ORIGINAL ARTICLE

Role of 18F-FDG PET/CT in the diagnosis of infective endocarditisin patients with an implanted cardiac device: a prospective study

Maddalena Graziosi & Cristina Nanni & Massimiliano Lorenzini & Igor Diemberger &

Rachele Bonfiglioli & Ferdinando Pasquale & Matteo Ziacchi & Mauro Biffi &Cristian Martignani & Michele Bartoletti & Fabio Tumietto & Giuseppe Boriani &Pier Luigi Viale & Stefano Fanti & Claudio Rapezzi

Received: 27 December 2013 /Accepted: 28 March 2014# Springer-Verlag Berlin Heidelberg 2014

AbstractPurpose Infective endocarditis (IE) is widely underdiagnosedor diagnosed after a major delay. The diagnosis is currentlybased on the modified DUKE criteria, where the only validat-ed imaging technique is echocardiography, and remains chal-lenging especially in patients with an implantable cardiacdevice. The aim of this study was to assess the incrementaldiagnostic role of 18F-FDG PET/CT in patients with an im-planted cardiac device and suspected IE.Methods We prospectively analysed 27 consecutive patientswith an implantable device evaluated for suspected device-related IE between January 2011 and June 2013. The diag-nostic probability of IE was defined at presentation accordingto the modified DUKE criteria. PET/CT was performed assoon as possible following the clinical suspicion of IE. Pa-tients then underwent medical or surgical treatment based onthe overall clinical evaluation. During follow-up, we consid-ered: lead cultures in patients who underwent extraction,direct inspection and lead cultures in those who underwentsurgery, and a clinical/instrumental reevaluation after at least6 months in patients who received antimicrobial treatment or

had an alternative diagnosis and were not treated for IE. Afterthe follow-up period, the diagnosis was systematicallyreviewed by the multidisciplinary team using the modifiedDUKE criteria and considering the new findings.Results Among the ten patients with a positive PET/CT scan,seven received a final diagnosis of “definite IE”, one of“possible IE” and two of “IE rejected”. Among the 17 patientswith a negative PET/CT scan, four were false-negative andreceived a final diagnosis of definite IE. These patientsunderwent PET/CT after having started antibiotic therapy(≥48 h) or had a technically suboptimal examination.Conclusion In patients with a cardiac device, PET/CT in-creases the diagnostic accuracy of the modified Duke criteriafor IE, particularly in the subset of patients with possible IE inwhom it may help the clinician manage a challenging situation.

Keywords Infective endocarditis . Pacemaker . Implantablecardioverter defibrillator . Positron emission tomography

Introduction

The diagnosis of infective endocarditis (IE) in patients withcardiac leads is still challenging [1]. The reported prevalenceof cardiac device infection is up to 0.8 % for implantablecardioverter defibrillators (ICD) and ranges from 0.13 to19.9 % for pacemakers (PM) [2, 3]. Distinguishing whetherthe infection is limited to the device pocket or extends to theintravascular portion of the leads and/or cardiac valve leafletscan be extremely difficult [4]. The diagnosis of IE is currentlybased on the modified DUKE criteria [5, 6], where the onlyconsidered imaging technique, which is often inconclusive, isechocardiography [6–8]. Indeed, even transoesophageal echo-cardiography (TEE) in this context is often suboptimal and islimited by a series of elements including: lead echoes and

M. Graziosi :M. Lorenzini : I. Diemberger : F. Pasquale :M. Ziacchi :M. Biffi :C. Martignani :G. Boriani :C. Rapezzi (*)Cardiology Unit, Department of Experimental, Diagnostic andSpecialty Medicine, Alma Mater-University of Bologna, and S.Orsola-Malpighi Hospital, Via Massarenti 9, 40138 Bologna, Italye-mail: claudio.rapezzi@unibo.it

C. Nanni :R. Bonfiglioli : S. FantiNuclearMedicine Unit, Department of Experimental, Diagnostic andSpecialty Medicine, Alma Mater-University of Bologna, and S.Orsola-Malpighi Hospital, Bologna, Italy

M. Bartoletti : F. Tumietto : P. L. VialeInfectious Diseases Unit, Department of Medical and SurgicalSciences, Alma Mater-University of Bologna, and S.Orsola-Malpighi Hospital, Bologna, Italy

Eur J Nucl Med Mol ImagingDOI 10.1007/s00259-014-2773-z

reverberations (especially in patients with more than onelead), atypical location of vegetations, and difficult distinctionbetween thrombi and vegetations [9–11]. This can result ininappropriate lead extraction or significant delays in treatment[12, 13]. In recent years, 18F-FDG PET/CT has been emergingas a promising diagnostic tool for IE and its complications[14–20]. FDG is a nonspecific tracer of increased glucosemetabolism developed for the detection of malignancy andinflammation that has been recently evaluated for detection ofIE on electronic devices in a limited number of retrospectivestudies [21–24]. The aim of this study was to assess theincremental diagnostic role of PET/CT in patients with animplanted cardiac device and suspected IE using a prospectivestudy design that included diagnostic reevaluation afterfollow-up.

Materials and methods

All patients admitted to our hospital between January 2011and June 2013 with a suspicion of device-related IE whocompleted a diagnostic work-up including clinical evaluation,blood cultures, echocardiography and PET/CT were includedin the study. Exclusion criteria were: patients who could notundergo PET/CT for logistical reasons or who had a criticalclinical condition.

The diagnosis/suspicion of IE was formulated at presenta-tion by an expert multidisciplinary team that included a car-diologist, infectious diseases specialist and a nuclear medicinespecialist according to modified DUKE criteria. Diagnosiswas qualified as “definite IE” in the presence of either twomajor criteria or one major and three minor criteria or fiveminor criteria, as “possible IE” in the presence of either onemajor and one minor criterion or three minor criteria, and as“IE rejected” when it either did not reach this score or analternative diagnosis was made [5]. All patients underwenttransthoracic echocardiogram (TTE) at presentation followedby TEE in selected patients (ambiguous findings, poor acous-tic window or high clinical suspicion of IE despite negativeTTE). Vegetations were defined as an oscillating mass at-tached to a valve structure or an intracardiac/intravasculardevice such as electrode leads, with independent motion [6,7], but also as a non-oscillating sessile mass. At least two setsof blood cultures were drawn from all patients before initiationof antimicrobial therapy, within 72 h of admission or afterappropriate wash-out (between 3 and 5 days depending on thetype of antibiotic and renal function).

The main clinical and laboratory findings were collectedfor the whole study population. Septic embolism diagnosed byangio-CT and tricuspid valve perforation diagnosed by echo-cardiography were considered IE-related complications.

PET/CT was performed as soon as possible following theclinical suspicion of IE. All patients underwent a PET/CTscan

with a new generation hybrid scanner (GE Healthcare). CTimages were acquired at 120 kVand 80 mA, pitch 1.75, 0.8 sper tube rotation, and slice thickness 3.75 mm. PET imageswere corrected for attenuation on the basis of the CT data, anditerative reconstruction was performed. The PET scan wasperformed 45 – 60 min after the administration of a standarddose of 18F-FDG (370 MBq). The radiation absorbed by thepatient was comparable to that of a routine CT scan. Alllesions identified by PET/CT were evaluated by two nuclearmedicine physicians and PET/CTwas considered positive forIE only when increased FDG uptake was documented alongthe lead course [25, 26].

Patients underwent lead extraction according to currentrecommendations [27–29], and cultures of the intravascularsegments of extracted leads were obtained. All patientsunderwent medical/surgical treatment based on the overallclinical evaluation of the multidisciplinary team. During fol-low up, we considered: lead cultures in patients whounderwent extraction, direct inspection and lead cultures inthose who underwent surgery, and a clinical/instrumentalreevaluation after at least 6 months in patients who receivedantimicrobial treatment. A 6-month follow-up was also per-formed in patients in whom IE was rejected. After the follow-up period, diagnosis was systematically reviewed by the mul-tidisciplinary team using the modified DUKE criteria on thebasis of all the follow-up data.

Statistical analysis

Categorical variables are reported as frequency (%) and con-tinuous variables are presented as either mean±SD or median(interquartile range, IQR).

Results

Patient population at presentation

A total of 27 patients were included in our study. Ten patientscould not undergo PET/CT and were excluded: six patientswere in critical condition and could not be transported, whilefour patients were transferred to another hospital for family orlogistical reasons before they could undergo PET\CT. Thepatients' mean age was 69±13 years and 89 % were men.Six patients (22 %) had an ICD, two (7 %) had a single leadPM, ten (37 %) had a dual chamber PM, eight (30 %) had abiventricular ICD (three leads) and one (4 %) had abiventricular ICD with a left ventricular endocardial lead.Other baseline characteristics are summarized in Table 1.

Seven patients (26 %) sought medical attention for signs ofinfection/inflammation of the pocket (pain, erythema, swell-ing), 5 (19 %) for decubitus, 14 (52 %) for fever. and 1 (4 %)did not present any signs or symptoms of infection (incidental

Eur J Nucl Med Mol Imaging

lead mass found on routine echocardiography). Nine patients(33 %) had a history of IE or had previously required multiplepocket revisions for decubitus or infection, and 6 (22 %) hadpreviously undergone multiple antibiotic treatments.

The main laboratory findings are listed in Table 1. Bloodcultures were positive in ten patients (37 %) and causativepathogens are listed in Table 1.

TEE was performed in seven patients (26 %). Echocardi-ography (TTE or TEE) was completely negative in 15 patients(56 %). Findings in the remaining patients were: an oscillatingmass attached to a lead (eight patients, 30 %, mean length of11 mm), unclear lead thickening (three patients, 11 %), ven-tricular lead and tricuspid valve involvement (two patients,7 %), sessile mass attached to a lead (one patient, 4 %) andaortic valve vegetation (one patient, 4 %). According to mod-ified Duke criteria, at presentation 5 patients (19 %) wereidentified with definite IE, 10 patients (37 %) were identifiedwith possible IE and 12 patients (44 %) had a score indicatingIE rejected. IE-related complications were documented atpresentation in a one patient (tricuspid valve perforation andseptic pulmonary embolism).

Patients underwent PET/CT as soon as possible after thesuspicion of IE but in 14 (52 %) in 14 cases (52 %) antimi-crobial treatment had already been started and in 5 cases hadbeen ongoing for at least 7 days.

Follow-up and final diagnosis

The mean follow-up was 11±5 months. The device and leadswere completely extracted in 14 patients; lead cultures led tothe identification of a pathogenic bacterium in eight patients.Two patients with severe chronic heart failure underwentcardiac transplantation after IE was excluded (within 1 monthof the suspicion of IE); the explanted hearts were examinedand IE was ruled out in both patients. Two patients developedovert endocarditis reaching the criteria for definite IE accord-ing to Duke score. In two patients IE was complicated byseptic pulmonary embolism that was visible on PET/CT andwas confirmed by angio-CT after clinical expression. Onepatient died during follow-up and death was related to a severepulmonary infection.

After the follow-up period, a final diagnosis of definite IEwas established in 11 patients (41 %), IE was rejected in 15patients (55 %) and one patient (4 %) maintained a diagnosisof “possible IE”. After follow-up, 13 patients (48 %) had adifferent Duke score category (Fig. 1).

Final diagnosis was based on pathological criteria (extract-ed lead culture/extracted heart pathology) in 12 patients(44 %), on an alternative diagnosis in 4 patients (15 %), andon Duke score reevaluation including new elements thatemerged during follow-up in 9 patients (33 %). In 2 patients(8 %) the clinical and morphological profile did not changeduring follow-up, and the diagnosis did not change. Of the 20patients who only underwent TTE, 14 were negative, and ofthese, 5 were false-negative. Two of these five patients hadinvolvement of the extracardiac portion of the leads (docu-mented with PET/CT), while three had a low clinical suspi-cion of IE and a technically good TTE. One patient with aninitial diagnosis of definite IE underwent lead extraction after6 weeks of antimicrobial therapy. Lead cultures were negative

Table 1 Main characteristics of the study population at presentation

Characteristic Value

No. of patients 27

Age (years), mean±SD 69±13

Male gender, n (%) 24 (89 )

Device, n (%)

Single chamber ICD 6 (22)

Single lead PM 2 (7)

Dual chamber PM 10 (37)

Biventricular ICD 8 (30)

Biventricular ICD with left ventricular endocardial lead 1 (4)

Clinical presentation, n (%)

Pocket infection 4 (15)

Signs of local inflammation (pain, erythema, swelling) 3 (11)

Decubitus 5 (19)

Fever 14 (52)

No symptoms 1 (4)

Positive blood cultures, n (%) 10 (37)

Coagulase-negative staphylococcus 2 (20)

Methicillin-sensitive Staphylococcus aureus 3 (30)

Methicillin-resistant Staphylococcus aureus 1 (10)

Streptococcus group D 1 (10)

Enterococcus 2 (20)

Polymicrobical 1 (10)

Echocardiography (tansthoracic or transoesophageal), n (%)

Negative 15 (56)

Sessile mass 1 (4)

Oscillating mass 8 (30)

Lead thickening 3 (11)

Tricuspid valve involvementa 2 (7)

Aortic or mitral valve involvementa 1 (4)

Laboratory (mean±SD)

Haemoglobin (g/dl) 12±3

PCR (mg/dl) 2±3

Ferritin (ng/ml) 686±610

Initial Duke score, n (%)

Definite IE 5 (19)

Possible IE 10 (37)

IE rejected 12 (44)

a Additional echocardiographic findings: two patients had an oscillatingmass+tricuspid valve involvement, and one patient had an oscillatingmass+aortic valve involvement

Eur J Nucl Med Mol Imaging

and this was considered a complete response to treatment. Inone patient, an initial diagnosis of definite IE was rejectedafter a long follow-up period (24 months) during which noprolonged antibiotic therapy was undertaken. The bacteraemia(Streptococcus bovis) and the pedunculate mass found onechocardiography were interpreted as transient bacteraemiaand coexisting fibrin strand in a patient with intestinal adeno-ma. One patient presenting with possible IE who underwentantibiotic treatment for 5 weeks did not develop any clinicalsigns of IE during a 12-month follow-up and was consideredas a resolved possible IE.

18F-FDG PET/CT study

Patients underwent PET/CTa median 6 days (IQR 2 – 8 days)after admission, and 16 days (IQR 9 – 25 days) after symptomonset. Figure 2 shows the individual PET/CT results in rela-tion to the initial and final diagnoses. In ten patients PET/CTshowed abnormal FDG uptake along the lead course. In threepatients uptake was along the extracardiac portion of the lead(echocardiogram was negative in all three) while itcorresponded to the intracardiac portion in the remainingseven. Among the ten patients with a positive PET/CT scan,seven received a final diagnosis of definite IE, one of possibleIE and two of IE rejected. Among the 17 patients with anegative PET/CT scan, four were false-negative and receiveda final diagnosis of definite IE. PET/CT performed best in thegroup of patients with possible IE in which the result coincid-ed with the final diagnosis in nine of ten patients (Fig. 2). Inthree patients abnormal FDG uptake was detected in the lungsuggestive of septic pulmonary embolism and one of thesepatients also had uptake along the lead course; the others didnot. Of the seven patients presenting with infection or inflam-mation of the device pocket, six had FDG uptake

corresponding to the pocket. Sensitivity, specificity andpositive/negative predictive values of PET/CT are reportedin Table 2.

Of the four patients with a false-negative PET/CT scan,three underwent the examination after having started antibiot-ic therapy (≥48 h) and one had a “hot myocardium”. Of thetwo patients with a false-positive PET/CTscan, one had a “hotmyocardium”.

Discussion

This is the first study to explore the possible contribution ofPET/CT to the diagnosis of IE in patients with a cardiac devicewith a prospective design that included diagnosis review afterfollow-up. The study showed that PET/CT increases the over-all diagnostic accuracy of the modified Duke criteria, partic-ularly in patients with a Duke score indicating possible IE(Fig. 2).

The characteristics of our study population are comparableto those of the other published series of patients with IE and acardiac device [30, 31]. Specifically, mean age was high andthere was frequently a history of pocket infections requiring

Fig. 1 Diagnosis of infective endocarditis at presentation and afterfollow up

Fig. 2 Duke score at presentation and after follow-up according to theresult of the PET/CT study

Table 2 PET/CT performance for detection of lead infection in patientswith suspected IE

True-positive 7 of 26 patients

False-positive 2 of 26 patients

True-negative 13 of 26 patients

False-negative 4 of 26 patients

Sensitivity 63 %

Specificity 86 %

Positive predictive value 77 %

Negative predictive value 76 %

Eur J Nucl Med Mol Imaging

multiple revisions, recurring decubitus and multiple orprolonged antibiotic treatments. The prospective design ofour study and diagnosis reevaluation after follow-up allowedus to note that a diagnosis of IE based on the Duke criteria atpresentation is extremely challenging and often imprecise, asshown by the high proportion of patients with a different finaldiagnosis after follow-up. In this context, the need for furtherimaging evidence of device infection is strong and PET/CThas been found to be promising in this role. Our study showedthat this technique can play a relevant part in the diagnosis,particularly in the subset of patients with an intermediateprobability of IE (possible IE) in whom clinical and echofindings are not conclusive and who therefore often have animportant therapeutic delay.

The added value of PET/CT derives from the possibility ofexamining the extracardiac portion of leads, the potentialability to differentiate between thrombus and vegetation, thepossibility of diagnosing septic emboli even when clinicallysilent and the potential early recognition of lead involvement.Specifically, echocardiography is not able to visualize theextracardiac portion of leads and abnormal FDG uptake wasfound at this level in three patients with a final diagnosis ofdefinite IE (Fig. 3). The potential ability to distinguish be-tween thrombus and vegetation can be hypothesized on thebasis of three patients in whom a lead-adherent mass detectedechocardiographically showed no FDG uptake and who wenton to a final diagnosis of IE rejected. Our series also confirmsthat PET/CT enables the physician to diagnose septic embo-lism despite the absence of clinical manifestations, therebystrengthening the diagnosis of IE. This happened in threepatients in our series, and all had a final diagnosis of definiteIE. Finally, PET/CT showed potential for early detection ofdevice infection in a patient with positive blood cultures andpersistently no vegetations during echo follow-up (possibleIE). PET/CT showed abnormal FDG uptake along the leadcourse and, after a long follow-up during which he underwentmultiple cycles of antibiotics, the patient finally developed a

vegetation. The device was extracted and subsequent leadcultures were positive.

In line with other published series [21–23], however, ourstudy also confirmed that PET/CT has someweaknesses whenused in patients with IE. Indeed, in many patients antimicro-bial treatment is started promptly upon suspicion of IE beforePET/CT, thus reducing inflammation and therefore FDG up-take, potentially leading to a false-negative result. In fact, inthis series, antibiotic therapy had been started at least 48 hprior to PET/CT in 80% of false-negative patients. Diagnosticaccuracy is also negatively influenced by the presence of a“hot myocardium” that was present in 20 % of false-negativepatients and one of the two false-positive patients in thisseries. The optimization of background myocardial FDG up-take remains an unresolved issue since it seems to be influ-enced by patient preparation (regarding which disagreementpersists) as well as other potentially contributing factors [26,32–34].

Limitations

This is a small hypothesis-generating study that requires val-idation through larger prospective investigations. A consider-able proportion of our patients underwent PET/CT after hav-ing started antibiotic treatment, and this could have influenceddiagnostic accuracy negatively. A limited number of patientscould not undergo PET\CT due to their critical condition andwere excluded; this is a potential cause of selection bias. Eventhough current guidelines were followed, only 27 % of pa-tients underwent TEE. However, we do not think this signif-icantly influenced the diagnostic contribution of echocardiog-raphy. Indeed, TEE could have contributed to achieving acorrect diagnosis only in a limited proportion of patients(3/20) since an important cause of false-negative echocardio-grams is the involvement of the extracardiac portion of theleads.

Fig. 3 PET/CT in a patient withICD-related IE. a Chestradiograph shows a dual chamberpacemaker system. b The 18F-FDG PET/CT image showsabnormal FDG uptake along theextracardiac portion of the lead(arrow) and at the level of thedevice pocket (asterisk)

Eur J Nucl Med Mol Imaging

Conclusion

In patients with a cardiac device, PET/CT increases the diag-nostic accuracy of the modified Duke criteria for IE, particu-larly in the subset of patients with possible IE in whom it mayhelp the clinician manage a challenging situation.

Conflicts of interest None.

References

1. Baddour LM, Bettmann MA, Bolger AF, Epstein AE, Ferrieri P,Gerber MA, et al. Nonvalvular cardiovascular device-related infec-tion. Circulation. 2003;108:2015–31.

2. Cabell CH, Heidenreich PA, Chu VH, Moore CM, Stryjewski ME,Corey GR, et al. Increasing rates of cardiac device infection amongMedicare beneficiaries: 1990-1999. Am Heart J. 2004;147:582–6.

3. Greenspon AJ, Patel JD, Lau E, Ochoa JA, Frisch DR, Ho RT, et al.16-year trends in the infection burden for pacemakers and implant-able cardioverter-defibrillators in the United States 1993 to 2008. JAm Coll Cardiol. 2011;58:1001–6.

4. Baddour LM. Cardiac device infection – or not. Circulation.2010;121:1686–7.

5. Li JS, Sexton DJ, Mick N, Nettles R, Fowler Jr VG, Ryan T, et al.Proposed modifications to the Duke criteria for the diagnosis ofinfective endocarditis. Clin Infect Dis. 2000;30(4):633–8.

6. Habib G, Hoen B, Tornos P, Thuny F, Prendergast B, Vilacosta I,et al.; ESC Committee for Practice Guidelines (CPG). Guidelines onthe prevention, diagnosis, and treatment of infective endocarditis(new version 2009): the Task Force on the Prevention, Diagnosis,and Treatment of Infective Endocarditis of the European Society ofCardiology (ESC). Endorsed by the European Society of ClinicalMicrobiology and Infectious Diseases (ESCMID) and theInternational Society of Chemotherapy (ISC) for Infection andCancer. Eur Heart J. 2009;30:2369–413.

7. Habib G, Badano L, Tribouilloy C, Vilacosta I, Zamorano JL,Galderisi M, et al. Recommendations for the practice of echocardi-ography in infective endocarditis. Eur J Echocardiogr. 2010;11(2):202–19.

8. Evangelista A, Gonzalez-Alujas MT. Echocardiography in infectiveendocarditis. Heart. 2004;90:614–7.

9. Vilacosta I, Sarriai C, San Romran JA, Jimenez J, Castillo JA,Iturralde E, et al. Usefulness of transesophageal echocardiographyfor diagnosis of infected transvenous permanent pacemakers.Circulation. 1994;89:2684–7.

10. Lo R, D’Anca M, Cohen T, Kerwin T. Incidence and prognosis ofpacemaker lead-associated masses: a study of 1,569 transesophagealechocardiograms. J Invasive Cardiol. 2006;18:599–601.

11. Victor F, De Place C, Camus C, Le Breton H, Leclercq C, Pavin D,et al. Pacemaker lead infection: echocardiographic features, manage-ment, and outcome. Heart. 1999;81:82–7.

12. Viganego F, O’Donoghue S, Eldadah Z, ShahMH, Rastogi M,MazelJA, et al. Effect of early diagnosis and treatment with percutaneouslead extraction on survival in patients with cardiac device infection.Am J Cardiol. 2012;109:1466–71.

13. Sohail MR, Uslan DZ, Khan AH, Friedman PA, Hayes DL, WilsonWR, et al. Infective endocarditis complicating permanent pacemakerand implantable cardioverter-defibrillator infection. Mayo Clin Proc.2008;83(1):46–53.

14. Kostakoglu L, Hardoff R, Mirtcheva R, Goldsmith S. PET-CT fusionimaging in differentiating physiologic from pathologic FDG uptake.Radiographics. 2004;24:1411–31.

15. Bertagna F, Bisleri G,Motta F,Merli G, Cossalter E, Lucchini S, et al.Possible role of F18-FDG-PET/CT in the diagnosis of endocarditis:preliminary evidence from a review of the literature. Int J CardiovascImaging. 2012;28:1417–25.

16. Thuny F, Gaubert JY, Jacquier A, Tessonnier L, Cammilleri S, RaoultD, et al. Imaging investigation in infective endocarditis: currentapproach and prospectives. Arch Cardiovasc Dis. 2013;106:52–62.

17. Millar BC, Prandegrast BD, Alavi A, Moore JE. 18FDG-positronemission tomography (PET) has a role to play in the diagnosis andtherapy of infective endocarditis and cardiac device infection. Int JCardiol. 2013;167(5):1724–36.

18. Bonfiglioli R, Nanni C, Morigi JJ, Graziosi M, Trapani F, BartolettiM, et al. 18F-FDG PET/CT diagnosis of ubexpected extracardiacseptic embolism in patients with suspected cardiac endocarditis. Eur JNucl Med Mol Imaging. 2013;40:1190–6.

19. Van Riet J, Hill EE, Gheysens O, Dymarkowski S, Herregods MC,Herijgers P, et al. 18F-FDG PET/CT for early detection of embolismand metastatic infection in patients with infective endocarditis. Eur JNucl Med Mol Imaging. 2010;37:1189–97.

20. Ozcan C, Asmar A, Gill S, Thomassen A, Diederichsen ACP. Thevalue of FDG-PET/CT in the diagnostic work-up of extra cardiacinfectious manifestation in infectious endocarditis. Int J CardiovascImaging. 2013;29:1629–37.

21. Sarrazin JF, Philippon F, TessierM,Guimond J,Molin F, ChampagneJ, et al. Usefulness of fluorine-18 positron emission tomography/computed tomography for identification of cardiovascular implant-able electronic device infections. J Am Coll Cardiol. 2012;59:1616–25.

22. Bensimhon L, Lavergne T, Hugonne F, Mainardi JL, Latremouille C,Maunoury C, et al. Whole body 18Ffluorodeoxyglucose positronemission tomography imaging for the diagnosis of pacemaker orimplantable cardioverter defibrillator infection: a preliminary pro-spective study. Clin Microbiol Infect. 2011;17:836–44.

23. Ploux S, Riviere A, Amraoui S, Whinnett Z, Barandon L, Lafitte S,et al. Positron emission tomography in patients with suspected pacingsystem infections may play a critical role in difficult cases. HeartRhythm. 2011;8:1478–81.

24. Amraoui S, Texier-Maugein J, Bordachar P. PET scan in suspectedbut unproven pacemaker endocarditis. Arch Cardiovasc Dis.2012;105:125–6.

25. Kumar R, Basu S, Torigian D, Anand V, Zhuang H, Alavi A. Role ofmodern imaging techniques for diagnosis of infection in the era of18F-fluorodeoxyglucose positron emission tomography. ClinMicrobiol Rev. 2008;21(1):209–24.

26. Balink H, Hut E, Pol T, Flokstra FJ, Roef M. Suppression of 18F-FDG myocardial uptake using a fat-allowed, carbohydrate-restricteddiet. J Nucl Med Technol. 2011;39:185–9.

27. Sohail MR, Uslan DZ, Khan AH, Friedman PA, Hayes DL, WilsonWR, et al. Management and outcome of permanent pacemaker andimplantable cardioverter-defibrillator infections. J Am Coll Cardiol.2007;49(18):1851–9.

28. Baddour LM, Epstein AE, Erickson CC, Knight BP, Levison ME,Lockhart PB, et al. Update on cardiovascular implantable electronicdevice infections and their management: a scientific statement fromthe American Heart Association. Circulation. 2010;121(3):458–77.

29. Wilkoff BL, Love CJ, Byrd CL, Bongiorni MG, Carrillo R, CrosselyGH, et al. Transvenous lead extraction: Heart Rhythm Society expertconsensus on facilities, training, indications, and patient manage-ment: this document was endorsed by the American HeartAssociation (AHA). Heart Rhythm. 2009;6(7):1085–104.

30. Athan E, Chu V, Tattevin P, Selton-Suty C, Jones P, Naber C, et al.Clinical characteristics and outcome of infective endocarditis involv-ing implantable cardiac devices. JAMA. 2012;307(16):1727–35.

Eur J Nucl Med Mol Imaging

31. Bongiorni MG, Tascini C, Tagliaferri E, Di Cori A, Soldati E,Leonildi A, et al. Microbiology of cardiac implantable electronicdevice infection. Europace. 2012;14:1334–9.

32. Kouijzer IJE, Vos FJ, Janssen JR, van Dijk APJ, Oyen WJG,Bleeker-Rovers CP. The value of 18F-FDG PET/CT in diagnos-ing infectious endocarditis. Eur J Nucl Med Mol Imaging.2013;40(7):1102–7.

33. Treglia G, Bertagna F. Factors influencing the sensitivity of 18F-FDGPET/CT in the detection of infective endocarditis. Eur Nucl MedMolImaging. 2013;40:1112–3.

34. Kouijzer IJ, Vos FJ, Janssen MJ, van Dijk AP, Bleeker-Rovers CP,OyenWJ. Reply to comment by Treglia and Bertagna: FDG PET/CTfor detection of infectious endocarditis. Eur J Nucl Med MolImaging. 2013;40:1114–5.

Eur J Nucl Med Mol Imaging