Change in luminal diameter of the left internal thoracic artery anastomosed to the totally occluded left anterior descending coronary artery

The postoperative 1-year CAG revealed no case of a string sign in the LITA graft anastomosed to the LAD; however, 1 case of string sign was found in the RGEA graft anastomosed to the OM in Group 1 and 1 case in PDA-anastomosed RA in Group 2. Patency rates of the RGEA and RA immediately after surgery were 96.2% (25/26) and 100% (32/32), respectively, for Group 1, while patency rates of the RGEA and the RA immediately after surgery were 100% (30/30) and 100% (63/63), respectively, for Group 2. Patency rates of the RGEA and the RA 1 year after surgery were 96.2% (25/26) and 96.9% (31/32), respectively, for Group 1. Patency rates of the RGEA and the RA 1 year after surgery were 100% (30/30) and 96.8% (61/63), respectively, for Group 2. On follow-up CAG, the first branch of the LITA was not seen in any of the patients.

No in-hospital mortality was observed. One case from Group 2 developed low cardiac output syndrome in the immediate postoperative period but recovered. No cases of neurologic complications or myocardial ischemia or infarction were noted during the early postoperative and follow-up periods.

One case in Group 1 was readmitted to the hospital for wound infection during the follow-up period. One case in Group 2 underwent percutaneous coronary intervention because of stenosis of the posterolateral branch anastomosis site. There was no cardiac-related death in either group.

As recent advances in percutaneous coronary intervention are associated with high success rates and reduced complications, CABG has largely become limited to patients with chronic diseases or severe multi-vessel coronary diseases [3]. Over time, patients tend to have a higher possibility of the need for a reoperation because of atherosclerotic changes in a graft rather than because of the native coronary disease itself. In order to reduce reoperation rates because of graft failure, it would be important to select a graft with good patency for primary CABG.

Owing to excellent long-term patency, arterial grafts have been increasingly used recently [4, 5]. In particular, the internal thoracic artery has the advantages of the low incidence of atherosclerosis, a functional arterial intima, an ideal vascular size that concurs with the coronary artery, and capacity for blood flow in accordance with changes in myocardial blood flow demand [1, 6]. For these reasons, anastomosis of the LITA to the LAD has been the basic surgical tenet of CABG.

Based on this fundamental procedure, TAR should be performed using various arterial grafts anastomosed with the LITA by using a composite Y- or T-graft configuration in cases requiring multiple grafting. A restrictive arterial graft could be utilized for TAR by using a composite graft, which has the advantage of a "no-touch technique" of the aorta. Nevertheless, it has the disadvantage of having the LITA as the only blood inflow source, and there are controversies regarding flow competition and inadequate myocardial blood flow in the immediate postoperative period [7, 8]. Thus, it is necessary to verify whether the composite LITA anastomosed to the LAD, which is the most important cardiac blood flow supply, could supply adequate blood volume.

Sakaguchi et al. asserted that a composite arterial Y-graft has less coronary flow reserve as opposed to that of independent grafts, and a small LITA size would call for precautions [2]. In contrast, Wendler et al. found that internal thoracic artery composite grafts could also provide adequate blood flow to the LAD [9]. Akasaka et al. insisted that an independent internal thoracic artery graft could supply adequate blood flow similar to that with saphenous vein grafts [10]. Markwirth et al. used a Doppler-guided wire to measure the flow of the proximal internal thoracic artery and coronary flow reserve after TAR through T-grafts, and observed that the functional and morphological adaptation capabilities of the internal thoracic artery were sufficient for higher flow volume requirements [11]. Citing the data measuring free flow in the operative field after formation of composite Y-graft with the LITA and the RA, Royces et al. reported that TAR with a composite conduit can also compensate for the myocardial blood flow requirement, as it shows a 2.3-fold reserve blood flow to the coronary vascular bed [12].

In patients subjected to TAR using the LITA as a composite graft, it is rare to find angiographic studies on whether sufficient blood flow could be supplied in the immediate postoperative period before achieving complete remodeling of the LITA anastomosed to the LAD. This study was based on patients with total occlusion of the LAD. The reasons were two-fold: (1) LITA flow to the LAD would not be affected by the degree of stenosis of the proximal LAD, and (2) LITA flow would be affected by myocardial blood flow demand alone.

The diameter of the LITA grafted to the LAD is closely related to the degree of proximal stenosis of the LAD, i.e. to the LAD flow volume via the LITA. The diameter of the LITA becomes larger when the coronary flow is LITA-dependent as a result of a severe proximal LAD stenosis. In contrast, the diameter of the LITA becomes smaller when the LAD flow is dependent on the native coronary artery because of a mild proximal LAD stenosis. Among children whose myocardial blood flow demand increases because of somatic growth, remodeling takes place along the length of the internal thoracic artery, as well as on its diameter [13]. That is, the internal thoracic artery graft is an active conduit with functional adaptability to myocardial flow demand, and not a passive conduit.

The remodeling of the internal thoracic artery is an integrated activity involving the endothelium, smooth muscle, fibroblasts, and extracellular matrix [14]. The internal thoracic artery is autoregulated to maintain the base shear stress of 15–20 dyne/cm² [15]. Remodeling takes place on the basis of the Hagen-Poiseuille equation (τ = 4ησ/πγ3. τ = shear stress, η = blood viscosity, σ = blood flow, γ = graft diameter). In the early phase, the LITA flow would depend on flow velocity increase. However, as the high shear stress continues by means of high flow velocity, the diameter of the LITA increases, implicating a remodeling process in which the extent of shear force would decrease to base levels [16, 17].

Although the time when remodeling is completed has yet to be determined, Barner reported that the internal thoracic artery flow measured again 6–8 h after reoperation because of bleeding, showed a 40% increase, and that such increase implicated that flow adaptation of the ITA could occur immediately after surgery [18]. Akasaka et al. revealed that the flow velocity within 1 month of surgery was high and the flow velocity 1 year after surgery was low, while the internal thoracic artery diameter increased [10]. Such phenomenon could be analyzed by the Hagen-Poiseuille equation. Tagusari et al. reported that the LITA diameter significantly increased on angiography performed 14 days on the average after CABG using a composite Y-graft of the LITA and the RA, and reported that the LITA may undergo early adaptation [19]. However, the study did not report diameter change following the LAD stenosis or results of changes in the late period. Nakayama et al. suggested, from the results of angiography performed within 1 month after surgery and a mean interval of 4.5 ± 1.5 years after surgery, that the LITA diameter increased as the LAD stenosis became severe in cases without total LAD occlusion; however, they also reported that the LITA diameters measured for cases of total LAD occlusion were 2.27 ± 0.32 mm and 2.32 ± 0.38 mm, respectively, showing little change [6]. This suggests that LITA remodeling could be completed and may be able to supply sufficient blood flow required for the LAD territory within 1 month. In contrast, Akasaka et al. reported, from cases of independent ITA anastomosed with a totally occluded LAD, that the ITA diameters changed by flow demands, showing them to be 2.4 ± 0.1 mm, 1 month after surgery, while being 2.9 ± 0.2 mm 1 year after surgery [10].

In this study, both groups showed significant increases in the LITA diameters on postoperative 1-year CAG, although the mean LITA diameter of Group 1 was consistently larger than that of Group 2 in both immediate postoperative CAG and postoperative 1-year CAG. The increase in mean diameter of LITA in Group 2 seems to be related, in a certain part, to the increase in flow demand due to the progression of the LAD. Sanidas EA, et al. reported about the natural history of untreated nonculprit lesions in 697 patients with acute coronary syndrome. On angiographic study, 44 patients experienced substantial lesion progression (≥20% angiographic diameter stenosis increase) [20]. Even in a healthy man with a normal coronary angiogram, new coronary lesions do develop [21]. Therefore, it seems to be reasonable to assume that the progression of LAD stenosis has occurred and it might affect the diameter of LITA, although we have not checked the progression of LAD stenosis in the postoperative 1-year CAG in group 2. Interestingly, the mean LITA diameter of group 1 in the immediate postoperative CAG was not different to that of group 2 in the CAG performed 1 year after surgery (2.09 ± 0.53 vs. 2.10 ± 0.45 mm, P = 0.46). It shows that the dilatation of LITA, which happened over 1 year according to the progression of the LAD stenosis in group 2, had occurred in an immediately postoperative period in group 1, indicating that maximum possible enlargement of LITA can be achieved by the LITA-dependent blood flow in the immediate phase.

The increase in mean diameter of the LITA in Group 1 seems to show the remodeling of the LITA reflecting the increase in myocardial demand for blood flow during the first year. This might imply that maximum possible enlargement of the LITA in early postoperative period may not fully satisfy the myocardial demand in a certain circumstance. In this regard, the patients who underwent CABG to totally occluded LAD by using composite Y-graft had good to avoid excessive exercises that could increase myocardial demand during the early postoperative period, even though no myocardial ischemia-related complication was noticed in this study.

Limitations of this study are as follows: being an angiographic study, the flow reserve of the LITA, which is an early compensatory mechanism for high blood flow demand, could not be measured; possible effects of collateral flow of the coronary artery were not considered; the domain of myocardial infarction was not considered although preoperative echocardiography showed significant differences in left ventricular ejection fraction; the degree of change in native coronary artery disease was not taken into account in Group 2; the location of the anastomosis of the LITA and LAD was not considered; and the degree of remodeling immediately after surgery could not be investigated because the size of the LITA was not measured preoperatively. Further studies are needed to address these aspects and involve a larger number of patients with LAD stenosis of various levels.

Changes in the diameters of the LITA on serial angiography, not considering functional LITA flow reserve, suggest that the LITA used as a composite graft during TAR undergoes remodeling, resulting in larger diameters, in order to provide adequate myocardial blood flow. The degree of change in luminal diameter varies according to the severity of the LAD stenosis.