Repeat resection for recurrent glioblastoma - does timing matter?

backgroundThe optimal timing of repeat surgical resection in patients with recurrent IDH-wildtype glioblastoma (rGB) remains unclear. We aimed to characterize temporal patterns between radiological suspicion of recurrence and repeat resection and to evaluate the impact of early versus delayed surgery on the extent of resection (EOR), functional outcomes, adjuvant therapy, and survival

methodsWe retrospectively analyzed a consecutive cohort of 150 patients who underwent resection for histopathologically confirmed rGB between 2015 and 2023 at a single tertiary care center. Assessment of contrast-enhancing preoperative and residual tumor volumes (RTV on early postoperative MRI) was performed using semi-automated segmentation. Based on the mean or median time between suspicion of recurrence and repeat resection, patients were stratified into early and late surgery groups. RANO Resect criteria and a 0.175-ml RTV threshold were used to classify EOR. Functional outcomes, postoperative treatment, as well as progression-free survival (PFS) after repeat resection, and overall survival (OS) after suspicion of recurrence were compared between groups.

resultsMean and median time from suspicion of recurrence to repeat resection were 54 and 24 days, respectively, with 75% of patients undergoing reoperation within 6 weeks. Applying the mean cut-off, early (n=120) and late (n=30) surgery groups showed comparable baseline demographics, performance status, tumor eloquence, and preoperative neurological deficits. Preoperative tumor volumes were significantly smaller in the early surgery group (12.7 vs. 25.9 ml, p=0.002). Late surgery was associated with a trend toward higher RTV and lower rates of gross total resection, though without statistical significance. Rates of transient and permanent postoperative neurological deficits were low (15% and 2%) and did not differ between groups. Adjuvant treatment patterns differed, with early surgery patients more frequently receiving CCNU-based chemotherapy, while late surgery patients more often received no further treatment. Median OS after suspicion of recurrence (12.4 vs 14.3 months) and PFS after repeat resection (4 months in both groups) were not significantly different between early and late surgery groups. A re-analysis using the median of 24 days as cut-off revealed similar results with regards to survival and functional outcomes.

conclusionMost patients with repeat resections for rGB underwent surgery shortly after radiological suspicion of recurrence. While delayed surgery was associated with larger tumor volumes and a trend toward less favorable EOR and adjuvant treatment, timing of surgery alone was not associated with functional outcomes or survival. These findings support individualized decision-making for repeat resection based on clinical and radiological factors rather than timing alone.