Introduction

While percutaneous coronary interventions (PCI) have tremendously evolved since the first balloon angioplasty in 1977 , with the development of bare metal stents in 1987, the subsequent introduction of three generations of drug-eluting stents (DES) between 2000 and 2011, and now the advent of bioresorbable as well as cell-selective DES , coronary surgery has remained basically the same. An overwhelming proportion of CABG procedures performed in the United States still involve the use of a single internal thoracic artery (ITA) (SITA) and saphenous veins (SVs) as secondary conduits. According to the 2014 STS database, cardiac surgeons use a SITA and vein grafts in 94.7% of their patients . The use of bilateral ITAs (BITAs) is infrequent and remains low, with an increase from 3.5% in 1999 to 4.2% in 2014 . North American surgeons have been reluctant to adopt multiple arterial grafting (MAG), despite potential advantages in graft patency and specific recommendations for a wider use of arterial conduits from the ACCF/AHA in 2011, EACT/ESC in 2014, and the STS in 2016 . Evidence suggests that MAG is beneficial in terms of long-term survival and major adverse cardiovascular event rates. MAG ranges from the use of BITA to the use of full arterial conduits, or total arterial revascularization (TAR).

The benefit of using the left ITA (LITA) on the left anterior descending (LAD) artery is no longer debated, but the advantages of revascularization with additional arterial grafts remain somewhat less recognized, despite growing evidence of their superiority over vein grafts. Suzuki et al. alleged a significant advantage of the ITA bypass in their 43-case series that included patients with BITA grafting . Later, Barner et al. demonstrated the superiority of long-term ITA conduit patency and thus recommended the routine use of BITA . However, Culliford et al. reviewed 2594 cases in 1976 and first suggested BITA as a risk factor for deep sternal wound infections . In the last decade, metaanalyses have pointed at the superiority of BITA over SITA grafting, consistently revealing a roughly 20% increase in long-term survival ( Table 13.1 ). However, retrospective studies are subject to patient selection bias, and prospective studies have yielded conflicting results. In ART trial (Bilateral vs Single Internal-Thoracic-Artery Grafts at 10 years), 3102 patients were randomized to undergo bilateral versus SITA grafting . In the intention to treat analysis, no differences were found either in the primary outcome of all-cause mortality or the secondary composite outcome of death from any cause, myocardial infarction or stroke. These results comforted many surgeons in their practice of avoiding MAG, especially since BITA grafting was associated with a slight but statistically significant increase in sternal wound complications at 6 months postoperatively. However, 14% of patients from the BITA group received only one ITA graft, and a second arterial conduit in the form of a radial artery (RA) was used in 22% of the patients from the SITA group. When analyzed as per their actual treatment group (single artery used vs MAG), a significant 19% reduction in mortality (and 20% reduction in the secondary outcome as well) was observed in the latter group, a magnitude that is consistent with results from retrospective studies. The potential positive impact of radial arteries in the SITA arm of the study is not surprising considering their repeatedly demonstrated superior patency over SVs . In a combined analysis of prospective trials, the use of radial arteries as a second conduit (compared to SV) translated into clinical benefits with a 50% reduction in repeat revascularization and 33% reduction in the composite endpoint of death, myocardial infarction, or repeat revascularization . Moreover, there appears to be an incremental benefit in using not only a second but also a third arterial graft , and ultimately, in aiming at TAR .

Failure to largely adopt BITA grafting may be explained by economic, institutional, and scientific reasons, mainly the potential of an increased risk in sternal wound infection. Critics of MAG often recognize the potential benefit of better long-term patency of arterial conduits, but question the safety and feasibility of this approach, and claim that positive results from retrospective studies may only be due to patient selection bias. Nevertheless, we believe we now have sufficient clinical evidence to propose the wide use of BITA in coronary artery bypass surgery. In this chapter we will describe an approach that has allowed us to safely apply MAG, and even TAR in a majority of cases, based on the use of both ITAs.

Characteristics of the internal thoracic artery

The ITA has unique histological and microscopic features compared to other available conduits. It is the only peripheral vessel to be elastic, thanks to the nonfenestrated internal elastic lamina and sparse smooth muscle cells of the tunica media. This elastic lamina can inhibit cellular migration responsible for media hypertrophy, intimal hyperplasia development, and atherosclerosis . The ITA has an irregular myocyte organization that is responsible for higher basal production of vasodilators (NO and prostacyclin) and better responds to vasodilator agents compared to the RA (mildly) and SVs (markedly) . Endogenous secretion of vasoactive substances is also enhanced by the ITA when stimulated . The ITA presents better endothelial coverage compared to SVs and significantly less thrombogenic intimal defects at the time of implantation . One can then conclude in a greater resistance of the ITA to the trauma of harvesting manipulations . Remodeling adaptation is also observed with an 81% increase in blood flow and 17% increase in diameter at 10 months after surgery .

Left and right ITAs have many similarities in their composition, anatomy, and trajectory. The average length of the LITA is 20.7 cm with a luminal diameter varying from 2.12 to 1.03 mm proximally to distally, before its bifurcation . The right ITA is reported mildly shorter with an average length of 20.1 cm and an average luminal diameter of 1.60 mm .

These properties not only confer ITA’s better patency than SV’s, they most probably also protect the native coronary circulation from the progression of atherosclerosis. Several clinical studies that evaluated the incidence of the progression of stenotic lesions in recipient coronary arteries with an ITA or an SV showed that the rate of occlusion of the primary stenotic lesion ranged from 12% to 39% following ITA grafting, compared to 38%–67% when an SV was used . In another study of over 772 coronary angiographies, Kaplan–Meier-estimated 10-year overall disease progression in territories with patent LITAs was 8%, compared to 11% with patent RA grafts, and 43% with patent SV grafts .

Internal thoracic artery surgical anatomy

The histologic characteristics and its easy access have made the ITA a vessel of choice for bypass surgery. Its anatomy is well described and its trajectory highly predictable. The LITA arises from the first tier of the subclavian artery, opposite to the thyrocervical trunk in more than 90% of the population. The origin of the LITA is as a single vessel in most cases, but in 5%–30% of patients the LITA originates from a common trunk with the thyrocervical trunk, inferior thyroid artery or suprascapular artery. The right ITA arises from the first tier of the subclavian artery and as a single artery in 96% of the population . The course of the LITA starts at 1.0–1.5 cm of the medial border of the first rib, posteriorly to the subclavian vein, and follows an almost straight path until the sixth rib. The ITA runsruns approximately at 10.5mm from the medial border of the sternum at the level of the first rib and at 20.0mm at the level of the sixth rib between the upper costal cartilages and the internal intercostal muscles. The right internal thoracic artery (RITA) is coursing slightly closer to the sternal edge at its bifurcation level (18.4±5.2 mm). Its length varies from 15.1 to 26.0 cm (mean 20.4±2.1 cm) with the LITA (mean 20.7±2.1 cm) and the RITA (20.1±2.0cm) having significantly different lengths . The ITA bifurcates into the superior epigastric artery and musculophrenic artery at the level of sixth rib, where it is covered by the transverse muscle of the thorax. The LITA is covered distally by the transverse muscle of the thorax for approximately 8 cm and the RITA for 6cm. The ITA is boarded by an internal thoracic vein on each side, and these two veins become confluent before connecting to the innominate vein. At each intercostal space the ITA gives rise to anterior intercostal, perforating, and sternal branches. They arise from the ITA in various patterns from single vessels to trunks of two or three branches. The sternal branches give blood supply to the periosteum of the posterior sternum and sternal marrow. The adult sternum is supplied in blood by its periosteal plexus only, which is solely fed by the sternal branches of the ITA. After ITA harvesting, neo-collaterals are developed by the anastomosis of the anterior intercostal branches and the musculophrenic artery to the posterior intercostal branches, branches of the thoracoacromial artery and branches of the lateral thoracic artery. Green advocates that many collaterals can supply the sternum after the ITA is harvested, and that the ITA branches should be divided as close as possible from their origins to promote the formation of these neo-vessels. Galbut et al. report similar findings of very low sternal infection rates with skeletonized ITA in BITA and diabetic patients. Preservation of the pericardiophrenic artery and the collaterals between the 7th and 10th ribs was associated with better healing of the sternum, and no difference in deep sternal wound infection between diabetic and nondiabetic patients undergoing BITA grafting .

When both ITA are harvested, caution must be taken to not injure the left phrenic nerve that will cross the LITA anteriorly while the right phrenic nerve will cross the RITA posteriorly in 12% of the time. The inverse situation is observed in 20% of the population. The point where the phrenic nerve and the LITA overlap is on average 1.5±0.7 cm from the subclavian artery. The right phrenic nerve crosses the RITA on average at 1.9±0.7 cm from its origin. The pericardiophrenic artery arises from the ITA in 99% of cases at 3.9 cm from the subclavian artery and is in part responsible for the blood supply of the phrenic nerve. Its dissection could compromise the vascular supply of the phrenic nerve and result in diaphragm palsy.

Harvesting technique

General considerations

Retractors: After skin incision, division of the sternum and hemostasis avoiding the use of bone wax, we apply a vancomycin paste on both sternal edges. This application will be repeated at the end of the procedure, immediately before sternal closure. A self-retaining internal mammary retractor is positioned to expose the ITA. In order to optimize the potential of using distal segments of our ITA’s as Y or T grafts, we try to extend the harvest as proximally as we can, clipping all side branches, and past the bifurcation distally ( Fig. 13.1 ). Care must be taken to prevent costal and sternal fractures, especially when both ITA’s will be harvested. Excessive separation can injure the brachial plexus or result into dislocation of the costosternal joint. While harvesting the RITA, overspreading the sternal leaves can result in hemodynamic instability. The retractor pushes the right ventricle and prevents it from appropriate filling. Tachycardia, low blood pressure, atrial fibrillation, or ST changes can be signs of hemodynamic instability. Urgent revascularization or poorly protected severe coronary artery disease are risk factors for hemodynamic instability while harvesting ITAs.

Proximal (A) and distal (B) extension of ITA harvesting. ITA , internal thoracic artery.

Harmonic scalpel

The electrocautery is widely used to harvest pedicled ITAs. The blade stays usually far from the artery and endothelial damage is less likely. For skeletonized harvesting the cautery tip is used to dissect directly the artery from the branches and surrounding tissue. Though usually used at a lower voltage, the heat of the electrocautery tip is more likely to injure the endothelium of the conduit. During dissection the heat from the electrocautery spreads to the surrounding tissue and can injure the artery. The ultrasonic scalpel (USS) is a device that limits electrical and thermal injury to the surrounding tissue. The blade can be used to divide the ITA from its branches, which yields a clipless ITA that may better suit the creation of sequential bypasses. The absence of metal clips also provides a more regular and smoother conduit. Kieser et al. compared more than 1000 skeletonized ITA harvests with USS or traditional electrocautery and found no difference in postoperative major adverse events . Their cohort included 74% of patients with BITA. Electron microscope analysis showed significantly less endothelial injury with USS when branch pedicles were less than 0.5 cm . Significant shorter operative time was reported compared to BITA harvested with an electrocautery and no-touch technique. USS reduces endothelial defect to the ITA and vasospastic response of the arterial wall . Paivendi et al. report saving 10 minutes per ITA harvest with the use of the USS compared to a no-touch technique. No difference was shown on sternal perfusion after pedicled harvesting of ITA with USS compared with electrocautery.

Our approach is to use pedicled ITAs to minimize the risk of trauma when patients have a body mass index (BMI)> (over) 30 kg/m ² and it is expected that the length of the conduits will not be an issue. In patients where distal segments of the ITAs are planned to be used as Y grafts, in those where a significant curve to sequentially graft a diagonal branch with a lateral position (more obtuse angle) and the LAD is expected, or in patients with a BMI>30 kg/m ² (or other risk factors for mediastinitis), we use skeletonized conduits. Most skeletonized arteries are harvested with the use of the electrocautery at low energy (25 W), using metal clips to separate the branches. In patients at higher risk of developing wound infections, we either use the harmonic scalpel when available, or fine-point De Martel scissors (completely avoiding the use of the electrocautery), with a no-touch technique ( Fig. 13.2 ).

ITA harvested with ultrasonic scalpel (harmonic): smooth and clipless vessel. ITA , internal thoracic artery.

Conduct of operations

Patient characteristics, the severity of coronary disease, and the distribution of stenotic lesions should all be considered when planning BITA grafting. All these variables can affect the final configuration of the montage and sequence of the procedure. Surgeons should take into consideration the typical risk factors for mediastinitis, including female gender, obesity, insulin-treated diabetes mellitus, chronic obstructive pulmonary disease, and immune suppressors/steroids intake . None of these factors constitutes an absolute contraindication to BITA grafting per se, but in the presence of very high–risk patients we often try to achieve MAG or even TAR using only the LITA and the RA. The distribution of the coronary lesions and their stenosis severity also impact the final configuration of the bypasses. To optimize the efficacy of arterial conduits, the risk of competitive flow with the native coronary circulation must be minimized. Therefore recommendations have been made to avoid using an arterial graft to bypass a coronary artery with a nonhemodynamically significant stenosis, with an angiographic cutoff of 70%, in order to lower the risk of graft failure . More recently, fractional flow reserve (FFR) has emerged as a potentially more reliable measure of the hemodynamic significance of a stenosis and can guide coronary artery bypass surgery as well. The IMPAG (Impact of Preoperative FFR on Arterial Bypass Graft Function) trial suggested an FFR cutoff of 0.78, which was associated with an anastomotic occlusion rate of 3% at 6 months .

In our practice, we find it useful to plan different strategies of revascularization and conduit disposition before entering the operating room, based on coronary angiogram evaluation. A correlation between angiogram assessment and actual anatomical features is confirmed once the heart is exposed. We perform the vast majority of our cases with cardiopulmonary bypass (CPB) and cross-clamping. After initiation of CPB, we mark the target coronary vessels with a fine blade, selecting sites exempt of significant calcifications and planning the configuration of the conduits. To get an estimation of the required length for the different conduits, a silk suture is often used to measure distances between planned anastomotic sites and cut to serve as gauges. Fig. 13.3 shows a silk suture being measured between the posterior descending artery (PDA) and the point where the RITA will plunge under the aorta. The silk is then cut and used to determine the adequate length of the RA that will be used to graft the PDA from the RITA, allowing the use of any remaining segment of the RA to create Y graft(s) ( Fig. 13.3 ). The basic principle that we follow is to try to use both ITAs as in situ grafts and thus have two separate inflows for our Y and/or T grafts. These ITAs act as tree trunks from which Y and T grafts will emerge as primary and sometimes secondary branches. In the most often used configuration, the LITA is used to bypass the LAD, the RITA passed through the transverse sinus to reach the most significant branch of the circumflex distribution, and the RA is directed to the right coronary system, anastomosed proximally as a Y graft from the RITA before it plunges beneath the aorta ( Fig. 13.4 ). Fig. 13.5 shows a preoperative angiogram from a patient who had undergone CABG with this particular configuration and presented with early aortic bioprosthesis degeneration requiring redo surgery. Diagonal branches are revascularized either sequentially in a side-to-side longitudinal fashion, when they lie parallel to the LAD ( Fig. 13.6A ), or with the use of short segments cut from the distal LITA, RITA, or RA used as Y grafts from the LITA when their origin is too proximal or their angulation from the LAD too obtuse to avoid anteroposterior and lateral kinking of the LITA ( Fig. 13.6B ), respectively. For the circumflex artery, if additional branches from the lateral wall require grafting, they are most often reached with similar distal segments of the LITA, RITA, or RA, which are anastomosed on the distal RITA, within 1 cm of the anastomosis to its target coronary vessel ( Fig. 13.7A ). Fig. 13.7B shows (left panel) that a significant length of the distal LITA can be cut for further use as a Y branch, and (right panel) excellent and equivalent flow in both branches of a short distal segment of RITA anastomosed on a donor LITA. When these segments are too short, the RA can be anastomosed as a Y graft on the very distal RITA and used to bypass marginal branches and the right coronary system sequentially ( Fig. 13.8 ), or the RITA used as a free graft, originating from the LITA, and deployed in a similar fashion ( Fig. 13.9 ). For the right coronary system, using the proximal RITA as an inflow for the RA usually allows to reach the PDA and posterolateral (PL) branches sequentially, when needed. In case a PL branch is very distally positioned, the proximal anastomosis should be done on the distal segment of the RITA and the RA deployed along the lateral wall, to reach first the PL and then the PDA sequentially ( Fig. 13.10 ). Alternately, the RITA can be used in situ to bypass the right coronary artery, especially if the lesion is proximal and/or the distal bed of target vessels from the lateral wall is poor. The RA can then be branched from the LITA to reach the circumflex distribution ( Fig. 13.11 ).

Use of a silk as a gauge to determine the required length of radial artery to be grafted from the RITA to the PDA. PDA , Posterior descending artery; RITA , right internal thoracic artery.

Stay updated, free articles. Join our Telegram channel

Join