MR-Linac (MRL) adaptive radiotherapy, when added weekly to temozolomide treatment, significantly reduced radiation exposure to healthy brain tissue in patients with high-grade glioma, according to findings from the phase 2 UNITED trial (NCT04726397). The study, presented at the 2024 American Society for Radiation Oncology Annual Meeting (ASTRO), suggests a potential for improved quality of life and reduced toxicity through precise radiation targeting.
The trial enrolled 108 patients with IDH-wildtype glioma between April 2021 and May 2023, with 98 receiving treatment per protocol. Participants received either long-course radiotherapy (60 Gy in 30 fractions, n=59) or short-course radiotherapy (40 Gy in 15 fractions, n=39). The study population included patients with MGMT methylated (n=52), unmethylated (n=41), and indeterminate (n=5) disease.
Reduced Target Volume and Marginal Failure Rates
After a minimum of one year of follow-up, the intent-to-treat analysis revealed a marginal failure rate of 4% (n=4/98), demonstrating non-inferiority compared to historical marginal failure rates (P < .001). Notably, 81% of patients experiencing radiographic failure also exhibited central failure components. The adaptive radiotherapy approach led to a substantial reduction in treated clinical target volume (CTV) margins. The mean reduction was 40% compared to the European Organisation for Research and Treatment of Cancer (EORTC) method, which uses a 15 mm CTV margin. When compared to the Radiation Therapy Oncology Group (RTOG) method, which involves T2-weighted fluid-attenuated inversion recovery (FLAIR) plus 2 cm, the mean reduction was an impressive 71%.
Survival Outcomes
In the long-course radiotherapy group, the median progression-free survival (PFS) was 11.6 months in the UNITED trial, compared to 9.2 months in the phase 3 SPECTRO GLIO study (NCT01507506), which included a contemporary cohort of 129 glioblastoma patients treated with radiotherapy (60 Gy in 30 fractions) plus temozolomide using standard margins with CBCT-Linac Image-Guidance (P = .35). Median overall survival (OS) was 18.5 months in the UNITED trial and 20.5 months in the SPECTRO GLIO study (P = .53).
For patients receiving short-course radiotherapy, the median PFS was 6.8 months in the UNITED trial and 5.3 months in the CE.6 cohort from a previous phase 3 trial (NCT00482677) that used radiotherapy plus temozolomide (P = .2). The median OS was 10.6 months and 9.3 months, respectively (P = .8).
Investigator Commentary
"This is the first trial to investigate MR-guided adaptive on-line radiation with margin reduction. We’ve shown it is safe and feasible; marginal failures are relatively low," said Jay Detsky, PhD, MD, FRCPC, associate director of Medical Residency at Odette Cancer Centre and assistant professor in the Department of Radiation Oncology at the University of Toronto. "We have dramatically reduced the CTV and normal brain tissue that is exposed to radiation, but the real question is [whether] that going to truly improve quality of life and reduce toxicity. That is best determined with a randomized trial."
The researchers are also working to standardize FLAIR usage beyond 5 mm and are evaluating tumor dynamics and advanced imaging biomarkers within this patient population.
Methodology
All patients were treated on a 1.5 Tesla integrated MR-Linac device, using a pre-planned reference with MR simulation. The CTV was defined as the gross tumor volume (GTV)—the surgical cavity plus any residual enhancing tumor—plus a 5 mm margin, with anatomical barriers and FLAIR involvement considered at the discretion of the radiation oncologist. The planning target volume (PTV) was the CTV plus 3 mm. An adapt-to-shape workflow was used at the first radiotherapy fraction and then weekly, involving Gadolinium-enhanced on-line T1 with contrast and FLAIR, along with real-time contouring and planning. The remaining four fractions each week employed an adapt-to-position workflow with non-contrast imaging.
The primary endpoint of the study was the risk of marginal failure. Secondary endpoints included PFS and OS stratified by radiotherapy schedule, PFS per RANO-HGG 2.0 criteria, tumor dynamics, dosimetry, and imaging biomarkers. Eligible patients had IDH-wildtype glioblastoma or IDH-mutated grade 4 astrocytoma.
