EORTC
Temozolomide Versus Radiotherapy as First-Line Therapy for Low-Grade Glioma: Mature Results of a Randomized Phase III Trial (EORTC 22033-26033/NCIC-CTG/TROG/MRC-CTU)
Bottom Line
In high-risk WHO grade 2 glioma, initial temozolomide and radiotherapy yield equivalent PFS and OS overall and within IDH-mutant subtypes, but IDH-wildtype tumors have a marked OS benefit with TMZ (HR 0.47) — supporting molecular stratification, not the choice of monotherapy modality, as the key prognostic driver.
Major Points
- Neither PFS nor OS differed by treatment arm in the overall ITT population.
- In IDH-mutant 1p/19q non-codeleted astrocytomas (n=178), median OS was similar (~6.6-6.7 yrs) with RT vs TMZ (P=.93).
- In IDH-mutant 1p/19q codeleted oligodendrogliomas (n=109), median OS was 12.9 yrs (RT) vs 14.9 yrs (TMZ), HR 0.88 (0.52-1.49), P=.63.
- In IDH-wildtype tumors (n=64), TMZ yielded significantly longer OS: 4.7 vs 2.5 yrs, HR 0.47 (0.27-0.82), P=.0068.
- Patients ≥40 years old had better outcomes than <40, reversing the historical view of age as a negative prognostic factor once molecular subtype is accounted for.
- 73% of enrolled tumors (351/478) had tissue available for 2021 WHO molecular reclassification.
- Combined-modality therapy (RT+chemotherapy) was not tested and has since become standard for IDH-mutant astrocytoma.
Design
Study Type: Randomized Phase III Open-Label Intergroup Trial
Randomization: 1
Blinding: Open-label (unblinded)
Enrollment Period: Registration Sep 2005 - Mar 2010; randomization Dec 2005 - Dec 2012
Follow-up Duration: Mature follow-up (median follow-up 48 months at primary report; long-term OS follow-up for this update)
Centers: 78
Countries: Netherlands, Belgium, Germany, France, United Kingdom, Italy, Spain, Switzerland, Austria, Portugal, Israel, Canada, Australia, United States
Sample Size: 478
Analysis: Intention-to-treat; molecular subgroup analyses post hoc using 2021 WHO classification
Inclusion Criteria
- Adults age ≥18 years
- Histologically confirmed supratentorial WHO grade 2 diffuse glioma (astrocytoma, oligodendroglioma, or oligoastrocytoma)
- At least one high-risk feature: age ≥40 years, radiologic tumor progression, new or worsening symptoms, or refractory seizures
- WHO performance status ≤2 and RTOG neurological function score 0-3
- Adequate hematologic (ANC ≥1500/mm³, platelets ≥100,000/mm³), renal (creatinine ≤1.5× ULN), and hepatic (bilirubin ≤1.5× ULN, LFTs <2.5× ULN) function
- Paraffin-embedded tumor tissue and heparinized blood sample available for central pathology review and 1p molecular testing
- Not candidate for surgery-only treatment
Exclusion Criteria
- Tumor transformation to a higher WHO grade
- Prior chemotherapy or radiotherapy
- Inadequate performance status (WHO >2) or neurological function (RTOG >3)
- Insufficient hematologic, hepatic, or renal function
- Ineligible tumor tissue for centralized molecular testing
Arms
| Field | Control | Temozolomide (dose-dense) |
|---|---|---|
| Intervention | 3D-conformal RT 50.4 Gy in 28 fractions of 1.8 Gy over 5-6 weeks (IMRT/stereotactic-guided permitted at same dose) | Temozolomide 75 mg/m² orally daily × 21 days every 28-day cycle, up to 12 cycles, or until progression/unacceptable toxicity |
| Duration | 5-6 weeks | Up to 12 months (12 × 28-day cycles) |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Progression-free survival (time from randomization to clinical or radiologic progression or death) — no significant difference between arms | Primary | Median PFS 46 months (RT) | Median PFS 39 months (TMZ) | 1.16 | 0.22 |
| Overall survival — overall population | Secondary | No significant difference | No significant difference | NS | |
| OS in IDH-mutant / 1p19q non-codeleted astrocytoma (n=178) | Secondary | Median 6.6-6.7 yrs (RT) | Median 6.6-6.7 yrs (TMZ) | 0.67-1.44 (range) | 0.93 |
| OS in IDH-mutant / 1p19q codeleted oligodendroglioma (n=109) | Secondary | Median 12.9 yrs (9.4-NR) (RT) | Median 14.9 yrs (10.1-NR) (TMZ) | 0.88 (0.52-1.49) | 0.63 |
| OS in IDH-wildtype tumors (n=64) | Secondary | Median 2.5 yrs (1.8-3.3) (RT) | Median 4.7 yrs (2.2-7.2) (TMZ) | 0.47 (0.27-0.82) | 0.0068 |
| Molecular reclassification per 2021 WHO | Secondary | 73% (351/478) with analyzable tissue | - | ||
| Grade 3-4 hematologic toxicity (TMZ) | Adverse | 13.7% (32/232) | |||
| Grade 3-4 hematologic toxicity (RT) | Adverse | Minimal (few cases) | |||
| Infections during TMZ | Adverse | 3.4% (8/232) including 1 Pneumocystis jirovecii pneumonia | |||
| Infections during RT | Adverse | 0.9% (2/220) including 1 Pneumocystis jirovecii pneumonia | |||
| Moderate-severe fatigue (TMZ) | Adverse | 6.8% (16/232), one grade 3 | |||
| Moderate-severe fatigue (RT) | Adverse | 3.3% (8/220), all grade 2 | |||
| Thromboembolic events | Adverse | 1.2% (3 patients: 2 RT, 1 TMZ) | |||
| Treatment discontinuation for toxicity (TMZ) | Adverse | 8% (18/232) | |||
| Treatment discontinuation for progression (TMZ) | Adverse | 9% (21/232) | |||
| TMZ dose reductions | Adverse | 15% (34/232) due to hematologic (8%), non-hematologic (4%), or other reasons | |||
| Treatment-related deaths | Adverse | 2 (one per arm: RT-arm encephalopathy with progression; TMZ-arm bone marrow failure) | |||
| Completion of 12 TMZ cycles | Adverse | 75% | |||
| Completion of planned RT | Adverse | 91% | |||
Subgroup Analysis
IDH-wildtype tumors derive significant OS benefit from TMZ over RT (HR 0.47). IDH-mutant subgroups (codel and non-codel) show no OS benefit from either monotherapy. Age ≥40 vs <40: older patients had better outcomes once molecularly stratified, challenging age as an independent negative prognostic factor. Original 2016 report also showed IDHmt/non-codel astrocytomas had significantly longer PFS with RT than TMZ (HR 0.53, P=.0043).
Criticisms
- Combined-modality therapy (RT + chemotherapy), now standard of care for IDH-mutant astrocytoma per RTOG 9802 and INDIGO, was not tested — limits contemporary applicability.
- IDH-wildtype subgroup was small (n=64) and molecularly heterogeneous; these tumors are now classified as glioblastoma or other entities under 2021 WHO criteria, so the finding of TMZ benefit reflects modern glioblastoma biology.
- Open-label design (impossible to blind RT vs oral chemotherapy) introduces potential ascertainment bias for PFS assessed by MRI.
- PFS as primary endpoint depends on imaging technique and interpretation; radiotherapy-induced imaging changes may confound progression calls in the RT arm.
- Molecular subgroup analyses were post hoc and underpowered — the trial was not stratified prospectively for 2021 WHO categories.
- Only 73% of tumors had analyzable tissue for reclassification, potentially biasing molecular subgroup results.
- Long time span (2005-2012 accrual) means practice patterns, imaging technology, and molecular diagnostic standards evolved during and after the trial.
Funding
Unrestricted educational grant and drug supply from MSD-Merck & Co; molecular analyses supported by Swiss-Bridge Award 2011 and Swiss Cancer Research Foundation (KFS-2949-02-2012, KFS-5555-02-2022); EORTC by NCI grants 5U10 CA011488-35 through 2U10 CA011488-41; UK by NIHR National Cancer Research Network and Cancer Research UK (CRUK/07/032); NCIC CTG by Canadian Cancer Society Research Institute (015469, 021039); TROG by NHMRC (509094); EORTC Cancer Research Fund
Based on: EORTC (Journal of Clinical Oncology, 2026)
Authors: Baumert BG, Hegi ME, van den Bent MJ, et al.
Citation: J Clin Oncol. 2026 Jul 2:JCO2502735. DOI: 10.1200/JCO-25-02735
Content summarized and formatted by NeuroTrials.ai.