Saphenous vein graft valves: “The bad guys”

Abstract

he report in this issue by Molina [I] is a substantial T departure from standard coronary artery bypass graft technique of using reversed saphenous vein for coronary artery bypass operations. This report deserves careful scrutiny by all practicing coronary artery bypass surgeons for following reasons. Flow physics clearly show inadequacies of what we have termed coronary-graft discrepancy, ie, a 6- to 7-mm distal diameter saphenous vein graft anastomosed to a 1.5-mm coronary artery with that graft having a 2- to 3-mm proximal anastomosis to aorta. Our 1-year patency rate of 72% with that graft configuration is significantly below our current average 1-year saphenous vein graft patency of 92%. We have abandoned using grafts in that fashion and have addressed that problem by two approaches. A 3- to 4-cm vein patch has been added to enlarge proximal anastomosis or, on occasion, vein graft has been used in a nonreversed fashion as described by Molina. Valves in saphenous vein grafts became suspect as the bad guys as early as 1971. At that time we noted our first case of distal embolism from a cast of thrombus liberated from a saphenous valve to cause a fatal left anterior descending myocardial infarction (Fig 1). This prompted a more careful assessment of saphenous vein graft valves on repeat angiography. The result in a surprising number of instances was finding of filling defects under sinuses of saphenous vein graft valves on repeat angiograms. This prompted development of a valvulotome [2] designed to divide central portion of each valve cusp, rendering valves incompetent as described and shown in Figure 3 in Molina's article [l]. We have found that turbulent flow is clearly diminished after valvulotomy. The valve remnant adheres to vein wall in reversed saphenous graft. Postoperative angiography in Molina's study of nonreversed saphenous grafts revealed that valve site could not be identified in 35% of grafts studied. In 28 remaining grafts only curled remnants of valve without obstruction or thrombus were found. Turbulent flow through a normally functioning saphenous vein valve has been identified as a source of thrombi and emboli for many years [3]. Whitney and associates [4], in an elaborate study using time-lapse angiography over an 11-year period, noted microscopically that majority of stenotic vein lesions occur at valve sites. Furthermore, this group proved that arterialized saphenous vein valve never fully opens more than 50% in systole. They showed that fibrotic valve stenosis is acquired by increased work of opening and closing Address reprint requests to Dr Mills, Cardiology Center, 4500 Wichers Drive, Marrero, LA 70072.