Comparison of Exogenous Ketone Administration Versus Dietary Carbohydrate Restriction on Myocardial Glucose Suppression: A Crossover Clinical Trial

Visual Abstract The ketogenic diet (KD) is the standard of care to achieve myocardial glucose suppression (MGS) for assessing inflammation using 18F-FDG PET. However, failure to suppress physiologic glucose uptake remains a significant diagnostic barrier. Although extending the duration of KD may be effective, exogenously delivered ketones may provide a convenient, reliable, and same-day alternative. The aims of our study were to determine whether exogenous ketone administration is noninferior to the KD to achieve MGS and whether serum β-hydroxybutyrate (BHB) levels can predict MGS. Methods: KEETO-CROSS (Ketogenic Endogenous versus Exogenous Therapies for myoCaRdial glucOse SuppresSion) is a crossover, noninferiority trial of the KD (endogenous ketosis) versus ketone ester ([KE] exogenous ketosis) drink. Twenty healthy participants were enrolled into 3 arms: weight-based KE drink, 24-h KD, and 72-h KD (n = 18 completed all arms). The primary outcome was achievement of complete MGS on PET (noninferiority margin 5%). The area under receiver-operating-characteristics (AUROC) of endogenous BHB levels (analyzed in a laboratory and by point-of-care device) for predicting MGS was analyzed in 37 scans completed on the KD. Results: The mean age was 30 ± 7 y, 50% were women, and 45% were nonwhite. The median achieved BHB levels (mmol/L) were 3.82 (25th–75th percentile, 2.55–4.97) (KE drink), 0.77 (25th–75th percentile, 0.58–1.02) (25th–75th percentile, 24-h KD), and 1.30 (25th–75th percentile, 0.80–2.24) (72-h KD). The primary outcome was achieved in 44% (KE drink), 78% (24-h KD), and 83% (72-h KD) of participants (noninferiority P = 0.97 and 0.98 for KE vs. 24-h and 72-h KD). Endogenous BHB levels robustly predicted MGS (AUROC, 0.88; 95% CI 0.71, 1.00). A BHB of 0.58 or more correctly classified 92% of scans. A point-of-care device provided comparable predictive value. Conclusion: In healthy volunteers, KE was inferior to KD for achieving MGS. Serum BHB is a highly predictive biomarker for MGS and can be clinically implemented upstream of 18F-FDG PET, with rapid facilitation by point-of-care testing, to reduce false-positive scans.