Figure 1.

Serum biomarkers of neurovascular unit and blood-brain barrier damage. Arterial and jugular venous blood draws were acquired simultaneously for both healthy controls (Ctr) and for patients with hypoxic ischemic brain injury (HIBI) with brain normoxia (NX; red) and hypoxia (HX; cyan). Data are presented as bars (median±interquartile range) with individual data overlaid for each group. Arterial serum biomarker concentrations for GFAP (glial fibrillary acidic protein; A), neurofilament-light (NF-L; B), total tau (C), NSE (neuron-specific enolase; D), and ubiquitin carboxy-terminal hydrolase L1 (UCH-L1; E) are presented on the top row. For arterial serum GFAP, NF-L, Tau, and NSE, both HIBI groups had elevated biomarkers compared with controls, while only patients with HIBI with brain hypoxia had higher UCH-L1 than controls. Simultaneous collection of arterial and jugular venous blood, and calculation of the consequent arterial-to-venous gradient (A-v) allows for determination of biomarker release from the brain, where a negative (−) A-v gradient indicates cerebral release. These data for each biomarker are depicted on the second row in F–J (note differing y axis scales). Note that left and right y-axes are used in F–J. An asterisk (*) indicates cerebral release (ie, A-v gradient) is significantly different from zero as determined by a one sample Wilcoxon Signed Ranks test (or 1-sample t test when data were normally distributed). The cerebral release of each biomarker in the normoxic HIBI group was not different from that of controls, whereas the patients with hypoxic HIBI had a greater cerebral release than controls and patients with normoxic HIBI for each biomarker. Significance testing between groups was conducted with a Kruskal-Wallis test. Following the detection of a significant effect, post hoc testing was conducted and corrected for multiple comparisons with a Dunn multiple comparisons test. P<0.05 was considered statistically significant.