ESTRO 2021 Abstract Book

S628

ESTRO 2021

correlated with the g% passing-rate (Figure 1a). For the 1mm-1% criteria, a traffic light protocol was adopted: plans with MCS≥0.35 can be considered as safe to treatment (green-light), plans with 0.15

Conclusion A model for the prediction of plan delivery accuracy has been successfully implemented using the correlation between the modulation complexity score and the gamma-index pass rate for VMAT fluence. A traffic light protocol is able to differentiate over-modulated plans before pre-treatment QA from high-modulated and moderate-modulated plans. PD-0793 Which metric quantifies dose fall-off for SRS treatments of brain lesions? T. Nano 1 , D. Capaldi 2 , C. Chuang 2 , L. Wang 2 , S. Soltys 3 , S. Braunstein 1 , L. Ma 1 1 University of California, San Francisco (UCSF), Radiation Oncology, San Francisco, USA; 2 Stanford University, Radiation Oncology, Palo Alto, USA; 3 Stanford University , Radiation Oncology, Palo Alto, USA Purpose or Objective Dose fall-off guidelines for stereotactic radiosurgery (SRS) are critical for ensuring normal brain sparing. Two well-known but distinctive dose fall-off metrics are (1) dose gradient index (DGI) and (2) Paddick gradient index (PGI). However, intrinsic correlations between these two indices has not been studied. We investigated such a correlation and the pros and cons of these two indices for SRS treatment of brain lesions. Materials and Methods Two cohorts of patient cases with multiple brain lesions from two institutions were treated with CyberKnife (n=72, 202 lesions, volume 2.94 ± 6.99 mL [range:0.01–41.0]) and GammaKnife (n=435, 1983 lesions, volume 1.23 ± 2.77 mL [range:0.01–28.89]). Dose metrics were extracted to compute DGI, where 100 is considered “ideal” and decreases with worst fall-off, and PGI, which increases with worst fall-off. Simple linear regression of DGI and PGI was compared for both CyberKnife and GammaKnife cases, and Fisher Z- transformation was used to compare the correlation coefficients. Results For CyberKnife cases, targets smaller than 1mL have DGI and PGI values of 89.96±17.67 and 8.51±4.14 respectively, and targets larger than 1mL have values of 58.18±19.59 and 4.26±2.41 respectively. GammaKnife cases, targets small than 1mL have a DGI and PGI value of 108.61±8.94 and 4.03±2.50 respectively and targets larger than 1mL have values of 78.73±18.40 and 3.08±0.85 respectively. DGI was found to positively correlated with PGI (slope=0.2626, R²=0.002, p=0.7039) for CyberKnife cases and negatively correlated with PGI for GammaKnife cases (slope=-0.9163, R²=0.009, p<0.0001). CyberKnife and GammaKnife plan relationship between DGI and PGI were found to be different (p=0.0139) indicating the DGI is flawed in quantifying dose fall-off. Conclusion PGI was found to best describes dose fall-off, whereas, DGI was found to be only weakly correlated with PGI and indicates that it is not a direct measure of dose fall-off. Guidelines for dose fall-off in SRS treatments of multiple brain lesions should be provided exclusively using PGI, with special consideration for small lesions and complex targets.

Poster discussions: Poster discussion 12: Patient-centred radiotherapy services

PD-0794 Effectiveness of Dynamic Prioritization for patient scheduling: A Discrete-Event Simulation Model M. Gurjar 1 , J. Lindberg 1,2,3 , E. Abel 4,5 , C. Olsson 1,3 1 Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Department of Radiation Physics, Gothenburg, Sweden; 2 Sahlgrenska University Hospital, Department of Medical Physics and Biomedical Engineering, Gothenburg, Sweden; 3 Regional Cancer Centre West, Western Sweden Healthcare Region, Gothenburg, Sweden; 4 Institute of Clinical Sciences, Sahlgrenska Academy, University of Gothenburg, Department of Oncology, Gothenburg, Sweden; 5 Sahlgrenska University Hospital, Department of Oncology, Gothenburg, Sweden