Alterations in Functional Connectivity and Network Topology During Remimazolam-Induced Unresponsiveness

Remimazolam, an ultrashort-acting intravenous benzodiazepine, is a safe and effective sedative agent. Previous studies have established a strong correlation between cortical network alterations and general anesthesia. However, the effects of remimazolam on the cortical network remain unclear.

Twenty-two patients were administered intravenous remimazolam. Recordings were obtained using a 32-channel electroencephalography across the baseline, anesthesia, and recovery states. Brain oscillatory activity during remimazolam anesthesia was assessed through spectral power analysis. Functional connectivity was assessed using the weighted and directed phase lag indices, with the former used to construct weighted brain networks. Network characteristics were analyzed using nodal metrics (nodal clustering coefficient and efficiency) and global metrics (average clustering coefficient, average path length, modularity, and small-worldness). In addition, hub nodes were identified using the largest betweenness centrality to investigate the network's hub structure across different states.

Remimazolam induced significant anteriorization of alpha power and markedly decreased alpha functional connectivity in both prefrontal-frontal and anterior-posterior regions ( P < .019). Remimazolam significantly affected the alpha-band functional brain network, characterized by reduced nodal clustering ( P < .001) and efficiency ( P < .001), but increased global clustering ( P < .001), average path length ( P = .022), and modularity ( P < .001). The small-world property-a network structure balancing high clustering with short path lengths-was preserved under remimazolam anesthesia (slightly increased, P = .028). After remimazolam anesthesia, the hub structure of the brain was reconfigured and characterized by hub node redistribution from the posterior to anterior regions.

Remimazolam induced reorganization of functional brain networks from highly connected, highly integrated complex networks to sparsely connected, locally modular cortical networks. These findings strengthen the notion that consciousness relies on networks capable of efficient information transmission that critically depends on the balance between global functional integration and segregation.