Neuropathology of Mild Traumatic Brain Injury: Correlation to Neurocognitive and Neurobehavioral Findings

The use of computed tomography (CT) and magnetic resonance imaging (MRI) in mild traumatic brain injury (mTBI) is overviewed in this chapter. Although in the majority of mTBI cases no abnormality will be shown, the common neuropathological changes that may be identified on CT and/or MRI are highlighted with an emphasis that such abnormalities provide only a macroscopic perspective of the pathology that may be viewed. Emphasis is placed on understanding the subtle nature of neuropathology that may accompany mTBI, the potential for dynamic changes that vary with time post-injury and that detection depends on which neuroimaging method is used. The role of advanced neuroimaging techniques that provide quantitative information about potential network-level damage using diffusion tensor imaging (DTI) and resting state functional MRI is overviewed with numerous examples provided that illustrate neuroimaging techniques that detect mTBI abnormalities. The common structural neuroimaging methods and findings in mTBI will be overviewed. The field of neuroimaging is expansive and the basics of neuroimaging will not be covered in this review. For the reader who would like additional background information in neuroimaging of TBI, Wilde et al. (2012) provide such a synopsis. The chapter will conclude with a section that relates the macroscopically identified pathologies using conventional and advanced neuroimaging techniques with the ultrastructure and underlying pathophysiology of mTBI. The neuroimaging investigation by Yuh et al. (2013) represents a comprehensive investigation of the common, visibly detected abnormalities observed in mTBI using day-of-injury (DOI) computed tomography (CT) followed by magnetic resonance imaging (MRI) during the subacute timeframe. In the Yuh et al. study, 135 mTBI patients were evaluated for acute head injury in three separate level 1 trauma centers in the United States and all were enrolled through an emergency department (ED) for prospective 3-month neurobehavioral outcome, assessed by the Extended Glasgow Outcome Scale (GOS-E). Although DOI CT imaging was done acutely, MRI on average was performed within 2 weeks postinjury. The National Institutes of Health has established the TBI Common Data Elements (TBI-CDEs; Haacke et al., 2010; Yue et al., 2013) for classifying both acute as well as chronic abnormalities, with all scan abnormalities identified by CDE criteria. CDE guidelines for pathoanatomical TBI findings on DOI CT or early MRI include skull fracture, hematoma (either epidural and/or subdural), traumatic axonal injury (defined as one to three foci), and diffuse axonal injury (DAI; defined as at least four foci). DOI CT foci are typically characterized as visibly identified contusions or intraparenchymally identified petechia. On MRI, such foci may take the form of white matter (WM) signal abnormality (hyperintense) and/or characteristic signal changes (hypointense) that reflect prior hemorrhage, often at the gray matter (GM)-WM interface. All of these types of macroscopic pathologies will be depicted in this chapter. Importantly in the Yuh et al. investigation, TBI-CDE features of more severe TBI such as midline shift ≥5 mm and partial/complete basal cistern effacement were not observed in any of the mTBI patients as part of that study. This is understandable and highlights that the visible abnormalities in mTBI do not reach the threshold associated with more severe TBI; nonetheless, very significant parenchymal injury may accompany mTBI. The 2013 Yuh et al. investigation was a subset of a much larger investigation (McMahon et al., 2013) that prospectively followed 375 mTBI patients at 3, 6, and 12 months. The McMahon et al. (2013) study found that by 1 year, 22.4% of mTBI patients were still below functional status as measured by the GOS-E. Although there was an association of positive CT findings with poorer 3-month outcome, by 1 year whether DOI CT was abnormal or not did not predict outcome. Clearly, mTBI results in lasting sequelae for some, but this is not necessarily predicted by DOI CT findings. As will be shown in this chapter, advanced neuroimaging studies provide additional information and insight into the neuropathological effects of mTBI potentially useful in better understanding mTBI sequelae as well as providing additional information in the assessment and treatment of mTBI. Figure 31.1 summarizes the findings of Yuh et al., which show that 44% of all mTBIs in this cohort of ED assessed individuals with mTBI had at least an identifiable neuroimaging abnormality. Clearly MRI was superior to CT in identifying abnormalities, especially those neuroimaging markers that infer axonal pathology. In fact, 27% of mTBI patients with normal head CTs had abnormal MRIs that were otherwise “missed” by DOI CT imaging. Of the 135 mTBI patients assessed in the Yu et al. investigation, only one had a Glasgow Coma Scale (GCS) of 13, with 26/135 (19%) assessed with a GCS of 14 and 108/135 (80%) with a GCS of 15. As such, the majority had a classically defined GCS score, yet almost half had some positive neuroimaging finding. This observation underscores the frequency with which MRI may identify structural pathology in mTBI, even with a GCS of 15. In terms of the frequency of CDE findings and their relation to outcome, presence of any type of CDE-identified TBI abnormality increased the likelihood of lower GOS-E at 3 months, supporting the importance of identifying neuroimaging-based MRI abnormalities because of increased sensitivity in detecting gross pathology (Bigler, 2013a,b). However, the majority scanned had negative conventional imaging. For those with DOI CT, presence of subarachnoid hemorrhage was associated with poorer 3-month GOS-E. For those with positive MRI findings in the subacute time frame, presence of contusion or DAI was found to predict lower GOS-E. Figure 31.2 from the Yuh et al. investigation depicts some common CT and MRI findings in mTBI during the acute to early subacute timeframe. What is depicted in this illustration shows many of the classic observable, macroscopic lesion types in mTBI, which will be discussed more fully throughout this chapter. Impressively, the Yuh et al. study shows that greater than 40% of mTBI patients initially evaluated within the ED will have positive CDE-identified abnormalities. However, the CDE technique requires visual identification of the abnormality from conventional clinical imaging (CT and/or MRI) studies and does not incorporate advanced magnetic resonance (MR) techniques to be discussed in subsequent sections. Nonetheless, it is important to understand what constitutes early neuroimaging identified abnormalities and how such findings relate to underlying neuropathology. Conventional CT and MRI clinical studies configure anatomical images with millimeter resolution, meaning they detect gross pathology at a similar level, although submillimeter MR resolution is now possible (Yassa et al., 2010; Heidemann et al., 2012). In contrast, the fundamental pathological changes that occur from TBI happen at the micron and nanometer cellular level (Bigler and Maxwell, 2011, 2012), with only the largest of lesions being visible with contemporary neuroimaging (Bigler, 2013b). This means for brain injuries in the mild range, with the subtlest of neural injury that the macroscopic lesions will not be observed. However, as will be discussed in this chapter, this does not mean that those with negative imaging have no underlying pathology.