Treatment of childhood central nervous system (CNS) tumors can lead to sensorineural hearing loss (SNHL), with prior research indicating associations between SNHL and cognitive difficulties. Infants (0-3 years) treated for CNS tumors are at particular risk for neurocognitive deficits due to increased vulnerability of the developing brain and missed developmental opportunities secondary to prolonged treatment. This study expands upon existing research by examining the association between treatment-related SNHL and later neurocognitive outcomes among infants.

Serial audiology and neurocognitive assessments were conducted as part of a prospective, multisite, longitudinal trial (SJYC07). Children with newly diagnosed CNS tumors were treated with chemotherapy, with or without focal proton or photon radiation therapy (RT). SNHL was dichotomized based on hearing in the better ear as present versus not present (Chang grade ≥1a vs. <1a). Neurocognitive assessments included intellectual functioning (IQ), and parent ratings of executive functioning and behavioral functioning. Demographic and clinical variables investigated included: sex, age at diagnosis (years), treatment type (chemotherapy only vs. chemotherapy + RT), risk group (low vs. intermediate vs. high), and socioeconomic status (SES, continuous). Logistic regression models were used to identify factors associated with SNHL. Change point longitudinal models were used to examine the effect of each covariate individually and the potential impact of SNHL on trajectories of neurocognitive outcomes.

Of 135 patients (median age at diagnosis= 1.5 years), 67% had mild-to-severe SNHL as defined by Chang grade ≥1a at last follow-up. SNHL occurred early after treatment with a 1-year cumulative incidence 63.0% ±4.3%. SNHL was associated with age at diagnosis (p <.001) but not sex, treatment exposure or study risk arm (p >.10). At pretreatment baseline, IQ was associated with age at diagnosis (older age= higher IQ) and SES (higher SES= higher IQ) with a change in the trajectory of IQ after SNHL (stable prior to SNHL and declined 1.46 points/year after SNHL), which was impacted by tumor location (patients with supratentorial tumors stable prior to SNHL and declined 2.84 points/year after SNHL; whereas, patients with infratentorial tumors increased 1.93 points/year prior to SNHL and were stable after SNHL). At pre-treatment baseline, adaptive functioning was associated with age at diagnosis (older age= higher skills) with a change in adaptive functioning after SNHL that varied by age. There was a change in trajectory of attention problems (stable before SNHL and worsening 1.39 points/year after SNHL). SNHL was not associated with parent report of emerging executive functioning.

Children with brain tumors experience SNHL and cognitive difficulties early in treatment that can worsen over time. Younger age at diagnosis is associated with greater risk for SNHL and cognitive difficulties. Analyses of the time course between the emergence of SNHL and cognitive late effects suggests even mild SNHL is associated with a clinically signficant decline in IQ and attention problems. These findings have notable implications with respect to refining monitoring guidelines, informing modifications to treatment, advocating for interventions, and helping educate parents, teachers, and providers about the significant impact of mild SNHL.