Very Urgent Carotid Endarterectomy Confers Increased Procedural Risk
Background and Purpose—Current Swedish guidelines recommend that carotid endarterectomy should be performed within 14 days of a qualifying neurological event, but it is not clear if very urgent surgery after an event is associated with increased perioperative risk. The aim of this study was to determine how the time between the event and carotid endarterectomy affects the procedural risk of mortality and stroke.
Methods—We prospectively analyzed data on all patients who underwent carotid endarterectomies for symptomatic carotid stenosis between May 12, 2008, and May 31, 2011, with records in the Swedish Vascular Registry (Swedvasc). Patients were divided according to time between the qualifying event and surgery (0–2 days, 3–7 days, 8–14 days, 15–180 days). Stroke rate and mortality at 30 days postsurgery were determined.
Results—We analyzed data for 2596 patients and found that the combined mortality and stroke rate for patients treated 0 to 2 days after qualifying event was 11.5% (17 of 148) versus 3.6% (29 of 804), 4.0% (27 of 677), and 5.4% (52 of 967) for the groups treated at 3 to 7 days, 8 to 14 days, and 15 to 180 days, respectively. In a multivariate analysis, time was an independent risk factor for perioperative complications: patients treated at 0 to 2 days had a relative OR of 4.24 (CI, 2.07–8.70; P<0.001) compared with the reference 3- to 7-day group.
Conclusions—In this study of patients treated for symptomatic carotid disease, it was safe to perform surgery as early as Day 3 after a qualifying neurological event in contrast to patients treated within 0 to 2 days, which has a significantly increased perioperative risk.
The risk of recurrent cerebrovascular events is extremely high within the first few days of a symptomatic event; meta-analyses have shown that the risk of recurrent stroke after a transient ischemic event (TIA) is 6.7% at 2 days and 10% at 7 days.1,2 The value of carotid endarterectomy (CEA) for symptomatic arteriosclerotic carotid stenosis has been clearly established through large prospective randomized trials.3–5 Pooled analyses from endarterectomy trials have shown that benefit from surgery is greatest in those randomized to CEA within 14 days from their most recent symptomatic event and that benefit fell rapidly with delay of surgery.6,7 These results have led to a change in treatment guidelines to promote early intervention.8–14 In Sweden, the current recommendation is for intervention within 14 days,15 and guidelines in the United Kingdom recommend intervention within 2 days for symptomatic severe carotid stenosis in neurologically stable patients.16 However, the benefit of very early CEA is questionable because of a potential risk of higher perioperative complications. Studies that report risk in urgent (0–2 days) CEA are very few and include only a small number of patients.17–19
The aim of this study was to determine the procedural risk of CEA for symptomatic carotid stenosis in relation to the time from the qualifying neurological event in a large population-based cohort. We also used a multivariate logistic regression analysis to find predictors of the risk of stroke and death within 30 days of CEA.
Subjects and Methods
The Swedish Vascular Registry (Swedvasc) has been nationwide since 1994 and covers all 22 vascular centers performing CEA in Sweden. The registry records precisely defined and standardized details of demography (age and gender), comorbidities, surgical technique, surgical center, responsible surgeon, and postoperative outcome at 30 days. In May 2008, the registry was upgraded to include a separate carotid module to prospectively study whether the time from a qualifying neurological event to CEA affects procedural risk. The registry now records standardized information on the date of the qualifying event (the neurological event that brought the patient to medical attention) as well as neurological symptoms that occurred before and after the qualifying event, modified Rankin Scale used in case of stroke, and the grade of ipsi- and contralateral stenosis are registered. The mortality data in the register are continuously crosschecked against the National Population Register and are thereby of the same high quality as official Swedish national statistics (100%). The Swedvasc steering committee recently performed a validation of the latest version, including the carotid module, in which a crosscheck against the Swedish In-Patient-Register was performed. These results have shown 93.7% to 100% completeness.20,21
We obtained data for patients who underwent intervention for symptomatic carotid stenosis in Sweden between May 12, 2008, and May 31, 2011, and were recorded in Swedvasc (n=2763). The study was approved by the regional ethical review board in Gothenburg.
Definition of Qualifying Event and Time to Intervention
The qualifying event was categorized as one of the following: amaurosis fugax (ocular TIA), TIA (restituted within 24 hours), crescendo TIA (daily ischemic events, including progressive stroke and stroke in evolution), minor stroke (symptoms restituted within 1 week or remaining minor dysfunction), or major stroke (disabling stroke). Time to intervention was defined as the number of days between the qualifying neurological event (Day 0) and the day of surgery (Day 0–180). Individuals who received surgery after Day 180 were considered to have asymptomatic carotid stenosis and were therefore not included in our analysis. The study population was divided into 4 subgroups depending on time to intervention: 0 to 2 days, 3 to 7 days, 8 to 14 days, and 15 to 180 days.
Definition of Procedural Risks
The surgeon or a certified neurologist registered the outcome data at the 30-day follow-up. Procedural adverse events were defined as mortality, minor stroke, or major stroke within 30 days of surgery. The register also includes neuroimaging-derived details regarding subtype of stroke divided in ischemic or hemorrhagic stroke. Stroke-related deaths are included as both mortality and major stroke.
Analyses were performed using SPSS (19.0). Categorical variables were compared using χ2 test significance analysis and continuous variables were compared using analysis of variance and expressed as mean with SD. Multivariate associations were assessed by a logistic regression analysis and results are presented as ORs. Significance was assumed at P<0.05.
We analyzed data from 2763 patients who underwent CEA for symptomatic carotid stenosis. We excluded 164 patients registered as receiving carotid artery stenting, ligature, exploration, transposition, or crossover bypass procedures because of the small numbers, and a further 3 patients were excluded because we did not have complete data at the 30-day follow-up. In total, 2596 patients were included in our final analysis.
Table 1 shows demographics, risk factors, surgical characteristics, and qualifying neurological events of the study population subdivided according to time to intervention. There were no significant differences between the subgroups regarding demographics except that the population treated 0 to 2 days after the qualifying event was younger and had a lower incidence of diabetes than the other groups. As expected, the incidence of crescendo TIA was higher in the 0- to 2-day subgroup and major stroke was more common in the group treated after 15 to 180 days.
Procedural Adverse Events
Table 2 shows the 30-day stroke and mortality rates in the study population subdivided according to time to intervention. The overall 30-day rate for mortality and any stroke was 4.8%. There was no significantly increased perioperative risk when intervention at ≤14 days was compared with later surgery (15–180 days; 4.5% versus 5.4%). When complication rate was stratified for all subgroups, there was a significantly higher risk of the combined end point of mortality and any stroke in patients undergoing CEA 0 to 2 days after the qualifying event compared with the groups treated after 3 to 7 days, 8 to 14 days, and 15 to 180 days (11.5% versus 3.6%, 4.0%, and 5.4%, respectively). There was no significant difference between subgroups when mortality alone was analyzed.
Analysis of preoperative modified Rankin Scale and its association to postoperative complications showed no significant difference in relation to timing of surgery.
The proportion of ischemic versus hemorrhagic stroke was also monitored and presented in online-only Supplemental Table I (http://stroke.ahajournals.org) and was not influenced by the delay of surgery.
Characteristics of the 17 patients who had a perioperative complication in the subgroup undergoing CEA 0 to 2 days after a qualifying event are presented in online-only Supplemental Table II. These data were crosschecked against the medical records at each hospital.
Median time to intervention ranged from 5 to 22 days at the 22 different surgical centers. To investigate if an ambitious early treatment strategy at the surgical unit level was associated with increased procedural risk, we identified 5 “early treatment” centers where the median time to intervention was ≤8 days. These centers performed 25% of the total interventions (643 of 2596) and 42% of the procedures at 0 to 2 days (62 of 148). The early treatment centers operated on 9.6% of their patients at 0 to 2 days (compared with 4.4% at the remaining centers). The overall complication rate (mortality and any stroke) did not differ between the early treatment centers and the remaining centers either when all the time-to-intervention subgroups were combined (4.4% versus 5.0%) or in the 0- to 2-day subgroup (11.3% versus 11.6%).
We also grouped the centers according to their procedural risk. We identified 6 centers that had a complication rate of >1.5 times the mean complication rate of the total population. These centers performed 15% of the total interventions (398 of 2596). Their mean procedural risk was 10.6% with no significant difference between the time-to-intervention subgroups: 0 to 2 days 12.5%, 3 to 7 days 11.0%, 8 to 14 days 9.4%, and 15 to 180 days 11.0%. The remaining centers had an overall complication rate of 3.8% with a significant difference between the time groups: 0 to 2 days 11.4%, 3 to 7 days 2.8%, 8 to 14 days 2.7%, and 15 to 180 days 4.2% (P<0.001).
Multivariable Predictors of Perioperative Outcome
We performed a multivariate analysis in which we controlled for demographics, clinical and surgical characteristics, and qualifying event and showed that time to intervention was a significant independent predictor of risk (Table 3). The subgroup treated 0 to 2 days after the qualifying event showed a 4-fold increased risk of having a perioperative complication compared with the reference subgroup that received surgery after 3 to 7 days. There was no increased risk for patients treated 8 to 14 days after the qualifying event compared with the reference group. However, the group that was treated 15 to 180 days after an event showed a significantly increased risk compared with the reference group (OR, 1.90).
Diabetes and female gender were also independent predictors of increased risk. Neither crescendo TIA as a qualifying event nor occurrence of contralateral occlusion was a significant independent factor of increased risk. However, amaurosis fugax as a qualifying event was associated with reduced risk.
Our prospective and population-based analysis of the perioperative risks associated with very early CEA is, to our knowledge, the largest such study reported to date. Our main objective was to determine how time to surgery from a qualifying neurological event affects the procedural risk of mortality and stroke by stratifying a large study population into 4 subgroups according to time to surgery. We showed that performing CEA as early as 3 to 14 days after the qualifying neurological event was safe compared with later surgery in agreement with recent studies.19,25–27 Our data thus support recent changes in guidelines advocating the need for early intervention.16 However, we also found that CEA performed 0 to 2 after the qualifying event was associated with a 4-fold higher procedural risk compared with surgery performed at 3 to 7 days. The increased risk was independent of other known and recorded confounding risk factors. Our study thus shows the importance of considering procedural complications associated with very early surgery. It is intriguing to speculate if it is worthwhile taking this increased procedural risk to protect against the very high risk of recurrent event in this time period.28 Randomized trials are necessary to determine whether the increased risk of recurrent events will outweigh the increased risk of complications. Such a trial will necessarily have to be large. If the same postoperative risk level like in our study is assumed, a randomized study will have to incorporate >1100 patients in each arm to reach a power of 80% (online-only Supplemental Table III). In the light of this, it is interesting to note that the group in which CEA was performed 0 to 2 days after a neurological event was small in relation to our whole study population (n=148 [5.7%]) but still represents the largest series of urgent CEAs so far reported.
We analyzed if the increased procedural risk in the subgroup treated 0 to 2 days after an event was associated with a high frequency of unstable symptoms in these patients before surgery. Examination of the medical records from patients who had a perioperative complication in the subgroup did not reveal any signs of instability; only 1 patient was classified as having a crescendo TIA in the registry. These data thus suggest, albeit indirectly, that the increased risk of very early surgery is not explained by a selection bias caused by aggregation of patients with unstable symptoms.
It is reasonable to assume that patients with an inherent perioperative risk are equally distributed among the different surgical centers. Thus, it can also be assumed that treatment centers with an early intervention strategy would partially overcome any potential selection bias caused by aggregation of patients with unstable symptoms at early time points by including a higher proportion of stable patients in the 0- to 2-day group compared with the other centers. Because the complication rate in the 0- to 2-day subgroup remained high in surgical centers with an expedited intervention strategy, it is likely that our results indicate an increased risk of the very early surgery itself.
Among the 19 different potential risk factors that we monitored in the registry, we identified 6 variables that were independently associated with the risk of perioperative stroke or death after CEA in a multivariate regression analysis. We confirmed earlier findings showing that patients with amaurosis fugax as a qualifying event had a substantially lower risk for complications.7,29 We also confirmed that procedural risks were higher in women and in patients with diabetes.30,31 In contrast, we found no significant risk associated with the presence of a contralateral occlusion or with a history of cardiac disease.29,30,32,33 Unstable neurological symptoms as a qualifying event (crescendo TIA) have previously been associated with an increased perioperative risk, but this was not evident from our data.29
The strength of our study is that we analyze a large population from a national registry. The external validity of the registry is high and covers >93% of all CEAs performed in Sweden.20 Because the data are prospectively gathered from a heterogeneous unselected population in everyday clinical practice, they are therefore well suited to monitor the effects of changes in guidelines on the timing of CEA. However, it is important to note that our study is not randomized and there may be differences between the subgroups that we did not identify. A potential weakness with the registry is self-reported data. However, the mortality data in Swedvasc are continually validated against official death certificates and a number of reports have validated other aspects of the registry (Swedvasc yearly report).22–24 Another potential limitation is that the registry is updated either by surgeons or neurologists in the clinical practice, although this factor did not affect the outcome in relation to delay of surgery in our analysis. It is not clear why there should be an increased risk with very early CEA, but potential explanations are that the brain is more vulnerable to ischemia so soon after a neurological event or that the carotid plaques are more vulnerable at early time points and thus hazardous to remove without complications.34,35
Prompt medical therapy has been shown to be beneficial in the acute phase, perhaps due to a plaque-stabilizing effect.28
We showed that perioperative risk is not increased when CEA is performed within 14 days of a qualifying neurological event but is significantly higher when surgery is performed within the first 2 days. On the basis of these results, we suggest that procedural risks of very urgent CEA should be cautiously monitored. More studies of procedural risks in this time period are required before general recommendations on extremely early intervention can be given.
Sources of Funding
This work was supported by the Swedish Foundation for Strategic Research, the Swedish Heart and Lung Foundation, LUA/ALF (Grants from the Swedish state under the agreement between the Swedish government and county councils concerning economic support for research and education of doctors), and Swedvasc.
We thank Dr Rosie Perkins (University of Gothenburg) for editing the article. This report had not been possible without the continuous registration efforts over the years by Swedish vascular surgeons. The Steering Committee of the Swedvasc: Anders Lundell (chairman), Ken Eliasson, Ingvar Jansson, Lars Karlström, Björn Kragsterman, Joakim Nordanstig, David Lindström, Maria Truedsson, and Marlene Hensäter. Senior advisors: David Bergqvist, Lars Norgren, and Thomas Troëng.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.111.639344/-/DC1.
- Received September 18, 2011.
- Revision received January 11, 2012.
- Accepted January 19, 2012.
- © 2012 American Heart Association, Inc.
- Pritz MB
- Sacco RL,
- Adams R,
- Albers G,
- Alberts MJ,
- Benavente O,
- Furie K,
- et al
- Werring DJ,
- Brown MM
Department of Health. The national stroke strategy. Available at: www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_083506. Accessed January 11, 2012.
- Bergqvist D,
- Troeng T,
- Elfstrom J,
- Hedberg B,
- Ljungstrom KG,
- Norgren L,
- et al
- Rerkasem K,
- Rothwell PM
- Rothwell PM,
- Eliasziw M,
- Gutnikov SA,
- Warlow CP,
- Barnett HJ
- Halm EA,
- Tuhrim S,
- Wang JJ,
- Rockman C,
- Riles TS,
- Chassin MR
- Rothwell PM,
- Giles MF,
- Chandratheva A,
- Marquardt L,
- Geraghty O,
- Redgrave JN,
- et al
- Bond R,
- Rerkasem K,
- Rothwell PM
- Rothwell PM,
- Slattery J,
- Warlow CP
- Peeters W,
- Hellings WE,
- de Kleijn DP,
- de Vries JP,
- Moll FL,
- Vink A,
- et al