2026-07-01

digest generated 2026-09-15

RTOG 1112: SBRT + sorafenib mOS 15.8 vs 12.3 mo, HR 0.77 (90% CI 0.59-1.01), not significant; mPFS HR 0.55 in locally advanced HCC.

TL;DR

RT was the question at both sites. RTOG 1112 adds SBRT to sorafenib in locally advanced HCC: OS missed its unadjusted test (HR 0.77, 1-sided P=.06), mPFS HR 0.55; accrual stopped early after systemic SOC changed. ReCOG HNSCC reirradiation consensus hit ≥85% agreement on 24/31 statements (no ENI, definitive IMRT ~66 Gy), none backed by randomised data.

Head & Neck

ReCOG ties curative reRT to Ward RPA class and to cumulative OAR EQD2 limits (carotid 120-125 Gy, mandible 114-119 Gy). Agreement was weakest on the cord and brainstem limits.

ReCOG HNSCC Reirradiation Consensus

TL;DR24 of 31 (77%) statements reached ≥85% agreement: no ENI, definitive IMRT ~66 Gy, postop ~60 Gy, SBRT ~40 Gy in five fractions.

Trials discussed

RTOG 9610RTOG 9911GORTEC 2008–01GORTEC 98–03GORTEC-GETTEC

Why it mattersRadiation oncology

Subsite, not only volume, gates SBRT: in Vargo's 291-pt series severe toxicity reached up to 50% for laryngeal or hypopharyngeal recurrences vs 6-20% at other subsites, which is the basis for the 94% caution statement. Postop SBRT data are thin (n=28), which is why IMRT is preferred after salvage surgery.

Monday clinic

For a recurrence 6-12 months after prior RT, this supports reirradiation only when disease is small, localised, and outside high-dose regions after MDT review; it does not extend to recurrence within 6 months outside urgent palliation, or to nasopharyngeal primaries. The longer read

8 details 1 trial watching

Design

International expert consensus: a ReCOG core group drafted statements, then a 17-member radiation oncology panel voted once on 31 statements. Thresholds were high ≥85%, moderate 70-84%, low <70%, with no-opinion votes counted in the denominator. 24 (77%) reached high consensus.

Population & inclusion

Recurrent or second primary HNSCC in a previously irradiated field, covering definitive and postoperative reirradiation. Nasopharyngeal cancer and rare histologies excluded. Evidence levels adapted from Oxford CEBM, level 1 (randomised) to level 4 (expert consensus).

Radiotherapy

Definitive: CTV = GTV + 5 mm, no ENI, IMRT ~66 Gy for >25 cc, IMRT or SBRT for ≤25 cc, SBRT ~40 Gy in five fractions. Postoperative: primary-RT CTV rules, IMRT preferred, ~60 Gy. Cumulative doses in EQD2 with α/β 2-3 Gy, voxel-wise summation with DIR when available, with conservative manual point-dose addition acceptable otherwise.

Safety

Carotid: avoid cumulative EQD2 >120-125 Gy (α/β=2); encasement >180° and prior ulceration add risk. Mandible: ORN risk rises above 114-119 Gy EQD2. Cord/brainstem: D0.1 cc ≤50 and ≤55 Gy as targets, ~60 Gy at the surface case by case, recovery factors (10-25%) only with caution.

vs leading data

Split-course chemo-reRT in RTOG 9610 and RTOG 9911 gave 2-year OS of 15.2% and 25.9%, with grade 3-4 toxicity up to 34% and treatment-related mortality around 8%. The IMRT-era Ward cohort (n=412) reported median OS 16.5 mo with fatal toxicity in 1%. GORTEC-GETTEC (n=130), the randomised postop trial, improved locoregional control and RFS with no OS benefit.

Applies to

recurrent or second primary non-nasopharyngeal HNSCC in an irradiated field treated with definitive or postoperative reirradiation

Does not represent nasopharyngeal cancer, rare histologies, or recurrence within 6 months outside urgent palliation.

Limitations

Guidance for RPA class II and III rests largely on one retrospective RPA model, since most reirradiation outcome data come from RPA class I pts. The IMRT vs SBRT split and the dose statements come from non-randomised multi-institutional cohorts. The 10-25% recovery factor has very low evidence.

Discussion

The seven moderate-consensus statements cluster on poor-prognosis subgroups, dose and fractionation, and SBRT, and the lowest agreement (70%) falls on cord and brainstem cumulative limits. The authors present dose thresholds and OAR constraints as pragmatic targets rather than validated limits. Sequencing reirradiation against first-line ICI has no randomised data.

Results

Statement	Consensus	Evidence level
Review prior RT plan: in-field vs marginal failure	High (100%)	3
Recurrence <6 mo: generally not advised, except urgent palliation	High (100%)	2
Recurrence 6-12 mo: only small, localised, outside high-dose regions, after MDT	High (88%)	2
RPA I with positive margins or ECE: consider postop reRT	High (100%)	2
R0 unlikely even without clearly positive margins: reRT can be considered	Moderate (82%)	4
RPA III: generally no curative-intent reRT	Moderate (76%)	3

Statement	Consensus	Evidence level
Postop: normofractionated IMRT preferred over SBRT	High (94%)	3
RPA II unresected ≤25 cc (cT1-T2): IMRT or SBRT similar	High (88%)	3
RPA II unresected >25 cc: IMRT over SBRT	High (88%)	3
SBRT with caution for unresected larynx or hypopharynx	High (94%)	3
Concomitant chemo case by case, large tumours in fit pts	High (88%)	3
Protons in selected cases to spare OARs	High (88%)	3
Brachy for <25 cc oral cavity or oropharynx, expert centres	High (88%)	3

Statement	Consensus	Evidence level
Definitive: CTV = GTV + 5 mm	High (94%)	4
Postop high-risk CTV: same rules as primary RT	Moderate (82%)	4
No elective nodal irradiation	High (100%)	3
Definitive IMRT: 66 Gy commonly used	Moderate (82%)	4
Hyperfractionated IMRT: an option, no clear benefit	Moderate (76%)	3
Postop: 60 Gy commonly used	High (94%)	3
SBRT: ~40 Gy in five fractions	Moderate (76%)	3

Statement	Consensus	Evidence level
Carotid: avoid cumulative EQD2 >120-125 Gy (α/β=2)	High (94%)	3
Carotid encasement >180° or prior ulceration raises blowout risk	High (100%)	3
Mandible: ORN risk rises above cumulative EQD2 114-119 Gy	High (100%)	3
Cord/brainstem recovery factors (10-25%): caution, document	High (88%)	4
Target cumulative EQD2 D0.1 cc ≤50 Gy cord, ≤55 Gy brainstem	Moderate (70%)	4
~60 Gy at cord or brainstem surface, case by case	High (94%)	4

RPA class	Definition	n	2-yr OS
I	Recurrence >2 yr after RT, salvage surgery	91	61.9%
II	Unresected >2 yr, or <2 yr without organ dysfunction	230	40.0%
III	<2 yr plus feeding tube or tracheostomy dependence	81	16.8%

Dose	2-yr OS
≥66.0 Gy	49.3%
60.0-65.9 Gy	34.2%
<60.0 Gy	30.4%

Open questions

Sequencing of reirradiation vs first-line immune checkpoint inhibitors

recruiting Re-Radiochemotherapy and Pembrolizumab vs. Immuno(Chemo)Therapy for Locoregionally Recurrent PD-L1 Positive (CPS≥1) HNSCC Phase 3

n=214 · primary completion 2032-12 · phase 3 re-RCT then pembro vs pembro ± chemo, OS
Proton therapy vs IMRT outcomes in reirradiation
Validated cumulative cord and brainstem constraints and recovery factors

Sourced from Julian Biau et al.

📚 Sources · 📄 1 paper

📄 PAPER Julian Biau; Arnaud Beddok; Manju Sharma et al. · The Lancet Oncology (2026-07)

Reirradiation for recurrent head and neck squamous cell carcinoma: international expert consensus recommendations endorsed by the Reirradiation Collaborative Group, the European Society for Radiotherapy and Oncology Reirradiation Focus Group, and the American Society for Radiation Oncology

Radiation Curative Consensus / guideline

Hepatobiliary

Randomised phase 3 testing what SBRT adds to sorafenib in HCC unsuitable for local-regional therapy, 74% with macrovascular invasion; sorafenib is no longer the systemic SOC.

Caveats dominate

NRG Oncology/RTOG 1112 NCT01730937

ForLocally advanced HCC, 74.0% MVI, local-regional tx unsuitable/refractory

Overall survival

HR 0.77

90% CI 0.59-1.01, 1-sided P=.06, not statistically significant

TL;DRmOS 15.8 vs 12.3 mo with SBRT + sorafenib vs sorafenib, HR 0.77 (90% CI 0.59-1.01), not significant; mPFS HR 0.55 in locally advanced HCC.

Why it mattersRadiation oncology

PFS carries the RT signal: HR 0.55 (95% CI 0.40-0.75) with personalized 27.5-50 Gy in 5 fractions, in a cohort with 74.0% macrovascular invasion and no rise in G3+ TRAE (47% vs 42%). In-liver control and failure patterns are not reported in source text, so the case for liver SBRT rests on PFS plus an adjusted OS HR.

Monday clinic

In locally advanced HCC with macrovascular invasion, unsuitable for or refractory to local-regional therapy, this supports weighing liver SBRT with first-line systemic therapy for progression control; it does not extend to pts on the newer systemic standard that ended accrual. The longer read

8 details 2 trials watching

Design

Phase 3 randomized, multicenter, 1:1, stratified by performance status, liver function, degree of metastases and macrovascular invasion. 193 randomized, 177 eligible; data collected April 2013 to March 2021. Accrual stopped early after a change in standard-of-care systemic therapy.

Population & inclusion

HCC unsuitable for or refractory to standard local-regional therapies, candidates for first-line systemic therapy. Median age 66 (IQR 60-72), 84.7% male, macrovascular invasion 74.0%.

Regimen

Sorafenib in both arms; the experimental arm received SBRT followed by sorafenib.

Radiotherapy

Personalized SBRT, 27.5 to 50 Gy in 5 fractions. Dose-selection rules, target volumes and OAR constraints not reported in source text.

Endpoints

Primary: overall survival. Secondary: PFS, adverse events, quality of life.

Results

Primary OS not statistically significant on the unadjusted test; significant after stratification-factor adjustment. PFS clearly improved with SBRT; values in the table above.

Endpoint	SBRT + sorafenib	Sorafenib	Effect size
Median OS (1° EP)	15.8 mo (90% CI 11.4-19.2)	12.3 mo (90% CI 10.6-14.3)	HR 0.77 (90% CI 0.59-1.01), 1-sided P=.06
OS, adjusted for stratification factors	n/a	n/a	HR 0.72 (95% CI 0.52-0.99), 2-sided P=.04
Median PFS	9.2 mo (95% CI 7.5-11.9)	5.5 mo (95% CI 3.4-6.3)	HR 0.55 (95% CI 0.40-0.75), P<.001
G3+ treatment-related AEs	39/83 (47%)	37/88 (42%)	P=.52
Improved QoL at 6 mo	6/17 (35%)	2/20 (10%)	not reported

Safety

G3+ treatment-related AEs not increased with SBRT. Treatment-related deaths: 2 with sorafenib (death NOS, liver failure), 1 with SBRT plus sorafenib (lung infection).

vs leading data

Radioembolization trials against or combined with sorafenib (SARAH, SIRveNIB, SORAMIC) did not show an OS gain. START, which tested TACE plus RT against sorafenib in macrovascular invasion, favored the radiation-containing arm, and RTOG 1112 points the same direction with stereotactic delivery.

Applies to

locally advanced HCC unsuitable for or refractory to local-regional therapy, 74.0% with macrovascular invasion, starting first-line sorafenib

Does not represent pts eligible for standard local-regional therapy or on newer first-line systemic regimens.

Limitations

Analysis restricted to 177 eligible of 193 randomized. Local control and in-liver failure patterns not reported in source text, and personalized 27.5-50 Gy dosing confounds any dose-response read. QoL rests on 17 and 20 evaluable pts.

Discussion

The authors call the OS gain clinically important but not statistically significant; adjusted OS and PFS point the same way, so underpowering after early closure is a more plausible reason for the miss than absence of effect. It does not settle whether SBRT adds to the systemic standard that replaced sorafenib.

Rationale

Primary OS missed after accrual stopped early; positive OS read depends on stratification-adjusted analysis, with PFS the only unadjusted win.

Open questions

Does SBRT add OS benefit over current first-line systemic therapy?

recruiting Testing Immunotherapy With or Without Stereotactic Body Radiation Therapy in Patients With Advanced Liver Cancer, HELIO-RT Trial Phase 3

n=226 · primary completion 2029-03 · phase 3 IO-based 1L ± SBRT, advanced HCC

not yet SBRT Combined with Camrelizumab and Apatinib for Conversion Therapy in Patients with Unresectable Hepatocellular Carcinoma. Phase 2/3

n=398 · primary completion 2030-12 · randomised 1L camrelizumab/apatinib ± SBRT
Local control and in-liver failure patterns after SBRT
Optimal dose within the personalized 27.5-50 Gy range

Sourced from Dawson et al.

📚 Sources · 📄 1 paper

📄 PAPER Dawson; Winter; Knox et al. · JAMA oncology (2025-02)

Stereotactic Body Radiotherapy vs Sorafenib Alone in Hepatocellular Carcinoma: The NRG Oncology/RTOG 1112 Phase 3 Randomized Clinical Trial.

PMID 39699905 → doi full text (PMC)

Abstract

IMPORTANCE: Most patients with locally advanced hepatocellular carcinoma (HCC) recur within the liver following systemic therapy. OBJECTIVE: To determine whether stereotactic body radiation therapy (SBRT) improves outcomes in patients with locally advanced HCC compared with sorafenib alone. DESIGN, SETTING, AND PARTICIPANTS: This multicenter phase 3 randomized clinical trial randomized patients with HCC 1:1 to sorafenib or SBRT followed by sorafenib, stratified by performance status, liver function, degree of metastases, and macrovascular invasion. Eligible patients had HCC unsuitable for or refractory to standard local-regional therapies and were candidates for first-line systemic therapy. Data were collected from April 2013 to March 2021, and data were analyzed from July 2022 to August 2023. INTERVENTION: Personalized SBRT, 27.5 to 50 Gy in 5 fractions. MAIN OUTCOMES AND MEASURES: The primary end point was overall survival (OS). Secondary end points were progression-free survival (PFS), adverse events, and quality of life. RESULTS: Of 193 patients randomized, 177 were eligible. Accrual was stopped early due to a change in standard-of-care systemic therapy. Of 177 included patients, 150 (84.7%) were male, and the median (IQR) age was 66 (60-72) years. Macrovascular invasion was seen in 131 (74.0%). As of July 1, 2022, the median OS was 12.3 months (90% CI, 10.6-14.3) with sorafenib vs 15.8 months (90% CI, 11.4-19.2) following SBRT and sorafenib (hazard ratio [HR], 0.77; 90% CI, 0.59-1.01; 1-sided P = .06). Adjusting for stratification factors, OS was improved with SBRT (HR, 0.72; 95% CI, 0.52-0.99; 2-sided P = .04). Median PFS was improved from 5.5 months (95% CI, 3.4-6.3) with sorafenib to 9.2 months (95% CI, 7.5-11.9) with SBRT and sorafenib (HR, 0.55; 95% CI, 0.40-0.75; 2-sided P < .001). Treatment-related grade 3 or higher adverse events were seen in 37 of 88 (42%) and 39 of 83 (47%) of patients treated with sorafenib vs SBRT and sorafenib, respectively (P = .52). There were 2 treatment-related deaths in the sorafenib group (death not otherwise specified and liver failure) and 1 in the SBRT and sorafenib group (lung infection). At 6 months, improved quality of life was seen in 2 of 20 (10%) and 6 of 17 (35%) of patients treated with sorafenib and SBRT and sorafenib, respectively. CONCLUSIONS AND RELEVANCE: In this phase 3 randomized clinical trial, among patients with locally advanced HCC, SBRT was associated with a clinically important but not statistically significant improved overall survival compared with sorafenib alone. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT01730937.

📝 https://jamanetwork.com/journals/jamaoncology/fullarticle/2827892

Radiation Palliative Phase 3 RCT Caveats dominate