Robust radiation therapy optimization using simulated treatment courses for handling deformable organ motion

Abstract

We describe a radiation therapy treatment plan optimization method that explicitly considers the effects of interfraction organ motion through optimization on the clinical target volume (CTV), and investigate how it compares to conventional planning using a planning target volume (PTV). The method uses simulated treatment courses generated using patient images created by a deformable registration algorithm to replicate the effects of interfraction organ motion, and performs robust optimization aiming to achieve CTV coverage under all simulated treatment courses. The method was applied to photon-mediated treatments of three prostate cases and compared to conventional, PTV-based planning with margins selected to achieve similar CTV coverage as the robustly optimized plans. Clinical goals for the CTV and healthy tissue were used in comparison between the two types of plans. Out of the two clinical goals for overdosage of the CTV, the three robustly optimized plans violated respectively 2, 2, and 0 goals in the mean over the scenarios, whereas none of the PTV plans violated these goals. Of the ten clinical goals for rectum, bladder, anal canal, and bulbus, the robustly optimized plans violated respectively 0, 1, and 1 goals in the mean, whereas the PTV plans violated 5, 7, and 4 goals. Compared to PTV-based planning, the inclusion of treatment course scenarios in the optimization has the potential to reduce the dose to healthy tissues while retaining a high probability of target coverage. This may reduce the need for adaptive replanning.

Effectiveness of robust optimization against geometric uncertainties in TomoHelical planning for prostate cancer

“…The target coverage (TC) and the homogeneity index (HI) 33 were calculated for all nominal and perturbed dose distributions in both the robust and PTV plans.…”

“…These were created from the original planning CT by simulating organ motion functionality based on a non-rigid deformable image registration algorithm in RayStation. 32 The dose distributions were recalculated for the six CT images using the same planning conditions for each nominal plan. Consequently, 12 perturbed dose distributions were created for each patient (six for the robust plan and six for the PTV plan).…”

Adaptive dose painting for prostate cancer

“…For the SV, we used the lower (LR: 5, AP: 7, SI: 7 mm) and upper limit (LR: 6, AP: 9, SI: 9 mm) of the anisotropic margins specified in the ESTRO reference ( 20 ) for the adaptive and conventional workflows, respectively. Homogeneous plans were optimized with standard CTV-to-PTV margins whereas DPBN plans were created using minimax optimization ( 24 ) with “robustness distances” set to the same values as the CTV-to-PTV margins; the minimax optimizer generates a set of treatment scenarios with patient setup displacements along three axes, and aims to find a plan which is optimal for the worst case of these scenarios (i.e., a plan which is robust to geometric uncertainties). For each objective, the software allows it to be set as ‘robust’ or not, i.e.…”

Time‐resolved MRI for off‐line treatment robustness evaluation in carbon‐ion radiotherapy of pancreatic cancer

“…We indeed plan to adopt the TR vCTs to better describe target motion for the selection of the proper motion compensation strategy and to evaluate additional uncertainties scenarios during robust optimization of the treatment plan. 37 , 38 , 39 , 40 In perspective, the proposed approach could be also replicated to perform 4D dose optimization 41 , 42 and dose accumulation, 43 , 44 as well as it could be integrated with dose variation models 45 , 46 to explicitly take into account TR 3D deformations. The approximate time required to estimate the vCTs used for dose recalculation was 40 min; this is reasonable for off‐line evaluations, but improvements in the implementation could speed up TR vCTs generation.…”

“…Overall, our work demonstrated the robustness of gated CIRT in pancreas tumors and suggests the use of TR vCTs derived from MRI to support treatment planning. We indeed plan to adopt the TR vCTs to better describe target motion for the selection of the proper motion compensation strategy and to evaluate additional uncertainties scenarios during robust optimization of the treatment plan 37–40 . In perspective, the proposed approach could be also replicated to perform 4D dose optimization 41,42 and dose accumulation, 43,44 as well as it could be integrated with dose variation models 45,46 to explicitly take into account TR 3D deformations.…”