KEY POINTS
- This dosimetric analysis included 16 patients enrolled in a prospective phase II trial of pencil beam scanning proton therapy for hepatocellular carcinoma. Patients received 58 or 67.5 Gy relative biological effectiveness in 15 fractions using exhale respiratory gating, robust optimization, and fiducial marker-based setup.
- Motion-inclusive dose reconstruction was completed for 76 fractions using 115 cone-beam computed tomography scans. The mean gating duty cycle was 58%, and mean cranio-caudal motion decreased from 17.6 to 8.2 mm during beam-on periods.
- Mean tumor-position error during spot delivery was 1.1 mm left-right, 2.9 mm cranio-caudal, and 1.6 mm anterior-posterior.
- At individual fractions, motion changed clinical target volume D98% by −1.9 ± 1.6 percentage points, D2% by +2.2 ± 1.5 percentage points, and homogeneity index by +4.1 ± 2.6 percentage points relative to the planned dose.
- When accumulated over 3–8 analyzed fractions, the corresponding changes decreased to −0.6 ± 1.0, +0.9 ± 1.4, and +1.5 ± 2.1 percentage points. Dose generally converged toward the plan after more than three fractions, while weekly control four-dimensional computed tomography scans showed static target-dose deviations below 0.2%.
CLINICAL TAKEAWAY
For 15-fraction proton therapy of hepatocellular carcinoma, exhale gating, robust planning, and fractionation substantially mitigated motion-related interplay effects despite meaningful single-fraction dose deviations. However, the study included only 16 patients, reconstruction was unavailable for approximately one-third of intended fractions, and the method is currently better suited to retrospective validation than routine per-fraction monitoring.
SOURCE
International Journal of Radiation Oncology, Biology, Physics