Cognitive and Adaptive Outcome in Low-Grade Pediatric Cerebellar Astrocytomas: Evidence of Diminished Cognitive and Adaptive Functioning in
http://www.100md.com
《临床肿瘤学》
the Cincinnati Children's Hospital Medical Center and The University of Cincinnati School of Medicine, Cincinnati, OH
Mailman Center for Child Development
University of Miami School of Medicine, Miami, FL
University of California, San Diego, San Diego, CA
St Jude's Research Hospital, Memphis, TN
Children's Oncology Group, Arcadia, CA
New York University Medical Center, New York, NY
ABSTRACT
PURPOSE: Clinicians often assume that children with posterior fossa tumors are at minimal risk for cognitive or adaptive deficits if they do not undergo cranial irradiation. However, small case series have called that assumption into question, and have also suggested that nonirradiated cerebellar tumors can cause location-specific cognitive and adaptive impairment. This study (1) assessed whether resected but not irradiated pediatric cerebellar tumors are associated with cognitive and adaptive functioning deficits, and (2) examined the effect of tumor location and medical complications on cognitive and adaptive functioning.
PATIENTS AND METHODS: The sample was composed of 103 children aged 3 to 18 years with low-grade cerebellar astrocytomas, who underwent only surgical treatment as part of Children's Cancer Group protocol 9891 or Pediatric Oncology Group protocol 9130. The sample was divided into three groups based on primary tumor location: vermis, left hemisphere, or right hemisphere. Data were collected prospectively on intelligence, academic achievement, adaptive skills, behavioral functioning, and pre-, peri-, and postsurgical medical complications.
RESULTS: The sample as a whole displayed an elevated risk for cognitive and adaptive impairment that was not associated consistently with medical complications. Within this group of children with cerebellar tumors, tumor location had little effect on cognitive, adaptive, or medical outcome.
CONCLUSION: We did not replicate previous findings of location-specific effects on cognitive or adaptive outcome. However, the elevated risk of deficits in this population runs contrary to clinical lore, and suggests that clinicians should attend to the functional outcomes of children who undergo only surgical treatment for cerebellar tumors.
INTRODUCTION
Although the cerebellum has long been known to have a role in motor coordination and timing, there is an increasing appreciation of its role in higher-level cognition. Focal lesions of the cerebellar hemispheres affect cognitive functions generally associated with contralateral cortical tissues (eg, right cerebellar lesions affect language), and lesions of the vermis have been linked to behavioral alterations and executive dysfunction similar to those produced by disruption of frontal-subcortical circuits.1-4 Furthermore, acquired and congenital cerebellar abnormalities of childhood have been linked to a subsequent cascade of abnormal cortical development.5,6
This evidence of cerebellar involvement in higher cognitive functions and cortical development is particularly interesting to clinicians who work with children with brain tumors. Two thirds of pediatric brain tumors arise in the posterior fossa. As treatments have become more effective, more children with cerebellar tumors are reaching adulthood without further disease progression. Children with low-grade astrocytomas (which comprise roughly one third of all pediatric cerebellar tumors) have a 10- to 25-year disease-free survival rate exceeding 90% with radical resection.7,8 Medulloblastomas, which comprise half of the tumors in this region, are treated more aggressively because of their malignancy, with 5-year survival rates of 50% to 80% after resection, radiotherapy, and chemotherapy in standard-risk cases.8,9
Medulloblastoma survivors experience broad declines in cognitive functioning relative to their healthy peers, largely because of craniospinal irradiation.10,11 Low-grade cerebellar astrocytoma survivors typically have a more favorable outcome, and are considered by many neurosurgeons and oncologists to be at very low risk of long-term cognitive or adaptive deficits. However, recently published data suggest that some deficits after tumor treatments follow relatively defined neuroanatomical contours within the cerebellum (rather than the diffuse boundaries of conventional radiotherapy), and that such deficits may be present in children who do not receive radiotherapy. One report on seven children with mixed pathology and treatments associated right cerebellar tumors with verbal and literacy deficits and left cerebellar tumors with spatial deficits.12 Two more studies confirmed this finding in groups of 19 and 26 surgically treated children, and additionally observed behavioral and affective dysregulation after vermis tumors were excised.13,14 These findings are consistent with the disruption of cerebro-medullary-cerebellar "feed-forward" circuits and cerebello-thalamic-cerebral "feedback" circuits.1,2,15 A recent study reported somewhat different findings: most of the 24 children who had received only excision of posterior fossa tumors displayed cognitive deficits, with the greatest deficits across verbal, visual, and behavioral domains observed among those patients who displayed localization in the left cerebellar hemisphere or the vermis.16
Each of these four studies used small samples and a retrospective research design that introduces concerns about referral and selection biases. Heterogeneous tumor types were studied and, in one case, five of seven patients were tested after irradiation.12 In consideration of the low base rate of pediatric brain tumors, such design issues are understandable in single-center studies but they complicate interpretation and raise concerns regarding the generalizability of findings. They also prevent analysis of potential mediators of effect, including medical complications, which some studies found to predict cognitive morbidity.17-19 In this article, we report on a large sample of children who had received only surgical treatment (no radiotherapy or chemotherapy) for low-grade cerebellar astrocytomas as part of national multisite collaborative research projects through the former Children's Cancer Group (CCG) and the Pediatric Oncology Group (POG). Our first goal was to assess whether surgically treated but not irradiated cerebellar tumors were associated with functional morbidity. This was accomplished by comparing published norms to prospectively gathered cognitive and adaptive data for the entire sample. The second goal was to assess the effect of tumor location and medical complications on functional morbidity. This was accomplished by associating functional outcome with specific tumor sites within the cerebellum, and with pre-, peri-, and postsurgical complications.
PATIENTS AND METHODS
Sample
The sample was composed of 103 children aged 3 to 18 years who were enrolled in CCG protocol 9891 or POG protocol 9130, between 1991 and 1996. This sample included only children who had their tumors resected but received no antineoplastic chemotherapies or radiotherapy before psychological testing. Because the psychological tests included in the collaborative protocols have poor technical qualities in very young children, only children aged 3 years and older were considered for the study (12 children younger than 3 years were excluded). This article focuses on the subgroup that had primary cerebellar tumors without brainstem involvement (a report on the larger sample is in preparation). Two children with severe postsurgical motor deficits or other complications (eg, obtundation) were excluded. Of the 292 remaining children with cerebellar astrocytomas who met these criteria, 105 patients (36%) underwent psychological testing within the first year after surgery; 14 others were tested before surgery or more than a year after surgery, and 173 children went unevaluated for reasons that were not consistently documented, but which included a lack of referral or lack of an available psychologist at a given study site. The 105 tested patients were divided by tumor location into four subgroups, based on site data provided by the operating neurosurgeons and verified via central review of pre- and postsurgical neuroimaging conducted using standardized protocols: (1) cerebellar vermis (n = 51), primary tumor site in the vermis; right cerebellar hemisphere (n = 25), primary tumor site in right cerebellar hemisphere or peduncles; left cerebellar hemisphere (n = 27), primary tumor site in left cerebellar hemisphere or peduncles; unclassified (n = 2), tumors for which inadequate localization data were available for central review.
No tumor appeared to have substantially invaded both the vermis and a hemisphere. Because the unclassified group was small, we excluded it from further analyses, leaving 103 subjects.
Among these 103 patients, psychological testing occurred on average 108 days (standard deviation [SD] 78 days) after surgery (median, 96 days), at an average age of 8.5 years (SD, 4.1 years; median, 7.6 years). Girls comprised 54% of the sample, which was largely white (80%), African American (12%), or Hispanic (4%). Of the 82 cases for whom parent educational information was available, the median level of parent education was 14 years (2 years college or technical training after high school). The three analyzed groups did not differ significantly in age, sex, ethnicity, time since surgery, or parent education (P > .50).
Procedure
Eligibility was based on histopathologic evidence of cerebellar low-grade astrocytoma, oligodendroglioma, mixed glioma, or ganglioglioma according to prevailing WHO standards, and was overseen by central review.20 On initial diagnosis, families were presented with the option of enrolling onto the collaborative research protocols. Those families that declined to enroll were offered the standard of care at that time, which was determined by the treating physician but often corresponded roughly to the CCG 9891/POG 9130 protocol. As such, nonparticipation was rare. The research design and consent procedures were reviewed by institutional review boards at each participating institution and were in accordance with the 1964 Declaration of Helsinki.
Enrolled children underwent maximal tumor resection. Follow-up supportive care included neurologic examination as needed; neuroimaging at 0, 6, and 12 months after surgery; the possible use of glucocorticoids in the first 6 weeks after surgery; and anticonvulsants, antiemetics, or analgesics as needed. No antineoplastic chemotherapy or radiotherapy was used before collecting psychological and neurologic data. All children were judged to have had no disease progression when psychologically tested.
Data on medical complications were gathered at three time points, yielding presurgical, perisurgical, and postsurgical composite scores. The presurgical composite reflected the sum of the following factors, coded based on presurgical radiologic studies and physical examination just before the surgery (median, 1 day before surgery): hydrocephalus (present = 1; absent = 0), seizures (present = 1; absent = 0), and level of consciousness (normal = 0; lethargic or somnolent = 1). The perisurgical composite reflected the sum of the following complications, coded by the neurosurgeon at the treating institution: CNS infection, aseptic meningitis, hematoma, new neurologic symptom, CSF leak, pseudomeningocele, or other specific complication (each present = 1; absent = 0). The postsurgical composite reflected the sum of the following ratings, made by the neurosurgeon 1 week after surgery compared with the presurgical baseline: level of consciousness and neurologic deficits (in each, better = –1; unchanged = 0; worse = +1). Higher composite scores reflected greater medical involvement or severity.
Psychological test data were gathered prospectively within the first year after surgery. Intelligence was assessed with the age-appropriate Wechsler Scale,21-23 yielding a Verbal Intelligence Quotient score (IQ; reflecting language-based skills), a Performance IQ score (reflecting spatial and visual-motor skills), and a Full-Scale IQ score. Visuomotor skills were further assessed by the Beery Test of Visual-Motor Integration.24 Academic skills were screened with the Wide Range Achievement Test–Revised,25 yielding scores in Reading, Spelling, and Arithmetic. Parents were interviewed using the Vineland Adaptive Behavior Scales,26 which assessed the following functional domains: Communication, Daily Living Skills, and Socialization. Parents of children younger than 6 years also answered Vineland questions regarding the child's motor skills. Finally, parents were asked to complete the Achenbach Child Behavior Checklist,27 which yields an overall problem behavior composite, as well as indexes of "internalizing" symptoms (eg, depression, anxiety, somatic concern) and "externalizing" symptoms (eg, conduct problems).
Overview of Data Analysis
To assess for selection bias, we first used 2 analyses to compare demographic and medical characteristics of the 103 children included in analyses with the 189 children who were not included. We then focused on the present sample of 103 children, using single sample t tests to compare mean scores on the psychological tests to the expected mean based on published norms (100 [SD, 15] for all tests except the Achenbach, for which the mean is 50 [SD, 10]). As a complement to this group-average analysis, we also indexed the percent of children in our sample who scored below normal, defined as falling at or below the 25th percentile based on norms (ie, standard score 90; T-score 57). This was compared with the population base rate of 25% via the normal approximation to the binomial test. Finally, we conducted group comparisons on the psychological variables (using ANOVA tests) and neurologic composites (using 2 tests). Because of the differences in sample sizes across measures, we elected to run univariate analyses rather than multivariate analyses that would require list-wise deletion of missing cases. To maximize statistical power, we elected to retain a testwise alpha of .05.
RESULTS
Comparing children included in final analyses with those who were not, there were no significant differences in age, sex, ethnicity, parental education, or pre-, peri-, or postsurgical neurologic composite indexes (P > .05, for all). Thus, there was no indication of selection bias. There was also no indication that age, sex, or time between surgery and psychological testing correlated substantially with the psychological test results; only one of 45 correlations reached P < .05 levels of significance, well within expectations for chance events.
Mean psychological test scores across groups and for the entire sample are listed in Table 1. In the sample as a whole, mean scores were generally within the average range. However, the mean score for the sample was significantly worse than the true population mean at the P < .005 level in Full-Scale IQ, Performance IQ, Spelling, and the adaptive domains of Social, Motor Functioning, and Composite Functioning; at the P < .01 level in Arithmetic and the adaptive domain of Communication; and at the P < .05 level in Verbal IQ. Although the Achenbach Internalizing subscale also indicated significant impairment (P = .002), it should be noted that this scale measures both physical/somatic and mood symptoms, complicating interpretation in a medically ill population.28
Figure 1 compares the obtained rate of subnormal psychological test scores in the entire sample with the population rate of 25%. The sample displayed high rates of subnormal scores across nearly all indexes, with the effect extending beyond the P < .05 level on six indexes; the P values for four more indexes were between .05 and .06. Taken together, these data suggest that whereas many children who have undergone resection of a cerebellar astrocytoma fall within the average range of cognitive and adaptive functioning, there is an unusual risk for problems within this population.
When comparing children by tumor location across the psychological tests (Table 1), only one effect reached the P < .05 level of significance (Vineland Motor Scale, F = 4.6; P = .02). Given the number of analyses run, this effect may reflect type I error. However, as detailed in the Discussion section, this finding is consistent with current knowledge of the cerebellum.
The three groups did not differ on the pre-, peri-, or postsurgical composites (P > .10, for all). Moreover, exploratory analyses using Spearman's rho indicated that these composites were not associated with cognitive or adaptive outcome. Of 45 correlations (three composites x 15 outcome indexes), only two reached the P < .05 level. This is within expectations for chance events.
At the request of an anonymous reviewer, we also reran these analyses, correlating the individual components of each risk factor with selected outcome variables (Full-Scale IQ, Vineland Adaptive Behavior Composite, and Achenbach Total Score) using a slightly more stringent alpha cutoff of .01. No correlation reached statistical significance. At the reviewer's request, we also related surgical approach to outcome: 98% of the sample had a midline or paramedical approach; these approaches did not differ on cognitive outcome or the risk composites.
DISCUSSION
Two main findings emerged from this large prospective study of children who underwent exclusively surgical treatment for low-grade cerebellar astrocytomas. First, although many of these children displayed normal cognitive and adaptive functioning, as a group they displayed an elevated risk for cognitive and adaptive deficits. These deficits could not be predicted by indices of presurgical medical state, perisurgical complications, or short-term postsurgical observations. Second, these data failed to replicate previous studies that reported topographically specific effects on cognitive and adaptive outcomes. Primary tumor location in the vermis or in either cerebellar hemisphere generally did not predict outcome and was not associated with pre- and postsurgical state or perisurgical complications.
Examining our sample as a whole, mean scores on cognitive and adaptive measures fell within one SD of norms, but most index means were significantly below true average. Follow-up analyses indicated that our sample displayed an elevated rate of subnormal scores on nearly all indices, the effect reaching statistical significance in about half. Such widespread effects, which would be difficult to explain by chance variation, may in part be attributable to motor deficits. The largest effect sizes were found in measures that required motor responses: Performance IQ, Spelling, Adaptive Motor Skills, Full-Scale IQ, and overall adaptive behavior. However, the Beery test, a drawing task that requires fine motor coordination, failed to indicate marked deficits. Moreover, mean scores on several indices that are unrelated to motor functioning (eg, Verbal IQ and Adaptive Communication Skills) were also significantly poorer than population norms. Demographic factors cannot account for this apparent risk of functional morbidity. Our sample was demographically similar to the normative samples presented in the test manuals,21-27 though our sample had a somewhat greater representation of white families with higher parental education; if anything, population trends suggest that our sample should have displayed slightly above-average cognitive and adaptive skills.29 Given this result, the most reasonable conclusion is that even in the absence of radiotherapy or chemotherapy children who have undergone surgery for cerebellar astrocytomas are at elevated risk of poor cognitive and adaptive outcomes.
In our clinical practices, we have observed a widely held belief that children treated surgically for low-grade cerebellar astrocytomas are at effectively no risk of long-term functional deficits. Typically these children are not referred for follow-up psychological or neuropsychological evaluation. Present data, which suggest heightened risk of deficits that cannot be predicted by pre-, peri-, or short-term postsurgical observations, call such clinical beliefs and practices into question. Moreover, our data are not unusual; other authors have reported similar findings, though they have not called attention to them. For example, some have studied children with surgically treated glial tumors as control subjects against whom to compare children who received radiotherapy for medulloblastomas, reporting incidentally that these control children scored below normative IQ levels.11 Studies that directly compared children with surgically treated cerebellar astrocytomas to non-CNS–affected controls found that the astrocytoma group displayed lower Performance and Full-Scale IQ scores and visual-motor slowing.10,30 The weight of the data then suggests more careful attention to the functional outcomes of children who have undergone cerebellar tumor resection.
Surprisingly, the location of the tumor within the cerebellum was not associated with cognitive or adaptive outcome in our sample, with one exception: vermis tumors were associated with poorer reported motor functioning in children younger than 6 years. Although this finding might be attributed to type I ("false-positive") error, it fits the functional anatomy of the cerebellum. Whereas the cerebellar hemispheres control the initiation, planning, and timing of motor activities, the vermis and flocculonodular lobe (not dissociated from the vermis in this study) are involved in balance, motor control, and ongoing execution of movement.31 The portion of the Vineland Adaptive Behavior Scales that deals with motor functioning includes many items that require the latter skills, especially balance (eg, climbing on play equipment, hopping on one foot, riding a bicycle), that are not assessed by the other measures used here.
Our finding of a general lack of cognitive and adaptive differences across tumor location groups is at odds with previously published findings of topographically specific effects of cerebellar tumors. The discrepancies across studies may relate to methodologic differences. Most of the patients reported in the article by Scott et al12 had undergone radiotherapy, which is particularly noteworthy given the young age of their sample (younger than 5 years at treatment). None of the cases reviewed by Riva and Giorgi14 and Levisohn et al13 underwent radiotherapy. However, these samples were small (< 10 children per group), raising the possibility that idiosyncratic findings in a few patients could affect results markedly. Lazaroff and Castro-Sierra32 found no link between functional outcome and tumor location within the cerebellum, though mean scores were poorer than those of healthy controls and statistical tests were severely underpowered.
Our finding that the pre-, peri-, and postsurgical complication indices did not differ across tumor location groups was not surprising. Aside from transient mutism following surgical removal of vermis tumors (unfortunately, mutism was not specifically assessed in this protocol), complications have not been reported to follow tumor removal differentially across locations within the cerebellum. Overall, the rates of complications in the current study (eg, 83% had presurgical hydrocephalus, more than 70% had no perisurgical complications) were similar to those reported by other investigators studying similar samples.7,9 Some investigators have reported a relationship between pre- and perioperative factors and neurobehavioral outcome in pediatric posterior fossa tumors,17-19 while others have observed no such relationship.33-35 We found no such relationship, though present data do not address the rare cases in which postsurgical complications are so extreme as to preclude valid psychological testing (eg, postsurgical obtundation).
The multi-institutional nature of this research introduced several limitations that warrant mention. First among these limitations is the reliance on broad-scoped cognitive and adaptive measures, such as intelligence tests, that are familiar to all pediatric psychologists. There is a pressing need to pair multi-institutional work with more sophisticated neuropsychological tools that may be more sensitive to focal neurologic injury. Similarly, more detailed data on pre- and postsurgical neurologic status, including neurologic features at the time of psychological evaluation, would have been optimal. Second, in this multi-institutional protocol, it was difficult to ensure compliance with nonmedical aspects of the study protocol, particularly psychological testing. This resulted in uneven sample sizes across measures and scattered missing data, which prevented detailed analysis of test profiles by which subjects might be clustered. Inconsistent institutional compliance with the research protocol also diminished the total sample size. We maximized statistical power by retaining a liberal alpha level in consideration of the overall number of analyses, but some analyses still involved group sizes of fewer than 10 patients. Third, no data were obtained on why many prospective subjects did not undergo psychological testing, though it was our impression that most prospective subjects who did not have such testing were either not referred or unable to be seen due to a lack of institutional resources. The fact that many children did not undergo psychological testing raises the possibility of differential selection or referral, but preliminary analyses did not uncover evidence of systematic sampling bias based on medical or demographic factors. Finally, the copious attention to case-by-case detail that characterizes smaller single-institution case series is impossible to reproduce in large multi-institutional protocols. For example, though the specificity of reported tumor location data in the present study allowed us to subgroup patients by gross cerebellar structures (ie, vermis, hemisphere), these groups still covered structurally heterogeneous areas.
These limitations, as well as inconsistencies in findings across data sets, suggest that further research is warranted. Certainly, previous studies documenting topographic specificity of effect in patients with acquired and degenerative cerebellar lesions justify further investigation, especially considering their elegant link to the developing scientific understanding of cerebro-cerebellar circuitry.2,15 Moreover, the mechanisms by which our sample displayed increased risk of functional deficits relative to norms remain speculative. Small-scale studies of children with posterior fossa tumors have suggested that cognitive and adaptive performance is unchanged or even improved by surgical treatment.32,33 If this finding holds true in larger-scale studies, then it will be important to explore other factors, such as prolonged exposure to increased intracranial pressure and highly specified topographically determined effects that may have gone undetected with present methodology. In the meantime, present data clearly suggest that clinicians pay increased attention to the functional outcome of children who undergo neurosurgical treatment for cerebellar astrocytomas, even if these children show no unusual perisurgical complications and do not undergo adjuvant chemotherapy or radiotherapy.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Presented in part at the 2001 Spring meeting of the International Neuropsychological Society, Chicago, IL, February 14-17, 2001, and the 2002 Spring meeting of the International Neuropsychological Society, Toronto, Ontario, Canada, February 13-16, 2002.
Participating institutions, principal investigators, and grant numbers for all studies in this report are listed in the Appendix.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Schmahmann J: An Emerging Concept: The cerebellar contribution to higher function. Arch Neurol 48: 1178-1187, 1991
Schmahmann J, Sherman J: The cerebellar cognitive affective syndrome. Brain 121: 561-579, 1998
Hoffmann M, Schmitt F: Cognitive impairment in isolated subtentorial stroke. Acta Neurol Scand 109: 14-24, 2004
Salman MS: The cerebellum: It's about time! But timing is not everything–New insights into the role of the cerebellum in timing motor and cognitive tasks. J Child Neurol 17: 1-9, 2002
Dennis M, Hetherington CR, Spiegler BJ, et al: Functional consequences of congenital cerebellar dysmorphologies and acquired cerebellar lesions of childhood, in, Broman S, Fletcher JM, (eds): The changing nervous system: Neurobehavioral consequences of early brain disorders. New York, NY: Oxford University Press, 1999, pp 172-198
Akshoomoff N, Pierce K, Courchesne E: The neurobiological basis of autism from a developmental perspective. Dev Psychopathol 14: 613-634, 2002
Pencalet P, Maixner W, Sainte-Rose C, et al: Benign cerebellar astrocytomas in children. J Neurosurg 90: 265-273, 1999
Pollack I: Pediatric brain tumors. Semin Surg Oncol 16: 73-90, 1999
Modha A, Vassilyadi M, George A, et al: Medulloblastoma in children: The Ottawa experience. Childs Nerv Syst 16: 341-350, 2000
Riva D, Pantaleoni C, Milani N, et al: Impairment of neuropsychological functions in children with medulloblastomas and astrocytomas in the posterior fossa. Childs Nerv Syst 5: 107-110, 1989
Mulhern R, Renier D, Palmer S, et al: Neurocognitive deficits in medulloblastoma survivors and white matter loss. Ann Neurol 46: 834-841, 1999
Scott R, Stoodley CJ, Anslow P, et al: Lateralized cognitive deficits in children following cerebellar lesions. Dev Med Child Neurol 43: 685-691, 2001
Levisohn L, Cronin-Golomb A, Schmahmann J: Neuropsychological consequences of cerebellar tumor resection in children: Cerebellar cognitive affect syndrome in a pediatric population. Brain 123: 1041-1050, 2000
Riva D, Giorgi C: The cerebellum contributes to higher functions during development. Brain 123: 1051-1061, 2000
Schmahmann J, Pandya D: The cerebrocerebellar system. Int Rev Neurobiol 41: 31-60, 1997
Steinlin M, Imfeld S, Zulauf P, et al: Neuropsychological long-term sequelae after posterior fossa tumour resection during childhood. Brain 126: 1998-2008, 2003
Kao G, Goldwein J, Schultz D, et al: The impact of perioperative factors on subsequent intelligence quotient deficits in children treated for medulloblastoma/posterior fossa primitive neuroectodermal tumors. Cancer 74: 965-971, 1994
Chapman CA, Waber DP, Bernstein JH, et al: Neurobehavioral and neurologic outcome in long-term survivors of posterior fossa brain tumors: Role of age and perioperative factors. J Child Neurol 10: 209-212, 1995
Ater JL, Moore BD, Francis DJ, et al: Correlation of medical and neurosurgical events with neuropsychological status in children at diagnosis of astrocytoma: Utilization of a neurological severity score. J Child Neurol 11: 462-469, 1996
Rorke L, Gilles F, Davis R, et al: Revision of the World Health Organization classification of brain tumors for childhood brain tumors. Cancer 56: 1869-1886, 1985
Wechsler D: Manual for the Wechsler Intelligence Scale for Children—Revised. San Antonio, TX: The Psychological Corporation, 1974
Wechsler D: Manual for the Wechsler Adult Intelligence Scale—Revised. New York, NY, The Psychological Corporation, 1981
Wechsler D: WISC-III manual. San Antonio, TX, The Psychological Corporation, 1991
Beery K: Revised administration, scoring, and teaching manual for the Developmental Test of Visual-Motor Integration. Cleveland, OH, Modern Curriculum Press, 1982
Jastak JF, Wilkinson GS: Wide Range Achievement Test—Revised. Wilmington, DE, Jastak Assessment Systems, 1984
Sparrow SS, Balla DA, Cicchetti DV: Vineland Adaptive Behavior Scales Interview Edition Survey Form manual. Circle Pines, MN, American Guidance Service, 1984
Achenbach TM, Edelbrock C: Manual for the Child Behavior Checklist and Revised Child Behavior Profile. Burlington, VT, University of Vermont, 1981
Perrin EC, Stein REK, Drotar D: Cautions in using the Child Behavior Checklist: Observations based on research about children with a chronic illness. J Pediatr Psychol 16: 411-421, 1991
Vanderploeg RD, Schinka JA, Baum KM, et al: WISC-III premorbid prediction strategies: Demographic and best performance approaches. Psychol Assess 10: 277-284, 1998
Taphoorn MJB, Schiphorst AK, Snoek FJ, et al: Cognitive functions and quality of life in patients with low-grade gliomas: The impact of radiotherapy. Ann Neurol 36: 48-54, 1994
Kandel ER, Schwartz JH, Jessell TM: Principles of Neural Science. New York, NY, Elsevier, 1991
Lazaroff J, Castro-Sierra E: Preoperative and postoperative analysis of visual and auditory memory in children with cerebellar tumors. Childs Nerv Syst 12: 81-86, 1996
Copeland DR, Moore BD: Neuropsychological outcome among children treated at M.D. Anderson Cancer Center. Cancer Bull 44: 509-517, 1992
Packer R, Sutton L, Atkins T, et al: A prospective study of cognitive function in children receiving whole-brain radiotherapy and chemotherapy: Two-year results. J Neurosurg 70: 707-713, 1989
Grill J, Renaux VK, Bulteau C, et al: Long-term intellectual outcome in children with posterior fossa tumors according to radiation doses and volumes. Int J Radiat Oncol Biol Phys 45: 137-145, 1999(Dean W. Beebe, M. Douglas)
Mailman Center for Child Development
University of Miami School of Medicine, Miami, FL
University of California, San Diego, San Diego, CA
St Jude's Research Hospital, Memphis, TN
Children's Oncology Group, Arcadia, CA
New York University Medical Center, New York, NY
ABSTRACT
PURPOSE: Clinicians often assume that children with posterior fossa tumors are at minimal risk for cognitive or adaptive deficits if they do not undergo cranial irradiation. However, small case series have called that assumption into question, and have also suggested that nonirradiated cerebellar tumors can cause location-specific cognitive and adaptive impairment. This study (1) assessed whether resected but not irradiated pediatric cerebellar tumors are associated with cognitive and adaptive functioning deficits, and (2) examined the effect of tumor location and medical complications on cognitive and adaptive functioning.
PATIENTS AND METHODS: The sample was composed of 103 children aged 3 to 18 years with low-grade cerebellar astrocytomas, who underwent only surgical treatment as part of Children's Cancer Group protocol 9891 or Pediatric Oncology Group protocol 9130. The sample was divided into three groups based on primary tumor location: vermis, left hemisphere, or right hemisphere. Data were collected prospectively on intelligence, academic achievement, adaptive skills, behavioral functioning, and pre-, peri-, and postsurgical medical complications.
RESULTS: The sample as a whole displayed an elevated risk for cognitive and adaptive impairment that was not associated consistently with medical complications. Within this group of children with cerebellar tumors, tumor location had little effect on cognitive, adaptive, or medical outcome.
CONCLUSION: We did not replicate previous findings of location-specific effects on cognitive or adaptive outcome. However, the elevated risk of deficits in this population runs contrary to clinical lore, and suggests that clinicians should attend to the functional outcomes of children who undergo only surgical treatment for cerebellar tumors.
INTRODUCTION
Although the cerebellum has long been known to have a role in motor coordination and timing, there is an increasing appreciation of its role in higher-level cognition. Focal lesions of the cerebellar hemispheres affect cognitive functions generally associated with contralateral cortical tissues (eg, right cerebellar lesions affect language), and lesions of the vermis have been linked to behavioral alterations and executive dysfunction similar to those produced by disruption of frontal-subcortical circuits.1-4 Furthermore, acquired and congenital cerebellar abnormalities of childhood have been linked to a subsequent cascade of abnormal cortical development.5,6
This evidence of cerebellar involvement in higher cognitive functions and cortical development is particularly interesting to clinicians who work with children with brain tumors. Two thirds of pediatric brain tumors arise in the posterior fossa. As treatments have become more effective, more children with cerebellar tumors are reaching adulthood without further disease progression. Children with low-grade astrocytomas (which comprise roughly one third of all pediatric cerebellar tumors) have a 10- to 25-year disease-free survival rate exceeding 90% with radical resection.7,8 Medulloblastomas, which comprise half of the tumors in this region, are treated more aggressively because of their malignancy, with 5-year survival rates of 50% to 80% after resection, radiotherapy, and chemotherapy in standard-risk cases.8,9
Medulloblastoma survivors experience broad declines in cognitive functioning relative to their healthy peers, largely because of craniospinal irradiation.10,11 Low-grade cerebellar astrocytoma survivors typically have a more favorable outcome, and are considered by many neurosurgeons and oncologists to be at very low risk of long-term cognitive or adaptive deficits. However, recently published data suggest that some deficits after tumor treatments follow relatively defined neuroanatomical contours within the cerebellum (rather than the diffuse boundaries of conventional radiotherapy), and that such deficits may be present in children who do not receive radiotherapy. One report on seven children with mixed pathology and treatments associated right cerebellar tumors with verbal and literacy deficits and left cerebellar tumors with spatial deficits.12 Two more studies confirmed this finding in groups of 19 and 26 surgically treated children, and additionally observed behavioral and affective dysregulation after vermis tumors were excised.13,14 These findings are consistent with the disruption of cerebro-medullary-cerebellar "feed-forward" circuits and cerebello-thalamic-cerebral "feedback" circuits.1,2,15 A recent study reported somewhat different findings: most of the 24 children who had received only excision of posterior fossa tumors displayed cognitive deficits, with the greatest deficits across verbal, visual, and behavioral domains observed among those patients who displayed localization in the left cerebellar hemisphere or the vermis.16
Each of these four studies used small samples and a retrospective research design that introduces concerns about referral and selection biases. Heterogeneous tumor types were studied and, in one case, five of seven patients were tested after irradiation.12 In consideration of the low base rate of pediatric brain tumors, such design issues are understandable in single-center studies but they complicate interpretation and raise concerns regarding the generalizability of findings. They also prevent analysis of potential mediators of effect, including medical complications, which some studies found to predict cognitive morbidity.17-19 In this article, we report on a large sample of children who had received only surgical treatment (no radiotherapy or chemotherapy) for low-grade cerebellar astrocytomas as part of national multisite collaborative research projects through the former Children's Cancer Group (CCG) and the Pediatric Oncology Group (POG). Our first goal was to assess whether surgically treated but not irradiated cerebellar tumors were associated with functional morbidity. This was accomplished by comparing published norms to prospectively gathered cognitive and adaptive data for the entire sample. The second goal was to assess the effect of tumor location and medical complications on functional morbidity. This was accomplished by associating functional outcome with specific tumor sites within the cerebellum, and with pre-, peri-, and postsurgical complications.
PATIENTS AND METHODS
Sample
The sample was composed of 103 children aged 3 to 18 years who were enrolled in CCG protocol 9891 or POG protocol 9130, between 1991 and 1996. This sample included only children who had their tumors resected but received no antineoplastic chemotherapies or radiotherapy before psychological testing. Because the psychological tests included in the collaborative protocols have poor technical qualities in very young children, only children aged 3 years and older were considered for the study (12 children younger than 3 years were excluded). This article focuses on the subgroup that had primary cerebellar tumors without brainstem involvement (a report on the larger sample is in preparation). Two children with severe postsurgical motor deficits or other complications (eg, obtundation) were excluded. Of the 292 remaining children with cerebellar astrocytomas who met these criteria, 105 patients (36%) underwent psychological testing within the first year after surgery; 14 others were tested before surgery or more than a year after surgery, and 173 children went unevaluated for reasons that were not consistently documented, but which included a lack of referral or lack of an available psychologist at a given study site. The 105 tested patients were divided by tumor location into four subgroups, based on site data provided by the operating neurosurgeons and verified via central review of pre- and postsurgical neuroimaging conducted using standardized protocols: (1) cerebellar vermis (n = 51), primary tumor site in the vermis; right cerebellar hemisphere (n = 25), primary tumor site in right cerebellar hemisphere or peduncles; left cerebellar hemisphere (n = 27), primary tumor site in left cerebellar hemisphere or peduncles; unclassified (n = 2), tumors for which inadequate localization data were available for central review.
No tumor appeared to have substantially invaded both the vermis and a hemisphere. Because the unclassified group was small, we excluded it from further analyses, leaving 103 subjects.
Among these 103 patients, psychological testing occurred on average 108 days (standard deviation [SD] 78 days) after surgery (median, 96 days), at an average age of 8.5 years (SD, 4.1 years; median, 7.6 years). Girls comprised 54% of the sample, which was largely white (80%), African American (12%), or Hispanic (4%). Of the 82 cases for whom parent educational information was available, the median level of parent education was 14 years (2 years college or technical training after high school). The three analyzed groups did not differ significantly in age, sex, ethnicity, time since surgery, or parent education (P > .50).
Procedure
Eligibility was based on histopathologic evidence of cerebellar low-grade astrocytoma, oligodendroglioma, mixed glioma, or ganglioglioma according to prevailing WHO standards, and was overseen by central review.20 On initial diagnosis, families were presented with the option of enrolling onto the collaborative research protocols. Those families that declined to enroll were offered the standard of care at that time, which was determined by the treating physician but often corresponded roughly to the CCG 9891/POG 9130 protocol. As such, nonparticipation was rare. The research design and consent procedures were reviewed by institutional review boards at each participating institution and were in accordance with the 1964 Declaration of Helsinki.
Enrolled children underwent maximal tumor resection. Follow-up supportive care included neurologic examination as needed; neuroimaging at 0, 6, and 12 months after surgery; the possible use of glucocorticoids in the first 6 weeks after surgery; and anticonvulsants, antiemetics, or analgesics as needed. No antineoplastic chemotherapy or radiotherapy was used before collecting psychological and neurologic data. All children were judged to have had no disease progression when psychologically tested.
Data on medical complications were gathered at three time points, yielding presurgical, perisurgical, and postsurgical composite scores. The presurgical composite reflected the sum of the following factors, coded based on presurgical radiologic studies and physical examination just before the surgery (median, 1 day before surgery): hydrocephalus (present = 1; absent = 0), seizures (present = 1; absent = 0), and level of consciousness (normal = 0; lethargic or somnolent = 1). The perisurgical composite reflected the sum of the following complications, coded by the neurosurgeon at the treating institution: CNS infection, aseptic meningitis, hematoma, new neurologic symptom, CSF leak, pseudomeningocele, or other specific complication (each present = 1; absent = 0). The postsurgical composite reflected the sum of the following ratings, made by the neurosurgeon 1 week after surgery compared with the presurgical baseline: level of consciousness and neurologic deficits (in each, better = –1; unchanged = 0; worse = +1). Higher composite scores reflected greater medical involvement or severity.
Psychological test data were gathered prospectively within the first year after surgery. Intelligence was assessed with the age-appropriate Wechsler Scale,21-23 yielding a Verbal Intelligence Quotient score (IQ; reflecting language-based skills), a Performance IQ score (reflecting spatial and visual-motor skills), and a Full-Scale IQ score. Visuomotor skills were further assessed by the Beery Test of Visual-Motor Integration.24 Academic skills were screened with the Wide Range Achievement Test–Revised,25 yielding scores in Reading, Spelling, and Arithmetic. Parents were interviewed using the Vineland Adaptive Behavior Scales,26 which assessed the following functional domains: Communication, Daily Living Skills, and Socialization. Parents of children younger than 6 years also answered Vineland questions regarding the child's motor skills. Finally, parents were asked to complete the Achenbach Child Behavior Checklist,27 which yields an overall problem behavior composite, as well as indexes of "internalizing" symptoms (eg, depression, anxiety, somatic concern) and "externalizing" symptoms (eg, conduct problems).
Overview of Data Analysis
To assess for selection bias, we first used 2 analyses to compare demographic and medical characteristics of the 103 children included in analyses with the 189 children who were not included. We then focused on the present sample of 103 children, using single sample t tests to compare mean scores on the psychological tests to the expected mean based on published norms (100 [SD, 15] for all tests except the Achenbach, for which the mean is 50 [SD, 10]). As a complement to this group-average analysis, we also indexed the percent of children in our sample who scored below normal, defined as falling at or below the 25th percentile based on norms (ie, standard score 90; T-score 57). This was compared with the population base rate of 25% via the normal approximation to the binomial test. Finally, we conducted group comparisons on the psychological variables (using ANOVA tests) and neurologic composites (using 2 tests). Because of the differences in sample sizes across measures, we elected to run univariate analyses rather than multivariate analyses that would require list-wise deletion of missing cases. To maximize statistical power, we elected to retain a testwise alpha of .05.
RESULTS
Comparing children included in final analyses with those who were not, there were no significant differences in age, sex, ethnicity, parental education, or pre-, peri-, or postsurgical neurologic composite indexes (P > .05, for all). Thus, there was no indication of selection bias. There was also no indication that age, sex, or time between surgery and psychological testing correlated substantially with the psychological test results; only one of 45 correlations reached P < .05 levels of significance, well within expectations for chance events.
Mean psychological test scores across groups and for the entire sample are listed in Table 1. In the sample as a whole, mean scores were generally within the average range. However, the mean score for the sample was significantly worse than the true population mean at the P < .005 level in Full-Scale IQ, Performance IQ, Spelling, and the adaptive domains of Social, Motor Functioning, and Composite Functioning; at the P < .01 level in Arithmetic and the adaptive domain of Communication; and at the P < .05 level in Verbal IQ. Although the Achenbach Internalizing subscale also indicated significant impairment (P = .002), it should be noted that this scale measures both physical/somatic and mood symptoms, complicating interpretation in a medically ill population.28
Figure 1 compares the obtained rate of subnormal psychological test scores in the entire sample with the population rate of 25%. The sample displayed high rates of subnormal scores across nearly all indexes, with the effect extending beyond the P < .05 level on six indexes; the P values for four more indexes were between .05 and .06. Taken together, these data suggest that whereas many children who have undergone resection of a cerebellar astrocytoma fall within the average range of cognitive and adaptive functioning, there is an unusual risk for problems within this population.
When comparing children by tumor location across the psychological tests (Table 1), only one effect reached the P < .05 level of significance (Vineland Motor Scale, F = 4.6; P = .02). Given the number of analyses run, this effect may reflect type I error. However, as detailed in the Discussion section, this finding is consistent with current knowledge of the cerebellum.
The three groups did not differ on the pre-, peri-, or postsurgical composites (P > .10, for all). Moreover, exploratory analyses using Spearman's rho indicated that these composites were not associated with cognitive or adaptive outcome. Of 45 correlations (three composites x 15 outcome indexes), only two reached the P < .05 level. This is within expectations for chance events.
At the request of an anonymous reviewer, we also reran these analyses, correlating the individual components of each risk factor with selected outcome variables (Full-Scale IQ, Vineland Adaptive Behavior Composite, and Achenbach Total Score) using a slightly more stringent alpha cutoff of .01. No correlation reached statistical significance. At the reviewer's request, we also related surgical approach to outcome: 98% of the sample had a midline or paramedical approach; these approaches did not differ on cognitive outcome or the risk composites.
DISCUSSION
Two main findings emerged from this large prospective study of children who underwent exclusively surgical treatment for low-grade cerebellar astrocytomas. First, although many of these children displayed normal cognitive and adaptive functioning, as a group they displayed an elevated risk for cognitive and adaptive deficits. These deficits could not be predicted by indices of presurgical medical state, perisurgical complications, or short-term postsurgical observations. Second, these data failed to replicate previous studies that reported topographically specific effects on cognitive and adaptive outcomes. Primary tumor location in the vermis or in either cerebellar hemisphere generally did not predict outcome and was not associated with pre- and postsurgical state or perisurgical complications.
Examining our sample as a whole, mean scores on cognitive and adaptive measures fell within one SD of norms, but most index means were significantly below true average. Follow-up analyses indicated that our sample displayed an elevated rate of subnormal scores on nearly all indices, the effect reaching statistical significance in about half. Such widespread effects, which would be difficult to explain by chance variation, may in part be attributable to motor deficits. The largest effect sizes were found in measures that required motor responses: Performance IQ, Spelling, Adaptive Motor Skills, Full-Scale IQ, and overall adaptive behavior. However, the Beery test, a drawing task that requires fine motor coordination, failed to indicate marked deficits. Moreover, mean scores on several indices that are unrelated to motor functioning (eg, Verbal IQ and Adaptive Communication Skills) were also significantly poorer than population norms. Demographic factors cannot account for this apparent risk of functional morbidity. Our sample was demographically similar to the normative samples presented in the test manuals,21-27 though our sample had a somewhat greater representation of white families with higher parental education; if anything, population trends suggest that our sample should have displayed slightly above-average cognitive and adaptive skills.29 Given this result, the most reasonable conclusion is that even in the absence of radiotherapy or chemotherapy children who have undergone surgery for cerebellar astrocytomas are at elevated risk of poor cognitive and adaptive outcomes.
In our clinical practices, we have observed a widely held belief that children treated surgically for low-grade cerebellar astrocytomas are at effectively no risk of long-term functional deficits. Typically these children are not referred for follow-up psychological or neuropsychological evaluation. Present data, which suggest heightened risk of deficits that cannot be predicted by pre-, peri-, or short-term postsurgical observations, call such clinical beliefs and practices into question. Moreover, our data are not unusual; other authors have reported similar findings, though they have not called attention to them. For example, some have studied children with surgically treated glial tumors as control subjects against whom to compare children who received radiotherapy for medulloblastomas, reporting incidentally that these control children scored below normative IQ levels.11 Studies that directly compared children with surgically treated cerebellar astrocytomas to non-CNS–affected controls found that the astrocytoma group displayed lower Performance and Full-Scale IQ scores and visual-motor slowing.10,30 The weight of the data then suggests more careful attention to the functional outcomes of children who have undergone cerebellar tumor resection.
Surprisingly, the location of the tumor within the cerebellum was not associated with cognitive or adaptive outcome in our sample, with one exception: vermis tumors were associated with poorer reported motor functioning in children younger than 6 years. Although this finding might be attributed to type I ("false-positive") error, it fits the functional anatomy of the cerebellum. Whereas the cerebellar hemispheres control the initiation, planning, and timing of motor activities, the vermis and flocculonodular lobe (not dissociated from the vermis in this study) are involved in balance, motor control, and ongoing execution of movement.31 The portion of the Vineland Adaptive Behavior Scales that deals with motor functioning includes many items that require the latter skills, especially balance (eg, climbing on play equipment, hopping on one foot, riding a bicycle), that are not assessed by the other measures used here.
Our finding of a general lack of cognitive and adaptive differences across tumor location groups is at odds with previously published findings of topographically specific effects of cerebellar tumors. The discrepancies across studies may relate to methodologic differences. Most of the patients reported in the article by Scott et al12 had undergone radiotherapy, which is particularly noteworthy given the young age of their sample (younger than 5 years at treatment). None of the cases reviewed by Riva and Giorgi14 and Levisohn et al13 underwent radiotherapy. However, these samples were small (< 10 children per group), raising the possibility that idiosyncratic findings in a few patients could affect results markedly. Lazaroff and Castro-Sierra32 found no link between functional outcome and tumor location within the cerebellum, though mean scores were poorer than those of healthy controls and statistical tests were severely underpowered.
Our finding that the pre-, peri-, and postsurgical complication indices did not differ across tumor location groups was not surprising. Aside from transient mutism following surgical removal of vermis tumors (unfortunately, mutism was not specifically assessed in this protocol), complications have not been reported to follow tumor removal differentially across locations within the cerebellum. Overall, the rates of complications in the current study (eg, 83% had presurgical hydrocephalus, more than 70% had no perisurgical complications) were similar to those reported by other investigators studying similar samples.7,9 Some investigators have reported a relationship between pre- and perioperative factors and neurobehavioral outcome in pediatric posterior fossa tumors,17-19 while others have observed no such relationship.33-35 We found no such relationship, though present data do not address the rare cases in which postsurgical complications are so extreme as to preclude valid psychological testing (eg, postsurgical obtundation).
The multi-institutional nature of this research introduced several limitations that warrant mention. First among these limitations is the reliance on broad-scoped cognitive and adaptive measures, such as intelligence tests, that are familiar to all pediatric psychologists. There is a pressing need to pair multi-institutional work with more sophisticated neuropsychological tools that may be more sensitive to focal neurologic injury. Similarly, more detailed data on pre- and postsurgical neurologic status, including neurologic features at the time of psychological evaluation, would have been optimal. Second, in this multi-institutional protocol, it was difficult to ensure compliance with nonmedical aspects of the study protocol, particularly psychological testing. This resulted in uneven sample sizes across measures and scattered missing data, which prevented detailed analysis of test profiles by which subjects might be clustered. Inconsistent institutional compliance with the research protocol also diminished the total sample size. We maximized statistical power by retaining a liberal alpha level in consideration of the overall number of analyses, but some analyses still involved group sizes of fewer than 10 patients. Third, no data were obtained on why many prospective subjects did not undergo psychological testing, though it was our impression that most prospective subjects who did not have such testing were either not referred or unable to be seen due to a lack of institutional resources. The fact that many children did not undergo psychological testing raises the possibility of differential selection or referral, but preliminary analyses did not uncover evidence of systematic sampling bias based on medical or demographic factors. Finally, the copious attention to case-by-case detail that characterizes smaller single-institution case series is impossible to reproduce in large multi-institutional protocols. For example, though the specificity of reported tumor location data in the present study allowed us to subgroup patients by gross cerebellar structures (ie, vermis, hemisphere), these groups still covered structurally heterogeneous areas.
These limitations, as well as inconsistencies in findings across data sets, suggest that further research is warranted. Certainly, previous studies documenting topographic specificity of effect in patients with acquired and degenerative cerebellar lesions justify further investigation, especially considering their elegant link to the developing scientific understanding of cerebro-cerebellar circuitry.2,15 Moreover, the mechanisms by which our sample displayed increased risk of functional deficits relative to norms remain speculative. Small-scale studies of children with posterior fossa tumors have suggested that cognitive and adaptive performance is unchanged or even improved by surgical treatment.32,33 If this finding holds true in larger-scale studies, then it will be important to explore other factors, such as prolonged exposure to increased intracranial pressure and highly specified topographically determined effects that may have gone undetected with present methodology. In the meantime, present data clearly suggest that clinicians pay increased attention to the functional outcome of children who undergo neurosurgical treatment for cerebellar astrocytomas, even if these children show no unusual perisurgical complications and do not undergo adjuvant chemotherapy or radiotherapy.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Presented in part at the 2001 Spring meeting of the International Neuropsychological Society, Chicago, IL, February 14-17, 2001, and the 2002 Spring meeting of the International Neuropsychological Society, Toronto, Ontario, Canada, February 13-16, 2002.
Participating institutions, principal investigators, and grant numbers for all studies in this report are listed in the Appendix.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Schmahmann J: An Emerging Concept: The cerebellar contribution to higher function. Arch Neurol 48: 1178-1187, 1991
Schmahmann J, Sherman J: The cerebellar cognitive affective syndrome. Brain 121: 561-579, 1998
Hoffmann M, Schmitt F: Cognitive impairment in isolated subtentorial stroke. Acta Neurol Scand 109: 14-24, 2004
Salman MS: The cerebellum: It's about time! But timing is not everything–New insights into the role of the cerebellum in timing motor and cognitive tasks. J Child Neurol 17: 1-9, 2002
Dennis M, Hetherington CR, Spiegler BJ, et al: Functional consequences of congenital cerebellar dysmorphologies and acquired cerebellar lesions of childhood, in, Broman S, Fletcher JM, (eds): The changing nervous system: Neurobehavioral consequences of early brain disorders. New York, NY: Oxford University Press, 1999, pp 172-198
Akshoomoff N, Pierce K, Courchesne E: The neurobiological basis of autism from a developmental perspective. Dev Psychopathol 14: 613-634, 2002
Pencalet P, Maixner W, Sainte-Rose C, et al: Benign cerebellar astrocytomas in children. J Neurosurg 90: 265-273, 1999
Pollack I: Pediatric brain tumors. Semin Surg Oncol 16: 73-90, 1999
Modha A, Vassilyadi M, George A, et al: Medulloblastoma in children: The Ottawa experience. Childs Nerv Syst 16: 341-350, 2000
Riva D, Pantaleoni C, Milani N, et al: Impairment of neuropsychological functions in children with medulloblastomas and astrocytomas in the posterior fossa. Childs Nerv Syst 5: 107-110, 1989
Mulhern R, Renier D, Palmer S, et al: Neurocognitive deficits in medulloblastoma survivors and white matter loss. Ann Neurol 46: 834-841, 1999
Scott R, Stoodley CJ, Anslow P, et al: Lateralized cognitive deficits in children following cerebellar lesions. Dev Med Child Neurol 43: 685-691, 2001
Levisohn L, Cronin-Golomb A, Schmahmann J: Neuropsychological consequences of cerebellar tumor resection in children: Cerebellar cognitive affect syndrome in a pediatric population. Brain 123: 1041-1050, 2000
Riva D, Giorgi C: The cerebellum contributes to higher functions during development. Brain 123: 1051-1061, 2000
Schmahmann J, Pandya D: The cerebrocerebellar system. Int Rev Neurobiol 41: 31-60, 1997
Steinlin M, Imfeld S, Zulauf P, et al: Neuropsychological long-term sequelae after posterior fossa tumour resection during childhood. Brain 126: 1998-2008, 2003
Kao G, Goldwein J, Schultz D, et al: The impact of perioperative factors on subsequent intelligence quotient deficits in children treated for medulloblastoma/posterior fossa primitive neuroectodermal tumors. Cancer 74: 965-971, 1994
Chapman CA, Waber DP, Bernstein JH, et al: Neurobehavioral and neurologic outcome in long-term survivors of posterior fossa brain tumors: Role of age and perioperative factors. J Child Neurol 10: 209-212, 1995
Ater JL, Moore BD, Francis DJ, et al: Correlation of medical and neurosurgical events with neuropsychological status in children at diagnosis of astrocytoma: Utilization of a neurological severity score. J Child Neurol 11: 462-469, 1996
Rorke L, Gilles F, Davis R, et al: Revision of the World Health Organization classification of brain tumors for childhood brain tumors. Cancer 56: 1869-1886, 1985
Wechsler D: Manual for the Wechsler Intelligence Scale for Children—Revised. San Antonio, TX: The Psychological Corporation, 1974
Wechsler D: Manual for the Wechsler Adult Intelligence Scale—Revised. New York, NY, The Psychological Corporation, 1981
Wechsler D: WISC-III manual. San Antonio, TX, The Psychological Corporation, 1991
Beery K: Revised administration, scoring, and teaching manual for the Developmental Test of Visual-Motor Integration. Cleveland, OH, Modern Curriculum Press, 1982
Jastak JF, Wilkinson GS: Wide Range Achievement Test—Revised. Wilmington, DE, Jastak Assessment Systems, 1984
Sparrow SS, Balla DA, Cicchetti DV: Vineland Adaptive Behavior Scales Interview Edition Survey Form manual. Circle Pines, MN, American Guidance Service, 1984
Achenbach TM, Edelbrock C: Manual for the Child Behavior Checklist and Revised Child Behavior Profile. Burlington, VT, University of Vermont, 1981
Perrin EC, Stein REK, Drotar D: Cautions in using the Child Behavior Checklist: Observations based on research about children with a chronic illness. J Pediatr Psychol 16: 411-421, 1991
Vanderploeg RD, Schinka JA, Baum KM, et al: WISC-III premorbid prediction strategies: Demographic and best performance approaches. Psychol Assess 10: 277-284, 1998
Taphoorn MJB, Schiphorst AK, Snoek FJ, et al: Cognitive functions and quality of life in patients with low-grade gliomas: The impact of radiotherapy. Ann Neurol 36: 48-54, 1994
Kandel ER, Schwartz JH, Jessell TM: Principles of Neural Science. New York, NY, Elsevier, 1991
Lazaroff J, Castro-Sierra E: Preoperative and postoperative analysis of visual and auditory memory in children with cerebellar tumors. Childs Nerv Syst 12: 81-86, 1996
Copeland DR, Moore BD: Neuropsychological outcome among children treated at M.D. Anderson Cancer Center. Cancer Bull 44: 509-517, 1992
Packer R, Sutton L, Atkins T, et al: A prospective study of cognitive function in children receiving whole-brain radiotherapy and chemotherapy: Two-year results. J Neurosurg 70: 707-713, 1989
Grill J, Renaux VK, Bulteau C, et al: Long-term intellectual outcome in children with posterior fossa tumors according to radiation doses and volumes. Int J Radiat Oncol Biol Phys 45: 137-145, 1999(Dean W. Beebe, M. Douglas)