Acetaminophen Liver Damage

>> okay, good afternoon. today's topic of drug induced liver injury's another kind of classic example of what we try to accomplish by having this series of talk. it's an example that bridges pharmacology and chemistry in the beginning.

and i guess you might call naturopathic medicine. and then it spills over into the use and development of drugs, the use of drugs for a variety of laudable purposes. and then realization what was initially thought to be safe wasn't safe, at least not for

everyone. and drug induced acute liver injury is a major problem in the united states. and for the most part of the world except perhaps the truly under developed world. so this is something that has been accelerated by the

commercialization of so-call natural foods, the sort of thing you buy in expensive bottles in the health stores. ancient remedies that are claimed to cure everything that one can think of but which has no known quantitative composition and bury one bottle

from the next. at any rate, are all of that separately and collectively has the capacity to produce serious liver injury and death. so you're going to hear today from leonard seefe. leonard graduated from medical school in south africa, came to

this country, trained in the washington area, was here at the nih and niddk for quite a few years, and is one of the leading figures in assessing the risk of drug-induced liver injury, how to predict who gets what, where. not so much how but the occurrence of it.

and what can be done. the experience is vast and he's a major player in a large national international network where cases some of which are quite rare with different drugs are collected and material is put in a reservoir of tissue, blood, history, everything else

for more reductionist translational, if you will, kind of studies. our second speaker is chris austin. chris graduated from harvard medical school, was trained in neurology, worked at merck for quite a few years.

some really fascinating work on studies of a genetic form of schizophrenia. came here to nih in the genome institute and has escalated up the stairway into more complex positions and the most complex of which she will tell us about is the director of the newest

institute, the national centerr translational research impact. so, quiz time. does anybody know who this person is? anybody else know who this okay, who is it? julian axelrod. now, are what is the structure?

that's tylenol, acetaminophen. why do we start with this? it gives historical background where we are. now julius axelrod is a ljened at the nih. he was trained as a chemist. he didn't get a ph.d. until he had written something like 50 or

60 major papers. he just didn't have time for that sort of stuff. the story is that the nih could not make him a lab chief if you didn't have a ph.d. so they arranged for one at georgetown and submitted some of his papers.

julius was a chemist, and he very early on worked on ways of studying the metabolism of drugs on the basis of chemical basis, based largely on solubility, lipid soluble versus water soluble and so forth and so on. as a result of that, he really discovered pathways of drug

metabolism. and i'm going to show you that in a moment. so he then became very influential in the pharmacology community with the discovery that for example one drug can influence the metabolism of the others.

but his real thing came when he applied that study of chemistry to the first detection of neurotransmitter. and that's what he won the nobel prize for in 1970. in his early work, at that time aspirin -- was the drug that people gave to their children if

they had high fever and people took if he had some joint pains or cold or whatever it was. it was like the universal panacea. years later, of course, it turned out that aspirin actually had rather serious toxicity, particularly when given the

young children with very high fevers. and they would develop mitochondrial defects in their liver and in their brain and some of them died. but nobody knew anything about the metabolism of -- or one of its derivatives.

so axelrod very early on gave it to himself. he was always walking around with a bucket of urine that he collected because he was the subject for most of his studies. and so they took, they followed the metabolism of -- and found this acetylated amino phenol in

the urine. that's the major metabolizer -- so julie looked at the formula and said wow this would be a better antipyretic an aspirin. they put it on the shelf and it sat on the shelf for many years, i think probably 30 or 40 before aspirin was shown to be toxic.

by that time there were no patent rights or anything and the company bausch i think that makes tylenol, commercialized on it and it becomes one of the most highly used drugs. now why would we start out something on drug induced liver injury by talking about tylenol.

the reason is because tylenol is one of the top ten causes of liver injury. it first came to attention when you take tylenol, most of the time you're taking a half a gram tablet, capsule or whatever at a time. if you take ten grams within

three days your liver will undergo massive necrosis, and you either have a liver transplant or you die. early on, if you happen to have evidence of tylenol liver toxicity, that is within the first few days, this can be treated rather successfully by

giving amount acid acetyl cysteine. so this you can buy over-the-counter and it's pretty cheap. the risk of liver injury is enhanced, if you drink alcohol. probably not a glass of wine or occasional cocktail.

but if you hang one on over the weekend, while you're taking more than five grams of tylenol, you're skirting with disaster. other drugs also can accelerate it. preexisting liver disease. it can require much much less than 10 grams of alcohol to

cause major liver injury. and it took a long time, close to ten years for the fda and the various groups to finally come around and put labels on tylenol, which warn against these pains and the product labeled often in very small print is still not fully

appreciated by most of the general public. now tylenol is one of the best examples of billi it's predictable. for those of you who are not particularly pharmacologists. to put this in the perspective how do we handle drugs.

what does 9 -- the liver do. this is the big picture. when you realize everything we eat whether we want to or not, ever that's added to food all kinds of stuff in our environment advertently or inadvertently, everything goes into the intestinal tract or

stuff that's made there of a toxic nature that's made there for example the micro flora we talked about, circulates from the blood to the liver. the liver is the major barrier that protects us from killing ourselves with all that stuff in the outside world and in our

food. so how does it do it? it does it by two major bio chemical processes. one an oxidated process located in the end or plasmic reticulum using oxygen and a series of cytochrome called site comb p450.

that's their absorption rate. they've been renamed. they're all the same. mixed function oxidase. what they do is they take like a chipping wire structure like a phenol, for example and they open up one of the bonds, insert an oxygen, then a hydroxyl.

so you have hydroxylated whatever it is, carcinogen or drug or whatever it is and that makes it more soluble. and then hydroxyl conjugated in reactions in the later in the liver with -- sulfate, some nitrogen and so forth. all of which rendered this stuff

highly soluble and substrates for transporters that secreted from the liver cell into the bile out of the body. and that's the major mechanism of protection against most drugs oxin. the second process is in the soluble, the site plasma of the

hepatocyte and that involves the try peptide glutathione. [indiscernible] it's the cysteine that reacts with any negatively charged compound. now when you think of it all the millions of structures of stuff that we could be exposed to, there isn't a simple separate

process for each one of those. these are garbage disposal system that operate with huge capacity but limited specificity, as little as a negative charge on a molecule will render it conjugated with glutathione and the glutathione ducts are secreted into the bile

by a specific transporter. so these adducts, the conjugates, most of which sits in the liver and much of which liver disease impairs. now the liver is also capable, these same responses of mediating disaster because when you form from a benzene ring the

hydroxyl tack on, the first step is to put [indiscernible]. you form that molecule which is highly reactive and it generates the production of reactive oxygen within the liver. now the more active the people of 50 is, the more this stuff is formed.

and most of this is protected by forming glutathione. if you don't eat a decent diet, if you're an alcoholic, if your liver is depressed in glutathione or if you've been taking a lot of tylenol which i'll show you that leads to glutathione then you a

re susceptible -- critical to cell function, particularly in the mitochondria. okay. now at this point, leonard was going to talk but he made an error, brought the wrong power stick so he went home to get the right one.

he lives only ten minutes from here and i'm sure will make an appearance very shortly. but meanwhile, we're going to change the order things and i introduce you to chris austin who is going to talk about a different but related area. >> yes.

so the first thing is seeing anything with liver in the title with everyone in the room is very intimidating so i hope you will bear with me here. i don't know if he's ever told you what his background is but he's really mr. liver or dr. liver, reverend liver, rabbi

liver, whatever you want to be. so yes, so the way we originally intended this was that you were going to hear more about liver injury and i was going to tell you about new ways that we're trying to deal with this problem. so just keep what i'm going to

tell you in your mind and then rewind. i just thought i would tell you what the impasse is all about before i dive into what we're doing with toxicity. our is this. it's all about [indiscernible] new methods and new technology

that will leap frog we hope over some of the road block hurdles in the translational process that enhance the development implementation of intervention diagnostic therapeutics devices, etcetera, that improve human health. and one of the roadblocks is

doing this is either prehoc or post hoc which is part of the -- and if you combine preclinical toxicity in animals and compounds are killed or only toxicities which are seen in animals, sorry in humans, and in the exowmed, it's about a third of the time that drugs fail.

and there's a related problem in the environmental world that is that there are about 80,000 chemicals that we are exposed to because of they're in commerce. they are never intended for human use and actually there's no human data that's required to have these chemicals be used in

human use because there's no intention for human ever to be exposed to these. and there are things like the flame retardant in your clothing or the lacquer in the chair you're sitting on or the styrofoam that's in your chair, all of those substances make it

into the ground water and people are exposed to them. in approximately 80,000 chemicals in commerce, only 2000 of them have any data whatsoever about their biological effects. because you're never intended to eat the chair you're sitting on. so there's no occipital human

dose of a chair. however, as we all know, people get exposed advantageously to all chemicals. so they came to us a number of years ago saying we're just dike to hear because we can do about 200 animal studies a year in an average of three million bucks a

pop. even when we do those, it's not predictive of human toxicity. and the chemical industry is coming up with a thousand new chemicals, a thousand pneumo -- new potential chemicals and we thought isn't there another way

to do that. this is the challenge we got confronted with and you just heard a beautiful example of tylenol figuring out the mechanism of tylenol which is really ultimately what this program do but to understand it in a general way to understand

the general principles are of chemicals that make them toxic broadly. and so to understand this, you have to understand that the conventional way that chemicals are tested and have been tested for decades and decades is that animals are given a certain

exposure to a given chemical either iv or orally. that gives you a certain tissue dose. to be a tad pejorative you close your eyes and wait for something to happen and that results in the toxicity. that tox world it's called

apical point. it's a tumor or something fails and it's obvious and you say that chemical was probably bad, let's go on and test another one and see if the same thing happens. if it doesn't, we'll say aye this is great, we'll go on and

develop that as a drug, put it in a chair. the problem with this is we don't learn anything so every time we go through this, the same thing happens. and we don't understand whatthe general rules are regarding toxicity so we don't get any

better and the drug keeps failing. the idea is pretty simple. shouldn't there be a way to understand what the molecular targets are, the cellular changes, the network changes, the system level changes that happens in response to chemicals

generally that would allow us to predict first of all understand toxicity and then ultimately predict it. so if you think about this from a sort of organism, you're giving each one of these chemicals to an animal and an end point.

you're dissecting the animal or human to the complement cell sites or pathways within the cell sites tweeting each of those pathways within those cell types with all of those thousands of different chemicals, watching the effects they v these are on target

pathways or phenotypes. and then computationally putting the rat back together or the human back together. so like the mother of all system biology experiments. this is a collaboration among four agencies and i mention three of them is us, part of the

eps and fda, these are the goals to identify first of all patterns -- figure out what the targets and pathways are they affect. and in a first instance to prioritize compounds for more extensive tox evaluations because the epa has to do this,

they have to make regular decisions. ultimately what we're interested in is predictive models for predictive response. four agencies came together with a very tight collaboration started five or six years ago across all these various levels

of areas of expertise, everything from very high group of screening which we do at our place through various more sophisticated assays, cell-based assays, tissue base the assays and put them all together. this is a slide to show how many people are involved in this.

the wonderful thing about this is this is four government agencies all working with each other. it's almost unheard of but it actually works extremely well because all of these are complementary. so how does this work?

how it works is that the testing system or the assays are nominated by somebody. we think this is relevant to toxicity for one reason or another. they can optimize within our place the nih genomic center in rockwell as part of n cast and

then screened in this case across 10,000 different chemicals, 10,000 different chemicals tested at 15 different concentrations. you haven't heard the old adage about dose makes the poison and peristalsis and all that stuff. talking about toxicity, it's all

about dose. you can't just test at one concentration. testing at 15 concentrations but because these compounds are going to be used by the public and used for regulatory potentially we don't just do it one time we test 15

concentrations three times across 10,000 chemicals in a different pathway assay every week. we're generating about 600,000 data points every single week and all that data gets put through a public data. we are obsessive compulsive in

the extreme about the optimization, validation, the data processing etcetera and then it goes into an internal, holding areas sort of purgatory here while all the various agencies pour over the data to make sure it's reliable before it's released.

this is just remind me to tell you one of the thing that's really important in chemical testing which believe it or not is not traditionally been done at least at the scale we're talking about is to make sure that the chemical you're testing is actually the chemical that

you think it is. this may sound trivial but those of you as molecular biologists have to make sure you rewednesday the clone and makes sure the airs. the history of the pharmaceutical industry is about ten years ago pharma companies

started looking into their massive compound collectsions of hundreds of thousands of compounds to find out they didn't find more act it was. upwards of 50% of compounds they thought were in the collection actually wcialt there. there was either nothing in the

well, nothing in the tube or it was the wrong compound or it was a break down product. so if we're having a project like this where we're going to try to figure out the fundamental basis of why chemicals are toxic, we can't have this happen.

so every one of these compounds goes through a very stringent process and we do this on a regular basis. so i mentioned this quantitative screening because those of you who aren't into this. everybody else in the world pretty much who does screening

does it in one concentration. this is done and nguyen can tell you how this is done. you can teak 5,000 tile falls with a few drinks your liver liquefies. this is no different, same principal. that's why we set it up ten

years ago we did everything in dose response. and so all of these compounds -- you can come up and see if you quantity this. but that's what everybody loves to see but really the hard part is this part. this is the infomatics big data

initiative. without going into it there's a huge amount of again ocd level validation of these data before the data is shared and then put into a number of public databases. what have we done here. keep in mind i just want to

remind you what the numbers are. in testing of molecules in animals in the environmental world, there's about 300 to 400 chemicals tested every year. and we only have data on good 2000 to 3,000. in the pharma world it's difficult to know how many is

tested but that's never disclosed but no more than 500, maybe a little bit more than that. in this case phase one which ended a number of years ago now, we actually screened this plus this about 2800, 2,800 chemicals across about total about 300

different assays. the whole idea was to get an idea what these compounds are actually doing. and a lot of the proof of principle what we're trying to do here is to make sure that the data that we got were actually reliable and interpretable.

if we did the same experiment twice, different days, different place etcetera we got the same results etcetera, etcetera. once we did that, we scaled this up to phase two including tylenol and aspirin. every drug ever approved for human uses in this collection

plus about 3,000 effective from cost continue and 3,000 chemicals -- things in your lucky charms. those lyme green cereal whatever the heck that stuff is. and then industrial chemicals dioxins. three industrial chemicals, pest

tides, food use chemicals and drugs. at this point just in the last couple years we put about 30 or 40 million data points so the idea is that this will be equivalent of a genome project genomic is what we're doing for chemical genomic.

and so to come out of the midst here now is what are the general principles. what are the reactive for toxical pours if you want to call them, characteristic of molecules -- what was the limitation of phase two. we did about two to 300 assays.

we're working on this now. you probably would be interested in this. i went to our infomatics people and i asked a very simple question. those of you who run labs will know this is just the kind of simple question that people like

me like to ask which i went to them and i said look, if we're going to figure out what the principles are by which compounds lead to toxicity. if you assume that any pathway that is involved in physiology, when you disrupt it might cause pathophysiology -- we first have

to enumerate every pathway operating in mammalian cells and then we have to identify how many assays which are not pathway assays, they are really network assays when you think of them. how many assays, what we would have to do to cover all the

biological pathways. i thought they might come back in two weeks. it took them like two years and they figured this out. i don't have time to show it to you but there's a three dimensional globe called the bioplanet.

every pathway represented on it and you can change the light. it's kind of like smithsonian and whatever pathway you want. but at this point we have limited pathway coverage. we're using reporter gene lines as artificial lines. we're focusing mainly a single

compounds. people don't usually take tylenol they take tylenol and alcohol and a big mac and so that's like probably 5,000 chemicals in a big mac alone. y o u

' re talking about mixture which is what people are talking about. this is a big deal. showing you tylenol by itself is not competent but metabolized it is and that's a frequent situation.

trying metabolism in a 1536 well placed in a 500,000 reactions a week is non-trivial. that's one of the technology development thing we're working on. so in phase three we're focusing on all of those things. but in addition, on like

everybody else, esl, ips cell lines, cell types etcetera etcetera, integration of metabolites. and we're also looking into this. a lot of people have done work on model organisms. it's unclear to us how really

useful this will be. if you look at what we're looking at in, you'll recognize a lot of things nguyen was talking about, ox day different stress, etcetera, etcetera. and this one glutathione depletion right here. we're using a lot now of human

cell lines for assessments in the hepatocyte world because as we're talking about this, there not only are liver hepatocytes one can get from donors. i love that donors because they voluntarily gave you their liver, they're usually dead donors but anyway they're dead

donors. but liver stem cells we are using as well. we actually published a paper a couple years ago of assessment of cytotoxicity using cryopreserve hepatocytes which you prebroadly available. pharma uses them every single

day to look at ma tab liz many. one of the things we looked at is the cytochrome activity of all of these and now you know -- across different donors. you'll notice each of these people is a did dorn and each of these is the cyp activity. and the cyp activity varies

among donors and it seems to be reasonably consistent but it's important it depends who the donor is you're doing and this is the basic of pharmacogenomics. you look at the toxins and say what are for killing these cells across all of these different

donors and you compare the -- you'll notice -- first of all there's a variation and variability across the donor and some compounds show a lot more variability according to the but others in some cases they look like better models compared to the primary donors than

the reason i show you this is not for you to memorize all these numbers but to realize this is not as simple as having the right cell type from the right species but you need the right genetic background and they've got to be cultured to actually get reliable data.

and i'll just show you an example of how we, one of the assays we've done is something called phospholiposis which is a classic toxicity not only in hepatocytes but lots of cell types. to show you we can reproduce this at a microscopic level and

electron microscopic level in response to a number of different environmental chemicals and drugs. and then using these data you could do what you really like to do which is if you have enough compounds you could predict algorithms.

predicting possible lipidosis using small molecules. now the chemists making these compounds, these are generally, they have no way to know if i can make compound a or b, which one should i make. these are the kind of data that will allow them we hope over

time to make compounds that are going to be safer a priori. in 15 concentrations in tip can you tell quite easily actually every week but the problem is how do we actually test, how do we actually get from this to the small number of compounds we can actually test in humans or in

rats. that means you need a lower throughput more still logical setting where you can test a hundred compounds at a time using gene expression and using high content screening assays to do that. and we're also developing a

different approach which i'm going to tell you now which is instead of a massively reductionist approach where we go down to target and pathways and cellular phenotypes and computationally put the rat back together again and that process i sometimes say suffers from the

humpty dumpty problem. that is it is possible that we will not be able to put humpty dumpty back together again. it's emergent properties that will simply not be modellable in this kind of reductionist way. almost certainly likely to be the case.

so another approach to this is this one. that is, instead of reducing the complexity of the system through a single cell level, what you do is you make a little tiny organ. so you make a little tiny liver or a little tiny intestine or a little tiny kidney, a little

tiny brain in washington so we have lots of people that have little tiny brains already and they're easy to find. and you put them on -- system for some bizarre reason they're called chips. this program is called organs on chips which seems to me you go

to a cocktail party and have organs on chips which is kind of bizarre. but anyway they're sometimes called tissue chips. what they are is essentially a little -- in a platform. that's the idea. could you take a multicellular

aggregate which represents each of the organs which are responsible for toxicity, metabolism toxicity in a stable platform and be able to infuse say artificial blood carrying a potential chemical, have it go to an artificial intestine and get absorbed or not, go to an

artificial liver, organoid and get metabolized or not, an artificial kidney organoid and have observe the toxicity or not. could you use that in a way that would test for toxicity much more quickly and accurately we hope than testing animals which

is very long very expensive and quite inept. so that's the idea. this is a approach that's classic end cast program in a number of ways. that is that it's a very technologically out there. we tend to really shoot for the

moon on almost everything we do because the problems in the translational space are enormous that we're looking for logarithmic change and this is changed -- fundamentally change the way toxicity testing is done. secondly it's a novel

collaboration. everything we do is a these are the people that brought you the gps and internet and all that kind of stuff. and so nih and -- put in $70 million here. they are focused on the engineering and nih not

surprisingly is focused on the biology. so here's the idea. it's very clear and a little terrifying for somebody like me who spent a lot of their career before i came whatever i am now when i used to be a scientist. it's a bit frightening to me

that everything we looked at shows you not only the cell culture don't translate into vivo system they don't transfer into 3d systems either. so the question is how much of what we've been doing for all these years is going to turn out to be relevant.

we'll find out. so it will be addressed this micro physiological system programs these organizations. so the organs are really multicellular aggregates of tissues, of cells. the place forms are what are call chips.

and we hope within five or maybe ten years to have this micro protolithic platform you could see here and the idea here is you wouldn't have to have all of them at once. they are built the modular system sort of like lego bricks where you can mix and match any

way you want. there are ten different organizations -- organs which are being looked at here. gastrointestinal means liver as well as intestines. they have to be modular or viable at least for a month and they have to be accessible.

the technologies have to be accessible. if you're interested in this because i'm obviously flying through this very quickly. there was an entire issue of stem cell research in therapy which was devoted to a review on every one of the 19 programs

that's going on developing tissue organnoids as part of this program. this is an example, a lung chip and a blood brain barrier chip. i will tell you when this program started being a skeptic, i thought this was nuts. i got to tell you.

i thought this whole idea was absolutely nuts. what i haven't told you is the end point is supposed to be acceptable to the fda to replace animal test within five years. that i think is nuts. i don't think the fda has done anything in five years.

but they're going to want a lot of proof of principle here. but why was i wrong that this was nuts? it's a really interesting object lesson in the translational phase. i under estimated the convergence of technology, that

this is a convergence of not only ips -- technology but a biosensor technology -- which are all converging at the same so every one of these programs, they're all milestone driven programs, every one of them is the head -- training compounds, things like tylenol and other

compounds both environmental economics and drugs are being sourced and given to these folks to positive negative controls. these are the toxicities we're interested in. you heard about the liver injury, functional defect, met biotic changes, etcetera,

etcetera. here's the idea. how do they actually do something like this. well you can't just take a bunch of cells and throw them in plarks -- plastic because i think you know that doesn't work well.

aggregates that are happy and functional for a month, you georgia other to start out with a scaffold that these cells are interested in growing on. you have to have a cell type. and they have to form gave structure. what's been remarkable to us and

remarkable to me is the developmental biologists is that these cells seem to remember an awful lot of what they are supposed to do. they are cooperative that informing the structures much more than i felt they would. this kind of spacial temple

patterning, sometimes you engineer in by the structure. and actually i think i have a kidney in my pocket. just my id. here's a kidney. this is not a real kidney. i'll pass this around. what you can see, this is a,

this is intended to be blood coming up this way and then there's urine this way and there's a little blood tubule interface there you can take a look at.'s one of the programs that's being developed i guess. this program, and in that case, there's very clear you'll see

there a spacial patterning forcing these endothelial cells -- focusing on bioreactors that will keep these organnoids together for a period of time. depending on what you're looking at these have to be innervated not only blood vessels but nerves.

sometimes you need a host response because a lot of toxic effects are actually generated by inflammation or immunity. and then you have to have a readout. what are you going to use as a and then you need some sort of core computational system.

that's what this program is doing. here's one of the livers. this comes from lance tailor who some of you probably know who has been around this world for a while. this high content screening world for a while.

i don't like going into any details because we don't have time for it bethey have designed these liver sinusoids that have all of the cell types supposed to be there. and these biochemical and metabolic readouts which come from these.

there's a paper here that you can't read but you're going to have these slides if you want to read about this, this will give you the reference from that paper a couple months ago. and these, the concept here is that you have all of these various cell types which are in

relation to each other in roughly the same form as they are in a tact liver and having an extracellular matrix they're living on. and then you have these -- to monitor cell functions death or what have you. and here's just an example of

this and i'll, i think this even has a, oops. anyway, this is a movie just showing that it's the bio sensors turn on and off. the point i want to make sheer is the kind of things you can do these days is to look for, if you go back -- so the question

is, do this, does this biosensor and phenocopies -- which allow you to do this. this one in particular cytochrome release from the mitochondria. this is initial compound testing and obviously this is these results come from the folks who

were funded to do this so this is the classic example where they're showing you the what did work and not showing you what didn't. but to show you there are examples where you can look at compounds or drugs with variation effects and you see

results which are either green negative when they should be negative, red that is that is positive when they should be positive or yellow which is equivocal in a human situation. these that we looked so far is predictive of the responses that you would see.

and that's a good thing. so future plans. this means micro physiological systems. congressman can't say micro physiological systems so that's obviously a sensor chip for but you all can say micro physiological systems or mps.

and so the plans now and what's going on is to take what's done in the liver that kidney that you will see coming around to you and the gi tract that is the small intestine mainly integrating these into one platform with these folks who are, have kidney and gi.

and the interesting thing about this result, about this kind of platform is immediately you can begin to think about incorporating the kind of variability that i showed you at that is that because what you're dealing with is modular just in time make cells from individual

donors or ips cells, they don't have to be dead people they could be live people even those sitting in this room that we would be able to make a chip which is relevant to each individual from the ips cells that are taken from those individuals.

and therefore be able to have, get a sense early on not only of the generic toxicity but the population of the toxicity while you're developing the drug. this is a big program like the other one -- a lot of people in darpa and fda. a lot of people from the fda --

and a but of people from virtually all of the institutes and centers. if you are interested in any of this, there is some stuff on our website about this but really the industry, the most easy thing to do because you're here and you're on the same e-mail

server is just e-mail these folks. so that's it. [applause] >> so while leonard is getting hooked in, i want to put in a little bit of information that i was actually going to put in after dr. seefe got finished and

now this is appropriate because this follows right after what chris was talking about. this is a system that we work with, with primary hepatocytes. one of the big problems with hepa they all are running out of gas. and they don't carry out the

normal function. they do for some but not for all. and so you have to put hepatocyte in quote. but primary hepatocytes don't have that sort of problem. i'll come back to that in a minute.

so for example, this is a non-dividing culture with rat, we've done this with mouse, some with humans. so these cells are put in what's commented a collagen sandwich which permits gradients to form and adhesions to form above and below.

now the interesting thing is that these cells. none of them are dying. you're looking at two stains, the green ones are tight junctions and they delineate the apical domain of this epithelial cell called the die -- diode canaliculus.

the thing with this is that systematically over the course of five to six days, this culture system which has become depolarized when you put it into the medium, polarizes to produce the network of the biocanaliculus which is the smallest branch of the biliary

tree. so each studied how that you can quantify it. and to our surprise, it was very interesting that this is associated with changes in mitochondrial dynamics and function. so here are mitochondria that

are stained with an outer membrane protein palm 20. and you can see on your left the electron micro graph and then up above the staining. all these mitochondria are very small. they're not dead. this is not [indiscernible]

they're just fragmented. they've undergone fisson. the mitochondria fuse and become connected. they are dynamic and moving all the time. fission and fuse. they goes on along day six. i forgot i had this one.

never mind. this is a movie that shows the mitochondria are under going fission and fusion. this is sort of in a net shell what happens. when fission occurs, you increase the number but you also decrease the efficiency of

mitochondria to carry out oxidative phosphorylation and but fusion leads to an increase in the mitochondrial mass. they've got these single mitochondria that are gigantic and all of this is controlled by two sets of protein. one, the fusion proteins mfm1 2

and [indiscernible] and also the fission proteins, drp1. and in this case of our system, if we look at the upper one on your right, the fusion protein activity increases as these mitochondria are fusing. the fission one that controls fission isn't changing.

so there's an activation of the mitochondrial fusion process and in the bottom are a bunch of structural proteins of the mitochondria which don't change. now, the fewest mitochondriaen fuses more. so in a that shows the membrane potential generated across the

mitochondria over day one to six. now the interesting thing is the atp level which is in b is practically all disappears when you first isolate cells and put them in culture. and over six days or so, it restores itself to a normal

level. but on day one and two, atp level is extremely low but nevertheless the cell preferentially uses that atp to repolarize to for this biocanaliculus. polarization is in the central feature of epithelial cells back

in the mesozoic era. that's what causes tissue to be formed that separate the inside from the outside. you have to have specialized structure and function to carry that out. now, at the initial time day one and two when atp is still very

low, where did the atp come from. and in c, you can see that the amount of atp that's sensitive to -- myosin which inhibits oxidative phosphorylation. it's almost entirely occurs in day one and two. so the damage mitochondria does

something to enhance oxidative phosphorylation. and then indeed the oxy glucose sensitive atp is glycolysis. beginning day four the cell switches over from oxidative phosphorylation glycolysis. now what happens if we interfere with this process.

well, you'll see in a, what happens if we add al gore myosin or fcccp which blocks the mitochondrial membrane potential. the atp level plummets. and the same thing happens if we add low concentrations to tylenol and a variety of other

drugs which are widely used and are on the hit list for hepatic toxicity. now this process of decline in atp is associated with the fact in b the mitochondria no longer diffuse. they quickly break down and form these inefficient small

mitochondria and c the cell loses polarity. these thing are directly linked. to the oxy glucose stimulates ampk. ampk is the metabolic sensor that drivers processes to make atp within the cell. when we stimulate with the oxy

glucose, this process doesn't occur. so you are by passed the initial effect. now where does the energy come from with ox dayive phosphorylation in a cell that has very low -- here's one example.

in a those are lipid drop puts. those are lipids that form denatured and so for when a cell is subjected to stress. either the isolation of cells or in the case of some drugs. what happens during the five to six day period is those things disappear and the fatty acids

that are released are transferred into the mitochondria by beta oxidation. and that drives atp levels which are in b and polarization which is in d. really dramatic. and if we put a drug in, it's talk mere that blocks beta

oxidation, these processes don't the other source of energy -- when a cell is stressed and organelles and membranes begin to break down, the proteins that are in that are accumulated in something called auto phagosome. that's the source that is transferred to this damaged

might condraw that is preferentially used to provide the building blocks to make mitochondrial proteins and now one of the things to think about, and leonard will [indiscernible] is if you list, we talked about tylenol which is an example pretty much of a

dose-related phenomena. and the problem you're going to hear about and the real challenge is that most of the drug that cause acute liver damage are not dose-related. they are called idiosyncratic. we don't know what the cause is but it's not very common.

and a big argument in the field i think is whether if i take the drug and my liver falls apart, is that because i have something missing in my genome or mutated or maybe it's something else i ate or whatever it happens to be. is that one in a million.

or, is it that all the drugs that cause hepatic toxicity basically affect mitochondria for example in a way that i showed you the stress of cell isolation i just mentioned in passing tylenol. and the question is not one in a million.

it's that everybody to some extent is affected by these effects, that's how drugs act. and i developed liver damage because i've got something else on top of it. that's the tip of the iceberg. now, this fits very directly to what chris was talking about.

why do you separate them. the last thing i would prepare because now you're going to hear the really story in a minute is to think about of the ten leading costs of drug induce the liver injury, the kind that doesn't happen to everybody when just by taking a drug.

nine of them are microbial derived antibiotic. most of our antibiotics came from [indiscernible] streptomycin, tetracycline or they came from bacteria. now bacteria of the proposed origin of mitochondria, maybe. they are the current, i mean

that's the thought of the review article. and it seems that if mitochondria are major targets of drug induced liver injury or the drugs -- it's interesting whether these three conceptual things are related in a fundamental way that we don't

understand maybe somebody particularly a bright young graduates doing the post doc will come up with a clue that will unlock this. okay, leonard. we're all set, i think. >> let me begin first of all by apologizing.

i was trying to download my slides this morning actually on updating the presentation and for some reason couldn't get it to come down. i had to run back and had to go through the traffic. anyway, you heard now two outstanding basic researches

talking about about problems regarding the liver and you'll have the mundane clinical aspect. this is the problem that we as clinicians faced when we find patient who have abnormal enenzymes and how do we in fact identify uninduce the liver

injury. one is a convention drug and very importantly because this is an area of great interest now and i am particularly interested about this, and that is the injury from herbables and dietary supplements. so win asked me to stop by and

show you a case. and see how we go through the whole process of assessing the possibility of drug induced so this is a caucasian male who was treated for [indiscernible] that is -- 2012. he was hospitalized from [indiscernible] his past history

included diabetes and [indiscernible] [indiscernible] and he had stretch marks because of the fluoride. he was slightly are tender in the right upper quadrant but that was it. the other drugs, you can see a list [indiscernible] and some of

the statins that indeed was taking for years and continued to take [indiscernible]. he had had a [indiscernible] earlier before that should really be 1, 7, 12. down to normal. when he came in you can see he had distinct elevation of his

alcs [indiscernible] liver elevated [indiscernible] moderately elevated and his bilirubin was up. this is very important. over the course of the next couple or several weeks his amino transferase levels dropped but they all came down and by

9/14 he was back to normal. bilirubin was now normal. now the question was, was this drug induced liver input. i'm not going to tell you what the answer is. i'll do that at the very end but let me tell you how we go about first of all some background.

i think as you heard idiosyncratic -- drug injury which is not direct toxicity. this is toxicity one in a thousand, one in 10,000, one in 100,000. this is among the most common regulatory reasons for [indiscernible] drug as well as

removing drugs from use. at one point, it was thought that liver injury was the leading cause, i think it is now cardiac but this is very close behind. it may prejudice as you huh-uh pathic cellular injury simulating hepatitis

[indiscernible] the actual fact is almost no diagnostic form of liver disease acute or chronic that is not limited by drug induce the liver injury. what about the frequency in the well, first of all it is rare. we talk about idiosyncratic drug injury is a rare phenomenon

[indiscernible] several reasons often people who are taking care of patients never considered the possibility that drug induced liver injury is the cause because they don't answer the right questions. they might do [indiscernible] but they mismanage it and i'll

show you. and then more than likely a true place is not [indiscernible] fda. then we're in error what the real frequency is. and review is 10% of drug induced liver injury from conventional drug, only about 1%

[indiscernible] are actually the food and drug administration. i don't want you to memorize this but this is just to remind you that there have been a lot of regulatory actions from the fda because of bilirubin over the years. some drugs were never approved,

some were approved elsewhere and with quran. there are a number of drugs for which there have been warnings. this is sort of a list that seems to grow as time goes by although it seems to be growing less frequently now because i think the fda is doing a very

good job of assessing the possibility of bi. b i li. it varies depending on the drug and its presentation. if the presentation is -- that is identified because the an

size are elevated but the belly ruben is normal, the prognosis is good particularly if the drug is withdrawn. we do face the problem of the issue of adaptation because there are many drugs in which you treat the patient, they develop abnormalities and

despite the fact you go on treating with the drug, is value of value is normal. this is ina and true for the statin drugs. so this is a real issue when we're trying to consider the possibility of what should we do, should we start the drug or

on the other hand, if it's hepato cellular that is associated with jaundice, this is a serious problem. and the mortality is about 10%. and you see there in parenthesis the words high lows. i'll show you a picture of the highs in a moment.

he was the leading expert in hepato toxicity in this country. he wrote brilliant textbooks. i will show you in a moment what [indiscernible] it's become very amusing to me because all over the world people go for highs lows and they have no interest on who high is.

but he is one of the founders of the whole area of hepato if you develop on the other hand you [indiscernible] object normal developing inself law they [indiscernible] tylenol is about 7% mortality. i think many of you may now of a study that's being done and has

been in progress now for some years supported by nid decay [indiscernible] form of hepatitis in the country. and it is of note that something like 65 to 75% of a plausible [indiscernible] is on drugs [indiscernible] hepatitis and other forms of bilirubin.

if you develop [indiscernible] the prognosis is not great although it's interesting it is less with acetaminophen. what he found when he reviewed thousands of cases all over the world particularly the fda on drug induced liver injury if the patient had a bilirubin elevated

2.5 mill grams per deciliter and the alt was elevated [indiscernible] there was about actually a 10 to 50% mortality. temple at the fda was so impressed with this he calls this highs lows and now this has become known throughout the world as hy's law.

and the pharmaceutical companies [indiscernible] are terrified of hj's law. in develop of a drug, if you have one there's a great deal of concern if you have two [indiscernible]. how do we diagnose hepato unfortunately there is no pie oh

marker for induced liver injury. we have to rely -- [indiscernible] and then excluding all other causes for these abnormalities. this is [indiscernible] how do you approach it and there are two ways of doing it. one is you recall clinical

judgment or expert opinion and the second is [indiscernible] instruments. i'll tell you about this in a little more detail. as it happens, the fda is having a big meeting the next few days probably tomorrow and the next day.

there's going to be a big debate about what approach is the best or assessing the potential causality of drug induced liver what are the initial [indiscernible] when you consider the possibility that there is dili. the [indiscernible] seems to be

concerned about it [indiscernible] jaundice elevation, it was what the concern was -- form of liver and this may or may thought be associated with symptoms and it has to, it is usually developed within six to nine months when you stop the drug.

then of course you have to as all clinicians should be doing is a systematic assembly of all historical clinical by i don't chemical serological and radiological who perform a liver biopsy in order to complete the assessment. as you know these are

non-specific measured and indicative of liver injury of any kind. as i mentioned they suggest about injury, the outsource of [indiscernible] there is now an algorithm or an equation i'll show you in a moment which helps you to distinguish between

[indiscernible] jaundice. what they're struggling with now is what is the level of al [indiscernible] of concern when somebody is taking the drug. often there are mild elevations of alt. at what level do you begin to worry?

some say it has to be three times the normal and others say it should be five the upper limit of levels and [indiscernible] the problem we're facing now is we don't really if of know what the upper limit is [indiscernible]. as you know the values for upper

limit of normal are established bid a large local population believed to be quote healthy and then selecting as a reference number the new value [indiscernible] standard deviation. that's how laboratories set their upper limit of normal.

the population may not. there's a big argument. often the report of upper limit of normals [indiscernible] but there are now wide variations being found in the upper limits of normal for alc among all these laboratories depending upon the population being

screened. note in 2002 in italy reported that the upper limit of normal for the [indiscernible] was 30 units for men and 19 for women. so there's now a distinct difference by gender. and since then there will be numerous papers in which people

from this country, from asia, from europe have looked at this issue and most of them agree [indiscernible] first began. and it appears that it differs not only by gender but by age, by race, by bmi, elevation of the bmi. that mans they have happy liver

disease and that accounts for the abnormalities cholesterol, triglyceride so on and so forth. [indiscernible] to use the manufactured upper limits of normal when you're assessing the possibility of [indiscernible] to determine whether abnormality is elevated by two or three or

four. so what are the next steps. you identified a signal. now you have to define the liver injury and you have to exclude all kinds of conditions that might be responsive to the liver injury and then you have to characterize the signature of

the drug-induced liver injury. [indiscernible] came up with an equation which they called an r value in which you point the alt divided by the upper limit of normal all of which is divided by the peak level divided by its upper limit of normal and depending what comes out

distinguishing hepatocellular this is used by the pharmaceutical companies. and what are the reasons for making this distinction is that other conditions that need to be excluded were will vary depreponderancing on the hepatocellular injury.

so we are working up and case ask we have to exclude everything you see here, hepatitis a, b, c and e and then you just focus on e. we never had any thoughts or thinking about e in the past, if you take the drug induced liver injury is now a number of papers

showing that papers with drug induced liver injury and people went back for hepatitis e hepatitis e now is much more common as you probably are already aware. you have to screen and exclude the various forms of hepatitis [indiscernible] stimulate

induced liver injury [indiscernible] that you see there where if the patient has [indiscernible] or pancreatitis. once you put all that together you now have to define the signature of this liver disease. and what are the components of the signature.

the reason for doing this was because then you can fit in the capillaries of known signatures that have made the other injuries. you look at the interval between the start of the medication and on start of the liver disease. you determine that it's

parastatic [indiscernible] if you're talked about the drug [indiscernible] doesn't go away, it may be that the patient always had underlying proclivities you didn't know about so it wasn't drug induced liver injury and should go away. sometimes it's [indiscernible]

before it goes away. then there's this issue of rechallenge. obviously that's the best way of identifying whether a particular drug is the responsible for the that is something that again should not be good. rechallenge a patient for the

injuries and often you need severe fatal disease. and sometimes what happens is inadvertent reexposure to a drug because people didn't realize that the drug had been used to full and they note then in the history that they've been used before and now you see a case in

which this particular drug seems to be responsible. let me just mention the fact that as i said the biomarker at the moment of the enzyme but there is a lot of work when you try to identify other biomarkers [indiscernible] rnas. and there's an interesting

[indiscernible] which is associated dili from a number of drugs including [indiscernible] which is one that i showed you. a lot of g was studies are being done [indiscernible] but there's still a lot of work being done. this is the big issue at the moment and this is going to be

discussed in detail at the fda meeting the next two days looking for biomarkers of liver now what about causality assessment using the standardized instrument? in 1993, the council for international organizations of medical scientists

[indiscernible] pharmaceutical company have international experts living in paris to work on the fund [indiscernible]. as you can see there are people from various country including people from from the u.s. [indiscernible] and they came up with a weighted causality assess

that they called [indiscernible] causality assessment method. now, what they found was that they awarded points for various items. time to answer the risk factors [indiscernible] search for non-drug causes [indiscernible] and you could

come up [indiscernible] from zero greater than eight and you can see how then it was classified. the problem is that it's been used by experts in the field and by the pharmaceutical companies. nobody that i know of [indiscernible] we actually when

i was here at the nih involved with the drug induced studies, we actually, we are [indiscernible] assessing causality using both expert opinion and [indiscernible] and then we compared the two. we found that the [indiscernible] was wanting.

nevertheless the europeans love the american don't particularly like it. but tomorrow there are going to be two europeans that the best way to assist causality for [indiscernible] and people in this country are going to say it probably is not the case.

so this is the study that i think you should all know about. this is a study that's been ongoing for ten years, studies supported by niddk. there are a number of vorgtsz around who are clicking all cases of drug induced liver inury both from conventional

drugs and from herbals and the important part of this is they wanted to address the problem of the fact that for clinical judgment is subjected, you one of the aims was to develop standardized definitions, writing systems and clinical instruments who

identify causality that case of the prospective dili. there was a grading system created for the likelihood which ranges from unlikely to definite. they are using legal terminology here but we have much more precise definitions specifying

the difference within these capillaries. and at the same time there was a grading system for liver disease severity. was it mild, moderate, was it moderate severe, was it see veer or was it fatal. the there are various criteria

to think of the definition. just a brief comment about the number of cases that have been collected thus far in the nihdili studies. we now have a total, there is now a total of 1300, 1352. i think you should note that 83.3% were conventional

medication. i will show you a list in a moment. that's almost 17% nowing believe that herbs are safe and herbs are good and so on and so forth. the data from a study we have here and i'll show you a little

bit more than that is unequivocal. these products are no more safe than conventional drugs. so here, and this actually -- i asked him for the top 16 conventional drugs that have been collected in the dili study for the liver injury.

as you can see the [indiscernible] turned out to be antibiotics [indiscernible] is the most common drug causing drug induced liver. it's the number one, one of the most common forms in these databases it's the i want to bring this to your

attention if you don't know about this and that is the brilliant es that [indiscernible] has done in developing a literature review which is in the national library of medicine for the liver talk. what they have done is they have pulled in hundreds and hundreds

of drugs that they have summarized giving every reference related to that particular drug. if anybody is interested in liver toxicity and you wonder where the data are to support it, this is the place to go to. the liver talk nih.gov and you

will filed everything that you want to know about this and all drugs that are available at the present time. now let me just quickly go into herbables and dietary supplements. herbable products have been utilized for centuries and

trying the authorities of the predominant mode of treatment of various illnesses. however there's a imrog interest in the use of western countries taken [indiscernible] c as we have learned and i'll show you some day, they're also used to lose or gain weight, body

weight. the other problem is many of these are used to treat [indiscernible] and that's where the problem is because there can be replaced conventional treatment then you run into big trouble and you're concerned about that.

what is the prevalence of hds [indiscernible] in the telephone surveys and reviews in the um of databases nhis, fda and so on that sales to indicate that somewhere between 30 and 70% of the population use alternative medicine. although many interviewed

individuals [indiscernible] disagree with it. most surveys have founded the use of [indiscernible] taken either with conventional medicine which is then complementary or on their own alternative, hence the term complement alternative medicine

and that this is increasing year by year. but analysis from the amount of money being spent in 1997, estimated considerably out of pocket expenses for herbable products was something like $27 billion. since then at the nih in 2007

estimated that these studies [indiscernible] equivalent to that one third of total out of pocket spending on prescription and at the cost of practitioners [indiscernible] out of pocket spending on physician visits. as you can see, it is a growing

issue the use of these products. and here there's a journal in the christian business journal that has been tracking the amount of money spent on drugs. as you can see in the middle and it gets higher and higher and higher. and the increase or decrease

year by year and as you can see, with the exception of two years the frequency of the use of supplements have been increasing every single year. now what kind of products are there? in the broader sense there are a few types of products.

one are the so-called crude herbables which are single products. these are the products that have been used traditionally for years by indigenous people [indiscernible] for various diseases and problems. all of this as you know comes

from the [indiscernible] it may come from a stem or roof or the flower or from the trees. the chemical composition is varied by geographical location. the climate raises and [indiscernible] often we just don't know what the constituents are and have never been

established. the whole problematic from my perspective are the commercial these are made up by companies that make these products. they contain numerous ingredients not only are identifiable. the contents and concentration

of chemical constituents may vary from batch to batch. the same product and we have done that. we've gone to purchase [indiscernible] and then tried to identify [indiscernible] and sometimes they don't even complain.

they also may be contaminated and [indiscernible] conventional i'll show you this in a second. the other thing of course is the effect of doses not established or sold over the internet. so contamination in the alteration. where do these, how do they deal

with these clones. the clones are cultivated and sprayed with pesticides or collected from the wild often are harvested dried [ [indiscernible] they may be con taminated by microbials or heavy metals such as lead and arsenic.

there are a number of papers showing this has been recorded. now what adulterants have been put in [indiscernible]. here's a list, there are probably more but i can show you, i've outlined [indiscernible] cortico steroid. when you think you're taking --

you may well be taking [indiscernible] cortico steroid doing the work they think it should be doing. so this is a real problem. and then of course the problem is its regulation. so in 19940, the fda did not permit food and dietary

supplement for the manufacture of products for disease prevention [indiscernible] but in 1994, congress wrote the dietary supplement and education act known as the this gave the responsibility of insurance safety of supplements any ingredients in the hands of

the manufacturer would not need to manufacture their products from the fda before producing or selling. they need to follow gmp. the fda can intervene only if the products are already on the market. manufacturing rights specifies

notifying the fda. this is believed that 1% adverse events are actually reported. there are some new fda guidelines regarding notification of new dietary [indiscernible] other countries such as the eu have somewhat different regulations.

but this remains a real concern for those of us who are interested in this that what people are taking are simply not regulated. what about their toxicity. well, we have a problem because you don't know often what is in these products that is

responsible. particularly if they take the multiple ingredient product. first of all when you go on-line to see what's in it, they sometimes cannot find what's in it and sometimes they don't tell you what's in it. they're not sure if there's

contamination [indiscernible] a bit of a problem especially in causality for herbables. and they're busy working on trying to develop a causality instrument to be more specific. somebody who has a great interest in capability of trying to identify the toxic components

in some of these problems of we've been working with these what are the chemical preparations being associated with hepatic toxicity. this is a list. some of you may know this. back cohosh used for menopause [indiscernible] the most

frequently [indiscernible] throughout the world. the problem with this is it's high concentration. and the data are unequivocal. we've got good data and [indiscernible] a number of the people in the study are evaluating and trying to

understand more about it but we have some very interesting data on that. but this is a list of the single here are a list of the products lipo kinetics, hydroxy cut and herballife. i'm showing you this because this is not the only list but

this is what we have been facing. look at the names of these products that people have been taking and that we had in fact identified what looks like drug-induced liver injury. i mean how about this has been at times very

complicated trying to understand what precisely is the cost of this and there's [indiscernible] look at this in more detail. what i'm showing you here, i'm almost done with this is we tried to categorize these what are they used for. and it turns out that the

product that is used for weight loss and body building were the leading products that are associated with dili. you can see there are a lot of others for various other conditions that are less likely, seem to be less frequent. now i can tell you by the way

that what we have been seeing over the ten years the proportion of cases have been equally also in our study year we don't know whether there's an increasing frequency because they don't have the, we don't have the baseline of how much of this is being used.

now there's a study that's ongoing now focusing on one state delaware. we gather all of this information to see whether we can determine the rate. so this here are the top 15 hbs products causing dili. they're trying to get this

getting the analysis done particularly the body building. the question is do they contain anabolic steroids that are being put in or is there a product that actually has antibiotic steroids-like effect. this is [indiscernible] i bring your attention to the

concern about herb-drug interactions. you may have trouble with [indiscernible] cyclosporin [indiscernible] when people are using some of these products one has to stake all of this into consideration. let me conclude with the hds.

herbable and dietary supplements is usually extensive and often exceeding the use of conventional medications. the products utilize largely for weight loss and muscle building but also improve well being as well as symptoms of chronic diseases including those of

liver origin. there's a general belief that because they are pure some having been used for centuries they must be effective and safe. their safety however is compromised by the fact that in my view they are not fda approved causing uncertainty

about their reported and you reported contents some of which may cause liver disease. the effectiveness of most of these products has not been scientifically proven resting largely on personal testimonials and advertising. the last slide this is the case

that are looking for all other causes of liver injury. as you can see he was seen for hiptitis a, b and d, cmv all negative. the anti-bodies were negative [indiscernible] but the patients were covered. and we did not use ultrasound

because we used it as part of the evaluation [indiscernible] ultrasounds or cp scans or mris in order to be sure there's no obvious evidence of obstruction. this was thought to be a classical case of augmenting amoxicillin.

i think i was trying to tell you the difficulties we have with trying to identify and distinguish the drug induced dili with other forms of liver injury in the absence of at this moment the specific and definitive biomarker. i think there's a lot more about

this and [indiscernible] very happy to be doing this study now. hepatitis c used to be the big problem but now [indiscernible] form an infectious disease [indiscernible] drug induced liver year seems to be let me stop at that point.

thank you. >> were you able to predict that your assays [indiscernible] tylenol [indiscernible] cytotoxic. >> i would imagine without knowing it [indiscernible]. >> the question was whether we can protect tylenol.

with tylenol you're sort of cheating because everybody knows what the mechanism is so we're able to test not only the parent compound but the tabloids that we know are toxic. but the problems is most of these we don't know what they are and so a positive may or may

not mean something and a negative may or may not need something depending on the >> let me ask a quick question here. driving force is to determine these things. it's not only to do with large numbers quickly efficiently and

specifically it's also economic. the fact that working with primary cell lines labor intensive. to work with animals gets into all sorts of complicated business. now the question comes how good, if those issues were not present

with regard to animals, what would you actually do? do you think that the value, is there, is this really is it just a matter of money and society that mice aren't used or rats or rabbits or monkeys. biologically is it relevant? >> well my understanding is that

when we come to idiosyncratic liver disease animal studies are not helpful. this is the interactive between the drug and the patient. if a million people take a drug or a thousand people take the drug and any one person gets injury in 999 taking the same

drug at the same dose doesn't, there has to be some specific interaction between the patient and the product. now i understand the possibility that everybody does get something but we just don't see it because we end up looking. this is not helpful in, i mean i

think it's the direct hepatic -- that animal studies [indiscernible] he doesn't see any value other than in making sure [indiscernible] could not be used but in point of view idiosyncratic disease the animal studies then seem to be terribly helpful.

i would just, when people have done this, there's some nice published papers about this where in retrospect if you look at drugs which are either safe or you been safe in human and look at what the results are in animals, it's varies around 50% whether the animal model is

predicted or not. it depends on the particular kind of toxicity. bunny marrow -- bone marrow is less and others are surprisingly hard to see in animals. it's about 50/50. >> i'm [indiscernible] regarding so reading some of the papers

drugs especially in the united states are the most important cause of active liver failure compared to other kinds of i was reading especially is one of the major cause [indiscernible] here in the states at least related to so i wonder if the results of

the dose is very important [indiscernible] and all the other drugs but within the paper especially william lee it's always taken for specific purposes. both to say and in england why in all the rest of the world [indiscernible] one of the most

important codes [indiscernible] so i don't know if this accounts for the difference. >> certainly in england we know this is the most common way to commit suicide. they tried to deal with that as you know in various ways by removing it from

over-the-counter and putting it behind the counter and putting it in bubble wrap because you had to get out 50 tablets to kill yourself. somebody want to commit suicide may not spend all the time doing that's what they were doing. in this country it's my

understanding it's less likely used for suicidal purposes. it does. it's used. alcohol for example, we're one of the first people to identify when i was here at the va with hy and we had patients that were taking what looked like

conventional doses of tylenol who had been taking it because they had teeth surgery or root canal or something and they were heavy alcoholics. we now have a lot of the information about the fact that the role of the alcohol inducing the toxic [indiscernible] so i

think it's some of the cases that are seen here are not suicidal. and there was a difference. >> the other thing that is seen in this country too is that with sometimes the most tragic is that a lot of the cases, they're acting out suicides but they're

not intended suicides. they are accidental suicides. as we all know a lot of this is done for, is a signal for help. and so they figure well tylenol is supposed to be relatively sake. instead of 20 pills of valium i'll take 20 pills of tylenol.

it turns out as nguyen doesn't mention this but dying of, if you're going to, if you want to commit suicide this is not the way to do it folks. it is a really bad way to die. and so frequently it's not an intentional, intentional suicide attempt, it's not meant to be.

>> [indiscernible]. >> one of the problems is that tylenol is in many drugs. it's in many drugs and people don't even realize this. for example nyquil it contains tylenol and it's got alcohol. so people may not realize how much they're taking.

so part of it is inadvertent overdose. >> so okay, thank you very very much, both of you. it's very exciting.