Abstract PS4-02-09: Symptom Clustering and Multi-Omics Signatures of Cancer-Related Fatigue in Breast Cancer: Findings from the China PERSEVERE Cohort

Abstract Background: Cancer-related fatigue (CRF) is a debilitating and underrecognized symptom in breast cancer, often preceding treatment and co-occurring with psychological and physical complaints. However, its pathobiological basis remains poorly understood. We leveraged the China PERSEVERE study (NCT07010939), a national prospective cohort of early breast cancer patients, to investigate baseline symptom clusters and identify molecular correlates of CRF using multi-omics profiling. Methods: PERSEVERE is an ongoing multi-center cohort study designed to evaluate treatment-related toxicities and quality of life in women with stage I-III breast cancer in China. Baseline assessments, all conducted prior to surgery or neoadjuvant treatment, include detailed sociodemographic and clinical data, biospecimen collection (plasma, PBMCs, and tumor tissue), and validated patient-reported outcome (PRO) measures (EORTC QLQ-C30/BR23, FA12, IPAQ, HADS, PSS-10, PSQI, SSRS). This cross-sectional analysis included 168 newly diagnosed breast cancer patients enrolled to date, all of whom had complete baseline PROs and biospecimen collection. Clinically significant CRF was defined as a QLQ-C30 fatigue score ≥40 or any FA12 fatigue domain score ≥40. Symptom interrelationships were analyzed using Spearman correlation and hierarchical clustering. Multivariable logistic regression identified predictors of baseline CRF. A nested case-control subgroup of 14 participants (7 with high fatigue scores and 7 matched controls based on age and clinical stage) underwent untargeted baseline serum proteomic and metabolomic profiling. Results: At baseline, 39/168 participants (23.2%) met the criteria for clinically significant CRF. CRF scores showed moderate correlation with poor sleep quality (PSQI; r=0.37, p0.001), anxiety (HADS-A; r=0.32, p0.05), and perceived stress (PSS-10; r=0.25, p=0.01), forming a distinct symptom cluster. Patients with CRF also exhibited significantly higher systolic blood pressure, breast-specific symptom burden (BR23), and elevated IL-6 levels (all p0.05). In adjusted models, BR23 symptom score and waist-to-hip ratio (WHR) remained independent predictors of CRF. In the multi-omics subgroup, 52 differentially expressed proteins were identified, enriched in cytokine signaling (e.g., IL-6 receptor activity) and mitochondrial pathways (e.g., oxidative phosphorylation). Metabolomic analysis revealed 76 altered metabolites, implicating disrupted kynurenine metabolism, glutathione balance, and neuroactive signaling. Integrated correlation analysis linked CRF severity with altered inflammatory markers (CRP, IL6R) and mitochondrial dysfunction (e.g., antimycin A1, tiopronin) across both omics layers. Conclusions: CRF is a core feature of a symptom cluster involving sleep and psychological distress, even before systemic therapy. This integrated clinical and molecular analysis from the PERSEVERE cohort highlights inflammation and mitochondrial dysfunction as plausible contributors to baseline fatigue. These findings inform early identification of high-risk individuals and suggest future intervention targets. Longitudinal follow-up is underway to validate predictive trajectories and biological correlates of CRF over time. Citation Format: Z. Qu, S. Pei, K. Li, F. YI. Symptom Clustering and Multi-Omics Signatures of Cancer-Related Fatigue in Breast Cancer: Findings from the China PERSEVERE Cohort abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS4-02-09.