[Experimental study on the evaluation of antineoplastic effects in the human cancer/nude mouse system]

Human tumor/nude mouse system is expected to be an antitumor screening system with higher predictability of the clinical effect. To establish this, it is quite necessary to fundamentally investigate the optimal treatment regimen in this new model, nevertheless few papers related to these subjects are currently, available. In this preliminary report, we examined the effect of several known antitumor agents against human breast tumor xenograft, MX-1, with reference to the dose, route and schedule of administration of drugs in comparison with the clinical regimens. As a result, the treatment regimens were desirable to be similar to that clinically done: the route of administration should be IV or PO, instead of IP often used in usual animal experiments. The selection of dose would be the most important factor, that is, the maximum tolerated doses (MTD) of mitomycin C, vincristine, etc. were higher than the clinical doses, even in terms of dose per square meter. Thus, for these drugs the risk to overestimate the effect might exist. On the other hand, the case was opposite as for 5-fluorouracil, because its MTD in the nude mouse was lower than the clinical dose. To evaluate the effect of a drug, it appeared to be reasonable to make account of not only the average percentage of tumor growth inhibition but also the statistical means such as Mann-Whitney's U-test.