Coronary Arteries Bypass Grafting as a Salvage Surgery in Ischemic Heart Failure

Abstract

Ischemic cardiomyopathy accounts for approximately two-thirds of all Heart Failure (HF) cases. Recent studies indicates that revascularization provides superior outcomes compared with optimal medical therapy (OMT) alone. Current European and American guidelines recommend an invasive approach in patients with reduced left ventricular ejection fraction (LVEF) less than 35% and with multivessel disease (MVD). Randomized controlled trials in these patients have proven that long-term survival is greater following coronary artery bypass grafting (CABG) than with OMT alone. Patients with ischemic cardiomyopathy and coronary artery disease that is amenable to surgical revascularization should undergo combination of surgical revascularization and medical therapy rather than medical therapy alone. In some cases, combined CABG with other surgeries are vital salvage procedures, such as atrial fibrillation, mitral valve, tricuspid valve, and LV remodeling. Based on small but, nontrivial, early mortality risk associated with CABG surgery as well as other post-CABG morbidities, patients may also reasonably choose medical therapy as initial treatment option. Revascularization remains an important treatment option for patients with ongoing anginal symptoms despite optimal medical therapy. In this chapter, we will highlight the role of CABG in heart failure treatment and when to use it as a salvage surgery before referring the patient for heart transplantation.

1. Introduction

There is no universally accepted definition of ischemic cardiomyopathy (ICM). However, the term ischemic cardiomyopathy generally refers to significantly impaired left ventricular function (left ventricular ejection fraction [LVEF] ≤35–40%) that results from coronary artery disease (CAD) [1, 2, 3]. In 2002, Felker et al. suggested that the symptomatic patients with LVEF ≤40% and presence of left main or proximal left anterior descending coronary artery stenosis ≥75% or two or more epicardial coronary artery stenosis ≥75% or a prior history of coronary artery revascularization [percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG)] or prior history of myocardial infarction should only be classified as having ICM [3].

Ischemic heart disease is a global pandemic, and its incidence continues to increase. In an estimate, 125 million people across the globe suffer from ischemic heart disease. In the United States itself, every year 720,000 people develop their first myocardial infarction (MI) resulting in hospitalization and/or death [4, 5]. Thirty-five percent of the patients who experience coronary event in a given year die due to it; and each death is associated with an average of 16 years of lost life. Patients who survive after the myocardial infarction are at an increased risk of developing ICM and eventually heart failure (HF). Etiopathogenesis of heart failure is multifactorial; however, ischemic cardiomyopathy is the single most common cause of heart failure. More than 64.3 million people across the world and 6 million people in the United States currently experience HF [3, 6]. In addition to increase in human toll, the estimated cost of HF exceeds $60 billion each year [7, 8].

2. Pathophysiology of ischemic cardiomyopathy

In patients with coronary artery disease, rupture of atherosclerotic plaque followed by in situ thrombus formation leads to sudden cessation of coronary blood flow. If the coronary blood flow is not established early enough either by spontaneous, pharmacological, or interventional recanalization, the death of ischemic myocytes ensues. With time, dead myocytes are replaced with fibrous tissue. Once the amount of scarred myocardium is significant enough after single or multiple episodes of MI, the left ventricle remodels with dilatation, regional deformation, and decrease in overall contractility. Remodeling and alteration of LV geometry especially the inferior wall may also lead to papillary muscle malalignment and mitral regurgitation (MR). Left ventricle volume overloading due to chronic MR in association with poor left ventricular contractility sets up a vicious cycle of worsening LV remodeling and MR [9].

The replacement of the dead myocardium with fibrous tissue is the most important mechanism in the development of ICM. Other pathophysiological processes such as myocardial stunning and hibernation that render the viable myocardial cells unable to perform their mechanical work and also contribute to the development of ICM. Both myocardial stunning and hibernation are reversible forms of myocardial contractile dysfunction that have the potential of mechanical work restoration if the blood flow supply can be improved [10]. In any given heart with ICM, all three stages of myocardium, i.e., normal, viable but hypocontractile and scarred myocardium often coexist within a single cross section of LV. Thus, ischemic cardiomyopathy is extremely heterogeneous and particularly challenging for accurate viability assessment with imaging studies [11].

The concept of hibernating myocardium is interesting as well as mysterious. Our present understanding about the hibernating myocardium is limited [12, 13, 14, 15, 16]. Rahimtoola [17] described the hibernating myocardium as “resting left ventricular dysfunction due to reduced coronary blood flow that can be partially or completely reversed by myocardial revascularization and/or by reducing myocardial oxygen demand.” Hibernating myocardium is usually limited to subendocardial tissues. Histologically, in hibernating myocardium, there is loss of contractile proteins and sarcoplasmic reticulum without the change in the cell volume. Presumably, hibernation is a protective dedifferentiation of myocardial cells or switch to a quiescent state of decreased mechanical work in times of chronically decreased oxygen supply [13]. This adaptive mechanism probably allows the myocytes to avoid the ischemic imbalance and remain alive in the milieu of decreased coronary blood flow that would otherwise lead to cell death. Alternative mechanism for ventricular dysfunction in ICM may be myocardial stunning. Myocardial stunning apparently occurs due to repeated episodes of ischemic insult that result in viable but chronically hypocontractile myocardium (i.e., repetitive stunning). Due to extremely low ischemic threshold of the myocytes, any decrease in coronary blood flow during stress leads to ischemia and ischemia–reperfusion changes in the myocytes despite normal or insignificantly decreased resting coronary perfusion [13, 18]. This repetitive stunning of the myocytes results in chronic LV dysfunction. Thus, in patients with ICM, territories with high numbers of cardiomyocytes with excess glycogen reserve and less fibrosis in all probabilities are reversible after revascularization. These myocytes also demonstrate higher blood flow and glucose uptake on positron emission tomography (PET) scan [19].

3. Preoperative considerations

Patients with ICM present with myriad of signs and symptoms depending upon the severity of heart failure and degree of physiological compensation. Some patients may be asymptomatic or minimally symptomatic with mild anginal chest pain and dyspnea on exertion while other patients may present with overt heart failure symptoms, e.g., dyspnea, orthopnea, poor exercise tolerance, and increased fatigability. Patients usually have a longstanding history of coronary artery disease and a prior history of myocardial infarctions. Physical examination can reveal bibasilar crackles, S3 gallop, displaced apical impulse, carotid bruits, jugular venous distension, positive hepato-jugular reflex, and bilateral lower extremity edema.

3.1 Diagnostic testing

In patients with ICM, multivessel disease, low LVEF, and increased LV end-systolic volumes are important prognostic factors. Therefore, all these factors must be taken into consideration when making the difficult decision regarding revascularization. Suitability of the patient for CABG depends upon: A) suitability of the diseased coronary arteries for bypass grafting; B) the amount of viable myocardium present and whether the viable myocardium is present in the territory of CAD; C) severity of right and left heart failure; and D) associated cardiac lesions. All the diagnostic investigations should be directed toward determining whether the patient is a suitable candidate for CABG or not.

Transthoracic echocardiography (TTE): Transthoracic echocardiography is an essential investigation in assessing myocardial viability in a patient with ICM. Echocardiography is useful in evaluating cardiac anatomy, valvular function, ventricular systolic/diastolic function, cardiac wall motion, and pericardial pathology. All this information is useful in diagnosing ischemic cardiomyopathy, especially in patients with HF and other high-risk features.

Coronary angiography: Coronary angiography allows direct visualization of the coronary arteries for assessment of severity of obstruction, collateralization, and the blood flow to the myocardium. Coronary angiography is most important in defining the extent and severity of coronary artery disease and whether the coronaries arteries are suitable for grafting. Computed tomography coronary angiography can also be performed in place of conventional coronary angiography to assess coronary arteries in patients with low to intermediate risk of CAD [20].

Cardiac stress test: There are different stress tests available depending on the patient’s health, functional status, baseline heart rhythm, and exercise tolerance. The goal of these stress tests is to assess for cardiac ischemia and myocardial viability. Late gadolinium enhancement cardiac magnetic resonance (LGE-CMR), dobutamine stress echocardiography, single-photon emission computed tomography (SPECT), and F-18- fluorodeoxyglucose positron emission tomography (FDG-PET) imaging can be used to assess myocardial viability [21]. Dobutamine stress echocardiography is widely used to assess myocardial contractility reserve and viability. With continuous dobutamine infusion, initially myocardial perfusion increases along with increased contractility. However, as the dobutamine dose increases, blood flow cannot be escalated further leading to reduced myocardial contractility. This phenomenon known as biphasic reaction can predict the recovery of the myocardial function after revascularization.

Late gadolinium enhancement cardiovascular magnetic resonance (LGE-CMR) can detect increase in extracellular space due to myocardial apoptosis and necrosis and can predict the reversibility of the myocardial contractility after successful revascularization while dobutamine stress CMR can detect the ischemic myocardium. In patients with ICM with transmural infarct, minimal LGE (<25%) in dysfunctional myocardial segment indicates a high likelihood of recovery while the chance of recovery is minimal in segments with >50% LGE.¹³ In segments with 25–50% LGE involvement, the recovery prediction is not consistent [22].

Single-photon emission computed tomography (SPECT) and positron emission tomography (PET) had been widely utilized in the past to assess myocardial viability. Thallium-based SPECT scan demonstrate delayed distribution but has increased risk of ionizing radiations while technetium-based SPECT has less risk of radiation, but it cannot demonstrate a delayed distribution. Another nuclear imaging modality to assess myocardial viability is cardiac PET. PET imaging is based on the principle that in an ICM, ischemic myocardium switches to glucose-based metabolism instead of fatty acids. ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) can detect this shift in viable but ischemic myocardium. PET has higher spatial resolution, lower risk of radiation, and better attenuation correction compared with SPECT. PET, however, cannot distinguish between normal and ischemic or hibernating myocardium in patients with insulin resistance, and results may be inaccurate in patients with variable uptake of FDG due to heart failure [23].

Brain natriuretic peptide (BNP) test: BNP is synthesized in the ventricles, and it is secreted when the myocardial muscle has a high wall tension. BNP is an important biomarker for heart failure patients. Increasing trend in BNP suggests worsening of heart failure; however, it cannot detect myocardial ischemia.

4. Clinical studies and randomized trials in patients with ischemic cardiomyopathy

Coronary artery bypass grafting for CAD started in the mid-1960s. Since then, numerous clinical trials and studies have tried to address different questions related to the management of CAD. All these trials and studies have established an undisputed role of surgical revascularization in patients with CAD in terms of improved survival, risk of reintervention, and quality of life [24, 25, 26, 27]. Nevertheless, prior to Surgical Treatment for Ischemic Heart Failure (STICH) trial [28], none of the studies specifically addressed the management of patients with ICM. The coronary artery surgery study (CASS) trial registry that followed the patients who were excluded from the main study reported that patients with LVEF <35% had better survival with CABG than with medical therapy, if they had associated three-vessel disease and if the presenting symptom was angina [29]. Similarly, a 25-year observational study involving 1391 patients (medical therapy (n = 1052) or CABG (n = 339)) from Duke Cardiovascular Disease Databank also reported an improved survival with CABG over medical therapy alone after 30 days to more than 10 years in patients with NYHA class ≥II, CAD with at least one vessel stenosis ≥75%, and LVEF <40%. The benefit with CABG was observed irrespective of the extent of coronary artery involvement (P < 0.001) [30].

These observational studies pointed toward the role of CABG in patients with ischemic cardiomyopathy; however, lack of randomized clinical studies in patients with ICM led to different therapeutic approaches driven by the physician bias regarding the potential benefit of myocardial revascularization [9]. The resulting equipoise formed the basis for the multiinstitutional STICH randomized controlled clinical trial [28]. STICH trial was the first and only large-scale randomized clinical trial to compare surgical revascularization with medical therapy in patients with LVEF ≤35% and CAD amenable to CABG. The STICH trial randomly assigned 1212 patients to three groups (medical therapy alone, medical therapy with CABG, and medical therapy with CABG and SVR). To evaluate the superiority of either procedure, two hypotheses were developed. In Hypotheses 1, the investigators evaluated medical therapy against medical therapy with CABG. All patients underwent coronary angiography to define the extent of CAD; patients with critical left main disease or unstable coronary syndromes were excluded from the trial. The primary outcome of the study was all-cause mortality, and secondary outcomes were cardiovascular mortality, combination of all-cause mortality and hospitalization for cardiac causes. At a median follow-up of 56 months, medical therapy plus CABG surgery resulted in a nonsignificant trend toward improvement in the primary outcome (36% vs. 41% with medical therapy alone) as well as significantly lower cardiovascular mortality and improved quality of life (at 4, 12, 24, and 36 months as assessed by the Kansas City Cardiomyopathy Questionnaire) [31]. However, this trial was fraught with certain limitations. First, during the study period, 9% patients in medical therapy plus CABG group crossed over to medical therapy group only while 17% patients in medical therapy alone group crossed over to medical therapy and CABG group. This crossover may have led to a diminished treatment benefit, thereby preventing the primary outcome from reaching statistical significance. Second, the STICH trial was designed to maximize both medical and surgical outcomes using strict criteria for surgical expertise (e.g., documented surgical expertise by volume and outcome criteria) and regular review of both surgical center conduct and intensity of medical therapy. Clinical equipoise had to be present, and both the surgeon and cardiologist had to believe revascularization was technically feasible. Both these issues may limit the generalizability of the trial to routine clinical practice.

In 2016, results of extended follow-up of STICH trial patients, i.e., the STICH Extension Study (STICHES), were published extending the median follow-up to 9.8 years [32]. After 9.8 years, the primary outcome (all-cause mortality) was significantly lower in the medical therapy and CABG group compared with medical therapy alone group (59% vs. 66%; hazard ratio [HR] 0.84; 95% CI, 0.73–0.97). Medical therapy and CABG group also experienced significant reductions in cardiovascular mortality (40.5% vs. 49.3%; HR 0.79; 95% CI, 0.66–0.93) and the combination of all-cause mortality and cardiovascular hospitalization (76.6% vs. 87%; HR 0.72; 95% CI, 0.64–0.82). Another large population based observational study related to CAD with LV systolic dysfunction was reported [33], it is recommended to do CABG and medical therapy for patients with ICM who have coronaries amenable to surgical revascularization.

5. Myocardial viability and treatment decisions

Observational studies done in early 2000s focused on the potential benefit of viable myocardium on the patient survival and LV function after the revascularization. Initial potential survival benefit from revascularization in patients with ICM and viable myocardium was reported in a meta-analysis published in 2002. This meta-analysis included 24 nonrandomized viability studies involving 3088 patients with CAD and LV dysfunction who had a mean LVEF of 32% [34]. Patients with myocardial viability had 80% reduction in annual mortality with revascularization (3.2% vs. 16% with medical therapy alone), while there was no significant change in annual mortality with revascularization in patients without myocardial viability (7.7% vs. 6.2% with medical therapy alone). Potential effect of viable myocardium on LVEF was also illustrated in a review published in 2004 that involved 29 observational studies including 758 patients [35]. In this review, LVEF increased after revascularization when myocardial viability was present (37–45%) but did not change significantly in the absence of viability. Further, studies have also demonstrated that 25–30% of the dysfunctional myocardium needs to be viable to result in improvement of LVEF. On the contrary, in a substudy of the STICH trial, 601 of the 1212 patients were evaluated for myocardial viability, and outcomes were analyzed according to those assigned to receive medical therapy plus CABG or medical therapy alone. Study showed minimal improvement in LVEF with revascularization (from 28% pre-CABG to 30% post-CABG). Following adjustment for differences in baseline variables and with follow-up extending beyond 10 years, there was no significant improvement in mortality with medical therapy plus CABG compared with medical therapy alone. Myocardial viability was associated with reduced mortality but did not predict a benefit from revascularization. This raises the question of whether viability assessment is needed prior to surgical revascularization. However, myocardial viability in STICH trial was assessed using stress echocardiography and SPECT radionuclide myocardial perfusion imaging; more contemporary techniques such as CMR and positron emission tomography (PET) were not studied and are an important limitation of the STICH findings [36]. Presence of myocardial viability does lead to improvement in contractility and myocardial thickness following revascularization subject to the presence of at least 25–30% of viable myocardium and scar burden <25% (as detected by LGE-CMR) [37]. However, inconsistencies in the criteria and the methods used to diagnose myocardial viability between various studies have led to blurring of the evidence of benefit of revascularization.

In the absence of firm evidence, routine viability assessment prior to consideration for CABG in patients with ICM is not recommended. However, situations that require greater precision in defining large infarcts either due to associated excessive surgical morbidity (e.g., renal failure) or risk of suboptimal outcome (e.g., evidence of LV remodeling, inability to achieve complete revascularization); viability assessment with more contemporary techniques such as LGE-CMR or FDG-PET may help further refine the potential risks and benefits.

6. Impact of left ventricular size and remodeling

Left ventricular size is an important determinant of outcome after surgical revascularization in patients with ICM. However, our present understanding of impact of preoperative LV size on postoperative LV function and survival is still limited. The impact of left ventricular enlargement on the improvement in LV function after revascularization was illustrated in a review of 61 patients with ischemic heart disease and a mean LVEF of 28%, all of whom had an evidence of substantial myocardial viability [38]. One-third of the patients had no significant improvement in the LVEF (≥5%). The study showed that the patients with a significant improvement in LVEF after CABG had a significantly smaller left ventricular end-systolic volume (LVESV) on preoperative echocardiography than those without improvement (121 mL vs. 153 mL). The observational data are in contrast with the findings from the STICH trial, which found greater benefit with respect to mortality in patients with greater baseline remodeling (e.g., larger left ventricle end-systolic volume index [LVESVI]) [28].

7. Percutaneous coronary intervention versus surgical revascularization

Percutaneous coronary intervention (PCI) is an established treatment for revascularization in acute myocardial infarction. Role of PCI in management of ICM is still unclear due to the lack of well-designed randomized studies. In the lack of randomized controlled study, best available data come from the observational study comparing PCI with CABG in 4616 patients with LVEF ≤35% who were enrolled in New York State registries (1351 underwent PCI with drug eluting stents and 3265 underwent CABG), from which 2126 patients were chosen for evaluation based on propensity score matching [39]. At a median follow-up of 2.9 years, there was no significant difference in mortality between contemporary PCI and CABG (HR 1.01; 95% CI 0.81–1.28). PCI was associated with a greater risk of myocardial infarction (HR 2.16; 95% CI 1.42–3.28) and need for repeat revascularization (HR 2.54; 95% CI 1.88–3.44), but a significantly lower risk of stroke compared with CABG (HR 0.57; 95% CI 0.33–0.97).

In a separate post hoc analysis of AWESOME trial, in which 454 patients who had medically refractory unstable or provocable ischemia were randomized to PCI or CABG. Ninety-four patients had LVEF <35% (mean 25%) [40]. Among patients with LVEF <35%, there was no difference in mortality between CABG and PCI. However, limitation of this trial was that all patients included in the study had angina and acute coronary syndromes and not heart failure.

8. Role of CABG in patients with ischemic cardiomyopathy

The mechanism of survival advantage conferred by CABG in patients with heart failure irrespective of myocardial viability still remains speculative, although, post hoc analysis of STICH trial has been able to shed some interesting insight on this topic. In STICH trial, a subanalysis evaluating cause-specific cardiac mortality in patients with ICM demonstrated that sudden cardiac death (SCD) was the most frequent mode of death and outnumbered pump failure deaths by approximately twofold [41]. Further, both SCD and death from HF were significantly reduced after the CABG (as was death from myocardial infarction). Predictors of increased risk of SCD in this analysis were increased LVESVI and elevated BNP level. Interestingly, same variables along with regional myocardial sympathetic denervation were found to be significant risk factors for SCD in patients with ICM in the Prediction of Arrhythmic Events with Positron Emission Tomography (PAREPET) Study [42, 43]. Thus, the survival benefit of CABG in patients with ICM is largely due to the significant effect of revascularization on reducing the death due to arrhythmia with a smaller contribution from reducing the deaths from pump failure and fatal MI.

9. Our approach to patients with ischemic cardiomyopathy

We suggest the combined CABG and medical therapy instead of medical therapy alone for patients with ICM and CAD that is amenable to surgical revascularization. This suggestion is based primarily on a 7% absolute reduction in overall mortality over 10 years (STICH trial) and superior relief of anginal symptoms following CABG. However, as significant morbidity and early mortality (compared with medical management alone) are associated with CABG in patients with ICM, patients may also reasonably choose medical therapy alone as the initial treatment option. Following initiation of medical therapy, patients should be reevaluated on an ongoing basis for any changes in clinical status or symptoms and consideration for surgical revascularization should be discussed with the patient.

Other clinical features that should be considered while tailoring the decision for any given patient are greater functional capacity (6-minute walk >300 m), greater burden of CAD (e.g., three-vessel disease), coexistent moderate to severe mitral regurgitation (MR), lower ejection fraction (e.g., LVEF <35%), and greater remodeling (e.g., LVESVI >79 mL/m²) (associated with improved outcomes in STICH trial).

Additionally, we do not recommend routine viability assessment prior to consideration for surgical revascularization and consideration should be case-to-case basis especially in patients in whom the risk-to-benefit profile is not as clear (e.g., patients with significantly elevated surgical risk). We believe that viability study may not aid in decision-making; however, the presence of significant viability and < 25–30% scar on LGE-CMR gives reassurance to the surgeon for improved surgical outcome.

Considering the advantage with CABG from the STICHES trial, it seems that patients with suitable targets for revascularization in the setting of an EF < 35% with two or three vessel CAD should be considered for CABG irrespective of the results of viability testing. However, competing risk factors such as severity of heart failure, age of the patient, and risks for noncardiac mortality need to be carefully weighed in considering the recommendation for revascularization and decision should be made on individual basis.

10. Preoperative optimization and perioperative temporary mechanical support

Factors that have been consistently associated with adverse outcomes after CABG for patients with ICM include preoperative renal dysfunction, advanced HF, recent myocardial infarction, and hemodynamic instability. Perioperative shock in this patient population more than doubles the rate of perioperative mortality [44, 45, 46]. Therefore, preoperative optimization of the patient status can improve the patient outcome after the surgery. The specific mode of optimization should be individualized to patients’ needs and driven by their response to initial therapy. If medical therapy alone is ineffective, more invasive measures should be considered. In the preoperative setting, prophylactic intra-aortic balloon pump (IABP) decreases afterload, increases coronary artery perfusion, provides a modest increase in cardiac output [47, 48]. In a variety of analyses, IABP therapy before the operation has been noted to result not only in improved patient condition before CABG, but also in reduced perioperative morbidity and mortality. Two meta-analyses of randomized clinical trials examining the utility of preoperative IABP therapy in patients with ICM demonstrated a strong association between preoperative use of IABP and reduced hospital mortality, lower incidence of low cardiac output syndrome, and shorter duration of ICU stay. Patients with high-risk profile including low LVEF, left main disease >70%, prior heart surgery, poor coronary artery targets, and unstable angina typically benefit from preoperative IABP [47, 48, 49, 50].

In patients who present with cardiogenic shock resulting from acute myocardial infarction or decompensated HF with end-organ dysfunction, IABP may be inadequate for stabilization or preoperative optimization. In these patients, transvalvular devices such as microaxial surgical heart pump can be used. These devices reduce left ventricular end-diastolic pressure (LVEDP) and volume workload and provide the circulatory support necessary to allow native heart recovery. In a recent analysis, the use of these micro-axial pump was associated with reduced mortality, without significant increase in device-related stroke, hemolysis, or limb ischemia [51, 52]. Finally, in patients with cardiogenic shock that is refractory to inotropic support, IABP, and/or microaxial pumps, ventricular assist device (VAD) implantation should be considered [47, 53, 54].

Patients with ICM with cardiogenic shock, who have organ dysfunction at the time of presentation, temporary VAD can be used as bridge to decision. Patients who reverse their organ dysfunction and acidosis after the insertion of temporary MCS and demonstrate an adequate contractile reserve and response to inotropic stimulation can successfully bridge to CABG. This is contingent to good coronary targets and absence of unfavorable anatomic and physiologic profiles [27]. Otherwise, they should be evaluated for heart transplant and should be considered for more durable VAD option as bridge to transplant.

11. Coronary artery bypass graft surgery strategy

11.1 On-pump arrested-heart CABG

The goal of CABG in patients with ICM is to achieve expeditious and complete revascularization. On-pump arrested-heart CABG is the most commonly used strategy that allows a bloodless and still field that facilitates complete revascularization [55]. Excellent myocardial protection especially right ventricle is paramount in the setting of ischemic cardiomyopathy as myocardial ischemia and injury are poorly tolerated when myocardial reserve is limited [56].

In patients undergoing on-pump CABG, controversy still remains about type of cardioplegic solution, temperature, and route of administration that provides the optimal myocardial protection. This becomes critical in patients with ICM as any amount of further myocardial damage may be deleterious. In a meta-analysis of 12 studies including 2866 patients, lower prevalence of perioperative myocardial infarction was found in patients who received blood cardioplegia [57]. Another meta-analysis of 41 randomized clinical trials (RCT) found that warm cardioplegia did not improve clinical outcomes but was associated with a mild reduction of cardiac enzyme release [58]. Single-dose cardioplegia benefit is limited to a reduction in ischemia and bypass time and does not translate into a major morbidity or mortality advantage [59]. There is no systematic comparison of different routes of cardioplegia administration (i.e., antegrade vs. retrograde vs. combined); however, isolated retrograde cardioplegia should be avoided due to its heterogeneous perfusion and unpredictable right ventricle myocardial protection [60]. On the other hand, retrograde cardioplegia may be useful in adjunct to antegrade cardioplegia in patients with severe CAD and in redo CABG to reach territories not otherwise reachable by antegrade delivery and to flush potential embolic debris from inadvertently manipulated diseased vein grafts [61, 62]. Although data are scarce, it has been reported that antegrade cardioplegia supplemented with venous graft perfusion can significantly improve myocardial protection. The most suitable myocardial protection strategy may be a combination of antegrade, retrograde, and delivery down the vein grafts.

11.2 Off-pump CABG

Utilization of off-pump CABG (OPCABG) is limited to few centers and selected patients in the developed countries. There have been no large RCTs comparing on-pump CABG versus OPCABG and small RCTs that did compare these two modalities have reported inferior or non-superior long-term outcome with OPCABG. Most of these studies are limited by smaller sample size, short duration of follow-up, and limited experience of the operator. This is of particular relevance given that OPCABG may lead to inferior long-term outcomes if performed by inexperienced operators and/or accompanied by incomplete revascularization [63]. In a meta-analysis of 23 individual nonrandomized studies published in 2011 that involved 7759 CABG patients with LVEF <40%, 2822 patients underwent OPCABG. Overall early mortality was significantly reduced (odds ratio [OR], 0.64; 95% CI, 0.51–0.81) in OPCABG group. Similar results were observed on subgroup analysis of 1915 patients with LVEF <30% (OR 0.61; 95% CI 0.47–0.80) [64]. A recent meta-analysis published in 2020 comprising 16 studies with 32,354 patients with LV dysfunction (defined as LVEF <40%) also reported a significant reduction in 30-day mortality (OR 0.84; 95% CI 0.73–0.97), perioperative complications, and transfusion requirements with OPCABG [65]. In a report published in 2016 from the Japan Adult Cardiovascular Surgery Database including 918 pairs of propensity-matched CABG patients with LVEF <30%, there was reduced perioperative and 30-day mortality with OPCABG (1.7% vs. 3.7%; P < 0.01) and reduced incidence of mediastinitis, reoperation for bleeding, and need for prolonged ventilation, but there was no difference in incidence of stroke or renal failure compared to on-pump CABG [66].

11.3 On-pump beating-heart CABG

On-pump beating-heart CABG has been proposed as an alternative strategy to on-pump cardioplegic arrest CABG, particularly in higher-risk patients including patients with impaired LV function [67]. This technique is more of historical significance as it is rarely used nowadays. In a review of 11 studies, comprising two RCTs and nine observational studies comparing on-pump beating-heart CABG and on-pump arrested heart CABG, lower mortality was reported with on-pump beating-heart CABG in five of the nine observational studies while mortality was similar with both techniques in two RCTs. However, due to the lack of randomization and the absence of propensity matching, the possibility of selection bias accounting for the difference in mortality cannot be discounted. Intraoperative myocardial injury with on-pump beating heart may increase due to inadequate coronary perfusion distal to areas of stenosis [68].

In the absence of more definitive evidence about the superiority of one technique of CABG over the other, the operative strategy should be tailored based on patient factors such as extent of CAD and associated comorbidities, surgeon’s expertise and comfort level of the cardiac anesthetist, and center experience. When off-pump technique is used, maintenance of appropriate perfusion pressure and when on-pump CABG is utilized, appropriate myocardial protection is imperative to minimize further myocardial injury.

12. Bypass conduits

Presently, use of left internal mammary artery (LIMA) for bypassing left anterior descending coronary artery and reverse saphenous vein grafts for bypassing rest of the coronary arteries is the standard of care across the globe. Evidence from the recent studies has shown the superiority of multi-arterial grafting in improving long-term patient survival after CABG. The impact on survival becomes even more significant with increasing duration of follow-up [69, 70, 71]. The evidence of beneficial effects of multi-arterial grafting in patients with ICM, however, is limited to few studies and a small number of patients [72, 73, 74]. Further, multi-arterial grafting in patients with ICM still remains controversial as the overriding priority in these patients is to mitigate the upfront risk of surgery and avoidance of perioperative myocardial ischemia. In a risk predictive model based on STS database review of patients operated for CABG, the HR for perioperative mortality after isolated CABG was 1.19 (95% CI, 1.17–1.22) for every 10% reduction in LVEF [75], and operative risk was further compounded with the addition of noncardiac organ dysfunction and other comorbidities.

There are four reasons why caution should be used when contemplating multi-arterial grafting in patients with ICM [56]. First, perioperative administration of high doses of vasopressors may be necessary in these patients, and this is an important predisposing factor for the development of spasm in the arterial grafts [76]. Radial and gastroepiploic arteries are particularly vulnerable to spasm compared with IMAs. Second, adequacy of blood flow in a fresh arterial graft may not be as robust as in a vein graft, with the potential for clinically significant perioperative coronary artery hypoperfusion [77, 78, 79]. Third, multi-arterial grafting usually adds to the complexity and length of the operation and prolongs myocardial ischemic time. This may not be well tolerated by the patients with ICM. Fourth, arterial grafts may not be of adequate length in massively dilated hearts, especially if sequential anastomoses are contemplated. A patient-level combined analysis of six RCTs associated radial artery grafts in addition to LIMA with improved clinical outcomes compared with venous grafts [80]. The benefit of radial artery grafting was persistent even on subgroup analysis of patients with severe LV dysfunction (LVEF <35%). However, the number of patients in subgroup were limited (25 (4.7%) and 32 (6.4%) in the radial artery and saphenous vein groups, respectively). The results of other observational studies have yielded mixed results with the use of multi-arterial grafting in patients with ICM [73, 81, 82, 83, 84]. The probable reason is variable cutoff for LVEF with different studies (lowest limit <30%), which adds to uncertainty regarding multi-arterial grafting benefits [85]. Observational evidence also suggests that the benefit of multi-arterial grafting is lost in patients with ICM with limited life expectancy or severe associated comorbidities [83, 86, 87, 88].

We believe that multi-arterial grafting should not be routinely recommended for patients with ICM. Patient selection for multi-arterial grafting should be based on patient factors and surgeon’s experience and comfort. Young patients with compensated HF having good target for bypass may be considered for multi-arterial grafting if the risk–benefit ratio is favorable and prolonged survival is anticipated after revascularization.

13. CABG combined with other procedures

13.1 Atrial fibrillation

Atrial fibrillation (AFib) is present in 5–10% of patients undergoing CABG. It is associated with increased risk of complications including stroke and renal failure, prolonged hospital stay as well as increased mortality despite adjustment for potential confounders [89]. Therefore, current North American and European guidelines for CABG recommend concomitant AFib ablation procedure in symptomatic patients or asymptomatic patients having low operative risk [90, 91]. The evidence supporting the surgical ablation of AFib in patients with ICM undergoing CABG is minimal and limited by selection bias [92]. Theoretically, patients with a reduced ejection fraction would benefit from the restoration of sinus rhythm and atrial contraction [93]. However, concomitant AFib ablation procedure adds to the technical complexity of the surgery and prolongs the duration of aortic cross clamp and cardiopulmonary bypass. Despite this, some studies reported that surgical AFib ablation is safe and effective in patients with heart failure [94, 95].

13.2 Mitral valve surgery

Up to 10% patients develop chronic moderate or severe MR following acute myocardial infarction. Chronic ischemic mitral regurgitation (CIMR) is associated with an increased incidence of heart failure and increased risk of mortality in patients with LV dysfunction [96]. Furthermore, LV dysfunction can lead to gradual dilatation and geometric change in the left ventricle that results in distortion of the mitral valve and worsening of MR. Although, there is a general consensus to repair or replace the mitral valve in patients with severe CIMR undergoing CABG, the management of moderate (Grade II) mitral regurgitation still remains controversial.

In the Cardiothoracic Surgical Trials Network study, adding surgical mitral valve repair to CABG in patients with moderate CIMR had no significant effect on survival or LV reverse remodeling at 2 years follow-up but was associated with increased duration of hospital stay and morbidity including neurological events and atrial arrhythmias [97]. Smaller RCTs have shown benefit in surrogate outcomes for CABG and mitral valve repair versus CABG alone in patients with moderate CIMR [98, 99]. However, none of the trials has specifically focused on patients with ICM. In patients with severe CIMR, mitral valve replacement has been shown to provide more reliable and durable relief of MR than repair, but without survival benefit [100]. Mitral valve replacement rather than repair is also favored in patients with LV basal aneurysm/dyskinesis or other potential risk factors for recurrent MR after repair, e.g., significant leaflet tethering and/or severe left ventricular dilatation (LV end-diastolic dimension >6.5 cm). Preserving the subvalvular apparatus is also strongly recommended when replacing mitral valve in these patients. Concerns about persistent tethering of the posterior leaflet and recurrent MR after CABG in patients with prior inferior wall MI have prompted some to combine mitral anuloplasty with a subvalvular procedure such as papillary muscle approximation and papillary muscle relocation. All these procedures result in improved echocardiographic and cardiovascular outcomes but fail to influence all-cause mortality or quality of life [101, 102, 103]. Therefore, this remains an area for further study and evaluation.

13.3 Tricuspid valve surgery

Tricuspid regurgitation (TR) is an established risk factor in patients undergoing CABG [104]. In patients with CIMR, although progression of unrepaired mild to moderate TR after revascularization is uncommon, presence and progression of moderate or greater TR are associated with increased incidence of clinical events [105]. The underlying etiology of TR in ICM includes tricuspid annular dilatation and leaflet tethering in the setting of RV remodeling due to right ventricle infarction with or without pulmonary hypertension, tricuspid annular dilatation associated with AFib, and iatrogenic or lead related injury to tricuspid leaflets. Current AHA/ACC guidelines assign class I recommendation for tricuspid valve repair at the time of left sided valve surgery for severe TR and class IIa for less than severe TR in the presence of annular dilatation (>4.0 cm) or right-sided HF [106].

Concomitant mitral valve repair can be considered in patients with ICM undergoing CABG in the presence of atrial arrhythmias, left atrial dilation, or in the setting of severe LV dilation. Replacement, rather than repair, should be considered in patients with limited viability in the posterolateral wall of the LV [97]. Tricuspid valve repair should be considered at the time of left sided valve surgery for severe TR and less than severe TR in the presence of annular dilatation (>4.0 cm), right-sided HF or iatrogenic, or lead-related injury to tricuspid leaflets. Severe TR in the presence of significant RV dysfunction is a marker of poor outcome after coronary revascularization and warrants evaluation and consideration for advanced HF therapies.

13.4 Surgical ventricular restoration

In patients with ICM, gradual dilatation of LV results in transition from elliptical to a more spherical geometry. This impairs the structure–function relationship of the left ventricle [107]. The concept of surgical ventricular restoration (SVR) procedure for the patients with ICM is more than four decades old; however, the procedure is yet to gain acceptance as not only the procedure is technically challenging but also, no study so far has been able to show consistent benefit with concomitant SVR. Doctrine of SVR operation assumes that resection of scarred myocardium, reducing the ventricular size, and restoring an anatomically elliptical shape can improve the left ventricular function [108]. However, studies so far have not been able to prove this assumption. A randomized study including 137 patients with LVEF <50% and LV end systolic volume index (LVESVI) >80 ml/m² showed that CABG alone was inferior to CABG with SVR in terms of improvement in LVEF, MR, and NYHA class. However, study was limited to only 2 years of follow-up [109]. Similarly, Prucz et al. reported this result [110]. Both these studies were limited by short duration of follow-up and failed to show any benefit of SVR procedure on survival. Consequently, the STICH trial was conducted to evaluate the long-term outcome of concomitant SVR procedure in patients with LV dysfunction, LV akinesis/dyskinesis, presence of scar, and LV dilatation [111]. To evaluate the benefit of SVR, patients enrolled in STICH trial in CABG arm were divided into two groups (medical therapy with CABG versus medical therapy with CABG and SVR). The study found no difference in mortality between the groups at median follow-up of 48 months (hazard ratio 1.00, 95% CI 0.79–1.26, P = 0.98) [111]. Results of these studies led to abandonment of the SVR procedure by majority surgeons [112].

It still remains uncertain which patients should receive SVR as part of CABG operation and what is its impact on long-term survival and functional outcome [112, 113, 114]. Therefore, consideration for SVR should still be given to patients with true large ventricular aneurysms who present with medically refractory heart failure or ventricular arrhythmias.

14. Postcardiotomy shock and temporary MCS

Patients with ICM undergoing CABG are at increased risk of postcardiotomy shock and the risk increases further in patients with ischemic MR and/or right ventricular infarct. Patients with postcardiotomy shock who are unable to separate from cardiopulmonary bypass or require high-dose inotropic therapy, MCS should be considered [115].

14.1 Intra-aortic balloon pump (IABP)

Intra-aortic balloon pump has been considered as first line therapy for PCS as it is safe, widely available, and easy to place. Intra-aortic balloon pump improves the coronary perfusion, decreases the left ventricular afterload, and improves the cardiac output by 0.5–1 L/min. However, the hemodynamic support provided by an IABP is usually insufficient in reversing cardiogenic shock [116, 117]. In a recent analysis of 4550 patients operated for CABG between 2004 and 2008, 5% patients required an intraoperative or postoperative IABP, with overall mortality of 37%. IABP was equally effective in patients with predominantly right-sided failure with 50% increase in cardiac index and associated mortality of 31%. This study specifically addressed the issue of IABP effectiveness in both right- and left-sided failure [118].

14.2 Impella

Impella is a percutaneous or surgically implanted axial-flow device that is used for all types of cardiogenic shock. Impella devices significantly reduce LV end-diastolic pressure and volume, reduce myocardial oxygen demand, and support the systemic perfusion while allowing the heart to recover. Engstrom and colleagues [119] reported their experience with Impella 5.0 for treating 46 postcardiotomy shock patients mostly after CABG at three European centers. Half of the patients received an IABP before the Impella placement. Overall survival was 40% at 30 days. More recently, David and colleagues [120] reported on use of the Impella 5.0/Impella LD in 29 patients (40% with isolated CABG) treated for PCS between 2010 and 2015. Mortality was nearly 40%, similar to the aforementioned study. The best results for PCS treatment were reported by Griffith and colleagues [121] in the RECOVER I study, wherein an Impella 5.0 was placed in 16 patients having difficulty weaning from cardiopulmonary bypass. Fifteen patients were successfully supported, with 30-day survival of 94%. Results of this study should however be interpreted carefully as all the patients in the study were on low level of inotropic support before the Impella placement as opposed to the study protocol requirement of high inotropic support prior to Impella placement.

14.3 Extracorporeal membrane oxygenation

Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) is second most commonly used device after IABP for postcardiotomy shock. Veno-arterial ECMO significantly unloads the right ventricle, improves the coronary perfusion, and supports the systemic perfusion while allowing the right heart to recover. However, VA-ECMO significantly increases the left ventricular afterload. Therefore, in patients supported with VA-ECMO, it is imperative to maintain left ventricular ejection either spontaneous or with inotropes. Otherwise, left side of the heart should be vented either by atrial septostomy, left atrial/left ventricular vent, or Impella [122]. There are no RCTs regarding the effectiveness of VA-ECMO in PCS, but several retrospective studies have shown 60–70% mortality in patients with PCS despite use of VA-ECMO [122, 123, 124, 125]. In a recent report of the European registry of 781 patients receiving VA-ECMO for PCS, institution of VA-ECMO was associated with increased mortality (odds ratio 1.54; 95% CI, 1.09–2.18), reoperation for bleeding/tamponade (odds ratio, 1.96; 95% CI, 1.37–2.81), and blood transfusion of >9 units (odds ratio, 2.42; 95% CI, 1.59–3.67). The authors also did a systematic review of 2491 patients with PCS who received VA-ECMO and reported 66.6% pooled prevalence of in-hospital/30-day mortality (95% CI, 64.7–68.4%), and lower in-hospital/30-day mortality in patients with peripheral ECMO (risk ratio, 0.92; 95% CI, 0.87–0.98). Switching the patients from central to peripheral cannulation appeared to provide close to a 10% mortality benefit [126]. Finally, studies evaluating the role of LV unloading during VA-ECMO for cardiogenic shock have reported 10–20% mortality benefit with LV unloading with either Impella or IABP [127, 128].

15. Post discharge management

In patients with ICM, the importance of adhering to guideline-directed medical therapy (GDMT), secondary prevention, and cardiac rehabilitation after revascularization cannot be overemphasized [129, 130]. Close follow-up of these patients is recommended for the titration of heart failure medications and continued assessment for needed additional interventions, including device implantation (e.g., automated implantable cardioverter-defibrillator (AICD)/Cardiac resynchronization therapy device (CRT) or advanced surgical therapies for persistent HF. In patients with ICM, initial 90 days after CABG are most vulnerable and associated with several-fold increase in HF-associated rehospitalization and mortality. Thus, these patients should undergo a close clinical monitoring after discharge. Initial post-discharge follow-up should be done at 7–14 days to review the volume status of the patient and titrate guideline-directed medications [131]. Although studies directly evaluating and comparing the impact of GDMT on ICM patients who have or have not undergone CABG are limited, conventional medical opinion supports that GDMT goals for post-CABG patients should not differ from those without CABG. Post hoc analysis has revealed that in patients with ICM, maintenance of optimal medical therapy after discharge is associated with best short-term and long-term outcomes [132].

16. Summary

Patients with ischemic cardiomyopathy and coronary artery disease that is amenable to surgical revascularization should undergo combination of surgical revascularization and medical therapy rather than medical therapy alone. This suggestion is based primarily on the long-term absolute reduction in mortality over the 10 years following CABG balanced against the early mortality risk of CABG. Routine assessment of viability to evaluate advisability of multivessel coronary revascularization to improve total mortality is not recommended. Based on the small but nontrivial early mortality risk associated with CABG surgery as well as other post-CABG morbidities, patients may also reasonably choose medical therapy as the initial treatment option. Revascularization remains an important treatment option for patients with ongoing anginal symptoms despite optimal medical therapy. For such patients, the relative efficacy of percutaneous coronary intervention (PCI) compared with CABG for revascularization is unknown. Nonrandomized registry suggests that there was no difference in mortality between CABG and PCI.