The impact of neuroprotection on brain metabolism during carotid endarterectomy

Overview

Objectives:
The primary objective of this study was to evaluate the impact of neuroprotection, administered during carotid endarterectomy, on brain metabolism. The secondary objective was to assess the impact of resulting changes in brain metabolism on the clinical outcome.

Methods:
A total of 35 patients underwent carotid endarterectomy with prophylactic combined neuroprotection (Sendai cocktail: Manitol, Phenhydan, Solumedrol, Tokoferol; Cerebrolysin; fraction of inspired oxygen (FiO₂) =1, middle arterial pressure (MAP) = 100 mmHg, total intravenous anesthesia - TIVA). The influence of neuroprotection on brain metabolism (S100B, glycaemia, lactate, pH, jugular vein bulb oxygen saturation - SvjO₂) was evaluated. Metabolic parameters were acquired from the jugular bulb during surgery, just before unclamping of the vessel. The clinical outcome was evaluated by NIHSS (National Institutes of Health Stroke Scale). There were 35 patients in the control group who where operated on without any neuroprotection. The results from both groups of patients were compared and statistically analyzed.

Results:
Postoperative NIHSS did not change in any patient in either group. An intraoperative shunt was not inserted in any patient in either group. In the group with neuroprotection there were significantly higher levels of S100B (median 0.117 vs. 0.088; p<0.0182), lactate (median 1.92 vs. 1.020; p<0.0006), glycaemia (median 9.5 vs. 8.2; p<0.0243), and SvjO2 (median 0.79 vs. 0.65; p<0.0001). There were no postoperative changes to NIHSS in either group.

Conclusions:
Neuroprotection administered before carotid endarterectomy influences some parameters of brain metabolism both positively and negatively, but with no impact on clinical outcome.

Key words:
neuroprotection, carotid endarterectomy, brain metabolism, cognitive functioning

Introduction

Carotid endarterectomy (CEA) is an effective operation in the primary and secondary prevention of stroke on the condition that morbidity and mortality (MM) do not exceed the recommended grade (1). The complications of CEA can be divided into 3 groups: neurological, internal and surgical. Intraoperative strokes, caused by embolisation or hypoperfusion, are largely accountable for neurological MM. Neurosurgeons try to minimize the risk of intraoperative stroke or neurological deficit in the event of a stroke. The occurrence of a stroke during an operation designed to prevent strokes is always a very frustrating complication.

One possibility of reducing the consequences arising from intraoperative stroke is to use neuroprotection. Neuroprotection is a strategy that antagonizes the sequence of injurious biochemical and molecular events that, if left unchecked, would result in ischemic injury (2). In spite of the demonstrable effects of many agents in animal models, until now, none of the tested neuroprotective agents have been shown to improve the outcome in a phase III clinical trial (3). The main cause of the failure of neuroprotection in stroke patients is an over-extended therapeutic window (3, 4). In preclinical research, neuroprotection becomes effective within a six hour therapeutic window. The most significant effect was achieved when administration was performed before the occurence of a stroke (5). This finding has led to a new concept of prophylactic neuroprotection (6). Prophylactic administration of neuroprotection prior to risk procedures (CEA is the typical example), could be beneficial for patients.

However, the incidence of intraoperative ischemic injury during CEA is very low. In big studies its frequency runs to anything from less than one to a few percent. It would be necessary to evaluate groups of thousands of patients if we wish to assess the impact of neuroprotection on the incidence of strokes. Such a high number is achievable only in prospective, international, multicentric studies. No such study has yet been published. From the definition of neuroprotection it is evident that biochemical changes precede a stroke. Therefore we evaluated the impact of neuroprotection on brain metabolism and its association with resulting clinical consequences.

Methods

Patients who underwent elective CEA due to symptomatic carotid stenosis were submitted to this prospective study. The criteria for acceptance to the study were a post TIA condition or minor ischemic stroke with a minimal residual neurological deficit (Institutes of Health Stroke Scale - NIHSS < 3), CT without new ischemia and symptomatic carotid stenosis of more than 50%.

Patients, whose actual clinical status, insufficient collateral supply or condition of cerebral tissue could influence the evaluated parameters (operation within 14 days of the origin of the stroke, neurological instability, significant residual neurological deficit, contralateral carotid obliteration, new hypodensity on CT), were not submitted to the study.

A total of 35 patients underwent carotid endarterectomy with combined prophylactic neuroprotection (Table 1). The neuroprotective strategy involved a Sendai cocktail (20% Manitol 150 ml, Phenhydan 500mg, Solumedrol 1g, Tokoferol 300 mg) and 50 ml Cerebrolysin administered at the beginning of anesthesia, fraction of inspired oxygen (FiO₂) =1 and middle arterial pressure (MAP) = 100 mmHg at the time of vessel clamping. Total intravenous anesthesia (Propofol) is a standard part of neuroprotection.

Microsurgical technique and intraoperative electrophysiological monitoring (EEG, SEP n. medianus) were used during operations.

There were 35 patients in the control group who where operated on without any neuroprotection under local anesthesia.

The primary objective of this study was to evaluate the impact of combined neuroprotection, administered during carotid endarterectomy, on brain metabolism (S100B, glycaemia, lactate, pH, jugular vein bulb oxygen saturation - SvjO₂₎. Metabolic parameters of the brain were ascertained during surgery from the homolateral internal jugular vein just before unclamping of the vessels. The secondary objective was to assess the impact of resulting changes in brain metabolism on clinical outcome, which was scored using NIHSS.

The results from both groups of patients were compared and statistically analyzed. Statistica 9.0 Software was used.

Results

CEA with neuroprotection was performed over a two year period on 35 patients (25 men, 10 women) aged 44 -77 (median 65). Nine patients had a mild neurological deficit (NIHSS 1-3) before surgery while the others had normal neurological status. CEA without neuroprotection was performed on a control group during the same two year period. Ten patients had a mild neurological deficit (NIHSS 1-3) before surgery, the others had a normal neurological status.

Statistically significant differences in metabolic parameters were demonstrated in S100B protein, lactate, glycaemia and SvjO₂.

In the neuroprotection group, compared with the control group, there were significantly higher levels of S100B (min 0,018, median 0,117, max 0,335 vs. min 0,022, median 0,088, max 0,262; p<0,0182), lactate (min 0,050, median 1,92, max 3,540 vs. min 0,670, median 1,020, max 3,580; p<0,0006), glycaemia (min 5,1, median 9,5, max 17,6 vs. min 4,8, median 8,2, max 16,0; p<0,0243), and SvjO₂ (min 0,56, median 0,79, max 0,97 vs. min 0,51, median 0,65, max 0,83; p<0,0001) (Graph 1 – 4).

No intraoperative stroke was detected in either group, preoperative and postoperative NIHSS did not change in any patients. We did not detect any statistically significant difference between the follow-up and control groups in either preoperative or postoperative neurological status (NIHSS 0 – 74% vs. 71%, NIHSS 1 – 17% vs. 17% - , NIHSS 2 – 3% vs. 9%, NIHSS 3 – 6% vs. 3%). Recognized differences in brain metabolic parameters did not influence clinical outcome.

Discussion

We had two reasons to do the presented research. The first was a very low incidence of symptomatic intraoperative ischemic stroke (0,5 %) and the low frequency of shunt insertion (2,75 %) in our cohort of CEA. We used shunt significantly less than in published literature (6-16%) (7-9). We interpret our positive results as a positive influence of the mentioned combination of neuroprotective strategies and general anesthesia that we have used over a long period in standard practice (10).

The second motivation to research was the idea of new concept of prophylactic neuroprotection, which is more beneficial, if it is administered before stroke (6). The combination of TIVA and prophylactic neuroprotective strategies is a typical example of mentioned concept.

The effect of our neuroprotective strategies was only recorded, as in all neuroprotective agents used until now, in experimental trials and in clinical trials of phase I and II (11-15). The combination of neuroprotective agents (cocktail), which work on more levels of ischemic cascade, is more effective (4). The most significant neuroprotective effect was demonstrated when administration was performed before stroke (5). Our method, which combines a number of neuroprotective strategies (pharmacological, physical) administered preventively before the occurrence of stroke, fulfills both requirements. General anesthesia is an irreplaceable part of the neuroprotective agents used and, moreover, general anesthesia is the condition for administration of a whole group of neuroprotective agents (FiO₂ = 1). There is much reliable evidence of the prophylactic effect of neuroanesthesia (16).

Patients of the neuroprotection and the control group did not significantly differ in regard to demographic characteristic, initila clinical status or timing of operation. All patients were monitored continuously for 24 hours in a neurosurgical intensive care unit, where NIHSS was also assessed first. We were able to detect reliably stroke as well as minor TIA and other manifestations related to intraoperative hypoperfusion or microembolisation (confusion, delirium, cognitive deterioration). In spite of the accuracy of clinical observation none of the aforementioned incidents were noted. NIHSS neither changed nor differed in either group.

With the exception of patients after a major stroke, we operated on all patients with symptomatic carotid stenosis as soon as possible according to recent guidelines (1). In no case was the timing of operations delayed to allow for the acceptance criteria of the study. The long time period from the stroke to surgery is due to the late transmission of patients from neurological departments and, paradoxically, due to no submission of patients operated on within the required 14 days of the stroke.

The parameters of brain metabolism were determined from jugular bulb blood acquired by puncture of the internal jugular vein during the operation and just before unclamping of the vessel. During this period the homolateral brain hemisphere is supplied by a more or less sufficient collateral blood supply and hence there is the most probable potential for hypoperfusion with a negative impact to brain metabolism.

Higher levels of 100B protein were surprisingly found in the neuroprotection group (Graph 1). S100B protein is a calcium linking protein which is used as a prognostic marker of brain injury. Its biological half-life is unknown (17). Lavicka et al mentioned that, in mild traumatic brain injury, 100B protein returns to normal levels within three days (18). Thus, to avoid having the 100B protein level influenced by their primary stroke, our patients were operated on within 14 days and the patients with major stroke were not submitted to the study. We did not record any correlation between S100B protein and timing of the surgery. The levels of S100B protein were significantly higher (p<0,0182), but in most cases (83%) still within standard range (0,001-0,14 microg/l). Extracerebral resources of S100B protein could not be of use in our study.

In the neuroprotection group there were also surprisingly significantly higher levels of lactate (p<0,0006) in spite of the time period of FiO₂ = 1 during surgery (Graph 2). The oxygen supply in arterial blood exceeded its consumption by more than under physiological conditions (19). Lactate originates from anaerobic glycolysis and its increase corresponds to the restriction of oxidative phosphorylation. The rise in lactate could be caused by hypoxia, arterial hypotension or significant intraoperative blood loss (20). Nevertheless, at the beginning of anesthesia, during anesthesia, and during surgery, none of the mentioned incidents were registered. Lactate, along with the rest of the parameters, was moreover tested at the end of the time period when the patient was ventilated with 100% O_2. The elevation of lactate was not accompanied by lactate acidosis.

A significantly higher level of glucose was also found in the neuroprotection group (p<0,0243) (Graph 3). It is important that there was no difference between the two groups in either the incidence of diabetes mellitus or in the correlation of higher glycaemia and the incidence of DM. The metabolism of glucose, lactate and oxygen are closely associated with each other. Hyperglycaemia exceeds the size of ischemia and induces anaerobic glycolysis which causes an elevation of lactate (12). Hyperglycaemia is an unambiguously harmful factor during brain ischemia (21). Elevated glucose accompanies stroke in 50% (stress hyperglycaemia) (13). This explanation can not be used in our study because none of the patients suffered intraoperative stroke. On the other hand, a higher level of glycaemia in blood from jugular vein can be connected with the lower rate of glucose (CMRGl) during lower energy consumption of the brain, which is influenced by neuroprotection. This interpretation is supported by significantly higher SvjO₂ in the neuroprotection group, which can be explained by lower oxygen consumption (reduced cerebral metabolic rate of oxygen - CMRO₂) or its elevated supply (Graph 4). Administered neuroprotection is probably involved in both mentioned mechanisms. The average oxygen consumption (O₂ER - oxygen extraction ratio) was lower in the neuroprotection group in comparison with the control group (21% vs. 33%).

The observed levels of metabolic markers can be interpreted as controversial (Table 2). Some markers were influenced positively by neuroprotection, others negatively. Surprisingly, higher S100B protein and lactate in the neuroprotection group indicate a negative impact of neuroprotection. There was also an elevated level of glucose in the neuroprotection group. In spite of the acceleration of pathophysiological changes and the enlargement of stroke volume by hyperglycaemia in the experiment, it can be considered as a manifestation of lower metabolic consumption of the brain. From this point of view hyperglycaemia can be interpreted as a positive consequence of neuroprotection. Higher SvjO₂ in the investigated group is an indisputably positive effect of neuroprotection.

We tried to detect a correlation between the metabolic markers in our study. However, a statistically significant correlation between the elevated levels of S100B protein, glucose and lactate was not proved.

Conclusion

The effect of combined neuroprotective strategies administered before CEA is controversial. Neuroprotection used in our study influenced some brain metabolic parameters, both positively and negatively, however without impact on the clinical outcome. We did not prove a statistically significant correlation between monitored metabolic parameters.

Our unexpected disappointment with the inconclusive results corresponds with the general disillusion from the failure of neuroprotection in clinical trials. According to the latest guidelines of the European Stroke Organisation (ESO) there is no recommendation for the administration of neuroprotection in patients with acute stroke (class I, level A). It is not expected that the administration of neuroprotection will become part of “evidence based medicine” in the near future. The use of neuroprotection in patients with acute stroke is based on experimental trials. It is necessary to take into account the cost benefit ratio when neuroprotection is used.

In spite of the failure of neuroprotection in the treatment of acute stroke, it is necessary to continue doing clinical research and importantly not to avoid publishing negative results.

Abbreviations

• FiO₂ - fraction of inspired oxygen
• MAP – middle arterial pressure
• TIVA – total intravenous anesthesia
• SvjO₂ – jugular vein bulb oxygen saturation
• NIHSS - National Institutes of Health Stroke Scale
• CEA – carotid endarterectomy
• MM - morbidity and mortality
• TIA – transient ischemic attack
• CT – computed tomography
• EEG - electroencephalogram
• SEP – somatosensory evoked potentials
• DM – diabetes mellitus
• CMRGl – cerebral metabolic rate of glucose
• CMRO₂ – cerebral metabolic rate of O₂
• O₂ER - oxygen extraction ratio
• ESO - European Stroke Organisation

Corresponding author contact:
Jan Mracek, MD
Department of Neurosurgery, Charles University Hospital Pilsen
Alej Svobody 80, Pilsen, 304 60, Czech Republic
Fax: +420377103963
mracek@fnplzen.cz