Keywords

FormalPara High Yield Facts

  • Coronary artery bypass grafting (CABG) has become one of the mainstays in the treatment of ischemic heart disease.

  • Both arterial and venous conduits and a variety of their combination can be used in CABG.

  • Arterial conduits available are left internal thoracic artery (LITA), right internal thoracic artery (RITA), radial artery (RA), and right gastroepiploic artery (RGEA).

  • Pedicled or skeletonized harvesting techniques have been described for arterial conduits. Skeletonization of ITAs has been suggested to reduce the risk of sternal wound complications.

  • Minimally invasive endoscopic harvesting can be applied to RA and long saphenous vein.

  • Use of a “no-touch” technique for saphenous vein grafts (SVG) preserving the tissue surrounding the vessel (no-touch technique) has been suggested to improve patency rate by minimizing vascular trauma but can increase risk of wound infection and bleeding.

  • Biological augmentation or physical reinforcement of SVG has been suggested but long term results are awaited.

  • Composite arteriovenous conduit obtained by anastomosing SVG to LITA to graft non-left anterior descending coronary (LAD) territories has been shown to have satisfactory short and mid-term outcomes, but more confirmatory evidence is required.

  • Arterial grafting, in particular LITA, to the LAD is well established on the basis of its prognostic benefit.

  • The choice of the second and further conduits for the remaining targets remains debatable, but on the basis of the evidence currently available and on the basis of the long-term results of previous and new randomized controlled trials (RCTs), a multiple arterial grafting strategy should be the preferred approach in the majority of the patients.

Introduction

Coronary artery bypass grafting (CABG) has progressively become one of the mainstays for the treatment of ischemic heart disease since its first introduction in 1967 by Rene Favaloro. Surgical myocardial revascularization can be achieved using different conduits, namely left internal thoracic artery (LITA), right internal thoracic artery (RITA), radial artery (RA), right gastroepiploic artery (RGEA) and saphenous vein (SV). A variety of their combinations have been described.

This chapter will individually examine the characteristics of each of the conduits and the results obtained over the course of 50 years of CABG experience.

Internal Thoracic Artery

The LITA originates from the subclavian artery and supplies the sternum and the breasts before continuing after its bifurcation as the superior epigastric and musculophrenic arteries.

From the histological standpoint, LITA is characterized by discontinuous internal elastic lamina and a thin media enriched by multiple elastic laminae in the absence of a significant muscular component. These characteristics, along with the biological findings of an increased production of vasoactive molecules and antiinflammatory cytokines [1], are thought to be at the root of the largely superior patency rate of this conduit when compared to the others. In fact, the reduced tendency for spasm due to the lack of significant muscle layer, and the relative protection from the development of atherosclerosis guaranteed by its biological properties, could explain the superior outcomes reported for this conduit [2].

The LITA can be harvested in a skeletonized or pedicled fashion, with the first technique being thought to carry the advantage of increasing the effective length of the conduit available for grafting (normal length ranging from 14.32 to 19.48 cm) [3] and of avoiding extensive devascularization of the sternum, with potential implications for the sternotomy healing process. In this regard, skeletonization of ITA has been reported to reduce the risk of deep sternal wound infections also in diabetic patients, normally more prone to this type of complications [4, 5].

More than 30 years after the pioneering work at the Cleveland Clinic on arterial CABG [6], the prognostic benefit of LITA grafting on the left anterior descending (LAD) artery and its superiority in comparison to other conduits is well established. Large studies on the long-term outcomes of LITA have shown more than 90% patency rate at 20 years [2, 7] using this conduit on LAD. These results were confirmed when LITA was anastomosed to circumflex artery (patency of 97% at 5 years, 92% at 8 years, 89% at 10 years and 91% at 15 years), whilst its patency was reduced to 84% at 15 years when grafting the right coronary artery [7]. The reasons of the performance discrepancy seen when LITA is anastomosed on the left or right system are probably due to the size mismatch of the vessels and the different rate of disease progression at the crux [8]. However, these results still outperform the patency rate of saphenous vein grafts (SVG), which dramatically falls below 75% at 10 years follow-up [9]. The right ITA demonstrated a patency rate spanning around 90% on the left system and between 80–90% on the right system at 10 years [10]. The histomorphological differences demonstrated between the right and the left ITA partly might explain this discrepancy in outcomes.

The Radial Artery

The radial artery (RA) arises from the bifurcation of the brachial artery in the antecubital fossa. It runs distally on the anterior part of the forearm ending in the deep palmar arch, where it joins with the deep branch of the ulnar artery. In the preoperative assessment, the anastomosis among the radial and ulnar system is assessed by the Allen’s test which proves the collateralization of the two systems and therefore the functional compensation of the ulnar artery if the radial artery is harvested as conduit for CABG. However, this test has been shown to be unreliable and more objective methods of ulnar compensation are needed, such as Doppler ultrasound and percutaneous oximetry [11, 12].

Histologically, the RA is characterized by a well-developed muscular profile which had initially cast some shadows on its use in CABG because of concerns regarding vasospasm once implanted in the coronary system. This issues conspired against RA use for a prolonged period of time since its initial introduction in the 70s [13], before being reintroduced in the early 90s [14]. For this reason calcium channel blockers with or without long-acting nitrates have been widely used in the postoperative management of these patients [15]. However, biological studies demonstrated a progressive remodeling towards a more elastomuscular phenotype after implantation as graft [16] and clinical reports denied a benefit of antispastic pharmacological therapy [17].

Angiographic studies have demonstrated patency rate of 80–90% at 7–10 years follow-up [18]. A more recent study reported an 84.4% patency rate at 20 years with a probability of graft failure at the same time point similar to LITA (19.0 ± 0.2% for LITA vs 25.0 ± 0.2% for the RA) [19].

The RA might be harvested in skeletonized or pedicled fashion with an open or endoscopic technique. Despite some proposing an advantage in conduit length and diameter if harvested in skeletonized fashion [20], others pointed out the longer harvesting time and the risk for endothelial damage, especially when harmonic scalpel is used [21]. Considering the lack of clear evidence of a significant improvement in patency rate using the skeletonization technique, this approach should be discouraged [11]. However, the success of RA use in CABG does not seem to be dependent either on the effect of an open incision or a minimally invasive endoscopic approach. Conversely, the degree of coronary stenosis influences the long-term patency rate, with vessel stenosis >70% being considered the only indication for RA grafting and stenosis of 90% being associated with even better results [22]. Despite apparent better results on the right coronary system, location of target vessel and site of proximal and distal anastomosis did not significantly influenced the patency rate in a large angiographic study at 6.5 years [23] and 20 years follow-up [19].

At least six randomized clinical trials and several observational studies have compared patency rate and outcomes of RA versus RITA and SVG and very recent evidence from a large patient-level meta-analysis has decreed the “renaissance” of RA [24]. When compared to SVG, randomized trials and meta-analyses have showed significantly higher patency rate and lower incidence of clinical events [22, 25,26,27,28,29,30,31,32]. With respect to RITA, the RAPCO (Radial Artery Patency and Clinical Outcomes) trial demonstrated no difference in the patency of the two conduits with a trend towards improved event-free survival in the RA group at the 6-year follow-up [27]. A meta-analysis including both RCTs and observational studies demonstrated comparable mortality but reduced incidence of cardiac events in the RA group [33]. However, a comparative network meta-analysis on angiographic outcomes demonstrated a non-significant 27% absolute risk reduction for late functional graft failure in the RITA as compared with RA [32]. From a practical perspective, it has been suggested that RA might be a better option in patients at increased risk for post-operative sternal complications. Hoffman et al. and Tranbaugh et al., in two separate propensity matched studies, supported the benefit of RA use in patients at risk of sternal complications [34, 35]. The easily tolerable harvesting also in frail patients and the avoidance of additional sternal devascularization are the perceived advantages of RA use [33]. Additionally, a sub-analysis of the RAPS (Radial Artery Patency Study) on diabetic patients even described a significant risk reduction of graft occlusion using RA in these patients [22], making the use of this conduit in diabetics particularly attractive.

The Right Gastroepiploic Artery

The right gastroepiploic artery (or right gastro-omental artery) is one of the two terminal branches of the gastroduodenal artery, running along the greater curvature of the stomach, between omental layers and eventually anastomosing with the left gastroepiploic artery from the splenic artery.

Histologically, despite sharing the same wall thickness with ITA [36], the RGEA contains many smooth muscle cells in the media, suggesting potential concerns of spasmogenicity during surgical manipulation [37].

First use of RGEA in CABG settings was independently described in 1987 by Pym et al. [38] and Suma et al. [39]. Early hemodynamic studies by Takayama et al. showed that harvested RGEA maintained its physiological behavior after exercise or digestion [40] and retained good flow capacity after endothelial stimulation with vasoactive substances [41]. Also, it was found to suffer very low incidence of severe atherosclerosis [42].

In clinical settings, RGEA is readily available and large angiographic studies on anatomical length and diameter of the conduit demonstrated 97% and 88% probability to reach right coronary and circumflex artery, respectively. Few contraindications have been described suggesting availability of RGEA for every case without the need for a preoperative angiographic evaluation [43]. Similarly, no increase in perioperative risk was identified when compared to patients in which RGEA was not utilized [44].

Because of the anatomical situation, the most favorable target for the in-situ RGEA graft is the distal right coronary artery, although distal circumflex can also be grafted with this conduit. However, a sub-occlusive (>90%) stenosis of the target vessel is recommended to maximize patency rates and avoid spasm or competitive flow issues [45]. In a recent review of evidence and experience with RGEA over 30 years, Suma et al. described patency rate of 97% in the short term and 90% at 10 years, especially using skeletonized harvesting technique and selecting target vessels with >90% stenosis [37, 46]. Despite a network meta-analysis of randomized controlled trials reporting a higher graft occlusion risk of RGEA compared to other conduit [32], 5-, 10- and 15-year actuarial survival rates have been described as 91.7%, 81.4%, and 71.3%, respectively, in a large series with a 20-years follow-up [47]. Moreover, superiority in terms of late survival with the use of RGEA instead of SVG has been reported in two separate series [48, 49]. Proponents of its use advocate the importance of the skeletonized technique [46] and the selection of targets as reliable precautions to achieve success [37], despite the majority of the currently available literature focusing on pedicled RGEA. Only few studies have compared RGEA with RITA [50, 51] providing limited information to draw comprehensive conclusions in this context.

The Saphenous Vein

The great saphenous vein (or “long saphenous vein”) is a large, subcutaneous, superficial vein of the leg, running along the length of the lower limb and returning blood to the deep femoral vein at the femoral triangle. Presence of valves necessitates its use in a reversed fashion when used as a coronary bypass graft. From the biological and histological point of view, several differences have been described in comparison to ITA including the intrinsic structure, the reduced production of endothelial nitric oxide and the impaired secretory pattern of cytokines after implantation in the coronary system [2, 52,53,54]. Since the initial studies by Motwani and Topol [52], the differences in biology and the impact of harvesting technique have been considered the main factors responsible for the incidence of graft failure (both acute and chronic) and the disappointing long-term patency rate (75% at 18 months [52, 55].

Over the years, a significant amount of experimental efforts have focused on minimizing trauma during harvesting or to enhance biological properties of the vein. Souza described the no-touch technique in 1996 in which the dissection of SVG is performed leaving its surrounding tissue in-situ [56]. This method is thought to prevent spasm associated with adventitial denudation during harvesting, inhibit smooth muscle cell activation [57], and preserve integrity of vasa vasorum and endothelial function [58]. The results of a recent randomized trial showed superiority of this technique over the standard harvesting in terms of long-term patency with rates comparable to ITA [59]. However, the advantage of this technique in terms of conduit integrity preservation and future patency rate is counterbalanced by the increased potential for wound infection and bleeding related to the greater “invasiveness” of the procedure [60].

The saphenous vein can also be harvested endoscopically, with the benefit of avoiding large leg incision, wound healing problems and pain, especially when the median saphenous nerve is accidentally damaged during harvesting [61, 62]. Initial concerns about patency of SVG harvested endoscopically were raised in a large observational study in which this technique was associated with vein graft failure and adverse clinical outcomes [63]. The reasons underlying these findings were found in potential structural damage during harvesting [64]. Despite a meta-analysis confirming these results [65], there is no consensus on the inferiority of endoscopic SVG harvesting [62, 66]. A sub-analysis of the large Veterans Affairs Randomized On/Off Bypass (ROOBY) trial also showed lower SVG patency rate in the off-pump group [67]. It is expected that the results of a large angiographic study, the REGROUP trial (Randomized Endo-Vein Graft Prospective) will potentially shed light on this issue.

Another interesting approach has been introduced by Kim et al. in the SAVE-RITA trial which compared the outcomes of RITA with SVG anastomosed as Y graft on the LITA and used to graft both the left and right system. The rationale underlying this study relies on the fact that SVG anastomosed to the LITA would undergo reduced circulatory stress compared to the anastomosis on the ascending aorta and would be continuously exposed to endothelium-protective substances such as nitric oxide produced by the LITA [68, 69]. Additionally, avoidance of aortic cross clamp for proximal anastomosis could reduce the risk of stroke and aortic dissection. Conversely, sparing of RITA might decrease the chances of sternal wound infection while preserving this conduit in case of future redo surgery. One-year angiographic results showed non-inferiority of the saphenous vein composite graft to the RITA composite graft [70]. Five-years results have demonstrated similar positive findings, however, as recently reported, more evidence is needed to confirm the non-inferiority of composite arteriovenous conduits for CABG [71].

Biological augmentation methods to improve SVG resistance to atherosclerosis and failure have also been proposed. In this context the PREVENT IV trial investigated the role of ex-vivo treatment with edifoligide, an E2F decoy that regulates expression of genes controlling smooth muscle cells proliferation and potentially prevents neointimal hyperplasia. However, this was ineffective in the prevention of early vein graft failure [72].

Another strategy available to improve SVG outcomes entails the use of external stenting of SVG with the aim of preventing pressure-induced wall stress and reactive neo-intimal hyperplasia. Despite clear preclinical evidences, the clinical results are conflicting as yet [73]. A nitinol-based mesh (SVS® external venous nitinol mesh, Kips Bay Medical, Minneapolis, USA) has not produced encouraging results at 1-year (patency rate of reinforced SVG 76% versus non reinforced SVG 100% and arterial grafts 100%) [74]. However, newer technologies, such as the VEST device, a cobalt chromium external stent (VEST, Vascular Graft Solutions, Tel Aviv, Israel) have been associated with comparable patency to non-stented grafts [75] and significant improvements in haemodynamic flow within the stents [76]. Ongoing studies will determine whether this translates into superior graft patency over the longer term.

Grafting Strategy and Conduit Choice

LITA is always anastomosed to the LAD on the basis of the widely established prognostic benefit of this strategy. As for the second and subsequent grafts, several combinations and strategies are available. These can be broadly grouped in to a single or multiple arterial strategy. In the first, vein grafts are used for the remaining targets of the left and right system, in the latter, a second arterial conduit such as RITA, RA or RGEA can be used to graft the remaining targets on both the right and left side, or just on the left side, using an SVG on the right system.

On the basis of the evidence currently available and in accordance with the long-term results of previous and new RCTs [77], a multiple arterial grafting strategy should be the preferred approach in the majority of the patients and an ad hoc decisional algorithm has been recently proposed (Fig. 13.1) [78].

Fig. 13.1
figure 1

Decisional algorithm for CABG. In the choice of grafting strategy particular attention should be given to technical, anatomic, and angiographic determinants of arterial conduit patency, as well as the clinical characteristics of the patient (morbid characteristics, diabetes, obesity, respiratory problems, steroid or immunosuppression treatments, especially if they occur simultaneously). Suggested conduit for the anterolateral wall are the RA and RITA, with the first being preferable in cases at risk for sternal complications and the second in patients without ulnar compensation. It is important to remember that RA can be used in sequential anastomosis being able to reach every distal vessel, however it is recommended for subocclusive target stenosis as more sensitive to competitive coronary flow than the RITA. The skeletonized in-situ RGEA can be used to graft the distal branches of the right coronary artery when critically stenosed (>90%). SVG is mainly considered for inferior wall lesions in case arterial conduit grafting is not feasible on clinical or technical grounds. (Note: In this algorithm no composite or elongated grafts are considered.) ∗Obesity, diabetes and severe chronic lung disease, especially in combination. CABG coronary artery bypass graft, RA radial artery, RGEA right gastroepiploic artery, ITA internal thoracic artery, SVG saphenous vein graft

Full size image