Mechanical signals play a key role in regulating cellular responses to pathological stressors, such as intracellular infections. In this work, we investigate how the spread of food-borne intracellular bacterial pathogen Listeria monocytogenes (Lm) through epithelial monolayers, depends on the initial shape of the infection focus (i.e., domain of infected cells). This shape modulates the ensuing mechanical competition between infected and uninfected cells, where surrounding uninfected cells forcefully migrate towards the infection focus, squeezing and eventually extruding infected cells. In our previous work [1], we developed a 2D hybrid computational model, integrating an agent-based model (ABM) and a finite element method (FEM). This approach allowed to simulate the shape heterogeneity of individual cells within a monolayer, and also the collective tissue dynamics in two scenarios, uninfected and infected epithelial monolayers. This model is used here to investigate how the shape of the infection focus influences the intercellular spread of bacteria. We found that the initial shape of the infection focus determines whether uninfected cells will mechanically outcompete infected ones; more rounded shapes facilitate the ability of uninfected cells to combat the infection. There was good agreement between the in vitro experiments and the in silico simulations in terms of area and stress distribution (Figure 1). We were also able to analyse a wide range of infection focus shapes, such as triangular, circular and quadrilateral. Thus, the hybrid framework serves as a robust predictive tool for studying epithelial mechanobiology, allowing in silico exploration of how physical factors influence the host's ability to prevent bacterial dissemination.