18F-FDG-PET/CT angiography in the diagnosis of infective ...

International Journal of Cardiology 248 (2017) 396–402

Contents lists available at ScienceDirect

International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard

18

F-FDG-PET/CT angiography in the diagnosis of infective endocarditis and cardiac device infection in adult patients with congenital heart disease and prosthetic material

María N. Pizzi a,i,⁎, L. Dos-Subirà a,i,k,l, Albert Roque b,h,i, Nuria Fernández-Hidalgo c,i,j,k,l,m, Hug Cuéllar-Calabria b,h,i, Antonia Pijuan Domènech a,k, María T. Gonzàlez-Alujas a, M.T. Subirana-Domènech i,k, B. Miranda-Barrio a,k, Ignacio Ferreira-González a,i,j, Juan J. González-López f,g,h,i, Albert Igual d, Olga Maisterra-Santos g, David García-Dorado a,i,l, Joan Castell-Conesa e,h,i, Benito Almirante c,i,j,k,l,m, Manuel Escobar Amores b,h, Pilar Tornos a,i, Santiago Aguadé-Bruix e,i a

Department of Cardiology, Hospital Universitari Vall d'Hebron, Barcelona, Spain Department of Radiology, Hospital Universitari Vall d'Hebron, Barcelona, Spain c Department of Infectious Diseases, Hospital Universitari Vall d'Hebron, Barcelona, Spain d Department of Cardiac Surgery, Hospital Universitari Vall d'Hebron, Barcelona, Spain e Department of Nuclear Medicine, Hospital Universitari Vall d'Hebron, Barcelona, Spain f Department of Microbiology, Hospital Universitari Vall d'Hebron, Barcelona, Spain g Department of Neurology, Hospital Universitari Vall d'Hebron, Barcelona, Spain h IDI (Institut de Diagnòstic per la Imatge), Spain i Universitat Autònoma de Barcelona, Spain j CIBER de Epidemiología y Salud Pública (CIBERESP), Spain k Integrated Adult Congenital Cardiac Unit of Vall d'Hebron and Sant Pau University Hospitals, Spain l CIBERCV, Spain m REIPI, Spain b

a r t i c l e

i n f o

Article history: Received 3 May 2017 Received in revised form 21 July 2017 Accepted 4 August 2017 Available online 9 August 2017 Keywords: Congenital cardiac disease Infective endocarditis 18 F-FDG-PET/CT Prosthetic material Cardiac computed tomography Adult

a b s t r a c t Objectives: Infective endocarditis (IE) and cardiac device infection (CDI) are a major complication in the growing number of patients with congenital heart disease (CHD) reaching adulthood. We aimed to evaluate the added value of 18F-FDG-PET/CT angiography (PET/CTA) in the diagnosis of IE-CDI in adults with CHD and intravascular or intracardiac prosthetic material, in whom echocardiography (ECHO) and modified Duke Criteria (DC) have limitations because of the patients' complex anatomy. Methods: A prospective study was conducted in a referral center with multidisciplinary IE and CHD Units. PET/ CTA and ECHO findings were compared in consecutive adult (≥18 years) patients with CHD who have prosthetic material and suspected IE-CDI. The initial diagnosis using the DC and the diagnosis with the additional PET/CTA data (DC + PET/CTA) were compared with the final diagnostic consensus established by an expert team at three months. Results: Between November-2012 and April-2017, 25 patients (15 men; median age 40 years) were included. Cases were initially classified as definite in 8 (32%), possible in 14 (56%) and rejected in 3 (12%). DC + PET/CTA allowed reclassification of 12/14 (86%) cases initially identified as possible IE. The sensitivity, specificity, PPV, NPV, and accuracy of DC at IE suspicion were 39.1%/83.3%/90.4%/25.5%/61.2%, respectively. The diagnostic performance increased significantly with addition of PET/CTA data: 87%/83.3%/95.4%/61.5%/85.1%, respectively. PET/CTA also provided an alternative diagnosis in 3 patients with rejected IE, and detected pulmonary embolisms in 3 patients. Conclusions: PET/CTA was a useful diagnostic tool in the complex group of adult patients with CHD who have cardiac or intravascular prosthetic material and suspected IE or CDI, providing added diagnostic value to the modified DC (increased sensitivity) and improving case classification. © 2017 Elsevier B.V. All rights reserved.

⁎ Corresponding author at: Department of Cardiology, Hospital Universitari Vall d'Hebron, Vall d'Hebron Institut de Recerca (VHIR), Universitat Autònoma de Barcelona, Passeig Vall d'Hebron 119-129, 08035 Barcelona, Spain. E-mail address: [email protected] (M.N. Pizzi).

http://dx.doi.org/10.1016/j.ijcard.2017.08.008 0167-5273/© 2017 Elsevier B.V. All rights reserved.

M.N. Pizzi et al. / International Journal of Cardiology 248 (2017) 396–402

1. Introduction A growing number of patients with congenital heart disease (CHD) are reaching adulthood, and infective endocarditis (IE) and cardiac device infection (CDI) are major complications in this group, with an IE-related risk higher than in the general population [1–3]. In addition, patients with CHD have a complex anatomy, and their surgical treatment often requires implantation of a large amount of prosthetic material. These factors make them a special IE population with clear predisposing factors and a different epidemiological profile and outcome than that of the overall IE population [4–5], especially when valve-containing prosthetics are present [6]. Use of the modified Duke criteria (DC) [7], which include echocardiography (ECHO) findings, is limited for diagnosing these patients, as the lesions visualized on ECHO are difficult to interpret. Hence, many cases of suspected IE are left without a conclusive diagnosis [8]. It would be of great value to have more accurate diagnostic imaging tools for use in this scenario. Positron emission computed tomography with 18F-fluordeoxyglucose (18 − F-FDG-PET/CT) associated with electrocardiogram (ECG)-gated cardiac CT-angiography (PET/CTA) combines the highest sensitivity to detect infection with the highest spatial resolution to define structural damage. Recent studies have reported a significant improvement in the diagnostic yield of the DC with the use of these techniques [9–11]. Therefore, the latest European guidelines have incorporated PET/CTA findings as major criterion in the diagnostic algorithms for IE [12]. The aim of this study was to evaluate the added value of PET/CTA in the diagnosis of IE and CDI in the specific population of adult patients with CHD and intravascular or intracardiac prosthetic material. To our knowledge, this is the first prospective evaluation of the utility of PET/ CTA in this group of complex patients.

397

of the region of interest; and 3) ECG-gated cardiac CTA (all patients in the cohort had creatinine level b2 mg/dL and were able to perform breath-holding). Metabolic images were fused and analyzed with CT images. Data were interpreted by an experienced team in cardiac imaging. In addition to the cardiac images, the whole-body PET/CT images were carefully examined to detect embolic events (a minor criterion), findings indicating diagnoses other than IE, and incidental neoplastic lesions. Studies were classified as positive or negative for active infection. The final diagnosis of IE-CDI (gold standard) was established by the IE Unit in possession of the complete clinical, microbiological, and imaging information and after a minimum follow-up of 3 months [9–10]. Patients were managed according to current local and international clinical guidelines [12] and were evaluated in the outpatient clinic at least at 30 days, 90 days, and 1 year after completing antimicrobial therapy. To preserve the assumption of independence of observations, only the first episode of IE recorded for an individual patient was included in the analysis. 2.1. Statistical analysis Continuous data are presented as the median and 25th–75th percentiles. Categorical data are expressed as percentages. PET/CTA results were analyzed and compared with the ECHO findings (kappa value). The reasons for discordant findings and the differences in the number of anatomical findings between ECHO and PET/CTA were recorded. The prosthesis-related SUVmax and SUVratio were compared between definite, possible, and rejected cases. We determined the performance of the DC at the time IE-CDI was suspected and the DC with addition of PET/ CTA information (DC + PET/CTA) for diagnosing IE (sensitivity, specificity, positive predictive value [PPV], negative predictive value [NPV], and accuracy by the area under the curve [AUC] with 95% confidence intervals [95% CI]). The results were compared with the final diagnostic consensus of the IE Unit, considered the gold standard in this study. Based on Bayes' theorem and according to data from our setting, an 80% prevalence of IE was used for the calculations. All tests were two-sided, and significance was set at a p value of b0.05. Statistical analyses were performed using the Stata/IC12.1 program (StataCorp, Lakeway, USA).

3. Results 3.1. Clinical and microbiological data

2. Methods A prospective study investigating the usefulness of PET/CTA in adult (≥18 years) patients with CHD who have intravascular or cardiac prosthetic material and suspected IE-CDI was conducted at a tertiary referral hospital with an adult CHD unit and a multidisciplinary IE unit including cardiologists, infectious diseases physicians, cardiac imaging specialists, heart surgeons, neurologists, and microbiologists. The study protocol was approved by the hospital ethics committee, and written informed consent was obtained from all the participating patients. Infective endocarditis was suspected based on the presence of at least one of the following criteria: 1) persistent or recurrent temperature N37 °C; 2) positive blood cultures or serology for microorganisms consistent with IE, and 3) suggestive findings on ECHO. All relevant clinical data related to the patient's heart disease history, including the details of cardiac surgery and data on the onset of symptoms, were collected when IE was suspected. The Charlson comorbidity index was used to stratify overall comorbidity [13]. The complexity of the patients' CHD was graded according to the Bethesda criteria [14], and patients were grouped based on their major underlying heart defect. Prosthetic material was defined as any intravascular or intracardiac implanted foreign material (biological or non-biological) in permanent contact with the bloodstream (e.g., valves, grafts, patches, pacemaker electrodes). Residual lesions included residual septal defects and moderate to severe valvular lesions. Infective cases were classified as healthcare-associated or community-acquired [15], and as early or late [16], according to previous publications. When the patient's condition allowed it, at least 4 sets of blood cultures (each containing one aerobic and one anaerobic bottle) were carried out on 2 consecutive days. If cultures were negative, serological tests were performed (Bartonella spp., Brucella spp., Coxiella burnetii, Chlamydophila spp. and Legionella). In all patients undergoing surgery, explanted tissues were cultured, and direct detection and identification of bacteria was carried out by 16S rRNA gene amplification and sequencing. Echocardiography was performed in all patients at the time IE-CDI was suspected and interpreted by experienced cardiologists, focusing on valve function, vegetations, and periannular complications (abscess, pseudoaneurysm, perivalvular leak), in accordance with international guidelines [17]. In patients with an initial doubtful or negative ECHO, a second study was performed one week later. All patients underwent whole-body PET/CT and cardiac PET/CTA. The method used has been fully described and validated previously [9]. Briefly, myocardial suppression was encouraged by a fasting period of at least 14 h and intravenous administration of heparin 15 min prior to 18F-FDG injection. After the patient had rested for 60 min, images were acquired using a SIEMENS Biograph mCT 64S PET/CT scanner, corrected for attenuation, and reconstructed using the iterative TrueX + TOF (ultraHD-PET) algorithm. The sequential acquisition protocol used was: 1) whole-body PET/CT at 2 min acquisition time per bed position; 2) localized 8-minute ECG-gated cardiac bed to improve evaluation

Between November 2012 and April 2017, 25 consecutive adult CHD patients who have prosthetic material and suspected IE-CDI were included (Supplemental Material-Table 1: Individual Patient Descriptions). Clinical and microbiological data are summarized in Table 1. Twentythree (92%) patients had undergone surgical repair and 2 (8%) had not been treated surgically but had a cardiac device (automatic implantable defibrillator/resynchronizer). Suspected IE-CDI were healthcare-associated in 12 patients (48%) and community-acquired in 13 (52%), early in 9 (36%) and late in 16 (64%), left-sided in 10 (40%) and right-sided in 15 (60%, 5 devicerelated). Blood cultures were positive in 20 patients (80%) and coagulase-negative Staphylococcus (CoNS) was the most common microorganism isolated in the cohort (10/25, 40%). Surgery was undertaken in 9 (36%) patients, 2 of whom (8%) died during the intervention. Another patient died due to IE-related neurological complications. Thus, the overall mortality rate was 12%. In the surviving patients (n = 22), median follow-up since the PET-CTA examination was 18.1 months (IQR 8.33–22.45). 3.2. PET/CTA and ECHO findings The median interval between ECHO and PET/CTA was 2 (IQR 1–5) days. Imaging findings are described in Table 2 and illustrated in Fig. 1. All patients underwent a transesophageal echo (TOE) except for 3 patients with a prosthetic pulmonary valve (right-sided cardiac structures) that were considered to be better assessable by transthoracic echo (TTE). Myocardial suppression was achieved in 19/25 (76%) patients. The calculated median effective radiation dose of PET/CT was 16.9 mSv (IQR: 13.8–19.7). ECHO was positive in 9 patients, negative in 11, and doubtful in 5 others. PET/CTA was positive in 19 patients and negative in 6. PET/CTA and ECHO diagnoses were concordant in only 12/25 (48%) patients (kappa: 0.2, CI: 0.07–0.4). In patients with discordant results between the 2 imaging techniques (13/25), PET/CTA was positive and accelerated

398


Table 1 Clinical and microbiological data.

Table 2 Echocardiography and PET/CTA findings according to the IE Unit final diagnosis. n = 25

Clinical characteristics Age, years Male sex Charlson comorbidity index Bethesda criteria [14] High complexity Moderate complexity Simple defects Major underlying heart defect Tetralogy of Fallot type Aortic/Subaortic stenosis Transposition type Simple septal defect - Miscellanea Type of prosthetic material Prosthetic valves and conduits Pulmonary Aortic Tricuspid Mitral Patches for VSD closure Shunt grafts Intracardiac devices-electrodes Pacemakers Automatic implantable defibrillator/resynchronizer Miscellanea‡ Time from last surgery/device implantation-to PET/CTA, years Symptoms Temperature N37° Local signs of infection (for external devices) Time from onset of symptoms to PET/CT, days Microbiology Positive microbiological findings at completion of follow-up, n (%) Coagulase negative Staphylococcusa Viridans-group Streptococcusb Propionibacterium acnes Coxiella burnetii Staphylococcus aureus Haemophilus parainfluenzae Treatment Days on antibiotics before PET/CT, median Valve surgery or device withdrawal Positive prosthetic material culture Positive polymerase chain reaction in negative material culture

40 (30–48) 15 (60) 2 (1–2) 9 (36) 7 (28) 9 (36) 8 (32) 7 (28) 5 (20) 3 (12) 2 (8) 62 21 11 7 2 1 15 6 9 5 4 11 3.6 (1.1–8.1) 21 (84) 3/9 (33) 11 (7–30) 24 (96) 10 6 3 2 2 1 5 (1–7) 9 (36) 5/9 (55) 1/9 (11)

Values are provided as the n (%) or median (IQR 25th–75th); ‡3 enlargement of patches for left ventricular outflow tract, 3 patches for right ventricular outflow tract, 2 pulmonary artery stents, 1 atrial septal defect closure device, 1 patch for pulmonary arteries enlargement, 1 tricuspid neo-chordae. VSD, ventricular septal defect; PET/CTA, 18F-FDG PET/CT angiography. a 8 Staphylococcus epidermidis, 1 Coagulase-negative Staphylococcus y 1 Staphylococcus caprae. b Including 2 Streptococcus mitis.

the IE diagnosis in 11 patients with false-negative or doubtful ECHO findings, and negative in 2 patients (1 false-positive ECHO due to a prosthetic thrombosis and 1 doubtful peritube aortic abscess). Nine patients ultimately diagnosed with definite IE by the IE Unit had an initially negative or doubtful ECHO study and a positive PET/CTA study. A repeat ECHO performed 1 week later identified lesions consistent with IE in 5 of these patients (vegetations in 4 and periaortic abscess in 1). PET/CTA enabled detection of a significantly larger number of abscesses/collections than ECHO (9 versus 3, p b 0.001), as well as 3 pseudoaneurysms that were not detected by ECHO. Semi-quantitative analysis of FDG uptake values showed significantly different SUVmax and SUVratio results in definite cases compared with possible or rejected cases (Table 2). There were 62 intravascular or intracardiac prosthetic implants in 25 patients, reflecting the large amount of prosthetic material required in CHD patients. After excluding material in close proximity to the infected elements, we evaluated the FDG uptake pattern of the remaining implants. We found that biological material had no metabolic activity, whereas non-biological material (Dacron®, Teflon®) can have some

Definite

Possible

Rejected

19 (76%)

2 (8%)

4 (16%)

0 0

1 (25%) 0

0 0

1 (25%) 1 (25%)

0 0

0 1 (25%)

NA 0

NA 1 (25%)

0.5& 0.3&

2.5 (0.5–6.23) 1.3 (0.33–4.05)

Anatomical lesions, n (%) Images consistent with vegetations Echocardiography 11 (58%) PET/CTA 8 (42%) Periannular complications Abscess/Collection Echocardiography 2 (11%) PET/CTA 8 (42%) Pseudoaneurysm Echocardiography 1 (5%) PET/CTA 3 (16%) Vascular access venous thrombosis Echocardiography NA† PET/CTA 0 Metabolic activity (PET/CTA) SUVmax‡ in prosthetic material 6.69 (5.14–9.85) SUVratio 4.41 (3.68–6.72) Peripheral findings Pulmonary embolism Suspected neoplastic lesions

3 (12%) 1 (4%)

Values are n (%), median (IQR 25th–75th); †Non-applicable; ‡Maximum standardized uptake value; &IQR not calculated due to N = 2. PET/CTA, 18F-FDG PET/CT angiography.

intrinsic FDG uptake (Fig. 1 and Supplemental Material-Table 2). The non-biological materials, which were considered not infected, regardless of some intrinsic tracer uptake, were not removed in patients who underwent surgery and no patient had a IE relapse in relation to these structures. PET/CTA also detected 3 (12%) episodes of pulmonary embolism. These occurred in 1 patient with tetralogy of Fallot repaired with a bioprosthetic pulmonary valve, who had right-sided Staphylococcus epidermidis IE, in 1 patient with a truncus arteriosus who had a bioprosthetic pulmonary valve IE due to Staphylococcus aureus, and in 1 patient with a partially repaired ventricular septal defect who had a tricuspid Streptococcus viridans IE. Furthermore, PET/CTA provided an alternative diagnosis in the 3 patients in whom IE-CDI was rejected: 1 infected pericardial collection, 1 acute axillary vein thrombosis and 1 mediastinitis. Whole-body PET/CT detected 1 unsuspected hypermetabolic thyroid nodule. Fine-needle puncture ruled out a neoplastic lesion. 3.3. Diagnostic performance of DC at IE-CDI suspicion and DC + PET/CTA Classification of cases using the DC at infectious suspicion and the consensus diagnosis by the IE Unit is shown in Fig. 2A. DC + PET/CTA findings enabled reclassification of 86% (12/14) of patients initially classified as possible IE or CDI by the DC and allowed establishing a conclusive diagnosis (definite/rejected) in 92% (23/25) of patients. The sensitivity, specificity, positive and negative predictive values, and the AUC (95% CI) of the DC at IE-CDI suspicion and DC + PET/CTA in the diagnosis of IE-CDI are reported in Fig. 2B. There was a substantial increase in diagnostic sensitivity—from 39.1% with DC at IE-CDI suspicion to 87% when the PET/CTA information was added—without affecting the specificity, mainly resulting from a significant reduction in the number of possible cases (from 56% to 8%). Although it was not the objective of this study, PET/CTA findings influenced the patients' clinical management in 8 (32%) cases: 3 surgery indications, 3 device extractions and the duration of the antibiotic therapy was changed in 2 patients. 3.4. PET/CTA follow-up studies in patients at high risk for surgery Three definite IE patients were declined for surgery due to a very high risk of death. PET/CTA control studies performed during antibiotic


399

Fig. 1. Two cases of right-sided IE confirmed by PET/CTA. A 37-year-old woman with a repaired pulmonary atresia and ventricular septal defect (VSD) (no 3). She had fever and positive blood cultures for Staphylococcus epidermidis, 1.1 years after her last surgery, which consisted on right ventricle to pulmonary artery homograft replacement. ECHO findings were inconclusive for a vegetation on the VSD patch. PET/CTA performed 3 days later showed a large (3 cm) vegetation on the pulmonary homograft (A, arrow) that was markedly hypermetabolic (SUVmax 9.93) in the fusion images, thus confirming IE (B, arrow). Another case of a 43-year-old man with tetralogy of Fallot-type CHD (no 12) who had early Staphylococcus epidermidis IE of a bioprosthetic pulmonary valve (Carpentier® Magna) 7 months after his fifth surgery. PET/CTA showed pulmonary vegetations and a periprosthetic abscess with a SUVmax of 19.28 (C, arrow) associated with septic pulmonary embolisms. He also had a right ventricular outflow tract enlargement patch of pericardium (D, asterisk), a Teflon® VSD patch (D, arrowhead), and Dacron® pulmonary artery plasties. None of them showed pathological FDG uptake or anatomical abnormalities.

therapy showed significant reductions in the metabolic activity, and stability or improvements in the anatomic lesions (periannular complications and vegetations). These imaging findings reinforced the conservative treatment decision and helped to establish the duration of antibiotic therapy. The first patient, a woman (patient no 3), had Propionibacterium acnes mechanical prosthetic aortic valve endocarditis complicated with a periprosthetic abscess and pseudoaneurysm, diagnosed in the first PET/CTA. She was treated with ceftriaxone for 16 weeks. Antibiotics were stopped due to toxicity and patient's favorable clinical evolution. At that time, a second PET/CTA showed a significant decrease in the perivalvular soft tissue thickening and a 50% reduction in the metabolic activity (SUVmax 6.47 vs 3.86 and SUVratio 3.66 vs 2.18). A third PET/ CTA performed 1 month later showed a greater reduction in the metabolic activity (SUVmax 1.86 and SUV ratio 1.05) and soft tissue thickening, as well as a decrease in the size of the pseudoaneurysm. Control blood cultures at 1, 2, 4, and 6 months after completing antibiotics were negative and the patient has remained asymptomatic up to the time of writing (23 months of follow-up). The second patient was a man (patient no 5) with Staphylococcus epidermidis bioprosthetic pulmonary valve IE complicated with a perivalvular abscess and possible VSD patch infection. He received 8 weeks of antibiotic treatment (vancomycin, gentamicin, and rifampin), with a good clinical outcome. Follow-up PET/CTA after completing antibiotic therapy showed reductions in the size of the vegetations and hypermetabolism of the valve and VSD patch: SUVmax 8.05 vs 5.14 and SUVratio 4.49 vs 2.26. Control blood cultures at 1, 2, and 3 months

after completing antibiotics were negative, and the patient has remained asymptomatic (17 months of follow-up). The third patient (no 16) had Propionibacterium acnes bioprosthetic pulmonary valve IE and possible VSD patch infection. The findings and clinical evolution of this patient are reported in the Supplemental Fig. 1. 4. Discussion In our population of CHD patients with intravascular or intracardiac prosthetic material with suspected IE and cardiac device infection, we found a considerable improvement in the diagnostic yield when the DC and PET/CTA were used in combination. Diagnostic sensitivity was particularly enhanced, increasing from 39.1% to 89%, mainly because of improvements in IE-CDI case classification. There was a significant reduction in the number of possible IE-CDI cases (from 56% to 8%), which were mainly reclassified as definite IE-CDI based on the additional PET/CTA information. A diagnosis of possible IE-CDI always poses a problem for therapeutic decision-making and can lead to a delay in starting appropriate treatment. By its ability to reclassify most possible IE cases, PET/CTA may have a major impact on the therapeutic strategies used and clinical outcomes in this population. The results of this study indicate that substantial benefits can be obtained by including PET/ CTA in the diagnostic workup of patients with CHD and prosthetic material in whom IE-CDI was suspected. Prosthetic material implanted during repair and palliations constitute additional IE targets. In this sense, a recently published study have shown that valve-containing prosthetics is the most important

400


Fig. 2. IE-CDI case classification according to the modified Duke Criteria at IE-CDI suspicion and the IE Unit consensus diagnosis at the end of follow-up (A). Diagnostic performance of the DC at IE-CDI suspicion and DC + PET/CTA (B). Definite (D). Possible (P). Rejected (R). Possible cases at admission (56%) were significantly reduced (6%). Addition of the PET/CTA findings provided a more conclusive diagnosis (definite/rejected) in 95% of cases. There was a significant increase in sensitivity (from 39% to 89%) with addition of the PET/CTA information. *Confidence interval; †Area under the curve.

risk factor for IE in patients with CHD and intravascular or intracardiac prosthetic material [6]. In agreement with this finding, 15 out of the 25 (60%) patients included in our series had a prosthetic valve or conduit-related IE (8 aortic valves, 6 pulmonary valves and 2 pulmonary tubes). Conversely and in spite of being considered a group of lower risk of IE [6], 4 (16%) patients of our group had a device-related IE. In the complex clinical scenario of CHD patients with prosthetic material, the sensitivity and specificity of the modified DC for the diagnosis of IE-CDI are greatly reduced, mainly because interpretation of the ECHO images is particularly challenging [4]. The diagnostic sensitivity of the DC was much lower in CHD patients than in our overall series of patients with IE-CDI (39.1% vs 52%) [9]. In this sense and, although ECHO must always the first image step before other advanced image studies, these data underscore the need to apply additional diagnostic tools in this population, in cases of doubtful clinical or echocardiographic diagnosis. Because of its high capacity for detecting inflammatory-infectious activity, 18F-FDG-PET/CT has shown considerable value in the diagnosis of cardiovascular infections [18]. Recent studies have reported promising results in prosthetic valve [9,11,19] and intracardiac device endocarditis [9,20–22]. Furthermore, the precise anatomical definition of cardiac CTA enables accurate detection of structural damage [9]. In our study, addition of PET/CTA to the initial DC allowed reclassification of 86% (12/14) of cases initially classified as possible IE-CDI, with a considerable increase in diagnostic sensitivity. The diagnostic accuracy of the DC in our overall IE-CDI population was 73% [9] whereas in our CHD patients it was 61.2%. Thus, the added diagnostic value of PET/CTA may have an even greater impact in this special population. Furthermore, PET/CTA enabled identification of a larger number of periannular complications than ECHO, which highlights the difficulty

of ECHO evaluation and the benefits of PET/CTA in these patients. Moreover, in patients with an indication for surgery, CTA provided useful preoperative information regarding coronary artery disease. In accordance with the findings of others [23], PET/CTA provided added value by detecting pulmonary embolisms (12%) in patients with rightsided IE and offered an alternative diagnosis in the 3 patients with rejected IE-CDI. Although these PET/CTA results are definitely encouraging, some false-positive and false-negative diagnoses occur with this technique. In our series of CHD patients, a false-negative diagnosis was given to a patient with a severely calcified Contegra conduit from the right ventricle to the pulmonary artery. False-negative diagnoses may be attributable to small lesions below the metabolic/spatial resolution of PET/CT, or to previous antibiotic therapy. False-positive diagnoses can occur because of postoperative inflammation when PET/CTA is performed too soon after surgery, particularly when a large amount of adhesive material is used [24], or in patients with severe, acute prosthetic thrombosis. In the specific CHD population, who receive a large amount of prosthetic material throughout their long surgical history, some of this material may retain intrinsic metabolic activity over the years [25]. In our series, biological prosthetic materials showed no metabolic activity unless they were in direct contact with the infected area. Although these findings should be interpreted with caution, they seem to suggest that biological prosthetic materials showing hypermetabolism on PET/CTA are highly suspicious of being infected, whereas interpretation of the same metabolic activity in an area of synthetic material is more challenging. Some CHD patients with IE are at high surgical risk because of their complex anatomy and number of previous sternotomies. The reduction


in metabolic activity and stability or improvement of the anatomic lesions observed on PET/CTA control studies reinforced the conservative treatment and helped to establish the duration of antibiotic therapy in selected patients. All had a good clinical evolution and were considered infection-free over follow-up. Although control PET/CTA studies are not a formal clinical indication, we believe that it can be of value to support clinical decisions in very complex scenarios, such as patients with highrisk CHD and IE. A final consideration has to do with the fact that PET/CTA implies radiation exposure, particularly important in the CHD population because many are young patients who may need to undergo multiple studies throughout their life. Cardiac magnetic resonance (CMR) could offer the advantage of a non-ionizing imaging technique. There is some limited data reporting its use in the diagnoses of IE [26,27]. PET/ MR technology could also be a promising option to reduce radiation dose. However, at present, PET/MR is not ready to be included in the diagnostic work-up of patients with suspicion of IE-CDI due to several unsolved aspects: very low availability of the equipments, the lower spatial resolution of MR in comparison with current CT scans, and the PET/MR susceptibility to artifacts in patients with prosthetic materials, a frequent situation in the specific congenital heart disease population. In addition, PET/MR would not be advisable in patients with permanent pacemakers or implantable cardioverter defibrillators (ICDs). 4.1. Limitations The use of a local strategy for the IE-CDI diagnosis as the gold standard in this study may be considered an important limitation. However, our IE Unit has a long clinical experience and the team based their judgment on the extensive information available from all sources at the end of the follow-up. Although this gold standard may be subject to criticism, no better alternative was available, since not all patients underwent surgery and cultures of the excised prosthetic material can be negative after a period of antibiotic treatment. The IE Unit was aware of the PET/CT results, since they provided extremely relevant information for patient management including the surgical indication in peri-tubular abscesses, septic embolisms that switched the diagnosis from possible to definite IE-CDI, and alternative diagnoses that resulted in rejection of low-probability cases. Besides, in the absence of a better gold standard option in the diagnosis of IE-CDI, the use of the IE Unit as the gold standard is completely in agreement with the current clinical guidelines [12]. Interpretation of the images in patients with CHD can be extraordinarily challenging due to the complexity of the anatomy and the presence of a large amount of prosthetic material, some of which may show persistent intrinsic metabolic activity. In our series, the combination of CTA and metabolic images was useful to overcome this limitation. However, each case was individually analyzed by a cardiac imaging team with considerable experience in CHD and this may constitute a limitation of the study. Such expertise is essential when assessing these patients, and for this reason, our results may not be extrapolable to centers with less experience. In this sense, a multicentre study is warranted. As previously mentioned, the radiation exposure derived from PET/ CTA examination is not negligible and should be addressed with radiation dose reduction measures. However, the potential advantages of PET/CTA overcome this drawback in seriously ill patients in whom a prompt diagnosis may have a critical impact on their prognosis. Finally, as this was a diagnostic accuracy study investigating the performance of PET/CTA in patients with CHD and IE, the clinical impact of the findings obtained with these techniques on the related clinical decisions and mortality requires further investigation. 5. Conclusions 18 F-FDG-PET/CTA proved to be a useful diagnostic tool in patients with suspected IE or CDI and CHD who have prosthetic materials in

401

whom the DC have a particularly low sensitivity for diagnosing IE-CDI. The additional information provided by this technique increased the diagnostic sensitivity of the modified DC from 39.1% to 87%, and resulted in a conclusive diagnosis in 92% of cases. Thus, it seems reasonable to encourage its use in this population in cases of doubtful clinical or echocardiographic diagnosis. This significant diagnostic improvement, which could have an important impact on outcomes, needs further assessment in larger series and should never replace clinical judgment. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijcard.2017.08.008. Funding sources This study was partially funded by the Integrated Excellence Project PIE-0013 and supported by Plan Nacional de I + D + i 2013–2016 and networks for cooperative research on cardiovascular (RECAVA RD16/0042/0021) and infectious diseases (REIPI RD16/0016/0003) of Instituto de Salud Carlos III, and co-financed by the European Development Regional Fund (FEDER) 2014–2020, “A way to achieve Europe”, Operative program Intelligent Growth 2014–2020. Disclosures None declared. References [1] P. Engelfriet, E. Boersma, E. Oechslin, et al., The spectrum of adult congenital heart disease in Europe: morbidity and mortality in a 5 year follow-up period - the Euro Heart Survey on adult congenital heart disease, Eur. Heart J. 26 (2005) 2325–2333. [2] J.H. Moller, R.C. Anderson, 1,000 consecutive children with a cardiac malformation with 26- to 37-year follow-up, Am. J. Cardiol. 70 (1992) 661–667. [3] K. Niwa, M. Nakazawa, S. Tateno, et al., Infective endocarditis in congenital heart disease: Japanese national collaboration study, Heart 91 (2005) 795–800. [4] J. Loureiro-Amigo, N. Fernández-Hidalgo, A. Pijuan-Domènech, et al., Infective endocarditis in adult patients with congenital heart disease. Experience from a reference centre, Enferm. Infecc. Microbiol. Clin. 34 (10) (2016 Dec) 626–632. [5] N. Fernández-Hidalgo, B. Almirante, P. Tornos, et al., Immediate and long-term outcome of left-sided infective endocarditis. A 12-year prospective study from a contemporary cohort in a referral hospital, Clin. Microbiol. Infect. 18 (2012) 522–530. [6] J.M. Kuijpers, D.R. Koolbergen, M. Groenink, et al., Incidence, risk factors, and predictors of infective endocarditis in adult congenital heart disease: focus on the use of prosthetic material, Eur. Heart J. (2017) http://dx.doi.org/10.1093/eurheartj/ ehw591. [7] J.S. Li, D.J. Sexton, N. Mick, et al., Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis, Clin. Infect. Dis. 30 (2000) 633–638. [8] L. Saiman, A. Prince, W.M. Gersony, Pediatric infective endocarditis in the modern era, J. Pediatr. 122 (6) (1993 Jun) 847–853. [9] M.N. Pizzi, A. Roque, N. Fernández-Hidalgo, et al., Improving the diagnosis of infective endocarditis in prosthetic valves and intracardiac devices with 18F-FDGPET/CT-Angiography: initial results at an infective endocarditis referral center, Circulation 132 (2015) 1113–1126. [10] L. Saby, O. Laas, G. Habib, et al., Positron emission tomography/computed tomography for diagnosis of prosthetic valve endocarditis increased valvular 18Ffluorodeoxyglucose uptake as a novel major criterion, J. Am. Coll. Cardiol. 61 (2013) 2374–2382. [11] G.M. Feuchtner, P. Stolzmann, W. Dichtl, et al., Multislice computed tomography in infective endocarditis: comparison with transesophageal echocardiography and intraoperative findings, J. Am. Coll. Cardiol. 53 (2009) 436–444. [12] G. Habib, P. Lancellotti, M.J. Antunes, et al., 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) and the European Association of Nuclear Medicine (EANM), Eur. Heart J. 36 (2015) (3075-23). [13] M.E. Charlson, P. Pompei, K.L. Ales, et al., A new method of classifying prognostic comorbidity in longitudinal studies: development and validation, J. Chronic Dis. 40 (1987) 373–383. [14] C.A. Warnes, R. Liberthon, G.K. Danielson, et al., Task force 1: the changing profile of congenital heart disease in adult life, J. Am. Coll. Cardiol. 96 (2001) 1170–1175. [15] N. Fernández-Hidalgo, B. Almirante, P. Tornos, et al., Contemporary epidemiology and prognosis of health care–associated infective endocarditis, Clin. Infect. Dis. 47 (2008) 1287–1297. [16] J. López, A. Revilla, I. Vilacosta, et al., Definition, clinical profile, microbiological spectrum, and prognostic factors of early-onset prosthetic valve endocarditis, Eur. Heart J. 28 (2007) 760–765. [17] G. Habib, L. Badano, C. Tribouilloy, et al., Recommendations for the practice of echocardiography in infective endocarditis, Eur. Heart J. 11 (2010) 202–219.

402


[18] F. Jamar, J. Buscombe, A. Chiti, et al., EANM/SNMMI guideline for 18F-FDG use in inflammation and infection, J. Nucl. Med. 54 (2013) 647–658. [19] F. Rouzet, R. Chequer, K. Benali, et al., Respective performance of 18F-FDG PET and radiolabeled leukocyte scintigraphy for the diagnosis of prosthetic valve endocarditis, J. Nucl. Med. 55 (2014) 1980–1985. [20] L. Bensimhon, T. Lavergne, F. Hugonnet, et al., Whole body [18F]fluorodeoxyglucose positron emission tomography imaging for the diagnosis of pacemaker or implantable cardioverter defibrillator infection: a preliminary prospective study, Clin. Microbiol. Infect. 17 (2011) 836–844. [21] J.F. Sarrazin, F. Philippon, M. Tessier, et al., Usefulness of fluorine-18 positron emission tomography/computed tomography for identification of cardiovascular implantable electronic device infections, J. Am. Coll. Cardiol. 59 (2012) 1616–1625. [22] M. Graziosi, C. Nanni, M. Lorenzini, et al., Role of 18F-FDG PET/CT in the diagnosis of infective endocarditis in patients with an implanted cardiac device: a prospective study, Eur. J. Nucl. Med. Mol. Imaging 41 (2014) 1617–1623.

[23] A. Asmar, C. Özcan, A.C. Diederichsen, et al., Clinical impact of 18F-FDG-PET/CT in the extracardiac work-up of patients with infective endocarditis, Eur. Heart J. Cardiovasc. Imaging 15 (2014) 1013–1019. [24] L.R. Schouten, H.J. Verberne, B.J. Bouma, et al., Surgical glue for repair of the aortic root as a possible explanation for increased F-18 FDG uptake, J. Nucl. Cardiol. 15 (2008) 146–147. [25] Z. Keidar, N. Pirmisashvili, M. Leiderman, S. Nitecki, O. Israel, 18F-FDG uptake in noninfected prosthetic vascular grafts: incidence, patterns, and changes over time, J. Nucl. Med. 55 (2014) 392–395. [26] I. Vilacosta, J. Gomez, Complementary role of MRI in infectious endocarditis, Ecocradiography 12 (1995) 673–676. [27] M. Dursun, S. Yilmaz, E. Yilmaz, et al., The utility of cardiac MRI in diagnosis of infective endocarditis: preliminary results, Diagn. Interv. Radiol. 21 (2015) 28–33.