Good morning I'm Harvey Clewell. I work at a company called ScitoVation. Just recently we formed this company but prior to that I was at the Hanover Institute for Health Sciences in Research Triangle Park, and I'm going to be talking about Biokinetic Modeling of Toxicokinetics. I have a number of different areas that I'm going to talk about related to that. I'll show you on the next slide the learning objectives for this particular lecture. First of all, I'm going to talk about biokinetic modeling itself and what it is, some examples of it, and focusing primarily on how it can be used when you're doing a risk assessment based on in vitro assays. The next block will be on understanding quantitative in vitro to in vivo extrapolation. We like to call that QIVIVE. And then we'll go on to talk about an approach that has been used recently by us and the EPA ToxCast group to apply biokinetic modeling, a very simple form to support in vitro high throughput screening. And also I'll talk a bit about how it can be used for more sophisticated analyses. And then, finally, we'll talk about the research that's still needed in order to improve the way that biokinetics can be incorporated into these kinds of assay extrapolation efforts. So let's go to the first section on Physiologically Based Biokinetic Modeling. And I do want to explain that there's actually multiple names for biokinetics, toxicokinetics, pharmacokinetics. I'll say about that in a minute, but whatever you want to call it, the role of kinetic modeling in the past was to relate animal doses to equivalent human exposures. This figure here demonstrates how those various things are related. On the bottom left you see human exposures to chemicals in the environment. They produce blood levels, and these days a lot of different organizations, CDC in particular with the enhanced data set, are collecting information on chemical concentrations in blood. And so one of the things that we need to be able to do is to be able to- if someone has a certain concentration of a chemical in their blood, can we estimate how much they've been exposed to? Another thing that we want to do is to be able to relate that concentration to animal studies. Unfortunately, animal studies are conducted with administered doses and very, very, very seldom have any measure of the chemical concentration that's achieved in the blood of the animals. So we have to use biokinetic modeling in order to estimate the blood concentrations in animals for a given dose in a study, so that we can compare effect levels in animals with the concentrations measured in biomonitoring studies. So this kind of work has been going on for many years in both pharma and in the environmental world. '70s and '80s was really when these kinds of modeling approaches were developed. But things are changing now because of the pressure, particularly in Europe, to try to find alternatives to live animal testing. And so, in the future, we hope that we will be using in vitro assays to determine what are the concentrations at which a particular chemical might interact with normal function of cells. And so what will be needed then will be some way of relating the concentration in the in vitro assay concentration in the media, for example, to a daily dose that a person would need to achieve that concentration so that we can give a basis for a person judging whether an exposure is safe or not. And so, as I said, this process is called quantitative in vitro to in vivo extrapolation. Now, the physiologically based biokinetic modeling or pharmacokinetic modeling or toxicokinetic modeling, it's all basically just a matter of what group is doing the modeling that determines what they like to call it. But the purpose of a physiologically based biokinetic model is to define the relationship between an external measure of exposure and an internal measure of exposure, which we hope is more directly related to a biological effect. And this diagram shows a simple illustration of what a physiologically based model looks like, and it uses the anatomy of the mammalian system. You can see that there is a lung for inhalation. There's rapidly and slowly profuse tissues. What that means is some have a richer blood flow and some have less of a blood flow per gram of tissue, and that makes a difference in terms of how quickly a chemical can get in and out of that particular compartment. That is a very lipophilic tissue, and so lipophilic chemicals, those that like fat, will build up in higher concentrations in that tissue. Liver is important tissue for clearance of chemicals. Unfortunately also, in some cases, metabolism produces a more toxic chemical than the one that was inhaled or ingested. Often, a reactive metabolite can be formed from oxidation, for example. And so you can see that the blood flows, the lines are showing in the blood flows and you can see it's parallel flow, which is the way that the mammalian system is designed. And so then these models are built up from physiological structure, from anatomy, from descriptions of metabolism, and transport processes that are known. And then the parameters have to be put in based on data that people have collected on the physiological volumes and flows and also biochemical data on how chemicals partition and how they're metabolized. And that all then becomes just a bunch of equations, a system of mass balance differential equations to be specific. And there's one equation for each tissue and it's connected by the equation for the blood. That's a system that has to be solved simultaneously. An example of how these equations look, and I won't have many equations so those of you who don't like equal science can take some comfort, this equation says that the change in the amount of the chemical in the liver over time is equal to the blood flow times the concentration coming in in the arterial blood minus the blood flow times the concentration leaving in the venous blood from that tissue, and then minus another term, which represents metabolism. And some of you will recognize this equation as the Michaelis-Menten equation. You have a V max, which is the maximum capacity for metabolism, times the concentration. You can see the concentration is always adjusted for the partitioning. And then divided by K M, the affinity constant, and plus a term, which is the concentration. So that's the Michaelis-Menten form. As I said, there won't be many more equations coming, but it's good for you to know how these models are built. The modeling process is useful for doing things in risk and safety assessment processes that involve trying to go from one system to another, whether it's from an animal system to a human system or from an in vitro system to an in vivo system. So in vivo toxicity is affected by biokinetic factors that are not appropriately reflected in in vitro toxicity tests. It's easy enough to think of things that don't happen in vitro that do in vivo. Absorption. Instead of pouring a chemical into the media, absorption in in vivo involves getting through the intestine into the blood, and then being distributed to the other tissues. Metabolism occurs in the liver. Many in vitro systems don't have metabolic capabilities. And then there's excretion both in the urine and in the bile and through the skin into the hair. There are many ways that the chemical can leave the body. Whereas, in vitro, it's pretty much stuck there. So we have to use models to go from the one environment to the other, and PBBK modeling serves as a tool for doing that. This is a busy slide, but I just want to show you that there have been some really important steps taken recently. On the left, you see a fairly complicated physiologically based biokinetic model that has a lot of compartments because it's an attempt to be a generic model, one that would work for different chemicals that might have different target tissues where the toxicity occurs, that might have different properties that mean that they will accumulate in different tissues. And so this is a general model that can be used by putting in parameters there for a particular chemical. A lot of work now is trying to develop things that people could use when they do in vitro assays that then could be applied to do the in vitro and in vivo extrapolation. And this was an example where the University of Utrecht Medical Hospital did a study to determine for acute releases whether these models could actually predict the kinetics in a short term exposure. And they had some chemicals where there had been human studies done, and the model worked very nicely. These are things like methylene chloride and toluene and benzene and acetone. In all the cases, the model does a pretty good job of predicting exposures over a period of a few hours. And so, for short term exposures where the concentrations rise and then fall, it's necessary to use this kind of a physiologically based model because you have to model all these processes in real simulated time. That's the end of the first section. Just background on PBBK and its uses for in vitro and in vivo extrapolation. When I come back I'll talk further about what do we mean by monetative in vitro to in-vivo extrapolation.