Proteomic blood-based biomarkers of brain damage in traumatic brain injury: are they suitable surrogate endpoints for cerebral pressure autoregulatory-guided therapy?

Management after severe traumatic brain injury (TBI) aims to prevent secondary injury by optimizing cerebral physiology, yet conventional metrics such as intracranial pressure (ICP) and cerebral perfusion pressure (CPP) incompletely capture the underlying pathophysiology. Impaired cerebral pressure autoregulation (CPA) is common and may exacerbate secondary injury. Proteomic blood-based biomarkers (PBBMs) of astrocytic (glial fibrillary acidic protein [GFAP], S100 calcium-binding protein B [S100B]), and neuronal/axonal injury (total tubulin associated unit [t-Tau], neurofilament light chain [NfL], ubiquitin C-terminal hydrolase-L1 [UCH-L1], neuron-specific enolase [NSE]) may provide global indicators of secondary injury. This study investigated temporal associations between PBBMs and cerebral physiological variables (ICP, CPP, pressure reactivity index [PRx], and CPP deviation from “optimal” CPP [ΔCPPopt]), and evaluated the PBBMs as surrogate short-term endpoints.

This retrospective, observational multi-center study, used prospective data from the CENTER-TBI cohort and included 151 patients with high-frequency cerebral physiological data and serial PBBM measurements during the first seven days post-injury. Associations were analyzed using Spearman correlations, univariate and multivariate linear mixed effects models (LMEMs), as well as cross correlation analyses adjusted for repeated measures and clinical confounders.

Elevated PBBM levels on day 1 were associated with a greater cumulative burden of high ICP, impaired CPA (positive PRx), and negative ΔCPPopt over the subsequent seven days. Throughout the monitoring period, higher median PBBM concentrations correlated with elevated ICP and impaired CPA, with strongest associations when cerebral physiological monitoring and PBBMs were analyzed from the same day. In multivariate LMEMs, elevated ICP and most PBBMs remained significantly associated, while particularly PRx amplified this association.

Cerebral physiological disturbances during intensive care after TBI were associated with elevated PBBM levels, with ICP showing the strongest influence and impaired CPA amplifying some PBBM responses. Associations were strongest in same-day analyses and appeared bidirectional. These findings suggest that PBBM may reflect the burden of secondary cerebral insults and may serve as candidate endpoints for future trials targeting cerebral physiology, pending validation in prospective high-resolution studies. Importantly, our results highlight the potential for PBBMs to complement multimodal neuromonitoring by tracking secondary injury trajectories.