Safety of Simultaneous Coronary Artery Bypass Grafting and Carotid Endarterectomy Versus Isolated Coronary Artery Bypass Grafting

Supplemental Digital Content is available in the text.


Stroke
October 2017 See related article, p 2650 C oronary artery bypass graft (CABG) surgery is the most commonly performed major cardiovascular operation. Carotid artery stenosis is present in ≈6% to 8% of all patients undergoing CABG and is associated with an increased risk of stroke during and after CABG. 1,2 Prophylactic treatment of asymptomatic concomitant carotid artery stenosis is managed in different ways, for example, by carotid artery angioplasty and stenting (CAS) or carotid endarterectomy (CEA), either simultaneously with CABG, before CABG, or delayed after CABG (staged or reverse staged). For many years, staged or synchronous CEA has been advocated by many cardiovascular surgeons in the attempt to reduce the perioperative and long-term risk of stroke associated with carotid artery stenosis but only very few patients with this disease entity have been included in controlled clinical trials. 3,4 Only data from uncontrolled studies with variable inclusion criteria and end point assessment are available in patients with coexisting cardiac and carotid atherosclerotic disease undergoing CABG without carotid revascularization (by CAS or CEA). 5 Some studies have even found that asymptomatic carotid stenosis did not increase the risk of post-CABG stroke. 6,7 Therefore, in the absence of any randomized controlled trial, no systematic high-level evidence exists that staged or synchronous CEA and CABG confer any short-term benefit over CABG without CEA. Moreover, improvements in medical therapy have considerably reduced the average long-term risk of ipsilateral stroke in patients with asymptomatic carotid stenosis. 8 Thus, any potential long-term benefit conferred by prophylactic CEA may be offset by the relatively high procedural risk reported in systematic reviews. 9,10 Therefore, the CABACS trial (Coronary Artery Bypass Graft Surgery in Patients With Asymptomatic Carotid Stenosis) aimed to compare the perioperative safety and long-term efficacy of synchronous CEA and CABG versus isolated CABG in patients with asymptomatic high-grade carotid artery stenosis.

Study Design
This investigator-initiated trial was designed as a multicenter, randomized (one-to-one), open, group sequential trial with 2 parallel arms and blinded end point adjudication. After initial commitment of 35 major German cardiac surgery centers, 25 German and 1 Czech center could be initiated and 17 centers finally recruited. Ethics approval was obtained from each center. The trial was conducted according to Good Clinical Practice guidelines and the principles stated in the latest revision of the Declaration of Helsinki. The trial protocol has been described previously. 11 Two amendments to the protocol were implemented in 2012 and 2014 and approved by the local ethics committees of each participating center (online-only Data Supplement).
Enrollment was terminated early because of withdrawal of funding following insufficient recruitment.

Participants
The inclusion criteria were as follows: asymptomatic (past 180 days) internal carotid artery stenosis ≥80% (following criteria of the ECST [European Carotid Surgery Trial], 12 main criterion: in-stenosis peak systolic velocity ≥300 cm/s, corresponding to ≥70% NASCET [North American Symptomatic Carotid Endarterectomy Trial] 13,14 ), carotid artery stenosis treatable with CEA, negative pregnancy test in premenopausal women, written informed consent and full legal capacity, ability of the patient to participate in follow-up examinations.
Exclusion criteria were as follows: nonatherosclerotic stenosis (eg, dissection, floating thrombus, fibromuscular dysplasia, tumor, and postradiation), complete occlusion or previous stenting of the carotid artery to be treated, additional higher grade intracranial or intrathoracic stenosis (tandem stenosis), recent (past 180 days) ischemic symptoms ipsilateral to carotid stenosis or occlusion, contralateral carotid occlusion or other known indication for carotid revascularization (apart from scheduled CABG), myocardial infarction (non-ST-segment-elevation myocardial infarction or ST-segmentelevation myocardial infarction) within the past 7 days (reduced to 48 hours for non-ST-segment-elevation myocardial infarction after the first amendment) or hemodynamically unstable patients, known high risk for cardiogenic embolism requiring anticoagulation (mechanical heart valve, chronic atrial fibrillation [omitted after the first amendment], left ventricular thrombus, left ventricular aneurysm), evidence for intracranial bleeding within the past 90 days, modified Rankin Scale score of >3 or severe aphasia, patients unlikely to survive >1 year because of concomitant diseases, planned combined cardiac valve replacement or any other cardiac surgery beyond CABG (±CEA) during the procedure, major surgery (apart from study procedures) planned within 8 weeks from randomization, and participation in another clinical trial. Written informed consent was obtained from each patient before trial participation.

Randomization
Eligible patients were randomized to either isolated CABG or synchronous CEA + CABG. To achieve comparable groups, patients were allocated in a concealed way by central preoperative randomization ≥1 day before surgery. To avoid unbalanced prognostic factor distributions, we used a web-based stratified block randomization (strata: center, age [(<60 or ≥60 years], sex [male or female], modified Rankin Scale [score of 0-1 or 2-3]) with randomly varying the block size.

Interventions
The eligibility of a patient was determined by both a certified study surgeon and a certified study neurologist with experience in cerebrovascular ultrasound examination. Additional imaging of the brain or cerebral circulation was not required or documented as part of the study. CABG with or without CEA under treatment with aspirin was to be performed as soon as possible (maximum within 7 days) after randomization. Standards for surgical treatment were formulated by a surgical quality subcommittee, which also had to approve every surgeon for the study. Each (cardio-) vascular surgeon had to meet the following standards to be certified: anonymous confirmation of the last 30 consecutively performed CEA surgeries, affirmed by the head of department, alone or in combination with anonymous confirmation of the last 150 consecutively performed CABG surgeries, affirmed by the head of department. With regard to carotid revascularization, no policies were prescribed for the use of a shunt, the type of patches, and the use of neurological monitoring or the method of carotid reconstruction (eversion endarterectomy or thromboendarterectomy). With regard to coronary revascularization, the use of extracorporeal circulation (on-or off-pump CABG) was left to the surgeon`s decision. Standards for medical treatment were formulated by a Best Medical treatment subcommittee.
Patients were followed up after 7 days (±1 day), 30 days (±3 days), and 1 year (±30 days) with a neurological examination including evaluation on the National Institutes Health Stroke Scale and the modified Rankin Scale at each visit as well as cerebrovascular ultrasound and DemTect at 30 days and 1 year. The DemTect is a generic dementia screening test, which consists of 5 subtests: a word list, a number transcoding task, a word fluency task, digit span reverse, and delayed recall of the word list. The transformed total score (maximum 18) is corrected for age and does not show any ceiling effect. 15 A parallel test version was used at 30 days to avoid retest effects. 16 In addition, a yearly telephone follow-up for 5 years after the operation was also prespecified and is still ongoing.

Outcomes
The primary composite outcome was the rate of any stroke or death of any cause up to 30 days after the operation (synchronous CEA and CABG versus isolated CABG) or 30 days after randomization for patients not receiving surgery (protocol violations). Stroke was clinically defined as any persistent focal or global neurological deficit lasting longer than 24 hours and presumed to be of no other than vascular origin. All perioperative stroke events were assessed by a study neurologist.
Secondary end points are listed in the Table I in the online-only Data Supplement. For the sake of comparability with previous research, we also included a post hoc secondary composite outcome event of any stroke, myocardial infarction, or death.
All suspected strokes, myocardial infarctions, and all deaths were adjudicated by an independent end point committee, which was blinded to treatment allocation.

Data Management
Data on paper case report forms were collected at the ZKSE in Essen. Good Clinical Practice compliant remote data capture (Oracle Clinical) was used for entering, managing, and validating the data from all centers.

Statistical Methods
We initially based the target sample size of 2 groups with 580 participants (total planned sample size 1160) on an estimate of 8.5% frequency of the primary outcome of stroke or death in the synchronous CEA and CABG arm, a clinically relevant 4.5% absolute reduction of this risk to 4.0% in the isolated arm, resulting in 84% power at a 2-sided level of significance of 0.05. The sample size of the first planned interim analysis was 550 (for details see 8 ). In accordance with the second amendment, the first interim analysis was performed after the first 100 patients, as recommended by the Data Safety Monitoring Board, and the final analysis was planned after inclusion of 300 patients. A generalized linear mixed-effects model that includes the fixed factors treatment group and the randomization factors age (<60 or ≥60 years), sex (male or female), modified Rankin Scale (score of 0-1 or 2-3), and a random (intercept) factor center was used to test and estimate the effect on the primary end point. This preplanned, confirmatory analysis of the primary end point was based on the Wald test statistic with corresponding 2-sided P value for the treatment effect and was applied to the intention-to-treat population. Because of the reduced sample size, we did not apply the originally planned group-sequential design with 1 interim analysis but instead performed 1 final confirmatory analysis at a significance level α of 5%. For robustness, we also ran sensitivity analyses using exact Monte Carlo estimation for χ 2 tests and analyses excluding major protocol violations (defined in the statistical analysis plan as not meeting any inclusion or exclusion criteria or not finishing the allocated therapy [including change of treatment group, an operation done by a noncertified surgeon, an end point event between randomization and surgical treatment]). For all other exploratory analyses, we also applied an exploratory significance level α of 5% for χ 2 , log-rank, and Wilcoxon-Mann-Whitney tests but focused more on point (mean, relative risk, and hazard ratio) and interval estimators of effects (95% confidence intervals). Details on the biometric analyses-especially those pertaining to the secondary end points-were defined in the

Stroke
October 2017 statistical analysis plan before data bank closure. Analyses were performed using SAS version 9.4.

Role of Funding Source
The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all data from the trial and had final responsibility for the results on submission for publication.

Results
Between December 2010 and December 2014, 129 patients were randomized in 17 centers with an average recruitment rate of 2.6 per year. Two patients withdrew consent between randomization and treatment and were excluded from the analysis. The intention-to-treat population consisted of 127 patients who were allocated to the 2 arms ( Figure 1). Baseline characteristics of both treatment groups are shown in Table 1. Two patients in the synchronous CEA and CABG group did receive neither CABG nor CEA and had no outcome events up to 1 year. One patient in the isolated CABG arm received aortic valve replacement instead of CABG and likewise had no outcome events up to 1 year. Two patients with perioperative ipsilateral ischemic events (1 stroke and 1 TIA) received CAS at 27 and 92 days after isolated CABG. Two patients in the isolated CABG group received CAS of asymptomatic carotid stenosis >30 days after CABG, and 3 patients received CEA of asymptomatic carotid stenosis >30 days after CABG with no subsequent outcome events. The rates of the primary composite end point of any stroke or death from any cause within 30 days after operation in the intention-to-treat population were 12/65 (18.5%) in the patients who underwent synchronous CEA and CABG and 6/62 (9.7%) in the patients who underwent isolated CABG (absolute risk reduction, 8.8%; 95% confidence interval, −3.2% to 20.8%; P WALD =0.12). The following major protocol violations resulted in exclusion from the per-protocol analysis: operation by a noncertified surgeon (n=11), myocardial infarction before CABG (n=1), informed consent not provided (n=1), no CABG performed (n=3), prior stenting of carotid stenosis to be treated according to the study protocol (n=1), and no high-grade carotid artery stenosis (n=2). In the per-protocol analysis, primary composite end point event rates at 30 days were 11 of 56 patients (19.6%) who received synchronous CEA and CABG and 6 of 53 patients (11.3%) who received isolated CABG (absolute risk reduction, 8.3%; 95% confidence interval, −5.1% to 21.8%; P WALD =0.21).
Similarly, there was no evidence for a significant treatmentgroup effect for all secondary end points at 30 days and 1 year ( Figure 2) although absolute event rates for isolated CABG were lower for most secondary end points. Sensitivity analyses, showing the same trend but no significant differences, in the per-protocol population are provided in the Table I in the  online-only Data Supplement. A complete clinical follow-up after 1 year could be obtained in 54 (83.1%) patients in the synchronous CEA and CABG arm and in 56 (90.3%) patients in the isolated CABG arm. Reasons for premature study termination by treatment group are displayed in Figure 1. All patients were on antithrombotic treatment, mostly aspirin, after 1 year and concomitant treatment was balanced among treatment groups (Table 2). Kaplan-Meier estimates of any stroke or vascular death-free survival after 1 year are shown in Figure 3. Pre-specified subgroup analyses as well as two post hoc analyses on the degree of ipsilateral and presence of contralateral carotid artery stenosis were in line with the main analysis (Table II online-only  Data Supplement).

Discussion
CABACS was conducted as a randomized trial to compare synchronous CEA of asymptomatic high-grade carotid artery stenosis versus no carotid operation in patients undergoing CABG surgery. Because of the low power of the trial following early termination, there was no evidence for a treatment-group effect although patients in the synchronous CEA and CABG arm had double the rate of stroke or death within 30 days or within 1 year compared with CABG without CEA. This observation was also found in predefined subgroups. All secondary end points were also more in favor of the isolated CABG arm but likewise failed to demonstrate a significant difference. Of note, we observed a >2-fold higher overall event rate compared with previously published data, which may result from the relatively high age of our study population, treatment quality, the systematic follow-up by study neurologists, or chance because of the relatively small study sample. 5,17,18 In contrast, only few patients showed postoperative worsening of cognitive functions as described previously, 19 but cognitive testing was not possible in all patients, particularly in those with stroke or severe physical impairment.
Our trial is the first multicenter randomized controlled trial with a rigorous design to investigate synchronous CEA versus no carotid operation in patients undergoing CABG. In addition, our trial provides data for perioperative events within 30 days and a 1-year follow-up including cognitive testing. The long-term 5-year follow-up is still ongoing. Two previous (bi-respectively monocentric) RCTs comparing synchronous CEA and CABG with delayed CEA after CABG suggested a lower perioperative risk of stroke in patients undergoing synchronous CEA and CABG compared with delayed CEA after CABG alone. 3,4 In these studies, however, the very low 30-day risk of stroke or death in the synchronous CEA and CABG arm (1% and 2.8%, respectively) is contradictory to a systematic review and large observational studies and therefore unlikely to represent routine clinical practice. 9,18,20 Conclusions on the safety of isolated CABG should be made with caution. Although a systematic review and meta-analysis suggested low complication rates in patients with unilateral asymptomatic 70% to 99% carotid artery stenosis undergoing CABG without CEA, 5 we found a considerably higher rate of perioperative stroke or death in the isolated CABG arm, which may

October 2017
raise concerns also about the isolated CABG approach. However, patient populations (and surgeons) may not have been comparable to other studies and therefore the only way to directly compare different operative strategies versus isolated CABG remains a headto-head randomized controlled trial. Compared with prior (staged) CEA, a synchronous CEA and CABG approach requires only 1 anesthesia and does not expose patients to the risk of myocardial infarction while waiting for CABG. On the contrary, overall stroke risk seems to be lower with prior staged than with simultaneous CEA and there seems to be no substantial difference in total vascular morbidity and mortality between the 2 approaches. 9,20 CAS has emerged as a possible alternative to CEA, but most clinical trials in patients with high-grade carotid artery stenosis have shown higher perioperative complication rates for CAS than for CEA, particularly in older men, which constitute the majority of patients undergoing CABG. [21][22][23] In a systematic review of cohort studies of predominantly asymptomatic patients with unilateral carotid disease undergoing staged CAS and CABG, the 30-day risk of any stroke or death was 9.1%. 24 More recent observational studies have even favored CAS before CABG. 20 Our trial has several limitations. The prespecified sample size was not reached because of withdrawal of funding after insufficient recruitment and therefore the study was underpowered to demonstrate statistically significant effects for the minimal, clinically relevant effect expected during the planning phase. 8 Moreover, investigators were not blinded to treatment allocation. However, main outcome events were adjudicated by blinded observers. Finally, few centers enrolled the large majority of patients, thus limiting the generalizability of our findings. Although all major German cardiovascular centers were invited and the majority participated in this trial, ≈90% of simultaneous CEA and CABG operations in Germany were performed outside of the trial. 17 A similar problem was encountered in the SPACE-2 trial, which was also stopped early because of slow enrolment. 27 Although all CABACS centers were required to keep screening logs, only a minority assessed all CABG patients for eligibility and thus bias by selective inclusion remains unknown. Reported reasons for the low study inclusion were surgery preferences of the referring physicians, lack of timely screening for carotid artery stenosis in patients scheduled for CABG, insufficient time to obtain informed consent for study participation, and symptomatic stenosis. Furthermore, an unknown number of patients with asymptomatic carotid artery stenosis scheduled for CABG received staged CEA or CAS before coronary bypass surgery but no registry data or complication rates are available for these patients.
In conclusion, although we cannot rule out a treatment effect, the very high rate of perioperative strokes does not seem to justify simultaneous CEA in patients with high-grade asymptomatic carotid artery stenosis undergoing CABG. Whether any carotid revascularization (CAS or CEA) is warranted in patients with unilateral asymptomatic high-grade carotid artery stenosis requiring CABG remains to be proven. Follow-up studies of the CABACS trial should test staged CAS or CEA followed by CABG versus isolated CABG and be performed in countries with a younger patient population and with reimbursement of staged or synchronous carotid revascularization only if a patient is treated within the trial.  To be allocated to trial (n = 1,160) To be analyzed (ITT) (n = 1,100)

Primary endpoint:
The primary efficacy endpoint is the event rate of nonfatal strokes or deaths from any cause (whatever occurs first) within 30 days after the intervention (either isolated CABG or synchronous CABG + CEA).

Secondary endpoints:
Secondary efficacy endpoints are single components of the primary endpoint, myocardial infarction, technical failures, duration of ventilatory support, change of cognitive performance on the Demtec scale and observations at different time points.

Statistical Analysis:
Efficacy: The null hypothesis of the primary endpoint is that the event rates for isolated CABG and synchronous CABG + CEA procedure are equal. This hypothesis will be tested with a generalized linear mixed effects model including covariates. The null hypothesis can be rejected if the p-value related to the Wald test statistic for the treatment effect is smaller than either α 1 = 0.0052 at the first planned interim analysis (550 patients randomized) or smaller than α 2 = 0.0480 at the final analysis of all data (O'Brien and Fleming group sequential plan). The total α is 0.05 (two-sided). The primary analysis will be performed on the intention-totreat analysis set, which includes all randomized patients. In addition, we will perform analyses on the per-protocol set to check the consistency of the claim. Safety: Safety analyses will be done for all adverse and serious adverse events reported and documented during the trial period. Details will be provided in the Statistical Analysis Plan (SAP).

Secondary endpoints:
All secondary analyses will be done exploratively, i.e., without adjustment for multiplicity. Details on the statistical techniques will be provided in the SAP.

Trial Duration and Dates:
Follow-up per patient: 1 year with an additional telephone follow-up over 4 years First patient in to last patient out: 8 years (3-year recruiting phase, 1-year followup; 4 years telephone follow-up) Total trial duration: 8 years Duration of treatment per patient: CABG with or without CEA is to be performed as soon as possible following randomization (see below). Best medical treatment continues throughout the follow-up period.  Flow Chart / Trial Schedule All patients scheduled for elective CABG surgery are routinely screened for carotid artery stenosis (CAS) in the participating centers. Neurological assessment before surgical intervention will further define study eligibility. Patients willing to participate as documented by written informed consent will be assigned to either treatment group by 1:1 block stratified randomization with variable block length and stratification factors (age, sex, and modified Rankin scale) for each center. Surgical intervention according to group allocation will be performed as soon as possible (but within less than 7 days) by a study-certified surgeon according to best practice. Adverse events occurring during or after surgery are continuously documented during the acute hospital stay.

Figure 1: Trial flow chart and visit plan
Patients will be followed-up by the study neurologist 7, 30 and 365 days after the procedure. In case of a suspected cerebrovascular event, cerebral imaging is considered clinical standard and therefore mandatory.

Scientific background
Annually, more than 60,000 coronary artery bypass graft (CABG) procedures are performed in Germany. 1 In patients without carotid artery stenosis (CAS) undergoing CABG, the perioperative risk of stroke or death is estimated at about 2-3%. 2, 3 Significant CAS is present in approximately 6-8% of all patients undergoing CABG and is associated with an increased risk of stroke during and after CABG. 4,5 Treatment in these patients (in Germany estimated at 4,000-5,000 per year) is handled controversially. Staged or synchronous carotid endarterectomy is advocated by many cardiothoracic surgeons to reduce the high perioperative and long-term stroke risk associated with polyvascular disease. However, no randomized trial has assessed whether a combined or staged procedure confers any overall benefit when compared with isolated CABG. While the perioperative risk of combined CABG and CEA remains high, the perioperative risk of isolated CABG estimated from single-center case series seems to be considerably lower. 6,7 Nevertheless, for a systematic review, only insufficient data from uncontrolled studies with variable inclusion criteria and endpoint assessment are available in patients with combined cardiac and carotid disease not undergoing staged or synchronous surgery. 6 The other point to consider in patients with asymptomatic CAS is the long-term risk of stroke of about 1-2% per year. 8 The long-term prognosis after CEA (without CABG) versus best medical treatment was investigated in the Asymptomatic Carotid Surgery Trial (ACST) and the smaller Asymptomatic Carotid Atherosclerosis Study (ACAS). 9,10 In a meta-analysis including 5,223 patients with asymptomatic CAS, patients undergoing CEA fared better than those treated medically for the outcome of perioperative stroke or death or any subsequent stroke over 5 years (relative risk 0.69; 95% CI 0.57 to 0.83). 8 Thus, there is a mild superiority of this intervention compared with standard medical therapy, provided that the CEA can be done with a perioperative morbidity and mortality of less than 3%. Even then, about 20 patients have to be treated to prevent one stroke over a period of 5 years, which also causes doubts about the cost effectiveness of this intervention. In addition, the majority of conservatively treated patients was included more than 10 years ago and was not treated according to current recommendations and guidelines. Therefore, two ongoing trials, "Transatlantic Asymptomatic Carotid Intervention Trial" (TACIT) and "Stent-Protected Angioplasty in asymptomatic Carotid artery stenosis" (SPACE-2), are currently re-evaluating therapy in patients with asymptomatic CAS. These trials will thus determine which perioperative morbidity and mortality rates are compatible with a long-term benefit of CEA in patients with asymptomatic CAS.

Evidence
CABG was introduced in 1968 and has become one of the most commonly performed major operations, with approximately 800,000 patients each year undergoing this procedure worldwide. CABG provides definite clinical advantage in patients with coronary artery disease, resulting in the relief of angina and an improved quality of life.
Advances in coronary surgery (e.g. off-pump CABG, smaller incisions, enhanced myocardial preservation, use of arterial conduits, and improved postoperative care) have reduced morbidity, mortality, and rates of graft occlusion. 11 Percutaneous coronary intervention (PCI) was introduced in 1977 and is an alternative definite option for treating patients with coronary artery disease. In the past few years, several randomized clinical trials 12 and observational studies 13 have examined outcomes for these interventions, but most, with few exceptions, were conducted before the availability of stenting. New stent devices and improvements in PCI techniques, however, have revolutionized the latter approach and challenge CABG in the treatment of coronary three-vessel disease and/or left main stem disease. Observational data from very large New York cardiac registries comparing short-term and long-term outcomes among patients with multivessel disease who underwent CABG (n=37,312) or stenting (n=22,102) showed that for patients with two or more diseased coronary arteries, CABG is associated with higher adjusted rates of long-term survival and lower rates for repeat revascularization than stenting. 14 In these days, PCI involving drug-eluting stents is increasingly used to treat complex coronary artery disease. In the Synergy between PCI with Taxus and Cardiac Surgery (Syntax) trial, the optimal revascularization strategy for patients with previously untreated three-vessel or left main coronary artery disease is being assessed. 15 Recently, the two-year results of this largest comparative randomized trial became available. These data demonstrated comparably good results in patients with less complex coronary artery disease between PCI and CABG, while patients with more complex coronary pathology had better results by surgical intervention. Hence, in these patients, CABG is regarded as the primary recommendation of revascularization. No ongoing or completed randomized trial has assessed whether a combined synchronous or staged CEA and CABG procedure confers any benefit compared with isolated CABG. 16 A recent systematic review of perioperative outcomes following synchronous CEA and CABG revealed a complication rate of 8.2% (95% CI 7.1-9.3%) for perioperative stroke or death in 5,386 patients. 6 Only limited data are available on other strategies: In 649 predominantly asymptomatic patients undergoing staged carotid artery stenting and CABG, the 30-day risk of any stroke or death was 9.4% (95% CI 7.0-11.8%), 17 whereas it was 6.1% (95% CI 2.9-9.3%) in 709 similar patients undergoing staged CEA and CABG. 6 However, when including myocardial infarction as an endpoint, the combined outcome rate for staged CEA and CABG rises to 10.2% (95% CI 7.4-13.1%). Another registry study that was not included in this review showed an in-patient death / stroke rate of 8.6% after synchronous CEA and CABG in 26,197 patients treated between 2000 and 2004. 18 Because screening for perioperative stroke is not routinely done by neurologists, the stroke rate is likely to be higher than the reported 3.9%. Similarly, propensity-score-matched studies comparing combined CABG and CEA and CABG without prophylactic CEA have reported conflicting results regarding overall morbidity and in-hospital mortality. 19,20 Because synchronous CEA and CABG procedures still constitute a frequently used approach in many German centers with a well documented periprocedural risk, it was chosen as comparator for isolated CABG and in the absence of a gold standard was defined as the control group.
Given an estimated annual frequency of about 4,000-5,000 CABG procedures in patients with significant CAS and the commitment of the 40 leading cardiothoracic surgery centers in Germany, recruitment of 10-20 patients annually per center seems feasible to meet the study size of 1,160 patients over 3 years.

Role of isolated CABG in patients with asymptomatic CAS
Given a prevalence of about 6-8% CAS in patients undergoing CABG surgery and a frequency of about 2% simultaneous operations in administrative databanks on CABG, a considerable number of CABG operations are performed without CEA in patients with asymptomatic CAS. 4 Nevertheless, data on safety and long-term outcome in these patients are very scarce. While no randomized studies have been performed in patients with asymptomatic CAS, the perioperative risk of CABG without CEA or carotid stenting derived from single-center case series was not increased in comparison with CABG in patients without CAS in retrospective case series. 6,7,21 In contrast, the only prospective observational study in patients with (symptomatic and asymptomatic) CAS found a higher stroke rate in patients undergoing isolated CABG compared with combined CABG and CEA but lower mortality and lower rate of myocardial infarction. 22

Role of combined CABG and carotid endarterectomy
During the synchronous operation, CEA is usually performed before CABG. Pooled data from single-center observational studies showed a perioperative rate of stroke or death of 8.2% (95% CI 7.1-9.3%), 6 whereas a nationwide US registry reported a risk of 8.6% for stroke or death after synchronous CABG / CEA in 26,197 patients treated between 2000 and 2004. 18 It remains unclear if this increased risk is due to the high vascular comorbidity or the combined CABG / CEA procedure. In a propensity score analysis, Ricotta et al. could not find an increased risk for the combined operation in the New York State Cardiac Data Base, 20 whereas risk adjustment in the nationwide inpatient sample revealed a 38% increased rate of complications in patients undergoing combined CABG / CEA. 23 Although the long-term benefit of CEA for asymptomatic CAS has been established, ongoing studies will have to demonstrate which perioperative complication rate is compatible with a long-term benefit of CEA (or carotid stenting) compared to modern medical treatment.

Role of medical management
Medical management of patients with asymptomatic carotid stenosis consists of risk factor modification, statin therapy, antiplatelet medication and careful monitoring of progression or hemodynamic relevance by ultrasound. These treatment regimens have been the standard therapy for asymptomatic carotid stenosis in the past and many treatment guidelines still recommend only medical therapy if surgical treatment cannot be provided with perioperative morbidity (stroke, MI) and mortality rates below 3%. 24

Trial rationale / justification
There is an ongoing debate as to which strategy is optimal with regard to the incidence of complications and long-term outcomes after CABG surgery. 3,25 In the absence of any randomized controlled trials, no systematic evidence exists that staged or synchronous operations confer any benefit over CABG without CEA. Conversely, it remains questionable whether the observed risk of the synchronous procedure is justified in any asymptomatic patient with high-grade carotid stenosis. 6 Thus, any potential (long-term) benefit conferred by prophylactic carotid intervention (CEA or stenting) may be offset by _______________________________________________________________________________________________________ Page 17 the relatively high procedural risks observed in systematic reviews. Evidence for isolated CABG, the least invasive and least expensive strategy, is limited to self-reported, uncontrolled, mostly retrospective case series. Only a randomized controlled trial will finally settle this open question and is expected to provide a base for defining an evidence-based standard. It will also enable development and evaluation of improved surgical techniques or strategies and thus have a wide impact on management of this disease. The fact that 40 cardiothoracic centers in cooperation with neurologists in Germany have committed to perform this study is further proof that a randomized trial is desperately needed.
Every individual in the trial will benefit from treatment by certified experts and the supervision of the medical treatment and risk factor management by neurologists. With an estimated 4,000-5,000 CABG procedures performed in patients with significant carotid artery stenosis in Germany, treatment of this condition is frequent. Estimated proceeds of 3,900 € (G-DRG F06A versus F06B) for an additional synchronous or staged CEA / carotid artery stenting performed in roughly 2,000-2,500 patients thus results in an annual burden of around 8-10 million €. This demonstrates that apart from preventing morbidity / mortality, there is a high economic interest in clarifying the optimal treatment strategy.

Benefit / risk assessment
The benefit of each of the two treatment options is the proposed reduction of perioperative and long-term cerebro-and cardiovascular events. However, each treatment option also carries specific risks. Most feared risks of both isolated CABG and combined CABG/CEA are periprocedural stroke and death. Differences between the risks of isolated CABG and combined CABG/CEA are based on technical details. CEA is usually performed before CABG, which contributes to the complication rate related to the combined procedure but possibly lowers the perioperative stroke risk during CABG surgery. In contrast, isolated CABG may put patients at risk of hemodynamic or embolic stroke distal to the CAS. In addition, these patients carry a continued long-term risk of stroke of approximately 1-2% per year. Thus, isolated CABG is expected to go along with possibly lower 30-day complication rates but possibly higher long-term ipsilateral stroke rates. The reverse is to be expected with the combined procedure. To minimize the periprocedural risks, every individual in the trial will be treated by named experts and a close monitoring for such events will be done. A quality subcommittee defines the rules for the treatment and the quality control criteria (see 10.4).

General aim / primary objective
The objective of this study is to compare safety and efficacy of isolated CABG with combined synchronous CABG and CEA in individuals with asymptomatic atherosclerotic carotid artery stenosis. _______________________________________________________________________________________________________ Page 18

Composite primary efficacy endpoint
The primary efficacy endpoint is the event rate of nonfatal strokes or deaths from any cause (whatever occurs first) within 30 days after the intervention (either isolated CABG or synchronous CABG + CEA). The primary effect expected from isolated or combined surgery is to reduce the rate of myocardial infarction. The clinically relevant figure, however, is the sum of periprocedural complications and ipsilateral stroke during a longer period. Isolated CABG is expected to go along with lower 30-day complication rates but higher long-term ipsilateral ischemia rates. The reverse is to be expected from the combined CABG / CEA procedure. The primary efficacy endpoint takes into account only the short-term effects. In contrast, ongoing studies will have to demonstrate which perioperative complication rate is compatible with a long-term benefit of CEA compared to modern medical treatment.

Assessment of safety
Safety is assessed as the rate of any nonfatal stroke or death from any cause within 30 days after CABG. This safety endpoint has been used in many trials concerned with treatment of carotid stenoses. 9,26,27 In addition, there will be a special focus on complications and serious adverse events within 30 days after CABG. The etiology of death within 30 days after CABG therefore should be verified by autopsy, if possible.

Secondary outcomes
Secondary efficacy endpoints are single components of the primary endpoint, myocardial infarction, technical failures, duration of ventilatory support, change of cognitive performance on the Demtec scale and observations at different time points. These are described in detail in chapter 6.3.

TRIAL DESIGN
CABACS is a randomized, controlled, open, multi-center group sequential trial with 2 parallel arms and blinded observers. All patients are treated with a best medical treatment (BMT) regimen tailored to their individual risk factor profile consisting in treatment of risk factors, lipid-lowering and anti-platelet medication. Rules for the BMT will be formulated by the BMT subcommittee (see Investigator site file). The overall duration of the trial is expected to be approximately 8 years. Results of the primary endpoints are expected after 4.5 years. Recruitment of subjects will start in October 2010. The actual overall duration of recruitment may vary. Patients will be allocated in a concealed way by central randomization before surgery. A clinical report of each outcome event will be prepared by the ZKSE (unblinded part), which removes all information about the allocated treatment. Subsequently, this report will be presented to the three independent members of the clinical adjudication committee who have not been involved in the treatment of the patient so far and who classify the event as blinded adjudicators / blinded observers. Each adjudicator will send his / her classification to the ZKSE. If the three classifications differ, the outcome event will be discussed by all members of the clinical adjudication committee considering the majority vote.

Number of subjects
As calculated in section 9.1 'Sample Size Calculation', 1,160 subjects should be enrolled in the clinical trial, 580 to the isolated CABG-arm and 580 to the combined CABG / CEA arm. Recruitment and treatment of subjects should be performed in around 35 -40 trial centers. The minimum / maximum number of subjects per trial centre should be 10 / 100.

General criteria for subjects' selection
In daily clinical practice most asymptomatic carotid stenoses are discovered by extracranial ultrasound techniques performed for screening purposes in patients with vascular disease in other vascular territories. Only few stenoses are diagnosed by extracranial MRA or CT-angiography. Due to the ultrasound-based screening procedure used in CABACS, the study population should closely resemble the general population with asymptomatic carotid stenoses undergoing CABG.

Inclusion criteria
Subjects meeting all of the following criteria will be considered for admission to the trial:  Asymptomatic (past 180 days) stenosis ≥ 80% (following criteria of the ECST 28 ) of the extracranial carotid artery in patients scheduled for CABG  Negative pregnancy test in pre-menopausal women  Written informed consent and full legal capacity  Carotid stenosis treatable with CEA  Ability of the patient to participate in follow-up examinations

Exclusion criteria
Subjects presenting with any of the following criteria will not be included in the trial:  Non-atherosclerotic stenosis (e.g. dissection, floating thrombus, fibromuscular dysplasia, tumor)  Complete occlusion of the carotid artery to be treated  Prior stenting of the carotid artery to be treated  Stenosis following radiotherapy  Additional higher grade intracranial or intrathoracic stenosis (tandem stenosis)  Recent (past 180 days) ischemic symptoms ipsilateral to carotid stenosis or occlusion  Contralateral carotid occlusion or other known indication for carotid revascularization (apart from scheduled CABG) _______________________________________________________________________________________________________ Page 20  Myocardial infarction (NSTEMI or STEMI) within the past 7 days or hemodynamically unstable patients  Known high risk for cardiogenic embolism requiring anticoagulation (mechanical heart valve, chronic atrial fibrillation, left ventricular thrombus, left ventricular aneurysm)  Evidence for intracranial bleeding within the past 90 days  Modified Rankin Scale score >3 or severe aphasia  Patients unlikely to survive more than 1 year due to concomitant diseases  Planned combined cardiac valve replacement or any other cardiac surgery beyond CABG (+/-CEA) during the procedure  Major surgery (apart from study procedures) planned within 8 weeks from randomization  Participation in another clinical trial No patient will be allowed to enroll in this trial more than once.

Withdrawal of subjects
A patient may be withdrawn from all trial related procedures (excluding follow-up visits and central follow-up by telephone) for the following reasons:  at his/her own request or at request of his legal representative  non-adherence to the trial-related requirements which may (have) influence(d) the validity of the trial data  if, in the investigator's opinion, continuation of the trial would be detrimental to the patient's well-being, e.g. if a patient was randomized into the isolated CABG arm and suffers ipsilateral neurological symptoms prior to CABG, patient treatment should be based on current guidelines  occurrence of exclusion criteria prior to CABG ± CEA.  if, in the investigator's opinion, protocol violations caused by the patient would lead to invalid data (e.g. non-compliance with planned study procedures) The Executive Committee (see chapter 10.2) decides about withdrawal of subjects from trial treatment in case of occurrence of criteria mentioned above. In all cases, the reason for withdrawal must be recorded in the CRF and in the patient's medical records. In case of withdrawal of a patient at his / her own request, the reason should be asked for as extensively as possible and documented, even though the patient is entitled not to answer. All efforts will be made to follow the patient and all examinations scheduled for the final trial day will be performed as far as possible and documented for all subjects. All ongoing Complications / Serious Adverse Events (SAEs) of withdrawn subjects have to be followed-up until no more signs and symptoms can be verified and the patient is in stable condition.

Replacement of Subjects
Subjects will not be replaced. _______________________________________________________________________________________________________ Page 21

Premature Closure of the Clinical Trial
If the 30-day complication rate (any stroke or death) of one or both treatments exceeds 10% (both point estimator and lower limit of 80% confidence interval are larger than 10%), the DSMB can give a recommendation to the steering committee to stop the trial. If the 30-day complication rate exceeds 15%, the DSMB should recommend to stop the trial early. The trial can be prematurely closed or suspended by the Steering Committee in case of new data about the risk-benefit of one of the treatment regimes becoming available, if the p-value for the treatment effect is smaller than α 1 = 0.0052 at the first planned interim analysis, or if the DSMB recommends study closure. The Ethics Committee (EC) must then be informed. Furthermore, the Ethics Committee(s) themselves may decide to stop or suspend the trial. Should the trial be closed prematurely, all trial material (completed, partially completed, and blank CRF, etc.) must be returned to the Center for Clinical Trials (ZKSE; Essen) All involved investigators have to be informed immediately about a cessation /suspension of the trial. The decision is binding to all trial centers and investigators.

Description of trial days
Study visits are scheduled as follows:  Screening  Randomization (Screening and randomization can be done on the same day)  Surgical treatment (day 0), as soon as possible (max. 7 days) after randomization  Day 7  1 after CABG  30  3 days after CABG (or randomization if surgical treatment was cancelled)  1 year  14 days after randomization  2 years  30 days after randomization (telephone follow-up)  3 years  30 days after randomization (telephone follow-up)  4 years  30 days after randomization (telephone follow-up)  5 years  30 days after randomization (telephone follow-up) An overview about the procedures to be done at specific visits can be found on page 9.

Screening and randomization
After initiation of the center, all patients will be consecutively screened and all eligible patients who are willing to participate will be included in the trial. To qualify for this trial, patients must have met all of the above described inclusion and none of the exclusion criteria. Screening and randomization can be done on the same day. CABG ± CEA must be done as soon as possible (max. 7 days) after randomization. Events occurring between randomization and operation are counted within the allocated treatment group. Events occurring between screening and randomization are not counted. In order to achieve comparable groups, patients will be allocated in a concealed way by central randomization ≥1 day before surgery. To address "concealment of allocation", the randomization will be done centrally by the Center for Clinical Trials, Essen ("Zentrum für Klinische Studien Essen", ZKSE; www.zkse.de) and each patient who is registered and randomized will be part of the ITT analysis set. The randomization list will be kept in safe and confidential custody at the ZKSE. Before randomization, the following procedures (5.2.1-5.2.6) have to be done:

Informed consent
The patient must be able to understand the nature of the trial and the related procedures and sign and date the informed consent form in person. Informed consent must be available before any study related procedures.

Neurological examination
Modified Rankin Scale (mRS) and NIH-Stroke Scale (NIHSS) are determined at screening as the basis for the follow-up-examinations. A mRS > 3 at baseline is an exclusion criterion. All neurologists should have a NIHSS certificate (http://asa.trainingcampus.net/uas/modules/trees/windex.aspx).

Prior and concomitant illnesses
Relevant additional illnesses present at the time of informed consent are regarded as concomitant illnesses and will be documented on the appropriate pages of the case report form (CRF). Patients with recent (last 180 days) ischemic symptoms ipsilateral to the territory of the vessel to be treated are not allowed to be included into the trial.

Prior and concomitant treatments
Relevant concomitant therapies (e.g. antiplatelets, antihypertensives, lipid-loweringdrugs) will be documented preoperatively, at 30 days, 1 year and thereafter yearly up to five years after CABG. If another major surgery (apart from CABG ± CEA) is planned during the first 8 weeks after randomization patients should not be included prior to this operation.

Ultrasound examination
The ultrasound examination including color-coded extracranial Duplex and examination of the intracranial circulation should be done by an experienced examiner (preferably with either DGKN or DEGUM ultrasound certificate). Documentation contains at least grading of the stenosis, plaque morphology and a measurement of the Intima-Media-Thickness in the common carotid arteries. Stenosis severity as measured with ultrasound is usually expressed following the method of the ECST. If a conversion from a NASCET measurement is necessary, the following equation will be used: %NASCET _______________________________________________________________________________________________________ Page 23 = (%ECST -43) x (100 / 57). 29 Grading of stenosis by CT angiography or MR angiography needs to be confirmed by ultrasound examination.

Allocation to treatment / Randomization
Eligibility of a patient will be determined by a study neurologist. A study surgeon needs to confirm that treatment is feasible. CABG with or without CEA must be performed as soon as possible (max. within 7 days) after randomization. In the case of a bilateral stenosis, randomization is performed only once, usually for the side with the higher grade stenosis. For this open trial, a centralized randomization will be set up. All eligible patients will be registered and randomized to achieve balanced prognostic factor distributions for the factors age (< 60 years or ≥ 60 years), sex (male or female), modified Rankin Score (0-1 or 2-3) and the factor centre.
The completed CRF Part Randomization is to be sent by fax to the

Center for Clinical Trials Essen (ZKSE), University of Duisburg-Essen
Hufelandstr. 55, 45147 Essen Fax: 0201 723 5933, Phone: 0201 723 4134 Documentation of all potentially eligible patients in a screening log is mandatory including reasons for non-inclusion into the trial. Inclusion of all patients meeting all inclusion criteria and none of the exclusion criteria is mandatory. Non-compliance with this rule may be a reason for exclusion of a site. For further details, e.g. concealment of allocation see 5.2.

Best medical treatment
All patients will be treated with up-to-date medication following national and international guidelines. Recommendations for this treatment will be given by the 'Best Medical Treatment' subcommittee (see TMF and investigator site file).

Isolated CABG without carotid endarterectomy
Standards for surgical treatment are formulated by the Surgical quality subcommittee (see TMF and investigator site file)

Combined CABG and carotid endarterectomy
Standards for surgical treatment are formulated by the Surgical quality subcommittee (see TMF and investigator site file)

Blinding
Due to the nature of the treatments involved, blinded treatment is not possible. On ultrasound upon follow-up, it is also obvious how the patient was treated. Therefore, the _______________________________________________________________________________________________________ Page 24 study has to follow an open design. To minimize a potential bias, the neurologist who is not directly involved in the operation is responsible for the follow-up examinations and end-point assessment. For endpoint adjudication, no information on treatment group will be made available (blinded observers).

Follow-up examinations
The duration of the trial for each patient is 5 years. After the periprocedural phase and a follow-up visit after one year, yearly telephone follow-ups are scheduled. Each visit consists of a neurological examination, an event screening, a risk factor screening, and an ultrasound examination. Telephone follow-ups consist of an event screening only. Events are screened using a standardized questionnaire. An event is suspected if one of the following questions has been answered positively:  "Ist (seit der letzten Untersuchung) eine (vorübergehende) Sehstörung auf einem Auge aufgetreten?"  "Ist (seit der letzten Untersuchung) eine Schwäche oder Empfindungsstörung auf einer Körperseite aufgetreten?"  "Ist (seit der letzten Untersuchung) eine Sprachstörung aufgetreten?"  "Haben sich (seit der letzten Untersuchung) neue Beschwerden entwickelt?" If a cerebrovascular event is suspected, appropriate information from the treating physician should be obtained. During the first year, an appropriate neuroimaging should be performed in case of a suspected cerebrovascular event to distinguish between possible types of stroke. In case of an ipsilateral cerebrovascular event in a patient randomized into the isolated CABG arm, the patient should be considered for CEA or carotid artery stenting following current guidelines for treatment of symptomatic stenosis. Risk factor screening, assessment of medication, and ultrasound examinations have to be done as described above (section 5.2) In addition, the mRS and NIHSS have to be assessed and documented after 30 days and 1 year.

Plan for medical treatment and care after the trial
Medical treatment with risk factor modification will be continued after the end of the follow-up period (5 years in each individual). The general practitioner of the patient will be responsible for control of these risk factors and antiplatelet therapy.

Definitions
Ischemic Stroke: Defined by at least one of the following criteria:  Sudden onset of focal neurological symptoms lasting more than 24 hours, with no apparent cause other than cerebral ischaemia.  Sudden onset of focal neurological symptoms with the presence of cerebral infarction in the appropriate territory on brain imaging (CT or MRI), regardless of the duration of symptoms (less than or more than 24 hours). Hemorrhagic Stroke: Sudden onset of focal neurological symptoms with the presence of cerebral haemorrhage in the appropriate territory on brain imaging (CT or MRI), regardless of the duration of symptoms (less than or more than 24 hours) and regardless of the cause of the haemorrhage (spontaneous or secondary to trauma, tumour or another cause).
Unknown type of stroke: In some cases, the type of stroke cannot be determined with certainty, particularly when brain imaging or an autopsy report or an operation report cannot be obtained, despite all efforts. This type of event will be classified as unknown type of stroke, if the symptoms last more than 24 hours. Ipsilateral stroke: Stroke within the territory of the treated vessel or higher grade stenotic carotid artery at screening. Disabling stroke: Stroke leading to a disability of at least 4 on the modified Rankin scale at day 30 after symptom onset. Myocardial infarction: Any one of the following criteria satisfies the diagnosis for an acute MI 30 1. Detection of rise and/or fall of cardiac biomarkers (preferably troponin) with at least one value above the 99 th percentile of the upper reference limit (URL) together with evidence of myocardial ischemia with at least one of the following  Symptoms of ischemia  ECG changes indicative of new ischemia (new ST-T changes or new left bundle branch block [LBBB])  Development of new pathological Q waves on the ECG  Imaging evidence of a new loss of viable myocardium or new regional wall motion abnormality 2. Pathological findings of an acute MI. 3. In patients who have undergone CABG (< 72h) with normal baseline troponin values: Elevation of biomarkers greater than 5 times the 99 th percentile URL together with either  New pathological Q waves  New left bundle branch block  Angiographically documented new graft or native coronary artery occlusion  Imaging evidence of a new loss of viable myocardium Vascular death:  death related to a cardiac or vascular cause.  death due to haemorrhage.  death due to pulmonary embolism.  sudden death: death occurring in less than 24 hours, unexpectedly in a subject in apparent good health and whose condition was stable or was improving.  death with no documented non-vascular cause.  fatal stroke: death occurring within 30 days of a stroke (ischaemic or haemorrhagic). Non-vascular death: Death due to a documented non-vascular cause (infection, cancer, accident, suicide, etc.). Technical failure: Inability to treat the stenosis with CEA or residual stenosis after CEA of at least 70% or occlusion following ultrasound-criteria at day 30.
Restenosis: Recurrent stenosis of at least 70% or occlusion following ultrasound-criteria. Observation period: 30 day endpoints for patients randomized into either group are evaluated after 30±3 days after the intervention (either isolated CABG and combined CABG + CEA). The long-term (up to 5 years) observational period starts after the intervention (either isolated CABG or combined CABG + CEA), annual assessment can be done in a time period ±14 days around the calculated date (±30 days from the second year on).

Evaluation of safety
Safety is assessed as the rate of any stroke within 30 days of treatment or death from any cause within 30 days after CABG+/-CEA.

Primary endpoint
The primary efficacy endpoint is the event rate of nonfatal strokes or deaths from any cause (whatever occurs first) within 30 days after the intervention (either isolated CABG or synchronous CABG + CEA).

Secondary endpoints
 Number of ischemic strokes ipsilateral to the initially higher grade, not occluded stenotic carotid artery within 30 days and 1 year  Any stroke or vascular death within 30 days, 1 year and 5 years  Deaths from any cause within 30 days, 1 year and 5 years  Number of disabling strokes (definition: stroke with resulting impairment >3 on the modified Rankin Scale) within 30 days and 1 year  Change of cognitive performance on the Demtec scale from randomization to 30 days and 1 year  Technical failure of intervention  Number of myocardial infarctions within 30 days, and from 30 days to 1 year  Duration of ventilatory support after operation (CABG ± CEA)

Additional analysis
 Total length of hospital stay and G-DRG for acute hospital stay.  Total length of ICU stay

Endpoint evaluation
All suspected strokes, as well as all myocardial infarctions, and all deaths will be evaluated by a committee of experts blinded to treatment, who will determine the final classification of these events. This committee, named 'clinical adjudication committee' is described in section 10.3. potential relationship to surgery (CABG +/-CEA). This change will be documented by an investigator. Surgical procedures themselves are not complications; they are therapeutic measures for conditions that require surgery. The condition for which the surgery is required may be a complication. Planned surgical measures permitted by the clinical trial protocol and the condition(s) leading to these measures are no complications, if the condition leading to the measures was present prior to inclusion into the trial.

Serious adverse event
All serious adverse events (irrespective of a possible causal relationship to the surgery) will be documented during the first 30 days after surgery only. A serious adverse event (SAE) is one that:  results in death  is life-threatening (the term life-threatening refers to an event in which the patient was at risk of death at the time of event and not to an event which hypothetically might have caused death if it was more severe)  requires patient hospitalization or prolongation of existing hospitalization  results in persistent or significant disability / incapacity or  is a congenital anomaly / birth defect.  requires ventilatory support beyond 30 days after CABG +/-CEA (if not documented already as SAE (e.g. prolongation of hospitalization))

Period of observation and documentation
Only surgical complications reported by the patient or detected by the investigator will be collected during the trial and must be documented on the appropriate pages of the CRF / BQS forms. Complications must also be documented in the patient's medical records.
In this trial, all complications that occur during 30 days after the operation (CABG ± CEA) will be documented on the pages provided in the CRF / BQS forms.

Reporting of SAEs / OE by investigator
All serious adverse events or outcome events must be reported by the investigator to the ZKSE within 7 days using the 'Serious Adverse Event / Outcome event' form. Forms are in the investigator site file. The report must be as complete as possible.

Center for Clinical Trials Essen (ZKSE), CABACS Trial
A report of an outcome event should include details concerning the type of stroke (ischemic vs. hemorrhagic, territory of stroke, and severity of stroke) or cause of death. All OEs are also evaluated by the CAC.

Expedited reporting
According to the Declaration of Helsinki every investigator should report all SAE to his Ethics Committee. In this trial with up to 40 sites, this reporting is counterproductive. A central reporting will improve the safety of all trial patients. With the Ethics Committee in Essen the following agreement was concluded: A report (analogous to the annual safety report) will be sent to the independent Ethics Committee (University of Duisburg-Essen) and the sites. This report includes all SAE and OE. This report will be sent every 3 months (4 times a year). SAEs resulting in death are to be reported separately to the Ethics Committee (University of Duisburg-Essen) within 14 days. These assessments will be done by the Trial coordinator Dr. med. Stephan Knipp, Dept. of Thoracic and Cardiovascular Surgery; University of Duisburg-Essen.

Data collection
The sponsor will provide blank CRF to the sites. Each CRF page has a unique identifier. This is the patient number. All findings will be documented in the patient's medical record and in the CRF. The investigator is responsible for ensuring that all sections of the CRF are completed correctly and that entries can be verified against source data. Any errors should have a single line drawn through them so that the original entry remains legible and the correct data should be entered at the side with the investigator's signature, date and reason for change. Self-explanatory corrections need not to be justified. The correctness of entries in the CRF will be confirmed by dated signature of the responsible investigator. After finalization of the visit after 30 days and 1 year, the original CRF will be copied for local safekeeping and the originals are transferred once a month to the data management of ZKSE.

Center for Clinical Trials Essen (ZKSE) c/o IMIBE CABACS Trial, Frau Anja Marr University of Duisburg-Essen, Hufelandstr. 55, 45147 Essen
The copies of the CRF will be kept by the investigator.

Data handling
Data handling is described in detail in the Data Management Plan and the Data Validation Plan.
To ensure data quality, a double data entry will be done. After completion of data entry, checks for plausibility, consistency, and completeness of the data will be performed. Based on these checks, queries will be produced combined with the queries generated by visual control. A responsible investigator will be obliged either to correct the implausible data or to confirm its authenticity and to give appropriate explanation. Further checks for plausibility, consistency, and completeness of the data will be performed after completion of the study. Queries will be generated on the basis of these checks combined with a visual control by a responsible monitor/data manager. All missing data or inconsistencies will be reported back to the sites and clarified by the responsible investigator. If no further corrections are to be made in the database, it will be declared closed and used for statistical analysis. All data management activities will be done according to the current Standard Operating Procedures (SOPs) of the ZKSE. To ensure the safety of participants in the study, and to ensure accurate, complete, and reliable data, the investigator must keep records of laboratory tests, clinical notes, and patient medical records in the patient files as original source documents for the study.

Storage and archiving of data
All important trial documents (e.g. original CRF) will be archived for at least 10 years after the end of the trial. The investigator(s) will archive all trial data (source data and Investigator Site File (ISF) including patient identification list, signed informed consent forms and relevant correspondence) according to the section 4.9 of the ICH Consolidated Guideline on GCP (E6).

Sample size calculation
Sample size calculations were performed for the primary efficacy endpoint. The sample size calculation is based on using the Brownian motion approximation to a 2x2 Χ 2 test (two-sided analysis) using the classical boundaries of O' Brien and Fleming (1979). Note that the maximum sample size of the group sequential plan is slightly larger than the sample size for the fixed sample size case which can e.g. be performed for the Χ 2 test using nQuery Advisor ® 6.0. The average sample size of the sequential plan is smaller than the sample size for the fixed sample size case. The prior assumptions are a frequency of 8.5% in the control group (synchronous CABG and carotid endarterectomy), and 4% in the experimental group (CABG without carotid endarterectomy) based on data from the literature. 7,17 To detect this difference with a total α=5%, a maximum sample size of n = 2 x 550 = 1,100 patients with complete information for the primary endpoint will lead to a power of 84%. Note that intra-surgeon correlation of the primary outcome may lead to an underestimation of the statistical power which is why a power >80% was chosen. The table below shows the impact on the power for five alternative scenarios (rates differing from those expected but ignoring intra-surgeon correlation) for n = 1,100 patients. In order to address a potential drop out rate of 5% overall (based on empirical data), another 60 patients have to be randomized. Thus, the total maximum sample size is n = 2 x 580 = 1,160 patients that have to be allocated to the treatment arms (scenario 1). With this sample size, the power of all scenarios is larger than 65% and except for scenario 3 also larger than 80%.

Definition of trial population to be analyzed
The primary confirmatory analysis will be performed on the intention-to-treat (ITT) population which includes all patients randomized and treated (full analysis set). In addition, a per-protocol (PP) analysis will also be performed after exclusion of patients who did not undergo the allocated operation, or other serious protocol violations. The following are considered major protocol deviations:  Not meeting one or more of the Inclusion Criteria and/or falling into one or more of the Exclusion Criteria  Not finishing the allocated therapy (change of treatment group)  Operation done by a non-certified interventionalist or using noncertified material  Endpoint event between randomization and surgical treatment In case of problems with the decision if a protocol violation is major, the blinded Clinical Adjudication Committee is responsible for the allocation.

Safety and Efficacy:
The analysis of the primary outcome will be a comparison of the experimental group (CABG without CEA; π experimental ) with the control group (synchronous CABG and CEA; π control ) with regard to differences in first nonfatal stroke or death within 30 days after CABG/CEA. In particular, the null (H 0 ) and the alternative hypotheses (H 1 ) are: H 0 : π experimental = π control H 1 : π experimental # π control where π · denotes the event rate for each group. Note that we consider a two-sided test with α=0.05 as both superiority of the experimental or superiority of the control group are of interest. The confirmatory analysis on the intention-to-treat analysis set (see above) will be performed by a generalized linear mixed effects model 31 including the prognostic factors of the randomization as covariates (fixed factors: age: < 60 years or ≥ 60 years; sex: male or female; modified Rankin Score: 0-1 or 2-3; random factor: centre). The null hypothesis can be rejected if the p-value related to the Wald test statistic for the treatment effect is smaller than either α 1 = 0.0052 at the first planned interim analysis (after 550 patients are randomized and followed-up for 30 days) or smaller than α 2 = 0.0480 at the final analysis of all data (O'Brien and Fleming group sequential plan) such that the total α is 0.05 (two-sided). In addition to the sequential multiple testing, no other multiple testing issues arise for the confirmatory analysis. Nonetheless, we will perform sensitivity analysis of the primary outcome. These include per-protocol analyses, subgroup analyses stratified by gender and worst/best case scenario analyses in case of missing data of the primary efficacy endpoint. Details on the biometric analyses will be defined in the statistical analysis plan which has to be authorized by the biometrician, the sponsor, and the LKP.

Interim analyses
The trial is designed as a group sequential plan with one planned interim analysis. The null hypothesis can be rejected if the p-value related to the Wald test statistic for the treatment effect is smaller than α 1 = 0.0052 at the first planned interim analysis (after 550 patients are randomized and followed-up for 30 days; O'Brien and Fleming group sequential plan).

STUDY ADMINISTRATION
All persons responsible for the study are listed in the investigator site file including name, address, phone and email. In order to monitor specific aspects of the current trial, the following Reference Committees will be established: Data Safety and Monitoring Board (DSMB), Steering Committee (SC), Executive Committee (ExC) and a Clinical Adjudication Committee (CAC). The work of these committees will be based on the 'Guideline on Data Monitoring Committees' EMEA/CHMP/EWP/5872/03

Data Safety and Monitoring Board (DSMB)
An independent Data Safety and Monitoring Board (DSMB) has been assembled. The DSMB is composed of independent experts in the field of Neurology, Vascular Surgery and Cardiovascular Surgery, Cardiology, and Biometry who assess the progress, safety of data and critical efficacy endpoints. The mission of the DSMB is to ensure the ethical conduct of the trial and to protect the safety interests of patients in this trial. In the context of overall patient safety the DSMB will receive periodic reports (also including center specific information) as well as any special reports as requested by the DSMB and to be prepared by the trial data center. The DSMB will have access to all trial data. The need and frequency of face-to-face meetings will be determined by the DSMB and the Steering Committee, taking into account the possibility of teleconferencing and other electronic conference options. DSMB safety data review meetings will regularly be held twice a year, with the first one scheduled 8 weeks after the first 200 patients have passed 30 days after therapy. Another meeting will be held after the interim analysis _______________________________________________________________________________________________________ Page 33 after 550 patients. The last meeting will be scheduled when all patients have passed this time limit. The DSMB will define details of their working procedures in an extra protocol to keep the independence from the study. The DSMB protocol will be part of the TMF.

Steering Committee (SC) and Executive Committee (ExC)
The SC consists of members of all involved disciplines that actively take part in the execution of this study. Because of the size of the steering committee, 3 members of the steering committee build an Executive Committee (ExC)

Clinical Adjudication committee (CAC)
A blinded clinical adjudication committee will confirm all outcome event notices, as it will make the final decision on whether a patient had any outcome event. All patients thought to have an outcome event will have their CRF and, if necessary, additional information examined by the members of this committee. The committee can also request additional ancillary information from an individual study investigator to assist their review. The CAC will also confirm the type (e.g. ischemic, hemorrhagic) and localization of each stroke or cause of death.

Quality Subcommittees
Two subcommittees with experts in this field have formulated rules and recommendations for diagnosis and treatment. There are recommendations for best medical treatment and cardiovascular surgery. These recommendations are appendices of this protocol. The subcommittees have also defined details of the quality criteria (see section 10.5.2) for participation of an investigator. The head of each sub-committee is also a member of the SC.

Centers
Each study center must consist of at least one neurologist and one cardiovascular / vascular surgeon. Certification of a participating center will follow upon compliance to all quality criteria. The steering committee can withdraw the certification to participate in the trial on recommendation of the DSMB and/or the monitoring agency.

Investigators
For all involved disciplines quality criteria are defined by the appropriate quality committee (see TMF). Every neurologist has to demonstrate his/her ultrasoundexpertise and should have a NIHSS-certificate. Cardiovascular / vascular surgeons must demonstrate their experience as defined in the appendix of the surgical quality subcommittee. The local principal investigator has to confirm this documentation. Together with the DSMB the quality committees will monitor local complication rates and can recommend the steering-committee or the executive committee to withdraw a certificate.

Investigator Site File
The investigator site file will contain the following documents

Good Clinical Practice
The procedures set out in this trial protocol, pertaining to the conduct, evaluation, and documentation of this trial, are designed to ensure that all persons involved in the trial abide by Good Clinical Practice (GCP) and the ethical principles described in the applicable version of the Declaration of Helsinki and the "Berufsordnung für Ärzte". The trial will be carried out in keeping with local legal and regulatory requirements.

Patient information and informed consent
Before being admitted to the clinical trial, the patient must consent to participate after the nature, scope, and possible consequences of the clinical trial have been explained in a form understandable to him or her. The patient and investigator must sign and date the informed consent. The signed Informed Consent Form will be filed by the investigator. A copy of the signed informed consent document must be given to the patient

Patient insurance
No patient or travel accident insurance will be taken out. Every hospital or surgeon is required to hold their own insurance so patients will be adequately covered in case of malpractice. Although desirable, no insurance can cover the risk inherent to either study procedure which moreover both constitute the current standard of care.

Confidentiality
The data obtained in the course of the trial will be treated pursuant to the Federal Data Protection Law (Bundesdatenschutzgesetz, BDSG).
To enable the 5 year follow-up, the contact data of the patient will be archived on paper in the clinical project management unit. These data include patient number, name, address, (cell) phone, email, contact data of the family doctor and, if possible, contact data of a family member. The permission to store these data in a database strictly kept separate from the clinical data will be obtained in the informed consent.
Other clinical data of the patients will be identified solely by means of their date of birth (month and year only), and an individual identification code (patient number). Trial findings stored on a computer will be stored in accordance with local data protection law and will be handled in strictest confidence. For protection of these data, organizational procedures are implemented to prevent distribution of data to unauthorized persons. The appropriate regulations of local data legislation will be fulfilled in its entirety. The patient consents in writing to release the investigator from his/her professional discretion in so far as to allow inspection of original data for monitoring purposes by health authorities and authorized persons (inspectors, monitors, auditors). Authorized persons (clinical monitors, auditors, inspectors) may inspect the patient-related data collected during the trial ensuring the data protection law. The investigator will maintain a patient identification list (patient numbers with the corresponding patient names) to enable records to be identified.

Responsibilities of investigator
The principal investigator should ensure that all persons assisting with the trial are adequately informed about the protocol, any amendments to the protocol, the trial treatments, and their trial-related duties and functions. The principal investigator should maintain a list of subinvestigators and other appropriately qualified persons to whom he or she has delegated significant trial-related duties.

Approval of trial protocol and amendments
Before the start of the trial, the trial protocol, informed consent document, and any other appropriate documents have been submitted to the independent Ethics Committee (EC) in Essen. A written favourable vote of each EC is a prerequisite for initiation of this clinical trial. The statement of EC should contain the title of the trial, the trial code, the trial site, and a list of reviewed documents. It must mention the date on which the decision was made and must be officially signed by a committee member. This documentation must also include a list of members of the EC present on the applicable EC meeting and a GCP compliance statement. For each site a separate approval must be obtained.
Before the first patient is enrolled in the trial, all ethical and legal requirements must be met. All planned substantial changes will be submitted to ECs in writing as protocol amendments. They have to be approved by the ECs.

Continuous information to independent ethics committee
The EC will be informed in case the risk / benefit assessment changes or any other new and significant hazards for subjects' safety or welfare occur. Furthermore, a report on all observed serious adverse events (SAEs) will be submitted four times a year. The EC will be informed at the end of the trial. Application for registration of the trial will be filed to http://www.controlled-trials.com. The protocol of the study will be published in a peer-reviewed journal.

Monitoring
The ZKSE will make periodic visits to the trial site (minimum once a year, about 4-5 visits per site total). During these visits, they check periodically a random sample of patient data recorded against source (SDV). During the on-site monitoring visits a review of source documents, informed consent, primary outcome, recruitment, SAE documentation and reporting of adverse events will be done. CRFs and queries will be discussed with the investigator.
If the data quality and the protocol compliance in the trial site is good or sufficient, the monitor will visit this centre according to schedule (one visit per year). If data quality and/or protocol compliance is poor, the number of visits and/or the patient number for SDV will be increased. This monitoring concept is in compliance with ICH GCP. Monitoring of clinical trials is often interpreted as requiring intensive on-site monitoring, but the following paragraph should be noted: "The determination of the extent and nature of monitoring should be based on considerations such as the objective, purpose, design, complexity, blinding, size and endpoints of the trial. In general there is a need for on-site monitoring, before, during and after the trial; however, central monitoring in conjunction with procedures such as investigators' training and meetings and extensive written guidance can assure appropriate conduct of the trial in accordance with GCP. Statistically controlled sampling may be an acceptable method for selecting the data to be verified." (ICH GCP 5.18.3) In this study central monitoring will be combined with on-site monitoring. The following issues will be implemented for this study: 1) In addition to the regular on-site visits, the site will be contacted through letters, web conferences, web seminars or telephone calls by the ZKSE to review study progress, investigator and patient compliance with requirements, and follow up on any issues to be addressed. 2) A central monitoring will be carried out including the following checks:  data collected are consistent with adherence to the trial protocol  no key data are missing  data appear to be valid (for example, range and consistency checks)  review of recruitment rates, withdrawals and losses to follow-up _______________________________________________________________________________________________________ Page 37 3) On-site monitoring including SDV for a random sample of patients 4) Increasing the monitoring frequencies in trial sites with poor quality 5) In this large multi-centre study a further central monitoring of data using statistical techniques will be carried out. This statistical monitoring is useful for identifying unusual patterns of data, and can be used to detect sites or individuals where there may be deviations from the protocol. 6) ZKSE may audit the study at any time. Investigators will be given notice before an audit occurs. They are obliged to permit these visits, assist visiting study personnel and auditors as necessary and make available any records required.

Inspections / Audits
DFG and an auditor authorized by the sponsor may request access to all source documents, CRF, and other trial documentation. Direct access to these documents must be guaranteed by the investigator who must provide support at all times for these activities.

Financing of the trial
The trial will be financed using funds of the DFG (WE 2585/3-1).

Publication
All information concerning the trial is confidential before publication. The data on safety and on the primary endpoint (30-day) results will be published first after the end of the trial. Publications will be prepared from a writing committee in the name of all CABACS investigators including the responsible trial statistician.