In Silico Digital Twins of Bone Metastasis Enable Investigation of Tumor Progression and Therapy Response

Luca Marsilio ^1,2, Sergio Barrios ³, Stefan Maksimovic ³

⁰Department of Electronics, Information and Bioengineering, Politecnico di Milano University, Milan, Italy.

Cancer Research +

|

September 18, 2025

View abstract on PubMed

Summary

Computational models of bone metastasis (BM) were developed to understand cancer progression and therapy response. These agent-based models accurately predict treatment outcomes for prostate and kidney cancers, aiding drug development.

Area Of Science

Computational biology
Cancer research
Biophysics

Background

Bone metastasis (BM) significantly impacts prostate and renal cancer patient outcomes.
Current in vivo models have limitations in capturing the complexity of BM.
Multiscale computational approaches are needed to understand tumor-bone interactions.

Purpose Of The Study

To develop spatially explicit, multicellular agent-based models of BM.
To simulate key processes like angiogenesis and bone resorption.
To validate models using in vivo data and predict therapy response.

Main Methods

Developed agent-based models inspired by in vivo bone metastasis.
Incorporated angiogenesis and bone resorption dynamics.
Calibrated models with prostate and kidney tumor data.

Main Results

Models successfully recapitulated tumor progression, angiogenesis, and bone resorption.
Simulations accurately predicted the effects of cabozantinib (antiangiogenic) and zoledronic acid (antiresorptive).
Demonstrated the predictive power of agent-based models for therapeutic outcomes.

Conclusions

Agent-based models provide a powerful tool for understanding BM dynamics.
These models can accelerate the evaluation of novel treatment strategies, including combination therapies.
The developed digital twins enhance the understanding of metastatic processes and therapy response in bone cancers.