Cerebrospinal fluid metabolic profiling reveals divergent modulation of pentose phosphate pathway by midazolam, propofol and dexmedetomidine in patients with subarachnoid hemorrhage: a cohort study

BACKGROUND: Agitation is common in subarachnoid hemorrhage (SAH), and sedation with midazolam, propofol and dexmedetomidine is essential in agitation management. Previous research shows the tendency of dexmedetomidine and propofol in improving long-term outcome of SAH patients, whereas midazolam might be detrimental. Brain metabolism derangement after SAH might be interfered by sedatives. However, how sedatives work and whether the drugs interfere with patient outcome by altering cerebral metabolism is unclear, and the comprehensive view of how sedatives regulate brain metabolism remains to be elucidated. METHODS: For cerebrospinal fluid (CSF) and extracellular space of the brain exchange instantly, we performed a cohort study, applying CSF of SAH patients utilizing different sedatives or no sedation to metabolomics. Baseline CSF metabolome was corrected by selecting patients of the same SAH and agitation severity. CSF components were analyzed to identify the most affected metabolic pathways and sensitive biomarkers of each sedative. Markers might represent the outcome of the patients were also investigated. RESULTS: Pentose phosphate pathway was the most significantly interfered (upregulated) pathway in midazolam (p = 0.0000107, impact = 0.35348) and propofol (p = 0.00000000000746, impact = 0.41604) groups. On the contrary, dexmedetomidine decreased levels of sedoheptulose 7-phosphate (p = 0.002) and NADP (p = 0.024), and NADP is the key metabolite and regulator in pentose phosphate pathway. Midazolam additionally augmented purine synthesis (p = 0.00175, impact = 0.13481) and propofol enhanced pyrimidine synthesis (p = 0.000203, impact = 0.20046), whereas dexmedetomidine weakened pyrimidine synthesis (p = 0.000000000594, impact = 0.24922). Reduced guanosine diphosphate (AUC of ROC 0.857, 95%CI 0.617–1, p = 0.00506) was the significant CSF biomarker for midazolam, and uridine diphosphate glucose (AUC of ROC 0.877, 95%CI 0.631–1, p = 0.00980) for propofol, and succinyl-CoA (AUC of ROC 0.923, 95%CI 0.785–1, p = 0.000810) plus adenosine triphosphate (AUC of ROC 0.908, 95%CI 0.6921, p = 0.00315) for dexmedetomidine. Down-regulated CSF succinyl-CoA was also associated with favorable outcome (AUC of ROC 0.708, 95% CI: 0.524–0.865, p = 0.029333). CONCLUSION: Pentose phosphate pathway was a crucial target for sedatives which alter brain metabolism. Midazolam and propofol enhanced the pentose phosphate pathway and nucleotide synthesis in poor-grade SAH patients, as presented in the CSF. The situation of dexmedetomidine was the opposite. The divergent modulation of cerebral metabolism might further explain sedative pharmacology and how sedatives affect the outcome of SAH patients.