Comprehensive circulating microRNA profile as a supersensitive biomarker for early-stage lung cancer screening

Area of Science:

• Biomarkers
• Genomics
• Oncology

Background:

• Early diagnosis of lung cancer is critical for patient survival.
• Conventional blood biomarkers have limitations in detecting early-stage lung cancer.

Purpose of the Study:

• To evaluate a serum comprehensive miRNA profile as a sensitive biomarker for early-stage lung cancer.
• To compare the diagnostic performance of the miRNA profile against conventional blood biomarkers.
• To utilize next-generation sequencing (NGS) and automated machine learning (AutoML) for biomarker discovery.

Main Methods:

• NGS was used to analyze miRNA profiles in 262 serum samples.
• AutoML screened 1123 miRNA-based diagnostic models from a discovery set.
• The best model's diagnostic performance was validated on independent patient samples.

Main Results:

• The miRNA-based diagnostic model achieved a high AUC score of 0.98 in validation.
• The model demonstrated high sensitivity (85.7%) for early-stage lung cancer.
• The miRNA model showed significantly higher sensitivity than CEA for early lung adenocarcinoma (77.8% vs. 27.8%).

Conclusions:

• Serum comprehensive miRNA profiles can serve as highly sensitive biomarkers for early-stage lung cancer.
• The developed miRNA-based diagnostic model shows promise for non-invasive early lung cancer detection.
• This approach offers a potential advancement over current blood-based diagnostic methods.