ESTRO 2020 Abstract Book

S403 ESTRO 2020

OC-0714 Feasibility of online adaptive MR-guided radiotherapy for rectal cancer on a 1.5T MR-linac. M. Intven 1 , S. De Mol van Otterloo 1 , S. Mook 1 , P. Doornaert 1 , R. Tijssen 1 1 UMC Utrecht, Radiation Oncology Department, Utrecht, The Netherlands Purpose or Objective The aim of this study was to assess the feasibility of an online adaptive workflow on a 1.5T MR-linac for patients with rectal cancer. Shape and volume of the clinical target volume (CTV) for radiotherapy in rectal cancer are influenced by bladder and rectal filling. Online adaptation of these daily variations may enable a reduction of the currently large treatment planning target volume (PTV) margins. This abstract reports on the clinical feasibility of daily online replanning in terms of technical feasibility, target coverage, treatment time, and patient compliance. Material and Methods Patients were treated on a 1.5T MR-linac (Unity, Elekta AB, Stockholm Sweden). At each fraction a 3D T2 weighted (T2w) MR-sequence was acquired on which the CTV was delineated after a deformable registration of the contours from the pre-treatment imaging. PTV margins for these first rectal cancer patients on the MR-linac were, conservatively, based on the standard PTV margins used in our clinic for conventional linac treatments, which were 1 cm isotropically around the mesorectum and 8 mm around the elective nodal regions. Based on the new contours a full online replanning was done after which a new 3D T2w MR-sequence was acquired for position verification purposes. Treatment delivery was done using a 5-field Intensity Modulated Radiotherapy (IMRT) technique during which another two 3D T2w scans were made. Another 3D T2w sequence was made after treatment delivery (figure 1). The duration of each step in the workflow was monitored by one of the RTTs.

in our department for combined elective EBRT + BT boost treatment planning) were aimed for (Table1). All dose levels include the elective EBRT dose (45Gy/25 fractions). For the (stereotactic) MRL boosts, isotoxic treatment planning was performed, i.e. the target dose to the PTV was restricted by hard OAR constraints. The eventual dosimetric effect of intrafraction motion was evaluated on the anatomy as seen on post treatment MRI for each fraction. Treatment response is measured on MRI and by clinical investigation at 3 months after treatment. Morbidity is scored according to the Common Terminology Criteria for Adverse Events Version 4.0.

Results Table1 shows the total doses (EQD2) achieved for targets and OAR for the 3 patients treated so far. All targets received doses ≥ 70Gy EQD2 and all OAR constraints were met. The target structures were under dosed in all three patients, which could be attributed to the unfavorable location between OARs and target (e.g. patient 3) (Figure 1). After recalculation on the post-treatments scans a small dose deviation is seen on CTV (-1.1, +1.6, -2.4Gy EQD2, resp.) compared to the doses anticipated at daily treatment planning. Patient 1 showed a complete remission after 3 months, Follow-up imaging of the other 2 patients has not been performed yet. No acute grade ≥ 3 toxicity was reported.

Results Twenty-five patients with rectal cancer were treated with 25 Gy in 5 fractions. 18 (72%) of patients was male and medium age was 63.0 (range 43.5-81.3). Most patients had intermediate-risk stage disease located in middle or distal rectum. 118 (94%) of the total 125 planned fractions were delivered on the MR-linac. Two fractions (1.6%) were delivered on a conventional accelerator due to logical reasons and three fractions (2.4%) due to technical problems with the MR-linac. Two fractions (1.6%) were not delivered at all due to an acute infectious disease of the patient. No fractions were aborted, missed or canceled due to intolerance of the patient. Median time per fraction, defined as the duration of time between the patient leaving and re-entering the dressing room, was 48 minutes. Minimal and maximal time per fraction was 32

Conclusion It is feasible to give EBRT boosts with the MRL in case BT boosts are not feasible in cervix cancer patients. With the MRL we achieve higher doses to the target than with CBCT- linacs, but compared to BT we miss the specific high dose volumes and the achieved target doses are lower. We strive to further optimize our treatment planning and fractionation scheme to improve the balance between target and OAR dose.