Advertisement

Current trends in bilateral internal thoracic artery use for coronary revascularization: Extending benefit to high-risk patients

Open ArchivePublished:February 22, 2018DOI:https://doi.org/10.1016/j.jtcvs.2018.01.094

      Abstract

      Background

      We sought to identify the trends in bilateral internal thoracic artery use and determine the degree to which the survival advantage of bilateral internal thoracic artery revascularization persists among perceived “high-risk” patients, compared with the use of left internal thoracic artery alone.

      Methods

      A retrospective review was conducted of patients who underwent isolated coronary artery bypass grafting for multivessel coronary artery disease at the Mayo Clinic between January 2000 and December 2015. Propensity score matching was performed between patients with bilateral internal thoracic artery and left internal thoracic artery alone grafts (1011 matched pairs). Effect of bilateral internal thoracic artery use on survival in “high-risk” patients (ejection fraction <40%, body mass index ≥30, age ≥70 years, diabetes, chronic lung disease, cerebrovascular accident) was evaluated.

      Results

      A total of 6468 isolated coronary artery bypass grafts were performed (5431 using left internal thoracic artery alone, 1037 using bilateral internal thoracic artery). There was an increasing trend in bilateral internal thoracic artery use (P value for linear trend = .005), with the percentage of coronary artery bypass grafting cases with bilateral internal thoracic artery doubling over the last 4 years (13% in 2012 to 27% in 2015). Propensity-matched comparisons showed a survival advantage for bilateral internal thoracic artery (hazard ratio, 0.81; 95% confidence interval, 0.66-0.99; P = .043). Risk of deep sternal wound infection, although higher in the bilateral internal thoracic artery group, was not significant (1.2% vs 0.5%; P = .088). None of the “high-risk” subsets of patients showed an adverse effect of bilateral internal thoracic artery on survival.

      Conclusions

      Bilateral internal thoracic artery use in coronary artery bypass grafting is increasing over time. There is a consistent survival benefit with bilateral internal thoracic artery use, extending to patients with higher-risk comorbidities, suggesting the need for further expansion in use of this technique.

      Key Words

      Abbreviations and Acronyms:

      ART (Arterial Revascularization Trial), BITA (bilateral internal thoracic artery), BMI (body mass index), CABG (coronary artery bypass grafting), CAD (coronary artery disease), CI (confidence interval), CLD (chronic lung disease), CPB (cardiopulmonary bypass), EF (ejection fraction), HR (hazard ratio), ICU (intensive care unit), ITA (internal thoracic artery), LAD (left anterior descending), LITA (left internal thoracic artery), OR (odds ratio), PH (proportional hazard), RITA (right internal thoracic artery), STS (Society of Thoracic Surgeons), SVG (saphenous vein graft)
      BITA use has a survival advantage over the use of LITA alone, and this benefit extends to the perceived high-risk groups.
      Perceived high-risk factors, such as low EF, obesity, CLD, advancing age, and diabetes, have prevented surgeons to adopt BITA use extensively. We used a propensity score model to compare BITA and LITA-alone use to identify if BITA use offers a survival advantage or not, especially in perceived high-risk patients.
      See Editorial Commentary page 2344.
      Coronary artery disease (CAD) accounts for approximately 1 of every 7 deaths in the United States, with approximately 660,000 Americans experiencing a coronary event every year.
      • Mozaffarian D.
      • Benjamin E.J.
      • Go A.S.
      • Arnett D.K.
      • Blaha M.J.
      • Cushman M.
      • et al.
      Heart disease and stroke statistics-2016 update: a report from the American Heart Association.
      Coronary artery bypass grafting (CABG) has been established as an optimal treatment strategy for patients with multivessel CAD disease.
      • Taggart D.P.
      CABG is still the best treatment for multivessel and left main disease, but patients need to know.
      The traditional revascularization strategy for a multivessel CABG involves placing a left internal thoracic artery (LITA) graft to the left anterior descending (LAD) artery with the use of reverse saphenous vein graft (SVG) to bypass any other diseased vessels. The use of LITA to LAD in CABG has been shown to offer a survival benefit to patients when compared with SVG.
      • Loop F.D.
      • Lytle B.W.
      • Cosgrove D.M.
      • Stewart R.W.
      • Goormastic M.
      • Williams G.W.
      • et al.
      Influence of the internal-mammary-artery graft on 10-year survival and other cardiac events.
      The angiographic patency rates of internal thoracic artery (ITA) at 10 years are higher (≈90%-95%) compared with vein grafts (≈50%-75%).
      • Loop F.D.
      • Lytle B.W.
      • Cosgrove D.M.
      • Stewart R.W.
      • Goormastic M.
      • Williams G.W.
      • et al.
      Influence of the internal-mammary-artery graft on 10-year survival and other cardiac events.
      • Cameron A.
      • Davis K.B.
      • Green G.
      • Schaff H.V.
      Coronary bypass surgery with internal-thoracic-artery grafts—effects on survival over a 15-year period.
      • Fitzgibbon G.M.
      • Kafka H.P.
      • Leach A.J.
      • Keon W.J.
      • Hooper D.
      • Burton J.R.
      • et al.
      Coronary bypass graft fate and patient outcome: angiographic follow-up of 5,065 grafts related to survival and reoperation in 1,388 patients during 25 years.
      Over the past decade, several observational studies have shown that the use of a bilateral internal thoracic artery (BITA) revascularization strategy offers a survival advantage over the use of LITA alone for multivessel CAD.
      • Taggart D.P.
      • D'Amico R.
      • Altman D.G.
      Effect of arterial revascularization on survival: a systematic review of studies comparing bilateral and single internal mammary arteries.
      • Takagi H.
      • Goto S.N.
      • Watanabe T.
      • Mizuno Y.
      • Kawai N.
      • Umemoto T.
      A meta-analysis of adjusted hazard ratios from 20 observational studies of bilateral versus single internal thoracic artery coronary artery bypass grafting.
      However, despite the growing evidence in support of this technique, only 4% to 12% of all CABG operations worldwide use a BITA approach.
      • El-Bardissi A.W.
      • Aranki S.F.
      • Sheng S.
      • O'Brien S.M.
      • Greenberg C.C.
      • Gammie J.S.
      • et al.
      Trends in isolated coronary artery bypass grafting: an analysis of the Society of Thoracic Surgeons adult cardiac surgery database.
      • Kappetein A.P.
      • Dawkins K.D.
      • Mohr F.W.
      • Morice F.C.
      • Mack M.J.
      • Russell M.E.
      • et al.
      Current percutaneous coronary intervention and coronary artery bypass grafting practices for three-vessel and left main coronary artery disease. Insights from the SYNTAX run-in phase.
      • Yan B.P.
      • Clark D.J.
      • Buxton B.
      • Ajani A.E.
      • Smith J.A.
      • Duffy S.J.
      • et al.
      on behalf of the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS), the Melbourne Interventional Group (MIG)
      Clinical characteristics and early mortality of patients undergoing coronary artery bypass grafting compared to percutaneous coronary intervention: insights from the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS) and the Melbourne interventional group (MIG) registries.
      This opposition to BITA grafting is largely based on the perception of increased perioperative risk, in particular the risk of sternal wound infection and myocardial and respiratory morbidity. The perceived “high-risk” factors described in the literature include older age, female sex, diabetes mellitus, obesity, chronic lung disease (CLD), and reduced ejection fraction (EF).
      • Kouchoukos N.T.
      • Wareing T.H.
      • Murphy S.F.
      • Pelate C.
      • Marshall Jr., W.G.
      Risks of bilateral internal mammary artery bypass grafting.
      • Vallely M.P.
      • Edelman J.J.B.
      • Wilson M.K.
      Bilateral internal mammary arteries: evidence and technical consideration.
      • Mohammadi S.
      • Dagenais F.
      • Doyle D.
      • Mathieu P.
      • Baillot R.
      • Charbonneau E.
      • et al.
      Age cut-off for the loss of benefit from bilateral internal thoracic artery grafting.
      • Taggart D.P.
      Respiratory dysfunction after cardiac surgery: effects of avoiding cardiopulmonary bypass and the use of bilateral internal mammary arteries.
      • Matsa M.
      • Paz Y.
      • Gurevitch J.
      • Shapira I.
      • Kramer A.
      • Pevny D.
      • et al.
      Bilateral skeletonized internal thoracic artery grafts in patients with diabetes mellitus.
      • Sofer D.
      • Gurevitch J.
      • Shapira I.
      • Paz Y.
      • Matsa M.
      • Kramer A.
      • et al.
      Sternal wound infections in patients after coronary artery bypass grafting using bilateral skeletonized internal mammary arteries.
      • Kurlansky P.A.
      • Traad E.A.
      • Dorman M.J.
      • Galbut D.L.
      • Zucker M.
      • Ebra G.
      • et al.
      Efficacy of single versus bilateral internal mammary artery grafting in women: a long-term study.
      The recently published midterm (<5 years) results of the Arterial Revascularization Trial (ART) showing no survival benefit with BITA has brought back into contention the need for an additional ITA in CABG.
      • Taggart D.P.
      • Altman D.G.
      • Gray A.M.
      • Lees B.
      • Gerry S.
      • Benedetto U.
      • et al.
      ART Investigators
      Randomized trial of bilateral versus single internal-thoracic-artery grafts.
      We sought to identify in our practice trends in BITA use over time and to evaluate the impact of patients' comorbidities on the relative benefit of BITA compared with the use of LITA alone.

      Materials and Methods

      A total of 9084 patients underwent CABG between January 2000 and December 2015 at our institute. Of these, 6468 patients had isolated CABG for multivessel CAD. A retrospective review of all isolated CABG for multivessel CAD was done after obtaining the necessary approval from the Institutional Review Board, Mayo Clinic. The requirement for individual patient consent was waived because of the study designation as a minimal-risk retrospective study.
      All patients who received 2 or more bypass grafts were included in the study. Patients with single bypass graft or any concomitant cardiac surgery were excluded. Patients were divided into 2 groups. The BITA group (n = 1037) included all patients in whom BITA grafts were used, with or without the use of SVG (55.8%) or radial artery (12.8%). The LITA-alone group (n = 5431) included all patients in whom LITA alone was used with no use of the right internal thoracic artery (RITA). In this group, SVG (96.3%) and radial artery (7.7%) accounted for the additional bypass grafts. Both skeletonized and pedicled techniques were used for ITA harvesting. Within the BITA group, surgical strategy for grafting included 4 patterns: (1) in situ LITA to LAD with free RITA anastomosed in a T-composite configuration off the side of the LITA to the circumflex and right coronary territories as sequential grafts; (2) in situ LITA to LAD, in situ RITA to circumflex territory via transverse sinus, with or without SVG or radial artery to right coronary territory; (3) in situ RITA to LAD with in situ LITA to circumflex territory, with or without SVG or radial artery graft to right coronary territory; and (4) in situ LITA to LAD, in situ RITA to right coronary territory, with or without SVG or radial artery to circumflex territory. The grafting strategy in the LITA-alone group was always LITA to LAD, with SVG or radial artery graft to the circumflex and right coronary territory. Most of the cases were performed with the use of cardiopulmonary bypass (CPB), whereas few were performed with the off-pump technique (4.7% and 5.8% in the BITA and LITA-alone groups, respectively).
      Patient data were analyzed according to the Society of Thoracic Surgeons (STS) National Cardiac Surgery Database guidelines and definitions.

      Society of Thoracic Surgeons. STS adult cardiac surgery database data specifications version 2.81. March 28, 2014. Available at: https://www.sts.org/sites/default/files/documents/ACSD_DataSpecificationsV2_81.pdf. Accessed March 9, 2018.

      The primary outcome of interest was long-term survival after surgery in the 2 study groups defined by BITA or LITA-alone use during CABG. To minimize treatment selection bias, propensity score matching of BITA and LITA-alone patients was performed to produce statistically comparable groups with respect to baseline characteristics: age, gender, smoking status, diabetes, body mass index (BMI), peripheral vascular disease, cerebrovascular accident, CLD, renal failure, EF, left main coronary disease greater than 50%, number of diseased coronary vessels, and nonelective surgery. A total of 1011 matched pairs were ultimately identified. Secondary outcomes included operative characteristics, including CPB and crossclamp times, given the potential for these variables to influence conduit selection, as well as postoperative complications. Analysis of both primary and secondary outcomes emphasized the comparison of matched groups, which were supplemented with risk-adjusted modeling to determine independent effects of BITA use. The perceived “high-risk” preoperative factors were defined as EF less than 40%, BMI 30 kg/m2 or more, age 70 years or more, diabetes, CLD, and cerebrovascular accident. BITA use in patients with “high-risk” factors was evaluated for a differential effect on survival.

       Statistical Methods

      Baseline characteristics are shown as number (percent) for categoric variables, median (interquartile range, range) for continuous variables, or mean (interquartile range, range) for ordinal variables. Baseline differences between BITA and LITA-alone surgery groups were determined using Pearson chi-square tests for categoric variables, Wilcoxon rank-sum tests for continuous variables, and Cochran–Armitage proportion trend tests for ordinal variables. Because any 1 of these factors could have influenced the decision to perform the surgery with BITA versus LITA-alone, differences in outcomes between the 2 surgery strategies are susceptible to treatment selection bias. To compensate for the potential confounding from nonrandom treatment selection, propensity score analysis was undertaken in the form of case-matching to produce comparable BITA and LITA-alone groups at the time of surgery. As the basis of matching, the propensity score is a probabilistic prediction of whether the patient underwent BITA and was derived using multivariable logistic regression. All available baseline covariates were included in this propensity model, with continuous covariates modeled flexibly with restricted cubic splines to permit nonlinear effects and any missing data imputed before the analysis. Specifically, we used single imputation based on the Markov chain Monte Carlo method to predict missing values of a target variable using all available data on other baseline covariates. The similarity of baseline characteristics between groups after propensity score matching was assessed to support the validity for comparing the 2 surgery strategies. To this end, absolute standardized differences are reported to express the group comparability both before and after matching.
      Comparison of clinical and patient outcomes between propensity-matched groups was achieved by the statistical tests indicated earlier (chi-square test, Wilcoxon rank-sum test) or by survival analysis based on the Kaplan–Meier method with the log-rank test. To assess differences in operative time and intensive care unit (ICU) length of stay or readmission, subset analysis was performed on calendar years 2002 to 2015 because these data were not collected in the STS before then. To assess the independent association of BITA with outcomes after adjustment for relevant baseline covariates, logistic regression models were fitted for binary outcomes and Cox proportional hazards (PH) regression models were fitted for survival outcomes. In the risk-adjusted model for survival time, we included all baseline variables used in the propensity score, in addition to the score itself, to determine whether our finding was robust to comprehensive adjustment. Evidence of heterogeneity in our study results for prespecified, clinically relevant subgroups was investigated by adding a BITA-by-subgroup interaction term to the outcome model and testing whether the treatment effect differed in these high-risk subsets of patients compared with all others. We further examined the sensitivity of our case-matched analysis to selection bias by refitting the survival models on the comprehensive series of patients with the propensity score used as a covariate to adjust for treatment selection. All analyses were performed using SAS software version 9.4 (SAS Institute Inc, Cary, NC).

      Results

      Among isolated CABG surgeries performed over the 16-year period, the percentage of CABGs with BITA use showed a significantly increasing temporal trend (P value for linear trend = 0.005) (Figure 1). Although the percentage of CABGs with BITA remained relatively constant from 2000 to 2012 (range from 12% to 18% of CABG cases per year), the percentage of CABGs with BITA nearly doubled in the last 4 years (2012-2015, 13%-27%, respectively).
      Figure thumbnail gr1
      Figure 1Trends in use of BITA and LITA (alone) in isolated CABG performed from 2000 to 2015. CABG, Coronary artery bypass grafting; LITA, left internal thoracic artery; BITA, bilateral internal thoracic artery.
      Baseline characteristics of the comprehensive and case-matched series of BITA and LITA-alone patients are shown in Tables 1 and 2, respectively. Distribution of propensity score by group in comprehensive and case-matched series of patients is shown in Figure 2. Significant differences were found between the original LITA-alone and BITA groups before matching. Patients receiving LITA-alone were older, more often female, and more likely to have diabetes, peripheral vascular disease, hypertension, reduced EF, renal failure, and nonelective CABG.
      Table 1Baseline characteristics in original sample
      VariableNLITA alone (n = 5431)BITA (n = 1037)P valueStandardized difference
      Age at surgery
      Continuous variables are described by percentile values: median (IQR) [range].
      646870.4 (63.3, 76.7) [28.9, 96.9]59.2 (53.3, 65.6) [24.6, 86.2]<.0011.063
      Female64681271 (23.4%)150 (14.5%)<.0010.230
      Current smoker6455686 (12.7%)226 (21.9%)<.0010.245
      Diabetes64682075 (38.2%)217 (20.9%)<.0010.386
      BMI
      Continuous variables are described by percentile values: median (IQR) [range].
      646429.4 (26.3, 33.3) [17.0, 69.4]29.1 (26.4, 32.6) [15.1, 58.3].0330.114
      Congestive heart failure6468679 (12.5%)33 (3.2%)<.0010.352
      Peripheral vascular disease6467889 (16.4%)107 (10.3%)<.0010.179
      Cerebrovascular accident6461377 (6.9%)40 (3.9%)<.0010.137
      Cerebrovascular disease6468936 (17.2%)85 (8.2%)<.0010.274
      CLD6465638 (11.8%)70 (6.8%)<.0010.173
      Hypertension64684571 (84.2%)771 (74.3%)<.0010.244
      Dyslipidemia64674975 (91.6%)983 (94.8%)<.0010.126
      EF
      Continuous variables are described by percentile values: median (IQR) [range].
      600556 (45, 63) [13, 89]59 (51, 64) [13, 87]<.0010.275
      Left main coronary disease (≥50%)63592038 (38.1%)352 (35.0%).0600.065
      Recent atrial fibrillation/flutter6363321 (6.0%)29 (2.8%)<.0010.156
      NYHA classification
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      62373.0 (2, 4) [1, 4]2.8 (2, 4) [1, 4]<.001
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      0.226
      Renal failure6467320 (5.9%)20 (1.9%)<.0010.206
      Dialysis646885 (1.6%)5 (0.5%).0060.108
      Year of surgery
      Continuous variables are described by percentile values: median (IQR) [range].
      64682006 (2002, 2010) [2000, 2015]2006 (2003, 2012) [2000, 2015].0050.105
      Any prior cardiac operations6468108 (2.0%)12 (1.2%).0690.067
      Previous CABG646888 (1.6%)9 (0.9%).0680.068
      Previous valve surgery646815 (0.3%)1 (0.1%).2860.042
      Creatinine level
      Continuous variables are described by percentile values: median (IQR) [range].
      64611.1 (1.0, 1.3) [0.3, 14.2]1.1 (0.9, 1.2) [0.6, 12.3]<.0010.179
      No. of diseased vessels6468.0050.094
       21015 (18.7%)233 (22.5%)
       34416 (81.3%)804 (77.5%)
      Mitral stenosis646738 (0.7%)0 (0.0%).0070.119
      Tricuspid regurgitation
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      64680.8 (0.0, 1.0) [0, 4]0.6 (0.0, 1.0) [0, 3]<.001
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      0.259
      Surgery status6468<.001
       Elective3192 (58.8%)705 (68.0%)0.192
       Urgent2103 (38.7%)325 (31.3%)0.155
       Emergency136 (2.5%)7 (0.7%)0.147
      N is the total number of nonmissing responses for each variable (proportion of subjects missing data ranged from 0% to 7% for all baseline variables). Unless stated otherwise, differences between groups were analyzed with Pearson chi-square test for categoric variables and Wilcoxon rank-sum test for continuous variables. Absolute standardized differences are used to express the baseline comparability between groups for each variable before propensity matching. LITA, Left internal thoracic artery; BITA, bilateral internal thoracic artery; BMI, body mass index; CLD, chronic lung disease; EF, ejection fraction; NYHA, New York Heart Association; CABG, coronary artery bypass grafting.
      Continuous variables are described by percentile values: median (IQR) [range].
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      Table 2Baseline characteristics in matched sample
      VariableNLITA alone (n = 1011)BITA (n = 1011)P valueStandardized difference
      Age at surgery
      Continuous variables are described by percentile values: median (IQR) [range].
      202259.4 (53.7, 65.6) [31.8, 91.0]59.6 (53.7, 65.8) [24.6, 86.2].9010.024
      Female2022154 (15.2%)149 (14.7%).7550.014
      Current smoker2017223 (22.1%)221 (21.9%).9240.004
      Diabetes2022215 (21.3%)217 (21.5%).9140.005
      Body mass index
      Continuous variables are described by percentile values: median (IQR) [range].
      201929.4 (26.2, 33.0) [18.1, 64.7]29.2 (26.5, 32.6) [15.1, 58.3].6590.030
      Congestive heart failure202236 (3.6%)33 (3.3%).7130.016
      Peripheral vascular disease202298 (9.7%)105 (10.4%).6040.023
      Cerebrovascular accident202141 (4.1%)39 (3.9%).8160.010
      Cerebrovascular disease202283 (8.2%)84 (8.3%).9360.004
      CLD202268 (6.7%)67 (6.6%).9290.004
      Hypertension2022777 (76.9%)758 (75.0%).3230.044
      Dyslipidemia2022962 (95.2%)960 (95.0%).8370.009
      EF
      Continuous variables are described by percentile values: median (IQR) [range].
      185358 (51, 64) [13, 89]59 (50, 64) [13, 87].9700.000
      Left main coronary disease (≥50%)1970353 (35.8%)346 (35.2%).7930.012
      Recent atrial fibrillation/flutter200230 (3.0%)28 (2.8%).7840.012
      NYHA classification
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      19222.8 (2, 4) [1, 4]2.8 (2, 4) [1, 4].546
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      0.028
      Renal failure202115 (1.5%)20 (2.0%).3960.038
      Dialysis20224 (0.4%)5 (0.5%).7380.015
      Year of surgery
      Continuous variables are described by percentile values: median (IQR) [range].
      20222006 (2002, 2011) [2000, 2015]2006 (2003, 2011) [2000, 2015].4330.034
      Any prior cardiac operations202214 (1.4%)12 (1.2%).6930.018
      Previous CABG202211 (1.1%)9 (0.9%).6530.020
      Previous valve surgery20221 (0.1%)1 (0.1%)>.9990.000
      Creatinine level
      Continuous variables are described by percentile values: median (IQR) [range].
      20181.1 (0.9, 1.2) [0.3, 9.5]1.1 (0.9, 1.2) [0.6, 12.3].7960.013
      No. of diseased vessels2022.1730.061
       2251 (24.8%)225 (22.3%)
       3760 (75.2%)786 (77.7%)
      Mitral stenosis20210 (0.0%)0 (0.0%)
      Tricuspid regurgitation
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      20220.6 (0, 1) [0, 4]0.6 (0, 1) [0, 3].825
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      0.010
      Surgery status2022.801
       Elective670 (66.3%)684 (67.7%)0.029
       Urgent334 (33.0%)320 (31.7%)0.030
      Emergency7 (0.7%)7 (0.7%)0.000
      N is the total number of nonmissing responses for each variable (proportion of subjects missing data ranged from 0% to 8% for all baseline variables). Unless stated otherwise, differences between groups were analyzed with Pearson chi-square test for categoric variables and Wilcoxon rank-sum test for continuous variables. Absolute standardized differences are used to express the baseline comparability between groups for each variable after propensity matching. LITA, Left internal thoracic artery; BITA, bilateral internal thoracic artery; CLD, chronic lung disease; EF, ejection fraction; NYHA, New York Heart Association; CABG, coronary artery bypass grafting.
      Continuous variables are described by percentile values: median (IQR) [range].
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      Figure thumbnail gr2
      Figure 2Distribution of propensity score by group in comprehensive and case-matched series of patients. PS, Propensity score; LITA, left internal thoracic artery; BITA, bilateral internal thoracic artery.
      The BITA group had significantly lower use of additional SVGs (56.0% vs 92.3%; P < .001) and radial arteries (12.5% vs 16.4%; P < .001). Compared with the LITA-alone group, crossclamp time was longer in the BITA group (P = .012), whereas CPB time was shorter (P = .011). The mean number of grafts used was 3.3 in the BITA group compared with 3.1 in the LITA-alone group (P < .001) and ranged from 2 to 6 grafts in both groups. In analysis restricted to the time frame with available data (82%), comparison of treatments in 2002 to 2015 suggested a longer operative time in patients who underwent BITA versus LITA alone (Table 3).
      Table 3Comparison of operative and postoperative measures between matched groups
      VariableNLITA alone (n = 1011)BITA (n = 1011)P value
      Use of SVG2022933 (92.3%)566 (56.0%)<.001
      Use of radial grafts2022166 (16.4%)126 (12.5%).011
      Crossclamp time
      Continuous variables are described by percentile values: median (IQR) [range].
      196853 (40, 66) [0, 128]55 (41, 68) [0, 143].012
      Perfusion time
      Continuous variables are described by percentile values: median (IQR) [range].
      196778 (63, 98) [0, 267]75 (59, 97) [0, 213].011
      Incision time (min)
      Continuous variables are described by percentile values: median (IQR) [range].
      1667254 (221, 293) [39, 1426]298 (251, 341) [121, 1134]<.001
      No. of grafts used
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      20223.1 (3, 4) [2, 6]3.3 (3, 4) [2, 6]<.001
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      Hospital length of stay, d
      Continuous variables are described by percentile values: median (IQR) [range].
      20225 (5, 6) [0, 75]5 (5, 6) [0, 41].898
      Hospital readmission within 30 d198979 (7.9%)75 (7.5%).742
      ICU length of stay, h
      Continuous variables are described by percentile values: median (IQR) [range].
      166723 (20, 28) [0, 1236]22 (19.5, 25) [5, 631].003
      ICU readmission within 30 d166717 (2.0%)14 (1.7%).575
      Complication-prolonged ventilation202244 (4.4%)43 (4.3%).913
      Complication-sternal infection, superficial or deep202214 (1.4%)31 (3.1%).010
      Complication-sternal infection, deep20225 (0.5%)12 (1.2%).088
      Death within 30 d or during hospitalization20226 (0.6%)3 (0.3%).316
      Survival over follow-up
      Survival analysis was by the Kaplan–Meier method with the log-rank test; tabulated values are percentages of patients alive (cumulative number of deaths) at each follow-up interval for each group.
      2022.031
       1 y98.0% (18)98.6% (13)
       5 y91.8% (65)93.8% (47)
       10 y79.5% (132)82.4% (107)
       Total No. of deaths203168
      N is the total number of nonmissing responses. The proportion of subjects missing data was minimal (range, 0%-3%), with the exception of incision time and ICU length of stay and readmission (STS data collection for these variables started in 2002, resulting in 18% of these subjects missing data). Unless stated otherwise, differences between groups were analyzed with Pearson chi-square test for categoric variables and Wilcoxon rank-sum test for continuous variables. LITA, Left internal thoracic artery; BITA, bilateral internal thoracic artery; SVG, saphenous vein graft; ICU, intensive care unit.
      Continuous variables are described by percentile values: median (IQR) [range].
      Ordinal variables presented as mean (IQR) [range] and analyzed by Cochran–Armitage proportion trend test.
      Survival analysis was by the Kaplan–Meier method with the log-rank test; tabulated values are percentages of patients alive (cumulative number of deaths) at each follow-up interval for each group.
      The percentage of patients who underwent prolonged ventilation after surgery was similar between the 2 groups. There was also no significant difference in the duration of hospital stay or in the number of patients readmitted within 30 days of surgery (Table 3). However, for the matched subset between 2002 and 2015, patients with BITA had a slightly shorter length of ICU stay (median, 22 vs 23 hours; P = .003).
      The incidence of overall sternal-site infection (superficial or deep) was higher in the BITA group than in the LITA-alone group (3.1% vs 1.4%; P = .010), with the odds of infection increased 2-fold with BITA (odds ratio [OR], 2.28; 95% confidence interval [CI], 1.21-4.32). This higher risk among the BITA group persisted after adjustment for suspected risk factors, including sex, BMI, diabetes, CLD, and year of CABG (OR, 2.39, 95% CI, 1.25-4.57; P = .008). However, the percentage of patients with deep sternal wound infections was not significantly higher in the BITA group compared with the LITA-alone group (1.2% vs 0.5%; P = .088).
      There was no significant difference in operative mortality between the matched BITA and LITA-alone groups (0.3% vs 0.6%; P = .316). However, survival over long-term follow-up was significantly improved in the BITA group (1, 5, and 10-year survival estimates: 98.6%, 93.8%, 82.4%, respectively) relative to the LITA-alone group (98.0%, 91.8%, 79.5%, respectively; P = .031) (Figure 3). In risk-adjusted models, BITA was associated with a 19% risk reduction in death after adjustment for propensity alone (hazard ratio [HR], 0.81; 95% CI, 0.66-0.99) and a 26% risk reduction after additional adjustment for all baseline covariates used in the propensity model (HR, 0.74; 95% CI, 0.60-0.92). Sensitivity analysis performed on the larger cohort with propensity score used as a covariate produced results similar to those from the case-matched analysis, with patients in the BITA group having favorable survival (20%-22% risk reduction). Survival analysis on the treatment interaction between BITA versus LITA-alone and “high-risk” patient subsets revealed no evidence of heterogeneity in our results. The relative efficacy of BITA versus LITA-alone on survival was not significantly different between high- and low-risk levels defined by age (P value for interaction = .952), EF (P = .330), BMI (P = .871), gender (P = .681), diabetes (P = .433), cerebrovascular accident (P = .651), CLD (P = .081), and nonelective surgery (P = .692) (Figure 4).
      Figure thumbnail gr3
      Figure 3Survival analysis of propensity score–matched groups. BITA, Bilateral internal thoracic artery; LITA, left internal thoracic artery; PS, propensity score; BL, baseline; CI, confidence interval.
      Figure thumbnail gr4
      Figure 4Association of BITA with survival time in clinically relevant subgroups. HR, Hazard ratio; CI, confidence interval; BMI, body mass index; EF, ejection fraction; CVA, cerebrovascular accident; CLD, chronic lung disease; CABG, coronary artery bypass grafting.

      Discussion

      This study, involving a large cohort of patients with isolated multivessel CAD, shows that the BITA use rate in our institution is increasing, with 27% of multivessel CABGs performed in 2015 using BITA grafts. Furthermore, the survival advantage of BITA compared with LITA-alone was shown with a 19% risk reduction in death. None of the “high-risk” subsets of patients showed an adverse effect of BITA on survival (Video 1).
      Figure thumbnail fx2
      Taggart and colleagues
      • Taggart D.P.
      • Altman D.G.
      • Gray A.M.
      • Lees B.
      • Gerry S.
      • Benedetto U.
      • et al.
      ART Investigators
      Randomized trial of bilateral versus single internal-thoracic-artery grafts.
      published the 5-year results of the randomized trial of BITA versus single ITA grafts. The results showed no significant difference between those receiving single ITA grafts and those receiving BITA grafts with regard to mortality or the rates of cardiovascular events at 5 years of follow-up. However, the BITA group had more sternal wound complications.
      • Taggart D.P.
      • Altman D.G.
      • Gray A.M.
      • Lees B.
      • Gerry S.
      • Benedetto U.
      • et al.
      ART Investigators
      Randomized trial of bilateral versus single internal-thoracic-artery grafts.
      This has again raised the debate about the superiority of BITA grafting and the perceived high perioperative risks that has prevented a more liberal use of BITA worldwide. This perception stemmed from previous reports showing BITA use to be associated with the need for prolonged postoperative ventilatory support
      • Knapik P.
      • Spyt T.J.
      • Richardson J.B.
      • McLellan I.
      Bilateral and unilateral use of internal thoracic artery for myocardial revascularization. Comparison of extubation outcome and duration of hospital stay.
      ; increased operative time
      • Taggart D.P.
      • Altman D.G.
      • Gray A.M.
      • Lee B.
      • Nugara F.
      • Yu L.
      • et al.
      Randomized trial to compare bilateral vs single internal mammary coronary artery bypass grafting: 1-year results of the arterial revascularisation trial (ART).
      ; higher sternal infection rates, especially in female patients and patients with diabetes, obesity, and CLD
      • Taggart D.P.
      CABG is still the best treatment for multivessel and left main disease, but patients need to know.
      • Kouchoukos N.T.
      • Wareing T.H.
      • Murphy S.F.
      • Pelate C.
      • Marshall Jr., W.G.
      Risks of bilateral internal mammary artery bypass grafting.
      • Matsa M.
      • Paz Y.
      • Gurevitch J.
      • Shapira I.
      • Kramer A.
      • Pevny D.
      • et al.
      Bilateral skeletonized internal thoracic artery grafts in patients with diabetes mellitus.
      • Sofer D.
      • Gurevitch J.
      • Shapira I.
      • Paz Y.
      • Matsa M.
      • Kramer A.
      • et al.
      Sternal wound infections in patients after coronary artery bypass grafting using bilateral skeletonized internal mammary arteries.
      ; and lack of survival benefit in elderly patients.
      • Mohammadi S.
      • Dagenais F.
      • Doyle D.
      • Mathieu P.
      • Baillot R.
      • Charbonneau E.
      • et al.
      Age cut-off for the loss of benefit from bilateral internal thoracic artery grafting.
      • Benedetto U.
      • Amrani M.
      • Raja S.G.
      Harefield Cardiac Outcomes Research Group
      Guidance for the use of bilateral internal thoracic arteries according to survival benefit across age groups.
      This has prevented surgeons worldwide from adopting BITA as a preferred grafting strategy, even in patients who could presumably benefit from it.
      • LaPar D.J.
      • Crosby I.K.
      • Rich J.B.
      • Quader M.A.
      • Speir A.M.
      • Kern J.A.
      • et al.
      Investigators for the Virginia Cardiac Surgery Quality Initiative
      Bilateral internal mammary artery use for coronary artery bypass grafting remains underutilized: a propensity-matched multi-institution analysis.
      The reported BITA use rates are 4% in the United States,
      • El-Bardissi A.W.
      • Aranki S.F.
      • Sheng S.
      • O'Brien S.M.
      • Greenberg C.C.
      • Gammie J.S.
      • et al.
      Trends in isolated coronary artery bypass grafting: an analysis of the Society of Thoracic Surgeons adult cardiac surgery database.
      12% in Europe,
      • Kappetein A.P.
      • Dawkins K.D.
      • Mohr F.W.
      • Morice F.C.
      • Mack M.J.
      • Russell M.E.
      • et al.
      Current percutaneous coronary intervention and coronary artery bypass grafting practices for three-vessel and left main coronary artery disease. Insights from the SYNTAX run-in phase.
      and 12.6% in Australia.
      • Yan B.P.
      • Clark D.J.
      • Buxton B.
      • Ajani A.E.
      • Smith J.A.
      • Duffy S.J.
      • et al.
      on behalf of the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS), the Melbourne Interventional Group (MIG)
      Clinical characteristics and early mortality of patients undergoing coronary artery bypass grafting compared to percutaneous coronary intervention: insights from the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS) and the Melbourne interventional group (MIG) registries.
      A Swedish study recently published that only 1% of the patients who underwent CABG between 1997 and 2008 received BITA grafts,
      • Dalén M.
      • Ivert T.
      • Holzmann M.J.
      • Sartipy U.
      Bilateral versus single internal mammary coronary artery bypass grafting in Sweden from 1997-2008.
      whereas in Canada, a survey found that only 40% of the surgeons used BITA in 5% to 25% of their CABG cases, whereas 37% used BITA in less than 5% of their cases.
      • Mastrobuoni S.
      • Gawad N.
      • Price J.
      • Chan V.
      • Ruel M.
      • Mesana T.J.
      • et al.
      Use of bilateral internal thoracic artery during coronary artery bypass surgery in Canada: the bilateral internal thoracic artery survey.
      A similar recent survey from India shows that 18% of surgeons do not use BITA at all, whereas 22% use BITA in 0% to 5% of their cases.
      • Sajja L.R.
      • Beri P.
      Bilateral internal thoracic artery grafting in India–time to raise the bar.
      Of the 43,823 patients identified from the Virginia Cardiac Surgery Quality Initiative data registry who underwent CABG, only 3% received BITA.
      • LaPar D.J.
      • Crosby I.K.
      • Rich J.B.
      • Quader M.A.
      • Speir A.M.
      • Kern J.A.
      • et al.
      Investigators for the Virginia Cardiac Surgery Quality Initiative
      Bilateral internal mammary artery use for coronary artery bypass grafting remains underutilized: a propensity-matched multi-institution analysis.
      In our institution, the total number of CABG cases declined from 649 in 2000 to 312 in 2011, with a corresponding decline in BITA numbers. However, between 2012 and 2015, the number of BITA cases increased 2-fold (13% to 27%) despite a continued decline in the number of CABG cases (Figure 1). The decline in the total number of CABG cases is a reflection of the general trend in coronary revascularization practice. This increasing trend (P = .005) in BITA use is a reflection of a deliberate change in our practice based on our published results in 2012, which showed that multiarterial grafting (>85% with BITA) conferred a significant survival benefit.
      • Locker C.
      • Schaff H.V.
      • Dearani J.A.
      • Joyce L.D.
      • Park S.J.
      • Burkhart H.M.
      • et al.
      Multiple arterial grafts improve late survival of patients undergoing coronary artery bypass graft surgery: analysis of 8622 patients with multivessel disease.
      In a matched cohort of patients, Grau and colleagues
      • Grau J.B.
      • Ferrari G.
      • Mak A.W.C.
      • Shaw R.E.
      • Brizzio M.E.
      • Mindich B.P.
      • et al.
      Propensity matched analysis of bilateral internal mammary artery versus single left internal mammary artery grafting at 17-year follow-up: validation of a contemporary surgical experience.
      found that use of BITA over LITA-only had a 10% survival advantage at 10 years and 18% at 15 years follow-up. A recent meta-analysis from 20 observational studies enrolling more than 70,000 patients reported that BITA grafting significantly reduced long-term mortality with a pooled HR of 0.80.
      • Takagi H.
      • Goto S.N.
      • Watanabe T.
      • Mizuno Y.
      • Kawai N.
      • Umemoto T.
      A meta-analysis of adjusted hazard ratios from 20 observational studies of bilateral versus single internal thoracic artery coronary artery bypass grafting.
      This correlates with our findings of reduced long-term mortality in the BITA group with an HR of 0.81 and 0.80 from case-matched and covariate-adjusted propensity score analyses. The survival benefit of BITA extends over longer follow-up and the curves appear to diverge more with time. Of note, in our previous report on a matched cohort of patients undergoing CABG between 1993 and 2009, the 5- and 10-year survivals of patients receiving multiarterial grafts versus those receiving LITA/SVG grafts were similar to patients treated with BITA versus LITA-alone in the current study, with the gap further increasing at 15 years of follow-up.
      • Locker C.
      • Schaff H.V.
      • Dearani J.A.
      • Joyce L.D.
      • Park S.J.
      • Burkhart H.M.
      • et al.
      Multiple arterial grafts improve late survival of patients undergoing coronary artery bypass graft surgery: analysis of 8622 patients with multivessel disease.
      The lack of survival benefit with BITA in the ART at 5 years may be related to asymptomatic vein-graft failure, improved medical therapy, possible differences in surgical techniques at different centers involved in this multicenter trial, and some methodological shortcomings.
      • Taggart D.P.
      • Altman D.G.
      • Gray A.M.
      • Lees B.
      • Gerry S.
      • Benedetto U.
      • et al.
      ART Investigators
      Randomized trial of bilateral versus single internal-thoracic-artery grafts.
      • Luthra S.
      Degrees of belief and the burden of proof: the ART Trial.
      Because the survival curves of patients operated on with vein grafts show a steeper decline after 7 years,
      • Fitzgibbon G.M.
      • Kafka H.P.
      • Leach A.J.
      • Keon W.J.
      • Hooper D.
      • Burton J.R.
      • et al.
      Coronary bypass graft fate and patient outcome: angiographic follow-up of 5,065 grafts related to survival and reoperation in 1,388 patients during 25 years.
      and the survival benefit of BITA tends to increase over time, the long-term results of the ART would better reflect on BITA use.
      In our study, the relative efficacy of BITA versus LITA-alone was not significantly different in specific “high-risk” subsets when compared with patients who were not high risk. Some recent studies, reporting BITA use rates of greater than 30%, have demonstrated a clear survival benefit of BITA in patients aged more than 70 years,
      • Kurlansky P.A.
      • Traad E.A.
      • Dorman M.J.
      • Galbut D.L.
      • Ebra G.
      Bilateral versus single internal mammary artery grafting in the elderly: long-term survival benefit.
      whereas other studies have advocated for an age cutoff at 60 years
      • Mohammadi S.
      • Dagenais F.
      • Doyle D.
      • Mathieu P.
      • Baillot R.
      • Charbonneau E.
      • et al.
      Age cut-off for the loss of benefit from bilateral internal thoracic artery grafting.
      or 69 years
      • Benedetto U.
      • Amrani M.
      • Raja S.G.
      Harefield Cardiac Outcomes Research Group
      Guidance for the use of bilateral internal thoracic arteries according to survival benefit across age groups.
      under the premise that BITA loses its survival benefit in elderly patients. We found no adverse effect of age 70 years or more on the survival benefit conferred by BITA over LITA-alone. Low use rates of BITA in some of these studies show a somewhat selective approach in patient selection, and thereby the differences in survival may not be as apparent.
      • Kurlansky P.A.
      • Traad E.A.
      • Dorman M.J.
      • Galbut D.L.
      • Ebra G.
      Bilateral versus single internal mammary artery grafting in the elderly: long-term survival benefit.
      Many surgeons are averse to using BITA as harvesting of an additional ITA increases the operative time. Taggart and colleagues
      • Taggart D.P.
      • Altman D.G.
      • Gray A.M.
      • Lee B.
      • Nugara F.
      • Yu L.
      • et al.
      Randomized trial to compare bilateral vs single internal mammary coronary artery bypass grafting: 1-year results of the arterial revascularisation trial (ART).
      found that there is approximately half an hour increase in operative time. This is similar to our practice wherein the median operative time increased by 44 minutes. Although the operative and crossclamp times increased with BITA, there was a reduction in CPB time, probably related to a lower number of proximal anastomoses performed in BITA cases with the composite-T sequential and in situ grafting techniques. We also did not find any increase in the length of hospital stay. BITA cases had a slightly shorter ICU stay. The increase in operative time should not be a deterrent in using BITA given the survival advantage observed in our study and others.
      Some studies have raised concerns about the need for prolonged postoperative ventilatory support
      • Knapik P.
      • Spyt T.J.
      • Richardson J.B.
      • McLellan I.
      Bilateral and unilateral use of internal thoracic artery for myocardial revascularization. Comparison of extubation outcome and duration of hospital stay.
      in BITA cases with CLD. However, later studies have shown that any concern for increased respiratory complications after BITA is unfounded.
      • Taggart D.P.
      Respiratory dysfunction after cardiac surgery: effects of avoiding cardiopulmonary bypass and the use of bilateral internal mammary arteries.
      Our study found no significant difference in the need for prolonged postoperative ventilatory support between BITA and LITA-alone groups, besides the absence of any adverse effect of CLD on the survival benefit of BITA.
      Diabetes and reduced EF are known risk factors for reduced survival.
      • Endo M.
      • Tomizawa Y.
      • Nishida H.
      Bilateral versus unilateral internal mammary revascularization in patients with diabetes.
      Endo and colleagues
      • Endo M.
      • Tomizawa Y.
      • Nishida H.
      Bilateral versus unilateral internal mammary revascularization in patients with diabetes.
      demonstrated that survival benefit may be limited in patients with diabetes and EF 40% or less because of an unavoidable high cardiac death rate. A recent study
      • Pevni D.
      • Medalion B.
      • Mohr R.
      • Ben-Gal Y.
      • Laub A.
      • Nevo A.
      • et al.
      Should bilateral internal thoracic artery grafting be used in patients with diabetes mellitus?.
      has shown an improved survival in diabetic patients with BITA grafts versus LITA-alone, with an HR of 0.73. Our study also found that reduced EF and diabetes have no adverse effect on the survival benefit conferred by BITA. Given the smaller size of coronary vessels and more severe coronary disease, diabetic patients are more likely to benefit from BITA grafting.
      The increased incidence of sternal infections in patients with diabetes, especially obese diabetic women
      • Matsa M.
      • Paz Y.
      • Gurevitch J.
      • Shapira I.
      • Kramer A.
      • Pevny D.
      • et al.
      Bilateral skeletonized internal thoracic artery grafts in patients with diabetes mellitus.
      and CLD,
      • Sofer D.
      • Gurevitch J.
      • Shapira I.
      • Paz Y.
      • Matsa M.
      • Kramer A.
      • et al.
      Sternal wound infections in patients after coronary artery bypass grafting using bilateral skeletonized internal mammary arteries.
      has somewhat prevented surgeons from adopting BITA grafting. The incidence of overall sternal-site infection in the BITA group in this study was approximately 2-fold higher (OR, 2.28) than matched patients in the LITA-alone group (3.1% vs 1.4%; P = .010). This higher risk among BITA cases persisted after adjustment for sex, BMI, diabetes, CLD, and year of CABG (OR, 2.39). Of note, there was no significant difference in the rates of deep-sternal infections between the 2 groups. The low number of deep-sternal infections in our study, although desirable, limited our statistical analysis. It has been shown that the use of skeletonized BITA reduces the incidence of sternal-site infections, particularly in patients with diabetes.
      • Deo S.V.
      • Shah I.K.
      • Dunlay S.M.
      Bilateral internal thoracic artery harvest and deep sternal wound infection in diabetic patients.
      Findings from the ART showed that using a skeletonized technique for harvesting BITA resulted in similar rates of sternal wound complications compared with harvesting a single ITA with the standard pedicled technique.
      • Benedetto U.
      • Altman D.G.
      • Gerry S.
      • Gray A.
      • Lees B.
      • Pawlaczy R.
      • et al.
      Pedicled and skeletonized single and bilateral internal thoracic artery grafts and the incidence of sternal wound complications: insights from the arterial revascularization trial.
      In our current practice, we aim to harvest the ITAs with an almost exclusive use of skeletonized technique.

       Study Limitations

      The study is limited by inherent disadvantages of a retrospective single-center design. In addition, even after controlling for all known variables in the propensity score model, we cannot exclude treatment selection bias based on nonmeasurable variables (eg, patient fitness, surgeon bias). Although most of our variables had little to no missing data, operative time and ICU stay were added to the STS data collection in 2002. We recognize this restricts those particular inferences to the years 2002 to 2015, although the missing subset is small (18%) and would likely not have negated these findings. Use of multiple grafting strategies and the likely unaccounted variability of target lesions might have biased our results. Nevertheless, our results show that despite the multiplicity of revascularization strategy, a second ITA offers a distinct survival advantage and remains the superior bypass graft. Finally, we have used an end point of all-cause mortality as the basis of treatment comparison because of the lack of information in our database regarding long-term follow-up of major adverse cardiac events or cause of death. This may limit the applicability of our findings even though all-cause mortality has been shown to be a better end point compared with cause-specific mortality in coronary interventions.
      • Holmes Jr., D.R.
      • Kip K.E.
      • Kelsey S.F.
      • Detre K.M.
      • Rosen A.D.
      Cause of death analysis in the NHLBI PTCA Registry: results and considerations for evaluating long-term survival after coronary interventions.

      Conclusions

      The study shows an increasing utilization rate of BITA in Mayo Clinic, Rochester. Expansion of the patient population selected for BITA revascularization seems to be supported by the favorable outcomes achieved even within the “high-risk” patient categories. Although the optimal use of BITA remains to be clearly defined, our data support more extensive use of BITA in patients with multivessel CAD.

       Conflict of Interest Statement

      Authors have nothing to disclose with regard to commercial support.

      Supplementary Data

      References

        • Mozaffarian D.
        • Benjamin E.J.
        • Go A.S.
        • Arnett D.K.
        • Blaha M.J.
        • Cushman M.
        • et al.
        Heart disease and stroke statistics-2016 update: a report from the American Heart Association.
        Circulation. 2016; 133: 38-60
        • Taggart D.P.
        CABG is still the best treatment for multivessel and left main disease, but patients need to know.
        Ann Thorac Surg. 2006; 82: 1966-1975
        • Loop F.D.
        • Lytle B.W.
        • Cosgrove D.M.
        • Stewart R.W.
        • Goormastic M.
        • Williams G.W.
        • 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
        • Cameron A.
        • Davis K.B.
        • Green G.
        • Schaff H.V.
        Coronary bypass surgery with internal-thoracic-artery grafts—effects on survival over a 15-year period.
        N Engl J Med. 1996; 334: 216-219
        • Fitzgibbon G.M.
        • Kafka H.P.
        • Leach A.J.
        • Keon W.J.
        • Hooper D.
        • Burton J.R.
        • et al.
        Coronary bypass graft fate and patient outcome: angiographic follow-up of 5,065 grafts related to survival and reoperation in 1,388 patients during 25 years.
        J Am Coll Cardiol. 1996; 28: 616-626
        • Taggart D.P.
        • D'Amico R.
        • Altman D.G.
        Effect of arterial revascularization on survival: a systematic review of studies comparing bilateral and single internal mammary arteries.
        Lancet. 2001; 358: 870-875
        • Takagi H.
        • Goto S.N.
        • Watanabe T.
        • Mizuno Y.
        • Kawai N.
        • Umemoto T.
        A meta-analysis of adjusted hazard ratios from 20 observational studies of bilateral versus single internal thoracic artery coronary artery bypass grafting.
        J Thorac Cardiovasc Surg. 2014; 148: 1282-1290
        • El-Bardissi A.W.
        • Aranki S.F.
        • Sheng S.
        • O'Brien S.M.
        • Greenberg C.C.
        • Gammie J.S.
        • et al.
        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: 273-281
        • Kappetein A.P.
        • Dawkins K.D.
        • Mohr F.W.
        • Morice F.C.
        • Mack M.J.
        • Russell M.E.
        • et al.
        Current percutaneous coronary intervention and coronary artery bypass grafting practices for three-vessel and left main coronary artery disease. Insights from the SYNTAX run-in phase.
        Eur J Cardiothorac Surg. 2006; 29: 486-491
        • Yan B.P.
        • Clark D.J.
        • Buxton B.
        • Ajani A.E.
        • Smith J.A.
        • Duffy S.J.
        • et al.
        • on behalf of the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS), the Melbourne Interventional Group (MIG)
        Clinical characteristics and early mortality of patients undergoing coronary artery bypass grafting compared to percutaneous coronary intervention: insights from the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS) and the Melbourne interventional group (MIG) registries.
        Heart Lung Circ. 2009; 18: 184-190
        • Kouchoukos N.T.
        • Wareing T.H.
        • Murphy S.F.
        • Pelate C.
        • Marshall Jr., W.G.
        Risks of bilateral internal mammary artery bypass grafting.
        Ann Thorac Surg. 1990; 49: 210-217
        • Vallely M.P.
        • Edelman J.J.B.
        • Wilson M.K.
        Bilateral internal mammary arteries: evidence and technical consideration.
        Ann Cardiothorac Surg. 2013; 2: 570-577
        • Mohammadi S.
        • Dagenais F.
        • Doyle D.
        • Mathieu P.
        • Baillot R.
        • Charbonneau E.
        • et al.
        Age cut-off for the loss of benefit from bilateral internal thoracic artery grafting.
        Eur J Cardiothorac Surg. 2008; 33: 977-982
        • Taggart D.P.
        Respiratory dysfunction after cardiac surgery: effects of avoiding cardiopulmonary bypass and the use of bilateral internal mammary arteries.
        Eur J Cardiothorac Surg. 2000; 18: 31-37
        • Matsa M.
        • Paz Y.
        • Gurevitch J.
        • Shapira I.
        • Kramer A.
        • Pevny D.
        • et al.
        Bilateral skeletonized internal thoracic artery grafts in patients with diabetes mellitus.
        J Thorac Cardiovasc Surg. 2001; 121: 668-674
        • Sofer D.
        • Gurevitch J.
        • Shapira I.
        • Paz Y.
        • Matsa M.
        • Kramer A.
        • et al.
        Sternal wound infections in patients after coronary artery bypass grafting using bilateral skeletonized internal mammary arteries.
        Ann Surg. 1999; 229: 585-590
        • Kurlansky P.A.
        • Traad E.A.
        • Dorman M.J.
        • Galbut D.L.
        • Zucker M.
        • Ebra G.
        • et al.
        Efficacy of single versus bilateral internal mammary artery grafting in women: a long-term study.
        Ann Thorac Surg. 2001; 71: 1949-1958
        • Taggart D.P.
        • Altman D.G.
        • Gray A.M.
        • Lees B.
        • Gerry S.
        • Benedetto U.
        • et al.
        • ART Investigators
        Randomized trial of bilateral versus single internal-thoracic-artery grafts.
        N Engl J Med. 2016; 375: 2540-2549
      1. Society of Thoracic Surgeons. STS adult cardiac surgery database data specifications version 2.81. March 28, 2014. Available at: https://www.sts.org/sites/default/files/documents/ACSD_DataSpecificationsV2_81.pdf. Accessed March 9, 2018.

        • Knapik P.
        • Spyt T.J.
        • Richardson J.B.
        • McLellan I.
        Bilateral and unilateral use of internal thoracic artery for myocardial revascularization. Comparison of extubation outcome and duration of hospital stay.
        Chest. 1996; 109: 1231-1233
        • Taggart D.P.
        • Altman D.G.
        • Gray A.M.
        • Lee B.
        • Nugara F.
        • Yu L.
        • et al.
        Randomized trial to compare bilateral vs single internal mammary coronary artery bypass grafting: 1-year results of the arterial revascularisation trial (ART).
        Eur Heart J. 2010; 31: 2470-2481
        • Benedetto U.
        • Amrani M.
        • Raja S.G.
        • Harefield Cardiac Outcomes Research Group
        Guidance for the use of bilateral internal thoracic arteries according to survival benefit across age groups.
        J Thorac Cardiovasc Surg. 2014; 148: 2706-2711
        • LaPar D.J.
        • Crosby I.K.
        • Rich J.B.
        • Quader M.A.
        • Speir A.M.
        • Kern J.A.
        • et al.
        • Investigators for the Virginia Cardiac Surgery Quality Initiative
        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
        • Dalén M.
        • Ivert T.
        • Holzmann M.J.
        • Sartipy U.
        Bilateral versus single internal mammary coronary artery bypass grafting in Sweden from 1997-2008.
        PLoS One. 2014; 9: e86929
        • Mastrobuoni S.
        • Gawad N.
        • Price J.
        • Chan V.
        • Ruel M.
        • Mesana T.J.
        • et al.
        Use of bilateral internal thoracic artery during coronary artery bypass surgery in Canada: the bilateral internal thoracic artery survey.
        J Thorac Cardiovasc Surg. 2012; 144: 874-879
        • Sajja L.R.
        • Beri P.
        Bilateral internal thoracic artery grafting in India–time to raise the bar.
        Indian Heart J. 2017; 69: 640-645
        • Locker C.
        • Schaff H.V.
        • Dearani J.A.
        • Joyce L.D.
        • Park S.J.
        • Burkhart H.M.
        • et al.
        Multiple arterial grafts improve late survival of patients undergoing coronary artery bypass graft surgery: analysis of 8622 patients with multivessel disease.
        Circulation. 2012; 126: 1023-1030
        • Grau J.B.
        • Ferrari G.
        • Mak A.W.C.
        • Shaw R.E.
        • Brizzio M.E.
        • Mindich B.P.
        • et al.
        Propensity matched analysis of bilateral internal mammary artery versus single left internal mammary artery grafting at 17-year follow-up: validation of a contemporary surgical experience.
        Eur J Cardiothorac Surg. 2012; 41: 770-776
        • Luthra S.
        Degrees of belief and the burden of proof: the ART Trial.
        Ann Thorac Surg. 2017; 104: 1441-1444
        • Kurlansky P.A.
        • Traad E.A.
        • Dorman M.J.
        • Galbut D.L.
        • Ebra G.
        Bilateral versus single internal mammary artery grafting in the elderly: long-term survival benefit.
        Ann Thorac Surg. 2015; 100: 1374-1382
        • Endo M.
        • Tomizawa Y.
        • Nishida H.
        Bilateral versus unilateral internal mammary revascularization in patients with diabetes.
        Circulation. 2003; 108: 1343-1349
        • Pevni D.
        • Medalion B.
        • Mohr R.
        • Ben-Gal Y.
        • Laub A.
        • Nevo A.
        • et al.
        Should bilateral internal thoracic artery grafting be used in patients with diabetes mellitus?.
        Ann Thorac Surg. 2017; 103: 551-558
        • Deo S.V.
        • Shah I.K.
        • Dunlay S.M.
        Bilateral internal thoracic artery harvest and deep sternal wound infection in diabetic patients.
        Ann Thorac Surg. 2013; 95: 862-869
        • Benedetto U.
        • Altman D.G.
        • Gerry S.
        • Gray A.
        • Lees B.
        • Pawlaczy R.
        • et al.
        Pedicled and skeletonized single and bilateral internal thoracic artery grafts and the incidence of sternal wound complications: insights from the arterial revascularization trial.
        J Thorac Cardiovasc Surg. 2016; 152: 270-276
        • Holmes Jr., D.R.
        • Kip K.E.
        • Kelsey S.F.
        • Detre K.M.
        • Rosen A.D.
        Cause of death analysis in the NHLBI PTCA Registry: results and considerations for evaluating long-term survival after coronary interventions.
        J Am Coll Cardiol. 1997; 30: 881-887