INTRODUCTION

Coronary artery bypass grafting (CABG) has a significant impact on survival of patients with coronary artery disease. The use of multiple arterial grafts (MAGs) plays an important role, due to its advantage related to long-term graft patency. Thus, guidelines recommend the use of a second arterial graft, mainly for the second best coronary artery^[1-4]. However, this choice is not yet recommended in the acute phase of myocardial infarction (AMI) mainly due to the lack of evidence.

Database studies have already shown that the use of bilateral internal thoracic artery (BITA) is superior to single internal thoracic artery graft in long-term follow-up, even when revascularization is incomplete^[1-4]. However, the time taken to harvest arterial grafts and the technical difficulty, with a consequent increase in surgical time, make most surgeons opt for the use of venous grafts in the AMI^[3,5,6]. In addition, glycemic decompensation, presence of kidney failure, and the use of antiplatelet agents, common in this phase, have an influence on decision-making^[6].

Believing that by using more venous grafts we are reducing surgical time may be a false impression since recent studies showed a decrease in surgery time and wound infection rates besides an excellent long-term patency when a radial graft is associated with the internal thoracic artery^[7]. Although the graft choice in CABG should always be individualized, we still have no evidence whether this can be modified by a situation of surgical urgency. In this scenario, groups with high-volume CABG surgeons have higher rates of use of MAGs with better results than groups with low-volume CABG surgeons^[8]. Thus, perhaps we are depriving patients of these benefits due to the lack of evidence in the AMI.

The objective of this study was to compare the mid-term clinical results of patients operated on in the AMI with MAGs vs. single arterial graft (SAG) based on data registered in the Registro Paulista de Cirurgia Cardiovascular II (REPLICCAR II).

METHODS

This is a retrospective cohort analysis of the REPLICCAR II database, a prospective, observational and multicenter registry with data from patients undergoing CABG consecutively in five hospitals in the state of São Paulo (Brazil) between July 2017 and June 2019 (n=4,053). REPLICCAR II followed the same variables and definitions of the Society of Thoracic Surgeons Adult Cardiac Surgery Database (STS ACSD) version 2.9 data collection system, through the partnership with the Harvard School of Public Health. All perioperative collection variables were performed online on a dedicated platform built on REDCap. Mid-term follow-up was carried out in a structured manner using a form filled out by telephone between October and December 2022. The AMI was considered 1-7 days of the diagnosis before the surgery, which followed the Fourth Universal Definition of Myocardial Infarction (2018)^[9].

This study included patients with AMI who underwent primary isolated CABG, while the exclusion criteria included elective procedures, combined surgeries, and reoperations. Definitions of the clinical status of the patient to guide revascularization follow the guidelines of the American societies of coronary revascularization of 2021^[10]. Thirty-six patients were operated on in the AMI using MAGs, one patient who died in the operating room was excluded because the purpose of this analysis was to evaluate the mid-term clinical follow-up of patients. Therefore, the final sample resulted in 35 patients operated using MAGs. Using the propensity score matching (PSM), these patients were compared with other 35 patients with a similar baseline profile who were operated on under the same clinical conditions, however, who received a SAG in the procedure. Therefore, the immediate and mid-term clinical results of patients operated on with ≥ 2 arterial grafts (MAG group) vs. patients with one arterial graft (SAG group) in the AMI were compared in relation to need for readmission, new AMI, reoperation, and death from all causes.

The methodology flowchart is shown in Figure 1.

Fig. 1 - Methodology flowchart. AMI=acute phase myocardial infarction; CABG=coronary artery bypass grafting; MAGs=multiple arterial grafts; SAG=single arterial graft; REPLICCAR II=Registro Paulista de Cirurgia Cardiovascular II.

Download

Fig. 1 - Methodology flowchart. AMI=acute phase myocardial infarction; CABG=coronary artery bypass grafting; MAGs=multiple arterial grafts; SAG=single arterial graft; REPLICCAR II=Registro Paulista de Cirurgia Cardiovascular II.

Statistical Analysis

R software version 4.0.2 was used to carry out all the analyzes conducted in this study.

In the descriptive analysis, continuous variables were expressed as mean, median, standard deviation, and quartiles (25th/75th percentiles), while categorical variables were expressed in terms of frequencies and percentages. Due to missing data, percentages were calculated using the number of responses obtained by variables instead of the total number of patients.

PSM was used to pair the groups using the GenMatch function, available in the MatchIt package of R software, and its quality was verified by using the standardized mean difference (SMD) method (Table 1). The following variables were used to match the groups: sex, age, diabetes management, creatinine level, left ventricular ejection fraction, and need for intravenous nitrates (in the last 24 hours before surgery).

Table 1 - Standardized mean difference (SMD) before and after propensity score matching (PSM).

PSM variable	SMD before PSM	SMD after PSM
Sex	0.144	0.085
Age	0.618	0.224
Diabetes management	0.536	0.144
Creatinine level	0.143	0.042
Left ventricular ejection fraction	-0.269	-0.078
Need for intravenous nitrates (in the last 24 hours before surgery)	0.247	0.085

SMD interpretation: 0 - 0.2, almost no difference; 0.2 - 0.5, little difference; 0.5 - 0.8, medium difference; 0.8 - 1, big difference

Table 1 - Standardized mean difference (SMD) before and after propensity score matching (PSM).

For the comparison of continuous variables from the two groups, t-test was used for normally distributed variables (Anderson-Darling test) and non-parametric tests were used for the other variables. Mann-Whitney test was used for homogeneous variables, and the Brunner-Munzel test was used for heterogeneous variables. Regarding categorical variables, Fisher's exact test or chi-square test was used. The significance level adopted in the tests was 0.05. Two-tailed hypotheses were considered. Furthermore, the constructed confidence intervals have a 95% confidence level.

RESULTS

In the AMI, 238 patients underwent CABG; of these, 35 patients received MAG (15%) in REPLICCAR II. After pairing with 35 patients from the SAG group, the groups did not present statistically significant differences between the preoperative variables (Table 2).

Table 2 - Propensity score matching preoperative variables.

Regarding the intraoperative period (Table 3), there was a significant statistical difference in surgery duration (period between surgical incision and referral to the intensive care unit), where the MAG group had a median of 4.78 hours while the SAG group had 4.11 hours (P=0.040), and there was no significant statistical difference in cardiopulmonary bypass time (P=0.560) or cross-clamping time (P=0.723). MAG group had a predominant use of BITA (62.86%), followed by radial graft associated with the use of left internal thoracic artery (28.57%) and combination of the three grafts (8.57%), with 28.57% not using venous grafts.

Table 3 - Propensity score matching intraoperative variables.

No significant differences were found in the immediate outcomes (Table 4).

Table 4 - Propensity score matching variables from the follow-up (up to 30 days).

Regarding the mid-term follow-up (3-5 years after surgery) in Table 5, 11.43% of patients were lost to follow-up, resulting in a total of 62 patients to be analyzed, 31 in each group. There was one fatal new AMI case in the SAG group, and two cases in the MAG group, where one patient died. With regard to hospital readmissions, non-cardiac causes (77%) were more frequent than cardiac causes (23%) in both groups, and the same applies to the causes of death, 75% and 25% respectively. There were two cases of angioplasty in the MAG group and one case of redo CABG in the SAG group.

Table 5 - Propensity score matching mid-term follow-up variables.

The authors present a Kaplan-Meier survival curve (Figure 2), where is observed a superior survival probability for the MAGs group (P=0.63).

Fig. 2 - Kaplan-Meier-survival curves for death outcome by the evaluated groups.

Download

Fig. 2 - Kaplan-Meier-survival curves for death outcome by the evaluated groups.

DISCUSSION

This study provides evidence that the use of MAGs during CABG in the AMI does not cause harm in the short and mid-terms when compared with the use of a SAG. This comes to fill a gap in relation to the use of the best grafts during the AMI in patients operated on urgently. Approximately 5-10% of patients in the AMI require CABG^[12,13] and represent a challenging subgroup due to their high-risk characteristics compared to patients undergoing elective CABG.

The advantages of the use of MAGs are related to the long term^[6], even with a higher rate of infection, technical difficulty, and increased surgical duration^[14,15]. The present analysis in patients in the AMI shows that the group with MAGs did not present a clinically significant difference in relation to the use of a SAG in the mid-term. Therefore, although there is a preference for not using MAGs in the AMI, this should be guided by thinking about the long-term benefits.

Taggart et al.^[16], in the Arterial Revascularization Trial (or ART), analyzed randomized patients regarding the impact of MAGs vs. SAG on CABG. In the mid-term follow-up, only the higher rate of deep sternal wound infection in the MAGs was pointed. Our study, which also included a radial graft in the MAG group in patients in the AMI, also showed no mid-term differences in relation to the clinical outcomes analyzed, with a tendency towards a higher rate of deep sternal wound infection.

In the immediate postoperative period, the SAG group had a slight increase in hours spent in the intensive care unit (77.71 ± 50.43 vs. 57.09 ± 26.23, P=0.135) and hospital stay (7.37 ± 3.46 vs. 6.94 ± 2.44, P=0.924). We believe that the potential for a reduction in the length of hospital stay in the MAG group is influenced by early postoperative ambulation due to the non-use of the saphenous vein, culminating in both clinical and psychological improvements already described in rapid recovery protocols^[17].

Although the groups had a similar baseline glycemic profile, a greater tendency for complications of surgical wounds was demonstrated in the MAG group, a cause already known for rejection of the multiple arterial technique by many health teams^[14,15]. The graft choice falls short of adequate glycemic control since the rigorous management of glycemic parameters is crucial for the positive evolution of patient’s outcomes. In our analysis, there was a less rigid interoperative management of glycemic parameters in the MAGs group, where the higher blood glucose in the group reached an average of 196.66mg/dL (P=0.247) followed by a higher frequency of infections outcomes.

Dorman et al.^[18] compared controlled diabetic patients undergoing CABG with SAG and MAGs in a 30-year follow-up and demonstrated that the use of MAGs did not increase morbidity or mortality rates; Zhou et al.^[19] also evaluated this patient profile and showed benefits in the use of the MAGs for well-controlled diabetic patients. This supports the hypothesis that, regardless of whether the diagnosis of diabetes, what influences surgical wound infection is the perioperative glycemic control.

About hospital readmission within 30 days of discharge, one patient in the MAG group had to return due to pleural effusion requiring thoracentesis. In the mid-term follow-up, five patients in the SAG group vs. eight patients in the MAG needed to return to the hospital, mostly due to non-cardiac causes. In the MAG group, the need for angioplasty was observed in two cases, while in the SAG there was a reoperation and a new CABG.

The SAG group had a higher mortality, although not statistically significant, with five (16.13%) deaths vs. three (9.68%, P=0.44). The occurrence of a new myocardial infarction with fatal outcome was observed in only one case in each group, given that deaths from non-cardiac causes were more frequent. Enezate et al.^[20] evaluated the difference between percutaneous revascularization and CABG for AMI patients, separating SAGs and MAGs, demonstrating that there were benefits related to a reduction in the mortality rate in the MAG group, similar to our findings.

Grothusen et al.^[21] retrospectively studied patients who had AMI complicated by cardiogenic shock, submitted to CABG hours after the diagnosis, showing that surgical treatment can bring benefits to patients, however, the techniques of SAG vs. MAG were not considered. In this scenario, more studies still need to be carried out to correlate the best AMI patient care strategy, providing satisfactory clinical evolution and longevity.

CONCLUSION

In this analysis, the use of MAGs, even in the AMI, did not bring disadvantages in the shortand mid-term follow-ups compared to patients with a SAG. Therefore, we encourage the use of MAGs given the long-term benefits for patients.