Dyadic communication in rare cancer: a registry-based study of patients with Erdheim-Chester disease and their caregivers

TO THE EDITOR:

A cancer diagnosis causes distress and instability not only to the affected patient but also to the psychosocial ecosystem of the patient, most notably their family and friend caregivers. Caregivers support patients in all components of their disease trajectory, from navigating the diagnosis to discussing the symptoms, treatments and side effects, and even advanced care planning. Investigating the dyad, that is, the patient-caregiver pair, is of critical importance. Open dyadic communication between a patient with cancer and their caregiver is associated with positive elements of dyadic coping, such as increased intimacy and emotional acceptance, fewer physical symptoms, and less psychological distress.^1,2 Conversely, avoidance of communication about illness is associated with adverse features of dyadic coping, such as decreased intimacy and relational satisfaction.³ 

Existing literature about dyadic communication in cancer has examined common cancers; there has been little examination of patient-caregiver communication about illness for dyads navigating a highly rare cancer. Although individual rare cancers are uncommon, rare cancers, in aggregate, constitute 20% of all cancer diagnoses in the United States,⁴ rendering them an important and distinct entity to investigate from the viewpoint of patient, caregiver, and dyadic experience.

Erdheim-Chester disease (ECD) is a very rare blood cancer, with ∼1500 cases since 1930. It is a disease with protean manifestations that can affect any structure, including bones, skin, eyes, and brain.⁵ Additionally, it is nearly invariably defined by delayed diagnosis exceeding 2 years.⁶ We have demonstrated dense, widely varied, and poor symptomatology, including frequent fatigue, pain, sadness, and anxiety in patients with ECD.^7,8 Among caregivers of individuals with ECD, we have identified a constellation of challenges emanating directly from the experience of providing support for a patient with a rare disease, such as heightened anxiety, distress, and unmet support needs related to protracted undiagnosed illness and scarcity of reliable information about ECD. Herein, we investigated communication about illness within the ECD patient-caregiver dyad; we hypothesized that factors salient to the experience of patients and caregivers with rare cancer would be associated with worse dyadic communication about illness.

This study consists of an institutional review board–approved registry study and caregiver study maintained at Memorial Sloan Kettering Cancer Center (#NCT03329274 and #NCT03990428). All data are contained in Research Electronic Data Capture, a platform for collecting data in a secure web-based interface.

Participants were patients with biopsy-confirmed ECD and their caregivers. Patient participants were self-referred or referred from within the Memorial Sloan Kettering Cancer Center cohort by their health care providers or from centers recognized by the ECD Global Alliance. Patients and caregivers provided informed consent to participate and complete self-reported assessments. Dyads of patient participants with ECD who enrolled from 2018 to 2021 and their caregivers with completed assessments were included.

The main outcome was caregiver-assessed dyadic communication, which was evaluated using 2 caregiver-reported measures: (1) The Cancer Communication Assessment Tool for Patients and Families (CCAT⁹) family communication subscale and (2) The Caregivers’ Communication with Patients about Illness and Death (CCID¹⁰) scale.

Patient and caregiver demographics and patient disease characteristics were recorded. For patients only, self-reported assessments included the Functional Assessment of Cancer Therapy–Cognitive assessment, ECD Symptom Scale, Instrumental Activities of Daily Life, Brief Fatigue Inventory, and Brief Pain Inventory. For caregivers only, self-reported assessments included the Duke–University of North Carolina Functional Social Support Questionnaire, University of California, Los Angeles Loneliness Scale, Difficulties in Emotional Regulation Scale, and Caregiver Reaction Assessment. For both patients and caregivers, assessments included the Supportive Care Needs Survey and the Hospital Anxiety and Depression Scale.

The CCAT family communication subscale and average CCID scale were correlated using Pearson correlation coefficient. Demographics, patient disease characteristics, patient-reported assessments, and caregiver-reported assessments were individually associated with caregiver-assessed dyadic communication, with univariable linear regression models using the CCAT family communication subscale and the CCID scale separately as outcomes. Tests were 2 sided with statistical level of significance <.05. Analyses were performed in SAS version 9.4.

Sixty-three dyads contributed to analyses (Table 1). Patients were aged 54.1 years, on average, at assessment, and caregivers were 52.9 years. Patients were mostly male (65%), and caregivers mostly female (76%). The mean and standard deviations (SD) of the CCAT family communication subscale and average CCID scale were 11.1 (SD, 4.0) and 2.3 (SD, 1.1), respectively, and the measures were moderately correlated (ρ = 0.45; 95% confidence interval [CI], 0.22-0.62).

Modeling the CCAT family communication subscale, caregivers of patients on treatment were more likely to report worse caregiver-assessed dyadic communication (Table 2). Caregivers of patients with unmet psychological needs (β_CCAT = 2.25; 95% CI, 0.27-4.23) or greater impact of perceived cognitive impairments (β_CCAT = −0.21; 95% CI, –0.42 to –0.01) were more likely to report worse caregiver-assessed dyadic communication, as were caregivers of patients rating their top 3 ECD symptoms occurring almost constantly (β_CCAT = 7.25; 95% CI, 1.69-12.80). If a patient reported more severe pain, the caregiver was more likely to report worse dyadic communication (β_CCAT = 0.44; 95% CI, 0.04-0.84). Worse caregiver anxiety (β_CCAT = 0.30; 95% CI, 0.08-0.52) and depression (β_CCAT = 0.41; 95% CI, 0.19-0.62) were associated with worse caregiver-assessed dyadic communication, as were more unmet caregiver work and social needs (β_CCAT = 2.75; 95% CI, 0.73-4.77) and informational needs (β_CCAT = 2.32; 95% CI, 0.35-4.29). Lower caregiver-perceived self-esteem (β_CCAT = −0.28; 95% CI, –0.55 to –0.005) and greater caregiver challenges with emotional regulation (β_CCAT = 0.12; 95% CI, 0.07-0.17) were associated with poorer caregiver-assessed dyadic communication.

Modeling the CCID scale mean, caregivers of patients with more anxiety reported worse communication (Table 2). Caregivers of patients with severe and interfering fatigue reported poorer communication (β_CCID = 0.14; 95% CI, 0.07-0.22; β_CCID = 0.10; 95% CI, 0.02-0.18, respectively). Similar to when the CCAT family communication subscale was the outcome, worse caregiver anxiety and depression as well as greater caregiver challenges with emotional regulations were all associated with worse caregiver-assessed dyadic communication (β_CCID = 0.09; 95% CI, 0.03-0.14; β_CCID = 0.08; 95% CI, 0.02-0.14; and β_CCID = 0.03; 95% CI, 0.02-0.04, respectively). Finally, higher caregiver financial burden was associated with worse caregiver-assessed dyadic communication (β_CCID = 0.09; 95% CI, 0.02-0.16).

In this cross-sectional study, we have identified several patient- and caregiver-related factors affecting dyadic communication in a large, paired patient-caregiver cohort of a very rare neoplasm, a field neglected by existing literature. This aligns with the Theory of Dyadic Illness Management of Lyons et al, which describes illness management, whether rare or common, as a dyadic experience.¹¹ The involvement of caregivers is inherent to the adult cancer experience, but the factors that influence dyadic illness management vary. A prior study by Vukcevic et al of patients and paired caregivers with diffuse large B-cell lymphoma demonstrated that patients experience caregivers as instrumental to their comprehension of treatment-related discussions, and, furthermore, that developing tailored educational materials for caregivers could aid in increasing the proportion of patients (71%) satisfied with their treatment experience.¹² In a dyadic study of patients with more common hematological cancers and their caregivers, Siminoff et al demonstrated that dyadic communication was most discordant for dyads with lower levels of education and income and for cancers diagnosed at later stages.¹³ In this current study, education and income level were not associated with communication, and, although ECD is not traditionally staged, ECD duration, a potential proxy, was also not associated with communication. We demonstrated in the rare cancer setting that dyads in which patients suffered more frequent disease symptoms, greater cognitive difficulties, and more severe pain and fatigue demonstrated worse dyadic communication. Caregivers with more frequent anxiety and unmet supportive care needs, higher financial burden, and less social support experienced poorer communication about illness.

Our results can be viewed in the context of prior psychosocial studies about patients with myeloproliferative neoplasms (MPNs). US-based and international MPN studies have evaluated patient-reported outcomes in patients with MPN and demonstrated reliance on caregivers, impact of MPN upon the patient-caregiver relationship, and the need for decisional and social support.^14-19 However, to our knowledge, the existing literature did not address dyadic communication. We investigated dyadic communication among patients with ECD and their caregivers and identified factors that may be unique to rare diseases, for example, unmet informational needs. The relatively young age of our caregivers and patients, in a disease now considered chronic, underscores the importance of durable interventions to improve dyadic communication in this population. A recent meta-analysis of 14 couple-based communication intervention studies in common cancers found improvements in individual functioning and mutual communication.²⁰ Furthermore, dyadic web-based interventions²¹ that can customize content have also shown to affect psychological health and the dyadic relationship.

Although our cross-sectional design yielded participants heterogeneous in disease trajectory and treatment, our study represents a large dyadic cohort for a rare disease and, to our knowledge, the largest dyadic cohort in ECD investigation. We have identified meaningful patient and caregiver factors associated with communication that are amenable to interventions.

Based on successes reported for web-based interventions of dyadic communication in common cancers, we propose that intensive disease-specific remote interventions focused upon disease education, specialized supportive care needs, and management of symptoms and anxiety be tested in ECD and other rare diseases. The current post-COVID era provides unprecedented opportunity to evaluate and implement such interventions.

Acknowledgments: This work was supported by National Institutes of Health/National Cancer Institute grant P30 CA008748, a Population Sciences Research Program award (E.L.D. and K.S.P.), and National Cancer Institute grant R37CA259260 (E.L.D. and K.S.P.). This work was also supported by the Frame Family Fund (E.L.D.), the Joy Family West Foundation (E.L.D.), the Applebaum Foundation (E.L.D.), and the Erdheim-Chester Disease Global Alliance (E.L.D.).

Contribution: A.S.R. was responsible for conceptualization, formal statistical analysis, interpretation, and drafting the manuscript; D.B., J.J.B., and A.M.S. were responsible for conceptualization, data collection, interpretation, and drafting the manuscript; D.F., K.B., J.G., T.M.A., J.J.M., and A.J.A. were responsible for conceptualization, interpretation, and drafting the manuscript; K.A.L. and K.S.P. were responsible for conceptualization, interpretation, drafting, and supervision; E.L.D. was responsible for conceptualization, data collection, interpretation, drafting the manuscript, and supervision; and all authors reviewed and edited the final version of the manuscript and agreed with the submission.

Conflict-of-interest disclosure: E.L.D. reports unpaid editorial support from Pfizer and serves on advisory boards for Day One Therapeutics, Opna Bio, and Springworks Therapeutics, outside the submitted work. A.J.A. reports paid consulting with Blue Note Therapeutics and BeiGene, outside the submitted work. The remaining authors declare no competing financial interests.

Correspondence: Eli L. Diamond, Department of Neurology, Memorial Sloan Kettering Cancer Center, 160 East 53rd St, 2nd Floor Neurology, New York, NY; e-mail: diamone1@mskcc.org.