Session Item

The main objective of this study was to compare two patient immobilisation devices in lung Stereotactic Body Radiation Therapy (SBRT): one based on a patient-customized body cushion and dedicated to SBRT and the other based on a simple arm support.

Twenty patients treated for early-stage lung cancer with SBRT were included in the study. Ten patients were installed in supine position with arms above their head in a Vacuum Cushion (VC) dedicated to SBRT (Bluebag, Elekta). Ten other patients were positioned with a simple support device consisting in an arm support (AS) (Posirest®, CIVCO). For each fraction, all patients had a pre-treatment 4-dimensional cone-beam computed tomography (4D-CBCT) and a post-treatment 3D-CBCT to compare positioning accuracy and immobilisation quality. For both CBCT, a first rigid registration with the planning CT was performed on the spine at the level of the target. Translations and rotations were reported. Treatment time, age, Performance Status (PS) and Body Mass Index (BMI) were also reported. 

The median number of fractions per treatment was 5 (range 3-10). Position deviations (translations and rotations) from 112 4D-CBCT pre-treatment and 111 3D CBCT post-treatment were estimated. Mean inter fraction errors were similar between both installations: in vertical and lateral direction, deviations were less than 1.2mm and inferior or equal to 1.1° in rotation. For longitudinal translations, mean inter-fraction errors were 1mm with VC and -4.0mm with AS

For the patients installed with VC, mean vertical, longitudinal and lateral intra-fraction errors, were 0.0mm, -0.2mm and -0.3mm respectively (SD 1.0, 1.5 and 1.8mm respectively). For the arm with AS, mean vertical, longitudinal and lateral intra-fraction errors were -0.2mm, -0.8mm and 0.1mm respectively (SD 2.3mm, 2.7mm and 2mm respectively). The time between the pre-treatment 4D-CBCT and the post-treatment 3D-CBCT was 14.3min and 15min for VC and AS installation respectively. No correlation between age, PS or BMI and immobilisation accuracy was found.

Our results were consistent with the literature. The use of a simple arm support allowed a intra fraction positioning error < 1 mm compatible with lung SBRT considering a GTV to PTV margin of 5mm.