KEY POINTS
- The crossover study included 21 treated breast sides across 328 fractions: 15 whole-breast cases, six partial-breast cases, and two sequential boosts. Treatment was delivered in free breathing, generally to 40 Gy in 15 fractions.
- Conventional skin-marker setup was used for 127 fractions, while ExacTrac Dynamic surface-guided positioning was used for 201 fractions. Subsequent orthogonal imaging provided the reference for residual setup error.
- Mean population setup offsets did not differ significantly between methods. With surface guidance, 76%, 89%, and 96% of fractions were within 5 mm vertically, longitudinally, and laterally, compared with 70%, 79%, and 86% using skin marks.
- Surface guidance reduced vertical random error from 3.5 to 2.5 mm (p=0.03) and longitudinal random error from 3.6 to 2.4 mm (p=0.01).
- Lateral systematic error decreased from 2.5 mm with skin marks to 1.6 mm with surface guidance (p=0.04), indicating more consistent positioning despite similar average accuracy.
- Intrafraction monitoring was analysed across 183 fractions. Median vector displacement was 0.9 mm, with a 95th percentile of 2.9 mm; only 0.1–1.1% of delivered monitor units exceeded 3 mm in an individual translational direction.
- A small vertical inter-field shift of −0.3 mm was statistically significant (p<0.01), while longitudinal and lateral shifts were not. Because approximately one-quarter of surface-guided fractions still exceeded 5 mm vertically, confirmatory IGRT remained necessary.
CLINICAL TAKEAWAY
Single-camera surface guidance can replace skin marks for initial free-breathing breast positioning and provides useful continuous motion monitoring. It should not replace image guidance for final alignment, particularly when surgical clips or chest-wall anatomy determine treatment position.