Noninvasive cardiac risk stratification of diabetic and nondiabetic uremic renal allograft candidates using dipyridamole-thallium-201 imaging and radionuclide ventriculography.

The ability of noninvasive risk stratification using dipyridamole-thallium-201 (Tl-201) imaging and radionuclide ventriculography to predict perioperative and long-term cardiac events (myocardial infarction or cardiac death) was evaluated in 36 uremic diabetic and 29 nondiabetic candidates for renal allograft surgery. Of the 35 patients who underwent renal allograft surgery 8 +/- 7 months after the study, none had transient Tl-201 defects (although 13 had depressed left ventricular ejection fraction) and none developed perioperative cardiac events. During a mean follow-up of 23 +/- 11 months, 6 (9%) patients developed cardiac events. Logistic regression analysis was used to compare the predictive value of clinical data (including age, sex, diabetes, chest pain history, allograft recipient) and radionuclide data. Presence of transient Tl-201 defect and left ventricular ejection fraction were the only significant predictors of future cardiac events (p less than 0.01). No other patient variables, including diabetes or receiving a renal allograft, had either univariate or multivariate predictive value. All 3 patients with transient Tl-201 defects had cardiac events compared with only 3 of 62 (5%) patients without transient Tl-201 defect (p less than 0.0001). Mean left ventricular ejection fraction was lower in patients with cardiac events (44 +/- 13%) compared with patients without cardiac events (57 +/- 9%, p less than 0.005). Overall, 5 of 6 patients with cardiac events had either transient Tl-201 defects or depressed left ventricular ejection fraction. Dipyridamole-Tl-201 imaging and radionuclide ventriculography may be helpful in identifying uremic candidates for renal allograft surgery who are at low risk for perioperative and long-term cardiac events.