|NCI/PDQ® Health professionals: Post-traumatic Stress Disorder (PDQ®)|
|National Cancer Institute|
| Last Modified: June 15, 2012
For a number of years, investigators have reported stress or trauma-related symptoms such as avoidant behaviors, intrusive thoughts, and heightened arousal in survivors of cancer. 1 2 3 4 These symptoms resemble those seen in persons who have experienced traumatic events such as military combat, violent personal assault (e.g., rape), natural disasters, or other threats to life and are referred to collectively as post-traumatic stress disorder (PTSD). 5 6 7 8 9 10 Acute stress disorder (ASD) is a Diagnostic and Statistical Manual of Mental Disorders (DSM) mental disorder with a profile similar to that of PTSD but a shorter time to onset, within 4 weeks of a traumatic event. Thus the occurrence of PTSD and trauma-related symptoms in patients with cancer has been under increasing study, influenced by changes in the diagnostic criteria for PTSD in the DSM, fourth edition (DSM-IV). 5 The DSM, third revised edition (DSM-III-R), 11 specifically excluded patients with medical illnesses such as cancer from PTSD. The diagnostic criteria for PTSD in the DSM-IV, text revision (DSM-IV-TR), however, specifically include being diagnosed with a life-threatening illness as one example of a traumatic event. 12 Thus, people with histories of cancer can now be evaluated and considered at risk for PTSD.
Reviews note that post-traumatic stress has been studied in a variety of cancers including melanoma, Hodgkin lymphoma, breast cancer, and mixed cancers. Studies have varied, however, in whether they assessed patients for the full syndrome of PTSD (i.e., all DSM-IV criteria met) or only some of the PTSD-related symptoms (e.g., intrusive thoughts as measured by the Impact of Event Scale [IES]). Thus, incidence rates have varied accordingly. The incidence of the full syndrome of PTSD (meeting full DSM-IV diagnostic criteria) ranges from 3% to 4% in early-stage patients recently diagnosed to 35% in patients evaluated after treatment. When incidence of PTSD-like symptoms (not meeting the full diagnostic criteria) is measured, the rates are higher, ranging from 20% in patients with early-stage cancer to 80% in those with recurrent cancer.
In one German study, patients with breast cancer (n = 127) were evaluated for PTSD immediately postsurgery and 6 months after the first assessment. 13 The assessments included screening instruments for ASD and PTSD, such as the IES-Revised (IES-R) and the PTSD Checklist-Civilian (PCL-C). First assessment also included a semistructured interview with the Structured Clinical Interview for DSM (SCID). On the basis of the SCID, 2.4% of participants met the criteria for mild to moderate cancer-related PTSD, and 2.4% were diagnosed with ASD. However, the screening instruments IES-R and PCL-C identified PTSD in 18.5% of participants at the first assessment and in 11.2% to 16.3% of participants at the second assessment. The study authors seem to suggest that unlike the SCID, the screening instruments IES-R and PCL-C measure diffuse emotional distress and adjustment problems and not precise PTSD symptoms. One of the main differences between symptom-based measures such as the PCL-C and an actual SCID-based diagnosis is the dysfunction caused by the symptoms. The symptoms are rather common, but only a very small percentage of people who have the symptoms are disabled by them.
Factors suggesting which patients might be at increased risk for the development of PTSD have not been extensively studied; however, one study of women with early-stage breast cancer 14 found that younger age, lower income, and fewer years of formal education were associated with PTSD-like symptoms. Another study of men and women treated with bone marrow transplant 15 found that lower levels of social support and the use of avoidance coping correlated significantly to a higher number of PTSD-like symptoms. One German study mentioned earlier 13 that evaluated patients with breast cancer for PTSD and ASD concluded that patients with lifetime PTSD (8.7%) were much more likely to experience cancer-related ASD or PTSD (odds ratio 14.1).
Although no specific therapies for PTSD in the cancer setting have been developed, treatment modalities used with other people with PTSD can be useful in alleviating distress in cancer patients and survivors.
In this summary, unless otherwise stated, evidence and practice issues as they relate to adults are discussed. The evidence and application to practice related to children may differ significantly from information related to adults. When specific information about the care of children is available, it is summarized under its own heading.
Reviews of the literature 1 note that post-traumatic stress has been studied in a variety of cancers, including melanoma, Hodgkin lymphoma, breast cancer, and mixed cancers. The incidence of the full syndrome of post-traumatic stress disorder (PTSD) (meeting the full Diagnostic and Statistical Manual of Mental Disorders, fourth edition [DSM-IV], diagnostic criteria) ranges from 3% to 4% in early-stage patients recently diagnosed to 35% in patients evaluated after treatment. When incidence of PTSD-like symptoms (not meeting the full diagnostic criteria) are measured, the rates are higher, ranging from 20% in patients with early-stage cancer to 80% in those with recurrent cancer.
The earliest research (predating DSM-IV) on PTSD among survivors of cancer concentrated on the prevalence and characteristics of the disorder in patients who had been or were undergoing treatment, adult and child survivors of cancer, and/or the family members of these patients and survivors. A wide variety of cancer types were studied, including leukemia, 2 breast cancer, and head and neck cancers. 3 Much of the earlier research dealt with survivors of Hodgkin disease, probably because diagnoses at an early age and higher rates of survival resulted in a larger population available for study. 4 These survivors were found to have a particularly high prevalence of intrusive thoughts and avoidance behaviors, even though they were many years posttreatment. 5 6 7 Most of these studies investigated PTSD-like symptoms, rather than the complete mental disorder with all diagnostic criteria.
The first study of cancer patients utilizing the current DSM-IV diagnostic criteria looked at 27 patients (mostly breast cancer), all at least 3 years postdiagnosis and no longer receiving any cancer treatments. In this study, a prevalence rate of 4% for current PTSD and 22% lifetime prevalence was found. 8 Those who met criteria for lifetime prevalence were noted to have higher levels of general psychologic distress, suggesting that individuals with a history of PTSD are at a substantial risk for continued emotional difficulties.
Studies using the Structured Clinical Interview for DSM (SCID) 9 have found prevalence rates of PTSD between 3% and 10% in adult cancer patients. Most of these studies looked at women with early-stage breast cancer, evaluated a few months to a few years after their cancer treatments. Similarly, in a prospective study of 115 patients with all stages of breast cancer being treated in a comprehensive cancer center, 4% met the full diagnostic criteria for PTSD; 41% met the subsyndromal criteria for PTSD (experiencing intense fear, helplessness, or horror after being diagnosed with cancer). This set of subsyndromal criteria was a weak predictor of PTSD (12%) but an equally useful predictor of major depressive disorder, global anxiety disorder, and past major depressive disorder, and it may better serve as a marker for elevated distress. 10 In a few studies of patients with bone marrow transplants, slightly higher prevalence rates have been reported, ranging from 5% 11 to 12% to 19% 12 to as high as 35%. 13 The range in prevalence appears to be influenced by time of assessment (higher rates occurring with more time since transplant) and the assessment method used. Studies reporting lower rates typically used a self-report questionnaire, 14 whereas those reporting higher rates 13 used the SCID and evaluated for symptoms at multiple times since diagnosis (i.e., lifetime prevalence).
As an illustration of the distinction between these tools, a German study evaluated patients with breast cancer (n = 127) for PTSD immediately postsurgery and 6 months after the first assessment. 15 The assessments included screening instruments for acute stress disorder (ASD) and PTSD, such as the Impact of Event Scale-Revised (IES-R) and the PTSD Checklist-Civilian (PCL-C). The first assessment also included a semistructured interview using the SCID. On the basis of the SCID, 2.4% of participants met the criteria for mild-to-moderate cancer-related PTSD, and 2.4% were diagnosed with ASD. However, the screening instruments IES-R and PCL-C identified PTSD in 18.5% of participants at the first assessment and in 11.2% to 16.3% of participants at the second assessment. Authors of the study seem to suggest that unlike SCID, the screening instruments IES-R and PCL-C measure diffuse emotional distress and adjustment problems and not precise PTSD symptoms. One of the main differences between symptom-based measures such as PCL-C and an actual SCID-based diagnosis is the dysfunction caused by the symptoms. The symptoms are rather common, but only a very small percentage of people who have the symptoms are disabled by them.
Post-traumatic stress disorder (PTSD) was initially characterized as an anxiety disorder that developed in response to a severe trauma in which an individual experienced, witnessed, or was confronted by actual or threatened death, injury, or loss of physical integrity of self or others. The Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV), stipulated for the first time that being diagnosed with a life-threatening illness or learning that one's child had such an illness qualifies as a stressful event. 1
In 1994, the application of PTSD to patients with cancer began with the redefinition of the trauma criteria in the DSM-IV to include life-threatening illness. 1 The essential feature of this disorder is the development of characteristic symptoms following exposure to an extreme traumatic stressor. 2 These events elicit responses of intense fear, helplessness, or horror and trigger three clusters of PTSD symptoms. Symptoms from each of the following three clusters must be present for an individual to meet the full criteria for a diagnosis of PTSD:
These symptoms must last for at least 1 month and cause clinically significant distress or impairment in social, occupational, or other important areas of functioning.
Symptoms that last for at least 1 day but less than 1 month and that cause significant distress or impairment in social, occupational, or other important areas of functioning might meet the diagnostic criteria for Acute Stress Disorder (ASD). ASD is often a prodrome to PTSD.
Conceptual and practical problems can arise in the application of PTSD to cancer patients and survivors. The basic concept of an extreme traumatic stressor has been described variously as an event involving direct personal experience that involves actual or threatened death or serious injury. 2 This event can be protracted and continuous but is more frequently a single, time-limited event (e.g., rape, natural disaster). In this context, for the person who has experienced a diagnosis of cancer, the exact nature of the trauma is unclear. Is it the actual diagnosis, aspects of the treatment process, information given about recurrence, negative test results, or some other aspect of the cancer experience? Identifying a discrete stressor within the multiple crises that constitute a cancer experience is much more difficult than it is for other traumas. In one study of breast cancer patients 3 who underwent autologous bone marrow transplant, more PTSD-like symptoms were reported at the time of initial diagnosis.
Another concern regarding conceptual fit is related to reexperiencing the trauma. Diagnostic criteria B require persistent reexperiencing of the traumatic event, implying that the patient would first encounter a trauma and then, at a later time, reexperience it in various ways. In a study of women with early-stage breast cancer, however, researchers 4 found that the traumatizing aspects of the cancer experience were receiving the diagnosis and waiting for test results from node dissection. Arguing that these information traumas are future oriented and tend to cause intrusive worry about the futurenot intrusive recollections of past eventsthe authors questioned whether cancer fits a conceptual model of PTSD trauma. Reexperiencing the trauma is often measured in terms of unwanted intrusive thoughts of the traumatic event. The cognitive processing of a current and ongoing health threat with uncertain outcome might differ significantly from unwanted intrusive thoughts about a single past event. Some researchers have argued that not all intrusive thoughts are negative or indicate reexperiencing a trauma, but rather that they might represent appropriate vigilance and attention to potential symptoms that could result in appropriate help-seeking. 5 6
Conversely, a unique study assessing the physiological reactivity of breast cancer patients to a personalized imagery script of their most stressful experiences with breast cancer found elevated physiologic responses that were comparable to those of PTSD patients who had experienced other (noncancer-related) traumas. This finding suggests a good fit between cancer patients and the PTSD trauma model, as it shows comparable symptoms of increased arousal in cancer patients. Also, in a factor analytic study 7 designed to confirm the presence of the three broad PTSD symptom clusters (reexperiencing, avoidance of reminders, and hyperarousal), researchers found some tentative support for the DSM-IV symptom clusters in a sample of breast cancer survivors.
In a study of 74 women breast cancer survivors interviewed at 18 months postdiagnosis via the Structured Clinical Interview for DSM (SCID), three groups were identified: one meeting the full criteria for PTSD (n = 12), another meeting partial but not full criteria for PTSD (i.e., subsyndromal, n = 5), and a no-PTSD group (n = 47). Further analyses investigated group differences. Some notable differences between the full-criteria PTSD group and the subsyndromal group include a significantly higher number of violent traumas (e.g., physical abuse, rape) and a higher number of anxiety disorders prior to a cancer diagnosis among the full-criteria PTSD group. 8
The PTSD group was also found to have more advanced disease (75% stage III vs. 7% in the subsyndromal group and 6% in the no-PTSD group), more extensive surgeries (83% modified radical mastectomy vs. 47% in the subsyndromal group and 38% in the no-PTSD group), and a higher lifetime prevalence of prior PTSD (42%) than the subsyndromal group (7%) and the no-PTSD group (9%). 8
Further research will be needed to continue to investigate the important question of how well the conceptual model of PTSD as an anxiety response to a major life trauma fits the life experience of patients with cancer. Reviews have argued both in favor of 9 and against 6 the continued use of trauma models for conceptualizing the experience of cancer. Others have proposed alternate conceptual models. 5 10
A variety of sociodemographic, disease-related, psychosocial, and psychological variables have been investigated to determine their relationship to post-traumatic stress disorder (PTSD). At present, no clear picture emerges about who is at increased risk of developing PTSD following diagnosis or treatment of cancer.
Few patient characteristics have been shown to predict the occurrence of PTSD. High levels of psychologic distress have been correlated with both stress symptoms 1 2 3 and full-syndrome PTSD diagnoses in adult survivors. 1 In addition, trait anxiety was found to predict post-traumatic symptoms in the parents of survivors of childhood cancer. 4 Women who are survivors of cancer and who have a diagnosis of lifetime PTSD tend to have a history of exposure to trauma. 1 5 Demographic characteristics such as age, sex, and education level at time of diagnosis have not been reliable predictors of stress symptoms. 1 6 7
Demographic variables that have been associated with a higher incidence of PTSD include younger age, 8 9 fewer years of formal education, and lower income; 10 11 however, one study 1 failed to find any significant differences between the mean age of cancer patients with or without PTSD. In noncancer community samples, PTSD occurs almost twice as often in women; however, the gender differences in incidence of PTSD in cancer patients has been mixed.
Disease-related variables that have been associated with a higher incidence of PTSD in patients who underwent bone marrow transplant include more advanced disease and a longer hospital stay. 12 Other studies, however, have found no association between time since diagnosis and treatment, severity of disease, or type of cancer treatment received. 1 9 11 The relationship between disease stage and post-traumatic symptoms has not been adequately studied. Most studies have not found an association; however, they typically include a limited range of disease stages or are studying early-stage cancer. 13
The time since diagnosis and treatment has been shown to correlate with and predict post-traumatic symptoms in survivors of osteogenic sarcoma 2 and Hodgkin lymphoma. 7 14 Specifically, persons who were farther from diagnosis and treatment tended to exhibit fewer symptoms. This effect, however, has not been found in studies of patients with recent recurrences, 15 survivors of breast cancer, 1 or survivors of childhood cancers. 16 Duration of treatment, rather than time since treatment, has been shown to predict stress symptoms in survivors of childhood cancer. 16 (Refer to the PDQ® summary on Pediatric Supportive Care for more information.)
The presence of pain and other physical symptoms has been shown to correlate with levels of intrusive thoughts. 2 Cancer recurrence has also been shown to increase the likelihood of stress symptoms in patients. 15
The experience of past traumatic events appears to be an important psychosocial risk factor associated with post-traumatic symptoms, 5 17 18 as was found in both early-stage 19 and metastatic breast cancer. 20 Previous trauma in combination with recent stressful life events was significantly related to post-traumatic symptoms. 21
Other psychosocial risk factors such as premorbid psychopathology, 22 23 high levels of general psychologic distress, 24 and dysfunctional coping and attributional styles 17 25 26 have been linked to a risk for PTSD in war veterans, Holocaust survivors, and other disaster victims. In addition, several investigators have linked predisposing genetic factors 27 and other biologic factors (e.g., overly reactive hormonal systems and reduced hippocampal volume) to risk for PTSD. 28 29 30 Among social factors, the quality of the recovery environment, often measured in terms of social support, has been shown to affect risk for PTSD following exposure to combat 22 and burn injury. 31 The effect of threat to life and body integrity has been documented in samples of adults and families 1 4 14 but not children. 16
Psychological variables that have been related to a higher incidence of PTSD include a history (precancer diagnosis) of PTSD, 5 10 increased use of avoidance coping, and lower levels of social support. 32
The availability and timeliness of accurate health-related information may also offer protection from stress response symptoms. Women who met the diagnostic criteria for Acute Stress Disorder (ASD) reported significantly less satisfaction with the communication of their cancer diagnosis; 33 similarly, women who were unaware of their cancer stage reported higher stress response symptoms than those who were more knowledgeable about the stage of their disease. 34 To the extent that adequacy of information reflects the quality of a patient's relationship with medical staff, another protective factor may be the quality of those relationships. Difficult patient-staff relationships have been reported to be predictors of stress response symptoms in women with cancer. 35
PTSD is precipitated by an intensely distressing event; however, this factor alone is not sufficient to explain the disorder. Not everyone exposed to a traumatic stressor develops the full-blown syndrome (or subsets of symptoms) or qualifies for the diagnosis. Attempts to explain these differences and to predict who is vulnerable have focused on psychologic (i.e., learning theory), biologic (especially hormonal), and social (i.e., social support) factors. Early studies of Vietnam War veterans suggested a two-factor learning theory to account for trauma-related pathology. 36 37 The same theory has also been applied to development of PTSD in patients with cancer. 38 39 40
PTSD symptoms develop as a function of both classical conditioning and instrumental learning. Classical conditioning accounts for the fear responses elicited by various stimuli that are associated with the original traumatic event. Neutral stimuli (e.g., smells, sounds, and visual images) previously paired with the aversive stimuli (e.g., chemotherapy or painful procedures) eventually evoke anxiety, arousal, and fear when presented alone, even after the trauma has ended. Higher order conditioning and stimulus generalization account for the exacerbation and extension of symptoms to additional stimuli. Once established, PTSD symptoms are maintained through instrumental learning, that is, avoidant responses are reinforced because avoidance of the stimuli prevents unpleasant feelings and thoughts.
Estimates from epidemiologic studies suggest that on average, 25% to 33% of individuals who are exposed to traumatic events, including cancer, develop PTSD. 28 41 Although the disorder appears to be a result of learning processes, many factors have been suggested to explain why one person develops PTSD and another does not.
A timely and careful assessment of cancer patients is critical to identify the symptoms of post-traumatic stress disorder (PTSD), to note the deleterious impact of the symptoms on functioning, and to plan interventions targeted at the most distressing symptoms. It is also critical that the assessment distinguishes between the full Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV), PTSD syndrome (meets all required diagnostic criteria) and PTSD-related symptoms only.
The most difficult aspect of PTSD assessment in the cancer setting is the determination of precisely when to evaluate the patient. Diagnosis is complicated because cancer is not an acute or discrete event, but is an experience marked by repeated traumas and indeterminate length. Thus, an individual may exhibit the symptoms of PTSD at any point from diagnosis through treatment, to treatment completion and, possibly, to recurrence. 1 Patients such as Holocaust survivors whose history of victimization causes PTSD or its symptoms can have the symptoms activated by any number of stimuli encountered during their treatment (e.g., clinical procedures such as being inside magnetic resonance imaging or computed tomography scanners). While such patients may have more difficulty in adjusting to cancer and cancer treatment, their PTSD symptomatology is likely to vary greatly according to their circumstances. The relative predominance of specific PTSD symptoms may wax and wane throughout the cancer experience and beyond. 2
The definition in the DSM-IV indicates that although PTSD symptoms usually begin within the first 3 months after trauma, there may be a delay of months or even years before symptoms appear. 2 3 These findings support the necessity for long-term monitoring of survivors of cancer and their family members.
At least one study found that individuals who have experienced a traumatic event may exhibit early symptoms without meeting the full criteria for a diagnosis of PTSD. 4 Nonetheless, the appearance of these early symptoms was found to predict later development of full PTSD syndrome. These results lend further credence to the need for both repeated and long-term follow-up of individuals exposed to the trauma of cancer. (Refer to the PDQ® summary on Adjustment to Cancer: Anxiety and Distress for further information.)
The difficulty in properly diagnosing PTSD may be compounded by the overlapping of PTSD symptoms with those of other psychiatric disorders and by the time-related aspects of normal adjustment. For example, irritability, poor concentration, hypervigilance, excessive fear, and disturbed sleep are also symptoms of generalized anxiety disorder. Other arousal and avoidance symptoms are common to PTSD, phobias, and panic disorder, but loss of interest, sense of a foreshortened future, avoidance of other people, and sleep impairment might suggest both PTSD and depressive disorders. Even normal reactions to the diagnosis and treatment of life-threatening disease can consist of responses such as intrusive thoughts, disassociation and depersonalization, sleep disturbances, and heightened arousal. Therefore, clinicians and researchers must be particularly attuned to the causes, duration, and severity of PTSD-like symptoms when considering PTSD among several diagnoses. For instance, in a study of women with breast cancer, 41% reported experiencing intense fear, helplessness, or horror (DSM-IV PTSD diagnostic criterion A2); however, on further comprehensive diagnostic interview, only 4% met the full PTSD criteria. Assessment must be able to distinguish between general psychological distress and symptoms of PTSD. 5
The accurate diagnosis of PTSD also requires the use of reliable and valid instruments. Many studies have used the PTSD module of the Structured Clinical Interview for DSM-III-RNonpatient Edition (SCID-NP). 6 This is a clinician-administered, structured clinical interview that is time intensive and may not be feasible in settings without adequately trained mental health professionals. However, one study 7 investigated the utility of a cost-effective screening tool, the PTSD Checklist-Civilian Version (PCL-C). 8 In this study of 82 women diagnosed with breast cancer assessed 6 to 72 months after cancer treatment, use of the PCL-C resulted in a sensitivity of .60 and specificity of .99. Other cutoff scores for the PCL-C that could be used were discussed, depending on the clinical resources available in specific cancer treatment settings. Most research studies have used the Impact of Event Scale, a self-report of intrusive thoughts; 9 however, it is important to note that this is not designed to be an assessment procedure for PTSD.
In attempting to diagnose PTSD, it is important to be aware that this disorder is often marked by comorbid psychopathology. Substance abuse, affective disorders, and other anxiety disorders are consistently encountered in samples of people with PTSD. 2 10 11 12 It has been reported that war veterans with PTSD exhibited substantial comorbid pathology that included major depression (32% to 72%), alcohol dependence (65%), drug dependence (40%), social phobia (50%), and obsessive-compulsive disorder (10%). 13 High rates of concurrent disorders have also been documented in other trauma victims. For example, 40% to 42% of disaster survivors with PTSD also qualified for a diagnosis of major depression, and 20% to 42% met the criteria for concurrent generalized anxiety disorder. 13 14 While this has not yet been studied in cancer patients or survivors, the presence of co-occurring psychiatric disorders in Vietnam War veterans and other trauma victims would indicate that cancer clinicians should be alert to the need to identify and treat such related syndromes in their patients.