Predicting clinical response to everolimus in ER+ breast cancers using machine-learning

Area of Science:

• Oncology
• Molecular Biology
• Bioinformatics

Background:

• Endocrine therapy is standard for ER+ breast cancer, but resistance is common.
• PI3K/AKT/mTOR pathway activation drives acquired resistance to endocrine therapy.
• Everolimus, an mTOR inhibitor, improves outcomes in metastatic ER+ breast cancer, but lacks predictive biomarkers.

Purpose of the Study:

• To develop and validate an integrative machine learning biomarker for predicting response to everolimus.
• To identify patients with advanced ER+ breast cancer who will most likely benefit from everolimus treatment.

Main Methods:

• Utilized gene expression signatures from ER+ breast cancer cell lines and patients treated with everolimus.
• Developed and validated an integrative machine learning model to distinguish responders from non-responders.

Main Results:

• The developed machine learning biomarker successfully distinguished between responders and non-responders to everolimus.
• The biomarker demonstrates potential for clinical application in guiding everolimus treatment decisions.

Conclusions:

• An integrative machine learning biomarker can predict everolimus response in ER+ breast cancer.
• This biomarker may help personalize treatment strategies for patients with advanced ER+ breast cancer, overcoming resistance.