Long-term cognitive outcomes after cerebral sinovenous thrombosis in childhood
Abstract
Aim
To assess long-term cognitive function in children after cerebral sinovenous thrombosis (CSVT).
Method
Children with CSVT, who had neuropsychological testing for intellectual ability, executive function, attention, language, or behavior, were included in a prospective observational study. Outcomes were compared with normative means using one-sample t-tests. Predictors of abnormal function were examined using logistic regression.
Results
Fifty children with CSVT were included (median age at diagnosis 2y 10mo, interquartile range 7d–6y 10mo; 35 males, 15 females). The median follow-up time was 4 years 2 months (interquartile range 2y 8mo–6y 4mo). Compared with normative means, children with CSVT had lower mean (± standard deviation) full-scale IQ, working memory, and processing speed scores (93.3±16, p=0.01; 93.6±16, p=0.04; 93.7±15.3, p=0.02 respectively). They also had lower scores in executive function, attention, and language domains. Refractory seizure at presentation was associated with a trend in behavioral problems (odds ratio [OR] 6.3, 95% confidence interval [CI] 0.9–46, p=0.07). Females were less likely to experience processing speed difficulties (OR 0.22, 95% CI 0.04–1.3, p=0.09). Incomplete recanalization was associated with a greater risk of abnormal verbal comprehension (OR 5.3, 95% CI 0.93–30.5, p=0.059).
Interpretation
Children with CSVT as a group performed below age expectations on standardized neuropsychological tests, although there was variability across individuals and cognitive domains. Larger studies are needed to evaluate predictors of cognitive deficits in children with CSVT.
Abbreviations
-
- BRIEF
-
- Behavior Rating Inventory of Executive Function
-
- CSVT
-
- Cerebral sinovenous thrombosis
What this paper adds
- Most children with cerebral sinovenous thrombosis (CSVT) exhibited emerging cognitive deficits.
- Children with CSVT had greater difficulties in intellectual, behavioral, executive, attention, and memory functions.
Cerebral sinovenous thrombosis (CSVT) in childhood is a rare but potentially serious condition with incidence rates ranging between 0.4 and 0.7 per 100 000 children a year.1 Early diagnosis of CSVT is challenging because of the non-specific neurological signs and symptoms. Advances in neurovascular imaging have led to increased recognition of this condition and reinforced the need for timely diagnosis and management, which is crucial for improved long-term outcomes.2 Risk factors for CSVT are age-dependent, with infections, systemic illnesses, and maternal factors being among the most common in neonates, while head/neck infections, chronic illnesses such as congenital heart disease, nephrotic syndrome, and autoimmune disorders are the most common in older children.2
Outcome studies in childhood CSVT have reported a broad range (sensorimotor, speech, visual, global developmental delay) of residual neurological deficits.3-5 However, few studies have described outcomes in greater depth, particularly from the cognitive and neurodevelopmental perspective. Although some studies have described normal intellectual outcomes after CSVT,6, 7 others have reported mild to severe cognitive impairments.8, 9 Nevertheless, all previous studies are limited not just by their small sample sizes but also by short duration of follow-up. Long-term follow-up is essential for more accurate outcome determination, as it takes time for emergence of cognitive deficits, particularly subtle ones, after early-life brain injury.
In our current study, we sought to evaluate the long-term cognitive outcomes and potential predictors in a prospectively enrolled cohort of children with CSVT. We hypothesized that most children with CSVT would exhibit impaired cognitive function on at least one domain (intellectual abilities, executive function, attention, learning abilities, or memory) compared with the general pediatric population. In addition, the presence of refractory seizures at presentation, abnormal neurological function, incomplete recanalization, or parenchymal brain injury may predict worse cognitive outcomes.
METHOD
Patient population
We conducted a single-center study including term neonates (0–28d) and children (29d to 18y) with radiologically confirmed CSVT prospectively enrolled into the Canadian Pediatric Ischemic Stroke Registry database in the Children’s Stroke Program at The Hospital for Sick Children, Toronto, Canada, between June 1996 and September 2017. Patients were included if they had formal neuropsychological testing after CSVT. Exclusion criteria included the presence of pre-existing or concurrent neurological conditions such as meningitis, arterial ischemic stroke, hypoxic–ischemic encephalopathy, brain tumors, and head injury at the time of CSVT diagnosis. This study was approved by the Research Ethics Boards at the Hospital for Sick Children. Written patient or parental informed consents were obtained from all study participants.
Clinical data
Demographic and clinical information at presentation and subsequent follow-up were obtained from registry data and supplemented with standardized retrospective chart review. Predictor variables of long-term cognitive outcomes selected a priori included age at presentation (neonate vs non-neonate), sex, refractory (requiring two or more anticonvulsant drugs to control) seizures at presentation, and neurological function (assessed using the Pediatric Stroke Outcome Measure). Active epilepsy and the use of antiseizure medication at the time of neuropsychological testing were also assessed. Head circumference was evaluated using centiles adjusted for age and sex as per published charts.10 ‘Normal’ size was considered between the 3rd and 97th centile; ‘microcephaly’ and ‘macrocephaly’ were defined as a head circumference not greater than the 3rd or at least the 97th centiles respectively. The Pediatric Stroke Outcome Measure is a standardized and validated neurological examination tool that includes five subscales (each scored from 0 to 2): right sensorimotor, left sensorimotor, language production, language comprehension, and cognitive/behavioral.11 Children with a maximum score of 0 to 0.5 on one or more of the Measure’s subscales were classified as having normal neurological function, otherwise they were considered to have abnormal function.12
Radiographic data
Magnetic resonance imaging/computed tomography scans were reviewed retrospectively by the study neurologists (ASA, MM). Investigators reviewed CSVT characteristics (superficial vessel vs deep vessel thrombosis) and the presence of parenchymal brain lesions and their characteristics (infarction, hemorrhage, location, extent).
Neuropsychological testing
Age-appropriate tests for evaluating intellectual ability were used. Preschool children completed the Wechsler Preschool and Primary Scale of Intelligence, Third Edition. This provides scores for overall intellectual ability (full-scale IQ), verbal and non-verbal abilities with a mean score of 100, and a standard deviation (SD) of 15. Scores no greater than 85 were indicative of abnormal function. School-aged children and older adolescents completed the Wechsler Intelligence Scale for Children, Fourth Edition or the Wechsler Adult Intelligence Scale, Fourth Edition. Both scales provide index scores for overall intellectual ability (full-scale IQ), verbal ability (Verbal Comprehension Index), non-verbal ability (Perceptual Reasoning Index), auditory attention and mental manipulation (Working Memory Index), and visual–motor speed (Processing Speed Index).13, 14
We assessed executive function using the Behavior Rating Inventory of Executive Function (BRIEF), a questionnaire designed to evaluate executive function as observed by parents and teachers. It is subdivided into the Behavioral Regulation Index and Metacognition Index, which constitute a global executive composite. Normative population scores have a mean of 50 and a SD of 10.15 Scores of at least 60 were indicative of abnormal function.
We used the Behavior Assessment System for Children parent form, Second Edition, which is designed to measure adaptive and maladaptive behaviors and self-perceptions of children.16 The Behavioral Symptoms Index is calculated using the externalizing problems composite, internalizing problems composite, and the adaptive skills composite scores. Normative population scores have a mean of 50 and a SD of 10. Scores of at least 60 were indicative of abnormal function.
To evaluate attention, we used the Test for Everyday Attention for Children; specifically, the test battery included the sky search subtest (evaluates focused visual attention and processing speed) and the score subtest (evaluates sustained auditory attention). Normative population scores within its subtests and overall test scores have a mean of 10 and a SD of 3.17 Scores no greater than 7 were indicative of abnormal function.
We evaluated verbal learning and memory using the California Verbal Learning Test for Children. This instrument assesses the child’s ability to learn, retain, and retrieve a long list of words read aloud by the examiner over multiple trials.18 We focused on list A total trials 1 to 5, which has a normative mean of 50 and a SD of 10 (scores not more than 40 were indicative of abnormal function), list B and long delay free recall, which both have normative means of 0 and SD of 1 (scores not more than −1 were indicative of abnormal function).
We studied the visual–motor integration using the Beery-Buktenica Developmental Test of Visuomotor Integration. This instrument has been widely used to detect visual–motor delays in school-aged children. The visual–motor integration provides a measure of the child’s skills on three subsets: visual–motor, visual–perceptual, and motor coordination. Normative population scores have a mean of 100 and SD of 15.19 Scores not more than 85 were indicative of abnormal function.
Information on whether the child had an individualized education plan at any time throughout the study period was obtained. An individualized education plan is typically prepared by a multidisciplinary team of teachers, school administrators, a school board psychologist, and the child’s parents. The plan describes the child’s needs and the services provided by the school to meet those needs.
Statistical analysis
Demographic, clinical, and radiographic data were examined using descriptive statistics. Continuous variables were presented as mean and SD or median and interquartile range (IQR), as appropriate. Qualitative variables were presented using frequency distributions and proportions. Cognitive outcome measures were compared between our cohort of patients with CSVT and the normative mean of each standardized test using one-sample t-tests or a Wilcoxon signed rank test for one sample, as appropriate.
Children with scores falling 1 SD away from the mean (toward the abnormal end of each scale) were classified as having impaired performance on the corresponding subscale. For each of the cognitive outcomes, we assessed the demographic, clinical, and radiographic predictors of impaired performance using univariable logistic regression models. Values of p<0.05 were considered statistically significant. Statistical analyses were conducted using SAS University Edition (SAS Institute, Cary, NC, USA).
RESULTS
Study cohort
A total of 77 children diagnosed with CSVT and who had neuropsychological testing were identified in the Canadian Pediatric Ischemic Stroke Registry, of whom 50 were eligible for our study. Reasons for exclusion included comorbid, pre-existing epilepsy (n=4), arterial ischemic stroke (n=4), hypoxic–ischemic encephalopathy (n=6), periventricular venous infarction (n=3), or other pre-existent neurological comorbidities (n=10).
Clinical and radiographic features
Demographic, clinical, and radiographic features of our study sample are summarized in Table 1. Median age at index CSVT event was 2 years 10 months (IQR 7d–6y 10mo). Two-thirds of children were non-neonates. Most were males (n=35; 70%). Twenty-one had seizures at presentation; eight of these were refractory seizures. Five out of the eight patients went on to have chronic epilepsy requiring long-term antiseizure medication, which were being administered at the time of neuropsychological assessment. Only one participant who did not experience refractory seizures at presentation (1 out of 13) went on to develop chronic epilepsy on follow-up and was on antiseizure medication at the time of neuropsychological assessment. Nine children had abnormal general neurological function after CSVT diagnosis at last clinical follow-up as evaluated using the Pediatric Stroke Outcome Measure. In terms of radiographic features, 40% had deep venous system involvement and 82% had multiple vessels involved. Parenchymal brain lesions and intracranial hemorrhage were present in 50% and 30% of study participants respectively. Compared with older children, neonates were more likely to have seizures at presentation (29%), deep venous structures involvement (53%), parenchymal brain lesions (71%), and intracranial hemorrhage (41%).
| Total, n=50 | Neonates, n=17 | Children, n=33 | |
|---|---|---|---|
| Demographic and neurological characteristics | |||
| Males, n (%) | 35 (70) | 15 (88.2) | 20 (60.6) |
| Median age at CSVT, y:mo (IQR) | 2:10 (7d–6:10) | 1d (0–7d) | 5:2 (3:1–7:10) |
| Median age at the time of initial neuropsychological testing, y:mo (IQR) | 7:0 (4:10–10:1) | 6:4 (4:7–7:0) | 8:1 (4:10–11:2) |
| Median time lag between CSVT onset and first neuropsychological assessment, y:mo (IQR) | 4:2 (2:8–6:4) | 6:4 (4:7–7:0) | 3:0 (2:0–4:4) |
| Refractory seizures at presentation, n (%) | 8 (16) | 5 (29.4) | 3 (9.1) |
| Chronic epilepsya, n (%) | 6 (12) | 4 (66.7) | 2 (33.3) |
| Abnormal neurological function, n (%) | 9 (18.8) | 4 (23.5) | 5 (16.1) |
| Radiological features | |||
| Depth of vessel involvement, n (%) | |||
| Superficial | 30 (60) | 8 (47) | 22 (66.7) |
| Deep | 4 (8) | 2 (11.8) | 2 (6.1) |
| Superficial and deep | 16 (32) | 7 (41.2) | 9 (27.3) |
| Presence of infarction, n (%) | 25 (50) | 12 (70.6) | 13 (39.4) |
| Infarction with white matter involvement | 22 (88) | 11 (91.7) | 11 (84.6) |
| Infarction with thalamus, deep grey matter involvement | 10 (40) | 3 (25) | 7 (53.9) |
| Presence of intracranial hemorrhage, n (%) | 15 (30) | 7 (41.2) | 8 (24.2) |
- a Receiving antiseizure medication at the time of neuropsychological testing. CSVT, cerebral sinovenous thrombosis; IQR, interquartile range.
Neuropsychological assessment
Overall, 50 children underwent formal neuropsychological testing (Fig. S1, online supporting information). The complete battery of neuropsychological tests was administered to 15 children. No differences were found between those who completed the seven tests (n=15) and those who did not (n=35) except for the age at initial neuropsychological testing (Table S1, online supporting information). Intellectual ability, full-scale IQ, working memory, and processing speed scores among children with CSVT as a group were lower than the normative means (Table 2), whereas verbal comprehension performance and perceptual reasoning fell within normal range. With respect to verbal learning and memory performance on the California Verbal Learning Test for Children, children with CSVT demonstrated weaker ability to learn and retain information; scores were lower than the normative expectations on list A (46.2±11, p<0.001), list B (−0.76±1.1, p=0.001), and long delay free recall (−0.55±1.2, p=0.02).
| Mean±SD | n | p | Children with impaired performance, n (%) | |
|---|---|---|---|---|
| Intellectual ability (standard scores) | ≤85 (1 SD) | |||
| Mean 100, SD 15 | ||||
| FSIQ | 93.3±15.8 | 39 | 0.011 | 10 (25.6) |
| Verbal Comprehension Index | 97±17.5 | 37 | 0.310 | 10 (27) |
| Perceptual Reasoning Index | 95±15.4 | 33 | 0.070 | 11 (33.3) |
| Working Memory Index | 93.6±16.1 | 29 | 0.043 | 9 (31) |
| Processing Speed Index | 93.7±15.3 | 33 | 0.024 | 13 (39.4) |
| Behavior Assessment System for Children | ≥60 (1 SD) | |||
| Mean 50, SD 10 | ||||
| Externalizing problems | 49.4±11.1 | 38 | 0.330a | 5 (13.2) |
| Internalizing problems | 52.8±13.7 | 37 | 0.579a | 8 (21.6) |
| Adaptive skills | 49.6±11.8 | 39 | 0.830 | 8 (20.5) |
| Behavioral Symptoms Index | 52.1±13.9 | 36 | 0.811a | 9 (25) |
| BRIEF (t-scores) – parent form | ≥60 (1 SD) | |||
| Mean 50, SD 10 | ||||
| Behavioral Regulation Index | 52.4±13.6 | 29 | 0.566a | 9 (31) |
| Metacognition Index | 56.9±13.4 | 28 | 0.011 | 11 (39.3) |
| Global Executive Composite | 55.4±13.9 | 28 | 0.050 | 9 (32) |
| BRIEF (t-scores) – teacher form | ||||
| Mean 50, SD 10 | ||||
| Behavioral Regulation Index | 56.9±14.9 | 22 | 0.099a | 5 (22.7) |
| Metacognition Index | 61.9±12.5 | 21 | <0.001 | 11 (50) |
| Global Executive Composite | 66.6±25.3 | 20 | 0.001 | 10 (50) |
| Test for Everyday Attention for Children | ≤7 (1 SD) | |||
| Mean 10, SD 3 | ||||
| Sky Search subtest attention | 8.8±2.9 | 26 | <0.001 | 7 (26.9) |
| Sky Search subtest targets | 8.8±2.9 | 25 | <0.001 | 7 (28) |
| Score! | 8.9±3.9 | 25 | <0.001 | 10 (40) |
| California Verbal Learning Test for Children | ≤40 (1 SD) | |||
| Mean 50, SD 10 | ||||
| T1–5 t-score | 46.2±11 | 28 | <0.001 | 8 (28.6) |
| Mean 0, SD 1 | ≤−1 (1 SD) | |||
| Long delay free recall z-score | –0.55±1.2 | 27 | 0.019 | 13 (48.2) |
| Short delay free recall z-score | –0.57±1.4 | 27 | 0.044 | 11 (40.7) |
| List B z-score | –0.76±1.1 | 27 | 0.001 | 12 (44.4) |
| Beery-Buktenica Developmental Test | ≤85 (1 SD) | |||
| Mean 100, SD 15 | ||||
| Visual–motor integration | 90.9±18.4 | 41 | <0.001a | 12 (29.3) |
| Motor coordination | 84.8±20.7 | 33 | <0.001a | 15 (45.5) |
| Visual perception | 92.7±21.9 | 35 | 0.071a | 11 (31.4) |
- a Wilcoxon signed rank test for one sample. SD, standard deviation; FSIQ: full-scale IQ; BRIEF, Behavior Rating Inventory of Executive Function.
Metacognition and global executive function scores were lower among children with CSVT as shown on both BRIEF parent and teacher reports. On the other hand, behavioral regulation index of the BRIEF and the Behavioral Symptoms Index evaluated using the Behavior Assessment System for Children, Second Edition were comparable to normative means. In addition, our study findings revealed that children with CSVT experience relative difficulties in their capacity to maintain visual and sustained auditory attention, as shown with the decreased Test for Everyday Attention for Children subset scores. Moreover, scores were lower for the CSVT group than for the normative sample on the three subtests of the Beery-Buktenica Developmental Test of Visuomotor Integration.
Prevalence of cognitive impairments
As shown in Table 2, a significant proportion of children with CSVT displayed impaired performance on various domains of the different cognitive measures. As a group, a normal cognitive profile was seen in only 20% of our patients. Most of our study sample performed less than expected compared with the normative population in at least one of the cognitive domains (80%), while 6% of our cohort performed below normative standards on all cognitive measures for which they were tested. The most prevalent impairments included processing speed of the intellectual ability (40%), metacognition of the BRIEF parent and the teacher form (39% and 50% respectively), global executive composite of the BRIEF teacher form (50%), short and long delay free recall and list B of the California Verbal Learning Test for Children (41%, 48%, 44% respectively), auditory attention of the Test for Everyday Attention for Children (40%), and the motor coordination of the Beery-Buktenica Developmental Test (46%). Thirty-four out of 45 children (75.5%) had an individualized education plan (missing data in five).
Predictors of impaired cognitive performance
None of the clinical and demographic factors predicted impaired performance across the different domains of the cognitive measures used in our study (Table S2, online supporting information). Nevertheless, we detected some signals for some of these factors. For instance, females were less likely to demonstrate impaired Processing Speed Index scores (odds ratio [OR] 0.22; 95% confidence interval [CI] 0.04–1.27, p=0.09) and they generally did slightly better on the different domains that were assessed; these findings were maintained for most of the domains even after excluding neonates most of whom were male (Table S3, online supporting information). The presence of refractory seizures at presentation is linked to poor Behavioral Symptoms Index scores (OR 6.3; 95% CI 0.85–46.1, p=0.07). The use of antiseizure medication at the time of assessment did not correlate with cognitive impairments. Abnormal neurological function (Pediatric Stroke Outcome Measure) after CSVT onset was associated with possibly impaired full-scale IQ (OR 4; 95% CI 0.77–20.8, p=0.09). At the level of radiographic factors, complete recanalization is probably associated with poor Verbal Comprehension Index outcomes (OR 5.3; 95% CI 0.93–30.5, p=0.06). Head circumference did not correlate with intellectual abilities and Behavior Assessment System for Children outcomes (Table S4, online supporting information).
DISCUSSION
This study, which aimed to examine the long-term cognitive outcomes in children after CSVT, showed that this population has a heterogeneous cognitive profile with significant differences compared with age-appropriate norms. The vast majority exhibited impaired cognitive performance in one or more of the assessed domains.
Our patient cohort showed lower overall intellectual ability compared with normative means. Our findings corroborated results documented in previous studies in which lower scores on IQ testing after neonatal and non-neonatal CSVT were found.9, 20 Interestingly, within the subtests of IQ testing, the Verbal Comprehension Index of our cohort did not differ compared with normative means. This finding has also been seen in other studies evaluating long-term neuropsychological outcome after a cerebrovascular insult.21, 22 Moreover, sparing of verbal ability after cortical lesions continues to be consistent across multiple studies irrespective of type of lesion, laterality, and age when insult occurred.23 A postulate for the sparing of language function comes from evidence of magnetoencephalography studies documenting early bilateral language representation that lateralize with maturity. Sparing of language function after insult is due to compensation of language by other intact cortical networks.24
At the level of the behavioral function, children with CSVT did not show elevated rates of externalizing problems, internalizing problems, or behavioral problems relative to the normative sample, and their adaptive skills were also rated as being age-appropriate. To our knowledge, behavioral function has not been investigated to date in children with CSVT. In a case series of four children with pediatric and lymphoma who developed CSVT during therapy, cognitive assessment showed the absence of any significant behavioral difficulties.25 On the other hand, our patient population exhibited behavioral manifestations of executive dysfunction in everyday life, as shown on the global executive composite reported by both parents and teachers. Parent- and teacher-reported BRIEF provide pertinent information about the functioning of children in everyday life, which may not be otherwise accessible through direct testing situations.26
Furthermore, children with CSVT displayed greater difficulties in verbal learning and memory as evaluated using the California Verbal Learning Test for Children. Compared with the normative scores, they had significantly lower scores on the overall index of verbal learning (list A), ability to learn competing information (list B), as well as long and short long delay recall recognition. These findings highlight the difficulties encountered in our sample of children with CSVT in encoding (learning fewer words), semantic clustering, and retrieval after long delay. In a cohort of 26 children with a history of stroke (hemorrhagic or ischemic), similar outcomes of impaired verbal and learning memory were reported; children with stroke experienced greater difficulties in organizing words into semantic clusters for better recall than typically developing comparison individuals.27 Performance on tests of attention, particularly sustained attention, was notably weaker in our sample of children with CSVT compared with the normative population. Recent research from functional magnetic resonance imaging studies of population norm volunteers highlighted that engaging in tasks requiring sustained attention activates multiple simultaneous cortical and subcortical areas bilaterally.28 The phenomenon of integrated network function that allows sustained attention may be susceptible to dysfunction in our patients, probably because of the disruption of localized brain modules leading to overall dysfunctional diffuse networks.29
Of interest, males tended to have slightly lower scores on multiple cognitive domains. These results were maintained for most of the domains after the exclusion of neonates, except for some intellectual abilities and behavioral regulation. Similar findings have been seen in other studies in children with brain injury (arterial ischemic stroke, hypoxic ischemic encephalopathy, intracranial hemorrhage, intraventricular hemorrhage), whereby males exhibited greater difficulties on neuropsychological testing.21 A possible explanation that can account for these differences is the increased lateralization for cognitive tasks in the male brain, with more widespread activation of female brains.30 A resultant disruption would render male brains more susceptible to adverse cognitive outcome owing to a lesser availability of diffuse cortical networks compared with their female counterparts for complex cognitive tasks. This study emphasizes the need to examine sex differences when studying cognitive outcome after CSVT.
We found no differences in cognitive outcomes associated with age at CSVT presentation (neonate vs non-neonate), neurological comorbidity, presence of seizure disorder, number of dural vessels involved (except for adaptive skills), and presence of parenchymal injury (infarct or hemorrhage). The lack of differences in cognitive outcomes in those with parenchymal injury contradicted our hypothesis of greater likelihood for poor outcomes in those who experience such injury. In addition, these results differed from previous studies, which reported poorer cognitive outcomes in those patients with parenchymal injury irrespective of the cause.21, 31
Our current study had several limitations. The availability of neuropsychological assessment in a subset of participants limited our ability to detect statistically significant predictors of poor long-term cognitive outcomes. It also prevented us from conducting multivariable analysis and testing for the effects of interactions between the different predictors. Nevertheless, our current study includes a relatively large cohort of children with CSVT and is the first to provide access to the long-term outcomes in this patient population. Furthermore, the lack of information on the specific rehabilitation programs (type, duration, and frequency) received by the study participants constitutes another limitation as these interventions represent an important confounding factor for our study findings. Nevertheless, our findings will be useful for hypothesis generation in future prospective clinical studies in children with CSVT.
It is important to note that, although significant declines were seen in our cohort compared with normative means, the overall cognitive outcome of our population was on average in a mild range. As such, children with CSVT are at risk for emerging cognitive deficits that may be classified as ‘normal’ on routine neurological assessment in a busy outpatient clinic setting, but they may eventually have subtle abnormalities detectable only by thorough neuropsychological testing. By examining the proportion of children with CSVT who had actual academic difficulties at public schools, it turned out that 75% of them received an individualized education plan compared with only 9% of the general population in the province of Ontario.32 This suggests that the use of adjustments to the academic curriculum was higher in our cohort than in the general population. However, this observation certainly needs more rigorous interrogation in future studies.
In summary, we found evidence of emerging cognitive deficits with comprehensive neuropsychological testing among children with CSVT, which highlights the importance of long-term outcome assessment. They tended to experience greater difficulties in intellectual, behavioral, executive, attention, and memory functions compared with typically developing, age-matched peers. Our study hints at the potential role of some clinical and radiographic factors in predicting long-term cognitive outcomes. Future studies with a larger sample size are, however, needed to confirm our findings and to delineate further robust predictors of outcome. Finally, given the resource constraints in healthcare in general and the fact that neuropsychological testing is not universally available in all centers, future studies of predictors should delve deeper into how patients can be screened for formal neuropsychological testing as part of the routine neurological clinical assessment.
Acknowledgments
The Auxilium Foundation funded this work. The authors have stated that they had no interests that might be perceived as posing conflict or bias.
