THE Bispectral Index (BIS) monitor (BIS® monitor; Aspect Medical System, Inc., Newton, MA) integrates electroencephalographic data via a proprietary algorithm based on empirical data into a nominal 100 to 0 scale, in which 100 represents the awake state and 0 represents the isoelectric electroencephalogram. 1The BIS® monitor is highly capable of correctly predicting both the loss and recovery of consciousness and has been reported to be more reliable in assessing the level of sedation and hypnosis during surgery than other available processed electroencephalogram algorithms. 2–4It also has been used clinically to guide anesthetic drug administration 5and to quantify the pharmacokinetic and pharmacodynamic action of anesthetic drugs in the laboratory setting. 6
We report a case of a rapid decrease in the BIS score during total intravenous anesthesia using propofol after cross-clamping the descending thoracic aorta for elective thoracic aortic aneurysm repair and suggest that the BIS® monitor provides useful information on altered propofol pharmacokinetics after aortic cross-clamping.
A 46-yr-old man was admitted for elective thoracic aortic aneurysm repair. His history was unremarkable except for hypertension. After routine monitoring of vital signs was supplemented with BIS® monitoring, anesthesia was induced with fentanyl and propofol and maintained with infusions of propofol and ketamine and intermittent injections of fentanyl via an upper limb vein. The infusion rate of propofol was varied to maintain the BIS between 40 and 60. A Swan-Ganz catheter was inserted into the right internal jugular vein for monitoring continuous cardiac output. The patient had consented to participate in a clinical investigation of the suppressive effects of propofol anesthesia on transcranial electrical motor evoked potentials during thoracic and thoracoabdominal aortic surgery. Blood samples for this purpose were drawn from the arterial line at defined time intervals. A stable level of anesthesia was maintained, with a BIS score of approximately 60. After isolation of both the right femoral artery and vein, the arterial cannula was inserted and then attached to the pump oxygenator. A long venous cannula (Long Cannula RK-28M, Toyobo, Osaka, Japan) was introduced into the right femoral vein, and the tip of this catheter was advanced between the inferior vena cava and right atrium. The tip was confirmed to be correctly positioned by transesophageal echocardiography. A standard roller pump and oxygenator to circulate oxygenated blood from the long venous cannula to the femoral artery were available for partial cardiopulmonary bypass of distal perfusion. During surgery, the patient was hemodynamically stable. A cross-clamp of the descending thoracic aorta at the level of Th 7 was performed, followed by the start of partial cardiopulmonary bypass of distal perfusion at a perfusion rate of 3–3.2 l/min under normothermic conditions. Guided by transesophageal echocardiographic images that revealed the left ventricular cavity size and function, perfusionists could regulate the rate of distal perfusion and level in the venous reservoir for adequate preload in the upper body. Approximately 15 min after cross-clamping, the patient's BIS level decreased rapidly from its steady-state value of 60 to a value of 0 (infusion rate of propofol, 6 mg · kg−1· h−1). Mean arterial pressure was 72 mmHg, and continuous cardiac output monitoring indicated 1.2–1.4 l/min proximal to the aortic cross-clamp. Pulsation of bilateral carotid arteries was clearly palpable, suggesting adequate blood supply to the brain. Although depth of anesthesia was assessed as very deep, infusion of propofol was discontinued until the BIS started to increase. Fifteen minutes after infusion of propofol was stopped, the BIS began to increase by degrees. At a BIS score of 30, infusion of propofol was begun at 6 mg · kg−1· h−1. Subsequently, the BIS score began to decrease again and was less than 10 at approximately 25 min after restart of propofol infusion. This phenomenon could be reproduced by “stopping and restarting” of propofol infusion (Fig. 1). An aortic graft was inserted, with a cross-clamp time of 105 min. Approximately 15 min after declamping, the patient's BIS level increased slowly from a value of 20 to more than 30. Surgical procedures were completed uneventfully, and this patient was admitted to the intensive care unit, with an anesthetic time of 7 h 45 min. The patient had recovered completely from anesthesia 6 h after admission to the intensive care unit and manifested no neurologic deficits. All blood samples drawn simultaneously from the radial artery and the left femoral artery, including those before cross-clamping of the aorta, at 3 time points during aortic cross-clamping, and 15 min after declamping, were retained for high-performance liquid chromatography (MC Medical, Inc., Osaka, Japan) analysis. High-performance liquid chromatography analysis revealed that the propofol concentrations of samples from the radial artery (proximal to the aortic clamp) during cross-clamping of the aorta were significantly elevated compared with pre–cross-clamping values. Also, propofol concentrations in samples from the radial artery were approximately 4–5 times higher than those from the femoral artery (Fig. 1).
The depth of anesthesia using propofol was assessed with the BIS® monitor, which appears to be a very promising tool for evaluating the adequacy of anesthesia. 7Therefore, the BIS® monitor can provide information on an adequate rate of propofol infusion in individual patients. In our case, before the aortic cross-clamp, the BIS® monitor indicated that 6 mg · kg−1· h−1of propofol was an adequate rate of infusion. However, despite steady infusions of propofol, a rapid decrease in BIS occurred approximately 15 min after the aortic cross-clamp. Hypoperfusion in the brain was an unlikely cause of this decrease, according to physical findings that included palpable pulses of bilateral carotid arteries and hemodynamic variables that included radial arterial blood pressure and cardiac output. Indeed, this patient had normal emergence from anesthesia and normal neurologic function within 24 h of the postoperative period. In this case, BIS values had altered, corresponding to the “stopping-restarting” of propofol infusion, and the plasma propofol concentration was much higher during the aortic cross-clamping than before. It was therefore likely that the increase in plasma propofol concentration (> 7 μg/ml) would induce an isoelectrical encephalogram resulting in a BIS of 0 during the aortic cross-clamp.
There are two explanations for such a decreased BIS value. First, cross-clamping of the thoracic aorta would change circulatory conditions, such as reducing circulating volume and decreasing cardiac output in the area proximal to the aortic cross-clamp, because blood flow from the heart should be only for perfusion of the head, neck, and upper limbs. It was reported that blood flow through the descending thoracic aorta was approximately 65–70% of the cardiac output in humans, 8meaning that only 30–35% of cardiac output may be essential for perfusion of the area proximal to the cross-clamp of the descending thoracic aorta. Therefore, it would be reasonable for approximately one third of the total blood volume to be circulated in the area proximal to the aortic cross-clamp. In this case, transesophageal echocardiographic images confirmed that the tip of the long venous cannula was positioned at the junction between the inferior vena cava and right atrium. This long venous cannula has 10 oval side-holes (10 × 5 mm) in the 20 cm that compose the tip. We believe that much of the venous blood in the inferior vena cava would be withdrawn from the side-holes of this cannula during the cardiopulmonary bypass. During the stage of cardiopulmonary bypass for distal perfusion, we used transesophageal echocardiography to monitor the left ventricular cavity size and function to allow adequacy of preload and optimal fluid resuscitation for maintenance of both upper and lower body perfusion. Although it is likely that blood from the inferior vena cava should enter the right atrium and mix with blood from the superior vena cava, we think that only a small amount of such blood entered the right atrium, with the result that an almost “split circulation” was produced by the thoracic aortic cross-clamp. This change would greatly reduce the distribution volume of propofol infused via the upper limb and result in much higher plasma propofol concentrations in the body proximal to the aortic cross-clamp when the infusion rate was set on the basis of body weight.
Second, the reduction of hepatic clearance of propofol should be considered. A clinical study 9of patients having orthotopic liver transplantation showed that the area under the time–blood propofol concentration curve was significantly greater than in patients having extrahepatic abdominal surgery. Furthermore, Mashimo et al. 10demonstrated that total plasma clearance of propofol in the anhepatic phase was significantly lower than that in the postanhepatic phase during liver transplantation in the pig. Because blood flow into the liver from the area proximal to the aortic cross-clamp theoretically vanishes, propofol clearance is ultimately decreased in that area, which would, in part, result in an increase in the blood concentration of propofol.
In our patient, ketamine, as well as propofol, was infused at 0.5 mg · kg−1· h−1during the aortic cross-clamp. Although we did not measure blood concentrations of ketamine in this patient, it is, of course, likely that with the infusion of ketamine, its blood concentration increased and deepened the level of anesthesia. 11However, ketamine does not influence either BIS or electroencephalogram variables during propofol infusion. 11,12Together with the series of observations in our patient, this dramatic change of BIS values should be associated predominantly with changes in propofol pharmacokinetics.
In our institution, monitoring of motor evoked potentials was used to monitor the functional integrity of the descending motor pathways during thoracic or thoracoabdominal aneurysm surgery under total intravenous anesthesia via an upper limb vein. It has been our experience in surgery of the descending thoracic aorta that motor evoked potentials, in the setting of steady-state propofol infusion, disappeared during the aortic cross-clamp and reappeared after propofol infusion was stopped. From this case and those experiences, it can be considered that in the cross-clamp of the thoracic aorta, a rapid change in BIS value may reflect an increase in propofol concentration in “the brain,” caused by a reduction of distribution volume for propofol and an ultimate reduction of its clearance. This change may induce undetectable motor evoked potentials without any spinal cord injury during thoracic aortic surgery and result in a “false-positive.”
In conclusion, we would like to emphasize that an unexpected, rapid decrease in the BIS score during cross-clamp of the thoracic aorta might reflect an unexpected increase in propofol concentration in the proximal area of the aortic cross-clamp. It was speculated that this increase in propofol concentration might be caused by alterations of the pharmacokinetics of propofol by the aortic cross-clamp. The BIS® monitor is a useful tool for providing information on the adequate infusion rate of propofol during descending thoracic aortic surgery.