 Thank you, well, I'm going to lead you away from the skin to the liver and talk about the drug induced liver injury network and a little bit about Stevens-Johnson syndrome that we've seen there. So drug induced liver injury accounts for about 3 to 10 percent of acute jaundice liver injury that presents to physicians in the United States. It's the single major cause of acute liver failure in the U.S. and most of the western world. It's a common cause for medication to be abandoned during development and a common cause for a medication to be withdrawn or restricted after it's approved. It used to be number one, now I think it's number two. And liver injury frequently accompanies Stevens-Johnson syndrome. Of course it's usually not a very major problem and the skin problem is somewhat overwhelming in the clinical presentation, but it's often there. Here's a network that we've been running for almost 20 years of acute liver failure in the United States. They've accumulated 2,000 cases and this is the etiology. The number one cause, the cause of about 50 percent of cases is acetaminophen, Tylenol, usually an overdose. Sometimes it's an unintentional overdose. The second most known cause are drugs and that's not one drug, it's not acetaminophen, but it's many drugs. And the other causes are shown there, hepatitis B for instance is disappearing as a cause of acute liver failure in America and hepatitis A is almost gone because of advances in those diseases. Over on the side are indeterminates where we don't know the cause, it's the real challenge that we have. But drugs are an important cause. When we talk about drug-induced liver injury, they're usually falling into two categories. One is a direct toxin and the other is, it's not really indirect, it's idiosyncratic. The direct toxin means that the drug is intrinsically toxic. Injury is frequent, it's dose related, it comes with high doses of the drug, it's reproducible in animal models, it's expected. So that's acetaminophen basically there. But the more challenging group here are those that are idiosyncratic, where you have a drug that's not intrinsically toxic. It causes liver injury, but the liver injury is rare, maybe 1,000 people or 1 in 10,000, maybe 1 in a million. It's not particularly dose related. It's a little bit dose related, you have to take enough of the drug. It's not reproducible in animals, it's unexpected. And of course there's the challenge, because how do you study something like this from a research point of view when it's so uncommon? So idiosyncratic is unexpected. Here's isonizid, the most common cause of fatal injury from a drug. It causes hepatitis in about 1 in 500 people. The most common cause of jaundice from drugs is amoxicillin-clavulonic acid, it's estimated about 1 in 2,000, 1 in 3,000. And declofenac, a very important cause is 1 in 30,000, it's quite unusual. The idiosyncracity, we talk about idiosyncratic, it can be, we think here of course hypersensitivity, immunologic. It may be metabolic, or it may be a combination of the two. It presents usually as an acute hepatitis that can be hepaticellular, or it can be colostatic or in between, so etiology is generally unknown. But one form of this is what we call immunologic hepatitis, it's our name for a drug hypersensitivity reaction that affects the liver. It usually occurs with rash or fever or one of these signs, typically with a short latency. The rash and fever may precede the liver manifestations, and the injury is typical hepaticellular, like viral hepatitis initially, but then can evolve into a colostatic picture. And it has many names, the syndrome, I've heard many of them today, can have hepatitis with a simple rash, immunologic hepatitis, drug induced hypersensitivity syndrome, which is probably a better name. The neurologists use the term anti-convulsant hypersensitivity syndrome, the dermatologists use the term DRESS, and of course the extreme forms are there. Now what we found about 10 years ago, I was asked to look at the NIH portfolio in liver disease in general, and I gathered grants from all the institutes and looked at them and put them in piles, you know, here's viral hepatitis, here's basic research, here's genetics. And one of the piles was drug induced liver injury, and we had a sizable portfolio. But looking at the grants, they're virtually all in direct hepatotoxicity. We were not funding anything on this idiosyncratic form. So what do you do when you have a cap on research? You do what you're doing today, you have a meeting. And here's the meeting that we held in 2000. I think you'll find a young Neil Shearer there somewhere. We brought by the clinicians, basic researchers, and said, what do we need to address this issue of idiosyncratic liver injury? And they came to the conclusion kind of what we need is a network that would collect cases that we can study. And so we created the drug induced liver injury network in 2003, and the dates are important because this is during the doubling of the NIH budget. You can't do this when your budget's flat, or it's very hard. But that offered a wonderful opportunity. My boss didn't think this was a good idea. He gave me three years of funding with the idea, if we don't show anything by three years, you're out. And we squeezed it. It's now been going on a little, we're in our 12th year, I think. It's a consortium of clinical centers. They've ranged a number from five to eight, a data coordinating center at Duke, a sample repository for the DNA and the serum and the urine and the cells, and a genetics core, which is also at Duke, North Carolina. And the aim of the network was to collect and fully characterize cases of clinically apparent drug induced liver injury to allow for mechanistic studies into its etiology and potential prevention treatment, so forth and so on. Here are the centers that have been involved. And we've gone through several re-competitions. They've been open, so we have new centers. These are the centers we have at present in black. There are six. If you notice, the one in California, it's floating out to sea. It's moving to the left there. But in the ones in green are the former ones. So the DCC is at Duke. It's been a great group. We have all the experts in drug induced liver disease have been involved. It's been very exciting. And here's the enrollment. We targeted to have two patients per center per month. And we've been right on target. We have a dip every time we recompete. So what we faced immediately was this problem, causality. How do you know the drug has caused the liver injury? Because drug induced liver injury basically can mimic any kind of liver disease, viral hepatitis, for instance. What you usually need is a compatible history. You need to rule out viral hepatitis, other causes of liver injury. Imaging is important to make sure there's not biliary obstruction or a mass. And what's most important is that you're dealing with a known cause, a drug that's known to cause liver injury, and it has a compatible signature, clinical signature, or phenotype. But that's not always possible. There are no specific tests to prove causality. So rely upon subjective determinations. We decided to use a five point scale. And assessing causality, one to five, definite, very likely, probable, possible, and unlikely. But what do those terms mean? We use them all the time. The FDA uses them. But what they mean? So what we tried to do was standardize them to give some objectivity to these subjective judgments. First of all, we use a percent. Definite isn't 100%, but it's greater than 95%. And very likely is 75 to 95, probable 50 to 7, so forth and so on. Actually, our Justice Department, the law courts have been dealing with us for a long, long time, actually. Is he guilty or is he not? Did he do it? And they have these very nice legal descriptions that are helpful. Definite, it means beyond a reasonable doubt. And that's a high level of evidence, difficult. Very likely, clear and convincing, probable the preponderance of the evidence is in favor. And if it's below that, it's not guilty. It's no. So that separates yes from no. Now, beyond a reasonable doubt, that's a tough level of evidence. And as you know, there have been famous law cases where a patient was declared not guilty. And it was because the level of evidence was so high. We don't have to do that here. We've had about 1,000 cases enrolled. And these were the first 1,899 were adjudicated as, in this area of yes. But they were caused by 250 different agents. So we're not dealing with an uncommon disease. We're dealing with 200 rare diseases. Prescription drugs were 84% herbals and dietary supplements. The 10 most common accounted for a third of the cases due to drugs. And here are the most common, the 20 most common I could fit on the slide. And you can see that several of these have come up here as causes of Stevens-Johnson syndrome. Sulfonomides, phenytoin, lomoctrigen, allopurinol. So they're there. And did we see Stevens-Johnson syndrome actually? Supposedly we have nine cases. I have to say that these were hepatologists making the diagnosis, not so. And looking over the cases myself, I think two of them aren't quite Stevens-Johnson. One is a pretty bad erythema multiformia. And one was Dress. But look at the most common causes here. Lomoctrogens up here. Only one case of carbamizapen was fatal, though. And the overall fatality rate with this diagnosis was high. So it's bad. And only one of those four deaths was due to liver disease. The other were due to multi-organ failure due to the complications of Stevens-Johnson. Here are the SGS cases. Here are all of them. You can see what do they look like. Are they different? Yes, they are different. They're younger, a little more likely to be woman. But here's a very interesting thing. And that's the racial diversity that's very striking in the Stevens-Johnson syndrome cases. A third were black. It's a small numbers. But in the whole population, more matches the population of the US. Short incubation period, sick, high degree of fatality. With drug-induced liver disease, the fatality is about 10%. This has been shown by other groups. This is our group from the United States, Indianapolis. This is a group from India. And you can see that Stevens-Johnson's was a more common cause in their series of drug-induced liver disease. But the clinical features were similar, different drugs. But those are drugs that are used more commonly in India. So lamocrogen accounted for 12 of our 900 cases. Looking at these, 11 of them had either dress, Stevens-Johnson syndrome, or hepatitis with rash. So this suggests to me that dress is a milder form of Stevens-Johnson syndrome. It's caused by the same drugs, 25% African-Americans. Again, with this drug as well. Latency is short, like with Stevens-Johnson. So the spectrum of drug-induced liver disease goes from acute liver failure, but below this are asymptomatic rise and serum enzymes. And the question is with the skin, are all drug hypersensitivity syndromes, are we talking about really just the tip of the iceberg? And would study of this also be helpful and reflect on it? We'll give you more cases, certainly. So the other thing I want to point out is the nine cases with Stevens-Johnson syndrome are also exposed to multiple other agents. The average was five. And look at the ones that are the also. These are the concomitant drugs, not the ones that are accused. Many of them have also been associated with Stevens-Johnson syndrome. So the causality is a challenge. So the other thing that our initiative in drug-induced liver disease has done is developed a website on drug-induced liver injury called liver tox. It's sponsored by NIDTK and the National Library of Medicine, in which you can go to this website. There's no ads. There's no price. You don't have to register. And you can look up a drug. We look up background of drug-induced liver disease for its association with liver disease. We include all drugs that cause liver injury. We include drugs that don't cause liver injury. Sometimes the negatives are striking. And this is the design of it. Each drug section has an overview, has representative cases. There's about 1,000 cases on the website. Have liver histology in a few, we're adding that now. The chemical structure, a link to the product label, and annotated references. Here's Declofenac, case report. Also, people can submit case reports to liver talks. And when they're submitted, you can search and find them from this like Declofenac thing. We'll search and find any cases of Declofenac that are in our 1,000, or it's actually 500 now in this, from the outside. Eventually, all of the cases from Dylan will be enrolled in this so that you can look at them yourself. References, there's the status now of liver talks. It was been released about 2 and 1 half years ago. We have about 100,000 unique visitors per month. We have 780 agents described. It's over a million words, 13,000 annotated references and 1,000 cases. So something to think about as a format to looking at a network and bringing attention to an area of serious drug abuse, serious drug adverse reaction. Thank you. Thank you. The floor is open for question. You made a comment yesterday that we need to stay focused and not do all adverse drug reactions. There's some clear commonality, or whatever the even bigger sense of the word is, between what you just talked about. I don't know whether this seems true with kidney disease or others, but where would you think the line should be drawn in terms of a network going forward that includes what you're doing, but also has some clear demarcation lines? Well, the problem is it requires the scientific expertise. And so I couldn't add skin problems here, because I'm not an expert. I'd have to get Neil. I'd have to be on the phone with Neil every day. So it does require a lot of input from someone who can dedicate themselves to this. I guess the issue, people have called me, why don't you do lung problems and kidney injury and so forth? And I don't have that expertise. With all these drugs, it's very difficult. And some of the most difficult things is to talk about drugs that don't cause liver injury, because you have to search the literature, and it's not easy to search when you're looking for a negative. I think your point, as well as some of the others, including from the advocacy groups, really highlighted this systemic disease. Oh, yes. Then you're looking at the liver, someone else is looking at the skin, someone else looking at whatever else. Maybe we do need to be more cross-cutting than we have been. Regardless of the way the NIH is currently designated. Right. Well, for instance, the mortality rate that we say of 10%, well, that's not liver, all liver deaths. Those are other multi-symptom deaths or other problems that come along. Even complications of the evaluation of the patient, unfortunately. Sometimes, it's a cause of death. Last question, please. Well, thank you very much for elucidating the dilly issues for us. It was very helpful for me to understand your perception of dilly and dress. But I'm not sure if I got you right. Did you say that you think dress is a minor type of Stevens-Johnson, or did I get that wrong? No. OK. So we are clear that dress is something completely different than SJSTN, and I'm happy. Thank you. Well, they're caused by the same drugs. I don't know. This is what the question I asked Neil, actually. That is true. They are caused by a number of drugs that are the same, allopurinol and carbamazepine, the most frequent ones we found in dress. I didn't talk about it because this was not the dress workshop, but I think in terms of clinical definition and disease entity, it's different conditions. Right. So the question is, are there drugs that cause dress that do not cause Stevens-Johnson syndrome? And that's what I'm talking about. When you're looking for a negative, it's sometimes very hard to do. And as the FDA said, they have these warnings, but they can't put out a statement that says, this drug doesn't cause prolongation of the QT interval. And that sometimes would be helpful to say that, no, this doesn't cause it. Or this doesn't cause drug-drug interactions. And it's difficult. Or you can also say, I don't know, which we don't like to say sometimes. Thank you. Because of time, we have to move on. So the last talk is given by Dave Finstra. He's a professor at the Department of Pharmacy University of Washington. And he's going to talk about the health, economics, policy, implementation, and cost-effectiveness research.