In terms of finding studies, I did go through a little bit this week -- I didn't know we'd discussed phase I eariler -- but we used to talk about a maximum tolerated dose, that issue of when the white count's too low to go any further. But now we're talking about potentially doing something called a target effect dose, when enough of the drug is being given. And there's some real important reasons to do that, to establish that relationship, because there are agents that may have no toxicity at any dose that can be given. {Endostatin} is an example of an agent that we really couldn't find a dose in a mouse that you could give that would make the mouse sick. It's a very safe drug.

For some expensive agents, like G3139, that are very expensive to make -- antibodies or antisense molecules, they may have useful effects at doses that aren't toxic and that would be a lot cheaper to give than if you gave ten times more of the dose. The typical relationship that a pharmacologist looks between how much drug there is and how much response is usually some version of this S-shaped curve. And monotonic means it steadily increases until you finally get to a maximum effect. Some drugs don't have any maximum effect -- the more you give, the more cells you kill but you start killing cells you don't want to kill, like white cells and immune cells and things like that.

But for interfueron, in fact, you get to the peak of the useful effect and if you keep increasing the dose, the effect actually drops off and you get some other effects that you don't want but you don't get more of the effect that you do want, you're actually getting less of it. So establishing what that right dose is difficult