ESTRO 2024 - Abstract Book

S3980

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2024

Inter-fraction variations are substantial in cervical cancer radiotherapy. To compensate for these variations and reduce the risk of target underdosage, a large planning target volume (PTV) margin is needed. To reduce PTV margins, Library of Plans (LoP), also known as Plan-of-the-Day, has been introduced [1-3]. In practice, a number of treatment plans mimicking a range of anatomical configurations are generated prior to the first fraction. During treatment, the best-fitting treatment plan is selected each fraction, based on the current anatomy as observed on the daily cone beam computed tomography (CBCT). The plan selection procedure, however, is subject to inter observer variation, which may influence the received radiation dose to the target and organs-at-risk (OAR). The aim of this study was to determine whether the inter-observer variation in plan selection led to significant dose deviations in PTVs and OARs.

Material/Methods:

For two different patients with limited and more extensive clinical target volume (CTV) movement, all fractions were retrospectively analyzed on plan selection and dose distribution. Patients were treated with a dose prescription of 45 Gy in 25 fractions and clinical goals according to EMBRACE II. Three plan selections were analyzed. Firstly, the plans selected during treatment, referred to as clinical selected plans, were compared. Secondly, the retrospectively repeated plan selection by a team of five experts, referred to as consensus selected plans was correlated. Thirdly, the full bladder plan was solely selected, thus simulating the situation in which no plan selection occurs (i.e. the default plan). To evaluate the influence of plan selection on the treatment dose, the dose of the selected plans in each fraction was calculated on the Hounsfield Units-corrected CBCT. To evaluate the total treatment, the doses of all fractions were accumulated on the planning CT. In order to determine whether the dose difference between the three different plan selections was significant, the uncertainty was calculated using the dose-volume histogram (DVH) overlay technique [4]. The fraction uncertainty was scaled to total treatment dose and the root-mean-squared sum of the scaled uncertainties of all fractions was used as uncertainty for the accumulated dose distribution over the complete treatment course.

Results:

In the fraction dose example of patient A, the default plan was selected in the clinically selected plan (Figure 1). Although the difference in the CTV DVH between the consensus and clinically selected plan was larger than the uncertainty, the V95% of the CTV was 100% with the three selections. In the shown fraction DVH of patient B, the CTV V95% in the CBCT was highest in the clinically selected plan and lowest in the default plan. In the OAR, the difference in DVH between the three selected plans was larger than the uncertainty. In the bladder DVH, the consensus selected plan resulted in a higher bladder dose compared to the clinically selected plan. The default plan resulted in the lowest bladder dose. The default plan resulted in the highest rectum dose and the consensus selected plan resulted in the lowest rectum dose.