Original Contributions The Use of Near-Infrared Cerebral Oximetry in Awake Carotid Endarterectomy Robert E. Carlin, MD,* Daniel J. McGraw, MD,† J. Robert Calimlim, MD,‡ Michael F. Mascia, MD, MPH§ Department of Anesthesiology and Department of Surgery, State University of New York Health Science Center at Syracuse, Syracuse, NY
*Surgical Resident †Assistant Professor of Surgery ‡Assistant Professor of Anesthesiology §Assistant Professor of Anesthesiology; Director, Department of Critical Care Anesthesiology Address correspondence and reprint requests to Dr. Mascia at the Department of Critical Care Anesthesiology, SUNY Health Science Center at Syracuse, 750 E. Adams St., Room 2146, Syracuse, NY 13210, USA. Received for publication April 22, 1997; revised manuscript accepted for publication October 16, 1997.
Study Objective: To determine the utility of cerebral oximetry for monitoring the adequacy of cerebral blood flow (CBF) during carotid cross-clamp. Design: Prospective study. Setting: University hospital. Patients: 16 consecutive ASA physical status III (or higher) patients for awake carotid endarterectomy (CEA). Interventions: Regional cerebral oxygen saturation (SaO2) was monitored continuously during CEA, which was performed by the same surgeon, and with standard regional anesthetic, sedation, monitoring, and operative techniques. Data were recorded and analyzed using repeated measures analysis of variance (ANOVA). Measurements and Main Results: 14 hemodynamically stable patients demonstrated significant decreases in cerebral SaO2 from baseline: 69 1 1.8% to 64 1 1.2% at carotid cross-clamp (p , 0.001). After 5, 10, and 15-minute cross-clamp time, cerebral SaO2 was 63 1 1.4%, 64 1 1.5%, and 63 1 1.4%, respectively (p , 0.001, vs. baseline). On cross-clamp removal, cerebral SaO2 rose significantly: 67 1 1.6% (p , 0.01 vs. 5, 10, and 15 min). Two hypotensive patients (mean arterial pressures of 40 and 43 mmHg) developed signs and symptoms of global cerebral ischemia, with a concomitant decrease in cerebral oximetry (40% and 48%, respectively). These changes resolved with correction of hypotension. Conclusion: Cerebral SaO2 decreased significantly on carotid cross-clamp in patients undergoing awake CEA. Hemodynamically stable patients demonstrated no evidence of regional brain failure when SaO2 decreased to 63% (mean decrease of 7.2%). Two hemodynamically unstable patients had evidence of global brain failure when SaO2 was less than 48% (mean decrease of 36%). Our findings suggest that cerebral oximetry reflects CBF, and it may be an effective, noninvasive method of monitoring regional cerebral oxygenation changes during CEA. Significant reductions in regional SaO2 may be tolerated without evidence of brain failure. Further studies are needed to define an SaO2 threshold that reflects regional brain failure. © 1998 by Elsevier Science Inc. Keywords: Carotid endarterectomy, awake; oximetry, cerebral.
Introduction Near-infrared spectroscopy is a noninvasive technology used to measure cerebral oxygenation. The INVOS 3100 cerebral oximeter (Somanetics, Troy, MI) is a
Journal of Clinical Anesthesia 10:109 –113, 1998 © 1998 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010
0952-8180/98/$19.00 PII S0952-8180(97)00252-3
Original Contributions
near-infrared spectroscopic device designed for cerebral oximetry, approved by the Food and Drug Administration for investigational use*. Carotid endarterectomy (CEA) is associated with variable decreases in cerebral perfusion at the time of cross-clamp and, hence, variable decreases in regional cerebral oxygen saturation (SaO2) would be expected. In an effort to examine the clinical utility of cerebral oximetry, we used the INVOS 3100 cerebral oximeter in patients undergoing CEA during regional anesthesia. CEA is now established as a prophylactic procedure for the prevention of stroke in patients with symptomatic and asymptomatic internal carotid artery stenosis.1–3 However, the benefit of CEA in both groups of patients requires the occurrence of minimal perioperative morbidity and mortality. One of the most serious complications of CEA is the occurrence, or aggravation, of a neurologic deficit. Two major causes of intraoperative cerebral ischemia are embolic events and decreased cerebral blood flow (CBF). Decreased CBF, secondary to inadequate collateral flow, may be detected by a variety of techniques, including stump pressure monitoring, determination of jugular venous oxyhemoglobin saturation, transcranial Doppler ultrasonography, electroencephalography (EEG), and somatosensory-evoked potentials. However, these techniques may be unreliable because they lack sensitivity and specificity, and the data may be difficult to interpret.4 –7 Therefore, neurologic monitoring of the awake patient during carotid cross-clamp has been suggested as the most sensitive indicator of the adequacy of CBF.8 Recently, several clinical studies of cerebral oxygenation have been published using several different nearinfrared spectroscopic instruments to monitor regional SaO2 in the brain. Near-infrared spectroscopy allows continuous, noninvasive measurement of cerebral tissue oxygenation, and its use has been advocated in settings of head trauma and acute stroke, and during head and neck surgery.9 –11 To date, most studies of cerebral oximetry during CEA have been in patients administered general anesthesia,12,13 while only two reports have been in patients undergoing awake CEA.**,14 One problem in using cerebral oximetry during CEA is that a critical SaO2 threshold, below which neurologic dysfunction would develop, remains to be defined. Therefore, we used cerebral oximetry in patients who underwent CEA during regional anesthesia, expanding our original patient population,4 all of whom were diagnosed and treated by the same surgeon for similar pathology. The purpose of this study was to determine whether changes in cerebral SaO2 could predict which patients might benefit from insertion of a shunt during CEA with general anesthesia.
* The device has also been recently approved for clinical use. ** Mascia MF, McGraw DJ, Camporesi EM: The use of near infrared cerebral oximetry in awake carotid endarterectomy [Abstract]. Anesthesiology 1994;81:A532. 110
J. Clin. Anesth., vol. 10, March 1998
Materials and Methods Following SUNY Health Science Center at Syracuse Institutional Board Review approval, a prospective study was carried out on 16 consecutive ASA physical status III (or higher) patients undergoing elective CEA. Patients were all male, and they ranged in age between 53 and 74 years. Anesthesia was accomplished with deep and superficial cervical plexus block supplemented by local anesthetic infiltration and intravenous (IV) analgesia. CEA was performed by the same surgeon using a standard technique under loupe magnification. No carotid shunting was used in any patient during the procedure. All patients had a radial artery cannula placed for continuous arterial pressure monitoring during the operation and in the early postoperative period. Pulse oximetry and ECG were continuously monitored. Cerebral monitoring during CEA included checking the neurologic status with regional anesthesia. Heavy sedation was avoided to maintain the patient’s normal verbal and motor responses to our questions and commands. Neurologic examination consisted of patients moving all four extremities and answering simple questions during CEA, squeezing a ‘‘clicker’’ on the side opposite the CEA, and monitoring consciousness throughout the procedure. Intracerebral SaO2 was measured continuously and noninvasively with the INVOS 3100 cerebral oximeter on the hemisphere ipsilateral to the CEA. The device consists of a sensor that contains a near-infrared light transmitter and two photodetectors, mounted 30 mm and 40 mm from the transmitter, on a flexible adhesive holder for attachment to the scalp. The sensor is applied below the hairline on either the left or right side of the forehead, matching the side of CEA (in accordance with the Somanetics 3100 guidelines). The transmitter emits near-infrared light at wavelengths between 730 and 810 nm, maximizing tissue penetration. Near-infrared light is reflected by tissues in two parabolic curves. The detector placed 30 mm from the transmitter detects reflected light primarily through scalp and calvarium, while the detector positioned at 40 mm from the transmitter receives reflected light through scalp, calvarium, and up to 15 mm of brain tissue. The oximeter computer mathematically separates the reflected signal of superficial tissue (scalp and calvarium) from that of deeper tissues (scalp, calvarium, and brain). The number calculated is the SaO2 of blood in the brain. In this study, cerebral SaO2 was continuously monitored and recorded digitally at 1-minute intervals for subsequent analysis. Mean values for cerebral oximetry, systolic (SBP), diastolic (DBP), and mean arterial blood pressures (MAP) were recorded at the following times during CEA: (1) twice before carotid artery cross-clamping; (2) immediately at cross-clamp; (3) 5 minutes after cross-clamp; (4) 10 minutes after cross-clamp; (5) 15 minutes after cross-clamp; (6) on clamp removal; and (7) 10 minutes after clamp removal. Patients were followed until the time of discharge (24 to 48 hours).
Cerebral oximetry in awake CAE: Carlin et al.
Figure 1. Oximetric response to carotid cross-clamp (Xclamp). Values are means 6 SEM. Significance of difference from baseline: *p ,0.01, **p ,0.001; n 5 14.
Statistical Analysis Statistical analysis of the results was done using repeated measures analysis of variance (ANOVA). All results are given as means 6 SEM. A p-value less than 0.05 was considered significant. The Student-Newman-Keuls posthoc test was performed for multiple comparisons at the different time intervals.
cerebral SaO2 was 66 6 1.3% (Figure 1). The mean decrease in cerebral SaO2 from baseline to cross-clamp was 7.2%. Despite the decrease in cerebral SaO2, all patients remained asymptomatic and neurologically intact in the OR and postoperatively. No patient required a temporary shunt. In the hemodynamically stable patients, SBP did not change significantly on cross-clamping, but it did increase significantly from a baseline of 149 6 5.6 mmHg to 172 6 5.4 mmHg at 5 minutes, 170 6 4.6 mmHg at 10 minutes, and 173 6 4.0 mmHg at 15 minutes (p ,0.01, vs. baseline). MAP also increased significantly after crossclamping from a baseline of 100 6 3.1 mmHg to 112 6 2.9 mmHg at 5 minutes, 111 6 3.0 mmHg at 10 minutes, and 112 6 2.4 mmHg at 15 minutes (p ,0.05, vs. baseline). When the cross-clamp was released, SBP and MAP returned toward baseline values. DBP did not change significantly on clamping or declamping. The two hemodynamically unstable patients developed profound hypotension and transient brain failure. The first patient had a low MAP of 40 mmHg and a decrease in cerebral SaO2 from 69% to 40% after 5 minutes of cross-clamping. After treatment with vasopressors, MAP returned to normal, symptoms resolved, and SaO2 increased to 64%. The second patient had a MAP of 43 mmHg and a decrease in SaO2 from 68% to 48% when the cross-clamp was released. After treatment with vasopressors, MAP returned to normal, symptoms resolved, and SaO2 rose to 66% (Figure 2). During the hypotensive episodes, both patients had manifestations of global cerebral ischemia (nausea and light headedness). Both patients remained neurologically intact postoperatively.
Discussion Cerebral protection during CEA is more important than ever, because recent standardized trials demonstrate better outcome in patients with symptomatic and asymptom-
Results All but two patients remained hemodynamically stable systemically before, during, and after the carotid crossclamp. (The two unstable patients will be considered separately.) Arterial saturation was greater than 95% in the hemodynamically stable patients at all times during the operation. Mean hemoglobin, hematocrit, and ASA scores were 13.6 6 0 g/dl, 39.7 6 1.9%, and 3, respectively. Under baseline conditions, and shortly after the hemodynamically stable group entered the operating room (OR), cerebral SaO2 was 69 6 1.8%, without significant variation. Baseline SBP, DBP, and MAP were 149 6 5.6 mmHg, 75 6 2.3 mmHg, and 100 6 3.1 mmHg, respectively. When the carotid artery was cross-clamped, cerebral oxygenation fell from 69 6 1.8% to 64 6 1.2% (p ,0.001, vs. baseline). After 5, 10, and 15 minutes of cross-clamp time, cerebral SaO2 was 63 6 1.4%, 64 6 1.5%, and 63 6 1.4%, respectively (p ,0.001, vs. baseline). When the cross-clamp was removed, cerebral oxygenation increased significantly to 67 6 1.6% (p ,0.01, vs. 5, 10, and 15 min). Ten minutes after releasing the cross-clamp,
Figure 2. Oximetric response during carotid endarterectomy for hemodynamically stable (n 5 14) and hypotensive (n 5 2) patients. J. Clin. Anesth., vol. 10, March 1998
111
Original Contributions
atic internal carotid artery stenosis treated with CEA rather than medical therapy.1–3 Of the current intraoperative cerebral monitoring techniques, it is generally accepted that neurologic examination in the awake patient is the most sensitive monitor of cerebral function and adequacy of CBF.8 For this reason, we evaluated nearinfrared cerebral oximetry in patients undergoing CEA with regional anesthesia, in an attempt to define the SaO2 threshold necessary to create a neurologic deficit. Cerebral oximetry during CEA is a relatively new technique that enables noninvasive, continuous measurement of cerebral SaO2. It has been recognized for more than two decades that near-infrared light may be used to evaluate oxygen transport to the brain.15 In 1985, cerebral SaO2 was first measured by near-infrared spectroscopy in preterm infants. The thin skull allowed light to be passed through brain tissue and detected on the contralateral side of the skull.16,17 Because the vascular space in cerebral tissue normally consists of 75% venous and 25% arterial or capillary blood,18 the oximeter value more closely approximates the venous saturation and changes with oxygen extraction by cerebral tissue. In this study, we observed an abrupt mean decrease of 7.2% in cerebral SaO2 (compared with baseline) on carotid cross-clamping. Cerebral SaO2 reached a new plateau within 5 minutes and it remained significantly decreased after 15 minutes of clamp time, followed by a return toward baseline when the cross-clamp was removed. Our findings are similar to those of Samra et al.14 There were no neurologic deficits intraoperatively, during decreased cerebral oxygenation, or postoperatively. Two patients had symptoms of global cerebral ischemia and profound decreases in SaO2 associated with profound hypotension. Limitations in our study are imposed by the constraints of patient safety in clinical research. Specifically, limited invasive monitoring restricted simultaneous measurements of CBF, arterial and jugular venous blood gases, and pH, as well as gases and pH of the intracranial fluids (blood, cerebrospinal fluid, intracellular fluids, and interstitial fluid) within and outside the brain region being examined by near-infrared light. In addition, none of the patients studied developed clinically detectable signs or symptoms of regional brain failure during uncomplicated carotid artery cross-clamping and unclamping. These limitations make it difficult to explain with absolute certainty the changes we have seen. Nevertheless, we may speculate that the decrease and plateau in cerebral SaO2 during carotid occlusion are most likely due to increased oxygen extraction by relatively poorly perfused cerebral tissue. When the cross-clamp was released, cerebral SaO2 returned toward baseline, probably indicating restoration of CBF and oxygen delivery to cerebral tissue. However, at 10 minutes post-clamp removal, cerebral SaO2 decreased slightly but significantly. The subsequent decrease in SaO2 is unexplained, but it may be due to cerebral autoregulation that occurred after the carotid artery was cross-clamped. It is possible that, during cross-clamp, cerebral vasodilation may have occurred, increasing 112
J. Clin. Anesth., vol. 10, March 1998
oxygen delivery and extraction. When the clamp is removed and CBF restored, there may be a state of luxury perfusion, and, hence, the rapid increase in cerebral SaO2. As normal cerebral vascular tone and oxygen extraction return, SaO2 may decrease slightly from the initial peak. If we are correct, and cerebral oximetry was followed for a longer period of time, SaO2 would probably slowly rise back to baseline. A previous study13 implied that large decreases in cerebral SaO2 indicate inadequate collateral blood supply. However, that study was done in patients undergoing CEA with general anesthesia.13 In our study, with hemodynamically stable patients undergoing CEA while awake, there were no neurologic deficits associated with significant decreases in cerebral SaO2 on carotid crossclamping. This finding may be explained by (1) the presence of adequate collateral circulation and (2) maintenance of normal to slightly elevated SBP and MAP. In these patients, we did not pharmacologically elevate SBP during cross-clamp or attempt to tightly regulate cross-clamp SBP to the patient’s baseline SBP. Although two patients developed symptoms of global cerebral ischemia accompanied by substantial decreases in cerebral SaO2, these changes may be attributed to transient hypotension leading to decreased CBF. Levy et al.19 examined near-infrared oximetry and EEG during ventricular fibrillation and concluded that near-infrared measurements reflect changes in cerebral oxygenation as indicated by EEG evidence of brain ischemia. Although this model probably maintains bone and soft tissues in a normoxic state, using EEG to detect cerebral ischemia is not as accurate as neurologic testing in the awake patient. In addition, with short bursts of ventricular fibrillation, interpretation of the EEG is difficult. The carotid occlusion model also maintains overlying tissue and bone in a normoxic state, implying that infrared-derived SaO2 reflects the underlying cerebral hypoxia. Although carotid occlusion may be associated with variable ischemia, it is a more practical model, especially in the setting of head and neck surgery. Also, neurologic examination in the awake patient allows an absolute determination of cerebral ischemia and will more reliably define the SaO2 threshold necessary to create brain failure. In summary, we used near-infrared cerebral oximetry in patients undergoing CEA while awake in an attempt to define the SaO2 threshold necessary to create regional ischemic brain dysfunction. Although there was no evidence of cerebral ischemia when mean cerebral SaO2 decreased to 63% (7.4% reduction from baseline), there was evidence of global cerebral ischemia during hypotension in two patients when SaO2 decreased to 48% in one patient and 40% in the other. We believe that awake CEA is a good model to use in determining the clinical utility of cerebral oximetry. However, further studies in the setting of awake CEA are needed to define the critical SaO2 threshold, which may better predict the need for temporary shunting.
Cerebral oximetry in awake CAE: Carlin et al.
Addendum The cost of the equipment used in this study, as incurred by our institution, was approximately $50.00 per sensor [we purchased the sensors, and the manufacturer (Somanetics) supplied the cerebral oximeter]. Total cost of this equipment was approximately $800.00.
Acknowledgments The authors wish to thank Reza Gorji, MD, and Enrico Camporesi, MD, for their support in recording intraoperative oximetric data.
References 1. Clinical Advisory: carotid endarterectomy for patients with asymptomatic internal carotid artery stenosis. Stroke 1994;25:2523–4. 2. Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade stenosis. North American Symptomatic Carotid Endarterectomy Trial Collaborators. N Engl J Med 1991; 325:445–53. 3. MRC European Carotid Surgery Trial: interim results from symptomatic patients with severe (70 –99%) or with mild (0 – 29%) carotid stenosis. European Carotid Surgery Trialists’ Collaborators Group. Lancet 1991;337(8752):1235– 43. 4. Larson CP Jr, Ehrenfeld WK, Wade JG, Wylie EJ: Jugular venous oxygen saturation as an index of adequacy of cerebral oxygenation. Surgery 1967;62:31–9. 5. Bass A, Krupski WC, Schneider PA, Otis SM, Dilley RB, Bernstein EF: Intraoperative transcranial Doppler: limitations of the method. J Vasc Surg 1989;10:549 –53. 6. Horsch S, De Vleeschauwer P, Ktenidis K: Intraoperative assessment of cerebral ischemia during carotid surgery. J Cardiovascular Surg 1990;31:599 – 602. 7. Green RM, Messick WJ, Ricotta JJ, et al: Benefits, shortcomings, and costs of EEG monitoring. Ann Surg 1985;201:785–91. 8. Evans WE, Hayes JP, Waltke EA, Vermilion BD: Optimal cerebral monitoring during carotid endarterectomy: neurologic response under local anesthesia. J Vasc Surg 1985;2:775–7.
9. Slavin KV, Dujovny M, Ausman JI, Hernandez G, Luer M, Stoddart H: Clinical experience with transcranial cerebral oximetry. Surg Neurol 1994;42:531– 40. 10. McCormick PW, Stewart M, Goetting MG, Dujovny M, Lewis G, Ausman JI: Noninvasive cerebral optical spectroscopy for monitoring cerebral oxygen delivery and hemodynamics. Crit Care Med 1991;19:89 –97. 11. McCormick PW, Stewart M, Goetting MG, Balakrishnan G: Regional cerebrovascular oxygen saturation measured by optical spectroscopy in humans. Stroke 1991;22:596 – 602. 12. Williams IM, Vohra R, Farrell A, Picton AJ, Mortimer AJ, McCollum CN: Cerebral oxygen saturation, transcranial Doppler ultrasonography and stump pressure in carotid surgery. Br J Surg 1994;81:960 – 4. 13. Williams IM, Picton A, Farrell A, Mead GE, Mortimer AJ, McCollum CN: Light-reflective cerebral oximetry and jugular bulb venous oxygen saturation during carotid endarterectomy. Br J Surg 1994;81:1291–5. 14. Samra SK, Dorje P, Zelenock GB, Stanley JC: Cerebral oximetry in patients undergoing carotid endarterectomy under regional anesthesia. Stroke 1996;27:49 –55. 15. Jobsis FF: Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science 1977;198:1264 –7. 16. Brazy JE, Lewis DV, Mitnick MH, Jobsis van der Vliet FF: Noninvasive monitoring of cerebral oxygenation in preterm infants: preliminary observations. Pediatrics 1985;75: 217–25. 17. Wyatt JS, Cope M, Delpy DT, Wray S, Reynolds EO: Qualification of cerebral oxygenation and hemodynamics in sick newborn infants by near infrared spectrophotometry. Lancet 1986;2(8515): 1063– 6. 18. Mchedlishvili GI: Arterial Behavior and Blood Circulation in the Brain. New York, Plenum Press, 1986:56 –7. 19. Levy WJ, Levin S, Chance B: Near-infrared measurement of cerebral oxygenation. Correlation with electroencephalographic ischemia during ventricular fibrillation. Anesthesiology 1995;83: 738 – 46.
J. Clin. Anesth., vol. 10, March 1998
113