Interesting information regarding the development of ALZA and Mylan’s patches.

MARIO A. GONZALEZ having first been duly sworn, was examined and testified as follows:

COURTROOM deputy: Please take the stand, Dr.

THE COURT: Good afternoon, Dr.

THE WITNESS: Good afternoon, Your Honor.

MS. DAVIS: should I proceed, Your Honor?

THE COURT: Yes

DIRECT EXAMINATION BY MS. DAVIS:

Q. Good afternoon, Dr. Gonzalez.

A. Good afternoon.

Q. Would you please tell the court where you reside?

A. I live in Pembroke Pines, Florida.

Q. what is your present employment, Dr. Gonzalez?

A. Well as of a few weeks ago, I was President of Global Max Americas. ICON clinical is a large CRO. They announced that they are purchasing us, that they have purchased us. And I will – and that was my old job.

I will continue consulting as a pharmacoki neti ci st.

Q. Could you just explain what a CRO – you used that term – I'm not sure we are all familiar with that acronym.

A. CRO is abbreviation for a Contract Research organization. Global Max Americas was part of the Global Max Group which was a group of about 80 scientists doing consulting work and running clinical studies and helping companies develop products. And ICON is about 3,000, scientists so they are bigger than us. They made an offer we couldn't refuse. And we are being bought.

Q. Would you describe briefly to us your areas of expertise or the areas in which you perform consulting?

A. I'm a pharmacokineticist. I consult primarily in pharmacokinetics and other issues related to clinical pharmacology. Meaning pharmacology applied to humans and the effect of drugs in humans.

My work in pharmacokinetics has primarily focused on control release delivery systems, both oral and transdermal. So most of the consulting I do usually involves the drug that is not an immediate release product.

Q. So an immediate release would be something that goes in once and immediately gets dispersed through the body, as opposed to controlled release where the drug comes out over time?

A. That's right. A controlled release product is, as the name implies, the drug is being physical, chemically controlled by the dosage form. A transdermal system is a perfect example of that.

Q. And you mentioned pharmacokinetics. Could you define for us or describe for us what aspects of the study of drugs and how they operate is covered under the definition of pharmacokinetics?

A. Yeah. There is the little – a little acronym 1 we always talk about pharmacokinetics, ADMS. And it stands for absorption, distribution, metabolism and excretion. So pharmacokinetics is really the study of how drugs move around in the body, sort of a cocktail party description of the term.

And what we do is we basically use mathematics to describe the absorption, or how the drug distributes, or how it gets eliminated. Elimination is a combination of both metabolisms and excretion. Excretion meaning that the drug comes out of the body without being metabolized. For example, Fentanyl – 10 percent of a Fentanyl dose comes out unchanged in the urine. That's excretion.

Q. And what happens to the other 90 percent?

A. The other 90 percent gets metabolized. So the body chews up the drug into an active form and excretes it that way.

Q. Dr. Gonzalez, could you describe briefly your educational background?

A. Well I have a Bachelor's in Pharmacy from the University of Texas in Austin. And then I pursued a Master's in Physical Pharmacy at the same institution. I got that in '66.

I completed my Master's in '66. And then I worked as a pharmacist for awhile. And then when I went on to graduate school at did my PhD at the university of California San Francisco with the major Professor Sid Riegelman and Tom Tocus, well known physicists.

My diploma, signed by Ronald Reagan, is January of 1975.

Q. After receiving your PhD, what employment did you pursue?

A. I became an Assistant Professor at Colorado. First off, University of Colorado. And then I jumped up to Purdue University, which was a much better school of pharmacy. I worked at Purdue for about four years, was involved in – in clinical research. That involved a drug for asthma called Theophylline.

Through contract research work that I was doing for a pharmaceutical company, I ended up getting a job with that pharmaceutical company and that was Key Pharmaceuticals. So in 1980 I left Purdue to go to Key Pharmaceuticals.

Q. So through some work you were doing for Key Pharmaceuticals, while you were employed at Purdue, you became acquainted with Key and then left Purdue to go to Key full time in 1980?

A. That's right.

Q. And could you describe for us what the business was, o'r what the focus was of Key Pharmaceuticals when you went there in 1980?

A. Key Pharmaceuticals was a fairly small company, but fairly research intensive. And the focus was to look at old molecules and improve their delivery by controlled release formulation work. And so at the time the product that I had been working on was the sustained relief Theophylline. Extended relief or controlled relief, all of those terms are interchangeable. Long acting. Okay. And they were working on a nitroglycerine patch at the time. And that became my first transdermal project in 1980.

Q. You mentioned, I think you used the term old molecules, what do you mean by that?

A. I meant drugs that had already found application and utility but not as sustained release product. For example, Theophylline had been a widely used drug for asthma, as an immediate release product before Theo-Dur came out. Theo-Dur was the first sustained formulation of Theophyiline. Nitroglycerine was available as sublingual tablets and as a nitroglycerine ointment. Actually 2 brands of nitroglycerine ointment. The ointments were approved in the 50's. The sublingual tablets much earlier than that. Actually I have no idea when it was approved, but it has to go way back.

And' then from the ointments it was – the transdermal concept rather was started – at Key was started by Alec Keith.

Q. And Dr. Keith was at Key Pharmaceuticals when you arrived there in 1980?

A. Yes. Dr. Keith was vice President of all R&D.

Q. When did you leave Key Pharmaceuticals?

A. In June of 1986 Key Pharm. was acquired, purchased, bought out, by Schering Plough.

Q. Did you remain with Schering Plough after the buyout?

A. Yes, I did.

Q. And what type of work were you doing at Schering Plough?

A. Basically the same thing. I,was a pharmacokineticist doing the – actually I did exactly the same work that I was doing for Key. I worked primarily on control release formulations, both oral and transdermal. I had a title of Director of Biopharmaceutics and Pharmacokinetics at Schering Plough Research.

Q. And how long were you employed at schering Plough approximately?

A. For about five more years.

Q. When you left Schering Plough, what did you do next?

A. I started a consulting company under the name P-Kinetics, Inc. Obviously P-Kinetics is an abbreviation for pharmacokinetics, and as the name implies I was consulting in pharmacokinetics.

Q. Can you describe what your consulting work consisted of when you were working with P-Kinetics?

A. Well my first client was Schering Plough. And one of the first things I did for them was help finish up some nitroglycerine studies we were doing. We were also working at – at modified formulations of nitroglycerine. And so I finished up that work for them and continued consulting on other things.

And then, of course, I started consulting for other companies. Among them Cygnus Therapeutic Systems and Novan Pharmaceuticals, both of which;are transdermal companies. And actually both of them are spinoffs of Key Pharmaceuti cals.

Q. Did your consulting work continue to focus on controlled release systems?

A. Pretty much. Yeah. I think Fluvoxamine is an anti-depressant. I think that's the only immediate release product that I can think of that I worked on.

Q. So with that exception, it was primarily focused on controlled release, including transdermals?

A. Correct. Unique delivery systems is my shtick, you kndw. A little niche, I guess, is the better word.

Q. Over the course of your career, have you published scientific articles in the area of pharmacoki neti cs?

A. Yes, ma'am.

Q. Do you know approximately how many articles you've published in that area?

A. The cv that I turned over to you guys about a year ago, I had 51 articles. I have 54 now.

Q. Let me ask you to turn in your binder actually to Defendant's Exhibit 1196. I'll just ask you if that's a copy of your cv?

A. Yes. That's my cv. Yes. It is.

Q. And does that reflect some of the articles that you've written in this area?

A. Yes. Yes, it does. It doesn't have the last 3 articles, but as I said, I know that I hadn't updated it.

Q. Could you estimate for me approximately how many of your published articles relate to the pharmacokinetics specifically of transdermal drug delivery?

A. At least 10 articles.

Q. Have you also presented symposia and lectures concerning the 'pharmacokinetics of transdermal drug delivery?

A. Quite a few. Invited lectures as well as submitted posters and presentations at different meetings. Yes.

Q. Have you consulted or worked with any government agencies concerning the pharmacokinetics of transdermal delivery?

A. I have been a paid consultant at the FDA. And then the Canadian equivalent of the FDA. I have been to the HPB in Canada to talk specifically about transdermal. I've been also to Therapeutic Good Administration which is the Australian equivalent of the FDA, and I lectured on transdermal products and the evaluating transdermal products. And those are English speaking countries.

Let's see, quite a few of the Spanish speaking countries also. I've lectured in Brazilia on transdermals, in Argentina, Mexico, of course, I'm skipping some countries here. I'm hitting a blank. I think the adrenaline is kicking in now. Panama, Uruguay, and – quite a few Latin American countries.

Q. Can you approximate for me about how many studies on transdermal products you worked on during your time at Key Pharmaceuticals and then at Schering Plough?

A. Yeah. I would say about 50 or 60 such studies. Right'.

Q. And have you also worked on studies of the pharmacokinetics of transdermal products since you left Schering Plough?

A. Yes. Quite a few.

Q. Approximately how many?

A. I would say about 20, 30.

MS. DAVIS: we would offer Dr. Gonzalez to be qualified as an expert in the field of pharmacokinetics, particularly pharmacokinetics from transdermal delivery systems.

THE COURT: Any objection?

MS. SANCHEZ: No objection, Your Honor.

THE COURT: So qualified.

BY MS. DAVIS:

Q. Dr. Gonzalez, you have been retained as an expert witness in the case on behalf of the defendant Mylan?

A. Yes, ma'am.

Q. Have you ever been retained as a consultant or expert before by Mylan?

A. NO.

Q. Have you ever worked for Mylan?

A. No.

Q. Has' any of your research work been funded by Mylan?

A. No.

Q. Do you have any relationship with Mylan apart from your work in this case?

A. I know Sharad Govil quite well, and I know John O'Donnell also. Not as well as Sharad, but I know Sharad extremely well.

I helped interview him when he interviewed at Key Pharmaceuticals, so we go back quite a ways.

Q. Are you receiving compensation from Mylan for the time that you are spending as an expert in this case?

A. Yes, ma'am.

Q. is your compensation related in any way to the outcome of the litigation?

A. No, of course not.

Q. Now before we address some of the specifics in this case, I would like to make sure that we all understand some of the terms that you might be using during the course of your testimony today. Could you explain, and I think you started to do this explanation, but could you explain pharmacokinetic terms, what is meant by the term absorption?

A. Okay. Well absorption is the input from a dosage form into the general circulation of the body. So absorption mean's the molecule leaves the dosage form and enters the body, enters the general circulation, and it goes to site of action.

Q. And by the general circulation there you're referring to actually being in the blood stream?

A. Correct. The systemic circulation is another way it's commonly referred to.

Q. And what is meant by the term elimination in the pharmacokinetic use of that term?

A. Okay. Elimination, I think I already said this, but elimination is actually a combination of excretion and metabolism, and you can quantify it. The easiest way is simply to look at the – at a plasma profile and follow the fall in the blood levels. And on a semi logarithmic plot, you calculate the slope of that fall, and from there, from that slope, we come up with a rate constant of elimination.

And that – the inverse of that rate constant and divide it into the natural log of two, that's where the half life terms comes off. Okay. And the simplest way if you look at a blood level profile, look at the blood level profile and you see how long it takes for the blood samples to drop to one half. That's a half life. Without getting into the slope. Sorry, about that.

THE COURT: That assumes the rate remains constant from the beginning until the end, is that – is that a scientific truth?

THE WITNESS: No. That's not correct at all. Because when you're talking about rate, you're actually talking about the rate constant times the concentration or the amount of drug in the body. Okay. That's what we always talk about.

Elimination being a first order process because the rate of elimination, the rate that you're talking about, is actually proportional to the amount of drug in the body. So the rate constant is the proportionality, the rate constant elimination is the proportionality constant. So a rate is proportional to an amount. You throw in a proportionality constant, now you have an equation. Rate equals constant times amount, right.

THE COURT: Okay. Well if you've got a slope, if you start tracking the elimination rate, is that slope relatively consistent through the half life so that basically you can figure out what the half life is from the top to the bottom?

THE witness: Correct.

THE COURT: Half life is right in the middle. And is that slope consistent?

THE WITNESS: okay. That straight line is a straight line only if your Y axis, the concentration, is a log scale. And then your time is a linear scale. Because it actually is a curve.

Somebody talked about exponential things. It's an exponential decay is another way to describe it. And so – but that line should be a straight line regardless of the concentration which is actually – that's a very good point, because sometimes that line is not linear all the time.

At a certain concentration it may be straight, but at a lower concentration it may not be straight, or vice versa. And in that case we are talking about dose dependent kinetics, in the case of Fentanyl it's pretty darn linear. So it's – it doesn't change.

The elimination half life doesn't seem to change regardless of what concentration you're at.

THE COURT: okay.

BY MS. DAVIS:

Q. At the risk of getting things a little complicated here, maybe we could try for a less equation- oriented explanation. But if you could explain to us the concept of steady state as that term is used in pharmacokinetics.

A. okay. Steady state. When I'm talking about steady state or when most pharmacokineticists are talking about steady state, we are talking about steady state plasma concentration, it's an adjective that usually appears in front of plasma concentration. And we were talking about half lives earlier, well it takes 5 half lives to get to steady state. 5 to 6 half lives. You can say four to 6 half lives, give yourself a range there.

But the thing is that if we think of a hundred molecules, and you dose a hundred molecules in one half life, the hundred molecules drops to 50. Right? And that keeps on dropping. Then 25, then 12 and-a-half, and then 6 and-a-half, so in about four or five half lives, you've got 97 percent of the drug out.

You remember I was talking about it was an exponential decay, when you're going up, if you're trying to reach steady state, it's also an exponential approach. And its complementary to that elimination phase. That's why if it takes 5 half lives to get rid of all the drug, it takes 5 half lives to get the drug to steady state when you have a constant input. So that's the steady state that I mean. Okay.

THE COURT: Well can I ask some c clarifying questions?

MS. DAVIS: Certainly, Your Honor.

THE COURT: I was under the impression that steady state – it's been used in this trial in two different ways, steady state in one way is the – exists when the rate of absorption equals the rate of elimination.

THE WITNESS: Correct.

THE COURT: so basically you get up to a certain level, and then it just – whatever comes in goes out. And assuming that it's the same amount, you end up at a steady state.

THE WITNESS: Correct.

THE COURT: The other way it's been used is in the way it inclines to a plateau level, in general. And when you get to the plateau level, which I guess is what you're saying here, you go through 5 half lives. You get to a plateau level, and that's when it levels out, and you do not get any greater amount into the blood, into the plasma?

THE WITNESS: The five half lives is to reach steady state in the body or in the general circulation. To get to steady state plasma concentration it takes 5 half lives.

The other steady state that has been talked about in here is the – is really, I would prefer to use the term constant input rate.

THE COURT: Right.

THE WITNESS: So if we talk about steady state, if transdermals we are talking about constant input from a transdermal, so if you have a constant input regardless of when the constant input started, it's still going to take 5 half lives of constant input to get to steady state plasma concentrations.

THE COURT: Okay. So it's – steady input – steady rate. So you get –

THE WITNESS: Constant input versus steady state plasma concentrations. Then we could avoid all that confusion, we talk about constant input versus steady state plasma concentration.

THE COURT: You know, you should have raised your hand from the back of the courtroom.

THE WITNESS: I wanted to.

THE COURT: You know, you just clarified the whole thing.

THE WITNESS: I wanted to. And Bill

Barr and I are very good friends. I wanted to hold up, Bill. You're mixing it up. But I didn't.

THE COURT: That would have been a classic. I have seen lawyers wave to you in the back.

THE WITNESS: I thought that was very friendly. Yeah, that was very nice.

THE COURT: Okay. Sorry.

BY MS. DAVIS:

Q. Dr. Gonzalez, you're familiar with the patent that issued to Dr. Keith and was assigned to Key Pharmaceuticals, that's Plaintiff's Exhibit 9, I believe in your binder. You're familiar with that patent generally, aren't you?

A. I'm familiar with the patent. Give me the Exhibit Number again, please.

Q. Exhibit 9.

A. Exhibit 9. Go ahead and ask your question while I find this thing here.

Q. My question is whether you know if any commercial products were made that used a gel diffusion matrix like the one that was described in that Keith patent?

A. Yes. Yes. Of course. Nitrodur was based on the Keith patent.

Q. And can you tell us what the Nitrodur product was?

A. Nitrodur was the gel matrix, a gel diffusion matrix. It was a monolith patch – there is the patent. I never did find it in here. There it is.

It was a gel matrix, very much as what has been described already in this case. It was round and white, it wasn't square as a patent describes, it wasn't so many inches by so many inches. It was circular, and it was manufactured in different surface areas. And I thought we had a slide for that. Or – there we go.

So that's a real Nitrodur photograph from the old days. The aluminum foil – this thing here, this aluminum foil you're looking at was the outside wrapper.

That's what the patient would see. Then when the patient removed it, then you had – this is the gel matrix there. This white gelatinous material. You see that aluminum foil. That's to keep the nitroglycerine from going up into the air.

Obviously, this is the surface that's applied to the patient'. And then you have – there is paper tape that was an overlay of hyperallergenic paper tape that went over the aluminum foil and over the matrix, and it held the patch to the skin. And this particular one, the gel matrix, is 20 centimeter squared. The entire patch was like 67 centimeter squared if I remember right.

And it was manufactured, we had a 5, 10, 15 and 20. But the overlay was always the same size.

Q. So the only thing that changed was the surface area of the actual active drug matrix?

A. The gel matrix, correct.

Q. And the way that worked was that the gel matrix would just be pressed directly up against the skin of the patient?

A. Correct.

Q. Was Nitrodur considered a success product for Key Pharmaceuticals?

A. it was for Key. Probably became our most successful product at the time. Yeah.

Q. Now are you familiar with the fact that certain matrices were made at Mylan at the university of Michigan according to example 6 of the Keith patent?

A. I became familiar as a consultant in this case, or expert witness in this case.

Q. Did you review the results of in vitro testing that was conducted on those matrices?

A. I reviewed all the in vitro data that was generated on those different patches. Yes.

Q. Let me show you what's been marked previously as Plaintiff's Exhibit 15.

A. Got it.

Q. is that one of the documents that you reviewed to provide you with data concerning in vitro test results on the Keith matrices?

A. Yes, that's the srini Venkateshwaran memo. Right. Srini sounds good to me.

Q. Let me ask you, Dr. Gonzalez, when you're conducting an in vitro test on a transdermal system, does the surface area of the patch that you are using to conduct the in vitro testing make a difference?

A. Not for in vitro purposes, no.

Q. Can you explain why not?

A. Well it's always tested on this Franz, f-r-a-n-z flux cells. That's the name of the inventor, Tom Franz. And Tom designed a certain cell, you know, and it's been modified over the years by different people. You saw a fancy diagram of a Franz cell earlier in this case.

And depending on the laboratory, the surface area of the Franz cell is fixed. So you put – if you put an intact patch' – like in the case of a duragesic which is sort of a plastic bag, you can't cut it. You can't trim it. So you put the whole thing in there. But it's only the surface area of that patch that's exposed to the cadaver skin that counts. And that's why the diffusion through the cadaver skin is always expressed in normalized to centimeter squared.

Q. So no matter what size patch you use in terms of the squared centimeters, what you're really looking at is the size of the amount of skin that you have in your diffusion cell, and then you're converting that into per square centimeter?

A. Correct. The flux is expressed per centimeter squared. Right.

Q. Dr. Gonzalez, I would like you to summarize the results of the in vitro testing that was done on the Keith matrices, and I think you've prepared a graphic to assist you in that?

A. Sure. Right.

Q. Can you just explain to the court what this reflects?

A. Even though one of the court reporters showed me how to use this ELMO, I'm having trouble with it. If somebody has got a laser pointer, I'll gladly take it off their hands.

Q. It just appeared in my hands.

A. I can talk about the colors and not even point, okay, well we see a green shading in here going from about point 5 on the Y axis, is the label, micrograms per centimeter squared per hour. So the green goes from point 5 to 10. That's right out of the Alza patent, Gale patent, point 5 to 10 centimeter squared per hour.

And the preferred range is orange which is 1 to 5 micrograms centimeter squared per hour. And the Keith matrix or the Keith monolith are the blue dots. They all fall within that orange or yellow – orange here, zone, which is a preferred – it's a window for the preferred or the best flux range.

Q. And does the blue dot reflect the flux over the entire 72-hour period that the in vitro test was conducted?

A. Yes. These dot are actually representing the flux actually through different cadaver skins at 72 hours. And each dot is a different experiment using different lots, in some cases, but in all cases using different cadaver skins.

Q. And is it fair to say that in every one of those in vitro studies, the flux rate fell within the range that was reflected in the '580 patent for flux rates?

A. Yes'. Because you see all the blue dots are all within the preferred range of 1 to 5.

Q. And are you able to determine from the results of this in vitro flux testing whether the Keith matrix would be able to deliver analgetically effective rates of Fentanyl from reasonably sized patches?

A. well as I understand, this preferred range is preferred because this is supposed to be an analgetically effective rate. So if all of them fall within the orange, then all of them would be capable of delivering an analgetically effective dose.

Q. And were there any studies in which the in vitro testing showed flux rates that were too low to deliver analgetically effective ratesof Fentanyl from reasonably sized patches?

A. No. All of these are within that range. So if you make – if you took a product over here, let's say this one, this one is kind of easy. Let's say this is about 2.5 micrograms per centimeter squared per hour.

That if I multiply that by 10, that would give me 25 micrograms per hour as a dosing rate, or what we can call a total flux dosing rate, is probably a better term there.

Q. So if you took that flux and put it in a patch that had an area of 10 centimeters squared, you would get 25 micrograms per hour?

A. Yes'. This should give us – this particular hypothetical product, the term is bounced a lot around here, this hypothetical product would give you something comparable to that lower – the lower approved dose of the duragesic.

Q. Have you also analyzed the fluxes that were obtained from the Keith matrix looking at the time points leading up to the final 72-hour average flux?

A. Sure. In the Srini memo they gave all of the times that are collected. In this particular case, I chose to present it as fluxed – amount fluxed at a mid point in time, because I had seen a memorandum prepared by Derm-Tech where the data had been plotted that way. So we chose to plot it in the same fashion, and so here you see the individual lots tested over time, and even in those cases where the – let's see, like this one, that one, that one, and that one, the first sampling time point the flux was below the preferred range. But soon after that, they were all within the preferred range.

And in some cases even at the first sampling time point they were already in the preferred range.

Q. Did you also analyze whether the administration rate from the Keith matrix was relatively constant over a substantial portion of the 72-hour period in which it was tested?

A. Yes'. We have a graphic to illustrate that, slide. What I did here is I took the three studies where the same sample points were – the same time points were collected, because not all the studies have the same time points. Some had sampling at 8, some had sampling at 18, at 10 hours. But then I looked at three of them that had the same sample time points, and those ones were averaged. You see the mean and standard deviation for duragesic is this orange curve, where it's relatively steady. It starts to drop off. And versus the Keith monolith where it's quite steady up here.

And the first point might be a little bit lower, probably not statistically different. And so I believe that this slide best illustrates how both products had a very steady delivery over the 72-hour period.

Q. And so do you have an opinion based upon your analysis of the in vitro flux data on the Keith matrices as to whether it delivers Fentanyl at a relatively constant rate through substantial portion of the 72 hours?

A. That's right. It's – I would say that it's relatively constant. And not too much fluctuation over the 72-hour time period.

Q. Now I want to turn to the bioavailability testing. Did you review materials concerning the bioavailability study that was conducted by Mylan on the Keith matrix?

A. Yes, I did.

Q. And what materials did you review to

famjliarize yourself with that study?

A. Everything that was contained in the Russ Rackley report. There was several volumes of clinical study reports that I looked at.

Q. Let me ask you to look – I think it's in your binder.

A. Part of it is, right.

Q. Plaintiff's Exhibit 269.

A. Exhibit 269. Correct.

Q. Is that the Russ Rackley report that you're referring to that you reviewed and relied on?

A. That's correct. Actually you've got quite a bit of it here. You might have everything in here.

Q. Dr. Gonzalez, let me ask you first as a general proposition, is comparing a product to a pre-existing known product a recognized way to demonstrate the effectiveness of that product?

A. I think I better hear that question again. I think I was still looking at things and – start again.

Q. Sure. Sure, well is it your understanding that in this bioavailability study there was a comparison between duragesic and the Fentanyl transdermal matrix, do you understand that?

A. Sure. Of course. Yeah. It was a cross over study. Yes.

Q. And I was just asking whether in your experience in doing studies of products over the years, whether it's a recognized way of demonstrating the effectiveness of a product to compare its performance to a known product?

A. The whole premise of the Office of Generic Drugs is to facilitate the approval of products. And so from the very beginning it was described as bioavailability studies or studies that studied the rate and extent of absorption of a drug, that bioavailability studies would be used in place of efficacy studies. And a bioavailability study has to have a reference, has to have an established reference.

In this case duragesic was a commercial product. And so that would be the established reference. The same – rather the same subjects are used to test both the reference product and the test product. And the subjects are crossed over in that the half of the population is dosed with one product; the other half is dosed with one product, and after a week you cross over the groups so the group that got treatment A then gets treatment B, et cetera. And so it's completely common practice to use' a reference product.

Q. Did the bioavailability study conducted by My!an note whether there were any side effects for either of the Fentanyl systems tested?

A. Yes. There were. I went through the case report forms specifically looking for that. And there were 8 adverse events. There were 8 subjects that complained about side effects, in other words.

Q. And on which products were those side effects noted?

A. The Mylan product had three subjects with side effects. Duragesic had 5.

Q. And were any of those any kind of major side effects, life threatening, or anything like that?

A. No. I think they were the typical common side effects that you expect to see in any bioavailability study where subjects are being stuck multiple times for blood samples. So there were complaints of nausea, headache, shakiness, you know, that kind of stuff.

These were healthy volunteers. So all of them were dosed with Naltraxone prior to being administered the narcotic, so you wouldn't expect to see any serious narcotic adverse event.

Q. Let's turn now to the results of that bioavailability study. You've reviewed the results of that study?

A. Yes, ma'am.

Q. And could you summarize for us what the results of the bioavailability study showed?

A. Well this next slide summarizes really the results of the study. This slide illustrates the mean plasma profiles for the Mylan product as well as the duragesic, the referenced product. You have drawn in a red line here for the minimum effective concentration as expressed by the PDR. But the interesting thing is you can see that the duragesic product has plasma concentrations that were on the whole, lower than duragesic. I'm sorry. Than the Keith matrix.

So duragesic had lower plasma concentrations averaging over the entire 72-hour period.than the Keith matrix product.

Q. Did you review a statistical analysis concerning the results of the bioavailability study?

A. Yes, of course.

Q. And –

A. This next table summarizes it.

Q. And is this the table? we have reprinted it, but this is a table that comes from the clinical study report; is that right?

A. Yes. That was a summary table of all of the pharmacokinetic parameters prepared by Dr. Rackley in the report. And this is an accurate reproduction of that table.

Q. Could you walk us through what the various terms used in this chart mean, and then we can talk about what the numbers are?

A. Okay. First time I want to point out that this is mean of – the mean pharmacokinetic parameters. And then in parenthesis, he has shown the percent coefficient of variation, which is a measure of the variability. Okay.

So – and this is CV, by the way, is calculated simply by taking the standard deviation and dividing it by the mean, by the mean parameter, and multiplying by a hundred. And this was from 9 subjects. Trying to find that. There it is. 9 healthy adult subjects.

It was a mixture, by the way, of males and females, but it was predominantly males. The first parameter is AUCL or meaning AUC to the last time point. So you sample throughout time, you look at the last plasma concentration that was collected. And you integrate the area up to that point.

And they are saying a very common integration technique – those of you who did calculus you may remember the trapezoidal rule. That's what's used to calculate the area of the curve very, very standard.

Q. Can I just slow you down, Dr. Gonzalez. Can you,just explain to us what AUC stands for? what that means?

A. Oh. That's area under the curve. AUC. Pharmacokineticists, we use very simple abbreviations. Area under the curve, AUC.

Q. For those of us that don't remember the trapezoidal – whatever it was.

A. And sub L right. So anyway, you can calculate the area under the curve to the last point. You see it came out higher for the Mylan product than for duragesic. The next AUC is AUC to infinity. The purpose for that is because the FDA recognized the AUC to infinity as being a measure of the amount of drug that was absorbed. And so that is a value that is AUC to the last time point with a segment added onto it which just extrapolation to infinity. It's just another math trick.

And you can see they parallel each other, of course, as you would expect. And then this one is CPEAK, which has also been called C-max in the discussions here, it's the maximum plasma concentration. Again you can see the Mylan product had a higher value. And then here is a ratio of the two LSMEANS, means least squared mean which comes from an analysis of variance. And then the 90 percent confidence interval comes from a statistical test, that is referred to as a 2, one sided test procedure which is a – the preferred FDA recommended methodology.

So the confidence interval to show bioavailable equivalents is 80 to 125 percent for lot transform pharmacokinetic parameters. In this case you can see that it was higher than 125. So this is statistically showing that the Mylan product was statistically higher than the duragesic for AUCL, AUC infinity, and C-max or CPEAK in all 3 cases, and the ratio was, you know, all of it right around a 40 percent higher for the Mylan product.

Q. Can you conclude from your analysis of the AUCL, the AUC infinity, and the CPEAK how much higher in terms of amount of drug absorbed the Keith matrix was than the duragesic product?

A. That's very easy because from this ratio, you can see that it was – in this case, AUCL was 41 percent higher, this was 39 percent higher this was 44 percent higher. That's why I said it was about 40 percent higher for the – the Mylan product was 40 percent higher for the 3 critical parameters asked for by the FDA.

Q. Below CPEAK there on this chart there is an abbreviation K-E-L. Can you explain what that is?

A. That's a rate of constant elimination that we were talking about earlier. You can see those values are very similar for both products. The reason is because that's a property of the drug in those subjects, the right constant elimination. And then this is converted over to half life. So you see that for Mylan was 20.3, for duragesic 24.75, 24.8. There is really no difference between 20.3 and 20.5, practically the same.

And then this one is the maximum, the average – the average maximum – no, average time to reach maximum concentration. So this is calculated by taking each individual value and averaging them out or taking the mean of that. And you can see that it was longer for the Mylan, just like the mean plot showed just like Dr. Barr pointed out. it was longer for the Mylan than for the duragesic product.

But you also see there is no stats here. And that wasn't my doing. That was Rackley's doing, but it was correct because the FDA does not ask for statistics on these parameters. If this study had been submitted to the FDA, this summary table would have been fine.

Q. Focusing specifically on the TPEAK, because I think we have heard some discussion about that in the earlier testimony. Does the fact that the duragesic product had a lower TPEAK value than the Keith matrix have any impact on your analysis of the performance of the Keith matrix?

A. Yes. It's just letting me know that it's – the maximum concentration was lower than the Keith matrix, and the maximum concentration is looked at by the FDA as a measure of maximum exposure. Or another way of saying it is, it's a measure of rate of absorption.

Q. I'm sorry. I think maybe – I think I may have misspoken.

A. okay.

Q. Because I think that it sounds like from your answer it sounds like you heard me to say the CPEAK. I was asking about the TPEAK.

A. oh, TPEAK. I'm sorry. Okay No. I was – I did hear CPEAK. The T-max or TPEAK it is different. And that's fine. I mean it's not a statistically critical parameter. It is not critical for FDA approval, it's not something that I would worry about significantly in a bioavailable study or bioavailable study.

Q. Now you mentioned the concept of bioequivalents. And my question is, based upon the data that you've looked at here, would you know what you needed to do to make the Keith matrix bioequivalent to the duragesic system?

A. well bioequivalency is achieved when two products have the same bioavailability. All right. So to have the same bioavailability, you have to have the same area, and the same C-max, or to say it better, you have to have areas in C-maxes that are not statistically different.

So since the Mylan product was 40 percent higher, I would want to reduce that patch size by 40 percent. So instead of a 9, I would reduce it by 40 percent. So 9 times.6, for example, if I take – if I make a patch that's only 60 percent of the other one, then I would come up with a patch that was about 9 times point 6 is 54, so I would come up with a patch that's 5 point 4 centimeter squared. And that patch should give me something that's going to be pretty darn close. Not statistically different.

Q. Based on your analysis of the results of the bioavailability study, do you have an opinion as to when during the course of the study the Keith matrix began delivering Fentanyl at an analgetically effective rate?

A. This red line marks the minimum effective concentration. So the green line, which is the Keith matrix, crossed that red line after some time – sometime after 10 hours. So somewhere – could be 12, 14 is the time when the Keith matrix started delivering enough or delivering at an analgetically effective rate concentration. It produced an analgetically effective concentration. There we go. That's better.

Q. And for what portion of the administration period did the Keith matrix maintain the concentration, the plasma concentration, at those analgetically effective rates?

A. From there to 72 hours, because it was above point 2. For that entire period. So 12 minus 72. 60. Something like that.

THE COURT: Dr., let me ask it, that particular diagram, you reached the high point at somewhere – 1.05 I would guess; is that right at 72 hours?

THE WITNESS: That's 72 hours.

THE COURT: is it that point at which the patch is removed?

THE witness: That's where the patch comes off. Correct.

THE COURT: I guess my question is why isn't there some – why isn't there some slippage in a sense, because there is some drug right underneath the skin, and it takes some time to absorb into the system.

THE WITNESS: Well part of the problem here is – and I think Bill Barr also referred to this, is the fact that we are looking at a mean profile. So it's sort of a little deceiving.

It's better to be looking at the individuals as far as trying to evaluate you're trying to look at this profile and trying to get a feeling for what the absorption and the elimination is. And it's very hard to see that from a plot, because as we are looking at this, we are looking not only at absorption, elimination is already going on.

Once the drug gets into the body, elimination is going on and distribution is going on. So it's very hard to really appreciate the absorption process by itself. So the patch comes off, you expect to see what a very – slower thing going on. But absorption is still going on as well as elimination. And if you look at the individual products, you don't see – the individual subjects, you don't see this rapid drop off as you see there.

THE COURT: Well perhaps I shouldn't ask you a question which would call for your speculation, but take the patch off at 72 hours and all of a sudden there is a turnaround. What happens if the patch wasn't taken off?

THE WITNESS: Well actually do you remember when we talked about achieving steady state? We were saying that it's going to take about 5 half lives. Can I go back to the previous – to the table with a summary of the parameters. Let's look at the half life here again quickly.

20 hours for the Mylan product. 24, 25 for the duragesic. Now let's go back to the plot. This is 72 hours. According to that 20 hour half life, if we multiply 20 by 5, that's going to be a hundred hours. It's going to take a hundred hours to reach steady state. If I don't take that patch off, and it continues delivering – and remember we are not really seeing only absorption going on in here. So you can't say that it's going to follow this line here continuously up.

But there will be continued absorption if I left that patch on. But it's going to reach a steady state somewhere around a hundred hours, so it's going to go up until it approaches that steady state, and then it's going to curve off and become a relatively flat line.

THE COURT: So it's clear according to your testimony they haven't reached steady state by this point.

THE witness: Absolutely not.

THE COURT: Steady state absorption versus elimination.

THE WITNESS: They haven't reached steady state plasma concentrations. They may have a constant rate of input, both products may have a constant rate of input.

THE COURT: Right. But not the elimination versus the absorption.

THE WITNESS: We haven't gotten to the place where elimination equals absorption. Which is when steady state happens.

THE COURT: So then go to the Y axis, and then to the X axis at 100 hours. Where are you going to be on the Y axis at a hundred hours when you reach steady state, if you know?

THE WITNESS: It can be – it –

THE COURT: Somewhere above the simulator.

THE WITNESS: It's going to be off of this scale. Yes. And Rackley did a simulation of that. And I don't think we have a graphic of that. But can we do a demo or not? Am I wasting the court's time here?

THE COURT: No. That's okay. You're not wasting my time. As long as you answer my questions. Then you're not wasting my time.

THE WITNESS: Go for it.

THE COURT: Well going up the Y axis, I mean theoretically, Fentanyl within the system becomes more and more dangerous. As you go north. would that be fair to say?

THE WITNESS: As a plasma concentration increase, it's going to become – the drug levels are going to increase the potential for serious adverse events, yes.

THE COURT: okay.

THE WITNESS: Does it become more dangerous? Yes.

THE COURT: Right. So where, if you can answer the question, where is the point at which the plasma level as it goes up the Y axis becomes unreasonably dangerous or becomes dangerous?

THE WITNESS: I was guessing it was going to be up here. And it's going to be 1.2 according to the Rackley simulation over here.

THE COURT: okay, well in particular, I assume that there is a difference between people who are sensitive to – people who are adjusted to narcotics versus people who are not adjusted to narcotics. And obviously there is a product that would be used by people primarily who are not adjusted to narcotics.

THE WITNESS: Addicts. Right.

THE COURT: Non addicts, right. So are you suggesting that once your blood plasma gets up to about 1.2, and if you are not insensitized to narcotics, that you would be in somewhat of a dangerous situation?

THE WITNESS: No. No. Definitely not. Even if I-look at Goodman and Gil man, I see the maximum plasma concentration – or the top of the therapeutic window listed as 3 nanograms per mil, and I probably misspoke when I said addicts. Because there is another group of people that become tolerant of narcotics, and that's the people that are using it chronically for pain. Cancer patients are going to be using it chronically, and they are going to become more tolerant, so they can tolerate much higher levels.

THE COURT: Oh sure. No. I understand that. But you're talking about first obviously you're not getting a bolus shot here at the beginning.

THE WITNESS: Correct.

THE COURT: This is probably an inexperienced person who begins to put the patch on. what you're suggesting is that after 72 hours you're at point 10, to point 11 on the average, and you're non addict at that point, and the question is how far up the scale do you go before, you know, you feel like – well before you're in a little bit of danger is my question.

THE WITNESS: Well again since you're achieving a steady state, and the – with this product, and this elimination half life, and this mean profile, the simulation here shows that you would be reaching a steady state at 1.2. And you're not going to go any higher than 1.2. That's page 43 of the Rackley report. We would like to show that.

And you see that the duragesic product also will go up to reach a steady state. it won't go up as high, of course. But it will go up to somewhere about point 8 or point 9.

THE COURT: Okay. So what you're suggesting is that as the steady state is that 1.2 –

THE WITNESS: Right.

THE COURT: That is not a dangerous situation to somebody who is non – who is a non addict and not sensitized to narcotics? is that your testimony?

THE WITNESS: I think that – well again, going by plasma concentration, the plasma concentration range and something by Goldman and Gil man, or looking at some of the studies where patients – postoperative patients – they weren't cancer patients, were administered some fairly high infusion rates, I think there was one as high as a hundred micrograms per hour IV infusion. Okay.

And some were done by stansky. It was an Alza-sponsored study, the plasma concentrations were higher. But that's a constant input.

THE COURT: okay.

THE WITNESS: The other thing, too, is when I say 1.2 it doesn't mean it's going to go to 1.2 and it's going to stay at 1.2. That's a misconception of steady state. You can be a steady state, but plasma concentrations will be fluctuating, right. And in this particular case the simulation shows that you reach a maximum of 1.2, but then they drop off until you give the next dose.

THE COURT: okay. I apologize.

MS. DAVIS: That's all right. I think I'll shift to a part of my outline that kind of addresses some of the issues that you have been raising, Your Honor, and we may have to go back a little bit.

THE COURT: You've got two minutes to shift, and then we will take a break.

BY MS. DAVIS:

Q. Okay. Dr. Gonzalez, how – let me ask the question this way. Do you have an opinion as to how the amount delivered by the Keith matrix that was tested in the bioavailability study, how that would compare to the amount delivered by the 50 microgram duragesics – I'm sorry, yeah – by the 50 microgram per hour duragesic system or the next higher dose duragesic?

A. it would be less. I calculated that the total flux from this patch is about 35 micrograms per hour. And so 50 would be less. A 50 should produce higher plasma concentrations.

Q. So the Keith matrix, as you analyze it, is producing lower plasma concentrations than you would expect to see in – from a 50 microgram duragesic?

A. Yes.

Q. Okay. And have you – got a little bit ahead of me. Have you looked at the issue of how the variability of plasma levels – plasma concentrations from the Keith matrix compares to the variability that is seen from the duragesic product in various clinical tests that have been done on the duragesic product?

A. Right. This was – this is a figure taken right off of the PDR, and it's based on the study that was reported on earlier here in this case. And this is showing you the subject with – a subject with a minimum, minimal – with minimal plasma concentrations versus subject one that had the highest plasma concentrations. And the range here is something like about 6. Going from here to here is about 6.

Q. So what you're saying there is that the ratio between the minimum plasma concentration you see from that one product is about 6 times less than what you see from the maximum, with the maximum person?

A. Yeah. Basically if I look at the maximum concentration in subject one and I come down here to the – at the same time point for this other subject, that range was about 6 fold. And if I do the same thing here, this is the subject that had the maximum concentration, and maybe I should clarify that this X axis is sort of confusing. Plus the fact that the Y axis are different scale.

This is only up to 3.5. This is up to 5 is up here. Okay. So this plasma concentration is much higher than this concentration because it was a larger dose that was administered. But the range – if we look at the value here which is somewhere, let's say, somewhere around point 5 and the value around 4.5, this range is 9 fold difference.

So the variation in any study, there is always variation, and it has been described that the skin flux alone can vary four to five times.

Q. So the skin flux – the flux of Fentanyl through skin if you used the exact same system on different people, can vary by a magnitude of 5 times, is that what you're saying?

A. Yeah. Even if nothing else is varying in that patient, we are just looking at skin variability, there should be easily a four to five fold variability.

Q. And I think you mentioned earlier that the plasma concentration curves like we see here are reflected in not just absorption but also other processes that are going on?

A. That's right, we are looking at the elimination process as well as distribution going on.

Q. And let me ask you this. is the rate at which Fentanyl is eliminated from the body something that is consistent among individuals or something that varies?

A. no. That also varies. The elimination rate constant varies from different people, and in fact, when we talk about clearance, that's describing – it's different than half life. Clearance is a parameter that includes volume of distribution as well as a rate constant of elimination and it comes from thinking about urinary clearance, and so in the world of pharmacokinetics we came up with this term to describe what volume of blood is cleared of the drug so the units are in volume per time.

Okay. And clearance can vary as much as 5 or 6 fold depending on some data that I saw in the Alza reports, so clearance is also very variable. And the interesting thing about the body is that you have an error from the skin absorption, you have an error introduced – not error but, you know, variability that's introduced by clearance. Well in the world of statistics error is not additive. And it may be multiplied, and so that's what causes this big variation between this plasma concentration and this other plasma concentration, because we don't know what this guy's skin absorption was. We don't know what his clearance was, and ditto for this other guy over here.

THE COURT: All right. Let's take our break at this point. At 3 o'clock. Be back in 15 minutes.

(A recess was taken)

THE court: All right. Ms. Davis, ready to –

MS. davis: All right.

THE COURT: – start?

MS. DAVIS: We will proceed.

BY MS. DAVIS:

Q. If I could provide – I don't think we had up – I think last the availability graphic.

Dr. Gonzalez, if you look at the right-hand side there for the duragesic product, the table from the PDR, do you see that?

A. Yes, ma'am.

Q. What does that reflect about the time frame over which the plasma concentrations continued to increase for the duragesic product that was applied in this particular study?

A. Well I think I should clarify this x axis a little further. Here we are looking – each tick mark represents 3 days as it goes from 1 to 4 and then 7 and then 10. So what they are showing you is patches – a patch has been applied here. It's taken off, and another one goes on. Taken off, and another one goes on. So we are seeing multiple applications. So here's 1, 2, 3 applications.

So on day 10, if we go up, we see that the plasma concentrations have climbed quite a bit. And interestingly enough, the plasma concentrations on day 10 are not that different from the concentrations on day 13.

And if we apply our rule of 5 half lives or so, you had to be at steady state already by day 10, so therefore between day 10 and day 13 you are at steady state, but the levels have been climbing, climbing, climbing, Okay.

Now the other thing on this day – let me a get back to the X axis again. Now all of a sudden the axis has changed from each tick mark being 3 days. So each tick mark is one day for this interval here. No, actually for the rest of the plot. Each tick mark is one day 24 hours. So now we see that over this 3 days this is one patch application. So it's one application. We are at steady state, but the plasma levels were not staying up at one point. It's reaching a maximum and then coming down.

And then the patch was removed here. And then they followed the decline after a patch removal. So the message from here is that it takes several applications to reach steady state which is what pharmacokinetics predicts.

Q. Now have you reviewed additional materials from Alza concerning whether duragesic causes plasma concentrations that rise for 72 hours in some individuals?

A. Yes. Data from Alza you said?

Q. Yes.

A. Yes, I have.

Q. And maybe you could put up graphic X 27502. Can you explain what this graphic shows, Dr. Gonzalez?

A. This is an individual subject from a number – number of subjects that were studied by – well, it was attached to a letter from Mr. Gale, from Bob Gale, and in his letter he's expressing concern that a study design that was meant to show reproducibility from day-to-day, so what they did is put on a patch for 24 hours, and then put on another patch for the next 24 hours, here's 48 here somewhere, and then a third patch. And instead of seeing the same kind of blood levels, what he saw was it was a 75 microgram per hour patch put on for only 24 hours. Taken off. Another one goes on, a third one goes on.

Instead of seeing the same plasma concentration over the three days, he saw accumulation. And they hadn't fully thought it through, and thought about the concept of steady state. So what you're seeing here is that it takes – you're seeing accumulation with the product because you're not at a steady state, and the plasma concentrations are continuing to rise.

Q. Does the fact that plasma concentrations are continuing to rise after 72 hours with a particular transdermal system mean that the amount being delivered by that system is rising over that 72-hour period in terms of the amount that's coming out of the system and into the patient?

A. Well intuitively I want to say yes. But the problem is that you have to remember what I said earlier. Is that not only do you have absorption going on, you also have metabolism and excretion going on, and distribution of the drug. So all of those parameters are working together to end up with a plasma concentration. Could it be that the drug is still absorbing, or could it be that the clearance is changing for that patient? A transient change in clearance, a dropping off of clearance is going to cause a rise in the plasma concentration. So it can be a little confusing.

Q. Am I correct that if you have – for individuals who clear Fentanyl at slower rates, you would expect to see higher plasma concentrations from the same input of drug?

A. correct.

Q. Okay. And in this particular study that was done by Alza, what was the plasma concentrations in terms of nanograms per milliliter that was reached by this particular subject?

A. Well you can read it right off the – of the Y axis here. It's about 8 nanograms per mil. For a – for a 75 microgram per hour patch.

Q. And Dr. Gonzalez, let me take you back, in response to one of the Judge's questions, you were discussing steady state. And I think the Judge was asking you about the plasma concentrations in which Fentanyl would become dangerous or you might expect serious adverse events, do you recall that exchange?

A. Yes, I do. Right.

Q. Is it your opinion that a plasma concentrations – concentration of 1.2 nanograms per milliliter would be a dangerous or a high adverse effects level?

A. No. What I was trying to say was that – what I said, I should have stated more clearly, is that the one point 2 that I was referring to was that would be the maximum steady state plasma concentration that would be reached. I was not saying that 1 point 2 would be toxic or lead to serious adverse events.

So your question was – His Honor's question was when is it going to become toxic. And I was focused on when are we going to get to steady state. So no. I'm not saying that 1.2 was or wasn't toxic. But if we go the average –

THE COURT: Is there a number, or is there just a scientifically accepted number of nanograms per milliliter of blood that would be toxic? I mean do you know or –

THE WITNESS: No. I think a way – the way that the PDR refers to it is that there is a higher incidents of side effects, so that PDR uses 2 nanograms per mil as an undesirable plasma concentration, because you're going to see more adverse side effects.

Doesn't mean toxic. Toxic to me means going to kill somebody. But if I see higher incidents of side effects, I don't want to be there either.

THE COURT: The thing is respiratory depression which is – an expression I picked up in this trial.

THE WITNESS: Respiratory depression is one of the toxic side effects. But remember, respiratory depression doesn't mean that you stop breathing. It just means that your breathing rate might get decreased. And then again, I want to say that Goodman and Gilman used the figure of three nanograms per mil. So – the PDR is more conservative. So I'll go with that.

But you know I'm not a M.D. so I'm not

THE COURT: okay.

BY MS. DAVIS:

Q. Dr. Gonzalez, you were here for Dr. Barr's testimony about the plasma concentration data that was – the plasma concentration data from the bioavailability study; correct?

A. Yes, ma'am. Sure. Yes, I was here.

Q. And did – let me strike that. Did you understand that Dr. Barr's position in his testimony was that you could look at the plasma curves from the bioavailability study and draw a conclusion about the rate at which the Keith system was delivering drug?

A. It was pretty confusing, especially when he had a long exchange with the Judge about steady state. It was fairly confusing. I'm not entirely sure that I understood exactly what he was saying.

But he was making assumptions that you could look at the plasma profiles and tell when steady state had been achieved or hadn't been achieved. I don't agree with that.

Q. Now you've talked in your testimony today about the fact that the plasma concentration curves include both absorption and elimination components?

A. Yes.

Q. Correct? Did – is there in pharmacokinetics, are there ways that you can analyze data to determine from the plasma concentration what the rate of absorption is in a rough way?

A. Yes. Yes, there are.

Q. Did Dr. Barr ever do an analysis along those lines that was designed to determine the delivery or absorption rates from the Keith matrix based upon an analysis of the plasma concentration data?

A. Yes. He calculated the absorption rate from both duragesic and the Keith matrix using a mathematical de-convolution. Just the way it sounds.

Q. What did Dr. Barr conclude from performing that mathematical de-convolution about whether the Keith matrix delivered Fentanyl at relatively constant administration rates?

A. He concluded that the Keith matrix delivered Fentanyl at a relatively constant input. And – but I have to say that I don't agree with his calculations. I have to say this first. That he made a number of assumptions which were incorrect, and to – in arriving at those calculations.

However, he still concluded that the Keith matrix was delivering at a constant input for 40 plus hours, if I remember right. 40 hours or 42 hours. I don't remember exactly.

Q. Just so – just so I understand, you don't agree with the way Dr. Barr performed his de-convolution, but even the way he performed it, he found constant delivery for that period from 40 hours through the 72 hours; correct?

A. Yes. Yes. It was from 40 hours, and now that I think about it, looking at his plot of the absorption profile, I was thinking that it looked linear from about – not linear, that it looked like it was a straight plot from about 30 hours all the way up to 72 hours.

Q. Going back to the issue that we have had some testimony about already, about when various products reach steady state, let me direct your attention to Defendant's Exhibit 1327. Which should be in your binder.

A. It's missing, but I can read it perfectly well there, so that's okay.

Q. All right. If you don't mind looking at the screen, that will be fine.

A. Okay.

THE COURT: Mine's missing too.

MS. DAVIS: Yours is missing too? Let me get you a copy.

THE WITNESS: I'm glad I'm not the only one who lost it.

THE COURT: That's okay.

MS. DAVIS: While I'm up here, you can have a copy too.

THE WITNESS: Thank you.

BY MS. DAVIS:

Q. And Dr. Gonzalez, in particular I want to direct your attention to the page that has Bates number 72272 on the bottom.

A. Got it.

Q. Do you have it? And I want to focus, in particular, on that slide in the middle that says pharmacokinetics. do you see that?

A. Yes. Yes, I see it.

Q. And this is an Alza document, a slide show concerning duragesic, is that your understanding?

A. Yes. I read through this after this first came up here in the trial earlier this week. And I got a copy of it. And it looks to me like it was a sales training document.

Q. Let me ask you to look down in that slide on pharmacokinetics. And first, let me ask you about the entry there that says steady state serum concentrations, do you see that?

A. Right.

Q. And what does that reflect about the time at which Alza thought that you would reach steady state serum concentrations with duragesic?

A. Several sequential is the Key thing to me there. I just felt very smug when I saw that because I felt that it is supporting what I have been lecturing here about. It takes 5 half lives to get there. So with 72-hour applications, after two applications, you're beyond a hundred hours, so after two applications you should be at steady state. And so with three applications, you're definitely there. So it's going to take several applications to reach a true steady state. That's what they are saying there.

Q. Below that there is reference to half life. Do you see that?

A. Yes.

Q. And does that reflect the range of half lives for the duragesic product?

A. well from this particular study that they are talking about, the problem with being a sales training document, it's not documented where the heck the values came from. But in this particular study they came up with an average half life of about – that squiggly line means approximately 17 hours. They are saying the range was 13 to 22 hours. That's really not too far. 17 is not too far from 20 hours.

Q. While you have that exhibit in front of you, and we have it up on the screen, does it also provide some information about the T-max? Where you would expect to find T-max for the duragesic product?

A. Well it says that the peak serum concentrations are occurring at 24 to 72 hours. So they are talking about maximum concentrations, but then they are talking about the time in which they occur. So this is actually TPEAK, T-max or TPEAK. So the answer is 24 to 72 hours is when they are saying that they expect to see the maximum plasma or serum concentrations.

Q. Now you mentioned, I think, earlier, Dr. Barr's testimony about the looking at the shape of the individual plasma curves for the Keith patch and for the duragesic patch?

A. Yes, I did.

Q. Did you do anything to look further at that issue after you heard Dr. Barr's testimony?

A. Yes. The afternoon after listening to Bill, I went back to your offices, and I plotted the same data from the Rackley report. I plotted all the individuals on a similar spaghetti plot type thing like he had shown. And I would like to see them up there.

Q. And let me ask you just as a preliminary question. Do you think it's a good way to analyze the issue of how a patch is delivering and the rate at which it's delivering just by looking at plasma concentrations?

A. No. It goes back to what I said earlier, that you have too many processes going on at the same time. What it is good to look at is to see how the individuals are really behaving.

Q. Well let's – and why don't we put on the screen X 27861. And can you tell – explain to the court what this reflects?

A. Yeah. Those are, Your Honor, those are the same individual profiles that Dr. Barr showed where the curves all were looking relatively flat down here. But the problem was the Y axis that had been used in that plot. So here this is the same data except that the Y axis is a reasonable Y axis.

I mean it's long enough so that the highest plasma concentration fits, but not so long that all of the data is going to be compressed. If I had plotted this from zero to 5, then this curve would all be compressed down on the bottom of the graph. And so, you know, you just look at it in an optimized plot, and you can see that, I believe, that Bill was saying oh, in about 24 hours, you know, all of them are – all of the curves are getting flat. All of the subjects are at a steady state. You can see the levels are going up and coming down. Just – which is really what even the Alza sales training manual here is telling their own sales reps. One subject here.

We could say that this one subject here, boy, this levels look awfully flat. Well and here's 72 hours, and the value is going up, and here's 48 hours when he reaches his maximum. But there is no way that we can look at this thing and say that steady state has been achieved during one dosing interval.

Q. And did you do the same sort of plot –

A. Excuse me. By steady state I mean plasma concentrations at steady state have not been achieved in one dosing interval.

Q. Can this kind of curve tell you anything about whether steady state in the sense of relatively constant input of drug has been achieved?

A. No. The only thing that this curve can tell me is that it looks like a lot of the subjects are reaching a maximum somewhere around 24 hours. But not all of them. Some of them take a little bit longer to get to maximum.

Q. And did you do the same sort of plot for the individual plasma profiles from the Keith matrix?

A. Of course. Yes. There they are. There's subject one, which is higher than the other subjects. But if we look at this they are taking longer to get up there. And look at that. It's like somewhere around between 60 and 72 is when they are reaching their maximum. You know, I can conjecture that it's because this product is continuing to deliver up to 72 hours, and the product – other product was not delivering up to 72 hours, and that's why you're reaching an earlier maximum, and they are starting to come down. But it would be conjecture. The profiles are different.

But I also don't see individuals that are just going up way crazy. You know, and oh, just continued absorption. One subject biases all of the data, and if you have – again, if the plots are not plotted carefully, you can make all kinds of erroneous assumptions from it.

Q. And I believe here in this plot it is subject one that's the highest of all the subjects in terms of the plasma concentrations?

A. Yes, it was.

Q. If you could flip back to the previous graphic. And I think you changed the colors on us here. But is it subject one that also reaches the highest maximum plasma concentration?

A. Yes, here's the code. It's a red square is subject one, and subject one also had the highest plasma concentration for duragesic.

Q. Dr. Gonzalez, do you recall Dr. Barr's testimony concerning the reliance that Dr. Flynn placed on certain studies in determining a therapeutic dose of Fentanyl?

A. Oh the – yeah, to arrive at the dosing rate. Right.

Q. Do you agree with Dr. Barr's criticism of Dr. Flynn's reliance on those particular studies, and referring particularly to the White study that I believe was discussed at length?

A. No. I don't agree.

Q. Can you explain why not?

A. Well again, I don't think Bill has ever set in with transdermal formulators to try to arrive at what would be a goal of an infusion rate, but I will tell you for Nitrodur, for instance, we relied on sublingual data to arrive at our goal for how much drug to put into the patch. And what the goal for the drug was going to be.

We used sublingual data to come up with a clearance value, and we used that clearance to calculate what kind of infusion rate we would get to get to a certain plasma concentration, we never had an IV infusion study before. Nitrodur had been tested a bunch of times. Well actually we never – well forgetting about Nitrodur, going to other products, we worked on Estradiol. we never had Estradiol IV infusion, we worked on nicotine. God forbid, we never had IV infusion with nicotine. I worked with Isosorbide Dinitrate, another cardiovascular drug. We used sublingual data for that.

Currently I'm working on a project with another company, that project is almost ready for FDA submission. We have never done an IV infusion, we relied on orally-administered drug to arrive at the transdermal flux rate. And' so in the real world you do not have to have intravenous infusion. What the luxury that Fentanyl had it was available as an intravenous drug, so there was data on IV infusion. The white paper suggested an infusion rate. There was a whole paper also that also had a flux rate, there was a paper by Andrews, a review article by Andrews, that had a whole list of different clearance values. All of those values could have been used to calculate infusion rates.

Q. Do you recall Dr. Barr testifying about the studies that have been conducted by Dr. Nimmo and Dr. Stansky that he characterized as being constant IV infusion studies?

A. Yes, I do remember that.

Q. And have you reviewed information about how the Nimmo study was conducted?

A. Yes, I've looked at that.

Q. Let me direct your attention to Defendant's Exhibit 1246.

A. I'm sorry. I said it was a published paper. I said it under my breath.

Q. I called up the wrong one. How about Defendant's Exhibit 1307.

A. That's the Nimmo article.

Q. And is that the article you're referring to that you reviewed to determine how the Nimmo study was conducted?

A. Yes.

Q. In the study by Dr. Nimmo, did the patient receive any bolus doses during the infusion period when the Fentanyl was being infused?

A. There was premedication, a bolus dose of Fentanyl 100 micrograms. Sorry, it's down over here. I can't read it from here.

Q. Maybe we could blow up the paragraph there that you're referring to.

A. Yeah. A bolus dose of Fentanyl, a hundred mikes was given intravenously, period. At the end of anesthesia each patient was prescribed – blah-blah-blah. But there was a hundred micrograms administered, and then there was an infusion. I don't see the infusion. They started the infusion before the induction of anesthesia. You know what, the infusion was earlier than this. And it was expressed in terms of per kilogram. 1.5 micrograms per kilogram and point 5, if I remember right. It was a range from.5 to 1.5.

Q. Just so I'm clear, the infusion of Fentanyl was started 1 to 2 hours before the patient was anesthetized?

A. Yes.

Q. And then as part of the process of the surgery and the anesthesia, a bolus dose of Fentanyl was given to those patients; correct?

A. Correct.

Q. And did the patients also have access to bolus doses of morphine to – if they had additional pain?

A. Right. I believe it was under the discretion of a nurse that things cyclimorphed, I wasn't familiar with that, it was a mixture of morphine and cycloserine intramuscular. It looks like they got 10 milligrams of intramuscular morphine when they requested it, or at the discretion of a nurse. I guess they would complain to the nurse, and the nurse would give it. I think that's later in the text where that's explained.

Q. And by IM there, it's referring to essentially a shot that you get in your muscles?

A. Yeah. Intramuscular injection. Correct.

Q. For how long did the patients in the Nimmo study have the constant infusion of Fentanyl?

A. For the duration of the anesthesia. For the duration of the surgery.

Q. Well if you look at the – maybe we need to back up and draw your attention to the summary on the upper right-hand corner. Let me ask you if that reflects for what period of time Dr. Nimmo studied the Fentanyl infusion?

A. 24 hours. Compared with placebo infusion an analgesic regimen for 24 hours. That's the infusion rate I was looking for. 1.5 micrograms per kilogram per minute or.5 micrograms per kilogram per minute.

Q. Does that mean that Dr. Nimmo varied the dose that he gave his subjects based on how much they weighed?

A. Well the point 5 or 1.5, there were two different doses administered, but that per kilogram means, yes, it was adjusted to the weight of the patient, and the average weight as I recall, was around 60 kilos. So for the low end you figure, that's about a 30 microgram per minute. Point 5 times 60. Yeah. That's 30 microgram per minute. And then this one is 90 – the other one would be 90 micrograms per minute. 1 point 5 microgram times 60. Yeah. 60 kilograms.

MR. FIGG: That's a typo in the summary.

MS. DAVIS: Okay. I knew there was something wrong with that, and I was just trying to figure out what it was.

THE WITNESS: Sounds pretty high when I think about it.

MS. DAVIS: I was going to say that's an awfully high dose.

BY MS. DAVIS:

Q. If you would please just look at the body of the paper and see if it reflects that actually those are the figures per hour. Actually I think it's right below ther summary in the patients and methods section.

A. Not per minute. It's per hour. There it is. Yup.

Q. I hope Dr. Nimmo went back and got that corrected before he got too confused. Now let me direct your attention to Defendant's Exhibit 1246. And let me direct you to page that has Alza 57930 on the bottom. That's probably the easiest way to find it.

A. Yeah. I've got it. Right. Yeah. I remember this one. Got it.

Q. OKay. And is this describing the study that Dr. Stansky conducted?

A. This was a study contracted by Alza to Dr. Stansky at Stamford, to do IV infusion in patients, and surgery patients. Yeah.

Q. And does this indicate that Dr. stansky gave the individuals involved in this study bolus doses, that he refers to as loading doses, during the time that the infusion was being studied?

A. Yes. He had a – used this as a technique to get the subjects up to – to get their plasma concentrations up into a therapeutic range as quickly as possible. He gave a loading dose of a hundred micrograms followed by an infusion of 25 micrograms per hour. Et cetera.

He's got the values here. 200 and 400. Here's the different infusion rates. This was for different ones –

Q. So he started things up by giving them the bolus does, and then started a continuous infusion after that?

A. Correct.

Q. And am I correct that in Dr. Stansky's study as with Dr. Nimmo's study, anesthesia and surgery occurred after the infusion had started?

A. Yes. This was – well let me check this carefully here before I answer that again. Because it's been awhile since I've looked at this. Okay, after 5 minutes of maintenance infusion. So first he gives the loading dose, then he starts the IV infusion, then anesthesia gets started. So yes, he had started it. in the case of Nimmo, I think he started earlier.

Q. But in both cases it was started before the surgery and the anesthesia occurred?

A. Yes.

Q. And look down at the bottom of that paragraph. For how long did Dr. Stansky administer the infusion of Fentanyl?

A. The Fentanyl infusion will be maintained at a predetermined rate into the postoperative period for a total of 24 hours.

Q. And in Dr. Stansky's study, the subjects also had the opportunity to have additional pain medication in bolus doses if they had pain?

A. I'm reading. I have to read this. Yeah. Morphine sulfate was administered in one milligram IV doses. He starts, if the patient complains, it's on page 3. Up here somewhere. If the patient complains of pain at a specific testing interval or between testing intervals, morphine sulfate in one milligram IV doses will be given at 10 minute intervals until the patient is comfortable and pain free.

Q. So what does that reflect about whether the individuals in Dr. stansky's study were given bolus doses of a narcotic analgesic during the time that they were also receiving the Fentanyl infusion?

A. They got morphine on top of the Fentanyl. And they got a bolus of Fentanyl up front, and then the infusion, and then they got morphine when they came out of anesthesia.

Q. Dr. Gonzalez, based on your review of the in vitro testing and the in vivo testing that was done on the Keith matrix, if the Keith matrix that was used in the bioavailability study were reduced in area to a 6.5 centimeter squared patch, would it deliver Fentanyl at anailgetically effective rates?

A. Yes. 6.5 is greater than what I had estimated at 5.4. I said 5.4 would be bioequivalent, so that should be analgetically because it would be comparable to duragesic. So 6.5 would be larger than the 5.4 that I calculated, so 6.5 would definitely be delivering at an analgetically effective rate.

Q. Let me ask you, Dr. Gonzalez, because we have heard some testimony this morning concerning the flux rates that were obtained in the very first flux study that Dr. Flynn did. Do you recall hearing some of that testimony this morning?

A. Yes. Very well.

Q. is there anything inconsistent about the flux rate that was seen in that study of 1 point something – I don't remember the precise figure, and your conclusion that a 6.5 centimeter squared Keith matrix would be expected to deliver Fentanyl at analgetically effective rates?

A. I'm glad I don't have to answer yes or no right now, because that's definitely a yes and know answer. Because it's not really in conflict, because the value that Dr.'Flynn was relying on was based on one cadaver skin. So he got this low flux from this one study. Right.

Then you multiply that by 6.5. You say oh, the flux from this patch is only 6.5 micrograms per hour. But our calculation over here shows that the flux was for a 9 centimeter squared patch. Was 35, and I said hey, we cut it back to 5.4. It's going to be about 25 micrograms per hour, so here I'm saying that the in vivo data shows 25 micrograms per hour flux. His in vitro data shows 6.5. That's because he used only one skin. We had 9 subjects over here. Right. So we have an average of 9 subjects.

If Dr. Flynn had used 9 cadaver skins and gotten an average flux, he probably would have gotten much better value, so I think his whole problem was stopping with one cadaver skin. Maybe that's the only sample he had.

Q. Well – and if you could – maybe to put this into context, maybe we could put back up the graphic on the in vitro flux. I think it's the first graphic that we used.

A. The first slide.

Q. It's –

A. Yeah, this one. I mean here again, well I guess are you going to ask me a question? Or do I continue lecturing here?

Q. My question is how do the various in vitro flux data points that you see here, how does that relate to your view about the question I just asked you about the results from Dr. Flynn's first test?

A. Continuation of the same question, right?

Q. Right.

A. We looked at Gordy's slides here, he's got a value too. Barely above one. And it was something like 1.13 or something like that. And it really was the lowest value that came out of all of these different things. Here's one that's kind of close. But that same lot repeated on a different skin was much higher. It was greater than 2.

Here's one, there was well, up to 5. Remember earlier we talked about the skin variation was easily four fold. Here we are seeing a 5-fold variation which is probably all due to skin variation. So an 5-fold difference in flux because of the skin. And that has been reported over and over again.

We showed the same thing in a publication with Nitrodur where we had a very large difference in flux which was about a 5-fold difference. Jane Shaw from Alza reported a 5-fold difference also. So it's not a surprise.

If we took an average of all of this, we might come up with a better estimate of what happened in the in vivo study.

Q. Based upon your review of the data concerning the Keith matrix, could someone, a person of ordinary skill in the art, who wanted to deliver a hundred micrograms per hour from a Keith matrix?

A. Well I'll use the flux data from the in vivo study, all right. And the easiest way to think about this for me would be that in that study, that 9 centimeter square patch delivered more than the 25 micrograms per hour duragesic. Right? So if I multiplied the 25 by 4 that gives me a hundred. So I multiply the 9 by 4, I come up with 36 centimeter squared. That 36 centimeter squared patch is going to deliver more than a hundred.

Q. And is 36 centimeters squared considered an unreasonable size for a transdermal patch?

A. No. Not at all. we had – we have the duragesic is the best example. The hundred microgram per hour patch is 40 centimeters squared.

Q. And looking back at the graphic that's up on the screen now, about the mean fluxes, the range of fluxes that are reflected there, how do those relate to the fluxes that are described in the '580 patent as you understand it?

A. Well how do they relate? You better repeat that question again.

Q. Maybe I should.

A. I must not have been paying attention or something.

Q. It's probably my fault. I probably wasn't clear. My question was, how do the fluxes and the range of fluxes here from a little over one to just about 5, how does that relate to what is described in the '580 patent concerning the flux of Fentanyl?

A. Well the preferred range is 1 to 5 micrograms per centimeter squared per hour, so all of this falls within that preferred range.

MS. DAVIS: No further questions, Dr. Gonzalez.

THE COURT: okay.

MS. SANCHEZ: Good afternoon, Your Honor.

THE COURT: Good afternoon.

CROSS EXAMINATION BY MS. SANCHEZ:

Q. Good afternoon, Dr. Gonzalez.

A. Good afternoon, Ms. Sanchez.

Q. You were just giving some testimony about the size of patch that you would pick in order to deliver analgetically effective rates. Do you remember that?

A. Right. You know, you're taller than Sharon, and you've got to get that microphone a little closer to you. There.

Q. Is that better?

A. That's much better, yes.

Q. So do you remember that testimony?

A. could you start again? Because I was trying to hear you, and I really didn't hear you very well.

Q. I'm sorry. You just gave some testimony about the size of the patch that you would use – you said you would lower it down to 5 point something; do you remember that testimony?

A. That's right. Yes.

Q. Now you understand that the patch that Dr. Flynn used was 6.5 centimeter squared; correct?

A. Yes.

Q. And you understand that that patch had 72 milligrams of Fentanyl in it; correct?

A. 72 milligrams. That's right. Yes. Okay.

Q. And the patch that was used in the clinical study that you've opined on today was a 9 centimeter squared patch; correct?

A. That's correct.

Q. You understand it had 145.5 milligrams of Fentanyl; correct?

A. That's right. Mr. Diskant drove that point across very well.

Q. So if you were going to reduce the size from 9 centimeters that was used in the clinical study down to the 6.5, would you use proportionally less than the 145 milligrams, the amount of drug put in the new 6.5 centimeter squared patch?

A. I would just cut that same patch down to 5 and-a-half centimeter squared. I would not lower anything. I wouldn't change the formula.

I think that we could take – in all honesty, I think we could take that original Gordy Flynn patch, Dr. Flynn's patch, excuse me, and take that 6.5, and put it into an in vivo study, and we would probably see the same results.

Q. Now if we took that same patch, that patch would have 72 milligrams of Fentanyl; correct?

A. Correct.

Q. And you understand that the only data we have that tests that patch in vivo or in vitro is Dr. Flynn's in vitro study; correct?

A. Dr. Flynn's what?

Q. In vitro study?

A. Oh, yes. Right. On one cadaver skin only. Right.

Q. Now Dr. Gonzalez, earlier in your testimony, you had testified that you worked at Key Pharmaceuticals in the early '80s; do you recall that?

A. Yes. From 1980 on.

Q. And that was when Dr. Alec Keith was there; correct?

A. correct.

Q. And your main transdermal project was on nitroglycerine, I think you said?

A. Correct.

Q. And that project used the Dr. Keith's formulation; correct?

A. Correct.

Q. And I think you testified that the commercial product Nitrodur was based on the Keith formulation; correct?

A. The commercial product you said, yes, it was based on Alec's formulation.

Q. Now Dr. Keith, he wanted to have a large excess of nitroglycerine so you could have a constant infusion, isn't that fair?

A. So you could have a constant infusion, transdermal infusion; correct.

Q. Now to your knowledge, the Keith formulation was never made using Fentanyl; is that right?

A. You know, Alec – you know, Dr. Keith was confined to a wheelchair. He had polio as an adult. So he would – he had a hard time traveling. So he kept a laboratory at – in College Town, Pennsylvania, as well as the labs at Key. So he had a very unique situation in that the university allowed him to keep his faculty slot while he was vice President at Key and all of that, so he did a lot of work at Penn state that I never got to see.

I had no idea every project that he worked on. But at Key I never tested a Fentanyl product, so I can only address what I know. I never saw – I never saw any in vitro flux data on Fentanyl, and I never ran any in vivo studies on Fentanyl.

Q. And that's why I carefully phrased my question to say to your knowledge.

A. okay.

Q. The Keith formulation was never made using Fentanyl, isn't that correct?

A. correct. To my knowledge. Correct.

Q. And you certainly were never involved in working on a Fentanyl transdermal patch at Key?

A. No.

Q. Now earlier in your testimony you stated that you relied on the clinical study authored by Russell Rackley; is that correct?

A. correct.

Q. And Russell Rackley is a Mylan employee, right?

A. Russell Rackley – oh, Russ Rackley is a Mylan employee, yes. I'm sorry.

Q. Now in your review of that clinical study which is this huge document, P.Ex. 269, do you have that in front of you?

A. Right. In fact, it's open here, part of it. Yes. I'm now at the front page of it. Yes.

Q. And you reviewed that study you said; correct?

A. Yes, I did. That's right.

Q. And you didn't find any indication in the entire study that approval was sought or gained from the FDA in order to conduct that study; correct?

A. There would be no need to obtain an FDA approval to run a study in the generic. The requirement from the Office of Generic Drugs is that you not dose a dose higher than the highest approved dose.

Q. And we are going to get there.

A. Okay.

Q. But my question right now is you found no evidence in there that approval was sought or received from the FDA; correct?

A. It wasn't necessary.

Q. So the answer would be no, right?

A. The answer is no, you're right. Yes, ma'am. Yes, ma'am. The answer is no.

Q. Now you'll agree that generally speaking the FDA requires approval for clinical studies?

A. Repeat that one again.

Q. You'll agree with me that generally the FDA requires approval for clinical studies?

A. No. I won't agree with you.

Q. okay. You would agree that there are certain exceptions to when the FDA requires approval for clinical studies; correct?

A. They require an approval for a clinical study for a new chemical entity for a product that is a dose that's – that hasn't been approved or a dose – if you're – if you want to test a tablet of a dose that's higher than anything that's normally dosed, then they require an approval for that.

Q. You understand when determining whether to get FDA approval, one of the things you have to do is weigh the risks and the benefits of the clinical study that you're about to engage in, isn't that fair?

A. I don't think I can answer that question. Do you want to paraphrase that?

Q. Do you want me to rephrase it?

A. Right.

Q. when you are seeking approval for a clinical study, if it's not of an already marketed drug, I understand that in order to make a determination about whether to seek FDA approval you have to weigh the risks and the benefits of that new product?

A. That statement is true. For a drug that hasn't been marketed. Fentanyl was already on the market, so there was no need to get an FDA approval.

Q. Let's stick with – so you agree with me there is a weighing of risks and benefits; correct?

A. For a non marketed drug, yes. You have to submit an investigational new drug application. You have to wait for that application to be reviewed. if you don't hear anything after you file a protocol, in 30 days after you file a protocol to an IND, then you can proceed with the study.

Q. And benefits that you would look at would be scientific and medical; correct?

A. For a new chemical entity, that would be true.

Q. They wouldn't be – a benefit wouldn't be one

for purposes of litigation; correct?

A. Well how can I answer that? If we had a new

chemical entity, and I was seeking approval to run a study for that and I file an IND, no, I doubt that I would get the IND approval based on litigation.

Q. Would you agree with me that you need an IND for a bioequivalent of an approved drug if you significantly alter the risk to benefit ratio?

A. No. No, I can't agree with that. I will tell you why. Because if you're dealing – if you're dealing with running a bioequivalents study for an approved drug, you're going to be running this thing under conditions such that the dose that's going to be delivered to the patients or to – sorry, to the healthy volunteers, those are going to be delivered is going to be with in the range of the approved doses.

Q. And if it's not, then you would have to look at whether the different doses would significantly alter the risk to benefit ratio; correct?

A. If it's not within the dose, I don't care what risk or benefit ratio, if it's not within the dose, the FDA guidance says that you cannot run that study.

Q. I would like to talk about a little bit about the risks that were involved in this study.

A. Okay.

Q. You'll agree that whenever you put a narcotic on a human being that that's a risk?

A. Yes. I think I could agree with that.

Q. And you would agree that there is a risk of putting 145 milligrams in a product on a human being?

A. If I was putting 145 milligrams of drug directly on the skin or directly through an IV infusion or an IM infusion, I would say that's a risk. But in this case it was 145 milligrams mixed in with polymer. And the 145 milligrams was not all available for absorption, so no, I think the risk was minimal.

Q. Well you understand that 145 milligrams is 14 times the highest approved dosage on the market; correct?

A. I understand that.

Q. And are you – would you also consider it a risk if the drug delivered more than what was approved in the prior dosage forms?

A. The 145 milligrams was not delivered, it was only that drug in solution.

Q. I'm not asking you that. I'm just asking you

A. what are you asking?

Q. would you agree that there is a risk, that there could be a risk, depending on how much drug is actually delivered? If that amount of drug is more than what is usually delivered in marketed products, wouldn't that be a risk?

A. It would not be an approvable study. If it was more than what's normally approved, it would not be an approvable study, whether there was a risk or not, you can't run that study as a – under the rules of the office of Generic Drugs.

Q. And on direct you testified a lot about in vitro flux studies; do you remember that?

A. Right.

Q. And you believe that flux studies are used to make predictions about what will happen in humans; correct?

A. I do agree that flux studies can be predictive of what you're going to see in vivo.

Q. It may not always be right, but they are a useful important tool to see – to determine what you're going to see in vivo?

A. Yes. we always rely on cadaver skin flux studies from a number of cadaver samples, prior to running any in vivo studies. In fact, I had to be convinced by the formulators that it was worth investing in a biostudy before we ran it. So therefore I always saw flux under different conditions, and under cadaver conditions.

Q. Companies like Mylan makes these predictions based on in vitro flux studies; correct?

A. There was some coughing.

Q. Companies like Mylan make predictions about what will happen in vivo based on in vitro flux studies; correct?

A. I think that's a fair assessment, yes.

Q. Now you're aware that in this case, Mylan predicted that the Keith matrix could deliver up to 4 times the amount of the duragesic 25 micrograms per hour patch; aren't you?

A. I wasn't familiar with that information, It was presented this afternoon. I hadn't seen that, and I didn't understand stand how Dr. Rackley arrived at those numbers.

Q. Let's take a look. I'm going to show you what's been marked as Plaintiff's Exhibit 539. Is this – I'm so used to using the binders, that I forgot this is different.

THE COURT: That's all right. The only reason I raised my finger is Mr. Lewis was telling you that you should send one up here.

MS. SANCHEZ: I don't know if this has been offered. If it hasn't, I offer it.

THE COURT: it has been offered. Because it was just discussed, just a short while ago. I assume it was offered. If not, it's admitted.

(The document marked Exhibit 539 was admitted into evidence.)

MS. SANCHEZ: Thank you.

BY MS. SANCHEZ:

Q. Now this is a memo from – compiled by Russell Rackley; correct?

A. Yes. I see that.

Q. And Russell Rackley again is the Mylan employee in charge of the clinical study; right?

A. He's a pharmacokineticist at Mylan. Yes.

Q. Now on the last page of the second page, I would like you to turn to that.

A. Can I just finish reading a paragraph here just one second?

Q. Sure.

A. Yeah. Okay. I've read it. I still don't understand how he arrived at his estimates.

Q. If you look at the last page, first you recognize this as individual flux data. Don't you?

A. Reference matrix and then a ratio. And so those numbers above are skin samples or lots?

Q. well do you understand that this is the individual data for the experiments that were reported in the Srini venkateshwaran memo that you testified about on your direct?

A. I see, okay. No. I didn't realize that.

There has been enough time points in here because in that Srini memo, there were more time points than that for the clinical lot. I remember at least four time points.

Q. Let's just –

A. It's confusing me here.

Q. Let's compare the two, so we can make sure we are all on the same page. You also have, I believe, the Srini memo up there?

A. Can you give me the number?

Q. Plaintiff's Exhibit 15.

A. It's plaintiff exhibit. Right? Okay.

Q. Let me draw your attention here. You understand that the reference –

A. Okay. I see now. This average values are what I had seen in the Srini memo, and the 8-hour values are missing from the Rackley summary memo.

Q. And when you say average value, you're talking about the values in this average column?

A. Yeah.

Q. 193.9, 90.5 and 25.6 and those are all reflected in the Srini memo, right?

A. Yes. I agree with you.

Q. So Srini reports the mean; correct?

A. Correct.

Q. And the document P.Ex. 539 is reporting the individual data, right?

A. Yes. So this – are different Franz cells. And I assume that this is – this number is assigned to the setup of the flux study. And I don't know whether it involved different cadaver skins or not. But I will assume that it did.

Q. well if you look at the Srini memo you understand that –

A. It was four donors, that's right.

Q. 4 donors?

A. I've got it.

Q. Go back to the first page of the Russ Rackley document. I want to point out a particular paragraph. He writes a 9-centimeter squared ftm. you understand ftm was the Keith matrix; right?

A. Okay.

Q. Was estimated to deliver increasing amounts of Fentanyl of 230, 584 and 931. As the FTM flux data was more available. You understand that means more variable than duragesic; correct?

A. Okay, before I agree, let me just see where you got those numbers. FTM was estimated to delivering increasing amounts of Fentanyl – where do you get these numbers that are in the paragraph? Where are those numbers in the table? 230, 584, 931.

Q. If you look over here again you have a reference, you have the matrix, and you have the different donor, 1, 2, 3, 4 and the average?

A. I saw that.

Q. Is 230, 585 and 931, which are the numbers we just went over on the other page, right?

A. Hang on a second. Right. Okay. I see them now. individual rate microgram patch area per 24 hours.

THE COURT: Dr., you have to be careful about speaking to yourself because I think it's – you're likely to find it on the record.

THE WITNESS: That was off the record. It was called mumbling. I apologize, Your Honor.

BY MS. SANCHEZ:

Q. Are you clear now where the numbers come from?

A. I see that now, yes.

Q. Let's go back to the first page of the memo. We left off as the FTM flux data was more variable, and you understand that means more variable than duragesic, right?

A. I think he's basing that on the coefficient of variation over here.

Q. whatever he's basing it on, it's more variable than duragesic; correct? That's the comparison?

A. Yes'.

Q. These estimates predict up to approximately 2,300 micrograms of Fentanyl might be delivered over 48 to 72 hours, and he refers to donor skin sample 2076, which is up to 4 times the average for duragesic. This might be considered a worst case scenario, do you see that?

A. I see his writing. Yes.

Q. And would you agree that four times the – four times would be equivalent to a dose you might expect from a hundred microgram per hour patch?

A. Four times the equivalent of that would be equal to a hundred microgram per hour patch, that's correct.

Q. Now you understand –

A. It's amazing that his calculations were so wrong from reality; isn't it? It didn't come anywhere close to the in vivo data.

Q. I'm sorry, shows you can't always rely on in vitro then. Or can you?

A. Well there is – it's always been very good for rank order. And so what this thing is showing me is that it was higher than the duragesic. It was more variable according to his thing, but it was higher than the duragesic, and that came true in vivo. It was higher than that.

Q. Okay.

A. But it wasn't as high as those absolute numbers that he's relying on. Yes.

Q. When you plan your in vivo studies, though, you're relying on the results you got in vitro; correct?

A. Yes.

Q. Now you understand that the hundred microgram per hour dose is a dose that the FDA had already told Mylan that it should not test in humans because of the danger to them?

A. That's true. It said that there was a letter from the FDA saying that they should not use reference duragesic of a hundred or something along those lines.

Q. Well let's just take a look at it to be sure.

A. okay.

Q. I'm going to show you what's been marked and is in evidence as Plaintiff's Exhibit 538.

A. Okay.

Q. Go to the second page. as you pointed out, I think already it's a letter to Mylan –

A. Sure.

Q. – from the FDA. And it's dated May 17, 2000 which is before the Keith matrix clinical study; correct?

A. Yes.

Q. And let's just read this paragraph. According to several box warnings of the duragesic labeling, because serious or life threatening hypo-ventilation could occur, duragesic is contraindicated in doses exceeding 25 miorograms per hour, at the initiation of opioid therapy. And the 50, 70 and a hundred microgram per hour dosages should only be used in patients who are already on and are tolerant to opioid therapy.

In a previous consultation with the FDA medical officer, Dr. Mary Fanning of the OGD, she also agreed that for safety reasons, only the lowest approved dose of Fentanyl transdermal patches should be used in the bioequivalents and skin irritation studies for this product. Therefore, the agency cannot recommend the use of the highest developmental dose of Fentanyl, a hundred micrograms per hour, for the bioequivalents study of the test product as proposed in the current protocol, do you see that?

A. Right.

Q. And as you know, for Mylan's ANDA testing, it did not use the 100 microgram per hour patch; correct?

A. They did not use the hundred microgram per hour duragesic patch, no.

Q. They used the 25 microgram per hour patch, right?

A. Yes, ma'am.

Q. Now' these factors using a narcotic with 145 milligrams of Fentanyl which is 14 times the amount in any approved dosage, and knowing that it might deliver 100 micrograms per hour, are risks that should be considered in deciding whether or not to do a clinical study in humans, isn't that fair?

A. I think it's a fair assessment. Again, if that 145 milligram was all loose and free to be absorbed, I would be very worried about it. But again, I maintain that it was only the drug in solution that was available for absorption. And also looking at the in vitro data, the way I would have looked at it, would be to look at the 72-hour – the 72-hour flux values, and how they compare to duragesic. And I look at them, and yeah, it may be more variable, but they are not that far off. It's 179 that has – 179 micrograms per centimeter scared that has flux at 72 hours. The other is 183.9.

I don't see that it's a hundred microgram per hour patch that they were testing.

Q. You accept that Mylan was concerned that it would deliver four times the amount of duragesic at 25 microgram per hour; correct?

A. I believe that Dr. Rackley was concerned as he expressed in this memo here.

Q. Now I would like to go through some of the other requirements for all clinical studies.

A. Some of the other requirements?

Q. I'm going to get to my question,

A. Are we done with the Rackley memo?

Q. You can put it away.

A. Okay. Can I point out that he suggests using Naltrexone?

THE COURT: Dr., there is not a pending question here. So –

THE WITNESS: okay. Sorry.

BY MS. SANCHEZ:

Q. Now you'll agree that for every study conducted on humans, regardless of whether or not you got FDA approval, there are additional safeguards?

A. Additional safeguards for every study?

Q. Let me give you an example.

A. Okay.

Q. Clinical studies of drugs are conducted by or supervised by a principal investigator; correct?

A. Yes, ma'am.

Q. And any person or entity who conducts clinical work in humans with a narcotic must be licensed to do so by the dea. You'll agree, right?

A. Yes.

Q. And the reason why they are licensed is so the DEA can assure itself of who is receiving these narcotics; correct?

A. Yes. Well the physician has a DEA license number, as I understand it.

Q. And he gets that license number after being – having a background check done by the DEA, and other things, so the DEA knows who in the country is allowed to receive narcotics; correct?

A. Yes.

Q. And in this study Dr. Thomas Clark was the principal investigator of the study, right?

A. As I understand. Yes.

Q. And you understand that Dr. Clark didn't know that the study was being conducted for litigation; correct?

A. I would have believed that before this morning when that letter was introduced. It wasn't introduced, or whatever happened to that letter? Because it appeared that the irb had been notified that it was being done for litigation purposes.

Q. Let me show you –

A. I'm looking here to see.

Q. The testimony of Dr. Clark at his deposition, I'm on page 25. Were you ever aware that this study was to be used in a lawsuit between Mylan and Alza? Answer: Not at the time the study was done.

MS. DAVIS: I object to the use of the deposition testimony. Dr. Clark is not Mylan. And is not impeaching this witness.

THE COURT: Well the question is, is he Mylan, at this point? In which case it's an admission, what technically is the relationship between Dr. Clark and Mylan, at least for purposes of this study?

MS. DAVIS: I do not believe that Dr. Clark would in any way meet any of the requirements for this deposition testimony to be used in this way.

THE COURT: what's the relationship between Dr. Clark and Mylan,; at least in connection with this particular study, is he an employee? Is he an agent? Acting on behalf of Mylan, or is he separate and independent?

MS. SANCHEZ: He says in his deposition he became employed by Mylan on January first, 2001. Up until that point he was a consultant.

He's also the person who Mylan put in charge of conducting this study. And I just want to make – I want to know if Dr. Gonzalez knew that this person –

THE COURT: If you've got a statement in there in which he indicated on the record that he was an employee of Mylan and that he was acting on behalf of Mylan at the time of either the testimony or the clinical test –

MS. SANCHEZ: I can show you. How long

THE COURT: You don't have to show me. Just if you can read a part of the – of the deposition which identifies him as an employee, then I'll ask Ms. Davis if that's correct or not.

MS. SANCHEZ: Page 5 to 6, he states that he's employed by Mylan as a Medical Director, and further on the page on page 6 he says I became employed by Mylan in 01/01/01. Until then I was a consultant, and I've held that title as Medical Director for greater than 25 years.

THE COURT: Okay. Why isn't he acting on behalf of Mylan either as an employee or an agent, in which case it's an admission? And it can be admitted as an admission.

MS. DAVIS: Your Honor, my understanding is that there was heightened requirement for using deposition testimony as hearsay rather than just him being an employee. If the question is, was he an employee when he testified, he was an employee when he testified.

MS. SANCHEZ: Your Honor, I'll also point out these sections were designated, and no objections were, as far as I can remember, asserted to these particular passages.

THE COURT: As far as you can remember.

MS. SANCHEZ: There were very few objections, and I don't have it in front of me so I don't want to misspeak.

THE COURT: First of all, it's not hearsay. It's an admission of a party, so therefore, it's not hearsay. And it can be admitted, so it's certainly – if it can be admitted, it can be used to impeach credibility which is the other ground to justify its use. So –

MS. DAVIS: I didn't understand that it was being used to impeach credibility.

THE COURT: I think that probably my guess is that this particular statement will then be followed up by another question which is meant to impeach the credibility of the witness. At least credibility of what he had said.

Regardless, it's an admission. In light of the fact that he's acknowledged being an employee. So – objection overruled. You can ask the question. Do you remember the question, Dr.?

THE WITNESS: No, I'm totally confused as to which question we are working on right now. She will remember.

THE COURT: Do you have a good grasp of the hearsay rules at this point?

THE WITNESS: Well it's –

MS. SANCHEZ: Let me just short circuit it.

BY MS. SANCHEZ:

Q. Your testimony today is that you weren't aware that Dr. Clark did not know that this study was used for litigation; correct?

A. That's correct.

Q. And you're also not aware whether the other Dr., Dr. Williams, knew that this study was being conducted for litigation; correct?

A. That's correct.

Q. You'll agree as an investigator, Dr. Clark must perform certain tasks; correct?

A. Dr. Clark would what?

Q. Must perform certain tasks. He has certain duties?

A. Oh, sure. Tasks, right.

Q. In conducting clinical studies; correct?

A. Yes.

Q. And I would like to take a look at a couple of those. If you'll take out your Plaintiff's Exhibit 269 and turn to the Bate stamp 4677.

A. We are talking about the study report right now?

Q. The big document.

A. And what page?

Q. Bate stamped 46777.

A. Yes, I'm there.

Q. And you recognize this as the protocol for this study; correct?

A. Yes. I do recognize this.

Q. And let's turn to the signature page of this protocol which is 46787.

A. It's where?

Q. 46787?

A. Okay. Signature page. I'm there.

Q. And you understand that both doctors Clark and Dr. Williams signed off on the protocol; correct?

A. That's correct.

Q. And that's normal procedure for a principal investigator; correct?

A. It's required procedure. Yes.

Q. And they make a statement: I, the undersigned, as the investigator for the study, have read the foregoing protocol and agree to conduct the study as outlined herein and in accordance with all local governmental and United States Food and Drug Administration regulations. Do you see that?

A. I see that.

Q. And any time a study is conducted, a principal investigator must make that assertion; correct?

A. Yes.

Q. Goes on to say applicable guidelines and regulations include, but are not limited to. Permission to allow the sponsor, and that would be Mylan in this case; correct?

A. Yes.

Q. And/or its agent or regulatory agencies, and regulatory agencies would include the FDA; correct?

A. Yes.

Q. To inspect study facilities and pertinent records at reasonable times, do you see that?

A. Yes. I see that.

Q. And that's something a principal investigator must allow every time they do a clinical study in humans; correct?

A. This is very routine, right.

Q. Go to number 3 of the submission. Proposed clinical investigation including the protocol and consent form to a duly constituted IRB for approval, and acquisition of written approval for each prior to study initiation, do you see that?

A. I see that.

Q. And that's something that's the responsibility of the principal investigator for every clinical study; correct?

A. To submit the protocol to the IRB for approval, that is the responsibility of the PI, you're right.

Q. And you'll agree with me, that regardless of whether or not you need FDA approval for any clinical study, you must obtain approval from an IRB; correct?

A. Absolutely.

Q. IRB is an Institutional Review Board; correct?

A. It's an ethics committee for the place that's

doing the study.

Q. It has to have at least 5 members; correct?

A. At least. Yeah.

Q. And those members are people who are qualified through their experience and expertise, to safeguard the rights and welfare of human subjects; correct?

A. And they're varied. Usually involves some clerical people. I'm sorry, clergy. Clergy is what I meant. Not clerical.

Q. Could involve clerical people.

A. Priests or ministers or rabbis is really what I meant to say. And it has to be a balance of males and females on the committee also. Right.

Q. And you understand that Dr. Clark submitted the protocol in this case; correct, to the IRB?

A. It's – definitely he would have. Yes.

Q. Now as part of your review of Plaintiff's 269, you also reviewed the correspondence to and from the irb; correct?

A. What was included in this report, I definitely looked at.

Q. And you mentioned already that there was an additional letter which we learned about this morning; correct?

A. Correct.

Q. You've only just seen this today, right?

A. I'm sorry.

Q. Have you seen the letter?

A. I saw the letter today. Last night it came in by fax when I was at the offices. I didn't see it last night.

Q. You saw it this morning for the first time?

A. I read it this morning for the first time. I saw – I saw this letter last night as it came from the fax machine. Yeah.

MS. SANCHEZ: And I only have one copy, Your Honor. If it's all right, I'll just put it on the screen.

THE COURT: That's fine.

BY MS. SANCHEZ:

Q. And this is the letter?

A. I think I misspoke. I think I might have seen this letter last night, not this morning.

Q. All right.

A. Yeah.

Q. For the record, this is marked Defendant's

Exhibit 1447.

A. I saw this last night. Yeah. I don't think it was ever on the screen this morning. Or if it was, I don't remember, but I think I saw it last night.

THE COURT: Ms. Sanchez, can you actually clear this up?

MS. SANCHEZ: How about if I go through it in pieces until it becomes clear?

THE COURT: Okay.

THE WITNESS: I remember this thing now. But for purposes of lawsuit.

BY MS. SANCHEZ:

Q. You're reading right here; for purposes of lawsuit?

A. Yeah. Yes.

Q. And the rest of the letter simply says enclosed are the protocols, consent forms, and information about the drug for the above referenced studies. The meeting is scheduled for Thursday August 16, 2001 at 5:30; correct?

A. Yeah. My problem is I'm trying to read here as you're talking, and I don't see Fentanyl – it says Fentanyl over there, and then it's got Metaxalone, Propranolol, Hydrochloride. So what's going on here is it concerns this study, and another study, and another study, is that what that is? I know I'm not supposed to ask questions, but I think that must be what this is.

Q. Your guess is as good as mine.

A. It's referring to 3 different protocols.

Q. Now other than this letter, you saw no other correspondence in P.Ex. 269, the clinical study, that refers to the purpose of the study being for litigation; correct?

A. This is the only place where I see that purposes of lawsuit. Correct.

Q. Now –

A. when you deposed me, I didn't think that litigation issue had come up anywhere.

Q. I'm sorry. Could you read back his response?

(The record was read as requested)

BY MS. SANCHEZ:

Q. In your review of the clinical study, you found nothing that disclosed that the Keith matrices contained 145 milligrams of drug per patch, did you?

A. When I reviewed what?

Q. When you reviewed the clinical study, Plaintiff's Exhibit 269, which by the way is also in evidence as Plaintiff's Exhibit 22, when you reviewed that study, you found nothing in there that said that the Keith matrix patches, each one had 145 milligrams of drug; correct?

A. No. No, that's not common to put a

formulation in a patch, a formula – I'm sorry. It's not common to put a formula of a product that you're testing into a protocol.

Q. Is it your testimony that it's not common to put in a protocol the amount of drug that's included in the particular dosage form?

A. I think that the dose that's going to be delivered, it's common to put that in there. In this particular case 145 milligrams is not going to be delivered. There is nothing in here that tells me how much is in the duragesic patch either. The duragesic patch is identified just as 25 micrograms per hour.

Q. So in your experience, all you would see is the 25 micrograms per hour, you wouldn't see the 145 milligrams, is that your testimony?

A. That's my experience with transdermal products.

Q. now in the correspondence with the IRB you saw nothing where Mylan told them their concern that the Keith patch might deliver four times the amount of duragesic, did you?

A. No. I do not recall seeing that. If you're referring to the Rackley calculations, no. I do not recall seeing that.

Q. And you saw nothing in there that informed the IRB or Dr. Clark or Dr. Williams that the FDA had recommended not' using 100 microgram per hour patch in clinical studies; correct?

A. I do not recall seeing any correspondence with IRB regarding FDA letter. No.

Q. And when you say FDA letter, that's the letter we looked at before where the FDA recommended not using 100 micrograms per hour; correct?

A. Yeah. Let me see. Plaintiff's Exhibit 538, FDA correspondence.

Q. Thank you.

A. I'm trying to see when it was received or sent. I guess it was sent by the FDA on May 17, 2000.

Q. That would be before the clinical study that is the subject –

A. Yes. The clinical face of the study, I don't start it in 2001. I don't recollect the exact date, but I remember it was 2001.

Q. Now you understand an additional safeguard is that you must attain informed consent from all of the human subjects that are in the study; correct?

A. That's very standard. Yes.

Q. And informed consent means that you have to provide the subjects with certain information and obtain their consent to participate in the study; correct?

A. Yes.

Q. You' must tell the subjects the purposes of the research; correct?

A. The purpose of the study. correct.

Q. You must disclose any foreseeable risks or discomforts; correct?

A. Yes.

Q. And you must disclose a description of the benefits to the subject or to others which may reasonably be expected from the research; correct?

A. Or if there is no benefit, then you let them know that. Right.

Q. And you'll agree that in this case the patients were not told that the study was for litigation; correct?

A. That the study was what?

Q. You'll agree with me that in the informed consent that all 9 patients signed for the Keith matrix, they were not told that the study was for litigation; correct?

A. I agree with that. All 14 subjects that entered the study. There were 14 that entered the study. 9 completed. All of them signed the consent forms, and the word litigation did not appear in any of the consent forms.

Q. And nor did the consent forms contain a statement that the patch might deliver four times the amount of the duragesic patch; correct?

A. No. They did not contain that.

Q. And they did not contain a statement that FDA recommended not using 100 microgram per hour patch; correct?

A. No. It did not contain that either.

Q. And the informed consent did not indicate that the amount of drug in the dosage form was 14 times the amount that was in an approved product; correct?

A. Again, that would be talking about the formulation, and it's typically not something that you tell the subject.

Q. And it wasn't in this informed consent; correct?

A. It was not in the informed consent. That's right.

Q. Okay. Now I would like to talk a little bit more detailed about the clinical study and the results as shown, okay?

A. Okay.

Q. Before I do that, before I discuss what your opinions about – what the study proves, I would like to discuss with you what it doesn't prove, okay?

A. okay.

Q. Now' you'll agree with me that one way to determine whether a product is analgetically effective is to compare it to a known analgetically effective product; correct?

A. That's the premise of the Office of Generic Drugs. That if you can demonstrate bioavailable equivalents, you do not have to conduct efficacy studies. And that's true for all drug categories, not just for narcotic antagonists.

Q. And in the Keith matrix study, Mylan compared the Keith matrix to duragesic; correct?

A. Yes.

Q. And duragesic is a known analgetically effective product; correct?

A. Yes.

Q. But you'll agree with me that the study demonstrated that the Keith matrix was not bioequivalent to the duragesic patch; correct?

A. The Keith matrix was – produced higher blood levels than the duragesic patch.

Q. You will agree with me that the Keith matrix was not bioequivalent to duragesic; correct?

A. It cannot be bioequivalent if it had – produces higher plasma concentration.

Q. So the answer would be no?

A. That's correct.

THE COURT: The answer is yes.

THE WITNESS: The answer would be yes. I agree with you.

MS. SANCHEZ: I'm glad I have you around.

THE WITNESS: I agree with you. Your statement not a question, was the Keith patch is not bioequivalent to the duragesic, and so I agree with that statement.

BY MS. SANCHEZ:

Q. And you'll agree that, in fact, for the first 24 hours of the delivery, the amount of drug that the Keith matrix delivered was statistically significantly less than duragesic; correct?

A. For the first 24 hours, you're talking now about point by point comparison?

Q. Yes, sir.

A. There was a statistic that was presented by Bill Barr, and he did not explain how this statistic was done, so I would have to say that I can't agree, because I don't understand how he did the stats. It was either a repeated measures analysis, was it simply an – I don't know, a T test, for instance, or did he do a – the two, one sided test between every point. So I don't know that.

Q. Well if you'll turn to page 46429 of the clinical study.

A. 464 –

Q. That's right. This document is very hard –

MS. DAVIS: Ms. Sanchez, where are you?

MS. SANCHEZ: 46429.

THE WITNESS: Okay, I'm getting there.

046429, right.

BY MS. SANCHEZ:

Q. Correct.

A. okay. Way on the front. okay.

Q. Do you see that?

A. okay. So it's a T test.

Q. It's a statistical analysis; correct?

A. Yeah. It's just a plain T test.

Q. what I've done here on the screen is put the hour mark down the side since we can't actually see the numbers. And you'll agree that from hour 4 to hour 24, the numbers are statistically significantly different; correct?

A. On the basis of the T test, yes, that's correct.

Q. And this is a test that was conducted by either Mr. Rackley or somebody under him; correct?

A. Right. It's not a very rigorous test. It's not something that normally is done. And I certainly would not have done it. But yes, it was done.

Q. And you'll agree that for the last 24 hours, the Keith matrix shows significantly more drug than the duragesic patch; correct?

A. Yeah. Up to 72 hours you're saying, right. And then after patch removal, I guess it doesn't really count. But up to 72 hours it did, yes.

Q. And even after patch removal, all except for one point was statistically significantly different; correct?

A. After patch removal, all the way down to 144 hours.

Q. Correct.

A. Yes. So it was statistically higher for the majority of the plasma concentration profile.

Q. Now understand that the Court has construed the claim term of the '580 patent analgetically effective rate; correct?

A. Could you repeat that again, please?

Q. Do you understand that the court has told us what analgetically effective rate will mean in this case?

A. Yes. The patent describes an analgetically effective rate. Is that what you're saying?

Q. You understand it's the Court's job to define what that term means; correct?

A. Okay.

Q. I'm about to start a new topic. If Your Honor wishes –

THE COURT: That's going to take more than 2 minutes?

MS. SANCHEZ: Absolutely.

THE COURT: Okay. All right. So let's stop at this point. That's fine.

THE WITNESS: You don't want to work late?

THE COURT: No. I would certainly go on, but there seems to be a consensus on the Plaintiff's side that we should stop at this point.

No. Actually, I mean a joke. But it's 2 minutes of 5. And I think the next topic is going to be quite lengthy. So let's take our break at this point.