Pallav J. Shah a, Manoj Durairaj a, Ian Gordon b, John Fuller c, Alex Rosalion a, Siven Seevanayagam a, James Tatoulis c, Brian F.

Transcription

1 European Journal of Cardio-thoracic Surgery 26 (2004) Factors affecting patency of internal thoracic artery graft: clinical and angiographic study in 1434 symptomatic patients operated between 1982 and 2002 q Pallav J. Shah a, Manoj Durairaj a, Ian Gordon b, John Fuller c, Alex Rosalion a, Siven Seevanayagam a, James Tatoulis c, Brian F. Buxton a, * a Department of Cardiac Surgery, Austin Health, Studley Road, Heidelberg, Vic Australia b Statistical Consulting Centre, University of Melbourne, Parkville, Vic., Australia c Epworth Medical Centre, Melbourne, Vic., Australia Received 4 September 2003; received in revised form 17 January 2004; accepted 10 February 2004; Available online 10 May 2004 Abstract Objective: The purpose is to define factors influencing long-term patency of the internal thoracic artery (ITA) to optimize the operative strategy. Methods: 1482 left internal thoracic artery (LITA) and 636 right internal thoracic artery (RITA) symptom-directed angiograms were studied in 1434 patients. Data were prospectively collected from patients who had primary coronary artery bypass surgery during the period The mean age of patients was 59 years; 85% were male. The mean period from operation to re-angiogram was 80 months. LITA was grafted to left anterior descending coronary artery (LAD) in 82% of cases, RITA to right coronary artery (RCA) in 40% and circumflex artery in 35% of cases. Graft failure was defined as $80% stenosis. Results: 96.3% of LITA and 88.1% of RITA grafts were patent. No patient variables were significantly associated with graft patency (age, gender, diabetes, hypertension, LVEF, NYHA, AMI). Target coronary artery was associated with patency of both LITA and RITA grafts with maximum patency when grafted to LAD ðp ¼ 0:02Þ: RITA had the worst patency to RCA, patency for the left system was identical to LITA. Proximal anastomosis to aorta (free RITA) had significantly better patency when compared with in situ RITA to RCA system ðp ¼ 0:005Þ while similar patency when grafted to left system. ITA diameter and target artery diameter were not associated with graft patency. Recent operations had better RITA patency ðp ¼ 0:03Þ: The interval from operation to angiogram was not associated with ITA patency (96% patency for LITA and 88% patency for RITA, remained stable when studied at,1, 1 4, 5 9, and.15 years). Conclusions: Even in a patient cohort that had adverse symptoms, excellent LITA and RITA patency was achieved which almost remained constant through all time intervals studied. q 2004 Elsevier B.V. All rights reserved. Keywords: Internal thoracic artery; Long-term; Factors; Patency 1. Introduction Since the early 1980s, the use of the left internal thoracic artery (LITA) for grafting of the left anterior descending (LAD) became the standard of care based on reports of superior graft patency, reduced cardiac events, decreased need for further intervention and enhanced long term survival when compared with patients receiving only q Presented at the joint 17th Annual Meeting of the European Association for Cardio-thoracic Surgery and the 11th Annual Meeting of the European Society of Thoracic Surgeons, Vienna, Austria, October 12 15, * Corresponding author. Tel.: þ ; fax: þ address: brian.buxton@austin.org.au (B.F. Buxton). venous conduits [1 6]. It has also been shown in previous studies that the internal thoracic artery (ITA) is relatively resistant to atherosclerosis in its native position [7], with the freedom from serious atherosclerosis being 20 years in some studies [8]. This led to more widespread use of both internal thoracic arteries and, subsequently, total arterial revascularization by developing different surgical strategies. These include in situ cross over grafting, free grafts, sequential anastomoses and the use of composite grafts [9 13]. Patients receiving bilateral ITA have better survival than the single ITA [14]. Although, the survival of patients with ITA grafts has been reported extensively, the factors affecting the late patency of these grafts have been studied less frequently [15 17] /$ - see front matter q 2004 Elsevier B.V. All rights reserved. doi: /j.ejcts

2 P.J. Shah et al. / European Journal of Cardio-thoracic Surgery 26 (2004) The purpose of this study is to analyze 20 years of clinical and angiographic data, and to delineate the pre and intraoperative factors associated with long-term patency of the ITA grafts thereby optimizing the operative strategy. 2. Materials and methods 2.1. Study population The study population was 1434 (9.8%) from a total of 14,659 patients who had undergone primary coronary artery bypass surgery between 1982 and Each patient had received at least one LITA or right internal thoracic artery (RITA) graft, and later presented with recurrence of symptoms (class III IV angina refractory to medical line of treatment) requiring coronary angiogram. Patients with a concomitant procedure during primary surgery were excluded. The preoperative patient characteristics and intraoperative variables were prospectively recorded in the database from Two thousand one hundred and eighteen ITA graft angiograms were studied in 1434 patients. One thousand, four hundred and eighty-two LITA grafts were studied in 1434 patients and 636 RITA grafts were studied in 626 (subgroup of the overall 1434) patients. The preoperative patient characteristics of 1434 patients are described in Table 1. The distribution of interval from operation to angiogram is described in Fig. 1. In LITA grafts, the mean interval from operation to reangiogram was 79 months and in RITA grafts the mean interval was 81 months. Aspirin was routinely given to patients from 1983 and lipid lowering agents from Angiographic analysis The native coronary artery and graft angiograms were performed using selective catherization. In patients who had two or more postoperative angiograms, the last angiogram was used for analytical purposes. Two cardiologists and a surgeon read the angiograms. The method of reporting was uniform. All the grafts were described in detail together with the degree of stenosis and coronary grafted. Table 1 Baseline patient characteristics of 1434 patients Variables Number % Age (mean) 60 years Sex (male) Diabetes NYHA III IV Hypertension LVEF,50% AMI Urgent operation Fig. 1. Number of angiograms performed in relation to time interval after the operation. A graft was considered patent when it had, 80% stenosis after visualization of the entire course of the graft including proximal anastomosis, distal anastomosis and distal target coronary artery. In sequential grafts, each segment was analyzed as a separate graft Statistical analysis The initial graft frequency and patency data were obtained from the entire group of patients. There were 1482 LITA grafts in 1434 patients and 636 RITA grafts in 626 patients. For analysis of the factors affecting graft patency, a reduced data set was created from patients in whom there were no missing data. This resulted in a dataset consisting of 1245 LITA grafts in 1209 patients and 541 RITA grafts in 537 patients. Simple logistic regression was used to analyze the factors affecting ITA graft patency. The time of graft failure was usually not recognizable and may have occurred any time between the operation and angiogram. We have, therefore, analyzed graft patency as a binary variable (failed or patent) as recorded at the time of the angiogram Surgery CABG was performed using a similar protocol by eight surgeons. All operations were performed on cardiopulmonary bypass with the help of antegrade cardioplegia prior to 1991 and antegrade/retrograde blood cardioplegia after ITA was harvested as an in situ grafts; none

3 120 P.J. Shah et al. / European Journal of Cardio-thoracic Surgery 26 (2004) Table 2 Target artery distribution and patency Target artery Number % Patency LITA LAD Diag Int, OM1, OM RITA LAD Diag Int, OM1, OM RCA PDA, PLV In situ RITA Free RITA Total % Patency Total % Patency LAD, Diag Int, OM1, OM PD, PLV, RCA Total Diag, diagonal; Int, intermediate; OM 1, obtuse marginal 1; OM 2, obtuse marginal 2; RCA, right coronary artery; PDA, posterior descending artery; PLV, posterior left ventricular branch. was skeletonized. The ITA grafts were dilated with a solution containing equal parts of Ringers Lactate and blood with 2 mm papaverine (80 mg) and 5000 u of heparin in 100 cc injected intraluminally with an arteriotomy cannula. The free ITA was stored in the same solution. There were 1482 LITA grafts in 1434 patients, 1434 standard end to side, 2 Y and 46 sequential grafts. There were 636 RITA grafts in 626 patients, 626 standard end-toside, 4 Y and 6 sequential grafts. T grafts and each segment of the sequential grafts were analysed as separate grafts. The LITA was used in-situ in 1466 grafts and free in 16 grafts. The free LITA was anastomosed proximally with the aorta in 14 grafts, to LITA in one graft and to RITA in one. The RITA was used as an in situ in 323 grafts and as a free graft in 313 grafts. Three hundred and nine free RITA grafts were proximally anastomosed to the aorta and four to LITA. 3. Results 3.1. Graft patency LITA (1461 of 1482 grafts were analysed for intraoperative predictors of graft failure 21 grafts had missing data). 1407/1461 LITA (96.3%) grafts were patent: 1383 (94.7%) grafts had 0 19% stenosis, 5 (0.3%) had 20 39% stenosis, 9 (0.6%) had 40 49% stenosis and 10 (0.7%) had 60 79% stenosis. Fifty four (3.7%) grafts failed: 10 (0.7%) had 80 99% stenosis and 44 (3.0%) were completely occluded. RITA (624 of 636 RITA were analysed statistically for intra-operative factors of graft patency 12 grafts had missing data). 550/624 (88.1%) RITA grafts were patent: 535 grafts (85.7%) had 0 19% stenosis, 2 (0.3%) had 20 39% stenosis, 8 (1.3%) had 40 49% stenosis and 5 (0.8%) had 60 79% stenosis. Seventy-four (11.9%) grafts failed: 19 (3.0%) had 80 99% stenosis and 55 (8.8%) were completely occluded. The target coronary arteries and the distribution of LITA and RITA graft patency are described in Table 2. The effect of patient and operative variables on LITA graft patency is described in Table 3 and RITA graft patency is described in Table 4. Relationship of patient variables to graft patency: Age, gender, diabetes, hypertension, previous MI, left ventricular ejection fraction, NYHA class and urgent Table 3 Factors affecting LITA graft patency Variable Baseline Comparison OR 95% CI P-value Age x xþ Sex Female Male AMI No Yes Hypertension No Yes Diabetes No Yes Ejection fraction $50%,50% NYHA Class 0, 1 or 2 Class 3 or Urgency of operation No Yes Target artery Non-LAD LAD Target artery diameter,2.0 mm $2.0 mm Target artery stenosis,60% 60 79% % Total Conduit diameter,2.0 mm $2.0 mm Interval between operation x xþ and angiogram (months) Year of operation x xþ

4 P.J. Shah et al. / European Journal of Cardio-thoracic Surgery 26 (2004) Table 4 Factors affecting patency of RITA grafts Variable Baseline Comparison OR 95% CI P-value Age x xþ Sex Female Male AMI No Yes Hypertension No Yes Diabetes No Yes Ejection fraction $50%,50% NYHA Class 0, 1 or 2 Class 3 or Urgency of operation No Yes Target artery LAD Circumflex Right Target artery diameter,2.0 mm $2.0 mm Proximal anastomosis and coronary stenosis In situ,,80% Aorta,,80% In situ, $80% Aorta, $80% Conduit diameter,2.0 mm $2.0 mm Interval between operation and angiogram (months) x xþ Year of operation operation were not significantly associated with ITA graft patency. Year of operation. For any given angiogram interval, recently performed operative procedures were associated with significantly better patency of RITA ðp ¼ 0:05Þ but this variable did not affect LITA ðp ¼ 0:09Þ patency. The interval from operation to angiogram did not affect ITA graft patency (LITA: P ¼ 0:11; RITA: P ¼ 0:09). Ninety-six percent patency for LITA and 88% patency for RITA, remained stable when studied at, 1, 1 4, 5 9, and. 15 years. (Table 5, Fig. 2). RITA. The variables that were significantly associated with graft patency were: Target artery (P ¼ 0.01). Grafts to RCA had the worst patency while those to the LAD had the best. The order of increasing patency was RCA, PDA, OM, INT, DIAG, LAD. Target coronary artery stenosis. There was a significant interaction between proximal anastomosis and coronary stenosis ðp ¼ 0:001Þ: This means that the association Relationship of operative variables to graft patency: LITA: The only variable that was significantly associated with LITA patency was the target coronary artery grafted ðp ¼ 0:012Þ: Maximum patency was seen when the LITA was grafted to LAD. Target artery diameter, stenosis or ITA diameter were not significantly associated with LITA graft patency. Table 5 Graft patency Interval(y) LITA RITA Grafts Studied % Patency Grafts Studied % Patency, All Fig. 2. Graft patency at,1, 1 4, 5 9, 10 14, and.14 years after the operation.

5 122 P.J. Shah et al. / European Journal of Cardio-thoracic Surgery 26 (2004) between patency and coronary stenosis varied according to whether the proximal end was in situ or anastomosed to the aorta. For this reason, a combined variable was created, containing the four combinations of the two variables. As can be seen from the results, relative to the baseline category of (attached,, 80% stenosed), the other three categories all had very similar odds ratios (0.26, 0.27, 0.31). The stenosis effect was very minor for aorta grafts but large and statistically significant for attached (in situ) grafts. Target artery diameter or ITA diameter was not significantly associated with graft patency. 4. Discussion This study was performed over 20 years and analysed the relationship between the patient and operative variables and graft patency. It demonstrated the durability and patency of ITA grafts through the second decade after implantation. In situ, LITA grafts maintained an almost uniform patency rate over the 20-year period. In situ, RITA grafts to the right coronary artery (RCA) initially resulted in an increased graft failure rate, which improved when in situ or free RITAs were grafted to the left coronary system. These findings are similar to other long-term angiographic patency studies [8,18,19]. The second important finding was that ITA grafts studied were almost always completely normal or totally occluded. There were only a few grafts with moderate or intermediate degrees of stenosis. These results are unlike our recent study on late patency of saphenous vein, [20] where there is almost an equal distribution of grafts through the entire spectrum of stenotic lesions. The target artery grafted affected patency of both LITA and RITA grafts, with maximum patency when grafted to the LAD. These findings were consistent with previous studies, demonstrating the benefits of grafting the left coronary system with ITA grafts [16,17]. The LAD supplies a large area of myocardium and, therefore, has a good runoff. Grafts to the non-lad arteries were at higher risk with the worst patency seen in the RCA territory. Insufficient graft length to bypass all disease without tension together with increasing right ventricular dilatation may have contributed to high failure of in situ RITA to RCA grafts. The overall patency of the RITA grafts to the left system is almost identical with that of LITA grafts. However, this is not surprising considering that both ITAs have identical histopathology. Low graft flow results from competitive flow or from grafting small vessels. Lower graft flow results in loss of shear force and decreased nitric oxide release, which may cause vasoconstriction and graft failure. The relationship between target coronary artery stenosis and ITA patency remains controversial [16,17,21,22]. Native vessels with low-grade stenoses were largely avoided in this study; therefore, it was not possible to analyze the relationship. Simple logistic regression was used to analyze the patient and operative variables affecting graft patency. The time between operation and angiogram was used as a continuous variable. It was not possible to identify the time of graft failure and, therefore, survival analysis techniques that are based on the assumption that graft failure occurs at the time of angiography are not valid. The patient variables studied were not significantly associated with ITA patency. However, data on cholesterol, triglycerides and cigarette smoking were incomplete and thus, they were excluded. Interestingly, diabetic patients who often have small diseased vessels and of poor quality, did not appear to adversely affect ITA graft patency over the average 7-year follow-up period. It is important to keep in mind that, although, the series is large, the number of failures is small, limiting the statistical power of the study to detect associations with patient characteristics. There was a bias towards better patency for free grafts in the series possibly because of the high failure rate between 1985 and 1989 when in situ RITA grafts were grafted to the right system which is similar to other reported studies [16,17]. In situ, RITA is rarely used to graft the RCA before crux where disease is common. If there is any suggestion of insufficient length to bypass disease in the distal RCA, the RITA was used as a free or a composite extension graft to PDA and PLV branches. Alternately, the RITA was used as a crossover graft to the left system. These modifications may have contributed to the increased RITA patency of the RITA in the later period. In the present study, ITA size and target artery diameter were not significantly associated with ITA graft patency, which suggests that the ITA suitable for grafting small vessels. This finding may explain some of the benefits of ITA grafting in patients with diabetes and diffuse small vessel disease. Our current strategy is total arterial revascularization, primarily using bilateral internal thoracic arteries as in situ, crossover, and free and composite grafts to the left system. The radial artery is used as the second line arterial graft originating from the aorta, or to extend the right ITA, or as a Y graft from the LITA. Composite graft patency, reduced cardiac events, decreased need for further intervention and survival require further monitoring and evaluation [10,23 25]. 5. Limitations Retrospective studies of graft patency have clear limitations. The selection of symptomatic patients might be expected to underestimate the true graft patency of patients having coronary artery bypass grafting. Our patient population with symptom-directed angiography was only

6 P.J. Shah et al. / European Journal of Cardio-thoracic Surgery 26 (2004) a small population of all patients undergoing surgery during that time. Although, the bias cannot be determined for certain, it seems plausible that those patients were more likely, if anything, to have failed grafts, given their symptoms. This bias will be offset to some extent because the asymptomatic population will include some patients with silent graft occlusion which will overestimate true graft patency. Some of the intra-operative and patient variables were missing, thus resulting in some patients being excluded from the statistical analysis. Measurements of operative variables were based on visual assessment and may have varied from the surgeon to surgeon. Different calendar periods may be associated with use of different methods of graft harvesting, preparation and anastomotic techniques. 6. Conclusion Even in a cohort with adverse symptoms, excellent patency of the ITA grafts was observed, which remained constant through all the time intervals studied. High failure rates of in situ RITAs to the right system early in the series have resulted in preference for the RITA being grafted to the left system or used as a free or composite extension graft to the right coronary system. References [1] Pick AW, Orszulak TA, Anderson BJ, Schaff HV. Single versus bilateral internal mammary artery grafts: 10-year outcome analysis. Ann Thorac Surg 1997;64: [2] Cameron AA, Green GE, Brogno DA, Thornton J. Internal thoracic artery grafts: 20-year clinical follow-up. J Am Coll Cardiol 1995;25: [3] Loop FD, Lytle BW, Cosgrove DM, Stewart RW, Goormastic M, Williams GW, Golding LA, Gill CC, Taylor PC, Sheldon WC. Influence of the internal mammary artery graft on 10-year survival and other cardiac events. N Engl J Med 1986;314:1 6. [4] Cosgrove DM, Loop FD, Lytle BW, Goormastic M, Stewart-Gill CC, Golding LR. Does mammary artery grafting increase surgical risk? Circulation 1985;72(Suppl II): [5] Dougenis D, Brown AH. Long-term results of reoperations for recurrent angina with internal mammary artery versus saphenous vein grafts. Heart 1998;80:9 13. [6] Cameron A, Davis K, Green G, Schaff HV. Coronary bypass surgery with internal thoracic artery grafts-effects on survival over a 15-year period. N Engl J Med 1996;334: [7] Sims FH. A comparison of coronary and internal mammary arteries and implications of the results in the etiology of atherosclerosis. Am Heart J 1983;105: [8] Barner HB, Barnett MG. Fifteen to twenty-one-year angiographic assessment of internal thoracic artery as a bypass conduit. Ann Thorac Surg 1994;57: [9] Loop FD, Lytle BW, Cosgrove DM, Golding LA, Taylor PC, Stewart RW. Free (aorta-coronary) internal mammary artery graft. Late results. J Thorac Cardiovasc Surg 1986;92: [10] Dion R, Verhelst R, Rousseau M, Goenen M, Ponlot R, Kestens- Servaye Y, Chalant CH. Sequential mammary grafting. Clinical, functional, and angiographic assessment 6 months postoperatively in 231 consecutive patients. J Thorac Cardiovasc Surg 1989;98: [11] Barner HB. Arterial grafting: techniques and conduits. Ann Thorac Surg 1998;66(S2-5):S25 8. (discussion). [12] Lev-Ran O, Paz Y, Pevni D, Kramer A, Shapira I, Locker C, Mohr R. Bilateral internal thoracic artery grafting: midterm results of composite versus in situ crossover graft. Ann Thorac Surg 2002;74: [13] Calafiore A, Contini M, Vitolla G, di Mauro M, Mazzei V, Teodori G, Giammarco G. Bilateral internal thoracic artery grafting: Long-term clinical and angiographic results of in situ versus Y grafts. J Thoracic Cardiovasc Surg 2000;120: [14] Lytle BW, Blackstone EH, Loop FD, Houghtaling PL, Arnold JH, Akhrass R, McCarthy PM, Cosgrove DM. Two internal thoracic artery grafts are better than one. J Thorac Cardiovasc Surg 1999;117: [15] Bernal JM, Rabasa JM, Lequerica MA, Echevarria JR, Herrera JM, Zueco J, Colman T, Pajaron A, Revulta JM. Factors affecting early graft patency after coronary grafts. Rev Esp Cardiol 1990;43(8): (in Spanish). [16] Buxton BF, Ruensakulrach P, Fuller J, Rosalion A, Reid CM, Tatoulis J. The right internal thoracic artery graft-benefits of grafting the left coronary system and native vessels with a high-grade stenosis. Eur J Cardiothorac Surg 2000;18: [17] Chow MST, Sim E, Orszulak TA, Schaff HV. Patency of internal thoracic artery grafts: comparison of right versus left and importance of the vessel grafted. Circulation 1994; [5 Pt 2]: II [18] Ivert T, Huttunen K, Landou C, Bjork VO. Angiographic studies of internal mammary artery grafts 11 years after coronary artery bypass grafting. J Thorac Cardiovasc Surg 1988;96:1 12. [19] Tatoulis J, Buxton BF, Fuller JA, Royse AG. Total arterial coronary revascularization: techniques and results in 3220 patients. Ann Thorac Surg 1999;68: [20] Shah P, Gordon I, Fuller J, Seevanayagam S, Rosalion A, Tatoulis J, Raman J, Buxton BF. Factors affecting patency of saphenous vein graft: 25-year clinical and angiographic study in 1402 symptomatic patients operated between 1977 and J Thor Cardiovasc Surg 2003;126(6): [21] Kawasuji M, Sakakibara N, Takemura H, Tedoriya T, Ushijima T, Watanabe Y. Is internal thoracic artery grafting suitable for a moderately stenotic coronary artery? J Thorac Cardiovasc Surg 1996; 112: [22] Gaudino M, Alessandrini F, Nasso G, Bruno P, Manzoli A, Possati G. Severity of coronary artery stenosis at preoperative angiography and midterm mammary graft status. Ann Thorac Surg 2002;74(1): [23] Tector A, Schmahl T. Purely internal thoracic artery grafts: outcomes. Ann Thorac Surg 2001;72: [24] Schmidt SE, John JW, Thornby JI, Miller CCIII, Beall Jr. A. Improved survival with multiple left sided bilateral internal thoracic artery grafts. Ann Thorac Surg 1997;64:9 14. [25] Munneretto C, Negri A, Manfredi J, Terrini A, Rodella G, ElQarra S, Bisleri G. Safety and usefulness of composite grafts for total arterial myocardial revascularization: a prospective randomized evaluation. J Thorac Cardiovas Surg 2003;125: Appendix A. Conference discussion Dr P. Mortensen (Odense, Denmark): It is always very interesting to see your extensive database. Do you have any comments on the difference between free ITAs and attached ITAs? Are there any results difference?

7 124 P.J. Shah et al. / European Journal of Cardio-thoracic Surgery 26 (2004) Dr Buxton: First of all, on the left system it does not apply, so it is only on the right, and we have found the patency of free and in situ internal thoracic arteries to be almost identical. Dr Y. Balbaa (Cairo, Egypt): I wanted to ask if any of the cases were done as composite Y grafts and T grafts and what is the impact of this on the results? My other question is, what is the relation between percentage target vessel stenosis and long-term patency, if you have analyzed this point? Dr Buxton: The first question. Most of these were simple grafts, and if you look at the number of patients and the number of angiograms, they were almost identical; there was perhaps a 5% use of Y and extension grafts. And second, with the LITA, I mentioned there was no relationship with stenosis, but I want to preempt the answer by saying we did not graft many patients that had low grade stenoses, and we tended to use 50 60% as a threshold. There was a relationship between the RITA and target artery vessel stenosis. It was rather complex and I did not present that here, but there was some relationship. But again, we did not graft many vessels with a low grade stenosis. Dr Balbaa: What is the cut point that you would suggest? Dr Buxton: We set a threshold of 70%, and that is a very naive statement, because collateral flow does not depend purely on native vessel stenosis, it depends on all sorts of other things, and the figure that is commonly quoted in the literature of 70% is a surgeons figure. This can be altered by nitroglycerine, drugs and other factors, and it is a figure that we use as a guideline only. Dr H. Mair (Leuven, The Netherlands): So you recommend a vein graft to the RCA? Dr Buxton: I think you are speaking to the wrong person. We have used total arterial grafts for years and years. We modified our practice based on early bad outcomes. I am now quite happy to use the free RITA graft or an extension graft to the RCA. I am very cautious about using in situ grafts to the right, particularly if there is any tension, because the anastomosis is difficult to do it when it is attached. Furthermore, the heart may enlarge with time. So we favor a free, or an extension graft if we are going to use the RITA to the right side. If we are going to use three grafts, I would put both internal thoracic arteries on the left system, and perhaps use a radial on the right no vein grafts.