Noninvasive quantification of regional myocardial perfusion with rubidium-82 and positron emission tomography. Exploration of a mathematical model.

Positron emission tomography (PET) centers without cyclotrons use generator-produced rubidium-82 (82Rb) for assessment of myocardial perfusion. The aim of the present study was to determine whether myocardial blood flow could be assessed quantitatively with 82Rb and PET. Because the myocardial extraction fraction of 82Rb varies inversely and nonlinearly with flow and cannot be measured conveniently with PET, we used an experimentally derived mathematical function defining the relation between single-pass extraction fraction of 82Rb and flow to obviate the necessity of measuring the extraction fraction directly. Myocardial blood flow in absolute terms (ml/g/min) was estimated from dynamic PET scans after intravenous administration of 82Rb in intact dogs and compared with flows measured with radiolabeled microspheres. In 36 comparisons in 13 dogs studied at rest, or after coronary occlusion, reperfusion, or after coronary hyperemia induced with intravenous dipyridamole, over the flow range from 0.2 to 2.0 ml/g/min, estimates of perfusion with rubidium correlated well with flows measured concomitantly with microspheres, although there was a slight underestimation of flow with rubidium (flow by 82Rb = 0.92 x flow by microspheres-0.021, r = 0.83). In general, estimates of flow in ischemic regions were less reliable than estimates for regions with normal flow. Thus, although the relation between myocardial extraction and retention of 82Rb and flow can vary under a variety of physiological and pathophysiological conditions, this study demonstrates the ability to obtain quantitative estimates of myocardial blood flow with 82Rb and PET under carefully defined conditions without measuring the extraction fraction directly.