Identification of Circulating Lipidomic Biomarkers of Malnutrition Risk among Oncology Patients in the Total Cancer Care (TCC) Study: A Cross-Sectional Analysis.

BACKGROUND

Early identification of malnutrition is critical for improving clinical outcomes in oncology patients. However, there are no established biomarkers for malnutrition screening.

OBJECTIVES

This study aimed to identify circulating lipid species associated with malnutrition risk among oncology patients through lipidomic analysis.

METHODS

A cross-sectional study was conducted using plasma samples from oncology patients classified as at risk (n = 90) or not at risk (n = 90) for malnutrition using the Malnutrition Screening Tool (Malnutrition Screening Tool score = 0 compared with ≥2). All participants had head and neck, lungs, or gastrointestinal cancer. Targeted lipidomics were conducted using liquid chromatography-mass spectrometry. Elastic net regression adjusted for confounding variables identified lipids associated with malnutrition risk. A weighted lipid malnutrition risk score was derived and evaluated using the receiver operating characteristic area under the curve. Conditional multivariable logistic regression assessed the association of the lipid score with malnutrition risk. Lipid enrichment analysis was performed using the lipid ontology enrichment framework.

RESULTS

Elastic net regression identified 12 lipids species that were inversely associated with malnutrition risk: cholesterol ester 20:0, ceramide 18:2;O2/26:0, lysophosphatidylcholine 26:0/0:0, lysophosphatidylinositol 18:2/0:0, phosphatidylcholine 34:5, phosphatidylcholine 40:8, phosphatidylethanolamine (PE) P-18:0/20:3, PE P-18:1/18:2, PE P-18:1/20:4, sulfated hexosylceramide 18:1;O2/16:0, sphingomyelin 18:2;O2/23:0, and triglyceride (O-50:1). One lipid, dihexosylceramide 18:1;O2/20:0, was positively associated with malnutrition risk. The weighted lipid malnutrition risk score was associated with increased risk for malnutrition [odds ratio: 3.57; 95% confidence interval (CI): 1.97, 6.47, P < 0.001]. Addition of the lipid score to established malnutrition risk factors improved model predictive performance, increasing the receiver operating characteristic area under the curve from 0.78 (95% CI: 0.71, 0.84) to 0.90 (95% CI: 0.86, 0.94). Lipid ontology enrichment analysis indicated downregulation of membrane structure and signaling lipids and upregulation of storage lipids.

CONCLUSIONS

This study highlights the potential of lipidomics to identify biomarkers of malnutrition risk among oncology patients. Large, prospective studies are warranted to validate these findings.