Background
The use of arterial grafts for coronary artery bypass grafting surgery (CABG), particularly bilateral internal mammary arteries (BIMA), is recommended due to the superior patency of these grafts compared with saphenous vein grafts (SV grafts) [
1]. In real-world practice, however, the utilization of total arterial revascularization (TAR) lags behind these recommendations [
2‐
5]. Reasons for reluctance to conduct total arterial CABG even in stable patients include the increased technical demand, the increased operation time, and fear of bleeding complications and impaired wound healing [
6‐
8]. In patients undergoing CABG for acute myocardial infarction (AMI), large-scale data on TAR rates is limited, and rates ranging from 2 to 58% have been described [
9,
10]. In the unstable situation of AMI, the above-mentioned arguments against total arterial CABG might play an even more important role for decision-making, as patients with AMI undergoing urgent or emergent surgery would be expected to benefit from short operation times and rapid revascularization afforded by use of venous grafting. Furthermore, AMI patients are frequently administered dual antiplatelet therapy (DAPT) preoperatively, resulting in increased risk of bleeding complications [
11‐
13].
It is currently unclear whether these concerns about the use of TAR in patients with AMI are valid in the current era of surgical myocardial revascularization. Furthermore, the possible effect of total arterial CABG on long-term outcome in AMI patients has never been explicitly investigated.
Methods
Study population
We conducted a retrospective, single-centre study comparing patients undergoing total arterial CABG (total arterial revascularization group [TAR group]) or CABG with a combination of one internal mammary artery (IMA) and saphenous vein grafts (saphenous vein graft group [SV group]). Adult patients with a diagnosis of AMI (non-ST-segment elevation myocardial infarction [NSTEMI] or ST-segment-elevation myocardial infarction [STEMI]) within a period of 5 days or less before CABG without concomitant procedures (e.g. valve surgery) between 01/2008 and 12/2014 were included in the analysis. Patients with low cardiac output syndrome (LCOS) or cardiogenic shock at the time of surgery were excluded. The local ethics committee approved the study.
Data collection, follow-up, definitions
Patients were identified according to the inclusion criteria from institutional patient records, and their baseline characteristics and perioperative data from the patient records and from data transferred to the nationwide quality assurance system (BQS Institute for Quality and Patient Safety, Hamburg, Germany) were analysed. Long-term follow-up was conducted via telephone interviews with the patients or their family physicians.
AMI was defined according to the Third Universal Definition of AMI [
14]. The time of AMI was defined as the time of symptom onset. ‘Complete revascularization’ was defined using the concept of anatomical complete numeric revascularization’ (bypassing of all vessels ≥1 mm with hemodynamically relevant stenosis, as assessed by coronary angiography) [
15]. We quantified the surgeon’s experience according to the years in practice since board certification as cardiac surgeon.
Endpoints
We compared intraoperative parameters (duration of surgery, completeness of revascularization), perioperative need for invasive ventilation, perioperative transfusion requirements and bleeding complications, acute kidney injury as defined by KDIGO (Kidney disease: improving global outcomes) [
16], sternal wound impairment requiring surgical therapy, postoperative duration of intensive care unit stay and hospitalization, as well as short- and mid-term survival between the groups.
Management strategy
Patients who underwent cardiac catheterization for AMI are referred to our unit immediately after completion of the angiographic diagnosis and the heart team-based decision for CABG. The timing of surgery is determined by the surgeon on duty. CABG with the goal of complete revascularization is routinely performed on-pump with cardioplegic arrest using cold-blood cardioplegia (Buckberg) [
17]. Acetylsalicylic acid is started 6 h postoperatively and continued lifelong at 100 mg/day. P2Y
12 inhibitors are started on the first postoperative day and are continued for 12 months.
Statistics
An inferential statistical analysis was performed using SPSS Version 24 (IBM, Armonk, NY, USA), GraphPad Prism version 6 software (GraphPad Software, Inc., La Jolla, CA, USA), and R version 3.1.2. Patient characteristics and outcomes were compared using Fisher’s exact test, Student’s t-test, or Wilcoxon-Mann-Whitney test, as appropriate. Continuous variables are presented as mean ± standard deviation (SD) unless stated otherwise.
In order to correct for potential confounding baseline parameters between the TAR group and the SV group, we carried out propensity score matching of the groups. Covariates included in the matching were age, gender, body-mass index, extent of coronary artery disease, preoperative left ventricular ejection fraction, diabetes mellitus (absence thereof, presence without insulin treatment, presence with insulin treatment), and EuroSCORE II. Nearest-neighbour matching in a 1:2 (TAR group vs. SV group) fashion was then performed. The maximum caliper between matched participants was set at 0.2. Long-term survival functions were determined using Kaplan-Meier estimation and compared using the log-rank test.
Discussion
The main finding of this analysis is that CABG using TAR is feasible in patients with AMI as it provides revascularization quality and patient safety like that of CABG using a combination of IMA and SV without increasing the time required for revascularization. Perioperative outcomes did not differ significantly between the groups. Bleeding complications and transfusion requirements were not higher after TAR than after revascularization using IMA/SV; in contrast, the proportion of patients who did not receive any red blood cell transfusion was higher in the TAR group. Postoperative atrial fibrillation was less frequent in the TAR group, possibly due to reduced red blood cell transfusion as demonstrated by previous studies [
18,
19]. Nevertheless, if transfusions were necessary, the amount of transfused erythrocyte units was rather high. This might be explained by the high rate of patients with DAPT at the time of surgery [
11].
The mean time of surgeon experience was slightly higher in the TAR group, probably reflecting that more experienced surgeons tend to perform this more challenging technique in urgent or emergent clinical settings. Moreover, the increased surgeon experience in the TAR group might result in better surgical results, although recent data did not confirm this assumption for CABG procedures [
20‐
22]. Surprisingly, analysis of procedural duration revealed that the total duration of the surgical procedures involving total arterial CABG and CABG using vein grafts were similar. Our data show that the surgeon’s experience has a significant influence on the duration of the procedure but that the amount is of questionable relevance. The longer phase of graft preparation in the TAR group was balanced by a shorter post-CPB phase in the TAR group. The reduction in reperfusion time and post-CPB time might be partly explained by the greater experience of the surgeons involved, leading to more efficient management at the end of the operation; however, the shorter time could additionally be explained by more rapid bypass graft function of arterial grafts compared with vein grafts, possibly resulting in quicker hemodynamic stabilization. Data on flow properties of arterial bypass grafts compared those of with vein grafts in the immediate intraoperative phase are limited: Spence et al. showed in a canine model that mammary artery graft flow is not impaired by competitive flow from the native vessel [
23]. As competitive flow from either the native vessel or collaterals is frequently observed in the early and late postoperative phase, resilience of the grafts may influence their immediate and long-term function [
24]. Concerning the immediate function, Weber et al. described improved intraoperative pulsatility indices and a tendency for reduced perioperative myocardial infarctions when using IMA grafts compared with vein grafts [
25]. We cannot substantiate our assumption of improved immediate bypass graft function, as flow measurements were not routinely carried out at our institution. Furthermore, although the postoperative increase in serum levels of cardiac biomarkers was somewhat less in the TAR group than in the SV group (possibly reflecting reduced cardiac injury resulting from improved bypass function), this difference was not statistically significant.
We were also surprised to observe that the most technically challenging phase of the procedure, the completion of the coronary anastomoses during cardioplegic arrest, required the same amount of time in the two groups, which is not in keeping with the reluctance to perform TAR due to more difficult and prolonged completion of coronary anastomoses. In fact, the present data should encourage surgeons to commit themselves early to TAR concepts, as these are feasible without loss of time in experienced hands.
Data from the postoperative follow-up period of up to 7 years did not show significant differences in survival between the groups; however, there was a tendency for improved survival in the TAR group from 4 years onwards. The rate of reported symptom-driven repeat coronary catheterizations were non-significantly lower in the TAR group. Unfortunately, there is no information available about the results of these coronary catheterizations and interventions performed. Redo-CABG occurred similarly in both groups. Previous studies have demonstrated that differences in graft patency between SV and IMA grafts become evident only after 4–8 years [
26,
27,
28]. A survival benefit after TAR in the mid- or long-term has been shown in pooled analyses [
29,
30]. Our observation is in accordance with the recently published work by Taggart et al. showing no significant survival benefit after bilateral IMA versus single IMA grafting after 5 years [
31]. A longer-term follow-up of the patients will be required to confirm these observations.
Several limitations of this study should be mentioned. First, patients with LCOS prior to surgery were excluded from this analysis, as CABG in these patients frequently does not follow the standardized sequence of operative steps. These patients are often placed on CPB prior to graft harvesting, and BIMA preparation is all but ruled out in these emergency situations, and hence we did not consider these exceptional, very individual situations to be suitable for a generalizable analysis. Therefore, the results of this study cannot be applied to patients who present with LCOS before CABG. Second, CPB with cardioplegic arrest was used in all procedures. Alternative approaches include off-pump CABG or on-pump CABG with beating heart [
32‐
34], which might reduce injury and inflammation associated with CPB and cardioplegic arrest. CPB with cardioplegic arrest, however, provides hemodynamic stability during the procedure with optimized exposure for accurate anastomosing. To date there are no data available showing superiority of one strategy over the other.
Although the logistical and technical aspects (procedural times, completeness of revascularization) of our study are not likely to be biased by the study design, outcome data of this retrospective, propensity-matched analysis should be considered with caution. Unknown confounders might reduce comparability of the groups and bias outcome data.
Acknowledgements
T The authors thank Irina Oswald for excellent assistance during patient follow-up and Dr. Elizabeth Martinson for language editing of the manuscript.