Rivera S(1), Ghodssighassemabadi R(2), Auzac G(3), Brion T(4), Kirova Y(4),

# 5-year results of hypofractionated locoregional radiotherapy in early breast cancer HypoG-01 (UNICANCER): a French multicentre, randomised, non-inferiority, phase 3, open-label, controlled trial *The Lancet 2026* Articles 5-year results of hypofractionated locoregional radiotherapy in early breast cancer HypoG-01 (UNICANCER): a French multicentre, randomised, non-inferiority, phase 3, open-label, controlled trial Sofia Rivera, Robabeh Ghodssighassemabadi, Guillaume Auzac, Thomas Brion, Youlia Kirova, Séverine Racadot, Mohamed Benchalal, Jean-Baptiste Clavier, Claire Charra Brunaud, Anais Groulier, Delphine Argo-Leignel, Karine Peignaux, Ahmed Benyoucef, David Pasquier, Philippe Guilbert, Aurore Goineau, Agnes Tallet, Marie Bergeaud, Assia Lamrani-Ghaouti, Stefan Michiels, on behalf of the HypoG-01 trialists* Summary Lancet 2026; 407: 976–87 Background Hypofractionated radiotherapy is standard for whole-breast radiotherapy, but 50 Gy in 25 fractions See Comment page 920 (5-week radiotherapy) is still standard in many countries when nodal radiotherapy is needed for morbidity concerns. *The HypoG-01 trialists are listed The UNICANCER HypoG-01 trial aimed to assess morbidity and efficacy of hypofractionated locoregional radiotherapy in appendix 1 (p 45) delivering 40 Gy in 15 fractions (3-week radiotherapy) versus 5-week radiotherapy. Radiotherapy Department, Institut Gustave Roussy, Methods This non-inferiority, open-label, multicentre, randomised phase 3 trial, conducted in 29 centres in France, Villejuif, France included female patients 18 years and older, with invasive breast carcinoma requiring nodal irradiation after complete (S Rivera MD PhD, G Auzac MSc); microscopic resection of the primary tumour. Patients were randomly allocated in a 1:1 ratio to either 3-week radiotherapy Biostatistics and Epidemiology, Gustave Roussy INSERM 1018, (experimental group) or 5-week radiotherapy (control group) to the regional nodes and thoracic wall or breast. The Villejuif, France primary endpoint was ipsilateral arm lymphoedema, defined as a 10% or greater increase in arm circumference at 15 cm (R Ghodssighassemabadi PhD, proximal, 10 cm distal, or both, to the ipsilateral olecranon relative to baseline and contralateral arm, with a non-inferiority S Michiels PhD); Radiation margin on the hazard ratio (HR) of 1·545. This trial was registered at ClinicalTrials.gov (NCT03127995) and is closed to Oncology, Institut Curie, Paris, France (T Brion MD, recruitment. Y Kirova MD PhD); Radiation Oncology, Centre Leon Berard, Findings Between Sept 26, 2016, and March 27, 2020, 1265 patients were enrolled, and 1221 were included in per- Lyon, France (S Racadot MD); Radiation Oncology, Centre protocol analysis (median follow-up 4·8 years [IQR 4·01–5·02]), 614 assigned to 3-week radiotherapy and 607 to Eugene Marquis, Rennes, 5-week radiotherapy. The median age was 58 years (IQR 49–68). Arm lymphoedema occurred in 275 (25%) patients France (M Benchalal MD); (143 with 3-week radiotherapy and 132 with 5-week radiotherapy). 3-week radiotherapy was non-inferior to 5-week Radiation Oncology, ICANS - radiotherapy regarding arm lymphoedema risk (HR 1·02 [95% CI 0·79–1·31] p <0·001), with a 3-year Institut de Cancérologie non-inferiority Strasbourg Europe, Strasbourg, cumulative incidence of 23·4% (95% CI 19·7–27·6) and 22·2% (95% CI 19·5–26·3), respectively. Safety profiles were France (J-B Clavier MD); similar between groups; grade 3 or worse adverse events frequencies were 8% and 13%, respectively. Following Radiation Oncology, Institut de French regulation, data on race and ethnicity were not collected. Cancérologie de Lorraine - Alexis Vautrin, Vandoeuvre-lès- Nancy, France Interpretation 3-week radiotherapy (40 Gy in 15 fractions) was found to be non-inferior to 5-week radiotherapy (50 Gy (C Charra Brunaud MD); in 25 fractions) for arm lymphoedema risk and was comparably safe regarding other late normal tissue effects for Radiation Oncology, Centre patients prescribed locoregional radiotherapy for early-breast cancer. Antoine Lacassagne, Nice, France (A Groulier MD); Radiation Oncology, Hôpital du Funding French National Cancer Institute. Scorff, Lorient, France (D Argo-Leignel MD); Radiation Copyright © 2026 Elsevier Ltd. All rights reserved, including those for text and data mining, AI training, and similar Oncology, Centre Georges technologies. François Leclerc, Dijon, France (K Peignaux MD); Radiation Oncology, Centre Henri Introduction Hypofractionated radiotherapy, delivering larger but Becquerel, Rouen, France Approximately one third of patients with breast cancer are fewer fractions with a lower total dose of 26 Gy in (A Benyoucef MD); Radiation diagnosed with regional-stage disease, a proportion that five fractions or 40 Gy in 15 fractions, resulting in a Oncology, Centre Oscar Lambret, Lille, France has remained stable over recent years, making it a reliable biologically equivalent dose slightly lower than the 50 Gy (Prof D Pasquier MD PhD); estimate for current clinical practice.1–3 These patients in 25 fractions, assuming an α/β ratio of 3–4 Gy for late- Radiation Oncology, Institut typically require regional nodal irradiation as part of their responding tissues, was proven non-inferior for Jean Godinot, Reims, France treatment to reduce the risk of recurrence and improve whole-breast irradiation, representing a major shift in (P Guilbert MD); Radiation Oncology, ICO - Institut de survival.4 However, the actual percentage of patients modern breast cancer management.6–9 The rationale for Cancerologie de l’Ouest - Site requiring nodal irradiation could be slightly higher, as hypofractionated radiotherapy is supported by radio- Paul Papin, Angers, France some node-negative patients with high-risk breast cancer, biological data indicating a low α/β ratio for both breast (A Goineau MD); Radiation and those in whom axillary lymph node dissection is cancer cells and normal breast tissue, suggesting Oncology, Institut Paoli- Calmettes, Marseille, France omitted, might also benefit from nodal radiotherapy.4,5 hypofractionated radiotherapy is as effective and safe as 976 Articles (A Tallet MD); Research & Research in context Development Department, Unicancer, Paris, France Evidence before this study controlled trials to publish 5-year outcome data specifically (M Bergeaud PhD, Approximately one-third of breast cancer patients are comparing locoregional radiotherapy over 3 weeks with 40 Gy A Lamrani-Ghaouti PhD) diagnosed with regional-stage disease and require regional in 15 fractions versus 5 weeks with 50 Gy in 25 fractions. Correspondence to: nodal irradiation as part of their treatment to reduce the risk of HypoG-01 shows non-inferiority of the shorter regimen in Sofia Rivera, Radiotherapy recurrence and improve survival. Hypofractionated terms of ipsilateral arm lymphoedema with a reassuring safety Department, Institut Gustave Roussy, 94800, Villejuif, France radiotherapy has been shown to be non-inferior to the profile. Additionally, HypoG-01 trial provides contemporary Sofia.rivera@gustaveroussy.fr historical 50 Gy in 25 fractions over 5 weeks for whole breast evidence confirming the safety of moderately hypofractionated See Online for appendix 1 irradiation with level 1A evidence, introducing a major shift in radiotherapy for comprehensive regional nodal volumes, modern breast cancer management when no regional including the internal mammary nodes (IMN) on survival irradiation was needed. The UK START trials established that endpoints: locoregional relapse-free survival, distant disease- 40 Gy in 15 fractions is non-inferior to 50 Gy in 25 fractions for free survival, breast cancer specific-survival, and overall survival. efficacy outcomes and is associated with lower rates of late While 40 Gy in 15 fractions has been the standard of care for normal tissue effects. We searched PubMed on Oct 2, 2025, nodal irradiation in the UK since NICE guidance in 2009, based using the search terms “breast cancer”, “regional nodal on the 5-year results of the START B trial, IMN irradiation was irradiation”, “hypofractionation”, and “randomised clinical not carried out in the trial. Hypofractionation adoption for IMN trials” and combinations thereof. We searched for original irradiation became widespread in the UK following the research articles and reviews published in English between 2016 consensus of the UK Royal College of Radiologists. Jan 1, 1980, and April 22, 2020. We found the primary endpoint HypoG-01 provides robust, prospective data from a phase 3 report of three randomised studies (one phase 2, two phase 3) trial specifically designed to assess the lymphoedema risk of this testing adjuvant locoregional hypofractionated radiotherapy modern treatment approach, confirming its safety and against 5-week regimens, ranging in sample size supporting its use as a global standard of care. from 50 to 820 patients. All offered consistent support for Implications of all the available evidence hypofractionation but only the smallest one included HypoG-01 provides evidence that moderately irradiation of the internal mammary chain and a small hypofractionated, 3-week, locoregional radiotherapy is a safe proportion of patients were irradiated on the lower axilla. alternative to historical 5-week radiotherapy for patients with Previously, the Early Breast Cancer Trialists’ Collaborative Group high-risk early breast cancer requiring nodal irradiation. meta-analysis has shown that regional irradiation improves Survival outcomes in HypoG-01 are encouraging and should be survival in women receiving effective local and systemic integrated with evidence from efficacy-focused trials. Our therapies. As none of the included trials used hypofractionation, findings contribute to a growing body of evidence supporting a the morbidity and risk/benefit of hypofractionated regional shift in clinical practice towards a 3-week radiotherapy standard irradiation remained to be evaluated. of care, offering substantial benefits to both patients and Added value of this study health-care systems, increasing treatment sustainability The UNICANCER HypoG-01 trial, alongside the Danish Breast without compromising oncological outcomes. Cancer Group Skagen trial 1, are the first phase III randomised the historical 5-week regimen for whole-breast a chronic, progressive condition that substantially affects irradiation.6–10 functional outcomes and quality of life. Whether the While the efficacy and safety of moderate favourable therapeutic ratio observed with whole-breast hypofractionation for whole-breast is well-established, hypofractionated radiotherapy extends to regional nodal high-level evidence for its use in comprehensive regional irradiation formed the primary justification for the nodal irradiation is scarce.6,7 Early data provided support, HypoG-01 patient-centred phase 3 trial. but a definitive trial was needed to confirm that a shorter Here, we report the 5-year outcomes of the HypoG-01 schedule does not increase toxicity, assuming similar trial, investigating moderately hypofractionated efficacy.11 The inclusion of nodal regions (eg, the radiotherapy in patients with early breast cancer who infraclavicular, supraclavicular, internal mammary, and have an indication for locoregional radiotherapy, potentially lower axilla) in the irradiated volumes including nodal irradiation. The aim was to reduce introduces additional complexity due to the proximity of treatment burden while ensuring similar or reduced critical organs, such as the brachial plexus, heart, lungs, toxicity and maintained efficacy. and thyroid, potentially increasing the risk of adverse events. Several systematic reviews and a recent meta- Methods analysis have identified regional nodal radiotherapy Study design and participants as a risk factor for arm symptoms, particularly HypoG-01 is a randomised, open-label, multicentre, lymphoedema.12 Treatment-related arm lymphoedema is phase 3, non-inferiority trial conducted across 29 French Articles centres (including 19 comprehensive cancer centres, Procedures four hospitals, and six radiotherapy centres). The study For all patients, CT-planning was mandatory; deep tested the safety and efficacy of a 15-fraction, 3-week inspiration breath hold (DIBH) was recommended; and schedule of adjuvant locoregional radiotherapy to the volume definition guidelines, dosimetric constraints, and regional nodes and whole breast or chest wall, compared patient setup verification procedures were detailed in the with a 25-fraction, 5-week schedule, with previously trial protocol and study manual (appendices 2 and 4). published acute toxicity data and radiotherapy quality Bolus use on the surgical scar was at investigator assurance.13,14 HypoG-01 was approved by the French discretion. Both 3D conformal radiotherapy (RT3D) and National Agency for Medicines and Health Products intensity-modulated radiotherapy (IMRT) techniques were Safety ([ANSM] 2016-A00702–49), the Île-de-France permitted as previously reported.13 Dose calculation was ethics committee (approval number CPP 16–012), and recommended using algorithms with lung inhomogeneity institutional review boards. It was conducted in correction (eg, Monte Carlo, anisotropic analytic algorithm, accordance with the principles of Good Clinical Practice. collapsed cone, or equivalent). Tumour-bed clip insertion The protocol, statistical analysis plan, and radiotherapy- was routine practice. Nodal boost was prohibited. See Online for appendices 2–4 planning manual can be found in appendices 2, 3, and 4, Recommended tumour-bed boost schemes according to respectively. The trial was registered at ClinicalTrials.gov treatment group were specified (appendices 2 and 4). (NCT03127995) and is closed to recruitment. All participating centres underwent a quality assurance Eligible participants were women aged 18 years or programme, including facility questionnaires and older with invasive breast carcinoma (T1–3, N0–3, M0) mandatory dummy-run validation with contouring and requiring nodal irradiation after complete microscopic radiotherapy planning before accrual. For the Dummy resection of the primary tumour, by mastectomy (with or run we used the HypoG-01 protocol recommendations as without immediate breast reconstruction) or breast- described in the trial protocol and study manual conserving surgery, with sentinel node biopsy, axillary (appendices 2 and 4). The dummy run was performed For the online platform see dissection, or both. Nodal irradiation was indicated using the ESTRO-Educase FALCON online platform. In https://estro.educase.com/ according to local guidelines based on pN+ status or pN0 addition, contour and radiotherapy plan reviews were with high-risk features, primarily including a performed randomly for one patient within the first combination of grade 3, lymphovascular invasion, and 3 months of trial initiation per centre, and for at least being younger than 40 years. Primary and adjuvant five patients per year per centre. Full Digital Imaging and chemotherapy, trastuzumab, and endocrine therapy Communications in Medicine for radiotherapy were permitted per institutional guidelines. Patients (DICOM-RT) data were requested and stored for all receiving chemotherapy were randomised within patients via the online platform Onco Place (Aquilab by 4 weeks of their last cycle. Trastuzumab, when indicated, Coexya). was continued during radiotherapy. Endocrine therapy, Patients were followed up at the end of treatment, when indicated, was recommended to start after 1 month after completing radiotherapy, then at 6 months, radiotherapy. If surgery was the last treatment before annually from years 1 to 5, and at year 10 after random radiotherapy, patients were randomised within 42 days allocation. At each visit, clinicians assessed adverse events; post-surgery. All patients provided written informed arm lymphoedema; shoulder range of motion; cosmetic consent. Following French regulation, data on race and outcomes; and oncological outcomes, including local, ethnicity were not collected. regional, contralateral, distant recurrence, and vital status. Adverse events were scored using the Common Termi- Randomisation and masking nology Criteria for Adverse Events (CTCAE version 4.0), At inclusion, patients were randomly assigned in a Radiation Therapy Oncology Group–European 1:1 ratio to receive nodal and breast radiotherapy or chest Organisation for Research and Treatment of Cancer wall radiotherapy in either the 3-week experimental (EORTC), and Late Effects Normal Tissue Task Force– group (40 Gy in 15 fractions of 2·67 Gy with or without a Subjective, Objective, Management, Analytic scales. tumour-bed boost) or the 5-week control group (50 Gy in All patients were invited to participate in health-related 25 fractions of 2 Gy with or without tumour-bed boost). quality of life (QoL) assessment using patient-reported Tumour-bed boosts (either sequential or simultaneous outcomes, collected at baseline and at each follow-up integrated) following breast-conserving surgery was visit, using the EORTC QLQ-C30, BR23 breast cancer permitted at investigator’s discretion. We conducted module, EQ-5D, morbidity, Body Image Score, and randomisation online using the Pocock and Simon protocol-specific questionnaires related to pain, sensory minimisation algorithm based on the following disturbances, and arm symptoms. All patients were minimisation factors: type of surgery (mastectomy invited to participate in a cost-effectiveness analysis. vs breast-conserving), axillary clearance (yes vs no), Transportation and sick leave questionnaires were radiotherapy centre, number of involved nodes collected at the end of treatment. (0 vs 1–3 vs ≥4), BMI (≤25 vs 25–30 vs >30). Treatment All patients treated with breast-conserving surgery allocation was not masked to clinicians or patients. were eligible for cosmetic assessment. The cosmetic 978 Articles score was independently assessed by patients and non-breast cancer death for breast cancer-specific clinicians using the Harris 4-point scale, and photographs survival). Adverse events of interest included cutaneous, were taken at baseline and at each follow-up visit. Digital subcutaneous, respiratory, cardiac, and endocrine photographs were scored by two independent observers adverse events, fatigue, pain, dysphagia, lymphoedema, (masked to treatment group). All assessable photographs and brachial plexopathy. were scored using BCCT.core software. All patients were invited to participate in the sarcopenia Statistical analysis assessment analysis based on baseline CT scan using The primary prespecified statistical test was a one-sided LITIS-QUANTIF group software. The different sub- log-rank test in the per-protocol population, stratified by studies, including analysis of patient-reported outcomes, the minimisation factors used in random allocation. A QoL, cost-effectiveness, cosmesis, and sarcopenia, will be sensitivity analysis using the intention-to-treat (ITT) reported separately. population was pre-planned. We estimated that among patients receiving axillary lymph node dissection, Outcomes regional nodes radiotherapy, and taxane-based systemic The primary outcome was the occurrence of arm therapy, 10% would have developed arm lymphoedema at lymphoedema after radiotherapy, defined as a 3 years after receiving 50 Gy in 25 fractions. Further 10% or greater increase in arm circumference at 15 cm assumptions were an exponential distribution, the proximal, 10 cm distal, or both, to the ipsilateral olecranon acceptance of up to a 5% increase in arm lymphoedema relative to baseline and contralateral arm. This formula at 3 years (ie, 15% of the patients having cumulative arm uses the absolute circumference, using measuring tapes lymphoedema at 3 years with 40 Gy in 15 fractions, on both patient arms, collected at baseline and follow-up corresponding to a non-inferiority limit on the hazard visits. The primary analysis used a time-to-event ratio [HR] scale of HR=1·545), with 80% power and a approach with a timepoint of interest at 3 years. To one-sided 5% significance level. Based on these account for patients who did not meet the primary assumptions, 131 lymphoedema events were required. A endpoint, patients without lymphoedema events were total of 1012 patients (506 patients in each randomisation censored either at the date of their last follow-up visit group) were needed over 3 years in the per-protocol where arm circumference was measured, or at their date population (primary analysis). Accounting for a of death, whichever came first. Secondary outcomes were 20% exclusion rate from the per-protocol population, defined as non-cancer-related loco-regional events: acute 1265 patients were required overall. The null hypothesis and late toxicities (arm motion and function, cosmetic was that the hazard rate of the occurrence of results, and lung and cardiac events) and cancer-related lymphoedema (as per primary endpoint definition) with events (locoregional relapse-free survival, distant disease- 40 Gy in 15 fractions was worse than with 50 Gy in free survival, overall survival, causes of death, 25 fractions; the alternative hypothesis was that the HR health-related quality of life, and cost-utility). Early of the occurrence of lymphoedema was non-inferior. toxicities have been reported previously.14 Arm function Stratified Cox regression models were used to calculate (measured with patient-reported outcome measures), HRs and 95% and 90% CIs. cosmetic results, health-related quality of life, and cost- The per-protocol population was defined as patients utility will be reported separately at a later stage. Range of meeting all inclusion criteria who received the allocated motion impairment was defined as a 25° or greater treatment with no major deviations predefined by the reduction in active abduction or flexion compared with Steering Committee or quality assurance procedures contralateral arm. Survival endpoint definitions followed (eg, incorrect eligibility, not receiving the allocated Definition for the Assessment of Time-to-event treatment, not receiving nodal irradiation, boost to one or Endpoints in Cancer trials (DATECAN) recommen- more regional nodes, and incomplete resection). The ITT dations.15 Locoregional relapse-free survival events population was defined as all patients who were randomly included invasive ipsilateral recurrence, locoregional allocated. recurrence, ipsilateral ductal carcinoma in situ (DCIS), A sensitivity analysis was also planned using the or death from any cause; invasive disease-free survival primary endpoint as per our sister trial the Danish Breast events included invasive ipsilateral recurrence, Cancer Group Skagen Trial 1 (NCT02384733), in which locoregional recurrence, distant relapse, contralateral arm lymphoedema was defined as a 10% or greater arm relapse, second primary cancer, ipsilateral or contralateral circumference increase measured to the olecranon at DCIS, or death from any cause; distant disease-free 15 cm proximal, 10 cm distal, or both on the treated side, survival events included distant relapse or death from compared to the contralateral arm (not taking baseline any cause; breast cancer-specific survival events included measurements into account). deaths attributable to breast cancer only. Overall survival Exploratory competing events analyses were performed was defined as the time from random allocation to the for distant metastasis, locoregional recurrence, and date of death from any cause. Patients alive without an death from breast cancer. For each endpoint, specific event were censored at the date of their last visit (or competing events were considered in Fine–Gray models Articles and are detailed in appendix 1 (p 27). Restricted mean Restricted mean time lost was defined as the expected survival time was defined as the expected event-free time lost due to the event before the 5-year horizon (the survival time up to 5 years (the area under the area under the cumulative incidence curve). Both Kaplan–Meier survival curve from baseline to 5 years). were calculated non-parametrically. The effect of 1316 patients assessed for eligibility 51 ineligible 11 exceeded authorised time between previous treatments (surgery or systemic treatments) and radiotherapy 8 had no eligible TNM stage 6 had changed their decision regarding trial participation or treating centre 4 had bilateral breast cancer, history of cancer, or thoracic radiotherapy 4 exceeded authorised time between randomisation and radiotherapy 3 were considered not suitable by their physician due to medical condition, COVID-19 or unmet dosimetric constraints 2 had concomitant potential radiosensitising treatments 13 had other reasons (eg, participating in another trial, unspecified screen failure) 1265 randomly assigned 5 withdrew consent, denying the use of their data 631 assigned to 3-week 629 assigned to 5-week radiotherapy* and included radiotherapy† and included in ITT population in ITT population 17 with major deviations 22 with major deviations 3 withdrew consent 12 withdrew consent 8 had exclusion criteria or did 5 had exclusion criteria or not match inclusion criteria did not match inclusion 2 had excessive delay between criteria randomisation and 4 had excessive delay treatment initiation between randomisation 3 did not receive allocated and treatment initiation treatment 1 did not receive allocated 1 had major dosimetric treatment deviation 565 with baseline and end of 553 with baseline and end of treatment lymphoedema treatment lymphoedema measurements included measurements included in ITT primary endpoint in ITT primary endpoint analysis analysis 3 did not match 2 did not match inclusion or exclusion inclusion or exclusion criteria criteria 614 without major deviations 562 without major deviations 607 without major deviations 551 without major deviations and included in and included in included in per-protocol included in per-protocol per-protocol population per-protocol primary population primary endpoint analysis endpoint analysis Figure 1: HypoG-01 trial profile ITT=intention-to-treat. *3-week radiotherapy (experimental group) comprised moderately hypofractionated radiation therapy (40 Gy in 15 fractions every 3 weeks, with or without a tumour bed boost). †5-week radiotherapy (standard group) comprised 50 Gy in 25 fractions every 5 weeks, with or without a tumour bed boost. 980 Articles Experimental group Standard group Experimental group Standard group (N=614)* (N=607)† (N=614)* (N=607)† Age, years (Continued from previous column) Mean (SD) 58·5 (13·1) 58·2 (12·8) Axillary exploration Median (IQR) 58 (49–68) 58 (48–68) Axillary clearance only 310 (50%) 326 (54%) Range 23–91 29–87 Axillary clearance and sentinel 195 (32%) 173 (29%) node or nodes biopsy Breast size‡ Sentinel node or nodes biopsy 109 (18%) 108 (18%) Small 69 (11%) 82 (14%) only Medium 241 (39%) 230 (38%) Radiotherapy technique Large 257 (41%) 236 (39%) IMRT 324 (53%) 314 (52%) Unknown 47 (8%) 59 (10%) RT3D 290 (47%) 293 (48%) Laterality Tumour-bed boost Left 309 (50%) 324 (53%) Total 293 (48%) 303 (50%) Right 305 (49%) 283 (47%) Integrated (SIB) 97 (16%) 95 (16%) Tumour size, mm Sequential 196 (32%) 208 (34%) Patients with tumour size 607 (97%) 600 (99%) Irradiated nodal levels§¶ available Mean tumour size, mm (SD) 26·2 (18·1) 26·1 (18·6) CTVn_L1–2§ 0 0 CTVn_L3–4 88 (14%) 72 (12%) Histology Ductal 491 (79%) 493 (81%) CTVn_L1–4§ 16 (3%) 21 (3%) CTVn_L3–4_IMN 133 (22%) 115 (19%) Lobular 87 (14%) 78 (13%) Other 35 (6%) 32 (5%) CTVn_L1–4_IMN§ 98 (16%) 99 (16%) CTVn_L2–4_IMN 8 (1%) 12 (2%) Unknown 1 (<1%) 4 (1%) Grade CTVn_L2–4_IMN§ 131 (21%) 137 (23%) Systemic treatment I 63 (10%) 57 (9%) Preoperative chemotherapy 130 (21%) 155 (26%) II 319 (51%) 351 (58%) Adjuvant Chemotherapy 389 (63%) 386 (64%) III 223 (36%) 191 (31%) Trastuzumab 106 (17%) 116 (19%) Unknown 9 (1%) 8 (1%) Preoperative endocrine therapy 13 (2%) 7 (1%) Breast cancer subtype Adjuvant endocrine therapy 496 (81%) 498 (82%) HER2+ 114 (18%) 125 (21%) HER2– with ER+ or PR+ 426 (68%) 420 (69%) Data are n (%) unless otherwise stated. The number of patients may vary for some characteristics due to missing data, because they are applicable only to a subset of HER2– and ER– and PR– 69 (11%) 61 (10%) patients, or because they are not mutually exclusive (eg, some patients had a Unknown 5 (1%) 1 (<1%) combination of several systemic treatments). Level 1=lower region of axilla (L1). T stage Level 2=upper region of axilla (L2). Level 3=infraclavicular region (L3). Level 4=supraclavicular region (L4). CTV=clinical target volume. CTVn_L=CTV of 0 19 (3%) 17 (3%) corresponding nodal level¶. ER=oestrogen receptor. ESTRO=European Society for 1 205 (33%) 202 (33%) Radiotherapy and Oncology. PR=progesterone receptor. HER2=Human epidermal 2 293 (48%) 285 (47%) growth factor receptor 2. IMN=internal mammary nodal level. IMRT=intensity 3 80 (13%) 83 (14%) modulated radiation therapy. RT3D=conformal 3D radiation therapy. SIB=simultaneous integrated tumour-bed boost. *Experimental group was 4 5 (1%) 4 (1%) moderately hypofractionated radiation therapy (40 Gy in 15 fractions every Unknown 12 (2%) 16 (3%) 3 weeks, with or without a tumour bed boost). †Standard group was 5-week N stage radiotherapy (50 Gy in 25 fractions every 5 weeks, with or without a tumour bed boost). ‡Breast size was categorised based on bra size as declared by the patient at 0 255 (42%) 236 (39%) baseline using chest circumference in cm and the depth of the bra cup (A–G). 1 292 (48%) 297 (49%) Small was a chest circumference <80 cm, <80 cm cup A or B, 85 cm cup A or B, or 2 40 (7%) 41 (7%) 90 cm with cup A. Medium was chest circumference 80 cm or 85 cm cup size C, 90 cm cup size B or C, or 95 cm cup sizes A–C. Large was counted as any chest 3 15 (2%) 16 (3%) circumference >95 cm or cup sizes of ≥D. Unknown refers either to not recorded Unknown 12 (2%) 17 (3%) at baseline or not measurable as a category. §Irradiated nodal levels including Breast surgery Rotter interpectoral levels. ¶CTV delineation was based on ESTRO consensus guidelines. Mastectomy 276 (45%) 274 (45%) Breast conservation 338 (55%) 333 (55%) Table 1: Baseline demographic, clinical, and treatment characteristics of the per protocol population (Table 1 continues in next column) Pre-planned subgroup analyses (see statistical analysis hypofractionation was quantified as the difference in plan in appendix 3) for the primary endpoint were restricted mean times between the two randomly reported according to stratification factors: type of surgery, allocated groups. centres grouped by number of patients (≤30, 30–70, and Articles 3-week 5-week HR (95% CI) 3-week 3-week radiotherapy, 5-week radiotherapy, radiotherapy n/N radiotherapy n/N vs 5-week radiotherapy 5-year rate (95% CI) 5-year rate (95% CI) Arm lymphoedema 143/562 132/551 1·02 (0·79–1·31) 33·1% (28·5–37·8) 32·6% (27·8–37·8) Range of motion impairment 205/609 199/604 0·97 (0·79–1·19) 39·2% (34·9–43·7) 39·5% (35·1–44·1) Locoregional relapse-free survival 43/614 58/607 0·62 (0·38–1·01) 92·7% (90·1–94·6) 89·6% (86·6–92·0) Invasive-disease-free survival. 71/614 89/607 0·60 (0·32–1·14) 87·8% (84·8–90·3) 83·6% (80·1–86·6) Distant disease-free survival 52/614 70/607 0·54 (0·31–0·96) 91·2% (88·5–93·3) 86·9% (83·7–89·6) Breast cancer-specific survival 24/614 35/607 0·53 (0·30–0·94) 95·9% (93·8–97·3) 93·9% (91·4–95·7) Overall survival 36/614 51/607 0·59 (0·37–0·93) 94·0% (91·6–95·7) 90·8% (87·9–93·1) Data are n=number of events and N=number of participants. HR=hazard ratio. Table 2: Time-to-event outcomes in the per protocol population trial. Five patients withdrew consent for the use of their data and were removed from the ITT population, leaving 1260 in the ITT analysis (631 assigned to 40 Gy in 15 fractions [experimental group]; 629 assigned to 50 Gy 60 in 25 fractions [standard group]; figure 1). After random allocation, 15 patients withdrew their consent from active 40 participation in the trial but agreed to allow their previously collected data to be retained and analysed. 20 13 patients were found to be ineligible after randomisation (six had a T4 stage, two were metastatic, two required a 0 nodal boost, two had no nodal irradiation, and one had a 0 1 2 3 4 5 Time since random allocation, years primary liver cancer). Six patients exceeded the Number at risk authorised time between randomisation and radiotherapy (censored) 3-week radiotherapy 562 415 343 293 210 92 initiation, four patients did not receive their allocated (0) (93) (137) (161) (223) (332) treatment, and one patient had a major dosimetric 5-week radiotherapy 551 394 330 276 214 99 (0) (100) (143) (176) (226) (325) deviation due to specific anatomy and was treated outside of the trial upon the decision of the treating investigator. Figure 2: Cumulative incidence of ipsilateral arm lymphoedema by fractionation schedule in the per-protocol A total of 39 patients were excluded from the per-protocol population 3-week radiotherapy (experimental group) comprised moderately hypofractionated radiation therapy (40 Gy in population, leaving 1221 in the per-protocol analysis 15 fractions every 3 weeks, with or without a tumour bed boost). 5-week radiotherapy (standard group) comprised (614 assigned to experimental group and 607 assigned to 50 Gy in 25 fractions every 5 weeks, with or without a tumour bed boost. HR=hazard ratio. the standard group; figure 1). Median age was 58 years (IQR 49–68 ; range 23–91). Baseline and end of treatment >70 patients per centre), number of lymph nodes involved lymphoedema measurements were available for (0, 1–3, and ≥4), BMI (≤25, 25–30, and >30), axillary lymph 1118 patients from the ITT population; among them, node dissection, age of participants (≤55 years vs >55 years), 1113 were included in the per-protocol analysis. Patient, radiotherapy technique (IMRT vs conformal RT3D), and tumour, and treatment baseline characteristics are irradiation of axillary nodal level 1. An exploratory presented in table 1 for the per-protocol population (for subgroup analysis by breast cancer subtype was included. the ITT population see appendix 1 pp 28–35). Axillary Lymphoedema rate per centre were reported. nodes irradiation depended on the type of axillary surgery Kaplan–Meier curves were used for visualisation, and and nodal burden, but appeared similar between the the Rothman method was used to calculate 95% CIs. standard and experimental groups independent of the All statistical analyses were conducted using SAS type of axillary surgery (appendix 1 pp 37–38). All analyses version 9.4 and R version 4.4.1.16 The trial had oversight were conducted on both the per-protocol and ITT from an independent Data Monitoring Committee. populations, yielding consistent results. Results in the ITT population are provided in appendix 1. Role of the funding source The median follow-up was 4·8 years (IQR 4·01–5·02). The funder had no role in study design, data collection, Of the 1221 patients in the per-protocol population, the data analysis, data interpretation, or manuscript median age was 58 years (IQR 49–68); for mean and preparation. range see table 1. Surgery included mastectomy (550 [45%] patients; among them 52 patients [9%] had Results immediate breast reconstruction) or breast-conserving Between Sept 26, 2016, and March 27, 2020, 1265 patients surgery (671 [55%] patients), and axillary clearance was were enrolled and randomly assigned in the HypoG-01 performed in 1004 patients (82%), with a mean of 982 % ,ecnedicni evitalumuC 3-week radiotherapy (intervention group) 5-week radiotherapy (standard group) 3-week vs 5-week radiotherapy: HR 1·02 (95% CI 0·79–1·31; 90% CI 0·82–1·26) p <0·001 non-inferiority Articles 12 removed nodes. A total of 1010 patients (83%) received (pp 11–15). For the ITT population see appendix 1 chemotherapy; among them 285 (28%) received (pp 16–20, 42). For distant disease-free survival, breast preoperative chemotherapy. 222 (18%) of 1221 received cancer-specific survival, and overall survival the 95% CI of trastuzumab. 994 (81%) received adjuvant endocrine the HR excluded the HR=1 value, favouring 3-week therapy (table 1). Radiotherapy plans (DICOM-RT plan, radiotherapy. In the per-protocol population, dose, and structures) were collected for 1093 (90%) of the 101 locoregional relapse-free survival events were per-protocol patients.13 DIBH was used in 80 (7%) of the recorded (43 in the 3-week radiotherapy group and 58 in per-protocol patients. A bolus on the mastectomy scar the 5-week radiotherapy group; table 2). Of these, there was used in 59 patients (5%), 33 of them in the were one and three ipsilateral breast tumour relapses experimental group compared with 26 in the standard and eight and nine regional relapses in the 3-week and group. Arm lymphoedema occurred in 275 patients (25%), of 1113 per-protocol patients with available lymphoedema Number of events/number of patients HR (95% CI) p value measurements, 143 (13%) in the experimental group and 3-week radiotherapy 5-week radiotherapy 132 (12%) in the standard group, respectively (table 2). In (intervention group) (standard group) the per-protocol population, the 3-week regimen was All patients 143/562 132/551 1·02 (0·79–1·31) non-inferior to the 5-week regimen for risk of arm Age, years lymphoedema (HR 1·02 [95% CI 0·79–1·31; ≤55 55/247 62/245 0·84 (0·56–1·27) 0·246 90% CI 0·82–1·26], p <0·001; figure 2, see non-inferiority >55 88/315 70/306 1·17 (0·83–1·67) appendix 1 p 3 for the ITT population). The 3-year BMI cumulative incidence of ipsilateral arm lymphoedema ≤25 72/277 63/275 1·07 (0·75–1·53) 0·225* was 23·4% (95% CI 19·7–27·6) with 3-week radiotherapy 25–30 32/164 37/157 0·73 (0·43–1·22) versus 22·2% (95% CI 19·5–26·3) with 5-week >30 39/121 32/119 1·28 (0·77–2·12) radiotherapy in the per-protocol analysis (figure 2, see Centre appendix 1 p 3 for the ITT population). The estimated ≤30 patients 15/66 17/61 0·46 (0·20–1·09) 0·639 cumulative incidence of ipsilateral arm lymphoedema up 30–70 patients 51/172 51/175 1·00 (0·66–1·53) to 5 years was 33·3% (95% CI 28·7–38·4) for 3-week >70 patients 77/324 64/315 1·18 (0·84–1·67) radiotherapy and 32·8% (95% CI 27·9–38·1) for 5-week Breast cancer subtype radiotherapy in the time-to-event analysis. The pre- HER2+ 33/110 32/118 0·98 (0·53–1·82) 0·424 HER2–, ER–, and PR– 20/64 13/55 1·18 (0·44–3·16) planned sensitivity analysis using the Skagen trial 1 HER2– and ER+ or PR+ 89/384 87/377 0·96 (0·69–1·32) definition showed a 3-year cumulative incidence of Surgical type ipsilateral arm lymphoedema compared with Lumpectomy 84/312 78/305 1·07 (0·77–1·49) 0·74 contralateral side, irrespective of baseline, of 26·8% Mastectomy 59/250 54/246 0·96 (0·65–1·41) (95% CI 23–31) with 3-week radiotherapy versus 24·9% Axillary lymph node dissection (21·1–29·1) with 5-week radiotherapy in the per-protocol No 20/96 19/96 1·37 (0·66–2·85) 0·392 population (appendix 1 pp 4–5). These results are Yes 123/466 113/455 0·98 (0·75–1·28) consistent with mean differences observed in arm Number of involved lymph nodes circumference in the ITT population at baseline and 0 18/58 10/51 1·73 (0·65–4·66) 0·974* 3-year follow-up, with no evidence of increased severe 1–3 91/360 89/364 1·07 (0·79–1·45) lymphoedema with 3-week radiotherapy (appendix 1 >4 34/144 33/136 0·75 (0·44–1·27) p 39). The cross-sectional lymphoedema incidence at Number of removed lymph nodes 3 years was 5% (23 of 453) and 4% (16 of 427) in the <10 58/221 57/242 1·21 (0·80–1·85) 0·342 ≥10 85/341 75/309 0·95 (0·68–1·33) 3-week and 5-week radiotherapy ITT groups, respectively. CTVn_L1 The estimated effect of 3-week radiotherapy on arm No 106/425 90/407 1·12 (0·83–1·51) 0·428 lymphoedema was consistent across all predefined Yes 37/136 42/143 0·85 (0·51–1·43) subgroups (figure 3; appendix 1 pp 6–7 for the ITT Type of radiotherapy population). No sign of treatment centre effect on IMRT 82/291 72/280 1·13 (0·79–1·61) 0·631 lymphoedema 3-year rate was observed in either the per- RT3D 61/271 60/271 0·98 (0·66–1·46) protocol or ITT groups (appendix 1 p 8). 0 1 2 3 4 In the per-protocol population, range of motion impairment occurred in 404 patients, with no obvious Favours 3-week radiotherapy Favours 5-week radiotherapy difference between treatment groups (HR 0·97 Figure 3: Hazard ratios for risk of arm lymphoedema by predefined factors in the per-protocol population [95% CI 0·79–1·19]; table 2 and appendix 1 pp 9–10, 41). Forest plot of subgroup analysis for the primary endpoint of arm lymphoedema. p-values correspond to tests for Regarding survival outcomes, the estimated treatment interaction, or for trend. Asterisks designate tests for trends. 3-week radiotherapy (experimental group) comprised effects on locoregional relapse-free survival, invasive moderately hypofractionated radiation therapy (40 Gy in 15 fractions every 3 weeks, with or without a tumour bed boost). 5-week radiotherapy (standard group) comprised 50 Gy in 25 fractions every 5 weeks, with or without a disease-free survival, distant disease-free survival, breast tumour bed boost. CTVn_L1=Clinical target volume of the nodal level 1. IMRT=intensity modulated radiation cancer-specific survival, and overall survival are presented therapy. RT3D=conformal 3D radiation therapy. HER2=human epidermal growth factor receptor 2. ER=oestrogen for the per-protocol population in table 2 and appendix 1 receptor. PR=progesterone receptor. Articles 5-week RT groups, respectively. No obvious differences in breast cancer requiring breast or chest wall and regional locoregional relapse-free survival were observed between nodal irradiation, with a median follow-up of 4·8 years. the two groups. The 5-year locoregional relapse-free Shoulder range of motion impairment and other adverse survival was 92·7% (95% CI 90·1–94·6) for the 3-week events up to 5 years were similar between schedules. No radiotherapy group and 89·6% (95% CI 86·6–92·0) for substantial detrimental effects of the 3-week radiotherapy the 5-week radiotherapy group. A total of 87 patients died were observed for locoregional relapse-free survival, during follow-up (36 in the 3-week radiotherapy group invasive disease-free survival, distant disease-free and 51 in the 5-week radiotherapy group), 59 of whom survival, or overall survival. The 95% CIs of the HRs died from breast cancer (24 in the 3-week radiotherapy favoured the 3-week radiotherapy group for distant group and 35 in the 5-week radiotherapy group). No sign disease-free survival, breast cancer-specific survival and of a detrimental effect of hypofractionated radiotherapy overall survival. However, this trial was designed and was observed on any of the survival outcomes. powered for a non-inferiority toxicity endpoint, not for Estimated 5-year differences in restricted mean time efficacy, and these secondary findings should be lost for arm lymphoedema and range of motion interpreted with caution. These findings are consistent impairment, and in 5-year restricted mean survival time with the 10-year START-B results demonstrating for locoregional relapse-free survival, invasive disease- significantly fewer distant relapses up to 10 years in the free survival, distant disease-free survival, breast 40 Gy group (HR 0·74 [95% CI 0·59–0·94]),6 with at least cancer-specific survival, and overall survival showed as good disease-free survival and overall survival. These consistent results in both the per-protocol and ITT findings support our hypothesis that 3-week radiotherapy, populations (appendix 1 pp 41–42). including regional nodal irradiation of axilla levels and In addition, the alternative modelling approach using the internal mammary chain, can reduce treatment competing risks models for locoregional recurrence, burden without compromising efficacy or increasing distant recurrence, breast cancer death, and other causes toxicity in patients with early breast cancer. of death showed no sign of a detrimental effect of While oncological efficacy remains paramount, hypofractionated radiotherapy (appendix 1 pp 21–23 for preventing treatment-related morbidity is equally crucial the per-protocol and pp 24–26 for the ITT population, for patients’ QoL and long-term functional outcomes. This respectively). The HR for breast cancer deaths was 0·66 motivated the selection of ipsilateral arm lymphoedema as (95% CI 0·39–1·11) for 3-week versus 5-week the primary endpoint. Our findings are clinically relevant, radiotherapy in the per-protocol population (appendix 1 given the well-established association between regional p 23). nodal irradiation and lymphoedema risk, one of the most The overall safety profile was similar between the feared long-term complications of early breast cancer randomly allocated groups (appendix 1 pp 43–44). In the treatment, substantially affecting patients’ functional per-protocol population, 32 serious adverse events were capacity and QoL.12,17,18 Breast-cancer treatment-related arm reported during follow-up (17 in the 3-week radiotherapy lymphoedema is typically a progressive and chronic group and 15 in the 5-week RT radiotherapy group), of condition, affecting more than one in five patients who are which two were related to radiotherapy, one in each in remission from breast cancer.19 However, reported rates group. One suspected unexpected serious adverse vary widely across studies, ranging from 0% to 94%, due to reaction (idiopathic thrombocytopenic purpura) was heterogeneous measurement methods (eg, circumferential reported in the 5-week radiotherapy group. Adverse arm measurements, bioimpedance spectroscopy, or events were otherwise mild and balanced across the perometry), subjective assessments (eg, patient-reported randomly allocated groups. Grade 3 or worse adverse outcomes or clinical evaluation), and follow-up periods events were reported in 158 of 1221 evaluable ranging from months to years.19–24 This underscores the patients (13%) with similar frequencies between the importance of our prospectively defined, standardised two randomly allocated groups. A similar pattern was assessment protocol and the substantial 5-year follow-up. observed in adverse events of interest (appendix 1 p 44). While arm circumference measurement is a pragmatic Besides radiodermatitis, the most frequent adverse surrogate for volumetry, it is a reliable and validated events were pain (838 [69%]), fatigue (791 [65%]), and approach for large trials, and a 2 cm change in fibrosis (578 [47%]; appendix 1 p 44). Notably, 55 and circumference predicts a change of 6·0% to 9·8% in relative 45 CTCAE grade 2 or worse lymphoedema events were arm volume.25 Bioimpedance spectroscopy, which reported in the 3-week and 5-week radiotherapy groups, measures tissue fluid content through electrical respectively, representing 8% of per-protocol patients. impedance, is an alternative, offering the potential for earlier detection of subclinical lymphoedema, but its Discussion increased sensitivity raises questions about capturing In this trial, we demonstrated the non-inferiority of clinically meaningful lymphoedema.26 Accordingly, the 3-week radiotherapy (40 Gy in 15 fractions) compared sister trials HypoG-01 and Skagen Trial 1 used the same with 5-week radiotherapy (50 Gy in 25 fractions) for common and pragmatic circumference-based method and ipsilateral arm lymphoedema in patients with early timepoints for arm lymphoedema assessment. HypoG-01 984 Articles included a rigorous, centralised quality assurance appropriate when regional nodes are included. Beyond programme for assessors’ training. An exploratory analysis reassuring safety and oncological outcomes, the use of of lymphoedema rates by centre did not reveal concerning ESTRO consensus guidelines for target definition, robust patterns across centres. In addition, randomisation was radiotherapy quality assurance, and the consistency of stratified by centre to account for potential centre effects. results between per-protocol and ITT analyses reinforces This, combined with the objective nature of the primary the robustness of the findings and supports their endpoint, provides confidence that the risk of measurement generalisability.13,35 The implications are substantial. A bias was minimised. To account for natural arm size 3-week schedule offers significant advantages for patients variation, measurements were considered relative to the and health-care systems, alongside a reduction in the contralateral side. To account for post-surgery but pre- treatment’s carbon footprint. For patients, it means fewer radiotherapy changes HypoG-01 also considered hospital visits, reduced travel time and costs, and less measurements relative to baseline, unlike Skagen Trial 1. disruption to their daily lives. For health-care systems, it Using both trials definitions, arm lymphoedema incidence can lead to increased treatment capacity and resource seems higher in HypoG-01, potentially reflecting higher efficiency. In light of global efforts to improve the value rates of axillary clearance (82%), although the patterns of and sustainability of cancer care, as highlighted by the arm lymphoedema cumulative incidence remain similar Lancet Breast Cancer Commission, adopting a 3-week and, consistently with the Skagen Trial 1, no detrimental standard for locoregional breast radiotherapy represents sign of hypofractionation was found. HypoG-01 and the a major advancement in addressing equity and Skagen Trial 1 are the largest phase 3 randomised trials sustainability challenges worldwide.36 evaluating 3-week radiotherapy using modern radiotherapy We must also acknowledge some limitations. First, few techniques in high-risk breast cancer patients requiring patients (7%) were treated with DIBH, a technique now regional nodal irradiation, including the internal widely adopted to reduce cardiac exposure. Second, only mammary nodes.27,28 Our findings build upon the 9% of mastectomy patients underwent immediate breast landmark UK START and Ontario trials,6,7 which reconstruction, which is increasingly common and might established the safety and efficacy of moderate influence the risk–benefit profile of different fractionation hypofractionated radiotherapy for the breast or thoracic schedules.13 Third, 82% of patients underwent axillary wall. However, the Ontario trial included only patients clearance, reflecting practice at the time of recruitment; without nodal involvement and excluded nodal current practice has moved toward significant radiotherapy.7 In START trials A and B, conducted over de-escalation of axillary surgery.37,38 This shift may reduce 20 years ago, 318 patients (14%) and 161 patients (7%), the baseline risk of lymphoedema, but is unlikely to respectively, received nodal irradiation, but contrary to differentially affect outcomes between schedules. Fourth, HypoG-01, no patients received internal mammary node 52% of patients received IMRT, while others received irradiation.10,29 Two additional phase 3 trials, in an Asian RT3D, as previously reported, reflecting technological population, demonstrated non-inferiority of moderate transition during our study.13 Current practice has largely hypofractionated radiotherapy but one used 2D shifted toward IMRT and volumetric modulated arc radiotherapy techniques in 97% of patients, and none therapy, which might offer improved dose distributions included internal mammary node irradiation, leaving a and different toxicity profiles. Finally, our study was crucial evidence gap, especially given increasing evidence conducted before the widespread adoption of of the survival benefit associated with internal mammary immunotherapy for triple-negative breast cancer, and no node irradiation.4,30–33 In HypoG-01, over 70% of patients patient received concomitant chemotherapy. The received internal mammary node irradiation (as previously interaction between modern systemic therapies, including reported) and no detrimental sign of hypofractionation on immunotherapy, and different radiotherapy fractionation safety, cardiac or lung side effects was observed, helping to schedules remains under investigation and could close this gap.13 Additionally, the UK IMPORT HIGH trial, influence the optimal integration of these treatments.39 although not designed to assess hypofractionated nodal In conclusion, the HypoG-01 trial provides evidence irradiation, provided reassuring data on the safety of this suggesting that moderately hypofractionated, 3-week approach.34 Together, this body of evidence solidifies the locoregional radiotherapy is a relatively safe and effective oncological safety of hypofractionated regional nodal alternative to historical 5-week radiotherapy for patients irradiation, allowing the focus to shift to minimising with early breast cancer requiring nodal irradiation. Our toxicity, which was the primary objective of HypoG-01. findings support a shift in clinical practice towards a Concurrently, the field is exploring even shorter, 1-week 3-week standard of care, offering significant benefits to (5 fraction) schedules, shown to be non-inferior for both patients and health-care systems, where the safety whole-breast irradiation alone in the FAST-Forward trial.8 of this regimen does not compromise oncological However, the safety and efficacy of 5-fraction schedules outcomes. for comprehensive regional nodal irradiation is not yet Contributors established in a phase 3 setting. Currently recruiting YK, SRi, SRa, MBen, J-BC, CCB, AGr, DA-L, KP, AB, DP, PG, AGo, AT, trials will determine whether a 1-week schedule is also TB, GA, RG, SM, AL-G, and MBer contributed to the review and editing Articles of the manuscript. RG, SM, SRi, MBen, AL-G, GA, and TB curated the 8 Murray Brunt A, Haviland JS, Wheatley DA, et al. Hypofractionated data. RG, SM, SRi, and TB performed the formal analysis. AL-G, and breast radiotherapy for 1 week versus 3 weeks SRi acquired funding. YK, SRi, SRa, MBen, JB-C, CCB, AGr, DA-L, KP, (FAST-Forward): 5-year efficacy and late normal tissue effects AB, DP, PG, AGr, AT, CM, AH, PR, and JP conducted the investigation. results from a multicentre, non-inferiority, randomised, phase 3 trial. Lancet 2020; 395: 1613–26. SM, RG, and SRi designed the methodology. AL-G, MBer, SRi, GA, SM, MBen, SRa, YK, and DP administered the project. AL-G, MBer, SR, GA, 9 Brunt AM, Haviland JS, Sydenham M, et al. Ten-year results of FAST: a randomized controlled trial of 5-fraction whole-breast SM, MBen, SRa, YK, and DP administered the project. YK, SRi, SRa, radiotherapy for early breast cancer. J Clin Oncol 2020; 38: 3261–72. MBen, J-BC, CCB, AGr, DA-L, KP, AB, DP, PG, AGr, AT, TB, and GA 10 The START triallists group. The UK Standardisation of Breast contributed resources. AL-G, MBer, SRi, SM, and GA supervised the Radiotherapy (START) Trial B of radiotherapy hypofractionation for project. AL-G, SM, and SRi validated the data. SM, RG, AL-G, MBer, and treatment of early breast cancer: a randomised trial. Lancet 2008; SRi visualised the data. SRi, SM, and RG wrote the manuscript. SM and 371: 1098–107. SRi conceptualised the study. The corresponding author had full access 11 Badiyan SN, Shah C, Arthur D, et al. Hypofractionated regional to all data and had final responsibility for the decision to submit for nodal irradiation for breast cancer: examining the data and potential publication. SM and RG also had full access to study data. for future studies. Radiother Oncol 2014; 110: 39–44. Declaration of interests 12 Liang Y, Zhou Y, Houben R, Verhoeven K, Rivera S, Boersma LJ. SM declares fees for Scientific Committee Study membership from A systematic review and meta-analysis of risk factors influencing Roche and for Data and Safety Monitoring of clinical trials membership patient-reported arm symptoms post-breast cancer treatment: accounting for radiotherapy impact. Breast 2024; 78: 103812. from IQVIA, Kedrion, Servier, and Yuhan, all outside the scope of the submitted work. All other authors declare no competing interests. 13 Brion T, Balde M, Allodji R, et al. Comprehensive radiotherapy quality assurance analysis from individual DICOM data in the Data sharing HypoG-01 phase 3 trial. Radiother Oncol 2026; 214: 111202. De-identified individual participant data, along with a data dictionary, 14 Brion T, Ghodssighassemabadi R, Auzac G, et al. Early toxicity of may be made available to researchers upon submission of a written, moderately hypofractionated radiation therapy in breast cancer detailed request to the HYPOG-01 trial steering committee at patients receiving locoregional irradiation: first results of the hypog-01@unicancer.fr. Trial documentation, including the protocol, is UNICANCER HypoG-01 phase III trial. Radiother Oncol 2025; available online. UNICANCER supports the broad dissemination of 207: 110849. research findings and encourages collaboration among investigators. 15 Gourgou-Bourgade S, Cameron D, Poortmans P, et al. Guidelines for Trial data are obtained, managed, stored, shared, and archived in time-to-event end point definitions in breast cancer trials: results of the DATECAN initiative (definition for the assessment of time-to- accordance with UNICANCER’s standard operating procedures to event endpoints in cancer trials). Ann Oncol 2015; 26: 2505–06. ensure their long-term quality, integrity, and utility. Formal data-sharing 16 R Core Team. (2024). R: a language and environment for statistical requests are considered in line with UNICANCER/UNITRAD computing. R Foundation for Statistical Computing, Vienna, procedures and in compliance with funder and sponsor guidelines. Austria. Data recipients must enter into a data-sharing agreement with 17 Hayes SC, Johansson K, Stout NL, et al. Upper-body morbidity after UNICANCER, specifying the conditions for data release and the breast cancer: incidence and evidence for evaluation, prevention, requirements for transfer, storage, archiving, publication, and and management within a prospective surveillance model of care. intellectual property. Requests are reviewed by the trial steering Cancer 2012; 118: 2237–49. committee based on scientific merit and ethical considerations, 18 Beaulac SM. Lymphedema and quality of life in survivors of early- including patient consent. Any restrictions related to patient stage breast cancer. Arch Surg 2002; 137: 1253. confidentiality and consent will be addressed by aggregating and 19 DiSipio T, Rye S, Newman B, Hayes S. Incidence of unilateral arm anonymising identifiable data. lymphoedema after breast cancer: a systematic review and meta- Acknowledgments analysis. Lancet Oncol 2013; 14: 500–15. We thank all the patients who participated in HypoG-01, the 20 Tsai RJ, Dennis LK, Lynch CF, Snetselaar LG, Zamba GKD, investigators and research teams at the participating centres, and Scott-Conner C. The risk of developing arm lymphedema among breast cancer survivors: a meta-analysis of treatment factors. UNICANCER. We also thank the UNITRAD team at UNICANCER and Ann Surg Oncol 2009; 16: 1959–72. the Institut National du Cancer (INCa) for providing the funds to 21 Armer JM, Stewart BR. Post-breast cancer lymphedema: incidence undertake this research (PHRC-K 15–146). We thank Tina Marie Zaarour, increases from 12 to 30 to 60 months. Lymphology 2010; 43: 118–27. Emilie Dasse, and Florence Villaret for their help with the coordination, 22 Erickson VS, Pearson ML, Ganz PA, Adams J, Kahn KL. Arm drafting and editing of the manuscript. edema in breast cancer patients. J Natl Cancer Inst 2001; 93: 96–111. References 23 Lee TS, Kilbreath SL, Refshauge KM, Herbert RD, Beith JM. 1 Giaquinto AN, Sung H, Newman LA, et al. Breast cancer statistics Prognosis of the upper limb following surgery and radiation for 2024. CA Cancer J Clin 2024; 74: 477–95. breast cancer. Breast Cancer Res Treat 2008; 110: 19–37. 2 Giaquinto AN, Sung H, Miller KD, et al. Breast cancer statistics, 24 Petrek JA, Heelan MC. Incidence of breast carcinoma-related 2022. CA Cancer J Clin 2022; 72: 524–41. lymphedema. Cancer 1998; (12 Suppl American): 2776–81. 3 Benitez Fuentes JD, Morgan E, De Luna Aguilar A, et al. Global 25 Ancukiewicz M, Miller CL, Skolny MN, et al. Comparison of relative stage distribution of breast cancer at diagnosis: a systematic review versus absolute arm size change as criteria for quantifying breast and meta-analysis. JAMA Oncol 2024; 10: 71. cancer-related lymphedema: the flaws in current studies and need 4 Taylor C, Dodwell D, McGale P, et al. Radiotherapy to regional for universal methodology. Breast Cancer Res Treat 2012; 135: 145–52. nodes in early breast cancer: an individual patient data meta- 26 Brunelle CL, Taghian AG. Lymphoedema screening: setting the analysis of 14 324 women in 16 trials. Lancet 2023; 402: 1991–2003. standard. Br J Cancer 2020; 123: 1–2. 5 Bartels SAL, Donker M, Poncet C, et al. Radiotherapy or surgery of 27 Rivera S, Ghodssighassemabadi R, Brion T, et al. 231O locoregional the axilla after a positive sentinel node in breast cancer: 10-year hypo vs normofractionated RT in early breast cancer: 5 years results results of the randomized controlled EORTC 10981–22023 of the HypoG-01 phase III UNICANCER trial. Ann Oncol 2024; AMAROS trial. J Clin Oncol 2023; 41: 2159–65. 35: S309. 6 Haviland JS, Owen JR, Dewar JA, et al. The UK Standardisation of 28 Offersen BV, Alsner J, Nielsen HM, et al. 5027 DBCG Skagen trial 1: Breast Radiotherapy (START) trials of radiotherapy phase III randomised trial of hypo- vs standard fractionated loco- hypofractionation for treatment of early breast cancer: 10-year regional radiotherapy in node-positive breast cancer patients. follow-up results of two randomised controlled trials. Lancet Oncol Radiother Oncol 2025; 206: S4430–32. 2013; 14: 1086–94. 29 START Triallists’ Group. The UK Standardisation of Breast 7 Whelan TJ, Pignol JP, Levine MN, et al. Long-term results of Radiotherapy (START) Trial A of radiotherapy hypofractionation for hypofractionated radiation therapy for breast cancer. N Engl J Med treatment of early breast cancer: a randomised trial. Lancet Oncol 2010; 362: 513–20. 2008; 9: 331–41. 986 Articles 30 Wang SL, Fang H, Song YW, et al. Hypofractionated versus 35 Offersen BV, Boersma LJ, Kirkove C, et al. ESTRO consensus conventional fractionated postmastectomy radiotherapy for patients guideline on target volume delineation for elective radiation therapy with high-risk breast cancer: a randomised, non-inferiority, open- of early stage breast cancer. Radiother Oncol 2015; 114: 3–10. label, phase 3 trial. Lancet Oncol 2019; 20: 352–60. 36 Coles CE, Earl H, Anderson BO, et al. The Lancet Breast Cancer 31 Wang SL, Fang H, Hu C, et al. Hypofractionated versus Commission. Lancet 2024; 403: 1895–950. conventional fractionated radiotherapy after breast-conserving 37 Man V, Duan J, Luk WP, Fung LH, Kwong A. Different strategies in surgery in the modern treatment era: a multicenter, randomized de-escalation of axillary surgery in node-positive breast cancer controlled trial from China. J Clin Oncol 2020; 38: 3604–14. following neoadjuvant treatment: a systematic review and meta- 32 Nielsen AWM, Thorsen LBJ, Özcan D, et al. Internal mammary analysis of long-term outcomes. Breast Cancer 2025; 32: 638–53. node irradiation in 4541 node-positive breast cancer patients treated 38 Cabıoğlu N, Koçer HB, Karanlık H, et al. De-escalation of nodal with newer systemic therapies and 3D-based radiotherapy surgery in clinically node-positive breast cancer. JAMA Surg 2025; (DBCG IMN2): a prospective, nationwide, population-based cohort 160: 257. study. Lancet Reg Health Eur 2025; 49: 101160. 39 Meattini I, Becherini C, Caini S, et al. International 33 Shaikh PM, Mulherkar R, Khasawneh MT, et al. Treatment of multidisciplinary consensus on the integration of radiotherapy with internal mammary nodes is associated with improved overall new systemic treatments for breast cancer: European Society for survival in breast cancer: a meta-analysis. Am J Clin Oncol 2024; Radiotherapy and Oncology (ESTRO)-endorsed recommendations. 47: 81–87. Lancet Oncol 2024; 25: e73–83. 34 Coles CE, Haviland JS, Kirby AM, et al. Dose-escalated simultaneous integrated boost radiotherapy in early breast cancer (IMPORT HIGH): a multicentre, phase 3, non-inferiority, open- label, randomised controlled trial. Lancet 2023; 401: 2124–37. --- [PDF原文](https://sci-net.xyz/storage/7932541/f81524d2c1d77b98c15276934ec5170d8341199e7fa040e708e78631ea21fb05/5-year-results-of-hypofractionated-locoregional-radiotherapy-in-early-breast-cancer-HypoG.pdf) DOI: 10.1016/S0140-6736(25)02597-8