Illness Understanding and Advance Care Planning in Patients with Advanced Lymphoma

Abstract

Background: The prognosis of an aggressive lymphoma can change dramatically following failure of first-line treatment. This sudden shift is challenging for the promotion of illness understanding and advance care planning (ACP). Yet, little is known about illness understanding and ACP in patients with aggressive lymphomas.

Objective: To examine illness understanding, rates of engagement in ACP, and reasons for lack of ACP engagement in patients with advanced B cell lymphomas.

Design: Cross-sectional observational study.

Setting/Subjects: Patients (n = 27) with aggressive B cell lymphomas that relapsed after first- or second-line treatment treated at a single urban academic medical center.

Measurements: Participants were administered structured surveys by trained staff to obtain self-report measures of illness understanding (i.e., aggressiveness, terminality, curability) and ACP (i.e., discussions of care preferences, completion of advance directives).

Results: The majority of patients reported discussing curability (92.6%), prognosis (77.8%), and treatment goals (88.9%) with their medical team. Yet, less than one-third of patients reported being terminally ill (29.6%) and having incurable disease (22.2%). Most patients had a health care proxy (81.5%) and had decided about do-not-resuscitate status (63%), but the majority had not completed a living will (65.4%) or discussed their care preferences with others (55.6%).

Conclusions: The accuracy of lymphoma patients' illness understanding following first-line treatment is difficult to determine due to the potential for cure following transplant. However, this study suggests that a large proportion of patients with advanced B cell lymphomas may underestimate the severity of their illness, despite discussing illness severity with their medical team. Providing patients with information on prognosis, and the ACP process may increase engagement in ACP.

Introduction

When patients are diagnosed with aggressive lymphomas, the goal of initial therapy is typically to cure the disease. However, up to 50% of patients develop disease that is refractory to first-line treatment¹ and will relapse within a year² with an overall survival of four to six months.^3–5 Only 20% of patients with refractory disease are alive two years later.⁴ As with many cancers, some patients with relapsed lymphoma will be cured. However, the likelihood of a cure is significantly reduced following disease progression after first-line treatment. Therefore, for many patients, while the initial treatment goal is cure, prognosis suddenly and dramatically worsens following disease progression.

Research in solid tumor patients indicates that most patients underestimate the severity of their illness^6–10 and less than half recognize they are terminally ill.^11,12 Patients who understand that their prognosis are more likely to complete advance directives^13,14 and prefer comfort care over aggressive care,¹⁵ are less likely to receive futile aggressive care at the end of life (EOL),^16,17 and are more likely to die in their preferred place of death (i.e., home).¹⁷

Less is known about illness understanding and completion of advance directives in patients with hematological malignancies. A recent study of patients with acute myeloid leukemia found that over half had an inaccurate understanding of the likelihood they would be cured, based on oncologists' ratings of curability.¹⁸ Rates of advance care planning (ACP) in patients with hematologic malignancies vary across studies. Up to 44% of patients with hematologic malignancies have documentation of advance directives in the medical record.^19–21 In another study of patients with mixed hematologic malignancies, 37% of patients with lymphoma had a health care proxy (HCP) and living will, a lower rate than patients with other hematologic malignancies.²² In a small sample of hematopoietic cell transplant survivors (n = 18), 50% had a living will and 72% had a formal HCP.²³

The abrupt and dramatic shift in prognosis in patients with advanced aggressive lymphomas is a little studied illness trajectory that may make understanding the illness and ACP particularly difficult for these patients. This study examined rates of illness understanding, engagement in ACP, and reasons for lack of engagement in ACP in patients with advanced lymphoma.

Materials and Methods

Participants and procedures

Participants were recruited from June 2016 to March 2018 from a single academic medical center in an urban setting. Approval was obtained from the Institutional Review Board and all participants provided informed consent. Participants received $20 for completing the baseline survey. Eligible participants were English-speaking and 21 years of age or older with a diagnosis of diffuse large B cell lymphoma (DLBCL), follicular lymphoma transformed to DLBCL, double/triple hit lymphoma, Burkitt Lymphoma (BL), or aggressive B cell lymphoma intermediate between BL and DLBCL that relapsed after first- or second-line treatment. These histologies were selected to enhance generalizability of study findings as they represent the most common aggressive lymphomas. Patients were excluded if they had a diagnosis of dementia, delirium, and/or endorsed active suicidal ideation on study screening measures. Participants were administered baseline study measures over the telephone by trained research staff.

Measures

Advance care planning

Patients were asked “Do you have a healthcare proxy (HCP) or durable power of attorney (DPOA)?” Those who reported having a HCP/DPOA indicated their relationship to their HCP/DPOA. Patients were also asked, “Have you signed a living will?” and “Have you completed a do-not-resuscitate order (DNR)?” Finally, patients were asked, “Have you written down or talked to someone (e.g., your doctor or a family member) about the care you would like to receive should you become critically ill?” and, if so, with whom? Finally, patients were asked whether they discussed curability, prognosis, and treatment goals (yes/no) with the medical team.

Illness understanding

Understanding of disease aggressiveness was assessed with the question, “Lymphomas can be considered aggressive (fast growing) or indolent (slow growing). Which type of lymphoma do you have?” Patient's understanding of the terminality of their illness was assessed with the question, “How would you describe your current health status?” Response options included the following: (1) Relatively healthy, (2) Relatively healthy and terminally ill, (3) Seriously ill but not terminally ill, and (4) Seriously and terminally ill.^24,25

Finally, patients' understanding of the curability of their cancer was assessed by asking, “Which of the following best represents your understanding of your cancer?” Response options included the following: (1) My cancer will be cured, (2) My cancer may be cured if my current treatment is successful, (3) My cancer cannot be cured, but we will try to control the cancer with treatment, (4) My cancer cannot be cured and I am not able to have any further treatment, and (5) I don't know.

Statistical analyses

Frequency and descriptive statistics were used to examine sample characteristics, rates of ACP, and the relationship between illness understanding and ACP.

Results

Sample characteristics

The sample had an average age of 63.85 (standard deviation = 16.09, Table 1) and was primarily white (n = 24, 88.9%) and non-Latino (n = 27, 100%) with a diagnosis of DLBCL (n = 20, 74.1%). Approximately half of the samples were female (n = 14, 51.9%).

Table 1.

Sample Characteristics (n = 27)

	Total samples
Age, years (M, SD)	63.85 (16.09)
30–39 years	4 (14.8)
40–49 years	1 (3.7)
50–59 years	4 (14.8)
60–69 years	6 (22.2)
70–79 years	9 (33.3)
80–89 years	3 (11.1)
Gender (n, %)
Female	14 (51.9)
Male	13 (48.1)
Ethnicity
Latino	0 (0)
Non-Latino	27 (100)
Race (n, %)
White	24 (88.9)
Other	3 (11.1)
Education (n, %)
Grade school	1 (3.7)
High school	1 (3.7)
Some college/college degree	11 (40.7)
Postgraduate	14 (51.9)
Marital status (n, %)
Married/partnered	14 (51.9)
Single	6 (22.2)
Divorced	6 (22.2)
Other	1 (3.7)
Annual income (n, %)
Less than $21,000	1 (3.7)
$21,000–39,999	3 (11.1)
$40,000–65,999	7 (25.9)
$66,000–105,999	2 (7.4)
$106,000+	13 (48.1)
Don't know	1 (3.7)
Religious affiliation (n, %)a
Jewish	14 (53.8)
Catholic	5 (19.2)
Atheist	1 (3.8)
Other	2 (7.7)
None	4 (15.4)
Primary cancer diagnosis
DLBCL	20 (74.1)
Double/triple hit lymphoma	2 (7.4)
Follicular lymphoma transformed to DLBCL	5 (18.5)
Years since diagnosis (M, SD)	3.79 (4.64)
Treatment lines received (n, %)
One	3 (11.1)
Two	17 (63.0)
Three or more	7 (25.9)

^aMissing: n = 1.

DLBCL, diffuse large B cell lymphoma; SD, standard deviation.

Illness understanding

Two-thirds of patients (n = 18, 66.7%; Table 2) described their lymphoma as aggressive or fast-growing. Less than one-third (n = 8, 29.6%) of patients reported being terminally ill and one-fifth (n = 6, 22.2%) described their illness as incurable.

Table 2.

Rates of Illness Understanding (n = 27)

	n (%)
Aggressive or indolent lymphoma
Aggressive (fast growing)	18 (66.7)
Indolent (slow growing)	3 (11.1)
I don't know	5 (18.5)
Refused	1 (3.7)
Current health status
Relatively healthy	12 (44.4)
Relatively healthy and terminally ill	8 (29.6)
Seriously ill but not terminally ill	7 (25.9)
Seriously and terminally ill	0 (0)
Curability of cancer
My cancer will be cured	4 (14.8)
My cancer may be cured if my current treatment is successful	16 (59.3)
My cancer cannot be cured but we will try to control the cancer with treatment	6 (22.2)
My cancer cannot be cured and I am not able to have any further treatment	0 (0)
I don't know	1 (3.7)

Advance care planning

The majority of the samples had designated a HCP (n = 22, 81.5%; Table 3). Reasons for not designating a HCP included that patients were considering who to choose (n = 2, 7.4%), had not thought about it (n = 2, 7.4%), and did not know the definition of a HCP (n = 1, 3.7%). Most of the samples (n = 17, 65.4%) had not completed a living will. Of note, 14.5% (n = 4) of the samples did not know the definition of a living will and almost half of patients who completed a living will (n = 4, 44.4%) did not know the care preference indicated in their living will.

Table 3.

Rates of Advance Care Planning (n = 27)

	n (%)
HCP/durable power of attorney
Yes	22 (81.5)
No: I am still considering who to choose	2 (7.4)
No: I haven't thought about it	2 (7.4)
I don't know what a HCP/DPOA is	1 (3.7)
Relationship of HCPa
Spouse/partner	11 (50.0)
Son/daughter	7 (31.8)
Parent	2 (9.1)
Clergy member	1 (3.7)
Other	1 (3.7)
Living will
Yes	9 (33.3)
No: I don't need a living will now	4 (14.8)
No: I am still considering whether I want one	3 (11.1)
I don't know what a living will is	4 (14.8)
I don't know	1 (3.7)
Other	6 (22.2)
Content of living willb
Comfort measures only	2 (22.2)
Limited medical interventions	2 (22.2)
No limitations on medical interventions	1 (11.1)
I don't know	4 (44.4)
DNR order
Yes: DNR order completed	14 (51.9)
No: I want resuscitation to be attempted	3 (11.1)
No: I am still considering whether I want one	1 (3.7)
No: I haven't thought about it	3 (11.1)
I don't know what a DNR is	3 (11.1)
Other	3 (11.1)
Discussed care preferences
Yes	12 (44.4)
No	13 (48.1)
I don't know	2 (7.4)
Recipient of care preferences discussionc
Oncologist	3 (11.1)
Social worker	1 (3.7)
Spouse/partner	7 (25.9)
Son/daughter	8 (29.6)
Other family member	2 (7.4)
Chaplain/clergy	2 (7.4)
I wrote it down	2 (7.4)
Other	6 (22.2)
Discussed curability with medical team
Yes	25 (92.6)
No	0 (0)
I don't know	2 (7.4)
Discussed prognosis with medical team
Yes	21 (77.8)
No	3 (11.1)
I don't know	3 (11.1)
Discussed treatment goals with medical team
Yes	24 (88.9)
No	2 (7.4)
Missing	1 (3.7)

^aPercentages based on total n = 22.

^bPercentages based on total n = 9.

^cPatients were permitted to choose more than one option.

DNR, do-not-resuscitate; DPOA, durable power of attorney; HCP, health care proxy.

Over one-third of patients (n = 10, 37%) had not made a decision about their DNR status and 11.1% (n = 3) did not have a DNR order because they wanted resuscitation to be attempted. Reasons for not making a decision about a DNR order included that patients did not know the definition of a DNR order (n = 3, 11.1%), had not thought about it (n = 3, 11.1%), were considering whether they wanted a DNR order (n = 1, 3.7%), and other (n = 3, 11.1%).

Less than half of the samples (n = 12, 44.4%) reported writing down or talking to someone about the care they would like to receive should they become critically ill, and only 11.1% (n = 3) of these patients discussed their preferences with a health care provider. However, most of the samples reported discussing curability (n = 25, 92.6%), prognosis (n = 21, 77.8%), and treatment goals (n = 24, 88.9%) with the medical team.

Discussion

Due to the nature of B cell lymphomas, determining prognosis and the outcome of additional treatments following relapse after first-line treatment is difficult. However, this study suggests that a large proportion of patients with recurrent aggressive lymphoma may underestimate the severity of their illness, even though most of the samples reported discussing these issues with their medical team. Further, despite experiencing disease progression after first-line treatment, most patients did not comprehensively document their care preferences.

The majority of the samples had identified a HCP and made a decision regarding their DNR status. However, less than half of the samples had discussed their care preferences with someone and only one-third had completed a living will. Patients may not associate HCPs with diagnosis of a terminal illness or may trust a designated HCP to make decisions consistent with their preferences,²⁶ making it easier to name a HCP than to complete other advance directives. However, an evaluation of next-of-kin surrogates in the Health and Retirement Study found that living will creation and discussion of EOL care preferences were associated with EOL care decisions while designation of a HCP was not.²⁷ These finding may suggest that HCPs benefit from the guidance provided by other advance directives.²⁸ The finding that few patients communicated about their care preferences with a health care provider is problematic given prior research suggesting that patients who discuss their EOL care preferences with a physician are more likely to receive care consistent with their preferences.²⁹

The current study sheds light on why patients may not complete advance directives. For example, ∼10% of patients had not completed a DNR order because they preferred resuscitation. These patients had considered their treatment preferences, similar to patients who completed a DNR order. Prior studies would have grouped them with patients who had not yet considered their preferences.³⁰ Inadequate knowledge of advance directives was a consistent reason for failure to complete advance directives, suggesting that patients may benefit from basic information on the definition and purpose of advance directives. Furthermore, the proportion of patients who did not know the contents of their own living will point to potential gaps in the process of completing a living will.

The primary study limitation is the small sample size that precludes examination of statistical significance and limits the generalizability of study findings. Information on patients' ultimate disease course was not available during the study timeline and oncologists' views of the patient's prognosis and the goals of treatment were not assessed. As a result, we are unable to determine the accuracy of patients' illness understanding. Indeed, a small minority of patients with chemorefractory aggressive lymphomas can achieve satisfying long-term outcomes. However, the overwhelming preponderance of data in this population suggests that the vast majority have terminal and noncurable disease.^2–5,31 The discord between probable outcome (i.e., death within four to six months) and possible outcome (20% survival at two years), the ability of physicians to convey that message without extinguishing hope, and the ability of patients and caregivers to absorb those nuances are likely the core causes of poor use of palliative care services and the often traumatic transition to EOL care in lymphoma patients. Future studies that conduct in-depth longitudinal assessment of patients' illness understanding, comparison of patients' and oncologists' views of the illness, and assessment of survival rates and care received through death will allow for an examination of illness understanding, ACP, and EOL care that considers the unique nuances of lymphoma.

Due to the sharp transition from likely curable at diagnosis to unlikely curable at relapse, patients may benefit from repeated discussions of prognosis to ensure awareness of changes in disease severity over time. Furthermore, these patients may benefit from information on the definitions and purpose of various advance directives, including the relevance of these documents for patients with advanced lymphoma.

Research suggests that the integration of palliative care into medical oncology improves illness understanding and patient outcomes.⁹ Multiple models for the provision of palliative care can be effective, including primary palliative care provided by the treating oncologist or specialist palliative care provided by a palliative care team.³² Greater integration of palliative care into hematologic oncology may help lymphoma patients and their families understand the complexities of lymphoma and articulate their treatment preferences.^33,34 Future research that examines the longitudinal relationship between illness understanding and ACP and EOL care received and the impact of palliative care on these relationships will inform strategies for increasing engagement in ACP to improve EOL care in this population.

Funding Information

This research was funded by the Weill Cornell Medicine Lymphoma Program (Dr. J.P.L.), the National Institute on Aging and American Federation for Aging Research (K23 AG048632, K.M.T.) and the National Cancer Institute (R35CA197730, H.G.P.). This work was supported by a grant from Ben and Ayalet Blaustein in memory of Henry Blaustein.

Author Disclosure Statement

Dr. P.M. serves as a consultant to AstraZeneca, Gilead, Janssen, Bayer, Seattle Genetics, and Kite; Dr. S.C.R. serves as a consultant for AstraZeneca, Celgene, Heron, Juno Therapeutics, Karyopharm, and Seattle Genetics; Dr. J.P.L. serves as a consultant to Sutro, Bayer, MEI Pharma, Gilead, AstraZeneca, Novartis, Celgene, Biotest, Merck, Morphosys, Beigene, Nordic Nanovector, Roche/Genentech, ADC Therapeutics; No other authors (K.M.T., C.M., D.J.O., and H.G.P.) have conflicts to disclose.