Abstract
Primary liver cancer remains a global health challenge due to rising incidence, limited curative options, and poor overall survival. Poor outcomes stem from tumor heterogeneity, limited efficacy of current therapies, and comorbid chronic liver disease. Despite recent advances in immunotherapy and combination treatments, response rates remain low, and predictive biomarkers are lacking. As a result, there is an urgent need for preclinical models that capture the molecular, cellular, and immune landscape of primary liver cancer. This review discusses the strengths and limitations of patient-derived models of liver cancer, including two-dimensional patient-derived cell lines (PDCL), three-dimensional (3D) patient-derived tumor organoids (PDTOs), and patient-derived xenografts (PDXs). While PDCLs and PDTOs enable high throughput studies, they lack a representative tumor microenvironment. PDXs, including PDXs in animals with humanized immune systems, may more effectively mimic tumor-environment interactions but are costly, complex, and still contain mouse stromal cells. Ex vivo tissue culture preserves tissue structure and cell-cell interactions in an immunocompetent environment; however, short duration of viable culture limits broader application. Continued innovation in the development of multicellular three-dimensional culture systems and in vivo humanization strategies will play a critical role in enabling the development of more personalized and effective therapies for primary liver cancer.