|
|
|
|
|
|
|
empiric use in classical toxicology (such as IP or repeated IV administration). This has resulted in both a reduction of the number of animals used and in some cases a substantial reduction in the number of patients needed for completion of the Phase I trials [8]. |
|
|
|
|
|
|
|
|
This system is not ideal since rodents (and other lower species) have quite different biological and biochemical profiles to humans. This problem may be circumvented when we know the mechanism of action of drugs, but potentially could be a major problem for drugs with unknown novel mechanisms. The best example of this phenomenon is in the clinical development of thymidylate synthetase inhibitors (TSI). Thymidylate synthetase is the major focus of activity for 5-fluorouracil, and many new agents have been synthesised to produce more efficient TS inhibition. The toxicity of TSI is dependent on extracellular fluid thymidine levels, and since this is 10-fold higher in mice compared to patients [9], the use of murine toxicology would be entirely inappropriate. |
|
|
|
|
|
|
|
|
As we head towards drugs that are not DNA-targeted and that have more specific enzyme inhibitory function, it is to be hoped that we will adapt our strategies even further. Conceivably, Phase I trials could be performed with dose escalations and starting doses based on expected enzyme inhibitory activity. |
|
|
|
|
|
|
|
|
It is clear that the days of extensive animal testing of cytotoxic drugs are now behind us. Our aim should be to limit this phase to the minimum needed for safety in human studies. |
|
|
|
|
|
|
|
|
IV. Methodological Issues in Phase I Trials |
|
|
|
|
|
|
|
|
Phase I trials in oncology are unique in that healthy volunteer studies are entirely inappropriate because of the cytotoxic and genotoxic nature of the drugs. This leads to a selection bias for patients that tends to reflect the investigator's practice or prejudices, although standard guidelines do exist that serve as a template for most Phase I protocols [7]. |
|
|
|
|
|
|
|
|
Although traditional Phase I trials are predominantly toxicology in humans, the temptation remains to bias patient accrual to those tumor types in which activity is expected (e.g., colon cancer for thymidylate synthase inhibition). The demonstration of a hint of activity in Phase I will often be the deciding factor in the progression to Phase II. Hence, there is an understandable desire to select those more likely to respond. This may be reasonable for a TSI, but what of a compound with an entirely unknown or novel mechanism? In my opinion it is more desirable to include as many tumor types as possible. This introduces an element of randomness that may not be inappropriate, since it is not inconceivable that a TSI may affect renal cancer or melanoma that may not traditionally be seen as 5-Fluorouracil sensitive. |
|
|
|
|
|