Almost two million Americans sustain a traumatic brain injury (TBI) every year, making TBI a significant public health concern (Faul, Xu, Wald, & Coronado, 2010). Over forty percent of these injuries are sustained by children and adolescents, resulting in approximately 630,000 emergency department visits, 60,000 hospitalizations, and 6,000 deaths per year within the 0-19 years age range (Faul et al., 2010). Of all neurological conditions in children, TBI has been the most common and has often resulted in diminished psycho-social functioning (Pierson & Noggle, 2010) as well as a variety of emotional and behavioral symptoms along with deficits in attention, working memory, processing speed, perceptual organization, and verbal memory (Wozniak, Krach, Ward, Mueller, Muetzel et al., 2007). Executive function is also commonly affected (Kinnunen, Greenwood, Powell, Leech, Hawkins et al., 2011). Long-term cognitive and motor dysfunction is common even in children with only mild to moderate head injuries, and the estimated annual economic burden of pediatric TBI in the US is between ten and fifteen billion dollars, not to mention additional expenses associated with long-term services (Adelson & Kochanek, 1998).
While the overall morbidity rate for children following TBI is lower than that of adults, the morbidity rate of younger children (less than age 4 years) and mid-to-late adolescents is higher than that of school-aged children. Children under four years of age often suffer more frequent, diffuse, and severe injuries, whereas adolescents more frequently suffer high impact injuries. Brain injuries in school-aged children tend to be more focal and less severe (Adelson & Kochanek, 1998).
Scientific advances have been made toward a better understanding of pediatric TBI; however, many questions have remained unanswered in the scientific literature such as regarding the underlying pathophysiology of TBI impairments in children and adolescents (Kinnunen et al., 2011). For example, few studies have examined brain morphology and function in pediatric TBI, particularly regarding the effects of diffuse axonal injury (DAI), also known as traumatic axonal injury (TAI). Within this framework, the purpose of the present study is to better understand the relationship between white matter integrity (as represented by the corpus callosum) and function (as represented by interhemispheric transfer time).
Mechanisms and Outcomes of Pediatric TBI
The mechanisms and sequelae of TBI are heterogeneous and thus a broad nosology is required. Among other important distinguishing factors, TBI can be classified as closed head injury (CHI) versus open or penetrating, and focal versus diffuse, in terms of damage to the brain. Diffuse CHIs represent 95% of all pediatric TBIs (Adelson & Kochanek, 1998). Focal injuries are those resulting in localized damage (typically to gray matter) such as cerebral contusions or intracranial hemorrhages; these injuries have usually been easier to detect with imaging techniques because they tend to be macroscopic (Wu, Wilde, Bigler, Li, Merkley et al., 2010). In contrast, DAI represents widespread (versus focal) damage to white matter via axonal shearing or stretching as a result of inertial forces that cause the brain to rapidly accelerate, decelerate, or rotate within the skull (Caeyenberghs, Leemans, Geurts, Linden, Smits-Engelsman et al., 2011). The widespread axonal damage is predominantly found in the 'parasagittal white matter of the cerebral cortex, corpus callosum, and the pontine-mesencephalic junction adjacent to the superior cerebellar peduncles' (Meythaler, Peduzzi, Eleftheriou, & Novack, 2001).
CHI severity is most commonly classified according to Glasgow Coma Scale (GCS) scores. This scale assesses the level of impaired consciousness in coma (Teasdale & Jennett, 1974). In a pediatric CHI study, Benavides, Fletcher, Hannay, Bland, Caudle et al. (1999) grouped patients' injury severity according to GCS scores: 13-15 represented mild CHI and <12 represented moderate to severe CHI. A score of <8 indicates severe brain injury with the patient unable to open his/her eyes, follow simple movement commands, or vocalize words.
Typical Brain Development
The study of the pediatric brain is complicated by developmental processes. TBI is itself a dynamic process, but in children pathological changes must be interpreted within the context of the dynamic changes that naturally occur through neurologic and cognitive stages of development. These include changes in overall brain size and in substructures such as the corpus callosum. In the absence of pathological processes, the CC begins to develop in utero and continues to develop into young adulthood. Although the number of callosal fibers (over 200 million) is fixed around birth, myelination, redirection, and pruning of fibers continue during postnatal development (Luders, Thompson, & Toga, 2010).
In children, the CC follows a predictable course of myelination (Wu et al., 2010). In normal development, this myelination process, particularly of the middle and posterior subregions of the CC, spans the first 18 years of life or more and results in increased area of the isthmus and splenium throughout childhood and adolescence (Rajapakse, Giedd, Rumsey, Vaituzis, Hamburger et al., 1996; Toga, Thompson, & Sowell, 2011). Myelination is especially rapid in the first ten years of life (Denckla, 1974; Levin et al., 2000), but continues throughout young adulthood (Pujol, Vendrell, Junqu??, Mart??-Vilalta, & Capdevila, 1993). Because the posterior CC continues myelination into adulthood, its vulnerability to injury is of particular concern.
In another study of healthy children (ages 4-18), Rajapakse et al. (1996) found that CC growth was greatest in the splenium and isthmus. Changes in the midbody regions were also noted. As these regions increased in area, they became more compact, with clearer boundaries increasing with age.
Functional Outcomes of DAI
DAI is the most common and problematic feature of TBI, with the highest occurrence in severe TBI, yet it is difficult to detect with most imaging techniques because it is primarily microscopic (Smith, Meaney, & Shull, 2003; Meythaler et al., 2001). Children are especially vulnerable to DAI 'because of their unique anatomical relationships of the head to body ratio, weak neck musculature, and the lack of myelination' (Adelson & Kochanek, 1998). Furthermore, functional recovery is less likely in DAI because neuroplasticity appears to apply more to focal damage and is thus more of a gray matter phenomenon, though damaged white matter axons can potentially mend if they did not disconnect entirely (Smith et al., 2003).
DAI and the CC. As a central fulcrum of the brain, the corpus callosum (CC) is the largest white matter bundle in the brain and, as such, is one of the most commonly damaged structures in DAI (Levin et al., 2000), especially in moderate-to-severe TBI (Wu et al., 2010). Focal hemorrhages and shear strains of the CC are common features of DAI, with CC lesions present in 20% of children suffering severe CHI, most of the damage being to the posterior body and splenium (Mendelsoh, Levin, Harward, & Bruce, 1992; Levin, Benavidez, Verger-Maestre, Perachio, Song et al., 2000) which are more susceptible to fiber disruption compared to anterior portions (Rutgers, Fillard, Parador, Tadi??, Lasjaunias et al., 2008).
The CC is the crossroads of widespread interhemispheric connections and its primary function is to mediate the communication of complementary and supplementary information between hemispheres, thus affecting a wide range of cognitive domains. As viewed from the midsagittal plane from anterior to posterior, the subregions of the CC include the genu, the rostrum (ventral to the genu and bending underneath the CC), the rostral body, anterior and posterior midbodies, the isthmus, and the splenium. The CC is organized topographically; fibers in the anterior portions connect frontal regions of both hemispheres and fibers in the posterior portions connect posterior cortices. This organization is functionally topographic as well; the rostrum likely mediates higher cognitive information, the anterior midbody mediates motor information, the posterior midbody mediates somatosensory information, the isthmus mediates auditory information, and the splenium mediates visual stimulation (de Lacoste, Kirkpatrick, & Ross, 1985; Fabri, Polonara, Mascioli, Salvolini, & Manzoni, 2011). Following extensive damage to the CC, the importance of this prominent commissural white matter tract is evident in callosal disconnection syndrome which includes but is not necessarily limited to hemianomia, agraphia, esthesia, apraxia, and hemialexia (Vuilleumier & Assal, 1995). However, CT and various MR techniques often underestimate the frequency of axonal injuries to the CC because these are typically microstructural (Rutgers et al., 2008).
The anterior portions of the CC have commonly been found to be atrophied following CHI as these portions connect to the frontal and temporal lobes which are often damaged (Wilde, Hunter, Newsome, Scheibel, Bigler et al., 2005). However, the posterior portions of the CC, particularly the posterior body and splenium, are typically the most affected in DAI. This is because the falx cerebri restricts and therefore places greater strain on the posterior regions of the CC, consistent with the shear strain gradient of DAI (Levin et al., 2000; Benevidez et al., 1999).
Neurocognitive Outcome Following CC DAI. Verger et al. (2001) studied the relationship between neurocognitive functioning and atrophy (ventricular volumes and CC area) in a group of children with a history of moderate-to-severe CHI and controls over six years post insult. Because of multiple comparisons, differences in neurocognitive functioning were considered significant only at p < 0.01. The investigators found that the patients' scores were significantly lower than controls in visual memory and a visuospatial measure (Rey Complex Figure). Though not statistically significant, patients' scores on measures of intelligence, a different visuospatial measure (Benton Line Orientation), verbal memory, phonemic fluency, sequencing (Trail Making test A), and reaction time were lower than controls' scores. Even after controlling for age, CC area was correlated with performance in intelligence, visuospatial processing, and two frontal measures (Trail Making Test A and B). Two other frontal measures (Wisconsin Card Sorting Task categories and perseverative errors), verbal memory, and visual reaction time differences were only statistically significant before controlling for age. Ventricular volume was less associated with neurocognitive functioning; significant differences at the p < 0.05 level were found only in verbal memory, Wisconsin Card Sorting Task (WCST) categories (an executive functions test), and visual reaction time (motor and mental slowness) without controlling for age. After controlling for age, only WCST categories scores were significantly different between groups. Consistent with past studies, findings indicated poor prognosis for diffuse lesions in the developing brain (Verger et al., 2001).
In a cross-sectional and longitudinal study of relationships between CC area, interhemispheric functioning (computerized Bimanual Coordination Test, or cBCT), and CC lesions in moderate-to-severe pediatric CHI, Hilleary (2010) found CC deterioration over time in the patients with macroscopic CC lesions as detected by midsagittal MRI. Longitudinally, decreased splenium size was related to decreased accuracy on the cBCT trials that required asymmetric movement and lacked visual control. The findings in this study were consistent with past findings that CC lesions are found in severe DAI (Adams et al., 1989; Gennarelli, et al., 1982), the CC shows longitudinal deterioration in severe injuries involving CC lesions (Levin et al., 2000), and both total and regional CC area is smaller in more severe pediatric TBI, resulting in interrupted development (Benevidez et al., 1999; Levin et al., 2000).
CC Morphology and Measurement
Research has firmly established that the central role of the CC is to mediate interhemispheric communication. However, histological subdivisions of the CC vary. As previously described, the CC is organized topographically in its connections with the corresponding brain regions surrounding it. The most commonly utilized topology to study the CC has been that of Witelson (1989) who divided it into five vertical regions with seven subregions on the midsaggital plane. In Witelson's scheme, the rostrum, genu, and rostral body subregions comprise the anterior third (region I) of the CC, with fibers connecting to the prefrontal, premotor, and supplementary motor cortices. The anterior midbody (region II) crosses fibers with the motor cortex while the posterior midbody (region III) crosses fibers with somatosensory and posterior parietal areas. The isthmus (region IV) and splenium (region V) comprise the posterior third of the CC, with fibers connecting with the temporal, parietal, and occipital cortices.
The high variability of callosal size in the healthy brain presents another challenging aspect to CC measurement. For example, Johnson et al. (1994) reported an average CC size of 667mm based on their comprehensive literature review of normative adult cases, but Parashos et al. (1995) found a range from 216.4 mm to 621.5 mm in normal callosal area. Pediatric studies are further complicated because of variability in growth rates of the CC and the resultant variability in area.
Manual tracing of CC area from midsagittal MRI represents the established morphological measurement approach in the literature as evidenced in the aforementioned studies (e.g., Johnson et al., 1996; Levin et al., 2000; Luna, 2005; Ross, 2008). Although volume measurement of brain structures is generally more accurate and comprehensive than area measurement, the latter is a more straightforward measure of the CC and is as informative as volume measurement (Ross, 2008). However, whole brain volume (or area, alternatively) should also be taken into consideration as a covariate due to normal variability in callosal size (Dorion, Capron, & Duyme, 2001; Johnson et al., 1996). Other studies have demonstrated the utility of this method in various populations including pediatric TBI (Wu et al., 2010, and Hilleary, 2010), pediatric intellectual functioning (Ganjavi et al., 2011), adult TBI (Johnson et al., 1994; Mathias et al., 2004), dementia (Hallam et al., 2008), and MS (Sanfilipo, Benedict, Zivadinov, & Bakshi, 2004).
EP-IHTT as a Functional Indicator of Callosal Damage
IHTT, which is the speed at which information crosses between hemispheres, was originally measured via a manual reaction time (RT) paradigm by Poffenberger (1912) who had subjects quickly press a button in response to a unilateral visual stimulus. Hs method, now known as the Poffenberger paradigm, included two conditions: uncrossed and crossed. The uncrossed condition was named as such because it involved an ipsilateral response such as the subject responding with his/her right hand to a right visual field stimulus. In this condition, no interhemispheric transmission is required because the response (right) hand is controlled by the contralateral (left) hemisphere to which the stimulus was initially relayed (because visual stimuli are transmitted as neural impulses to the contralateral side of the brain). The crossed condition was so named because it involved a contralateral response to the stimulus, such as a left hand response to a right visual field stimulus, therefore requiring interhemispheric transmission. The left hemisphere (which receives the right visual field stimulus) must relay the stimulus to the right hemisphere (which controls the left hand).
In the Poffenberger paradigm, IHTT is measured by calculating the difference between the RT of the crossed condition and the uncrossed condition. This calculation is known as the crossed vs. uncrossed reaction time difference, or more simply as the crossed-uncrossed difference (CUD). Clark and Zaidel (1989) found that RTs in the uncrossed condition tended to be shorter than RTs in the crossed condition, presumably because of the additional time required for the signal to cross via the CC.
In addition to the CUD, callosal transfer time can be reliably estimated by evoked potentials (EPs) or event-related potentials (ERPs) via EEG, which has been found to be a more valid and direct estimate of IHTT than RT measures (Rugg, Milner, & Lines, 1985; Saron & Davidson, 1989; Brown, Bjerke, & Galbraith, 1998). The RT method is limited in that it takes into account only about ten percent of callosal fibers, cannot measure responses from both hemispheres simultaneously, and requires a manual response (Saron & Davidson, 1989). EPs have none of these limitations and more directly measure neural potentials. EPs can estimate interhemispheric transfer time (IHTT) to somatosensory, auditory, or visual stimuli.
EP-IHTT has also been investigated in a variety of conditions such as multiple sclerosis (MS), schizophrenia, and dyslexia. For example, Polich, Romine, Sipe, Aung, and Dalessio (1992) and Burnison (1995) found that adult participants with MS demonstrated significantly slower IHTT compared to controls. Endrass, Mohr, and Rockstroh (2002) and Barnett, Corballis, and Kirk (2005) found that controls demonstrated a faster right-to-left hemispheric transfer that was absent in schizophrenics, though Whitford, Kubicki, Ghorashi, Schneiderman, Hawley et al. (2011) found no such differences. Markee, Brown, Moore, Theberge, and Zvi (1996) found that IHTT was faster in normal participants than in dyslexic participants. IHTT has also been found to be significantly correlated with brain metabolites of the posterior CC in a pediatric TBI population (Babikian, et al., 2010) and with diffusion properties (isotropic and prolate diffusion elipsoids) in the adult population of the Whitford et al. (2011) study.
In visual EP paradigms, IHTT is indexed by presenting unilateral visual stimuli (i.e., isolated to one visual field, either the right or the left) and then recording the resulting brain activity from both hemispheres via bilateral electrodes placed on the scalp. The unilateral stimuli produce early visual EP components (P1 and N1) in the contralateral cerebral hemisphere (because of the lateralization of visual processing) followed by corresponding components in the other hemisphere (ipsilateral to the stimuli) via the corpus callosum. Simply put, IHTT is the speed at which information crosses between hemispheres. This speed is operationalized as the difference in latency of the contralateral (direct pathway) versus ipsilateral (callosal pathway) P1 or N1 components. The latency difference (IHTT) is typically between 10 and 20 milliseconds (msec) when recorded from parietal sites.
Studies have confirmed that IHTT is a reliable measure of callosal conduction. For example, Brown, Jeeves, Dietrich, and Burnison (1999) used the EP method to examine the interhemispheric integration of visual information in a sample that included adult subjects with complete agenesis of the corpus callosum (ACC), partial agenesis, total commissurotomy, and normals. They found that the commissurotomy and acallosal patients were unable to produce ipsilateral visual EPs. The investigators concluded that P1 and N1 interhemispheric transmission is mediated exclusively by the CC, particularly the posterior callosum (see also Brown, Larson, & Jeeves, 1994; Theberge, 1997; Westerhausen et al., 2006). Similar studies with ACC and commissurotomy subjects (Bayard, Gosselin, Robert, & Lassonde, 2004; Brown, Bjerke, & Galbraith, 1998; Brown & Jeeves, 1993; Kutas, Hillyard, Volpe, & Gazzaniga, 1990; Mangun, Luck, Gazzaniga, & Hillyard, 1991; McWain, 1993; Rugg, Milner, & Lines, 1985) have demonstrated this finding, i.e., that without a corpus callosum, interhemispheric transmission of visual EPs is not possible. However, Barr, Hamm, Kirk, and Corballis (2005) conducted a RT study using a high-density 128-channel system and found evidence of ipsilateral activation in all three of their participants with callosal agenesis. The investigators speculated that the transfer detected in their study must have been projected from subcortical-to-cortical pathways or via some other unknown sensory pathway, and that the reason previous studies failed to detect ipsilateral activation was because they used too few electrodes.
Very few studies exist of interhemispheric functioning in pediatric populations, but IHTT does appear to be related to age and typical brain development. Salamy (1978) found a significant negative correlation between EP-IHTT and age in a pediatric population (ages 3.5 to 20) and concluded that IHTT becomes faster as the CC matures. Likewise, in a study of healthy kids, Hagelthorn, Brown, Amano, and Asarnow (2000) found a statistical trend for EP-IHTT to decrease with age. Like Salamy, they concluded that this faster interhemispheric transfer was likely the result of callosal maturation such as increased myelination during late childhood. Marion, Babikian, Newman, Brown, Giza, & Asarnow (2008) investigated the relationship between IHTT (as measured by a visual EP protocol utilizing a computerized pattern-matching task) and FA (as measured by DTI) in a sample of pediatric TBI participants at post-acute and chronic stages as well as longitudinally. Their findings indicated that callosal functioning (e.g., longer IHTT) is compromised in moderate-to-severe brain injury, but with some degree of recovery over time. The post-acute group demonstrated significantly longer IHTT compared to the chronic (one year post-injury) group, and post-acute IHTT was significantly negatively correlated with post-acute FA. These findings suggest that lower FA is related to slower interhemispheric transfer in pediatric TBI.
Although both P1 and N1 latencies have been used for measuring IHTT, Brown and Jeeves (1993) found that IHTT was correlated with direct visual path P1latencies. Thus, interhemispheric latency differences were confounded by P1speed and efficiency, thereby making it an indeterminable measure of IHTT. However, no such correlation was found with N1 latencies. Because the N1 component was not confounded with direct pathway EP speed and efficiency, it could be considered a unique IHTT measure. Larson and Brown (1997) and Theberge (1997) had similar findings and therefore only used N1-IHTT. Other investigators have followed suit in using only the N1 component to index IHTT (Babikian, 2010; Brown, Larson, & Jeeves, 1994; Moes, Brown, & Minnema, 2007).
Interhemispheric Functioning Following CHI. In the first study of interhemispheric dysfunction following TBI in children, Benavidez et al., (1999) examined the relationship between CC atrophy and/or lesions and functional hemispheric disconnection in mild to severe TBI. On a verbal dichotic listening task, investigators found that the severe injury group demonstrated a significant association between increased right ear advantage (more efficient left hemisphere processing of verbal information) and CC lesions and atrophy. Similarly, on a verbal tachistoscopic test, children with the most severe atrophy of total CC as well as midbody and posterior body atrophy demonstrated a greater absolute difference in recognition between right and left visual fields. In contrast, no group differences in asymmetry were observed in either the nonverbal dichotic listening task or nonverbal tachistoscopic tasks. Because the role of the splenium is to interhemispherically transfer visual information, damage to the splenium contributes to asymmetries in tachistoscopic performance, yet few of the children had extensive splenial damage. The investigators concluded that future studies involving more cases of splenium damage would find stronger associations with nonverbal tachistoscopic performance. As noted above, the relationship between brain metabolites and IHTT was found to be significant only in the posterior CC (Babikian et al., 2010).
In a study of electrophysiological and neurocognitive correlates of moderate to severe pediatric TBI, Ellis et al. (2013) found that a subset (15 of 35) of the children with TBI demonstrated normal IHTT (near the average value of control group IHTT). Thus, the sample was divided into three groups accordingly: slow IHTT TBI, fast IHTT TBI, and healthy controls. Significant differences were found between the slow IHTT TBI group and control group in IHTT, processing speed, working memory, and verbal learning.
Source: Essay UK - http://www.essay.uk.com/free-essays/science/traumatic-brain-injury.php