A Single-Center Experience in Low Ejection Fraction Coronary Artery Bypass Surgery

Background . Coronary artery bypass graft (CABG) in patients with an ejection fraction (EF) ≤ 35% predisposes them to higher complications and mortality risks. Given the usually compromised status of other end organs in low EF patients, ONCAB, involving cardiopulmonary bypass (CPB) and aortic cross-clamping, might intuitively pose more complications than OPCAB. Objective . To explore short-and long-term outcomes between ONCAB and OPCAB procedures in patients with EF ≤ 35%. Methods . A retrospective and observational analysis was conducted in 196 patients with EF ≤ 35% who underwent ONCAB ( n =58) or OPCAB ( n =138) procedures at a single center between January 2015 and May 2023. Baseline characteristics were well matched using the stabilized inverse probability treatment weighted matching technique. Results . After matching, ONCAB and OPCAB had comparable 30-day mortality and 30-day cardiac mortality. OPCAB exhibited signifcantly shorter length of hospital and ICU stays, with a trend towards more discharges to home. Rates of composite complication and its individual components such as acute kidney injury, reoperation bleeding, stroke, pneumonia, GI disease, and atrial fbrillation were similar between the two groups. Rates of sepsis, liver dysfunction, and blood transfusion were signifcantly lower in the OPCAB group. As assessed by EF and LVDD, neither procedure showed superiority in improving cardiac function. Median follow-up time was 4.9 (interquartile range: 2.1–7.2) years. After matching, long-term overall survival (1, 3, 7years) and cardiac mortality rates were comparable between OPCAB and ONCAB. Cumulative rates of cardiac arrest, heart failure, myocardial infarction (MI), atrial fbrillation (Afb), renal disease, and readmission (overall and cardiac) at 7years were similar. Conclusion . Tis study demonstrates comparable short-term and long-term outcomes between ONCAB and OPCAB in patients with reduced EF, with OPCAB favoring faster recovery. OPCAB appears as a safer and equally efective option for low EF CABG patients. Larger samples and longer follow-ups are needed for conclusive clinical evidence.


Introduction
Coronary artery disease (CAD) is the most common cause of heart failure leading to the highest number of deaths in the USA and worldwide [1,2].Coronary artery bypass grafting (CABG) is a well-established surgical procedure for the management of CAD and its associated complications [3,4].Over the years, there has been a growing interest in comparing the outcomes of on-pump (ONCAB) and of-pump (OPCAB) CABG surgeries [5][6][7][8][9][10][11], particularly in patients with reduced EF [12,13].Reduced EF, defned as an EF of 35% or less, is a strong predictor of adverse cardiac events and is associated with increased morbidity and mortality following cardiac surgeries [14,15].Terefore, determining the optimal surgical approach for patients with reduced EF is of paramount importance.However, few trials have compared outcomes from OPCAB and ONCAB procedures in CAD patients with severely reduced EF [6,12,16].Current evidence on the superiority of one procedure over the other remains inconclusive.
We studied short-term (30-day) and long-term (7-year follow-up) outcomes between the two procedures in our CABG patients with severely reduced EF (≤35%).Given the usually compromised status of other end organs in low EF patients and systemic infammatory response syndrome (SIRS) caused by the use of cardiopulmonary bypass (CPB) in ONCAB, one would intuitively expect higher complication rates postoperatively.In addition, coronary collateral circulation may be disrupted due to transient changes in ventricular anatomical geometry during the use of CPB [17], which may provide further challenges on the functionality of global ischemia that was already compromised due to CAD [18].

Design and Study
Population.We conducted a retrospective, single-center observational study including patients who underwent CABG surgery between January 2015 and May 2023.Te study focused on patients who underwent either onpump coronary artery bypass grafting or of-pump coronary artery bypass grafting procedures and had an EF ≤ 35%.Patients who underwent concomitant valve procedures with CABG or minimally invasive coronary artery bypass surgery were excluded from the study.Te research ethics board at BSWH approved this study and waived the requirement for individual patient consent, as the data collected were not identifable.

Data Collection.
We used the STS registry data for collecting baseline demographics, historical risk factors, perioperative, and postoperative short-term attributes.For the short-term cohort, the primary outcome was 30-day mortality of any cause, and the secondary outcomes were postoperative complications, hospital stay, ICU, and ventilation times.All variables were defned according to the STS standards.For the 7-year follow-up cohort, we used our institutional database.Te primary outcome was 7-year mortality of any cause, and the secondary outcomes were stroke, infection of any causes, cardiac arrest, myocardial infarction, renal failure, readmit for any reasons, cardiac readmission, atrial fbrillation, and heart failure.Te identifcation of these variables was retrieved from the operative notes written by the surgeon or physician assistant.

Statistical Analysis.
Te statistical analyses were conducted using SAS EG 9.4.Quantitative parametric variables were summarized as the mean ± standard deviation, while nonparametric variables were summarized as the median (interquartile range: IQR).Group comparisons were performed using standard t-tests, Mann-Whitney U tests, or quantile regression tests as appropriate.Categorical variables were presented as numbers (%) and analyzed using the chi-square test or Fisher exact test.Te design for the presurgery and postsurgery continuous variables involved repeated measures ANCOVA.
For the 7-year follow-up data, time-to-event analysis (Cox regression) was utilized.Kaplan-Meier survival curves were used to present the primary outcome (mortality), and the log-rank test was used for subgroup comparisons with the use of stabilized inverse probability treatment weighted sample.Failure plots, hazard ratios, and 95% confdence intervals were used to present the secondary outcomes, and comparisons between groups were performed using the log-rank test of inequality with the stabilized IPTW rounded sample.
To address baseline imbalances and confounding factors between the OPCAB and ONCAB groups (as shown in Table 1), we employed the stabilized inverse probability treatment weighting (IPW) matching technique.Tis approach assigns weights to individuals based on their propensity score to receive a specifc treatment, resulting in improved balance and reduced bias [19].Covariates such as age, BMI, gender, diabetes, chronic lung disease, liver dysfunction, kidney dialysis, stroke, peripheral vascular disease, myocardial infarction, hypertension, arrhythmia, number of deceased vessels, sleep apnea, hemoglobin, creatinine, and EF were incorporated in the matching process.Te balance achieved through weighted matching was confrmed through examination of the distribution of logit propensity scores plots (Figure 1) and statistical diferences in each covariate between the treatment groups (Table 1).All statistical signifcance was defned as a two-sided p value of less than 0.05.

Results
2,150 patients had cardiac surgery at our cardiovascular clinic during seven years, from January 1, 2015, to May 31, 2023.1,022 of them had undergone isolated coronary artery bypass grafting (CABG) operations.To create the study cohort, we further chose 196 patients (13.8% of the CABG patients) from this group who had an EF ≤ 35%.We fnally split these patients into two groups to examine the outcomes: on-pump CABG (58 patients, or 29.6%) and of-pump CABG (138 patients, or 70.4%).Following a rigorous process of stabilized IPTW matching, we successfully obtained a weighted cohort of 54 patients from the ONCAB surgery and 137 patients from the OPCAB group.Te postoperative and 7-year follow-up outcomes of these patients are meticulously presented in the following analysis.

7-Year Follow-Up Mortality
Outcomes.Te median follow-up period was approximately 2 years in the OPCAB group and 5 years in the ONCAB group.Troughout the 7year follow-up, a total of 50 patients (25.5% of the entire cohort) passed away, with 29 of them (58%) succumbing to cardiac disease.

Discussion
Te present study aimed to compare the outcomes of ONCAB and OPCAB surgeries in patients with reduced EF (≤35%).Te analysis included a well-matched cohort of 54 ONCAB and 137 OPCAB patients with similar baseline characteristics that include patients' age, BMI, sex, comorbidities, previous myocardial infarction, or history of heart failure.Most of our patients in both groups had multivessel disease.Operative features showed that the OPCAB group used fewer number of grafts.In the context of grafting, our surgical team prioritizes the use of high-quality arterial grafts over their quantity.Low EF patients typically experience signifcant collateral damage, and our philosophy holds that, in this context, one IMA is better than two veins.Tis is evident in the comparable long-term outcomes observed between the two cohorts, despite the lower graft count in the OPCAB group.Further, the OPCAB group demonstrated a higher utilization of bilateral and skeletonized IMAs.It is noteworthy, at the beginning of this discussion, that these discrepancies may be refective of the surgical team's philosophy rather than a generalized approach.Te rates of cardiogenic shock cases were comparable between the two groups.
Regarding short-term postoperative outcomes, the 30-day mortality rate was signifcantly lower in the OPCAB group compared to the ONCAB group.However, after matching, the rates of 30-day mortality and 30-day cardiac mortality were similar between the two groups.A recent research work found no signifcant diference in short-term mortality between the two surgical approaches in patients with reduced EF, which aligns with our matched analysis [16].Tis suggests that the initial disparity in 30-day mortality may be attributed to diferences in baseline characteristics that were subsequently mitigated through matching.
Few other studies further support our fndings, demonstrating no signifcant diference in 30-day mortality between OPCAB and ONCAB groups [6,11,20].Tis indicates that the surgical approach itself may not signifcantly impact short-term mortality rates.Instead, other factors such as patient characteristics, surgical expertise, and perioperative care may play a more substantial role.Furthermore, we found that the rates of 30-day cardiac mortality were similar between the OPCAB and ONCAB groups, with rates of 6.6% and 7.4%, respectively.Tese results align with other studies [21,22] that showed no signifcant diference in 30-day cardiac mortality rates between the two approaches.Tis consistency in fndings across diferent studies suggests that both OPCAB and ONCAB procedures have comparable efcacy in terms of reducing cardiac mortality within the early postoperative period.
Te convergence of 30-day mortality rates between the OPCAB and ONCAB groups after matching aligns with other studies, demonstrating comparable outcomes in case of short-term mortality.Te absence of signifcant diferences in 30-day cardiac mortality further supports the notion that both approaches are equally efective in reducing cardiac-related mortality within the early postoperative period.Tese fndings emphasize the safety of performing low EF CABGs by experienced OPCAB teams.
Te OPCAB group showed shorter lengths of hospital stay, shorter ICU stay, and a trend towards lower rates of complications such as AKI, stroke, pneumonia, GI complications, and atrial fbrillation.In addition, the OPCAB group showed higher rates of discharge-to-home meaning 10.6% lower rates of discharge to acute care facilities compared to the ONCAB group (Table 2).Notably, sepsis and blood transfusion rates were signifcantly lower in the OPCAB group.Furthermore, the OPCAB group had a lower

Survival Probability
All-cause mortality before matching AFib before matching AFib after matching HR = 1.4 (0.9-2.2); Log-rank P = .145HR = 1.5 (0.9-2.4); Log-rank P = .078 Heart failure after matching Heart failure before matching HR = 0.9 (0.6-1.4); Log-rank P = .612HR = 1.1 (0.7-1.7); Log-rank P = .714 Readmission before matching HR = 0.9 (0.5-1.5); Log-rank P = .604HR = 1.4 (0.7-2.7); Log-rank P = .327HR = 1.3 (0.7-2.5); Log-rank P = .437HR = 0.96 (0. Figure 4: Kaplan-Meier survival curves were plotted to compare the secondary outcomes between OPCAB and ONCAB, incorporating hazard ratio (HR) and log-rank statistics for both procedures prior to and following the application of stabilized weighted matching.For each outcome, the fgures also include the count of patients at risk for each procedure at various time points.incidence of postoperative liver dysfunction.Tese fndings align with the results of a multicenter trial, which reported similar advantages of OPCAB over ONCAB in terms of shorter hospital stays and reduced rates of postoperative complications [23].Another study reported similar results, highlighting a shorter hospital stay in OPCAB patients compared to ONCAB patients [21].Although our study showed no statistically signifcant diference in the median ventilation time between the OPCAB and ONCAB groups, the OPCAB group exhibited about 3 times shorter (∼12 hours) stay on the ventilation support (Table 2).However, we observed a signifcantly shorter median ICU stay in the OPCAB group compared to the ONCAB group.Tis aligns with the results of other research studies that reported a shorter ICU stay in OPCAB patients [24,25].Te shorter hospital stays, reduced ICU stay, or a trend of reduced ventilation hours in the OPCAB group may refect the lower incidence of postoperative complications and faster recovery, leading to a shorter duration of critical care management.
Furthermore, our study found that the rates of acute kidney injury (AKI) were similar between the OPCAB and ONCAB groups.Tis is consistent with a recent research reported by Phothikun et al. 2023 and by a meta-analysis conducted by Parissis et al., who showed comparable rates of AKI in both groups [26,27].However, we observed signifcantly lower rates of sepsis and postoperative liver dysfunction in the OPCAB group compared to the ONCAB group.Tese fndings are in line with another study that also reported lower rates of sepsis and liver dysfunction in OPCAB patients [28].Te lower incidence of sepsis and liver dysfunction in the OPCAB group may be attributed to the reduced need for blood transfusion and less systemic infammatory response associated with of-pump surgery.
Regarding mechanical assist devices, a vast majority (94.3% of all devices and 33.8% of all cases) consisted of preoperative IABPs with 9% higher in the OPCAB group.Tis is worth mentioning here that the OPCAB group had signifcantly lower preoperative EF.In addition, the higher rate of IABP utilization in the OPCAB procedure is a common phenomenon aimed at preventing hypotension during surgery and avoiding conversion to ONCAB and does not result in extended ICU stay (our conversion rate was <2.2%).Notably, these IABPs were promptly removed in the early postoperative phase.Furthermore, all 8 Impella devices were used postoperatively, 1 in ONCAB and 7 in OPCAB, with 6 of the latter group having IABPs preoperatively.
Te study also assessed the impact of the two procedures on the improvement of left ventricular diastolic diameter (LVEDD) and EF.Te results showed that both OPCAB and ONCAB surgeries led to signifcant improvements in postdischarge EF, while the changes in postdischarge LVEDD were not signifcant.Few studies have examined the impact of OPCAB and ONCAB on left ventricular function, particularly EF.For example, a study conducted by Letsou et al. found a similar improvement but no signifcant differences in postoperative EF between OPCAB and ONCAB patients [29].Similarly, Youn et al. reported comparable results in terms of overall left ventricle function after OPCAB and ONCAB procedures [30].It is important to note that our study and previous literature have shown a signifcant improvement in postdischarge EF from the presurgical stage in both OPCAB and ONCAB groups.Tis indicates that regardless of the surgical approach, coronary artery bypass grafting leads to an overall improvement in left ventricular function.Te improvement in EF may be attributed to the restoration of blood fow to the ischemic myocardium and reactivation of hibernating myocardium.
In terms of long-term outcomes, the 7-year follow-up revealed that the survival rates and cardiac mortality rates were similar between the OPCAB and ONCAB groups.Tis fnding is consistent with a recent study by Neumann et al., which analyzed long-term outcomes in patients with reduced EF (≤35%) undergoing CABG [16].Tey did not fnd any signifcant diference in survival rates between ONCAB and OPCAB groups over a 3-year follow-up period.Tese fndings are consistent with several previous studies that have also reported comparable long-term survival rates between OPCAB and ONCAB patients.For instance, a study by Jiang et al. conducted a meta-analysis of randomized controlled trials and found no signifcant diferences in longterm survival between the two groups [11].Another study by Matkovic et al. reported similar results, supporting the notion that both OPCAB and ONCAB procedures provide comparable long-term (6 years) survival in patients with poor left ventricle function [31].Tese fndings were in line with another meta-analysis conducted by Zhang et al., which reported similar long-term mortality between the two procedures in patients with redo CABG settings [32].Furthermore, our analysis of mortality due to cardiac causes also revealed similar rates between the OPCAB and ONCAB groups at 1 year, 3 years, and 7 years.Tis indicates that the choice of surgical approach does not signifcantly impact the risk of mortality specifcally attributed to cardiac disease.Tese fndings align with previous studies that have consistently shown comparable cardiac mortality rates between OPCAB and ONCAB patients.For example, a study by Youn et al. found no signifcant diference in cardiac mortality between the two procedures during a long-term follow-up in patients with reduced EF (≤35%) [30].Similarly, a recent study by Bassano et al. reported similar results, highlighting the equivalent 14-year cardiac mortality outcomes between OPCAB and ONCAB surgeries [33].Other studies have reported similar fndings between the two groups [30,[34][35][36][37].
Te overall hazard of CVA events was notably higher in OPCAB.Nevertheless, the wide confdence interval of 111.2 indicates that this signifcance is accompanied by a low precision of the estimate, possibly due to an insufcient sample size and hence a lack of the occurrence of CVA events in the proportional hazard analysis.Importantly, no signifcant diference in the occurrence of CVA events was observed over a 7-year period between the ONCAB and OPCAB groups in cases involving bilateral IMA with total arterial grafts (log-rank p � 0.752).Te distinction between single and bilateral IMA, as well as the choice between arterial and venous grafts, may play a crucial role in infuencing the occurrence of long-term CVA events.Terefore, this study calls for further assessment of longterm CVA diferences between the two procedures particularly focusing on the distinctions between single and bilateral IMA and the choice between venous and arterial grafts.However, there were no signifcant diferences in the rates of cardiac arrest, heart failure, myocardial infarction, atrial fbrillation, renal disease, and overall or cardiac readmission.
From a physiological perspective, the underlying mechanism behind the comparable long-term outcomes can be attributed to the shared goal of revascularization achieved by both on-pump and of-pump CABG procedures.Both techniques aim to restore blood fow to the ischemic myocardium, thereby improving cardiac function and reducing the risk of adverse cardiac events [38,39].
Overall, the fndings of this study suggest that both ONCAB and OPCAB surgeries can be performed safely in patients with reduced EF.Although the short-and long-term mortality was comparable between the two procedures accomplishing the primary objective of successful revascularization of the myocardium and the improvement of overall cardiac function, the OPCAB group showed favorable short-term outcomes with shorter hospital stays, shorter ICU stays, and lower rates of complications such as sepsis, blood transfusion, and postoperative liver dysfunction.Tese fndings are well supported by previous studies that compare the outcomes between the OPCAB and ONCAB procedures in low EF patients [12,16,40].
Although our results are comparable with several previous studies and hence may add value to the scientifc community in the feld, the study has certain limitations.First, it has a retrospective design and has the potential for unmeasured confounders.We believe that stabilized inverse probability treatment weighted matching could mitigate this bias [19].A randomized controlled trial may be warranted to address the potential confounders and selection bias.Second, further prospective studies with larger sample sizes and longer followup periods are warranted to validate these fndings and provide more comprehensive insights into the long-term outcomes of CABG surgeries in this high-risk patient population.
In conclusion, our results are consistent with previous studies, indicating similar short-and long-term mortality rates between the two procedures.Furthermore, our study highlights the advantages of OPCAB in terms of shorter hospital stays, reduced postoperative complications, and comparable long-term outcomes.Terefore, the OPCAB is a safer and equally efective option for the low EF CABG patients.

Figure 2 :
Figure 2: Repeated-measures ANCOVA tests of left ventricular (LV) diastolic diameter and ejection fraction between presurgery and postsurgery periods for the OPCAB and ONCAB groups.Here, the error bars are the standard errors of means.

1 . 0 Figure 3 :
Figure3: Kaplan-Meier survival curves were plotted to compare the mortality outcomes between OPCAB and ONCAB, incorporating hazard ratio (HR) and log-rank statistics for both procedures prior to and following the application of stabilized weighted matching.Te fgures also include the count of patients at risk for each procedure at various time points.