Chapter 24: Cardiac Amyloid Imaging and ¹⁸F-FDG PET/CT for Imaging Sarcoidosis and Cardiovascular Infection

Cardiac amyloidosis, sarcoidosis, and cardiovascular infections are major causes of morbidity and mortality. Early diagnosis and timely application of specific therapy are critical to improve clinical outcomes.^1,2 Echocardiography, cardiac magnetic resonance imaging (CMR), and cardiac CT can detect increased wall thickness, expanded extracellular volume, diffuse late gadolinium enhancement, wall motion abnormalities, fibrosis, vegetations, and abscesses resulting from cardiac amyloidosis, sarcoidosis, or infection. However, by the time the above cardiac structural changes are evident, the disease is at a fairly advanced stage. Moreover, some of these changes are nonspecific and may represent sequelae of other forms of heart diseases and are not disease specific. Radionuclide imaging methods are evolving as specific disease markers in amyloidosis, sarcoidosis, and infection imaging.

Introduction

Cardiac amyloidosis is a subset of systemic amyloidosis, which is caused by deposition of insoluble nonbranching protein aggregates of amyloid fibrils in the extracellular space of the myocardium. This results in left ventricular thickening and heart failure. Heart failure is a frequent cause of death in these patients. Two major forms of systemic amyloidosis have cardiac involvement^3,4: (1) Light chain amyloidosis (AL) where amyloid fibrils are formed from immunoglobulin light chains produced by clonal population of plasma cells and (2) Transthyretin amyloidosis (ATTR) where misfolded monomers or dimers from either mutant or wild type TTR deposit as fibrils in the myocardium. The mutant type (ATTRm) is an autosomal dominant disorder while the wild type disease (ATTRwt) is associated with aging (senile systemic amyloidosis). Some of the ATTRm diseases have predominant cardiomyopathic manifestations while others have coexistent or predominant neuropathy, including carpal tunnel syndrome and/or autonomic neuropathy along with cardiomyopathy. ATTRwt may be significantly underdiagnosed clinically, as a recent autopsy study showed a prevalence of up to 30% of myocardial amyloid deposits in individuals over 75 years with heart failure and preserved ejection fraction.⁵

Distinction between the two types of cardiac amyloidosis is critical as the treatment and prognosis are vastly different. For AL amyloidosis, chemotherapy is used to inhibit the plasma cell dyscrasia which results in the abnormal light-chain production. Overall mortality as well as treatment-related mortality for AL patients is markedly increased in individuals with cardiac involvement, therefore demonstrating cardiac involvement can guide treatment in these patients. AL cardiac amyloidosis can be rapidly progressive (median survival, if untreated, is <12 months after heart failure onset).³ On the other hand, median survival after heart failure onset is better for ATTR patients (median survival 75 months)⁴ and within this subset, individuals with ATTRwt do better than those with ATTRm. In recent years, new therapies for ATTR have emerged including drugs that inhibit TTR synthesis, stabilize TTR, or degrade proteins.⁶ Hence, there is an urgent need to image and quantify myocardial amyloid burden for early diagnosis and for assessment of response to therapy with these new therapies.

Imaging with echocardiography and CMR typically raises the suspicion of cardiac amyloidosis in individuals with heart failure.⁷ However, imaging features on structural imaging, though highly suggestive, may not be definitively diagnostic for cardiac amyloidosis. Importantly, no specific imaging feature on echocardiography or CMR can distinguish AL from ATTR amyloidosis.⁷ Several radiotracers have been used for imaging cardiac amyloidosis including bone imaging compounds, ¹²³I metaiodobenzylguanidine (mIBG, to image myocardial denervation in familial amyloidosis), and amyloid binding agents (¹²³I-serum amyloid p-component SAP, ¹⁸F-florbetapir, flutemetamol, florbetaben, ¹¹C-Pittsburgh B compound).⁷ In this chapter, we will focus on ^99mTc bone imaging compounds, which are widely used clinically and are highly sensitive and specific for imaging cardiac ATTR amyloidosis.

Indications

Diagnosis

Radionuclide imaging plays an important role in noninvasive diagnosis of cardiac ATTR amyloidosis. ^99mTc-pyrophosphate (PYP) and ^99mTc-3,3-diphosphono-1,2-propranodicarboxylic acid (DPD) are two radiopharmaceuticals that are used for imaging cardiac amyloidosis. In a recent multicenter study, ^99mTc-PYP has been shown to identify ATTR amyloidosis with 97% sensitivity and 100% specificity.^{8–10 99m}Tc-DPD, available in Europe but not approved for clinical use in the United States, has a specificity of 88% and a sensitivity of 100%.^10–12

Several other amyloid binding radiotracers, in particular PET agents that have been used for beta-amyloid imaging in the brain (¹¹C Pittsburgh B-compound,^{13,14 18}F-florbetapir,^{15,16 18}F-flutemetamol, and ¹⁸F-florbetaben¹⁷) are being investigated for cardiac amyloid imaging.

¹²³I-MIBG had been used for the diagnosis of early myocardial denervation in ATTRm.^18–21 mIBG imaging is discussed in Chapter 23.

Echocardiography and CMR have well-established roles in the diagnosis and follow-up of cardiac amyloidosis and a proposed algorithm incorporating radionuclide imaging is shown in Figure 24-1.

In patients with CHF and unexplained left ventricular thickening with suspected cardiac amyloidosis who either have echo findings that are classic for cardiac amyloid and no evidence of plasma cell dyscrasia or patients with suggestive but not classical echo findings and a negative cardiac MRI (or cannot get an MRI) without plasma cell dyscrasia, ^99mTcPYP/DPD scan can be performed to assess for ATTR amyloidosis. Patients with familial ATTR and suspected cardiac amyloidosis can also be evaluated with ^99mTcPYP/DPD scan. Patients with a typical echocardiographic appearance and strongly positive (grade 2 or 3 myocardial radiotracer uptake ≥ rib uptake) ^99mTcPYP/DPD scan can be considered to have ATTR amyloidosis (Fig. 24-2). In these patients, biopsy is not necessary, but may be performed for typing of amyloid. A negative scan, however, does not exclude AL amyloidosis and further evaluation is necessary.⁷ In these patients, bone marrow, fat pad, or endomyocardial biopsy can be performed to confirm the diagnosis of AL. Early studies of cardiac PET imaging for amyloidosis such as with ¹⁸F-florbetapir show potential for detection of AL cardiac amyloidosis. However, this is not yet currently in routine clinical use.

Response to Therapy

Currently, response to therapy for cardiac ATTR is mainly monitored using echocardiography and CMR. However, they are limited due to the temporal delay in detection of change in myocardial wall thickness. Extracellular volume fraction on CMR is emerging as a novel and quantitative marker to diagnose changes in cardiac amyloidosis burden. SPECT/CT with ^99mTcPYP/DPD has the potential to provide quantitative imaging for monitoring response. Results to date, with planar imaging, however, have been disappointing.²² However, it is unclear whether the therapy was not successful or the imaging technique was limited at detecting changes. Amyloid imaging using PET amyloid tracers is inherently quantitative and offers the promise to detect changes in amyloid burden following response to therapy.²³

Prognosis

For patients with ATTRm, it has been shown that the heart-to-whole body ratio (>7.5) combined with LV wall thickness (>12 mm) was associated with the highest rate of major adverse cardiovascular events (defined as cardiovascular death, hospitalization or stroke).¹² A heart-to-contralateral lung ratio of ≥1.6 on planar ^99mTc-PYP scan is associated with worse survival.⁹ Among individuals with V30M mutation ATTRm a heart to mediastinal (H/M) ratio of <1.6 on late ¹²³I-mIBG imaging was associated with a significantly worse 5-year mortality (42% vs. 7% for H/M ratio of <1.6 and ≥1.6, respectively, p < 0.001).

Imaging Technique

Protocol

^99mTc-PYP/DPD

The standard dose for ^99mTc-PYP/DPD is 20 to 25 mCi injected intravenously. No special patient preparation is required and imaging is performed at 1 to 2.5 hours after radiotracer injection. Protocol options include: (1) whole-body or planar cardiac with chest/cardiac SPECT if planar is abnormal and (2) planar and chest/cardiac SPECT or (3) chest/cardiac SPECT-only images. SPECT-only imaging is adequate and provides quickest imaging times. Images are reconstructed with standard ^99mTc protocols and displayed in standard short-axis, vertical long-axis, and horizontal long-axis views.

Image Interpretation ^99mTc-PYP

Qualitative

Images are evaluated qualitatively with regard to uptake in the myocardium when compared to bone (ribs) as follows:

• Negative/grade 0: no uptake

• Mild/grade 1: less than rib uptake

• Moderate/grade 2: equal to rib uptake

• Severe/grade 3: greater than rib uptake

Grade 2 or greater is considered positive for ATTR amyloidosis.²⁴

Quantitative Analysis

Regions of interest over the heart and in the contralateral lung can be used to assess heart-to-contralateral-lung ratio.⁶ The heart-to-contralateral-lung ratio is derived by drawing a region of interest (ROI) over the heart and copying it over to the right chest. Visual uptake in the heart is compared to visual radiotracer uptake in the ribs on planar images and preferably on SPECT images.

Other Findings

Normal physiologic distribution of ^99mTc-PYP/DPD is similar to a standard nuclear medicine bone scan with uptake in the osseous structures and excretion in the genitourinary system.

Pitfalls

A potential pitfall is myocardial blood pool uptake in early images or in individuals with renal dysfunction. Myocardial radiotracer uptake in ATTR amyloidosis is usually clear-cut with excellent contrast between myocardium and blood pool, if there is any question on early images, more delayed images can be obtained to increase the contrast between the myocardium and the blood pool.

Reporting

Report should contain the presence and degree of myocardial uptake, as well as other incidental findings on whole-body imaging and SPECT/CT, if performed. If qualitative analysis is requested, the heart-to-contralateral-lung ratio on planar imaging can be reported as well. Readers are referred to American Society of Nuclear Cardiology (ASNC) practice points on ^99mTc-PYP imaging for more details.²⁵

Conclusion/Future Directions

Radionuclide imaging of cardiac ATTR amyloidosis with ^99mTc-PYP/DPD is easy to perform with essentially no contraindications. Image analysis is relatively straightforward and nuclear imaging can provide important adjunct diagnostic and prognostic information to echocardiography and CMR. ¹²³I-mIBG imaging can be useful to identify myocardial denervation and stratify risk of adverse clinical outcomes in individuals with ATTRm. Amyloid-binding PET radiotracers that are approved for beta-amyloid brain imaging are currently under investigation and show promise, particularly for quantitation and for characterization of the AL subtype of amyloidosis.

Positron emission tomography (PET) with ¹⁸F-2-fluoro-2-deoxy glucose (¹⁸F-FDG) is emerging as a robust technique for early diagnosis of focal myocardial inflammation and infection, particularly prosthetic material infection. ¹⁸F-FDG is a cyclotron-produced glucose analog with a half-life of 110 minutes. Due to its long half-life, ¹⁸F-FDG can be shipped as unit doses to sites without a cyclotron on-site. ¹⁸F-FDG is a surrogate marker for cellular glucose uptake. Tissue hypoxia is a potent stimulus for increased glucose utilization by the cardiomyocytes. High glycolytic activity from infiltrates of active inflammatory cells also increases glucose utilization in inflammation and infection. GLUT1 synthesis and cell membrane expression of GLUT1 receptors are upregulated in activated macrophages facilitate increased glucose utilization.²⁶ In addition, circulating cytokines and growth factors are thought to increase the affinity of glucose transporters for ¹⁸F-FDG.²⁷ Once transported into the cells through GLUT1, GLUT3, and GLUT4 receptors,^{28 18}F-FDG is phosphorylated by hexokinase to ¹⁸F-FDG-6-phosphate which is not metabolized any further. The trapped ¹⁸F-FDG-6-phosphate within the cell²⁸ provides the imaging signal for metabolically active inflamed tissues. This chapter will focus on imaging of cardiac sarcoidosis and cardiovascular prosthetic infection using ¹⁸F-FDG PET. (See Chapter 3 for further review of FDG.)

¹⁸F-FDG PET/CT Imaging of Cardiac Sarcoidosis

Introduction

Sarcoidosis is an idiopathic, multisystem inflammatory disorder characterized by accumulation of noncaseating granulomas. These organized collections of macrophages, epithelioid cells, and lymphocytes are formed in response to inciting antigens and can eventually lead to organ damage. Both cardiac and neurologic sarcoidosis can occur in isolation without involvement of other organs.²⁹

Cardiac sarcoidosis can involve virtually any part of the heart from endocardium to pericardium and both ventricles and atria.³⁰ The myocardium is most frequently affected, in particular, the left ventricular free wall at the base of the heart, followed by the basal interventricular septum.³¹ Clinical manifestations range from arrhythmias to congestive heart failure to sudden death. Cardiac involvement represents the cause of death in up to 25% of fatal sarcoidosis in the United States. Early diagnosis is important as these complications are potentially preventable with early treatment. Until recently cardiac sarcoidosis was diagnosed using the criteria proposed by the Japanese Ministry of Health and Welfare, that included a histological diagnosis by endomyocardial biopsy or a clinical diagnosis by extracardiac biopsy-proven sarcoidosis in conjunction with major and minor criteria including findings on ECG, echocardiography, myocardial perfusion imaging with ²⁰¹thallium, ^99mtechnetium, ⁶⁷gallium, and CMR (Table 24-1). ¹⁸F-FDG PET/CT, though not included in the Japanese criteria, plays a major role in the contemporary diagnosis and management of cardiac sarcoidosis.³²

Indications for ¹⁸F-FDG PET/CT

Diagnosis of Cardiac Sarcoidosis

The most definitive diagnosis of cardiac sarcoidosis, currently, is based on a positive endomyocardial biopsy. Endomyocardial biopsy is usually a blind procedure with several small samples taken from the right ventricular aspect of the interventricular septum (usual location). As sarcoidosis is a focal disease, endomyocardial biopsy is more prone to sampling error. In sarcoidosis, different parts of the myocardium may harbor granulomas in different stages of inflammation, fibrosis, or an admixture of inflammation and fibrosis. As only a minute portion of the myocardium from the interventricular septum is sampled randomly, assessment of disease activity in the whole heart is limited. ¹⁸F-FDG-PET/CT overcomes these limitations. Inflammation is assessed in the entire heart. Furthermore, as shown in Table 24-2, an abnormal ¹⁸F-FDG PET/CT clinches a diagnosis of cardiac sarcoidosis, without an endomyocardial biopsy, in individuals with a histological proof of extracardiac sarcoidosis and abnormal cardiac symptoms, examination findings, ECG findings, or LVEF <50%.³³

Despite these advantages, when cardiac sarcoidosis is suspected, a CMR study is generally considered before ¹⁸F-FDG PET/CT due to its high negative predictive value to exclude sarcoidosis⁷ and the absence of ionizing radiation. ¹⁸F-FDG PET/CT may be considered as the initial test, if CMR is contraindicated (ferromagnetic devices, glomerular filtration rate, GFR <30 mL/min), or if CMR is abnormal and a confirmation of active inflammation may change clinical management or if clinical suspicion of cardiac sarcoidosis remains high despite a normal CMR.³² The sensitivity and specificity of ¹⁸F-FDG PET for the diagnosis of cardiac sarcoidosis are summarized in Table 24-3.³⁴ Figure 24-3 outlines a proposed algorithm for CMR and ¹⁸F-FDG PET/CT to diagnose cardiac sarcoidosis.

Assessment of Response to Treatment

The results of ¹⁸F-FDG PET/CT can be useful to initiate, as well as to monitor response to anti-inflammatory therapy (Fig. 24-4). In individuals with ongoing or increasing inflammation an increase in dose or addition of new medications may be considered, while in those with decreased ¹⁸F-FDG uptake on subsequent scans, medication dose may be tapered.³⁵ However, the utility of ¹⁸F-FDG PET/CT to guide changes in anti-inflammatory therapy has not been validated in prospective clinical trials.

Prognosis

Cardiac involvement in sarcoidosis carries a poor prognosis. In a study of 118 patients with suspected or known cardiac sarcoidosis, the presence of both a perfusion defect and myocardial ¹⁸F-FDG uptake had the strongest association with death or sustained ventricular tachycardia when compared to those with normal studies or perfusion defects alone or myocardial ¹⁸F-FDG uptake alone.³⁶ In addition, of the patients with PET/CT abnormalities, those with focal uptake in the right ventricle had an even higher adverse event rate. Cardiac PET/CT imaging of perfusion and metabolism, therefore, has the potential to risk stratify patients with cardiac sarcoidosis.³⁶

Imaging Technique

Protocol

Radionuclide imaging protocols for cardiac sarcoidosis comprise of a rest myocardial perfusion imaging study along with cardiac and a limited whole-body ¹⁸F-FDG imaging. Myocardial perfusion imaging can be performed with SPECT using ^99mTc-sestamibi or ^99mTc-tetrofosmin or with a PET using ¹³N-ammonia or ⁸²rubidium. Table 24-4 outlines a summary protocol for the ¹⁸F-FDG PET/CT study for cardiac sarcoidosis. Historically, ⁶⁷gallium SPECT was the mainstay for inflammation imaging and is included in the Japanese diagnostic criteria for cardiac sarcoidosis. But, due to the poor image resolution, long scan duration and limited sensitivity, in practice, this has largely been supplanted by ¹⁸F-FDG PET/CT. PET-only imaging can also be performed on dedicated PET cameras with smaller field of views, but can be challenging to localize focal hotspots of ¹⁸F-FDG. Fusion imaging with myocardial perfusion may help localize ¹⁸F-FDG uptake. As with PET/CT, images should be reconstructed and displayed as short axis, horizontal long axis, and vertical long axis to correlate with myocardial perfusion imaging.

Dietary Preparation

The myocardium is a metabolic omnivore that consumes glucose or fatty acids for its metabolic needs, depending on the substrate abundance. The goal of dietary preparation is to suppress normal myocardial glucose utilization to be able to detect abnormal signal within the myocardium. This is accomplished by shifting myocardial metabolism to fatty acid use. Several approaches exist including prolonged fasting, specific diet (a high fat, low to zero carbohydrate, protein permitted diet), and intravenous unfractionated heparin (Table 24-5).

A retrospective study³⁷ from the Netherlands compared three methods: (1) fasting alone (>6 hours), (2) low carbohydrate, protein and fat permitted diet for 12 hours followed by 12-hour fasting, and (3) low carbohydrate diet, protein and fat permitted diet for 12 hours followed by 12-hour fasting and IV unfractionated heparin 50 IU/kg, 15 minutes prior to injection of ¹⁸F-FDG. This study showed that dietary preparation (method 1) outperformed fasting alone while the addition of IV unfractionated heparin to diet and fasting (method 3) outperformed diet and fasting (method 2) in terms of suppression of myocardial glucose utilization. However, this study did not include a high fat diet. A high fat, low to zero carbohydrate, protein permitted diet for two meals on the day prior followed by an overnight fast prior to the ¹⁸F-FDG PET examination currently remains the standard protocol for ¹⁸F-FDG PET for suppressing physiological myocardial glucose utilization for sarcoid, infection/inflammation protocol ¹⁸F-FDG PET.¹⁵ Table 24-6 lists some sample foods for the sarcoid ¹⁸F-FDG PET diet.

Image Interpretation

Assessment/Diagnosis

Interpretation of the perfusion and ¹⁸F-FDG integrated study involves four steps: (1) interpretation of perfusion images, (2) interpretation of cardiac ¹⁸F-FDG PET/CT, (3) integration of perfusion and ¹⁸F-FDG PET/CT data, and (4) interpretation of limited whole-body ¹⁸F-FDG PET/CT. Images should be interpreted in conjunction with clinical data, ECG findings, and other results of imaging studies such as echocardiography and CMR. Interpretation of perfusion images is discussed separately in this book. For the purposes of cardiac sarcoidosis, perfusion imaging is approached the same way as for coronary artery disease.

Cardiac ¹⁸F-FDG PET/CT images are reviewed in the standard cardiac imaging planes of short axis (SA), vertical long axis (VLA), and horizontal long axis (HLA) as well in the standard radiological imaging planes of axial, sagittal, and coronal planes. ¹⁸F-FDG uptake in the right ventricle (RV) and atria are evaluated on the fused ¹⁸F-FDG emission and the CT transmission images. In individuals with adequate suppression of physiological myocardial glucose uptake, focally increased ¹⁸F-FDG uptake with or without corresponding perfusion defects may represent myocardial inflammation. Lack of focal myocardial ¹⁸F-FDG uptake may represent noninflamed tissue, including normal myocardium and scar. Patterns of myocardial perfusion and ¹⁸F-FDG uptake and their interpretations are summarized in Figure 24-5. In a multivariate analysis, Blankstein et al.³⁶ found that a pattern of both abnormal perfusion and abnormal FDG had the strongest association with death or VT (Fig. 24-6).

Notably, hibernating myocardium could cause a similar pattern of perfusion–metabolism mismatch. Hence, significant epicardial coronary artery disease needs to be excluded, preferably, by coronary angiography (CT or invasive) or by rest and stress imaging. Indeed, local tissue hypoxia from microvascular perfusion abnormalities related to endothelial dysfunction or microvascular compression from sarcoid granulomas may account for some of the increased myocardial ¹⁸F-FDG signal in active sarcoidosis. The limited whole-body ¹⁸F-FDG PET/CT can be a valuable adjunct by providing evidence of active extracardiac sarcoidosis (or lack thereof) and potentially identifying a biopsy site. Individuals with sarcoidosis have a twofold higher risk of malignancy³⁸; hence, the whole-body images are evaluated also for coexisting undiagnosed malignancy (Fig. 24-7). Knowledge of the normal biodistribution of ¹⁸F-FDG is essential to understand normal variants and identify pathological uptake on the whole-body images. The spectrum of extracardiac ¹⁸F-FDG PET/CT findings in sarcoidosis is beyond the scope of this text and readers are referred to an excellent review on this topic.³⁹

Assessing Response to Treatment

Due to the serious side-effect profile with long-term use of anti-inflammatory medications (steroids, methotrexate, or the novel TNF alfa inhibitors), ¹⁸F-FDG is frequently repeated several months after initiation of anti-inflammatory therapy to tailor medication dose. Interval decrease or complete resolution of ¹⁸F-FDG uptake in the myocardium, compared to baseline study, could either mean resolution of inflammation or progression to fibrosis. Comparison with myocardial perfusion imaging is very important as a concomitant improvement in perfusion would support resolution of inflammation while worsening perfusion indicates a scar.⁴⁰ It is important to recognize that the images are normalized to the hottest pixel and hence a purely visual assessment could be potentially misleading. Semi-quantitative metrics of standardized uptake value (SUV) (defined as the radioactivity activity concentration [kBq/mL] measured by the PET scanner within an ROI divided by decay-corrected amount of injected radiolabeled ¹⁸F-FDG [kBq] per unit patient weight [g]), offer the advantage of absolute quantitation of tissue uptake. That being said, semi-quantitative analyses using SUVmax or SUV mean, though widely used clinically, are influenced by a number of factors,⁴¹ and need further evaluation. With studies performed at the same institution with the same protocol, suggested methods include comparing mean (SUV mean) and maximum SUVs (SUVmax), measuring volume of inflamed tissue using an SUV threshold, and measuring target-to-background ratio comparing myocardial SUV to blood pool or cerebellum as outlined in Table 24-7.

Artifacts and Pitfalls

Several artifacts need to considered in interpreting ¹⁸F-FDG PET/CT scans for cardiac sarcoidosis. A major potential pitfall in interpretation is insufficient suppression of physiological glucose utilization by normal myocardium. Diffuse myocardial ¹⁸F-FDG uptake is more likely due to insufficient suppression of physiological uptake rather than cardiac sarcoidosis. Likewise, focal increase in ¹⁸F-FDG uptake in the lateral wall, especially with normal perfusion in the lateral wall, may represent a normal variant.³⁶ In these scenarios, one should investigate possible dietary noncompliance, remedy accordingly and repeat the study if needed. Otherwise, only data from perfusion images and any extracardiac evidence of sarcoidosis on the ¹⁸F-FDG PET/CT portion of the study can be used for interpretation. As mentioned above, there is evidence that the addition of IV unfractionated heparin to dietary preparation could result in improved suppression and the study can be redone using this protocol, if there are no contraindications to heparin.

Patient motion during image acquisition is another potential cause of artifact and care must be taken to ensure proper motion correction and registration to ensure accurate attenuation correction. Focal hotspots may pose a problem particularly in individuals with intracardiac devices, especially ICD. In those cases both the attenuation-corrected and non–attenuation-corrected images need to be reviewed or metal artifact reduction applied.

Some of these patients may be on corticosteroid therapy which could alter the biodistribution of ¹⁸F-FDG. In this case, using target-to-background ratio or ratio of SUV uptake in different organs can be helpful.

Other factors that alter biodistribution such as hyperglycemia and hyperinsulinemia should also be taken into consideration during interpretation. In general, patients with blood glucose >120–200 mg/dL based on various guidelines^42,43 should be rescheduled if possible as altered biodistribution from hyperglycemia could decrease the diagnostic specificity of ¹⁸F-FDG PET/CT.

Reporting

The report should include all relevant clinical information (including prior echocardiography, CMR, ¹⁸F-FDG PET/CT results, dose and type of anti-inflammatory medications), technique (including imaging protocol, dose of ¹⁸F-FDG and uptake period, interval between injection and scan), the quality of the study (including adequacy of suppression of physiological glucose utilization), results of perfusion (including LV size, RV size, tracer uptake, perfusion defects), gated SPECT findings (including rest LVEF, LV volumes, wall motion, RV function), ¹⁸F-FDG PET/CT cardiac and extracardiac findings as well as a final impression (presence of myocardial inflammation and concordance with perfusion defects, change from prior, LV function, evidence of extracardiac sarcoidosis).⁴⁴

Conclusions

¹⁸F-FDG PET/CT plays a critical role in the diagnosis and treatment of individuals with suspected or known cardiac sarcoidosis. Diagnosis of active inflammation and assessment of response to treatment are the two primary indications for ¹⁸F-FDG PET/CT in cardiac sarcoidosis. While ¹⁸F-FDG PET/CT is sensitive to detect myocardial inflammation, it is not specific; other noninflammatory or neoplastic processes that increase myocardial glucose utilization or even normal myocardium may exhibit increased ¹⁸F-FDG uptake. Further studies are warranted to optimize patient preparation methods to suppress physiological myocardial ¹⁸F-FDG uptake and improve the specificity of ¹⁸F-FDG PET/CT in cardiac sarcoidosis. Finally, several studies are underway to evaluate more specific radiotracers including somatostatin receptor–binding agents, as well as novel more specific radiotracers for inflammation (¹¹C-PBR28).

Introduction

Cardiovascular prosthetic material/device infection significantly increases the risk of mortality. In the setting of suspected infection in cardiovascular devices such as prosthetic valves, vascular grafts, and implantable pacemakers/defibrillators, accurate diagnosis can be difficult due to atypical clinical presentations, frequently negative blood cultures, and limitations of anatomical imaging with echocardiography or cardiac CT. Several recent studies have shown the usefulness of ¹⁸F-FDG PET/CT in detecting infections associated with devices and grafts as discussed in the recent European guidelines for the management of infective endocarditis.^45–49

Indications

Pacer/ICDs

¹⁸F-FDG PET/CT can be helpful in evaluating patients with suspected infection from an implantable pacer/defibrillator when anatomic imaging studies are uncertain or negative. In a prospective study of 66 patients, ¹⁸F-FDG SUV values were found to be helpful in differentiating between active cardiac device infection and normal postoperative inflammatory changes 1 to 2 months' post implantation (SUVmax 4.4 ± 1.6 vs. 1.2 ± 1.4, p < 0.001) with a sensitivity of 89% and a specificity of 86%.⁴⁸ In the same study, they found that inflammatory changes and residual uptake typically resolve after 6 months. Figure 24-8 shows a proposed diagnostic and treatment algorithm for suspected cardiac device infection.

Left Ventricular Assist Devices

Use of left ventricular assist devices (LVAD) has been increasing due to shortage of donor hearts. They are used both as a bridge to transplant and as destination therapy. Infectious complications in these patients are common with sepsis developing in 42% of patients after 1 year and 52% of patients after 2 years.¹⁹ All areas of the LVAD can be involved (Fig. 24-9). A recent retrospective study of 31 patients (total of 40 studies) with and without external signs of driveline infection showed 100% sensitivity and 80% specificity in detection of LVAD infection. Management was altered by ¹⁸F-FDG PET/CT findings in 85% of patients.^44,50 This study also showed decreased FDG uptake related to response to therapy.

Prosthetic Valves

Prosthetic valve endocarditis (PVE) is associated with high morbidity and mortality. The current diagnostic gold standard for infective endocarditis, the modified Duke criteria,⁵¹ has limited sensitivity in the presence of prosthetic valves.⁵² This is in part due to difficulty of echocardiographic assessment in these patients due to acoustic shadowing from the valve and difficulty in detecting small vegetations and abscesses. ¹⁸F-FDG PET/CT is particularly helpful for reclassifying patients that fall under the possible PVE category under the modified Duke criteria⁴⁷ and identifying peripheral embolic and metastatic infectious foci (except in the brain, where high physiological uptake of ¹⁸F-FDG may limit visualization of metastatic infectious foci).⁴⁵ Pizzi et al.⁵³ showed that the combination of modified Duke criteria and ¹⁸F-FDG PET with cardiac CT angiography compared to cardiac CT without contrast reclassified an additional 20% of cases of possible endocarditis.

Vascular Grafts

Infection of vascular grafts is a serious complication that occurs in 0.5% to 5% of prostheses.⁵⁴ No consensus currently exists regarding diagnosis and management but a few studies have found ¹⁸F-FDG PET/CT to be useful in diagnosing vascular prosthetic infection, particularly low-grade infections. In a prospective study of 76 patients, focal ¹⁸F-FDG uptake and irregular graft boundary were found to be significant predictors of infection.⁴⁹

Imaging Technique

Protocol and patient preparation including diet is identical to ¹⁸F-FDG PET/CT for cardiac sarcoidosis.⁴⁴ For the whole-body images, scan from head to toes is recommended to evaluate for potential secondary embolic sources of infection. Myocardial perfusion images are not necessary for imaging cardiovascular infection. Imaging with hybrid PET/CT is preferred to localize region of focal ¹⁸F-FDG uptake, and imaging with dedicated PET cameras can be challenging to localize the region of abnormal radiotracer uptake. Ideally, whole-body images should be done to also evaluate for potential secondary embolic sources of infection.

Image Interpretation

Assessment/Diagnosis

¹⁸F-FDG accumulates in inflamed and infected tissue due to increased intracellular glucose metabolism in inflammatory cells that are responding to the infection (such as leukocytes and macrophages). Infection in general manifests higher SUV values than postprocedural inflammation. However, given the variability of SUV based on different factors there is currently no established cutoff for distinguishing postprocedural inflammation from infection. In general, the further out the PET/CT is from surgery/implantation (particularly >8 weeks), the more likely that the radiotracer uptake represents true infection in the proper clinical setting.

It is very important to evaluate the non–attenuation-corrected images. Devices have relatively higher attenuation than normal soft tissue or myocardium on CT. Attenuation correction with CT tends to overestimate focal radiotracer uptake in areas of high attenuation and artifactually increase tracer uptake in these regions. Non–attenuation-corrected images or metal artifact reduction methods may be used to distinguish inflammation from artifact. In instances where high specificity is desired, radiolabeled white blood cell scanning is recommended.⁴⁵

Extracardiac Findings

As with sarcoidosis, the remainder of the scan should be evaluated for other possible embolic areas of infection or other abnormalities such as undiagnosed malignancy.

Response to Treatment

In general, treatment includes removal of infected devices followed by antibiotic therapy. In rare instances wherein the devices are not removed (e.g., aortic stent, prosthetic material in complex congenial heart disease patients), response to treatment can be evaluated by decreased ¹⁸F-FDG uptake in the area of concern (Fig. 24-10). It is recommended to evaluate both the uptake (SUVmax) as well as the target-to-background ratio to account for possible differences in biodistribution between scans.

Artifacts and Pitfalls

Apart from attenuation artifact from the device itself, false-positive findings have been reported in Dacron grafts^48,50 and bland thrombi.⁴⁷ Patients who are immunosuppressed or on long-term antibiotics can present as false negative due to relatively decreased immune cells in the region of infection.

Small and mobile infected lesions (endocarditis) may not be detected on ¹⁸F-FDG PET/CT due to limited temporal resolution; in those instances a TEE may be superior to identify infection on the leads. Factors that alter biodistribution such as hyperglycemia and hyperinsulinemia should also be taken into consideration during interpretation. In general, patients with blood glucose >120–200 mg/dL based on various guidelines^42,43 should be rescheduled if possible as altered biodistribution from hyperglycemia could decrease the specificity of ¹⁸F-FDG PET/CT. Although less of a concern, as with the sarcoid protocol, inadequate myocardial suppression can render a study uninterpretable if the area of concern is close to the myocardium.

Reporting

The report should include the indication, relevant clinical information (including date of surgery/device implantation and medications such as antibiotics or corticosteroids), technique (including type and dose of radiopharmaceuticals), the quality of the study, ¹⁸F-FDG PET/CT cardiac and extracardiac findings as well as a final impression (presence of suspicious ¹⁸F-FDG uptake, any unexpected findings).⁴⁴

Conclusions

The main strength of ¹⁸F-FDG PET/CT is its high sensitivity to detect infection, while its main limitation is its limited specificity. It is not useful in the early postoperative period (4–8 weeks) due to expected postoperative inflammation that could lead to false-positive findings on ¹⁸F-FDG PET/CT. Certain types of devices and grafts can also cause false-positive results and low-grade infections can be missed. Development of more specific radiotracers for imaging infection⁵⁵ will be useful to improve the specific diagnosis of device infection.