Quantitative analysis of myocardial metabolic heterogeneity is superior to visual assessment for the detection of active cardiac sarcoidosis by F-18 FDG PET-CT imaging

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background Myocardial inflammation may occur in the context of a multisystem disease such as sarcoidosis, adversely affecting prognosis. A definitive diagnosis of cardiac sarcoidosis (CS) is essential to implementing life-saving treatment but this is complicated by the invasive nature of endomyocardial biopsy (EMB) and its low accuracy. Positron emission tomography (PET) assists in diagnosis, which relies on visual interpretation of myocardial F-18 FDG uptake. The value of quantitative analysis and its application to clinical practice remain uncertain. Purpose To investigate the power of quantitative F-18 FDG PET-CT imaging analysis for detecting CS in patients with suspected disease. Methods All patients underwent F-18 FDG PET-CT after a 24-hour low-carbohydrate diet and 15-hour fasting as part of their diagnostic work-up for suspected cardiac inflammation. Cardiovascular magnetic resonance acted as gatekeeper to PET-CT in 8 of every 10 scans. Myocardial F-18 FDG uptake was assessed qualitatively and quantitatively using both manually drawn regions of interest and automatic polar maps to measure global and segmental standardised F-18 FDG uptake values (SUV). The coefficient of variation (CoV) was calculated to determine uptake heterogeneity. To confirm diagnosis, follow-up data regarding disease progression, further testing and treatment were collected. To allow for sufficient follow-up time, the first 40 consecutive patients from a prospective registry (n= 214; Sep 2017-Jun 2020) were included. Results A comprehensive clinical picture was obtained successfully in 37 patients (median [IQR], 17 [13.5] months) and a final diagnosis of CS reached in 7 (disease prevalence, 19%). EMB was performed in 2 patients only while 3 underwent PPM/ICD implantation. Significant predictors of CS were fulfilment of Japanese Ministry of Health and Welfare criteria (Wald, 6.44; p = 0.01) and left ventricular dysfunction (Wald 6.72; p = 0.01). Qualitative F-18 FDG PET-CT had a high negative (95%) but low positive (45%) predictive value for CS (sensitivity, 83%; specificity, 77%). F-18 FDG SUV CoV was the strongest imaging predictor (Wald, 6.77; p = 0.009) and was significantly higher in CS than non-CS (CoV median [quartiles], 0.26 [0.21, 0.36] and 0.12 [0.11, 0.14] respectively; p = 0.004). As per ROC curve analysis (AUC, 0.84), a CoV threshold of 0.20 was highly specific (93%) and sensitive (86%) for CS. Conclusion In a referring population with a low prevalence of cardiac sarcoidosis, F-18 FDG PET-CT imaging is sensitive for the detection of myocardial inflammation with active disease unlikely in patients with a negative scan. Quantitative evaluation of metabolic heterogeneity within the myocardium provides a strong, independent marker of active disease and should be considered alongside visual assessment.