Abstract 19783: Performance of Ultra-Low Dose Tc-99m Sestamibi Myocardial Perfusion Imaging Using Solid-State SPECT: Comparison to Coronary Angiography

Abstract

Introduction: Ultra-low dose (ULD) SPECT is feasible using solid-state detector systems, with advantages of lower dosimetry and lower radionuclide costs. Thus far, there have been no studies validating accuracy of ULD SPECT against coronary angiography. We therefore performed a 2-expert consensus blinded review of 37 patients who had coronary angiography and ULD SPECT (D-SPECT, Spectrum Dynamics) within 30 days, with a stress-first protocol utilizing < 4 mCi Tc-99m sestaMIBI. Methods: Pts were imaged both supine & upright for ∼10 6 myocardial counts; rest images were acquired for equivocal or abnormal stress images using ∼4X stress dose. Mean stress dose was 3.9 mCi +0.47 mCi and mean rest dose was 14.8 mCi +1.8 mCi (n=23). Patient mean age was 62 yrs; 32% were female; 43% underwent pharmacologic stress. Each study was interpreted as ischemic vs non-ischemic, and also evaluated semi-quantitatively (4-point scale) for perfusion and function using a 17 segment model. Significant CAD was defined as a stenosis >70% on coronary angiography. Results: Coronary angiography was abnormal in 25 and normal in 12 pts; 11 had single vessel (SVD) & 14 had multi-vessel disease (MVD). ULD SPECT sensitivity was 88% (22/25), specificity was 92 % (11/12), PPV was 96%, and NPV was 79%. Sensitivity for detection of SVD was 91% and for MVD was 92%. Detection of a 2 nd or 3 rd diseased vessel in pts with MVD was 54%. Conclusions: The results of this study are encouraging that ULD-SPECT with patient dosimetry ∼10% of typical rest/stress SPECT is diagnostically accurate for identification of significant CAD. Correct identification of MVD is currently suboptimal.