Utilization of Vein Grafts in Coronary Artery Bypass Grafting: Reasons and Outcomes in a Bilateral Mammary Artery First Center

Objectives . Despite guideline recommendations for use of bilateral internal mammary artery (BIMA) in coronary artery bypass grafting (CABG), a large proportion of patients still receive saphenous vein grafts (SVG). We herein aimed to identify reasons for SVG use at a center with a BIMA utilization rate between 60 and 70% and compare outcomes of patients undergoing CABG with either BIMA or left internal mammary artery (LIMA) plus SVG. Methods . Between 2013 and 2022, 4145 consecutive patients underwent isolated CABG at our center. Of those, 2067 patients received BIMA (group 1) and 1206 patients received LIMA/SVG (group 2). A propensity score-matched analysis was performed to adjust for baseline diferences. Results . Group 2 presented with higher age, more female patients, and more patients with acute coronary syndrome including NSTEMI/STEMI with more urgent/emergency CABG. In unadjusted analysis group 2 presented adverse 30-day outcomes compared to group 1 with a higher mortality (18/2067, 0.9% vs. 34/1206, 2.8%; p < 0 . 001), higher rate of re-revascularization (52/2067, 2.5% vs. 50/1206, 4.1%; p < 0 . 001), more stroke (20/2067; 1.0% vs. 33/1206, 2.7%; p < 0 . 001), and more postoperative renal failure (17/2067, 0.8% vs. 27/1206, 2.2%; p � 0 . 001). After adjustment for baseline characteristics, 30-day outcomes were comparable. Conclusions . After adjustment for baseline characteristics no diferences in outcomes were found between groups suggesting a safe applicability of BIMA even in patients with acute coronary syndrome undergoing urgent/emergency CABG. Reasons for SVG use were higher age, female gender, and acute coronary syndrome with urgent/emergency CABG. Outcomes of both groups were excellent with low rates of primary endpoints.


Introduction
According to recent guidelines for myocardial revascularization coronary artery bypass grafting (CABG) is recommended for treatment of complex coronary artery disease (CAD) in patients with intermediate to high SYN-TAX (Synergy between Percutaneous Coronary Intervention with TAXUS and Cardiac Surgery) score and with type 2 diabetes [1,2].Although results of randomized controlled trials and retrospective studies are partly contradictory, there is a wide consensus for utilization of the left (LIMA) and right internal mammary artery (RIMA), usually referred to as bilateral mammary artery (BIMA), in CABG since long-term patency rates were shown to be superior compared to saphenous vein grafts (SVG) or the radial artery [3,4].Especially patients with diabetes mellitus (DM) and reduced left ventricular function beneft from utilization of BIMA during CABG in terms of long-term survival and freedom from re-revascularization [5,6].However, despite clear recommendations for BIMA use in CABG, when appropriate and anatomically and clinically suitable, there is still a large proportion of patients receiving LIMA plus SVG in the real world as refected by a BIMA use of 11.3% in North America as documented in the Society of Toracic Surgeons Adult Cardiac Surgery registry and around 30% in Europe as documented in a large multicenter registry [7,8].Reasons for these low rates of BIMA utilization are likely multifactorial and difer according to investigated countries but main considerations may include longer procedure times with BIMA use in CABG, higher risk of wound healing disorders or deep sternal wound infections (DSWI), and a technically more demanding procedure [9,10].In our center, a BIMA frst strategy is implemented, leading to an average BIMA rate in CABG of 63% over the last ten years with a clear trend towards a further increasing utilization rate (78.1% in 2021).Despite this strategy, SVG is still applied in selected patients and particular clinical circumstances at our center.We herein aimed to retrospectively identify reasons for SVG use at our center and compare short-term outcomes of patients undergoing CABG with either BIMA or LIMA plus SVG over the last nine years.A patient inclusion fowchart is shown in Figure 1.

Patients and Methods
Primary endpoint for this study is all-cause mortality.Secondary outcomes consist of adverse events during 30 days after the index procedure including postoperative myocardial infarction, major stroke, acute renal failure (kidney disease improving global organization: KDIGO > 3), re-revascularization (percutaneous coronary intervention or surgical revision due to (combination of) signifcant changes in the ECG, rising troponin levels, wall motion abnormalities, hemodynamic instability), resternotomy for bleeding, number of transfused red blood cell (RBC) units, length of intensive care unit (ICU) and hospital stay, and postoperative wound infection.
Complete revascularization was defned as revascularization of all coronary segments with a stenosis of ≥50% supplying viable myocardium [11].DSWI was defned as postoperative infection involving the sternum and mediastinal space.

Statistical Analyses.
Continuous variables are reported as mean ± standard deviation or median with interquartile range.Categorical variables are presented as proportions.Baseline diferences between patients undergoing CABG using BIMA and LIMA plus SVG were detected using the Student's t-test for normally distributed continuous variables, the Mann-Whitney test for non-normally distributed continuous variables and the Fisher's exact test for categorical variables.Te efect of the second bypass graft was analyzed in a univariable model and subsequently with propensity score matching analysis.Te propensity score was estimated using a nonparsimonious logistic regression model, including age, sex, myocardial infarct, number of diseased coronary arteries, extracardiac artheropathy, chronic obstructive lung disease, previous transient ischemic attack or stroke, left ventricular ejection fraction (LVEF) < 35%, insulindependent diabetes mellitus, emergency procedure and preoperative GFR <60 ml/min.One-to-one propensity score matching was performed using the nearest-neighbor method and a caliper width of 0.2 of the standard deviation of the logit of the propensity score.Balance between propensity scorematched groups was assessed by evaluation of standardized diferences which was considered nonsignifcant <0.10.Te Student's t-test for continuous variables and the McNemar test for dichotomous variables were used to evaluate diferences in baseline characteristics and adverse events of propensity scorematched pairs.Post hoc power analysis presented an adequate sample size for detection of diferences in the primary endpoint.
All tests were two tailed and p < 0.05 was considered statistically signifcant.All statistical analyses were performed using the statistical software SPSS (IBM SPSS Statistics for Windows, Version 27.0,IBM Corp., Armonk, NY, USA).
After propensity score matching, 646 patients in each patient group remained for further analysis.Signifcant diferences between matched groups persisted regarding age, BMI > 30 kg/m 2 , and EuroSCORE II.
Detailed patient demographics of unmatched and matched groups are summarized in Table 1.

Periprocedural Data.
In the unmatched patient cohorts, signifcant diferences in periprocedural data were found regarding procedure time (255 [220-300]  Detailed 30-day outcome parameters are summarized in Table 3. Te distribution of key 30-day outcome parameters for matched patient cohorts is shown in Figure 2.

Discussion
Main fndings of the herein conducted study are: (I) difering baseline characteristics of patients provided with BIMA or LIMA and SVG suggesting that main reasons for utilization of vein grafts at our center are higher age, female gender, acute coronary syndrome with an urgent or emergency CABG procedure and prevalence of comorbidities with an overall higher surgical risk profle, (II) outcomes of both groups were excellent with low rates of mortality, stroke, renal failure and DSWI, and (III) when comparing groups after adjustment for baseline characteristics no signifcant diferences in 30-day outcomes were found suggesting a safe applicability of BIMA even in patients with acute coronary syndrome undergoing urgent or emergency CABG.
Commonly, SVG use in CABG is considered to present certain drawbacks such as high failure rates in the mid and long term, rapid progression of intermediate stenoses, limitations in applicability of percutaneous intervention in SVG stenoses, and an even higher rate of graft failure when providing multiple vessel targets with SVG compared to single SVG bypass [12,13].Although these drawbacks are described in several studies and guidelines for myocardial revascularization clearly recommend use of BIMA or the radial artery instead of SVG, vein grafts are still the most frequently used bypass grafts worldwide [1,14].Described reasons for application of SVG in CABG consist of female gender, obesity, older age, and presence of diabetes [15].4

While association of SVG use and operation time with
Journal of Cardiac Surgery regards to economization of clinical resources is not well investigated, reluctance in application of BIMA in diabetic patients may be related to describe higher incidences of DSWI with BIMA [9].Furthermore, SVG use in predominantly female patients may be connected to anatomical considerations regarding smaller target vessel diameters [16].Te herein presented data from a BIMA frst center largely confrms previously described reasons for SVG use in CABG, although SVG is overall signifcantly less utilized over a period of nine years compared to describe utilization rates.In addition, our investigated data suggest that patients with acute coronary syndrome undergoing urgent or emergency CABG are more likely to receive SVG for intended fast and sufcient reestablishment of coronary blood fow by sparing preparation of a second arterial graft.However, although this strategy is plausible, there is no data comparing BIMA and LIMA plus SVG use in the setting of acute coronary syndrome.Te herein presented outcome data of baseline-adjusted groups indicate that utilization of BIMA even in patients undergoing emergent CABG in acute coronary syndrome may be safe and feasible.On the other hand, utilization of LIMA and SVG may be advantageous in the short term in older and frailer patients due to the reduced operative trauma and less DSWI as herein shown.A patient-specifc approach should be established to weigh possible long-term benefts of BIMA utilization against short-term benefts of SVG.Although clear drawbacks in outcomes of the LIMA and SVG group were seen in unadjusted analysis, it needs to be emphasized that outcomes of both groups were excellent with low rates of mortality, stroke, rerevascularization, and renal failure which are in line or even lower when compared to national and international registries [17,18].Terefore, strategy for graft choice should be based not only on the clinical setting but also on surgeons`expertise, experience in utilization of BIMA and graft quality.Whenever SVG is used key factors for prolonged graft patency should be taken into consideration including intraprocedural measures such as applying SVG to target vessels with a larger diameter, implementing a no touch harvesting technique [19], and avoid bypassing of hemodynamically insignifcant stenosis [20] as well as postoperative management including administration of dual antiplatelet therapy [21].
In this study, a trend towards a higher incidence of DSWI in patients receiving BIMA after adjustment for baseline diferences was seen.However, DSWI after CABG is a multifactorial process including impact of the patient's nutritional status, technique of IMA harvesting (skeletonized vs. pedicled), and/or presence of metabolic syndrome and obesity [22][23][24].Since not all risk factors for DSWI were investigated in the context of this study and CABG using BIMA still results in decreased hospital mortality, cerebrovascular events, and re-revascularization rates, especially in patients with reduced left ventricular function and DM [25], BIMA should be considered the frst line grafting strategy in the majority of patients, and complete revascularization should be the goal in every patient.
4.1.Limitations.Limitations are inherent in the retrospective, single-center study design with limited patient numbers: patients were not randomized to a specifc treatment; therefore, patient preselection with hidden confounders may apply, especially in terms of selection bias regarding surgeons' choice of LIMA vs. BIMA and on-pump CABG vs. OPCAB.Furthermore, no data were available on how many patients were converted from OPCAB to on-pump CABG or beating heart procedures.Also, a propensity score matching approach was applied to adjust for baseline diferences; diferences regarding BMI and EuroSCORE II persisted after matching, suggesting imbalances between patients receiving BIMA or SVG which may hamper comparability.However, major diferences in outcome parameters before and after matching suggest at least a mostly sufcient abolishment of major baseline imbalances.Furthermore, operative strategies were diferent between groups, suggesting an imbalance of operating surgeons between groups, and NSTEMI/STEMI patients were merged to one subgroup, which may hamper interpretation of results.Journal of Cardiac Surgery

Conclusions
After adjustment for baseline characteristics no diferences in outcomes were found between groups suggesting a safe applicability of BIMA even in patients with acute coronary syndrome undergoing urgent/emergency CABG.Difering baseline characteristics of patients provided with BIMA or LIMA and SVG suggesting that main reasons for utilization of vein grafts in our center are higher age, female gender, acute coronary syndrome with an urgent or emergency CABG procedure, and prevalence of comorbidities with an overall higher surgical risk profle.Utilization of LIMA and SVG may be advantageous in older and frailer patients due to the reduced operative trauma and less DSWI as herein shown.When taking the higher rate of DSWI in patients provided with BIMA into consideration, a patient-specifc approach should be established to weigh possible long-term benefts of BIMA utilization against short-term benefts of SVG.
2.1.Ethical Statement.Data acquisition was performed anonymized and retrospectively.Terefore, in accordance with German law, no ethical approval is needed and informed patient consent was waived.2.2.Patients and Defnitions.Between 01/2013 and 12/2022, 4145 consecutive patients underwent isolated CABG at our center.Exclusion criteria were utilization of no internal mammary artery, utilization of BIMA plus SVG, single anastomosis and application of the radial artery.Overall, 2067 patients with BIMA use and 1206 patients with LIMA plus SVG use remained for analysis.Choice of surgical strategy (e.g., on-pump CABG vs. OPCAB) was left to operators' discretion.

Table 1 :
Baseline data of unmatched and propensity score-matched patients.

Table 2 :
Periprocedural data of unmatched and propensity score-matched patients.