RADI-04. PRETREATMENT VOLUME OF MRI-DETERMINED WHITE MATTER INJURY (WMI) PREDICTS COGNITIVE DECLINE AFTER HIPPOCAMPAL AVOIDANT (HA) WBRT FOR BRAIN METASTASES: SECONDARY ANALYSIS OF NRG ONCOLOGY RTOG 0933

Abstract PURPOSE: RTOG 0933 demonstrated benefits to memory following HA-WBRT, supporting the hypothesis of hippocampal radiosensitivity and associated memory specificity. However, some patients demonstrated cognitive decline, suggesting mechanisms outside hippocampal radiosensitivity playing a role. WMI has been implicated in RT-induced cognitive decline. This secondary analysis explored the relationship between pre-treatment WMI and memory following HA-WBRT. METHODS AND MATERIALS: 113 patients received HA-WBRT. Standardized cognitive assessments were performed at baseline, 2, 4, and 6 months. The primary endpoint was Hopkins Verbal Learning Test Delayed Recall (HVLT-DR) at 4 mos. Secondary endpoints included HVLT Total Recall (HVLT-TR) and Recognition (HVLT-Recog). Of 113 patients, 34 underwent pre-treatment and 4-month post-treatment HVLT testing and pre-treatment post-contrast volumetric T1 and axial T2/FLAIR MRI. Volumetric analysis of metastatic disease burden and disease-unrelated WMI was conducted on the pre-treatment MRI. Correlational analyses were performed examining the relationship between pre-treatment WMI and HVLT outcomes following HA-WBRT. RESULTS: Correlation was found between larger volumes of pre-treatment WMI and decline in HVLT-Recog (r=.54, p< .05) and a correlational trend was observed between larger volume of pre-treatment WMI and decline in HVLT-DR (r=.31, p=.08). Patients with higher pre-treatment disease burden experienced a greater magnitude of stability or positive shift in HVLT-recall and –delayed recall following HA-WBRT. (r=-.36 and r=-.36, p’s < .05), compared to the magnitude of stability/positive shift in those with lesser disease burden. CONCLUSION: In patients receiving HA-WBRT, pre-treatment-WMI predicts memory decline, suggesting white matter integrity pre-treatment contributes to the pathogenesis of post-WBRT cognitive toxicity independent of hippocampal stem cell radiosensitivity. Less decline or improvement in HVLT following HA-WBRT for patients with higher pre-treatment intracranial metastatic burden supports the importance of WBRT-induced intracranial control on cognition. These imaging biomarkers for cognitive toxicity will be further explored on NRG CC001 and CC003, phase III trials of WBRT with or without HA.