The choice of experimental models in cancer research: the key to ultimate success or failure?

The technologies available for assessing the size and physiological status of malignant tumours are becoming more and more sophisticated. Some of these detect phenotypic biochemical alterations resulting from the genetic changes associated with malignant transformation. Many, however, monitor the abnormal relationship between the tumour mass and the host, detecting pathophysiological changes which result from the host/tumour interaction. The most important of these involve the microregional heterogeneities of nutrients resulting from diffusion patterns around abnormal, newly-formed vascular networks, or inflammatory or immune responses of the host to the tumour. In order to validate new technologies for introduction to clinical medicine they are usually extensively tested in experimental models for cancer. The relevance of these models (whether in vitro cell cultures or experimental tumours) must be considered carefully at the outset, with the factors that influence the parameter being measured, e.g., oxygen or energy status, being given special attention. The genetic similarity of tumour and host, the site of tumour growth, the size of the tumour, the burden it imposes and the immunocompetence of the host are all important features of the experimental model. Inappropriate models may give totally misleading information which cannot be extrapolated to human cancers.