Location May Not Affect IQ and Adaptive Outcome in Pediatric Cerebellar Tumors D.W. Beebe, M.D. Ris, and E. Holmes

1
Location May Not Affect IQ and Adaptive
Outcomein Pediatric Cerebellar TumorsD.W.
Beebe, M.D. Ris, and E. Holmes

• Three recent studies suggested that pediatric
  cerebellar tumors result in neuropsychological
  impairment, even in cases of surgical treatment
  without radiation or chemotherapy. These studies
  further suggested that tumor location (vermis v-
  hemispheric) affects outcome. However, sample
  sizes have been small (ns7, 19 26) and
  referral biases may have affected the results.
  The present study examines the intellectual and
  adaptive outcomes of 96 children who underwent
  surgical resection for low-grade cerebellar
  astrocytomas as part of national collaborative
  research projects (CCG 9891/POG 9130). Children
  were classified based upon tumor location vermal
  (n37), mixed vermal-hemispheric (19), right
  hemispheric (22), and left hemispheric (18). Data
  on intellectual ability, academic skills, and
  adaptive functioning were prospectively gathered
  within one year of surgery, as were medical
  records of pre-, peri-, and post-surgical medical
  complications. Despite demographics that would
  favor positive functioning (81 White, median
  parent ed.14 yrs), mean scores were
  significantly below average on most measures of
  intellectual, academic and adaptive outcome (10
  of 15 pslt.05, 6 of 15 pslt.005). However,
  intellectual, academic, and adaptive outcomes
  were not associated with tumor location (13 of 15
  psgt.05, no plt.01), nor were these variables
  associated with pre-, peri-, or post-surgical
  complications (44 of 45 psgt.05, no plt.01). Thus,
  despite a large sample size and evidence of
  impairment, we did not find an association
  between cerebellar tumor location or
  complications and intellectual and adaptive
  outcome. However, definitive conclusions await
  further large-scale studies involving specialized
  neuropsychological measures.

• Medical complication measures
• Pre-surgical Tally of pre-surgical presence of
  hydrocephalus, seizures, and lethargy /
  somnolence.
• Peri-surgical Tally of surgical complications,
  including CNS infection, aseptic meningitis,
  hematoma, pseudomeningocele.
• Post-surgical Composite of surgeons ratings of
  change in level of consciousness and neurological
  deficits one week post-surgery.
• Psychological measures gathered within the first
  year (112 77 days) post-surgery
• Age-appropriate Wechsler Intelligence Scale
• Beery Developmental Test of Visual Motor
  Integration (VMI)
• Age-appropriate Wide Range Achievement Test
  (WRAT-R)
• Vineland Adaptive Behavior Scales, Survey Form
• Achenbach Child Behavior Checklist Parent Form
  (CBCL)

• IN THE SAMPLE AS A WHOLE
• There was no evidence of referral bias. Children
  who were tested did not differ from those who
  were eligible but not tested in age, sex,
  ethnicity, parent education, or the three medical
  composites (psgt.05).
• There was an increased risk of cognitive and
  adaptive problems.
• Mean scores were significantly lower than
  population norms on 10 of the 15 indexes (Figure
  1). In contrast, the expected IQ is 105 for a
  sample of this demographic background
  (Vanderploeg et al, 1998).
• Figure 2 shows the rate at which children in the
  sample scored below average as compared to a
  population base rate of 25. Nonparametric tests
  confirm elevated risk of below-average
  performance on 7 of 15 indexes.
• There was no significant correlation between the
  psychological test data and the pre-, peri-, or
  post-surgical composites. Of 45 bivariate
  nonparametric correlations (15 tests X 3
  composites), only 1 reached the plt.05 level.
  This is at expectation for chance events.
• COMPARISONS ACROSS TUMOR LOCATION YIELDED
• No demographic differences. There was no
  difference across groups in age, time since
  surgery, sex, ethnicity, or parent education
  (psgt.05).
• No differences in pre-, peri-, or post-surgical
  complications. All comparisons of these
  composite scores across groups had psgt.05.
• Table 1 The four groups differed on only 2
  (related) indexes of psychological functioning.
  Motor functioning and overall adaptive skills
  were reported to be lower in the vermis group
  than the others.

• Beyond simple motor functioning, the cerebellum
  has a role in the quality of language, attention,
  and spatial processing (Dennis, 1999).
• Three studies of pediatric cerebellar tumors
  (Levisohn et al., 2000 Riva Giorgi, 2000
  Scott et al., 2001) have suggested
• Verbal deficits with right cerebellar lesions
• Spatial deficits with left cerebellar lesions
• Affective/behavior dysregulation or mutism with
  vermis lesions
• These deficits seem to fit with known
  cortico-pontine-cerebellar and cerebello-thalamic-
  cortico systems (Schmahmann Sherman, 1998).
• However, the three studies were small (ns7, 19,
  26), relied on restrospective review of
  referred children (introducing possible referral
  biases), assessed children with heterogeneous
  tumor types.
• We examined whether tumor location within the
  cerebellum affects cognitive and adaptive outcome
  in a large prospectively-gathered data set in
  which children received only surgical treatment
  for low-grade astrocytomas.

• More detailed neuropsych tests may reveal subtle
  or specific areas of deficit that were not
  evident on the more gross measures used here.
• There is a pressing need to integrate such
  sophisticated assessments into multi-institutional
  research, and to minimize attrition.
• We need to understand the mechanism by which
  these children are more susceptible to low scores
  on IQ and adaptive measures.
• We need to assess the long-term functioning of
  these children , as deficits may not be evident
  in first year post-treatment.
• It is important to consider the possible effect
  of tumor and surgical excision when interpreting
  data on samples with more complex treatments
  (i.e., chemotherapy and radiotherapy).

• 96 children treated for low-grade cerebellar
  astrocytomas
• Enrolled in national collaborative research study
  (CCG9891/POG9130)
• Surgical resection, but no antineoplastic
  radiotherapy or chemotherapy
• Age 8.4 4.1 years (range 3 - 18 years)
• 54 girls, 81 Caucasian, mean parental education
  of 14 years
• Divided into four groups based upon primary tumor
  site
• 1. Cerebellar vermis (n37 not hemisphere,
  peduncle, or brain stem)
• 2. Right cerebellar hemisphere (n22 not vermis
  or brain stem)
• 3. Left cerebellar hemisphere (n18 not vermis
  or brain stem)
• 4. Mixed vermis-hemisphere (n19 vermis
  hemisphere/peduncle)