Dosimetric impact of surface-measured anatomical changes in breast radiotherapy with SGRT

This study quantified the dosimetric impact of anatomical changes in whole-breast irradiation (WBI) using surface-guided radiotherapy (SGRT) data and derived thresholds for adaptive decision-making.

Seventeen patients (57 treatment days) were analyzed using daily 3D AlignRT (VisionRT) surface captures. An in-house SGRT-to-CT deformable registration pipeline, SurVolT, replaced the planning CT body contour with the daily surface as a deformation constraint. The original plan was reapplied to each deformed CT and recalculated without re-optimization (AcurosXB, dose-to-medium). Two surface-change metrics were computed: (i) signed radial thickness change (cm) and (ii) signed swelling/shrinkage volume (cm³). Dosimetric outcomes were target coverage (PTV V95 %, CTV D95 % and D98 %) and hotspot burden (V107 % > 5 cm³)..Predictive accuracy and thresholds were determined via ROC analysis.

Breast swelling consistently caused under-coverage, while shrinkage correlated with superficial hotspots. A swelling threshold of 1.11 cm (95 % CI: 1.11-1.74) or 9.9 cm³ (95 % CI: 9.9-22.8) perfectly predicted V95 % failures (AUC = 1.00, sensitivity = 1.00, specificity = 1.00). For hotspots, shrinkage > 0.93 cm (95 % CI: 0.93-1.59) yielded AUC = 0.92, and a shrinkage-volume > 4.9 cm³ achieved AUC = 0.97.

SGRT-derived metrics, particularly radial thickness change, provide fast, interpretable estimates of dose degradation. Swelling > ≈ 1.1 cm should prompt coverage reassessment, whereas shrinkage > ≈ 0.9 cm signals elevated hotspot risk. These pragmatic thresholds can be derived from routine SGRT workflows and support adaptive decision-making in WBI, with complementary CTV analyses supporting their clinical consistency.