The Radiation Oncology Feed For radiation oncologists · physicists · RTTs · dosimetrists
technical-relevance Protons treatment-planning

Beam-deflection compensation restored proton dose delivery inside an in-beam MR scanner

Energy-layer-wise compensation restored millimetre-level proton targeting despite magnetic-field beam shifts approaching 30 mm in an MR-integrated prototype.

RadOncToday · ·1 min read

KEY POINTS

  • The study developed an energy-layer-wise proton beam-deflection compensation method for an MR-integrated proton therapy prototype using a 0.32-T transverse magnetic field. A research version of RayStation with magnetic-field-aware Monte Carlo calculation was used to model energies from 100–220 MeV and generate the required steering corrections.
  • Magnetic deflection was clinically substantial: previous measurements in the same prototype showed lateral shifts of approximately 33 mm at 100 MeV and 21 mm at 220 MeV at the extended isocenter. The new method compensates each energy layer by modifying pencil-beam scanning steering positions before delivery.
  • Experimental validation used three spread-out Bragg-peak box fields across different energy ranges, with measurements inside the actual in-beam MR scanner. Under 3%/3-mm gamma with a 10% threshold, all configurations exceeded 97% passing, and most remained above 90% even with the stricter 2%/2-mm criterion.
  • Absolute dose agreement at the center of the spread-out Bragg peaks was within 1.2–2.2%, while measured and calculated proton ranges agreed within 0.8–1.1 mm, indicating that lateral compensation did not materially compromise depth-dose delivery.
  • A clinically shaped plan was then tested in an anthropomorphic head phantom with a range shifter. All measurement planes achieved >98% gamma passing at 3%/3 mm, while the deepest plane still reached 92.4% at 2%/2 mm. Spatial agreement between calculated and measured dose was within ±1.6 mm.
  • In the compensated head plan, CTV D95 was 58.1 Gy(RBE), or 96.8% of the 60-Gy(RBE) prescription, and D98 was 57.4 Gy(RBE), or 95.7%. Brainstem, ocular and other evaluated organ-at-risk constraints remained satisfied.
  • The method still does not perform full inverse optimization directly in the magnetic field; plans are first optimized without the field, then spot positions are corrected and dose recalculated with the magnetic field included. Full TPS commissioning, dedicated QA and first-in-human validation remain necessary before routine clinical use.

CLINICAL TAKEAWAY

A major obstacle to MR-guided proton therapy is straightforward physics: the MRI magnetic field bends the proton beam. This work shows that energy-specific steering can compensate centimetre-scale deflection while retaining millimetre-level dosimetric accuracy in realistic measurements. It is an important translational step, but not yet evidence that MR-integrated proton therapy improves patient treatment.

SOURCE

International Journal of Radiation Oncology, Biology, Physics