A Comparison of Dipyridamole Stress Lower-Limb Thallium-201 Scintigraphy with Ankle-Brachial Index in Assessing Peripheral Arterial Occlusive Disease of Diabetic Patients

Background: Peripheral arterial occlusive disease (PAOD) has become more important in the treatment of diabetic patients due to it accounts many morbidities and mortalities. This study investigates the feasibility of dipyridamole stress lower-limb thallium scintigraphy (LLTS) in assessing the peripheral arterial occlusive disease of DM patients by comparing with the results of ankle-brachial pressure index (ABI). Methods: A total of 371 diabetic patients (203 males, 158 females, mean age 62.6, SD 13.0) with one or more cardiovascular risk factor (hypertension, dyslipidemia, aging, obesity) were prospectively recruited. The criteria for inclusion were: (1) symptomatic PAOD or asymptomatic subjects in outpatient clinics; (2) Subjects with poor control of DM and hospitalization. All patients received dipyridamole stress LLTS started right after the myocardial perfusion SPECT image (MPI). The mean pixel counts of ROI of the three levels-posterior aspects of buttock, thigh and calf were calculated. The right-to-left ratio of mean pixel counts of three levels with value less than 0.9 or greater than 1.1 were defined as abnormal LLTS. The ABI data was collected for comparison. The result of angiographic significant coronary artery disease shown in MPI was interpreted by two nuclear physicians, and a semiquantitative MPI with summed stress score (SSS) was generated by QPS 3.0 (GE) software for reference in 234 patients. A SSS greater than 8 was considered to have positive CAD. Whether abnormal LLTS can predict CAD were analyzed. Results: The association of abnormal LLTS and abnormal ABI is statistically significant especially in the calf level. When asymmetric of any of the three levels is considered positive LLTS, it can predict the result of ABI with a sensitivity of 63% and a specificity of 67% by Fisher's exact test (P＜0.0001). There was no statistically significant relationship between LLTS and visually interpreted CAD (P=0.8858) or between LLTS and positive CAD defined by SSS greater than 8 (P=0.6862). Conclusions: The dipyridamole stress LLTS coupled with myocardial perfusion scintigraphy is convenient and has the potential to access PAOD in DM patients. However, the sensitivity it provided (63%) was not sufficient to replace current available tools, such as ABI, for routine clinical screening use. Using LLTS as a predictor for CAD seems to be inconclusive.