Integrative Analysis and Machine Learning based Characterization of Single Circulating Tumor Cells.

Area of Science:

• Oncology
• Genomics
• Immunology

Background:

• Circulating tumor cells (CTCs) are crucial biomarkers for cancer detection, prognosis, and treatment monitoring.
• Understanding the heterogeneity of CTCs, particularly their epithelial-to-mesenchymal transition (EMT) status, is vital for effective cancer management.
• The interplay between PD-L1 and MHC expression in CTCs may influence response to cancer immunotherapy.

Purpose of the Study:

• To analyze the transcriptomic profiles of CTCs across various cancer types.
• To investigate the relationship between EMT and gene expression patterns, including PD-L1 and MHC, in CTCs.
• To develop and validate a classifier for accurate CTC identification using gene expression data.

Main Methods:

• Collated and analyzed publicly available single-cell RNA sequencing data of CTCs from diverse cancers.
• Performed integrative analysis of CTC transcriptomes to identify key gene expression patterns.
• Trained a machine learning classifier using CTC and peripheral blood mononuclear cell (PBMC) expression profiles.
• Validated the classifier using CTCs captured by a novel label-free microfluidic enrichment system.

Main Results:

• CTCs from different cancers demonstrate a continuous spectrum of EMT.
• An inverse correlation between PD-L1 and MHC gene expression was observed in CTCs.
• A classifier was successfully trained to recognize diverse CTC phenotypes with high accuracy.
• The classifier validated the presence of circulating breast tumor cells isolated via the new microfluidic device.

Conclusions:

• The EMT continuum in CTCs provides a unified model for understanding CTC biology across cancers.
• The identified PD-L1/MHC expression pattern offers insights into immune evasion mechanisms and potential immunotherapy targets.
• The developed classifier represents a promising tool for CTC detection and characterization.
• The study validates a novel microfluidic system for CTC enrichment and analysis.