Incremental prognostic value of stress myocardial perfusion imaging in asymptomatic diabetic patients

Objective: Stress myocardial perfusion single-photon emission computed tomography (MPS) variables are robust estimators of prognosis. No data are available on the comparative ability of stress MPS risk markers using varied iterative and risk classification approaches in asymptomatic diabetic patients. We compared analytical approaches to estimate the added value of MPS variables in estimating coronary artery disease (CAD) outcomes in asymptomatic diabetic patients. We also evaluated the temporal characteristics of cardiac risk according to MPS findings. Methods: A total of 436 consecutive asymptomatic diabetic patients who underwent stress-rest gated MPS were prospectively enrolled. Multivariable Cox proportional hazards model was employed to estimate cardiac death and nonfatal myocardial infarction (MI). Risk reclassification was calculated and parametric survival analysis was used to predict time to events. Results: At multivariable analysis, post-stress left ventricular ejection fraction (LVEF) and stress MPS ischemia were independent predictors of CAD death or MI (both p < 0.01). The net reclassification improvement by adding MPS results to a model including pre-test CAD likelihood was 0.25 (95% confidence interval 0.06-0.44; p < 0.01). Parametric survival analysis showed the highest probability of CAD death or MI and the major risk acceleration in time in patients with stress MPS ischemia and post-stress LVEF <= 45%. Conclusion: In asymptomatic diabetic patients, analytical approaches that establish the reclassification of events may serve for estimation of improved outcomes for stress MPS. Post-stress LVEF and stress-induced ischemia by gated MPS influence the temporal characteristic of the patient's risk at long-term follow-up. (C) 2013 Elsevier Ireland Ltd. All rights reserved.