Pediatric brain tumor (PBT) survivors are at risk for speech (e.g., articulation, prosody, fluency) and language (e.g., vocabulary, grammar, narratives, pragmatics) difficulties (Hodges et al., 2020). It is important to understand what treatment and/or demographic factors are associated with language functioning soon after diagnosis, and what factors are associated with language functioning years after treatment completion. This study characterizes longitudinal language functioning for clinically referred PBT survivors diagnosed in early childhood.

Participants were 48 PBT patients (54% supratentorial, 6% disseminated), 21% with NF-1, who were diagnosed by age 6 (M = 43.2 months, SD 24.5) and received tumor-directed intervention including surgery (85%), chemotherapy (69%), and/or radiation therapy (50%). Hearing concerns existed for 29% of the patients. Age at first neuropsychological evaluation was 2-15 years (M=7.6, SD=3.63), age at second neuropsychological evaluation was 5-19 years (M=12.04, SD=3.86), with an average of 4.42 years (SD=2.37) between evaluations. Patients were 63% male, 77% White, 94% non-Hispanic, and fluent English speakers. Verbal IQ, working memory, fluencies, comprehension, memory, and parent-reported functional communication outcomes were assessed as part of comprehensive batteries. Rates of weak performance (1 SD<Mean) were compared to the expected base rate of 16%.

Group means significantly diverged from age-expected performance by the second evaluation in all domains except semantic fluency. Weakness was identified on at least 1 verbal subtest for 79% of the sample at the first evaluation, and for 85% of the sample at the second evaluation. As a group, patients showed a significant increase in the number of weaknesses identified on performance-based measures from the first to second evaluation [t(47) = -3.60, p <.001]. Over half of the sample showed an increase in the rate of verbal weaknesses identified (56.3%). Those with more weaknesses over time had lower IQ at the initial evaluation [t(36) =-2.61, p=.013]. An increase in the number of weaknesses from first to second evaluation was not associated with tumor type/location, treatment modality, or demographic variables.

Brain tumor diagnosis in early childhood during rapid language development is associated with language impairments soon after diagnosis, and years after treatment completion. Causes for continued and increased impairment are multifactorial and risk cannot clearly be identified by demographic and treatment variables alone. Any early language weakness identification should signal need for intervention as the causes for difficulty are complex and these weaknesses are likely to persist and increase over time.