Quantitative analysis of stress thallium-201 myocardial scintigrams: a multicenter trial

Abstract

Previously we validated a method for quantification of 201TI myocardial stress distribution and washout in which the patient's stress and washout circumferential profiles are compared with observed normal limits. The present study reports the results of a multicenter trial in which this method, utilizing normal limits from our institution, was employed to evaluate the presence, location, and extent of coronary artery disease (CAD). The normal limits utilized were generated from 49 patients having a low likelihood of CAD. The study population included 157 patients from four centers in the United States and Canada as well as a comparative prospective population from Cedars-Sinai Medical Center (CSMC) of 51 patients with CAD, 30 patients with normal coronary arteriograms, and 30 additional low-likelihood normals. The results in the combined centers regarding overall detection of CAD revealed a sensitivity of 84% and a frequency of test normality in the patients with low likelihood of CAD of 88%, compared to a sensitivity of 82% and true normalcy rate of 83% obtained in the prospective CSMC population. The sensitivity for detecting disease increased according to the extent of angiographic CAD in both the multicenter sites and the prospective CSMC group. Regarding localization of disease, similar sensitivities and specificities for detecting disease in individual coronary arteries were found in the multicenter sites and the prospective CSMC population. The results indicate that our method for quantifying 201TI stress-redistribution scintigrams utilizing standard normal limits can be applied at other institutions using a variety of scintillation cameras with similar accuracy to that currently obtained at our institution.