Reduction of bispectral index value associated with clinically significant cerebral air embolism

A 38-year-old man with Marfan syndrome presented to the operating room for an aortic valve replacement and aortic annulus enlargement (Nicks procedure). He had undergone a David procedure 3 years before this admission. Since the previous procedure, he developed 3 aortic regurgitation associated with symptoms of mild congestive heart failure. On this admission, he had no neurologic abnormality. No premedication was administered. Anesthesia was induced with midazolam (1 mg), fentanyl (2 g/kg), and thiopental (4 mg/kg), and tracheal intubation was facilitated with vecuronium (120 g/kg). Another 2 g/kg of fentanyl was given at the time of skin incision. Anesthesia was maintained with 1.0% to 3.0% sevoflurane in 100% oxygen and 0.5 g/kg/h of fentanyl infusion. No bolus doses of intravenous anesthetics were given after skin incision. A BIS monitor (Aspect A-1050 EEG monitor; Aspect Medical Systems Inc., MA) was used to assess hypnotic level throughout the procedure. Cardiopulmonary bypass (CPB) was established with arterial cannulation in the ascending aorta distal to the graft from the previous surgery and bicaval venous cannulae. Two liters per minute of carbon dioxide insufflation was applied continuously to the surgical wound through an intravenous tube (inner diameter 2 mm) sutured to the caudal end of the wound throughout the CPB period. One percent sevoflurane was administered throughout the CPB. PaO2 was maintained higher than 250 mmHg, and the PaCO2 was maintained within the range of 30mmHg to 38 mmHg using alpha-stat management. The minimal esophageal temperature during the CPB period was 27.2°C. No significant metabolic acidosis developed. The BIS value stayed at 45 to 50 after the anesthetic induction until the initiation of CPB. It then decreased to 35 as the patient was cooled and stayed between 35 and 40 until the rewarming started. The aorta was cross-clamped distal to the graft from the previous surgery. The aortic graft was opened, the aortic annulus was enlarged, and a 23-mm bileaflet mechanical valve (St. Jude Medical, Inc, St. Paul, MN) was inserted into the enlarged aortic annulus. During the period of rewarming, the aortotomy was closed, and a needle was stuck into the aortic graft to release the air inside the aorta proximal to the aortic cross-clamp site. The left ventricular venting was discontinued, and the lungs were ventilated several times to fill the left atrium and ventricle with blood to release air inside them as well as in the proximal aorta. After the temporary slowing down of the bypass pump flow, the aortic cross-clamp was released with the patient placed in a steep Trendelenburg position and his bilateral carotid arteries temporarily compressed manually by the anesthesiologist. Before the aortic cross-clamp was released, the BIS value was increasing steadily to the mid-50s as the rewarming progressed; but after the aortic cross-clamp was released, it decreased rapidly to 30. Rectal temperature was 35.2°C and mean arterial pressure was 55 mmHg at the time of sudden decrease in the BIS value. Using transesophageal echocardiography, a large amount of air was found in the aortic arch and descending aorta (Fig 1). No air was seen inside the left ventricle and the left atrium. The surgical team was informed of the findings, and they tried to evacuate as much air as possible from the needle through the aortic graft; but air was still seen inside the descending aorta for the next minute. The patient was hemodynamically stable, and he was weaned from CPB easily with a moderate catecholamine infusion. After the conclusion of CPB, the BIS number remained in the mid-30s, even though no inhalation or intravenous anesthetic agents were given for more than an hour. The total doses of intravenous anesthetics during the 6-and-one-half-hour surgery were fentanyl, 400 g; midazolam, 1 mg; and thiopental, 240 mg. The patient did not receive any neuromuscular blockade after the conclusion of CPB and had 4 strong twitches with the train-of-4 stimuli from a neuromuscular blockade-monitoring device. After the completion of the surgical procedure, the patient was transferred to the intensive care unit and his trachea remained intubated. When he arrived in the intensive care unit, he opened his eyes spontaneously. However, he did not respond to verbal commands, and no movement of the extremities was seen. Even though no sedatives were given in the intensive care unit, over the next 12 hours his level of consciousness did not improve. A computed tomography scan of the head was obtained, and it showed no abnormality. On postoperative day 1, he started to move the right-side extremities without apparent weakness after a verbal command, but the left-side extremities remained paralyzed. He started to move his left-side extremities on postoperative day 2, and he regained normal muscle strength on both sides in the next 3 days. His level of consciousness also kept improving, and his trachea was extubated on the third postoperative day. A repeat computed tomography scan of the head showed no cerebral infarction. He was transferred to a regular ward on postoperative day 5.