Local anaesthesia for carotid endarterectomy: Con: decrease the stress for all.

Abstract

This Editorial is part of a Pro and Con debate and is accompanied by the following articles: Licker M. Regional or general anaesthesia for carotid endarterectomy. Does it matter? Eur J Anaesthesiol 2016; 33:241–243. Cedergreen P, Swiatek F, Nielsen HB. Local anaesthesia for carotid endarterectomy. Pro: protect the brain. Eur J Anaesthesiol 2016; 33:236–237. There is still insufficient evidence to support the use of regional anaesthesia (RA) over general anaesthesia (GA) in patients undergoing carotid endarterectomy (CEA). Advantages for both patient and surgeon mean that for GA, the benefits outweigh the disadvantages. The evidence for a better outcome after RA is lacking. Advantages of general anaesthesia For the surgeon, general anaesthesia facilitates optimal positioning of the patient and offers a motionless field with better visualisation during CEA. For the patient, GA is easier to tolerate particularly for those who are restless, anxious or claustrophobic, and is the best option for those with joint problems or neurologic disorders.1 GA avoids all the potential risks and complications associated with the use of local anaesthetics and deep cervical block.1 It also avoids agitation in the awake patient caused by the ‘tent’ of drapes around the face and the need to overcome poor access to secure the airway urgently. One of its main advantages is the secure airway throughout the procedure, preventing the adverse effects of hypocarbia and hypercarbia on cerebral blood flow by controlling gas exchange and oxygenation by artificial ventilation.2 GA can be maintained with inhalational or intravenous agents that have neuroprotective effects, and offers the potential benefits of cerebral preconditioning and postconditioning.3 If required, GA may facilitate management of complications such as cerebral ischaemia with induced hypertension and pharmacologic suppression of electroencephalographic activity.3 It has been shown that propofol anaesthesia provides better haemodynamic conditions during carotid artery clamping,4 but there are insufficient data to suggest that any single anaesthetic technique is superior. Disadvantages of general anaesthesia The main objection to GA for CEA is that it does not allow optimum monitoring of cerebral function during cross clamping.1 It is accepted that the direct neurological monitoring of the awake patient during RA is the ‘gold standard’, and it eliminates the necessity for equipment for that purpose.1,5 There are numerous cerebral function monitoring devices and each has its own problems with detecting inadequate cerebral perfusion during GA.5 Furthermore, general anaesthetics have the potential to impair cerebral autoregulation, although newer inhalational agents have less effect if concentrations are kept below 1 Minimum alveolar concentration.2 The persistence of residual effects from GA in the early postoperative period may mask the signs or symptoms of neurological complications related to surgery. Genuine neurological deficit could be wrongly attributed to the effects of anaesthesia because its initial manifestation is often nondistinct agitation or sedation, rather than a clear motor deficit. Outcome after general anaesthesia (death, stroke and myocardial infarction, barorepector sensitivity) Studies list considerable benefits resulting from the use of RA. These include a reduction in mortality and major morbidity such as stroke, myocardial infarction (MI) and pulmonary complications, decreased surgery time and length of hospital stay, less use of intensive care facilities and potential cost savings. Unfortunately, these claims are often based on retrospectively collected data or nonrandomised prospective studies,6 as indeed are those studies that have failed to demonstrate a difference in the incidences of stroke and death.7 In 2007, Guay7 prospectively analysed data collected from six randomised controlled trials with a total of 468 procedures, eight nonrandomised trials with 1407 procedures and retrospective data from 34 studies with 15 153 procedures. He found that with RA there were statistically significant reductions in the incidences of stroke (P < 0.00001), stroke and/or death (P < 0.0001), death (P < 0.004), any new neurologic impairment (P < 0.00001), MI (P < 0.0001), arrhythmia, respiratory complications and a reduction in the length of surgery. But when only randomised trials and prospective studies were considered, there was no statistically significant difference in outcomes between the two anaesthesia techniques.7 The main disadvantage of this meta-analysis was that there were insufficient patients in the randomised trials or even in the prospective studies to avoid a type-β error and justify any conclusion. The minimal number of patients needed to eliminate a type-β error was 6622.7 Also, the quality of this ‘evidence of efficacy’ cannot be considered high because retrospective studies are more prone to bias for various reasons.7 A prospective, peer-controlled database8 with 24 716 patients showed no difference in the 30-day incidences of composite stroke, MI and death between GA and RA groups. A propensity score analysis of 8050 patients in the RA cohort showed no differences in the major outcomes based on the anaesthetic type, but with RA there were shorter average operative times (RA 99 ± 36 min vs. GA 119 ± 53 min; P < 0.0001) and anaesthetic times (RA 52 ± 29 min vs. GA 64 ± 37 min; P < 0.0001), and a higher percentage of patients discharged the next day (RA 77.0% vs. GA 64.4%; P < 0.0001).8 Although the P values might impress, the clinical significance of a 30-min advantage in duration of anaesthesia and surgery is questionable, and whatever the choice of anaesthesia, the majority of patients were discharged on the first postoperative day.8 Although this study included a large number of patients, it was not randomised and consequently subject to selection bias. Also, propensity score matching cannot be considered equivalent to randomisation. The General anaesthesia versus local anaesthesia (GALA) study, a large randomised multicentre study with 3526 patients, was intended to provide an answer to the question of which anaesthesia technique is better for CEA.9 It also failed to find a significant difference between the two anaesthetic techniques in terms of the primary outcome, although in the RA group there was a marginally lower stroke rate (GA 4.0% vs. RA 3.7%) and mortality rate between the time of randomisation and 30 days after anaesthesia (GA 1.5% vs. RA 1.1%).9 Other differences were, with RA, a nonsignificant increase in the rate of MI (RA 0.5% vs. GA 0.2%) and in the 30-day rate of transient ischaemic attacks (2.3 vs. 0.8%; P = 0.42).9 There was an apparent divergence of survival curves of the RA and GA patients at 1 year, with more patients of the GA group dying, but that difference was not statistically significant (log-rank test, P = 0.094).9 Although the largest randomised study at present, the GALA study was underpowered to detect a difference in the incidence of stroke and death. The study would have needed about 25 000 patients (α = 0.05 two-tailed, β = 0.2), to detect a significant difference in mortality.7 A recent Cochrane review that included 14 randomised trials and 4596 operations came to a similar conclusion.10 In general, methodology of included studies was poor; there was no statistically significant difference in the incidences of stroke (RA 3.2%; GA 3.5%), stroke or death within 30 days of surgery (RA 3.6%; GA 4.2%) and MI (RA 0.6%; GA 0.4%).10 There was a nonsignificant trend towards lower operative mortality with RA [RA 0.9%; GA 1.5%; odds ratio 0.62 (95% confidence interval 0.36 to 1.07)]. The pooled analysis of randomised studies was not adequately powered to reliably detect an effect on mortality. Cardiovascular lability in patients undergoing anaesthesia and surgery for CEA is a well documented problem. Several studies that show that GA is associated with a more labile blood pressure and the need for vasopressors to maintain mean arterial pressure during surgery have suggested that use of RA may lead to a better outcome.1 During GA, there is an increased likelihood of intraoperative hypotension and postoperative hypertension.11 In contrast, in the RA group, more patients have intraoperative hypertension and postoperative hypotension. The GALA study reported that significantly more GA than RA patients had their blood pressure supported (43 vs. 17%), and more RA patients had their blood pressure reduced or not manipulated at all (74 vs. 41%).9 Both hypertension and hypotension may provoke myocardial ischaemia by increasing myocardial oxygen demand during the former and reducing coronary perfusion pressure during the latter. Also both require administration of intravenous vasoactive medication that is associated with increased perioperative mortality, stroke, MI, heart failure and also 1-year mortality.12 Although RA might appear to have an advantage with regard to a more stable arterial pressure and a reduced vasopressor requirement, a higher incidence of tachycardia and increased blood levels of catecholamines occurs.13 In a randomised study of 75 patients with RA, Takolander et al.13 demonstrated higher intraoperative values of plasma noradrenaline and adrenaline, and found a positive correlation between noradrenaline and adrenaline, and mean blood pressure and heart rate. Marrocco-Trischitta et al.,14 in a prospective nonrandomised study of 113 consecutive CEAs, observed that the intraoperative stress response (cortisol and adrenocorticotropic hormone levels) was significantly greater in the RA group (P < 0.001 for both).14 After CEA, the stress response was abolished within 24 h regardless of the type of anaesthesia.14 This may mean that RA exposes patients to undue stress and pain during surgery, resulting in an increased risk of myocardial ischaemia. Experience with general anaesthesia in the authors’ hospital Our hospital is a high-volume centre for CEA, with more than 800 procedures per year. Most operations (99.4%) are performed under GA.15 Our surgical colleagues have published data from our 20-years’ experience of CEA in 9897 patients: contralateral neurologic deficit occurred in 0.92% and stroke-related mortality in 0.52%.15 If the use of GA confers disadvantage, our experience is that this has not been translated into poor outcomes.15 Taking our own experience together with the evidence from all publications to date, we find no reason to support a change in our practice to the use of RA. A further mega-trial is needed that is adequately powered to confirm, or to refute, a relevant difference in mortality. We intend to continue using GA for CEA until strong evidence of a relevant difference in major outcomes becomes available. Acknowledgements related to this article Assistance with the Editorial: none. Financial support and sponsorship: none. Conflicts of interest: none. Comment from the Editor: This Editorial is part of a Pro and Con debate. The Editorial was checked and accepted by the editors but was not sent for external peer review.