Early cognitive function tests predict early progression in glioblastoma.

Abstract

Early progression of glioblastoma prevents patients from completing the standard chemoradiation protocol. Given that cognitive function is associated with prognosis in glioblastoma, we investigated the usefulness of preoperative cognitive function tests for predicting the early progression of glioblastoma. Consecutive patients who underwent glioma surgery were preoperatively evaluated with cognitive function tests including the Mini Mental State Examination, digit span tests, the Controlled Oral Word Association Test, the Trail Making Tests (TMT, parts A, B, and C), and the Stroop test. Glioblastomas were treated with a standard protocol using radiation and temozolomide, and 6-month progression-free survival (PFS-6) was analyzed retrospectively. Among 126 patients who underwent glioma surgery, 55 patients were diagnosed with glioblastoma, and 50 patients were eligible for the PFS-6 analysis. Thirty-four patients (68%) achieved PFS-6. No significant differences were observed in demographics or tumor characteristics between patients without progression (PFS-6) or patients with progression (no-PFS-6). In the cognitive function tests, the PFS-6 patients exhibited better performance in TMT-A and TMT-B. In a multivariate logistic regression, TMT-B was the only independent predictor for PFS-6, whereas age, years of education, gross total or near total resection, concomitant chemoradiation, and TMT-A were not predictors. Patients with good TMT-B performance exhibited better early prognosis in the Kaplan-Meier survival analysis and had better recursive partitioning analysis classes. Our results indicated that preoperative TMTs can be useful for rapid evaluation of early prognosis in patients with glioblastoma.