Bilateral internal mammary artery grafting: in situ versus Y-graft. Similar 20-year outcome

Transcription

1 European Journal of Cardio-Thoracic Surgery 50 (2016) doi: /ejcts/ezw100 Advance Access publication 25 March 2016 ORIGINAL ARTICLE Cite this article as: Di Mauro M, Iacò AL, Allam A, Awadi MO, Osman AA, Clemente D et al. Bilateral internal mammary artery grafting: in situ versus Y-graft. Similar 20-year outcome. Eur J Cardiothorac Surg 2016;50: a Bilateral internal mammary artery grafting: in situ versus Y-graft. Similar 20-year outcome Michele Di Mauro a, Angela L. Iacò b, Ahmed Allam c, Mohammed O. Awadi d, Ahmed A. Osman e, Daniela Clemente a and Antonio M. Calafiore b, * L Aquila University, L Aquila, Italy b Prince Sultan Cardiac Center, Riyad, Saudi Arabia c Ain Shams Universities, Cairo, Egypt d Benha Universities, Banha, Egypt e Cairo Universities, Giza, Egypt * Corresponding author. Department of Research, Department of Adult Cardiac Center, Prince Sultan Cardiac Center, Riyadh, Saudi Arabia. Tel: ; fax: ; am.calafiore@gmail.com (A.M. Calafiore). Received 27 August 2015; received in revised form 25 December 2015; accepted 22 February 2016 Abstract OBJECTIVE: The aim of this study was to evaluate the 20-year clinical outcome of patients undergoing coronary artery bypass grafting with bilateral internal mammary arteries (BIMAs) using two different configurations, in situ versus Y-graft. METHODS: From September 1991 to December 2002, 2150 patients with multivessel coronary artery disease underwent isolated myocardial revascularization with BIMA grafting. BIMA was used as an in situ or Y-configuration in 1332 and 818 cases, respectively. A propensity score model was applied to calculate a standardized difference of 10% between groups (BIMA in situ vs BIMA Y-graft), and a cohort of 1468 matched patients was identified (734 in each group). Death, non-fatal myocardial infarction and the need for repeat revascularization were defined as major adverse cardiac events. RESULTS: Late mortality was 24.3% (n = 357) [BIMA in situ vs BIMA Y-graft: 26.9% (n = 197) vs 21.8% (n = 160)]; in 11.6% (n = 170) of cases death was due to cardiac causes [11.9% (n = 87) vs 11.3% (n = 83)]. The rate of major adverse cardiac events was 37.1% (n = 545) [40.8% (n = 299) vs 33.5% (n = 246)]. The 20-year survival was 59 ± 6% and the event-free survival was 45 ± 7%. CONCLUSIONS: The clinical outcome of BIMA grafting is independent of surgical configuration. Y-grafting increases the flexibility of BIMA grafting and should be taken into account when a surgical strategy for myocardial revascularization needs to be planned. Keywords: Myocardial revascularization Bilateral internal mammary artery Y-grafting INTRODUCTION The use of bilateral internal mammary arteries (BIMAs) is scarcely diffused in the real world [1, 2], despite being supported by robust evidence over the last decade [3]. Lack of standardization for the appropriate use of the right internal mammary artery (RIMA) has limited its routine use in myocardial revascularization. RIMA can be anastomosed in situ to the left anterior descending artery (LAD) or to the proximal branches of the circumflex artery, but its length may pose technical challenges. Although skeletonization is a technique that provides better internal mammary artery (IMA) length, it is not widely accepted. As a result, it is not always possible to choose the best anastomotic site, and we are forced to graft the RIMA according to its length rather than the most suitable coronary artery segment. Alternatively, the RIMA can be used as a free graft, which is anastomosed to the left internal Presented at the 29th Annual Meeting of the European Association for Cardio- Thoracic Surgery, Amsterdam, Netherlands, 3 7 October mammary artery (LIMA) as a composite conduit. This option is more technically demanding and can expose to higher rates of competitive flow with increased graft failure [4]thoughnodefinitive data are available to confirm this finding [5 7]. Mid-term clinical outcome seems to be similar regardless of BIMA configuration, either in situ or as a composite conduit [5, 6]. The aim of this study was to report our very long-term results in propensity-matched patients in whom both strategies were used, and to evaluate whether there is any difference in survival and major adverse cardiac event rates. MATERIALS AND METHODS Study population From September 1991 to December 2002, 2150 patients with multivessel coronary artery disease underwent isolated myocardial ADULT CARDIAC The Author Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

2 730 M. Di Mauro et al. / European Journal of Cardio-Thoracic Surgery Table 1: Preoperative data BIMA in situ (n = 1332) BIMA Y-graft (n = 818) SD (%) BIMA in situ (n = 734) BIMA Y-graft (n = 734) SD (%) Propensity score (%) 34 ± ± ± ± Age (years) 61 ± 9 62 ± ± 9 62 ± Female gender 184 (14%) 106 (13%) (13%) 96 (13%) 0.4 Urgency 357 (27%) 112 (13%) (15%) 112 (15%) 0.4 EF (%) 59 ± ± ± ± LMT disease 218 (16%) 104 (13%) (14%) 96 (13%) 2.9 IDDM 62 (5%) 46 (6%) (6%) 41 (6%) 0.6 NIDDM 218 (16%) 144 (18%) (19%) 132 (18%) 1.4 Previous stroke 65 (5%) 33 (4%) (4%) 30 (4%) 0 Previous MI 740 (55%) 380 (46%) (56%) 365 (50%) 7.2 COPD 48 (4%) 54 (7%) (6%) 39 (5%) 1.1 ECV 274 (21%) 144 (18%) (20%) 137 (19%) 3.9 Unstable angina 442 (33%) 198 (24%) (24%) 176 (24%) 1 Heart failure 23 (2%) 10 (1%) (2%) 9 (1%) 5 Chronic renal failure 28 (2%) 33 (4%) (4%) 25 (3%) 0.7 Ventricular arrhythmias 19 (1%) 7 (1%) (1%) 6 (1%) 1.5 Hypercholesterolaemia 560 (42%) 339 (41%) (43%) 307 (42%) 1.4 Smoke addiction 597 (45%) 313 (38%) (39%) 289 (39%) 1.4 Hypertension 602 (45%) 375 (46%) (48%) 335 (46%) 4.1 Preoperative atrial fibrillation 20 (2%) 9 (1%) (1%) 8 (1%) 1.3 EF: ejection fraction; LMT: left main trunk; IDDM: insulin-dependent diabetes mellitus; NIDDM: non-insulin-dependent diabetes mellitus; MI: myocardial infarction; COPD: chronic obstructive pulmonary disease; ECV: extracardiac vasculopathy; SD: standard deviation; BIMA: bilateral internal mammary artery. Table 2: Perioperative data BIMA in situ (n = 1332) BIMA Y-graft (n = 818) SD (%) BIMA in situ (n = 734) BIMA Y-graft (n = 734) SD (%) Redo 43 (3%) 43 (5%) (5%) 37 (5%) 0.6 OPCAB 367 (28%) 277 (34%) (35%) 238 (32%) 5.5 No. of anastomoses 2.9 ± ± ± ± No. of arterial anastomoses 2.6 ± ± ± ± BIMA sequential 333 (25%) 360 (44%) (35%) 305 (42%) 16.2 Grafted territories 2.5 ± ± ± ± Complete myocardial revascularization 1108 (83%) 692 (85%) (83%) 619 (84%) 3 BIMA and SVG 365 (27%) 357 (44%) (39%) 296 (40%) 2.5 BIMA and RA 114 (9%) 39 (5%) (9%) 37 (5%) 14.4 BIMA and RGEA 240 (18%) 58 (7%) (14%) 53 (7%) 27.1 OPCAB: off-pump coronary artery bypass; BIMA: bilateral internal mammary artery; SVG: saphenous vein graft; RA: radial artery; RGEA: right gastroepiploic artery; SD: standard deviation. revascularization with BIMA grafting; in particular, BIMA was used as an in situ or Y-configuration in 1332 and 818 cases, respectively. The decision to use a Y-graft was taken after careful observation of the coronary angiography, selecting territories with a similar expected run-off. If any doubt, in situ grafting was preferred. As the decision was taken before surgery, there was no between groups cross-over. A propensity score model was applied to calculate a standardized difference of 10% between groups (BIMA Y-graft vs BIMA in situ), and a cohort of 1468 matched patients was identified (734 in each group). Preoperative data are listed in Table 1. Surgical technique Patients were anaesthetized as previously described [8]. From October 1991 to May 1994, BIMAs were harvested pedicled, whereas from June 1994 the conduits were dissected skeletonized. The harvesting techniques have recently been reported [9]. The technique used to perform end-to-side and end-to-end anastomoses has already been described [8]. To avoid graft distortion, the IMA needs to be merely placed over the heart. The internal blood pressure will force the graft to maintain always the right orientation. After heparinization, regardless of the harvesting technique used, IMAs were distally clipped, injected with 10 ml of diluted papaverine (1 mg/ml) and allowed to pharmacologically dilate. When a composite graft was constructed [5], the proximal anastomosis of the free arterial graft to the IMA was performed before starting cardiopulmonary bypass. Briefly, the conduits are positioned on a folded pad put between the branches of the sternal retractor, to assure a relatively stable operative field. Operative details are reported in Table 2.

3 M. Di Mauro et al. / European Journal of Cardio-Thoracic Surgery 731 Follow-up All patients were clinically followed up for a median of 200 months ( months). The most recent information was obtained by telephone interviews with the patients or referring cardiologists. Follow-up was 100% complete and ended in June Statistics and end-points Data are expressed as mean ± standard deviation or median. Categorical variables were reported as counts and percentages. Differences between groups were assessed, using independent t-test or Mann Whitney U-test for continuous variables and χ 2 test for categorical variables. A saturated propensity score model was built with an area under the curve of A 1:1 sample matching was performed. All variables entered into the model are listed in Tables 1 and 2 (but BIMA-RA, BIMA-RGEA and BIMA sequential). Standardized differences in means before and after matching are plotted in Fig. 1. A standardized mean difference of 10% was considered as the threshold for a good balance. For all tests, statistical significance was set at P < The SPSS 20 software (SPSS, Inc., Chicago, IL, USA) was used; an integration with R-software was used to obtain automatically a good sample matching. The primary end-point was to assess the 20-year survival with BIMA Y-graft versus BIMA in situ. Early results and late event-free survival were also analysed. Major adverse cardiac events included death, non-fatal myocardial infarction and the need for repeat revascularization. Early major complications included death, myocardial infarction, low output syndrome, acute respiratory failure, acute renal failure, cerebrovascular accidents and gastrointestinal bleeding or ischaemia. RESULTS Early results In-hospital mortality was 1.2% (n = 18), 30-day mortality was 2.0% (n = 29) and 1-year mortality was 3.1 (n = 45), without difference between groups (Fig. 2). Long-term results Late mortality was 24.3% (n = 357) [BIMA in situ vs BIMA Y-graft: 26.9% (n = 197) vs 21.8% (n = 160)]; in 11.6% (n = 170) of cases, death was due to cardiac causes [11.9% (n = 87) vs 11.3% (n = 83)]. The rate of major adverse cardiac events was 37.1% (n = 545) [40.8% (n = 299) vs 33.5% (n = 246)] (Fig. 3). The 20-year survival and the event-free survival were 59 ± 6 and 45 ± 7%, respectively. Differences between groups are plotted in Figs 4 and 5. Cardiac survival was 75 ± 6% (73 ± 8% Y-graft vs 77 ± 5% in situ, P = 0.48). DISCUSSION Adding another anastomosis to a technically demanding surgical procedure can increase the possibility of failure. However, as we ADULT CARDIAC Figure 2: Early outcomes: death, cardiac death, problems, major complications. BIMA in situ (red column); BIMA Y-graft (blue column). DSW: deep sternal wound; BIMA: bilateral internal mammary artery. Figure 1: Standardized differences before and after propensity score matching. Figure 3: Late outcomes: death, cardiac death, MI, new revascularization, cardiac event, any event. BIMA in situ (red column); BIMA Y-graft (blue column). BIMA: bilateral internal mammary artery; MI: myocardial infarction.

4 732 M. Di Mauro et al. / European Journal of Cardio-Thoracic Surgery Figure 4: Twenty-year survival. BIMA in situ (red line); BIMA Y-graft (blue line). BIMA: bilateral internal mammary artery. Figure 5: Twenty-year event-free survival. BIMA in situ (red line); BIMA Y-graft (blue line). BIMA: bilateral internal mammary artery. perform the Y anastomosis as a bench surgery, any possible error becomes unlike, if not impossible. Anastomosing the RIMA to the LIMA has many advantages: it results in increased effective RIMA length, especially if the graft is harvested in a skeletonized fashion, and offers the possibility of reaching distal targets, including the posterior descending artery. Arterial revascularization of the left side becomes an easy task, the only decision left being whether to use another arterial graft for the right side. As with all surgical techniques, there are possible drawbacks. Grafting two territories with different run-off is the Achilles heel of Y-grafting (e.g. 90% proximal LAD stenosis of a large territory and 50% proximal circumflex artery stenosis). A different expected run-off can be associated with lower patency rates [9], given that competitive flow or a steal phenomenon may predispose to graft failure [3, 9, 10]. The relative flow distribution would be dependent on the sizes of the respective vascular beds and grafted vessels as well as the degree of native coronary stenosis [11]. Indeed, the real danger is grafting a coronary territory where intracoronary pressure is not significantly lower than the driving pressure. The conductance of arterial grafts is generally lower due to the smaller diameter and greater length, resulting in some pressure drop along the graft. The longer the arterial assembly, the lower the pressure at the distal anastomosis, a phenomenon aggravated by normal distal tapering of coronary arteries. This goes with the difficulty in identifying the severity of a coronary stenosis by a simple visual assessment. There are many studies that highlighted discrepancies between the angiographic grade of a coronary stenosis and its functional severity assessed with fractional flow reserve. Hamilos et al.[12] showed that 23% of angiographically non-critical left main stenoses were functionally critical when fractional flow reserve was evaluated. Tonino et al. [13] showed that 50 70, and >90% angiographic stenoses were not functionally significant in 65, 20 and 4% of the cases, respectively. An interesting study from Botman et al. [14] showed that in patients with a stenosis severity of 50 70% on visual estimation, graft occlusion occurred in 9.8% of the functionally significant lesions and in 20.2% of the functionally non-significant lesions. These findings reflect the difficulties in following strict and objective rules when a composite conduit is built. On the other hand, many reports did not show any drawback when BIMA was used as a Y-graft to the left system. Our group [15] demonstrated that intraoperative injection of dobutamine increases blood flow in the Y-graft by more than two times, not only in the main stem but also in each branch. Mid-term patency rates were found to be similar in patients who received BIMA in situ or as Y-graft [4 6]. Other studies reported different results. Sakaguchi et al. [16] showed that, 2 weeks after surgery, although arterial composite Y-graft improved myocardial blood flow at rest, it was not as effective as an independent graft for improving coronary flow reserve soon after coronary artery bypass grafting. Manabe et al. [9] showed that composite grafting was associated with a higher incidence of graft occlusion or string sign than individual grafting in the presence of moderate (50 75%) coronary stenosis, a finding consistent with previous studies [3, 14]. Results of this study can be helpful in encouraging more surgeons to use BIMA grafting. Since the basic paper from Loop et al. [17] demonstrates the overwhelming superiority of LIMA over the saphenous vein when grafted to the LAD, LIMA to the LAD has become the gold standard for myocardial revascularization. However, another study from the same institution [18] that showed the superiority of IMA grafting over single LIMA did not get the same impact on clinical practice. The use of BIMA in the general population was found to be superior in a number of observational or propensity-matched studies and in several meta-analyses. Increased long-term survival has also been reported in high-risk patients undergoing BIMA grafting when compared with that undergoing single IMA grafting. Obese [19], diabetic [20] and elderly [21] patients as well as patients with significant renal impairment [22], all have the benefit of a longer survival when BIMA grafting is used. Notwithstanding this, surgeons are reluctant to embrace BIMA grafting. Analysing data from the State of Virgina, LaPar et al. [2] found that, from 2001 to 2013, BIMA use was 3% in the overall bypass population and 6% in a subgroup of patients considered low risk for BIMA use. A recent analysis of the Society of Thoracic Surgeons database showed that the use of BIMA was 3.5% in 1999 and 4.1% in 2009 [1]. There are two main reasons for this attitude. The first reason is the fear of sternal wound complications. A large body of evidence suggests that IMA skeletonization minimizes wound infections [23]. This finding has a solid anatomical basis. The selection of the most appropriate technique of IMA harvesting is crucial to preserve sternal vascularization. Most of the side branches of the IMA arise as a trunk. If it is cut before the bifurcation, inversion of the blood flow allows maintenance of sternal revascularization from other sources (perforating branches to pectoralis muscle and intercostal branches). This aspect has been confirmed by functional studies. Moreover, aggressive blood glucose control in diabetic patients further reduces the incidence of sternal wound infections even in this patient subset. Skeletonization has another added value as it provides extra

5 M. Di Mauro et al. / European Journal of Cardio-Thoracic Surgery 733 length to the conduit and higher internal diameter, both characteristics that can allow for a higher number of IMA anastomoses. In addition, a recent propensity-matched study from our group [8] reported a better 17-year survival in patients with skeletonized BIMA grafts than in patients with pedicled BIMA grafts. The second reason is the lack of experience. Most surgeons are not accustomed to using the RIMA. The right coronary artery is the easiest target vessel, but it is often diseased at the point where the RIMA could be grafted and the impact on survival of a second arterial graft is by far superior if used on the left side rather than on the right side of the heart. Although the outcome of RIMA to LAD is similar to that of LIMA to LAD [24], the retrosternal position of the conduit may increase the risk for re-entry injuries. The retroaortic route to the lateral territory has some limitations, as it enables the proximal branches of the circumflex artery to be reached easily, but makes grafting of more distal targets difficult. The only alternative is to cut the RIMA and to use it as a free graft, anastomosed, when possible, to the LIMA. Results of our study show that, when BIMA is used, Y-grafting is safe and can achieve the same long-term clinical results as in situ grafting, making RIMA utilization easier (and friendly). We always tried to prevent competitive flow situations avoiding Y-grafting when two territories with different expected run-off had to be grafted. However, as routine functional studies cannot be obtained in the real world, where visual assessment of coronary stenoses is the rule, the effectiveness of our strategy is more theoretical than practical. Nonetheless, we believe that competitive flow due to functionally non-significant stenoses has little clinical relevance, as these patients will not experience ischaemia and will not need repeat revascularization on that territory [14]. The main limitation of our study is its retrospective nature, which is partially overcome by applying propensity matching. Additionally, the sample size of the two study groups is relatively small and clinical outcomes are not supported by angiographic visual assessment of anastomoses. Notwithstanding this, the two groups of patients as well as their risk profile throughout the study period are quite similar. Furthermore, when BIMA was indicated, all patients were operated on consecutively, using the same surgical principles and the same indications though the procedures were performed by different surgeons. An important strength of this study is the length of follow-up, with all patients being operated on during the 1990s. We are aware that propensity-matched studies are not as strong as randomized controlled studies. However, the latter would last so many years to exceed the length of our professional life. Our study includes approximately 1500 propensity-matched patients and, despite the single-centre design, strongly suggests that using BIMA as a composite conduit increases the possibility of grafting the lateral wall and provides long-term results similar to the in situ configuration. Definitively, Y-grafting increases the flexibility of BIMA grafting and should be taken into account when a surgical strategy for myocardial revascularization needs to be planned. Our experience is supported by ESC/ECTS 2014 guidelines that suggests to use BIMA in patients with reasonable life expectancy (Class II A) and to skeletonize mammary artery when BIMA is used (Class I B). Moreover, Y-graft can implement the use of BIMA which in turn increases the long-term outcome of patients with multivessel disease; this is surely the best way to increase the number of patients with three-vessel disease undergoing unjustified percutaneous coronary intervention [25]. We strongly recommend that anatomical and physiological aspects be carefully evaluated by the operating surgeon on an individual basis to identify possible contraindications to Y-grafting, independent of any theoretical attempt to schematize or predict flow dynamics. Conflict of interest: none declared. REFERENCES [1] ElBardissi AW, Aranki SF, Sheng S, O Brien SM, Greenberg CC, Gammie JS. Trends in isolated coronary artery bypass grafting: an analysis of the Society of Thoracic Surgeons adult cardiac surgery database. J Thorac Cardiovasc Surg 2012;143: [2] LaPar DJ, Crosby IK, Rich JB, Quader MA, Speir AM, Kern JA et al. Bilateral internal mammary artery use for coronary artery bypass grafting remains underutilized: a propensity-matched multi-institution analysis. Ann Thorac Surg 2015;100:8 15. [3] Pevni D, Hertz I, Medalion B, Kramer A, Paz Y, Uretzky G et al. Angiographic evidence for reduced graft patency due to competitive flow in composite arterial T-grafts. J Thorac Cardiovasc Surg 2007;133: [4] Calafiore AM, Contini M, Vitolla G, Di Mauro M, Mazzei V, Teodori G et al. Bilateral internal thoracic artery grafting: long-term clinical and angiographic results of in situ versus Y grafts. J Thorac Cardiovasc Surg 2000;120: [5] Glineur D, Hanet C, Poncelet A, D Hoore W, Funken JC, Rubay J et al. Comparison of bilateral internal thoracic artery revascularization using in situ or Y graft configurations: a prospective randomized clinical, functional, and angiographic midterm evaluation. Circulation 2008;118:S [6] Fukui T, Tabata M, Manabe S, Shimokawa T, Takanashi S. Graft selection and one-year patency rates in patients undergoing coronary artery bypass grafting. Ann Thorac Surg 2010;89: [7] Calafiore AM, Vitolla G, Iaco AL, Fino C, Di Giammarco G, Marchesani F et al. Bilateral internal mammary artery grafting: midterm results of pedicled versus skeletonized conduits. Ann Thorac Surg 1999;67: [8] Di Mauro M, Iaco AL, Acitelli A, D Ambrosio G, Filipponi L, Salustri E et al. Bilateral internal mammary artery for multi-territory myocardial revascularization: long-term follow-up of pedicled versus skeletonized conduits. Eur J Cardiothorac Surg 2015;47: [9] Manabe S, Fukui T, Shimokawa T, Tabata M, Katayama Y, Morita S et al. Increased graft occlusion or string sign in composite arterial grafting for mildly stenosed target vessels. Ann Thorac Surg 2010;89: [10] Glineur D, Hanet C, D Hoore W, Poncelet A, De Kerchove L, Etienne PY et al. Causes of non-functioning right internal mammary used in a Y-graft configuration: insight from a 6-month systematic angiographic trial. Eur J Cardiothorac Surg 2009;36:129 35; discussion [11] Sugimura Y, Toyama M, Katoh M, Kotani M, Kato Y, Hisamoto K. Outcome of composite arterial Y-grafts in off-pump coronary artery bypass. Asian Cardiovasc Thorac Ann 2011;19: [12] Hamilos M, Muller O, Cuisset T, Ntalianis A, Chlouverakis G, Sarno G et al. Long-term clinical outcome after fractional flow reserve-guided treatment in patients with angiographically equivocal left main coronary artery stenosis. Circulation 2009;120: [13] Tonino PA, Fearon WF, De Bruyne B, Oldroyd KG, Leesar MA, Ver Lee PN et al. Angiographic versus functional severity of coronary artery stenoses in the FAME study fractional flow reserve versus angiography in multivessel evaluation. J Am Coll Cardiol 2010;55: [14] Botman CJ, Schonberger J, Koolen S, Penn O, Botman H, Dib N et al. Does stenosis severity of native vessels influence bypass graft patency? A prospective fractional flow reserve-guided study. Ann Thorac Surg 2007;83: [15] Gaudino M, Di Mauro M, Iaco AL, Canosa C, Vitolla G, Calafiore AM. Immediate flow reserve of Y thoracic artery grafts: an intraoperative flowmetric study. J Thorac Cardiovasc Surg 2003;126: [16] Sakaguchi G, Tadamura E, Ohnaka M, Tambara K, Nishimura K, Komeda M. Composite arterial Y graft has less coronary flow reserve than independent grafts. Ann Thorac Surg 2002;74: [17] Loop FD, Lytle BW, Cosgrove DM, Stewart RW, Goormastic M, Williams GW 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. [18] Lytle BW, Blackstone EH, Loop FD, Houghtaling PL, Arnold JH, Akhrass R et al. Two internal thoracic artery grafts are better than one. J Thorac Cardiovasc Surg 1999;117: ADULT CARDIAC

6 734 M. Di Mauro et al. / European Journal of Cardio-Thoracic Surgery [19] Benedetto U, Montecalvo A, Kattach H, Amrani M, Raja SG, Harefield Cardiac Outcomes Research Group. Impact of the second internal thoracic artery on short- and long-term outcomes in obese patients: a propensity score matched analysis. J Thorac Cardiovasc Surg 2015;149:841 7 e1 2. [20] Kajimoto K, Yamamoto T, Amano A. Coronary artery bypass revascularization using bilateral internal thoracic arteries in diabetic patients: a systematic review and meta-analysis. Ann Thorac Surg 2015;99: [21] Kinoshita T, Asai T, Suzuki T, Kuroyanagi S, Hosoba S, Takashima N. Off-pump bilateral skeletonized internal thoracic artery grafting in elderly patients. Ann Thorac Surg 2012;93: [22] Kinoshita T, Asai T, Suzuki T. Off-pump bilateral skeletonized internal thoracic artery grafting in patients with chronic kidney disease. J Thorac Cardiovasc Surg 2015;150: e3. [23] Fouquet O, Tariel F, Desulauze P, Mevel G. Does a skeletonized internal thoracic artery give fewer postoperative complications than a pedicled artery for patients undergoing coronary artery bypass grafting? Interact CardioVasc Thorac Surg 2015;20: [24] Raja SG, Benedetto U, Husain M, Soliman R, De Robertis F, Amrani M, Harefield Cardiac Outcomes Research Group. Does grafting of the left anterior descending artery with the in situ right internal thoracic artery have an impact on late outcomes in the context of bilateral internal thoracic artery usage? J Thorac Cardiovasc Surg 2014;148: [25] Kolh P, Windecker S, Alfonso F, Collet JP, Cremer J, Falk V et al ESC/ EACTS Guidelines on myocardial revascularization: the Task Force on Myocardial Revascularization of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS). Developed with the special contribution of the European Association of Percutaneous Cardiovascular Interventions (EAPCI). Eur J Cardiothorac Surg 2014;46: APPENDIX. CONFERENCE DISCUSSION Scan to your mobile or go to to search for the presentation on the EACTS library Dr M. Wyler Von Ballmoos (Milwaukee, WI, USA): Dr Di Mauro and his team present their data on the use of bilateral mammary artery grafting in nearly 1500 patients over a decade with a mean follow-up of 16 years. They addressed the question whether outcomes are different when right internal mammary artery plus left internal mammary artery strategy is used in two different configurations: Once using the right internal mammary artery as an in situ graft and alternatively as a Y-graft off the left internal mammary artery strategy. Based on their data, the authors conclude that early and late outcomes, including mortality, nonfatal MI and revascularizations, are similar with either use of the right internal mammary artery. While it has been shown consistently that bilateral inferior mesenteric artery revascularization is superior to single artery bypass grafting plus vein, it s likely much more challenging to demonstrate a difference between two surgical strategies in the subpopulation where both inferior mesenteric artery are being used. I commend the authors on their attempt at answering this question with a large study that provided long-term follow-up in a group of patients that also received a significant number of additional arterial grafts from the radial and the gastroepiploic artery. The success of revascularization is typically measured in survival, freedom from symptoms, cardiac events and repeat revascularization, and you suggest no difference here. But some of these outcomes may be very difficult to ascertain in a cohort study that uses a posteriori data collection over 20 years. Further, I would propose that for your specific study question using both internal mammary arteries and simply comparing two surgical strategies- graft patency at different time intervals might be a more suitable end point. Finally, you have chosen to use a propensity score matched analysis of your data with a C-statistic of 0.81 which suggests that the two groups indeed are quite different. In addition then you opted not to adjust for other covariates in your analysis of treatment effect. With that being said, I would like to ask you two questions, one about distal targets and one about the impact of time and surgeon proficiency as they relate to your study. My first question is: Was the inferior mesenteric artery ever used for revascularization of the right coronary circulation; and if so, to what extent? If not, how often was the inferior mesenteric artery sequenced when the right internal mammary artery was used as a Y-graft? Dr Di Mauro: Iaddressthefirst question. In Y-configuration we use left internal mammary artery mainly to left anterior descending coronary artery and the right internal mammary artery mainly to lateral wall. We use just in 3% of cases sequential graft to re-vascularise with right internal mammary artery lateral wall and inferior wall sequentially. In in- situ configuration we use in 50% of cases left internal mammary artery to left anterior descending coronary artery crossing the chest, and in 50% to right coronary artery and in 60% to, left anterior descending coronary artery and just sometimes to reach the lateral wall because the right internal mammary artery was not so length to reach the lateral wall. This was our target coronary revascularization. Dr Wyler Von Ballmoos: So if I understand correctly, you predominantly used free right internal mammary artery or Y-graft right internal mammary artery for the left coronary circulation? Dr Di Mauro: Sure. Dr Wyler Von Ballmoos: So, secondly, it s more a technical question about also how you analysed your data. We know that technically more skilled surgeons tend to have better outcomes, and medical management changed drastically over 10 years. So my question is: Were all surgeons equally likely to contribute cases to the study in both treatment arms, and was there a trend over time in using either configuration and how eventually did you adjust for that in your analysis? Dr Di Mauro: Yes there was a trend, you re right, because we started in 1991 using mainly in situ graft, and in the last part of this study we used mainly Y-graft. This was a problem to adjust even with propensity score, because in that case we could obtain just a small size of cohort, so it s very difficult to adjust for this data. But this is our real life. I m here just to report the result of a group of surgeons who started with the idea to use a bilateral internal mammary artery in situ and then over time switched this idea to a Y-configuration and especially to achieve a left side revascularization with the bilateral mammary artery. To answer your question about the strategy of the group, in Italy we have usually a chief, not a consultant. And the chief shared his idea, his strategy, with the assistant and then we all started to perform the same strategy. So all the surgeons across the study adopted the same strategy. Dr Wyler Von Ballmoos: If I may just make one recommendation based off the manuscript you sent me, sit down with your statistician again and see how you can get the most out of your study.