Routine clinical reproducibility of hyperemic myocardial blood flow measurements with 15O-water-PET

Abstract Background Hyperemic myocardial blood flow (MBF) has been shown to provide the most accurate non-invasive diagnosis of hemodynamically significant CAD. A high reproducibility of hyperemic MBF measurements is important for accurate stratification of patients. Previous studies have shown that whole-myocardium hyperemic MBF measurement with 82Rb has a repeatability coefficient (RC) of 50% (1.3 mL/g/min) for a two-day test-retest using adenosine in healthy controls 1 and 21% (0.51 mL/g/min) and 38% (0.90 mL/g/min) for a single-session and two-day test-retest using dipyridamole, respectively, in patients in clinical routine 2. For 15O-water and adenosine, an RC of 25% (0.90 mL/g/min) has been reported in 1999 for healthy controls, whereas RC at the regional level was 50% 3. Purpose The aim of the current work is to assess the reproducibility of hyperemic MBF measurements with 15O-water using state of the art technology. Methods Ten patients referred for assessment of ischemia underwent a same day test-retest protocol consisting of 4-min dynamic rest and stress scans starting simultaneously with controlled bolus injection of 400 MBq 15O-water on a digital PET-CT scanner. Patients left the scanner between test and retest sessions and time between test and retest scan sessions was circa 1 h. Adenosine infusion at 140 μg/kg/min was started 2 min prior to each stress scan and continued during the entire scan. Images were analysed using fully automated software. RC (1.96 x SD of test-retest differences), mean relative test-retest difference and within-subject coefficient of variation (wCV) were calculated at the whole myocardium, regional and segmental levels. Results Whole myocardium hyperemic MBF ranged from 1.3 to 4.6 mL/g/min. RC was 15% (0.43 mL/g/min), 23% (0.67) and 27% (0.80) at whole myocardium, regional and segmental levels, respectively. Mean test-retest difference was 7%, 10% and 11%, and wCV was 5.5%, 8.3% and 9.8%, respectively. Figure 1 shows scatter plots of MBF values and figure 2 shows stress MBF polarmaps of a typical patient. Conclusion Automated data analysis and use of modern PET scanners results in a large improvement of reproducibility of hyperemic MBF measurements with 15O-water. Reproducibility of hyperemic MBF with 15O-water is superior to previously published values for 82Rb.Scatter plot of test-retest MBF values Stress MBF polar maps of one subject