Since the first coronary artery bypass was performed on a human in New York in 1961,1 conduits for coronary artery bypass grafting (CABG) remain in a state of evolution. During the early years of CABG development, saphenous venous grafts (SVGs) were standard conduits for CABG. However, it soon became apparent that vein grafts in the majority of patients have limited longevity.2,3 The internal mammary artery (IMA) soon became the ‘gold standard’ with which all conduits are still compared. The left IMA (LIMA) has been definitively established as the conduit of choice in CABG because of superior survival and event-free survival.4 Better clinical outcomes and superior long-term patency with multiple arterial grafts have led to growing interest in the use of total arterial revascularisation in CABG.3–11 This article details the current status of the conduits used to achieve total arterial myocardial revascularisation.
Internal Mammary Artery
Routine use of the IMA has risen from 3% in the early 1970s to above 99%. IMA is very resistant to atherosclerosis, making it superior to any other conduit for CABG. Multiple reasons have been put forward for the low incidence of atherosclerosis in IMA. These include the developed internal elastic lamina, blood supply from the vasa vasorum, the small amount of smooth muscles in the media and its perivascular lymphatic drainage. IMA is currently used in patients of all ages requiring myocardial revascularisation and patients with unstable angina.12 Spasm of distal IMA, usage in non-critical coronary artery stenosis, low flow in IMA and use in hypertrophied left ventricle still remain controversial. Should IMA be used for myocardial revascularisation in these situations? Harvesting of the IMA without any damage or dissection remains very important and utmost care should be taken. If damage to the IMA occurs in the superior or inferior portion and sufficient length is still available for grafting, it can still be used as a free graft. However, the long-term patency of free IMA is not as good as that of the pedicled IMA. The excellent clinical outcomes, long-term patency and fewer repeat procedures of this graft provides a standard with which all other grafts can be compared, and it should be the mainstay of any cardiac surgeon’s coronary artery bypass operation.4,5,13,14
Right Internal Mammary Artery – Bilateral Internal Mammary Artery
The right internal mammary artery (RIMA) can be used as a pedicled graft to bypass the right coronary artery (RCA), the acute marginal or, if long enough, the posterior descending artery (PDA). Pedicled RIMA has also been used for left anterior descending (LAD) coronary artery bypass, and diagonally and through the transverse sinus or in front of the aorta for ramous intermedius and obtuse marginal arteries. Mobilisation of the pedicle, ligation of the branches and transaction of the lower end are identical to those for the procedure on the left side. RIMA can also be used as free graft and can be anastomosed to any of the target coronary vessels, and the proximal anastomoses can be carried out on the aorta or the LIMA. Lytle et al. analysed the results of bilateral versus single IMA and found survival of the bilateral IMA group was 94, 8 and 67% and of the single IMA group 92, 79 and 64% at five, 10 and 15 post-operative years, respectively (p<0.001). They concluded that patients who received bilateral IMAs had decreased risk of death, re-operation and angioplasty.5 However, concerns have been raised about the devascularisation of the sternum with harvesting of BIMA. The length of pedicled RIMA is always a matter of concern if it is to be anastomosed to distal right RCA or PDA. Both of these issues can be addressed by skeletonisation of IMAs.
Harvesting IMAs (left or right) without any accompanying veins, lymphatics, muscles and fascia, etc., is known as skeletonisation. The IMA branches in the anterior intercostal spaces are divided close to the IMA to preserve the collateral supply of the sternum, thereby reducing the risk of sternal wound complications.15 In addition, skeletonisation provides extra length and is associated with a lower incidence of pulmonary complications.16
Sequential and Composite Internal Mammary Artery Grafts
The number of coronary arteries bypassed with IMA grafts can be increased by using free IMA grafts, sequential grafts and composite (T or Y) grafts. In most cases total arterial myocardial revascularisation can be achieved by using T or Y grafts. Joining the end of one IMA to the side of another IMA increases the reach and allows even the posterior marginal branches of the circumflex and right coronary arteries to be bypassed. The proximal end of the free RIMA is anastomosed perpendicularly into the side of the LIMA at the level of the left atrial appendage. Side-to-side sequential anastomosis with RIMA can be performed to bypass vessels on the lateral wall (see Figure 1).
Free RIMA can also be anastomosed to the aorta. However, arterial free conduits anastomosed to the aorta have a significantly lower patency rate than the pedicled or composite grafts. The dp/dt of the ascending aorta is much greater than that of second-generation arteries, e.g. IMAs.17 Increased dp/dt results in a hyperplastic process at the proximal aortic anastomotic site. Therefore, it is recommended that free arterial graft should be anastomosed to the upper third of the IMA in an inverted Y or T fashion to avoid this problem by using both the internal mammaries in this fashion. Four to five anastomoses can be constructed. Technical expertise, kinking/rotation of the IMAs, relying only on one IMA for blood supply and increasing the risk because of proximal subclavian or IMA atherosclerosis remain limitations for the routine use of this technique.
The radial artery (RA) was first used by Carpentier et al. in 1971.18 After four years he reported a 30% occlusion rate of RA grafts, and use of the RA was completely abandoned.19,20 In the early 1990s there was resurgence of the RA as a conduit, as RA graft was found to be patent after 14–18 years post-operatively.21–23 Length and accessibility make the RA an excellent conduit for CABG. To avoid ischaemia, care must be taken to assess the adequacy of the ulnar collateral circulation in the forearm and hand before harvesting the RA. Total arterial revascularisation can be achieved in most patients if RA is combined with IMAs (see Figures 2a and 2b). The proximal end of the RA can be anastomosed to the IMA in an inverted ‘Y’ fashion and it almost always has sufficient length to reach all target vessels on the lateral wall, including the distal PDA, or it can be anastomosed to the aorta as a free graft. Contraindications for the harvesting of the RA include inadequate ulnar artery collaterals as detected by the Allen’s test, injury during harvesting, Raynaud/Burger disease, previous arm arterial catherisation, known subclavian bruits and certain occupations. Age is not a consideration.
Hand ischaemia, numbness of the hand, spasm of the RA and atherosclerosis remain concerns for the harvesting and usage of the RA as a CABG conduit. Few of these shortcomings can be minimised by en bloc harvesting of the RA with satellite veins, avoidance of instrumental dilation of the RA, administration of antispasmodic drugs such as diltiazem and usage of the RA in critically stenosed (>80%) coronary arteries.24
Usage of the RA in non-critically stenosed target vessels may result in competitive flow and a subsequent ‘string sign’ on the angiogram. Early graft patency of RA is 98%. At one year patency of RA grafts was 90%, significantly superior to the 58% patency rate of vein grafts anastomosed to similar target vessels. Follow-up at 5.6 years showed RA patency to be 84% compared with 90% of the LIMA.25
Right Gastroepiploic Artery
In 1987 Pym et al. reported the first clinical application of the right gastroepiploic artery (RGEA) for CABG.26 Since then, pedicled RGEA has become an additional conduit in the surgeon’s armamentarium for bypassing coronary arteries. It has sufficient length to reach any of the target coronary arteries. With experience, the RGEA can be harvested without any subsequent increase in surgical morbidity and mortality. Although it can reach all target vessels, most centres have found that this graft is best suited for bypass to branches of the RCA system. As atherosclerosis has a propensity to form in the RCA/PDA junction, it is preferable to anastomose the graft to the PDA itself and not to the main RCA. One of the main reasons for the RGEA not being used as routine is the small calibre of the artery and the thinness of its wall. When small, thin RGEA is sutured to coronary arteries there is a tendency to kink and stenose at the anastomotic neck due to the size of the pedicle that carries the thin-walled vessel. Other reasons for the RGEA not being a regular conduit are the increased time required to harvest and the reluctance of many surgeons to enter the abdomen. However, in the centres that have the expertise to use the RGEA, it can definitely be used to achieve total arterial myocardial revascularisation.
Inferior Epigastric Artery
In 1990, Puig used the inferior epigastric artery (IEA) as a conduit for CABG.27 This artery arises from the inferior end of each external iliac artery just above the inguinal ligament. It is bilateral and thus provides two segments of conduit. IEA length, which is about 11–13cm, is a limiting factor. Atherosclerosis of the IEA is not very common. The IEA is probably best used as a composite graft with the IMA. Total arterial myocardial revascularisation can thus be accomplished by using the IEA as a simple side arm or an extension. Even now, the IEA is considered as an alternative graft when both IMAs, RAs and the RGEA are not available or have been exhausted, and the patient is scheduled to have an arterial myocardial revascularisation.
Rarely Used Arterial Grafts
A number of additional human arteries can be used as a conduit in specific rare circumstances. The susceptibility to atherosclerotic changes and long-term patency of these grafts is unknown. These alternative arterial conduits include the ulnar artery, the splenic artery, the subscapular artery and the left gastric artery.
There has been increasing interest in total arterial revascularisation because of the superior long-term results. Long-term graft patency and freedom from recurrent angina appear to be the main reasons for the adoption of this strategy by an increasing number of centres; however, many surgeons still feel that the effectiveness of arterial grafting, beyond the use of the LIMA to the LAD, remains unproved, so they have not yet reverted to a ‘total arterial revascularisation strategy’.28 All arterial grafting is technically more demanding and there is undoubtedly a learning curve for the procedure. Two thousand nine hundred and thirty patients with critical proximal lesions may be at early risk if the arterial conduits used do not ‘deliver’ immediately due to small size or liability to vasospasm. In addition, the selection of patients for this grafting strategy and which conduit should be used at which target coronary vessel needs to be completely understood before resorting to total arterial myocardial revascularisation. Native coronary artery stenosis has a major influence on arterial conduit patency. For arterial grafts the best patency is achieved when placed to tightly stenosed or occluded coronary vessels.25,30–32 All conduits (LIMA, RIMA, RGEA and RA), whether pedicled or free, behave similarly. The pedicled LIMA is the most versatile and least sensitive. If the native coronary artery is less than 80%, stenosed patency is affected.
The current strategy for total arterial myocardial revascularisation is to use IMA grafts to revascularise the most important left-sided vessels, sometimes as a composite LIMA and RIMA ‘Y’ graft. Composite IMA grafting allows more options in the use of free RIMA. Pedicled RIMA is to be used for right-sided coronary vessels and sometimes as a cross-over for left-sided vessels. The RA is used for obtuse marginal branches and sometimes for RCA and PDA. The RGEA is best used for critically stenosed PDA.
Recently, a 20-year follow-up of multiple IMA (MIMA) grafts has confirmed and demonstrated clinical benefits of multiple arterial grafts. As long as MIMAs were placed next to the two largest coronary systems, no significant differences in long-term results were observed whether both IMAs were anastomosed to the LAD and left circumflex system or to the LAD and RCA systems.33
Most of the studies that have demonstrated benefits of total arterial myocardial revascularisation are observational and retrospective. These have their own limitations. Ideally, the value of multiple arterial grafts would be clarified by a prospective randomised trial, but such a study would require a large number of patients and would face enormous practical difficulties in proving anything beyond doubt. Variables other than choice of conduits influence patency, and a detailed follow-up of clinical status and graft patency would be necessary for at least 10 years to demonstrate benefits of total arterial myocardial revascularisation. Nevertheless, based on the current body of evidence, multiple arterial grafting should be performed liberally in patients with multivessel coronary disease.