ESTRO 2021 Abstract Book

S708

ESTRO 2021

The optimization problem was encoded in a cost function to minimize with respect to the beamlet weights, consistently to what is usually done in inverse planning. The function used is the following:

with i running over the number of voxels of object r, target or OAR, and j the total number of beamlets. a_ij represents the elements of the influence matrix, x_j the beamlet weights, D_i^(P) the dose prescription to each voxel and gamma_i its assigned priority. The cost-function was mapped to a Hamiltonian, where the x_j were discretized and codified as a systems of long-range interacting qubits (similarly to what will happen with quantum computers). The role of the TN algorithm is to look for the ground state of the system, namely the minimum of the cost-function. Preliminary studies were performed on simple geometries such as sperical targets in a box and then the algorithm was applied to a prostate cancer case. The problem was approached also using Quadratic Programming (QP) and Simulated Annealing (SA) to compare the solutions. Results Results on the prostate cancer case for a simple beam geometry and a small number of beamlets (64) are shown in the figure. The upper panels show the DVHs obtained with TNMs (blue), SA (green) and QP (pink) for the bladder (dotted line), the rectum (dashed line) and the prostate (solid line). In the lower panels, the green line represents TN - SA, while the pink line TN - QP, both for the rectum (left) and the prostate (right). The three algorithms provide solutions which are consistent within the 4% of volume coverage. Despite this is not a clinical result, it shows that TNMs can solve effectively the underlying optimization problem.

Conclusion This study showed that TNMs might become and alternative to standard algorithms for IMRT planning. This opens interesting scenarios about the application of these methods to more sophisticated cost functions to embed dosimetric, geometrical and radiobiological constraints, as well as about the future use of quantum computing in radiotherapy treatment planning.

Poster discussions: Poster discussion 25: Lung 2

PD-0872 Second In-field course of stereotactic body radiotherapy for thoracic tumors: a multicentre analysis C. John 1 , R. Dal Bello 1 , N. Andratschke 1 , M. Guckenberger 1 , J. Boda-Heggemann 2 , E. Gkika 3 , F. Mantel 4 , H.M. Specht 5 , C. Stromberger 6 , F. Zehentmayr 7 , O. Blanck 8 , P. Balermpas 1 1 Zurich University Hospital, Radiation Oncology, Zurich, Switzerland; 2 Medical Faculty Mannheim, Heidelberg University, Radiation Oncology, Mannheim, Germany; 3 University Medical Center Freiburg, Radiation Oncology, Freiburg, Germany; 4 University Hospital Würzburg, Radiation Oncology, Würzburg, Germany; 5 Clinic for Radiation Oncology Freising, Radiation Oncology, Freising, Germany; 6 Charite – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Radiation Oncology, Berlin, Germany; 7 Paracelsus Medical University, Radiation Oncology, Salzburg, Austria; 8 University Medical Center Schleswig-Holstein, Radiation Oncology, Kiel, Germany Purpose or Objective Data of in-field thoracic re-irradiation with two courses of stereotactic radiotherapy (SBRT) is scarce. Aim of this study is to investigate patterns of care, efficacy and safety and to analyze cumulative dose distributions. Materials and Methods Patients with a second SBRT-course in the lung, planning target volume (PTV) overlap and available treatment plans were analyzed in this retrospective, multicenter study. All plans and clinical data were centrally collected. Radiotherapy plans were subjected to summation and dosimetric evaluation (figures 1 and 2). All dosimetric and volumetric parameters were extracted using the software MIM (version 6.9.2, MIM Software Inc., Cleveland, USA). The applied dose was converted to the corresponding 2 Gy equivalent dose (EQD2) distribution using the parameters alpha/beta = 10 Gy (EQD2/10) for the tumor and alpha/beta = 3 Gy