Deficits in Functional Connectivity of Hippocampal and Frontal Lobe Circuits After Traumatic Axonal Injury

ORIGINAL CONTRIBUTION Deficits in Functional Connectivity of Hippocampal and Frontal Lobe Circuits After Traumatic Axonal Injury

Carlos D. Marquez de la Plata, PhD; Juanita Garces, BS; Ehsan Shokri Kojori, MSc; Jack Grinnan, BS; Kamini Krishnan, MS; Rajesh Pidikiti, MS; Jeffrey Spence, PhD; Michael D. Devous Sr, PhD; Carol Moore, MA; Rodderick McColl, PhD; Christopher Madden, MD; Ramon Diaz-Arrastia, MD, PhD

Objective: To examine the functional connectivity of tive outcomes were assessed approximately 6 months af- hippocampal and selected frontal lobe circuits in pa- ter injury. tients with traumatic axonal injury (TAI). Main Outcome Measures: Interhemispheric func- Design: Observational study. tional connectivity, spatial patterns of functional connectivity, and associations of connectivity measures with Setting: An inpatient traumatic brain injury unit. Imaging functional and neurocognitive outcomes. and neurocognitive assessments were conducted in an outpatient research facility. Results: Patients showed significantly lower interhemispheric functional connectivity for the hippocampus and Participants: Twenty-five consecutive patients with brain ACC. Controls demonstrated stronger and more fo- injuries consistent with TAI and acute subcortical white cused functional connectivity for the hippocampi and matter abnormalities were studied as well as 16 healthy ACC, and a more focused recruitment of the default mode volunteers of similar age and sex. network for the dorsolateral prefrontal cortex ROI. The interhemispheric functional connectivity for the hippo- Interventions: Echo-planar and high-resolution T1- campus was correlated with delayed recall of verbal in- weighted images were acquired using 3-T scanners. Re- formation. gions of interest (ROI) were drawn bilaterally for the hippocampus, anterior cingulate cortex (ACC), and Conclusions: Traumatic axonal injury may affect inter- dorsolateral prefrontal cortex and were used to extract hemispheric neural activity, as patients with TAI show time series data. Blood oxygenation level–dependent data disrupted interhemispheric functional connectivity. More from each ROI were used as reference functions for cor- careful investigation of interhemispheric connectivity is relating with all other brain voxels. Interhemispheric func- warranted, as it demonstrated a modest association with tional connectivity was assessed for each participant by outcome in chronic TBI. correlating homologous regions using a Pearson correlation coefficient. Patient functional and neurocogni- Arch Neurol. 2011;68(1):74-84

RAUMATIC BRAIN INJURY ing from TAI,4,5 but more novel neuroim- (TBI) is a major public aging modalities have shown sensitivity to- health problem in modern ward white matter injury.6-9 Author Affiliations: Center for Brain Health, University of societies, with an inci- Neuroimaging studies have found Texas at Dallas, Richardson dence in the United States that integrity of white matter after TAI is (Dr Marquez de la Plata); and estimated between 92 and 250 per 100 000 correlated with injury severity and out- T 10-17 the Departments of Psychiatry persons annually; approximately 50 000 come. The neurocognitive effect of (Dr Marquez de la Plata, Messrs individuals each year are left with long- TAI has been documented by Kraus et Shokri Kojori and Grinnan, and term physical and psychological limita- al,18 who found that reduction in the in- Ms Krishnan), Neurology tions that limit their independence and tegrity of various white matter structures (Dr Diaz-Arrastia, Mss Garces ability to work.1,2 Diffuse axonal injury, was associated with poorer performance and Moore, and Mr Pidikiti), more recently referred to as traumatic axo- on measures of attention, memory, and Clinical Sciences (Dr Spence), nal injury (TAI), is a common subtype of executive function. It is not yet known Radiology (Neuroradiology) (Drs Devous and McColl), and TBI occurring in most motor vehicle col- whether degree of white matter injury Neurological Surgery lisions in which deceleration and rota- (ie, structural integrity) is associated (Dr Madden), University of tional forces cause shearing of the brain’s with impairment in neuronal (ie, func- Texas Southwestern Medical white matter.3 Computed tomography is tional) activity between highly intercon- Center, Dallas. insensitive to white matter lesions result- nected cortical regions.

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 74

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 Functional-connectivity magnetic resonance imaging (Siemens AG, Erlangen, Germany) or a General Electric Signa (MRI) is a technique for analyzing functional MRI data Excite 3T (General Electric Healthcare, Milwaukee, Wiscon- to determine the functional relatedness of selected brain sin) scanner. A time series of 128 echo-planar image volumes regions. It is based on determining brain regions that dem- was acquired at 36 axial slice locations throughout the whole onstrate temporally correlated blood oxygenation level– brain. Participants were asked to direct their attention to 19,20 crosshairs projected onto a screen during image acquisition and dependent (BOLD) signal. This technique demon- not think of anything. Each scanner acquired images from con- strates differential patterns of functional hippocampal trols and patients. The Siemens scanner acquired images from connectivity during the resting state between patients with 7 controls and 9 patients, and the GE scanner acquired images Alzheimer disease and healthy volunteers, as controls from 9 controls and 16 patients. The echo-planar image data showed diffuse cortical and subcortical connectivity and acquired by the Siemens scanner were obtained with single- patients demonstrated reduced connectivity including ab- shot gradient-recalled pulse sequence with time to repetition sence of connectivity with the frontal lobes.21,22 While pa- (TR) of2 seconds; echo time (TE),25 milliseconds; flip angle, tients with Alzheimer disease demonstrate a reduction 90°; matrix, 64 ϫ 64; field of view (FOV), 210 mm; and 3.5- in hippocampal functional connectivity, the association mm slice thickness. High-resolution T1-weighted structural im- between functional connectivity and behavioral mea- ages acquired by the Siemens scanner were acquired using mag- netization prepared rapid acquisition gradient echo with slice sures is not well understood, as the aforementioned stud- thickness of 1.0 mm; FOV, 240 mm; TE, 4 milliseconds; in- ies examined patients documented to have poorer per- version time, 900 milliseconds; TR, 2250 milliseconds; flip angle, formance on tasks of memory ability than controls but 9°; and number of excitations, 1. Echo-planar image data ac- did not correlate their memory performance to connec- quired by the GE scanner were obtained with single-shot gra- tivity measures. dient-recalled pulse sequence with TR,2 seconds; TE,25 mil- The goal of the present study is to examine whether hip- liseconds; flip angle, 90°; matrix, 64 ϫ 64; field of view, 210 pocampal and frontal lobe circuits of patients with sus- mm; and slice thickness, 3.5 mm). High-resolution T1- pected TAI differ from those of healthy individuals during weighted structural images acquired by the GE scanner were the resting state. Given that hippocampal and frontal lobe acquired using fast spoiled gradient recall with slice thick- injuries are common after TAI23,24 and subsequent memory ness, 1.3 mm; FOV, 240 to 280 mm; TR, 8 milliseconds; TE, 2.4 milliseconds; flip angle, 25°; and number of excitations, 2. and executive function deficits are frequently re- 25,26 Patients’ neuroimaging data were acquired between 6 and 10 ported, we hypothesize that patients with TAI will dem- months after injury, which coincides with the day their out- onstrate distinct hippocampal and frontal lobe connectiv- come evaluations were conducted. ity patterns and weaker bilateral connectivity than controls. Furthermore, we will examine whether the degree of func- PREPROCESSING OF FUNCTIONAL tional connectivity in these regions correlate with test per- IMAGING DATA formance in their respective neurocognitive domains. The images were first converted from DICOM (digital imaging METHOD and communications in medicine) to an Analysis of Functional NeuroImages (AFNI; National Institutes of Mental Health, PARTICIPANTS Bethesda, Maryland)–readable format. The AFNI software was used for selected preprocessing steps. Slice-time correction was Twenty-five patients with TBI were recruited from Parkland Me- performed to adjust for varying acquisition time for slices. Time morial Hospital, Dallas, Texas. Inclusion criteria required that series data were corrected for motion and linear drift artifacts. patients (1) sustain closed-head traumatic brain injury through The amount of movement observed on a frame-by-frame basis a mechanism consistent with TAI (such as high-speed motor did not exceed 1 voxel in size for any participant in this study. vehicle collision) and (2) were at least 16 years old. Exclusion Given that coherence in BOLD signal fluctuations occurs at low 19,27,28 criteria were (1) preexisting neurologic disorders or history of frequencies, high-frequency components were removed prior TBI; (2) presence of focal lesions (including contusion, extra- to analysis of functional connectivity by setting a low-pass filter axial hematoma, and/or intraparenchymal hemorrhages) with at 0.12 Hz. The signal to noise ratio was then increased by spa- volume greater than 10 mL visible on cranial computed to- tially smoothing the data with a 3-dimensional Gaussian taper- mography; (3) conditions that may result in abnormal MRI find- ing function (5 mm full-width at half maximum). Functional im- ings and compromise cognitive functions (ie, prior brain tu- ages were coregistered to high-resolution T1 images for each mor, epilepsy, multiple sclerosis, encephalitis/meningitis, participant, and masks of regions containing cerebrospinal fluid 29 Parkinson disease, Alzheimer disease/mild cognitive impair- was created using FSL (FMRIB Software Library) 4.1.4. These ment, human immunodeficiency virus encephalopathy, vas- masks were used to obtain averaged time series data from re- cular malformation, and psychiatric disease); or (4) being a pris- gions unlikely to contribute variance of neuronal origin. The time oner, homeless patient, or pregnant woman. All patients series from these regions were later regressed out from func- demonstrated subcortical white matter lesions visible on T2 fluid- tional data of interest. attenuated inversion recovery MRI. Sixteen healthy volunteers of similar age and sex were recruited as controls. All con- SEED REGIONS OF INTEREST trols had good general health and no known neurocognitive disorders. Informed consent was obtained from all partici- Six anatomical regions of interest (ROI) corresponding to the pants or their legally authorized representative. hippocampus, anterior cingulate, and dorsolateral prefrontal cortex were hand drawn bilaterally using AFNI’s graphical user IMAGE ACQUISITION AND PROCESSING interface by trained research assistants who used a human brain atlas for reference.30 These ROIs were used as seed volumes to Functional and anatomical magnetic resonance images were ob- extract average time series data in subjects’ native space tained for each participant using either a Siemens Trio 3T (Figure 1). Owing to low spatial resolution involved in func-

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 75

Figure 1. Location of manually drawn reference seed regions of interest: hippocampus (A), anterior cingulate cortex (B), and left dorsolateral prefrontal cortex (C) (radiologic convention).

tional MRI, seeding the anterior cingulate unilaterally may in- STATISTICAL ANALYSIS clude signal from both left and right hemispheres and ulti- mately make interpreting interhemispheric connectivity difficult. The anterior cingulate cortex seed ROI was drawn by exclud- Demography ing slices on either side of the midline. Between-group differences in age and interhemispheric functional connectivity were examined using an independent- OUTCOME MEASURES samples t test, as these data were suitable for analyzing with parametric tests. Group differences for sex were examined using Functional Outcomes a ␹2 test for independence.

Functional and neurocognitive outcomes were assessed the same Functional Connectivity Analyses day the neuroimaging scans were obtained (ie, at least 6 months after injury). Functional outcome was assessed using the Glasgow Outcome Scale–Extended (GOSE).31 TheGOSEisa For each ROI, the averaged time courses of BOLD signal were commonly used structured interview that assesses functional used as the seed reference time series for calculating correlation with all other brain voxels’ time series. A false discovery abilities in multiple domains following a head injury. Total GOSE Ͻ scores range from 1 to 8, with higher scores associated with rate–corrected P .05 was considered statistically significant better outcome. to reduce the occurrence of multiple comparison-related false- positive results. The result is a spatial map of correlation coefficients for every voxel in the brain representing an individu- Neurocognitive Outcome al’s pattern of functional connectivity with the seed region. Fisher z transformation was applied to the individual correlation maps Information processing speed, learning and memory, and ex- to adjust the variance of correlation coefficients for subse- 32-36 ecutive function deficits are common after TBI. Neurocog- quent group-level comparisons. The results were then trans- nitive outcome assessments were conducted by a research co- formed into Talairach space, and separate group maps were gen- ordinator who completed standardized training, was supervised erated for patients and controls. A significant cluster of by a neuropsychologist, and was blinded to imaging results. De- correlation was defined as a group of at least 200 adjacent vox- mographically adjusted scores for neurocognitive measures were els. Spatial statistical analyses were conducted using AFNI. used when applicable. Processing speed was assessed using the Trail Making Test A,37 and the digit symbol and symbol search subtests of the Interhemispheric Connectivity Analysis Wechsler Adult Intelligence Scale–Third Edition.38 Memory functioning is commonly affected after TBI and was Interhemispheric functional connectivity examined synchro- also assessed approximately 6 months after injury using the Cali- nicity of BOLD signal fluctuations of bilateral regions over time. fornia Verbal Learning Test–II.39 Total items learned across 5 This analysis used the BOLD data extracted from the ROIs from trials were used to measure learning, while short and long delay- each participant. Averaged left and right BOLD fluctuations from free recall trials were used to measure memory. each ROI across 124 time points (excluded first 4 frames) were Various neurocognitive tests were used to evaluate execu- tested for significant associations using a Pearson correlation tive functions, which are largely influenced by the frontal lobes. coefficient. The Fisher z transformation was applied to the cor- Trail Making Test B37 was used to measure patients’ ability to relation coefficients to allow for group comparisons between shift mental sets efficiently. The Dodrill-Stroop Color- patients and controls. Between-group t tests were used to de- Naming condition40 was used to measure ability to selectively tect significant differences in degree of interhemispheric con- attend to meaningful information while inhibiting a prepotent nectedness between the respective groups for each ROI. Am- response. The Controlled Word Association Test41 was used to plitude of BOLD fluctuation was inspected and not determined assess verbal generativity. to be statistically different between groups.

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 76

Controls Patients With TAI (n=16) (n=25)

Characteristic Mean (SD) Median Mean (SD) Median Age, y 37.06 (14.16) 36.00 30.28 (13.80) 24.00 Education, yb 15.38 (2.72) 16.00 12.63 (3.05) 12.00 Male sex, % 63 80 Handedness right, % 100 96 White race, % 81 76 GCS 8.04 (5.01) 8.00 Days in ICU 8.28 (7.68) 4.00 Days in hospital 13.28 (10.09) 12.00 Time from injury to scan, mo 7.13 (1.12) 7.00 GOSE 5.97 (1.97) 7.00 Digit Symbol T score 39.00 (11.01) 40.00 Symbol Search T score 43.78 (13.27) 46.67 COWAT T score 35.47 (13.36) 36.00 TMT A T score 42.87 (19.40) 43.50 TMT B T score 44.53 (20.12) 48.50 Stroop color-naming T score 44.97 (19.72) 53.50 CVLT total learning T score 40.42 (20.29) 46.00 CVLT short delay T score 38.65 (19.16) 45.00

Abbreviations: COWAT, Controlled Oral Word Association Test; CVLT-II, California Verbal Learning Test–Second Edition; GCS, Glasgow Coma Scale; GOSE, Glasgow Outcome Scale Extended; ICU, intensive care unit; TAI, traumatic axonal injury; TMT A, Trail Making Test A, TMT B, Trail Making Test B. a No significant differences between groups in age, sex, handedness, or ethnicity. b Denotes a significant difference between group difference in education, PϽ.05.

Outcomes Analysis The results of 2-sample independent t tests showed that all 3 interhemispheric connectivity measures were simi- Spearman correlations were used to test the association be- lar across scanners (hippocampi, P=.06; DLFPC, P=.82; tween connectivity measures and functional outcome, as the ACC, P=.13), suggesting that the data from the 2 scan- GOSE is an ordinal measure. Pearson correlations were used ners are comparable. to examine associations between connectivity measures and neurocognitive outcome. Statistical analyses were performed using Statistical Package for Social Sciences (SPSS v11.0; SPSS Inc, HIPPOCAMPAL CONNECTIVITY Chicago, Illinois). Connectivity Between Hippocampi RESULTS Figure 2 illustrates the fluctuation of BOLD signal in DEMOGRAPHY the bilateral hippocampi for a representative control relative to a representative patient with TAI. The degree of As expected, no differences were found between pa- interhemispheric hippocampal connectivity was signifi- tients and controls in age (mean [SD], 30 [14] and 37 cantly greater for controls than for patients with TAI [14] years, respectively) or sex (80% and 63% male, re- (P=.04) (Table 2). The degree of interhemispheric hip- spectively). Patients had traumatic brain injuries rang- pocampal connectivity among patients was negatively cor- ing in severity from complicated mild to severe, as the related with delayed recall of verbal information mean (SD) Glasgow Coma Scale score was 8 (5). Pa- (Table 3). A scatterplot of this association is displayed tients’ average GOSE scores were in the upper moderate in Figure 3. recovery range (Table 1). Additionally, their neurocognitive outcomes ranged from mildly impaired to low av- Whole-Brain Connectivity erage impairment, with the lowest scores on the Digit Sym- bol subtest, Controlled Word Association Test, Stroop Generally, the spatial distribution of hippocampal connec- word reading condition, and California Verbal Learning tivity among healthy controls showed a strong, focused sig- Test-II short delay recall. nal bilaterally within the body of the hippocampi (Figure 4A). Controls also demonstrated connectivity in INTERSCANNER VARIABILITY the septal and subthalamic nuclei. While the hippocampus appeared to have connectivity with bilateral parahip- To investigate whether there is significant variability be- pocampal gyri, most of the correlated signal among con- tween the two scanners used, we examined the inter- trols occurred in the anterior medial temporal lobes. In hemispheric connectivity measures among controls in a contrast, patients demonstrated more abundant connec- between-scanner fashion. Nine controls were scanned tivity with the parahippocampal gyrus and posterior cin- using the GE magnet and 7 using the Siemens magnet. gulate, as well as more diffuse connectivity in the tempo-

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 77

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 brains, as the average strength of bilateral anterior cin- 1.5 r = 0.850 gulate interconnectivity was greater for controls than for patients with TAI (P=.02) (Table 2). The degree of in- 1.0 terhemispheric anterior cingulate connectivity among patients was not significantly associated with outcome but 0.5 showed a trend toward significance with the GOSE (P=.10) (Table 3). 0.0 Whole-Brain Connectivity

–0.5 BOLD Signal Change, % Anterior cingulate connectivity was similar for both left and right seeds. The pattern of anterior cingulate cortex –1.0 connectivity for controls is displayed in Figure 5A. Among controls, synchronous areas include focused sig- –1.5 nal in the anterior cingulate bilaterally, bilateral ventral posterior cingulate cortex, and bilateral caudate. Ante- 1.5 rior cingulate connectivity for patients showed diffuse r = 0.355 correlations surrounding the anterior cingulate bilater- 1.0 ally, bilateral caudate and thalamus, bilateral dorsal posterior cingulate cortex, and cingulate cortex connecting 0.5 the anterior and posterior cingulate cortices (Figure 5B). Patients also demonstrated negatively correlated signal

0.0 in the occipital-temporal gyrus.

–0.5 DORSOLATERAL PREFRONTAL BOLD Signal Change, % CORTEX CONNECTIVITY

–1.0 Connectivity Between Bilateral Dorsolateral Prefrontal Cortex –1.5 0 50 100 150 200 250 Time, s There was no significant difference in connectivity for bilateral dorsolateral prefrontal cortex (DLPFC) be- Figure 2. Bilateral hippocampal blood oxygen level–dependent (BOLD) tween controls and patients with TAI (P=.35) (Table 2). fluctuations over time. Additionally, the degree of bilateral DLPFC connectivity among patients was not significantly associated with outcome but demonstrated a trend toward significance Table 2. Connectivity Between Bilateral Regions of Interest for 2 outcome measures (P=.10) (Table 3).

Mean (SD) Whole-Brain Connectivity Patients Region of Interest Controls With TAI P Value Left DLPFC connectivity for healthy individuals Hippocampus 0.78 (0.11) 0.59 (0.26)a .04 (Figure 6A) includes the ipsilateral inferior frontal gy- Anterior cingulate cortex 0.85 (0.11) 0.73 (0.19)a .02 rus, middle temporal gyrus, anterior cingulate, poste- Dorsolateral prefrontal 0.49 (0.21) 0.54 (0.25) .35 rior cingulate, bilateral angular gyrus, dorsal aspect of cortex superior and middle frontal gyri, and contralateral precuneus. The pattern of left DLPFC is similar between Abbreviation: TAI, traumatic axonal injury. a PϽ.05. healthy volunteers and patients, but patients demonstrate stronger correlations in bilateral angular gyri, and more diffuse correlations in contralateral frontal and tem- ral and frontal lobes, basal forebrain, and subthalamic nuclei poral lobes, occipito-temporal region, and parahippo- than controls. Furthermore, in areas of bilateral hippocampal gyri. Additionally, while negative correlations are campal connectivity, patients demonstrated weaker con- found in the contralateral precuneus in patients and contralateral connectivity than controls (Figure 4B). Healthy trols, patients also demonstrate a relatively greater num- right hippocampal connectivity showed a similar pattern ber of negatively correlated voxels in the ipsilateral pre- of connectivity as the left hippocampus. cuneus (Figure 6B). The pattern of right DLPFC is similar to that of the left DLPFC within patients and controls. ANTERIOR CINGULATE CONNECTIVITY

Connectivity Between Bilateral Anterior Cingulate COMMENT

Interhemispheric connectivity for the anterior cingulate The synchronicity of BOLD signal fluctuations through- was significantly different between healthy and injured out the brain has been useful in understanding function-

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 78

Interhemispheric Connectivity

Outcome Measure Hippocampus ACC DLPFC Functional Outcome GOSE r −0.16 0.35 0.35 P value .45 .09 .09 Processing Speed Digit Symbol T score r −0.27 0.07 −0.06 P value .22 .73 .77 Symbol Search T score r 0.02 −0.10 0.04 P value .93 .66 .87 TMT A T score r 0.06 0.26 0.36 P value .80 .22 .09 Executive Function TMT B T score r −0.09 0.06 0.30 P value .69 .78 .15 COWAT T score r −0.28 −0.35 0.04 P value .20 .10 .85 Stroop color naming T score r 0.10 0.02 0.16 P value .66 .92 .46 Learning and Memory CVLT-II total learning T score r −0.34 −0.25 −0.10 P value .13 .26 .67 CVLT-II short delay recall T score r −0.50b −0.04 −0.04 P value .02 .86 .86

Abbreviations: ACC, anterior cingulate cortex; COWAT, Controlled Oral Word Association Test; CVLT-II, California Verbal Learning Test–Second Edition; DLPFC, dorsolateral prefrontal cortex; GOSE, Glasgow Outcome Scale Extended; TMT A, Trail Making Test A, TMT B, Trail Making Test B. a Pearson correlations between measures of interhemispheric functional connectivity and neurocognitive outcome. A Spearman correlation was used for the GOSE, as it is an ordinal measure. b PϽ.05.

ally related brain regions/networks.42-44 Functional con- 70.00 nectivity patterns in healthy individuals demonstrate a r = – 0.50 functional link between various regions known to com- P = .02 municate during various tasks and at rest. In contrast, 60.00 connectivity patterns among clinical populations devi- ate considerably from those observed in healthy brains. 50.00 For example, patients with Alzheimer disease have demonstrated disrupted hippocampal and frontal lobe con- 40.00 nectivity throughout the brain compared with healthy 21,22 30.00 peers. Furthermore, significant functional connec- Short Delay Recall CVLT-II tivity between hemispheres is found in healthy individu- 20.00 als,22,45-47 whereas interhemispheric functional connectivity is significantly reduced in clinical populations with 0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 Bilateral Hippocampal Connectivity compromise in the corpus callosum (CC).48,49 The relationship between CC integrity and functional connec- 48 Figure 3. Association between interhemispheric hippocampal connectivity tivity was demonstrated by Quigley and colleagues, who and verbal memory. CVLT-II indicates California Verbal Learning found that patients with agenesis of the CC showed sig- Test–Second Edition. nificantly reduced interhemispheric connectivity compared with controls. Likewise, Johnston et al49 demonstrated dramatic reductions in interhemispheric functional was relatively preserved. These results implicate the CC connectivity of various functional systems after a com- as having a significant role in the degree of interhemi- plete callosotomy, while intrahemispheric connectivity spheric functional connectivity observed using func-

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 79

H H BF FL BF FL HT HT

TL TL SN SN

1.00

STN STN PG PG

– 1.00 PC PC

Figure 4. Average left hippocampus connectivity in controls (A) and patients with traumatic axonal injury (B). BF indicates basal forebrain; FL, frontal lobe; H, hippocampi; PC, posterior cingulate; PG, parahippocampal gyrus; SN, septal nuclei; STN, subthalamic nuclei; and TL, temporal lobe.

A B

RC ACC RC ACC CC PC CC PC

1.00 T T

LC – 1.00 LC Figure 5. Average left anterior cingulate connectivity in controls (A) and patients with traumatic axonal injury (B). ACC indicates anterior cingulate cortex; CC, cingulate cortex; LC, left caudate; PC, posterior cingulate; RC, right caudate; and T, thalamus.

tional MRI and suggest investigation of the influence of MacDonald et al52 that demonstrated compromised hip- corpus callosum damage in functional connectivity in pocampal connectivity in a patient with a TBI. To our other clinical populations with axonal damage. knowledge, this investigation is the first to examine func- Given that the CC is the most commonly injured white tional connectivity differences between a group of pa- matter structure in traumatic axonal injury,4,23,50,51 and tients with TAI and healthy controls. that compromise to the integrity of the CC results in re- The general pattern of hippocampal functional con- duced interhemispheric functional connectivity in other nectivity in controls included stronger bilateral hippo- clinical populations, patients with TAI should demon- campal and greater connectivity in the septal nuclei near strate reduced functional connectivity as well. Indeed, the anterior commissure than the septal nuclei com- the interhemispheric connectivity results in this study pared with patients. In contrast, the pattern of hippo- are commensurate with findings in other clinical popu- campal connectivity in patients with TAI demonstrated lations, as patients with TAI demonstrated significantly reduced contralateral strength of correlation, but gener- reduced interhemispheric functional connectivity in the ally preserved correlation ipsilaterally, and greater cor- bilateral hippocampi and anterior cingulate relative to con- relation in the subthalamic nuclei near the posterior com- trols. The results are consistent with a those of a case study missure, parahippocampal gyrus, and posterior cingulate

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 80

PG PG OG OG C C

ACC ACC 1.00

PCC PCC SFG SFG AG AG

P – 1.00

Figure 6. Average left dorsolateral prefrontal cortex connectivity controls (A) and patients with traumatic axonal injury (B). ACC indicates anterior cingulate cortex; AG, angular gyrus; C, caudate; IFG, inferior frontal gyrus; MTG, middle temporal gyrus; OG, occipito-temporal gyrus; P, precuneus; PCC, posterior cingulate cortex; PG, parahippocampal gyrus; and SFG, superior frontal gyrus.

cortex than controls. Although contralateral hippocam- inefficiency and may represent the brain’s attempt to re- pal connectivity was significantly reduced in patients, it store interhemispheric hippocampal connectivity by using was not absent. This is consistent with a prior study of less direct connections in light of injured subcortical white functional connectivity before and after a complete sur- matter. gical corpus callosotomy.49 Their observation of limited The functional connectivity pattern of the anterior cin- interhemispheric connectivity despite the complete tran- gulate cortex demonstrated interesting spatial differ- section of the main commissural fiber suggests that in- ences between groups, as healthy brains showed greater terhemispheric connectivity also occurs through other correlation with the ventral posterior cingulate cortex commissural fibers such as the anterior and/or the pos- compared with injured brains, and injured brains had terior commissures. Patients in our study undoubtedly greater connectivity with most of the cingulate cortex, underwent varying degrees of subcortical white matter particularly the dorsal aspect of the posterior cingulate injury including injury to the CC, as evidenced by fluid- cortex, than healthy brains. An altered connectivity pat- attenuated inversion recovery MRI; therefore, they pre- tern within the cingulate cortex has been found in clini- sumably have varying degrees of healthy callosal axons cal populations. Castellanos et al54 described compro- allowing some (albeit reduced) functional connectivity mised connectivity between the precuneus and posterior with the contralateral hemisphere through this most par- cingulate and areas of the default mode network includ- simonious route. However, it is also possible that pa- ing the anterior cingulate. Furthermore, Wang et al22 tients in this study demonstrate some interhemispheric found that patients with early-stage Alzheimer disease connectivity via the use of the anterior or posterior com- showed compromised resting state connectivity be- missures or the dorsal hippocampal commissure in lieu tween the posterior cingulate and the hippocampus, areas of the CC. Studies of both animals and humans have sug- involved in the default mode network. Taken together, gested that these smaller commissures play a role in in- these studies and the results from the present investiga- terhemispheric hippocampal connectivity owing to their tion suggest that the connectivity between anterior and proximity to the hippocampus.49,53,54 Our data show that posterior cingulate may be sensitive to compromise in interhemispheric connectivity occurs near both the an- clinical populations with functional or neurocognitive terior and posterior commissures in healthy brains. In deficits. Furthermore, the results of this study support contrast, interhemispheric connectivity in injured brains examining the functional connectivity between various occurs more posteriorly and may rely on posterior com- brain regions involved in the default mode network, as missural fibers (ie, posterior commissure and/or dorsal their connectedness may be a marker of cerebral integ- hippocampal commissure) to a greater degree than healthy rity or compromise. brains. Additionally, patients demonstrated more dif- Interestingly, functional connectivity of the DLPFC for fuse but weaker correlations throughout the medial tem- patients and controls demonstrate a similar pattern ob- poral and frontal lobes and posterior cingulate cortex com- served in the default network (ie, medial superior frontal pared with controls. This appears to be evidence of neural lobe, posterior cingulate and bilateral inferior parietal lobes,

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 81

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 parahippocampal gyrus). Furthermore, patients demon- with heterogeneous injury profiles including mechanism strate significantly greater negative connectivity in occipital- of injury, injury severity, and most importantly, location temporal and parahippocampal gyri than controls, which of brain lesions. Although every patient in this sample was may suggest that patients are suppressing brain activity to selected based on a head injury consistent with traumatic a greater degree than controls. While it is unclear whether axonal injury, the degree of injury to particular white mat- the DLPFC is suppressing brain activity in the aforemen- ter structures undoubtedly varied widely. For example, Ben- tioned regions or if these regions are suppressing brain ac- son et al10 examined the integrity of whole-brain white mat- tivity in the DLPFC and other regions involved in the de- ter of 20 patients with TAI using a histographic analysis fault network, the fact that the frontal lobes play a role in and demonstrated that the distribution of white matter frac- modulating and coordinating complex behaviors sug- tional anisotropy (ie, measure of the directionality of wa- gests that the DLPFC is more likely modulating activity in ter diffusion along axons) for individual patients was sig- other regions. A greater amount of negative correlations nificantly more variable than the distribution for controls. observed when using the DLPFC as a seed among pa- Variability of white matter integrity in various interhemi- tients may suggest that they are less efficient in quieting spheric structures may, in part, explain how interhemi- their minds during rest, demonstrating that the default mode spheric connectivity can be reduced without reducing cog- network is sensitive to changes after TAI. Subsequent in- nitive or functional ability, as it is possible that certain vestigation into this matter should use a time-lag analysis patients had damage to the CC and subsequently re- of connectivity, as this may help determine whether there routed neuronal signal between hippocampi through less is a causal relationship between the DLPFC and the nega- direct but more intact commissural fibers (ie, posterior cin- tively correlated brain regions. gulate, hippocampal commissure), thereby lowering their Relatively few studies have examined the association degree of interhemispheric hippocampal connectivity but between measures of functional connectivity and cogni- maintaining enough contralateral connectivity to approxi- tive ability. Their findings generally suggest that func- mate the desired behavior. tional connectivity in various brain regions have a sig- A limitation of our study is that the degree of compro- nificant relationship with certain cognitive abilities.55,56 mise to interhemispheric white matter is not described. Dec- In this study, the degree of functional connectedness be- rement in interhemispheric structural connectivity may ac- tween hippocampi is negatively associated with perfor- count for degree of functional connectivity in patient mance on an auditory verbal memory task, such that pa- populations, but DTI studies must be performed to mea- tients with less bilateral connectivity recalled more words sure the degree of structural compromise. It is highly rec- after a delay than patients with greater interhemi- ommended that diffusion tensor tractography (or similar spheric connectivity. Given the role of the hippocam- analysis of diffusion tensor imaging data) be incorporated pus in learning and memory, we hypothesized that in- into the research design to more directly examine the as- terhemispheric hippocampal connectivity is associated sociation between the integrity of certain white matter struc- with performance in this cognitive domain. Interest- tures, functional connectivity, and outcome. Another limi- ingly, the observed negative association between hippo- tation of the current study, and any neuroimaging study campal connectivity and memory suggests that cogni- of TBI, has to do with heterogeneity inherent in TBI, as in- tive functioning is not completely dependent on the jury profiles including mechanism of injury, injury sever- integrity of structural connections. Additionally, these ity, and most importantly, location of brain lesions can vary results hint at neural plasticity, as the data suggest that from patient to patient. Therefore, the results of this in- hippocampal signal is rerouted to reach its homologous vestigation may not be generalizable across traumatic brain contralateral region through more indirect posterior con- injuries with different profiles. nections (ie, posterior commissure/thalamic nuclei) af- Given that the functional connectivity measures for this ter TAI, and the degree of inefficiency (ie, diffuse con- study are obtained from a resting-state fMRI paradigm, it nectivity) seen in the pattern of hippocampal connectivity is not known whether patterns of connectivity during rest in patients may be evidence of this plasticity. However, should correlate to functional or neurocognitive tasks. Con- while this plasticity eventually results in neuronal sig- sequently, it is possible that the association between func- nal reaching its contralateral destination and a relative tional connectivity measures and outcome could be sig- increase in interhemispheric hippocampal connectiv- nificantly different had synchronicity of BOLD signal ity, it occurs at the expense of memory ability. between regions been measured during a cognitive task While the interhemispheric connectivity for the ante- rather than during rest. Future studies may benefit from rior cingulate cortex and the DLPFC was significantly lower incorporating both resting state and task-related func- in patients than controls, the correlation between degree tional connectivity measures in their design. It is also im- of connectivity in these regions and outcome only trended portant to note that the results of this study are specific to toward significance with outcome. Though the associa- functional connectivity patterns present 6 months after in- tions between interhemispheric functional connectivity and jury, as results may be different in a more acutely or more outcome observed in our study do not fully support the chronically brain-injured sample. hypothesis that frontal lobe functional brain synchronic- The use of MRI scanners from different manufactur- ity is associated with executive functions after TBI, the re- ers in this study may be perceived as a limitation despite sults are not entirely surprising, as resting state interhemi- demonstrating equivalence in interhemispheric connec- spheric connectivity should not be assumed to have a strong tivity in controls between scanners. However, novel neu- association with functional or neurocognitive outcome in roimaging modalities must show robust differences be- this clinical population. Patients with TBI or TAI present tween control and clinical populations across scanners

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 82

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 to be useful as a clinical biomarker, as hospitals/medical 2. Langlois JA, Rutland-Brown W, Thomas KE. Traumatic brain injury in the United centers use scanners from various manufacturers. While States: emergency department visits, hospitalization, and deaths. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Injury Preven- attempts were made to limit the influence of multiple com- tion; 2006. parisons on the results of the statistical analyses (ie, using 3. Adams JH, Doyle D, Ford I, Gennarelli TA, Graham DI, McLellan DR. Diffuse axo- false discovery rate ␣ correction and cluster threshold- nal injury in head injury: definition, diagnosis and grading. Histopathology. 1989; ing), this study may still be affected by false positives ow- 15(1):49-59. ing to the large number of voxel ϫ voxel comparisons. 4. Meythaler JM, Peduzzi JD, Eleftheriou E, Novack TA. Current concepts: diffuse axonal injury-associated traumatic brain injury. Arch Phys Med Rehabil. 2001; This study provides support for the use of functional 82(10):1461-1471. connectivity MRI in clinical populations, including pa- 5. Mittl RL, Grossman RI, Hiehle JF, et al. Prevalence of MR evidence of diffuse tients with compromised anatomical connectivity such as axonal injury in patients with mild head injury and normal head CT findings. AJNR TAI. The results support the hypothesis that the hippo- Am J Neuroradiol. 1994;15(8):1583-1589. 6. Garnett MR, Cadoux-Hudson TA, Styles P. How useful is magnetic resonance campus and frontal lobe circuits of patients with TAI have imaging in predicting severity and outcome in traumatic brain injury? Curr Opin distinct patterns of interconnectedness and less connec- Neurol. 2001;14(6):753-757. tivity with their contralateral homologue compared with 7. Levine B, Fujiwara E, O’Connor C, et al. In vivo characterization of traumatic brain those of healthy individuals. Additionally, the degree of bi- injury neuropathology with structural and functional neuroimaging. J Neurotrauma. lateral connectivity in hippocampal circuits appears to cor- 2006;23(10):1396-1411. 8. Tong KA, Ashwal S, Holshouser BA, et al. Diffuse axonal injury in children: clini- relate with patients’ memory-related outcome after TAI. cal correlation with hemorrhagic lesions. Ann Neurol. 2004;56(1):36-50. 9. Marquez de la Plata C, Ardelean A, Koovakkattu D, et al. Magnetic resonance imaging Accepted for Publication: July 30, 2010. of diffuse axonal injury: quantitative assessment of white matter lesion volume. Correspondence: Ramon Diaz-Arrastia, MD, PhD, De- J Neurotrauma. 2007;24(4):591-598. 10. Benson RR, Meda SA, Vasudevan S, et al. Global white matter analysis of diffu- partment of Neurology, University of Texas Southwest- sion tensor images is predictive of injury severity in traumatic brain injury. ern Medical Center, 5323 Harry Hines Blvd, Dallas, TX J Neurotrauma. 2007;24(3):446-459. 75390-9036 (ramon.diaz-arrastia@utsouthwestern 11. Bazarian JJ, Zhong J, Blyth B, Zhu T, Kavcic V, Peterson D. Diffusion tensor imaging .edu). detects clinically important axonal damage after mild traumatic brain injury: a Author Contributions: Drs Garces and Kojori contrib- pilot study. J Neurotrauma. 2007;24(9):1447-1459. 12. Wang JY, Bakhadirov K, Devous MD Sr, et al. Diffusion tensor tractography of uted equally to this study. Study concept and design: Mar- traumatic diffuse axonal injury. Arch Neurol. 2008;65(5):619-626. quez de la Plata, Devous, and Diaz-Arrastia. Acquisition 13. Arfanakis K, Haughton VM, Carew JD, Rogers BP, Dempsey RJ, Meyerand ME. of data: Marquez de la Plata, Garces, Grinnan, Krishnan, Diffusion tensor MR imaging in diffuse axonal injury. AJNR Am J Neuroradiol. Devous, Moore, and Diaz-Arrastia. Analysis and interpre- 2002;23(5):794-802. 14. Gupta RK, Saksena S, Agarwal A, et al. Diffusion tensor imaging in late post- tation of data: Marquez de la Plata, Garces, Kojori, traumatic epilepsy. Epilepsia. 2005;46(9):1465-1471. Krishnan, Pidikiti, Spence, Devous, McColl, Madden, and 15. Inglese M, Makani S, Johnson G, et al. Diffuse axonal injury in mild traumatic Diaz-Arrastia. Drafting of the manuscript: Marquez de la brain injury: a diffusion tensor imaging study. J Neurosurg. 2005;103(2):298- Plata, Kojori, Grinnan, Krishnan, and Diaz-Arrastia. Criti- 303. cal revision of the manuscript for important intellectual con- 16. Salmond CH, Menon DK, Chatfield DA, et al. Diffusion tensor imaging in chronic head injury survivors: correlations with learning and memory indices. Neuroimage. tent: Marquez de la Plata, Garces, Kojori, Pidikiti, Spence, 2006;29(1):117-124. Devous, Moore, McColl, Madden, and Diaz-Arrastia. Sta- 17. Huisman TA, Schwamm LH, Schaefer PW, et al. Diffusion tensor imaging as po- tistical analysis: Marquez de la Plata, Kojori, Grinnan, tential biomarker of white matter injury in diffuse axonal injury. AJNR Am J Krishnan, Pidikiti, Spence, and Devous. Obtained fund- Neuroradiol. 2004;25(3):370-376. ing: Marquez de la Plata, Devous, and Diaz-Arrastia. 18. Kraus MF, Susmaras T, Caughlin BP, Walker CJ, Sweeney JA, Little DM. White matter integrity and cognition in chronic traumatic brain injury: a diffusion ten- Administrative, technical, and material support: Garces, sor imaging study. Brain. 2007;130(pt 10):2508-2519. Kojori, Devous, Moore, McColl, Madden, and Diaz- 19. Biswal B, Yetkin FZ, Haughton VM, Hyde JS. Functional connectivity in the mo- Arrastia. Study supervision: Marquez de la Plata, Devous, tor cortex of resting human brain using echo-planar MRI. Magn Reson Med. 1995; and Diaz-Arrastia. 34(4):537-541. Financial Disclosure: None reported. 20. Peltier SJ, Noll DC. T(2)(*) dependence of low frequency functional connectivity. Neuroimage. 2002;16(4):985-992. Funding/Support: This study was supported by grants 21. Allen G, Barnard H, McColl R, et al. Reduced hippocampal functional connectiv- K23 NS060827 (Dr Marquez de la Plata) and R01 ity in Alzheimer disease. Arch Neurol. 2007;64(10):1482-1487. HD48179 (Dr Diaz-Arrastia) from the National Insti- 22. Wang L, Zang Y, He Y, et al. Changes in hippocampal connectivity in the early tutes of Health and Department of Education grant H133 stages of Alzheimer’s disease: evidence from resting state fMRI. Neuroimage. A020526. 2006;31(2):496-504. 23. Adams JH, Graham DI, Murray LS, Scott G. Diffuse axonal injury due to non- Additional Contributions: We thank Evelyn Babcock, missile head injury in humans: an analysis of 45 cases. Ann Neurol. 1982;12 PhD, of the University of Texas Southwestern Medical (6):557-563. Center for her assistance with image acquisition. Thanks 24. Gentry LR, Godersky JC, Thompson B. MR imaging of head trauma: review of to Kaundinya Gopinath, PhD, for his image processing the distribution and radiopathologic features of traumatic lesions. AJR Am J Roentgenol. 1988;150(3):663-672. and statistical advice. We are also very thankful to Caryn 25. Wallesch C-W, Curio N, Kutz S, Jost S, Bartels C, Synowitz H. Outcome after Harper, MS, for coordinating and conducting many of mild-to-moderate blunt head injury: effects of focal lesions and diffuse axonal the outcome assessments. injury. Brain Inj. 2001;15(5):401-412. 26. Fork M, Bartels C, Ebert AD, Grubich C, Synowitz H, Wallesch CW. Neuropsy- chological sequelae of diffuse traumatic brain injury. Brain Inj. 2005;19(2): REFERENCES 101-108. 27. Cordes D, Haughton VM, Arfanakis K, et al. Frequencies contributing to func- 1. Thurman DJ, Alverson C, Dunn KA, Guerrero J, Sniezek JE. Traumatic brain in- tional connectivity in the cerebral cortex in “resting-state” data. AJNR Am J jury in the United States: a public health perspective. J Head Trauma Rehabil. Neuroradiol. 2001;22(7):1326-1333. 1999;14(6):602-615. 28. Lowe MJ, Mock BJ, Sorenson JA. Functional connectivity in single and mul-

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 83

©2011 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 tislice echoplanar imaging using resting-state fluctuations. Neuroimage. 1998; tional connectivity: mapping, assessment of stability, and relation to Alzhei- 7(2):119-132. mer’s disease. J Neurosci. 2009;29(6):1860-1873. 29. Smith SM, Jenkinson M, Woolrich MW, et al. Advances in functional and struc- 44. van den Heuvel MP, Mandl RC, Kahn RS, Hulshoff Pol HE. Functionally linked tural MR image analysis and implementation as FSL. Neuroimage. 2004;23 resting-state networks reflect the underlying structural connectivity architec- (suppl 1):S208-S219. ture of the human brain. Hum Brain Mapp. 2009;30(10):3127-3141. 30. Mai JK, Assheuer J, Paxinos G. Atlas of the Human Brain . 2nd ed. Amsterdam, 45. Stein T, Moritz C, Quigley M, Cordes D, Haughton V, Meyerand E. Functional con- The Netherlands: Elsevier Academic Press; 2004. nectivity in the thalamus and hippocampus studied with functional MR imaging. 31. Wilson JT, Pettigrew LE, Teasdale GM. Structured interview for the Glasgow Out- AJNR Am J Neuroradiol. 2000;21(8):1397-1401. come Scale and the Extended Glasgow Outcome Scale. J Neurotrotrauma. 2007; 46. Greicius M. Resting-state functional connectivity in neuropsychiatric disorders. 15:573-585. doi:10.1089/neu.1998.15.573. Curr Opin Neurol. 2008;21(4):424-430. 32. Mathias JL, Wheaton P. Changes in attention and information-processing speed 47. Damoiseaux JS, Beckmann CF, Arigita EJ, et al. Reduced resting-state brain ac- following severe traumatic brain injury: a meta-analytic review. Neuropsychology. tivity in the “default network” in normal aging. Cereb Cortex. 2008;18(8):1856- 2007;21(2):212-223. 1864. 33. Levin HS. Memory deficit after closed head injury. J Clin Exp Neuropsychol. 1990; 48. Quigley M, Cordes D, Turski P, et al. Role of the corpus callosum in functional 12(1):129-153. connectivity. AJNR Am J Neuroradiol. 2003;24(2):208-212. 34. Brooks N, Campsie L, Symington C, Beattie A, McKinlay W. The five year out- 49. Johnston JM, Vaishnavi SN, Smyth MD, et al. Loss of resting interhemispheric come of severe blunt head injury: a relative’s view. J Neurol Neurosurg Psychiatry. functional connectivity after complete section of the corpus callosum. J Neurosci. 1986;49(7):764-770. 2008;28(25):6453-6458. 35. Ylvisaker M, Feeney TJ. Executive functions after traumatic brain injury: supported 50. Ng HK, Mahaliyana RD, Poon WS. The pathological spectrum of diffuse axonal cognition and self-advocacy. Semin Speech Lang. 1996;17(3):217-232. injury in blunt head trauma: assessment with axon and myelin strains. Clin Neu- 36. McDonald BC, Flashman LA, Saykin AJ. Executive dysfunction following trau- rol Neurosurg. 1994;96(1):24-31. matic brain injury: neural substrates and treatment strategies. NeuroRehabilitation. 51. Amaral DG, Insausti R, Cowan WM. The commissural connections of the mon- 2002;17(4):333-344. key hippocampal formation. J Comp Neurol. 1984;224(3):307-336. 37. Reitan RM. Validity of the Trail-making Test as an indicator of organic brain damage. 52. MacDonald CL, Schwarze N, Vaishnavi SN, et al. Verbal memory deficit follow- Percept Mot Skills. 1958;8:271-276. ing traumatic brain injury: assessment using advanced MRI methods. Neurology. 38. Wechsler D. Wechsler Adult Intelligence Scale. 3rd ed. San Antonio, TX: Psy- 2008;71(15):1199-1201. chological Corporation; 1997. 53. Gloor P, Salanova V, Olivier A, Quesney LF. The human dorsal hippocampal com- 39. Delis DC, Kramer JH, Kaplan E, Ober BA. California Verbal Learning Test- missure: an anatomically identifiable and functional pathway. Brain. 1993;116 Second Edition (CVLT-II). San Antonio, TX: Psychological Corporation; 2000. (Pt 5):1249-1273. 40. Dodrill CB. A neuropsychological battery for epilepsy. Epilepsia. 1978;19(6):611- 54. Castellanos FX, Margulies DS, Kelly C, et al. Cingulate-precuneus interactions: a 623. new locus of dysfunction in adult attention-deficit/hyperactivity disorder. Biol 41. Benton AL, Hamsher K. Multilingual Aphasia Examination. Iowa City, IA: AJA As- Psychiatry. 2008;63(3):332-337. sociates; 1983. 55. Bosma I, Douw L, Bartolomei F, et al. Synchronized brain activity and neurocog- 42. Greicius MD, Krasnow B, Reiss AL, Menon V. Functional connectivity in the rest- nitive function in patients with low-grade glioma: a magnetoencephalography study. ing brain: a network analysis of the default mode hypothesis. Proc Natl Acad Sci Neuro Oncol. 2008;10(5):734-744. USA. 2003;100(1):253-258. 56. Song M, Zhou Y, Li J, et al. Brain spontaneous functional connectivity and 43. Buckner RL, Sepulcre J, Talukdar T, et al. Cortical hubs revealed by intrinsic func- intelligence. Neuroimage. 2008;41(3):1168-1176.

(REPRINTED) ARCH NEUROL / VOL 68 (NO. 1), JAN 2011 WWW.ARCHNEUROL.COM 84