ESTRO 2024 - Abstract Book

S4102

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2024

1 Vejle Hospital, University Hospital of Southern Denmark, Department of Oncology, Vejle, Denmark. 2 Vejle Hospital, University Hospital of Southern Denmark, Radiotherapy Research Team, Department of Oncology, Vejle, Denmark

Purpose/Objective:

A study is presented of the interfraction variation in the delivered dose to target in prostate cancer radiotherapy assessed through CBCT-based dose calculation for all 39 fractions for each of 22 patients. The results are analysed to evaluate the appropriateness of planning margins and treatment plan robustness to interfraction anatomical changes.

Material/Methods:

22 intermediate and high-risk prostate patients receiving external radiotherapy to the prostate, proximal seminal vesicles and pelvic lymph nodes at Vejle Hospital, Denmark, were randomly selected for this study. CBCT scans were obtained daily as part of patient setup (prostate seed match) on Elekta linacs using the XVI imaging system (Elekta, United Kingdom). RayStation 11B (RaySearch Laboratories AB, Sweden) was used for planning and retrospective dose evaluation of the treated 6 MV VMAT plans.

Table 1: Target volumes and planning margins

Target

CTV-to-PTV margins

Prostate (CTVp; PTVp)

L-R: 7 mm; S-I: 10 mm; P-A: 7 mm

Proximal seminal vesicles (CTVp_SV; PTV_SV)

L-R: 7 mm; S-I: 10 mm; P-A: 7 mm

Elective pelvic lymph nodes (CTVn_E; PTVn_E)

L-R: 5 mm; S-I: 8 mm; P-A: 7 mm

From the 39 CBCT scans for each patient, a set of corrected CBCT scans (corrCBCTs) were created in RayStation to allow for accurate dose calculation[1]. Target volumes were propagated from the planning CT scan to the corrCBCTs, and relevant dose characteristics, including the average dose to target and the target volume coverage ( V 95% ), were calculated. The 858 corrCBCTs were analysed to quantify set-up errors, prostate motion, and abdominal girth variation.

Results:

A simple propagation based on a seed-based rigid image registration was in most cases found to give reasonable contouring of the CTVp on the corrCBCTs. Whilst this strategy ignores deformation, it is useful for evaluating dosimetric effects of prostate motion, abdominal girth variation, and setup errors affecting the distance from the skin to the prostate. For all 22 patients, the median of the fraction average dose to the CTVp is within 1.5% of the planned dose (Figure 1). The abdominal girth is for a small sub-group of the patients seen to increase or decrease systematically during the course of the treatment, causing the fraction average dose to the prostate to decrease or increase, respectively, here by up to 3% (data not shown).