153 - Assessing Atrophy and Neurocognitive Decline in Hippocampal Subfields after Fractionated Brain Radiation Therapy

H. Nguyen¹, A. B. Hopper², R. Karunamuni³, S. Unnikrishnan⁴, M. A. Salans⁵, A. Reyes³, A. Stasenko⁶, C. McDonald⁶, A. Christofferson⁷, and J. A. Hattangadi-Gluth³; ¹UCSD School of Medicine, La Jolla, CA, ²Department of Radiation Medicine and Applied Sciences, UC San Diego, La Jolla, CA, ³Department of Radiation Medicine and Applied Sciences, University of California San Diego, La Jolla, CA, ⁴Winship Cancer Institute of Emory University, Atlanta, GA, United States, ⁵University of California San Francisco, San Francisco, CA, ⁶Department of Psychiatry, University of California San Diego, La Jolla, CA, ⁷University of California, San Diego, La Jolla, CA, United States

Purpose/Objective(s): Radiation therapy (RT) is critical for primary brain tumors, yet many patients will experience neurocognitive decline. Bilateral hippocampi are largely implicated, with relative sparing of these areas as a cognitive protective strategy with whole brain. We have previously described that dose-dependent atrophy is found in the hippocampus (HC), yet just as the brain is a complex and interconnected organ the HC itself has many subfields which subserve different components of memory. Here we examine changes in hippocampal subfield volume after RT, in relation to dose and time from treatment, as well as the association between these changes and decline in verbal and visuospatial memory. Materials/

Methods: Data were analyzed from a prospective longitudinal clinical trial. Patients (n=85) with primary brain tumors receiving fractionated RT completed high-resolution volumetric brain MRI and neurocognitive evaluation at baseline and 3-, 6-, and 12-month intervals. Image processing using robust, validated automated parcellation segmented the bilateral HC and the following subfields: parasubiculum, presubiculum, subiculum, corpu ammonis (CA) 1-4, granule and molecular cell layers of the dentate gryus (GC-ML-DG), hippocampal-amygdala transition area (HATA), fimbria, molecular layer, HC fissure, and HC tail. Neurocognitive testing was performed by certified neuropsychologists and included the Brief Visuospatial Memory Test (BVMT) and Hopkins Verbal Learning Test (HVLT). Multivariable linear mixed-effects models assessed longitudinal changes in whole HC (left, right) and nuclei volumes as well as associations between dose, volume, and verbal and visuospatial memory performance.
Results: We found significant atrophy in the left HC at 6 months (P = 0.032). Atrophy was dose-dependent in the left HC at 12 months (P = 0.025), and the right HC at 3 months (P = 0.018), 6 months (P = 0.016) and 12 months (P = 0.009). Worse verbal memory on the HVLT total and delayed recall was associated with decreased volume in the left hippocampal tail (P = 0.0098, P = 0.0090, respectively) and left hippocampus–amygdaloid transitional area (P = 0.013, P = 0.026, respectively). Poor BVMT total recall was associated with atrophy of the right body of the subiculum (P = 0.024), right head of the presubiculum (P = 0.045), right body of the molecular layer (P = 0.031), and right parasubiculum (P = 0.024). Atrophy of the left hippocampal tail (P = 0.0044) was associated with poor BVMT delayed recall.
Conclusion: Significant overall atrophy was only noted in the left hippocampus at 6 months; however, dose-dependent atrophy was found in the bilateral hippocampi. Decrease in verbal and visuospatial memory was generally associated with left-sided hippocampal subfield atrophy, while decrease in visuospatial memory was associated with right-sided hippocampal subfield atrophy. Relative specificity of the HC subfields may allow more directed memory-preserving cognitive sparing strategies.