Piepgras DG, Morgan MK, Sundt TM et al. Intracerebral haemorrhage after carotid endarterectomy .... Blake FG, Cecil RL. Studies on experimental pneumonia. I.
Br. J. Surg. 1993, Vol. 80, December, 1523- 1527
A. R . Naylor, M. R . Whyman, J. A. W. Wildsmith*, J. H. McClure*, A. McL. Jenkins, M. V. Merrickt and C. V. Ruckley Departments of Vascular Surgery, *Anaesthetics and ?Nuclear Medicine, The Royal Infirmary and Western General Hospital, Edinburgh, UK Correspondence to: Mr A. R. Naylor, Aberdeen Royal Infirmary, Aberdeen AB9 228, UK
Factors influencing the hyperaemic response after carotid endarterectomy Transcranial Doppler ultrasonography was used to evaluate serially the changes in middle cerebral artery bloodflow velocity ( M C A V ) in 37 consecutive patients during the first 72 h after carotid endarterectomy to identfy factors that may predispose towards postoperative hyperaemia. Within 6 h of endarterectomy, median M C A V in the operated hemisphere was 48 per cent (95 per cent confidence interval 37-60 per cent) above that on admission and remained 27 per cent (95 per cent confidence interval 19-37per cent) higher at 72h. There was a similar, but less marked, increase in M C A V in the contralateral middle cerebral artery during the same time period. There was no association between the postoperative increase in M C A V and clinical presentation, admission M C A V , the presence or absence of a residual neurological deficit or infarction on computed tomography before operation, carotid clamp time, shunt usage, internal carotid artery stump pressure or M C A V during clamping. The greatest increase in M C A V was observed in patients with internal carotid artery stenosis 2 50 per cent and, more particularly, in those with preoperative evidence of impaired cerebrovascular reserve. In the latter patients, M C A V was 100 per cent above the admission level within 12h of operation and was still 50 per cent raised at 72 h.
Immediately after carotid clamp release and restoration of blood flow, there is usually a surge in middle cerebral artery blood flow velocity (MCAV) that then decreases within 10 min of clamp release, although not back to the preoperative level'. Evidence suggests that this initial phenomenon represents a short-lived hyperaemic response to the effects of carotid clamping and may reflect the ability of the collateral circulation to maintain cerebral blood flow (CBF) during carotid occlusion'. After this initial surge, some patients progress to a secondary hyperaemic phase that may persist for up to 5 days after surgery'. The true incidence of this secondary hyperaemic phase is unknown, but it is much commoner than the associated clinical symptoms. Fewer than 1 per cent of patients will suffer a stroke, usually due to intracerebral haemorrhage3, but a greater proportion will have headache or di~orientation'.~.It is generally thought that this phenomenon results from transient impairment or failure of cerebral a u t ~ r e g u l a t i o n ' ~secondary ~,~, to a chronic state of cerebral vasodilatation in an attempt to maintain blood flow. The clinical manifestations tend to become apparent 24-72 h after operation, although evidence suggests that the initial increase in CBF develops well before this, and CBF may even be falling by the time of onset of symptoms'. This prospective study used transcranial Doppler (TCD) ultrasonography to evaluate serially changes in MCAV in 37 consecutive patients during the first 72 h after carotid endarterectomy with the aim of identifying factors associated with this phenomenon.
Patients and methods Assessment before operation Twenty-three of the 37 patients presented with a transient ischaemic attack (TIA) or amaurosis fugax. Fourteen presented with a completed stroke, 11 of whom had a persisting neurological deficit at the time of operation. No operation was carried out on stroke patients within 3 months of onset of symptoms. All patients underwent carotid angiography, computed tomography (CT), TCD ultrasonography and assessment of cerebrovascular reserve (CVR) before operation.
0007-1323/93/121523-05
Q 1993 Butterworth-Heinernann Ltd
The degree of stenosis of the internal carotid artery was determined as the maximal biplane diameter reduction expressed as a percentage of the normal internal carotid artery diameter above the stenosis6. Thirty of the 37 patients had internal carotid artery stenosis >, 50 per cent, eight of whom had stenosis between 50 and 7 0 per cent. Seven patients had internal carotid artery stenosis < 50 per cent and had been randomized to surgery in the European Carotid Surgery Trial before publication of the interim results. Ten patients had a localized region of hypodensity on CT consistent with an area of old infarction. Five recordings of MCAV were measured in both the symptomatic and asymptomatic hemispheres in the manner described by Aaslid and colleagues7together with measurements of blood pressure and end-tidal carbon dioxide. End-tidal carbon dioxide was measured, using an infra-red capnograph, from continuous samples of air drawn from a tight-fitting face mask. The measurement was made because of the known dependency of MCAV on arterial partial pressure of carbon dioxide (PaC0J8. CVR was evaluated to assess the haemodynamic state of the cerebral circulation before operation. Above a critical internal carotid artery stenosis, the brain attempts to maintain normal CBF by progressive pial vasodilatation, which increases the cerebral blood volume (CBV). Thus, while CBF may be maintained at near-normal levels, the progressive increase in CBV means that the cerebral arterioles have a reduced capacity to undergo further vasodilatation when stressed (the concept of impaired CVR). Conventionally, CVR is evaluated by measuring the increase in either blood flow or blood flow velocity in response to a vasodilatory stimulus, such as inhaling carbon d i ~ x i d e ~ . ~ . Using positron emission tomography", it has been shown that the most sensitive indicator of CVR is the ratio of CBF to CBV. This ratio is an alternative expression for the reciprocal ofmean cerebral transit time (MCTT), and Gibbs et ~ 1 . ' proposed ~ that measurement of MCTT could provide a simpler method of evaluating CVR. The present study used a novel technique for imaging regional changes in MCTT to evaluate CVR. The theory, experimental confirmation and validation have been detailed elsewhere' '*". In summary, a bolus of 99"Tc-sodium pertechnetate was injected rapidly into an antecubital vein, with the patient lying supine in a head-extended position with the skull vertex over a small field of view y camera, equipped with a high-sensitivity collimator. Sequential frames were acquired using a 32 x 32 matrix at a rate of 3 s-' for 1 min, by which time the examination was complete. The method generates a colour parametric image of regional MCTT, while the computer calculates the mean transit time for each hemisphere. The normal range of hemispheric
1523
Hyperaemic response after carotid endarterectomy: A. R . Naylor et al.
MCTT (2.2-7.7 s) reflects the wide interindividual variability and the dependency of MCTT on end-tidal carbon dioxide''. Interhemispheric MCTT symmetry is, however, strictly maintained throughout the (median interhemiphysiological range of end-tidal carbon spheric MCTT difference 0.2 (95 per cent confidence interval ( c i ) (M.6) s). In normal healthy volunteers an increase in end-tidal carbon dioxide causes a significant reduction in MCTT in both hemispheres (while maintaining interhemispheric MCTT symmetry) that correlates directlyi2 with simultaneous increases in MCAV. Specifically, blood flow velocity changes12in direct proportion to the reciprocal of MCTT. Thus MCTT not only changes in the manner predicted through changes in MCAV, but also, because the reciprocal of MCTT is an alternative expression for CBF: CBV, the increase in MCAV (with increasing end-tidal carbon dioxide) measures changes in CBF: CBV, i.e. CVR. In patients with haemodynamically significant extracranial internal carotid artery stenosis causing impaired CVR, the CBV increases to maintain CBF and there is a corresponding increase in MCTT. This, therefore, produces significant interhemispheric MCTT asymmetry, reflecting the change in CBF:CBV. Impaired CVR can thus be recognized by prolongation" of hemispheric MCTT beyond 7.7 s or by the development' of interhemispheric MCTT asymmetry 20.7s. In this study, eight of the 37 patients had preoperative evidence of impaired CVR in the symptomatic hemisphere.
remained 27 per cent (95 per cent c.i. 19-37 per cent) above the admission level in the operated hemisphere. This hyperaemic phase was not simply a consequence of concomitant changes in either mean arterial pressure or end-tidal carbon dioxide (Table 2) and could not be predicted from the preoperative MCAV value. Two patients recovered from anaesthesia with a transient minor neurological deficit. Both had an initial surge in MCAV on clamp release but neither developed a hyperaemic response over the ensuing 72 h. Table 3 details the relationship between the percentage increase in MCAY and various parameters that might be associated with an increased likelihood of postoperative hyperaemia, if this were mediated by intraoperative ischaemic
250
Surgical technique Endarterectomy was performed under normotensive normocarbic general anaesthesia using systemic heparinization. All but one arteriotomy was closed with a vein patch and the nerve to the carotid sinus was not anaesthetized or divided in any patient. Systemicblood pressure and heart rate were measured continuously from a radial artery cannula and electrocardiogram respectively. A temporary indwelling shunt was employed in 19 patients using the criteria of slowing of the cerebral function monitor or internal carotid artery stump pressure < 50mmHg. Systolic and mean ipsilateral MCAV were measured continuouslv throughout the operation13.
0
r Clamp
L 9:45
A 1O:OO
1 ' 1
I
I
I
I
I
1
1
2 4 6 12 24 48 10:30 11:OO 10:15 10:45 11:15
Operation time
I
72
Time after operation ( h )
Assessment after operation On recovering from anaesthesia, two patients suffered a minor transient neurological deficit. One was thought to be embolic in origin and the other haemodynamic' '. Both patients were confirmed to have a patent internal carotid artery using duplex ultrasonography. Serial measurements of MCAV, mean arterial blood pressure and end-tidal carbon dioxide were made every hour for the first 6 h and then again at 12, 24, 48 and 72 h in all patients. On each occasion, five measurements were taken from both the operated and non-operated middle cerebral artery, in addition to measurements of arterial blood pressure and end-tidal carbon dioxide. These values were then averaged and a mean calculated for each time period.
Figure 1 Serial monitoring of systolic (-) and mean (---) middle cerebral artery blood j?ow velocity ( M C A V ) in the operated hemisphere during, and for 72 h after, carotid endarterectomy in a patient with preoperative evidence of impaired cerebrovascular reserve. Note the initial transient surge in M C A V immediately after carotid clamp release, followed by a more pronounced secondary hyperaemic phase over the next 24 h. A t I 2 h, MCA V was looper cent above the preoperative level
Statistical analysis MCAV was not normally distributed (Filliben's testI4) and so non-parametric analysis was used. Values in the text refer to the median (95 per cent c.i.).P
