Daily adaptive MR-guided stereotactic ablative re-irradiation (re-SABR) in oligometastatic liver disease: A single-institution retrospective analysis.

Objective: Stereotactic ablative body radiotherapy (SABR) is an established treatment option in oligometastatic liver disease. Limited data exist on liver SABR re-irradiation (re-SABR). MR-guided SABR with daily plan adaptation and deformable image registration (DIR) allows organ-at-risk (OAR) sparing and dose accumulation in a re-irradiation setting. We report the safety and efficacy of MR-guided re-SABR in oligometastatic liver disease, alongside a DIR-based workflow for cumulative OAR dose calculation. Methods: MR-guided re-SABR to oligometastatic liver disease was retrospectively analysed. Key inclusion criteria included: prior liver SABR, MR-guided re-SABR to ≤ 4 liver metastases, Child Pugh Score (CPS) ≤ B7, and a minimum prescribed dose of 30 Gy in 5 fractions. Lesions were classified according to the ESTRO-EORTC re-irradiation consensus. Acute and delayed toxicity, local control (LC), local progression-free survival (LPFS), and overall survival (OS) were reported, overall and by primary tumour type. Cumulative OAR doses were estimated using a DIR-based as compared to a non-DIR-based workflow. Results: Between October 2020 and April 2024, twelve patients underwent MR-guided re-SABR to 18 liver metastases. While the majority (12/18) were colorectal in origin, other primaries included pancreas, breast, oesophagogastric, and ovary. 50% of lesions were type-1 re-irradiation (including 3 repeated target irradiations).Median prescription BED10 was 100 Gy (range 72-151 Gy) for initial SABR and 100 Gy (range 48-132 Gy) for re-SABR. 50 Gy in 5 fractions (BED10 100 Gy) was delivered in 8 out of 12 re-SABR cases. With a median follow-up of 10 months (range 3-33) from re-SABR, no acute ≥ G2 toxicity was seen. 12-month PFS was 92 % (95 % CI 54-99).Median, 1-year, and 2-year OS were 36 months (range 12-37), 100 % (95 % CI 54-99) and 91 % (95 % CI 51-99) for SABR and 22 months, 68 % (95 %CI 28-89) and 34 % (95 % CI 5-67) for re-SABR. DIR-based workflow estimates predicted significantly higher MLD and D700cc (p < 0.01) and smaller uninvolved liver volumes (p < 0.05). Conclusion: With the limitation of relatively low patient numbers and mixed tumour histology, MR-guided re-SABR to oligometastatic liver disease appeared well tolerated, achieving high LC rates. DIR-based workflow predicted higher cumulative OAR doses, potentially further improving the safety of liver re-SABR.