Brain regions and pathways implicated in stressor appraisals

In the context of medical conditions, one particularly salient cognitive appraisal construct—illness uncertainty—has been proposed as a common phenomenon characterizing the experience of a life-threatening or chronic diseases, like cancer.Reference Mishel^25–Reference Wright, Afari and Zautra²⁸ Illness uncertainty includes appraisals such as the ambiguity and unpredictability of the disease course and prognosis, issues related to treatment complexity, and lack of information regarding the illness and treatment. The brain–body pathways related to meanings of uncertainty were recently described by Peters et al.Reference Peters, McEwen and Friston²³ Accordingly, a stress response results when the person perceives environmental changes and she/he is uncertain of a strategy that can be used to protect physical, mental, or social well-being. This appraisal process (ie, what is threatening to the individual) is originated in the brain after receiving both sensory and viscerosensory inputs. The lateral prefrontal cortex (lPFC), pre-supplementary motor area (pre-SMA), ventromedial prefrontal cortex (vmPFC), and orbitofrontal cortex (OFC), respectively, are where the subjective meanings about the current state of the world and the body seem to be represented.Reference Peters, McEwen and Friston²³

How are physiological, behavioral, cognitive, and emotional responses initiated following the appraisal of the perceived stressor? In accordance with this model, the anterior cingulate cortex (ACC) “makes” the risk assessment of a certain state and selects a strategy. If this strategy “resolves” the current state of uncertainty, the pre-SMA and primary motor cortex initiate the respective behavioral response. If the individual has a state of uncertainty about what has to be done and the situation (ie, changes in the internal body or the external environment) remains threatening, the ACC stimulates the amygdala (Amg), which increases Locus coeruleus (LC), sympathetic nervous system (SNS), and hypothalamic–pituitary–adrenal axis (HPA axis) activity; this initiates the biological stress response. The ACC-Amg complex is crucial since it stimulates 2 important descending projections: (1) the pathway to the LC, which releases norepinephrine, which then causes a hypervigilant state, increasing “attention”; and (2) the pathway to the ventromedial hypothalamus (Hy) and the paraventricular nucleus, stimulating SNS and HPA axis, secreting glucocorticoids (cortisol in humans) from the adrenal cortex and epinephrine from the adrenal medulla in response to stress.

Physiological sequelae of stress responses

Both acute and chronic stress alter brain structure and function and can produce peripheral changes in the cardiovascular, immune, and endocrine systems.Reference Karatsoreos, McEwen, Fink, Pfaff and Levine^29–Reference Davidson and McEwen³² The brain is considered the key organ of the response to stress, because it determines both the meaning of the threat and the physiological and behavioral responses.Reference McEwen³³ Two key stress-response subsystems mediating these effects are the autonomous nervous system (ANS) and the HPA axis.Reference McEwen and Gianaros³⁴ Moreover, as was referred to above, the brain complex formed by the ACC and the Amg has a main role in the initial activation of these subsystems.Reference Peters, McEwen and Friston²³ A study with rhesus monkeys supports the crucial role of ACC, particularly the subgenual area, in HPA output.Reference Jahn, Fox and Abercrombie³⁵ Jahn et al.Reference Jahn, Fox and Abercrombie³⁵ suggested that individuals with elevated activity in the subgenual anterior cingulate cortex (sgACC) may show heightened cortisol levels and may be at risk for stress-related HPA dysregulation.

Processing of stressful events is, thus, mediated by the nervous system, which regulates physiology through the “participation” of the immune, neuroendocrine, and autonomic systems. These brain–body pathways linking cognitive (stressor) appraisal processes to changes in brain structure and function, and reactions in the cardiovascular, immune and endocrine systems, may increase risk for physical disease or may aggravate existing disease. According to the uncertainty stress model, some brain regions are proposed as having a key role in the stress response, such as the LC, insula (Ins), prefrontal cortex (PFC), cingulate cortex (CC) (mainly the ACC and the subgenual area), and Amg. Neuroimaging studies in a variety of medical populations have now accrued to the point that one can observe reliable patterns of association between distress states and changes in the structure and functional activity of specific brain regions and networks. One subset of these studies in cancer patients that has examined the association of distress states with brain activity has identified specific brain regions that overlap with those proposed in the uncertainty stress model. This research is reviewed below.

Distress and brain function (¹⁸F-FDG-PET studies)

The studies presented in Table 1 aimed to assess associations between distress (depression and/or anxiety) and rCMRglc in specific brain regions of patients with various types of cancer. Only 1 studyReference Fang, Yao and An⁵⁰ examined the association of distress with the metabolism of brain networks.

Most of studies were conducted by the same team of researchers.Reference Tashiro, Kubota and Itoh^36, Reference Tashiro, Itoh and Kubota^51–Reference Inagaki, Yoshikawa and Kobayakawa⁵⁶ The participants were cancer patients with various types of cancer, and the control group was composed of patients with benign diseases. To assess levels of distress (anxiety and/or depression), investigators used questionnaires or interviews such as the Self-Rating Depression Scale,Reference Zung⁵⁷ Taylor’s Manifest Anxiety Scale,Reference Taylor⁵⁸ Structured Clinical Interview for DSM-IV-Axis I Disorder, Hamilton Depression Rating Scale,Reference Hamilton⁵⁹ and the State and Trait Anxiety Inventory.Reference Spielberger⁶⁰ Across all these studies the brain regions that revealed abnormal metabolism in cancer patients in association with higher distress scores were the OFC, Ins, posterior and ACC, Th, and basal ganglia.

A study by another teamReference Castelli, Tonello and D’Agata⁶¹ used the National Comprehensive Cancer Network’s Distress Thermometer (DT)⁶² and the Hospital Anxiety and Depression Scale (HADS)Reference Zigmond and Snaith⁶³ to assess levels of distress in 21 patients with various types of cancer. The DT has been adopted internationally as a screening measure to identify and address psychological distress in individuals with cancer. The study provided provocative evidence that distress levels assessed by the DT were the best predictor, among all the other clinical and demographic variables, of hypothalamic metabolism, a key indicator of the HPA axis. Specifically, DT scores were positively correlated with rCMRglc in the hypothalamic and midbrain areas immediately below the Th, especially the periventricular areas. In fact, the hypothalamic metabolism was the best indicator to classify patients as distressed according to the DT cut-off score of 4.

Another line of neuroimaging research examines the association of distress with the metabolism of brain networks as opposed to specific brain regions. This follows emerging views in neuroscience that emotional processes are a function of the whole brain and that research should consider the biological organization in the brain, operationalized as large-scale distributed networks.Reference Pessoa^64–Reference Barrett and Satpute⁶⁷ From this perspective, brain structures such as the Amg or the Ins function within a complex network of other brain structures with which they are connected. Accordingly, many neuroimaging studies now aim to examine the properties of whole-brain networks using a graph perspective.Reference Bullmore and Sporns⁶⁸ For example, “nodes” are distinct brain regions and “edges” represent functional connectivity among them. Many conditions, such as depression,Reference Sheline, Price and Yan^69, Reference Kaiser, Andrews-Hanna and Wager⁷⁰ can be described not only as influencing the function and structure of specific brain regions but also as “dysconnectivity” syndromes: disrupting the connectivity patterns among spatially distributed regions of the brain that support normal functioning.

In one recently published connectivity study,Reference Fang, Yao and An⁵⁰ researchers examined abnormalities in resting-state metabolic brain networks using graph analysis in pre-chemotherapy cancer patients with depression (n = 78; 44 had minimal depressions, 16 mild depressions, and 18 moderate/severe depressions) relative to matched healthy non-depressed normal controls (n = 80). The hypothesis was that cancer patients would show altered “small-world properties” and “topological architecture” in the metabolic brain network. Specifically, investigators examined differences in regional metabolism in cancer patients vs controls, and the potential association between regional hypo-metabolism and severity of depression symptoms, assessed by the Beck Depression Inventory (BDI-II).Reference Beck and Steer⁷¹ The whole brain was segmented into 90 regions (45 regions in each hemisphere) to construct a metabolic brain network. Metabolic connections were defined as statistical associations in normalized glucose metabolism between each possible pair of brain regions (metabolic connectivity aims to detect functionally interacting brain regions based on rCMRglc). Four global network properties (clustering coefficient, characteristic path length, small-world attribute, and connectivity patterns) and one regional nodal property (betweenness centrality) were combined to investigate the topological architecture of the metabolic brain network in control and cancer groups. Possible whole-brain metabolism differences between cancer patients and healthy subjects were evaluated on a voxel-by-voxel basis.

The main results were separated into 4 areas:

1. 1. Regions of abnormal glucose metabolism in cancer patients vs controls: a significant decrease in metabolism in cancer patients in the right inferior temporal gyrus, left middle frontal gyrus, left fusiform gyrus, right Th, caudate nucleus, left superior temporal gyrus, Ins, right postcentral gyrus and left superior frontal gyrus, which suggests that the brain networks of cancer patients had declining global efficiency, or the brain regions had less neuronal activity.

2. 2. Correlation between depressive symptoms and metabolism in specific brain regions: Depression severity was negatively correlated with PFC (left inferior frontal gyrus), Th, and cuneus metabolic levels and positively correlated with regional metabolic activity in the parietal, limbic, and temporal lobes, such as the rolandic operculum, Hi, para-Hi, Ins, and Amg. It was suggested that these metabolic abnormalities associated with depression might affect processing in the whole-brain network and thereby influence the social and emotional impairments in the daily life of cancer patients.

3. 3. Global network measures in cancer patients vs controls: The strength of connectivityFootnote ² was lower in many brain regions in the parietal and frontal lobes in cancer patients, which predominantly belonged to long-range connectivity. These changes in connectivity might lead to an increase in regional efficiency and long path length disruption in cancer.

4. 4. Regional network measures in cancer patients vs controls: Cancer patients displayed decreased connectivity (nodal betweenness centralitiesFootnote ³) in several regions of frontal, temporal, and limbic lobes, including the triangular part of the inferior frontal gyrus, the orbital part of the middle frontal gyrus, the olfactory cortex, the Heschl gyrus, the caudate nucleus, the bilateral Hi, the right para-Hi, the Amg, and the ACC. According to the researchers, this pattern might lead to lessened outflow to the thalamus and fusiform gyrus and subsequent hypometabolism in cancer patients. Increased nodal betweenness centralities in cancer patients were mainly located in the middle frontal gyrus, Ins, superior occipital gyrus, and pallidum, and were positively correlated with activity in others (Amg and Hi).

It was concluded that the topological organization of whole-brain metabolic networks may be disrupted in cancer and that this disruption may be more severe in cancer patients with comorbid states of emotional distress such as depression.

Distress and brain structure (MRI studies)

Another set of studiesReference Nakano, Wenner and Inagaki^72–Reference Inagaki, Matsuoka and Sugahara⁷⁴ investigated the correlations between distress states and anatomical changes of neural structures based on the assumption that chronic functional abnormality might be associated with morphological abnormalities. Two studiesReference Nakano, Wenner and Inagaki^72, Reference Matsuoka, Yamawaki, Inagaki, Akechi and Uchitomi⁷³ examined the association between distressing cancer-related recollections (DCRR), regarded as a component of posttraumatic stress disorder (PTSD), and hippocampal and Amg volume, respectively, in breast cancer (BCa) patients. In the first study, 67 patients revealed DCRR versus 39 without, and, in the second, 35 BCa patients had DCRR versus 41 without. In the first study, the volume of the left Hi was significantly smaller in the subjects with a history of DCRR. In the second study, the same authors found that the total (left and right) volume of the Amg was significantly smaller in patients with DCRR vs those without, even after controlling for age, height, and major depressive disorder.

Two other MRI-based studiesReference Inagaki, Matsuoka and Sugahara^74, Reference Yoshikawa, Matsuoka and Yamasue⁷⁵ examined associations between hippocampal, PFC, and Amg volumes and a first major depressive episode (MDE) after cancer diagnosis in female cancer survivors who had undergone BCa surgery 3 or more years earlier. One of these studies (n = 68; 17 with MDE) showed no significant differences in the left or right hippocampal volumes of cancer survivors who did and did not experience a first MDE after cancer diagnosis. The other study (n = 51; 11 with MDE) showed that left Amg volumes in patients with a first minor and/or MDE were significantly smaller than in those with no incident of depressive episode, though there were no significant differences in the volume of the Hi among groups, which is consistent with the results of previous work.Reference Inagaki, Matsuoka and Sugahara⁷⁴ Thus, the authors suggested that Amg volume was associated with a first minor and/or MDE after cancer diagnosis, which is similar to the findings noted above for DCRR.

Thus, this set of studies utilizing MRI to determine changes in the volume (enlargement or atrophy) of certain brain regions associated with distress (depression or intrusive thoughts) suggested that among diagnosed cancer patients, distress may be associated with reduced left Hi and Amg volumes. Interestingly, these 2 regions were referred to, in functional neuroimaging studies, as being involved in metabolic abnormalities in cancer patients with distress.

Conclusions from the review

Strengths

The findings of these structural and functional neuroimaging studies with cancer patients suggest that “distress” factors may relate to changes in function and structure of specific brain regions and networks. While chemotherapy and paraneoplastic factors also affect the brain, the effect of psychological factors may be independent of these.Reference Tashiro, Itoh and Kubota⁷⁶ There are multiple points of convergence in the findings reported in the studies reviewed. First, the key brain regions identified by these neuroimaging studies were similar to those demonstrated in previous imaging studies of non-cancer patients with major depressionReference Tashiro, Itoh and Kubota⁷⁶ and with those proposed in the uncertainty stress model. Second, distress states appear to be associated with the metabolism of brain networks in cancer patients, in studies suggesting that depression might affect processing in the whole-brain network. Third, the severity of depressive symptoms was associated with metabolic activity in brain regions overlapping with those referred to in the other studies with cancer patients, such as the Amg, ACC, Hi, Ins, PFC, and Th. Interestingly, these key brain regions are all included in the brain regions of an important brain network: the interoceptive network (that is made up of 2 overlapping networks: the salience network and the default mode network).Reference Kleckner, Zhang and Touroutoglou^65, Reference Barrett^66, Reference Barrett and Simmons⁷⁷ As will be shown below, this network has crucial functions, such as the control of body metabolism (eg, cardiac output, breathing rhythm, cortisol release) and interoception, which refers to the sensing of the internal state of one’s body.Reference Strigo and Craig⁷⁸ Moreover, this network regulates the autonomic, neuroendocrine, and immune systems, and, as such, its dysregulation may be involved in some physical illnesses such as cancer, on the one hand, and mood disorders on the other.Reference Barrett⁶⁶

Based on the areas of convergence in these studies, we can enumerate key brain regions that potentially might be involved in emotional distress processes in cancer patients. We think that it is reasonable to propose the interoceptive network as a key brain network that should be included in studies investigating brain-mediated biobehavioral processes (Figure 1).

Figure 1.

Key brain regions and brain systems that potentially might be involved in emotional distress processes in cancer patients. These brain regions and the interoceptive network, are involved in the regulation of autonomic, neuroendocrine and immune systems and, as such, its dysregulation may be involved in some physical illnesses such as cancer, on the one hand, and mood disorders on the other.

Curiously, neuroimaging studies about the effects of psychotherapy in the brains of depressed persons have observed changes in many brain regions that overlap with most of those proposed in Figure 1, such as the ACC, prefrontal cortices, Hi, and Amg.Reference Weingarten and Strauman⁷⁹ As we will note below, cognitive-behavior therapy (CBT), a specific model of psychotherapy commonly used with cancer patients, is associated with changes in the activity of the interoceptive system, which may be a critical consideration in studying the brain mechanisms underlying the effects of such interventions on emotional distress and related biobehavioral processes of cancer patients going forward.

Top-down and bottom-up pathways

The top-down paradigm, although it can produce leads and hypotheses for further studies, is restrictive in relation to all the variables and body systems that are involved in adaptation to cancer. To date we have a very limited knowledge about the brain changes that account for the influence of distress in cancer processes. This information could ultimately inform the development of theoretically driven distress management behavioral interventions, which could contribute to enhanced clinical cancer care and improved health outcomes. Studies in this research area should progress to a paradigm that includes multilevel integrative analyses and in which the formulation of structure–function relationships should integrate “top-down” and “bottom-up” information,Reference Cacioppo, Berntson, Berntson and Cacioppo⁸⁵ that is, the study of bidirectional mind–body interactions, including behavioral and social factors, brain activity and structure, neural peripheral systems, and the endocrine and immune systems.

Brain networks and body metabolism

The executive homeostatic network was proposed to explain the bidirectional bottom-up and top-down pathways between “mind” and body and is an example of an integrative model.Reference Taylor, Goehler and Galper⁸⁶ In this network, and in a simplified description, the bottom-up pathways have their influence via parallel ascending projections within the spinal cord and brain. Vagal sensory pathways and spinal visceral and somatic sensory pathways provide information regarding inflammation, pain, and other important conditions to regulatory brain regions including the Hy and Th, which ultimately reach components of the executive homeostatic network, comprised of the ACC, PFC, and insular cortex. Accordingly, information regarding bodily states gives rise to the experience of symptoms associated with illness and injury.

In turn, top-down pathways influence bodily functions via modulation of sympathetic, vagal, spinal, and adrenal output, which may interact with peripheral tissue, including immune and tumor cells in the circulation and the tumor microenvironment, via neuroendocrine (norepinephrine (NE), epinephrine (E), and cortisol) ligands. Changes in the state of these cells (eg, inflammation) can feed back to the brain (bottom-up), for instance, via cytokines. For example, interleukin-1 (IL-1) receptors were found in the Hi and Hy and brain stem.Reference Goshen and Yirmiya⁸⁷ This cytokine has an important role in memory loss and suppressed neural plasticity in cases of infection, injury, and severe emotional distress.Reference Yirmiya and Goshen⁸⁸ Empirical evidence indicates that that there is communication between peripheral cytokines and cytokines inside the central nervous system (CNS) that stimulate neuronal and supportive cells to release cytokines in the brain.Reference Kronfol and Remick^89–Reference Quan and Banks⁹¹ Cytokines produced in those brain areas activate the HPA axis and SNS, participating in a network of reverberating feedback loops within and between the multiple levels of brain–body system (eg, peripheral sensory-to-brain-to-peripheral-sensory).

More recently this network was called the interoceptive network. Reference Kleckner, Zhang and Touroutoglou^65, Reference Barrett and Simmons⁷⁷ As noted previously, distress-related brain regions (Amg, ACC, Hi, Hy, Ins, PFC, and Th) are included in this important network. In fact, it includes (1) cortical regions, such as the CC, medial PFC (including both the ventral and dorsal sectors), anterior Ins, mid- and posterior Ins (primary interoceptive cortex), inferior frontal gyrus, supplementary motor cortex, superior temporal sulcus and temporal pole, and the parahippocampal gyrus; and (2) subcortical regions, such as the Amg, ventral striatum, Th, Hy, periaqueductal gray (PAG), parabrachial nucleus (PBN), and nucleus tractus solitarius.Reference Kleckner, Zhang and Touroutoglou⁶⁵ These brain regions are proposed to constitute the functional and anatomical substrates of an allostatic-interoceptive system, that is, a system that continually anticipates the body’s energy levels and demands, and prepares to meet those needs before they arise.Reference Barrett and Simmons⁷⁷

Considering the brain as a predictive internal model of the world, in which mental events are generated as predictions, not reactions,Reference Hoemann, Gendron and Barrett⁹² the cortical regions that are the substrates of the allostatic-interoceptive system make visceromotor predictions through their connections to the Hy, the periaqueductal gray (PAG), and other brainstem nuclei (eg, PBN, Nucleus Tractus Solitarius (NTS)), which control the body’s internal milieu to regulate the autonomic, neuroendocrine, and immune systems.Reference Kleckner, Zhang and Touroutoglou⁶⁵

Thus, this allostasis model assigns a central role to the brain, considered an organ for predictive regulation. Contrary to the classical model in which the regulation of the internal milieu is based on the preservation of its constancy (homeostasis), in allostasis the maintenance of an efficient regulation requires anticipating needs and preparing to satisfy them before they arise.Reference Sterling⁹³ These brain predictions are top-down influences. Bottom-up influences are ascending sensory inputs from the body’s organs and systems (eg, levels of insulin or cortisol, heart rate, inflammatory cytokine levels), which are interoceptive sensations about the state of the body.Reference Barrett and Simmons⁷⁷ When body metabolism is chronically unbalanced, like when individuals are chronically anticipating a threat (anxiety) or loss (depression), having chronic perceptions of social adversity, or other experiences that may accompany cancer diagnosis and treatment, the brain might regularly predict that individuals need more energy than their body requires. This can cause a problem: these predictions, based on a chronic “false” alarm, cause the body to release hormones, like cortisol, which over time promotes leukocyte glucocorticoid receptor resistanceReference Miller, Cohen and Ritchey⁹⁴ and exacerbates inflammatory signaling, causing a vicious cycle in prediction.Reference Barrett^66, Reference Sterling⁹³ Dysregulation within the interoceptive system may be involved in some physical illnesses and mood disorders as well, since they share a common neural substrate.Reference Barrett and Simmons⁷⁷

Interestingly, as noted previously, CBT, a therapeutic intervention designed to modify cognitive appraisals in order to improve affect and behavior, is associated with changes in the activity of the interoceptive system, such as decreased activity in agranular cortex in depressionReference Goldapple, Segal and Garson⁹⁵ and enhanced ACC activity and decreased activity of the insular cortex, dorsolateral prefrontal cortex (dlPFC), and Amg in PTSD.Reference Aupperle, Allard and Simmons⁹⁶ These changes in the activity of brain regions of the interoceptive network accompanying CBT may, in part, explain the effects of psychosocial interventions on psychological adaptation, stress-related biobehavioral processes, and disease progression in cancer patients.Reference Antoni⁶ This follows from the notion that the interoceptive system not only controls the body’s internal milieu to regulate the autonomic, neuroendocrine, and immune systems but is also the main source of our phenomenology and feelings.Reference Barrett^66, Reference Damasio and Carvalho⁹⁷

Interoception and sickness behaviors

Interoception refers to the perception of autonomic, hormonal, and immunological homeostatic signals that describe the physiologic state of the body.Reference Barrett and Simmons^77, Reference Craig^98, Reference Ceunen, Vlaeyen and Van Diest⁹⁹ Interoceptive information is conveyed to the CNS through spinal, cranial, and humoral homeostatic pathwaysReference Craig, Lewis, Haviland-Jones and Barrett^100, Reference Critchley and Harrison¹⁰¹ and is integrated with learned associations, memories, and emotions, creating a subjective experience (ie, feeling).Reference Ceunen, Vlaeyen and Van Diest⁹⁹ Thus, feelings are subjective experiences that accompany a change in body stateReference Barrett^66, Reference Damasio and Carvalho⁹⁷ and have an intensity and a valence (positive or negative, pleasant or unpleasant, or neutral), influencing motivation.Reference Barrett⁶⁶ The experience of unpleasant feelings resulting from a change in a body state can be illustrated by immune-to-brain communication, namely the neural and humoral pathways that transduce immune signals through mediators called pro-inflammatory cytokines, from the periphery (bottom-up influences) to brain regions such as the nucleus of the tractus solitarius, PBN, the hypothalamic paraventricular and supraoptic nuclei, central Amg, and the bed nucleus of the stria terminalis.Reference Dantzer, O’Connor and Freund¹⁰²

These cytokines include IL-1α, IL-1β, TNF-α, and IL-6, and are responsible for a constellation of behaviors called sickness behaviors, which influence emotionReference Critchley and Garfinkel¹⁰³ through motivational changes that include anhedonia, reductions in social affiliative behavior, and irritability.Reference Dantzer, O’Connor and Freund^102, Reference Dantzer and Kelley¹⁰⁴ Thus, when we get a viral or bacterial infection, and possibly cancer, we experience symptoms (eg, sad, nauseated, reduced appetite and motivation). These cytokines may induce depressive symptoms and are also associated with compromised cognitive performance (eg, impaired attention and short-term memory).Reference Dantzer, O’Connor and Freund^102, Reference Walker, Kavelaars and Heijnen¹⁰⁵ In fact, there is evidence for the crucial role of cytokines in mediating the effects of cancer itself and cancer-related therapy on cognitive dysfunction (ie, the subjective experience when one has deficits in his or her cognitive function) and cognitive impairment (ie, objectively measured cognitive deficits).Reference Wang, Walitt and Saligan¹⁰⁶ This post-treatment cancer-related cognitive dysfunction or impairment has been called “chemobrain” due to the attribution of these symptoms to chemotherapy. However, chemobrain may be caused by multiple factors such as cancer itself, chemotherapy, secondary medical conditions (ie, comorbidities), and emotional distress, which may affect patient cognition and brain function and structure.Reference Bompaire, Durand and Léger-Hardy¹⁰⁷ In fact it has been proposed that chemobrain would be more accurately labeled “cancer- or cancer-therapy-associated cognitive change.”Reference Hurria, Somlo and Ahles¹⁰⁸ One can view chemobrain symptoms as a type of sickness behavior in cancer patients, and one that may be mediated by inflammatory signaling. The role of pro-inflammatory cytokines has been enlightened, indicating their crucial role in bottom-up and top-down influences occurring in chemobrain syndrome.Reference Cheung, Ng and Shwe¹⁰⁹ Actually, there is evidence that underlying chemobrain cognitive dysfunction are peripheral cytokines that stimulate neuronal and supportive cells that signal the release of central cytokines, which act to alter neuronal plasticity and brain functioning,Reference Wang, Walitt and Saligan^106, Reference Cheung, Ng and Shwe¹⁰⁹ which, in turn, affects peripheral biobehavioral processes such as the activation of the HPA-axis, resulting in the production of high levels of cortisol to down-regulated immune activity.Reference Yirmiya and Goshen⁸⁸

Some brain areas and networks were proposed as being associated with chemobrain and some (CC, PFC, and Hi) overlap with distress-related brain regions. A review of breast cancer chemotherapy-related cognitive dysfunction studies found evidence for reduced connectivity in the default mode network (DMN), suggesting that the cognitive dysfunction may represent a brain network disorder and that altered DMN activity may serve as a potential biomarker of chemotherapy-related brain injury.Reference Kesler¹¹⁰ This network includes brain regions such as the precuneus, posterior cingulate, medial frontal, middle temporal, and lateral parietal regions, as well as Hi,Reference Buckner, Andrews-Hanna and Schacter¹¹¹ and is part of the interoceptive network, which can indicate that this network can be also potentially disrupted.

A reviewReference Simó, Rifà-Ros, Rodriguez-Fornells and Bruna¹¹² of structural and functional neuroimaging chemobrain studies concluded that although there were many contradictory results across studies, there was a convergence in some findings in cancer patients: (1) a diffuse decrease of gray matter (GM) and white matter (WM) volume together with a decrease of the overactivation in frontal regions in chemotherapy-treated patients compared to controls and (2) a long-term persisting decrease in GM and WM volumes together with a predominantly frontal cortex hypoactivation in only a subgroup of chemotherapy-treated patients. However, due to the complexity of chemobrain syndrome and the multiple factors involved, more studies are needed to better understand the mechanisms underlying the cognitive changes reported by cancer survivors.Reference Hurria, Somlo and Ahles¹⁰⁸

In conclusion, there are bidirectional bottom-up and top-down pathways between brain and the periphery that allow the communication among bodily systems in an integrated way. The research about the effects of psychological and emotional factors on peripheral stress-related biobehavioral pathways known to contribute to cancer progression and long-term health outcomes should progress to a more integrative analysis. Developing novel ways to integrate data from psychosocial assessments, neuroimaging studies of brain functioning, and peripheral indicators of neuroendocrine and immune/inflammatory processes will be key to the development of this line of work. The next logical step is to examine the association of distress and affect-related brain region activity with these peripheral immunologic processes relevant for the pathophysiology of cancer, in order to determine the relevance of specific brain regions and networks in the progression and/or metastasis of cancer. Establishing an evidence base for brain regions that are predictive of biobehavioral processes and disease outcomes may be relevant for future cancer research. As will be presented next, there is a growing body of evidence, in the field of psychoneuroimmunology (PNI), showing that distress and other indicators of adversity are associated with SNS- and HPA-mediated biobehavioral processes that contribute to biological activities that can promote cancer growth and metastasis.

Concepts and applications

The last decade has produced substantial evidence that psychosocial variables like distress, depression, lack of social support, and trauma history are associated with biological alterations that could promote cancer progression.Reference Antoni^6, Reference Costanzo, Sood and Lutgendorf^113, Reference Lutgendorf and Andersen¹¹⁴ While there is some evidence for the contribution of these distress-related biobehavioral factors to cancer initiation,Reference Lutgendorf and Andersen¹¹⁴ there is a stronger evidence base for links between such biobehavioral factors and cancer progression once a tumor has been established.Reference Costanzo, Sood and Lutgendorf^113, Reference Lutgendorf and Andersen¹¹⁴ Biobehavioral factors affect cellular immunity and modulate fundamental processes in cancer growth, such as inflammation, angiogenesis, invasion, and metastasis.Reference Lutgendorf and Andersen¹¹⁴

Studies documenting bidirectional communication among the different body systems (psychosocial, neural, neuroendocrine, and immune) have been reported over the last 4 decades in the field of PNI,Reference Ader, Cohen and Felten¹¹⁵ a research domain that has more recently been applied to cancer research.Reference Lutgendorf and Costanzo^116, Reference McDonald, O’Connell and Lutgendorf¹¹⁷ A growing body of PNI work has demonstrated that the interaction between psychosocial processes (eg, interpretation of events, social support, negative affect), sociodemographic factors (eg, sex, age, race, ethnicity, socioeconomic status), and health behaviors (eg, sleep, diet, exposure to viruses, screening behaviors) may influence cancer progression, treatment-associated symptoms, recovery, and survival after treatment, as well as long-term quality of life.

PNI research has identified some of the processes underlying these connections involving neuroendocrine (eg, cortisol and norepinephrine changes) and immune/inflammatory mechanisms (eg, using measures of leukocyte cytokine and chemokine gene expression and circulating pro-inflammatory cytokines such as IL-1β, IL-6 and TNF-α). This body of work has been used to generate multiple biobehavioral models of cancer that can be used to guide integrative analyses of distress, biobehavioral processes, and health outcomes in cancer patients and survivors.Reference Costanzo, Sood and Lutgendorf^113, Reference Lutgendorf and Andersen^114, Reference McDonald, O’Connell and Lutgendorf^117, Reference Antoni, Lutgendorf and Cole¹¹⁸

Biobehavioral models of cancer control emphasize that perceptions of events that are evaluated as threatening or stressful are associated with activation of limbic-cortical structures of the CNS, dysregulation of the HPA axis, and increased signaling of the SNS.Reference Cole, Nagaraja and Lutgendorf¹¹⁹ For example, studies report HPA axis abnormalities such as flattening of the circadian rhythm of cortisol secretion and elevated plasma cortisol levels in women with metastatic BCa (mBCa),Reference Abercrombie, Giese-Davis and Sephton¹²⁰ and other reports indicate that mBCa patients with flattened or abnormal diurnal cortisol rhythms had earlier mortality.Reference Sephton, Sapolsky and Kraemer¹²¹ Other work indicates an association between flattened or abnormal diurnal cortisol patterns and increased risk of mortality in patients with Stage IV renal cell carcinomaReference Cohen, Cole and Sood¹²² and lung cancer.Reference Sephton, Lush and Dedert¹²³ Thus, among patients with advanced cancers, abnormalities in HPA axis functioning may promote hastened mortality. Importantly, increased SNS signaling and HPA axis dysregulation may produce alterations not only in cortisol but also in catecholamines such as norepinephrine, which can act as ligands for immune cells, to down-regulate cellular immune function and up-regulate pro-inflammatory signaling.Reference Antoni^6, Reference Cole, Nagaraja and Lutgendorf¹¹⁹

Some of this work has focused on the context of primary BCa treatment, given that BCa is one of the most prevalent cancers worldwideReference Siegel, Miller and Jemal¹²⁴ and since women diagnosed at this point in the disease can showed marked variability in clinical outcomes over time. Among early stage BCa patients in the weeks after surgery, negative affect and depressive symptoms have been associated with greater levels of serum IL-1β, IL-6, and TNF-α,Reference Bouchard, Antoni and Blomberg¹²⁵ and with up-regulated leukocyte gene expression for these cytokines.Reference Antoni, Lutgendorf and Blomberg¹²⁶ Greater depressive symptoms in the weeks after surgery in this cohort were shown to predict shorter survival over an 11-year median follow-up.Reference Antoni, Jacobs and Bouchard¹²⁷ Conversely, greater decreases in leukocyte pro-inflammatory and increases in anti-viral gene expression over the first year of treatment predicted a longer disease-free interval over this 11-year follow-up.Reference Antoni, Bouchard and Jacobs¹²⁸ Thus, one way that stress factors (negative affect and depressive symptoms) can influence cancer progression is through their influence on the expression of adversity-associated genes in circulating immune cells and associated cytokines.Reference Cole and Sood¹²⁹

Distress, regulation of leukocyte and tumor cell gene expression, and cancer progression

Studies from the recently developed field of human social genomics Reference Cole¹³⁰ show that changes in the expression of hundreds of genes can occur as a function of the physical and social environments that we live in.Reference Slavich and Cole^131, Reference Miller, Chen and Cole¹³² These effects can result from physical threats but can also be a response to symbolic or imagined adversities. That is, our subjective meanings about our socio-environmental conditions can regulate the expression of broad sets of genes. We now know, for example, that extended periods of distress emanating from subjective evaluation of social-environmental threats or adversity are linked to co-occurring changes in patterns of SNS and HPA axis activity and alterations in the transcriptional programs expressed in immune and tumor cells, among others. This control of human gene expression by psychological and social stress is characterized by a pattern of upregulated proinflammatory activity and down-regulated antiviral immune response activity in leukocytes referred to as the conserved transcriptional response to adversity (CTRA).Reference Cole, Nagaraja and Lutgendorf^119, Reference Cole¹³⁰ These changes, when chronically expressed, can result in a variety of inflammation-related diseases such as cardiovascular disease and depression, caused by excessive proinflammatory immune response gene expression, and/or vulnerability to viral infections such as upper respiratory disease, caused by insufficient antiviral immune response gene expression.Reference Cohen, Doyle and Skoner¹³³

Emerging work suggests that psychological adversity may be related to an increased CTRA pattern of leukocyte gene expression in cancer patients. One study of women with non-metastatic BCa showed that greater negative affect (relative to positive affect) was associated with greater leukocyte pro-inflammatory (IL1, IL6, TNF) and pro-metastatic (eg, MMP9) gene expression in the weeks after surgery.Reference Antoni, Lutgendorf and Blomberg¹²⁶ These alterations can be the result of both the disease and its treatment-related effects on endocrine regulation, as well as the elevated stress of diagnosis and treatment of BCa.Reference Antoni, Lutgendorf and Cole^118, Reference Antoni, Lechner and Diaz^134, Reference Van Der Pompe, Antoni and Heijnen¹³⁵

Conversely, facilitating adaptation with group-based cognitive behavioral stress management (CBSM) intervention has been shown to decrease leukocyte pro-inflammatory and pro-metastatic gene expression and increase anti-viral gene expression (interferon response genes) in women undergoing primary treatment for BCa.Reference Antoni, Lutgendorf and Blomberg¹²⁶ Importantly, this intervention was also shown to reduce negative affect and depressive mood; increase positive affect, social support, and relaxation skills; decrease afternoon serum cortisol levels; and increase cellular immune functional responses to in-vitro challenge.Reference Antoni, Lutgendorf and Blomberg^126, Reference Antoni, Lechner and Diaz^134, Reference Antoni, Lechner and Kazi^136, Reference Phillips, Antoni and Carver¹³⁷ Subsequently, bioinformatics analyses of these leukocyte gene expression changes implicated decreases in the NFκB transcriptional activity and increases in glucocorticoid receptor (GR) sensitivity as potentially mediating the pro-inflammatory and pro-metastatic gene expression changes over the 12-month period after CBSM.Reference Antoni, Lutgendorf and Blomberg¹²⁶ Increases in GR sensitivity were hypothesized to follow from the decreased circulating cortisol levels documented in women who received CBSMReference Phillips, Antoni and Lechner¹³⁸ and may herald a reversal in the negative-affect–associated HPA dysregulation.

As noted above, there was also evidence that women assigned to CBSM showed a leukocyte gene expression profile consistent with an increased expression of key anti-viral immune genes such as Type I and II interferons, which is in line with prior reports that CBSM increases lymphocyte production of IFN-γ from stimulated lymphocytes in these women.Reference Antoni, Lechner and Diaz¹³⁴ This suggests that the intervention may have hastened a recovery of cellular immune signaling relevant to defending against opportunistic infections and micrometastatic tumor cells during the first year of primary treatment.

Together this series of analyses converges on the notion that greater negative affect and lower positive affect states are related to a CTRA pattern identified in other populations, and that stress management interventions can reverse the CTRA pattern during a crucial point in cancer treatment.Reference Slavich and Cole¹³¹ But do these forms of intervention produce measurable improvements in clinical outcomes over the long term, and can these effects be attributable to intervention-associated biobehavioral changes?

Distress management interventions

Studies developed by a small number of teams, supporting biobehavioral models of cancer, demonstrated that psychosocial interventions shown to help cancer patients learn stress management may not only reduce depression and anxiety and modulate biobehavioral processes but are associated with greater survival and a longer time until recurrence.Reference Antoni, Lutgendorf and Blomberg^126, Reference Antoni, Bouchard and Jacobs^128, Reference Antoni, Lechner and Kazi^136, Reference Andersen, Farrar and Golden-Kreutz^139–Reference Stagl, Lechner and Carver¹⁴¹ These data have utility because they can culminate ultimately in the development of distress management interventions with the potential to improve both quality of life and clinical disease outcomes through changes in brain structure and function that have implications in peripheral biobehavioral processes that, in turn, will feed back into the brain. For instance, one team showed that a CBSM intervention targeting anxiety-related affective and behavioral processes reduced negative affect and increased positive affect while also reducing CTRA-related leukocyte gene expression over the initial 12 months of BCa treatment.Reference Antoni, Lutgendorf and Blomberg¹²⁶ In a long-term follow-up of this cohort, investigators found that patients assigned to CBSM had longer 11-year overall survival and disease-free survival (longer period until recurrence),Reference Stagl, Lechner and Carver¹⁴¹ and greater reductions in the CTRA leukocyte expression profile during the 12 months of primary treatment predicted a longer disease-free survival over the 11-year follow-up.Reference Antoni, Bouchard and Jacobs¹²⁸

Another team conducted a clinical trial (a 12-month CBT-based intervention with a total of 26 sessions of 1.5 hours in small cohort groups of 8–12 patients) in early-stage BCa patients, before adjuvant therapy, and analyzed the biobehavioral, immune, and health benefits of psychological intervention over a period of 11 years.Reference Andersen, Farrar and Golden-Kreutz^139, Reference Andersen, Yang and Farrar^140, Reference Andersen, Thornton and Shapiro^142–Reference Thornton, Andersen and Schuler¹⁴⁴ They reported that the intervention improved affect and immune function over the initial year of treatmentReference Andersen, Farrar and Golden-Kreutz¹³⁹ and was associated with a reduced risk of BCa recurrence and death from BCa and from all causes over an 11-year median follow-up.Reference Andersen, Yang and Farrar¹⁴⁰ Lower levels of inflammation during the follow-up period predicted lower odds of recurrence.Reference Thornton, Andersen and Carson¹⁴³ Moreover, a subgroup of patients that, on accrual, reported clinically elevated depressive symptoms and were assigned to the intervention showed a significant reduction in depressive symptoms, pain, fatigue, and inflammation biomarkers.Reference Thornton, Andersen and Schuler¹⁴⁴ Finally, among the patients in both arms who did have a recurrence, intervention-arm patients had a reduced risk of BCa death vs controls.Reference Andersen, Thornton and Shapiro¹⁴² Therefore, across 2 independent RCTs of CBT-based distress reduction in non-metastatic BCa patients, changes in biobehavioral processes preceded and were related to long-term clinical outcomes. What sorts of psychological processes or skill changes in these cohorts could have accounted for these biobehavioral alterations?

What psychological changes occurring during stress management interventions can account for changes in body physiology?

Hypothesized positive psychological changes occurring during CBSM that may explain its effects on physiological status include changes in the personal meanings that cancer patients have, reductions in perceived threats, and feeling less vulnerable via increased confidence (self-efficacy) in using personal (eg, relaxation and CBT skills) and interpersonal resources (social support) to cope with the distress.Reference Antoni¹⁴⁵ For example we know that BCa patients assigned to CBSM show increases in finding benefit in the cancer experience (ie, benefit finding),Reference Antoni, Lehman and Kilbourn¹⁴⁶ and greater increases in benefit finding after CBSM are associated with greater reductions in afternoon serum cortisolReference Cruess, Antoni and McGregor¹⁴⁷ and greater increases in cellular immune functioning.Reference McGregor, Antoni and Boyers¹⁴⁸ In addition, BCa patients reporting greater increases in confidence in using relaxation and CBT skills showed greater reductions in afternoon cortisol levels.Reference Phillips, Antoni and Carver¹³⁷

Actually, a basic assumption of CBT and psychotherapy, more broadly construed, is that the meaning a person attaches to a situation plays a key role not only in how that person will feel and behave but also has consequences on their physiology.Reference Beck and Haigh^149–Reference Beck¹⁵³ Thus, according to CBT intervention assumptions, to become more functional and adaptive, a person can reflect or become more aware of her/his mental processes in order to modify her/his interpretations of social and environmental conditions. There is a primacy of cognition in behavior and emotional change: whatever psychosocial method is used to promote patient adaptation (eg, relaxation, exposure, cognitive restructuring, teaching social skills, mindfulness meditation), there is a change in cognition.Reference Beck and Dozois¹⁵¹ Also the administration of antidepressant medication may have an effect on cognition, namely reducing the negative biases in information processing.Reference Harmer, O’sullivan and Massey-Chase¹⁵⁴

The empirical literature has provided support for the efficacy of CBT in the treatment of several psychological disorders (eg, panic disorder, unipolar depression, generalized anxiety disorder, social phobia), as well a physiological benefit in medical patients (eg, changes in inflammatory parameters like reductions in C-reactive protein levels in rheumatoid arthritis patients).Reference Dobson and Dobson^155, Reference Sharpe, Sensky, Timberlake and Ryan¹⁵⁶ Research on patients’ illness perceptions have also shown their relevance in the self-regulation processes of health and illnessReference Leventhal, Phillips and Burns¹⁵⁷ and as a main mediator of more positive adaptation to illness in cancer patients.Reference Jim, Richardson and Golden-Kreutz^5, Reference Collie, Bottorff and Long^158–Reference Lee, Cohen, Edgar and Laizner¹⁶¹ For example, a CBSM intervention tested among 199 women newly treated for stage 0-III BCa reduced reports of cancer-specific thought intrusions, clinician-rated anxiety, and negative affect compared with the control condition.Reference Antoni, Wimberly and Lechner¹⁶² Thus, subjective meanings (eg, threat, loss) of social and environmental conditions may also play a central role in leukocyte transcriptional alterations through the CNS-mediated changes in the regulation of neural signaling and neuroendocrine production previously observed in this cohort.Reference Slavich and Cole¹³¹ With this in mind, it is paramount to understand how stress perceptions and concomitant negative affect alter activities within specific brain regions and/or patterns of activity across brain regions,Reference Geuter, Lindquist, Wager, Cacioppo, Tassinary and Berntson⁸⁰ which in turn contribute to the regulation of SNS and HPA axis hormones in cancer patients.

Distress management effects on comorbidity

Another potential beneficial effect of distress management in cancer patients is on mitigating the risk of and effects of comorbidity or multimorbidity.Reference Ritchie, Kvale and Fisch^163, Reference Sarfati, Koczwara and Jackson¹⁶⁴ Depression, fatigue, sleep disturbances, and cognitive dysfunction are common in cancer patients and have an adverse impact on quality of life and cancer survival.Reference Fu, Axelrod and Guth^165, Reference Søgaard, Thomsen and Bossen¹⁶⁶ For example, insomnia associated with cancer has been linked to decreased quality of life,Reference Fu, Axelrod and Guth¹⁶⁵ impaired immune function,Reference Blask, Hill and Dauchy^167, Reference Payne, Piper, Rabinowitz and Zimmerman¹⁶⁸ cancer-related fatigue,Reference Roscoe, Kaufman and Matteson-Rusby¹⁶⁹ and may have implications for tumor progressionReference Cash, Sephton and Chagpar^170–Reference Filipski, Li and Lévi¹⁷² and greater mortality risk.Reference Steel, Terhorst and Collins¹⁷³

These secondary medical conditions (eg, sleep disruption) can aggravate inflammatory processes and stimulate the release of central cytokines, which act to alter neuronal plasticity and brain functioning, affecting peripheral biobehavioral processes such as the activation of the HPA axis and the SNS, resulting in the production of high levels of cortisol. Thus, stress management interventions might help to reduce, mitigate, or prevent chemobrain syndrome indirectly by mitigating the risk of and effects of comorbidity or multimorbidity.

Other important comorbidities are those associated with metabolic syndrome (MetS) (ie, dyslipidemia, hypertension, central obesity, and insulin resistance),Reference Basaria, Muller and Carducci^174, Reference Felicetti, Fortunati and Brignardello¹⁷⁵ which are risk factors of cardiovascular disease (CVD)Reference Mottillo, Filion and Genest¹⁷⁶ and should be considered in cancer survivors. In fact, a newly published statement from the American Heart Association notes that CVD and associated risk factors (obesity and dyslipidemia) are increasing in BCa survivors, and the development of these risk factors is higher in older BCa survivors than the risk of tumor recurrence.Reference Mehta, Watson and Barac¹⁷⁷ Distress management interventions might have a role in preventing, reducing, or mitigating the effects of these comorbidities, since chronic psychosocial stress, which causes dysregulation of HPA axis, is a risk factor for the development of the MetS.Reference Chrousos^178–Reference Miller, Ancoli-Israel and Bower¹⁸⁰ For example, a 1-year, CBT-based stress management of 20 2-hour sessions over 1 year with cardiac patients showed less fatal and nonfatal first recurrent CVD events and fewer recurrent acute myocardial infarctions than the control group.Reference Gulliksson, Burrell and Vessby¹⁸¹ Given the increase in CVD risk factors after BCa treatment, it is reasonable that distress management interventions might offer the dual benefit of decreased BCa recurrenceReference Antoni, Bouchard and Jacobs^128, Reference Stagl, Lechner and Carver¹⁴¹ and less risk for developing comorbid CVD-related clinical outcomes, thereby contributing to longer survival and better quality of life.Reference Antoni⁶ Since CVDReference Libby, Ridker and Maseri¹⁸² and BCa recurrenceReference Cole, Nagaraja and Lutgendorf¹¹⁹ may both be hastened by shared inflammatory processes (eg, via myeloid cells in the arterial lumen and tumor microenvironment, respectively), it is important to understand the CNS and peripheral stress physiology pathways (neuroendocrine and immune) underlying these pathological processes in order to develop tailored psychosocial and pharmacologic interventions targeting those at greatest risk.

Pharmacologic interventions to block SNS signaling and inflammation in cancer patients

Because distress-associated alterations in SNS and inflammatory processes have been implicated in cancer progression,Reference Lutgendorf, Sood and Antoni^19, Reference Antoni, Lutgendorf and Cole¹¹⁸ and since related hormones and cytokines are elevated during surgeryReference Ben-Eliyahu and Opp¹⁸³—a common curative intervention used in cancer patients—a novel line of work uses pharmacologic interventions to block SNS signaling (eg, non-selective beta adrenergic blockade) and inflammation (eg, COX2 inhibitors) in patients undergoing cancer surgery. A recently published double-blind, placebo-controlled trial showed that peri-surgical administration of a combined propranolol (beta adrenergic blocker) and etodolac (COX2 inhibitor) regimen downregulated leukocyte pro-inflammatory/pro-metastatic GATA and EGR family transcriptional factors and buffered increases in serum IL-6 and CRP, and prevented a decline in anti-viral immune responses to challenge (IL-12 and IFN-γ production) in BCa patients.Reference Shaashua, Shabat-Simon and Haldar¹⁸⁴ The pattern of down-regulated pro-inflammatory and pro-metastatic, and up-regulated anti-viral signaling is very similar to what has been observed in BCa patients receiving CBSM after surgery, which reflects a reversal of the CTRA.Reference Antoni, Lutgendorf and Blomberg^126, Reference Antoni, Bouchard and Jacobs^128, Reference Antoni, Lechner and Diaz¹³⁴

This work suggests that psychosocial and pharmacologic distress management interventions are capable of reducing distress-related biobehavioral processes (eg, neuroendocrine and inflammatory signaling), which may have significant beneficial effects on long-term clinical outcomes that are proportional to the biobehavioral changes induced during primary treatment and the skills that persist throughout cancer survivorship. Is there a potential for pharmacological agents to help regulate brain functioning by modulating SNS and inflammatory signaling, reversing the effects of stress? The future challenge in cancer research and neuroscience is to examine how these pharmacologic treatments, combined with behavioral (distress management) interventions, help to change brain structure and function in cancer patients promoting their well-being, quality of life, and better health outcomes.