Dosimetric robustness of tumour- and fiducial-based tracking in Radixact lung stereotactic body radiotherapy: a four-dimensional CT-based simulation using imaging-derived tracking-error scenarios

Abstract Objectives To compare the dosimetry and robustness of two Radixact Synchrony lung stereotactic body radiotherapy (SBRT) tracking strategies: tumour-based lung tracking with respiratory modelling (LwR) and fiducial-based tracking with respiratory modelling (FwR). Methods Ten stage I non–small–cell lung cancer patients were retrospectively analysed. SBRT plans (42 Gy in 4 fractions) were generated on maximum-exhalation computed tomography (CT) using planning target volumes (PTVs) created by 3–7 mm isotropic gross tumour volume (GTV) expansion. Each plan was mapped to all four-dimensional CT (4DCT) phases via isocentre shifts from phase-specific GTV (LwR) or individual fiducial displacement (FwR). Deformable dose accumulation yielded LwR, nearest FwR, and farthest FwR simulated plans. GTV/PTV V100%, representative organ-at-risk (OAR) doses, and robustness versus target localisation error, tumour–fiducial distance, and tumour volume were evaluated. Results With a 3-mm PTV margin, FwR showed larger losses in GTV coverage and PTV robustness than LwR (maximum decreases of 9.0% and 26.0%). FwR was less robust, particularly with unfavourable tumour–fiducial geometry. Larger PTV margins progressively mitigated FwR coverage degradation while maintaining acceptable OAR doses. Conclusion For lung SBRT delivered with Radixact Synchrony, LwR was more robust than FwR in this cohort, for which all tumour–fiducial distances exceeded 20 mm. When FwR is required, particularly under suboptimal tumour–fiducial geometry, a 3-mm margin should be used cautiously, and patient-specific evaluation of 5–7 mm margins within OAR dose constraints may be necessary. Advances in Knowledge A 4DCT-based simulation using imaging-derived tracking-error scenarios reveals dosimetric robustness differences between LwR and FwR tracking.