A Metabolomics-Based Screening Proposal for Colorectal Cancer

Area of Science:

• Oncology
• Metabolomics
• Biochemistry

Background:

• Colorectal cancer (CRC) presents significant morbidity and mortality.
• Current screening methods like fecal occult blood tests and colonoscopy have limitations in accuracy and invasiveness.
• Early diagnosis of CRC is crucial for improving patient outcomes.

Purpose of the Study:

• To investigate the potential of serum metabolomic profiles for developing a novel, non-invasive screening approach for colorectal cancer.
• To identify metabolic alterations associated with colorectal cancer.
• To evaluate the diagnostic performance of machine learning models applied to metabolomic data.

Main Methods:

• Untargeted serum metabolomic profiling was performed on 100 CRC patients, 50 healthy controls (CTRLs), and 50 individuals with benign colorectal disease.
• Machine learning models, including an ensemble model, were developed to differentiate CRC patients from controls.
• Metabolite-set enrichment analysis and metabolic pathway analysis were conducted to identify key metabolic differences.

Main Results:

• An ensemble machine learning model achieved 100% accuracy in classifying CRC patients and controls using a subset of the data.
• Metabolite-set enrichment analysis revealed significant differences in cell glucose metabolism between CRC patients and controls.
• The study identified a distinct metabolomic signature associated with colorectal cancer.

Conclusions:

• Serum metabolomic signatures show potential as a non-invasive screening tool for colorectal cancer.
• Metabolic pathway analysis provides insights into the pathophysiology of CRC.
• Further validation in larger, blinded studies is warranted to confirm these findings.