Nadia Laack of the Mayo Clinic will present Stereotactic Body Radiation Therapy (SBRT) and Comprehensive Metastasis-Directed Therapy for Ewing Sarcoma Patients Enrolled on AEWS1221: A Report from the Children's Oncology Group on September 27 at 2:10 pm ET in the Grand Ballroom, during the ASTRO 2026 Clinical Trials session.
The parent trial was negative. Its radiotherapy component may prove the more durable contribution, and not for the reason usually given.
The trial and its radiotherapy design
AEWS1221 randomly assigned 298 patients with newly diagnosed metastatic Ewing sarcoma between December 2014 and March 2019, to interval-compressed vincristine, doxorubicin and cyclophosphamide alternating with ifosfamide and etoposide, with or without the IGF-1R antibody ganitumab. All patients were to receive 14 cycles. Local control of the primary tumor was planned after six cycles. Radiotherapy to metastatic sites followed completion of all 14 cycles. Ganitumab did not improve event-free survival, and three-year event-free survival on the trial was 38.3 percent.
Embedded in the protocol was a distinct strategy. Patients were to receive definitive local treatment of all metastatic sites present at diagnosis, provided they had not progressed. Pulmonary metastases were addressed with whole lung irradiation. Bone metastases up to 5 cm were candidates for SBRT at a recommended 40 Gy in five fractions, with conventionally fractionated radiotherapy for larger lesions or where organs at risk made ablative treatment unsafe. This was the first COG trial to evaluate SBRT.
The analysis being presented was written into the protocol from the outset as exploratory objective four: to describe the feasibility of, and local failure rates following, hypofractionated SBRT directed at bone metastases. Prespecification distinguishes it from the retrospective metastasis-directed therapy literature against which it will be compared.
Why SBRT was introduced, and what that implies
The rationale was operational rather than radiobiological. Treating every metastatic site with conventional fractionation in a patient who has already completed months of intensive chemotherapy and definitive treatment of a primary tumor is a substantial undertaking: weeks of additional treatment per site, larger integral dose, and delay to recovery or subsequent therapy. Five fractions per lesion was intended to make comprehensive consolidation practically achievable.
This matters for how the results should be read. The case for SBRT here was never that it controls Ewing metastases better than conventional radiotherapy. The available external data do not support that claim and, if anything, point the other way. Casey and colleagues reported a cumulative local failure incidence of 6.6 percent at one year and 9 percent at three years using SBRT for 49 bone metastases in 22 Ewing patients. Talleur and colleagues reported a ten-year cumulative local failure rate of 4.4 percent using conventionally fractionated radiotherapy in 45 Ewing patients. Ewing sarcoma is radioresponsive, and conventional fractionation already achieves excellent local control at treated sites.
The proposition being tested is therefore about deliverability: whether shortening treatment per lesion converts comprehensive metastatic consolidation from a theoretical recommendation into something that actually happens.
The preliminary signal, and why the denominator is the story
An earlier analysis of this cohort was presented at CTOS in 2023. Of the 298 eligible patients, 169 completed consolidation chemotherapy without relapse and had discrete metastatic disease beyond bone marrow involvement. Eighty-five had bone metastases. Among 74 patients whose osseous lesions were considered eligible for SBRT, 28 received it, or 38 percent. The reported three-year cumulative incidence of failure was 8.4 percent at bone sites treated with SBRT and 20.1 percent at bone sites not treated with SBRT.
These figures come from a conference poster and should be treated as preliminary until the mature analysis is presented.
The uptake number is the more important of the two, and it points somewhere uncomfortable. SBRT was introduced into this protocol specifically to solve a feasibility problem. In a cooperative-group trial that recommended it, at centers that had agreed to the protocol, fewer than four in ten eligible patients received it. Whatever the reasons, and they are worth hearing, the intervention designed to make comprehensive treatment achievable did not by itself make it achievable.
The local failure comparison requires more care than it will receive. The relevant question is what constitutes the comparator group. If bone sites "not treated with SBRT" includes sites that received conventionally fractionated radiotherapy, then a 20.1 percent failure rate sits well above the published conventional benchmarks and demands explanation. If it includes sites that received no local therapy at all, the comparison is radiotherapy against its absence rather than one schedule against another, and it says nothing about SBRT specifically. Which composition applies will determine whether the figure means anything.
The structural limitation
Metastatic site radiotherapy was delivered after 14 cycles of chemotherapy, roughly seven to eight months into treatment. Only patients who survived that interval without progression received it.
Any comparison between patients who received comprehensive metastasis-directed therapy and those who did not therefore compares favorable disease biology against early progression. In a population with 38 percent three-year event-free survival driven by early distant failure, this is not a confounder to be adjusted for. It is a mechanism capable of generating an apparent survival benefit of nearly any size, and no adjustment available in a 298-patient cohort recovers randomization. The existing retrospective literature carries the same defect: Grewal and colleagues found a non-significant trend favoring metastasis-directed therapy in 27 patients, with an odds ratio of 0.24 and a p-value of 0.38, which is exactly the pattern a small selected series produces.
The prespecified endpoints avoid this, because feasibility and local failure within treated sites are descriptive rather than comparative. The risk is in the discussion that follows.
What the report should be examined for
The proportion receiving genuinely comprehensive treatment. Treating all sites and treating at least one site are different strategies with different rationales. The clinical value of consolidation depends on the former, and the achievable rate is the number the field needs.
Why eligible patients did not receive SBRT. If the reasons are logistical, that is a solvable problem and identifying it is the most actionable output of the analysis. If the reasons are anatomic or dosimetric, the eligibility criteria themselves need revision.
Who could not be treated at all. Bone marrow involvement is not amenable to site-directed therapy. The proportion of the metastatic population for whom comprehensive treatment was never possible defines the reach of the strategy and is easily lost when the denominator is restricted to SBRT recipients.
How local failure was estimated. With high early mortality from distant progression, death competes with local failure. Cumulative incidence with death as a competing risk and Kaplan-Meier estimation give materially different answers. Which was used should be stated.
Patterns of failure. This is the analysis with the most information in it. If treated sites remain controlled while progression arises at untreated baseline metastases, the biological case for comprehensive consolidation strengthens considerably without requiring randomization. If progression arises predominantly as new disease at new sites, local therapy is working as intended and the limitation is systemic.
Late skeletal toxicity. Forty gray in five fractions to bone in children and adolescents raises questions absent from adult SBRT: growth arrest, deformity, fracture, and marrow reserve. Acute toxicity will likely be acceptable and is not the concern. Whether follow-up is long enough to address the rest should be stated rather than left implicit.
Assessment
Two questions are entangled in this presentation and they deserve to be separated.
The technical question is whether 40 Gy in five fractions reliably sterilizes appropriately selected Ewing bone metastases. The trial can address this, and the answer is likely to be yes, though it should be read against conventional fractionation benchmarks that are at least as good rather than against no treatment.
The strategic question is whether consolidating every known metastatic site improves the course of the disease. The trial cannot address this, and the design forecloses it in a way that no analysis can repair.
The value of the report lies elsewhere, in the operational data. Pediatric oncology has incorporated SBRT slowly and for defensible reasons, and this trial is the first cooperative-group attempt to find out what happens when it is offered at scale. A trial that recommended comprehensive local therapy and achieved it in a minority of eligible patients has produced a genuinely useful finding about the gap between protocol intent and delivery.
Expected findings
We anticipate that local control at SBRT-treated bone sites will be high, in the range suggested by both the preliminary analysis and the external Ewing SBRT literature, and that acute toxicity will be acceptable with late skeletal outcomes reported as immature. We anticipate an association between comprehensive metastasis-directed therapy and better survival, and we expect it to be presented with appropriate caution and received without it.
The finding that would most change practice is not on the efficacy side. It is a clear account of why 62 percent of SBRT-eligible patients did not receive it. If the barriers turn out to be logistical, that is directly fixable in the next protocol and would do more for comprehensive consolidation in this disease than any local control figure. If they turn out to be clinical, the eligibility criteria were wrong and should be rewritten before the strategy is exported.
The comparison to be wary of throughout is survival in patients who completed comprehensive metastasis-directed therapy against survival in those who did not. In a disease where reaching consolidation is itself a marker of favorable biology, that curve will separate regardless of whether the treatment works.
We will report on the presentation from ASTRO 2026.
This is the fourth article in the RadOncToday series previewing the Clinical Trials and Plenary sessions at ASTRO 2026 over the three weeks before the meeting.