Angiographic 5-Year Follow-up Study of Right Gastroepiploic Artery Grafts Sari Voutilainen, MD, Kalervo Verkkala, MD, PhD, Antero J~irvinen, MD, PhD, and Pekka Keto, MD, PhD Departments of Thoracic and Cardiovascular Surgery and Diagnostic Radiology, Helsinki University Central Hospital, Helsinki, Finland Background. The right gastroepiploic artery (RGEA) has been used from 1987 in coronary artery bypass grafting in several clinical studies. However, the published 1- to 5-year patency rates have been dependent on the selection of patients for angiography. Methods. In our study, the RGEA was used from March 1987 to May 1990 for coronary artery bypass grafting in 31 consecutive patients, 25 male and 6 female. All but 1 patient had triple-vessel disease, and the mean number of distal anastomoses was 3.9 (range, 2 to 5). Internal thoracic artery grafts were used concomitantly in all patients. Results. One early and two late deaths occurred. All but I of the 28 surviving patients underwent clinical and angiographic follow-up examinations 3 months and 5 years after the operation. The 5-year patency of RGEA grafts was 82.1%, with a 95% confidence interval of 63.1% to 93.9%. In 4 of the 5 nonvisualized cases, the recipient coronary artery showed proximal stenosis of up to 70%, allowing substantial competitive flow. The 5-year patency of the RGEA graft was near that of the left internal thoracic artery, at 90.3%, and the right internal thoracic artery, at 94.4%; and superior to the 66.7% patency of venous grafts. Conclusions. At 5-year follow-up, angiography of RGEA grafts showed good function and a smooth lumen, especially if the proximal stenosis was more than 70%. (Ann Thorac Surg 1996;62:501-5) here are two principal reasons for using artery grafts T instead of vein grafts in coronary artery bypass grafting (CABG). The first is the high failure rate of saphenous vein grafts due to degenerative changes. Follow-up studies have shown that only 50% to 60% of vein grafts are patent after 10 years [1, 2]. In contrast, good long-term patency has been demonstrated for internal thoracic artery (ITA) grafts in several series [3, 4]. The second reason is the lack of suitable veins for use as grafts. They may be of poor quality, or they may have been used in previous bypass operations or stripped away because of varicosity. In triple-vessel disease, the use of both ITAs may not be enough to revascularize the whole myocardium. Our earlier study showed the right gastroepiploic artery (RGEA) to be a promising choice as a bypass graft [5]. Since March 1987, we have used the RGEA in addition to both ITA and vein grafts to perform CABG in triple-vessel disease. Here we report the 5-year follow-up results of our first 31 patients. Material and Methods From March 1987 to May 1990, the RGEA was used for CABG in 31 patients, aged 55.1 _+ 9.04 years (mean _+ standard deviation). Patients with triple-vessel or left main disease and with a life expectancy of more than 10 Accepted for publication March 25, 1996. Address reprint requests to Dr Verkkala, Department of Thoracic and Cardiovascular Surgery, Helsinki University Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland. years, excluding possible coronary death, were included in this study (Table 1). One patient had had a previous CABG; one of his two earlier venous grafts was occluded. All the patients underwent elective operations. Patients with a history of abdominal problems were not included in the present series. Of the total, 30 patients had triplevessel disease and 1 had left main disease. The preoperative left ventricular ejection fraction was 0.59 _+ 0.14 (range, 0.38 to 0.90). Functionally, 8 of the patients were in New York Heart Association (NYHA) class IV, 19 were in class III, and 4 were in class II. None had congestive heart failure. Both ITAs were investigated preoperatively using a selective angiographic technique. The suitability of the RGEA was studied intraoperatively by estimating the free flow that was considered sufficient in every patient. The indication for using the RGEA was a lack of suitable veins in 7 patients. The RGEA was used in the rest of the patients to avoid using vein grafts, especially when the patient was young. The RGEA was combined with both ITAs in 30 cases and once with only the right ITA. Vein grafts were used concomitantly in 14 patients, in I of whom the vein was taken from the right arm. The total number of venous grafts was 15, however, because 1 patient received two venous grafts. In 1 case, the inferior epigastric artery was used as a free graft to the left circumflex artery. The ITA was used as a free graft four times and the RGEA once. The ITA was used as a bifurcated graft ten times and as a sequential graft twice. Thus, the total number of left ITA anastomoses was 34, and RITA anastomoses numbered 39. The RGEA was anastomosed to the right coronary artery in 28 cases; in 10 1996 by The Society of Thoracic Surgeons 0003-4975/96/$15.00 Published by Elsevier Science Inc PII S0003-4975(96)00279-2
502 VOUTILAINEN ET AL Ann Thorac Surg GASTROEPIPLOIC ARTERY: 5-YEAR PATENCY 1996;62:501-5 Table 1. Patient Characteristics (n 31) Characteristic No. Percent Male/female 25/6 80.6/19.4 Diabetes 4 7.7 Arterial hypertension 15 48.4 Family histo~ 20 64.5 Smoking 18 58 Previous AMI One 15 48.4 Two 5 16.1 AMI = acute myocardial infarction. of them the anastomosis was to a peripheral branch because the whole trunk of the right coronary artery was severely atherosclerotic. The anastomosis for the RGEA was twice to the left anterior descending artery and once, because of insufficient length, as a free graft to the left diagonal artery. The left and right ITAs were used to bypass the left anterior descending artery and the left circumflex coronary artery and their branches. The right ITA was also anastomosed once to the right coronary artery. The mean number of grafts per patient was 3.9, (range, 2 to 5). A follow-up examination was performed at 3 to 5 months and again 5 years after the operation. It consisted of a clinical examination and coronary angiography with left ventricular cineangiography, using conventional technique. In addition, angiography of the celiac axis and selective angiography of both ITAs were performed. Confidence intervals (CIs) for graft patency were calculated using the binomial distribution; CIs for differences between graft patencies were calculated using the normal distribution. Confidence intervals for differences in median NYHA class were calculated using the Wilcoxon approach [6]. Operative Technique After harvesting the ITAs, we dissected the RGEA as a pedicle graft from the great curvature of the stomach using 3-0 polyglycolic sutures. After systemic heparin administration, the grafts were transected distally, and dilute papaverine solution was injected into them to facilitate blood flow. The free flow was measured only if it was suspected to be insufficient. In these 5 cases, the RGEA flow was 86.8 _+ 34 ml/min (range, 50 to 140 ml/min). The free flow in 1 minute was estimated by collecting the flow volume from a free artery during 15 seconds and then multiplying this volume by 4. The RGEA graft was brought up anterior to the pylorus in 27 cases and behind it in 3 cases. In 20 cases, a hole was made in the fibrous part of the diaphragm with blunt dissection using a long clamp; in 10 cases, a slit was made with a knife, either because of a voluminous pedicle or to achieve a better position for the RGEA graft. One RGEA graft was used as a free graft with proximal anastomosis into the ascending aorta. The diameter of the bypassed coronary arteries was measured by flexible probes. The median diameter was 1.5 mm (range, 1 to 2 mm). The standard technique of cardiopulmonary bypass with a membrane oxygenator was used. The mean aortic cross-clamp time was 62.4 minutes (range, 45 to 88 minutes), and the mean cardiopulmonary bypass time was 110.7 minutes (range, 76 to 183 minutes). Mild systemic hypothermia and crystalloid cardioplegia were used for myocardial protection. The prepared vein with ligated tributaries was maintained in situ with continuous flow until just before the bypass was to be performed. Coronary anastomoses were performed with either 7-0 or 8-0 continuous polypropylene (Johnson & Johnson) sutures. All patients received prophylactic antibiotic treatment with vancomycin for 48 hours, beginning at the induction of anesthesia. Results There was one hospital death, due to cardiac arrhythmia, on the fourth postoperative day. This patient had had two previous myocardial infarctions. He had also undergone bilateral femoral amputation after an accident 20 years earlier. At autopsy, all the grafts were patent and no acute myocardial infarction was found. Important early complications among survivors developed in 7 patients. Reexploration for postoperative bleeding was needed in 2 patients. Sternal wound infection developed in 3 patients (9.7%) (95% CI, 2.0% to 25.8%) in whom the ITAs had been used bilaterally. In addition, 1 of them had insulin-dependent diabetes mellitus. Two of them were treated successfully with sternal refixation and mediastinal lavage, and 1 with conservative therapy. One patient experienced hemiplegia on the left side on the first postoperative day. Seven days postoperatively, 1 patient suffered a perforated gastric ulcer, which was treated operatively. The 3-month clinical follow-up examination showed improvement in NYHA class in all but 1 patient. The angiograms (3 to 5 months) showed patent RGEA grafts in 80% (24 of 30) of the patients. Two late deaths occurred during the 5-year follow-up period. A woman, who showed no improvement in NYHA class at the 3-month visit, died 5 months after the operation, probably of myocardial infarction. No autopsy was done. One patient died 5 years after the operation because of rupture of an abdominal aortic aneurysm. At autopsy, all the grafts were patent. One patient underwent laparotomy 2 years after CABG. She had a perforated appendix, which was treated successfully without complications. At follow-up examination 5 years postoperatively, all patients but I were in NYHA class I or II. In a comparison of the preoperative versus 5-year NYHA class, the median improvement was 2 classes (95% CI, 1.5 to 2 classes). Coronary angiography was performed on all but I of the surviving patients. In addition, graft patency of I patient was obtained at autopsy. Five RGEA grafts, anastomosed to the right coronary artery, could not be visualized at angdography. In 4 of them, the proximal stenosis was up to 70% (Fig 1). Three left ITA and two right ITA anastomoses, one of which was a free
Ann Thorac Surg VOUTILAINEN ET AL 503 1996;62:501-5 GASTROEPIPLOIC ARTERY: 5-YEAR PATENCY Proximal stenosis (%) 100 80 Or 60 40 20 Non visualized (5) Patent (23) Fig 1. Association of right gastroepiploic artery graft patency with preoperative angiographic proximal stenosis. *Intraoperative stenosis in this artery was 50% to 60%, and so it was bypassed. right ITA graft, could not be seen. Five of 15 venous grafts, including one brachial vein graft, were occluded (Table 2). The difference between the RGEA and left ITA anastomosis patency was -8.2% (95% CI, -25.8% to 9.4%) and between the RGEA and right ITA was -12.3% (95% CI, -28.3% to 3.7%), with both results slightly in favor of ITAs. The RGEA graft patency was, however, better than the venous graft patency, with a difference of 15.5% (95% CI, -12.3% to 33.8%). In addition, significant narrowing was observed in five of the ten patent venous grafts when compared with the 3-month angiography results (Fig 2). In contrast, no noticeable atherosclerosis was observed in the patent RGEA grafts (Fig 3). The left ventricular ejection fraction was 0.63 + 0.20 (n - 19) at 5 years postoperatively. This was not significantly different from the preoperative ejection fraction of 0.59 _+ 0.14 (n = 29). Fig 2. Patent free right internal thoracic artery (Free RITA) anastomosed to the first left oblique marginal artery, and a significantly narrowed saphenous vein graft (VG) anastomosed to the second left oblique marginal artery. the anatomic location of the coronary arteries. By using the RGEA as an arterial graft in addition to both ITAs, it is possible to perform CABG with only arterial grafts [5, 7-12]. We started to use the RGEA in 1987 as a third arterial graft. The in situ pedicle is long enough to reach nearly any part of the heart, and harvesting of the RGEA is well tolerated by the patient. Early results, including our own [5], using the RGEA in CABG have been Comment In triple-vessel disease, safe and complete revascularization using only ITA grafts is often impossible because of Table 2. Five-Year Anastomosis Patency Patency Graft N1/N Percent 95% CI (%) RGEA 23/28 82.1 63.1-93.9 LITA 28131 90.3 74.2-98.0 RITA 34/36 94.4 81.3-99.3 VG 10/15 66.7 38.4-88.2 CI - confidence interval; L[TA - left internal thoracic artery; N number of anastomoses; N1 number of patent anastomoses; RGEA - right gastroepiploic arted,; RITA = right internal thoracic artery; VG - venous graft. Fig 3. Patent right gastroepiploic artery graft (RGEA) anastomosed to the left anterior descending artery.
504 VOUTILAINEN ET AL Ann Thorac Surg GASTROEPIPLOIC ARTERY: 5-YEAR PATENCY 1996;62:501-5 Table 3. Results and Percentage of Patients Having Angiography in Different Studies of Right Gastroepiploic Artery Grafts Angiography First Author/Year Operated (n) Time Done Patency a Grandjean [8] (1994) 300 1-25 mo 88 (29%) 77% --~ 95% c Perrault [14] (1993) 51 1 y 5 (10%) 80% Suma [12] (1993) 200 1-5 y 40 (20%) 95% (83.1%-99.4%) Mills [13] (1993) 90 1-36 mo 22 (24%) 100 (84.6%-100%) Voutilainen (1996) 31 5 y 27 + I b (93%) 82.1% (63.1%-93.9%) 95% confidence interval in parentheses, b One patency was obtained at autopsy. * The patency was improved according to the learning curve. promising [8-14]. Here we report our systematic late follow-up angiographic results. In our series, the incidence of postoperative sternal infections was higher (9.7%) than is usually seen in patients undergoing CABG [15]. Diabetes was a risk factor in 1 of our patients with mediastinitis in whom bilateral ITAs were also used. Diabetes is a risk factor that has been noticed earlier [3, 16]. In selected cases such as diabetics, use of the RGEA graft and only one ITA graft might be indicated instead of bilateral ITA grafts to avoid disturbing the blood supply of the sternum, as occurs when both ITAs are harvested. Other reasons for the high incidence of mediastinitis could be the greater operative trauma and the longer operative time because of harvesting the RGEA graft in addition to both ITAs. Postoperative gastric ulceration has not been found previously in relation to use of the RGEA [17]. It is therefore notable that in 1 of our patients, perforation of a gastric ulcer developed 7 days postoperatively. This patient had no history of abdominal problems. Whether the ulceration was caused by disturbed blood supply to the gastric mucosa or by postoperative stress itself remains unclear. The only in-hospital death was apparently not associated with the use of an RGEA graft. This patient died of cardiac arrhythmias 4 days postoperatively. At 3 months postoperatively, angina was relieved in all but 1 patient. This woman, who stayed at the NYHA lii level, had no well-functioning grafts at follow-up angiography and died at home 5 months after the operation, probably of myocardial infarction. No new complications due to RGEA grafts developed during 5 years. At 5-year follow-up, all of the patients but I were symptom free and had remained in NYHA class I to II. One patient who showed improvement in NYHA class at 3 months fell back to NYHA class III at 5 years. All but 1 of the surviving patients underwent angiographic reexamination at 5 years; the patency rate was 82.1% for RGEA grafts. The patency rate was not as good as that of ITA grafts. However, it was better than the vein graft 5-year patency rate of 66.7%, which was similar to the saphenous vein graft patency in other arterial graft studies [3, 18]. There are several possible reasons for nonfunctioning of the RGEA grafts. These 31 patients represent our first experience with the RGEA. Perhaps we were still on the rising part of the learning curve with this new graft [8]. Another reason could be the difficulty of visualization of the RGEA grafts using angiography. In our study, we used the RGEA mainly to bypass the right coronary artery and its distal branches, which tend to be smaller in diameter than the coronary arteries bypassed with ITA grafts. In addition, the degree of the proximal stenosis may influence blood flow through the graft. It has been reported that possible competitive flow should be avoided, especially when using a pedicled RGEA graft [13]. We observed that among the 5 cases in which the RGEA could not be visualized, in only 1 case was the proximal stenosis of the recipient coronary artery more than 70% of its diameter. We consider that the competitive flow from the bypassed coronary artery itself could be the reason for the poor function of the RGEA graft in the noncritically stenosed cases. No such clear association between graft patency and proximal stenosis was observed in ITA grafts and venous grafts. In 1 case, the function of the RGEA graft had become better at the 5-year angiography, probably because of progression of the coronary artery stenosis. This graft could not be visualized at early angiographic examination, but was widely patent after 5 years. The same observation has been reported earlier for ITA grafts [19]. There are some studies of midterm results on the patency of the RGEA graft, but the follow-up times are still short [8, 12-14]. In addition, the proportion of the patients undergoing midterm angiographic examination seems to be low ('Fable 3). In this study, all but 1 of the surviving patients underwent the 5-year angiographic examination. No complication associated with using the RGEA and causing permanent deficiency was observed in this study. Patency of the RGEA was better than venous graft patency, and no noticeable atherosclerosis was found at B-year angiography, in contrast to venous grafts. In cases without substantial competitive flow, the RGEA patency was as good as that of ITA grafts. After receiving long-term results, we started to use RGEA grafts again in April 1995 in a new prospective study, in which we compare them with other arterial grafts and venous grafts. The indications for using the RGEA are young patients with triple-vessel disease and no history of abdominal problems, whose survival is dependent on well-functioning grafts. When using the RGEA, we avoid competitive flow, ie, proximal stenosis of less than 70%. Probably because of the learning curve,
Ann Thorac Surg VOUTILAINEN ET AL 505 1996;62:501-5 GASTROEPIPLOIC ARTERY: 5-YEAR PATENCY the incidence of mediastinitis and of other complications seems to be lower now. References 1. Carnpeau L, Enjalbert M, Lesp6rance J, Vaislic C, Grondin CM, Bourassa MG. Atherosclerosis and late closure of aortocoronary saphenous vein grafts: sequential angiographic studies at 2 weeks, I year, 5 to 7 years and 10 to 12 years after surgery. Circulation 1983;68(Suppl 2):1-7. 2. Grondin CM, Campeau L, Thornton JC, Engle JC, Gross FS, Schreiber H. Coronary artery bypass grafting with saphenous vein. Circulation 1989;79(Suppl 1):24-9. 3. Galbut DL, Traad EA, Dorman MJ, et al. Seventeen-year experience with bilateral internal mammary artery grafts. Ann Thorac Surg 1990;49:195-201. 4. Lytle BW, Cosgrove DM, Saltus GL, Taylor PC, Loop FD. Multivessel coronary revascularization without saphenous vein: long-term results of bilateral internal mammary artery grafting. Ann Thorac Surg 1983;36:540-7. 5. Verkkala K, J/irvinen A, Keto P, Virtanen K, Lehtola A, Pellinen T. Right gastroepiploic artery as a coronary bypass graft. Ann Thorac Surg 1989;47:716-9. 6. Gardner MJ, Altman DG, ed. Statistics with confidence. 1st ed. London: Br Med J, 1989:28-30, 70-6. 7. Attum AA. The use of the gastroepiploic artery for coronary artery bypass graft: another alternative. Tex Heart Inst J 1987;14:289-92. 8. Grandjean JG, Boonstra PW, den Heyer P, Ebels T. Arterial revascularization with the right gastroepiploic artery and internal mammary arteries in 300 patients. J Thorac Cardiovasc Surg 1994;107:1309-16. 9. Lytle BW, Cosgrove DM, Ratliff NB, Loop FD. Coronary artery bypass grafting with the right gastroepiploic artery. J Thorac Cardiovasc Surg 1989;97:826-31. 10. Mills NL, Everson CT. Right gastroepiploic artery: a third arterial conduit for coronary artery bypass. Ann Thorac Surg 1989;47:706-11. 11. Pyre J, Brown PM, Charrette EJP, Parke JO, West RO. Gastroepiploic-coronary anastomosis. A viable alternative bypass graft. J Thorac Cardiovasc Surg 1987;94:256-9. 12. Suma H, Wanibuchi Y, Terada Y, Fukuda S, Takayama T, Furuta SI. The right gastroepiploic artery graft. Clinical and angiographic midterm results in 200 patients. J Thorac Cardiovasc Surg 1993;105:615-23. 13. Mills NL, Hockmuth DR, Everson CT, Robart CC. Right gastroepiploic artery used for coronary artery bypass grafting. Evaluation of flow characteristics and size. J Thorac Cardiovasc Surg 1993;106:579-86. 14. Perrault LP, Carrier M, H6bert Y, et al. Clinical experience with the right gastroepiploic artery, in coronary artery bypass grafting. Ann Thorac Surg 1993;56:1082-4. 15. Verkkala K, J~irvinen A. Mediastinal infection following open-heart surgery. Treatment with retrosternal irrigation. Scand J Thorac Cardiovasc Surg 1986;20:203-7. 16. Kouchoukos NT, Wareing TH, Murphy SF, Pelate C, Marshall WG Jr. Risks of bilateral internal mammary artery bypass grafting. Ann Thorac Surg 1990;49:210-9. 17. Suma H, Wanibuchi Y, Furuta S, Takeuchi A. Does use of gastroepiploic artery graft increase surgical risk? J Thorac Cardiovasc Surg 1991;101:121-5. 18. Lytle BW, Loop FD, Cosgrove DM, Ratliff NB, Easley K, Taylor PC. Long term (5 to 12 years) serial studies of internal mammary artery and saphenous vein coronary bypass grafts. J Thorac Cardiovasc Surg 1985;89:248-58. 19. Dincer B, Barner HB. The "occluded" internal mammary graft. Restoration of patency of apparent occlusion associated with progression of coronary disease. J Thorac Cardiovasc Surg 1983;85:318-22.