Abstract
1597
Objectives: Quantification of myocardial blood flow (MBF) has become central in clinical application of rubidium-82 (82Rb) PET myocardial perfusion scans. The current joint statements from the SNMMI Cardiovascular council and the American Association of Nuclear Cardiology (ASNC) suggest bolus injections of 30-40 mCi of 82Rb when using LSO-based PET systems. However, the injection of high doses of 82Rb poses a potential risk of PET system saturation on most 3D PET/CT systems currently being used, which results in an artifactual increase in MBF measurements. We aimed to evaluate the impact of PET system saturation with a 3D PET/CT system and to test the potential use of a half-dose (HfD) acquisition protocol.
Methods: This study comprised of twenty patients who underwent repeated rest 82Rb scans in the same imaging session, with one scan employing a full-dose (FD) and the other a HfD protocol. All patients were scanned on a Siemens Biograph-64 TruePoint PET/CT. Datasets were evaluated for saturation based on visual assessments of input functions and sinograms. Saturation of the datasets was declared when single-slice rebinned sinograms (3D sinograms), obtained for the bolus arrival in the heart, had visually reduced countrates in combination with speckle noise in the corresponding PET reconstructions. We compared FD and HfD MBF measurements using coefficients of variation (CV) and paired t-tests. By comparison with the datasets without saturation, a correction factor between MBF was obtained with FD (MBFFD) and HfD protocols (MBFHfD) were established, thus, permitting preservation of MBF estimates obtained for FD protocol when using a HfD imaging protocol.
Results: The HfD protocol resulted in dose reduction of 46.7% compared to FD (FD: 33.7±5.3mCi, HfD: 17.9±3.1mCi). Saturation effects were observed in 4/20 (20%) FD scans, with none observed in the 20 HfD scans. Assessment of global MBFs for FD and HfD protocols revealed a bias in the MBF assessments of 0.09 ml/g/min (global MBFFD= 1.03±0.29 vs MBFHfD = 0.94±0.22 ml/g/min [p=0.001]) (Figure 1 A, C). Exclusion of patients with identified saturation effects (N=4) reduced the bias to 0.05 ml/g/min ( MBFFD= 0.97±0.28 vs MBFHfD = 0.92±0.23 ml/g/min [p=0.02]). Following exclusion of the saturated datasets, an excellent correlation of FD and HfD was observed (R=0.97), as well as high test-retest reproducibility (CV = 6.0%). This allowed correction for the linear bias: Corrected MBFHfD =1.09*MBFHfD-0.03 ml/g/min. MBFFD and MBFHfD did not differ following the bias correction (MBFFD=0.97±0.28 vs MBFHfD,corrected=0.98±0.25 ml/g/min, p=0.77) (Figure 1, B and D).
Conclusions: Saturation effects can be problematic in 82Rb MBF studies using the recommended FD protocols for 3D PET/CT scanners. The use of HfD protocol eliminates the risks of saturation, with comparable MBF assessments as obtained in FD imaging if a simple correction is applied. We propose to use HfD imaging protocols to avoid saturation of the 3D-PET systems.