S1532
Interdisciplinary - Quality assurance and risk management
ESTRO 2026
References: 1. Kim H et al. Radiat Oncol2020;15:83.2. Rossi M et al. Med Dosim 2022;47(1):59-65.3. enninkhof J et al. Tech Innov Patient Support Radiat Oncol 2022;23:16- 23.4. Parsons D et al. J Appl Clin Med Phys 2023;24(9):e14200.5. Rudat V et al. J Appl Clin Med Phys 2024;25(4):e14232.6. Mankinen M et al. Radiat Oncol 2024;19:137.7. Huijskens S et al. Radiother Oncol 2024;195:110229. Keywords: QA, Dose prediction ,Intrafraction Learnings and insights applying systems engineering hazard analysis to real-time radiotherapy Jonathan Hindmarsh 1 , Scott Crowe 2 , Jemma Walsh 2 , Tanya Kairn 2 , Sonja Dieterich 3 , Jeremy Booth 4,5 , Paul Keall 1 1 Image X Institute, University of Sydney, Eveleigh, Australia. 2 Cancer Care Services, Royal Brisbane and Women’s Hospital, Herston, Australia. 3 Department of Radiation Oncology, UC Davis Health, Sacramento, USA. 4 Northern Sydney Cancer Centre, Royal North Shore Hospital, St Leonards, Australia. 5 Institute of Medical Physics, University of Sydney, Camperdown, Australia Digital Poster Highlight 3679 Purpose/Objective: Real-time adaptive radiotherapy is increasingly accessible. Therefore, the need to assess the safety of the systems and the processes enabling them is even more important. Historically, failure modes and effects analysis (FMEA) has been the recommended and predominant method (1, 2). However, FMEA is limited by its theoretical underpinnings (3) and linear approach that struggles to handle complex systems (4). Given the complexity of real-time adaptive systems, other hazard analysis techniques are necessary. One example is system theoretic process analysis (STPA), a systems-based approach suited to complex systems (5). The goal of this research was to use STPA to evaluate the implementation of real-time adaptive treatment on a helical tomotherapy platform. Material/Methods: The latest evolution of the helical tomotherapy platform incorporates upgrades that facilitate real- time motion monitoring and treatment adaptation (6). An STPA of real-time adaptive radiotherapy treatment on this system was conducted in collaboration between a remote team with experience in hazard analysis and a clinical team from the radiation oncology department of a large public hospital. Figure 1 provides an overview of the STPA methodology.
residual shifts, dose reproducibility, and Δin vivo as a QA surrogate. Material/Methods: Eighty left-sided breast cancer patients treated with VMAT (two partial arcs) were prospectively assigned to four arms (n=20): (1) ABC-Gating + SGRT monitoring (reference); (2) ABC manual control + SGRT beam- gating (3 mm/2° thresholds); (3) SGRT-Only DIBH and beam-gating (3 mm/2° thresholds); (4) Hybrid ABC manual control + SGRT beam-gating with intrafraction correction between arcs. Pre- and post-treatment CBCTs were done for all patients. Mean 6D residual errors were calculated from translational and rotational shifts (X, Y, Z, pitch, roll, yaw) extracted from ΔCBCT and Align-RT logs. Changes in PTV D95, heart Dmean (<5Gy), and lung (V16 Gy <30%) were quantified by simulating shifts on Monaco TPS. Nanodot OSLDs In-vivo dosimetry put on patient reference points and analyzed relative to a unified baseline in the first fraction (Δ=0). SPSS v28 used for statistical analysis. Results: All datasets showed normal distribution (Shapiro–Wilk p> 0.05). One-way ANOVA—Bonferroni correction— yields significant inter-arm differences (p < 0.001) and strong Pearson correlations (p<0.05) between geometric residuals, dosimetric variations, and Δin- vivo. The hybrid Arm4 ABC + SGRT-gating with intrafraction correction achieved the highest geometric precision, showing mean translational and rotational shifts of 0.3 ± 0.2 mm and 0.4 ± 0.2°, respectively, versus 1.6 ± 0.6 mm and 0.9 ± 0.4° in the reference ABC + SGRT monitoring (Arm1) (p< 0.01). In Arm1, maximum per-fraction drifts reached 2.5 mm / 1.6°. Geometric drift was significantly reduced in Arm 4 (p<0.01), and dosimetric stability was obtained with deviations below 5%, 4%, and 3% for PTV D95, heart Dmean, and lung V16, respectively. Residual shifts of ΔCBCT has moderate to strong correlation with Allig-rt (R² = 0.74-0.45). Δin-vivo variations were minimal (<1% mean, max 2.2%), while the reference arm1 exceeded 3% on average (max 5%). Δin-vivo and mean residual shifts showed strong Linear correlation (R² = 0.78, p< 0.001). Conclusion: Integrating SGRT beam-gating and intrafraction correction within ABC-controlled DIBH significantly enhances geometric accuracy and dosimetric reproducibility in left-sided breast radiotherapy. AlignRT can be used as an alternative for CBCT setup and ABC DIBH control. Strong correlation between Δin- vivo deviations and residual shifts implements OSLD in-vivo dosimetry as a practical tool as real-time quality-assurance surrogate for verifying DIBH treatment consistency.
Made with FlippingBook - Share PDF online