ESTRO Clinical Practice Guideline: SBRT for Spine Metastases
ForUnirradiated spine mets, solid tumors, no symptomatic MSCC
TL;DR22 recommendations, 5 statements: spine SBRT dose above 1x18Gy (BED10=50Gy10) for LC; 2x12Gy conditionally preferred for painful mets; PTV margin ≤3mm.
The dose floor is the actionable line: above 1x18Gy (BED10=50Gy10), with 2x12Gy preferred for painful mets, because RTOG 0631's 16 or 18Gy single fraction lost to 8Gy on 3mo pain response (41.3% vs 60.5%). Single fractions >20Gy carry high-level VCF evidence, so 24Gy x1 for oligomets trades LC against fracture.
In an unirradiated painful spine met that is not frankly unstable, Bilsky 0-1, up to 3 contiguous segments, with prolonged life expectancy, this supports considering SBRT over conventional RT; it does not extend to symptomatic MSCC, re-irradiation, or short life expectancy, where conventional RT stays preferred.
The dose floor is the actionable line: above 1x18Gy (BED10=50Gy10), with 2x12Gy preferred for painful mets, because RTOG 0631's 16 or 18Gy single fraction trailed 8Gy on 3mo pain response (41.3% vs 60.5%). Single fractions >20Gy carry high-level VCF evidence, so 24Gy x1 for oligomets trades LC against fracture.
10 details 3 trials watching
ESTRO clinical practice guideline built on a PRISMA systematic review of PubMed, Embase and Cochrane (Jan 2005-Sep 2021), with RTOG 0631 added Apr 2023. ASTRO methodology graded evidence and recommendation strength; ≥75% agreement defined consensus.
Vertebral metastases from solid malignancies, painful or oligometastatic, previously unirradiated. Re-irradiation and symptomatic MSCC excluded from scope.
Dose above 1x18Gy (BED10=50Gy10) for durable LC; 2x12Gy conditionally preferred palliative schedule. PTV margin ≤3mm (minimum 1mm), intensity-modulated delivery, daily IGRT, cord PRV prioritized over coverage.
VCF is the most frequent AE (6-14% in two multicenter analyses); single fractions >20Gy raise VCF risk (high-level evidence). Myelopathy 0.4% in a pooled de novo series of >1,000 pts; routine prophylactic steroids for pain flare not recommended.
The pain recommendation leans on SC.24 (24Gy/2fx vs 20Gy/5fx: CR 35% vs 14%, p=0.0002). RTOG 0631 (16 or 18Gy x1 vs 8Gy x1) showed no pain gain, which the panel reads as a dose-floor problem rather than a failure of technique.
Search closed Sep 2021 apart from RTOG 0631; ROBOMET is abstract-only and the TwiCs trial was underpowered with non-standardized dosing. Panel is radonc-weighted with one spine surgeon and no medical oncologist, yet addresses concurrent targeted/IO toxicity.
Dose, not technique, organizes the document: low single-fraction SBRT matched or trailed 8Gy, while higher doses improved CR and LC at a fracture cost. 2-5 fraction schedules are the hedge where fracture management capacity is limited.
| Trial | SBRT | CRT | Pain result |
|---|---|---|---|
| Sprave, rand phase II (n=55) | 24Gy/1fx | 30Gy/10fx | Response 6mo 73.7% vs 35%, p=0.015 |
| CCTG SC.24 phase II/III | 24Gy/2fx (n=114) | 20Gy/5fx (n=115) | CR 3mo 35% vs 14%, p=0.0002 |
| RTOG 0631 phase 3 (n=339, 2:1) | 16 or 18Gy/1fx | 8Gy/1fx | Response 3mo 41.3% vs 60.5%, favoring CRT |
| ROBOMET phase III, abstract (n=126) | 20Gy/1fx | 8Gy/1fx | CR 1mo 37% vs 25%, ns (ITT) |
| Statement | Level of evidence |
|---|---|
| Fractionated SBRT: no increased VCF risk vs CRT | High |
| Single-fraction >20Gy: increased VCF vs CRT and vs fractionated SBRT | High |
ESTRO practice guideline from a PRISMA systematic review with ≥75% panel consensus; most recommendations rest on moderate or expert-opinion evidence, not new outcome data.
- Does 2x12Gy transfer beyond high-volume SC.24-type centers?
- Spine SBRT safety with concurrent targeted therapy or immunotherapy recruiting Vertebroplasty With Radiation Therapy for Spine Metastatic Cancer Patients With Indeterminate Lesion (SINS Criteria) Phase NAn=200 · primary completion 2026-12 · spine SRS ± immunotherapy, side effects tracked
- Role of separation surgery before SBRT in high-grade epidural disease n=26 · primary completion 2026-05 · phase 2 SBRT alone to avoid surgery, high-risk spine metsrecruiting Efficacy and Safety of Stereotactic Body Radiation Therapy for Spinal Metastases With Bilsky Grade 1c-3 Phase NAn=50 · primary completion 2028-09 · single-arm SBRT alone in Bilsky 1c-3 epidural disease
📚 Sources · 📄 1 paper
The longer read
The guideline's center of gravity is a dose argument, and the randomized data under it are less uniform than the recommendations read. SC.24 carries the pain recommendation: 24Gy in 2 fractions beat 20Gy in 5 on 3-month complete response (35% vs 14%), and the Sunnybrook subcohort extended the advantage into local failure and re-irradiation. RTOG 0631, the largest trial, went the other way, with 16 or 18Gy single-fraction SBRT showing no pain improvement over 8Gy and a higher response rate with conventional RT. The panel reconciles the two by dose rather than by technique, which is why the floor sits above 1x18Gy and why 16 or 18Gy single fraction is explicitly not recommended as a replacement for 8Gy when pain relief is the goal. That reading is coherent, but it is an inference across trials with different populations, pain definitions and fractionation, not a head-to-head dose comparison.
The oligometastatic case is thinner. The 80-95% LC at 1-2 years from the systematic review is drawn largely from non-randomized series, and the one randomized dose comparison cited (24Gy single fraction vs 27Gy in 3 fractions, 117 pts, 56% spinal) favored the higher single dose for local control. The guideline sets that directly against its fracture statements: single fractions above 20Gy carry a high-level association with VCF versus both CRT and fractionated SBRT, while fractionated SBRT does not raise fracture risk over CRT. The consequence is a choice the document deliberately leaves to institutions. 24Gy in 1 fraction buys local control at a fracture cost, and 2 to 5 fraction schedules are the hedge for centers less able to manage fractures.
Selection criteria do much of the work and deserve more weight than the dose tables. Only 2% of SC.24 SBRT pts had high-grade epidural disease, and RTOG 0631 mandated a 3mm cord gap. The recommended population (not frankly unstable, Bilsky 0-1, up to 3 contiguous segments, prolonged life expectancy) mirrors who was actually studied, and the guideline states plainly that conventional RT remains preferred for short life expectancy, inability to tolerate MRI or immobilization, or pain too severe to wait for MR-based planning. Separation surgery followed by SBRT is described as an emerging trend, not endorsed broadly, which is the appropriate weight given no randomized data were cited for it.
The technical recommendations (MRI-based delineation, ≤1.5mm simulation slices, PTV margin ≤3mm with a 1mm minimum, cord PRV prioritized over coverage, daily image guidance) are framed as the conditions under which the existing evidence base should reproduce, including technology the trials predate. Their value is in drawing a boundary: a center that cannot hold an end-to-end tolerance under 2mm sits outside that evidence base whatever dose it prescribes.
What the document cannot settle is whether 2x12Gy, the conditionally preferred palliative schedule, transfers beyond the high-volume programs that generated SC.24, and how SBRT interacts with concurrent targeted therapy or immunotherapy, where the recommendation is only to discuss the risk with the treating medical oncologist. Re-irradiation and symptomatic MSCC, two of the harder spine decisions in practice, are outside its scope entirely.