A statistical framework for detecting therapy-induced resistance from drug screens

Area of Science:

• Cancer Research
• Mathematical Biology
• Evolutionary Dynamics

Background:

• Therapy resistance in cancer is a major challenge, often driven by stem-like cells and induced plasticity.
• Anticancer therapies can inadvertently promote resistance through epigenetic changes and de-differentiation.
• Understanding the interplay between anti-tumor effects and induced resistance is crucial for effective treatment strategies.

Purpose of the Study:

• To develop a statistical framework for characterizing tumor cell population dynamics under therapy.
• To enable the detection and quantification of therapy-induced resistance using multi-type branching process models.
• To analyze high-throughput drug screening data without needing subpopulation counts.

Main Methods:

• Utilized multi-type branching process models to simulate tumor cell population evolution.
• Developed a statistical framework to analyze total cell count data from drug screening.
• Validated the model using both in silico simulations and in vitro experimental data.

Main Results:

• The framework successfully detected and quantified therapy-induced resistance in simulated and experimental data.
• Demonstrated the ability to predict evolutionary dynamics of tumor cell populations.
• Provided a robust method for analyzing drug screening results.

Conclusions:

• The developed statistical framework offers a powerful tool for understanding and combating cancer therapy resistance.
• This approach can guide the development of more effective cancer treatment strategies by accounting for induced resistance.
• The model's ability to use total cell counts simplifies resistance analysis in drug screening.