Validation of an algorithm for predicting cardiac events in renal transplant candidates

A 2-tiered noninvasive cardiac risk stratification algorithm was first evaluated in a test population with planar thallium myocardial perfusion imaging and subsequently in a validation population using single-photon emission computed tomographic (SPECT) thallium myocardial perfusion imaging. This study examines if SPECT imaging was as predictive of cardiac death as planar imaging and also if SPECT imaging predicted nonfatal cardiac events in the patient population. Renal transplant candidates were evaluated using a 2-tiered noninvasive cardiac risk stratification algorithm. The first tier of risk assessment utilized 5 variables: age >50 years, insulin-dependent diabetes mellitus, abnormal electrocardiogram, and a history of either angina or congestive heart failure. Patients without risk factors were considered low risk and underwent no further cardiac evaluation. Patients with > or =1 risk factor were considered high risk and underwent a second tier of risk assessment with planar (n = 95) or SPECT (n = 112) imaging. In the test population, 13 of 16 cardiac deaths (81%) occurred in high-risk patients with abnormal planar studies. In the validation group, all cardiac deaths (5 of 60) and nonfatal cardiac events (13 of 60) occurred in high-risk patients with abnormal SPECT studies. SPECT imaging was at least as predictive as planar imaging and also predicted nonfatal as well as fatal cardiac events. Pretransplant risk stratification by clinical variables identified low-risk patients who may not require further cardiac evaluation and high-risk patients with normal SPECT imaging who may not require angiography.