Neurocognitive outcomes following 30.6-Gy whole-ventricular radiotherapy with 54-Gy total dose focal tumor bed boost for CNS non-germinomatous germ cell tumors: A Children’s Oncology Group study (ACNS1123).

Abstract

10036 Background: COG ACNS1123 stratum 1 was a prospective, phase II trial treating children with localized non-germinomatous germ cell tumors of the brain with chemotherapy followed by whole ventricular (WV) radiotherapy (RT, 30.6 Gy) and a focal tumor bed boost (54 Gy total dose) Given the potential adverse impact of RT on survivors’ neurocognitive development, ACNS1123 had a co-primary objective to evaluate and longitudinally model cognitive and behavioral outcomes. Methods: Cognition was prospectively examined at 9, 30, and 60-months post-diagnosis via COG protocol ALTE07C1. Attention/concentration, processing speed (PS), and estimated intelligence quotient (est. IQ) were assessed by Wechsler Intelligence tests. Survey-reported attention and executive functioning (EF) were also obtained. Multivariable linear mixed-effects models examined trends over time, as well as associations and interactions with age, gender, tumor location (pineal gland versus other) and insurance type (private versus other and private versus public). Results: Seventy patients were evaluable and received WVRT followed by focal RT boost. Mean age at initiation of RT was 11.8 ± 4.3 years; 74% were male. A total of 56, 60, and 61 patients had at least one valid assessment score across 3 time points for attention/concentration, est. IQ and PS, respectively. Nine, 20 and 20 patients had data at all 3 time points for attention, est. IQ and PS, respectively. The average PS scores fell below the normative mean at all three time points. The average Est.IQ was below the normative mean at 9 months only. Males had higher attention scores on average (p < .01), while patients with other insurance had lower attention compared to the private group (p < .01). For the model-estimated longitudinal trajectories and their interactions with clinical factors, younger children had declining processing speed (p = 0.012), females had declining est.IQ scores (p < .01) and patients with tumors at pineal gland had improving est.IQ scores (p = 0.010) over time. For survey-reported outcomes, there was no change in attention while younger children had improved EF over time (p = 0.045). Conclusions: Treatment associated with the ACNS1123 stratum 1 protocol had variable impact on patient’s longitudinal neurocognitive functioning. Younger age at irradiation and female gender were risk-factors for worse neurocognitive outcomes on examiner-administered measures. Only average PS was significantly below the normative mean at all 3 times in the collective sample. Patients should be proactively monitored for cognitive and educational supportive care services. These data could potentially serve for comparison of neurocognitive outcomes in future clinical studies that modify treatment approaches; however, the notable attrition must be considered as a limitation.