Multiscale modelling of vascular dependent tumor growth in hepatocellular carcinoma (HCC) under therapy with transarterial chemoembolization (TACE)

Background and Aims: TACE is an established option in advanced HCC. Despite complete disruption of blood flow by imaging, a distinct number of tumor cells survive that are responsible for tumor regrowth. To investigate the influence of different parameters that determine TACE efficacy, a multiscale model was developed to study the evolution of vascular supply and angiogenesis. This mathematical model is informed, parameterized and directly validated using clinical data, based on 43 patients. The aim was to get new hypothesis to improve the efficacy of TACE.

Methods: Development of a 3-dimensional multiscale model, which couples blood flow, angiogenesis, vascular remodelling, nutrient/growth factor transport, movement of, and interactions between, normal and tumour cells, and nutrient-dependent cell cycle dynamics. In 48 patients volume perfusion CT was performed within 24 hours before and after TACE with doxorubicin loaded beads (DEB) for a total of 73 lesions. CT was repeated 90 days after treatment. Response to TACE was classified as complete (no residual perfusion), partial (incomplete embolization) or no response (no change).

Results: The multiscale modeling, based on 73 TACE treated HCC lesions, simulates the occlusion of tumor-supplying blood vessels by DEBs, leading to a local break down of the vascular system and induction of large hypoxic regions with survival of only few cancer cells. DEBs release doxorubicin that additionally targets cancer cells. The high degree of hypoxia and the induced release of VEGF stimulates angiogenesis, leading to a stepwise tumor revascularization and regrowth, which is dependent on several identified parameters. Subsequently, data of patients with a second TACE were included. The simulations define relevant parameters that determine the outcome of TACE, and, of note, already after two TACE cycles the tumour is subdivided into clusters, that should be considered in further therapeutic strategies.

Conclusions: We have set up a simulation model supported by clinical data from 73 TACE treated HCC lesions in 43 patients to study TACE-induced hypoxia, necrosis and stimulation of angiogenesis. Based on these results an improvement of current TACE application strategies will be discussed.