Advances in radionuclide imagingradionuclide imaging of cardiac sarcoidosis
Kouranos V,1Wells AU,1 Sharma R,2 Underwood SR,3,4WechalekarK3 K4
1Interstitial Lung Disease Unit, RoyalBromptonHospital, London
2Cardiology Department, RoyalBromptonHospital, London
3National Heart & Lung Institute, ImperialCollegeLondon
4Nuclear Medicine Department, Royal Brompton& Harefield Hospitals, London
Address for correspondence
Dr Kshama Wechalekar
Department of Nuclear Medicine
RoyalBromptonHospital, London SW3 6NP
E-mail:
Short title
Radionuclide imaging for of cardiac sarcoidosis
Keywords
cardiac sarcoidosis, radionuclide imaging, FDG-PET, SPECT
Abstract
Introduction: Radionuclide imaging for the diagnosis and monitoring of cardiac involvement in sarcoidosis has advanced significantly in recent years.
Sources of data: This article is based on published clinical guidelines, literature review and our collective clinical experience.
Areas of agreement: Gallium-67 scintigraphy is established as part of among the diagnostic criteriafor cardiac involvement in systemic sarcoidosis and it is strongly associated with response to treatment. However, fluorine-18 2-fluoro-deoxyglucose (FDG) positron emission tomography (PET) is now preferred both for diagnosis and for assessing prognosis.
Areas of controversy: Most data are currently from small observational studies that are potentially biased.
Growing points: Quantitative imaging to assess changes in disease activity in response to treatment may lead to FDG PET having an important routine role in managing cardiac sarcoidosis.
Areas timely for developing research:Larger prospective studies are required, particularly to assess the effectiveness of radionuclide imaging in improving clinical management and outcome.
Background
Sarcoidosis is a multi-system disease of unknown aetiologyaetiology. The diagnosis is confirmed when non-caseatinggranulomata granulomata are identified in tissue biopsies, predominantly from the lung or from mediastinal lymph nodes and is supported by compatible clinical and radiological features [1]. Cardiac involvement can occur in isolation or it can precede, follow or occur alongside the involvement of other organs. All cardiac structures can be involved and involvement can present in many forms, but the commonest are conduction tissue involvement leading to brady-arrhythmia and syncope and the myocardial involvement leading to ventricular arrhythmia and heart failure [2, 3]. Variation in the location and extent of involvements accounts for the different cardiac manifestations. Myocardial granulomata granulomata and fibrosis can lead to abnormal automaticity and re-entrant tachy-arrhythmias or atrioventricular block. Widespread diffuse infiltration can cause ventricular remodeling and impairment of left ventricular (LV) systolic function. Histological and imaging studies suggest that there are three myocardial stages: granulomatous infiltration, oedema and fibrosis, and regional functional abnormalities or aneurysm formation [4].
The diagnosis of cardiac sarcoidosis is often circumstantial, such as when there are systemic manifestations of the disease in the presence of electrical or structural cardiac abnormalities. Histological diagnosis from endomyocardial biopsy is not routine because it is invasive and because of the heterogeneity of disease, which leads to poor sensitivity [5, 6]. The Japanese Ministry of Health (JMH) criteria for the diagnosis of cardiac sarcoidosis are widely accepted [2, 5], (table 1) and. tThey include gallium-67 scintigraphy and magnetic resonance imaging (MRI) using gadolinium myocardial enhancement [2]. However, only 5-10% of patients with sarcoidosis present with clinical evidence of myocardial involvement based on the JMH criteria while post mortem studies indicate that it is present in up to 50% of cases, implying that there is significant sub-clinical involvement that is not readily detected [1, 7].
Advances in radionuclide imaging of cardiac sarcoidosis have allowed the identification of inflammation at an early stage before structural changes are apparent and it can provide measurements of activity that might be helpful in monitoring the effects of therapy. Although gallium-67 scintigraphy, using both planar imaging and single photon emission computed tomography (SPECT) has traditionally been used, positron emission tomography (PET) using 18F-2-fluoro-deoxyglucose (FDG) combined with myocardial perfusion scintigraphy (MPS) is more sensitive and reduces radiation exposure to the patient [8, 9]. FDG-PET is now included as a diagnostic criterion in the most recent expert consensus statement on the diagnosis of cardiac sarcoidosis (table 2). [9].
Gallium-67 scintigraphy
Gallium-67 is a single photon gamma emitter suitable for planar imaging and for SPECT. It is injected intravenously as gallium citrate and it is a non-specific marker of inflammation, infection and rapid cell division such as in neoplasia. It is partly bound to transferrin with high affinity and partly complexed with hydroxyl ions forming gallate. Gallium-transferrin complex enters cells that express the transferrin receptor, such as macrophages and neoplastic cells. Although it is a non-specific marker of inflammation, in the presence of the typical imaging distribution of sarcoidosis it is highly specific for sarcoidosis, such as when the lambda and panda patterns are present (see fFigure 1) [10]. When cardiac sarcoidosis is suspected clinically, gallium uptake is a major diagnostic criterion and it correlates well with inflammatory findings at biopsy [2, 3, 9]. In some studies gallium scintigraphy has been sensitive for cardiac involvement with abnormal images in 80% of symptomatic patients with either atrioventricular (AV) block or ventricular tachycardia [11]. However, in other studies it has been less sensitive and in a study of patients with cardiac sarcoidosis according to JMH criteria, gallium scintigraphy was abnormal in only 3 of 12 cases [12]. Diagnostic accuracy can be improved by combining gallium SPECT with CT [13].
There is evidence that tTreatment of sarcoidosis with and clinical regression are is associated with reduced gallium activity although the technique has not been widely used to monitor therapy [11, 14]. In one study there were more frequent abnormalities in patients with ventricular tachycardia (VT) than in those without, and after steroid therapy the majority of patients with prior VT but no recurrence had reduced gallium activity [15]. Another study showed that steroids decreased myocardial gallium activity and that this was associated with reduced AV block but no reduction in ventricular ectopy [11]. The absence of gallium activity in myocardial regions with scar by thallium MPS has been assumed to indicate inactivity that is unlikely to respond to steroids but very few patients have been included in prognostic studies to assess this.
Gallium-67 scintigraphy is not without its difficulties. Uptake is seen in conditions that mimic cardiac sarcoidosis such as endocarditis, myocarditis, myocardial infarction, abscess, histiocytic lymphoma and infective and non-infective pericarditis and so clinical context is important [16-19]. The relatively high non-specific background activity leads to low contrast to noise and difficulties of localization, even when SPECT is combined with CT. There is also a relatively high radiation exposure for patients (~15mSv) and the need for imaging up to 48 hours after injection making it a protracted procedure.
Myocardial Perfusion Imaging
MPS has been used in patients with suspected cardiac sarcoidosis for many years. The longest established tracer is thallium-201, which has a main X-ray emission at 80keV, a physical half-life of 72 hours and an effective dose to the patient of 11mSv for 80MBq. Injected as thallous chloride, it is a combined tracer of myocardial viability and perfusion. Areas of reduced uptake after a resting injection indicate loss of viable myocardium. This is most commonly seen in ischaemic heart disease but it can suggest myocardial replacement by fibrosis or granulomatous tissue if ischaemic heart disease is unlikely or excluded. Differentiation between ischaemic and inflammatory scarring is difficult, but the former commonly has profound defects corresponding with coronary vascular territories.Coronary angiography remains the gold standard of excluding significant coronary artery disease. Even in the absence of ischaemic heart disease, thallium defects are not specific for sarcoidosis and they may be the result of other infiltrative disorders, inflammation or cardiomyopathy and so clinical context is important for interpretation. Invasive coronary angiography is commonly used to exclude anatomically significant coronary artery disease.
Other tracers of viability and perfusion are the technetium-99m labeled compounds such as methoxy-isobutyl-isonitrile (MIBI) and tetrofosmin. These have some disadvantages for imaging viability and perfusion but they have other advantages such as a higher gamma photon energy of 140 keV and a shorter half-life of 6 hours with a lower effective dose for the patient of 3 to 4 mSv for 400MBq.[11].
In a study of 10 patients with sarcoidosis without known cardiac involvement and six patients with presumed involvement there were more frequent MPS abnormalities in the latter group (32% and 60% respectively) and both were significantly more frequent that in control subjects (7%) [20]. Left ventricular involvement was associated with heart block and congestive heart failure and right ventricular involvement with tachyarrhythmia.
The combination of MPS and gallium SPECT is helpful but there is less experience in monitoring therapy. In a study of 25 patients with sarcoidosis and no known ischaemic heart disease, six patients had resting thallium defects [13]. Four of these underwent gallium imaging and two had myocardial uptake. After steroid treatment there was clinical and scintigraphic improvement in two patients but no improvement in the other two. In another study, myocardial gallium uptake was seen in 9 of 14 patients [13]. In seven of these, the site of gallium uptake corresponded with a MIBI defect and the gallium uptake was suppressed by steroid therapy in all. Of the five patients without myocardial gallium uptake, two had MIBI defects and both had already received steroids.
Positron Emission Tomography
Molecular basis of inflammation imaging
PET is a radionuclide imaging technique using radiopharmaceuticals that emit positrons and it has advantages compared with SPECT in that the images are of higher resolution and it is possible to quantify tissue activity. It is almost always combined with X-ray computed tomography (CT) for attenuation correction and for anatomical localization. The commonest PET radiopharmaceutical is fluorine-18 labelled FDG, which is a glucose analogue that is transferred into living cells by membrane-based glucose transporters. These are densely present in inflammatory cells such as neutrophils and macrophages and hence areas of inflammation appear with FDG activity that is dependent upon the degree of inflammatory cell infiltrate [8, 21]. In addition, the release of cytokines from active granulomata granulomata leads to increased FDG uptake in the cells of involved tissues.
The myocardium uses a variety of substrates in the fed state for energy including fatty acids, glucose and lactate. When fasted, free fatty acids become the preferred substrate and there is reduced metabolism of glucose. However, ischaemia and inflammation both lead to an increase in glucose metabolism and this is seen as myocardial FDG activity when there should be none in the fasted state. Different dietary protocols have been used including a prolonged fast (18 hours), with or without a prior zero carbohydrate diet, and some investigators have also given unfractionated heparin, which increases fatty acid metabolism [22]. This is in contrast to FDG PET in oncology patients where there is usually six to eight hours of fasting without a prior specific diet leading to generalized myocardial FDG activity that would obscure pathology.
It is desirable to combine PET with resting MPS in the assessment of cardiac sarcoidosis and the images can be interpreted alongside each other using oblique myocardial reconstructions, when MPS defects indicate myocardial scar and FDG activity indicates active inflammation (fsee Figure 2). Early studies using PET alone classified abnormal FDG activity as focal, diffuse or focal on diffuse, but the combination with MPS allows more detailed classification with stages of involvement as early (inflammation without scar), progressive (combined inflammation and scar) and fibrotic (scar without inflammation) (fsee Figure 32) [23-25]. In a recent study using this classification the extent of mismatch between inflammation and scar of more than 6% was strongly associated with active disease [26]. Figures 34 and 5-5 demonstrate other examples of the use of FDG-PET along with MPS in the diagnosis of cardiac sarcoidosis and the evaluation of the extent and severity of heart cardiac involvement.
Diagnosis of cardiac sarcoidosis
A systematic review of the diagnostic accuracy of FDGPET in cardiac sarcoidosis is promising. In 164 patients with a prevalence of 50% based on the JMH criteria FDG PET it had a sensitivity of 89% (95% CI 0.79-0.96) and a specificity of 78% (95% CI 0.68-0.86) [27]. The findings of individual studies are variable mainly because they are small and because of the heterogeneous nature of the disease. In a study from Okumura and colleagues ten of 16 patients with systemic sarcoidosis had abnormalities suggesting cardiac involvement and, of these, all had abnormal myocardial FDG activity whereas only five had abnormal gallium uptake and eight had myocardial scarring by MPS [28]. The regions with abnormal FDG activity did not necessarily correspond with MPS defects, consistent with fact that different processes are being imaged (fsee Figures 4 and 5).
It would be particularly valuable to quantify FDG activity as a marker of disease activity but there is not yet a standardized method of quantification. Some studies have described FDG activity as either normal or abnormal and have described myocardial activity as focal or diffuse [25, 28-33]. The standardized uptake value of FDG (SUV) is widely used in oncology imaging and it has also been applied to myocardial activity [34, 35]. Another study used the coefficient of variation of FDG activity in each myocardial voxel as a marker of heterogeneity and found this is be elevated in patients with cardiac sarcoidosis [25]. A further study quantified FDG activity using a parameter called cardiac metabolic activity (CMA), which was derived from SUV using a threshold for abnormality and is analogous to the measurement of total lesion glycolysis in oncologyic PET [36]. CMA was associated with lower LV ejection fraction and with adverse clinical events.
Despite its accuracy, PET is not recommended for screening patients with systemic sarcoidosis in the absence of symptoms, ECG or echocardiographic abnormalities suggestive of cardiac involvement because of the lack of evidence that such screening will influence management and outcome. However, in the presence of such abnormalities PET is now recommended as a diagnostic criterion [8]. The available imaging modalities used for diagnosis and monitoring of cardiac sarcoidosis are presented in table 3 and an example of their use in a patient with biopsy proven sarcoidosis is shown in fFigure 6. Subsequently the use of our The diagnostic clinical algorithm used in our own institution is presented in ffFigure 7.
Comparison of PET with other imaging modalities
FDG PET is now favored over gallium SPECT for imaging systemic sarcoidosis because it is more sensitive (97% compared with 88% in a recent study), it has greater spatial resolution, it is quicker and it exposes patients to less radiation [37]. For detecting cardiac involvement the increase in sensitivity is greater. One comparative study of 22 patients showed a sensitivity of 100% for PET compared with 36% for gallium SPECT [28]. A larger study of 76 patients showed sensitivity of 85% for PET compared with only 15% for gallium SPECT although SPECT sensitivity was 64% when combined with MPS SPECT [37]. In another study abnormalities on PET were not detected by gallium SPECT [38].
Echocardiographic findings of abnormal regional wall endocardial motion abnormalities, wall myocardial thinning and/or thickening and other morphological abnormalities can be non-specific but such findings in patients with known extra-cardiac sarcoidosis help suggest cardiac involvement. Early disease may manifest as shows diastolic dysfunction before systolic impairment. Thinning of the interventricular septum, regional wall motion abnormalities in a non-coronary distribution or focal intra-cardiac mass (large granuloma) can be seen. Impaired longitudinal strain and strain rate have also been suggested as clues for early diagnosis abnormalities but . There is however no single feature is specific to diagnose cardiac sarcoidosis. [39].
Magnetic resonance imaging (MRI) has also been used for the assessment of cardiac sarcoidosis. It has the advantage of not involving ionizing radiation but it cannot be used in patients with implanted pacemakers or defibrillators unless they are MRI compatible, and care is required in the use of gadolinium contrast in patients with impaired kidney function. Specifically gadonium It is usually avoided in patients with GFR <30ml/min and is used with caution in patients who have a if GFR is between 30 and 60 mlL/min. [3940]. Devices are commonly required in patients with cardiac involvement and so this is an important limitation. PET also has the advantage of readily assessing imaging the whole body distribution of disease, which is more difficult with MRI. Late gadolinium enhancement is very sensitive for the detection of myocardial fibrosis but it is less well validated in the detection of active inflammation. STIR and T2-weighted images can show the oedema associated with inflammation but not at an early stage and PET can be considered to be is more sensitive for early disease.
Ohira and colleagues studieds 21 patients and found that PET was more sensitive for detecting cardiac involvement (88% vs 75%) but it was less specific (39% vs 77%) [401]. In practice, MRI and PET are both valuable because they show related aspects of the disease. MRI is an excellent initial technique for identifying the presence and extent of myocardial fibrosis and for excluding other pathologies, whereas FDG PET is an important adjunct for assessing disease activity and whole body distribution (figure fFigures 3 and 5). However, it should be recognized that using MRI without PET risks may increase the risk of failing to detect active myocardial involvement until it results in fibrotic changes. T2- weighted cardiac MRI is images are more related to inflammatory activity shows the oedema associated with inflammation but it is less their sensitive sensitivity remains low compared to than PET , and while a degree of inflammation in areas of fibrosis that would lie among fibrotic areas is not easily detected. [412]. After a diagnosis and management plan have been achieved, PET is also more useful for monitoring changes in activity and treatment responseto treatment (figure Ffigure 76). [33].
Role of PET in assessing prognosis of in cardiac sarcoidosis