Introduction
The first operation of carotid endarterectomy (CEA) was performed by M. DeBakey in 1953 [1]. Over a 70-year period, it has become one of the most commonly performed vascular procedures in the world. However, one of the most serious complications of CEA is perioperative ischemic stroke (IS).
According to the national clinical guidelines, the incidence rate of intraoperative stroke in departments performing CEA should not exceed 3% for patients with transient ischemic attack and 5% for patients with a previous history of acute cerebral ischemia (ACI) [2]. The probability of the development of stroke during surgery in a significant lesion of the contralateral internal carotid artery (ICA) considerably increases [3].
Intraoperative ischemic stroke develops through one of two main mechanisms: hemodynamic insufficiency and arterioarterial atheroembolism. Measures on cerebral protection from ischemia are aimed at preventing namely these mechanisms of stroke.
Over the decades-long history of performing CEA, several methods of preventing stroke have been suggested. They may be subdivided into “anesthesiological” and “surgical”. These methods are aimed at decreasing energy demands, maintaining adequate perfusion of cerebral tissue, and preventing atheroembolism. Of the general measures, to them belong craniohypothermia, narcosis with moderate and deep degree of narcotic sleep, artificially created controlled systemic arterial hypertension for the period of carotid artery clamping, the use of cerebroprotective and antioxidant agents [4-7].
Of “surgical” techniques, intraluminal shunt placement in the ICA is used, with the strictly specified sequence of both clamping of vessels and restoration of blood flow through them. The attitude towards placing intraluminal shunts in various surgical teams differs. Some surgeons during CEA use intraluminal shunt placing routinely in all patients [8-14]. Others decide upon shunt placement intraoperatively with due regard for changes of the parameters of intracranial hemodynamics and functional activity of the brain [15-17].
Negative aspects of shunt placing are also well known: the occurrence of air or material embolism with a fragment of carotid arteries atheroma, formation of parietal parafistulous thrombotic masses, additional difficulties during creation of an anastomosis, traumatization of the ICA wall.
It was reported that while using intraluminal shunts, the incidence rate of intraoperative strokes may reach 4.0-4.8% [10-12, 18-21], whereas in patients with multiple lesions of precerebral arteries and occlusion of the contralateral ICA, the incidence of stroke with the use of intraoperative shunting is also high, amounting to as much as 9.8% [22].
Objective. The study was aimed at determining efficacy and safety of the technology of perioperative management and the technique of carotid endarterectomy with no intraluminal arterial shunt.
Patients and methods
The study included a total of 3615 patients successively operated on at the same department. Of these, there were 2845 (78.7%) men and 770 (21.3%) women, with the mean age amounting to 65±7.2 years.
Our clinical investigation was a single-center cohort retrospective-and-prospective study.
Inclusion criteria:
1) elective nature of the operation;
2) presence of chronic cerebrovascular insufficiency caused by a significant (≥70%) atherosclerotic lesion of the carotid artery bifurcation;
3) presence of chronic cerebrovascular insufficiency caused by a ≥50% ICA stenosis with complicated atheroma.
Exclusion criterion: patients in acute period of ischemic stroke.
The endpoints of the study were intraoperative ischemic stroke and intraoperative IS + patient death.
During preparation of patients for surgery, an obligatory condition was examination of each patient by a neurologist, cardiologist, and cardiovascular surgeon.
All patients prior to operation underwent computed tomography (CT) of the brain. Ischemic stroke before hospitalization for surgery had been experienced by 1861 (51.5%) patients. 3113 (86.1%) patients had a pronounced clinical picture of encephalopathy (stage II and III). The structure of the forms of cerebrovascular insufficiency in the operated patients is presented in Table 1.
&hide_Cookie=yes)
According to the criteria of the Russian national guidelines on managing patients with brachiocephalic arteries diseases (2013), 1139 (31.5%) patients were symptomatic.
The structure of concomitant diseases is shown in Table 2.
&hide_Cookie=yes)
The leading method of diagnosis of the state of precerebral arteries was Doppler ultrasonography with color duplex scanning. If clinically indicated, the patients underwent CT angiography or X-ray contrast angiography of brachiocephalic arteries. A criterion for selection of patients for surgery was an instrumentally revealed finding of a ≥70% stenosis (according to the NASCET criteria) [3301 (91.3%) patients]. Besides, the operations were performed in symptomatic patients with a 50-69% stenosis in the presence of signs of instability of carotid artery atheroma in the basin of acute ischemic stroke [314 (8.7%) patients]. These data are shown in Table 3.
&hide_Cookie=yes)
889 (24.6%) patients were found to have a significant lesion (>70% or occlusion) of the contralateral ICA. A hemodynamically significant lesion (>70% stenosis, occlusion, steal syndrome) of the ipsilateral vertebral artery (VA) was revealed in 511 (14.1%) patients, that of the contralateral VA in 358 (9.9%) patients. Vertebrosubclavian steal syndrome was detected in 215 (5.9%) patients. The incidence and severity of lesions of the contralateral and ipsilateral ICA, VA and subclavian artery (SCA) are shown in Fig. 1 and 2, respectively.
&hide_Cookie=yes)
In patients with multiple lesions of precerebral arteries, the leading criterion for selection of the side of the first operation included a more pronounced degree of vessel stenosis, the presence of unstable atheroma, and clinical manifestations in the vascular basin involved.
All 3615 patients were operated on without the use of an intraluminal shunt.
The main stages of perioperative treatment were standardized and consisted of the following:
1. In the absence of contraindications, we performed preoperative preparation with obligatory administration of β-blockers. The patient was taken for the operation with the target heart rate of 50-60 bpm. To rule out hypovolemia, the day before surgery the patients were not restricted in drinking liquids.
2. We did not focus on the Matas test for determining the indications for intraluminal shunt placement.
3. All operations were carried out under general combined anesthesia with artificial lung ventilation. The anesthesiologist assessed the presence and degree of hypovolemia and corrected it at the very beginning of the operation [23, 24]. Infusion therapy was carried out with crystalloid and colloid solutions, taking into account the patient’s volemic status.
4. All operations were performed using operating enlarging optics with 2.5-4.5× magnification. The carotid arteries were accessed along the front edge of the nodding muscle. At the stage of artery exposure, the patient was given 5000 IU heparin intravenously. Holders on carotid arteries were not used.
5. Prior to carotid artery clamping, we increased the level of systolic AP by 20-30% above the level of the “working” level (individual for each patient) AP [23].
6. To prevent embolism we strictly observed the sequence of clamping of carotid arteries: first the external, then common and ICA.
7. To control sufficiency of cerebral blood supply at initial stages we measured the level of retrograde pressure. We then completely abandoned this method of control and began to perform only visual assessment of the pattern of retrograde blood flow from the cut-off ICA according to the principle “present” - “absent”, “pulsing” - “not pulsing”. But it was obligatory to use cerebral oximetry with the Somanetics INVOS 3100 unit. In case of a decrease of blood flow in the cerebral hemisphere by more than 20%, we additionally increased the level of AP by 10-15% to obtain the target parameters of cerebral oximetry (interhemispheric difference of not more than 20% of the baseline values).
8. 3145 (87%) patients underwent modified CEA with creation of high divarication: cutting off the ICA from the ostium was flowed by dissection of the internal and external carotid arteries along the inner edge to the atheroma-free lumen. We sought to attain so that from the distal line of removal of atheroma to the end of carotid arteries dissection there were not less than 1.5 diameters of the ICA. Complete removal of atheromas from the ICA and ECA was performed under visual control, and the plaque from the CCA was removed by means of eversion, to be followed by establishing an extended "end-to-side" anastomosis with creation of high divarication of carotid arteries by means of sewing the ICA area to the ECA-ICA area (Fig. 3). In all patients we used only 7/0 polypropylene suture [25].
&hide_Cookie=yes)
470 (13%) patients at the initial stage of the study due to length of atheroma in the ICA exceeding 4 cm underwent classical CEA with ICA and CCA plasty using an autovenous patch.
The structure of the interventions and mean time of carotid artery clamping are shown in Table 4.
&hide_Cookie=yes)
To prevent material embolism and carotid artery restenoses in the remote period, the main rule during CEA was meticulous removal of all fragments of atheroma in the zone of operation to the adventitia, paying particular attention thereto, although these manipulations increased the time of carotid artery clamping.
9. An obligatory stage of the plastic stage of the operation was prevention of air and material embolism by means of washing and filling the cavity of the reconstructed arteries with normal saline and heparin, as well as consecutive removal of clamps from the ICA then from the ECA. Before removal of clamps from the CCA, we again temporarily clamped the ICA at the ostium in order to prevent migration of possible material particles from the reconstruction zone. Prior to removal of clamps from carotid arteries, the anesthesiologist decreased the AP level to the baseline “working” values.
All patients were withdrawn from narcosis in the operating room, with the assessment of the neurological and somatic status, registering electrocardiograms (ECG). They were all immediately after the operation compulsorily examined by the neurologist and cardiologist.
10. During the first 24 hours after surgery, we monitored such parameters as AP, HR, SpO2, ECG, neurological status, controlling and correcting the AP level.
The drainage was removed on POD 2.
During the first 5 days, the patients were given 5000 IU heparin 3 times daily subcutaneously under control of the activated partial thromboplastin time. Besides, from POD 1, the patients received disaggregants, analgesics, by indications hypotensive and antiarrhythmic therapy. From POD 2, we performed extension of the motor activity and dietary regimen.
Revealing a clinical picture of intraoperatively occurring acute cerebral ischemia was followed by ultrasonographic assessment of carotid artery patency without removing the patient from the operating table, as well as by studying structural lesions of the brain by means of CT performed immediately after surgery and in dynamics.
The criteria for assessing the mechanism of neurological problems were as follows:
· atheroembolic mechanism of ischemic stroke was determined when the data of CT imaging demonstrated conformity of stroke with zones of vascularization of separate branches of the middle cerebral artery (MCA), as well as evidence of wedge-shaped cerebral infarction in embolism to large branches of the MCA or small subcortical foci (Fig. 4, A);
· hemodynamic mechanism of ischemic stroke was determined when CT visualization findings revealed extensive cerebral infarction involving the territory of the anterior cerebral artery and MCA on the side of artery clamping; subcortical infarctions localizing in the zones of adjacent circulation (Fig. 4, B);
· lacunar stroke was determined in a long history of hypertension and diabetes mellitus, as well as events of microangiopathy; susceptibility to arterial hypertension during the operation; the presence of focal deficit on the background of normal recovery, preserved speech and intellectual functions. According to the CT imaging data: ischemic foci of less than 30 mm with localization in the caudate nucleus, putamen or thalamus on the side of the performed operation (Fig. 4, C).
&hide_Cookie=yes)
All patients with ACI underwent treatment in conditions of resuscitation according to the standards of managing the pathology concerned.
Methods of statistical processing: the findings of the study were processed by means of the applied software package for statistical processing in the Microsoft Office Excel.
Results
After CEA, 3566 (98.64%) patients had no evidence of ACI.
Forty-nine (1.36%) of the 3615 patients developed intraoperative stroke confirmed by CT. Based on the findings of ultrasonographic examination of precerebral arteries, thrombosis of carotid arteries in these patients was excluded. In each of the 49 patients, ACI subtype was defined according to the presented criteria, with determining possible causes of its development.
The distribution of patients by the pathogenetic mechanism of the onset of intraoperative stroke is shown in Table 5.
&hide_Cookie=yes)
The incidence rate of atheroembolic stroke amounted to 1.13% (in 41 of the 3615 patients) and to 83.65% of all ischemic strokes. Hence, the leading mechanism of the development of intraoperative stroke turned out to be arterial atheroembolism.
Five (0.14%) of the 3615 operated patients developed hemodynamic stroke.
The incidence of ischemic stroke by the years of the study is shown in Figure 5.
&hide_Cookie=yes)
At the in-hospital stage, eleven (0.3%) patients with intraoperative ischemic stroke died. Of these, ten (91%) had atheroembolic stroke.
Discussion
While performing CEA, one of the main tasks is prevention of ischemic stroke, especially in patients with multiple lesions of precerebral arteries.
Using an intraluminal shunt during CEA is aimed at preventing cerebral hypoperfusion during coronary artery clamping and at avoiding the development of hemodynamic-type ischemic stroke. The use of this method of cerebral protection is primarily considered in patients with severe lesions of the contralateral carotid artery. Mention should be made that the presence of a hemodynamically significant ICA stenosis suggests a decrease in the vessel’s lumen by more than 2/3. A stable condition of the patient admitted to undergo an elective operation is suggestive of the formed and functioning ways of collateral compensation of cerebral circulation. Another matter is whether this collateral compensation would be enough for viability of brain tissue during surgery with complete clamping of the ICA.
The obtained findings showed that in 41 of the 49 patients who developed stroke it was atheroembolic. Obviously, the use of an intraluminal shunt not only would have failed to contribute to preventing stroke in these patients, but, probably, would have posed additional risks for stroke development. That is why more important for preventing atheroembolic stroke in these patients is the necessity of a meticulous preoperative evaluation of the morphological condition of atheroma (in the assessment of its embologenic potential) and an extremely sparing attitude towards carotid arteries while exposing them during surgery.
Hemodynamic stroke turned out to be a rarer event encountered in five of the 3615 patients, thus accounting for 0.14%. All these patients had been operated on more than 10 years before (the initial stage of the study). Analyzing their anesthesiological records showed that they had had inadequate levels of artificial controlled arterial hypertension during carotid artery clamping. We suppose that a way out of this complicated situation may be more adequate control of the AP level for an individual patient during the operation. An especially important factor is considered to be the fact that the patient during surgery should have no hypovolemia.
Based on the obtained findings, we believe that correction of the factors influencing the hemodynamic mechanism of the development of ACI during carotid artery clamping is the task of an anesthesiologist, whereas the task of an angiosurgeon has been and remains that of preventing arterial embolism at various stages of the operation.
Comparing the obtained results with previously published data concerning operations with or without intraluminal shunts is suggestive of appropriateness of the technology we used (Table 6).
&hide_Cookie=yes)
Conclusion
While performing CEA, in the overwhelming majority of patients it is possible and necessary to refuse from using an intraluminal shunt as a factor increasing the risk for the development of intracranial artery embolism. The task of protection of the brain from hypoperfusion during CEA should be recognized as anesthesiological.