Significant Correlation between Cerebral Oximetry and Carotid Stump Pressure during Carotid Endarterectomy
Article Outline
Limited information on a correlation between carotid stump pressure and cerebral oximetry changes associated with cross-clamping of carotid vessels during carotid endarterectomy (CEA) prompted us to prospectively evaluate 38 consecutive CEAs in 37 patients. The authors used the INVOS-4100 cerebral oximeter to measure cerebral oximetry (cerebral oxygen saturation) before (t1) and after (t2) cross-clamping along with carotid stump pressure. All patients had CEA under general anesthesia with the routine use of a Javid shunt. Cross-clamping (t1 vs. t2) resulted in statistically significant changes (p < 0.0001) on the operated side of 6.03 units or a percent change of 9.2% when analyzed using the nonparametric signed-rank test. The nonoperated side had insignificant change (p = 0.71). Spearman correlation analysis revealed significant correlation (r = -0.63) between cerebral oximetry changes on the operated side and carotid stump pressure such that a larger change in cerebral oximetry due to cross-clamping was strongly and significantly correlated with lower carotid stump pressure. Using regression analysis, stump pressures of 25 and 50 mm Hg were predicted by cerebral oximetry changes of 28.5 or 8.8 units, respectively. This is equivalent to a percent change from baseline (t1) of 41.1% or 13.1%, respectively. Taken together, these findings suggest that cerebral oximetry can be used as an alternative to carotid stump pressure to provide noninvasive, inexpensive, and continuous real-time monitoring during CEA.
Introduction
Carotid endarterectomy (CEA) is a commonly performed vascular procedure. The procedure requires clamping of the carotid vessels, and as a result, cerebral ischemia and perioperative stroke are of major concern. There are three major views on methods to minimize cerebral ischemia. These are routine shunting, selective shunting in high-risk patients, and no shunting. Currently, there is no conclusive study to determine which of these three techniques dealing with cerebral ischemia is most desirable.1 Nonetheless, surgeons interested in monitoring cerebral perfusion have a variety of methods available to assess the need for shunting and to optimize hemodynamics.
Previous studies have shown adequacy in identifying patients at ischemic risk and assessing the need for shunting using carotid artery stump pressure measurements, transcranial Doppler, intraoperative electroencephalographic (EEG) measurements, neurological assessment in awake patients, and cerebral oximetry. Recently introduced, cerebral oximetry has the benefits of noninvasive continuous monitoring of regional cerebral oxygen saturation (rSO2) at relatively low cost. Historically, stump pressure has been used as a benchmark when comparing other methods of ischemic risk assessment. However, few data are available on the correlation that may exist between stump pressure and cerebral oximetry.
The purpose of the study was to evaluate the potential correlation between the two means of risk assessment. This task was achieved by analyzing cerebral oximetry measurements before and after clamping on the operated side with measured stump pressure for a statistically significant correlation.
Methods
Thirty-seven patients undergoing 38 CEAs between November 2005 and June 2006 were prospectively evaluated. Critical rSO2 on the operated side was available for 36 of the 38 procedures. As a result, only 36 CEAs were included in the data analysis. Patients were considered symptomatic if they presented with a recent onset of transient ischemic attack (TIA), amaurosis fugax, or stroke.
Preoperative evaluation consisted of duplex carotid scans in all patients. Many patients, but not all, underwent magnetic resonance angiography. No patient in this study underwent a standard carotid angiogram.
All CEAs were done under general anesthesia by the same surgeon. A Javid (BARD, Tempe, Arizona) shunt was routinely used in all cases, and bovine pericardial patch (Vascuguard, Synovis, St. Paul, Minnesota) was used for arteriotomy closure. The INVOS-4100 oximeter (Somanetics, Troy, MI) was used to measure rSO2. Sensors were placed bilaterally and monitored by the anesthesiologist. The cost of each sensor was $60, and the oximeter was provided by Somanetics for a total cost of $120/procedure.
Radial artery pressure lines were inserted in all patients prior to induction of anesthesia. Various anesthesiologists were involved in the care of these patients. The involved anesthesiologists predominantly cared for patients undergoing cardiothoracic and vascular procedures. There were no specific anesthetic or hemodynamic protocols involved. Episodes of hypotension, hypertension, bradycardia, or tachycardia were managed by standard methods. Peripheral oxygen saturation was continuously monitored.
rSO2 was measured continuously and recorded for this study on two separate occasions: prior to clamping (t1) and immediately after clamping but before the shunt placement (t2). t1 was recorded <30 sec prior to clamping, and t2 was recorded in the 30-90 sec required for measurement of stump pressure, arteriotomy, and shunt insertion. In addition, immediately after clamping of the proximal common carotid artery and external carotid artery, internal carotid stump pressure was measured. All CEAs were performed with routine Javid shunt placement. All patients awoke in the operating room at the end of the procedure and were assessed for obvious neurological deficits.
Statistical analysis of collected data was carried out by a biostatistician. Since changes in rSO2 observed with clamping were not normally distributed, the nonparametric signed-rank test was used to determine statistical significance. Results were considered significant at p < 0.05. Spearman's correlation was used to investigate the correlation between changes in rSO2 on the operated side and measured carotid stump pressure. Simple linear regression was performed to predict a stump pressure as a function of change in cerebral oxygen saturation. SAS 9.1 for Windows (SAS Institute, Cary, NC) was utilized to carry out all calculations.
Results
From November 2005 until June 2006, 38 consecutive CEAs were studied. However, only 36 of them were used for the data analysis due to missing data. Fifty-six percent of patients were male, with a mean age of 68.3 years. Seventy-five percent of the CEAs were performed for asymptomatic disease in comparison to 25% for symptomatic disease. Patients had risk factors that would be expected of this patient population and are listed in Table I. Postoperatively, there was one complication of TIA 6 hr after the CEA. No mortality was observed in this series.
Table I. Patient characteristics
| Characteristics | n | % |
|---|---|---|
| Mean age (years) | 68.3 ± 8.4 | |
| Patients | 36 | 100 |
| Male | 20 | 55.6% |
| Female | 16 | 44.4% |
| Symptomatic | 9 | 25.0% |
| Asymptomatic | 27 | 75.0% |
| Coronary disease | 14 | 38.9% |
| Diabetes | 8 | 22.2% |
| COPD | 2 | 5.6% |
| Hypertension | 32 | 88.9% |
| Hyperlipidemia | 31 | 86.1% |
| Current smoker | 7 | 19.4% |
| Previous smoker | 17 | 47.2% |
Table II shows rSO2 of preclamping (t1) and postclamping (t2) values on both the operated and nonoperated sides. The changes in rSO2 between t1 and t2 expressed in terms of units and percent change from baseline (t1) are also listed. The results revealed that rSO2 of the nonoperated side showed no statistically significant change in relation to clamping, with a mean change in rSO2 of 0.11 ± 1.65 (p = 0.44). This is equivalent to a percent change from baseline (t1) of 0.05 ± 2.53% (p = 0.71).
Table II. Comparison of operated and nonoperated sides in regard to rSO2 differences and respective percent changes resulting from clamping (t1 vs. t2, mean ± SD)
| Side | t1 | t2 | rSO2 difference (% ±SD) | Percent difference (% ±SD) | pa |
|---|---|---|---|---|---|
| Operated | 65.17 ± 8.77 | 59.14 ± 11.23 | 6.03 ± 7.93 | 9.20 ± 11.92 | <0.0001 |
| Nonoperated | 69.37 ± 9.12 | 69.26 ± 8.67 | 0.11 ± 1.65 | 0.05 ± 2.53 | 0.71 |
ap values reflect percent change from preclamping value (t1). |
In contrast, rSO2 values on the operated side revealed a statistically significant change when comparing t1 and t2. The mean change in rSO2 was 6.03 ± 7.93 (p < 0.0001), equivalent to a percent change of 9.20 ± 11.92% (p < 0.0001). Taken together, these findings indicate that clamping of the carotid artery induces a significant change in rSO2 on the operated side, while the nonoperated side is unaffected.
The correlation between changes in cerebral oxygen saturation on the operated side and stump pressure was analyzed using Spearman's correlation. The result showed a high correlation value of -0.63 between the variables (p < 0.0001). This finding indicates that a larger drop in rSO2 that occurs after clamping on the operated side correlates with lower stump pressure.
Using regression analysis, stump pressure could be predicted as a function of changes in cerebral oxygen saturation. Applying this technique, stump pressure readings of 25 and 50 mm Hg were predicted with a clamping-associated drop in rSO2 of 28.5 and 8.8 units on the operated side, respectively. In terms of relative percent change from baseline (t1), this is equivalent to decreases of 41.1% and 13.1%, respectively. The regression analyses between absolute difference and stump pressure and relative percent difference and stump pressure are depicted in Fig. 1, Fig. 2, respectively.
Fig. 1
Regression analysis of mean stump pressure and relative percent difference of cerebral oximetry.
Fig. 2
Regression analysis of mean stump pressure and relative percent difference of cerebral oximetry.
Discussion
The first reported successful CEA was performed by Carrea et al. in 1951.2 Since then, CEA has become one of the most commonly performed vascular procedures. Even with over 50 years of experience, certain details surrounding the procedure are still debated. To minimize the risk of perioperative stroke, intraluminal shunting has been advocated by certain groups of surgeons; but no conclusive study exists that substantiates the need for such shunting or the benefit over no shunt placement.1 Furthermore, no consensus exists regarding whether routine shunting is superior to selective shunting due to contradictory study findings. Nonetheless, for clinicians who prefer a selective shunting technique, there are a number of available methods that have shown efficacy in evaluating the need for shunting. These include carotid artery stump pressure measurements,3, 4 transcranial Doppler,5 intraoperative EEG measurements,6, 7, 8 neurological assessment in awake patients,9 and cerebral oximetry.10, 11, 12, 13, 14
The value of carotid artery stump pressure measurements in assessing the need for shunt was first reported in 1969 by Moore and Hall.3 Since then it has become one of the most commonly utilized techniques. The measurement reflects back pressure in the internal carotid artery from collateral cerebral flow after cross-clamping has been applied to the common and external carotid arteries. It provides an inexpensive and simple method of assessing crude collateral circulation. A study by Hays et al.4 advocates 50 mm Hg as a safe stump pressure level without the need for shunting, while Moore and Hall3 endorse 25 mm Hg as the safe cut-off.4, 5 The disadvantage of the stump pressure technique is that it provides only one time reading during the procedure and fails to provide continuous monitoring. In addition, incomplete occlusion can erroneously increase stump pressure readings.
To circumvent these problems, other modalities were developed. EEG allows continuous or intermittent monitoring with low number of false negatives but is plagued with a significant number of false-positive readings and considerably increased cost of the operation.6, 7, 8 The technique of using local anesthesia allows neurological examination to be completed on awake patients and is seen as the most sensitive gauge on cerebral perfusion and function. However, most CEAs are still done using general anesthesia. Finally, employing intraoperative transcranial Doppler allows monitoring of blood flow velocity changes in the middle cerebral artery associated with cross-clamping. It is a reliable indicator for assessing the necessity for shunting and is unique in its ability to detect microembolization. However, this technique is troubled by the difficulty of isolating the middle cerebral artery even when used by experts. 5
In contrast, cerebral oximetry is a relatively new modality that provides an inexpensive and noninvasive method of continuous monitoring of cerebral oxygen saturation. The device consists of a monitor and two probes that are placed over a patient's forehead in a hair-free area, one on each side. The probes transmit near-infrared light of 730 and 805 nm wavelengths into the patient's forehead. The light will enter the tissue and refract. A portion of the light will refract back to the skin and is recorded by the two light detectors located on the probe. By analyzing the amount of returning photons in relation to the wavelength, average oxygenation of overlying primary gray matter blood and some white matter blood is inferred.15
Cerebral oximetry has previously been evaluated in patients undergoing CEA under regional anesthesia. Carlin et al.14 measured cerebral oxygen saturation in 16 patients and found that 14/16 who remained hemodynamically stable did not demonstrate neurological symptoms despite a drop of rSO2 from 69% to 64% after carotid clamping. Two of their patients developed hypotension and had decreases in rSO2 from 69% to 40% and from 68% to 48%, and both developed neurological symptoms, which improved when the hypotension was reversed. Samra et al.10 compared changes in rSO2 in patients who underwent CEA under regional anesthesia and who did (symptomatic) or did not (asymptomatic) manifest neurological changes after carotid clamping. Symptomatic patients had a decrease in rSO2 from 63.2% to 51.0% after carotid clamping compared to a drop of only 65.8% to 61.0% in asymptomatic patients. They did not find a critical rSO2 value that predicted ischemia, but they did determine that patients who had <20% decrease in rSO2 remained asymptomatic.
Using EEG, Rigamonti et al.12 found that a relative percent rSO2 drop of >15% was associated with a 20-fold increase in the odds of developing severe cerebral ischemia. Lastly, Cho et al.11 determined that a decrease in rSO2 of greater than 10 units was associated with a significant change in somatosensory evoked potential amplitude. The study concluded that an rSO2 drop of greater than 10 units from baseline or an rSO2 reading of <50% is associated with cerebral ischemia.
The current study attempted to correlate rSO2 with carotid stump pressure readings. We found a significant decrease in rSO2 from the preclamping value on the operated side, while no such statistically significant change was observed on the nonoperated side. This finding is consistent with a previous study that examined cerebral oximetry values at various stages of 42 consecutive CEAs.13 In that study, rSO2 was measured on both the operated and nonoperated sides at the following stages: before clamping (t1), after clamping but prior to shunt placement (t2), 5 min after shunting (t3), and after patch closure with reestablished flow (t4). On the operated side, clamping (t1 vs. t2) resulted in an rSO2 drop of -12.3%, while shunt placement (t2 vs. t3) increased rSO2 by +10.9%. When compared to baseline, shunting resulted in a -3.9% drop (t1 vs. t3), suggesting improved saturation but incomplete recovery to baseline. Lastly, there was no significant change after reestablishment of flow when compared to shunt placement (t3 vs. t4). As in the current study, the nonoperated side showed no statistically significant changes at all four data collection points (t1, t2, t3, t4).
In addition, our study found a strong correlation between stump pressure and change in rSO2 such that a larger drop in rSO2 was associated with lower stump pressure readings. Using regression analysis, we determined that stump pressures of 50 and 25 mm Hg were predicted with an rSO2 drop of 9 and 28.5 units on the operated side, respectively. It is of interest to note that the significant rSO2 cut-off stated by Cho et al.11 of 10 units closely resembles the predicted value of 9 units at 50 mm Hg from the current study. In terms of relative percent change from baseline (t1), our study showed that 13.1% and 41.1% decreases in rSO2 from baseline are predictive of stump pressure of 50 and 25 mm Hg, respectively. The 13.1% drop predictive of 50 mm Hg by our study is comparable to Rigamonti et al.'s12 15%, while it is lower than the value of 20% determined by Samra et al.10
In conclusion, we have shown that there was a significant correlation between changes detected by cerebral oximetry and stump pressure associated with cross-clamping during CEA. In addition to cerebral oximetry being an inexpensive and noninvasive monitoring tool, it is valuable in monitoring perfusion during shunting such that failure of rSO2 to return to near baseline after shunt placement should alert the surgeon to the possibility of shunt malfunction or inadequate placement.13 However, since a conclusive cut-off value does not exist for cerebral oximetry, the authors do not advocate using it as the only mean of evaluating ischemic risks and the need for selective shunting. A future study involving a larger patient population with an attempt to define safe cut-offs with sensitivity and specificity is strongly encouraged. Such study will provide the guideline needed for cerebral oximetry as a safe, inexpensive, and noninvasive alternative to carotid stump pressure in monitoring the potential for cerebral ischemia.
References
- . Routine or selective carotid artery shunting for carotid endarterectomy (and different methods of monitoring in selective shunting). Stroke. 2003;34:824–825
- . Surgical treatment of spontaneous thrombosis of internal carotid artery in the neck: carotid-carotidal anastomosis: report of a case. Acta Neurol. Latinoam. 1955;1:71–78
- . Carotid artery back pressure: a test of cerebral tolerance to temporary carotid occlusion. Arch Surg. 1969;90:702–710
- . Intraoperative measurement of carotid back pressure as a guide to operative management for carotid endarterectomy. Surgery. 1972;72:953–960
- . Relationship between middle cerebral artery blood flow velocity and stump pressure during carotid endarterectomy. Stroke. 1992;23:1439–1445
- . Relationship of intraoperative EEG monitoring and stump pressure measurements during carotid endarterectomy. Circulation. 1980;62:I4–I7
- . EEG as a criteria for shunt need in carotid endarterectomy. Ann Vasc Surg. 2002;16:756–761
- Benefits, shortcomings and costs of EEG monitoring. Ann Surg. 1985;201:785–792
- . Carotid endarterectomy in the awake patient. Am J Surg. 1985;150:159–165
- . Evaluation of a cerebral oximeter as a monitor of cerebral ischemia during carotid endarterectomy. Anesthesiology. 2000;93:964–970
- . Cerebral monitoring by means of oximetry and somatosensory evoked potentials during carotid endarterectomy. J Neurosurg. 1998;89:533–538
- . Clinical evaluation of near-infrared cerebral oximetry in the awake patient to monitor cerebral perfusion during carotid endarterectomy. J Clin Anesth. 2005;17:426–430
- . Cerebral oximetry monitoring during carotid endarterectomy: effect of carotid clamping and shunting. Vasc Endovasc Surg. 2003;37:407–413
- The use of near-infrared cerebral oximetry in awake carotid endarterectomy. J Clin Anesth. 1998;10:109–113
- INVOS Cerebral Oximeter. Principles of operation. Somanetics Product Information, Somanetics Corporation, Troy, MI.
PII: S0890-5096(07)00305-6
doi:10.1016/j.avsg.2007.07.022
© 2008 Annals of Vascular Surgery Inc. Published by Elsevier Inc All rights reserved.
