Ongoing myocardial thickening after aortic valve closure (postsystolic thickening = epsilonPST) is an established marker for the presence of segmental ischemia. However, epsilonPST may also be present in late activated segments and can be induced by pharmacological interventions or left ventricular pressure overload. The aim of this study was to determine if it is possible to distinguish between ischemic and nonischemic epsilonPST. In an experimental pig-model (n = 11) regional radial deformation was measured in the inferolateral wall during either normal perfusion or regional ischemia using ultrasonic strain rate imaging. Ischemia was induced by active hypoperfusion of the circumflex coronary artery territory. Measurements were made at 1. baseline, and during 2. theodrenalin infusion, 3. dobutamine infusion 4. esmolol infusion and 5. during a preload increase induced by saline infusion. In all segments where thickening was ongoing after aortic valve closure, the amount of epsilonPST was calculated as the difference of maximal strain minus systolic strain. In addition, peak strain rate during the isovolumetric relaxation period was extracted. During normal coronary perfusion, 73% of all segments (n = 40) developed epsilonPST. This physiological epsilonPST averaged 5 +/- 2% and was most frequently induced during the esmolol infusion (n = 11). Peak isovolumetric strain rate averaged -2.1 +/- 0.5 s(-1) in segments with physiological epsilonPST. During coronary hypoperfusion, 96% of the "at risk" segments developed epsilonPST. EpsilonPST in the ischemic segments averaged 14 +/- 3%, and was highest during the dobutamine infusion (25 +/- 4%) and lowest during the esmolol infusion (5 +/- 1%). In contrast to normally perfused segments, peak isovolumetric strain rate was positive in the ischemic segments and averaged 2.0 +/- 0.5 s(-1) in these pathologic segments with postsystolic strain. Using a cut-off value of > or = 0 s(-1) for isovolumetric strain rate, pathologic epsilonPST was detected with a sensitivity of 100% and a specificity of 87%. These experimental findings were confirmed by a subsequent clinical study with 6 patients with acute myocardial infarction (ischemic group) and 6 patients with arterial hypertension or aortic stenosis (nonischemic group). Ischemic and nonischemic postsystolic thickening can be precisely differentiated by extracting the polarity of the peak isovolumetric strain curve.