Is routine duplex examination after carotid endarterectomy justified? James M. Cook, MD, Bernard W. Thompson, MD, and R o b e r t W. Barnes, M.D, Little Rock, Ark. Routine follow-up of patients after carotid endarterectomy with duplex scanning is commonly practiced, yet the clinical significance of identifying those with asymptomatic restenosis is unclear. To address this issue we reviewed 120 consecutive patients who underwent 143 carotid endarterectomies from August 1983 to December 1988. One hundred one patients (118 operations) were available for clinical follow-up, and the overall incidence o f recurrent symptoms was 6% (6/101). Sixty-three of these patients (78 carotid endarterectomies) had postoperative duplex examination. Two had evidence of residual disease from the time of surgery and were not included in further analysis. Significant recurrent stenosis (greater than 50% diameter reduction) developed in 14 of the remaining 76 arteries (18.2%). Twelve of 14 stenoses remained asymptomatic during follow-up from 18 to 72 months (mean 47.0 months) and did not undergo reoperation. Recurrent ipsilateral hemispheric symptoms developed in two patients with restenosis (14.3%). Four of the 62 arteries without significant recurrent stenosis developed ipsilateral symptoms (6.5%), but none required reoperation during follow-up from 1 to 71 months (mean 31.6 months). Life-table analysis showed no increased risk of transient ischemic attack, stroke, or death in patients with restenosis. This study supports regular clinical followup after carotid endarterectomy with emphasis on patient education in the recognition of symptoms. Although duplex scanning may be useful to follow known contralateral asymptomatic disease or evaluate those with recurrent symptoms, its routine use to identify patients with asymptomatic restenosis after carotid endarterectomy may be unnecessary. (J VAsc Sting 1990;12:334-40.)
Many surgeons recommend routine noninvasive follow-up of all patients after carotid endarterectomy to identify recurrent asymptomatic stenosis.l-4 However, the role of carotid restenosis in the pathogenesis of stroke after endarterectomy is unclear. The risk of stroke has been estimated in the population with hemodynamically significant atherosclerotic plaque, but has not been rigorously studied in patients with recurrent lesions. This uncertainty has led some surgeons to recommend reoperation on asymptomatic high-grade recurrent stenosis, whereas others believe that this should be reserved for those with recurrent symptoms. Several factors have influenced us to not reoperate on any patient with asymptomatic carotid restenosis. First, it is known that only approximately 1% to 4% From the Departmentof Surgery,UniversityofArkansasfor Medical Sciences, John L. McClellan Memorial Veterans Administration Hospital. Presented at the Fourteenth Annual Meeting of the Southern Association for Vascular Surgery, Acapulco,Mexico,Jan. 2427, 1990. Reprint requests: JamesM. Cook,MD, VascularSurgerySection, 112-PV,John L. McClellanMemorialVeteransAdministration Hospital, 4300 West 7th St., Little Rock,AR 72205. 24/6/22759 334
of patients have focal neurologic symptoms (transient ischemic attack [TIA] or stroke) associated with recurrent carotid artery stenosis, s8 Second, despite excellent protection from stroke, long-term survival rates after carotid endarterectomy are disappoint-' ingly low, primarily as a result of coronary atherg;, sclerosis. 9,1° Finally, the morbidity of reoperation may be several times that of the original procedure.3,11 The results of our nonoperative approach to asymptomatic carotid restenosis and the implications' for routine postoperative duplex examination arc the basis of this report. PATIENTS AND METHODS The records of all patients undergoing carotid endarterectomy from August 1983 to December 1988 were retrospectively reviewed from the University Hospital of Arkansas and John L. McClellan Memorial Veterans Administration Hospital. All operations were performed or directly supervised by the two senior authors (B.W.T. or R.W.B.). Important technical aspects were similar in each case. General anesthesia was used, with shunting reserved for those with carotid artery back pressures less than 50 mm~
Volume 12 Number 3 September 1990
Duplex examination after carotid endarterectomy 335
100
,-I
m l . . . ~
8o
_>
l m
n,. m~"
so-
u.
40-
m
0
Io. >. --
93
87
78
67
54
45
0
32
24 I
15
12 i
MONTHS
Fig. 1. Survival free of ipsilateral neurologic symptoms in 101 patients followed clinically (n = number of patients available for follow-up at the corresponding time point).
Hg. Endarterectomy was performed with loupe magnification, and the end point was visualized in every case. Tacking sutures were used in the minority where a smooth end point could not be achieved. patching of the endarterectomy site with vein or synthetic material was rarely done. Assessment of the immediate technical result by use of a sterile continuous-wave Doppler probe or intraoperative arteriography was performed in most cases. Each patient was prescribed aspirin, 325 mg daily. Duplex scans (Acuson, Inc., Model 128, 'Mt. View, Calif.; A.T.L., Inc., Ultramark 4, Bellvue, Wash.; Biosonics, Inc., Indianapolis, Ind.) were obtained by registered vascular technologists in our vascular laboratory and reviewed by two of us (B.W.T. or R.W.B.). Criteria for categorizing the degree of carotid stenosis have been previously validated and hre based on the peak systolic and end diastolic fiequencies or velocities in conjunction with observed spectral changes. 12 An internal carotid:common carotid artery systolic velocity ratio of 2:1 generally indicates a stenosis greater than 50% diameter reduction. Clinic and hospital notes were reviewed; those patients not seen within the past 12 months were contacted by telephone. Recurrent neurologic symptoms (TIA, stroke) and death were the study end points. These were correlated with the presence or absence of hemodynamically significant carotid re~stenosis. The incidence ofipsilateral neurologic events and Jeath was analyzed by the life-table method accord-
ing to Kaplan and Meier. Fisher's exact test was used to determine differences in stroke rate in the group with significant recurrent stenosis versus those without restenosis. Student's t test was used to test the difference in mean length of follow-up between the groups. RESULTS Of 120 patients who underwent 143 endarterectomies during the period, 101 (118 operations) were available for clinical followup. This group included 92 men and 9 women, ranging in age from 48 to 82 years (mean 66.7 years). The overall incidence of recurrent ipsilateral TIA and stroke was 6% (6/101). The initial event in three patients was TIA, whereas three had an unheralded stroke. Actuarial survival free ofipsilateral symptoms was 71.7% at 5 years and is depicted in Fig. 1. Sixty-three patients (78 operations) followed clinically had at least one postoperative duplex scan. Two undergoing unilateral endarterectomies had evidence of residual disease from the time of surgery and were not included in further analysis. One developed an asymptomatic internal carotid artery occlusion 2 weeks after endarterectomy and was treated nonoperatively. Another had evidence of residual disease by duplex examination 1 week after operation and underwent reoperation only after the development of TIAs 3 months later. Significant recurrent stenosis (greater than 50% diameter reduction) developed in 14 of the remaining 76 arteries (18.4%) from 6 to 70 months after
336
Journal of VASCULAR SURGERY
Cook, Thompson, and Barnes
100
UJ UJ
n-
80
RESTENOSIS
NO RESTENOSlS
O
I--
6o
40
I'-
Z
UJ I1.
0
14
14
13
12
12
11
11
7
4
2
. . . . . 54
52
46
39
31
25
16
12
10
8
36 '
12
48 '
go
MONTHS
Fig. 2. Ipsilateral neurologic symptoms in patients with and without recurrent carotid stenosis (n = number of patients available for follow-up at the corresponding time point)
operation (mean 23.0 months). By duplex criteria, 10 recurrent lesions remained stable (71,4%), and four regressed (28.6%); none progressed to occlusion. Twelve of fourteen stenoses remained asymptomatic during follow-up from 18 to 72 months (mean 47.1 months) and did not undergo angiography or reoperation. Ipsilateral hemispheric TIAs developed 19 and 70 months after surgery in two patients with recurrent stenosis (14.3%). After confirmation of restenosis by angiography, the former underwent successful reoperation whereas the latter suffered a stroke awaiting operation 1 month after the onset of symptoms. Four (6.5%) of the 62 arteries without significant recurrent stenosis developed ipsilateral symptoms during follow-up from 1 to 71 months (mean 31.6 months). Length of follow-up was not statistically significant between the two groups (Student's t test). Three patients had strokes at 18, 24, and 37 months after operation whereas the fourth returned with hemispheric TIAs after 21 months and has been free of symptoms on aspirin for the past 12 months. None has had reoperation. The incidence of ipsilateral hemispheric symptoms in the group with hemodynamically significant stenosis was not significantly different from that in the group with mild or no stenosis (p = 0.30, Fisher's exact test). By life-table analysis, the risk of ipsilateral stroke and TIA was the same in both cohorts after 5 years (Fig. 2; log-rank p = 0.74; Wilcoxon p = 0.97). Also, 5-year survival without stroke or
TIA ("symptom-free survival") in patients with significant restenosis did not differ significantly from that in patients without recurrent stenosis (Fig. 3; log-rankp -- 0.89; Wilcoxonp = 0.78). DISCUSSION
This observational, case-control study demonstrates that the natural history of patients with hemodynamically significant recurrent carotid stenosis is not different from the outcome of those with mild or no recurrent disease. Life-table data show no increased risk of TIA, stroke, or death in patients with restenosis, justifying our policy of selective r e o p e ~ ation based on recurrent symptoms. Therefore routine postoperative duplex scanning to detect patients with asymptomatic restenosis may be unnecessary. Others have also failed to find a consistent re!ation between restenosis and recurrence of neuro-' logic symptoms. Unlike the original atherosclerotic plaque, most early recurrent stenoses (less than 2 years after endarterectomy) are due to neointimal hyperplasia. 13"16 It has been hypothesized that the low incidence of intraplaquc hemorrhage or ulceration in recurrent lesions accounts for the low frequency of' symptomatic recurrence. 17 Using life-table methods, Healy et al.ls reported no difference in the incidence of symptoms in patients with restenosis than in those without significant recurrent disease. Their rate of symptomatic recurrent stenosis (14%) was similar to ours. In a series of 142 endarterectomies, Glover et al. 19 found that 3 of 23 patients with greater than
Volume 12 Number 3 September 1990
Duplex examination a~er carotid endmcerectomy 337
100
Many surgeons recommend routine noninvasive follow-up of all patients after carotid endarterectomy to identify recurrent asymptomatic stenosis.l-4 However, the role of carotid restenosis in the pathogenesis of stroke after endarterectomy is unclear. The risk of stroke has been estimated in the population with hemodynamically significant atherosclerotic plaque, but has not been rigorously studied in patients with recurrent lesions. This uncertainty has led some surgeons to recommend reoperation on asymptomatic high-grade recurrent stenosis, whereas others believe that this should be reserved for those with recurrent symptoms. Several factors have influenced us to not reoperate on any patient with asymptomatic carotid restenosis. First, it is known that only approximately 1% to 4% From the Departmentof Surgery,UniversityofArkansasfor Medical Sciences, John L. McClellan Memorial Veterans Administration Hospital. Presented at the Fourteenth Annual Meeting of the Southern Association for Vascular Surgery, Acapulco,Mexico,Jan. 2427, 1990. Reprint requests: JamesM. Cook,MD, VascularSurgerySection, 112-PV,John L. McClellanMemorialVeteransAdministration Hospital, 4300 West 7th St., Little Rock,AR 72205. 24/6/22759 334
of patients have focal neurologic symptoms (transient ischemic attack [TIA] or stroke) associated with recurrent carotid artery stenosis, s8 Second, despite excellent protection from stroke, long-term survival rates after carotid endarterectomy are disappoint-' ingly low, primarily as a result of coronary atherg;, sclerosis. 9,1° Finally, the morbidity of reoperation may be several times that of the original procedure.3,11 The results of our nonoperative approach to asymptomatic carotid restenosis and the implications' for routine postoperative duplex examination arc the basis of this report. PATIENTS AND METHODS The records of all patients undergoing carotid endarterectomy from August 1983 to December 1988 were retrospectively reviewed from the University Hospital of Arkansas and John L. McClellan Memorial Veterans Administration Hospital. All operations were performed or directly supervised by the two senior authors (B.W.T. or R.W.B.). Important technical aspects were similar in each case. General anesthesia was used, with shunting reserved for those with carotid artery back pressures less than 50 mm~
Volume 12 Number 3 September 1990
Duplex examination after carotid endarterectomy 335
100
,-I
m l . . . ~
8o
_>
l m
n,. m~"
so-
u.
40-
m
0
Io. >. --
93
87
78
67
54
45
0
32
24 I
15
12 i
MONTHS
Fig. 1. Survival free of ipsilateral neurologic symptoms in 101 patients followed clinically (n = number of patients available for follow-up at the corresponding time point).
Hg. Endarterectomy was performed with loupe magnification, and the end point was visualized in every case. Tacking sutures were used in the minority where a smooth end point could not be achieved. patching of the endarterectomy site with vein or synthetic material was rarely done. Assessment of the immediate technical result by use of a sterile continuous-wave Doppler probe or intraoperative arteriography was performed in most cases. Each patient was prescribed aspirin, 325 mg daily. Duplex scans (Acuson, Inc., Model 128, 'Mt. View, Calif.; A.T.L., Inc., Ultramark 4, Bellvue, Wash.; Biosonics, Inc., Indianapolis, Ind.) were obtained by registered vascular technologists in our vascular laboratory and reviewed by two of us (B.W.T. or R.W.B.). Criteria for categorizing the degree of carotid stenosis have been previously validated and hre based on the peak systolic and end diastolic fiequencies or velocities in conjunction with observed spectral changes. 12 An internal carotid:common carotid artery systolic velocity ratio of 2:1 generally indicates a stenosis greater than 50% diameter reduction. Clinic and hospital notes were reviewed; those patients not seen within the past 12 months were contacted by telephone. Recurrent neurologic symptoms (TIA, stroke) and death were the study end points. These were correlated with the presence or absence of hemodynamically significant carotid re~stenosis. The incidence ofipsilateral neurologic events and Jeath was analyzed by the life-table method accord-
ing to Kaplan and Meier. Fisher's exact test was used to determine differences in stroke rate in the group with significant recurrent stenosis versus those without restenosis. Student's t test was used to test the difference in mean length of follow-up between the groups. RESULTS Of 120 patients who underwent 143 endarterectomies during the period, 101 (118 operations) were available for clinical followup. This group included 92 men and 9 women, ranging in age from 48 to 82 years (mean 66.7 years). The overall incidence of recurrent ipsilateral TIA and stroke was 6% (6/101). The initial event in three patients was TIA, whereas three had an unheralded stroke. Actuarial survival free ofipsilateral symptoms was 71.7% at 5 years and is depicted in Fig. 1. Sixty-three patients (78 operations) followed clinically had at least one postoperative duplex scan. Two undergoing unilateral endarterectomies had evidence of residual disease from the time of surgery and were not included in further analysis. One developed an asymptomatic internal carotid artery occlusion 2 weeks after endarterectomy and was treated nonoperatively. Another had evidence of residual disease by duplex examination 1 week after operation and underwent reoperation only after the development of TIAs 3 months later. Significant recurrent stenosis (greater than 50% diameter reduction) developed in 14 of the remaining 76 arteries (18.4%) from 6 to 70 months after
336
Journal of VASCULAR SURGERY
Cook, Thompson, and Barnes
100
UJ UJ
n-
80
RESTENOSIS
NO RESTENOSlS
O
I--
6o
40
I'-
Z
UJ I1.
0
14
14
13
12
12
11
11
7
4
2
. . . . . 54
52
46
39
31
25
16
12
10
8
36 '
12
48 '
go
MONTHS
Fig. 2. Ipsilateral neurologic symptoms in patients with and without recurrent carotid stenosis (n = number of patients available for follow-up at the corresponding time point)
operation (mean 23.0 months). By duplex criteria, 10 recurrent lesions remained stable (71,4%), and four regressed (28.6%); none progressed to occlusion. Twelve of fourteen stenoses remained asymptomatic during follow-up from 18 to 72 months (mean 47.1 months) and did not undergo angiography or reoperation. Ipsilateral hemispheric TIAs developed 19 and 70 months after surgery in two patients with recurrent stenosis (14.3%). After confirmation of restenosis by angiography, the former underwent successful reoperation whereas the latter suffered a stroke awaiting operation 1 month after the onset of symptoms. Four (6.5%) of the 62 arteries without significant recurrent stenosis developed ipsilateral symptoms during follow-up from 1 to 71 months (mean 31.6 months). Length of follow-up was not statistically significant between the two groups (Student's t test). Three patients had strokes at 18, 24, and 37 months after operation whereas the fourth returned with hemispheric TIAs after 21 months and has been free of symptoms on aspirin for the past 12 months. None has had reoperation. The incidence of ipsilateral hemispheric symptoms in the group with hemodynamically significant stenosis was not significantly different from that in the group with mild or no stenosis (p = 0.30, Fisher's exact test). By life-table analysis, the risk of ipsilateral stroke and TIA was the same in both cohorts after 5 years (Fig. 2; log-rank p = 0.74; Wilcoxon p = 0.97). Also, 5-year survival without stroke or
TIA ("symptom-free survival") in patients with significant restenosis did not differ significantly from that in patients without recurrent stenosis (Fig. 3; log-rankp -- 0.89; Wilcoxonp = 0.78). DISCUSSION
This observational, case-control study demonstrates that the natural history of patients with hemodynamically significant recurrent carotid stenosis is not different from the outcome of those with mild or no recurrent disease. Life-table data show no increased risk of TIA, stroke, or death in patients with restenosis, justifying our policy of selective r e o p e ~ ation based on recurrent symptoms. Therefore routine postoperative duplex scanning to detect patients with asymptomatic restenosis may be unnecessary. Others have also failed to find a consistent re!ation between restenosis and recurrence of neuro-' logic symptoms. Unlike the original atherosclerotic plaque, most early recurrent stenoses (less than 2 years after endarterectomy) are due to neointimal hyperplasia. 13"16 It has been hypothesized that the low incidence of intraplaquc hemorrhage or ulceration in recurrent lesions accounts for the low frequency of' symptomatic recurrence. 17 Using life-table methods, Healy et al.ls reported no difference in the incidence of symptoms in patients with restenosis than in those without significant recurrent disease. Their rate of symptomatic recurrent stenosis (14%) was similar to ours. In a series of 142 endarterectomies, Glover et al. 19 found that 3 of 23 patients with greater than
Volume 12 Number 3 September 1990
Duplex examination a~er carotid endmcerectomy 337
100
