Typically, cells are grown in monolayer on flat polystyrene or glass dishes in two dimensions (2-D culture) in growth medium at an appropriate temperature and gas mixture (typically, 37 °C, 5 % CO₂ for mammalian cells) in an incubator. Growth medium can vary in pH, glucose concentration, nutrients, and growth factors (often derived from calf serum). Adherent cells can be genetically manipulated, treated with drugs, and observed using microscopy. A major criticism of 2-D culture is the assumption that accurate physiology is reproduced in a cell monolayer. Molecular and cellular responses differ significantly from the organ or tissue of origin due to a lack of extracellular matrix (ECM), which mediates cell-cell and cell-matrix communication in vivo.

Three dimensional (3-D culture) is more physiological allowing cells to grow and respond to environmental cues within an “organotypic” structure resembling ECM (based on collagen, elastin, and other ECM proteins). 3-D spatial cell-cell and cell-matrix interactions emulate in vivo cell growth and response to drug therapies. Spheroid culture is the simplest 3-D culture technique and does not require an ECM scaffold. By single culture or co-culture of more than one cell line (e.g. hanging-drop, rotating culture, and concave plate methods), 3-D cell spheroids can be generated to model solid tumour growth. A number of biocompatible synthetic and naturally-derived scaffolds with specific surface chemical properties can be used to model the ECM environment, and examine in vitro cancer cell biology. Long-term maintenance of large biological cellular structures requires bioreactors, which continuously supply nutrients/oxygen and remove waste products/metabolites. Although, co-culture models facilitate the study of interactions between different cell types (e.g. epithelial-fibroblast interactions), these models lack additional cellular and humoral factors in vivo.

Tumour tissue or immortalised human cell lines from one species can be ectopically (in another tissue or organ) or orthotopically (in organ of origin) transplanted into another species (usually an immune-deficient mouse). Common ectopic sites include subcutaneous flank, mouse tail vein, and renal capsule. Although xenografts allow rapid and facile assessment of in vivo tumour dynamics, these models have several shortcomings: For example, cell lines ectopically-transplanted in immune-compromised hosts tend to grow more rapidly than human cancers, yielding inaccurate modelling of response to anti-cancer agents. Orthotopic transplantation is thought to be more physiological and enhances the chances of metastatic spread. However, ectopically- or orthotopically-transplanted tumours are not genetically heterogeneous (due to clonal expansion or over-representation of a particular cell type) and lack normal architecture or microenvironment observed with autochthonous tumours (reviewed in [3, 4]).

Cancers can be induced in animal models, typically mice, by a variety of environmental carcinogens including radiation, chemicals, and pathogens. Tumour development is highly-reproducible, so these models can be used to screen potential carcinogens and chemo-preventative agents. However, these animals exhibit variable tumour susceptibility (dependent on the genetic background/strain) and latency, and do not always develop the full range of pathological tumour subtypes.

Genetically-engineered models allow the spontaneous development of cancer in an immune-competent host, which accurately mimics the molecular, pathological and phenotypic features of human disease. Cancer can be caused by an accumulation of genetic insults over time, and these models allow individual analysis of oncogenic pathways. By integrating (or crossing) different genetically-engineered models, the combinatorial effects of multiple genetic defects can be studied in vivo. These models may also be used to test novel therapies specifically targeted to molecular pathways of interest, representative of genetic changes in human disease.

Cell lines derived from or embryonic tissue is isolated from a particular species is genetically manipulated, and subsequently transplanted ectopically or orthotopically into a genetically identical and immune-competent adult host. Advantages of this model include the genetically-identical backgrounds of graft and host, and an intact immune system required for anticancer immunotherapy experiments.

Transgenic animals, usually mice, express oncogenes or dominant-negative tumour suppressor genes (transgenes) under the control of promoter and enhancer elements that drive gene expression in the tissue of interest following random integration into the genome. Specific gene expression within cells constituting the tissue of interest subsequently drives a particular phenotype. By incorporation specific genetic elements (operons) into the transgenic system, transgene expression can be reversibly switched on and off using exogenous ligands (e.g. tetracycline/doxycycline or tamoxifen). The use of synthetic promoters/enhancers and random integration into the genome can result in incomplete penetrance and phenotypic effects that are unrepresentative of endogenous gene. Transgenic animals are typically created by the following methods: