Introduction

Infective endocarditis (IE) including native valve endocarditis (NVE) as well as intracardiac prosthetic material-related IE is a serious condition affecting about 3–10 per 100,000 persons per year [1, 2]. Despite major advances in diagnostic and therapeutic procedures, the prognosis is poor with a 1-year mortality approaching 30% and high complication rates at long term [3]. Thus, there is a need for improved early diagnosis and optimal management of IE which still remains challenging in daily clinical practice.

Recent guidelines recommend a multidisciplinary Endocarditis Team comprising cardiologists, infectious disease specialists, microbiologists, cardiac surgeons, radiologists, and nuclear medicine physicians to cope with the complexity of the disease [4]. The current standard of assessment is the revised Duke criteria. These criteria comprise the detection of the causative pathogen and the characterization of endocardial involvement by means of echocardiography as the most important factors for the diagnosis of IE [5]. However, echocardiographic findings often remain ambiguous particularly in the early phase of the disease. In addition, diagnosis is often difficult when foreign material such as cardiac implantable electronic devices or prosthetic valves is present. Whereas echocardiography in general has a specificity > 90%, the sensitivity of transthoracic echocardiography (TTE) for the diagnosis of vegetations ranges from 20 to 65% and can be improved to up to 90% using transesophageal echocardiography (TEE) [6]. Although the overall level of diagnostic performance is high, there are a number of pitfalls in the analysis of vegetations leading to false positive or negative findings. As a result, other imaging techniques such as positron emission tomography (PET) with the glucose analogue [18F]fluoro-2-deoxy-d-glucose (18F-FDG) have recently been proposed in guidelines to improve diagnostic accuracy in cases of suspected IE and diagnostic difficulties [4]. Recent data have underlined the usefulness of 18F-FDG-PET/CT in diagnosing and monitoring infectious conditions [7]. However, even this methodology exhibits several limitations and randomized controlled trials are lacking.

The goal of our endocarditis cohort study was to identify the specific indications to perform 18F-FDG-PET/CT and to evaluate the diagnostic relevance of this technique.

Materials and methods

Patients

This is a retrospective analysis of prospectively collected data. In our single-center cohort study, all patients were included that were standardly evaluated by a multidisciplinary Endocarditis Team at the University Hospital of Cologne, Germany between July 2013 and December 2016 for definite or possible native valve (NV), prosthetic valve (PV), or cardiac implantable electronic device (CIED)-related endocarditis according to the revised Duke criteria [5].

Clinical history and findings from detailed physical examination as well as results of echocardiography, blood cultures, and serum markers of inflammation (leukocytes, C-reactive protein, and erythrocytes sedimentation rate) were collected.

TTE was used as the primary cardiac imaging technique in all patients. Additional TEE was performed when (a) initial TTE was positive, (b) there was high clinical suspicion of endocarditis despite negative TTE, (c) a poor quality of TTE, or (d) a suspected prosthetic valve endocarditis and/or intracardiac devices. TEE was standardized and interpreted by an experienced cardiologist. It was performed in all except one patient who had esophageal stenosis.

All patients received a bedside consultation of an experienced infectious disease consultant and were entered in the Endocarditis Registry of the University Hospital of Cologne in case of possible or definite infective endocarditis according to the revised Duke criteria, which were applied as the gold standard in our study. Further imaging with 18F-FDG-PET/CT was undertaken when requested by at least one member of the multidisciplinary team for diagnostic clarification. When more than one 18F-FDG-PET/CT or TEE per patient was performed, only the first diagnostic imaging result was used for further evaluation in this study.

The study was approved by the Ethics Committee of the medical faculty at the University Hospital of Cologne (vote 14–221). Since diagnosis and treatment correspond to the quality standard of current guidelines, no written informed consent of the patients was considered necessary. However, patients were asked for written informed consent to the 18F-FDG-PET/CT examination. This study was registered at the U.S. National Library of Medicine [NCT02388893, Endocarditis Registry of the University Hospital of Cologne (ER-UHC)].

18F-FDG-PET/CT

18F-FDG-PET and combined low-dose CT scans were obtained using a Siemens Biograph until end of 2014 and since then a Siemens Biograph mCT Flow 128 Edge (Siemens Medical Solutions, Erlangen, Germany). All patients were asked to fast for an extended period of at least 12 h before scanning. During the study period, a diet with a meal rich in fat and low in carbohydrates in the days prior to the exam was not yet routinely implemented and also heparin was not part of the routine protocol. Blood glucose levels were checked before 18F-FDG injection. Approximately 60 min after injection of 350 MBq18F-FDG, image acquisition in 3D-mode was commenced. All emission data were corrected for attenuation, randoms, scatter, and decay. Attenuation correction was performed using an unenhanced low-dose CT scan (120 kV, mA modulation, pitch 1.2, and slice thickness 5.0 mm). Reconstruction was conducted with an ordered subset expectation maximization (OSEM) algorithm with 4 iterations and 12 subsets and Gauss-filtered to a transaxial resolution of 5 mm at full-width at half-maximum (FWHM). Images were viewed on a Siemens workstation, permitting simultaneous viewing in all three planes with easy cross-referencing between planes.

All cases were reviewed by an experienced nuclear medicine physician. Visual analysis determined whether the examination was positive when focal areas of increased uptake of 18F-FDG in the valve area were seen and the uptake was confirmed on non-attenuation corrected images.

Results

A total of 235 patients with possible (n = 43) or definite (n = 192) IE were prospectively registered during the study period. In 29 cases, 18F-FDG-PET/CT was performed as an additional diagnostic imaging method, and 20 patients were eligible for further evaluation within this study (see Fig. 1).

Fig. 1
figure 1

Diagram of patient selection

Full size image

Patient characteristics are summarized in Table 1. Of the 20 included patients, 12 individuals were classified as definite and eight as possible IE according to the revised Duke criteria. 15 out of 20 patients had a history of valve or aortic replacement/reconstruction or implantation of intracardiac devices.

Table 1 Baseline characteristics of the study population
Full size table

Table 2 gives a detailed overview of the major and minor criteria according to the revised Duke criteria together with the indications and results for the 18F-FDG-PET/CT images. Indications for additional nuclear medicine imaging were diagnostic uncertainty due to artificial heart valves (n = 9), followed by overall inconclusive clinical assessment (n = 5) and search for other foci than cardiac (n = 3) or a combination of these factors (n = 3). Eventually, one single case of vertebral osteomyelitis, knee empyema as well as upper and lower leg abscesses were detectable as other foci than cardiac. In the subgroup of patients with definite IE, 18F-FDG-PET/CT was considered helpful as an additional tool especially due to diagnostic uncertainties in TEE (n = 5, artificial heart valve; n = 1, assist device; n = 2, degenerative heart valve) and the suspicion of other infectious foci (n = 2) or inconclusive clinical assessment (n = 2). With regard to the latter, one patient suffered from recurrent bacteremia without any signs of IE in repeated TEE and had a history of periprothetic hip infection. The second patient suffered from recurrent bacteremia and back pain. On this background, uncertainties remained regarding discreet abnormalities in TEE.

Table 2 Revised major and minor criteria according to Duke and 18F-FDG-PET/CT indications/results of the study population
Full size table

18F-FDG-PET/CT images showed cardiac uptake in only 6 out of 12 patients classified as definite IE and in one out of 8 patients classified as possible IE. Thus, PET/CT scan was confirming concerns of the multidisciplinary Endocarditis Team in 35% (7/20) of all cases. Figure 2 shows the overlap of positive PET/CT images with fulfilled major criteria of echocardiography and microbiology. The positive PET/CT results (n = 7) were exclusively observed when blood culture results fulfilled major criteria according to the revised Duke criteria either with (n = 5) or without (n = 2) positive echocardiographic findings. In four cases, other infectious foci could be detected with either positive (n = 1) or negative (n = 3) 18F-FDG-PET/CT results regarding IE. In the other 13 patients, nuclear imaging was not helpful for further clarification of the IE diagnosis.

Fig. 2
figure 2

Diagnostic overlap in definite and possible endocarditis according to the revised Duke criteria

Full size image

More importantly, in 5 out of 20 cases (25%), 18F-FDG-PET/CT results helped us to modify the management of patients by confirmation of IE diagnosis (n = 1), identification of knee empyema (n = 1), rejection of IE (n = 1), confirmation of IE diagnosis with identification of vertebral osteomyelitis (n = 1) and by diagnosing a drive-line associated infection (n = 1). This led to a prolongation of antibiotic therapy in three patients, shortening in one patient and surgical intervention in two patients.

Discussion

In this endocarditis cohort study, in only a minority of 20 out of 235 patients, 18F-FDG-PET/CT was performed. With help of 18F-FDG-PET/CT, one patient with possible IE diagnosis could be reclassified to definite IE, and in six patients with definite IE, the diagnosis could be confirmed against the background of the previous diagnostic uncertainties. Furthermore, 18F-FDG-PET/CT led to a modification in the management of five patients.

In recent years, the use of 18F-FDG-PET/CT has proven helpful in diagnosing inflammatory and infectious diseases and a few reports have even shown promising results in the field of IE [8]. Both the American Heart Association and the European Society of Cardiology propose the usefulness of 18F-FDG-PET/CT to reduce the number of misdiagnosed IE classified in the possible IE category of the revised Duke criteria and to visualize peripheral embolic events [4, 9]. However, to the present day, there is no specific or formal recommendation in the current guidelines for routine use in daily clinical practice. Especially, the AHA statement emphasizes that more studies are needed to determine the role of 18F-FDG-PET/CT in the management of patients with IE [9]. In our cohort, particularly, when positive results in TEE or of blood cultures were reported, but overall diagnosis was inconclusive, 18F-FDG-PET/CT contributed to the clarification of the diagnosis of IE and had the potential to detect extracardiac manifestations of the disease.

The high number of patients (12/20) who were additionally tested with 18F-FDG-PET/CT despite definite IE according to the revised Duke criteria reflects the challenge to diagnose IE [10]. One of the causative factors is that the aforementioned criteria have a sensitivity of approximately 80% when they are evaluated at the end of patient follow-up in epidemiological studies, but show a notably lower diagnostic accuracy for early diagnosis in clinical practice [4, 11, 12]. Therefore, it is reasonable to involve a multidisciplinary team including mainly cardiologists, cardiac surgeons, infectious disease specialists, nuclear scientists, and microbiologists in the management of IE which is already recommended by recent guidelines [4, 13]. In a recent observational study with 196 patients, the involvement of an interdisciplinary team led to earlier initiation of specific antibiotic therapy and was an independent predictor of 1-year survival in patients without surgery (HR 0.24, 95% CI 0.07–0.87; p = 0.03) [14]. In our cohort, the interdisciplinary decision to perform 18F-FDG-PET/CT leads to an improvement in the management of five patients (25%).

Moreover, recent guidelines emphasize the significance of 18F-FDG-PET/CT in the setting of IE related to foreign material [4]. Previous studies have notably shown the incremental diagnostic value of this nuclear imaging technique in IE especially in case of PVE and devices. For instance, an abnormal FDG uptake around a prosthetic valve as an additional criterion apart from the revised Duke criteria increased the diagnostic sensitivity from 70 to 97%, reducing the number of patients with possible IE from 56 to 32% [15]. In our cohort, 15 out of 20 patients (75%) had prosthetic valve or device-related IE and additional nuclear imaging was particularly beneficial in these patients, since one of our main indications was diagnostic insecurity due to artificial heart valves. Additional indications comprised the detection of extracardiac sites of infection, which is in line with aforementioned studies and guidelines [4, 8, 9]. Thus, in our relatively small cohort, with 18F-FDG-PET/CT, we detected further extracardiac infectious foci in three patients.

However, there are also a number of limitations to 18F-FDG-PET/CT. It is difficult to detect small vegetations (< 5 mm) below the spatial resolution of the PET/CT system. Thus, small vegetations and the inability to correct for cardiac and respiratory motion during routine acquisition as well as a high glucose level especially in diabetic patients may lead to false negative results. False positive results may occur shortly after cardiac surgery or rarely because of concomitant diseases such as primary and secondary tumors of the heart. Moreover, with regard to prosthetic valves or devices, non-specific perivalvular uptake was described years after valve replacement in the absence of infection. Intracardiac leads can be a reason for artifacts, requiring specific corrections during diagnostics [16]. In addition, “false negative” findings have been reported after prior administration of antimicrobial therapy, rather demonstrating the ability of FDG-PET/CT to monitor a desired response [17]. In recent years, various strategies such as prolonged fasting and unfractionated heparin intravenous administration have been proposed to suppress the physiological uptake of 18F-FDG throughout the heart to improve accurate diagnosis of inflammatory cardiac diseases [18,19,20]. The rationale behind these strategies is to increase plasma free fatty acid level, which result in a suppression of glucose metabolism that facilitates the detection of areas of myocardial inflammation. In our study, diet with a meal rich in fat and low in carbohydrates in the days prior to the examination was not yet routinely implemented and also heparin was not part of the routine protocol. However, we did not overserve physiological uptake of 18F-FDG in a manner that might have hampered the differentiation of pathological nuclide uptake in the heart.

Conclusion

The final diagnosis of IE often remains difficult. Especially, the growing number of patients with implanted prosthetic material is posing an ever-greater challenge in the future. Particularly, in patients with suspected PVE, the use of 18F-FDG-PET/CT has recently proven to increase diagnostic sensitivity. Our findings support the utility of this nuclear imaging technique as an adjunctive diagnostic tool especially in the evaluation of prosthetic valve-/cardiac device-related IE and for the detection of extracardiac foci. However, due to the remaining limitations of both TTE/TEE and 18F-FDG-PET/CT, the interdisciplinary clinical evaluation still represents the essential basis for the diagnostic assessment.