The Use of Dexmedetomidine Infusion for Awake Craniotomy

Anesthesia for intracranial procedures requiring patient cooperation presents a challenge to the anesthesiologist. Drugs administered during the procedure should provide an adequate level of sedation and analgesia for bone flap removal, but must not interfere with functional testing and electrocorticography. In this case report, we describe the use of dexmedetomidine in combination with nitrous oxide and sevoflurane for bone flap removal and dexmedetomidine alone for brain mapping of the cortical speech area. Dexmedetomidine is a highly specific α2-adrenoreceptor agonist with sedative, analgesic, and anesthetic-sparing effects (1,2). It does not suppress ventilation. Small-dose infusion of this drug in healthy volunteers provided sedation that could be easily reversed with verbal stimuli (3). We anticipated that the patient treated with dexmedetomidine would be sedated and comfortable but easily arousable to tolerate a prolonged awake craniotomy. Case Report A 38-yr-old man was scheduled for resection of a left temporal brain neoplasm. The patient presented with a 14-mo history of increasing difficulty expressing himself and poorly controlled partial complex seizures. His medical history was otherwise unremarkable. The patient received no preoperative medication. In the operating room the patient was monitored with a noninvasive blood pressure cuff, a pulse oximeter, and electrocardiogram. The approximate level of sedation was measured by using a Bispectral electroencephalogram index (BIS). Anesthesia was induced with propofol 200 mg and fentanyl 100 μg IV. A laryngeal mask airway (LMA) was then placed. A right radial arterial line and Foley catheter were inserted after induction of anesthesia. The scalp nerves were blocked with 0.5% bupivacaine (4). The patient was then positioned in the right decubitus position. Anesthesia was maintained with 70% nitrous oxide, dexmedetomidine infusion (initial dose 1 μg/kg IV over 30 min followed by continuous infusion of 0.4 μg · kg−1 · hr−1 IV), and sevoflurane (0.3%–0.7%). The sevoflurane concentration was adjusted to maintain a BIS index between 50 and 60. We observed a decrease in heart rate from 62–78 bpm to 50–60 bpm and blood pressure from 95–110/55–65 mm Hg to 90–100/52–60 mm Hg after an initial dose of dexmedetomidine. The patient was breathing spontaneously throughout the procedure. Skin incision, bone flap removal, and dissection of dura were uneventful. Twenty minutes before the intended awakening, ondansetron 4 mg IV was administered, sevoflurane was discontinued and dexmedetomidine infusion was reduced to 0.2 μg · kg−1 · hr−1. The LMA was removed, and oxygen was administered via nasal cannula. Vital signs remained stable. At a dexmedetomidine infusion rate of 0.2 μg · kg−1 · hr−1 the patient was arousable, but too sedated to complete counting or sentence completion tasks (BIS 75–80). The infusion rate was reduced to 0.1 μg · kg−1 · hr−1. BIS increased to 95 within 15 min. Over 2 h were spent performing language localization. In addition to dexmedetomidine, the patient received fentanyl 100 μg. The tumor adjacent to the language areas was dissected from the normal brain while the patient was undergoing continual testing to assess language as the resection progressed. The BIS was approximately 95 while the patient was awake. An arterial blood gas sample taken during this time revealed a carbon dioxide partial pressure (Paco2) of 42 mm Hg and oxygen partial pressure (Pao2) of 156 mm Hg. The remainder of the surgical procedure was uneventful. The patient was sedated with propofol 150 mg IV and a LMA was reinserted. Dexmedetomidine infusion was discontinued and anesthesia was maintained with nitrous oxide and sevoflurane. On completion of the procedure, sevoflurane and nitrous oxide were stopped. The patient was responsive to verbal commands immediately after the LMA removal. He remembered that he was awake during the tumor resection, but could not recall any details. Discussion Anesthesia for awake craniotomy has varied from local to general with intraoperative awakening during language mapping and tumor resection (5). The combined use of droperidol and one of the lipid soluble opioids termed neuroleptanalgesia has been traditionally used to produce a state of indifference, immobility, and analgesia (6). Common intraoperative complications associated with this technique include agitation, drowsiness, pain, and seizure (7,8). Moreover, neuroleptanesthesia may lead to respiratory depression. Paco2 may reach 45–60 mm Hg (9). Several studies have endorsed the use of propofol for awake craniotomy (10,11). The use of propofol reduced the incidence of perioperative seizures and agitation, but led to a frequent incidence of respiratory depression (10). More recently, the combination of propofol and remifentanil has been successfully used for brain tumor mapping (12,13). Remifentanil’s context-sensitive half-life is short (<5 minutes) and independent of infusion duration, which allows a rapid modulation of analgesia and sedation required during the course of the surgery. The side effects of remifentanil, however, are similar to those of all fentanyl congeners. The patients in both studies had episodes of respiratory depression, airway obstruction, and desaturation. This case report describes the use of dexmedetomidine infusion for an awake language mapping and left temporal tumor resection. Dexmedetomidine is a selective α2-adrenoceptor agonist with centrally mediated sympatholytic effects. It significantly reduces the intraoperative and postoperative anesthetic requirements for various surgical procedures (14,15). A lack of respiratory depression offers a distinct advantage over other anesthetic techniques for an awake craniotomy. The manufacturer’s recommended maintenance dose for intensive care unit sedation is from 0.2 to 0.7 μg · kg−1 · hr−1. We found that our patient was excessively sedated even at 0.2 μg · kg−1 · hr−1. Although the patient was arousable and responded to simple commands, he could not perform counting or sentence completion tests. He could complete all required tests and remained cooperative at 0.1 μg · kg−1 · hr−1. Coadministration of sevoflurane, midazolam, fentanyl, or propofol may lead to enhancement of α2-adrenoceptor agonist sedative effects (16,17). It is unlikely, however, that the administration of these drugs at the beginning of the surgery (100 minutes before awakening) led to excessive sedation. It is more likely that the dose suggested for extubation of the mechanically ventilated intensive care unit patients is too large to perform the complex tasks of cortical language mapping. The BIS scores may decrease by 31% after a 60-min dexmedetomidine infusion at 0.2 μg · kg−1 · hr−1 in healthy volunteers (3). Once stimulated, however, the subject’s alertness returned to a baseline level. Continuous verbal stimulation explains only a minimal BIS index reduction during cortical speech mapping. Dexmedetomidine produces dose-dependent decreases in blood pressure and heart rate as a result of its agonistic effect at the α2-adrenoreceptors (1,15). Although heart rate and blood pressure decreased after the administration of a dexmedetomidine initial dose, the changes were minimal. The patient did not receive any vasoactive drug during the procedure. The observed minimal hemodynamic changes in our case are consistent with the reported effect of small dose dexmedetomidine infusion on cardiovascular function in healthy adults (3). After IV infusion, dexmedetomidine exhibits the following pharmacokinetic variables: a rapid distribution phase with a distribution half-life (t1/2α) of approximately 5 minutes and a terminal elimination half-life (t1/2β) of approximately 120 minutes (18,19). Although a context-sensitive half-life after infusions of different durations is not known, Talke et al. (18) report that dexmedetomidine plasma concentration was halved within 20 minutes after 60 minutes of continuous infusion at the rate of 1.15 μg/hr. We had to reduce the infusion rate from 0.4 μg · kg−1 · hr−1 to 0.2 μg · kg−1 · hr−1 and then from 0.2 μg · kg−1 · hr−1 to 0.1 μg · kg−1 · hr−1 during the course of the operation. In both cases, a waiting interval of 15 minutes was sufficient to observe changes in the sedation level from a Ramsey score of 6 (asleep with no response) to a Ramsey score of 3–4 (asleep with response to simple command) and, subsequently, to a Ramsey score of 2 (patient cooperative, oriented, and tranquil) (20). In summary, we report the first successful application of dexmedetomidine combined with BIS monitoring in an awake craniotomy setting. The pharmacology of dexmedetomidine allowed us to achieve a level of sedation and analgesia sufficient to complete the neuropsychiatric testing required for the mapping of the cortical language area, as well as to perform an awake tumor resection. The patient remained hemodynamically stable and cooperative during the “awake” portion of the procedure.