A multicenter randomized phase III study for newly diagnosed, maximally resected glioblastoma comparing carmustine wafer implantation followed by chemoradiotherapy with temozolomide versus chemoradiotherapy alone: Japan Clinical Oncology Group study JCOG1703 (MACS study).

Abstract

2001 Background: Three randomized phase III trials evaluated carmustine wafers (CWs) before temozolomide (TMZ) became standard therapy. To date, no randomized study has assessed CWs with standard TMZ chemoradiotherapy in newly diagnosed glioblastoma (GB). A post hoc analysis of Westphal’s phase III trial, which evaluated CWs versus placebo, suggested a survival benefit with CWs in GB patients with ≥90% resection. We conducted a randomized phase III trial to determine whether adding CWs to standard chemoradiotherapy improves overall survival (OS) in newly diagnosed GB with ≥ 90% resection. Methods: This multicenter trial (38 institutions; JCOG1703) enrolled patients with newly diagnosed GB considered resectable at ≥90%. After intraoperative pathological confirmation of malignant glioma and ≥90% removal, patients were randomized (1:1) to CW implantation (Arm B) or no implantation (Arm A). Arm B received intraoperative placement of up to 8 CWs. All patients received 60 Gy radiotherapy with concomitant daily TMZ followed by 12 cycles of maintenance TMZ. The primary endpoint was OS; secondary endpoints included progression-free survival (PFS), loco-regional PFS (LPFS), and adverse events. DNA methylation array analysis and molecular testing included MGMT promoter methylation and copy number profiling. Results: Among 234 registered patients, 221 (Arm A, n=110; Arm B, n=111) were randomized between June 2019 and November 2022. Baseline characteristics were well balanced. Median OS for all randomized patients was 26.5 months (1-year 86.4%, 2-year 54.3%, 5-year 16.6%), with no significant difference between Arm A and B (21.7 vs 27.7 months; hazard ratios (HR) 0.907, 91.5% confidence interval (CI) 0.691–1.192; one-sided p=0.269 by the stratified log-rank test). Median PFS was 9.8 months (1-year 43.4%, 2-year 16.3%), also without significant difference (9.5 vs 10.4 months; HR 0.867, 95% CI 0.658–1.141). Median LPFS was 11.6 months (1-year 48.4%, 2-year 25.1%), with no difference between arms (10.6 vs 12.7 months; HR 0.886, 95% CI 0.669–1.173). MGMT promoter methylation was a significant favorable factor for OS, PFS, and LPFS. Early postoperative complications and adverse events were comparable between groups, and no treatment-related deaths occurred. Conclusions: CW implantation added to standard TMZ chemoradiotherapy did not significantly improve OS in newly diagnosed GB after ≥90% resection. Based on these findings, CW implantation does not provide significant therapeutic or survival benefit when standard TMZ chemoradiotherapy is feasible in this population. Clinical trial information: jRCT1031190035.