The Scientific Program of the Society for Vascular Surgery and the International Society for Cardiovascular Surgery, North American Chapter, was held in Boston, Mass., on June 3-5, 1991. Selected abstracts from that meeting are published here in order of presentation at the meeting.
From the Scientific Program of the Thirty-ninth Annual Meeting of the North American Chapter of the International Society for Cardiovascular Surgery
of new methods of expressing aortic
dimensions, whereas 26 of 100 elective AAAs (26%) had ratios less than 1 : 1. These data suggest that diameter alone is not the most useful parameter in assessing an aortic aneurysm. A parameter indexed to the body size appears to be a better discriminator, such as the ratio of the AAA diameter to the vertebral body transverse dimension. Use of this method of expression may more accurately predict the risk of rupture, reserving elective aneurysm resection for patients with a ratio approaching 1 .OO.
The natural tic aneurysms
Peter M. Brown, John R. Gutelius,
Ruth Pattenden, MD, and Queen’s University, Enngston,
Kenneth Ouriel, MD, R. M. Green, MD, Carlos E. Donayre, MD, J. A. DeWeese, MD and Janice Elliott, RTR, University of Rochester Medical Center, Rochester, AT
This prospective review was designed to assess the natural history of small abdominal aortic aneurysms and thus the appropriateness of surgical replacement. Our unit is the only tertiary care unit for 85% of 400,000 people and virtually the only care center for approximately 150,000. Thus our prospective study has a defined population base. Our protocol of observing all aneurysms under 5 cm in maximal diameter was established in 1976. In late 1982 this policy was applied to all patients seen by the vascular unit. By Dec. 31, 1987,216 patients had been entered into the study. They were followed until June 30,1990, and had at least two aneurysm sizings by ultrasonography, CT scanning, or both. The minimum follow-up period is 30 months. Operation was performed (94 patients) if the aneurysm reached 5 cm (75%), expanded by greater than 0.5 cm in 6 months (15%), or developed significant aneurysmal or occlusive symptoms (10%). In this group the average annual increase in diameter was 0.8 cm, and about 45% met the criteria for surgery within 2 years of entry. One hundred twenty-two patients have been followed without operation for an average of 45 months. No aneurysm rupture has occurred in this group. The average annual increase in diameter was 0.25 cm. Thirty patients of this group have died of other causes during the observation period. This study adds additional information concerning the natural history of abdominal aortic aneurysms. Unlike many other studies it reflects a defined population and regional base. These results support a policy of observation for aneurysms under 5 cm in maximal transverse diameter.
It has been commonly assumed that the most significant risk factor predicting rupture of an abdominal aortic aneurysm (AAA) is its diameter, usually expressed as the maximum transverse dimension in centimeters. The variability of patient body size, normal aortic size, and saccular versus fusiform aneurysm configuration made us uncomfortable when using raw diameter measurements as the sole anatomic factor in the determination of whether a patient should undergo aortic aneurysm resection. We reviewed preoperative CT scans of 100 patients undergoing the elective resection of AAAs and compared these with CT scans of 18 patients who were admitted with ruptured AAAs. Precise measurements were made of the aneurysm diameter and craniocaudad length. The transverse diameter of the L-3 vertebral body was used as an index of patient body size, and the diameter of the aorta at the supraceliac, infradiaphragmatic location (SCA Dia) was used as an index of aortic size. Abdominal aortic aneurysm diameter did not accurately discriminate between nonruptured and ruptured A&As. The diameter of one ruptured AAA was only 3.5 cm, and 3 ruptured AAAs (17%) had diameters of between 4.0 and 4.9 cm. Saccular aneurysms (high AAA diameter: AAA length ratio) were not more common in the ruptured group. By contrast, longer, more fusiform (low AAA diameter: AAA length ratio) AAAs were more prone to rupture. The ratio of AAA diameter to the vertebral body transverse dimension was the best discriminator between the two groups. No ruptured AAAs were observed with a ratio of less than 1: 1 between the AAA and vertebral body 902
Volume 13 Number 6 June 1991
Wound complications of the retroperitoneal approach to the aorta and iliac vessels Mark P. Honig, MD, Robert A. Mason, MD, and Fabio Giron, MD, PhD, Univemity Hospital SUNY, Stony Brook, NY
Repeated complaints of incisional pain prompted this review of 113 consecutive vascular operations involving a retroperitoneal approach to the aorta and/or iliac vessels. A flank muscle splitting incision was used in 53 patients when the procedure involved the terminal aorta or iliac arteries. Two types of muscle-dividing incisions (MDI) were also used. A twelfth rib or eleventh intercostal space incision (MDI-1) was carried out in 41 patients to approach the infrarenal aorta. In 19 additional patients with suprarenal aortic disease, an incision through the eighth, ninth, or tenth interspace with division of the diaphragm (MDI-2) was used. All patients were examined (follow-up period 3 to 48 months) to evaluate the incidence of prolonged incisional pain, lumbosacral neuritic pain, incisional hernia, or deforming abdominal bulge. The two types of MD1 for exposure of the aorta were associatedin a high percentage of caseswith an abdominal bulge and postoperative disabling pain. Although a retroperitoneal exposure may be the preferred or the safest approach to certain lesions of the aorta, its routine use should not be recommended when the proposed operation can be carried out equally well by the conventional midline transperitoneal incision. The role of beta-lactamase vascular surgery: cetioxime
stability in prophylaxis versus cefazolm
William H. Edwards, Jr., MD, Allen B. Kaiser, MD, Douglas S. Kernodle, MD, et al., St. Thomas Hospital 8 Vanderbilt
School of Medicine,
Cefazolin has proved inferior to cefamandole in prophylaxis of Staphylococcus aweus wound infections after cardiac surgical procedures. The susceptibility of cefazolin to hydrolysis by S. ltztrew B-lactamasesis a theoretic explanation of the observed differences. Accordingly, we undertook a prospective, randomized trial of cefirroxime, a highly B-lactamase-stablecephalosporin, versuscefazolin, the current c‘standard” of prophylaxis in clean vascular surgery, Prophylaxis was administered just before, and for 24 hours after surgery. Pharmacokinetic studies on the first 30 patients verified the adequacy of the dosing regimens. Three deep wound infections (of which two were S. aweus) occurred among 315 cefazolin recipients versus seven infections (five S. aureus) among 301 cemroxime recipients (infection rates of 0.95% and 2.3%, respectively, p = 0.2). During the time period of the study (July 1988-November 1989), a total of 15 S. aweus wound isolates were identified. All isolates were susceptible to both cephalosporins by standard criteria, however, MICs to cefazolin were significantly lower than cefuroxirne, 0.5 versus2.0 mc&nl (p < 0.01 Mann-Whimey). Superior in vitro activity of cefazolin was also reflected in time-kill kinetic studies.
In conclusion, the trend of infection rates suggested cefazolin to be more effective, although we noted no significant difference in efficacy of prophylaxis. The B-lactamasestability of the cephalosporins did not prove clinically important. Subtle in vitro differences in the susceptibility of S. aweus may prove to be of clinical relevance. The impact of selective use of dipyridamole-thallium scans and surgical factors on the current morbidity aortic surgery
Richard P. Cambria, MD, David C. Brewster, MD, William M. Abbott, MD, Gilbert J. L’Italien, BS, Joseph J. Megerman, PhD, and Glenn M. LaMuraglia, MD, Mamzcbusetts General Hospital, Boston, Mass. Cardiac risk stratification with dipyridamole-thallium scanning (D-THAL) has increased the safety or aortic reconstruction (AR). Prior work in our unit suggested such testing could be limited to patients with certain clinical markers of coronary artery disease (CAD), but others continue to advocate its routine use before AR. The present study evaluates the efficacy of this selective policy and assessedthe combined impact of CAD clinical markers and surgical factors on the current morbidity of AR. Two hundred two ( 15 1 abdominal aortic aneurysm, 5 1 aortoiliac occlusivedisease)consecutively performed, elective AR from l/1/89-6/30/90 were reviewed. Demographic data, previously defined clinical markers of CAD (diabetes, age 2 70 yrs, Q wave on ECG, history of ventricular arrhythmia (VEA) , history of angina), details of preoperative cardiac testing and surgery were recorded. End points included all major cardiopulmonary complications and/or deaths, and increased surgical morbidity (2 2 days in ICU and/or postoperative stay > 12 days). Surgical factors included prolonged ( 2 5 hr) operation or cross-clamp ( z 1.5 hr) times, complex (suprarenal clamp or added aortorenal graft) operation, and excess( 2 1500 ml) blood loss. CAD clinical markers and surgical factors were then analyzedwith stepwise logistic regression for the prediction of end points with significance assigned at the p < .05 level. Preoperative D-THAL was performed in 29% of all patients. Twenty percent of 146 patients (72% total) with O-l CAD markers had D-THAL, but > 50% of these scans were normal. Fifty percent of the remaining 55 patients with z 2 CAD markers had D-THAL of which 2/3 were positive. Treatment of CAD (coronary artery bypass grafting or percutaneous transluminal coronary angioplasty) was performed before operation ( 5 1 yr) in 9% patients. After such treatment overall mortality rates of AR was 2Oh with one cardiac related death, and major CP complications occurred in an additional 8.5%. Variables attaining significance in predicting death or CP complications included operation for aortoiliac occlusive disease, history of VEA, prolonged operative time, and excess blood loss. Increased surgical morbidity was strongly correlated with surgical factors alone. The combination of
Journal of VASCULAR SURGERY
L 2 CAD markers and 2 2 surgical factors predicted both death&P complications (p c 0.02) and increased surgical morbidity (p < 0.001). Selective use of the D-THAL scan based on certain clinical markers of CAD will identify the approximately 10% of patients admitted with AR in whom preoperative invasive treatment of CAD is appropriate. Thereafter the overall morbidity of AR, albeit minimal, is dominated by surgical factors rather than the extent of antecedent CAD.
Silent myocardial ischemia monitoring predicts late as well as perioperative cardiac events in vascular surgery patients Peter F. Pasternack, MD, Eugene A. Grossi, MD, F. Gregory Baumann, PhD, et al. Nap 2?& University Medical Center, New Ywk, AT Perioperative silent myocardial ischemia (MI) monitoring has been shown to be a predictor of perioperative myocardial infarction in patients undergoing vascular surgery, but the value of silent ischemia as a predictor of late cardiac events in such patients is unknown. Therefore after perioperative silent ischemia monitoring was performed on 424 patients undergoing elective vascular surgery, the survivors were followed for cardiac events for a mean interval of 23 months. At follow-up there were 19 cardiac deaths and 27 noncardiac deaths, and 378 patients were still alive. Nineteen patients suffered a nonfatal MI, and 5 patients suffered a fatal MI during this period. Angina occurred in 77 patients, a new onset in 16. Fifty patients underwent cardiac catheterization with 13 patients undergoing coronary artery bypass grafting and 18 patients undergoing coronary angioplasty during the follow-up period. A significantly higher incidence was found of fatal and nonfatal late MIS in those patients who suffered perioperative silent ischemia lasting 2 1 hour (n = 71) or who had 2 10 episodes of silent ischemia (n = 75) (p < 0.001). In addition, actuarial survival from late cardiac death was significantly worse for patients who met either of these criteria for perioperative silent ischemia compared to patients who suffered less silent ischemia (p c 0.01). Thus silent myocardial ischemia monitoring during the perioperative period in vascular surgery patients appears to provide valuable prognostic data for both perioperative MI and late MI and cardiac death.
The incidence of perioperative myocardial infarction in general vascular surgery Lloyd M. Taylor, Jr., MD, Richard A. Yeager, MD, Gregory L. Moneta, MD, Donald B. McConnell, MD, and John M. Porter, MD, oregmt Health Sciences University, Purtland, Ore. In recent years considerable attention has focused on preoperative detection and prophylactic treatment of coronary artery disease in patients with peripheral vascular disease (PVD) in an effort to reduce the incidence of perioperative myocardial infarction (MI), which reportedly
occurs after as many as lo%-15% of PVD procedures. It is curious that the modern incidence of perioperative MI in general vascular surgery is poorly documented. Such information from a large patient group appears critical for the accurate evaluation of the results of aggressive coronary intervention programs and forms the basis for this report. All patients undergoing general vascular surgery for 1 year on our combined University Hospital/Veterans Hospital vascular service were prospectively monitored with serial CPK isoenzymes and ECGs. Only patients being considered for elective surgery with unstable symptomatic coronary disease underwent detailed preoperative cardiac screening ( < 5% of patients). Less than 2% of patients underwent preliminary coronary bypass. Eleven of the 21 MIS including all 5 fatal MIS that occurred after urgent or emergent operations, which precluded screening for coronary disease and preliminary coronary bypass. The suprisingly low overall incidence of MI (2.2% asymptomatic, 0.7% symptomatic, and 1% fatal) and the very low rate of possibly preventable MI (10 of 21: 1.8%) found in this unselected general vascular surgery population is of considerable interest, and indicates to us that little additional reduction in the incidence of perioperative MI can result from any preoperative coronary screening program no matter how vigorously applied. We conclude that modern surgical-anesthetic management has already reduced perioperative MI incidence to a near minimum. Increasingly expensive and invasive preoperative coronary screening programs are unlikely to confer additional significant benefit in the prevention of perioperative MI and must be viewed with caution.
Selective deep hypothermia of the spinal cord prevents paraplegia after aortic cross-clamping in the dog model Ramon Berguer, BS, and Ljubisa Detroit, Mich.
MD, Jose Porto, MD, Brian Fedoranko, Dragovic, MD, Wayne State Universiv,
We tested the hypothesis that sehctive deep hypothermia (19 to 12” C) of the spinal cord protects it from the ischemia that follows double aortic cross-clamping in the dog. The extracorporal perfusion system consisted of a heat exchanger and a pump, infusing saline at 5 to 6” C into the subarachnoid space (L5) and draining it through the cisterna magna. This system, after 30 minutes, cools a normally perfused spinal cord to a stable temperature gradient of 18” C (lumbar) to 13” C (cervicothoracic) Proximal and distal intraspinal, proximal and distal aortic, and central venous pressures were continuously recorded. Rectal temperature was maintained between 36.5” C and 38.5” C. Eight control dogs had crossclamping of the aorta below the left subclavian artery and above the diaphragm without cord hypothermia. Nine experimental dogs had cord hypothermia initiated 50 minutes befor systemic heparinization (100 units/kg) and double cross-clamping of the aorta. Cross-clamping was maintained for 45 minutes. The aorta was then unclamped, heparin reversed, cord cooling discontinued, and the dura
Volume 13 Number 6 June 1991
closed. Hindlimb function of animals was graded by use of Tarlov’s scale at recovery and 24 hours later. The dogs were then killed, and the cords were removed and fixed for microscopy. All control animals were paraplegic and had histologic confirmation of spinal cord infarction. All experimental animals had intact hindlimb function and normal appearing cords on histologic examination. A two-tailed Fisher’s exact test (chi square) shows this difference to be significant to p = 0.00004. In the dog, selective deep hypothermia of the cord avoids the paraplegia induced by aortic cross-clamping. The incidence of venous leg ulceration in relation to ambulatory pressure Andrew N. Nicolaides, FRCS, Maher K. Hussein, MD, Gabriel Szendro, MD, Dimitri Christopoulos, MD, and Spiro Vasdekis, MD, St. Mavys Hospital, London, United fin&ma The purpose of our study was to determine the incidence of ulceration in relation to different grades of venous hypertension resulting from abnormalities of either the superficial or deep venous system. Two hundred fifty-eight unselected consecutive patients (300 limbs) with chronic venous problems referred by their family doctor have been studied with duplex scanning and venous pressure measurements so that they could be classified as having superficial disease only (primary varicose veins) or deep venous disease. The ambulatory venous pressure (AVP) was defined as the lowest pressure recorded during a 10 tiptoe exercise with a canula in a vein on the dorsum of the foot. In 18 limbs with small spider venules, both the superficial or deep veins were competent (group A). In 183 limbs reflux was confined to the superficial veins only (group B), and in the remaining 99 limbs reflux was present in both superficial and deep veins (group C) The incidence of ulceration was 0 in group A, 49 (27%) in group B, and 50 (51%) in group C. The results indicate that for AVP < 30 the incidence of ulceration is nil, that there is a linear relationship between the incidence of ulceration and AVP for pressure between 30-79, and an incidence of 73% for AVP > 79. This finding is irrespective of whether the high AVP is the result of superficial or deep venous disease. Renal revascularkation for recurrent pulmonary edema in patients with poorly controlled hypertension and azotemia: a distinct subgroup of patients with renal artery occlusive disease Louis M. Messina, MD, Gerald B. Zelenock, MD, Katharine A. Yao, BA, and James C. Stanley, MD, University ofMichigan Medical Center, Ann Arbov, Mtib. Recurrent acute pulmonary edema in patients with poorly controlled hypertension and azotemia appears to be a marker of severe bilateral renal artery disease. These
patients have few therapeutic options since drug therapy often results in worsening of renal insufficiency. To determine the effectiveness of renal revascularization in this distinct subgroup of patients with renal artery occlusive disease we report the outcome of 17 consecutive patients who underwent renal revascularization for recurrent pulmonary edema and poorly controlled hypertension. This subgroup represents 8.9% of the 191 patients who underwent renal revascularization at our institution between 1984 and 1990. Mean preoperative blood pressure was 207/111 mm Hg. Serum creatinine was elevated in 16/17 patients (mean, 3.8 m&U). Pulmonary edema occurred despite the finding of normal ventricular function in 65% of the patients. Three patients required dialysis before operation, two of these and one other developed malignant hypertension and encephalopathy; two required intubation, one had a cardiac arrest. Angiography showed severe bilateral renal artery occlusive disease in 94% of patients; 54% had an occluded renal artery. Renal revascularization was accomplished by ileorenal bypass (41%), aortorenal bypass (29%), and endarterectomy (24%). Contralateral nephrectomy (41%), contralateral transluminal angioplasty (6%), concomitant aortic reconstruction (24%) were also required. No patient died after operation. At discharge, hypertension was cured in 2 patients and improved in 15 patients. Renal function was improved ( > 0.5 mg/dl decrease creatinine) in 13 patients, two of the three patients requiring preoperative dialysis had recovery of renal function. Postoperative angiography in 14 patients showed all grafts were patent, one mild residual stenosis, and an occlusion of a small lower pole vessel. Long-term follow-up (mean, 2.4 years) was achieved in all patients, one patient was cured of hypertension (6%), and control of hypertension was improved in 16 patients (94%). One late episode of pulmonary edema occurred in a patient with rheumatic heart disease. Renal function (mean creatinine 1.6 mg/dl) was improved in 71%, stable in 18%, and worsened in two patients, one requiring dialysis. Patients with recurrent pulmonary edema, severe hypertension, and azotemia should undergo angiography. Renal revascularization can result in effective and durable resolution of pulmonary edema, control of hypertension, and improvement in renal function. Extraanatomic carotid reconstruction: the subdavian carotid bypass William R. Fry, MD, John D. Martin, MD, G. Patrick Clagett, MD, and William J. Fry, MD, University of Texas Medical Center, Dallas, Texas Whereas the predominant location of symptomatic carotid occlusive disease is the carotid bifurcation, proximal common carotid lesions cause similar symptomatology. Common carotid lesions occur as isolated disease or in tandem with carotid bulb disease. Restoration of carotid inflow from subclavian based extraanatomic bypasses should provide adequate reconstruction of these lesions. To evaluate subclavian to carotid artery bypass, a
retrospective review of all patients undergoing this procedure from Jan. 1 1977, to Feb. 20, 1989, was performed. Twenty patients (14 men, 6 women) with a mean age of 60 years were treated. Fifteen patients (75%) were admitted with transient ischemic attacks. Five (25%) had nonfocal symptoms (e.g., dizziness, syncope). Arteriographic evaluation demonstrated severe proximal occlusive disease of the common carotid in all cases. Reconstruction bypasses were performed to the carotid bulb (459/o), internal carotid artery (30%), and external carotid artery (25%). Four patients underwent endarterectomy of the internal carotid artery in conjunction with subclavian-carotid bypass. Bypass conduits included saphenous vein (75%) and prosthetic grafts (25%). Asymptomatic phrenic nerve neuropraxia was identified by postoperative chest radiographs in four cases, with no resultant respiratory morbidity. No perioperative strokes occurred. One postoperative death (5%) resulted from a myocardial infarction. Long-term results were available for 18 patients (90%) with a mean follow-up of 50 months (range, 1-122 months). Four patients have died of causes unrelated to carotid vascular pathology. Serial duplex scans have documented graft patency in all 18 patients. A single patient returned with focal neurologic symptoms caused by a posterior circulation infarct. From this experience subclavian to carotid artery bypass appears to be a safe, durable, and well-tolerated procedure for the reconstruction of symptomatic proximal carotid stenosis when the ipsilateral subclavian artery is an appropriate inflow source. For this procedure, saphenous vein appears to be an suitable conduit. Long-term follow-up demonstrates excellent patency and protection against further anterior circulation neurologic events. Amaurosis fugax: is it innocuous? David Rosenthal, MD, John C. Hungerpillar, MD, Mark Crispin, MD, Michael D. Clark, MD, and Pano A. Lamis, MD, Georgia Baptist Medical Center, Atlanta, Ga. There is no concensus on the most appropriate management of the patient with amaurosis fugax (AF). It is simply a minor occular problem to be treated with antiplatelet medications, or should it be regarded as a warning signal of possible monocular blindness or stroke? To evaluate this problem, a lo-year study of 103 patients with AF was carried out. Sixty-two patients with symptoms of AF underwent arteriography that demostrated an ulcerated carotid bifurcation lesion in 37 and hemodynamically sign&ant stenoses ( > 75% diameter reduction) in 25; all of these patients underwent carotid endarterectomy (CE). The other 41 patients with symptoms of AF and arteriographically established carotid artery disease (33 ulcerated plaque, 8 hemodynamic stenoses) were not operated on, but were treated with aspirin/dipyridamole (ASA/DIP) or
Journal of VASCULAR SURGERY
warfarin and served as a control series. There were no strokes or deaths after CE in the immediate 30-day postoperative period. In follow-up of the 62 patients operated on extending to 10 years (mean, 4.2 years), one (1.6%) patient developed recurrent AF symptoms, two (3.2%) TIAs, and one (1.6%) sustained a stroke in the operated hemisphere. These patients underwent repeat arteriography and CT brain scanning, which demonstrated a carotid restenosis in one patient experiencing TIAs and a thrombotic infarct in the patient who suffered a late stroke. In the nonoperated group, despite AU/DIP or warfarin treatment, four (9.7%) patients had ongoing AF symptoms, and two (4.8%) developed TIAs, which led to CE. One (2.4%) other patient developed sudden, permanent monocular blindness, and two (4.8%) suffered hemispheric strokes, one of which was fatal. The cumulative morbidity (ongoing occular or TIA symptoms, perioperative and late stroke) in the operated group was 6.4% (four patients), whereas the cumulative morbidity in the nonoperated group was 21.9% (nine patients) (p = 0.020). When patients are admitted with symptoms of AF and have demonstrable ipsilateral extracranial carotid bifurcation disease, CE in appropriate patients is recommended. Amaurosis fugax is not an innocuous symptom; it should be regarded seriously as a potential harbinger of monocular blindness and/or stroke. Aneurysmal enlargement of the abdominal aorta during regression of experimental atherosclerosis Christopher K. Zarins, MD, Chengpei Xu, MD, and Seymour Glagov, MD, University of Chicago, Chicago, Ill. Recent observations of aortic aneurysms in patients on cholesterol-lowering drugs suggest a relationship between regression of atherosclerosis and aneurysm formation. We explored this relationship in a controlled study of cholesterol lowering in experimental atherosclerosis. Atherosclerosis was induced in 17 cynomolgus monkeys by feeding a 2% cholesterol/25% peanut oil diet for 6 months (group I, n = 6 and group III, n = 6) or 12 months (group II, n = 5). Regression was induced in group III by feeding a no cholesterol “regression” diet for 6 months after the 6-month dietary induction period. No drugs were used. Serum cholesterol after 6 months in group I was 788 * 80 mg/dl (normal 107 + 5 mg/dl) and after 12 months in group II was 508 2 53 mg/dl. In group III, cholesterol was 810 ? 52 mg/dl at 6 months and 198 + 15 mg/dl at 12 months. Plaque deposition was increased in group II compared to group I but group III had the same plaque area as group I. Regression of abdominal aortic plaque was more prominent than thoracic plaque in group III with an abdominal/thoracic lesion area ratio of 0.3 compared to 1.3 in group II @ < 0.05). After 12 months, regression monkeys had a twofold enlargement of the abdominal aorta (both lumen area and artery size [IEL area],p < 0.05) and
Volume 13 Number 6 June 1991
Joint Annual Meeting Abstracts
a reduction in media thickness from 0.18 +- 0.02 to 0.13 ” 0.01 (p < 0.05). No significant enlargement was found of the thoracic aorta. Our finding of early aneurysmal enlargement of the abdominal aorta in monkeys undergoing dietary cholesterol lowering supports the hypothesis that atherosclerotic plaque evolution may play a role in aneurysm formation. Histologic evidence reveals that during progression of atherosclerosis, intimal plaque deposition may be accompanied by atrophy and thinning of the underlying media. Under these circumstances, intimal plaque may contribute to structural support of the aorta. Reduction of plaque volume during regression may leave an aortic wall which is too thin to support mural tension with subsequent aneurysmal dilation. Localization of these changes to the abdominal aorta in this experiment is consistent with the known pattern of distribution of aneurysms and atherosclerosisin man.
vein mapping (n = 21), originally used to evaluate saphenous vein length and diameter, correctly identified TW, CAL, VAR, and PSO veins in 61% of cases.Histology of diseased veins demonstrated recanalized sclerotic vein segments with intraluminal clot, calcifications, and preexisting smooth muscle cell proliferation. Duplex criteria and histologic features will be described. From these data, we conclude that (1) unsuspected preexisting saphenous vein pathology occurs in approximately 12% of cases; (2) detection is possible with duplex scanning and intraoperative methods; and (3) veins that irrigate well and appear normal, but are identified to be PSR, CAL, or TW, should not be used if alternative vein is available. The lesser saphenous vein: an underappreciated source of autogenous vein Benjamin B. Chang, MD, Dhiraj M. Shah, MD, Robert P. Leather, MD, and Jeffrey L. Kaufman, MD, Albany Medical
Unsuspected preexisting saphenous vein pathology: an unrecognized cause of vein bypass failure Thomas F. Panetta, MD, Kurt R. Wengerter, MD, Michael L. Marin, MD, Jamie Goldsmith, RN, Sushi1 K. Gupta, MD, and Frank J. Veith, MD, Montejiure Medical Center, Albert Einstein Collge
New Ymk, NT.
Early and late graft failures occur with all techniques for autogenous vein bypass. Our prior anecdotal experience with unsuspected preexisting saphenous vein pathology prompted us to study its incidence, its relation to graft failure, and to develop techniques for its detection. We reviewed infrainguinal reconstructions in which postphlebitic sclerotic recanalized (PSR), thick walled (TW), varicose (VAR), calcified (CAL), and postphlebitic sclerotic occluded (PSO) vein lesions were detected. During a 7-year experience with 529 cases,preexisting saphenous vein pathology was detected in 63 cases(12%). Fifty of these veins or portions thereof were used for bypasses.Thirty-three veins with lesions (29 TW, 2 VAR, 2 VAR and TW) were used in their entirety. Seventeen postphlebitic veins (13 PSR, 3 PSO, 1 CAL) required either excision, patching, composite, or interposition grafting for diseasedsegments. Thirteen veins with severe, diffise disease (8 PSO, 2 CAL, 2 VAR, 1 TW and VAR) were discarded. PSR and PSO vein segments were identified by palpable abnormalities; obstruction to irrigation, catheter, or valvulotome passage;or by abnormal intraoperative arteriography. Early graft failures occurred in 7 out of 50 cases( 14%). Three of these were PSR veins, which irrigated well but occluded intraoperatively and were replaced with synthetic grafts. Cumulative primary patency for pathologic veins at 30 months was 32%. This was significantly less than the 68% cumulative primary patency for veins without detectable pathology (p < 0.05). Retrospective evaluation of preoperative duplex scan
College, Albany, N.Y.
Use of the ipsilateral greater saphenouse vein (GSV) for arterial bypass procedures is frequently limited by previous stripping, bypassoperations, or anatomic unsuitability. In such cases,the contralateral GSV or arm veins are often resorted to. However, over the past 5 yearswe have used the lesser saphenous vein (LSV) as a preferred alternative for autogenous vein, as this usually requires no extra preparation, and the LSV has characteristicsidentical to the GSV. Duplex scanning has been used in 3 11 casesfor preoperative mapping and assessmentwith excellent correlation with actual anatomy found at operation. Harvest of the LSV has been facilitated by the use of a medial subfascialapproach not requiring specialpositioning of the leg. A total of 91 LSVs have been used for peripheral arterial bypass procedures; 66 of these were re-do cases. Vein utilization was 88%. In 58 of these cases,the LSV was used asthe entire conduit, including 10 in situ, 38 reversed vein (including 18 for coronary artery bypass), and 10 orthograde vein bypasses. Thirty-day patency was 92% (45/49). In the remaining 33 cases,the LSV was spliced together with another vein to complete a bypassprocedure. These data suggest that the LSV should be the primary alternative to ipsilateral GSV for arterial bypass becauseof its ready availability, high utilization rate, easeof harvesting and preparation, and ideal handling characteristics. A blinded comparison of arteriography, angioscopy, and duplex scanning for the intraoperative evaluation of in situ saphenous vein bypass g&Is Jeffrey J. Gilbertson, MD, Daniel B. Walsh, MD, Robert M. Zwolak, MD, et al. Dartmmth-Hitchcock Medical Center, Havwveq N.H. Intraoperative confirmation of technical adequacy of in situ saphenous vein bypassgrafts is mandatory to avoid early graft failure. Arteriography, angioscopy, and duplex
scanning have each been advocated as the preferred technique, but no study comparing the three modalities in the same gr% had been reported. Nineteen femoralinfragenicular m situ saphenous vein bypass grafts were each studied during operation in sequence with arteriography, duplex scanning, and angioscopy. Each study was performed and interpreted by a different surgeon blinded to the results of the other studies. Abnormalities requiring intervention were categorized as (1) patent vein branches, (2) residual valve cusps, or (3) anastomotic defects. Modality-specific criteria corresponding to each category were defined at study outset. Seventeen residual valve cusps and 53 patent vein branches were identified by at least one modality. Only 9 residual valve cusps and 30 patent vein branches were actually found (and corrected) by direct inspection. No anastomotic abnormalities requiring intraoperative revision were encountered. Study duration did not differ among the three modalities. To date (mean, 6-month follow-up) no stenosis or arteriovenous fistula has been detected in any graft by postoperative duplex surveillance. Arteriography and angioscopy were superior to duplex scanning for detecting patent vein branches. Arteriography or duplex scanning, used alone, would have missed > 75% of residual valve cusps. Although the impact of residual valve cusps on long-term patency of the graft is unknown, our results suggest that angioscopy is the preferred method of intraoperative graft evaluation. lo-year experience with popliteal-to-distal artery bypasses: the significance and management of proximal disease Kurt R. Wengerter, MD, Paul Yang, MD, Sushi1 K. Gupta, MD, Thomas F. Panetta, MD, and Frank J. V&h, MD, Montefime Medical Center, Albert Einsteha College of Medicine, New York, N.Y. The value of popliteal-to-distal artery bypasses (PDBs) in limb salvage has been documented. However, the influence of progression of proximal disease and the role of superficial femoral artery (SFA) or popliteal artery (PA) percutaneous transluminal angioplasty (PTA) befor bypass have not been adequately assessed. To evaluate these and other factors we reviewed 143 nonsequential PDBs performed since 1979. Limb salvage was’the indication for all procedures, and 87% of the patients had diabetes. Nineteen patients underwent prebypass PTA of the SFA or PA proximal to the graft for stenotic lesions ranging from 24% to 85% diameter reduction. Of those cases in which preoperative PTA was not performed, 22 limbs were found on preoperative arteriograuhv to have one or more stenoses in the SFA or PA Y I s with a diameter reduction of 20% to 60%. Postbypass intraoperative pressure measurement at the proximal anastomosis was used to assess the adequacy of inflow when an inflow stenosis existed. The postbypass pressure gradient was I 1.5 mm Hg in all cases in which an inflow stenosis
Journal of VASCULAR SURGERY
was accepted; when a larger gradient was present an inflow (femoropopliteal extension) procedure was performed. The data show that PDBs are durable procedures. Moderate disease of the inflow arteries does not appear to affect graft patency or limb salvage when the postbypass inflow gradient is I 15 mm Hg. Preoperative PTA of stenotic inflow vessels does not adversely affect the late (4-year) patency or limb salvage results of PDBs performed distal to the PTA. Causes of primary graft failure after in situ saphenous vein bypass Magruder C. Donaldson, MD, John A. Mannick, MD, and Anthony D. Whittemore, MD, [email protected]
6 Wopnenr Hospital, Boston, Mass. Further improvement of methods for infrainguinal revascularization depends in part on analysis of causes of bypass graft failure. Four hundren fifty-five in situ saphenous vein grafts originating in the groin were performed by use of the Mills valvulotome and routine completion angiography. Cumulative primary graft patency was 70% and secondary patency 79% at 5 years. Primary graft failure occurred in 90 grafts, including 21 gratis (4.6%) with nonocclusive stenosis and 69 grafts (15.2%) with occlusion. The cause for failure could not be determined in 7 grafts; 103 contributory causes were identified in the remaining 83 grafts. Among the 103 causes identified, 62 (60%) were intrinsic to the graft itself and contributed to failure of 51 of 455 grafts (11.2%). These causes included vein stricture (20), anastomotic stenosis (14), valve site scar (7), valvulotome injury (9), kink (5), retained valve leaflet (3), residual arteriovenous fistula (3), and intimal flap ( 1). Among these intrinsic causes, 20 (19.4%) were directly related to the in situ technique, contributing to failure of 18 of 455 grafts (4.0%). Eleven of 29 graft failures within 30 days of surgery were related to causes specific to the in situ method. Forty-one (40%) of the 103 causes identified were extrinsic to the graft, including compromised inflow (1) or outflow (23), hypercoagulability (9), systemic hypotension (6), and graft sepsis (2). Seventy-three percent of known causes arose within 6 months of surgery. Hypothetically, improved technique and perioperative management could have eliminated 48 of 103 causes of primary graft failure. Myointimal hyperplasia and late progression of outflow disease remain resistant to modern therapy. Treatment of reversed vein graft stenosis by percutaneous transhminal angioplasty Henry D. Berkowitz, MD, and Andrew D. Fox, MD, University of Pennsylvania Hospital, Philadelphia Pa. Conscientious surveillance of intrainguinal bypass grafts is mandatory to detect vein graft stenoses which, if uncorrected, can lead to graft occlusion., It is now widely accepted that noninvasive vascular lab studies are the best way to detect these lesions. However, there is still
VolLlme 13 Number 6 June 1991
Joint Annual Meeting Abstracts 909
controversy over treatment, specifically whether balloon angioplasty is an acceptable substitute for surgery (patch angioplasty or short jump grafts) in the treatment of these lesions. We have always favored balloon angioplasty as primary treaUnent and have summarized our experience with treating 62 stenotic reversed femoral-popliteal (FP) and femoral-tibial (FT) veingrafts, which represent 12% of 521 femoral-popliteal and femoral-tibia1 bypass grafts performed at our institution. Prosthetic and in situ grafts are specificallyexcluded from this report, as well as occluded grafts, which are found to have stenotic lesions after lytic therapy. The most common stenotic lesion occurred within 4 cm of the proximal anastomosis (33/62,53%). The other siteswere near the distal anastomosis (18/62,29%), and in the middle of the graft (1 l/62,18%). Eighty-four percent (52/62) of the lesions were treated initially by PTA with a 34.6% recurrence. None of the grafts treated surgically by vein patch angioplasty or short jump grafts experienced recurrence. Sixty-one percent of the recurrent stenoseswere again treated by PTA with a 38.9% recurrence.Most of the tertiary recurrences were then treated surgically. Overall 63% (39/62) of the stenotic grafts were treated by PTA alone. The 5-year life-table graft patency after correction of stenoses was 60%. No statistical difference was found between the 54% FP and the 67% FT patency, nor was there any statistically significant difference between the 5-year patency of the three types of lesions: proximal (66%), midgraft (54%), and distal (50%). Since 63% of grafts can be treated by PTA alone, we recommend this modality for the primary treatment of graft lesions. However, once a recurrent lesion develops, surgical repair is warranted. Vascular disease in the antiphospholipid syndrome: a comparison to the atherosclerotic patient population Cynthia K. Shortell, Kenneth Ouriel, MD, John J. Condemi, MD, Richard M. Green, MD, and James A. DeWeese, MD, University of Rochester Medical Center, Rochestep;
The antiphospholipid syndrome (APS) is a condition characterized by arterial and venous occlusions, thrombocytopenia, and recurrent fetal loss. Antiphospholipid syndrome was diagnosed in 16 of 1075 patients admitted to our vascular service between 1987 and 1990. All patients with APS had either anticardiolipin antibody (14/16) or the lupus anticoagulant (9/16); in three patients the platelet count was < 150 K, and in seven patients the partial thromboplastin time was prolonged (20.8% above control). The erythrocyte sedimentation rate was elevated in 9 of the 12 patients in whom the test was performed, and degree of elevation did not correlate with presentation. The most common site of involvement was the cerebral
circulation (nine patients) manifested by transient ischemic attacks or stroke, each in the presence of a normal carotid angiogram. Upper extremity disease was seen in eight patients, and was characterized by symptoms of Raynaud’s disease but with &red angiographic lesions involving the brachial, radial, ulnar and/or digital arteries. Lower extremity disease occurred in six patients with a clinical and angiographic presentation similar to that of atherosclerosis. There were two patients with coronary artery diseaseand one patient with renovascular hypertension. Antiphospholipid syndrome should be suspected in young patients with vascular diseasewho do not smoke, have upper extremity involvement with fixed angiographic findings, have cerebrovasculardiseasewith a normal carotid angiogram, or are admitted with premature vascular graft failure. The economic cost of a femorwrural reconstrwtion policy with and without autologous vein Nicholas J. W. Cheshire, FRCS, Marie Noone, RGN, Linda Davios, BA, Felix Eastcott, FRCS, and John H. N. Wolfe, MD, St. [email protected]
Hospital, Lmmim, United Kingdom It is well established that successful primary arterial reconstruction, even to crural vessels is cheaper than amputation. Reintervention increasesexpenditure and may produce costs exceeding primary amputation. Furthermore, secondary amputation may eventually become necessary. Femorocrural grafts have the highest average ‘creconstruction policy” cost (i.e., primary procedure and all further operations necessaryduring follow-up). We must therefore seek evidence to support this potentially expensive policy. In conjunction with health economists we have compared average policy costs of 125 reconstructions using grafts exceeding 70 cm in length (89 vein grafts, 36 PTFE grafts with collar) and 67 amputations at mean follow-up of 3 years. One-month mortality rate for reconstruction was < 1% and was 10% for amputation. At 3 years, however, 20% of both groups were dead. Overall 3-year patency is 65% (72% for vein grafts, 48% for PTFE). Ninety-eight percent of irreversible graft occlusions resulted in amputation in these patients. It is important to note that community costs incurred by immobile amputees are not included in the following costs. Seventy percent of amputees were confined to the home compared to 9% of reconstructions, incurring costs estimated at between L6000 and L21000 ($12 to 42000) per year. These figures have been used to formulate a decision tree showing cost increase and likelihood of successat each decision mode. We have also calculated primary and secondarygraft patency combinations that will produce an average reconstruction policy cost cheaper than the cost of primary amputation (S-year primary patency of 46% is cost equivalent, primary patency 40% requires a 20% improvement from secondary intervention to achieve equivalence).
Journal of VASCULAR SURGERY
From the Scientific Program of the Forty-fi&h Annual Meeting of the Society for Vascular Surgery Accuracy (LEADM)
Gregory L. Moneta, MD, Richard A. Yeager, MD, Ruza Antonovich, MD, Lee D. Hall, MD, and John M. Porter, MD, Veterans Administration Medical Center, Portland,
Lower extremity arterial duplex mapping (LEADM) has been suggested as an alternative to angiography before arterial reconstruction. To date, however, no study has been performed comparing the accuracyof LEADM to that of angiography in a large patient group. The present prospective blinded study was performed for this purpose. One hundred fifty consecutive patients (286 limbs) underwent LEADM, segmental Doppler pressures,and arteriography before contemplated arterial reconstruction. Based on clinical data and Doppler pressures,limbs were classified as normal, aortoiliac disease (A), femoral-popliteal-tibial disease (B), or multilevel inflow-outflow disease (C) to determine if the clinical diseasecategory affected LEADM accuracy. Lower extremity arterial duplex mapping (LEADM) was performed with use of velocity criteria for stenosis previously published by us and others. Lower extremity arterial duplex mapping (LEADM) was performed on each lib from the aorta through the tibials. Ninety-nine percent of 2036 arterial segments from the popliteal artery cephalad were visualized, as were 95% of 540 anterior (AT) and posterior (PT) tibial artery segments. In vesselsproximal to the tibial arteries, LEADM was evaluated for ability to detect 2 50% angiographic stenosis and distinguish stenosis from occlusion. In the tibial arteries, LEADM was assessedfor ability to predict continuous tibial artery patency from origin to ankle. Lower extremity arterial duplex mapping accurately distinguished >50% stenosis from occlusion in 97% of cases. We conclude LEADM is highly accuratein comparison to angiography in patients before operation with lower extremity ischemia and in all commonly encountered clinically determined patterns of lower extremity atherosclerosis.Potential substitution of lower extremity arterial duplex mapping for angiography before arterial reconstruction deservescontinued evaluation. Vascular laboratory vein grafts
Robert M. Zwolak, MD, Jack L. Cronenwett, MD, Martha D. McDaniel, MD, Dartmwth-Hitchock Medical Center, Hanover, N.H.
The value of vascular laboratoy studies in the postoperative surveillance of lower extremity vein grafts is controversial. We analyzed 985 laboratory evaluations of 310 vein grafts (250 primary, 60 secondary) in 239 patients for which at least 6-month follow-up was available after the
last evaluation. Studies were performed at 3 to 6-month intervals (mean 107 f 26 days), increasing with longer follow-up, for an average of 3.2 sequential studies/graft. Each study included Doppler measured ankle-brachial index (ABI) and duplex ultrasound measurement of peak systolic velocity (PSV) and volumetric resting blood flow. Criteria for ( + ) studies were: ABI decreaseby 2 0.15 from previous study, PSV < 45 cm/set, and flow < 60 ml/min. These were tested for their ability to predict graft failure, as defined as thrombosis (n = 38) or stenosis requiring surgical revision (n = 53). Sensitivity, specificity, and accuracywere calculated. Positive predictive value (PPV) is the likelihood of graft failure within 6 months of a ( + ) study. Negative predictive value (NPV) after a ( - ) study is the likelihood of graft patency until the next study or for 6 months. Accuracy, PPV, and NPV are influenced by the prevalence of graft failure in our patient population. Ankle-brachial index was most specificof the individual tests, whereas PSV and flow were more sensitive for detecting impending graft failure. Overall accuracy was improved by combining the three tests. If all tests were ( - ), only 3% of grafts failed before the next evaluation. Alternatively, if all tests were ( + ), 79% of grafts failed within 6 months. Seventy-six percent of the 38 graft thromboses in this experience were actually predicted by a ( + ) study, which was clinically ignored. We conclude that vascularlaboratory surveillanceprovides an accurate way to detect impending vein graft failure. Popliteal entrapment - an uncommon presentation the chronic compartment syndrome (CCS)
William D. Turnipseed, MD, and Myron A. Pozniak, MD, Unive&g Hospittal and Chits, Madison, Wis. Claudication in young adults may occur becauseof an unusual form of popliteal entrapment associated with chronic compartment syndrome (CCS) . Patients complain of calf cramping, and quick limb fatigue. Eleven of 120 patients (9%) (mean age, 27 years) evaluated for possible CCS had these complaints. Although results of resting and postexercise pulse volume recording examinations were normal, these patients demonstrated complete obliteration of the popliteal pulse with forced plantar flexion of the foot. (Screening of 36 asymptomatic age group controls demonstrated popliteal compression in 29 of 72 limbs [40%], and no compression in 43 of 72 limbs [60%]). Duplex examination of the popliteal artery in the 11 symptomatic patients demonstrated complete arterial occlusion with plantar flexion (7 bilateral [64%], 4 unilateral [36%]). Intravenous digital angiograms demonstrated no intraluminal defects or anatomic deviation of the popliteal artery at rest, but with plantar flexion, occlusion occurred. Magnetic resonance imaging studies demonstrated lateral compression of the neurovascular bundle by the medial head of the gastroc and plantaris muscle, forcing it against the lateral condyle of the femur and the solealsling. Twelve symptomatic limbs were treated by surgical excision of the plan&s muscle, take down of the soleus muscle, and resection of its fascialband. Eight of 11 (72Oh)patients had
Volume 13 Number 6 June 1991
complete relief of symptoms, and 3 (28%) were significantly improved. Entrapment of the popliteal vesselsmay result from compression at the entrance of the deep posterior (soleal) compartment. Symptomatic entrapment is uncommon. Surgical releaseof the soleus muscle and its sling is effective treatment. Color flow duplex scanning for the surveillance diagnosis of acute deep venous thrombosis
Mark A. Mattes, MD, Gregg L. Londrey, PhD, Kim J. Hodgson, MD, Darr W. Leutz, MD, E. Shannon Stauffer, and David S. Sumner, MD, So&wn Illinois University Schoolof Medicine, Spin&ld, Ill. Compared to conventional duplex imaging, color flow scanning (CFS) facilitates the identification of veins (especially below the knee), decreasesthe need to assessDoppler flow patterns and venous compressibility, and allows veins to be surveyed longitudinally. These advantages translate into a lessdemanding and time-consuming examination. This study was designed to determine the accuracyof CFS for detecting acute deep venous thrombosis (DVT) in patients in whom the diagnosis is clinically suspected and in asymptomatic patients after operation at high risk for developing DVT. The diagnostic group (D) included 65 limbs of 60 patients, and the surveillance group (S) included 152 limbs of 79 patients undergoing total hip or knee replacement. All patients were prospectively examined with CFS and phlebography. Patients in the diagnostic group had significantly more above-knee (AK) and popliteal (POP) thrombi than those in the surveillance group (p < 0.002). In symptomatic patients, CFS was more accurate above the knee and most accurate in the superficial femoral segment. All AK thrombi were identified, and only two below-knee (BK) thrombi were missed. In contrast, accuracywas relatively poor in the surveillance group. Color flow scanning was significantly lesssensitive (p < 0.01) for detecting clots in the AK, BK, and tibioperoneal (TP) segments of the surveillancegroup. Negative predictive values were good in both groups, exceeding 90% in most segments. Color flow scanning effectively excludes DVT in symptomatic and asymptomatic high-risk patients and predicts the presence of AK thrombi in diagnostic patients with reasonable accuracy. We conclude that CFS is as accurateas conventional duplex imaging and, becauseof its advantages, is the noninvasive method of choice for evaluating patients with suspected DVT. Its role in the surveillance of high-risk patients remains to be determined and awaits further clinical evaluation. Valvular reflux after deep vein thrombosis and time of occurrence
Arie L. Markel, MD, Richard A. Manzo, BS, CCVT, Robert 0. Bergelin, MS, and D. Eugene Strandness, Jr., MD, Univern’ty of Washington, Seattle, Wash. From December 1986 to November 1990, 349 patients with acute deep vein thrombosis (DVT) were
studied in our laboratory. From this group 107 patients (123 legs with DVT) were placed in our long-term follow-up program. The documentation of valvular reflux and its site was documented by duplex scanning. The duplex studies were done on days 1, 7, and 30. Repeat studies were done every 3 months during the first year and yearly thereafter. The mean follow-up time for these patients was 341 days (11.4 months). At the time of the initial study valvular incompetence was noted in 19 limbs (15%). Reflux was absent in 104 limbs (85%). In this last group reflux was documented in 17% of the limbs by day 7. By the end of the first month 40% demonstrated reflux. By the end of year 2, two thirds of the involved limbs had developed valvnlar incompetence. By sites, the following was noted at the end of the first year of follow-up: (1) popliteal vein-58%; (2) superficial femoral vein-40%; (3) greater saphenous vein-25%; and (4) posterior tibial vein-18.2%. Of great interest was the observation that reflux in the deep venous system tends to develop in those venous segments that had been previously occluded and then underwent lysis.After 1 year follow-up reflux developed in the superficial femoral vein in 46% of the limbs where this segment was thrombosed initially. The posterior tibial vein showed reflux in 33% of the caseswith initial thrombus at this segment. In contrast only 17% of the superficial femoral vein and 8% of posterior tibial veins that were not occluded developed subsequent reflux. The implication of these findings will be discussed. Infected femorodistal moval mandatory?
Kenneth J. Cherry, Jr., MD, Christopher F. Roland, MD, Peter C. Pairolero, MD, et al. tiyo C&c, Rocbestm,M&an. Infected lower extremity bypass grafts have been associated with high rates of limb loss. Treatment traditionally had included graft excision.To compare aggressive local treatment, without graft removal, with more conventional graft excision, we reviewed 38 consecutive patients with 39 infected lower extremity bypasses (33 femoropopliteal, 6 femorotibial) treated during the last 10 years. The grafts were expanded polytetrafluoroethylene (ePTFE) in 27 cases,Dacron in 6, vein in 4, and composite (vein and ePTFE) in 2. Median follow-up was 2.7 years. Twenty-eight infected grafts were treated with either complete (14) or partial (14) graft removal. Nine of these patients had new grafts placed. Recurrent infection developed in five cases(17.9%), and two patients (7.4%) died of complications of graft infection. Ten of 20 limbs at risk (50%) were lost. Eleven patients with patent bypasses (4 vein, 2 composite, 5 prosthetic) were treated without graft excision. Six patients were admitted with abscessesadjacent to the graft, two with systemic signs and symptoms and local cellulitis, two with infected sinus tracts, and one with pseudoaneurysm. Five patients were treated with initial graft and wound debridement followed by wound closure
Journal of VASCULAR SURGERY
Joint Annual Meeting Abstracts
with muscle transposition. Six patients with infection at the popliteal level were treated with incision and drainage of abscesses (5), or excision of a persistent sinus tract involving the graft (1). Five-year limb salvage rate was 100% in patients treated without graft removal. This was significantly higher (p = 0.005) than in patients treated with graft excision. One patient (9.1%) died of complications of graft sepsis, and no recurrent infections developed. These were not significant differences compared to those having graft excision. We conclude that infected lower extremity bypass grafts pose a major threat of limb loss. Aggressive local treatment, including drainage of localized infection, debridement of devitalized tissue, and muscle transposition may definitively treat the infection in selected patients without the need for graft removal. Heparin-induced thrombocytopenia in the newborn Donald Spadone, MD, Frank Clark, MD, Elizabeth James, MD, Jerry Laster, MD, and Donald Silver, MD, University of Missouri-Columbia, Columbia, MO. Heparin-induced thrombocytopenia (HIT), which occurs in about 5% (0.9% to 30%) of adult patients receiving heparin, has not been reported in the newborn. Seriously ill neonates have been noted to develop thrombocytopenia associated with thrombotic or embolic complications while receiving dilute solutions of heparin to maintain catheter patencies. This study was initiated to determine&e relationship of HIT to these thromboembolic complications. Infants in the newborn intensive care unit receiving heparin in whom who thrombocytopenia developed (< 70,000/mm3) (n = 23), thromboses (n = 6), or precipitous falls in platelet count (n = 5) were studied. Thirty-four patients meet these criteria. By use of platelet aggregation test the presence of heparin-dependent antibodies, was demonstrated in 14 infants. The infants’ average gestational age (29 * 6 weeks); birth weight (1300 +- 945 gm); and platelet count at birth (234,000 ? 11 l,OOO/mn?) did not differ statistically from infants without heparin-dependent platelet aggregation (HDPA). Bleeding was not associated with the thrombocytopenia. Aortic thrombosis was documented by abdominal ultrasonography in 11 of 13 (85%) of infants with HDPA. (One patient died with a midgut volvulus before the aorta could be examined.) Infants with HDPA had a lower average platelet count (54,000 versus 89,000/mm3, p < 0.05) at time of diagnosis than the 20 without HDPA. Only 5 aortic thromboses were detected in the 20 remaining infants with negative aggregation studies; all of these patients had umbilical artery catheters and thrombocytopenia developed in only one. One patient with previously demonstrated thrombocytopenia and HDPA had recurrent thrombocytopenia when reexposed to heparin; her platelet count subsequently recovered after heparin withdrawal. Heparin-induced thrombocytopenia does occur in preterm and term infants receiving heparin and can be associated with arterial thromboses. All patients receiving heparin must be carefully monitored for HIT.
Preoperative carotid artery screening in elderly patients undergoing cardiac surgery Eric S. Berens, MD, Nicholas T. Kouchoukos, Suzan F. Murphy, RN, and Thomas H. Wareing, Jewish
Hospital, St. Louis, MO. The role of preoperative screening for carotid artery disease in elderly patients undergoing cardiac surgical procedures (CSP) is not clearly established. A prospective study was designed to identify preoperative variables that would define populations of patients at higher risk for carotid artery disease. During a 54-month interval, 1094 patients of a consecutive series of 1181 patients over 65 years of age who underwent CSP (89.3% had coronary artery disease) were preoperatively evaluated with carotid duplex ultrasonography. The prevalence of disease was 16.7% for > 50% stenosis and 5.5% for > 80% stenosis. By use of a stepwise, logistic regression model of 10 variables, six were found to be independent predictors of carotid disease: female gender (p = O.OOOl), peripheral vascular disease (p = 0.0006), cigarette smoking (p = 0.0006), older age (p = O.OOl), previous transient ischemit attack and/or stroke (p = 0.002), and left main coronary artery disease (p = 0.01). Whereas 67% of patients with all risk factors had > 50% stenosis, only 5% of patients with no risk factors were found to have similar carotid disease. By use of combinations of these risk factors, the probability of carotid stenosis can thus be predicted. Carotid endarterectomy (CEA) was performed on 45 of the patients who were either neurologically symptomatic ( 17), had > 80% stenosis on screening (26), or had carotid artery ulcerations by angiography (2). Carotid endarcterectomy was combined with CSP in 43 of 45 cases. The overall stroke rate for the 1094 patients was 1.8% (20 patients), and the 30-day postoperative mortality rate was 5.4% (59 patients). The post-CEA stroke rate was 6.6% (3 of 45 patients; all with severe bilateral disease). If only hemispheric strokes associated with moderate or severe carotid stenosis are considered, the stroke rate was 0.4% (4 of 1094 patients). The remaining 16 strokes were related to aortic and cardiac emboli or nonembolic events. In summary, by use of the above risk factors, subgroups of elderly patients undergoing CSP have been identified who should be preoperatively screened for carotid artery disease. Furthermore, by performing combined CEA-CSP in all symptomatic and in nonsymptomatic patients with high-grade carotid obstruction, an acceptable postoperative stroke rate has been achieved. Carotid replacement in conjunction with resection of squamous cell carcinoma of the neck M. Kathleen Reilly, MD, Malcolm 0. Perry, MD, James L. Netterville, MD, and Patrick W. Meacham, MD,
Vanderbilt University Schoolof Medicine, Nash&,
Squamous cell carcinoma can invade the carotid artery. The treatment options then include irradiation, “palliative peeling” of tumor from the artery, and carotid resection with ligation or in-line grafting. From 1987 to 1990, 12 patients underwent carotid
Volume 13 Number 6 June 1991
artery resection and replacement bypass grafting in conjunction with tumor resection. Before operation all patients had CT scans and/or magnetic resonance imaging of the head and neck. Carotid arteriography was performed in eight patients. All patients underwent en bloc resection of the carotid after frozen sections showed cancer cells in the adventitia. Eleven patients underwent reversed saphenous vein bypass grafting and one patient had in-line grafting performed by use of the suprascapular artery. There were no neurologic events perioperatively nor during follow-up of 6 months to 3 years. Two patients subsequently died of distant metastases, and one was lost to follow-up. Of the nine patients available for follow-up none have developed local recurrence involving the bypass. No combination of tests accurately predicted tumor invasion of the carotid artery. Carotid replacement can be safely performed in conjunction with resection of the squamous cell carcinoma. Whether the en bloc resection made possible by carotid grafting will improve long-term survival or not will depend on the results of follow-up. Acute traumatic aortic aneurysm: The Duke experience from 1970 to 1990 Francis G. Duhaylongsod, MD, Donald D. Glower, MD, and Walter G. Wolfe, MD, Duke Univers$y Medical Center, Durham, N. C. Improvements in the operative management of acute traumatic thoracic aortic aneurysm (TAA) have resulted in safe and expeditious repair, nonetheless, multisystem injuries continue to inflict significant mortality. From 1970 to 1990, 108 patients with TAA, representing the second largest reported series, were evaluated at our institution. Mean injury severity score (ISS), excluding aortic injury, was 17.5. Ninety-three patients (86%) survived beyond initial resuscitation and came to operation. Median interval from injury to aortic repair was 8 hours (range, 2 hours to 19 days); there were five operative deaths. Lethal nonaortic injuries included 18 closed head injuries, four myocardial contusions, two intraabdominal vascular injuries, and one pulmonary contusion. The overall mortality rate was 39% of total admissions (42 of 108), and 29% of survivors of resuscitation (27 of 93). Significantly, only 11 of the 42 deaths (26%) were directly attributable to TAA. Adjuncts to prevent spinal cord ischemia (shunt/bypass) were used in 76 patients, whereas 12 underwent clamp/repair. Postoperative paraplegia developed in 5 of 73 patients (6.8%), including 4 of 62 (6.5%) repaired with shunt/bypass and 1 of 11 (9.1%) repaired with clamp/repair (p = NS). Among those who developed paraplegia, the mean ISS was 27.0, and the median interval from injury to repair was 4.9 hours (range, 2 to 6.5). Death in patients with TAA is due primarily to associated injuries. Overall injury severity, intraoperative hypotension, and extensive aortic tissue destruction appear to correlate with the development of postoperative paraplegia.
Proximal aortic pseudoaneurysms after aortic bypass grafting James M. Edwards, MD, Sharlene A. Teefey, MD, R. Eugene Zierler, MD, and Ted R. Kohler, MD, Seattle Veterans Administration Medical Centa, SeattLe, Wash. Although the reported incidence of proximal aortic pseudoaneurysm (PAP) after abdominal aortic bypass grafting ranges from 1% to 5%, the true incidence is unknown because few studies have routinely used a test capable of reliably detecting PAP. Over the last 3 years it had been our policy to perform yearly abdominal ultrasound examinations on all clinic patients who have an aortic graft. Our experience with these patients forms the basis of this report. From July 1,1988, to Dec. 1,1990,131 patients with aortic grafts underwent ultrasound examination. Complete data was available on 100 patients. Eight patients had PAPS ranging in size from 4.0 to 6.2 cm (mean 4.76 2 0.9 cm). The average time between operation and detection of the PAP was 128 & 76 months (range, 35 to 265). There was a bimodal distribution, with three PAPS detected at 36 to 38 months, and the remaining five detected at 120 to 265 months. All three patients whose PAP was detected early had one or more risk factors associated with pseudoaneurysm formation (urgent operation, steroid use, postoperative sepsis, renal failure, and reeoperation) . There were no differences in the type of graft or suture material, type of anastomosis, or indication for surgery between those who developed early or late PAPS or between patients who did or did not develop PAPS. Two large PAPS (5.1 and 6.0 cm) have been repaired by graft replacement. The remaining six PAPS have remained asymptomatic after 13 2 9 months of follow-up. These data support the use of ultrasonography for routine follow-up of aortic grafts. The frequency at which this follow-up should occur remains to be determined; however, it seems reasonable to obtain yearly abdominal ultrasound examinations, particularly in patients with known risk factors for pseudoaneurysm development. Comparison of cardiac morbidity in aortic and infrain&al vascular operations William C. Krupski, MD, Elizabeth Layug, MD, Linda M. Reilly, MD, Joseph H. Rapp, MD, and Dennis T. Mangano, PhD, MD, San Francisco Veterans Administration Medical Center, San Francisco, Calif: We prospectively compared the differences in perioperative cardiac ischemic events between major abdominal versus infiainguinal vascular operations in 140 patients requiring revascularization. Fifty-three patients underwent aortic operations, and 87 had lower extremity vascular reconstructions. We also assessed the predictive value of preoperative dipyridamole thallium cardiac scintigraphy in 38 patients with treating physicians blinded to the test results. Myocardial ischemia was determined during operation by use of 12-lead electrocardiography (EGG) and transesophageal echocardiography. Continuous 2-lead ambulatory ECG (Holter monitoring) was performed before,
Journal of VASCULAR SURGERY
during, and after operation for 3 days. Outcome events were unstable angina, ventricular tachycardia, congestive heart failure, myocardial infarction, and cardiac death. Most demographic variables (e.g., age, hypertension, cigarette smoking, serum cholesterol, surgical indication) were comparable between patients having aortic or infrainguinal arterial operations. However, significant differences were found in baseline clinical features between groups. Abnormalities in preoperative Holter monitoring, ECGs, and thallium scanswere equivalent between groups. During operation, whereas Holter and ECG abnormalities were comparable, more patients undergoing aortic procedures suffered ischemia as determined by transesophageal echocardiography (26% versus lO%,p = 0.019). Although one third of patients in each group who had preoperative thallium scintigraphy had reversible defects, such defects did not reliably predict which patients were at greatest risk for postoperative cardiac morbidity. Because preoperative cardiac diseaseand adverse cardiac outcomes occurred with similar or even greater frequency in both groups of patients, we conclude that the risk for postoperative cardiac ischemic events in lower extremity vascular operations is at Least as great as for aortic operations. Therefore, cardiac complications should be anticipated in patients having infrainguinal revascularization, and perioperative monitoring and care should be as intensive asfor patients having aortic procedures. Experience with more than 300 operations for nonspecific aortoarteritis (Takayasu’s disease) A. V. Pokrovsky, MD, A. K Vislmevsky Institutefbr Swvery, Moscow, USSR Sponswed by: Lany H. Ho&r, MD, New Orleans, La.
We have reviewed the Soviet Union experience with 300 cases of operative intervention in patients with Takayasu’sarteritis. Presentation occurred within the first two decades of life, and angiographic changes were localized to the aortic arch branches, the thoracoabdominal aorta, and the renal arteries. Primary indications for surgery were renovascular hypertension and/or cerebrovascular insufficiency. Acute inflammation is believed to be a contraindication to surgical intervention. Operative repair was undertaken for stenosis of the aortic arch branches in 100 patients. Sixty-six percent had bifurcation grafts placed to revascularize carotid and subclavian arteries; 33% had extra thoracic grafts for cerebral revascularization. Revascularization of the descending thoracic and thoracoabdominal aorta was performed in 50 patients; the usual procedure was graft replacement, with visceralrevascularization being included in 50%. One hundred patients underwent renal artery reconstruction, with 80% having concomitant aortic graft replacement. Transaortic endarterectomy was usually possible and was the preferred method of revascularization for visceral and renal arteries. The incidence of complications and the long-term results of treatment are discussed. Comparative studies
document that surgical treatment is better than nonoperative medical treatment. Surgical management tremity ischemia
Richard A. Yeager, MD, Gregory L. Moneta, MD, Lloyd M. Taylor, Jr., MD, Daniel W. Hamre, MD, Donald B. McConnell, MD, and John M. Porter, MD, Pwthnd Veterans Administrattim
Medkcal Center, Porthd,
Optimal management of acute lower extremity ischemia (ALEI) remains controversial. Thrombolytic and/or interventional radiologic therapies are of unproven efficacy, may be associatedwith serious complications, and may lead to intolerable delays in revascularization. Previous reports of early surgical management of ALE1 are flawed by a consistent lack of objective documentation of ischemic severity. In our institution we use initial anticoagulation and revascularization in ALEI. Thrombolytic and therapeutic endovascular interventions are avoided. This review was conducted to determine the efficacy of this approach. During 1986 to 1990 we treated 70 patients (66 men, 4 women; mean age, 64 years)who were admitted with < 2 week history of ALE1 and absent pedal Doppler signals in 79 limbs. The causesof ALE1 were in situ thrombosis in 64 patients (91%) and embolism is six (9%). Patient management included immediate anticoagulation (87%), early angiography (679/o), and either emergent (33%) or urgent (57%) revascularization. Early amputations in 16 patients (23%) were performed either primarily (n = S), after inflow reconstructions to permit amputation healing (n = 7), or after graft occlusion (n = 1). Early deaths (n = 11[15.7%]) were caused by preoperative (n = 4) or postoperative (n = 2) myocardial infarction or metastatic cancer (n = 5). Graft patency and survival rates were calculated by life-table analysis and summarized below. These data emphasize that thrombosis, not embolism, is the overwhelming cause of ALE1 at the present time. Anticoagulation and revascularization resulted in excellent graf? patency and excellent salvageof severelyischemic but viable extremities. Amputations were essentiallylimited to patients admitted with nonviable limbs. Early and longterm patient survival was related to coexisting medical conditions and not complications of ALEI. These data should serve as a standard for comparison with series utilizing thrombolytic therapy and/or interventional radiologic management of patients with severe acute lower extremity ischemia. Peripheral vascular bypass in juvenile onset diabetes mellitus: Are aggressive revascularization attempts justified?
Christopher J. Kwolek, MD, Frank B. Pomposelli, Jr., MD, Gary A. Tannenbaum, MD, et al. New England Deaconess Hospital, Boston, Mass.
The incidence of diabetes melhtus is bimodal with peaks occurring at puberty (juvenile onset) and in the fifth decade of life (adult onset). Do reported differences in the
Volllme 13 Number 6 June 1991
clinical, biochemical, autoimmune, and genetic features between these two groups extend to the outcome of peripheral vascular reconstruction in patients with juvenile onset diabetes mellitus (JODM)? The results of 67 lower extremity bypass procedures performed on 60 patients with JODM between Jan. 1, 1984, and Dec. 31,1989, were reviewed. Patients ranged in age between 29 and 59 years (mean, 44.4 years), with an average age of onset of JODM of 9.8 years (range, 1 to 19). These procedures comprised 4.1% (67/1632) of all bypass procedures and 5.5% (67/1214) of bypasses performed on patients with diabetes during the same time period at a single institution. Ninety-one percent (54/67) were done for limb salvage. Fifty-four (8 1%) procedures were primary infrainguinal bypasses with saphenous vein (femoropopliteal 19, femorodistal or popliteal-distal 35). Six procedures (9%) were reoperations with vein, two (3%) were infrainguinal procedures with polytetrafktoroethylene, and five (7%) were in-flow procedures. Perioperative (30&y) mortality and morbidity rates were 0% and 31%, respectively. Actuarial patency and limb salvage of the primary vein graft group were 75.4 k 9.7% and 83.4 k 8.0%, respectively at 24 months. Cumulative survival of the entire group at 2 years was 84.1%. These results in patients with JODM compare favorably with published reports of peripheral bypass operations in adult onset diabetic and nondiabetic populations. The presence of JODM should not diminish the vascular surgeons’s expectations of successful outcome and justifies an aggressive approach toward lower extremity revascularization in this group of patients. Is long vein bypass grafting from groin to ankle a durable procedure? An analysis of outcome Dhiraj M. Shah, MD, R. Clement Darling, MD, Robert P. Leather, MD, Benjamin B. Chang, MD, Jeffrey L. Kaufman, MD, and Kathleen M. Fitzgerald, BS, Albany Medical College, Albany, N.Y. Long vein bypass grafting from the femoral artery to the level of the ankle may be performed with good initial success despite extreme bypass length and limited outflow tracts. However, the long-term performance of these bypasses remains to be defined. During the last 10 years we have performed single greater saphenous vein in situ bypass grafting to the ankle level in 265 patients. There were 184 male and 81 female patients, and 62% of the patients had diabetes. All bypasses were performed for limb salvage. Operative mortality rate was 3.4%. Cumulative bypass patency was 79% at 3 years and 74% at 5 years. Similarly, limb salvage was 93% at 3 years and 89% at 5 years. The patency rate was similar for various inflow arteries (common femoral, 93 cases; proximal superficial femoral, 133 cases; and deep femoral, 39 cases) and outfIow tracts (dorsal pedal, 69 cases; anterior tibial, 58 cases; posterior tibia& 72 cases; and peroneal, 66 cases). Short bypasses, composite bypasses, free-vein grafts, and bypasses proximal to 10 cm above the ankle were excluded from this analysis.
These data show that a long bypass to the ankle level for limb salvage is a durable procedure. The basic concept of bypassing all occlusive disease to the distal open artery in patients undergoing limb salvage should be an acceptable dictum. Excellent long-term patency and limb salvage rates are achievable by following this principle. The dorsal pedal bypass -Moderate success in difKcuIt situations Eliazbeth B. Harrington, MD, Martin E. Harrington, MD, Harry Schanzer, MD, Julius Jacobson, II, MD, and Moshe Haimov, MD, Mount Sinai Medical Center, New York, N.Y. Recent reports have documented excellent results for inframalleolar reconstructions. We reviewed our outcomes for dorsal pedal bypass and report a more modest rate of success. We analyzed reasons for failure. Sixty-six patients underwent 70 dorsal pedal bypass procedures between 1984 and 1990. Seventy-five percent of the patients had diabetes. Inflow was from the femoral artery in 34, popliteal artery in 33, and tibial artery in 3. The operative indication was gangrene in 54%, ulcer in 25%, and rest pain in 21%. Forty-two percent of limbs had foot infection, with five requiring minor amputation before the bypass and 19 requiring minor amputation after bypass. There was one perioperative death. Twenty-seven grafts failed over the course of the series. The patency rate at 2 years was 61%. The limb salvage rate was 71%. Of the 10 perioperative failures, four were due to continued foot infection, four to marginal vein quality, and two to skin necrosis of the bypass incisions. Graft failure occurred at 3 to thirty months in 10 of 11 patients who required digital angiography, Doppler exam or intraoperative angiography to identify a patent dorsal pedalivation artery. These 11 patients had deficient anterior arches with segmental occlusion of the dorsal pedal or its branches. Five of the six patients with extensive infections of the forefoot or heel required amputation with patent bypasses. In dorsal pedal bypass poor outcome occurred in patients with extensive foot infection, marginal vein quality, and deficient anterior pedal arches. Primary amputation should be considered in these patients. From the Society Research Forum
Future alternatives to heparin: low molecular weight heparin and hirudin Carlos E. Donayre, MD, Kenneth Ouriel, MD, Robert Y. Rhee, MD, Victor J. Marder, MD, Richard M. Green, MD, and James A. DeWeese, MD, Univershp of Rochester Medical Center, Rochester, N.Y. Great advances in vascular surgery would not have been possible without the advent of heparin therapy, yet the complications of bleeding, platelet aggregation, and thrombocytopenia have compelled investigators to contin-
ually serarch for a safer antithrombotic agent. We investigated the dose-responsive relationships of heparin, low molecular weight heparin (LMWH), and hirudin in an in vitro perfusion system. Fresh human blood was treated with varying doses of the agents and perfused for 180 seconds over 70 everted, deendotheliahzed porcine venous segments at a shear rate of 100 set-‘. The effectiveness of the agents as anticoagulants was assessed by three parameters: (1) the dose required to double the activated clotting time, (2) the anti-Xa activity, and (3) the amount of iodine 125-labeled fibrin deposited on the vessel wall during the perfusion period. The effectiveness of the agents as antiplatelet agents was assessed by a single parameter, the anount of indium Ill-labeled platelets accumulating on the vessel wall during perfusion. The fibrin deposition data demonstrated that hirudin was the most effective anticoagulant, with heparin’s anticoagulant activity between that of hirudin and LMWH. The low aXa associated with hirudin reiterates its purported direct effect on thrombin, in contrast to the indirect, antithrombin III mediated actions of heparin and LMWH. LMWH was the only agent to exihibit selectivity between anticoagulant and antiplatelet effects. Intermediate doses of LMWH (2.0 mcg/ml) inhibited 70% of FIB but only 43% of PLATE. We conclude that hirudin is a powerful anticoagulant and antiplatelet agent. As such, it can be used as a heparin substitute when complete antithrombotic activity is imperative. LMWH may be more appropriate for those instances where anticoagulation without antiplatelet activity is desired, as is advantageous during aortic procedures. These and other heparin alternatives will provide a variety of agents with discrete indications for specific clinical situations, maximizing effective antithrombotic activity and patient safety. New RGD-analog eliminates platelet deposition on vascular grafts in thrombosis-prone dogs Brian G. Rubin, MD, Daniel J. McGraw, MD, Gregorio A. Sicard, MD, and Samuel A. Santoro, MD, PhD, Washington University School of Medicine, St. his, MO. Platelet adhesion and aggregation are mediated by fibrinogen via the receptor glycoprotein IIbflIIa, which recognized the Arg-Gly-Asp (RGD) sequence. We investigated the ability of 8-guanidino-octanoyl-Asp-Phe, an intravemusly infused, stable RGD analog, to inhibit platelet deposition on prosthetic grafts. Eight thrombosis-prone dogs had polytetrofluoroethylene (PTFE) femoral artery grafts. Dogs received RGD analog (1 mg/kg bolus then 50 ug/kg/min) or normal saline infusion during their first graft procedure. One week later a second contralateral femoral graft with infusion of the other agent was performed. Indium III-labeled autologous platelets were introduced, and either normal saline or the RGD-analog was intravenously infused for 2 hours. Platelet counts and specific activity (counts/platelet) were monitored initially and every 30 minutes. Aggregometry
Journal of VASCULAR SURGERY
during RGD analog infusion demonstrated inhibition of induced aggregation, whereas normal saline infusion had no effect. After infusion, indium III platelets deposited on the explanted graft were measured in gamma counter. Platelets per cm’ were calculated: plts/cm2 = total graft counts/specific activity x graft surface area. Platelet deposition was reduced by S-guanidinooctanoyl-Asp-Phe on vascular grafts by more than 90% (p = 0.0006, log transformed data, paired t test). Graft histology demonstrated deposited platelets within the interstices of the graft, not on the luminal surface. 8-guanidino-octanoyl-Asp-Phe represents a novel class of glycoprotein IIbflIIa inhibitors, which may have a broad range of clinical applications. Differential response of arteries and vein grafts to blood flow reduction Spencer W. Gait, MD, Robert J. Wagner, BS, Jeffrey J. Gilbertson, MD, and Robert M. Zwolak, MD, PhD, Dartmouth-Hitchcock Medical Center, Hanover, N.H. Shear stress variation caused by alteration in blood flow may affect artery and autogenous vein graft architecture differently. Jugular vein grafts were placed in both common carotid arteries of 12 rabbits. Blood flow was reduced on one side without alteration in blood pressure by external carotid artery ligation resulting in flow rates of 13 2 4 (mean t SEM) ml/min (ligated) versus 21 + 4 ml/min (control; p < 0.05). Grafts and proximal arteries were harvested after 4 weeks by pressure perfusion fixation. The carotid artery underwent media thickening with reduced lumen diameter and shear stress normalization, but no intima thickening, smooth muscle cell (SMC) proliferation or change in wall area. In vein grafts, an augmented SMC proliferative response occurred in both the intima and the media, with an increase in wall area but no reduction in lumen diameter. We conclude that arteries and vein grafts respond to blood flow reduction by wall thickening, but that the mechanisms differ substantially. The arterial response is one of media remodeling and shear stress normalization. Whereas SMC proliferation and wall area enlargement characterize the vein graft response. SaraIasin inhibits intimal hyperplasia Xian M. Pan, MD, Nicolas Nelken, MD, Nicolas Colyvas, MD, and Joseph H. Rapp, MD, San Francisco Medical Center, San Francisco, Cal& Sponsored Ly Jerty Goldstone, MD, San Francisco, CalijY Increasing evidence points toward local production of renin and angiotensinogen in the artery wall. Because angiotensin converting enzyme (ACE) inhibitors have been shown to block intimal hyperplasia after aortic injury in the rat, it has been suggested that angiotensin II is an important mediator of the proliferative response to arterial injury. To prove that previous observations using ACE inhibitors are a result of effects on local angiotensin levels versus nonspecific drug effects, we tested the ability of an
v01umc 13 Number 6 June 1991
unrelated drug, the angiotensin II receptor antagonist saralasin, to similarly block intimal hyperplasia after aortic injury in the rat. We performed balloon catheter aortic denudation in 28 rats pharmacologically treated from 6 days before to 14 days after surgery and split into four groups: group 1, saralasin 360 &kg/hr intravenously (IV); group 2 (negative control), normal saline 0.5 p$hr IV; group 3 (positive control) captopril 100 mg/kg/d orally; and group 4 (positive control), heparin 50 U/kg/hr IV. Weight gain was similar in all groups. Blood pressure in the captopril treated group decreased from 107.4 * 4.3 to 96.3 + 4.7 mm Hg @ < 0.01) after 6 days, whereas saralasin and heparin had no effect on blood pressure. Animals were killed and aortas were perfusion fixed at physiologic pressure 14 days after denudation. Cross-sectional intima to media ratios were calculated by computerized planimetry. Compared with saline controls, saralasin inhibited intimal hyperplasia 45% (p < O.OOl), captopril 59% (p < O.OOl), and heparin 68% (p < 0.001). This work indicates that attenuating the angiotensin system both at the level of ligand production and receptor blockade interferes with the proliferative response to vascular injury. These results in combination with evidence of local production in the artery wall provide strong support for the hypothesis that angiotensin II contributes to smooth muscle cell proliferation after aortic injury in the rat. PDGF production of canine aortic grafts seeded with endothelial cells Bram R. Kaufman, MD, Paul L. Fox, PhD, and Linda M. Graham, MD, Veterans Administration Medical Center, Cleveland, Ohio. An endothelial cell (EC) surface on prosthetic grafts decreases thrombogenicity, but growth factor production by ECs may increase graft failure caused by smooth muscle cell proliferation. The present study assessed plateletderived growth factor (PDGF) production by cells lining prosthetic grafts and by adjacent arteries. Thirteen beagles underwent placement of 20 to 22 cm long, 8 mm ID, expanded polytetrafluoroethylene thoracoabdominal aortic grafts. In one group of dogs (n = 6), grafts were seeded with autologous jugular vein ECs, whereas in remaining animals (n = 7) grafts were unseeded. Grafts and adjacent arteries were explanted after 4 weeks and studied in organ culture. Platelet-derived growth factor production by all segments was measured after 72 hours by use of a radioreceptor assay. Measurement of protein synthesis by C3H]leucine incorporation demonstrated continued viability of segments in organ culture. Platelet-derived growth factor production correlated directly with EC coverage on graft segments as measured by scanning electron microscopy (Y = 0.63, p = 0.02), but inversely with platelet deposition (r = 0.46, p = 0.01). The results of this study suggest that ECs on prosthetic vascular grafts may be a significant source of
PDGF. Futhermore, the finding that PDGF production by arteries distal to the grafts is significantly higher than the thoracic aorta (p < 0.0001) may offer an explanation for the clinical finding of more severe intimal hyperplasia adjacent to the distal anastomosis. Shear stress modulates the proliferation rate, protein synthesis, and mitogenic activity of arterial smooth muscle cells (SMC). Antonio V. Sterpetti, MD, Alessandra Cucina, PhD, Luciana Santoro, PhD, Barbara Cardillo, PhD, and Antonino Cavaliaro, I Clinica Chkqtia-Universita Di Romu La Sapienut, ha, Italy, Sponswed by: Joseph l? Arc&e, MD, Rule&b, N.C. To define the correlation between flow dynamics and the abnormal proliferation of vascular SMCs that occurs during the development of atherosclerosis, bovine arterial SMCs were subjected to increasing laminar flow shear in an original in vitro system. Smooth muscle cells from passages 6 through 10 were seeded in a fibronectin coated polystyrene cylinder at 5 x lo5 cells/tube and allowed to reach confluence and to adhere for 48 hours. The experimental groups were subjected to a laminar flow of 150 ml/min (9 dyne/cm’), 100 ml/min (6 dyne/cm’), and 50 ml/min (3 dyne/cm’) for 24 hours. The control group was subjected to similar incubation conditions without flow. The cells in the experiments remained attached and viable. All experiments were performed at least in triplicate. Shear stress reduced significantly (p < 0.001) the 24-hour incorporation of tritiated thymidine and cell proliferation. This effect was proportional to the level of shear stress and was still evident 24 hours after flow cessation. Flow cytometry confirmed a lower percentage (p < 0.01) of SMCs in S phase with increasing shear stress (20% 0 ml/min; 11% 50 ml/min; 8% 150 ml/min). Synthesis of cell-associated proteins was increased twofold @ cc 0.01) in SMCs subjected to laminar flow. Smooth muscle cells subjected to shear stress released a higher quantity of mitogens, including a platelet-derived growth factor (PDGF)-like substance as detected by immunologic tests. Fifty percent vol/vol conditioned serum-free medium from SMCs subjected to shear stress increased tritiated thymidine uptake in PDGF receptor-bearing Swiss 3T3 cells threefold as compared to conditioned serum-free medium from control SMCs not subjected to shear stress, and twelvefold as compared to standard control. The release of mitogens was proportional to the level of shear stress and was still evident 48 hours after flow cessation. The mitogenic activity was partially reduced (30%; p < 0.01) by an excess of monospecific anti-PDGF antibody. We conclude that (1) increasing shear stress inhibits SMC proliferation and stimulates the synthesis of cellassociated proteins and release of mitogens; (2) decreasing shear stress facilitates proliferation of SMCs, which maintain, even if reduced, the capability of mitogens release. Thus in situations of arterial flow separation, the increased release of mitogens from SMCs subjected to high shear
Journal of VASCULAR SURGERY
stressand the increased proliferation rate and susceptibility to mitogens of SMCs subjected to very low shear stress, may generate a critical condition that predisposes to the development of atherosclerosis, with early plaque formation in regions of low flow shear stress. Anastomotic intimal flow induced?
Hisham S. Bassiouny, MD, Scott White, MS, Eric Choi, MS, Don I’. Giddens, PhD, Seymour Glagov, MD, and Christopher K. Zarins, MD, Uniperrity of Chicago, Chkgo, Ill. AU anastomotic intimal thickening (AIT) may not be the same, and the underlying mechanism(s) regulating the different types may vary. The purpose of this study was to determine whether different types of AIT occur, quantitatively assessAIT in both vein and polytetraf-luoroethylene (PTFE) anastomoses, and relate these findings to known mechanical and hemodynamic factors. In vivo studies: Bilateral iliofemoral bypasses(saphenous vein and PTFE) were implanted in 10 mongrel dogs. The distal end-to-side anastomosis geometry was standardized with a ratio of hood length to vesseldiameter of 4 : 1, and the flow parameters were measured. After 8 weeks 7 animals (group I) were killed, and the anastomoses (n = 14) were pressureperfusion fixed. Histologic sections (n = 22) from each anastomosis were studied with light microscopy, and regions of AIT were mapped and quantitated using oculomicrometry. Model flow studies: To characterize the flow patterns in regions of AIT, transparent silicon models were constructed from castings of the distal anastomoses (n = 6) of three animals (group II). Flow was visualizedusing He-Ne laserilluminated particles under conditions simulating the in vivo pulsatile flow parameters. Histologic sections (n = 154) revealed two separate and distinct regions of AIT. The first occurred at the site of direct mechanical injury; suture line AIT was greater in PTFE anastomoses (0.35 ? 0.23 pm) than in vein anastomoses (0.15 ? 0.10 krn,P < 0.03). The seconddistinct type of AIT developed on the arterial floor of the distal anastomosis and was no different in PTFE (0.11 + 0.11 pm) and vein anastomoses (0.12 + 0.03 prn,P > 0.50). Model flow visualization studies revealed flow reattachment, low and oscillatory shear along the arterial floor where AIT developed. High shear and short particle residence time were observed along the hood of the graft, an area not associatedwith AIT. Regions of relatively low
shear and short particle residence time formed along the lateral walls, toe, and heel of the anastomoses and were also not specifically related to suture line AIT. These studies suggest that at least two different types of AIT exist, one induced by mechanical and other by hemodynamic factors. Suture line AIT ia a feature of vascularhealing; greater prominence with prosthetic grafts may be related to compliance mismatch. Arterial floor AIH develops in regions of flow stagnation and oscillation, and is unrelated to graft type. Prevention of graft failure caused by AIT in the clinical setting requires preciseunderstanding of the control mechanisms of each different type. The effect of thrombin on LDL permeability by an arterial endothelial-smooth muscle
and uptake cell bilayer
J. Jeffrey Alexander, MD, Remedios Miguel, MS, Debra Graham, MD, and Joseph Piotrowski, MD, Cheland Metropolitan General Hospital, Cleveland, Ohio, Sponsored by: Linda Graham, MLI, Cleveland, Ohio
Thrombin has been shown to influence the rate of replication and matrix element production by arterial endothelial cells (EC) and smooth muscle cells (SMCs). The following study was designed to determine the effect of thrombin on the permeability and cellular uptake of low density lipoprotein (LDL) by an EC-SMC bilayer. Confluent bovine aortic EC-SMC bilayers were established by use of a transwell coculture system.Thrombin (50 units/ml) was added to the media of the upper (EC) chambers. After 24 hours, iodine 125-LDL (10 mg protein/ml) was also added. After a 3 hour incubation period, the “‘1-LDL content of the lower well (SMC) media was determined. Membrane bound and intracellular LDL was measured for both cell types, and the result was compared with those of untreated cells. Thrombin was found to enhance EC permeability to LDL in the bilayer, increasing its influx from 44.7% +- 2.8% (control) to 53.3% 2 2.3% (treated; p < 10e7). Endothelial cell binding and uptake of LDL was reduced by thrombin. Uptake of SMCs was significantly increased, whereas SMC binding was not affected. Thrombin increased the LDL uptake/binding ratio for SMCs, but resulted in a reduced ratio for ECs. The results indicate that thrombin may increasevascular permeability of LDL in a manner independent of EC binding or uptake, and that it may subsequently enhance SMC uptake of LDL in a nonreceptor mediated fashion. Such findings could implicate a role of thrombin in the promotion of atherosclerotic plaque formation.