Tailoring Cognitive-Behavioral Therapy to Treat Anxiety Comorbid with Advanced Cancer

Abstract

Patients with advanced cancer often experience debilitating anxiety symptoms that interfere with quality of life and relate to worse medical outcomes. Although cognitive behavioral therapy (CBT) is an empirically-validated, first-line treatment for anxiety disorders, clinical trials of CBT for anxiety typically exclude patients with medical comorbidities in general, and those with terminal illnesses, such as advanced cancer, in particular. Moreover, CBT has generally targeted unrealistic fears and worries in otherwise healthy individuals with clinically significant anxiety symptoms. Consequently, traditional CBT does not sufficiently address the cognitive components of anxiety in patients with cancer, especially negative thought patterns that are rational but nonetheless intrusive and distressing, such as concerns about pain, disability and death, as well as management of multiple stressors, changes in functional status and burdensome medical treatments. In this paper, we describe a treatment approach for tailoring CBT to the needs of this population. Three case examples of patients diagnosed with terminal lung cancer are presented to demonstrate the treatment methods along with outcome measures for anxiety and quality of life.

Introduction

Epidemiology and Etiology of Anxiety and Cancer

According to the CDC National Center for Health Statistics, 559,888 persons died from malignant neoplasms in 2006, making cancer the second leading cause of mortality in the United States (Heron et al., 2009). Although the majority of individuals will be understandably distressed or anxious in response to the diagnosis of terminal cancer, a significant minority will experience persistent debilitating anxiety symptoms that are maladaptive (Kadan-Lottick, Vanderwerker, Block, Zhang, & Prigerson, 2005). In their review of studies, Miovic and Block (2007) report that the prevalence of diagnosable anxiety disorders in individuals with advanced cancer at a given point in time ranges from approximately 2%–14%. Investigators using screening instruments have observed rates of clinically significant anxiety symptoms as high as 30%–40% (Delgado-Guay, Parsons, Li, Palmer, & Bruera, 2009; Hopwood & Stephens, 2000).

Elevated anxiety comorbid with advanced cancer is associated with a number of distressing symptoms, including dyspnea (Bruera, Schmitz, Pither, Neumann, & Hanson, 2000), fatigue (Tchekmedyian, Kallich, McDermott, Fayers, & Erder, 2003), nausea and pain (Andrykowski, 1990; Delgado-Guay, Parsons, Li, Palmer, & Bruera, 2009), as well as poor quality of life (Smith, Gomm, & Dickens, 2003). Given the symptom burden and incurable nature of metastatic disease, medical treatments for advanced cancer typically involve chemotherapy and/or radiation for the purpose of preventing tumor progression as well as for palliation. Anxiety may complicate such treatments for advanced cancer, with researchers observing relationships with decreased adherence to chemotherapy (Greer, Pirl, Park, Lynch, & Temel, 2008), longer hospital stays (Prieto et al., 2002), and more aggressive care at the end of life (Temel et al., 2008), though studies to date have largely been correlational in design.

Considering the high lifetime prevalence of anxiety disorders in the general population (i.e., 28.8%; Kessler et al., 2005), many individuals presenting to their oncologists will have a premorbid history of a primary DSM-IV anxiety disorder (American Psychiatric Association, 2000), which may be exacerbated by the stress of the cancer diagnosis. For others without such a history, adjusting to the major transitions of the disease state, such as diagnosis, medical appointments, stressful procedures, follow-up imaging studies and threat of pain and functional incapacity, may cause significant anxiety symptoms. Additionally, conditioned responses (e.g., anticipatory nausea) may develop during the course of treatment (Andrykowsky, 1990). Mental health clinicians in collaboration with the oncology team must be careful to rule out organic causes for symptoms related to disease (e.g., blood clot in the lung) or medications that may be incorrectly attributed to primary anxiety (Roth & Brietbart, 1996). Consideration of these various etiological factors is essential for differential diagnosis and appropriate treatment of anxiety.

Psychosocial Interventions for Anxiety Comorbid with Cancer

Over the last several decades, researchers have conducted numerous investigations examining the utility of various psychosocial interventions (e.g., group support, individual psychotherapy, psychoeducation, relaxation training, among others) to treat psychological distress, manage pain, improve quality of life, and even extend survival in patients with cancer (Meyer & Mark, 1995; Newell, Sanson-Fisher, & Savolainen, 2002; Sheard & Maguire, 1999; Tatro & Montgomery, 2006). In a meta-analysis of 19 randomized trials to estimate the impact of psychological interventions on anxiety in patients with cancer, Sheard and Maguire (1999) found a moderate effect size (i.e., standardized mean difference = 0.36). However, only three studies restricted inclusion to individuals experiencing clinically significant anxiety, hence diminishing the potential for an effect size that would be compatible to those in anxiety treatment studies. The authors therefore concluded that interventions targeting patients at risk of or suffering from marked anxiety symptoms would likely yield greater treatment effects.

Although cognitive-behavioral therapy is considered an effective first-line treatment for diagnosable anxiety disorders in the general population (Otto, Smits, & Reese, 2004), the use of CBT for anxiety in patients with cancer can only be tentatively recommended based on the present scientific literature. In a rigorous and systematic review of psychosocial intervention studies, Newell and colleagues (2002) did not find consistent evidence for the efficacy of CBT for anxiety comorbid with cancer, noting poor methodological quality, small sample sizes, and short-term follow-up periods among the few existing trials. Only 15% of reviewed studies included individuals with advanced disease, underscoring the lack of attention researchers give to patients with later-stage cancers. Moreover, this population is confronted with realistic fears regarding disease progression, disability, and death, which existing cognitive-behavioral interventions for anxiety disorders fail to address. More recent meta-analyses have demonstrated that CBT is effective for reducing distress (ES = 0.31) in patients with breast cancer (Tatro & Montgomery, 2006) as well as treating anxiety (ES = 1.99) and improving quality of life (ES=.91) in cancer survivors (Osborn, Demoncada, & Feuerstein, 2006), though such individuals may more closely resemble the general population than patients with advanced disease undergoing palliative treatment. Taken together, these findings highlight the need for developing and testing cognitive-behavioral interventions that target the concerns of patients with advanced cancer experiencing complex medical symptoms and decrements in functioning.

Rationale for Tailoring Cognitive-Behavioral Therapy for Anxiety Comorbid with Cancer

Traditional CBT techniques are generally based on the premise that anxious individuals unrealistically overestimate negative outcomes (Beck & Emery, with Greenberg, 1985), which leads to avoidance and other maladaptive coping behaviors. Restructuring distortions in perceptions, combined with exposure-based therapies, therefore allows individuals to overcome anxiety through using more adaptive thinking, habituation, and decatastrophization of feared predictions (Barlow, 2002). However, cancer, with its related symptoms and treatment side effects, is a disease that genuinely may cause severe pain, functional impairment, and death. Thus, mental health clinicians may find that classic CBT techniques, such as cognitive restructuring, are inadequate or even inappropriate for patients with realistic fears related to the cancer diagnosis and treatment, particularly among individuals with advanced disease or shortened life expectancy. Yet, intrusive worries regarding medical treatments, imaging procedures, and poor prognosis still cause marked impairment and disruption in quality of life, requiring alternate and innovative cognitive interventions.

As noted, the strong relationships between anxiety and cancer-related symptoms are well-documented in the literature. Patients often report an escalation of anxiety when experiencing somatic symptoms like nausea, fatigue, or pain, which may be functionally limiting and difficult to interpret. This anxiety in turn, through the physiologic changes associated with the acute stress response (Cannon, 1929), can heighten the perceived severity of disease and treatment-related symptoms. Comprehensive cognitive-behavioral approaches to the management of anxiety in cancer patients must therefore address salient medical illness concerns. To bridge this gap in the treatment literature, several authors have begun publishing clinical manuals tailored to individuals coping with various diseases including cancer (Moorey & Greer, 2002; Taylor, 2006; White, 2001), though in large measure these approaches have yet to be empirically validated.

In this paper, we describe a brief CBT protocol that targets anxiety and quality of life in patients with terminal cancer, which is currently being tested in a randomized control trial funded by the National Cancer Institute (R03CA128478, PI: Greer). Considering the legitimate threat to survival and changes in self-schema due to the diagnosis and treatment of advanced cancer, the proposed CBT intervention has been tailored to incorporate strategies for managing the complex interplay between the psychological and medical concerns. As depicted in the treatment algorithm presented in Figure 1 (Greer, Graham, & Safren, 2009), the components of the CBT intervention in the current study were selected according to research on cognitive-behavioral treatment outcome for anxiety in general (Barlow, 2002, 2008), approaches used with patients with chronic illnesses (e.g., Barlow, Wright, Sheasby, Turner, & Hainsworth, 2002; Keefe, Abernethy & Campell, 2005), and approaches used in coping with cancer in particular (Moorey & Greer, 2002; Given, Given, Champion, Kozachik, & Devoss, 2003). Various features of the treatment are geared toward advanced cancer and differ from traditional CBT. The rationale for these differences is interwoven into the description of the treatment modules.

Figure 1.

CBT Treatment Algorithm for Anxiety Comorbid with Advanced Cancer

Description of Treatment Approach

Overview of Approach

As shown in Table 1, four core modules comprise the cognitive-behavioral intervention for individuals with anxiety and advanced cancer: 1) psychoeducation and goal setting; 2) relaxation training; 3) coping with cancer fears; and 4) activity planning and pacing. The goal of the proposed treatment approach is to help patients with advanced cancer learn coping strategies for reducing anxiety and worry, while simultaneously developing skills for managing cancer-related symptoms and treatment side effects. The treatment is brief, consisting of only 6–7 sessions, which is of heightened importance in this context given the short life-expectancy of many patients. Also, the format is modular, to allow for maximum flexibility in the application of the intervention and relevance for the specific problem areas of each patient.

Table 1.

Components of CBT Manual for Anxiety and Advanced Cancer

Module

Psychoeducation & Goal Setting Review CBT model of anxiety within context of advanced cancer Establish treatment goals Relaxation Training Discuss associations between anxiety, physiologic stress-response, and cancer-related symptoms and treatment side effects Teach (modified) diaphragmatic & pursed-lips breathing Practice and audio record autogenic relaxation exercise Coping with Cancer Fears Identify automatic thoughts related to having advanced cancer Examine evidence to differentiate realistic from unrealistic worries Gather additional data from credible sources if needed Utilize cognitive restructuring skills for cognitive errors Apply either problem-focused or acceptance-based/adaptive, emotion-focused strategies for realistic concerns Activity Planning & Pacing Assess and prioritize daily activities Categorize days by functional capacity Plan and pace activity schedule by time rather than by task

Module 1: Psychoeducation & Goal Setting (approximately one session)

The primary aims of this module are to 1) elicit patients’ understanding of their anxiety symptoms in the context of having advanced cancer; 2) provide an overview of the cognitive-behavioral model of therapy including the ways people with advanced cancer may experience anxiety physiologically, cognitively, and behaviorally; and 3) increase motivation for treatment by clarifying patients’ goals and role for CBT to aid in achieving these goals. Module 1 begins with an open-ended discussion about patients’ cancer experiences, current treatments, and illness-related concerns, in order both to help patients appreciate the healthy aspects of anxiety, such as increasing motivation to seek treatment and social support, as well as identify any maladaptive, distressing, and interfering symptoms.

To illustrate the negative cycle of anxiety, the therapist interactively reviews the CBT model, eliciting from patients the cognitive, behavioral and physiological symptoms of their anxiety. The cognitive component of the CBT model may elucidate the types of cancer-related fears or worries that the patient experiences. For example, many individuals with advanced cancer monitor and interpret (at times inaccurately) somatic symptoms, such as pain or dyspnea, as confirmation of worsening disease and imminent death. Managing changes in role functions finances, or family issues may also be of concern (Hill, Amir, Muers, Connolly, & Round, 2003), as many patients will be unable to work and may worry about the care of their loved ones.

Behavioral manifestations of anxiety are typically expressed as avoidance of situations or activities that trigger threat (Barlow, 2002). In the context of cancer, such behavioral avoidance may involve refusal to adhere to medical treatments or screenings due to excessive fears about the procedures, results, or potential side effects. Alternatively, patients may engage in excessive monitoring of somatic symptoms, reassurance seeking from clinicians, and ineffective symptom management resulting in activity withdrawal (Stark & House, 2000; Stark et al., 2004).

Common physical symptoms of anxiety include rapid breathing, heart palpitations, muscle tension, abdominal changes, and lightheadedness, etc. Though part of the body’s natural and adaptive defense against threat (Nesse, 1999), these symptoms may be exacerbated by the disease process of cancer and related treatments. For example, dyspnea, fatigue, pain, and gastrointestinal symptoms are also frequent complaints of people undergoing chemotherapy and radiation. The overlap in symptoms is often challenging to disentangle, such that the patient (and therapist) may be unsure whether to attribute symptoms to anxiety, disease progression, or treatment side effects (Massie & Greenberg, 2005). Regardless of etiology, many of the CBT techniques are helpful for managing both anxiety and medical symptoms. Reviewing the CBT model of anxiety helps to demonstrate with patients how these symptoms function together to form a negative cycle, each bidirectionally influencing one another. For example, dyspnea may trigger thoughts or worries about disease progression and lead to withdrawal from social activities, personal responsibilities or self-care. Such reduction in activity level may cause further deconditioning and ongoing problems with breathlessness. Thus, the goal of CBT is to break this negative cycle with interventions that are tailored to cognitive, behavioral, and physiological domains, reinforcing adaptive coping with both anxiety and medical symptoms.

To conclude the discussion of the CBT model of anxiety, patients examine their reasons for seeking treatment, list two to three, clearly-defined treatment goals (e.g., managing worry more effectively in anticipation of medical scans), and explore potential obstacles to achieving such goals (e.g., treatment side effects interfering with attendance). The remaining modules may then be flexibly applied to address patients’ specific presenting concerns and goals.

Module 2: Relaxation Training (approximately one session)

The primary aims of this module are to: 1) help patients understand the nature of the acute stress response and ways that physiological anxiety may exacerbate cancer-related symptoms; 2) build mastery in coping with dyspnea and other somatic symptoms through breathing retraining and practice of autogenic relaxation. Such techniques play an important role for self-management of chronic conditions and cancer-related symptoms (Barlow, Wright et al., 2002; Bausewein, Booth, Gysels, & Higginson, 2008), not simply for anxiety reduction.

Part of relaxation training involves psychoeducation aimed at helping patients identify the symptoms of the acute stress response (Cannon, 1929), clarify interpretations of these bodily changes and how they relate to cancer symptoms, and learn about the mechanisms and utility of activating the relaxation response (Benson, 1975). Many of the physiological symptoms associated with the stress response clearly overlap with cancer progression and treatment side effects. For example, such cancer-related symptoms may include: breathlessness (e.g., due to malignancy invading lung tissue); palpitations (e.g., due to coticosteroids or beta-2 agonists); peripheral tingling/numbness (e.g., due to neuropathy secondary to chemotherapy toxicity); nausea (e.g., due to chemotherapy); restlessness and akathesia (e.g., due to anti-emetic medications) (Holland, Greenberg, & Huges, 2006). Prior to initiating CBT, the therapist should ensure that patients have reviewed any physiological symptoms with their physicians to rule out life-threatening causes. Once medically cleared, relaxation techniques may be helpful for symptom management regardless of whether the distress is primarily related to the physiological stress response, cancer progression, anticancer therapy, or a combination of all three.

Breathlessness is common among individuals with advanced cancer, with approximately half of patients reporting clinically significant symptoms (Bruera et al., 2000; Tanaka, Akechi, Okuyama, Nishiwaki, & Uchitomi, 2002a). This module includes an assessment of the times and places in which patients experience breathlessness, noting whether symptoms are at rest, with exertion, or at moments of emotional distress. Although not all patients will report dyspnea at baseline, some will develop this symptom as their disease progresses or as a side effect from anticancer therapy. During breathing retraining, some patients may find it helpful to lean slightly forward (e.g., 45 degree angle) when in the seated position to achieve the optimal mechanics of breathing (O’Neill & McCarthy, 1983).

Instruction in diaphragmatic and pursed-lips breathing are intended to manage symptoms like dyspnea as well as reduce patients’ general level of stress. Individuals who are unable to take full, deep breaths due to severe lung disease may modify diaphragmatic breathing by focusing on slowing the breath and extending the expiratory phase rather than taking deep inhalations. Practice of this exercise should occur in a quiet, comfortable place when calm, rather than during acute attacks of dyspnea. By contrast, pursed-lips breathing may be used for symptomatic relief during episodes of breathlessness, especially after exertion, such as climbing stairs (Nield, Soo Hoo, Roper, & Santiago, 2007). Commonly used for chronic obstructive lung disease (Gosselink, 2004), pursed-lips breathing provides natural resistance to lengthen the exhalation phase, strengthen respiratory muscles, and release “trapped” air from the lungs.

Module 2 concludes with in-session practice of autogenic relaxation. As the name implies, autogenic training involves a self-generated state of relaxation that combines visual imagery and body awareness to reduce stress (Schultz & Luthe, 1969). The exercise is safe for use in a medically-ill population experiencing acute and chronic pain, as it does not include any tensing of muscles. To develop these skills further and monitor changes in anxiety, therapists may provide patients with an audio recording of the exercises for practice between sessions.

Module 3: Coping with Cancer Fears (approximately three sessions)

The primary aims of the third module are: 1) to help patients identify types of automatic thoughts and differentiate worries that are realistic from those that are biased or too negative; 2) to teach adaptive thinking skills for managing various kinds of worries using traditional cognitive restructuring techniques as well as present-focused coping strategies; and 3) to identify and problem-solve any avoidance that may be interfering with the patient’s functioning or cancer treatment. Figure 1 provides an overview of an algorithm for coping with cancer-related worries (Greer, Graham, & Safren et al., 2009; partially adapted from Moorey & Greer, 2002).

Identifying Automatic Thoughts and Worry in Individuals with Cancer

Similar to the physiological changes of the “fight-or-flight” response (Cannon, 1929), individuals experience cognitive changes during periods of heightened stress, bringing their attention to possible sources of danger (Beck & Emery, with Greenberg, 1985; Craske & Barlow, 2006). While this cognitive response is automatic and adaptive for survival, it may become problematic and functionally disabling for individuals experiencing prolonged stresses related to cancer treatment because it perpetuates a future-oriented focus rather than directing attention to the present situation. Such anxious preoccupation is particularly challenging for patients with cancer, who must undergo repeated consultations, medical scans, and procedures with delayed or unknown future outcomes. In this context, the patient’s negative thought patterns will likely include a combination of cognitive distortions as well as realistic concerns regarding the physical, emotional, social, and existential challenges related to having advanced cancer and undergoing debilitating treatments. Although many of the worries associated with advanced cancer have a basis in reality, the intensity of the worry may nonetheless be magnified by unrealistic, catastrophic predictions and negative beliefs about coping abilities, compounding the existing burden. Negative cognitions common to individuals with cancer include thoughts that the development of cancer was deserved, that others will not understand what they are going through, and that they will become a burden to family members (White, 2001). Traditional cognitive restructuring techniques (Beck, 1995) that help patients identify and challenge cognitive errors are valid for worries that clearly reflect errors or biases in perception, though a different approach is needed for realistic worries.

Managing Realistic Worries

Since many serious concerns of patients with advanced cancer do not result from cognitive errors, these individuals will benefit from learning strategies for differentiating realistic from unrealistic worries, applying either problem-focused or adaptive emotion-focused techniques for managing the realistic concerns. Specifically, to differentiate the realistic fears patients must first learn to examine the evidence that either supports or disconfirms their thought patterns. If the therapist and patient are unable to determine whether the automatic thoughts are realistic, obtaining data from credible sources, such as consultation with the oncology team or the results of medical scans, may be an important first step.

When it becomes evident that the patient’s worry is indeed realistic, the next step would be to apply the worry algorithm depicted in Figure 1 (Greer, Graham, & Safren, 2009). Patients would apply a problem-focused approach in instances when they are able to execute some action in addressing the worry versus an acceptance-based or emotion-focused approach when no further action will lead to resolution. Specifically, by employing problem-solving skills, patients with advanced cancer learn methods for breaking down overwhelming tasks into manageable steps with the goal of reducing cognitive and behavioral avoidance (D’Zurilla, 1986). Typically, these steps include defining the problem, articulating possible solutions, selecting the best possible solution, and developing an action plan (Nezu, Nezu, Friedman, Faddis, & Houts, 1998).

In contrast, patients who have already taken every action they can to resolve their realistic worry must shift their attention and target the distress from the anxiety rather than addressing the problem itself. Acceptance-based interventions (e.g., mindfulness meditation) or adaptive emotion-focused techniques (e.g., self-soothing or pleasurable activity) are intended help an individual become less focused on the future with its negative images of potential threats and instead become more aware of the present moment and capacity for coping (Hamilton, Kitzman, & Guyotte, 2006). Of course, patients should identify those activities from their own lives which are most likely to achieve these goals, recognizing that their participation may need to be modified to account for any functional limitations.

One realistic and powerful fear for many patients with cancer, especially those with advanced disease, is the fear of dying. Typically, patients’ concerns pertain to the fear of pain and suffering; fear of loneliness and concerns about how loved ones will be cared for after death; as well as fears of the unknown and existential worries (Moorey & Greer, 2002; Sigal et al., 2008). Acceptance of having a terminal illness is an evolving process that occurs over time, or in some cases, never. The therapist’s role is not to confront the patient in these moments, but rather to provide safety and support for patients to discuss their fears about death openly when they are willing; to clarify and perhaps adjust priorities; to set specific short-term goals; and to help frame the experience of the dying patient in terms of increasing quality rather than length of life.

The three sessions of Module 3 dedicated to coping with cancer fears are tailored to individual patient concerns and may include some or all of the aforementioned strategies. Therefore, homework assignments typically involve having patients complete self-monitoring forms that help them first differentiate the types of worries they experience and then take an appropriate course of action, such as cognitive restructuring, employing the worry algorithm, seeking additional information or consultation to clarify concerns, or completing any action plans formulated in session based on problem-solving exercises.

Module 4: Activity Planning and Pacing (approximately one session)

Fatigue and physical pain are among the most common and distressing symptoms experienced by individuals with advanced cancer (Butt et al., 2008; Tanaka, Akechi, Okuyama, Nishiwaki, & Uchitomi, 2002b), causing significant disruption in daily life activities. Perhaps due to the physical limitations, role interference, and exacerbation of cancer-related worries, somatic symptoms such as fatigue and pain are strongly correlated with anxiety and poor quality of life in patients with cancer (Theobald, 2004; Stark et al., 2002). Thus, the primary aims of this module are: 1) to help patients prepare and adjust for fluctuations in stamina and cancer-related symptoms; 2) to teach skills for prioritizing activities, including daily responsibilities as well as pleasurable events; and 3) to plan and pace activities according to patients’ priorities and required exertion with the goal of maintaining function and engagement.

Although individuals with advanced cancer receive medical treatments that may be significantly debilitating, many patients at times overexert themselves by undertaking daily tasks in the manner to which they were traditionally accustomed. A comprehensive assessment of activities would ideally review both daily responsibilities and pleasurable events, with particular focus on individuals’ evolving personal priorities and goals. Additionally, this discussion may examine the role of social supports for maximizing patient participation in personally valued life activities, considering that concerned family and friends may either take on too much or too little responsibility in helping their ill loved one. For example, a 60-year-old woman with advanced pancreatic cancer may want to prepare dinner for the family, but perhaps feel smothered by her overly solicitous children who refuse to allow her to do any cooking or housework. She may lack the necessary communication skills to assert herself with loved ones and clarify her wishes.

After evaluating participation in daily activities, the patient and therapist are able to then categorize days by functional capacity, accounting for the fluctuations in stamina and disease-related symptoms. For this intervention, we use the following categorization scheme: “A”-Day: Able to participate in and complete at least 75% of usual activities, which may include both daily tasks and pleasurable events; “B”-Day: Able to participate in and complete approximately 50–75% of usual activities, with equal emphasis on self-care as on daily tasks or responsibilities; “C”-Day: Able to participate in or complete less than 50% of daily tasks or responsibilities, requiring greater emphasis on activities directed toward self-care. The therapist may want to have the patient generate recent examples of each type of day, highlighting and affirming instances in which the patient engaged in adaptive behaviors. Although the patient may not be functioning at the level desired prior to cancer diagnosis, preparing plans for how to respond to such decrements helps build a sense of self-efficacy in managing physical symptoms.

Another aim of activity planning and pacing is to conserve energy while maximizing function. For many individuals with advanced cancer, the notion of changing their traditional activity schedule is undesirable, especially when having an “A”-Day. However, when patients overexert themselves by working until the point of exhaustion, they are likely to experience more extreme fatigue or functional impairment in the following days, reducing their ability to engage in desired activities. Alternatively, individuals who under-exert themselves may become physically deconditioned, also leading to worse fatigue and pain symptoms. Therefore, activity pacing requires intermittent rests during physically demanding tasks in order to accomplish more and maintain optimal levels of energy and functioning within and across days (Otis, 2007).

To help patients optimally participate in daily activities, time-based pacing involves breaking activities down by time rather than by task (Caudill, 2008; Otis, 2007; Turk & Gatchel, 2002). Specifically, the patient and therapist first estimate the amount of time it takes to engage in an strenuous activity (e.g., doing laundry, dishes, etc.) until the patient’s pain or fatigue level reaches a 6 on a ten-point scale, in which 1=no fatigue/pain and 10=the most severe fatigue/pain. The patient then plans to undertake these activities according to the estimated time frame, and paces the tasks by alternating them with intermittent periods of rest, which may involve simple, relaxing breaks (e.g., reading) or sedentary but productive activities (e.g., paying bills). Module 4 concludes with the therapist and patient applying time-based pacing strategies and developing schedules for “A,” “B,” and “C” days, incorporating a variety of pleasant and self-care activities that are meaningful and interesting to the patient in order to conserve energy and reduce stress.

To elucidate our treatment approach further, we next present case descriptions of three participants who were diagnosed with terminal lung cancer from the open pilot phase of our ongoing RCT. The details of the cases have been altered to protect patient confidentiality.

Case Examples

Participants

Adult patients diagnosed with metastatic cancer who were at least six weeks post cancer diagnosis and experiencing clinically significant anxiety symptoms (i.e., a Hamilton Anxiety Rating Scale score of ≥ 14) were able to enroll in the open trial. To increase generalizability, the sample eligibility criteria did not require a diagnosis of a particular anxiety disorder or the presence of psychotropic medication. Rather, we selected a cutoff on the HAM-A denoting marked anxiety symptoms and impairment in functioning.

Measures

The primary outcome was assessed with the Hamilton Anxiety Ratings Scale (HAM-A; Hamilton, 1959). Widely used in psychiatry research to evaluate somatic and psychic anxiety symptoms, the HAM-A consists of 14 items that are scored on scale from 0 (not present) to 4 (very severe) and summed to a total value ranging from 0–56. A structured interview guide was created for the scale (Shear et al., 2001), which possesses strong psychometric properties (test-retest reliability = 0.89; α = 0.82) and was administered in the present study.

The Beck Anxiety Inventory (BAI; Beck, Epstein, Brown, & Steer, 1988) is a 21-item self-report questionnaire that measures severity of cognitive and somatic anxiety symptoms in the past week, with a total score ranging from 0–63. The instrument demonstrates good test-retest reliability (r = 0.75) and internal consistency (α = 0.92).

The Functional Assessment of Cancer Therapy-Lung (FACT-L; Cella et al., 1995) is a 34-item questionnaire that assesses health-related quality of life across multiple domains including physical, social, emotional, and functional wellbeing, as well as seven symptoms specific to lung cancer, such as shortness of breath and chest discomfort. Items are measured on a 5-point scale, yielding a total score of 136, with higher scores indicating better overall quality of life. Internal consistency estimates for the different subscales range from 0.56 to 0.80.

The Clinical Global Impression Scale (CGI; NIMH, 1985) is widely used in psychiatric research to measure overall illness severity and impairment. The CGI is a one-item scale that ranges from 1 (not at all ill) to 7 (extremely ill).

Procedure

Participants were referred to the study by oncology clinicians at a large academic cancer center. After providing informed consent, participants underwent a baseline evaluation with the study therapist, completed the six-session CBT protocol over the course of two months, and then received a post-assessment with an independent assessor who was blind to study condition. A licensed clinical psychologist served as the study therapist and administered all clinician-rated assessments at baseline. A master’s-level psychology trainee served as the independent assessor for the post-assessment, rating anxiety symptoms with the HAM-A as well as assigning a corresponding CGI score. Participants received up to $50 as remuneration for their time in completing the clinical interviews and self-report measures. For each of the three case examples, we describe the history of the presenting anxiety symptoms, case conceptualization, course of CBT, and outcome data. The study was approved by the institutional IRB prior to initiation.

Case Study 1

History of Presenting Problem

“Mary,” a 59-year-old, married Caucasian woman living with stage IV non-small cell lung cancer, was referred to the study for evaluation and treatment of anxiety symptoms, especially dyspnea. At the time of the referral, she had been diagnosed with advanced lung cancer for over a year and was receiving palliative chemotherapy to prevent disease progression. The patient acknowledged a history of mild to moderate anxiety symptoms in response to life stressors in the past, but she had never sought mental health treatment prior to the study. During the baseline evaluation, she reported symptoms of excessive worry, muscle tension, fatigue, concentration difficulties, and significant breathlessness, scoring in the moderate to severe range on the HAM-A (Total Score = 25). Her low score on the FACT-L (Total Score = 66) was consistent with these changes in physical and functional wellbeing.

Case Conceptualization

Mary was living with advanced lung cancer with uncertain life expectancy. Although medical evaluation revealed no notable impairments in lung function, the breathlessness caused marked distress, triggering fears that the cancer was progressing. Moreover, Mary’s dyspnea was coupled with debilitating fatigue, an expected consequence of chemotherapy, which she initially responded to by pushing herself too hard followed by withdrawing physically and socially. For example, she had been an avid golfer, though at the time of study enrollment she was avoiding strenuous activities due to concerns about overexertion and feeling out of breath. Such withdrawal led to greater deconditioning that only exacerbated the physical symptoms, reinforcing her fears that the cancer was worsening.

Treatment Course and Outcome

The focus of Mary’s initial session was on introducing the CBT model and defining her goals for treatment. Mary reported that her primary aims were to learn more effective strategies for coping with anxiety and breathlessness as well as to resume involvement in activities that she enjoys but had been avoiding, such as golf.

Considering Mary’s reported difficulties with dyspnea as well as pain and tightness in the muscles around her lower rib cage, we focused less on diaphragmatic breathing in the second session and emphasized slowing the breath by counting and using pursed-lips breathing, especially upon exertion. Mary also enjoyed learning the autogenic relaxation exercise in session, which we audio recorded for daily practice at home. At next follow up, she reported that, although the frequency of the dyspneic episodes remained similar, she found them less distressing, such as when walking.

Subsequent sessions focused on cognitive restructuring skills and self-monitoring of automatic thoughts. Mary’s thought records revealed worries centered on her cancer, medical treatment, side effects, and role limitations. For example, a particularly anxiety-provoking event occurred when her oncologist recommended that she terminate participation in a clinical trial and switch chemotherapy regimens. Mary initially refused because, although the anticancer therapy had stopped working, she appreciated that the trial required follow-up scans to be scheduled every three weeks, allowing close monitoring of her disease. Off trial, she would only have scans every six weeks and feared the cancer would grow uncontrollably in that time. Using the worry algorithm depicted in Figure 1, she decided to obtain further medical advice and learned that the frequency of scans would not alter her medical outcome. Mary was then able to challenge these fears with evidence from her oncologist and develop more rational responses, noting that to remain on an ineffective chemotherapy because of the scan cycle was not in her best interest.

Although Mary achieved some gains in applying traditional cognitive restructuring techniques to her worries, it became clear over the course of CBT that her fatigue was not simply the result of chemotherapy but also likely because of disease progression. For this recurrent and intrusive worry, cognitive restructuring was not sufficient to reduce her distress. We therefore drew on other acceptance-based and adaptive emotion-focused strategies to help manage her anxiety. Using Figure 1, Mary appropriately and realistically recognized that the fatigue may not remit entirely and no further action could be taken to resolve this particular problem. In such moments, we made a shift in treatment strategy, encouraging Mary to tolerate her fears by participating in activities that help her orient to and accept her present-moment experience, such as practicing mindfulness meditation and self-soothing techniques. Specifically, she would try to bring her conscious awareness to her bodily sensations, noticing her feelings and somatic symptoms in a non-judgmental manner as part of mindfulness practice (Baer, 2003). In addition, Mary used reading, music, and guided imagery as strategies for letting go of her fears about disease progression, in order to maintain the highest quality of life in the present moment.

The final sessions addressed activity planning and pacing, with the goal of returning to golf outings with her husband. For this skill, Mary began to overcome her avoidance of activities that involved physical exertion by recognizing that she could still participate in some of her hobbies as long as she took planned, intermittent breaks. She modified the golf game by playing fewer holes and bringing a folding chair for resting between turns. Such accommodation and pacing allowed her to continue engaging in meaningful life events while also conserving energy.

Mary responded well to CBT during the two-month study period, as noted in Table 1. Specifically, per an independent assessor, her anxiety decreased from the moderate to the mild range (CGI=3), with a corresponding 11-point reduction on the HAM-A from baseline to post-treatment. Though still symptomatic, her overall quality of life also improved somewhat, as evidenced by a five-point gain on the FACT-L. In a qualitative exit interview, Mary noted that the breathing and relaxation strategies were consistent with techniques she had used previously and were therefore accessible and beneficial. She also reported that the cognitive strategies and activity scheduling/time-based pacing raised her awareness of pessimistic thoughts and tendency to overexert herself as well as achieve greater balance in her perceptions and activity level.

Case Study 2

History of Presenting Problem

“David” was a 56-year-old, married Caucasian man who presented for evaluation in the study after experiencing heart palpitations, shortness of breath, and chest tightness when climbing stairs at his home. He had been diagnosed with incurable lung cancer approximately 18 months earlier, and for the most part, had experienced few symptoms related to his disease. He maintained full-time work as an owner of a small business and physically exercised six days per week, including lifting weights. Despite the gravity of his medical diagnosis, he appeared healthy with minimal impairment in functioning. During the baseline evaluation, he endorsed mild anxiety symptoms on the HAM-A. He was anxious and worried that recent somatic symptoms, such as heart palpitations and fatigue, reflected a change in his physical status, namely that his current chemotherapy regimen was no longer effective.

Case Conceptualization

David’s anxiety was fairly circumscribed and primarily related to his interpretations of somatic symptoms. Although the severity of distress was mild overall, recurrent cancer-related triggers, such ambiguous physical symptoms, medical appointments and follow-up imaging studies, would nonetheless heighten anxious preoccupation with his disease. During these times, he tended to worry excessively, become irritable, and isolate from loved ones. He otherwise functioned at a high level, remaining consistently adherent to medical treatment recommendations and engaged in work activities and his daily exercise regimen.

Treatment Course and Outcome

Over the duration of the CBT sessions, David appeared to achieve the greatest benefit from psychoeducation regarding the nature of the stress response and processes of sympathetic arousal along with a review of cognitive restructuring skills. A self-proclaimed “hot head,” David recognized that at times he would jump to conclusions about the status of his cancer and makes catastrophic predictions. Such thoughts were particularly intrusive in the days before receiving the results of medical scans. As part of the Module 3 (Coping with Cancer Fears), David learned to practice techniques to help him remain focused on the present moment rather than a future-oriented apprehensive state. Specifically, he found much solace in maintaining a regular schedule of exercise in the mornings, which helped to clear his head as he prepared for the day, as well as taking his dog for walks and going fishing with friends. Fortunately, subsequent medical scans revealed that his disease was stable, with no evidence of progression. These results were pivotal in helping David understand that unexpected somatic symptoms in the context of cancer treatment cannot be accurately self-assessed or diagnosed without proper medical examination. As he monitored his daily automatic thoughts, he began to realize that stressors at work had been mounting for several weeks since losing one of his employees. He described how he had been working extra hours and pushing himself physically, which may have accounted for the recent heightened physiological arousal. These alternative explanations served as powerful rational responses to his cancer-related fears.

As part of a qualitative exit interview, David noted that each of the core modules enhanced his ability to manage stress more effectively by increasing his awareness to pause and think before reacting. He reported feeling generally more calm and improved communication with family and friends. Upon termination, David’s anxiety symptoms had abated considerably. As noted in Table 1, his HAM-A scores decreased by 11 points to the remission range, and his Clinical Global Impression for anxiety was a designated a 1 (not at all ill) by an independent assessor at the post-treatment evaluation. Finally, although he reported fairly good quality of life at baseline, his FACT-L scores also improved by 10 points.

Case Study 3

History of Presenting Problem

“Ellen,” a 61-year-old, divorced mother of three and retired paralegal, was referred to the study by her psychiatrist for treatment of generalized anxiety worsened by the diagnosis of terminal non-small cell lung cancer approximately six months earlier. She reported a long history of anxiety and dysthymia that had been previously treated with psychotropic medication and counseling, though she continued to experience residual symptoms. During the baseline evaluation, she reported on the HAM-A anxious mood; excessive worry about her health, finances and family; muscle tension, aches and pains; difficulty concentrating; stomach discomfort; and frequent headaches. Her anxiety symptoms were sufficiently severe to warrant a CGI rating of 4 (moderately ill). In addition, the patient’s cancer had metastasized to her tibia (lower leg) as well as her back, causing significant pain that compromised her quality of life and impaired her ability to complete household tasks among other responsibilities, as evidenced by the low FACT-L score (Total = 63) at baseline.

Case Conceptualization

Unlike the first two case examples, Ellen’s anxiety pertained less to fears of disease progression, but rather to stressors associated with changes in functioning. Though she prided herself on being an independent, self-sufficient woman, she now found herself struggling to complete what she described as simple activities, such as organizing her home, going out with friends, or taking walks through her neighborhood. In response to these losses, she experienced a mix of worry and sadness about how she was going to care of herself, especially because she couldn’t imagine requesting help from others. She was becoming increasingly sedentary, which only seemed to compound her pain and related anxiety symptoms.

Treatment Course and Outcome

Ellen’s goals for therapy were to enhance her quality of life by learning to cope with the physical limitations secondary to cancer and related pain symptoms. To that end, she observed considerable benefit in applying the breathing techniques in her daily life, especially the use of pursed-lips breathing when dyspneic upon exertion.

Ellen’s thought records revealed that she experienced much anxiety and worry about her ability to finish household tasks. Automatic thoughts included concern about how her children and friends would perceive her, given that they were accustomed to the patient maintaining a beautiful and comfortable home. Using cognitive restructuring techniques, we discussed her tendency to engage in cognitive errors, such as “mind reading,” and ways to challenge this distortion examining evidence that substantiate alternative interpretations. Additionally, Ellen began to adjust her priorities, noting that she would rather spend quality time with her daughters and grandchildren instead of cleaning the house, given her limited life expectancy. At the same time, Ellen did not simply want to ignore her responsibilities, so we used the activity planning and pacing module to help her discern how to alternate physically demanding tasks with periods of rest. She would judiciously choose when and where she wanted to expend energy on activities ranging from cleaning the kitchen to taking a short walk, visiting with a friend or volunteering to read books to children at the local preschool. She would punctuate these activities with 15–20 minutes of quiet relaxation exercises. This approach allowed her to gain a sense of self-efficacy, to value her current priorities, and to live more fully in the present moment.

As detailed in Table 1, Ellen’s HAM-A, BAI, and CGI scores remained unchanged from baseline to post evaluation, with the independent assessor rating her anxiety still in the moderate range (CGI = 4) after six sessions of CBT. Although she expressed appreciation for the different study interventions in a qualitative exit interview, she seemed to reap the most benefit from the activity planning and pacing module, which she suggested may have been more effective had we started with those skills. Also, despite persistent decrements in functioning at post assessment, her FACT-L scores increased by 19 points, the greatest improvement of all three participants.

Conclusions and Future Directions for Research

Many individuals with cancer experience significant symptoms of anxiety that are associated with poor coping and worse psychosocial and medical outcomes. Empirically-validated cognitive-behavioral therapy for anxiety has generally excluded individuals with medical comorbidities, thus diminishing its impact on addressing the clinically-relevant problems of patients with advanced cancer, such as fear of dying, management of toxic medical treatments, and consequent functional impairments. Moreover, evidence to date on the utility of psychosocial interventions to treat anxiety comorbid with cancer has been mixed, with notable flaws in the design of studies, focus on non-distressed patients, and limited accessibility of interventions to patients in need (Coyne, Lepore, & Palmer, 2006).

In developing and testing cognitive-behavioral interventions for anxiety, researchers have presupposed that negative thought processes reflect distortions in which individuals overestimate the likelihood of negative outcomes and underestimate their ability to cope (Beck & Emory, with Greenberg, 1985; Barlow, 2002). Therapists cannot necessarily make such assumptions when working with medically-ill patients who may have significant disability and potentially progressive disease. As highlighted in the three case examples, realistic concerns about medical symptoms such as breathlessness, fatigue, pain, and loss of functioning as well as worries about shortened life span understandably exacerbate anxious preoccupation. Incorporating techniques for managing psychic and somatic symptoms, ranging from relaxation strategies to mindfulness and activity pacing, our tailored treatment approach helped patients like Mary gain a sense of personal control and improve quality of life in the face of an uncertain future and unpredictable disease course. Also, such an approach offers clinicians concrete tools for helping patients with terminal illness for whom cognitive restructuring or behavioral exposure may not be feasible.

Although the case examples describe three individuals diagnosed with incurable lung cancer, patients with advanced malignancies are nonetheless a heterogeneous group with variable prognoses, treatment regimens and side effects, performance statuses, and medical comorbidities (American Cancer Society, 2009). For example, individuals with metastatic lung cancer have a five-year survival rate of 3.5% (versus 23.3% for advanced breast cancer) and often experience high symptom burden from the time of diagnosis (Horner et al., 2009). Anticancer therapy may provide some degree of palliation for dyspnea, fatigue, and pain, but such improvements are temporary and insufficient for relieving suffering. Comprehensive care for advanced lung cancer therefore requires adjunctive interventions for these symptoms (Temel, Pirl, & Lynch, 2006). Of the three case examples, Ellen was the most physically ill and debilitated. Given her progressive decline in functioning, the mere stability of her symptoms over time was considered a good outcome, and the improvement in quality of life as measured by the FACT-L was remarkable. These findings also emphasize the importance of conducting RCTs with patients who have significant medical comorbidities, as psychosocial interventions may buffer against worsening symptoms rather than significantly improving outcomes compared to a control group.

To cope with stress and anxiety caused by changes in functioning from fatigue and other somatic symptoms, a variety of non-pharmacological interventions (e.g., exercise, stress management) may be beneficial for patients with cancer (Mustian et al., 2007). Researchers have also begun to test the efficacy of energy conservation and activity management in adult patients receiving cancer treatment (Barsevick et al., 2007). The goals of such activity planning and pacing techniques are to teach individuals how to prioritize their daily tasks, to identify their limits for energy expenditure, and to alternate between physically demanding and less strenuous activities. While these strategies seem likely to alleviate stress and optimize productivity, follow-up studies are needed to test whether such integrative cognitive-behavioral models are effective and to discern the optimal format, timing, dose, and delivery of intervention.

Given the limitations of previous research on psychosocial interventions for cancer, clinicians need to begin developing therapies that can be easily and flexibly disseminated in the medical setting to patients diagnosed with terminal illness and comorbid anxiety. The current phase of our research involves a pilot randomized control trial to test this CBT intervention for anxiety in patients with advanced cancer. We have learned firsthand the challenges of conducting research and facilitating psychotherapy with such medically-ill individuals. For example, though we initially planned only to sample individuals with metastatic lung cancer, we decided to broaden the eligibility criteria to respond to the myriad recruitment challenges in this patient population, such as rapidly declining health and poor performance status (Schofield et al., 2008). We have therefore enrolled patients with diverse metastatic cancers, including breast, prostate, colorectal, lung, and brain. The trial is still underway, but anecdotal evidence from qualitative exit interviews suggests that patients find the tailored CBT approach relevant and useful in addressing their specific concerns regardless of cancer type. Moreover, to enhance participant motivation and retention, we have made efforts to keep the protocol as brief as possible (i.e., 6–7 sessions), to administer assessments over the phone when necessary, and to schedule CBT in tandem with other medical appointments, at times delivering treatment simultaneously with chemotherapy infusion, in order to reduce the burden of multiple office visits.

Despite the potential strengths of brief, tailored CBT in treating patients with advanced cancer, some qualifications regarding the therapeutic approach deserve mention. First, the treatment requires patients to be able to attend outpatient counseling sessions in an ambulatory clinic setting, limiting its utility for individuals in the last months of life who have become homebound due to progressive illness. Second, in the face of a late-stage cancer diagnosis, some patients may engage in cognitive and behavioral avoidance, a maladaptive strategy for managing thoughts and distress associated with the illness (Manne, Glassman, & Du Hamel (2000), which may need to become the target of therapy. Specifically, if the anxiety and any associated avoidance compromises adherence to medical treatment (e.g., refusing chemotherapy, biopsy, scans, etc.), this behavioral pattern ought to become the focus of attention and addressed in close collaboration with the oncology team. For patients with generally good health status and expected prognosis of at least one year, standard graded exposure therapy may be the treatment of choice (Craske, Antony & Barlow, 2006). However, for those experiencing significant morbidity, a pragmatic, problem-solving approach to resolving the anxiety may be necessary (Felgoise, Nezu, & Nezu, 2002). As described previously, in such cases, the therapist and patient should mutually define the problem that avoidance creates, brainstorm alternative behaviors, choose and optimal solution, and develop a realistic action plan. Such plans may involve interventions that are not typical components of behavioral exposure programs (such as adjunctive psychotropic medication, relaxation strategies), but nonetheless increase the likelihood that the patient will follow through with the steps of the action plan.

Lastly, recognizing the distress caused by anxiety in patients with cancer, the National Comprehensive Cancer Network (NCCN), a not-for-profit alliance of the leading cancer centers across the world, has published general guidelines for the evaluation and treatment of anxiety to improve the quality of care delivered to patients with cancer experiencing significant psychological distress. These guidelines outline various causes for anxiety and recommend psychotherapy along with close monitoring of response to treatment (NCCN, 2006). Hopefully, the NCNN’s acknowledgment of anxiety treatment as a salient component to comprehensive cancer care will spur development of novel therapies that are theoretically-driven and empirically-based, targeting the clinical features and concomitant psychosocial sequelae of advanced cancer and its treatment.

Table 2.

Baseline and Post-Assessment Results for Open Pilot Case Series

	Case Study 1: Mary		Case Study 2: David		Case Study 3: Ellen
	Baseline	Post	Baseline	Post	Baseline	Post
HAM-A	25	14	14	3	28	28
BAI	19	15	8	2	21	23
FACT-L	66	71	119	129	63	82
CGI	4	3	3	1	4	4