ESTRO 2024 - Abstract Book

S4091

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2024

2551

Digital Poster

Improve CTV coverage with a dose-guided plan-of-the-day selection in adaptive RT of cervical cancer

Delphine Lebret 1 , Caroline Lafond 1 , Julie Leseur 2 , Anaïs Barateau 1 , Diane Chan Sock Line 1 , Karine Peignaux 3 , Nathalie Mesgouez-Nebout 4 , Magali Le Blanc-Onfroy 5 , Chantal Hanzen 6 , Nedjla Allouache 7 , Sophie Renard-Oldrini 8 , Florence Le Tinier 9 , Renaud De Crevoisier 1 , Antoine Simon 1 1 Univ Rennes, CLCC Eugène Marquis, LTSI – UMR 1099, Rennes, France. 2 Centre Eugène Marquis, Radiation Oncology, Rennes, France. 3 Centre Georges-François Leclerc, Radiation Oncology, Dijon, France. 4 Institut de Cancérologie de l'Ouest–Site Paul Papin, Radiation Oncology, Angers, France. 5 Institut de Cancérologie de l'Ouest-Site Saint-Herblain, Medical Physics, Nantes Saint-Herblain, France. 6 Centre Henri Becquerel, Radiation Oncology, Rouen, France. 7 François Baclesse Cancer Centre, Radiation Oncology, Caen, France. 8 Centre Alexis Vautrin, Radiation Oncology, Vandoeuvre-lès-Nancy, France. 9 Oscar Lambret Cancer Center, Radiation Oncology, Lille, France

Purpose/Objective:

For locally advanced cervical cancer, one adaptive image guided radiotherapy (ART) strategy consists in a treatment plan library with different target volumes depending on bladder filling. In this case, plan-of-the-day (PoD) is usually selected by radiation oncologist only based on projection of the planned clinical target volumes (CTV) on daily CBCT (Manual-ART). We proposed to evaluate the impact on dose for OARs and CTV if the daily selection of PoD is performed based on dosimetric coverage of the daily CTV (D95-ART) compared to Manual-ART and to single plan treatment (No ART).

Material/Methods:

Data from 49 patients with cervical cancer included in a phase II ART clinical trial (ARCOL) were collected from 7 clinical centers. For each patient, three planning CT scans (158 CT scans in total) were acquired with different bladder fillings (empty, intermediate and full). A treatment plan was performed on each CT, with IMRT of 1.8 Gy per session for a total dose of 45 Gy to the pelvis. The clinical target volume included: the gross tumour volume defined on MRI; cervix ; parametria; uterus; at least upper half of the vagina according to primary clinical extend; the utero-sacral ligaments and lymph nodes (common, external, internal, and presacral). The planning target volume (PTV) was defined as 1cm isotropic expansion of the primary CTV and a marge between 0.5 and 1 cm isotropic for the nodal CTV. Daily CBCT was acquired and rigidly registered to the CT, 3 PTVs (empty, intermediate and full bladder plans) were projected on this image. The optimal treatment plan was chosen daily by the radiation oncologist on visual criterion: best uterus coverage and sparing of OARs, in particular the bladder. If there was a doubt between 2 plans, the dose planning with intermediate bladder was chosen. A deep learning model (nnU-Net) was trained, on a separate database, to segment CTV, rectum and bladder in CBCT scans. Each resulting automatically segmented daily CTV was assessed by a radiation oncologist into three categories: good, intermediate and poor. CBCT scans associated to a poor segmentation were excluded. To estimate the daily dose distributions, planned dose distributions were rigidly propagated to CBCT scans using bone structures. For each CBCT scan, three approaches were simulated, using: (i) the plan corresponding to intermediate bladder (“no-ART”); (ii)