of Aorta-to-Coronary Artery

Transcription

1 Reoperation Following Clinical Failure of Aorta-to-Coronary Artery Bypass Vein Grafts Maurice Adam, M.D., Gerald F. Geisler, M.D., Cary J. Lambert, M.D., and Ben F. Mitchel, Jr., M.D. ABSTRACT In the past four years the authors have used 1,360 saphenous vein grafts (SVG) in the treatment of 600 patients with coronary artery disease. In 19 patients reoperation was carried out. The SVG occlusions were related to distal obstructions or to fibrotic and inflammatory changes around the SVG and at times were unrelated to detectable causes. The presence of both closed and open vein grafts in some of the patients suggests that SVG obstructions are probably due to local causes and, it is hoped, are nonrecurring in most instances. Primary SVG failure may be related to local vein factors such as ischemic or mechanical trauma, and efforts to avoid these are strongly recommended. Of the 17 survivors, 14 had excellent results or were improved clinically following reoperation, and we have concluded that reoperation is safe and effective in achieving ultimate success in those patients in whom distal arterial runoff is adequate. Recurring angina demands prompt restudy and corrective operation when possible. Treatment of postpericardiotomy syndromes with corticoids and avoidance or evacuation of mediastinal hematomas are important. R eversed autogenous saphenous vein bypass grafting has become the procedure of choice in the treatment of myocardial ischemia in both established angina and preinfarction syndromes [l, 2, 81. It serves to alleviate angina, it improves myocardial work load tolerance, and it may protect against later myocardial infarction. Some subsequent clinical failures must inevitably occur on the basis of progression of the patient s arterial disease. However, primary failure of the saphenous vein graft (SVG) also occurs. The prevention of such failures is an important surgical goal, but solution awaits clear definition of the etiology of SVG closure. Studies done over a number of years in femoral-to-popliteal artery bypass graft procedures [3-61 indicate that though clinical failure may often be due to pathological conditions within the vein graft rather than to progressive arterial disease, vein bypass grafts in fact have considerable resistance to failure under such adverse circumstances as distal arterial obstruction. More recently, Johnson, Auer, and Tector [71 have described changes of intimal thickening within vein grafts leading to their obstruc- From the Departments of Thoracic and Cardiovascular Surgery, Baylor University Medical Center, St. Paul Hospital, and The University of Texas Southwestern Medical School, Dallas, Tex. Presented at the Eighth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, Calif., Jan , Address reprint requests to Dr. Adam, 3434 Swiss Ave., Dallas, Tex THE ANNALS OF THORACIC SURGERY

2 Reoperation Following Vein Graft Failure tion, and this has been suggested by Vlodaver and Edwards [lo] to be due to a water hammer effect secondary to poor runoff under conditions of unaccustomed pressure within the veins. Other modes of failure include SVG thrombosis, suture granulomas, and technical errors. Clinical Material In the past four years the authors have used 1,360 aorta-to-coronary artery bypass vein grafts in the treatment of 600 patients with coronary artery disease. Of these, 82% have had good long-term results [2]. Interestingly, some of these patients have one or more occluded vein grafts but have maintained a good clinical status on the basis of residual patent grafts or infarction. The exact number is not known because the asymptomatic patients are not routinely studied by repeat coronary angiography. On the other hand, those becoming symptomatic again are restudied. In a few of these, all grafts were found to be open and the condition of the arteries unchanged. In others, vein graft narrowing or occlusion had appeared in one of several grafts or in all the grafts. Arterial disease had progressed in some patients. In 19 of the 600 patients we believed reoperation was indicated. The purpose of this study is to analyze this group of 19 patients in order to establish the results obtainable by reoperation techniques, to try to get some idea of the incidence of arterial disease progression, and to determine the reasons for graft failure when possible. The patients, 3 women and 16 men, ranged in age from 34 to 64 years. The reoperations were done from 1 to 28 months following the initial procedure. The criteria for operability were the same as those established for the initial procedure in these patients [I, 81, namely, the presence of angina, the presence of significant anatomical arterial or SVG obstruction, and the demonstration of an adequate runoff below an area suitable for grafting. Most of the patients had good ventricular contractility and were in reasonably stable condition, but there were exceptions. Table 1 groups the patients according to what was considered their most noteworthy characteristic immediately preceding the second operation. Reoperation was performed in a manner basically identical to the ini- TABLE 1. STATUS BEFORE REOPERATION IN 19 PATIENTS WITH FAILURE OF CORONARY SVG No. of Status Patients Reasonably stable, good ventricular contractility 13 Reasonably stable, poor ventricular contractility 1 Preinfarction syndrome 1 Acute myocardial infarction 12 to 24 hours previously 2 Healed dehiscence of sternotomy 1 VOL. 14, NO. 3, SEPTEMBER,

3 ADAM ET AL. tial procedure. This consisted of using total cardiopulmonary bypass, with hemodilution technique maintaining the hematocrit at 30% or better; moderate hypothermia of 30" to 32 C.; intentional electric fibrillation; the use of left ventricular apical venting as necessary; and anastomosis with either continuous 6-0 Ethiflex or interrupted 6-0 silk sutures. Aortic cross-clamping for periods up to 20 minutes was also used. The operating time and the time spent on bypass are both increased in reoperation. Besides the presence of the expected adhesions and the care necessary to insure protection of the functioning vein grafts, there may be a problem in reopening the sternum. The right atrium or right ventricle may be densely adherent to the sternum, leading to severe blood loss from opening these chambers. By prior exposure of the femoral artery and heparinization, cannulation of the femoral artery may be done, if needed, to return all the collected blood encountered at this awkward moment. Such a system was necessary in 3 of our patients. The findings at reoperation were interesting. In some, only light adhesions were present. Occasionally, dense scar had formed, possibly representing organized mediastinal hematoma. Those veins which were patent at the time of reoperation were soft and very similar in appearance to fresh veins. Some were thicker, yet soft and flexible. Closed vein grafts were firm, full-feeling, and rubbery or sometimes were found to be flattened and soft, as if collapsed. The entire length of an occluded vein was not always demonstrable. The sites of anastomosis were usually patent, especially at the veinto-coronary artery end. Restoration of flow in closed grafts was by interposition of a new vein graft into the patent stub of the occluded graft or into a new site on the same artery as before. The choice was dictated by arteriographic and operative findings. Treatment after operation included continuing endotracheal intubation and respirator support overnight, antibiotic coverage, digitalis and diuretic therapy, and antiarrhythmic management as necessary. After oral intake was resumed, aspirin and Persantine were used for their action in minimizing platelet agglutination and associated thrombosis. The course of events in the 19 patients was extremely varied and complex. In order to properly assess and study the problems leading to reoperation, one must consider the status of the original 58 arteries and the initial management of the 38 diseased arteries, then follow the changes which took place with the passage of time and at the time of reoperation (Table 2). Of 12 arteries that had been considered normal at the original operation, 5 demonstrated advancing obstructive disease which led to treatment by vein grafting at the second operation (see Table 2). Eight arteries listed as untreated were recognized as having disease thought to be subcritical or which was treated by internal mammary artery implantation. At the subsequent operation, 7 of these had either demonstrated progression of the obstructive phenomena or were still considered to need a proper revascular- 274 THE ANNALS OF THORACIC SURGERY

4 Reoperation Following Vein Graft Failure TABLE 2. FLOW SHEET OF GRAFT AND ARTERY CHANGES IN 19 REOPERATED PATIENTS Initial Management Preoperative Management Status Just Prior at Second Status of at First to Second Operation Artery Operation Operation If Treated 12 normal arteries No treatment 7 normal 5 critical stenoses 4 vein grafts; 1 vein graft & distal endarterectomy 8 subcritical stenoses No treatment 1 subcritical (< 75%) stenosis 7 critical stenoses 8 vein grafts 38 critical stenoses 35 vein grafts 12 open vein (> 75%) grafts 23 closed vein 21 vein grafts; 2 grafts not treatedarteries too diseased 2 endarterectomies 1 open endarterectomy 1 closed 1 vein graft endarterectomy 1 endarterectomy 1 open vein graft and proximal into distal vein graft endarterectomy ization procedure, and these were treated in 6 instances with vein bypass grafts and in 1 instance by distal endarterectomy with proximal vein bypass graft. Of the 35 arteries initially treated by vein grafting, 12 had remained open and 23 showed vein graft closure. Twenty-one of these were treated by repeat vein grafting and 2 were left untreated due to severely advancing distal arterial disease. Two vessels initially had been endarterectomized; 1 of these had remained open and 1 had closed, and the latter was managed successfully with a vein graft. The 1 vessel originally treated by endarterectomy and proximal vein grafting was open at the time of the second operation. Thus, failures were attributable to advancing arterial disease in untreated arteries, advancing arterial disease in treated arteries, and to vein graft failure occurring primarily. It was noted frequently that in a single patient, vein graft closure had occurred in only 1 of 2 or in 2 of 3 vein grafts. Among the 23 SVG failures, it was frequently noted that the aortic and arterial ends of the vein grafts were still widely patent; and, in most instances, the closure of the SVG could and did occur in such a way as to leave uncompromised the preexisting arterial situation. Although occluded veins VOL. 14, NO. 3, SEPTEMBER,

5 ADAM ET AL. TABLE 3. TYPE OF VEIN GRAFT FAILURE AND RELATED FACTORS IN 19 REOPERATED PATIENTS FIP & Encased Pathological Condition Thrombosis Thrombosis in Scar Obstruction distal to vein graft Severe fibrosis & inflammation around graft Neither of above FIP = fibrous intimal proliferation. usually were obstructed throughout their entire length except at each end, there was some variation in the gross and microscopical appearance of the veins. Two veins were encased in extremely heavy mediastinal scar tissue and were not identifiable within this mass of scar. Of the remaining vessels, 7 appeared to show thrombosis as the major occluding factor and 14 showed various degrees of fibrous intimal proliferative change, usually with some thrombus centrally. Table 3 lists the number of patients having each observed occlusive process, divided into three categories representing associated phenomena observed in the patients, One sees that both distal runoff obstruction and severe inflammatory disease or fibrosis around the graft appear to be related to the formation of severe fibrous intimal proliferative changes. R esu 1 ts The results of reoperation (Table 4) are comparable to the results obtained with the initial Of the 19 patients, 14 showed excellent or improved results and 2 were unchanged; there were 2 operative deaths and 1 late death. Both operative deaths occurred in patients who had had acute infarction 12 to 24 hours previously and who thus would have been difficult patients for an initial operation as well as a reoperation. The late death occurred in a patient with severe congestive heart failure who had improved clinically following triple bypass vein grafting but in whom failure had recurred; restudy had demonstrated closure of 1 of the 3 grafts. Reoperation was followed by sternal dehiscence, continuing congestive heart failure, and death three months postoperatively. A postmortem examination was not obtained, but it was believed that a myocardiopathy had probably been a significant component in the patient s clinical difficulties. The 2 patients not helped by reoperation were both men with early, severe, and rapidly progressive arteriosclerosis. One, aged 34, had had early closure of 2 vein grafts at two months postoperatively. An endarterectomy, initially open, had closed at twenty-one months. Severe progression of his distal arterial disease was shown at restudy, and at reoperation, severe mediastinal fibrosis was observed and reaction about the graft was extreme. 276 THE ANNALS OF THORACIC SURGERY

6 Reoperation Following Vein Graft Failure TABLE 4. RESULTS OF REOPERATION IN 19 PATIENTS WITH FAILURE OF CORONARY SVG No. of Result Patients Remarks Excellent 11 Asymptomatic without medication & working Improved 3 Working, some symptoms, some Unchanged 2 medication Rapid progression of arteriosclerotic heart disease Operative deaths 2 Acute preoperative infarctions Late deaths 1 Congestive heart failure Total 19 After reoperation, mediastinitis and sternal dehiscence occurred, and very early reclosure of his grafts was observed. The other unaided patient was 47 years old. Restudy showed extreme stenosis distal to the left anterior descending coronary artery graft with patency of the right coronary and circumflex grafts. Reoperation with bridging of the distal lesion was done, but a later study showed marked further progression of distal disease and closure of the grafts. Treatment with streptokinase was undertaken; it failed to reopen the grafts but, surprisingly, did partly reopen 2 of the 3 blocked arteries. One wonders if disseminated intravascular coagulation in a mild form might not have been present. Comment When autogenous vein grafts are placed in the arterial tree in human beings, they undergo changes of thickening, or arterialization. This thickening tends to involve all the coats of the vessel but is primarily intimal and chiefly composed of fibrous tissue [5]. Wyatt and Taylor [11] have clearly shown the deleterious effects of trauma, either of a mechanical nature or that produced by 90 minutes of irrigation with saline solution, in producing increased amounts of fibrous intimal proliferation. Overton and Owen [91 have demonstrated that ischemia can produce fibrous intimal proliferation and fibrous medial proliferation as well. Downs [6], in examining 7 patients undergoing femoropopliteal reoperation for stenosis and obstruction of vein grafts, found fibrous intimal proliferation to be the cause of obstruction in 2 of his 7 patients, while in 5 others the cause was listed as valve stenosis and described as being fibrosis of the valve cusps. This would seem to us to be a localized form of the same process of fibrous intimal proliferation. Interestingly, the degree of fibrous intimal proliferation seen in arterialized vein grafts has not frequently been reported as the chief cause of failure except by Johnson and associates [7]. Downs [6] believes this is a more likely primary VOL. 14, NO. 3, SEPTEMBER,

7 ADAM ET AL. cause when the veins are small, and he, as well as others, tends to find a degree of fibrous intimal proliferation together with a final thrombosis as the ultimate cause for obstruction. Vlodaver and Edwards [lo] have described the pathological features of the process of intimal proliferation, and from their study of postmortem specimens and the consistent finding of distal arterial disease, they have concluded that the distal obstruction produces a water hammer effect which aggravates the effect of the arterial pressure to produce unusual degrees of fibrous intimal proliferation and subsequent vein obstruction. If this theory is true, subsequent study should show an increased incidence of fibrous intimal proliferation in hypertensive patients. On theoretical grounds only, it is our belief that injury to the vein wall-chiefly to the intima-is the primary cause of serious intimal proliferative obstructive phenomena and that probably ischemia is the chief type of trauma encountered in coronary surgery patients. We propose that ischemia may be produced by drying and by excessively long soaking of the vein in saline solution prior to use or that it may be produced by scarring around the SVG through interference with the vasa vasorum and the partial failure to develop an adequate blood supply to the vein wall. Distal obstructive disease may also be contributory, but the data seem weak in this regard. In line with these thoughts, we have increasingly tried to avoid trauma in the harvesting of veins for use as grafts. We try to avoid drying the veins, we make every attempt to tie branches in such a way that the closed branch provides no pocket or puckering, and we try to see that valves are not cut in any way when the veins are trimmed for anastomosis. In order to minimize inflammatory changes around the SVG, we treat diagnosed postpericardiotomy syndrome with steroids in a vigorous manner and try to avoid mediastinal hematoma by extremely careful hemostasis and by using minimal dissection upon entering. There is some likelihood that increasing attention should be given to arteries which have less than critical stenosis present at the time of operation. Placing a graft about such a stenosed vessel before the condition become more critical may minimize the number of patients needing to be reoperated upon for advancing arterial changes in untreated subcritically stenotic vessels. Those patients requiring only single and double bypass grafts and with apparently normal remaining vessels will undoubtedly continue to show some degeneration due to advancing arterial disease in the untreated vessels. Reoperation fortunately is available for these people and can prevent possibly disastrous consequences when done promptly. Early restudy of all patients in whom angina recurs is strongly recommended. The use in the postoperative period of antiplatelet-agglutination agents such as aspirin and Persantine may contribute to a diminution in thrombotic failure of vein grafts and may minimize local buildup of platelet and 278 THE ANNALS OF THORACIC SURGERY

8 Reoperation Following Vein Graft Failure fibrous tissue at the site of local vein injury, particularly if used during the period prior to establishment of a completely smooth inner lining of the vein. References 1. Adam, M., Mitchel, B. F., and Lambert, C. J. Immediate revascularization of the heart. Circulation (Suppl. II):73, Adam, M., Mitchel, B. F., Lambert, C. J., and Geisler, G. F. Long-term results with aorta-to-coronary artery bypass vein grafts. Ann. Thorac. Surg. 14:1, Beebe, H. G., Clark, W. F., and DeWeese, J. A. Atherosclerotic change occurring in an autogenous venous arterial graft. Arch. Surg. 101:85, DeWeese, J. A., and Rob, C. G. Autogenous venous bypass grafts five years later. Ann. Surg. 174:3, DeWeese, J. A., Terry, R., Barner, H. B., and Rob, C. G. Autogenous venous femoropopliteal bypass grafts. Surgery 59:28, Downs, A. R. Repair of late vein graft occlusions. Arch. Surg. 103:639, Johnson, W. D., Auer, J. E., and Tector, A. J. Late changes in coronary vein grafts. Am. J. Cardiol. 26:640, Mitchel, B. F., Adam, M., Lambert, C. J., Sungu, U., and Shiekh, S. Ascending aorta-to-coronary artery saphenous vein bypass grafts. J. Thorac. Cardiovase. Surg. 60:457, Overton, J. H., and Owen, E. R. The successful replacement of minute arteries. Surgery 68 : 7 13, Vlodaver, Z., and Edwards, J. E. Pathologic changes in aortic-coronary arterial saphenous vein grafts. Circulation 44:719, Wyatt, A. P., and Taylor, G. W. Vein grafts: Changes in the endothelium of autogenous free vein grafts used as arterial replacements. Br. J. Surg. 53:945, Discussion DR. ROBERT G. CARISON (New York, N.Y.): It takes a brave surgeon to stand up and talk about his bad results. Actually, he had very good results with reoperations. At Cornell University we had one lesson that we would like to share with you. My SVG patency rate at two weeks is 90% in coronary arteries with a lumen of 1.5 mm. or greater. One patient who had patent double vein grafts two weeks postoperatively had routine follow-u angiograms six months later. This asymptomatic patient demonstrated a pseu Ioaneurysm f of the ascending aorta at the site of the aorta-to-vein graft anastomosis, and the left coronary artery was occluded. I excised the pseudoaneurysm. When the aneurysm and the segment of occluded SVG had been removed, the distal portion of that same graft to the anterior descending coronary artery was widely patent. Copious blood flowed retrogradely through this previously occluded SVG. A 1-inch-long vein graft was anastomosed between the ascending aorta and the previously occluded vein graft. At the same time, I bypassed a mild stenosis of the right vein graft. Two weeks after reoperation, the grafts were patent by angiogram. This pseudoaneurysm may have been due to slippery polyethylene sutures used for the anastomosis or to weakening of the aortic wall by the partially occluding clamp. Another possibility was that the angiography catheter may have weakened or pierced the anastomosis site two weeks postoperatively and allowed the pseudoaneurysm to form. VOL. 14, NO. 3, SEPTEMBER,

9 ADAM ET AL. DR. WILLIAM J. KERTH (San Francisco, Calif.): There is no doubt that aortato-coronary artery bypass SVG is a quantum step in coronary artery surgery. Until they are eliminated, however, late occlusions will probably preclude the extension of this procedure to the essentially asymptomatic patient who is at high risk of sudden death from severe coronary atherosclerosis. Information like that presented by the authors may well give clues to the reasons for late closure and lead to its prevention. I have some data on 8 patients who were reoperated upon at Pacific Medical Center in San Francisco following venous aorta-to-coronary artery bypass grafting. When we first began doing this work in February, 1968, we excised the valves in the veins of the first 7 patients, and I think that was the cause of occlusion in 1 patient, who had to be operated upon twice. In the second reoperation a reversed graft was used, and I believe this patient s vein is still open, as she has a normal treadmill test. One cause of proximal stenosis, particularly of a left bypass, is inadequate size of the proximal anastomosis. Initially we did not excise a large button of aortic tissue but only a small triangle. In January of 1970 we began to do arteriograms about two weeks after operation on all patients who had this operation, followed a year later by another bypass arteriogram and coronary arteriogram. Therefore, in all these patients except the first, we have early postoperative studies showing the technical adequacy of both the proximal and the distal anastomosis. One patient, the same one whose bypass was wide open both proximally and distally, developed a very severe proximal stenosis of the vein, down to only a pinpoint opening. The distal vein was wide open on early restudy; the distal anastomosis was fine. There was, however, intense intimal hyperplasia of the aortic anastomosis site eight months postoperatively. For the past two years we have examined control veins. We take a segment of vein at the time of coronary bypass and send it to the pathologist so we have something to compare the graft with if there is subsequent trouble. Also, by analyzing the early and late postoperative studies on patients whose grafts have remained opened, we have found that the vein does tend to shrink and that the left anterior descending SVG generally gets pulled up to some extent. On the right side we believe that the SVG does tend to tent up the distal right coronary artery unless the vein is made redundant at the initial operation. This tenting then may well lead to some funneling or turbulence at the distal anastomosis. We now make, as I indicated earlier, a very large proximal anastomosis, as large as the diameter of the vein or perhaps a little larger. We have seen at least 1 patient with a very intense external fibrous reaction around the proximal site of the aortic anastomosis, so much that the vein was encased in cement and was, I believe, kinked by intense scar. There are other reasons which we are not aware of now for this condition. We hope that with further study this will be a solvable problem. We are currently treating all our patients with aspirin postoperatively to attempt to prevent platelet deposition on the anastomosis. We have abandoned the use of heparin and Coumadin in the postoperative period, as they did not seem to make any difference. We also have a series of more than 20 patients in whom we kept the vein in situ until just before the anastomosis was made, thinking that ischemia may have been the cause for this reaction. The postoperative studies in these patients really are quite comparable as far as the amount of narrowing or shrinking of the veins, so I don t believe that that is the problem. DR. ADAM: As far as this search for the cause of SVG failure, I see that we are not the only ones still baffled. The pseudoaneurysm demonstrated by Dr. Carlson was well managed. Anas- 280 THE ANNALS OF THORACIC SURGERY

10 Reoperation Following Vein Graft Failure tomotic bleeding more often results in early reexploration for suture control or, as in 1 of our patients, in death from hemorrhage. Dr. Kerth s demonstration of excised valves and other local problems tends to strengthen my feeling about vein trauma being the cause of difficulties. With regard to the proximal anastomosis, we use a rather large cobra-head anastomosis with a slit in the aorta, and we think two factors are important in this. One is not to have the incision too long in the aorta, because that stretches the cobra head flat when it opens up; so one has to make it short enough to suit if the vein is small. Second, one can see when it is finished that the full lumen of the vein is available for entry, and we have not had much trouble with proximal anastomoses. We too believe that shrinkage of the vein is likely and have therefore used the loop application of vein with a generous length. We also use aspirin and Persantine postoperatively (for an indeterminate period as yet) and hope this will prevent early failure until the veins develop a smooth inner lining. VOL. 14, NO. 3, SEPTEMBER,