Fear of cancer recurrence in patients with multiple myeloma: Prevalence and predictors based on a family model analysis

Abstract Objective Fear of cancer recurrence (FCR) is a common psychosocial sequela among cancer survivors, but data on patients with multiple myeloma are scarce. This study calculated the prevalence of FCR and identified family and social factors that predict FCR in the study population. Methods We recruited 127 myeloma patients and their partners to participate in a cross‐sectional survey from a regional tertiary cancer centre in China. The questionnaires included items on demographic characteristics and from the fear of disease progression simplified scale, family hardiness index and Social Support Scale. Univariate and multivariate regression was used to identify predictors of FCR. Results Of the participants, 56.4% patients reported high‐level FCR, which was similar to the partner‐reported proportion. The partners' FCR was positively associated with the patients' FCR, while family hardiness and social support were statistically significant, negative predictors. Conclusions Interventions to mitigate partners' FCR and improve family hardiness and social support may help with the psychological adjustment and well‐being of myeloma patients.

well-being and global quality of life. Cancer impacts patients and their caregivers and family members negatively. 6,7 Mellon and colleagues proposed a family based model to analyse predicting factors of FCR in patients and their family caregivers. 8 According to Mellon's model, both individual and dyadic factors were associated with FCR.
Personal factors, such as a partner's age, individual and concurrent family stressors and a patient's appraisal of the illness, could predict individuals' degree of FCR. A patient's FCR was found to be interdependent with that of the caregiver. The model is critical for addressing potential factors influencing patients' and their family members' FCR. Using this family model as a conceptual framework, we conducted the present study with the aims of (1) investigating myeloma survivors' residual FCR issues after completing conventional treatment and (2) uncovering relationships between demographic data, partners' FCR, family hardiness, social support and the FCR of myeloma patients in China.

| Participants
Convenience sampling was used to recruit participants. The inclusion criteria for patients were: (1) previously diagnosed with multiple myeloma according to the Multiple Myeloma Diagnosis and Treatment Guidelines 9 ; (2) awareness of disease diagnosis and condition; (3) completed conventional treatment, in the maintenance phase, currently in remission or disease control without a history of recurrence; (4) able to understand the contents of the questionnaires; able to speak and read in Chinese Mandarin or Cantonese and (5) provided consent to participate. We excluded patients with concurrent major physical or mental health problems (e.g., dementia, acute myocardial infarction). For patients who met the inclusion criteria, further inclusion reviews for their spouses included: (1) married and living with the patient, (2) willing to participate in this study, (3) able to understand the contents of the questionnaires and (4) able to speak and read in Chinese Mandarin or Cantonese. During the study period from October 2018 to September 2019, 127 pairs of myeloma patients and partners were enrolled at Sun Yat-sen University Cancer Center. All patients were married.

| Measures
We took a dyadic approach to investigate patients and their spouses.
All patients completed four in-person questionnaires to gather personal information and assess FCR, family hardiness and social support.
Their spouses completed two questionnaires to gather personal information and assess FCR. After completing the questionnaires, inperson interviews were conducted with the patients and their spouses separately. Missing items were confirmed with the participants, using questions, such as 'Did you have any problems in answering this question?' and 'What do you think about this question?'. The investigators encouraged the participants to provide details and to clarify their statements. Demographic data, including age, gender, education level, residence, occupation, income level, quantity of children, religion and healthcare coverage, were assessed via participant self-reports. 'Need for psychological consultation' was evaluated using questions such as 'Did you ever consider of seeking advice from psychological consultants?' and 'How often do you think about this question?'. The investigators collected clinical data, including myeloma stage, duration since diagnosis and comorbidity. FCR was measured with the Fear of Progression Questionnaire short form (FoP-Q-sf), which is a 12-item, bidimensional (physical health and social-familial) scale 10 that was simplified from the original 43-item version. 11 The items are scored on a 5-point Likert scale ranging from 1 ('never') to 5 ('always'). The potential total score ranged from 12 to 60. A higher score indicates a higher level of FCR.
According to a previous study by Herschbach et al., 12 the cut-off for dysfunctional FCR is ≥ 34. The Chinese version of the questionnaire demonstrated high internal and test-retest reliability and high consistency in Singapore. 13 The partners' FCR was measured with the FoP-Q-sf/P, the only instrument available to evaluate partners' FCR at present. It was developed by Zimmermann 14 and is based on the FoP-Q-sf; thus, it has the same cut-off value.
Family resilience was assessed using the Family Hardiness Index (FHI), a 20-item, tridimensional (commitment, challenge and control) scale. It was developed by McCubbin in 1986 to evaluate the internal strengths of family members. 15 The FHI totals all the responses to every item and has a possible score range of 0-60. There is no cut-off value to avoid discriminating against participants. A higher score indicates a higher level of family hardiness.
Social Support Rating Scale (SSRS) assesses the strength of social support. This 10-item, tridimensional scale was developed by Xiao 16 and has been widely used in China. For the classification of low, moderate, good and high levels of social support, the cut-off values are 26, 36 and 46, respectively.

| Ethics
The present study was approved by the Research Ethics Committee of Sun Yat-sen University Cancer Center (Approval Number: B2020-054-01) and conducted following the Declaration of Helsinki.
Data collection permission was acquired from hospital administrators. Written consent was obtained before collection, and anonymity was ensured. Trained investigators performed the data collection and assistance with completing questionnaires when needed.

| Statistics
Data analysis was conducted using SPSS version 20. Frequency, percentage, mean and standard deviation were used to describe demographic and clinic data. The mean and standard deviation were calculated to explore partners' and patients' FCR and FHI and SSRS scores. Pearson correlations were used to measure the relationship HU ET AL. between the variables and FCR. We performed multivariate regression analysis on the variables that were significantly associated with FCR (p < 0.05).

| Demographics, clinical characteristics and FCR
The ages of the 127 patients ranged from 28 to 80 years, with a mean age of 58.09 years (SD ¼ 9.52). Of the patients, 61.4% (n ¼ 54) were male, and 25.2% (n ¼ 32) had graduated college or above. Of the patients, 41.8% (n ¼ 53) lived in a village and worked in agriculture  Table 1. Age of participant, time since diagnosis, comorbidity, household income and self-reported need for psychological consultation were statistically significantly associated with participants· FCR levels (p < 0.05; Table 1).

| Correlations between family factors and participants' FCR
Pearson correlation was computed to evaluate whether there were relationships between patients' FCR, FHI and SSRS scores and partners' FCR. The results indicated that patients' FCR was highly correlated with that of their partners, while higher family hardiness and social support were significantly negatively related to the patients' FCR (see Table 3). FHI was significantly associated with SSRS.
Partners' FCR was independent of FHI and SSRS.

| Predictive and protective factors of participants' FCR
Multivariate linear regression analysis was conducted to identify the variance in FCR accounted for by demographic, clinical and family factors and is presented in Tables 1-3. Based on the significant results between candidate predictors and FCR, these variables were validated in multivariate linear regression (see Table 4 Addressing the issue of whether FCR occurs differently according to cancer type, it has been suggested that having skin, colon or haematological cancer might predict FCR. 17 The occurrence and intensity of FCR are greatly impacted by perceived controllability, a consequence of recurrence, treatment modality and the time-course of cancer. 18 Multiple myeloma, as an incurable haematological malignancy, 19 has several characteristics that can cause fear. A relatively longer survival period than other advanced stage cancers can result in a persistent confrontation with potential progression.
Asymptomatic relapse or progression before symptomatic recurrence can mean recurrence might be invisible. Bone pain, which is the most common, obvious symptom, can contribute to depression and anxiety. Novel agents that compose the standard regimen are still expensive for the majority of Chinese patients, and recurrence might mean a continually updated and/or extended use of novel agents.
According to a survey on patients' lived experience of myeloma, 20 participants had a distinctive experience in living with this form of cancer, and their fears need to be addressed. In our cohort of 127 myeloma patients, the mean FCR score was 35.05 (SD ¼ 11.09), with 56.4% classified as high-level or maladaptive FCR. In line with other studies conducted in China on participants with breast cancer 21 and gynaecologic cancer, 22 FCR was highly prevalent in myeloma survivors. FCR, to a certain degree, is a natural concern due to a real threat, and it may even be adaptive. Moderate FCR may increase 178 -T A B L E 1 Demographic characteristics and comparisons of FCR among subgroups  24 Severe FCR has a negative impact on survival in the lymphoma population. 25 The present study highlighted the differential need of the myeloma population and provided a wealth of data to inform the planning and implementation of targeted interventions.
Concerns over recurrence may also affect spouses and care- givers. This study corroborated the findings of previous studies that FCR is not restricted to cancer survivors but also affects partners. 8 For instance, Marieke and colleagues 26  This study failed to find a statistically significant impact of time since diagnosis on FCR. We found that patients within the first year of diagnosis reported higher levels of FCR than those who had been diagnosed 1-2 years earlier. However, in patients 3 years after diagnosis, FCR levels were higher than those who had been diag- This study indicates that FHI scores are negatively related to FCR. The FHI was developed to measure family stress resistance and adaption resources, which refer to a family's ability to work together, confidence in handling problems, approach and attitude to new experiences, and sense of being in control of family life. 31 The more resilient a patient's family is, the less fear the patient feels toward cancer recurrence. Evidence of the alleviation of family hardiness on the stress response of family members has been supported in previous studies. 32, 33 Walsh 34 suggested that the capacity to handle problems as a family is crucial to facing a crisis. A family's internal strengths and resilience augment and contribute to the entire family's appraisal and sense of meaning. Family hardiness serves as an important resource in predicting a patient's appraisal of illness. 35,36 The entire family's maintenance of a positive, optimistic attitude might boost a patient's confidence and, thus, mitigate his FCR.
The second resource, social support, had a significant and direct negative effect on patients' FCR. Other studies have also shown that patients who report more social support have a lower level of FCR. 37 Support received from family members, friends and health professionals might reduce patients' stress, improve their confidence and compliance, and help their rehabilitation. 38 As an available external resource, 39,40 social support plays an important role in keeping family functions in balance.

| CONCLUSIONS
Most multiple myeloma survivors completing conventional treatments report fear of recurrence. Several demographic and medical factors are helpful in predicting FCR. To mitigate FCR, partner factors, family hardiness and social support should be addressed during rehabilitation and follow-up care.