| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Click here to read this article as a CME activity
a Department of Cardiothoracic Surgery, Royal Melbourne Hospital, Melbourne, Victoria, Australia
b University of Melbourne, Melbourne, Victoria, Australia
c Department of Cardiothoracic Surgery, Epworth Hospital, Melbourne, Victoria, Australia
Accepted for publication March 25, 2009.
* Address correspondence to Prof Tatoulis, Private Medical Centre, Ste 28, Royal Melbourne Hospital, Melbourne, Victoria, 3050, Australia (Email: james.tatoulis{at}mh.org.au).
Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
| ADULT CARDIAC SURGERY:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org]. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Methods: Two independent observers evaluated 1108 consecutive postoperative RA conduit angiograms performed between January 1997 and June 2007 for cardiac symptoms. Mean time to postoperative angiography was 48.3 months (range, 1 to 132 months). An RA graft was considered failed (nonpatent) if there was stenosis exceeding 60%, string sign, or occlusion. Patency was determined over time, by coronary territory grafted and by the degree of native coronary artery stenosis (NCAS).
Results: At a mean of 48.3 months, 982 of the 1108 RA grafts (89%) were patent. RA patencies for the left anterior descending were 96% (24 of 25), diagonal/intermediate, 90% (121 of 135); circumflex marginal, 89% (499 of 561); right coronary, 83% (38 of 46); posterior descending, 89% (253 of 286); and left ventricular branch/posterolateral, 86% (47 of 55). Patency was 87.5% (56 of 64) for NCAS of less than 60% compared with 89% (926 of 1044; p = 0.89) for NCAS exceeding 60%. Of 318 RAs in place more than 5 years, 294 (92.5%) were patent, and for 107 RAs in place for more than 7 years, 99 were patent (92.5%). Patency was consistent through each year of the decade. Mechanisms of failure did not involve development of atherosclerosis. Patent RA grafts were smooth, with no angiographic evidence of atheroma.
Conclusions: Late patencies of RA grafts are excellent and justify continuing use of the RA in coronary operations.
The objective of coronary artery bypass grafting (CABG) is excellent long-term results through complete revascularization and optimum patencies of the conduits used. Internal thoracic artery (ITA) graft use in CABG is associated with better patencies and short-term and long-term results [1–4]. It seemed logical to use a further arterial graft to avoid the intimal hyperplasia, atherosclerosis, and occlusions that develop in saphenous vein grafts (SVG) [5].
Reports of long-term radial artery (RA) patencies by Acar and colleagues [6] and the further early work by Calafiore and colleagues [7], Brodman and colleagues [8], and others encouraged us to use the RA as the next graft of choice after the left ITA (LITA) and right ITA (RITA) to avoid the limitations and complications of SVGs. The perceived advantages of the RA were an arterial wall structure, comparable diameter, excellent length, robustness, and superior healing of the forearms compared with leg wounds. Hence, we have developed a large experience with the use of the RA artery [9].
However, because the RA is a muscular artery, there has been concern over its biologic behavior and its vulnerability to spasm and to competitive flow. Although most clinical and patency reports regarding the RA have been favorable [10–15], some are not [16]. Numerous early-term (1 to 3 years) and medium-term (5 years) reports have been published, including randomized trials [10–19]; however, few reports have detailed the long-term patency of the RA as a coronary graft [20]. The aim of this study was to document the medium-term to long-term patency of a large number of RA grafts through our 12-year experience. Secondary aims were to note any changes in patency over time, to correlate RA graft patency with the coronary territory grafted and native coronary artery stenosis, and to document the mode of RA graft failure.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
We studied 1108 consecutive postoperative RA conduit angiograms performed between January 1997 and June 2007 for possible postoperative cardiac symptoms. Symptoms included typical angina, atypical chest pain, new myocardial infarction, and unexplained shortness of breath. During this time, 14,074 patients underwent CABG in which a RA graft was used as part of the coronary reconstruction.
Radial Artery Graft Procurement and Spasm Prophylaxis
An Allen's test and finger plethysmography were always performed. The RA, preferably from the nondominant forearm, was always harvested with an atraumatic, open technique, using sharp dissection and vascular clips for branches, supplemented by minimal low power cautery. The RA was harvested with its venae-comitantes. Vascular clips and suture ligatures were used to secure the distal and proximal stumps. Operative details have been previously published [9, 15].
After distal transection, 2 to 3 mL of papaverine solution (1 mg/mL in heparinized arterial blood at 37°C) was atraumatically injected intraluminally. The distal RA was clipped, and left to beat under arterial pressure for 5 minutes before the proximal end was divided and the stump secured. The RA conduit was stored in an identical solution until use.
All patients received intravenous nitroglycerine (30 to 100 µg/min) intraoperatively and for the first 24 hours. All received aspirin (100 mg daily) commencing the first postoperative day and the calcium channel antagonist amlodipine (5 mg orally daily) for 6 months [9, 15].
Grafting Strategy
The LITA was used as an in situ graft to the left anterior descending (LAD) artery. In younger patients (< 65 years), the RITA was used to the next most important vessel, usually the circumflex marginal, and the RA was used as a supplementary graft. In older patients (> 65 years), the LITA was again used to the LAD. Use of the RITA was less frequent, and in general, the RA was used as the second graft of choice, particularly to the circumflex or posterior descending coronary artery. Supplementary grafts included a second RA or SVG.
Since 2000 when the influence of competitive flow on RA patency became apparent, RA conduits were usually deployed only to arteries with stenosis exceeding 80%. Moderate coronary stenoses, particularly in large vessels, were managed by stenting in younger patients or with SVG in older patients.
Most RA grafts were deployed to the circumflex marginal, followed by the posterior descending. There was a mean of 3.2 distal anastomoses per patient, and 81% were constructed with arterial grafts. RA grafts were used to the LAD when an ITA graft was not appropriate (damaged during harvesting or poor flow) or not available (prior use or a string sign at reoperation). Supplementary SVGs were used to achieve complete myocardial revascularization. The right gastroepiploic artery, ulnar, and inferior epigastric arteries were infrequently used.
All CABG procedures were performed using cardiopulmonary bypass and single aortic clamping for all distal and proximal anastomoses together with combined antegrade and retrograde blood cardioplegia. RA conduits were routinely used as aortocoronary grafts. Sequential anastomoses in the circumflex system were infrequently used [9, 15].
All postoperative coronary graft angiograms were performed by conventional angiography by using a transfemoral route under local anesthetic. All angiograms were reported by 2 observers independently. RA graft failure was defined as complete occlusion, a string sign (conduit 
1 mm diameter for some or all of its length), or a discrete stenosis exceeding 60% anywhere within the conduit or at either anastomosis.
Data Collection and Analysis
Demographic and angiographic data were collected prospectively and entered into computer database programs. Intraoperative data, including data related to the conduits, were entered at the time of operation. Postoperative angiographic patency data were entered at the time of angiographic review and of the angiogram report. Values are reported as the mean ± standard deviation. Percentages are given where appropriate. Statistical analysis was done with SPSS PC-Plus software (SPSS Inc, Chicago, IL). The 
2 test was used for categoric variables. A value of p < 0.05 was considered significant. Long-term RA conduit survival and patency was calculated by the Kaplan-Meier method but is also presented as patency within each individual time period.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
60%, string sign, total occlusion).
RA Conduit Patency: Influence of Territory Grafted
RA graft patency was best when the RA was grafted to the LAD (96%) and worst when grafted to the main right coronary artery (RCA) usually between the acute margin and the crux (83%). Details are presented in Table 1.
|
When NCAS was less than 60%, 56 of 64 grafts were patent (87.5%) compared with 926 of 1044 (89%; p = 0.89) for grafts placed to arteries where NCAS exceeded 60%. However, conduits with less than 60% NCAS comprised only 5.8% of all RA conduits.
When NCAS was less than 80%, 265 of 306 (86.6%) were patent compared with 717 of 802 (89.4%) when NCAS exceeded 80% (p = 0.76). Only 27% of RA grafts were placed to targets with NCAS of less than 80%. Although there was a trend to better patency with the more severe NCAS, this was not significant (Table 2).
|
|
There were 318 RA conduits in place for more than 5 years, and 294 (92.5%) of these were patent. Of the 107 RA conduits that were in place for more than 7 years, 99 (92.5%) were patent.
Modes of Conduit Failure–Qualitative Observations
Three modes of failure were observed: total occlusion, a string sign, and localized stenosis. String signs almost always occurred when the NCAS was moderate and especially when the coronary artery grafted was of a large caliber (Fig 2). Localized stenoses within the RA conduit were also seen, usually over a short segment of approximately 1 cm (Fig 3). These were usually managed by balloon angioplasty and stenting. In this large observational study, no instance of diffuse, irregular, endoluminal abnormalities consistent with atheroma was noted. All patent RA conduits had smooth, uniform endoluminal appearances on angiography (Fig 4). In addition, those RA conduits that had discrete stenoses had smooth uniform endoluminal appearances in all the other portions of the graft.
|
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This large observational study of postoperative RA conduit angiograms includes more than 300 RA grafts in situ for 5 years or more. The overall RA patency of 89% is in keeping with results published by others ranging from 88% to 90% at 5 years [9–15, 17–19] to 88% at 9 years [20]. RA patency results published thus far are excellent, but there is one report of poor patency, and this questions the process of harvesting, the muscular nature of the RA, management of spasm, and strategies of RA deployment [16].
We avoid RAs where there is calcification and have preferred open, scissor dissection, and discrete clipping of branches, with minimal use of low-power diathermy. The RA is harvested together with is venae-comitantes, ensuring a margin from the RA vessel wall and minimizing interference with its adventitial integrity. Some authors, however, have had good results harvesting the RA as a skeletonized graft or using an endoscopic approach [21]. The RA is always harvested by the cardiac surgeon or a senior cardiac surgery trainee. We also preferentially use the proximal part of the RA, because distally it is more muscular, has a smaller diameter, and is more prone to calcification [22].
Spasm always occurs, despite a no-touch technique, and we always use spasm prophylaxis (as described) that achieves a large, dilated conduit when the cross-clamp is released and the RA graft fills under systemic pressure. The poor results documented by Koht and colleagues [16] may reflect differences in the approach to harvesting, spasm prophylaxis, and deployment.
Postoperative use of oral once-daily amlodipine, a calcium channel antagonist, is empiric and based on our observations of isolated incidents of RA graft spasm up to several months postoperatively. A randomized trial and a meta-analysis of postoperative calcium channel antagonist use failed to show any clinical benefit; however, it is possible that graft spasm or closure may occur without necessarily having a clinical effect [23, 24].
The RA patency to the circumflex marginal and posterior descending arteries is excellent, at approximately 90% and in keeping with the findings reported by other authors, and some report identical patencies for the RA and RITA when used to the circumflex marginal [10–13]. Lower RA patencies to the main RCA may reflect competitive flow in a relatively large vessel. The excellent, late patency results suggest the RA is versatile and can be used in any territory if so required—even to the LAD (patency 96%) if for some reason the LITA is unavailable or damaged [15].
There was a tendency for lower RA patency, including string signs, when used to graft coronaries with moderate stenoses, most likely because of competitive flow that appears to exert a greater influence than for the ITAs [10–15, 17–20, 25]. We, and others have recognized this, and now avoid deploying the RA in such a situation. Hence, in this observational study, only 5.6% of all RA grafts were deployed to a target vessel with less than 60% stenosis. Many workers now report that the more severe the NCAS stenosis, the better the RA patency. The best RA patencies (> 90%) are achieved when the NCAS is 90% or higher [18, 26], and RA patencies are similar whether used as aortocoronary or LITA-RA composite grafts [10–13, 21, 25, 27].
Patency results over time were consistent, and patencies of more than 90% for the 318 and 107 RA conduits in situ of more than 5 and 7 years, respectively, are extremely encouraging and consistent with the findings of Possati and colleagues [20]. The observation of slightly lower patency rates in the early annual periods may reflect early graft failure causing symptoms and prompting early angiography, whereas angiograms done after 5 years may have resulted from symptoms due to late closure of supplementary vein grafts or development of problems unrelated to graft closure, particularly aortic stenosis.
No diffuse intraluminal irregularities were noted in the almost 1000 patent RA conduits or in those with discrete stenoses to suggest atherosclerosis. The endoluminal RA outline was smooth and uniform. We may speculate from these observational findings that once an arterial graft is in place and functioning well, it will continue to do so in the long-term. Indeed, when we have reoperated on patients 5 to 10 years later (usually for new aortic stenosis), we have noted in situ functioning RA grafts to be pristine, with normal walls, free of any wall thickening or atheroma.
A further implication is that if reintervention is required for a failed graft, there is less risk of atheromatous embolization and less danger to the patient.
Enhanced graft patency is associated with improved outcomes and results in fewer reoperations. Randomized, propensity-matched, and observational studies where large numbers of RAs have been used to achieve arterial revascularization have all shown a survival benefit [9–14, 17–20].
The dilemma of the "best strategy" for the RCA system is ever present. Assuming 2 arterial grafts on the left side, what is the role of the RA? For younger patients, we place the RA to the posterior descending coronary artery if there is a tight (> 80%) proximal stenosis or advise balloon angioplasty and stenting if the RCA stenosis is moderate. For older patients, depending on conduit availability, an RA if NCAS is tight, or an SVG if moderate. If the RCA territory has already sustained a significant infarction but is still judged worth grafting, then graft choice is probably optional.
This study has a few limitations. This is an observational study of angiograms in symptomatic patients; however, the study makes data on a large number of postoperative RA angiograms available. Care is required in interpretation. Because these patients were symptomatic, patency results may be expected to be better than those reported here. Symptom-directed observational studies appear to have twice the graft failure rates of asymptomatic/trial studies [28, 29].
Randomized trials of RA against SVG or RITA to allow a more scientific answer are being conducted [17–19]. However, it may take a further 5 years to derive definitive answers given that the benefits of bilateral ITA vs LITA/SVG took more than 10 years to become apparent. It could be argued that if it takes more than 10 years to show possible benefits of RA over SVG, is it worthwhile? Our response is yes, because the mean age of patients undergoing CABG in our practice is 68 years, improved 10-year results would translate into less recurrent angina, myocardial infarction, and less need for reintervention in patients aged 78 to 85 years old in the future. Indeed, Muneretto and colleagues [13] have shown improved outcomes in a randomized trial in elderly patients when multiple arterial grafts (including RA) are used.
The exact time of conduit occlusions is not known. The mode of reporting assumes they occurred at the time of angiography, but all occlusions probably occurred at various times before angiography.
This angiographic cohort is less than 10% of all patients undergoing CABG with at least 1 RA graft during the time of the study, and we need to consider how the results may relate to the other 90%. However, these proportions are similar to randomized trials where typically only 5% to 10% of potential candidates are enrolled [17–19]. The overall patency results would probably be even better if the asymptomatic patients (> 90%) were added.
In conclusion, the late patencies of RA grafts are excellent, similar to reported RITA patencies, particularly to the circumflex, and superior to reported SVG patencies. A significant mode of failure is due to competitive flow rather than atherosclerosis as in SVGs, and functioning RA grafts seem to remain patent in the long-term. The findings support the use of the RA as a graft of second or third choice, in preference to SVGs.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
I have a comment that I'd like to have your thoughts about this. We all know about the competitive flow, but I think that looking at it just simply from the perspective of the percent stenosis may be somewhat simplistic. I thought about it, and I think that the two important elements in considering competitive flow is the size of the target vessel, the maximum percent stenosis, vis-a-vis the size of the conduit. And I think that the critical element to consider is the cross-sectional area of the conduit vis-a-vis the cross-sectional area at the site of the maximum stenosis. Because I think that at 50% stenosis and a 4-mm RCA [right coronary artery] with a 2-mm radial artery is a very different phenomenon than a 50% stenosis and a 2-mm RCA and the same size 2-mm radial artery. So I'd like to have your comments on that. Thank you.
DR TATOULIS: Thank you for your comment and question. I agree entirely. The problem of competitive flow is something that we have learned. We have been using radial artery grafts for 14 years now, and so we have learned that lesson and that is why we avoid that situation. Hence, there were very few right coronary arteries grafted.
To take your idea further, we should consider the residual internal diameter of the native coronary artery at the stenosis, because, as you say, a 50% stenosis in a 4- or 5-mm right coronary artery is different to a 50% stenosis in a 2-mm coronary artery. So I would propose the concept of a "residual internal diameter" of the native coronary as a better, or more absolute objective measure, and then matching it to the size of the conduit. Unfortunately, in setting up our database, we had not forethought all those issues and so we could not analyze this method of determining the residual lumen diameter as a measure of native coronary artery stenosis and its effect on competitive flow and radial artery patency. I agree entirely with your comments.
DR ALFONSO CHISCANO (San Antonio, TX): I had the pleasure to be a fellow with Brian Buxton in Houston. I hope he's doing fine. I don't see him around. My question is: Please say "Hi" to him. Second, the proximal anastomosis of the radial artery to the ascending aorta sometimes is a pain in the neck. Can you tell us the little tricks that you use? Thank you.
DR TATOULIS: Thank you. My colleague, Brian Buxton, and I have worked together since 1974. He is currently sailing his yacht around the Pacific! The overwhelming majority of radial arteries were used as aortocoronary grafts, from the ascending aorta. We don't have any special tricks. We use a single cross-clamp technique for all distal and proximal anastomoses. We don't use side-biting clamps, and that makes access to the ascending aorta and construction of the proximal anastomoses easier. We use a 3.5-mm punch. The proximal radial artery is quite a large size, usually 3.5 to 4 mm in diameter and very robust. We use continuous 6-0 polypropylene to construct the proximal anastomosis. We ensure that the radial artery is (pharmacologically) dilated and has a large hood. If we do use a vein graft in an older patient as well, the hood of the radial artery is often larger than the hood of the vein graft at the proximal anastomosis.
DR ALI KHOYNEZHAD (Omaha, NE): Dr Tatoulis, thank you for presenting your excellent results. They are very encouraging and add up to the growing literature supporting increase use of the radial artery. I have advocated and used the radial artery in my practice for the many reasons summarized today. I also think there are important factors that influence the long-term patency of the radial artery, such as postoperative vasodilatory medication and the use of radial artery in certain patient cohorts, or grade of target vessel stenosis. For example, I am concerned about using the radial artery on patients who are active smokers or have poorly controlled diabetes. I wonder if you have looked into these risk factors and if that had affected the patency of your conduits.
Also, please comment on your choice of vasodilatory and spasmolytic medication in immediate postoperative and later period. Thank you.
DR TATOULIS: Thank you for those questions. I cannot answer your first question. We have not examined in detail our patients with either type 1 or type 2 diabetes in terms of either patency or survival. That is the next project when we evaluate the 15-year results, so I will take that on board.
With respect to spasm prophylaxis, we use 1 mg/mL of papaverine in arterial heparinized blood at 37°C, and we inject 2 to 3 mL intraluminally in the radial artery (distally) and then clip the distal end while it is beating against itself for several minutes in the forearm before we secure the proximal end. We then store it in the same solution until grafting. We use a nitroglycerin infusion or milrinone infusion intraoperatively and for 24 hours to achieve maximal dilatation of all the arterial conduits and also of the coronaries as well. We then use amlodipine (5 to 10 mg, orally), which is a once-daily calcium channel blocker, for 6 months, based on the premise that we have seen at least four instances of radial artery spasm on angiography in the months early after surgery. And as I showed you, even one example at 11 years post-op, which was a very big surprise!
DR DARRYL HOFFMAN (New York, NY): Like you, we have been fans of the radial artery and have followed your lead. In the 12 years we have done 1600, and the last 1000 have been endoscopically harvested. We followed the standard protocols, and we have recatheterized 200 patients and have found similar results to yours.
I would like you to reference, since every patient is his own control, if possible, your patency rates for the vein grafts, left and right mammaries, since we found the radial to be superior to an admittedly small number of free right internal mammary artery grafts used in the same group of patients.
DR TATOULIS: Thank you. Yes, we do have that information. At 5 years, there is no difference between the right internal thoracic artery and the radial artery, both with patency rates of 90%. To go beyond that would be difficult because, as you saw, our mean follow-up is only 48 months for the radial, whereas it is almost 10 years for the right internal thoracic artery.
My colleague, Brian Buxton, is performing a randomized trial, as you probably know, in younger patients randomizing the second graft to the right mammary vs the radial artery. Thus far for the free right internal thoracic artery and the radial artery, the results are identical, but that is only for 5 years. So, hopefully, we will have the longer-term answer over the next 5 years. My personal feeling is the right internal thoracic artery is probably a better conduit if applied in the correct manner, keeping in mind that it is also subject to competitive flow.
In the 318 radials studied 5 to 10 years postoperatively, their patency was significantly superior to that of vein grafts.
DR FRIEDRICH MOHR (Leipzig, Germany): Jim, can I give you a patient? A patient who had 3-vessel disease and had a drug-eluting stent in the right, which has restenosed by 60%, PCI [percutaneous coronary intervention] was done 6 months ago. And you need to do a radial artery. Would you consider to snare down the right coronary artery?
DR TATOULIS: I have never done that. The scenario you outlined represents a difficult and controversial issue. I would use the right internal thoracic artery to the right coronary artery if the patient were young. In an older patient (> 70 years), I would use a vein graft. If you really press me, if I had to use the radial as the only available graft, then, possibly I would make that stenosis a bit tighter to minimize competitive flow.
DR MOHR: There are several surgeons who routinely do that just to avoid competitive flow.
DR TATOULIS: Exactly right.
DR MARK BURLINGAME (Lancaster, PA): Do you use any special preservation techniques from the time of harvest to the time of implantation, since this is a rather thick-walled vessel?
DR TATOULIS: Not specifically. We harvest the radial artery. We inject the papaverine/blood solution intraluminally and clip it at the distal end and let it beat in situ for 5 minutes before we disconnect the proximal end. Some surgeons leave the radial artery in situ in the forearm until they actually use it. For practicality and to minimize infection, we remove the radial and close the forearm. But we do store the radial in a solution of arterial heparinized blood at 37° with 1% papaverine, for approximately 15 to 30 minutes until we use it. That is the only storage technique that we use.
Doing that, we have shown in the lab that the radial artery maintains its reactivity and nitric oxide production, so we are quite happy with that type of storage, as it preserves the radial artery endothelial and wall function.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
