One versus three weeks hypofractionated whole breast radiotherapy for early breast cancer treatment: the FAST-Forward phase III RCT

Breast cancer and the multidisciplinary team

Breast cancer is globally the most commonly occurring cancer in women with over 2 million new cases in 2018, including 58,000 in the UK. It is rare in men, with around 300–400 cases per year in the UK. Since the early 1990s, breast cancer incidence rates in females have increased by around a quarter (24%). 5 Despite increased incidence, UK mortality rates from breast cancer have fallen, due to advances in all aspects of breast cancer diagnosis and treatment, including radiotherapy (RT).

Breast cancer usually presents as symptoms such as a lump, nipple changes or a finding of an abnormal mammogram. Patients are generally assessed in a specific Breast clinic where they have a triple assessment involving clinical assessment, imaging (mammography/ultrasound) and a biopsy. A plan for treatment is then made in a multidisciplinary meeting which nearly always involves surgery to resect the cancer. In the modern era this is a wide local excision (breast-conserving therapy), or a mastectomy, along with nodal surgery. The resected tissue is examined by a pathologist to determine the size, grade, type, receptor profile and resection margins of the cancer along with other features. The lymph nodes may be selectively sampled (sentinel node biopsy) or all removed if involved (axillary clearance).

In larger, or more biologically aggressive cancers [triple negative and human epidermal growth factor receptor-2 (HER-2) positive] chemotherapy is now often given prior to surgery to downstage the cancer in the breast and the axilla, and to assess response to therapy, which may well alter subsequent treatment, including surgery, post-operative drug therapy and RT. Similarly in larger hormone receptor-positive HER-2 negative cancers, upfront antihormonal therapy may be given with similar intent. However, surgery is still used after the neoadjuvant therapy to remove any remaining cancer and to assess residual cancer burden.

Benefits and adverse effects of radiotherapy for patients with breast cancer

Radiotherapy uses high-energy X-rays to destroy cancer cells remaining in the breast after the main lump has been removed. Many studies have shown that this treatment substantially reduces the risk of cancer recurring in the breast and or lymph glands and improves overall survival. A meta-analysis from the Early Breast Cancer Trials Collaborative Group (EBCTCG) showed that after breast-conserving surgery, RT reduces relapse and breast cancer death. 6 Increased risk of local relapse is associated with age under 50 years and grade 3 cancers. Other risk factors include cancer size, cancer biology, negative hormonal receptor status, HER-2 positive status, the presence of lymphovascular invasion and axillary node involvement. The absolute benefits of RT are greatest in those with the highest risk factors. The EBCTCG meta-analysis included individual patient data for 10,801 women in 17 randomised trials of RT versus no RT after breast-conserving surgery, 8337 of whom had pathologically confirmed node-negative (pN0) or node-positive (pN+) disease. Overall, RT reduced the 10-year risk of any (i.e. locoregional or distant) first relapse from 35.0% to 19.3% [absolute reduction 15.7%, 95% confidence interval (CI) 13.7 to 17.7; 2p < 0.00001] and reduced the 15-year risk of breast cancer death from 25.2% to 21.4% (absolute reduction 3.8%, 1.6–6.0; 2p = 0.00005). In women with pN0 disease (n = 7287), RT reduced the first relapse risk from 31.0% to 15.6% (absolute relapse reduction 15.4%, 13.2 to 17.6; 2p < 0.00001) and risk of breast cancer death from 20.5% to 17.2% (absolute mortality reduction 3.3%, 0.8 to 5.8; 2p = 0.005), respectively. In these women with pN0 disease, the absolute relapse reduction varied according to age, grade, oestrogen-receptor (ER) status, tamoxifen use and the extent of surgery. These characteristics were used to predict large (≥ 20%), intermediate (10–19%), or lower (< 10%) absolute reductions in the 10-year relapse risk. The absolute reductions in 15-year risk of breast cancer death in these three prediction categories were 7.8% (95% CI 3.1 to 12.5), 1.1% (–2.0 to 4.2) and 0.1% (–7.5 to 7.7), respectively (trend in absolute mortality reduction 2p = 0.03). In the smaller number of women with pN+ disease (n = 1050), RT reduced the 10-year relapse risk from 63.7% to 42.5% (absolute reduction 21.2%, 95% CI 14.5 to 27.9; 2p < 0.00001) and the 15-year risk of breast cancer death from 51.3% to 42.8% (absolute reduction 8.5%, 1.8–15.2; 2p = 0.01). Overall, about one breast cancer death was avoided by year 15 for every four relapses avoided by year 10. The mortality reduction did not differ significantly from this overall relationship in any of the three prediction categories for pN0 disease or for pN+ disease.

Stepwise improvements in many aspects of patient management have led to a steady reduction in the absolute risk of relapse and death for patients over the decades. This reduction is due to many factors such as improved screening, better surgery with closer adherence to guidelines on achieving negative margins and advances in systemic therapy and RT techniques. A local relapse risk of around 2% at 5 years is a reasonable target to aim for in the current era, although the absolute risk will be higher in younger patients with grade 3 tumours and may be lower in elderly patients with grade 1 tumours. Currently patients at extremely low risk of relapse following surgery are commonly recommended no adjuvant RT based on studies such as PRIME. 7

The absolute benefit of RT for each patient is considered by the multidisciplinary team based on the biological and pathological staging of the tumour, type of surgery and patient factors, including age and comorbidities. The main short-term side effects of RT are tiredness, skin colour changes, discomfort and swelling (oedema). Long-term side effects may develop over many years following RT and these are thought to be due to damage to the small blood vessels supplying tissue that has been irradiated. Long-term side effects include the following: (1) skin changes, where the treatment area appears permanently tanned after treatment has finished, although this is not harmful. Later, the skin might appear to have very tiny broken veins in the skin called telangiectasia. (2) Breast shrinkage or distortion; RT can make the breast tissue contract so that the breast gradually gets smaller. This can happen to natural breast tissue or a reconstructed breast. An implant in a reconstructed breast can become hard (capsular contracture) and may need replacing. (3) Breast induration, where the breast feels hard and less stretchy; this is due to a side effect called radiation fibrosis. (4) Cough and breathlessness may occur in some patients who have RT to the chest area although this is not common. The problems are due to changes in the lung tissue called chronic radiation pneumonitis. They might start many months or a few years after treatment. The chances of this happening increase if the volume of lung irradiated increases, or if the patient has pre-existing lung conditions. (5) Irradiation of the ribs or clavicle may lead to radiation osteitis or rarely to osteoradionecrosis which may cause non-healing fractures. (6) Radiotherapy may cause nerve damage in the arm on the treated side, which can develop many years after treatment. Symptoms include tingling, numbness, pain, and weakness, and in some people it may cause some loss of movement in the arm and shoulder. This is extremely rare with modern RT. (7) Radiotherapy treatment may cause another type of cancer in many years’ time. These include lung cancers, especially in smokers, and the rare angiosarcoma.

Partial breast radiotherapy in low relapse risk patient subgroups

Instead of irradiating the whole breast, partial-breast RT restricted to the region of the original tumour in low relapse risk patients reduces the morbidity of RT without compromising its ability to cure the cancer. This technique is based on international reports of reductions in local relapse incidence, and the recognition that the majority of ipsilateral local relapses occur close to the region of the index tumour (the so-called tumour bed). Several studies have assessed partial-breast RT compared with whole breast RT. 8–12 Most of them show comparable rates of local control in patients receiving partial-breast RT compared with whole breast RT. However, most of the studies not only compared partial versus whole breast RT, but also used differing RT techniques and doses in the treatment groups, making toxicity comparisons difficult.

IMPORT LOW was a multicentre, randomised, controlled, phase 3, non-inferiority trial comparing the safety and efficacy of standard whole breast RT (control, whole breast group) with experimental schedules of RT to the whole breast and partial breast (reduced-dose group), and to the partial breast only (partial-breast group). 12 Patients assigned to whole breast RT (control) received 40 Gy in 15 fractions (daily doses) to the whole breast. Those assigned to the reduced-dose group received 36 Gy in 15 fractions to the whole breast and 40 Gy in 15 fractions to the partial breast containing the tumour bed, and those assigned to the partial-breast group received 40 Gy in 15 fractions to the partial breast only. Women who were aged 50 years or older who had breast-conserving surgery for unifocal invasive ductal adenocarcinoma (excluding invasive carcinoma of classical lobular type) of any grade (1–3) were recruited. Other inclusion criteria were pathological tumour size of 3 cm or less (pT1–2), axillary lymph-node negative or one to three positive lymph nodes (pN0–1), and a minimum microscopic margins of non-cancerous tissue of 2 mm or more. A total of 2018 patients were randomly assigned to the three groups, with around 670 patients per group. Patients were clinically assessed yearly for local relapse and toxicity, and around half completed questionnaires included self-assessments of side effects and health-related quality of life. After a median follow-up of 72.2 months, local relapse had been reported for 18 patients, nine (1%) of whom were in the whole breast group, three (< 1%) in the reduced-dose group, and six (1%) in the partial-breast group. The 5-year estimated cumulative incidence of local relapse was 1.1% (95% CI 0.5 to 2.3) in the whole breast group, 0.2% (95% CI 0.02 to 1.2) in the reduced-dose group, and 0.5% (95% CI 0.2 to 1.4) in the partial-breast group. At 5 years patients reported fewer moderate or marked normal tissue effects (NTE) in terms of skin change, change in overall breast appearance, breast being smaller, and breast being harder or firmer to touch in the partial-breast group than in the whole breast group although this reduction was statistically significant for change in breast appearance only (p < 0.0001). At 5 years, change in breast appearance had the highest cumulative incidence of items reported as moderate or marked by patients in all groups. Reports of breast becoming harder or firmer were significantly reduced in both the reduced-dose group (p = 0.002) and partial-breast group (p < 0.0001) compared with the whole breast group. The conclusions from IMPORT LOW were that for patients over 50, with smaller, generally lymph node-negative breast cancer, partial-breast RT was non-inferior to whole breast RT with regards local control and had less toxicity. Reduced-volume breast cancer treatment will also reduce lung and cardiac doses. Follow-up to 10 years is ongoing in the IMPORT LOW trial to collect data on longer-term efficacy and toxicity outcomes.

The control arm in IMPORT LOW was 40 Gy in 15 fractions to whole breast, which has been adopted as the standard of care for partial-breast RT. The control arm for the FAST-Forward Trial was the same. Therefore, if the 40 Gy/15 fractions schedule was found to be isoeffective for either of the test arms, then it was deemed that this would also be the same for partial-breast RT. It would make no biological sense to have different fractionation schedules with reduced-field RT (partial-breast). For FAST-Forward whole breast RT was still the standard-of-care when the study was conceived and recruiting.

Simplifying radiotherapy dose schedules for patients with breast cancer

For many years, and still in some countries, the international standard regimen for whole breast RT delivers a total dose of 50 Gy in 25 fractions over 5 weeks following surgical resection of primary tumour in women with early breast cancer. Attempts to reduce the number of fractions in the 1970s made inadequate downward adjustments to total dose, resulting in unacceptable rates of late complications. 13 These miscalculations inhibited further research in breast RT fractionation for decades, but interest in fewer larger fractions delivered over a shorter overall treatment time has been rekindled by randomised clinical trials based on a better understanding of normal tissue and tumour responses (radiobiology). Fractionation sensitivity describes responses of normal and malignant tissues to RT fraction size and is quantified in terms of the α/β ratio, expressed in Gy. The lower the α/β ratio, the greater the effect on normal and malignant tissues of changes in fraction size.

Four randomised trials involving a total of > 8000 women have compared a lower total dose in fewer larger fractions against 50 Gy in 25 fractions, and all have reported favourable results in terms of local tumour control and late adverse effects (AE). 14–18 The Royal Marsden Hospital/Gloucestershire Oncology Centre (now sometimes referred to as START-P) and Ontario Clinical Oncology Group (OCOG) trials totalling 2644 women with mainly axillary lymph node-negative tumours < 5 cm diameter were the subject of a 2008 Cochrane review of altered RT fractionation in early breast cancer. 19 Radiotherapy fractions larger than 2.0 Gy did not appear to affect: (1) local relapse-free survival (absolute difference 0.4%, 95% CI −1.5% to 2.4%), (2) breast appearance [risk ratio (RR) 1.01, 95% CI 0.88 to 1.17; p = 0.86], (3) survival at 5 years (RR 0.97, 95% CI 0.78 to 1.19; p = 0.75), (4) late skin toxicity at 5 years (RR 0.99, 95% CI 0.44 to 2.22; p = 0.98 or (5) late radiation toxicity in subcutaneous tissue (RR 1.0, 95% CI 0.78 to 1.28; p = 0.99). The review concluded that the use of unconventional fractionation regimens did not affect breast appearance or toxicity, nor appear to affect local cancer relapse. The results of the UK START trials (N = 4451) were published too late to be included in the overview but were consistent with the findings. The START Trials tested the effects of RT schedules using fraction sizes larger than 2.0 Gy. The START-A Trial tested two dose levels of a 13-fraction regimen delivered over 5 weeks in order to measure the sensitivity of normal and malignant tissues to fraction size. Patients in START-A were randomly assigned to either 50 Gy in 25 fractions (control group) or 41.6 Gy in 13 fractions or 39 Gy in 13 fractions over 5 weeks. Patients in the START-B trial were randomly assigned to either 50 Gy in 25 fractions (control group) over 5 weeks or 40 Gy in 15 fractions over 3 weeks. The START-A trial (N = 2236) showed that the estimated absolute differences in 5-year local-regional relapse rates compared with the control schedule of 50 Gy in 2.0 Gy fractions were 0.2% (95% CI −1.3% to 2.6%) after 41.6 Gy and 0.9% (95% CI −0.8% to 3.7%) after 39 Gy. In START A, photographic and patient self-assessments suggested lower rates of late AE after 39 Gy than with 50 Gy, with a hazard ratio (HR) for late change in photographic breast appearance of 0.69 (95% CI 0.52 to 0.91; p = 0.01). In the START-B trial (N = 2215) the estimated absolute difference in 5-year local-regional relapse rates for 40.05 Gy compared with 50 Gy was −0.7% (95% CI −1.7% to 0.9%), and the HR for late change in photographic breast appearance was 0.83 (95% CI 0.66 to 1.04). Therefore, the START trials reported similar local tumour control with some evidence of lower rates of late AE after schedules with fraction sizes larger than 2.0 Gy compared with the international standard 25-fraction regimen. 18

A 15-fraction schedule has been the UK standard-of-care recommended by the National Institute for Health and Care Excellence (NICE) since 2009, but it was thought to be unlikely to represent the useful limits of hypofractionation for whole breast RT. There is a history of prescribing once-weekly fractions of whole breast RT for women too frail or otherwise unable to attend for conventional schedules. In a French series of 115 patients undergoing primary RT without surgery for non-metastatic breast cancer from 1987 to 1999, the whole breast was treated with two tangential fields and received five once-weekly fractions of 6.5 Gy. 20 One hundred and one were given additional tumour bed boost doses, 7 with 1 fraction, 69 with 2 fractions and 25 with 3 once-weekly fractions of 6.5 Gy using electrons. Kaplan–Meier estimates of late effects in the breast were 24% grade 1, 21% grade 2 and 6% grade 3 at 48 months. The 5-year local progression-free rate was 78% (95% CI 66.6 to 88.4). In a separate French series, five once-weekly fractions of 6.5 Gy to the whole breast with no boost were given to 50 women after local tumour excision. 21 Grade 1 or 2 induration was reported in 33% of the patients at a median follow-up of 93 months (range 9–140). The 7-year local relapse-free survival was 91%.

The UK FAST trial (N = 915) tested two dose levels of a 5-fraction regimen delivering one fraction per week against a control schedule of 50 Gy in 25 fractions, defining RT AE as the primary endpoint. 22 The two test dose levels delivered 5 fractions of 5.7 Gy or 6.0 Gy (total dose 28.5 Gy or 30 Gy), estimated to be isoeffective with the control regimen assuming α/β values of 3.0 Gy or 4.0 Gy, respectively. Nine hundred and fifteen patients were recruited from October 2004 to March 2007. The mean age of participants was 62.7 years. Only 17 patients (5.2%) developed moist desquamation (12 after 50 Gy, 3 after 30 Gy, 2 after 28.5 Gy) out of 327 with Radiation Therapy Oncology Group (RTOG) skin toxicity data available. At a median follow-up of 28.3 months [interquartile range (IQR) 24.1–33.6], 729 patients had 2-year photographic assessments available, with mild and marked change in breast appearance in 19.3% and 1.7% after 50 Gy, 26.2% and 9.3% after 30 Gy, and 20.3% and 3.7% after 28.5 Gy. RRs for mild and marked change for 30 Gy versus 50 Gy were 1.48 (95% CI 1.06 to 2.05) and 6.06 (2.14 to 17.20); p < 0.001 for trend, favouring 50 Gy; and for 28.5 Gy versus 50 Gy were 1.07 (0.75 to 1.54) and 2.25 (0.70 to 7.18); p = 0.26 for trend, favouring 50 Gy. Any clinically assessed moderate or marked AE in the breast were increased for 30 Gy compared with 50 Gy (HR 2.19, 95% CI 1.46 to 3.29; p < 0.001), but similar for 28.5 Gy (HR 1.33, 95% CI 0.86 to 2.08; p = 0.19). At a median follow-up of 37.3 months two local tumour relapses had been recorded.

The 10-year results from the FAST trial were subsequently published in 2020,23 confirming that as far as late toxicity is concerned, the 28.5 Gy schedule was isoeffective to the 50 Gy schedule. The 30 Gy schedule had an increased rate of late toxicity compared with 50 Gy. Any moderate or marked physician-assessed NTE in the breast (shrinkage, induration, telangiectasia, oedema) was reported for 92/774 (11.9%) at 5 years and 55/392 (14.0%) at 10 years. The most prevalent individual effect was breast shrinkage. Five-year prevalence of any moderate or marked breast NTE was estimated to be 10% higher (95% CI 5% to 16%) for 30 Gy versus 50 Gy (p < 0.001), with no statistically significant difference between 28.5 Gy and 50 Gy (2%, 95% CI −2% to +7%; p = 0.349). At 5 years, RRs for moderate or marked breast shrinkage versus 50 Gy were 2.03 (95% CI 1.15 to 3.58; p = 0.017) for 30 Gy and 1.20 (95% CI 0.63 to 2.27; p = 0.604) for 28.5 Gy. There were no statistically significant differences between schedules in 5-year prevalence of moderate or marked breast induration, telangiectasia and breast oedema, nor in 10-year prevalence of any moderate/marked effects, with few marked events. At 10 years, the estimated absolute differences in prevalence of any moderate or marked breast NTE compared with 50 Gy were 9% (95% CI 1% to 18%; p = 0.032) for 30 Gy and 5% (95% CI −2% to +13%; p = 0.184) for 28.5 Gy.

Five- and 10-year cumulative incidence rates of moderate or marked NTE in the breast were higher for 30 Gy compared with 50 Gy, with statistically significant differences for any NTE in the breast, breast shrinkage, breast induration, and breast oedema. Cumulative incidence rates of any moderate or marked NTE in the breast and breast induration were significantly higher for 28.5 Gy versus 50 Gy. Modeling all annual physician assessments over follow-up, rates of moderate or marked effects were statistically significantly higher for 30 Gy compared with 50 Gy [odds ratio (OR) for any breast NTE 2.12, 95% CI 1.55 to 2.89; p < 0.001], but with no significant difference between 28.5 Gy and 50 Gy (OR 1.22, 95% CI 0.87 to 1.72; p = 0.248; Statistically significant differences between the test schedules were found for breast shrinkage, telangiectasia, and breast oedema, with higher rates for 30 Gy compared with 28.5 Gy. The prevalence of breast shrinkage and telangiectasia increased over time, with a decline in breast oedema.

By the time of the 10-year analysis of the FAST trial only 11 local relapses had been reported (50 Gy: 3, 30 Gy: 4, 28.5 Gy: 4), hence analysis of tumour control outcomes would be underpowered.

A gain in local tumour control due to shortening treatment time to 1 week from longer 3- to 5-week schedules is theoretically possible. Evidence based on retrospective studies for an influence of treatment time on local tumour control is conflicting with recent systematic reviews drawing different conclusions. 24,25 Even without a gain in tumour control, accelerated RT is likely to be more convenient for patients, and may ease scheduling with other treatment modalities. A pilot study (N = 30) tested 30 Gy in 5 fractions of 6.0 Gy in 15 days to the whole breast in terms of acute AE and late effects at 2 years. 26 In this series, 23/30 (77%) patients scored no change in post-operative photographic breast appearance at 2 years, 7/30 (23%) scored mild change and none scored marked change. The acute skin reactions were mild, with no reaction more severe than grade 2 erythema, scored in 9/30 (27%) patients. In conclusion, it is fair to say that after decades of resistance to evaluating larger RT fraction sizes in breast cancer, expert opinion is responding to an accumulating body of evidence supporting the safety and effectiveness of this approach.

Against this background, the FAST-Forward phase III randomised trial was designed, with the primary aim of testing local tumour control in women with early breast cancer following a 5-fraction schedule of adjuvant RT delivered in 1 week.

Lymph node radiotherapy for breast cancer

This sub-study to the Main Trial tested the safety of 5-fraction regimens in the context of lymphatic RT. The model of breast cancer spread that was dominant in earlier decades envisaged a limited role for regional therapy beyond protection of quality of life, typically secured by surgery. Systematic overviews of RT effects by the Early Breast Cancer Trialists Collaborative Group (EBCTG) provide level 1A evidence that prevention of local-regional relapse has a major impact on breast cancer mortality. 6,27 An important conclusion to be drawn is that even heavily axillary lymph node-positive axillary patients can be cured by effective local-regional treatment, whether this is achieved by surgery, RT or systemic therapy. 27 The traditional model of breast cancer spread still has some adherents. The Z11 trial randomised 891 out of a planned 1900 patients with clinically lymph node-negative, sentinel lymph node-positive disease to axillary clearance versus no further axillary treatment. 28 Twenty-seven per cent allocated axillary clearance had additional positive lymph nodes. The axillary relapse rate at 5 years was 0.5% after axillary clearance and 0.9% after no axillary clearance, and there was no difference in breast cancer mortality. For some, this result reinforces the traditional model of breast cancer spread that discounts a role of nodal metastases in determining cancer spread. This interpretation fails to take account that standard post-operative tangential beam RT includes at least lower axillary lymph nodes and may be needed to eradicate residual disease. The same issue has been raised in discussion of the IBCSG 23-01 trial that compared axillary dissection versus no further axillary surgery in 929 sentinel lymph node-positive patients, 91% of whom were treated by breast conservation followed by whole breast RT. 29 The disease-free survival events, including axillary relapses, were non-inferior in the group spared axillary dissection, where standard whole breast RT will have included at least level I axillary lymph nodes. Although this remains a contested area, there is a wide consensus that control of axillary disease, whether by surgery, systemic therapy or RT, is an important component of curative therapy.

The AMAROS trial is informative in determining the role of surgery or RT for axillary management. 30 The study randomised 1425 patients with positive sentinel lymph nodes to either axillary lymph node dissection in 744 patients or axillary RT including photon beams to axillary apex and medial supraclavicular fossa (SCF) in 681 patients. The axillary relapse rates were low in both groups; 1.19% after RT and 0.43% after surgery, too low to assess non-inferiority. The main difference is that clinically reported arm swelling was less of a problem after RT than after surgery; 13.6% versus 28.0%, respectively, at 5 years; p < 0.0001. The implication is that lymphatic RT may increasingly be used as an alternative to surgery in this context.

Very little internal mammary chain (IMC) RT has been given in the UK in recent decades, but a modest reduction in breast cancer mortality is suggested by the NCIC MA20 trial and confirmed by the European Organisation for Research and Treatment of Cancer (EORTC) 229922 trial. 6,31 Both tested IMC and SCF RT, so it is possible that therapeutic effects are attributable solely to the SCF component. This does not seem likely since, if SCF RT is needed in order to enhance cure of patients with positive IMC nodes, it is reasonable to assume that the latter require managing too. The same argument might apply to the infraclavicular (ICF) lymph nodes (level III axilla), which are usually included in unshielded (rectangular) fields to the SCF. In the EORTC trial, 4004 patients with axillary lymph node-positive disease or had cancers that were centrally or medially situated but axillary lymph node-negative were randomised to receive IMC/medial SCF RT or not. At a median of 10.9 years of follow-up, the primary endpoint of overall survival improved from 80.75% to 82.3% with the addition of IMC/SCF RT (HR 0.87, 95% CI 0.76 to 1.00; p = 0.0556; p = 0.0496 after adjustment). It is not currently clear how these results will impact on practice. In conclusion, there was a need to test the safety of a 5-fraction schedule of lymphatic RT if the FAST-Forward Trial is to remain relevant to the 25% of patients referred for treatment with lymph node-positive disease. Whereas most are currently referred following axillary dissection, international and UK practices are changing, and more patients are likely to be referred in future for RT to axillary, SCF/ICF and perhaps IMC lymph node groups. When hypofractionation for breast cancer was first introduced in the 1960s, inappropriate dose regimens and uncertain dosimetry combined to cause unacceptably high rates of brachial plexopathy in patients with early breast cancer. 32–44 Even with hindsight, it is difficult to be sure how much of the morbidity was related to technical factors, especially beam overlap, and how much to dose-time factors. The only series describing brachial plexopathy after total dose ≤ 50 Gy delivered in 1.8–2.0 Gy fractions to breast and axilla/SCF reported 3/724 (0.6%) affected patients treated between 1968 and 1985, of whom two resolved and one progressed at 6.5 years median follow-up. 44 All patients were treated at the Joint Center, Boston, in a supine treatment position using a three-field technique with hanging block. A review of all published evidence available in 2005, suggested that brachial plexopathy after local-regional RT for early breast cancer is uncommon (< 1%) at doses < 55 Gy in 2.0 Gy equivalents. 45 Dose regimens were normalised using a linear quadratic model, assuming α/β value of 2.0 Gy.

The dose–response relationships are not difficult to reconcile with current practices in head and neck cancer, where total doses of 60 Gy in 30 fractions are standard. One recent example modelled the relationship between total dose in 2.0 Gy fractions and probability of brachial plexopathy in 330 patients systematically screened for evidence of sensorimotor symptoms a median of 56 months (range 6–135) after radical RT for head and neck cancer. Patients treated with definitive RT received a median dose of 70 Gy, and for those treated post-operatively the median dose was 60 Gy. Intensity-modulated RT was used on 62% cases, and 40% had concurrent chemotherapy, usually cisplatin. The brachial plexus was outlined using RTOG criteria on X-ray computerised tomography (CT) scans. 46

Against this background, the FAST-Forward Trial was extended beyond its original target accrual in order to test the safety of hypofractionated lymphatic RT. It is realistic to test only the common dose-limiting AE, including arm swelling and overall arm function. Very uncommon AE, including brachial plexopathy, cannot be formally tested in such a protocol, since a non-inferiority trial wishing to exclude an excess 1% risk with standard statistical power would require tens of thousands of patients.

Embedded sub-studies

Eligibility for Nodal Sub-Study

Protocol Version 3.0 onwards included additional requirement for lymphatic RT. Typical examples of indications for lymphatic RT include (1) Patients with positive lymph nodes removed by axillary clearance who require RT to level I–III axilla and/or level IV axilla (SCF). (2) Patients with sentinel node-positive axillary disease not proceeding to axillary dissection and who require RT to level I–III axilla and/or level IV axilla (SCF). (3) Patients treated by pre-operative systemic therapy who are recommended post-operative RT to level I–III axilla and/or level IV axilla (SCF).

Eligibility criteria for the Nodal Sub-Study from protocol version 4.0 (24 February 2017) onwards are as follows:

Eligibility for PRO and photographic assessment sub-studies

For the PRO sub-study all the patients at those centres participating were eligible (and all patients in the Nodal Sub-Study). All patients who had BCS were eligible for the photographic sub-study at participating centres, with the intention that the same group of patients were included in the PRO and photographic studies.

Participating centres

All hospitals in the UK were invited to participate in the trial and were designated as treating (RT) and referring (non-radiotherapy) centres. All centres had to obtain the necessary regulatory approvals and RT centres completed a rigorous radiotherapy quality assurance (RT QA) approval process (see Chapter 3 for details of RT QA programme). A site initiation visit was performed either in person or by telephone prior to opening to recruitment. Radiotherapy centres identified eligible patients, obtained informed consent, treated patients and performed assessments according to the schedule. Non-radiotherapy hospitals performed the same activities except for the RT and acute toxicity assessments. Centre staff are listed in Table 30.

Patient information and informed consent

Eligible patients (see eligibility criteria) were identified during breast cancer multidisciplinary meetings or from clinic lists at participating sites. Patients were invited to participate in the FAST-Forward Trial during consultations in oncology clinics, where treatment options were discussed. Here the local principal investigator, coinvestigator or other trained healthcare professionals, discussed the trial with the patient and provided them with a copy of the patient information sheet. Patients were given at least 24 hours to consider participation in the trial, to discuss this with their family and friends and to ask questions of the clinical team. Patients who were willing to participate were asked to provide written informed consent to the principal investigator, co-investigator or other trained individual. All patients from the start of the trial were asked to provide consent for the tissue and blood sub-studies. Other optional sub-studies were included in subsequent versions of the consent form.

There were two consent forms available at the start of the trial:

1. For centres participating in the acute toxicity sub-studies (seven centres).

2. For centres participating in the PROMS and photographic sub-studies.

There were two consent forms available from Protocol v2.1:

1. For centres opting to participate in the PRO and photographic assessment sub-studies.

2. For all other centres.

There was only one consent form from Protocol v3.0 (inclusion of Nodal Sub-Study) as the PRO assessments were mandatory, since the primary endpoint of this sub-study was patient-assessed arm/hand swelling.

Randomisation

Patients in the Main Trial and Nodal Sub-Study were allocated to 40.05 Gy in 15 fractions, 27.0 Gy in 5 fractions or 26.0 Gy in 5 fractions in a 1 : 1 : 1 ratio. The 27.0 Gy group in the Nodal Sub-Study was closed to recruitment from Protocol v5.0 (14 December 2017) onwards (as described in Protocol amendments and Study Conduct sections). Treatment allocation was not blinded to patients or clinicians. Randomisation was performed by recruiting centres contacting The Institute of Cancer Research-Clinical Trials and Statistics Unit (ICR-CTSU) by telephone or fax. The randomisation allocation method was computer-generated random permuted blocks [mixed block sizes 6 and 9 (changed to 4 and 6 following the amendment of Nodal Sub-Study to 2-group design) to avoid predictability]. Patients in the Main Trial were stratified according to RT centre and risk group (high: < 50 years or grade 3 vs. low: ≥ 50 years and grade 1 or 2). Patients in the Nodal Sub-Study were stratified according to RT centre and whether or not the patient had a level II/III axillary clearance. A unique trial identifier was assigned to each patient comprising components for centre number, patient number and designated stratification.

Data collection

All participating centres were supplied with NCR case report form (CRF) booklets and paper copies of the baseline PRO booklets. Site staff were trained in trial procedures during the site initiation visit which was performed either in person or by telephone. Data were collected according to the schedule in Table 1 and used to populate the CRFs. All CRFs and baseline PRO booklets were returned to the ICR-CTSU where they were logged into a tracking database. The clinical data were entered into a clinical database (MACRO).

Data queries

Visual checks were performed on all CRFs received at ICR-CTSU. Consistency checks of the data were written into the clinical database to identify any discrepancies in the data entered. Data queries were raised for these anomalies, discrepancies and omissions and were sent to centres on a regular basis. Queries were resolved by centres and returned to ICR-CTSU.

Central statistical data monitoring

Details of data monitoring were described in the FAST-Forward Central Statistical Data Monitoring Plan (v1.0 May 2018). Checks of the data stored on the MACRO database were performed regularly and on an ongoing basis, including prior to each analysis for an IDMC report, and queries raised with the study sites. In addition, a formal comparison between FAST-Forward CRFs received (clinical CRFs and PRO booklets) and data entered onto the MACRO database was done for 5% of Main Trial patients for all CRFs from baseline up to and including year 5 follow-up. The error rate found in this data quality control exercise was very low (0.2% of items checked).

Site monitoring

The quality of the data received from sites was assessed on a regular basis by the number of data queries generated per centre. Early intervention with the centre was initiated to discuss the queries and to avoid repeat of the errors. Central statistical monitoring of the data was also performed on a regular basis as described above. No centres were identified as needing an on-site monitoring visit.

Data assumptions/coding

Adverse effects reported under ‘other’ on the annual follow-up forms were reviewed (blind to randomised treatment schedule) by the Chief Investigator and coded according to whether they were likely to be radiotherapy-related. Prior to the Main Trial primary analysis all reported local and regional relapses were reviewed (blind to randomised treatment schedule) along with pathology reports to check coding of the relapses, including location and whether invasive or DCIS.

Efficacy endpoints

The protocol specified that ipsilateral tumour relapse and contralateral primary tumour must be confirmed by cytological/histological assessment. Distant relapse was to be determined by an appropriate combination of clinical, haematological, imaging and pathological assessment, recognising that pathological confirmation was not always possible. Patients remained evaluable for local relapse after distant relapse. Patients without an event were censored at the last follow-up assessment or death. Patients were to have annual clinical assessments for 10 years and annual mammograms for 5 years or until screening age if younger (as per NICE guidelines). 50

Endpoints of late adverse effects of radiotherapy

Patient questionnaires were to be completed at baseline, 3, 6, 12, 24 and 60 months after randomisation. Clinical assessments of radiotherapy-related late AE were carried out annually from the date of randomisation up to at least 5 years.

Sample size

Statistical methods

Patient and public involvement

FAST-Forward was part of a broader portfolio of patient-centred ICR-CTSU and National Cancer Research Institute (NCRI) Breast Group trials where patient representatives with lived experience of breast cancer have been partners from concept, formalising their role as members of the protocol working group and Trial Management Group (TMG). In FAST-Forward patient representatives helped shape the trial with active involvement from trial concept, through recruitment, reviewing trial documents, discussing trial results prior to publication and helping to write lay summaries of the results for trial participants. A link to the lay summary of the Main Trial results that was distributed to trial participants following publication of the primary analysis in 2020 can be found on the ICR-CTSU FAST-Forward webpage (www.icr.ac.uk/fastforward). Our patient and public involvement (PPI) partnerships enabled design of an efficient and cost-effective follow-up schedule that truly reflects the patient experience with minimal burden and is patient-centred.

In addition, ongoing links have been forged with local (Royal Marsden) and national (Independent Cancer Patients’ Voice, UK Breast Intergroup, NCRI) patient advocate groups, embedding patient representation in the whole clinical trial lifecycle. PPI work in breast RT trials co-ordinated by ICR-CTSU has included development of alternative patient information formats, focus groups where patients emphasised the need for information on risks/benefits of RT at all stages of treatment and recovery, and future plans to collaborate within a PhD project optimising data capture of AE from trial participants.

The TMG had preferentially at least two patient representatives which enabled mutual support and empowerment. All TMG members were included in any email discussions, encouraging PPI input at all times. A mentoring programme for new TMG members was also initiated which included conversations with the Chief Investigator and PPI lead at ICR-CTSU.

Trial oversight committees

The TMG, TSC and IDMC each worked to a charter issued by ICR-CTSU and met annually until the time of Main Trial publication, with the TMG and TSC continuing to meet thereafter.

Pre-trial QA

1. A facility questionnaire and process document collating details of techniques, equipment and processes to be used within the centre for FAST-Forward patients.

2. A dummy-run (centre-selected sample patient) plan containing target and organ-at-risk (OAR) structures outlined by the principal investigator and used to create a clinical plan, was assessed to ensure adherence to the trial protocol.

3. Once the nodal study was introduced through Protocol v3.0, each centre was required to undergo credentialing for lymphatic nodal outlining and planning through completion of an outlining and a planning benchmark case (RTTQA Group provided cases). Centres were also asked to submit an amended process document detailing additional information in relation to treatment of patients randomised to the nodal study.

A visit by the RTTQA Group was also performed prior to centre participation to validate, independently, the technique used against the information given in the process document. In particular, the following parameters were assessed:

1. Target volume and treatment technique used.

2. Confirmation of plan modulation – implementation, for example IMRT forward planning.

3. Planning of dose distribution across the treatment volume for homogeneity and prescription points.

4. Assessment of routine quality checks (QC) performed by the centre compared with current IPEM guidelines. 57

5. Dose measurements across the treatment volume within a purpose-made phantom, if not performed for the same technique within the last 3 years.

6. Assessment of imaging verification protocol and technique.

On-trial QA

1. A prospective review of the outlines and RT plan was performed for the first breast and first chest wall patients recruited from each centre prior to the patient starting treatment. On introduction of the nodal study a prospective review of the nodal CTV outlines and RT plan was conducted for the first patient recruited at each centre. Feedback was provided to the investigator site within 48 hours of receipt of data. If any major protocol variations were identified, additional reviews were requested.

2. All patients’ RT planning CT, plan file, dose file, structure set and plan assessment forms were anonymised by the centre and submitted electronically to the RTTQA Group.

3. Retrospective outlining and planning reviews were performed to monitor continued protocol compliance. Any protocol variations were reported and discussed with the Chief Investigator (CI) and ICR-CTSU.

Patients and centres

The acute toxicity sub-studies were conducted in a subset of centres (Royal Marsden Hospital, Royal Cornwall Hospital, University Hospitals of North Midlands, Derriford Hospital Plymouth, Norfolk and Norwich University Hospitals and Addenbrookes Hospital) that had the infrastructure necessary to carry out weekly toxicity assessments. The same centres participated in both sub-studies, with the addition of Torbay Hospital in the second sub-study. Patients receiving a RT tumour bed boost were eligible for the first acute toxicity sub-study but were excluded from the second acute toxicity sub-study since the objective of this sub-study was to quantify the toxicity of 5-fraction schedules relative to control, effects that are independent and additive to those of the boost. At centres participating in the acute toxicity sub-studies, consecutive eligible patients were invited to participate at the time of consent to the Main Trial.

Assessments

Acute reactions of the skin of the treated breast were graded using RTOG criteria for the first sub-study (Protocol version 1.0). Following discussions with both the FAST-Forward IDMC and TSC, it was agreed that this may not be the most appropriate scoring system in this context due to the inclusion of oedema in the RTOG system, since the primary concern was to score moist desquamation. Therefore, it was agreed that a second sub-study would be undertaken using standard common toxicity criteria for adverse effects (CTCAE) criteria (v4.03) (Protocol versions 2.1 and 2.2). Toxicity assessments were made by a healthcare professional at each participating centre using predesigned forms. The assessments were scheduled to be carried out weekly during treatment and for 4 weeks following the end of RT. In the second sub-study, if moist desquamation beyond skin folds or creases (CTCAE grade 3) were seen during this time, weekly assessments were to continue until the reaction resolved to CTCAE grade 1 or less. If any assessment was missed, the centre contacted the patient by telephone to assess and grade acute skin reactions by asking the patient to describe the skin appearance and supplementing this by direct questions. In the first acute toxicity sub-study patient self-assessments of acute AE (breast soreness, reddening, swelling and blistering) were recorded using patient diary cards completed weekly during RT and for at least 4 weeks post-RT. The scores were recorded as ‘none’, ‘a little’, ‘quite a bit’ or ‘very much’. If symptoms persisted, patients were asked to continue scoring their AE on a weekly basis until all scores were graded as ‘none’ or ‘a little’. The submitted patient diaries showed low data completion to the extent that the results were not considered of value to be included in the submitted manuscript.

Acute skin reactions scoring scales

Endpoints

The primary endpoint for each of the sub-studies was the proportion of patients within each treatment group with grade ≥ 3 toxicity (RTOG and CTCAE, respectively) at any time from the start of RT to 4 weeks after completion of RT. The secondary endpoints were clinical assessment of (1) any acute skin toxicity, defined as the worst grade reported from the start of RT to 4 weeks post-RT, (2) acute skin toxicity during RT, defined as the worst grade reported from the start to the end of RT, (3) acute skin toxicity post-RT, defined as the worst grade reported at completion of RT until at least 4 weeks post-RT and (4) adherence to the acute toxicity assessments.

Statistical considerations

Each sub-study aimed to recruit approximately 150 evaluable patients (50 per group) in order to exclude a within-group rate of grade ≥ 3 acute skin reactions of over 11% compared to target rate of under 3% (89% power and one-sided 7.9% significance level). The sample size was informed by data from the FAST pilot trial. 26

Principal analyses were based on the evaluable population, defined as all patients randomised into the study receiving at least one fraction of RT (regardless of whether they were later found to be ineligible or a protocol violator) and with complete data, or, at most, one missing toxicity assessment. For the first sub-study which included some patients receiving a boost, only non-boost assessments were used to define inclusion in the evaluable population. The required number of assessments for a 40 Gy/15 fractions patient was seven (three during and four post-RT) and five for 27 Gy/5 fractions and 26 Gy/5 fractions patients (one during and four post-RT). Given the focus of the sub-studies on safety patients who switched to receive a different trial treatment after randomisation were analysed according to treatment received rather than randomised treatment.

For the primary endpoint, the proportion of patients within each treatment group with grade ≥ 3 RTOG toxicity (first sub-study) or grade ≥ 3 CTCAE toxicity (second sub-study) was estimated with associated upper one-sided 95% CI. Secondary endpoints were estimated as frequencies and percentages, with the prevalence of grades ≥ 1, ≥ 2 and ≥ 3 toxicity at each time point presented graphically. Adherence to toxicity assessments was estimated as frequencies and percentages. Toxicity data were included as reported, and there were no restrictions based on the date of assessment. The acute toxicity sub-study was designed to be non-comparative so no statistical comparisons were made across the treatment groups.

Clinician assessments

At least one annual clinical assessment of NTE was available for 3975/4096 (97%) patients. Frequencies of clinician-assessed NTE according to year of follow-up (from 1 to 4 years) and fractionation schedule are shown in Appendix 2, Table 41. Bar charts of clinician-assessed NTE from 1 to 5 years are presented in Appendix 3, Figures 27–33. Cross-sectional analyses of clinician-assessed NTE at 5 years are in Table 8. At 5 years, any moderate/marked clinician-assessed NTE in the breast/chest wall was reported for 98/986 (10%) patients in the 40 Gy group, 155/1005 (15%) for 27 Gy and 121/1020 (12%) for 26 Gy (see Table 8, Figure 11), with a statistically significant difference between 40 Gy and 27 Gy (p = 0.0003) but not between 40 Gy and 26 Gy (p = 0.17). Breast shrinkage was the most prevalent moderate/marked effect at 5 years, reported in 50/916 (6%) for 40 Gy, 78/948 (8%) for 27 Gy and 65/954 (7%) for 26 Gy (see Table 8). Longitudinal analysis of all annual clinical assessments of NTE over follow-up showed a statistically significant increased risk of any moderate/marked effect in the breast/chest wall for the 27 Gy group compared with 40 Gy (OR 1.55, 1.32 to 1.83; p < 0.0001), with no statistically significant difference between 26 Gy and 40 Gy (OR 1.12, 0.94 to 1.34; p = 0.20; see Table 9). This pattern was similar for the individual effects of breast distortion, shrinkage, induration and breast/chest wall oedema, with statistically significant higher risk for 27 Gy compared with 40 Gy but not for 26 Gy (see Table 9, Appendix 3, Figures 34–40). Comparing the two 5-fraction schedules, 26 Gy had statistically significantly lower risk of any moderate/marked breast/chest wall NTE (p = 0.0001) and breast shrinkage (p = 0.0018) compared with 27 Gy. Estimates of 5-year cumulative incidence of any moderate/marked clinician-assessed NTE in the breast/chest wall were 26.8% (95% CI 24.4 to 29.4) for 40 Gy, 35.1% (32.4 to 37.9) for 27 Gy and 28.5% (26.0 to 31.1) for 26 Gy (see Table 10). Results for comparison of schedules from the analyses of time to first moderate/marked effect were similar to those from the longitudinal modelling of all annual clinical assessments.

FIGURE 11.

Kaplan–Meier plot of any moderate/marked clinician-assessed NTE in breast/chest wall (Main Trial).

Patient self-assessments

One thousand seven hundred and ninety-six patients consented to the PRO sub-study, of whom 10 withdrew consent immediately after randomisation and eight were not given the baseline booklet. Questionnaires returned from those expected (patients alive and well, not withdrawn) totalled 1771/1778 (99%) at baseline, 1668/1733 (96%) at 3 months, 1622/1722 (94%) at 6 months, 1599/1707 (94%) at 1 year, 1531/1669 (92%) at 2 years and 1334/1589 (84%) at 5 years. Of the 1774 patients with at least one completed questionnaire, 1634 had BCS and 140 mastectomy.

Frequencies of patient-assessed breast/chest wall and arm/shoulder/hand symptoms are shown for each time point from baseline to 2 years and according to fractionation schedule in Appendix 2, Table 41. Cross-sectional analyses of patient-assessed symptoms at 5 years are in Table 11. Bar charts of patient-assessed late AE up to 5 years are presented in Appendix 3, Figures 41–51.

Change in breast appearance had the highest 5-year prevalence, with moderate/marked change reported in 140/432 (32%) for 40 Gy, 158/440 (36%) for 27 Gy and 136/429 (32%) for 26 Gy. There were no statistically significant differences in 5-year prevalence of patient-reported AE between the schedules (see Table 11). There was some evidence of an increase in patient-reported moderate/marked breast hardness/firmness at 5 years for 27 Gy compared with 40 Gy and more breast swelling in both 5-fraction schedules, but these were not statistically significant at the pre-specified cut-off of p = 0.005. Longitudinal analyses of all patient assessments from baseline to 5 years showed a statistically significantly higher risk of moderate/marked breast hardness/firmness for 27 Gy compared with 40 Gy (OR 1.42, 1.17 to 1.72; p = 0.0003), and less change in breast appearance for 26 Gy compared with 27 Gy (p = 0.0018), but no statistically significant differences between schedules for the other NTE (see Table 12).

Photographic assessments of late adverse effects in the Main Trial

Of the 1737 patients (BCS and post-mastectomy) who consented to the photographic sub-study, baseline photographs were received for 1634 (94%), and 2- and/or 5-year photographs were available for 1385 (80%). The vast majority (1309) were patients who had BCS; for these patients, 2- and 5-year photographs were assessed in 1267 and 875, respectively (see Table 13). A total of 226 patients died or withdrew from the photographic sub-study by year 5, for the remainder the most common reasons for photographs not being taken were appointments not made due to clerical errors at the centres, patients not attending clinic visits, and patients withdrawing consent from the sub-study.

At 2 years, mild/marked change in photographic breast appearance was reported in 35/411 (8%) for 40 Gy, 67/429 (16%) for 27 Gy and 46/427 (11%) for 26 Gy; corresponding figures at 5 years were 34/283 (12%) for 40 Gy, 83/308 (27%) for 27 Gy and 37/284 (13%) for 26 Gy (see Table 13). Modelling 2- and 5-year photographic assessments together, 27 Gy had a statistically significantly increased risk of mild/marked change in breast appearance compared with 40 Gy (OR 2.29, 1.60 to 3.27; p < 0.0001), with no statistically significant difference between 26 Gy and 40 Gy (OR 1.26, 0.85 to 1.86; p = 0.24; see Table 13). Twenty-six Gy had a statistically significantly lower risk of change in photographic breast appearance compared with 27 Gy (p = 0.0006).

At 2 years, mild/marked retraction/distortion was reported in 14/411 (3.4%) for 40 Gy, 27/429 (6.3%) for 27 Gy and 14/427 (3.3%) for 26 Gy; corresponding figures at 5 years were 3/283 (1.1%) for 40 Gy, 25/308 (8.1%) for 27 Gy and 14/284 (4.9%) for 26 Gy (see Table 14). Modelling 2- and 5-year photographic assessments together, 27 Gy had a statistically significantly increased risk of mild/marked change in breast appearance compared with 40 Gy (OR 2.83, 1.50 to 5.34; p = 0.001), with no statistically significant difference between 26 Gy and 40 Gy (OR 1.59, 0.79 to 3.18; p = 0.190; see Table 14). There was no statistically significantly difference in risk of retraction/distortion for 26 Gy compared with 27 Gy (p = 0.056).

Model structure and patient population

The target population was UK adults who have undergone breast-conserving surgery or mastectomy for early breast cancer (stage I/II/IIIa). The patient population for the model matched the inclusion criteria for FAST-Forward as described in Chapter 2. This included individuals with tumour grades 1–3, ER positivity and negativity, HER-2 positivity and negativity and those with or without regional lymph node metastasis.

A Markov model was built to estimate the total QALYs and costs for each treatment alternative. The structure of the economic model is illustrated in Figure 16. Circular arrows indicate the possibility of staying in the same state for multiple cycles, and linear arrows represent the possible transitions between states. The model structure was developed in discussion with clinical experts. A single health state was used to capture locoregional relapse, which includes local relapse and regional relapse as these relapses have similar prognostic and cost implications. The Markov model cycle length was 1 year as outcomes occur over a long period. A half-cycle correction was applied.

FIGURE 16.

Markov model structure of early breast cancer.

Patients entered the model following local tumour excision and received RT in either 15 fractions or 5 fractions. The RT schedules differed in terms of resource use and the likelihood of inducing acute skin reactions. Following RT, patients began in the ‘alive and disease-free’ health state and over time were at risk of locoregional relapse, distant (metastatic) relapse or death. Radiotherapy modality determined how patients move through states over time.

Radiotherapy treatment period

This section concerns costs of delivering RT and the costs of managing acute side effects. Patients were assumed to receive treatment as specified in the FAST-Forward protocol. 2 Radiotherapy costs were applied at model entry and are calculated as shown in Table 19. As RT delivery resource activity was not recorded in the trials, resource use was informed by expert opinion. A proportion of patients are instructed to follow breath hold techniques while receiving RT in order to protect the heart. This extends the duration of the RT appointment, and we estimated the consequent increase in cost.

We included costs relating to treating acute adverse skin reactions during treatment. These were assumed to be one-off costs which depend on the severity of AE as measured by worst RTOG score observed during treatment. 3 Preference-based measures of health-related quality of life (HRQoL) were collected at baseline and at 3 months, but not during receipt of RT when acute skin reactions would be present. Literature review did not identify any studies that linked HRQoL measures to RTOG scores. Quality of life was assumed to be the same across 5 and 15 fractions during the treatment period due to an absence of preference-based quality of life data for this period. This assumption was viewed to be conservative towards the 26 Gy schedule given the numerically lower number of events compared to 40 Gy.

Post radiotherapy period

Time to locoregional relapse, distant relapse and some maintenance costs and HRQoL parameters were estimated from FAST-Forward individual patient data. Stata 16 was used to carry out all individual patient data analysis. 75 The remaining model parameters were informed by searching the wider literature. All model inputs for this period are presented in Table 20.

Transition probabilities

Treatment effects

To model the rates of locoregional relapse for 5 fractions relative to 15 fractions, we applied the HR for locoregional relapse estimated in FF to the baseline rate of locoregional relapse estimated in the survival model. For both 5 fractions and 15 fractions, a common rate of transition from alive and disease free to distant relapse was assumed. This assumption was based on the clinical rationale that RT is a local treatment and therefore its causal impact on developing distant relapse (if any) can only occur through reducing locoregional relapses. The impact of this assumption was explored in a sensitivity analysis. For the base case, treatment effects were assumed to persist over time.

Costs

Costs for the remaining model states

These costs were sourced from the wider literature as there were insufficient observations to estimate them from FAST-Forward. Locoregional relapse was associated with one-off mastectomy costs in addition to supportive care costs in the first year of the event. 50,82–84 Following this first year supportive care was assumed to consist of one GP visit and one mammogram per year. 50 A UK study was used to estimate supportive care and treatment costs for distant relapse. 85

To make them consistent with the inclusion of related and unrelated costs in the FAST-Forward resource use questionnaire, costs unrelated to breast cancer were added to the breast cancer costs for locoregional and distant relapse health states. 86

Health-related quality of life

Uncertainty analyses

The distributions described in Table 20 were applied to each of the parameters to represent uncertainty in the model. A probabilistic sensitivity analysis (PSA) was used to reflect the impact of this joint uncertainty on outcomes. This involved taking 10,000 draws from the distributions and calculating model outcomes for each iteration. One-way sensitivity analysis was used to explore the sensitivity of results to one-at-a-time changes in individual inputs and assumptions, see Table 21.

Subgroups

For the low-risk population (subgroup 1), 5 fraction therapy was expected to dominate with expected cost savings of £1881 (95% CI £1252 to £2648) and higher expected QALYs: 0.03 (95% CI −0.01 to 0.07). Across simulations there was a 99.9% chance that 5 fractions either dominated WB15F or had an ICER below £15,000/QALY.

For the high-risk population (subgroup 2), 5 fractions therapy was expected to dominate with expected cost savings of £2102 (95% CI £1230 to £3093) and higher expected QALYs: 0.05 (95% CI −0.01 to 0.11). Across simulations there was a 99.9% chance that 5 fractions either dominated WB15F or had an ICER below £15,000/QALY.

Sensitivity analysis

Based on expected outcomes, 5 fractions dominated all other options except when using the distant relapse HR results estimated in the trial (see Table 21). In this one scenario, 5 fractions compared with 15 fractions was expected to be less expensive with incremental costs of −£908 (95% CI −£2689 to £975) but less effective, with −0.14 incremental QALYs (95% CI −0.43 to 0.12). This placed the expected ICER in the southwest quadrant. With a threshold of £15,000/QALY 5 fractions was expected to have 33.8% chance of being cost-effective in this scenario.

Symptoms in arm, shoulder or hand

Frequencies of PRO relating to arm/shoulder/hand symptoms from baseline to 24 months are shown in Table 24 for 40 Gy compared with 26 Gy (all randomised patients), and in Table 25 for 40 Gy and 27 Gy patients randomised up to the end of 2017. For all outcomes relating to arm, shoulder or hand, the number of patients reporting marked effects was small at all time points in all treatment groups. Appendix 2, Table 55 shows baseline arm/shoulder/hand symptoms according to maximal extent of axillary surgery and fractionation schedule.