Saphenous vein patch rupture carotid endarterectomy

after

Patrick J. O 'Hara , MD, Norman R. Hertzer, MD, Leonard P. Krajewski, MD, and Edwin G. Beven, MD, Cleveland, Ohio

From January 1983 to September 1990, 2731 carotid endarterectomies were performed at The Cleveland Clinic. Patch angioplasty with autogenous saphenous vein was used for arteriotomy closure during 1691 (62%) of these procedures and was associated with eight postoperative ruptures (0.5%) of the central portion of the patch in seven patients. This complication occurred in three men and in four women (mean age, 69 years), all of whom were hypertensive and all but one were smokers. Two patients (29%) had diabetes. In each case of patch rupture the vein had been harvested from the leg distal to the knee. Although the harvest site could not be determined retrospectively for every patient in this series, no patch ruptures were encountered among 370 procedures for which it could be documented that the saphenous veins had been obtained from the groin. All ruptures occurred within 5 days of the primary operations (including four during the first 24 hours) and were urgently corrected by primary closure of the original arteriotomy in two cases and by replacement of the ruptured patch in the remaining six. Two (29%) of the seven patients either died or sustained a permanent neurologic deficit. Central rupture of a saphenous vein patch is a rare but devastating complication after carotid endarterectomy. Since vein harvested from the lower leg or ankle may be marginally more likely to rupture than proximal vein from the thigh or groin, it should not be used indiscriminately for carotid patch angioplasty. (J VAsc SURG 1992;15:504-9.)

It is widely recognized that attention to technical detail is necessary to minimize the incidence of complications after carotid reconstruction. In an effort to improve surgical outcome, several reports have encouraged the use of patch angioplasty for arteriotomy closure in conjunction with carotid endarterectomy.l7 In a prospective but nonrandom- ized study from our own center, the merit of this approach was assessed in a series of 917 carotid endarterectomies (801 patients) from 1983 through 1985. Patch angioplasty was performed with a segment of greater saphenous vein that usually was harvested from the ankle during 434 (47%) of these procedures, and the remaining 483 arteries (53%) were closed primarily. The operative mortality rate (0.5%) was identical for both the patched and nonpatched groups, but vein patching was associated with significantly lower risks for perioperative stroke

From the Department of Vascular Surgery, The Cleveland Clinic Foundation, Cleveland.

Reprint requests: Patrick J. O'Hara, MD, Department of Vascular Surgery, The Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44106.

24/1/31964

504

(0.7% vs 3.1%;p = 0.0084), early thrombosis of the internal carotid artery (0.5% vs 3.1%; p = 0.0027), and late recurrent stenosis (4.8% vs 14%; p = 0.014)?

On the basis of these favorable results, we have since used vein patch angioplasty as a routine precaution. Nevertheless, acute false aneurysms that required urgent reoperation developed in three patients in our original patch group (0.7%) within the first few postoperative days. Two of these events were caused by anastomotic suture line disruptions, but a single patient had an unexplained rent in the center of the saphenous vein patch. Eikelboom et al. 8 also described one patch rupture (1.5%) among 67 carotid endarterectomies closed with saphenous vein obtained from the ankle, and although Riles et al.9 never have encountered central rupture in more than 600 carotid patches harvested from the groin, they have reported this complication in 4% of patients in whom ankle vein was used. Because of these obser- vations and the fact that additional patch ruptures eventually occurred in our own series, we have reexamined our experience with this serious problem to place our earlier advice concerning vein patch angioplasty into perspective.

Volume 15 Number 3 March 1992 Carotid vein patch rupture 505

PATIENTS AND METHODS

From January 1983 to September 1990 the authors performed 2731 carotid endarterectomies (Table I), a figure that includes reoperations for recurrent carotid occlusive disease but excludes all procedures that were limited to the external carotid artery. A total of 904 arteriotomies (33%) were closed primarily, but some form of patch angioplasty was used in the remaining 1827. Because of our traditional preference for autogenous material, we used saphenous vein for 1691 (93%) of the patches and either polytetrafluoroethylene (PTFE) or Da- cron for only 136. Our previous preference for primary closure declined steadily throughout the study period (Fig. 1). In 1983, the first year of our vrospective investigation of patch angioplasty, ap- proximately two thirds of all carotid arteriotomies were closed primarily, and just one third were patched. By the time our initial results were published in 1987, however, we elected to use patch closure during 86% of our carotid endarterectomies, and by September 1990 only 3.5% of our arteriotomies were closed primarily. We currently use saphenous vein patches in nearly 78% of carotid endarterectomy operations, but the fact that we now use prosthetic material in approximately 18% reflects our reluctance to accept a substandard saphenous vein for patching purposes.

Although saphenous vein patches were harvested preferentially from the ankle from 1983 through 1985, we subsequently have tended to use more groin veins because of a relative concern regarding the tensile strength of the distal vein in at least some patients. Although it was difficult to determine retrospectively whether the saphenous vein was obtained from the thigh or the lower leg in each of the patients undergoing carotid endarterectomy be- fore 1989, this information has been recorded ha a departmental computer registry for the 496 carotid endarterectomies with patch closure performed by the authors since that time (Fig. 2). During this recent interval, 74% of all carotid patches (370) were constructed with greater saphenous vein from the proximal thigh, and ankle veins were used in 48 procedures (10%). The remaining 78 patients (16%) received other types of patches consisting of PTFE in 59, Dacron in 13, and autogenous arterial segments or miscellaneous materials in six.

RESULTS

Patch angioplasty with autogenous saphenous vein was used for arteriotomy closure during 1691 carotid endarterectomies and was associated with

eight postoperative ruptures (0.5%) of the central portion of the patch in seven patients. Although no evidence of local infection was observed in any of these patients, each was found during an urgent reoperation to have a dilated patch and a central, transverse laceration despite the presence of an intact suture line. The rupture group consisted of three men (43%) and four women (57%) ranging in age from 61 to 76 years (mean, 69 years). All were hyperten- sive, all but one were smokers, and two (29%) had diabetes. Each had ~;ndergone carotid endarterec- tomy with use of general anesthesia and routine intratuminal shunting. All patients were neurologi- cally intact after the carotid procedure and before the vein patch rupture. In every case the arteriotomy had been closed with a patch of saphenous vein harvested from the leg distal to the knee.

The distribution of carotid patch ruptures dur- ing the entire survey period is illustrated in Fig. 3. Six ruptures occurred after primary carotid proce- dures and two after reoperations for recurrent carotid occlusive disease (Table II). One woman whosc ankle vein patch had ruptured sustained the same complication with a replacement patch har- vested from the calf. Her original arteriotomy then was closed primarily to avoid placing synthetic material into a wound that potentially was con- taminated, and she fortunately had no neurologic complications.

The mean interval between patch construction and rupture was 2.3 days. Four of the eight ruptures occurred within the first 24 hours, but the others were evenly distributed throughout the next 4 postoperative days. All patients required urgent reoperations because of sudden, severe cervical hem- orrhage. Two arteriotomies were closed primarily, and the remaining six were managed by secondary patch closure with use of additional saphenous vein harvested either from the groin (N = 2) or the calf (N = 2), an endarterectomized segment of superfi- cial femoral artery that previously was occluded (N = 1), or PTFE (N = 1). As already indicated, one of the replacement vein patches obtained from the calf also ruptured and required primary closure for definitive repair.

All seven patients were known to be hypertensive before endarterectomy. Although the recorded blood pressure measurements obtained from the hospital records varied widely, blood pressure in two of the five patients was always well controlled with no recorded systolic pressures over 170 mm Hg or diastolic pressures over 90 mm Hg. In contrast, blood pressure in the remaining five patients was

506 O'Hara et al.

~ournal of VASCULAR

SURGERY

100

• 80l 631 829 0 81.7 .~.Z8 4

SS" "0 ~ 66.5 63.6/"

-~",,,,,~56 8 ,,'"* Primary closure

~ 8 60 ~ " ~ - . . . ~ . 8 ,,," Saphenous vein patch R . S

~ _ X . . . . Other patch

- _.,....,.,.'~ "N~.9

- \138 1_.8:4 ~ 9 . 7 . - .13~0....-."

e., o

"0 e-

I l l

40

20

1.3 1.4 1.5 4 . 5 . - - 0 0~s . . . . I . . . . . ~ . . . . . I . . . . . . - 1 17.2 I 1983 1984 1985 1986 1987 1988 1989 1990

Y e a r

January 1983 to September 1990

Fig. 1. Graphic representation of the method of arteriotomy closure and the type of patch material used in 2731 carotid endarterectomies from January 1983 to September 1990.

Table I. Method of carotid arteriotomy closure used from January 1983 to September 1990

Primary closure Vein patch Other patch

Year Total No. % No. % No. %

1983 400 266 66.5% 132 ,33.0% 2 0.5% 1984 370 210 56.8% 155 41.9% 5 1.3% 1985 356 188 52.8% 163 45.8% 5 1.4% 1986 393 137 34.9% 250 63.6% 6 1.5% 1987 377 52 13.8% 308 81.7% 17 4.5% 1988 319 31 9.7% 265 83.1% 23 7.2% 1989 315 13 4.1% 261 82.9% 41 13.0% 1990 201 7 3.5% 157 78.1% 37 18.4% Total 2731 904 33.1% 1691 61.9% 136 5.0%

more difficult to control, with occasional systolic pressures exceeding 200 mm Hg or diastolic pres- sures exceeding 90 mm Hg at some time during the hospitalization before patch rupture. Nevertheless, blood pressure control was achieved before each of the eight patch ruptures, with the closest prerupture systolic recordings ranging from 100 to 170 mm Hg and diastolic recordings ranging from 52 to 90 mm Hg.

One death (14%) occurred among these seven patients. Carotid patch rupture also was responsible for three postoperative strokes (43%), one of which involved the patient who died. One survivor sus- tained a permanent ipsilateral neurologic deficit, and another had a transient deficit that resolved corn-

pletely within 5 days. Fatal complications or a permanent stroke thus occurred in two (29%) of the seven patients, representing 25% (2/8) of the patch ruptures.

R E C E N T EXPERIENCE

From January 1989 to September 1990 the authors performed 418 carotid endarterectomics in which saphenous vein patch angioplasty was used and its source was specifically recorded. Of these, 370 patches were harvested from the groin, and 48 others were taken from the ankle (Fig. 2). None of these patches ruptured, and in fact, we have not encoun- tered this complication in any patient since May 1988.

Volume 15 Number 3 March 1992 Carotid vein patch rupture 507

Thigh Vein 370

,o .,:W;n . . . . . . . .

78 . . . . . . . 5 9

All Carotid Patches (496)

Non-Vein Patches (78)

sc. 3

Dacron 13

January 1989 to September 1990

Fig. 2. Graphic representation of the type of patch material used in the closure of 496 carotid endarterectomies from January 1989 to September 1990.

500

400

300

200

100

400 -- ,~_ _ _ 393 377

308

Total operations Vein patches

* Patch ruptures

315

1983 1984 1985 1986 1987 1988 1989 1990

Years

January 1983 to September 1990

Fig. 3. Graphic representation of the distribution of vein patch ruptures from January 1983 to September 1990.

D I S C U S S I O N

Lingering doubts about its operative risk and durability are responsible for much of the current controversy regarding carotid endarterectomy. The initiation of several prospectively randomized trials to reassess the efficacy of this procedure has made these issues even more timely, but it always has been abundantly clear that stroke and recurrence rates must be low to justify the surgical treatment of carotid disease, especially in asymptomatic patients. To meet these objectives, patch angioplasty with autogenous saphenous vein has become the preferred

method for arteriotomy closure at some centers because it facilitates the reconstruction of small vessels and provides a nonthrombogenic endothelial surface that may discourage both perioperative thrombosis and recurrent hyperplastic stenosis. 2,s,7,1°

Our published restflts in a series of more than 900 carotid endarterectomies added further support to these perceived advantages of vein patch angioplasty and suggested that its routine use enhances surgical outcome even in experienced hands. 3 Although Clagett et al.ll subsequently failed to demonstrate the same perioperative superiority for vein patch closure

508 OJHara et al.

Journal of VASCULAR

SURGERY

Table II. Distribution of saphenous vein patch ruptures (January 1983 to September i990)

Vein patches Ruptured patches Total

Year (Total) Primary procedures Carotid reoperations No. %

1983 132 2 0 2 1.5% 1984 155 0 0 0 0% 1985 163 0 1 1 0.6% 1986 250 1 1 2 0.8% 1987 308 2 0 2 0.6% 1988 265 1 0 1 0.4% 1989 261 0 0 0 0% 1990 157 0 0 0 0% Total 1691 6 2 8 0.5%

in a small, randomized group of patients, the conclusions of this study are limited by the fact that it comprised only men who had large (_> 5 mm) internal carotid arteries and smooth endarterectomy planes. Rosenthal et alY likewise found no statistical differences in the incidence of either postoperative stroke or late recurrence among 1000 patients who had primary carotid closure or patch angioplasty with saphenous vein, Dacron, or PTFE. The median follow-up interval for this investigation was 26 months (mean, 37.8 months), however, and vein patch angioplasty nevertheless was associated with the lowest risk for neurologic complications or recurrent stenosis in any of the treatment cohorts. Fietsam et al.~s determined in laboratory animals that primary arteriotomy closure causes an obligatory stenosis of approximately 15% to 20% that can be prevented by the application of a patch measuring 5 mm in diameter. Conversely, they also discovered that larger (10 mm) patches produce turbulent blood flow, 13 a liability that may account for conflicting reports regarding the clinical effectiveness of q~otid patch angioplasty. 2,7,14 It seems quite likely that several factors may influence the relative benefits of patch closure, including the size of the patch, the material with which it is constructed, and if saphe- nous vein is used, the site from which it is harvested.

Despite the intuitive advantages of autogenous carotid patching, central patch rupture is a rare but catastrophic event that seems to be encountered only in conjunction with saphenous vein segments ob- tained from below the knee in our experience. The reported incidence of this complication ranges from 0.7% to 4%, 8'9'1s and it occurred in seven patients (three men and four women) after 1691 vein patch angioplasties (0.4%) performed at our center from 1983 to i990. The combined mortality and/or permanent stroke morbidity rate in these patients was 29%, and most of them had certain common

characteristics. All were hypertensive, all but one were smokers, and each had a vein patch harvested from the ankle or calf that ruptured transversel), within the first 5 postoperative days.

Similar results were reported by Tawes et al.ls in a review of 1760 saphenous patch carotid closures performed by 48 surgeons in the Western Vascular Society. Patch rupture occurred in 13 patients (0.7%), resulting in a 30.7% mortality rate. It is interesting to note that all but one of the ruptured patches in this series were harvested from the ankle. Eighty-five percent (11/13) of the patch ruptures occurred within the first 3 postoperative days2 s The cause of patch rupture is unclear but may be related to the geometry of the patch, the diameter of the saphenous vein, and possibly to undefined differences in the tensile properties of ankle and groin veins.

Yu et al)6 observed in vitro that the mean bursting pressure for groin saphenous vein was 94.5 +_ i5.1 psi, whereas it was 75.5 _-_ 8.9 psi for ankle saphenous vein. Although they are suggestive of a trend, these differences apparently were not statistically significant, probably because of the small sample size (N --- 10) in each group. It seems likely that hypertension increases the risk of rupture in susceptible patches from a purely mechanical per- spective by increasing the wall stress. Donovan et al.,~7 however, have also demonstrated that the preoperative diagnosis of hypertension itself, as well as age, diabetes, and gender may adversely influence the inherent circumferential tensile strength of sa- phenous vein harvested from the leg distal to the knee. These observations are further arguments that support efforts to maintain intraoperative and peri- operative normotension.

Archie and Green ~8 calculated a positive linear correlation between the diameter of intact veins and rupture pressures in vitro, a feature that appears to favor the use of proximal saphenous segments for

Volume 15 Number 3 March 1992 Carotid vein patch rupture 509

patching purposes as long as the biomechanical properties of the fileted patch are comparable to those of cylindrical veins. They also demonstrated that the zone of a vein patch that is most likely to rupture is located where the radius of curvature is greatest and the circumferential wall stress is maximal. This observation is consistent with our clinical findings and implies that a narrow vein patch confers an element of early safety as well as long-term hemody- namic superiority.

We still are convinced that vein patch closure enhances the outcome of carotid endarterectomy by minimizing the risks for postoperative thrombosis, stroke, and early recurrent stenosis. Although we have continued to use greater saphenous segments from below the knee in occasional, selected patients , asually men with excellent ankle veins on physical examination) and have not been confronted with a ruptured patch since May 1988, the use of proximal saphenous vein harvested from the groin and trimmed to construct a patch measuring approxi- mately 5 mm in diameter may provide the safest autogenous reconstruction if applied routinely under all circumstances. When removal of the saphenous vein from the thigh is for some reason contraindi- cated, primary arteriotomy closure remains an ac- ceptable alternative for patients who have large internal carotid arteries and smooth endarterectomy surfaces, and it clearly seems preferable to patching with an inadequate vein. Synthetic patch material does not provide viable endothelium, but it may be preferable to primary closure in women because of the relatively small size of their extracranial arteries and their recognized predilection for recurrent ca- rotid stenosis. In conclusion, our experience with central ruptures seems sufficiently compelling to discourage the indiscriminate use of ankle veins for carotid patching, but it clearly is not a legitimate indication to abandon vein patch angioplasty alto- gether.

REFERENCES

1. Archie IP. Prevention of early restenosis and thrombosis- occlusion after carotid endarterectomy by saphenous vein patch angioplasty. Stroke i986;17:901-5.

2. Deriu GP, Ballotta E, Bonavina L, et al. The rationale for patch-graft angioplasty after carotid endarterectomy: early and long-term follow-up. Stroke 1984;115:972-9.

3. Hertzer NR, Beven EG, O'Hara PJ, Krajewsld LP. A prospective study of vein patch angioplasty during carotid endarterectomy. Three-year results for 801 patients and 917 operations. Ann Surg 1987;206:628-35.

4. Imparato AM. The role of patch angioplasty after carotid endarterectomy [Editorial]. l VASC SURG 1988;7:715-6.

5. Lord RSA, Raj TB, Stary DL, Nash PA, Gragam AR, Goh KH. Comparison of saphenous vein patch, polytetrafluoro- ethylene patch, and direct arteriotomy closure after carotid endarterectomy. Part I. Perioperafive results. J VASC SURG 1989;9:521-9.

6. Shultz GA, Zammit M, Sauvage LR, et al. Carotid artery Dacron patch graft angioplasty: a ten-year experience. J VASC SUING 1987;5:475-8.

7. Ten Holter JBM, Ackerstaff RGA, Thoe Schwartzenberg GWS, Eikelboom BC, Vermuelen FEE, Van Den Berg ECJM. The impact of vein patch angioplasty on long-term surgical outcome after carotid endarterectomy. J Cardiovasc Surg 1990;31:58-65.

8. Eikelboom BC, AckerstaffRGA, Hoeneveld H, et al. Benefits of carotid patching: a randomized study. J VASC SURG 1988;7:240-7.

9. Riles TS, LampareUo PJ, Giangola G, Imparato AM. Rupture of the vein patch: a rare complication of carotid endarterec- tomy. Surgery 1990;107:10-2.

10. Govostis DM, Bandyk DF, Bergamini TM, Towne IB. Biochemical adaptation of venous patches placed in the carotid bifurcation. Arch Surg 1989;124:490-3.

11. Clagett GP, Patterson CB, Fisher DF, et al. Vein patch versus primary closure for carotid endarterectomy. A randomized prospective study in a selected group of patients, l Vasc SURG 1989;9:213-23.

12. Rosenthal D, Archie JP, Garcia-Rinaldi R, et al. Carotid patch angioplasty: immediate and long-term results. J VASC SURG 1990;12:326-33.

13. Fietsam R, Ranval T, Cohn S, Brown OW, Bendick P, Glover JL. Hemodynamic effects of primary closure versus vein patch angioplasty of the carotid artery (Submitted for publication).

14. Curley S, Edwards WS, lacob TP. Recurrent carotid stenosis after autologous tissue patching. J Vase SURG 1987;6:350-4.

15. Tawes RL, Treiman RL. Vein patch rupture after carotid endarterectomy: a survey of the Western Vascular Society members. Ann Vasc Surg 1991;5:71-3.

16. Yu A, Dardik H, Wolodiger F, et al. Everted cervical vein for carotid patch angioplasty. J VASC SURG 1990;12:523-6.

17. Donovan DL, Schmidt SP, Townshend SP, Njus GO, Sharp WV. Material and structural characterization of human saphenous vein. J Vase SUR6 i990;12:531-7.

I8. Archie JP, Green[IJ. Saphenous vein rupture pressure, rupture stress, and carotid endarterectomy vein patch reconstruction. Surgery 1990;107:389-96.

19. Archie JP. The geometry and mechanics of saphenous vein p~/tch angioplasty after carotid endarterectomy. Texas Heart J 1987;14:395-400.

Submitted Mar. 14, 1991; accepted June 17, 1991.