Abstract

Mohammad Varasteh¹, Asmaa Mohammad Ali², Sergio Esteve³, Fabian Göpfert⁴, Prakash Jeevanandam³, Alan R Hounsell^3,1, Conor K McGarry^3,1

¹Queen's University Belfast, Patrick G Johnston Centre for Cancer Research, Belfast, United Kingdom; ²Queen's University Belfast, School of Mathematics and Physics, Belfast, United Kingdom; ³Belfast Health and Social Care Trust, Radiotherapy Physics, Belfast, United Kingdom; ⁴PTW-Freiburg, The Dosimetry Company, Freiburg, Germany

To implement a comprehensive methodology for individual pre-treatment estimation of the maximum dosimetric discrepancies caused by the interplay effect. This moving-platform-free approach can further be used for verification of VMAT plans.

In-house software has been developed to convert breathing position into a translation file, which in conjunction with a novel 3D dose reconstruction algorithm (alpha version) of the plan verification software VeriSoft (PTW-Freiburg, Germany), can be applied to measurements from a stationary OCTAVIUS 4D phantom (PTW-Freiburg, Germany) to allow dose reconstruction as if the phantom was moving. The developed software gives the possibility to define motion patterns with exact beam starting phases, enabling comparisons between reconstructed doses with different beam starting points (essential for interplay investigations). To validate our approach, 5 VMAT lung plans were measured with the stationary OCTAVIUS 4D. The same plans were then measured using the Varian TrueBeam developer mode (DM) to continuously move the treatment couch following a sinusoid wave (amplitude of 25 mm, period of 4s and 20s) with beam starting at 0% (end-inhalation) and 50% (end-exhalation) respiratory phases. Using the novel dose reconstruction method and the in-house developed software, static measurements were reconstructed with the same sinusoidal breathing pattern and beam starting phases. Gamma analysis of the motion-corrected stationary measurements was performed against those moving with a criterion of 2% dose difference and 2mm distances to agreement. To estimate the interplay effect, 13 lung SABR patients were investigated by comparing the motion-corrected static measurements starting at 0% respirational phase versus 50% using simple sinusoidal motion patterns. This was repeated using four different QUASAR breathing patterns representing more realistic tumour motions.

The validation analysis showed ≥99% of voxels passing a gamma criterion of 2%/2mm comparing the virtually moved dose distributions with the ones affected through real DM motion (Figure 1). For the interplay effect estimation using sinusoidal breathing patterns, the gamma results disclosed that >92% of points in the dose matrices meet the 2%/2mm criteria for >25 number of breaths (NBs) per delivery time (Figure 2). In case of the realistic breathing patterns, the threshold was found to be lower (>18 NBs).

An approach that gives individual pre-treatment estimation of the potential dosimetric impact of the interplay between organ motion and the dynamic beam was introduced. Although the interplay effect assessment disclosed the impact to be more severe for patients with longer breathing periods, this may be patient and plan dependent. Our approach can further be used for verification of ITV-based treatment plans using the patient respiratory pattern in order to identify whether the clinical goals can be achieved without explicit use of a respiratory management technique.