The clinical course of residual carotid arterial disease R i c h a r d M. Green, M D , JoAnne McNamara, R N - C , MS, RVT, K e n n e t h Ouriel, M D , and James A. DeWeese, M D , Rochester, N.Y. Over a 5-year period 642 patients underwent 686 carotid endarterectomies with patch closure and intraoperative surveillance with continuous-wave Doppler. The perioperative stroke rate was 1.5%. Patients were screened with duplex scans immediately after operation for the presence o f residual carotid lesions, and followed every 3 to 6 months for either the development o f a true recurrent lesion or a change in a residual one. Five hundred thirty-nine arteries (84%) had no postoperative abnormalities. The incidence of recurrent carotid lesions in this groups was 1.5%, 3.4%, and 5.2% at 1, 2, and 3 years, respectively. The incidence o f symptoms in this group was 0,2%, 0.7%, and 1.4% at 1, 2, and 3 years, respectively. The earliest recurrence or s y m p t o m occurred 8 months from operation. One hundred forty-seven arteries had residual lesions that were more c o m m o n w h e n either a temporary shunt was used or the operation was carried out above the hypoglossal nerve or below the o m o h y o i d muscle. Sixty-one patients w h o had plaque proximal to the arteriotomy without a significant stenosis were followed an average o f 21 months. There were no changes in plaque morphology and no proven symptoms
related to the residual lesion. Fifty-six patients with both plaque and significant hemodynamic abnormalities in the carotid bulb were followed an average of 18 months. Seven of these patients (12.5 %) had either a significant deterioration of the lesion or a s y m p t o m from it. Each event occurred within 6 months of operation. Thirty patients had significant flow abnormalities but no visible plaque. N o n e o f these lesions deteriorated. Although c o m m o n , residual carotid lesions are benign unless the lesion is characterized by both plaque within the artery and a hemodynamically significant stenosis. These lesions should be further investigated and treated w h e n discovered. (J VAS¢ SURG 1991;13:112-20.)
The incidence of recurrent carotid stenoses after endarterectomy varies from 1.7% to 42% depending on the definition of recurrence, the method of surveillance, and the time interval from operation. 1 A significant percentage of recurrent lesions have been found to actually be present immediately after operation and therefore should be considered residual rather than recurrent, z Residual lesions are due either to technical errors or to inadequate plaque removal. Major technical errors are usually recognized at the time of operation, and if not, they lead to serious neurologic morbidity? The minor technical errors are not easily recognized without sophisticated intraoperative imaging and may not cause any increase in perioperative morbidity rates. Even if recognized, residual lesions may be left despite the best efforts From the Universityof Rochester MedicalCenter Rochester. Presented at the Forty-fourthAnnualMeeting of the Societyfor Vascular Surgery, Los Angeles, Calif.,June 4-6, 1990. Reprint requests: Richard M. Green, MD, 125 Lattimore Rd., Rochester, NY 14620. 24/6/25558 112
of the surgeon. Atherosclerotic disease may extend more proximal than the operator can conveniently reach or the distal end point may require tacking sutures for residual plaque. This study attempts to define the natural history of these abnormalities left behind knowingly or unknowingly and recognized shortly after operation. METHODS From January 1984 to July 1989, 642 patients underwent 686 carotid endarterectomies at the University of Rochester Medical Center. Four hundred eleven patients (60%) were operated on for transient ischemic attacks or amaurosis fugax. One hundred twenty-four patients (18%) were operated on fbr stroke, 69 patients (10%) for vertebrobasilar symptoms, and 82 patients (12%) for hemodynamically significant asymptomatic stenoses. Patients were monitored during surgery with continuous 12-lead electroencephalographic analysis with a neurologist in attendance. Temporary indwelling shunts were placed when the baseline elec-
Volume 13 Number 1 January. 1991
troencephalogram was abnormal in the ipsilateral hemisphere, when abnormalities occurred after carotid clamping, or when the indication for operation was an acute stroke? Attention was paid to the presence of unusual technical factors at the time of operation, and these were noted in the operative report. Of particular interest was the presence of atherosclerotic disease below the bulb requiring division of the omohyoid muscle, the presence of disease extending above the hypoglossal nerve requiring mobilization of the nerve and division of the digastric muscle, and the need for tacking sutures. These were considered to be "nonroutine" or extended operations. Arteriotomies were closed with knitted Dacron velour patches. All patients had intraoperative continuouswave Doppler spectral analysis and revision of the closure if significant abnormalities were detected. Intraoperative studies were performed by the operating surgeon using a 9.8 MHz Doppler (model 811; Parks Medical Electronics, Inc., Aloha, Ore.). Postoperative studies were performed within 4 weeks of operation with a duplex scanner (Advanced Technology Laboratories; Bothell, Wash.). A 5 MHz pulsed Doppler probe placed at center stream and held at a 60 degree angle with respect to the vessel axis was used. Internal carotid artery stenoses were graded in six categories: (A) Normal peak systolic frequency 50%. Residual carotid atherosclerotic disease was categorized as type I: plaque proximal to the carotid bulb with spectral broadening of the Doppler signal and peak systolic flows less than 4.0 kHz; type II:
Carotid arterial disease 113
plaque proximal to the carotid bulb with peak systolic flows of greater than 4.0 kHz, or type III: no plaque seen but hemodynamic changes similar to type II lesions. All patients with residual carotid atherosclerotic disease were followed at 3-month intervals for 1 year and then every 6 months. Patients with normal postoperative studies were seen at 6 months, 1 year, and 2 years unless a contralateral lesions of significance was present. Arteriograms were obtained when a significant change in the lesion in question occurred or when the patient complained of symptoms in the appropriate cerebral hemisphere or eye. Data analysis. The major variables examined in this report were the presence of hemodynamic abnormalities or plaque or both immediately after operation, neurologic morbidity in both the perioperative and postoperative periods, mortality rates, shunt use, and extent of operation above and below typical sites on the carotid artery. Life-table analysis was used to determine the effects of plaque at the end of operation on long-term mortality rates. These data were compared with the log-rank test. s RESULTS Operative results. Intraoperative revisions were performed on 17 arteries (2.5%) because of abnormalities in the ultrasonic spectrum. Three of these were due to intimal or atherosclerotic flaps at the distal end point of the endarterectomy and were treated by extending the arteriotomy, removing the flap, and repatching. The other 14 reopenings were due to abnormalities in the external carotid artery. These were treated by continuing the dissection up to the lingual artery and making a small transverse arteriotomy at this point. Every carotid problem was due to inadequate plaque removal. The diagnosis of external carotid occlusion was obvious from palpation of the artery, whereas the internal carotid artery abnormalities were not clinically apparent. Neurologic morbidity after these 686 operations was 1.5%. Ten patients either awoke with a neurologic deficit or a deficit developed within the early postoperative period. The recovery room OPG-Gee was normal in all 10 patients. Duplex scanning or immediate arteriography was performed within the first week in 8 of the 10 patients. Two of these patients had significant residual disease (type II). Neurologic events developed in both several days after operation. One patient had a severe residual stenosis proximal to the endarterectomy site with a probable clamp injury. At operation the presence of plaque but not the severity of it was recognized. The other
Journal of VASCULAR SURGERY
114 Green et al.
6% I ~204 .
•
_
.:/
32 128
/ o%i 0
10
20
30
40
Follow-up (months) Fig. 1. Life table shows the incidence of true recurrent disease. These patients had normal duplex scans immediately after operation and were studied serially by use of the same imaging technique and examiner. At 24 months the incidence of true recurrence is 3.4%.
Fig. 3. A, Type I plaque. The residual plaque appears as a shelf proximal to the endarterectomy site. B, Doppler spectrum shows peak flow velocities less than 4.0 kHz, and is typical of the flows associated with the common carotid shelf. patient did not have obvious residual disease at the time of operation. Six of these deficits persisted after hospital discharge, and four resolved completely. No recognized immediate postoperative internal carotid artery occlusions occurred in the neurologically normal patients. Two patients were immediately returned to the operating room because they were bleeding. The normal g r o u p (no residual disease). Five hundred thirty-nine arteries (84%) had no residual atherosclerotic disease or significant hemodynamic abnormalities at the operative site (categories A and B). The operation was carried out either above th.-
digastric muscle or below the omohyoid muscle in 86 of these patients (16%). Shunts were used during 60 operations (11%). These patients were followed for 23 + 0.8 months. Eight patients (1.5%) had neurologic events during follow-up. Five of these patients had symptoms in the opposite hemisphere; two had ipsilateral lacunar infarcts, and one suffered an ipsilateral intracranial hemorrhage. Fifty-three patients died (8%) during the period of observation, but only one died of a stroke (the patient with the intracranial bleeding). Twenty-one (3.8%) of these patients exhibited a significant change in ultrasonic pattern of the operated artery during follow-up (Fig.
Volume 13 Number 1 lanuary 1991
Carotid arterial disease 115
1.6%
204
1.4%
f
1.2%F
0.8% 0.6% I
i
0.4%I
493 S - J
J
10
0
i
~
20
30
40
Follow-up (months) Fig. 2. Life table shows the incidence of symptoms from recurrent carotid stenoses. At 24 months the incidence of symptomatic recurrence is 0.7%. No patient had a neurologic symptom within 8 months of operation.
¢
B
C
Fig. 4. A, Doppler spectrum shows peak diastolic flows of greater than 4.5 kHz. The realtime image showed soft plaque in the vessel. These changes were present in both patients whose angiograms are shown below. B, Arterogram shows the soft material filling the internal carotid artery. 4C, The same phenomenon at 1 month after operation in the common carotid artery. The external carotid artery has also occluded.
1.) No patient whose condition was normal immediately after operation developed changes earlier than 8 months after operation. The incidence of true recurrence calculated by life-table methods was 1.5% at 12 months, 3.4% at 24 months, and 5.2% at 36 months. The incidence of symptomatic recurrence in these same arteries was 0.2% at 12 months, 0.7% at 24 months, and 1.4% at 36 months (Fig. 2).
Residual disease type I: proximal plaque, no stenosis. Sixty-one patients had residual plaque appearing as a shelf immediately proximal to the arteriotomy (Fig. 3, A). None of these patients had Doppler abnormalities other than mild spectral broadening with peak systolic flows less than 4.0 kHz (Fig. 3, B). Twenty-one of these patients had extension of their lesion proximally on their preoperative
116
Journatof VASCULAR SURGERY
Green et al.
Table I. Summary of the groups and relevant factors
Normal Type I: CCA plaque Type II: CCA plaque + stenosis Type III: no plaque stenosis
Frequency
Follow-up mo +- SEM
Progression of lesion
Neurologic cymptoms ~ ~
Deaths
539 (79%) 61 (9%)
23 -+ 0.8 21 -+ 2.2
21 (3.8%) 0
8 (1.5%) 1 (1.6%)
53 (8%) 5 (8.2%)
56 (8%)
18 _+ 2.2
4 (7%)
5 (9%) *
8 (14%)
30 (4%)
22 + 3.5
0
0
0
CCA, contralateral carotid artery. ~p < 0.05 when compared to all other groups. **Neurologic symptoms: in the normal group five were in contralateral hemisphere, two were ipsilateral lacunar infarcts, and one was an ipsilateral hemorrhagic infarct. In t-ype I the event was a series of ipsilateral sensor), transient ischemic attacks. In group II all five were ipsilateral and related to the operated artery.
arteriogram. The operation was extended below the omohyoid muscle in 16 (26%) of these patients. Shunts were used in 18 patients (30%). These patients were followed for 21 -+ 2.2 months. No signiticant changes occurred in plaque morphology or flow characteristics. Transient, ipsilateral neurologic deficits occurred in one patient ( 1.6%) 7 months after operation, and an arteriogram was obtained. A nonstenotic waist was seen at the proximal extent of the patch, but no obvious source for the emboli was identified, and the patient was placed on warfarin anticoagulation for control of symptoms. Five of these patients died during this interval. Residual disease type II: proximal plaque with stenosis. Fifty-six patients had both plaque and significant hemodynamic abnormalities at the carotid bulb characterized by peak systolic frequencies greater than 4.0 kHz. Three patients had severe changes as seen in Fig. 4, A. Two of these patients had angiograms (Figs. 4, B and C). The operation was carried out below the omohyoid muscle in 22 patients (39%). Shunts were used in 16 patients (29%). These patients were followed for 18 -+ 2.2 months. During that period one patient had a minor stroke, three patients had transient ischemic attacks and one patient had a retinal stroke. Two patients with symptoms and two others without symptoms had significant morphologic changes in their lesions. All seven patients underwent reoperation. Four of the patients had thin layers ofclo~ and atherosclerotic debris loosely adherent to the arterial wall. When the debris was removed the endarterectomized surface appeared normal. Each of these patients had an embolic event. One patient had a discrete atherosclerotic plaque with ulceration proximal to the prior endarterectomy, and the other two patients had nonthrombotic, fibrotic changes at the carotid bifurcation at the site of the septum dividing the internal and ex-
ternal carotid arteries and extending upward into the internal carotid artery. Eight of these patients died during the follow-up period. Residual disease type III: no visible plaque but significant stenosis. Thirty patients had significant flow abnormalities consisting of peak systolic flow velocities of greater than 4.0 kHz. No visible plaque was identified in these patients, but eight had difficult studies because of neck size, and nine patients had kinks at the distal end of the artery. Positive OPGGee measurements were obtained in 12 patients, the test could not be done in 7 patients, and 11 were normal. Four of these patients had operations beyond the digastric muscle (13%), and seven others had tacking sutures placed at the distal end of the arteriotomy for control of the end point of the endarterectomy. These patients were followed for 22 -+ 3.5 months. During that period none of the lesions changed, no one suffered any neurologic symptoms, and no one died. Table I summarizes factors relevant to each group of arteries. Development o f symptoms caused by residual disease. Seven patients had neurologic symptoms during the period of follow-up (Fig. 5). Symptoms developed in two patients while in hospital, both resolved within 3 weeks. The remaining four patients had symptoms within 7 months of operation. Effect o f shunting. Shunts were used in 102 operations (15%). No difference was found in the neurologic outcome between shunted and nonshunted patients. The incidence of residual abnormalities was significantly higher in the group of patients that required shunting. Shunts were used in 60 of the 539 (11%) operations with normal anatomic and hemodynamic outcomes. (Table II). One hundred forty-seven patients had residual abnormalities, and shunts were used in 42 of these operations (29%), p < 0.05. This increase in residual disease
Volume 13 Number 1 January 1991
Carotid a~eriald#eme
117
0.05
0.04
- /
.........
123
48
0.03
0.02
/
0.01
0
--
-
~
0
5
3 _ _ _
10
15
20
25
Follow-up (months) Fig. 5. Life table shows the development of symptoms in patients with residual carotid lesions. M1 seven patients had problems within 7 months of operation.
Table II. The effect o f temporary shunting on residual disease Carego~y No residual disease Residual disease Type I Type II Type III
Tempora~ shunting 60 42 18 16 8
(11%) (29%)* (30%) (29%) (27%)
Table III. The effect o f a non routine operation on residual disease Category No residual disease Residual disease Type I Type II Type III
Non routine operation 86 49 16 22 11
(16%) (33%)* (26%) (39%) (37%)
*p < 0.05 when residual groups are compared to no residual group,
*p < 0.05 when residual groups are compared to no residual group.
after using a shunt was observed equally in all three groups with residual disease. Effect o f a " n o n r o u t i n e " operation o n residual disease. Operations were carried out below the omohyoid muscle, above the hypoglossal nerve, or with tacking sutures to secure the distal end point in 135 arteries (Table III). Eighty-six o f the 539 normal arteries (16%) underwent a "nonroutine" operation. Forty-nine o f the 147 arteries (33%) with residual disease had "nonroutine" operations. This difference is statistically significant, p < 0.05.
ences in the definition o f recurrence, the surveillance techniques used, and the time interval arteries are studied after operation. Many factors such as patching, shunting, gender, tacking sutures, and artery size have been shown both to influence and not to influence the development o f a recurrent carotid stenosis. 7"s'2 We separated 686 endarterectomized carotid arteries into those with residual disease and those without, and followed them serially with duplex scans for an average o f 2 years. We found that residual and recurrent carotid stenoses are each distinct clinical entities. Residual carotid lesions may be left despite the best efforts o f the surgeon. The most common reasons for residual lesions in our series were the socalled extended or nonroutine operation and the use o f temporary indwelling shunts. Residual lesions were more than twice as common in these settings compared to a standard operation without shunting. The physical presence o f the shunt makes visualization o f the end points more difficult and therefore
DISCUSSION Technical errors that result in incomplete endarterectomy have been incriminated as a cause ofneurologic morbidity after operations on the carotid artery. 3 N o w some claim that leaving residual disease is also a major cause o f recurrent carotid stenosis. 2'6 The topic o f the recurrent carotid stenosis is a confusing one. The wide variation in incidence among the numerous studies on this entity is due to differ-
7OUrpk2[ O ~"
118
Green et al.
impedes complete plaque removal. Our primary concern was removal of all distal plaque, but we often knowingly left residual disease proximally when extending the dissection presented problems. The use of tacking sutures for unsatisfactory end points also leaves residual disease and if possible should be avoided if further distal dissection is possible. We used continuous-wave Doppler analysis to assess flow patterns and to repair major technical errors. We recognized three internal carotid arteries with problems and repaired them. Although this technique may be criticized for being totally dependent on the operator's abili~ to recognize audible patterns, it was successful in avoiding any occlusions of the internal carotid artery and resulted in a consistently low stroke rate in a large series of patients. Barnes et alJ report a similar satisfactory experience with this technique. More sophisticated and expensive intraoperative imaging techniques, such as duplex scanning or arteriography, can identi~ residual disease, but the proper approach to those abnormalities not causing hemodynamic alterations is not clear. Flanigan et al. 9 were the first to use the duplex scanner in this setting. They found a 28% incidence of intraluminal defects, but repaired only 25% of them. No adverse clinical sequelae occurred as the stroke rate was 3.8% with a normal ultrasound and 3.3% with an abnormal study. In a follow-up report of the unrepaired arterial defects Sawchuk et a1.1° found no statistical relationship between an unrcpaircd arterial defect and subsequent recurrent carotid artery stcnosis. On the other hand, Reilly ct al, 2 found that significant, unrepaired technical defects were a major factor in the development of recurrent carotid stenoses. Defects causing a stcnosis of greater than 41% or having a flap length of greater than 26% of the lumen size were associated with the development of recurrent lesions. Hypertension, smoking, and cholesterol levels, however, also played a major role in this series. Bandyk et al. 6 used pulsed Doppler spectral analysis and concluded that moderate to severe flow disturbances were associated with a higher incidence of recurrent disease over a 2-year period. We would maintain that these phenomena were not true recurrenccs but represented residual disease that progressed in an atherogenic environment. When the pathologic make up of the recurrent carotid lesion was examined by Clagett et al. H most of the lesions were located within the endarectomized segment, 65% of the lesions were located beyond the origin of the internal carotid artery, and most were associated with a long primary artcriotomy. Ninety
VA~Ct LAI~ S1 ~RGER'~
percent of the arteries had surface or intraplaquc thrombus, which formed the bulk of the lesion in many patients. These lesions were identical to our type II group. The high frequency of thrombus tbrmation on the endarterectomized surface may explain why other investigators report that early "recurrences" regressJ 2 Wc did not find regression of residual lesions, but did find thrombus in those that progressed or became symptomatic. O'Donncll et al. ~3 studied the same problem and found that the earliest "recurrent" lesions were located in the common carotid artery and actually represented residual disease similar to our type I lesions. Clagctt did describe three cases of "retained shelves" like this as an exception to the rest of his cases. A strong desire exists to put the entire problem of carotid stenoses after endarterectomv into a continuum of changes as a result of local healing factors acting in concert with systemic risk factors. There is abundant data that show that recurrent disease is the result of a progression of events beginning with ongoing thrombogcnesis occurring at the cndartercctomy site. Organization of this thrombus and proliferation of smooth muscle cells result in neointimal hyperplasia that with time undergoes atherosclerotic degeneration. 11 Our series presents no evidence to support or refute that hypothesis. Instead, we offer data that suggest that true recurrence starts somewhere beyond 8 months from operation, and at 3 years occurs in 5.2% of arteries that were normal immediately after operation. This incidence is considerably lower than others have reported when residual lesions were not excluded from the analysis and similar to those when residual disease was excluded. 1"14Furthermore, the incidence of symptoms from these true recurrences is even rarer. At 3 years only 1.4% of patients who were healthy after operation developed symptoms from their operated arteries. It is important not to confuse recurrent lesions with residual ones. Like recurrent lesions, most residual lesions arc benign. The common carotid shelves with minimal flow disturbances do not progress and rarely cause symptoms. We do not believe that extensive proximal mobilization and a long arteriotomy is necessary to remove nonstenotic, nonulcerated common carotid plaque. The one case of recurrent symptoms we observed in this type of patient probably came from somewhere else. The lesions with significant flow disturbances but no visualized plaque are also benign. We routinely patch and do not hesitate to use tacking sutures when a perfectly feathered end point cannot be achieved. If
Volume 13 Number 1 Januaq, 1991
w e are confident that t h e e n d p o i n t is secure, w e are n o t b o t h e r e d b y flow velocity elevations after surgery. T h e velocity elevations are usually kinks o r abnormalities caused b y a distal plaque that has been tacked d o w n . T h e o n l y residual lesion that is o f concern is the o n e w i t h b o t h plaque (usually soft) and significant flow disturbances. W e are h o p e f u l that these lesions will be r e c o g n i z e d in the o p e r a t i n g r o o m . W e c o n t i n u e to believe t h a t the c o n t i n u o u s w a v e - D o p p l e r analysis is a d e q u a t e for this surveillance since s y m p t o m s a n d p r o g r e s s i o n usually occur w i t h i n 6 m o n t h s o f o p e r a t i o n . These patients should u n d e r g o a r t e r i o g r a p h y w h e n this t y p e o f lesion is identified, a n d the lesion s h o u l d be repaired.
Carotid arterial disease 119
6.
7. 8.
9. 10.
REFERENCES
1. Barnes RW, Nix ML, Nichols BT, Wingo JP. Recurrent versus residual carotid stenosis. Incidence detected by Doppler ultrasound. Ann Surg 1986;203:652-60. 2. ReilJy LM, Okuhn SP, Rapp JH, et al. Recurrent carotid stenosis: a consequence of local or systemic factors? The influence of unrepaired technical defects. I VASCSURG 1990; 11:448-60.
3. Callow AD. Restenosis after carotid arteu surgeqT.Arch Surg 1982;117:1082-5. 4. Green RM, Messick WJ, Ricotta JR, et al. Benefits, shortcomings and costs of EEG monitoring. Ann Surg 1985; 201:785-92. 5. Anderson S, Auquier A, Hauck W, et al. Techniques for
DISCUSSION Dr. R o b e r t Barnes (Little Rock, Ark.). Dr. Green and his colleagues have expanded on a subject first reported by Dr. Blaisdell, who used intraoperative and postoperative arteriography to identify perioperative and postoperative carotid defects. From the present study, we learned that 10 patients with perioperative neurologic deficits were not operated on again because of normal OPG studies, yet two proved to have significant residual common carotid stenosis on subsequent carotid duplex examination. Second, we learned that in their hands significant residual carotid stenosis correlated with the use of a shunt. Third, we learned that residual disease also correlated with a nonroutine operation, which extended below the omohyoid, above the digastric, or required tack down sutures. Fourth, we learned that residual stenosis was more likely to cause postoperative neurologic deficits if it was proximal than if it was distal. Finally, we learned that significant residual carotid stenosis was detected in the early postoperative period in 12% of their patients despite the use of intraoperative
11. 12. 13.
14.
bias reduction. Statistical methods for comparative studies. New York: John Wiley and Sons, 1980:205-11. Bandyk DF, Kaebnick HW, Adams MB, et al. Turbulence occurring after carotid bifurcation endarterectomy: a harbinger of residual and recurrent stenosis. J VASC SURG 1988; 7:261-74. Ouriel K, Green RM. Clinical and technical factors influencing recurrent carotid stenosis and occlusion after endarterectomy. J VASCSURG 1987;5:702-6. Ackroyd N, Lane R, Appleberg M. Carotid endarterectomy. Long-term follow-up with specific reference to recurrent stenosis, contralateral progression, mortality,, and recurrent neurological episodes. J Cardiovasc Surg 1986;27:418-25. Flanigan DP, Douglas DJ, Machi J, et al. Intraoperative ultrasonic imaging of the carotid arte~ during carotid endarterectomy. Surgery, 1986;100:893-9. Sawchuk AP, Flanigan DP, Machi J, et al. The fate of unrepaired minor technical errors detected by intraoperative ultrasound during carotid endarterectomy. J VASC SURG 1989;9:671-6. Clagett CG, Robinowitz M, Youkey Jr et al. Morphogenesis and clinico-pathologic characteristics of recurrent carotid disease. J VASCSURG 1986;3:10-23. Zierler RE, Bandyk DF, Thiele BL. Intraoperative assessment of carotid endarterectomy. J VASCSURG 1984;1:73-83. O'Donnell TF Jr, Callow AD, Scott G, et al. Ultrasound characteristics of recurrent carotid disease: hypothesis explaining the low incidence of symptomatic recurrence, j"VASC SURG 1985;2:26-41. Blaisdell FW, Lim R Jr, Hall JD. Technical result of carotid endarterectomy. Am J Surg 1967;114:239-46.
Doppler monitoring. This is higher than the 1% incidence noted by Blaisdell or the two percent in our study. This newfound knowledge prompts me to ask the following four questions. First, do you plan to change your method of invasive evaluation of patients with perioperative neurologic deficits? Second do you plan to change your method o f endarterectomy in patients who require shunts? Third, do you plan to change your approach to patients requiring nonroutine, more extensive proximal or distal dissection? Fourth and finally, how do you propose to identify and correct these severe residual carotid lesions during surgery to reduce the 12% incidence of early postoperative disease, which in seven patients led to reoperation? Dr. R o b e r t Courbier (Marscillc, France). In our group we used routinely pcrioperative angiography since 1982 and during that period performed more than 1800 perioperative angiograms. In 4% o f cases we had to correct a technical fault, and in 5.5% we accepted residual disease. Over the past 3 years we began a prospective study on 280 patients who accepted long-term fbllow-up with du-
120
Green et al.
plex scanning every 6 months and a digital subtraction angiogram 1 year after operation. Fifteen patients had accepted residual disease. Among those who had internal carotid artery stenosis less than 50%, two developed an occlusion with transient ischemic attack, two remain asymptomatic. And three, who had transient ischemic attack were reoperated for high-grade stenosis. On the common carotid artery the shelf aspect disappeared in six cases and remained unchanged and asymptomatic in two cascs. Dr. Green, I want to underline that the pathologic significance o f this residual disease has to be known and that chiefly now is when noninvasive techniques, like angioscopy, demonstrate an increasing number o f technical faults without any clinical consequences in the future. Dr. Charles Branfigan (Denver, Colo.). A group of patients in your study had velocity elevations but no changes in the gray scale image. Do you have any angiographic confirmation that this indeed represents a stenosis? Is it possible that this velocity elevation is some kind o f an artifact introduced by surgery or introduced by the change in Doppler angle caused by a change in position o f the artery after surgery? Dr. Wesley M o o r e (Los Angeles, Calif.). I have just one other comment. I think the information that Dr. Green has shared with us today with respect to residual stenosis in the common carotid artery is extremely valuable in that he has pointed out that this is a lesion o f relatively low morbidity. If that were the only location for residual stenosis, I think we would be in great shape. Unfortunately, there are other sites such as the distal end point in the proximal internal carotid artery. More recently, we have identified a problem with residual lesion in the external carotid artery. We have now gathered three patients who have developed thrombus in association with the flap o f the external carotid artery with retrograde propagation and subsequent embolization in the internal carotid. Therefore, I believe that intraoperative surveillance deserves more attention. Beyond just looking for a proximal lesion, it is more important that we look for the distal lesions. I would recommend the use o f some sort o f intraoperative surveiUance for this particular problem, such as completion angiography or B-mode ultrasound. Dr. John R i c o t t a (Buffalo, N.Y.). Dr. Green, it seems to me that the problem you are having is failure to detect or treat proximal disease. I wonder if you could tell us whether this is a function o f your preoperative angiography, perhaps more use o f intraarterial digital subtraction angiography versus cut films, and whether you went back and looked. I know these patients are studied on duplex scanning before surgery. Could you have predicted by the
Journalof VASCULAR SURGERY
preoperative duplex scanning whether there was more common carotid disease than we saw, and would you suggest that we spend a little more time and effort looking at the common carotid as well as the bifurcation? Dr. Green. Dr. Barnes, I cannot explain the differences between our residual stenosis rate o f 12% and your incidence and Dr. Blaisdell's incidence of 1% to 2%. Some o f this may reflect selection criteria since many o f these patients had extended operations, and some may represent your skills with the continuous-wave Doppler examination. Since our overall stroke rate after operation is quite low we do not plan any major changes in the way we evaluate and treat these patients. We are not prepared to routinely use intraoperative duplex scanning or arteriography. We will certainly be more attentive to leaving proximal disease when we use a shunt. We will be more aggressive with patients in the early postoperation period with both residual plaque and significant flow velocity, elevations. I appreciate Dr. Courbier's remarks. I wish that our operating room had fixed angiographic equipment that would allow us to easily study our patients. Our intraoperative carotid studies leave much to be desired and would not with our current equipment provide us with useful information about hemodynamically insignificant residual disease. Dr. Brantigan's remarks arc very valid. I think that the 30 patients who had velocity elevations alone with no symptoms and no progression probably to a largc degree represent kinks and difficult studies. This is a very soft group o f patients and I would not make a whole lot of" conclusions about them. If we do not sec a filling defect we are not concerned about the vclocity elevation. Dr. Moore, we have not seen symptoms from external carotid stenoses but have ccrtainly seen a number o f these lesions. I think that the reason we are not seeing internal carotid residual stenoses is because we are concerned with going high enough on thc internal carotid to remove all disease. We have not been as compulsive going proximally, and perhaps if we did our ovcrall incidence o f residual lesions would be lower. Dr. Ricotta, we looked at the arteriograms of every patient to see whether we could have predicted lcaving a residual stenosis and we could not. However some patients had disease extending down the common carotid artc~, and we made every effort to remove it all. Other patients whose angiograms did not raise our suspicions had more disease than we could remove and were left with residual lesions.
related to the residual lesion. Fifty-six patients with both plaque and significant hemodynamic abnormalities in the carotid bulb were followed an average of 18 months. Seven of these patients (12.5 %) had either a significant deterioration of the lesion or a s y m p t o m from it. Each event occurred within 6 months of operation. Thirty patients had significant flow abnormalities but no visible plaque. N o n e o f these lesions deteriorated. Although c o m m o n , residual carotid lesions are benign unless the lesion is characterized by both plaque within the artery and a hemodynamically significant stenosis. These lesions should be further investigated and treated w h e n discovered. (J VAS¢ SURG 1991;13:112-20.)
The incidence of recurrent carotid stenoses after endarterectomy varies from 1.7% to 42% depending on the definition of recurrence, the method of surveillance, and the time interval from operation. 1 A significant percentage of recurrent lesions have been found to actually be present immediately after operation and therefore should be considered residual rather than recurrent, z Residual lesions are due either to technical errors or to inadequate plaque removal. Major technical errors are usually recognized at the time of operation, and if not, they lead to serious neurologic morbidity? The minor technical errors are not easily recognized without sophisticated intraoperative imaging and may not cause any increase in perioperative morbidity rates. Even if recognized, residual lesions may be left despite the best efforts From the Universityof Rochester MedicalCenter Rochester. Presented at the Forty-fourthAnnualMeeting of the Societyfor Vascular Surgery, Los Angeles, Calif.,June 4-6, 1990. Reprint requests: Richard M. Green, MD, 125 Lattimore Rd., Rochester, NY 14620. 24/6/25558 112
of the surgeon. Atherosclerotic disease may extend more proximal than the operator can conveniently reach or the distal end point may require tacking sutures for residual plaque. This study attempts to define the natural history of these abnormalities left behind knowingly or unknowingly and recognized shortly after operation. METHODS From January 1984 to July 1989, 642 patients underwent 686 carotid endarterectomies at the University of Rochester Medical Center. Four hundred eleven patients (60%) were operated on for transient ischemic attacks or amaurosis fugax. One hundred twenty-four patients (18%) were operated on fbr stroke, 69 patients (10%) for vertebrobasilar symptoms, and 82 patients (12%) for hemodynamically significant asymptomatic stenoses. Patients were monitored during surgery with continuous 12-lead electroencephalographic analysis with a neurologist in attendance. Temporary indwelling shunts were placed when the baseline elec-
Volume 13 Number 1 January. 1991
troencephalogram was abnormal in the ipsilateral hemisphere, when abnormalities occurred after carotid clamping, or when the indication for operation was an acute stroke? Attention was paid to the presence of unusual technical factors at the time of operation, and these were noted in the operative report. Of particular interest was the presence of atherosclerotic disease below the bulb requiring division of the omohyoid muscle, the presence of disease extending above the hypoglossal nerve requiring mobilization of the nerve and division of the digastric muscle, and the need for tacking sutures. These were considered to be "nonroutine" or extended operations. Arteriotomies were closed with knitted Dacron velour patches. All patients had intraoperative continuouswave Doppler spectral analysis and revision of the closure if significant abnormalities were detected. Intraoperative studies were performed by the operating surgeon using a 9.8 MHz Doppler (model 811; Parks Medical Electronics, Inc., Aloha, Ore.). Postoperative studies were performed within 4 weeks of operation with a duplex scanner (Advanced Technology Laboratories; Bothell, Wash.). A 5 MHz pulsed Doppler probe placed at center stream and held at a 60 degree angle with respect to the vessel axis was used. Internal carotid artery stenoses were graded in six categories: (A) Normal peak systolic frequency 50%. Residual carotid atherosclerotic disease was categorized as type I: plaque proximal to the carotid bulb with spectral broadening of the Doppler signal and peak systolic flows less than 4.0 kHz; type II:
Carotid arterial disease 113
plaque proximal to the carotid bulb with peak systolic flows of greater than 4.0 kHz, or type III: no plaque seen but hemodynamic changes similar to type II lesions. All patients with residual carotid atherosclerotic disease were followed at 3-month intervals for 1 year and then every 6 months. Patients with normal postoperative studies were seen at 6 months, 1 year, and 2 years unless a contralateral lesions of significance was present. Arteriograms were obtained when a significant change in the lesion in question occurred or when the patient complained of symptoms in the appropriate cerebral hemisphere or eye. Data analysis. The major variables examined in this report were the presence of hemodynamic abnormalities or plaque or both immediately after operation, neurologic morbidity in both the perioperative and postoperative periods, mortality rates, shunt use, and extent of operation above and below typical sites on the carotid artery. Life-table analysis was used to determine the effects of plaque at the end of operation on long-term mortality rates. These data were compared with the log-rank test. s RESULTS Operative results. Intraoperative revisions were performed on 17 arteries (2.5%) because of abnormalities in the ultrasonic spectrum. Three of these were due to intimal or atherosclerotic flaps at the distal end point of the endarterectomy and were treated by extending the arteriotomy, removing the flap, and repatching. The other 14 reopenings were due to abnormalities in the external carotid artery. These were treated by continuing the dissection up to the lingual artery and making a small transverse arteriotomy at this point. Every carotid problem was due to inadequate plaque removal. The diagnosis of external carotid occlusion was obvious from palpation of the artery, whereas the internal carotid artery abnormalities were not clinically apparent. Neurologic morbidity after these 686 operations was 1.5%. Ten patients either awoke with a neurologic deficit or a deficit developed within the early postoperative period. The recovery room OPG-Gee was normal in all 10 patients. Duplex scanning or immediate arteriography was performed within the first week in 8 of the 10 patients. Two of these patients had significant residual disease (type II). Neurologic events developed in both several days after operation. One patient had a severe residual stenosis proximal to the endarterectomy site with a probable clamp injury. At operation the presence of plaque but not the severity of it was recognized. The other
Journal of VASCULAR SURGERY
114 Green et al.
6% I ~204 .
•
_
.:/
32 128
/ o%i 0
10
20
30
40
Follow-up (months) Fig. 1. Life table shows the incidence of true recurrent disease. These patients had normal duplex scans immediately after operation and were studied serially by use of the same imaging technique and examiner. At 24 months the incidence of true recurrence is 3.4%.
Fig. 3. A, Type I plaque. The residual plaque appears as a shelf proximal to the endarterectomy site. B, Doppler spectrum shows peak flow velocities less than 4.0 kHz, and is typical of the flows associated with the common carotid shelf. patient did not have obvious residual disease at the time of operation. Six of these deficits persisted after hospital discharge, and four resolved completely. No recognized immediate postoperative internal carotid artery occlusions occurred in the neurologically normal patients. Two patients were immediately returned to the operating room because they were bleeding. The normal g r o u p (no residual disease). Five hundred thirty-nine arteries (84%) had no residual atherosclerotic disease or significant hemodynamic abnormalities at the operative site (categories A and B). The operation was carried out either above th.-
digastric muscle or below the omohyoid muscle in 86 of these patients (16%). Shunts were used during 60 operations (11%). These patients were followed for 23 + 0.8 months. Eight patients (1.5%) had neurologic events during follow-up. Five of these patients had symptoms in the opposite hemisphere; two had ipsilateral lacunar infarcts, and one suffered an ipsilateral intracranial hemorrhage. Fifty-three patients died (8%) during the period of observation, but only one died of a stroke (the patient with the intracranial bleeding). Twenty-one (3.8%) of these patients exhibited a significant change in ultrasonic pattern of the operated artery during follow-up (Fig.
Volume 13 Number 1 lanuary 1991
Carotid arterial disease 115
1.6%
204
1.4%
f
1.2%F
0.8% 0.6% I
i
0.4%I
493 S - J
J
10
0
i
~
20
30
40
Follow-up (months) Fig. 2. Life table shows the incidence of symptoms from recurrent carotid stenoses. At 24 months the incidence of symptomatic recurrence is 0.7%. No patient had a neurologic symptom within 8 months of operation.
¢
B
C
Fig. 4. A, Doppler spectrum shows peak diastolic flows of greater than 4.5 kHz. The realtime image showed soft plaque in the vessel. These changes were present in both patients whose angiograms are shown below. B, Arterogram shows the soft material filling the internal carotid artery. 4C, The same phenomenon at 1 month after operation in the common carotid artery. The external carotid artery has also occluded.
1.) No patient whose condition was normal immediately after operation developed changes earlier than 8 months after operation. The incidence of true recurrence calculated by life-table methods was 1.5% at 12 months, 3.4% at 24 months, and 5.2% at 36 months. The incidence of symptomatic recurrence in these same arteries was 0.2% at 12 months, 0.7% at 24 months, and 1.4% at 36 months (Fig. 2).
Residual disease type I: proximal plaque, no stenosis. Sixty-one patients had residual plaque appearing as a shelf immediately proximal to the arteriotomy (Fig. 3, A). None of these patients had Doppler abnormalities other than mild spectral broadening with peak systolic flows less than 4.0 kHz (Fig. 3, B). Twenty-one of these patients had extension of their lesion proximally on their preoperative
116
Journatof VASCULAR SURGERY
Green et al.
Table I. Summary of the groups and relevant factors
Normal Type I: CCA plaque Type II: CCA plaque + stenosis Type III: no plaque stenosis
Frequency
Follow-up mo +- SEM
Progression of lesion
Neurologic cymptoms ~ ~
Deaths
539 (79%) 61 (9%)
23 -+ 0.8 21 -+ 2.2
21 (3.8%) 0
8 (1.5%) 1 (1.6%)
53 (8%) 5 (8.2%)
56 (8%)
18 _+ 2.2
4 (7%)
5 (9%) *
8 (14%)
30 (4%)
22 + 3.5
0
0
0
CCA, contralateral carotid artery. ~p < 0.05 when compared to all other groups. **Neurologic symptoms: in the normal group five were in contralateral hemisphere, two were ipsilateral lacunar infarcts, and one was an ipsilateral hemorrhagic infarct. In t-ype I the event was a series of ipsilateral sensor), transient ischemic attacks. In group II all five were ipsilateral and related to the operated artery.
arteriogram. The operation was extended below the omohyoid muscle in 16 (26%) of these patients. Shunts were used in 18 patients (30%). These patients were followed for 21 -+ 2.2 months. No signiticant changes occurred in plaque morphology or flow characteristics. Transient, ipsilateral neurologic deficits occurred in one patient ( 1.6%) 7 months after operation, and an arteriogram was obtained. A nonstenotic waist was seen at the proximal extent of the patch, but no obvious source for the emboli was identified, and the patient was placed on warfarin anticoagulation for control of symptoms. Five of these patients died during this interval. Residual disease type II: proximal plaque with stenosis. Fifty-six patients had both plaque and significant hemodynamic abnormalities at the carotid bulb characterized by peak systolic frequencies greater than 4.0 kHz. Three patients had severe changes as seen in Fig. 4, A. Two of these patients had angiograms (Figs. 4, B and C). The operation was carried out below the omohyoid muscle in 22 patients (39%). Shunts were used in 16 patients (29%). These patients were followed for 18 -+ 2.2 months. During that period one patient had a minor stroke, three patients had transient ischemic attacks and one patient had a retinal stroke. Two patients with symptoms and two others without symptoms had significant morphologic changes in their lesions. All seven patients underwent reoperation. Four of the patients had thin layers ofclo~ and atherosclerotic debris loosely adherent to the arterial wall. When the debris was removed the endarterectomized surface appeared normal. Each of these patients had an embolic event. One patient had a discrete atherosclerotic plaque with ulceration proximal to the prior endarterectomy, and the other two patients had nonthrombotic, fibrotic changes at the carotid bifurcation at the site of the septum dividing the internal and ex-
ternal carotid arteries and extending upward into the internal carotid artery. Eight of these patients died during the follow-up period. Residual disease type III: no visible plaque but significant stenosis. Thirty patients had significant flow abnormalities consisting of peak systolic flow velocities of greater than 4.0 kHz. No visible plaque was identified in these patients, but eight had difficult studies because of neck size, and nine patients had kinks at the distal end of the artery. Positive OPGGee measurements were obtained in 12 patients, the test could not be done in 7 patients, and 11 were normal. Four of these patients had operations beyond the digastric muscle (13%), and seven others had tacking sutures placed at the distal end of the arteriotomy for control of the end point of the endarterectomy. These patients were followed for 22 -+ 3.5 months. During that period none of the lesions changed, no one suffered any neurologic symptoms, and no one died. Table I summarizes factors relevant to each group of arteries. Development o f symptoms caused by residual disease. Seven patients had neurologic symptoms during the period of follow-up (Fig. 5). Symptoms developed in two patients while in hospital, both resolved within 3 weeks. The remaining four patients had symptoms within 7 months of operation. Effect o f shunting. Shunts were used in 102 operations (15%). No difference was found in the neurologic outcome between shunted and nonshunted patients. The incidence of residual abnormalities was significantly higher in the group of patients that required shunting. Shunts were used in 60 of the 539 (11%) operations with normal anatomic and hemodynamic outcomes. (Table II). One hundred forty-seven patients had residual abnormalities, and shunts were used in 42 of these operations (29%), p < 0.05. This increase in residual disease
Volume 13 Number 1 January 1991
Carotid a~eriald#eme
117
0.05
0.04
- /
.........
123
48
0.03
0.02
/
0.01
0
--
-
~
0
5
3 _ _ _
10
15
20
25
Follow-up (months) Fig. 5. Life table shows the development of symptoms in patients with residual carotid lesions. M1 seven patients had problems within 7 months of operation.
Table II. The effect o f temporary shunting on residual disease Carego~y No residual disease Residual disease Type I Type II Type III
Tempora~ shunting 60 42 18 16 8
(11%) (29%)* (30%) (29%) (27%)
Table III. The effect o f a non routine operation on residual disease Category No residual disease Residual disease Type I Type II Type III
Non routine operation 86 49 16 22 11
(16%) (33%)* (26%) (39%) (37%)
*p < 0.05 when residual groups are compared to no residual group,
*p < 0.05 when residual groups are compared to no residual group.
after using a shunt was observed equally in all three groups with residual disease. Effect o f a " n o n r o u t i n e " operation o n residual disease. Operations were carried out below the omohyoid muscle, above the hypoglossal nerve, or with tacking sutures to secure the distal end point in 135 arteries (Table III). Eighty-six o f the 539 normal arteries (16%) underwent a "nonroutine" operation. Forty-nine o f the 147 arteries (33%) with residual disease had "nonroutine" operations. This difference is statistically significant, p < 0.05.
ences in the definition o f recurrence, the surveillance techniques used, and the time interval arteries are studied after operation. Many factors such as patching, shunting, gender, tacking sutures, and artery size have been shown both to influence and not to influence the development o f a recurrent carotid stenosis. 7"s'2 We separated 686 endarterectomized carotid arteries into those with residual disease and those without, and followed them serially with duplex scans for an average o f 2 years. We found that residual and recurrent carotid stenoses are each distinct clinical entities. Residual carotid lesions may be left despite the best efforts o f the surgeon. The most common reasons for residual lesions in our series were the socalled extended or nonroutine operation and the use o f temporary indwelling shunts. Residual lesions were more than twice as common in these settings compared to a standard operation without shunting. The physical presence o f the shunt makes visualization o f the end points more difficult and therefore
DISCUSSION Technical errors that result in incomplete endarterectomy have been incriminated as a cause ofneurologic morbidity after operations on the carotid artery. 3 N o w some claim that leaving residual disease is also a major cause o f recurrent carotid stenosis. 2'6 The topic o f the recurrent carotid stenosis is a confusing one. The wide variation in incidence among the numerous studies on this entity is due to differ-
7OUrpk2[ O ~"
118
Green et al.
impedes complete plaque removal. Our primary concern was removal of all distal plaque, but we often knowingly left residual disease proximally when extending the dissection presented problems. The use of tacking sutures for unsatisfactory end points also leaves residual disease and if possible should be avoided if further distal dissection is possible. We used continuous-wave Doppler analysis to assess flow patterns and to repair major technical errors. We recognized three internal carotid arteries with problems and repaired them. Although this technique may be criticized for being totally dependent on the operator's abili~ to recognize audible patterns, it was successful in avoiding any occlusions of the internal carotid artery and resulted in a consistently low stroke rate in a large series of patients. Barnes et alJ report a similar satisfactory experience with this technique. More sophisticated and expensive intraoperative imaging techniques, such as duplex scanning or arteriography, can identi~ residual disease, but the proper approach to those abnormalities not causing hemodynamic alterations is not clear. Flanigan et al. 9 were the first to use the duplex scanner in this setting. They found a 28% incidence of intraluminal defects, but repaired only 25% of them. No adverse clinical sequelae occurred as the stroke rate was 3.8% with a normal ultrasound and 3.3% with an abnormal study. In a follow-up report of the unrepaired arterial defects Sawchuk et a1.1° found no statistical relationship between an unrcpaircd arterial defect and subsequent recurrent carotid artery stcnosis. On the other hand, Reilly ct al, 2 found that significant, unrepaired technical defects were a major factor in the development of recurrent carotid stenoses. Defects causing a stcnosis of greater than 41% or having a flap length of greater than 26% of the lumen size were associated with the development of recurrent lesions. Hypertension, smoking, and cholesterol levels, however, also played a major role in this series. Bandyk et al. 6 used pulsed Doppler spectral analysis and concluded that moderate to severe flow disturbances were associated with a higher incidence of recurrent disease over a 2-year period. We would maintain that these phenomena were not true recurrenccs but represented residual disease that progressed in an atherogenic environment. When the pathologic make up of the recurrent carotid lesion was examined by Clagett et al. H most of the lesions were located within the endarectomized segment, 65% of the lesions were located beyond the origin of the internal carotid artery, and most were associated with a long primary artcriotomy. Ninety
VA~Ct LAI~ S1 ~RGER'~
percent of the arteries had surface or intraplaquc thrombus, which formed the bulk of the lesion in many patients. These lesions were identical to our type II group. The high frequency of thrombus tbrmation on the endarterectomized surface may explain why other investigators report that early "recurrences" regressJ 2 Wc did not find regression of residual lesions, but did find thrombus in those that progressed or became symptomatic. O'Donncll et al. ~3 studied the same problem and found that the earliest "recurrent" lesions were located in the common carotid artery and actually represented residual disease similar to our type I lesions. Clagctt did describe three cases of "retained shelves" like this as an exception to the rest of his cases. A strong desire exists to put the entire problem of carotid stenoses after endarterectomv into a continuum of changes as a result of local healing factors acting in concert with systemic risk factors. There is abundant data that show that recurrent disease is the result of a progression of events beginning with ongoing thrombogcnesis occurring at the cndartercctomy site. Organization of this thrombus and proliferation of smooth muscle cells result in neointimal hyperplasia that with time undergoes atherosclerotic degeneration. 11 Our series presents no evidence to support or refute that hypothesis. Instead, we offer data that suggest that true recurrence starts somewhere beyond 8 months from operation, and at 3 years occurs in 5.2% of arteries that were normal immediately after operation. This incidence is considerably lower than others have reported when residual lesions were not excluded from the analysis and similar to those when residual disease was excluded. 1"14Furthermore, the incidence of symptoms from these true recurrences is even rarer. At 3 years only 1.4% of patients who were healthy after operation developed symptoms from their operated arteries. It is important not to confuse recurrent lesions with residual ones. Like recurrent lesions, most residual lesions arc benign. The common carotid shelves with minimal flow disturbances do not progress and rarely cause symptoms. We do not believe that extensive proximal mobilization and a long arteriotomy is necessary to remove nonstenotic, nonulcerated common carotid plaque. The one case of recurrent symptoms we observed in this type of patient probably came from somewhere else. The lesions with significant flow disturbances but no visualized plaque are also benign. We routinely patch and do not hesitate to use tacking sutures when a perfectly feathered end point cannot be achieved. If
Volume 13 Number 1 Januaq, 1991
w e are confident that t h e e n d p o i n t is secure, w e are n o t b o t h e r e d b y flow velocity elevations after surgery. T h e velocity elevations are usually kinks o r abnormalities caused b y a distal plaque that has been tacked d o w n . T h e o n l y residual lesion that is o f concern is the o n e w i t h b o t h plaque (usually soft) and significant flow disturbances. W e are h o p e f u l that these lesions will be r e c o g n i z e d in the o p e r a t i n g r o o m . W e c o n t i n u e to believe t h a t the c o n t i n u o u s w a v e - D o p p l e r analysis is a d e q u a t e for this surveillance since s y m p t o m s a n d p r o g r e s s i o n usually occur w i t h i n 6 m o n t h s o f o p e r a t i o n . These patients should u n d e r g o a r t e r i o g r a p h y w h e n this t y p e o f lesion is identified, a n d the lesion s h o u l d be repaired.
Carotid arterial disease 119
6.
7. 8.
9. 10.
REFERENCES
1. Barnes RW, Nix ML, Nichols BT, Wingo JP. Recurrent versus residual carotid stenosis. Incidence detected by Doppler ultrasound. Ann Surg 1986;203:652-60. 2. ReilJy LM, Okuhn SP, Rapp JH, et al. Recurrent carotid stenosis: a consequence of local or systemic factors? The influence of unrepaired technical defects. I VASCSURG 1990; 11:448-60.
3. Callow AD. Restenosis after carotid arteu surgeqT.Arch Surg 1982;117:1082-5. 4. Green RM, Messick WJ, Ricotta JR, et al. Benefits, shortcomings and costs of EEG monitoring. Ann Surg 1985; 201:785-92. 5. Anderson S, Auquier A, Hauck W, et al. Techniques for
DISCUSSION Dr. R o b e r t Barnes (Little Rock, Ark.). Dr. Green and his colleagues have expanded on a subject first reported by Dr. Blaisdell, who used intraoperative and postoperative arteriography to identify perioperative and postoperative carotid defects. From the present study, we learned that 10 patients with perioperative neurologic deficits were not operated on again because of normal OPG studies, yet two proved to have significant residual common carotid stenosis on subsequent carotid duplex examination. Second, we learned that in their hands significant residual carotid stenosis correlated with the use of a shunt. Third, we learned that residual disease also correlated with a nonroutine operation, which extended below the omohyoid, above the digastric, or required tack down sutures. Fourth, we learned that residual stenosis was more likely to cause postoperative neurologic deficits if it was proximal than if it was distal. Finally, we learned that significant residual carotid stenosis was detected in the early postoperative period in 12% of their patients despite the use of intraoperative
11. 12. 13.
14.
bias reduction. Statistical methods for comparative studies. New York: John Wiley and Sons, 1980:205-11. Bandyk DF, Kaebnick HW, Adams MB, et al. Turbulence occurring after carotid bifurcation endarterectomy: a harbinger of residual and recurrent stenosis. J VASC SURG 1988; 7:261-74. Ouriel K, Green RM. Clinical and technical factors influencing recurrent carotid stenosis and occlusion after endarterectomy. J VASCSURG 1987;5:702-6. Ackroyd N, Lane R, Appleberg M. Carotid endarterectomy. Long-term follow-up with specific reference to recurrent stenosis, contralateral progression, mortality,, and recurrent neurological episodes. J Cardiovasc Surg 1986;27:418-25. Flanigan DP, Douglas DJ, Machi J, et al. Intraoperative ultrasonic imaging of the carotid arte~ during carotid endarterectomy. Surgery, 1986;100:893-9. Sawchuk AP, Flanigan DP, Machi J, et al. The fate of unrepaired minor technical errors detected by intraoperative ultrasound during carotid endarterectomy. J VASC SURG 1989;9:671-6. Clagett CG, Robinowitz M, Youkey Jr et al. Morphogenesis and clinico-pathologic characteristics of recurrent carotid disease. J VASCSURG 1986;3:10-23. Zierler RE, Bandyk DF, Thiele BL. Intraoperative assessment of carotid endarterectomy. J VASCSURG 1984;1:73-83. O'Donnell TF Jr, Callow AD, Scott G, et al. Ultrasound characteristics of recurrent carotid disease: hypothesis explaining the low incidence of symptomatic recurrence, j"VASC SURG 1985;2:26-41. Blaisdell FW, Lim R Jr, Hall JD. Technical result of carotid endarterectomy. Am J Surg 1967;114:239-46.
Doppler monitoring. This is higher than the 1% incidence noted by Blaisdell or the two percent in our study. This newfound knowledge prompts me to ask the following four questions. First, do you plan to change your method of invasive evaluation of patients with perioperative neurologic deficits? Second do you plan to change your method o f endarterectomy in patients who require shunts? Third, do you plan to change your approach to patients requiring nonroutine, more extensive proximal or distal dissection? Fourth and finally, how do you propose to identify and correct these severe residual carotid lesions during surgery to reduce the 12% incidence of early postoperative disease, which in seven patients led to reoperation? Dr. R o b e r t Courbier (Marscillc, France). In our group we used routinely pcrioperative angiography since 1982 and during that period performed more than 1800 perioperative angiograms. In 4% o f cases we had to correct a technical fault, and in 5.5% we accepted residual disease. Over the past 3 years we began a prospective study on 280 patients who accepted long-term fbllow-up with du-
120
Green et al.
plex scanning every 6 months and a digital subtraction angiogram 1 year after operation. Fifteen patients had accepted residual disease. Among those who had internal carotid artery stenosis less than 50%, two developed an occlusion with transient ischemic attack, two remain asymptomatic. And three, who had transient ischemic attack were reoperated for high-grade stenosis. On the common carotid artery the shelf aspect disappeared in six cases and remained unchanged and asymptomatic in two cascs. Dr. Green, I want to underline that the pathologic significance o f this residual disease has to be known and that chiefly now is when noninvasive techniques, like angioscopy, demonstrate an increasing number o f technical faults without any clinical consequences in the future. Dr. Charles Branfigan (Denver, Colo.). A group of patients in your study had velocity elevations but no changes in the gray scale image. Do you have any angiographic confirmation that this indeed represents a stenosis? Is it possible that this velocity elevation is some kind o f an artifact introduced by surgery or introduced by the change in Doppler angle caused by a change in position o f the artery after surgery? Dr. Wesley M o o r e (Los Angeles, Calif.). I have just one other comment. I think the information that Dr. Green has shared with us today with respect to residual stenosis in the common carotid artery is extremely valuable in that he has pointed out that this is a lesion o f relatively low morbidity. If that were the only location for residual stenosis, I think we would be in great shape. Unfortunately, there are other sites such as the distal end point in the proximal internal carotid artery. More recently, we have identified a problem with residual lesion in the external carotid artery. We have now gathered three patients who have developed thrombus in association with the flap o f the external carotid artery with retrograde propagation and subsequent embolization in the internal carotid. Therefore, I believe that intraoperative surveillance deserves more attention. Beyond just looking for a proximal lesion, it is more important that we look for the distal lesions. I would recommend the use o f some sort o f intraoperative surveiUance for this particular problem, such as completion angiography or B-mode ultrasound. Dr. John R i c o t t a (Buffalo, N.Y.). Dr. Green, it seems to me that the problem you are having is failure to detect or treat proximal disease. I wonder if you could tell us whether this is a function o f your preoperative angiography, perhaps more use o f intraarterial digital subtraction angiography versus cut films, and whether you went back and looked. I know these patients are studied on duplex scanning before surgery. Could you have predicted by the
Journalof VASCULAR SURGERY
preoperative duplex scanning whether there was more common carotid disease than we saw, and would you suggest that we spend a little more time and effort looking at the common carotid as well as the bifurcation? Dr. Green. Dr. Barnes, I cannot explain the differences between our residual stenosis rate o f 12% and your incidence and Dr. Blaisdell's incidence of 1% to 2%. Some o f this may reflect selection criteria since many o f these patients had extended operations, and some may represent your skills with the continuous-wave Doppler examination. Since our overall stroke rate after operation is quite low we do not plan any major changes in the way we evaluate and treat these patients. We are not prepared to routinely use intraoperative duplex scanning or arteriography. We will certainly be more attentive to leaving proximal disease when we use a shunt. We will be more aggressive with patients in the early postoperation period with both residual plaque and significant flow velocity, elevations. I appreciate Dr. Courbier's remarks. I wish that our operating room had fixed angiographic equipment that would allow us to easily study our patients. Our intraoperative carotid studies leave much to be desired and would not with our current equipment provide us with useful information about hemodynamically insignificant residual disease. Dr. Brantigan's remarks arc very valid. I think that the 30 patients who had velocity elevations alone with no symptoms and no progression probably to a largc degree represent kinks and difficult studies. This is a very soft group o f patients and I would not make a whole lot of" conclusions about them. If we do not sec a filling defect we are not concerned about the vclocity elevation. Dr. Moore, we have not seen symptoms from external carotid stenoses but have ccrtainly seen a number o f these lesions. I think that the reason we are not seeing internal carotid residual stenoses is because we are concerned with going high enough on thc internal carotid to remove all disease. We have not been as compulsive going proximally, and perhaps if we did our ovcrall incidence o f residual lesions would be lower. Dr. Ricotta, we looked at the arteriograms of every patient to see whether we could have predicted lcaving a residual stenosis and we could not. However some patients had disease extending down the common carotid artc~, and we made every effort to remove it all. Other patients whose angiograms did not raise our suspicions had more disease than we could remove and were left with residual lesions.
