INTRODUCTION
Breast cancer is the most frequently diagnosed cancer in women, with an estimated incidence of 2.3 million new cases in 2022 and it is the leading cause of cancer-related deaths worldwide in females[1]. Early detection has increased due to screening improvements and survival rates are high in localized breast cancer. However, up to 4-6% of patients will be metastatic from the diagnosis and a non-negligible percentage of patients will have a distant relapse during the course of the disease[2]. Multimodality treatment is crucial in breast cancer, with a combination of surgery, radiotherapy and different systemic therapies depending on the clinical and pathological stage and immunohistochemistry[3,4]. Systemic treatments such as hormonotherapy or chemotherapy are often administered during prolonged periods, even in early stages, and can cause secondary effects continued over time. For these reasons, beyond specific breast cancer treatments, it should not be forgotten the psychological impact secondary to the fear and worry during the diagnosis process and treatments, the uncertainty of the future, body changes, side effects and quality of life alterations[5].
Depression is estimated to be diagnosed in more than 30% of breast cancer patients, with the highest prevalence in the Eastern Mediterranean Region[6]. Besides, depression appears to be more frequent in recently diagnosed patients, in those under treatment, in advanced stages of the disease and it increases with patients’ age[7]. It has also been suggested that family support is a protective factor for major depression unlike pain severity[5]. In the subgroup of younger patients, other concerns must be taken into account. For example, the impact of treatments on their fertility and family worries, the economic burden due to the disease with the discontinuity of their careers or sexuality alterations[8-11]. Furthermore, a recent meta-analysis indicates that depression and anxiety have an influence on cancer recurrence and all-cause mortality in breast cancer patients associated, among other causes, with poorer adherence to treatments[12].
Therefore, appropriate screening tools that ensure an early diagnosis and approach of depressive symptoms are crucial.
A NOVEL NOMOGRAM PREDICTION MODEL
Mao et al[13] recently published a paper identifying factors affecting depressive symptoms in middle-aged and young breast cancer patients and a nomogram prediction model was constructed based on them. This cross-sectional study included 360 breast cancer patients aged between 18 and 60 years from two tertiary hospitals in China. Patients with severe comorbidity, other tumors and those with a history of depression were excluded. For the purpose of the study, demographic data were collected and patients had to complete the Visual Analogue Scale (VAS) and three questionnaires: Patient Health Questionnaire-9 (PHQ-9), Perceived Social Support From Family Scale (PSS-Fa) and International Physical Activity Questionnaire (IPAQ).
The incidence of depressive symptoms was 38.61%. The authors found several characteristics with significant differences between the group of patients with depressive symptoms and those without them, and a multivariable logistic regression model showed that tumour grading, patient's monthly income, pain score, physical activity score, and family support score were significant predictors of depression. Based on these variables, a nomogram prediction model was constructed and showed strong predictive capability with an area under the receiver operating characteristic curve of 0.852, showing considerable accuracy in properly identifying patients at risk of depression. Besides, high sensitivity and specificity values were reported (86.80% and 89.50% respectively).
Screening for depressive symptoms and PHQ-9
The PHQ-9[14] is a depression screening tool with nine items scored from 0-3 that evaluates how often patients have each Diagnostic and Statistical Manual major depression symptom in the past two weeks. It usually has a standard cutoff value of ≥ 10 for major depression. Even though this questionnaire was not designed for oncological patients, the Cronbach's α coefficient was 0.85 in the study of Mao et al[13], showing acceptable reliability. In agreement, Park et al[15] recently published a study demonstrating acceptable validity and reliability of the PHQ-9 in a cohort of 327 breast cancer patients. Hinz et al[16] showed a Cronbach's α coefficient of the PHQ-9 of 0.84 in the cohort of patients with different cancers included in their study. Despite previous results, there is a lack of standardization of its use in breast cancer patients, with little evidence supporting it and further studies are needed in order to corroborate the reliability in this population.
Furthermore, the PHQ-9 is not the only questionnaire utilized for screening of depression in oncological patients. The Hospital Anxiety and Depression Scale (HADS) was described by Zigmond et al[17] in 1983 and has also been widely used in cancer patients. It consists of a questionnaire of 14 items to assess anxiety and depression by two subscales (HADS-A for anxiety and HADS-D for depression) and was initially validated as a reliable screening tool for clinically significant anxiety and depression in patients attending a general medical clinic. Vodermaier et al[18] published a meta-analysis suggesting specifical cut-off numbers as did Annunziata et al[19] in their study published in 2020 suggesting values for the HADS-A and HADS-D in cancer patients. This questionnaire has also been used in breast cancer studies. The Patient-Reported Outcome Measures[20], a sub-study of the IMPORT LOW, used the HADS as one of the questionnaires included and showed that women at risk of anxiety and depression assessed by this questionnaire were more likely to report adverse effects.
As different screening questionnaires for depressive symptoms are used in clinical practice and studies, it seems relevant to compare the reliability of the main depression screening tools in breast cancer patients in order to determine which is the most suitable in this context before standardizing its use.
Physical activity and the IPAQ
Physical activity has been increasingly recognized as an important factor in managing depression in breast cancer patients, as regular activity can alleviate depressive symptoms. In this sense, Perna et al[21] accomplished a randomized controlled trial, demonstrating that an individualized structured exercise intervention reduced depression in breast cancer patients. Rogers et al[22] performed a multicenter randomized study in which breast cancer survivors received, after primary treatment, a multicomponent physical activity intervention [Better Exercise Adherence after Treatment for Cancer (BEAT Cancer)] or usual care. Depressive symptoms were evaluated using the HADS questionnaire and the study concluded that BEAT Cancer reduced anxiety and depressive symptomatology up to three months post-intervention. Mostafaei et al[23] showed similar benefits of exercise in the group of breast cancer patients undergoing chemotherapy. These results are in concordance with those in the meta-analysis conducted by Patsou et al[24] which included 14 randomized controlled trials. The authors evaluated the effect of different types of interventions on depressive symptoms and identified improvements in those breast cancer patients undergoing exercise. For all these reasons it is of greater importance that Mao et al[13] included an exercise measurement tool in the study.
The IPAQ is useful for monitoring physical activity and inactivity and has shown acceptable reliability and validity in a group of age similar to that of the study of Mao et al[25] (between 18 and 69 years)[25]. This questionnaire includes 27 items in its long version and 7 items in the short one. However, despite the good reliability, the IPAQ was not designed initially for cancer patients and it should be mentioned that, for example, 7 items are related to work. Patients undergoing cancer treatments may have a job but be on sick leave and during acute interventions the answers about the transportation items can also vary so, different interpretation of the questions between patients may have incur biases.
Social support and the PSS-Fa
Perceived social support, specially from family, should be important during an oncological process, considering the potential of mitigating psychological distress and enhancing quality of life. However, there is little evidence of the real impact in breast cancer patients. Mao et al[13] included a family support evaluation in the study assessed by the PSS-Fa and it was found, in the multivariable logistic regression, a significant correlation of depression with family support score. These results are in concordance with those reported by Shen et al[26] in breast cancer patients or Jeong and An[27] in gastric cancer patients even though the screening tools for family support were heterogenous. Interestingly, Katapodi et al[28] remarked that depressive symptoms in young breast cancer patients can negatively influence the perceived support received from relatives, showing a reciprocal relationship in which depression may also affect their families.
The PSS-Fa was described by Procidano and Heller[29] in 1983. It originally consisted of 20 questions with a “Yes”, “No” or “Don´t know” answer. The authors also developed another scale called Perceived Social Support From Friends (PSS-Fr). The PSS-Fa originally demonstrated lower rates of psychopathology and distress when the results of the scale were higher. It should be mentioned that the PSS-Fa was not designed for oncological patients and, as mentioned, there is little evidence in this population. Carpenter et al[30] designed a cross-sectional study with the aim of determining the influence of social support on the relation of physical and psychological outcomes in gynaecological patients. The PSS-Fa, the PSS-Fr and the Interpersonal Support Evaluations List subscales were the questionnaires used to evaluate social support. The study found an inverse correlation between perceived social support and depressive symptoms.
A standardized implementation of patients’ social support when evaluating depression in oncology seems crucial, but the available scales are not originally designed nor validated yet in this population and further investigation is needed.
Pain evaluation and the VAS
The VAS assesses the severity of pain scoring from 0 (no pain) to 10 (the most intense pain). The multivariate logistic regression model in the study of Mao et al[13] showed significant correlation of pain score and depression utilizing this tool.
It seems essential to include pain evaluation when creating a nomogram for depression in oncological patients as the relationship between cancer pain severity and its duration with the risk of depression has been highlighted by several studies. Laird et al[31] conducted a systematic review including 14 studies with various types of cancer in which patients’ depression and pain were assessed. A statistically significant association was found between pain and depression as well as a relationship with pain intensity. The cross-sectional study by Fitzgerald et al[32] embraced 487 patients with different advanced tumors, including stage IV breast cancer. The authors evaluated the relationship between physical symptoms measured by the Memorial Symptom Assessment Scale and the risk of depression. The prevalence of pain was 55% and it was significantly correlated with depression. In accordance, Khemiri et al[33] accomplished a cross-sectional study evaluating the relationship of pain by VAS and the presence of anxiety and depression by HADS. A correlation was found between depression and pain intensity, the need for opioids and an incomplete response to analgesics. Unseld et al[34] also demonstrated a correlation between depression and pain when evaluated by the VAS in oncological patients.
Limitations of the study
Mao et al[13] show promising outcomes in this study. However, some limitations should be noticed. As mentioned before, most of the questionnaires used in the study were not designed for oncological patients and the suitability of some of them for breast cancer may be uncertain. The cross-sectional design makes it difficult to establish clear causal relationships between risk factors and the presence of depressive symptoms. The sample size is relatively small for breast cancer (360 patients). The sensitivity was 86.80% and the specificity was 89.50% but the authors propose a maximum Jordan index of 0.973 for the model which might be overestimated. It should also be mentioned that the odds ratio of 0.552 is not concordant with the model’s reported good predictive performance. Besides, one of the major limitations is the fact that the study includes patients only from two hospitals in Jiangsu Province, China. Depression is influenced by the socio-cultural and economical context which can vary between regions and countries. For this reason, generalizing the results to other populations may be incorrect and further investigation that corroborates the applicability in different contexts and with larger samples is needed.
CONCLUSION
The significant prevalence of depressive symptoms in breast cancer patients makes it necessary to implement early detection tools as part of a comprehensive management. The nomogram prediction model of Mao et al[13] shows strong predictive capability and high sensitivity and specificity values. It also demonstrates significant correlation of family support, patient’s economy, pain and physical activity with the risk of depression. The results are in concordance with those available in the literature. However, different screening tools for depression are currently available and it has not been determined which is the most suitable in breast cancer patients. The tests used to assess depression, physical activity or family support were not originally designed for oncology and, even though there is some evidence supporting their use, further investigation to validate them is advisable. Finally, larger samples of patients from different socio-cultural contexts should be included in order to generalize the results.
