Cancer initiation, progression, recurrence, and metastasis are intricate mechanisms that depend on various components. Those components include genetic alterations, proliferation, vascularization, invasion, embolization, and evasion of apoptosis. When a tumor is established, its growth, metastatic spread, or eventual recovery, heavily depend on interactions with the microenvironment and psychological state of a patient (Armaiz-Pena et al., 2009). Despite the prevalence of cancer recurrence, research of primary psychological causes that might trigger the disease relapse are limited. Stress and fear of recurrence can also negatively affect biological processes that regulate immune and endocrine systems, influence the survival rate and the quality of life. Approximately 7% of patients develop disabling emotional condition including intrusive thoughts and misinterpretation of mild and unrelated symptoms. Patients with breast cancer (BC) have up to 30% of recurrence rate, and the timing of relapse varies considerably, influenced by classic prognostic factors (the axillary lymph node status, the tumor size, and the nuclear grade and histological grade) as well as the choice of adjuvant treatment strategies (Colleoni et al., 2016).

The new promising microbiological development proposes to suppress cancer relapse and metastasis by inhibiting cancer stem cells (Li et al., 2015). Not less important or complex research line exploring tumor development is related to the study of behavioral stress and immune or neuroendocrine biomarkers' changeability in response to stress. Psychoneuroimmunology (PNI) investigates in detail the connection between the psychological stress, its impact on neuro-immune system and cancer incidence or progression. Psychological stress and adverse life events can impact cellular immune response dysregulating homeostatic functionality on a cellular level and harm the protective functions of the immune system (Glaser and Kiecolt-Glaser, 2005). Specific signaling pathways are activated through neuroendocrine stress response resulting in promotion of tumor growth and metastasis. Neuroendocrine and immunological mediators of stress impact sympathetic nervous system and diverse interaction has been identified between malignant tissues and immune-cites (Batty et al., 2017). Oncologists now implicate psychological functioning in the prediction of cancer outcomes.

Finding a “gold standard” biomarker to measure body's response to allostatic load or experiential input (social, emotional, and physical experiences) has been proven to be challenging, given its complex etiology and highly individual manifestations. Cortisol, synthesized from cholesterol and released into the body fluids after activating hypothalamic-pituitary-adrenal (HPA) axis, is the main glucocorticoid hormone of the adrenal cortex that can be used as a peripheral indicator of hypothalamic neural activity (Dubey and Boujoukos, 2005) in the process of physiological response to endogenous and exogenous stressors. Evidence of recent research in molecular biology shows relationship between dysregulation of HPA axis, abnormal secretion of cortisol and protumorigenic pathways within breast cells (Figueira et al., 2022). When the HPA axis is chronically activated due to stress events and fears (major contributors being negative lifestyle, fast paced jobs, existential fears, negative emotional experiences, anxiety, depression), cortisol levels rise, and inflammatory responses are downregulated. As a response to biochemical stress, its secretion contributes to the suppression of the “stress axis” on mental health, where more inflammatory ligands (cytokines) are released into circulation in the vicinity of tumor cells (Levine et al., 2007; Lee et al., 2015). This proliferation of cytokines may affect disease progression by causing angiogenesis and endothelial-to-mesenchymal transition (EndMT) leading to metastasis (Morris et al., 2022).

Some molecular studies also suggest that cortisol's immunosuppressive effect may result in inability to combat tumor formation, reduced immunosurveillance of early-stage cancer, facilitating the immune escape and acquisition of further oncogenic mutations (Antonova et al., 2011; Coutinho and Chapman, 2011; Al Sorkhy et al., 2018). Additionally, cortisol has hyperglycaemic and obesogenic effects where both dysregulations lead to increased risk of malignancies (Nead et al., 2015). More specifically, stimulation of BC development by prolonged or elevated presence of cortisol during periods of stress may also partially occur as a result of augmented estrogen biosynthesis (Aromatase activity) (O'Neill et al., 1988). Thus, both clinical and molecular research reveals a positive correlation of high cortisol levels and the progression of BC, resulting in worse prognosis.

A small fraction of unbound free cortisol is demonstrated to be biologically active, coming out of the mitochondrion, whereas in general cortisol binds to cytosolic receptors. Cortisol migrates out of the cell into the extracellular space and into the bloodstream. Due to its low molecular weight and lipophilic nature, unbound cortisol enters the cells through passive diffusion, which makes it feasible to measure free cortisol in many body fluids. Several studies explored the potential of using cortisol as an indicator to assess cumulative biological risks, predict chronic stress, reduce the incidence of pre-phase of chronic illnesses (Dubey and Boujoukos, 2005; Levine et al., 2007; Lee et al., 2015).

Among its various manifestations, cortisol parameters are compounded of both serum (blood) and salivary measures. Levels of circulating cortisol reflect the activity of central and peripheral pathways that are responsive to experiential input. It has also been observed that 70% of BC patients show high levels of cortisol (De Brabander and Gerits, 1999). Plasma and salivary cortisol levels rise due to circadian influences as well as perturbations by stressors of the organism's environment that makes it possible to detect rather robust experimental effects (Levine et al., 2007; Lee et al., 2015).

Clinical studies dealing with the relationship between behavioral tendencies and the progression of a disease in patients that were already diagnosed with cancer have found that factors such as lack of social support, negative emotions and their suppression, fear of recurrence, hopelessness and denial are associated with poorer survival (Batty et al., 2017). Several systematic reviews and meta-analysis dedicated to causal relationship of stress, depression with incidence or progression of tumor have been inconclusive or showing only borderline association (De Brabander and Gerits, 1999; Petticrew et al., 1999; Antoni et al., 2006; Batty et al., 2017). Other studies have confirmed the adverse effect of this relationship on cancer incidence and survival (Duijts et al., 2003; Chida et al., 2008; Santos et al., 2009). Finally, in the study of Satin et al. (2009), depression in cancer patients has been exposed to be a predictive factor of mortality but not progression. These clinical data parallel results from animal-based studies, demonstrating that experimentally imposed stress can modulate cancer progression (Mustafa et al., 2013).

When determining the predictive power of psychosocial factors for the course of BC, additional biological factors and markers suggesting progression including tumor-free interval (TFI), site of recurrence, number of metastases, menopausal status and use of different medications should be taken into consideration (Levy et al., 1988; Carter et al., 1989; Lin et al., 2013; Mustafa et al., 2013).

BC is regarded to be one of the most appropriate cancers for studying psychosocial influences on disease progression due to its relatively variable course, high prevalence, and hormone sensitivity (Thaker and Sood, 2008).

Thus, data that would illustrate a relationship between behavioral risk factors and cancer initiation and progression have been inconclusive. To differentiate our research and bring novelty to the subject, this review and meta-analysis is aimed to consolidate existing data up to date (2022) and test the hypothesis if psychosocial interventions (and their variations that are subject of this study) can reduce neuroendocrine biomarker (cortisol) and can be considered as a plausible adjuvant therapy to attenuate chronic stress. The objective of this study is to provide a summary of the studies where primary or secondary outcome measure was cortisol in the population of females with BC in various stages of disease progression. Further, to present an overview of methodological and reporting aspects of these studies and to execute a quantitative analysis of association between the applied intervention and cortisol levels. We focused solely on BC and narrowed the scope of psychosocial interventions to be structured and methodical. We also considered other potential moderators, including characteristics of samples and trial quality by pooling relevant data from studies that met eligibility criteria. Although this review has focused on investigating Changes in the variable of cortisol under the influence of applied psychosocial intervention, neuroendocrine and immunological mechanisms that modulate cancer growth and might prevent cancer progression, metastasis or recurrence will remain outside of the scope of this review.

To our knowledge no other quantitative study devoted to this narrow scope of interventions and outcome measures has been conducted so far.

We used the checklist of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Guidelines as a validated methodology of investigation (Page et al., 2021) to perform the selection procedure, study identification, synthesis methods and critical appraisal.

The present meta-analysis aimed to test if the effect of certain types of psychosocial interventions on stress reduction can be traced by measuring cortisol in saliva or blood. We have based our research on a narrow niche of specific parameters by using cortisol as a biomarker to measure the effect size of specific psychosocial interventions (that are subject of this study) on chronic stress reduction.

Psychosocial interventions in mental health use non-pharmaceutical means to alter a person's behaviors and relationships with society, to reduce the impact of the person's disorder or condition and manage the causal issues better. Interventions associated with psychosocial support fall under two main umbrellas of therapy: cognitive therapy and behavioral therapy, or collectively - cognitive behavioral therapy and its sub-categories.

In our research we determined that the applied interventions within these two major arms providing non-invasive care had overall impact on the BC patients' level of cortisol. The scope of these interventions included: Mindfulness Based Stress Reduction, Cognitive Behavioral Stress Management, Mind Body Spirit integrated therapy, cognitive restructuring, cognitive based compassion therapy, hypnotic guided imagery program, coping skills development, psychoeducation, supportive expressive therapy, behavioral modification program, spiritual and psychosocial support, various relaxation techniques, biofeedback, complementary medical therapies involving movement in meditation modalities. All these programs emphasized positive aspects and emotionally uplifting assertions or techniques. In the studies measuring salivary cortisol (Matchim et al., 2011; Baker et al., 2012; Hsiao et al., 2012) and BMS Intervention arm of Chan et al. (2006) study the applied interventions decreased the level of cortisol in patients. In the studies of Schedlowski (1994), Cruess et al. (2000), and Phillips et al. (2011) using plasma cortisol measurements, the reduction of cortisol due to intervention was even more noticeable, thus providing stronger evidence of this relationship. These results confirmed our hypothesis that psychosocial support and mental adjustment can decrease stress marker (cortisol) in patients with BC. However, as described in Results section, interventions [study of Dodds et al. (2015) and SE intervention arm of the study of Chan et al. (2006)] that dealt partially or entirely with negativity aspects like fears of recurrence, death, dying, or accentuating distress had no effect or had adverse effect on the level of cortisol in patients with BC.

Studies of Schedlowski (1994), Cruess et al. (2000), and Matchim et al. (2011) provided information that cortisol level can be reduced rather at the intervention completion, thus showing short-term effect. Yet the study of Baker et al. (2012) registered positive results of intervention at 6 months follow up, whereas the studies of Chan et al. (2006), Phillips et al. (2011), Hsiao et al. (2012) have presented that CBSM and BMS interventions strongly predict decreased cortisol levels up to 8 or 12-months follow up respectively. The evidence that intervention is effective in applying stress reducing and relaxation habits after 6 months and plateauing afterwards suggests that there might be a need for maintenance sessions.

According to our knowledge no other meta-analysis has dealt with this specific and comparable biomarker of cortisol for a specific target population and grouping certain types of psychosocial interventions that dealt with determined focus of positive assertion, hope, encouragement, and self-regulation techniques (as opposed to emphasizing fears). The study's limitations conclude small sample size of each study, not following randomization principle in some studies, high risk of bias, some heterogeneity of the groups and quality of support in interventions. Due to small sample size, the source of heterogeneity through sensitivity analysis, meta regression and subgroup analysis was not possible to perform. This implies the necessity of further investigation in this field to be able to provide a solid base for the preventive care. Finally, to optimize patient care through understanding the relationship between the influence of glucocorticoid signaling on cancer biology to be able to implement adequate psychosocial interventions and revert the disease progression will require stronger coordinated efforts between oncologists and psychology researchers.

Our research is supported by previous similar findings that cannot be ignored falling into three main pillars of scientific evidence.

Firstly, exposure to physiological stress has been proposed to increase cancer risk and recurrence (van der Pompe et al., 1997, 2001), however combined effects of stress accompanying psychological depression and disruption of normal social life that seriously impair BC females' quality of life and wellbeing, when demonstrated, have been inconclusive. Certain personality types may enhance or degrade immune response (Oswald et al., 2006). Chronic stress has been associated with suppression of immune function and elevations in inflammatory activity, and there is evidence that the immune system may not adapt over time, while cortisol's immunosuppressive effect might be involved in cancer development, causing further oncogenic mutations (Thornton et al., 2009). Some studies have revealed that abnormal diurnal cortisol rhythms are related to poor prognoses in BC patients (Spiegel et al., 1998; Sephton et al., 2000). Thus, stress, playing a mediating role in immune modulation, might be involved in cancer development (including recurrence), while psychosocial interventions may conversely reduce stress induced suppression of immunity, hence improving cancer survival by buffering the biological effects of stress (Fife et al., 1996).

Secondly, cortisol is included within inflammatory and immune processes, acting as an anti-inflammatory agent by causing apoptosis through the mitochondrial pathways. It has been extensively used for measuring HPA axis functioning and body's immunity, and can equally demonstrate the link of psychological stimulation, emotions, and physiological changes (both hormonal and immunological). The study of Matousek et al. (2010) indicates, that in conjunction with other biological measurements, the use of cortisol levels as a physiological marker of stress may be useful to validate even self-reported results of attributed intervention. Molecular studies demonstrate a clear overlap between intracellular stress signaling and protumorigenic pathways within breast cells but need to be integrated with other stress-BC research in order to obtain an unambiguous assessment of the potential cause-effect relationship (Antonova et al., 2011). The findings regarding specific pathways of molecular mechanisms and increased glucocorticoid signaling on the immune system and systemic metabolism related to cancer progression are described elsewhere (Volden and Conzen, 2013).

Comprehensive study of cortisol biomarker's variability due to stress modulating psychological interventions might be used as a novel prognostic factor for disease development (progression, recurrence, or recovery). Cortisol is a reliable tool to assess the adrenocortical response, yet easy and non-invasive means to perform a repeated and “non-disturbing” sampling. Salivary cortisol represents the biologically active form of the hormone, and a strong association between levels in blood and saliva has been described in animal studies. Peeters et al. (2011) indicated that total blood cortisol concentrations might account for 80% of salivary cortisol concentrations and vice versa.

A strong basal diurnal rhythm of cortisol exists: it has a tendency to surge in the morning upon waking up, increase 50–60% in the first 30–45 min after awakening, drop rapidly over the first few hours after waking, and then decline more slowly across the day to reach a low point around midnight. Therefore, it requires certain methodological prerequisites in sampling process to avoid confusion factors (circadian variations, age, gender, smoking) (Weibel, 2003). As seen in the results of this study, the temporal variability of salivary cortisol is higher than that in the plasma cortisol added to the overall diurnal tendency of fluctuation, and it is more complicated to assess the accuracy of the measurements. In our research most of the included studies followed rigorous methods in collecting salivary or plasma cortisol (using high sensitivity cortisol enzyme immunoassay or enzyme-linked immune-absorbent assay kit or intra-interassay sensitivity kit) and providing demographic overview of the participants. Also, to control cortisol circadian rhythm in the reviewed studies, several measures have been taken. In our meta-analysis, we have considered the mean cortisol value throughout the day. In this way we tried to control for diurnal changes in cortisol levels. In the studies where blood cortisol was measured, only one cortisol measurement was collected as this procedure is more invasive than the saliva sample. However, in all of them the blood was taken at the same or nearly the same time of the day. This also decreases the effect of variability in blood cortisol levels throughout the day.

Thirdly, past research in PNI has been investigating various sides of this large topic: there are some extensive narrative reviews dedicated to the subject of PNI in general (Mulder et al., 1992; Bauer, 1994; Garssen and Goodkin, 1999; Schleifer, 2007; Wahbeh et al., 2009; Subnis et al., 2014; Jassim et al., 2015; Hulett and Armer, 2016). Further, the effect of psychological and psycho-behavioral interventions on cancer in general and on BC patient's immune function (Tong et al., 2014; Zhang et al., 2016, 2019; Guarino et al., 2020). Finally, the impact of psychosocial interventions on survival time in cancer patients (Chow et al., 2004; Smedslund and Ringdal, 2004; Oh et al., 2016). Some of these published meta-analyses used specific neuroendocrine biomarkers as outcome variable and as predictors for recurrence of BC (Sephton et al., 2000; Tong et al., 2014; Zhang et al., 2016, 2019).

To understand the ambiguity of this large and complex topic, we have further examined that several previous reviews show little or no evidence that psychosocial interventions improve survival time (Jassim et al., 2015) or demonstrate only short-term improvement in survival (Chow et al., 2004). Although there is a considerable evidence of psycho-behavioral interventions' effect on immune indicators, the effect size of meta-analysis of Tong et al. (2014) is modest and the authors highlighted that intervention focus and comparable immune measures should be taken into consideration for future research. Due to heterogeneity in various studied aspects, meta-analysis in some of these reviews was not possible (Mulder et al., 1992; Hulett and Armer, 2016).

Even though the results of the reviews of Tong et al. (2014), Hulett and Armer (2016), and Zhang et al. (2016) are inconclusive, the authors suggest that interventions with mindfulness-breathing-stretching components, cognitive–behavioral therapies or relaxation training appear to offer potential improvement or stabilization of neuroendocrine-immune activity through PNI interactions in patients with cancer compared to control groups. There is growing evidence that psychosocial interventions improve the wellbeing of cancer patients, producing favorable effects on some psychological outcomes, such as anxiety, depression, and mood disturbance. Most of the studies included in the above meta-analyses suffered from small sample size and poor compliance, especially in studies involving patients with advanced disease. Although psychosocial interventions appear to be promising adaptation therapy against many of the psychophysiological impacts of cancer, according to the review of Smedslund and Ringdal (2004) they should be thoroughly defined, focus on a single diagnosis, and consider known risk factors.

If chronic psychological distress is associated with health-relevant behavioral and biological processes in cancer patients, then it is plausible that certain types of psychosocial interventions aimed to reduce distress and promote psychological adaptation could possibly influence health outcomes particularly in BC patients (McGregor and Antoni, 2009). The effect of psychological factors has been more related to cancer progression, recurrence and metastasis rather than cancer initiation (Mulder et al., 1992). The preliminary evidence of mindfulness based, or similar psychosocial supportive care, more convincingly in interventions, shows the reduction of pro-inflammatory cytokine production response in cancer patients that may strengthen coping capacity and improve BC survivorship. Relating stress biology and pro-metastatic molecular processes, clinical researchers are now looking to pioneer further development in research to mitigate adversity of cancer recurrence in part by down-regulating similar pathways (Lutgendorf et al., 2010) and link effects of chronic stress on BC progression (Sloan et al., 2010).

Even though it is still premature to conclude the effectiveness of these interventions among cancer patients, there is a call to acquire evidence of association between interventions that improve emotional wellbeing, quality of life, and objective indicators like survival /mortality/relapse/recurrence and immune responses. The evidence for efficacy of psychosocial interventions in cancer care is unequivocal, however, there is a need for an objective marker of improvement in research evaluating specific psychosocial programs standardizing content, duration, frequency, follow up and measurement methodology. Additional caveats should include analysis of the type, focus and quality of patient's support network and more personalized treatment plan. Further studies of high quality and large sample sizes are required for more solidified conclusions. We continue to explore this domain of PNI of cancer where immune reactions are more sensitive and responsive (yet measurable) to psychological influences aimed at improving overall health outcome. Results of this meta-analysis could be a promising steppingstone to grant evidence-based foundation in routine medical care.

Our study has concluded that certain types of psychosocial interventions reduce cortisol (indicator of chronic stress) in patients with BC. Application of specific psychosocial support as adjuvant non-invasive therapy for affected females with BC at all phases of treatment could contribute to more cost-effective health care.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

EM, NG-C, and RL-G contributed to conception and design of the study and wrote the first draft of the manuscript. EM-L and MA-S wrote sections of the manuscript. AM organized the database. MG-B performed the statistical analysis. All authors contributed to manuscript revision, read, and approved the submitted version.