Abstract
BACKGROUND AND PURPOSE: Stereotactic ablative body radiotherapy (SABR) achieves excellent local control in medically inoperable non-small cell lung cancer. However, treating central and ultra-central tumours carries significant risks to adjacent critical structures, notably the central airways. This study evaluated the suitability of average value intensity projection (AvIP) imaging for airway delineation and assessed the adequacy of a 5 mm planning risk volume (PRV), by quantifying the influence of respiratory motion on dose to the airways using four-dimensional (4D) dose accumulation. MATERIALS AND METHODS: Eighteen patients (11 ultra-central, 7 central) treated with 60 Gy in 8 fractions were retrospectively analysed. Clinical 4D computed tomography (CT) datasets were used to capture 10 respiratory phases. Central airway contours (trachea, main bronchi, bronchial tree) were delineated on the AvIP and deformably propagated across all phases. Dose was recalculated per phase and accumulated back into a single distribution using the deformable registration for comparison with the AvIP-based plan. RESULTS: Maximum differences of 5 Gy were observed for the airway D(0.1 cm3) , with an average absolute difference of 0.95 ± 1.7 Gy. Notably, 28% of patients had differences >2 Gy, and 17% exceeded 3 Gy. The greatest dose discrepancies occurred in cases with tumours inferior to the first-order bronchi and with breathing motion >6 mm. Five patients' plans exceeded dose-volume constraints when respiratory motion was accounted for. CONCLUSIONS: Dose to the airways can be significantly underestimated when motion is not incorporated into treatment planning. A uniform 5 mm PRV may be inadequate in some central/ultra-central SABR cases. Patient-specific internal risk volumes merit investigation.