P01.03.B A quantitative comparison of cognitive performance and patient-reported symptoms in preoperative lower-grade glioma patients from two Dutch Hospitals

Abstract

Abstract Background Protocols for assessment of (neuro)psychological outcomes in lower-grade glioma patients vary between hospitals. This potentially complicates generalization of these outcomes. We compared standardized scores on tests of two frequently impaired cognitive domains (attention and executive functioning (EF)), and two relevant patient-reported outcomes (PROs; depression and fatigue) of two neuro-oncological hospitals that use different measurement instruments. Material and Methods Data were used from preoperative assessments of patients with (IDH-mut) WHO grade II/III glioma tested between 2007 and 2021 at Amsterdam UMC (AMS) or at Elisabeth-Tweesteden Hospital Tilburg (ETZ). AMS patients were referred for (neuro)psychological assessment based on physician and patient preference (paper and pencil tests), whereas all ETZ patients routinely undergo screening (computerized tests). To compare scores of the different attention and EF tests we converted patients’ performances to z-scores based on normative data. For cognitive performance, we compared scores of different cognitive flexibility tests (CST vs SAT), processing speed tests (SDC vs LDMT), and Stroop tests (Stroop I and Stroop III). PROs included the CES-D vs HADS-D and the CIS-fatigue vs MVI-general fatigue (AMS vs ETZ, resp.). Differences were tested using Fisher's, χ 2, and Mann-Whitney U tests. Results Assessments were done median 4 weeks (AMS, n=97, range 19-0 weeks) and 1 day (ETZ, n=106; range 14-0 days) preoperatively. Age, sex, tumor location and histology were comparable between cohorts (p>0.05), but the AMS cohort showed significantly more grade III tumors (36% vs 16%) and more awake surgeries (84% vs 46%). Z-scores measuring attention and EF (n=94 and n=95, AMS vs ETZ) were not significantly different (CST vs SAT, percentage with a disorder (z <-1.5SD) 15% vs 13%; SDC vs LDMT 13% vs 14%; Stroop I 11% vs 18%; Stroop III 13% vs 16% at AMS and ETZ, resp.). Percentages of patients with possible depression (CES-D≥16, n=88 and HADS-D≥8, n=106) did not differ significantly between hospitals (28% vs 26%), nor did percentages of patients with severe fatigue (CIS-fatigue≥35, n=88 and MVI-general fatigue (z <-1.5SD), n=38, 42% vs 24% at AMS and ETZ, resp.). Conclusion Standardized scores of glioma patients on cognitive domains (attention and EF) and PROs (depression and fatigue) did not differ between two centers with slightly different samples using different testing protocols. This cautiously suggests that study findings on cognitive functioning and symptoms could be generalized. For research purposes, conjoint use of pooled populations for outcome evaluation could be explored with different samples from other centers using different instruments.