The influence of coronary collateral flow on the assessment of myocardial perfusion by videodensitometry

OBJECTIVES: Coronary collateral flow often mitigates the effects of coronary artery obstruction and has a significant impact on the prognosis of patients with coronary artery disease. In the presence of variable degrees of coronary collateral flow, digital radiographic assessment of myocardial blood flow has not been quantitatively validated. METHODS: A distal coronary arterial collateral path was created into the left anterior descending coronary artery (LAD) bed in 8 anesthetized pigs. Both LAD and collateral paths were pump-perfused and corresponding flows measured. A number of commonly used digital indices and parametric images of myocardial perfusion were then extracted from the sequence of images filmed before and during the injection of contrast. Data were acquired at 5 levels of total flow (LAD flow + collateral flow): 100, 85, 70, 55 and 40% of maximally vasodilated, baseline flow. At each level of total flow, data were acquired at 4 levels of collateral flow ratios (collateral flow/total flow): 0, 10, 25 and 50%. RESULTS: Regional percent segment shortening, reflecting myocardial blood flow, decreased as total flow fell, and remained unaltered when coronary collateral ratio alone was altered without change in total flow. On the other hand, linear regression between total flow and digital indices at 10, 25 and 50% coronary collateral flow ratios, compared with 0%, showed a successive and significant downward displacement, documenting an underestimation of flow by all digital indices in the presence of collateral flow. CONCLUSIONS: In the absence of a collateral pathway and during maximal coronary vasodilation with adenosine, digital radiographic indices of myocardial perfusion, based upon indicator dilution theory, show a relatively good correlation with regional transmural myocardial blood flow. However, due to underestimation of total transmural blood flow, these indices have limited utility when myocardial perfusion is provided in part by a collateral pathway. The effect is probably related to an alteration in the regional vascular volume into which iodinated contrast is injected.