ESTRO 2024 - Abstract Book

S5597

RTT - Patient care, preparation, immobilisation and IGRT verification protocols

ESTRO 2024

1 Azienda USL-IRCCS di Reggio Emilia, Medical Physics Unit, Reggio Emilia, Italy. 2 Azienda USL-IRCCS di Reggio Emilia, Radiotherapy Unit, Reggio Emilia, Italy

Purpose/Objective:

Surface-guided radiotherapy (SGRT) is a relatively new technique for initial patient positioning and in-treatment movement monitoring. SGRT is completely non ‐ invasive since it uses visible light to determine the position of the patient's surface. SGRT continuously evaluates the difference between the real-time three-dimensional surface of the patient and the reference patient's body surface, providing a substantial amount of information. Recently, several publications have shown the validity of SGRT to monitor inter-fraction errors for different sites or quantify intra-fraction errors [https://doi.org/10.1038/s41598-023-44320-2]. However, given the large amount of data provided by this system, there are still many unexplored points and further in-depth analysis should be performed to identify factors that can affect both intra- and inter-setup errors.

In this study, we retrospectively analysed SGRT signal on a large cohort of breast cancer patients with the aim of investigating if inter-fraction errors, treatment time, and the number of fractions can affect intra-fraction accuracy.

Material/Methods:

A total of 366 breast cancer patients (2400 fractions) who underwent non-gated free-breathing tangential fields radiotherapy in AUSL-IRCCS of Reggio Emilia were selected. Two fractionation schemes were used for treatment 40 Gy/15 fr and 39 Gy/13 fr. Patients were positioned and monitored with AlignRT system (London, UK) as SGRT without the use of skin markers. On board cone beam CTs (CBCTs) image set was acquired before treatment to verify inter-fraction setup following institute-defined frequencies. In the standard protocol, when the corrections detected by CBCT were within the tolerance levels (TLs), inter-fraction errors ≤ TLs, CBCT was acquired in the first three fractions and a fourth image was set in the middle of the treatment (N=4). If, instead, CBCT shifts resulted outside the TLs in the first three treatment fractions, inter-fraction errors > TLs, additional CBCTs were acquired during treatment. Patients were divided into two groups based on the number of CBCTs: well-setup (N=4, small inter-fraction errors) and bad-setup (N>4, large inter-fraction errors) groups respectively. In addition, the whole dataset (all) was also considered. The patient intra-fraction error (mean error) was estimated as the average SGRT drift during treatment time (from first treatment field beam-on to last treatment field beam-off). Maximum error was defined as the maximum difference within 10-90 percentiles of patient drift of each session. For each group (all, well-setup and bad-setup), the intra-fraction error magnitude (mean and maximum) was quantified and correlated with treatment session time (within 2, between 2-3, between 3-4, greater than 4 minutes) and with the treatment fraction number.

Results:

Mean (maximum) intra-fraction error over the whole patient cohort (all) resulted in (mean± 1SD) 2.1 ± 1.7 mm (2.1 ± 3.8 mm) of magnitude. 90° percentile resulted within 1.5 mm in all directions.