Anesthesiology. 2021 Jul 1;135(1):69-82. doi: 10.1097/ALN.0000000000003774.
Despite the well-known clinical effects of midazolam and ketamine, including sedation and memory impairment, the neural mechanisms of these distinct drugs in humans are incompletely understood. The authors hypothesized that both drugs would decrease recollection memory, task-related brain activity, and long-range connectivity between components of the brain systems for memory encoding, pain processing, and fear learning.
In this randomized within-subject crossover study of 26 healthy adults, the authors used behavioral measures and functional magnetic resonance imaging to study these two anesthetics, at sedative doses, in an experimental memory paradigm using periodic pain. The primary outcome, recollection memory performance, was quantified with d' (a difference of z scores between successful recognition versus false identifications). Secondary outcomes were familiarity memory performance, serial task response times, task-related brain responses, and underlying brain connectivity from 17 preselected anatomical seed regions. All measures were determined under saline and steady-state concentrations of the drugs.
Recollection memory was reduced under midazolam (median [95% CI], d' = 0.73 [0.43 to 1.02]) compared with saline (d' = 1.78 [1.61 to 1.96]) and ketamine (d' = 1.55 [1.12 to 1.97]; P < 0.0001). Task-related brain activity was detected under saline in areas involved in memory, pain, and fear, particularly the hippocampus, insula, and amygdala. Compared with saline, midazolam increased functional connectivity to 20 brain areas and decreased to 8, from seed regions in the precuneus, posterior cingulate, and left insula. Compared with saline, ketamine decreased connectivity to 17 brain areas and increased to 2, from 8 seed regions including the hippocampus, parahippocampus, amygdala, and anterior and primary somatosensory cortex.
Painful stimulation during light sedation with midazolam, but not ketamine, can be accompanied by increased coherence in brain connectivity, even though details are less likely to be recollected as explicit memories.
尽管咪达唑仑和氯胺酮具有众所周知的临床作用，包括镇静和记忆损害，但这些药物在人类中的神经机制仍不完全清楚。作者假设这两种药物都会降低回忆记忆、与任务相关的大脑活动以及记忆编码、疼痛处理和恐惧学习的大脑系统组件之间的长程连接。
在这项针对 26 名健康成年人的随机、自身对照、交叉研究中，作者使用行为测量和功能磁共振成像，在周期性疼痛的实验记忆范式中研究了这两种麻醉剂在镇静剂量下的作用。主要结局指标是回忆记忆表现，用 d'（成功识别与错误识别之间的 z 分数差异）来量化。次要结局指标包括熟悉记忆表现、连续任务反应时间、与任务相关的大脑反应以及来自 17 个预选解剖种子区域的基础大脑连接。所有测量均在盐水和药物稳定状态浓度下进行。
与盐水相比（d'=0.73[0.43 至 1.02]），咪达唑仑降低了回忆记忆（中位数[95%CI]）（d'=1.78[1.61 至 1.96]）和氯胺酮（d'=1.55[1.12 至 1.97]；P<0.0001）。在盐水条件下，大脑中与记忆、疼痛和恐惧相关的区域，特别是海马体、岛叶和杏仁核，都能检测到与任务相关的大脑活动。与盐水相比，咪达唑仑增加了与 20 个大脑区域的功能连接，减少了与 8 个大脑区域的连接，这些大脑区域的种子区域位于楔前叶、后扣带和左侧岛叶。与盐水相比，氯胺酮减少了与 17 个大脑区域的连接，增加了与 8 个大脑区域的连接，这些大脑区域的种子区域包括海马体、海马旁回、杏仁核以及前和初级体感皮层。
在轻度镇静下接受咪达唑仑的疼痛刺激，即使不太可能作为明确的记忆被回忆起来，也可能伴随着大脑连接的协同性增加。
