Every exercise bout matters:

Linking systemic exercise responses to breast cancer control

Christine Dethlefsen, Katrine Seide Pedersen, and Pernille Hojman

AFFILIATION

Centre of Inflammation and Metabolism (CIM) and Centre for Physical Activity Research (CFAS), Rigshospitalet, Faculty of Health Science, University of Copenhagen, Denmark

CORRESPONDING AUTHOR

Pernille Hojman, PhD, MSc.

Senior researcher, group leader

The Centre of Inflammation and Metabolism (CIM) and the Centre for Physical Activity Research (CFAS),

Copenhagen University Hospital, 7641, University of Copenhagen

Blegdamsvej 9, DK-2100 Copenhagen, Denmark

E-mail:

Phone: +45 35457544

ARTICLE INFORMATION

Word count:

Abstract: 211

Article:3495

Figures/boxes: 1 figure and 2 boxes

KEYWORDS

Breast cancer, acute exercise, chronic training,systemic factors

Abstract

Cumulative epidemiological evidence shows that regular exercise lowers the risk of developing breast cancer and decreases the risk of disease recurrence. The causality underlying this relation has not been fully established, and the exercise recommendations for breast cancer patients follow the general physical activity guidelines, prescribing 150 minutes of exercise per week. Thus, elucidations of the causal mechanisms are important to prescribe and implement the most optimal training regimen in breast cancer prevention and treatment. The prevailing hypothesis on the positive association within exercise-oncology has focused on lowering of the basal systemic levels of cancer risk factors with exercise training. However, another rather overlooked systemic exercise response is the marked acute increases in several potential anti-cancer components during each acute exercise bout. Here, we review the evidence of the exercise-mediated changes in systemic components with ability to influence breast cancer progression. In the first part, we focus on systemic risk factors for breast cancer, i.e. sex hormones, insulin, and inflammatory markers, and their adaptation to long-term training. In the second part, we describe the systemic factors induced acutely during exercise, including catecholamines and myokines. In conclusion we propose that the transient increases in exercise-factors during acute exercise appear to be mediating the positive effect of regular exercise on breast cancer progression.

Introduction

Within the last years, interest in exercise and physical training of cancer patients has exploded, driven by consistent epidemiological evidence, proving that regular physical activity is associated with decreased risk of a range of cancers[3]. Moreover, epidemiological studies show reduced risk of recurrence of several cancer diagnoses, including breast cancer, in physically active compared to inactive cancer survivors [5]. As a consequence, huge efforts are being put into conducting large-scale exercise intervention trials in cancer patients and survivors[6].However, there are still many challenges in the design of these studies. This is for instancereflected in the enormous range of endpoints included in the conducted exercise intervention trials. In a review of more than 80 published exercise intervention trials in cancer patients, 60 different endpoints were identified as outcome measures [8]. These endpoints ranged from direct physiological adaptations to training, such as fitness levels, oxygen consumption, muscle mass and strength, across exercise-related functional outcomes, i.e. functional capacity and body composition, to biological and psychosocial outcomes, including quality of life, cancer-related fatigue, anxiety and self-esteem.

This diversity in the methodology and selection of endpoints reflects the genuine lack of understanding of the biological mechanisms behind the protective effect of exercise on cancer risk and progression. To this end, numerous factors have been suggested to be implicated in the protection. In 2008, McTiernan proposed that exercise-dependent regulation of the systemic levels of known risk factors, i.e. sex hormones, insulin, inflammatory markers and immune cell function, could be linking exercise to cancer protection, and these factors have since been considered the main candidates as mediators of the exercise-dependent protectionagainst cancer [12]. This assumes that the factors are both directly driving cancer, as well as regulated by long-term training. Yet, the causality of this relationship has not been experimentally established. In contrast, we recently published a study challenging this idea, as training-dependent reductions in known risk factors did not translate into any control of breast cancer cell viability,when tested in serum incubation cell culture studies [13]. Oppositely, our study highlighted the importance of understanding the systemic changes occurring during each individual bout of exercise, as serum collected immediately after cessation of exercise decreased viability of the breast cancer cells[13].

In this review, we describethe role of systemic factors in the exercise-dependent control of breast cancer. First, we focus on systemic breast cancer risk factors and their basal adaptation to long-term training. Secondly, we highlight exercise factors, which are induced acutely during exercise, and where accumulating evidence underscores their importance in exercise-dependent regulation of breast cancer.Finally, the clinical perspectives of these two distinct exercise responses are discussed. To narrow the scope of the review, we only focus on the systemic effects of endurance exercise (physiological adaptations are discussed in box 1).

SYSTEMIC BREAST CANCER RISK FACTORS AND THEIR ADAPTATIONS TO LONG-TERM TRAINING

Regular exercise has been suggested to protect against breast cancer through lowering ofsystemic levels ofknown risk factors, and clinical exercise studies in breast cancer patients are thus aiming at reducingbasal levels of these with regular training. Conceptually, reducing the levels of circulating growth factors for breast cancer cells may improve cancer prognosis, and here we review the literature regarding the direct exercise-mediated effects on baseline values of these systemic components in healthy people and breast cancer patients.

Sex steroid hormones

In premenopausal women, the majority of estrogens are produced in the ovaries, while postmenopausal women primarily produce estrogens in the adipose tissue through aromatization of androgen precursors. Thus in the latter group of women,systemic sex hormone levels and body composition are tightly correlated [14].In postmenopausal women, elevated systemic levels of sex hormonesare associated with increased risk (HR: 2.58 between highest and lowest quartiles) of breast cancer independently of BMI [15]. In premenopausal women, some studies find associations similar to those seen in postmenopausal women [16], while others suggest that the association between elevated levels of sex hormones and breast cancer risk is limited to testosterone [17].

At the cross-sectional level, studies have shown that high physical activity levelsare inversely correlated with estradiol and testosterone levels in premenopausal women [18-20]. However, in a large exercise randomized controlled trial (RCT) involving 319 women,no regulation of sex hormones levels could be demonstrated with training, and the authors explained this by a lack ofweight loss[21].In postmenopausal women, the effect of exercise on sex hormone levels is tightly linked to their production in the adipose tissue, and training-dependent reductions in sex hormone levels have primarily been observed in overweight women, who lose weight during the exercise intervention. This is illustrated by a 1-year training intervention in 170 overweight postmenopausal women, where only women, who lost body weight during the intervention, showed significant reductions in estrone(-3.8%) and free estradiol (-8.2 %). Alsothe levels of free testosteronewere dependent on weight loss with an overall decrease of -6.5 % compared to -2.1 % in the control group [22,23]. At the cross-sectional level, physical activity in postmenopausal women is in some studies correlated with sex hormone levels [24-28], howeverthese correlations are only evident before adjustment for BMI [24,28], stressing the importance of fat mass in controlling sex hormone levels in postmenopausal women.The overall training-induced decreases in sex hormone levels, regardless of menopausal status, were analysed in a meta-analysis, where modest decreases in estradiol and testosterone (total estradiol: -0.12 pmol/l, free estradiol: -0.2 pmol/l,testosterone: -0.18 nmol/l) were found in theexercise intervention groups compared tocontrol groups [29].

Insulin

People suffering from type 2 diabetes and/or obesity show increased incidence of many cancers, and have higher cancer-related mortality [30]. This has partly been explained by insulin resistance with resulting hyperinsulinemia, and elevated plasma levels of Insulin-like growth factor (IGF) family members[31,32]. Insulin controls blood glucose levels by inducing peripheral glucose uptake, but also exerts direct anabolic and anti-apoptotic effects on normal and malignant cells [31,32]. Several studies have shown correlations between elevated plasma insulin and increased incidence of various cancers, including postmenopausal breast cancer [33]. In continuation, high levels of insulin have been associated with increased recurrence in breast cancer survivors [34]. IGF-1 resembles insulin in its stimulatory effects on cell proliferation [35]. Most IGFs in the bloodstream are bound to proteins such as IGF-binding protein 3 (IGFBP-3), but a small fraction of IGF-1 is bioavailable [35]. Accordingly, systemic levels of IGF-1 and its binding proteins have been related to various cancers including breast cancer [36].

A recent meta-analysis showed that breast cancer patients may reduce fasting insulin levels with a mean difference of -3.46 µU/mlafter exercise interventions, however this decrease is dependent on weight loss [37]. Another large meta-analysis of 160 randomized controlled trials, including data from >7000 subjects, showedno effect of exercise training on fasting insulin in healthy people without co-morbidities (type 2 diabetes, metabolic syndrome etc.) [38]. In line with this, other studies have shown that fasting insulin levels do not change with exercise training, although the training interventionsresult in weight loss [39,40]. A few studies have investigated the IGF-1 axis in relation to exercise in cancer survivors, but the results are inconsistent [41-43].

Inflammatory markers

Cancer-related inflammation has been included as the seventh hallmark of cancer [44], and the most prominent inflammatory markers are C-reactive protein (CRP), IL-6,and TNF-α. CRP is an acute phase protein widely recognized as a sensitive biomarker of systemic inflammation, whereas IL-6 and TNF-α are pro-inflammatory cytokines stimulating CRP production and a range of other inflammatory processes [45]. Studies have shown that elevated CRP levels are associated with increased risk of cancer, and increased levels of CRP are associated with early death after a cancer diagnosis [46]. Moreover, CRP is significantly elevated in breast cancer patients compared to people without a cancer diagnosis [47-49]. Less conclusive evidence exists for the involvement of IL-6 and TNF- in cancer risk, but plasma IL-6 has been reported to be augmented in breast cancer patients [50].

Exercise has shown to reduce systemic CRP levels [51-53], as inextremely well-trained male ultrarunners displaying66 % lower CRP levels compared to sedentary male controls [54]. Generally, trained individuals without a cancer diagnosis have between 18-60 % lower plasma CRP levels compared to controls [51,53-55]. The response in CRP levels to exercise training is most prominent if the exercise intervention exceeds 4 months, and larger reductions are observed if the subjects are obese [56], diseased with low-grade inflammation [57], or if the intervention is combined with diet restrictions [58].Limited data in breast cancer patients are available, and most studies do not show any changes in CRP during a training period [51,59-62].The effects of regular exercise on IL-6 and TNF-α levels are more varied, with studies reporting attenuated levels with training[63-65], while other studies show no training effect on IL-6 and TNF-α concentrations [52,66]. As for CRP, studies of longer duration have shown the most pronounced effects on IL-6, suggesting that prolonged training interventions are needed for reducing pro-inflammatory cytokine levels. A recent meta-analysis of 160 exercise intervention studies showed no effect on levels of CRP,IL-6 or TNF-α[38]. Little information is available on the regulationTNF-α and IL-6 in breast cancer patients

, but a meta-analysis in breast cancer survivors has shown no effect of exercise training [37].The average duration of the included trials in the two meta-analyses were 12 and 16 weeks, and it can be speculated that these interventions were of too short a duration to have an effect.

Overall, long-term training may decrease systemic levels ofthe above-mentionedbreast cancer risk factors. However, the reductions aremodest and closely related to weight loss.Importantly, no direct link between exercise-dependent reductions in their circulating levels and breast cancer progression have been established. Notwithstanding the impact of exercise on body composition, diet restrictions are much more potent in reducing excessive body weight. In this context, controlling caloric intake may be a more feasibleapproachforloweringthe systemic levels of these risk factors.

ACUTE SYSTEMIC RESPONSES IN CIRCULATINGFACTORS DURING ACUTE EXERCISE

During the performance of exercise, majorbut short-lasting alterations occur in several circulating components, whichin magnitude by far surpass the adaptations seen with long-term training. In people, who regularly exercise, these continuous “boluses” of exercise factors have the potential to impact breast cancer cell biology and viability. Yet, the anticancereffect of these acute systemic responsesto exercise remains a neglected areain the exercise-oncology research field.Preclinical cancer studies have shown that the changes induced by acute exercise are capable atdirectly inhibiting breast cancer viability, as evident through stimulation with exercise-conditioned serum or exercise-induced muscle-derived peptides [13,67,68]. Studies in other cancer diagnoses add to this effect and stress the importance of this acute response, showing reduced cancer growth byexercise-mediated increases in immune cells, epinephrine and muscle-derived factors [69,70]. Here, we discuss the acute regulations of the risk factors reviewed above, as well as other exercise factorsdemonstrating strong systemic regulation during acute exercise,in light of their potential effect in breast cancer regulation.

RISK FACTORS

Sex hormones

Increases in serum estradioland testosteronelevelsare seen in both pre- and postmenopausal women during exercise, with changes in their concentrations being approximately 35 pmol/l for estradiol and 0.2 nmol/l for testosterone, as reported in an endurance study including 30 females between 19-69 years of age[71]. Both sex hormones return to baseline levels within 30 minutes after exercise cessation [71,72].

Insulin

Generally, insulin levels are recognized to decrease during both moderate-and high-intensity endurance exercise[73,74].A study in well-trained athletes found a progressive decrease in insulin levels from baseline concentrationsduring incremental exercise, ending at 4-fold lower levels[75]. However in the recovery period, insulin rapidly increased, overshooting baseline levels 2-fold5 minutes post exercise [75]. Plasma IGF-1doesnot appear to be regulated during acute exercise [71,76]. However, a couple of studies have shown increases in plasma IGF-1 immediately after high-intensity exercise with decreases 30-90 minutes into recovery comparable to or below baseline values [77,78].

Inflammatory markers

Exercise directly affects the systemic levels of inflammatory markers. During acute exercise, plasma IL-6 levels increase >10-fold[79], and this is followed by an acute induction in the plasma concentration of anti-inflammatory markers such as IL-1ra and IL-10. This anti-inflammatory surge impedes CRP and TNF-, lowering the systemic inflammation. Indeed, ithas been shown that acute exercise is capable of inducing and maintaining an anti-inflammatory milieu several hours after the exercise bout[80]. Although exercise has also been described to elicit pro-inflammatory responses and induce increases inTNF-α, IL-1β, and CRP, this only occurs after strenuous exercise with associated muscle damage [81], and exercise is in general considered anti-inflammatory.

EXERCISE FACTORS

Myokines

During contractions, skeletal muscle releasespeptides, known as myokines. Thebestcharacterized myokine is IL-6 [82], and the increase in plasma IL-6 seen during exercise can largely be attributed to release from the contracting muscles. The exercise-induced secretion of IL-6 is closely associated with low muscle glycogen content, and exercise of high intensity or long duration, and intramuscular glycogen depletion triggers augmented IL-6 secretion from the muscle[83]. Since the discovery of IL-6, numerous other myokines induced by acute exercise have been identified, including ANGPTL-4, MCF-1, CCL2, CX3CL1, IL-8, IL-15, Irisin, and SPARC [69,84].The list of exercise-induced myokines is continuously growing, and large-scale omics-based strategies are aiming at elucidating the entire muscle secretome.

Data on the role of myokines in cancer protection is still limited, however a few preclinical studies have been conducted, demonstrating that muscle-derived OSM and Irisin can inhibit breast cancer cell viability,while SPARC reduces tumorigenesis in the colon of exercising mice[67-69].Myokines belong to a number of distinct protein classes, and their potential in control of breast canceris reflected by their ability to either activate tumor suppressor pathways, or antagonize cellular ligands involved in oncogenic pathways, e.g. TGF-β or Wnt signalling.

Stress hormones

Exercise is associated with induction of stress hormones in an intensity-dependent manner. Plasma epinephrine and norepinephrine increase rapidly within the first 15 minutes of high-intensity exercise[85], reaching levels up to >20 times of basal concentrations [85,86]. Both epinephrine and norepinephrine levels rapidly return to baseline levels after cessation of exercise with the half-life of epinephrine being within minutes. Cortisol, on the other hand, has mainly been described to increase with exercise of long duration, stimulating hepatic gluconeogenesis for maintenance of blood glucose levels. Thus, short term exercise is not thought to increase plasma levels of this glucocorticoid[74,87,88].Unlike the catecholamines, cortisol exerts its effect over several hours as its half-life is around60 minutes [88].

The impact of the acute transient peaks in epinephrine and norepinephrine has not been studied directly in breast cancer patients, butpreclinical studies of human breast cancer cell lines indicate a dual role of epinephrine, which at high concentrations inhibitscancer cell growth, whileat low concentrations stimulates it [89].Contrary to the exercise-mediated transient spikes of stress hormones, chronic stress is characterized by chronically elevated cortisol and catecholamine levels, and observational studies show that chronic stress might promote breast cancer onset and progression [90].In line with this, treatment with β-blockers was recently showed to be positively correlated with breast cancer specific survival in a large meta-analysis [91]. Noticeably, the chronic stress-induced augmentations in catecholamine levels areonly modest compared to the increases reported during acute exercise, which may explain the opposite effects on breast cancer progression of chronic stress and the spikes in stress hormone levels induced during acute exercise.