Chapter 1 - Preclinical Drug Development: Translating Basic Research into Clinical Work
Once studies in cell lines have shown that a new agent has anti-proliferative properties and is able to inhibit its target, in vivo studies in experimental animal models are undertaken to further define the anti-tumour activity and provide pharmacology and toxicology data needed for the subsequent clinical development. The anti-tumour effect of a new agent must be evaluated in vivo and, for those agents in which the target is known (or believed to be known), efforts should be made to show that the observed anti-tumour effect is related to target modulation and to establish if a dose-dependent relationship exists between target inhibition and the observed anti-tumour effect. Pharmacodynamic endpoints used to define target inhibition in vivo may vary based on the target and the mechanism of action of the drug (e.g. measurement of substrate phosphorylation for kinase inhibitors, measurement of mRNA and protein levels for small oligonucleotides, etc.). More recently, imaging techniques have also been used to detect the effect of some new agents in animal models (e.g. detection of angiogenesis inhibition using dynamic contrast-enhanced magnetic resonance imaging).
In addition to pharmacodynamic measurements, pharmacokinetic studies provide information about drug absorption, metabolism, excretion and plasma-protein binding. Safety pharmacology and toxicology studies are also performed in animals. The objective is to estimate a safe starting dose for first-in-human phase I studies, assess toxic effects with respect to target organs and help to select different dosing regimens and dose-escalation schemes for clinical studies.
The choice of starting dose for first-in-human studies is usually based on toxicology evaluation in both rodent and non-rodent (dog or monkey) species and the most sensitive species is chosen for safe starting dose determination. The US Food and Drug Administration (FDA) and the European Medicines Agency (EMA) recommend that new anti-cancer agents be evaluated in both rodent and non-rodent species before undergoing human phase I evaluation. One tenth of the lethal dose to 10% of mice (LD10) and one sixth to one third of the lowest dose that results in no toxicity (TDL) in non-rodent species are some of the parameters that have been most frequently used to select a starting dose for many anti-cancer agents. While these methods have been widely used for cytotoxic agents, their ability to predict a safe starting dose for molecularly targeted agents is debated, and thus far it is not clear which animal models could better predict a safe starting dose. In addition, no standard parameter exists and for most new agents a multitude of parameters has been used to determine a safe starting dose. In our recent review of first-in-human studies of molecularly targeted agents, only 3.7% of phase I trials had a starting dose that exceeded the maximum tolerated dose, providing evidence that, with the exception of a very small proportion of new agents, the choice of the starting dose has been generally safe. A better understanding of the target and mechanism of action of a new agent can help to select the animal models that would best predict for toxicities in humans.
The most important question regarding the in vivo study of a new anti-cancer agent, however, is represented by the likelihood that the anti-tumour activity observed in animals may translate into a clinically significant efficacy. Substantial controversy exists as to the best animal model that would positively predict for anti-tumour activity in humans. Generally, there is no single system that is considered the best positive predictor of anti-tumour activity for human tumours. Xenograft tumours implanted in immunodeficient mice by subcutaneous or orthotopic inoculation of tumours (grown in vitro or obtained from patients’ tumour biopsies), have served as a model for the evaluation of a large number of anti-cancer agents. Xenografts have several limitations (e.g. a low tumour establishment rate for many human tumours, low reproducibility of “real” cancer with respect to surrounding tumour environment, and growth rates that do not mimic the ones in human cancer, among others), but they represent a valid model and they have contributed to the identification and development of many new agents. More recently, the possibility to obtain genetically engineered mice that recapitulate a specific cancer genotype has opened new horizons in the preclinical evaluation of new compounds. The discovery that genes with either oncogenic or tumour-suppressor activity may be altered in human cancer, and the possibility to introduce these changes by various techniques into mice, raise the possibility to study the anti-tumour activity of a new agent against tumours that more closely recapitulate the biology of human cancer.