Further Evaluation of the Circular Sequential Vein Graft Technique of Coronary Artery Bypass Joseph C. Cleveland, M.D., Ira M. Lebenson, M.D., Robert J. Twohey, M.D., Joseph G. Ellis, M.D., Daniel B. Nelson, M.D., Raymond J. Suchor, M.D., Aldred A. Heckman, Jr., M.D., David W. Morse, M.D., and John R. Dague, B.S.N. ABSTRACT Our study reports a series of circular sequential vein grafts in 1 patients with highly symptomatic triple-vessel coronary artery disease. Four or more distal anastomoses were done in each patient. Thirteen of the patients were restudied, and the results revealed a 97% patency rate for distal anastomoses ( out of 60) at 4 to 1 months after operation. One patient died months after operation. Postmortem examination revealed a desmoplastic, fibrotic reaction at the proximal anastomosis of the circular graft, with of 4 distal anastomoses patent. Twenty of the 1 patients in this series are now alive with asymptomatic cardiac status 14 to months after operation. The finding by Grondin and associates [l] of increased patency rate with this technique for distal anastomoses is confirmed. The circular sequential vein graft represents a particularly advantageous technique for patients in whom 4 to 6 distal anastomoses are needed for complete revascularization and in whom one or more vessels have limited runoff. The obvious disadvantage of this technique is that all distal anastomoses depend on a single proximal anastomosis. There is virtually no debate that coronary artery bypass operation ameliorates the symptoms of angina pectoris in the vast majority of patients [, 1. Hurst and co-workers [1 concluded that we have adequate evidence that bypass surgery prolongs the life of patients with left main arterial obstruction and triple or double vessel From the Departments of Cardiovascular Surgery and Cardiology, Carle and Christie Clinics, Urbana-Champaign, IL. Presented at the Sixteenth Annual Meeting of The Society of Thoracic Surgeons, Jan 1-, 1980, Atlanta, GA. Address reprint requests to Dr. Cleveland, c/o Carle Clinic Association, 60 W University Ave, Urbana, IL 61801. obstruction. Additionally, it has been shown that exercise-induced myocardial dysfunction is usually abolished [4]. Relief of symptoms and adequacy of myocardial revascularization are related to graft patency [51. Various studies have demonstrated saphenous vein patency to range between 75 and 90% in the first 6 to 1 months following operation [, 4, 61. Since the advent of coronary artery bypass operation, multiple techniques have been used and advocated. In the last two or three years, bypass to smaller branches of the coronary tree, as well as the major trunks, has been practiced. Thus, complete or total myocardial revascularization and high patency rates of the bypass conduits are key technical points of this operation. The experience reported by Grondin and colleagues [l] with the circular sequential vein graft seemed to achieve these goals with four to five distal vessels being bypassed per patient and a reported patency of 96%. In February, 1978, we began to use the circular sequential saphenous vein bypass graft technique on selected patients with symptomatic triple-vessel coronary artery disease (Fig 1). Twenty-one such patients underwent operation during 1978, and 4 to 6 distal anastomoses were constructed in each. One patient died months after operation, and postmortem examination showed severe scarring and near occlusion of the proximal anastomosis, even though of 4 distal anastomoses were patent. After this event, the remaining 0 were contacted and angiographic restudy was recommended. Thirteen patients were restudied. Material and Method The 1 patients ranged from 7 to 7 years old (mean age, 57 years). There were 19 men and 6 000-4975/80/1006-06$01.5 @ 1980 by The Society of Thoracic Surgeons
7 Cleveland et al: Circular Sequential Vein Graft Technique Fig 1. Circular sequential saphenous vein graft with distal anastomoses to (1) the posterior descending right coronary artery, () the second marginal circumflex and () the first marginal circumflex coronary arteries, (4) the diagonal branch, and (5) the left anterior descending coronary artery. women. All had severe triple-vessel coronary artery disease with disabling angina. Eight patients had unstable angina, i.e., unprovoked angina, or an accelerating pattern in terms of frequency and severity. Four patients had left main coronary artery disease (stenosis greater than 70%). All except of these patients were on a regimen of maximal medical therapy at the time operation was advised. In every patient, there was some distal vessel atherosclerotic disease and less than ideal size in one or more arteries to be grafted. All of the patients were operated on using standard cardiopulmonary bypass techniques with moderate hypothermia and hemodilution. Myocardial protection was accomplished with cold cardioplegic solution infused into the aortic root to achieve electromechanical rest of the myocardium. The distal anastomoses were done first in the following sequence: An endto-side anastomosis was created on the left anterior descending coronary artery. The anastomosis was usually about 6 mm in length, and the largest segment of the reversed, trimmed, and spatulated saphenous vein was used. The next coronary arteries to be grafted were one or more diagonal branches. These were side-toside anastomoses with a.5 mm opening in the diagonal coronary artery and a.5 mm oblique opening in the saphenous vein. For the circumflex marginal coronary arteries, a vertical.5 mm opening in the coronary artery and a.5 mm transverse opening in the saphenous vein were made. Again, a side-to-side anastomosis was created. Usually the posterior descending aspect of the right coronary artery was used for anastomosis, and a vertical opening was made here and an oblique or vertical opening in the saphenous vein, again.5 mm in length. The final anastomosis was made to the highest anterior part of the ascending aorta near the middle of that structure. It was done in a vertical manner and was about 7 mm long. This technique varies from the technique described by Grondin and co-workers [11 in that we did not use the diamond shaped side-toside anastomosis or start at the same point on the heart. All coronary anastomoses were gently probed with 1 mm, malleable, metallic probes prior to completion to make sure that both proximal and distal anastomoses were patent. A continuous suture technique was used in all anastomoses with 7-0 polypropylene cardiovascular suture for distal anastomoses and 5-0 polypropylene cardiovascular suture for proximal anastomoses. A two-needle technique was employed with tying of the single suture at the midpoint of the second side of the anastomosis. This particular technique has been used by our group for all coronary anastomoses and has been shown to be quite satisfactory on numerous reexaminations. Results Tables 1 and show the important data for the entire patient group. Flow rates in the vein grafts were measured with an electronic flowmeter just beyond the aortic anastomosis,
8 The Annals of Thoracic Surgery Vol 0 No 4 October 1980, Table 1. Preoperative and Postoperative Data on the Patients Who Underwent Recatheterization Left Preop Main Preop Systolic Coronary Functional Ejection Patient Age Disabling Unstable Artery Class Fraction No. (Yd Angina Angina Diseasea (NYHA) ( 70) 1 4 5 6 7 8 9 10 11 1 1 71 51 5 69 66 8 6 7 6 5 + =Greater than 70% stenosis. NYHA = New York Heart Association. 48 4 4 51 4 54 50 71 6 80 66 prior to chest closure. The range was from 100 to 400 ml per minute (mean, 197 ml per minute). There was a zero operative mortality for this series, and a single late death occurred at months. This patient was readmitted to the hospital because of chest pain and died shortly after readmission. Postmortem examination revealed a severe, fibrotic, desmoplastic reaction of the proximal anastomosis of the vein graft and aorta. Three of 4 distal anastomoses were patent. That finding prompted a restudy of the remaining patients. All were contacted and asked to undergo reevaluation by clinical examination and cardiac recatheterization. Six patients declined angiographic restudy. All had an asymptomatic cardiac status. They did submit to clinical reexamination and were found to be in New York Heart Association Functional Class I. One patient agreed to the restudy, but severe hives had developed at his first cardiac catheterization. Because he was extremely active and asymptomatic, it was thought the risk of recatheterization exceeded the benefits to be gained. The remaining 1 patients were restudied between 4 and 1 months after operation (mean, 8 months). Eleven were completely asymptomatic and active. One patient had been very active but mild recurrent angina was developing on heavy exertion. Restudy of this patient showed a 95% proximal stenosis of the sequential vein graft. This patient was promptly reoperated on, and a saphenous vein bypass to the previous bypass graft was constructed. Now, he is again very active and completely asymptomatic. Another patient was admitted to the hospital with anxiety and chest wall complaints. The graft was patent despite widespread diffuse disease of all coronary arteries. Twenty of the 1 patients are alive and essentially asymptomatic in terms of cardiac status 14 to months following operation. In the 1 patients who underwent recatheterization, there were 60 distal anastomoses, of which were patent at the time of recatheterization. Two patients each had a single marginal coronary artery of small size which could not be refilled at the time of repeat injection of the sequential circular graft. This is a 97% distal patency rate. With the additional information gained at postmortem examination of the 1 patient who died, the distal patency rate becomes 61 of 64 (95%). Systolic ejection fractions were slightly increased postoperatively but were not significantly different from
9 Cleveland et al: Circular Sequential Vein Graft Technique Postop Systolic Postop Flow in No. of Grafts Ejection Time to Functional No. of Grafts Patent on Fraction Restudy Class Anastomoses (mumin) Restudy ( OIo 1 (mo) (NYHA) 5 50 4 61 1 1 5 150 5 46 10 1 5 10 5 6 10 1 5 10 4 48 9 1 5 90 5 67 8 1 4 70 4 6 11 1 4 180 4 47 4 1 4 140 4 49 8 1 4 0 4 51 6 1 5 180 5 66 6 1 4 165 4 48 8 1 5 15 5 77 4 1 5 10 5 64 8 1 Table. Data on the 8 Other Patients in the Study Lefta Preop Main Preop Systolic Postop Coronary Functional Ejection No. of Flow in Functional Patient Age Disabling Unstable Artery Class Fraction Anasto- Grafts Class No. (yr) Angina Angina Disease (NYHA) (%) moses (mumin) (NYHA) 45 + 59 + 48 + 55 + 50 + 69 + 65 + 7b + agreater than 70% stenosis. bthis was the patient who died months after operation. NYHA = New York Heart Association. 51 55 67 60 49 4 70 61 150 140 N.A. 00 50 150 400 100 preoperative values (p > 0.05). There was a mean systolic ejection fraction of 5% preoperatively, and the mean ejection fraction postoperatively was 56%. All of the proximal anastomoses in the remaining 1 patients studied were widely patent. Comment The most salient finding of our study is the confirmation of a distal anastomotic patency rate of about 96% as reported by Grondin and co-workers [ll. Their patients were restudied within six weeks after operation. Our patients were restudied at a mean of 8 months after operation. Thus, the immediate and intermediate distal patency rate of about 96% in such a group of surgical patients is demonstrated and confirmed. Flow rate is a major factor in short-term and long-term patency rates of vein grafts. Further, patency rate and completeness of myo-
40 The Annals of Thoracic Surgery Vol 0 No 4 October 1980 cardial revascularization are major determinants of successful long-term surgical results. Thus, the circular sequential graft meets these requirements for myocardial revascularization quite well. The problem of the dependency of all distal anastomoses on one proximal connection was demonstrated in our series. We believe that in both patients who had stenosis of the proximal anastomosis, the problems were due to technical considerations that can be resolved and that are not inherent weaknesses of the method. The patient who died was the eldest patient in the series (7 years old). The aorta was highly atherosclerotic with a concomitant inflammatory response. To try to avoid an area of heavy plaque formation, we shifted the proximal anastomosis to the right side of the aorta rather than place it in the usual position (anterior aorta). The saphenous vein was also somewhat small compared with a usual-sized vein. Placing a large vein patch on the aorta and the circular graft into this vein patch with the use of a larger segment of saphenous vein (necessitating a splice) might have prevented the problem that later developed. In the second patient with difficulty, the saphenous vein bifurcated (going distad) just above the knee into two equal-sized vessels. This meant that the proximal anastomosis would have only a marginal size. A larger vein used proximally (by means of a splice) would have been preferable. The situation was later repaired with the saphenous vein from the opposite leg as a new bypass to the existing graft using a large segment for a proximal generous cobrahead anastomosis. Although our surgical technique differs somewhat from that described by Grondin and colleagues El], we find only one difference that we consider crucial. They stated that the arteriotomies for the side-to-side anastomoses should be 4 to 6 mm in length. We find that a.5 mm opening is appropriate, but that more than 4 mm will always consume so much surface area of the vein that it will pucker in the saphenous vein, thereby creating a potential obstruction to flow. Recently a vein patch angioplasty was necessary when the arteriotomy for a side-to-side anastomosis was inadver- Fig. Angiogram made at restudy of a patient with a circular sequential saphenous vein graft. The distal vessels bypassed as well as the course of the saphenous vein are demonstrated. tently extended to 5 mm. This puckered in too much of the remaining saphenous vein, and repair was done by placing a large diamondshaped vein patch on the opposite side. Other than that, we think that either technique is satisfactory for construction of the circular graft. In any event, the circular sequential graft is technically more demanding than the usual double or triple bypass using single veins. We certainly recommend that patients be carefully selected for circular sequential grafts. Probably no more than 15% of patients having coronary bypass operation are suitable candidates. We consider patients who have the most advanced degree of coronary artery disease to be the most suitable for this technique (we recognize that no patient is truly inoperable). Even if the coronary anatomy is as described but the saphenous vein is not of good size and character, other techniques will be more appropriate. However, there exists a subgroup of patients who seem to be ideal for this technique. They
41 Cleveland et al: Circular Sequential Vein Graft Technique are highly symptomatic patients with triplevessel coronary artery disease, with or without left main involvement, who need four to six distal vessels bypassed to achieve adequate to complete myocardial revascularization. The functioning of a good circular graft is impressive in seeing and stripping the graft at operation, measurement of the flow rate in the graft at operation, and visualization of the graft on repeat angiography (Fig ). Our study confirms the early patency rate of 96% obtained by Grondin and co-workers [ll and demonstrates that this rate is present up to 1 months after operation. We know that grafts patent at one year stay patent and functioning for upwards of seven years [7]. We believe that all patients in whom this technique of myocardial revascularization is used should be studied by repeat coronary and graft arteriography within 6 months after operation until the results of the technique are more firmly established. Potential problems can be identified and corrected by this approach. As in most surgical endeavors, problems tend to lessen with experience as results improve. References 1. Grondin CM, Vouhe P, Bourassa MG, et al: Optimal patency rates obtained in coronary artery grafting with circular vein grafts. J Thorac Cardiovasc Surg 75:161, 1978. Mathur VS, Guinn GA, Anastassiades LC, et al: Surgical treatment for stable angina pectoris. N Engl J Med 9:709, 1975. Hurst JW, King SB, Logue RB, et al: Value of coronary bypass surgery. Am J Cardiol4:08, 1979 4. Kent KM, Barer JS, Green MV, et al: Effects of coronary artery bypass operation on global and regional left ventricular function during exercise. N Engl J Med 98:144, 1978 5. Hartman CW, Kong Y, Margolio JR, et al: Aortocoronary bypass surgery: correlation of angiographic, symptomatic and functional improvement at 1 year. Am J Cardiol 7:5, 1976 6. Sheldon WC, Rincon G, Pichard AD, et al: Bypass graft operation, with a study of 741 patients followed -7 years. Frog Cardiovasc Dis 18:7,1975 7. Seides SF, Borer JS, Kent KM, et al: Long-term anatomic fate of coronary artery bypass grafts and functional status of patients five years after operation. N Engl J Med 98:11, 1978 Discussion DR. CLAUDE M. GRONDIN (Montreal, Que, Canada): The number of patients and the patency rate reported by Dr. Cleveland are similar to what my colleagues and I reported three years ago. There are some differences, however. We did not encounter stenosis on the proximal graft, which Dr. Cleveland attributes to a faulty technique. I am concerned that in his series there were two instances of severe narrowing of the proximal portion of the graft, one said to be of 95%. These, of course, endanger patency of all the distal anastomoses. Therefore, the reported 96% patency rate may be misleading if this is not taken into account. The second point of difference is in the occurrence of isolated occlusion of a side-to-side anastomosis without concomitant occlusion of the graft beyond such an occlusion. We did not see this phenomenon. I relate it to the use of parallel side-to-side anastomoses instead of diamond side-to-side anastomosis, as described in our series. We believe there are some important technical points in a circular vein graft. The thigh vein should be used. It is important to measure the distance between anastomoses before clamping the aorta and to use diamond side-to-side anastomoses. At times it may be necessary to use an additional segment of graft from the aorta to the primary graft. Like Dr. Cleveland and his group, we have been concerned with the risk of having all the distal 4 or 5 anastomoses depend on a single aortic orifice. Therefore, recently, particularly in instances when the segment of graft next to the aorta is of small caliber or when there is any doubt concerning excess length or kinks, we have used a segment of graft between the aorta and that segment of the graft comprised between the left arterior descending coronary artery and the first diagonal branch. This anastomosis, by the way, must be on the medial aspect of the primary graft because of the danger of adding too much height to that graft. This could lead to compression and occlusion of the secondary graft. DR. LEBENSON: Thank you, Dr. Grondin, for your comments. Our problems with the proximal anastomoses were encountered early in our experience (the steep portion of our learning curve). We believe now that we have overcome these technical problems. We certainly appreciated your comments about the additional graft and will take that into consideration. We are comfortable in performing side-to-side anastomoses in any configuration that is comfortable for us. As the patency rates for all techniques appear to be equal, we will probably continue doing what seems most convenient and comfortable for us. We stopped performing the circular grafts while we restudied the patients. Since the results were satisfactory, we performed this operation in an additional 15 patients, all of whom are asymptomatic at this time.