P01.09.A Prevalence and predictors of cognitive impairment in adult glioma survivors after multimodal therapy

Abstract

Abstract Background Long-term survival can be achieved in an increasing number of glioma patients after treatment. Therefore, safeguarding these survivors’ quality of life (QoL) is essential. Neurocognitive decline arises in many young patients, placing a heavy burden on the social and economic aspects of the patients’ lives. A lot of debate is currently ongoing regarding the prevalence of neurocognitive impairment and individual predictors of whom is susceptible for such side effect. Material and Methods In this cross-sectional study, 37 WHO grade 2-3 adult glioma survivors, at least one year after multimodal therapy, were tested using a comprehensive neurocognitive test battery covering multiple cognitive domains. Neurocognitive test scores were converted into z-scores using country-specific normative data. Cognitive impairment was defined as a z-score lower or equal to -1.50 for each subtest. Age, time since multimodal therapy, radiotherapy treatment and tumour location were included as predictors in a linear regression model per outcome (n=12). Results In this cohort, 29 patients (78%) showed a test score below the predefined cutoff on at least one cognitive test. The percentage of patients who showed test-specific cognitive impairment ranged from 8.1% to 56.76% per test. Fine motor skills, verbal memory, processing speed and executive functioning were the most commonly affected cognitive domains. In this study, the variability in processing speed performance was associated with age (TMT A, p=0.03), time since therapy (WAIS-IV coding, p=0.02) and tumour location. In these measures, poorer outcomes were observed with increasing age, longer time since therapy and in patients with gliomas located in the left frontal lobe. Moreover, age showed to be a significant predictor of verbal memory, with poorer outcomes on the HVLT-R delayed recall task with increasing age (p=0.04). Tumour location predicted working memory performance, as patients with right parietal tumours (p=0.03) showed significantly worse on the WAIS-IV digit span task. Conclusion These preliminary data underline the various alterations of neurocognitive functioning in glioma survivors after multimodal therapy. Therefore, future research needs to shift towards a patient-tailored approach. The next step in this study will be to link these neurocognitive data to advanced neuroimaging data to explore the potential predictive value of imaging markers for neural damage and cognitive outcomes, paving the path to innovative treatment planning techniques.