Background

Breast cancer is the most common cancer in women in the UK, with over 55,000 new cases diagnosed each year.¹ Breast cancer incidence has increased by 20% since the early 1990s.¹ Despite increasing incidence, survival rates have improved dramatically as a result of advances in early diagnosis and treatment.¹ Breast cancer survival has doubled in the UK over the last 40 years; now, nearly 8 in 10 women (78%) treated for invasive breast cancer survive for ≥ 10 years.¹ Treatments are complex and can be toxic, causing side effects that persist in the long term. There is increased recognition of the benefits of providing supportive care for people living with and beyond cancer treatment.²

Surgical treatment of breast cancer

Surgery is the mainstay of treatment for breast cancer, supplemented with chemotherapy, radiotherapy and biotherapy, with or without reconstruction surgery.³ Treatment decisions are based on clinical criteria, tumour stage, lymphatic spread and patient preference. Surgery to the breast consists of either mastectomy or breast-conserving surgery (BCS), with newer oncoplastic conserving procedures increasingly being used.⁴ Breast-conserving procedures, such as lumpectomy or wide local excision, combined with whole-breast radiotherapy, aim to achieve disease control with minimal morbidity. These breast-conserving treatments have been demonstrated to be as effective as mastectomy in increasing long-term overall survival in patients with early breast cancer.^5,6 Conservative surgery followed by radiation therapy allows for the preservation of the breast, which can improve patient quality of life (QoL) and satisfaction with treatment.⁷ Sentinel lymph node biopsy (SLNB) has largely replaced axillary lymph node dissection (ALND) for disease staging, and also reduced the need for extensive axillary node clearance (ANC). There is good evidence that 10-year survival among women receiving SLNB only is equivalent to that among women receiving ANC after SLNB.^8,9

Treatment-related side effects

Although largely curative, breast cancer treatments have negative sequelae. Surgery and radiotherapy can affect the upper body, especially the shoulder joint and upper limb, causing restricted shoulder range of movement (ROM), impaired strength and functional limitations. Arm morbidity has been strongly associated with the extent of axillary node surgery. Although arm lymphoedema can affect up to 20% of women, systematic reviews report higher rates of lymphoedema after ALND than after SLNB up to 2 years after surgery [20%, 95% confidence interval (CI) 14% to 28%, n = 18 studies, n = 3599 participants, vs. 6%, 95% CI 4% to 9%, n = 18 studies, n = 3583 participants].^10,11

A systematic review¹¹ of upper limb problems after surgery and radiotherapy (32 observational studies) reported prevalence estimates for restricted shoulder ROM (up to 67%), arm weakness (< 28%) and shoulder/arm pain (< 68%). Prevalence estimates vary widely, in part because of differences in definitions, methods of measurement and timing of postoperative follow-up. Other common postoperative complications include wound infection, seroma and axillary web syndrome (cording) and chronic pain.^11,12

A nationwide Danish study¹³ of 2500 women undergoing breast cancer surgery found that over one-third of women reported persistent pain and half reported sensory disturbances up to 7 years after treatment. Persistent upper limb dysfunction and pain are debilitating, affecting sleep quality, QoL and physical and emotional function. These enduring adverse sequelae of cancer treatment are burdensome and associated with increased health-care utilisation.

Risk factors for persistent post-treatment complications

Research has examined patient- and treatment-related risk factors associated with upper body problems after breast cancer treatment.^11,14,15 Women undergoing mastectomy have higher odds of postoperative shoulder restriction than those undergoing BCS [odds ratio (OR) 5.67, 95% CI 1.03 to 31.2].¹¹ More invasive axillary surgery is associated with greater impairments of abduction ROM and strength than SLNB, up to 7 years post treatment.¹⁶ Radiotherapy to the axilla or chest wall, compared with no radiotherapy, slightly increases the odds of shoulder ROM restriction (pooled OR 1.67, 95% CI 0.98 to 2.86) and lymphoedema (pooled OR 1.46, 95% CI 1.16 to 1.84).¹¹ A higher body mass index (BMI) was found to be an independent risk factor for shoulder external rotation problems up to 7 years after treatment.¹⁶ Higher BMI (overweight or obese) is also a known risk factor for lymphoedema¹⁰ and for development of chronic post-surgical pain (six studies, pooled OR 1.34, 95% CI 1.08 to 1.67).¹⁷

Evidence for the effect of exercise on shoulder dysfunction

A Cochrane systematic review,¹² published in 2010 [24 randomised controlled trials (RCTs), 2132 participants], reported that exercise and/or physiotherapy may help to prevent shoulder and arm morbidity after breast cancer treatment. This review¹² found that physiotherapy, compared with usual care or control, improved shoulder flexion only within the first 2 weeks and at 3 and 6 months postoperatively. The timing of starting postoperative physiotherapy may also be important for shoulder ROM and upper limb function. Early exercise, started on the first postoperative day, was beneficial in improving flexion and abduction at 1 week postoperatively, and flexion at 4–6 weeks postoperatively, when compared with delayed exercise (exercise that started after the fourth postoperative day).¹²

A more recent systematic review,¹⁴ published in 2015 (18 RCTs, 2389 participants), compared different exercise modalities (multifactorial therapy, passive mobilisations, stretching and exercise therapy) and the timing of application.¹⁴ The overall findings were similar, suggesting that early exercise improved upper arm ROM in the short and long term after breast cancer treatment. However, exercising in the first postoperative week also increased the risk of greater wound drainage volume and seroma formation.¹⁴ Regarding physiotherapy modalities, adding stretching to an exercise programme may improve postoperative ROM.¹⁴

Although these reviews suggest that physiotherapy may prevent postoperative shoulder problems, the majority of trials conducted to date are small, methodologically weak and with short-term follow-up. Many trials investigated exercise delay prescription until after completion of adjuvant therapy.¹² Few fully report details of prescribed regimes; hence there is a lack of knowledge regarding the optimum content, frequency, intensity, timing or safety of exercise prescription. Another limitation is the exclusion of patients with existing shoulder problems, the very population who may benefit the most from targeted postoperative support.¹⁸

Rationale for the PRevention Of Shoulder ProblEms tRial

We designed the PRevention Of Shoulder ProblEms tRial (PROSPER) to address the evidence gap and to investigate whether or not an early supervised exercise programme, compared with usual care, could prevent musculoskeletal shoulder conditions in patients undergoing treatment for breast cancer. This research was commissioned in 2013–14 by the National Institute for Health Research (NIHR) Health Technology Assessment (HTA) programme, with specifications to design an exercise intervention for women identified as being at higher risk of developing shoulder problems as a consequence of their breast cancer treatment. At the time of funding, the UK National Institute for Health and Care Excellence (NICE) recommended that all breast cancer patients should be provided with instructions on functional exercises to start doing from the first postoperative day. Each breast cancer centre should have written local guidelines for postoperative physiotherapy, but NICE recommended that patients be referred to physiotherapy services only if they experienced persistent shoulder restrictions after cancer treatment.³

Literature update

We reviewed literature to identify new trials investigating exercise after breast cancer surgery published since the commissioned call in 2014. We sought RCTs comparing exercise and/or physiotherapy with standard or usual care (i.e. no active intervention), regardless of the type of outcome. Our search strategies were adapted from previous systematic reviews^12,14 and applied to MEDLINE (via Ovid), EMBASE (via Ovid), PEDro (Physiotherapy Evidence Database) and LILACS (Latin American and Caribbean Health Sciences Literature). We also searched for trials registered on the World Health Organization (WHO) search portal, the European Union clinical trials register and www.clinicaltrials.gov (US National Library of Medicine). We searched for citations published from 1 January 2014 to 10 December 2019. Of 439 potentially eligible studies identified, after screening titles and abstracts, 12 trials were included, eight of which were published trials and four of which were registered as ongoing. One trial reported ROM and grip strength data across two separate publications.^19,20

Recent evidence: physiotherapy compared with usual care

Of eight published trials, four were pilot RCTs and all studies were single centred with small sample sizes (mean 79 participants), although one trial recruited 153 participants.²¹ Type of exercise varied and included aquatic-based,²² aerobic²³ and resisted exercises.^{19–21,24–26} Interventions were delivered either in the clinic setting^{19–23,25,26} or using an online interface.²⁴ Exercise programmes varied widely in terms of duration and frequency, ranging from 3 to 9 months (Table 1).^23,26 Outcomes also varied, but the most commonly reported were health-related quality of life (HRQoL), function and lymphoedema. Five studies reported improvements favouring the intervention group for the majority of outcomes (n = 427 participants). Three studies reported no differences between groups for function¹⁹ (n = 59 participants), lymphoedema²² (n = 29 participants) or limb volume²⁵ (n = 35 participants).

TABLE 1

Randomised controlled trials comparing physiotherapy with usual care published since PROSPER was commissioned

Forthcoming studies: registered trials

At the time of writing, we found four registered trials, all overdue for reporting, from Spain (n = 90 participants²⁷ and n = 84 participants²⁸), Brazil (n = 38 participants²⁹) and the USA (n = 568 participants³⁰) (Table 2). These trials have different primary outcomes and postoperative follow-up points: ROM at 1 month,²⁸ pain and fatigue after 7 weeks of exercise sessions,²⁹ functional capacity at 12 months²⁷ and presence of lymphoedema at 18 months.³⁰ The American lymphoedema trial³⁰ has provided interim data on the clinicaltrials.gov website suggesting early benefit on lymphoedema outcomes; final results are pending. We present an overview of findings regarding the content and safety of exercise interventions in Chapter 3, which describes the development of the PROSPER exercise intervention.

TABLE 2

Registered RCTs comparing physiotherapy with usual care

Aims and objectives of PROSPER

The overall aim of PROSPER was to investigate whether or not an early supervised exercise programme compared with best practice usual care was clinically effective and cost-effective for women at high risk of shoulder problems after breast cancer treatment on outcomes of upper limb function, complications and QoL.

The study objectives were to:

• develop and refine a complex intervention of physiotherapy-led exercises, incorporating behavioural strategies, for women at risk of developing musculoskeletal problems after breast cancer treatment
• assess the acceptability of the structured exercise programme and outcome measures, to optimise participant recruitment and refine trial processes during a 6-month internal pilot phase
• use findings from the internal pilot phase to undertake a definitive, full RCT in approximately 15 UK NHS breast cancer centres.

A health economic analysis and a qualitative substudy were embedded within the trial. Qualitative research was undertaken throughout to inform intervention development and gain insight into the experiences of both women and physiotherapists taking part in trial interventions.

Overview of report

The report is structured across seven subsequent chapters. We present the methods and describe intervention development and trial results, followed by separate chapters reporting the qualitative findings and the health economic evaluation. Finally, we present an overarching discussion and conclusion.