Portal de Periódicos da CAPES

Comparative efficacy and safety of remifentanil and fentanyl in ‘fast track’ coronary artery bypass graft surgery: a randomized, double-blind study

2001; Elsevier BV; Volume: 87; Issue: 5 Linguagem: Inglês

10.1093/bja/87.5.718

1471-6771

T. Mo ̈llhoff, Luc Herregods, Anneliese Moerman, D. Blake, Charles MacAdams, R. Demeyere, Klaus Kirnö, T Dybvik, Shiraj Shaikh, the Remifentanil Study Group,

Anesthesia and Neurotoxicity Research

This multi-centre, parallel group, randomized, double-blind study compared the efficacy and safety of high-dose remifentanil administered by continuous infusion with an intermittent bolus fentanyl regimen, when given in combination with propofol for general anaesthesia in 321 patients undergoing elective coronary artery bypass graft surgery. A significantly lower proportion of the patients who received remifentanil had responses to maximal sternal spread (the primary efficacy endpoint) compared with those who received fentanyl (11% vs 52%; P<0.001). More patients who received remifentanil responded to tracheal intubation compared with those who received fentanyl (24% vs 9%; P<0.001). However, fewer patients who received remifentanil responded to sternal skin incision (11% vs 36%; P<0.001) and sternotomy (14% vs 60%; P <0.001). Median time to extubation was longer in the subjects who received remifentanil than for those who received fentanyl (5.1 vs 4.2 h; P=0.006). There were no statistically significant differences between the two groups in the times for transfer from intensive care unit or hospital discharge but time to extubation was significantly longer in the remifentanil group. Overall, the incidence of adverse events was similar but greater in the remifentanil group with respect to shivering (P<0.049) and hypertension (P<0.001). Significantly more drug-related adverse events were reported in the remifentanil group (P=0.016) There were no drug-related adverse cardiac outcomes and no deaths from cardiac causes before hospital discharge in either treatment group. This multi-centre, parallel group, randomized, double-blind study compared the efficacy and safety of high-dose remifentanil administered by continuous infusion with an intermittent bolus fentanyl regimen, when given in combination with propofol for general anaesthesia in 321 patients undergoing elective coronary artery bypass graft surgery. A significantly lower proportion of the patients who received remifentanil had responses to maximal sternal spread (the primary efficacy endpoint) compared with those who received fentanyl (11% vs 52%; P<0.001). More patients who received remifentanil responded to tracheal intubation compared with those who received fentanyl (24% vs 9%; P<0.001). However, fewer patients who received remifentanil responded to sternal skin incision (11% vs 36%; P<0.001) and sternotomy (14% vs 60%; P <0.001). Median time to extubation was longer in the subjects who received remifentanil than for those who received fentanyl (5.1 vs 4.2 h; P=0.006). There were no statistically significant differences between the two groups in the times for transfer from intensive care unit or hospital discharge but time to extubation was significantly longer in the remifentanil group. Overall, the incidence of adverse events was similar but greater in the remifentanil group with respect to shivering (P<0.049) and hypertension (P 0.3, undergoing elective CABG surgery were randomized to receive either remifentanil or fentanyl as part of a total i.v. anaesthesia technique with propofol. A total of 321 patients were recruited, of whom 24 were treated on an open label basis as 'pilot' patients and 297 were randomized to treatment. All patients included in the study were potentially eligible for early extubation/ICU discharge. Patients with severely impaired left ventricular function (ejection fraction 15 mm Hg above preoperative baseline or >140 mm Hg for ≥1 min; arterial pressure (AP) >80 mm Hg for ≥1 min during bypass. Heart rate (HR) >90 beats min−1 for ≥1 min (pre-bypass); non-paced HR >100 beats min−1 for ≥1 min (post-bypass). Somatic responses: gross movement, swallowing, grimacing, eye opening. Autonomic responses: lachrymation, sweating. Inadequate anaesthesia was treated in a sequential manner beginning with simultaneous bolus dose administration (remifentanil 1 μg kg−1, fentanyl 2 μg kg−1) and infusion increments equivalent to 0.5 or 1.0 μg kg−1 min−1 remifentanil (unless there were intervening decreases). A maximum of three to six maintenance rate increases were allowed up to a maximum infusion rate of remifentanil 4 μg kg−1 min−1 or placebo equivalent. All infusion rate increases were preceded by a bolus dose. If a response was not controlled within 5 min of adjusting the opioid infusion rate or if the maximum infusion rate was reached, the propofol infusion rate was increased as required. If the response was still not controlled, other agents (e.g. sodium nitroprusside, beta-blockers) could be given. Once the patient had achieved a stable, non-responding status, the propofol infusion rate was then titrated back down but the opioid maintenance infusion was left at the higher rate unless hypotension occurred. Excessive anaesthesia was defined as one or more of the following: SBP <80 mm Hg for ≥1 min; AP <40 mm Hg for ≥1 min during bypass. HR 7.25, PaCO2 ≤7.33 kPa and ventilatory frequency adequate to maintain oxygenation. If the patient had not met the criteria for beginning the extubation sequence by 4.5 h after entry into the ICU, the analgesic opioid bolus was administered at this time and the down-titration of the study opioid was then started 30 min later. Patients were eligible for transfer from ICU if stable conditions continued. Vital signs were recorded immediately before induction of anaesthesia (baseline values) and at regular intervals throughout surgery and in the ICU. These consisted of SBP, DBP, MAP, HR, CVP, and SpO2. During the pre-bypass period, patients' vital signs were recorded 1 min before and every minute for 5 min after each major surgical stress event (MSE), then every 15 min from the MSE until the next MSE. MSEs were defined as intubation, sternal skin incision, sternotomy, maximal sternal spread (MSS), and sternal wire placement. Arterial pressure was recorded at the start of cardiopulmonary bypass and every 15 min until the patient went off bypass. After entry into the ICU, patients' vital signs were recorded immediately and at 15 min intervals thereafter until extubation. In addition, vital signs, pain, and sedation scores were recorded every 10 min during the down-titration of study opioid. Pain was assessed using a four-point patient self-rated scale (0=no pain, 1=mild, 2=moderate, 3=severe), while sedation was assessed using a five-point scale (1=fully awake and orientated, 5=eyes closed and cannot be roused by mild physical stimulation). Vital signs and pain and sedation scores were also recorded at 15, 30, 45, 60, 90, and 120 min after discontinuation of study opioid infusion. 12-lead ECG recordings were obtained within 14 days before surgery and on postoperative day 1, and at the end of day 5 or on the morning of day 6 (or day of hospital discharge if prior to this). Blood samples were obtained at 8, 16, 24, and 48 h after aortic cross clamp release for analysis of plasma cardiac enzyme concentrations (CK-CKMB). The primary efficacy endpoint was response to MSS. Secondary efficacy endpoints included responses to intubation, sternal skin incision, sternotomy, and sternal wire placement. Efficacy was evaluated by the number of patients who showed signs of inadequate anaesthesia and by the number of treatments for responses indicative of inadequate or excessive anaesthesia. Overall use of study drug, propofol and other medications were noted. Patients were continually assessed for occurrence of adverse events (including negative cardiac outcomes) throughout the perioperative period and up to the end of postoperative day 5 (or up to hospital discharge if this occurred earlier). Negative cardiac outcomes were defined as ventricular failure (requirement for postoperative intra-aortic balloon pump or ventricular assist device), myocardial infarction (CK-MB >50 unit litre−1 and presence of new Q-waves on 12-lead ECG) or death from cardiac causes (e.g. ventricular failure or congestive heart failure) before hospital discharge. The primary endpoint assumed a response rate of 7% and 18% for patients receiving remifentanil and fentanyl, respectively. One hundred and thirty five patients per treatment would have a power of 80% to detect this difference at a two-sided 5% level of significance. To compensate for withdrawals and to obtain additional safety data, up to 300 randomized patients were to be included in the study. The study was stopped when 297 patients had been recruited because of time restriction. Patients were randomized to receive either remifentanil or fentanyl by assignment of treatment randomization details contained in a hidden entry envelope, which had been computer-generated before the study set-up. The hidden entry envelopes were allocated sequentially to each subject eligible for study entry and the corresponding treatment allocation and drug preparation was performed by a named person, independent to the conduct of the study. In most instances this was the hospital pharmacist. All tests of statistical significance were two-sided and carried out at the 5% level. Logistic regression analyses were used to analyse the proportions of patients with responses to MSS. Estimates of the odds ratios and 95% confidence intervals were calculated. The weighted mean pain and sedation scores during the ICU down titration were analysed using the Wilcoxon rank sum test. The treatment differences in overall use of alternate analgesics until extubation, were analysed using logistic regression analysis. The incidences of the most commonly reported adverse events (defined as occurring in at least 5% of patients in any treatment group) were analysed using Fisher's exact test. Three hundred and twenty one patients entered the study, 24 were treated on an open label basis and were not included in the intent-to-treat population for efficacy analyses although they were included in the population for safety data. The intent-to-treat population consisted of 297 patients, of whom 148 patients received remifentanil and 149 received fentanyl. The two groups were well matched with regard to gender, ethnic origin, age, height, weight, and ASA status distribution (Table 1). The mean ejection fraction was similar in the remifentanil (0.62) and fentanyl (0.63) groups. Most patients required three arteries to be grafted (41% in the remifentanil group and 45% in the fentanyl group). Two patients in each group had previous CABG surgery. The median durations of surgery, of bypass, and of aortic cross clamping were comparable between the two groups. The median (sd) time from the start of study drug to skin closure was 3.67 h (1.85, 6.3) in the remifentanil group and 3.58 h (1.97, 8.7) in the fentanyl group. The median (sd) durations of bypass and of aortic cross clamping were 1.31 (0.45, 2.4) and 0.75 h (0.22, 3.6), respectively, in the remifentanil group and 1.23 (0.17, 2.85) and 0.73 h (0.08, 2.33), respectively, in the fentanyl group.Table 1Patient characteristics (safety population)Remifentanil (n=172)Fentanyl (n=149)GenderMale147 (85%)132 (89%)Female25 (15%)17 (11%)Ethnic originCaucasian/white169 (98%)147 (99%)Oriental01 (<1%)Asian (not oriental)2 (1%)0Other1 (<1%)1 (<1%)Age (yr)Mean (range)62 (37–78)63 (38–76)sd8.88.4Height (cm)Mean (range)172 (150–191)172 (152–190)sd7.97.8Weight (kg)Mean (range)79 (50–115)80 (54–117)sd11.911.4ASA statusII38 (22%)33 (22%)III124 (72%)103 (69%)IV10 (6%)13 (9%) Open table in a new tab The incidence of inadequate anaesthesia responses to MSS (the primary efficacy endpoint) was significantly lower in the remifentanil group (11%) compared with the patients in the fentanyl group (52%; P<0.001, odds ratio remifentanil:fentanyl= 0.09, 95% CI 0.05, 0.18). The mean remifentanil infusion rate at the time of MSS was 1.42 μg kg−1 min−1 and the mean propofol infusion rate in the remifentanil group was 3.14 mg kg−1 h–1. In the fentanyl group, the cumulative fentanyl bolus dose administered at this point was 19.67 μg kg−1 and the mean propofol infusion rate was 3.94 mg kg−1 h−1 . Hypertension was the most common type of response in those patients who had an inadequate anaesthesia response, occurring in seven patients (5%) in the remifentanil group and 75 patients (50%) in the fentanyl group. There was a mean increase in SBP (from the pre-MSS value to the maximum within 5 min of MSS) of 6 mm Hg in the remifentanil group compared with a mean increase of 15 mm Hg in the fentanyl group (P<0.001). A tachycardic response was recorded in 10 patients (7%) in the remifentanil group and eight patients (5%) in the fentanyl group. Overall, the weighted mean infusion rate for remifentanil (R) during maintenance of anaesthesia (from pre-bypass to end of surgery) varied between 1.29 (0.69) μg kg−1 min−1 pre-bypass; 1.23 (0.73) μg kg−1 min−1 during bypass; 1.21 (0.77) μg kg−1 min−1 post-bypass. Fentanyl (F) cumulative bolus doses were 7.38 (3.63) μg kg−1 pre-bypass; 3.16 (1.56) μg kg–1 during bypass; and 2.92 (1.63) μg kg−1 post-bypass. Propofol varied between (R) 2.98 (0.69) vs (F) 3.58 (0.99) mg kg−1 h−1 pre-bypass; (R) 2.86 (0.45) vs (F) 3.46 (1.44) mg kg−1 h−1 bypass; (R) 2.8 (0.78) vs (F) 3.45 (1.4) mg kg–1 h−1 post-bypass; (R) 2.72 (0.85) vs (F) 3.32 (1.17) mg kg−1 h−1 end of surgery. During tracheal intubation, more patients in the remifentanil group showed signs of inadequate anaesthesia compared with those in the fentanyl group (R 24%, F 9%, P<0.001). Response to sternal skin incision (R 11%, F 36%, P<0.001), sternotomy (R 14%, F 60%, P<0.001) and MSS (R 11%, F 52%, P<0.001) were significantly and clinically less in the remifentanil group. During sternal wire placement no statistical or clinically significant difference was noted (R 10%, F 13%). Figures 1 and 2 present the SBP and HR profiles during key stages during surgery for both treatment groups. SBP fell after pre-induction (baseline) but remained stable within each treatment group compared with baseline values. SBP values were higher in the fentanyl group pre- and post-skin incision compared with remifentanil values at similar time intervals. HR remained stable and similar between treatment groups compared with pre-induction values.Fig 2Mean (95% CI) intraoperative HR in the remifentanil and fentanyl anaesthesia groups (intent-to-treat population).View Large Image Figure ViewerDownload (PPT) More patients in the fentanyl group (92%) received medication to treat inadequate anaesthesia responses compared with the patients who received remifentanil (53%). Conversely, more patients in the remifentanil group (82%) received medication to treat excessive anaesthesia responses compared with the patients who received fentanyl (60%). Table 2 summarizes the postoperative recovery times in the remifentanil and fentanyl anaesthesia groups. The median times to eligibility for weaning for extubation, and for the actual start of the extubation sequence, were longer in the remifentanil group than in the fentanyl group (P<0.05). The median times to eligibility for, and the actual start of extubation were also significantly longer in the remifentanil group than in the fentanyl group (P<0.05). The median times to eligibility for, and for actual transfer from the ICU to less intense monitoring were similar in the remifentanil and fentanyl groups. The median times to hospital discharge were also similar in the remifentanil and fentanyl groups and a similar proportion of patients in each group (17% R, 20% F) were discharged by the end of postoperative day 5 or morning of day 6.Table 2Postoperative recovery times in the remifentanil and fentanyl anaesthesia groups (intent-to-treat population) *LIM=less intense monitoringRemifentanil (n=148)Fentanyl (n=149)P-valueTime (h) to eligibility for starting extubation sequenceMedian2.41.8Range0.0–4.60.0–4.80.038(n=143)(n=149)Time (h) to actual start of extubation sequenceMedian2.41.8Range0.0–17.40.0–7.00.048(n=145)(n=149)Time (h) to eligibility for extubationMedian4.94.1Range0.2–48.60.4–20.20.016(n=145)(n=148)Time (h) to actual extubationMedian5.14.2Range0.2–48.60.4–20.20.006(n=145)(n=148)Time (days) to eligibility for transfer to LIM*Median0.80.7Range0.0–6.60.1–3.9n.s.(n=145)(n=147)Time (days) to actual transfer to LIM*Median1.10.9Range0.0–6.70.1–3.9n.s.(n=145)(n=148)Time (days) to hospital dischargeMedian6.96.1Range0.0–270.1–58n.s.(n=139)(n=140) Open table in a new tab Pain and sedation scores were assessed at scheduled times during the ICU down titration period and following discontinuation of study drugs. The weighted mean pain score during the ICU down-titration/transition period was lower in the remifentanil group (0.21) compared with the fentanyl group (0.48). However, no significant difference in the change from baseline in pain scores between the two groups (P=0.174) was noted. A significantly higher proportion of patients in the remifentanil group required additional analgesia during the period from the end of surgery until extubation compared with the fentanyl group (81% vs 69%; P=0.02). During the ICU transition period, the weighted mean sedation score was slightly but significantly higher in the remifentanil group (score 3.61) compared with the fentanyl group (score 2.92; P<0.001). Both anaesthetic regimens were well tolerated. In total, 138 patients (80%) in the remifentanil group and 113 patients (76%) in the fentanyl group experienced at least one adverse event (P=0.347). The most common adverse events recorded during the study (defined as events occurring in ≥5% of patients in either treatment group) are listed in Table 3. Overall, the incidences of these events were similar in the remifentanil and fentanyl anaesthesia groups, except for hypertension and shivering which had a significantly higher incidence in the remifentanil group compared with the fentanyl group (P<0.001 and P=0.049, respectively). Both of these events were most commonly reported during the ICU period. Significantly more drug-related adverse events were reported in the remifentanil group compared with the fentanyl group (R 44% vs F 31%, P=0.016). The most common drug-related adverse events were nausea (20% of patients in each group), vomiting (R 6%, F 8%) and shivering (R 11%, F 5%). The only significant differences in drug-related events between groups were hypertension during the immediate postoperative period (R 5%, F 0%, P=0.008) and postoperative ache(s) (R 2%, F 0%; P=0.016).Table 3Adverse events occurring in ≥5% of patients in all study phases in either treatment group (Safety Population). * Fisher's exact testAdverse eventRemifentanil (n=172)Fentanyl (n=149)P value*Nausea63 (37%)58 (39%)n.sVomiting22 (13%)20 (13%)n.sAtrial fibrillation16 (9%)15 (10%)n.sShivering21 (12%)8 (5%)0.049Fever17 (10%)9 (6%)n.s.Postoperative complication12 (7%)9 (6%)n.s.Hypertension15 (9%)1 (<1%)<0.001Ache(s)10 (6%)5 (3%)n.s.Haemorrhage10 (6%)5 (3%)n.s.Hypotension11 (6%)4 (3%)n.s. Open table in a new tab A total of 13 patients (remifentanil five, fentanyl eight) were withdrawn from the study because of adverse events. Of the remifentanil-treated patients who were withdrawn, two were withdrawn during the ICU anaesthesia phase, that is, from end of surgery until first down titration of the opioid infusion. In these patients anaphylactic reaction and bleeding from the chest wall, respectively, were given as the reason for patient withdrawal. Neither of these events was assessed as related to study drug. In a further three patients the phase of the adverse event was not specified. The events leading to withdrawal in these three cases were postoperative haemorrhage, left ventricular failure, and case of bleeding leading to haemodynamic instability, respectively. In the latter case, the event was considered possibly related to study drug. Of the fentanyl-treated patients who were withdrawn, one patient was withdrawn during the ICU anaesthesia phase, six during the ICU transition phase, that is, from first down-titration of opioid infusion until the infusion was discontinued and for one patient the phase was not specified. In four patients, the reason for withdrawal included bleeding or postoperative haemorrhage and two patients were withdrawn because of reasons which included myocardial ischaemia; a further two patients were