Postconcussional disorder and PTSD symptoms of military-related traumatic brain injury associated with compromised neurocircuitry

Abstract

Traumatic brain injury (TBI) is a common combat injury, often through explosive blast, and produces heterogeneous brain changes due to various mechanisms of injury. It is unclear whether the vulnerability of white matter differs between blast and impact injury, and the consequences of microstructural changes on neuropsychological function are poorly understood in military TBI patients. Diffusion tensor imaging (DTI) techniques were used to assess the neurocircuitry in 37 U.S. service members (29 mild, 7 moderate, 1 severe; 17 blast and 20 nonblast), who sustained a TBI while deployed, compared to 14 nondeployed, military controls. High-dimensional deformable registration of MRI diffusion tensor data was followed by fiber tracking and tract-specific analysis along with region-of-interest analysis. DTI results were examined in relation to post-concussion and post-traumatic stress disorder (PTSD) symptoms. The most prominent white matter microstructural injury for both blast and nonblast patients was in the frontal fibers within the fronto-striatal (corona radiata, internal capsule) and fronto-limbic circuits (fornix, cingulum), the fronto-parieto-occipital association fibers, in brainstem fibers, and in callosal fibers. Subcortical superior-inferiorly oriented tracts were more vulnerable to blast injury than nonblast injury, while direct impact force had more detrimental effects on anterior-posteriorly oriented tracts, which tended to cause heterogeneous left and right hemispheric asymmetries of white matter connectivity. The tractography using diffusion anisotropy deficits revealed the cortico-striatal-thalamic-cerebellar-cortical (CSTCC) networks, where increased post-concussion and PTSD symptoms were associated with low fractional anisotropy in the major nodes of compromised CSTCC neurocircuitry, and the consequences on cognitive function were explored as well.

Keywords: blast injury; concussion; diffusion tensor imaging; neurocircuitry; post-traumatic stress disorder; traumatic brain injury.

Figure 1

Maps of significant group differences of voxel‐wise analyses of FA (A (FWE=5%, TFCE P ≤ 0.05), B (FWE=5%, TFCE P ≤ 0.06)), trace (C left (light blue, FWE=5%, c > 2.3, P ≤ 0.05)), radial diffusivity (C middle (yellow, FWE=5%, c > 2.5, P ≤ 0.05)), and parallel diffusivity (C right (blue uncorrected P ≤ 0.05)) and reconstructed tracts (D) using lower FA (A) as seeds in axial, sagittal and coronal views overlaid on the mean FA template (1 × 1 × 1 mm³). acr, anterior region of corona radiata; alic, anterior limb of internal capsule; bCC, body of corpus callosum; cst, corticospinal tract in brainstem; cyg, cingulum; gp, globus pallidus; pu, putamen; ec, external capsule; fx, fornix; ilf, inferior longitudinal fasciculus; pct, pontine crossing tracts; scr, superior region of corona radiata; slf, superior longitudinal fasciculus; ss, sagittal stratum, including inferior longitudinal fasciculus and inferior fronto‐occiptal fasciculus; tha, thalamus; unc, uncinate fasciculus.

Figure 2

Tract‐specific analysis of fractional anisotropy using the method of continuous medial representations. Significant low FA clusters of the TBI group are shown in regions of superior longitudinal fasciculi (T statistics (A) and p values (B), P ≤ 0.05), where the color indicates the significance of T statistics and P values.

Figure 3

Examples of tractography using the low FA clusters as seeds (red), identified by comparing the individual TBI participant to the mean of healthy controls (Z < −3.5), in TBI participants (left of each subfigure), and in the gender and age matched controls using the same seeds (right of each subfigure). A: A mild blast TBI participant (male, 29 years old, 59 days post‐injury) injured by rocket‐propelled grenade with complaints of dizziness and problems of balance, hearing, vision, memory and concentration. Close‐up view of fiber “breaks” (discontinuity) in middle cerebellar peduncles in the right lower corner of TBI patient (A, Left), compared with the control (A, Right). B: A mild nonblast TBI participant (male, 24 years old, 156 days post‐injury) injured by mine‐resistant ambush‐protected vehicle rollover with complaints of problems with memory, mood, and cognition. Low FA clusters were mainly at prefrontal and parietal regions, as well as pyramidal tracts (red). C: A moderate blast TBI patient (male, 25 years old, 65 days post‐injury) injured by Improvised Explosive Devices (IED) blast with complaints of attention, irritability and cognition. Low FA clusters were found at forceps minor, DLPFC and fornix (arrows). Close‐up view shows fiber “breaks” in the DLPFC in the left upper corner, and in the fornix in the left lower corner (C. Left). D: A moderate blast TBI patient (male, 24 years old, 32 days post‐injury) injured by IED blast with complaints of anxiety, sleep problems, and hearing loss. Close‐up view shows fiber “breaks” in fornix in the left upper corner (D, Left), compared to the intact fornix of one control participant in the left upper corner (D, Right) (see Supporting Information Fig. 3 for more illustrations of the same TBI subject).

Figure 4

Results of the negative associations between FA and NBSI (A), PCLC (B) total scores, days post‐injury (C), and their scatter plots of the mean FA of all significant clusters (RFT P ≤ 0.05) vs. NBSI (A), PCLC (B) total scores, and days post‐injury (C) in TBI participants, with T statistics of significant clusters (RFT P ≤ 0.05) overlaid on the mean FA image. A: Negative association between NBSI and FA at the ventral striatum and ventral cingulum bundles. B: Negative association between PCLC and FA at the ventral striatum, ventral cingulum bundles, genu of the internal capsules, superior corona radiata, and fronto‐parietal association fibers. C: Negative association between post‐injury duration and FA at the superior corona radiata.

Figure 5

Results of the negative associations between FA and NBSI (A, C) and PCLC (B, D) total scores in nonblast TBI (A, B) and blast TBI (C, D) patients, and their scatter plots of the mean FA of all significant clusters (RFT P ≤ 0.05) vs. NBSI (A, C) and PCLC (B, D) scores with T statistics (color bars) of significant clusters (RFT P ≤ 0.05) overlaid on the mean FA image. Although similar to the findings of Figure 4, there are more anterior‐posteriorly oriented clusters in nonblast TBI (A,B), but more centrally and inferior‐superiorly oriented clusters in blast TBI (C,D).

Figure 6

Results of hemispheric asymmetry TBSS of FA between 3 groups: HC vs. nonblast TBI (A), HC vs. blast TBI (B), and nonblast TBI vs. blast TBI (C). Significant hemispheric asymmetry clusters (RFT P ≤ 0.05, dilated for better viewing, red for greater left‐minus‐right asymmetry and blue for lesser left‐minus‐right asymmetry) are overlaid on the mean FA skeleton (green) and MNI FA template. TBI differed from HC in the regions of bodies of CC, forceps major/anterior corona radiata, the genu of internal capsule, inferior temporal and cerebellar white matter (A, B). In addition, nonblast TBI had more (a)symmetry clusters over the anterior‐posteriorly oriented fibers such as anterior limb of internal capsule, superior longitudinal fasciculus and sagittal stratum (A,C), while blast TBI had more over the centrally, inferior‐superiorly oriented fibers such as superior corona radiata, pyramidal tracts, and brainstem fibers (B, C).

Figure 7

Results of the associations between FA and neuropsychological (NP) testing of Wechsler Test of Adult Reading (WTAR) (A), Conners' Continuous Performance Test (CPT)‐II Hit Reaction Time (B) and Response Style (C) in TBI patients. Significant clusters (t values in (A) and (B), and mask in (C, red)) are overlaid on the mean FA image (A, B) and MNI T1 template (C) (RFT P ≤ 0.05) with their scatter plots of the mean FA vs NP scores. A: WTAR score positively correlated with FA in the anterior thalamic radiation (atr), bodies of corpus callosum (bcc), and occipital (occ) white matter. B: CPT‐II Hit Reaction Time correlated negatively with FA in bilateral corona radiata (SCR), but negatively with FA in the fornix (fx), posterior dorsal cingulum bundle (dCy) / bodies of corpus callosum, and precuneal (Pcu) white matter. C: CPT‐II Response Style correlated positively with FA in the ventral striatum.