Early Metabolic Changes Following Focal Traumatic Brain Injury in Rats Measured Using 1 H MRS

3 voxel. The voxel covered immediate pericontusional zone, all layers of the hippocampus, and superior thalamic structures. Data were acquired before injury (baseline), at 3-hours, and at 5-hours after injury in both pericontusional voxel (Fig 1A) and the corresponding contraleteral side (Fig 1B). For each spectrum, 300 acquisitions were averaged for a total of 13 min. At all times during the experiment, the animal was under 1-2% isoflurane anesthesia and 1 L/min oxygen administration. Respiratory monitoring was performed and the animal was maintained at 36-37 o C during the entire experiment. Proton MRS data was fitted using the LC Model package, and only metabolites with standard deviations (SD) % < 20 were included for further analysis. The in vivo mean metabolite concentrations relative to tCr at each time point were subjected to paired one-tail Student t-test in comparison with the control time point. Results Fig 1 demonstrates the bilaterally anatomic images with the voxel located in the axial view (A and B) and the corresponding bilateral spectra (C and D) at 3-hours after TBI from a rat brain. The in vivo 1 H spectra demonstrate good spectral resolution and sensitivity both at the pericontusional side and the contraleteral side. Among the metabolic ratios of, NAA/tCr, Glu/tCr and Cho/tCr demonstrated significant changes over the five hours following injury as shown in Fig 2. No statistically significant differences were found in glutamine, myo-inositol, and taurine concentrations among the three time points in either the pericontusional voxel itself or in comparison to the contraleteral side. Significant reduction of 32 % and 33 % NAA was observed in the pericontusional voxel at 3-hours and 5-hours after TBI respectively compared to the baseline. Although the contraleteral voxel also exhibited significant reduction in NAA this reduction was much lower compared to the pericontusional side. No significant differences in NAA were found in the pericontusional side between the 3 and 5 hours. In addition to NAA, our results showed that Glu significantly decreased at 3- hours after TBI in the pericontusional voxel, compared to the baseline (0.922 ± 0.137 vs. 1.155 ± 0.202, p<0.03) and the contraleteral side (0.922 ± 0.137 vs. 1.12 ± 0.08, p<0.04). As with NAA, we did not observe a significant difference between 3-hours and 5-hours in Glu level in the pericontusional side. Cho in the pericontusional voxel was significantly lower than the contraleteral side (0.166 ± 0.014 vs. 0.179 ± 0.013, p<0.05) at 3-hours after TBI. Although, the signal intensities of Lac were undetectable during the baseline, varying levels (0.115 - 2.098) of increased Lac signal intensity was observed in the pericontusional voxel at 3-hours and 5-hours after the injury, but not in the corresponding contraleteral voxel. Discussion This study shows that there exists a temporal window of brain vulnerability after TBI in rat, which is in line with previously studies (2). Furthermore, our investigation demonstrates that the neuro-metabolic changes following TBI associated with NAA, Glu and Cho may have their most significant changes as early as three hours after the injury. Since the pericontusional voxel chosen in this study is of special clinical interest for neuroprotective treatment strategies, our finding may indicate a temporal window of about 3 hours for planning interventions that target cerebral energy metabolism. Acknowledgement