ESTRO 2024 - Abstract Book

S2159

Clinical - Upper GI

ESTRO 2024

PS; 0 HR 2.74, p<0.001), a smaller PTV volume (PTV>130cc vs. PTV<45cc; HR 1.92, p<0.001) and a higher BED10 (BED10≥100 Gy vs. BED10<100 Gy; HR 0.74, p=0.025)

LC data could be retrieved for 345 lesions. LC rates at one and two years were 70% and 52% respectively. In univariable analysis, a better LC was found for HCC compared to colorectal metastases (HR 0.40, p=0.002). A better LC was observed for lesions in patients who did not receive prior systemic therapy (p<0.001) and where personalised immobilisation (p=0.012) was used. The use of Cyberknife was associated with a better LC as compared to 3D-CRT (HR 0.46, p=0.007). A higher BED10 and a smaller PTV volume were also associated with a superior LC. After multivariable analysis, however, better LC was only found for smaller PTV volumes (PTV>130cc vs. PTV<45cc; HR 2.13, p=0.002), higher BED10 (BED10≥100 Gy vs. BED10<100 Gy; HR 0.61, p=0.011) and for lesions in patients who were not priorly treated with systemic therapy (HR 1.93, p=0.007). Histology, radiotherapy techniques and the use of personalised immobilisation were no longer influencing LC.

Fig. 1 Cumulative number of centres performing liver SBRT from 2013-2019 (left panel) and prescribed biologically effective dose (BED10) from 2013-2019 (right panel).

Table 1. Patient, tumour and treatment characteristics