Early postoperative recovery of neurologic and cognitive functions is advantageous after neurosurgical procedures because it expedites the diagnosis should a life-threatening complication develop . An approach for combining stable haemodynamics with early predictable recovery of consciousness is to use opioid analgesic agents that are easily titratable, rapidly reversible and allow changes in the depth of anaesthesia during surgery while providing a short recovery time regardless of how long the operation lasts [2,3]. Commonly used opioid analgesic agents that provide stable intraoperative haemodynamic control and fast-track awakening after cardiac surgery and neurosurgery are remifentanil and sufentanil [4,5].
No published studies have investigated early postoperative cognitive recovery after remifentanil- and sufentanil-based anaesthesia in a neurosurgical setting. Early postoperative (postanaesthesia) cognitive recovery is difficult to evaluate owing to the lack of consensus-based tests. As primary outcome measures for evaluating early postoperative cognitive recovery, from the numerous scales available for cognitive assessment we decided to use a translated Italian version of the Short Orientation Memory Concentration Test (SOMCT), a patient-based test designed to assess cognitive function that has already been used in the perioperative setting [6,7], and the Rancho Los Amigos Scale (RLAS), a simple and quick physician-based scale designed to quantify behavioural and cognitive long-term recovery after acute brain injuries and also used in the postoperative intensive care setting [8,9].
This double-blind study was primarily designed to compare early postoperative (postanaesthesia) cognitive recovery after total intravenous anaesthesia induced with propofol as the hypnotic agent, and remifentanil or sufentanil as opioid analgesics in patients undergoing craniotomy for resection of supratentorial expanding lesions. To ensure comparability of the primary outcome variable in the two groups (postoperative cognitive recovery measured by the SOMCT and RLAS), we assessed intraoperative and postoperative haemodynamics, emergence times (recovery and extubation times) and postoperative pain.
The institutional review board approved the study procedures and written informed consent was obtained from all patients. Sixty consecutive patients (18–75 yr) classified as ASA I–III, Glasgow Coma Scale (GCS) score 15, undergoing craniotomy for supratentorial expanding lesions resection were enrolled in this prospective, randomized, double-blind study (neither the cognitive status assessor nor the patient knew the assigned treatment). All patients were randomly assigned (computer-derived randomization list) to one of two groups to receive anaesthesia with propofol plus remifentanil or sufentanil. Patients were excluded if they had a GCS score of <15, a body mass index >30 (obese patients), a history of drug abuse, preoperative aphasia or postoperative aphasia and neurologic deficit and preoperatively foreseen delayed extubation. An isotonic crystalloid saline solution (10 mL kg−1 preoperatively and 5 mL kg−1 h−1 intraoperatively) was infused through a peripheral intravenous catheter, and a second venous line was inserted for anaesthesia induction.
Perioperative monitoring included, in all patients, continuous two-lead ECG, heart rate (HR), invasive (arterial line) systolic, diastolic and mean arterial pressure (MAP), SaO2, end-tidal CO2, diuresis and rectal or oesophageal temperature. The depth of anaesthesia was monitored with bispectral index (BIS) (target 40–50) through four skin electrodes (Zipprep; Aspect Medical System, Inc, Natick, MA, USA) placed in a two-channel referential montage on patients' foreheads. All leads were connected to an electroencephalographic monitor (A-1000; Aspect Medical System International B.V., Leiden, The Netherlands). Data were sampled 128 times s−1, and high-frequency (70 Hz) and low-frequency (2 Hz) filters were used with version 3.22 of the BIS. Intraoperative normothermia was actively maintained with a forced air warming system (Bair Hugger, USA).
In the remifentanil–propofol group, anaesthesia was induced with remifentanil (0.5 μg kg−1 min−1) infusion, 1 min later, propofol (2 mg kg−1) was administered. After loss of consciousness, manual mask ventilation was begun, and propofol continuous infusion (150 μg kg−1 min−1) was started. Cisatracurium (0.1 mg kg−1) was administered to facilitate endotracheal intubation. Remifentanil was subsequently titrated to lower doses according the clinical need defined as the minimal dose required to maintain MAP within 10% of baseline. Propofol was administered at the discretion of the anaesthesiologist to keep the BIS between 40 and 50. Propofol infusion was reduced to 30–60 μg kg−1 min−1 when the bone flap was replaced, and stopped at the beginning of skin closure. Remifentanil infusion was discontinued after removal of the pinhead holder.
In the sufentanil–propofol group, anaesthesia was induced with sufentanil (0.5 μg kg−1) and propofol (2 mg kg−1). After loss of consciousness, manual mask ventilation was begun, and propofol continuous infusion (150 μg kg−1 min−1) was started. Cisatracurium (0.1 mg kg−1) was administered to facilitate endotracheal intubation. Sufentanil was subsequently titrated to lower doses according to the clinical need defined as the minimal dose required to maintain MAP within 10% of baseline. Propofol was administered at the discretion of the anaesthesiologist to keep the BIS between 40 and 50. Sufentanil was stopped after dural closure. Propofol infusion was reduced to 30–60 μg kg−1 min−1 when the bone flap was replaced, and stopped at the beginning of skin closure.
Intraoperatively, HR and MAP were recorded at five time-points: before anaesthesia induction (baseline), at placement of Mayfield head-frame, and dural incision, during maintenance of anaesthesia after complete tumour resection, and on recovery (awakening). Haemodynamic monitoring was continued until the study ended. Emergence times were measured as recovery time, defined as the time elapsing from the end of propofol infusion to eye opening, and extubation time, defined as the time elapsing from the end of propofol infusion to extubation. Patients were extubated when respiratory function was adequate (frequency <25 and >10 breaths min−1; FiO2 < 0.6; SaO2 > 95%), haemodynamics were stable and upper airway reflexes fully recovered.
Postoperative pain was prevented in all patients with morphine bolus (0.02 mg kg−1) administered at the beginning of skin closure followed by continuous morphine infusion (0.2 mg kg−1 per day for 24 h, in 50 mL of saline at 2 mL h−1 infusion rate) with elastomeric pump (Accufuser, Diaco Biomedicali, Trieste, Italy); before skin dressing the surgical wound was infiltrated with local anaesthetic (10–15 mL 0.5% bupivacaine). Postoperative shivering was prevented with nefopam, a drug that also has analgesic effects . Postoperative pain was evaluated with a 10-step visual analogue scale (with 0 indicating no pain and 10 the worst possible pain) 30 min after extubation and then every 30 min for 3 h. Patient who required additional postoperative analgesia received 1 g i.v. paracetamol.
Postanaesthesia recovery was evaluated according to the Aldrete score  5 min after extubation and every 5 min until the patient reached an Aldrete score of 9–10 and could undergo cognitive testing. If MAP increased more than 20% during awakening patients received esmolol (loading bolus of 500 μg kg−1 min−1, followed by continuous infusion titrated stepwise). Postoperative nausea was prevented with 4 mg i.v. ondansetron administered before extubation.
The primary outcome variable, postoperative (postanaesthesia) cognitive function, was evaluated with two test batteries administered individually to all patients at four time-points: before surgery (baseline), and 15 min after extubation when they reached an Aldrete score of 9–10 and 45 min and 3 h thereafter. The SOMCT is a patient-based test designed to assess cognitive function in terms of level of orientation, memory and concentration [6,7]. The SOMCT requires subjects to recall the current year and month, and one sentence, and to repeat in numerical order and reverse order the sequence of the months through the year. These six variables yield scores ranging from 0 to 28, with higher scores indicating better cognitive function (Table 1). The RLAS is a physician-based scale designed to quantify behavioural and cognitive status after acute brain injuries also used in the postoperative intensive care setting [8,9]. Each state corresponds to one of eight levels, ranging from I to VIII, with higher levels indicating better behavioural and cognitive status (Table 2). Both scales were evaluated pre- and postoperatively by two experienced physicians blind to group assignment. If the observers' score disagreed, the opinion of a third blind observer was considered discriminant.
Assessment of treatment was based on intention-to-treat criteria. We determined that a sample size of 28 patients would be required for each group (remifentanil and sufentanil) to detect a 20% difference in the SOMCT at 45 min after extubation (using alpha and beta values of 0.05 and 0.2). Data were entered into a database and checked by double entry and visual inspection. Paired t-test was used to detect changes in haemodynamics and monitored variables between baseline and subsequent measurements. Intra and intergroup analysis of variance (ANOVA) corrected for repeated measurements was used to test cognitive differences. All data were analysed with SAS (release 8.2). Continuous data were compared using t-tests or Kruskal–Wallis' test when appropriate. For categorical data the χ2 statistic or the Fisher's exact test were used. Results are given as mean ± 1 SD or where appropriate, as mode. A P < 0.05 was considered statistically significant.
Patient characteristics data, including age, gender and weight, were similar in the remifentanil–propofol and sufentanil–propofol groups and no significant difference was found between the two groups in baseline haemodynamics (Table 3). The overall incidence of intraoperative hypotension, hypertension, bradycardia and tachycardia during induction and maintenance of anaesthesia (after placement of the Mayfield head-frame, dural opening and complete tumour resection) was similar in the two groups (Table 4). One patient assigned to the sufentanil group became hypothermic during surgery and was therefore excluded from the study. A total of 59 patients therefore completed the study (30 in the remifentanil–propofol group and 29 in the sufentanil–propofol group).
Patients in the two groups received intraoperatively an equivalent amount of opioid analgesics and hypnotics: the remifentanil–propofol group received a total 5270 ± 2110 μg remifentanil (range 0.1– 0.5 μg kg−1 min−1), and a total 2321 ± 681 mg propofol (range 150–500 μg kg−1 min−1); the sufentanil–propofol group received a total 672 ± 268 μg sufentanil (range 0.01–0.05 μg kg−1 min−1), and a total 2355 ± 691 mg propofol (range 150–500 μg kg−1 min−1).
The mean recovery time was similar in the two groups (13 ± 5 vs. 19 ± 6 min) as was the time of extubation (15 ± 6 vs. 21 ± 8 min), showing that stopping sufentanil infusion during bone flap placement and remifentanil after applying skin dressing led to similar awakening times. In the first 3 h no patients required additional analgesia. Although postoperative Aldrete scores were similar in the two groups, a significantly larger number of patients in the remifentanil group than in the sufentanil group required antihypertensive medication to maintain MAP within 20% of the baseline values (18/30 vs. 4/29; P = 0.0004 by Fisher's exact test). In all cases, postoperative hypertension was transient, responded to esmolol, and esmolol was discontinued within 3 h (mean esmolol infusion time 36 ± 15 min). At 15 min after extubation all patients had an Aldrete score of 9 or 10, and underwent cognitive testing.
In the remifentanil–propofol group the SOMCT score was significantly lower 15 min after extubation than at baseline (22 ± 3 vs. 28 ± 1; P < 0.0001, by t-test) and returned to normal at 45 min and 3 h (27 ± 1 vs. 28 ± 1; 28 ± 1 vs. 28 ± 1) (Figs 1 and 2). In the sufentanil–propofol group the SOMCT score was significantly lower at 15 and 45 min after extubation than at baseline (16 ± 3 vs. 28 ± 1, P < 0.0001; 22 ± 4 vs. 28 ± 1, P < 0.0001) and returned to normal at 3 h (26 ± 2 vs. 28 ± 1). Intergroup analysis showed no differences in baseline SOMCT scores (28 ± 1 vs. 28 ± 1) whereas the mean SOMCT score at 15, 45 min and 3 h after extubation was significantly higher in the remifentanil group (30 patients) than in the sufentanil group (29 patients) (22 ± 3 vs. 16 ± 3; P < 0.0001 and 27 ± 1 vs. 22 ± 3; P < 0.0001; 28 ± 1 vs. 26 ± 2; P = 0.0126). Intragroup and intergroup analysis showed a similar trend for RLAS ratings in the two groups. The 30 patients who received remifentanil had RLAS Grade VIII at baseline (preoperative: 100%), at 15 min after extubation 25 patients (83%) reached RLAS Grade VIII (100% vs. 83%; P = 0.05, by Fisher's exact test), no differences were detectable at 45 min and 3 h after extubation. The 29 patients who received sufentanil had RLAS Grade VIII at baseline (preoperative: 100%), at 15 min after extubation eight patients (28%) reached RLAS Grade VIII, at 45 min after extubation 16 patients (55%) reached RLAS Grade VIII (100% vs. 28%; P < 0.0001 and 100% vs. 55%; P < 0.0001), no differences were detectable at 3 h after extubation.
Intergroup analysis showed no differences in baseline RLAS levels (VIII and VIII) whereas at 15 min and 45 min after extubation RLAS levels was higher in the remifentanil group (30 patients) than in the sufentanil group (29 patients). At 15 min after extubation 25 patients (83%) in the remifentanil group had RLAS Grade VIII and 18 patients (62%) in the sufentanil group had RLAS Grade VII (P < 0.0001). At 45 min after extubation 30 patients (100%) in the remifentanil group had RLAS Grade VIII and 13 patients (44%) in the sufentanil group had RLAS Grade VII (P < 0.0001). At 3 h after extubation the RLAS Grade was similar in the two groups: 30 patients (100%) in the remifentanil group had RLAS Grade VIII and 27 patients (93%) in the sufentanil group had RLAS Grade VIII. The two blinded observers disagreed on fewer than 10% of the ratings (interobserver disagreement <10%). The prevalence of postoperative vomiting was similar in the two groups (two patients in the sufentanil group and one in the remifentanil group). Postoperative visual analogue pain scores were acceptably low and similar in the two groups: 3.2 ± 0.5 and 3.1 ± 0.4 at 30 min after extubation and 2.7 ± 0.9 and 2.9 ± 1.0 at 3 h after extubation. None of the patients had postoperative shivering or neurosurgical complications requiring early postoperative reoperation. One patient in the sufentanil group had respiratory depression at about 30 min after extubation that promptly responded to i.v. naloxone (0.4 mg).
The most distinctive finding in this double-blind study of patients undergoing craniotomy for supratentorial expanding lesions who received neuroanaesthesia with propofol as the hypnotic agent and remifentanil or sufentanil as opioid analgesics is that patients receiving remifentanil–propofol anaesthesia had quicker early postoperative (postanaesthesia) recovery of cognitive function as indicated by the SOMCT and RLAS ratings that returned to baseline values within 45 min after extubation. Whereas in patients receiving sufentanil–propofol cognitive function returned to baseline values 3 h after extubation. This difference in cognitive recovery is clinically advantageous and in a neurosurgical setting could allow surgical complications to be detected earlier.
In our study the remifentanil and sufentanil groups had similar extubation times. This finding is in line with the only other studies that compared these drugs, in cardiac anaesthesia and neuroanaesthesia [4,5]. In our experience, whereas sufentanil makes it easier to obtain stable intraoperative and postoperative haemodynamics, the ultra short-acting opioid remifentanil provides faster cognitive recovery. Having a brighter patient immediately after extubation facilitates early clinical evaluation.
Another clinically relevant finding in this study was that despite similar intraoperative haemodynamics, emergence times (recovery and extubation times), and postoperative pain scores, the need for antihypertensive medication differed significantly in the two treatment groups. Patients receiving remifentanil-based anaesthesia more frequently needed antihypertensive medication immediately after awakening because the MAP rose by more than 20% from baseline values (18/30 vs. 4/29 patients; P = 0.0004). Because we aggressively prevented postoperative pain by giving morphine bolus followed by continuous infusion, infiltrating the wound with local anaesthetics, and infusing nefopam none of our patients required additional analgesia in the first 3 postoperative hours. Hence we consider it unlikely that the hypertensive episodes in the early postoperative period arose from postoperative pain. Whether the higher frequency of hypertensive episodes in the remifentanil group depended on higher postoperative sympathetic tone [5,12,13] – possibly due to acute tolerance and a related withdrawal syndrome after long-term remifentanil administration [14,15] – remains unclear.
Our study confirms the data on intraoperative haemodynamics and awakening times reported by Gerlach and colleagues in patients undergoing remifentanil–propofol or sufentanil–propofol anaesthesia for craniotomy , but provides new clinically important information on postoperative (postanaesthesia) cognitive recovery. The return of cognitive values to baseline in both groups at 3 h after anaesthesia ended, strongly suggest that the transient cognitive impairment was related to anaesthesia rather than to other factors (surgical manoeuvres, pain, hypothermia or cerebral damage inducing persistent postoperative cognitive impairment).
In this study we quantified postoperative cognitive impairment with two scales taken from well-validated tests of cognitive function [6–9]. Both scales (SOMCT and the RLAS) appropriately rated our patients' postoperative cognitive status in less than a minute, far more quickly than other conventional time-consuming scales such as the Mini Mental State Examination (MMSE) and Self Psychological Test (SPT). Rapidity is a distinct advantage in the neurosurgical setting when cognitive status needs to be graded repeatedly. The SOMCT and RLAS combined could therefore provide a useful means for the early assessment of postoperative (postanaesthesia) cognitive function in future studies.
In conclusion, for total intravenous anaesthesia during surgery for supratentorial expanding lesions remifentanil and sufentanil both provide adequate intraoperative haemodynamic stability and analgesia. Although remifentanil needs close titration and is more frequently associated with treatable postoperative hypertension than sufentanil it provides faster cognitive recovery. Having a bright patient with near normal cortical function may facilitate postoperative neurosurgical evaluation and expedite the diagnosis should a complication develop.
This work was partly financed by departmental funding with no corporate sponsor. The Authors wish to thank Prof. Concezione Tommasino for her constructive review of the article.
1. Bruder NJ. Awakening management after neurosurgery for intracranial tumors. Curr Opin Anesthesiol
2. Coles JP, Timothy SL, Monteiro JN et al
anesthesia for craniotomy: a double-blind comparison of remifentanil
, alfentanil and fentanyl. J Neurosurg Anesthesiol
3. Balakrishnan G, Raudznes P, Samra SK et al
. A comparison of remifentanil
and fentanyl in patients undergoing surgery for intracranial mass lesion. Anesth Analg
4. Ergoren H, Luther G, Fenn-Buderer N. A comparison of fentanyl, sufentanyl and remifentanil
for fast-track cardiac anesthesia. Anesth Analg
5. Gerlach K, Uhlig T, Huppe M et al
anaesthesia for supratentorial craniotomy: a randomized trial. Eur J Anaesth
6. Davous P, Lamour Y, Debraud E et al
. A comparative evaluation of the short orientation memory concentration test of cognitive impairment. J Neurol Neurosurg Psych
7. Katzman R, Brown T, Fuld P, Peck A, Schechter R, Schimmel H. Validation of a short orientation-memory-concentration test of cognitive impairment. Am J Psychiat
8. Perry J, Garrett M, Gronley JK, Mulroy SJ. Classification of walking handicap in the stroke population. Stroke
9. Beauchamp K, Baker S, McDaniel C et al
. Reliability of nurses' neurological assessments in the cardiothoracic surgical intensive care unit. Am J Crit Care
10. Bilotta F, Pietropaoli P, La Rosa I et al
. Effects of shivering prevention on haemodynamic and metabolic demands in hypothermic postoperative neurosurgical patients. Anaesthesia
11. Aldrete JA. The post-anaesthesia recovery score revisited. J Clin Anesth
12. Muellejans B, Lopez A, Cross MH et al
versus fentanyl for analgesia based sedation to provide patient comfort in the intensive care unit: a randomized double blind controlled trial. Crit Care
13. Gelb AW, Salevsky F, Chung F et al
with morphine transitional analgesia shortens neurological recovery compared to fentanyl for supratentorial craniotomy. Can J Anesth
14. Hood DD, Curry R, Eisenach JC. Intravenous remifentanil
produces withdrawal hyperalgesia in volunteers with capsaicin-induced hyperalgesia. Anesth Analg
15. Delvaux B, Ryckwaert Y, Van Boven M et al
in the intensive care unit: tolerance and acute withdrawal syndrome after prolonged sedation. Anesthesiology
Keywords:© 2007 European Society of Anaesthesiology
NEUROANAESTHESIA; POSTOPERATIVE COGNITIVE RECOVERY; TOTAL INTRAVENOUS ANAESTHESIA; PROPOFOL; REMIFENTANIL; SUFENTANIL