JOURNAL OF SURGICAL RESEARCH &,379-384 (1990) Nonuniform Blood Flow in the Canine Left Ventricle1p2 ARTHUR E. FLYNN, M.D., DWAIN L. COGGINS, M.D., RICHARD E. AUSTIN, M.D., DEREK D. MUEHRCKE, M.D., GABRIEL S. ALDEA, M.D., MASAMI GOTO, M.D., JOSEPH W. DOUCETTE, AND JULIEN I. E. HOFFMAN, M.D. Department of Surgery and Cardiovascular Research Institute, Uniuersity of California, San Francisco, California 94143 Submitted for publication November 20, 1989 In order to investigate the relationship between cor- onary perfusion pressure and blood flow distribution in the left ventricle (LV), we measured myocardial blood flow in small regions using radioactive microspheres in six anesthetized, open-chest dogs. Mean coronary per- fusion pressure (CPP) was controlled with a femoral ar- tery to left main coronary artery shunt which included a pressurized, servo-controlled blood reservoir. In each dog, we measured flow in 192 regions of the LV free wall (mean weight per region = 206 + 38 mg) at different perfusion pressures. At CPP = 80 mm Hg, blood flow to individual regions varied fourfold (0.30 to 1.18 ml/min/ g; relative dispersion (RD) = 21.8 +_2.3%). At CPP = 50 mm Hg, flow varied over sevenfold (0.08 to 0.60 ml/min/ g; RD = 42.8 -t 10%; P < 0.01 vs 80 mm Hg). This re- lationship between flow variability and CPP was present within individual LV layers as well between layers and is much higher than the error associated with the micro- sphere technique. We conclude that blood flow to small regions of the LV is markedly nonuniform. This heter- ogeneity becomes more profound at lower CPP. These findings suggest that (1) global measurements of coronary flow must be interpreted with caution, and (2) even in hearts with normal coronary arteries some regions of the LV are more susceptible to ischemia than others. In addition, these findings may help explain the patchy na- ture of myocardial damage that occurs following periods of low coronary pressure or inadequate myocardial pro- tection during cardiopulmonary bypass. o ISSO Academic Press, Inc. INTRODUCTION Blood flow to the left ventricle is remarkably nonuni- form. This variability is most apparent between the dif- ferent layers of the left ventricular wall. Clinically, acute myocardial infarction [l, 21 or inadequate myocardial protection during cardiopulmonary bypass [3, 41 fre- 1 Presented at the Annual Meeting of the Association for Academic Surgery, Louisville, Kentucky, November 15-18,1989. * Sponsor: William P. Schecter, M.D. quently results in myocardial injury limited to the sub- endocardial layer. Experimentally, the vulnerability of the subendocardium has been demonstrated by the observa- tion that coronary reserve is exhausted in the deep layer at a higher perfusion pressure than in the more superficial layers of the left ventricle [5,6]. In addition to variability between layers, blood flow to small regions within the same layer is also nonuniform [7-121. This topographical variability in the left ventricle has been termed “spatial heterogeneity” and is present even in hearts with normal coronary arteries [13]. We hypothesized that low coronary perfusion pressure may accentuate the nonuniformity of myocardial flow within layers of the left ventricle, and therefore render some regions of the left ventricle more susceptible to injury than others. In order to test this hypothesis, we measured myocardial blood flow in small regions of the left ventricle during normal and low cor- onary perfusion pressures. METHODS Experimental Preparation Mongrel dogs of either sex (n = 6, weight = 30 to 49 kg) were premeditated with 2 cc droperidol-fentanyl (In- novar) im before being anesthetized with sodium 20 mg/ kg pentobarbital iv. All experiments were performed in accordance with the National Research Council’s guide for the care and use of laboratory animals. Anesthesia was maintained with a mixture of oxygen, halothane (I%), and nitrous oxide (60%) delivered by a Harvard respirator through a cuffed endotracheal tube. Arterial pH and blood gases were measured frequently and kept in the normal range by adjusting inspired oxygen concentration and minute ventilation. A heating pad was used to maintain normal body temperature. A fluid-filled catheter was placed in the descending thoracic aorta via the femoral artery and connected to a strain gauge transducer (Sta- tham P32Db) for measurement of aortic pressure. A tho- racotomy was performed in the left fourth or fifth inter- costal space, and the heart was suspended in a pericardial cradle. A solid-state transducer (Gaeltec Instruments, Model 16CT) was introduced into the left ventricle (LV) 379 cxx?z-4804/90 $1.50 c opynght 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.