Brain-derived tau in plasma and cerebrospinal fluid after surgery: a prospective cohort substudy.

Abstract

Background: Postoperative increases in plasma total tau (t-tau) protein have raised concerns that surgery can cause CNS injury. Brain-derived tau (BD-tau) is a novel biomarker that quantifies CNS-derived tau. We assessed BD-tau concentrations in plasma and CSF after surgery to investigate neuronal injury and mechanisms driving increases in plasma t-tau.

Methods: This was a substudy of a previous prospective cohort study enrolling adult patients undergoing hip and knee arthroplasty. We measured plasma and CSF BD-tau and t-tau before neuraxial anaesthesia (baseline) and at 4, 8, 24, 32, and 48 h after surgery and assessed cognitive performance at baseline and 3 months. Biomarker concentrations are reported as median (IQR) percentage change from baseline.

Results: Plasma BD-tau peaked at 4 h with a 254% increase (137-382%, P<0.001) and remained elevated by 31% (6-75%, P<0.001) at 48 h. CSF BD-tau was unchanged at 4 h but increased by 51% (21-82%, P=0.004) at 48 h. Plasma BD-tau was associated with plasma t-tau (β=1.09, 95% CI 0.95-1.23, P<0.001), whereas plasma BD-tau and CSF BD-tau were inversely associated (β=-0.71, -0.97 to -0.45, P<0.001). Higher postoperative plasma BD-tau was associated with both higher CSF albumin and worse cognitive performance at 3 months.

Conclusions: Early postoperative increases in plasma t-tau are consistent with a CNS origin but not concomitant neuronal injury. Instead, the early plasma BD-tau increase and the inverse association between CSF and plasma BD-tau could indicate a redistribution of tau from the CNS to plasma.