Annual meeting abstract

Annual meeting abstract

ANNUAL MEETING ABSTRACTS From the Eastern Vascular Society The Sixth Annual Meeting of the Eastern Vascular Society was held in New York, N.Y., on ...

1MB Sizes 0 Downloads 3 Views

Recommend Documents

No documents
ANNUAL

MEETING

ABSTRACTS

From the Eastern Vascular Society The Sixth Annual Meeting of the Eastern Vascular Society was held in New York, N.Y., on April 30 to May 3, 1992. Selected abstracts from that meeting are published here. Twenty-year trends using autogenous venous bypasses Carlos Donayre, MD, Richard M. Green, MD, Kenneth Ouriel, MD, and James A. DeWeese, MD, Unipersity of RochesterMedical Center, Rochester, N.T. The status of 103 patients (G-I) 5 years after 113 autogenous venous bypass grafts (AVBGs) performed between 1957 to 1964 at our institution was reported in 1971. This report compares these resultswith the status of 100 patients (G-II) who underwent 125 AVBGs from 1982 to 1985 and were monitored in a similar fashion. Sign&ant changes occurred in patient demographics. G-II patients were signiIicantly older, 51% of the patients were more than 70 yearsof age in G-II compared with only 29% in G-I. The male : female ratio was 1.43 in G-II compared with 2.96 in G-I. Hypertension was noted in 51% of patients in G-II compared with 29% of patients in G-I. Finally, 38% of patients in G-II were diabetic compared with 24% of G-I. More patients in G-II had a prior history of myocardial infarction (42% vs 330/b), but this difference did not reach statistical significance. Similarly, more patients in G-II were operated on for salvage(70% vs 59%) to arteries below the knee (70% vs 62%). Equal numbers of patients were alive after 5 years in both groups (52%, G-I; 5196, G-II), and in each group mortality rates were significantly higher in the salvage population than in patients with claudication (62% vs2696, G-I; 75% vs 1196, G-II). Both groups were monitored before the introduction of duplex scanning for graft surveillance with either oscillometrics or Doppler-derived ankle pressuresat regular intervals. The 30&y mortality rate was 2.6% for G-I, whereas there were no perioperative deaths in G-II. The only significant difference in patency rates between the two groups occurred at 30 days. Twenty-nine of 113 (26%) grafts in G-I occluded compared with only 9 of 125 grafts (7%) in G-II, p < 0.05. Thereafter, interval failure rates were similar, and by 24 months the graft patency rateswere similar (6196, G-I; 69%, G-II). At 5 yearsthe patency rates were 57% and 53%, respectively. Twenty-eight (25%) extremities were amputated in G-I compared with only 18 (14%) in G-11,~ < 0.05. The 20-year trends indicate that patients undergoing AVBG are older, more are women, and more have associated medical problems. Fewer operations are done for claudication to the above-knee popliteal artery. The striking improvements in early patency rates and the higher limb salvage rate in G-II is testimony to improved judgement and technique. Unfortunately, the long-term factors adverselyaffecting graft patency have not changed. Strategies for reducing the inevitable attrition of AVBGs remain our challenge in the 1990s.

The relative risks of limb revascnlarization and amputation in the modern era Michael J. Schina, Jr., MD, Robert G. Atnip, MD, Dean A. Healy, MD, and Brian L. Thiele, MD, The Penn$vania State

Universi~,

Milton

S. Hershey Medial

[email protected], Hershey,

Pa.

A retrospective review of 266 patients who underwent infrainguinal revascularization for limb salvage and/or major amputation (transmetatarsal, below knee, or above knee) from 1984 to 1990 was conducted to determine procedure-specific 30-day operative morbidity (defined as all postoperative complications) and mortality rates. The incidence of hypertension, diabetes mellitus, and coronary artery disease was similar in both groups. Two hundred eleven patients (61% men, 39% women, mean age, 68) underwent 295 infrainguinal vascular reconstructions. There were 195 primary and 100 secondary vascular reconstructions. There were 122 major amputations in 98 patients (47% men, 53% women, mean age, 71,29 above knee, 70 below knee, 23 transmetatarsal). Only 9% of patients were denied revascularization because of medical reasons, and there was one postoperative death in this group. Most other amputations were required because of unreconstructible vascular disease, including 39 patients (41 extremities) who underwent amputation after failed infrainguinal revascularization. Procedure-specificcomprehensive morbidity and mortality rates were 48% and 2% for primary revascularization, 35% and 2% for secondary revascularization, and 37% and 4%, respectively, for amputation. The difference in mortality rate between revascularization and amputation approached but did not achievestatistical significance. Cardiac and wound complications were the major cause of morbidity in all groups. Deaths were a result of cardiacevent (8 patients), sepsis(2), or metabolic derangement (1). Revascularization can be offered to virtually ail patients with advanced limb ischemia and carries a mortality rate equivalent to or perhaps lower than that of amputation. When limb amputation r;F required, it can be performed with a mortality rate remarkably lower than that ascribed to amputation in the older literature. Magnetic resonance angiography: A new gold standard Jeffrey P. Carpenter, MD, Rodney S. Owen, MD, Richard A. Baum, MD, Clyde F. Barker, MD, Henry D. Berkowitz, MD, Constantin C. Cope, MD, Michael A. Golden, MD, and Leonard J. Perloff, MD, TheDepartments of SurBeyand Radkh~, Pbiti~bia,

Unbm’ty Pa.

of

Penmylvania

School of Medicine,

Planning lower limb salvage procedures demands 297

298

Annual

meeting abstracts

detailed imaging of extremity arteries. To date, contrast arteriography has been the gold standard for diagnosis and treatment decisions. This study was undertaken to determine whether magnetic resonance angiography (MRA) would provide images with comparable detail in the noninvasive evaluation of the anatomy of the infracrural arteries. The leg vessels of 51 patients (55 limbs) with vascular occlusive disease were studied before operation by conventional angiography and MRA. The arteries were divided into 14 segments and scored as normal, diseased, or occluded. Independent interventional plans were developed on the basis of each technique. Validation of MRA and conventional angiograms was provided by intraoperative arteriography and direct operative exploration. The two studies differed in 26 limbs (48%). Not only did MRA demonstrate all of the vessels identified by conventional angiography, it also identified an additional 106 patent segments occult to conventional arteriography, yielding an increased sensitivity of 24% overall (p < 0.01). MRA identified additional patent segments at every level. This increased sensitivity was most apparent in the distal segments, with 17 additional patent dorsal pedal arteries detected by MRA (52% greater sensitivity). Greater specificity of MRA was demonstrated by the correct identification of vessels and stenoses misidentified by conventional angiography. As a result, operative management decisions were modiied in nine patients (18%). Bypass procedures were performed successfully in seven patients who would have otherwise undergone amputation. In conclusion, MRA is a noninvasive technique with greater sensitivity and specificity than conventional arteriography. MRA resulted in increased limb salvage through improved demonstration of patent but angiographically “occult” target vessels in seven of 51 patients. Prebypass saphenous vein angioscopy: A valuable method to detect unsuspected venous pathology Clifford M. Sales, MD, Michael L. Marin, MD, Frank J. Veith, MD, Wiiiam D. Suggs, MD, Kurt R. Wengerter, MD, Thomas F. Panetta, MD, and Ronald E. Gordon, MD, [email protected] Medical Center/Albert Einstein Collge of Medicine and Mount Sinai School of Medicine, New York, N.Y. The presence of preexisting saphenous vein lesions adversely affects graft patency. Despite careful preoperative venous duplex examination and meticulous intraoperative evaluation, clinically significant saphenous vein abnormalities may remain undetected. We evaluated angioscopy as a means to better detect these saphenous vein lesions. Saphenous vein remnants from 90 patients, obtained at the time of bypass surgery, were later examined angioscopically. The specimens were then sectioned longitudinally allowing for direct luminal inspection and histologic examination. The angioscopic and histologic appearances were recorded by independent examiners blinded to each other’s results.

Journal of VASCULAR SURGERY

The overall sensitivity and specificity of angioscopy for identifying saphenous vein disease were 65% and lOO%, respectively. Angioscopic findings of irregular white plaques reliably identified sclerotic vein segments. Images of multiple lumens, fibrous strands, and thickened valve cusps were all verified histologically and thought to be indicative of a postphlebitic vein. Absence of an angioscopic lumen was confirmed histologically in occluded veins. However, thick walled veins and varicosities could not be distinguished from normal veins on the basis of their angioscopic appearance. The use of angioscopy to detect preexisting saphenous vein disease in our last 17 patients has identified five unexpected diseased vein segments requiring excision or repair before use of the vein as a graft. Since the use of angioscopy is a reliable means of prospectively assessing the vein for most preexisting lesions, its routine use may ultimately improve graft patency. l?TFE versus HW in above-knee femoropopliteal bypass: Long-term results of a randomized clinical trial Theo J. van Vroonhoven, MD, PhD, and Gijs Aalders, MD, Department of Surdely University Hospital, Utrecht, and Universily Hospital Mamtricbt, The Netherlands Controlled studies comparing polytetrafhroroethylene (PTFE) and human umbilical vein (HUV) are remarkably few, that is, absent. We undertook a prospective randomized trial comparing PTFE and HUV in a well-defined setting: above-knee femoropopliteal bypass. All patients undergoing above-knee femoropopliteal bypass during 1983 to 1984 were candidates for entry into the study. In patients with claudication a prosthetic graft was used intentionally; in limb salvage cases prosthetic grafts were used only when autologous vein was insufficient. Preoperative evaluation included complete clinical examination, Doppler studies at rest and during reactive hyperemia, and complete arteriography. Standard surgical techniques were used. All patients received oral warfarin (Coumadin) therapy after operation indefinitely. Ninety-three extremities were randomized, three of which were excluded, leaving 90 (45 PTFE, 45 HW) for follow-up and analysis. The two groups showed no differences in age, gender, indication, diabetes mellitus, Doppler indexes, and arteriographic outflow characteristics. Preoperative and postoperative Doppler indexes were 60.9% and 99.596, respectively, in the PTFE and 60.3% and 95X%, respectively, in the HUV group. Median follow-up was 76 months (47 to 91), during which 23 patients died with functioning grafts, (malignancies, 10; cardiovascular causes, 8). Primary patency rates in the PTFE group at 2, 3 and 5 years were 62.6%, 52.5%, and 38.7%, respectively, and in the HW group 91.4%, 86.4%, and 75.196, respectively, showing a statistically significant difference (log rank test p < 0.0005) in favor of the HUV graft. When only patients with claudication were considered, the difference

Volume 16 Number 2 August 1992

again was statistically different (log rank test p < 0.01. Secondary patency at 5 yearswas 5 1.4% in the PTFE group and 80.2% in the HUV group (log rank testp < 0.01). Secondary patency at 5 yearswas 5 1.4% in the PTFE group and 80.2% in the HUV group (log rank testp < 0.005). When a prosthetic graft is necessaryor desirable in the above-knee position HUV is to be preferred over PTFE grafts. Lib salvage in patients with a palpable popliteal pulse Robert A. Brigham, MD, and Eric C. Jaxheimer, MD, The Reading Hospital and Medical Center, Temple University School of Medicine, Uniformed ServicesUniversity of the Health Services,West Read& and Philadelphia, Pa., and Bethesda, Md.

Salvage of severely ischemic limbs with common femoral-to-tibial artery bypasses represents a continuing challenge to the vascular surgeon. A subset of patients, usually diabetic, present yet another challenge. These patients are admitted with limb-threatening ischemia and a palpable popliteal artery pulse. Angiographically there is no demonstrable hemodynamicaily significant disease proximal to the popliteal artery with severe infrapopliteal occlusive disease. Between February 1984 and September 1991, 34 popliteal-to-distal tibial or peroneal artery bypasseswere performed on 30 patients: 18 men and 12 women. All procedures were performed for limb-threatening ischemia. Ages ranged from 34 to 81 years (mean, 69.1 years). Twenty-six (87%) had adult onset diabetes mellitus, 21 (70%) had hypertension, 13 (43%) had coronary artery disease,and 16 (53%) had a past history of tobacco abuse. AU patients underwent successfulbypass grafting with no operative deaths and minimal morbidity generally related only to slow wound healing. Follow-up ranges from 1 to 46 months. Patency rates have been excellent with only one graft thrombosis. One- and 3-year limb salvage rates were 94%.

We conclude that bypassgrafting originating from the popliteal artery is an excellent procedure producing results equal to those for any method of limb salvage. Transfusion guidelines for cardiovascular surgery. LAS sons learned from bloodless surgery in Jehovah’s Witnesses R. K. Spence, MD, J. B. Alexander, MD, A. J. DelRossi, MD, A. Cemaianu, MD, J. Cilley, MD, M. J. Pello, MD, U. Atabek, MD, R. C. Cam&ion, MD, Department of Surgery, Robert Wood Johnson Medical School at Camden, UMDNJ Camdtxz, N.J Cardiovascular surgical patients are among the top usersof homologous blood transfusion (HBT). Awareness of the risks of disease transmission and immune system modulation from HBT has prompted us to 6nd alternatives such as autologous predonation (APD) and intraoperative autotransfusion (IAT). However, guidelines for both HBT

Annual

meetin& abstracts

299

and APD are based on maximum surgical blood ordering schedules (MSBOS) derived from vague indications for transfusion and insistence on a perioperative 10 gm/dl hemoglobin (Hgb) level. We reviewed our experiencewith 59 patients who were Jehovah’sWitnesseswho underwent 63 elective procedures, either coronary artery bypassgrafts (CABG) (N = 27), aortic valve replacement (AVR) (N = S), or major vascular surgery (N = 28) to develop revised guidelines for MSBOS. All patients refused both HBT and APD for religious reasons. Three of 59 patients died (5.1%), two after AVR, one after PC shunt, but only one of operative bleeding complications. We conclude the following: (1) elective cardiovascular surgery can be done safely without the use of either HBT or APD, (2) HBT is not necessaryin leg bypass procedures, and (3) MSBOS guidelines for HBT in major cardiovascularsurgery can be reduced to near zero by the use of intraoperative autotransfusion and acceptanceof a postoperative Hgb nadir of 7.0 gm/dl. Treatment of iatrogenic femoral artery injuries with ultrasound guided compression Gerald Patton, MD, R. Anthony Carabasi,MD, Rick Feld, MD, Archibald Alexander, MD, and Lawrence Needleman, MD, [email protected]&v Hospital, Pbikdelph, Pa.

Color Doppler imaging has become the preferred method to evaluate patients with suspectedfemoral artery injuries after angiographic procedures. During an S-month period, 14 femoral artery injuries were identified and referred for nonsurgical treatment of their abnormality. There were nine men and five women, ranging in age from 44 to 81 years. Twelve patients had undergone cardiac catheterization, one had a cerebral arteriogram and one patient had had multiple femoral punctures to obtain blood studies. The injuries included 13 pseudoaneurysms (PA) and one arteriovenous fistula (AVF). The PASranged in sizefrom 1.4 to 7.8 cm (greatest dimension). The time from injury to diagnosis ranged from 6 hours to 16 days. Two patients had spontaneous thrombosis of their PAS at 2 and 3 days, respectively. Ultrasound guided compression (USGC) was used to treat the remaining 12 patients. The neck of the PA or AVF was identified. During color Doppler ultrasound visualization, pressure was applied with the transducer until flow into the neck of the PA or AVF ceased.At 10 minutes, pressurewas releasedand flow assessed (20 minutes for anticoagulated patients). If necessary,pressure was reapplied for lOminute intervals (or 20 minutes for anticoagulated patients) until the PA or AVF thrombosed. The common femoral artery flow was not interrupted by compression (confirmed by color Doppler). Total compression time ranged from 10 to 150 minutes. Successfulobliteration of the lesion was accomplished in 10 of 12 patients and verified by follow-up ultrasound examinations done 24 to 72 hours after procedure. No complication of the technique was encoun-

300

Annual

Journal of VASCULAR SURGERY

vneetin. absimcts

tered, and discomfort of the procedure was slight, requiring little or no sedation. Two PAScould not be occluded and were repaired surgically without incident. Successfulocclusion occurred in two of three anticoagulated patients. USGC is a safe, simple noninvasive technique that can be used to treat femoral artery injuries that traditionally were treated surgically. The technique can be applied by any laboratory possessingthe necessaryultrasound equipment and is currently the method of choice for treating these arterial injuries at our institution. Iliofemorai versus femorofemoral an individualized approach

bypass: The case for

Martin E. Harrington, MD, Elizabeth B. Harrington, MD, Moshe Harmon, MD, Harry Schanger, MD, and Julius H. Jacobson, MD, Depamnt of SUYJ~, Mt. Sinai Medical Center, New Ymk, N.Y.

The treatment of unilateral iliac occlusion remains controversial. Seventy-four iliofemoral and 144 femorofemoral bypasseswere performed over a 25-year period. Demographic characteristics of the two groups were similar. Operative indications included claudication in 21.6% iliofemoral, 29.2% femorofemoral; rest pain in 40.5% iliofemoral, 26.7% femorofemoral; and gangrene in 25.7% iliofemoral, 13.9% femorofemoral. In 16% of the femorofemoral group the indication was an occluded limb of a previous aortic procedure. Concomitant distal revascularizations were performed in 17.6Ohof the iliofemoral group and 5.6% of the femorofemoral group. Five-year primary and secondary patencies for iliofemoral were 75%, and 79%. Five-yearprimary and secondary patencies for femorofemoral were 59% and 68%. Early closure was more common after femorofemoral (14% vs 5%). Primary patency for 23 femorofemoral grafts performed for aortic occlusion was 62%, but secondary patency was 81% at 1 year with one subsequent closure. Proximal endarterectomy adversely affected secondary patency of iliofemoral (62% versus 92% at 1 year). Operative mortality rate was 6% for femorofemoral and 2% for iliofemoral. Nine wound complications occurred in the femorofemoral group, resulting in two deaths. Four patients underwent graft removal without limb loss. One wound problem in the iliofemoral group led to graft removal with limb loss. Iliofemoral had a better patency rate than femorofemoral and avoided operation on an asymptomatic limb. Femorofemoral avoids entry into the abdomen or retroperitoneum and can be performed with the patient under local anesthesia. In patients where both iliofemoral or femorofemoral are applicable, the choice between these two procedures should be individualized with these factors in mind. Transfemoral endovascular aneurysm repair

graft

placement

for aortic

Timothy Chuter, MD, Kenneth Ouriel, MD, William Fiore, MD, James A. DeWeese, MD, and Richard M.

Green, MD,

University of Rochester Medical Center,

Rochester, N.Y.

Transfemoral placement of an endovascular graft for treatment of infrarenal abdominal aortic aneurysm has become feasible as a result of advances in catheter delivery systems, intravascular stems, and intraoperative digital vascular imaging technology. Successful repair requires precise and permanent seating of the graft to isolate the aneurysm without occluding the orifices of the renal or iliac arteries. Furthermore, the insertion apparatus must be small and flexible enough to traverse tortuous iliac arteries. Our apparatus was developed with use of experimental canine and human cadaveric models. The prosthesis consists of barbed self-expanding stems attached to a woven Dacron fabric. The current miniaturized model was tested in six dogs. Grafts were inserted via the right femoral artery. Digital subtraction fluoroscopy equipment was used to guide insertion and record angiographic images. Angiograms obtained immediately after insertion revealed accurate placement of the graft, within 4.6 + 1.6 mm of the intended level. Follow-up angiograms at 1 month and 3 months showed no migration, no leakageof contrast, and patency of all six grafts. Grafts were also inserted into six cadaveric aortas (5 atherosclerotic, 1 aneurysmal; age of patients, 68.7 + 14.3 years) using the same technique. Traction was applied to sutures on the caudal end of the grafts in 50 fl increments to a maximum of 2000 8. Measurements were performed three times for each specimen. The force required to displacethe grafts 10 mm from their original position was 1388 + 124 8, which far exceedsthe expected sheer force of approximately lg on an endovascular graft, held in intimate contact with the walls of the infrarenal aorta by low profile stents. These preliminary results show that endovascular grafts can be positioned accurately and securelyin the abdominal aorta, without subsequent leakage, migration, or thrombosis. A video of the insertion procedures of both a tube and a bifurcated graft will be shown. Major

complications

after abdominal

aortic

aneurysm

sT9-Y

Jon R. Cohen, MD, and Howard Franklin, MD, Department of Surgery, Long l&and Jewish Medical Center and Albert Einstein College of Medicine, N.Y. In the last 7 years, eight surgeons repaired 342 abdominal aortic aneurysms (AAAs); 288 elective AAAs, and 54 ruptured AAAs. The purpose of this study was to review these complications and to identify specific risk factors for their occurrence. The charts of alI patients sustaining a major complication after AAA surgery in the last 7 yearswere analyzed. Of the total group, 58 patients (17%) developed major complications. The complication rate was 8% for elective AAA versus 65% for ruptured AAA (p < 0.001). The mortality rate associatedwith a complication in ruptured AAAs was 94% versus 47% for elective aneurysms

Volume Number August

16 2 1992

AnnuuJ

@ < 0.001). The elective AAA complications were as follows: limb ischemia (n = 7), postoperative bleeding (n = 6), dehiscence (n = 3), cardiac (n = 5), gastrointestinal (n = 3), and miscellaneous (one patient each; cerebrovascular accident, acute renal failure, respiratory failure, and hepatic failure). The most morbid complication was lib &hernia associatedwith a 71% mortality rate. In reviewing the possible risk factors for each specific complication, four of the sevenpatients with limb ischemia had preoperative distal infrainguinal occlusivedisease.Two of the sevenpatients requiring reexploration from bleeding had a history of previous bleeding episodes, although they had normal prothrombin/partial thromboplastin time values before operation. No specific factor was identified for the development of dehiscence or gastrointestinal problems. All of the patients who suffered cardiaccomplications had abnormal preoperative cardiograms. Elective AAA repair in our seriescarried a surprisingly high mortality rate (47%) when associatedwith any major complication. Sign&ant risk factors that could possibly identify patients with potentially preventable highly morbid complications include the following: (1) chronic infrainguinal vascular disease and (2) a history of prior bleeding episodes. A postoperative major complication after ruptured AAA is almost uniformly fatal. Aneurysm

proximal

to an in&renal

aortic graft

G. Richard Curl, MD, Gian Luca Faggioli, MD, A. Stella, MD, M. D’Addato, MD, John J. Ricotta, MD, State Universiiy of New Tork at [email protected] and University of Bologna, Buff&,

N.Y., and Bologna, Italy

Over the last 5 years we have operated on 21 patients in whom aneurysmal dilation developed, without evidence of graft infection, proximal to an aortic prosthesis. Patients were predominantly male (19 of 21) with mean age of 66 years (range, 54 to 77). Indications for the original graft were aneurysm in 13 patients and occlusive diseasein 8. Patients were admitted a mean of 10 years (range, 3 to 23 years) after their original operation with an asymptomatic mass (14), symptoms of expansion (3), frank rupture (3), or thrombosis (1). At operation 12 patients (57%) showed evidence of pseudoaneurysm whereas nine demonstrated proximal arterial degeneration alone. Operative approach was transperitoneal (13), extended retroperitoneal (4), or throacoabdominal (5). Infrarenal control was possible in seven patients, seven had clamps placed between the renal and superior mesenteric arteries, and seven had clamps placed above the celiac axis. Repair was accomplished by graft-to-graft anastomosis in 12 patients, a new graft was placed in nine cases, and seven patients required a suprarenal anastomosis with visceralreconstruction. There were five postoperative deaths (24%), including two of three ruptures, and two patients in the suprarenal group who had reexploration for bleeding. One patient died (9%) of the 11 patients operated on electively with infrarenal anastomosis. Two late deaths occurred (myocardial tiarc-

meetin

abstracts

301

tion, pneumonia), one patient required perioperative dialysis, and a graft infection developed in one at 2 years. Although aneurysmal dilation above an aortic graft is a challenging problem, elective operation can be undertaken with acceptable results. Factors associated with increased deaths include rupture and the need for suprarenal reconstruction. Good results can be expected in those patients who are admitted with false aneurysms and a normal visceralaorta. The choice of approach was dictated by the patient’s anatomy and surgeon preference and did not affect operative outcome. Treatment of upper extremity ischemia by revascularization

angioaccess-induced

Harry Schanzer, MD, Milan Skladany, MD, Moshe Haimov, MD, Depa-t of Surpry, Mount Sinai School of Medicine, New Yurk, N.Y.

Upper extremity ischemia related to arteriovenous fistula (AVF) or bridge AVF (BAVF) constructed for hemodialysis is a relatively infrequent but serious complication. Distal steal, induced by the fistula, is the underlying pathophysiologic mechanism. Methods directed at improving distal perfusion by increasing fistula resistance (banding, elongation) often are unsuccessful in correcting the ischemia or result in accessthrombosis. This work presents our experience with a new technique that consist of (1) ligation of the arterv iust distal to the takeoff of the AVF/BAVF and (2j ‘arterial bypass from the artery proximal to the takeoff of the A*/BAVF to the arte& &&to th.- limt ion. Thirteen patients (mean age, 58 + 18 as_“6” years, all diabetic, 9 women) underwent this procedure. Eleven patients had polytetrafluoroethylene (FIFE) brachioaxillary BAVF, one patient had a radiocephalic AVF, and one patient had a brachiocephalic AVF. All patients had rest pain, and five of them had gangrenous changes. Symptoms were present immediately after construction of the accessin nine cases.The remaining four caseshad late onset of ischemia (range, 5 months to 5 years). Bypass material used was great saphenous vein in 12 casesand PTFE in one case.Twelve patients had complete resolution of symptoms immediately after the procedure and eventual healing of the gangrenous lesions. One patient with extensive gangrene before the procedure required hand amputation after 10 months, the bypass having failed at 3 months (thrombectomized) and 10 months. One patient died 7 months after the revascularization, with both access and bypass patent. The remaining patients have their bypass patent (mean follow-up, 16.8 months, range, 1 to 51 months). One patient had her BAVF thrombosed at 1 month, and in the remaining, the accesscontinues functioning unaffected by the revascularization. In conclusion, this technique produces significant and durable hemodynamic and clinical improvement in arms affected by access-inducedischemic steal with minimal morbidity and does not affect the normal life of the angioaccess.It is the technique of choice for correction of AVF/BAVF-induced ischemic steal.

302

Annzral

Journal of VASCULAR SURGERY

meeting abstracts

Experience with vascular reconstruction for dialysis in patients who are HIV positive compared with patients who are HIV negative James S. Brock, MD, Thomas S. Riles, MD, Patrick J. Lamperello, MD, Gary Giangola, MD, Mark A. Adeltnan, MD, Ronnie Mintzer, BS, F. Gregory Baumann, PhD, New York University Medical Center, New York, N.Y. It has been suggested that patients who are HIV positive may have a significantly increased incidence of infection and occlusion of arteriovenous reconstructions performed for dialysisaccess.This study examined hospital and follow-up data on 108 patients who had such reconstructions before undergoing dialysis between January 1989 and June 1991. Patients were divided into three groups: 18 (16.7%) who were HIV positive but asymptomatic (group I), 11 who were HIV positive and symptomatic (group II), and 79 (73.2%) who were HIV negative (group III). There were no significant differences between groups in the incidence of diabetes, but in groups I and II a history of intravenous drug abuse (83.3% and 63.6%) and use of prosthetic grafts (72.2% and 63.6%) were significantly more frequent than in group III (drug abuse 3.8%, prosthetic graft 55.7%). Infection of the vascular reconstruction was significantly more common in patients who were HIV positive (group I, 42.1%; group II, 27.3%) compared with patients who were HIV negative (group III, 13.9%) (p < 0.05). However, vascularreconstruction occlusion rates were high overall (39.896, 43 of 108) and did not differ significantly between group I (44.4%), group II (27.3%), and group III (40.5%) @ > 0.05). These results show that although the rate of infection of vascular reconstructions for dialysis in patients who are HIV positive is significantly higher than for patients who are HIV negative, the rates of occlusion are not significantly different. Therefore, the diagnosis of HIV in a patient should not of itself preclude the possibility of vascular reconstruction for dialysis access. Therapeutic and prophylactic vena caval interruption for pulmonary embolism: Caval and venous insertion site patency Ali F. AbuRahma, MD, Patrick A. Robinson, MD, Yale D. Conley, MD, Kevin R. Snodgrass, MD, Todd A. Witsberger, MD, Daniel J. Wood, MD, JamesP. Boland, MD, and Robert C. Co&ran, MD, West ‘virginiu University Health Sciences Centm/Cbarlestun Area Mediuzl Center, Charleston, W Va., ryiur Central Research, Groi%n, Cmn.

Ninety-two filters and 69 Adams-DeWeese dips were used in 161 patients. Fifty-four percent were done prophylactically. The operative mortality and morbidity rates were 0% and 3.3% for the filter and 8.7% and2.9% for the clip, with no procedure-related deaths. The late caval patency rate (mean, 20 months) as documented by duplex ultrasonography/venography was 100% for patients with filters and 88% for patients witb clips @ = 0.011). Seven percent of the patients with filters and 20% of the patients

with clips had late limb swelling after operation (p = 0.05). The incidence of recurrent late pulmonary embolism (PE) was 2.5% in the filter group and 1.9% in the clip group. In the filter group, 10% of patients had postoperative thrombosis at the femoral vein insertion site and 0% at the jugular vein insertion site. Eighty-eight high-risk patients underwent prophylactic caval interruption (54 clips and 34 filters) with major abdominal operations. The operative mortality and morbidity rates in this group were 5.7% and O%, with no procedure related-deaths or postoperative PE. The late caval patency rate was 100% for patients with filters and 91% for patients with clips. Eighteen percent of the patients with clips and 6% of the patients with filters had late limb swelling after operation. In conclusion, (1) the filter has a better caval patency than the clip, (2) the jugular vein has a better patency than the femoral vein for filter insertion, and (3) the incidental use of filter for PE prophylaxis in major abdominal operations in high risk patients is safe and preferable; however, the use of the clip must be weighed against the overaIl morbidity of this device. Natural history, duplex characteristics, and histopathologic correlation of arterial injuries in a canine model: Which lesions require operation CliffordM. Sales,MD, Michael L. Schwartz, MD, Thomas F. Panetta, MD, Michael L. Marin, MD, Anne M. Jones, BSN, Kurt R. Wengerter, MD, and Frank J. Veith, MD, Montefiore Medical Centeq/Alb& cine, New Ymk, N. Y.

Einstein

Collge

of Medi-

The role of nonoperative management of arterial trauma remains controversial. Duplex ultrasonography (DUS) has been advocated to define the nature of traumatic arterial lesions and to help select patients for operation. However, the validity of this method and its value in defining indications for operation remain questionable. We, therefore, studied a canine model to correlate the natural history, DUS findings, and histopathology of different arterial injuries to define criteria for operative intervention. Fifty-two canine femoral and carotid arteries were randomized to surgically created intimal flaps (n = 15), crush injuries (n = 15), lacerations (n = 15), or control groups (n = 7). An experienced sonographer, blinded to the presence or type of injury, evaluated the vesselsweekly for 1 month. All vesselswere perfusion &red for pathologic examination at the time of sacrifice 1 month after injury. Duplex ultrasonography was accurate for detecting intimal flaps (sensitivity, 93%). Thirteen vessels with intimal flaps remained patent (87%). Three flaps were incorporated into newly formed stenotic areas that were histologically identified as neointimal hyperplasia. All arteries subjected to crush injuries demonstrated abnormal duplex findings (velocity change, 100%; intimal defects, 94%; dilation, 89%), and none of these vesselshealed. Crush injuries resulted in adventitial hemorrhage, intimal

Volume 16 Number 2 August 1992

Annual

and medial hyperplasia, and endotheliahzation of irregular surfaces. Seventy-three percent of arterial lacerations appeared normal on DUS within 5 daysof injury. AU but two (13%) were normal, microscopicalIy, at the time of sacrifice. Overah, DUS 8ndings correlated weIl with histopathology. The natural history of arterial trauma varies with the mechanism of injury. This study confirms the accuracyof DUS in defining various injuries. Arterial lacerations, without active bleeding or pseudoaneurysm formation, may be safelyobserved becausethese injuries generally heal without stenosis.Arterial injuries with DUS characteristics of healing rarely occlude and may not require repair; however, those with significant or progressive stenosis necessitateoperative intervention. These data suggest a role for nonoperative management of arterial injuries based on DUS characteristics and indicate a need for clinical trials. Mapopliteal arterial injury: Prompt revascul~tion &ords opportunity for limb salvage Frank T. Padberg, MD, Robert W. Hobson, MD, Joseph J. Rubelowsky, MD, Sing C. Lee, MD, Vincent J. MiIazzo, MD, and Juan J. Hernandez-Maidonado, MD, UX4DNJNew Jersey Medical

School, Newark, NJ

Fifty patients with inti-apophteal arterial injury (28 penetrating, 22 blunt) were evaluated to determine the factors involved in limb salvage.Management consisted of primary repair (3), bypass with reversed saphenous vein (24), ligation (9), observation (lo), and other (2). Twenty-seven penetrating injuries were from gunshot wounds and frequently involved isolated fib&r fracture. Patients with penetrating injuries have a good prognosis; despite limb-threatening ischemia, treatment resulted in one amputation (4%). Twenty-two blunt injuries resulted in 10 amputations (44%), two of which were performed primarily. There was no limb loss from observation or ligation of single infi-apophteai arterial injury. Fifteen of 26 ischemic limbs were salvaged by revascularization; six for blunt injury. Outcome after graft failure was amputation in aII cases.Infrapopliteal vascular injuries are associatedwith high amputation rates, particularly from blunt injury. Although primary amputation is indicated when severe ischemia is associated with fractures and soft tissue or neural injury, an aggressiveapproach to revascuiarization with autogenous bypass grafts is recommended. The preferential use of the external iliac artery as an inflow source fix redo femoropopiiteal and infiapopliteal bypass operations Enrico Aster, MD, Jon Kirwin, MD, Chittur Mohan, MD, and Mark Gennaro, MD,M&um~ssM Center, State Univm’@ of New T&r&,Brootiy, N.Y. It is weil known that reoperations involving the inguinal region are associatedwith a sigr&cant morbidity, which includes infection, lymphatic obstruction, lymphorrhea, and neurovascuku injury. Furthermore, we have

metin&

ahtvacts

303

previously shown that reoperations on prosthetic grafts carry up to a fourfold increasein graft infection rate when the groin is reexplored. To prevent these potentially important complications we have avoided a redo groin incision in 38 severelysymptomatic patients who experienced primary (23 cases)or secondary (15 cases) femoropophteal bypass thrombosis over the last 4 years. AIl patients were candidates for prosthetic bypassesbecauseof lack of a suitable vein. Twenty-nine external ihac-topopliteal bypasses (18 above-knee; 11 below-knee) and nine external iliac-to-infi-apopliteal bypasses (5 anterior tibial; 2 posterior tibial; 2 peroneal) were performed with 6 mm polytetratluoroethylene ringed grafts in 28 patients. Adjunctive distal arteriovenous fist&s were constructed in ail infrapopliteal bypasses.Our surgical technique to avoid the scarred groin is simple and can be expeditiously performed. The external iliac artery was exposed via a standard retroperitoneal approach. The second incision was placed just below the scarredarea and deepened to the level of the medial border of the sartorius muscle. A tunnel connecting both incisions was created by blunt dissection alongside the lateral border of the femoral artery. Four pophteal bypassesoccluded at 4, 6, 10, and 28 months, after operation. The remaining 25 casesare patent (mean, 24 months), Three of the infrapophteal bypassesoccluded at 0,2, and 3 months, after operation. The remaining six casesare patent from 2 to 18 months (mean, 12 months). Only one patient developed a superficialwound infection at the below-knee pophteal incision that healed with conservative treatment. All other patients had an uneventll postoperative course. Thus we believe this approach to be simple, safe and durable and should be used preferentially to avoid the di&uIt and hazardous dissection of a previously operated groin.

Central and regional hemodynamics determine optimal management of major venous ‘injuries P. V. Pathanjali Sharma, MD, Rao Ivatury, MD, Ronald Simon, MD, and Alexander Vinzons, MD, Lincoln Medic& andMental N.T.

Health Cente$New York Medical CoUge, Bronx,

Controversy continues in the selection of optimal management of major venous injuries. We analyzed our experience with 191 major venous injuries in 163 patients (1986 to 1991). The mechanism of injury was gunshot wounds (112),stabwounds (44), blunt (6), andiatrogenic (1). One hundred five of the 191 injuries (54.9%) were repaired (lateral repair 76, end-to-end anastomoses15, vein patch 4, vein grafts 5, Gore-Tex [w. L. Gore and Assoc., Elkton, Md.] graft 4, compilation graft 1). Twenty-one of 22 died before treatment. Twelve of 17 had arterial injury. Forty-two of the 163 (25.8%) patients died of severeassociatedtrauma. The decision to ligate or repair venous injury correlated with hemodynamic stability and in stable patients with (1) extent of injury and (2) clinical and measured venous hypertension (stump pres-

304

Annual

meetin

Journal of VASCULAR SURGERY

abstracts

sure). Venous ligation in hemodynamically stable patients did not increasethe need for fasciotomy or amputation. It is concluded that optimal treatment choice primarily depends on hemodynamic stability. In stable patients the extent of venous and associated injury and venous hypertension should dictate the need for repair. Management of peripheral allograft recipients

vascular problems

in cardiac

Alan I. Benvenisty, MD, George Todd, MD, Craig Smith, MD, Eric Rose, MD, and Keith Reemtsma, MD, Columbia Uniiwsi~ CoUge of Physicians and Surgemrs, Department of Surgery, New York, NY.

Five hundred twenty consecutive cardiac transplants (458 adult, 62 pediatric) were reviewed to assessthe impact of peripheral vascular (PV) problems in cardiac allograft recipients. Peripheral vascular complications occurred in three children. Ischemia as a result of intraaortic balloon pump (IABP) in one resolved with removal of the device. In a second, a large falseaneurysm of the external iliac artery was repaired after routine postoperative angiography. Another child required repair of a percutaneous internal jugular venotomy. Four adults required operations for complications of intraaortic balloon pump or femoral puncture (1 fasciotomy, 1 thrombectomy, 3 direct artery repair). Two of 16 atherosclerotic adults with PV disease before operation had pretransplant operative procedures (resection abdominal aortic aneurysm and iliofemoral bypass. Three had abdominal aortic aneurysms repaired 0.3, 1, and 5 years after operation. Three of 12 patients with preoperative claudication had chronic aortic occlusions, one required axillobifemoral grafting. Two patients required infrainguinal arterial reconstruction (one ischemic rest pain and one disabling claudication 40 and 6 months after transplant). The remainder of the patients did not require any reconstruction after transplant with improved or stable symptoms. There were no limb losses and no infections. Although there were sixdeaths in the PV disease group, none were related directly to PV disease. We conclude that despite the rigors of strenuous immunosuppression, patients with manifestations of PV disease including aneurysmal and symptomatic occlusive disease may be treated by cardiactransplantation without increased cost to morbidity directly related to the PV disease. Lipoprotein, is a risk factor for peripheral vascular disease (ND): The influence of coronary artery disease, cigarette smoking, and diabetes

Mark Widmann, MD, and Bauer E. Sumpio, MD, PhD. Department of Sqpy, Tide Unihmity Schoolof Medick, New Hapen, Corm.

An association between coronary artery disease(CAD) and elevated levelsof lipoprotein a (LP,) hasbeen reported, but the importance of Lp, as a marker for peripheral vascular disease (PVD) is unknown. Furthermore, the effects of coexistent risk factors, such as CAD and diabetes

mellitus (DM), which may independently affect LP, values, needs to be assessed.To study this, Lp, (mg/dl * SD) was measured by ELISA in triplicate in white male patients with (n = 50; mean age, 67.6) and without PVD (n = 50; mean age, 68.3 years). Patients with PVD had a significant elevation (p = 0.04) of LP, (27.3 t 3.9) compared with patients without PVD (20.0 * 2.9). The incidence of cigarette smoking (CIG), CAD, or DM was significantly higher (p < 0.03) in patients with PVD, but there was no significant difference in LP, in patients with CIG (26.2 -C 2.9, n = 77) compared with those without CIG (20.4 -+ 4.0, n = 23) or in patients with CAD (29.9 2 4.1, n = 41, p = 0.3) compared with those without CAD (21.4 * 3.0, n = 59, p = 0.08). Patients with DM had significantly Cp = 0.04) lower levelsof LP, (17.8 + 3.5, n = 24) than those without DM (27.1 rf: 3.0, n = 76). We conclude (1) Lp, is increasedin patients with PVD and may be a stronger marker for PVD than for CAD; (2) Despite the high incidence of CAD and CIG in patients with PVD, these factors did not significantly affect LP, in this study; and (3) patients with DM may have a lower distribution profile of Lp, and may represent a novel finding. Comparison of descending phlebography with quantitative photoplethysmography, air plethysmography, and duplex quantitative valve closure time in assessing deep venous reflm

Harold J. Welch, MD, Eleni C. Faliakou, MD, Robert L. McLaughlin, RVT, Susan E. Umphrey, RVT, Michael Belkin, MD, Thomas F. O’Donnell, MD, Tufts-Neau England Medi&

Center, Bostun, Mass.

The descending phlebogram is the “‘gold standard” for determining the degree of deep venous reflux and is necessaryin the potential selection of patients for deep venous reconstruction. We compared the invasive “gold standard” with three available noninvasive tests; quantitative photoplethysmography (QPG), venous tilling index (VFI) as measured by air plethysmography (APG), and valve closure time (VCT). Descending phlebography was performed with the patient in 60-degree semiupright position with Valsalva. We examined a total of 25 limbs; eight limbs in four healthy, normal volunteers (group N [normal]), nine limbs in eight patients with grade l-2 phlebographic re5ux (group M [mild]), and eight limbs in sevenpatients with grade 3-4 phlebographic reflux (group S [severe]). SVS/ICVS clinical stage (O-normal, 3-ulcer) did not differ between group M and group S. Similarly, QPG, although normal in group N, did not differ between group M and group S. VFI as measured by APG was unable to differentiate mild from severe reflux, but did show a significant difference between group N and group S. Duplex VCT in the superficial femoral vein did not show a difference between the three groups. There was, however, a statisticallysignificant difference in the popliteal valve and total (sfv + pop) VCI between group M and group S.

Volume 16 Number 2 August 1992

,

@ < 0.001 by analysisof variance and Duncan’s Multiple Range Test) A receiver operating characteristics curve analysis, however, shows TVCT to be the best test to identify severe venous reflux. A total (sfv + pop) valve closure time of 2 4 seconds is the optimal discriminating value in identifying severe deep venous reflux with a specificity of 9496, sensitivity of 88% and accuracyof 92%. In the global assessmentof lower extremity venous reflux, QPG and APG are able to differentiate normal limbs from those with reflux, but fail to determine the degree of reflux. By contrast, TVCT as measured by duplex, is a sensitive and specific noninvasive method for determining the degree of reflux and thus is able to selectthose patients who would benefit from deep venous reconstruction. Complications of bypass operation for radiationinduced arterial disease Gordon Phillips, MD, Richard M. Peer, MD, James F. Upson, MD, John Ricotta, MD, State Uniaersity @Nets York, at Bujhb,

N.Y.

Vascular reconstruction for radiation-induced arterial diseasehas the potential for serious complication. Over a 14-year period 20 patients, mean age 55 years, underwent revascularization for radiation-induced arterial obstructive disease.Three to 13 years (mean, 7) lapsed between initial radiation therapy and the development of arterial insufliciency. An average of 5000 rads was delivered for malignancies of the following origin: gynecologic (8), lymphoma (5), testicular (l), head and neck (6), and lower extremity sarcoma (1). Arterial occlusive diseaseoccurred in the iliac (7), common femoral (5), super&ial femoral (l), renal (l), and carotid (6) arteries. Presenting symptoms were claudication (7), rest pain and nonhealing ulcers (7), renal insufhciency (l), and transient ischemic attacks (4). Arterial reconstruction included aortobifemoral bypass (6)) femoral-femoral bypass (2)) femoral patch angioplasty (2), aortorenal bypass (l), axillofemoral bypass(2), carotid endarterectomy (3), carotid patch angioplasty (3), and femoral popliteal (1). In this group of patients no major wound complications or other major early morbidity occurred. Five patients developed late graft infections manifested 2 to 5 yearsafter operation. All occurred in areaswhere the bypass graft passed through previously irradiated tissues. Presenting symptoms included a draining groin sinus (3) and soft tissue abscess (2). There were three infected aortofemoral grafts; one femoral femoral graft and one femoral popliteal bypass. In all casesthe graft did not incorporate into the surrounding tissues when passing through the irradiated area. Treatment included graft excision and extraanatomic bypass through nonirradiated tissues.One patient died of systemic sepsis. From this experience we conclude that vascular reconstructive surgery can safely be performed for radiationinduced arterial disease.However, there is a high incidence of late graft infection, 25% in this series,when the bypass traverses previously irradiated tissue. If possible, arterial

Ann&

meetin. abs~acts 305

reconstruction should avoid the irradiated areasby use of extra anatomic bypass if necessary. Use of autologous conduits and aggressivesurveillance may decreaseassociated morbidity. Carotid endarterectomy: A safe, approach Paul E. Collier, MD, Se&blq: Pa.

Cost-eflicient

The DRGs have encouraged physicians to become more efficient in the care of their patients; often, however, raising the question of safety. For 3 years all patients undergoing carotid endarterectomy were monitored in the intensive care unit (ICU) for 24 hours, and most were discharged on the second postoperative day. After review of these patients’ hospital records and direct patient interviews, it was clear that many patients did not require the ICU and could be discharged on the first postoperative day. In January 1991 a prospective policy was established to evaluate the safety and efficacy of outpatient arteriography, sameday admission, selective use of the ICU, and early discharge on first postoperative day when feasible. Over a lo-month period all patients undergoing carotid endarterectomy at our institution were evaluated (N = 52). Eleven patients had suffered a prior stroke (2 1%), 3 1 had either amaurosis fugax or transient ischemic attack (60%), and 10 were asymptomatic (19%). Fortynine of the patients had their arteriogram performed as an outpatient or during a prior admission and were admitted on the day of operation. Nine patients had general anesthesiaand had shunts placed, whereas 43 had cervical block, 8 of whom had shunts (19%). Only five patients required the ICU for either hypertension or neurologic complication (1 transient ischemic attack and 1 minor stroke). Forty-six patients (88%) were discharged on the first postoperative day, average length of stay, 1.29 days/patient. All patients were seen 3 to 5 days after discharge. There were no readmissions for hypertension, hemorrhage, or cardiac or neurologic events. No deaths occurred in the 30-day period. The hospital charge was over $1900 below the DRG reimbursement on average. From this series it is concluded that by following this protocol, a short stay with selectiveuse of the ICU is safe and cost-effective for patients undergoing carotid endarterectomy. Routine postendarterectomy duplex surveillance: Does it prevent late stroke? William C. Mackey, MD, Michael Belkin, MD, Rakesh Sindhi, MD, Harold J. Welch, MD, Thomas O’Donnell, MD, Tufts-New En&and Medical Center, Boston, Mass. A recent finding that lessthan 50% of late postendarterectomy strokes are related to recurrent carotid stenosis (RCS) led us to question the utility of routine postendarterectomy duplex surveillance (RPCEADS) as a means of late stroke prevention. To evaluate our program of RPCEADS we reviewed our postoperative duplex scans

306

Annd

meeting abstracts

and correlated their resultswith clinical data. A total of 949 postoperative scanswere carried out on 306 carotid arteries (219 patients) (3.1 * 0.1 studies/artery) over an average follow-up of 52.9 + 2.4 months. Lessthan 50% RCS was documented throughout follow-up in 258 (84.3%) arteries. Recurrent carotid stenosis greater than 50% in either the common or internal carotid artery was noted in only 48 arteries (15.7%). The interval to maximal RCS was 5 36 months in 23 (47.9%) and >36 months in 25 (52.1%) cases. Thus 41 out of 48 (85.4%) arteries with RCS 2 50 remained asymptomatic or underwent reoperation for recurrent symptoms over long-term follow-up. Only two patients suffered unheralded stroke as a result of RCS, and five patients underwent prophylactic reoperation. Assuming that a repeat endarterectomy would have prevented these two strokes and was effective prophylaxis in the additional five, at most 7 of 306 arteries (2.3% of the entire group), could have derived identifiable benefit from RPCEADS. Given the expense, resource use, and meager benefit in late stroke prevention, routine duplex scan surveillance after carotid endarterectomy is not justified. Plaque progression and onset of neurologic symptoms in asymptomatic patients with significant carotid stenosis: VA cooperative study No. 167 Wiiard Johnson, MD, Robert Hobson, MD, J. Dennis Baker, MD, David Weiss, MD, and the VA Cooperative Aymptomutic

Carotid

Artery

Stenosis Study Group, Cooper-

ative Studies Proyam, Deparhnent of Veterans A&k, Periypoint,

Md.

Journal of VASCULAR SURGERY

The origin of neurologic symptoms in patients with significant carotid stenosis remains an enigma. A randomized prospective clinical trial was initiated in 1983 to evaluate the efficacy of carotid endarterectomy in the reduction of the risk of neurologic symptoms in asymptomatic patients with a significant carotid stenosis (diameter reduction greater than 50%). Patients were randomized to either carotid endarterectomy and aspirin or only aspirin. The outcome of this study has been presented elsewhere and suggests a significant reduction in the incidence of neurologic events in patients treated by carotid endarterectomy. During this study, 57 neurologic events occurred in 233 patients treated with only aspirin. We have chosen to evaluate the degree of stenosisand its progression over time and the relationship to the development of neurologic symptoms. All patients had serial Gee-OPG studies. As an indicator of stenosisand its progression, we have used the index of eye systolic pressure/arm systolic pressure, a decreasein the index of 0.05 was considered indicative of progression of plaque stenosis. Plaque progression was not a reliable predictor of neurologic events in patients treated with aspirin. Hence, a policy to restrict carotid endarterectomy to those patients with plaque progression is not as efficacious as an early endarterectomy once a significant carotid stenosis is identified.