Effect of carotid revascularization on cerebral autoregulation in combined cardiac surgery

Abstract

OBJECTIVES

Combined carotid artery endarterectomy (CEA) and coronary artery bypass grafting surgery is considered to reduce long-term stroke risk for patients with severe carotid artery stenosis. The benefits of CEA for improving cerebral perfusion during subsequent cardiopulmonary bypass (CPB) are unclear. The purpose of this pilot study was to assess cerebral autoregulation and cerebral oximetry in patients undergoing combined CEA and cardiac surgery with those undergoing cardiac surgery without significant carotid artery stenosis or with uncorrected stenosis.

METHODS

Cerebral autoregulation was monitored continuously in 257 patients with the cerebral oximetry index (COx). COx represents a moving Pearson's correlation coefficient between low-frequency changes in regional cerebral oxygen saturation (rScO₂) and mean arterial pressure that has been validated in previous investigations. Impaired autoregulation was defined as a value of COx ≥0.3.

RESULTS

Nineteen patients had prior CEA, 8 underwent combined CEA and cardiac surgery, 8 had uncorrected stenosis >70% and 197 had stenosis <50%. Combined, patients with stenosis >70% had a higher COx before CPB compared with those with stenosis <50% (median, 0.26, 25th percentile and 75th percentile [p25–p75], 0.18–0.33 vs 0.18, p25–p75, 0.07–0.27, respectively, P = 0.054). Patients who underwent combined CEA and cardiac surgery had a higher COx before surgery compared with those with prior CEA (P = 0.027) and stenosis <50% (P = 0.026). There were no differences in average COx or rScO₂ during CPB in patients undergoing combined CEA and cardiac surgery compared with those with prior CEA (P = 0.53, 0.27) and those with stenosis <50% (P = 0.71, 0.19), respectively. During CPB, patients with uncorrected stenosis had an average COx of 0.36 (p25–p75, 0.28–0.56) indicating cerebral autoregulation impairment, and lower rScO₂ compared with patients with prior CEA (P = 0.006) and stenosis <50% (P = 0.005).

CONCLUSIONS

While higher at baseline, patients undergoing CEA immediately before cardiac surgery had COx and rScO₂ measurements during CPB similar to those with non-significant stenosis in contrast to those patients with uncorrected stenosis who had evidence of impaired autoregulation and lower rScO₂. These preliminary results suggest the potential utility of COx, possibly for complimenting patient selection for CEA as well as for individual patient management during surgery.

INTRODUCTION

The prevalence of carotid artery stenosis >50% has been estimated to be 12–17% in patients undergoing coronary artery bypass grafting (CABG) surgery and 6–8.5% of these patients have stenosis >80% [1–4]. The management of concomitant severe carotid artery stenosis for patients in need of cardiac surgery remains controversial. In a systematic review, it was reported that patients undergoing cardiac surgery without significant carotid artery stenosis have a perioperative stroke risk of <2%, whereas the risk increased to 3 and 5% for patients with unilateral and bilateral arterial stenosis of 50–99%, respectively [5]. Factors that are considered for combined carotid artery revascularization at the time of cardiac surgery include (i) severity of carotid artery stenosis (>50% for male and >70% for female), (ii) complexity of morphological characteristics of the carotid lesion including ulcer, (iii) the presence of related symptoms and (iv) the individual surgical team's 30-day combined stroke and death rate [6–8].

A steady supply of oxygenated blood is normally ensured by physiological processes that keep cerebral blood flow (CBF) constant over a range of blood pressures. When blood pressure is outside the limits of autoregulation, CBF becomes pressure passive. In these situations, cerebral hypoperfusion and risk for stroke may occur with low blood pressure, whereas cerebral hyperperfusion and risk for delirium may occur with high blood pressure [9, 10]. Monitoring of CBF autoregulation in real time during cardiac surgery can be performed by processing regional cerebral oxygen saturation (rScO₂) data obtained using near-infrared spectroscopy (NIRS) in relation to changes in mean arterial pressure (MAP) [11, 12]. We have previously found that impaired autoregulation using these methods is associated with a higher risk for stroke after cardiac surgery [13]. There are little data regarding the effects of carotid artery revascularization on CBF autoregulation for patient undergoing cardiac surgery. The purpose of this pilot study was to assess CBF autoregulation and rScO₂ in patients undergoing combined CEA and cardiac surgery with those undergoing cardiac surgery without significant carotid artery stenosis or with uncorrected stenosis.

MATERIALS AND METHODS

From April 2008 to March 2014, 257 patients undergoing cardiac surgery requiring cardiopulmonary bypass (CPB) were enrolled in a prospective observational study at the Johns Hopkins Hospital (clinical trial registration no. NCT 00981474). All procedures received the approval of the Institutional Review Board of The Johns Hopkins Medical Institutions, and all patients were provided with written informed consent. The patients included in this study were those who underwent preoperative carotid artery ultrasound for clinical indications.

Perioperative care

The patient had routine monitoring that included arterial pressure measured from a radial artery catheter placed for perioperative monitoring. General anaesthesia was induced and maintained with midazolam; fentanyl, isoflurane and pancuronium or vecuronium were given for skeletal muscle relaxation. CPB was initiated after administration of heparin to achieve an activated clotting time >480 s. The CPB flow was non-pulsatile flow and maintained between 2.0 and 2.4 l/min/m². Isoflurane concentration was managed at 0.5–1.0% through oxygenator gas inflow, and arterial pressure was controlled using normal institutional practices by adjusting CPB flow and the administration of phenylephrine. The patients were managed with alpha-stat pH management and continuous in-line arterial blood gas monitor that was calibrated hourly.

NIRS-based autoregulation monitoring

Two near-infrared signal (NIRS) monitor sensor pads were placed on the patient's forehead prior to induction of anaesthesia (INVOS 5100C, Covidien, Mansfield, MA, USA). The methodology and the analysis of these signals with MAP signals have been described [11, 14]. In summary, analogue continuous arterial pressure signals were processed with a data acquisition module (DT9800, Data Translation, Inc., Marlboro, MA, USA) and, along with the digital NIRS signals, were analysed using the ICM+ software (University of Cambridge, Cambridge, UK). The signals were filtered as a non-overlapping 10-s mean value that were time-integrated, which is equivalent to having a moving average filter with a 10-s time window and resampling at 0.1 Hz. This eliminates high-frequency components described by respiration and pulse waveforms. Additional high-pass filtering was applied with a DC cut-off set at 0.003 Hz to remove slow drifts associated with haemodilution at initiation of CPB. A continuous, moving Pearson's correlation coefficient between MAP and rScO₂ was calculated rendering a variable cerebral oximetry index (COx). Consecutive average COx within 10-s window was collected at 30 data points to monitor each COx in a 300 s window. COx approaches 1 when CBF autoregulation is impaired, whereas COx approaches 0 or is negative when CBF autoregulation is functional. A COx of 0.3 was considered as the limit of autoregulation based on the results of prior laboratory and clinical studies [12, 15, 16].

Data analysis

The normality of the distribution of the data was assessed by the Kolmogorov–Smirnov test. Continuous data that were normally distributed were analysed by Student's t-test for comparing two groups and one-way ANOVA for comparing more than two groups. Data that were not normally distributed were analysed by the Mann–Whitney U-test for comparing two groups and Kruskal–Wallis test for comparing more than two groups. For all categorical variables, more than 20% of the cells had expected values below 5. Fisher's exact test was used for the analysis. The patients were grouped based on the severity of carotid artery stenosis determined by preoperative carotid artery ultrasound findings. The patients were classified as having non-significant carotid artery stenosis when the luminal narrowing was <50%, and significant artery stenosis for luminal narrowing of >70%. The patients were then grouped into four categories based on their degree of carotid artery stenosis and the surgical procedure performed: combined CEA and cardiac surgery; cardiac surgery with uncorrected significant carotid artery stenosis (>70%); cardiac surgery with prior CEA and cardiac surgery with non-significant carotid artery stenosis (<50%). Comparison of COx and rScO₂ was made between these four groups in the periods: before CPB (after induction of anaesthesia and initiation of CPB) and during CPB. Changes in COx and rScO₂ between these periods were also analysed within each four groups using the Wilcoxon-signed rank test. Two-sided P-values of less than 0.05 are considered statistically significant. No adjustment for multiple comparisons was performed.

RESULTS

Among the 257 patients, 8 patients underwent combined CEA and cardiac surgery, 8 had uncorrected significant carotid artery stenosis (>70%), 19 had undergone prior CEA and 197 had non-significant carotid artery stenosis (<50%). The median period between prior CEA and cardiac surgery was 36 (p25–p75; 12–80) months. Twenty-five patients with carotid artery stenosis of 50–69% were not included in this analysis. In patients with prior CEA, 2 patients (10.5%) had carotid artery stenosis of 50–69% and 17 (89.5%) had carotid artery stenosis of <50%. The 2 patients with carotid artery stenosis of 50–69% in the prior CEA group were excluded from the analysis. All patients who underwent combined CEA and cardiac surgery had carotid artery stenosis of >70%. The characteristics of the patients are presented in Table 1. A history of peripheral vascular disease was lower in the patients with carotid artery stenosis of <50% (P < 0.001). Arterial pressure, arterial blood gas results and duration of each period are listed in Table 2. There were no differences in average arterial pressure before CPB (P = 0.44) and during CPB (P = 0.75). The gas measurements during CPB were not different among the four groups: PaO₂ (P = 0.73); PaCO₂ (P = 0.13); pH (P = 0.79) and haemoglobin (P = 0.87).

Table 1:

Patient's demographic and medical information for the patients with combined carotid endarterectomy (CEA) and cardiac surgery, uncorrected significant carotid artery stenosis (>70%), prior CEA and patients with non-significant carotid artery stenosis (<50%)

	CEA and cardiac surgery	Carotid artery stenosis (no CEA)	Prior CEA	Carotid artery stenosis <50%	P-value
	N = 8	N = 8	N = 17	N = 197
Age (year)a	67 ± 12.4 (57–78)	69 ± 7.6 (62–75)	72 ± 8.6 (68–77)	71 ± 8.0 (70–72)	0.43
Male (%)	5 (62.5%)	8 (100.0%)	11 (64.7%)	139 (70.6%)	0.24
Hypertension (%)	6 (75.0%)	6 (75.0%)	16 (94.1%)	168 (85.3%)	0.37
Diabetes (%)	5 (62.5%)	5 (62.5%)	9 (52.9%)	93 (47.2%)	0.69
CHF (%)	0 (0.0%)	1 (12.5%)	4 (23.5%)	29 (14.7%)	0.51
Peripheral vascular disease (%)	3 (37.5%)	6 (75.0%)	12 (70.6%)	21 (10.7%)	<0.001
COPD (%)	2 (25.0%)	2 (25.0%)	6 (35.3%)	19 (9.6%)	0.006
Prior cerebral vascular event (%)	2 (25.0%)	1 (12.5%)	4 (23.5%)	17 (8.6%)	0.070
Aspirin (%)	7 (87.5%)	5 (62.5%)	15 (88.2%)	149 (75.6%)	0.49
Statins (%)	7 (87.5%)	6 (75.0%)	12 (70.6%)	126 (64.0%)	0.65
Angiotensin-converting enzyme inhibitors I (%)	2 (25.0%)	3 (37.5%)	5 (29.4%)	71 (36.0%)	0.92
Calcium channel blocker (%)	2 (25.0%)	1 (12.5%)	5 (29.4%)	52 (26.4%)	0.90
Beta-blocker (%)	5 (62.5%)	6 (75.0%)	11 (64.7%)	122 (61.9%)	0.97
Diuretics (%)	3 (37.5%)	2 (25.0%)	6 (35.3%)	82 (41.6%)	0.84
Current smoker (%)	0 (0.0%)	1 (12.5%)	2 (11.8%)	17 (8.6%)	0.74
Previous smoker (%)	6 (75.0%)	6 (75.0%)	8 (47.1%)	88 (44.7%)	0.15
Surgery
CABG (%)	7 (87.5%)	6 (75.0%)	11 (64.7%)	105 (53.3%)	0.55
CABG + AVR/MVR (%)	1 (12.5%)	2 (25.0%)	1 (5.9%)	38 (19.3%)
AVR/MVR (%)	0 (0.0%)	0 (0.0%)	4 (23.5%)	45 (22.8%)
Others (%)	0 (0.0%)	0 (0.0%)	1 (5.9%)	9 (4.6%)
Bypass time (min)b	95 (75–114)	112 (85–147)	86 (66–130)	101 (83–132)	0.53
Total clamp time (min)b	52 (46–73)	61 (52–99)	54 (46–70)	63 (52–83)	0.30

Data are listed as numbers and percent of patients for dichotomous variables with the exception of age listed as ^amean ± SD (95% confidence interval), and duration of CPB and aortic cross-clamping that are listed as ^bmedian (25th percentile to 75th percentile). The P-value represents comparisons between all four groups.

AVR: aortic valve replacement; CABG: coronary artery bypass graft; CHF: chronic heart failure; COPD: chronic obstructive pulmonary disease; MVR: mitral valve replacement or repair; statins: HMG-CoA reductase inhibitors.

Table 2:

Blood pressure and blood gas measurement during cardiopulmonary bypass

	Combined CEA and cardiac surgery	Uncorrected carotid artery stenosis (>70%)	Prior CEA	Non-significant carotid artery stenosis (<50%)	P-value
	N = 8	N = 8	N = 17	N = 197
Blood pressure (mmHg)
Before CPBa	85 ± 10.3 (76–93)	79 ± 8.5 (72–86)	79 ± 10.1 (74–84)	80 ± 8.7 (79–81)	0.45
CPBa	76 ± 8.1 (69–83)	77 ± 9.5 (69–84)	77 ± 7.6 (74–81)	80 ± 8.7 (74–77)	0.75
PaO2 (mmHg)b	284 (240–334)	258 (244–283)	267 (254–279)	260 (238–288)	0.73
PaCO2 (mmHg)b	39.5 (38.0–41.3)	41.0 (40.4–41.5)	41.0 (40.0–42.0)	41.8 (40.0–43.0)	0.13
pHb	7.39 (7.38–7.40)	7.39 (7.39–7.41)	7.38 (7.37–7.41)	7.39 (7.37–7.41)	0.79
Haemoglobin (g/dl)b	8.2 (7.3–9.2)	8.5 (8.1–9.4)	8.5 (8.2–9.1)	8.6 (7.7–9.8)	0.87
Time duration (min)
Before CPBb	77 (72–93)c	144 (135–175)	96 (60–142)	117 (93–145)	0.012
CPBb	95 (75–114)	112 (85–147)	86 (66–130)	101 (83–132)	0.53

The P-value represents comparisons between all four groups.

CPB: cardiopulmonary bypass.

^aMean ± SD (95% confidence interval).

^bMedian (25th percentile to 75th percentile).

^cTime duration from induction of anaesthesia to start of CEA.

The results of COx analysis before CPB for all patients based on the severity of carotid stenosis are shown in Fig. 1. Before CPB, patients with carotid artery stenosis of >70%, including patients with uncorrected carotid artery stenosis and those with planned concomitant CEA, had a higher COx compared with those with carotid artery stenosis of <50% (P = 0.054). The COx results for each group of patients are shown in Fig. 2. Before CPB, at baseline, COx in patients subsequently undergoing combined CEA and cardiac surgery was significantly higher than in those with prior CEA (P = 0.027) and in those with non-significant carotid artery stenosis (<50%; P = 0.026). There were no significant differences between the groups in mean COx during CPB (P = 0.48): Patients with uncorrected carotid artery stenosis (>70%) were not different from those with prior CEA (P = 0.073) and those with non-significant carotid artery stenosis (<50%; P = 0.11). Within each group, there was a significant increase in COx in patients with significant stenosis (P = 0.037), prior CEA (P = 0.029) and non-significant stenosis (P < 0.001) compared with the measurement obtained before CPB. Only patients who underwent combined CEA and cardiac surgery had no significant difference between the periods (P = 0.68). During CPB, only the uncorrected stenosis group (>70%) had an average COx of 0.36, indicating cerebral autoregulation impairment.

Figure 1:

Box and whisker plots comparing COx before cardiopulmonary bypass in patients with significant stenosis of >70% and non-significant stenosis of <50% (median, 0.26, p25–p75 0.18–0.33 vs 0.18, p25–p75, 0.07–0.27, P = 0.054), respectively. The horizontal line in the shaded box represents the median value, and the shaded box represents the interquartile range. The error bars below and above the shaded area represent ±1.5× the interquartile range; points beyond the error bar are outliers. COx: cerebral oximetry index; CPB: cardiopulmonary bypass.

Figure 2:

Box and whisker plots comparing mean COx before cardiopulmonary bypass (CPB), after carotid endarterectomy (CEA) and during CPB among the surgical groups. The horizontal line in the shaded box represents the median value, and the shaded box represents the interquartile range. The error bars below and above the shaded area represent ±1.5× the interquartile range; points beyond the error bar are outliers.

The rScO₂ result for each patient group is shown in Fig. 3. Before CPB, there were no differences in mean rScO₂ between the groups (P = 0.71): rScO₂ in patients subsequently undergoing combined CEA and cardiac surgery was not different from that in those with prior CEA (P = 0.23) and in those with non-significant carotid artery stenosis (<50%; P = 0.15). During CPB, rScO₂ was not different between patients undergoing combined CEA and cardiac surgery and those with prior CEA (P = 0.27) or non-significant carotid artery stenosis (<50%; P = 0.19). However, patients with uncorrected carotid artery stenosis (>70%) had lower rScO₂ during CPB compared with those with prior CEA (P = 0.006) or non-significant carotid artery stenosis (<50%; P = 0.005). Within each group, there was a significant decline in rScO₂ during CPB in patients with significant stenosis (P = 0.002), prior CEA (P < 0.001) and non-significant stenosis (P < 0.001) compared with the measurements obtained before CPB. Only patients who underwent combined CEA and cardiac surgery had no significant difference between the periods (P = 0.074).

Figure 3:

Box and whisker plots comparing mean regional cerebral oxygen saturation (rScO₂) before cardiopulmonary bypass (CPB), after carotid endarterectomy (CEA) and during CPB among the surgical groups. The horizontal line in the shaded box represents the median value, and the shaded box represents the interquartile range. The error bars below and above the shaded area represent ±1.5× the interquartile range; points beyond the error bar are outliers.

DISCUSSION

In this study, we found that the patients who subsequently underwent combined CEA and cardiac surgery had a significantly higher COx before surgery compared with those with prior CEA and non-significant stenosis (<50%). While there were no differences in average COx during CPB between groups, COx was higher during CPB compared with before CPB in those with non-significant stenosis (<50%), uncorrected significant stenosis (>70%) and those with prior CEA. During CPB, only the uncorrected stenosis group (>70%) had an average COx of 0.36, indicating cerebral autoregulation impairment. There was a significant decline in rScO₂ during CPB in all groups compared with the measurements obtained before CPB except for those who underwent combined CEA and cardiac surgery. Patients with uncorrected carotid artery stenosis (>70%) had lower rScO₂ during CPB, compared with those with prior CEA and non-significant carotid artery stenosis (<50%).

Patients with previous transient ischaemic attack (TIA), or severe carotid artery stenosis, are at high risk for perioperative stroke. Prior history of TIA or stroke, for example, is associated with an 8.5% (95% CI 4.9–12.1) frequency of perioperative stroke compared with 2.2% (95% CI 1.4–3.1) in neurologically asymptomatic patients undergoing CABG [6]. Limited vasodilatory reserve in vascular territories distal to severe carotid artery stenosis may be manifest as impaired CBF autoregulation. In this situation, CBF is pressure passive, which may result in cerebral hypoperfusion during hypotension or cerebral hyperperfusion during hypertension possibly leading to brain injury. Impaired autoregulation is associated with risk for ipsilateral ischaemic events in non-surgical patients with symptomatic or asymptomatic carotid artery stenosis [17, 18]. While we observed that patients with severe carotid artery stenosis had a higher COx indicating dysfunctional autoregulation than those without significant stenosis, our pilot study has too small of a sample size to infer any relationship between impaired autoregulation and stroke.

During CPB, we observed that patients undergoing concomitant CEA and cardiac surgery had similar COx measurements as those without significant carotid disease, whereas patients with uncorrected carotid artery stenosis at the time of CPB had an average Cox of >0.3 during CPB. In prior studies, we found that impaired regulation of similar magnitude (i.e. Cox >0.3) during CPB is associated with major postoperative complications such as stroke, prolonged mechanical ventilation, acute kidney injury and mortality [13, 15, 19, 20]. We also noted that rScO₂ decreased from baseline during CPB in all groups except for those who underwent combined CEA and cardiac surgery. Furthermore, patients with uncorrected carotid stenosis had lower rScO₂ during CPB compared with those with prior CEA and carotid artery stenosis <50%. While our data must be viewed as preliminary tempering conclusions, these results are consistent with improved cerebral perfusion in patients undergoing CEA before CPB compared with those with uncorrected severe carotid artery stenosis.

In this study, we observed that patients with carotid stenosis of >70% had a higher COx indicating perturbed CBF autoregulation compared with those with non-significant stenosis (Fig. 1). In a study of 165 patients with internal carotid stenosis of >70% or occlusion, transcranial Doppler measurement of CBF autoregulation correlated with CO₂ reactivity, and was more robust in identifying risk for subsequent ischaemic events [21]. Thus, measuring CBF autoregulation might provide a means for evaluating the extent of vascular compromise in patients with carotid artery stenosis, possibly complimenting decisions on determining candidates for CEA. We have previously reported that non-invasive NIRS measured rScO₂ can serve as a clinical surrogate of CBF for autoregulation monitoring [11, 12]. These methods involve signal processing of rScO₂ in relationship with MAP focusing on the low-frequency components (20 s–3 min) in the bandwidth of vasoreactivity mediating CBF autoregulation [22]. Haemodynamic management based on real-time monitoring of cerebral autoregulation rather than the current standard of care where MAP targets are empirically chosen might provide a means for reducing neurological complications after cardiac surgery.

There are several limitations that should be considered in interpreting these results. First, as mentioned, the sample size of the study is small precluding any inferences on the relationship between impaired autoregulation and patient outcomes. Our preliminary study provides a rationale for future studies to address this question, as well as the data to perform sample size estimates. Our use of a COx of >0.3 as indicative of impaired autoregulation is rather arbitrary. While there is no consensus on what value of COx represents pressure passive CBF, a COx of >0.3 was found to have the highest sensitivity and specificity for detecting the lower limit of autoregulation in animal studies [23]. Furthermore, this value has been previously shown to be predictive of adverse patient outcomes after cardiac surgery [10, 12, 15, 20]. Impaired cerebral autoregulation as defined by a COx of >0.3 is reported to occur in 12–18% of the patients undergoing CPB [20, 24]. Similarly, impaired CBF autoregulation, monitored by transcranial Doppler, occurs in 20% of the patients undergoing CPB.

In conclusion, using COx monitoring, we found that patients with severe carotid artery stenosis have evidence of dysfunctional autoregulation prior to CPB. Patients undergoing CEA immediately before cardiac surgery had COx and rScO₂ measurements during CPB similar to those with non-significant carotid stenosis in contrast to those patients with uncorrected severe carotid stenosis who had evidence of impaired autoregulation and lower rScO₂. These preliminary results suggest the potential utility of COx monitoring possibly for complimenting patient selection for CEA as well as individual patient management during surgery.

Funding

This work was supported in part by grant-in-aid number (103363 from the Mid-Atlantic Affiliate of the American Heart Association) and grants (R01HL092259 from the National Institute of Health).

Conflict of interest: Daijiro Hori received funding from the Japan Heart Foundation/Bayer Yakuhin Research Grant Abroad. Charles W. Hogue received research funding from Covidien, Inc. (Boulder, CO, USA), the makers of the near-infrared spectroscopy monitors used in this study.

APPENDIX. CONFERENCE DISCUSSION

Scan to your mobile or go to http://www.oxfordjournals.org/page/6153/1 to search for the presentation on the EACTS library

Dr P. Ogutu (Augsburg, Germany): I think it's a very interesting idea to use autoregulation to see what happens or what the rate of stroke is for combined CEA and CABG, for example.

My first question is do you think there would have been some space for you to design the whole study in a randomized manner because if you look at most of the literature that compares the outcomes of carotid or concomitant procedures, most of the time they're very small groups. We don't really have a lot of evidence to show that what they have shown is actually something that we should go on to do in our practice. And at the same time, the guidelines also actually say that we need randomized studies.

So do you think there would have been some space or some idea of randomizing the whole study?

Dr Hori: This study is just looking at patients undergoing combined CEA and cardiac surgery, but we are currently doing a randomized study in patients who are undergoing cardiac surgery.

In this randomized study, we randomized the patients into intervention and non-intervention group. For both group of patients, we monitor cerebral autoregulation before they go on cardiopulmonary bypass. For the patients who were randomized to intervention, we tell the perfusionist and the anaesthesiologist the optimal blood pressure measured by cerebral autoregulation monitoring, and we follow up the patient for postoperative outcome.

So we're actually doing a randomized study right now at Hopkins.

Dr Ogutu: Okay. That's encouraging.

The second question is I realize that you had a cutoff of significant and non-significant stenosis, so you had below 50% and then above 70%. I was wondering why did you neglect the ones who had stenosis between 50 and 60%?

Dr Hori: In one of the guidelines, there is a gender difference in the definition of significant stenosis. Carotid stenosis of more than 69% is considered to be significant in females while carotid stenosis greater than 50% is considered significant for males.

So I did not include the patients with carotid stenosis of 50 to 69% in this study. However, I did do the analysis of the patient with 50 to 69% stenosis, and the results were very similar to those with a stenosis of less than 50%.

Dr Ogutu: Were these stenoses bilateral or unilateral?

Dr Hori: For the analysis, I used the side that had stenosis and did not use the side which didn't have a stenosis.

Dr Ogutu: Do you think that would have made a difference if you had patients who had bilateral stenosis?

Dr Hori: I think so. Collateral blood flow in patients with carotid stenosis should be associated with cerebral perfusion. I think we should look into that as well.

Dr Ogutu: So do you think the autoregulation would have been more impaired in these patients?

Dr Hori: If they had bilateral stenosis, I think they would be more dysfunctional. The arteries would be more likely to be pathological and dysfunctional to cerebral autoregulation despite the collateral flow.

Dr Ogutu: And the last question is, were there any asymptomatic patients in your cohort?

Dr Hori: The ones who didn't have combined CEA and cardiac surgery were asymptomatic.

Dr Ogutu: No. I mean, the ones who had carotid stenosis.

Dr Hori: So there were 16 patients who had severe stenosis, and 8 of them had combined CEA and cardiac surgery because they were symptomatic. But the other 8 did not have combined CEA and cardiac surgery because they didn't have any symptoms.

Dr Ogutu: But you're referring to symptoms that were cardiac related, but I'm thinking about the symptoms that were carotid related.

Dr Hori: The symptoms that I am referring to are carotid related. So the patients who were grouped into uncorrected stenosis had no symptoms of the carotid stenosis.

Dr A. Hassouna (Cairo, Egypt): I have one question. You have four groups or four subgroups of your study?

Dr Hori: Right.

Dr Hassouna: Did you make a comparison for the four groups together, altogether before proceeding to subgroup comparison? I mean, did you compare the four groups first?

Dr Hori: Yes, I did.

Dr Hassouna: And then you proceeded to the subgroup comparison?

Dr Hori: Yes.

Dr Hassouna: But I imagine that most of the comparison of the four groups were non-significant?

Dr Hori: Yes that is correct.

Dr Hassouna: And we have to criticize the subgroup comparison if the main comparison was not significant.

Dr Hori: Yes.

Dr Hassouna: You are aware of this?

Dr Hori: Yes.

Dr Hassouna: The second point, you have proceeded to make multiple subgroup comparisons. This, as you know, would inflate the P value. Did you put any penalty to correct for this inflation or not?

Dr Hori: No, I am afraid we did not.

Dr Hassouna: You didn't make. Okay. Thank you.