Significant attention has been given to the advantages of complete revascularization and maximizing arterial graft anastomoses in the treatment of multivessel disease. However, graft harvesting itself can be a surgical trauma, potentially involving up to three or four surgical fields on the patient’s skin. Not all patients have a complete theoretical graft pool available, or their usage may be contraindicated, leading to a need to balance graft availability with morbidity.

In this context, myocardial revascularization surgery offers a range of graft configurations. Two major technical variants multiply the number of anastomoses dependent on a donor graft: composite and sequential grafts.

In the first approach, composite grafts utilize Y or inverted T configurations, where one branch serves as the donor and another as the recipient from a side-to-end anastomosis toward another coronary vessel. The common trunk of the graft should not be restrictive, enabling sufficient flow to two territories. This configuration has been suggested for T-configured mammary arteries (the so-called “tector”). While this setup is commonly used with both internal mammary arteries, other configurations are also possible, including using saphenous vein grafts or a free arterial graft (either the mammary or radial artery), where the donor anastomosis is performed on another graft (typically the proximal saphenous vein) instead of the aorta due to length or fragility concerns. This configuration allows for anastomosis between right and left coronary territories, as the composite graft arms are independent yet connected by the common trunk. The distal anastomoses are often side-to-side longitudinal, enabling greater graft length. An exotic variation extends a graft using a saphenous vein or radial artery with a mammary artery donor, allowing for an “no-touch” strategy on the ascending aorta. Sequential anastomoses are then possible with the recipient graft, gaining protection from endothelial factors downstream from the IMA donor graft.

In the second approach, sequential grafting covers multiple epicardial vessels with a single graft, incorporating at least two anastomoses. The final distal anastomosis is side-to-end, while intermediate ones are side-to-side. To maximize graft length and match graft trajectory with epicardial vessels, these may be performed as longitudinal or 90º “diamond” anastomoses. The latter, despite initial permeability, result in higher shear forces due to perpendicular inflow, which may impact patency. This configuration uses graft length most efficiently, though its construction poses a higher technical risk in length measurement and anastomotic geometry. Connecting right and left coronary territories is generally less recommended due to differing physiology, and the progressive drop in flow exposes distal anastomoses to greater native bed competition, necessitating severely diseased vessels, particularly for the right coronary artery. Many studies highlight the benefits of sequential grafts in covering parallel resistances, as the resulting resistance is lower than that of individual grafts.

These configurations are not mutually exclusive, as the first is a proximal anastomosis variant and the second is a distal one. They can coexist in the same revascularization procedure, even as a composite proximal graft with sequential branches.

Despite all these doses of creativity, most of the available evidence, particularly that which relies on clinical trials, barely considers these types of configurations. The majority of grafts used are direct, single grafts, and the composite configuration of both mammary arteries is only considered in a few studies focused on the use of multiple arterial grafts. As we mentioned, there are more than a dozen ways to achieve complete revascularization in multivessel disease. If we add these technical variants, the possibilities approach a hundred, which, aside from being poorly standardized, logically should not yield the same results in all patients.

To this end, this Australian group analyzed the patency of the internal mammary artery (IMA) graft anastomosed to the left anterior descending (LAD) artery in different configurations. It is well known that the proper functionality of this anastomosis is the main prognostic marker for the revascularization of multivessel disease. For this purpose, they conducted a retrospective study of angiographic records from 2002 to 2020 of patients who had previously undergone revascularization surgery. A total of 84% of the patients were referred due to angina symptoms, although only 5.7% showed signs compatible with anterior ischemia. They selected 570 direct internal mammary artery grafts, 100 sequential, and 129 composite grafts. They excluded angiograms where no internal mammary artery graft was anastomosed to the LAD, those that combined both composite and sequential configurations simultaneously, and those composite grafts covering vessels from both the left and right coronary arteries simultaneously. This thorough case selection for analysis led to a reduction from the initial 1,256 angiograms to a final 799.

For direct grafts, 90.7% showed a proximal LAD lesion >70%, and this was <1.5 mm in 3.3%. The failure rate for these grafts was only 3.7%. For sequential grafts, 89% had a >70% proximal lesion in the LAD, with a caliber <1.5 mm in 8%. In 93% of the cases, the initially bypassed vessel was a diagonal branch, with a >70% lesion in 86% of the cases and a caliber <1.5 mm in 10%. The failure rate for the LAD bypass was 9%. For composite Y grafts, the LAD had a stenosis >70% in 92.9% of cases and a caliber <1.5 mm in 3.1%. In 45.7% of cases, the other arm of the anastomosis was used to bypass a diagonal branch, and in 54.3%, it was used for an obtuse marginal branch. Failure of the internal mammary artery anastomosis to the LAD occurred in 6.2% of cases.

The authors performed a logistic regression analysis, concluding that female sex and the presence of a non-significant proximal lesion in the LAD led to worse graft patency. Regarding the configuration, they concluded that direct grafts have better patency and that the failure rate, although not reaching statistical significance, is higher for sequential configurations compared to composite ones. Failure in the sequential configuration occurred distally, prior to the LAD anastomosis, while in the composite configuration, it tended to occur at the level of the Y-branch anastomosis. Therefore, the authors recommend, whenever possible, the use of simple grafts, and if alternative configurations are required—as a compromise between graft availability and revascularization needs—composite configurations should be preferred over sequential ones.

COMENTARY:

The analyzed study stands out for its originality, addressing a previously unexamined assumption: that different graft configurations, as long as they are permeable and function normally intraoperatively, are equivalent. Although this is a retrospective experience with most patients studied due to angina recurrence, it is known that a significant proportion of grafts fail without clinical repercussion, which inherently biases the study’s design. Nevertheless, the “one bridge, one graft” principle is validated, allowing for a standardized revascularization pattern with coronary grafts that are direct and individual, as previously discussed. From this model, other potential variants may yield similar or inferior results, which may be accepted based on graft availability and the reasonable morbidity associated with their harvesting. Consequently, revascularization remains a custom-fit strategy… but now with a framework to tailor adjustments to each patient’s needs. Although occasionally complex, this technique may provide the best solution for complete revascularization. Alternative configurations also present greater risks of anastomotic error, particularly when not frequently practiced.

This study brings us closer to enhancing knowledge in revascularization techniques. The outcomes provided for IMA grafts might not replicate with saphenous vein or radial artery grafts or different combinations. Ideally, future studies will resolve some of these doubts. For now, we understand that we have a one-size-fits-all framework adaptable to a procedure that remains bespoke. Should alternative graft configurations be needed, it is advisable that the IMA-LAD graft remains intact.

REFERENCE:

Singh B, Singh G, Tripathy A, Larobina M, Goldblatt J, Tatoulis J. Comparing the patency of the left internal mammary in single, sequential, and Y grafts. J Thorac Cardiovasc Surg. 2024 Jan;167(1):176-182. doi: 10.1016/j.jtcvs.2022.02.025.