Proteomic profiling in cerebrospinal fluid reveal biomarkers for shunt outcome in idiopathic normal-pressure hydrocephalus

作者：Yuqi Ying, Jingquan Lin, Weijun Gao, Liang Yue, Qingze Zeng, Katrina Bartas, D. M.O. Cheong, Hongjie Jiang, Zhe Zheng, Ligen Shi, Ping An, Yuanjian Fang, Feng Yan, Tiannan Guo, Jianmin Zhang, Hemmings Wu, Kevin Beier, Junming Zhu, Zhoule Zhu · 发表于：Journal of Advanced Research · 年份：2025 · DOI：10.1016/j.jare.2025.04.043 · 被引用次数：3 · 研究领域：Cerebrospinal fluid and hydrocephalus、Traumatic Brain Injury and Neurovascular Disturbances、Intracerebral and Subarachnoid Hemorrhage Research

BACKGROUND: The pathophysiology of idiopathic normal pressure hydrocephalus (iNPH) remains unclear, and the treatment strategy remains suboptimal. This study aims to identify biomarkers for shunt prognosis by cerebrospinal fluid (CSF) proteomic profiling. METHODS: CSF samples collected from 37 iNPH patients from the discovery cohort and 12 iNPH patients from an independent validation cohort (71.9 ± 6.1 years (mean ± SD)), and 16 age-balanced controls (69.9 ± 7.6 years (mean ± SD)) were collected from September 2020 to December 2023. 53 CSF samples were analyzed using a mass spectrometry-based proteomic workflow. Clinical evaluations were performed on all iNPH patients, and 44 patients underwent ventriculoperitoneal shunting. Postoperative CSF were also collected from 10 iNPH patients who underwent shunting surgery. Bioinformatics, machine learning, and enzyme-linked immunosorbent assay (ELISA) were performed to identify CSF proteome changes related to pathophysiology in iNPH, and screen for biomarkers associated with shunt response. RESULTS: 39 and 285 proteins significantly increased and decreased in iNPH CSF compared to the control group. Gene ontology analysis revealed that the noticeably increased proteins were mainly associated with myeloid leukocyte migration and extracellular matrix organization, and significantly decreased proteins were primarily associated with axon development and synapse organization. Machine learning identified 6 candidate biomarkers that potentia...