Coronary artery bypass grafting using the radial artery : influence of proximal anastomosis site in mid-term and long-term graft patency

Objective: To determine whether the proximal anastomosis interferes or not in midand long-term patency


Conclusion:
The site of proximal anastomosis of the RA coronary grafts does not interfere in mid-and long-term graft occlusion and patency.

INTRODUCTION
Even nowadays, the radial artery (RA) graft is a controversial alternative, but with wide use in coronary artery bypass grafting.Since the first publications in the 1970s [1], with disappointing short-term results [2][3][4], until its "resurgence" in the 1990s [5], the results indicated by numerous series of patients founded well established concepts about this graft.It is known, for example, about the effect of the degree of obstructive lesion of the target coronary on evolution of its flow [6,7].However, only recently, other aspects of RA have been studied such as, for example, the different techniques for graft dissection or the site of proximal anastomosis [6,[8][9][10].
The aim of this study is to determine whether the site of proximal anastomosis influences on patency in mid-and long-term of RA grafts.

METHODS
Between 1994 and 2006, 481 patients underwent CABG at the Heart Institute of Clinics Hospital of Faculty of Medicine, University of São Paulo (InCor / FMUSP), operated by the same surgical team (surgeon and first assistant).In all surgeries, was used at least one RA graft.Of these patients, 123 (25.57%) underwent coronariography restudy in the postoperative period, according recommended by the Clinics of Coronary Disease of the institution.The data -including the coronariography exams performed and their reports -were obtained by evaluating the records, with approval of the Ethics Committee for
Of this total, 150 arteries were grafted with RA by means of single end-to-side or side-to-side anastomosis (sequential) in more than one coronary branches.The RA dissection in all cases was the same, using single incision through the ventral side of the forearm selected, from the distal portion of the vessel, with manual dissection and ligation of branches, thus preserving adjacent tissues and extending up to the junction of the interosseous branch.The left marginal branches (LMB) -or equivalent from the same territory -were the most prevalent (48.67%).The other coronary branches, in decreasing order of use of RA, were diagonal (DB) (30.67%), right coronary artery (RCA) (15.33%) and anterior interventricular branch (AIB) (5.33%) (Figure 1).
In this group of patients, the proximal anastomosis of RA was performed directly in the aorta in 50 (40.65%)patients, through traditional end-to-side technique and using 7-0 polypropylene thread.In the remaining 73 (59.35%) patients, the RA was used as composed graft, through end-to-side Y-shaped anastomosis to the left internal thoracic artery (LITA) and right internal thoracic artery (RITA), using 7-0 polypropylene thread.There were no complications during the postoperative period and all patients were discharged from hospital with maintenance of outpatient follow-up.behavior could be observed similarly in both groups of proximal anastomoses, with statistical significance only in the Y-shaped proximal anastomosis (P=0.0031)(Figures 5 and 6).
Coronary angiography was performed in 123 patients studied, within a mean period of 5.36 + 3.21 years, also with injection in aorto-coronary grafts and LITA and/or RITA, if present.
The data were divided into two categories, as the proximal anastomosis (aorta/ composed graft) and the graft patency (occluded/pervious).For comparison among variables, we used chi-square test for two proportions (or linear trend chi-square for comparison among more than two proportions), with confidence interval (CI) of 95%.

RESULTS
Considering the evolution of the RA grafts, irrespective of the proximal anastomosis, we noted a patency rate of 82.11%.The grafts with total obstructive lesion in anastomoses (distal/proximal) and/or on the graft were considered as "occluded" (17.89%).
There was no statistically significant difference in distribution of obstructive lesions of grafted coronaries between the two groups of proximal anastomosis, and the most prevalent obstructions were between 90% and 99% (Figure 2).
We noted a predominance of graft patency in the most severe obstructive lesions, particularly in groups with obstructions of 90% or more (P=0.0003)(Figure 4).This

DISCUSSION
It is known the historical evolution of the use of RA in CABG.Since the pioneering proposal and comments of Carpentier et al. [1] in 1973, investigations were made in various aspects, aiming to evaluate its effectiveness.The first studies, even in the 1970s, showed unfavorable results [2,3], which raised interesting in different studies to determine possible factors -perhaps at the histopathological level -that may justify the behavior of early occlusion of RA [4,11,12].In the early 1990s, numerous technical and pharmacological advances [6.8] raised again the interest in this graft [5,13,14].With this new approach, there were very satisfactory results, but always in series of patients restudied early or in short-term [5], between 14 and 18 months [6,10,13,15].

Fig. 6 -Evolution of RA grafts patency as the obstructive lesion of the grafted coronay -group "Y-shaped" anastomosis (n=73)
From this time and since then there has been preference for Y-shaped proximal anastomosis, usually involving the LITA or RITA [7,[15][16][17][18].This trend was based on the favorable adaptation shown by in situ internal thoracic arteries by appropriate supplying of blood for Y-shaped composed grafts [7.19].Furthermore, some authors have shown good results with RA, using the proximal anastomosis in the aorta, by always considering the shortterm restudies [9,20,21].
Good results have also been showed in recent seriesof which the follow-up time reaches 6 years [22.23].In these studies, there is no evidence that the proximal anastomosis affects RA patency at mid-term [24].Currently, the site of proximal anastomosis is recommended, considering, for example, the less or even no manipulation of the ascending aorta [21].It is also considered that the topography of the target coronary could interfere with the RA patency [25] data that have not yet been shown by other authors.
The degree of preoperative obstructive lesion interferes with the development of RA patency.The possibility of competition between the flow and the graft is described, especially when the obstructive lesion of the targetcoronary is not subocclusive [22].
In a series of 54 patients restudied with one year of postoperative, there was also apparent vulnerability of the RA grafts under situations of "theft of flow."The authors describe 50% of graft occlusion for preoperative obstructions less than 60% [26].In postoperative angiographic restudy (mean 32 months) of 123 patients undergone surgery using ITA and RA grafts was observed patency of 99.2% and 92% respectively.The highest patency rates of RA were recorded in the "target coronary" with obstructions of 90% or more (98%) in relation to the less obstructive lesions (83.3%, P<0.05) [27].Similar results were shown in retrospective assessment of 600 patients, where 93 (15.5%) were restudied, with 92.5% of RA grafts patency rate, and all occluded grafts were related to the coronary arteries with less severe obstructions, 56.3% + 15.4 (P <0.001) [28].
It should be considered further, the progress of the RAPCO study (Radial Artery Patency and Clinical Outcome), that compares the RA grafts with free RITA and saphenous vein graft (SVG).Preliminary results of this study include mean time of follow-up of 3.5 years.The authors note that although with tendency to be better than PVS, the free arterial graft does not maintain patency comparable to in situ LITA or RITA.Clear differences on the influence of the proximal anastomosis in the results were not shown [29].
The number of patients we studied who had undergone CABG using RA, presents midterm evolution compatible with the data available in recent literature.
Still, considering the time of follow-up of these patients and time of restudy, it is noted that the favorable results described herein extend from mid-to long-term.

CONCLUSION
It is concluded that the site of proximal anastomosis (aorta/ "Y-shaped" with in situ LITA or RITA) did not interfere with the mid-and long-term patency of the RA grafts.
The degree of obstructive lesion interferes with the patency of RA grafts with a higher number of pervious grafts in obstruction of 90% or more.
CARNEIRO, LJ ET AL -Coronary artery bypass grafting using the radial artery: influence of proximal anastomosis site in mid-term and long-term graft patency Rev Bras Cir Cardiovasc 2009; 24 (1): 38-43

Fig. 2 -Fig. 1 -
Fig. 2 -Distribution of different degrees of obstructive lesion in both groups of proximal anastomosis

Fig. 5 -
Fig. 5 -Evolution of RA grafts patency as the obstructive lesion of the grafted coronay -group proximal anastomosis in the aorta (n=50)