Characterisation of serum total tau following paediatric traumatic brain injury: a case-control study

Abstract

Traumatic brain injury (TBI) is a major health problem in children. Blood-based biomarkers interpreted by use of normative values might improve the accuracy of diagnosis. Ultrasensitive assays can quantify serum concentrations of the neuronal microtubule-associated protein tau, which is increased in adult brains following TBI. We aimed to determine if serum total tau correlates with TBI diagnosis, severity, and radiological findings on CT scans in children younger than 18 years. In this case-control study, we included venous blood samples from healthy control children in the Canadian Laboratory Initiative on Pediatric Reference Intervals (CALIPER) biobank. For TBI cases, we recruited children (aged 0-17 years) who presented to the emergency department within 24 h of a TBI in three tertiary-care paediatric hospitals (Toronto, Vancouver, and Melbourne). Children were eligible if they required hospital observation for a minimum of 4 h or admission to the intensive care unit, and were excluded if they had had hospital treatment for a previous TBI, had birth trauma, or their parents could not speak English or French and therefore could not readily give consent. All available control samples were used and a case-control match was therefore not done. Venous and arterial blood samples were collected from patients with TBI within 28 h of injury (day 1). We used an ultrasensitive single-molecule immunoassay to measure serum total tau in blood samples. We first generated reference intervals of serum total tau from the control group, and used these normative data to interpret injury-associated changes in serum total tau in children with TBI. Concentrations of serum tau were measured in all CALIPER participants and patients with TBI, and no participants were excluded before analysis. We included samples from 416 control participants from the CALIPER cohort. Median total tau concentrations did not differ between sexes (p=0·12), but three significant reference intervals based on age groups were identified (1-3 years [0·88-19·2 pg/mL], 4-15 years [0·93-5·31 pg/mL], and 16-19 years [0·79-4·20 pg/mL]). Blood samples were obtained from 158 patients with TBI recruited between April 30, 2011, and June 28, 2013. Serum total tau on day 1 of TBI was negatively associated with Glasgow Coma Scale (GCS) score (rs=-0·42, 95% CI -0·55 to -0·28, p<0·0001). Median total tau was 2·86 pg/mL (IQR 1·52-4·83) in patients with GCS score 13-15 points (n=114), 7·08 pg/mL (3·75-41·1) in those with GCS score 9-12 points (n=13), and 8·48 pg/mL (2·53-70·6) in those with GCS score 3-8 points (n=31). Notably, participants who had GCS scores of 15 points had median total tau concentrations (2·57 pg/mL [1·50-4·61]) indistinguishable from those of control participants (2·46 pg/mL [1·77-3·42]), whereas those with GCS score 13-14 points had elevated total tau (6·41 pg/mL [2·97-42·5]). Serum total tau was not strongly associated with CT findings in patients with mild TBI. Serum total tau might help to differentiate between patients with mild TBI (GCS 13-14 vs GCS 15), but larger studies are needed to validate these results before this biomarker can be used for diagnosis and prognosis. Canadian Institutes of Health Research, Ontario Neurotrauma Foundation, and Victoria Neurotrauma Foundation.