Dynamic and static circulating cancer microRNA biomarkers - a validation study

Area of Science:

• Biochemistry
• Oncology
• Genomics

Background:

• Early diagnosis of cancers significantly improves patient survival rates.
• Longitudinal cohort profiling offers valuable insights into biomarker expression trajectories.
• MicroRNAs (miRNAs) are emerging as critical biomarkers in various pathologies.

Purpose of the Study:

• To investigate the diagnostic potential of circulating microRNAs in serum for early cancer detection.
• To identify both static and dynamic microRNA biomarkers for colon, lung, and breast cancers.
• To evaluate the efficacy of computational models in classifying cancer patients based on miRNA profiles.

Main Methods:

• Serum samples from 240 individuals (cancer patients and controls) were analyzed for microRNA expression.
• Longitudinal sampling (pre- and post-diagnosis) with a median interval of 11.6 years was employed.
• Computational models, including gradient boosted trees, were used to analyze 21 microRNAs and classify samples.

Main Results:

• Two sets of diagnostic biomarkers were identified: static (absolute difference) and dynamic (change over time).
• miR-99a-5p showed significant differences across all three cancer types compared to controls.
• miR-155-5p demonstrated predictive value for lung and colon cancers.
• Gradient boosted trees achieved an average classification accuracy of 79.9% for cancer vs. non-cancer patients.

Conclusions:

• Individual changes in microRNA levels over time or absolute values can predict cancer with high specificity and sensitivity.
• Circulating microRNAs hold promise as non-invasive biomarkers for early cancer diagnosis.
• Dynamic miRNA profiling offers enhanced diagnostic information content.