Variability of myocardial blood flow and reserve measurements using quantitative dynamic SPECT perfusion imaging

Introduction: Dedicated CZT cardiac cameras provide accurate measurements of absolute myocardial blood flow (MBF) and flow reserve (MFR). Reproducibility of MBF and MFR remains unclear but critical for reliable clinical use. In this study, we evaluated the reproducibility of SPECT MBF and MFR measurement in patients referred for coronary artery disease (CAD) screening.

Methods: We retrospectively analyzed 28 patients referred for Myocardial Perfusion Imaging (MPI) for CAD screening included in a prospective trial (CFR-OR). SPECT data were acquired on a CZT-based pinhole cardiac camera (Discovery NM530c, GE Healthcare, Haïfa, Israël) in list-mode using stress (251 ± 13 MBq) / rest (512 ± 27 MBq) one-day Tc-99m-tetrofosmin protocol. Kinetic analyses were done with Corridor4DM^TM software using a 1-tissue-compartment model and converted to MBF using a previously determined extraction fraction correction. Motion detection was applied; no attenuation correction was applied.

Results: 28 patients (14 male, 14 female) were included, and SPECT acquisitions were twice reconstructed by two nuclear medicine board-certified physicians. Mean global MFR was 2.63 ± 1.12; 8 patients had impaired MFR (using a threshold of 2). Angiographically significant coronary stenosis were found at invasive coronary angiography in all 8 patients. Reproducibility of MBF diverse between measurements: mean difference (95% CI) -0.08mL/min/g (- 0.77 to 0.48), P = 0.01 for stress MBF, and -0.03mL/min/g (- 0.26 to 0.10), P = 0.03 for rest MBF. However, we noticed a better reproducibility for MFR: mean difference was -0.003 (-0.71 to 1.51), P=0.84. No significant difference was found between regional and global MFR measurements. Using the MFR threshold of 2, only 1 patient remained incorrectly classified, with a global MFR of 1.79 at first evaluation and 2 on the second read; however, regional MFR remained correctly classified for this patient.

Conclusions: Reproducibility of global and regional stress and rest MBF varied for the different measurements. However, no significant difference was found for MFR between different measurement, showing a better reproducibility. These results need to be confirmed on a larger scale but could confirm the clinical reliability for SPECT MFR evaluation.