Bispectral index (BIS) has been developed as a monitor of hypnotic drug effect. It produces a scale from 0 to 100 by processing the electroencephalogram (EEG). A BIS of 100 indicates an awake and responsive subject. When hypnotics are administered, BIS decreases, and a BIS value of 0 represents EEG silence (1). During surgery, BIS may decrease to nearly 0 due to deep anesthesia, hypothermia, and cerebral ischemia. There have been several case reports (2–4) that have represented BIS to be an indicator of inadequate cerebral perfusion. We describe a patient in whom BIS decreased to 0 during general anesthesia, possibly because of cerebral hypoperfusion.
A 42-yr-old man with chronic renal dysfunction was scheduled to undergo construction of an arteriovenous shunt on his left forearm under general anesthesia. A month earlier, he had been found unconscious in his house and transferred to our hospital. Examination showed that the patient had an acute cerebral hemorrhage. Urgent craniotomy and removal of the hematoma were performed on the same day. He had untreated chronic renal dysfunction and hypertension. After surgical removal of the hematoma, his level of consciousness became almost normal, but his renal dysfunction was exacerbated, and hemodialysis was performed twice before the shunt surgery. Blood chemistry showed creatinine 8.1 mg/dL and blood urea nitrogen 80 mg/mL. Severe hypertension was also treated by several antihypertensive drugs; however, the systolic blood pressure before surgery was approximately 180 mm Hg. Computed tomography before surgery showed no residual hematoma in the brain. No neurologic deficit was observed.
He received nifedipine 20 mg by mouth (PO), carvedilol 20 mg PO, and doxazosin mesylate 1 mg PO on the day of surgery. An A-1050 monitor (Version 3.4; Aspect Medical Systems, Natick, MA) was applied in addition to the standard monitors. The EEG signal from the raw EEG port of the A-1050 was introduced into a personal computer with Microsoft Windows ME (Microsoft Corp., Redmond, WA). Then EEG, BIS, and burst suppression ratio (BSR) were recorded with EEG analysis software (BSA, Bispectral Analyzer) that we have developed and previously reported (5). Anesthesia was induced with propofol 100 mg and sevoflurane 3%. After a laryngeal mask was inserted anesthesia was maintained with nitrous oxide 60% in oxygen and sevoflurane 3% with spontaneous respiration.
Figure 1 shows the changes in BIS and mean arterial blood pressure during the surgery. Before the induction of anesthesia, BIS was 91 and mean arterial blood pressure was 140 mm Hg (systolic blood pressure, 180 mm Hg). After insertion of the laryngeal mask, BIS suddenly decreased to near 0, and BSR increased to more than 90. Spontaneous respiration disappeared; therefore, mechanical ventilation was started. Because overdose of anesthetics was suspected, the inspired sevoflurane concentration was decreased from 3% to 1%. However, BIS remained less than 20, and the BSR was more than 60. The electrode impedance was <3000 Ω. Although his systolic blood pressure and mean arterial blood pressure were approximately 110 and 80 mm Hg, respectively, etilefrine 2 mg IV was given to increase the blood pressure. As soon as the mean arterial blood pressure increased to 100 mm Hg (systolic blood pressure, 140 mm Hg), BIS increased to 40. When surgery started, BIS was 43, and anesthesia was maintained with 1.5% sevoflurane (inspired concentration). Blood pressure was 135 mm Hg (systolic) and 100 mm Hg (mean). During surgery, BIS decreased to 0 again, and BSR increased to 100 when systolic blood pressure and mean arterial blood pressure decreased to 110 and 80 mm Hg, respectively. The EEG showed an almost flat or burst suppression pattern at that time (Fig. 2). Etilefrine 2 mg IV and phenylephrine 0.1 mg IV were given to increase his blood pressure. BIS increased to 40 when the systolic blood pressure increased to 130 mm Hg and the mean arterial blood pressure to 100 mm Hg. The total time the BIS value was less than 20 was approximately 5 min. After those two events, we attempted to keep his mean arterial blood pressure more than 100 mm Hg. BIS was more than 40, and BSR was less than 10. Arterial blood gas analysis (fraction of inspired oxygen, 0.4) showed an arterial oxygen tension of 131 mm Hg, a carbon dioxide tension of 45 mm Hg, a pH of 7.33, and a hemoglobin concentration of 9.1 g/dL. After surgery, sevoflurane was discontinued, the patient opened his eyes at a BIS value of 70, and the laryngeal mask was removed. The postoperative course was uneventful. The patient showed no new neurologic deficit.
In our patient, we were confronted with a sudden decrease in BIS to 0 with an increase in BSR to 100, possibly because of cerebral perfusion disturbances. During surgery, there are several possible causes for a decrease in BIS, including deep anesthesia, hypothermia, and cerebral ischemia. In our patient, decreasing sevoflurane concentration had little effect on the BIS during the first episode of BIS decrease. When BIS decreased to 0 shortly after the start of surgery, the delivered sevoflurane concentration was 1.5%. Katoh et al. (6) reported that BIS decreased linearly with an end-tidal sevoflurane concentration increasing from 0.2% to 1.4%. However, sevoflurane concentrations >1.4% produced a limited further reduction in BIS up to 2.4%. It is therefore unlikely that deep anesthesia was the reason for the decrease in BIS in this case.
Hypoxia, hyperventilation, and severe anemia, all of which might cause cerebral dysfunction, were not present in this case. Hypotension, although the mean arterial blood pressure remained above the normal limit of autoregulation, was the likely cause of the decreased BIS value, because increasing the blood pressure resulted in an increased BIS and decreased BSR.
In a patient with normal autoregulation, cerebral blood flow (CBF) should be maintained at a mean arterial blood pressure of 80 mm Hg. However, in this case, BIS decreased at this level of mean arterial blood pressure. Chronic hypertension is accompanied by a rightward shift of the autoregulation curve (7); the lower limit of autoregulation is higher in hypertensive patients. Our patient had untreated hypertension before the onset of cerebral hemorrhage, and antihypertensive therapy was not yet effective at the time of operation. The mean arterial blood pressure of 80 mm Hg was approximately a 40% decrease from preinduction values in this patient. Therefore, CBF might have decreased to the level at which cerebral function was disturbed and BIS decreased. Recent cerebral hemorrhage might have promoted the decrease in CBF.
Typical EEG changes during cerebral ischemia have been reported as progressive slowing of the signal accompanied by a decrease in high-frequency activity and a generalized attenuation of voltage and, ultimately, becoming an isoelectric EEG with prolongation of ischemia (8). EEG slowing should decrease the spectral edge frequency that might be accompanied by a decrease in BIS, because the changes in BIS correlated well with spectral edge frequency during surgical levels of anesthesia (9). Isoelectricity of the EEG is reflected in an increase in BSR (the percentage of the suppression period against the time of analysis) (10). We have also reported that BSR values >40% are linearly and inversely correlated with BIS values in the range of 30 to 0 (9). A BSR of 100 indicates EEG silence, which results in BIS value of 0. This suggests that a BIS value <30 is an indicator of EEG suppression when BSR is high. Therefore, EEG changes during cerebral ischemia should result in decreased BIS.
During 0.6%–1.2% sevoflurane in 50% nitrous oxide anesthesia, similar to the dose used in our case, the CBF value at which 50% of patients demonstrated EEG evidence of cerebral ischemia with carotid cross-clamping during carotid endarterectomy was reported to be 11.5 ± 1.4 mL · 100 g−1 · min−1 (11). The decreased BIS values in our case suggest that CBF decreased to this range.
BIS has been suggested to be an indicator of inadequate cerebral perfusion in cases of cardiac arrest (2), perioperative stroke (3), and pediatric cardiac surgery (4). Similar to our case, these reports suggest that BIS intended to monitor depth of anesthesia may also detect incidental ischemic brain insults. Thus, the patients with risk factors such as a history of ischemic cerebral diseases may especially benefit from use of a BIS monitor. BIS has some limitations as a cerebral perfusion monitor. The monitoring area is only the unilateral frontal cortex. Thus, the possibility of false-negative values for other parts of the brain is large. However, BIS was helpful in our case, because we would not have expected cerebral hypoperfusion at a systolic blood pressure of 110 mm Hg. Although it is clear that BIS was not designed to be a monitor of cerebral ischemia, it might detect cerebral hypoperfusion or ischemia when anesthesia is stable. It is also stressed that meticulous care should be exercised to maintain optimal blood pressure in the patient with severe uncontrolled hypertension.
1. Johansen JW, Sebel PS. Development and clinical application of electroencephalographic bispectral monitoring. Anesthesiology 2000;93:1336–44.
2. Billard V. Brain injury under general anesthesia: is monitoring of the EEG helpful? Can J Anaesth 2001;48:1055–60.
3. Welsby IJ, Ryan JM, Booth JV, et al. The bispectral index in the diagnosis of the perioperative stroke: a case report and discussion. Anesth Analg 2003;96:435–7.
4. Hayashida M, Chinzei M, Komatsu K, et al. Detection of the cerebral hypoperfusion with bispectral index during paediatric cardiac surgery. Br J Anaesth 2003;90:694–8.
5. Hagihira S, Takashina M, Mori T, et al. Practical issues in bispectral analysis of electroencephalographic signals. Anesth Analg 2001;93:966–70.
6. Katoh T, Suzuki A, Ikeda K. Electroencephalographic derivatives as a tool for predicting the depth of sedation and anesthesia induced by sevoflurane. Anesthesiology 1998;88:642–50.
7. Joshi S, Ornstein E, Young W. Cerebral and spinal cord blood flow. In: Cottrell JE, Smith DS, eds. Anesthesia and neurosurgery. 4th ed. St. Louis: Mosby, 2001:19–68.
8. Clute HL, Levy WJ. Electroencephalographic changes during brief cardiac arrest in humans. Anesthesiology 1990;73:821–5.
9. Morimoto Y, Hagihira S, Koizumi Y, et al. The relationship between bispectral index and electroencephalographic parameters during isoflurane anesthesia. Anesth Analg 2004;98:1336–40.
10. Rampil IJ. A primer for EEG signal processing in anesthesia. Anesthesiology 1998;89:980–1002.
11. Grady RE, Weglinski MR, Sharbrough FW, et al. Correlation of regional cerebral blood flow with ischemic electroencephalographic changes during sevoflurane-nitrous oxide anesthesia for carotid endarterectomy. Anesthesiology 1998;88:892–7.