Diagnostic value of transient dilatation of the left ventricle in negative dipyridamole-thallium imaging.

To evaluate the diagnostic value of a transient dilatation of the left ventricle during dipyridamole-thallium imaging (DTI) for detecting significant coronary artery disease (CAD) in patients with negative DTI results, 81 consecutive patients were studied. Twenty one patients (26%) had CAD and 60 patients had normal coronary anatomy (NCA). The initial/delayed ratio of the left ventricular dimension, which was measured as the distance between the 2 peaks of a count profile curve on a 45 degrees left anterior oblique planar image, was defined as the dilatation ratio (DR) of the left ventricle. Patients with CAD had a higher incidence of chest pain after dipyridamole infusion (35 vs 13%; p < 0.05), and ST depression during exercise testing (50 vs 25%; p < 0.05) than those with NCA. DR was significantly greater in CAD patients than in NCA patients (1.08 +/- 0.10 vs 0.97 +/- 0.03; p < 0.0001). DR was considered abnormal (> 1.03) when it was greater than the mean +2 standard deviations of the DR in NCA patients. Seventy-six percent of CAD patients had an abnormal DR. A stepwise discriminant analysis revealed that an abnormal DR alone had the same ability to predict CAD (sensitivity 76%, specificity 98%, chi-square 80.9, p < 0.0001) as the best combination of abnormal DR, chest pain during exercise testing, age and gender (sensitivity 76%, specificity 98%, chi-square 98.5, p < 0.0001). When abnormal DR was excluded from this analysis, the best combination of the variables showed a reduced ability to predict CAD (sensitivity 81%, specificity 77%, Wilks' Lambda 0.71, chi-square 26.7, p < 0.0001).(ABSTRACT TRUNCATED AT 250 WORDS)