利用脂质组学识别结直肠癌患者癌症相关疲劳的新型治疗靶点：ColoCare研究的结果

Background: Cancer-related fatigue (CRF) is the most frequently reported symptom among patients with colorectal cancer (CRC), with limited therapeutic options. Alterations in metabolic pathways related to lipid metabolism have been shown to play a role in non-cancer fatigue-associated diseases, and are hypothesized to influence CRF. The purpose of the present study was to investigate serum lipidomic biomarkers to identify longitudinal predictors of CRF in a prospective cohort of patients with CRC. Methods: The ColoCare Study enrolled men and women ages 18 to 89 with newly diagnosed primary stage I-IV CRC at six U.S. sites and one German site. CRF was measured using the fatigue subscale of the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire C30 (EORTC QLQ-C30) at T0 (CRC surgery/baseline), T1 (6 months post-surgery), T2 (12 months), and T3 (24 months). Using blood collected at each time point, we performed targeted lipidomics following comprehensive protocols for measurement and quality control. For the present study, participants with stage I-III disease and at least one measurement of CRF and lipidomic profiling (N=863) were included. Using linear mixed effects models with an interaction term with time point and subject-specific random intercepts, we assessed associations between individual lipids and CRF at each time point. We adjusted for multiple testing using the Benjamini-Hochberg correction for false-discovery rate. Models also adjusted for age, sex, tumor site, stage, body mass index, chemotherapy, radiation, and study site. We used elastic net regression on a training subset (n=176) to identify lipids at T1 (when first-line treatment was nearing completion) that were predictive of fatigue at T2. Further analyses validating prediction models and using metabolic pathway analyses are ongoing. Results: Mean age was 61.6 years (SD: 12.7). N=305 total lipids were identified. In linear mixed effects models, no lipids were associated with CRF at T0 or T1. At T2 and T3, higher levels of ceramides (20), monohexosylceramides (11), gangliosides (4), trihexosylceramides (1), and sphingomyelins (16) were associated with lower CRF after adjustment for multiple testing. Predictive modeling identified higher levels of three sphingolipids at T1 associated with lower CRF at T2 and higher levels of two lipids (one diglyceride and one ceramide) at T1 associated with higher CRF at T2. Conclusions: Specific sphingolipids, including ceramides, monohexosylceramides, and sphingomyelins, were inversely associated with CRF, suggesting a protective role. Predictive modeling supports their potential as targetable biomarkers of fatigue. These findings highlight lipid metabolism as a promising target for understanding and mitigating fatigue in cancer survivors, warranting external validation and mechanistic research.