The Unique Burden of Rare Cancer Caregiving: Caregivers of Patients with Erdheim-Chester Disease

Abstract

Research examining the experience of informal caregivers (ICs) for patients with rare cancers is limited. This was a mixed-methods pilot study of 14 ICs for patients with Erdheim-Chester disease (ECD), an ultra-rare neoplasm. Participants were predominantly female and over half provided at least 60% of their loved one’s care. Participants completed measures of the impact of caregiving, caregiver burden, unmet needs, quality of life, anxiety, and depression. Participants reported substantial impact of caregiving, including limiting (50%) or discontinuing (21%) paid employment, and exhausting financial savings (43%). ICs reported a moderate level of burden with five (38%) reporting risk for burnout. While participants reported anxiety (64%) and depression (14%), their overall quality of life was favorable. Semi-structured interviews highlighted factors related to the distress and isolation of navigating a rare cancer. ECD ICs report burden and distress shaped by the experience of providing care for a patient with a rare cancer.

Introduction

Informal caregivers (ICs) of individuals with cancer provide uncompensated care that is physically, emotionally, and financially demanding [1]. ICs help patients navigate symptoms, medical encounters, and decisions about treatment, and often perform medical or nursing tasks without formal training [2]. There is increasing evidence of psychological morbidity experienced by cancer ICs, among whom rates of anxiety and depression are often higher than those found among patients [3–5]. Burden – a construct that refers to the broad negative impact of the caregiving role on the caregiver [6]– is well documented among cancer ICs [7,8]. Diminished well-being of ICs not only compromises the quality of care they provide to patients, but has also been found to be associated with depressed patient mood [9] and poorer patient health outcomes [10].

The hardship of providing care to someone with cancer is accentuated when the patient has a rare malignancy, posing challenges such as delayed diagnosis, a scarcity of treating specialists, and potentially limited therapeutic options [11,12]. Rare cancers, in aggregate, account for approximately 20% of all cancer diagnoses in Europe and the United States [13,14]. As such, the collective burden of rare cancer caregiving is substantial. However, there is limited empiric literature on the unique experience of ICs of patients with rare cancers. Studies of ICs for individuals with rare diseases suggest significant need for information and emotional support [15], as well as psychological distress related to uncertainty about the future [11]. A systematic review of the experiences of ICs of patients with glioblastoma multiforme, a rare brain cancer, highlighted isolation and existential distress in the caregiving experience [16]. Additionally, one study found that ICs of patients with rare cancers reported greater decline in their own health compared to ICs for more common cancers [12]. Despite a growing interest and focus on the supportive care needs of cancer ICs [2], there remains limited research focused on the supportive care needs of ICs of patients with rare cancers [11,15,17,18].

Erdheim-Chester disease (ECD) is an orphan blood cancer with approximately 750 cases reported in the literature. ECD is marked by somatic mutations of the mitogen-activated-protein-kinase (MAPK) pathway, whereby histiocytes accumulate in tissues and cause illness [19]. ECD presents a wide variety of clinical phenotypes from limited to life-threatening forms and can involve multiple organ systems, including bones, brain, skin, kidneys, eyes, lungs, and heart [19]. As a result of ECD’s protean manifestations and rarity, delayed diagnosis is common; the median time elapsed from first presentation to ECD diagnosis is 3–5 years [19]. Furthermore, the hallmark manifestation of ECD is chronic and widely-varied symptomatology caused by tumors compressing organs and inflammatory cytokines circulating in the blood [20]. ECD causes generalized symptoms of fevers and fatigue as well as localized symptoms such as bone pain, ocular problems, cardiac dysfunction, neurologic deficits, and others. The rarity, clinical heterogeneity, and complex symptoms endured by ECD patients prompted us to investigate the experience of ECD ICs.

The purpose of the present study was to understand the unique experience and needs of ICs of patients with ECD in order to improve comprehensive care for patients and families living with this rare disease. We collected mixed-methods data regarding supportive care needs, burden, psychological distress, and quality of life among ICs of patients with ECD and explored whether their experiences would reflect the challenges of providing care to a patient with a rare cancer more broadly.

Method

Participants and Procedures

Participants were recruited from the Department of Neurology at Memorial Sloan Kettering Cancer Center (MSK) and the Erdheim Chester Disease Global Alliance (a patient advocacy group dedicated to ECD). Recruitment took place between March 7, 2018 and November 2, 2018 as a part of an Institutional Review Board (IRB)-approved exempt prospective study. Participants were individuals designated by ECD patients as friends or family members who provide unpaid help or support. They completed questionnaires via a secure REDCap server. A convenience subsample of participants was selected to partake in semi-structured interviews via telephone. No protected health information was collected.

Measures

Family Impact.

Participants completed the 16-item Covinsky Family Impact Survey to assess the impact their caregiving experience had on family finances, roles, and plans for the future [21]. Items include dichotomous and Likert-type response options to examine the extent to which the patient’s illness has altered the dynamic within the family unit.

Caregiver Burden.

Burden was assessed using the Caregiver Burden Inventory (CBI) and the Caregiver Reaction Assessment (CRA). The CBI [22] is a 24-item questionnaire used to assess five areas: Time Dependence, Developmental, Physical, Emotional, and Social Burden. Items are rated on a five-point Likert scale from 1 to 5 (i.e. never to nearly always). Coefficient alphas for each subscale ranged from .73 to .86. The CRA [23] contains 24 items, also rated on a five-point Likert scale from 5 to 1 (i.e., strongly agree to strongly disagree) in five dimensions: Caregiver Esteem, Lack of Family Support, Impact on Finances, Impact on Schedule, and Impact on Health. Domain and total burden and reaction scores were calculated according to published methods [24]. In validation studies with ICs of cancer patients, total score coefficient alpha was .74 and subscale coefficient alphas ranged from .57 to .85 [24].

Caregiver Unmet Needs.

Unmet supportive care needs were assessed for the subset of ICs who were in partnered relationships with patients using the Cancer Survivors’ Partners Unmet Needs (CaSPUN) assessment [25] which asked ICs to rate their level of 36 needs (i.e., not applicable to high need) within five areas of relationships, information, partner impact, comprehensive cancer care, and emotional support. Responses were dichotomized as the absence (N/A or satisfied) versus the presence (Low, Moderate, or High) of a need. Coefficient alpha for the total score was .86 and ranged between .48 and .81 for individual subscales [25].

Psychological Well-being.

Psychological well-being was assessed with the McGill Quality of Life Questionnaire-Revised (MQOL-R) and the Hospital Anxiety and Depression Scale (HADS). MQOL-R is a 15-item measure [26] that generates scores for four domains of IC QOL (Physical, Psychological, Existential, and Social), as well as a single overall QOL score. Cronbach’s alpha for the total scale was .94, with subscale alphas ranging from .66 to .87. The HADS is a 14-item questionnaire [27] measuring overall psychological distress with seven questions each for anxiety and depression. Items are rated with a 4-point Likert scale from 0–3 (i.e., definitely to not at all). Each subscale score ranges from 0–21, with ranges representing absent (0–7), mild (8–10), moderate (11–14), or severe (≥15) depression or anxiety [28]. Measure validation in 106 ICs of patients with terminal cancer strong internal consistency for both subscales, with alphas of .85 for anxiety and .84 for depression [29].

Semi-structured interview.

A semi-structured interview guide was developed by the Patient Reported Outcomes and Community Engagement and Language (PRO-CEL) Core of the Department of Psychiatry and Behavioral Sciences at MSK (Table 1). Interviews were conducted by a study investigator (A.J.A) via telephone. The interviews lasted 20–30 minutes, were recorded using encrypted devices, and transcribed verbatim for analysis.

Table 1.

Qualitative Interview Guide

Questions
1. What has it been like to be a caregiver for your loved one with ECD? How has the fact that ECD is so rare impacted your experience?
2. Tell me a little bit about how your loved one came to be diagnosed with ECD. Was it difficult to find a physician who understands this illness?
3. From your perspective, how long was it from the time your loved one experienced symptoms to his/her receiving the diagnosis?
4. What do you understand about your loved one’s treatment? Do you have a sense of his/her prognosis? What do you expect to happen after this treatment is completed?
5. What have your responsibilities been as a caregiver to your loved one with ECD?
6. Tell me a bit about the types of symptoms your loved one is experiencing. For example, does he/she have pain? Is this pain (or other symptoms) controlled? How have these symptoms impacted your role as a caregiver?
7. Have you noticed any changes in your loved one’s cognition, e.g., memory, attention, changes in personality? If so, what? And how have these impacted your experience and role as a caregiver?
8. Is your loved one able to work? What new responsibilities at home do you have now that you are taking care of him/her?
9. How have you been impacted emotionally by your caregiving role? Have you experienced anxiety? Depression? Sleep difficulties? Do you feel burned out?
10. How have you been impacted physically by your caregiving role? Have you been able to take care of yourself as well as your loved one with ECD?

Analysis:

Descriptive statistics were used to summarize the quantitative measures, with proportions, means, standard deviations and ranges as appropriate. Study staff completed qualitative analysis of interview transcripts through two rounds of independent and collaborative inductive thematic content analysis [30]. The code names and definitions in the inductive codebook were developed from in-depth review of the transcripts, meaning that the code categories of interest were grounded in participant experience. Codebook fidelity was established through an iterative process of independent, open coding and consensus. First, three members of the research team independently reviewed transcripts to identify key narrative content based on the study aims and to develop descriptive and interpretive codes. The team met weekly to review the coded content, resolve discrepancies on code assignment, and build consensus on identification of emergent themes. A second round of analysis was led by a senior qualitative methods specialist (K.L), who independently coded each transcript and reviewed the first round coded narrative content with a focus on the ways disease rarity may influence the experience and burden of ICs. This involved a round of open coding and consensus session with the study leaders (A.J.A, E.L.D) and the first round coding team, after which a subsequent coding round was completed using a consolidated codebook. After the codebook was established, the qualitative methodologist conducted a quality assurance review of all coded transcripts to ensure appropriate code application. Major themes included in this manuscript are those which were identified in more than half of interviews (n≥4). The team determined whether saturation was reached through two separate rounds of coding, with saturation defined as the point at which identified themes were felt to be fully understood, with no new themes identified in analyzing two additional interviews. Consistent with qualitative methodologies employed to identify preliminary themes in pilot studies [31,32], we anticipated that 5–10 participants would be sufficient to achieve saturation.

Results

Participant Characteristics

The 14 participants were predominantly female (71%, n = 10; see Table 2) and about half (53.4%, n = 8) were providing 60% to 100% of their loved one’s care, with 83% (n = 13) identifying as the sole IC. ECD involvement was most commonly reported in the bone (n = 11, 79%), brain (n = 10, 71%), heart (n = 8, 57%) and kidney (n = 8, 57%).

Table 2.

Participant Characteristics (N=14)

Characteristic		N (%)
Gender
	Male	4 (29)
	Female	10 (71)
Highest Education Level
	High School	3 (21)
	Undergraduate	1 (7)
	Graduate	5 (36)
	Post-Graduate	5 (36)
Employment Status
	Part-Time	1 (7)
	Full-Time	8 (57)
	Unemployed	1 (7)
	Retired	4 (29)
Household Gross Annual Income
	$25,000 – $50,000	1 (7)
	$50,000 – $75,000	5 (36)
	$75,000   $100,000	3 (21)
	$100,000 +	5 (36)
Site(s) of Patient’s ECD Involvement
	Bone	11 (79)
	Skin	4 (29)
	Brain	10 (71)
	Lungs	3 (21)
	Heart	8 (57)
	Kidney	8 (57)
	Eyes	6 (43)
	Spine	6 (43)
	Other	4 (29)
Percent of Patient’s Informal Care Provided
	0–19%	5 (36)
	20–39%	2 (14)
	40–59%	0 (0)
	60–79%	1 (7)
	80–100%	6 (43)
Patient’s Current ECD Treatment
	Targeted (kinase inhibitor) therapy	8 (57)
	Conventional (chemotherapy, immunosuppression) therapy	2 (14)
	Combined targeted/conventional	1 (7)
	No current therapy	3 (21)

Quantitative Data

Family Impact.

Diagnosis and subsequent treatment of ECD had a significant impact on family roles, finances, and future plans (Table 3). Most ICs (n = 8, 57%) agreed or strongly agreed that caregiving was a positive experience, and eight (n = 8, 57%) believed they were carrying out their responsibilities well or extremely well. However, since the patient’s diagnosis, almost all participants reported the patient’s needing at least some help from an IC (n = 13, 97%). As such, half (n = 7, 50%) reported cutting back at work, and three (21%) discontinued working altogether to provide care. Importantly, participants reported a negative financial impact of ECD diagnosis and treatment; six (43%) participants endorsed having used most or all of their family’s savings on care for the patient, and three (21%) reported moving to a less expensive home to meet the financial needs of ECD care.

Table 3:

Family Impact (N=14)

Item	N	%
Percentage of patient’s informal care you provide
0–19%	5	36
20–39%	2	14
40–59%	0	0
60–79%	1	7
80–100%	6	43
Is there anyone else providing informal care for the patient
No	13	93
Yes	1	7
Percentage of care this person provides (for caregiver answering yes to previous item; N=1)
Don’t know	1	100
Has caregiving been a positive experience for you
Strongly Agree	3	21
Agree	5	36
Disagree	4	29
Strongly Disagree	1	7
I don’t know	1	7
How well do you think you provide care
Extremely well	2	14
Very well	6	43
Adequately	5	36
Not well	1	7
Since cancer diagnosis, how much help has s/he needed from a family member
Completely	2	14
A great deal	3	21
Somewhat	8	57
Not at all	1	7
How many hours a week did you work prior to diagnosis
60 hours	3	21
50 hours	2	14
40 hours	4	29
35 hours	1	7
24 hours	1	7
None/homemaker/retired	3	21
Have your work hours each week diminished
Yes	7	50
No	7	50
How many hours less do you work each week (only for those answering yes to previous item; N=7)
5 hours	1	14
10–15 hours	1	14
12 hours	1	14
20 hours	1	14
35 hours	1	14
40 hours	2	29
Since cancer diagnosis, has anyone quit work/major changes to provide care
Yes	3	21
No	11	79
Has anyone become ill/stopped functioning normally since diagnosis
Yes	1	7
No	13	93
Has diagnosis required using all or most of family’s savings
Yes	5	36
No	9	64
Has diagnosis meant loss of major source of family’s income
Yes	7	50
No	7	50
Has cost of care required a move to a less expensive place to live
Yes	3	21
No	11	79
Has cost of care required putting off important medical care for another family member
Yes	2	14
No	11	79
Unanswered	1	7
Has cost of care required putting off education plans/changing plans for another family member
Yes	6	43
No	8	57

Caregiver Partners Unmet Needs.

Nine ICs were spouses to ECD patients and provided partner-specific data regarding unmet needs (Table 3). Of 12 items related to the dyadic relationship, eight were unmet in 50% or more of participants, including needing help with the impact cancer has had on their relationship (n = 7, 78%), with changes cancer has caused in the patient (n = 6, 67%), with knowing how to support the patient (n = 5, 56%) and how to address problems with their sexual/intimate life (n = 5, 56%). Other frequent unmet needs were related to needing assistance with decision making in the context of uncertainty (n = 6, 67%), emotional support (n = 5, 56%), and wanting to ensure the members of the medical team were communicating to coordinate care (n = 5, 56%).

Caregiver Burden.

Participants’ responses to the CBI reflected a generally low to moderate level of burden (mean (M) = 31.1, standard deviation (SD) = 21.6), with scores ranging from 8 to 72 (possible: 0–96) (Table 4). Seven (54%) ICs scored above 24, indicating a need for supportive services or respite care, and five (38%) reported scores indicative of increased risk for burnout. Subscale scores similarly indicated moderate levels of burden in the domains of Time-dependence (M = 8.9, SD = 6.0, range 1–20; possible: 0–20), Physical (M = 5.2, SD = 4.7, range 0–16; possible: 0–16), and Social (M = 3.9, SD = 3.9, range 0–12; possible: 0–20) burden. Participant responses to the CRA similarly indicated moderate levels of burden (mean total score 68.8 (SD = 4.8; range 59–79, possible: 24–120). Individual subscale scores are suggestive of low confidence in the IC role (M = 3.6, SD = 0.8; range 2.1–5, possible: 1–5), moderate lack of family support (M = 2.4, SD = 0.5; range 1.4–3.4, possible: 1–5) financial impact (M = 2.4, SD = 1.2; range 1–4.7, possible: 1–5), impact on their daily schedules (M = 3.2, SD = 0.9; range 2–4.6, possible: 1–5) and physical health (M = 2.0, SD = 0.5; range 1.3–3, possible: 1–5).

Table 4:

Unmet Caregiver Partner Needs (N=9)

Category	Unmet Need	N	%
Relationship	I need help to deal with the impact that cancer has had on my relationship with my partner	7	78
	I need help dealing with changes that cancer has caused my partner	6	67
	I need help to reduce stress in my partner’s life	5	56
	I need help to manage ongoing side effects and/or complications that my partner experiences as a result of their treatment	5	56
	I need help communicating with my partner and/or family	5	56
	I need help to know how to support my partner and/or family	5	56
	I need help adjusting to changes to my partner’s body	5	56
	I need help to address problems with our sex life	5	56
	I need help to manage my concerns about the cancer coming back	3	33
	I need to talk to others who have had a partner experience cancer	3	33
	My partner needs help to move on with their life	3	33
	I need help to cope with others not acknowledging the impact that having had a partner experience cancer has had on my life	3	33
Information	I need up to date information	3	33
	I need information provided in a way that I can understand	2	22
	I need information relevant to my needs as a partner	1	11
Partner Impact	I need help with trying to make decisions about my life in the context of uncertainty	6	67
	I need more accessible hospital parking	4	44
	I need help to deal with changes to my partner’s working life and/or usual activities	4	44
	I need help to deal with the impact that my partner’s cancer has had on my working life and/or usual activities	4	44
	I/we need help to find out about financial support and/or government benefits to which we are entitled	4	44
	I need help to manage additional responsibilities as a result of my partner’s illness	4	44
	I need help to make my life count	4	44
	I need help to explore my spiritual beliefs	3	33
	I need help with my own health problems	2	22
Comprehensive Cancer Care	I need to know that all my partner’s doctors talk to each other to coordinate my partner’s care	5	56
	I need to feel like I am managing my partner’s health together with the medical team	4	44
	I need local health care services that are available when my partner requires them	3	33
	I need my complaints regarding my partner’s care to be properly addressed	2	22
Emotional Support	I need emotional support to be provided for me	5	56
	I need emotional support to be provided for my loved ones	4	44
Other Needs	I need my partner to have an ongoing case manager to whom I can go to find out about services whenever they are needed	4	44
	Due to my partner’s cancer, we need help getting life and/or travel insurance	4	44
	I need help to handle the topic of cancer in social and/or work situations	3	33
	Due to my partner’s cancer, I/we need help accessing legal services	2	22
	We need help with having a family due to fertility problems	0	0

Psychological Well-Being.

The MQOL-R total score suggested favorable overall QOL (M = 6.7, SD = 2.6, range 0–10; possible: 0–10). Subscales demonstrated generally good Physical QOL (M = 7.5, SD = 2.8, range 2–10; possible: 0–10), and adequate Psychological (M = 6.5, SD = 3.1, range 0.5–10; possible: 0–10), Existential (M = 6.4, SD = 2.4, range 1.8–9.5; possible: 0–10) and Social (M = 6.7, SD = 2.9, range 1.7–10; possible: 0–10) QOL. Nonetheless, 21% (n=3) of ICs reported mild, 14% (n = 2) moderate, and 29% (n=4) severe levels of anxiety as measured by the HADS. Additionally, 14% (n = 2) reported severe levels of depression.

Qualitative Data

Thematic analysis of our semi-structured interviews indicated that saturation was achieved in our sample of 7 participants. Five themes emerged (see Table 5 for representative direct quotations):

Table 5:

Mood, Burden, Quality of Life, Caregiver Reaction (N=14)

HADS Anxiety Total Score (N=14)	N	%
0–7 (normal)	5	36
8–10 (mild)	3	21
11–14 (moderate)	2	14
15+ (severe)	4	29
Depression Total Score (N=14)	N	%
0–7 (normal)	12	86
8–10 (mild)	0	0
11–14 (moderate)	0	0
15+ (severe)	2	14
CBI Burnout and Needing Respite (N=13)	N	%
≤36 (Not at burnout risk)	8	62
>36 (Burnout risk)	5	38
≤24 (Not needing to seek respite care)	6	46
>24 (Needing to seek respite care)	7	54
McGill total and subscale scores (Number of items) (N=14)	Mean	St.Dev.	Min	Max	Completing all items (N/%)
Overall QOL (14 items)	6.7	2.6	0.0	10	14	100
Physical (3 items)	7.5	2.8	2	10	14	100
Psychological (4 items)	6.5	3.1	0.5	10	14	100
Existential (4 items)	6.4	2.4	1.8	9.5	14	100
Social (3 items)	6.7	2.9	1.7	10	14	100
Total Score (mean of 4 subscales)	6.8	2.4	1.5	9.4	14	100
Caregiver Reaction Assessment Total and Subscales (N=14)	Mean	St.Dev.	Min	Max	N	%
Total (24 items)	2.9	0.2	2.5	3.3	13	93
Factor 1: Caregiver’s Esteem (7 items)	3.6	0.8	2.1	5.0	14	100
Factor 2: Lack of Family Support (5 items)	2.4	0.5	1.4	3.4	14	100
Factor 3: Impact on Finances (3 items)	2.4	1.2	1.0	4.7	14	100
Factor 4: Impact on Schedule (5 items)	3.2	0.9	2.0	4.6	13	93
Factor 5: Impact on Health (4 items)	2.0	0.5	1.3	3.0	14	100
Caregiver Burden Inventory Total and Subscales (N=14)	Mean	St.Dev.	Min	Max	N	%
Total (24 items)	31.1	21.6	8	72	13	93
Time Dependence Burden (5 items)	8.9	6	1	20	14	100
Developmental Burden (5 items)	10.5	5.9	3	20	14	100
Physical Burden (4 items)	5.2	4.7	0	16	14	100
Emotional Burden (5 items)	5.4	5.1	0	15	14	100
Social Burden (5 items)	3.9	3.9	0	12	13	93

Theme 1: Delayed diagnosis and treatment led to unique responsibilities, burden, and distress.

Participants described an invariably arduous and protracted process of achieving an ECD diagnosis, involving repeated consultations, invasive procedures, and misdiagnoses. During this process, ICs described the burden of being the only person pursuing all measures necessary to achieve a diagnosis, such as advocating for additional expert opinions. Furthermore, the patient often clinically deteriorated during this period. As such, in addition to feeling responsible for managing the diagnostic process, ICs endured distress and guilt in witnessing the patient’s worsening health and symptoms.

Theme 2: Lack of reliable information about ECD engendered a sense of responsibility in ICs to gain scientific expertise.

First, participants described a lack of accessible and reliable information about ECD from both physicians and public sources. Specifically, information on the Internet was scarce and conflicting, and the process of finding guidance was discouraging. ICs therefore often felt pulled to read technical literature and independently gain an understanding of the science and treatment of ECD. Relatedly, they found themselves serving as ECD “experts” to friends and family, providing information and answering questions, which was an added challenge and responsibility.

Theme 3: Anxiety and isolation are prominent features of the ECD IC experience because of poorly understood ECD symptomatology and prognostic uncertainty.

Participants explained that ECD patients frequently encounter new and unexplained symptoms, and it was difficult for the healthcare team to discern whether they represented worsening of disease, side effects of treatment, or an unrelated issue. This uncertainty generated persistent worry about disease progression when facing unclear symptoms. Furthermore, ICs described anxiety about the patient’s prognosis, including concerns about how long a current treatment would be effective, what would happen if treatment failed, and regarding the patient’s life expectancy overall. The additional financial burden of assessing and treating unclear symptoms was also a source of anxiety. Finally, participants elaborated upon the particularly isolating manifestation of ECD that is unexplained cognitive impairments. ICs described the phenomenon of patients’ having difficulties with thinking, memory, personality changes, or behavior in the absence of classic neurologic symptoms (motor weakness, imbalance), and often in the absence of their seeming generally ill. While ICs recognized that these impairments were observed by friends and family, they remained unexplainable and hence contributed to feelings of anxiety and isolation.

Theme 4: The caregiving role was not central to participants’ sense of identity.

Even though participants reported significant and negative impact of the caregiving role, many did not identify as an IC. In the setting of chronic disease, once the period of diagnosis and initial treatment was complete, participants conceptualized their responsibilities as an extended function of their pre-existing relationship with the patient (predominantly spousal) rather than as a unique caregiving role. For some, this was explained in light of the patient’s intact functional status which did not require active care on part of ICs.

Theme 5: Psychosocial support from other ECD ICs allowed participants to feel understood and receive helpful advice.

Participants found that the community of ECD ICs was a critical source of support and adaptive coping. Specifically, they noted that fellow ECD ICs offered empathy that felt genuine and helpful because of their shared experiences and challenges. Social media was observed to have greatly facilitated connections among ECD ICs, particularly via the Erdheim Chester Disease Global Alliance. Facebook and other social media channels allowed ICs to connect and help each other navigate their experiences in ways not imaginable before online communities. Finally, participants endorsed the unique opportunity of ECD ICs to provide more in-depth support for one another in an IC-only environment, in the absence of patients, where they could express their feelings without exposing them to the patient.

Discussion

This study is the first investigation of the experience of ICs of patients with ECD. Despite our small sample, a clear picture emerged through the quantitative and qualitative data. Similar to other cancer ICs [34], those providing care to patients with ECD reported that caregiving significantly impacted family roles and responsibilities, the ability to plan for the future, and financial well-being. As has also been found in the context of other cancers [7,35], several ICs reported that entire family savings had been used to care for the patient with ECD, and changes to employment or residence were needed to accommodate caregiving needs. Interestingly, ICs here did not identify primarily with the caregiving role but instead, viewed their responsibilities as an extension of their relationship with the patient. This finding, which has been documented elsewhere in the caregiving literature [36–38], may reflect the fact that the majority of ICs interviewed were the partners of patients, and as such the caregiving role was syntonic with the spouse/partner role (as compared to adult child ICs, for example, whose caregiving responsibilities are outside the boundaries of the role of child). However, the nine ICs who identified as partners reported greater unmet partner needs than has been previously reported among partner ICs of cancer patients, such as knowing how to support the patient (6.4%), decision making in the context of uncertainty (8.2%), ensure medical team communication (9.8%)) in other studies [39].

The levels of burden reported here were generally similar to investigations of ICs of patients with more common cancers, such as breast, prostate, and lung [40,41], and of ICs of patients with advanced cancers and those receiving palliative care [24][42]. This low to moderate level of burden is initially surprising given the significant challenges ICs described in the in-depth interviews. We hypothesize that these data reflect the timing of the assessment: most ICs interviewed long after the period of initial diagnosis and treatment and had adjusted to the ECD caregiving context. It is possible that higher burden scores would have emerged had we interviewed ICs earlier on in the caregiving trajectory. It is also possible that this level of burden reflects the relatively low amount of care provided by some of our participants, which was below 20% of the patient’s overall informal care in 5 (36%). Notably, 64% of participants endorsed at least mild - and more than one quarter reported severe - anxious symptomatology, and 14% reported severe depressive symptomatology—rates that are comparable to that observed in one study of 84 cancer ICs [33]. Importantly, studies of ICs for individuals with more common cancers have shown that anxiety and depressive disorders develop across the caregiving trajectory when early mild symptoms are untreated [43]. Therefore, among ECD ICs, the provision of psychosocial support early in the disease course may be beneficial to diminish anxiety.

Despite the familial and financial burden, and the psychiatric symptoms described above, our quantitative assessments indicated that participants had a favorable overall QOL, felt that caregiving was a positive experience, and that they were carrying out the responsibilities of this role well. This duality has been documented previously among ICs of patients with more common cancers [33], who despite the burden and distress they experience concurrently report a sense of accomplishment, growth, and satisfaction in their role. Indeed, a growing body of literature documents the concurrence of meaning and suffering resulting from caregiving [33,44], and the possibility of personal growth despite the challenges of caregiving. As such, ECD ICs may benefit from interventions such as Meaning-Centered Psychotherapy for Cancer Caregivers (MCP-C) that helps ICs to find meaning and purpose in the caregiving role and to foster a sense of resilience and growth [33,45].

Several themes emerged from our qualitative data to suggest that our quantitative findings can be explained by the rarity of ECD. First, the burden and impact of caregiving felt by participants resulted from the protracted process of obtaining an ECD diagnosis and from the lack of robust and consistent medical information available about a rare and poorly understood disease. The significant financial impact reported, specifically, is understandable considering the need for extensive and repeated diagnostic evaluations and consultations. Anxiety that was frequently observed among our participants derives from many aspects of navigating a rare disease, including (1) witnessing the patient’s clinical deterioration in the absence of a clear diagnosis, and (2) continuous worry about disease progression brought about by an abundance of clinical symptoms. Insofar as high burden of symptoms has been found to be present across rare cancers [46], our results raise the question of whether anxiety related to symptomatology may be a consistent feature of rare cancer caregiving.

Uncertainty about ECD prognosis was also a source of anxiety for participants. This finding resonates with studies of rare disease caregivers, for whom uncertainty about the future is a significant source of distress [11]. To date, research in other cancer ICs has found that the receipt of accurate prognostic information – even when such information indicates a poor prognosis – is not associated with anxiety, depression, or other poor psychosocial outcomes, but instead helps facilitate productive healthcare communication [47]. In the case of ECD ICs, where the dynamic state of scientific knowledge renders accurate prognostic information inherently difficult to provide, worry about the future stemming from uncertainty may be uniquely severe. These data suggest that anxiety among ECD ICs may represent an important target for supportive intervention, and they also raise the question of whether ICs for other rare cancers have a similar unmet need.

Interviews revealed a particularly challenging aspect of ECD caregiving: cognitive, personality and behavioral changes, at times in the absence of conventional neurologic deficits. Neuro-oncologic diseases such as malignant brain tumors cause a constellation of cognitive, personality and behavioral changes together with overt impairments such as language or motor dysfunction. This constellation of impairments allows friends and family to understand the cognitive changes to be part of a broader neurologic illness, and to provide empathy and support accordingly. In ECD, ICs described the difficult experience of navigating cognitive symptoms, particularly in the social environment, without associated deficits that provide implicit context and explanation for the neurologic changes. Participants endorsed feeling isolated and diminished social support as a result, and difficulty communicating about the patient’s health to friends and family. This finding raises the important question of how isolation may be an important contributing factor to the distress experienced by ICs of other rare cancers, depending on the ways that the disease departs from more common cancers and affects the dynamics of social support.

Several limitations of this study must be noted. Most importantly, our sample size is small, and this limits the generalizability of our findings. However, we believe our study is informative as ECD is an ultra-rare condition and there have been no previous studies on the experiences of ECD ICs. While our findings are not exhaustive on this topic, this pilot study has identified important areas to explore further in future studies of the ECD IC experience. Within our small sample, our participants were primarily women, however this is consistent with a male predominance of ECD (70% of patients) that has been observed repeatedly [48,49]. Also, we did not robustly capture or analyze characteristics about the ECD patients for whom our participants provided care, such as the duration of their illness, which could have informed the caregiving experience at the time it was reported. Nonetheless, our mixed-methods approach allowed us to glean rich data from our participants and highlight opportunities to improve IC care by, for example, offering supportive interventions to ameliorate anxiety, helping to foster the positive experience of caregiving, and by taking particular notice of ICs of patients with cognitive impairments. Our study also suggests directions for future investigation. First, larger studies of ECD ICs would be informative to confirm and expand the findings of this study. Also, research in this domain, and the domain of rare cancer caregiving, may benefit from a focus upon the ramifications of delayed diagnosis, factors leading to anxiety and isolation, and unmet needs for disease-specific information and psychosocial support. Further study of this area holds the promise to improve outcomes for patients with rare cancers and their ICs.

Table 6.

Qualitative Themes and Representative Quotations (N=7)

Theme	Experiences Supporting Theme	Reference Quotes
Theme 1: Burden and distress related to delayed diagnosis and treatment (N = 6)	ICs endured the burden of advocating for the patient through misdiagnosis and skepticism about the legitimacy of the patient’s illness.	“They… they kept asking me are you sure he doesn’t do drugs, are you sure he doesn’t do drugs? So - - no, he’s never done drugs in his life. It’s not that. […] They misdiagnosed him […] I got mad actually at the doctors here and decided to drive him to Mayo Clinic. And the guy up there, he knew his stuff, he knew right away what it was.”
	ICs experience emotional distress witnessing the patient’s deteriorating health during the period of diagnosing ECD.	“Then her condition got worse and worse all through the Fall, the blood pressure problem was worse until finally – I guess around December or so and we had get her to the hospital because of the blood pressure and it was so bad that her kidney shut down. She went on dialysis. Meanwhile, they still didn’t know what the – what was going on. I guess they thought the brain problem was separate from the blood pressure problem, but it wasn’t.”
Theme 2: Lack of reliable information about ECD and the burden of caregiver expertise (N = 5)	ICs feel obligated to become an expert about ECD in order to reassure the patient and family members, despite the lack of available reliable information.	“There’s not much information to draw on. You know? So when you have questions, you’re just really limited to how much information is out on the internet or you know how you get very much information so that’s difficult.”
		“I know that from all the reading I’ve done, and I just told him, I’m in charge, you don’t have to worry, I’ve got it all covered.”
Theme 3: Anxiety and isolation resulting from poor understanding of ECD symptomatology and prognosis (N = 7)	ICs experience tremendous anxiety as a result of not knowing the prognosis of ECD, i.e. how long the patient is expected to survive and what the future holds. They find it difficult to prepare for the future.	“So… it’s scary. It’s been terrifying, it’s not knowing like -- before he got diagnosed what was wrong with him and then once he got diagnosed… you know you hear that word and it scares you and you don’t know what the future holds and it’s… it makes me anxious like… I don’t know if he’s going to be around in ten years and… so all our plans we had for the future have are just kind of toast -- connotation like that’s what we’re going to plan to do, assuming I still have that ten years or assuming he’s still alive in five years.”
	Cognitive symptoms cause worry about neurodegenerative disease.	“He’s fixated on certain things and yeah, I don’t know [if it’s] ECD or it’s his aging process […] It’s always been there it’s just so much more now. I don’t know if anybody can say it’s this disease, I mean, right? […] I will tell you, because I think it’s of note, that his father died of complications from Alzheimer’s, a specific type of Alzheimer’s and I hope – you know that I really pray we don’t have – we have enough down the road and I hope the absentmindedness is not early symptoms of that.”
	ECD patients are often not recognizable as ill to others, and ICs receive less support and concern from friends and family who are not aware of the seriousness of the illness.	“We went to a birthday party yesterday for the family and she was a full participant. When people look at her, they can’t understand […] It doesn’t affect your appearance at all and all her… when we go to the annual meeting, that’s something that people say [about having ECD], ‘you look fine, you talk fine, what’s wrong with you?’ […] When you look at her, she looks fine. The neighbor – I saw him out on Friday. He goes, ‘was that your wife out front yesterday? ‘Cause she never goes outside hardly at all, out front’ and – I go, yeah and he goes, ‘I thought she was sick. She looks good; must be doing really good.’ I go, well, you know….”
Theme 4: Dynamic sense of Caregiver identity (N = 6)	Rather than embracing the identity of a “caregiver,” some ICs conceptualize their responsibilities as an extended function of their pre-existing relationship with the patient; especially after the initial treatment period has ended.	“I’m 71 now, so let’s say I was 64 or so when this began, and we were married for I don’t know – 30 something years by then. So it was just like part of the deal.”
		“I don’t know if I really qualify because my husband’s pretty self-sufficient…”
	Although a caregiver may have taken on an “active” role during diagnostic process, once symptoms are managed and the patient is stable, their role may have shifted from physical to emotional support.	“But it’s out and about and in terms of her daily needs, there’s not as much need for that. I’ll offer her my arm if she’s not using a cane if we’re walking somewhere but it’s not extreme in terms of the amount of caregiving.” “You know my caregiving consists of consoling, because now she says this is not the life I expected. “
Theme 5: Need for psychosocial support that is specific to the ECD caregiving experience (N = 5)	ICs found support in the community of other ECD ICs. This highlights the importance of social media as an avenue of support for rare cancer ICs	“I think certainly the ECD caregiver chat has been helpful in sharing experiences and […] there’s a couple of different ways people can offer support. One is just by kind of empathizing, hey, I’ve been there. I know what it’s like. Keep hanging in there. And the other is kind of understanding what they’re going through”
	Supportive groups and experiences attended only by ICs were critical in their allowing ICs to express their fears and vulnerabilities.	“That was extremely powerful to me to be able to go in a room where every -- with all the other caregivers and you know it’s never your -- my husband is never going to hear what I have to say. He’s never going to hear that it’s hard ‘cause I don’t let him see that. I try not to let him see that, you know what I mean?”