|Home | About | Journals | Submit | Contact Us | Français|
The advent of percutaneous intervention has made surgical treatment of coronary artery disease less favored by patients though the evidence that supports CABG in certain patient subsets is strong.
Literature review was done using Pubmed, Scopus, Google and Google Scholar with MeSH terms-coronary artery bypass grafting, internal mammary artery, drug eluting stent, stroke, myocardial revascularization.
The adoption of evolving techniques like anaortic off pump grafting, bilateral internal mammary artery use, hybrid and minimally invasive coronary revascularization techniques, intra-operative graft assessment, and heart team approach can lead to better outcomes following surgery as is evidenced by recent literature.
Though the adoptability of the newer strategies may vary between centers a close coalition between coronary surgeons and cardiologists would ensure that the management of coronary artery disease is based on evidence for the benefit of the patient.
Coronary artery disease (CAD) is predicted to be the cause for 14.2% deaths by 2030.1 The advances made in the percutaneous treatment of CAD and the advent of various generations of stents has broadened the indications of PCI in recent times. However there is overwhelming evidence in the literature comparing CABG with medical therapy and variations of PCI in specific patient subsets showing the superiority of CABG in terms of the long term survival benefits and repeat revascularization rates.2,3 The inherent invasiveness of conventional CABG with the attendant risks of peri-operative stroke and morbidity have proved decisive in the decline of its popularity in the present era of absorbable stents.
Evolving surgical strategies likely aortic no touch technique/total arterial coronary revascularization [TACR], bilateral internal mammary artery (BIMA) use, multiple arterial grafting (MAG), minimally invasive CABG, hybrid coronary revascularization (HCR) techniques and intra-operative graft assessments have to routinely become part of coronary surgeon's armamentarium to stay competitive and take the practice to the next level. At the same time, the emphasis on close coalition with the core team of cardiologists with input from the family physicians who have the first level contact with the patient and the patient himself/herself should guide the decision making process. Hence the approach to treating CAD with a “Heart Team (HT)” concept seems to be the need of the hour. This review discusses the evidence based, adoptable strategies for the coronary surgeon to remain competitive yet co-operative and potentially alter the way CAD is treated for better patient outcomes.
Although the use of single internal mammary artery (SIMA) has become part of a standard practice protocol in CABG, the routine use of BIMA has found a patronage of only 4–12% as per the statistics based on Society of Thoracic Surgeons Database.4 There are consistent reports in the literature quoting the superiority of BIMA over SIMA grafting. The advantage of two arterial grafts supplying vital areas of myocardium afford better protection from long term adverse events especially considering the inferior patency rates of saphenous venous grafts (SVG) for the same territories. Weiss et al in a recent meta-analysis have documented that the long term survival of BIMA group was significantly higher than SIMA group.5
As observed by Lytle et al in a previous landmark study, the incremental benefit of using BIMA for the first 12 years in the younger subset of patients would be in its value in the prevention of repeat revascularizations and the real survival benefit of BIMA vs SIMA use would be apparent beyond this time-frame.6 In a follow up study by the same author, it was noted that the general survival benefit associated with BIMA grafting extended for twenty post-operative years. As expected from the pilot study the increasing advantage of a BIMA graft became apparent only in the second decade after surgery.7
Concerns about deep sternal wound infections [DSWI] have limited BIMA use by surgeons. The reported incidence of DSWI with BIMA use vary from 0.3% to as high as 16% in different series. The risk factors associated with DSWI are female sex, obesity, and diabetes.8 The use of skeletonized harvest technique for BIMA with the proposed advantage of the preservation of sternal blood supply has clearly shown reduction in the incidence of DSWI by some groups.9 However the same reduction in the incidence of DSWI could not be replicated with a pedicled BIMA harvest technique.
Thus though the retrospective and other observational studies have registered a long term survival benefit for BIMA grafting, a randomized prospective trial to validate this is the need of the hour. The ART trial was designed to this end.
The ART, a multi-centric prospective trial comparing the long-term survival benefits of BIMA versus SIMA grafting is now underway involving 3102 patients either assigned to BIMA or SIMA arms. As expected the short term outcomes of BIMA and SIMA grafting with respect to mortality appear to be comparable in this trial. In concordance with the previous studies BIMA and SIMA groups were comparable in the other short term outcomes like, major complications, myocardial infarctions, stroke, and repeat revascularization rates. However there was an increased incidence of DSWI and sternal wound reconstructions in BIMA group [1.9% (BIMA) vs 0.6% (SIMA)].10
The long term results of ART is expected to resolve the ongoing debate on the choice of BIMA vs SIMA and the attendant risks.
One of the major limitations of CABG is the higher incidence of stroke peri-operatively and during early follow up when compared to PCI in similar patient subsets. Peri-operative stroke hikes the risk for peri-operative mortality, extends the hospital stay and adds to the economic burden. In SYNTAX trial the higher incidence of peri-operative stroke was found to even out with the PCI group at the end of 2 years. By the end of 5 years, the incidence of stroke in PCI group was 2.4% as against 3.7% in the CABG group which was not statistically significant. In the early SYNTAX follow up data the incidence of stroke was higher in the CABG group especially in patients with a higher SYNTAX score. This difference evened out at mid-term follow up.2 Despite the fact that most of the peri-operative strokes documented in the SYNTAX trial occurred in the waiting period for surgery and might have contributed to the higher early stroke rates in the CABG cohort, another well planned randomized trial (FREEDOM)3 also drew similar conclusions on this, with a higher stroke rate in the peri-operative period. At 5 years, while the CABG group was still trending towards a higher incidence of stroke (5.4% as against 2.4% in the PCI group). This trial was done on patients with multi-vessel disease (increased disease burden) and diabetes mellitus [DM] (independent risk factor for stroke) and returned similar results as in SYNTAX. The results with regard to the incidence of stroke after CABG were similar in the ASCERT study also with a higher early stroke rate in the CABG arm and a decreasing incidence of the complication with longer duration of follow up.11 These studies thus expose the disadvantage of CABG in the incidence of early and peri-operative period.
Etiology of peri-operative stroke after CABG is multifactorial. Embolism and the water shed infarcts resulting from the low perfusion pressures are implicated frequently. Roach et al in a large multicenter study had classified the neurological outcomes after elective coronary artery bypass and found that the Type 1 group which included the peri-operative stroke subsets to be strongly associated with ascending aortic atherosclerosis.12 A prospective study spanning 30 years involving various modalities of surgical revascularization reported the risk of intra-operative stroke to be higher with increasing atherosclerotic disease burden.13
The avoidance of aortic manipulation has been thus proposed to decrease the incidence of peri-operative stroke. Although off pump CABG (OPCAB) has been found to have lower incidence of stroke rate when compared to the conventional CABG, its utility may be limited if partial clamping of aorta is practiced. Wolf et al had studied the incidence of solid micro-emboli by transcranial Doppler in patients undergoing CABG with or without aortic clamping. The reduction in the number of solid micro-emboli was significant in the clampless group. They recommended a CABG with no touch aortic technique to minimize the peri-operative neurological events.14
Borgermann et al compared clampless OPCAB (anaortic/aortic no touch OPCAB) with conventional CABG by a propensity matched analysis and reported lower incidence of stroke and mortality with the clampless off pump technique.15 In a subgroup analysis of SYNTAX trial the incidence of stroke in the conventional CABG arm at 3 years follow up was compared with that of anaortic OPCAB arm. The percentage of stroke was trending towards a lower but non-significant value with the aortic no touch technique (1.3% in the no touch group vs 3.4% in the SYNTAX group). Moreover, the no-touch group results compared favorably with the stroke percentage in the PCI arm of SYNTAX trial, thus making the technique as safe as PCI with respect to the post-operative stroke rate. (1.3% (no-touch) vs 2%(PCI arm).16
Incidence of peri-operative strokes in anaortic OPCAB vs conventional CABG was studied by Misfeld et al. This was one of the largest studies comparing the aortic no touch technique with the conventional CABG. The investigators performed either anaortic no touch total arterial coronary revascularization (TACR) (using in situ arterial grafts) or clampless proximal anastomosis for SVG to the ascending aorta with the Heart String (HS) device in the anaortic OPCAB group. The stroke rates were significantly lower for the TACR/HS OPCAB group when compared to the conventional CABG group.17
On the other hand Emmert et al reported that the incidence of peri-operative stroke remained the same regardless of off-pump or on pump grafting techniques if partial aortic clamping was practiced. The stroke rate associated with the HS device in this study was lower than that of conventional CABG (0.7% vs 2.3%). Also, stroke rates achieved by utilizing either HS or no touch TACR were reported to be similar in the study.18
Thus, it may be safely inferenced that in patients with higher risk of post-operative neurological complications the technique of anaortic OPCAB utilizing TACR or the proximal anastomotic systems should be the practice recommendation for surgical coronary revascularization. This could normalize the gap between the CABG and PCI in terms of both in-hospital and early post-operative stroke rates.
SVG are the most common conduits used for CABG. Available literature reports their poor patency rate of 50% at 10 year follow up and about three quarters get occluded at 15 years.19 The inferior patency rates of SVG harvested via the conventional techniques have prompted the surgeons to look for alternate arterial conduits with better patency rates. As already stated BIMA grafting has demonstrated superiority over SIMA in angina relief and a survival benefit.5 In recent times, the addition of a third or multiple arterial grafts to the BIMA has become a routine practice in many centres performing CABG. A considerable amount of evidence has been generated off late reporting better long term outcomes in patients receiving multiple arterial grafts. Technique wise the left coronary (for utilization of arterial grafts) is given priority over the right system unless the right coronary vessels are tightly stenosed to prevent competitive flow related occlusion or stringing. Use of both the internal mammary arteries on to the left system (BIMA grafting) would require a third arterial graft onto the right system if TACR is planned. The most common conduits used for the right system are right internal mammary artery (RIMA), radial artery (RA) and right gastroepiploic artery (RGEA).20,21
RIMA use has been widespread and current evidence suggests that it works best when anastomosed onto the left coronary system-LAD and its branches or the circumflex system. Although the reported patency of RIMA is comparable to LIMA when used for grafting LAD,22 the same results could not be replicated when used as a conduit for the right system. Previous evidence has shown that grafting RIMA onto the right coronary system returned inferior patency rates as compared to grafting onto the left system and was prone to stringing if the stenosis was <90%.23
This could have been due to the deployment of the conduit onto the distal right coronary where it is easier to graft the vessel with an in situ RIMA for want of length. The high propensity of the development of the right coronary crux disease as it is a branching point24 and the competitive flow pattern could have resulted in the inferior patency rates. Hence it is wiser to use RIMA on the left side of the heart and preferably distally onto the PDA when used on the right coronary system to decrease the possibility of graft failure. On the left side of heart, in situ RIMA can easily reach a ramus intermedius or the high obtuse marginals. It can be positioned behind the thymus and anterior to aorta to reach this position. To afford more length it can be tunneled behind the ascending aorta with the obvious risks of rotation and non-accessibility in case of bleeding. RIMA is best used as a free graft if mid and distal territories are being grafted for want of adequate length.25
RA was first used for CABG by Carpentier et al as early as in 1974,26 but later was abandoned due to adverse results attributed to narrowing and spasm. The interest in RA use was revived by Acar et al in the nineties.27
RA as a conduit does not match with IMA as the atherosclerotic changes occur 5 times more in it when compared to the IMA grafts-(5.3% vs 0.7%).28 Still it fares better than the SVG in patency rates. The RVSP trial assessed the mid-term patency of RA grafts over SVG and found it to be significantly higher (98.3% vs 86.4% at 5 years).29 The RAPCO trial in contrast to these studies reported a similar midterm patency rates for RA grafts and SVG (90% vs 87%).30 Tranbaugh and colleagues recently analyzed the long term outcomes of RA grafting vs SVG with 14 year follow up. They found that the SVG group had 52% higher mortality than the RA group irrespective of the territory grafted. Particularly, RA grafting added survival benefit to younger men.31
RA can be harvested either using an open technique or with the help of an endoscope. It can be used as a composite graft or a direct aorto-coronary graft with no difference in the long term patency. The sequential use of RA for multiple target grafting also can be done on pump or off pump with superior patency rates than a LIMA–SVG sequential graft.
The major drawback of RA is its vasospastic tendency due to a thick medial layer with tightly packed smooth cells.28 The vasospastic tendency of RA can be counteracted with the use of topical agents like papaverine, phenoxybenzamine and verapamil with nitroglycerine which are commonly used in clinical practice. Recently fasudil a rho kinase inhibitor has been found to be useful intra-radially to relieve its spasm.32
In short, the current evidence states that SVG may compare well with RA in the patency rates during short term follow up, but definitely RA scores better in the long term. Also, RA use should be planned on the right coronary system only if there is high grade stenosis of the target vessel which supplies a large area of viable myocardium33 and in younger men where they yield better results.
RGEA is the least used arterial conduit among the three. Although it has shown superior patency rates compared to SVG in some studies, its use has never become popular. Suzuki et al found better outcomes with RGEA use when compared with SVG on to the RCA especially when a skeletonised technique of harvest was utilized.34
In contrast, De Mauro et al in a long term follow up analysis criticized the use of RGEA on the right coronary system quoting an increased rate of adverse events at the end of 8 years. They advocated the use of supplementary venous grafting rather than RGEA onto the right coronary system when BIMA grafting is done.35
The short length of RGEA generally permits its use only as an in situ pedicled graft onto the right coronary system. SVG when grafted is functionally a first or second branch of aorta in contrast to the RGEA which originates as the fourth branch of the visceral arterial tree. The resultant difference in the flow dynamics could be the reason for the inferior patency rates and its unpopularity.20
The utility of intra-operative graft flow assessment is useful in predicting the technical perfection of the anastomosis objectively. The incidence of intra-operative graft flow failure has been studied and found to average about 5% for IMA and 11% for the SVG. The anastomotic problems or the limited distal flow in the native coronaries could produce flow compromise in the grafts.36 Although coronary angiography is considered the gold standard for assessing the graft patency, the logistic difficulties make its use impractical peri-operatively. Various other innovations namely transit time flow measurement (TTFM), thermal angiography, sonography based modalities (Doppler velocity measurements, epicardial colour Doppler) and intra-operative fluorescence imaging (IFI) have come in vogue.
The earliest attempts to quantify the graft functionality were the use of myocardial pH electrodes to assess metabolically the regional perfusion.37 During the same period Mohr et al popularized the intra-operative thermal angiography technique for graft assessment and reported its utility.38
The most common modalities used in the current era to this end are TTFM and IFI. The graft patency ideally should be confirmed based on its anatomical and functional imaging. TTFM is an excellent tool to assess the functionality of the graft. A low pulsatility index with high flow rate by TTFM confirms graft patency. A pulsatility index of >5 usually indicates graft occlusion by TTFM. But the incidence of false negatives-grafts with poor flow and low pulsatility index limits its use.39 Giammarco et al recently reported the utility of a combined approach using TTFM and epicardial ultrasound. The addition of the high resolution epicardial ultrasound to TTFM was shown to eliminate the false negatives and comprehensively assesses the graft patency both morphologically and functionally yielding an accuracy of 100%.40
IFI angiography is a semi-quantitative tool for the assessment of intra-operative graft patency. It uses laser activated indocyanine green dye for the assessment of the grafts. Its limitation is that it does not allow the exact flow through the graft to be quantified and thus is only useful in assessing the morphology of the grafts. Also, the laser beam used for the quantification has a penetration of only 1 mm limiting its utility in certain situations where the anastomosis is in the depth.41 IFI delineates the morphology of grafts clearly but the functional quantification of the graft patency is indirectly measured based on the dye wash out delay in terms of heart beats (HB) (≤15 HB-good flow, >15 HB-low flow, dye pooling in the graft-occlusion).42
The routine use of intra-operative graft assessment provides early identification of failed grafts, buys lead time for the surgeon to adopt corrective measures and consequently decreases the rates of early graft failure.
Patients are often disillusioned by the invasiveness of on/OPCAB and prefer to have a percutaneous procedure if the option is available, although the phenomenon of temporal discounting is operational-the emphasis on the short term benefits of PCI overlooking the full range of benefits of CABG.2 The recent advances in the minimally invasive cardiac surgery (MICS) with the improved retractors and modified instruments have made these options viable for the surgeons and patients. The 2011 ACC/AHA guidelines predict that in the future CABG will need to evolve into less invasive techniques for elective coronary revascularization.43
Minimally invasive direct coronary artery bypass (MIDCAB) avoids sternotomy and its attendant complications. Despite the increase in post-operative pain due to the amount of chest spreading involved in the left anterior thoracotomy approach, it has added benefits like earlier discharge, cosmetic superiority, improved patient satisfaction and better quality of life.
Holzhey et al reported long term follow up data of MIDCAB procedures which were mostly single anastomoses to the anterior left coronary system in a low Euroscore cohort. They reported survival rates of 88.3% and 76.6% at 5 and 10 years of follow up respectively.44
As MIDCAB offers a limited exposure of mostly the anterior surface of the heart its utility is mostly restricted to grafting LIMA- LAD. But with the advanced minimally invasive techniques there are centers performing multi-vessel grafting even in the lateral and inferior surfaces of the heart employing MIDCAB techniques.45 The long term results of multi-vessel MIDCABS are still awaited.
Recently BIMA grafting employing MIDCAB techniques have been proposed. Nambiar et al had proposed a technique of harvesting BIMA via a left anterior thoracotomy without the use of thoracoscope or robot under direct vision. TACR was achieved by fashioning a LIMA–RIMA Y-composite conduit.46 The feasibility of combining BIMA use into MIDCAB procedures thus achieving MAG can give an added edge to the coronary surgeon.
Totally endoscopic coronary artery bypass (TECAB) and Robot assisted CABG are other lesser invasive approaches. Though the initial results of TECAB were discouraging, the recent reports from highly experienced centers show comparable mortality rates of 1–2% as compared with the conventional procedures.47 The high degree of technical expertise involved, the infrastructure and equipment costs coupled with the steep learning curve limit the widespread applicability of these techniques.
Hybrid coronary revascularization (HCR) is yet another avenue which facilitates the co-operation between cardiologists and cardiac surgeons. There is strong evidence supporting the survival benefit and symptom relief provided by a LIMA- LAD anastomosis in the literature.4 Also there are reports claiming the superiority of drug eluting stents (DES) over the SVG at least in the short term when used for treating non-LAD lesions.48
The basic indications for HCR is a suitable coronary anatomy where we can blend the two modalities (PCI and CABG) conferring the benefit of a LIMA–LAD anastomosis to the critically stenosed LAD, performed through a minimally invasive approach and the advantages of a percutaneous DES to the significant non-LAD targets. Short term results of this approach have been satisfactory with survival rates reported with HCR similar to those of traditional CABG. The relative contraindications for HCR include poor LAD target, hemodynamic instability, redo surgery, pulmonary disease with contraindication to single-lung ventilation, morbid obesity and non-LAD disease not amenable to PCI.49 Randomized trials to confirm the results from the available retrospective data on HCR are overdue as it presents a durable and less invasive treatment option for multi-vessel CAD.
In the Indian context, the number of PCI done in acute coronary syndrome (ACS) is rapidly increasing. Patients having culprit vessel PCI to non-LAD targets may have a longer term benefit if their LAD lesions are later treated with a LIMA–LAD anastomosis using a left thoracotomy minimally invasive approach. The cost implications of HCR may still break even with multi-vessel PCI considering the repeat revascularizations,2,3,11 the accompanying hospitalizations and follow up visits associated with PCI. This might make HCR an economically viable option in India and other developing countries.
Although endoscopic vein harvesting [EVH] has the purported advantages of early recovery, lesser post-operative pain and reduced incidence of harvest site complications, the data from PREVENT IV trials reported inferior patency rates when this technique was employed with an associated increase in mortality.50
A large observational study comparing EVH vs open venous harvest technique recently has reported comparable outcomes with regard to mortality rates, incidence of major adverse cardiac or cerebrovascular events and revascularization rates. The harvest site complications were reported to be lower in the EVH group. This has resulted in a paradigm shift for reconsideration of the technique.51
The concept of no touch vein harvest technique i.e. harvesting the saphenous vein with an intact periadventitial tissue was introduced on the premise that it produces lesser intimal damage. Recent reports suggest that the adherence to the no touch technique of vein harvest resulted in better graft patency rates angiographically.52
The value of these observations lies in the utility of SVG in supplementary grafting-i.e. the deployment of the arterial grafts on the left coronary system where they yield better long term results, can leave right coronary system potentially amenable for SVG use. SVG has been reported to have similar patency rates when compared with RA grafts when used on the right coronary branches without high grade stenosis at 8 years follow up.53
These techniques of SVG harvest along with the concept of supplementary venous grafting may revive the interest in SVG use although, further trials may reveal their true value.
This concept has proven its viability with the advent of Trans-aortic valve replacement [TAVR]. The efficiency of this team approach demonstrates that in the management of CAD which demands multipronged treatment options a synergistic effect as evidenced in HCR may be useful. The multidisciplinary heart team should comprise of clinical/interventional cardiologist, cardiovascular surgeons, anesthetists and the primary care physicians interacting cohesively to tackle the decision making dilemma and provide comprehensive care in CAD. The importance of the Heart Team (HT) Approach has been stressed in SYNTAX trial and it should be integrated into all coronary care facilities.2 A proposed algorithm for management of CAD with the “HT approach” in Indian scenario is shown in Fig. 1.
Anatomical SYNTAX score has become a major tool for the HT in assessing the risk for MACCE and can guide management decisions. However the burden of comorbidities like age, COPD, peripheral vascular disease, DM, smoking and poor general health do play a significant role in the decision making on the surgical candidacy. The advent of risk scoring systems like SYNTAX II and functional SYNTAX afford to devise a more individualized treatment plan and have furthered the risk assessment ability of the HT by combining both the anatomical, functional and comorbid factors for predicting the risk of intervention. Although SYNTAX II does not take into account DM as a predictor of peri-operative risk, it is an important tool in assisting the choice of intervention. Moreover SYNTAX II has been validated in the DELTA registry involving Asian population which can augment its applicability in the Indian scenario.54 Though fractional flow reserve (FFR) based calculation of SYNTAX score (culprit vessel scoring) i.e. the functional SYNTAX has been proposed after the FAME trial, the feasibility of routinely incorporating it into clinical practice in a developing country may be limited by cost constraints.55
The adoptability of the surgical strategies outlined here vary between surgeons, centers and target populations. However their utility in ensuring better patient outcomes has been reasonably well established. The weak spots in the practice of CABG could potentially be rectified by the adoption of these strategies although they often involve unlearning and relearning. The evolution of catheter based techniques in the treatment of CAD should be integrated with enthusiasm by the surgical community as it would provide the necessary impetus in the proliferation of HCR and other minimally invasive techniques with proportionally better patient outcomes. The practice of cardiovascular surgery and cardiology pertaining to the treatment of CAD should thus be competitive at all levels at the same time mutually complementary. To conclude-“If you want to be incrementally better: Be competitive. If you want to be exponentially better: Be co-operative.”
The authors have none to declare.