RADIOSA NCT03940235
ForMetachronous oligorecurrent hormone-sensitive prostate cancer, ≤3 lesions
TL;DRcPFS 32.2 vs 15.1 mo, HR 0.43, adding 6 months of ADT to ablative SBRT in metachronous oligorecurrence.
Surfaced from a review's discussed trials
The RT is nearly free here: one G3 event (left ureter stenosis) and no late toxicity across 105 pts treated to a stated BED >100 Gy, so the entire cost of this decision is 6 months of castration. SBRT alone still gave 15.1-mo cPFS, the benchmark for deferring systemic therapy in a selected pt.
In a man with metachronous oligorecurrence, three or fewer nodal or bone lesions after radical local treatment, this supports adding a short ADT course to ablative SBRT over SBRT alone; it does not extend to synchronous, higher-volume, or castration-resistant disease.
The RT side of the decision is nearly free: one G3 event (left ureter stenosis) and no late toxicity at a stated BED >100 Gy. The 15.1-mo SBRT-alone cPFS is the benchmark that ablative prescription buys, so a softer nodal dose is not this control arm.
The systemic question is duration, not agent: 6 months of LHRH started within a week of SBRT, 22 G1 events, all resolved. cPFS 32.2 vs 15.1 mo (HR 0.43) with no OS or castration-resistance readout leaves open whether this modifies disease or defers detection.
8 details 5 trials watching
Single-centre randomised open-label phase 2 at the European Institute of Oncology, Milan. 105 pts assigned 1:1 between Aug 1, 2019 and April 30, 2023; median follow-up 31 months (IQR 16-36). Modified intention-to-treat analysis, 3 pts lost to follow-up.
Metachronous oligorecurrent hormone-sensitive disease: biochemical progression after radical local treatment with ≤3 lesions (pelvic nodal, extra-regional nodal, or bone) on next-generation imaging, ECOG 0-1. Median age 70 (IQR 65-75). Stratified by doubling time (≤3 vs >3 mo), node vs bone, and PET vs MRI.
30 Gy in 3 fractions every other day to all oligometastatic sites, or equivalent regimens by site. Stated EQD2 98.6 Gy (α/β 1.5 Gy) and BED >100 Gy, an ablative prescription rather than a symptomatic one, which is the dose the control-arm result belongs to.
6 months of ADT with an LHRH analogue in the combination arm, begun within 1 week before SBRT; no ADT in the control arm. The RT backbone is fixed across arms, so the randomised variable is the short systemic course alone.
Primary: clinical progression-free survival, assessed in 51 pts per group. No overall survival, ADT-free survival, or castration-resistance readout is reported in source.
Radiation contributed one G1 GI event; the only G3 was a left ureter stenosis in the combination arm, with no late toxicity in either group. The 22 G1 events were ADT-attributed and had resolved by last follow-up, so the harm ledger sits with the systemic course, not the RT.
STOMP and ORIOLE established MDT against observation in metachronous oligorecurrence, partly on the strength of keeping men off systemic therapy. EXTEND tested the mirror-image addition, MDT layered onto hormone therapy. This is the first randomised trial in this setting to report improved cPFS for the combination.
Progression was assessed unmasked, and PSA suppression from the randomised ADT gates the restaging that defines the event. Median follow-up sits below the combination arm's median cPFS of 32.2 mo, so the tail after testosterone recovery is thin. No biomarker separates the pts who would do well on SBRT alone.
Local and systemic therapy read as additive here rather than redundant, but the trade is priced in cPFS only. Whether 6 months is the right duration, and for whom SBRT alone suffices, is exactly what the investigators leave to future biomarker work.
CONSORT flow
Randomised, prespecified primary cPFS hit (HR 0.43). Single-centre, N=105, surrogate composite endpoint, so it supports adding short ADT to MDT rather than establishing it.
- Optimal ADT duration alongside metastasis-directed SBRT active Testing the Addition of the Drug Relugolix to the Usual Radiation Therapy for Advanced-Stage Prostate Cancer, The NRG Promethean Study Phase 2n=194 · primary completion 2029-02 · SBRT +/- relugolix vs placebo in oligomet CSPCn=162 · primary completion 2031-04 · randomises RDT +/- ADT in radiorecurrent oligomet CSPC
- Biomarkers identifying pts who do well with SBRT alone active Immune Response Evaluation in Oligorecurrent and Oligoprogressive Prostate Cancer Patients Treated With SBRT Phase NAn=40 · primary completion 2026-07 · immune profiling of SBRT + ADT in oligorecurrence
- Whether cPFS benefit translates to OS or delayed castration resistance active Prostate-cancer Treatment Using Stereotactic Radiotherapy for Oligometastases Ablation in Hormone-sensitive Patients Phase 3n=550 · primary completion 2026-06 · phase 3, n=550, SBRT + SOC in oligomet HSPCrecruiting Metastasis Directed Stereotactic Body Radiotherapy for Oligo Metastatic Hormone Sensitive Prostate Cancer Phase NAn=118 · primary completion 2031-12 · phase 3 MD-SBRT vs standard tx, failure-free survival
📚 Sources · 📄 1 paper
Abstract
The longer read
The trade this trial proposes is months of castration for months of progression-free time, and the value of the result depends almost entirely on how a reader prices that trade. Metastasis-directed therapy earned its place in metachronous oligorecurrence partly because it kept men off systemic therapy; RADIOSA hands six months of that back and gets an HR of 0.43 on clinical PFS in return. EXTEND approached the same intersection from the opposite side, adding MDT to hormone therapy rather than hormone therapy to MDT, and it points the same direction: local and systemic treatment look additive in this disease state rather than redundant. What none of this settles is the exchange rate itself.
The endpoint is where confidence should move most. Clinical PFS in an unmasked trial where one arm receives six months of LHRH suppression is partly a measurement of that suppression. PSA is what triggers restaging in this population, castration lowers PSA, and the imaging event that defines progression is therefore gated in part by the randomised intervention. That is not a flaw peculiar to this trial, but it means the hazard ratio describes time to a detected event rather than a demonstrated change in disease trajectory. With no overall survival or castration-resistance readout in the source, whether the delay is durable or merely deferred detection stays open.
Follow-up compounds it. The median of 31 months sits below the combination arm's median cPFS of 32.2 months, whose upper confidence bound was not reached. The separation is being read close to the point where the combination arm's events begin to accrue, and the segment of the curve after testosterone recovery is thin. A temporary systemic therapy layered on a fixed local backbone will separate early almost by construction; whether the gap survives washout is the part this dataset cannot yet show.
For a radiation oncologist the transferability question concerns the control arm, not the experimental one. Thirty gray in three fractions with a stated BED above 100 Gy is an ablative prescription, and the 15.1-month SBRT-alone result belongs to that prescription. Anyone treating nodal recurrence more softly is not running this control arm and should not assume the same benchmark. Toxicity settles the other half: one G3 event across 105 pts and no late events means the marginal harm attributable to RT was close to nothing, which puts the whole cost of the decision on six months of castration and the 22 low-grade events that came with it.
The claim the trial supports is narrow but real: in one centre, in pts selected down to three or fewer lesions, adding a short ADT course to ablative MDT delays clinical progression. It does not establish six months as the correct duration, and the investigators concede that carefully selected pts might still do well with SBRT alone without naming who those pts are. That missing selection rule, more than the effect size, is what stands between this result and a general recommendation.