Background: Whether general anesthetics induce neurotoxicity is unclear. Objectives: We hypothesized that serum neurofilament light (NfL) release, a marker of neural injury, would not be different in patients receiving general anesthesia (GA) compared to hypnosis. Design: Prospective, non-randomized. Setting: Tertiary university hospital. Methods: Patients undergoing breast cancer or thyroid/parathyroid surgery were enrolled. Propofol and remifentanil were used for GA. For hypnosis, only low dose remifentanil was allowed. Main outcome measures: NfL was measured at baseline, at day 1 and at postoperative visit. At preoperative and postoperative visit, patients performed a Montreal Cognitive Assessment (MoCA) test. Primary outcome was absolute change in NfL at day 1 between groups. Results: Among 100 included patients, 51 were in GA and 49 in hypnosis group. There were 26/49 (53%) breast cancer patients in hypnosis group and 13/51 (26%) in GA arm, p = 0.005. Baseline NfL was higher in the hypnosis group (p = 0.050). We subtracted log-transformed baseline NfL from postoperative NfL to analyze normalized values as absolute change. The mean normalized value at day 1 (primary endpoint) was 0.02 (± 0.11) in the GA and 0.03 (± 0.20) in the hypnosis group; p = 0.979 [d = 0.16, 95% CI: -0.40 to 0.39]. In a linear mixed-effects model including variables that influenced baseline NfL (advanced age, higher creatinine values, breast cancer), mean difference (95 % CI) in absolute change of log NfL at day 1 between both groups was 0.007 (- 0.04 to 0.06); p = 0.791. A change in MoCA as compared to baseline was not different between groups (p = 0.761). Conclusions: Our results show that there was no statistical or clinical difference in postoperative NfL release between GA and hypnosis. GA with propofol was not associated with a different risk of neural injury compared to hypnosis. Trial registration: NCT04500236.
Read full abstract