 3 February 2016
Accepted: 12 January 2017
Published in print: March 2017
Published online: 15 March 2017
Keywords
Subjects
Authors
Disclosures
None.
Metrics & Citations
Metrics
Citations
Download Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Select your manager software from the list below and click Download.
- Radionuclide Imaging of Infective Endocarditis, Heart Failure Clinics, 21, 3, (443-459), (2025).https://doi.org/10.1016/j.hfc.2025.01.008
- 18F-fluorodeoxyglucose positron emission tomography computed tomography in patients with prosthetic heart valves: Variations in imaging findings: A review of the literature, Journal of Nuclear Cardiology, (102247), (2025).https://doi.org/10.1016/j.nuclcard.2025.102247
- Infective endocarditis: it takes a team, European Heart Journal, (2025).https://doi.org/10.1093/eurheartj/ehaf219
- [18F]FDG PET/CT in cardiovascular infections: a practical approach, The Quarterly Journal of Nuclear Medicine and Molecular Imaging, 69, 1, (2025).https://doi.org/10.23736/S1824-4785.25.03616-7
- Infection vs inflammation in prosthetic heart valves by [18F]FDG-PET/CT: A practical tutorial for accurate interpretation, Journal of Nuclear Cardiology, 45, (102053), (2025).https://doi.org/10.1016/j.nuclcard.2024.102053
- Cardiac Imaging in the Diagnosis of Infective Endocarditis, Infective Endocarditis, (121-134), (2025).https://doi.org/10.1007/978-3-031-65839-6_11
- Contemporary Multi-modality Imaging of Prosthetic Aortic Valves, Reviews in Cardiovascular Medicine, 26, 1, (2025).https://doi.org/10.31083/RCM25339
- Comparison between Duke, European Society of Cardiology 2015, International Society for Cardiovascular Infectious Diseases 2023, and European Society of Cardiology 2023 criteria for the diagnosis of transcatheter aortic valve replacement-related infective endocarditis, European Heart Journal - Cardiovascular Imaging, 26, 3, (532-544), (2024).https://doi.org/10.1093/ehjci/jeae310
- Novedades en la endocarditis infecciosa, Revista Española de Cardiología, 77, 9, (779-787), (2024).https://doi.org/10.1016/j.recesp.2024.03.011
- New developments in infective endocarditis, Revista Española de Cardiología (English Edition), 77, 9, (779-787), (2024).https://doi.org/10.1016/j.rec.2024.03.016
- See more
Loading...
View Options
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginPurchase Options
Purchase this article to access the full text.
eLetters(0)
eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.
Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.