- Email: [email protected]
The radial artery in coronary surgery: a 5-year experience—clinical and angiographic results
The Radial Artery in Coronary Surgery: A 5-Year Experience—Clinical and Angiographic Results James Tatoulis, FRACS, Alistair G. Royse, FRACS, Brian F. Buxton, FRACS, John A. Fuller, FRACP, Peter D. Skillington, FRACS, John C. Goldblatt, FRACS, Robin P. Brown, FRACS, and Michael A. Rowland, FRACS Department of Cardiothoracic Surgery, Royal Melbourne and Epworth Hospitals, The University of Melbourne, Melbourne, Victoria, Australia
Background. The radial artery (RA) has been used extensively by us as a way of reducing the use of the saphenous vein. It has been hoped that the RA will maintain greater late patency than the saphenous vein. We evaluated our initial 5-year experience with the RA in coronary surgery. Methods. We studied 6,646 consecutive patients who had a single RA (4,872), or bilateral RA (1,774), as coronary grafts, from June 1995 to June 2000. Angiograms were performed mostly in symptomatic patients, or as part of a research project in asymptomatic patients. Results. The patients’ mean age was 65.1 years; 23% had diabetes, 14% had unstable angina, and 42% had prior myocardial infarction. An average of 3.3 grafts per patient were performed, 87% from arterial conduit. Con-
duits used were RA (8,420), left internal thoracic artery (6,296), and right internal thoracic artery (1,076). Operative mortality occurred in 58 (0.9%) patients, stroke in 92 (1.4%), deep sternal infection in 97 (1.4%), reoperation for hemorrhage in 56 (0.9%), and myocardial infarction in 52 (0.8%). Peak mean postoperative creatine kinase MB (CKMB) was 16.5 IU/L. Two patients developed fingertip ischemia. Postoperative angiographic RA patency was 90.2% (333 of 369 distal anastomoses). Conclusions. Good early clinical and angiographic results can be achieved by using the RA in coronary surgery.
T
patients underwent isolated CABG in which one or both RAs were used. These patients form the study group. Patients having combined procedures (eg, valve and CABG) were excluded from this analysis to obtain a uniform study group. The patient characteristics are detailed in Table 1.
he incidence of failure of saphenous vein grafts in the late period [1, 2] prompted our progressive expansion of arterial conduit use. Initially, we used bilateral internal thoracic artery (ITA) [3]. However, it was commonly noted that the right ITA would not reach distal branches in the circumflex and right coronary territories, and these areas remained a challenge in terms of arterial revascularization. After the revival of radial artery (RA) use by Acar and associates [4], we used the RA [5], and subsequently made more efficient use of the conduit with sequential grafting [6] and composite grafting [7]. Others have reported similar operative strategies [8 –10]. The purpose of this study was to review our initial experience, more than 5 years, in isolated coronary artery bypass grafting (CABG) surgery where the RA was used.
Material and Methods Patients We began to use the RA as a conduit in CABG in December 1994, and within 6 to 18 months, its use became routine [5]. From June 1995 to June 2000, 6,646 Presented at the Thirty-seventh Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 29 –31, 2001. Address reprint requests to Dr Tatoulis, Suite 28, Private Medical Centre, Royal Melbourne Hospital, Parkville, Victoria 3050, Australia; e-mail: [email protected].
© 2002 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2002;73:143– 8) © 2002 by The Society of Thoracic Surgeons
Assessment of Ulnar Collateral Circulation All patients underwent a modified Allen’s test preoperatively. Both the RA and ulnar arteries were compressed while the hand was rendered ischemic by five strong fist clenching and relaxing actions for more than 30 seconds. The ulnar artery was then released. Reperfusion of the thenar eminence, thumb, and index finger within 5 seconds was considered excellent, within 10 seconds satisfactory, and beyond 10 seconds not acceptable. Details of technique have been previously reported [5]. Finger plethysmography or ultrasound examination was used sometimes as part of a research protocol. Contraindications to RA use were abnormal Allen’s test, prior trauma or surgery to the relevant upper limb, known subclavian or brachial artery stenosis, Raynaud’s phenomenon, scleroderma, and RA calcification.
Graft Procurement The left RA (nondominant) was the preferred conduit. After sternotomy, the left internal thoracic artery (LITA) and left RA harvest proceeded simultaneously. If both RAs were used, these were harvested first, the forearms 0003-4975/02/$22.00 PII S0003-4975(01)03290-8
144
TATOULIS ET AL RADIAL ARTERY IN CORONARY SURGERY
Table 1. Patient Demographics Characteristic Age (years)a Male/female Diabetes Angina classd III or IV Unstable anginac Left main stenosis ⬎ 50% Prior Q wave myocardial infarction Left ventricular ejection fraction ⬍ 0.5 Reoperation
No. of Patients (n ⫽ 6,646) 65.1 ⫾ 9.9 (29 – 89)b 5,182:1,464 (78:22) 1,547 (23) 3,455 (52) 916 (14) 864 (13) 2,778 (42) 238 (4) 434 (7)
a
b Standard deviation and range shown. Numbers in parentheses are c Society of Thoracic Surgeons definipercentages (except for age). d Canadian Cardiovascular Society Classification for tions used. angina.
closed and placed by the patient’s side, and covered with a further large sterile drape before sternotomy. In long operations (combined, reoperations), a small suction drain was sometimes placed.
Radial Artery Harvesting The distal one-third of the RA was usually exposed first to exclude calcification. The RA was harvested using a combination of low-power cautery [5] or sharp dissection without much cautery [11], dependent on surgeon preference. Small vascular clips were used to secure branches. Care was taken to avoid injury to the superficial radial and the lateral cutaneous nerve of the forearm.
Antispasm Prophylaxis Papaverine 1 mg/mL in heparinized arterial blood was introduced intraluminally via the distal end of the RA (5 mL) with a blunt-ended vascular needle, and the distal end of the RA was secured with a vascular clip. The RA was allowed to pulsate for 5 minutes before the proximal end was secured, and the RA was placed in an identical solution until its use. The ITAs similarly had intraluminal papaverine and a papaverine-soaked gauze placed around until use. Intravenous nitroglycerine (30 to 100 g/min) or Milrinone (0.1 g/kg/min) (SanofiSynthelabo, Sydney, Australia) were commonly used perioperatively for 24 hours according to surgeon’s preference. Oral calcium channel blockers (amlodipine or felodipine) were commenced postoperatively, and continued for a variable period (usually 6 months) after discharge.
Patient Monitoring Where bilateral RAs were harvested, a femoral arterial monitoring line was used. Swan-Ganz catheters, and transesophageal echocardiography were used routinely in all patients.
Surgical Strategy Grafting varied according to surgeon’s preference. Initially, total arterial revascularization was most frequently used in younger patients (⬍ 65 years), but later, all patients were considered suitable. Sequential grafting
Ann Thorac Surg 2002;73:143– 8
became increasingly popular due to the additional length and flexibility afforded by the RA. Similarly, composite “Y” grafting became more popular. The most common patterns of grafting were the LITA to the left anterior descending (LAD), the right internal thoracic artery (RITA) (if used) to the proximal circumflex or posterior descending artery (PDA), and the RAs to the distal circumflex, left ventricular branch of the right coronary artery (RCA), or PDA. Four thousand eight hundred seventy-two patients had one RA used, and 1,774 patients had both RAs harvested. The arterial conduits used are detailed in Table 2. Eighty-seven percent of all distal anastomoses were performed with arterial grafts, and the rest with saphenous vein.
Surgical Procedure This has been consistent for the duration of the study. Cardiopulmonary bypass was established by cannulation of the ascending aorta and right atrium. The patient was cooled to 34°C, with flows of 2.5 L/min/m2 and a mean blood pressure of 80 mm Hg maintained. Myocardial protection with antegrade (700 mL) and retrograde (300 mL) blood cardioplegia at 20°C was used to maintain a myocardial temperature of 25°C. Further doses of retrograde cardioplegia (300 mL) were given after each anastomosis. An aortic vent was used. If there was aortic regurgitation, a left ventricular vent via the right superior pulmonary vein was employed. All anastomoses were completed during one period of cross-clamping, with the use of a partial occlusion aortic clamp being rare. A “hot shot” dose (1,000 mL) of aspartate-enriched warm blood cardioplegia was usually given before cross-clamp release. If a Y graft was constructed, the proximal LITA/RA anastomosis was completed after heparinization but before bypass. The distal anastomoses were performed with continuous 7/0 polypropylene. The proximal anastomoses were performed directly onto a 3.5-mm punched opening in the ascending aorta with continuous 6/0 polypropylene. Sequential distal anastomoses were performed in a manner to ensure the best lie of the graft [7, 11]. Preference was for parallel sequential anastomoses if possible. In general, the LITA to LAD anastomosis was performed last. Pericardial cold slush was not used. The intraoperative details are shown in Table 3. Table 2. Conduits Used in 6,646 Patients Conduit LITA RITA RA Single RA Bilateral RA Total Total anastomoses Total arterial anastomoses
Patients
Conduits
n 6,296 1,076 8,420
4,892 1,774 6,646
4,872 3,548 8,420
LITA, RITA ⫽ left and right internal thoracic artery; artery.
21,931 18,079 (87%) RA ⫽ radial
Ann Thorac Surg 2002;73:143– 8
Data Collection and Analysis
Table 4. Operative Results
All data were collected prospectively and entered into a computer database program. Intraoperative data were entered at the time of operation. Each patient was reviewed and data were collected daily until hospital discharge. All surviving patients were reviewed at 30 days or more after operation by office, outpatient clinic, or cardiologist visit. The Society of Thoracic Surgeons definitions were used. Values are reported as mean ⫾ standard deviation and range. Percentages are given where appropriate. Comparisons between proportions were by the Fischer’s exact test (statXact 3.1; Cytel Software Corp, Cambridge, MA). A p value of less than 0.05 was considered significant.
Variable
Postoperative angiography was performed as part of an ethics-approved study (Royal Melbourne Hospital Research and Ethics Committee), or in response to symptoms [12]. Grafts with string signs were considered not patent.
Results Operative Mortality Fifty-eight patients died in hospital or within 30 days of CABG, or of associated complications. The operative mortality was 0.9%. The majority of the deaths were from cardiac causes or from multiorgan failure associated with low cardiac output.
a
n 58 (0.9)a 52 (0.9) 16.5 ⫾ 22 80 (1.2) 92 (1.4) 56 (0.9) 79 (1.4) 21 (0.3) 14 (0.2) 2
Numbers in parentheses are percentages. IABP ⫽ intraaortic
Fifty-six patients (0.9%) required return to the operating room for control of postoperative hemorrhage. In no instance was the source of bleeding from an RA branch. Ninety-seven patients (1.4%) developed a deep sternal wound infection, defined as any sternal infection that required intravenous antibiotics, or a further operation on the sternum. The most common pathogen was Staphylococcus aureus (92 patients, 95%). The mean time to extubation was 8.5 ⫾ 8.6 hours (range 0 to 226 hours), and the mean hospital stay was 7.7 ⫾ 5.4 days (range 4 to 126 days).
Hand and Forearm Complications
Perioperative Morbidity The perioperative major morbidity was low (Table 4). Fifty-two patients (0.9%) developed perioperative myocardial infarction by electrocardiogram (new Q-wave) and enzyme criteria (creatine kinase MB /cardiac isoenzyme [CKMB]). The mean postoperative peak level of CKMB was 16.5 ⫾ 22 IU/L (normal laboratory values, 0 to 25 IU/L). In 80 patients (1.2%), an intraaortic balloon pump was used for vascular support in the perioperative phase. In 21 (25%), the intraaortic balloon pump had been placed prebypass for unstable hemodynamics or because of poor left ventricular function. Ninety-two patients (1.4%) developed a stroke, defined as any clinically detectable neurologic abnormality producing a motor, speech, or sensory deficit. Six patients died after a severe stroke, and 7 died from intraabdominal catastrophes (five mesenteric ischemia, two abdominal aortic aneurysms). Table 3. Intraoperative Data for 6,466 Patients Variable
n
Aorto-coronary RA Y grafts (LITA-RA, RA-RA) Distal anastomoses per patient Aortic cross-clamp time (min)a
5,583 (84%) 1,063 (16%) 3.3 ⫾ 1.0 65 ⫾ 25
All distal and proximal anastomoses during one period of cross-clamp.
LITA ⫽ left internal thoracic artery;
Operative mortality Perioperative myocardial infarction Maximum CKMB, IU/L IABP use Stroke Reoperation for hemorrhage Deep sternal infection Forearm infection Forearm hematoma (drainage) Fingertip ischemia
CKMB ⫽ creatine kinase cardiac iso-enzyme; balloon pump.
Postoperative Angiography
a
145
TATOULIS ET AL RADIAL ARTERY IN CORONARY SURGERY
RA ⫽ radial artery.
Objective upper limb morbidity was minimal. Twentyone (0.3%) infections required intravenous or oral antibiotics, or some form of operative drainage. All except one were due to S aureus. Fourteen patients (0.2%) required evacuation of a forearm hematoma, most commonly between the 3rd and 5th postoperative days. Two patients developed fingertip (index finger) ischemia. Both had scleroderma, Raynaud’s phenomenon, severe peripheral vascular disease, and were early in our experience [13]. There was no difference in complications between patients with excellent (⬍ 5 seconds) and those with satisfactory (5 to 10 seconds) ulnar collaterals.
Radial Artery Graft Patency Two hundred seventy-one patients from both institutions underwent RA coronary graft angiography, a mean of 14.4 ⫾ 10.4 months postoperatively (range 1 to 35 months). Early patency data have been previously published [11–14]. Two hundred eighty RA conduits were studied with 369 RA to coronary anastomoses. RA segments to 36 (10%) anastomoses were either occluded or had string signs. Three hundred sixty-three of 369 (90.2%) anastomoses were widely patent (Table 5). Patency of the RA in the LAD and circumflex territories was 90.7% and 92.5%, respectively. Patency was lowest in the right coronary artery territory (86.7%), and is consistent with previous reports for ITA and RA [9, 12]. Additionally, string signs were associated with RA grafts to
146
TATOULIS ET AL RADIAL ARTERY IN CORONARY SURGERY
Ann Thorac Surg 2002;73:143– 8
Table 5. Comparison of Radial Artery Graft Patency Coronary Territory Left anterior descending Circumflex Right coronary artery Total
Anastomoses
Patent
Patency (%)
p Valuea
33
30
90.7
0.769b
201 135
186 117
92.5 86.7
0.725c 0.093d
369
333
90.2
a
b Fisher’s exact test. Compared with right coronary artery. d Compared with left anterior descending. Compared with circumflex.
c
coronary arteries of lesser degrees of coronary stenosis [12, 15].
Comment At the inception of coronary artery bypass surgery, aortacoronary saphenous vein grafting (SVG) was a major advance in the management of coronary artery disease. However, its full potential was often limited by the development of intimal hyperplasia and occlusive disease due to progressive conduit atheroma or thrombosis. Loop and associates [2] clearly demonstrated the clinical and angiographic superiority of the LITA grafted to the left anterior descending artery compared with SVG. Bilateral ITA grafting did not show an early advantage, mainly due to the concomitant use of SVG in both single and bilateral ITA groups. Recent studies suggest improved outcome with bilateral ITA [16, 17]. However, many surgeons may have been concerned by longer operation times or the potential for increased morbidity in the obese, diabetic, or those with chronic airways disease. Additionally, the RITA often cannot reach the distal circumflex and PDA territories. The revival of the RA has made available an additional arterial conduit, which can be used in addition to the ITA to achieve predominantly or total arterial coronary revascularization. Significant advantages for the RA are potential availability of two in most patients, excellent length, and excellent handling characteristics for constructing anastomoses. It is also resistant to kinking; and avoidance of leg wounds allows rapid ambulation [11, 13]. As well as being available for routine use, the RA is extremely important when there is conduit shortage: stripped saphenous vein, reoperations, peripheral vascular disease, or in patients with chronic airways disease or obese diabetics where RITA use may be avoided. Conversely, the RA may have distal calcification, or distal fibrosis from prior cannulation [11]. With a combination of the ITA and RA, we were able to achieve a very high rate (87%) of total arterial revascularization. Objective forearm and hand morbidity (hematoma and infection) was very low, at less than 1%. Discomfort, pain, paraesthesia, and wound tightness are more common complaints but settle with time [5]. If only one RA is to be used, the nondominant arm (usually left) should be preferred, as the efficiency of the surgery is enhanced
and there is less potential for disturbance of function. There was no relationship between the degree of ulnar collateral flow and postoperative forearm complications or function. RA spasm is not commonly seen during surgical procedure. Intraoperative prophylactic antispasmodic agents varied according to surgeon preference. All surgeons used intraluminal papaverine during harvest as previously used for the LITA. No topical “ice slush” was used in any of these patients. This may also protect against RA spasm. Most also administered low-dose glycerine Trinitrate, as was previously used for the LITA. Some surgeons preferred to administer intravenous Milrinone for its potent vasodilating effect. Calcium channel antagonists were normally prescribed postoperatively, and these were usually ceased within 3 to 6 months. These strategies have not been tested scientifically. The clinical role of postoperative oral calcium antagonists is not defined. However, isolated instances of discrete postoperative RA spasm have been reported [4]. RA calcification can be problematic, though rare [11]. When it occurs, it is more pronounced distally, on the superficial surface, and may be so severe that the RA is rigid and cannot be used. Difficulties include technical problems and hemostasis of the distal anastomosis, potential disruption of calcified plaques, intima, and possible thrombosis. Minimal calcification where the RA wall remains supple is acceptable. There are no data relating presence or absence of RA calcification to outcome. Overall, RA patency was 90.2% beyond 1 year and is similar to other reports [18 –20]. We have also found a lower patency when the RA was anastomosed to coronaries with lesser stenosis [12]. The patency findings are similar to those with the free RITA grafts [3]. The moderately stenosed coronary artery at the time of coronary operation poses a dilemma with respect to conduit choice. The patency was least in RA grafts to the RCA territory. Others have also noted a lower patency in this territory [9, 12, 15]. Relevant factors may be a higher incidence of coronary wall disease, the length of the RA conduit required, steal into the circumflex territory by sequential anastomoses, and potential for competitive flow in the setting of moderate (lesser) stenoses in large dominant RCA systems. Intuitively, it is expected that those RAs that are patent and are functioning well, beyond 1 year, should continue to do so into the long term. As a proportion of RA angiograms were performed for possible cardiac symptoms, the patency results may possibly underestimate the “true” patency of RA grafts. Current strategies in RA deployment include routine use to the circumflex and distal RCA/PDA territories. The favored in-flow is from the aorta, though in-flow from the LITA to create pedicled grafts is routinely used by one author (A.G.R.) and favored by all authors for off-pump CABG, or for conventional CABG where there is graft shortage or aortic wall disease and where clamps on the aorta are best avoided. Conduit choice for the moderately
Ann Thorac Surg 2002;73:143– 8
stenosed (⬍ 60%) dominant RCA is problematic. The pedicled RITA may be the most appropriate graft. Limitations of the study include the fact that it is observational, and that the RA was routinely used in our CABG group. However, in a study comparing the earlier half of this experience when SVG was still commonly used, radial artery use and total arterial revascularization were associated with significantly improved operative survival [6]. In conclusion, widespread routine use of the RA (in addition to the ITAs) is possible. The RA has proven to be an extremely versatile conduit and its use has been associated with extremely satisfactory clinical and angiographic results. Further evaluation into the long term is required. We thank Professor John Ludbrook for statistical advice, and secretarial staff, Tonia Iacuone, Jan Matthews, and Margaret Rizzo for preparation of the manuscript.
References 1. Grondin C, Campeau L, Lesperance J, Enjalbert M, Bourassa M. Comparison of late changes in internal mammary artery and saphenous vein grafts in two consecutive series of patients 10 years after operation. Circulation 1984;70: 1208–12. 2. Loop F, Lytle B, Cosgrove D, et al. Influence of the internalmammary-artery graft on 10-year survival and other cardiac events. N Engl J Med 1986;314:1– 6. 3. Tatoulis J, Buxton B, Fuller J. Results of 1,454 free right internal thoracic artery-to-coronary artery grafts. Ann Thorac Surg 1997;64:1263– 8. 4. Acar C, Jebara V, Portoghese M, et al. Revival of the radial artery for coronary artery bypass grafting. Ann Thorac Surg 1992;54:652–9. 5. Royse AG, Royse CF, Shah P, et al. Radial artery harvest technique, use and functional outcome. Eur J Cardiothorac Surg 1999;15:186–93. 6. Royse AG, Royse CF, Tatoulis J. Total arterial coronary revascularization and factors influencing in-hospital mortality. Eur J Cardiothorac Surg 1999;16:499 –505.
TATOULIS ET AL RADIAL ARTERY IN CORONARY SURGERY
147
7. Royse AG, Royse CF, Raman JS. Exclusive Y graft operation for multivessel coronary revascularization. Ann Thorac Surg 1999;68:1612– 8. 8. Calafiore A, Teodori G, Di Giammarco G, et al. Coronary revascularization with the radial artery: new interest for an old conduit. J Card Surg 1995;10:140– 6. 9. Sundt TM 3rd, Barner HB, Camillo CJ, Gay WA, Jr. Total arterial revascularization with an internal thoracic artery and radial artery T graft. Ann Thorac Surg 1999;68:399 – 404. 10. Tector A, Amundsen S, Schmahl T, Kress D, Peter M. Total revascularization with T grafts. Ann Thorac Surg 1994;57: 33– 8. 11. Tatoulis J, Buxton BF, Fuller JA, Royse AG. The radial artery as a graft for coronary revascularization: techniques and follow-up. In: Karp R, Lacks H, Weschler A, eds. Advances in cardiac surgery, vol. 11. St Louis: Mosby, 1999:99 –128. 12. Royse AG, Royse CF, Tatoulis J, et al. Postoperative radial artery angiography for coronary artery bypass surgery. Eur J Cardiothorac Surg 2000;17:294 –304. 13. Tatoulis J, Buxton BF, Fuller JA. Bilateral radial artery grafts in coronary reconstruction: technique and early results in 261 patients. Ann Thorac Surg 1998;66:714–9. 14. Tatoulis J, Buxton BF, Fuller JA, Royse AG. Total arterial coronary revascularization: techniques and results in 3,220 patients. Ann Thorac Surg 1999;68:2093–9. 15. Buxton BF, Ruengsakulrach P, Fuller J, et al. The right internal thoracic artery graft: benefits of grafting the left coronary system and native vessels with a high grade stenosis. Eur J Cardiothorac Surg 2000;18:255– 61. 16. Pick A, Orszulak T, Anderson B, Schaff H. Single versus bilateral internal mammary artery grafts: 10-year outcome analysis. Ann Thorac Surg 1997;64:599 – 605. 17. Lytle BW, Blackstone EH, Loop FD, et al. Two internal thoracic artery grafts are better than one. J Thorac Cardiovasc Surg 1999;117:855–72. 18. Possati G, Gaudino M, Alessandrini F, et al. Midterm clinical and angiographic results of radial artery grafts used for myocardial revascularization. J Thorac Cardiovasc Surg 1998;116:1015–21. 19. Acar C, Ramsheyi A, Pagny JY, et al. The radial artery for coronary artery bypass grafting: clinical and angiographic results at five years. J Thorac Cardiovasc Surg 1998;116: 981–9. 20. Calafiore A, Di Giammarco G, Teodori G, et al. Radial artery and inferior epigastric‘ artery in composite grafts: improved midterm angiographic results. Ann Thorac Surg 1995;60: 517–23.
DISCUSSION DR MARK W. CONNOLLY (New York, NY): I commend the authors on probably putting together the largest experience in the world in radial artery harvesting. At our institution, we use it in approximately 60% of the patients, but one of the things that we have difficulty with, particularly in females, is selling the cosmetic end of it. We are actually exploring endoscopic removal through small incisions. I wonder if you could comment on how you could get so many patients enrolled, particularly bilateral arm incisions, to buy into the procedure. It is clear that there is probably going to be a benefit from its patency rate compared with vein grafting, but I would be interested in how the authors do that with their patients. DR ROBBIN G. COHEN (Los Angeles, CA): I just have one question. It is a huge experience in radial arteries, but if you look at your angiographic follow-up, it turns out to be about 5% or less. Don’t you have a little bit of trouble extrapolating patency
data when your follow-up is in such a limited number of patients? DR SAMUEL RUSSELL VESTER (Cincinnati, OH): I was just curious if any of these were done off-pump and if there was an off-pump versus on-pump differential in the patency of your radial artery grafts? DR DOUG ADAMS (Merrillville, IN): How does the performance of radial angiography preoperatively affect your use of the radial artery? DR JOHN D. PUSKAS (Atlanta, GA): Let me ask one other question, and that relates again to the angiographic patency that Dr Cohen alluded to. There were 87% total arterial revascularization cases, therefore, 13% of patients had one or more saphenous vein grafts done. Can you compare for us the patency
148
TATOULIS ET AL RADIAL ARTERY IN CORONARY SURGERY
of those saphenous veins compared with the radial arteries at 1 year? DR TATOULIS: I would like to thank the discussants for their questions. Dr Connolly, with regard to endoscopic harvest, we have not attempted that. I guess we have adopted this technique of radial artery use fairly readily and I think most of the patients have come to accept it. The wounds heal very well, and an additional bonus is the ambulation is very rapid because you do not have leg incisions, and in fact, that is a bonus that we had not expected but it certainly is. But the wounds heal well and the patients are certainly accepting. With regard to the angiographic follow-up, I do acknowledge that 336 anastomoses followed up is small by comparison with 6,000-odd patients, however, we do have a prospective ongoing ethics-approved study to see these angiograms at 1 month, 1 year, 3 years, and 5 years. So we do have that as an ongoing study. And as you may well appreciate, it is not easy to get angiographic follow-up on patients in a routine fashion. There was a question regarding off-pump surgery. We have done 450 off-pump cases, almost all involving the radial. The mean number of anastomoses was 2.6 in that series, and we mainly used the Y graft for that series because it allows us to totally avoid the aorta. I do not have a differential between the patency of the off-pump cases and the regular cases. My under-
Ann Thorac Surg 2002;73:143– 8
standing was that they were very similar. So there was no statistical significance at this point in time. Otherwise, we would not be doing it. With regard to radial arteries that had been used for angiography, our cardiologists do not use the radial artery as a site for angiography or for angioplasty, so that has not been an issue. What can sometimes be an issue is the use of the radial artery in reoperations where the radial artery has been previously cannulated for the first coronary operation or for a major vascular procedure. In that case, we certainly explore the artery distally if the Allen’s test is appropriate, and sometimes it is normal; sometimes the distal 2 or 3 cm is partly fibrosed or dissected and that can be discarded and the proximal segment used. With regard to saphenous veins, yes, we do. All these patients obviously had all their grafts studied, and indeed there was something like 60-odd saphenous veins in this group, and the patency at 1 year was 87%. So it was very similar. I think the point being is we know that the patency of the saphenous veins declines over time, but we would hope once the technical and other issues relating to the implantation of the radial arteries are overcome, then we would hope that the patency then would plateau and stay the same. I thank the Association for the privilege of presenting this information.
Background. The radial artery (RA) has been used extensively by us as a way of reducing the use of the saphenous vein. It has been hoped that the RA will maintain greater late patency than the saphenous vein. We evaluated our initial 5-year experience with the RA in coronary surgery. Methods. We studied 6,646 consecutive patients who had a single RA (4,872), or bilateral RA (1,774), as coronary grafts, from June 1995 to June 2000. Angiograms were performed mostly in symptomatic patients, or as part of a research project in asymptomatic patients. Results. The patients’ mean age was 65.1 years; 23% had diabetes, 14% had unstable angina, and 42% had prior myocardial infarction. An average of 3.3 grafts per patient were performed, 87% from arterial conduit. Con-
duits used were RA (8,420), left internal thoracic artery (6,296), and right internal thoracic artery (1,076). Operative mortality occurred in 58 (0.9%) patients, stroke in 92 (1.4%), deep sternal infection in 97 (1.4%), reoperation for hemorrhage in 56 (0.9%), and myocardial infarction in 52 (0.8%). Peak mean postoperative creatine kinase MB (CKMB) was 16.5 IU/L. Two patients developed fingertip ischemia. Postoperative angiographic RA patency was 90.2% (333 of 369 distal anastomoses). Conclusions. Good early clinical and angiographic results can be achieved by using the RA in coronary surgery.
T
patients underwent isolated CABG in which one or both RAs were used. These patients form the study group. Patients having combined procedures (eg, valve and CABG) were excluded from this analysis to obtain a uniform study group. The patient characteristics are detailed in Table 1.
he incidence of failure of saphenous vein grafts in the late period [1, 2] prompted our progressive expansion of arterial conduit use. Initially, we used bilateral internal thoracic artery (ITA) [3]. However, it was commonly noted that the right ITA would not reach distal branches in the circumflex and right coronary territories, and these areas remained a challenge in terms of arterial revascularization. After the revival of radial artery (RA) use by Acar and associates [4], we used the RA [5], and subsequently made more efficient use of the conduit with sequential grafting [6] and composite grafting [7]. Others have reported similar operative strategies [8 –10]. The purpose of this study was to review our initial experience, more than 5 years, in isolated coronary artery bypass grafting (CABG) surgery where the RA was used.
Material and Methods Patients We began to use the RA as a conduit in CABG in December 1994, and within 6 to 18 months, its use became routine [5]. From June 1995 to June 2000, 6,646 Presented at the Thirty-seventh Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 29 –31, 2001. Address reprint requests to Dr Tatoulis, Suite 28, Private Medical Centre, Royal Melbourne Hospital, Parkville, Victoria 3050, Australia; e-mail: [email protected].
© 2002 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2002;73:143– 8) © 2002 by The Society of Thoracic Surgeons
Assessment of Ulnar Collateral Circulation All patients underwent a modified Allen’s test preoperatively. Both the RA and ulnar arteries were compressed while the hand was rendered ischemic by five strong fist clenching and relaxing actions for more than 30 seconds. The ulnar artery was then released. Reperfusion of the thenar eminence, thumb, and index finger within 5 seconds was considered excellent, within 10 seconds satisfactory, and beyond 10 seconds not acceptable. Details of technique have been previously reported [5]. Finger plethysmography or ultrasound examination was used sometimes as part of a research protocol. Contraindications to RA use were abnormal Allen’s test, prior trauma or surgery to the relevant upper limb, known subclavian or brachial artery stenosis, Raynaud’s phenomenon, scleroderma, and RA calcification.
Graft Procurement The left RA (nondominant) was the preferred conduit. After sternotomy, the left internal thoracic artery (LITA) and left RA harvest proceeded simultaneously. If both RAs were used, these were harvested first, the forearms 0003-4975/02/$22.00 PII S0003-4975(01)03290-8
144
TATOULIS ET AL RADIAL ARTERY IN CORONARY SURGERY
Table 1. Patient Demographics Characteristic Age (years)a Male/female Diabetes Angina classd III or IV Unstable anginac Left main stenosis ⬎ 50% Prior Q wave myocardial infarction Left ventricular ejection fraction ⬍ 0.5 Reoperation
No. of Patients (n ⫽ 6,646) 65.1 ⫾ 9.9 (29 – 89)b 5,182:1,464 (78:22) 1,547 (23) 3,455 (52) 916 (14) 864 (13) 2,778 (42) 238 (4) 434 (7)
a
b Standard deviation and range shown. Numbers in parentheses are c Society of Thoracic Surgeons definipercentages (except for age). d Canadian Cardiovascular Society Classification for tions used. angina.
closed and placed by the patient’s side, and covered with a further large sterile drape before sternotomy. In long operations (combined, reoperations), a small suction drain was sometimes placed.
Radial Artery Harvesting The distal one-third of the RA was usually exposed first to exclude calcification. The RA was harvested using a combination of low-power cautery [5] or sharp dissection without much cautery [11], dependent on surgeon preference. Small vascular clips were used to secure branches. Care was taken to avoid injury to the superficial radial and the lateral cutaneous nerve of the forearm.
Antispasm Prophylaxis Papaverine 1 mg/mL in heparinized arterial blood was introduced intraluminally via the distal end of the RA (5 mL) with a blunt-ended vascular needle, and the distal end of the RA was secured with a vascular clip. The RA was allowed to pulsate for 5 minutes before the proximal end was secured, and the RA was placed in an identical solution until its use. The ITAs similarly had intraluminal papaverine and a papaverine-soaked gauze placed around until use. Intravenous nitroglycerine (30 to 100 g/min) or Milrinone (0.1 g/kg/min) (SanofiSynthelabo, Sydney, Australia) were commonly used perioperatively for 24 hours according to surgeon’s preference. Oral calcium channel blockers (amlodipine or felodipine) were commenced postoperatively, and continued for a variable period (usually 6 months) after discharge.
Patient Monitoring Where bilateral RAs were harvested, a femoral arterial monitoring line was used. Swan-Ganz catheters, and transesophageal echocardiography were used routinely in all patients.
Surgical Strategy Grafting varied according to surgeon’s preference. Initially, total arterial revascularization was most frequently used in younger patients (⬍ 65 years), but later, all patients were considered suitable. Sequential grafting
Ann Thorac Surg 2002;73:143– 8
became increasingly popular due to the additional length and flexibility afforded by the RA. Similarly, composite “Y” grafting became more popular. The most common patterns of grafting were the LITA to the left anterior descending (LAD), the right internal thoracic artery (RITA) (if used) to the proximal circumflex or posterior descending artery (PDA), and the RAs to the distal circumflex, left ventricular branch of the right coronary artery (RCA), or PDA. Four thousand eight hundred seventy-two patients had one RA used, and 1,774 patients had both RAs harvested. The arterial conduits used are detailed in Table 2. Eighty-seven percent of all distal anastomoses were performed with arterial grafts, and the rest with saphenous vein.
Surgical Procedure This has been consistent for the duration of the study. Cardiopulmonary bypass was established by cannulation of the ascending aorta and right atrium. The patient was cooled to 34°C, with flows of 2.5 L/min/m2 and a mean blood pressure of 80 mm Hg maintained. Myocardial protection with antegrade (700 mL) and retrograde (300 mL) blood cardioplegia at 20°C was used to maintain a myocardial temperature of 25°C. Further doses of retrograde cardioplegia (300 mL) were given after each anastomosis. An aortic vent was used. If there was aortic regurgitation, a left ventricular vent via the right superior pulmonary vein was employed. All anastomoses were completed during one period of cross-clamping, with the use of a partial occlusion aortic clamp being rare. A “hot shot” dose (1,000 mL) of aspartate-enriched warm blood cardioplegia was usually given before cross-clamp release. If a Y graft was constructed, the proximal LITA/RA anastomosis was completed after heparinization but before bypass. The distal anastomoses were performed with continuous 7/0 polypropylene. The proximal anastomoses were performed directly onto a 3.5-mm punched opening in the ascending aorta with continuous 6/0 polypropylene. Sequential distal anastomoses were performed in a manner to ensure the best lie of the graft [7, 11]. Preference was for parallel sequential anastomoses if possible. In general, the LITA to LAD anastomosis was performed last. Pericardial cold slush was not used. The intraoperative details are shown in Table 3. Table 2. Conduits Used in 6,646 Patients Conduit LITA RITA RA Single RA Bilateral RA Total Total anastomoses Total arterial anastomoses
Patients
Conduits
n 6,296 1,076 8,420
4,892 1,774 6,646
4,872 3,548 8,420
LITA, RITA ⫽ left and right internal thoracic artery; artery.
21,931 18,079 (87%) RA ⫽ radial
Ann Thorac Surg 2002;73:143– 8
Data Collection and Analysis
Table 4. Operative Results
All data were collected prospectively and entered into a computer database program. Intraoperative data were entered at the time of operation. Each patient was reviewed and data were collected daily until hospital discharge. All surviving patients were reviewed at 30 days or more after operation by office, outpatient clinic, or cardiologist visit. The Society of Thoracic Surgeons definitions were used. Values are reported as mean ⫾ standard deviation and range. Percentages are given where appropriate. Comparisons between proportions were by the Fischer’s exact test (statXact 3.1; Cytel Software Corp, Cambridge, MA). A p value of less than 0.05 was considered significant.
Variable
Postoperative angiography was performed as part of an ethics-approved study (Royal Melbourne Hospital Research and Ethics Committee), or in response to symptoms [12]. Grafts with string signs were considered not patent.
Results Operative Mortality Fifty-eight patients died in hospital or within 30 days of CABG, or of associated complications. The operative mortality was 0.9%. The majority of the deaths were from cardiac causes or from multiorgan failure associated with low cardiac output.
a
n 58 (0.9)a 52 (0.9) 16.5 ⫾ 22 80 (1.2) 92 (1.4) 56 (0.9) 79 (1.4) 21 (0.3) 14 (0.2) 2
Numbers in parentheses are percentages. IABP ⫽ intraaortic
Fifty-six patients (0.9%) required return to the operating room for control of postoperative hemorrhage. In no instance was the source of bleeding from an RA branch. Ninety-seven patients (1.4%) developed a deep sternal wound infection, defined as any sternal infection that required intravenous antibiotics, or a further operation on the sternum. The most common pathogen was Staphylococcus aureus (92 patients, 95%). The mean time to extubation was 8.5 ⫾ 8.6 hours (range 0 to 226 hours), and the mean hospital stay was 7.7 ⫾ 5.4 days (range 4 to 126 days).
Hand and Forearm Complications
Perioperative Morbidity The perioperative major morbidity was low (Table 4). Fifty-two patients (0.9%) developed perioperative myocardial infarction by electrocardiogram (new Q-wave) and enzyme criteria (creatine kinase MB /cardiac isoenzyme [CKMB]). The mean postoperative peak level of CKMB was 16.5 ⫾ 22 IU/L (normal laboratory values, 0 to 25 IU/L). In 80 patients (1.2%), an intraaortic balloon pump was used for vascular support in the perioperative phase. In 21 (25%), the intraaortic balloon pump had been placed prebypass for unstable hemodynamics or because of poor left ventricular function. Ninety-two patients (1.4%) developed a stroke, defined as any clinically detectable neurologic abnormality producing a motor, speech, or sensory deficit. Six patients died after a severe stroke, and 7 died from intraabdominal catastrophes (five mesenteric ischemia, two abdominal aortic aneurysms). Table 3. Intraoperative Data for 6,466 Patients Variable
n
Aorto-coronary RA Y grafts (LITA-RA, RA-RA) Distal anastomoses per patient Aortic cross-clamp time (min)a
5,583 (84%) 1,063 (16%) 3.3 ⫾ 1.0 65 ⫾ 25
All distal and proximal anastomoses during one period of cross-clamp.
LITA ⫽ left internal thoracic artery;
Operative mortality Perioperative myocardial infarction Maximum CKMB, IU/L IABP use Stroke Reoperation for hemorrhage Deep sternal infection Forearm infection Forearm hematoma (drainage) Fingertip ischemia
CKMB ⫽ creatine kinase cardiac iso-enzyme; balloon pump.
Postoperative Angiography
a
145
TATOULIS ET AL RADIAL ARTERY IN CORONARY SURGERY
RA ⫽ radial artery.
Objective upper limb morbidity was minimal. Twentyone (0.3%) infections required intravenous or oral antibiotics, or some form of operative drainage. All except one were due to S aureus. Fourteen patients (0.2%) required evacuation of a forearm hematoma, most commonly between the 3rd and 5th postoperative days. Two patients developed fingertip (index finger) ischemia. Both had scleroderma, Raynaud’s phenomenon, severe peripheral vascular disease, and were early in our experience [13]. There was no difference in complications between patients with excellent (⬍ 5 seconds) and those with satisfactory (5 to 10 seconds) ulnar collaterals.
Radial Artery Graft Patency Two hundred seventy-one patients from both institutions underwent RA coronary graft angiography, a mean of 14.4 ⫾ 10.4 months postoperatively (range 1 to 35 months). Early patency data have been previously published [11–14]. Two hundred eighty RA conduits were studied with 369 RA to coronary anastomoses. RA segments to 36 (10%) anastomoses were either occluded or had string signs. Three hundred sixty-three of 369 (90.2%) anastomoses were widely patent (Table 5). Patency of the RA in the LAD and circumflex territories was 90.7% and 92.5%, respectively. Patency was lowest in the right coronary artery territory (86.7%), and is consistent with previous reports for ITA and RA [9, 12]. Additionally, string signs were associated with RA grafts to
146
TATOULIS ET AL RADIAL ARTERY IN CORONARY SURGERY
Ann Thorac Surg 2002;73:143– 8
Table 5. Comparison of Radial Artery Graft Patency Coronary Territory Left anterior descending Circumflex Right coronary artery Total
Anastomoses
Patent
Patency (%)
p Valuea
33
30
90.7
0.769b
201 135
186 117
92.5 86.7
0.725c 0.093d
369
333
90.2
a
b Fisher’s exact test. Compared with right coronary artery. d Compared with left anterior descending. Compared with circumflex.
c
coronary arteries of lesser degrees of coronary stenosis [12, 15].
Comment At the inception of coronary artery bypass surgery, aortacoronary saphenous vein grafting (SVG) was a major advance in the management of coronary artery disease. However, its full potential was often limited by the development of intimal hyperplasia and occlusive disease due to progressive conduit atheroma or thrombosis. Loop and associates [2] clearly demonstrated the clinical and angiographic superiority of the LITA grafted to the left anterior descending artery compared with SVG. Bilateral ITA grafting did not show an early advantage, mainly due to the concomitant use of SVG in both single and bilateral ITA groups. Recent studies suggest improved outcome with bilateral ITA [16, 17]. However, many surgeons may have been concerned by longer operation times or the potential for increased morbidity in the obese, diabetic, or those with chronic airways disease. Additionally, the RITA often cannot reach the distal circumflex and PDA territories. The revival of the RA has made available an additional arterial conduit, which can be used in addition to the ITA to achieve predominantly or total arterial coronary revascularization. Significant advantages for the RA are potential availability of two in most patients, excellent length, and excellent handling characteristics for constructing anastomoses. It is also resistant to kinking; and avoidance of leg wounds allows rapid ambulation [11, 13]. As well as being available for routine use, the RA is extremely important when there is conduit shortage: stripped saphenous vein, reoperations, peripheral vascular disease, or in patients with chronic airways disease or obese diabetics where RITA use may be avoided. Conversely, the RA may have distal calcification, or distal fibrosis from prior cannulation [11]. With a combination of the ITA and RA, we were able to achieve a very high rate (87%) of total arterial revascularization. Objective forearm and hand morbidity (hematoma and infection) was very low, at less than 1%. Discomfort, pain, paraesthesia, and wound tightness are more common complaints but settle with time [5]. If only one RA is to be used, the nondominant arm (usually left) should be preferred, as the efficiency of the surgery is enhanced
and there is less potential for disturbance of function. There was no relationship between the degree of ulnar collateral flow and postoperative forearm complications or function. RA spasm is not commonly seen during surgical procedure. Intraoperative prophylactic antispasmodic agents varied according to surgeon preference. All surgeons used intraluminal papaverine during harvest as previously used for the LITA. No topical “ice slush” was used in any of these patients. This may also protect against RA spasm. Most also administered low-dose glycerine Trinitrate, as was previously used for the LITA. Some surgeons preferred to administer intravenous Milrinone for its potent vasodilating effect. Calcium channel antagonists were normally prescribed postoperatively, and these were usually ceased within 3 to 6 months. These strategies have not been tested scientifically. The clinical role of postoperative oral calcium antagonists is not defined. However, isolated instances of discrete postoperative RA spasm have been reported [4]. RA calcification can be problematic, though rare [11]. When it occurs, it is more pronounced distally, on the superficial surface, and may be so severe that the RA is rigid and cannot be used. Difficulties include technical problems and hemostasis of the distal anastomosis, potential disruption of calcified plaques, intima, and possible thrombosis. Minimal calcification where the RA wall remains supple is acceptable. There are no data relating presence or absence of RA calcification to outcome. Overall, RA patency was 90.2% beyond 1 year and is similar to other reports [18 –20]. We have also found a lower patency when the RA was anastomosed to coronaries with lesser stenosis [12]. The patency findings are similar to those with the free RITA grafts [3]. The moderately stenosed coronary artery at the time of coronary operation poses a dilemma with respect to conduit choice. The patency was least in RA grafts to the RCA territory. Others have also noted a lower patency in this territory [9, 12, 15]. Relevant factors may be a higher incidence of coronary wall disease, the length of the RA conduit required, steal into the circumflex territory by sequential anastomoses, and potential for competitive flow in the setting of moderate (lesser) stenoses in large dominant RCA systems. Intuitively, it is expected that those RAs that are patent and are functioning well, beyond 1 year, should continue to do so into the long term. As a proportion of RA angiograms were performed for possible cardiac symptoms, the patency results may possibly underestimate the “true” patency of RA grafts. Current strategies in RA deployment include routine use to the circumflex and distal RCA/PDA territories. The favored in-flow is from the aorta, though in-flow from the LITA to create pedicled grafts is routinely used by one author (A.G.R.) and favored by all authors for off-pump CABG, or for conventional CABG where there is graft shortage or aortic wall disease and where clamps on the aorta are best avoided. Conduit choice for the moderately
Ann Thorac Surg 2002;73:143– 8
stenosed (⬍ 60%) dominant RCA is problematic. The pedicled RITA may be the most appropriate graft. Limitations of the study include the fact that it is observational, and that the RA was routinely used in our CABG group. However, in a study comparing the earlier half of this experience when SVG was still commonly used, radial artery use and total arterial revascularization were associated with significantly improved operative survival [6]. In conclusion, widespread routine use of the RA (in addition to the ITAs) is possible. The RA has proven to be an extremely versatile conduit and its use has been associated with extremely satisfactory clinical and angiographic results. Further evaluation into the long term is required. We thank Professor John Ludbrook for statistical advice, and secretarial staff, Tonia Iacuone, Jan Matthews, and Margaret Rizzo for preparation of the manuscript.
References 1. Grondin C, Campeau L, Lesperance J, Enjalbert M, Bourassa M. Comparison of late changes in internal mammary artery and saphenous vein grafts in two consecutive series of patients 10 years after operation. Circulation 1984;70: 1208–12. 2. Loop F, Lytle B, Cosgrove D, et al. Influence of the internalmammary-artery graft on 10-year survival and other cardiac events. N Engl J Med 1986;314:1– 6. 3. Tatoulis J, Buxton B, Fuller J. Results of 1,454 free right internal thoracic artery-to-coronary artery grafts. Ann Thorac Surg 1997;64:1263– 8. 4. Acar C, Jebara V, Portoghese M, et al. Revival of the radial artery for coronary artery bypass grafting. Ann Thorac Surg 1992;54:652–9. 5. Royse AG, Royse CF, Shah P, et al. Radial artery harvest technique, use and functional outcome. Eur J Cardiothorac Surg 1999;15:186–93. 6. Royse AG, Royse CF, Tatoulis J. Total arterial coronary revascularization and factors influencing in-hospital mortality. Eur J Cardiothorac Surg 1999;16:499 –505.
TATOULIS ET AL RADIAL ARTERY IN CORONARY SURGERY
147
7. Royse AG, Royse CF, Raman JS. Exclusive Y graft operation for multivessel coronary revascularization. Ann Thorac Surg 1999;68:1612– 8. 8. Calafiore A, Teodori G, Di Giammarco G, et al. Coronary revascularization with the radial artery: new interest for an old conduit. J Card Surg 1995;10:140– 6. 9. Sundt TM 3rd, Barner HB, Camillo CJ, Gay WA, Jr. Total arterial revascularization with an internal thoracic artery and radial artery T graft. Ann Thorac Surg 1999;68:399 – 404. 10. Tector A, Amundsen S, Schmahl T, Kress D, Peter M. Total revascularization with T grafts. Ann Thorac Surg 1994;57: 33– 8. 11. Tatoulis J, Buxton BF, Fuller JA, Royse AG. The radial artery as a graft for coronary revascularization: techniques and follow-up. In: Karp R, Lacks H, Weschler A, eds. Advances in cardiac surgery, vol. 11. St Louis: Mosby, 1999:99 –128. 12. Royse AG, Royse CF, Tatoulis J, et al. Postoperative radial artery angiography for coronary artery bypass surgery. Eur J Cardiothorac Surg 2000;17:294 –304. 13. Tatoulis J, Buxton BF, Fuller JA. Bilateral radial artery grafts in coronary reconstruction: technique and early results in 261 patients. Ann Thorac Surg 1998;66:714–9. 14. Tatoulis J, Buxton BF, Fuller JA, Royse AG. Total arterial coronary revascularization: techniques and results in 3,220 patients. Ann Thorac Surg 1999;68:2093–9. 15. Buxton BF, Ruengsakulrach P, Fuller J, et al. The right internal thoracic artery graft: benefits of grafting the left coronary system and native vessels with a high grade stenosis. Eur J Cardiothorac Surg 2000;18:255– 61. 16. Pick A, Orszulak T, Anderson B, Schaff H. Single versus bilateral internal mammary artery grafts: 10-year outcome analysis. Ann Thorac Surg 1997;64:599 – 605. 17. Lytle BW, Blackstone EH, Loop FD, et al. Two internal thoracic artery grafts are better than one. J Thorac Cardiovasc Surg 1999;117:855–72. 18. Possati G, Gaudino M, Alessandrini F, et al. Midterm clinical and angiographic results of radial artery grafts used for myocardial revascularization. J Thorac Cardiovasc Surg 1998;116:1015–21. 19. Acar C, Ramsheyi A, Pagny JY, et al. The radial artery for coronary artery bypass grafting: clinical and angiographic results at five years. J Thorac Cardiovasc Surg 1998;116: 981–9. 20. Calafiore A, Di Giammarco G, Teodori G, et al. Radial artery and inferior epigastric‘ artery in composite grafts: improved midterm angiographic results. Ann Thorac Surg 1995;60: 517–23.
DISCUSSION DR MARK W. CONNOLLY (New York, NY): I commend the authors on probably putting together the largest experience in the world in radial artery harvesting. At our institution, we use it in approximately 60% of the patients, but one of the things that we have difficulty with, particularly in females, is selling the cosmetic end of it. We are actually exploring endoscopic removal through small incisions. I wonder if you could comment on how you could get so many patients enrolled, particularly bilateral arm incisions, to buy into the procedure. It is clear that there is probably going to be a benefit from its patency rate compared with vein grafting, but I would be interested in how the authors do that with their patients. DR ROBBIN G. COHEN (Los Angeles, CA): I just have one question. It is a huge experience in radial arteries, but if you look at your angiographic follow-up, it turns out to be about 5% or less. Don’t you have a little bit of trouble extrapolating patency
data when your follow-up is in such a limited number of patients? DR SAMUEL RUSSELL VESTER (Cincinnati, OH): I was just curious if any of these were done off-pump and if there was an off-pump versus on-pump differential in the patency of your radial artery grafts? DR DOUG ADAMS (Merrillville, IN): How does the performance of radial angiography preoperatively affect your use of the radial artery? DR JOHN D. PUSKAS (Atlanta, GA): Let me ask one other question, and that relates again to the angiographic patency that Dr Cohen alluded to. There were 87% total arterial revascularization cases, therefore, 13% of patients had one or more saphenous vein grafts done. Can you compare for us the patency
148
TATOULIS ET AL RADIAL ARTERY IN CORONARY SURGERY
of those saphenous veins compared with the radial arteries at 1 year? DR TATOULIS: I would like to thank the discussants for their questions. Dr Connolly, with regard to endoscopic harvest, we have not attempted that. I guess we have adopted this technique of radial artery use fairly readily and I think most of the patients have come to accept it. The wounds heal very well, and an additional bonus is the ambulation is very rapid because you do not have leg incisions, and in fact, that is a bonus that we had not expected but it certainly is. But the wounds heal well and the patients are certainly accepting. With regard to the angiographic follow-up, I do acknowledge that 336 anastomoses followed up is small by comparison with 6,000-odd patients, however, we do have a prospective ongoing ethics-approved study to see these angiograms at 1 month, 1 year, 3 years, and 5 years. So we do have that as an ongoing study. And as you may well appreciate, it is not easy to get angiographic follow-up on patients in a routine fashion. There was a question regarding off-pump surgery. We have done 450 off-pump cases, almost all involving the radial. The mean number of anastomoses was 2.6 in that series, and we mainly used the Y graft for that series because it allows us to totally avoid the aorta. I do not have a differential between the patency of the off-pump cases and the regular cases. My under-
Ann Thorac Surg 2002;73:143– 8
standing was that they were very similar. So there was no statistical significance at this point in time. Otherwise, we would not be doing it. With regard to radial arteries that had been used for angiography, our cardiologists do not use the radial artery as a site for angiography or for angioplasty, so that has not been an issue. What can sometimes be an issue is the use of the radial artery in reoperations where the radial artery has been previously cannulated for the first coronary operation or for a major vascular procedure. In that case, we certainly explore the artery distally if the Allen’s test is appropriate, and sometimes it is normal; sometimes the distal 2 or 3 cm is partly fibrosed or dissected and that can be discarded and the proximal segment used. With regard to saphenous veins, yes, we do. All these patients obviously had all their grafts studied, and indeed there was something like 60-odd saphenous veins in this group, and the patency at 1 year was 87%. So it was very similar. I think the point being is we know that the patency of the saphenous veins declines over time, but we would hope once the technical and other issues relating to the implantation of the radial arteries are overcome, then we would hope that the patency then would plateau and stay the same. I thank the Association for the privilege of presenting this information.