The Absolute Effect / Evidence / RAPID
RAPID
After breast-conserving surgery for DCIS or node-negative invasive breast cancer, is twice-daily external-beam accelerated partial-breast irradiation non-inferior to whole-breast irradiation for ipsilateral breast tumour recurrence?
Registry: NCT00282035 · One underlying randomized study; aliases and follow-up reports consolidated.
Design, population and eligibility
Multicentre, randomized, non-inferiority phase III trial in 33 centres in Canada, Australia and New Zealand.
2,135 women aged ≥40 years; 1,070 assigned APBI and 1,065 whole-breast irradiation. Approximately 82% invasive disease and 18% DCIS only.
DCIS or node-negative invasive cancer after breast-conserving surgery; age ≥40 years.
Median follow-up 8.6 years; primary recurrence outcome reported at eight years.
Treatment and comparator
External-beam APBI 38.5 Gy in 10 fractions, twice daily over 5–8 days.
Whole-breast irradiation 42.5 Gy in 16 once-daily fractions or 50 Gy in 25 once-daily fractions.
Primary endpoint: Ipsilateral breast tumour recurrence (IBTR); non-inferiority margin expressed as upper 90% CI for HR <2.02.
Other endpoints: Acute and late radiation toxicity, cosmesis and treatment burden.
Key efficacy results
Eight-year cumulative IBTR was 3.0% with APBI and 2.8% with whole-breast irradiation; HR 1.27 (90% CI 0.84–1.91). The prespecified non-inferiority criterion was met.
Grade ≥2 acute toxicity within three months: 28% APBI versus 45% whole breast (P<0.0001).
Grade ≥2 late toxicity after three months: 32% versus 13% (P<0.0001); grade 3 late toxicity was 4.5% versus 1.0%.
Adverse cosmesis was more frequent with APBI: absolute difference 17.7 percentage points at seven years (95% CI 12.9–22.3).
Source-reported outcome visualization
Eight-year ipsilateral breast tumour recurrence · 8 years
Cumulative incidence estimates; the non-inferiority design and toxicity/cosmesis trade-off are essential to interpretation.
Absolute effect
Eight-year IBTR differed by 0.2 percentage points (3.0% versus 2.8%), with non-inferiority demonstrated. The 19-point increase in grade ≥2 late toxicity is not presented as an NNH because the abstract reports a broad post-three-month window rather than a single fixed-time risk estimate.
NNT: Not assigned; the absolute recurrence difference is small and the trial tested non-inferiority, not superiority.
NNH: Not assigned for late toxicity because the endpoint window is not a fixed time horizon; report the 19-point observed difference directly.
Toxicity, harms and treatment burden
Acute grade ≥2 toxicity was lower with APBI (28% versus 45%). Late grade ≥2 toxicity was higher (32% versus 13%); grade 3 late toxicity 4.5% versus 1.0%.
Adverse cosmesis (fair/poor) increased by 11.3 points at three years, 16.5 at five years and 17.7 at seven years.
The twice-daily schedule over 5–8 days is a treatment-burden advantage, but late induration and cosmetic deterioration may offset convenience.
Quality of life and patient-reported outcomes
Cosmesis is a major patient-relevant outcome. The observed late cosmetic disadvantage should be discussed alongside fewer acute reactions and fewer treatment visits; do not infer overall QoL superiority from fraction count.
Subgroups, maturity and limitations
The trial enrolled a broad population for APBI; many participants had low-risk, hormone-receptor-positive disease. Applicability to modern, more selectively chosen APBI populations is limited.
The non-inferiority margin allowed a relatively wide upper HR bound; a non-inferiority conclusion does not mean identical local control.
The late toxicity and cosmesis signal may be related to twice-daily fractionation; once-daily APBI schedules require evidence from their own trials.
Clinical interpretation
Current-practice context · 1 November 2026 (full guideline freshness review pending): Contemporary partial-breast recommendations favour careful patient selection and evidence-based once-daily or other validated regimens. Do not use RAPID to claim equivalence of every APBI technique.
Primary sources and registry
Source hierarchy: primary randomized report and prespecified follow-up first. Missing numerical toxicity, discontinuation or QoL data are labelled rather than imputed.