Bifurcation in the Healing or Fibrotic Response in a Network Model of Fibrosis: Role of the Initial Injury Structure

Ethan Israel¹Joseph Hall¹Yuqing Deng²Jason H.T. Bates³Béla Suki^1*

• ¹Biomedical Engineering, Boston University, Boston, MA, United States
• ²Mechanical Engineering, Boston University, Boston, MA, United States
• ³Vermont Lung Center, Larner College of Medicine, University of Vermont, Burlington, Vermont, United States

Pulmonary fibrosis (PF) is a heterogeneous progressive lung disease characterized by excessive extracellular matrix (ECM) deposi on and cross-linking, leading to irreversible ssue s ffening and loss of func on. Previous evidence suggests that percola on behavior, where increasing local s ffness facilitates the emergence of s ff regions that span the ssue, underlies the s ffening of the ECM, and drives irreversible mechanical dysfunc on. However, it is not fully understood how percola on emerges from the complex interac ons between cells and the ECM. In this study, we inves gated a previously published agent-based spring network model of PF that exhibited bifurca on behavior between healing and fully developed fibrosis as network members were gradually s ffened. By systema cally analyzing the configura on of the ini al ssue injury, we iden fy key structural determinants that govern whether an injury heals or transi ons into fibrosis. Results demonstrate that fibrosis is strongly associated with increased ini al clustering of injured springs, reduced inter-cluster distances, and the presence of cri cal s ffening sites that act as bifurca on points for disease progression. Furthermore, we show that selec vely modifying the s ffness of pivotal network regions at the me of injury can shi the network's trajectory from fibrosis to healing, highligh ng poten al interven on targets. These findings suggest that the network structure of ssue injury may serve as a predic ve marker for fibrosis suscep bility and provide a mechanis c basis for understanding the nonlinear progression of PF.