Age-Specific Transcriptomic Signatures for Classification of Progression-Free Survival Outcomes in Luminal A Breast...

Area of Science:

• Genomics and Bioinformatics
• Oncology
• Translational Research

Background:

• Progression-free survival (PFS) is a key endpoint in Luminal A breast cancer (BC).
• Molecular drivers of BC recurrence are not fully understood.
• Understanding age-specific recurrence patterns is important for personalized treatment.

Purpose of the Study:

• To identify gene expression alterations associated with recurrence in Luminal A BC across different age groups.
• To evaluate the predictive value of these transcriptomic alterations using machine learning.
• To develop an integrated model combining clinical and molecular data for recurrence prediction.

Main Methods:

• Analysis of gene expression profiles and clinical data from the METABRIC cohort.
• Stratification of patients into premenopausal, postmenopausal non-geriatric, and geriatric subgroups.
• Differential gene expression analysis, feature selection, and machine learning (XGBoost) model development.

Main Results:

• Fifteen genes with significant differential expression were identified, with KMT2D, RFNG, IGF1, and CDKN2C showing consistent recurrence association.
• KMT2D emerged as the most significant molecular predictor, while age, NPI, and tumor size were key clinical predictors.
• An integrated clinicopathological-transcriptomic XGBoost model demonstrated superior predictive performance (AUC=0.71, accuracy=0.78) compared to clinical models alone.

Conclusions:

• Transcriptomic biomarkers, particularly KMT2D, combined with clinical variables, show promise for predicting recurrence in Luminal A BC.
• The findings support further investigation into integrated models for improved recurrence risk stratification.
• External validation in independent cohorts is necessary to confirm the clinical utility of these biomarkers.