Anxiety Shapes Expectations of Therapeutic Benefit in Phase I Trials for Patients With Advanced Cancer and Spousal Caregivers

PURPOSE:

Advanced cancer patients (ACP) hope to receive significant therapeutic benefit from phase I trials despite terminal disease and presumed symptom burdens. We examined associations between symptom burdens and expectations of therapeutic benefit for ACP and spousal caregivers (SC) during phase I trials.

PATIENTS AND METHODS:

A prospective cohort of ACP-SC enrolled in phase I trials was assessed at baseline and one month using symptom burden measures evaluating depression, state-trait anxiety, quality of life, global health, post-traumatic coping, and marital adjustment. Interviews evaluated expectations of benefit.

RESULTS:

Fifty-two phase I ACP and 52 SC (N = 104) were separately assessed and interviewed at baseline and one month. Total population demographics included the following: median age 61 years (28-78), 50% male, 100% married, 90% White, and 46% ≥ college education. At T1, ACP reported symptoms of mild state anxiety, mild trait anxiety, poor global health, and quality of life. SC reported moderate state and mild trait anxiety and good global health with little disability at baseline. State anxiety was a significant predictor of ACP expectations for phase I producing the following therapeutic benefits: stabilization (P = .01), shrinkage (P < .01), and remission (P = .04). Regression analyses also revealed negative associations between SC expectation for stabilization and SC anxiety: state (P = .01) and trait (P = .02). ACP quality of life was also negatively associated with SC expectations for stabilization (P = .02) and shrinkage (P = .01).

CONCLUSION:

Anxiety, both state and trait, impacts couples’ beliefs regarding the likelihood of therapeutic benefit from phase I trial participation.

INTRODUCTION

In 2020, a total of 606,520 million Americans are expected to die from cancer, the second leading cause of death in the United States.¹ These deaths occur with significant anticipation, with expected median survival of 9-12 months for terminally ill, advanced cancer patients (ACP).^1,2 Given this survival and elevated symptom burden,^3-6 oncologist-patient communication regarding this terminal prognosis prior to ACP death should enable an informed decision for appropriate end-of-life care (eg, hospice). Yet, some ACP, as a means to sustain hope, continue efforts with experimental chemotherapy, ie, phase I trials.^7-19 As described in the Position Statement on phase I trials by ASCO,²⁰ evidence indicates that phase I trials have the potential to deliver direct medical and quality-of-life benefits; however, they are primarily designed to determine the recommended dosage and safety of the agent, rather than establish efficacy. Phase I trials may have therapeutic intent, yet the likelihood for long-term benefits, such as remission, remains limited. Hence, ethical concerns arise about these ACP who, despite a rigorous informed consent process, frequently overestimate the probability of a positive therapeutic response.^7-20 Evidence has confirmed that ACP possess inadequate, measured understanding of phase I informed consent including the following: (1) research purpose (dose determination), (2) likelihood of therapeutic benefit, and (3) trial alternatives (eg, hospice).^7-17 ACP motivations for clinical benefit and oncologist communication also further contribute to this misunderstanding.^9-20 However, the potential impact that physical and psychological symptom burdens have on ACP ability to consent for trial participation remains unknown.

Phase I ACP gain significant psychological benefits from participation.¹² This treatment-related optimism, although unrealistic, serves as a protective mechanism against the development of psychological distress, providing hope despite a dismal prognosis.^12,13 To date, research has failed to explore this unique relationship between treatment-related optimism and psychological adjustment.¹⁷ Rather, study has centered on identifying patient characteristics correlated with unrealistic expectations of benefit such as comprehension, knowledge, and risk-seeking.^15-17 Despite the fact that phase I ACP report greater symptom severity (pain, fatigue, and distress) than other ACP populations,^3-5 and evidence revealing that functional status is correlated with expectations of benefit,¹¹ critical longitudinal symptom assessment, including potential associations with ACP expectations of benefit, has not been conducted. Presently, a single-event, individual instrument has been used to evaluate multiple symptom burdens for phase I ACP, challenging the accuracy of symptom assessment.^3-5

Additional concerns surround the symptom burdens that spousal caregivers (SC) experience during the advanced cancer trajectory.^21-25 Approximately 38% SC report symptoms meeting criteria for psychiatric disorders, ie, anxiety and depression.^24,25 SC report significant physical impairments, financial burdens, and poor social support, thereby accessing medical and mental health services more frequently than other caregiver counterparts.^21-25 SC aid in patient treatment decision making and at the end of life assume the role of proxy decision maker.^21,24,25 As caregivers, SC make stressful familial and occupational adjustments threatening their ability to cope resulting in marital distress.^21-25 Yet, the symptom burdens SC experience and their corresponding expectations of ACP benefit have not been identified.

Therefore, this original study investigated phase I ACP-SC symptom burdens defined as multiple, concurrent physical (pain, fatigue, and global health) and psychological (cognition, depression, anxiety, marital adjustment, and post-traumatic coping) symptom expression associated with phase I ACP cancer treatment and/or toxicity and SC caregiving. Potential associations between symptom burdens and expectations of therapeutic benefit for phase I ACP-SC were examined. We hypothesized the following: (1) symptoms will intensify over time; (2) ACP and SC will report significant symptom severity and high expectations of benefit in forms of stabilization, shrinkage, remission, and cure over time; (3) symptoms are negatively associated with ACP and SC expectations of benefit; (4) one partner’s symptoms are associated with the other partner’s expectations of benefit.

PATIENTS AND METHODS

Study Design

Participants were recruited from a Midwestern Comprehensive Cancer Center’s Developmental Therapeutics Clinic at The University of Chicago. This prospective cohort study was approved by the Institutional Review Board.

Participants

ACP eligible to participate in this study were as follows: (1) English-speaking, (2) able to give consent, (3) age ≥ 18 years, (4) prognosis of at least months, (5) Karnofsky performance status ≥ 60%, (6) documented diagnosis of advanced cancer proven refractory to standard therapy or for which no identifiable standard exists, and (7) consented within the past week for phase I trial participation. ACP self-identified SC who were spouses or domestic partners of ≥ 6-month duration.

Procedure

ACP and their SC underwent phase I trial discussions with four primary investigators who disclosed key consent elements: (1) purpose as dosage/toxicity, (2) likelihood of therapeutic benefit, (3) risks and benefits, and (4) trial alternatives. ACP-SC symptoms and expectations of benefit were assessed by a research assistant after ACP trial consent, one week prior to the receipt of investigational agents.^7,8,13,14,18 Both sets of participants independently completed multiple measures with interviews in the clinic at two time points: enrollment (T1: baseline) and one month (T2), each lasting 1 hour.

MATERIALS

Demographics

Sociodemographic and clinical information was recorded per ACP-SC self-report.

Psychological Measures

Cognition.

Folstein Mini-Mental State Examination (MMSE)²⁶ is a 30-item cognitive function measure testing five domains: orientation, registration, attention/calculation, recall, and language. Scores range from 0 to 30; scores < 24 reveal impairment.

Depression.

Center for Epidemiological Studies Depression Scale (CES-D)²⁷ is a 20-item affective depression measure. ACP rate items using a 4-point Likert scale (0 = only rarely/none of the time to 3 = most of the time). Scores range from 0 to 60; higher scores indicate greater depression.

Anxiety.

State-Trait Anxiety Inventory (STAI-S/T)²⁸ is a 40-item instrument with 20 items measuring state anxiety on a 5-point Likert scale (1 = not at all to 4 = very much so) and 20 measuring trait anxiety on a 5-point Likert scale (1 = almost never to 4 = almost always). Scores range from 20 to 80; higher scores reveal greater anxiety.

Post-traumatic coping.

Post-Traumatic Growth Inventory (PTGI)²⁹ is a 21-item scale measuring participant ability to find meaning after a traumatic cancer experience across five subscales (personal strength) on a 6-point Likert scale (0 = I did not experience this change to 5 = I experienced this change to a great degree). Scores range from 0 to 105; high scores indicate greater change.

Marital adjustment.

Dyadic Adjustment Scale (DAS)³⁰ is a 32-item measure assessing marital functioning, satisfaction, and distress. Responses were provided in multiple-choice formats. Scores range from 0 to 151; higher scores represent greater adjustment.

Physical Symptom Burdens

Global health.

Medical Outcomes Form-36 (SF-36)³¹ is a 36-item general health measure evaluating eight domains (eg, energy/vitality). Domains form two health concepts: physical and mental health. Scores range from 0 to 100; high scores indicate best health.

Quality of life.

Functional Assessment of Chronic Illness Therapy-Palliative Care (FACIT-Pal)³² is a 46-item quality-of-life (QoL) measure comprised of the FACT-General (FACT-G) with a palliative symptom subscale (Pal). ACP rate symptom expression in the past week on a 5-point Likert scale (0 = not at all to 4 = very much). FACIT-Pal yields FACT-G, Pal-subscale, and FACIT-Pal-Total scores, with low scores revealing poor QoL.

Structured Interview

ACP-SC expectations of benefit were obtained by interviews evaluating the understanding of the phase I informed consent process.^7,8,13,14,18 Quantitative inquiry explored beliefs regarding the likelihood of phase I producing anticancer benefit in forms of stabilization, shrinkage, remission, and cure. ACP-SC rated items using Likert scores 1-10: On a scale from 1 to 10, 1 = not likely to succeed to 10 = very likely to succeed, how successful do you think your current therapy will be in STABILIZING your cancer?

Statistical Analyses

Data were analyzed using statistical software (STATA).³³ A priori power analysis was completed. Descriptive statistics are reported as proportions, means, standard deviations, and frequencies. Repeated measures analysis of variance (ANOVA) evaluated symptom changes over time comparing groups. Paired t-tests determined ACP-SC differences for expectation of benefit variables over time. Two regression analyses examined the following: (1) predictability of expectations from overall symptom assessment at T2 and (2) one partner’s symptom association with other partner’s expectations. Regressions were separately completed for ACP and SC. For analyses, T2 expectation of benefit was the criterion variable.

RESULTS

Recruitment

A convenience sample of 154 ACP was approached for study participation. Of these, 61 ACP were deemed ineligible (eg, three did not speak English) and 17 ACP refused participation. Seventy-six couples provided study consent with 24 couples discontinuing phase I participation at both time- points because of cancer progression. Final sample consisted of 33.7% ACP approached and 68.4% ACP consented.

Sample Characteristics

Total sample consisted of 104 participants (52 ACP and 52 SC) consented, completing measures with interviews at T1 and T2 (Table 1). The total population demographics include the following: median age 61 years (28-78), 50% male, 100% married, 90% White, and 46% ≥ college education. All, 52 couples were married (mean = 37 years; 15-50 years), 10 ACP were in a second marriage, 50 ACP were diagnosed 2-3 years earlier, and ACP median survival was 7.2 months (range, 0.41-18.2). For the population as a whole, 53% had an income < $65,000 per year in US dollars, with 45% ACP employed full-time or part-time and 55% retired. For SC, 66% were employed full-time or part-time with 34% retired.

TABLE 1.

Demographics

Descriptive Statistics

Tables 2 and 3 present phase I ACP-SC symptoms and expectations of benefit over time. ACP reported mild state and trait anxiety and high marital quality at T1 and T2. They reported poor global health and QoL. ACP believed phase I had a high likelihood of producing cancer stabilization, shrinkage, remission, and cure. Additional ACP symptoms did not change significantly over trial: depression, marital adjustment, FACIT-Pal-Total, global health, cognition, PTG, state anxiety, and trait anxiety (Table 2).

TABLE 2.

SB Repeated Measures ANOVA

TABLE 3.

Paired ACP-SC Differences for Expectations of Benefit

SC reported moderate state anxiety, mild trait anxiety, high global health, and high marital quality at T1. For SC, phase I had a high likelihood of producing cancer stabilization, shrinkage, remission, and cure. Repeated measures ANOVA revealed that only two SC symptoms significantly improved at T2: marital adjustment F(1,51) = 4.42, P = .04 and trait anxiety F(1,51) = 3.82, P = .03. No significant differences were found between ACP-SC expectations at T1 or T2 (Table 3).

Anxiety Shapes Expectations of Benefit

State anxiety was a significant predictor of ACP expectations for producing cancer stabilization, shrinkage, and remission at T2 (Table 4). A negative association existed between ACP expectation for stabilization and ACP state anxiety F(1,51) = 6.52, P = .01. ACP with mild state anxiety reported higher expectations for phase I stabilization. As state anxiety increased, ACP expectation for stabilization decreased. Similarly, state anxiety was negatively associated with ACP expectations for shrinkage, F(1,51) = 8.44, P < .01; and remission, F(1,51) = 4.30, P = .04. No additional associations existed between other ACP symptoms (eg, depression, and pain) and their expectations.

TABLE 4.

Regression Analyses of T2 Expectation of Benefit and SB for ACP-SC

At T2, several associations were identified between SC symptoms (state-trait anxiety, marital adjustment, and global health) and their expectations of benefit (Table 3). SC expectation for stabilization was negatively associated with SC state anxiety F(1,51) = 6.52, P = .01; and trait anxiety F(1,51) = 5.46, P = .02. As SC state and trait anxiety increased, SC expectations for stabilization decreased. State anxiety was a significant predictor of their expectations for benefit in the form of shrinkage: F(1,51) = 3.02, P < .01. High SC marital adjustment was associated with elevated expectations for stabilization F(1,51) = 4.14, P = .04; shrinkage F(1,51) = 4.46, P = .04; remission F(1,51) = 4.39, P = .04; and cure F(1,51) = 4.22, P = .03. Finally, SC global health was positively associated with beliefs for shrinkage, F(1,51) = 3.97, P = .03. SC with heightened physical global health symptoms tiredness F(1,51) = 3.55, P = .04 and nervousness F(1,51) = 3.19, P = .02 reported high expectations for tumor shrinkage.

Associations Between ACP Symptoms and SC Expectations of Benefit

Regressions between ACP symptoms and SC expectations of trial benefit revealed significant associations at T2. ACP state anxiety was associated with SC expectation for cure, F(1,51) = 3.01; P = .03. ACP post-traumatic coping was negatively associated with SC shrinkage beliefs, F(1,51) = 4.46; P < .01. ACP global health was a negative significant predictor of T2 SC expectations for shrinkage F(1,51) = 3.97, P = .01 and stabilization F(1,51) = 6.74, P = .02. No statistically significant associations were identified between SC symptom burdens and ACP expectations at T2.

Why State Anxiety Is Associated With Trial Expectations: A Qualitative Analysis

ACP and SC qualitative statements depicting thematic relationships between anxiety and trial expectations reveal why state anxiety is associated with trial expectations. ACP-SC responses include the following: (1) PT: “The doc told me to get my will 1 year ago. I beat that but now in research. It worries me. Chances of the trial shrinking it is 50/50.” (2) PT: “It's promising and can't give up. It’s scary, a slow tick tock. Drs said you want to know how long. I said No. I’m fighting to live, plain scared. This trial will stabilize it, put it in remission.” (3) SC: “This is frightening—the likelihood that an experimental drug will work is slim. I try to be positive most of the time. I’m also very apprehensive, fearful.” ACP-SC themes associated with hope include the following: (4) PT: “I’m hoping for the best. It’ll stop the tumor growth but will be there. Need to live with the fear.” (5) SC: “I’m terrified, lonely, scared but have high hopes for this trial will benefit him.”

DISCUSSION

For over 2 decades, research overwhelmingly has documented ACP hope to receive significant therapeutic benefit from phase I trial participation despite the presence of terminal disease and presumed significant symptom burdens.^7-20 Nevertheless, data were lacking what specific symptoms impact both ACP and SC expectations of phase I trial benefit. Our study suggests that ACP and SC expectations of phase I trial benefit are influenced by not only physical but also psychological symptoms. Primarily, state-trait anxiety shapes couples’ beliefs about the likelihood of a therapeutic benefit for phase I trials.

To our knowledge, this is the first study to formally assess symptoms and expectations of benefit for ACP and SC in phase I oncology trials using multiple quantitative measures during the trial.¹⁷ Research to date has documented the unrealistic expectations of benefit of phase I trial ACP, however, only examining previously identified symptoms or patient characteristics at one time point.^7-20 Our study results are unique given the specific focus on and identification of elevated symptoms that shape phase I ACP-SC expectations over time.

Anxiety was identified as a significant predictor of ACP and SC expectations of benefit. ACP with mild state anxiety reported higher expectations for phase I to produce cancer stabilization, shrinkage, and remission. As ACP anxiety increased, their expectations for benefit decreased. Similarly, for SC, mild trait and state anxiety predicted higher expectations of ACP benefit. These results are striking in that prior research has only identified associations between mood and trial decision making.^12,34 Our data support evidence suggesting that ACP arousal, triggered by oncologist communication of serious news, may hinder cognitive processing of disease and treatment information.^35-37 Although we assessed ACP one week after the phase I discussion, it may be that ACP-SC anxiety levels were not sufficiently reduced to improve information recall (eg, comprehension of trial purpose), leading to unrealistic expectations of benefit.^35-37 Cognitive bias may be promoted by the way information is delivered by diverse group of investigators.³⁸ Accordingly, oncologist communication regarding phase I trials should be a supportive communication process, uniquely tailored to reduce ACP negative affect to enhance recall pertaining to benefit. Anxiety may arise due to a situational, systematic, analytical thought process resulting from elevated symptoms or due to decisional uncertainty of phase I trial participation.^35-37 Moreover, these results confirm prior critical research revealing that not only do ACP with higher anxiety need greater assistance understanding or accepting disease and scan information, but also the presence of anxiety in one partner (either patient or caregiver) is associated with a greater likelihood of anxiety disorders in the other partner.^39,40

Indeed, these data challenge current debate that distinguishes optimism and unrealistic optimism by refocusing attention on the role of anxiety that potentially is responsible for avoidance of negative information viewed as life-threatening.^15-17 Such emotional responses are not extraordinary, may be psychologically protective in the short term, and are necessary to maintain hope.^{12,14,16,35-37} However, future study on the complex relationship between anxiety and perception of information should be explored empirically. For SC, poor quality of life and marital adjustment were associated with higher expectations for shrinkage, stabilization, and cure. Such factors may affect SC measured responses to queries about the success of chemotherapy and when, or if, SC believe ACP will die. These data support prior work proposing symptom severity, especially anxiety, leads to poor coping, adversely affecting both individuals and dyad.^21-25,39

Additional analyses supported the hypothesis that ACP symptoms were associated with SC expectations of phase I benefit. Negative associations existed between ACP symptoms (PTG, QoL, and state anxiety) and SC expectations for cancer stabilization, shrinkage, and cure at T2. Theoretically speaking, SC expectations for a positive outcome may be in response to witnessing ACP decline, thus engaging in active avoidance of information to facilitate coping or a form of empathic anxiety.⁴⁰ In contrast, no additional associations were detected between SC symptoms and ACP expectations of benefit suggesting ACP are not fully aware of the significant anxiety SC experience during the cancer trajectory. Consequently, these data confirm evidence that ACP-SC were motivated by hope for the possibility of benefit from phase I participation.^7-20

Study Limitations

The questionnaires were conducted with a selected group of ambulatory, competent ACP with good functional status despite advanced disease. ACP were white, male, middle class, and actively participating in phase I clinical trials at a single Midwestern institution and not representative of all ACP. Self-reported symptoms may differ in ACP assessed prior to trial enrollment or ACP with disease progression during the trial. Longitudinal assessment would monitor changes in anxiety as ACP entered hospice. An early time point during phase I as ACP become symptomatic (eg, 2 weeks) may provide vital information regarding overall symptom incidence and severity. Despite adequate power, a larger sample would strengthen marginal statistical associations between variables. The participation rate suggests a potential lack of generalizability given the significant sample selection bias, which requires the need for confirmation in more representative cancer dyads. Finally, our results may not be representative of all couples with cancer considering end-of-life care; yet, these associations provide hypotheses for future study.

Clinical Implications

Oncologists should routinely assess ACP symptoms, including psychological distress, using patient-reported outcome measures to track symptom severity during phase I trials to improve the quality of cancer care. Psychological distress, especially anxiety, should be assessed in both ACP and their caregivers. The presence of ACP anxiety not only affects the decision making regarding end-of-life care but poses a warning sign that anxiety is also likely present in the caregiver, therefore warranting critical intervention. Consequently, the central focus of the oncology community should concentrate on delivery of early, simultaneous supportive, team-based palliative care concurrent with investigational therapy targeting both physical and psychological symptoms of ACP and SC.^41,42 ACP-SC preferences for supportive care including psychological support (eg, individual psychotherapy, couple-based cognitive behavioral therapy interventions, and mindfulness stress reduction) must be considered to address this anxiety over end-of-life fears and informational needs. Hence, the application of early, team-based, simultaneous supportive care, couple interventions, with a dyadic focus, should be incorporated as a fundamental, vital component in advanced cancer care at the end of life.

AUTHOR CONTRIBUTIONS

Conception and design: Fay J. Hlubocky, Tamara G. Sher, Christopher K. Daugherty

Administrative support: Fay J. Hlubocky, Christopher K. Daugherty

Collection and assembly of data: Fay J. Hlubocky, Christopher K. Daugherty

Data analysis and interpretation: Fay J. Hlubocky, David Cella, Kristen E. Wroblewski, Jeffery Peppercorn, Christopher K. Daugherty

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Anxiety Shapes Expectations of Therapeutic Benefit in Phase I Trials for Patients With Advanced Cancer and Spousal Caregivers

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/op/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).

David Cella

Stock and Other Ownership Interests: FACIT.org

Consulting or Advisory Role: Abbvie, GlaxoSmithKline, Pfizer, Astellas Pharma, Novartis, PledPharma, IDDI, Bristol-Myers Squibb, Asahi Kasei, Ipsen, Mei Pharma

Research Funding: Novartis, Genentech, Ipsen, Pfizer, Bayer, GlaxoSmithKline, PledPharma, Bristol-Myers Squibb, Abbvie, Regeneron, Clovis Oncology

Travel, Accommodations, Expenses: Ipsen, PledPharma

Jeffery Peppercorn

Employment: GlaxoSmithKline

Stock and Other Ownership Interests: GlaxoSmithKline

Consulting or Advisory Role: Athenex, Abbott Laboratories

Research Funding: Pfizer

Christopher K. Daugherty

Consulting or Advisory Role: Daiichi Sankyo

No other potential conflicts of interest were reported.