QOL-13. Impact of hearing loss on neuropsychological functioning in children treated for medulloblastoma: A report from the Children’s Oncology Group (COG)

Abstract

BACKGROUND/OBJECTIVE: We prospectively examined neuropsychological outcomes and ototoxicity in children with average-risk medulloblastoma. METHODS: Eligible patients included those treated on COG protocol ACNS0331 who completed audiograms at end of therapy or one-year off-therapy, and neuropsychological assessments between 2- and 5-years post-diagnosis. Conventional pure-tone audiometric evaluations (0.25-8kHz) were assigned an ototoxicity grade based on the International Society of Pediatric Oncology (SIOP) grading scale. Grade for the better hearing ear was used for analyses. Participants were divided into two groups: SIOP grade≥3 hearing loss (HL) versus SIOP grade<3. Cutoff score of 60 on BASC-2 was used to dichotomize parent-reported anxiety and depression scores as ‘low’ or ‘high’. RESULTS: Data were available for 113 children (66% male; 86% white), aged 3.0-18.5 at diagnosis (Mean=9.1). One-quarter (24.8%, n=28) had at least moderate HL (≥ SIOP grade 2), and 12.3% (n=14) had severe HL (≥ SIOP grade 3). After controlling for radiation exposure and age, children with severe HL showed significantly higher levels of anxiety (OR=5.9, 95%CI 1.3-26.0, p=0.0195) and borderline differences in depression (OR=4.0, 95%CI 1.0-16.5, p=0.0563), but no differences in cognitive functioning when compared to other participants. When moderate and severe HL were combined in exploratory analyses, significantly greater anxiety (OR=9.0, 95%CI 2.1-37.4, p=.0027) and depression (OR=4.6, 95%CI 1.3-15.7, p=.0165) were observed. CONCLUSIONS: Survivors of pediatric medulloblastoma with moderate to severe HL evidenced greater psychosocial, but not neurocognitive, difficulties compared to those with no or mild HL. It may be that modern treatment protocols generally preserve cognitive functioning such that associations between HL and cognitive impairment are no longer significant. It is also possible that neurocognitive risk associated with HL may not manifest until survivors are further from diagnosis. In contrast, survivors with HL may be at greater risk for negative psychosocial adjustment, suggesting that increased monitoring of mental health outcomes is warranted.