Precision Combination Therapies Based on Recurrent Oncogenic Coalterations

Area of Science:

• Oncology
• Bioinformatics
• Computational Biology

Background:

• Cancer is driven by multiple genetic alterations.
• Drug combinations can suppress oncogenic effects.
• Personalized medicine requires therapies tailored to specific molecular profiles.

Purpose of the Study:

• To create a resource of precision combination therapies for cancer.
• To develop a computational method for identifying therapies targeting recurrent oncogenic co-alterations.
• To validate the efficacy of therapies selected by the developed method.

Main Methods:

• Developed the Recurrent Features Leveraged for Combination Therapy (REFLECT) pipeline.
• Integrated machine learning and cancer informatics with multiomic data.
• Validated the pipeline using patient-derived xenografts, in vitro screens, and clinical trial data.

Main Results:

• REFLECT identified therapeutically actionable co-alteration signatures in patient cohorts.
• REFLECT-selected combination therapies demonstrated improved efficacy, synergy, and survival outcomes.
• Validated improvements were observed in preclinical and clinical settings.

Conclusions:

• REFLECT provides a data-driven framework for designing precision combination therapies.
• The platform optimizes therapeutic benefit by matching drug combinations to tumor molecular signatures.
• This approach can guide the design of future clinical trials and preclinical studies.