Advanced PET and MRI identify additional glioblastoma volumes beyond conventional 20-mm margins

In 153 glioblastoma patients, combined 18F-DOPA PET and advanced MRI identified additional imaging-defined volumes beyond a 20-mm conventional MRI expansion in 20%.

KEY POINTS

  • This prospective imaging analysis was conducted within a randomized phase II glioblastoma trial (NCT05781321) across three Mayo Clinic tertiary centers and associated regional sites. Between April 2023 and July 2025, 170 patients were enrolled, of whom 153 underwent complete advanced imaging and treatment. Eligible patients had newly diagnosed, histologically confirmed IDH-wildtype glioblastoma, with postoperative or postbiopsy enhancing disease plus surgical bed measuring ≤6 cm. The parent trial's primary overall survival results remain pending.
  • All patients underwent conventional planning MRI, 18F-DOPA PET/CT, diffusion-weighted MRI, and dynamic contrast-enhanced perfusion MRI. Advanced imaging subvolumes were delineated independently and combined with the surgical cavity and contrast-enhancing tumor to generate the treatment target. PET segmentation used a tumor-to-normal uptake ratio >2.0, while diffusion and perfusion abnormalities were identified using patient-specific thresholds with physician review. Eleven patients (7%) had no PET abnormality above the prespecified threshold, and one had no qualifying diffusion abnormality.
  • Advanced imaging identified additional volumes outside conventional MRI-based treatment expansions in a substantial proportion of patients. Combined advanced imaging extended beyond the standard contrast-enhancing MRI volume plus 5, 10, 15, and 20 mm margins in 86%, 56%, 38%, and 20% of patients, respectively. The analysis classified any additional volume >0.03 cm³ as outside the reference expansion. These findings demonstrate spatial discordance but do not establish that all additional regions contain viable microscopic or macroscopic tumor.
  • PET, diffusion MRI, and perfusion MRI provided complementary rather than fully redundant information. Perfusion-defined abnormalities extended outside the combined PET and diffusion volumes in 152/153 patients (99%). Diffusion-defined volumes extended beyond PET in 146 patients (95%), while PET-defined abnormalities extended beyond diffusion in 138 patients (90%). Among 126 patients with all three advanced imaging subvolumes available for the overlap analysis, 89% had regions not completely shared between the three techniques.
  • Advanced imaging was incorporated into both conventional and accelerated radiotherapy strategies. Seventy-one patients received standard-fractionation RT and 82 received accelerated treatment. Standard schedules delivered 60 Gy in 30 fractions or 40 Gy in 15 fractions to the combined imaging-defined high-dose GTV, with lower-dose treatment to an anatomically constrained 10–20 mm expansion. Accelerated schedules delivered 40 Gy in ten fractions or 35 Gy in five fractions to PET/advanced-MRI-defined targets, with lower-dose treatment incorporating a 5-mm CTV expansion. Photon or proton therapy was permitted.
  • Conventional MRI and advanced imaging also identified different portions of the postoperative target. No patient's conventional MRI GTV was entirely contained within the advanced PET/MRI-defined regions, reflecting the inclusion of postoperative cavity components without corresponding advanced imaging abnormalities. The median conventional MRI GTV volume not overlapping the advanced imaging subvolumes was 10 cm³. This illustrates why advanced imaging cannot simply replace conventional postoperative anatomical delineation.
  • The study evaluates imaging geometry, not validated tumor extent or treatment efficacy. Although earlier biopsy studies support the biological relevance of selected advanced imaging abnormalities, the current analysis did not systematically obtain pathological confirmation of the additional volumes. It also did not compare recurrence patterns, neurological toxicity, progression-free survival, or overall survival between planning strategies. Moreover, the inclusion of extremely small additional volumes in the overlap analysis means that the frequency of discordance does not directly quantify clinically meaningful geographic miss.

CLINICAL TAKEAWAY

Combining amino acid PET with diffusion and perfusion MRI identifies additional imaging-defined abnormalities beyond conventional contrast-enhancing glioblastoma volumes, including beyond standard anatomical expansions in selected patients. The modalities provide complementary information that may improve biological target characterization and support future individualized dose painting or hypofractionation strategies. However, clinical benefit and the accuracy of these additional target regions require validation against tumor pathology, recurrence patterns, and mature prospective trial outcomes.

SOURCE

International Journal of Radiation Oncology, Biology, Physics

Browse more research Suggest a correction