Early changes in thallium distribution. Effect on quantitative analysis

Thirty-two patients with coronary artery disease and an abnormality on an initial anterior view thallium scan had repeat images obtained after delays of 30 and 240 minutes. Scans were analyzed by quantitative criteria. Comparison of the initial stress study with the 30-minute redistribution scan showed no significant change in 11 patients, defects becoming smaller in 13 patients, and defects becoming larger in eight patients. When comparing the stress or the early redistribution images with the late redistribution scans, the diagnosis (eg, scar vs. ischemia) would have been affected in 14 cases. Analysis of the sources of variability showed that all the apparent worsening but only part of the defect resolution could be explained by variability inherent to repositioning the patient. Thus, the size of an initial defect is very sensitive to the time between the end of exercise and the onset of data collection and the nature of changes in scan appearance is complex.