ESTRO 2021 Abstract Book

S1570

ESTRO 2021

Materials and Methods Fourteen HA SRS plans were included. Plans consisted of 4-5 non-coplanar VMAT arcs treating solitary lesions (3 cases) and multiple brain metastases with a single isocenter (11 cases, 2 to 35 targets per case). The target diameter ranged from 4 mm to 33 mm. Photon beams of 6 MV from a TrueBeam linac with a Millennium 120 MLC were used. Plans were computed with the Acuros XB v. 16.1.0 algorithm with a calculation grid size of 1 mm. It was configured using the Varian TrueBeam representative beam data and succesfully commissioned against measurements for clinical use. Plans were recalculated using three software systems for independent calculation: 1) PRIMO Monte Carlo software v. 0.3.64.1800 (www.primoproject.net), using the phase-space files (v. 2, Feb. 27, 2013) provided by Varian for 6 MV photon beams of a TrueBeam; 2) Varian Mobius 3D software (M3D, v. 3.1) with its preconfigured beam model, with the exception of the 1×1 cm2 output factor which was changed to match our measurements; and 3) Sun Nuclear DoseCheck v. 3.1, using the default beam model (v. 2016-09-30) for a 6 MV beam from a TrueBeam. For each plan, a 3D global (G) gamma analysis focused on each target was done (Eclipse vs PRIMO, Eclipse vs. M3D and Eclipse vs. Dosecheck). Gamma passing rates (GPRs) for a total of 83 targets and 2%(G)/2 mm, 3%(G)/1 mm and 5%(G)/1 mm criteria were collected. Action levels (AL= 100% - 1.96 SD, according to definition of TG-119 report) for the GPR metric were calculated for each criteria/independent software combination. A minimum AL of 90% is considered as adequate in this work. Results The table shows the GPRs (mean, SD and AL) for the comparisons of Eclipse with each software. The action levels derived for DoseCheck (all criteria) and PRIMO (for 5%/1 mm and 2%/2 mm criteria) were > 90%. For M3D, the action levels were <90%, showing that the M3D default beam model requires tuning for HyperArc SRS plans.

Conclusion Both DoseCheck and PRIMO (with the 6 MV Varian phase-space file) agree with Eclipse HyperArc calculations for a TrueBeam, with no need for the user to fine-tune the calculation parameters. The Mobius 3D default model, however, would need tuning to match HyperArc dose distributions PO-1842 Introduce a new rotational robust optimized Spot-scanning Proton Arc (SPArc) framework X. Ding 1 , S. Chang 2 , G. Liu 1 , L. Zhao 1 , W. Zheng 1 , A. Qin 3 , Y. Di 1 , X. Li 1 1 Beaumont Health, Radiation Oncology, Royal Oak, USA; 2 Wuhan Renmin Hospital, Radiation Oncology, Wuhan, China; 3 Beaumont Health, Radiation Onoclogy, Royal Oak, USA

Purpose or Objective Develop a new rotational robust optimization SPArc algorithm (SPArc rot

) to mitigate dosimetric impact from the rotational

setup error. Materials and Methods The SPArc rot

incorporate multi-CT robust optimization framework by taking into account such rotation setup error. The initial planning CT was rotated ±5° simulating the worst-case setup error in roll direction so it includes three CT image sets (initial, +5° and -5° planning CT). Five cases representing different disease sites such as brain, lung, head neck (HN), prostate, and liver cancer were selected to evaluate the plan robustness against the rotational error. Both conventional SPArc and SPArc rot plans were generated using the same translational robust optimized parameters. Both SPArc and SPArc rot plans were recalculated using a series of pseudo-CT introducing ±1°, ±2°, ±3°, and ±5° rotational setup error To quantitatively investigate the effectiveness of the rotational error mitigation effect. Dosimetric metrics such as D98%, D95% of CTV and 3D gamma analysis was used to assess the dose distribution changes in the target and patient body. Results The magnitudes of dosimetric changes in the targets due to rotational error were significantly reduced by the SPArc rot in all the cases. The uncertainties of the max dose to the brainstem and optic nerves in the brain case and max dose to spinal cord and esophagus in lung cases were significantly reduced using SPArc rot (Figure 1). The uncertainties of the mean dose to the OARs such as the liver and oral cavity, parotid is comparable between the two planning techniques. The gamma pass- rate (3%/3mm) was significantly improved for CTV of all tumor sites after SPArc rot corrections (Figure 2).