in Endarteredomized Coronary Arteries

Transcription

1 Patency and Flow Response in Endarteredomized Coronary Arteries M. Laxman Kamath, M.D., Donald H. Schmidt, M.D., Pablo M. Pedraza, M.D., Fred M. Blau, M.S., A. Sampathkumar, M.D., Linda L. Grzelak, B.S., and W. Dudley Johnson, M.D. ABSTRACT Ninety patients, operated on from May, 1978, through June, 1979, underwent coronary endarterectomy and early recatheterization. Patency of grafts to endarterectomized arteries was 103 of 118 (87.3%) and patency of conventional vein grafts in the same patients was 217 of 233 (93.1%). Myocardial blood flow using xenon 133 washout, at rest and with isoproterenol-induced stress, was measured in 7 normal coronary arteries, 28 conventional saphenous vein grafts, and 33 saphenous vein grafts to endarterectomized coronary arteries. The increase in myocardial blood flow, from rest to isoproterenolinduced stress, was comparable for the three groups. The endarterectomized group was divided further by separating out the 10 patients with heavy scarring or residual disease. The remaining patients had a flow response identical to those with conventional saphenous vein grafts. The rate of perioperative infarction in patients receiving endarterectomy was 3 of 113 (2.6%), as measured by appearance of new persistent Q waves on the serial postoperative electrocardiogram. Positive pyrophosphate scans were noted in 13 of 105 (12.4%) patients. It is concluded that, in the early stages at least, grafts to endarterectomized coronary arteries stay open and perfuse the myocardium as well as conventional saphenous vein grafts unless the myocardium is heavily scarred or unless residual disease remains. It is well established that successful coronary bypass operation relieves angina pectoris [ll and improves left ventricular function [2, 31. The conditions, if any, under which bypass operation prolongs life, are part of a major medical From the University of Wisconsin-Mount Sinai Medical Center, Milwaukee, WI. Presented at the Sixteenth Annual Meeting of The Society of Thoracic Surgeons, Jan 21-23, 1980, Atlanta, GA. Address reprint requests to Dr. Kamath, 3112 W Highland Blvd, Milwaukee, WI controversy of our time. We believe that the preponderance of available data favors surgical treatment for two- or three-vessel disease. When the disease is segmental, surgical bypass is simple and straightforward, the graft being placed into a relatively undiseased section of artery, distal to all of the disease. If, however, the disease is diffuse and no section of undiseased artery can be demonstrated, conventional bypass can be difficult or impossible. In these diffusely diseased vessels, with total or partial occlusion of the lumen, we carry out endarterectomy both proximally and distally to remove the majority of diseased intima. Then the vein graft is routinely attached to the open artery. We believe this is the only practical way to handle these "inoperable" coronary arteries. During the last two years, about 30% of our patients received at least l coronary endarterectomy. Therefore, we can offer here short-term results only. The initial questions about coronary endarterecto my include the following: Is it technically feasible? In our experience, yes, in almost all patients. This report describes our technique in some detail. Do the grafts remain patent? This report demonstrates that early patency of grafts to endarterectomized arteries is comparable to that of conventional saphenous vein grafts in the same patients. Do endarterectomized vessels perfuse myocardium effectively? Using xenon 133 washout, we measured regional myocardial perfusion at rest and with isoproterenolinduced stress. It appears that endarterectomized vessels perfuse the myocardium as well as nonendarterectomized vessels, unless the endarterectomy is incomplete or the myocardium is heavily scarred by The Society of Thoracic Surgeons

2 29 Kamath et al: Endarterectomized Coronary Arteries Materials and Methods From May 1, 1978, through June 30, 1979, 438 patients underwent revascularization for coronary artery disease at Mount Sinai Medical Center, Milwaukee. There were 23 (5.25%) hospital deaths. Of the 438 patients, 122 (27.85%) had coronary endarterectomy. There were 7 hospital deaths (5.7%) in this group. Of the 115 hospital survivors, 90 (78.3%) were recatheterized. The other 25 refused recatheterization for personal reasons. The study group consisted of all 122 patients undergoing endarterectomy during the 14-month period. Sixty-seven other patients, chosen on the basis of availability of facilities and personnel, were studied with the xenon 133 washout technique during rest and isoproterenol-induced stress. Group 1 had 7 patients suspected to have coronary artery disease. All coronary arteries appeared normal on catheterization, and xenon flow studies were made of 1 coronary artery. Group 2 consisted of 28 patients who underwent myocardial revascularization but not coronary endarterectomy. Recatheterization and xenon flow studies of 1 graft were done. Group 3 had 32 patients who underwent myocardial revascularization with coronary endarterectomy. Recatheterization with xenon flow studies of 1 graft to an endarterectomized artery was done. Group 3 partially overlaps the study group of 122 patients. Our endarterectomy technique is simple mechanical removal of diseased intima. A longitudinal arteriotomy is made dividing the adventitia and media. Usually, the plane between the media and intima is easily identifiable with blunt dissection (Fig 1). Once the core of atheroma is dissected, gentle traction is applied to the distal end of the core, and countertraction is applied to the media distally (Fig 2). Extraction of the proximal core is done in the same manner. A clean, tapered end with multiple side branches from secondary and tertiary coronary arteries is usually seen (Fig 3). When atheroma breaks in the center or when the end of the atheroma is not clean, the arteriotomy is extended further down into the distal part of the vessel or a separate distal arteriotomy is used to complete the endarterectomy. The procedure is repeated until a satisfactory and com- plete endarterectomy is accomplished, or until the artery is opened for 6 to 8 cm (Fig 4). A vein is attached to the full length of the arteriotomy and connected to the aorta (Fig 5). Endarterectomy is particularly suited to revascularization of the septum. Figure 6 shows a core of atheroma emerging from the septa1 ostium and proximal left anterior descending coronary artery simultaneously. The measurement of regional myocardial perfusion from the washout of xenon 133 has been described previously 14, 51. In brief, following diagnostic arteriography, a catheter is positioned at the origin of the coronary artery or bypass graft to be studied. Ten to 12 mci of xenon 133 is injected into the vessel. Xenon diffuses into the myocardium very rapidly, and the disappearance of radioactivity from the myocardium is a measure of capillary perfusion. The myocardial washout rate of xenon 133 is monitored with a multiple-crystal scintillation camera. Capillary blood flow (milliliters per minute per 100 gm of tissue) is calculated using the Kety-Schmidt formula (Fig 7). Following the resting study, isoproterenol is infused to produce an increase of approximately 50% in heart rate, and the flow study is repeated. In the final analysis, a diagram of the coronary artery under study is superimposed on the computer-output matrix of the flow, and the regional perfusion is calculated for the area of myocardium perfused by the artery (Fig 8). Results Patency Since the internal mammary artery is used only rarely with endarterectomy, we have analyzed patency of vein grafts only. These recatheterizations were all performed during the surgical hospitalization, 8 to 14 days postoperatively. Table 1 compares the patency of grafts to endarterectomized arteries with conventional vein grafts. Patency rates are similar, 87.2% for endarterectomy and 93.1% for bypasses to relatively undiseased distal coronary arteries. Serial Electrocardiography and Radionuclide Scan Table 2 shows the results of serial 12-lead electrocardiography. Three patients exhibited new, significant, persistent Q waves indicative of

3 30 The Annals of Thoracic Surgery Vol 31 No 1 January 1981 Fig 1. Core of atheroma separated from media and adventitia. Fig 2. Traction on the core and countertraction on the media. Fig 3. Core of atheroma removed from the artery. Fig 4. Arteriotomy area following the endarterectomy.

4 31 Kamath et al: Endarterectomized Coronary Arteries Fig 5. Vein anastomosis to a long segment of the artery. perioperative transmura1 myocardia1 infarction- One of these 3 did not manifest the Q waves until the tenth postoperative day. The Minnesota Code [6] 1-1 or 1-2 was used to define significant Q waves Results of radionuclide scanning are demonstrated in Table 2 also. Positive scans were either 3+ or 4+ in any view, or 2+ in all three views of a discrete area. Thirteen of 105 scans (12.4%) were positive. ACTMTY 3o Fig 6. Simultaneous removal of atheromatous core from proximal left anterior descending coronary artery and Myocardial Perfusion Table 3 shows the response of heart rate and systolic blood pressure in the three groups to isoproterenol-induced stress. Systolic blood pressure changed very little in any group. Change in heart rate was higher in Group 1 and similar in Groups 2 and 3. In general, the amount of stress was slightly higher for Group 1 and essentially equal for Groups 2 and 3. Table 4 shows the response of regional myocardial blood flow to isoproterenol-induced stress. Group 1 showed the greatest change (79%) with Group 2 slightly higher (56%) than Group 3 (43%). Table 5 shows flow response of Group 3 by presence or absence of distal residual disease and scarring. In general, the normal coronary arteries underwent higher stress and had a higher flow response than grafted arteries. Conventional grafts and grafts TIME (Sec.) Fig 7. Xenon washout calculated by the Kety-Schmidt formula: flow (mlll00gmlmin = K x (Alp) x 100, where K = slope, A = partition coefficient (0.72), and p = specific gravity of the myocardium (I.05). (CPS = counts per second.) to completely endarterectomized arteries had the same stress and the same flow response. The flow response of grafts to incomplete endarterectomized arteries or heavily scarred muscle was lowest. Comment There has been considerable controversy about the benefit of coronary operation in those

5 32 The Annals of Thoracic Surgery Vol 31 No 1 January 1981 REST / \\ ISOPROTERENOL S I\ \&* 16s 12l U >/ hF12kS FLOW 63 ml/min.100g HEARTRATE 90 F L 0 W 134 ml/min HEARTRATE 125 Fig 8. Example of computer printout of xenon flow studies for the right coronary artery during rest and isoproterenol-induced stress. patients who have severely diffuse coronary artery disease. Often, bypass is not considered for individual arteries, and occasionally, operation is not done because of diffuse disease. It has long been our impression that no matter how diffuse the coronary artery disease, patients can benefit from operation as long as the ventricular contraction is reasonably good. We have used endarterectomy routinely in recent years for diffusely diseased vessels. This study confirms the findings of others that early patency of grafts to endarterectomized vessels equals that of conventional vein grafts [7-91. The long-term patency of these grafts is unknown at this time. Although grafts remain open with endarterectomy, the technique has been criticized Table I. Early Patency of Grafts to Endarterectomized Coronary Arteries versus Conventional Spahenous Vein Bypass Graftsa Group Total Closed Patency (Yo) Conventional b 93.1 Endarterectomized " 87.2 RCA LAD Circ "Ninety patients receiving 351 vein grafts. bp = RCA = right coronary artery; LAD = left anterior descending coronary artery; Circ = circumflex coronay artery. because of the "snowplow" phenomenon which is said to close off small side branches and prevent effective perfusion [lo]. In order to assess the validity of this concept, we measured regional myocardial perfusion using xenon 133. Decreased response to isoproterenol-induced stress, seen in some patients, is related to residual scar tissue in the myocardial wall and to the amount of residual disease in the distal coronary artery, not to endarterectomy per se. Coronary endarterectomy also is said to be associated with a high incidence of perioperative myocardial infarction. Our incidence was 2.7% by electrocardiographic evidence and 12.4% with radionuclide scan. If the snowplow effect were important, one would anticipate a higher infarction rate in these patients, especially in the areas perfused by endarterectomized vessels. The surgical (30-day) mortality of 5.7% in patients undergoing endarterectomy was virtually identical to the 5.1% mortality in patients with conventional grafts only, operated on during the same time period. A review of the 7 hospital deaths shows nothing unusual. Two patients died on the operating table of severe left ventricular failure. One died on the fourth postoperative day. This patient had had a preoperative impending infarction, a postoperative myocardial infarction, and a nonocclusive mesenteric thrombosis. Another patient died of respiratory dysfunction of unknown etiology on postoperative day 7. One died on postoperative day 9 of left ventricular failure and extension of a preoperative myocardial infarction. One died of a cerebrovascular accident and multiple organ failure on postoperative day

6 33 Kamath et al: Endarterectomized Coronary Arteries Table 2. Results of Serial Electrocardiography and Technetium 99m-labeled Pyrophosphate Myocardial Scanning Postoperatively Patient Status Electrocardiography Scanning Records Technically Total Available New Q Adequate Positive Patients for Review Waves Scan Done Scana Refused recatheterization /23 (4.3%) 20 4/20 (20%) Recatheterized /86 (2.3%) 82 9/82 (llo/o) Died in hospital Total (2.6%) (12.4%) apositive scan defined as 4+ or 3+ in any view, or 2+ in all three views of a discrete area. Table 3. Response of Heart Rate and Systolic Blood Pressure to lsoproterenol-induced Stress" Variable Group1 Group2 Group3 (n = 7) (n = 28) (n = 33) Heart rate (beatdmin) Rest Stress Change f SD 47 f f 9 34? 11 Systolic blood pressure (mm Hg) Rest Stress us Change f SD 4 f 11 2f8 5 f 12 ;'n = number of vessels studied. SD = standard deviation. Table 4. Flow Response to lsoproterenolinduced Group 1 Group 2 Group 3 Variable (n = 7) (n = 28) (n = 33) Rest Stress Change f SD 63 f k f 24 aregional myocardial perfusion (mumid100 gm of tissue). 'In = number of vessels studied. SD = standard deviation. 11. The seventh patient died of stroke, renal failure, and sepsis on postoperative day 27. An overall surgical mortality of 5.3% may seem somewhat high by today's standards until one considers that this mortality includes all patients operated on for coronary artery disease including those with severe left ventricular dysfunction, left ventricular aneurysmectomy, repeat coronary artery operation, combined proce- Table 5. Flow Response in Group 3 as Shown by Scarring and Residual Distal Diseasea Result Rest Improtereno1 Change" Without scar and with com f 23" plete endarterectomy (n = 22) With scar or incomplete f 20c endarterectomy (n = 10) aall values mumid100 gm of tissue. bchange f standard deviation. 'p < dures, and emergency procedures. No patient has been excluded for any reason. The details of our surgical technique contribute much to a successful endarterectomy. Patience and persistence are two essential elements in the surgical treatment of diffuse coronary artery disease. Frequently, the atheromatous core emerges intact. When this occurs, less than a minute is added to the total operating time. Often, however, the core breaks off every few millimeters. Then it is necessary to extend the arteriotomy as much as 8 cm in order to remove a major portion of diseased intima. One endarterectomy and the subsequent 16 cm suture line can take 45 minutes, occasionally even longer. Two or 3 difficult endarterectomies plus 3 or 4 conventional grafts add up to long pump runs and much ischemia time. Under these conditions, all details become important, but myocardial preservation is crucial. We use intermittent ischemic arrest with 34 C total-body hypothermia. Hematocrit is kept in the range of 20 to 25%, mean arterial blood pressure is kept

7 34 The Annals of Thoracic Surgery Vol 31 No 1 January 1983 at greater than 60 torr, preferably 70, and all other variables are maintained in the physiological range. If an anastomosis is to take longer than 20 minutes, the aortic clamp is removed every 15 minutes for 5 minutes of reperfusion. Total clamp time averages 100 minutes. Occasionally, three hours of clamp time are necessary. Pump runs of five hours are common. We have little experience with cardioplegia, but do know that our incidence of perioperative myocardial infarction and CPK-MB isoenzyme rise are comparable to the best reported series with cardioplegia. There has been some speculation that the cardioplegic technique may offer inadequate protection where the diffuseness of disease prevents the cardioplegic solution from cooling muscle adequately. The results of the present study lead us to the following conclusions: Early patency of vein grafts to endarterectomized vessels is comparable to conventional vein grafts. Flow response of grafts to endarterectomized vessels is equal to conventional grafts provided there is no residual distal disease and the myocardium is not heavily scarred. The rate of perioperative infarction is not increased with endarterectomy. Within the limits of this study, coronary endarterectomy is shown to be an effective method for treating diffuse coronary artery disease. References Geha AS, Baue AE, Krone RJ, et al: Surgical treatment of unstable angina by saphenous vein and internal mammary artery bypass grafting. J Thorac Cardiovasc Surg 71:348, 1976 Kolibash AJ, Goodenow JS, Bush CA, et al: Improvement of myocardial perfusion and left ventricular function after coronary artery bypass grafting in patients with unstable angina. Circulation 59:66, 1979 Hellman CK, Kamath ML, Schmidt DH, et al: Improvement in left ventricular function after myocardial revascularization: assessment by first-pass rest and exercise nuclear angiography. J Thorac Cardiovasc Surg 79:645, 1980 Cannon PJ, Dell RB, Dwyer EM: Measurement of regional myocardial perfusion in man with 133 xenon and a scintillation camera. J Clin Invest 51:964, Cannon PJ, Dell RB, Dwyer EM: Regional myocardial perfusion rates in patients with coronary artery disease. J Clin Invest 51:978, Blackbum H, Keys A, Simonson E, et al: The electrocardiogram in population studies. Circulation 21:1160, Cheanvechai C, Groves LK, Reyes EA, et al: Manual coronary endarterectomy. J Thorac Cardiovasc Surg 70:524, Groves LK, Loop FD, Silver GM: Endarterectomy as a supplement to coronary artery-saphenous vein bypass surgery. J Thorac Cardiovasc Surg 64:514, Hochberg MS, Merrill WH, Michaelis LL, McIntosh CL: Results of combined coronary endarterectomy and coronary bypass for diffuse coronary artery disease. J Thorac Cardiovasc Surg 75:38, Effler DB, Grove LK, Sones FM Jr, et al: Endarterectomy in the treatment of coronary artery disease. J Thorac Cardiovasc Surg 47:98, 1964 Discussion DR. D. CRAIG MILLER (Stanford, CA): To my knowledge, this work by Dr. Kamath and his colleagues from Milwaukee represents the first attempt to objectively assess the hemodynamic functional adequacy of coronary endarterectomy. I have a few specific questions for Dr. Kamath. This analysis of 90 patients undergoing endarterectomy represented only a portion of the total patients who underwent angiography early after operation. To eliminate any possible selection bias, in what proportion of the total 438 patients operated on during this interval was coronary endarterectomy carried out? Were the results in the overall group comparable to those presented today? Second, we are all aware of the vagaries inherent in the inert gas washout technique of measuring regional blood flow. I wonder how you verified the reproducibility of your flow measurements. Were count matrices from multiple injections averaged? Third, did the variability of your perfusion measurements preclude statistically significant differences between the treatment groups? For example, were the isoproterenol-induced changes in perfusion significantly different between Groups 2 and 3? Furthermore, was Group 2 divided into subsets of patients (as was Group 3) with and without residual scarring or distal disease; indeed, was there any statistically significant difference in flow response between Group 2 patients who underwent bypass grafting alone and those patients with residual scar or distal disease who required endarterectomy? In other words, did the presence of old infarction, incomplete endarterectomy, or distal coronary artery disease really have any significant impact on the flow response?

8 35 Kamath et al: Endarterectomized Coronary Arteries Fourth, I would be interested to know if the results and conclusions would be different if the data had been segregated according to each individual regional coronary perfusion bed; that is, were the data analyzed separately according to each individual coronary artery with and without endarterectomy? Dr. Kamath's manuscript alluded to "excellent" late functional results, but did not include any supporting data. I would like to submit some information which I believe will help answer these questions. The long-term clinical results in two concurrent, matched cohorts of consecutive patients undergoing myocardial revascularization with and without right coronary artery endarterectomy during a remote period ( ) at Stanford University Medical Center were analyzed. Eighty patients underwent adjunctive right coronary endarterectomy and bypass grafting, in addition to 1 or more other bypass grafts. These were compared with a control group of 592 patients who underwent right coronary artery bypass grafting without endarterectomy in addition to 1 or more other bypass grafts. The two groups were statistically comparable with no significant differences in any preoperative clinical or hemodynamic variables. The early and late clinical results of endarterectomy were reviewed. Dr. Kamath's remarkably low perioperative transmural infarction rate of 2.6% in patients undergoing endarterectomy is superior to our perioperative (transmural and subendocardial) infarction rates of 16% for patients undergoing endarterectomy and 8% for those without endarterectomy. Nevertheless, this difference in the incidence of perioperative myocardial infarction was the only variable in our analysis that attained statistical significance between the two subsets. Operative mortality was 3.8% in patients with endarterectomy, and 2.9% in those without it. Follow-up averaged 5.7 years for the endarterectomy subset and 4.5 years for the control subset: there was no difference between the subsets in the linearized incidence of late postoperative angina, myocardial infarction, or congestive heart failure. As for overall actuarial survival for the two subgroups, there was no significant difference in sur- vival between patients with and those without endarterectomy. Actuarial survival rates 5 years postoperatively were 88 k 4%(SEM) for the endarterectomy subgroup versus 85 k 2% for the control subgroup. In distinct contrast to these gratifying functional and survival results and to Dr. Kamath's opinion that late graft occlusion occurs rarely, 15 of our 80 patients undergoing endarterectomy later experienced recrudescence of angina and were restudied. In this selected, symptomatic subpopulation, the patency rate of the right coronary artery bypass grafts (at an average of 3.2 years postoperatively) was only 33%. In the control subgroup, 114 symptomatic patients underwent late angiography; the patency rate of the right coronary artery bypass grafts (at an average of 1.9 years postoperatively) in this selected subpopulation was also low, being 52%. However, the difference between these two relatively low patency rates is not statistically significant (p > 0.05). Because only very small, symptomatic minorities of patients were restudied, it is obvious that no conclusive general statements concerning late graft patency are possible. In closing, I congratulate Dr. Kamath and his associates for producing this objective report. It adds to our understanding of the effects of coronary endarterectomy. DR. KAMATH: Thank you Dr. Miller for your kind re- Marks. The incidence of endarterectomy was 122 of 438 (27.85%). The results of the overall group were similar to the subgroup who consented to postoperative angiography. Because of the time involved and the radiation hazard, we deem it imprudent to do multiple injections on the same patient. This is an inherent limitation of the technique. The only statistically significant difference in flow response was between patients with complete endarterectomy and little or no scarring versus those with incomplete endarterectomy, scarring, or both. We did not investigate the effect of scarring on the flow response of conventionally grafted arteries. The data were not analyzed by individual coronary arteries because the subgroups would be too small.