The Role of Transcranial Doppler in Carotid Artery Surgery*
R o b e r t o C h i e s a I , F a b i o M i n i c u c c i 2, G e r m a n o M e l i s s a n o I , G iu l io T ruc i 2, G i a n c a r l o C o m i 2, G i a n m a r i a Pao l i l lo 1 and A d a l b e r t o G r o s s i 1
l lstituto Malattie Apparato Cardiovascolare e Respiratorio and 2Clinica Neurologica, UniversiM degli Studi di Milano, IRCCS Ospedale San Raffaele, Milano, Italy
In order to evaluate the usefulness of transcranial Doppler (TCD) in carotid artery surgery, 90 patients who underwent carotid endarterectomy were studied. Transcranial Doppler was performed preoperatively with digital compression of the carotid artery and intraoperatively at the time of cross-clamping, together with routine EEG monitoring. The percentage residual middle cerebral artery velocity ( %MCA V) during carotid artery compression or cross-clamping was considered. The intraoperative %MCA V at cross-clamping was compared with the values obtained in the preoperative examination during the compression test and with the EEG data; the EEG records were blindly reviewed by a neurologist and divided into three groups: group A, normal; group B, moderate anomalies and group C, major anomalies. There were no significant differences between preoperative %MCAV (compression test) and intraoperative %MCA V (cross-clamping) by separate analysis of the data related to systolic, diastolic and mean velocity. The systolic, diastolic, mean %MCA V and the Gosling index after cross-clamping were then compared with the EEG data: no significant differences were seen between groups A and B; on the other hand, in all patients with major EEG anomalies (group C) the %MCAV at TCD was reduced to O. In conclusion, preoperative TCD associated with carotid artery compression appears a safe and useful way of identifying patients at risk of cerebral ischaemia during carotid artery cross-clamping. Patients with no residual flow at intraoperative TCD need to be considered as presenting a very high risk of ischaemia and in this group the use of a shunt is mandatory.
Carotid endarterectomy is one of the most commonly performed operations in vascular surgery, however the indications, particularly in asymptomatic patients, and the efficacy of the procedure in the pre- vention of stroke are still controversial and are cur- rently being evaluated by means of prospective ran- domised trials. <2 Intraoperative cerebral perfusion monitoring and the prevention of ischaemic cerebral morbidity due to carotid artery cross-clamping are commonly debated issues. 3
It is well known that cerebral ischaemia during carotid cross-clamping occurs in only a small percent- age of cases, and in most surgical series where a shunt was not used, neurological complications were
* Presented at the 5th Annual Meeting of the European Society for Vascular Surgery, Warsaw, September 1991.
Please address all correspondence to: R. Chiesa, c/o Istituto Malat- tie Apparato Cardiovascolare e Respiratorio, Ospedale San Raf- faele, Via Olgettina, 60, 20132, Milano, Italy.
reported in less than 5% of cases. 4 Cerebral ischaemia can be prevented by using an intraluminal shunt, however its use carries several potential risks such as embolisation of air or of atheromatous fragments, damage to the vessel wall and distal intimal dissec- tion. The presence of the shunt increases the techni- cal difficulty of performing the endarterectomy and prolongs the operation. The morbidity associated with the use of the shunt is confirmed by the fact that the results of the surgical series in which a shunt was routinely used are not significantly different from the ones without shunting. 4 We agree with the general opinion 3-6 that the most appropriate strategy is the selective use of a shunt: it is therefore mandatory to correctly identify the patients at risk of cerebral ischaemia.
The most commonly employed criteria for selec- tive shunting are:
(1) clinical, such as evaluation of the history, the angiography and the brain computed tomography (CT) scan or observation of the awake patient oper- ated on under regional anaesthesia;
212 R. Chiesa et aL
(2) haemodynamic, such as evaluation of back pressure, cerebral blood flow (CBF) Xe133, conjuncti- val pO2 and transcranial Doppler (TCD);
(3) electrophysiological such as electro-encepha- lography (EEG), CT brain mapping (CTBM), somato- sensory evoked potentials (SEP).
None of these methods is 100% accurate and different techniques can be used synchronously in order to improve the reliability of monitoring.
The application of Doppler velocimetry to the study of the intracerebral arteries was introduced by Aaslid in 1982; 7 the TCD device that he developed employed a low frequency probe capable of insonat- ing through thin zones of the skull (i.e. transtemporal windows) to measure blood flow velocity in the arter- ies of the circle of Willis. This technique allows not only a record of interruptions of blood flow, but also of the changes following physiological or pathological stimuli (i.e. inhalation of CO2, breath holding, com- pression test etc.8).
The recent introduction of transcranial Doppler into clinical practice has opened up interesting pros- pects for its pre- and intraoperative use in carotid artery surgery. 9-12 The aim of this study is to evaluate the usefulness of TCD in this field and to establish the relationship between changes in EEG and TCD during carotid artery surgery.
Patients and Methods
Between January 1989 and December 1990, 90 con- secutive patients, 79 men and 11 women, underwent carotid endarterectomy (Table 1). All patients were operated on by the same surgical team and under- went preoperative duplex scanning of the extra- cranial cerebral arteries, complete cerebral angio- graphy of the extra- and intracranial vessels using the digital subtraction technique and brain CT scan. In addition preoperative TCD and intraoperative moni- toring with EEG and TCD were also performed in all patients.
All the surgical procedures were performed under general anaesthesia: after induction with 2~gkg-1 fentanyl, 5mgkg-1 thiopentone and 1.5 mgkg -1 succinylcoline, an endotracheal tube was positioned; anaesthesia was maintained with nitrous oxide and oxygen in combination with 1% isoflurane. Blood pressure was monitored by means of an intra- arterial line and was controlled with sublingual nife- dipine or i.v. nitrates. The end-tidal carbon dioxide (ETCO2) was continuously monitored by capnometry, serial arterial blood gas (ABG) analyses were carried
Table 1. Profile of patients
90 patients
79 males (86.1%),
11 females (13.9%)
Demographic data Mean age 63.0 years (range 51-74 years)
* Transient ischaemic attack. t Other symptoms include vertigo, drop attacks, visual blurring, diplopia, hoarseness, disphasia, confusion, numbness.
out and the patient's ventilation adjusted in order to maintain PaCO2 between 35 and 40mmHg. Prior to carotid artery cross-clamping, l m g k g -1 heparin i.v. was administered.
A Javid shunt was selectively used; the decision to shunt was made in cases where EEG changes 13 occurred at the time of carotid artery cross-clamping and it was not influenced by the TCD data during the course of this study. A completion angiogram was performed in all cases to ascertain the patency of the endarterectomised segment.
Electro-encephalographic monitoring was per- formed using a San-Ei 1A92 14 channels device; 14 scalp electrodes were applied with collodion accord- ing to the International Ten-Twenty system. At the end of the study the EEG records were reviewed blindly and interpreted by a single electro-encephalo- grapher, and divided into three groups according to Blume's classification: 13
group A: no EEG changes;
Eur J Vasc Surg Vol 6, March 1992
TCD in Carotid Artery Surgery 213
group B: moderate changes (clearly persisting 8- 15 Hz activity where the amplitude was attenuated by at least 50% and/or obvious and persistent increase of delta activity at greater than 1Hz ipsilateral or bilateral);
group C: major EEG changes (an attenuation of 8-15 Hz activity to minimal or nil and/or a two-fold or more increase of the delta activity at 1 Hz or less ipsi- lateral or bilateral).
Transcranial Doppler was performed using a 2MHz pulsed wave probe and an EME 3D (Eden Medical Electronics) ultrasound device. The probe was positioned in the temporal area, ipsilateral to the carotid artery to be operated upon, at a mean insona- tion depth of 51.1mm. A satisfactory velocimetric wave of the middle cerebral artery (MCA) was obtained in 82 patients (91%).
We considered the percentage residual MCA vel- ocity (%MCAV) calculated as the ratio of the MCAV after compression or cross-clamping of the carotid artery in relation to the basal MCAV. This ratio should overcome the problem of the insonation angle of the vessel and is a good index of the flow vari- ations, as reported by Aaslid.14
The pulsatility index (Gosling index) [(peak to peak height of sonogram waveform)/(mean height over one cardiac cycle)] was also calculated and is independent of probe to vessel angle, carrier fre- quency and velocity of sound in the tissues traversed. 15
The preoperative TCD examination was asso- ciated with a carotid artery compression test (Matas test): the common carotid artery was previously eva- luated by duplex scanning to ensure that it was free of atherosclerotic plaque; ECG monitoring was started a n d the common carotid artery gently pressed, to screen for carotid sinus hypersensitivity; the common carotid artery was then compressed and kept occluded for 10 s. Prior to and during the com- pression test the MCAVs were evaluated and the %MCAV was then calculated. During the surgical procedure the MCAV was monitored before and
during carotid artery cross-clamping, the %MCAV was calculated between the MCAV before and 10 s after cross-clamping. When an intraluminal shunt was used, the MCAV was monitored for the whole of the procedure to verify proper function of the shunt.
Intraoperative %MCAV at cross-clamping was compared with the preoperative % MCAV obtained at the compression test and with the results of the EEG monitoring.
Paired and unpaired Student's t-tests were used when appropriate. As a quantification of reproducibi- lity the kappa coefficient, appropriate for discrete var- iables, 16 was used. Statistical significance was accepted at p < 0.05.
Results
Eight of the 90 patients admitted to the study were excluded because a satisfactory MCAV wave could not be obtained. Thirteen patients were shunted because of abnormal electro-encephalograms. A Dac- ron velour patch was used in 29 cases and the carotid artery mean clamping time was 28min (range 10- 45 min). Completion angiography showed a patent internal carotid artery in all cases and no surgical mortality occurred in this series. Two patients suf- fered a postoperative neurological complication: one patient fully recovered from the motor deficit after 3 months, while the other has residual disability in his upper limb. No morbidity related to the compression test was recorded.
As far as the comparison between the pre- operative (compression test) and intraoperative (cross-clamping) %MCAV is concerned, the data relating to systolic, diastolic and mean velocity were separately analysed and no significant differences were found in any of the three groups (Table 2). The data relating to the preoperative and intraoperative diastolic %MCAV seems to be more homogeneous.
In four of the 82 patients the MCAV recorded
Table 2. Comparison between systolic, diastolic and mean %MCAV recorded preoperatively, at compression test and intraoperatively at carotid artery cross-clamping (before shunting)
Table 3. Intraoperative systol ic , diastol ic , mean + S.D. % M C A V and G o s l i n g index after carotid artery cross-c lamping in the three EEG groups
EEG Systolic Diastolic Mean Gosling group % MCAV % MCAV % MCAV index
Group A (n = 64) 52.6 +_ 20.5 68.1 _+ 25.5 60.8 +_ 20.2 0.57 +_ 0.17
Group B (n = 1.4) 51.6 + 23.0* 63.4 _+ 25.0* 59.2 +_ 20.7* 0.48 _+ 0.12"
Group C (n = 4) No residual flow No residual flow No residual flow - -
* N.S. v s . group A.
intraoperat ively at cross-clamping was 0 and in three cases the preoperat ive examinat ion correctly identi- fied this situation, while in one instance the pre- operat ive diastolic %MCAV was 27% which was thus a false negative. The calculated kappa coefficient was 0.8508 (p < 0.05).
According to the EEG data, 14 patients (17%) deve loped modera te changes and were allocated to group B, four patients (5%) deve loped major changes and were allocated in group C, while in the remain- ing 64 cases (78%) the EEG was unchanged (group A). The %MCAV (systolic, diastolic, mean, Gosling index) after cross-clamping was then compared with the EEG data and the results are shown in Table 3. No significant differences were seen be tween the %MCAV of group A and group B which were diffu- sely scattered in both groups; on the other hand the four patients with major EEG anomalies (group C) all had a reduct ion to 0 of the %MCAV at TCD (Table 3, Fig. 1).
Fig. 1. lntraoperative diastolic %MCAV after carotid artery cross- clamping plotted according to the EEG pattern.
Two neurological complications occurred in our series, both in patients belonging to group C. In one case the EEG disturbance was unde tec ted intraopera- tively, while in the other case a high carotid bifurca- tion and small internal carotid artery made it difficult to insert a shunt resulting in a 3 min delay. These technical complications could be avoided with increasing experience of the operators.
In the shun ted cases the TCD showed that an adequate blood flow was restored and maintained th rough the intraluminal shunt (Fig. 2).
Discussion
Preoperat ive TCD associated with a compress ion test is a reliable way of identifying those patients at risk of cerebral ischaemia at carotid artery cross-clamping and in the decision-making and preparat ion for the use of an intraoperative shunt. Howeve r it does not fully reproduce the intraoperat ive findings at cross- clamping: differences be tween the pre- and intra- operat ive findings in a single pat ient may be due to the effects of anaesthesia, to different haemodynamic situations and to different conditions of carotid artery blood flow interruption. The possibility of false nega- tives must be considered.
The occurrence of a false negative test (i.e. a pat ient with a good residual %MCAV during the compress ion test who deve loped EEG changes and no residual flow at TCD intraoperatively) could be due to perfusion from collaterals f rom the external carotid artery, unoccluded dur ing the compress ion test, bu t cross-clamped intraoperatively.
Transcranial Doppler provides a quantitat ive assessment of collateral cerebral circulation at the MCA and is a major advance in our ability to charac- terise non-invasively cerebrovascular disease. The usefulness of TCD monitor ing in the detection of intraluminal shunt malfunct ioning and of immediate restenosis due to intimal flap have been previously
Eur J Vasc Surg Vol 6, March 1992
TCD in Carotid Artery Surgery 215
Fig. 2. Intraoperative MCAV wave in a patient of group C: the MCAV drops to nil after carotid artery cross-clamping but an adequate flow is promptly resumed after insertion of an intraluminal shunt.
described. 17 Transcranial Doppler may also detect embolisation of intracranial vessels. Emboli were not detected in this series of patients but were noticed in the previous experience of the authors and by othersl~, 18, 19
From the comparison between TCD and the EEG data, a peculiar subset of patients emerged in whom the attenuation to minimal or nil in the amplitude of the 8-15Hz activity was associated with the disap- pearance of flow in the MCA. On the other hand uni- or bilateral increases in the delta activity were asso- ciated with different haemodynamic patterns. Our data shows that in patients with an increase of the delta activity but without attenuation of the 8-15 Hz activity, during carotid artery cross-clamping, the MCAV reduction was not significantly different from those patients with a normal EEG and shunting in this group of patients is likely to be unnecessary. In patients in whom an attentuation to minimal or nil of the 8-15 Hz activity occurs and the MCAV drops to nil, a shunt must be inserted without delay to prevent cerebral ischaemic damage.
In conclusion, TCD appears to be a reliable tool prior to and during carotid artery surgery. The use of TCD is limited by the significant number of patients in whom the thickness of the skull does not allow a satisfactory signal to be obtained. In addition, the quality of the examination is strictly operator-depen- dent. In spite of these disadvantages, the association of TCD with electrophysiological or clinical monitor-
ing techniques allows us to obtain more complete information on cerebral perfusion and function during carotid artery cross-clamping, and may help us in getting closer to the elimination of neurological morbidity on a haemodynamic base during carotid artery surgery.
Acknowledgements
We wish to thank Dr Laura Gatli, H. S. Raffaele--Milan--Depart- ment of Statistics and Biometry, for her statistical advice.
References
1 NORTH AMERICAN SYMPTOMATIC CAROTID ENDARTERECTOMY TRIAL (NASCET) STEERING COMMITTEE. North American Symp- tomatic carotid endarterectomy trial: methods, patients' charac- teristics and progress. Stroke 1991; 22: 711-720.
2 NORTH AMERICAN SYMPTOMATIC CAROTID ENDARTERECTOMY TRIAL (NASCET) INVESTIGATORS. Clinical alert: benefit of carotid endarterectomy for patients with high-grade stenosis of the internal carotid artery. Stroke 1991; 22: 816-817.
3 MOoREWS, QUINONES-BALDRICHWJ. Extracranialcerebrovascu- lar disease: the carotid artery. In: MOORE WS. Vascular Surgery. 3rd edn. Philadelphia: W. B. Saunders, 1991; 452.
4 KIEEEER E, CORMIER F, DAUNE B, FRANCO E. L'ufilisation d 'un shunt temporaire en chirurgie carotidienne. In: KIEFrER E, NATALI J. Aspects Techniques de la Chirurgie Carotidienne. Paris: AERCV, 1987; 103-109.
5 IMPARATO AM. Extracranial vascular disease: advances in oper- ative indications and technique. World J Surg 1988; 12: 756-762.
Eur J Vasc Surg Vol 6, March 1992
216 R. Chiesa etaL
6 GREENHALGH RIM~. Selective shunting during carotid endarterec- tomy. In: GREENHALGH RIVI. Vascular Surgical Techniques. 2nd edn. Philadelphia: W. B. Saunders, 1989; 89-93.
7 AASLID R, MARKWALDER TM, NORNES H. Noninvasive transcra- nial Doppler ultrasound recording of flow velocity in the basal cerebral arteries. J Neurosurg 1982; 57: 769-774.
8 RATNATUNGA C, ADISESHIAH M. Increase in middle cerebral artery velocity on breath holding: a simplified test of cerebral perfusion reserve. Eur ]Vasc Surg 1990; 4: 519-523.
10 SCHNEIDER PA, ROSSMAN ME, TOREM S, OTIS SM, DILLEY RB, BERNSTEIN EF. Transcranial Doppler in the management of extracranial cerebrovascular disease: implications in diagnosis and monitoring. J Vasc Surg 1988; 7: 223-231.
11 THIEL A, Russ W, ZEILLER D, DAPPER F, I-[EMPELMANN G. Trans- cranial Doppler sonography and somatosensory evoked poten- tial monitoring in carotid surgery. Eur J Vasc Surg 1990; 4: 597-602.
12 BERNSTEIN EF. Role of transcranial Doppler in carotid surgery. Surg Clin North Am 1990; 70: 225-234.
13 BLUME WT, FERGUSON GG, McNEIL DK. Significance of EEG changes at carotid endarterectomy. Stroke 1986; 17: 891-897.
14 AASLID R. Visually evoked dynamic blood flow response of the human cerebral circulation. Stroke 1987; 18: 771-775.
15 GOSLING RG, KING DH. Arterial assessment by Doppler shift ultrasound. Proc R Soc Med 1974; 67: 447-449.
16 KELSLEY JL, DOUGLAS THOMPSON W, EVANS AS. Methods in Obser- vational Epidemiology. New York: Oxford University Press, 1986; 290-291.
17 LAMAN DM, VOORWINDE A, DAVIES G, VAN DUIJN H. Intra- operative internal carotid artery restenosis detected by transcra- nial Doppler monitoring. Br J Surg 1989; 76: 1315-1316.
18 SPENCER MP, THOMAS GI, NICHOLLS SC, SAUVAGE LR. Detection of middle cerebral artery emboli during carotid artery endarter- ectomy, using transcranial Doppler ultrasonography. Stroke 1990; 21: 415-423.
19 STUMP DA, STEIN KS, TEGELER CH, et al. Validity and reliability of an ultrasound device for detecting carotid emboli. J Neuroi- maging 1991; 1: 18-22.
