Trigger-based adaptive proton therapy and changes in target coverage and cardiac dose in esophageal cancer

Inter-fractional anatomical variations during proton therapy for esophageal cancer may compromise target coverage and increase dose to surrounding organs. Adaptive replanning can mitigate these effects, but its cumulative dose-based and clinical impact remains insufficiently characterized. We evaluated offline adaptive intensity-modulated proton therapy (IMPT) triggered by target underdose or cardiac overdose using deformable dose accumulation.

A retrospective cohort of 192 esophageal cancer patients treated with IMPT was analyzed. Eighty-five underwent adaptive replanning and 107 were treated non-adaptively. Dose accumulation was performed by assigning each fraction to the closest weekly repeat computed tomography (CT) scan and deformably mapping doses to the planning CT. Accumulated doses from adapted and non-adapted plans were compared for internal target volume (ITV) and gross tumor volume (GTV) coverage, target hotspots, heart and lung dose, and normal tissue complication probability (NTCP) for 2-year post-treatment mortality.

Compared with the non-adaptive treatment scenario, the adaptive scenario increased ITV D_98% by 1.9 percentage points (pp) and GTV D_98% by 0.8 pp. in the target underdose subgroup. In the cardiac overdose subgroup, mean heart dose decreased by 1.0 Gy and heart V_40Gy by 1.7 pp., with no increase in mean lung dose. NTCP for 2-year post-treatment mortality decreased by 1.5 pp. in this subgroup.

Trigger-based adaptive IMPT improved cumulative target coverage and cardiac sparing without increasing lung dose. Target underdose and cardiac overdose triggers provided complementary benefits, supporting their use in structured adaptive replanning.