| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Duke University Medical Center, Box 3808, Durham, NC
27710.
2 Departments of Radiology, Oncology, and Biomedical Engineering, Emory
University School of Medicine, Atlanta, GA.
3 Mallinckrodt Institute of Radiology, Washington University School of Medicine,
St. Louis, MO.
4 Department of Pediatrics, Duke University Medical Center, Durham, NC.
5 Department of Pediatrics, Carolina Institute for Developmental Disabilities,
University of North Carolina, Chapel Hill, NC.
Received February 22, 2008;
accepted after revision August 14, 2008.
Address correspondence to J. M. Provenzale.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Nine infants with Krabbe's disease underwent a total of 19 MR studies during the first year of life as well as tests of mental development, gross motor skills, and fine motor skills (score range: 0-100) within 1 month of imaging. MR scans were scored using the Loes severity scale based on signal abnormality and atrophy, ranging from 0 (best) to 32. We performed three comparisons (Student's t test): each test versus total brain Loes score, fine motor and gross motor tests versus Loes score for the pyramidal tract, and fine motor and gross motor tests versus Loes score for the internal capsule.
RESULTS. Mean test results were 65 ± 31 for mental development, 48 ± 39 for gross motor score, 57 ± 35 for fine motor score, and mean total brain score was 7.79 ± 6.20. Correlations for total Loes score were -0.78 (p = 0.003) for mental development, -0.74 (p = 0.003) for gross motor function, and -0.80 (p < 0.001) for fine motor function. Correlations for pyramidal system Loes scores were -0.73 (p = 0.003) for fine motor function and -0.58 (p = 0.028) for gross motor function. Correlation between Loes scores for internal capsule and fine motor function was -0.38 (p > 0.05) and between Loes scores for internal capsule and gross motor function was -0.35 (p > 0.05).
CONCLUSION. The very good correlation between testing results and Loes scores for the entire brain and moderately good correlation between test results and scores for specific brain regions indicate the Loes scoring system likely provides a reasonable means for assessing prognosis and therapeutic response for infants with Krabbe's disease.
Keywords: brain • Krabbe's disease • leukodystrophy • neurobehavioral development • stem cell • white matter
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The major forms of Krabbe's disease include an early-onset (infantile) form and a late-onset form. The early-onset form, which is seen in young infants, is the more severe type and is characterized by rapidly progressive neurologic deterioration resulting in a vegetative state and, typically, death within the first few years of life [2]. Until recently, no treatment for Krabbe's disease existed. However, in the past few years, hematopoietic stem cell transplantation has emerged as a treatment that can substantially improve outcome by delivering β-galactocerebrosidase through donor cell engraftment to various tissues, thereby mitigating the effects of enzyme deficiency on the central and peripheral nervous systems [3]. As a result, neurologic function is substantially improved and life expectancy is increased.
With the advent of a therapy for this previously untreatable disease, greater attention has been focused on early screening as well as evaluation of degree of neurologic disability to assess candidates for therapy and to monitor therapeutic response. Recently, New York became the first state to initiate a pilot neonate screening program for Krabbe's disease that involves a serologic test for the missing enzyme followed by mutational analysis to confirm diagnosis [4, 5]. MRI plays a fundamental role in assessing disease status and progression in patients with Krabbe's disease but may be especially important because infants can be expected to be relatively asymptomatic at birth, when screening is conducted. Furthermore, the complete extent of correlations with genetic and clinical phenotypes has not been elucidated. Therefore, a need exists for a better understanding of the manner in which MRI findings correlate with clinical outcomes. Although the aim of this study was not to assess response to therapy, we have previously documented that higher initial Loes scores [6] correlate with worse initial neurodevelopmental status and subsequent poorer response to therapy [3]. In prior studies, the timing of transplantation was found to be important to improvement of clinical status of these patients [3, 7]. We refer readers who are interested in the effects of stem cell transplantation in this patient population to our previous articles reporting our findings [3, 7]. Infants who are diagnosed after the first month of life are identified on the basis of clinical neurologic symptoms; after transplantation (usually at 4-7 months old), these infants do not attain meaningful neurologic function. Infants who are diagnosed in the first month of life are identified on the basis of an affected older sibling and then subjected to laboratory testing for the characteristic enzyme deficit. These infants typically are asymptomatic at the time of transplantation and have substantial improvement in neurologic outcome [3, 7].
The Loes severity scale is an MR scoring protocol that was originally devised for measuring the degree of brain abnormality in X-linked adrenoleukodystrophy (ALD) [6]. It has been shown to correlate with clinical outcomes in ALD. However, it has subsequently been applied to the evaluation of children with Krabbe's disease and is considered a well-accepted imaging method of assessing Krabbe's disease [8]. Nonetheless, to our knowledge, the findings obtained using the Loes severity scale have not been correlated with neurologic and cognitive development parameters in Krabbe's disease. Such a correlation is needed for the severity scale to be further validated as a test for neuroradiologic monitoring of patients with Krabbe's disease.
We set out to compare Loes scores reflecting MRI findings with functional (clinical) scores in a series of young children with Krabbe's disease. The intent of the study was not to describe the results of stem cell transplantation, which are described elsewhere [3, 9], but instead to determine how closely Loes scores for the entire brain and for specific white matter regions compare with clinical outcomes for corresponding brain regions.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All infants were treated with allogeneic hematopoietic stem cell transplantation using unrelated cord blood donors who were known to express normal levels of β-galactocerebrosidase at our institution between October 1999 and August 2004. A total of 16 MRI examinations were per formed in nine children. One child underwent four MR examinations (at 2 weeks old and at ages 3, 4, and 10 months), two children had three scans (one child at 2 weeks old and at ages 2 and 4 months; another child at 1 week old and at ages 4 and 7 months), and six children had one scan (at the mean age of 5.2 months). Ages at the time of transplantation were 3 weeks (n = 2), 5 weeks (n = 2), 5 months (n = 1), 6 months (n = 2), 7 months (n = 1), and 9 months (n = 1). The six children with a single MR study underwent solely pretransplantation MRI and the other children had a single pretransplantation examination and one or more posttransplantation examinations. Serial imaging was typically performed to obtain an independent assessment as to whether myelination was proceeding.
The goal of this study was not to assess the therapeutic benefit of the transplantation but to compare clinical evaluations with MRI findings as reflected by Loes scores. Thus, all the infants are included in this study group regardless of whether their imaging was performed before or after transplantation. Loes scores (reported in the following text) for these 16 MR examinations were compared with scores on clinical testing (described later). The institutional review board at our medical center approved the study, and informed consent was obtained from the families of all patients.
Scoring of Clinical Parameters
All patients underwent serial clinical evaluations within 1 month of each
MRI study. Evaluations consisted of assignment of three clinical scores: a
mental development score, a gross motor skills score, and a fine motor skills
score. We used the Capute scale to provide a mental development quotient
(equivalent to intelligence quotient [IQ] in older children); similar to an IQ
score, a mental development score of 100 is the mean for age
[10]. In this article, we use
the term "mental development score" rather than the term
"developmental quotient" because we believe it will be more easily
understood by our intended audience.
The developmental quotient of the Capute scale is based on a combined assessment of visual motor skills and language skills and takes into account that cognitive and language function cannot be assessed in the neonate in the same manner as in older children.
The Mullen Scales of Early Learning were used to calculate fine motor age-equivalent quotients [11]. Gross motor age-equivalent quotients were calculated using the gross motor section of the Mullen Scales of Early Learning or the Peabody Developmental Motor Scales [12]. The final gross motor and fine motor quotients were calculated by dividing the age-equivalent score by the child's actual age. This ratio provides an age-adjusted measure of functioning.
Both the Mullen and the Capute scales consist of a checklist of age-appropriate actions that are documented by an observer. At the earliest ages, these actions are spontaneous or coaxed responses to stimuli provided by the evaluator. Examples of the Mullen gross motor tasks include, in the first 1-4 months of life, head rotation and vigorous movement of the arms and legs to sitting; and for ages 13-17 months, rolling over, standing, and walking. Examples of the fine motor test on the Mullen scale include flexing of both arms and hands to a midline orientation at 1-3 months as well as, at age 13-17 months, manipulating blocks. The Capute scales consist of language and visual motor problem-solving subscales. Examples of the language tasks include reacting to sound, smiling, and cooing at age 1-3 months; and orienting to sounds, babbling, using words, and following simple commands at age 14 months. Examples of the visual motor problem-solving tasks include visual fixation on an object and lifting the chin or chest off a table at 1-3 months, interacting, and, by age 14 months, inspecting objects in their environment, placing objects in containers, and using actions indicative of an understanding of object permanence (e.g., finding a cube hidden under a cup).
MRI Data Acquisition
A 1.5-T clinical MRI scanner (Signa, GE Healthcare) with a standard head
coil was used for imaging. All MRI examinations were performed using sedation.
MR sequences included trans verse unenhanced T1-weighted images (TR/TE,
600/20; number of excitations, 2) and transverse intermediate-weighted and
T2-weighted images. The sequence parameters for proton density-weighted and
T2-weighted images were 2,800/30 with 2 excitations and 2,800/100 with 2
excitations, respectively; field of view, 22 x 22 cm; matrix size, 256
(frequency direction) x 192 (phase direction); slice thickness, 5 mm;
and interslice gap, 2.5 mm.
Scoring of MRI Studies
Two neuroradiologists who were blinded to the clinical status of patients
graded the MR images using the Loes MRI severity scale in a consensus manner
[8]. The scoring scale is
discussed in detail next.
The Loes severity scale for Krabbe's disease uses a scoring system for T1-, proton density-, and T2-weighted images [8]. This scoring system divides the entire brain into nine regions (parietooccipital white matter [WM], anterior temporal WM, frontal WM, corpus callosum, visual pathway, pyramidal system, cerebellum, basal ganglia, and anterior thalamus), which are themselves divided into 23 subregions. As an example, one of the brain regions, termed the "pyramidal system," is divided into three subregions, which are the corona radiata, the internal capsule, and the brainstem. Each of the 23 subregions is awarded a score of 0 (normal), 0.5 (unilateral abnormality), or 1 (bilateral abnormality) on the basis of the presence of abnormal signal intensity, for a possible total of 23 points if all subregions have abnormal signal intensity. In addition to signal abnormality, degree of atrophy is scored on a scale of 0-1 for four brain regions (parietooccipital WM, anterior temporal WM, frontal WM, cerebellum) and two brain sub regions (genu and splenium of the corpus callosum) as well as on a 0-3 scale for the brain as a whole, providing up to 9 points for atrophy. Thus, the range of scores for the Loes system, when combining possible points for signal abnormality (up to 23 points) and atrophy (up to 9 points), is 0 (normal) to 32 (markedly abnormal). An example of the use of this scoring system to evaluate serial MR scans in a study patient is shown in Figures 1A, 1B, 1C, 1D, 1E, 1F, 1G, 1H, 1I, 1J, 1K, and 1L.
|
|
|
|
|
|
|
|
|
|
|
|
For comparison 2, we assessed one of nine brain regions designated by the Loes scale, that is, the pyramidal system. This brain region is especially relevant to two clinical functions that we assessed (i.e., fine motor age-equivalent quotient and gross motor age-equivalent quotient) by virtue of serving as the location of motor pathways. As mentioned earlier, in the Loes scoring system the pyramidal system is composed of three subregions—the corona radiata, the internal capsule, and the brainstem. The score for the pyramidal system region is the sum of scores for the three subregions (each of which can have a score ranging from 0 to 1), and thus the score for the pyramidal system can range from 0 to 3.
In comparison 2, we correlated the Loes score of the pyramidal system (individually) with the gross motor age-equivalent quotient and fine motor age-equivalent quotient. We did not compare the remaining clinical component, that is, the mental development score, with a specific brain region because very limited information about brain colocalization of cognitive and language functions exists in infants; these functions are not as regionally specialized in infants as in older children.
In comparison 3, we further analyzed the pyramidal system region by correlating the Loes score for one of its subregions (i.e., the internal capsule) with both the fine motor age-equivalent quotient and the gross motor age-equivalent quotient. The internal capsule was chosen because it is the component of the pyramidal system that is most widely represented by the motor fiber pathway and thus most appropriate to compare with gross motor and fine motor scores.
Statistical Analysis
We used mixed models to adjust for multiple observations per person and
then calculated standardized coefficients, which are equivalent to
correlations, using p < 0.05 as statistically significant.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
MRI (Loes) Scores
Mean Loes score for the entire brain for the 16 MR examinations was 7.79
± 6.20 (range: 0-22.5). Mean Loes score for the pyramidal system was
1.18 ± 1.10 (range: 0-3.00), and mean Loes score for the internal
capsule was 0.47 ± 0.48 (range: 0-1.00).
Comparison of Clinical Scores and MRI (Loes) Scores
Comparison 1—Correlations between each clinical score and
total Loes score were as follows: for mental development index, -0.78
(p = 0.003); for gross motor age-equivalent quotient, -0.74
(p = 0.003); and for fine motor age-equivalent quotient, -0.80
(p < 0.001). All comparisons achieved statistical
significance.
Comparison 2—Correlations between clinical scores and Loes scores for the pyramidal system were as follows: fine motor age-equivalent quotient, -0.73 (p = 0.003); and gross motor age-equivalent quotient, -0.58 (p = 0.028). Both comparisons achieved statistical significance (Table 2).
|
Comparison 3—Correlations between clinical scores and Loes score for the internal capsule were fine motor age-equivalent quotient, -0.38 (p = 0.22); and gross motor age-equivalent quotient, -0.35 (p = 0.24). Neither comparison was statistically significant (Table 3).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
These findings suggest that MRI findings using Loes scores in Krabbe's disease reflect fundamental neurologic processes associated with those regions, but only to a limited degree. One explanation is that the neurologic functions assessed by the gross motor and fine motor evaluations are represented by other brain regions than solely those in the pyramidal system. Another explanation is that this measure does not capture the degree of abnormality in these regions, especially in infancy. Finally, comparison 3 showed that Loes scoring results in the very specific brain region of the internal capsule correlated poorly with relevant motor functions.
One explanation for the difference in the strength of the correlations across our analyses may be the differences in range of possible scores in various components of the Loes scoring system. The range of whole-brain Loes scores is relatively wide, that is, 0-32, whereas it is much narrower for individual components: 0-3 for the pyramidal system and 0-1 for the internal capsule. These differences in scoring scales may, in part, account for the reduced correlation for comparisons 2 and 3. An alternative measure that provided a finer measure of brain abnormalities with a broader range of scores may find stronger correlations between MRI findings and function for smaller areas of the brain.
The Loes severity scale was initially developed for assessment of brains of children with ALD and has been shown to correlate with severity of neurologic deficits and to be predictive of neuroradiologic disease progression as well as survival in patients with ALD [13]. In one study of patients with ALD who ranged in age from early childhood to older than 40 years, investigators found a good correlation between Loes severity score on MRI and neurologic dysfunction [14]. That study differed from ours in a number of ways (other than addressing a disease other than Krabbe's disease). First, that study did not compare components of the Loes severity score with corresponding neurologic function scores, but only compared total Loes score with total neurologic score. Second, that study dealt primarily with adults rather than infants as in our study; the components of the neurologic examination and the detailed standardized assessment performed in our study differ substantially in the two patient populations. Another study showed that, in ALD patients undergoing umbilical cord stem cell transplantation, pretransplantation Loes scores were strong predictors of posttransplantation cognitive and motor outcomes [15].
Recently, the Loes severity scale has been adapted for assessment of Krabbe's disease. In one study, the authors found that MRI findings, as judged by the Loes scale, were valuable for distinguishing infants with early-onset Krabbe's disease from those with late-onset disease [8]. The data indicated that the Loes scores on the initial MR examinations for the two groups do not substantially differ, but disease progression patterns differed for the two groups. Infants with early-onset disease typically had initial involvement of the corticospinal tracts, deep gray matter nuclei (i.e., dentate nuclei, thalamus, and basal ganglia), and cerebellar WM, whereas infants with late onset usually had involvement of the corticospinal tracts, corpus callosum, and parietooccipital WM [8]. The Loes scale has also been used in another study for serial assessment of a small number of Krabbe's patients after stem cell transplantation [9]. The results of the scoring scale were reported in four patients with late-onset Krabbe's disease; mild improvement was seen in two children and no change in the other children. One limitation of the Loes scale in early-onset Krabbe's disease has been reported to be that the transient causes of abnormal WM signal intensity not directly related to WM demyelination (e.g., transependymal resorption of cerebrospinal fluid in hydrocephalus) can lead to spuriously high scores [8].
To our knowledge, the Loes severity scale has not yet been widely correlated with clinical features in a serial manner. In one study, investigators performed four types of electrophysiologic testing (visual evoked potentials, brainstem evoked potentials, electroencephalography, and testing of nerve conduction velocity) in children with Krabbe's disease [16]. Electrophysiologic test results were compared with scores on an MRI scoring scale derived from the Loes severity scale, using a 5-point scale in which the brain was divided into four regions (cerebral WM, basal ganglia, cerebellum, and brainstem) and an additional point could be given for atrophy. The investigators found that visual evoked potential and brainstem evoked potential abnormalities significantly correlated with Loes scores but that electroencephalography and nerve conduction velocity findings did not correlate in a statistically meaningful way with Loes scores.
A number of limitations were apparent in our study. We did not rate intra- or interoperator variability in the assessment of Loes scoring. Thus, this study cannot determine the degree of reproducibility of the scoring performed in our study. However, previous studies have reported that interobserver variability, at least with regard to ALD, is very low [13]. Second, limitations are imposed by the Loes scoring scale in that degrees of abnormality in a structure may not be accounted for beyond a point. For instance, a maximum score of 1 is assigned when definite signal abnormality is present in a structure. However, if the signal abnormality substantially worsens, the same score is still applied. Thus, the system does not adequately account for degrees of severity beyond a certain point. Third, ours is a retrospective study. These observations need to be confirmed and validated in a prospective clinical trial.
In summary, we found good correlation between clinical scores and Loes scores for the entire brain. Correlation between clinical scores and a specific corresponding brain region were also moderately good. These findings support the use of the Loes scoring system for evaluation of patients with Krabbe's disease.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
