Continuous versus intermittent cerebrospinal fluid drainage after severe traumatic brain injury in children: effect...

Drainage of cerebrospinal fluid (CSF) is routinely used in the treatment of severe traumatic brain injury (TBI), either continuously or intermittently in response to increases in intracranial pressure (ICP). There has been little study of the effect of CSF drainage method on the biochemistry, pathophysiology or outcome of TBI in adults or children. Having previously reported that a variety of markers of injury or repair increase in CSF after severe TBI, we chose to evaluate directly the effect of CSF drainage method on the biochemistry and volume of CSF drained as well as ICP. We hypothesized that concentrations of these markers would be similar in CSF drained continuously vs intermittently. We compared CSF levels of markers of neuronal injury (neuron specific enolase, [NSE]), glial injury (s100B), inflammation (interleukin-6 [IL-6]), and regeneration (vascular endothelial growth factor [VEGF]) (measured by ELISA) in 80 CSF samples from 19 severely injured children whose CSF was drained continuously (n = 13) versus intermittently (n = 6) as part of standard care in two institutions. Compared to continuous CSF drainage, intermittent drainage of CSF was associated with twofold greater CSF concentrations of NSE, s100B, IL-6 and VEGF (p < 0.05) and with about half the volume of CSF removal than continuous drainage (p = 0.002). The resulting elimination (concentration x volume) of these biochemicals, however, was not influenced by drainage method. Patients treated with continuous drainage had lower mean ICPs than those with intermittent drainage (13.6 +/- 0.69 vs. 21.8 +/- 0.95 mm Hg, p < 0.0001). We conclude that the method of CSF drainage greatly affects concentrations of CSF markers after TBI and may influence ICP. The influence of method on CSF marker concentration must be kept in mind when interpreting studies of CSF biomarkers. The striking difference in biomarker concentration, CSF volume drained, and ICP suggests the need for a randomized trial directly comparing these two approaches in infants and children with severe TBI.