Cancer-related fatigue (CRF) is a significant problem for cancer patients. This prospective, basic science, observational study will evaluate for changes in CRF associated with molecular characteristics prior to, during, and at the completion of non-investigational, standard-of-care, combined chemotherapy and radiation therapy (CCRT) and to develop and assess predictive models for CRF severity.
Primary Objective For mean, morning and evening CRF: Aim 1. Evaluate for associations between phenotypic characteristics and initial levels and the trajectories of CRF. Aim 2. Evaluate for associations between changes in CRF severity and changes in gene expression levels prior to the initiation and at the end of CCRT. Aim 3. Evaluate for associations between changes in CRF severity and changes in circulating free cytokine levels prior to the initiation and at the end of CCRT. Aim 4. Develop and assess predictive models for CRF severity midway, at the end of, and at least six months post-CCRT using demographic, clinical, and molecular characteristics collected prior the initiation of CCRT. Secondary Objectives For the commonly co-occurring symptom of chemotherapy-induced peripheral neuropathy (CIPN): Secondary Aim 5. Evaluate for associations between phenotypic characteristics and initial levels and the trajectories of CIPN. Secondary Aim 6. Evaluate for associations between changes in CIPN severity and changes in gene expression levels prior to the initiation and at the end of CCRT. Secondary Aim 7. Evaluate for associations between changes in CIPN severity and changes in circulating free cytokine levels prior to the initiation and at the end of CCRT. Secondary Aim 8. Develop and assess predictive models for CIPN severity midway, at the end of, and at least six months post-CCRT using demographic, clinical, and molecular characteristics collected prior the initiation of CCRT. Exploratory Aim 1 - Evaluate the feasibility of the protocol for the collection of stool samples. Exploratory Aim 2 - Evaluate the feasibility of processing and storing stool samples. Exploratory Aim 3 - Evaluate the feasibility of processing and storing performing blood samples and performing Cytometry by time of flight (CyTOF) assays.
Study Type
OBSERVATIONAL
Enrollment
125
Blood samples will be obtained throughout the course of the study
Stool samples will be obtained throughout the course of the study
Surveys will be given throughout the course of the study.
University of California, San Francisco
San Francisco, California, United States
RECRUITINGMeasure associations between changes in cancer-related fatigue (CRF) and changes in gene expression over time
Association between phenotypic characteristics and initial levels and trajectories of CRF severity will be assessed using a hierarchical linear model (HLM) approach.
Time frame: Up to 34 weeks
Measure associations between changes in CRF and changes in cytokine levels over time
Association between changes in CRF severity and biomarker levels prior to the initiation and at the end of CCRT. Linear regression will be used to evaluate for associations between fatigue changes and biomarker levels at baseline controlling for covariates identified in the initial primary outcome. Adjustments for multiple comparisons will be conducted using the Benjamini-Hochberg (BH) procedure at a false discovery rate (FDR) of 10%.
Time frame: Up to 34 weeks
Measure associations between changes in CRF and changes in gene expression over time
Association between changes in CRF severity and gene expression prior to the initiation and at the end of CCRT. Linear regression will be used to evaluate for associations between fatigue changes and biomarker levels at baseline controlling for covariates identified in the initial primary outcome. Adjustments for multiple comparisons will be conducted using the Benjamini-Hochberg (BH) procedure at a false discovery rate (FDR) of 10%.
Time frame: Up to 34 weeks
Evaluate the predictive utility of gene expression and cytokine data
A validated prediction model of CRF severity will be generated using machine learning (ML) methods to minimize the error between predicted and observed levels of fatigue midway through CCRT, at the completion of CCRT, and at least six months following the completion of CCRT. Evaluation of common ML algorithms for prediction accuracy and evaluation of model performance as compared to simple linear regression. Separate training and testing sets will be created, cross-validated, and repeated and impact of each variable will be determined.
Time frame: Up to 34 weeks
Evaluate for associations between changes in chemotherapy-induced peripheral neuropathy (CIPN) and changes in gene expression
The association between phenotypic characteristics and initial levels and trajectories of CIPN severity will be evaluated using a hierarchical linear model (HLM) approach.
Time frame: Up to 34 weeks
Evaluate for associations between changes in CIPN and changes in cytokine levels
The association between phenotypic characteristics and initial levels and trajectories of CIPN severity will be evaluated using a hierarchical linear model (HLM) approach.
Time frame: Up to 34 weeks
Evaluate the predictive utility of gene expression and severity of CIPN
The association between changes in CIPN severity and gene expression prior to the initiation and at the end of CCRT. Linear regression will be used to evaluate for associations between CIPN changes and gene expression at baseline controlling for covariates identified in previous objectives/endpoints. Adjustments for multiple comparisons will be performed using the Benjamini-Hochberg (BH) procedure at a false discovery rate (FDR) of 10%.
Time frame: Up to 34 weeks
Evaluate the predictive utility of cytokine levels and severity of CIPN
The association between changes in CIPN severity and cytokine levels prior to the initiation and at the end of CCRT. Linear regression will be used to evaluate for associations between CIPN changes and cytokine levels at baseline controlling for covariates identified in previous objectives/endpoints. Adjustments for multiple comparisons will be performed using the Benjamini-Hochberg (BH) procedure at a false discovery rate (FDR) of 10%.
Time frame: Up to 34 weeks
Evaluate the predictive model of severity of CIPN
The predictive utility will be assessed through a validated prediction model of CIPN severity using machine learning (ML) methods to minimize the error between predicted and observed levels of CIPN midway through CCRT, at the completion of CCRT, and at least six months following the completion of CCRT. We will evaluate common ML algorithms for prediction accuracy and evaluate their performance as compared to simple linear regression
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
Time frame: Up to 34 weeks