Early Repeat Residual Resection Versus Adjuvant Therapy for Incompletely Resected Glioblastoma: A Case-Control Study

BACKGROUND AND OBJECTIVE: Maximal safe resection of newly diagnosed glioblastoma (GBM) optimizes clinical outcomes. For patients who receive biopsy or subtotal resection (STR), early repeat resection (ERR) to improve extent of resection can be considered but is controversial. Oncological outcomes of patients undergoing ERR for residual GBM to patients receiving upfront at least near-total resection were compared. METHODS: This case-control study including patients with GBM treated at a single institution identified 3 treatment groups: (1) ERR patients underwent biopsy or STR (<95% cytoreduction), followed by ≥NTR (≥95% cytoreduction) within 8 weeks; (2) control patients underwent upfront ≥NTR; and (3) biopsy with chemoradiation only. ERR and control patients were 1:1 case-control matched by age, performance status, O6-methylguanine-DNA methyltransferase methylation status, isocitrate dehydrogenase mutation status, and completion of standard chemoradiation. Patient demographics, treatments, and oncological outcomes were analyzed. A P -value of <.05 was considered significant. RESULTS: A total of 22 ERR patients (16 biopsies and 6 STR) were identified, and each was matched with a control patient. Baseline patient characteristics at presentation did not differ between matched patients. Time from first surgery to chemoradiation was longer in the ERR cohort (1.8 vs 1.1 months, P < .001). Median overall survival (mOS) (17.7 vs 20.3 months, P = .87) and progression-free survival (5.5 vs 4.5 months, P = .25) did not differ between ERR and control groups, respectively. In the biopsy-only group, mOS was 4.1. Univariate Cox-regression analysis suggested age, failure to complete chemoradiation, isocitrate dehydrogenase-wt, and hemorrhage at presentation were independent predictors of mOS, whereas only age and failure to complete chemoradiation remained independent predictors of mOS after multivariate analysis. Surgical complications were similar between cohorts. CONCLUSION: ERR achieving ≥95% extent of resection within 8 weeks of initial surgery results in similar oncological outcomes to upfront ≥NTR in a case-control–matched analysis despite delayed initiation in chemoradiation for the ERR cohort.