17 Comparing Cognitive Patient-Reported Outcomes with Neuropsychological Impairment in Patients with Diffuse Glioma.

Abstract

Objective:Cognitive difficulties among diffuse glioma survivors are common in survivorship due to cancer treatment effects (i.e., surgery, chemotherapy, and/or radiation therapy), which can diminish quality of life. Routine monitoring of cognitive symptoms in survivorship is recommended and can help address patient needs and inform clinical interventions (e.g., cognitive rehabilitation). While several patient-reported outcome (PRO) measures have been used in brain tumor populations, there has been few studies comparing the performance of these PROs in patients with diffuse glioma. In order to better understand the value of different PROs, we conducted preliminary analyses associating cognitive PROs with neuropsychological impairment in a well-characterized sample of patients with diffuse glioma.Participants and Methods:23 glioma patients (mean aged 44.26 ± 12.24), six or more months after completing cancer treatment, underwent comprehensive psychosocial and neuropsychological assessments. The neuropsychological battery included the Hopkins Verbal Learning Test - Revised, Brief Visuospatial Memory Test - Revised, Wechsler Adult Intelligence Scale-IV tests of Coding and Digit Span, Trail-Making Test, Stroop Test, FAS, Animals, Boston Naming Test, and Rey-Osterrieth Complex Figure (copy). Completed cognitive PROs included the Functional Assessment of Cancer - Cognitive Function and Brain questionnaires (FACT-Cog; FACT-Br), the European Organization for the Research and Treatment of Cancer Quality of Life Questionnaire for Brain Neoplasms (EORTC QLQ-BN20), and the Multidimensional Fatigue Symptom Inventory, short form (MFSI-SF) Mental subscale. Based on published norms, we divided the sample into cognitively impaired and non-impaired groups (two or more primary neuropsychological test scores <= -2 z-score). We compared PRO scores between impaired and non-impaired groups using Mann-Whitney U tests. Higher medians equate to better cognitive functioning for all PROs, except for the MSFI-SF.Results:We found significantly worse scores in the impaired group compared to non-impaired group on the FACT-Cog subscales of perceived cognitive ability (PCA), [Non-Impaired (Mdn = 21, n = 11), Impaired (Mdn = 10, n = 12), U = 22.5, z = -2.68, = 0.007], perceived cognitive impairment (PCI), [Non-Impaired (Mdn = 59, n = 11), Impaired (Mdn = 44, n = 12), U = 32.5, z = -2.06, p=0.039]. The impaired group also trended towards worse scores on the FACT-Br additional concerns subscale [Non-Impaired (Mdn = 79.5, n = 10), Impaired (Mdn = 61, n = 12), U = 32.5, z = -1.81, p=0.07]. Group differences were not observed on the MSFI-SF [Non-Impaired (Mdn = 5, n = 11), Impaired (Mdn = 7, n = 12), U = 40.5, z = -1.57, p=0.12], or EORTC Cognitive Functioning subscale [Non-Impaired (Mdn = 83.33, n = 10), Impaired (Mdn = 75, n = 12), U = 42, z = -1.23, p=0.218].Conclusions:The preliminary findings suggest that the FACT-Cog, especially the PCA and PCI correspond with neuropsychological impairment among diffuse glioma survivors better than other cognitive PROs. The FACT-Br subscale was somewhat effective. The MFSI-SF Mental and EORTC Cognitive Functioning subscales did not correspond to impairment status. The FACT-Cog is a promising instrument and future work is needed to better determine relative utility of cognitive PROs in this population.