Mechanobiological Modeling of Lumbar Interbody Fusion: Sensitivity Analyses on Cell-Specific Parameters, Loading Conditions, and Cage Material.

Despite decades of clinical experience, lumbar interbody fusion (LIF) still leads to pseudoarthrosis in 5%-28% of patients. Although tissue-differentiation algorithms offer a powerful predictive framework, their application to spinal fusion modeling is currently limited. In this study, we present one of the first mechano-regulated tissue differentiation algorithms incorporating cell-phenotype-specific activities, with the goal to simulate the fusion process after posterior-LIF. The influence of the model input parameters (n = 23) was first assessed in 2D through fractional factorial analyses to identify the most influential parameters and quantify their effect. Second, we simulate a 3D idealized case study of posterior-LIF at the L4-L5 spinal level. The effects of interbody cage material (Ti6Al4V vs. PEEK) and loading conditions (standing vs. daily activity) on bone concentration and distribution were investigated. The fractional factorial analyses identified the rates of bone formation and degradation as the most influential parameters, followed by the osteoblast migration rate and key parameters governing mesenchymal stem cell activities. From the LIF model, simple standing led to higher bone formation in the posterior disc region, whereas daily activity promoted anterior bone formation, in agreement with clinical observations. The stiffer Ti6Al4V cage induced stress shielding and resulted in sub-optimal tissue stimulation, whereas the softer PEEK cage promoted appropriate mechanical stimulation, leading to both internal and lateral bony bridging formation. This work demonstrated the potential of our novel mechano-regulated tissue-differentiation algorithm to predict clinically observed fusion patterns and support future assessments of surgical procedures.