CTNI-50. NEUROCOGNITIVE FUNCTION (NCF) OF THE PHOTON COHORT IN NRG-BN001

Abstract NRG BN001 is an ongoing randomized phase II trial of dose-intense (DI-RT) versus standard dose photon-based radiation therapy (SD-RT) with temozolomide (TMZ) for newly diagnosed glioblastoma. We report preliminary results of Group 1, DI-RT delivered with IMRT while SD-RT could be 3DCRT or IMRT. Group 2, DI-RT delivered with proton therapy, continues to accrue. From 10/2014 to 7/2018, 229 patients were eligible and randomized. Differences in overall survival following DI-RT versus SD-RT were not significant. Patients were scheduled to complete NCF testing at baseline, cycle 3 (within 7 days of cycle 4), and cycle 12 (day 22–28 of cycle 12, or 60 weeks from completion of chemoradiation). At baseline, 93–94% of eligible patients completed NCF testing. Compliance for evaluable patients at cycle 3 and cycle 12 was 66–68% and 51–54%, respectively, across the battery of NCF tests. The most common reasons for missing data were patient refusal (cycle 3: 14%, cycle 12: 22–23%) and institutional error (cycle 3: 9–10%, cycle 12: 14–15%). A prespecified secondary endpoint analysis was conducted to evaluate differences in NCF between SD-RT and DI-RT based on the Clinical Trial Battery Composite (CTB COMP), which is the mean of the standardized scores from the NCF test battery (HVLT-R, TMT, COWA). There was no statistically significant between arm difference in change from baseline on the CTB COMP at cycle 3 (DI-RT vs SD-RT mean/SD, 0.0 +/- 1.3 vs -0.3 +/- 1.5, p=0.370, Cohen’s d=0.22) or cycle 12 (DI-RT vs SD-RT mean/SD, 0.2 +/- 1.7 vs 0.2 +/- 1.1, p=0.977, Cohen’s d=0.01). A mixed effects longitudinal model of the CTB COMP yielded a non-significant time by treatment effect interaction (p=0.216). There were no significant differences in change scores between arms on any NCF test at cycles 3 or 12. NCF outcomes were similar for photon-based SD-RT and DI-RT.