Division of Neurosurgery, University of Texas Medical Branch (HSL, HH, KAC, HME), Galveston, Texas, and Departments of Radiology (DM) and Neurosurgery (DB), University of Texas Southwestern Medical School, Dallas, Texas Neurosurgery 31; 1117-1121, 1992 ABSTRACT: REVERSIBLE CEREBRAL ATROPHY in humans has been documented by computed tomography in alcoholics and has been described as an incidental finding after head injury in children. Two children were studied who had sustained a severe closed head injury, 1 and 5 years previously, after which cerebral atrophy had developed, according to subacute computed tomography. Reversible cerebral atrophy was seen on magnetic resonance images of both patients. Despite normal appearance on magnetic resonance images more than 1 year after injury, both patients exhibited residual neuropsychological impairment on a broad range of cognitive and memory tests. KEY WORDS: Central nervous system trauma; Head injury; Magnetic resonance imaging; Plasticity Evidence of neuroplasticity within the human central nervous system after injury includes remyelination (6) , synaptic reorganization (7), and reversible cerebral atrophy (3,10,13). Reports of reversible cerebral atrophy demonstrated by serial computed tomography (CT) in abstinent alcoholics have supported the capacity of the mature brain for recovery (3). Studies utilizing magnetic resonance imaging (MRI), including analysis of relaxation times (13), and pathological findings (10) have confirmed the CT evidence for reversible atrophy in abstinent alcoholics, and raised doubt about rehydration as the underlying mechanism (13) . Apart from morphological changes associated with abstinence in alcoholics, studies of multiple sclerosis (6) , uncal herniation (7), and normal pressure hydrocephalus (12) have supported the potential for recovery of the compromised brain. In a biopsy of a 15-year-old with multiple sclerosis in whom MRI had revealed a resolved lesion, Ghatak et al. (6) found that newly formed myelin sheaths were present in a small proportion of demyelinated axons. Grady et al. (7) studied two cases of uncal herniation in which reorganization of synaptic connections in the dentate gyrus was reflected by intensification of acetylcholinesterase staining (presumed secondary to proliferation of the cholinergic axon terminals).
PATIENTS AND METHODS Patient 1 At the age of 4 years and 10 months, this girl sustained a severe closed head injury that produced immediate loss of consciousness when she was struck by a car in June 1989. The postresuscitation Glasgow Coma Scale (GCS) score was seven (eye opening = 1, motor = 5, verbal = 1), and the patient's consciousness remained impaired for 2 weeks (14). Her pupils reacted to light, and the neurological examination revealed right hemiplegia. Associated injuries included fractures of the right femur and pelvis. A right frontal ventriculostomy was performed to monitor intracranial pressure, which remained below 20 mm Hg. The initial CT scan on the day of injury showed a small amount of subarachnoid blood within the left ambient cistern without evidence of associated intraparenchymal hematoma. The ventricular system was normal, with no shift of the midline structures nor extra-axial fluid collections (Fig. 1A). The ventricular size and sulci were within normal limits for the patient's age, particularly in light of the fact that there was no evidence of raised intracranial pressure. A follow-up CT scan without contrast material performed 20 days later showed a marked increase in the size of the third and lateral ventricles as well as prominence of the sylvian fissures and convexity sulci (Fig. 1B). This dramatic change implies a loss of brain volume, or generalized cerebral atrophy. These changes were not thought to represent the end result of anoxia or diffuse brain swelling, as there was no evidence to suggest an anoxic event, nor was there evidence of raised intracranial pressure. In November 1990, at the age of 6 years (17 months after injury), the patient was examined by
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
AUTHOR(S): Levin, Harvey S., Ph.D; Mendelsohn, Dianne, M.D.; Bruce, Derek, M.D.; Harward, Harriet, M.A.; Culhane, Kathleen A., M.A.; Eisenberg, Howard M., M.D.
Grady et al. inferred that sprouting of connections from entorhinal cortex could potentially support behavioral recovery in nonfatal injuries. The potential for recovery of the brain compromised by normal pressure hydrocephalus was supported by a recent case study (12) documenting reduced ventricular volume and increased cerebral metabolism 1 year after shunt surgery in a 78-year-old demented woman. In a serial CT study of 63 children who developed diffuse cerebral swelling secondary to a severe closed head injury, Bruce et al. (2) found that 22 patients (35%) had reversal of mild cerebral atrophy within 6 to 12 months that was not related to their clinical recovery. However, these investigators did not report neurobehavioral findings, nor was it clear whether reversible atrophy was restricted to children with initial brain swelling. In an ongoing investigation employing MRI and tests of neurobehavioral functioning in head-injured patients, we have discovered evidence of reversible cerebral atrophy in 2 of the 55 patients studied to date. We report on these patients to document the morphological changes in their brains and to raise questions concerning the underlying mechanism of these changes. Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Neurosurgery 1992-98 December 1992, Volume 31, Number 6 1117 Reversibility of Cerebral Atrophy after Head Injury in Children Case Report
NEUROPSYCHOLOGICAL TESTS Frontal lobe functioning was evaluated with tests of fluency, problem solving, and inhibitory control. Verbal fluency (i.e., production of words beginning with a specified letter under timed conditions) (1) and semantic memory (i.e., production of animal names under timed conditions) were assessed, as was design fluency (11) (i.e., creation of nonsense designs under timed conditions). Two computerized tasks were administered, including the Wisconsin Card Sorting Test (9) (a measure of problem-solving ability and response flexibility), and the go-no-go task (5), which measures inhibitory control of responses. Memory, including immediate and delayed recall and recognition, was assessed via the California Verbal Learning Test (4). Four subtests of the Wechsler Intelligence Scale for Children--Revised (WISC-R) (15) (including two verbal and two performance tasks) were also administered. NEUROPSYCHOLOGICAL FINDINGS The neuropsychological findings for both patients and their respective control groups are summarized in Table 1. Comparison of the findings for Patient 1 with a group of age-matched control subjects revealed diminished verbal fluency (i.e., associating words with initial letters) and reduced semantic memory as reflected by animal naming. Patient 1 also created fewer nonsense designs than her control group. Patient 1 used almost twice the number of trials to reach criterion on the go-no-go test as did her control group, although her performance on this response inhibition task was not impulsive. The Wisconsin Test was discontinued because Patient 1 was unable to sort the cards correctly according to even a single category. Intellectual functioning, as estimated from four subtests of the WISC-R, was also impaired, particularly in the areas of abstract verbal reasoning and visuospatial construction abilities. Her prorated Verbal and Performance IQ scores were 66 and 61, respectively. Patient 1's memory was impaired, as reflected by her poor recall of the first list of words across five trials and lower delayed recall as compared with the control group (Fig. 3). The verbal fluency findings for Patient 2 indicate that he invented fewer unique designs under timed conditions relative to his control group, but his verbal fluency was within one standard deviation of that of the control subjects (Table 1). However, Patient 2 exhibited no consistent pattern of perseverations. In contrast to Patient 2's low productivity on fluency tests, his semantic memory (i.e., animal naming) was
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Patient 2 At the age of 1 year and 8 months, this boy sustained a severe closed head injury that produced immediate loss of consciousness when he was hit by a truck in August 1985. His postresuscitation GCS score was 8 (eye opening = 2, motor = 5, verbal = 1). His pupils were equal and reactive, and a left-sided gaze preference was present. He began following commands (albeit inconsistently) at about 1 hour after admission, and became increasingly alert over the ensuing 5 days. However, a focal motor seizure 5 days after injury resulted in right hemiparesis that persisted for 2 weeks. There were no extracranial injuries and no surgical procedures. A transaxial CT scan of the head without contrast material on the day of admission showed a normal ventricular system with preservation of the cisterns and no evidence of any intraparenchymal hematoma, edema, mass effect, or shift of the midline structures (Fig. 2A). A transaxial CT scan without contrast material performed 17 days later showed a marked change, with enlargement of the ventricular system and convexity sulci, suggesting an overall loss in the volume of the brain parenchyma (Fig. 2B). There were no associated features to suggest dehydration, nor was there a documented episode of anoxia or raised intracranial pressure. Five years later, at the age of 7 years and 9 months, the patient underwent an MRI scan of the head on a 0.5 Toshiba system with both T1- (600/20/2) and T2weighted (2000/40/120/1) transaxial images and the spin-echo technique. Once again, the angulation of scan was was chosen to approximate the gantry angulation of the previous scout CT study. The ventricles and sulci, as well as the surrounding brain parenchyma, all appeared within normal limits, with no evidence to suggest cerebral atrophy (Fig. 2C). This MRI scan approximated the initial admission CT study. Patient 2 was attending regular second grade classes at the time of this study. He has exhibited
difficulty in verbal comprehension and attention span since kindergarten, but no evidence of behavioral disturbance. Apart from fatigability and restlessness, Patient 2 was cooperative with the neuropsychological examination. Uninjured children were recruited as control subjects from the Dallas metropolitan area by advertising in the community newspaper. Groups of control children were selected to match each of the patients on the basis of demographic features (Table 1).
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
MRI using T1- (600/20/2) and T2-weighted (2000/40/120/1) transaxial images and the spin-echo technique on a 0.5 Tesla Toshiba magnet (Toshiba America MRI, Inc., South San Francisco, CA). The angle of the MRI scan was specifically chosen to approximate the gantry angulation obtained on the scout CT study, to provide the most comparable transaxial views for assessing any changes. The MRI studies showed the ventricular system and sulcal pattern to resemble more closely the patient's normal initial CT study, with no evidence to suggest cerebral atrophy (Fig. 1C). There were no abnormal foci of increased signal intensity within the white matter or elsewhere to suggest any shear injuries or additional pathological features. Patient 1 was preparing to enter the first grade at the time of this study. Her mother reported that she had had no learning problems or behavioral difficulty in kindergarten during the previous year. This child was highly cooperative during the neuropsychological examination, but she tended to forget the instructions.
ACKNOWLEDGMENTS This study was supported by grant NS-21889 from the National Institute of Neurological Diseases and Stroke and by a grant from The Greenery, Dallas, Texas. We are indebted to Melanie Meiselbach and Angela Thompson for assistance in word processing. Received, September 24, 1991. Accepted, January 29, 1992. Reprint requests: Harvey S. Levin, Ph.D., Division of Neurosurgery D-73, The University of Texas Medical Branch, Galveston, TX 77550. REFERENCES: (1-15) 1. 2.
Benton AL, Hamsher K: Multilingual Aphasia Examination. Iowa City, The University of Iowa, 1976. Bruce DA, Alavi A, Bilaniuk L, Dolinskas C,
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
DISCUSSION Follow-up MRI documented reversal of cerebral atrophy in both children surviving severe closed head injury. This conclusion is supported by comparing axial slices of their previous CT scans with MRI scans obtained at long intervals after injury. To mitigate concern about comparing the results obtained with different imaging techniques, the angle of the MRI scan was specifically chosen to approximate the gantry angulation obtained on the scout CT study. The thinner slices of the follow-up MRI study as compared with the CT study also diminish the possibility that the present findings underestimate the extent of cerebral atrophy or relate to volume averaging. A follow-up CT scan was not indicated because this procedure would have exposed these children to unnecessary radiation. Although the findings in our patients bear some resemblance to the reversible cerebral atrophy in abstinent alcoholics documented by CT (3), MRI (13), and pathological (10) studies, the mechanism is unclear. Neurohistological studies have shown reorganization of synaptic connections (plasticity) in necropsy and biopsy tissue from patients who had uncal herniation (7) and in plaques associated with multiple sclerosis (6). Reversal of mild cerebral atrophy was reported by Bruce and co-workers (2) as occurring in about one-third of a series of children who had sustained severe closed head injury and had initial CT findings indicative of diffuse cerebral swelling. They reported that the ventricles of these children returned to normal size with a decrease in the size of the sulci within 6 months to 1 year after trauma. Although neither of the children described
herein had early CT evidence of diffuse brain swelling, our finding of residual neurobehavioral deficits agrees with the observation by Bruce et al. (2) that "there was no close correlation between clinical recovery and return of the ventricles to normal size." Review of the intensive care records indicated that neither dehydration nor hypoxia were complications in either child. Although the mechanism underlying the reversal of cerebral atrophy in our patients remains unclear, several causes are plausible. While the changes demonstrated by MRI may not represent rehydration, shifts in the fluid content of the brain parenchyma could have produced changes in ventricular size similar to the acute increase in parenchymal bulk after shunting in hydrocephalus. Patient 1 could have developed ventricular enlargement and increased cerebrospinal fluid spaces resulting from increased outflow pressure of the cerebrospinal fluid secondary to subarachnoid hemorrhage. Although this mechanism is unlikely to explain the initial atrophy of Patient 2, who had no CT evidence of subarachnoid hemorrhage, the possibility cannot be eliminated. It is also conceivable that our MRI findings reflect continued brain growth that refilled the skull, despite the initial atrophy. The immediate loss of consciousness following acceleration-deceleration injury is consistent with the possibility that diffuse axonal injury contributed to the initial degeneration of the periventricular white matter in both of our patients (8) . Consequently, remyelination should also be considered as a plausible mechanism to account for their reversible atrophy. Despite the resolution of cerebral atrophy in both children, both exhibited residual neuropsychological impairment on a broad range of cognitive and memory tests. In view of their normal preinjury functioning, it is unlikely that the neuropsychological deficits can be attributed to factors other than traumatic brain injury. In contrast to the anatomic evidence for reversible atrophy, it is plausible that functional brain imaging would reveal residual cerebral dysfunction in both children.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
relatively intact (Table 1). Patient 2 used an excessive number of trials to reach criterion on the go-no-go task as compared with the normal control subjects. Although Patient 2's recall of a list of related words over five trials fell below that of his control group, he performed relatively well in cued and delayed recall (Fig. 3). In contrast, his concept formation (Wisconsin Card Sorting) was unimpaired relative to his control group. Patient 2's verbal intellectual functioning (WISC-R subtests) was impaired, particularly in abstract reasoning, whereas his visuospatial ability was in the borderline range. His prorated Verbal and Performance IQ scores were 59 and 75, respectively. Taken together, the neuropsychological findings revealed contrasting patterns in these patients, despite their common intellectual impairment on the Wechsler Scale. Patient 1 had deficient performance on tests of cognition that are sensitive to frontal lobe dysfunction, but she exhibited an inconsistent pattern of perseverative errors. In addition, she had deficient recall under both short and long delay conditions. Patient 2 exhibited reduced verbal skills (i.e., on tests of verbal fluency, immediate verbal recall, and Verbal IQ) relative to his control group, while showing no consistent evidence of frontal lobe dysfunction (note his relatively intact Wisconsin Card Sorting) or excessive memory interference (i.e., short and long delayed recall).
5. 6.
7.
8. 9.
10.
11.
12.
13.
14. 15.
COMMENTS The authors of this work describe two children in whom brain volume decreased (cerebral atrophy) after severe closed head injury and in whom this atrophy resolved on late follow-up magnetic resonance imaging scans. Neuropsychological data is presented showing moderate disabilities. Several possible mechanisms of this phenomenon, which has been previously reported in abstaining alcoholics, are
Harold L. Rekate Phoenix, Arizona REFERENCES: (1,2) 1.
2.
Carolan PL, McLaurin RL, Towlin RB, Towlin JA, Egelhoff JC. Benign extraaxial collections of infancy. Pediatr Neurosci 12:140-144, 1986. Rekate HL, Brodkey JA, Chizeck HJ, El Seikka W, Ko WH: Ventricular volume regulation: A mathematical model and computer simulation. Pediatr Neurosci 14:7784, 1988.
The concept of reversible cerebral atrophy after head injury is intriguing, especially since both cases reported here were in children. The authors' discussion of the potential explanation or mechanism is excellent. The mystery of why this occurs remains, but the possibilities of several mechanisms are discussed. I wonder, however, if the authors are correct in their statement regarding Patient 2. They have noted that transitory "external" hydrocephalus could not have happened because of the absence of subarachnoid hemorrhaging on the initial study.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
4.
discussed. These include changes in the water content of the white matter as seen after shunting and loss of brain parenchymal elements with subsequent regrowth. The volume of the brain is in a dynamic equilibrium with the other fluid within the intracranial compartment (blood and cerebrospinal fluid). An increase in resistance to outflow at the sagittal sinus level would lead to an increase in all cerebrospinal fluid compartments at the expense of the brain. Changes in intracranial pressure would depend on the intrinsic stiffness or turgor of the brain (2) . The most likely explanation for the events described in this paper involve a terminal cerebrospinal fluid absorptive problem, probably related to subarachnoid blood (even when not seen on a computed tomographic scan), as well as a change in the turgor of the brain as a result of the damage done by impact or subsequent swelling. This is somewhat analogous to and radiologically indistinguishable from the benign extracerebral collections of infancy, which are also reversible (1). While continued brain growth may explain the resolution of the atrophy in the 1-year old, it is very difficult to ascribe the displacement of large volumes of cerebrospinal fluid to brain growth in a 4-year-old. This paper is thought-provoking, and may help us to understand the outcome of a subset of children with diffuse pediatric head injury. This whole subjectits pathophysiology, treatment, and outcome-remains to some extent con-fusing. Careful study of these patients as done in this work will help to clarify these issues.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
3.
Obrist W, Uzzell B: Diffuse cerebral swelling following head injuries in children: The syndrome of "malignant brain edema. " J Neurosurg 54:170- 178, 1981. Carlen PL, Wilkinson DA: Reversibility of alcohol-related brain damage: Clinical and experimental observations. Acta Med Scand 717:19- 26, 1987. Delis DC, Kramer JH, Kaplan E, Ober BA: The California Verbal Learning Test (research edition). New York, Psychological Corporation, 1987. Drewe EA: Go-no-go learning after frontal lobe lesions in humans. Cortex 11:8-16, 1975. Ghatak NR, Leshner RT, Price AC, Felton WL: Remyelination in the human central nervous system. J Neuropathol Exp Neurol 48:507-518, 1989. Grady MS, Jane JA, Steward O: Synaptic reorganization within the human central nervous system following injury. J Neurosurg 71:534- 537, 1989. Graham DI, Adams JH, Doyle D, Lawrence AE, McLellan DR, Ng HK: Fatal head injury in children. J Clin Pathol 42:18-22, 1989. Grant DA, Berg EA: A behavioral analysis of degree of reinforcement and ease of shifting to new responses in a Weigl-type card-sorting problem. J Exp Psychol 38:404-411, 1948. Harper CG, Kril JJ, Daly JM: Brain shrinkage in alcoholics is not caused by changes in hydration: A pathological study. J Neurol Neurosurg Psychiatry 51:124-127, 1988. Jones-Gotman M, Milner B: Design fluency: The invention of nonsense drawings after focal cortical lesions. Neuropsychologia 55:653-674, 1977. Kaye JA, Grady CL, Haxby JV, Moore A, Friedland RP: Plasticity in the aging brain: Reversibility of anatomic, metabolic, and cognitive deficits in normal-pressure hydrocephalus following shunt surgery. Arch Neurol 47:1336- 1341, 1990. Schroth G, Naegele T, Klose U, Mann D, Petersen D: Reversible brain shrinkage in abstinent alcoholics, measured by MRI. Neuroradiology 30:385-389, 1988. Teasdale G, Jennett B: Assessment of coma and impaired consciousness: A practical scale. Lancet 2:81-84, 1974. Wechsler D: WISC-R manual. Wechsler Intelligence Scale for Children--Revised. New York, Psychological Corporation, 1974.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Marion L. Walker Salt Lake City, Utah
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
Although this may decrease the likelihood that external hydrocephalus may happen, it certainly does not eliminate that possibility. It makes intuitive sense that "external" hydrocephalus may be a part of the mechanisms seen in many of these children, or that further brain growth eventually fills the space created in the small child with a growing brain. It is obvious that functional imaging of the brain will be required in order to make more accurate predictive assessments of patients with head injuries, especially young children. The current imaging modalities may reflect a return to more normal anatomy, but certainly have not been correlated with a return to normal brain function.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Figure 2. Imaging studies of Patient 2: A, a transaxial CT scan through the level of the lateral and third ventricles discloses nothing abnormal on the day of injury. B, a transaxial CT scan without contrast material performed 17 days later through the same level as Figure 2A shows an increase in ventricular size as well as prominence of the convexity sulci, consistent with generalized atrophy. C, T2-weighted (2000/120/1) transaxial MRI scan shows return of the ventricular size and sulci to an appearance more closely approximate to that seen on the patient's initial admission CT scan. The MRI study was performed approximately 5 years after the injury and demonstrates return of the patient's brain volume to normal with no evidence to suggest generalized cerebral atrophy.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
Figure 1. Imaging studies of Patient 1: A, a transaxial CT study without contrast material performed at admission shows a normal ventricular system with no evidence of any intraparenchymal pathology nor fluid collection. B, a transaxial CT study performed 20 days later shows prominence of the ventricular system and sulci, consistent with generalized cerebral atrophy. C, T2-weighted (2000/40/120/1) axial MRI scan performed approximately 1.5 years after the initial injury demonstrates a normal ventricular system and sulcal pattern that closely approximate the normal findings of the admission CT scan. There has been interval reexpansion or volume increase of the brain parenchyma without evidence of atrophy.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
Figure 3. Performance of Patients One and Two on the California Verbal Learning Test, which involves recall of a 15-word list over 5 trials, with delayed recall and recognition measures. The performance of each patient is compared with the mean for the respective control group. Short delay recall is assessed immediately after presentation of a distractor word list, and long delay recall is tested after a 20-minute interval.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
Table 1. Demographic and Neuropsychological Findings in Patients and Control Subjects
PATIENTS AND METHODS Patient 1 At the age of 4 years and 10 months, this girl sustained a severe closed head injury that produced immediate loss of consciousness when she was struck by a car in June 1989. The postresuscitation Glasgow Coma Scale (GCS) score was seven (eye opening = 1, motor = 5, verbal = 1), and the patient's consciousness remained impaired for 2 weeks (14). Her pupils reacted to light, and the neurological examination revealed right hemiplegia. Associated injuries included fractures of the right femur and pelvis. A right frontal ventriculostomy was performed to monitor intracranial pressure, which remained below 20 mm Hg. The initial CT scan on the day of injury showed a small amount of subarachnoid blood within the left ambient cistern without evidence of associated intraparenchymal hematoma. The ventricular system was normal, with no shift of the midline structures nor extra-axial fluid collections (Fig. 1A). The ventricular size and sulci were within normal limits for the patient's age, particularly in light of the fact that there was no evidence of raised intracranial pressure. A follow-up CT scan without contrast material performed 20 days later showed a marked increase in the size of the third and lateral ventricles as well as prominence of the sylvian fissures and convexity sulci (Fig. 1B). This dramatic change implies a loss of brain volume, or generalized cerebral atrophy. These changes were not thought to represent the end result of anoxia or diffuse brain swelling, as there was no evidence to suggest an anoxic event, nor was there evidence of raised intracranial pressure. In November 1990, at the age of 6 years (17 months after injury), the patient was examined by
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
AUTHOR(S): Levin, Harvey S., Ph.D; Mendelsohn, Dianne, M.D.; Bruce, Derek, M.D.; Harward, Harriet, M.A.; Culhane, Kathleen A., M.A.; Eisenberg, Howard M., M.D.
Grady et al. inferred that sprouting of connections from entorhinal cortex could potentially support behavioral recovery in nonfatal injuries. The potential for recovery of the brain compromised by normal pressure hydrocephalus was supported by a recent case study (12) documenting reduced ventricular volume and increased cerebral metabolism 1 year after shunt surgery in a 78-year-old demented woman. In a serial CT study of 63 children who developed diffuse cerebral swelling secondary to a severe closed head injury, Bruce et al. (2) found that 22 patients (35%) had reversal of mild cerebral atrophy within 6 to 12 months that was not related to their clinical recovery. However, these investigators did not report neurobehavioral findings, nor was it clear whether reversible atrophy was restricted to children with initial brain swelling. In an ongoing investigation employing MRI and tests of neurobehavioral functioning in head-injured patients, we have discovered evidence of reversible cerebral atrophy in 2 of the 55 patients studied to date. We report on these patients to document the morphological changes in their brains and to raise questions concerning the underlying mechanism of these changes. Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Neurosurgery 1992-98 December 1992, Volume 31, Number 6 1117 Reversibility of Cerebral Atrophy after Head Injury in Children Case Report
NEUROPSYCHOLOGICAL TESTS Frontal lobe functioning was evaluated with tests of fluency, problem solving, and inhibitory control. Verbal fluency (i.e., production of words beginning with a specified letter under timed conditions) (1) and semantic memory (i.e., production of animal names under timed conditions) were assessed, as was design fluency (11) (i.e., creation of nonsense designs under timed conditions). Two computerized tasks were administered, including the Wisconsin Card Sorting Test (9) (a measure of problem-solving ability and response flexibility), and the go-no-go task (5), which measures inhibitory control of responses. Memory, including immediate and delayed recall and recognition, was assessed via the California Verbal Learning Test (4). Four subtests of the Wechsler Intelligence Scale for Children--Revised (WISC-R) (15) (including two verbal and two performance tasks) were also administered. NEUROPSYCHOLOGICAL FINDINGS The neuropsychological findings for both patients and their respective control groups are summarized in Table 1. Comparison of the findings for Patient 1 with a group of age-matched control subjects revealed diminished verbal fluency (i.e., associating words with initial letters) and reduced semantic memory as reflected by animal naming. Patient 1 also created fewer nonsense designs than her control group. Patient 1 used almost twice the number of trials to reach criterion on the go-no-go test as did her control group, although her performance on this response inhibition task was not impulsive. The Wisconsin Test was discontinued because Patient 1 was unable to sort the cards correctly according to even a single category. Intellectual functioning, as estimated from four subtests of the WISC-R, was also impaired, particularly in the areas of abstract verbal reasoning and visuospatial construction abilities. Her prorated Verbal and Performance IQ scores were 66 and 61, respectively. Patient 1's memory was impaired, as reflected by her poor recall of the first list of words across five trials and lower delayed recall as compared with the control group (Fig. 3). The verbal fluency findings for Patient 2 indicate that he invented fewer unique designs under timed conditions relative to his control group, but his verbal fluency was within one standard deviation of that of the control subjects (Table 1). However, Patient 2 exhibited no consistent pattern of perseverations. In contrast to Patient 2's low productivity on fluency tests, his semantic memory (i.e., animal naming) was
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Patient 2 At the age of 1 year and 8 months, this boy sustained a severe closed head injury that produced immediate loss of consciousness when he was hit by a truck in August 1985. His postresuscitation GCS score was 8 (eye opening = 2, motor = 5, verbal = 1). His pupils were equal and reactive, and a left-sided gaze preference was present. He began following commands (albeit inconsistently) at about 1 hour after admission, and became increasingly alert over the ensuing 5 days. However, a focal motor seizure 5 days after injury resulted in right hemiparesis that persisted for 2 weeks. There were no extracranial injuries and no surgical procedures. A transaxial CT scan of the head without contrast material on the day of admission showed a normal ventricular system with preservation of the cisterns and no evidence of any intraparenchymal hematoma, edema, mass effect, or shift of the midline structures (Fig. 2A). A transaxial CT scan without contrast material performed 17 days later showed a marked change, with enlargement of the ventricular system and convexity sulci, suggesting an overall loss in the volume of the brain parenchyma (Fig. 2B). There were no associated features to suggest dehydration, nor was there a documented episode of anoxia or raised intracranial pressure. Five years later, at the age of 7 years and 9 months, the patient underwent an MRI scan of the head on a 0.5 Toshiba system with both T1- (600/20/2) and T2weighted (2000/40/120/1) transaxial images and the spin-echo technique. Once again, the angulation of scan was was chosen to approximate the gantry angulation of the previous scout CT study. The ventricles and sulci, as well as the surrounding brain parenchyma, all appeared within normal limits, with no evidence to suggest cerebral atrophy (Fig. 2C). This MRI scan approximated the initial admission CT study. Patient 2 was attending regular second grade classes at the time of this study. He has exhibited
difficulty in verbal comprehension and attention span since kindergarten, but no evidence of behavioral disturbance. Apart from fatigability and restlessness, Patient 2 was cooperative with the neuropsychological examination. Uninjured children were recruited as control subjects from the Dallas metropolitan area by advertising in the community newspaper. Groups of control children were selected to match each of the patients on the basis of demographic features (Table 1).
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
MRI using T1- (600/20/2) and T2-weighted (2000/40/120/1) transaxial images and the spin-echo technique on a 0.5 Tesla Toshiba magnet (Toshiba America MRI, Inc., South San Francisco, CA). The angle of the MRI scan was specifically chosen to approximate the gantry angulation obtained on the scout CT study, to provide the most comparable transaxial views for assessing any changes. The MRI studies showed the ventricular system and sulcal pattern to resemble more closely the patient's normal initial CT study, with no evidence to suggest cerebral atrophy (Fig. 1C). There were no abnormal foci of increased signal intensity within the white matter or elsewhere to suggest any shear injuries or additional pathological features. Patient 1 was preparing to enter the first grade at the time of this study. Her mother reported that she had had no learning problems or behavioral difficulty in kindergarten during the previous year. This child was highly cooperative during the neuropsychological examination, but she tended to forget the instructions.
ACKNOWLEDGMENTS This study was supported by grant NS-21889 from the National Institute of Neurological Diseases and Stroke and by a grant from The Greenery, Dallas, Texas. We are indebted to Melanie Meiselbach and Angela Thompson for assistance in word processing. Received, September 24, 1991. Accepted, January 29, 1992. Reprint requests: Harvey S. Levin, Ph.D., Division of Neurosurgery D-73, The University of Texas Medical Branch, Galveston, TX 77550. REFERENCES: (1-15) 1. 2.
Benton AL, Hamsher K: Multilingual Aphasia Examination. Iowa City, The University of Iowa, 1976. Bruce DA, Alavi A, Bilaniuk L, Dolinskas C,
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
DISCUSSION Follow-up MRI documented reversal of cerebral atrophy in both children surviving severe closed head injury. This conclusion is supported by comparing axial slices of their previous CT scans with MRI scans obtained at long intervals after injury. To mitigate concern about comparing the results obtained with different imaging techniques, the angle of the MRI scan was specifically chosen to approximate the gantry angulation obtained on the scout CT study. The thinner slices of the follow-up MRI study as compared with the CT study also diminish the possibility that the present findings underestimate the extent of cerebral atrophy or relate to volume averaging. A follow-up CT scan was not indicated because this procedure would have exposed these children to unnecessary radiation. Although the findings in our patients bear some resemblance to the reversible cerebral atrophy in abstinent alcoholics documented by CT (3), MRI (13), and pathological (10) studies, the mechanism is unclear. Neurohistological studies have shown reorganization of synaptic connections (plasticity) in necropsy and biopsy tissue from patients who had uncal herniation (7) and in plaques associated with multiple sclerosis (6). Reversal of mild cerebral atrophy was reported by Bruce and co-workers (2) as occurring in about one-third of a series of children who had sustained severe closed head injury and had initial CT findings indicative of diffuse cerebral swelling. They reported that the ventricles of these children returned to normal size with a decrease in the size of the sulci within 6 months to 1 year after trauma. Although neither of the children described
herein had early CT evidence of diffuse brain swelling, our finding of residual neurobehavioral deficits agrees with the observation by Bruce et al. (2) that "there was no close correlation between clinical recovery and return of the ventricles to normal size." Review of the intensive care records indicated that neither dehydration nor hypoxia were complications in either child. Although the mechanism underlying the reversal of cerebral atrophy in our patients remains unclear, several causes are plausible. While the changes demonstrated by MRI may not represent rehydration, shifts in the fluid content of the brain parenchyma could have produced changes in ventricular size similar to the acute increase in parenchymal bulk after shunting in hydrocephalus. Patient 1 could have developed ventricular enlargement and increased cerebrospinal fluid spaces resulting from increased outflow pressure of the cerebrospinal fluid secondary to subarachnoid hemorrhage. Although this mechanism is unlikely to explain the initial atrophy of Patient 2, who had no CT evidence of subarachnoid hemorrhage, the possibility cannot be eliminated. It is also conceivable that our MRI findings reflect continued brain growth that refilled the skull, despite the initial atrophy. The immediate loss of consciousness following acceleration-deceleration injury is consistent with the possibility that diffuse axonal injury contributed to the initial degeneration of the periventricular white matter in both of our patients (8) . Consequently, remyelination should also be considered as a plausible mechanism to account for their reversible atrophy. Despite the resolution of cerebral atrophy in both children, both exhibited residual neuropsychological impairment on a broad range of cognitive and memory tests. In view of their normal preinjury functioning, it is unlikely that the neuropsychological deficits can be attributed to factors other than traumatic brain injury. In contrast to the anatomic evidence for reversible atrophy, it is plausible that functional brain imaging would reveal residual cerebral dysfunction in both children.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
relatively intact (Table 1). Patient 2 used an excessive number of trials to reach criterion on the go-no-go task as compared with the normal control subjects. Although Patient 2's recall of a list of related words over five trials fell below that of his control group, he performed relatively well in cued and delayed recall (Fig. 3). In contrast, his concept formation (Wisconsin Card Sorting) was unimpaired relative to his control group. Patient 2's verbal intellectual functioning (WISC-R subtests) was impaired, particularly in abstract reasoning, whereas his visuospatial ability was in the borderline range. His prorated Verbal and Performance IQ scores were 59 and 75, respectively. Taken together, the neuropsychological findings revealed contrasting patterns in these patients, despite their common intellectual impairment on the Wechsler Scale. Patient 1 had deficient performance on tests of cognition that are sensitive to frontal lobe dysfunction, but she exhibited an inconsistent pattern of perseverative errors. In addition, she had deficient recall under both short and long delay conditions. Patient 2 exhibited reduced verbal skills (i.e., on tests of verbal fluency, immediate verbal recall, and Verbal IQ) relative to his control group, while showing no consistent evidence of frontal lobe dysfunction (note his relatively intact Wisconsin Card Sorting) or excessive memory interference (i.e., short and long delayed recall).
5. 6.
7.
8. 9.
10.
11.
12.
13.
14. 15.
COMMENTS The authors of this work describe two children in whom brain volume decreased (cerebral atrophy) after severe closed head injury and in whom this atrophy resolved on late follow-up magnetic resonance imaging scans. Neuropsychological data is presented showing moderate disabilities. Several possible mechanisms of this phenomenon, which has been previously reported in abstaining alcoholics, are
Harold L. Rekate Phoenix, Arizona REFERENCES: (1,2) 1.
2.
Carolan PL, McLaurin RL, Towlin RB, Towlin JA, Egelhoff JC. Benign extraaxial collections of infancy. Pediatr Neurosci 12:140-144, 1986. Rekate HL, Brodkey JA, Chizeck HJ, El Seikka W, Ko WH: Ventricular volume regulation: A mathematical model and computer simulation. Pediatr Neurosci 14:7784, 1988.
The concept of reversible cerebral atrophy after head injury is intriguing, especially since both cases reported here were in children. The authors' discussion of the potential explanation or mechanism is excellent. The mystery of why this occurs remains, but the possibilities of several mechanisms are discussed. I wonder, however, if the authors are correct in their statement regarding Patient 2. They have noted that transitory "external" hydrocephalus could not have happened because of the absence of subarachnoid hemorrhaging on the initial study.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
4.
discussed. These include changes in the water content of the white matter as seen after shunting and loss of brain parenchymal elements with subsequent regrowth. The volume of the brain is in a dynamic equilibrium with the other fluid within the intracranial compartment (blood and cerebrospinal fluid). An increase in resistance to outflow at the sagittal sinus level would lead to an increase in all cerebrospinal fluid compartments at the expense of the brain. Changes in intracranial pressure would depend on the intrinsic stiffness or turgor of the brain (2) . The most likely explanation for the events described in this paper involve a terminal cerebrospinal fluid absorptive problem, probably related to subarachnoid blood (even when not seen on a computed tomographic scan), as well as a change in the turgor of the brain as a result of the damage done by impact or subsequent swelling. This is somewhat analogous to and radiologically indistinguishable from the benign extracerebral collections of infancy, which are also reversible (1). While continued brain growth may explain the resolution of the atrophy in the 1-year old, it is very difficult to ascribe the displacement of large volumes of cerebrospinal fluid to brain growth in a 4-year-old. This paper is thought-provoking, and may help us to understand the outcome of a subset of children with diffuse pediatric head injury. This whole subjectits pathophysiology, treatment, and outcome-remains to some extent con-fusing. Careful study of these patients as done in this work will help to clarify these issues.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
3.
Obrist W, Uzzell B: Diffuse cerebral swelling following head injuries in children: The syndrome of "malignant brain edema. " J Neurosurg 54:170- 178, 1981. Carlen PL, Wilkinson DA: Reversibility of alcohol-related brain damage: Clinical and experimental observations. Acta Med Scand 717:19- 26, 1987. Delis DC, Kramer JH, Kaplan E, Ober BA: The California Verbal Learning Test (research edition). New York, Psychological Corporation, 1987. Drewe EA: Go-no-go learning after frontal lobe lesions in humans. Cortex 11:8-16, 1975. Ghatak NR, Leshner RT, Price AC, Felton WL: Remyelination in the human central nervous system. J Neuropathol Exp Neurol 48:507-518, 1989. Grady MS, Jane JA, Steward O: Synaptic reorganization within the human central nervous system following injury. J Neurosurg 71:534- 537, 1989. Graham DI, Adams JH, Doyle D, Lawrence AE, McLellan DR, Ng HK: Fatal head injury in children. J Clin Pathol 42:18-22, 1989. Grant DA, Berg EA: A behavioral analysis of degree of reinforcement and ease of shifting to new responses in a Weigl-type card-sorting problem. J Exp Psychol 38:404-411, 1948. Harper CG, Kril JJ, Daly JM: Brain shrinkage in alcoholics is not caused by changes in hydration: A pathological study. J Neurol Neurosurg Psychiatry 51:124-127, 1988. Jones-Gotman M, Milner B: Design fluency: The invention of nonsense drawings after focal cortical lesions. Neuropsychologia 55:653-674, 1977. Kaye JA, Grady CL, Haxby JV, Moore A, Friedland RP: Plasticity in the aging brain: Reversibility of anatomic, metabolic, and cognitive deficits in normal-pressure hydrocephalus following shunt surgery. Arch Neurol 47:1336- 1341, 1990. Schroth G, Naegele T, Klose U, Mann D, Petersen D: Reversible brain shrinkage in abstinent alcoholics, measured by MRI. Neuroradiology 30:385-389, 1988. Teasdale G, Jennett B: Assessment of coma and impaired consciousness: A practical scale. Lancet 2:81-84, 1974. Wechsler D: WISC-R manual. Wechsler Intelligence Scale for Children--Revised. New York, Psychological Corporation, 1974.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Marion L. Walker Salt Lake City, Utah
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
Although this may decrease the likelihood that external hydrocephalus may happen, it certainly does not eliminate that possibility. It makes intuitive sense that "external" hydrocephalus may be a part of the mechanisms seen in many of these children, or that further brain growth eventually fills the space created in the small child with a growing brain. It is obvious that functional imaging of the brain will be required in order to make more accurate predictive assessments of patients with head injuries, especially young children. The current imaging modalities may reflect a return to more normal anatomy, but certainly have not been correlated with a return to normal brain function.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Figure 2. Imaging studies of Patient 2: A, a transaxial CT scan through the level of the lateral and third ventricles discloses nothing abnormal on the day of injury. B, a transaxial CT scan without contrast material performed 17 days later through the same level as Figure 2A shows an increase in ventricular size as well as prominence of the convexity sulci, consistent with generalized atrophy. C, T2-weighted (2000/120/1) transaxial MRI scan shows return of the ventricular size and sulci to an appearance more closely approximate to that seen on the patient's initial admission CT scan. The MRI study was performed approximately 5 years after the injury and demonstrates return of the patient's brain volume to normal with no evidence to suggest generalized cerebral atrophy.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
Figure 1. Imaging studies of Patient 1: A, a transaxial CT study without contrast material performed at admission shows a normal ventricular system with no evidence of any intraparenchymal pathology nor fluid collection. B, a transaxial CT study performed 20 days later shows prominence of the ventricular system and sulci, consistent with generalized cerebral atrophy. C, T2-weighted (2000/40/120/1) axial MRI scan performed approximately 1.5 years after the initial injury demonstrates a normal ventricular system and sulcal pattern that closely approximate the normal findings of the admission CT scan. There has been interval reexpansion or volume increase of the brain parenchyma without evidence of atrophy.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
Figure 3. Performance of Patients One and Two on the California Verbal Learning Test, which involves recall of a 15-word list over 5 trials, with delayed recall and recognition measures. The performance of each patient is compared with the mean for the respective control group. Short delay recall is assessed immediately after presentation of a distractor word list, and long delay recall is tested after a 20-minute interval.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Downloaded from https://academic.oup.com/neurosurgery/article-abstract/31/6/1117/2752146 by university of winnipeg user on 23 January 2019
Table 1. Demographic and Neuropsychological Findings in Patients and Control Subjects
