NCOG-21. TESTING AND VALIDATION OF A DIFFUSE GLIOMA COGNITIVE RESERVE MODEL

Abstract

Abstract Behavioral and cognitive symptom severity varies widely across patients with glioma because of unknown mechanisms. The concept of cognitive reserve (CR) recognizes that there are individual differences in each person’s brain reserve capacity and this theory accounts for the common disconnection between the degree of brain injury and its symptom manifestations. CR is a fixed variable throughout an individual’s life and works by recruitment of dormant yet redundant neural networks, and by efficient information modulation. New strategies to treat cognitive impairments have been encumbered by a lack of models that accurately reflect human disease. Here, we measure the Cognitive Reserve Index (CRI) in a registry of 450 patients with WHO grade 2-4 glioma. Transient behavioral impairments were induced by anesthetic medications propofol, dexmedetomidine, and remifentanil during standard of clinical care. We administered and analyzed 48 picture naming (PN) and 32 auditory stimulus naming (AN) trials per patient both before and after drug administration with impairments defined as a decline in task performance by more than one standard deviation from the mean. We discovered that the CRI in patients with glioma follows a normal distribution, like the general US population. Next, we found a significant decline in task performance in a subgroup of patients in both PN and AN. Electrode arrays in vivo to record human local field potentials from glioma infiltrated cortex demonstrated electrophysiological biomarkers of cognitive decline. Further elucidation and validation of CR measures in patients may significantly influence the delivery and timing of cancer directed therapies.