Is the Internal Thoracic Artery the Cond,uit of Choice to Replace a Stenotic Vein Graft? Daniel Navia, MD, Delos M. Cosgrove 111, MD, Bruce W. Lytle, MD, Paul C. Taylor, MD, Patrick M. McCarthy, MD, Robert W. Stewart:, MD, Eliot R. Rosenkranz, MD, and Floyd D. Loop, MD The Cleveland Clinic Foundation, Cleveland, Ohio

Reoperative coronary artery bypass grafting secondary to saphenous vein graft (SVG) stenosis is a mushrooming problem. The internal thoracic artery graft (ITA) pro- vides superior long-term patency, but its flow is limited and may be inadequate to meet large myocardial de- mands. To evaluate the efficacy of the ITA as a replace- ment conduit for a stenotic SVG, 387 consecutive pa- tients undergoing reoperative bypass grafting from 1985 to 1990 with a stenotic SVG to a totally obstructed left anterior descending coronary artery (LAD) were ana- lyzed. The patients were divided into four groups ac- cording to the management of the previously placed SVG. Group I (n = 155) underwent graft replacement with a new SVG. Group I1 (n = 90) received an ITA with the old SVG left intact. In group I11 (n = 37), an ITA was placed to the LAD with an SVG to the diagonal (old graft interrupted). Group IV (n = 104) had an ITA only to the LAD (old graft interrupted). There were 14 deaths (3.6%). Mortality rate was 7.9% for group IV and 2.1% for groups

ecurrent ischemia after myocardial revascularization R is a burgeoning problem, which has resulted in an increasing number of patients undergoing reoperative myocardial revascularization [ 11. Vein graft atherosclero- sis is the most common angiographic indication for reop- erative myocardial revascularization [2]. Management of atherosclerotic stenotic vein grafts is a challenging techni- cal problem because of the potential of atheroembolism [3]. Marshall and colleagues [4] d.emonstrated a high incidence of vein graft atherosclerosis even in patients with angiographically normal-appearing grafts. Grondin

For editorial comment, see page 8. ~~ ~~

and associates [5] demonstrated improved results by ligating old saphenous vein grafts as early as possible during dissection to help avoid perioperative myocardial infarction caused by atheroembolism. The choice of con- duit to replace divided grafts is a subject of controversy.

Internal thoracic artery grafts have a superior long-term patency rate compared with vein grafts and are associated

Presented at the Twenty-ninth Annual Meeting of The Society of Thoracic Surgeons, San Antonio, TX, Jan 25-27, 1993.

Address reprint requests to Dr Cosgrove, The Cleveland Clinic Founda- tion, 9500 Euclid Ave, Cleveland, OH 44195.

I through I11 ( p = 0.01). Multivariate analyses identified advancing age (p = 0.001), [TA only (p = 0.001), and female sex ( p = 0.04) as independent predictors of operative mortality. Evidence of hypoperfusion in the distribution of the LAD wais present in 19 patients, all of whom were in group IV (18.9%). Predictors of hypoper- fusion were moderatelsevere left ventricular function (p = 0.02) and ITA to the LAD with interruption of the old graft (p = 0.0001). Hypoperfusion syndrome was treated with a new SVG to the LAD in 11 patients, and all but 1 survived; 8 were treated with an intraaortic balloon pump, for a 63% mortality ( p = 0.01). We conclude that replacing a stenotic SVG to a totally occluded LAD with an ITA is associated with increased mortality and in- creased incidence of hypoperfusion syndrome. Hypoper- fusion syndrome is best treated with supplemental vein grafting.

(Ann Thorac Surg 1994;57:404)

with improved survival and {freedom from cardiac events [6]. Several authors have advocated the use of the internal thoracic artery in reoperation [7-111; however, blood flow through this arterial graft is less than that through a saphenous vein graft and may not be adequate to meet requirements during times of peak myocardial demands [12-141. Hypoperfusion syndrome is the clinical entity reported secondary to inadequate flow via the internal thoracic artery [15]. Therefore, the question arises as to the best choice of conduit to replace an atherosclerotic saphenous vein graft to a totally obstructed vessel. This study was designed to evaluate the results of various treatment options when dealing with a stenotic vein graft to a totally occluded left anterior descending coronary artery.

Patients and Methods From 1985 through 1990, 2,798 patients underwent iso- lated reoperative myocardial revascularization at The Cleveland Clinic Foundation. Review of these patients identified 387 who underwent reoperation and who had a stenotic vein graft to a totally obstructed left anterior descending coronary artery as the only source of circula- tion to the anterior wall.

Mean patient age was 64.5 +. 9.2 years; 89% were men.

0 1994 by The Society of Thoracic Surgeons 0003-4975/94/$7.00

Ann Thorac Surg 1994;5740-4

NAVIAETAL 41 ITA GRAFT AND REOPERATION

Fig 1. There were four methods of managing these pa- tients. In group l , the atherosclerotic vein graft was divided and replaced with a new saphenous vein graft (SVG) to the left anterior descending coronary artery (LAD); in group 11, the old vein graft was left in place and flow to the LAD was supplemented with a new internal thoracic artery UTA) graft to the LAD; in

group 111, the old SVG was divided and replaced with an 1TA graft to the LAD, and an SVG was also placed to the diagonal corona ry system; and in group IV, the old vein graft was divided and replaced with an ITA graft to the LAD

I I1 I11 IV ‘1 ’I I

I ), 1; ‘ 1

I

N=155 N=90 N=37 N=105

The patients were divided into four groups according to the operative management of the stenotic vein graft to the left anterior descending coronary artery (Fig 1). During this interval, the surgical staff was constant and the therapeutic option was selected by the individual sur- geon. Nitroglycerin patches were used in all patients, but there was no consistent use of calcium-channel blockers. The groups were as follows:

Group I (n = 155): old vein graft replaced by a new vein graft distally

Group I1 (n = 90): old vein graft left intact and supplemented by an internal thoracic artery graft

Group I11 (n = 37): old vein graft interrupted and replaced by internal thoracic artery graft and new vein graft constructed to the diagonal

Group IV (n = 105): old vein graft interrupted and replaced by internal thoracic artery graft

Groups were analyzed for the incidence dE death, myocardial infarction, and hypoperfusion syndrome. Hy- poperfusion syndrome was defined by clinical evidence of ischemia, which was characterized by a sudden onset of hypotension, low cardiac output, increased filling pres- sure, decreased contractility, and ventricular arrhythmias.

A x2 or Fisher’s exact test was used to evaluate univari- ate relationships of preoperative and intraoperative risk factors with mortality, myocardial infarction, and hypo- perfusion. The relationship between age and outcome was examined with an unpaired t test. Factors found to be univariately related (p < 0.20) were included in a stepwise logistic regression analysis to determine those factors independently associated with the outcome.

Results There were 14 deaths, for an overall mortality rate of 3.6%. Significant morbidity included myocardial infarc- tion (7.8%), respiratory insufficiency (6.2%), renal failure (2.8%), stroke (2.8%), and wound complication (0.8%). When operative mortality was analyzed by group, mor- tality was 3.8%, 0%, 0%, and 7.9% in groups I through IV,

respectively. Patients in group IV who had only an internal thoracic artery graft to the left anterior descend- ing coronary artery had a 7.9% mortality compared with a 2.1% mortality for patients in groups I through I11 who had a vein graft to the left anterior descending coronary artery system (p < 0.001).

The small size of the four groups and the relatively small number of deaths and events made further statisti- cal analysis impractical. Multivariate analysis of the entire group was employed to identify risk factors for deaths and events.

A multivariate analysis identified advancing age ( p = 0.001), internal thoracic artery graft only (p = 0.001), and female sex (p = 0.04) as the independent predictors of operative mortality.

There was no significant difference in the incidence of myocardial infarction among treatment options. Myocar- dial infarction rate was 9.7%, 1.1%, 10.9%, and 9.9% for groups I through IV, respectively (not significant). No univariate predictors of myocardial infarction were iden- tified.

Evidence of hypoperfusion syndrome in the distribu- tion of the left anterior descending coronary artery was present in 19 patients, all of whom were in group IV (18.9%). All episodes of hypoperfusion syndrome oc- curred late in the series (between 1987 and 1990). Hemo- dynamic deterioration occurred in the operating room in 14 patients and in the surgical intensive care unit within 2 hours of arrival in 5 patients. Seventeen patients were men. Multivariate analysis identified internal thoracic artery graft only (group IV) (p = 0.0001) and moderate/ severe left ventricular function (p = 0.02) as predictors of hypoperfusion syndrome.

Treatment of hypoperfusion syndrome varied. Six pa- tients underwent reoperation with an additional saphe- nous vein graft placed distal to the internal thoracic artery graft. Five patients had reconstruction of the previously divided saphenous vein graft to the left anterior descend- ing coronary artery. Of the 11 patients in these two groups, 10 survived. Eight patients were treated with an intraaortic balloon pump, and 5 (63%) died. Patients who had construction of an additional saphenous vein graft to

42 NAVIAETAL ITA GRAFT AND REOPERATION

Ann Thorac Surg 1994;574&4

the left anterior descending coronary artery had signifi- cantly better survival than patients who were treated with an intraaortic balloon pump ( p = 0.01.).

Autopsies were obtained in 9 of the 14 patients who died (64%). Eight of the 9 patients had evidence of an acute myocardial infarction. No evidence of atheroemboli was found in any of the autopsies. All 5 patients who died with only an internal thoracic artery graft to their left anterior descending coronary artery had an acute antero- lateral myocardial infarction.

Comment Recurrent ischemia after myocardial revascularization is common, resulting in an increasing incidence of reopera- tive coronary artery bypass grafting. It is projected that by 1995, 50,000 reoperative coronary revascularization proce- dures will be performed annually in the United States [l]. This exponentially increasing number of reoperations results from a number of factors including an increasing number of patients at risk, improved survival in patients who have been revascularized, the progressive nature of the atherosclerotic disease process in the native circula- tion, and the accelerated atherosclerotic disease process in saphenous vein grafts. In patients undergoing their first reoperation at The Cleveland Clinic Foundation, vein graft atherosclerosis was an angiographic indication in 85% of the patients [Z].

Frequently, patients requiring reoperations have multi- ple sources for their myocardial blood supply including atherosclerotic and stenotic vein grafts. The risk of embo- lization from these grafts is ever present and carries the potential for catastrophic results. Telchniques most likely to help avoid this complication are dissection using a no-touch technique for the grafts and complete interrup- tion of the vein graft as early as possible. If the grafts are interrupted, the most appropriate conduit to replace the vein graft must be carefully considered. When using the internal thoracic artery, the potential for imbalance be- tween supply and demand exists.

Several findings in this analysis point out the danger associated with replacement of patent vein grafts with an internal thoracic artery graft. It is of interest that all episodes of hypoperfusion syndrome occurred in the latter part of the series (between 1987 and 1990) when the surgical approach was altered to include division of old vein grafts as early as possible to avoid atheroembolism and the use of internal thoracic artery grafts to enhance long-term patency. Consistent findings in this syndrome include technically satisfactory internal thoracic artery anastomosis, division of a vein graft without a vein graft replacement, and sudden onset of hypoperfusion. The absence of this syndrome in patients whose saphenous vein grafts were left in place further points to the flow limitations of the internal thoracic artery as the cause of this clinical entity.

Further proof of this relationship is the dramatic change in clinical outcome in patients who received additional vein grafts placed distal to the internal thoracic artery for treatment of their hypoperfusion syndrome. Although

the number of patients identified with this problem is small, the high incidence recorded in group IV and its reversal with an additional vein graft are convincing evidence.

The imbalance in the supply/demand ratio may occur secondary to reduced supply or large demands. When the internal thoracic artery is small, the supply may be inad- equate. The myocardial blood supply requirement may surpass even normal or large internal thoracic arteries if the myocardium is hypertrophied, left ventricle is dilated, or the entire septa1 blood supply arises from the left anterior descending coronary artery because of complete occlusion of the right cororiary artery.

Although the incidence of perioperative deaths and hypoperfusion syndrome was increased in patients in group IV, who had a totally occluded left anterior de- scending coronary artery and a patent vein graft that was interrupted and replaced by an internal thoracic artery graft, the majority of the patients tolerated this procedure well. The difference in results may be a function of the amount of myocardium supplied by the left anterior descending coronary artery. Unfortunately, we do not currently have a method to quantitate the amount of myocardium supplied by am individual artery or the oxygen demands of that tissue. If such a method were available, it would be a vallunble guide in the application of the internal thoracic artery graft in this circumstance.

Recognizing the potential for inadequate flow via the internal thoracic artery to meet myocardial blood flow demands requires consideration of alternative ap- proaches. These include replacing atherosclerotic vein grafts with new vein grafts. This has the advantage of reducing the potential for atheroembolism while provid- ing adequate flow. The second alternative involves divid- ing the previously placed vein graft and placing a new vein graft to the diagonal or left anterior descending coronary artery to supplement the internal thoracic artery flow. Tlik has the advantage of providing adequate flow without the potential for atheroembolism but has the potential disadvantage of competitive flow resulting in a "string sign." The third alterinative is to leave the old vein graft in place and augment bllood flow to the left anterior descending coronary artery with an internal thoracic artery graft. This has the theoretical advantage of provid- ing adequate blood flow but potential disadvantages of increased incidence of atheroembolism and competitive blood flow causing atrophy of the internal thoracic artery graft. Although the latter approach is of concern because of the potential for atheroemboli, Galbut and associates [16] have reported using tlhis approach without noting an increase in myocardial infarctions. The fact that we have not experienced an increased incidence of myocardial infarction or death in this group suggests that this may be a viable alternative.

When evaluating the best approach to take in reopera- tion, we must consider mulbiple factors. These include life expectancy, potential for atl~eroembolism, and the long- term significance of competitive flow and its effect on the internal thoracic artery. When life expectancy is great, the use of arterial grafts become imperative. If life expectancy

Ann Thorac Surg 1994;574C-4

NAVIAETAL 43 ITA GRAFT AND REOPERATION

is short, the advantages of the internal thoracic artery are diminished.

There is little doubt that atheroembolization is a reality and can result in a major myocardial infarction. Even the slightest manipulation may produce this event, and noth- ing can reverse it. This puts a premium on preventing this complication. It, therefore, seems reasonable to interrupt the graft as early as possible. The fact that the periopera- tive myocardial infarction incidence in this group was only 1.1% and there were no deaths may be the result of good surgical judgment for interrupting grafts with a high potential for embolization. One should not conclude from this observation that this management technique is al- ways safe or necessarily the preferred management alter- native. Retrograde cardioplegia has made this surgical approach much safer.

The key unsettled issue revolves around competitive flow and its effect on the internal thoracic artery graft. It has been reported that the internal thoracic artery be- comes atretic, resulting in a “string sign” angiographi- cally, when the internal thoracic artery graft flow is exceeded by another source [17]. This unquestionably occurs, but the clinical implications of this phenomenon are not clear. Increasing evidence suggests that an atretic- appearing internal thoracic artery graft may be able to dilate and function normally if competitive flow is de- creased and demand increased [18-201. If this is true, the use of a vein graft and an internal thoracic artery graft to the left anterior descending system after occluding the old vein graft may be the best alternative. Replacing the vein graft protects against atheroembolism and hypoperfusion syndrome, and the internal thoracic artery graft enhances the potential for long-term graft patency.

The policy of interrupting the old vein graft to a totally obstructed left anterior descending coronary artery and replacing it with only an internal thoracic artery graft is associated with increased incidence of death and hypo- perfusion syndrome.

References 1. Cosgrove DM, Loop FD, Lytle BW, et al. Predictors of

reoperation after myocardial revascularization. J Thorac Car- diovasc Surg 1986;92:811-21.

2. Lytle BW, Loop FD, Cosgrove DM, et al. Fifteen hundred coronary reoperations: results and determinants of early and late survival. J Thorac Cardiovasc Surg 1987;93:847-59.

3. Keon WJ, Heggtveit HA, Leduc J. Perioperative myocardial

DISCUSSION

DR ROBERT M. LUST (Greenville, NC): I am both privileged and saddened to have this opportunity today. My chief, Dr Chitwood, was originally invited to discuss this report, a task he was looking forward to immensely. Unfortunately, we received word Sunday evening that his father had passed away. Knowing that Dr Chitwood and I collaborate, Dr Cosgrove kindly invited me to discuss this report on his behalf.

This excellent study of reoperative grafting strategies demon- strated an overall 4.9% incidence of hypoperfusion syndrome,

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

infarction caused by atheroembolism. J Thorac Cardiovasc Surg 1982;84:849-55. Marshall WG Jr, Saffitz J, Kouchoukos NT. Management during reoperation of aortocoronary saphenous vein grafts with minimal atherosclerosis by angiography. Ann Thorac Surg 1986;42:16>7. Grondin CM, Pomar JL, Hebert Y, et al. Reoperation in patients with patent atherosclerotic coronary vein grafts. A different approach to a different disease. J Thorac Cardiovasc Surg 1984;87379-85. Loop FD, Lytle BW, Cosgrove DM, et al. Influence of the internal-mammary-artery graft on 10-year survival and other cardiac events. N Engl J Med 1986;314:1-6. Loop FD. A 20-year experience in coronary artery reopera- tion. Eur Heart J 1989;10:78-84. Cameron A, Kemp HG Jr, Green GE. Reoperation for coro- nary artery disease. Ten years of clinical follow-up. Circula- tion 1988;78:158-62. Galbut DL, Traad EA, Dorman MJ, et al. Seventeen-year experience with bilateral internal mammary artery grafts. Ann Thorac Surg 1990;49:195-201. Green GE. Reoperations and the internal mammary artery. Adv Cardiol 1988;36:158-61. Cameron A, Green GE, Kemp HG Jr. Role of internal mammary artery in reoperations for coronary artery disease. Adv Cardiol 1988;36:84-9. Flemma RJ, Singh HM, Tector A], Lepley D Jr, Frazier BL. Comparative hemodynamic properties of vein and mammary artery in coronary bypass operations. Ann Thorac Surg 1975;20:61%72. Hamby RI, Aintablain A, Wisoff BG, Hartstein ML. Compar- ative study of the postoperative flow in the saphenous vein and internal mammary artery bypass grafts. Am Heart J

I + . - 1977;93:30&15. Singh H. Flemma R1. Tector AT. Leulev D Tr. Walker IA. Di&t myocardial r&ascularization.’ Determinants in the choice of *vein graft or internal mammary artery. Arch Surg 1973; 107~699-703. Jones EL, Lattouf OM, Weintraub WS. Catastrophic conse- quences of internal mammary artery hypoperfusion. J Thorac Cardiovasc Surg 1989;98:902-7. Galbut DL, Traad EA, Dorman MJ, et al. Bilateral internal mammary artery grafts in reoperative and primary coronary bypass surgery. Ann Thorac Surg 1991;52:2&8. Barner HB. Double internal mammary-coronary artery by- pass. Arch Surg 1974;109:627-30. Dincer B, Barner HB. The ”occluded” internal mammary artery graft. Restoration of patency after apparent occlusion associated with progression of coronary disease. J Thorac Cardiovasc Surg 1983;85:31%20. Aris A, Borras X, Ramio J. Patency of internal mammary artery grafts in no flow situations. J Thorac Cardiovasc Surg 1987;93:62-4. Kitamura S, Kawachi K, Seki T, Sawabata N, Morita R, Kawata T. Angiographic demonstration of no-flow anatomi- cal patency of internal thoracic-coronary artery bypass grafts. Ann Thorac Surg 1992;53:15&9.

which was concentrated entirely in the internal thoracic artery- only group, so that nearly 1 in 5 patients in this subset experi- enced hypoperfusion. Mortality was more than twice as high in this group. When hypoperfusion was treated immediately with supplemental vein grafting, patient outcome was good, but when hypoperfusion syndrome was managed with balloon pumping, the mortality was well in excess of 50%.

First, the data clearly highlight how poorly we continue to understand the pathophysiology of perioperative ischemia. For

44 NAVIAETAL ITA GRAFT AND REOPERATION

Ann Thorac Surg 1994;574M

example, in group I, in which only isolated vein grafts were used, the infarct rate was identical to when internal thoracic artery grafts were used, yet hypoperfusion syndrome was not ob- served. In those patients who received emergent supplemental grafts, it is interesting to wonder whether the patency of the internal thoracic artery was confirmed when the supplemental graft was placed. Had the implantation flow been maintained? In other words, was the problem truly an anatomic flow capacity or was there an acute vasoactive limitation?

Second, how do these results compare with results with isolated internal thoracic artery grafts to totally occluded left anterior descending coronary arteries at the time of first opera- tion? If these incidences are different, then the problem is not so much internal thoracic artery flow capacity per se, but its capacity only as a replacement graft in the setting of reoperation. If such is the case, this may be a diminishing problem, because internal thoracic artery use at first operation is greatly increased over 5 to 7 years ago, and the availability of the internal thoracic artery at the time of reoperation is almost certainly decreasing.

Third, it must be reemphasized that this is an acute phenom- enon. There is little evidence to suggest chronic underperfusion to totally occluded myocardial regions revascularized with inter- nal thoracic artery grafts. In fact, significant exercise tolerance would suggest not only adequate flow but significant recruitable flow reserve. Therefore, it would seem that strategies that could successfully enhance internal thoracic artery flow perioperatively would be useful and perhaps preferable to extra conduit use to construct supplemental and possibly redundant grafts. Also, the mortality associated with hypoperfusion and the success with supplemental grafting highlights the necessity for improved real-time ability to monitor adequacy of perfusion and progres- sion of ischemia intraoperatively.

Fourth, there remains the possibility that high flow from the vein graft, particularly inserted distal to the internal thoracic artery anastomosis, will promote closure of the internal thoracic artery, obviating its clear patency advantages. Experimental data from our group suggest that internal thoracic artery patency might be retained, at least in competition with residual native flow. Even if early internal thoracic artery closure recurs, Navia and associates cite several references in their report with anec- dotal experience that would suggest potential recanalization of the internal thoracic artery with progression of native disease. However encouraging, these reports represent a combined ob- servation of fewer than 10 patients total, certainly a limited basis upon which to project the ultimate fate of these competing grafts. Additional clinical and experimental studies are needed.

Again, I congratulate Navia and associates on a fine study. I thank them for the privilege of reviewing the manuscript and discussing their results.

DR ALFRED J. TECTOR (Milwaukee, WI): I would like to compliment Dr Navia on an excellent presentation of a very dangerous problem that confronts cardiovascular surgeons when attempting reoperations. I think the best treatment for any problem is its prevention, and the way to prevent this problem is to avoid placing a saphenous vein graft in a critical left anterior descending coronary artery the first time! Some of the options for approaching this problem include using another vein graft, anastomosing a vein in conjunction with an internal thoracic artery, leaving the old vein graft in, and placing an internal thoracic artery. In some cases, we will replace the vein graft with

an internal thoracic artery and ligate the old vein graft. If there is any evidence of hypoperfusion syndrome, we just remove the ligature. This has some risk of em.bolization from the vein graft just like leaving in the old vein graft does. So none of the treatments are perfect.

The other thing that is very important in this problem is very early recognition of hypoperfusion syndrome. Intraaortic balloon pumps and vasotonic drugs do mot correct the problem. The only way you correct the problem is to i:evascularize the ischemic area of myocardium. Several things that have helped us in detecting this immediately after we remove the aortic clamp are electrocar- diographic changes and also wall motion abnormalities detected on echocardiography. As we do more and more reoperations and we use transesophageal echocardiography more, this becomes probably the most helpful adjunct in early detection of this serious problem.

DR DIMITRI NOVITZKY (Tampa, FL): Have you carefully reviewed the coronary angiograms comparing the different groups and found difference in thle diagonal system for stenosis of 50% or less? Sometimes the coronary angiogram is misleading and the narrowing seems to be significant. However, the hemo- dynamic impact can be significant and impair the flow. Maybe the whole issue is the incomplete revascularization of the left ventricle. Actually, implantation of a vein graft may result in complete revascularization rath.er than an insufficient blood sup- ply using only one graft.

DR DENTON A. COOLEY (Houston, TX): That was a very provocative report, Dr Navia. I must say for myself I would be inclined to put in a new vein graft in the stenotic cases and do an internal thoracic artery bypass distally or proximally as well, accepting the criticism of competitive flow. I am more inclined to go for “belt and braces,” as they say in England.

DR NAVIA: Answering the last question, we did check all the angiograms, and the stenosis on the vein graft was more than 30%. We prefer the graft to the ].eft anterior descending artery, and we think that the incidence of complete revascularization was really high.

I think the most important message of this study is that you can safely use one of these three conduit configurations, but defi- nitely I would not recommend use of the fourth one. Dividing the old vein graft and replacing it with an internal thoracic artery is associated with a high morbidity and mortality.

What is the best surgical approach to deal with this problem? If the long-term patency is not the issue, for example, in elderly patients, I would recommend dividing the old vein graft and placing a new vein graft. This (eliminates potential problems with atheroembolism, and the flows will be adequate. In young patients, less than 60 years old, in whom use of the arterial graft is advantageous, dividing the mold vein graft and placing an internal thoracic artery graft to the left anterior descending artery and another vein graft to the diagonal artery to provide supple- mental flow to the anterior wall is an attractive approach. The potential problem with atheroernbolism is reduced by dividing the old vein graft, and the additional vein graft ensures adequate flow. We did not have an increased incidence of myocardial infarction when the old vein graft was left intact and we added an internal thoracic artery graft. We do not know the fate of the graft long-term. This study is currentl~y underway.