Local Versus General Anesthesia for Carotid Endarterectomy

Health-Related Quality of Life

Although the overall 30-day all-cause mortality did not vary between the 2 groups, there was a trend toward lower 30-day mortality rates with LA compared with GA (1.9% vs 2.3%, respectively; P = .1). ¹ There was also a trend toward lower major stroke rates with LA compared with GA (0.8% vs 3.9%, respectively; P = .115).¹ Although not significant, the relatively large difference in stroke rates with GA (3.9%) compared with LA (0.8%) should be considered. The GALA trial also reported a stroke rate ranging from 3.7% (LA) to 4.0% (GA) ² ; these rates are probably too high. Many institutions have stroke rates ≤1.0% (including the authors themselves who reported a 0.8% stroke rate for CEA under LA). ¹ This difference in major stroke rates was counterbalanced by a nonsignificant higher incidence of 30-day transient ischemic attack (TIA) rates after LA compared with GA (2.3% vs 0.8%, respectively; P = .42). ¹ The TIAs, however, have a considerably smaller impact on the patient’s health-related quality of life ([HRQoL]; decreased working capacity, decreased physical and social functioning, depression, cognitive impairment, and general well-being) compared with a minor/major stroke. ^3–5 The decrease in the patient’s HRQoL after a TIA/stroke is proportional to the degree of disability following the cerebrovascular event. ^3–5

Similar results were also observed in the GALA trial. ² Slightly more (but not significantly so) patients in the GA group had a stroke (70 vs 66 patients, or 4.0% vs 3.7%, respectively). Death occurred between randomization and 30 days after anesthesia in slightly more (but not significantly so) patients under GA than under LA (26 vs 19 patients or 1.5% vs 1.1%, respectively). Therefore, CEA under LA may be associated with marginally improved outcomes and a better HRQoL compared with GA although this remains to be proven.

Shunt Usage in CEA Under LA Versus GA

An earlier commentary on the GALA ⁶ trial posed a question regarding the clinical significance in the difference in shunt usage between the 2 procedures (14% vs 43%, for CEA under LA vs GA, respectively). ² A similar (but considerably larger) difference in shunt usage was noted in the study by Sideso et al (9.2% vs 82.2%, for CEA under LA vs GA, respectively). ¹ Several other studies comparing CEA under LA with GA also demonstrated a higher rate of shunt usage after GA compared with LA. ^7–11

Some of the advantages of performing CEA under LA compared with GA include the more appropriate and less frequent shunt use, fewer cardiorespiratory complications, and preserved cerebrovascular autoregulation. ¹² As mentioned in the GALA trial, ² although shunts protect the brain from stroke due to low cerebral blood flow during carotid clamping, they can damage the arterial wall, thus causing embolism to the brain.

The use of LA may help identify more accurately which patients need a shunt. Selective shunting can also be used with GA, but only if a continuous electroencephalograph and other expensive equipment is available. Furthermore, specialized personnel are required to interpret the readings of the equipment used. Finally, there is always the chance of false-positive and false-negative signals.

A recent comprehensive review evaluated the role of cerebral monitoring during CEA. ¹³ Cerebral monitoring during CEA (electroencephalography, transcranial Doppler, stump pressure, and sensory-evoked potentials) allows detection of the main causes of perioperative stroke, that is, embolism, intraoperative hypoperfusion, and postoperative hyperperfusion syndrome. ¹³ Near-infrared spectroscopy is a relative novel and noninvasive technique which has already demonstrated some positive results ¹⁴ ; these however, need to be validated in larger studies. Although some physicians may question the value of cerebral monitoring and consider it a costly and time-consuming procedure, others consider it essential as it provides direct information regarding new neurological deficits, which would otherwise be missed. ^13,14 The immediate feedback provided allows the physician to promptly correct any intraoperative defects. Unfortunately, although several modalities exist, no single test is definitive. Thus, a combination of several monitoring modalities may be essential.

It was recognized almost 2 decades ago that if brain ischemia after carotid clamping does not occur (due to adequate collateral circulation from the contralateral carotid artery), a much higher stroke rate with shunting is observed (8.8% vs 2.1%, for shunt usage vs nonusage, respectively; P < .0001). ¹⁵ An independent study verified this theory by showing that patients undergoing electrophysiological monitoring with selective shunting are >7 times less likely to experience a perioperative stroke (odds ratio [OR]: 0.05; 95% confidence interval: 0.01-0.40; P < .01). ¹⁶ Most patients have adequate collateral circulation to tolerate up to at least 30 minutes of unilateral carotid clamping without suffering ischemic damage. ^6,15 In these individuals, shunting is not only unnecessary but it is also an important risk factor for postoperative stroke. ^6,15 Selective shunting may be preferable based on hemodynamic/electrophysiological monitoring, the presence of poor internal carotid artery back bleeding, or contralateral carotid artery critical stenosis/occlusion for the detection of those few patients where considerable cerebral ischemia occurs. ⁶

Cost-Effectiveness and Length of Hospitalization

The CEA under LA may be a more cost-effective option compared with GA. Carotid endarterectomy under LA is associated with lower intensive care unit (ICU) requirement, length of in-hospital stay, and consequently decreased costs compared with CEA under GA. ⁹ A subgroup analysis of the GALA trial showed that CEA under LA was more cost-effective by a mean £178 per patient compared with GA. ¹⁷ The cost-effectiveness of CEA under LA compared with GA was verified in an independent study ($885.71 ± 78.57 vs $1007.14 ± 135.71, respectively; P < .0001). ¹⁸ Carotid endarterectomy under LA was also associated with a reduced hospitalization period compared with GA (2.4 ± 1.1 vs 4.1 ± 1.9 days, respectively; P < .0001). ¹⁸ Another study comparing the outcomes after CEA under LA versus GA showed that patients operated under LA had a considerable shorter length of in-hospital stay compared with those receiving GA (median [range]: 3 [2-10] vs 4.5 [3-14] days, respectively; P < .001). ¹⁰ These results were verified in a prospective study from Bath, United Kingdom, showing that patients undergoing CEA under LA had a reduced in-hospital stay (median [interquartile range]: 2 [1-2] vs 3 [1-4] days, respectively) and a cost saving of £235/CEA procedure compared with CEA under GA (for both associations, P < .001). ⁸

A well-designed study investigated ways to improve the cost-effectiveness of CEA. ¹⁹ A total of 63 CEAs performed in 60 patients were entered into a clinical protocol that included avoidance of cerebral angiography, preferential use of LA, selective use of ICU, and early hospital discharge. The outcomes of these 63 CEAs were then compared with 45 standard CEAs. Implementation of the CEA protocol resulted in reductions in the use of angiography (13% vs 74%, respectively; P < .001), GA (24% vs 100%, respectively; P < .001), ICU use (30% vs 98%, respectively; P < .001) and mean in-hospital length of stay (2.0 vs 5.8 days, respectively; P < .001). ¹⁹ These changes resulted in a 41% reduction in mean total hospital cost ($5699 vs $9652, respectively; P < .001) and a 124% reduced mean net hospital costs ($1804 vs $4039, respectively; P < .001) per CEA procedure. For the 39 patients (62%) who achieved all elements of the CEA protocol, the mean hospital length of stay was 1.3 days, the mean hospital cost was $4175 and the mean hospital savings ¹⁹ were $4327. This study demonstrated the cost-effectiveness of CEA performed under LA compared with GA. ¹⁹

Postoperative Delirium, Cognitive Function, and Cardiovascular Outcomes

The majority of patients undergoing CEA are >70 years of age. Elderly patients undergoing surgery under GA have increased rates of postoperative delirium and cognitive impairment. ^20,21 Postoperative delirium may result in increased morbidity and mortality, delayed functional recovery, prolonged hospital stay, and consequently increased hospital costs. ^20,21 The incidence of postoperative delirium is increased in patients undergoing vascular and orthopedic surgery. ²¹ Thus, employment of LA instead of GA for CEA should reduce the risk of postoperative delirium.

Carotid endarterectomy under GA is also associated with early postoperative cognitive dysfunction compared with LA. In a subgroup analysis of the GALA trial, CEA under LA was associated with improved postoperative neurocognitive performance based on the results of the neuropsychological Trail Making Test (ie, less cognitive dysfunction) compared with GA. ²² Thus, performance of CEA under LA may protect the patient from early postoperative cognitive dysfunction.

A study comparing the outcomes of CEA under GA with LA showed that more patients operated under GA have hypertensive events (systolic blood pressure >180 mm Hg) on the first postoperative day compared with individuals operated under LA. ²³ Postoperative high systolic blood pressure (>150 mm Hg) after CEA is implicated in the development of grave complications, such as elevated intracranial pressure, intracerebral hemorrhage, and the cerebral hyperperfusion syndrome. ²⁴ Postoperative hypertension after CEA is also associated with an increased incidence of stroke or death (P = .04). ²⁵

It has been suggested that high-risk patients may be more appropriate candidates for the less invasive CAS compared with CEA (performed under GA). ^26,27 A well-performed study compared the perioperative complication rates between high-risk (American Society of Anesthesiology [ASA] 4 classification, “hostile neck,” recurrent internal carotid artery stenosis, contralateral internal carotid artery occlusion and age ≥80 years) and low-risk patients undergoing CEA under LA. ²⁸ This study demonstrated that with the use of LA, CEA was not associated with worse outcomes in high- compared with low-risk patients. The conclusion reached was that “it does not appear justified to refer high-risk patients principally to CAS when LA can be chosen to perform CEA.” ²⁸ An independent study verified that CEA under LA can be safely performed in the elderly patients and in high-risk patients. ²⁹

Conclusions

Although CEA under LA may be associated with similar stroke, myocardial infarction, and death rates compared with GA, ^1,2 several additional outcomes may need to be taken into account when comparing the 2 procedures. Examples include the patient’s HRQoL, the possible clinical significance, and long-term effects of the difference in shunt usage between the 2 procedures, the cost-effectiveness and length of hospitalization after each technique, as well as the incidence of postoperative delirium, cognitive dysfunction, and cardiovascular outcomes (eg, postoperative blood pressure differences). According to a recent report, the hypothesis regarding the benefit of CEA under LA compared with GA cannot be answered.³⁰

As the incidence of stroke and death after both procedures is low, it is unlikely that any trial will show any meaningful difference between LA and GA. It may therefore be appropriate to consider other additional parameters (like the ones discussed in this commentary) when comparing the outcomes of the 2 techniques. These issues remain to be clarified in future trials.