ESTRO 2024 - Abstract Book

S3428

Physics - Dose calculation algorithms

ESTRO 2024

1651

Digital Poster

The impact of ground-glass opacities on dose calculation accuracy for lung SBRT

Carla Cases, Marc Ballestero, Francesc Casas, Gabriela Antelo, Meritxell Mollà, Carles Gomà

Clinic Barcelona, Radiotherapy, Barcelona, Spain

Purpose/Objective:

The Anisotropic Analytical Algorithm (AAA) is, by far and large, the most widely used dose calculation algorithm in lung stereotactic body radiation therapy (SBRT), despite failing at accurately computing the dose in the density gradient region between the solid lung nodule and the less dense lung tissue surrounding it. Many studies have reported these deficiencies as a function of the size of the lesion or the location. Nevertheless, it is unclear whether these deficiencies still hold when ground-glass opacities (GGOs)—which are denser than lung— are present in the target region. The purpose of this study is therefore to investigate whether the dose calculation accuracy is correlated with the consolidation-to-tumor ratio (CTR) of the lung lesion.

Material/Methods:

To evaluate the impact of the algorithm used, we analysed the dose distribution inaccuracies in two data sets: a static physical phantom with a defined set of manually contoured spherical lesions and a series of 42 patients.

First, we used the CT scan of a thorax phantom (CIRS, Sun Nuclear, Norfolk). We contoured a set of spherical lesions in the lung region with diameters of 1 cm, 1.5 cm, 2 cm, 3 cm, and 4 cm; and CTRs of 0.00, 0.25, 0.50, 0.75 and 1.00 for every diameter. We manually assigned an AXB Physical Material Table value to each region inside the PTV: water (0 HU) for the solid kernel and lung (-400 HU) for the GGO shell (Figure 1). Regarding the clinical cases, we studied a cohort of 42 patients treated between 2017 and 2022 at our institution. We studied two scenarios: (i) the average reconstruction of a 4DCT scan simulating an ITV approach, and (ii) one of the respiratory phases of the 4DCT scan, to simulate a gated treatment. To evaluate dose inaccuracies in both datasets, we calculated volumetric modulated arc therapy (VMAT) dosimetries with a dose prescription of 55 Gy in 5 fractions. We optimized and calculated the dose distribution using AAA and then recalculated these plans using AXB (transport in medium, dose to medium) with fixed MUs. We analysed the relative differences between relevant dosimetric parameters: D98% (Gy) and D2% (Gy) on the PTV and Dmean on the PTV-ITV, in which the differences are expected to be more appreciable. We then analysed whether these differences were correlated with the CTR value.