Abdominal airbag matched plate compression but increased lateral setup error

KEY POINTS

• This retrospective study included 54 patients with primary or metastatic liver or pancreatic tumours: 16 were treated during free breathing, 21 with a rigid abdominal compression plate, and 17 with an inflatable airbag beneath a thermoplastic torso mask.
• The airbag was inflated to a patient-specific end-inspiratory pressure of 10–20 mmHg. Respiratory motion was measured from liver-centroid displacement between the 0% and 50% phases of ten-phase four-dimensional CT rather than from individual tumour motion.
• Mean three-dimensional respiratory motion was 0.88 ± 0.20 cm during free breathing, 0.61 ± 0.24 cm with plate compression, and 0.57 ± 0.19 cm with the airbag (p<0.001). No significant difference was identified between the two compression devices.
• Compared with free breathing, the airbag reduced mean superior–inferior motion from 0.83 to 0.51 cm, a reduction of approximately 39%, and anterior–posterior motion from 0.21 to 0.14 cm, a reduction of approximately 33%.
• Setup reproducibility was evaluated across 331 treatment fractions using daily pretreatment CBCT. The airbag produced the lowest anterior–posterior absolute error at 0.15 ± 0.11 cm, compared with 0.26 ± 0.20 cm during free breathing and 0.20 ± 0.15 cm with the plate (overall p=0.010).
• Left–right setup error was significantly larger with the airbag at 0.32 ± 0.22 cm, compared with 0.18 ± 0.15 cm during free breathing and 0.19 ± 0.15 cm with the plate (p<0.001). Overall vector setup error did not differ significantly among groups: 0.46, 0.44, and 0.51 cm, respectively.
• CBCT images appeared qualitatively sharper with compression, particularly with the airbag, but image quality was not formally measured. Treatment assignment was non-randomized, different immobilization systems were used between groups, and liver-centroid motion may not represent tumour-specific displacement.

CLINICAL TAKEAWAY

SOURCE

Strahlentherapie und Onkologie