Neuroanatomical Correlates of Cognitive Phenotypes in Patients with Primary Brain Tumors

2117 

Abstract Submission 

Jiwandeep Kohli¹, Anny Reyes¹, Austin Hopper¹, Alena Stasenko¹, Natalia Menendez¹, Divya Prabhakaran¹, Roshan Karunamuni¹, Jona Hattangadi-Gluth¹, Carrie McDonald¹

¹University of California, San Diego, San Diego, CA

Jiwandeep Kohli  
University of California, San Diego
San Diego, CA

Anny Reyes  
University of California, San Diego
San Diego, CA

Austin Hopper  
University of California, San Diego
San Diego, CA

Alena Stasenko  
University of California, San Diego
San Diego, CA

Natalia Menendez  
University of California, San Diego
San Diego, CA

Divya Prabhakaran  
University of California, San Diego
San Diego, CA

Roshan Karunamuni  
University of California, San Diego
San Diego, CA

Jona Hattangadi-Gluth  
University of California, San Diego
San Diego, CA

Carrie McDonald  
University of California, San Diego
San Diego, CA

Patients with primary brain tumors exhibit substantial variability in neurocognitive profiles, likely related to multifactorial etiologies and different neuroanatomical locations. Cognitive phenotyping offers a patient-centered approach toward characterizing patterns of impairment and has shown utility for determining risk for disease progression in neurological disorders (Hancock et al., 2023; Hermann et al., 2007). Investigating the neuroanatomical correlates of cognitive phenotypes in patients with brain tumors may lead to a better understanding of the mechanisms underlying different patterns of cognitive impairment while also elucidating specific risk factors for cognitive decline.

Patients with primary brain tumors were recruited for a prospective, observational study examining the effects of fractionated, partial brain radiotherapy on cognition between 2014 and 2021. Neurocognitive and structural MRI data were available for 79 participants prior to radiation treatment. Patients were cognitively phenotyped using latent profile analysis in a prior study, which revealed three groups: those with generalized impairments (17.7%), those with isolated verbal memory impairments (12.7%), and those with minimal impairments (69.6%). MRI scans were acquired on a 3.0T 750 GE scanner. Anatomical images were acquired using a T1-weighted inversion recovery spoiled gradient echo sequence and diffusion data were acquired with a single-shot pulsed-field gradient spin EPI sequence with b = 0, 500, 1500, and 4000 s/mm2, with 1, 6, 6, and 15 unique gradient directions for each b-value respectively. Anatomical scans were processed using FreeSurfer version 5.3.0. Diffusion tensors were calculated using mono-exponential fitting from b=0, 500, and 1500 s/mm2 to extract estimates of fractional anisotropy (FA) and mean diffusivity (MD). DTI metrics were calculated from the co-registered DTI maps by sampling up to 5 mm below the white matter surface normal at each vertex and then averaging within each ROI volume. Tumor, necrotic tissue, and regions of edema were manually censored for each patient and excluded from final ROIs prior to analysis. Cognitive phenotypes were compared using ANCOVAs for CT, FA, and MD in each ROI while controlling for age, with follow up pairwise comparisons. Pearson correlations with neurocognitive test performance were examined in regions exhibiting significant differences between groups on structural metrics.

Compared to the minimal impairment group, the verbal memory impairment group showed significantly increased CT in the left temporal pole (p = 0.032) and right parahippocampal gyrus (p = 0.022), along with reduced MD bilaterally in the parahippocampal gyri (left p = 0.008; right p = 0.078). Greater MD in the left parahippocampal gryus was also significantly associated with poorer HVLT Learning (r = 0.784; p < 0.001) and Delayed Recall (r = 0.528; p < 0.01) scores. The generalized impairment group showed decreased CT in the left cuneus (p = 0.039) left frontal pole (p = 0.049), right pars orbitalis (p = 0.033) and right superior parietal region (p = 0.049). Neither CT nor FA was associated with cognitive performances within the phenotypes.

Cognitive phenotypes in patients with primary brain tumors showed unique patterns of brain pathology, suggesting different underlying mechanisms of impairment profiles. These results demonstrate the utility of examining neuroanatomical correlates of cognitive phenotypes for identifying areas of vulnerability that may inform treatment decisions for individual patients based on patterns of neurocognitive performance, with significant brain-cognition correlations supporting the biological relevance of this approach. Examining how phenotypes evolve over the course of treatment could be additionally helpful for understanding the effects of different intervention approaches on cognition, and for identifying patients at risk for treatment-related decline.

Learning and Memory Other

Diffusion MRI Modeling and Analysis

Cortical Anatomy and Brain Mapping ¹

White Matter Anatomy, Fiber Pathways and Connectivity ²

Memory

Neoplastic Disease

STRUCTURAL MRI

WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC