Different ischemia size estimation on exercise and dipyridamole quantitative databases with 99mTc-sestamibi myocardial SPECT

Background. Frequently, dipyridamole (DIP) myocardial perfusion SPECT is analyzed quantitatively using an exercise (EXE) database. Objective. To compare a group of patients with known coronary artery disease, processed with DIP and EXE polar maps. Method. We studied 20 males and 19 females with DIP Tc99m-Sestamibi myocardial SPECT. After similar processing, the DIP SPECT images were read using both EXE and DIP polar map databases. We analyzed defect and reversibility extension (in pixels) and severity in standard deviations. Results. Agreement to classify patients as having necrosis, ischemia or both was 92% (kappa: 0.859). Stress defect size was similar with both databases in males and females. However, in males reversibility resulted larger (156.4±107.7 versus 128.6±113.6 pixels; p=0.0251) and more severe (699.6±665.8 versus 486.6±551.9 SD; p=0.0003) with EXE. In females, reversibility resulted smaller (92.8±110.3 versus 113.0±121.9 pixels; p=0.0024) and less severe (278.5±356.7 versus 363.8±432.7 SD; p=0,0009) with EXE. Conclusion. Patients undergoing DIP stress were correctly classified as presenting ischemia, necrosis or both by EXE polar map database. However, EXE database underestimated ischemia extension in females and overestimated it in males