Eur J Vasc Endovasc Surg 20, 153 157 (2000) doi:10.1053/ejvs.1999.1072, available online at http://www.idealibrary.com on Eversion Carotid Endarterectomy Generates Fewer Microemboli than Standard Carotid Endarterectomy M.-Y. Gao 1, H. H. Sillesen 2, J. E. Lorentzen 1 and T. V. Schroeder 1 1 Department of Vascular Surgery, Rigshospitalet, The National University Hospital, and 2 Department of Vascular Surgery, Gentofte University Hospital, Copenhagen, Denmark Objectives: to test whether the occurrence of microembolism differed between eversion and standard carotid endarterectomy (CEA). Design: prospective, non-randomised transcranial Doppler (TCD) monitoring study of 61 patients. Materials and methods: eversion CEA was performed in 27 and standard CEA in 34 patients. Surgery was performed under general anaesthesia. Three (5%) patients had a shunt inserted based on continuous EEG monitoring. Continuous middle cerebral artery TCD monitoring (EME, TC-4040) was performed intraoperatively and for 45 min postoperatively on day 1, day 2 3, day 4 5 and after 3 months. Unidirectional signals lasting >25 ms, having intensities of >9 db were considered to represent embolic events. Results: intraoperative embolic events were detected in 50 (93%) of 54 patients in whom successful intraoperative TCD monitoring was achieved. Events occurred most frequently immediately following clamp release (85%), without difference between the two techniques. Embolic events were encountered postoperatively in four (15%) and 16 (48%) patients having eversion and standard CEA, respectively (p < 0.02). Four patients developed new neurological symptoms equally distributed between eversion and standard CEA. Two (3%) deficits were permanent and two transient. The patency of the carotid bifurcation was confirmed in all instances with duplex scanning. Conclusion: we identified a surprisingly high number of postoperative embolic events as detected with transcranial Doppler in patients who had undergone carotid surgery using standard endarterectomy, as compared to patients who underwent eversion endarterectomy. Whether this difference has any clinical implication has yet to be proved. Key Words: Carotid endarterectomy; Eversion; Doppler; Embolism. Introduction a number of recent clinical, non-randomised studies 4 6 as well as in a randomised Italian multicentre study. 7 Carotid endarterectomy (CEA) has become one of All studies, including the latter the Everest Trial the most commonly performed operations in vascular suggested that EE is a safe and rapid procedure with surgery. In the International Trials, the risk of having low major complication rates. an ipsilateral stroke following CEA in symptomatic The principal cause of intraoperative neurological patients was 5% 7%, 1 and improvement of surgical morbidity during CEA is considered to be particulate technique may possibly reduce their number. A number embolisation. In order to investigate whether EE dif- of technical modifications and monitoring pro- fered from TEA, in terms of microembolism, individual cedures have been suggested in order to reduce this risk. These include routine use of shunts or patch closure as well as TCD monitoring, etc. 1,2 A modification of the surgical procedure was originally suggested by Etheredge 3 and eversion thromboendarterectomy (EE) has been developed on the basis of the conventional thromboendarterectomy (TEA). The benefit of EE has lately been compared to TEA in microembolic rates were monitored using transcranial Doppler (TCD) ultrasound during and after carotid surgery. Materials and Methods From February 1997 to January 1998, 81 consecutive patents underwent CEA, of whom 61 were included Please address all correspondence to: M.-Y. Gao, Department of Vascular Surgery RK, Rigshospitalet 3112, Blegdamsvej 9, 2100 in the present study. Two patients refused, in five Copenhagen Ø, Denmark. patients a bone window was not obtained and in 13 1078 5884/00/080153+05 $35.00/0 2000 Harcourt Publishers Ltd.
154 M.-Y. Gao et al. monitoring was not possible due to lack of staff. The Table 1. Number of patients with embolic events pre-, intra- and postoperatively. median age of the remaining 41 men and 20 women was 65 years (range 47 89 years). The study was No. pre- intra- post- approved by the medical ethics committee for Copen- TEA 34 7/34 (21%) 28/30 (93%) 16/33 (48%) hagen and Frederiksberg counties (KF 01-062/94). All EE 27 3/27 (11%) 22/24 (92%) 4/26 (15%) patients of the study gave informed consent. All Total 61 10/61 (16%) 50/54 (93%) 20/59 (34%) patients had experienced ipsilateral focal neurological p Value NS NS p < 0.02 symptoms: 13 (21%) patients had experienced amaur- TEA: conventional thromboendarterectomy. EE: eversion thromboosis fugax, 32 (52%) patients had experienced transendarterectomy. pre-: preoperative, intra-: intraoperative and post-: itory ischaemic attacks (TIA) and 16 (26%) had suffered postoperatively. stroke. All patients had an ipsilateral internal carotid artery stenosis of 76% assessed by duplex, plus arteriography in five cases. Operations were per- signals; HITS). 10,11 HITS were only accepted as embolic formed at a median of one month (range 0.5 3 months) events, as opposed to artefacts, if two experienced after the most recent neurological symptom. examiners agreed. Carotid surgery was performed under general anwas In most patients, the preoperative TCD monitoring aesthesia. Endarterectomy was performed either as a performed on the day before surgery. During TEA (n=34), through an arteriotomy in the common surgery the following events were marked: clamping carotid artery and extending through the plaque into as well as declamping of the common carotid artery, the internal carotid artery, or as an EE (n=27), beand and a shunt was inserted the point of introduction ginning with dividing the carotid bulb and performing removal was noted. Postoperative monitoring was an eversion endarterectomy of the internal carotid done on the first day, second or third day and fourth artery (ICA). 4 The ICA blood pressure was measured or fifth day, as well as after 3 months. All patients directly before and after clamping of the common and were treated by low molecular weight heparin 3500 external carotid arteries. Continuous EEG monitoring IU once every evening from the day before the oper- was used to determine whether a shunt should be ation until the day of discharge. Thereafter all patients inserted. The arteriotomy was closed with a running were recommended to take ASA 100 150 mg daily. suture. The median duration of the operations was 68 There was no difference in the treatment between two (range 48 95) min and of clamping 25 (15 41) min. groups. Duplex beyond 30 days as a surveillance Three (5%) patients had a shunt inserted and four modality was not a part of our routine protocol in (7%) patients had the TEA closed by a vein patch, Copenhagen and the 3-month follow-up was only which displays the information according to the op- TCD-based. erative approach. Non-parametric rank tests were used for statistical Continuous TCD monitoring of both middle cerebral comparison of differences between samples. Proarteries (MCA) was performed pre- and postwere performed two-tailed, and p<0.05 was considered portions were compared by Fisher s exact test. All tests operatively, as well as the ipsilateral MCA during surgery. TCD monitoring was performed for 45 minutes significant. pre- and postoperatively, with a pulsed Doppler TCD machine (TC-4040 EME, Überlingen, Germany), using two 2-MHz probes fixed in the patient s temporal regions bilaterally. Each of these probes, consisting of Results one transmitting and two receiving channels, were During preoperative monitoring 7 of 34 (21%) patients capable of simultaneous insonation at two different who had a TEA and three of 27 (11%) who had ES depths of the same vessel. 8 Emboli detection was per- had HITS corresponding to the symptomatic side (NS, formed with specially designed software (Pioneer Ver- Table 1). There was no correlation between presion 2.10 EME) that counts events defined as a operative HITS and type of preoperative neurological sudden increase in the power of the ultrasound signal symptoms, e.g. three of 13 (23%) patients with AF, five expressed in decibels, rejecting artefacts. 9 Uni- of 32 (16%) with TIA, and two of 16 (13%) with stroke. directional from baseline, lasting 25 ms or more, having Successful intraoperative TCD monitoring was intensities of 9 db or higher than that of the sur- achieved in 54 operations (30 in TEA, 24 in EE). rounding blood, and associated with a chirping Intraoperative embolic events were detected in 50 sound on the audio output, signals were considered (93%) patients without difference between the two to represent embolic events (high intensity transient endarterectomy techniques. HITS were detected dur-
Eversion Carotid Endarterectomy and Cerebral Microemboli 155 Table 2. Number of patients with embolic events detected intraoperatively. Pressure Dissection measurement Clamp on Clamp off Table 3. Symptoms and embolic events postoperatively (1 5 days postoperatively). Post-op. embolic events TEA 12/30 (40%) 10/25 (40%) 8/30 (27%) 25/30 (83%) + Total EE 9/24 (38%) 14/23 (61%) 1/23 (4%) 20/23 (87%) Total 21/54 (39%) 24/48 (50%) 9/53 (17%) 45/53 (85%) Symptoms Headache 11 14 25 p Value NS NS p < 0.05 NS New neurological symptoms 2 2 4 TEA: conventional thromboendarterectomy. EE: eversion thromboendarterectomy. No symptoms 7 25 32 Total 20 41 61 ing carotid dissection before clamping of the vessels in 21/54 (39%) patients, during pressure meas- urements in 24/48 (50%), and at reopening in 45/53 (85%) patients. In all instances, there were no differences between the two original techniques (Table 2). HITS were encountered postoperatively in 16 of 33 (48%) patients having TEA vs. four of 26 (15%) patients having EE (p < 0.02) (Table 1). Three of four (75%) patients who had their arteriotomy closed with a patch had embolic events lasting for at least 5 days postoperatively. Another patient showed more than five embolic events once and more than 10 twice for 5 days postoperatively. Following surgery, four (7%) patients developed new neurological symptoms equally distributed between EE and TEA (two patients on the day of operation, one patient on the day after operation and one patient 5 days after the operation, probably as part of a hyperperfusion syndrome). Two (3%) deficits were permanent and two transient. No patient died in the present series. A non-significant tendency was noted towards a relationship between postoperative headache and postoperative embolic events (p=0.10) (Table 3). At follow-up examination 3 months postoperatively no HITS were encountered. The patency of the carotid bifurcation was confirmed in all instances with carotid duplex scanning. Discussion Intraoperative embolic events occurred in 93% of patients, typically related to reopening of the endarterectomised artery (85%), during pressure measurement (50%) or during dissection (39%). Ackerstaff et al. 12 also found embolic transients after release of the cross-clamp at the end of the endarterectomy in 53% of procedures and in 25% during dissection. Though intraoperative microembolic events did not result in clinical sequelae, they noticed that the subgroup of patients who revealed more than 10 embolic events during surgery had a significantly higher number of lesions on magnetic resonance (MR) scan, as well as experiencing more clinical events. 12 A similar re- lationship between intraoperative microembolic signals and the occurrence of cerebrovascular complications has been reported in a number of series. 13 17 Our present experience, i.e. that all four patients who developed new neurological symptoms presented more than 10 embolic events intra- operatively, is in line with these reports, but our limited experience precludes any definite conclusions. Clearly, the high number of embolic events occurring during reopening, reported by several authors, could be gas- eous rather than solid. 8,18 20 However, events noticed in most series during dissection cannot arise from gaseous emboli. Though not employed in the present series, methods have been developed to separate par- ticulate from air emboli. 19 Several reports have indicated a relationship between the number of embolic signals detected during preoperative monitoring and whether patients had experienced neurological symp- toms or not, though the proportion of signal-positive patients have varied from 20 to 90%. 21 23 Other series have reported a reduction in frequency of embolic signals after carotid endarterectomy, corroborating the pathophysiological concept of most carotid procedures, i.e. to eliminate a source of embolism. 9,24 The main finding of the present series was that postoperative embolic events were noticed more frequently after TEA as compared to EE; 48% vs. 15% (p < 0.02). One reason could be that the eversion technique makes it easier to remove the stenotic plaque, leaving a smooth endarterectomised surface with a better-defined distal endpoint. In TEA, the longitudinal arteriotomy sometimes makes a patch closure necessary; though in the present series only in four (12%) cases, potentially increasing the thrombogenicity of the surface. In our study three (75%) patients in whom the arteriotomy was closed with a patch experienced embolic events
156 M.-Y. Gao et al. postoperatively as compared to 12 (41%) who had their References arteriotomy closed primarily (not significant (NS)). Previous reports on EE have only involved clinical comoperative thrombotic stroke after carotid endarterectomy: The 1Lennard N, Smith J, Dumville J et al. Prevention of post- parison, as was the case with the large Italian role of transcranial Doppler ultrasound. J Vasc Surg 1997; 26: randomised study on eversion versus standard carotid 579 584. endarterectomy. 7 No difference was found. The Leicesarterectomy by eversion technique its safety and durability. 2Shah DM, Darling RC III, Chang BB et al. Carotid end- ter group has a large volume of experience on intra- Annals of Surgery 1998; 228: 471 478. and postoperative TCD emboli monitoring. 1,19,25 Their 3Etheredge SN. A simple technique of carotid endarterectomy. studies indicate a high incidence of particulate em- Am J Surg 1970; 120: 275 278. 4Darling RC 3rd, Paty PS, Shah DM, Chang BB, Leather RP. bolisation during the first few postoperative hours that Eversion endarterectomy of the internal carotid artery: technique diminished with dextran therapy. 1,25 In the present and results in 449 procedures. Surgery 1996; 120: 635 639. study, monitoring was concentrated on the intraresults obtained with carotid eversion endarterectomy and with 5Entz L, Jaranyi Z, Nemes A. Comparison of perioperative operative period and on intermittent monitoring during conventional patch plasty. Cardiovasc Surgery 1997; 5: 16 20. the first postoperative days. 6Peiper C, Nowack J, Ktenidis K et al. Eversion endarterectomy Another question is whether the difference in emtreatment of internal carotid artery stenosis. Eur J Vasc Endovasc versus open thromboendarterectomy and patch plasty for the bolic events between the two surgical methods was a Surg 1999; 18: 339 343. reflection of patient or surgeon selection, rather than 7Cao P, Giordano G, Rango PD et al. Collaborators of the procedure-related. In the present series, three surgeons EVEREST Study Group. A randomized study on eversion versus standard carotid endarterectomy: study design and preliminary performed all procedures. Although they did not perresults: The Everest Trial. J Vasc Surg 1998; 27: 595 605. form an equal number of procedures and techniques, 8Nabavi DG, Georgiadis D, Mumme T, Zunker P, Ringelstein there was no indication that results were surgeoncerebral artery stenosis by means of a bigate probe. A pilot EBR. Detection of microembolic signals in patients with middle dependent; neither was there any indication that the study. Stroke 1996; 27: 1347 1349. operating surgeon selected TEA in case a particularly 9Van Zuilen EV, Moll FL, Vermeulen FEE et al. Detection of difficult procedure was anticipated. The vast majority cerebral microemboli by means of transcranial Doppler mon- of carotid endarterectomies in our clinic are performed itoring before and after carotid endarterectomy. Stroke 1995; 26: 210 213. on the basis of duplex imaging alone. Only in cases 10 Consensus Committee of the Ninth International where the distal part of the internal carotid artery was Cerebral Hemodynamic Symposium. Basic identification cri- difficult to visualise was arteriography performed. teria of Doppler microembolic signals. Stroke 1995; 26: 1123. 11 Forteza AM, Babikian VL, Hyde C, Winter M, Pochay V. This was the case in five patients, and even four of Effect of time and cerebrovascular symptoms on the prevalence them had an EE procedure. Thus we feel convinced of microembolic signals in patients with cervical carotid stenosis. that, although the study was not performed as a Stroke 1996; 27: 687 690. 12 Ackerstaff RGA, Jansen C, Moll FL et al. The significance randomised study, there was no bias in patient se- of microemboli detection by means of transcranial Doppler lection. ultrasonography monitoring in carotid endarterectomy. J Vasc Finally, the method used for emboli detection can Surg 1995; 21: 963 969. 13 Gaunt ME, Martin PJ, Smith JL et al. Clinical relevance of be debated. Traditionally, emboli have been detected intraoperative embolization detected by transcranial Doppler by more or less cumbersome offline analysis of re- ultrasonography during carotid endarterectomy: a prospective corded data, yielding more accurate and sensitive study of 100 patients. Br J Surg 1994; 81: 1435 1439. 14 Spencer MP. Transcranial Doppler monitoring and causes of methods. However, automated detection software has stroke from carotid endarterectomy. Stroke 1997; 28: 685 691. been developed, facilitating the procedure at the cost 15 Levi CR, Roberts AK, Fell G et al. Transcranial Doppler microembolus detection in the identification of patients at high risk of sensitivity. In a test of different equipment for emboli detection, van Zuilen 27 concluded that the equipment of perioperative stroke. Eur J Vasc Endovasc Surg 1997; 14: 170 176. 16 Muller M, Behnke S, Walter P, Omlor G, Schimrigk K. used for the present study was reasonably accurate Microembolic signals and intraoperative stroke in carotid endarterectomy. Acta Neurol Scand 1998; 97: 110 117. though less sensitive as compared to manual HITS detection. 17 Cantelmo NL, Bigian VL, Samarawerra RN et al. Cerebral microembolism and ischaemic changes associated with carotid In conclusion, we identified a surprisingly high endarterectomy. J Vasc Surg 1998; 27: 1024 1031. number of postoperative embolic events in patients 18 Georgiadis D, Goeke J, Hill M, König M, Nabavi DG. A novel who had undergone carotid surgery using standard technique for identification of Doppler microembolic signals based on the coincidence method. In vitro and in vivo evaluation. endarterectomy, as compared to patients who under- Stroke 1996; 27: 683 686. went eversion endarterectomy. Whether this difference 19 Smith JL, Evans DH, Fan L et al. Interpretation of embolic has any clinical implication remains to be proved. phenomena during carotid endarterectomy. Stroke 1995; 26: 2281 2284. 20 Gerraty RP, Bowser DN, Infeld B, Mitchell PJ, Davis SM. Acknowledgement Microemboli during carotid angiography. Association with stroke risk factors or subsequent magnetic resonance imaging The TCD recordings were kindly reviewed by, and discussed with, changes? Stroke 1996; 27: 1543 1547. Karsten Olesen, M.Sc.EE, Cephalon, Denmark. 21 Markus HS, Thompson ND, Brown MM. A symptomatic
Eversion Carotid Endarterectomy and Cerebral Microemboli 157 cerebral embolic signals in symptomatic and asymptomatic directed dextran therapy in the prevention of carotid thrombosis. carotid artery disease. Brain 1995; 118: 1005 1011. Three hour monitoring is as effective as 6 hours. Eur J Vasc 22 Markus HS, Harrison MJ. Microembolic signal detection using Endovasc Surg 1999; 17: 301 305. ultrasound. Stroke 1995; 27: 17 19. 26 Sitzer M, Müller W, Siebler M, Hort W, Kniemeyer HW. 23 Siebler M, Nachtmann A, Sitzer M, Rose G, Kleinschmidt Plaque ulceration and lumen thromboses are the main sources MEDTA et al. Cerebral microembolism and the risk of ischaemia of microemboli in high-grade internal carotid artery stenosis. and asymptomatic high-grade internal carotid artery stenosis. Stroke 1995; 26: 1231 1233. Stroke 1995; 26: 2184 2186. 27 Van Zuilen EV, Mess WH, Jansen C et al. Automatic embolus 24 Siebler M, Sitzer M, Rose G, Bendfeldt D, Steinmetz H. Silent detection compared with human experts. A Doppler ultrasound cerebral embolism caused by neurologically symptomatic highstudy. Stroke 1996; 27: 1840 1843. grade carotid stenosis: event rates before and after carotid endarterectomy. Brain 1993; 116: 1005 1015. 25 Lennard N, Smith JL, Hayes P et al. Transcranial Doppler Accepted 14 December 1999