Brain Fluid Clearance After Traumatic Brain Injury Measured Using Dynamic Positron Emission Tomography

Abstract

Brain fluid clearance by pathways including the recently described paravascular glymphatic system is a critical homeostatic mechanism by which metabolic products, toxins, and other wastes are removed from the brain. Brain fluid clearance may be especially important after traumatic brain injury (TBI), when blood, neuronal debris, inflammatory cells, and other substances can be released and/or deposited. Using a non-invasive dynamic positron emission tomography (PET) method that models the rate at which an intravenously injected radiolabeled molecule (in this case ¹¹C-flumazenil) is cleared from ventricular cerebrospinal fluid (CSF), we estimated the overall efficiency of brain fluid clearance in humans who had experienced complicated-mild or moderate TBI 3-6 months before neuroimaging (n = 7) as compared to healthy controls (n = 9). While there was no significant difference in ventricular clearance between TBI subjects and controls, there was a significant group difference in dependence of ventricular clearance upon tracer delivery/blood flow to the ventricles. Specifically, in controls, ventricular clearance was highly, linearly dependent upon blood flow to the ventricle, but this relation was disrupted in TBI subjects. When accounting for blood flow and group-specific alterations in blood flow, ventricular clearance was slightly (non-significantly) increased in TBI subjects as compared to controls. Current results contrast with past studies showing reduced glymphatic function after TBI and are consistent with possible differential effects of TBI on glymphatic versus non-glymphatic clearance mechanisms. Further study using multi-modal methods capable of assessing and disentangling blood flow and different aspects of fluid clearance is needed to clarify clearance alterations after TBI.

Keywords: PET scanning; adult brain injury; cerebrospinal fluid; clearance; glymphatic system.

FIG. 1.

Ventricular clearance measured using dynamic PET. A single subject's PET images, superimposed over co-registered MRI, show decreasing radiotracer in lateral ventricles over the 1-h scan period. Image from Li and colleagues. MRI, magnetic resonance imaging; PET, positron emission tomography

FIG. 2.

Clearance compartment models evaluated.

FIG. 3.

TBI versus control differences in ventricular clearance and the dependence of ventricular clearance on blood flow to the ventricle. (A) Plot showing the relationship between tracer/blood delivery to the ventricle (k1_blood) and ventricular clearance (k_clearance) in controls and TBI subjects. In controls, ventricular clearance is highly, linearly dependent upon tracer/blood delivery to the ventricle (R = 0.95, p < 0.001). In TBI subjects, the correlation between ventricular clearance and tracer/blood delivery to the ventricle is not significant (R = 0.69, p = 0.084). (B) Bar plot of estimated marginal means (controlling for k1_blood and the k1_blood × group interaction) of ventricular clearance in TBI subjects and controls. Ventricular clearance was numerically (though not significantly) greater in TBI subjects as compared to controls when accounting for these factors. SE, standard error; TBI, traumatic brain injury.