Participants

Men with prostate cancer undergoing ADT (orchiectomy, luteinising hormone-releasing hormone agonists or antagonists) are invited to participate in the exercise programme when attending their ambulant prostate cancer controls at the Department of Urology, Herlev and Gentofte University Hospital, Denmark. ADT must be continued for at least 1 year forward. Thus, the cohort consists of patients receiving ADT as an adjuvant treatment in combination with external beam radiation with curative intent and patients with advanced metastatic prostate cancer. Patients undergoing chemotherapy or second-line hormonal therapy with enzalutamide or abiraterone in addition to ADT are also eligible for participation. Exclusion criteria are opioid-demanding treatment for skeletal pain, Eastern Cooperative Oncology Group (ECOG) performance status above 2 or the inability to perform floor and machine exercises independently. Patients with osteoporosis or a larger bone metastatic burden assessed by the referring physician undergo a modified resistance training programme. Prospective data collection began in August 2014 and is expected to continue until April 2017. The number of participants, who have given consent to the use of their healthcare data, was 169 in December 2015.

Intervention

The programme is a standard of care (implemented at our facility in December 2013) being offered to all our patients receiving ADT who meet the criteria for participation. The intervention consists of an educational session followed by 12 weeks of supervised group-based exercise, ending with a structured out-of-hospital transition to continued exercise at a local training facility.

The educational session lasts 1½ hours, with lectures from a urologist, nurse, dietitian and a physiotherapist specially assigned to this task. Patients are invited to bring a person of confidence and to attend with the group they are going to start the training programme with. During the educational session, the natural history of prostate cancer is covered as well as the rationale, indications and methods of action for ADT. Furthermore, known physical and psychological side effects of ADT, including the implications of decreased libido and risks of osteoporosis, diabetes and potential CVD, are addressed. The dietitian lectures on the latest dietary recommendations according to the Danish Health Authorities.

The following exercise programme consists of 1½ hours of supervised group-based training two times a week for 12 weeks combining aerobic and resistance training. This exercise routine has shown positive effects on physical fitness, body composition and QoL in several randomised clinical trials.15–17 ,19 ,21 The main focus of this programme is progressive resistance training using machines, following the guidelines for progression recommended by the American College of Sports Medicine (ACSM).22 Each session is started with 15 min of warm-up on a bike or stair walking followed by 60 min of core, upper and lower body resistance training using machines. The programme consists of seven exercises targeting the major muscle groups of the body—leg press, knee extensions, hamstring curls, chest press, low row, sit-ups and back extension exercises. The progression model outlined in table 1 is used as a general guideline to the individualised training progression. The resistance training programme is supplemented by aerobic exercise—∼6 min of exercise on a treadmill and 500 m of ergometer rowing. The resistance training is tailored to the individual participant by the physiotherapist to accommodate both his strength level and potential hazards such as osteoporosis and major regions of metastatic disease. Close attention is paid if the physician has highlighted training restrictions in the patient file due to comorbidity. In the presence of multiple bone metastases or osteoporosis, a symptom-dependent and very conservative approach to the progression of the resistance training programme is adopted—that is, very small increments in resistance. In addition to the supervised exercise, participants are encouraged to perform one exercise session at home each week according to a given programme provided in the patient educational material; furthermore, both the participants and, if possible, their partners are instructed to stay generally physical active, defined as daily non-strenuous activity of 10 000 paces.

The main behaviour change techniques according to the the ‘Coventry, Aberdeen, and London—Refined’ (CALO-RE) taxonomy23 ,24 used in this programme are: ‘provide information on consequences of behaviour in general’, ‘goal setting (behaviour)’, ‘prompt practice’, ‘prompting generalisation of a target behaviour’, ‘provide instruction on how to perform the behaviour’, ‘model/demonstrate the behaviour’ and ‘plan social support/social change’. The supervised group-based training allows for verbal and visual instruction on how to exercise and creates a platform where social support and bonds between the participants are created. Furthermore, the last two supervised training sessions are performed at a local commercial fitness centre to familiarise participants with available facilities in the community and to make a structured transition to out-of-hospital training. Participants are encouraged to stay in contact with one another and to continue the group-based training. For motivational purposes and to monitor the effect on physical fitness, the participants are tested at baseline and after 12 and 24 weeks. Concurrently, the participants are seen by a nurse to monitor changes in body composition as well as to address any concerns or questions the participant might have regarding both his cancer treatment and the exercise programme.

Measurements

All measurements of primary and secondary outcomes are validated and widely used in clinical research. All participants will be assessed at baseline, postintervention and at the 24-week follow-up by specially trained nurses and physiotherapists. A loss to follow-up of ∼10% at 24 weeks is expected primarily due to symptomatic disease progression. Handling of missing data is covered under sample size and statistical analysis.

Primary study endpoints

The primary endpoints of this study are a change in workload (Watt) during the Graded Cycling Test with Talk Test (GCT-TT) and a change in the number of repetitions using the 30 s Chair-Stand Test (30s-CST). Both endpoints are measured as the within-person change from baseline. GCT-TT is a submaximal exercise test used to assess aerobic fitness. It is not dependent on the measurement of heart rate and is thus not influenced by medications or comorbidities potentially affecting the heart rate variability. Excellent test/retest reliability as well as an acceptable measurement error and good responsiveness have been observed in cardiac patients.25 ,26 Although not directly validated in patients with prostate cancer, the GCT-TT is expected to be feasible and reliable in this population because the patients are expected to resemble the populations of cardiac patients regarding age and comorbidity. The GCT-TT is performed on a cycle ergometer as described by Nielsen et al.25 A constant pedalling rate of 60 rounds per minute is maintained throughout the test. After an initial 2-min warm-up period with the exercise intensity set to 15 W, a ramp protocol is applied with 15 W increases every minute. During the last 10 s of each exercise step of the test, a standard, 30-word text passage is recited by the patients. Subsequently, they are asked, ‘Are you still able to speak comfortably?’. The following predefined answers can be given: ‘yes’, ‘unsure’ or ‘no’, corresponding to a positive talk test (TT+), equivocal talk test (TT±) and negative talk test (TT−). When the answer ‘no’ is given, the test ends and the exercise intensity is recorded.

The 30s-CST is used to assess functional performance/lower extremity muscle strength and endurance.27 The 30s-CST was chosen because it has previously been used in patients with cancer undergoing a similar exercise intervention mainly consisting of progressive resistance training.28 The improvement in the 30s-CST result (≈15–16%) corresponded well with the improvement in isokinetic knee extension strength (≈12–18%) measured by an isokinetic dynamometer.28 The 30s-CST is performed according to the test manual described by Jones et al,27 as follows: a chair with a seat height of 43 cm is used. With arms crossed over the chest as many rises as possible is performed in 30 s. The starting position is sitting on the chair. A complete rise is recorded when the knees and hips are fully extended in a standing position and the next rise can begin when the patient has assumed a sitting position on the chair seat without touching the back rest. The command ‘ready, set, go’ is used. The clock is started on ‘go’, and the number of complete rises in 30 s is counted. The modified version of the 30s-CST, in which the patients actively use the armrests, is used in patients who are unable to stand up only using the legs.29

Secondary study endpoints

Questionnaire-based assessments are conducted using the validated European Organisation for Research and Treatment of Cancer Quality of Life Core Questionnaire (EORTC QLQC-30 V.3.0) and Quality of Life Prostate-specific Questionnaire (EORTC QLQ-PR25).30 ,31 Specific endpoints of the EORTC QLQC-30 questionnaire are changed from baseline in global health status, functional scales (physical functioning, role functioning, emotional functioning, cognitive functioning and social functioning) and fatigue measured by the symptom scale. The EORTC QLQ-PR25 questionnaire is used to measure changes from baseline in sexual activity and sexual functioning. The latter will only be analysed for patients who are sexually active at baseline. Furthermore, the change in ADT-related symptoms is assessed by the hormonal treatment-related symptoms scale of the EORTC QLQ-PR25.

To examine changes in body composition, nurses measure the body mass index (BMI, kg/m²) and the waist circumference (WC) and hip circumference (HC) in centimeters. A secondary end point is the change from baseline in BMI, WC, HC and waist–hip ratio. These changes do not reflect the benefits of exercise on muscle mass. However, waist–hip ratio and WC are used to assess change in fat distribution and abdominal adipose tissue, which are risk factors of the metabolic syndrome, diabetes mellitus and CVD.32 WC is measured according to the WHO guidelines midway between the inferior margin of the last rib and the crest of the ilium, in the horizontal plane.33 Similarly, HC is measured around the buttocks in the horizontal plane at the level of maximum extension.

Blood pressure (systolic and diastolic) is measured at each visit after a 10 min seated rest. Changes in systolic and diastolic blood pressure from baseline will be investigated separately.

Safety of exercise is measured by the prospective recording of any serious adverse events in relation to training. A serious related adverse event is defined in accordance with the European Medicine Agency (EMA) as any undesirable experience associated with exercise that results in death, is life-threatening, requires hospitalisation or prolongation of existing hospitalisation, results in persistent or significant disability or results in some other medically important condition. Minor adverse events such as muscle strains or muscle soreness after exercise will not be reported.

Compliance with the programme will be reported as attendance at the supervised group training sessions. The ability of participants to perform the exercises as prescribed, that is correct intensity and number of repetitions, will be reported as percentage of exercises during attended sessions completed as prescribed. Patient self-reports, completed during each session, on intensity and the number of repetitions done for each exercise will be used for this analysis. Furthermore, continued exercise after the ended intervention will be reported as either yes or no and also relies on patient self-reports.

Sample size and statistical analysis

An inclusion cut-off date has been set to 15 October 2016 with expectations of having included ∼230 participants. Previous larger randomised trials investigating the effects of exercise in men with prostate cancer have included ∼100 participants.16 ,21 ,34 As we mean to evaluate the safety of exercise and there is no control arm, we have chosen a substantially larger sample size. Minimal sample size calculations to detect a clinically significant change have been carried out for the primary endpoints. Data from an in-house project have shown clinical significance within-person change in the GCT-TT at 13 W and the SD of the change at 18 W.26 As we expect to perform multiple tests, the α-level in the sample size calculation is set at 0.005, two-sided, to allow correction for multiple testing. With a power at 0.80, a minimal sample size of 30 participants is required. For the 30s-CST, the clinically significant change was set at 2.6 repetitions with an SD at 2.2 in accordance with a previous study by Wright et al.35 With a power at 0.80 and a significance level at 0.005, a sample size of 13 is necessary.

Linear mixed-effect models or Friedman tests will be used for analyses of data with multiple time points. Clinically relevant covariates such as age, exercise compliance (high compliance is predefined as 75% attendance), continued exercise postintervention, duration of ADT, baseline BMI, treatment for castration-resistant prostate cancer other than ADT and known comorbidities (CVD, diabetes mellitus, hypertension, dyslipidaemia and osteoporosis) will be included in the analyses to assess effect modification (interaction). If a significant interaction is found we will perform subgroup analysis on the covariate. Missing data will be handled by the mixed-effect models, assuming the data are missing at random. A sensitivity analysis for the missing-at-random assumption will be performed. Secondary, and more explorative, analyses will include comparisons of the baseline to the 12-week follow-up and of the 12-week to the 24-week follow-up.

p Values will be two-sided and p values below 0.05 will be considered significant. Statistical calculations will be performed using SAS 9.4 (SAS Institute, Cary, North Carolina, USA) or a similar software. Statistical planning was conducted in cooperation with a biostatistician.