High-dose hyperfractionated SIB RT vs standard-dose RT for limited-stage SCLC NCT03214003
ForLS-SCLC (VALSG), age 18–70, ECOG 0–1, PET-CT and brain MRI staged
HR 0.55
95% CI 0.37–0.72, p=0.003 · mOS 60.7 vs 39.5 mo
TL;DRmOS 60.7 vs 39.5 mo, HR 0.55 (0.37–0.72), p=0.003 with 54 Gy/30 fx BID SIB vs 45 Gy/30 fx BID in LS-SCLC; toxicity similar.
Escalation is a GTV/IGTV-only SIB (PTV held at 45 Gy) within the same 30 fx, 3-week BID course on PET-defined involved fields. Local PFS separated (HR 0.51) while MFS did not (HR 0.81), fitting an in-field effect. The decision it moves is BID dose, 45 vs 54 Gy, not BID vs once daily.
In PET-CT and brain-MRI staged LS-SCLC, age 18–70 and ECOG 0–1, planned for concurrent platinum-etoposide with BID thoracic RT, this supports 54 Gy SIB as an alternative to 45 Gy; it does not extend to pts over 70, ECOG 2, or once-daily schedules.
A GTV/IGTV-only boost to 54 Gy with PTV at 45 Gy keeps 30 fx BID over 3 weeks on involved PET-defined fields, with G3–4 acute oesophagitis 13% vs 12%. The actionable choice is BID dose, 45 vs 54 Gy, for fit PET-staged pts; once-daily escalation is not addressed.
8 details 1 trial watching
Open-label randomised phase 3, 1:1, 16 public hospitals in China, N=224 (108 vs 116) enrolled June 2017 to April 2021. Stratified by ECOG, stage, prior chemo courses and platinum choice. Stopped early by the DSMB at the interim after 88 deaths; median follow-up 46 mo.
Age 18–70, ECOG 0–1, LS-SCLC by VALSG staging, with mandatory whole-body PET-CT and brain MRI. Untreated or after 1–2 courses of platinum-etoposide. Median age 64; 86% stage III; 79% current or former smokers.
Cisplatin 75 mg/m² or carboplatin AUC 5 plus etoposide 100 mg/m² days 1–3, 4 courses every 3 weeks; almost 99% received all 4. PCI 25 Gy/10 fx for responders (82% vs 81%).
54 Gy/30 fx SIB to GTV/IGTV with PTV at 45 Gy vs 45 Gy/30 fx, both VMAT, BID at least 6 h apart, 5 days/wk for 3 weeks, starting 0–42 days after chemo cycle 1. Involved field: PET-positive primary and nodes, CTV = GTV + 5 mm, no elective nodal irradiation. Constraints included MLD ≤15 Gy, lung V20 <30%, cord max ≤40 Gy.
Primary: OS in ITT, measured from chemo start. Secondary: PFS, local PFS, MFS, DCR, acute and late toxicity (CTCAE v5.0), HRQOL (to be reported separately). Powered on HR 0.65 at 81% power, target 326 pts.
Primary endpoint crossed the interim boundary and held at the January 2023 cutoff, with 40 (37%) vs 66 (57%) deaths. Arm-level estimates are in the efficacy table.
| Endpoint | 54 Gy SIB | 45 Gy | HR (95% CI), p |
|---|---|---|---|
| Median OS (primary) | 60.7 mo (49.2–62.0) | 39.5 mo (27.5–51.4) | 0.55 (0.37–0.72), p=0.003 |
| 2-yr OS | 76% | 54% | p=0.001 |
| Median PFS | 30.5 mo (14.5–46.5) | 16.7 mo (10.7–22.7) | 0.70 (0.49–0.99), p=0.044 |
| Median local PFS | Not reached | 40.6 mo (22.8–58.4) | 0.51 (0.33–0.78), p=0.002 |
| Median MFS | 40.2 mo (26.6–53.8) | 23.8 mo (10.1–37.5) | 0.81 (0.57–1.16), p=0.248 |
| DCR | 97% | 93% | p=0.218 |
| Event | 54 Gy | 45 Gy | p |
|---|---|---|---|
| Oesophagitis | 14 (13%) | 14 (12%) | 0.84 |
| Pneumonitis | 5 (5%) | 7 (6%) | 0.663 |
| Neutropenia | 48 (44%) | 48 (41%) | 0.64 |
| Thrombocytopenia | 14 (13%) | 17 (14%) | 0.71 |
| Trial | Escalated schedule | Endpoint | Escalated arm | 45 Gy BID arm |
|---|---|---|---|---|
| This trial (phase 3) | 54 Gy/30 fx BID SIB | mOS · 2-yr OS | 60.7 mo · 76% | 39.5 mo · 54% |
| CONVERT (phase 3) | 66 Gy/33 fx QD | mOS · 2-yr OS | 25 mo · 51% | 30 mo · 56% |
| CALGB 30610/RTOG 0538 (phase 3) | 70 Gy QD | Median survival · 5-yr OS | 28.5 mo · 29% | 30.1 mo · 32% |
| Northern European randomised phase 2 | 60 Gy/40 fx BID | 2-yr OS | 74.2% | 48.1% |
Acute G3–4 RT toxicity similar between arms. No late G3 oesophagitis or stricture in either arm; late G3 pneumonitis in 2 vs 3 pts, no G3 pulmonary fibrosis. One treatment-related death (myocardial infarction) in the 54 Gy arm.
INT-0096 set 45 Gy/30 fx BID as the reference. CONVERT (66 Gy/33 fx QD) and CALGB 30610/RTOG 0538 (70 Gy QD) escalated dose without acceleration or hyperfractionation and did not beat it; the Northern European randomised phase 2 of 60 Gy/40 fx BID was the only prior BID escalation signal. The authors describe this as the first phase 3 to escalate while keeping BID delivery.
Control mOS of 39.5 mo beat the 24 mo design assumption and prior 45 Gy BID arms (20.8–30 mo), marking a selected cohort. Tumour and RT volumes were not controlled, and the interim analysis is listed as the only protocol amendment despite being described as preplanned. Reporting inconsistencies: median age given as both 64 and 62, and the OS CI upper bound (62.0) sits just above the 60.7 mo median.
The authors credit local control for the OS gain, but first-relapse rates were near identical (59 [55%] vs 65 [56%]) and chest-first failure did not differ significantly, so patterns of failure do not yet show where the benefit arises. Subgroup and HRQOL results are unreported.
CONSORT flow
Stopped at interim for benefit with 224 of 326 planned accrued; truncated RCTs tend to overstate effect. Single-country, open-label, no central RT QA, age ≤70.
- Does the 54 Gy SIB benefit hold in pts older than 70 or ECOG 2?
- 54 Gy/30 fx SIB BID vs 60 Gy/40 fx BID or high-dose once-daily? recruiting Dose-Escalation Radiotherapy in Limited-Stage Small Cell Lung Cancer: A Phase III Randomized Trial Phase 3n=300 · primary completion 2028-09 · phase 3: SIB 45-54 Gy vs 60 Gy vs 45 Gy BID, LS-SCLC
- Does higher thoracic dose add benefit alongside consolidation immunotherapy?
📚 Sources · 📄 1 paper
The longer read
For a radiation oncologist who already treats limited-stage disease twice daily, the escalation question has mostly been tested sideways. INT-0096 made 45 Gy in 30 BID fractions the reference, and the two phase 3 attempts to improve on it, CONVERT at 66 Gy in 33 once-daily fractions and CALGB 30610/RTOG 0538 at 70 Gy once daily, raised dose while giving up acceleration and hyperfractionation. Neither improved survival, which left unresolved whether dose was the wrong lever or whether abandoning the short BID course cancelled any gain. This trial isolates the variable: same 30 fractions, same three weeks, same 45 Gy PTV, with only the GTV/IGTV boosted to 54 Gy. The Northern European randomised phase 2 of 60 Gy in 40 BID fractions pointed the same way, so the direction is now supported by two randomised datasets.
The magnitude deserves less trust than the direction. The DSMB stopped the trial after 88 deaths with 224 of 326 planned pts accrued, when the interim HR was 0.50; with follow-up to January 2023 it sat at 0.55. Trials truncated for benefit tend to overstate effect, and drift toward the null as events accrue is the expected pattern. An upper CI bound of 0.72 leaves room for a real benefit even if the estimate shrinks further, but the 21.2-month median gain the authors cite should not be carried to tumor board as a stable figure.
The control arm is the second caution. Standard-dose pts reached a median OS of 39.5 mo, against 20.8–30 mo in earlier 45 Gy BID cohorts and a 24-month design assumption. The authors point to VALSG staging, mandatory PET-CT and brain MRI, and exclusion of ECOG 2 and anyone over 70, in a disease whose NCDB median age is 67. Selection does not bias the randomised contrast, but absolute medians will not transfer, and the relative benefit is untested in the older, less fit pts seen in practice. Open-label assessment matters little for OS but more for investigator-read PFS and local PFS, and with no central RT quality assurance across 16 centres, contour and delivered-dose consistency are unverified.
The mechanism is plausible but not closed. Local PFS favoured 54 Gy (HR 0.51) while MFS did not separate (HR 0.81, p=0.248), which fits an in-field effect. Yet first-relapse rates were 55% vs 56% and chest-first failure did not differ significantly, so the local-control story is not visible in first-site failure data, which record where relapse began rather than later in-field progression. Toxicity parity is the other half of the case: G3–4 acute oesophagitis 13% vs 12%, against the 32% the authors cite from INT-0096, delivered with VMAT, 4D-CT, cone-beam verification and involved fields. That parity likely depends on a GTV-only boost within PET-restricted volumes and may not hold with elective nodal fields or larger boost volumes.
What it moves: for a fit, PET-staged pt already planned for concurrent chemo with BID thoracic RT, a 54 Gy SIB is a defensible alternative to 45 Gy. What it leaves open: whether 54 Gy in 30 fractions matches or beats 60 Gy in 40 BID fractions, whether benefit survives beyond age 70, and how a higher GTV dose behaves alongside consolidation immunotherapy, which was not part of this chemo-only protocol.