Reversible ischemia in severe stress technetium 99m-labeled sestamibi perfusion defects assessed from gated single-photon emission computed tomographic polar map Fourier analysis.

Abstract

Detection of reversible ischemia and regional viability by inference from the presence of preserved resting myocardial thickening on dynamic electrocardiographic-gated tomographic (GSPECT) images obtained after stress injection could potentially obviate the need for a separate resting injection. This would decrease the cost, effort, duration, and radiation exposure of diagnostic myocardial perfusion imaging. The aim of this study was to determine whether functional images derived from GSPECT stress myocardial perfusion images, which represent indices of regional wall thickening, could predict the pattern of reversibility of perfusion defects in myocardial segments with severe perfusion defects on stress 99mTc-labeled sestamibi (99mTc-sestamibi) images. Reversible ischemia in myocardial segments with severe hypoperfusion (< or = 50% of normal activity) on stress 99mTc-sestamibi images was assessed with GSPECT indexes of myocardial thickening, as reflected by an increase in regional count density during systole. GSPECT bullseye plots were generated for each of eight frames acquired after stress 99mTc-sestamibi injection in 44 patients with coronary artery disease and at least one severe perfusion defect on summed (ungated) SPECT images. With first harmonic Fourier amplitude (AMP) and AMP/perfusion ratio (APR) images, regional myocardial systolic thickening was assessed according to a nine-segment model as absent, markedly reduced, mildly reduced, or normal thickening. These data were compared on a regional basis with defect reversibility determined with resting sestamibi images. Of 124 severe stress defects, 32 showed absent, 53 minimal, 35 partial, and four complete reversibility on resting images. AMP and APR scores had 75% and 82% agreement on the presence or absence of reversibility on resting images, respectively (both p < or = 0.00005), and both had significant agreement (p = 0.0072 and p < or = 0.00005, respectively) with resting reversibility grades by kappa analysis. AMP correctly identified 83% of the defects that were reversible on resting images, whereas the APR identified 96% (p = 0.0014 for sensitivity of APR vs AMP). On analysis of the triad of segment scores, the AMP slightly underestimated the degree of resting reversibility (p = 0.0002), whereas APR images indicated more reversibility than did resting images (p < or = 0.00005). AMP and APR indicated a greater degree of reversibility than did resting images in at least one myocardial segment in 13 (30%) and 27 (61%) of the 44 patients, respectively. The presence and degree of reversible ischemia in severe stress sestamibi perfusion defects seen on resting images can be detected reliably by the pattern of regional myocardial thickening demonstrated by Fourier analysis of GSPECT polar maps. When indexed for degree of perfusion, the APR images predict a greater degree of reversibility than do resting images.