SBRT for liver metastases (Seminars in Radiation Oncology review)
TL;DRBED ≥100 Gy10 tied to 3-year LC 93% vs 65% in liver mets; review maps fractionation to OAR distance and liver dose constraints.
MAESTRO
The dose argument rests on one pooled analysis: BED above 100 Gy10 tracked with 3-year LC 93% vs 65% in mets, while primary liver tumors showed no gain above 100 Gy10. That separates the dose target for liver mets from HCC practice. For lesions near bowel or the central biliary tree, the review protracts fractionation rather than lowering BED.
In a non-cirrhotic pt with a liver metastasis abutting bowel or the central biliary tree, this supports protracted fractionation over a 3-fraction ablative course; it does not extend to cirrhotic livers, where the cited limits are tighter.
BED above 100 Gy10 tracked with 3-year LC 93% vs 65% in mets but not in primary liver tumors, so the dose target differs from HCC practice. Near bowel or the central biliary tree, the review protracts to 10-15 fractions instead of lowering BED.
11 details 3 trials watching
Narrative topic review of liver SBRT for metastases. It covers prospective series, pooled analyses, prognostic scores, QOL, dose and fractionation, OAR tolerance and motion management. No search strategy or inclusion criteria are stated in the source.
Most series use 3-5 fractions (range 1-6). When maximizing LC is the priority the review recommends BED ≥100 Gy10 to the PTV. Fractionation is chosen to respect OARs in priority order: luminal GI, central biliary tree, chest wall, then liver. An SIB with PRV is suggested for tumors 1-2 cm from OARs.
QUANTEC: MLD limits of 15 Gy in 3 fractions and 20 Gy in 6 fractions keep the risk of G3 liver enzyme toxicity under 20%. In 6-fraction SBRT, G3 liver enzyme risk was 0% at MLD ≤15Gy but 6-18% at MLD ≤21Gy, depending on D800cc, so MLD alone is insufficient.
Against RFA, a 2238-pt meta-analysis showed 2-year LC 84% vs 60% favoring SBRT in liver mets studies, with no OS difference and G3+ toxicity <5% for both. MAESTRO, a randomized phase II trial of MR-guided adaptive vs ITV-based SBRT, is cited as the randomized technique question.
The BED >100 Gy10 threshold comes from a pooled analysis of published series, open to selection by lesion size and location. Several cited constraints are borrowed from the ASTRO guideline for liver primaries, not derived in metastases.
The review positions SBRT for lesions percutaneous ablation handles poorly: periampullary, perihilar, subcapsular, perivascular or large tumors. Optimal dose and fractionation remain unknown, with no consensus guideline. Whether CRC histology is radioresistant at sufficiently high dose is unresolved.
| Distance to OAR | Dose range | Fractions |
|---|---|---|
| >2 cm from all OARs | 45-75 Gy | 3 |
| 1-2 cm from any OAR | 50-60 Gy | 5-6 |
| <1 cm from luminal GI or central biliary tree | 60-67.5 Gy | 10-15 |
| Metric | 3 fractions | 5 fractions |
|---|---|---|
| Mean liver dose | <12-15 Gy | <15-18 Gy |
| Critical volume ≥700cc | <19 Gy | <21 Gy |
- Optimal dose and fractionation for liver metastases n=110 · primary completion 2026-04 · standard vs radiobiologically-guided SBRT doserecruiting Ultra-high Dose Radiation for Liver Metastasis Using MR-guided TReatment With Stereotactic Ablative Single-fraction Phase 3n=114 · primary completion 2031-09 · phase 3 RCT, ultra-high vs high single-fraction dose
- Randomized comparison of SBRT vs RFA for liver metastases
- Benefit of MR-guided adaptive SBRT over ITV-based planning recruiting Magnetic Resonance-guided Adaptive Stereotactic Body Radiotherapy for Hepatic Metastases Phase 2n=90 · primary completion 2029-01 · MRgSBRT vs ITV-based SBRT, 1-3 hepatic mets
📚 Sources · 📄 1 paper
The longer read
The practical weight of this review rests on its dose-selection framework, and it helps to see how thin the evidence under that framework is. The claim that liver metastases need more dose than primary liver tumors comes from a pooled analysis of 13 published series. Metastases had lower 3-year LC than primaries (76% vs 86%), and lesions treated above a BED of 100 Gy10 did better than those below it (93% vs 65%). The finding is biologically plausible and consistent with the single-fraction phase 1 experience. But it is a cross-series comparison, and the confounding runs in a predictable direction. Lesions that can safely reach a BED above 100 Gy10 are the small ones far from bowel and the central biliary tree, and the review itself lists smaller tumor volume as an independent predictor of LC. Part of the dose effect may therefore be a size and location effect. A reader should treat 100 Gy10 as a reasonable planning target for a lesion that allows it, not as a proven threshold below which control collapses.
The fractionation tiers keyed to OAR distance follow from that reasoning. They are an argument for keeping BED high by protracting delivery, rather than de-escalating dose when anatomy is unfavorable. That is sensible, but the review offers no outcome data for the 10-15 fraction tier in liver metastases. Its support comes from a review of external beam RT and from dose-escalated pancreas practice. For a lesion within 1 cm of duodenum, the reader is borrowing technique from pancreas SBRT, and should say so when counseling.
The SBRT vs RFA comparison deserves the most caution. A meta-analysis favoring SBRT on 2-year LC with no OS difference is what one would expect when nonrandomized cohorts are compared. Tumor size, location relative to vessels and referral pattern all differ between the two modalities, and the heat-sink effect the review describes biases RFA failures toward exactly the perivascular lesions SBRT is often chosen for. The LC gap may be real, but its size cannot be read from pooled observational data. The absence of an OS difference is the more defensible finding, because local therapy for one lesion rarely drives survival when extrahepatic progression dominates.
On toxicity, the review is candid that liver SBRT constraints are conservative values that proved safe on trials, not tolerance doses. Several of the most-cited limits come from the ASTRO guideline for liver primaries. Applying them to a non-cirrhotic oligometastatic liver is likely conservative rather than dangerous. The point the review makes well is that mean liver dose alone does not predict enzyme toxicity once the spared critical volume is small.
What the review does not settle is the question that matters most for an RT reader: which patients with liver oligometastases gain survival from ablation at all. The prognostic scores it cites are retrospective and unvalidated. Until randomized data separate patient selection from treatment effect, the dose and constraint guidance here is a sound way to deliver SBRT safely, not evidence of when to deliver it.