A range of organisms including viral, bacterial, and fungal can infect the endocardium, cardiac prosthetic devices, pericardium, and myocardium. Currently, most evidence on the use of FDG PET/CT regards its use in prosthetic cardiac components and valves. It should also be noted that echocardiography, CT, and MRI play a useful role in the investigation of potential or proven cardiac infections. A good overview of these modalities, together with helpful descriptions of the pathophysiology of cardiac infections, has been described in a recent review article [1].

8.1 Preparation for Cardiac PET/CT

Normal myocardium has variable uptake of glucose and the standard 4–6 h fast prior to a PET/CT does not consistently suppress physiological cardiac FDG uptake to allow accurate evaluation of pathological uptake. However, the combination of a high fat, low carbohydrate diet followed by prolonged fasting (e.g. 18 h) as been shown to significantly suppress physiological uptake [2,3,4]. Protocols using this type of preparation are required when assessing for cardiac infections. The use of unfractionated heparin to further supress myocardial uptake has also been described [5] and is used in some centres [6].

8.2 Native Valve Infective Endocarditis

FDG PET/CT is not typically used for the diagnosis of native valve infectious endocarditis (IE), diagnosis is based on the modified Duke criteria which includes clinical, echocardiographic, and biological findings, as well as the results of blood cultures and serology [7]. Recent European Society of Cardiology (ESC) guidelines on endocarditis [8] have maintained this approach and have also provided further modified diagnostic criteria.

The rationale for this is due to the published low sensitivity of FDG PET/CT in detecting native valve IE. A recent systematic review and meta-analysis [9] demonstrated a sensitivity of 61%, lower than the modified Duke criteria of 80% [8]. Additionally, a recent cross-sectional study found that PET/CT was false-negative in all cases of suspected native valve IE (n = 21) [6].

However, FDG PET/CT has proved to be useful in detecting septic emboli in patients with native valve IE. In a prospective cohort study of patients with definite IE, the systematic use of PET/CT in 47 patients compared to 94 controls who did not have IE showed that PET/CT diagnosed significantly more infectious complications (septic emboli) (18% vs. 57.4%, p = 0.0001) and was associated with a two-fold reduction in the number of relapses (9.6% vs. 4.2%, p = 0.25) [10]. The ESC guidelines [8] have included imaging for embolic events, which can be performed with FDG PET/CT, in the further diagnostic workup of patients who fall into the “possible” category of the modified Duke criteria.

8.3 Prosthetic Valve Endocarditis

Diagnosis of prosthetic valve IE (PVE ) is more challenging than native valve IE due to more atypical presentation and more frequently negative blood cultures and echocardiography [8]. The use of PET/CT in the evaluation of patients with possible PVE, according to Duke criteria, is now recommended in recently published ESC guidelines on IE [8]. In these guidelines, abnormal activity around the site of implantation detected by 18F-FDG PET/CT (if the prosthesis was implanted >3 months previously) is considered a major criterion. This is based on a study of FDG PET/CT in PVE which found a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 73% (95% CI: 54–87%), 80% (56–93%), 85% (64–95%), 67% (45–84%), and 76% (63–86%), respectively. Use of FDG PET/CT increased the sensitivity of the modified Duke’s criteria at admission from 70% (52–83%) to 97% (83–99%), p = 0.008 [11].

8.4 Cardiac Implantable Electronic Device Infections

Cardiac implantable electronic devices (CIED ) include pacemakers and implantable cardioverter defibrillators. When considering infections of these devices, it is important to be aware of the distinction between an infection limited to the generator pocket versus an infection extending to the cardiac leads, valves, or endocardium although differentiating between the two types of infection can be difficult.

Similarly to PVE, diagnosis can be challenging. Echocardiography plays a lead role in the diagnosis of CIED infection; however, a normal echo does not rule it out. There is increasing evidence of the usefulness of FDG PET/CT in suspected CIED infection. The recent ESC guidelines [8] state: “Radiolabelled leucocyte scintigraphy and 18F-FDG PET/CT scanning may be considered additive tools in patients with suspected cardiac device related IE, positive blood cultures and negative echocardiography”. In addition, recent evidence-based guidelines on the use of PET/CT in the United Kingdom [12] state PET/CT is indicated in the “evaluation of… cardiac implantable device related infection in selected cases provided sufficient time has elapsed since surgery”.

These recommendations are based on a number of studies. One study [13] compared three groups of patients: group A consisted of 42 patients with suspected CIED infection, group B included 12 patients without infection who underwent PET/CT 4–8 weeks post-implant and group C included 12 patients implanted for >6 months without infection who underwent PET/CT for another indication. A ratio between maximal suspicious cardiac uptake and lung parenchymal uptake (cardiacmax:lungmean) was calculated. In Group A, 32 patients had positive PET/CT, 24 of these underwent extraction with excellent correlation. In 7 patients with positive PET/CT, 6 were treated as superficial infection with clinical resolution. One patient with positive PET/CT but negative leukocyte scan was considered false positive due to a Dacron pouch. Ten patients with negative-PET/CT were treated with antibiotics and none had relapsed at 12.9 ± 1.9 months. Group B patients had mild uptake seen at the level of the connector, group C patients did not demonstrate any abnormal uptake. Median cardiacmax:lungmean was significantly higher in Group A (2.02) vs. B (1.08) vs. C (0.57); p < 0.001.

A later study used similar methods to analyse the use of FDG PET/CT to detect pocket infection in CIEDs [14]. It used a ratio of maximum count rate around the CIED to mean normal lung parenchymal count rate (pocketmax:lungmean) in patients with and without (controls) suspected CIED pocket infection, patients were managed independently of PET/CT result. Patients with suspected generator pocket infection that required CIED extraction (n = 32) had significantly higher 18F-FDG activity compared with those that did not require extraction (n = 14) or controls (n = 40). The pocketmax:lungmean was: 4.80 (3.18–7.05) vs. 1.40 (0.88–1.73) vs. 1.10 (0.98–1.40), respectively; p < 0.001. The area under the receiver operator characteristic analysis for the pocketmax:lungmean was 0.98 for those ultimately needing extraction with an optimal cut-off value of >2.0 (sensitivity 97%, specificity 98%).

A smaller study consisted of 27 patients [15] with suspected CIED infection who had an initial PET/CT and a final diagnosis made via culture in those who had leads removed or clinical/instrumental follow-up after 6 months in those who did not. 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 patients with a negative-PET/CT scan, four were false-negative and received a final diagnosis of definite IE. These patients underwent PET/CT after having started antibiotic therapy (≥48 h) or had a technically suboptimal examination.

Figure 8.1 shows an example of a non-infected CIED. Figure 8.2 shows increased uptake in an infected prosthetic aortic valve prosthesis. Figure 8.3 shows infection of a device pocket with extension to the lead.

Fig. 8.1
figure 1

No CIED infection. There is no uptake in the device pocket or lead to suggest infection. Most modern PET /CT scanners have provisions for metal artefact reduction allowing accurate assessment of metabolic activity

Full size image
Fig. 8.2
figure 2

Aortic valve endocarditis with splenic abscess. There is focal uptake in the prosthetic aortic valve suggesting infective endocarditis. There is a large photopaenic area in the spleen with a rim of metabolic activity suggesting abscess. Whole body imaging allows evaluation of ectopic source of infection

Full size image
Fig. 8.3
figure 3

Device pocket infection with extension to the lead. There is uptake in the device pocket with tracer uptake tracking along the intravascular part of the lead stopping short of the intra cardiac insertion

Full size image

8.5 Pericarditis

It is important to remember there are mulitple non-infectious causes of pericarditis , some of the most common being autoimmune and malignant (particularly lung, breast, and lymphoma) [16] and PET/CT is frequently used in investigating these pathologies. Viruses are the most common cause of pericarditis in the developed world and TB in the developing world [16].

ESC guidelines on pericarditis are not specific on the use of PET/CT stating [16]: “PET/CT can be indicated to depict the metabolic activity of pericardial disease” (p. 2942). However, the guidelines do further elaborate that it is useful in distinguishing TB from idiopathic pericarditis, this is based on a study of 15 patients which retrospectively evaluated the ability of PET/CT to distingish between the two aetiologies [17]. This found that the mean and standard deviation (SD) of pericardial thickness and SUVmax of acute tuberculous pericarditis was significantly higher than in acute idiopathic pericarditis (5.1 SD [1.0] vs. 3.4 [0.9], p < 0.05; 13.5 [3.9] vs. 3.0 [0.7], p < 0.05, respectively). The mean and SD SUVmax of mediastinal and supraclavicular lymph nodes of acute tuberculous pericarditis (5.3 [1.8]) was significantly higher than that of acute idiopathic pericarditis (2.8 [0.6], p < 0.05). There was no significant difference in the mean size of the mediastinal and supraclavicular lymph nodes between acute tuberculous and idiopathic pericarditis.

8.6 Myocarditis

A variety of infectious and systemic diseases can also cause myocarditis . Viruses, often enteroviruses, are the most important causes in developed countries [18, 19]. Electrocardiogram, myocardiocytolysis markers, and transthoracic echocardiography are the standard first-line investigations, cardiac MRI is useful to provide further structural and functional information and endomyocardial biopsy is the gold standard diagnostic test [19].

ESC guidelines state (p. 2642) [19]: “Nuclear imaging is not routinely recommended in the diagnosis of myocarditis, with the possible exception of suspected cardiac sarcoidosis” and a relatively recent review on FDG PET in cardiac infections also does not describe any specific role for PET/CT [18]

Conclusion

  • There are recommedations and evidence that FDG PET/CT is of use in evaluating patients with suspected prosthetic valve or CIED infection when initial investigations/diagnostic criteria are inconclusive.

  • cardiacmax:lungmean or pocketmax:lungmean has been shown to be a useful parameter when assessing for CIED infection.

  • FDG PET/CT should not currently be used to directly investigate for native valve IE. However, in suspected native valve endocarditis, it can be used to look for septic emboli, the knoweldge of which is useful when managing patients with suspected IE.

  • FDG PET is currently not routinely used to assess patients with pericarditis or myocarditis; however, it has been shown to help distinguish between idiopathic and TB pericarditis.

Key Points

  • There are recommendations and evidence that FDG PET/CT is of use in evaluating patients with suspected prosthetic valve or CIED infection when initial investigations/diagnostic criteria are inconclusive.

  • FDG PET/CT should not currently be used to directly investigate for native valve IE. However, in suspected native valve endocarditis, it can be used to look for septic emboli, the knowledge of which is useful when managing patients with suspected IE.

  • FDG PET is currently not routinely used to assess patients with pericarditis or myocarditis however it has been shown to help distinguish between idiopathic and TB pericarditis.