Joint annual meeting abstract

Joint annual meeting abstract

JOINT ANNUAL The Surgery Surgery, Ill., on meeting meeting. MEETING Scientific Program of the Society for Vascular and the International Society for...

3MB Sizes 1 Downloads 26 Views

JOINT ANNUAL The Surgery Surgery, Ill., on meeting meeting.


Scientific Program of the Society for Vascular and the International Society for Cardiovascular North American Chapter, was held in Chicago, June 7-11, 1992. Selected abstracts from that are published here in order of presentation at the

From the Scientific Program of the Forty-sixth Annual Meeting of the Society for Vascular Surgery Is the iliac artery a suitable inflow conduit for iliofemoral occlusive disease: An analysis of 514 aortoiliac reconstructions R. Clement Darling III, MD, Angela Wellhausen, Benjamin B. Chang, MD, Robert I?. Leather, MD, and Dhiraj M. Shah, MD, Albany Medical College, Albany, N.Y. The aorta is the conventional inflow source for reconstructions in patients with aortoiliofemoral occlusive disease. In patients with unilateral iliac or femoral disease, femoral-to-femoral bypasses have been used but with less favorable patency rates. The purpose of this study is to evaluate the performance of the unobstructed iliac artery as an inflow source for ipsilateral, contralateral or bilateral reconstructions in iliofemoral occlusive disease. Over the past 6 years, 322 reconstructions have been performed with use of the iliac artery as the donor vessel. Patients were evaluated for proximal hemodynamically significant lesions by augmented pullout pressures during aortography. Patients who had balloon angioplasty were excluded. Results were compared with 192 patients who underwent conventional aortodistal bypass for occlusive disease during the same period. Both groups were similar in risk factors, age, sex, and indications for surgery. For the iliac group, the operative mortality rate was 3.4%, and the 30-day patency rate was 97%, similar to those in the aortic group (4% and 99%, respectively). Cumulative patency rates at 5 years by life-table analysis were 79% for iliac artery inflow and 85% for aortic inBow reconstructions. Our experience suggests that unobstructed iliac artery is a reasonable inflow source for reconstructions in iliofemoral occlusive disease. The long-term patency rate is comparable to aortodistal bypasses and superior to other extraanatomic bypasses. Ultrasound can be used to guide and monitor balloon angioplasty Scott R. Cluley, MD, Bruce J. Brener, MD, Larry Hollier, MD, Donald K. Brief, MD, Joseph Alpert, MD, Robert Goldenkranz, MD, David Eisenbud, MD, Jan Huston, MD, Victor Parsonnet, MD, and Michelle Ferrara-Ryan, BS, Newark Beth Ipael Medical Center, Newark, N.J.


Percutaneous balloon angioplasty is an accepted technique for the treatment of short segmental stenoses of the iliac and superficial femoral arteries. Surgeons have not embraced this technique because of lack of training, inability to use a radiologic facility, and poor quality radiographic equipment in the operating room. A new technique, ultrasound directed balloon angioplasty, enables the vascular surgeon to guide and evaluate the progress of the procedure without the use of radiographic imaging. This technique eliminates the risk of ionizing radiation and contrast nephropathy, and allows a physiologic assessment of the procedure. The new catheter is integrated externally into a standard duplex scanner. A spherical brass bead positioned on the catheter shaft in the middle of the balloon is covered with a piezoelectric material. This omnidirectional receiver is connected through a wire to a catheter interface system that allows the exact position of the balloon to be represented on the screen of a duplex unit. The stenotic lesion can be identified by duplex scanning. The same unit can be used to guide the catheter to the lesion and to monitor and verify the success of the dilation. The positional accuracy and efficacy of this system has been successfully demonstrated previously in animal studies in our laboratory. The catheter has been used to perform balloon angioplasty in 12 lesions in 8 patients (8 superficial femoral, 3 iliac, 1 popliteal arteries). Seven procedures were performed percutaneously; in one patient bilateral angioplasties were performed during operation. The balloon was successfully positioned in each case by use of ultrasonography alone; each procedure was confirmed by angiography. Angioplasty was accomplished in 11 of 12 attempts. Mean velocity was reduced from 3.7 m/set to 1.5 m/set. Patency has been maintained in all but one patient (mean follow-up, 6 months). This technique records physiologic and anatomic data in real time so that the progress of the angioplasty can be monitored. In three cases the inadequacy of the angioplasty was recognized, leading to correction. This new procedure allows vascular surgeons to use equipment with which they are familiar and to monitor angioplasty with physiologic parameters. This may result in better long-term results.

Infrainguinal reconstruction for patients with chronic renal failure Anthony D. Whittemore, MD, Magruder C. Donaldson, MD, and John A. Mannick, MD, [email protected] and W&men’s Hospital, Boston, Mass. While infrainguinal reconstruction yields highly acceptable results in most patients, revascularization of ischemic limbs in those with renal failure remains challenging. The purpose of this review was to determine the 1057

Joournal of 1058







No. of Total Dialysis Diabetes

(+ ) C-1 (+ ) (-1

No. of



55 11 45 28 27

69 15 54 37 32

(Whittemore, SO-day Mortality 12% 20% 9% 14% 9%

efficacy of autogenous infrainguinal reconstruction in patients with chronic renal failure. During the past 15 years, 55 patients with chronic renal failure underwent 69 autogenous vein bypass procedures. All patients had serum creatinine levels greater than 2 mg/dl, 11 (22%) were on maintenance dialysis and 28 (54%) were diabetic. The majority of procedures (84%) were carried out for limb salvage; 36 (52%) consisted of femoropopliteal and 33 (48%) femorotibial reconstmctions. During the same period, 34 patients with renal failure underwent 45 primary major amputations. Primary amputation was associated with a 30-day operative mortality rate of 13%, and major morbidity occurred in 10 patients (22%). Results from primary reconstruction are in the table. Morbidity and death resulting from primary amputation in patients with chronic renal failure was virtually identical to that associated with primary reconstruction. Patients with chronic renal failure who do not require dialysis achieve patency and limb salvage results comparable to an unaffected population, yet their operative morbidity and mortality rates are higher and survival rate substantially lower, especially for those with diabetes. Patients requiring hemodialysis demonstrated minimal patency and limb salvage rates, and none survived 3 years. Temporizing measures using nonoperative endovascular techniques may be more appropriate for diabetic patients with renal failure and for patients on maintenance hemodialysis to minimize the high morbidity and mortality rates associated with both primary amputation and reconstruction.

Percutaneous transhninal renal angioplasty vs. reconstructive renovascular surgery in atherosclerotic unilateral renal artery stenosis-A prospective raudo&d study Henrik Weibull, MD, Sven-Erik Bergentz, MD, David Bergqvist, MD, Lennart H&hen, MD, Per Manhem, MD, Kjell Jonsson, MD, and Fred& Norstrom, MD, Lund University, MahuY, Sweden The purpose of this study was to results, primary and secondary patency of significant restenosis), and effects on renal function between percutaneous angioplasty (PTRA) and reconstructive

compare technical (defined as absence blood pressure and transluminal renal renovascular sur-


Mannick) Ptimavy


30-day Morbidity



23% 27% 22% 16% 31%

79% 22% 86% 73% 84%

74% 0 81% 55% 84%

Limb salvage 84% 75% 83% 63% 91%

(3 yr)

Smvi9al 5YY 40% 0 46% 21% 57%

gery (RVS) in patients with atherosclerotic unilateral juxtaaortic RAS in a prospective randomized study. Fifty-eight nondiabetic patients (35 men; 5 70 years) with severe hypertension, normal renal function, and hemodynamically significant RAS ( 5 2 mm diameter and/or renal vein renin ratio 2 1.5) were randomized to PTRA or RVS. Renal angiography was performed 10 days, 1 year, and 2 years after PTRA/RVS, and simultaneously blood pressure and renal function were evaluated. Results were as follows: (1) Technical results: PTRA was successful in 24 of 29 (83%) patients and RVS in 28 of 29 (97%). (2) Primary patency of technically successful cases: At end of follow-up (24 months) 18 of 24 (75%) in the PTRA group and 27 of 28 (96%) were patent. (3) Secondary patency: After additional intervention (rePTRA or reRVS), patency in the PTRA group was 26 of 29 (90%) and in the RVS group 28 of 29 (97%). (4) Blood pressure and renal function: After additional treatment 27 of 29 (93%) of the patients in the PTRA group and 28 of 29 (97%) in the RVS group were normotensive or had improved blood pressure. The corresponding figures for improved or unchanged renal function were 25 of 29 (86%) and 26 of 29 (90%), respectively. Technical failures and restenoses were significantly more frequent after PTRA than after RVS. Secondary patency was similar in the two groups, and effects on blood pressure and renal function did not differ. PTRA with close follow-up and reintervention is acceptable as the initial treatment in atherosclerotic unilateral renal artery stenosis.

Impact of a color-flow duplex surveillance program on infrainguinal vein graft patency. A five-year experience Mirza Idu, MD, Peter de Gier, MD, Ellie Truyen, RVT, and Jacob Buth, MD, Catharina Hospital, Eindhmen, The Netherlands Sponsored by: R. Clement Darling m, M.D., Albany, N. Y. Color-flow duplex scanning for the surveillance of infrainguinal vein grafts has been demonstrated to be effective to detect preocclusive stenotic lesions. The value of this advanced screening method in improving overall bypass patency by early repair of graft lesions, however, is uncertain. In addition, doubts remain about which stenoses should be repaired and which not. Two hundred one infrainguinal bypass procedures

Volume Number

15 6


June 1992


Difference between grafts (Mattos et al.)









0.97 0.96 0.85

0.90 0.96 0.76

0.83 0.88 0.68

24 39

100 100

with autologous vein were performed in 187 patients during a ii-year period. Only grafts that were open after the first postoperative month were considered. One hundred sixty bypasses were monitored with color-flow duplex scanning of the entire graft and its adjacent arterial segments at 3-month intervals during the first year (group A). Forty-one bypassesdid not undergo color-flow duplex surveillance for various reasons (group B). In this latter group follow-up comprised only regular conventional assessment.Group A and group B were comparable with regard to most pertinent clinical factors. Stenotic lesions were identified in 58 bypassesand its severity determined by intraarterial digital subtraction angiography. A more optimal use of revisional procedures in group A was reflected by a 3-year secondary patency of 92% compared with a patency of 69% in group B (log rank, p = 0.004). Eighteen bypasseswith stenoses did not undergo a revision because of doctors, hospital, or patient reasons. The occlu.sionrate of revised and nonrevised stenotic grafts was compared for lesions of different severity. In the category with stenosesof 30% to 49% diameter reduction (DR) none of the grafts failed during follow-up. In stenoseswith 50% to 69% DR, 57% of nonrevised and 9% of revised bypassesoccluded. Stenoses70% or greater DR were associatedwith graft failure in all nonrevised bypasses and in 10% of revised grafts. Overall graft patency can be improved by color-flow duplex surveillance and repair of all sign&ant stenotic lesions. Does correction of stenoses identified with color duplex scanning improve infrainguinal graft patency?

Mark A. Mattos, MD, Paul S. van Bemmelen, MD, Kim J. Hodgson, MD, Don E. Ramsey, MD, Lyrme D. Barkmeier, MD, and David S. Sumner, MD, Southemz12linois Univemity School





Although duplex scanning provides a method for identifying stenoses in infrainguinal bypass grafts, little data exist to show that correcting these stenoses prolongs graft survival. Over a 39-month period, 546 color duplex scans(CDS) were performed on 205 grafted limbs. Grafts were scannedwithin 3 months of operation and then every 6 months thereafter. Autogenous veins were used in 170 and polytetrafluoroethylene (PTFE) or composite PTFEvein in 35. Locations of distal anastomoses were: aboveknee popliteal in 41, below-knee popliteal in 46, and tibioperoneal in 118. Doubling of the velocity at any point in the graft-arterial systemcompared with that immediately above or below was the criterion adopted for identification



0.83 0.88 0.68

of a stenosis.A total of 122 stenoseswere detected in 72 or 35% of the limbs, of which 11 (9%) were in native vessels, 35 (29%) were at the anastomoses, and 761(62%) were in the graft itself. At the discretion of the attending surgeon, 27 (38%) of the grafts with positive scanswere revised. During follow-up, occlusions occurred in I9 (14%) of the 133 grafts with negative scans (NEG), in 3 (11%) of the positive revisedgrafts (PR), and in 10 (22%) of the positive nonrevised grafts (PNR). Log rank tests of life tables for the vein grafts showed a significant difference between the patency ratesof NEG and PNR grafts at 1 year @ = 0.04) and 3 years (p = 0.05). Differences approached significance at 1 year @ = 0.090) between the PR and PNR grafts shown in the table. Flow velocities lessthan 45 cm/set and ankle-brachial indexes did not discriminate between grafts with or without stenosis or identify those grafts that subsequently occluded. The results of this study suggest that revision of grafts with stenosesidentified with color duplex scanning prolongs patency, but that most grafts with stenosesremain patent even when they are not revised. Stenoses,therefore, appear to be only one of many factors that affect patency. Surgical treatment in women

of iufrainguinal



Joseph G. Magnant, MD, JackL. Cronenwett, MD, Daniel B. Walsh, MD, Sharon R. Besso,RN, Robert M. Zwolak, MD, and Joseph R. Schneider, MD, Dartmouth-Hitchcock Medical

Center, Lebanon, N.H.

Although infrainguinal bypass grafts are performed frequently, their utility in female patients has not been separately analyzed. Since women are likely to be older at presentation and have smaller blood vesselsthan their male counterparts, they might experience different treatment outcomes. We reviewed the records of 100 consecutive women who underwent infrainguinal bypasses in 115 limbs from 1984 to 1990 for limb-threatening ischemia (94%) or disabling daudication (6%). These women ranged in age from 39 to 95 years (mean, 72); 54% were diabetic, 68% were hypertensive, but only 59% had a history of smoking. Infrainguinal disease distribution necessitated bypass to the above-knee popliteal artery in 15%, to the below-knee popliteal artery in 26%, and to the tibial arteries in 59%. Autogenous vein grafts were used in 80% (ISSVG, 78%; other, 22%) and polytetrafluoroethylene or polytetrafluoroethylene-vein composite gratis in 20%. Nearly all grafts were primary procedures, while 10% were secondary. Prior percutaneous transluminal angioplasty (PTLA) had been performed in 11% of the arteries




that subsequently required bypass. Inflow procedures (PTLA or bypass)had preceded 20% of these infrainguinal grafts. The perioperative (30-day) mortality rate was 2% (myocardial infarction, 1; stroke, 1). However, life-table analysisdemonstrated l- and 3-year patient survival of only 87% and 59%, respectively.Primary graft patency at I and 3 years was 66% and 55%, respectively. Secondary graft patency was 83% and 75%, respectively, at the same intervals. Cumulative limb salvagewas 89% and 85% at 1 and 3 years, respectively. Five limbs (4%) required amputation within 1 month of bypass. Ankle-brachial indexes improved from a mean of 0.38 before bypass to 0.83 after bypass. Limb-threatening ischemia requiring revascularization developed in the contralateral lower extremity in 15% of patients during follow-up. Compared with contemporary seriesof male patients, women who underwent infrainguinal bypass were more elderly, more often diabetic, but less frequently smokers. They underwent surgery almost exclusivelyfor limb threat, and the majority required tibial bypass. Limb salvagewas achieved in 85% of patients. Despite an acceptable perioperative mortality rate, average survival was only 4 years after revascularization, considerably less than the expected 14-year life span for age-matched disease-free controls. We conclude that infrainguinal revascularization provides effective limb salvagefor elderly women in their last years of life. Mesenteric tive study






Raymond W. Lee, MD, Gregory L. Moneta, MD, John D. Caster, RN, Cary A. Cummings, RN, and John M. Porter, MD, Oregon Health Sciences Dzivevsity, Portland, Ore. The clinical utility of duplex scanning of the celiac(CA) and superior mesenteric (SMA) arteries has been evaluated in several preliminary retrospective studies. However, neither the ability of mesenteric duplex scanning (MDS) to visualize the SMA and CA nor duplex criteria for highgrade SMA and CA stenosis have been prospectively validated, a critical requirement to determine the usefulness of MDS. To develop and validate MDS criteria, we performed 383 mesenteric duplex scanson vascularsurgical patients over a a-year period. After 6 months, initial duplex data were compared with angiography. This retrospective review suggested an SMA peak systolicvelocity (PSV) 275 or greater cm/set or no flow signal and a CA PSV 200 or greater cm/set or no flow signal indicated a 70% or greater stenosisin the SMA and CA. Over the next 18 months we blindly and prospectively evaluated these MDS criteria in 100 patients with anteroposterior and lateral angiograms of their CA and SMA. MDS was evaluated for its ability to visualize the CA and SMA and to detect a 70% or greater stenosis by the above criteria. One hundred percent of the CAs and 99% of the SMAs were satisfactorily visualized by angiography. Of these, 93% of the SMAs and 83% of the CAs were visualized by

MDS. Comparisons between MDS and angiography, in vesselsvisualized with MDS, for detection of 70% or greater stenosis are presented in the table. With the proposed criteria, our blinded, prospective study indicates MDS sensitivity, specificity, and positive and negative predictive values for detection of 70% or greater stenosisto be lOO%, 96%, 81%, and 100% for the SMA and 91%, 80%, 63%, and 96% for the CA. These are the first prospectively validated MDS criteria and indicate MDS is feasiblein most patients and sufficiently accurate to be useful asa screening examination to detect SMA and CA stenosis. Long-term to 21-year

results of venous valve reconstruction: follow-up

A 4-

Elna M. Masuda, MD, and Robert L. Kistner, MD, [email protected] Pacijc Health Fomm!attin, Honohlu,


The ultimate test of treatment in chronic venous insufficiency (CVI) is the function of the extremity over the very long follow-up. This report describes the clinical, physiologic, and imaging results after venous valve reconstruction for reflux diseasein 51 extremities (48 patients) who have had long-term follow-up of 4 to 21 years, with a mean of 10.5 years. CVI caused by reflux diseasewas classified into three groups: Pure primary valve insufficiency (PVI) 47%, pure secondaryor postthrombotic syndrome (PTS) 29%, and a mixed category of proximal PVI with distal PTS (PVIPTS) in 24%. With use of strict clinical criteria, 59% overall were permanently converted from class3 to class0, 1, or 2 status. Thirty percent of the patients had permanent excellent results in that they were restored to full activity without elastic support, recurrent ulcerations, pain, or severe swelling (class 0 or 1). Clinical results after valvuloplasty for PVI or PVI-PTS disease (32 patients) yielded 70% long-term good to excellent results, as compared with 37% after transposition for pure PTS disease (I7 patients). Results of venous valve reconstruction were clearly superior in the PVI or PVI-PTS group when compared with those with pure PTS; this may reflect a difference in the basic diseaseprocess, or a difference in the durability of the valvuloplasty versus the transposition operation. Ambulatory venous pressure studies showed improvement or normal results in 50% (10 of 20) of those with PVI or PVI-PTS disease,as opposed to 8% (1 of 12) in the pure PTS group. Postoperative descending venography and/or duplex scanning showed full competence or mild reflux in 67%. Competence or mild reflux was found in 94% (16 of 17) of the excellent clinical group. When perforators were present but left untreated, the ulcer recurrencerate was 60%. This was significantly higher than the recurrence rate of 30% found in those with no perforators or with perforators treated by interruption. This report highlights a difference in the long-term results between class3 patients whose reflux state is treated surgicallyfor PVI (70% long-term control) and PTS (37% long-term control). It also indicates that better results are

Volume 15 Number 6 June 1992



between MDS and angiography SMA

Duplex Duplex

70%-100% < 70%

(Lee et al.)

(N = 92)

Angiography 70%-100% 13 0

obtained when both deep vein reflux incompetence are controlled surgically.


and perforator

Early detection of aortic graft infection with 99mTo-hexametazime labeled leukocytes Paolo Fiorani, MD, Francesco Speziale, MD, Vittorio Faraglia, MD, Luigi Rizzo, MD, Francesco De San&, MD, Gregory Joseph Massimi, BSc, Maurizio Taurino, MD, Mauro Liberatore, MD, and Marta Clemente, MD, Universiq of Rome %a Sapienza-5: Rome, Italy Surgical treatment of aortic graft infection (AGI) is associated. with high rates of morbidity and mortality. A major challenge in treating AGI is early and accurate diagnosis. Surgical treatment is rarely performed in early stages of AGI because of the nonspecific symptoms and of the inadequacy of diagnostic techniques (ultrasound, CT, MR). Current isotopic techniques (67-Ga, 111-In) have also several diagnostic limitations. A new method of leukocyte labeling using 99m-Technetium (99m-Tc) may allow quicker, higher definition imaging. The ability of 99m-Tc to detect AGI was studied. Patients and methods. 37 patients (1985 to 1991) with clinically suspected AGI were divided into three groups: group A: 10 patients (27%) groin abscess/graft exposure. Group B: 18 patients (49%) Nonspecific symptoms ( ‘T’ESR and WBC, general malaise). Group C: 9 patients (24%) Anastomotic aneurysm. CT scanning and 99m-Tc were performed and interpreted according to accepted methods. The reliability of 99m-Tc to detect AGI was evaluated using microbial culture results obtained from intraoperative graft and perigraft exudate material. Follow-up data (18 months; range 11 to 42) was used to evaluate 99m-Tc reliability in patients not undergoing surgery. Results. (Group A) Concordance between CT, 99mTc, and intraoperative bacteriologic cultures was found. (Group B) CT scan was negative in all cases; 99m-Tc identified four cases of low-grade infection confirmed by bacteriologic cultures at elective surgery. Of the 14 patients not operated on, a graft-independent infection was confirmed in nine. No signs of AGI were identified during follow-up in the 14 patients not operated on. (Group C) All patients were operated on. CT scanning suggested AGI in five cases confirmed by intraoperative bacteriologic cultures. 99m-Tc showed one false positive, likely caused by platelet contamination. Conclusions. The 99m-Tc can [email protected] the absence of AGI. In patients with nonspecific symptoms 99m-Tc can


(Iv = 83)

Angiography < 70% 3 76

detect a low-grade surgery.

70%-100% 20 2

i 70% 12 49

AGI and permits its treatment in elective

Determinants of successful peroneal artery bypass Edward Plecha, MD, Gary R. Seabrook, MD, Dennis F. Bandyk, MD, and Jonathan B. Towne, MD, Medical College of Wisconsin, Milwaukee, Wis. Bypass to the peroneal artery has been considered the most challenging of the infrageniculate reconstructions that has led some surgeons to prefer pedal bypass when available. This report evaluates the influence of venous conduit modification, patency of the dorsal pedal or posterior tibia1 artery at the ankle, and the quality of the peroneal collaterals and pedal arch. From 1981 to 1989,73 consecutive peroneal artery bypasses were performed for critical limb ischemia (n = 71) or popliteal aneurysm (n = 2). The average follow-up was 34 months (range, 1 to 92 months). By life-table analysis, the S-year primary (first degree) and secondary (second degree) patency rates were 59% and 92%, respectively. The vein graft was modified at the original procedure in 24 patients (30%) for iatrogenic injury (n = 7), inadequate length (n = 9), or a sclerotic segment (n = 8). This resulted in a reduced 4-year 1 degree patency of 36%, compared with 76% for unmodified conduits (p < 0.001). Of the five grafts that eventually failed, four venous conduits had been modified at the initial operation (p < 0.03). Operative results improved with surgeon experience. The 5-year 2 degree patency of grafts placed before 1985 was 80% compared with 97% for subsequent grafts (p < 0.02). The original postoperative mean ankle-brachial index (ABI) for grafts revised for hemodynamic failure or thrombosis (n = 21) was 0.84 compared with 0.96 (e = 52) in grafts that did not require revision (p < 0.02). The mean ABI of the revised grafts increased to 1 .O after the revision (p < 0.03). This suggests that the vein graft may have harbored a sclerotic segment, residual competent valve, or arteriovenous fistula from the initial operation and did not develop a de novo lesion in the follow-up period. To compare the value of peroneal artery bypass with pedal bypass, the patency rates were calculated in patients with a graftable inframalleolar artery in whom a pedal bypass was an alternative. The presence of a patent dorsal pedal or posterior tibial artery at the ankle and the status of the pedal arch did not significantly influence first-degree or second degree patency. The 5-year second degree graft patency for patients with a patent dorsal pedal or posterior tibial artery at the ankle was 86%. The peroneal artery should be



selected for outflow when it is the single tibial runoff vessel and is preferable to a bypass to an inframalleolar arterial segment. The quality of venous conduit and the technical skill of the surgeon are the two most important factors in the success of bypasses to the peroneal artery. A twelve-year follow-up of axillofemoral grafts Sherif El-Massry, MD, Ehab Saad, MD, Lester R. Sauvage, MD, Michael Zammit, MD, Christopher C. Davis, MD, James C. Smith, MD, and Edward A. Rittenhouse, MD, The Hope Heart Institute, Seattle, Wash. Seventy-nine consecutive axillofemoral bypass graft operations were performed on 77 patients from January 1978 to April 1990, with use of 8 mm, externally supported knitted Dacron grafts. The mortality rate within 30 days of operation was 5%. In this 12-year period (mean, 42 months), only three patients were lost to follow-up. Assessment was made by regular clinical and Doppler examinations; in most instances, follow-up visits occurred every 6 months. The primary patency rate was 78% at 5 years and 73% at 7 years, with no change thereafter. Neither the graft configuration, that is, axillounifemoral (n = 50) versus axillobifemoral (n = 29), nor the patency of the superficial femoral artery were found to have an impact on the primary patency rate. Patients who underwent surgery for claudication (n = 30 grafts) had a primary patency rate of 80% at 6 years compared with 65% for those who required surgery for limb salvage (n = 49 grafts). On the basis of these findings, axillofemoral bypass grafts may be used appropriately for severe, high-risk patients with severe claudication with advanced aortoiliac disease considered not amenable to aortofemoral bypass or to angioplasty. Refining the indications for arteriography in penetrating extremity trauma: a prospective analysis Michael R. Schwartz, MD, Fred A. Weaver, MD, Albert E. Yellin, MD, and Madeline Bauer, PhD, LAC @ USC Medical Center, Los Angeles, Calif: The indications for diagnostic arteriography in penetrating extremity trauma remain controversial. To investigate this issue, 509 consecutive patients over a 12-month period with an isolated upper or lower extremity penetrating injury were entered into a prospective study designed to refine the indications for diagnostic arteriography. Twenty-three (4%) patients with obvious limb-threatening ischemia, which required immediate operation, and 19 (3%) who refused arteriography were excluded from subsequent analyses. The remaining 467 patients were stratified into high, intermediate, or low risk groups on the basis of criteria previously developed and reported. The low risk group comprised 208 (41%) patients and included all patients without hard or soft physical findings of arterial injury and an ankle- or arm-brachial index (ABI) 2 1.00. These patients were observed for 24 hours, discharged, and followed as outpatients. No patient in this group developed


complications of a missed injury. The intermediate risk group included 153 (30%) patients who had one or more of the following findings: fracture, bruit, hematoma, decreased capillary refill, hemorrhage or hypotension, major soft tissue injury, or an ABI of < 1.00. The high risk group included 106 (21%) patients who had either a pulse deficit, neurologic deficit, or shotgun injury. Arteriography was performed in all patients in the high and intermediate risk groups. Seventy-nine (31%) injuries were identified, 62 minor and 17 major (limb threatening). Twelve major injuries required operative repair or arteriographic embolization. The remaining five injuries were observed without sequelae. A logistic regression analysis of high and intermediate risk patients was performed to determine potential predictors’of arterial injury. Only pulse deficit (p = 0.01) and/or an ABI < 1.00 (p = 0.03) were found to be significant predictors for arterial injury. The presence of either of these two variables successfully predicted 100% of the major artery injuries. This prospective study definitively demonstrates that risk stratification of patients as described above, with arteriography reserved for only those patients who have either a pulse deficit or ABI < 1 .OO will successfully predict all significant arterial injuries. The adoption of this simple algorithm eliminates unnecessary arteriography in the patient with penetrating extremity injury. The natural history of femoral pseudoaneurysms and AV fistiae induced by arterial puncture K. Craig Kent, MD, Cohn McArdle, MD, Bernice Kennedy, RDMS, Donald Bairn, MD, Elaine Anninos, RN, and John J. Skillman, MD, Beth Israel Hospital, Boston, Mass. We prospectively followed 22 consecutive patients who were found immediately after femoral artery puncture to have either a false aneurysm (FA) (16) or an arteriovenous fistula (AVF) (6). After an initial duplex scan, all patients were followed with serial scans, either in hospital or weekly as outpatients, depending on the stability of the process. Operative repair was performed for the following indications: (1) a greater than 100% increase in size of a FA by duplex scan; (2) the development of symptoms; (3) continued patency of the lesion after 2 months of observation. Nine of 16 FA and four of six AVFs spontaneously closed, Size as a predictor of necessity of FA repair approached statistical significance if 6 cm3 (1.8 cm in diameter) was used as the transition point (Fischer’s Exact Testp = 0.06). Size was not an absolute predictor as two small aneurysms (1.6 and 0.7 cm3) remained patent (although both had been followed safely as outpatients), and two large aneurysms (13.2 and 10.7 cm”) spontaneously thrombosed. Anticoagulation, length of FA neck, velocity in the FA cavity, size of original arterial puncture, or velocity in AVE did not correlate with thrombosis. We conclude (1) all FA do not spontaneously thrombose; (2) some FA remain potentially dangerous lesions (> 100%

Volume 15 Number 6 June 1992


expansion in four of the lesions eventually requiring repair); (3) repair of FA greater than 6 cm3 will result in a safe and cost-effective approach, allowing either immediate repair or d.ischarge of patients after their initial diagnosis; (4) AVF should be observed for a minimum of 2 months before considering repair with the expectation that the majority will spontaneously thrombose. Innominate artery trauma - a thirty-year experience Robert H. Johnston, Jr., MD, Matthew J. Wall, Jr., MD, and Kenneth L. Mattox, MD, Baylor College of Medicine, Houston, Texas Injury to the innominate artery may represent a zone I cervical, thoracic outlet, or intrathoracic vascular injury and poses diagnostic exposure and management problems for general, vascular, and thoracic surgeons. This complex injury becomes a new learning experience with each infrequent encounter. Between 1960 and 1990 35 patients with innominate artery injuries were managed. Penetrating injuries were from gunshot wounds in 17, stab wounds in 6, and shotgun wounds in 1 patient. Blunt injury accounted for seven of the patients. In nine patients there were multiple injuries, including multiple vascular injuries. Median sternotomy was used in all planned emergency operations, and bilateral transternal anterolateral thoracotomy was used in patients undergoing urgent emergency center thoracotomy. Blunt injury resulted in tears near the aortic arch, with intimal dissection. Bypass grafting without hypothermia, shunts, or systemic heparinization was used in all patients. Twenty-six patients survived to leave the hospital with no complications related to the procedure. Postoperative neurologic complications were associated with preexisting neurologic deficits, and gastrointestinal complications were associated with systemic infections. Patients with innominate artery injury and stable vital signs can usually be transported without difficulty and managed without complex perioperative adjuncts. These patients can be revascularized by use of simple vascular surgical techniques. These patients should expect an uncomplicated postoperative course unless there has been associated central nervous system injury or other injuries leading to systemic infection. Synthetic conduits have been used with success and have not required systemic heparinization or complex temporary shunting. Carotid endarterectomy with reconstruction niques tailored to operative findings Joseph P. Archie, Jr., MD, Carolina Cardiavascular Associates, Raleigh, N. C.


Reconstruction techniques tailored to operative findings were applied to 400 consecutive carotid endarterectomies (CEAs) between 1987 to 1991. There were 392 primary arid 8 redo CEAs on 351 patients (199 men, 152 women) age 67 f 9 years. Indications were transient ischemic attack in 111 (28%), amaurosis fugax in 52


(13%), reversible ischemic neurologic deficit in 10 (2%), stroke in 56 (14%), global ischemia in 46 (12%), and asymptomatic > 75% diameter stenosis in 125 (31%). A shunt was used selectively in 42 (10%) CEAs. The internal carotid arteriotomy was routinely carried 5 to 7 mm distal to the plaque. Distal end point feathering was complete in 371 (94%) of CEAs. Patch reconstruction was performed in 366 (91%) CEAs in which the arteriotomy extended distal to the internal carotid bulb. Of these, 326 (81%) had saphenous vein patch, 25 (6%) polytetrafluoroethylene patch, and 16 (4%) Dacron patch reconstruction. A saphenous vein patch (thigh vein only in females) was used when available, and the distended vein diameter was > 3.5 mm. Before patching, redundant endarterectomized internal carotid segments were shortened by transverse plication to prevent kinking in 13 (3%). A saphenous vein interposition graft was used in the seven (2%) CEAs in which the endarterectomized internal carotid had severe adventitial disease. The other 27 (7%) CEAs had an arteriotomy that did not extend distal to the bulb and were primarily closed. There were two (0.5%) hospital deaths, both from myocardial infarction and two (0.5%) strokes, one with a permanent deficit, for a hospital mortality and permanent stroke rate of 0.75%. Five patients had a hyperperf&ion syndrome with seizures. One occurred in the hospital without complications and the other four after discharge, of which three had a stroke and two died. All five had > 90% carotid stenosis and a low stump-back pressure. CT or MRI scan showed ipsilateral cerebral edema in four and hemorrhage in one. All hyperperfusion syndrome and stroke patients had normal carotid duplex scans. Two patients died after hospitalization ofmyocardial infarctions for a 30-day mortality rate (6,1.5%) and permanent stroke (2,0.5%) rate of 2%. Only one stroke was potentially due directly to intraoperative management. No patch ruptures or false aneurysms occurred. Postoperative duplex scans of 364 (91%) CEAs from 3 to 6 months showed no internal carotid occlusions, and three (1%) had greater than 50% diameter stenosis. These results support CEA reconstruction techniques based on operative findings. Recognition of patients at risk for and treatment of the hyperperfusion syndrome is a clinical challenge. Correlation of North American symptomatic carotid endarterectomy trial definition of 70-99% internal carotid stenosis with duplex criteria Gregory L. Moneta, MD, James M. Edwards, MD, Raymond, W. Lee, M,D, Lloyd M. Taylor, Jr., MD, Richard W. Chitwood, MD, and John M. Porter, MD, &eHon Health Sciences University, P&and, Ore. The North American Symptomatic Carotid Endarterectomy Trial (NASCET) recently showed benefit for carotid endarterectomy (CEA) for symptomatic 70% to 99% internal carotid (ICA) stenosis. In the trial, ICA stenosis was classified angiographically: percent ICA stenosis = 1 - (narrowest ICA diameter/diameter normal distal ICA) x 100%. However, widely used duplex scan







et al.)

DS vatiable


PSV 2265 cm/set PSV 2 370 cm/set EDV 2 100 cm/set EDV 2 185 cmjsec PSV/CCA PSV 2 4.2 PSV/CCA PSV t 7.5



90% 67% 85% 28% 90% 40%

73% 89% 71% 84% 77% 89%

(DS) criteria for ICA stenosis correlate with different angiographic categories (50% to 79%, > SO%), and are based on estimated bulb diameter. Given NASCET results, it is clearly important to determine the accuracy with which DS can detect a 70% to 99% ICA stenosis as calculated above. We therefore blindly evaluated with separate observers carotid angiograms from 100 randomly selected patients who also underwent carotid DS in our laboratory over 4 years. Angiographic ICA stenosis was calculated as above. Duplex scanning was assessed for ICA peak systolic velocity (PSV) , ICA end diastolic velocity (EDV) , and the ratio of ICA PSV to common carotid (CCA) PSV. Receiver operator curves (ROC) were constructed for sensitivities, specificities, positive (PPV) and negative (NPV) predictive values of DS variables to identify a 70% to 99% ICA stenosis with specific attention to obtaining high sensitivities and PPVs ( 2 85% to 90%). One hundred eighty-four patent ICAs were present in the 100 randomly selected patients with carotid DS and carotid angiography. 58/184 patent ICAs (32%) had a 70% to 99% angiographic stenosis. The table shows ROC evaluation. We conclude an ICA PSV/CCA PSV ratio of 4.2 offers the best combination of sensitivity and PPV for determining a 70% to 99% ICA stenosis as calculated in the NASCET trial. It is clear DS determination of 70% to 99% stenosis as defined in the NASCET study requires adoption of significantly modiied DS criteria from that in current use in most vascular laboratories. The natural history of asymptomatic carotid artery disease Richard W. Bock, MD, Anthony C. Gray-Weale, MB, Philip A. Mock, PhD, David A. Robinson, MB, and Robert J. Lusby, MD, Untie&y of Louisville, Louisville, ICy. Sponsored by: J. David Richardson, M.D., Louisville, Ky. The natural history of asymptomatic carotid artery disease remains poorly understood despite the widespread acceptance of duplex ultrasonography and its ideal suitability for accurate, serial examination of asymptomatic patients. We report a prospective study of 242 asymptomatic, unoperated patients followed by annual duplex scans, using both standard duplex spectral criteria as well as plaque morphology analysis. During follow-up, which ranged to 72 months (mean, 27.4), 15 ischemic strokes (6.2%), and 20 transient ischemic attacks (TIAs) (8.3%) developed in


85% 96% 84% 98% 87% 98%


95% 86% 92% 75% 95% 78%

34 patients. Annual stroke, TIA, and overall event rates were 2.7%, 3.6%, and 6.2%, respectively. Although patients with 80% to 99% lesions had a 20.6% annual event rate, most events occurred contralateral to these lesions, and reanalysis by vessel rather than by patient showed a 5.1% annual vessel event rate for this level of disease. Only one of 15 strokes occurred downstream of an 80% to 99% stenosis. Overall, vessels ipsilateral to TIA and stroke were less diseased than their contralateral counterparts. Plaque echolucency on B-mode ultrasonography correlated with TIA and stroke (5.7% annual vessel event rate versus 2.4% for echodense plaques,? = 0.03). Disease progression was highly correlated with TIA and stroke @ < O.OOOl), but it usually occurred in association with rather than before ischemic events, thus proving more useful in explaining pathogenesis than in predicting future events. There was no association between aspirin use and TIA rate, but patients on aspirin did have a threefold increase in annual stroke rate (1.6% versus 4.8%, p = 0.027). This study, while confirming a high risk for asymptomatic patients with critical stenosis and/or echolucent plaque, demonstrates the importance of contralateral disease and the absence of orderly progression from minimal disease through high-grade stenosis to symptomatic cerebral ischemia. Transient ischemic attack and stroke occur in association with abrupt, unpredictable, quantum changes in carotid artery disease. Nonoperative treatment of superficial femoral artery disease: long-term follow-up Geoffrey S. Cox, MD, Norman R. Hertzer, MD, Jess R. Young, MD, Patrick J. O’Hara, MD, Leonard P. Krajewski, MD, and Edwin G. Beven, MD, The Cleveland Clinic Foundation, Cleveland, Oh Between 1977 and 1991,405 patients with atherosclerotic occlusive disease of the superficial femoral artery (SFA) underwent clinical as well as noninvasive laboratory evaluation and were recommended nonoperative treatment. Limbs with coexisting aortoiliac occlusive disease, aneurysmal degeneration, or previous femoropopliteal bypass were excluded, leaving 568 involved extremities. Complete follow-up was available in 377 patients (93%) with 527 limbs (93%) and forms the basis of this study. As patients were followed with serial noninvasive testing, changes in the ankle-brachial index (ABI) and waveforms on pulse volume recordings (PVRs) were available for

Volume 15 Number 6 June 1992

comparison with the symptomatic changes reported by the patients. Patients were followed for a minimum period of 2 years (range, 24 to 158 months; median, 66 months), returning annually or every other year for review. Those patients who returned beyond this time interval were thought to be no longer undergoing routine surveillance and were excluded from this report. There were 25 1 men and 126 women, with a mean age of 63.9 years (range, 37 to 91 years). Three hundred forty-one patients (88%) had a history of tobacco use, and 36% continued to smoke during the period of surveillance. Hypertension was present in 300 patients (SO%), whereas 115 patients (31%) had diabetes mellitus. Coexistent vascular disease was common. Nearly 50% of patients had significant coronary artery disease (CAD), 29% had cerebrovascular disease, and 2% had renal failure requiring hemodialysis. During the surveillance period, 45 limbs (8.5%) in 42 patients (11.1%) received arterial intervention. This entailed surgery in 39 cases and endovascular treatment in 6. By means of life-table analysis, the cumulative risk for intervention with 5 years was 5.2%. A total of 14 limbs (2.7%) in 14 patients (3.7%) ultimately required major amputation either after failed bypass or as a primary procedure. Analysis of risk factors revealed that only diabetes mellitus (p = 0.0011) or multilevel disease (p = 0.0009) was associated with an increased risk for both intervention and amputation. Other factors, including age, gender, current or prior smoking history, hypertension, or the presence of CAD or cerebrovascular disease, failed to reach statistical significance. These data support the continued conservative approach to patients with SFA occlusive disease without limb-threatening symptoms. Patients with multilevel disease or diabetes mellitus are at increased risk and should be followed closely. Risks and benefits of femoral-popliteal percutaneous balloon angioplasty Maria GM. Hunink, MD, Michael F. Meyerovitz, MD, Magruder C. Donaldson, MD, Krishna Kandarpa, MD, Joseph F. Polak, MD, Anthony D. Whittemore, MD, Clement J. Grassi, MD, John Army, MD, and John A. Mannick, MD, Brlghanz C? Women’s Hospital, Boston, Mass. The risks and long-term efficacy of femoral-popliteal balloon angioplasty (PTA) remain incompletely understood, and its role is controversial. From 1980 to 1991 angioplasty was performed on 170 femoral-popliteal lesions during 127 procedures in 106 patients. Sixty-six patients underwent a single angioplasty, and angioplasty was performed on 40 multiple sites during one or more procedures. Five lesions were recurrences after previous angioplasty. Forty-three percent of the angioplasties were performed for critical ischemia and 57% for claudication. Twenty-four common femoral, 116 superficial femoral, and 30 popliteal artery lesions were treated. Ninety-one percent of the treated lesions were stenoses and



Five-year cumulative patency (Hunink et al.) NO.


Claudication Critical &hernia Stenosis Occlusion Good runoff Poor runoff

170 97 73 155 15 92 78

5-y patency


47% 59% 25% 48% 38% 60% 24%

5% 6% 8% 5% 13% 6% 9%

9% were occlusions. In 55% of cases runoff to the distal vessels was good (2 or 3 vessels), and in 45% it was poor (0 or 1 vessel). Primary patency was defined as maintenance of hemodynamic and symptomatic improvement after angioplasty without further intervention. The mean follow-up time was 2 years. Life-table analysis was conducted on all sites subjected to angioplasty. Major systemic morbidity occurred in 1.4% of procedures performed for claudication and in 10.9% of those performed for critical ischemia. Local morbidity including technical failure occurred in 7.1% of sites undergoing angioplasty. Thirty-day procedural mortality occurred in 0.8%. The 5-year cumulative primary patency is shown in the table. The results suggest that femoral-popliteal angioplasty is a low-risk procedure with acceptable long-term results except in patients with critical ischemia who have an occlusion and/or poor runoff. The etiology and anatomic distribution of lesions causing reversed vein graft failure: A five-year prospective study Joseph L. Mills, MD, Roy M. Fujitani, MD., and Spence M. Taylor, MD, Wilfoord Hall USAF Medical Cents, LAFB, Texas. Sponsored by: John M. Potier, MD, Portland, @e. Previous reports of autogenous vein graft failure focused primarily on the origin of early ( < 30 days) and late graft failure. Little data are available concerning vein graft failure in the intermediate postoperative period (3 to 18 months). To delineate the incidence, cause, and anatomic distribution of lesions causing graft failure in this critical interval, 222 consecutive infrainguinal reversed vein grafts (IRVGs) performed at a single institution from July 1986 to November 1991 were prospectively entered into a duplex scan surveillance protocol, the details of which have been previously reported. Duplex surveillance with arteriographic confirmation identified 28 patent, hemodynamically Failing IRVGs during a mean follow-up of 21 months (range, 1 to 60 months). An additional 17 grafts thrombosed before detection of any underlying abnormality; thrombolytic therapy and/or reoperation uncovered the origin of occlusion in 10 of these grafts. The cause of graft failure (failing as well as failed) was therefore clear in 38 of 45 (84.4%) instances. The causes of failure were the following: intrinsic graft



stenosis (n = 24; 53.3%); inflow failure (n = 6; 13.5%); outflow failure (n = 4; 8.9%); muscle entrapment (n = 2; 4.4%); and hypercoagulable state (n = 2; 4.4%). The most common intrinsic graft lesion was focal intimal hyperplasia (17 lesions/l6 grafts) in the juxtaanastomotic position, occurring solely in the vein graft itself, with equal frequency, immediately distal to the proximal, or proximal to the distal anastomosis. Only rarely (a = 3) did this involve the juxtaanastomotic artery. Focal midgraft valvular stenoses (n = 4) and diffuse myointimal hyperplasia (n = 4) were also detected. The peak incidence of graft failure was 4 to 12 months after operation (65% within 12, 77% within 18 months). We conclude that duplex surveillance of IRVGs is warranted by the 21% incidence of potentially remediable graft failure. A significant portion of these failures occur during the intermediate postoperative period (3 to 18 months), most commonly a result of focal intrinsic vein graft lesions. With reversed vein conduits, these lesions arise predominantly in the vein graft itself, in the juxtaanastomotic position. Biology of long-term autogenous vein grafts- a dynamic evolution Thomas Reifsnyder, MD, Jonathan B. Towne, MD, Gary R. Seabrook, MD, Jean F. Blair, MD, and Dennis Bandyk, MD, Medical College of Wisconsin, Milwaulzee, Wis. The venous conduit as a living arterial substitute has a dynamic biology that affects its durability. This study was conducted to document the evolution of lesions which may threaten these grafts over their lifetime. Seventy lower extremity vein grafts (in situ [n = 601, reversed saphenous [n = 81, and cephalic vein [n = 21) functioning at 4.5 to 21.6 years (median, 6.0 years) were evaluated by color duplex ultrasonography to determine the morphologic characteristics of the long-term graft. The entire graft was imaged and than classified by segments as normal (grade I), abnormal but not graft-threatening (grade II), or abnormal and graft-threatening (grade III) for the proximal, middle, and distal thirds. Thirty-three (47%) grafts were normal, and 37 (53%) harbored 54 grade II and 13 grade III abnormal segments. Nineteen grafts had 40 regions of nonstenotic wall irregularity or plaque, IO had 11 discrete stenoses, and 8 conduits had 16 segments with aneurysmal degeneration. Nonstenotic wall plaques (20 proximal, 12 mid, 8 distal) and aneurysms (5 proximal, 7 mid, 4 distal) occurred throughout the graft, but discrete stenoses occurred at the proximal and distal ends (10 of 11) and were more severe and potentially graft threatening, with 9 of 11 having diameter reduction > 50% (grade III). The five grafts that required thrombectomy in the distant past (mean, 38 months) eventually developed aneurysmal segments, suggesting that thrombosis predisposes to late aneurysm formation. There were 66 revisions of 37 (53%) grafts. Twelve (18%) were performed within the first 30 days, 23 (35%) between 1 and 24 months, and 31(47%) after 2 years. The average postoperative interval of grafts revised after 2 years


was 5.4 years (range, 2.1 to 17.5 years) and corrected inflow in nine, outflow in nine, anastomotic or midgraft stenosis in nine, anastomotic pseudoaneurysm in two, and nongraft related thrombosis in two. Grafts with at least one abnormal segment by duplex imaging (s = 37) were more likely to have undergone postoperative revision than those found to be normal @ < 0.05). No grafts that were normal at the time of duplex imaging needed revision of the vein conduit after the first year, although four had seven revisions for disease progression in the inflow or outflow arteries. The autogenous vein remains the most durable arterial conduit, but vigilant surveillance is essential. The atherosclerotic environment continually produces lesions that may imperil the longevity of the graft. Regional adenosine prevents postischemic spinal cord injury P. Scott Seibel, MD, Curtis G. Tribble, MD, Irving I. Kron, MD, and Pierre Theodore, MD, University of [email protected] Health Sciences Center, Charlottesville, Va. Operations on the thoracic aorta requiring periods of cross-clamping are associated with postoperative paraplegia ranging from 3% to 40%. Attempts to reduce this complication have focused on decreasing the energy needs of the spinal cord with adjuncts such as hypothermia. The present study explores the hypothesis that spinal cord substrate enhancement with adenosine will prevent paraplegia. New Zealand white rabbits underwent 40 minutes of aortic occlusion. Control rabbits sustained aortic occlusion without any protective measures. The remaining .animals were randomized into three groups which underwent aortic cross-clamp with various protective adjuncts. Group I received regional aortic hypothermic perfusion, group II received systemic adenosine, and group III received regional aortic perfusion with high dose adenosine. Neurologic function was graded according to the Tarlov scale (0 = no movement, 1 = slight movement, 2 = sits with assistance, 3 = sits alone, 4 = weak hop, 5 = normal hop). After 96 hours animals were killed and spinal cords were harvested for histologic examination. Results are summarized in the table. The nonparaplegic animals in group I had mean Tarlov scores of 1.5, while those in group III had mean scores of 3.4 (p < 0.05). Histologic studies on spinal cord tissue could reveal no significant differences between the study groups. Hypothermic regional permsion provided’ some cord protection due to a slowing of cellular metabolism. Although the exact cellular mechanisms are still unclear this data demonstrates that substrate enhancement with regional adenosine prevents paraplegia associated with aortic occlusion in this experimental model. Cigarette smoke alters chylomicron metabolism in rats Xian Mang Pan, MD, Ilona Staprans, PhD, and Joseph H. Rapp, MD, VA Medical Centev, Saw Francisco, Cal&-f It has been proposed that smoking delays the plasma clearance of dietary fat and cholesterol allowing more time

Volume 15 Number 6 June 1992


Results of study (Seibel et al.) Paraplgic Control Group I Group II Group III




6 10

4 0


10 (p < 0.05, ANOVA)

for their interaction with the artery wall. All dietary fat and cholesterol enter the circulation as chylomicrons (CM). CM attach to vascular endothelial surfaces where they are hydrolyzed by lipoprotein lipase (LPL) to cholesterol-rich CM remnants that are rapidly cleared by the liver. These remnants are known to be atherogenic in both humans and animals. We examined the effects of smoking on CM metabolism by exposing Sprague-Dawley rats to an atmosphere of 10% cigarette smoke for 10 minutes, 4 times per day for 10 days. In addition, smoke-treated CM were prepared by passing cigarette smoke through CM containing lymph at constant negative pressure for 10 minutes. Mesenteric lymph was collected from rats gavage fed corn oil and milk (1: 2) containing 14C-cholesterol and 3H-triolein. CM were concentrated by centrifugation at 25,000 rpm for 1 hour. Control CM (CCM) were then used either without further alteration or exposed to smoke (SCM) as noted above. Ten minutes after the injection of CCM, the livers and hearts were removed and examined for label. We found that in smoke-treated rats there was a slower clearance of CCM cholesterol to the liver, 47.77% 2 1.45% in control rats and only 40.85% + 2.29% in smoke-treated animals @ < 0.01). In control animals 0.69% ? 0.2% of the cholesterol label remained in the heart tissue, presumably in CM bound to LPL on the endothelial surface. This increased to 0.96% t 0.16% in the smoke-treated rats. When SCM were injected into normal (non-smoke exposed) rats, their hepatic uptake of the labeled cholesterol was 29.45% * 4.97% compared with 43.32% * 6.06% for CCM (p < 0.05). The endothelial binding in the normal rat hearts was increased from 1.48% t 0.29% (CCM) to 2.51% ? 0.77% with the infusion of SCM (p < 0.05). The CCM and SCM recovered in the liver were examined to determine the degree of triglyceride (TG) hydrolysis. We found that 62.79% + % of TG were hydrolyzed in control CM, but only 17.73 rfr % of TG were hydrolyzed in smoke-treated CM (p = 0.001). In conclusion, these data demonstrate for the first time that exposure to cigarette smoke delays CM clearance in rats and reduces the rate of TG hydrolysis in smoked CM. There appears to be separate and additive effects of exposing the animal and the CM particle to smoke. These findings are consistent with a reduction in LPL activity both by a direct effect on this endothelial bound enzyme in smoke exposed animals and by rendering the CM a poor substrate for LPL in SCM. Delayed clearance of the relatively cholesterol-rich CM remnants could create a more atherogenic postprandial milieu,


Distribution of the C-MYC oncoprotein in normal and arteriosclerotic human carotid arteries Michael L. Marin, MD, Frank J. Veith, MD, Ronald E. Gordon, PhD, Natalie Tulchin, PhD, and Thomas F. Panetta, MD, Monte&ye Medical Cente$Albert Einstein College of Medicine, New York, N.Y. Smooth muscle cell (SMC) proliferation in arteriosclerosis is widely regarded as a reactive process in response to vessel wall injury or inflammation. Pa fundamentally different pathogenesis of arteriosclerosis is described by the monoclonal hypothesis that proposes that proliferating SMCs derive from a minority of transformed SMCs. The presence of oncogenes, which play an important role in the control of cell growth and differentiation, represents a possible underlying basis for such proliferation. We have compared the distribution of the c-myc oncogene product (c-myc oncoprotein) in normal and arteriosclerotic human carotid arteries to determine its cellular and tissue localization. Samples of six carotid plaques from six patients were rapidly frozen in liquid nitrogen at the time of carotid endarterectomy. Nondiseased human carotid arteries (3) were similarly prepared at the time of organ harvest from brain dead donors. Frozen sections were labeled with a polyclonal rabbit anti-c-myc antibody, which recognizes a 64KD c-myc human protein. C-myc positive cells were identified by peroxidase activity within nuclei of medial smooth muscle cells. One hundred cells were analyzed from each specimen, and the percentages of positive cells in each group were calculated. Normal human carotid artery demonstrated minimal staining (mean number of positive cells = 14.7 rfr 3.5/100 cells) for the c-myc oncoprotein, with positive cells located primarily in the immediate subendothelial media. In comparison, carotid plaques demonstratI:d a marked increase in staining for c-myc oncoprotein (mean number of positive cells = 89.8 + 4.0/100 cells,p < 0.001). These findings support a monoclonal hypothesis in which a small population of activated SMCs are involved in the cell proliferative response seen in arteriosclerosis. The underlying mechanism(s) that initiates or controls this process remains to be determined. Endothelin production by hypoxic human endothelium Jonathan P. Gertler, MD, Jonathan P. Gertler, MD, and Victor H. Ocasio, BS, Massachusetts General Hospital, Boston, Mass., and State University N.Y. - Health Science Center, Brooklyn, N. T. The physiologic significance of endothelin remains incompletely defined. Procoagulant and antifibrinolytic activity is increased in hypoxic human endothelium (HUVEC). We examined the effect of hypoxia on HUVEC endothelin production in vitro to determine if a correlation existed. Methods. Cultured HUVEC were rendered hypoxic (Paz = 40 mm Hg) or normoxic (PO, = 120 mm Hg) for 24 hours. Media were either standarcl, high glucose (5 gm/L), (High Glu), or contained 500 units/ml super-




of study results (Gertier, Gertler, and Ocasio)

Summary croup


4 4 4 4 3 3









0 24 0 24 0 24

in tension


standard standard High Glu High Glu SOD SOD


( i: SEM)


for drug

9% 39% 18% dose effect,

740 1308 804 3244 664 2168





oxide dismutase (SOD). Endothelin-like immunoreactivity (ET-IR) in conditioned media was measured by radioimmunoassay and expressed as mean fmol/ml ? SD per 100,000 cells. Viability was assessed by trypan blue exclusion. Significance was determined by Student’s t test shown on the table. All groups retained >95% viability assessed by trypan blue. Conclusion. Twenty-four hours of hypoxia causes an increase in conditioned media Endothelin-like immunoreactivity in human endothelium. Neither increasing glucose concentration in the media to allow for ongoing substrate availability despite hypoxia, nor free radical inhibition blunted the response. This suggests a signal for hypoxic induction of endothelin production or secretion that is unrelated to cell toxicity. Differences in vasoreactivity of in situ and reversed vein grafts Joseph G. Magnant, MD, Daniel B. Walsh, MD, Joseph R. Schneider, MD, Timothy W. James, MD, Robert J. Wagner, BS, and Jack L. Cronenwett, MD, Da-z&Hitchcock Medical Center, Lebanon, N.H. Vein graft vasoreactivity offers a functional assessment of the interaction and integrity of endothelial and smooth muscle cells. We compared responses of in situ vein grafts (ISVG) and reversed vein grafts (RVG) to endotheliumdependent and -independent vasoactive mediators and correlated these responses with a quantitative analysis of graft wall morphology. Twenty-three New Zealand white rabbits underwent unilateral replacement of the carotid artery with 3 cm segments of autologous ipsilateral external jugular vein performed either in situ (n = 12) or reversed (n = 11). After 4 weeks, isometric tension responses were studied by use of midgraft circumferential vein rings from 12 rabbits (ISVG, 6; RVG, 6). Control, normal vein (CV), ISVG, and RVG rings were first exposed to increasing concen-

80 188 30 1336 112 736

0.004 0.027 0.025


5% iz 2.4 26% i- 3.2*# 13% k 1.3t *ISVG

-c f 2 -t -+ 2

p vs control

et al.)


f 1.2 2 7.5*?? 2 2.2t p < 0.01:




12% 4% 0.1%

-t 2.8 i- l.OX# k 3.067



SNP 18% 30% 10%

t 6.4 t 1.4# k 1.2t


trations of norepinephrine (NE, lo-* - IOm3 mol/L) or serotonin (5E-IT, lo- lo - 10m4 mol/L). These contracted rings were then treated sequentially with acetylcholine (ACh,lO-lo - 10e4 mol/L), and sodium nitroprusside (SNP, 10m9 - 10m5 mol/L). Vasoreactivity was measured (grams) and expressed as percent change from baseline tension after initial equilibration to 0.5 gm. Vein grafts from 11 perfusion fixed rabbits (ISVG, 6; RVG, 5) underwent morphometric analysis and scanning electron microscopy. There was no significant difference in luminal circumference, intimal or medial area, or intimal or medial thickness between ISVG and RVG. Scanning electron microscopy of ISVG and RVG revealed similarly normal endothelial cell appearance. Mean responses to the range of drug doses are shown in the table. After 4 weeks, rabbit ISVG are significantly more reactive to exogenous vasoconstrictors (NE and 5-m) than RVG or CV. Endothelium-dependent relaxation (ACh) is better preserved in ISVG than in RVG, but still less than CV. Endothelium-independent relaxation (SNP) occurs to a significantly greater degree in ISVG than RVG. Although the mechanism of this enhanced vasoreactivity of ISVG is not yet clear, the similar histologic appearance of RVG suggests functional rather than morphologic differences in endothelial and/or smooth muscle cells. Human saphenous veins grafted into arterial circulation demonstrate altered smooth muscle and endothelial responses Thomas C. Park, MD, Carla Bowman, BS, Christian T. Harker, PhD, James M. Edwards, MD, Lloyd M. Taylor, Jr., MD, and John M. Porter, MD, Oregon Health Sciences University, Potiland, Ore. Attempts have been made in the rabbit model to assess the effects of grafting venous segments into the arterial circulation on vascular smooth muscle and endothelial

Volume 15 Number 6 June 1992

function of the grafted segments. Such studies suggest a selective i.ncrease in the responsiveness of the grafted segments to serotonin (5-HT). The relevance of this potentially important observation to human saphenous vein grafting is unknown. We have obtained both human saphenous veins (HSV, n = 7) and human vein grafts (HVG,n = 8) from patients undergoing leg bypass grafts or graft revisions. The HVG had been placed with use of end-to-side anastomoses to the common femoral arteries proximally and to the popliteal or infrapopliteal arteries distally for periods ranging from 10 to 26 months before removal (mean, 18 months). There was no evidence of graft thrombosis in any segment of the vein used in these experiments. The vessels were immediately cut into rings and suspended in organ chambers for recording isometric contractions. The maximal tension elicited by both norepinephrine (NE) and. 5-HT was reduced (p < 0.05) in HVG in comparison to HSV [NE,, = 1.39 gm ? 0.29 (HVG) versus 4.59 gm ? 1.05 (HSV); 5-HT,, = 2.45 gm t 0.55 (HVG) versus 4.72 ? 1.11 (HSV)]. HVG was less responsive to NE than HSV (ED,, - 5.93 -C 0.09 and - 6.84 -C 0.24, respectively;p < 0.05). After precontraction with NE (ED& HSV but not HVG, demonstrated endothelium-dependent relaxation (EDR) to acetylcholine (ACh; maximum relaxation = 27.4% * 6.8%). In contrast to data from the rabbit model, our human data demonstrate vigorous responsiveness to both 5-HT and NE in the HSV as well as the HVG. EDR is lost in HVG. The relevance of rabbit vein graft data to human bypass graft appears questionable. Future studies of vein-artery graft contractility should focus on human tissue insofar as possible, as we may not be able to extrapolate animal results.

Heparin reduces tissue-type plasminogen activator gene expression in serum-stimulated human vascular smooth muscle cells Masahiko Yamaguchi, MD, Scott L. Diamond, PhD, David F. Cruess, PhD, Wendy E. Baur, MS, and John B. Sharefkin, MD, New E&mad Medical Center, Boston, Mass. Sponsored by: William C. Mackey, MD, Boston, Mass. Heparin inhibits both smooth muscle cell (SMC) proliferation and migration in vitro and postinjury intimal hyperplasia (IH) in vivo, but its mechanism of action is unknown. Since tissue plasminogen activator (tPA) acts to facilitate SMC migration by inducing early plasmin degradation of SMC matrix, we studied the effect of heparin on &‘A gene expression in SMCs. Cultured saphenous vein SMCs made quiescent by 72 hours of incubation in DMEM + 0.5% serum were stimulated with 10% serum 5 heparin (50 j&ml). Changes in tPA mRNA levels were measured after 4 hours by Northern blot as densitometric ratios of tPA signal to those for a control gene (glyceraldehyde 3 phosphate dehydrogenase) (GAPDH). In a second study tPA protein in conditioned



media after 24 hours was measured by ELISA as a function of response to variable heparin doses. The data suggest that heparin reduces IH at least partly via a dose dependent inhibition of SMC tPA gene expression to curtail initial SMC migration from media to intima. Efficacy and toxicity of differently charged poly cationic protamine-like peptides for heparin anticoagulation reversal Alphonse DeLucia III, MD, Thomas W. Wakefield, MD, Philip C. Andrews, PhD, Brad J. Nichol, BS, Amy M. KadeIl, Shirley K. Wrobleski, BS, and James C. Stanley, MD, University of Michigan Medical Center, Ann Arbor, Mich. It is hypothesized that protamine sulfate reversal of heparin anticoagulation relates to its total cationic charge. The aim of the present study was to examine the relationship of total cationic charge of synthetic protaminelike peptides to heparin reversal and adverse hemodynamic effects. Five protamine variants (lysine-containing peptides) were synthesized by FMOC procedures, having specific totalchargesof(+8),(+16),(+18),(+20),and(+21). These peptides and control salmine protamine (argininecontaining, +21 charge) were each studied in five dogs who received heparin 150 III/kg intravenously (IV) followed by either the synthesized peptide or control protamine, 1.5 mg/kg IV over 10 seconds. Coagulation studies obtained 3 minutes and 30 minutes after reversal included the following: activated clotting time (ACT), activated partial thromboplastin time (aPTI), thrombin clotting time (TCT), and factor Xa inhi.bition (FXa). Hemodynamic studies included the following: mean systemic arterial blood pressure (MAP), heart rate (HR), cardiac output (CO), and systemic oxygen consumption (VO,). Total toxicity scores (ITS) were developed for each peptide involving the above hemodynamic parameters. Anticoagulation reversal was more effective with peptides of greater cationic charge. For example, ACT reversal 3 minutes after peptide administration was: 6% ( + 8), 60% (+16), 81% (+18), 92% (+20), 81% (+21), and > 100% (protamine). Similar analysis for aPTT reversal at 3m was: 0% ( + 8), 61% (+ 16), 79% ( + 18), 91% (+20), 88% ( + 21), and 100% (protamine). Reversal of heparin anticoagulation was strongly correlated with total cationic charge (r value at 3 minutes, 30 minutes): ACT (0.97+, 0.99+), aPTI (0.99+, 0.78),TCT (0.84*, 0.85*), andFXa (0.87*, 0.75), *p I 0.05, ‘p .ZZ0.01. Maximum declines in MAP were more severe with peptides of greater charge: -1 mm Hg (t-S), -3 mm Hg (+16), -31 mm Hg (+18), -31 mm Hg (-!-20), -35 mm Hg (+21), and - 34 mm Hg (protamine). Maximum decreases in MAP, CO and $0, correlated @ 5 0.05) with total cationic charge (rvalue): MAP (0.87), CO (0.87), andV0, (0.86). ITS were greater with increasing charge: - 1.9 -C 1.1 (+X), -2.7 -r- 0.8 (+16), -6.6 ? 3.3 (+18), -6.1 t






and Lee)




Mean gge

266 260 226


Preop. AX

61yr 62yr 63yr



0.44 0.43 0.47



27% 31% 41%

occlusion 30.6% 25.0% 28.8%


p value NS NS NS

and Lee)

1 YV





76% 80% 79%

72% 74% 73%

62% 70% 65%

42% 69% 59%


3.5(+20), -6.9 * 3.8(+21), -7.0 * 5.2(protamine). TTS also correlated with total cationic charge (r = 0.89, p < 0.05). These data suggest for the first time that effective protamine-like alternatives to salmine protamine can be synthesized that will reverse heparin anticoagulation. Furthermore, total cationic charge appears to be an important determinant for both anticoagulation reversal and toxicity accompanying these protamine-like peptides.

From the Scientific Program of the Fortieth Aunual Meeting of the North American Chapter of the International Society for Cardiovascular Surgery Comparative evaluation of PTFE, HUV and saphenous vein bypasses in fern-pop AK vascular reconstruction Willard C. Johnson, MD, Kelvin K. Lee, PhD, and VA Cooperative Study Group Number 141, Baton VAMedical Center, Boston, Mass. Between June 1983 and June 1988,752 patients at 18 Medical Centers, who had superficial femoral artery occlusive disease, were randomized to receive either an externally supported PTFE, human umbilical vein (HUV), or saphenous vein (when available) bypass in femoral popliteal (above-knee) revascularization. Doppler-derived ankle brachial indexes (ABIs) were determined before operation and serially after operation. Patients were seen every 3 months for the first year and then every 6 months thereafter. All patients were instructed to take 650 mg of aspirin each day for the duration of the study. For primary patency evaluation, a graft was considered to be patent if the Doppler-derived postoperative ABI remained siginficantly improved (0.15 units above their preoperative value). Times to graft failure were compared with the Kaplan Meier survival procedure. The demographics and outcome are presented in the two tables.

p value

In summary, the overall results of this prospective randomized study suggest that the physical properties of a bypass graft are not a major factor in the prediction of graft patency in patients having femoral-popliteal above-knee arterial reconstruction. Femoropopliteal bypass with ipsilateral reversed saphenous vein when available and adequate and PTFE when not Joseph P. Archie, Jr., MD, Carolina Cardiwascular [email protected] Associates, Ralei&, N. C. The controversy continues over femoropopliteal bypass with a strict saphenous vein (SV) policy versus liberal use of polytetrafluoroethylene (PTFE) as is widely applied in general vascular practice. Between 1981 and 1991, 312 femoropopliteal bypasses were performed on 285 patients (158 men, 127 women) with ipsilateral reversed greater SV when available and adequate and with PTFE when not. Indications were daudication in 121, rest pain in 95, and nonhealing ulceration or gangrene in 96 legs. There were 235 SV bypasses (75%) ofwhich 157 were above knee and 78 below, and 77 PTFE (25%) bypasses, 58 above knee and 19 below. Of these, 232 SV (79%) and 62 PTFE (21%) bypasses were primary procedures. Infrapopliteal runoff was graded by the SVS/ISCVS unit scale. Runoff resistance was measured in 272 (87%) operations. In 17 (5.4%) operations the SV was exposed or harvested but not used because of small diameter, prior thrombosis or varicosities. End point graft status was confirmed by ultrasonography. Life-table exposure was 272 (87%) at 1, 151 (48%) at 2, 94 (30%) at 3, 59 (19%) at 4, and 41 (13%) at 5 years. Five SV grafts failed in the first 72 hours, two were revised and three replaced by PTFE. The 30-day mortality rate was 3% (7 of 285) and amputation rate 2% (6 of 312). Overall, 54 grafts failed (24 SV, 30 PTFE). The cumulative patency was 77% * 3% (M * 1SE) (85% to 69%, 95% confidence interval) at 3 years and 75% I- 3% (86% to 64%) at 5 years. Saphenous vein patency was superior to PTFE at 3 years, 87% 2 3% (97% to 77%) versus 50% 1?17% (62% to 38%) (p < 0.001) and at 5

Volume 15 Number 6 June 1992

years, 85% 2 3% (99% to 71%) versus 44% + 8% (59% to 29%) (p < 0.001). There was little difference in SV patency for the three indications (91% to 84% 3 years, 87% to 83% 5 years). However, PTFE had a 55% + 8% 1 year and 40% 2 9% 3-year patency when used for ulcer or gangrene. Above-knee primary SV bypass patency was slightly better than below knee at 3 years, 89% 2 4% versus 84% * 4%, and at 5 years, 87% * 4% versus 80% * 6%. Above-knee primary PTFE patency was 61% ? 8% at 3 years and 54% t 10% at 5 years (p < 0.01 when compared with SV). Below-knee PTFE patency was 32% * 12% at 3 years. Runoff units (U) and resistance (R) did not correlate with SV patency at 3 years, but PTFE had a a-year patency of 75% 2 12%, with 0 to 4 U runoff (p < 0.01 versus SV) and 41% to 59% with 5 to 10 U runoff. Regression for R versus U was R = 0.95 U + 0.11, mm Hg/ml/min,p < 0.001, butr = 0.26 and the M ? 1SD was 1.5 ? 1.2 for R and 5.4 -C 2.8 for U. These results strongly support SV for all femoropopliteal bypasses including primary above-knee bypasses with good runoff where the PTFE failure rate is 3 times that of SV. a mandate for total graft Vein graft rupture: replacement Kurt R. Wengerter, MD, Michael L. Schwartz, MD, Frank J. Veith, MD, William D. Suggs, MD, Thomas F. Panetta, MD, Clifford M. Sales, MD, and Michael L. Marin, MD, Montejiore Medical Center/Albert Einstein College of Medicine, Bronx, N.Y. Although rupture of carotid vein patch angioplasty and infected vein graft anastomoses are well described, rupture of the body of autologous vein grafts has not been adequately addressed. In a series of 952 lower extremity bypasses with vein, performed between 1981 and 1991,22 episodes of midgraft rupture of nonvaricose veins were identified in 16 patients. Sixteen episodes of initial rupture occurred 1 to 72 days after operation (mean, 23 days), and six episodes of rerupture occurred at 2 to 20 days (mean, 8 days) after initial repair. In five patients initial rupture occurred with no evidence of infection and negative cultures. The vein was noted to be dilated with a linear disruption in two cases, and microscopic calcification was present in the vein wall in three cases. In the remaining 11 patients, initial rupture was associated with positive bacterial cultures, although in only two of these cases was gross evidence of infection present. During this time period, another 35 vein grafts associated with infected wounds were managed with operative debridement and meticulous wound care, and the grafts were salvaged. Management of ruptured vein grafts consisted of the following: (1) local repair, including suture repair or local excision and interposition (7 cases), (2) excision of the involved segment (9 cases), or (3) subtotal graft replacement through a separate uninvolved tunnel (6 cases). Appropriate intravenous antibiotics were used in all cases. Treatment with local repair resulted in an 86% (6 of 7) rerupture rate, and a 14% limb salvage rate. Excision of the



graft without revascularization led to an 11% limb salvage rate. In patients treated with subtotal graft replacement routed through a separate tunnel, there was an 83% graft salvage rate and 50% limb salvage rate. Midgraft rupture of autologous vein is an ominous complication that may occur with or without bacterial contamination, although the former may not be evident at the time of rupture. Whether organisms are present or not, local repair or excision of the involved segment is doomed to failure. Total or subtotal excision with replacement via an uninvolved tunnel can result in maintenance of distal circulation and limb salvage. Crescendo TIAs: Samuel E. Wilson, Yatsu, MD, and Cooperative Study and VA Cooperative

a surgical imperative MD, Mark R. Mayberg, MD, Frank David G. Weiss, PhD and the VA 309 Trial&, UC Imilze Medical Center Study Group Nunabw 309> Ovan.e, CaliF

In a randomized, prospective, multicenter trial at 16 medical centers, 189 of 5000 patients screened with cerebrovascular disease were identified as having angiographic internal carotid artery stenosis greater than 50% corresponding to presenting symptoms of transient ischemit attacks (TIAs), transient monocular blindness (TMB), or recent small completed stroke. After one year there was a significant reduction in stroke or crescendo attacks in the 91 patients who received carotid endarterectomy (7.7%) compared with 98 nonsurgical patients (19.4%) (p = 0.011). This report examines the 19 nonsurgical patients who failed medical treatment, of whom 13 had crescendo TIAs, four had minor strokes, and three had major strokes. Crescendo TIAs were identified as recurrent transient cerebral or retinal ischemia characterized by an increased frequency, increased duration, or increased severity of events. The 13 of 98 (13.3%) symptomatic patients treated medically in whom crescendo TIAs developed had the following disabling characteristics: three or more episodes of TMB in 24 hours (six patients), prolonged retinal ischemia (two patients), loss of speech for more than 30 minutes (three patients), and more than three episodes of hemiparesis in 24 hours (four patients). One of the 13 patients had a new deficit on CT scan. The average time from randomization until the onset of crescendo TIAs was 2.5 months. Seven of 13 patients in whom crescendo TIAs developed had greater than 90% stenosis, one greater than 80% stenosis and four between 70% and 80% stenosis, with one of these four having contralateral occlusion. Patients in whom crescendo TIAs developed were offered carotid endarterectomy, and 12 of the 13 had an uncomplicated, urgent procedure. Complete thrombosis developed in the one patient who postponed surgery. On follow-up all 13 patients were free of symptoms at the study conclusion. Conclusion. Crescendo TIAs are disabling symptoms that occur in patients with high-grade carotid stenosis within 3 months of identification of symptomatic cerebrovascular disease. The 13% of patients on medical




Comparison of global reimbursement

in dollars per hour (Fillinger et al.)



Pati B reimbws.

National Local

$156/hr $148/hr

$144/hr $134/hr

regimens in whom crescendo TIAs will develop require urgent operation. Carotid endarterectomy is highly effective treatment. Influence of aspirin in the management of asymptomatic carotid stenosis Robert W. Hobson, IL-MD, William C. Krupski, MD, David G. Weiss, PhD, and the VA Cooperative Asymptomatic Carotid Artery Stenosis Study Group, UmwNew Jersey Medical School, Newayh, N.J., and IrA Cooperative Study Croup Number 167, Perrypoint; Md. Use of aspirin in cardiovascular disease has been reported as beneficial in many clinical settings. Although administration of aspirin is common in the treatment of symptomatic extracranial carotic occlusive disease, its role in the management of asymptomatic carotid stenosis has not been reported previously. During the period 1983 to 1991,444 adult male patients with significant asymptomatic carotid stenosis ( 2 50% arteriographically) were randomized in 11 clniical centers to optimal medical management including aspirin plus carotid endarterectomy versus comparable medical management alone. Patients initially received 650 mg aspirin twice daily. During a clinical follow-up of up to 8 years (mean, 47.9 2 27.9 months), patients intolerant to the initial dose of aspirin could be changed to a lower dose protocol (80 to 325 mg daily) or in some cases had the drug discontinued. At the conclusion of the study, at the time of a patient’s death, or at occurrence of a neurologic end point, 16% of patients (n = 72) were off all aspirin, 51% (n = 226) had been converted to enteric coated aspirin, and only 33% (n = 146) were on plain aspirin (27% having been placed on a reduced dosage regimen). Neurologic event rates among patients in the medically treated group were proportionately higher in patients off aspirin (41%) as compared with those on aspirin (21%). These data suggest that patients intolerant of aspirin have a higher incidence of neurologic events than those patients able to tolerate the drug. Vascular laboratory cost analysis and the impact of the RBRVS payment system Mark F. Fillinger, MD, Robert M. Zwolak, MD, Anne M. Musson, RVT, and Jack L. Cronenwett, MD, DamnouthHitchcock Medical Center, Lebanon, N.H. A computer model was developed to calculate the actual cost per hour to operate a contemporary noninvasive vascular laboratory, including realistic limitations on billable hours generated/total laboratory hours (operating

RBRVS reimburs. ’

$116/hr $llO/hr




- 18%

efficiency). This model also analyzed reimbursement under both Medicare part B and under the RBRVS guidelines as of Jan. 10, 1992, including conversion of case volumes to the entirely new set of CPT codes. The analysis was performed on both a national and a local level. National case volumes and average Medicare reimbursement allowance for each CPT code under Medicare part B for 1909 were obtained from the HCFA Bureau of Policy Development. “Local” values for case volumes and Medicare reimbursement were obtained from our own laboratory. Using representative costs for equipment, personnel, facilities and overhead for the national model, we project an average operating cost of $156 per laboratory hour, which is similar to the “local” value of $148 per hour calculated from the actual costs of our own laboratory. The table compares global reimbursement (technical + professional fees) in dollars per hour for a laboratory performing outpatient studies under Medicare part B and the new RBRVS system, along with the resulting decrease in Medicare revenue. The costs shown are based on a realistic operating efficiency of 80% nationally and an actual value of 77% locally. If billable hours could be generated for each available minute of technologist time (an unachievable 100% efficiency), the theoretic minimum operating cost is $125 per hour, which is still greater than RBRVS reimbursement. Laboratories with a case mix including inpatients would fare substantially worse, requiring up to 130% efficiency to meet costs. By eliminating physician costs and revenue from the model, an efficiency of 122% would be required to meet technical costs alone. When the technical and professional fees are examined separately, it is clear that most of the reduction in revenue under RBRVS is in the professional component. These results are for a best case scenario assuming neutral volumes. We conclude that the RBRVS reimbursement system is unrealistic in terms of laboratory costs and oeprations. Under these guidelines, vascular laboratories whose payor mix includes a high proportion of Medicare patients will not meet operating expenses. Captopril renal scintigraphy: an advance in outcome prediction in renovasdar hypertension George H. Meier, MD, Bauer E. Sumpio, MD, ColleenM. Brophy, MD, John F. Setaro, MD; Henry R. Black, MD, and Richard J. Gusberg, MD, Yale University School of Medicine, New Haven, Corm. Since Goldblatt’s description of experimental renovascular hypertension in the 1930s prediction of response to revascularization has been difficult. The advent of clinical

Voklme 15 Number 6 June 1992

renin determination promised improved outcome prediction in revascularization. Nonetheless, a significant percentage of renovascular patients have hypertension improvement without abnormal renin studies. Thus the prediction of hypertension improvement with revascularization for renovascular disease remains elusive. Captopril renal scintigraphy (CRS) is a nuclear medicine technique for evaluating each kidney independently for changes in GFR and perfusion induced by captopril associated alterations in vascular tone. Captopril renal scintigraphy is noninvasive, independent of renal function, requires no significant alteration in antihypertensive regimen, and is an accurate method of screening for renal artery stenoses. Since its role in predicting the response to revascularization is uncertain, the following study was undertaken. A group of 107 patients with refractory hypertension were evaluated prospectively by CRS and angiography for the detection of renovascular disease. Of these patients, a subgroup of 41 patients were revascularized by surgery (21 patients) or balloon angioplasty (20 patients) for hypertension control (28 patients) and renal salvage (16 pacaptopril-induced scintigraphic tients) . Preoperative changes were present in 24 of the 41 patients. Hypertension was evaluated at 3 to 6 months after operation with use of using AHA criteria. Of the patients with captoprilinduced scintigraphic changes, 21 of 24 patients had cure or improvement in their hypertension. When captoprilinduced changes were not present, only 1 of 17 patients improved (p > 0.001). On the basis of these data, CRS appears to reliably predict hypertension response to revascularization in patients with renovascular disease. Ancrod vs heparin for anticoagulation during vascular surgical procedures C. William Cole, MD, Janis Bormanis, MD, Kenneth A. Harris, MD, Gregory K. Luna, MD, and George G. Hajjar, MD, Civic Hospital, Ottawa, Canada and I’ictotia Hospital, London, Canada On occasion vascular surgeons must seek an alternative to heparin as a means of anticoagulation. Depletion of fibrinogen to low levels (


Results comparison (Cole et al.)

Mean age (SEM) Mean runoff score patency l-month

Ancrod n = 14

Heparin n = 14

75 ( ZL 2.3) 8.6 (f 0.3) 13

73 (f 2.2) 8.1 (2 0.3) 14

tively by intravenous infusion (70 units over 12 hours) until the serum fibrinogen was between 0.3 and 0.5 gm/L, and the prothrombin time was 1 or 2 seconds longer than control. The infusion was discontinued and rhe operative procedure was carried out without further anticoagulation. Ancrod was restarted after operation (70 units over 24 hours) to maintain the serum fibrinogen < 0.5 gm/L for 48 hours. All grafts were assessed after operation by objective, noninvasive means to document patency. The results are in the table. No clotting of blood within the graft or native vessels was noted during the conduct of the surgical procedure. The operative procedure, complication rate and hospital course were no different whether anticoagulation was established with ancrod or heparin. Conclusion: Fibrinogen depletion with ancrod provides anticoagulation for the conduct of infra-inguinal vascular reconstructions which is as effective as heparin. When heparin is contraindicated, ancrod is an effective and safe alternative. Interruption of vascular graft thrombosis and occlusion by antiplatelet glycoprotein IIb/IIIa receptor monoclonal antibody William C. Krupski, MD, Arie Bass, MD, Andrew B. Kelly, DVM, Zaverio M. Ruggeri, MD, Laurence A. Harker, MD, and Stephen R. Hanson, PhD, University of Colorado Health SciencesCenter, Denver, Cob., and Emory University School of Medicine, Atlanta, Ga. Since platelets cause thromboocclusion of arterial endarterectomies and small-diameter vascular grafts, we have determined the effect of blocking the platelet glycoprotein (GP) IIb/IIIa receptor-dependent platelet recruitment by intravenous bolus infusions of the murine monoclonal antibody (LJ-CP8) directed against the platelet GP IIb/IIIa complex in baboons. Two models of thrombosis were evaluated. In the first, 1 cm segments of 4 mm inner diameter PTFE grafts were interposed into common carotid arteries of five control and four animals treated with LJ-CPS. Indium- 11 l-labeled platelet deposition on PTFE grafts is shown in the table. After 30 days, four of four (100%) grafts in LJ-CP8treated animals were patent, compared with two of five (40%) grafts in control animals (p = 0.06). In the second model, Indium-11 l-labeled platelet deposition on endarterectomized and unendarterectomized aortic segments incorporated into chronic arteriovenous shunts was measured by quantitative gamma camera imaging. Standard 1





Control LJ-Cl’8 p value *Total tRatio





et al.)

30 Minutes*

60 Minutes*







1.93 0.14

2.85 -t 0.20 0.07 +- 0.02 0.0001

2.57 0.16

t 0.43 + 0.06 0.004

6.25 0.79

t 1.35 F 0.29 0.01

4.98 0.77

+ 1.76 ” 0.29 0.09

f 0.35 ? 0.05 0.006

platelets deposited x 109/cm. of radioactivity in graft to blood ( 1 ml).

cm long endarterectomies were performed at the midportion of the aortic segments (EA) with use of an eversion technique. After 60 minutes, acute platlet deposition on untreated control aortas (n = 6) was 0.89 -C 0.26 x 109/cm; on EA (n = 6), 4.40 + 0.89 x lo9 platelets/cm were deposited. Infusion of LJ-Cl?8 (n = 4) reduced deposition to 0.23 t 0.01 x lo9 platelets/cm (p < 0.05 compared with controls, p < 0.005 compared with EA.) Blocking of the platelet GPIIb/IIIa receptor by single bolus injection of anti-GPIIb/IIIa monoclonal antibody interrupts platelet thrombus formation is clinically relevant models of thrombosis. Prevention of early platelet deposition by such short-acting monoclonal antibodies also appears to protect small caliber vascular grafts from thromboocclusion long term.

Evaluation of a new treatment strategy for Paget Schroetter’s syndrome Herbert I. Machleder, MD, University of Califbmia, Los Angeles, Los AnJeles, Calif Fifty consecutive patients with spontaneous axillosubclavian vein thrombosis were prospectively entered into a previously described staged multidisciplinary program of therapy. The diagnosis was established by contrast venography in all cases. The protocol included initial thrombolytic therapy and anticoagulation, followed by surgical decompression of the thoracic outlet and transluminal angioplasty of residual stenosis. At the time of the thrombotic event 38 patients were engaged in strenuous and 12 in sedentary activity. Fourteen were laborers and 21 were engaged in competitive sports. Of 31 men (mean age, 23.7 years) and 19 women (mean age, 37.5 years) 28 were initially treated with thrombolytic and 7 were untreated. agents, 15 with anticoagulants, Thirty-three underwent secondary surgical treatment, and all patients were followed for a mean of 1.6 years after therapy. Twelve veins remained occluded and 38 were restored to patency. Fourteen of 17 (82%) treated with urokinase achieved durable patency. Of 29 patients with follow-up studies of the contralateral vein, 16 had a similar compressive abnormality, and five occluded during the course of the study. Of 23 surgically treated patients with serial follow-up venographic studies, 22 (96%) remained patent at a mean restudy period of 25 months.

Forty patients returned to their usual work and avocation, five workers remained disabled or retired, and five laborers were retrained for other occupations. This staged multimodal approach results in superior restoration of function, relief of symptoms, and venous patency compared with previously reported algorithms of management. Sympathectomy for reflex sympathetic dystrophy: factors affecting outcome Ali F. Abu Rahma, MD, Patrick A. Robinson, MD, Melissa Powell, MD, Deniz F. Bastug, MD, and James P. Boland, MD, West Viginia University Health Sciences Centev/Cbarleston Area Medical Center, Charleston, W. IT This is a report on a 12-year experience with sympathectomy for causalgic pain of reflex sympathetic dystrophy (RSD). Twenty-one patients had lumbar sympathectomies and seven had cervical dorsal sympathectomies. The mean duration between injury and chemical and surgical sympathectomy was 10 months and 11.4 months, respectively. Ten patients (36%) had overt extremity trauma as the precipitating event. Ten patients (36%) had a lumbar laminectomy. Twenty-four patients were diagnosed as stage II RSD (Drucker classification). There were no operative deaths, and the operative morbidity was 7%. However, 25% had transient postoperative sympathetic neuralgia. Twenty-three patients (82%) were satisfied (complete or partial relief of symptoms), and five patients (18%) were unsatisfied early (30 days) after surgical sympathectomy; however, 20 patients (71%) were satisfied, and eight (29%) were unsatisfied later ( 2 6 months). Patients with stage I/II were significantly more likely to have satisfactory early and late postoperative results than stage III (92% and 80% vs 0% and O%, respectively p = 0.017). Patients with an excellent response to chemical sympathectomies (symptom relief of > 6 hours) were more likely to have satisfactory early and late surgical outcomes. The time between injury and chemical and surgical sympathectomies was significantly shorter in patients who had satisfactory early and late surgical outcomes (p < 0.0001). There was no signiftcant association between type of injury and outcome. Surgical sympathectomy should be confined to patients with stage II disease who have had a good response to chemical sympathectomy and when relief from repeated

Volume 15 Number 6 June 1992



sympathetic block becomes less effective and the response is dramatic but of shorter duration. *p ( 0.05

Heparin prevents postischemic endothelial cell dysfunction by a mechanism independent of its anticoagulant activity W. Charles Sternbergh, III, MD, Raymond G. Makhoul, MD, and Burt Adelman, MD, H.H. M&the VA Medical Center and Medical Collge of Vigiba, Richmond, Va. Sponsored by: Michael Sobel, MD, Richmond, Va. Heparinization has reduced skeletal muscle necrosis in several skeletal muscle ischemia/reperfusion models, but the mechanism of action has not been elucidated. We hypothesized that heparin may have direct endothelial protective effects that are distinct from its anticoagulant, antiplatelet, or anticomplement functions. To explore this possibility, endothelium-derived relaxing factor (EDRF) release was measured from the intact vasculatnre of isolated rat hindlimbs with or without 60 minutes of ischemia and 10 minutes of reperfusion. Decreases in EDRF release are a sensitive marker for endothelial cell dysfunction. Potential heparin modulation of plasma proteinor cellular interactions was eliminated by perfusing the hindlimb with a nonrecirculating albumin-enriched crystalloid buffer. Fifty limbs were divided into seven groups: two nonischemic groups (one with heparin), and five ischemia/reperfusion (I/R) groups treated with increasing doses of heparin (0 to 1.0 units/ml). Endothelial cell function was assessed by subjecting phenylephrine-preconstricted limbs to increasing log dose infusions of acetylcholine and nitroprusside to measure endothelial-dependent (EDRF) and endothelialindependent vasorelaxation, respectively. Total vascular resistance (TVR) was calculated from the flow rate and the constant perfusion pressure of 70 mm Hg. Endothelial-dependent vasorelaxation of the rat hindlimb was significantly attenuated after injury from ischemia and repetfusion (closed circle vs open circle in the figure). Endothelial-independent vasoreactivity was unchanged (data not shown). Pretreatment with heparin 0.5 units/ml provided complete protection from this endothelial injury (closed triangle) but had no effect on nonischemic limbs (open triangle). Heparin’s protective effect was dose dependent. Pharmacologic -doses of heparin provided complete endothelial protection from ischemia/reperfusion injury via a mechanism independent from its interactions with blood cells or plasma proteins. A direct protective interaction of heparin with the endothelium was suggested.

After the blue toe: prognosis of noncardiac arterial embolization in the lower extremities Kara H.V. Kvilekval, MD, Jonathan P. Yunis, MD, Robert A. Mason, MD, and Fabio Giron, MD, Health Sciences Center, Stony Brook, N.Y.

Endothelial-dependent vasorelaxation of the rat hindlimb was significantly attenuated after injury from ischemia and reperfusion (0 vs 0) (Sternbergh et al.).


(U/ml) in POST-ISCHEMIC 0 0.i 025 0.5 1.0


J~hA l

I2 (n-12)


k i,;*

* !


Pretreamrent with heparin 0.5 units/ml provided complete protection from this endothelial injury (A) but had no effect on nonischemic limbs (A). Heparin’s protective effect was dose dependent (Sternbergh et al.).

To better’ understand the prognosis of arterioarterial embolic disease, we reviewed the outcomes of 41 patients with embolization to the viscera and lower extremities that originated from a radiographically identified proximal arterial source. All patients were admitted within the past 6 years. There were 31 men and 10 women, with a mean age of 65 years and a mean follow-up of 2 years. The overall mortality rate was 17% (7 of 41) and recurrent embolization 15% (6 of 41). To compare outcomes associated with supradiaphragmatic versus subdiaphragmatic disease, two groups were defined: group I comprised patients (n = 5) in whom the identified embolic source extended above the diaphragm, and group II comprised patients (n = 36) in whom the source remained below the diaphragm. The mortality rate in group I was 60% (3 of 5) as opposed to 11% (4 of 36) in group II (p < 0.05). Recurrent embolization was also significantly higher in group I (60% vs 8% [p < 0.0251). There were two amputations in group I and six in group II (p = NS) . Group II patients were then



divided into two subgroups: those with limited disease (n = 19) in which the emboli had a single, radiographically identifiable source (i.e., aneurysm or single area ofulcerated plaque) and those with diffuse disease (n = 17) in which the emboli had multiple, radiographically identifiable potential sources. In the subgroup of patients with limited disease, no deaths or episodes of recurrent embolization occurred, whereas four deaths and three episodes of recurrent emboliiation occurred in the subgroup of patients with diffuse disease. The differences in these outcomes, however, were not statistically significant. Thirty selected patients (1 from group I and 29 from group II) had surgery on, or to bypass, the imputed lesion. Only 1 (7%) of these 30 patients had recurrent embolization. In contrast, recurrent embolization was noted in 4 (36%) of the 11 patients who did not have surgery directed at the lesion @ c 0.025). There was no significant different in mortality rate between patients who underwent surgery and those who did not. In conclusion, patients with arterioarterial emboli have a substantial mortality rate and risk of recurrent embolization, especially if the disease process extends above the diaphragm. In selected patients, however, surgery may decrease the frequency of recurrence without increasing mortality. Surgical aspects of arterial occlusive disease in young adults Pave1 J. Levy, MD, Daniel S. Rush, MD, William 0. Holloway, MD, John R. Hobson, Jr., MD, Joseph G. Fine, MD, and James L. Haynes, MD, Uziveys$ of South Carolina Medical School, Columbia, S. C. Surgical management of 103 young adults with lower extremity atherosclerotic arterial occlusive disease (AAOD) was retrospectively studied. Of these patients 61 men and 42 women, 57 were white and 46 black. The mean age at onset of symptoms was 36 years (range, 25 to 40 years) and at first surgical intervention was 39 years (range, 26 to 48 years). Of 38 patients with rapidly progressing ischemia, 32% had hypercoagulable states. Intervention was needed within 5 years of initial symptoms in 90% of patients. Sixty patients had severe limb-threatening ischemia, and 43 had claudication. Of 275 procedures performed 36 patients required only 1 procedure; 67 patients had multiple procedures (mean, 3.6 per patient; range, 2 to 15), and in 55 patients these were bilateral. The 144 primary interventions were balloon angioplasty, 46; inflow or infrainguinal revascularization, 39 each; amputation, 19; and sympathectomy, 1. Life-table analysis of aortofemoral and femoropopliteal bypasses indicated a greater than expected graft failure rate. Fifty-one patients required 131 secondary interventions; of these, only 12% had distal progression of AAOD. Multiple ( > 3) interventions were more frequent in patients with unique risk factors (hypoplastic arteries and hypercoagulable states) compared with the other patients with typical risk factors for AAOD (Fisher’s Exact Test,


p < 0.05). Overall, 40 patients have required major amputations. Surgical intervention for ‘AAOD in young adults has a high incidence of early failure and need for multiple procedures and amputations. Of concern are the adverse effects of unique risk factors and the use of infrainguinal revascularization despite rapidly progressive inflow, rather than distal, AAOD. Descending thoracic aorta to femoral bypass: ten years’ experience with a durable procedure Walter J. McCarthy, MD, Charles L. Mesh, MD, William D. McMillan, MD, William R. Flinn, MD, William H. Pearce, MD, and James ST. Yao, MD, Nmthwestem U&e&y Medical School, Chicag-o, Ill. Arterial bypass from the descending thoracic aorta to the femoral system is known to provide initially successful perfusion when an approach to the abdominal aorta is undesirable. This review of a large series and a decade’s experience suggest safety and patency comparable to standard aortofemoral reconstruction. During a lo-year period ending in January 1992, 20 patients, 18 men and two women, (mean age, 59 years) underwent descending thoracic aorta to left femoral bypass with femorofemoral bypass. These were placed to correct multiply failed axillopopliteal or axillofemoral grafts (11 patients). The axillary bypasses had been previously placed to allow removal of infected aortic grafts and had undergone a total of 19 revisions or thrombectomies before conversion. In addition, patients had this procedure after multiple failures (greater than two) of aortofemoral grafts (five patients) or to avoid exploring the abdomen after extensive retroperitoneal dissection or radiation (four patients). Surgical technique involves a seventh interspace thoracotomy with a transdiaphragmatic retroperitoneal tunnel to the groin. Partial occlusion of the supradiaphragmatic thoracic aorta allows an end-to-side anastomosis with a 10 mm polytetrafluoroethylene prosthesis. There were no perioperative deaths. The mean hospital visit was 13.5 days and intensive care stay 4.7 days. There were no perioperative myocardial infarcts, strokes, or renal failure requiring dialysis. With follow-up of 1 to 111 months (mean, 39), the 4-year patency was 100%. A single graft failed at 49 months but was renewed by thrombectomy and distal bypass. Thus the 9-year patency was 82%, but the secondary patency remains 100% throughout. During follow-up, patients required five femoropopiiteal, four femorotibial, and one profunda reconstruction. Four patients died of myocardial infarction during follow-up (mean, 42.5 months) with patent grafts, one related to infection to the thoracic graft. Two patients eventually required limb amputation, but neither of these was due to inflow failure. Descending thoracic aorta to femoral grafting is a safe, extremely durable arterial bypass configuration. It is an excellent reconstruction for survivors of aortic graft infec-

Vohlme 15 Number 6 June 1992

tion, those with multiply failed aortic grafts, and patients for whom abdominal exploration would be hazardous. In situ allograft replacement of infected infrarenal aortic prosthetic grafts: results in 37 patients Edouard Kieffer, MD, Amine Bahnini, MD, Fabien Koskas, MD, Carlo Ruotolo, MD, Dominique Le Blevec, MD, and Didier Plissonnier, MD, Pitie-Saketriere University Hospital, Paris, France. Spomwed by: Rmnon Beguer, M.D., Detroit, Mich. From August 1988 to November 1991, 37 patients with infected infrarenal aortic prosthetic grafts underwent in situ replacement by means of preserved allografts obtained from cadavers as part of a multiorgan transplant tissue retrieval program. Twenty-eight patients had isolated prosthetic infections whereas nine had aortoenteric fistulas. The allograft revascularized both lower extremities in 32 patients and only one in five patients. All infected prosthetic material was removed at the same operation. Associated hypogastric or visceral arterial reconstructions were performed in 14 patients. Sixteen. patients had nonvascular procedures, including 11 intestinal procedures. Four patients (11%) died after operation: three of general causes, one of persistent infection as a result of incomplete debridement of the native aorta. Two patients had successful reoperation for allograft-related complications. All surviving patients were discharged with patent allografts and have been followed with serial duplex and CT scans for a mean follow-up of 13.5 months (1 to 30). Fourteen patients have been monitored for more than 1 year. There were three late deaths: two unrelated to the vascular operation and one possibly caused by late recurrent or persistent infection. Two patients have had pathologic changes in the allograft, with one requiring a reoperation. There were no early or late amputations. Although late deterioration may be expected, in situ allograft replacement appears to be a major advance in the management of infected infrarenal aortic prosthetic grafts as it avoids most of the infectious and vascular complication of conventional methods. Experience with 1509 patients undergoing thoracoabdominal aortic surgery Lars G. Svensson, MD, E. Stanley Crawford, MD, Kenneth R. Hess, MD, Joseph S. Coselli, MD, and Hazim J. Safi, MD, Baylor College of Medicine, Houston, Texas From 1960 to 1991, 1509 patients underwent 1679 thoracoabdominal aortic repairs. The median age as 66 years (range, 1.5 years to 86 years) and aortic dissection was present in 276 (18%). The extent of the first repair performed included 378 (25%) type I (proximal descending to upper abdominal aorta); 442 (29%) type II (proximal descending to below the renal arteries); 343 (23%) type III (distal descending and abdominal aorta);



and 346 (23%) type IV (most of the abdominal aorta). The median total aortic clamp time was 43 minutes. The 30-day survival rate was 92% (1386 of 1509) for the 32-year period. On univariate analysis, the variables associated with death included (p < 0.05) age greater than 65 years, hypertension, respiratory, cardiac and renal disease, and postoperative complications. The overall incidence of paraplegia or paraparesis was 16% (234 of 1509). By use of logistic regression analysis, the best predictors of paraplegia or paraparesis developing were (p < 0.05): Clamp time (odds ratio [OR] = 1.98 per 30-minute increment); aneurysm extent type II (OR = 1.85 versus type I); type III (OR = 0.31); type IV (OR = 0.18); and increasing age (OR = 1.47 per 20-year increment). Use of atriofemoral bypass had no effect on the incidence. Renal failure (postoperative creatinine > 3 mg/dl or dialysis) occurred in 18% (269 of 1509) patients, dialysis was required in 9% (136 of 1509), and gastrointestinal complications manifested in 7% (101 of 1509). In conclusion, although the survival rate has improved, paraplegia/paraparesis, and renal failure continue to be vexing problems that require further research to reduce the associated morbidity of thoracoabdominal aortic surgery. Adventitial elastolysis is a primary event in aneurysm formation John V. White, MD, Kent Haas, MD, Steven Phillips, MD, and Anthony J. Comerota, MD, Temple University Philadelphia, Pa. Adventitial elastin degradation is a hallmark of human abdominal aortic aneurysm (AAA) formation. However, the quantitative relationship between elastin loss and AAA formation and growth is unknown. This study was undertaken to quantitate the reduction of adventitial elastin for small AAAs, determine whether the loss of this structural component parallels aneurysm growth, and examine the ultrastructure of the remaining elastin elements. Longitudinal and transverse strips of anterior aneurysm wall were taken from 12 patients having elective repair of small (diam 5 cm, n = 4), moderate (diam 5 to 7 cm, n = 4), or large (diam 7 cm, n = 4) AAAs and from six normal controls at autopsy. Specimens were prepared with Weigert-van Gieson elastin stain for histologic examination or formic acid (which digests all nonelastin arterial wall components) for scanning electron microscopic evaluation of elastin architecture. Adventitial elastin content was quantitated with a calibrated video microscopy system by determining the number of elastin fibers within a 1.6 x lOma mm2 grid. Ten separate fields of greatest elastin stain uptake on five longitudinal and five frontal histologic sections (100 measurements) were evaluated for each patient. AAA and control elastin content were compared by Wilcoxon rank sum test for unpaired samples. The three-dimensional adventitial elastin architecture of






30 E


r iii 2






! m


Comparison of angioscopy and angiography for monitoring infrainguinal bypass vein grafts: results of a prospective randomized trial Arnold Miller, MB, Edward Marcaccio, MD, Gary Tannenbaum, MD, Christopher Kwolek, MD, Peter Stonebridge, MB, Philip T. Lavin, PhD, Gary W. Gibbons, MD, David R. Campbell, MD, Frank B. Pomposelli, MD, Dorothy V. Freeman, MD, and Frank W. LoGerfo, MD, New En&and Deaconess Hospital, Bostoq Mass.

, moNTALsEcnoN











elastin (White et al.).

12 10 e 6 4 2 0 (eS.0) (6.0-7.0) (87.0) AORTIC DIAMETER (IN CU) p >0.05





Elastin content versus aortic size (White

et al.).

AAA and controls was determined by SEM of longitudinal and frontal sections oriented to the directon of blood flow. Adventitial elastin content of all AAAs was reduced 81.6% ? 2.1% (p < 0.001). There was no significant difference in elastin content between small, moderate, and large AAAs (p > 0.05). SEM of controls demonstrated multiple lamellae of advent&J elastin alternating with layers of collagen. The elastin lamellae were composed of fibers oriented parallel to the direction of blood flow. The elastin work of AAA was grossly disrupted. No discrete lamellae were present, and remaining fibers were fractured, curled, and without a uniform axis of orientation. These data strongly suggest that elastolysis is a primary event in AAA formation, occurring before overt loss of adventitial structural integrity and the development of small aneurysms. AAA growth is independent of elastin degradation and not the result of elastase on the aortic wall. Data shown on figures.

This study was undertaken to compare the efficacy of completion intraoperative angioscopy (IAS) and intraoperative angiography (IAG) in the “optimal” quality primary infrainguinal vein bypass graft with regard to the sensitivity of identifying correctable technical defects (clinical decisions), effects on graft patency, operation time (costeffectiveness), fluid administration, and safety. Between February 1990 to November 1991, 282 consecutive primary saphenous vein infrainguinal bypass grafts (fern-pop 83, fern-distal 116, pop-distal 83) in 273 patients were prospectively randomized and monitored with either IAS (n = 122) or IAG (n = 117). Forty-three bypasses (IAG 24, IAS 19) were excluded from the study for lack of adequate greater saphenous vein, surgical preference, or unavailability of angioscopy. Clinical parameters, indications for surgery (limb-threatening ischemia in 93%), graft extent, and configuration (reversed, nonreversed and in situ) were evenly matched in both IAS and IAG, and the exclusion groups, confirming success&l randomization. Thirty-seven clinical decisions were made during the completion studies: only one graft (IAG) failed in less than 30 days where a decision was made. In IAS, 30 decisions (conduit 27, distal anastomosis 2, distal artery 1) in 122 bypasses and in IAG, 7 decisions (conduit 1, distal anastomosis 5, distal artery 1) in 117 bypasses resulted in interventions Primary patency was 121 of 122 (99.2%) for IAS and 110 of 117 (94%) for IAG at less than 30 days (p = 0.033, Fisher’s Exact Test), and 82.9% and 77.3% (life-table analysis) at 18 months, respectively. Mean total intraoperative fluid administered (IAS = 1876 ml, IAG = 1784 ml), mean operation time (IAS = 245 mins, L4G = 235 mins), and perioperative mortality (TAS = 3 of 122 (2.46%), LAG = 2 of 117 (1.71%)) was no different in the two groups. Our study shows that completion IAS is not only as safe and cost-effective as IAG, but is more effective in monitoring infrainguinal vein bypass grafts, resulting in significant improvement in early graft patency, even in vein grafts at least risk for early occlusion.