Health-related quality of life in children with chronic myeloid leukemia in the chronic phase

This study aimed to explore the health-related quality of life (HRQoL) and associated variables in children with chronic myeloid leukemia in the chronic phase (CML-CP) receiving tyrosine kinase inhibitors (TKIs). A cross-sectional questionnaire was given to children with CML and their parents, who were < 18 years at diagnosis of CML and < 19 years at study. The questionnaire comprised three parts, including demographic information, clinical information, and the Chinese version of Pediatric Quality of Life Inventory™ (PedsQL™) Cancer Module 3.0 as HRQoL questionnaire. A total of 240 respondents data were analyzed. Multivariate analysis showed that children with symptoms had worse pain (− 10.2; P < 0.001), nausea (− 17.3; P = 0.001), more treatment anxiety (− 7.2; P = 0.005), worse self-assessment appearance (− 7.1; P = 0.001), communication problems (− 6.4; P = 0.001), and worse HRQoL (− 7.0; P < 0.001). Children with mothers having low educational qualifications had worse pain (− 6.0; P = 0.014), more worried about future (− 5.4; P = 0.042), worse cognition problems (− 7.1; P = 0.002), worse communication problems (− 5.5; P = 0.008), and worse HRQoL (− 4.3; P = 0.005). Younger age children at study had more procedural anxiety (2.7; P = 0.001), treatment anxiety (− 1.7; P = 0.014) and cognition problem (3.6; P < 0.001), as well as worse HRQoL (1.8; P = 0.008). However, older age children at diagnosis were more worried about future (− 2.8; P = 0.001), worse self-assessment appearance (− 1.1; P = 0.042) and worse HRQoL (− 1.8; P = 0.007). Other variables significantly associated with worse HRQoL included female gender, rural household registration and their father’s low education level. Parents reported more gastrointestinal disorders, were worried about the future and had less concern about appearance than their children. Female gender, older age at diagnosis, younger age at study, lower mother’s education level, and TKI-related symptoms are significantly associated with worse HRQoL in Children with CML. Children and their parents have different priorities in the HRQoL.


Introduction
Chronic myeloid leukemia (CML) is very rare in children, accounting for 2 ~ 3% of all pediatric leukemia cases. The incidence rate of children with CML is 1/1,000,000 per year worldwide (Suttorp et al. 2021). Tyrosine kinase inhibitors (TKIs) have significantly saved the lives of children suffering from CML over the past 2 decades (Suttorp et al. 2021;Fuente et al. 2014). In addition to leukemia control, a key treatment goal is to improve children's health-related quality of life (HRQoL) (Efficace et al. 2011). Many studies showed that children with cancer had worse HRQoL than Fangyuan Zheng and Xuelin Dou contributed equally to this article. healthy children. Female gender, living alone, severity of the malignant disease, aggressive treatment, comorbidities, and poor response to treatment negatively impacted HRQoL in childhood cancer survivors (Sedmak et al. 2020;Corella et al. 2019;Duran et al. 2020;Kowaluk et al. 2019;Fardell et al. 2017;Fakhry et al. 2013). Moreover, the high education qualification of the patients and good communication with their parents were related to better HRQoL (Pan et al. 2017;Yeh et al. 2016;Vlachioti et al. 2016). Unlike other cancer patients, children with CML in the chronic phase (CML-CP) mainly receive long-term oral TKIs as outpatients, enable them to participate normally in family and school life; therefore, their HRQoL and the associated variables might be different. In adults with CML, female gender, older age, lower education level, co-morbidities, concomitant medication, moderate or severe symptoms and higher annual out-of-pocket expense of TKI were significantly associated with worse HRQoL (Unnikrishnan et al. 2017;Breccia et al. 2016;Yu et al. 2019;Flynn and Atallah 2016;Efficace and Cannella 2016;Jiang et al. 2017). Since children with CML are actively growing during TKI treatment and may experience unique adverse effects such as growth abnormalities (Hijiya and Suttorp 2019), possibly variables associated with their HRQoL will be different from those in adults with CML. However, there are rare data on HRQoL in children with CML. Therefore, a cross-sectional study was designed to assess HRQoL in children with CML-CP receiving TKI-therapy and explore the associated variables.

Survey
From July 2018 to December 2019, a cross-sectional study was conducted in 43 hospitals in China. A Chinese-language electronic or hard copy questionnaires were distributed nationwide to children with CML < 18 years at diagnosis and < 19 years at the time of study and their parents via WeChat, through patient advocacy meetings or at the out-patient clinic at Peking University People's Hospital, respectively. CML diagnosis, disease phase, monitoring, and response assessment to TKI-therapy were carried out according to the European LeukemiaNet 2013 recommendations (Baccarani et al. 2013). A complete cytogenetic response (CCyR) was defined as 0% Ph + metaphases and > 20 metaphases in a bone marrow sample. A BCR-ABL1 expression of ≤ 0.1% was defined as a major molecular response (MMR). The study protocol was approved by the Ethics Committee of Peking University People's Hospital. Patients' guardians who agreed to participate in the study provided informed consent.

Questionnaire
The questionnaire is comprised of three parts. The first part included 9 questions on demographic information such as gender, age at diagnosis and study, household registration (urban or rural), educational qualification of their parents, and hospital information. The second part included 16 questions assessing disease-and therapy-related data, including date and disease phase at diagnosis of CML, TKI use, TKI-therapy duration, response to therapy, and patient self-reported discomforts. The third part was the HRQoL questionnaire. We adopted the Chinese version of Pediatric Quality of Life InventoryTM (PedsQL ™ ) Cancer Module 3.0, which has shown great reliability and validity in Chinese children (Ji et al. 2011). The HRQoL questionnaire is comprised of two sections: one is a child self-report for children by current age (5-7 years, 8-12 years and 13-18 years, respectively); the other one is a parent proxy-report for parents by the age of their children (2-4 years, 5-7 years, 8-12 years and 13-18 years, respectively). Each section has eight dimensions: D1 (pain and hurt), D2 (nausea), D3 (procedural anxiety), D4 (treatment anxiety), D5 (worry), D6 (cognition problems), D7 (perceived physical appearance), and D8 (communication problems). Each dimension contains 2-5 items. Each item has a five-point scale ranging from 'never' to 'almost always'. Scale scores are computed as the sum divided by the number of items answered, ranging from 0 to 100. The higher the score, the better the HRQoL.

Statistical analysis
Data were represented using mean, mean ± standard deviation, or median (range) for continuous variables, frequencies and proportions were used for categorical variables. The median was used to determine the boundary value of continuous variables of the age at diagnosis and duration of TKI-therapy. Pearson correlation analysis were performed to assess the correlation of the score of each dimension between child self-report and parent proxy-report. The paired sample t test was used for the score of each dimension between child self-report and parent proxy-report. Demographic and clinical variables were included in univariate analysis. Variables associated at a level of P < 0.2 in univariate analysis were included in a multivariate linear regression model. P < 0.05 was considered statistically significant. All the statistical analysis were conducted using SPSS v19.0 (IBM Corporation, New York, USA).

Results
Subjects 105 hard copy questionnaires were delivered with a response rate of 100%, and 184 electronic questionnaires were collected. Finally, data of 240 questionnaires were included in this study with the exclusion of those in the advanced phase (N = 6), with no data of CML diagnosis (N = 8) or therapy information (N = 12), not completing the HRQoL questionnaire (N = 20), or having comorbidity(ies) (such as mental retardation, cardiac insufficiency or disability) (N = 3). A flowchart of patient enrollment is shown in Fig. 1.

HRQoL profile
Among the 240 respondents, 7 younger than 5 years were unable to complete a child self-report on HRQoL, 55 at the age of 5-18 years failed to complete the HRQoL questionnaire. A total of 178 child self-reports and 240 parent proxyreports were included in the HRQoL analysis ( Fig. 1).
Based on the child and parent self-reports, the HRQoL scores were 80.9 ± 10.7 and 79.1 ± 12.3, respectively. The score of the 8th dimension (communication problems) was the highest, with a mean score of > 92, while that of the 3rd dimension (procedural anxiety) was the lowest, with a mean score of < 60. HRQoL profile is presented in Table 2.
In the 178 respondents with both child self-report and parent proxy-report, the scores of child self-report were higher than those of parent proxy-report in dimensions of nausea and worry (P < 0.001). The score of child self-report was higher than that of parent proxy-report in the dimension of perceived physical appearance (P < 0.001). There was no significant difference in total scores and other dimensions. However, we found an inconsistency of the scores between the child self-report and parent proxy-report (Fig. 3), and an intermediate or high correlation (r = 0.576 ~ 0.934, P < 0.001) between each dimension in the 178 respondents.

Discussion
To our knowledge, this is the first study to assess HRQoL in children with CML-CP and explore the associated variables. We found some variables, including female gender, age, and TKI-related symptoms that were associated with HRQoL, consistent with the previous studies about HRQoL of children with cancers (Sedmak et al. 2020;Duran et al. 2020;Kowaluk et al. 2019;Fardell et al. 2017;Fakhry et al. 2013). In addition, we found that the lower the mother's education level, the worse HRQoL the children had, and parents reported more gastrointestinal disorders and worries about the future and less concerns about the appearance than children's reports.
Age at the time of diagnosis was related to the HRQoL of childhood cancer survivors, but the results are different. Fradell et al. reported that HRQoL of survivors with acute lymphoblastic leukemia at an older age was worse (Fardell et al. 2017), while Alessi reported that HRQoL of survivors with malignant tumors at an older age was better than those diagnosed at a younger age (Alessi et al. 2007). In our study, children who were older at the time of diagnosis were more likely to be worried about the future and dissatisfied with their appearance, often complained about skin rash and edema, and had worse HRQoL. In addition, the younger age of the children at study was associated with the worse HRQoL, which was mainly manifested for procedural anxiety, treatment anxiety and cognitive problems. It is easy to understand that the frequent hospital visits and the need for invasive procedures (such as blood tests every month in the first 3 months and every 3 months after that, and bone marrow puncture every 3 months before CCyR) may make the younger children feel scared. Moreover, younger children may experience more difficulties in cognitive problems. Mulder SF reported that vascular endothelial growth factor receptor tyrosine kinase inhibitors (VEGFR TKI) had a negative impact on cognitive function, specifically on learning and memory, and executive function (Mulder et al. 2014). There was no data on the effect of imatinib or dasatinib on patient's cognitive function, especially in children during nervous system development.
We found that mother's educational levels were associated with the worse HRQoL of their children. This result was similar to Rosalie Power's finding on the HRQoL of Fig. 2 The self-reported discomforts of the patient and the proportion in the 240 respondents. D1-D8: dimension 1 to dimension 8. D1: pain and hurt; D2: nausea; D3: procedural anxiety; D4: treatment anxiety; D5: worry; D6: cognition problems; D7: perceived physical appearance; D8: communication problems adolescents with cerebral palsy in Bangladesh (Power et al. 2020). The lower the education level of the father and mother, the more it affects the children's communication, similarly to Neeraj Kaushal' report that the higher the parents' education level, the more well-being and the more helpful the children' recovery from chronic health conditions (Kaushal 2014).
In our study, the first three most common adverse effects reported were gastrointestinal disorders, skeletal muscle pain and fatigue. In addition, the children complaining a discomfort of TKI-therapy reported more treatment anxiety, dissatisfaction with appearance, and impaired communication problems. The worse HRQoL associated with TKI-related symptoms is consistent in adult patients with CML (Unnikrishnan et al. 2017;Yu et al. 2019).
Although there is a strong correlation between children self-report and parent proxy-report, we found that there were inconsistencies between them. Children reported more significant self-negation of appearance than their parents, whereas parents reported worse HRQoL in terms of gastrointestinal disorders and worried about the future. That is to say, the children may pay more attention to their present conditions and feelings, like appearance, caused by skin rash, change of skin color, short stature due to the TKI-therapy than their parents. While the basic concern of all parents was the health and future of their children. The life experiences and inner lives of parents might be different from those of the children.
There are some limitations to our study. First, respondents needed to be old enough to complete the self-report part independently. Some parents refused their children to answer the questionnaire because of privacy concerns, which significantly reduced the number of child self-report subjects. Second, the patients' self-reported discomfort and degree were subjective feelings with individual differences. Third, parents usually pay different attention to the child's growth, and the mother is more delicate in the relationship, but we did not restrict the selection of father and mother in answering the parent proxy-report.
In conclusion, our study in children with CML showed that female gender, older age at the time of diagnosis, younger age at study, low mother's educational qualification, and TKI-related symptoms had a negative effect on their HRQoL. Moreover, parents did not pay attention to the children's real feelings, which may affect the children's HRQoL. This requires pediatricians not only to provide patients with the most feasible measures, such as the adjustment of type and dose of drugs to reduce adverse effects, but also to pay more attention to psychological counseling for children with CML to carry out disease management.  . 3 The comparison of HRQoL scores of the eight dimensions and total score between child self-report and parent proxy-report in the 178 respondents with both child self-report and parent proxyreport