Editor s key points. M. S. Green*, P. Green, S. N. Malayaman, M. Hepler, L. J. Neubert and J. C. Horrow

Transcription

1 British Journal of Anaesthesia 109 (5): (2012) Advance Access publication 24 July doi: /bja/aes233 Randomized, double-blind comparison of oral aprepitant alone compared with aprepitant and transdermal scopolamine for prevention of postoperative nausea and vomiting M. S. Green*, P. Green, S. N. Malayaman, M. Hepler, L. J. Neubert and J. C. Horrow Department of Anesthesiology, Drexel University College of Medicine, Hahnemann University Hospital, Philadelphia, PA, USA * Corresponding author: Department of Anesthesiology and Perioperative Medicine, 245 North 15th Street, Mail Stop 310, Philadelphia, PA 19102, USA. michael.green@drexelmed.edu Editor s key points Postoperative nausea and vomiting (PONV) can be a challenging problem, especially in high-risk patients. Multimodal antiemetic therapies may provide a way to reduce PONV. This study examines the effects of combining oral aprepitant and transdermal scopolamine on PONV. There was no improvement in PONV nor increase in side-effects by combination therapy. Further research is needed in this area. Background. Aprepitant blocks the emetic effects of substance P. Scopolamine antagonizes muscarinic type 1 and histamine type 1 receptors. This study compares monotherapy and multimodal therapy by looking at complete response, nausea, vomiting, and rescue medication in patients at high risk for postoperative nausea and vomiting (PONV) treated with oral aprepitant with or without scopolamine. Methods. We enrolled 120 patients in this randomized, double-blind trial. Inclusion criteria were:.18 yr old, ASA I III, two or more Apfel four-point risk factors, undergoing an elective surgical procedure with a high risk of PONV expected to last at least 60 min. The primary outcome variable was complete response, that is, no emesis and no rescue therapy from 0 to 24 h. The outcomes measured included the incidences of nausea, vomiting, their composite, and the need for rescue medication. Results. The aprepitant alone and aprepitant with scopolamine did not differ in complete responses (63% vs 57%, P¼0.57) or net clinical benefit (26% vs 19%, P¼0.38). The number who did not experience PONV and who used rescue medication did not differ. The incidence of PONV in the post-anaesthesia care unit did not differ nor did the use of rescue medications. Conclusions. This trial evaluating the effectiveness of aprepitant alone and in combination with scopolamine showed no difference between treatment groups. The primary objective, complete response, and secondary objectives, incidences of nausea, vomiting, their composite, and the need for rescue medication, all showed no statistical difference. Keywords: autonomic agents, antiemetics; central nervous system agents, adjuvants anaesthesia; postoperative nausea and vomiting; receptors, neurokinin-1; tropanes, scopolamine Accepted for publication: 14 April 2012 Aprepitant, a selective antagonist of neurokinin-1 (NK-1) receptors, blocks the emetic effects of substance P. 1 NK-1 receptors are found on vagal afferents in the gastrointestinal tract and in the nucleus tractus solitarius in the brain. Aprepitant is equally as effective as in the prevention of postoperative nausea and rescue antiemetic use and has better control of vomiting at 24 and 48 h when compared with conventional therapies. 2 Scopolamine antagonizes muscarinic type 1 and histamine type 1 receptors in the central nervous system, hypothalamus, and vomiting centre. The noradrenergic system is also suppressed resulting in a diminished response to vestibular stimulation. 3 A consensus panel from Society of Ambulatory Anesthesia recommended scopolamine as an effective therapeutic agent for postoperative nausea and vomiting (PONV). 4 Multimodal therapy with transdermal scopolamine (TDS) in combination with ondansetron (OND) has been show to be superior to monotherapy with OND alone with no difference in the incidence of anticholinergic-related side-effects. 5 PONV risk in adults undergoing general anaesthesia with inhalation anaesthetic agents can be predicted by four factors: female sex, history of PONV or motion sickness, non-smoking status, and the use of postoperative opioids. The frequency of PONV is 10% with zero, 21% with one, 39% with two, 61% with three, and 79% with four risk factors. 67 The primary objective of this study was to test the hypothesis that in patients with high risk for PONV, the use of & The Author [2012]. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved. For Permissions, please journals.permissions@oup.com

2 Aprepitant monotherapy to prevent postoperative nausea aprepitant alone is as effective at producing a complete antiemetic response (no emesis and no rescue therapy from 0 to 24 h) compared with aprepitant and TDS in combination in the first 24 h after operation. Secondary objectives included the incidences of nausea, vomiting, their composite, and the need for rescue medication. Methods After approval from the institutional review board, all patients provided written informed consent for this randomized, double-blind trial (clinicaltrials.gov registry # NCT ). Study inclusion criteria consisted of age at least 18 yr, ASA physical status I III, elective high-risk PONV surgery expected to last at least 60 min requiring general anaesthesia, and two or more Apfel four-point risk factors for PONV. Exclusion criteria consisted of pregnant or breast-feeding patients, those treated with antiemetic medications within 24 h of their procedure, history of vomiting from identified non-surgical causes, and possession of an allergy or other contraindication to study medications. Patients having ambulatory surgical procedures and patients requiring admission were included. A computer-generated process randomly assigned subjects using a sealed envelope technique to either the oral aprepitant or the combined treatment group. Each subject received a transdermal patch and an aprepitant 40 mg pill. Patients ingested the pill and a study investigator applied the transdermal patch over the mastoid area, at least 1 h before anaesthesia induction. Depending on the assigned group, the transdermal patch was either a placebo or active medication. All clinicians were blinded to the assigned group. Premedication with i.v. midazolam, 1 2 mg, preceded transfer to the operating suite. Propofol 2 3 mg kg 21 induced general anaesthesia. Tracheal intubation, facilitated by rocuronium mg kg 21 or placement of a laryngeal mask airway established ventilation. Volatile anaesthetics (sevoflurane, desflurane, or isoflurane) maintained general anaesthesia. Clinical judgement guided titration of volatile anaesthetic concentrations. Fentanyl boluses of mg kg 21 provided analgesia. Neostigmine 70 mg kg 21 and glycopyrronium 10 mg kg 21 provided reversal from neuromuscular block if necessary. Nitrous oxide was prohibited. In the post-anaesthesia care unit (PACU), patients received supplemental oxygen via a nasal cannula. According to the clinical judgement of the anaesthesiologist, fentanyl boluses of mg kg 21 and morphine boluses of 50 mg kg 21 controlled postoperative pain. As rescue medication, OND 4 mg i.v. treated postoperative emesis or nausea if symptoms lasted longer than 15 min or if the patient requested treatment. Anaesthesiologists discretion guided further necessary treatment. Oral administration of OND 4 mg and metoclopramide 10 mg provided post-discharge nausea and vomiting treatment. TDS or placebo remained in place for 24 h from PACU arrival time. Study investigators removed the patch for admitted patients during 24 h postoperative follow-up visits. Discharged patients received written instructions for the correct timing of patch removal and received a follow-up phone call by study investigators at 24 h. The primary outcome variable was complete response, that is, no emesis and no rescue therapy from 0 to 24 h. Other outcomes measured included the incidences of nausea, vomiting, their composite, and the need for rescue medication. Trained investigators blinded to the treatment group collected nausea and vomiting information at PACU admission and every 15 min for 1 h, at 2 h, and again at 24 h after operation. The time from PACU admission until achieving a modified Aldrete score of 9, assessed every 15 min, determined the duration of PACU stay. For those patients admitted to the hospital, the investigator collected data on nausea, vomiting, rescue medication use, and sideeffects in the hospital at 24 h. For discharged patients, telephone contact at 24 h assessed nausea, vomiting, or both, use of rescue medication, and other listed side-effects after discharge. The patients were asked to rate their nausea on a 0 10 point scale, with 0 being no nausea and 10 being severe nausea. Data on side-effects specifically queried visual disturbances, dry mouth, dizziness, and agitation, and then asked for additional volunteered symptoms. Statistical analysis plan The two pivotal trials establishing the efficacy of aprepitant for PONV prophylaxis demonstrated a 63.8% and 44.8% complete response, respectively, with mean 54%. 8 Scopolamine alone also yields a complete response of 54%. 9 Based on an expected incidence of 54% for the primary outcome variable complete response in the aprepitant-alone-treated group, with a two-sided type I error of 5% and type II error of 80%, each group needed to collect valid data on 54 patients to demonstrate a minimally detectable increase of 1.5-fold the 54% individual rates, to 81% absolute complete response, with combined therapy. 10 Assuming a dropout rate of 10%, each group needed to enrol 60 patients to obtain 54 with valid data. We pre-specified the net clinical benefit of aprepitant compared with combined therapy as the fraction of patients free of the composite outcome of nausea, vomiting, visual disturbances, dry mouth, dizziness, or agitation. The primary outcome variable was complete response, that is, no emesis and no rescue therapy from 0 to 24 h. Student s t-test was used to compare continuous patient characteristic variables of the groups, while the Fisher exact test compares frequency data and outcome variable incidences. All comparisons utilize a two-sided 5% significance level. Standard formulae provide relative risk reduction and number needed to treat for outcome variables and for the net clinical benefit. Results Of 120 patients randomized, 119 received study medications and 115 completed the trial. Figure 1 illustrates the patient 717

3 Green et al. 1 patient excluded due to prior participation 120 eligible patients randomized 119 received study medications 58 patients received aprepitant alone 61 patients received aprepitant and scopolamine 1 patient excluded due to use of neuroaxial anaesthesia 2 patients excluded due to use of neuroaxial anaesthesia and 1 postponed due to medical condition 57 patients per protocol population 58 patients per protocol population Approximately 4 patients screened for each entered in the study Fig 1 Patient screening and randomization flow chart. screening and random assignment process. Prior participation in the study excluded one randomized subject before medication administration. The use of a neuraxial anaesthesia technique after medication administration excluded three patients. One patient had surgery postponed due to an upper respiratory tract infection after medication administration. The safety analysis set included 119 patients and the efficacy analysis set included 115 patients with outcome data. The two treatment groups did not differ with regard to patient characteristic variables, including number and type of risk factors, and clinical characteristics (Table 1). The type of surgical procedures did not differ between groups. Efficacy The aprepitant alone group and the aprepitant with scopolamine group did not differ in complete response (63% vs 57%, P¼0.57) or net clinical benefit (26% vs 19%, P¼0. 38; Table 2). The total number of patients who did not experience PONV (51% vs 41%, P¼0.35) and the number of patients who utilized rescue medication (37% vs 40%, P¼0.84) did not differ between the monotherapy and two drug therapy groups. The incidence of PONV in the PACU did not differ between two groups (68% vs 60%, P¼0.44) nor did the use of rescue medications in the PACU (19% vs 28%, P¼0.38). The total incidence of vomiting did not differ between the two treatment groups (3.5% vs 8.6%, P¼0.44). There was no difference between the groups in the incidence of home nausea, home vomiting, or home rescue medication (Table 2). The incidence of nausea, vomiting, and the number of patients receiving rescue medication did not differ at any given time point during data collection (Table 3). As expected, the incidence of PONV increased with increasing number of risk factors (P¼0.0017). Likewise, the complete response rate declined with the increasing number of risk factors (P¼0.0011): for two or three risk factors, a complete response occurred in 77% and 71% of the patients, respectively; four risk factors cut this response to 39% (Table 4). The overall incidence of all adverse events did not differ between treatment groups (21% vs 26%, P¼0.48). Nor did the overall incidence of the anticholinergic side-effects dry mouth, blurry vision, dizziness, and agitation differ between treatment groups (P¼0.54; Table 5). Additional side-effects included three people experiencing difficult urination, three people experiencing dyspepsia, two people affected by memory loss, two people with itching at the site of transdermal patch application, one person with constipation, and one person with throat pain. Discussion When aprepitant was ingested orally at least 1 h before induction of anaesthesia, the result was as effective as treating 718

4 Aprepitant monotherapy to prevent postoperative nausea with both oral aprepitant and TDS at least 1 h before induction of anaesthesia. The primary outcome variable, complete response, defined as no emesis and no rescue therapy from 0 to 24 h, showed no statistically significant difference Table 1 Patient characteristic data by treatment group and clinical characteristics Variable Aprepitant alone, mean (SD) (N557) Aprepitant and scopolamine, mean (SD) (N558) P-value (t-test) Age (yr), range Height (cm) (7.5) (9.7) Weight (kg) 84.3 (24.82) 78.7 (21.36) ASA class [n (%)] N¼58 N¼61 I 9 7 II III 8 15 Risk factors [n (%)] Female sex 52 (91) 51 (88) History of PONV/ 32 (55) 33 (54) motion sickness Non-smoker 41 (72) 38 (66) Planned postoperative 55 (96) 57 (98) opioid utilization Total risk factors per patient, mean (SD) 3.14 (0.76) 3.07 (0.75) Clinical variables, mean (SD) Fluid given (ml) 1940 (1159) 2210 (1292) 0.25 Procedure duration (83.9) (90.7) 0.76 (min) Urine output (ml) 306 (264) 264 (144) 0.38 Blood loss (ml) 142 (276) 159 (300) 0.76 Aldrete score (PACU) 8.42 (0.706) 8.60 (0.815) 0.20 Aldrete score at 1 h 9.11 (0.698) 9.25 (0.739) 0.34 Admitted patients [n (%)] 34 (60) 37 (64) between treatment groups. The use of rescue medication and the need for unplanned hospital admission did not differ between the monotherapy and combination therapy groups. The two patient populations did not differ with respect to patient characteristic variables, ASA classification, clinical variables, the specific type of procedures performed, or their PONV risk factor profiles. Aprepitant, an NK-1 antagonist, works both peripherally and centrally to prevent PONV. Stimulation from peripheral and central afferents, through the lateral reticular formation, yields the sequence of events within the medulla oblongata to produce vomiting. 11 Inputs from the vestibular system, higher cortical centres, the cerebellum, glossopharyngeal nerve stimulation, and vagal afferents can produce central vomiting activation. 11 The NK-1 receptors are located in the area postrema, the central pattern generator for vomiting and/or afferent relay station, and nucleus tractus solitarius. These areas play important roles in emesis via these receptors. 12 Aprepitant has proven efficacy in the prevention of PONV. Habib and colleagues 13 found the combination of aprepitant and dexamethasone to be more effective than the combination of OND and dexamethasone for prophylaxis against PONV in adult patients undergoing craniotomy. However, no difference occurred between the groups in the incidence or severity of nausea alone, the need for rescue antiemetics, or in complete response. 13 In 805 patients receiving general anaesthesia for open abdominal surgery, aprepitant monotherapy was superior to OND monotherapy for the prevention of vomiting in the first 24 and 48 h, but no different for nausea control, use of rescue therapy, or complete response. 2 Diemunsch and colleagues found aprepitant monotherapy to be non-inferior to OND for achieving complete response for 24 h after surgery in 922 patients undergoing open abdominal surgery. It was also more effective than OND for preventing vomiting 24 and 48 h after surgery and in reducing nausea severity in the first 48 h after surgery. 14 The efficacy of TDS in the prevention of PONV is well established Three recent studies support its use in the Table 2 Complete response, net clinical benefit, time and incidence of nausea and vomiting, use of rescue antiemetic, and unplanned hospital admission. PONV, postoperative nausea and vomiting; PACU post-anaesthesia care unit Variable Aprepitant alone (n557) Group % Aprepitant and scopolamine (n558) Group % Complete response Net clinical benefit No PONV in PACU PACU rescue medication No PONV from 0 to 24 h Total vomiting, 0-24 h Total rescue medication required Home nausea Home vomiting Home medication P-value 719

5 Green et al. Table 3 Incidence of nausea, vomiting, and antiemetic treatment at different time points. *t, time point Variable Aprepitant alone, N557 Per cent Aprepitant and scopolamine, N558 Per cent Nausea, t*¼ Vomiting, t¼ Rescue antiemetic Nausea, t¼15 min Vomiting, t¼15 min Rescue antiemetic Nausea, t¼30 min Vomiting, t¼30 min Rescue antiemetic Nausea, t¼45 min Vomiting, t¼45 min 0 0 Rescue antiemetic Nausea, t¼60 min Vomiting, t¼60 min 0 0 Rescue antiemetic Nausea, t¼2 h Vomiting, t¼2 h 0 0 Rescue antiemetic P-value Table 4 Number of risk factors and the incidence of postoperative nausea and vomiting and complete response. *For total postoperative nausea and vomiting, the Jonckheere Terpstra test for ordered alternatives, two-sided P¼ For complete response, the Jonckheere Terpstra test for ordered alternatives, two-sided P¼ Risk factors Number Postoperative nausea and vomiting (%)* Complete response (31) 20 (77) (53) 36 (71) (71) 15 (39) prevention of PONV in combination with other therapies. Gan and colleagues 5 concluded that TDS in combination with OND reduced PONV compared with OND alone after outpatient surgery and was associated with greater patient satisfaction. Sah and colleagues studied high-risk patients undergoing outpatient plastic surgery. OND+TDS provided a significant reduction in postoperative nausea between 8 and 24 h compared with OND alone. 15 Lee and colleagues used TDS and dexamethasone for the prevention of PONV in patients with epidural patient-controlled analgesia after major orthopaedic surgery. That combination more effectively prevented PONV than dexamethasone alone or than dexamethasone plus ramosetron. 16 The TDS delivery system allows constant uptake of medication through the skin for 72 h. This slow release provides a more constant medication level, thus minimizing medication-related anticholinergic side-effects. 5 This current study supports this premise because no significant statistical difference in anticholinergic side-effects between the two groups was found. This is in contrast to a meta-analysis done by Kranke and colleagues. In their review of 23 randomized, controlled trials published between 1984 and 1996, they found for every 100 patients who receive TDS, 17 will not experience postoperative vomiting. However, 18 of 100 patients will have visual disturbances, 8 dry mouth, 2 dizziness, and 1 agitation. 9 In another meta-analysis completed by Apfel and colleagues including 25 randomized controlled trials and 3298 patients, TDS proved to be valuable in the prevention of both early and late PONV during the first 24 h after the start of anaesthesia. 17 In this meta-analysis, TDS was associated with a higher incidence of visual disturbances at h but no other adverse events, compared with placebo. 17 Gan and colleagues studied 618 patients in two groups. The incidence of anticholinergic side-effects between the group that received TDS and those who did not were not different. 5 Our current study found similar results with regard to anticholinergic side-effects. Multimodal therapy for the prevention of PONV is a widely accepted treatment approach. 18 In 2004, Apfel and colleagues published a large multicentre study of more than 5000 patients showing each intervention to act independently while decreasing the incidence of PONV by 25 30% in an additive manner but not synergistic. 19 Other studies have drawn similar conclusions. Singla and colleagues studied multimodal therapy using the NK-1 receptor antagonist casopitant in over 700 patients with a history of PONV or motion sickness undergoing a laparoscopic or laparotomic gynaecological surgical procedure or laparoscopic cholecystectomy with general anaesthesia. 20 Compared with OND monotherapy, casopitant and OND provided superior 720

6 Aprepitant monotherapy to prevent postoperative nausea Table 5 Number of side-effects Anticholinergic side-effect Aprepitant alone, N557 Per cent Aprepitant and scopolamine, N558 Per cent P-value Dry mouth Dizziness Visual disturbance Agitation Composite of above emesis prevention during the first 24 h after operation, thus supporting multimodal therapy while using NK-1 receptor antagonists. 20 Our study found that two drug therapy did not prove more efficacious in the prevention of PONV. Several factors might explain the different outcomes: inherent differences between the two NK-1 receptor antagonists, different second therapies, OND and TDS, different data collection, surgical procedures, or opioid administration. Different study designs may have contributed. The time of administration, both medications take prolonged periods of time to reach maximal effect, may have contributed to lack of difference. A trend towards a relative risk reduction in PONV did exist when comparing the two groups. The current study may have been underpowered to show a significant difference. The combination of two long-acting medications, TDS and aprepitant, may be a poor choice of combination therapy. A quicker onset shorter acting medication might complement aprepitant better. In summary, this randomized, double-blind trial evaluating the effectiveness of aprepitant alone and in combination with TDS for the prevention of PONV showed no difference between treatment groups. The primary objective, complete response, and secondary objectives, incidences of nausea, vomiting, their composite, and the need for rescue medication, all showed no statistical difference. The results of this study question the combination of aprepitant and TDS for prophylaxis of PONV. Funding This work was supported by a small project grant received from Merck & Co. (M.S.G.). Declaration of interest The concept of the project was originated by M.S.G., DO, who served as the Principal Investigator. Merck & Co. reviewed the protocol before issuing any funding. Funds received paid for medications administered, research personnel, pharmacy administrative fees, Western IRB fees, and statistical analysis. Merck & Co. had no involvement in the analysis of the data or the preparation of the manuscript. The manuscript was reviewed and approved by Merck & Co. before submission with no changes made. M.S.G. has received a small project grant from Merck & Co. for the completion of this research project. P.G. s husband has received a small project grant from Merck & Co. for the completion of this research project. References 1 Massaro AM, Lenz KL. Aprepitant: a novel antiemetic for chemotherapy-induced nausea and vomiting. Ann Pharmacother 2005; 39: Gan TJ, Apfel CC, Kovac A, et al. Aprepitant-PONV Study Group. A randomized, double-blind comparison of the NK1 antagonist, aprepitant, versus ondansetron for the prevention of postoperative nausea and vomiting. Anesth Analg 2007; 104: Takeda N, Morita M, Horii A, Nishiike S, Kitahara T, Uno A. Neural mechanisms of motion sickness. J Med Invest. 2001; 48: Gan TJ, Meyer TA, Apfel CC, et al. Society for ambulatory anesthesia guidelines for the management of postoperative nausea and vomiting. Anesth Analg 2007; 105: Gan TJ, Sinha AC, Kovac AL, et al. TDS Study Group. A randomized, double-blind, multicenter trial comparing transdermal scopolamine plus ondansetron to ondansetron alone for the prevention of postoperative nausea and vomiting in the outpatient setting. Anesth Analg 2009; 108: Koivuranta M, Laara E, Snare L, Alahuhta S. A survey of postoperative nausea and vomiting. Anaesthesia 1997; 52: Apfel CC, Laara E, Koivuranta M, Greim CA, Roewer N. A simplified risk score for predicting postoperative nausea and vomiting: conclusions from cross-validations between two centers. Anesthesiology 1999; 91: EMEND w Product Label, Merck 9 Kranke P, Morin AM, Roewer N, Wulf H, Eberhart LH. The efficacy and safety of transdermal scopolamine for the prevention of postoperative nausea and vomiting: a quantitative systematic review. Anesth Analg 2002; 95: Available from (accessed 14 February 2007) 11 Hornby PJ. Central neurocircuitry associated with emesis. Am J Med 2001; 111: 106S 112S 12 Saito R, Takano Y, Kamiya H. Roles of substance P and NK1 receptor in the brainstem in the development of emesis. J Pharmacol Sci 2003; 91: Habib AS, Keifer JC, Borel CO, White WD, Gan TJ. A Comparison of the combination of aprepitant and dexamethasone versus the combination of ondansetron and dexamethasone for the prevention of postoperative nausea and vomiting in patients undergoing craniotomy. Anesth Analg 2011; 112: Diemunsch P, Gan TJ, Philip BK, et al. Aprepitant-PONV Protocol 091 International Study Group. Single-dose aprepitant vs ondansetron for the prevention of postoperative nausea and vomiting: a randomized, double-blind phase III trial in patients undergoing open abdominal surgery. Br J Anaesth 2007; 99:

7 Green et al. 15 Sah N, Ramesh V, Kaul B, Dalby P, Shestak K, Vallejo MC. Transdermal scopolamine patch in addition to ondansetron for postoperative nausea and vomiting prophylaxis in patients undergoing ambulatory cosmetic surgery. JClinAnesth2009; 21: Lee HK, Lee JH, Chon SS, Ahn EK, Kim JH, Jang Y. The effect of transdermal scopolamine plus intravenous dexamethasone for the prevention of postoperative nausea and vomiting in patients with epidural PCA after major orthopedic surgery. Korean J Anesthesiol 2010; 58: Apfel CC, Zhang K, George E, et al. Transdermal scopolamine for the prevention of postoperative nausea and vomiting: a systematic review and meta-analysis. Clin Ther 2010; 32: Chandrankantan A, Glass PS. Multimodal therapies for postoperative nausea and vomiting, and pain. Br J Anaesth 2011; 107(Suppl. 1): i Apfel CC, Korttila K, Abdalla M, et al. IMPACT Investigators. A factorial trial of six interventions for the prevention of postoperative nausea and vomiting. N Engl J Med 2004; 350: Singla NK, Singla SK, Chung F, et al. Phase II study to evaluate the safety and efficacy of the oral neurokinin-1 receptor antagonist casopitant administered with ondansetron for the prevention of postoperative and postdischarge nausea and vomiting in highrisk patients. Anesthesiology 2010; 113: