The influence of lifestyle changes (diet, exercise and stress reduction) on prostate cancer tumour biology and patient outcomes: A systematic review

Abstract Background The mostly indolent natural history of prostate cancer (PCa) provides an opportunity for men to explore the benefits of lifestyle interventions. Current evidence suggests appropriate changes in lifestyle including diet, physical activity (PA) and stress reduction with or without dietary supplements may improve both disease outcomes and patient's mental health. Objective This article aims to review the current evidence on the benefits of all lifestyle programmes for PCa patients including those aimed at reducing obesity and stress, explore their affect on tumour biology and highlight any biomarkers that have clinical utility. Evidence acquisition Evidence was obtained from PubMed and Web of Science using keywords for each section on the affects of lifestyle interventions on (a) mental health, (b) disease outcomes and (c) biomarkers in PCa patients. PRISMA guidelines were used to gather the evidence for these three sections (15, 44 and 16 publications, respectively). Evidence synthesis For lifestyle studies focused on mental health, 10/15 demonstrated a positive influence, although for those programmes focused on PA it was 7/8. Similarly for oncological outcomes, 26/44 studies demonstrated a positive influence, although when PA was included or the primary focus, it was 11/13. Complete blood count (CBC)‐derived inflammatory biomarkers show promise, as do inflammatory cytokines; however, a deeper understanding of their molecular biology in relation to PCa oncogenesis is required (16 studies reviewed). Conclusions Making PCa‐specific recommendations on lifestyle interventions is difficult on the current evidence. Nevertheless, notwithstanding the heterogeneity of patient populations and interventions, the evidence that dietary changes and PA may improve both mental health and oncological outcomes is compelling, especially for moderate to vigorous PA. The results for dietary supplements are inconsistent, and although some biomarkers show promise, significantly more research is required before they have clinical utility.

obesity, 5 high alcohol intake, 6 smoking 7 and dietary constituents and supplements 8 (fried food, 9 cadmium, 10 high and low vitamin D, 11 vitamin E and selenium, 12 lycopenes, 13 green tea, cruciferous vegetables, pomegranate 14 and phytoestrogens 15 ) have all been found to have either negative or positive effects on PCa development or progression. 1 Although several studies have investigated individual dietary factors, a broader approach has been to assess the influence of a reduction in obesity, body mass index (BMI) and chronic stress. The molecular biology underlying this has been explored with details of how obesity and stress affect proinflammatory cytokines, growth factors and the immune response both systemically and within the emerging tumour microenvironment (TME). 16,17 Evidence from these studies would suggest appropriate changes in lifestyle including diet, exercise and stress reduction may improve both disease outcomes and patient's mental health 18 as well as reducing the economic cancer burden on global health systems. 19 Moreover, there may be biomarkers that have utility for clinicians and patients to assess whether these lifestyle changes are having a positive benefit. 20 As research into lifestyle programmes gains momentum, this article aims to review the current evidence on the benefits of lifestyle programmes for PCa patients aimed at reducing obesity and stress as well as other dietary modifications and supplements, explore how tumour biology may be affected by such interventions and highlight any biomarkers that have clinical utility.

| OBESITY, SYSTEMIC INFLAMMATION AND PCa TUMOUR BIOLOGY
Obesity may promote the development and progression of several cancers, including prostate, via the induction of systemic inflammation and its local influence on the TME. 21 In a state of obesity, adipose tissue exceeds its blood supply leading to tissue hypoxia, triggering a chain of events ultimately causing sustained inflammatory signalling. Local ischaemia activates the cytokine MCP-1, causing macrophage proliferation and coalescence resulting in crown like structures around dying adipose cells. 22 After macrophage phagocytosis, released free fatty acids stimulate toll like receptor 4 and nuclear factor kappa B, with increased expression of proinflammatory genes cyclooxygenase 2 (COX-2), TNF-alpha and IL-1Beta. 23 Further release of free fatty acids with associated TNF-alpha and IL-6 cytokine signalling results in sustained adipose inflammation. Elevated circulating levels of TNF-alpha and IL-6 result in a switch to a systemic pro-inflammatory state, with adipose tissue in obese patients providing a source of these proinflammatory mediators for a developing TME. In PCa, IL-6 is also known to activate the androgen receptor and so may specifically promote PCa cell proliferation and survival. 24 Obesity-induced insulin resistance is also associated with elevating circulating levels of insulin like growth factor (ILGF) that may promote carcinogenesis as may other signalling pathways associated with obesity such as the adipocyte specific hormone leptin (stimulating cell proliferation), adipokines (increasing TNF-alpha secretion), reduced adiponectin (anticancer activity by inhibition of vascular endothelial growth factor [VEGF] and increased insulin sensitivity), ceruloplasmin (promoting angiogenesis) and the kynurenine pathway (activating the aryl hydrocarbon receptor [AHR] with suppressed T cell effector functioning and increased Treg activity). 25 Clinical studies have also investigated the association between obesity and PCa. 26 Examining tumour and benign tissue in overweight and normal weight men with PCa, nearly 95% of whom had localised disease, 15 gene sets were found to be overexpressed in obese men, 5 of which functioned for chromatin modification and remodelling, which are linked to DNA mutation burden. Moreover, obese patients had worse Gleason grades and a poorer prognosis independent of disease stage. 26 Another study, investigating weight loss in presurgical PCa patients, found weight loss-induced mixed effects on tumour gene expression, with increased expression of the proliferative marker Ki67 in malignant epithelium as well as reduced expression of genes related to insulin secretion (EFNA5), and increased expression of immune response genes (MRC1, HLA-PB1 and CD86) and DNA repair genes. 27 The genetic changes affecting insulin metabolism, immune responses and DNA repair are biologically advantageous, but elevation of Ki67 in tumour tissue is not. This may be related to the 'Obesity paradox' seen in advanced disease when significant tumour burden and the catabolic state results in obesity and excess adipose tissue providing a biological advantage. 28 A follow-up study from the same group noted weight loss was associated with loss of lean muscle and muscle catabolism, activating mitochondrial rather than glycolytic pathways. The increased Ki67 was highly linked to loss of lean muscle mass and so could potentially be reversed by combining lifestyle programmes with both weight loss and resistance muscle training to maintain lean muscle. 29 Another clinical trial evaluated the effect of exercise in obese patients with advanced PCa on circulating tumour cells (CTCs). CTCs are of prognostic value in advanced PCa, and platelets have been shown to interact with CTCs in a process known as 'cloaking' that facilitates immune escape from NK cells. 30 A positive correlation was found between CTC and platelet count, which persisted in control and overweight groups but less so in the exercise group and not at all in the normal weight group. These results may give insight into the role of obesity in disease progression by a platelet cloaking CTC mechanism, 31 although whether this is relevant to localised and locally advanced disease requires further study.

| STRESS RESPONSES AND PCa TUMOUR BIOLOGY
Chronic stress has been implicated in the development of several cancers, in part by similar mechanisms to obesity. Through activation of the sympathetic nervous system (SNS) and the hypothalamic-pituitary-adrenal (HPA) axis, chronic stress can not only promote oncogenesis but also produce systemic inflammation and suppress the immune system. The proposed oncogenic effect of catecholamines is mediated by the Beta 2 adrenergic receptor (ADRB2) enhancing tumour growth and angiogenesis by activation of the c-AMP protein kinase A (PKA) pathway with subsequent elevation of matrix metalloproteinases and VEGF, as well as p53 breakdown and DNA damage. 32 Additionally, Scr phosphorylation by the same pathway activates Ras-related protein 1 (Rap1) that inhibits extracellular signal regulated kinases (ERKs) promoting tumour cell invasion and migration. 33 By elevating corticosteroids and catecholamines, chronic stress may stimulate pro-tumorigenic immune cells to produce IL-10, IL-6, TNF-alpha and MCP-1 as well as activating the COX-2/PGE2 pathway to produce VEGF, all of which may influence TMEs to suppress tumour immunity. Moreover, a reduction in IL-12 and elevation of IL-10 selectively inhibit Th1, subsequently reducing cytotoxic T lymphocytes (CTL)-mediated interferon production and cellular immunity. 32 Bio-behavioural studies of stress have also demonstrated its influence on gene expression in the TME with stressinduced activation of the SNS inhibiting programmed cell death and the immune response as well as promoting angiogenesis, inflammation, epithelial to mesenchymal transformation and tumour invasion. [34][35][36][37] A cancer diagnosis itself increases stress levels and so may potentially exacerbate disease progression by the above mechanisms. 38 Specific to PCa, an early clinical trial, albeit with small numbers, investigated whether 6 months of mindfulness-based stress reduction (MBSR) (with a plant-based diet) could influence prostate specific antigen (PSA) levels in localised PCa patients following surgery or radiotherapy. 39 Of the 14 patients recruited, 8 had intermediate or highrisk disease, and with an average time from treatment to trial recruitment of 43 months, a significant reduction in post intervention PSA levels and PSA doubling times was seen compared with patients' PSA dynamics before the intervention.
A similar study investigating MBSR in a lifestyle intervention also including diet and physical activity (PA) examined changes in prostate gene expression in a cohort of men under active surveillance (AS) for low-risk PCa. 40 In the 30 recruits, there was reduced expression of 453 genes and increased expression of 48 genes all relating to tumorigenesis, including intracellular protein transport, phosphorylation and metabolism as well as down regulation of RAN and SHOC2 oncogenes. However, with both studies, separating the influence of MBSR from the dietary and exercise aspects of the lifestyle intervention is difficult.
The influence of stress on tumour biology and the racial disparity seen in PCa incidence and outcomes has also been investigated. 41 Studies have shown from early in life, children raised in a low socioeconomic environment develop resistance to glucocorticoid signalling, leading to exaggerated inflammatory and adrenocortical responses, which may then contribute to oncogenesis as adults. 42 Later in life, chronic stress induced by racist behaviour may give rise to increased b-adrenergic and corticosteroid signalling, 43 increased expression of pro-inflammatory genes 44 and increased smoking and alcohol consumption 45 creating an environment that also promotes cancer incidence and progression. 41

| EVIDENCE ACQUISITION/METHODS
Evidence for this review was obtained from PubMed and Web of Science using keywords specific to each section on the effects of lifestyle interventions on (a) mental health, (b) disease outcomes and (c) biomarkers, metabolic imaging, surgical pathology and genetics in PCa patients. The search strategy aimed to include all potential terms that may be used to describe a 'Lifestyle Program' focused on any aspect of diet (including supplements), PA and MBSR, as well as their influence on mental health outcomes, oncological outcomes, biomarkers and tumour biology. PRISMA guidelines were used to gather the evidence for these three sections. All publications including systematic reviews, qualitative, quantitative, experimental and mixed method design studies were included. References of the chosen articles were also screened and relevant missing studies added. Any article not available in English was excluded. The PRISMA flow charts are shown in Figures 1 and 2, and the PRISMA checklist is provided in the Supporting Information. Abstracts and full papers were reviewed by two qualified individuals by title, abstract and then complete manuscript to avoid selection bias. The data were then extracted from the chosen articles as per Tables 1-3 and analysed in the text accordingly.

| SPECIFIC EFFECT OF LIFESTYLE INTERVENTIONS ON MENTAL HEALTH IN PCa PATIENTS
For the increasing survivorship of PCa patients, balancing a positive mental outlook in the present against worries for the future is crucial, 122 110 41 men with localised PCa that had undergone prostatectomy or radiation therapy Emotional expression and processing Emotional processing predicted lower inflammatory markers IL-6, sTNF-RII and CRP. Emotional expression was significantly associated with higher levels of sTNF-RII. Expression of emotion may be associated with higher inflammation only in the context of low emotional processing.

Physical activity
One exercise session on the day prior to radical prostatectomy is insufficient to effect NK cell infiltration into prostatic tissue.

male patients with intermediate or high-risk PCa scheduled for radical prostatectomy
Physical activity Physical exercise mobilises and redistributes NK cells, particularly those with cytotoxic phenotype.
No increase of NK cell tumour infiltration.
Djurhuus et al. 113 30 patients with localised PCa undergoing radical prostatectomy Physical activity The per-protocol analysis (HIIT 4 times per week) showed a significant increase in tumour NK-cell infiltration.
The total number of training sessions was positively correlated with the change in NK-cell infiltration Kaushik et al. 114 29 men with newly diagnosed PCa Physical activity/stress management (yoga) Increased numbers of circulating CD4+ and CD8+ T-cells and production of interferon-gamma by natural killer cells.
Increased Fc receptor III expression in natural killer cells. Decrease in numbers of regulatory T-cells, myeloid-derived suppressor cells, suggesting antitumour activity Reduction in inflammatory cytokine levels (granulocyte colony-stimulating factor, monocyte chemoattractant protein and FMS-like tyrosine kinase-3 ligand (Continues)

T A B L E 3 (Continued)
Reference Cohort size and demographics

Type of lifestyle intervention
Biomarker activity outcomes Hanson et al. 115 11 men with PCa on ADT, 14 men with PCa not on ADT and 8 healthy controls Physical activity PCa survivors have NK cell mobilisation comparable with healthy controls following acute exercise.
CD56 total cell egress tended to be attenuated in ADT.
ADT also consistently showed a less mature phenotype with greater proportion of IFNγ expression and possibly lower cytotoxicity. Though perforin levels remained unchanged.
The proportion of perforin expressing NK cells in PCa was reduced, suggesting these cells may be more prone to degranulation. By 24 h, all NK and leukocyte populations returned to baseline levels, suggesting that consecutive training sessions could be used without adverse immune system effects.
Ornish et al. 116 Hvid et al. 119 25 patients with biochemical recurrence of PCa or on active surveillance

Physical activity
The intervention group showed significant improvements in plasma triglycerides, IGF-1, IGFBP-1, adiponectin and fasting glucose levels.
Galvão et al. 120 10 PCa patients on ADT for at least 2 months Physical activity Serum growth hormone (GH), dehydroepiandrosterone (DHEA), interleukin-6, tumour necrosis factor-alpha and differential blood leukocyte counts increased following acute exercise.
Cohen et al. 121 159 PCa patients scheduled for radical prostatectomy Stress management/supportive attention Men in the stress management group had significantly higher levels of natural killer cell cytotoxicity and circulating proinflammatory cytokines 48 h post-surgery than men in the supportive attention group and higher levels of natural killer cell cytotoxicity and IL-1β than men receiving standard care.
within PCa patients' broader oncological care. 123 Not surprisingly, PCa patients are more likely to be prescribed antidepressants compared with their cancer-free counterparts, and some studies have reported suicide rates up to 6.5 times higher in PCa patients compared with controls. 124 There have been several studies investigating the effect of lifestyle interventions on mental health in PCa patients or cohorts including a significant fraction of PCa patients. A summary of these is provided in Table 1 with the results represented in Figure 3.
As is evident from Table 1, cohorts and outcome assessments are heterogenous making uniform conclusions challenging. There are 15 studies presented in Table 1 incorporating approximately 3000 patients with PCa. In Figure 3, the schematic mosaic plot reflecting these studies suggests those with a focus on PA may be more likely to have a positive influence on mental health.

| Studies showing a positive influence on mental health
Of the 10 studies demonstrating a positive influence on mental health, 2 focused on broader interventions combining PA with diet and stress management and 1 on a more esoteric approach using a smart shirt and mobile app solution for depression self-management. 46 and SF-36 form. 53

| Studies showing no influence on mental health
Of the 5 studies in Table 1  The molecular mechanism behind which exercise improves mental health may relate to increasing serotonin, endothelial growth, insulinlike growth and brain-derived neurotrophic factors within the central nervous system (CNS). 125

| Diet only
There are 31 studies focused on diet, 48.4% (15) of which were successful at improving oncological outcomes [67][68][69]71,73,84,85,[91][92][93]95,[97][98][99]102 and 51.6% (16) of which were not. 70 145 In other studies, the accuracy of NLR as a prognostic marker of disease progression in localised PCa or response to treatment (e.g., docetaxel, cabazitaxel or radium-223) in advanced PCa has been inconclusive suggesting the need for further study. [146][147][148][149] Investigating specifically SII in a retrospective multicentre cohort of over 6000 men treated surgically for localised PCa, elevated presurgical SII was found to be significantly associated with both adverse surgical pathology and longer term BCR. 150 The predictive value of the SII is supported by other studies in localised PCa post surgery examining not only BCR but also long term survival. [151][152][153] Focusing on PCa oncogenesis, Fowke and Motley 154 studied links between metabolic dysregulation, obesity and systemic inflammation in 160 men with high grade prostatic intraepithelial neoplasia (HGPIN), regarded in high volume as a PCa risk factor, who were then diagnosed with PCa at subsequent biopsy. They found statin use was linked to overall PCa diagnosis, but obesity parameters and markers of systemic inflammation were not. In another study, no clear relationship was found between systemic inflammatory or oxidative stress markers in blood or urine with prostate size or lower urinary tract symptoms (LUTS) in patients with BPH but not PCa. 155 By contrast, an earlier study examining data from the PCa Prevention Trial, an RCT assessing whether Finasteride reduced PCa risk, found markers of systemic inflammation (in this case c-reactive protein) or lower levels of soluble receptors that bind inflammatory cytokines (such as TNF and IL-6) did increase BPH risk. 156 The inference of these studies of HGPIN and BPH is that they may represent overall 'prostate health' and draw on the belief that the path to PCa is via histological inflammation with BPH included in this process. These results are, however, inconclusive.
Studies have also assessed inflammatory cytokines as biomarkers of PCa oncological outcomes. In a retrospective analysis of 237 men undergoing surgery for localised PCa, high TNF-α levels were found to be positively correlated with upgraded Gleason scores post surgery, whereas, counterintuitively, high levels of serum IL-6 were found to be positively related with Gleason score downgrades. 157 In another study, men were randomly selected from a prospective cohort derived from the Osteoporotic Fractures in Men (MrOS) study, with a view to investigating correlation between serum inflammatory cytokine levels and risk of subsequent PCa. After 6 years of follow up, IL-10 levels were found to be associated with lower risk of PCa, but there was no association with PCa risk and either TNF alpha or IL-6. 158 Studies where biomarkers of inflammation or immunity were tracked as a function of intervention are summarised in Table 3