Intraoperative arteriography and carotid surgery A n t h o n y J. R o o n , M D , FACS, and D e b o r a h H o o g e r w e r f , Everett, Wash. To minimize morbidity in carotid surgery we have adopted the practice of routine operative arteriography. In our total experience of 692 carotid artery reconstructions done since 1979 no arteriography was used for the first 157 procedures. The combined morbidity and mortality rate in this group was 4.5% (4 strokes and 3 deaths as a result of stroke). After 157 carotid operations we used intraoperative arteriography on all cases, and the neurologic morbidity and mortality rate for this part of the series was 1.3% (6 strokes and I fatal cerebral hemorrhage). One death was due to myocardial infarction. The difference in ischemic neurologic sequellae between these two groups was statistically significant by chi-square analysis (p < 0.01). Although the difference in these two cohorts could be related to the "learning curve" effect, it is unlikely that this is the only reason for the difference. Intraoperative arteriography has demonstrated unsuspected stenoses, occlusions, flaps, and kinks. There was also a statisticaUy significant difference (p < 0.01) in the number of reoperations fi~r recurrent stenosis: 3.8% in the first group and 0.8% in the second group. No complications were thought to be directly associated with the arteriogram. Since the development of better digital imaging systems for the operating suite, this technique is much more easily used than when it was first described. (J VASC SURG 1992;16:239-43.)
In the recent past many studies have been done on the efficacy of carotid surgery in the prevention of stroke. Surgical intervention has been shown to be effective when the morbidity of surgery is less than the morbidity of the lesion. 1 Despite this intense reevaluation of carotid surgery, most surgeons are not assessing the adequancy of their surgical results with intraoperative completion arteriography or any other form of intraoperative imaging. Completion carotid arteriography is not a new concept. In 1967 BlaisdeU et al.2 demonstrated that in 25% of cases an unsuspected defect was noted on completion arteriograms. Other authors have confirmed unsuspected defects on routine intraoperative arteriography, a-5 To diminish the incidence of neurologic complications in carotid surgery in our practice we have instituted a policy of routine intraoperative carotid arteriography. The results of the entire series of carotid surgery patients seen in our clinical practice of From the Divisionof VascularSurgery,the EverettClinic,Everett. Dr. Roon is assistantclinicalprofessorof surgeryat the University of WashingtonSchool of Medicine, Seattle. Presented at the Ninth AnnualMeeting of the PacificNotxhwest Vascular Society,Portland, Ore., Nov. 7, 1991. Reprint requests: Anthony I. Roon, MD, Division of Vascular Surgery, the Everett Clinic, 1330 RockefellerAve., Suite 520, Everett, WA 9820i. 24/1/36909
peripheral vascular surgery is presented to demonstrate the effectiveness of this technique in the prevention of immediate and late complications of carotid surgery. PATIENTS AND METHODS All carotid operations performed between August 1979 to May 1991 were included in this series. There were 692 carotid operations. The operations were all performed by the first author (A.J.R.) in._a community hospital setting. In the first 157 operations (group 1) no intraoperative arteriography was performed. In an attempt to better assess the operative results, the subsequent 535 patients (group 2) had arteriography on completion of the endarterectomy. This was initially done with plain x-ray film, but more recently digital angiography has been done with a portable image intensifier. This change in technique has reduced the amount of time required for evaluation and allowed the carotid bifurcation to be evaluated in multiple projections. The patients in these two groups were analyzed for differences in their preoperative status and the indications for their surgical procedures. The findings are represented in Table I. It should be noted that there were fewer patients in the second group who had suffered strokes before surgery (13.5% in group 2 versus 19.7% in group i). This difference was 239
Journal of VASCULAR SURGERY
240 Roon and Hoogerwerf
Table I. Preoperative assessment of patients in each group Percent male Percent diabetic Percent previous MI Indication for surgery Asymptomatic TIA CVA Amaurosis fugax
Group I
Group 2
56.7% 19.7% 17.8%
56.8% 18.9% 16.3%
22.3% 31.8% 19.7% 5.9%
26.9% 35.70,6 13.5% 9.3%
M/, Myocardial infarction; T/A, transient ischemic attack; CVA, cerebrovascular accident.
statistically significant by chi-square analysis (p < 0.01). The remaining factors evaluated, as well as the indications for the surgical procedures, showed minor differences that demonstrated no significance by statistical analysis. Surgical procedures included primary and secondary operations for carotid stenosis. The surgical procedures were all performed with general anesthesia. An arterial catheter was inserted before the induction of anesthesia to continuously measure arterial pressure. Carotid shunts were used on almost every operation. Initially Javid shunts were used exclusively. Our present method is to select the appropriate sized Argyle shunt and to use no clamps on the carotid artery to hold it in place. Arteriotomy repairs were done without patches unless the artery appeared small, the operation was for recurrent stenosis, or if a stenosis was noted on operative arteriography. At this time we are using polytetrafluoroethylene (PTFE) exclusively for carotid patches. Patch angioplasty was done in only 41 patients, 5.1% of the patients in group i and 6.2% of the patients in group 2. When the arteriotomy repair has been completed, an arteriogram is clone by clamping the common carotid and instilling 8 to 10 ml of contrast into the common carotid artery above the clamp with use of a 21-gauge scalp vein needle. One or more views of the carotid bifurcation and distal vessels are obtained to evaluate the carotid artery for residual stenosis and assure the stability of the end point of the endarterectomy. We drain all wounds with suction drainage. After operation the patient's blood pressure is controlled with fluids, inotropic agents, and antihypertensives to keep the pressure in the range of 100 to 170 mm Hg. All patients who can tolerate it are placed on 100 mg baby aspirin twice a day and 50 mg dipyridimole twice a day. Patients are observed for one night in an intensive care unit and one night on a surgical floor, and they are usually discharged from the hospital 2 days after operation.
All patients in this series were asked to return for routine follow-up noninvasive tests on a yearly basis. In group 1 61% of patients had longitudinal followup, and in group 2 94% had routine follow-up. In the early part of our experience we used the ocular pneumoplethysmography (OPG)-Gee device. Since 1983 a duplex scanner has been used to reassess the carotid bifurcation. Reoperations have been performed for asymptomatic preocclusive recurrent stenosis and for patients with flow-significant stenosis and focal hemispheric symptoms. RESULTS
The results of the procedures in the two groups were evaluated for postoperative stroke, death, and recurrent stenosis requiring reoperation. The comparison of the results in the two groups can be seen in Table II. There were three deaths out of 157 patients (1.9%) in group 1 and two deaths out of 535 patients (0.4%) in group 2. Although there is a difference between the two groups, it is not a statistically significant difference according to chi-square analysis. Two of the three deaths in group 1 were due to stroke and one was due to hemorrhage from the operative site. In group 2 one death was a result of cerebral hemorrhage in the contralateral hemisphere, and one was a result of myocardial infarction. There was a statistically significant difference (p < 0.01) between the number of patients in each series having permanent neurologic deficits (lasting greater than 30 days): seven patients (4.5%) in group 1 and six patients (1.1%) in group 2 suffered permanent ischemic neurologic deficits or died of stroke. Of note is that in the last 400 patients the incidence of neurologic deficit was 0.5%. Of the two strokes that occurred in these 400 patients, one was the result of cerebral hemmorhage after endarterectomy of the contralateral carotid artery. The incidence ofischemic infarction in the last 400 patients was then 0.25%. Because significantly more patients were in the first cohort than the second with a preoperative history of stroke, we stratified the series into those who had suffered a preoperative stroke and those who had not. The results can be seen in Table III. Thirty-one of the patients in group 1 had suffered a preoperative stroke. Four of these patients (14.8%) had a cerebral infarct with surgery. Only two of the 72 patients in group 2 who had suffered a preoperative infarct had strokes with surgery (2.8%). This difference was statistically significant by chi-square analysis (p < 0.05). Although a trend indicated a reduction in postoperative stroke from 2.4% to 0.9% in those patients in groups 1 and 2 who had not suffered a preoperative stroke, the difference was not statistically significant in this series.
Volume 16 Number 2 August 1992
The use of patch angioplasty has been evaluated for its effectiveness in preventing stroke and recurrent stenosis. None of our patients who were treated with patch closure had postoperative strokes. The difference between this portion of the series and the remainder was not statistically different. Since the percentage of patients treated with a patch in groups 1 and 2 were nearly identical, it is unlikely that this had any influence on the results. Although none of our patients who had patches placed had strokes, one patient who had a vein patch angioplasty required reoperation for a third time with a PTFE patch and has subsequently not had a recurrence. The number of patients in this series with patch angioplasty is not large enough to derive statistically relevant information about its use in preventing stroke or recurrent stenosis.. To determine whether the reduction in the incidence of stroke was due to arteriography or the "learning curve effect," the results of the first 100 patients in group 2 were compared with the last 100 patients in group 2. In the first 100 patients four cerebral infarctions (4.0%) occurred, and in the last 100 patients there were none (0%). This difference was not statistically significant by the chi-square method. A statistically significant difference occurred between the number of patients in group 1 and group 2 needing reoperation: six patients in group 1 (3.8%) and four patients in group 2 (0.8%) (p < 0.01). The mean interval between the primary and secondary operations was 19 months (23.8 months in group 1 and 9 months in group 2). Although there may seem to be a major difference in the mean intervals, the median intervals were 13 months in group 1 and 9 months in group 2. The shorter intervals in group 2 may be related to our frequency of follow-up with use of the duplex scanner versus the OPG-Gee machine and the sensitivities of the two types of' instruments. One patient in group 1 had a 108-month interval between the primary and secondary operations. The findings in this case were that the patient had recurrent atherosclerosis. All of the remaining operations were for fibrointimal hyperplasia. When the last 88 patients in the series were eliminated to evaluate only those patients in whom we had a 24-month follow-up period, there was still a significant difference between groups 1 and 2 in the number of patients requiring reoperation @ < 0.02). Arteriography did appear to influence the outcome of surgery directly as well as indirectly. In the 535 patients in group 2 who had intraoperative arteriography, 11 abnormal arteriograms demonstrated a defect likely to occlude the internal carotid
lntraoperative arteriography and carotid ,surgery 241
Table II. Results of surgical procedures Percent death Percent stroke Percent reoperation
Group 1
Group 2
1.9% 4.5% 3.8%
0.4% 1.1% 0.8%
Table III. Effect of preoperative history of stroke on outcome of carotid endarterectomy
Group 1 No. of patients Postoperativestrokes Strokes and deaths Group 2 No. of patients Postoperativestrokes Strokes and deaths
Patients with prior CVA
Patients with no CVA
31 4 (14.8%) 4 (14.8%)
126 3 (2,4%) 3 (2.4%)
72 2 (2.8%) 3 (4,3%)
463 4 (0.9%) 5 (1.I%)
artery (2.1% of the arteriograms). Because of the abnormal arteriograms the arteries were reopened, the defect was corrected, and the arteriotomy was then closed with a patch. The abnormalities included five angulations causing stenosis, three stenoses resulting from small vessels with tight arteriotomy closures, and two flaps. Some of the patients had minimal defects noted on arteriography that were corrected by further dissection of the carotid artery to release adventitial adhesions. The number of patients who had these minor corrections was not recorded, In comparison with group 1, there was a 3% reduction in stroke and a 3 % reduction in reoperation in the second series. It would appear that many of the "recurrent" stenoses noted in the first group may,well have been residual unrecognized stenoses inadequately treated by the initial operation. There were no obvious compfications from the use ofangiography. Initially the procedure took 10 to 15 minutes because of the use of x-ray film and development. At present we are using digital fluoroscopy with an image intensifier. The procedure now takes 5 to 10 minutes. Further advantages of this procedure are the ability to easily obtain multiple views of the operative site. Continuous observation after the injection of contrast is used to determine if there are any flow-restricting lesions within the vessel. An illustrative case is presented to demonstrate the utility of continuous fluoroscopy in comparison with taking single-film operative angiograms, CASE REPORT
The patient is a 54-year-old man with right hemispheric transient ischemic attacks and an 80% to 90% right
242 Roon and Hoogerwerf
~ournal of VASCULAR SURGERY
Fig. 1. Early image in a sequence of inttaoperative completion arteriograms demonstrates an impediment to flow of contrast up the internal carotid artery.
Fig. 2. Later film in the sequence of completion intraoperative arteriograms obscures findings of defect seen initially.
internal carotid stenosis noted on arteriography. A standard carotid endarterectomy was performed. The distal end point appeared to be adherant to the arterial wall on initial closure of the arteriotomy. Two sequential completion arteriogram films are shown (Figs. 1 and 2). The first image in the sequence (Fig. 1) demonstrates the contrast being restricted in passage up the internal carotid artery. The second image (Fig. 2) shows the contrast now progressing past this point up the carotid artery. Had this been a solitary image the abnormality might not have been noticed. This patient's arteriotomy was reopened, and a large flap was found and removed. The arteriotomy was subsequently closed with a patch. The subsequent arteriogram (Fig. 3) demonstrated an adequate result. The use of an image intensifier fluoroscopic unit in surgery may have resulted in better results in the latter part of our serieS.
this prophylactic procedure is purely the difference between the morbidity of the procedure and the morbidity of the carotid artery lesion. Since the margins are relatively small, every effort should be made to eliminate the possibility of complications from carotid operations. Chambers and Norris 6 have indicated that it is necessary to have a combined morbidity and mortality rate less than 5% in order to derive a preventive effect from carotid artery surgery. It would seem reasonable to try to assess the end result of this operation before leaving the operating room. Intuitively this would be the most direct approach to reducing the number of unfortunate results. It is always surprising how few surgeons have adopted this method. Although the potential dangers of air embolus, contrast media reaction, and the like, have been considered, we have not seen any increase in unexplained morbidity with the use of this technique. This study suffers from several obvious limitations: the patients were not randomized; this was a longitudinal series, with the poorest results seen
DISCUSSION
To decide on the justification for surgical intervention in the carotid system, one must know the morbidity and mortality rates of the procedure in each hospital and with each surgeon. The benefit of
Volume 16 Number 2 August 1992
Fig. 3. Completion arteriogram shows resolution of previously noted defect after reclosure and patch angioplasty.
in the earliest part of the study; and, during the study there were certain changes in technique. Despite these real deficiencies, we did show that intraoperarive arteriography will demonstrate unsuspected problems that can be corrected before inadvertant carotid occlusion and stroke. At this point it does not seem justified to randomize patients to prove whether the differences remain statistically significant between two groups of patients. It is our belief that the use of intraoperative assessment of surgical results has been shown to be associated with a lower morbidity of carotid surgery in our practice. With the use of an intraoperative digital fluoroscopic device, the surveillance of our operative results has become more easily accomplished and more complete. This has subsequently diminished the number of patients who required reoperation. The use of operative angiography has alerted us to the pitfalls in technique that can be avoided. This may account for the decline in morbidity in the course of the series.
Intraoperative arteriography and carotid surgery 243
Other methods of intraoperative assessment have been used, including continuous-wave Doppler, duplex scanning, and angioscopy, s,7,8,9 Continuouswave Doppler is not sensitive enough to reveal an unstable flap or stenotic closure that is not preocclusive. We have tried duplex scanning but it has been very cumbersome and time consuming. The sensitivity of the instrument is very good, but it is hard to evaluate small defects and their significance. We have had no experience with angioscopy, but it is unlikely that this will adequately demonstrate the defects caused by the arteriotomy closure. Certain defects, such as kinks, may not be evident until the edges of the arteriotomy are reapproximated and blood flow reestablished. Angioscopy may also result in arterial trauma or embolJzation if one is not very adept at the technique. The use of completion angiography is considered a standard practice with femoropopLiteal reconstruction but it has never been universally accepted in carotid surgery. This is the case despite the fact that the brain is a much more sensitive end organ than the leg. The results of this series of patients support ~ e use of intraoperative arteriography, since it can eliminate unsuspected problems that might otherwise result in stroke and death. The perceived potential risks of the technique have not materialized in our series. Routine surveillance can be accomplished without great diffimlty or significant morbidity. REFERENCES 1. North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effect of carotid endarterectomy in syanptomatic patients with high-grade carotid stenosis. N Engl J Med 1991;325:445-53. 2. Blaisdell FW, Lim R, Hall AD. Technical result of carotid endarterectomy. Am J Surg 1967;114:239-46. 3. Alpert l, Brener BJ, Parsonnet V, et al. Carotid endarterectomy and completion contact arteriography. J VAsc SurvG 1984;1:548-54. 4. Courbier R, Jausseran 1, Reggi M, et al. Routine intraoperadve carotid arteriography: its impact on operative morbidity and carotid restenosis. J VAsc SuR~ 1986;3:343-50. 5. Zierler R~E,Bandyk DF, Thiele BL. Inn'aoperative assessment of carotid endarterectomy. ~ VASC SuRe 1984;1: 73-83. 6. Chambers BR, Norris DV. The case against surgm7 for asymptomatic carotid stenosis. Stroke 1984;15:964-7. 7. Gonzalez LL, Partusch L, WitrAiP. Noninvasive carotid artery evaluation following endarterectomy. 1 VASC S~It~G1984;1: 403-8. 8. Olcott C. Clinical applications of video angioscopy. J VAsc SURG 1987;5:664-6. 9. Mehigan JT, Olcott C. Video angioscopy as an alternative to angiography. Am J Surg 1986;152:139-45.
Submitted Dec. 2, 1991; accepted Feb. 3, 1991.
In the recent past many studies have been done on the efficacy of carotid surgery in the prevention of stroke. Surgical intervention has been shown to be effective when the morbidity of surgery is less than the morbidity of the lesion. 1 Despite this intense reevaluation of carotid surgery, most surgeons are not assessing the adequancy of their surgical results with intraoperative completion arteriography or any other form of intraoperative imaging. Completion carotid arteriography is not a new concept. In 1967 BlaisdeU et al.2 demonstrated that in 25% of cases an unsuspected defect was noted on completion arteriograms. Other authors have confirmed unsuspected defects on routine intraoperative arteriography, a-5 To diminish the incidence of neurologic complications in carotid surgery in our practice we have instituted a policy of routine intraoperative carotid arteriography. The results of the entire series of carotid surgery patients seen in our clinical practice of From the Divisionof VascularSurgery,the EverettClinic,Everett. Dr. Roon is assistantclinicalprofessorof surgeryat the University of WashingtonSchool of Medicine, Seattle. Presented at the Ninth AnnualMeeting of the PacificNotxhwest Vascular Society,Portland, Ore., Nov. 7, 1991. Reprint requests: Anthony I. Roon, MD, Division of Vascular Surgery, the Everett Clinic, 1330 RockefellerAve., Suite 520, Everett, WA 9820i. 24/1/36909
peripheral vascular surgery is presented to demonstrate the effectiveness of this technique in the prevention of immediate and late complications of carotid surgery. PATIENTS AND METHODS All carotid operations performed between August 1979 to May 1991 were included in this series. There were 692 carotid operations. The operations were all performed by the first author (A.J.R.) in._a community hospital setting. In the first 157 operations (group 1) no intraoperative arteriography was performed. In an attempt to better assess the operative results, the subsequent 535 patients (group 2) had arteriography on completion of the endarterectomy. This was initially done with plain x-ray film, but more recently digital angiography has been done with a portable image intensifier. This change in technique has reduced the amount of time required for evaluation and allowed the carotid bifurcation to be evaluated in multiple projections. The patients in these two groups were analyzed for differences in their preoperative status and the indications for their surgical procedures. The findings are represented in Table I. It should be noted that there were fewer patients in the second group who had suffered strokes before surgery (13.5% in group 2 versus 19.7% in group i). This difference was 239
Journal of VASCULAR SURGERY
240 Roon and Hoogerwerf
Table I. Preoperative assessment of patients in each group Percent male Percent diabetic Percent previous MI Indication for surgery Asymptomatic TIA CVA Amaurosis fugax
Group I
Group 2
56.7% 19.7% 17.8%
56.8% 18.9% 16.3%
22.3% 31.8% 19.7% 5.9%
26.9% 35.70,6 13.5% 9.3%
M/, Myocardial infarction; T/A, transient ischemic attack; CVA, cerebrovascular accident.
statistically significant by chi-square analysis (p < 0.01). The remaining factors evaluated, as well as the indications for the surgical procedures, showed minor differences that demonstrated no significance by statistical analysis. Surgical procedures included primary and secondary operations for carotid stenosis. The surgical procedures were all performed with general anesthesia. An arterial catheter was inserted before the induction of anesthesia to continuously measure arterial pressure. Carotid shunts were used on almost every operation. Initially Javid shunts were used exclusively. Our present method is to select the appropriate sized Argyle shunt and to use no clamps on the carotid artery to hold it in place. Arteriotomy repairs were done without patches unless the artery appeared small, the operation was for recurrent stenosis, or if a stenosis was noted on operative arteriography. At this time we are using polytetrafluoroethylene (PTFE) exclusively for carotid patches. Patch angioplasty was done in only 41 patients, 5.1% of the patients in group i and 6.2% of the patients in group 2. When the arteriotomy repair has been completed, an arteriogram is clone by clamping the common carotid and instilling 8 to 10 ml of contrast into the common carotid artery above the clamp with use of a 21-gauge scalp vein needle. One or more views of the carotid bifurcation and distal vessels are obtained to evaluate the carotid artery for residual stenosis and assure the stability of the end point of the endarterectomy. We drain all wounds with suction drainage. After operation the patient's blood pressure is controlled with fluids, inotropic agents, and antihypertensives to keep the pressure in the range of 100 to 170 mm Hg. All patients who can tolerate it are placed on 100 mg baby aspirin twice a day and 50 mg dipyridimole twice a day. Patients are observed for one night in an intensive care unit and one night on a surgical floor, and they are usually discharged from the hospital 2 days after operation.
All patients in this series were asked to return for routine follow-up noninvasive tests on a yearly basis. In group 1 61% of patients had longitudinal followup, and in group 2 94% had routine follow-up. In the early part of our experience we used the ocular pneumoplethysmography (OPG)-Gee device. Since 1983 a duplex scanner has been used to reassess the carotid bifurcation. Reoperations have been performed for asymptomatic preocclusive recurrent stenosis and for patients with flow-significant stenosis and focal hemispheric symptoms. RESULTS
The results of the procedures in the two groups were evaluated for postoperative stroke, death, and recurrent stenosis requiring reoperation. The comparison of the results in the two groups can be seen in Table II. There were three deaths out of 157 patients (1.9%) in group 1 and two deaths out of 535 patients (0.4%) in group 2. Although there is a difference between the two groups, it is not a statistically significant difference according to chi-square analysis. Two of the three deaths in group 1 were due to stroke and one was due to hemorrhage from the operative site. In group 2 one death was a result of cerebral hemorrhage in the contralateral hemisphere, and one was a result of myocardial infarction. There was a statistically significant difference (p < 0.01) between the number of patients in each series having permanent neurologic deficits (lasting greater than 30 days): seven patients (4.5%) in group 1 and six patients (1.1%) in group 2 suffered permanent ischemic neurologic deficits or died of stroke. Of note is that in the last 400 patients the incidence of neurologic deficit was 0.5%. Of the two strokes that occurred in these 400 patients, one was the result of cerebral hemmorhage after endarterectomy of the contralateral carotid artery. The incidence ofischemic infarction in the last 400 patients was then 0.25%. Because significantly more patients were in the first cohort than the second with a preoperative history of stroke, we stratified the series into those who had suffered a preoperative stroke and those who had not. The results can be seen in Table III. Thirty-one of the patients in group 1 had suffered a preoperative stroke. Four of these patients (14.8%) had a cerebral infarct with surgery. Only two of the 72 patients in group 2 who had suffered a preoperative infarct had strokes with surgery (2.8%). This difference was statistically significant by chi-square analysis (p < 0.05). Although a trend indicated a reduction in postoperative stroke from 2.4% to 0.9% in those patients in groups 1 and 2 who had not suffered a preoperative stroke, the difference was not statistically significant in this series.
Volume 16 Number 2 August 1992
The use of patch angioplasty has been evaluated for its effectiveness in preventing stroke and recurrent stenosis. None of our patients who were treated with patch closure had postoperative strokes. The difference between this portion of the series and the remainder was not statistically different. Since the percentage of patients treated with a patch in groups 1 and 2 were nearly identical, it is unlikely that this had any influence on the results. Although none of our patients who had patches placed had strokes, one patient who had a vein patch angioplasty required reoperation for a third time with a PTFE patch and has subsequently not had a recurrence. The number of patients in this series with patch angioplasty is not large enough to derive statistically relevant information about its use in preventing stroke or recurrent stenosis.. To determine whether the reduction in the incidence of stroke was due to arteriography or the "learning curve effect," the results of the first 100 patients in group 2 were compared with the last 100 patients in group 2. In the first 100 patients four cerebral infarctions (4.0%) occurred, and in the last 100 patients there were none (0%). This difference was not statistically significant by the chi-square method. A statistically significant difference occurred between the number of patients in group 1 and group 2 needing reoperation: six patients in group 1 (3.8%) and four patients in group 2 (0.8%) (p < 0.01). The mean interval between the primary and secondary operations was 19 months (23.8 months in group 1 and 9 months in group 2). Although there may seem to be a major difference in the mean intervals, the median intervals were 13 months in group 1 and 9 months in group 2. The shorter intervals in group 2 may be related to our frequency of follow-up with use of the duplex scanner versus the OPG-Gee machine and the sensitivities of the two types of' instruments. One patient in group 1 had a 108-month interval between the primary and secondary operations. The findings in this case were that the patient had recurrent atherosclerosis. All of the remaining operations were for fibrointimal hyperplasia. When the last 88 patients in the series were eliminated to evaluate only those patients in whom we had a 24-month follow-up period, there was still a significant difference between groups 1 and 2 in the number of patients requiring reoperation @ < 0.02). Arteriography did appear to influence the outcome of surgery directly as well as indirectly. In the 535 patients in group 2 who had intraoperative arteriography, 11 abnormal arteriograms demonstrated a defect likely to occlude the internal carotid
lntraoperative arteriography and carotid ,surgery 241
Table II. Results of surgical procedures Percent death Percent stroke Percent reoperation
Group 1
Group 2
1.9% 4.5% 3.8%
0.4% 1.1% 0.8%
Table III. Effect of preoperative history of stroke on outcome of carotid endarterectomy
Group 1 No. of patients Postoperativestrokes Strokes and deaths Group 2 No. of patients Postoperativestrokes Strokes and deaths
Patients with prior CVA
Patients with no CVA
31 4 (14.8%) 4 (14.8%)
126 3 (2,4%) 3 (2.4%)
72 2 (2.8%) 3 (4,3%)
463 4 (0.9%) 5 (1.I%)
artery (2.1% of the arteriograms). Because of the abnormal arteriograms the arteries were reopened, the defect was corrected, and the arteriotomy was then closed with a patch. The abnormalities included five angulations causing stenosis, three stenoses resulting from small vessels with tight arteriotomy closures, and two flaps. Some of the patients had minimal defects noted on arteriography that were corrected by further dissection of the carotid artery to release adventitial adhesions. The number of patients who had these minor corrections was not recorded, In comparison with group 1, there was a 3% reduction in stroke and a 3 % reduction in reoperation in the second series. It would appear that many of the "recurrent" stenoses noted in the first group may,well have been residual unrecognized stenoses inadequately treated by the initial operation. There were no obvious compfications from the use ofangiography. Initially the procedure took 10 to 15 minutes because of the use of x-ray film and development. At present we are using digital fluoroscopy with an image intensifier. The procedure now takes 5 to 10 minutes. Further advantages of this procedure are the ability to easily obtain multiple views of the operative site. Continuous observation after the injection of contrast is used to determine if there are any flow-restricting lesions within the vessel. An illustrative case is presented to demonstrate the utility of continuous fluoroscopy in comparison with taking single-film operative angiograms, CASE REPORT
The patient is a 54-year-old man with right hemispheric transient ischemic attacks and an 80% to 90% right
242 Roon and Hoogerwerf
~ournal of VASCULAR SURGERY
Fig. 1. Early image in a sequence of inttaoperative completion arteriograms demonstrates an impediment to flow of contrast up the internal carotid artery.
Fig. 2. Later film in the sequence of completion intraoperative arteriograms obscures findings of defect seen initially.
internal carotid stenosis noted on arteriography. A standard carotid endarterectomy was performed. The distal end point appeared to be adherant to the arterial wall on initial closure of the arteriotomy. Two sequential completion arteriogram films are shown (Figs. 1 and 2). The first image in the sequence (Fig. 1) demonstrates the contrast being restricted in passage up the internal carotid artery. The second image (Fig. 2) shows the contrast now progressing past this point up the carotid artery. Had this been a solitary image the abnormality might not have been noticed. This patient's arteriotomy was reopened, and a large flap was found and removed. The arteriotomy was subsequently closed with a patch. The subsequent arteriogram (Fig. 3) demonstrated an adequate result. The use of an image intensifier fluoroscopic unit in surgery may have resulted in better results in the latter part of our serieS.
this prophylactic procedure is purely the difference between the morbidity of the procedure and the morbidity of the carotid artery lesion. Since the margins are relatively small, every effort should be made to eliminate the possibility of complications from carotid operations. Chambers and Norris 6 have indicated that it is necessary to have a combined morbidity and mortality rate less than 5% in order to derive a preventive effect from carotid artery surgery. It would seem reasonable to try to assess the end result of this operation before leaving the operating room. Intuitively this would be the most direct approach to reducing the number of unfortunate results. It is always surprising how few surgeons have adopted this method. Although the potential dangers of air embolus, contrast media reaction, and the like, have been considered, we have not seen any increase in unexplained morbidity with the use of this technique. This study suffers from several obvious limitations: the patients were not randomized; this was a longitudinal series, with the poorest results seen
DISCUSSION
To decide on the justification for surgical intervention in the carotid system, one must know the morbidity and mortality rates of the procedure in each hospital and with each surgeon. The benefit of
Volume 16 Number 2 August 1992
Fig. 3. Completion arteriogram shows resolution of previously noted defect after reclosure and patch angioplasty.
in the earliest part of the study; and, during the study there were certain changes in technique. Despite these real deficiencies, we did show that intraoperarive arteriography will demonstrate unsuspected problems that can be corrected before inadvertant carotid occlusion and stroke. At this point it does not seem justified to randomize patients to prove whether the differences remain statistically significant between two groups of patients. It is our belief that the use of intraoperative assessment of surgical results has been shown to be associated with a lower morbidity of carotid surgery in our practice. With the use of an intraoperative digital fluoroscopic device, the surveillance of our operative results has become more easily accomplished and more complete. This has subsequently diminished the number of patients who required reoperation. The use of operative angiography has alerted us to the pitfalls in technique that can be avoided. This may account for the decline in morbidity in the course of the series.
Intraoperative arteriography and carotid surgery 243
Other methods of intraoperative assessment have been used, including continuous-wave Doppler, duplex scanning, and angioscopy, s,7,8,9 Continuouswave Doppler is not sensitive enough to reveal an unstable flap or stenotic closure that is not preocclusive. We have tried duplex scanning but it has been very cumbersome and time consuming. The sensitivity of the instrument is very good, but it is hard to evaluate small defects and their significance. We have had no experience with angioscopy, but it is unlikely that this will adequately demonstrate the defects caused by the arteriotomy closure. Certain defects, such as kinks, may not be evident until the edges of the arteriotomy are reapproximated and blood flow reestablished. Angioscopy may also result in arterial trauma or embolJzation if one is not very adept at the technique. The use of completion angiography is considered a standard practice with femoropopLiteal reconstruction but it has never been universally accepted in carotid surgery. This is the case despite the fact that the brain is a much more sensitive end organ than the leg. The results of this series of patients support ~ e use of intraoperative arteriography, since it can eliminate unsuspected problems that might otherwise result in stroke and death. The perceived potential risks of the technique have not materialized in our series. Routine surveillance can be accomplished without great diffimlty or significant morbidity. REFERENCES 1. North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effect of carotid endarterectomy in syanptomatic patients with high-grade carotid stenosis. N Engl J Med 1991;325:445-53. 2. Blaisdell FW, Lim R, Hall AD. Technical result of carotid endarterectomy. Am J Surg 1967;114:239-46. 3. Alpert l, Brener BJ, Parsonnet V, et al. Carotid endarterectomy and completion contact arteriography. J VAsc SurvG 1984;1:548-54. 4. Courbier R, Jausseran 1, Reggi M, et al. Routine intraoperadve carotid arteriography: its impact on operative morbidity and carotid restenosis. J VAsc SuR~ 1986;3:343-50. 5. Zierler R~E,Bandyk DF, Thiele BL. Inn'aoperative assessment of carotid endarterectomy. ~ VASC SuRe 1984;1: 73-83. 6. Chambers BR, Norris DV. The case against surgm7 for asymptomatic carotid stenosis. Stroke 1984;15:964-7. 7. Gonzalez LL, Partusch L, WitrAiP. Noninvasive carotid artery evaluation following endarterectomy. 1 VASC S~It~G1984;1: 403-8. 8. Olcott C. Clinical applications of video angioscopy. J VAsc SURG 1987;5:664-6. 9. Mehigan JT, Olcott C. Video angioscopy as an alternative to angiography. Am J Surg 1986;152:139-45.
Submitted Dec. 2, 1991; accepted Feb. 3, 1991.
