Abstract
BACKGROUND: The neurochemical alterations in cerebrospinal fluid (CSF) associated with the typical symptomatology in idiopathic normal pressure hydrocephalus (iNPH) and their association with outcome after shunt surgery are unsettled. AIM: To explore associations between concentrations of CSF biomarkers reflecting amyloid- and tau pathology, neuronal degeneration as well as astrocytic activation and the characteristic symptomatology in iNPH and to examine whether these biomarkers can predict the postoperative outcome in all patients and in patients without evidence of Alzheimer's disease (AD) pathology. METHODS: This explorative study included 81 patients diagnosed with iNPH at the Hydrocephalus research unit, Sahlgrenska. Symptoms were assessed using the iNPH-scale and standardized clinical tests measuring gait, balance, cognition and urinary incontinence before and median 8 months after shunt surgery. Pre-operative lumbar CSF concentrations of Aβ38, Aβ40, Aβ42, ratio Aβ42/Aβ40, sAPPα, sAPPβ, T-tau, P-tau, MCP-1, and NFL were analyzed. A low Aβ42/Aβ40 ratio defined patients with AD pathology. Correlation and regression analyses between biomarker concentrations and clinical symptoms at baseline as well as postoperative change in symptoms after surgery, were performed. RESULTS: Higher NFL correlated with more pronounced impairment in all clinical tests, i.e. included measures of gait, balance, cognition and urinary incontinence (r(p)=0.25-0.46, p < 0.05). Higher T-tau and P-tau correlated with poorer performance in cognitive tests (r(p)=0.26-0.39, p < 0.05). No biomarker could differentiate between improved and unimproved patients in the whole sample or in AD-pathology negative patients. Low ratio Aβ42/Aβ40 lacked predictive value. A higher preoperative P-tau was weakly correlated with less pronounced overall clinical improvement (r(p) = -0.238, p = 0.036). CONCLUSIONS: Axonal degeneration, as indicated by elevated NFL, is probably involved in the generation of the full iNPH tetrade of symptoms and tau pathology more specifically with iNPH cognitive impairment. No CSF biomarker could identify shunt responders. CSF evidence of Alzheimer pathology should not be used to exclude patients from shunt surgery.