ESTRO 2023 - Abstract Book

S109

Saturday 13 May

ESTRO 2023

179 treatment fractions in Phase Three over 5 months; 5 cases were marked acceptable but with suggestions for next fraction (4 involved contours, 1 the adapted plan), all other cases were deemed acceptable. Conclusion A training strategy was developed to enable a RT-driven workflow for prostate adaptive radiotherapy for MR-Linac treatment. This model may be adapted for other anatomical sites to maximize efficiencies for MR-Linac radiotherapy. MO-0147 Plan library approach to optimize online adapted MRgRT workflow in cervical cancer patients C. Votta 1 , L. Boldrini 1 , A. Romano 1 , G. Chiloiro 1 , G. Panza 1 , G. Turco 1 , L. Placidi 1 , M. Nardini 1 , D. Cusumano 1 , R. Autorino 1 , P. Cornacchione 1 , L. Indovina 1 , V. Valentini 1 , M.A. Gambacorta 1 1 Fondazione Policlinico Universitario A. Gemelli IRCCS, Diagnostica per Immagini, Radioterapia Oncologica ed Ematologia, Roma, Italy Purpose or Objective Cervical-uterine structure (CUS) motion represents the main uncertainty during radiation therapy (RT) delivery for locally advanced cervical cancer (LACC). The set up of adequate planning target volume (PTV) margins and using effective replanning strategies is therefore strongly needed. Different studies suggest the plan-library (PL) approach to manage daily anatomy changes in LACC. The introduction of Magnetic Resonance-guided Radiotherapy (MRgRT) could represent an improvement to the current standard, especially thanks to the possibility to perform online adaptive radiotherapy (oART). However, the oART can significantly prolong the treatment time and pose a challenge for daily clinical use. This study aims to propose a new hybrid approach based on prior knowledge represented by the PL including oART plans generated during treatment (oAR- PL). Materials and Methods 14 patients affected by LACC undergoing chemoradiotherapy (CRT) were retrospectively analysed to evaluate the feasibility of this approach. Each patient received 50.6 or 45 Gy MRgRT treatment delivered in 22 or 25 daily fractions using a 0.35 T MR Linac. All daily positioning MR-scans were registered with the planning MR-scan (MRP) by rigidly aligning the bony anatomy. A PTVsim was obtained by adding a 5 mm margin to the CUS of MRP. For each daily MR-scan, the CUS was retrospectively delineated and compared to the PTVsim: if the CUS was outside the PTVsim, a new PTVfx was generated to optimize CUS dose coverage for the fraction in exam, as currently done in oART, creating a dedicated PL on adapted daily target volumes. The CUS of the following fractions was then compared with the PTVsim and with all the PTVfx collected in the PL, identifying the PTV which covered at best the position of the CUS (Figure1).

Figure 1. Example of a possible application of oART- PL. MR-sagittal plane of Fraction 6 scan (a): all PTVfx obtained from the previous oART fractions are visible (b.); the green PTVfx1 entirely covers the position of the CUS fx6 (c.); the user may therefore select the plan generated during fraction 1 oART. Results A total of 332 pre-treatment MR scans were reviewed. With the proposed PL hybrid approach, the median number of fractions requiring oART resulted to be 5 (range: 1 - 9). The remaining fractions can be therefore managed using the PL obtained with the previous oART plans ( median: 54%) or with the Original Plan (median: 18%). Figure2 shows the plan selection over the treatment course.

Figure2. oART is highlighted in blue. The different plan selection percentages are reported: original plan (OP), adaptive online (oART) and plan library (PL). Conclusion