Psychosomatics 43:259-281, August 2002
© 2002 The Academy of Psychosomatic Medicine
Stress Response Syndromes and Cancer: Conceptual and Assessment Issues
Maria Gurevich, Ph.D.,
Gerald M. Devins, Ph.D., C.Psych., and
Gary M. Rodin, M.D., FRCP (C)
Received June 15, 2001; revised December 17, 2001; accepted December 28, 2001. From Psychosocial Oncology, University Health Network; University Health Network Mental Health Program and Psychosocial Oncology; Culture, Community and Health Studies, Centre for Addiction and Mental Health and Psychosocial Oncology, Toronto, Canada. Address for correspondence and reprints: Dr. Maria Gurevich, Assistant Professor, Department of Psychology and Justice Studies, Ryerson University, 350 Victoria St., Toronto, Ontario, M5B 2K3; mgurevic{at}ryerson.ca (e-mail).
|
ABSTRACT
|
Stress response syndromes have begun to receive research attention in cancer, including melanoma, Hodgkin's disease, breast cancer, and mixed-diagnosis samples. This paper focuses on conceptual and assessment issues in the application of the trauma model to adaptation to cancer. Among the central considerations: differentiating cancer from other traumatic events, the utility of conceptualizing stress responses along a continuum in this population, diagnostic issues, and preliminary recommendations for therapeutic interventions.
Key Words: Stress • Cancer
|
INTRODUCTION
|
The personal tragedy of serious medical illness is not necessarily captured within the bounds of psychiatric illness. The recent acceptance of posttraumatic stress disorder (PTSD) as a psychiatric diagnostic category indicates a renewed recognition of the role of the environment in triggering psychopathology.1,2 This role is part of a complex interplay of biological, psychological, and social forces that contribute to persistent distressing responses to overwhelming stressors.3
Although stress response syndromes have been studied extensively after natural and human-caused disasters, war, and physical and sexual assault,1 their systematic study in association with cancer4,5 and other medical illnesses6,7 is more recent. Over the past two decades, it has been observed that cancer may trigger stress response syndromes in both adults8,9 and children.10–12 However, not until the DSM-IV13 was the threat to life and bodily integrity associated with illness, such as cancer, recognized as a traumatogenic stressor in diagnostic criteria for PTSD.
Two recent reviews on cancer and PTSD14,15 have examined the evidence for including cancer as a traumatogenic stressor, although both were based on a subset of the existing literature. Neel14 included three adult and four pediatric cancer studies, and Smith et al.15 reviewed five pediatric cancer studies and four on females with breast cancer. Neel's predominant focus was on the historical evolution of the PTSD diagnosis, whereas Smith et al. considered a variety of assessment, diagnostic, and risk factors. This paper extends these reviews by including 18 other studies that were reported in the literature prior to 1999 and 20 subsequent studies. We will address the unique characteristics of cancer as a traumatic stressor and the implications for diagnosis, treatment, and prevention. This review is restricted to the adult cancer literature, because the familial, developmental, and medical issues that distinguish pediatric cancer require separate consideration.
|
EPIDEMIOLOGY
|
Posttraumatic stress has now been studied in a range of cancers, including melanoma,16,17 Hodgkin's disease,18 breast cancer,19–26 and mixed-diagnosis samples.27–29 The incidence of full-syndrome PTSD ranged from 3% to 4% in patients recently diagnosed with early-stage cancer23 to 35% in patients after treatment.30 Posttraumatic symptoms were much more common, ranging in frequency from 20% in patients with early-stage cancer23 to 80% in those with a recent recurrence of cancer.9 Symptom incidence and severity was affected by many factors, including the type and stage of cancer, treatment severity, prior trauma, current and past social support, and the personality and coping style of the affected individuals.
|
CONCEPTUAL CONSIDERATIONS
|
Full Syndrome Versus Subthreshold Disorders
The usual psychological response to trauma involves oscillation between avoidance of the event's emotional impact and the intrusion, or re-experiencing, of the emotional response to it.31 The failure to resolve or assimilate traumatic experience32 is associated with persistent stress response symptoms, which are diagnosed as a stress disorder when they meet specified criteria.13 Trauma is defined in the DSM-IV as experiencing, witnessing, or learning about an event that involves either actual or threatened death or severe injury (criterion A1) when the individual responds with intense fear, helplessness, or horror (criterion A2). The DSM-IV provides a nonexhaustive list of typical PTSD stressors, which includes life-threatening illness. The DSM-IV cardinal symptom clusters are:
- Re-experiencing/intrusion (i.e., nightmares, flashbacks, intrusive thoughts, emotions, or images) (criterion B);
- Avoidance/numbing (i.e., detachment, restricted affect, and avoidance of reminder thoughts and activities) (criterion C); and
- Hyperarousal (i.e., hypervigilance, decreased concentration, increased startle response, insomnia, and irritability) (criterion D).
Symptoms that persist from 2 days to 1 month after a traumatic event and that cause significant social or occupational distress may meet DSM-IV criteria for an acute stress disorder (ASD). ASD often appears as a prodrome of PTSD,33 which can be diagnosed when these symptoms persist for at least 1 month. The symptom clusters of ASD correspond to those of PTSD, with the key exception of the dissociative symptom cluster, which is unique to ASD.34 However, ASD and PTSD may represent only the "tip of the distress iceberg."35 That is, subthreshold or partial PTSD36–38 is associated with significant social, interpersonal, and physical impairment38–42 and psychiatric comorbidity43 and is likely to be more common than full syndrome disorders in nonpsychiatric populations. Appropriate thresholds for clinically significant stress responses have not been established for populations of patients with cancer.
Continuum Versus Category
Some authors advocate a dichotomous approach to stress response syndromes that differentiates "cases" from "noncases."44 Evidence cited for considering PTSD as a distinct category includes its relatively low incidence after trauma, the associated premorbid and concurrent psychiatric comorbidity, and its biological specificity.44 However, the validity and utility of dichotomizing essentially continuous clinical phenomena has been questioned,45 and some view posttraumatic stress disorders as part of a continuum.1,46,47 In this paper, we consider the continuum of posttraumatic stress responses and restrict the term PTSD to cases in which DSM criteria have been met.
Differentiating Cancer From Other Traumatic Stressors
Cancer is a relatively common disease, but its consideration as a traumatic stressor stems not from its rarity but from its potential to overwhelm ordinary adaptive capacities.4,48 The risk, diagnosis, treatment, progression, and recurrence of a medical illness such as cancer may constitute profoundly stressful events.49,50 The threat to life and bodily integrity is often considerable,51 and the experience of disfiguration, disability, pain, and loss of social and occupational roles can trigger overwhelming feelings in a significant minority of affected individuals.52,53 The perceived lack of control imposed by the illness and the suddenness of the diagnosis may trigger intense fear, helplessness, or terror.5 Indeed, those with prior exposure to traumatic events often regard cancer as the worst trauma experienced.27
Although in many respects the trauma of cancer resembles that of other traumatic stressors, some distinctions are described below:
Chronicity of threat. The cyclical course of cancer parallels that of other traumatic exposures, such as may be found with war, family violence, and incest. Although a key feature of posttraumatic stress is the persistence of the stress response beyond the termination of the stressor,54 in the case of cancer and other protracted stressors, it may be difficult to distinguish symbolic repetition from a new or current threat. The cumulative response to such multiple, prolonged, and intermittent traumatic experiences may be qualitatively and quantitatively different from those that occur after more discrete, time-limited events,5,55 Triggers for such responses may include diagnostic and treatment procedures, witnessing the adverse course of other patients, psychoeducational interventions, or even routine follow-up. Recurrence of cancer may be even more traumatic than its initial presentation,9 and the entire cycle of stress responses may then be repeated. This "accumulated burden of adversity"56 may have a significant adverse effect on later psychological functioning.57
Uncertainty, intangibility, and anticipatory nature of threat. Cancer represents, in part, an "informational threat," whereby the mere knowledge that one has the disease may be regarded as the stressor.5 The intangible nature of this threat as well as the uncertainty regarding the outcome of the disease may amplify the stressful nature of the illness experience. For instance, waiting for the results of diagnostic testing has been cited as one of the most stressful aspects of the cancer experience.23 Furthermore, the trajectory of cancer is often unpredictable, with long periods of remission interrupted by acute episodes that require palliation,58 and the anticipated "death encounter" lingers throughout the course of the disease.5 In this sense, in contrast to traditional traumatic stressors, anticipatory threats may be just as salient as past or current threatening events in the case of cancer. In other words, the traumatic stressor may not be the memory of the initial diagnosis or the aversiveness of current or past treatments; rather, the chief preoccupation is likely to center on possible recurrence, progressive deterioration, and ensuing death.
Internality of threat. In contrast to external events such as sexual and physical assault, combat, and natural disasters, cancer originates from within the body.5 This internal origin may affect the perception and meaning of the threat, including its perceived inescapability. Furthermore, bodily signs or symptoms such as posttreatment scars, pain, or physical limitations imposed by the illness or treatment may also serve as persistent reminders of the disease. The internal and progressive nature of cancer may account for some of the overwhelming distress associated with it. However, comparative studies that have assessed the differential impact of internal versus external events have not been conducted.
|
ASSESSMENT CONSIDERATIONS
|
There is ongoing debate regarding the appropriate diagnostic criteria and assessment methods for stress response syndromes.59 Some of the conceptual and methodological issues that affect the assessment of stress responses in cancer are addressed below.
Cancer as a Criterion A Stressor
There is growing consensus among trauma stress researchers that the measurement of both index (i.e., precipitating) and nonindex (i.e., prior) traumatic events is critical to the evaluation of stress response syndromes.59,60 (There are a few reports of clinically significant PTSD symptoms in the absence of an identifiable traumatic event, leading to the suggestion that an additional stress disorder, prolonged duress, should be acknowledged.61) However, establishment of the most relevant index traumatic event(s) is difficult with multidimensional trauma, such as cancer, because numerous differentially threatening experiences may characterize each stage of the disease. Furthermore, comparability across studies is difficult, because there is substantial intra- and interstudy variability in the designation of the stressor event. Some studies have centered on diagnosis, others on pre- or posttreatment, and still others have examined recurrence. More important, these junctures are not similarly defined across studies. Posttreatment latency has been labeled as short as 6 weeks in some studies26 and as long as an average of 6.7 years in other studies.62 Some have collapsed the trauma of diagnosis and treatments19,21,63,64 to capture the full spectrum of cancer-related experiences. However, this approach obscures the differential threat associated with specific events in the disease and treatment process. Numerous events along the illness trajectory (e.g., cancer detection and confirmatory diagnosis, diagnostic and treatment procedures, and recurrence) can elicit a traumatic response. The only study that has assessed the range of cancer-related index traumatic events23 revealed that the three most stressful experiences, in order, were receiving the cancer diagnosis, waiting for the results of diagnostic testing, and waiting for surgery.
Measurement of Prior Traumatic Events
The term trauma is used variously to refer to a stimulus (e.g., traumatic event), a response (e.g., traumatic distress), and an intervening state (e.g., appraisal of a traumatic event), all three of which are included in the DSM criteria. However, a prior history of trauma should also be assessed because a robust positive relationship has been found among cumulative lifetime trauma, recent stressful events, and current posttraumatic stress responses, across a range of trauma populations.65–69 Stress-response symptoms related to earlier index events may be reactivated by subsequent exposure to traumatic events,70 and cumulative trauma is associated with a greater propensity for dissociation.71,72 Only five cancer and stress response studies have included a measure of prior trauma exposure,19,21,23,27,73 although such data would help in identifying individuals at highest risk.
Assessment Tools
Most studies of stress response syndromes in cancer have relied exclusively on self-report measures. The most frequently used measure has been the Impact of Events Scale (IES).74 This 15-item scale of the impact of stressful events is based on the frequency of distressing experiences, rather than the degree of experienced distress, which is more in keeping with the intended purpose of measuring the impact of traumatic events. The IES contains avoidance and intrusion subscales, and its most recent revision, the IES-R,75,76 which has not been used with cancer patients, also contains six new hyperarousal items and one item that assesses dissociative reexperiencing; it also assesses the degree of distress. These modifications bring its content and format closer to current diagnostic criteria, but it is still considered a non–DSM-correspondent measure.59
A less frequently used measure is the PTSD Checklist—Civilian Version (PCL-C),77 which consists of 17 items that correspond to specific DSM-IV PTSD symptoms. The PCL-C has recently been adapted for patients with breast cancer and includes an open-ended component that inquires about the nature of the symptoms.22 Andrykowski et al.20 compared the diagnostic value of the PCL-C with the PTSD module of the Structured Clinical Interview for DSM-IV (SCID),78–80 using 82 posttreatment patients with breast cancer. They concluded that the PCL-C is a cost-effective screening tool for PTSD, with diagnoses subsequently confirmed by a structured diagnostic interview. Evaluations of cancer-related PTSD symptom structure that have used the PCL-C63,80 have indicated that these stress symptoms reflect structurally concordant PTSD symptomatology. The PCL-C contains an item pertaining to a foreshortened future, which is a realistic concern for cancer patients. However, this may not capture the degree of preoccupation with an uncertain future, despite the fact that future concerns are often paramount for cancer patients, even during lengthy periods of remission.17,81 Uncertainty about recurrence is a primary concern for patients with both early-stage82 and late-stage breast cancer83 and is as an even greater concern than dying.82 Intrusive thoughts or images in this population may focus on future-oriented ruminations about possible recurrence, progressive physical and psychological decline, deterioration of social and occupational roles, the dying trajectory (i.e., the course leading toward death), and the prospect of death itself.5
Current stress response scales assess the arousal symptom of hypervigilance only in relation to external cues, whereas for cancer patients internal sensations (e.g., pain) and physical signs (e.g., new lumps) may be more threatening because of their potential association with metastases or recurrence. Indeed, there is evidence that patients with breast cancer exhibit an amplified preoccupation with physical sensations and bodily changes.5,64 This vigilance may be a realistic response to danger, a hypochondriacal or catastrophizing response, and/or a cancer-specific manifestation of posttraumatic arousal. Self-report measures and structured diagnostic interviews may need to be revised to take into account such symptoms of internal hypervigilance and to establish when such behavior is triggered by anxiety and when it has an anxiolytic effect.
Structured diagnostic interviews add to the reliability and validity of the assessment because they reduce response bias, use standardized diagnostic criteria, and permit the observation of behavioral manifestations of stress responses that are not revealed in self-report measures.59,84 However, structured diagnostic interviews have been used in only six studies of stress response syndromes in patients with cancer.19,23,27,30,85,86 Naidich and Motta86 used the Clinician Administered PTSD Scale,87,88 which is currently considered to be the best structured interview for diagnosing PTSD.59 The remaining studies relied on the SCID PTSD module. Only Andrykowski et al.19 reported comparison scores for the self-report measures versus the structured interview. In that study, the SCID and the PCL-C identified the same rate of current PTSD (6%) in patients with breast cancer. However, subsyndromal levels were identified more frequently by the PCL-C (13%) compared with the SCID (4%). This difference may reflect the more stringent criteria of the SCID or the willingness of participants to report symptoms in the more anonymous context of the self-report measure.
Confounding Symptoms
Stress responses may be confounded with illness- and treatment-related symptoms of cancer.5 For example, sleeplessness, irritability, and impaired memory and concentration may reflect the disease process of cancer or the pattern of distress that makes up a stress response syndrome or both. Also, medications (e.g., some sedative-hypnotics) can act as behavioral disinhibitors, amplifying or mimicking stress-response symptoms.89 These confounding symptoms may lead to overdiagnosis or misdiagnosis of stress response syndromes in cancer, particularly because the sensitivity and specificity of the self-report cutoff levels to screen for PTSD have not been determined in cancer and other medical populations. This difficulty can be addressed by clinical assessment, the inclusion of structured diagnostic interviews in research protocols, and the establishment of the sensitivity and specificity of self-report measures.
Stress Response Syndromes in Cancer: Risk and Protective Factors
The trauma literature suggests that sociodemographic variables,44 characteristics of the stressor,90 prior history of traumatic stress,91 and social factors92,93 act as significant determinants of the nature and severity of stress response syndromes. However, the relative contribution of each of these factors to stress response symptom development in cancer has received little empirical attention. Although most studies have treated stress responses as an outcome variable, several studies treat them as an independent variable. The design, sample characteristics, and major findings for each of the studies reviewed below are summarized in Table 1 .
Sociodemographic characteristics. Although a twofold preponderance of PTSD in women has been reported in the community,94,95 the findings in cancer have been mixed. Four studies have found female sex to be correlated with greater stress response symptomatology,16,29,96,97 whereas four others did not reveal such differences.9,17,18,28 However, inferences from these data are limited because of the predominant focus on female patients both within and across studies.
Younger patients typically exhibit higher rates of22–24,26,97–100 and more persistent stress response symptoms than older patients over a 1-year period.21 However, several studies have found no relationship between age and symptom severity.18,19,25,86,101 A few studies have also found stress response symptoms to be associated with lower income18,22,23 and less education.18,22,98 The association with lower socioeconomic class may be related to the higher rates of prior trauma exposure in that group, which is a well-established risk factor for stress response disorders.102,103
Disease and treatment variables. In general, greater proximity to diagnosis and treatment, greater treatment intensity, and recurrence have been associated with a higher frequency or severity of stress response symptoms. The impact of disease stage has not been sufficiently studied, although a few studies have found this variable to be related to stress responses. In a study of 82 women after breast cancer treatment, Andrykowski and Cordova19 found that higher stress response frequency was associated with more advanced disease. Among 43 disease-free women who underwent autologous bone marrow transplantation for stage II–IV breast cancer, the severity of stress response symptoms was associated with more advanced disease at transplantation and longer hospitalization after transplantation.104 Most studies have not found an association between disease stage and stress response symptoms; however, these typically have a restricted range in disease stage and/or focus on disease-free early-stage patients, several years posttreatment. For instance, Alter et al.27 did not find an effect of disease stage among 27 disease-free female cancer survivors, who were on average 4.6 years posttreatment and 5.4 years postdiagnosis. Tjemsland et al.26 did not find a difference in stress responses between breast cancer patients with and without axillary-node metastasis, which may have been due to the relative homogeneity of disease stage in this sample (i.e., stage I and II only). Similarly, Baider et al.16 did not find an effect of disease stage in a group of early localized malignant melanoma patients, 73% of whom had stage I disease and 27% stage II disease. Disease stage was also not a significant predictor of intrusion and avoidance in a mixed sample of cancer patients undergoing palliative radiotherapy for pain management due to metastases, 63% of whom were inpatients.29
In a self-report study of patients with breast cancer, stress response symptoms were more common soon after treatment.22 Open-ended questions revealed that the most common concerns of those with intrusive symptoms were treatment side-effects and fears of recurrence. Similarly, Andrykowski and Cordova19 found that higher stress response frequency was associated with less time since treatment completion in patients with breast cancer. Kornblith et al.18 also found that greater time since treatment completion predicted lower self-reported stress response symptoms in a group of patients with advanced Hodgkin's disease. However, several other studies found no relationship between time since treatment completion and stress responses.21,23,27,30
The frequency and severity of symptoms may also be affected by the extent to which there is sufficient time for adaptation. A cross-sectional study of 95 patients with malignant melanoma17 across four disease stages (I–IV) found that the highest IES scores were exhibited by patients with stage III disease. This stage, which is associated with more extensive treatments and may signal the transition to a less favorable prognosis, can be the most traumatic juncture in the illness course.17 Although stage IV reflects greater disease severity, there may be some accommodation to the illness by this time. This may explain the negative correlation of stress symptoms with time elapsed since diagnosis in this study and the high frequency of stress responses after recurrence9,72 or metastases.29
The meaning of the treatment may also be as important as its intensity in producing stress responses. Although some studies have shown higher rates of stress responses in patients with cancer who have undergone bone marrow transplantation (BMT)101 compared with other forms of treatment,22,23,27 no differences were found between current levels of PTSD between 17 women who underwent autologous BMT for breast cancer compared with 20 women who had received conventional treatments.30 The women in the latter study viewed BMT as life-affirming, despite its invasiveness and the high risk of treatment failure.105,106 Notably, when retrospective accounts of PTSD in relation to the entire spectrum of cancer experiences were elicited, 41% of the BMT group (76% stage III and IV at diagnosis) and 30% of the conventional treatment group (15% stage III and IV at diagnosis) reported PTSD symptoms, although this difference was not statistically significant. Tjemsland et al.26 found that women undergoing adjuvant radiation treatment, an active form of treatment, reported lower levels of intrusive symptomatology than those undergoing less active treatment. Although the authors attribute this finding to the possible social affiliation facilitated by such treatment, it is also possible that the reassurance provided by engaging in ongoing treatment may be a more (or equally) significant factor. However, treatment type was not found to be related to stress response symptoms in several other studies.21,23,27,72,101,107,108
Factors that increase stressor severity in other populations include length of the event, perceived life threat, degree of actual physical harm, and the extent to which the trauma constitutes an interpersonal event.55,91,109 Although cancer type and stage and proximity to treatment have been used as proxy indicators, these objective disease variables may not reflect the subjective experience of life threat or the interpersonal meaning of cancer. This may account for the lack of association between treatment intensity and stress responses in some studies.26,30 Notably, Hampton and Frombach96 found that higher stress response scores were related to high perceived life threat, low illness meaning, and high perceived treatment intensity in female cancer patients. Global meaning (i.e., one's existential belief system) has also been found to moderate the relationship between intrusive thoughts and psychological distress in both breast cancer survivors 2–15 years postdiagnosis99 and in leukemia patients who had undergone BMT.110
Prior traumatic events. Prior trauma appears to diminish stress tolerance, leading to a rapid escalation of symptoms in response to subsequent traumatic events.111 Although a few studies have not found a relationship between trauma history and stress responses,27,107,130 greater prior exposure to traumatic events is typically associated with clinically significant stress response symptoms in both early-stage19 and metastatic breast cancer.73 Green et al.23 did not initially assess the impact of trauma history, although these data were collected. In a later report,108 using the same sample of 160 patients with early-stage breast cancer, they found that both prior traumatic event exposure and recent life events were significantly related to PTSD symptoms. Similarly, Baider and Kaplan De-Nour112 reported that past traumatization (being a Holocaust survivor) and concurrent additional stressors (recent immigration) were both positively related to symptoms of intrusion and avoidance in a sample of 283 patients with early-stage breast cancer who participated in follow-up treatment. In a later study, second-generation Holocaust survivors, 8 months to 8 years following breast cancer diagnosis, reported higher stress response symptoms than women whose parents had not been through the Holocaust.113 The priming effect of earlier trauma is supported by other research that shows that Holocaust survivors with cancer exhibit more distress than other cancer patients.114–117 Similarly, distress after cancer is predicted by a history of multiple life stressors predating the illness.118,119 These findings are compatible with the view that prior trauma may serve as an "organizer" of preexisting and subsequent experiences,120 which then filters the impact of a current traumatic event.68 The association with prior trauma may also account for the comorbidity of PTSD with other psychological sequelae, such as affective disorders and substance abuse,121,122 with anxiety and depression being the most common.123
Temporal course of stress responses. Stress responses may be immediate, delayed, or persistent after trigger events.124–127 However, only five published studies have explored the temporal course of stress response symptoms after the diagnosis of cancer.21,26,98,107,128 These studies indicate that symptom resolution does not always occur with the passage of time. Bleiker et al.107 assessed stress responses in 170 patients with early-stage breast cancer 2 months and then again on average 21 months after surgery. Intrusive thought levels remained relatively stable across the two time periods, with 16% of women reporting high levels of distress at both times and 31% reporting moderate levels. High levels of avoidance were reduced, from 15% at time 1 to 8% at time 2; likewise moderate avoidance levels were reduced from 31% at time 1 to 25% at time 2. The best predictors of distress at time 2 were the following factors at time 1: intrusive thoughts, health problems, sleep problems, and trait anxiety. Similarly, in a 6-month longitudinal study of 80 women with newly diagnosed, early-stage breast cancer, intrusive thoughts and avoidance were moderately high at the time of diagnosis, but only the former decreased over the subsequent 6 months.98 A sample of 129 patients with mixed-diagnosis advanced-stage cancer exhibited similar mean intrusion scores from the first assessment to the 3-month follow-up.128
In a 12-month prospective study of 46 patients with early-stage breast cancer (on average nearly 30 months posttreatment), overall PTSD levels remained relatively stable from initial assessment to a 1-year follow-up.21 However, older women showed a decrease in stress response symptoms between the two assessment points, and reductions in symptom severity scores in 25% of the sample were associated with comparatively fewer precancer traumatic stressors and greater social support. Tjemsland et al.26 assessed stress responses in 106 patients with early-stage breast cancer 1 day before surgery and again at 6 weeks and 1 year postsurgery. Medium to high intrusion (81%) and avoidance (75%) scores were common at time 1. Symptoms were significantly reduced by 6 weeks postsurgery, and a further diminishment occurred 1 year postsurgery. The most important risk factors for stress responses 1 year after surgery were premorbid health problems, high emotional reactivity, and high intrusion levels at 6 weeks. Additional variance was accounted for by younger age, poor presurgery psychosocial functioning, and presurgery negative life events.
Social responses and resources. Support from others attenuates the response to traumatic events and facilitates affect regulation,93 and the relative absence of prior or current support is associated with pathological stress response symptoms.129 Greater perceived social support is associated with fewer stress response symptoms in early-stage breast cancer19,108 and in patients who had received BMT,130 although at least one study107 did not find this association. As indicated above, symptom resolution over a 1-year period is also associated with fewer precancer traumatic stressors and greater social support in patients with early-stage breast cancer.21 Green et al.108 found that social support was among the main predictors of stress responses in 160 patients with early-stage breast cancer. In a study of 64 breast cancer survivors, on average 6.7 years posttreatment, social support moderated the relationship between intrusive thoughts and quality of life.62 Notably, aversive social interactions have a deleterious effect on stress response symptom development. As such, they may be more appropriately conceptualized as additional current stressors.73 Butler et al.73 found that aversive social support (e.g., excessive demands or criticism) interacted with past life stressors to increase intrusion and avoidance in a sample of 125 patients with metastatic breast cancer. Similarly, in a sample of 129 patients with advanced mixed-diagnosis cancer, Manne128 found that spouse criticism moderated the association between intrusive thoughts at initial assessment and psychological distress 3 months later. In a study of second-generation Holocaust survivors with breast cancer,113 married women reported higher intrusion levels compared with a comparison group of married women with breast cancer whose parents were not in the Holocaust. Although no information was collected on the nature of the marital relationship for either sample, it is notable that Holocaust survivors may perceive marriage as an additional stressor rather than as a source of mutual support.
The availability of health-related information is also a common concern for cancer patients9,131 and may reflect the responsiveness of the medical or broader social environment. The lack of appropriate information may amplify stress response symptoms in patients with cancer. In recently diagnosed patients with cancer,97 women with ASD reported significantly less satisfaction with the communication of their cancer diagnosis. Similarly, women who were unaware of their cancer stage reported higher levels of stress response symptoms compared with women with diagnostic knowledge.86 The adequacy of information may also reflect the quality of the relationship with medical caregivers. In that regard, Hampton and Frombach96 found that difficulty with healthcare professionals was among the main predictors of stress response symptoms in women with cancer. These results are consistent with evidence that emotional and instrumental support from family, friends, and medical caregivers may influence physical and mental health outcomes among patients with cancer.132–134
Coping and personality styles. Premorbid personality characteristics may be important determinants of stress responses. Avoidant coping styles have been associated with increased stress response symptoms in motor vehicle accident victims,135 war veterans,136 and patients with burn injury.137 However, it may be difficult to distinguish responses to a current trigger event, such as illness, from premorbid coping style. In patients with cancer, close monitoring of bodily changes138 may be a coping strategy or may reflect reexperiencing and/or hyperarousal.64 Both anxious preoccupation and avoidant coping were shown to correlate with stress response symptoms in breast cancer,64 BMT,130 and a mixed-diagnosis sample of cancer patients.176 In a sample of patients with mixed-diagnosis cancer, cognitive avoidance coping was related to PTSD for women, whereas behavioral avoidance was the best predictor of PTSD in men.96 Coping was also unrelated to stress responses in at least two studies.16,98 In this regard, Baider et al.113 found that anxious preoccupation was unrelated to intrusive thoughts in second-generation Holocaust survivors, whereas a positive relationship was found in a comparison group of patients with breast cancer who had no Holocaust history. This is consistent with the finding that the functional adjustment of first-generation Holocaust survivors is comparable to that of other cancer patients, but their emotional distress levels are considerably higher.115 Furthermore, a recent study26 found that the severity of stress symptoms 1 year after surgery was significantly associated with high emotional reactivity as measured by the Eysenck Personality Inventory.139 However, antecedent characteristics are difficult to distinguish from outcomes, because individual coping styles and personality characteristics become inevitably intertwined in the response to illness. Although the coping measures used in these studies were illness-specific (e.g., Mental Attitude to Cancer Scale), illness-related coping styles will undoubtedly reflect preexisting coping orientations to some degree and are, therefore, not easily disentangled in the absence of baseline precancer data.
Empirical validity of trauma model for cancer. Preliminary evidence for the construct validity of cancer-related stress responses is also provided by the finding that women with cancer have reported greater interference effects in response to cancer-related information on the emotional Stroop compared with controls without cancer.86 Psychophysiologic support has also been provided by a recent study that showed that, compared with patients with breast cancer who did not have PTSD, patients with current cancer-related PTSD exhibited higher heart rate, skin conductance, and corrugator electromyogram responses when presented with imagery of their own cancer experiences.140 In addition, high rates of stress response syndromes are reported by patients with cancer who exhibit low levels of depression and/or anxiety,30,86 which attests to discriminant validity. Incremental validity is provided by the finding that stress response symptoms are more directly related to disease stage than to anxiety and depression.17 The predictive validity of cancer-related stress responses is supported by the following: 1) intrusion is a better predictor of distress at 3 months after diagnosis than anxiety and depression;98 2) intrusion 6 weeks to 2 months postsurgery is a better predictor of stress responses 1 year to 21 months postsurgery than global distress;26,107 and 3) avoidance near the time of diagnosis is a better predictor of disease status 1 year after diagnosis than global distress.28 With respect to the latter, the possible role of avoidance in decreased compliance with cancer treatments and surveillance necessitates much greater attention. In this regard, avoidance was found to mediate the relation between intrusive thoughts and later psychological distress over a 6-month period in patients with advanced-stage but not early-stage cancer.141 For the latter group, intrusive thoughts were directly related to increased distress. It is possible that avoidance may be more maladaptive in the case of greater cancer threat, with a more unpredictable outcome and a longer time frame, as is the case with advanced cancer.
In summary, these findings provide provisional empirical support for the utility of the trauma model in cancer. Most known risk and protective factors in other trauma contexts are relevant to cancer, although the findings regarding sex and age are equivocal in this population. Stress responses in cancer are also associated with communication with health care providers, diagnostic knowledge, and treatment intensity and proximity. The data on cancer stage are mixed and insufficient. Preliminary evidence has also been provided for construct, discriminant, and predictive validity of cancer-related stress response syndromes.
|
TREATMENT IMPLICATIONS
|
Identifying stress responses and risk and protective factors in those diagnosed with cancer has practical significance, because early psychosocial and pharmacological interventions are effective in other populations,142 and early stress responses may identify patients with cancer who are at risk for later distress.21,26,98 A variety of interventions, including cognitive behavior and exposure therapies, psychodynamic therapies, pharmacotherapy, and eye movement desensitization and reprocessing (EMDR), have been used in the treatment of PTSD.143–149 Systematic treatment-outcome studies are limited.3 Cognitive-behavioral, psychopharmacological, and EMDR approaches have been the central focus of published treatment studies.147 Few studies have examined the efficacy of psychodynamic therapies, although these constitute the central treatment modalities used by clinicians with traumatized patients.150 Currently, a multidimensional approach is frequently needed, given the multidimensional nature of PTSD, its length and severity, and the reticence of many PTSD sufferers to seek treatment.151
There is little systematic evidence for the benefit of interventions to treat stress response syndromes in cancer, although there is a substantial literature on the prevention and treatment of psychological distress in individuals with cancer.132,133,152,153 The most commonly used interventions fall broadly under the following overlapping categories: educational techniques,133,153 behavioral training/coping skills training,154,155 emotional and social support,156–160 and interpersonal psychotherapy.161–163 Individual and group approaches have been fairly evenly applied to all of these interventions. All of these treatments have been shown to be effective in alleviating distress.
A single case study of a trauma-focused intervention in cancer that used a male patient with PTSD, nearly 3 years after BMT, found a reduction of symptoms after a 10-session cognitive-behavioral intervention.85 A combination of education, relaxation training, cognitive coping strategies, and relapse prevention and generalizability techniques were used, and the treatment effect was maintained at the 6-month follow-up. A nonrandomized study that used supportive group psychotherapy, coping, and relaxation strategies for 60 women with primary breast cancer also showed a reduction in intrusive symptoms.101 Similarly, a 6-week intervention with 86 cancer patients relied on progressive muscle relaxation and guided imagery to reduce stress response symptoms.170 Cancer-related stress response syndromes were also reduced in a recent randomized trial with 125 patients with metastatic cancer.164 Women who received 1 year of weekly supportive-expressive group therapy along with educational materials exhibited significantly greater reductions in stress responses and mood disturbance compared with women in the control group, who received educational materials only. In another randomized trial with 312 patients with early-stage breast cancer, women who participated in an educational group or in a combined peer discussion and education group showed a positive effect on overall adjustment compared with women in the no-intervention and peer-support-only groups.165 Unfortunately, the IES was only administered after the intervention, so no inferences can be made regarding treatment-related stress response changes. In contrast, a randomized study that used emotional expression writing sessions in 44 women with early-stage breast cancer did not find a diminishment in intrusive thoughts, avoidance, or overall psychosocial adjustment.166 This is consistent with the larger literature on cancer-related distress, which shows that most robust and enduring efficacy is demonstrated when several treatment modalities are combined.167–169
The oncology clinic setting provides a powerful opportunity for instituting early interventions to prevent or diminish maladaptive stress response syndromes. Specifically, the healthcare staff can be poised to intervene with high-risk patients before or soon after many illness- and treatment-related traumatic events have occurred. Although mental health professionals may have more specific training in psychotherapeutic techniques, patients with cancer frequently prefer obtaining such support from their oncologists.171 However, the opportunity to provide support is lost in many medical settings. Dissatisfaction with medical caregivers' communication of illness-related information is common among patients with cancer,172–174 and stress response symptoms in women are associated with interactional difficulties with healthcare providers.96
It may not be feasible, or even desirable, for oncologists to spend more time with patients in settings with high case volumes and large clinical demands. However, attention to the way in which brief medical interactions are constructed and the regularization of a team approach to patient support may help create a more supportive environment for patients. This approach should parallel the cancer trajectory from diagnosis to treatment to surveillance. A multidisciplinary health care delivery model with a focus on information provision and psychosocial support has been shown to increase psychosocial adjustment, physical functioning, and satisfaction with health care providers in patients with breast cancer.134 Likewise, organizational changes that lead to greater integration of health-service provision within an oncology setting have been shown to increase patients' sense of control over health care and decrease distress.175 Systematic research is needed to determine to what extent such approaches diminish distress and the occurrence of problematic stress response syndromes in patients with cancer.
|
SUMMARY AND CONCLUSIONS
|
Although the PTSD diagnosis captures many of the symptoms associated with the response to severe trauma, this diagnosis may not encompass the multidimensionality or continuum of persistent responses to overwhelming experience.3 This continuum may be particularly relevant for patients with cancer and for other nonpsychiatric populations in whom full-syndrome PTSD is relatively uncommon. The assessment of stress response symptoms in cancer may also be complicated by the multiplicity, internality, and indeterminate nature of the stressor event(s) and by the confounding effect of symptoms related to the illness or treatment.
The reported incidence and correlates of cancer-related stress response syndromes have also been affected by a variety of factors, including variability in the nature and quality of measurement tools, stressor identification, and study designs. However, there is growing empirical evidence that significant stress response symptoms are common in cancer. Most of the risk and protective factors that have been identified in other trauma populations are applicable to patients with cancer. Although communication with health care providers, diagnostic knowledge, and treatment intensity and proximity are specific correlates in this population, these are aspects of social support in the case of the first two factors and trauma severity in the case of the latter two factors. With regard to trauma severity, factors that affect the meaning of the illness and perceived life threat as potential mediators of distress require further empirical investigation.
Future research should also place greater emphasis on the evaluation of prior traumatic events and the range of cancer-related traumatic events. Structured diagnostic interviews in conjunction with multiple self-report screening tools with a greater symptom range may help clarify some of the phenomenological and diagnostic issues. Longitudinal prospective studies may also help determine the extent to which specific symptoms vary over time and how they are affected by the illness trajectory. Most important, the prognostic validity of cancer-related stress response syndromes in terms of physical and mental health outcomes requires further empirical attention. In that regard, the cancer setting offers considerable opportunity to devise preventive and treatment measures that may help alleviate maladaptive stress response symptoms and improve health, well-being, and adjustment to illness.
|
ACKNOWLEDGMENTS
|
The preparation of this manuscript was facilitated by the Edith Kirchmann Research Fellowship awarded to the first author and administered by the Psychosocial Oncology Program at Princess Margaret Hospital (University Health Network).
|
REFERENCES
|
- Briere J: Psychological Assessment of Adult Posttraumatic States. Washington, DC, American Psychological Association, 1997
- Wilson JP, Raphael B: Theoretical and conceptual foundations of traumatic stress syndromes, in International Handbook of Traumatic Stress Syndromes. Edited by Wilson JP, Raphael B. New York, Plenum Press, 1993, pp 1-10
- van der Kolk BA, McFarlane AC: The black hole of trauma, in Traumatic Stress: The Effects of Overwhelming Experience on Mind, Body and Society. Edited by van der Kolk BA, McFarlane AD, Weisaeth L. New York, Guilford Press, 1996, pp 3-23
- Baum A, Posluszny DM: Traumatic stress as a target for intervention with cancer patients, in Psychosocial Interventions for Cancer. Edited by Baum A, Andersen BL. Washington, DC, American Psychological Association, 2001, pp 143-173
- Green BL, Epstein SA, Krupnick JL, Rowland JH: Trauma and medical illness: assessing trauma-related disorders, in Assessing Psychological Trauma and PTSD. Edited by Wilson JP, Keane TM. New York, Guilford Press, 1997, pp 160-191
- Kelly B, Raphael B, Judd F, et al: Posttraumatic stress disorder in response to HIV infection. Gen Hosp Psychiatry 1998; 20:345-352[CrossRef][Medline]
- van Driel RC, Op den Velde W: Myocardial infarction and post-traumatic stress disorder. J Trauma Stress 1995; 8:151-159[CrossRef][Medline]
- Cella DF, Tross S: Psychological adjustment to survival from Hodgkin's disease. J Consult Clin Psychol 1986; 54:616-622[CrossRef][Medline]
|
TOP
ABSTRACT
INTRODUCTION