Comprehensive transcriptomic analysis to identify biological and clinical differences in cholangiocarcinoma

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Cholangiocarcinoma (CC) is a rare, aggressive biliary tract cancer with poor prognosis and limited treatment options.
• Intrahepatic cholangiocarcinoma (ICC) and extrahepatic cholangiocarcinoma (ECC) represent distinct subtypes requiring tailored therapeutic approaches.
• Comprehensive molecular classification is crucial for understanding CC heterogeneity and improving patient outcomes.

Purpose of the Study:

• To develop a gene expression-based classification of cholangiocarcinoma (CC) with clinical relevance.
• To identify distinct biological subgroups within ICC and ECC based on transcriptomic data.
• To create predictive models for stratifying CC patients by prognosis and biological features.

Main Methods:

• Utilized transcriptomic data from seven public datasets (543 patients) for discovery analysis.
• Applied unsupervised clustering to identify molecular subgroups in ICC and ECC.
• Developed and validated class predictors using machine learning on an independent cohort (131 patients).

Main Results:

• Identified four distinct subgroups in both ICC and ECC, characterized by unique immune infiltrates and signaling pathways.
• Developed short gene list signatures for class prediction.
• Validated the ability to distinguish ICC subgroups with different prognoses in an independent dataset.

Conclusions:

• The developed class-predictor accurately identifies CC subgroups at the single-sample level.
• These subgroups exhibit specific biological features and clinical behaviors.
• This classification serves as a foundation for further molecular characterization and clinical application of CC genomics.