bypass: A twenty-two-

Transcription

1 Carotid-subclavian year experience bypass: A twenty-two- Michael J. Vitti, MD, Bernard W. Thompson, MD, Raymond C. Read, MD, Paul J. Gagne, MD, Gary W. Barone, MD, Robert W. Barnes, MD, and John F. Eidt, MD, Little Rock, Ark. Purpose: A retrospective review of 124 patients who underwent carotid-subclavian bypass from 1968 to 1990 was done to assess primary patency and symptom resolution. Methods: Preoperative data included age, atherosclerosis risk factors, and indications for surgery. Perioperative data included mortality and morbidity rates and graft conduit. Postoperative follow-up assessed graft patency, resolution of symptoms, and late survival. Results: Age ranged from 42 to 78 years (mean 57.9). Indications for surgery were vertebrobasilar insufficiency in 24 (19%), extremity ischemia (EI) in 33 (27%), transient ischemic attacks (TIAs) in 13 (11%), both vertebrobasilar insufficiency and E1 in 31 (25%), and both TIAs and EI in 23 (18%) patients. Graft conduits were polytetrafluoroethylene in 44 (35%) and Dacron in 80 (65%) cases. Concomitant ipsilateral carotid endarterectomy was done in 32 (26%) patients. During operation, death occurred in one patient (0.8%), and complications occurred in 10 (8%) patients. Thirty-day primary patency and symptom-free survival rates were 100%. Long-term follow-up ranging from 5 to 164 months was available for the 60 cases done between 1975 and Three grafts occluded at 30, 36, and 51 months after surgery for a primary patency rate of 95% at 5 and 10 years. Twenty-two patients died, yielding survival rates of 83% at 5 years and 59% at 10 years. Symptom recurrence occurred in six (10%) patients from 9 to 66 months after surgery. The symptom-free survival rate was 98% at I year, 90% at 5 years, and 87% at 10 years. Symptoms recurred in three patients with occluded grafts and three with patent grafts. The preoperative symptoms of drop attacks and TIAs did not recur. EI recurred in 5% and was noted only in the presence of graft occlusion. Dizziness recurred in 17% of patients admitted with this symptom and was observed despite graft patency. Conclusion: Carotid-subclavian bypass was a safe and durable procedure for relief of symptomatic occlusive disease of the subclavian artery. Long-term symptomatic relief appeared particularly likely in patients with drop attacks or upper extremity ischemia. (J VASC SISRG 1994;20:411-8.) The management of occlusive disease affecting the innominate or subclavian arteries has evolved considerably over the past three decades. In 1962, Crawford et al.1 reported their initial series of 62 revascularizations of the innominate and subclavian arteries with transthoracic bypass grafting or endarterectomy. Although the therapeutic efficacy of transthoracic revascularization was satisfactory, operative mortality rates of From the Department of Surgery, University of Arkansas for Medical Sciences, and John L. McClellan VA Medical Center, Little Rock. Presented at the Eighteenth Annual Meeting of the Southern Association of Vascular Surgery, Scottsdale, Ariz., January 26-29, Reprint requests: Michael J. Vitti, MD, Department of Vascular Surgery, 4301 W. Markham St., Slot 520, Little Rock, AR /6/56842 this and subsequent series ranged from 6% to 19%. 2.4 As a result of these mortality rates, extrathoracic methods of revascularization were investigated. In 1967, Dietrich et al.5 introduced carotid-subclavian bypass (CSB) and reported symptom resolution in 93% of 125 patients, with an operative mortality rate of 4.8%. In 1964, Parrot 6 introduced the carotidsubclavian transposition whereby the subclavian artery is transected distal to the stenosis/occlusion and anastomosed to the side of the common carotid artery. Both techniques have been shown in subsequent studies to have comparable efficacy to transthoracic revascularization with considerably less operative morbidity and mortality rates.3,4,712 Axilloaxillary bypass, first introduced for treatment of occlusive disease in the subclavian or innominate arteries in 1971, has also been shown to be efficacious 411

2 412 Vitti et al. ~/OURNAL OF VASCULAR SURGERY " September 1994 Table I. Atherosclerosis risk factors (n = 124) No. patients (%) Cigarette smoking % Hypertension 87., 70% Hyperlipidemia 16 13% Diabetes meuitus 21 17% Coronary artery disease 84 68% in the relief of symptoms with minimal operative morbidity. 13,14 In 1980, Backman and Kim 15 described the use of percutaneous transluminal angioplasty as an alternative to surgical therapy for subclavian artery occlusive disease. The technical feasibility and low complication rate of percutaneous angioplasty have been demonstrated in subsequent reports, but the longterm durability of this treatment has yet to be fully evaluated. ~6-~9 To aid in selecting from this array of treatment options for a relatively uncommon disease entity, it was our purpose to assess both the immediate and long-term therapeutic efficacy of CSB in a study population accumulated over three decades of surgical experience. Follow-up of this patient population focused on assessments of patient mortality, graft patency, and resolution of preoperative symptoms. MATERIAL AND METHODS From January 1960 to January 1990, 140 patients (132 men, 8 women) underwent revascularization for symptomati c occlusive disease of the innominate or subclavian arteries at the John L. McClellan VA Medical Center in Little Rock, Ark. Transthoracic bypass grafting or endarterectomy was done in 16 patients who were excluded from this study. Because of the 19% operative mortality rate associated with the transthoracic approach, 2 extrathoracic reconstruction with CSB has been the preferred method of treatment at our institution since The 124 patients who underwent CSB from June 1968 to January 1990 constitute the study population. Hospital records were retrospectively reviewed for demographic data, atherosclerosis risk factors, and presenting clinical signs and symptoms. Also noted were the location and extent of occlusive disease on four-vessel arteriography, and the details of the operations. Data noted from the perioperative period (i.e., 30 days or less after surgery) included death, complications of surgery, and resolution of preoperative symptoms caused by occlusive disease of Table II. Presenting symptoms of vertebrobasilar insufficiency and upper extremity ischemia No. patients (%) Vertebrobasilar insuffidency (n = 55) Dizziness 21 39% Drop attacks 15 28% Ataxia 11 20% Vertigo 8 13% Syncope 8 13% Diplopia 6 11% Upper extremity ischemia (n = 87) Arm/hand weakness 34 42% Claudication 38 44% Acute ischemia 12 14% Dizziness (n = 21) Arm/hand weakness 8 38% Claudication 7 33% Drop attacks 6 29% "the subclavian artery. Postoperative follow-up focused on assessments of graft patency and resolution of preoperative symptoms made through periodic clinic visits. Graft patency was assessed by physical examination and blood pressure measurements, with duplex examination or arteriography reserved for patients with recurrent symptoms or a suspected graft failure. Age of the study patients ranged from 42 to 78 (mean 57.9) years. The prevalence of atherosclerosis risk factors is listed in Table I. All patients were white and 120 (97%) were cigarette smokers. Hypertension was present in 87 (70%) patients, whereas diabetes was present in 21 (17%) patients. Hyperlipidemia was detected in 16 (13%) patients, whereas a history of symptomatic coronary artery disease was present in 84 (68%) patients. The indications for surgery included symptoms of vertobrobasilar insufficiency in 24 (19%) patients, upper extremity ischemia in 33 (27%) patients, and hemispheric transient ischemic attacks (TIA) in 13 (10%) patients. Combined symptoms of vertebrobasilar insufficiency and upper extremity ischemia were present in 31 (25%) patients, whereas both TIA and upper extremity ischemia existed in 23 (19%) patients. The specific symptoms of the 55 patients with vertebrobasilar insufficiency and the 87 patients with upper extremity ischemia are listed in Table II. The most frequent presenting symptom of upper extremity ischemia was arm claudication, whereas dizziness was most frequent among vertebrobasilar symptoms. However, although arm claudication was often the sole indication for surgery, no patient underwent CSB for the isolated symptom of dizzi-

3 Volume 20, Number 3 V#ti et al. 413 Table III. Carotid-subclavian bypass: perioperative results (n ) No. patients (%) Mortality 1 0.8% Morbidity 10 8% Myocardial infarction 3 Gastrointestinal bleeding i Phrenic nerve palsy 2 Recurrent laryngeal nerve palsy 2 Marginal mandibular nerve palsy 1 Thoracic duct fistula 1 Resolution of symptoms % Graft patency % Neurologic deficit 0 0% ness. All patients with dizziness had additional symptoms, which prompted surgical intervention as listed in Table II. On physical examination, all patients exhibited a greatly diminished or absent brachial pulse on the affected side. A mean systolic brachial blood pressure gradient of mm Hg was also found between the upper extremities of study patients. All patients were evaluated before operation by a staff neurologist to confirm the diagnosis of symptomatic "subclavian steal" syndrome. Four-vessel arteriography demonstrated occlusive disease to involve the left subclavian artery in 99 (80%), the right subclavian artery in 21 (17%), and the proximal common carotid artery in 4 (3%) patients. Among those with subclavian artery involvement, stenosis of 80% or greater was seen in 67 (58 % ) patients, whereas occlusion was detected in 49 (42%) patients. In addition, concomitant stenosis of 50% or greater of the ipsilateral carotid artery was found in 32 (26%) patients (23 left carotid, 9 right carotid). Four patients were noted to have stenosis 50% or greater in the carotid artery contralateral to their subclavian artery occlusive disease. CSB was performed with the patient r&eiving general anesthetic with 8 mm Dacron as the graft conduit in 80 (65%) patients; 8 mm polytetrafluoroethylene was used in 44 (35%) patients. When symptomatic 50% or greater stenosis of the ipsilateral internal carotid artery was present, this lesion was corrected first at the same operation. When symptomatic contralateral internal carotid artery stenosis was found, surgical correction was done 6 weeks before bypass of the proximal subclavian or innominate artery lesion. Concomitant ipsilateral carotid endarterectomy was performed in 32 (26%) patients. RESULTS Perioperative period. Perioperative data were available for all 124 study patients. Perioperative morbidity, mortality, graft patency, and resolution of symptoms are listed in Table IIi. Within 30 days of surgery, one death occurred as a result of an acute myocardial infarction 3 days after an uneventful left CSB. Nonfatal complications occurred in 10 (8%) patients and consisted predominantly of myocardial infarctions (n = 3) and peripheral nerve injuries (n = 5) sustained at the time of surgery. Although one phrenic nerve palsy persisted for 4 months and was associated with a perioperative right lower lobe pneumonia, the remaining nerve injuries all resolved clinically within 2 months of follow-up. The thoracic duct fistula occurred after a left CSB and closed spontaneously 19 days after surgery. Aside from the functional deficits caused by the peripheral nerve injuries, the incidence of postoperative neurologic deficits and stroke was 0%. The addition of ipsilateral carotid endarterectomy to the CSB did not increase the likelihood of postoperative neurologic complications. Resolution of preoperative symptoms was uniform in all patients after CSB and the mean systolic brachial blood pressure gradient after operation ranged from 0 to 22 (mean ) mm Fig. Therefore, in this series, the operative mortality rate for CSB was 0.8%. The 30-day primary patency and symptom free survival rates were both 100%. Long-term follow-up. Because of inadequacies in the hospital records of most patients who underwent operation before 1975, long-term follow-up data were available only for patients receiving CSB between 1975 to 1990 (n = 60). For these 60 patients, follow-up ranged from 5 to 176 (mean 91.5) months. Twenty-two patients died during follow-up. Causes of death were myocardial infarction/congestive heart failure in 12 (55%) patients, malignancy in five (23%) patients, and unknown causes in the remaining five patients. The compilation of cumulative long-term survival rates by life-table analysis is illustrated in Table IV. The

4 414 Vitti et al. September 1994 Table IV. Carotid subclavian bypass: long-term survival Withdrawn Interval (yr.) Patients at start Duration Lost to follow-up Death Interval survival Cumulative survival Table V. Carotid-subclavian bypass: long-term primary patency Withdrawn Interval (yr.) Patients at start Occlusions Duration Lost to follow-up Death Interval patency Cumulative patency survival rate was 83% at 5 years, 59% at 10 years, and 51% at 15 years. Three PTFE grafts occluded during follow-up at 30, 36, and 51 months after surgery. The cumulative duration of primary patency was calculated by life-table analysis. Observations were censored at the last follow-up visit for grafts that remained patent. Observations were also censored for patients lost to follow-up, who died of causes unrelated to the graft, or who had patent grafts at the end of the study. This analysis of long-term primary CSB patency is illustrated in Table V. The primary patency rate of CSB was 94% at both 5 and 10 years after surgery. Recurrence of preoperative symptoms occurred in six (10%) patients from 9 to 66 months after surgery. In three patients, the recurrent symptoms were those of upper extremity ischemia (arm claudication with arm/hand weakness). Duplex examination of these patients revealed graft occlusion in all three cases. In the remaining three instances of symptom recurrence, dizziness was the returning complaint. Duplex examination of these patients revealed patent CBS grafts. In addition, these three patients had no recurrence of any of their other preoperative vertebrobasilar or extremity symptoms. Life-table analysis of symptom-free survival is illustrated in Table VI. The symptom-free survival rate was 98% at 1 year, 90% at 5 years, and 87% at 10 years. Therefore symptom recurrence occurred in the form of upper extremity ischemia in the presence of graft occlusions, whereas recurrence of dizziness did not signal graft occlusion. The preoperative symptoms of drop attacks and hemispheric TIAs did not recur during the long-term follow-up of this study population. DISCUSSION Arterial occlusive disease affecting the proximal innominate or subclavian arteries may be asymptomatic or may produce disabling symptoms of upper extremity ischemia or vertebrobasilar insufficiency. Atherosclerotic occlusive disease involving the proximal subclavian or innominate arteries is relatively uncommon because only 17% of patients with extracranial cerebrovascular disease were noted to have stenosis of 30% or greater within these vessels.

5 Volume 20, Number 3 Vitti et al. 415 Table VI. Carotid-subclavian bypass: symptom-free survival Withdrawn Interval Patient Recurrent symptonu (yr.) at start (no.) Duration Lost to fbllow-up Interval Cumulative Death asymptomatie asymptomatic Of these patients, only 24% had clinical symptoms. 2 Moreover, atherosclerosis does not progress rapidly in the proximal subclavian artery, because only 17% of patients demonstrated any occlusive progression during 2 years of serial ultrasound surveillance. 21 Asymptomatic subclavian steal has also been shown to be associated with a minimal risk of subsequent stroke and at most a 2% to 8% chance of progression to clinical symptoms over 2 yearsy,23 Therefore surgical treatment of subclavian artery obstruction is only indicated in the presence of disabling symptoms of upper extremity ischemia or vertebrobasilar insufficiency to improve quality of life. The treatment selected should therefore be safe, effective, and long lasting in its benefits. Currently, carotid-subclavian bypass, subclaviancarotid transposition, axilloaxillary bypass, and percutaneous angioplasty have all been advocated as the primary treatment for symptomatic subclavian steal syndrome. However, because of the low incidence of symptomatic disease, most series reporting on the efficacy of these treatment modalities are relatively small in size with limited long-term follow-up. Small sample sizes have also compromised the reliability of trials that compare one form of treatment to another. These trials have, for the most part, been retrospective, nonrandomized, and therefore vulnerable to selection bias. This study presents the results of CSB from an institution in which it constitutes the primary treatment for symptomatic subclavian steal. This uniform application of CSB over a long period of time for all cases of symptomatic subclavian steal permits a reliable assessment of the safety, efficacy, and durability of this operation. CSB can be carried out by direct side-to-side anastomosis or by interposing a short graft between the carotid and subclavian arteries. Because of the documented superiority of prosthetic grafts compared with saphenous vein grafts, Dacron or PTFE was used for all CSBs in this series. The small diameter of the vein, as well as its potential for kinking and shortening because of neck motion, may explain the inferior performance of vein grafts as compared to the wider, more rigid prosthetics. 9 Regarding safety, because of the high mortality rate (18.7%) of transthoracic revascularization early in our experience, extrathoracic reconstruction in the form of CSB was adopted in 1968: The operative mortality rate of 0.8% attests to the minimal operarive risk posed by CSB, and compares favorably with other series for CSB (0% to 5%), subclavian-carotid transposition (0% to 3%), and axilloaxillary bypass (0% to 3%). TM In this series, no strokes or neurologic deficits occurred after CSB. The addition of ipsilateral carotid endarterectomy in the presence of symptomatic 50% or greater stenosis of the carotid artery had no effect on the incidence of postoperative neurologic deficits. The risk of stroke from common carotid artery cross-clamping and endarterectomy can thus be minimal and need not argue against the use of CSB. The initial concern about the creation of a carotid steal, manifested by ipsilateral hemispheric TIAs, no longer exists. A steal will only occur in the presence of a significant stenosis in the carotid artery proximal to the bypass or at the bifurcation. Regarding efficacy, resolution of preoperative symptoms occurred in 100% after surgery with symptom-free survival rates of 90% at 5 years and 87% at 10 years. These symptom-free survival rates compare favorably with other CSB series (80% to 85%) and is superior to those reported for axilloaxillary bypass (65% to 80%). 814,24 Longterm durability of CSB is supported by the 94% 5- and 10-year patency rates, which also compare favorably with the 66% to 89% patency rates reported for axilloaxillary bypass. 13'14,24 Carotidsubclavian transposition has been reported to have

6 416 Vitti et al. IOURNAL OF VASCULAR SURGERY September 1994 a 100% 7-year patency rate in one series, 1 but extensive experience with this procedure has yet to be reported. Percutaneous transluminal angioplasty of the subclavian artery appears considerably less durable than surgical treatments because 2-year patcncy rates of 46% to 80% have been reported, ls,2s-27 The performance of concomitant endarterectomy did not appear to affect CSB graft longevity because two of the three grafts that occluded after operation had undergone CSB alone. Concomitant endarterectomy was not associated with any increase in the incidence of neurologic sequelae either during operation or during long-term follow-up. Therefore, in the presence of symptomatic occlusive disease affecting both the subclavian and carotid arteries, this study suggests that a CSB with concomitant carotid endarterectomy may be undertaken with a neurologic morbidity and therapeutic efficacy rate comparable to CSB alone. No concomitant endarterectomy was done in this study population for asymptomatic ipsilateral carotid artery stenosis. In the event of symptom recurrence, this study showed that recurrent extremity symptoms were associated with CSB graft occlusion, whereas recurrent dizziness occurred in the presence of patent grafts. Recurrent symptoms in the presence of patent grafts was reported in another series,12 but no specific symptoms were noted to be more likely to recur in this setting. In this study, no preoperative symptom other than dizziness recurred in the presence of a patent CSB graft. A cause other than symptomatic subclavian steal syndrome may have been responsible for the symptom of dizziness in these patients before operation or during follow-up. However, no definitive data to this effect could be ascertained from this retrospective review. Hemispheric TIAs and drop attacks were not observed to recur in this series of CSBs. Therefore in this study CSB appeared to be a safe, highly efficacious, and durable procedure in the treatment of symptomatic occlusive disease of the proximal innominate and subclavian arteries. This series, although not a comparative trial, still attests to the value of CSB in the surgical management of arterial occlusive disease of the innominate and subclavian arteries. REFERENCES 1. Crawford ES, DeBakey ME, Morris GC, Howell IF. Surgical treatment of occlusion of the innominate, common carotid, and subclavian arteries: a 10-year experience. Surgery 1969; 65: Thompson BW, Read RC, Campbell GS. Operative correction of proximal blocks of the subclavian or innominate arteries. J Cardiovasc Surg 1980;21: Raithel D. Our experience of surgery for innominate and subclavian lesions. J Cardiovasc Surg 1980;21: Vogt DP, Hertzer NR, O'Hara PJ, Beven EG. Brachiocephalic arterial reconstruction. Ann Surg 1982;196: Dietrich EB, Garrett HE, Ameriso J, Crawford ES, EI-Bayar M, DeBakey ME. Occlusive disease of the common carotid and subclavian arteries treated by carotid subclavian bypass: analysis of 125 cases. Am J Surg 1967;114: Parrot JD. The subclavian steal syndrome. Arch Surg 1969; 88: Moore WS, Malone JM, Goldstone 1. Extrathoracic repair of branch occlusions of the aortic arch. Am J Surg 1976;132: Crawford ES, Stowe CL, Powers RW Jr. Occlusion of the innominate, common carotid, and subclavian arteries: longterm results of surgical treatment. Surgery 1983;94: Ziomek S, Quifiones-Baldrich WJ, Busuttil RW, Baker JD, Machleder HI, Moore WS. The superiority of synthetic arterial grafts over autogenous veins in carotid-subclavian bypass. J VASe SURG 1986;3: Sterpetri AV, Schultz RD, Farina C, Feldhans RJ. Subclavian artery revascularization: a comparison between carotidsubclavian bypass and subclavian-carotid transposition. Surgery 1989;106: Sandman W, Kneimeyer HW, Jeaschock R, Hennerici M, Aulick A. The role of subclavian-carotid transposition in surgery for supraaorfic occlusive disease. J VASC SURG 1987;5: Perler BA, Williams GM. Carotid-subclavian bypass-a decade of experience. J VASC SURG 1990;12: Rosenthal D, EDison RG Jr, Clark MD, et al. AxiUoaxillary bypass: is it worthwhile? J Cardiovasc Surg 1988;29: Schanzer H, Chung-Loy H, Haimov M, Jacobson JH. Evaluation of axilloaxillary artery bypass for the treatment of subclavian or innominate artery occlusive disease. J Cardiovasc Surg 1987;28: Bachman DM, Kim RM. Transluminal dilatation for subclavian steal syndrome. Am J Radiol 1980;135: Motarjime A, Keifer JW, Kuske AJ. Percutaneous transluminal angioplasty of the brachiocephalic arteries. Am J Radiol 1982;138: Galicki JP, Bajai AK, Vine DL, Roberts RW. Subclavian artery stenosis treated by transhiminal angioplasty: six cases. Cardiovasc Intervent Radiol 1983;6: Burke DR, Gordon RL, Mishkin JD, McLean GK, Meranje SG. Percutaneous transhiminal angioplasty of subclavian arteries. Radiology 1987;164: Jaschke W, Menges HW, Ockert D, Huck K, George M. PTA of the subclavian and innominate arteries: short- and longterm results. Ann Radiol 1989;32: Fields WS, Lemak NA. Joint study of extracranial arterial occlusion: subclavian steal-a review of 169 cases. JAMA 1972;222: Ackerman H, Diener HC, Siboldt H, Huth C. Ultrasonographic followup of subclavian stenosis and occlusion: natural history and surgical treatment. Stroke 1988;19: Bornstein NM, Norris JW. Subclavian steal: a harmless hemodynamic phenomenon? Lancet 1986;2:303-5.

7 Volume 20, Number 3 Vitti et al Moran KT, Zide RS, Persson AV, Jewell ER. Natural history of subclavian steal syndrome. Am Surg 1988;54: Abu Rahma AF, Robinson PA, Kahn MZ, Khan JH, Boland JP. Brachiocephalic revascularization: a comparison between carotid-subclavian bypass and axilloaxillary bypass. Surgery 1992;112: Farina C, Mingoli A, Schultz RD, et al. Percutaneous transluminal angioplasty versus surgery for subclavian artery occlusive disease. Am J Surg 1989;158: Erbstein RA, Wooley RA, Smoot S. Subclavian artery steal syndrome: treatment by percutaneous transluminal angioplasty. Am J Roentgenol 1988;151: Hebrang A, Maskoic J, Tomac B. Percutaneous transluminal angioplasty of subclavian arteries: long-term results in 52 patients. Am J Roentgenol 1991;156: Submitted Feb. 3, 1994; accepted April 17, DISCUSSION Dr. Joseph L. Mulherin, Jr. (Nash~lle, Tenn.). In reading the numbers in their study, follow-up was available on less than half of these patients. There were 60 patients in whom adequate records were available for follow-up and at the time of this retrospective review 32 of those 60 were dead, leaving 28 patients alive for follow-up. Of these, six patients had recurrent symptoms and three of these patients, or 11%, had documented occlusion of their grafts. I suggest that there is a better way. William Edwards, Jr, reported our experience with operations for subclavian disease in 190 patients. In only 12 of these patients was a CSB performed, and on late follow-up 50% of these bypasses had occluded. I admit we are biased in our approach to subclavian artery disease and reserve bypass only for those patients with severe disease and disease involving the subclavian artery distally. We also freely admit that, by not counting these patients in our carotid-subclavian transpositions, we may favorably weigh the outcomes because these patients obviously were included in the current series. Eight of our patients early in the experience underwent a side-to-side carotid-subclavian anastomosis, which we no longer do, and 18 underwent subclavian transposition associated with other procedures such as coronary artery bypass, vertebral carotid artery transposition, and simultaneous carotid endarterectomy. Of 145 patients undergoing an isolated subclavian to carotid transposition, we had a 1.4% mortality rate and a 0% neurotogic morbidity rate. Of the 145 patients, only five were lost to follow-up, giving us follow-up on 97% of the patients, with a mean follow-up of 46 months and 74 patients in whom follow-up was more than 5 years. There was one occlusion in this group of patients which occurred at 26 months and was due to occlusion of the carotid artery. We believe carotid-subclavian transposition is a better operation. There are a number of advantages. It is an easier operation. Subclavian artery distal to the vertebral artery is in a busy place; behind the scalene muscle, the phrenic nerve is superficial to it, there are branches going north and south, as the trunks of the brachial plexus become involved, and the clavicle is in the way. Transposition requires no foreign body; it is a simple matter of sewing two blood vessels together. There is no compliance mismatch to favor the development of intimal hyperplasia. And finally, in patients in whom the subclavian is not completely occluded, by removing the disease from the flow, it completely alleviates the possibility of embolism. It cannot be done in everybody. In patients who have disease distal to the vertebral artery, bypass may be preferable. It is also important to remember because subclavian-carotid transposition requires the simultaneous occlusion of the carotid artery and the vertebral flow on that side, in patients who have occlusion of the contralateral internal carotid and vertebral artery we have been reluctant to apply this technique without dealing with it in some other fashion. Homer's syndrome is frequent, and we warn all of our patients about this before operation. A transient Homer's syndrome will probably develop in approximately 20% of patients. This has not been mentioned in this study. I think this is probably because of the dissection in the area of the vertebral artery necessary to free the vertebral up to allow the transected subclavian artery to be brought up to the carotid artery, which results in this Homer's syndrome, but this has not been a problem and usually resolves within 1 or 2 months. I want to mention bleeding. In dissecting the proximal subclavian artery, although it is an easier dissection than dissecting an area in the subclavian artery distal to the vertebral artery, it creates a space that is deep in the neck and the trachea is just medial to it. When you close the cervical fascia in closing the incision, you create a space where if bleeding occurs it frequently is not immediately obvious, and I would warn that respiratory symptoms in the recovery room in patients having had this procedure should be dealt with aggressively. In discussing the problem of subclavian artery occlusive disease, a number of studies in the recent literature have mentioned carotid-subclavian transposition, and frequently the comment has been made that good results have been reported. However, long-term follow-up is lacking, and the number of patients is small. At this time, with our recent report included in the recent literature there are 453 cases of carotid-subclavian transpositions having been performed with a 99% patency rate. I suggest that this is gradually becoming the procedure of choice for this problem.

8 418 Vitti et al. September 1994 Dr. Bruce A. Pealer (Baltimore, Md.). I think whether you bypass it, or translocate it, the long-term results in our experience, like yours, are better than with percutaneous transluminal angioplasty, and probably equally safe. Despite that, our radiologists lobby very hard to dilate these vessels. I am just wondering what role you see for percutaneous transluminal angioplasty, if any, in the treatment of these patients. Dr. Michael J. Vitti. To supplement some of the information presented, I add that we had two phrenic nerve injuries, but those were transient in nature and resolved within 4 months of follow-up. Patients who underwent operation before 1975 were not presented because their records were not of sufficient detail to permit life-table analysis, which was carried out for the subsequent 60 patients. There is, however, adequate follow-up to note that their survival rates at 5 years was 87% and at 10 years 70%. Similarly, their primary patency rate at 5 years was 96% and at 10 years 92%. Both subsets of the total study population therefore showed a consistently high primary patency rate, so those 60 patients were in no way a distortion of the entire study population. As far as percutaneous transluminal angioplasty is concerned, looking at some recent literature citing fairly high recurrent rates at 5 years of anywhere from 20% to 40%, I would reserve that treatment for patients at very high risk, especially those with severe coronary artery disease.