- Email: [email protected]
Correlation Between Cerebral and Mixed Venous Oxygen Saturation During Moderate Versus Tepid Hypothermic Hemodiluted Cardiopulmonary Bypass
Correlation Between Cerebral and Mixed Venous Oxygen Saturation During Moderate Versus Tepid Hypothermic Hemodiluted Cardiopulmonary Bypass Anis Baraka, MD, FRCA, Maud Naufal, MD, and Mohamad El-Khatib, PhD, FAARC Objective: This study was undertaken to compare cerebral oxygen saturation (RsO2) and mixed venous oxygen saturation (SvO2) in patients undergoing moderate and tepid hypothermic hemodiluted cardiopulmonary bypass (CPB). Design: Prospective study. Settings: University hospital operating room. Participants: Fourteen patients undergoing elective coronary artery bypass graft surgery using hypothermic hemodiluted CPB. Interventions: During moderate (28°-30°C) and tepid hypothermic (33°-34°C) hemodiluted CPB, RsO2 and SvO2 were continuously monitored with a cerebral oximeter via a surface electrode placed on the patient’s forehead and with the mixed venous oximeter integrated in the CPB machine, respectively. Measurements and Main Results: Mean ⴞ standard deviation of RsO2, SvO2, PaCO2, and hematocrit were determined prebypass and during moderate and tepid hypothermic phases of CPB while maintaining pump flow at 2.4 L/min/m2 and mean arterial pressure in the 60- to 70-mmHg range. Compared with a prebypass value of 76.0% ⴞ 9.6%, RsO2 was significantly decreased during moderate hypothermia to 58.9% ⴞ 6.4% and increased to 66.4% ⴞ 6.7% after slow rewarming to tepid hypothermia. In contrast, compared with a prebypass value of 78.6% ⴞ 3.3%, SvO2 significantly increased to 84.9% ⴞ 3.6% during moderate hypothermia and decreased to 74.1% ⴞ 5.6% during tepid
hypothermia. During moderate hypothermia, there was poor agreement between RsO2 and SvO2 with a gradient of 26%; however, during tepid hypothermia, there was a strong agreement between RsO2 and SvO2 with a gradient of 6%. The temperature-uncorrected PaCO2 was maintained at the normocapnic level throughout the study, whereas the temperature-corrected PaCO2 was significantly lower during the moderate hypothermic phase (26.8 ⴞ 3.1 mmHg) compared with the tepid hypothermic phase (38.9 ⴞ 3.7 mmHg) of CPB. There was a significant and positive correlation between RsO2 and temperature-corrected PaCO2 during hypothermia. Conclusions: During moderate hypothermic hemodiluted CPB, there was a significant increase of SvO2 associated with a paradoxic decrease of RsO2 that was attributed to the low temperature-corrected PaCO2 values. During tepid CPB after slow rewarming, regional cerebral oxygen saturation was increased in association with an increase with the temperature-corrected PaCO2 values. The results show that during hypothermic hemodiluted CPB using the alpha-stat strategy for carbon dioxide homeostasis, cerebral oxygen saturation is significantly higher during tepid than moderate hypothermia. © 2006 Elsevier Inc. All rights reserved.
D
oxygenation of the whole brain12 and partial brain tissue oxygen pressure.6 In addition, data are not clear as to how low the cerebral oxygen saturation can fall before clinicians should become concerned; hence, a baseline saturation should be taken when the patient is still awake and resting. Previous studies have reported the lack of correlation between SvO2 and regional cerebral oxygen saturation (RsO2) in animals and pediatric patients.2 The current study was designed to assess the correlation between cerebral oxygen saturation as monitored by the INVOS 5100 and SvO2 during moderate versus tepid hypothermic hemodiluted CPB in adult patients undergoing coronary artery bypass graft surgery.
URING CARDIOPULMONARY BYPASS (CPB), routine monitoring of mixed venous oxygen saturation (SvO2) is used to reflect the whole-body oxygen supply-demand balance.1 However, it might not reflect the adequacy of specific end-organ perfusion such as the brain.2 Cerebral hypoxia may result in brain injury and subsequent neurologic and neuropsychologic dysfunction.3-5 The INVOS 5100 cerebral oximeter (Tyco Healthcare, Mansfield, MA) is a device based on near-infrared spectroscopy and provides noninvasive and continuous real-time monitoring of cerebral oxygen saturation that indirectly assesses the cerebral oxygen-supply balance.6-8 It is based on the principle of transmission of 2 wavelengths of near-infrared light through tissues and absorption by oxygenated versus deoxygenated hemoglobin.8,9 However, it does have limitations because it detects regional cerebral oxygenation.8 With electrodes placed on the patient’s forehead, the INVOS 5100 signal might be affected by the patient’s skin and skull. However, because this device uses 2 photodetectors, the saturation displayed on the monitor is the difference between the skin and skull and the cerebral measurement. This excludes extracranial contamination and isolates changes in oxygen saturation in the brain.10,11 In addition, the optical path length is difficult to quantify and does not distinguish arterial from venous changes. However, the majority of blood in the brain is venous, and hence changes in the INVOS values reflect predominantly the SvO2, which predicts the critical balance between cerebral oxygen delivery and consumption.8 The oxygen saturation measured by the near-infrared spectroscope is closely related to the oxygen saturation in the jugular bulb, which represents the venous
KEY WORDS: cerebral oximetry, hypothermia, hemodilution, cardiopulmonary bypass, alpha-stat
MATERIALS AND METHODS The study was approved by the institutional review board at the American University of Beirut. Because the study involved no alterations in routine patient management, the institutional review board waived the need for a prior informed consent. Fourteen patients (American Society of Anesthesiologists II-III; mean age, 61 ⫾ 4.2 years; mean weight, 77 ⫾ 8.8 kg) of both genders who were undergoing elective coronary artery bypass graft surgery with hemodilution and
From the Department of Anesthesiology, School of Medicine, American University of Beirut, Beirut, Lebanon. Address reprint requests to Anis Baraka, MD, FRCA, Department of Anesthesiology, American University of Beirut, PO Box 11-0236, Beirut, Lebanon. E-mail: [email protected] © 2006 Elsevier Inc. All rights reserved. 1053-0770/06/2006-0012$32.00/0 doi:10.1053/j.jvca.2005.04.015
Journal of Cardiothoracic and Vascular Anesthesia, Vol 20, No 6 (December), 2006: pp 819-825
819
820
BARAKA, NAUFAL, AND EL-KHATIB Table 1. Mean ⴞ SD of RsO2, SvO2, PaCO2, and Hematocrit Precardiopulmonary Bypass and During the Moderate and Tepid Phases of CPB Prebypass
Moderate Hypothermia 28-30°C
Tepid Hypothermia 33-34°C
RsO2 (%) SvO2 (%) PaCO2 (mmHg)
76.0 ⫾ 9.6 78.6 ⫾ 3.3 43.5 ⫾ 8.9
Hct (%) CO (L/min) Pump Flow (L/min/m2)
38.2 ⫾ 5.1 4.6 ⫾ 0.9 NA
58.9 ⫾ 6.4* 84.9 ⫾ 3.6* Temp. Corrected: 26.8 ⫾ 3.1* Temp. Uncorrected: 38.9 ⫾ 3.7 27.7 ⫾ 4.1* NA 2.4
66.4 ⫾ 6.7*† 74.1 ⫾ 5.6*† Temp. Corrected: 37.9 ⫾ 3.1‡ Temp. Uncorrected: 42.1 ⫾ 2.4‡ 28.5 ⫾ 3.8* NA 2.4
Abbreviation: NA, not applicable. *p ⬍ 0.05 versus prebypass. †p ⬍ 0.05 versus moderate hypothermia. ‡p ⬍ 0.05 versus moderate hypothermia (temperature corrected).
hypothermic CPB using a membrane oxygenator were included in the study. Patients were premedicated with 10 mg of oral diazepam. In all patients, while awake and throughout the whole study, cerebral oxygen saturation was monitored by a cerebral oximeter (INVOS 5100, Somanetics) using disposable sensors placed on the patients’ forehead. Anesthesia was induced with midazolam (2 mg), thiopental (3 mg/ kg), lidocaine (2 mg/kg), sufentanil (25 g), and rocuronium (1.2 mg/kg) followed by orotracheal intubation and positive-pressure ventilation. Anesthesia was maintained with subsequent doses of sufentanil (1 g/kg/h), midazolam (0.1 g/kg/h), cisatracurium (0.1 mg/kg/h); and intermittent positive-pressure ventilation was initiated using 100% oxygen supplemented with 1% to 2% isoflurane to achieve an end-tidal CO2 in the 35- to 40-mmHg range. Patients were monitored with a 5-lead electrocardiogram, a radial artery cannula, and a pulmonary artery catheter. Before CPB, patients were given Ringer’s lactate at a rate of 10 mL/kg and an additional 1,500 mL was used to prime the membrane oxygenator (Medtronic, Minneapolis, MN). No blood or colloid was added to the prime. After full heparinization (heparin of 400 U/kg to achieve an activated coagulation time ⬎450 seconds), CPB was started using a nonpulsatile roller pump (Sarns 8000; 3M Health Group, Ann Arbor, MI) at a flow of 2.4 L/min/m2. Oxygen flow delivered to the membrane oxygenator was adjusted according to the blood gas results using alpha-stat strategy for carbon dioxide homeostasis. Mean arterial pressure (MAP) during CPB was maintained between 60 to 70 mmHg using incremental doses of phenylephrine whenever needed. The site for SvO2 measured by the Bentley Oxy-Stat Meter (American Bentley, Irvine, CA) and for body temperature was the venous blood at the entrance of the pump oxygenator. Body temperature was gradually decreased to 28° to 30°C, and the heart was arrested after aortic cross-clamping using cold crystalloid cardioplegia. After 20 minutes, the patients were slowly rewarmed to a temperature of 33° to 34°C. The surgeon started working on the coronary arteries immediately after establishing CPB and cross-clamping of the aorta and continued his work during and after rewarming. RsO2, SvO2, MAP, hematocrit, and arterial blood gases were recorded before CPB and during the moderate and tepid hypothermic phases of CPB. Temperature equilibration for 10 minutes was allowed at each measurement interval. Arterial blood gases were immediately subjected to duplicated measurements by a bench blood gas analyzer (ABL700; Radiometer, Copenhagen, Denmark), and both temperature-corrected and temperature-uncorrected PaCO2 values were determined. At the end of the surgical procedure, all patients were slowly rewarmed to 37°C and heparin was neutralized with protamine. A power analysis using type I and type II errors of 5% and 10%, respectively, considering a clinically significant change in RsO2 of 7.5% and a standard deviation of 8.5% as reported in previous studies,
revealed that 14 patients were needed for this study. The mean and standard deviation of RsO2, SvO2, temperature-corrected and temperature-uncorrected partial pressure of arterial carbon dioxide, and hematocrit levels were determined pre-CPB and during moderate (28°30°C) and tepid (33°-34°C) hypothermic phases of the CPB. These mean values were compared with the analysis of variance and student t test. Also, the degrees of agreement between RsO2 and SvO2 during moderate and tepid hypothermic phases of the CPB were determined with the Bland-Altman analysis.13 Statistical significance was considered at p ⬍ 0.05. RESULTS
Mean ⫾ standard deviation of RsO2, SvO2, PaCO2, and hematocrit pre-CPB and during the moderate (28°-30°C) and tepid phases (33°-34°C) of CPB are presented in Table 1. The hematocrit level significantly decreased from the prebypass value of 38.2% ⫾ 5.1% to 27.7% ⫾ 4.1% during the moderate hypothermic phase and to 28.5% ⫾ 3.8% during the tepid hypothermic phase of CPB. There was no significant difference between the hematocrit levels during moderate and tepid hypothermia (Table 1). The values of SvO2 and RsO2 during the prebypass period and during moderate and tepid hypothermic CPB phases are presented in Figure 1. RsO2 was significantly decreased during the moderate hypothermic phase (58.9% ⫾ 6.4%) compared with the prebypass phase (76.0% ⫾ 9.6%). During the tepid hypothermic phase, the RsO2 significantly increased to 66.4% ⫾ 6.7% compared with the moderate hypothermic phase. In contrast, SvO2 significantly increased during the moderate hypothermic phase (84.9% ⫾ 3.6%) compared with the prebypass phase (78.6% ⫾ 3.3%); during the tepid hypothermic phase, SvO2 significantly decreased to 74.1% ⫾ 5.6% compared with the moderate hypothermic phase. During the moderate hypothermic phase of CPB, there was no significant correlation (r ⫽ 0.07) between RsO2 and SvO2 (Fig 2). Furthermore, the Bland-Altman analysis revealed a poor agreement between RsO2 and SvO2 (Fig 3) with a gradient of 26%. After rewarming to the tepid hypothermic phase of CPB, the correlation between RsO2 and SvO2 improved (r ⫽ 0.14) (Fig 4) without reaching statistical significance. Also, the Bland-Altman analysis revealed a strong agreement between RsO2 and SvO2 (Fig 5) with a gradient of 6%.
CEREBRAL AND VENOUS OXYGEN SATURATION
821
Fig 1. Mean ⴞ standard deviation of SvO2 and RsO2 during the prebypass period and during moderate and tepid hypothermic CPB phases. *p < 0.05 versus prebypass; ¶p < 0.05 versus moderate hypothermia.
The PaCO2 was 43.5 ⫾ 8.9 mmHg prebypass. Going on CPB, the temperature-uncorrected PaCO2 remained unchanged throughout the study, whereas the temperature-corrected PaCO2 decreased to 26.8 ⫾ 3.1 mmHg during the moderate
Fig 2.
hypothermic phase and then increased to 37.9 ⫾ 3.1 during the tepid hypothermic phase. Also, there was a significant positive correlation between the RsO2 and the temperature-corrected PaCO2 during CPB (Fig 6).
Correlation between RsO2 and SvO2 during moderate hypothermic hemodiluted CPB (r ⴝ 0.07, not significant).
822
BARAKA, NAUFAL, AND EL-KHATIB
Fig 3.
Difference in SvO2 and RsO2 against the mean of the 2 measurements during the moderate hypothermic CPB phase.
Fig 4. Correlation between RsO2 and SvO2 during the tepid hypothermic hemodiluted CPB (r ⴝ 0.14, not significant).
CEREBRAL AND VENOUS OXYGEN SATURATION
Fig 5. Difference in SvO2 and RsO2 against the mean of the 2 measurements during the tepid hypothermic CPB phase.
Fig 6.
Correlation between RsO2 and temperature-corrected PaCO2) during the hypothermic hemodiluted CPB (r ⴝ 0.41, p < 0.05).
823
824
BARAKA, NAUFAL, AND EL-KHATIB
DISCUSSION
The current report confirms previous findings that in patients undergoing hemodiluted hypothermic CPB, SvO2, which reflects the whole-body oxygen supply-demand relationship, is significantly increased during moderate hypothermia (28°30°C) and is gradually decreased by slow rewarming to tepid hypothermia (33°-34°C).1 These results were previously attributed to the decrease of whole-body oxygen consumption by cooling according to the Q10 principle that reflects the variation in metabolic activity produced by a 10°C change in temperature.14 In contrast to the SvO2 findings, the present report shows that RsO2, which reflects the regional cerebral oxygen supply-demand relationship, was significantly decreased on CPB at 28° to 30°C (moderate hypothermia) as compared with the prebypass level. However, after slow rewarming to 33° to 34°C (tepid hypothermia) with the patients still on bypass, the RsO2 significantly increased. In these patients, slow rewarming to 33° to 34°C while still on bypass resulted in a significant increase of RsO2, despite maintenance of the pump flow, MAP, and the hematocrit at the same levels. Also, it has been previously shown that hemodilution is associated with decreased viscosity and cerebral vasodilatation, which can partially offset the effect of hemodilution,15 suggesting that another factor, such as the PaCO2, might have contributed to the low RsO2 during the moderate hypothermic CPB phase. Murkin et al,16 using the clearance of xenon radioisotope and a cerebrograph for measurement of cerebral blood flow, showed that the alpha-stat strategy for carbon dioxide homeostasis during hypothermia maintains cerebral autoregulation and perfusion-metabolic coupling. In the present patients, the authors used the alpha-stat strategy of carbon dioxide homeostasis as evidenced by the maintenance of the temperature-uncorrected PaCO2 at the normocapnic levels throughout CPB, whereas the temperature-corrected PaCO2 was significantly lower (26.8 ⫾ 3.1 mmHg) during the moderate hypothermic phase of CPB and significantly increased (37.9 ⫾ 3.1 mmHg) by slow rewarming to 33° to 34°C. PaCO2 is a major determinant of the cerebral blood flow, and hypocarbia can significantly decrease the cerebral blood flow and subsequently decrease the cerebral oxygen supply. However, Czinn et al17 have shown in an experimental model of dogs anesthetized with halothane and using radioactive microspheres for measurement of cerebral blood flow that severe hemodilution (he-
matocrit ⬍20%) either attenuates or completely abolishes the vasoconstrictor responses within the brain and spinal cord during hypocapnia. In the present study, only moderate hemodilution to a hematocrit level ⬎25% was maintained throughout the hemodiluted hypothermic CPB and was associated with significant correlation between the RsO2 and the temperature-corrected PaCO2 values. The significant decrease of temperature-corrected PaCO2 secondary to the alpha-stat strategy during moderate hypothermia might have resulted in a decrease of cerebral blood flow with a subsequent decrease of cerebral oxygen supply. This decreased cerebral oxygen supply could have been aggravated by both hypothermia and hypocarbia, which shift the oxyhemoglobin dissociation curve to the left and increase the difficulty in unloading the oxygen. Slow rewarming to 33° to 34°C was associated with an increase of the temperature-corrected PaCO2, which can result in cerebral vasodilatation and thus can explain the significant increase in the observed RsO2. These results show that during hypothermic hemodiluted CPB using a pump flow of 2.4 L/min/m2 and adopting the alpha-stat strategy for carbon dioxide homeostasis, mild hypothermia to the 33° to 34°C temperature range is associated with a higher RsO2 than that achieved during moderate hypothermia in the 28° to 30°C temperature range. Previous studies have shown during a similar setup of CPB that a 2° to 3°C temperature decrease confers significant protection in the setting of transient cerebral ischemia and that active cooling to 28° to 30°C does not provide any additional benefits.18 Thus, it would seem a reasonable practice to allow the temperature to “drift” during CPB by 2° to 3°C only, which will offer brain protection while avoiding prolonged rewarming and bypass times and eliminating the possibility of cerebral hyperthermia resulting from rewarming from a lower temperature range.18-20 In conclusion, the current report suggests that cerebral oximetry by near-infrared spectroscopy is valuable for noninvasive continuous monitoring of the cerebral oxygen supply-demand relationship during CPB. The results show that during hypothermic hemodiluted CPB using the alpha-stat strategy, RsO2 was significantly higher during tepid hypothermia (33°-34°C) as compared with moderate hypothermia (28°-30°C). ACKNOWLEDGMENT The authors are grateful to TYCO Healthcare for providing the INVOS 5100 cerebral oximeter and the disposable sensors.
REFERENCES 1. Baraka A, Baroody M, Haroun S, et al: Continuous venous oximetry during cardiopulmonary bypass: Influence of temperature changes, perfusion flow, and hematocrit levels. J Cardiothorac Vasc Anesth 1:35-38, 1992 2. Yeh T, Gouldman J, Auden S, et al: Mixed venous saturation does not adequately predict cerebral perfusion during pediatric cardiopulmonary bypass. J Thorac Cardiovasc Surg 122:192-193, 2001 3. Roach GW, Kanchuger M, Mangano CM, et al: Adverse cerebral outcomes after coronary bypass surgery. Multicenter Study of Perioperative Ischemia Research Group and the Ischemia Research and Education Foundation Investigators. N Engl J Med 335:18571863, 1996
4. Newman MF, Kirchner JL, Phillips-Bute B, et al: Longitudinal assessment of neurocognitive function after coronary artery bypass surgery. N Engl J Med 344:395-402, 2001 5. Richard FM, Bill IW: Normothermic versus hypothermic cardiopulmonary bypass: Central nervous system outcome. J Cardiothorac Vasc Anesth 10:45-53, 1996 6. Holzschuh M, Woertgen C, Metz C, et al: Dynamic changes of cerebral oxygenation measured by brain tissue oxygen pressure and near-infrared spectroscopy. Neurol Res 19:246-248, 1997 7. Edmonds HL, Rodriguez RA, Audenaert SM, et al: The role of neuromonitoring in cardiovascular surgery. J Cardiothorac Vasc Anesth 10:15-23, 1996
CEREBRAL AND VENOUS OXYGEN SATURATION
8. Lozano S, Mossad E: Cerebral function monitors during pediatric cardiac surgery: Can they make a difference? J Cardiothorac Vasc Anesth 18:645-656, 2004 9. Yao FS, Tseng CC, Ho CY, et al: Cerebral oxygen desaturation is associated with early postoperative neuropsychological dysfunction in patients undergoing cardiac surgery. J Cardiothorac Vasc Anesth 18: 552-558, 2004 10. Jobsis FF: Noninvasive, near-infrared monitoring of cellular sufficiency in vivo. Adv Exp Med Biol 191:833-841, 1985 11. Samra SK, Stanley JC, Zelenock GB, et al: An assessment of contributions made by extracranial tissues during cerebral oximetry. J Neurosurg Anesth 11:1-5, 1999 12. Daubeney PE, Pilkington SN, Janke E, et al: Cerebral oxygenation measured by near-infrared spectroscopy: Comparison with jugular bulb oximetry. Ann Thorac Surg 61:930-934, 1996 13. Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet I:307-310, 1986 14. Larach DR, High KM, Derr JA, et al: Carbon dioxide elimination during total cardiopulmonary bypass in infants and children. Anesthesiology 69:185-191, 1988
825
15. Duebener LF, Hagino I, Schmitt K, et al: Effects of hemodilution and phenylephrine on cerebral blood flow and metabolism during cardiopulmonary bypass. J Cardiothorac Vasc Anesth 18:423-428, 2004 16. Murkin JM, Farrar JK, Tweed WA, et al: Cerebral autoregulation and flow/metabolism coupling during cardiopulmonary bypass: The influence of PaCO2. Anesth Analg 66:825-832, 1987 17. Czinn EA, Salem MR, Crystal GJ: Hemodilution impairs hypocapnia-induced vasoconstrictor responses in the brain and spinal cord in dogs. Anesth Analg 80:492-498, 1995 18. McLean RF, Wong BI: Normothermic versus hypothermic cardiopulmonary bypass: central nervous system outcomes. J Cardiothorac Vasc Anesth 10:45-53, 1996 19. Cook DJ, Orszulak TA, Daly RC, et al: Cerebral hyperthermia during cardiopulmonary bypass in adults. J Thorac Cardiovasc Surg 111:672-676, 1996 20. Shaaban A, Harmer M, Vaughn RS, et al: Changes in cerebral oxygenation during cold (28°C) and warm (34°C) cardiopulmonary bypass using different blood gas strategies (alpha-stat and pH-stat) in patients undergoing coronary artery bypass graft surgery. Acta Anaesthesiol Scand 48:837-844, 2004
hypothermia. During moderate hypothermia, there was poor agreement between RsO2 and SvO2 with a gradient of 26%; however, during tepid hypothermia, there was a strong agreement between RsO2 and SvO2 with a gradient of 6%. The temperature-uncorrected PaCO2 was maintained at the normocapnic level throughout the study, whereas the temperature-corrected PaCO2 was significantly lower during the moderate hypothermic phase (26.8 ⴞ 3.1 mmHg) compared with the tepid hypothermic phase (38.9 ⴞ 3.7 mmHg) of CPB. There was a significant and positive correlation between RsO2 and temperature-corrected PaCO2 during hypothermia. Conclusions: During moderate hypothermic hemodiluted CPB, there was a significant increase of SvO2 associated with a paradoxic decrease of RsO2 that was attributed to the low temperature-corrected PaCO2 values. During tepid CPB after slow rewarming, regional cerebral oxygen saturation was increased in association with an increase with the temperature-corrected PaCO2 values. The results show that during hypothermic hemodiluted CPB using the alpha-stat strategy for carbon dioxide homeostasis, cerebral oxygen saturation is significantly higher during tepid than moderate hypothermia. © 2006 Elsevier Inc. All rights reserved.
D
oxygenation of the whole brain12 and partial brain tissue oxygen pressure.6 In addition, data are not clear as to how low the cerebral oxygen saturation can fall before clinicians should become concerned; hence, a baseline saturation should be taken when the patient is still awake and resting. Previous studies have reported the lack of correlation between SvO2 and regional cerebral oxygen saturation (RsO2) in animals and pediatric patients.2 The current study was designed to assess the correlation between cerebral oxygen saturation as monitored by the INVOS 5100 and SvO2 during moderate versus tepid hypothermic hemodiluted CPB in adult patients undergoing coronary artery bypass graft surgery.
URING CARDIOPULMONARY BYPASS (CPB), routine monitoring of mixed venous oxygen saturation (SvO2) is used to reflect the whole-body oxygen supply-demand balance.1 However, it might not reflect the adequacy of specific end-organ perfusion such as the brain.2 Cerebral hypoxia may result in brain injury and subsequent neurologic and neuropsychologic dysfunction.3-5 The INVOS 5100 cerebral oximeter (Tyco Healthcare, Mansfield, MA) is a device based on near-infrared spectroscopy and provides noninvasive and continuous real-time monitoring of cerebral oxygen saturation that indirectly assesses the cerebral oxygen-supply balance.6-8 It is based on the principle of transmission of 2 wavelengths of near-infrared light through tissues and absorption by oxygenated versus deoxygenated hemoglobin.8,9 However, it does have limitations because it detects regional cerebral oxygenation.8 With electrodes placed on the patient’s forehead, the INVOS 5100 signal might be affected by the patient’s skin and skull. However, because this device uses 2 photodetectors, the saturation displayed on the monitor is the difference between the skin and skull and the cerebral measurement. This excludes extracranial contamination and isolates changes in oxygen saturation in the brain.10,11 In addition, the optical path length is difficult to quantify and does not distinguish arterial from venous changes. However, the majority of blood in the brain is venous, and hence changes in the INVOS values reflect predominantly the SvO2, which predicts the critical balance between cerebral oxygen delivery and consumption.8 The oxygen saturation measured by the near-infrared spectroscope is closely related to the oxygen saturation in the jugular bulb, which represents the venous
KEY WORDS: cerebral oximetry, hypothermia, hemodilution, cardiopulmonary bypass, alpha-stat
MATERIALS AND METHODS The study was approved by the institutional review board at the American University of Beirut. Because the study involved no alterations in routine patient management, the institutional review board waived the need for a prior informed consent. Fourteen patients (American Society of Anesthesiologists II-III; mean age, 61 ⫾ 4.2 years; mean weight, 77 ⫾ 8.8 kg) of both genders who were undergoing elective coronary artery bypass graft surgery with hemodilution and
From the Department of Anesthesiology, School of Medicine, American University of Beirut, Beirut, Lebanon. Address reprint requests to Anis Baraka, MD, FRCA, Department of Anesthesiology, American University of Beirut, PO Box 11-0236, Beirut, Lebanon. E-mail: [email protected] © 2006 Elsevier Inc. All rights reserved. 1053-0770/06/2006-0012$32.00/0 doi:10.1053/j.jvca.2005.04.015
Journal of Cardiothoracic and Vascular Anesthesia, Vol 20, No 6 (December), 2006: pp 819-825
819
820
BARAKA, NAUFAL, AND EL-KHATIB Table 1. Mean ⴞ SD of RsO2, SvO2, PaCO2, and Hematocrit Precardiopulmonary Bypass and During the Moderate and Tepid Phases of CPB Prebypass
Moderate Hypothermia 28-30°C
Tepid Hypothermia 33-34°C
RsO2 (%) SvO2 (%) PaCO2 (mmHg)
76.0 ⫾ 9.6 78.6 ⫾ 3.3 43.5 ⫾ 8.9
Hct (%) CO (L/min) Pump Flow (L/min/m2)
38.2 ⫾ 5.1 4.6 ⫾ 0.9 NA
58.9 ⫾ 6.4* 84.9 ⫾ 3.6* Temp. Corrected: 26.8 ⫾ 3.1* Temp. Uncorrected: 38.9 ⫾ 3.7 27.7 ⫾ 4.1* NA 2.4
66.4 ⫾ 6.7*† 74.1 ⫾ 5.6*† Temp. Corrected: 37.9 ⫾ 3.1‡ Temp. Uncorrected: 42.1 ⫾ 2.4‡ 28.5 ⫾ 3.8* NA 2.4
Abbreviation: NA, not applicable. *p ⬍ 0.05 versus prebypass. †p ⬍ 0.05 versus moderate hypothermia. ‡p ⬍ 0.05 versus moderate hypothermia (temperature corrected).
hypothermic CPB using a membrane oxygenator were included in the study. Patients were premedicated with 10 mg of oral diazepam. In all patients, while awake and throughout the whole study, cerebral oxygen saturation was monitored by a cerebral oximeter (INVOS 5100, Somanetics) using disposable sensors placed on the patients’ forehead. Anesthesia was induced with midazolam (2 mg), thiopental (3 mg/ kg), lidocaine (2 mg/kg), sufentanil (25 g), and rocuronium (1.2 mg/kg) followed by orotracheal intubation and positive-pressure ventilation. Anesthesia was maintained with subsequent doses of sufentanil (1 g/kg/h), midazolam (0.1 g/kg/h), cisatracurium (0.1 mg/kg/h); and intermittent positive-pressure ventilation was initiated using 100% oxygen supplemented with 1% to 2% isoflurane to achieve an end-tidal CO2 in the 35- to 40-mmHg range. Patients were monitored with a 5-lead electrocardiogram, a radial artery cannula, and a pulmonary artery catheter. Before CPB, patients were given Ringer’s lactate at a rate of 10 mL/kg and an additional 1,500 mL was used to prime the membrane oxygenator (Medtronic, Minneapolis, MN). No blood or colloid was added to the prime. After full heparinization (heparin of 400 U/kg to achieve an activated coagulation time ⬎450 seconds), CPB was started using a nonpulsatile roller pump (Sarns 8000; 3M Health Group, Ann Arbor, MI) at a flow of 2.4 L/min/m2. Oxygen flow delivered to the membrane oxygenator was adjusted according to the blood gas results using alpha-stat strategy for carbon dioxide homeostasis. Mean arterial pressure (MAP) during CPB was maintained between 60 to 70 mmHg using incremental doses of phenylephrine whenever needed. The site for SvO2 measured by the Bentley Oxy-Stat Meter (American Bentley, Irvine, CA) and for body temperature was the venous blood at the entrance of the pump oxygenator. Body temperature was gradually decreased to 28° to 30°C, and the heart was arrested after aortic cross-clamping using cold crystalloid cardioplegia. After 20 minutes, the patients were slowly rewarmed to a temperature of 33° to 34°C. The surgeon started working on the coronary arteries immediately after establishing CPB and cross-clamping of the aorta and continued his work during and after rewarming. RsO2, SvO2, MAP, hematocrit, and arterial blood gases were recorded before CPB and during the moderate and tepid hypothermic phases of CPB. Temperature equilibration for 10 minutes was allowed at each measurement interval. Arterial blood gases were immediately subjected to duplicated measurements by a bench blood gas analyzer (ABL700; Radiometer, Copenhagen, Denmark), and both temperature-corrected and temperature-uncorrected PaCO2 values were determined. At the end of the surgical procedure, all patients were slowly rewarmed to 37°C and heparin was neutralized with protamine. A power analysis using type I and type II errors of 5% and 10%, respectively, considering a clinically significant change in RsO2 of 7.5% and a standard deviation of 8.5% as reported in previous studies,
revealed that 14 patients were needed for this study. The mean and standard deviation of RsO2, SvO2, temperature-corrected and temperature-uncorrected partial pressure of arterial carbon dioxide, and hematocrit levels were determined pre-CPB and during moderate (28°30°C) and tepid (33°-34°C) hypothermic phases of the CPB. These mean values were compared with the analysis of variance and student t test. Also, the degrees of agreement between RsO2 and SvO2 during moderate and tepid hypothermic phases of the CPB were determined with the Bland-Altman analysis.13 Statistical significance was considered at p ⬍ 0.05. RESULTS
Mean ⫾ standard deviation of RsO2, SvO2, PaCO2, and hematocrit pre-CPB and during the moderate (28°-30°C) and tepid phases (33°-34°C) of CPB are presented in Table 1. The hematocrit level significantly decreased from the prebypass value of 38.2% ⫾ 5.1% to 27.7% ⫾ 4.1% during the moderate hypothermic phase and to 28.5% ⫾ 3.8% during the tepid hypothermic phase of CPB. There was no significant difference between the hematocrit levels during moderate and tepid hypothermia (Table 1). The values of SvO2 and RsO2 during the prebypass period and during moderate and tepid hypothermic CPB phases are presented in Figure 1. RsO2 was significantly decreased during the moderate hypothermic phase (58.9% ⫾ 6.4%) compared with the prebypass phase (76.0% ⫾ 9.6%). During the tepid hypothermic phase, the RsO2 significantly increased to 66.4% ⫾ 6.7% compared with the moderate hypothermic phase. In contrast, SvO2 significantly increased during the moderate hypothermic phase (84.9% ⫾ 3.6%) compared with the prebypass phase (78.6% ⫾ 3.3%); during the tepid hypothermic phase, SvO2 significantly decreased to 74.1% ⫾ 5.6% compared with the moderate hypothermic phase. During the moderate hypothermic phase of CPB, there was no significant correlation (r ⫽ 0.07) between RsO2 and SvO2 (Fig 2). Furthermore, the Bland-Altman analysis revealed a poor agreement between RsO2 and SvO2 (Fig 3) with a gradient of 26%. After rewarming to the tepid hypothermic phase of CPB, the correlation between RsO2 and SvO2 improved (r ⫽ 0.14) (Fig 4) without reaching statistical significance. Also, the Bland-Altman analysis revealed a strong agreement between RsO2 and SvO2 (Fig 5) with a gradient of 6%.
CEREBRAL AND VENOUS OXYGEN SATURATION
821
Fig 1. Mean ⴞ standard deviation of SvO2 and RsO2 during the prebypass period and during moderate and tepid hypothermic CPB phases. *p < 0.05 versus prebypass; ¶p < 0.05 versus moderate hypothermia.
The PaCO2 was 43.5 ⫾ 8.9 mmHg prebypass. Going on CPB, the temperature-uncorrected PaCO2 remained unchanged throughout the study, whereas the temperature-corrected PaCO2 decreased to 26.8 ⫾ 3.1 mmHg during the moderate
Fig 2.
hypothermic phase and then increased to 37.9 ⫾ 3.1 during the tepid hypothermic phase. Also, there was a significant positive correlation between the RsO2 and the temperature-corrected PaCO2 during CPB (Fig 6).
Correlation between RsO2 and SvO2 during moderate hypothermic hemodiluted CPB (r ⴝ 0.07, not significant).
822
BARAKA, NAUFAL, AND EL-KHATIB
Fig 3.
Difference in SvO2 and RsO2 against the mean of the 2 measurements during the moderate hypothermic CPB phase.
Fig 4. Correlation between RsO2 and SvO2 during the tepid hypothermic hemodiluted CPB (r ⴝ 0.14, not significant).
CEREBRAL AND VENOUS OXYGEN SATURATION
Fig 5. Difference in SvO2 and RsO2 against the mean of the 2 measurements during the tepid hypothermic CPB phase.
Fig 6.
Correlation between RsO2 and temperature-corrected PaCO2) during the hypothermic hemodiluted CPB (r ⴝ 0.41, p < 0.05).
823
824
BARAKA, NAUFAL, AND EL-KHATIB
DISCUSSION
The current report confirms previous findings that in patients undergoing hemodiluted hypothermic CPB, SvO2, which reflects the whole-body oxygen supply-demand relationship, is significantly increased during moderate hypothermia (28°30°C) and is gradually decreased by slow rewarming to tepid hypothermia (33°-34°C).1 These results were previously attributed to the decrease of whole-body oxygen consumption by cooling according to the Q10 principle that reflects the variation in metabolic activity produced by a 10°C change in temperature.14 In contrast to the SvO2 findings, the present report shows that RsO2, which reflects the regional cerebral oxygen supply-demand relationship, was significantly decreased on CPB at 28° to 30°C (moderate hypothermia) as compared with the prebypass level. However, after slow rewarming to 33° to 34°C (tepid hypothermia) with the patients still on bypass, the RsO2 significantly increased. In these patients, slow rewarming to 33° to 34°C while still on bypass resulted in a significant increase of RsO2, despite maintenance of the pump flow, MAP, and the hematocrit at the same levels. Also, it has been previously shown that hemodilution is associated with decreased viscosity and cerebral vasodilatation, which can partially offset the effect of hemodilution,15 suggesting that another factor, such as the PaCO2, might have contributed to the low RsO2 during the moderate hypothermic CPB phase. Murkin et al,16 using the clearance of xenon radioisotope and a cerebrograph for measurement of cerebral blood flow, showed that the alpha-stat strategy for carbon dioxide homeostasis during hypothermia maintains cerebral autoregulation and perfusion-metabolic coupling. In the present patients, the authors used the alpha-stat strategy of carbon dioxide homeostasis as evidenced by the maintenance of the temperature-uncorrected PaCO2 at the normocapnic levels throughout CPB, whereas the temperature-corrected PaCO2 was significantly lower (26.8 ⫾ 3.1 mmHg) during the moderate hypothermic phase of CPB and significantly increased (37.9 ⫾ 3.1 mmHg) by slow rewarming to 33° to 34°C. PaCO2 is a major determinant of the cerebral blood flow, and hypocarbia can significantly decrease the cerebral blood flow and subsequently decrease the cerebral oxygen supply. However, Czinn et al17 have shown in an experimental model of dogs anesthetized with halothane and using radioactive microspheres for measurement of cerebral blood flow that severe hemodilution (he-
matocrit ⬍20%) either attenuates or completely abolishes the vasoconstrictor responses within the brain and spinal cord during hypocapnia. In the present study, only moderate hemodilution to a hematocrit level ⬎25% was maintained throughout the hemodiluted hypothermic CPB and was associated with significant correlation between the RsO2 and the temperature-corrected PaCO2 values. The significant decrease of temperature-corrected PaCO2 secondary to the alpha-stat strategy during moderate hypothermia might have resulted in a decrease of cerebral blood flow with a subsequent decrease of cerebral oxygen supply. This decreased cerebral oxygen supply could have been aggravated by both hypothermia and hypocarbia, which shift the oxyhemoglobin dissociation curve to the left and increase the difficulty in unloading the oxygen. Slow rewarming to 33° to 34°C was associated with an increase of the temperature-corrected PaCO2, which can result in cerebral vasodilatation and thus can explain the significant increase in the observed RsO2. These results show that during hypothermic hemodiluted CPB using a pump flow of 2.4 L/min/m2 and adopting the alpha-stat strategy for carbon dioxide homeostasis, mild hypothermia to the 33° to 34°C temperature range is associated with a higher RsO2 than that achieved during moderate hypothermia in the 28° to 30°C temperature range. Previous studies have shown during a similar setup of CPB that a 2° to 3°C temperature decrease confers significant protection in the setting of transient cerebral ischemia and that active cooling to 28° to 30°C does not provide any additional benefits.18 Thus, it would seem a reasonable practice to allow the temperature to “drift” during CPB by 2° to 3°C only, which will offer brain protection while avoiding prolonged rewarming and bypass times and eliminating the possibility of cerebral hyperthermia resulting from rewarming from a lower temperature range.18-20 In conclusion, the current report suggests that cerebral oximetry by near-infrared spectroscopy is valuable for noninvasive continuous monitoring of the cerebral oxygen supply-demand relationship during CPB. The results show that during hypothermic hemodiluted CPB using the alpha-stat strategy, RsO2 was significantly higher during tepid hypothermia (33°-34°C) as compared with moderate hypothermia (28°-30°C). ACKNOWLEDGMENT The authors are grateful to TYCO Healthcare for providing the INVOS 5100 cerebral oximeter and the disposable sensors.
REFERENCES 1. Baraka A, Baroody M, Haroun S, et al: Continuous venous oximetry during cardiopulmonary bypass: Influence of temperature changes, perfusion flow, and hematocrit levels. J Cardiothorac Vasc Anesth 1:35-38, 1992 2. Yeh T, Gouldman J, Auden S, et al: Mixed venous saturation does not adequately predict cerebral perfusion during pediatric cardiopulmonary bypass. J Thorac Cardiovasc Surg 122:192-193, 2001 3. Roach GW, Kanchuger M, Mangano CM, et al: Adverse cerebral outcomes after coronary bypass surgery. Multicenter Study of Perioperative Ischemia Research Group and the Ischemia Research and Education Foundation Investigators. N Engl J Med 335:18571863, 1996
4. Newman MF, Kirchner JL, Phillips-Bute B, et al: Longitudinal assessment of neurocognitive function after coronary artery bypass surgery. N Engl J Med 344:395-402, 2001 5. Richard FM, Bill IW: Normothermic versus hypothermic cardiopulmonary bypass: Central nervous system outcome. J Cardiothorac Vasc Anesth 10:45-53, 1996 6. Holzschuh M, Woertgen C, Metz C, et al: Dynamic changes of cerebral oxygenation measured by brain tissue oxygen pressure and near-infrared spectroscopy. Neurol Res 19:246-248, 1997 7. Edmonds HL, Rodriguez RA, Audenaert SM, et al: The role of neuromonitoring in cardiovascular surgery. J Cardiothorac Vasc Anesth 10:15-23, 1996
CEREBRAL AND VENOUS OXYGEN SATURATION
8. Lozano S, Mossad E: Cerebral function monitors during pediatric cardiac surgery: Can they make a difference? J Cardiothorac Vasc Anesth 18:645-656, 2004 9. Yao FS, Tseng CC, Ho CY, et al: Cerebral oxygen desaturation is associated with early postoperative neuropsychological dysfunction in patients undergoing cardiac surgery. J Cardiothorac Vasc Anesth 18: 552-558, 2004 10. Jobsis FF: Noninvasive, near-infrared monitoring of cellular sufficiency in vivo. Adv Exp Med Biol 191:833-841, 1985 11. Samra SK, Stanley JC, Zelenock GB, et al: An assessment of contributions made by extracranial tissues during cerebral oximetry. J Neurosurg Anesth 11:1-5, 1999 12. Daubeney PE, Pilkington SN, Janke E, et al: Cerebral oxygenation measured by near-infrared spectroscopy: Comparison with jugular bulb oximetry. Ann Thorac Surg 61:930-934, 1996 13. Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet I:307-310, 1986 14. Larach DR, High KM, Derr JA, et al: Carbon dioxide elimination during total cardiopulmonary bypass in infants and children. Anesthesiology 69:185-191, 1988
825
15. Duebener LF, Hagino I, Schmitt K, et al: Effects of hemodilution and phenylephrine on cerebral blood flow and metabolism during cardiopulmonary bypass. J Cardiothorac Vasc Anesth 18:423-428, 2004 16. Murkin JM, Farrar JK, Tweed WA, et al: Cerebral autoregulation and flow/metabolism coupling during cardiopulmonary bypass: The influence of PaCO2. Anesth Analg 66:825-832, 1987 17. Czinn EA, Salem MR, Crystal GJ: Hemodilution impairs hypocapnia-induced vasoconstrictor responses in the brain and spinal cord in dogs. Anesth Analg 80:492-498, 1995 18. McLean RF, Wong BI: Normothermic versus hypothermic cardiopulmonary bypass: central nervous system outcomes. J Cardiothorac Vasc Anesth 10:45-53, 1996 19. Cook DJ, Orszulak TA, Daly RC, et al: Cerebral hyperthermia during cardiopulmonary bypass in adults. J Thorac Cardiovasc Surg 111:672-676, 1996 20. Shaaban A, Harmer M, Vaughn RS, et al: Changes in cerebral oxygenation during cold (28°C) and warm (34°C) cardiopulmonary bypass using different blood gas strategies (alpha-stat and pH-stat) in patients undergoing coronary artery bypass graft surgery. Acta Anaesthesiol Scand 48:837-844, 2004