Assessment of neuropsychological function during early treatment of diffuse glioma

Background: Cognitive function in patients with diffuse glioma is frequently impaired and can have a profound impact on quality of life. Accurate, reproducible and accessible tools to assess cognition are mandatory to understand the effects of the tumour and treatment. Our hypothesis was that an app-based assessment would be complementary to traditional neuropsychological testing, thereby aiding in defining cognitive profiles and trajectories during early treatment of diffuse glioma. Methods: Seventeen subjects with diffuse low-grade gliomas completed a traditional neuropsychological assessment battery before and after surgery. In addition an app-based tablet assessment (OCS-Bridge) was administered pre- and post-operatively as well as longitudinally at 3- and 12-month follow-up. Deficit rates, mean performance, and changes over time were compared using standardized z-scores between the two testing methods. Unsupervised k-means clustering was performed on individual cognitive tests in each battery. Results: Preoperative testing showed an average of 2.88 deficits and 1.18 deficits per patient on neuropsychological testing and the tablet-based OCS-Bridge assessment, respectively. Digit span testing demonstrated agreement between testing modalities, but otherwise there was no significant correlation (Pearson9s correlation: p=0.7723, r-value = 0.0758, df = 16). Longitudinal assessment revealed dynamic changes in attention and nonverbal skills. Traditional assessment was more sensitive to memory deficits, showing 22 preoperative deficits within the cohort vs. 1 for the app-based assessment, while app-based assessment was more sensitive to nonverbal skills, showing 8 deficits preoperatively vs. none in the traditional assessment. Clustering analysis did not create clusters along the predetermined domains, indicating that certain individual tests may test more than one cognitive function. Conclusions: These data suggest app-based assessment is reliable and complementary to data obtained from traditional neuropsychological testing. Advantages include efficiency, facilitation of longitudinal testing, and increased sensitivity in domains of non-verbal skills and attention. Patients with diffuse glioma show subtle neuropsychological impairments, unique cognitive profiles, and discrete trajectories during early treatment, therefore judicious assessment is imperative.