Cardiovascular Surgery. Routine Use of Bilateral Skeletonized Internal Thoracic Artery Grafting Long-Term Results

Transcription

1 Cardiovascular Surgery Routine Use of Bilateral Skeletonized Internal Thoracic Artery Grafting Long-Term Results D. Pevni, MD; G. Uretzky, MD; A. Mohr, BSc; R. Braunstein, PhD; A. Kramer, MD, PhD; Y. Paz, MD; I. Shapira, MD; R. Mohr, MD Background Skeletonized harvesting of the internal thoracic artery (ITA) decreases the severity of sternal devascularization, thus reducing the risk of postoperative sternal complications in patients undergoing bilateral ITA grafting. Methods and Results Between 1996 and 2001, 1515 consecutive patients underwent skeletonized bilateral ITA grafting. Of the 1179 male and 336 female patients, 641 (42.3%) were 70 years of age, and 519 (34.2%) had diabetes mellitus. Operative mortality was 2.8%. Early postoperative morbidity included sternal infection (1.6%), cerebrovascular accident (3%), and perioperative myocardial infarction (1%). Multiple regression analysis showed chronic obstructive pulmonary disease (odds ratio, 11.3; 95% confidence interval [CI], 4.45 to 28.55), repeat operation (odds ratio, 12.7; 95% CI, 3.25 to 49.56), and diabetes mellitus (non insulin dependent: odds ratio, 4.64; 95% CI, 1.85 to 11.59; insulin dependent: odds ratio, 6.9; 95% CI, 1.35 to 35.27) to be associated with increased risk of sternal infection. Follow-up (between 5 and 12 years) revealed 305 late deaths. Kaplan-Meier 10-year survival rates for patients 65, 65 to 74, and 75 years of age were 87%, 75%, and 52%, respectively. Cox regression analysis revealed increased overall mortality (early and late) in patients with peripheral vascular disease (hazard ratio [HR], 1.8; 95% CI, 1.39 to 2.33), patients 75 years of age (HR, 7.23; 95% CI, 4.16 to 12.55), those undergoing repeat operations (HR, 2.22; 95% CI, 1.27 to 3.89), patients with preoperative congestive heart failure (HR, 1.64; 95% CI, 1.29 to 3.75), and those with chronic renal failure (HR, 1.52; 95% CI, 1.11 to 2.01). Operations performed without cardiopulmonary bypass were associated with better postoperative survival (HR, 0.66; 95% CI, 0.49 to 0.87). Conclusions Bilateral ITA grafting is associated with low morbidity and good long-term results. Use of skeletonized bilateral ITA is appropriate for the elderly and most patients with diabetes; however, it is not recommended for repeat operations or for patients with chronic obstructive pulmonary disease. (Circulation. 2008;118: ) Key Words: coronary disease grafting revascularization Enhanced long-term survival has been shown when the left internal thoracic artery (LITA) is grafted to the left anterior descending artery (LAD) rather than the saphenous vein graft (SVG). 1 The ITA has been recognized as the optimal conduit because of its superior patency rate and freedom from arteriosclerosis. 2,3 Clinical Perspective p 712 Recent studies have shown that survival benefit and freedom from reintervention are even better with the use of 2 ITAs. 3,4 Survival advantage of 2 ITAs has been related especially to bilateral ITA (BITA) grafting of the myocardium supplied by the LAD and circumflex arteries (ie, the left coronary artery system). 5 In most centers, the ITA is isolated from the chest wall as a pedicle, together with the vein, muscle, fat, and accompanying endothoracic fascia. 4 6 Harvesting is relatively quick because cautery is used to separate the pedicle from the chest wall. However, cauterization damages the blood supply to the sternum, which in turn impedes sternal healing and exposes the sternum to the risks of early dehiscence and infection in operations involving both ITAs. The risk of sternal infection is particularly high in patients with preoperatively limited sternal blood supply such as the elderly and those with diabetes mellitus. 7,8 In an effort to reduce the risk of sternal infection, a surgical technique was developed in which the ITA is dissected as a skeletonized vessel. 9,10 The skeletonized artery is isolated gently with scissors and silver clips without cauterization. Skeletonized ITA dissection leaves the vein, muscle, and accompanying tissue in place. The advantages are that the dissected ITA is longer and that its spontaneous blood flow is Received November 29, 2007; accepted May 30, From the Department of Cardiothoracic Surgery, Tel Aviv Sourasky Medical Center and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. Correspondence to Professor R. Mohr, MD, Department of Cardiothoracic Surgery, Tel Aviv Sourasky Medical Center, 6 Weizman St, Tel Aviv 64239, Israel. goldmarion@hotmail.com 2008 American Heart Association, Inc. Circulation is available at DOI: /CIRCULATIONAHA

2 706 Circulation August 12, 2008 Figure 1. The 2 main surgical techniques for BITA grafting. A, Composite T grafting with LITA to the LAD and free RITA to the marginal branches of circumflex (Cx-Marg). B, In situ grafting with RITA to the LAD and LITA to the circumflex. PDA indicates posterior descending artery. greater than that of the pedicled ITA, 11 allowing the use of both ITAs as grafts to all necessary coronary vessels. Harvesting the ITA as a skeletonized artery preserves the collateral blood supply to the sternum, enabling more rapid healing and a lower risk of infection. 9 In a previous study describing our experience with 545 patients, we showed that the use of skeletonized BITAs was not associated with an increased risk of sternal infection. 12 Increased risk of sternal infection in this preliminary report 12 was noted only in emergency cases or in patients with chronic obstructive pulmonary disease (COPD). More recent reports have described early and midterm results of the routine use of the skeletonized BITA grafting as the preferred revascularization procedure in our center between 1996 and These reports suggest that this surgical approach is safe, with a sternal infection rate comparable to that reported with a single ITA. 13,14 Most studies reporting results of BITA grafting include preselected patients operated on over a relatively long period. 3 5 Most patients in these series were not diabetic; they were preselected for this procedure according to their life expectancy; and only a few of them were 70 years of age when offered the option of BITA grafting. Unlike those series, in this series, we describe results in a nonselected group of patients. Complete arterial grafting with BITAs was the preferred method of myocardial revascularization for all ages during the 6-year study period. BITA grafting was performed in 58% of the patients referred for coronary artery bypass grafting (CABG) during this time period, and 40% of them were 70 years of age. The purpose of the present report is to describe a larger series with longer follow-up compared with our previous reports. Part of this cohort of the first 1000 patients was described in our previous reports. 13 We therefore expect this cohort to better reflect the advantages of BITA grafting and to define predictors of long-term survival more precisely. Methods From 1996 to 2001, 1515 consecutive patients underwent myocardial revascularization with BITAs that were dissected as skeletonized arteries. 15 BITAs made up 58% of the 2612 isolated CABG procedures performed in the Tel Aviv Sourasky Medical Center during this time. Surgical Technique Operations were performed with standard cardiopulmonary bypass (CPB) or off-pump coronary artery bypass. Myocardial preservation during CPB involved intermittent, antegrade, or retrograde blood cardioplegia (30 C to 32 C). Coronary stabilization during off-pump coronary artery bypass was facilitated by use of CTS stabilizers (Guidant, Curpentino, Calif) or the Octopus system (Medtronic, Minneapolis, Minn). ITAs were mobilized from the chest wall as skeletonized vessels. In all cases, BITAs were used to graft the left coronary system, ie, the myocardial territory supplied by the LAD and left circumflex arteries. Two arrangements were implemented: a free right ITA (RITA) attached proximally end to side on the LITA in a T-graft configuration (Figure 1A) and an in situ BITA with an anteaortic crossover RITA (Figure 1B). The choice of configuration was determined by previously detailed technical considerations. 15,16 We prefer to use BITAs as in situ grafts for myocardial revascularization. The 2 ITAs, combined with a right gastroepiploic artery, SVG, or radial artery, provide 3 sources of blood supply. When the right coronary artery (RCA) system is unsuitable for arterial grafting such as in cases with a potential for high competitive flow in the

3 Pevni et al Long-Term Results of BITA 707 RCA, we select the SVG as the conduit for revascularization of the RCA. We believe that more blood sources are associated with an improved long-term outcome. The cross arrangement (Figure 1B) is based on the assumption that the patency rate of the in situ RITA on the LAD coronary artery is similar to that of the in situ LITA on the LAD. 17 We do not use the cross technique in patients who have a short RITA, a very long ascending aorta, an enlarged right ventricle, or an LAD anastomotic site that is too distal or unpredictable. In most of the reported cases (1003 of the 1515 patients), we used the composite arterial grafting technique. The composite graft can be prepared before the patient is connected to CPB or before the first anastomosis (LAD) is constructed in off-pump coronary artery bypass cases. Most of the composite grafts included end-to-side anastomosis of the free RITA on an in situ LITA (Figure 1A). The free RITA can sometimes reach the anastomotic site of the RCA (posterolateral or patent ductus arteriosus). However, in most patients who required sequential anastomoses, the RITA was not long enough to reach the posterior descending artery, and we preferred using a third conduit (right gastroepiploic artery, SVG, or radial artery). The type of conduit selected for RCA grafting was not related to the configuration of the ITAs. Our strategy was to use RITAs, right gastroepiploic arteries, and radial arteries as grafts to the RCA branches only in the presence of a significant stenosis (ie, 80%). 18 To decrease the risk of spasm of the arterial grafts, we treated all of our patients with a high-dose intravenous infusion of isosorbide dinitrate (Isoket, Schwarz Pharma AG, Monheim, Germany; 4 to 20 mg/h) during the first 48 hours postoperatively. Calcium channel blockers (diltiazem 90 to 180 mg/d orally) were given to patients operated on using the right gastroepiploic artery or radial artery from the second postoperative day for at least 3 months. 15 Definition of Terms and Data Collection Patient data were collected retrospectively and analyzed according to the Society of Thoracic Surgeons National Cardiac Surgery database guidelines and definitions. A perioperative myocardial infarction was defined by the appearance of new Q waves in the ECG associated with elevated levels of creatine phosphokinase-mb fraction 50 U/L. A cerebrovascular accident was defined as a new permanent neurological deficit and computed tomographic evidence of cerebral infarction. Chronic renal failure was diagnosed if the creatinine level exceeded 1.8. Our definition of emergency operation is based on Society of Thoracic Surgeons guidelines and includes patients with ongoing angina, failed percutaneous transluminal coronary angioplasty, acute evolving myocardial infarction, pulmonary edema, and cardiogenic shock. For patients who needed emergency surgery and were not stabilized after intraaortic balloon counterpulsation, we usually used 1 ITA combined with SVGs. Deep sternal infection was defined as the sum of deep infection and late dehiscence requiring sternectomy. Follow-up was obtained after institutional board approval and patients consent by a telephone questionnaire and the national registry database. Follow-up was 100% complete. Statistical Analysis Data are expressed as mean SD or as a proportion. The 2 test and 2-sample t tests were used to compare discrete and continuous variables, respectively. Multivariable logistic regression analysis was used to predict early mortality, sternal infection, and stroke by various risk factors. The odds ratios (ORs) and 95% confidence intervals (CIs) are given. The Cox proportional-hazards model was used to evaluate the influence of preoperative variables, operative data, and early major postoperative complications on late and overall mortality. The Cox model also was used to compare adjusted survival between the various age groups after controlling for differences between groups in preoperative and operative characteristics. Results of Cox analysis are expressed as hazard ratios and 95% CI. Postoperative survival also is expressed by the Kaplan-Meier method, and survival curves were compared by the log-rank test. All analyses were performed with SPSS 14 software (SPSS Inc, Chicago, Ill). The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written. Results The patients preoperative and operative characteristics are presented in Tables 1 and 2. The mean number of grafts per patient was 3.1. Three hundred fifty-nine patients (23.6%) were operated on without CPB. The average CPB time for patients operated on with CPB was minutes, and aortic cross-clamping time was minutes. Data describing the technical aspects of the operative procedures performed are given in Table 2. Operative mortality (30 days postoperatively) in the cohort of 1515 consecutive BITA patients was 2.8%. Mortality of intraaortic balloon pump supported patients was 10.7% (8 of 75; P 0.001), and that of patients with COPD was 9.8% (8 of 82) (Table 3). Multivariable analysis (Table 4) revealed significantly increased operative mortality in emergency, COPD, and elderly patients and in patients referred to CABG with preoperative congestive heart failure. Reduced operative mortality was noted among BITA patients in whom right system revascularization was performed with an SVG. Sixteen perioperative infarcts occurred, and the occurrence of perioperative myocardial infarction was associated with increased operative (30-day) mortality (P 0.001). Sternal wound infection occurred in 24 patients (1.6%), and the mortality rate for patients with mediastinitis was 29% (7 of 24). Increased risk of sternal infection was found among patients with COPD (8 of 82, 9.8%), repeat operations (3 of 26, 11.5%), insulin-dependent diabetes mellitus (2 of 34, 5.9%), and non insulin-dependent diabetes mellitus (14 of 485, 2.9%). These preoperative patient characteristics also were found to be independent predictors for deep sternal infection in multivariable logistic regression analysis (Table 4). Forty-five patients sustained postoperative stroke. Independent predictors of postoperative strokes were peripheral vascular disease, left main disease, old age, and non insulindependent diabetes mellitus (Table 4). Follow-up between 60 and 141 months postoperatively revealed 305 late deaths. Ten-year actuarial survival (Kaplan- Meier) was 76 1%. Long-term survival of patients with composite T grafts was similar to that of patients with in situ grafts (Kaplan-Meier 10-year survival, 75 1% versus %). Ten-year survival rates of patients 55, 55 to 64, 65 to 74, and 75 years of age were %, %, %, and %, respectively. The relation between preoperative and operative patient characteristics and crude overall mortality rate estimates (ie, the number of dead divided by the total number of patients in the subgroup) is demonstrated in Tables 1 and 2. Cox-adjusted overall (early and late) survival for the different age groups of patients is demonstrated in Figure 2. Adjusted 10-year survival of patients 60 years of age is 90%. Significantly lower survival was noted in patients 75 years of age. Cox analysis of overall (early and late) mortality (Table 5) showed age, non insulin-dependent diabetes mellitus, ejection fraction 25%, congestive heart failure, chronic renal

4 708 Circulation August 12, 2008 Table 1. Patient Characteristics and Overall (Early and Late) Crude Mortality* Prevalence Mortality With Factor Mortality Without Factor Factor n % n % n % P Age 70 y Women Left main stenosis NS Acute MI 1 wk NS Hypertension NS Old MI PTCA EF 35% NS EF 25% CHF Preoperative IABP NS IDDM NS NIDDM PVD Chronic renal failure Severe COPD Emergency surgery NS 3-Vessel disease (vs 2-vessel disease) MI indicates myocardial infarction; PTCA, percutaneous transluminal coronary angiography; EF, ejection fraction; CHF, congestive heart failure; IABP, intraaortic balloon pump; IDDM, insulin-dependent diabetes mellitus; NIDDM, non insulin-dependent diabetes mellitus; and PVD, peripheral vascular disease. *The number of overall deaths was 348. The population evaluated was failure, COPD, peripheral vascular disease, emergency operation, and old myocardial infarction to be associated with increased late and overall mortality. Operations performed without CPB were associated with better long-term survival. Postoperative coronary angiography performed in 252 patients (most of them symptomatic) revealed an ITA patency rate of 90.8%. Discussion This report describes early and long-term results of skeletonized BITA grafting. The cohort evaluated includes patients typical of an urban population (relatively older), and this Table 2. Patient Operative Data and Overall Crude Mortality Prevalence (n 1515) technique was used in most of the patients (58%) who underwent CABG in our institution over the 6-year study period. In the early years (1996 to 1997) of our experience, the only contraindications for use of ITA grafts were emergency operations with hemodynamic instability requiring rapid connection to CPB. After careful analysis of our initial results, 19 COPD patients, diabetic females, and obese diabetics were added as absolute contraindications because of an increased risk of sternal infection in these subgroups. The immediate operative results in these early reports were comparable to those described in procedures in which 1 ITA was used. 20 Mortality With Factor Mortality Without Factor Factor n % n % n % P Repeat CABG Composite T graft NS Use of RGEA NS Use of SVG NS Off pump Sequential grafting NS 3 Grafts (vs 2 grafts) NS Perioperative MI RGEA indicates right gastroepiploic artery; MI, myocardial infarction.

5 Pevni et al Long-Term Results of BITA 709 Table 3. BITA and Early Mortality Prevalence* Early Mortality With Factor Early Mortality Without Factor Factor n % n % n % P Age 70 y CHF Acute MI IABP EF 25% EF 35% Emergency PVD COPD CRF CHF indicates congestive heart failure; MI, myocardial infarction; IABP, intraaortic balloon pump; EF, ejection fraction; PVD, peripheral vascular disease; and CRF, chronic renal failure. *The population evaluated was 1515 patients. The number of early deaths was 43. Later retrospective analyses of a larger cohort of BITA patients, including patients operated on during our early experience, 13,14 confer significant clinical approval of the basic assumption about the skeletonized ITA technique: It probably causes less damage to the sternal blood flow, 7,21,22 and therefore the rates of sternal infections and complications are low and not significantly different from those reported by others. 23 As in our recently published report defining independent predictors for sternal wound infection in 1000 BITA patients, we found in the present report that repeat operations, chronic lung disease, and diabetes mellitus are the major risk factors for deep sternal infection. 14 Explanations for the latter findings in patients with chronic lung disease might include high suture line pressure or collagen abnormalities described in smokers. 24 Despite a relatively small number of repeat operations in our report, we observed a surprisingly high incidence of deep sternal wound infections in the subset of patients undergoing repeat operation. This may be related to a decreased blood supply to the healing zone of a sternum undergoing resternotomy, but a definitive scientific explanation for the increased occurrence of sternal infection in redo procedures incorporating 2 skeletonized ITAs does not exist. Diabetes mellitus is generally considered to be a major risk factor for sternal complications, especially when bilateral internal mammary artery grafting is used. The risk in these circumstances was estimated to be 5 times greater than in other patients. 25 We found no evidence of this relationship in patients who received BITAs harvested as skeletonized vessels. Our results are even more definitive, considering the fact that 34% of our patients had diabetes mellitus. No difference was found in the occurrence of deep sternal wound infection among female and elderly patients compared with those without these risk factors. In a previous study, we Table 4. Multivariate Logistic Regression Analysis: Predictors of Early Mortality, Sternal Infection, and Postoperative Cerebrovascular Accident* Early Mortality Sternal Infection Postoperative CVA Variable OR 95% CI P OR 95% CI P OR 95% CI P Age 70 y Emergency operation COPD PVD Left main disease Repeat CABG IDDM NIDDM SVG to RCA CHF CVA indicates cerebrovascular accident (stroke); PVD, peripheral vascular disease; IDDM, insulin-dependent diabetes mellitus; NIDDM, non insulin-dependent diabetes mellitus; RCA, right coronary artery; and CHF, congestive heart failure. *The number of early deaths was 43. There were 24 sternal infections and 45 CVAs.

6 710 Circulation August 12, 2008 Figure 2. Cox-adjusted survival for the age groups. showed that Increased risk of infection among diabetic patients was noted only among obese patients (8.3% versus 1.1% in nonobese patients; P 0.03), especially obese females (15% versus 1.4% in diabetics without these risk factors; P 0.001). 19 In another study, the occurrence of sternal infection among insulin-treated patients with BITA was similar to that of insulin-treated patients with a single ITA (4% versus 2.7%; P 1.00). On the other hand, they had significantly lower rates of leg infection in the SVG harvesting site (0% versus 24%; P 0.000). 26 Unlike the results of this early report, in our present report with a larger number of Table 5. Cox Proportional-Hazards Model for Death (Early Plus Late) at Long-Term Follow-Up Variable HR 95% CI P Age y Age y Age 75 y Peripheral vascular disease COPD Repeat operation On/off pump CHF CRF Ejection fraction 25% Diabetes Old MI Emergency CHF indicates congestive heart failure; CRF, chronic renal failure; and MI, myocardial infarction. patients, insulin-treated diabetes emerged as an independent predictor of sternal infection. In a later report comparing BITA with single ITA grafting in an orally treated subset of diabetics, early outcome of BITA and single ITA was comparable, including the incidence of deep sternal infections (1.8% in both groups). During follow-up (range, 4 to 7.5 years; median, 5 years), fewer repeat revascularizations (4.4% versus 12.3%; P 0.025) were performed in the BITA group, and major adverse cardiac events also declined (11.2% versus 36.8%; P ). At 7 years, survival (Kaplan-Meier) (75% versus 59%; P 0.006, log-rank test), freedom from cardiac mortality (92% versus 68%; P ), and freedom from major adverse cardiac events (70% versus 59%; P 0.004) were superior in the BITA group. Multivariate analysis identified the use of BITA as a protective factor against the occurrence of late cardiac mortality (OR, 0.2; 95% CI, 0.06 to 0.6) and major adverse cardiac events (OR, 0.3; 95% CI, 0.1 to 0.66). 27 In our opinion, these better long-term results compared with CABG with a single ITA justified selective referral (excluding obese and female diabetics) of diabetics for skeletonized BITA grafting despite slightly but significantly increased risks of sternal infection, stroke, and late mortality. Our study clearly defines 3 subgroups of patients with increased operative and overall (early and late) mortality. Patients with peripheral vascular disease and diabetes probably have a diffuse and advanced form of atherosclerotic involvement of the heart and peripheral vessels. This may explain in part their unfavorable late results. 13,28 The increased early and overall mortality in the subset of patients with COPD is related in part to the increased early mortality observed in COPD patients who underwent sternectomy for sternal infection.

7 Pevni et al Long-Term Results of BITA 711 An interesting subgroup of the above cohort was the elderly patients. Six-hundred forty-one patients were 70 years of age (70 to 92 years of age). Three hundred ten of them (20.5%) were 75 years of age. Ten-year actuarial survival of patients 75 years of age was 52%, and age 75 years was found to be an independent predictor (Cox model) for overall (early plus late) mortality. Assessing the late survival of elderly patients undergoing CABG with 2 ITAs requires comparisons with an agematched population undergoing operation with a single ITA or percutaneous interventions. Such comparisons are the subject of planned future studies. The use of BITAs in the elderly is controversial. He et al 8 reported an operative mortality of 24% in elderly patients ( 70 years of age) who underwent BITA. Moreover, use of BITAs in the older patients in their report was found to be a major risk factor for operative mortality because mortality in the patients receiving 1 ITA was only 6.8% (P 0.007). It is important to note that the ITA in their report was used as a pedicled conduit, and as the authors stated, the fact that only 4% of the patients were grafted with BITAs might have explained the higher operative mortality and increased use of postoperative intraaortic balloon pump (16.2% versus 5.9%; P 0.015). In the study by Lytle and Loop, 3 the number of patients 60 years of age operated on with BITA was relatively small; however, BITA grafting improved survival of this subset of middle-aged and older patients compared with patients 60 years of age with a single ITA graft. The only large series (1467 patients) comparing BITA with a single ITA in the elderly was reported by Galbut et al. 24 In this study, patients with BITAs had lower hospital mortality rates (3.1%) compared with patients with a single ITA (6.4%), and the late survival rate (mean follow-up, 43 months) was also better (69.7% versus 60.7%). Little information is available in the literature about longterm results of BITA grafting without CPB. In a large series of 1802 skeletonized BITA patients, mid-term (42 months) mortality, cardiac mortality, reinterventions, and cardiac events were not significantly different between patients operated on with and without CPB. 29 In patients with atheromatous disease of the aorta (3000 patients propensity matched to 3000 conventional CPB patients), off-pump operations were associated with reduced risk of stroke and death. 30 The use of BITA grafting without aortic clamping can eliminate almost completely the risk of stroke in this subset of CABG patients with atheromatous disease of the aorta. 31,32 Our report is the first to show that avoiding CPB during BITA grafting is an independent predictor of better long-term survival. Study Limitations This is a retrospective cohort study, and several potential advantages of the BITA grafting technique could not be illustrated without a comparison of the mortality and morbidity with those of single ITA to the LAD (skeletonized and nonskeletonized) and SVG of other myocardial territories. Important patient data variables (eg, obesity and body mass index 30 kg/m 2 ) were not evaluated because of incomplete information on patient height and weight. Conclusions Routine use of skeletonized BITA for left-sided arterial grafting seems to be a safe technique for patients undergoing CABG, in terms of not only early and long-term mortality but also morbidity, especially sternal infection. Long-term results are better for BITA patients operated on without CPB. However, in patients with chronic lung disease, those with repeat operations, and obese and female diabetic patients, the risk of sternal infection is still unacceptably high; for them, we advocate the use of a single ITA plus SVGs instead of BITAs. None. Disclosures References 1. 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: Barner HB, Standeven JW, Reese J. Twelve-year experience with internal mammary artery for coronary artery bypass. J Thorac Cardiovasc Surg. 1985;90: Lytle BW, Loop FD. Superiority of bilateral internal thoracic artery grafting: it s been a long time comin. Circulation. 2001;104: Buxton BF, Komeda M, Fuller JA, Gordon I. Bilateral internal thoracic artery grafting may improve outcome of coronary artery surgery. Circulation. 1998;98(suppl II):II-1 II Schmidt SE, Jones JW, Thornby JI, Miller CC III, Beal AC Jr. Improved survival with multiple left-sided bilateral internal thoracic artery grafts. Ann Thorac Surg. 1997;64: Lytle BW, Blackstone EH, Loop FD, Houghtaling PL, Arnold JH, McCarthy PM, Cosgrove DM. Two internal thoracic artery grafts are better than one. J Thorac Cardiovasc Surg. 1999;117: Carrier M, Gregoire J, Tronc F, Cartier R, Leclerc Y, Pelletier YC. Effect of internal mammary artery dissection on sternal vascularization. Ann Thorac Surg. 1992;53: He GW, Acuff TE, Ryan WH, Mack MJ. Risk factors for operative mortality in elderly patients undergoing internal mammary artery grafting. Ann Thorac Surg. 1994;57: Sauvage LR, Wu HD, Kowalsky TE, Davis CC, Smith JC, Rittenhouse EA, Hall DG, Mansfield PB, Mathisen SR, Usui Y. Healing basis and surgical techniques for complete revascularization of the left ventricle using only the internal mammary arteries. Ann Thorac Surg. 1986;42: Cunningham JM, Gharavi MA, Fardin R, Meek RA. Considerations in the skeletonization technique of internal thoracic artery dissection. Ann Thorac Surg. 1992;54: Gurevitch J, Kramer A, Locker C, Shapira I, Paz Y, Matsa M, Mohr R. Technical aspects of double-skeletonized internal mammary artery grafting. Ann Thorac Surg. 2000;69: Sofer D, Gurevitch J, Shapira I, Paz Y, Matsa M, Kramer A, Mohr R. Sternal wound infection in patients after coronary artery bypass grafting using bilateral skeletonized internal mammary arteries. Ann Surg. 1999; 229: Kramer A, Mohr R, Lev-Ran O, Braunstein R, Pevni D, Locker C, Uretzky G, Shapira I. Midterm results of routine bilateral internal thoracic artery grafting. Heart Surg Forum. 2003;6: Pevni D, Mohr R, Lev-Ran O, Locer C, Paz Y, Kramer A, Shapira I. Influence of bilateral skeletonized harvesting on occurrence of deep sternal wound infection in 1,000 consecutive patients undergoing bilateral internal thoracic artery grafting. Ann Surg. 2003;237: Pevni D, Mohr R, Lev-Ran O, Paz Y, Kramer A, Frolkis I, Shapira I. Technical aspects of composite arterial grafting with double skeletonized internal thoracic arteries. Chest. 2003;123: 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:

8 712 Circulation August 12, Dion R, Etienne PY, Verhelst R, Khoury G, Rubay J, Bettendorf P, Hanet C, Wyns W. Bilateral mammary grafting. Eur J Cardiothorac Surg. 1993;7: Pevni D, Hertz I, Medalion B, Kramer A, Paz Y, Uretzky G, Mohr R. Angiographic evidence for reduced graft patency due to competitive flow in composite arterial T-grafts. J Thorac Cardiovasc Surg. 2007;133: Matsa M, Paz Y, Gurevitch J, Shapira I, Kramer A, Pevny D, Mohr R. Bilateral skeletonized internal thoracic artery grafts in patients with diabetes mellitus. J Thorac Cardiovasc Surg. 2001;121: Lytle BW, Cosgrove DM. Coronary artery bypass surgery. Curr Probl Surg. 1992;29: Arnold M. The surgical anatomy of sternal blood supply. J Thorac Cardiovasc Surg. 1972;64: Parish MA, Asai T, Grossi EA, Esposito R, Galloway AC, Colvin SB, Spencer FC. The effects of different techniques of internal mammary artery harvesting on sternal blood flow. J Thorac Cardiovasc Surg. 1992;104: Kouchoukaos NT, Wareing TH, Murphy SF, Pelate C, Marshall WG Jr. Risk of bilateral internal mammary artery bypass grafting. Ann Thorac Surg. 1990;49: Galbut DL, Traad EA, Dorman MI, DeWitt PL, Larsen PB, Weinstein D, Ally JM, Gentsch TO. Twelve-year experience with bilateral internal mammary artery grafts. Ann Thorac Surg. 1985;40: Lytle BW, Cosgrove DM, Loop FD, Borsh J, Goormastic M, Taylor PC. Perioperative risk of bilateral internal mammary artery grafting: analysis of 500 cases from 1971 to Circulation. 1986;74(suppl III):III- 37 III Lev-Ran O, Mohr R, Kramer A, Matsa M, Nesher N, Locker C, Uretzky G. Bilateral internal thoracic artery grafting in insulin-treated diabetics: should it be avoided? Ann Thorac Surg. 2003;75: Lev-Ran O, Braunstein R, Nesher N, Ben-Gal Y, Bolotin G, Uretzky G. Bilateral versus single internal thoracic artery grafting in oral-treated diabetic subsets: comparative seven-year outcome analysis. Ann Thorac Surg. 2004;77: Pevni D, Uretzky G, Paz Y, Ben-Gal Y, Shapira I, Nesher N, Braunstein R, Mohr R. Revascularization of the right coronary artery in bilateral internal thoracic artery grafting. Ann Thorac Surg. 2005;79: Calafiore AM, Di Mauro M, Canosa C, Cirmeni S, Iaco AL, Contini M, Mazzei V. Myocardial revascularization with and without cardiopulmonary bypass in multivessel disease: impact of strategy on midterm outcome. Ann Thorac Surg. 2003;76: Mishra M, Malhotra R, Karlekar A, Mishra Y, Trehan N. Propensity case-matched analysis of off-pump versus on-pump coronary artery bypass grafting in patients with atheromatous aorta. Ann Thorac Surg. 2006;82: Lev-Ran O, Braunstein R, Sharony R, Kramer A, Paz Y, Mohr R, Uretzky G. No-touch aorta off-pump coronary surgery: the effect on stroke. J Thorac Cardiovasc Surg. 2005;129: Lev-Ran O, Pevni D, Nesher N, Sharony R, Paz Y, Kramer A, Mohr R, Uretzky G. Off-pump coronary artery bypass grafting: single center experience with 1,000 consecutive cases. Isr Med Assoc J. 2004;6: CLINICAL PERSPECTIVE Recent studies have shown survival benefit and freedom from reintervention with the use of 2 internal thoracic arteries (ITAs) compared with a single ITA. However, in most of these studies, bilateral ITA (BITA) grafting is offered to only a selected group of nonurgent, nondiabetic young patients. Unlike those reports, our study describes long-term results of BITA grafting in nonselected patients. The study includes many elderly, emergency, and diabetic patients who would not otherwise be referred for BITA grafting. In most centers, the ITA is isolated from the chest wall as a pedicle, together with the vein, muscle, fat, and accompanying endothoracic fascia. This technique damages blood supply to the sternum, which in turn impedes sternal healing and exposes the sternum to the risks of early dehiscence and infection in operations involving both ITAs. The risk of sternal infection is particularly high in patients with preoperatively limited sternal blood supply such as the elderly and those with diabetes mellitus. Harvesting the ITA as a skeletonized artery preserves sternal collateral blood supply, thus enabling more rapid healing and lower risk of infection. We have found that skeletonized BITA grafting is associated with low morbidity and good long-term results. Use of skeletonized BITA was found to be an appropriate technique for the elderly and most patients with diabetes mellitus. However, in patients with chronic lung disease, in repeat operations, and in obese and female diabetic patients, the risk of sternal infection is still unacceptably high; for these patients, we advocate operations incorporating only a single ITA.