Jørgen Frøkiær

Background: Dynamic PETwith 11C-acetate can be used to assess myocardial oxygen use which in turn is usedto calculate myocardial external efficiency (MEE), anearly marker of heart failure. MEE is defined as the ratio of total work (TW) and total energy use (TE). Calculation of TW and TE requires additional stroke volume (SV) and myocardial mass data, respectively, which are typically derived from a separate cardiovascular magnetic resonance (CMR) scan. Dual scanning is logistically problematic and different loading conditions during PET and CMR scans can cause errors in MEE estimates. The aim of this study was to develop and validate an automated method of calculating MEE from a single dynamic 11C-acetate PETscan.
Methods: 21 subjects underwent a dynamic 27 min 11C-acetate PETscan on a Siemens Biograph TruePoint 64 PET/CTscanner. Using cluster analysis, the LV-aortic time-activity curve (TACLV) was obtained automatically from the dynamic images and used to calculate parametric images of 11C-acetate uptake rate K1 and washout rate k2. SV was estimated from the area of the first-pass peak of TACLV. Using parametric K1 images, the left-ventricular wall was segmented automatically and myocardial mass was derived from the volume of the obtained regions. Gold standard SV and mass measurements were performed using cardiovascular magnetic resonance (CMR). TW, TE and MEE were estimated using the standard combination of PET and CMR data and using PET data alone.
Results: Both cluster analysis and LV segmentation were performed successfully in all patients. The range of MEE from CMR was 15.3+2.3% and for PET 19.6+4.3%. For all measures, a high correlation was found between PETand CMR derived values although significant differences were found for TW and MEE but not for TE (intraclass correlation coefficient [ICC] =0.85, bias of 23+15%, p=0.009 for TW; ICC=0.95, bias of -4+9%, p=0.74 for TE; ICC=0.78, bias of 29+17%, p=0.002 for MEE). Correcting for scannerdependent differences in SV eliminated biases in TW and MEE (ICC=0.85, bias of 0.2+14.3%, p=0.96 for TW, ICC=0.74, bias of 4.9+16.4%, p=0.26 for MEE).
Conclusion: Myocardial efficiencycanbe derived directly andautomatically froma single dynamic 11C-acetate PET scan. This eliminates the need for a separate CMR scan and eliminates any potential errors due to different loading conditions between CMR and PETscans.