Home-based physical activity in breast cancer patients and cardiorespiratory tness: during and/or after chemotherapy? A three-arm randomized controlled trial (APAC)

Adapted physical activity (APA) program is recommended for breast cancer care. However, their modalities have not been dened. The aim of this study was to determine the best time to begin APA. This randomized controlled trial evaluated at 12 months the effect of home-based APA performed during and/or after treatment on cardiorespiratory tness. The primary endpoint was peak oxygen consumption (VO 2peak ) compared at 12 months (group A vs C and B vs C). Secondary endpoints included the 6-min walking test (6MWT), assessment of muscular strength, fatigue, quality of life, anxiety, and depression, and a questionnaire of PA levels. All tests were evaluated at baseline, 6 months, and 12 months. The combined one and aerobic per Analysis of was used for repeated or Mann–Whitney for continuous variables, and χ 2 test for binary or categorical variables.


Page 6/31
Cardiovascular training was performed on a bicycle ergometer at constant wattage. At baseline, all patients performed a cardiopulmonary exercise test to determine VO 2peak , maximum aerobic power, and heart rate at ventilatory threshold. Participants began to pedal three series of 8 min at 60% of their maximum aerobic power obtained at ventilatory threshold with 1-min rest intervals, and gradually rode 30 min continuously at 70%. The patients could also choose to perform brisk walking in addition to the bicycle. Resistance training was performed once a week on ve muscle groups, including abdominal, hamstring, quadriceps, triceps, and surae and gluteus maximus using elastic bands. Each resistance training session consisted of two sets of 8 to 12 repetitions. A total of 27 or 54 resistance sessions were performed.
All patients received the same nutrition counselling.

Study outcomes
The primary objective of the APAC trial was to evaluate the effects of the training program performed for 6 or 12 months on VO 2peak at 12 months. The VO 2peak results were compared between group A (APA 6 months during speci c treatment) versus C (APA for 12 months) and B (APA 6 months after speci c treatment) versus C (APA for 12 months).

Secondary objectives
The following secondary objectives were included: Comparison at 12 months of VO 2peak between group A and B.
Comparison at 6 months of VO 2peak between group B and A + C.
Comparison of functional capacity, muscle strength, 6-min walking test, fatigue, QOL, anxiety or depression, and anthropometric (body mass and body mass index [BMI]) measures of body composition based on impedance and PA evaluation at 6 months between group B and groups A + C and at 12 months between A versus B, B versus C, and A versus C.

Primary and secondary endpoint measures Cardiopulmonary exercise tests
To determine VO 2peak , an incremental supervised cardiopulmonary exercise test with 12-lead electrocardiogram monitoring (Corina; GE Medical Systems IT Inc., Milwaukee, WI, USA) was performed according to cardiopulmonary exercise test guidelines for clinical and cancer populations (23).

Six-minute walking test
Under the supervision of a respiratory physiologist, patients were instructed to walk as quickly as possible for up to 6 min, and the total distance was recorded. Patients were allowed to stop at any time during the 6-min test. Age-and sex-predicted 6-min walking test was calculated from Enright's Eq. (24). Body composition BMI was obtained from height and body mass using the formula BMI = mass (kg)/height² (m). Fat and lean mass were assessed using dual-energy X-ray absorptiometry.
Peripheral muscular strength Muscular strength of the quadriceps was determined using the best of three repetitions on an isometric bench with a strain gauge (Globus System). For a valid measurement, it was necessary that the patient's strength reached a plateau and was supported for at least 0.5 s. Each test was followed by 1 min of rest (three trials). Force was measured in kg.

Fatigue
Fatigue was assessed using the Multidimensional Fatigue Inventory . The MFI-20 is a 20-item questionnaire consisting of ve dimensions: general fatigue, physical fatigue, mental fatigue, reduced activity, and reduced motivation. Scores of the subscales range from 4 to 20, and a high score indicates signi cant fatigue (25).
Quality of life QOL was assessed using the EORTC QLQ-C30 (26). This questionnaire assesses ve functional scales (physical, role, cognitive, emotional and social) and nine symptoms caused by cancer or its treatments (fatigue, nausea and vomiting, pain, dyspnea, insomnia, loss of appetite, constipation, diarrhea, and nancial di culties) and a global health and quality of life scale. For functions, higher scores represent a better QOL; for symptoms, higher scores represent a worse QOL. This questionnaire has been validated for individuals with cancer (27) and more speci cally with breast cancer.

Anxiety and depression
Anxiety and depression were evaluated using the Hospital Anxiety and Depression Scale (HADS) (28), a self-administered questionnaire of 14 items rated from 0 to 3. A higher score is related to greater anxiety or depression.
Assessment of exercise performance APA program performance was monitored using two methods: Polar monitor and exercise diary.
Participants were provided with a heart rate monitor (Polar RS400SD; Polar Electro, Kempele, Finland) and asked to wear it during exercise. Results measured using the polar monitor provided a calculated metabolic equivalent of tasks (MET) estimated based on caloric expenditure (29). APA programs were considered valuable if the patient realized ≥ 85% of the scheduled sessions while wearing the Polar monitor. Participants were also asked to record in a diary the number of minutes and kilometers of exercise performed without wearing the monitor.

Results
From March 2013 to May 2015, 105 patients were eligible and 94 patients were enrolled in the APAC trial ( Fig. 1). All patients completed assessments upon admission to the study (T0). One patient in group B and one in group C refused to participate in the program after baseline assessment (T0). For primary objective analysis at 12 months, 21 patients were excluded.

Participant characteristics at baseline
Patient characteristics at baseline are presented in Table 1. The three groups were mostly homogenous for tumor characteristics and treatments. The average age for patients in the groups was 53.0 ± 8.9 years. No statistical difference was observed between the three groups. For BMI and when subcategories were Rheumatological symptoms (n = 6) 2 (6.2) 2 (6.5) 2 (6.5) Effect of home-based activity on aerobic and functional capacity At T0, VO 2peak did not differ among the three groups, with relatively low values of 20.8 ± 5, 20.9 ± 4.1, and 21.1 ± 4 mL min − 1 kg − 1 in group A, B, and C, respectively (p = 0.31) ( Table 3). Regarding criteria for maximal efforts, a plateau in VO 2peak was achieved in all participants. At T2, VO 2peak increased in the three groups without signi cant differences (B vs. C: p = 0.78, A vs. C: p = 0.64) ( Table 3). T 0 At T1 (secondary objective), the VO 2peak decreased in group B (patients under chemotherapy without an APA program) and increased in groups A and C (patients included in an APA program), but changes between the groups were not signi cant (Table 4). To study the VO 2peak in patient populations in function of PA performed, we created three subgroups whose patients participated in ≥ 85% of the sessions in each randomized group: A' (n = 25), B' (n = 23), and C' (n = 21). The mean values of the VO 2peak compared to the baseline (T2 − T0) were 0.97 ± 3.65, 0.84 ± 2.69, and 1.95 ± 2.56, respectively, for A', B', and C'(no statistical difference), with a trend towards an increase in group C', which was twice as high as group A' (Table 5).  (T1  -T0)   T0  T1  T1-T0   T0  T1  T1-T0   T0  T1  T1-T0   T0  T1  T1-T0 A + C vs B   (T2-T0)   T0  T2  T2  -T0   T0  T2  T2  -T0   T0  T2  T2  -T0   A

Adherence assessment
Aerobic exercise program ≥ 85% was performed by 91%, 80%, and 77% of patients in group A, B, and C, respectively ( Table 2). The results measured in METs were determined using the polar monitor, as described in the Methods. Adherence measured using the monitor ( Table 2) was registered in less than 50% of patients who performed ≥ 85% of the program because of di culties using the monitor; 90% of patients performed moderate PA. Quantity and duration of walking and biking were registered with diary recording independently of polar wearing and did not differ between group A and B (

Changes in the 6-min walking test
The results of the 6-min walking test are shown in Table 3. At baseline (T0), no difference was observed between the three groups (p = 0.87). At 6 months (T1), patients in group A and C signi cantly increased their distance compared to group B, who decreased their distance (p = 0.042) ( Table 6). At 12 months (T2), all groups increased their performance from baseline without a statistical difference (Table 3). In group A, patients had a delta (T1 − T0) at 17 m ± 48.9 and a lower increase but continued to increase after T1 with a delta (T2 − T1) at 5.8 m ± 32. In group B, the distance performed by patients decreased at T1 and increased after starting APA, and T2 was signi cantly higher than T1 (p < 0.0001). In group C, the increase in distance was slow until T1 (p = 0.48), and continued to increase until T2 with a signi cant change (p = 0.001), and was twice as high as the values obtained in group A and B. Table 5 shows the stability in body composition variables at 6 and 12 months across the three groups, without difference between the groups.

Quadriceps strength
No change in quadriceps strength was observed between the three groups at T2 (Table 5). At T2, the values were about the same among the three groups, but with an increase in all patients compared to values obtained at baseline in groups A (3.6 ± 7.9), B (2.6 ± 6.5), and C (4.4 ± 7.7). At 6 months, decreased strength was observed in group B, with a difference between T1 and T0 of − 0.17 ± 2.8, while there was an increase in group A (1.52 ± 3.4) and C (0.02 ± 3.2) ( Table 5).
Level of physical activity estimated from the IPAQ questionnaire The assessment of PA with the IPAQ questionnaire did not reveal a difference between the three groups in terms of classes of Met-min/w low or moderate and high PA at T2 or T1 (Table 5). More than 50% of patients were globally considered to have moderate activity, with a small percentage in high activity (from T0 to T2: 3 to 12% in group A, 6 to 13% in group B, and 6 to 4% in group C).

Quality of life, symptoms, and functions from the EORTC QLQ-C30
The EORTC QLQ-C30 results are presented in Tables 4 and 5. Global score of QOL measured by QLQC30 was stable during the protocol. No signi cant decrease was observed during chemotherapy performance.
No signi cant difference between groups was observed at the various times when the questionnaires were assessed. When the different functions were studied, at 6 months group B showed decreases in all functional areas but only emotional state showed a signi cant difference between groups A + C versus B, in favor of APA groups (p = 0.010). At 12 months, group B showed a decreased global score and functional score, but no signi cant difference was observed between groups.

Anxiety and depression from the HADS questionnaire
The results showed an overall reduced symptomatology of anxiety in all groups during the protocol and a decrease in depressive symptoms in group C, but without a signi cant difference between the groups at 6 or 12 months (Tables 4 and 5).
Fatigue measured using the MFI scale At T0, no difference was observed between the three groups with an MFI score of 59.7 ± 5.6, 60.9 ± 5.7, and 59.4 ± 5 in group A, B, and C, respectively (Tables 4 and 5). Fatigue global score decreased at T1 in group B, who did not perform APA (− 0.9 ± 6.4), but with no signi cant difference between groups A + C versus B. Fatigue global score was stable with no between-group differences at T2.

Adverse events
No grade 3 or 4 toxicity was observed in patients in relation with APA, but two types of adverse events were reported for whom it was di cult to determine their origin (cancer, chemotherapy, or APA). Fatigue was reported during the APA program in 21 patients in group A, 10 in group B, and 21 in group C. Myalgia or arthralgia was observed in ten, ve, and eight patients in group A, B, and C, respectively. More speci cally, tendinitis in two patients of group B and C and a calf snap in one patient in group C may have been associated with AP.

Discussion
The APAC study does not show different physiological functional impacts between three strategies of a home-based APA program during and/or after speci c adjuvant treatment in patients with early-stage breast cancer. The between-group difference in VO 2peak observed in the APAC study was not signi cant.
Although many APA studies involving breast cancer patients have been performed, few randomized controlled studies have been conducted during or after breast cancer chemotherapy. In addition, these studies showed that the performance of a home-based APA program increases VO 2peak , but intensity, duration, and schedule programs vary among studies. During cancer-speci c treatment, APA may increase treatment effectiveness to limit secondary effects, maintain physical tness preventing muscle loss, fat gains, and fatigue, and improve QOL (32). Exercise post-treatment aims to accelerate recovery, improve physical tness and QOL, and reduce fatigue. The aim of our study was to assess changes in VO 2peak , an indicator of cardiorespiratory tness, after an APA program started at different times in breast cancer patients.
In this trial, we compared the feasibility and bene ts of a home-based PA program during or after speci c cancer treatment to provide recommendations for patients undergoing breast cancer treatment. This AP program combined aerobic and resistance exercises, as proposed for the majority of trials for cancer patients (33). We did not compare the VO 2 values with a control group without AP because it would not have been ethical to perform breast cancer adjuvant treatment without proposing an APA program, even if the modality is not precise. Previous reports have mostly assessed the impact of AP at the end of the AP program. In our study, the rst objective was home-based exercise training impact on VO 2max evaluated at 12 months after starting the AP program. In this trial, we measured cardiorespiratory tness with VO 2max using a cycle ergometer with breath-by-breath expired gas analysis, while many studies on home-based PA apply the 6-month walking test. Secondary objectives included exhaustive assessments on physical capacity, body composition, QOL, and anxiety and depression.
Breast cancer survivors have been reported as having VO 2max values 22-25% lower compared to their age-matched healthy, sedentary non-cancer peers (11,34). Low cardiorespiratory tness is known to be inversely associated with breast cancer-related deaths, cardiovascular, and all-cause mortality (10,35). In this trial, the signi cant increase in VO 2 after APA was con rmed. However, at 12 months, APA did not increase the VO 2max differently between the three groups A, B, and C. Group A maintained at T2 the improvement in VO 2 obtained after 6 months of APA (as shown previously in SAPA trial), whereas Group B increased VO 2 after their APA program despite a decrease in T1, and this increase recovered the value obtained in group A and C.
At T1, patients received chemotherapy and radiotherapy and comparison between groups A + C versus B showed the classical decrease in VO 2 with chemotherapy alone and an increase when AP was performed concomitantly with chemotherapy. These changes were signi cantly different within each group but were not different between groups. These results support the ndings of previous studies, but the VO 2 improvement of 0.9 ± 2.7 mL min − 1 kg − 1 was lower than those obtained by Courneya et al. (36) (2.7 ± 2.6 mL min − 1 kg − 1 ) and in our previous SAPA protocol (2.26 ± 1.53 mL min − 1 kg − 1 in intention-to-treat analysis and 3.49 ± 1.64 mL min − 1 kg − 1 based on per-protocol analysis) (22). Typical curves of VO 2peak evolution during the protocol were observed, and the lack of difference among the groups may be explained based on the following hypotheses. The number of patients included in the protocol was calculated based on a planned difference in VO 2peak too high among the groups; in group C, the low adherence to APA from T0 to T1 may have been due to the high proportion of patients who were overweight or obese. The explications of the coach to patients highlighted the importance of PA and the aim of this protocol may explain why the majority of patients in group A maintained their adherence to PA after T1, contrary to published series (37,38); the majority of patients in the three groups performed exercise at moderate levels based on the Polar monitor or questionnaires, which are commonly used for home-based exercise programs. These differences showed a heterogeneity of program performance and VO 2 peak status was initially low in the majority of patients despite the average age being younger than normal in patients with breast cancer. Because we expressed the results as VO 2peak means, we cannot discuss the within-group and between-group heterogeneity of results: 66% of patients in group A showed increased VO 2 compared to 83% of group C.
The same variations were obtained in the 6-min walking tests than in VO 2peak , supporting the concordance of these tests. A signi cant difference was observed at T1 between group A and C compared to group B, with a decrease in group B during chemotherapy performed without the AP program. A continued increase in walking meters was obtained in group C from T0 and T2, even if no signi cantly different values were present at T2 compared to group A and B.
Under any APA program performed by patients that was assessed with questionnaires, patients were considered to perform a moderate level of PA during the week. However, we found no associations between post-intervention changes in VO 2max and changes in self-reported moderate to vigorous PA, revealing some limitations of these questionnaires. With a more accurate evaluation of the percentage of the program performed by patients, we found that the majority of patients performed 85% or more of the APA program, but it was di cult to measure the true expended calories because of di culties using the Polar monitor. The APA program was performed in accordance with international PA recommendations for adults (39)(40)(41).
Cancer-related fatigue has been reported in up to 90% of people with cancer during adjuvant treatment with radiation therapy, chemotherapy, and endocrine therapy (42). Meta-analysis has shown that APA has a signi cant positive effect on fatigue (43)(44)(45)(46)(47) and QOL (48). In our study, fatigue evaluated based on MFI was stable without aggravation despite chemotherapy, except in group B, in which fatigue increased during chemotherapy with no difference between groups. A positive effect on QOL without deterioration was present in the three groups but was smaller than expected. A bias in evaluation in these questionnaires highlights the meaning of the personal self-evaluation, with changes in internal standards values and conceptualization of QOL, as reported previously (49). Only emotional state was considered at T1 to differ signi cantly when patients performed APA during speci c treatments. All patients decreased their anxiety based on the HADS questionnaires, as described previously.
No change was observed in BMI and body composition based on absorptiometry. This stability was signi cant, as described previously (37), and may be explained by the absence of diet control. Muscle strength increased after APA, and a decrease was only observed in group B at T1 but reached the other groups at T2. This result was important because resistance training was not supervised and performed only once a week. The same increase was described in a study by Wanderson (50); two or three sessions were recommended in other previous reports (51) and the maintenance of muscular strength is known to have repercussions on QOL. In a study by Shilz et al. (53), 60% of breast cancer survivors suffered from signi cant decreases in muscle strength, thereby reducing QOL.
These results on VO 2peak and muscular strength are encouraging to establish recommendations because they are known to facilitate PA behavior. It has been shown that exercise programs that improve or at least maintain physical tness during breast cancer chemotherapy improve long-term exercise adherence (52); VO 2peak can predict aerobic exercise behaviors and muscular tness resistance. The moderate PA performed in these three groups may affect long-term exercise behavior since previous studies reported controversial results depending on the PA intensity with no impact of PA level (53) or intensity (54). A strength of our study is the exhaustive assessments with validated measures and addressing areas of physical performance, body composition, symptoms, and QOL.

Conclusions
A home-based program combining aerobic and resistance training is feasible without serious adverse events, leading to a maintenance of QOL, stable body composition, and improvement in fatigue, VO 2peak , and muscular strength. Adherence to an APA program is highly variable and novel tools are required to evaluate motivation and rate of participation in speci c programs that may need to be supervised. The timing of the PA program did not strongly affect well-being 12 months after diagnosis, but this trial con rms the negative impact of no APA program during chemotherapy and highlights that it may be performed at home from the beginning of treatment.

Declarations
Ethics approval and consent to participate Clinical data were collected in accordance with French bioethics laws regarding patient information and consent. The study was performed in accordance with the Declaration of Helsinki. Data collection and use were approved by Limoges Hospital Ethics Committee (approval number no. 2012-A01401-42).

Consent for publication
Not applicable.
Availability of data and materials The datasets used or analyzed during the current study are available from the corresponding author upon reasonable request.

Competing interests
The authors declare that they have no competing interests.   VO2peak changes between T0, T1, and T2

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. TIDieRChecklistWord.docx