Symptom clusters among cancer survivors: what can machine learning techniques tell us?

The association between health-related quality of life and five-year overall survival among head and neck cancer patients: A prospective cohort study

OBJECTIVE

To estimate associations between health-related quality of life (HRQOL) at baseline (T0), six months after treatment (M6), and change in HRQOL (T0-M6) and five-year overall survival (OS) among head and neck cancer (HNC) patients, after adjusting for potential confounders. Furthermore, we explored whether personal, clinical, physical, psychological, social, lifestyle, HNC-related and biological factors moderate the association.

METHODS

Data of a prospective multi-center cohort study (NET-QUBIC) was used. In this specific study, patients with HRQOL data at T0 (n = 596), M6 (n = 489), and T0 and M6 (n = 463) were included. HRQOL was operationalized by the EORTC QLQ-C30 global quality of life subscale (QL) and summary score (SumSc). Cox regression analyses investigated associations between HRQOL and OS, adjusted for confounders, and explored which variables moderate the association.

RESULTS

Adjusted models showed that higher baseline QL (HR: 0.85 (95% CI: 0.76-0.96)) and SumSc (HR: 0.90 (95% CI: 0.81-0.99)) were associated with longer OS. Adjusted M6 models and adjusted T0-M6 models found no such association. The association between QL and OS was moderated by sex (significant among males) and mean arterial blood pressure (BP) (significant for patients with high BP). The association between SumSc and OS was moderated by coping (significant for patients with no avoidant coping style) and systemic BP (significant for patients with normal BP).

CONCLUSION

Higher HRQOL at baseline (how patients enter the cancer trajectory) was associated with longer OS, but (change in) HRQOL at 6 months (how they overcome cancer treatment) was not. This knowledge is important to personalize treatment plans.

Psychometric Properties of the Depression Anxiety Stress Scales (DASS-42 and DASS-21) in Patients with Hematologic Malignancies

Background/Objectives: Patients with hematologic malignancies undergo prolonged, intensive treatments involving frequent hospitalizations and experience debilitating side effects. Consequently, they are at increased risk of developing symptoms of depression, anxiety, and stress, which can undermine their quality of life. However, there is a scarcity of instruments validated for the simultaneous assessment of depression, anxiety, and stress within hematologic cancer populations. The aim of this study is to examine the construct validity, convergent and discriminant validity, and reliability of the Depression Anxiety Stress Scales (DASS-42 and DASS-21) among hematologic cancer patients. Methods: We collected data from 452 hematologic cancer patients across three studies. Results: Confirmatory factor analyses indicated that the theoretical correlated three-factor model and bifactor model for DASS-42 and DASS-21 responses were sufficient to explain the underlying factor structure of the scales in hematologic cancer patients. However, the bifactor model for DASS-42 and DASS-21 fit better with the data compared to the theoretical correlated three-factor model. In addition, we found the correlated three-factor model and the bifactor structure to exhibit scalar measurement invariance across gender for DASS-42 and DASS-21. DASS-42 and DASS-21 subscales demonstrated weak to strong negative correlations with measures of psychological well-being (happiness, well-being, life satisfaction) and strong positive correlations with measures of similar constructs (depression, anxiety, and stress), thereby supporting their convergent and discriminant validity with theoretically and empirically expected correlations with external criteria. The reliability analyses demonstrated that both DASS-42 and DASS-21 subscales exhibited strong internal consistency and test-retest reliability when assessing symptoms of depression, anxiety, and stress among patients with hematologic malignancies. Moreover, the item-scale convergent and discriminant validity analyses demonstrated that items exhibited higher corrected item-total correlations with their intended subscales than with other subscales across the DASS-42 and DASS-21, providing evidence for the distinct measurement properties of each subscale. Conclusions: The findings suggest that the DASS-42 and DASS-21 are psychometrically robust instruments for use in Turkish hematologic cancer patients.

Heterogeneity of pain-fatigue-sleep disturbance symptom clusters in lung cancer patients after chemotherapy: a latent profile analysis

PURPOSE

Pain-fatigue-sleep disturbance symptom cluster (PFS) was common in patients with lung cancer and seriously affected the life quality of patients. However, the heterogeneity and subgroups of PFS were unclear in lung cancer patients after chemotherapy. This study was conducted to identify distinct subgroups of PFS in patients with lung cancer after chemotherapy, and explore the differences and risk factors of PFS subgroups.

METHODS

Lung cancer patients after chemotherapy were recruited. Data were collected using the Chinese version of the Brief Pain Inventory, the Cancer Fatigue Scale, and the Pittsburgh Sleep Quality Index. The latent profile analysis (LPA) was used to identify the subgroups of PFS. Univariate analysis was used to identify the differences among all subgroups. Logistic regression and restricted cubic splines were used to investigate predictors of the PFS subgroups.

RESULTS

Based on LPA, 512 participants were divided into four subgroups (Class 1: low pain-fatigue-sleep disturbance; Class 2: moderate pain-moderate fatigue-low sleep disturbance, Class 3: low pain-high fatigue-high sleep disturbance, and Class 4: high pain-fatigue-sleep disturbance). The univariate analysis showed that gender, body mass index (BMI), Eastern Cooperative Oncology Group (ECOG) performance status, leukocyte, neutrophils, platelet, C-reactive protein, stress, anxiety, depression, and social support were associated with PFS. The logistic regression analysis revealed that patients in Class 2 and Class 3 were more likely to experience great stress than those in Class 1. Additionally, compared to Class 1, females, lower BMI, stress, anxiety, and depression were independent predictors of Class 4.

CONCLUSION

This study successfully identified subgroups of PFS in patients with lung cancer after chemotherapy. Based on the results of this study, medical workers can identify patients with high risks for PFS and conduct more targeted interventions to improve symptom management.

Latent profile analysis for assessing symptom clusters in women with breast cancer

PURPOSE

To identify symptom clusters among breast cancer survivors and investigate differences in health-related quality of life (HRQoL) and distress based on these discerned symptom clusters using latent profile analysis.

METHODS

We enrolled 655 adult breast cancer survivors aged 19 years and older, registered with the Cancer Survivor Integrated Supportive Center from May 2020 to July 2022. The study measured five symptoms-pain, fatigue, insomnia, anxiety, and depression-using a Visual Analogue Scale ranging from 0 to 10 points. Distress was measured using the National Comprehensive Cancer Network Distress Thermometer, with scores ranging from 0 to 10 points. HRQoL was determined using the EuroQol-5 Dimension questionnaire. Data analysis was conducted using the Jamovi and Mplus 8.8 software programs.

RESULTS

The Cluster with Few Symptoms (46.8%) was the most common, whereas the Psychological Cluster with a very high degree of depression and anxiety accounted for 20.0%, and the Moderate symptom cluster with symptoms of 3 or more points accounted for 14.4%. Distress scores were relatively high in the Psychological Cluster and the Pain-Fatigue-Insomnia Cluster, and were lowest in the Cluster with Few Symptoms (F = 103.92, p < 0.001). HRQoL scores were highest in the Cluster with Few Symptoms and lowest in the Pain-Fatigue-Insomnia Cluster (F = 177.62, p < 0.001).

CONCLUSIONS

Approximately half of breast cancer survivors who had completed the major treatment experienced persistent high symptoms such as depression and anxiety or pain, fatigue, and insomnia.

IMPLICATIONS FOR CANCER SURVIVORS

These findings provide foundational data for developing tailored intervention strategies and programs based on symptom experiences.

Factors influencing psychological distress among breast cancer survivors using machine learning techniques

Breast cancer is the most commonly diagnosed cancer among women worldwide. Breast cancer patients experience significant distress relating to their diagnosis and treatment. Managing this distress is critical for improving the lifespan and quality of life of breast cancer survivors. This study aimed to assess the level of distress in breast cancer survivors and analyze the variables that significantly affect distress using machine learning techniques. A survey was conducted with 641 adult breast cancer patients using the National Comprehensive Cancer Network Distress Thermometer tool. Participants identified various factors that caused distress. Five machine learning models were used to predict the classification of patients into mild and severe distress groups. The survey results indicated that 57.7% of the participants experienced severe distress. The top-three best-performing models indicated that depression, dealing with a partner, housing, work/school, and fatigue are the primary indicators. Among the emotional problems, depression, fear, worry, loss of interest in regular activities, and nervousness were determined as significant predictive factors. Therefore, machine learning models can be effectively applied to determine various factors influencing distress in breast cancer patients who have completed primary treatment, thereby identifying breast cancer patients who are vulnerable to distress in clinical settings.

Does medicinal cannabis affect depression, anxiety, and stress in people with cancer? A systematic review and meta-analysis of intervention studies

INTRODUCTION

Medicinal cannabis might have a role in supporting the mental health of people with cancer. This systematic review and meta-analysis examined the efficacy and safety of medicinal cannabis, compared with any control, as an intervention for depression, anxiety, and stress symptoms in people living with cancer. A secondary aim was to examine the effect of low versus high Δ9-tetrahydrocannabinol (THC) dose on these outcomes.

METHODS

Five databases were systematically searched, and complemented with a snowball search from inception to May 2023, for any type of interventional study that included humans of any age with any cancer type. Primary outcomes were incidence and severity of depression, anxiety, and stress symptoms. Secondary outcomes were mood, cognition, quality of life, appetite, nutrition status, gastrointestinal symptoms, and adverse events. Data were pooled using Review Manager. Evidence was appraised using Cochrane risk of bias tools. Confidence in the estimated effect of pooled outcomes was assessed using Grading of Recommendations, Assessment, Development and Evaluation (GRADE).

RESULTS

Fifteen studies (n = 11 randomized trials, n = 4 non-randomized trials) of 18 interventions (N = 1898 total participants; 100 % ≥18 years of age) were included. Ten studies examined THC (70 % synthetic), two synthetic cannabidiol with or without THC, and six whole-plant extracts. No clinically significant effects of medicinal cannabis were found on primary outcomes. The likelihood of anxiety events increased with higher-dose synthetic THC compared with a lower dose (OR: 2.0; 95 % CI: 1.4, 2.9; p < 0.001; Confidence: very low). Medicinal cannabis (THC, cannabidiol, and whole-plant extract) increased the likelihood of improved appetite (OR: 12.3; 95 % CI: 3.5, 45.5; p < 0.001; n = 3 interventions; Confidence: moderate) and reduced severity of appetite loss (SMD: -0.4; 95 % CI: -0.8, -0.1; p = 0.009; Confidence: very low). There was very low confidence that higher doses of synthetic THC increased the likelihood of any adverse event (OR: 0.5; 95 % CI: 0.3, 0.7; p < 0.001). Medicinal cannabis had no effect on emotional functioning, mood changes, confusion, disorientation, quality of life, and gastrointestinal symptoms. Confidence in findings was limited by some studies having high or unclear risk of bias and imprecise pooled estimates.

CONCLUSIONS

There was insufficient evidence to determine the efficacy and safety of medicinal cannabis as a therapeutic intervention for depression, anxiety, or stress in people with active cancer. Further research should explore whether medicinal cannabis might improve and maintain appetite and if high-dose synthetic THC might increase the incidence of side-effects, including anxiety. To inform clinical practice, well-powered and rigorously designed trials are warranted that evaluate the effects of medicinal cannabis prescribed to target anxiety, depression, and stress.

Symptom cluster among cancer survivors from a nationally representative survey: a network analysis

PURPOSE

To identify the symptom cluster among cancer survivors and examine their subgroup differences via network analysis based on nationally representative data.

METHODS

This cross-sectional study included 2966 survivors participating in the 2020 National Health Interview Survey (NHIS). Participants self-reported the presence of 14 symptoms capturing four clusters (physical, somatic, sleep, and psychologic problems). Network analysis models were used to reveal the relationships between symptoms and those interactions. Network comparison tests were applied to compare subgroups.

RESULTS

The core symptoms of the symptom cluster were fatigue (Bet = 33, Clo = 0.0067, Str = 0.9397), pain (Bet = 11, Clo = 0.0060, Str = 0.9226), wake up well rested (Bet = 25, Clo = 0.0057, Str = 0.8491), and anxiety (Bet = 5, Clo = 0.0043, Str = 0.9697) among cancer survivors. The core symptoms, network structure, and global strength were invariant between time since diagnoses (< 2 years vs. ≥ 2 years) or between numbers of cancers (1 vs. ≥ 2), yet varied between the comorbidity group and non-comorbidity group (≥ 1 vs. 0).

CONCLUSIONS

Fatigue would be a potential target for alleviating other symptoms through a negative feedback loop of other related symptoms of cancer survivors. In particular, cancer survivors with other chronic diseases should be the focus of attention and strengthen targeted intervention.

Symptom clusters and their impact on quality of life in multiple myeloma survivors: Secondary analysis of BMT CTN 0702 trial

Autologous haematopoietic cell transplantation (autoHCT) and continuous post-transplant maintenance therapy are the standard of care in transplant-eligible multiple myeloma (MM) patients. We sought to describe symptom burden and identify symptom clusters occurring in MM patients after autoHCT using data from the BMT CTN 0702 randomized controlled trial comparing the outcomes of three treatment interventions after an autoHCT in 758 MM patients. We analysed individual transplant-related symptoms assessed via the FACT-BMT questionnaire at enrolment and annually for 4-year post-autoHCT. We also described the effect the individual symptoms and symptom clusters have on quality of life (QoL). We identified three stable symptom clusters: malaise symptom cluster (lack of energy, feeling ill, having pain, experiencing nausea, loss of appetite), physical symptom cluster (having skin problems, tremors, worsening eyesight, change in taste, shortness of breath, frequent colds) and emotional symptom cluster (feeling sad, being nervous, experiencing sleep problems). Malaise and emotional symptom clusters have a greater impact on QoL than the physical symptoms cluster. Identifying these symptoms warrant additional support in terms of psychosocial support, in addition to treatment of the physical symptoms themselves.

Toward an improved conceptualization of emotions in patients with cancer

Cancer and its associated treatment is a major stressor, leading to emotions such as anxiety or depressive mood. Human emotions have developed through the course of evolution because they facilitate adaptation to important events, such as cancer and its associated treatment. On the other hand, emotions can be maladaptive and interfere with adaptation to cancer. Emotions are maladaptive if they are disproportionally severe or persistent, and if they interfere with functioning. We aim to expand the conceptualization of adaptive and maladaptive emotions in patients with cancer. We draw on major theories in the field of mental disorder and mental health, and apply these theories to conceptualize adaptive and maladaptive emotions in patients with cancer. (i) Maladaptive emotions have two essential features: mental dysfunction and patient harm. Maladaptive emotions are characterized by a network of strongly associated emotional symptoms, which may include cancer-related somatic symptoms. The dysfunctional symptom network is hypothesized to be the result of disturbance of life goal pursuit caused by cancer. (ii) Adaptive emotions have two essential features: ability to deal with cancer and functioning well. The ability to use emotions in an adaptive way depends on skills to recognize, express, and regulate emotions in a flexible manner. A secure attachment style facilitates adaptive emotional responses to cancer. The present conceptualization of adaptive and maladaptive emotions is expected to contribute to better understanding and management of emotions in patients with cancer.

Network analysis used to investigate the interplay among somatic and psychological symptoms in patients with cancer and cancer survivors: a scoping review

PURPOSE

Patients with cancer often experience multiple somatic and psychological symptoms. Somatic and psychological symptoms are thought to be connected and may reinforce each other. Network analysis allows examination of the interconnectedness of individual symptoms. The aim of this scoping review was to examine the current state of knowledge about the associations between somatic and psychological symptoms in patients with cancer and cancer survivors, based on network analysis.

METHODS

This scoping review followed the five-stage framework of Arksey and O'Malley. The literature search was conducted in May, 2023 in PubMed, APA PsycINFO, Embase Cochrane central, and CINAHL databases.

RESULTS

Thirty-two studies were included, with eleven using longitudinal data. Seventeen studies reported on the strength of the associations: somatic and psychological symptoms were associated, although associations among somatic as well as among psychological symptoms were stronger. Other findings were the association between somatic and psychological symptoms was stronger in patients experiencing more severe symptoms; associations between symptoms over time remained rather stable; and different symptoms were central in the networks, with fatigue being among the most central in half of the studies.

IMPLICATIONS FOR CANCER SURVIVORS

Although the associations among somatic symptoms and among psychological symptoms were stronger, somatic and psychological symptoms were associated, especially in patients experiencing more severe symptoms. Fatigue was among the most central symptoms, bridging the somatic and psychological domain. These findings as well as future research based on network analysis may help to untangle the complex interplay of somatic and psychological symptoms in patients with cancer.

Changes in Sexuality and Sexual Dysfunction over Time in the First Two Years after Treatment of Head and Neck Cancer

The aim of this study was to investigate changes in sexuality and sexual dysfunction in head and neck cancer (HNC) patients in the first two years after treatment, in relation to the type of treatment. Data were used of 588 HNC patients participating in the prospective NETherlands Quality of life and Biomedical Cohort Study (NET-QUBIC) from diagnosis to 3, 6, 12 and 24 months after treatment. Primary outcome measures were the International Index of Erectile Function (IIEF) and the Female Sexual Function Index (FSFI). The total scores of the IIEF and FSFI were dichotomized into sexual (dys)function. In men, type of treatment was significantly associated with change in erectile function, orgasm, satisfaction with intercourse, and overall satisfaction. In women, type of treatment was significantly associated with change in desire, arousal, and orgasm. There were significant differences between treatment groups in change in dysfunctional sexuality. A deterioration in sexuality and sexual dysfunction from baseline to 3 months after treatment was observed especially in patients treated with chemoradiation. Changes in sexuality and sexual dysfunction in HNC patients were related to treatment, with an acute negative effect of chemoradiation. This effect on the various domains of sexuality seems to differ between men and women.

A Scalable Radiomics- and Natural Language Processing-Based Machine Learning Pipeline to Distinguish Between Painful and Painless Thoracic Spinal Bone Metastases: Retrospective Algorithm Development and Validation Study

BACKGROUND

The identification of objective pain biomarkers can contribute to an improved understanding of pain, as well as its prognosis and better management. Hence, it has the potential to improve the quality of life of patients with cancer. Artificial intelligence can aid in the extraction of objective pain biomarkers for patients with cancer with bone metastases (BMs).

OBJECTIVE

This study aimed to develop and evaluate a scalable natural language processing (NLP)- and radiomics-based machine learning pipeline to differentiate between painless and painful BM lesions in simulation computed tomography (CT) images using imaging features (biomarkers) extracted from lesion center point-based regions of interest (ROIs).

METHODS

Patients treated at our comprehensive cancer center who received palliative radiotherapy for thoracic spine BM between January 2016 and September 2019 were included in this retrospective study. Physician-reported pain scores were extracted automatically from radiation oncology consultation notes using an NLP pipeline. BM center points were manually pinpointed on CT images by radiation oncologists. Nested ROIs with various diameters were automatically delineated around these expert-identified BM center points, and radiomics features were extracted from each ROI. Synthetic Minority Oversampling Technique resampling, the Least Absolute Shrinkage And Selection Operator feature selection method, and various machine learning classifiers were evaluated using precision, recall, F1-score, and area under the receiver operating characteristic curve.

RESULTS

Radiation therapy consultation notes and simulation CT images of 176 patients (mean age 66, SD 14 years; 95 males) with thoracic spine BM were included in this study. After BM center point identification, 107 radiomics features were extracted from each spherical ROI using pyradiomics. Data were divided into 70% and 30% training and hold-out test sets, respectively. In the test set, the accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve of our best performing model (neural network classifier on an ensemble ROI) were 0.82 (132/163), 0.59 (16/27), 0.85 (116/136), and 0.83, respectively.

CONCLUSIONS

Our NLP- and radiomics-based machine learning pipeline was successful in differentiating between painful and painless BM lesions. It is intrinsically scalable by using NLP to extract pain scores from clinical notes and by requiring only center points to identify BM lesions in CT images.

Experience of living with symptom clusters in postoperative pancreatic cancer patients

OBJECTIVE

The purpose of this study was to describe symptom clusters (SCs), and symptom experiences in early postoperative patients with pancreatic cancer who are recovering at home.

METHODS

From October 2021 and April 2022,15 patients following pancreatic cancer surgery were recruited from a tertiary hospital in Jiangsu Province by maximum variation sampling. Semi-structured interviews were performed to collect data and thematic analysis was conducted to analyze the data.

RESULTS

Three themes were extracted from the data. The first theme "difference in symptom perception and cognition" illustrated factors such as patient knowledge, expected recovery status, and risk perception could influence patient symptom experience. Most patients actually reported fewer and less severe symptoms than previous studies. The second theme "the results of symptom cluster" demonstrated anorexia-distress SC, bowel-digestive related SC and sleep disturbance related SC, and anorexia-distress SC should be considered as the priority SC given its multidimensional significance for patients. The last theme described the patient's positive attitudes, behaviours and barriers to coping with symptoms, namely "symptom self-management experience".

CONCLUSION

There are differences in the perception and interpretation of SCs in postoperative pancreatic cancer patients. Understanding the meaning of individual differences in the experience of symptoms can assist patients in the management of SCs. Medical staff should combine patient self-management strategies and evidence-based data to provide appropriate support at different stages to improve patient symptom management and quality of life.

Artificial intelligence empowered digital health technologies in cancer survivorship care: A scoping review

Objective

The objectives of this systematic review are to describe features and specific application scenarios for current cancer survivorship care services of Artificial intelligence (AI)-driven digital health technologies (DHTs) and to explore the acceptance and briefly evaluate its feasibility in the application process.

Methods

Search for literatures published from 2010 to 2022 on sites MEDLINE, IEEE-Xplor, PubMed, Embase, Cochrane Central Register of Controlled Trials and Scopus systematically. The types of literatures include original research, descriptive study, randomized controlled trial, pilot study, and feasible or acceptable study. The literatures above described current status and effectiveness of digital medical technologies based on AI and used in cancer survivorship care services. Additionally, we use QuADS quality assessment tool to evaluate the quality of literatures included in this review.

Results

43 studies that met the inclusion criteria were analyzed and qualitatively synthesized. The current status and results related to the application of AI-driven DHTs in cancer survivorship care were reviewed. Most of these studies were designed specifically for breast cancer survivors' care and focused on the areas of recurrence or secondary cancer prediction, clinical decision support, cancer survivability prediction, population or treatment stratified, anti-cancer treatment-induced adverse reaction prediction, and so on. Applying AI-based DHTs to cancer survivors actually has shown some positive outcomes, including increased motivation of patient-reported outcomes (PROs), reduce fatigue and pain levels, improved quality of life, and physical function. However, current research mostly explored the technology development and formation (testing) phases, with limited-scale population, and single-center trial. Therefore, it is not suitable to draw conclusions that the effectiveness of AI-based DHTs in supportive cancer care, as most of applications are still in the early stage of development and feasibility testing.

Conclusions

While digital therapies are promising in the care of cancer patients, more high-quality studies are still needed in the future to demonstrate the effectiveness of digital therapies in cancer care. Studies should explore how to develop uniform standards for measuring patient-related outcomes, ensure the scientific validity of research methods, and emphasize patient and health practitioner involvement in the development and use of technology.