Introduction

Sarcoidosis is an inflammatory disorder of unknown etiology characterized by non-necrotizing granulomas involving one or more organ systems. Approximately 25% of patients with pathology-proven extracardiac sarcoidosis have evidence of cardiac involvement by advanced cardiac imaging,1 a rate that is consistent with a previously published U.S.-based autopsy study.2

Because glucose metabolism is markedly increased in inflammatory cells (such as in granulomas), myocardial inflammation can be detected by 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET). Cardiac 18F-FDG PET has become an essential tool in the evaluation of suspected or established cardiac sarcoidosis (CS).3 Since normal cardiomyocytes also use glucose for energy production, suppression of physiologic myocardial 18F-FDG uptake is critical for visualization of 18F-FDG uptake by inflammatory cells. The recently published joint SNMMI/ASNC expert consensus document on the role of 18F-FDG PET in CS detection and therapy monitoring underscored the importance of patient preparation aimed to achieve suppression of physiologic myocardial 18F-FDG uptake in patients undergoing PET for CS detection.3 This may be accomplished using several approaches: a fat-enriched diet lacking carbohydrates for 12-24 hours prior to the scan,4,5 a 12-18-hour fast,6 and/or the use of intravenous unfractionated heparin approximately 15 minutes prior to 18F-FDG injection.7 The expert consensus document acknowledges the substantial variability in these preparation protocols as well as their success rates in achieving suppression of physiologic myocardial 18F-FDG uptake in published studies. In a recent review of 31 studies assessing various suppression protocols, success rates ranged between 10 and 100%, with the majority of studies having a small sample size.8 Based on published literature and expert opinion, the joint expert consensus document advocates a combined approach of at least two high-fat (> 35 g), low-carbohydrate (< 3 g) (HFLC) meals the day before the PET study followed by a fast of at least 4-12 hours. The purpose of this study was to compare the differences in the suppression of physiologic 18F-FDG myocardial uptake between (1) a diet protocol consisting of only HFLC meals and (2) a structured protocol of combined HFLC meals and prolonged fasting with compliance reinforcement and detailed patient instructions.

Methods

Patient Population and Preparation Protocols

The study included consecutive patients who underwent cardiac PET imaging for detection of CS between March 2010 and September 2017 at a single tertiary academic medical center (Mayo Clinic, Rochester, Minnesota). Prior to August 1, 2016 (Group 1), the preparation protocol consisted of only HFLC meals, and patients were asked to eat up until 2 hours before the 18F-FDG PET study. Patients were contacted only 24 hours prior to the 18F-FDG PET study. Only one example of an acceptable meal was provided and no instructions were given regarding exercise activities. Starting on August 1, 2016, a new suppression protocol with stricter and earlier instructions was implemented.9 Differences between the old and new preparation protocols are outlined in Table 1. The new protocol instructed patients to consume at least two HFLC meals the day prior to the 18F-FDG PET study, and fast for at least 15 hours after the meals before the PET study. Detailed examples of HFLC meals and permissible vs undesirable foods were provided. Patients were also instructed not to exercise for 24 hours prior to the PET examination. An alternative option to the combined HFLC meals and fasting was an 18-hour fast, but this approach was discouraged by nursing staff when the patients were called 48 hours prior to testing. On the day of the study, nursing/technical staff ensured that patients have strictly followed the recommended instructions and reviewed the food diary questionnaire. Compliance with the new preparation protocol was ensured by a combination of detailed written instructions (Appendix 1), a phone call from nursing staff 48 hours before the PET study, and a food diary questionnaire (Appendix 2), which was reviewed by healthcare providers at the time of the study. Both outpatients and inpatients were included in the study and the same dietary instructions were provided to both. In the case of inpatients, we worked closely with the dietetics team and nursing staff to ensure dietary compliance. Based on review of the food diary questionnaire, patients were categorized as fully compliant, partially compliant, or non-compliant. Non-compliant patients were not scanned and were instructed to return after following the protocol. Partial compliance was defined as (1) limited carbohydrate consumption (1-2 carbohydrate containing options described in Appendix 2), (2) a last food intake past 5 p.m. (but not past 7 p.m.) the day prior to PET scan, or (3) exercise the day prior to PET scan. Non-compliance was defined as more than 3 carbohydrate items listed in Appendix 2 or late food intake (after 7 p.m. the day prior to PET scan).

Table 1 Comparison between old and new preparation protocols
Full size table

Data Collection

Data were retrospectively collected, prospectively analyzed, and included all consecutive PET scans performed at our institution for the assessment of CS between March 2010 and September 2017. In patients who had multiple studies either before or after application of the new protocol, only the first of the studies before August 2016 and the first of the studies after August 2016 were included in the analysis. This was done to avoid data duplication and patient over-representation in the study sample. Other clinical variables were obtained via manual review of the electronic medical record.

PET Sarcoidosis Protocol

PET scans were performed using the GE Advance PET camera (GE Medical Systems, Milwaukee, WI, USA) between March 2010 and January 2016. The GE Discovery PET/CT camera (GE Medical Systems, Milwaukee, WI, USA) was used for all tests performed starting in February 2016 through September 2017. A 10-minute transmission scan utilizing 68Germanium rod sources was used for attenuation correction with the GE Advance camera. Computed tomography was used for attenuation correction in scans performed with the GE Discovery camera (helical scanning, slice thickness 3.75 mm, interval 3.27 mm, 120 kVp with arms elevated (default) or 140 kVp with arms along the side). Initial images included myocardial perfusion assessment using N13 ammonia (10-20 mCi) administered via intravenous injection. Following acquisition of perfusion images, inflammation was assessed using 18F-FDG (15-20 mCi) administered via intravenous injection. The resting 18F-FDG scan was performed after a period of 60 minutes following 18F-FDG administration to allow for 18F-FDG uptake. With the GE Discovery PET/CT camera, image acquisition was performed in 3D for 10 minutes. PET/CT fusion images were reviewed for accurate co-registration to ensure accuracy of attenuation correction. Gated perfusion as well as static perfusion and 18F-FDG series were transferred to an in-house developed nuclear cardiology viewing software10 as well as a nuclear medicine dedicated software (MIM Software Inc, Cleveland, Ohio). Myocardial perfusion and 18F-FDG images were displayed in 3 orthogonal planes (short axis, vertical long axis, and horizontal long axis). Myocardial perfusion and 18F-FDG uptake were assessed in each of the standard 17 myocardial segments as described previously.11 Attenuation corrected as well as non-corrected images were also reviewed using a general nuclear medicine hybrid display software (MIM Software Inc., Cleveland, OH, USA).

Image Interpretation

PET images of patients who underwent the study using the old preparation protocol (group 1) vs the new preparation protocol (group 2) were evaluated for suppression of physiologic myocardial 18F-FDG uptake. All studies were interpreted by consensus of two staff cardiologists with specialized training in advanced nuclear cardiac imaging. Interpretation of myocardial 18F-FDG uptake was performed as per the 2017 SNMMI and ASNC guidelines as follows: 1) No 18F-FDG uptake, 2) Focal 18F-FDG uptake, 3) Diffuse 18F-FDG uptake, and 4) Focal on diffuse 18F-FDG uptake (Figure 1).3 Patients with focal 18F-FDG uptake or no 18F-FDG uptake were considered to have complete suppression of physiologic 18F-FDG uptake. Patients with diffuse 18F-FDG uptake were considered to have suboptimal suppression of physiologic 18F-FDG uptake. Patients with focal on diffuse myocardial 18F-FDG uptake were reviewed by HJ and if 4 or more myocardial segments had normal perfusion without 18F-FDG uptake, physiologic 18F-FDG suppression was considered adequate. We based this on the assumption that physiologic myocardial 18F-FDG uptake is uniform and it would be very unlikely to have differential suppression of myocardial 18F-FDG uptake between various segments. Isolated basal lateral 18F-FDG uptake has been reported to be a normal variant in the literature and was not considered to represent inadequate 18F-FDG suppression.

Figure 1
figure 1

Patterns of myocardial 18F-FDG uptake on cardiac PET. Patterns of myocardial 18-fluorodeoxyglucose (FDG) uptake were interpreted per established guidelines as follows: (A) No 18F-FDG uptake, (B) diffuse myocardial 18F-FDG uptake, (C) focal 18F-FDG uptake, and (D) focal on diffuse 18F-FDG. Patients presented in panels (A) and (C) were considered to have appropriate suppression of physiologic myocardial 18F-FDG uptake while (B) had suboptimal suppression of myocardial 18F-FDG uptake. Patients in panel (D) had either appropriate or suboptimal suppression of myocardial 18F-FDG uptake based on discretion of the cardiac imager (see Methods)

Full size image

Statistical Analysis

Continuous variables were expressed as means (standard deviation) and compared using student’s t test. Categorical variables were expressed as numbers (percentage) and compared using the Chi-square test. Comparison of matched pairs was performed using the McNemar’s test. A two-tailed P value of < .05 was used to determine statistical significance. All analyses were performed using JMP software version 9.0 (SAS institute, Cary, NC, USA).

Results

Of the 366 patients that underwent sarcoidosis PET scans, 134 had the old preparation protocol (group 1) and 232 had the new preparation protocol (group 2). Patient characteristics are summarized in Table 2. There were no statistically significant differences in age, sex, body mass index, diabetes mellitus, coronary artery disease, or left ventricular ejection fraction between the two groups. Glucose levels at the time of PET scan and were 94 ± 24 mg·dL−1 (group 2) and 101 ± 21 mg·dL−1 (group 1) (P = .04). Inpatients represented 5% of the total patient population and were overrepresented in group 1 (P = .03). 18F-FDG myocardial suppression was noted in 78% of PET scans in group 1 compared to 91% of PET scans in group 2 (P < .001, Table 2). Among patients in group 1, “diffuse” myocardial 18F-FDG uptake (indicative of suboptimal physiologic 18F-FDG suppression) was noted in 12% of these studies compared to only 2% of studies in group 2, (P < .001). In group 2, prolonged fasting (at least 18 hours) was the preparation of choice in 13 patients (6%) and was associated with adequate 18F-FDG suppression in 11 patients (85%).

Table 2 Patient characteristics and PET findings
Full size table

Regarding the new preparation protocol (group 2), complete compliance and partial compliance were noted in 70% and 30% of the patients, respectively. No major differences were noted in compliance rates between males (71%) vs females (68%) (P = .63). Exercise prior to PET scanning was the cause of partial compliance in 2% of patients. Partial compliance in group 2 was associated with adequate 18F-FDG suppression in 88% of patients. Patients who had suboptimal physiologic suppression of 18F-FDG uptake in group 2 had compliance rates of 50% (10 out of 20 patients). Patients who had partial compliance had physiologic 18F-FDG suppression in 88% of cases and patients who had complete compliance had physiologic 18F-FDG suppression in 93% of cases (P = .31).

There were 47 patients who underwent 2 PET scans in the study period; one utilizing the old preparation protocol and the later scan utilizing the new preparation protocol. In these patients, physiologic myocardial 18F-FDG suppression was achieved in 85% using the old protocol compared to 96% using the new protocol, albeit this difference was not statistically significant (P = .18, likely due to lack of statistical power). Of the 47 patients, 7 patients had inadequate 18F-FDG suppression that changed to adequate 18F-FDG suppression, 2 patients had adequate 18F-FDG suppression that changed to inadequate 18F-FDG suppression, and 38 patients had adequate 18F-FDG suppression before and after the protocol changed.

Discussion

The present study shows that implementation of a structured and detail-oriented patient preparation protocol in line with the SNMMI/ASNC expert consensus document was successful in suppressing physiologic myocardial 18F-FDG uptake in most patients undergoing cardiac PET for CS. To the best of our knowledge, our study is the largest in the literature to assess appropriate suppression of myocardial 18F-FDG uptake using a structured patient preparation protocol. The frequency of diffuse myocardial 18F-FDG uptake, associated with false positives, decreased 6-fold using the new protocol and complete suppression of physiologic myocardial 18F-FDG uptake was achieved in 91% of patients appropriately following the new protocol compared to 78% using the old protocol.

The optimal preparation for sarcoidosis PET scan is not well established. Different strategies to suppress physiologic myocardial 18F-FDG uptake have included prolonged fasting, dietary manipulation, intravenous heparin, or a combination of the above methods. Prolonged fasting for at least 12-18 hours has been successful in suppressing physiologic 18F-FDG uptake. However, it is laborious and is associated with decreased patient compliance.8 Dietary manipulation is the most common preparation method due to its efficacy and patient tolerability. It most commonly includes a low-carbohydrate diet or a low-carbohydrate/high-fat diet, often in combination with fasting. The exact amount of fat or carbohydrate intake has not been standardized although more than 35 grams of fat and less than 5 g of carbohydrates were suggested in a prior study.12 Several studies have trialed a high-fat meal or beverage in the last hours prior to the scan without significant benefit.13,14,15 Lastly, low-dose intravenous heparin administration increases lipolysis and plasma free fatty acid levels in healthy individuals, and has been utilized, alone or in combination with dietary manipulation, to suppress 18F-FDG uptake, with variable results.7,16

Prior research shows great variability in the adequacy of physiologic myocardial 18F-FDG uptake suppression. In a recent review of 31 studies by Osborne and colleagues, success rates ranged between 10 and 100%.8 Studies with unrestricted diet and short fasting periods (< 12 hours) tended to be less successful achieving proper suppression of myocardial 18F-FDG uptake (10-60%). However, studies that implemented prolonged fasting (> 12 hours), with or without dietary manipulation prior to the fasting period yielded the highest success rates (60-100%). Manabe et al reported a 0% diffuse uptake rate in 24 patients who underwent preparation with a low-carbohydrate diet, an 18-hour fast, as well as intravenous heparin.6 However, the small patient sample size does not allow for definitive conclusions. Coulden and colleagues reported a 98% success rate with a low-carbohydrate diet “Atkins diet” prior to >8 hours of fast in 94 patients.17 However, Coulden et al defined successful suppression of myocardial 18F-FDG uptake as myocardial maximum standard uptake values < 3.6 without describing the pattern of myocardial 18F-FDG uptake. Of note, a large proportion of their intervention group (n = 26, 22%) were non-compliant with the recommended dietary instructions and were excluded from the study. In the largest currently available study by Blankstein and colleagues, success rate was 87% in 118 patients with applications of a high-fat, no-carbohydrate diet for two meals with a 4-hour fast.18 Similar results and a 90% success rate were reported by Harisankar et al in 60 patients who had a similar preparation.4 Lastly, a high-fat, low-carbohydrate combination diet for > 24 hours with heparin achieved an 84% success rate in 89 patients as reported by Nensa et al 19

Based on prior literature and anecdotal experience, our institution modified the patient preparation protocol for cardiac sarcoidosis in August 2016. We identified that the major culprit for inadequate 18F-FDG suppression was patient non-compliance. As such, the new protocol aimed to enhance timely patient education and re-enforcement. Providing a list of dietary options is presumably superior to patient-driven options. Moreover, having the patients log their food in the preparation timeframe likely increases food awareness and overall compliance. A prolonged fasting option may be suitable for select patients (vegetarian/vegan patients or patients with difficulty logging their food items). Since the evidence for benefit of high-fat meals in the 12 hours prior to PET scan is lacking, those interventions were eliminated from the new preparation protocol. It is possible that eliminating the HFLC morning meal on the day of the scan also improves PET imaging quality, since it eliminates the possibility of inadvertent carbohydrate intake close to the time of testing. Exercise prior to scanning has been postulated to result in an increase in skeletal muscle and myocardial FDG uptake via a norepinephrine-mediated mechanism. Accordingly, patients were instructed not to exercise for 24 hours prior to PET scan although the contribution of this intervention was hard to quantify (only two patients reported exercise with the new protocol and were termed partially compliant). Lastly, heparin administration prior to the scan was not used given the inconsistent evidence to support its use as well as institutional constraints regarding heparin use in the outpatient setting. It should be noted that although the new preparation protocol was designed to promote compliance, complete dietary adherence is likely difficult to achieve and 30% of the group 2 patients had only partial compliance.

Our study demonstrated a success rate of 91% in suppressing physiologic myocardial uptake. It is noteworthy that 50% of cases with suboptimal physiologic myocardial 18F-FDG suppression in patients using the new preparation protocol were only partially compliant with the provided instructions, which could possibly underestimate the protocol’s true success rate when the dietary instructions are followed completely. Conceivably, our study’s protocol could be implemented at other tertiary centers with high volumes of CS studies to improve myocardial suppression of physiologic 18F-FDG uptake. This can improve the specificity of positive myocardial 18F-FDG uptake which has significant diagnostic and therapeutic implications in patients with CS. Additionally, it may also decrease cost and patient radiation dose related to repeat testing.

Several important limitations warrant discussion: (1) The present study was performed at a single tertiary academic institution; therefore, care should be exercised when extrapolating our results to other centers. (2) By study design, our PET image analysis used in this study was semi-quantitative. Inter-observer variability in image interpretation and classification is possible and could be subjective to some degree.20 However, error from misclassification would be random rather than systematic and having two reviewers for each study minimizes the odds of misinterpretation or misclassification. (3) We did not quantify fat and carbohydrate amounts consumed in our study since this quantification is difficult. We presumed that a patient strictly adhering to the suggested meals would have consumed < 3 grams of carbohydrate (amount suggested by SNMMI/ASNC consensus document). (4) Patient over-representation bias from including sequential PET scans in the study could potentially skew results in favor of the new preparation protocol, particularly since patients requiring repeat follow-up testing can learn what dietary habits are most appropriate and effectively suppress physiologic 18F-FDG uptake in the subsequent scans. However, only 7 of the 47 patients that were included in both groups 1 and 2 in our study had a change from inadequate to adequate physiologic suppression. (5) We did not assess the clinical implications of CS presentation and treatment on PET myocardial 18F-FDG imaging findings, as this was beyond the scope of this study.

In conclusion, we describe an efficacious structured patient preparation protocol that included clear communication of dietary instructions between healthcare providers and patients, in agreement with the SNMMI/ASNC expert consensus document. This protocol was highly successful in achieving complete suppression of physiologic myocardial 18F-FDG uptake and should be strongly considered to promote patient adherence and improve PET image quality.

New Knowledge Gained

  • A structured protocol of combined high-fat low-carbohydrate meals and prolonged fasting with compliance reinforcement and detailed patient instructions as recommended by the joint SNMMI/ASNC expert consensus document can significantly suppress physiologic 18F-FDG uptake in positron emission tomography for cardiac sarcoidosis.

  • Partial compliance with a strict protocol can result from inadvertent carbohydrate consumption, decreased time interval of prolonged fasting, and exercise, is only achieved in 70% patients, and could potentially result in failure to suppress physiologic myocardial 18F-FDG uptake.