Set-up uncertainty during breast radiotherapy: Image-guided radiotherapy for patients with initial extensive variation

Purpose: The aim of this work was to establish a customized strategy for image-guided radiotherapy during whole breast irradiation. Risk factors associated with extensive errors were assessed. Methods and materials: A series of 176 consecutive breasts in 174 patients were retrospectively assessed. Electronic portal images from 914 medial and 807 lateral directions were reviewed. On the basis of the chest wall, the deviations between the simulation and each treatment were measured. The systematic (Σ) and random error (σ) of population, and the planning target volume (PTV) margin (2 Σ + 0.7σ) were calculated for each direction. Extensive set-up errors were defined as the fraction over the PTV margins in any direction. For extensive set-up errors, χ² tests and logistic regression analyses were conducted. Results: The medial and lateral PTV margins for the right-left, superior-inferior, and anterior-posterior axes and the rotation of collimator were 2.6 and 2.4 mm, 4.6 and 4.6 mm, and 3.1 and 3.3 mm and 2.8 and 2.9 and cut-off values for extensive errors were 3, 5, and 4 mm and 3, respectively. In χ² tests, tumor in upper outer quadrant (p = 0.012) and chest wall thickness ≥ 2.0 cm (p = 0.003) for medial portals and age group (p = 0.036) for lateral portals were associated with extensive errors. In multivariate tests, the extensive error on the initial fraction had a high probability of extensive set-up errors in both medial (OR = 4.26, p < 0.001) and lateral portals (OR = 3.07, p < 0.001). Conclusion: In terms of the set-up uncertainty during breast irradiation, patients with extensive error in the initial treatment should be closely observed with serial image-guided radiotherapy.