| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesiology, University of Washington, School of Medicine, Seattle, Washington
Address correspondence and reprint requests to Andrius P. Skucas, MD, Department of Anesthesiology, Box 356540, University of Washington, School of Medicine, Seattle, WA 98195-6540. Address e-mail to askucas{at}u.washington.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
At our institution, awake craniotomies for epileptogenic focus resection are performed with an unprotected airway. A uniform protocol used since 1993 entails deep sedation by propofol infusion for injection of local anesthetic and craniotomy. During this time, spontaneous ventilation is preserved and no endotracheal tube, laryngeal mask airway (LMA), or similar device is used. Sedation is discontinued for mapping of language, motor, and/or sensory function and for ECoG. Patients are then re-sedated for resection of the lesion and closure. Our propofol-based awake craniotomy technique was first described as a case series of 9 patients premedicated with fentanyl 50-100 µg and droperidol 1.25-2.5 mg (5). No complications were noted. Since that time we have simplified our management to using a propofol infusion as the sole sedative drug and avoiding premedications as much as possible. We continue to avoid the use of endotracheal intubation or LMA placement.
Commonly called the "asleep-awake-asleep" (AAA) technique for craniotomy, concerns have been raised in the anesthesia literature including the lack of a secured airway, poor patient cooperation, patients experiencing nausea or pain during the procedure, and others (1,4,6–11). We performed a retrospective chart review of AAA cases for epileptogenic focus resection at our institution to examine the incidence of intraoperative complications. A chart review of epileptogenic focus resection performed under GA with endotracheal intubation was performed for comparison.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Anesthetic management in the AAA group was as outlined above. In the lateral or semi-lateral position, patients received a propofol infusion for the "asleep" portions; discontinuation of propofol for the "awake" portions; no use of an endotracheal tube, LMA, or similar device; and spontaneous ventilation. Propofol was dosed as an initial IV bolus of 0.5 mg/kg, then subsequent bolus doses of 0.25 mg/kg until the desired level of sedation was achieved, followed by continuous infusion at 75-250 µg·kg-1·min-1. Room air was supplemented by oxygen 3-6 L/min via nasal prongs or facemask. Use of a nasal trumpet was a standard part of the AAA technique and was not considered a "secured airway." Anesthesia in the GA group was induced with one or more IV drugs (thiopental, propofol, lidocaine, and/or opioid), an endotracheal tube was inserted, the patient's lungs were ventilated mechanically, and anesthesia was maintained using a volatile anesthetic (isoflurane or sevoflurane in oxygen) supplemented with an IV drug (fentanyl, sufentanil, or remifentanil).
Airway/ventilation complications in AAA cases were defined as any maneuver beyond placement of an oral or nasal airway required to permit unobstructed breathing and/or adequate ventilation. Airway/ventilation complications in GA cases were defined as difficulty (>3 attempts by more than 2 anesthesia care providers) or inability to insert an oral endotracheal tube, or difficulty with ventilation once the patient was tracheally intubated. In both AAA and GA, moderate oxyhemoglobin desaturation was defined as Spo 2 91%-95% and severe oxyhemoglobin desaturation was defined as Spo 2 
90%. The proportion of cases with desaturation events was noted. Paco 2 was also noted for those cases in which arterial blood gas tensions were measured.
Hemodynamic complications included abnormal arterial blood pressure and heart rate. Hypertension was defined as systolic blood pressure more than 150 mm Hg, hypotension was defined as systolic blood pressure <90 mm Hg, tachycardia was defined as heart rate more than 110 bpm, and bradycardia was defined as heart rate <45 bpm. The proportion of cases with hemodynamic complications was noted.
Other complications noted were intraoperative seizures, new intraoperative neurologic deficits, nausea without vomiting, nausea with vomiting, patient movement, brain swelling, bleeding, local anesthetic toxicity, pulmonary aspiration, air embolism, and death. We also noted the frequency of re-do operations for both types of anesthetic technique versus first-time craniotomies and the frequency of invasive arterial blood pressure monitoring (i.e., insertion of a catheter into an artery, usually a radial artery).
Two-tailed unpaired Student's t-tests for data collected as means (age, obese weight, and Paco 2) and 
2 tests for data collected as proportions, were performed using GraphPad Prism (version 3.02 for Windows 98; GraphPad Software, San Diego, CA). A P value <0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
Airway/ventilation complications occurred in 6 AAA cases (1.8%). One required temporary closure of the craniotomy 2 h after incision for Spo 2<93% with subsequent oral endotracheal intubation (via fiberoptic bronchoscope) after induction of GA; the case was then resumed. The second case required placement of a LMA for Spo 2 88%. The third patient developed apnea after a remifentanil bolus was given in addition to propofol infusion; as a result of poor mask ventilation, a LMA was placed, nasal 2% lidocaine and transtracheal 4% lidocaine were applied, and a nasal endotracheal tube was placed (via fiberoptic bronchoscope). All three patients who required their airways to be secured were obese. The fourth incident resulted in a nasal airway being placed and attached to the anesthesia circuit for Spo 2 93% while the fifth incident had a similar management but with an endotracheal tube blindly placed through the nares to a position above the glottis (in both cases the patients continued to ventilate spontaneously); the latter patient was obese. The sixth case was complicated by intraoperative laryngospasm but the specific management was not noted (it was noted on the anesthetic record, however, that the patient did not require tracheal intubation.) All six of these incidents occurred before the awake testing portion of the case and, besides obesity, these patients did not have underlying diagnoses of asthma, tobacco use, obstructive sleep apnea, or other pulmonary disease. The single airway/ventilation issue in the GA group (0.8%) was the development of an endotracheal tube cuff leak in one patient. Air was replaced into the cuff eight times and the patient did not require reintubation. None of the GA cases had any other difficulty with ventilation or inability to intubate.
AAA cases had statistically significantly more frequent incidences of oxyhemoglobin desaturation, hypertension, hypotension, and tachycardia, and significantly higher levels of Paco 2 (Table 2). GA cases had a statistically significantly more frequent incidence of bradycardia (Table 2). There was more frequent use of invasive arterial blood pressure monitoring (54.3% versus 13.6%) in GA cases (Table 1). Incidences of other complications did not differ between AAA and GA (Table 2). There were no reported instances of local anesthetic toxicity, pulmonary aspiration, air embolism, or death.
|
One significant complication occurring in the AAA group appeared to be related to the anesthetic technique. The patient was a 59-yr-old female who refused blood transfusion for religious reasons. After craniotomy and removal of the bone flap, marked dural swelling was noted (Paco 2 at that time was 52 mm Hg). Attempts were made to mask-ventilate the patient to decrease the Paco 2. Nonetheless, with dural opening there was a large transverse sinus tear with a resultant significant hemorrhage (postoperative hematocrit was 18%). Although she survived the hospitalization, extensive ischemia was noted in the left temporal region with persistent right-sided hemiparesis at the time of discharge.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In addition to uniform adherence to a propofol infusion without other sedative drugs, AAA craniotomies at our center are performed with no instrumentation of the airway besides placement of a nasal trumpet. Techniques described for more secure airway protection include endotracheal intubation and awakening without removal of the endotracheal tube (disallowing vocalization testing), extubation and reintubation with an endotracheal tube after testing, and methods using placement, extubation, and replacement of an LMA or cuffed oropharyngeal airway (7,8,14,19–21). Noninvasive positive pressure ventilation has also been described (22). In our series of patients with unsecured airways, airway compromise during AAA requiring urgent or emergent securing of the airway was relatively rare and occurred only in obese patients; there was no association with asthma, tobacco use, or obstructive sleep apnea. Only 3 of 332 AAA patients required airway control with an endotracheal tube or LMA, including one patient in whom remifentanil was used in addition to propofol (a departure from our usual protocol). Two other patients experiencing desaturations simply had nasal airway devices placed and connected to the anesthesia circuit with adequate oxygen saturations (previously described as a means of improving oxygenation with spontaneous breathing and with controlled ventilation by closing the mouth and applying pressure to the reservoir bag) (23). Obesity (body mass index >30 kg/m2) appeared to be a consistent risk factor that led to oxygen desaturations requiring a secure airway. There were no negative sequelae to patients who had respiratory complications. The GA group had no significant respiratory complications besides a slowly leaking endotracheal tube cuff.
Higher Paco 2 was more common in the AAA group, as would be expected in a population of sedated patients breathing spontaneously. Brain swelling was noted in only 2 of those 332 patients, and may have contributed, in one of those two patients, to the occurrence of a clinically significant hemorrhage with dural opening. In retrospect, that patient may have benefited by more definitive airway control to allow hyperventilation before dural opening.
Hypertension, hypotension, and tachycardia were more frequent in the AAA group. In most cases, recognition of these hemodynamic changes prompted initiation of appropriate treatment and in no cases did isolated hemodynamic variability result in negative sequelae to a patient. Bradycardia was relatively rare in both groups. Reports of combining a propofol infusion with a remifentanil infusion and clonidine suggest smoother hemodynamic conditions, although we believe that this benefit may not outweigh the risk of respiratory complications from the use of opioids (10,16).
Differences in incidence of intraoperative seizures were not statistically significant between the two groups. In the AAA group, half of the seizures were tonic-clonic seizures and half were focal seizures of one or two extremities. Half occurred during the asleep portion and half during awake testing. Eight seizures were so brief in duration that no treatment was administered. One seizure during awake testing was treated with a small propofol bolus, another with a midazolam bolus. None of these patients required their airways to be secured with an endotracheal tube or LMA.
In the AAA group, nausea without vomiting, nausea with vomiting, and patient movement were relatively rare and were managed without apparent difficulty. Patients did not receive prophylactic antiemetic medications although when nausea did occur, antiemetics were administered. The infrequent nausea in the AAA group is at least partially attributable to the antiemetic properties of propofol and the lack of opioid administration.
The incidence of complications when using our anesthetic technique is compared with those reported in previous studies in Table 3. Although direct comparisons of respiratory complications with the prospective trials are difficult to interpret because of small sample sizes, our rates compare favorably. One retrospective trial had less respiratory complications but did not separate complications by anesthetic protocol (24). In the Herrick et al. study (16), airway complications in the propofol group occurred when opioids were added. Berkenstadt et al. (10) noted a decrease in respiratory complications over time as they became accustomed to their anesthetic protocol; our frequency of complications did not vary significantly over a 12-year period. The highest rates of seizures and nausea occurred in predominantly opioid-based protocols (13). Our protocol had the highest rates of hypotension, although this variable was not examined in most trials.
|
Limitations of this study relate to those inherent to a retrospective chart review. The quality of the data collected is limited to the accuracy of the original medical chart and an intraoperative complication that was not reported would not be reflected in this study. Data regarding securing of the airway, oxygen desaturations, Paco 2, and hemodynamic variations are considered more reliable in being reflected in the anesthetic record compared with other variables studied. Several other factors may have contributed to the less frequent incidence of many categories of complications during GA as compared with AAA. One factor may have been our preoperative anesthetic assessment of the patient and selection of GA in patients determined to not be good candidates for the AAA technique. Use of controlled mechanical ventilation during GA permits the patient to be hyperventilated to relieve brain swelling. Control of brain swelling, in turn, decreases the risk of tearing a cerebral vessel or sinus at the time of dural opening. Also, arterial catheters were placed in a larger proportion of GA patients and their use facilitates more frequent sampling of arterial blood for measurement of Paco 2. Despite a larger proportion of re-do craniotomies in the GA group, these interventions may have decreased risks associated with dural adhesions and other scarring resulting from previous surgeries.
In summary, in this series of 332 epilepsy patients who underwent a propofol-based AAA technique with an unsecured airway, there were relatively infrequent clinically significant complications. Only one serious complication appears to have been related to the anesthetic technique.
The authors thank Karen Posner, PhD, and John Campos, MA (Department of Anesthesiology, University of Washington, Seattle, Washington) and George Ojemann, MD (Department of Neurological Surgery, University of Washington, Seattle, Washington) for assistance in obtaining patient case lists.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Dinsmore Anaesthesia for elective neurosurgery Br. J. Anaesth., July 1, 2007; 99(1): 68 - 74. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
