Comparison of Microembolism Detected by Transcranial Doppler and Neuropsychological Sequelae of Carotid Surgery and Percutaneous Transluminal Angioplasty
Background and Purpose—Percutaneous transluminal angioplasty (PTA) is currently being assessed for the treatment of carotid stenosis. In comparison with carotid endarterectomy (CEA), there is evidence of an increased risk of cerebral microembolism during the procedure. We have sought evidence of any neuropsychological sequelae of carotid PTA and compared it with CEA to demonstrate the relative safety of the 2 treatment options.
Methods—The neuropsychological outcomes after CEA and PTA were compared in 2 matched groups of patients with severe symptomatic carotid stenosis, 96% of whom had been randomized in the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS), at a single center. Transcranial Doppler insonation of the middle cerebral artery was used to measure cerebral reactivity in response to carbon dioxide inhalation before treatment and then to detect microembolization of the ipsilateral cerebral hemisphere and measure changes in blood flow velocity during the procedures. The performance on a neuropsychological test battery administered before, 6 weeks after, and 6 months after the procedure was compared in 20 patients undergoing PTA and 26 having CEA.
Results—At 6 weeks, 5 patients in each group showed a similar decline in neuropsychological performance; global measures showed no significant difference between the 2 procedures, despite a significantly higher incidence of microemboli during PTA. Both groups showed a marked reduction in anxiety after treatment.
Conclusions—The findings provide some reassurance that PTA is not associated with greater cerebral complications than CEA, despite the higher embolic load recorded by transcranial Doppler ultrasonography during angioplasty.
Severe symptomatic carotid stenosis is associated with a significant risk of subsequent ipsilateral stroke when treated medically. Multicenter trials have shown that surgical treatment by carotid endarterectomy (CEA) significantly reduces this risk.1 2 Carotid percutaneous transluminal angioplasty (PTA) is being assessed as an alternative procedure to carotid surgery.3 The advantages of carotid PTA over CEA include a shorter hospital stay, avoidance of a general anesthetic, and avoidance of an incision in the neck, which may be complicated by hematoma formation, wound infection, and cranial nerve palsies, although these are frequently mild when they do occur.4
Various studies have addressed the effect of CEA on cognitive function. We have recently reviewed these.5 Just over 50% of the studies reported an improvement in cognition after CEA. The remainder found no change. Heyer et al6 studied 112 patients undergoing elective CEA. They reported that although some neuropsychological test scores improved after surgery, 80% of patients showed a decline in 1 or more tests. They speculated that this was due to global hypoperfusion and that improvement during follow-up, which they also documented, was probably due to relief of hemodynamically significant stenosis. They recommended, however, that in future studies the role of microembolism also be studied.
There is no fully published study of the effect of carotid PTA on cognitive function. Carotid PTA has been criticized on the grounds that the risk of stroke during the procedure may be similar to the risk with CEA, and PTA might result in more neuropsychological deficit7 8 from hemodynamic ischemia during balloon inflation and microembolism during disruption of plaque. However, we have previously demonstrated that carotid PTA results in significantly less hemodynamic ischemia during the procedure than does CEA.9 On the other hand, we did confirm that carotid PTA causes significantly more cerebral microembolism (as inferred from transcranial Doppler [TCD] evidence of more microembolic signals) than CEA.9 10 In the context of cardiac surgery, there is good evidence that neuropsychological outcome is related to microembolic load11 ; it therefore remained possible that microembolism during carotid PTA would cause disproportionate neuropsychological deficit.
Cerebral hemodynamics may be impaired distal to a stenosed carotid artery, although compensatory mechanisms such as increased cerebrovascular responsiveness and sufficient collateral flow may normalize flow. Some patients, however, do not appear to have sufficient compensatory mechanisms. This situation can be assessed by measuring the increase in cerebral blood flow in response to an increase in inspired CO2 concentration.12 13 Impaired reactivity to hypercapnia is improved by both CEA and PTA.14 15 16 The relationship of cerebral reactivity to neurological complications occurring during carotid PTA is unknown, but Thiel et al15 showed no association between impaired preoperative cerebral reactivity and stroke during CEA. There are no published data comparing cerebral reactivity with neuropsychological deficit after CEA or carotid PTA.
We have therefore studied the effect of CEA and carotid PTA on cognition and compared any changes to embolic load and hemodynamic ischemia sustained during the procedure. We have also compared changes in cognition with cerebral reactivity measured before the procedure.
Subjects and Methods
Fifty patients were recruited from those undergoing CEA or carotid PTA at St George’s or Atkinson Morley’s Hospitals, London, between September 1995 and January 1997. All had severe (>70%) symptomatic carotid stenosis. Forty-eight were randomized between the 2 procedures as part of the Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS). Exclusion criteria included patient refusal and inability to obtain a TCD signal. Seventeen such patients were excluded during the period of recruitment. Four patients had to be withdrawn from the study: 1 had a major stroke unrelated to the procedure, 1 developed cerebral metastases, 1 developed a primary malignancy requiring major surgery, and test results were incomplete in 1 because of equipment failure. Of the remaining 46 patients, 7 patients in the PTA randomized group had preprocedural ipsilateral symptoms of stroke, 5 of transient ischemic attack (TIA), and 5 of amaurosis fugax, compared with 9 of stroke, 8 of TIA, and 7 of amaurosis fugax in the CEA group. Neuropsychological testing was performed on 26 patients before CEA and 20 before carotid PTA and was repeated at 6 weeks and at 6 months after the procedure. A battery of standard neuropsychological tests was performed on each occasion by 1 of 2 neuropsychologists at the Middlesex Hospital, London (J.S., S.L.). These tests were chosen because our previous work had shown that they were sensitive to the effects of microembolism during cardiac surgery.17 Measures of attention and concentration (Trail-Making A and B, Letter Cancellation, Two Choice Reaction Time, Symbol Digit, Displaced Reaction Time), verbal memory (Rey Auditory Verbal Learning Test), nonverbal memory (Nonverbal Memory Test), visuomotor skills (Finger Tapping, Grooved Pegboard) were administered both before and after the procedure. Assessments of mood state were also conducted before and after the intervention and at each testing session; the Beck Depression Inventory was used as an indication of depressed mood, and the Spielberger State Anxiety Inventory was employed to assess state anxiety (Table 1⇓). Premorbid intelligence was assessed before the procedure by the National Adult Reading Test. Current IQ was estimated by 3 subsets of the Wechsler Adult Intelligence Scale (Vocabulary, Picture Completion, Block Design). The Spielberger Trait Anxiety Inventory was administered on the first visit only to assess dispositional anxiety. Only 4 patients were left-handed. The neuropsychologist was blinded to the procedure as much as possible, but a surgical incision made this difficult.
TCD (Pioneer 4040, Nicolet-EME) was used to estimate cerebral reactivity before the procedure in all individuals.12 16 The middle cerebral artery (MCA) was insonated. The patient was then instructed to breathe room air while exhaling via a mask into an end-tidal CO2 monitor (Normocap 200, Datex). When the end-tidal CO2 and MCA blood velocity had both been steady for 2 minutes, the mean MCA blood velocity was recorded. The patient then inhaled a mixture of air and 8% CO2. Mean MCA blood flow velocity was recorded when end-tidal CO2 had been stable for 2 minutes. The ability of the cerebral vasculature to vasodilate was assessed by the change in mean MCA blood flow velocity per kilopascal change in end-tidal CO2 (cm/s per kilopascal).
Normal cerebral reactivity was defined by studying cerebral reactivity in each hemisphere in a group of 9 volunteers of similar age (mainly spouses) without cerebrovascular events and in whom carotid Doppler studies were normal. Patients were defined as having impaired ipsilateral cerebral reactivity if reactivity measured in the MCA ipsilateral to the carotid being treated was >1 SD below the mean in the control group. The control mean was 11 cm/s per kilopascal and the SD 5.5 cm/s per kilopascal.
Patients were all monitored with TCD during either CEA or carotid PTA. Monitoring was started before the procedure and continued for 20 minutes after the procedure whenever possible. The TCD tracing was recorded onto digital audiotape for subsequent analysis. Recordings during CEA were divided into a period of dissection (before the artery was opened, when all microembolic signals must be solid) and later (after the artery was opened, when microembolic signals could be air or solid). The use of a shunt was recorded. The times at which contrast was injected during PTA were recorded, and all times when movement of the catheter produced signals that sounded and looked like potential microembolic signals were also noted. The time of balloon inflation and deflation was recorded. Mean MCA velocity at the start of either procedure was considered the baseline velocity.
Ischemic time was defined as the period for which the mean MCA flow velocity was reduced to less than a third of baseline, either by clamp without shunt, shunt malfunction, excess pressure on the carotid artery, or cardiovascular changes during CEA, or by catheter obstructing the internal carotid artery lumen, balloon inflation, or cardiovascular changes during PTA.
Microembolic signals were identified as previously defined.18 Contrast injection during PTA results in a shower of microembolic signals that cannot be counted individually.19 Therefore, only microembolic signals that were detected in the MCA while the catheter or balloon was being manipulated (including balloon deflation) were included.
All tapes were analyzed with the investigator blinded to the name of the patient and stage of the procedure. It was not possible to be blinded as to whether the procedure was carotid PTA or CEA since the TCD recording showed frequent contrast injections during carotid PTA, which did not occur during CEA. Intraobserver and interobserver variability in tape interpretation was checked, as previously described.9
CEA was performed under general anesthesia by 1 of 2 vascular surgeons. One used a shunt routinely, the other selectively. One used a patch routinely. Both directed flow into the external carotid on restitution of flow to minimize passage of debris into the cerebral circulation. PTA was performed by 2 radiologists working together. A guidewire was introduced into the origin of the internal carotid artery from the femoral artery in the groin under x-ray control. It was then exchanged for a flexible guidewire, which was manipulated across the stenosis. A balloon was then passed over the guidewire to lie across the stenosis, where it was inflated manually. Thereafter, an angiogram was performed to assess the result. Balloon dilation was repeated up to 5 times until a satisfactory result was detected. In 1 case a stent was inserted when it proved impossible to achieve an adequate dilation of the vessel.
Independent t tests were performed to compare baseline characteristics such as IQ, mood state, neuropsychological performance, and CO2 reactivity. Paired t tests were used to compare mood state before and after the procedure. The normality of distribution of each of the neuropsychological tests was examined by means of the Kolmogorov-Smirnov goodness-of-fit test. Trail-Making A, Two-Choice Reaction Time Test, Displaced Reaction Time, Letter Cancellation, and the Grooved Pegboard Dominant and Nondominant were found to not be normally distributed. Square root transformations rendered these tests to a normal distribution.
We have applied a conventional definition of neuropsychological deficit that has been widely used in studying the impact of cardiac surgery.17 A SD unit for each test is computed from all the preoperative scores. A deficit occurs in a test when the patient’s postoperative score has dropped by ≥1 SD unit from the preoperative score. For the deficit to be significant, it must occur in ≥2 tests.
This type of deficit analysis is, however, relatively insensitive because it collapses continuous measures into 2 groups (deficit and nondeficit). Therefore, as has been suggested elsewhere,20 a measure of the overall postprocedural change (reflecting both the potential preserved learning ability on the neuropsychological tests coupled with the potential deterioration on the tests) was applied to the data. To perform this analysis, a change score was calculated for each subject’s score for each test by subtracting the postprocedural score from the preprocedural performance. By dividing these difference scores by the SD of the preprocedural group performance of all patients in the study, these differences were converted into a standardized change score. These standardized scores therefore reflect the relative change in performance from before to after surgery or angioplasty. Improved performance in any test is revealed by a lower score. If necessary, the directional data were reversed so that all improvements gave rise to positive differences. Because of the potential for all neuropsychological tests to show learning with repetition, a greater improvement in neuropsychological scores on this analysis will reflect a combination of greater learning and less deficit.21 22
The sum of all the standardized change scores between before the procedure and 6 weeks after the procedure and between before the procedure and 6 months after the procedure was then correlated with the following: (1) the number of presumed solid microembolic signals detected during the procedure (ie, the number of microembolic signals detected during dissection during CEA or catheter movement and balloon deflation during PTA); (2) total ischemic time; and (3) ipsilateral cerebral reactivity before the procedure.
Cerebral reactivity correlations were performed with the use of the actual cerebral reactivity initially. Cognitive change was then compared between individuals with normal cerebral reactivity and individuals with abnormal cerebral reactivity.
Ethical Approval for TCD and Neuropsychological Studies
Ethical approval was obtained from the district ethics committee. Each patient was given an information sheet, and consent for these studies was obtained from all patients.
Baseline characteristics, including age, sex, vascular risk factors, and clinical presentation of the patients, are shown in Tables 2⇓ and 3⇓. There were no significant differences between the group treated with CEA and that treated with carotid PTA. Data were collected at 2 additional time points after the procedure in 40 patients. In 6, data were missing from either the 6-week or 6-month follow-up. The right side was treated in 24 patients (16 CEA, 8 PTA) and the left in 22 (10 CEA, 12 PTA). One patient in the PTA group had a minor left hemisphere stroke at the time of the procedure but was able to complete the protocol. No patient in the CEA group had a periprocedural stroke.
Table 3⇑ shows the National Adult Reading Test raw scores, together with Verbal, Performance, and Full-Scale IQ prorated from the tests performed. In addition, the scores for anxiety and depression are shown. There were no intergroup differences found in these preprocedural assessments. Mean CO2 reactivity was not significantly lower in the surgery group, but more operated patients were deemed to show abnormal reactivity compared with controls, although again the difference was not significant (14/26 versus 4/20; χ2=1.4).
The Beck Depression Inventory did not change from preprocedural assessments and did not differ between the groups at either follow-up time point. State anxiety, in contrast, showed significant reductions in both groups at 6 weeks, and this decline was maintained to 6 months (Figure⇓). There were no significant differences between the groups at any time point.
Table 4⇓ describes the neuropsychological scores for the 46 patients obtained at each assessment. No significant differences were found between the groups on any test at any of the assessment times. At the 6-week postprocedural neuropsychological assessment, 5 of the 20 patients in the PTA group and 5 of the 25 patients in the CEA group were classified as having deficits. Neuropsychological deficits were also found in 7 of the 18 patients in the PTA group and 4 of the 22 in the CEA group at the 6-month follow-up assessment. Table 5⇓ describes the standardized change scores obtained at the 6-week assessment for each task and standardized global change score. There were no significant differences between the 2 groups on any task or on the global score.
The standardized change scores for the 6-month assessment are shown in Table 6⇓. One significant difference was found with the Grooved Pegboard (dominant hand), with CEA patients performing better at the 6-month assessment than PTA patients. No significant difference was found between the groups on the global score.
A mean of 178 presumed solid emboli was detected by TCD in patients during PTA compared with a mean of 10 in patients undergoing CEA. An independent t test revealed that there was a significant difference (t=6.93, df=41, P<0.001) between the 2 procedures. If the number of emboli generated in the PTA group during balloon deflation is subtracted (mean, 40.2; SD, 93.6) to produce a figure representing definite solid emboli in PTA (ie, only during the insertion of the guidewire), then the difference between the groups (PTA=137.4; CEA=10) remains significant (t=6.6, df=41, P<0.001). There was no significant correlation between emboli and standardized neuropsychological change scores at either the 6-week (r=−0.001, P=1.0) or the 6-month (r=0.017, P=0.92) time periods.
Ten patients in the PTA group and 6 in the CEA group had measurable ischemic times. There was no evidence of a relationship between the presence or absence of such a hemodynamic event and neuropsychological test results as measured by standardized change scores at 6 weeks (t=0.88, df=39, P=0.38) or at 6 months (t =−0.79, df=35, P=0.44). There was also no correlation between cerebral CO2 reactivity and cognitive outcome (standardized change scores) when we adjusted for age.
Neuropsychological deficit may not be as apparent to the clinician as neurological deficit but can be similarly disruptive to the life of an individual. Neuropsychological deficit has been shown to occur after other vascular procedures; for example, a deficit can be detected in approximately 60% of patients 8 days after coronary artery bypass grafting (CABG) or after other major vascular surgery and is still present in up to 33% of patients up to 8 weeks after either procedure.6 23 The use of an arterial line filter during CABG has been shown to reduce both the microemboli that can be detected during the procedure and the neuropsychological deficit following it, suggesting that microemboli may contribute to such a deficit.11 If this is the case, carotid PTA may result in more neuropsychological deficit than CEA.
This study, however, has shown no difference in the incidence of neuropsychological impairment after PTA or CEA, despite the greater incidence of microembolic signals in the TCD recordings during angioplasty.9 10 The implication that microembolism during the procedure does not result in neuropsychological deficit after CEA or carotid PTA is supported by the lack of correlation between the numbers of emboli and the test scores. There is, however, a risk of a type II error because of the numbers involved.
It has recently been suggested24 that patients with low perfusion pressures may be more likely to develop symptoms from cerebral embolism because of impaired washout. We found no evidence of this effect in our 2 groups of patients. Thus, neither impaired hemodynamics before treatment, as assessed by CO2 cerebral reactivity, nor ischemia during the procedure, as assessed by a fall in MCA blood flow velocity, was associated with inferior cognitive outcome.
We know of only 1 other study of neuropsychological outcome after carotid PTA.25 This was performed concurrently with the present study and involved 116 patients also within CAVATAS. It too found no significant difference in neuropsychological performance between those having surgery or PTA, although the results are only available in abstract form at this time. Taken with the finding of the main CAVATAS trial (that PTA was associated with no greater stroke complication rate than CEA), these combined results provide reassurance that PTA is not responsible for any excess of cerebral complications compared with CEA. However, the number of patients in CAVATAS was relatively small, and the rate of stroke and death was 10% in both the surgical and angioplasty groups. This high complication rate has caused concern. It is likely that this resulted from the inclusion of patients with higher risks than most published surgical series. Thus, larger trials would be needed before carotid PTA should be introduced as standard therapy. Primary carotid stenting is rapidly replacing simple balloon angioplasty as the interventional technique of choice for the treatment of carotid stenosis, and it is possible that this will improve the safety of the procedure. The results of this study provide some reassurance that even if stenting is accompanied by a similar incidence of embolization,10 this is unlikely to be accompanied by serious neurological sequelae; however, catheterization of the aorta is involved, which may result in larger emboli. A study comparable to our own should thus be included in the assessment of the safety and efficacy of carotid stenting in order that any neuropsychological consequences may be detected early on in the use of this procedure.
Further devices are being developed to be placed in the distal carotid before stenting and PTA. Our studies suggest that the pore size of such filters or traps should be designed to prevent macroemboli related to the risk of stroke but not necessarily all microembolization.
The present results contrast with those of CABG. Previous studies of CABG from our group using similar methodology have shown that approximately 20% to 30% of patients show evidence of a deficit in neuropsychological test performance 6 weeks after surgery, which relates to the microembolic burden. The difference may relate to the numbers of emboli (higher during CABG), their constitution, the hemodynamic circumstances of bypass during the period of embolization, or the nature of the emboli, which may be larger when originating in the aorta, as would be the case in CABG.
The finding that patients showed a decline in levels of anxiety from before to after the procedure, which was maintained at the 6-month assessment, is consistent with much other work.26 27 The decline appears to reflect specific anxieties regarding the threat of the procedure and relief at the outcome.
This study was supported by a UK (North Thames) National Health Service (NHS) Research and Development grant. Dr Crawley was funded by a grant from the NHS Management Executive. CAVATAS was funded by the British Heart Foundation and the NHS Management Executive. The ultrasound laboratory was funded by grants from the Wellcome Trust and the Neurosciences Research Foundation. We thank Diana Colquhoun for assistance with the ultrasound measurements. Drs A. Clifton and T. Buckenham performed the carotid PTA procedures, and R.S. Taylor and T. Loosemore performed the CEAs.
- Received November 15, 1999.
- Revision received March 3, 2000.
- Accepted March 3, 2000.
- Copyright © 2000 by American Heart Association
Barnett HJM, Taylor DW, Eliasziw M, Fox AJ, Ferguson GG, Haynes RB, Rankin RN, Clagett GP, Hachinski VC, Sackett DL, Thorpe KE, Meldrum HE, for the North American Symptomatic Carotid Endarterectomy Trial Collaborators. Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. N Engl J Med. 1998;339:1415–1425.
Brown MM, for the CAVATAS Investigators. Results of the carotid and vertebral angioplasty study (CAVATAS) Cerebrovasc Dis. 1998; 8(suppl iv):21. Abstract.
Ferguson GG, Eliasziw M, Barr HWK, Clagett GP, Barnes RW, Wallace MC, Taylor DW, Haynes RB, Finan JW, Hachinski VC, Barnett HJM, for the North American Symptomatic Carotid Endarterectomy Trial (NASCET) Collaborators. The North American Symptomatic Carotid Endarterectomy Trial: surgical results in 1415 patients. Stroke. 1999;30:1751–1758.
Heyer EJ, Adams DC, Solomon RA, Todd GJ, Quest DO, McMahon DJ, Steneck SD, Choudhri TF, Connolly ES. Neuropsychological changes in patients after carotid endarterectomy. Stroke. 1998;29:1110–1115.
Crawley F, Clifton A, Buckenham T, Loosemore T, Taylor RS, Brown MM. Comparison of hemodynamic cerebral ischemia and microembolic signals detected during carotid endarterectomy and carotid angioplasty. Stroke. 1997;28:2460–2464.
Louwerse ES, Esselink RAJ, Ernst JMPG, Overtoom TTC, Van den Berg JC, Mauser HW, Bistervels JHGM, Mast EG, de Valois JC, Heesewijk JPM, Kelder JC, Morshuis WJ, Schepens MAAM, van de Pavoordt HDWM, Moll FL, Ackerstaff RGA. Transcranial Doppler monitoring in carotid angioplasty and stenting. Cerebrovasc Dis. 1999;9(suppl 2):7. Abstract.
Pugsley W, Klinger L, Paschalis C, Treasure T, Harrison M, Newman S. The impact of microemboli during cardiopulmonary bypass on neuropsychological functioning. Stroke. 1994;25:1393–1399.
Silvestrini M, Troisi E, Matteis M, Cupini LM, Caltagirone C. Transcranial Doppler assessment of cerebrovascular reactivity in symptomatic and asymptomatic severe carotid stenosis. Stroke. 1996;27:1970–1973.
Reivich M. Arterial pCO2 and cerebral hemodynamics. Am J Physiol. 1964;206:25–35.
Hartl WH, Janssen I, Furst H. Effect of carotid endarterectomy on patterns of cerebrovascular reactivity in patients with unilateral carotid artery stenosis. Stroke. 1994;25:1952–1957.
Markus HS, Clifton A, Buckenham T, Taylor RS, Brown MM. Improvement in cerebral hemodynamics after carotid angioplasty. Stroke. 1996;27:612–615.
Consensus Committee, Criteria of Doppler Microembolic Signals. Basic identification criteria of Doppler microembolic signals. Stroke. 1995;26:1123.
Arrowsmith JE, Harrison MJG, Newman SP, Stygall J, Timberlake N, Pugsley WB. Neuroprotection of the brain during cardiopulmonary bypass: a randomized trial of remacemide during coronary artery bypass in 171 patients. Stroke. 1997;29:2357–2362.
Newman SP, Stygall J. Neuropsychological outcome following cardiac surgery. In: Newman SP, Harrison MJG, eds. The Brain and Cardiac Surgery: Causes of Neurological Complications and Their Prevention. London, England: Harwood Academic; 2000:22–49.
Sivaguru A, Gaines PA, Beard J, Venables GS. Neuropsychological outcome after carotid angioplasty: a randomised controlled trial. J Neurol Neurosurg Psychiatry. 1999;66:262. Abstract.
Newman SP. Anxiety, surgery, and hospitalisation. In: Fitzpatrick R, Hinton J, Newman S, Scambler G, Thompson J, eds. The Experience of Illness. London, England: Tavistock; 1984:132–153.