Strict carbohydrate restriction enables volume-based FDG-PET/CT to outperform SUVmax for risk stratification in cardiac sarcoidosis

Background FDG-PET/CT SUVmax reflects focal metabolic intensity but not the patchy inflammatory burden of cardiac sarcoidosis (CS). We evaluated whether volume-based total cardiac metabolic volume (tCMV) and total cardiac metabolic activity (tCMA), measured after optimized suppression, predict long-term major adverse cardiac events (MACE) better than SUVmax. Methods Forty-eight consecutive patients with CS classified by JCS 2016 criteria underwent FDG-PET/CT after 24-h strict carbohydrate restriction without high-fat supplementation and delayed cardiac imaging at 120 min. tCMV, tCMA, and SUVmax were quantified at a fixed SUV threshold of 1.70. MACE was assessed over a median of 7.3 years. DeLong tests and Firth-penalized Cox regression evaluated prognostic performance. Results MACE occurred in 18 of 48 patients (37.5%). tCMA (AUC, 0.766) and tCMV (AUC, 0.764) outperformed SUVmax (AUC, 0.606; DeLong P = 0.008 and 0.013). High tCMA predicted higher MACE rates (log-rank P = 0.0003). In multivariate Firth-penalized Cox regression, tCMA (HR per unit, 1.0010 [95% CI, 1.0005–1.0016]; HR per IQR, 1.4093 [1.1671–1.7019]; P = 0.0002) and NYHA class (HR, 2.7460 [1.6209–4.6521]; P = 0.0002) were independently associated with MACE. Adding tCMA to NYHA class and troponin T provided modest incremental discrimination (ΔC-index, 0.081; 95% CI, 0.001–0.206). Conclusions Under optimized suppression, tCMA captured global inflammatory burden rather than focal intensity and improved MACE risk stratification in CS. Volume-based quantification may complement SUVmax and conventional risk markers, pending external validation.