Tumor Size and Dose Impact Recurrence after Stereotactic Body Radiotherapy (SBRT) for Liver Metastases (LM) from Colorectal Cancer (CRC)

Abstract

<h3>Purpose/Objective(s)</h3> SBRT is used to treat LM from CRC. Prior studies demonstrate acceptable in-field control (IFC) but have limited evaluation of treatment characteristics. We sought to determine the effect of tumor and treatment-related variables. <h3>Materials/Methods</h3> We reviewed an IRB-approved prospectively collected database of SBRT treatments between 2013-2021 at an NCI-designated comprehensive cancer center. Post-SBRT MRI every 3-4 months was fused with RT plan and evaluated for in-field (IFR), marginal (MR), and regional (RR) liver recurrences. MR = low dose region adjacent to high dose region. RR = elsewhere in liver. Retreated lesions were excluded. Tumor and dosimetric variables were compared by lesion recurrence status using Wilcoxon rank sum tests and Fisher exact tests. IFC (time to IFR) and IFC/MC (time to IFR or MR) were evaluated using Kaplan Meier methods with log rank tests and/or Cox proportional hazards methods using within-patient robust standard errors. <h3>Results</h3> The study included 33 patients with 41 lesions. Most (72%) were metastatic at presentation and had a single lesion treated (n=25), while the remaining had 2 treated simultaneously (n=5) or sequentially (n=3). Median age was 59 years (34-88). Median tumor volume was 35.5 cc, with 25 (61%) from 10 to 50 cc. Median total dose (TD) was 55 Gy (24-60 Gy) in 5 fractions (3-5); 19 lesions (46%) received 60 Gy. Median follow-up was 18.4 months in those who died (n=10) and 20.2 in those alive (n=13). IFR and MR were diagnosed in 7 (17%) and 4 (9%) lesions, respectively. Nineteen (46%) had RR. 1-year IFC was 86% (CI 70-94). IFR was associated with lower median TD (45 vs 60 Gy; p=0.008), biologically effective dose (BED) (85 vs 132; p=0.008), prescription isodose line (IDL) (68.0 vs 77.7%; p=0.007), and larger tumor size (155 vs 30.8 cc; p=0.002). Differences were maintained when comparing lesions with either IFR or MR vs. neither. IFC was better for smaller tumors, with 1-year rate of 96% (CI 73-99) for tumors <50 cc compared to 70% (CI 37-87) for ≥50 cc (p=0.005). TD of 50-60 Gy had better 1-year IFC of 96% vs 64% for 24-47.5 Gy (p<0.001). Higher IDL also had better 1-year IFC (86% for 80-85.7 vs 83% for 70-79 vs. 33% for 63.3-69.5; p=0.048). Occurrence of either IFR or MR was higher at TD 24-55 Gy vs. 60 Gy (41% vs 10%; p=0.006), with IFC/MC at 1-year of 65% vs 94% (p=0.006). Using Cox models, IFR/MR was also higher at larger tumor size (HR 11.08; p=0.001) and IDL (for 1% difference, HR 0.91; p=0.008). Estimated IFC/MC is 80% (CI 62-90) at 1 year vs 69% (CI 49-82) at 2 years. <h3>Conclusion</h3> IFC after SBRT for LM in CRC was excellent with 50 to 60 Gy. Higher BED and TD reduced IFR and MR after SBRT for liver metastases. Larger tumor size yielded a higher failure rate and may warrant additional dose. Although durable IFC is frequently achieved, there are opportunities for intensifying therapy.