Because sarcoidosis is an inflammatory disease characterized by aggregation of T-lymphocytes, macrophages, and noncaseating granulomas, the metabolic activity of inflamed myocardial tissue is important in determining treatment strategy. Previously, gallium-67 imaging was used to detect myocardial inflammation, and abnormal uptake of gallium-67 is a major diagnostic criteria in both the JMHW recommendations and the HRS expert consensus.5)6) However, the diagnostic sensitivity of gallium-67 is low due to its low image resolution. Myocardial FDG-PET, which has higher sensitivity and better accuracy than gallium-67 scintigraphy,10)11)12) has been increasingly utilized to evaluate CS. FDG, a glucose analog, is useful in detecting active inflammatory lesions, as activated proinflammatory macrophages show a higher metabolic rate and glucose utilization.2) Due to its sensitivity and accuracy in diagnosing CS, as well as its usefulness in guiding CS therapy, the importance of FDG-PET in patients with CS has been further emphasized.2)12)

Although FDG-PET is highly useful for diagnosing CS, it should be noted that glucose is used as an energy source by normal cardiomyocytes as well as by inflammatory cells in CS-affected myocardium. However, the mode of action for glucose uptake differs between normal cardiomyocytes and inflammatory cells in CS-affected myocardium (Figure 1). In normal myocardium, more than 90% of fasting energy metabolism is due to fatty acids, and the remaining less than 10% of metabolism involves other substances, including glucose.13) Because the lipid bilayer of the plasma membrane is impermeable to glucose, glucose transport across the plasma membrane is mediated via glucose transporters (GLUT). GLUT4, the main isoform in cardiomyocytes, is located in intracellular membrane compartments and translocated to the cell surface in response to stimuli including increased workload, ischemia, catecholamines, and insulin.14)15) While increased blood glucose and insulin levels enhance glucose utilization in cardiomyocytes (Figure 1A), increased serum fatty acid levels during fasting inhibit glucose uptake via insulin-dependent GLUT4 (Figure 1B).16)17)18)19) In contrast, inflammatory cells, such as neutrophils, monocytes, and macrophages, express high levels of GLUT1 and GLUT3 in cell membranes (Figure 1C).20)21) Increased glucose metabolism is a hallmark of the immune cell activation involved in both innate (e.g., monocytes and macrophages) and adaptive (e.g., T and B cells) immunity.22)23)24) Because glucose metabolism of inflammatory cells is sensitive to proinflammatory cytokines and noxious signals, glucose uptake is enhanced in activated inflammatory cells even in conditions in which blood glucose and insulin levels are not increased.

Because glucose uptake mechanisms differ in cardiomyocytes and inflammatory cells, myocardial FDG-PET can be utilized in diagnosis of cardiac inflammatory diseases, such as CS.12)25)26) Of note, it is important to differentiate pathologic FDG uptake of affected myocardium from physiologic FDG uptake of normal myocardium. Numerous studies have aimed to differentiate cardiac involvement of sarcoidosis using heterogeneous myocardial uptake patterns of CS. For example, Tahara et al.27) compared FDG-PET images between 24 patients with CS and 8 age-matched patients with dilated cardiomyopathy. They found that FDG uptake is distinctly heterogeneous in patients with CS, and patients with CS had a significantly higher coefficient of variation (standard deviation divided by the mean) for myocardial FDG uptake than patients with dilated cardiomyopathy. The authors also reported that the value of the coefficient of variation in myocardial FDG uptake decreased after corticosteroid therapy, suggesting that heterogeneity in myocardial FDG uptake indicates the disease activity of CS. Sperry et al.28) retrospectively analyzed 203 patients who underwent FDG PET imaging to evaluate CS and showed that heterogeneity in FDG uptake assessed by coefficient of variation provides an incremental prognostic advantage. However, the degree of heterogeneity in FDG uptake is not intuitive and is less suitable for quantitative assessments of treatment response than other quantitative measures, such as maximal standardized uptake value (SUV_max) or volume of FDG uptake.29)30) Therefore, effective suppression of physiologic glucose uptake in normal myocardium is a key factor in the evaluation of CS using FDG-PET.

Typical FDG-PET imaging in patients with known or suspected malignancy requires at least 4 hours of fasting to ensure low serum levels of glucose and insulin, as elevated glucose or insulin is directly responsible for glucose uptake by non-tumor cells and low diagnostic yield.31)32) In patients with suspected CS, the contrast in FDG uptake between affected and normal myocardium must be distinct to allow identification of inflammatory lesions in the myocardium. Thus, prolonged fasting prior to imaging is recommended for patients with suspected CS.9)12)33) While many previous studies utilized 12 hours of fasting,10)12)27)34)35)36) more recent studies suggest that fasting for more than 18 hours further suppresses physiologic myocardial glucose uptake and improves the specificity of FDG-PET for CS. Langah et al.37) assessed the diagnostic accuracy of an 18-hour fasting protocol in 76 patients with suspected CS and reported sensitivity of 85% and specificity of 90%.37) They also noted that the FDG myocardial-to-blood pool ratio decreased more significantly after 18-hour fasting than after shorter fasting durations. However, imaging after prolonged fasting remains limited due to variation in suppression of glucose utilization by normal cardiomyocytes (Figure 2).

The use of unfractionated heparin enhances plasma lipolytic activity, and increased lipoprotein lipase and hepatic lipase activity leads to increased plasma free fatty acids.38)39) The consequent elevation in serum free fatty acid inhibits glucose uptake via GLUT4.17) Based on these findings, a heparin pre-administration protocol has been suggested for diagnosing CS by FDG-PET: unfractionated heparin is administered 15 minutes before FDG administration at a dose of 50 IU/kg.35)36) However, it should also be noted that heparin pre-administration effectively suppresses physiologic myocardial FDG uptake in only some patients, and the relationship between heparin dose and degree of physiologic myocardial FDG uptake has not been established.12) Although pre-administration of unfractionated heparin prior to FDG-PET imaging is theoretically appropriate, further investigations are required to establish an appropriate FDG-PET protocol for the evaluation of CS.

In addition to prolonged fasting and pre-administration of unfractionated heparin, dietary modifications have been suggested to suppress physiologic myocardial FDG uptake. As FDG uptake in normal myocardium is dependent on serum levels of glucose and free fatty acid, a low-carbohydrate and high-fat diet prior to FDG-PET imaging would theoretically improve diagnostic yield. Consistent with this, Lum et al.40) reported that clinically significant myocardial FDG uptake on FDG-PET was more commonly observed in patients who consumed carbohydrates compared to those who did not consume carbohydrates prior to scanning. A small randomized trial by Cheng et al.41) further clarified the effect of dietary modifications: compared to patients with an unrestricted diet, a low-carbohydrate diet resulted in significant suppression of physiologic myocardial FDG uptake. Although no well-established study has reported improved diagnostic accuracy with a low-carbohydrate diet, many institutions advocate carbohydrate content of less than 5 g per meal prior to scanning.41)42)43)

Furthermore, adding a high-fat diet to a low-carbohydrate diet 3-6 hours prior to FDG-PET scanning has been suggested to significantly reduce FDG uptake in the entire myocardium.43)44) Consistently, a prospective study by Harisankar et al.45) demonstrated consistent and significant suppression of myocardial FDG uptake in most patients who maintained a low-carbohydrate high-fat protein-permitted diet the night before and 4 hours prior to scanning. However, in a small randomized trial conducted by Cheng et al.41), the addition of a high-fat diet 1 hour prior to FDG injection did not result in greater reduction of physiologic myocardial FDG uptake compared with a low-carbohydrate diet alone. In a similar randomized trial by Demeure et al.46), volunteers who consumed a low-carbohydrate high-fat diet followed by 12 hours of fasting demonstrated effective suppression of physiologic myocardial FDG uptake, whereas those who consumed additional free fatty acids 1 hour prior to FDG injection failed to show additional suppression of myocardial FDG uptake over that achieved in those who did not undergo this additional preparation. Considering these findings, a low-carbohydrate high-fat diet followed by prolonged fasting may be expected to effectively suppress the physiologic glucose uptake of normal myocardium and improve specificity for diagnosing CS (Figure 2).

Although the clinical usefulness of FDG-PET for patients with known or suspected CS is well-known, a detailed FDG-PET protocol has not been established. The various protocols utilized by previous studies differ from each other in terms of composition and calorie content of the dietary preparation, duration of fasting, and dosage and timing of heparin pre-administration. For example, the Japanese Society of Nuclear Cardiology Recommendations suggested that patients have a low-carbohydrate diet the night before FDG-PET (< 5 g) followed by at least 12 hours of fasting and a high-fat diet 3-6 hours before the FDG injection, but also recognized that 18 hours of fasting would improve the specificity of CS diagnosis (Figure 3A).12) These recommendations did not include heparin pre-administration. The Society of Nuclear Medicine and Molecular Imaging (SNMMI) and American Society of Nuclear Cardiology (ASNC) expert consensus statement of 2017 recommended starting a low-carbohydrate (< 3 g/meal), high-fat (> 35 g/meal) diet about 24 hours prior to scanning followed by 4–12 hours of fasting (> 18 hours for patients who cannot follow the dietary recommendations) (Figure 3B).33)47) Additionally, the Joint SNMMI-ASNC expert consensus statement suggested the use of unfractionated heparin (50 IU/kg) 15 minutes prior to FDG administration.

Our institution has adopted a protocol with a low-carbohydrate, high-fat diet followed by at least 18 hours of fasting in addition to intravenous pre-administration of unfractionated heparin prior to FDG injection (Figure 3C). Specifically, hospitalized patients receive a low-carbohydrate (< 5 g/meal), high-fat (> 35 g/meal) lunch and dinner developed by specialists from the nutritional support team (Figure 4A). For those in outpatient clinics, a diet modification manual explaining food types and caloric content is provided, and patients are requested to submit a detailed diet diary (Figure 4B). For patient convenience, myocardial FDG-PET scans are usually scheduled around 1 PM for hospitalized patients. Patients undergo 18 hours of fasting starting at 6 PM immediately after consuming a low-carbohydrate high-fat dinner. Subsequently, patients are given two 1,000 IU doses of unfractionated heparin at 30 minutes and 15 minutes before FDG administration. Although there is limited head-to-head evidence comparing protocols, it is assumed that stricter patient preparation results in better diagnostic performance of myocardial FDG-PET scanning. Indeed, better specificity is easily observed in real-world practice in patients who consume a low-carbohydrate high-fat diet prior to scanning (Figure 2).

Even with the establishment of a standardized patient preparation protocol, it is not easy to ensure complete suppression of physiologic FDG uptake by the normal myocardium. Because contrast in FDG uptake between normal versus inflamed myocardium is critical for diagnosing CS, it is helpful to monitor the degree of myocardial FDG uptake suppression. Suppression of physiologic FDG uptake in normal myocardium is dependent on increasing free fatty acid level. A randomized trial by Demeure et al.46) compared several protocols for the optimization of FDG-PET, and serum free fatty acid levels were higher in volunteers with better suppression of physiologic FDG uptake.

There was also an inverse correlation between serum free fatty acid level and SUV_max, suggesting that serum free fatty acid level predicts the degree of myocardial FDG uptake suppression. Furthermore, our group assessed the relationship between serum free fatty acid level and heart SUV_max in 147 patients who underwent FDG-PET to evaluate malignancy (Figure 5). Because FDG-PET scanning was performed according to the oncology protocol, patients were only required to fast for at least 4 hours without dietary modification or heparin pre-administration. As shown in Figure 5A, the heart and brain demonstrated higher SUV_max values compared to the liver or muscle. Although the heart demonstrated the widest variability in SUV_max values, serum free fatty acid level at the time of FDG-PET scanning showed an inverse correlation with heart SUV_max (Figure 5B). According to preliminary data from oncology patients with normal myocardium, physiologic myocardial FDG uptake was sufficiently suppressed at a serum free fatty acid level of 1,000 μEq/L. This result indicates that patients with suspected CS should have serum fatty acid > 1,000 μEq/L at the time of myocardial FDG-PET, and increased myocardial FDG uptake in patients with a serum free fatty acid level > 1,000 μEq/L at the time of FDG-PET can be regarded as pathologic FDG uptake. Thus, measurement of serum free fatty acid at the time of FDG-PET is expected to be useful to monitor the degree of myocardial FDG suppression and interpret results. Of course, further studies are required to establish an optimal cutoff value.

For evaluation of CS, CMR with an LGE protocol can reveal the development of fibrosis, whereas myocardial PET can detect increased metabolic activity of glucose uptake in affected myocardium. Because these two diagnostic modalities characterize different features of CS, hybrid imaging using CMR and myocardial PET is useful for diagnosing CS.56) Combined use of CMR and myocardial PET has distinct strengths, as it co-localizes lesions with increased metabolic activity while also detecting lesions with irreversible scar changes (i.e. presence of LGE without increased FDG uptake) and early forms of myocardial inflammation (i.e. increased FDG uptake in absence of LGE). Fusion images of CMR and myocardial PET can be simultaneously obtained using a hybrid system, but can also be reconstructed from separate images using post-processing software, as shown in Figure 6. Thus, combined images from CMR and myocardial PET may further contribute to a higher diagnostic yield for patients suspected to have CS.