Mental Health in Persons With Chronic Myeloid Leukemia During the SARS-CoV-2 Pandemic: The Need for Increased Access to Health Care Services

Mental health problems in the general population have been reported during the SARS-CoV-2 pandemic; however, there were rare data in persons with chronic myeloid leukemia (CML). Therefore, we performed a cross-sectional study on mental health evaluated using the 9-item Patient Health Questionnaire (PHQ-9; depression), the 7-item Generalized Anxiety Disorder (GAD-7; anxiety), and the 22-item Impact of Event Scale—Revised (IES-R; distress), including subscales of avoidance, intrusion, and hyper-arousal in persons with CML, non-cancer persons, and immediate family members of persons with cancer as controls (≥16 years) by an online survey. Data from 3,197 persons with CML and 7,256 controls were collected. In multivariate analyses, CML was significantly associated with moderate to severe depression (OR = 1.6; 95% Confidence Interval [CI], 1.4, 1.9; p < 0.001), anxiety (OR = 1.4 [1.1, 1.7]; p = 0.001), distress (OR = 1.3 [1.1, 1.5]; p < 0.001), and hyper-arousal (OR = 1.5 [1.3, 1.6]; p < 0.001). Moreover, delay in regular monitoring was significantly associated with depression (OR 1.3 [1.0, 1.7]; p = 0.024), anxiety (OR = 1.3 [1.0, 1.8]; p = 0.044), avoidance (OR = 1.2 [1.0, 1.4]; p = 0.017), and intrusion (OR = 1.2 [1.0, 1.4]; p = 0.057); tyrosine kinase-inhibitor dose reduction or discontinuation, depression (OR = 1.9 [1.3, 2.8]; p = 0.001), distress (OR = 2.0 [1.4, 2.8]; p < 0.001), avoidance (OR = 1.6 [1.2, 2.1]; p = 0.004), intrusion (OR = 1.6 [1.1, 2.1]; p = 0.006), and hyper-arousal (OR = 1.3 [1.0, 1.8]; p = 0.088). We concluded that persons with CML during the SARS-CoV-2 pandemic have worse mental health including depression, anxiety, and distress symptoms. Decreasing or stopping monitoring or dose resulted in adverse mental health consequences.

The mental health of persons with CML is of concern because they require long-term tyrosine kinase-inhibitor (TKI) therapy response monitoring and TKI dose adjustments (22). Data from before the pandemic indicate that persons with CML experience more severe depression and/or anxiety to daily life challenges compared with controls (23)(24)(25)(26)(27). Several studies reported prevalence and/or severity of SARS-CoV-2-infection and of coronavirus infectious disease 2019  in persons with CML (28-31) compared with controls. There are no data on the mental health of persons with CML during the SARS-CoV-2 pandemic. Therefore, we performed a crosssectional survey to explore the prevalence of depression, anxiety, and distress symptoms in 3,197 subjects with CML during the pandemic. Results were compared with concurrent controls, which were non-cancer persons and immediate family members of persons with cancer including CML. We were especially interested in the effects on mental health of adjustments in the frequency of monitoring response to TKI therapy and/or dose.

Study Design
From April 28 to May 12, 2020, a cross-sectional online survey was done using the WeChat-based survey program Wenjuanwang. Persons diagnosed with CML, non-cancer persons, and immediate family members of persons with cancer including CML as controls (≥ 16 years) in China were invited to participate in the survey. Persons diagnosed with CML after February 2020 and those with COVID-19 were excluded. The studies involving human participants were reviewed and approved by the Ethics Committee of Peking University People's Hospital who waived the requirement for informed consent because of the outbreak of SARS-CoV-2.

Questionnaire
The questionnaire consisted of five dimensions (Supplementary Contents 1, 2). The first dimension included a brief introduction of the survey and a question about whether having filled out the questionnaire before. The second dimension included 10 questions assessing demographics. The third included 15 questions about CML-related data such as diagnosis, therapy, and response. The fourth included 11 questions about behavior and experience at the peak of the COVID-19 pandemic in China, the end of January to end of March (32)(33)(34). The fifth included three self-reported validated measurement tools in Chinese version including the 9-item Patient Health Questionnaire (PHQ-9), the 7-item Generalized Anxiety Disorder (GAD-7), and the 22-item Impact of Event Scale-Revised (IES-R) to assess the prevalence and severity of depression, anxiety, and distress during the end of January to April. The questionnaire was the same for age and subjects with CML except for CML-related questions (the third dimension). Comorbidities were defined as coexisting diseases except CML and COVID-19.

Statistical Analyses
Descriptive analysis results were presented as median (range) or number (percent). Categorical variables were compared with the chi-square test and continuous variables with the Mann-Whitney test. The k-means clustering method was used to cluster the three subscales of distress including avoidance, intrusion, and hyper-arousal symptoms derived from the IES-R. To identify the variables associated with the  poor mental health in responses with CML and controls, covariates including age, sex, education level, marital status, household registration (rural vs. urban), comorbidity(ies), residence in Hubei province or elsewhere, cohabitation, behavior and experience (such as following pandemic information frequently, sharing feelings actively), and having an acute respiratory symptom during the pandemic were used to analyze the association with the moderate to severe depression, anxiety, and distress, as well as the presence of avoidance, intrusion, and hyper-arousal by the k-means clustering analysis. To identify the covariates associated with the poor mental health in persons with CML, CML-related data including disease phase, CML and TKI therapy duration, TKI used, TKI therapy line, treatment response, interruption or delay in disease monitoring, and interruption or dose reduction of TKI therapy were used to analyze. Covariates with p < 0.2 in univariate analysis were included in the multivariate binary logistic regression analyses. P < 0.05 was considered statistically significant. Analyses were conducted with SPSS Version 22.0 software.

RESULTS
From April 28 to May 12, 2020, data of 3,581 respondents with CML and 7,556 controls from 32 provinces and municipalities across China were collected. Questionnaires from subjects < 16 years (n = 121), duplicates (n = 513), subjects with CML diagnosed after February 2020 (n = 42), and persons with COVID-19 (n = 8) were excluded. The updated dataset consisted of 3,197 respondents with CML and 7,256 controls.

Comparison of Mental Health Between the Respondents With CML and Controls
Data from respondents with CML indicated a higher prevalence of depression (PHQ-9 score ≥ 5, 33%, 95% confidence interval [CI], [ [19, 21%], p < 0.001) by k-means clustering analyses (Figure 1, Table 2 ; p < 0.001). In addition, having comorbidity(ies), following the pandemic information frequently, and having an acute respiratory symptom were significantly associated with moderate to severe depression, anxiety, distress, avoidance, intrusion, and hyper-arousal. Female sex, increasing age, unmarried, divorced, or widowed marital status, and living in Hubei province were significantly associated with ≥1 symptom of mental health described above; however, for those who were sharing feelings actively, there was no or mild depression ( Table 3).
We collected data from subjects with CML on the disease phase at diagnosis, with the interval from diagnosis to starting TKI therapy, CML duration, TKI therapy duration, current therapy, the number of TKI therapy lines given, response (MMR vs. <MMR), delay in regular monitoring, and tyrosine kinase-inhibitor dose reduction or discontinuation. Univariate analysis results of CML persons and controls are shown in Supplementary Tables 2, 3; in multivariate analyses, delay in monitoring response to TKI therapy was significantly associated with reporting moderate to severe depression (OR = 1. . CML disease phase, TKI, TKI therapy duration, and therapy response were not significantly associated with mental health. Other covariates associated with mental health in subjects with CML were similar to those in controls (Tables 4, 5).

Comparison of Mental Health Between Low-and High-Risk Respondents With CML and Controls
Next we categorized CML respondents into low-risk (no risk covariate; n = 1,664, 52%), intermediate-risk (1 risk; n = 1,412, 44%), or high-risk (2 risks, n = 121, 4%) cohorts based on delay in monitoring response to TKI therapy, TKI therapy interruption, or dose reduction or both during the pandemic. There were significant differences in mental health among the three CML risk cohorts or between the low-risk CML cohort and controls. The high-risk CML cohort had the highest prevalence and most severe depression (PHQ-9 score ≥ 5, 47% [      ( Figure 2, Table 2

DISCUSSION
The high-and intermediate-risk CML cohorts defined by harboring both or either of the two CML-related risk covariates including delay in monitoring response to TKI therapy and TKI therapy interruption or dose reduction during the pandemic had significantly higher prevalence of depression, anxiety, distress, avoidance, intrusion, and hyper-arousal compared with lowrisk CML cohort harboring no-risk covariate. Even in the lowrisk CML cohort, hyper-arousal was more common compared with controls. These data suggest that persons with CML are psychologically vulnerable during the SARS-CoV-2 pandemic. Delay in monitoring response to TKI therapy, TKI therapy interruption, or dose reduction in persons with CML was associated with worse mental health in our study, consistent with the recent findings that treatment interruption, delay in cancer care, or reduced therapy intensity was associated with mental health problems and worse HRQoL in persons with cancer or lymphoma (12,13,(16)(17)(18)(40)(41)(42)(43). Fear of being infected with SARS-CoV-2 in the hospital or during travel as the common reason causing them not to follow the regular monitoring or cannot get TKI drugs in the hospital also reflected that they exaggerated the implementation of containment measures for avoidance of SARS-CoV-2 infection due to their psychological fragility.
In our study, having CML was significantly associated with moderate to severe depression, anxiety, distress, and hyperarousal. Even in the low-risk CML cohort, hyper-arousal was more common compared with controls. Although fewer respondents with CML reported being exposed to someone with COVID-19 or a family member experiencing acute respiratory symptoms, more respondents reported that they experienced an acute respiratory symptom and went to a hospital and had a lung CT scan to evaluate COVID-19 compared with controls. These data suggest that persons with CML are psychologically more vulnerable with trauma-related distressing memories and persistent negative emotions resulting from the pandemic.
During the pandemic, although the likelihood of developing a COVID-19 infection in persons with CML is very low (28)(29)(30)(31), negative impact on different aspects of CML management including TKI therapy response monitoring, TKI therapy, and enrollment in and compliance with clinical trials is also reported (29). These highlighted the importance of adequate access to health care services such as patient education of having appropriate personal self-protection equipment, establishing a safe area in the hospital or clinic, and telemedicine and mailed medicine to avoid monitoring and therapy interruptions 1,2,3 (44)(45)(46)(47)(48)(49)(50).
Our study has several limitations. First, it is cross-sectional. We lacked baseline pre-pandemic data so we cannot be certain that the observed changes in mental health were related to the pandemic. Second, there were potential selection biases in the respondents. It is not possible to assess the participation rate since it is unclear how many subjects received the link for the online survey. Third, it was impossible to confirm the accuracy and authenticity of the information provided by the respondents in the online survey. Last, the survey was conducted when the pandemic in China was mostly controlled, which might result in recall bias.
Our data indicate that persons with CML are more vulnerable than controls to mental health problems. This results in delay in monitoring TKI therapy response and in TKI therapy interruptions and dose reductions. The results of our study may help physicians identify vulnerable persons with CML and help them by increasing access to health care services.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Ethics Committee of Peking University People's Hospital. Written informed consent from the participants' legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements.