Noninvasive Lung Cancer Subtype Classification Using Tumor-Derived Signatures and cfDNA Methylome.

Area of Science:

• Oncology
• Genomics
• Molecular Diagnostics

Background:

• Accurate lung cancer subtyping is crucial for effective treatment decisions.
• Histologic examination of tumor biopsies is the current standard for lung cancer subtyping.
• Liquid biopsy using cell-free DNA (cfDNA) shows promise for cancer detection and classification.

Purpose of the Study:

• To investigate the potential of cfDNA methylome for noninvasive classification of lung cancer histologic subtypes.
• To develop a robust classification model for distinguishing lung adenocarcinoma and lung squamous cell carcinoma using cfDNA methylation profiles.

Main Methods:

• Identified subtype-specific methylation markers from tumor samples using a fragment-based approach.
• Validated markers in independent cohorts and assessed their association with transcriptional activity.
• Constructed a classification model based on cfDNA methylation profiles.

Main Results:

• Achieved an AUC of 0.808 in cross-validation and 0.747 in independent validation for subtype classification.
• Identified robust, subtype-specific methylation markers.
• Inferred tumor copy-number alterations from cfDNA methylome analysis for potential treatment selection.

Conclusions:

• cfDNA methylome analysis is a promising tool for noninvasive lung cancer subtyping.
• This approach offers potential for improved cancer monitoring and early detection.
• Findings support the use of cfDNA methylation for guiding lung cancer treatment decisions.