University of Florida Proton Therapy Institute Jacksonville, FL
R. Mailhot Vega1, D. J. Indelicato2, J. A. Bradley1, E. M. Mobley3, C. G. Morris1, A. Markatia1, N. P. Mendenhall1, A. M. Crisp4, and M. D. Miller5; 1Department of Radiation Oncology, University of Florida College of Medicine, Jacksonville, FL, 2Department of Radiation Oncology, University of Florida College of Medicine, Gainesville, FL, 3Department of Surgery, University of Florida College of Medicine Jacksonville, Jacksonville, FL, 4Center for Data Solutions, University of Florida College of Medicine Jacksonville, Jacksonville, FL, 5Collaborative Assessment and Program Evaluation Services & School of Human Development and Organizational Studies , University of Florida College of Education, Gainesville, FL
Purpose/Objective(s): Radiotherapy causes cognitive deficits in pediatric brain tumor survivors (PBTS). This has traditionally been measured using exams such as serial IQ tests administered after diagnosis. Scholastic data provides pre-diagnostic measurements, is practical for patients and families, and is directly aligned with the learning of students before and after treatment. We sought to evaluate the association between dose to organs at risk (OARs) historically associated with cognitive decrement and scholastic achievement in PBTS pre- and post-radiotherapy treatment. Materials/
Methods: With IRB approval, we retrospectively analyzed scholastic achievement in children (<21) with primary brain tumors treated with radiation at our institution. Eligible PBTS all resided within the state of the institution and were treated from 2007 to 2021. The state’s Department of Education (DOE) provided scholastic data for grade promotion, transcripts, accommodations, and state-mandated assessments of mathematics and reading, which were merged with institutional clinical data. The DOE also provided scholastic data on healthy children matched 3:1 by year, school district, age, and free or reduced lunch (FRL) eligibility. A general linear mixed-effects model was performed with the above scholastic dependent variables from the post-treatment phase and the independent variable of mean OAR dose to the hippocampus, corpus callosum (CC), and frontal lobe (FL). A sensitivity analysis was conducted with the primary model evaluating robustness, limited to only children in the treatment group. The coefficient for the mean dose was the outcome of interest, with a=0.1 for this pilot. Results: A total of 200 students were available for analysis: 50 PBTS and 150 matched controls with a median age of 11.6 years at treatment and 7 years of follow-up. Overall, 52% were FRL eligible, and 56% received craniospinal irradiation. In the primary model, increasing mean dose to the hippocampus (p = 0.048), CC (p = 0.053), and FL (p = 0.097) were negatively associated with standard grade promotion. In the secondary model restricted to PBTS, only hippocampal dose remained significantly associated with grade promotion (p = 0.092). OAR mean doses were not significantly associated with mathematics or reading performance nor the odds of receiving accommodations. Conclusion: We present the first report evaluating associations between dosimetry and scholastic performance. Our work identifies an association between increasing OAR dose associated with higher odds of not receiving a standard grade promotion (such as being held back), particularly for hippocampal dose. Evaluating scholastic success is an unmet need for PBTS. We demonstrate a novel method using scholastic performance data as a patient-centered metric, leveraging prospectively collected scholastic outcomes already recorded by states.
