SIB-CRT vs Standard CRT for LARC (NCT02195141) NCT02195141
ForStage II/III rectal adenocarcinoma, neoadjuvant chemoRT candidates
15.2% vs 18.4%
P=0.695, primary endpoint not met
TL;DR9yr LC 87.1% vs 70.1% (HR 0.40, P=0.038) and OS 74.3% vs 48.9% (HR 0.43) favoring SIB dose escalation.
The boost was 56 Gy to PGTV and 60 Gy to involved lateral nodes on a 50 Gy/25 fx pelvic base, a deliverable prescription with acute G3 toxicity 14.5% vs 19.6%. LC 87.1% vs 70.1% is the mechanistically coherent signal; the OS and MFS separation on 106 pts is not, and gates any move to boost outside a trial.
In stage II/III LARC where perioperative chemotherapy is not planned or not tolerated, this supports testing an integrated boost to gross disease and involved lateral nodes; it gives no read on pts receiving modern TNT, where the subgroup showed no added benefit.
The prescription is deliverable and specified: 56 Gy to PGTV, 60 Gy to involved lateral nodes, on a 50 Gy/25 fx pelvic base, with acute G3 toxicity 14.5% vs 19.6% and no late toxicity reported. LC 87.1% vs 70.1% is the endpoint a boost can own; treat the OS and MFS separation as hypothesis-generating for a boost trial, not license to escalate off-protocol.
The benefit was confined to pts who did not receive perioperative chemotherapy (9yr DFS 70.8%, HR 0.343, P=.014), with no additional benefit in those who did, and chemo receipt was not randomized. Read as a signal about RT dose in chemotherapy-ineligible pts rather than any argument against systemic intensification.
9 details 5 trials watching
Prospective randomized phase 2, 1:1, N=106 (55 SIB-CRT, 51 CRT), accrued August 2013 to February 2015. Median follow-up 116.6 months, which is the study's real asset.
Stage II/III rectal adenocarcinoma. Radical surgery planned 6 to 8 weeks after chemoradiotherapy. Perioperative chemotherapy was not randomized, and its receipt defines the subgroup that carries the result.
Both arms received 50 Gy/25 fx to the pelvis. The experimental arm added a simultaneous integrated boost of 56 Gy to PGTV and 60 Gy to lateral metastatic nodes where present, so the escalation targets gross primary and involved lateral nodes rather than the elective volume.
Primary: pCR rate. Secondary: DFS, OS, MFS, LC, CSS and toxicity. Every result the conclusion rests on is secondary.
Acute grade 3 toxicity 14.5% (SIB-CRT) vs 19.6% (CRT), primarily radiation dermatitis. Late toxicity is not reported in the source, which matters most for a boost delivered near bowel and for pts followed nearly a decade.
Neoadjuvant dose escalation in LARC has repeatedly raised pCR without translating to survival; this trial reports the mirror image, a null pCR (15.2% vs 18.4%) with survival separation. The result also sits against the TNT era, where systemic intensification rather than RT dose is the current lever, and the chemo-receiving subgroup here showed no additional benefit.
The survival claim is a secondary-endpoint result from 51 vs 55 pts, so a handful of events moves each HR, and no confidence intervals are reported in the source. The driving subgroup is defined by non-randomized chemotherapy receipt, and the higher rate of curative treatment in the SIB arm (89.1% vs 78.4%) is itself a plausible route to the OS difference independent of dose.
A pelvic boost has a defensible mechanism for LC 87.1% vs 70.1% and essentially none for MFS 70.8% vs 47.2%. The authors attribute the distant benefit to intensified control of micrometastases; the simpler reading is that a small trial with a large curative-treatment imbalance produced correlated secondary endpoints.
| Endpoint | SIB-CRT | CRT | Effect |
|---|---|---|---|
| DFS | 70.8% | 47.4% | HR 0.46, P=0.013 |
| OS | 74.3% | 48.9% | HR 0.43, P=0.008 |
| MFS | 70.8% | 47.2% | HR 0.48, P=0.017 |
| LC | 87.1% | 70.1% | HR 0.40, P=0.038 |
| CSS | 77.4% | 57.2% | P=0.027 |
| pCR | 15.2% | 18.4% | P=0.695 |
CONSORT flow
Randomized phase 2, N=106, primary endpoint (pCR) missed; survival gains are secondary endpoints with wide implied precision and a non-randomized chemo subgroup driving them.
- Does SIB add anything on a total neoadjuvant therapy backbone n=37 · primary completion 2026-05 · SIB-SCRT plus CAPOX and PD-1 in high-risk LARCn=156 · primary completion 2026-12 · randomises GTV SIB 58.75 vs 50 Gy/25f for CRrecruiting THeragnostic Utilities for Neoplastic DisEases of the Rectum by MRI Guided Radiotherapy Phase NAn=179 · primary completion 2027-01 · RT dose escalation gated by early regression index
- Late GI and GU toxicity of the boost beyond 9 years active Safety of a Boost (CXB or EBRT) in Combination With Neoadjuvant Chemoradiotherapy for Early Rectal Adenocarcinoma Phase 3n=148 · primary completion 2023-06 · safety endpoint for CXB vs EBRT boost after nCRTrecruiting Standard Dose Versus High Dose of Radiotherapy in Rectal Preservation With Chemo-radiotherapy in Rectal Cancer Patients Phase 3n=162 · primary completion 2026-12 · 62 Gy vs 50.4 Gy phase 3, dose-escalation toxicity
- Whether the curative-treatment rate imbalance explains the OS gap
📚 Sources · 📄 1 paper
Abstract
The longer read
The honest way to read this trial is to separate the endpoint that dose escalation can plausibly move from the endpoints it cannot. A simultaneous integrated boost of 56 Gy to gross primary and 60 Gy to involved lateral nodes, layered on a conventional 50 Gy/25 fx pelvis, is a mechanistically coherent way to improve local control, and local control at 9 years was 87.1% versus 70.1%. That number is the trial's defensible contribution. Metastasis-free survival of 70.8% versus 47.2% is a different kind of claim. Nothing about a pelvic boost acts on micrometastatic disease except through the pathway of preventing a local recurrence that later seeds distant sites, and in rectal cancer that pathway is real but far too narrow to carry a 23-point absolute separation. When the distant endpoint moves as far as the local endpoint, the more parsimonious explanation is that both are being pulled by something upstream of the randomization.
The candidate is visible in the paper's own numbers. Curative treatment, radical surgery or watch-and-wait, was achieved in 89.1% of the boost arm and 78.4% of the control arm. That comparison was not statistically significant at P=0.135, but with 51 and 55 pts a non-significant difference of that size is not evidence of no difference, and it describes exactly the mechanism by which an arm accumulates better survival for reasons other than dose. Nine pts who never reached curative treatment in one arm against six in the other is a small enough margin to be noise and a large enough one to generate the observed gap.
The primary endpoint failing sharpens rather than softens this concern. pCR was 15.2% versus 18.4%, numerically favoring the control arm. Dose escalation trials in rectal cancer have generally shown the opposite pattern, a measurable pCR gain that does not convert into survival, and the usual interpretation is that pCR is an imperfect surrogate. This trial reports a pattern with no dose-response signal at all in the tissue the boost irradiated, alongside large gains in every time-to-event endpoint. A surrogate that fails to move while its downstream endpoints move dramatically should lower confidence in the causal chain, not raise it.
The subgroup analysis is where the authors place the most weight and where it is least supportable. Benefit was concentrated in pts who did not receive perioperative chemotherapy, 9-year DFS 70.8% with HR 0.343 and P=.014, with no additional benefit in those who did. Chemotherapy receipt was not randomized; it is a post-treatment characteristic that reflects fitness, tolerance, and pathology. Splitting a 106-pt trial on such a variable produces subgroups where a few events determine the estimate, and it selects the healthier pts into whichever arm they land. The conclusion that dose escalation is most valuable in chemotherapy-ineligible pts is a reasonable hypothesis and not a finding.
What would have to be true for this to be right is a boost that improves local control enough to prevent a meaningful share of distant failures, in a population where local failure remains a dominant driver of death. That is more plausible for a 2013 to 2015 Chinese cohort than for a contemporary TNT-treated population, which is the reason this result does not transfer cleanly. For a radiation oncologist, the actionable read is narrow: the local control number and the acute toxicity profile, 14.5% versus 19.6% grade 3, support testing an integrated boost prospectively, particularly where systemic intensification is not on the table. The survival numbers should be treated as the reason to run that trial, not as its result.