White-Matter Informed Tumour Spread Mapping and Diffusion/Perfusion MRI for Personalized GBM Targeting
Abstract
Purpose
Standard glioblastoma (GBM) radiotherapy planning uses uniform margins to expand the gross tumour volume (GTV) to the clinical target volume (CTV) and planning target volume (PTV). This approach does not account for patient-specific microscopic infiltration or biological heterogeneity, potentially undertreating infiltrated tissue while increasing dose to normal brain. This study extends our Tumour Spread (TS) mapping method, which models microscopic infiltration along white matter pathways to generate patient-specific CTVs, by incorporating a biological target volume (BTV) derived from diffusion and perfusion MRI. The aim was to assess the feasibility of further personalizing target volumes by comparing three planning strategies: (1) conventional GTV with fixed-margin CTV/PTV, (2) GTV with TS-derived personalized CTV and PTV (pCTVG/PTVG), and (3) BTV with TS-derived personalized CTV and PTV (pCTVB/PTVB), evaluating target coverage, normal tissue sparing, and spatial overlap with recurrence.
Methods
MRI data from ten newly diagnosed GBM patients were analyzed. BTVs were delineated using Fractional Anisotropy, Apparent Diffusion Coefficient, and relative cerebral blood volume maps. TS maps were generated based on both the GTV and BTV. VMAT plans were developed for three planning strategies, and comparisons were made across target volumes, organ-at-risk (OAR) doses, and spatial overlap with recurrence regions identified on follow-up imaging.
Results
Both personalized CTVs (pCTVG and pCTVB) were smaller than the standard CTV, reducing treated volume while maintaining tumour coverage. pCTVG and pCTVB reduced the standard CTV margin volume by 59% and 49%, respectively. All plans met conformality criteria, and both personalized strategies lowered mean OAR doses. Recurrence coverage was similar across strategies: 98% (standard PTV), 95% (PTVG), and 97% (PTVB); ≥80% dose covered 99% of recurrence with the standard PTV and 98% with both PTVG and PTVB.
Conclusion
This study demonstrated the feasibility and potential benefits of MRI-derived, biology-informed personalized target volumes for radiotherapy planning of glioblastoma patients.