Journal of Vascular Surgery
Volume 17, Issue 1 , Pages 141-151, January 1993

Carotid endarterectomy with reconstruction techniques tailored to operative findings☆

Presented at the Forty-sixth Annual Meeting of the Society for Vascular Surgery, Chicago, Ill., June 8-9, 1992.

Raleigh and Chapel Hill, N.C

From the Wake Medical Center, Raleigh, and The University of North Carolina, Chapel Hill

Received 12 June 1992; accepted 22 August 1992.

Article Outline

Abstract 

Purpose: Reconstruction techniques tailored to operative findings were applied to 466 consecutive carotid endarterectomies (CEA) performed on 408 patients over 5 years. The choice of reconstructive technique was based on the extent of the arteriotomy incision required to obtain a complete internal CEA endpoint, the ability to obtain a complete endpoint, and the quality or redundancy of the endarterectomized internal carotid artery (ICA) segment. The hypothesis was that a complete internal CEA endpoint and a tapered, smooth, nonkinked reconstruction minimize complications. Methods: Complete distal endpoint feathering was obtained in 437 (94%) CEA. Patch reconstruction was performed in the 429 (92%) CEA in which the arteriotomy extended distal to the ICA bulb. Before patching, 16 (3.4%) redundant endarterectomized ICA segments were shortened by transverse-eversion suture plication to prevent kinking. A saphenous vein interposition graft was used in nine (1.9%) CEA. The other 28 (6.0%) CEA had an arteriotomy that did not extend distal to the bulb and were primarily closed. Results: Two (0.4%) patients died of myocardial infarction in the hospital; one of these patients also had a stroke. Three (0.6%) patients had nonfatal strokes in the hospital. Five patients had hyperperfusion syndrome after CEA, one of which occurred without complications in the hospital. The other four occurred after discharge from the hospital. Three of these patients had a stroke, and two with strokes died. The combined 30-day mortality and nonfatal stroke rate was 2.1% for CEA and 2.4% for patients. There were no patch ruptures, false aneurysms, or ICA occlusions. Three patients had a >50% diameter carotid artery stenosis 6 months after CEA. Conclusions: These results support an aggressive attempt to obtain a complete or optimal ICA endpoint with reconstruction techniques based on operative findings. Recognition of patients at risk for and treatment of hyperperfusion syndrome after CEA remains a clinical challenge. (J VASC SURG 1993;17:141-51.)

 

Excellent results with carotid endarterectomy (CEA) require attention to multiple details of perioperative management. Early postoperative internal carotid artery (ICA) thrombosis remains a major determinant of CEA neurologic morbidity and death, whereas recurrent or residual stenosis seems to be a lesser stroke threat.1, 2 Patch reconstruction is reported to decrease the incidence of both early thrombosis and recurrent stenosis.3, 4, 5, 6, 7, 8, 9, 10 However, routine patching may not be sufficient to ensure a hemodynamically sound reconstruction that is resistant to thrombosis and restenosis. Patch reconstruction alone may not protect against the risk of thrombosis caused by kinking, residual distal plaque, or a thrombogenic segment of endarterectomized ICA. Many of these potential problems are recognized only during operation and may require modified or unusual reconstruction techniques to eliminate. The hypothesis of this study was that a complete ICA endarterectomy endpoint and a hemodynamically tapered, smooth, and nonkinked reconstruction minimizes the probability of early postoperative thrombosis, early and late restenosis, and perioperative stroke. Such a CEA reconstruction theoretically minimizes local blood flow turbulence and flow irregularities, and provides a minimally thrombogenic surface. This is a report of a large consecutive series of CEA for which a protocol designed to achieve these goals was followed.

Back to Article Outline

Material and methods 

During the 5-year period, 1986 to 1991, 466 consecutive CEA were performed by the author in a large tertiary care community hospital. All patients were operated upon under general anesthesia. There were 456 primary endarterectomies and 10 (2%) CEA for recurrent stenosis in 408 patients, of whom 231 (57%) were men and 177 (43%) were women. Three male and seven female patients underwent CEA for recurrent stenosis. A second operation was performed on one patient 4 years after the first CEA to correct a 90% common carotid artery stenosis at the proximal end of a saphenous vein patch. Staged bilateral CEA was done on 28 men and 29 women. Patients undergoing ICA bypass from the subclavian, axillary, or contralateral common carotid arteries were not included in this series, nor were patients with ICA occlusion undergoing isolated common or external CEA. The age was 67.5 ± 8.8 years (mean ± 1 SD) (mean 67.2 years for men, 67.8 years for women) with a range of 39 to 90 years. CEA was performed concomitantly with coronary artery bypass in 31 patients and with abdominal aortic surgery in four. Associated diseases and risk factors were coronary artery disease in 71%, hypertension in 64%, and diabetes mellitus in 15%. Indications for CEA were transient ischemic attack (TIA) in 125 (26.8%), amaurosis fugax in 57 (12.2%), reversible ischemic neurologic deficit in 12 (2.6%), stroke in 64 (13.7%), nonlateralizing transient cerebral ischemic symptoms such as syncope, gait disturbance, or dizziness in 61 (13.1%), and >75% diameter stenosis in 147 (31.6%) asymptomatic carotid arteries. In the asymptomatic subset, 54 (37.4%) CEA were staged operations after contralateral CEA for symptomatic disease in 47 patients and asymptomatic >75% stenosis in seven patients. All patients underwent standard or arterial-digital two-view arteriography, with selective carotid artery views in 98%. Percent diameter stenosis was calculated from arteriograms with use of the equation: (1—Minimum internal or common carotid artery lumen diameter/Normal distal ICA lumen diameter) × 100. The stenosis on the operated side was 83.4% ± 13.9% (mean 83.6% for men and 82.9% for women), and that on the contralateral side was 29.7% ± 37.4% (mean 29.5% for men and 29.9% for women). Forty-one (8.7%) patients had contralateral ICA occlusion.

Perioperative management 

All patients had a radial artery cannula placed for on-line systemic arterial pressure monitoring during CEA and the early postoperative period. Systemic systolic arterial pressure was maintained at less than 160 mm Hg during operation and early after the operation with nitroglycerin, nitroprusside, or both. Phenylephrine (Neo-synephrine) was used to keep systemic systolic arterial blood pressure above 100 mm Hg. Before carotid artery occlusion, 5000 units heparin was given intravenously. Heparin effects were reversed with protamine in 20% of CEA when there was excessive intraoperative or early postoperative bleeding. Surgical technique included placement of elastic loops for temporary occlusion of the common carotid artery proximal to the significant plaque, the common trunk of the external carotid artery, and usually the superior thyroid artery. A soft bulldog clamp was used on the ICA unless a shunt was necessary. The ICA was softly pinched on a routine basis with a vascular forceps distal to the plaque seen with angiography to determine the extent of the plaque and to determine whether a distal tongue of atherosclerotic plaque was present. The ICA was exposed 1 to 2 cm distal to the plaque. When necessary, the diagastric muscle was divided, the vagus and hyperglossal nerves were mobilized and retracted, and occasionally the occipital artery was divided and ligated. Special exposure techniques such as physiologic mandibular subluxation were not used. No patient was rejected for CEA because of lack of exposure during the operation or perceived inability to obtain an adequate reconstruction. A shunt was used selectively in 47 (10.1%) of the CEA based on carotid stump back-pressure and collateral cerebral perfusion pressure criteria.11, 12 Stump back-pressure accuracy was confirmed by techniques previously described.12

Management protocol for completion endarterectomy and reconstruction 

No consideration was given for patient sex or ICA diameter when decisions were made about the type of reconstruction. The internal carotid arteriotomy was carried 5 to 10 mm distal to the endarterectomy endpoint to obtain adequate endpoint visualization. This also provides a 360-degree circumferential endothelial surface at the tapered distal end of the ICA if vein patch reconstruction was used, or a partial distal endothelial surface if a synthetic patch was used. The endarterectomy was usually begun in the distal common carotid artery at the junction of the deep media and adventitia. Occasionally the deep media was left in place if it was firmly adherent to the adventitia. Care was taken to develop a deep to shallow media transition plane in the ICA distal to the major plaque bulk, with tapering to the subintima and finally a transition of the endarterectomy to the intima-endothelial film in a smooth manner.

Primary closure was performed when the arteriotomy did not extend distal to the ICA bulb segment as shown in Fig. 1.

When the arteriotomy incision extended distal to the bulb and patch reconstruction was indicated, as illustrated in Fig. 2, autogenous greater saphenous vein was used when adequate and available. Endpoint tacking sutures were used when necessary. When an intact, fixed, circumferential transition of adventitia to media to intima layer could not be obtained, resulting in a circumferential ringlike defect and the possibility of dissection, a circumferential running 6-0 Prolene suture (Ethicon, Inc., Somerville,N.J.) was used to secure the end-point. When a localized long tongue of thick plaque extended distally and could not be removed with optimal exposure, the plaque was cut sharply in cross section 5 to 10 mm proximal to the end of the arteriotomy and a double layer of running 6-0 Prolene was used to secure the cut end. The adjacent proximal endarterectomized adventitia was invaginated circumferentially to cover the cut end, as illustrated in the left panel of Fig. 3.
  • View full-size image.
  • Fig. 3. 

    Method of management of long tongue of thick plaque in ICA with use of inversion tacking sutures over cut tongue (left view) with patch reconstruction (middle views), or saphenous vein interposition graft reconstruction with suturing of vein graft over tongue of plaque (right view).

It is important to recognize this problem before disrupting the intimal and endothelial layers of the tongue of plaque too far distally so that this technique can be used to obtain an adequate endpoint. When a long tongue of plaque could not be removed and when greater saphenous vein was not available or was of too small a diameter for an interposition graft, then either a vein patch with the smaller diameter vein or a synthetic patch was used to reconstruct the artery. The patch was placed 5 to 10 mm distal to the sutured-tacking endpoint of the retained tongue, as shown in the middle panel of Fig. 3. Interposition grafting with autologous reversed greater saphenous vein, if available and adequate, was used to deal with the problem of a long tongue of plaque. The vein was sutured over the tongue, as illustrated in the right panel of Fig. 3. An interposition graft was also used when the endarterectomized ICA adventitia was grossly uneven, exceedingly rough, as can be caused by deep calcification, and seemed to be significantly thrombogenic. When necessary, redundant endarterectomized ICA segments were shortened by transverse-eversion plication to prevent kinking at the distal end of the patch reconstruction (Fig. 4).

When planning vein patch reconstruction with a completed patch width of 7 to 9 mm, it is necessary to trim greater saphenous veins with 6 mm or larger distended diameter. This is done after the veins are opened longitudinally. Autogenous greater saphenous vein was used for patching when the distended diameter was 3.5 mm or larger.13 All greater saphenous veins from female patients were obtained from the thigh and met these criteria. Most saphenous veins from male patients were obtained from below the knee, but 12 below-knee veins were rejected because the diameter was less than 3.5 mm. Saphenous vein rupture pressure was measured in 369 (98%) of the CEA vein patch reconstructions with use of a previously described technique.13 When saphenous vein was either not available or not adequate, either woven Dacron or polytetrafluoroethylene (PTFE; Gore-Tex*patches were used. Synthetic patches were cut 9 to 13 mm wide.

Confirmation and follow-up 

After reconstruction and restoration of blood flow, continuous-wave Doppler ultrasonography was used to interrogate the common, internal, and external carotid arteries. Isolated repair of the external carotid artery was performed when indicated.14 No ICA required revision. All patients were monitored in the recovery room and overnight in an intensive care unit. Neurologic examination was performed by the author in the recovery room and later in the intensive care unit. Perioperative neurologic evaluation was also performed by a neurologist in 110 (27%) patients and by an internist or cardiologist in 207 (51%) patients. In addition, experienced recovery room and intensive care unit nursing personnel routinely performed neurologic assessment. However, it is possible that subtle or transient neurologic deficits were missed. Routine carotid artery duplex scanning was done at 3 months and 6 months after CEA (Ultra Mark 8 or Ultra Mark 9; Advanced Technology Laboratories, Bothell, Wash.). Duplex scanning was performed in the early postoperative period on patients in whom a TIA, stroke, or hyperperfusion syndrome developed. All data were prospectively stored in a computer vascular registry. Descriptive statistics are expressed as mean ± 1 SD. Statistical comparisons are by unpaired t testing.

Back to Article Outline

Results 

The CEA reconstruction methods used are given in Table 1.

Table I. Distribution of the reconstruction methods and ICA endarterectomy endpoint completion
Patch reconstruction
VeinDacronPTEFAllPrimary closureInterposition vein graftTotal
Complete endpoint3522823403286437 (93.8%)
Incomplete endpoint
“Ring”1701180018 (3.9%)
“Tongue”71080311 (2.3%)
Total3762924429 (92.1%)28 (6.0%)9 (1.9%)466 (100%)
Early in this series, PTFE was used when a synthetic patch was necessary, but because of significant bleeding at the suture holes that required blood transfusion in three patients, woven Dacron patches were used more recently. The endarterectomized ICA segment was shortened by eversion plication in 16 (3.4%) CEA, 14 of which had a complete endarterectomy endpoint and patch reconstruction. During operation the measured mean carotid artery stump back-pressure was 46.1 ± 17.9 mm Hg (mean 47.6 mm Hg for mean and 44.2 mm Hg for women, p < 0.05) and the cerebral perfusion pressure (mean back-pressure minus mean jugular venous pressure) was 39.9 ± 18.0 mm Hg (mean 41.5 mm Hg for men and 37.8 mm Hg for women, p < 0.05). The occlusion time was 28.1 ± 6.1 minutes for the 419 CEA done without a shunt. In the 47 CEA in which a shunt was used, the occlusion plus shunt time was 34.3 ± 7.0 minutes. The distended diameter of the saphenous vein was 4.6 ± 0.72 mm for men and 5.0 ± 1.2 mm for women (p < 0.001). The rupture pressure of the saphenous vein13 was 4.3 ± 1.4 atm for men and 3.1 ± 1.3 atm for women (p < 0.0001).

The hospital and 30-day mortality and morbidity rates are given in Table II.

Table II. Mortality and morbidity rates for 466 CEA
Hospital30 Day
No.%No.%
Death
 Cardiac10.230.6
Stroke
Ischemic1*0.21*0.2
Hyperperfusion syndrome, ischemic0—10.2
Hyperperfusion syndrome, hemorrhagic0—10.2
Total20.461.2
Neurologic disease
Transient ischemia20.440.9
Stroke
Ipsilateral ischemic20.420.4
Contralateral ischemic10.210.2
Hyperperfusion syndrome, ischemic0—10.2
Hyperperfusion syndrome, no stroke10.220.4
Other disease
Myocardial infarction30.640.9
Cervical hematoma requiring reoperation30.830.6
Patch disruption0—0—
Cranial nerve injury20.420.4
Leg vein harvest incision, deep infection30.6112.3
*Died of myocardial infarction.
There were two deaths in the hospital; both patients died of myocardial infarction. One of the deaths occurred on postoperative day 26 after a mild motor stroke on postoperative day 2. There were three nonfatal strokes in the hospital. One stroke occurred 18 hours after operation and was completely resolved by 2 months. The second occurred on the first postoperative day and resulted in a mild residual speech and motor deficit. The third stroke occurred 36 hours after concomitant CEA and coronary artery bypass in the contralateral cerebral hemisphere. The contralateral ICA had a 60% diameter stenosis. All four patients who had stroke while in the hospital had normal postoperative duplex scans on the CEA side. The combined hospital mortality and nonfatal stroke rate was 1.1% (5/466) for CEA and 1.2% (5/408) for patients. There were two hospital TIA. One TIA was arm weakness that lasted less than 1 hour and occurred in the recovery room in a patient who underwent operation for crescendo TIA. The other TIA was a brief period of expressive dysphagia on the third postoperative day.

Five patients had hyperperfusion syndrome with severe ipsilateral headache and petit mal or grand mal seizures. In one patient the seizure occurred on postoperative day 4 just before planned discharge from the hospital; no complications ensued. Hyperperfusion syndrome occurred in the other four patients after hospital discharge on postoperative days 5, 5, 6, and 7, respectively. Three patients had a stroke and two died. All five patients had ≥95% diameter carotid artery stenosis before CEA. The interoperative collateral cerebral perfusion pressure for these five patients was 20.4 ± 7.1 mm Hg (p < 0.01 when compared with the other 461 CEA). Two patients underwent magnetic resonance imaging; one scan was normal and one was consistent with ipsilateral hyperperfusion. Three patients underwent computed tomography (CT), which showed one ischemic stroke, one patchy edema, and one hemorrhage, all on the ipsilateral side. All five patients with hyperperfusion syndrome had normal carotid artery duplex scans.

Two deaths from myocardial infarction occurred after discharge from the hospital within 30 days after CEA. The three strokes in patients with hyperperfusion syndrome were the only strokes that occurred after discharge from the hospital within 30 days after CEA. All six ipsilateral strokes occurred in patients with saphenous vein patch reconstruction. The indications for CEA in the three patients who had a stroke and died were TIA in the patient who had a stroke while in the hospital, previous completed stroke in the patient with hyperperfusion syndrome and ischemic stroke, and an asymptomatic 95% carotid artery stenosis in the patient with hyperperfusion syndrome and hemorrhagic stroke. The indications for CEA in the four surviving patients who had a stroke were asymptomatic 90% stenosis but an old ipsilateral stroke on CT in one, previous completed stroke in one, TIA in one, nonlateralizing ischemia in one, and 95% asymptomatic stenosis in the patient with a contralateral stroke and coronary artery bypass. There were no patch ruptures or false aneurysms.

Duplex scans were obtained from 3 to 6 months after 423 (91%) CEA. No ICA occlusions were found. Three arteries (0.6%) had 50% to 75% diameter stenosis at 6 months after CEA, one at the distal end and one at the proximal end of a vein patch and one at the distal end of a PTFE patch. The only ≥75% restenosis in this series was found 4 years after the primary CEA and was corrected with a CEA. Approximately 90-degree kinking of the ICA just past the distal end of one vein and one PTFE patch reconstruction were noted, but neither of these CEA had plication shortening of the ICA. No false or true aneurysms were identified.

Back to Article Outline

Discussion 

The theoretic goal of this approach to CEA was to minimize the probability of early postoperative ICA thrombosis, residual distal ICA stenosis as a result of retained plaque, technically produced stenosis as a result of intima or media endpoint dissection or primary closure, and recurrent stenosis. The hypothesis was that a complete internal CEA endpoint and a tapered, smooth, nonkinked ICA reconstruction with a partial endothelial surface minimize the probability of early postoperative thrombosis and early and late restenosis. The technical goals were achieved most frequently with CEA saphenous vein patch reconstruction. These principles seem to be beneficial in preventing ICA occlusion and minimizing early postoperative stroke and residual or recurrent stenosis. In this series no ICA occlusions and no strokes occurred in the first 18 hours after operation. The three ipsilateral strokes that occurred in the hospital had no abnormality identified on duplex scan. However, the hospital and 30-days mortality and morbidity rates as a result of myocardial infarction and hyperperfusion syndrome remain a major problem.

The use of a patch to reconstruct CEA is as old as the operation itself15 and is the mainstay of the reconstruction methods used in this series. Recommended indications for selective use of a patch include patients with an ICA diameter less than 3 to 3.5 mm, incomplete or difficult ICA endpoints, and repeated operations for recurrent stenosis, as well as female sex.4, 8, 16 This study takes a different approach, namely, to use patch reconstruction if the arteriotomy required to obtain a complete endpoint endarterectomy extends distal to the ICA bulb, and to use a greater saphenous vein patch if it is available and adequate. This protocol resulted in patch reconstruction in 92% of CEA, of which 81% were saphenous vein patches. The choice of patch material remains controversial. Several studies show no significant difference between saphenous vein and synthetic patches in the early thrombosis rate and early and late restenosis rates.7, 9 Patch rupture is a major concern with CEA patch reconstruction utilizing greater saphenous vein. Other reasons for not using greater saphenous vein are leg incision wound problems and the “spare the vein” concept. Most patch ruptures occur in women with small diameter veins obtained from below the knee.3, 6, 17, 18, 19 Veins less than 3.5 mm in diameter may have inadequate mechanical properties to withstand the increased wall stress caused by arterialization and the increased radius of curvature.13 In this study 11 below-knee veins from male patients were rejected for use because of small diameter. This suggests that if one wants to use veins with a minimum 3.5 mm distended diameter, it is advisable to use thigh veins in men as well as women to ensure adequate diameter. All veins from female patients were obtained from the thigh and all had diameters of at least 3.5 mm. No vein patch ruptures occurred using this protocol, which suggests that the use of veins with a minimum 3.5 mm distended diameter has merit.13

There are both theoretic and practical reasons to consider autogenous saphenous vein the patch material of choice. One theoretic advantage of the use of autogenous saphenous vein for a patch is an endothelial nonthrombogenic surface. However, it has been shown that adjacent saphenous vein does not accelerate or aid reendothelialization of endarterectomized carotid artery segments.20 In fact, some question exists whether endothelialization of the endarterectomized segment occurs at all after CEA.21 On the practical side, saphenous vein is easy to use as a patch and can provide an interposition graft if needed. Synthetic patches are more difficult to suture and can result in increased blood loss during operation, particularly from suture holes in PTFE grafts, and may result in unnecessary blood transfusion.

A thick tongue of plaque extending up the ICA can be a difficult management problem. If the plaque is identified before arteriotomy, adequate distal ICA exposure can be obtained without prolonging the occlusion or shunt time. Exposure can be more difficult if the tongue of plaque is found after arteriotomy. Even good-quality, biplane, selective carotid arteriograms frequently fail to identify the presence or extent of a thick tongue of atherosclerotic plaque. Contingency planning for management of this problem is imperative, and when plaque is found, the availability of a good-quality saphenous vein for a patch or bypass is a great help.

There are theoretic advantages of a tapered, smooth, nonkinked ICA reconstruction. Arterial thrombosis occurs in regions of flow irregularities, high shear stress, and local turbulence.22, 23 When ICA thrombosis occurs early after CEA, it is usually supposed that there was a technical problem. However, the endarterectomized carotid artery is known to be thrombogenic,24 and this, plus a focal zone of stenosis or kinking, may be enough to produce thrombosis, particularly if blood flow rate is abnormally high, as can occur in some patients early after CEA.25

Although patch reconstruction has been proposed to reduce the probability of kinking,26 this was not true in this series. If there is a redundant endarterectomized ICA segment, patch reconstruction tends to straighten this out, producing a kink just past the distal end of the patch. This problem can be reduced or eliminated by shortening the endarterectomized ICA segment by eversion plication as described above. However, this potential problem is not easy to identify, because two 90-degree kinks were found on postoperative duplex scans in this series.

High-resolution B-mode ultrasonography and pulsed Doppler velocity spectral criteria are the current standards for short-term and long-term follow-up after CEA. In this study 91% of all CEA and 92% of the 30-day survivors underwent postoperative duplex scanning that showed no ICA occlusions and a 0.7% incidence of residual or recurrent stenosis >50%. These results support the reconstruction methods used.

After CEA, hyperperfusion syndrome is characterized by ipsilateral headache followed by petit mal or grand mal seizures. Patients do well early after CEA except that ipsilateral headache may occur. This is followed by worsening of headache and the onset of seizures on the first to the tenth postoperative days.25, 27, 28 This syndrome may be the result of loss of cerebral autoregulatory mechanisms because of chronic cerebral ischemia.25 Patients typically had a high-grade ipsilateral carotid artery stenosis before CEA. Early postoperative regional cerebral blood flow may be three to four times normal.25 Similar mechanisms have been postulated to cause postoperative intracranial hemorrhage.27 CT scans are usually normal but may show patchy edema25 or hypodense white matter.25 Neurologic deficits can occur, although most are transient.25 Overall the clinical outcome of the five patients in this study seems to be worse than that of other series.25, 27, 28 Four of the patients had an abnormal CT or magnetic resonance imaging scan, two had an ischemic stroke, and one had an intracranial hemorrhage. This experience suggests that this syndrome is far from benign. However, it could be argued that a stroke produced the seizure and that one or more of these patients did not have a true hyperperfusion syndrome. Control of hypertension and avoidance of antiplatelet and anti-coagulant drugs is advisable. The development of severe ipsilateral headache after CEA in a patient with a low carotid artery stump back-pressure and a ≥90% diameter carotid artery stenosis may warrant prophylactic anticonvulsant medication, continued hospitalization, and strict blood pressure control. The presence of a mild ipsilateral headache is not infrequent after CEA but may be an indication for close observation.

Back to Article Outline

Discussion 

Dr. William C. Mackey (Boston, Mass.). Dr. Archie has presented a fine personal series of 466 consecutive CEA performed over a 5-year period. He has presented excellent short-term results, along with the details of his technique for achieving these results. Almost all surgeons who operate on carotid arteries would agree with many of Dr. Archie's technical points. The importance of complete plaque removal with perfect visualization of a smooth endpoint and liberal use of tacking sutures when necessary to smooth the endpoint are basic prerequisites for safe carotid artery surgery in any hands.

Other techniques espoused by Dr. Archie are less uniformly endorsed. Selective shunting based on stump pressures and cerebral perfusion pressures, routine patch closure of arteriotomies extending beyond the carotid artery bulb, liberal use of carotid artery plication, and lack of use of mandibular subluxation to aid distal exposure are among the more controversial aspects of his technical protocol.

Despite these areas of controversy, it is hard to argue with Dr. Archie's short-term results. A 30-day combined stroke/mortality rate of 2%, no carotid artery thromboses, no patch ruptures or pseudoaneurysms, and a 0.7% significant restenosis rate at 6 months are results of which any surgeon would be justifiably proud.

The major question raised by Dr. Archie's report, then, is whether his results justify making this technical protocol a standard for all surgeons who operate on carotid arteries. Should all who use shunts routinely and all electroencephalograph enthusiasts revert to stump pressure measurement? Should all “nonpatchers” become “patchers”? Should we abandon mandibular subluxation?

Although I admire Dr. Archie's achievements as presented in this report and although I believe his techniques merit serious consideration by all surgeons who operate on carotid arteries, I do not believe that his report is sufficiently compelling to raise his techniques to the level of a gold standard. First, there was no independent assessment of neurologic outcome. Second, no long-term follow-up data were presented. And third and most important, the surgeons participating in North American Symptomatic Carotid Endarterectomy Trial (NASCET), who used a variety of techniques, achieved a 30-day major stroke and mortality rate of a little more than 2%, and their patients were subjected to the most meticulous scrutiny by our nonsurgical colleagues. The NASCET short-term results are not different from Dr. Archie's results.

What is the message of your report, Dr. Archie? What would you have us learn from your results?

Shunt use rate as presented in your article is 10% and is somewhat low compared with most reports. Ours is about 17% with electroencephalographic criteria. Exactly what are your criteria for shunt placement? Are any criteria other than stump and cerebral perfusion pressure used?

I would like to know how you managed your small number of postoperative TIA? Did patients undergo CT scanning?

Your patients with hyperperfusion syndrome were noteworthy. We are under increasing pressure to discharge our patients as quickly as possible. A series of patients who underwent endarterectomy and were discharged on post operative day I was presented at the most recent Eastern Vascular Society meeting. How do you identify patients at risk for hyperperfusion syndrome, and what do you do to prevent this devastating complication in patients who are at risk?

Dr. Joseph P. Archie, Jr. The major message is to get a complete ICA endarterectomy endpoint and a hemodynamically sound reconstruction. If you do this and do the other things right, you may get good results. I have routinely used the carotid stump back-pressure and cerebral perfusion pressure methods for more than a decade. I shunt when the back-pressure is less than 25 mm Hg or the perfusion pressure is less than 18 mm Hg. I have not deviated from this protocol even for patients with completed stroke or a contralateral carotid artery occlusion. This criterion produces an 8% to 12% shunt rate. Most electroencephalography criteria have a higher rate of shunting. If a patient had a postoperative transient ischemic attack, I routinely obtain a duplex scan and, if the symptoms recur, a CT scan.

The hyperperfusion syndrome is a perplexing problem. If a patient has a severe ipsilateral headache and a high-grade stenosis was present before CEA, do you administer any convulsant medications? And does it help if you do so? I have asked this of a number of neurologists, and no one seems to know the right answer. If a patient has a bad ipsilateral headache, I now keep them in the hospital, make sure their blood pressure is well controlled, and wait. Literature is lacking in this area and I frankly do not know what the right answer is.

Dr. Norman R. Hertzer (Cleveland, Ohio). I do not disagree with Bill Mackey on many things, but it may be a leap of faith to assume that the stroke risk for CEA across the country is comparable to the NASCET surgical results. The centers that participated in NASCET were very carefully selected to exclude surgical risk as an arbiter of outcome. A 5% perioperative risk in patients with symptoms would be quite acceptable if it were true everywhere, but I do not believe it is. A number of retrospective surveys have reported stroke rates approaching double digits, and that is an important reason that CEA has been in trouble in the past.

I suspect many of these strokes are caused by postoperative thrombosis of the ICA, perhaps involving surgeons who perform relatively few of these operations each year. I believe that patch angioplasty may avoid this problem. Aside from the loss of a segment of vein that might eventually be used for something else, vein patching has a single liability, and that is patch rupture. You have either had none or relatively few of those; how do you select your vein? Do you have any reservations about patching with something other than vein, and if an autogenous patch is unavailable, what material would you select?

Dr. Archie. I have encountered three vein patch ruptures in approximately 400 prior vein patch reconstructions. I believe that it is small-diameter veins that rupture. In this series all of the veins from female patients were obtained from the thigh and were at least 3.5 mm in distended diameter. I still take veins from male patients from below the knee if they seem to be of good size, but I have been disappointed in some veins obtained from male patients and have not been able to use them because they were too small by my criteria. I am moving more toward using greater saphenous veins from above the knee in all patients. When saphenous vein was not available or not adequate, I used thin-walled PTFE patches in the early part of the series, but I had some trouble with bleeding from suture holes. The latter patches were of woven Dacron.

Dr. Frans L. Moll (Nieuwegein, The Netherlands). We have used the transcranial Doppler recordings during CEA for the last 1 ½ years, and in 152 CEA we discovered two patients with a postoperative flow velocity that exceeded 120 cm/sec. In those two patients an intracerebral hematoma developed on the third and the fifth postoperative days, respectively. This may be a direction in which we have to look in future research to prevent postoperative hematomas in those patients.

Dr. James S. T. Yao (Chicago, I11.). Would you care to comment? Do you use the transcranial Doppler?

Dr. Archie. I only use continuous-wave Doppler interrogation. We have also recently started using a transit time Doppler flowmeter. Some information suggests that patients with significantly increased high ICA blood flow, by whatever method you assess it, early after CEA are at increased risk for the hyperfusion syndrome. I believe if you have a patient with a 95% stenosis, a low stump back-pressure that indicates a poor circle of Willis, and a high ICA flow after endarterectomy, you have probably identified a patient who is at increased risk for the hyperfusion syndrome.

Dr. Larry H. Hollier (New Orleans, La.). Hyperperfusion syndrome that may occur after CEA is a definite entity. We previously reviewed our experience with hyperperfusion syndrome, identifying 32 patients in whom cerebral hyperperfusion developed among 2439 patients who underwent CEA. A select group of patients seemed to be at higher risk: those with bilateral high-grade stenoses and those with a unilateral high-grade stenosis and a contralateral occlusion. These patients seem to have a loss of autoregulation of blood flow in the brain, probably because of severe chronic ischemia.

Hyperperfusion syndrome is heralded by severe postoperative headache, often occurring unilaterally. On electroencephalography these patients will be noted to have periodic lateralizing epileptiform discharges. From a practical standpoint, a patient in whom severe nonrelenting headache develops after CEA should have electroencephalography performed; if periodic epileptiform discharges are noted on electroencephalography, these patients should probably be started on anticonvulsant therapy.

It is interesting to note that if one measures blood flow in the operating room at the time of CEA, one may note more than a threefold increase in blood flow after endarterectomy in those patients in whom hyperperfusion syndrome subsequently develops. It may be possible to identify some of these patients by changes identified on transcranial Doppler studies, although we have not yet been able to document this thus far.

Dr. Henry D. Berkowitz (Philadelphia, Pa.). We also have seen this hyperperfusion syndrome, and it is most distressing. If patients have a unilateral headache lateral to the CEA, they are considered to be at high risk for an intracerebral hemorrhage. We use very aggressive measures to control blood pressure. We have found that propranolol hydrochloride (Inderal) is especially effective for correcting the symptoms. The patients are given Inderal and other antihypertensive medications as needed, and they are not discharged until we are assured that their blood pressure is under control and that they have no symptoms with respect to headaches.

Dr. Vincent A. Piccone (Staten Island, N.Y.). Dr. Archie, what were your criteria for using transverse plication to prevent kinking? Did you use preoperative noninvasive tests such as oculoplethysmography to show physiologic need?

Dr. Archie. When I finish the endarterectomy and the ICA segment is floppy and the artery is kinked, tortuous, or semilooped, I shorten it by transverse plication. It adds very little operating time. I have no specific criteria other than just observation of what you have when you finish the endarterectomy.

Back to Article Outline

References 

  1. Stoney RJ, String ST. Recurrent carotid stenosis. Surgery. 1976;80:705–710
  2. Cossman D, Callow AD, Stein A, et al.  Early restenosis after carotid endarterectomy. Arch Surg. 1978;113:275–278
  3. Archie JP. Prevention of early restenosis and thrombosis-occlusion after carotid endarterectomy by saphenous vein patch angioplasty. Stroke. 1986;17:901–905
  4. Eikelboom BC, Ackerstaff RG, Hoeneveld H, et al.  Benefits of carotid patching: a randomized study. J Vasc Surg. 1988;7:240–247
  5. 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;115:972–979
  6. Hertzer NR, Beven EG, O'Hara PJ, Krajewski LP. A prospective study of vein patch angioplasty during carotid endarterectomy: three-year results for 801 patients and 917 operations. Ann Surg. 1987;206:628–635
  7. Lord RSA, Raj TB, Stary DL, Nash PA, Gragam AR, Goh KH. Comparison of saphenous vein patch, polytetrafluoroethylene patch, and direct arteriotomy closure after carotid endarterectomy, part I: perioperative results. J Vasc Surg. 1989;9:521–529
  8. Clagett GP, Patterson CG, Fisher DF, et al.  Vein patch versus primary closure for carotid endarterectomy: a randomized prospective study in a selected group of patients. J Vasc Surg. 1989;9:213–223
  9. Rosenthal D, Archie JP, Garcia-Rinaldi R, et al.  Carotid patch angioplasty: immediate and long-term results. J Vasc Surg. 1990;12:326–333
  10. Ten Holter JBM, Ackerstaff RGA, Thoe Schwartzenberg CWS, 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
  11. Archie JP, Feldtman RW. Determinants of cerebral perfusion pressure during carotid endarterectomy. Arch Surg. 1982;117:319–322
  12. Archie JP. Technique and clinical results of carotid stump back-pressure to determine selective shunting during carotid endarterectomy. J Vasc Surg. 1991;13:319–327
  13. Archie JP, Green JJ. Saphenous vein rupture pressure, rupture stress, and carotid endarterectomy vein patch reconstruction. Surgery. 1990;107:389–396
  14. Archie JP. Management of the external carotid artery during routine carotid endarterectomy. J Cardiovasc Surg. 1991;33:62–64
  15. DeBakey ME, Crawford ES, Morris GC, Cooley DA. Patch graft angioplasty in vascular surgery. J Cardiovasc Surg. 1962;3:106–111
  16. Hans SS. Late follow-up of carotid endarterectomy with venous patch angioplasty. Am J Surg. 1991;162:50–54
  17. Riles TS, Lamparello PH, Giangola G, Imparato AM. Rupture of the vein patch: a rare complication of carotid endarterectomy. Surgery. 1990;107:10–12
  18. O'Hara PH, Hertzer NR, Krajewski LP, Beven EG. Saphenous vein patch rupture after carotid endarterectomy. J Vasc Surg. 1992;15:504–509
  19. Tawes RL, Treiman RL. Vein patch rupture after carotid endarterectomy; a survey of the Western Vascular Society members. Ann Vasc Surg. 1991;5:71–73
  20. Govostis DM, Bandyk DF, Bergamini TM, Towne JP. Biochemical adaptation of venous patches placed in the carotid bifurcation. Arch Surg. 1989;124:490–493
  21. Clagett GP, Robinowitz M, Youkey JR, et al.  Morphogenesis and clinicopathologic characteristics of recurrent carotid disease. J Vasc Surg. 1986;3:10–23
  22. Stain PD, Sabbah HN. Measured turbulence and its effect on thrombus formation. Circ Res. 1974;35:608–614
  23. Goldsmith HL, Turitto VT. Rheological aspects of thrombosis and haemostasis: basic principles and applications. Thromb Haemost. 1986;55:415–435
  24. Dirrenberger RA, Sundt TM. Carotid endarterectomy: temporal profile of the healing process and effects of anticoagulant therapy. J Neurosurg. 1978;48:102–119
  25. Reigel MM, Hollier LH, Sundt TM, Piepgras DG, Shar-brough FW, Cherry KJ. Cerebral hyperperfusion syndrome: a cause of neurologic dysfunction after carotid endarterectomy. J Vasc Surg. 1987;5:628–634
  26. Lord RSA. Carotid endarterectomy. In: Lord RSA. Surgery of occlusive cerebrovascular disease. St Louis: CV Mosby; 1986;p. 227–280
  27. Pomposelli FB, Lamparello PL, Riles TS, et al.  Intracranial hemorrhage after carotid endarterectomy. J Vasc Surg. 1988;7:248–255
  28. Harrison PB, Wong MJ, Belzberg A, Holden J. Hyperperfusion syndrome after carotid endarterectomy. Neuroradiology. 1991;33:106–110
  • * Gore-Tex is a registered trademark of W.L. Gore & Assoc., Flagstaff, Ariz.

☆ Reprint requests: Joseph P. Archie, Jr., MD, 3020 New Bern Ave., Raleigh, NC 27610.

PII: 0741-5214(93)90018-H

doi:10.1067/mva.1993.42068

Journal of Vascular Surgery
Volume 17, Issue 1 , Pages 141-151, January 1993

Article Tools

* Image Usage & Resolution