Prognostic value of automated quantification of 99mTc-sestamibi myocardial perfusion imaging

Although human interpretation of 99mTc-sestamibi SPECT myo-cardial perfusion imaging has been repeatedly validated in the diagnostic and prognostic assessment of coronary artery dis-ease, it remains unclear if automated computer-derived quan-titative indices of perfusion have similar or independent prog-nostic information. Methods: We studied 718 patients referred for 99mTc-sestamibi SPECT myocardial perfusion imaging who were followed for 5.6 1.1 y (mean SD). The SPECT studies were initially interpreted visually without benefit of computer-based analysis and were then subjected to a blinded reprocess-ing to extract quantitative indices of perfusion. Follow-up was through the Manitoba Population Health Research Data Repos-itory. Acute myocardial infarction or cardiac death occurred in 79 individuals (11.0 % of the cohort). Results: Visual and quan-titative categorization of scan perfusion abnormalities showed similar prognostic value for predicting acute myocardial infarc-tion or cardiac death. Discordance between the visual and quantitative categorizations defined a group at intermediate risk. There was a gradient of risk with increasing severity of the summed stress score (SSS) or summed difference score (SDS). The automated SSS and SDS provided incremental prognostic information over that obtained from visual interpretation. Con-clusion: Automated quantification of 99mTc-sestamibi SPECT myocardial perfusion scans provides objective prognostic infor-mation and may complement the conventional visual image interpretation. Key Words: coronary artery disease; prognosis; 99mTc; sesta-mibi; quantification; myocardial perfusion imagin