Radial Artery Free and T Graft Patency as Coronary Artery Bypass Conduit Over a 15-Year Period
Background—The radial artery is often used as the second arterial graft for coronary artery bypass grafting. Little is known about the differences in long-term patency between radial free and T grafts. This study was performed to determine long-term radial artery patency over a 15-year period.
Methods and Results—Radial arteries were used as free grafts or T grafts for coronary artery bypass grafting over a 15-year period. Patients were contacted to determine if postoperative cardiac catheterization was performed and examination of any reports and films was performed. Grafts were graded as patent, luminal irregularity, or occluded. Each sequential graft was counted separately. Between September 1993 and December 2008, 13 926 patients underwent isolated coronary artery bypass grafting and 3248 patients had at least one radial artery graft used as a conduit. Catheterizations were performed at a mean of 7.4±3.8 years (range, 3 days to 14.4 years) on 372 radial artery grafts (103 free and 269 T) in 215 patients. Kaplan-Meier freedom from occlusion for radial free and T grafts at 1 and 10 years was 97.1% and 75.4% and 99.6% and 62.9%, respectively (P=0.146 free versus T). Kaplan-Meier survival to 15 years was not statistically different between free and T graft patients (P=0.5).
Conclusions—In 215 patients with postoperative catheterization after coronary artery bypass grafting with a radial artery graft, radial free and T grafts had similar and acceptable long-term patency to support their use as a coronary artery bypass graft conduit.
The radial artery is often used as the second arterial graft for coronary artery bypass grafting (CABG). Its use remains controversial due to perceived issues by individual surgeons. Proponents of the use of the radial artery note its ease of harvest, absence of development of atherosclerosis and intimal hyperplasia seen in saphenous vein grafts, the benefit of reduced sternal and leg infections (particularly in obese and diabetic patients), its ideal length, and its superior patency (particularly in the long term) to saphenous vein grafts.1–3 In contrast, critics of its use stress its relatively smaller size in women, its ability to spasm, its reduced patency in target vessels with competitive flow, warn of the risk of hand complications, and stress the speed and ease in which saphenous vein may be harvested.
Mid- to long-term radial artery graft patency (≥5 years) has been reported to be between 70.7% and 98.3%.3–12 Reports have noted patency differences with respect to target vessel location and target vessel stenosis; however, little is known about the differences in long-term patency between radial artery free grafts and radial artery T grafts.
This study was designed to review radial artery graft patency over a 15-year period and to specifically compare patency of radial artery free and T grafts.
With Washington University Institutional Review Board approval, all patients undergoing isolated CABG at Barnes Jewish Hospital (September 1993 and December 2008) and Christian Northeast Hospital (September 1993 and December1998) were queried for the use of the radial artery as conduit. During this timeframe, one surgeon almost exclusively performed radial artery T grafting using the radial artery originating from the left internal mammary artery using cardiopulmonary bypass. Other surgeons typically performed radial artery free grafts with or without cardiopulmonary bypass.
All patients in whom the radial artery was used as a conduit for CABG as a free (anastomosed to the ascending aorta, the hood of a vein graft anastomosed to the aorta, or anastomosed to the descending aorta) or T graft (originating from the left internal mammary artery) were contacted by telephone or mailed survey to determine if postoperative cardiac catheterization was performed. In addition, electronic medical records were queried for postoperative cardiac catheterization. In the majority, catheterizations were symptom-directed but also included preoperative clearance for noncardiac surgery or other significant stressors. Cardiac catheterization reports (and films when appropriate in 2.8% of patients) were collected and reviewed by 2 surgeons (H.B.B. and J.S.L.). All bypass grafts were graded as patent (0%–50% stenosis), presence of luminal irregularity (51%–99% stenosis), or occluded (including grafts with “string sign”). Each sequential graft was graded separately. Mortality was updated by the Social Security Death Index as of November 15, 2011. Mean patient follow-up time was calculated from the date of the patient's initial surgical procedure until the date of last confirmed contact or date of death. Mean graft follow-up time was determined from the date of the surgical procedure until the date of the cardiac catheterization procedure that confirmed either the patency or occlusion of the specific graft.
Continuous variables are expressed as mean±SD. Categorical variables are expressed as frequencies and percentages. Categorical outcomes were compared by using either the χ2 or Fisher exact test, and continuous outcomes were compared by using the t test for means of normally distributed continuous variables and Wilcoxon rank-sum nonparametric test for skewed distributions. Kaplan-Meier method was used to estimate freedom from graft occlusion and survival. Estimates were compared using the log-rank test when stratified between free and T grafts. All data analyses were done using SPSS (SPSS 11.0 for Windows; SPSS Inc, Chicago, IL).
Between September 1993 and December 2008, 13 926 patients underwent isolated CABG at Barnes Jewish Hospital and Christian Northeast Hospital (September 1993 and December 1998). A total of 3248 patients (23.3%) had a radial artery used for at least one bypass graft. Of these patients, 215 patients (6.6% of those patients with a radial artery graft) had a postoperative cardiac catheterization. A total of 372 radial artery grafts (103 free grafts and 269 T grafts) were used in these in 215 patients. The mean time from CABG until time of catheterization was 7.5±3.5 years and 6.6±4.0 years for radial artery T and free graft patients, respectively (range, 3 days to 14.4 years). Kaplan-Meier freedom from occlusion for radial free and T grafts at 1, 3, 5, and 10 years was not statistically different (P=0.146 free versus T; Figure 1).
Demographics for radial artery T and free graft patients are listed in Table 1. More patients in the radial artery T graft group had preoperative renal failure, 3-vessel coronary artery disease, and use of cardiopulmonary bypass. Only 4 patients with free radial artery grafts had CABG without cardiopulmonary bypass. More patients in the radial artery free graft group had hypertension and previous cardiac surgery (Table 1).
The majority (65.2%) of radial artery T graft patients received 2 grafts with the radial artery, whereas the majority (74.0%) of radial artery free graft patients received one graft with the radial artery. The majority (88.4%) of patients received one internal mammary artery graft (left internal mammary artery [IMA] in 185 and right IMA in 5 patients). A total of 130 radial artery T graft patients received one IMA and 59 radial artery free graft patients received one IMA. A total of 15 patients (8 radial artery T graft patients and 7 radial artery free graft patients) received bilateral IMA grafts and 163 patients received all arterial revascularization (98.5% of T graft patients and 36.7% of free graft patients).
The patency of radial artery T versus free grafts at the time of angiogram as well as graft patency for each sequential anastomosis is listed in Table 2. Overall radial artery graft patency was 69.3% at the time of angiogram with overall patency of T grafts of 65.8% and that of free grafts 78.6%. Radial artery T graft patency declined with each successive anastomosis, whereas radial artery free graft patency remained similar or improved.
Radial artery graft patency by target vessel is listed in Table 3. Patency was highest for radial artery grafts to the diagonal followed by the obtuse marginal branches and then the right coronary artery branches. There were only 10 radial artery grafts (3 T and 7 free) to the left anterior descending artery with patency of 80.0%, and there were only 14 radial artery grafts (9 T and 5 free) to the ramus with patency of 64.3%.
The mean patient follow-up time (time from initial surgical procedure until the date of last confirmed contact or date of death) was 14.0±3.3 years and 12.7±3.3 years for radial artery T and free graft patients, respectively. Kaplan-Meier survival for radial free and T graft patients was not statistically different (P=0.5 free versus T; Figure 2).
Patency of T Versus Free Grafts
The present study uniquely documents long-term follow-up (7.5±3.5 years) in a large cohort of patients (N=138) with 269 radial artery T graft segments. In this study, no significant difference was noted between radial artery T graft and radial artery free graft freedom from occlusion. These findings are consistent with others as well as Yie and colleagues who reported that the methods used for proximal anastomosis failed to affect radial artery patency by univariate analysis with follow-up angiography at a mean of 32.7 months in a group of 123 patients.13,14 The majority of radial artery T grafts were used as sequential grafts to 2 targets (N=90), and the majority of radial artery free grafts were used as single grafts to one target (N=57). Aortocoronary and composite T radial artery grafts have inherently different flow characteristics. Patency of the radial artery T graft is dependent on flow from the left IMA (and subclavian artery), the anastomosis to the left IMA, the resistance of the sequential grafts, the target vessel quality, and target vessel runoff. Patency of the radial artery free graft is dependent on the anastomosis to the aorta, the target vessel quality, and target vessel runoff.
Patency in Sequential Grafts
Each sequential radial artery graft was evaluated independently for patency in this study. Radial artery free graft patency remained similar or improved with each sequential anastomosis. Similar findings were reported by Schwann and colleagues15 who noted similar patency for radial artery sequential segments in radial artery grafts primarily anastomosed to the aorta. In addition, Achouh and colleagues12 reported a trend (P=0.08) toward better patency when radial artery grafts were used as sequential grafts when compared with single grafts.
In contrast, the radial artery T graft patency was superior in the first segment and declined in the second and third sequential anastomoses in the present study. In the majority of cases, the first target for the radial artery T graft was a branch of the left circumflex and the second target was a branch of the right coronary artery. Therefore, these results may reflect reduced patency noted in the right coronary artery distribution (and perhaps due to a degree of target vessel stenosis or reduced runoff area) as noted by other authors.2,14
Patency Based on Target Vessel
Patency for all radial artery grafts was best in grafts to the diagonal followed by the obtuse marginal and then the right coronary artery. In radial artery T grafts, patency was equivalent in diagonal and obtuse marginal targets and superior to right coronary artery targets. Patency in radial artery free grafts was best with a diagonal target and superior to both obtuse marginal targets and right coronary artery targets, which were similar.
Similar variable patency rates have been documented by other investigators suggesting radial artery patency is best when grafted to the left anterior descending artery > diagonal > obtuse marginal branches > right coronary artery branches.2,12–14 Others have suggested that the right coronary artery and its branches may be perceived as smaller targets with smaller runoff territory when compared with the left circulation (particularly in nonright-dominant patients).11 In contrast, in a randomized trial (Radial Artery Patency Study) of saphenous vein versus radial artery graft, Desai and colleagues found no difference in patency based on the bypassed vessel.16,17 Similarly, Hadinata and colleagues11 found no difference in absolute patency rates between radial artery and saphenous vein grafts to the right coronary artery territory suggesting no benefit of the arterial conduit to this territory.
Longevity of Patent Radial Grafts
Freedom from occlusion for radial artery T and free grafts was not significantly different at 10 years (Figure 1). The freedom from occlusion remained stable and acceptable out to 10 years and suggests continued use of this arterial conduit for CABG. Similarly, patency remained consistent in patients followed over a decade by Tatoulis and colleagues2 who reported 92.5% patency of radial artery grafts in place for >5 years and 92.5% patency in grafts in place for >7 years. The authors also reported a lack of atherosclerosis in radial artery grafts. Indeed, the real benefit of the radial artery may be in the long term in which low attrition rates may be similar to other arterial grafts as longer-term data become available.
In a recent meta-analysis, 35 studies reporting graft patency for radial artery and saphenous vein grafts were reviewed and categorized by mean follow-up (early patency was ≤1 year, medium-term patency was 1–5 years, and long-term patency was >5 years).18 Early radial artery patency was similar to vein graft patency, medium-term follow-up demonstrated deterioration in vein graft patency, and radial artery patency was superior in long-term follow-up. The long-term radial artery patency ranged from 85% to 94.9% and the saphenous vein patency ranged from 64.5% to 91.7%; the pooled OR for 7 studies indicated better patency with the radial artery conduits.18 The authors suggested a nonhomogeneous comparative patency of radial arteries and saphenous vein grafts over time and recommended preferential radial artery use over saphenous vein use for CABG conduit in all patients who have a life expectancy >5 years. This meta-analysis lends support for the durability of radial artery grafts after the initial postoperative period.
Survival was not statistically different in radial artery T and free graft patients up to 15 years despite different grafting strategies. It is well established that use of the left IMA improves long-term survival after CABG as does the addition of the right IMA.19 In addition, complete arterial revascularization ideally provides excellent long-term patency of CABGs. When comparing survival between the radial artery T and the radial artery free graft groups, it is therefore important to note that all arterial revascularization was accomplished in 134 radial artery T graft patients (98.5%) and in 29 radial artery free graft patients (36.7%) in this study. Only 15 patients (8 radial artery T graft and 7 radial artery free graft) had revascularization using both IMAs. In addition, all of the radial artery T graft patients received at least one IMA, whereas only 85.7% of radial artery free graft patients received at least one IMA graft.
This retrospective study is limited due to the lack of information on postoperative medication use (aspirin, β-blockers, statin use, tobacco use, Plavix, etc). Similarly, the routine use of calcium channel blockers was not done or often stopped by referring physicians at the first follow-up visit. Also, preoperative vessel stenosis, conduit quality, and target vessel size and runoff are not known, all of which may influence conduit patency. In addition, the study spans multiple years and advances in operative and postoperative care make comparisons difficult.
The Kaplan-Meier method for estimating freedom from occlusion is not ideal because actual time to graft failure is unknown.20 Symptom-directed angiograms will overestimate graft failure rates, miss asymptomatic graft occlusions, and miss patients who have silent ischemia.20
We thank the tireless efforts of Tina Burmeister and Patricia Buckley for data collection and management.
- © 2012 American Heart Association, Inc.
- Barner HB
- Collins P,
- Webb CM,
- Chong CF,
- Moat NE
- Possati G,
- Gaudino M,
- Prati F,
- Alessandrini F,
- Trani C,
- Glieca F,
- Mazzari MA,
- Luciani N,
- Schiavoni G
- Cameron J,
- Trivedi S,
- Stafford G,
- Bett N
- Zacharias A,
- Habib RH,
- Schwann TA,
- Riordan CJ,
- Durham SJ,
- Shah A
- Achouh P,
- Isselmou KO,
- Boutekadjirt R,
- D'Alessandro C,
- Pagny J,
- Fouquet R,
- Fabiani J,
- Acar C
- Yie K,
- Na C,
- Oh SS,
- Kim J,
- Shinn S,
- Seo H
- Desai ND,
- Naylor CD,
- Kiss A,
- Cohen EA,
- Feder-Elituv R,
- Miwa S,
- Radhakrishnan S,
- Dubbin J,
- Schwartz L,
- Fremes SE
- Athanasiou T,
- Saso S,
- Rao C,
- Vecht J,
- Grapsa J,
- Dunning J,
- Lemma M,
- Casula R