Purpose: To assess the impact of anatomical target location on intrafracton target motion for SBRT lung treatments.
Methods: 4DCBCTs from 89 SBRT lung patients treated between 2018-2019 were analyzed retrospectively. Patients were grouped based on target location (50 upper lobe (UL), 8 middle lobe (ML), 31 lower lobe (LL)). GTV defined at phase-0 of planning CT was propagated to phase images of pre-delivery (with couch correction applied) and post-delivery 4DCBCTs using deformable image registration (Elekta/Admire). Target centroid was used to calculate motion amplitude and mean target position (averaged over 10-phases). Differences in amplitude and mean target position between pre-delivery and post-delivery 4DCBCT was used to assess intrafraction motion variation (IFV). Results were compared to institutional PTV margins of 5mm (anteroposterior, mediolateral) and 6mm (craniocaudal) and statistical significance evaluated with ANOVA.
Results: Absolute intrafraction target motion amplitude differences in mediolateral and anteroposterior directions was similar between lobes. In the craniocaudal direction, significant differences were seen (p<0.005) between lobes, with mean±sd [range] of 1.15±1.21 [0.0-6.0]mm, 0.75±0.63 [0.0-2.6]mm, and 0.57±0.52 [0.0-2.7]mm for LL, ML and UL. Absolute mean target position differences were statistically significant in mediolateral (p=0.03) and craniocaudal (p<0.005) directions, while anteroposterior was not (p=0.09). In the craniocaudal direction mean±sd [range] differences were 2.00±1.78 [0.01-13.13]mm, 1.56±1.32 [0.06-4.79]mm, and 1.35±1.44 [0.01-8.77]mm for LL, ML, and UL. Comparison of mean target position differences to PTV margins found 0% of treatments had ≥5mm difference in the mediolateral direction for all lobes, whereas 5.26%(LL), 3.13%(ML) and 0.42%(UL) had ≥5mm difference in anteroposterior direction and 1.50%(LL), 0.0%(ML) and 1.26%(UL) had ≥6mm difference in craniocaudal direction.
Conclusion: Absolute IFV difference in target motion amplitude and mean target position were largest in the craniocaudal direction for LL lesions, with some lesions surpassing existing PTV margins. Further investigation is needed to determine the dosimetric impact of the IFV differences observed.