Resting-State Functional Connectivity in Neurosurgical Patients Under Propofol Anesthesia: Detectability and Variability Between Patients and Between Sessions

Abstract

Intraoperative assessment of functional connectivity (FC) provides a new possibility for mapping the eloquent brain region before, during, and after tumor resection. The aim of this study was to perform a systematic analysis of detectability of FC and its variation between subjects and sessions. FC was analyzed in 10 patients with pituitary adenoma under propofol anesthesia before and after tumor resection. FC of each session (20 sessions total) was correlated to a reference matrix of a group of healthy subjects to evaluate variations of overall, interhemispheric, and intrahemispheric FC between sessions. Resting-state patterns could be detected during anesthesia (F1,9= 112.14, P < 0.001). There was a significant effect of session (F1,9= 19.401, P= 0.002), which included a reduction in resting state from the first to second session. There was no effect of connection type (F2,8= 1.498, P= 0.280), and there was no interaction between connection type and session (F2,8= 0.187, P=0.833). The correlation between the observed reduction in resting-state activity between the sessions and the time span between sessions was not significant (r= 0.25, P= 0.29). FC of the first session showed a significant correlation to the initial dose of anesthesia (r= 0.7, P=0.007). However, there was no significant correlation between the total dose of propofol and FC of the second session (r= 1.7, P= 0.6). Significant FC could be detected under anesthesia but showed a significant decrease in the second session. To implement FC intraoperative brain mapping, further studies are required to optimize the depth sedation to obtain stable FC between sessions.