Predictors of cardiopulmonary fitness in cancer-affected and -unaffected women with a pathogenic germline variant in the genes BRCA1/2 (LIBRE-1)

Physical activity (PA) helps prevention and aftercare of sporadic breast cancer (BC), cardiopulmonary fitness (CPF) being an age-independent predictor of tumor-specific mortality. Therefore, we wanted to identify predictors of CPF (represented by peak oxygen uptake: VO2peak) in BRCA1/2 mutation carriers whose risk of developing BC is high. We used cross-sectional data from 68 BRCA1/2 germline mutation carrying women participating in the randomized, prospective, controlled clinical study LIBRE-1. Assessments included cardiopulmonary exercise testing, medical and lifestyle history plus socioeconomic status. Additionally, the participants completed a psychological questionnaire regarding their attitude, subjective norms, perceived behavior control and intention towards PA. A multivariate logistic regression model was used to identify predictors for participants reaching their age- and sex-adjusted VO2peak reference values. 22 participants (median age: 40 years, interquartile range (IQR) 33–46) were cancer-unaffected and 46 cancer-affected (median age: 44 years, IQR 35–50). The strongest predictor for reaching the reference VO2peak value was attitude towards PA (Odds Ratio 3.0; 95% Confidence Interval 1.3–8.4; p = 0.021). None of the other predictors showed a significant association. A positive attitude towards PA seems to be associated with VO2peak, which should be considered in developing therapeutic and preventive strategies. Trial registrations: NCT02087592; DRKS00005736.

www.nature.com/scientificreports/ Cardiopulmonary fitness (CPF). In addition to the high risk of developing cancer, there is also increasing evidence that BRCA1/2 mutation carriers are more likely to suffer from cardiovascular diseases 2 . Initial studies indicate, that VO 2peak could be a predictor of overall mortality 3 and breast cancer specific mortality 4 in sporadic breast cancer patients. Cardiopulmonary fitness, measured by peak oxygen consumption (VO 2peak ), is an assessment of cardiovascular function. A recent investigation found an inverse association between VO 2peak and several biomarkers linked to tumorigenesis 5 . A reduced VO 2peak favors tumor-related side effects, e.g. cancer related fatigue or reduced quality of life 6 . A high level of physical activity leads to a higher VO 2peak and thus better cardiopulmonary fitness. In a meta-analysis involving 571 cancer patients, a supervised exercise training was associated with significant improvements in VO 2peak 7 . Therefore, strategies to prevent or restore a weak VO 2peak in the large and fast-growing population of cancer survivors are of great clinical relevance. Hence, we wanted to identify factors, which influence physical activity behavior and consequently the VO 2peak . Another two components are relevant in this undertaking, which are explained below.
Physical activity in youth. Studies indicate that the high risk of cancer in BRCA1/2 germline mutation carriers may be mitigated by physical activity, especially during adolescence and early adulthood 8,9 . In a retrospective, longitudinal case-control study of 886 BRCA1/2 mutation carriers, Lammert et al. (2018) showed that women moderately active between the ages of 12-17 years had a 38% lower risk of developing premenopausal breast cancer compared to less active women 10 . A further retrospective cohort study (725 BRCA1/2 mutation carriers) reported a 42% risk reduction for developing breast cancer among participants with increasing levels of sports activity prior to, but not after, age 30 11 .
Theory of planned behavior (TPB). The theory of planned behavior (TPB) includes the determinants attitude, subjective norm, perceived behavioral control and intention, and is validated for predicting physical activity behavior in breast cancer survivors. Studies have shown an association between attitude, the subjective norm, perceived behavioral control, and intention with regard to the current physical activity behavior in breast cancer survivors [12][13][14] .
We wanted to examine the association of the aforementioned factors (physical activity in youth and TPB) in particular, adjusted for other available co-factors in this study with VO 2peak in this high-risk group of BRCA1/2 mutation carriers.

Methods
The LIBRE-1 study (Lifestyle Intervention study in women with hereditary BREast and ovarian Cancer, 1 = pilot) is a multi-centric, prospective, randomized and controlled clinical trial. It is registered in the German Study Register for Clinical Trials (DRKS No.: DRKS00005736), as well as the study registry of the National Institutes of Health (NCT No.: NCT02087592) on 14/03/2014. The study was conducted according to the requirements of national laws and ICH E6 Good Clinical Practice (GCP) of June 1996. The recommendations of the Declaration of Helsinki in its current version were followed as well as the German Federal Data Protection Act (BDSG). The ethics review board of the Klinikum Rechts der Isar of the Technical University of Munich has approved the study protocol (Reference 5686/13). All participating centers and all participants provided written informed consent and the study design and methods are reported in detail elsewhere 15 . Study population. In the LIBRE-1 study, 68 cancer-unaffected or previously breast and/or ovarian canceraffected women with a BRCA1/2 germline mutation were recruited from three centers of the German Consortium for Hereditary Breast and Ovarian Cancer (GC-HBOC, www. konso rtium-famil iaerer-brust krebs. de). Inclusion criteria were: pathogenic germline variant in the BRCA1 or BRCA2 gene, age ≥ 18 years and written informed consent. Exclusion criteria were: presence of metastatic tumor disease, life expectancy < 3 years, Body Mass Index (BMI) < 15 kg/m 2 , clinically limiting cardiopulmonary disease and Karnovsky-Index (describes the general condition of a patient) < 60%, blood pressure at rest > 160/100 mmHg. Further exclusion criteria were significant orthopedic or psychological problems that would not allow the participant to partake in a group intervention, or a current pregnancy 15 . Assessment of cardiopulmonary fitness. Cardiopulmonary fitness was determined by the peak oxygen uptake (VO 2peak) and assessed by cardiopulmonary exercise testing (CPET). The CPET was a ramp protocol (3 min sitting on the bicycle, 3 min steady state at 30 watts, continuous individual increase in wattage with the aim of achieving a maximal workload on the test-person within 8 to 12 min, 5 min recovery after exercise) with the target of exhausting them with a respiratory exchange ratio (RER) > 1.05. During the CPET, the participants were asked about their received perception of exertion (RPE) based on the Borg scale every 2 min. The Borg scale ranges between 6 (no exertion at all) to 20 (maximal exertion) 16 . The VO 2peak indicates the maximal number of milliliters of oxygen the body can utilize per minute when under workload. The VO 2peak can be used as a criterion for evaluating a person's physical endurance. In the present analysis, the aim was to determine which participants could reach their VO 2peak reference value. The VO 2peak reference value was calculated with the SHIP study formula (Study of Health in Pomerania). This formula considers the gender, age, height and weight of the participant in calculating their VO 2peak reference value 17 . In addition to the VO 2peak , the determination of the oxygen uptake (VO 2 ) at the first ventilatory threshold (VT1) as a criterion for the aerobic capacity was also assessed 18 . www.nature.com/scientificreports/ Questionnaire assessment of physical activity in youth, the attitude and rating towards physical activity (TPB) and the socioeconomic status. The physical activity in youth was determined with a clinical baseline interview using an international questionnaire, clinical baseline questionnaire. Participants' activity ages 10 to 19 years were categorized into two groups: (1) Much less active and somewhat less active than their peers were classified as Inactive in youth and (2) participants who indicated that they were exactly as active or more active than their peers were classified as Active in youth.
The questionnaire for the assessment of the rating towards physical activity was developed based on the theory of the planned behavior of Icek Ajzen 19 . The TPB is validated for predicting physical activity behavior in breast cancer survivors 20,21 . The TPB assumes that positive attitudes such as "exercise is fun" and subjective norms "my family thinks I should exercise more", as well as perceived behavior control "to exercise regularly is for me very feasible", influence intentions. Intentions such as "I intend to exercise regularly" are assumed to be mediators of the influence of subjective norms and attitudes on behavior. TPB's summary statement is that people intend to behave, if they rate it positively, believe that important other individuals feel that they should do it, and that they perceive it to be under their own control 19 . Participants completed a psychological questionnaire "Rating physical activity and nutrition" (Bewertung körperlicher Aktivität und Ernährung = BKAE) based on the TPB. The 9 sections of the exercise-related part of the BKAE questionnaire include 44 questions on a 7-step scale and were standardized for the analysis so that the 4 TPB determinant scores (attitude, subjective norms, perceived behavior control and intention) would be comparable.
In the clinical baseline questionnaire the socioeconomic status was also determined 22 . For the analysis we used an adaptation of the Winkler-Stolzenberg Index (WSI = ((education-1)/6) + employed − smoker) based on the following items with values between 0 and 1. Statistical analysis. First all data were examined graphically and descriptively. T-Tests were carried out to compare the groups of cancer-unaffected and cancer-affected participants. Then, in order to determine possible parameters that influence the VO 2peak performance in the study sample, a multivariate logistic regression was carried out, where the dependent variable was the dichotomized VO 2peak (i.e. whether the participant reached at least their VO 2peak reference value, which is explained in detail above in the section "Assessment of cardiopulmonary fitness" 17 ). The independent variables were the TPB determinants (standardized continuous score values from the BKAE questionnaire as explained in detail above), cancer-affected status (binary), physical activity in youth (binary) and the socioeconomic status' value (as explained in detail above). In order to illustrate the aims and methods of this analysis, we have added a graphical abstract (Fig. 1).
All statistical analyses were carried out using R version 3.4.2 (R Core Team, 2017. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https:// www.rproje ct. org). For all analyses, an alpha level of 0.05 was used to determine statistical significance (two-tailed).
Ethical standards. The 24 . Results from the CPET showed a median VO 2peak of 26 ml/min/kg in both groups, while median VO 2 at VT1 as its percentage was 61.9% in the cancer-affected group and 60.9% in the cancer-unaffected group (Table 1). Of the cancer-affected participants, 28 (61%) considered themselves as active as their peers during their youth in comparison to 19 (86%) of the cancer-unaffected participants.
Cancer-unaffected participants had lower standardized median attitude and subjective norms scores in comparison to cancer-affected participants (attitude: 0.04 versus (vs) 0.30, subjective norms: − 0.37 vs 0.01), while their median intention and perceived behavior control scores were higher compared to cancer-affected group (intention: 0.39 vs − 0.63, perceived behavior control: 0.50 vs − 0.55) ( Table 1). Table 2, participants with higher attitude scores showed significantly higher odds of reaching the VO 2peak reference val- Table 1. Characteristics of the participants. Data are presented as median (interquartile range); Active in youth = those who indicated that they were exactly as active or more active than their peers from ages 10 to 19 years; BMI Body Mass Index; VO 2peak Maximal Oxygen Uptake; VT1 ventilatory threshold one; RER respiratory exchange ratio; RPE rate of perceived exertion/RPD rate of perceived dispnoe. *Adaptation of the WSI to describes the socioeconomic status (compare method part).  www.nature.com/scientificreports/ ues (Odds ratio (OR) 3.0; 95% confidence interval (95% CI) 1.3-8.4; p = 0.021). The intention (OR 1.9; 95% CI 0.7-5.3; p = 0.192) scores showed a trend for higher chances of reaching the VO 2peak reference values, whereas perceived behavior control (OR 0.4; 95% CI 0.2-1.0; p = 0.067) and subjective norms (OR 0.8; 95% CI 0.4-1.5; p = 0.491) scores showed the opposite tendency. Cancer-unaffected participants and those with a higher socioeconomic status had higher odds of reaching their VO 2peak reference values (OR = 1.8; CI 0.4-7.9; p = 0.438 and OR = 1.4; CI 0.7-2.9; p = 0.417 respectively). Those who indicated they were active in their youth had lower odds of reaching their VO 2peak reference values (OR = 0.6; CI 0.2-2.1; p = 0.398).

Discussion
The purpose of this analysis was to identify predictors of cardiopulmonary fitness (VO 2peak) in cancer-unaffected and cancer-affected BRCA1/2 mutation carriers.
Of all determinants, a positive attitude towards physical activity (a TPB component) was the only significant predictor found in our analysis (OR 3.0; 95% CI 1.3-8.4; p = 0.021) for reaching the participant's VO 2peak reference value. A previous meta-review across 72 studies suggests that generally people's attitudes, seem to be the key influences in forming intentions to participate in physical activity 25 . According to the TPB it is necessary to have an intention, such as the intention of "walking half an hour every day" to carry out the intended behavior in accordance with the attitudes, subjective norms, and perceived behavioral control. Courneya et al. established that attitude and subjective norm were significant determinants of intention in a study of 164 sporadic breast cancer patients, and intention and perceived behavioral control in turn were significant factors of physical activity behavior during breast cancer treatment 20 .
In our analysis, the intention, subjective norms, and perceived behavioral control played a subordinate role for reaching the VO 2peak reference value, which may indicates that attitude towards physical activity should be considered when implementing physical activity programs, which consequently influences the cardiopulmonary fitness. A qualitative study by Smith et al. reported that participants' attitudes toward physical activity were negatively influenced by receiving little information from the oncologist and health professionals. A lack of social support and structured exercise programs were also identified as barriers to physical activity implementation 26 . Therefore, treatment strategies for BRCA1/2 mutation carriers should contain information dissemination and education about the beneficial effects of physical activity before and after tumor occurrence to improve the participants' attitude towards physical activity. In addition, cancer patients should be educated that physical activity is feasible and safe during tumor therapy. For this, it can be advantageous to include, inform and educate people around the participant accordingly.
It has been shown that age, sex, health status, self-efficacy and motivation are associated with higher levels of physical activity. Baumann et al. described that the personal attitude towards exercise and physical activity behavior in youth and adolescence and the intention to maintain a physical activity behavior are important components, which influence current physical activity 27 . Therefore, we included physical activity in youth in our analysis, however no association was found. It should be noted that the activity in youth was only surveyed retrospectively in our analysis and did not include any information on the duration and intensity of the respective physical activity. VO 2peak is determined by genetics, gender, body composition, physical activity behavior and age 28 . The VO 2peak decreases by about 8-10% per decade from the 30th year of life 29 . In cancer patients, VO 2peak has been found to be about 30% lower compared to healthy peers 7 . The reference values of VO 2peak were age-adjusted so that the age-related decrease in cardiopulmonary fitness was implicitly included in our analysis. We found no significant association between cancer status and reaching the VO 2peak reference value in our study.
We had expected that previously cancer-affected BRCA1/2 mutation carriers would reach their VO 2peak reference value less often compared to the cancer-unaffected participants. Usually, there are great differences between cancer-unaffected and previously cancer-affected women depending on the respective tumor therapy 29,30 . Various tumor therapies in combination with a sedentary lifestyle additionally led to marked impairments in cardiopulmonary fitness and may not recover after treatment 31,32 . This circumstance has not yet been investigated in BRCA1/2 mutation carriers. However, some studies have shown that the knowledge of an increased cancer risk or a positive BRCA genetic test induces lifestyle changes [33][34][35] . Positive lifestyle changes, which promote a higher amount of physical activity can lead to a better cardiopulmonary fitness. This could be even more important especially after a cancer diagnosis and might explain our results.
For sporadic breast cancer patients, the TPB might be a viable framework to promote physical activity successfully during cancer treatment. However, the salient beliefs of breast cancer patients concerning physical activity were different from those of the cancer-unaffected population 20 . Whether this also applies to cancer-affected BRCA1/2 mutation carriers and has an influence on the VO 2peak still needs to be investigated. If these findings are validated, the attitude should be taken into account in the treatment and preventive measures for BRCA1/2 mutation carriers with regard to cancer risk and prognosis.
Limitations. Physical activity in youth and adolescence was recorded retrospectively and only in comparison with the peer group. This meant that no quantitative information was available on the extent, duration, intensity and type of physical activity. In future, this information should also be collected in order to ensure better comparability and significance with regard to physical activity in youth.
As our results come from a feasibility study it should be noted that the informative value is limited by the small sample size and must be interpreted accordingly.

Conclusion
In conclusion, a positive attitude towards physical activity seems to play a role in the current VO 2peak status in cancer-unaffected and cancer-affected BRCA1/2 mutation carriers. Physical activity during adolescence and the health and socioeconomic status did not influence VO 2peak levels. These relationships need to be investigated in a larger sample size, for instance in the ongoing LIBRE-2 study. If these results are verified, future concepts and studies should consider the attitude towards being physically active in managing BRCA1/2 mutation carriers.