Meta-analysis
SBRT for Ultra-Central Lung Tumors Meta-Analysis
ForUltra-central lung tumors (primary NSCLC or mets), PTV abutting/overlapping PBT
TL;DRPooled 1-yr LC 92% and 2-yr LC 89% after ultra-central lung SBRT; G5 toxicity 4%, mostly hemoptysis, across 27 mostly retrospective studies.
The toxicity signal follows PBT hot spots, not the prescription: HILUS's 15% G5 came with a 67% IDL and median main bronchus Dmax 97 Gy. ISRS response: prescribe to 80-85% IDL, cap PBT Dmax at 133-150 Gy BED3 (EQD23 80-90 Gy), undercover PTV before overdosing airway.
In an ultra-central lung tumor without endobronchial extension or ILD, this supports 60 Gy/8 or 60 Gy/15 with a strict PBT hot-spot limit; it does not extend to endobronchial disease, where ISRS favors 60 Gy/15 with PBT Dmax <100 Gy BED3. The longer read
Airway hot spots drive fatal toxicity: HILUS's 15% G5 came with a 67% IDL and median main bronchus Dmax 97 Gy. ISRS asks for 80-85% IDL prescriptions, PBT Dmax <133-150 Gy BED3, and 60 Gy/15 with PBT Dmax <100 Gy BED3 for endobronchial disease.
Concurrent systemic agents are a modifiable risk: VEGF inhibitor within 30 days of SBRT carried fatal hemoptysis HR 16.9 in one 88-pt series. ISRS asks for 3-4 wk around VEGF inhibitors, 2-3 days for targeted agents, ≥2 days for IO, and holding antiplatelet/anticoagulation when safe.
14 details 4 trials watching
Systematic review and meta-analysis (PRISMA/MOOSE) of 27 studies, 1183 pts, published 2010-2022; random-effects pooling with meta-regression. All retrospective except HILUS (prospective phase II); none randomized. Median f/u ranged 7.6-44.5 mo.
Ultra-central lung tumors treated with SBRT (≥5 Gy/fx, photons), ≥5 eligible lesions per study. Excluded reirradiation and hilar/mediastinal nodal targets. Histology mixed; 9 studies (33%) were entirely primary NSCLC.
Dose 30-70 Gy in 3-10 fx, most often 50 Gy/5, 60 Gy/8, 60 Gy/12. Motion management in 96% of studies, mostly 4DCT. OAR dosimetry reporting was variable.
1-yr LC 92% (95% CI 86-96), 2-yr LC 89% (82-94); primary NSCLC subset 95% at both timepoints. Higher BED tracked with 1-yr LC (HR 2.14, p=0.03), not 2-yr (p=0.66).
| Endpoint | Pooled | 95% CI | Range | I² |
|---|---|---|---|---|
| 1-yr LC (21 studies) | 92% | 86-96 | 54-100% | 93.2% |
| 2-yr LC (19 studies) | 89% | 82-94 | 57-100% | 94.3% |
| Any G3-4 toxicity | 6% | 3-10 | 0-35% | 71% |
| Any G5 toxicity | 4% | 2-6 | 0-22% | 68% |
| G5 hemoptysis | 1% | 0-3 | 0-15% | 65% |
| G3-4 pneumonitis | 1% | 0-2 | 0-11% | 5% |
| Category | Recommendation |
|---|---|
| Delivery/dosing | 60 Gy/8 or 60 Gy/15; prescribe to 80-85% IDL; global Dmax far <150% |
| PBT constraint | Dmax <133-150 Gy BED3 (EQD23 <80-90 Gy); lower if risk factors or near trachea/main bronchi |
| Antiplatelet/anticoagulation | Hold when pausing is low-risk; else de-escalate PBT Dmax or use non-SBRT schedule |
| Targeted/IO | Gap 2-3 days targeted, 3-4 wk VEGF inhibitors, ≥2 days IO |
| Endobronchial tumor | Non-SBRT (60 Gy/15); PBT Dmax <100 Gy BED3 (EQD23 <60 Gy) |
| ILD | Lung V20 <10% strict cutoff; treat on trial |
Any G3-4 6% (3-10), any G5 4% (2-6), G5 hemoptysis 1% (0-3). G5 ≥10% in 7 studies; HILUS 15%. Risk factors: endobronchial tumor, antiplatelet/anticoagulant use, VEGF inhibitor within 30 days (HR 16.9), ILD.
Pooled 2-yr LC sits beside RTOG 0813 (89.4% in central/UC tumors). A prior 250-pt systematic review reported higher toxicity (treatment-related mortality 5%, G≥3 10%); the authors attribute the lower rates here to more contemporary, conformal series.
Fig 3 dose-event mapping drew on only 37 events from 9 studies, and PBT substructure dosimetry was rarely reported, so constraints beyond HILUS's could not be derived. Toxicity attribution in retrospective series is open to misclassification.
Efficacy is not the question; airway dose is. The ISRS constraint (PBT Dmax <133-150 Gy BED3) is built from cross-series convergence of fatal-hemoptysis doses, not a modeled threshold, and is pending SUNSET and LUNGTECH.
Pooled retrospective series (one prospective, none randomized) with I² >90% for LC, publication bias on every endpoint, and inconsistent ultra-central definitions. Guideline layer is expert opinion.
- Safe PBT dose limits by substructure (trachea, main vs lobar bronchi)
- Prospective toxicity of 60 Gy/8 vs 60 Gy/15 for ultra-central tumors n=30 · primary completion 2021-04 · phase 1 escalation from 60 Gy/8, G3 toxicity endpoint
- Role of MR-guided adaptive SBRT for ultra-central targets recruiting MR-informed Stereotactic Radiotherapy for Treatment of Ultracentral Lung Tumours Utilising a Dedicated MR-simulator for Daily Adaptation Followed by CBCT-guided Treatment Delivery Phase NAn=11 · primary completion 2028-12 · MR-sim daily adaptation, CBCT delivery, ultracentralactive Magnetic Resonance Guided Adaptive Stereotactic Body Radiotherapy for Lung Tumors in Ultracentral Location Phase 1n=38 · primary completion 2030-11 · phase 1 MTD of MR-linac SBRT, daily plan adaptationnot yet Multicentre Phase II Trial Evaluating Stereotactic Body Radiotherapy on Ultra-central Lung Tumors on MR-Linac Phase NAn=59 · primary completion 2031-11 · multicentre MR-linac SBRT safety, ultracentral tumors
📚 Sources · 📄 1 paper
Locoregional recurrence after neoadjuvant chemotherapy: pooled analysis of nine trials
ForPrimary breast cancer after anthracycline/taxane NACT, incl. cT4a-d
TL;DR5-yr LRR 4.1% with pCR vs 9.5% without (multivariate HR 3.33); TNBC non-pCR 18.5%. Surgery type not independent after NACT.
Residual disease interacts with subtype: TNBC non-pCR LRR was 18.5% at 5y versus 4.2% after pCR, and HR-/HER2+ 14.8% vs 3.1%. With RT use unrecorded, these rates show where post-neoadjuvant RNI/PMRT escalation trials should focus. Within pCR, only cN+ trended toward risk (HR 1.61, p=0.054), relevant to node-based de-escalation.
In TNBC or HR-/HER2+ breast cancer with residual disease after NACT, this backs a high-LRR-risk read when weighing PMRT and nodal coverage; after pCR it shows low LRR across subtypes but cannot say whether omitting RT is safe, because RT delivery was not recorded. The longer read
Residual-disease TNBC (18.5%) and HR-/HER2+ (14.8%) carry the 5-yr LRR burden, while pCR flattens every subtype to 3.1-4.8%. RT delivery was unrecorded, so this sets risk strata for PMRT/RNI decisions after NACT rather than proving that RT can be omitted after pCR.
Mastectomy's crude LRR excess (11.3% vs 7.8% for BCS) disappears on multivariate analysis, and surgery type predicted LRR in no subtype. After NACT, choosing mastectomy over BCS did not independently improve local control; response and biology did.
13 details 4 trials watching
Exploratory individual-patient-data pooled analysis of 9 prospective GBG/AGO-B neoadjuvant trials (GeparDuo through GeparSepto, PREPARE, TECHNO), 1998-2013. N=10,075; median follow-up 67 mo. Pathologic response was analysed with landmarking (multivariate set N=9552).
Histologically confirmed primary invasive breast cancer, tumour mostly ≥2 cm; locally advanced cT4a-c and inflammatory cT4d eligible in all trials except GeparDuo. 63.6% HR-positive, 28.1% HER2-positive; pCR (ypT0 ypN0) in 20.4%.
Anthracycline/taxane-based NACT in all trials. Anti-HER2 therapy varied by trial (trastuzumab, lapatinib, pertuzumab in later trials; none in GeparDuo, GeparTrio, PREPARE). Endocrine therapy ≥5 years for ER/PgR-positive pts.
Protocols recommended adjuvant RT per national guidelines, with RNI indicated mainly by nodal stage. No patient-level RT data were available, so receipt, fields and dose are unknown.
LRR as first event (breast/chest wall, axillary, infra/supraclavicular nodes) measured from surgery; distant recurrence, second malignancy and death were competing events.
5-yr LRR 7.8% BCS vs 11.3% mastectomy, but surgery type was not an independent predictor in multivariate analysis. Age, cN, grade, subtype and pCR were independent; non-pCR HR 3.33 and TNBC HR 2.72.
| Stratum | pCR | Non-pCR |
|---|---|---|
| HR+/HER2- | 3.9% | 5.9% |
| HR+/HER2+ | 4.8% | 8.1% |
| HR-/HER2+ | 3.1% | 14.8% |
| TNBC | 4.2% | 18.5% |
| BCS | 4.0% | 8.8% |
| Mastectomy | 3.9% | 11.3% |
The authors place this alongside the EBCTCG meta-analysis, which found higher local recurrence after NACT than after adjuvant chemotherapy, especially with suboptimal local therapy. The NSABP B-51 question (RNI after nodal pCR) is the RT decision these data frame but cannot answer.
Anti-HER2 therapy was inconsistent across eras, which the authors use to explain the lower LRR in HR+/HER2+ disease. gBRCA status incomplete; the univariate surgery effect is likely confounded by selection.
Biology and response drive LRR after NACT more than surgery type. The non-pCR TNBC and HR-/HER2+ groups are where locoregional escalation has the most room to show benefit. With RT unrecorded, the low post-pCR rates are not evidence that RT can be omitted.
Large mature IPD pool of prospective trials confirming known LRR predictors after NACT. Exploratory, prognostic only, and blind to RT delivery, so it cannot test an RT decision.
- Is RNI or PMRT safely omitted after nodal pCR in initially cN+ pts? active Standard or Comprehensive Radiation Therapy in Treating Patients With Early-Stage Breast Cancer Previously Treated With Chemotherapy and Surgery Phase NAn=1636 · primary completion 2023-09 · randomized regional nodal XRT vs none after chemo + surgeryn=100 · primary completion 2028-06 · RNI de-escalation in biopsy-proven cN1 → ypN0 after NATrecruiting Axillary Management in Breast Cancer Patients With Needle Biopsy Proven Nodal Metastases After Neoadjuvant Chemotherapy Phase NAn=1900 · primary completion 2030-02 · omit ALND + axillary RT if SLN negative post-NACT, cN1
- Does locoregional RT escalation reduce LRR in non-pCR TNBC? recruiting Internal Mammary Lymph Nodes Irradiation in High-risk Breast Cancer After Neoadjuvant Chemotherapy Phase 3n=722 · primary completion 2032-07 · phase 3 IMN RT vs none, ypN+ after NACT, DFS endpoint
- Do post-neoadjuvant T-DM1 or capecitabine lower LRR?
📚 Sources · 📄 1 paper
EBCTCG 20-year recurrence after 5 years of endocrine therapy
ForER+ T1-T2, 0-9 nodes, disease-free after 5y adjuvant endocrine therapy
TL;DRAfter stopping endocrine therapy at 5y, distant recurrence in years 5-20 ran from 13% (T1N0) to 41% (T2N4–9), at a steady annual rate.
In ER+ T1-T2 early breast cancer disease-free at 5 years, this frames the extension discussion around TN status, with T1N0 still carrying 13% distant risk over years 5 to 20; it does not address genomic-assay-selected or trastuzumab-treated HER2+ pts. The longer read
11 details
EBCTCG individual patient data meta-analysis of 88 trials. Kaplan-Meier and Cox regression, stratified by trial and treatment, from year 5 to year 20.
62,923 women with ER-positive T1-T2 disease, 0-9 positive nodes, diagnosed before age 75, scheduled to stop endocrine therapy at 5 years and event-free at year 5. 63% tamoxifen, 17% AI, 20% a tamoxifen-AI sequence; all had lymph-node dissection.
Distant recurrence, any breast-cancer event (distant, locoregional, contralateral), and death from breast cancer by log-rank subtraction.
Recurrences continued at a steady rate from year 5 to 20 in every TN category. Given TN status, grade and Ki-67 carried only moderate independent value after year 5; PR and HER2 status were not predictive.
| Nodal status | T1 | T2 |
|---|---|---|
| N0 | 13% | 19% |
| N1–3 | 20% | 26% |
| N4–9 | 34% | 41% |
| Grade | Distant recurrence | Any recurrence or contralateral BC |
|---|---|---|
| Low | 10% | 17% |
| Moderate | 13% | 22% |
| High | 17% | 26% |
| Nodal status | Distant recurrence | Death from breast cancer |
|---|---|---|
| N0 | 22% | 15% |
| N1–3 | 31% | 28% |
| N4–9 | 52% | 49% |
The source restates EBCTCG data that 5y tamoxifen cut recurrence by about 30% in years 5-10 versus no therapy. Extension trials such as ATLAS and MA.17 test the intervention; this analysis supplies the untreated baseline risk that their proportional benefit acts on.
Grade and Ki-67 were often locally assessed, and Ki-67 was available in only 7692 pts. Few pts received trastuzumab (2%), and treatment eras predate modern staging and systemic therapy.
No TN subgroup had a late risk low enough to dismiss extension outright, even low-grade T1N0 at 10% distant recurrence. Whether extension actually lowers mortality, and in whom it outweighs toxicity, is left to randomised data.
Large multi-trial IPD prognostic analysis, not a randomised test of extension. Quantifies baseline late risk; extended-therapy benefit must come from separate randomised trials.
- Does a genomic low-risk score identify pts safe to stop at 5 years?
- Does extended endocrine therapy reduce breast-cancer mortality, not just recurrence?
- How much do modern-era treatments lower these late risks?
📚 Sources · 📄 1 paper
NSABP B-15/B-16/B-18/B-22/B-25 IBTR and LRR prognosis
ForNode-positive breast, lumpectomy + WBI + adjuvant chemo and/or tamoxifen
TL;DRAfter lumpectomy + WBI in node-positive pts, IBTR (HR 2.58) and other LRR (HR 5.85) predicted death; 5-yr OS 59.9% vs 24.1%.
Supraclavicular failure is the RT-relevant read: it made up 81 of 165 oLRRs and carried 5-yr DDFS 12.1% vs 31.5% after axillary recurrence. oLRR reached 14.6% at 10 yrs with ≥10 positive nodes, so nodal burden frames the regional-field question after BCS; field design was not reported in source.
In a node-positive pt with an isolated in-breast or regional recurrence after lumpectomy and WBI, this supports treating the event as a marker of high distant risk when weighing restaging and systemic options; it does not describe recurrence risk under taxane or HER2-directed therapy. The longer read
Supraclavicular failure dominated oLRR (81 of 165) and had 5-yr DDFS 12.1%, while oLRR hit 14.6% at 10 yrs with ≥10 nodes. This frames the regional-field question after BCS in node-positive disease, though WBI technique and nodal coverage were not reported in source.
An oLRR carried mortality HR 5.85 and an IBTR 2.58, with 5-yr DDFS 18.8% and 51.4% after each, so a locoregional relapse signals high distant risk. Whether chemotherapy at relapse helps was unresolved; regimens predate taxanes and HER2-directed therapy.
11 details
Pooled secondary analysis of 5 NSABP node-positive adjuvant trials (B-15, B-16, B-18, B-22, B-25). Cumulative incidence for IBTR and oLRR; IBTR and oLRR entered Cox models as time-varying covariates. Median time on study 13.3 yrs.
Of 10,010 women entered, 2,669 had breast-conserving surgery and were eligible. Negative margins were required; about two thirds had 1-3 positive nodes. B-18 contributed only its postoperative chemotherapy arm.
AC, CMF, PF/PAF and tamoxifen-based adjuvant regimens from 1984-1994, including AC cyclophosphamide intensification in B-22 and B-25. Pts aged >50 received tamoxifen in several protocols; premenopausal women did not receive tamoxifen irrespective of ER status.
All pts received whole-breast irradiation after lumpectomy. Dose, fractionation, boost and any regional nodal coverage are not reported in source.
Cumulative incidence of IBTR and oLRR; DDFS and OS after each event; mortality modeled with clinical factors plus IBTR or oLRR.
oLRR carried roughly double the mortality hazard of IBTR (HR 5.85 vs 2.58), and supraclavicular recurrence had the poorest DDFS. Arm-level values are in the detail table.
| Endpoint | IBTR | oLRR |
|---|---|---|
| Pts with event | 259 (9.7%) | 165 (6.2%) |
| 10-yr cumulative incidence | 8.7% | 6.0% |
| Events within 5 yrs | 62.2% | 80.6% |
| 5-yr DDFS after event | 51.4% | 18.8% |
| 5-yr OS after event | 59.9% | 24.1% |
| Mortality HR (time-varying Cox) | 2.58 (95% CI 2.11-3.15) | 5.85 (95% CI 4.80-7.13) |
| Adjusted HR, distant disease | 2.72 | 6.68 |
| Protocol | Pts at risk | 10-yr IBTR % | 10-yr oLRR % |
|---|---|---|---|
| B-15 | 624 | 10.0 | 7.6 |
| B-16 | 340 | 4.8 | 5.4 |
| B-18 | 210 | 9.1 | 8.7 |
| B-22 | 596 | 10.1 | 6.1 |
| B-25 | 899 | 8.3 | 4.5 |
| Total | 2,669 | 8.7 | 6.0 |
Consistent with the NSABP B-06 IBTR literature the authors cite and with the EBCTCG overview linking local failure to breast cancer mortality. The supraclavicular-dominant oLRR pattern is the population later addressed by randomized RNI trials such as MA.20 and EORTC 22922.
Lumpectomy was not the randomized variable, so the 2,669-pt subset is self-selected. Site-specific DDFS estimates rest on small numbers with wide CIs, and no HER2 or RT technique data are available.
oLRR clusters with nodal burden and receptor negativity, the same drivers of distant failure, so it may flag disseminated disease rather than cause it. Later IBTRs fared better, consistent with the authors' view that many are new primaries. The role of chemotherapy at recurrence was unresolved at publication.
Post-hoc pooled analysis of a non-randomized lumpectomy subset; LRR is a prognostic marker here, not a tested intervention target. Pre-taxane, pre-HER2 era therapy.
- Does regional nodal RT reduce supraclavicular failure in this population?
- Does chemotherapy at isolated locoregional recurrence improve survival?
- Can molecular profiling separate true IBTR from new ipsilateral primaries?
📚 Sources · 📄 1 paper
Abstract
MARCAP Consortium
ForLocalized prostate cancer, NCCN intermediate to high risk, definitive RT
TL;DRIPD meta-analysis, 10,853 pts: ADT added to RT improved MFS HR 0.83 and OS HR 0.86; dose escalation improved bRFS only.
Reported via UroToday →
The dose-invariance result is the load-bearing one: ADT's benefit was identical at ≥74 Gy and <74 Gy (both HR 0.83), so escalating dose does not buy out the ADT decision. Adjuvant prolongation carries the benefit (MFS HR 0.84), neoadjuvant prolongation does not (HR 0.95), which moves sequencing, not just duration.
In NCCN intermediate- and high-risk localized prostate cancer receiving definitive RT, this supports adding and prolonging ADT on the adjuvant side rather than substituting dose escalation; it does not address node-positive or post-prostatectomy salvage settings. The longer read
ADT's benefit was identical at ≥74 Gy and <74 Gy (HR 0.83 both), so escalation does not buy out the hormone decision, and HEAT showed escalation to 79.2 Gy improved biochemical RFS but not MFS or OS. Adjuvant prolongation carries the survival benefit; neoadjuvant prolongation (HR 0.95) does not.
The duration and sequencing read: extending adjuvant ADT from 4-6 to 18-36 months gave MFS HR 0.84 and OS HR 0.85, while extending neoadjuvant ADT from 3-4 to 6-9 months gave HR 0.95 for both. Relative benefit held across NCCN groups, so risk stratification changes NNT, not effect size.
9 details 5 trials watching
Individual patient data meta-analysis from the MARCAP Consortium, 12 eligible trials, 10,853 patients, median follow-up 11.4 years (IQR 9.0-15.0). Three intensification questions analyzed separately: ADT use, neoadjuvant ADT prolongation, adjuvant ADT prolongation.
Localized prostate cancer treated with definitive radiotherapy, stratified by NCCN prognostic risk group. The RT ± ADT comparison ran 2,557 control vs 2,579 experimental (NCCN high 32.6% vs 32.8%, intermediate 47.6% vs 48.7%); the adjuvant-prolongation comparison was high-risk enriched at 71.5% vs 71.1% NCCN high across 1,900 vs 1,874 pts.
Dose was analyzed as a binary ≥74 Gy vs <74 Gy split, with the companion HEAT analysis using 64 to <74 Gy vs ≥74 Gy and escalation up to 79.2 Gy. Brachytherapy boost is discussed as an escalation route alongside external-beam dose.
ADT addition improved MFS HR 0.83 (0.77-0.89) and OS HR 0.86 (0.80-0.92). Adjuvant prolongation improved MFS HR 0.84 (0.78-0.91) and OS HR 0.85 (0.78-0.94); neoadjuvant prolongation did neither (MFS HR 0.95, OS HR 0.95). 10-yr MFS 61% (59-63) with ADT vs 52% (50-54) without; 10-yr OS 65% (63-67) vs 57% (55-59).
| Intervention | Intermediate risk | High risk |
|---|---|---|
| ADT addition | 18.0 | 8.4 |
| Adjuvant ADT prolongation | 16.1 | 10.4 |
The 2022 HEAT network meta-analysis (13 trials, 11,862 pts) found dose escalation improved biochemical RFS but not MFS or OS, and named high-dose RT plus long-term ADT the best strategy. PROG 05/09 and GETUG-18 are invoked to explain the gap: a 21% 5-year biochemical benefit in low risk where metastasis is rare, versus 3% in high risk where it is not.
The pooled trials span 30+ years of practice, so the RT technique behind the <74 Gy stratum is not the technique a reader delivers today. The stated optimal durations (closer to 12 months high risk, 24+ months very high risk) come from DHARMA, accepted but unpublished, so that specific number is not yet auditable.
The framing claim is that dose escalation and ADT are not interchangeable levers: escalation buys biochemical control, ADT buys metastasis-free and overall survival, and the two do not substitute. The dogma that dose escalation obviates ADT is named as widely repeated and never demonstrated.
IPD meta-analysis of randomised trials, large N, long f/u. Reinforces ADT intensification over dose escalation; no new randomisation, and duration guidance rests on unpublished DHARMA.
- Optimal adjuvant ADT duration by prognostic risk group active A Study of Shorter Course Hormone Therapy and Radiation for High-risk Prostate Cancer Phase 2n=50 · primary completion 2026-10 · shorter-course ADT with brachy + hypofx EBRT, high-riskrecruiting Artificial Intelligence Driven Personalisation of Radiotherapy and Concomitant Androgen Deprivation Therapy for Prostate Cancer Patients (the HypoPro Trial) Phase 2n=30 · primary completion 2026-10 · MMAI classifier individualises ADT duration, high-risk
- Does dose escalation change ADT benefit with modern brachytherapy boost active Comparative Study of Radiotherapy Treatments to Treat High Risk Prostate Cancer Patients Phase 3n=307 · primary completion 2028-12 · phase 3 HDR brachy boost vs dose-escalated RT + ADTrecruiting Androgen Suppression Combined With Nodal Irradiation and Dose Escalated Prostate Treatment Phase 3n=710 · primary completion 2032-10 · phase 3 SBRT vs EBRT + brachy boost, all on ADTrecruiting PRO-BOOST-LC: Whole-Gland Boost Strategies Versus SBRT Monotherapy in PSMA-Staged Localized and Locally Advanced Prostate Cancer Phase 2/3n=1000 · primary completion 2033-12 · randomised brachy or SBRT boost vs SBRT alone, + ADT
- Whether intermediate-risk subsets warrant long-term ADT
📚 Sources · 📄 1 paper
Abstract
POSEIDON
ForPost-RP prostate cancer receiving PORT, median pre-PORT PSA 0.3 ng/mL
HR 0.87
95% CI 0.76–1.01, p=0.06; not met
TL;DRIPD meta-analysis of 6 RCTs: adding HT to PORT did not improve OS (HR 0.87, 95% CI 0.76–1.01, p=0.06); benefit confined to pre-PORT PSA >0.5 ng/mL.
MFS improved (HR 0.79) but never cleared the ICECaP 0.81 surrogate threshold at any pre-PORT PSA, so the bPFS/MFS gains behind GETUG-AFU 16 and RTOG 0534 were not expected to translate to OS. Duration showed no interaction (p=0.17); long-term HT OS benefit on splines appears only at PSA ≥1.6 ng/mL, above most modern early salvage.
In PET-negative BCR after RP heading to early salvage PORT at PSA ≤0.5 ng/mL, this questions adding HT for survival; above 0.5 ng/mL it supports HT, with no OS evidence favoring 24 over 4–6 mo; it does not extend to PSMA PET-positive nodal disease. The longer read
For early salvage PORT at pre-PORT PSA ≤0.5 ng/mL, HT added no OS benefit, so RT alone is a defensible default in PET-negative BCR. Above 0.5 ng/mL the OS HR falls to 0.72 to 0.69; long-term HT OS benefit on splines appears only at PSA ≥1.6 ng/mL.
Duration showed no OS interaction (p=0.17) and prolonging short to long course did not improve OS (HR 0.89), so 4–6 mo carries the supported benefit. MFS gains did not clear the ICECaP threshold, weakening MFS as a basis for systemic intensification with PORT.
14 details 5 trials watching
IPD meta-analysis of six randomised phase 3 trials via the MARCAP consortium (PROSPERO CRD42019134376), one-stage Cox stratified by trial. 6057 pts, median follow-up 9.0y (IQR 7.2–10.7).
Men receiving PORT after radical prostatectomy for recurrent prostate cancer. Median pre-PORT PSA 0.3 ng/mL in the any-HT and short-term sets, 0.5 ng/mL in the long-term set. Race and ethnicity data not available.
Short-term HT 4–6 mo (GETUG-AFU 16, NRG/RTOG 0534, RADICALS-HD two-way and three-way), predominantly GnRH agonist. Long-term HT 24 mo (RADICALS three-way, RTOG 9601); RTOG 9601 used high-dose bicalutamide monotherapy.
Primary: overall survival. Secondary: MFS, event-free survival, biochemical recurrence; prespecified interaction tests for pre-PORT PSA and HT duration.
No OS benefit overall (HR 0.87, p=0.06) despite MFS gain (HR 0.79, p<0.001). OS benefit limited to pre-PORT PSA >0.5 ng/mL (interaction p=0.02), with NNT 22 at 0.51–1.00 and 12 at >1.00 ng/mL.
| Comparison | N | OS HR (95% CI) | MFS HR (95% CI) |
|---|---|---|---|
| Any HT + PORT vs PORT | 5026 | 0.87 (0.76–1.01), p=0.06 | 0.79 (0.70–0.89), p<0.001 |
| Short-term HT (4–6 mo) | 3938 | 0.93 (0.77–1.11) | 0.82 (0.71–0.95) |
| Long-term HT (24 mo) | 1088 | 0.79 (0.63–1.00), p=0.049 | 0.74 (0.60–0.91), p=0.004 |
| Short → long prolongation | 1523 | 0.89 (0.68–1.16), p=0.37 | 0.76 (0.61–0.95), p=0.02 |
| Pre-PORT PSA (ng/mL) | OS HR (95% CI) | 10y OS PORT vs PORT+HT | 10y NNT (death) |
|---|---|---|---|
| ≤0.20 | 1.14 (0.83–1.57) | 89.9% vs 87.4% | negative |
| 0.21–0.50 | 0.94 (0.74–1.19) | 84.9% vs 84.5% | negative |
| 0.51–1.00 | 0.72 (0.54–0.96) | 76.4% vs 80.9% | 22 |
| >1.00 | 0.69 (0.48–0.98) | 72.4% vs 80.5% | 12 |
GETUG-AFU 16 and RTOG 0534 established HT with salvage RT on biochemical endpoints; per the ASCO Post write-up only RTOG 9601, a late-salvage trial, had shown an OS benefit. POSEIDON pools these trials and finds the OS signal tracks pre-PORT PSA rather than HT duration.
Duration comparison is confounded: long-term HT trials enrolled higher-PSA, higher-risk pathology. The PSA interaction is model-dependent (continuous PSA: one-stage p=0.19, two-stage p=0.04) and no multiplicity correction was applied. RT dose and field detail not reported in source excerpt.
The trial-level MFS gains held but did not clear the ICECaP surrogate threshold, so the historical case for routine HT with low-PSA salvage rests on endpoints that do not predict OS here. Short-term HT may suffice where HT is used; NRG GU006 (PAM50, apalutamide) tests biomarker selection.
Pooled randomised IPD with prespecified OS primary not met, undercutting routine HT with low-PSA PORT. The PSA >0.5 benefit rests on subgroup interaction, not a primary comparison.
- Does PSMA PET staging change which pts benefit from HT with PORT?
- Can a genomic classifier outperform pre-PORT PSA for selecting HT?
- Does ARSI-based short-term HT change the OS read? active Radiation Therapy With or Without Apalutamide in Treating Patients With Recurrent Prostate Cancer, the BALANCE Trial Phase 2n=298 · primary completion 2025-03 · ph2 SRT ± apalutamide, placebo-controlled, post-RPn=586 · primary completion 2026-11 · ph3 apalutamide added to ADT + RT, pN+ post-RPactive Combined Apalutamide, Radiotherapy, and LHRH Agonist in Prostate Cancer Patients After Prostatectomy Phase 3n=490 · primary completion 2028-09 · ph3 apalutamide + SRT + LHRHa vs SRT + LHRHa post-RPactive A Study of Adding Apalutamide to Radiotherapy and LHRH Agonist in High-Risk Patients With Hormone-Sensitive Prostate Cancer Phase 3n=691 · primary completion 2029-08 · ph3 apalutamide + RT + LHRHa post-RP, PSMA-PET 1° EPrecruiting Treating Prostate Cancer That Has Come Back After Surgery With Apalutamide and Targeted Radiation Based on PET Imaging Phase 3n=804 · primary completion 2032-12 · ph3 apalutamide+abiraterone added to SRT + short ADT
📚 Sources · 📄 2 papers
Abstract
STAMPEDE abiraterone ± enzalutamide in high-risk non-metastatic prostate cancer (meta-analysis) NCT00268476
ForHigh-risk non-metastatic prostate cancer (N+ or high-risk N0) starting ADT
HR 0.53
95% CI 0.44-0.64, p<0.0001; 6-yr MFS 82% vs 69%
TL;DRMFS HR 0.53 (0.44-0.64) and OS HR 0.60 (0.48-0.73) adding 2yr abiraterone ± enzalutamide to ADT in high-risk non-metastatic prostate cancer.
The MFS benefit held whether or not RT was planned: HR 0.54 (0.44-0.67) with RT vs 0.51 (0.34-0.76) without, interaction p=0.67. So abiraterone adds to prostate RT and does not stand in for it. The RT was 74 Gy/37 fx to prostate + SVs, and the source does not describe pelvic nodal fields for the 39% who were N+.
In N+ or high-risk N0 non-metastatic prostate cancer starting long-course ADT with prostate RT, this supports 2 yr of abiraterone + prednisolone without enzalutamide. It does not extend to men treated with prostatectomy, who were not studied. The longer read
The abiraterone benefit is independent of planned RT (MFS HR 0.54 with vs 0.51 without, interaction p=0.67), and RT was mandated for N0. It stacks on 74 Gy/37 fx to prostate + SVs with 3 yr ADT. The source does not describe pelvic nodal fields, and only 71% of N+ pts had RT planned.
Two years of abiraterone + prednisolone added to 3 yr ADT improved OS (HR 0.60). Adding enzalutamide gave no extra MFS benefit (interaction HR 1.02) and raised grade 3+ AEs to 58% vs 32%. Enzalutamide adds cost and toxicity without added efficacy.
Also covered Oct 24
11 details 4 trials watching
Prespecified IPD meta-analysis of two open-label, randomised (1:1) phase 3 trials in the STAMPEDE platform, run at 113 UK and Swiss sites. N=1974, randomised 2011-2016, with no overlapping controls. Median follow-up was 72 mo.
High-risk non-metastatic prostate cancer on conventional imaging: N+, or N0 with at least two of T3/T4, Gleason 8-10, or PSA ≥40 ng/mL. Pts relapsing with high-risk features were also eligible. Median age was 68, and 774 (39%) were node-positive. Clinically significant cardiovascular disease was excluded.
All pts received 3 yr ADT. The combination arm added abiraterone 1000 mg + prednisolone 5 mg daily for 2 yr, and the second trial also added enzalutamide 160 mg. If RT was omitted, treatment could continue until progression.
Per local guidelines: 74 Gy/37 fx to prostate + SVs or a hypofractionated equivalent. RT was mandated for N0 and encouraged for N+. RT was planned in 85% overall: 99% of N0, 71% of N+ and 7% of previously treated pts.
Primary: MFS, which was changed from OS before any efficacy inspection and approved by the steering committee on Dec 2, 2019. Secondary endpoints were OS, prostate cancer-specific survival, biochemical failure-free survival, PFS and toxicity.
Both trials showed the benefit on their own: MFS HR 0.54 in the abiraterone trial and 0.53 in the abiraterone + enzalutamide trial, with no between-trial heterogeneity (I² p=0.90).
| Endpoint | Combination | Control | HR (95% CI) |
|---|---|---|---|
| MFS (1°) | 6-yr 82% | 6-yr 69% | 0.53 (0.44-0.64), p<0.0001 |
| OS | 6-yr 86% | 6-yr 77% | 0.60 (0.48-0.73), p<0.0001 |
| Prostate cancer-specific survival | median not reached | median not reached | 0.49 (0.37-0.65), p<0.0001 |
| Biochemical failure-free survival | median not reached | median 86 mo | 0.39 (0.33-0.47), p<0.0001 |
| PFS | median not reached | median not reached | 0.44 (0.36-0.54), p<0.0001 |
| Subgroup | HR (95% CI) | p interaction |
|---|---|---|
| RT planned: yes | 0.54 (0.44-0.67) | 0.67 |
| RT planned: no | 0.51 (0.34-0.76) | 0.67 |
| N0 | 0.60 (0.46-0.78) | 0.22 |
| N+ | 0.49 (0.38-0.64) | 0.22 |
| WHO PS 0 | 0.47 (0.38-0.58) | 0.00656 |
| WHO PS 1-2 | 0.86 (0.58-1.28) | 0.00656 |
| NSAID/aspirin: no | 0.62 (0.51-0.77) | 0.0052 |
| NSAID/aspirin: yes | 0.32 (0.21-0.48) | 0.0052 |
| AE | Abi trial: combo | Abi trial: control | Abi+enza trial: combo | Abi+enza trial: control |
|---|---|---|---|---|
| Any grade 3+ | 37% | 29% | 58% | 32% |
| Hypertension | 5% | 1% | 14% | 2% |
| ALT increase | 6% | <1% | 13% | 1% |
Grade 3+ AEs in the first 24 mo were 37% vs 29% in the abiraterone trial and 58% vs 32% in the abiraterone + enzalutamide trial. The main drivers were hypertension and ALT elevation. There were 7 grade 5 events, all in the combination arms.
The backbone is long-course ADT with prostate RT, which the authors frame as standard for this population. Adding enzalutamide to abiraterone also gave no extra benefit here, consistent with the modest rPFS gains and absent OS benefit the authors cite from that combination in mCRPC.
Staging used conventional imaging only. Toxicity was reported only for the first 2 yr, and data on subsequent therapy are unreliable. Heterogeneity by PS 1-2 (HR 0.86, 0.58-1.28) makes the benefit less certain in less fit men.
Abiraterone intensification improved MFS and OS on top of ADT whether or not RT was planned. Enzalutamide added toxicity without extra efficacy. Open questions are the right duration, the benefit after prostatectomy, and the benefit in PSMA-staged disease, none of which were tested.
Pooled IPD from two randomised phase 3 trials, 72-mo f/u, with both MFS and OS improved. The OS→MFS switch was prespecified before efficacy was seen. Authors call it a new standard.
- Is 2 years the optimal abiraterone duration? recruiting Reduced Length of ADT and ARTA With XRT in High-Risk Prostate Cancer (RELAX): A Randomised Trial Phase 2n=206 · primary completion 2035-01 · 9mo ADT + 6mo ARTA vs 2y ADT with RT, high-risk N0M0
- Benefit after radical prostatectomy untested recruiting Treating Prostate Cancer That Has Come Back After Surgery With Apalutamide and Targeted Radiation Based on PET Imaging Phase 3n=804 · primary completion 2032-12 · post-RP BCR: apalutamide+abiraterone added to SRT+ADT
- Benefit in relapse after prior local therapy underrepresented active Finite Androgen Ablation With or Without Abiraterone Acetate and Prednisone in Treating Patients With Recurrent Prostate Cancer Phase 2n=310 · primary completion 2027-02 · recurrent PCa: finite ADT ± abiraterone, randomisedrecruiting Treating Prostate Cancer That Has Come Back After Surgery With Apalutamide and Targeted Radiation Based on PET Imaging Phase 3n=804 · primary completion 2032-12 · phase 3 ARSI intensification in post-RP BCR
📚 Sources · 📄 1 paper
HYDRA
ForLocalised prostate cancer, definitive external-beam RT
TL;DRNo PFS difference for either isodose (HR 0.92) or dose-escalated MHFRT (HR 0.94), but escalation raises late G2+ GI (OR 1.48).
The split that matters is isodose vs dose-escalated MHFRT, not hypofractionation itself: escalation adds no PFS (HR 0.94, 0.82-1.09) and costs bowel on both physician grading (OR 1.48) and patient report (OR 1.68). GU was unchanged in both strata. The schedule decision lands on 60 Gy in 20 fractions.
In a man with localised prostate cancer starting definitive prostate-only EBRT, this supports an isodose moderately hypofractionated schedule over a dose-escalated one; it does not extend to five-fraction ultrahypofractionation, post-prostatectomy salvage, or whole-pelvis treatment. The longer read
The decision this moves is schedule selection, not modality: escalated MHFRT shows no PFS gain (HR 0.94, 0.82-1.09) and raises late G2+ GI (OR 1.48) plus patient-reported bowel decrement (OR 1.68), while isodose shows neither. 60 Gy in 20 fractions is the defensible default for prostate-only volumes.
7 details 5 trials watching
IPD meta-analysis of randomised phase 3 CFRT vs MHFRT trials via the MARCAP consortium. Searches on Dec 15, 2023 and re-run Jan 8, 2025 screened 1696 records down to 7 eligible trials. Three separate analyses: efficacy, physician-scored late toxicity, and patient-reported outcomes.
Localised prostate cancer on trials that published patient-level efficacy AND late toxicity data. 3454 pts across three isodose trials, 2426 pts across four dose-escalated trials. Trials whose CFRT arm fell below modern dose were excluded.
The intervention split is the whole point: isodose MHFRT (same equivalent dose in fewer fractions, eg 60 Gy in 20 fractions) versus dose-escalated MHFRT. The CFRT comparator had to deliver ≥70 Gy in 2 Gy equivalents.
Primary (efficacy): progression-free survival. Co-primary toxicity endpoints: late grade 2 or higher GU and GI. Co-primary PRO endpoints: clinically-significant decrement in urinary or bowel quality of life.
The GI signal sits entirely in the dose-escalated stratum and shows up on both physician grading and patient report; the isodose stratum carries neither. GU odds ran above 1 in both comparisons with intervals crossing unity.
CHHiP and PROFIT established 60 Gy in 20 fractions as non-inferior to conventional fractionation. The escalated schedules were built on the premise that a higher equivalent dose in fewer fractions would improve control; pooled here that premise fails on PFS while adding bowel toxicity.
Toxicity scales and PRO instruments were not uniform across the seven trials, and the dose-escalated stratum pools four schedules that are not interchangeable, so the OR describes escalation as a class, not one regimen. Follow-up also differs between strata (5.4 vs 7.1 yrs).
With efficacy answered as a null, the schedule decision turns entirely on toxicity, and the toxicity difference runs one way. Escalating per-fraction dose beyond isodose buys no measurable PFS while adding bowel morbidity that pts themselves report, which leaves little argument for an escalated MHFRT schedule in intact localised disease.
Pooled IPD from seven randomised phase 3 trials, aligned with existing moderate-hypofractionation practice; refines which regimen rather than establishing a new modality or population.
- Does the escalation bowel signal extend to five-fraction ultrahypofractionation? recruiting Comparing Moderately Ultra Hypofractionated Radiation Treatments for Prostate Cancer Phase 2n=204 · primary completion 2030-11 · randomised 20fx vs 5fx post-op bed +/- pelvisrecruiting Salvage Moderate Hypofractionated Versus Ultrahypofractionated Radiotherapy for Biochemical Recurrence After Radical Prostatectomy in Prostate Cancer Phase 3n=270 · primary completion 2034-12 · phase 3 moderate vs ultrahypo salvage, toxicity EP
- Do rectal spacers and daily IGRT narrow the dose-escalated GI gap? n=500 · primary completion 2027-12 · SpaceOAR Vue for late GI toxicity under SBRTn=84 · primary completion 2027-12 · Barrigel anterior rectal sparing in post-op RT
- Same toxicity read when MHFRT covers whole-pelvis nodal volumes? n=18 · primary completion 2026-08 · 20/16/12fx pelvic nodal RT with prostate SIB
📚 Sources · 📄 1 paper
Abstract
WOLVERINE
ForOligometastatic prostate cancer, up to 5 mets, mostly castration-sensitive
TL;DRIPD meta-analysis of 6 randomised phase 2 trials, 472 pts: MDT improved PFS (HR 0.44, 0.35-0.56), OS non-significant (HR 0.63, 0.39-1.00, p=0.051).
Surfaced from a review's discussed trials
The endpoint that survives every sensitivity cut is PFS (HR 0.44, and 0.46 excluding the observation-SOC trials), while OS stops at HR 0.63 (0.39-1.00, p=0.051). CRFS 0.58 in CSPC is the more decision-relevant signal, since delaying castration resistance is what MDT is being asked to buy. Nothing here resolves dose, target, or which oligo burden benefits.
In castration-sensitive oligometastatic prostate cancer with up to five lesions and a treated primary, this supports adding MDT for progression-free and castration resistance-free benefit; it does not establish an OS benefit, and the CRPC-only and untreated-primary populations are thinly represented. The longer read
The endpoint that survives every sensitivity cut is PFS (HR 0.44, and 0.46 excluding the observation-SOC trials), and CRFS 0.58 in CSPC is what MDT is being asked to buy. None of the constituent trials' dose, fractionation or target-selection choices are resolved by pooling, and rPFS carried I²=50% heterogeneity, so how to deliver MDT stays a local decision.
MDT deferred castration resistance (CRFS HR 0.58) in the castration-sensitive subset (n=257), which is the sequencing-relevant read: the question is whether local therapy buys time before ARPI escalation. Note the SOC arm actually got MORE second-generation ARPI (59.8% vs 50.4%), so the systemic backbone was not favouring MDT.
10 details 5 trials watching
Systematic review and individual patient data meta-analysis (X-MET collaboration), PROSPERO CRD42023479078. Searched Embase, PubMed, CENTRAL, MEDLINE and ClinicalTrials.gov to Nov 3 2023, updated May 4 2025; dual independent screening in Covidence, Cochrane RoB 2.0. Of 2975 studies screened, 7 phase 2 trials randomising 574 men were included.
Published randomised prospective trials in oligometastatic (up to five metastases) prostate cancer with data sufficient for PFS and OS. Most patients were castration-sensitive (n=375, 65%); ARTO enrolled entirely CRPC and the two EXTEND baskets a CRPC subset. The primary tumour had received prior definitive local therapy in 491 patients (85.5%), required in every trial except the EXTEND baskets and ARTO.
Co-primary: progression-free survival and overall survival. Secondary: radiographic PFS and castration resistance-free survival. Primary analysis restricted to the six trials randomising MDT plus SOC versus SOC, with both a random-effects trial-level analysis and a patient-level analysis stratified by trial.
Effect sizes are in the endpoint table. Trial- and patient-level estimates agreed closely on all four endpoints, and OS showed HR<1 in every individual trial without reaching significance in aggregate.
The constituent trials are the ones that currently drive MDT practice: STOMP and ORIOLE (small randomised phase 2, ADT-free intervals and progression), ARTO (MDT added to abiraterone in CRPC), the two EXTEND hormone baskets, and the prostate subgroup of SABR-COMET (16 men). Pooling them raises precision on PFS but cannot add the phase 3 evidence none of them supply, and the ongoing randomised phase 3 programmes remain the gate.
The SOC arm was not one thing: observation in all or part of STOMP, ORIOLE and COMET-SABR, and second-generation ARPI use differed between arms (59.8%, n=134 in SOC vs 50.4%, n=125 in MDT), which cuts against MDT rather than for it. Four abstract-only randomised primary analyses could not supply IPD and were excluded, and the prostate contribution from SABR-COMET is 16 men.
PFS is where the estimate is tight and consistent; OS is where the question stays open, and at p=0.051 the honest read is an underpowered signal, not a negative result. The endpoint most likely to matter to practice is CRFS (HR 0.58) in castration-sensitive disease, because deferring castration resistance is the outcome MDT is being asked to deliver.
Pools only phase 2 trials with non-blinded randomisation and mixed SOC; co-primary OS missed (p=0.051). Supports existing MDT practice rather than establishing level 1 evidence.
- Does MDT extend overall survival in a phase 3 population active Prostate-cancer Treatment Using Stereotactic Radiotherapy for Oligometastases Ablation in Hormone-sensitive Patients Phase 3n=550 · primary completion 2026-06 · phase 3 SBRT to all oligomets, 550 pts, mHSPCrecruiting Veterans Affairs Seamless Phase II/III Randomized Trial of STAndard Systemic theRapy With or Without PET-directed Local Therapy for Oligometastatic pRosTate Cancer Phase 2/3n=464 · primary completion 2026-09 · seamless ph2/3, CRPC-free survival, PET-directedrecruiting Metastasis Directed Stereotactic Body Radiotherapy for Oligo Metastatic Hormone Sensitive Prostate Cancer Phase NAn=118 · primary completion 2031-12 · randomised ph3 MD-SBRT vs standard tx, 1-3 mets
- Optimal dose, fractionation and target selection for prostate MDT recruiting OligoCare TwiCs (Trials Within Cohorts) Trial Comparing Acute Toxicity in Single-fraction vs Multiple-fraction SBRT for Metastasis-directed Treatment (SPRINT) Phase NAn=302 · primary completion 2029-02 · single- vs multi-fraction SBRT, acute toxicity 1° EPrecruiting SBRT Versus Hypofractionated Radiotherapy for Biochemically Recurrent or Oligometastatic Prostate Adenocarcinoma Phase 3n=118 · primary completion 2030-01 · ph3 SBRT vs hypofx RT in oligomet/BCR prostate
- Upper bound on metastasis number that still benefits
📚 Sources · 📄 1 paper
Abstract
FIRESTORM
ForHigh-risk meningioma (WHO gr 2 post-STR, recurrent gr 2, any gr 3), postop RT
HR 0.40
95% CI 0.24-0.69, P = .001 (MVA); IPTW HR 0.45 (0.24-0.83)
TL;DR5-yr PFS 65.8% vs 38.8% with dose-escalated postop RT (BED ≥79.2 Gy) in high-risk meningioma; MVA HR 0.40.
The benefit survives the obvious confounders: excluding the 35 single-fraction SRS cases from SD-RT, DE-RT still gave 5-yr PFS 65.8% vs 41.7% (HR 0.56, 0.36-0.86), and photon-alone escalation matched carbon on PFS (68.3% vs 61.1%) with 0% grade ≥3 radionecrosis. That makes 66-70 Gy conventionally fractionated photons the practical escalation route, not particle referral.
In a subtotally resected WHO grade 2 or any grade 3 meningioma being planned for postoperative RT, this supports considering escalation beyond 60 Gy/30 fx rather than defaulting to it; it does not speak to gross-totally resected grade 1 disease. The longer read
Photon-alone escalation to 70 Gy/35 fx or 66 Gy/33 fx matched carbon on 5-yr PFS (68.3% vs 61.1%) with zero grade ≥3 radionecrosis, versus 9.1% in the carbon subgroup. Boost PTV margins were 0.5 to 2 mm off cavity and residual tumor. That makes escalation a standard photon department decision, not a particle referral.
14 details
Individual patient-level meta-analysis pooling 248 patients from 7 international institutions, all retrospective except one prospective trial. Median follow-up 67.1 months (range 2.13-178). PFS was the primary outcome, analyzed by Kaplan-Meier, Cox MVA, and IPTW propensity weighting.
High-risk meningioma per RTOG 0539: newly diagnosed WHO grade 2 after STR/biopsy, any recurrent grade 2, or any grade 3. 188 (75.8%) were grade 2, 103 (41.5%) recurrent, and 182 (75.2%) had Simpson grade 4/STR. Median age 60; 55 (22.2%) had prior RT.
DE-RT was defined by biologically effective dose ≥79.2 Gy (equivalent to 66 Gy in 33 fractions) or receipt of a carbon-ion boost; anything below that threshold counted as SD-RT. Photon DE-RT used a sequential or simultaneous integrated boost with a 0.5 to 2 mm PTV margin, versus 1-2 cm CTV margins in the SD-RT arm. Carbon DE-RT boosted 16 Gy/8 fx after 50 Gy/25 fx photons with a 6 mm CTV margin.
Whole-cohort 3- and 5-yr PFS were 62.8% and 45.0%. DE-RT improved 3-yr (86.4% vs 55.6%) and 5-yr PFS (65.8% vs 38.8%), P = .0022, holding on stratified Cox by grade (HR 0.40, 0.22-0.73) and after IPTW (3-yr 84.7% vs 55.8%). OS was not improved on MVA or IPTW.
| Subgroup | 3-yr PFS DE-RT vs SD-RT | 5-yr PFS DE-RT vs SD-RT | MVA HR (95% CI) | P |
|---|---|---|---|---|
| Simpson 1-3 | 87.5% vs 55.9% | 70.0% vs 40.0% | 0.31 (0.08-1.14) | .08 |
| Simpson 4-5 | 86.3% vs 55.4% | 63.3% vs 38.4% | 0.55 (0.36-0.84) | .006 |
| Cohort | Any-grade RN | Grade ≥3 RN |
|---|---|---|
| DE-RT overall | 20 of 59 (33.9%) | 3 of 59 (5.1%) |
| DE-RT mixed carbon/photon | 15 of 33 (45%) | 3 of 33 (9.1%) |
| DE-RT photon-alone | 5 of 26 (19.2%) | 0% |
| SD-RT | 25 of 189 (13.2%) | 6 of 189 (3.2%) |
Any-grade radionecrosis was 33.9% with DE-RT vs 13.2% with SD-RT (P = .001), but grade ≥3 RN did not differ (5.1% vs 3.2%, P = .47). All 3 grade ≥3 events in the DE-RT arm occurred in the mixed carbon-photon cohort; the photon-alone DE-RT cohort had none, and its any-grade RN rate was not significantly higher than SD-RT (P = .41). One grade 5 event occurred in each arm.
Published series anchored on RTOG 0539 report 5-yr PFS of roughly 40% to 60% with 60 Gy/30 fx, which is what the SD-RT arm reproduces (45.0% overall). The escalation signal draws on MARCIE (carbon boost) and the Zeng et al. photon series, both contributors to this pool, so the comparison is partly internal rather than independent.
DE-RT was delivered at only 2 of 7 institutions, one carbon-only and one photon-only, so treatment arm is nearly collinear with center, and unmeasured practice differences (DOTATATE PET planning, response assessment, supportive care) travel with it. Molecular classification was unavailable, and P values were not adjusted for multiple testing across the subgroup analyses.
The Simpson 1-3 subgroup showed absolute separation as large as Simpson 4-5 (5-yr PFS 70.0% vs 40.0%) without reaching significance (P = .08 on MVA), which the authors attribute to a smaller GTR subset and fewer events rather than an absent effect. Whether escalation belongs in fully resected disease is the open question, and the PFS-only benefit means the case rests on avoiding local progression and its neurologic morbidity, not on survival.
Retrospective IPD pooling, DE-RT confined to 2 centers, no randomization; IPTW cannot remove selection. Direction consistent across every sensitivity analysis, but prospective randomization still needed.
- Benefit of dose escalation after gross total resection
- Whether molecular subgroups predict DE-RT benefit
- Prospective randomized confirmation of the PFS signal
📚 Sources · 📄 1 paper
Abstract
NRG/RTOG 9804 + E5194 combined analysis (tamoxifen after lumpectomy, DCIS)
ForGood-risk DCIS, lumpectomy without RT, low/int grade, ≤2.5 cm, margins ≥3 mm
TL;DR15-yr IBR 11.4% vs 19.0% with vs without tamoxifen in good-risk DCIS treated by lumpectomy without RT; invasive IBR HR 0.43.
The association was confined to invasive events (DCIS-IBR p=0.089), which made up 65 of 117 IBRs. For the RT-omission discussion, the 11.4% 15-yr IBR on tamoxifen has no RT arm in this analysis to benchmark against, and which good-risk subgroups drive it sits in Fig 3, with subgroup HRs not in source text.
In low/intermediate-grade DCIS ≤2.5 cm with margins ≥3 mm managed without RT after lumpectomy, this supports tamoxifen as an invasive-IBR risk-reduction option in the omission discussion; it does not extend to high-grade DCIS or pts receiving RT. The longer read
For RT omission in good-risk DCIS, tamoxifen alone still left 15-yr IBR at 11.4%, and this analysis has no RT arm to benchmark it against. The reduction tracked invasive IBR (HR 0.43), while DCIS-IBR did not separate (p=0.089).
In good-risk DCIS managed without RT, tamoxifen associated with lower invasive IBR (HR 0.43, 0.24-0.77) but not DCIS-IBR (p=0.089). Dose, duration and adherence are not reported, and contralateral breast event results, the other half of the endocrine rationale, are not in source text.
9 details 1 trial watching
Ancillary exploratory analysis combining the non-RT arm of NRG/RTOG 9804, which randomized good-risk DCIS to RT or no RT, with the good-risk E5194 cohort observed without RT. Tamoxifen was optional and not randomized in either trial. Univariate and multivariable Fine-Gray models; median follow-up 14.85 yr.
N=878 (317 RTOG 9804, 561 E5194); median age 59 (28-88). Good-risk: low- or intermediate-grade, ≤2.5 cm, excision margins ≥3 mm, lumpectomy without RT.
Tamoxifen, optional per trial; used by 43.1% overall (65.6% RTOG 9804 vs 30.3% E5194). Dose, duration and adherence not reported in source.
No RT in the analyzed population. The RTOG 9804 RT arm is not part of this analysis, so tamoxifen alone is not compared against RT here.
IBR (any), invasive IBR, DCIS-IBR, contralateral breast event and overall survival. CBE and OS results are not reported in the source text.
117 IBRs (65 invasive, 52 DCIS). The reduction tracked invasive IBR, not DCIS-IBR.
| Endpoint | Tamoxifen | No tamoxifen | Effect |
|---|---|---|---|
| Any IBR, 15-yr estimate | 11.4% (7.9-15.5) | 19.0% (15.3-22.9) | p=0.001 |
| Any IBR, multivariable | n/a | n/a | HR 0.54 (0.35-0.83), p=0.0045 |
| Invasive IBR | not reported | not reported | p=0.0048; MVA HR 0.43 (0.24-0.77), p=0.0042 |
| DCIS-IBR | not reported | not reported | p=0.089 |
| Characteristic | Tamoxifen | No tamoxifen | p |
|---|---|---|---|
| Enrolled on NRG/RTOG 9804 | 55.0% | 21.8% | <0.0001 |
| Non-white | 13.8% | 9.2% | 0.034 |
| Negative re-excision | 27.2% | 10.6% | <0.0001 |
| Pathologic size ≤5 mm | 57.0% | 41.3% | <0.0001 |
NSABP B-24 randomized tamoxifen after lumpectomy plus RT, and the UK/ANZ DCIS trial randomized tamoxifen and RT in a factorial design. Here tamoxifen is examined only in good-risk pts managed without RT, and without randomization.
Multivariable covariates are not listed in source, and trial membership also differs in follow-up (13.87 vs 16.15 yr). The DCIS-IBR result (p=0.089) rests on 52 events and does not exclude an effect.
Tamoxifen use tracked lower invasive recurrence in good-risk DCIS managed without RT, consistent in direction with randomized tamoxifen data. It does not settle whether tamoxifen alone approaches RT, or how much of the gap reflects the lower-risk profile of users.
Exploratory ancillary analysis; tamoxifen optional, not randomized, and imbalanced by trial, re-excision and size. Parent RCT tested RT, not tamoxifen, so residual confounding limits causal read.
- Does tamoxifen alone approximate RT for IBR control in good-risk DCIS? recruiting Omission of Adjuvant Radiotherapy in Low- and Intermediate-Risk Ductal Carcinoma In Situ After Breast-Conserving Surgery Phase 3n=636 · primary completion 2031-06 · ph3 NI RT omission in ER+ low/int-grade DCIS, tamoxifen
- Which good-risk DCIS subgroups gain enough to justify tamoxifen?
- Is the lack of DCIS-IBR reduction real or underpowered?
📚 Sources · 📄 1 paper
Abstract
Single-fraction SABR pooled analysis, 1687 pts
ForPrimary NSCLC or pulmonary oligomets selected for single-fraction SABR
TL;DRLocal control 90-93% at 2yr and G3+ AEs 2.9% across 1687 single-fraction SABR pts at 3 centres.
The oligomet read is the gap between local control and PFS: 90-93% LC at 2yr against median PFS 11 mo, so distant failure, not the treated lesion, drives the course. For primary NSCLC the same LC sits with median PFS 30 mo, which is the split that should decide whether one-visit ablation is offered as definitive treatment or as a break from systemic therapy.
In early-stage primary NSCLC where visit burden drives the fractionation choice, this supports single fraction as a durable local option (LC 90-93% at 2yr, G3+ 2.9%); tumour location and operability are not reported, so who it represents stays open. The longer read
Local control holds at 90-93% at 2yr for a primary and for a metastasis alike, so the single fraction is not the variable separating outcomes; median PFS is (30 vs 11 mo). Tumour size and location go unreported, so the target selection behind that number is unmeasured.
For a pt with pulmonary oligometastases, one-visit ablation gave 90-93% local control at 2yr but median PFS of 11 mo, so it clears the treated lesion without changing the systemic course. That frames referral as a local step between systemic lines, not a substitute for one.
| Cohort | n | Median PFS | Median OS |
|---|---|---|---|
| Primary NSCLC | 1200 | 30 mo | 3.5 yrs |
| Pulmonary oligometastases | 487 | 11 mo | >4 yrs |
+2 more figures
| Endpoint | Primary NSCLC | Oligometastases |
|---|---|---|
| 1yr OS | 84% (95% CI 82, 86) | 90% (95% CI 86, 92) |
| 2yr OS | 67% (95% CI 64, 69) | 75% (95% CI 71, 79) |
| Median OS | 40 mo (36, 43) | 51 mo (42, 58) |
| Adverse event (n=789) | n (%) |
|---|---|
| Any AE | 215 (27%) |
| Grade 2+ | 124 (15.7%) |
| Grade 3+ | 23 (2.9%) |
| Chest wall pain | 114 (14%) |
| Pneumonitis | 52 (7%) |
| Fatigue | 29 (4%) |
| Dyspnea | 13 (2%) |
6 details
Pooled analysis of 1687 pts treated with single-fraction SABR at three centres (Peter MacCallum, Cleveland Clinic, Roswell Park): 1200 primary NSCLC and 487 pulmonary oligometastases. Whether the contributing cohorts were prospective or retrospective is not stated in source.
Eligibility, operability, tumour size and central vs peripheral location are not reported in source. Cohort mix differs sharply by centre: Roswell Park supplied 401 of the NSCLC pts but only 34 oligomet pts, while Peter Mac supplied 283 of 487 oligomet pts.
Single fraction throughout, but the prescribed dose is not reported in source. Without it the outcome cannot be mapped onto a schedule a reader could write, which is the one parameter that would carry this into planning.
No primary endpoint is stated in the source. Reported outcomes are local control, freedom from local failure, PFS, OS and adverse events, each descriptive rather than tested against a comparator.
Local control 90-93% at 2 years across both cohorts, with isolated local or locoregional failure described as very uncommon. Survival separates by cohort while local outcome does not.
| Centre | Primary NSCLC | Pulmonary oligomets |
|---|---|---|
| Cleveland Clinic | 576 | 170 |
| Peter MacCallum | 223 | 283 |
| Roswell Park | 401 | 34 |
AE reporting covers 789 primary NSCLC pts only, with no Roswell Park data and no oligometastasis toxicity in source. Within that subset chest wall pain and pneumonitis dominate and G3+ events stay at 2.9%.
Single-fraction SABR already carries randomised support: RTOG 0915 in peripheral early-stage NSCLC and SAFRON II in pulmonary oligometastases, both randomised single against multi-fraction schedules. This series adds scale and follow-up at three high-volume centres, which is what a non-randomised dataset can contribute, and no comparator.
Toxicity rests on 789 of 1687 pts, with one centre absent from the AE table and the oligometastatic cohort not represented in it at all. Centre mix is uneven, so pooled rates carry each centre's own selection rather than a common one.
The question the thread raises, whether one-stop SABR should be used more often, is not the question this dataset answers. What it does show is that local control near 90-93% and G3+ toxicity near 3% hold at scale outside a protocol, which is the usual worry about a schedule with no second chance. The unreported dose sits between that reassurance and a prescription.
Pooled uncontrolled series across three centres, no multi-fraction comparator and no stated design; dose unreported, so outcomes cannot be tied to a prescription.
- Whether single-fraction outcomes hold for central tumours
- Durability of single-fraction ablation for pulmonary oligometastases beyond first progression
- Toxicity of single-fraction SABR in the oligometastatic cohort
📚 Sources · 🐦 1 tweet
👏🏽👏🏽👏🏽@neildwallaceie at #ESTRO26 - 1687 patients receiving single fraction SABR for #lungcancer and pulmonary oligomets, @PeterMacRadOnc / @ClevelandClinic / @RoswellPark. Fantastic local control, and low adverse rates. Should we be using “one stop” SABR more often #radonc ? pic.twitter.com/w2IlGKRU5o
— Shankar Siva (@_ShankarSiva) May 18, 2026
Multinational HCC EBRT IPD Cohort
ForVery early / early-stage HCC (BCLC-0 or A), incl. treatment-naive
TL;DRMedian OS 6.8y BCLC-0 and 4.6y BCLC-A across 4,913 EBRT-treated HCC pts, comparable to resection and ablation.
The modifiable RT variable in the Cox model is ablative dose, which was associated with reduced mortality, so the transferable read is that dose, not simply delivering EBRT, tracks with the outcome. Fractionation and dose thresholds are not given in the abstract. This is the citation for putting EBRT on the BCLC-0/A allocation discussion.
In BCLC-0 or A HCC where resection, transplant, or ablation is not feasible or is declined, this supports discussing ablative-dose EBRT as a locoregional option; it does not establish EBRT over resection or ablation in a pt eligible for either. The longer read
Ablative dose was the modifiable variable associated with reduced mortality, so the transferable read is dose, not simply offering EBRT. Dose thresholds, fractionation, and modality mix are not given in the source abstract. This is the citation for putting EBRT into the BCLC-0/A allocation discussion at tumor board.
Child-Pugh B or C, performance status, and tumor burden drove mortality in the multivariable model, which is the gating frame for who gets locoregional therapy at all. For the med onc coordinating a BCLC-0/A pt, EBRT enters the locoregional menu alongside ablation rather than as a last resort.
Median OS of 6.8 y (BCLC-0) and 4.6 y (BCLC-A) in EBRT-treated pts is a cross-literature benchmark against resection and ablation, not a head-to-head. It bears on referral for the medically inoperable or anatomically difficult pt, and does not establish EBRT over resection in a fully resectable candidate.
9 details 2 trials watching
Systematic review of EBRT publications meeting prespecified HCC technical standards (search date December 15, 2022), with corresponding authors invited to contribute individual patient data. Kaplan-Meier OS and RMST stratified by BCLC stage and treatment status; random-effects Cox for covariates. No comparator arm.
4,913 pts treated with EBRT, median follow-up 5.0 years, multinational. Analyses split by BCLC stage and by treatment-naive vs treatment-experienced; the headline read sits in BCLC-0 and BCLC-A.
EBRT delivered per each contributing series, gated by the prespecified technical standards rather than one protocol. Ablative dose was associated with a reduced risk of death; specific dose levels, fractionation, and modality mix are not given in the source abstract.
Overall survival by Kaplan-Meier and restricted mean survival time, stratified by BCLC stage and treatment status. Covariate associations from multivariable random-effects Cox modeling.
Median OS 6.8 y (95% CI 5.7-8.7) for BCLC-0 and 4.6 y (95% CI 4.1-5.1) for BCLC-A. Treatment-naive: not reached (95% CI 8.6-NR) for BCLC-0, 5.4 y (95% CI 4.5-6.7) for BCLC-A.
| Cohort | BCLC-0 | BCLC-A |
|---|---|---|
| All pts | 6.8 y (95% CI 5.7-8.7) | 4.6 y (95% CI 4.1-5.1) |
| Treatment-naive | NR (95% CI 8.6-NR) | 5.4 y (95% CI 4.5-6.7) |
The authors frame these medians as comparable with resection, thermal ablation, and other ablative locoregional therapies, a cross-study benchmark rather than a randomised comparison. EBRT's exclusion from BCLC has rested on the absence of OS evidence, which is the gap this cohort is built to fill.
IPD came only from authors who published and agreed to share, so contributing centers are self-selected and unmeasured selection at the patient level (who was routed to EBRT rather than resection) is unrecoverable. More recent year of treatment predicting survival mixes technique gains with stage migration and modern systemic salvage over a multi-decade accrual window.
The claim is an allocation claim, not an efficacy claim: EBRT belongs in the BCLC decision tree as an option to be weighed. It does not settle sequencing against ablation in a pt eligible for both, and the ablative-dose signal makes the quality of the RT, not its mere availability, the operative variable.
Largest EBRT IPD cohort argues for a BCLC allocation change, but it is pooled non-randomised data with no head-to-head comparator against resection or ablation.
- Ablative dose threshold and fractionation driving the survival association
- EBRT vs thermal ablation head-to-head in early HCC recruiting Stereotactic Radiosurgery Versus Radiofrequency Ablation for Primary Liver Cancer Phase 2n=130 · primary completion 2025-01 · phase 2 SBRT vs RFA, inoperable primary liver caactive Stereotactic Body Radiotherapy Versus Radiofrequency Ablation for Unresectable, Small (≤ 3 cm) HCC Phase NAn=178 · primary completion 2026-12 · randomised SBRT vs RFA, unresectable HCC <=3cm
- Selection differences between EBRT-treated and resected early-stage pts