Background

Previous studies indicate that isoflurane could be useful for the sedation of patients in the intensive care unit (ICU), but prospective studies evaluating isoflurane's efficacy have been small. The aim of this study was to test whether the sedation with isoflurane was non-inferior to sedation with propofol.

Methods

This phase 3, randomised, controlled, open-label non-inferiority trial evaluated the efficacy and safety of up to 54 h of isoflurane compared with propofol in adults (aged ≥18 years) who were invasively ventilated in ICUs in Germany (21 sites) and Slovenia (three sites). Patients were randomly assigned (1:1) to isoflurane inhalation via the Sedaconda anaesthetic conserving device (ACD; Sedana Medical AB, Danderyd, Sweden; ACD-L [dead space 100 mL] or ACD-S [dead space 50 mL]) or intravenous propofol infusion (20 mg/mL) for 48 h (range 42–54) using permuted block randomisation with a centralised electronic randomisation system. The primary endpoint was percentage of time in Richmond Agitation–Sedation Scale (RASS) range –1 to –4, assessed in eligible participants with at least 12 h sedation (the per-protocol population), five or more RASS measurements, and no major protocol violations, with a non-inferiority margin of 15%. Key secondary endpoints were opioid requirements, spontaneous breathing, time to wake-up and extubation, and adverse events. Safety was assessed in all patients who received at least one dose. The trial is complete and registered with EudraCT, 2016–004551–67.

Findings

Between July 2, 2017, and Jan 12, 2020, 338 patients were enrolled and 301 (89%) were randomly assigned to isoflurane (n=150) or propofol (n=151). 146 patients (97%) in each group completed the 24-h follow-up. 146 (97%) patients in the isoflurane group and 148 (98%) of patients in the propofol group were included in the per-protocol analysis of the primary endpoint. Least-squares mean percentage of time in RASS target range was 90·7% (95% CI 86·8–94·6) for isoflurane and 91·1% (87·2–95·1) for propofol. With isoflurane sedation, opioid dose intensity was 29% lower than with propofol for the overall sedation period (0·22 [0·12–0·34] vs 0·32 [0·21–0·42] mg/kg per h morphine equivalent dose, p=0·0036) and spontaneous breathing was more frequent on day 1 (odds ratio [OR] 1·72 [1·12–2·64], generalised mixed linear model p=0·013, with estimated rates of 50% of observations with isoflurane vs 37% with propofol). Extubation times were short and median wake-up was significantly faster after isoflurane on day 2 (20 min [IQR 10–30] vs 30 min [11–120]; Cox regression p=0·0011). The most common adverse events by treatment group (isoflurane vs propofol) were: hypertension (ten [7%] of 150 vs two [1%] of 151), delirium (eight [5%] vs seven [5%]), oliguria (seven [5%] vs six [4%]), and atrial fibrillation (five [3%] vs four [3%]).

Interpretation

These results support the use of isoflurane in invasively ventilated patients who have a clinical need for sedation.

Funding

Sedana Medical AB.

中文翻译：

---

在德国和斯洛文尼亚，通过麻醉保存装置吸入异氟醚与丙泊酚对重症监护病房有创通气患者的镇静：一项开放标签、3 期、随机对照、非劣效性试验

背景

先前的研究表明，异氟醚可用于重症监护病房 (ICU) 患者的镇静，但评估异氟醚疗效的前瞻性研究很少。本研究的目的是检验异氟醚的镇静作用是否不劣于异丙酚的镇静作用。

方法

这项 3 期随机、对照、开放标签非劣效性试验评估了异氟醚与丙泊酚在德国 ICU （21 个地点）进行有创通气的成人（≥18 岁）中长达 54 小时的疗效和安全性和斯洛文尼亚（三个地点）。患者通过 Sedaconda 麻醉保存装置（ACD；Sedana Medical AB，Danderyd，瑞典；ACD-L [死腔 100 mL] 或 ACD-S [死腔 50 mL]）随机分配（1:1）吸入异氟醚组或静脉输注丙泊酚（20 mg/mL）48 小时（范围 42-54），使用带集中电子随机化系统的置换区组随机化。主要终点是里士满激动-镇静量表 (RASS) 范围 –1 到 –4 的时间百分比，在至少 12 小时镇静的合格参与者（符合方案人群）中进行评估，5 次或更多 RASS 测量，且无重大协议违规，非劣效性边际为 15%。关键的次要终点是阿片类药物的需求、自主呼吸、醒来和拔管的时间以及不良事件。在接受至少一剂的所有患者中评估了安全性。该试验已完成并在 EudraCT 注册，2016-004551-67。

发现

2017 年 7 月 2 日至 2020 年 1 月 12 日期间，338 名患者入组，301 名 (89%) 患者被随机分配至异氟醚 (n=150) 或丙泊酚 (n=151)。每组146名患者（97%）完成24小时随访。异氟醚组的 146 名 (97%) 患者和丙泊酚组的 148 名 (98%) 患者被纳入主要终点的符合方案分析。异氟醚在 RASS 目标范围内的最小二乘平均时间百分比为 90·7% (95% CI 86·8–94·6)，丙泊酚为 91·1% (87·2–95·1)。使用异氟醚镇静时，在整个镇静期间，阿片类药物的剂量强度比丙泊酚低 29%（0·22 [0·12–0·34] vs0·32 [0·21–0·42] mg/kg/h 吗啡当量剂量，p=0·0036)，第 1 天自主呼吸更加频繁（优势比 [OR] 1·72 [1·12– 2·64]，广义混合线性模型 p=0·013，估计异氟醚观察率为 50% ，丙泊酚观察率为37%）。异氟醚后第 2 天拔管时间短且中位唤醒明显更快（20 分钟 [IQR 10-30]对比30 分钟 [11-120]；Cox 回归 p=0·0011）。治疗组最常见的不良事件（异氟醚vs丙泊酚）是：高血压（150 人中有 10 人 [7%]对151 人中有 2 人 [1%]）、谵妄（8 人 [5%]对7 人 [5%]）、少尿（七 [5%]对六 [4%]）和心房颤动（五 [3%]与四个 [3%]）。

解释

这些结果支持在临床需要镇静的有创通气患者中使用异氟醚。

资金

塞达纳医疗 AB。