Hypotension after induction of anesthesia with remimazolam or etomidate: a non-inferiority randomized controlled trial in patients undergoing coronary artery bypass grafting
Article information
Abstract
Background
Remimazolam is a novel ultra-short-acting benzodiazepine known for its hemodynamic stability over propofol. However, its hemodynamic effects compared to those of etomidate are not well established. This study aimed to determine whether the use of remimazolam is non-inferior to etomidate with regard to the occurrence of post-induction hypotension in patients undergoing coronary artery bypass grafting.
Methods
Patients were randomly assigned to either the remimazolam group (6 mg/kg/h) or the etomidate group (0.3 mg/kg) for induction of anesthesia. Anesthetic depth was adjusted based on the bispectral index. Primary outcome was the incidence of post-induction hypotension, defined as a mean arterial pressure less than 65 mmHg within 15 min after endotracheal intubation, with a non-inferiority margin of 12%.
Results
A total of 144 patients were finally analyzed. Incidence of post-induction hypotension was 36/71 (50.7%) in the remimazolam group and 25/73 (34.2%) in the etomidate group, with a rate difference of 16.5% (95% CI [3.0–32.6]) between the two groups that was beyond the prespecified non-inferiority margin of 12.0%. The number of patients who needed vasopressors was similar in the two groups.
Conclusions
In this non-inferiority trial, remimazolam failed to show non-inferiority to etomidate in terms of post-induction hypotension when used as an induction drug for general anesthesia in patients undergoing coronary artery bypass grafting. However, different doses or infusion techniques of remimazolam should be compared with etomidate in various patient groups to fully assess its hemodynamic non-inferiority during induction of anesthesia.
Introduction
Post-induction hypotension, generally defined as hypotension after induction of general anesthesia and before starting the operation procedure, occurs frequently [1,2] and has been linked to unfavorable postoperative outcomes. However, unlike intraoperative hypotension, post-induction hypotension is independent of surgical factors [3] and is known to be associated with the anesthetic drug selected and the dose used for induction of anesthesia [2,4]. Therefore, selection of anesthetic agents that provide hemodynamic stability is essential to reduce the occurrence of post-induction hypotension, particularly in patients at increased cardiovascular risk.
Etomidate is the preferred choice of induction drug in patients at risk of anesthesia-induced acute cardiovascular instability because of its favorable hemodynamic profile [5,6]. In patients with severe aortic stenosis, the use of etomidate was reported to reduce the incidence of hypotension by more than half compared to propofol, along with a reduction in the severity of hypotension [7]. Also, anesthetic induction with etomidate was shown to maintain blood pressure better than sevoflurane inhalation induction in patients undergoing coronary artery bypass grafting (CABG) [8]. Although a recent meta-analysis suggested that a single dose of etomidate did not have a detrimental impact on the outcomes of patients who underwent cardiac surgery [9], concerns regarding its potential suppression of adrenal function still limits its widespread use as an induction drug [9].
As an alternative, remimazolam, a short-acting benzodiazepine that has minimal effects on the cardiovascular system, has been recently approved for general anesthesia. Similar to etomidate, hypotension developed in less than half of the patients who received a continuous infusion of remimazolam compared to a bolus injection of propofol, irrespective of the infusion dose [10]. Moreover, in a recent study of patients undergoing CABG, remimazolam reduced post-induction hemodynamic instability by 23% compared to propofol and required less vasopressor use [11]. The stable hemodynamic properties of remimazolam have led to an increasing interest in its use in cardiac surgery [12]. However, only one study has compared the hemodynamic stability of remimazolam in cardiac surgery to that of a more established induction drug, etomidate [13]. The study found that low-dose remimazolam (0.2 mg/kg) and etomidate (0.3 mg/kg) showed less hemodynamic fluctuations than high-dose remimazolam (0.3 mg/kg). Because the results varied depending on the dose of remimazolam, it is still insufficient to draw a conclusion regarding whether etomidate and remimazolam are hemodynamically equivalent. In this study, we conducted a randomized, non-inferiority clinical trial to determine whether remimazolam is non-inferior to etomidate in preventing post-induction hypotension in patients undergoing CABG.
Materials and Methods
Ethics
Ethical approval for this study (SMC 2021-11-159-003) was provided by the Institutional Review Board on January 14, 2022. The trial was registered prior to patient enrolment at the Clinical Research Information Service (https://cris.nih.go.kr/: KCT0007074). We obtained written informed consent from each patient before their participation. Study protocols were conducted in accordance with the principles of the 2013 Declaration of Helsinki.
Study subjects
From March 23, 2022, to February 27, 2023, we included 148 adult patients undergoing CABG (Fig. 1). Exclusion criteria were as follows: patients with a history of hypersensitivity to the study medications, severe hepatic dysfunction, low left ventricular ejection fraction (≤ 40%), presence of valvular heart disease, glaucoma, and diagnosed adrenocortical insufficiency.
Consolidated Standards of Reporting Trials (CONSORT) diagram.
Intervention
We randomly assigned all eligible patients to either the etomidate or remimazolam group using computer-generated block randomization (www.randomizer.com) with an allocation ratio of 1:1 in blocks of four. An independent investigator and an attending anesthesiology resident, who were not blinded to the group allocation, prepared the induction drugs according to group allocation and started induction of anesthesia. Because the study design stipulated the initiation of data recording immediately after tracheal intubation with continuation of recording for the next 15 min, an investigator, who was blinded to group allocation, entered the operating room immediately after the tracheal intubation. At this point, the study medication had already been discontinued, and anesthesia was maintained with sevoflurane. The blinded investigator, who entered the operating room after the intubation, recorded all parameters and decided whether to give rescue medications.
Standard monitoring procedures, including five-lead electrocardiography, non-invasive blood pressure, pulse oximetry, and bispectral index (BIS) assessment, began as soon as the patient arrived in the operating room. Invasive arterial catheterization was performed through the radial artery before induction of anesthesia. If there was a greater than 30% difference between the baseline mean arterial pressure (MAP) in the operating room and the MAP measured on the day before the surgery day, the patient was allowed to relax for 10 min. In accordance with our institutional preoperative preparation standards, angiotensin receptor blockers and angiotensin-converting enzyme inhibitors that are associated with intraoperative hypotension [14] were withheld on the morning of surgery for all patients.
We induced general anesthesia in the etomidate group using an intravenous bolus injection of etomidate (0.3 mg/kg, etomidate; B. Braun Korea), and continuously infused remimazolam (6 mg/kg/h, Byfavo®; Hana Pharmaceutical) in the remimazolam group until the BIS reached 60. The infusion rate of remimazolam was determined according to the manufacturer’s recommendations. Except for the choice of anesthesia induction drug, patient management was consistent between the etomidate and remimazolam groups. After loss of consciousness, 0.5 µg/kg of sufentanil (sufentanil citrate; BC World Pharmaceutical) and 0.6-0.8 mg/kg of rocuronium were administered intravenously to provide appropriate analgesia and neuromuscular blockade, respectively. Sevoflurane was initiated when the BIS reached 60 in both groups and titrated to maintain a BIS between 40 and 60. After confirming that the train-of-four decreased from 1.0 to 0.0, endotracheal intubation was performed with a video laryngoscope (Pentax-AWS®; Pentax). No additional opioids were used during the study period. Nil per os (NPO) was applied from midnight, and no fluid was infused before induction of anesthesia in the operating room. We administered a balanced crystalloid at 10 ml/kg/h during the study period.
Outcomes
Change in MAP following induction of anesthesia was used as the principal indicator in this study. Primary outcome was the incidence of post-induction hypotension, defined as a MAP less than 65 mmHg during the first 15 min immediately following endotracheal intubation [15]. Secondary outcomes included the incidences of alternate definitions of post-induction hypotension as follows: MAP less than 60 mmHg and MAP reduction of more than 20% and 30% as compared to the baseline MAP measured the night before surgery [16,17]. During the 15-minute study period, patients were maintained in a supine position, and only sterile preparation or femoral artery cannulation was performed.
Incidence of post-induction hypertension, time to loss of consciousness, time to reach BIS 60, and the incidence of injection pain and myoclonus were also compared between the two groups. Post-induction hypertension was defined as a systolic blood pressure greater than 140 mmHg [18]. Time to loss of consciousness was defined as the time taken for the patient to become unresponsive to light tapping on the shoulder or verbal commands after administrating an anesthesia induction drug [19]. Time to reach BIS 60 was defined as the duration of the interval between the administration of the anesthesia induction drug and the reduction of the BIS to 60. Myoclonus was defined as an involuntary, short contraction of some muscle fibers occurring for a period not longer than 100 milliseconds [20]. Injection pain was measured using a graded scale (Grade 0: no pain; Grade 1: verbal complaint of pain; Grade 2: withdrawal of the arm; and Grade 3: verbal complaint and withdrawal of the arm) [21]. A vasoactive drug (vasopressor) was administered if the patient’s MAP dropped below 60 mmHg for more than 5 min or if the MAP dropped below 50 mmHg [22]. In these cases, ephedrine 5 mg or phenylephrine 30-50 μg were administered depending on whether the heart rate was below 60 beats/min or not, respectively. Vasoactive drug was administered every minute until the blood pressure goal was achieved. All hemodynamic parameters were automatically recorded using real-time data acquisition software (Vital Recorder, Vital DB) during the study period [23].
Statistical analysis
We used the non-inferiority approach to analyze the primary outcome [24]. Sample size was calculated based on the findings of a previous study in which the incidence of post-induction hypotension after etomidate administration was approximately 7% [4]. To test equivalence with CIs, we considered any difference in the primary outcome < 12% to be clinically acceptable based on FDA guidelines that indicate differences of 10% to 12% to be clinically relevant [25]. At the time we planned the present study, no previous study had reported the incidence of post-induction hypotension with remimazolam, and we therefore hypothesized that the incidence of post-induction hypotension would be equivalent following remimazolam or etomidate induction. This non-inferiority hypothesis for the primary outcome was tested using a one-sided t-test at a significance level of 2.5%. With an alpha error of 0.025 and a power of 80%, the required sample size was 71 per group. We therefore recruited 74 patients per group.
SPSS® 25.0 (IBM® Corp.) software package was used for all analyses except for evaluation of the primary outcome that was performed using GraphPad Prism 9.4.1 (GraphPad Software). Continuous variables are expressed as mean ± SD or as median (Q1, Q3) and were analyzed using Student’s t-test or the Mann-Whitney U test as appropriate depending on the normality of the data as assessed by the Kolmogorov-Smirnov test. The chi-square test and Fisher’s exact test were used to compare categorical variables that are presented as numbers (%). A P value of < 0.05 was considered statistically significant. Two-sided tests were used, except for the one-sided t-test of non-inferiority.
Results
Of the 150 patients screened for eligibility, 148 met the inclusion criteria and were randomly allocated into the etomidate (n = 74) or remimazolam (n = 74) groups. All patients received the allocated interventions without study violations, but four patients were dropped from the analysis (Fig. 1): one patient in the etomidate group and two patients in the remimazolam group were excluded after the intervention because of a malfunction in the vital sign recording system, and one patient allocated to the remimazolam group was excluded because persistent arrhythmia occurred after endotracheal intubation. Baseline clinical characteristics of the patients in the two groups were similar (Table 1). Cumulative dose of remimazolam infused until the BIS reached 60 was approximately 0.3 mg/kg.
Comparison of Patient Characteristics
Post-induction hypotension (MAP < 65 mmHg) occurred in 36/71 (50.7%) patients in the remimazolam group and 25/73 (34.2%) patients in the etomidate group, with a rate difference in the incidence of post-induction hypotension between the two groups of 16.5% (95% CI [0.39–31.5], Table 2). Thus, non-inferiority was not established because the upper limit of the 95% CI was higher than the prespecified non-inferiority margin of 12.0% (Fig. 2). We also failed to establish non-inferiority when post-induction hypotension was defined as a MAP less than 60 mmHg and MAP reduction of more than 20% compared to the baseline MAP. In contrast, when hypotension was defined as a MAP reduction of more than 30% compared to the baseline MAP, the estimated hypotension incidence rate difference was 32.8% (90% CI [17.2–48.4], Fig. 2), indicating the inferiority of remimazolam versus etomidate.
Comparison of Hemodynamic Changes between Groups during the Study Period
Non-inferiority diagram illustrating the absolute risk reduction in the incidence of hypotension between the etomidate and remimazolam groups over the first 15 minutes after anaesthetic induction. A vertical dotted line indicates a non-inferiority margin of 12%. Reversed triangle, triangle, square, and rhombus indicates hypotension incidence differences, and the error bars indicate the 95% CI of the incidence difference between groups. MAP: mean arterial pressure.
The number of patients who required vasoactive drugs and the percent MAP decrement was not different between the two groups (Table 2). In the etomidate group, the incidence of post-induction hypertension was greater than in the remimazolam group during the study period (58 [81.7%] vs. 36 [51.4%] patients, P < 0.001, Table 2). The highest and lowest MAPs during the 15-minute study period are shown in Fig. 3.
Comparison of the highest and the lowest MAP during the 15-minute study period between the etomidate group (gray, circle) and the remimazolam group (blue, square). Initial = MAP before anaesthetic induction; Highest = highest MAP value during the 15-minute study period; Lowest = lowest MAP value during the 15-minute study period. *P value was statistically significant. MAP: mean arterial pressure.
Regarding other secondary outcomes, both the time to loss of consciousness and the time to reach BIS 60 were significantly longer in the remimazolam group than in the etomidate group (48 [36, 59] vs. 97 [85, 115] s, P < 0.001; and 86 [77, 105] vs. 178 [157, 208] s, P < 0.001, respectively, Table 1). In terms of adverse events related to the study medications, only patients in the etomidate group (22 patients [30.1%]) experienced myoclonus (P < 0.001). No patient complained of injection pain in either group (Table 1).
Discussion
In this clinical trial, remimazolam infusion at 6 mg/kg/h failed to show non-inferiority to bolus etomidate of 0.3 mg/kg in terms of post-induction hypotension when used as a general anesthesia induction drug in patients undergoing CABG. Post-induction hypotension incidence rate difference was 16.5% (95% CI [3.0–32.6]) that exceeded the predefined equivalence threshold of 12.0%. However, there was no difference in the requirement for vasoactive drugs between the two groups. Although remimazolam may be less effective at preventing post-induction hypotension than etomidate in patients undergoing CABG, most cases of post-induction hypotension in the remimazolam group were not severe and resolved without the need for vasoactive drugs.
Previous studies have shown that both remimazolam and etomidate reduce the incidence of hypotension by nearly half when compared to propofol [7,10]. While numerous studies have compared the hemodynamic effects of propofol or remimazolam induction, as well as propofol or etomidate induction, few studies have directly compared the hemodynamic profiles of remimazolam and etomidate during anesthesia induction [13,26,27]. Moreover, although remimazolam has gained attention as an induction drug for cardiac surgery [12], only one study conducted a clinical trial to compare the hemodynamic stabilities of etomidate and remimazolam in cardiac surgery. We addressed this gap in knowledge by comparing these two agents in patients with compromised cardiovascular reserve.
The only comparative study in patients undergoing cardiac surgery compared two bolus doses of remimazolam with etomidate for induction of anesthesia. In this study by Hu et al. [13], low-dose remimazolam (0.2 mg/kg) yielded a post-induction hypotension incidence similar to that of etomidate, while high-dose remimazolam (0.3 mg/kg) resulted in a hypotension incidence double that of etomidate. These findings indicate that the remimazolam infusion dose is a critical factor affecting hemodynamic stability and it is important to choose a proper infusion dose. However, using a bolus remimazolam injection for the induction of general anesthesia is still an off-label use of this drug, which limits the applicability of available information to everyday clinical practice. Therefore, before recommending remimazolam as an alternative to etomidate, a clinical trial comparing etomidate with the on-label use of remimazolam should be conducted first. This study is the first to investigate whether the on-label use of remimazolam is at least as effective as etomidate in preventing post-induction hypotension or whether the higher incidence of post-induction hypotension associated with the use of remimazolam is acceptable within the specified inferiority margins.
In our non-inferiority clinical trial, the incidence of post-induction hypotension following remimazolam induction was up to twice as high as that reported in recent studies using continuous infusion of remimazolam at an equivalent dose of 6 mg/kg/h [10,28–30]. This discrepancy may be explained by the average cumulative dosage of remimazolam administered during induction. In our study, the cumulative dose of remimazolam infused throughout the induction period was approximately 0.3 mg/kg, whereas in the previously mentioned recent studies, patients received a cumulative dose of remimazolam ranging from 0.1 mg/kg to 0.26 mg/kg [10,28,30]. With regard to the average age of patients in the remimazolam group, 0.3 mg/kg is higher than the recommended age-adjusted bolus injection dose of remimazolam that is 0.19 to 0.25 mg/kg in 60- to 80-year-old patients [31]. This may have led to a higher incidence of post-induction hypotension in the remimazolam group irrespective of the way hypotension was defined and the failure of this study to demonstrate the non-inferiority of remimazolam in preventing post-induction hypotension compared with etomidate.
However, the requirement of vasoactive drugs and the absolute MAP decrease during the study period did not differ significantly between the two groups. These findings suggest that some post-induction hypotension cases in the remimazolam group were mild and resolved without having to use rescue vasoactive drugs (Table 2). Therefore, if the continuous infusion dose of remimazolam was adjusted according to the patient’s age, or if the time point of remimazolam discontinuation had been different, our findings may have been quite different.
Regarding secondary outcomes, we found that the incidences of myoclonus and post-induction hypertension were greater in the etomidate group than the remimazolam group. Myoclonus usually increases patient discomfort during induction [32], while intraoperative hypertension is associated with increased 30-day mortality after cardiac surgery [33]. However, concomitant use of opioids could reduce the incidence of myoclonus and post-induction hypertension following etomidate use [34]. Despite the concurrent use of sufentanil in the current study, a larger dose of analgesic may be needed after etomidate induction to avoid myoclonus or abrupt blood pressure changes.
Our study had several limitations. First, continuous remimazolam infusion was compared to a bolus etomidate injection. Although different forms of infusion were compared, a previous study demonstrated no significant differences in hemodynamic parameters when anesthesia was induced by continuous infusion or single-bolus injection of remimazolam [35]. Thus, we reasoned that comparing the two drugs using the approved infusion method for each would provide more meaningful information for real-world clinical practice than comparing these drugs using the same infusion method that was not approved for one or the other. Second, considering the lag time between the effective site concentration and the BIS values dropping below 60, the remimazolam group may have been overdosed. Because there is no clearly defined method to monitor the depth of anesthesia during remimazolam administration, discontinuation of remimazolam infusion was standardized to an objective target value of BIS 60. However, further studies are needed to develop a pharmacokinetic-pharmacodynamic model for remimazolam and to devise an appropriate monitoring system for the depth of anesthesia during remimazolam-induced anesthesia. Third, hypotension was recorded after endotracheal intubation due to blinding of the study authors, and not immediately after administering the anesthesia induction drugs. Therefore, we were unable to evaluate any hemodynamic changes that occurred between the initiation of anesthesia induction and the completion of endotracheal intubation. Fourth, we were unable to control all cardiovascular medications taken by the patients prior to entering the operating room. However, the initial blood pressure did not differ between the remimazolam and etomidate groups. Finally, patients with a reduced left ventricular ejection fraction (< 40%) were not included in our study to ensure patient safety. Given that the criteria for non-inferiority were not met even in patients without impaired cardiac function, a comparison between etomidate and remimazolam within a patient group with low ejection fractions will likely result in greater blood pressure differences during anesthesia induction.
In conclusion, remimazolam use was associated with a greater incidence of post-induction hypotension than etomidate in patients undergoing CABG, and we failed to prove the non-inferiority of remimazolam to etomidate. However, the number of post-induction hypotension cases that required vasoactive drugs did not differ between the two groups. Therefore, various doses and infusion techniques of remimazolam should be compared with etomidate across different patient groups to fully assess the hemodynamic changes induced by etomidate versus remimazolam during anesthesia induction.
Notes
Correction
This article was corrected on April 15, 2024, for an author affiliation.
Funding
This study was supported by Hana Pharm Co., Ltd., Seoul, Republic of Korea (grant number, PHO021529). None of the authors have a personal financial interest in this study and the sponsor was not involved in the data analyses or interpretation.
Conflicts of Interest
No potential conflict of interest relevant to this article was reported.
Data Availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Author Contributions
Jeong-Jin Min (Conceptualization; Formal analysis; Methodology; Supervision; Writing – original draft; Writing – review & editing)
Eun Jung Oh (Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Visualization; Writing – original draft; Writing – review & editing)
Hyunsung Cho (Data curation; Supervision)
Chungsu Kim (Data curation; Supervision)
Jong-Hwan Lee (Methodology; Supervision; Validation; Writing – review & editing)