Session Item

To evaluate the dosimetric accuracy of Eclipse HyperArc (HA) plans to perform SRS of multiple brain targets. To investigate the feasibility of using a Monte Carlo free software for independent verification of HA plans.

A slab RW3 phantom (16 cm x16 cm x 14 cm) was CT scanned and imported into the Eclipse v 16 TPS. Three spherical targets with the same diameter were outline: one at the center and the other two located at 7 cm distance from the phantom center (superior-right , and inferior-left, Fig 1). The centers of three targets were placed at the middle coronal plane. Contouring was done for three diameter values: 5, 8 and 10 mm, resulting three target sets. For each one, a single-isocenter HA plan was designed to simultaneously treat the three targets. Each HA plan consisted of 4 non-coplanar VMATs of 6 MV photons beams from a Varian TrueBeam linac equipped with a Millennium 120 MLC. The Accuros XB v 16.1.0 algorithm with a calculation grid size of 1.25 mm was used. Plans were also simulated with the PRIMO v 0.3.64.1800 Monte Carlo software (www.primoproject.net), using the phase space files provided by Varian for 6 MV photon beams of a TrueBeam (www.myvarian.com/montecarlo).

The dosimetric accuracies of the HA plans and PRIMO simulations were evaluated using EBT3 radiochromic film (RCF) dosimetry. Film pieces of 4 cm x 4 cm (same film batch) were sandwiched in the RW3 phantom at the planned positions of the targets. Delivery of each HA plan was repeated three times on different days. RCF dosimetry was performed using the Radiochromic.com v 3.3.6 web application. 2D global gamma analysis (Eclipse vs. film and PRIMO vs. film) was focused in the region around each target defined by the 20% dose threshold. Local dose differences at the center of each target (DD%) were also evaluated. 

1.       2D global (G) gamma passing rates (GPRs) for comparisons of Eclipse and PRIMO dose calculations vs RCF dosimetry are shown in Tables 1 and 2, respectively.

2.       Eclipse-EBT3 GRPs were > 95% for 3%(G)/1 mm, while PRIMO-EBT3 GRPs were > 95% for majority of cases for 5%(G)/1 mm.

3.       Peak dose differences (%DD) were within 3.5% for Eclipse-EBT3 comparison. %DD differences up to 8% were observed for PRIMO-EBT3 comparison.

4.       PRIMO dose statistical uncertainty (U) was within 2% (k=2), while the estimated U for our RCF dosimetry method was 4% (k=2).

Excellent dosimetric agreement with film dosimetry was found for the phantom-based HA plans calculated by Eclipse, including targets down to 5 mm-diameter (0.07 cm³). This study validates the use the PRIMO software for independent verification of the HA plans using a 5%(G)/1 mm criteria.