Management of first recurrence from IDH-wildtype glioblastoma with re-resection or stereotactic radiosurgery: A retrospective, international, multicenter study.

Abstract

2063 Background: Optimal management of recurrent glioblastoma remains ill-defined. While re-resection is frequently provided as local salvage therapy, stereotactic radiosurgery (SRS) might represent a treatment approach for small, focal recurrences. We comparatively evaluated outcomes after re-resection and SRS for first recurrence of IDH-wildtype glioblastoma. Methods: Adult patients with first radiographic recurrence of IDH-wildtype glioblastoma (WHO 2021 classification) were identified. Information on patients managed with re-resection were collected through the international eight-center database of the RANO resect group, and patients managed with SRS were collected by twelve centers from the IRRF study group. Survival after recurrence was analyzed using Kaplan–Meier estimates and Cox proportional hazards models. Inverse probability of treatment weighting (IPTW) was applied to mitigate baseline imbalances. Results: Overall, 567 patients were included (re-resection: n = 310; SRS: n = 257). Compared with SRS, patients managed with re-resection were younger and had larger contrast-enhancing recurrence volumes. Despite those baseline differences, overall survival did not differ between treatment modalities (10.2 [95%CI 9.3–11.7] vs. 11.1 [10.3–12.8] months; HR 1.12, 95%CI 0.93–1.35; p = 0.231). Also, no survival differences between treatment groups were observed when effects of potential baseline confounders including tumor volume were minimized using IPTW. Notably, among patients typically not considered SRS candidates due to tumor volumes >10 cm³ ( n = 234), maximal re-resection of the enhancing tumor (RANO class 1–2) was associated with survival comparable to SRS-treated low-volume tumors ≤10 cm³ (9.1 [7.7–11.5] vs. 11.0 [9.9–12.8] months; HR 1.27, 0.95–1.69; p = 0.106) whereas incomplete re-resection (RANO class 3) was associated with less favourable survival. In the subgroup of patients with focal recurrences in whom local physicians from the IRRF consortium perceived clinical equipoise between treatment modalities but eventually performed SRS ( n = 112), IPTW-adjusted analyses demonstrated no significant difference in overall survival compared with re-resected patients (HR 0.90, 0.65–1.25; p = 0.539). Conclusions: While complete re-resection is prognostic in recurrent glioblastoma, SRS might represent a reasonable therapeutic approach for selected individuals with small low volume tumor recurrence. Given substantial baseline differences between patients treated with re-resection and SRS and potential treatment provider bias, prospective trials are warranted to guide nuanced decision-making.