Traumatic Brain Injury

TBI is the leading cause of death and morbidity in young, otherwise healthy young adults and children. Due to variability in the form and extent of the initial injury, as well as individual variation in response to injuries, prediction of outcome following TBI is challenging and generally inadequate [16]. Common clinical markers, such as the Glasgow Coma Scale (GCS) are linked statistically with outcome in large samples, but are less useful in predicting cognitive functioning in individual patients. CT is generally the preferred neuroimaging technique in the acute evaluation of TBI, while MRI studies are especially valuable in subacute or chronic forms due to its greater sensitivity to diffuse axonal injury. However, quantitative estimates of the extent of brain trauma with these methods have proven of only modest value in predicting concurrent cognitive status or cognitive outcome [17]. The most important limitation of imaging studies of TBI is a relative insensitivity to neuronal dysfunction as opposed to neuronal death.

We have noted in samples of both children and adults with TBI that NAA and Cho concentrations are highly predictive of cognitive deficits. In our initial study, nineteen adults with TBI (mean age = 32.1 years, SD = 13.2, 5% female, mean acute Glasgow Coma Scale = 8.4, SD = 3.4) were compared with 28 normal controls (mean age = 26.6 years, SD = 11.1, 39% female [18]). Most TBI patients were assessed at three time points (median = 38 days, 125 and 184 days post-injury). Figure 1 shows the location of the gray matter (GM) and white matter (WM) voxels studied, as well as representative spectra. Spectroscopic data were analyzed with MRUI (Leuven, Belgium) and normalized for CSF within the voxel, providing an estimate of absolute neurometabolite concentrations. A large battery of neurocognitive tests were administered, including measures of attention, memory, processing speed, and "executive" functioning. To reduce the number of variables examined and provide an overall measure of cognitive status, each test score was standardized and then averaged to create a "composite z-score". Additional methodological details are provided in [19, 20].

Greater GM NAA predicted better overall cognitive functioning at each time point (r's = .63, .70, and .57). Relevant to the clinical utility of 1H-MRS, initial (i.e., sub-acute) GM NAA also predicted overall neuropsychological function at outcome, roughly six months post injury (r = .70): moreover, better cognitive performance was seen in patients whose GM NAA improved over time. For a smaller sample of patients we also assessed frontal WM NAA. This measure was highly correlated with overall cognitive function at the initial time point (r= .89), but did not predict subsequent neuropsychological performance or long-term outcome [unpublished data].

We believe these results suggest that MRS may be able to detect the neurobiological substrate of diffuse axonal injury producing neurocognitive deficit relatively early in the disease process. It is important to note that the voxels studied were free from overt abnormality, as determined by independent clinical neuroradiological examination. Consistent with this observation, Gasparovic and colleagues [21] concurrently demonstrated in the weight-drop animal model of TBI that NAA/Cre ratios were notably reduced in brain tissue remote from the impact site that prior histological studies have shown to be free from any cellular damage.

Cre NAA

Cre NAA

Canavan Disease
Good outcome

Impairment

Figure 1. White matter (WM) and gray matter (GM) voxels analyzed in studies of adults with traumatic brain

Good outcome

Good outcome

Impairment

Impairment

Figure 1. White matter (WM) and gray matter (GM) voxels analyzed in studies of adults with traumatic brain

Reduced NAA concentrations can reflect neuronal death, but this mechanism cannot account for the reversible decreases in NAA observed in multiple sclerosis [22] and stroke [23], or the possible increases over time, post-injury, observed in our TBI sample [18]. Given Gasparovic's results, and the fact that our NAA analyses were undertaken in normal appearing tissue, it may well be the case that individual variation in metabolism, rather than loss of neurons, underlies the well-established relationship between NAA and concurrent and long term cognitive function. Consistent with this possibility, Positron Emission Tomography analysis has shown that NAA concentrations correlate with overall metabolic rate [24], which is often reduced after TBI.

Our recent studies with children have utilized Spectroscopic Imaging (SI), rather than the single voxel methodology described above [25]. A spectroscopic imaging slice was selected above the lateral ventricles to extend from the frontal lobe to the occipital and parietal lobes sampling both white and gray matter. Water suppressed PRESS localization with outer voxel suppression bands was used to excite parenchyma and avoid lipid artifact from the skull. See Figure 2 (panel A) for an example MRI image with superimposed SI grid. Patient recruitment and data analyses are ongoing. Here we report on NAA/Cre ratios averaged across many voxels (predominantly white matter), with separate estimates for anterior and posterior halves of the data matrix. A large battery of cognitive tests was administered; individual scores were standardized and averaged into these composites: overall cognitive (as above), motor, language, visuo-motor, and working memory. The TBI children (N = 28, mean age = 13.6, SD = 3.6, 21% female; mean GCS = 7.8, SD = 4.7) were compared to healthy controls (N = 14, mean age = 15.2, SD = 1.7, 56% female). As expected, the TBI children had much lower NAA/Cre and injuries.

greater Cho/Cre, overall and in each half of the brain. They also scored lower on each cognitive composite.

Table 1. Correlations of NAA/Cre from total supraventricular MRS slab, and anterior and posterior halves of the MRS slab, with age adjusted neurocognitive domain scores in pediatric TBI patients (N = 28).

Was this article helpful?

0 0

Post a comment