Baseline cognitive function to predict survival in patients with glioblastoma.

Background: Overall survival of patients with glioblastoma (GBM) remains dismal with range from 12-14 months. In addition to the oncologic burden that these patients face during the disease trajectory, there are demonstrable neurologic and cognitive challenges. These cognitive challenges can occur at baseline and throughout the disease course. We sought to understand the contribution of baseline cognition on the overall survival of patients with newly diagnosed GBM. Methods: In this prospective single-center study, newly diagnosed GBM patients were evaluated with computerized neurocognitive testing including testing of domains of cognitive flexibility, complex attention, executive functioning, psychomotor speed, processing speed, verbal and visual memory. Demographic data and known prognostic factors (age, extent of resection, and KPS) were obtained. Associations of each cognitive test with survival were assessed using adjusted Cox proportional hazards models. Patients were dichotomized to cognitive function groups as performing poorly on 4 or less modalities or greater than 4 modalities and overall survival of each group was estimated using Kaplan Meier methods. Results: 61 newly diagnosed GBM patients were initially included with a mean age of 58.01 yrs (sd = 10.22 yrs) with KPS ≥ 90 for 45 patients (47.9%) and KPS 70-80 for 49 patients (52.1%). Overall survival of 61 GBM patients that had a baseline cognitive function showing poor performance on 4 modalities or less was 23.9 mos (95% CI 14.8 mos, 30.7 mos) whereas the overall survival of GBM patients that had a cognitive function showing poor performance on greater than 4 modalities was 14.1 mos (95% CI 11.8 mos, 20.3 mos). This difference neared significance with log-rank of p = 0.0273. Conclusions: These observations highlight the importance of baseline cognitive function as it pertains to survival in GBM patients. Understanding baseline cognition and continued measures of cognition in primary brain tumor patients remains important and should be taken into context in regards to prognostication and quality of life of patients and caregivers.

Duke Scholars

Author Katherine Barnett Peters Neurosurgery, Neuro-Oncology

Author James Emmett Herndon II Biostatistics & Bioinformatics, Division of Biostatistics

Author Henry Seth Friedman Neurosurgery, Neuro-Oncology