Prognostic value of automated vs visual analysis for adenosine stress myocardial perfusion SPECT in patients without prior coronary artery disease: A case-control study

Abstract

PurposeWe aimed to evaluate the prognostic value of automated quantitative hypoperfusion parameters derived from adenosine stress myocardial perfusion SPECT (MPS) for predicting sudden or cardiac death (CD) in case-controlled patients with suspected coronary artery disease (CAD). MethodsWe considered patients with available adenosine stress Tc-99m sestamibi MPS scans and follow-up information. 81 CD patients from a registry of 428 patients documented by the National Death Index were directly matched in a retrospective case-control design to patients without CD by key clinical parameters (age by deciles, gender, no early revascularization, pre-test likelihood categories, diabetes, and chest pain symptoms). Multivariable analysis of stress MPS total perfusion deficit (STPD) and major clinical confounders were used as predictors of CD. Visual 17-segment summed stress segmental scores (VSSS) obtained by an expert reader, were compared to STPD. ResultsCD patients had higher stress hypoperfusion measures compared to controls [STPD: 7.0% vs 3.6% (P < .05), VSSS: 5.3 vs 2.1 (P < .05)]. By univariate analysis, STPD and VSSS have similar predictive power (the areas under receiver operator characteristics curves: STPD = 0.64, VSSS = 0.63; Kaplan-Meier models: χ2 = 7.59, P = .0059 for STPD and χ2 = 11.10, P = .0009 for VSSS). The multiple Cox proportional hazards regression models with continuous perfusion measures showed that STPD had similar power to normalized VSSS as a predictor for CD (χ2 = 4.92; P = .027) vs (χ2 = 8.90; P = .003). ConclusionsQuantitative analysis is comparable to expert visual scoring in predicting CD in a case-controlled study.