Myocardial blood flow quantitation (MBFQ) with 99mTc-sestamibi dynamic SPECT (dSPECT) for prediction of angiography documented coronary artery disease (CAD): The impact of missing attenuation correction (AC)

1728 Objectives MBFQ with AC dSPECT previously demonstrated clinical value in CAD detection (Chen et al, RSNA 2013). The adequateness of dSPECT without AC (No-AC) for MBFQ is still unknown. We aim to evaluate the impacted diagnostic capability of No-AC dSPECT using invasive coronary angiography (CAG) as the reference standard. Methods Patients (n=25) who received a dSPECT scan under 2-head SPECT and separate CT or SPECT/CT scanners with CAD verified by CAG were included (CAD=14, non-CAD=11, Age=69.7, BMI=24.5, M=15). Rest/dipyridamole-stress dSPECT images were reconstructed with physical corrections and No-AC or AC by using a single CT scan for AC. Blood flow in heart was quantified with 1-tissue compartment model with corrections of spillover, tracer extraction and rest condition. Values and heterogeneity (SD/mean*100%) of flow and coronary flow reserve (CFR=stress flow/rest flow) obtained from both image sets were analyzed. Receiver operating characteristics (ROC) statics was further employed to assess the CAD detection of No-AC and AC using the criterion of ≥50% stenosis as CAD positive. Results No-AC dSPECT significantly had higher stress K1 (21%), stress flow (49%), CFR (243%) with 2.0 times (1.6-2.4) elevated heterogeneity than AC dSPECT. ROC analysis further revealed No-AC dSPECT underperformed AC SPECT for CAD detection although insignificantly. Conclusions Missing AC in dSPECT reduced the accuracy of MBFQ and amplified flow heterogeneity. Attenuation correction for dSPECT is mandatory to warrant the diagnostic performance of MBFQ.