2726 - Stereotactic MR-Guided Adaptive Radiotherapy (SMART) for Liver Metastases

Purpose/Objective(s): Local treatment for liver oligometastases has been shown to improve overall survival. SBRT has emerged as an important treatment modality with recent technological advances. Stereotactic MR-guided Adaptive Radiotherapy (SMART) offers advantages including superior soft-tissue visualization, online on-table adaptive planning, and real-time target tracking during treatment delivery. We aim to report our institutional outcomes of SMART for liver metastases.

Materials/Methods: This IRB-approved retrospective study included 140 patients (252 lesions) with SMART for liver metastases between 2018-2025. All treatment were delivered using a 0.35 T MR-linac using respiratory-gated delivery. Clinical and dosimetric outputs were analyzed. Overall survival (OS) was calculated on per-patient basis, distant progression-free survival (DPFS) and liver progression-free survival (LPFS) were calculated on a course basis and local control (LC) was calculated on a per-lesion basis. Toxicities were graded per CTCAE v5.0.

Results: The median patient age was 61.5 years (range, 20-85) and 50.7% were female. ECOG performance status was 0 in 121 patients (86.4%). Primary tumor types were colorectal cancer in 33.6%, breast cancer in 19.3%, and lung cancer in 16.4% of patients. Prior to the first SMART treatment, 68.6% of patients had liver metastases, and 40 of these patients had received previous local treatments for liver metastases. The median time from diagnosis to first liver metastases was 13.0 months (range, 0-177.3 months). The median number of treated lesions per patient was 1 (range, 1-10 lesions). Median lesion diameter was 17 mm (range, 4-96 mm), median GTV volume was 5.4 cc (range, 0.2-520.1 cc) and the median PTV margin was 3 mm (range, 2-8 mm). Median BED10 value was 100 Gy (range, 48-151.2 Gy). Total fraction number was 1026 and in 93.3% adaptive online plans were used. Median follow-up time from SMART was 21.9 months (range, 1.8-83.5 months). One- and two-year local control rates were 94.7% and 92.9%, respectively. One- and two-year LPFS rates were 41.6% and 36.6%, DPFS rates were 46.2% and 25.3%, and overall survival (OS) rates were 80% and 68.4%, respectively. Median overall survival was 53.4 months. On MVA, presence of >2 hepatic metastases during SMART were associated with worse LPFS (p=0.038). Having received IO in the last three months (p=0.028) and existence of extrahepatic metastases during SMART (p=0.006) was associated with inferior DPFS. Non-lung primary tumors (p=0.001) and total GTV volume <12.3 cc were found to be associated with improved OS (p=0.002). No acute or late grade =3 side effects were observed.

Conclusion: Herein we present the largest series of liver metastases treated with SMART to date in the literature. SMART offers favorable local control with promising survival rates with an acceptable toxicity profile in the treatment of liver metastases. These findings support SMART as a robust local treatment modality for carefully selected patients with liver metastases.