Pharmascience Inc. v. Meda AB et al - Nov 19, 2021 - Part 2

6:21PM Nov 19, 2021

Speakers:

MR. SKYODYN

Keywords:

noc

generic

paper

market

sales

data

case

correct

firms

paragraph

fair

report

page

hollis

pharma

quarter

entrant

ontario

designation

gruden

This is a

test for the live transcription test.

123 examine resume. So yes good at our end. Thank you I'm ready to go recording in progress Thank you.

Yes, thank you justice in that so Dr. Wallace remember exactly where we left off but I think at some point anyways, in my cross examination we were talking about the real world. And you mentioned it a bit in your direct examination. And, and so in terms of the timing of what happened in the real world, Apotex had about a seven quarter Head Start Over farmer science, is that right? Give or take?

Yes, can I actually return since you've mentioned the last question where we stopped? I had a chance to think about your last question and wondered if I could give a more complete answer to it.

Ah

normally the way these things work Dr. Hollis is that I asked the questions you give the answers it may be that Mr. stainsby would go to this and re examination anyways, so I suppose I'll save him the time. And sure, what is it you had to add

your previous question just before we broke for lunch? Was is it possible that Apotex actually began to fill up the essentially began to supply wholesalers in the distribution channel with with Apple zolpidem in the in January 2015. And I guess the piece of information which I didn't look at and consider carefully was the actual volume of sales, which IMS records for APA zolpidem in February. It is it's very small relative to what they saw in March. So I think that the implication of that I mean, just to give a context basically, they sold around 5010 milligram units in Ontario in February, and they were up to 94,000, which is more or less where they stayed by the subsequent month. It seems to suggest that they weren't really in a position in January to have made substantial sales into the pharmacies in February. And it looks more like they actually really started filling up the distribution channels in February and that's why basically so few sales got through to CDH in February.

That's, that's a, you're just trying to extrapolate from the information that you can see on the CDH. Right,

that's correct. So I would hold with my previous answer that it's not possible for me to say that Apotex didn't begin to you know, sell in January of 2015. But certainly it doesn't really look as though they did, certainly not in any substantial way.

I see. In terms of

being able to fix, in your opinion, being able to fix the date of first sale of Apotex to say February 13 as opposed to January 25, or February 5 or February 12. You can provide an opinion to assist just as in and knowing which one of those is more likely. Is that fair?

I think about which one is more likely, is probably not

very, since.

As I said there were so few sales in February compared to the subsequent month. That it looks as though they really only started having significant sales, you know, by late February,

right but I'm just talking about the very very first sale that that first sale so they've got their NOC in December of 2014. Right yeah. And they're doing things in January of 2015. We looked at one of them, but I think, you know, it fair to say that Apotex would be doing other things commercialization wise in the month after it gets its NOC and in your experience, is that fair? Yes, yes. And one of those things might be that very first sale out to a wholesaler, but you don't know is what you're saying?

That's correct. Right.

And so if we go back to the real world well that was the real world actually. So we're still in the real world. But where, where? Real world visa V pharma science as well. So in pharma science, you know, gets its notice of compliance in December of 2016. About seven quarters after Apotex gets its NOC and and makes its first sale, right? Yes. And again, we talked about this right at the beginning, and that's the 20 more or less the 27 2017 2018 time period. It starts a little bit in December of 2016. doesn't run quite to 2018. But it's it's 23 months in better almost all in 2017 2018, right? Yes, yeah. And I had called that the proxy period and you made a clarification as to who that applies to, but it's or as to what portions of your analysis that applies to but it that period was what you used in determining the total XOpen a market Correct.

Sorry, not the proxy period, but the actual period,

yes, the actual period 2015 to 2016. Yes, right. Right. Correct. And so then that sorry, yes, you're quite right. So to the 2017 to 2018 period, is the 23 month period, where Apotex and pharma science are first in direct competition with each other. Yes. And I believe in your testimony this morning. You said that that Apotex that I believe the word you use was dominate dominated the market in that period. Yes. And you said that

I believe that there was something

why they dominated it that much. It could have been something specific to Apotex. It could have been something specific to zolpidem or it could have been a combination of those two things. Is that fair? Yes. And then I think later, you may have said this, but I will add it and ask you if you agree with that. It could have been something in respect of the Apotex farmer science dynamic in this market.

Yeah, I mean, it's hard to know, you know, without more sort of detailed information about exactly what was going on in these circumstances,

right and and I think you said in your in your evidence in chief that or at least suggested it that the the level of dominance for zolpidem that Apotex had over pharma science in that first 23 months of competition was I think you said really quite high. Yes. And you were a little surprised by how high it was, which is why you were trying to figure out why it might be that high.

At least it was higher than we would have expected from our hijab the paper,

right? Yeah, so that's exactly the question I was going to ask you that if you would put the Zopa market into the the iGEM model. And, and and seeing what the projections would have been, Apotex would have had a probably a lower share than it did in the real world and pharma science a higher one. Yes. But that's not what happened in the real world. It was higher than you would have expected. Yes. And at paragraph 30 of your report, which is in Section G, at page eight of your report you say that you relied on factual experience and data from the actuals open a market where possible, correct? Yes. And that

it from a

from the perspective of the reliability of the predictions using data from the real world for the specific molecule we're talking about, is beneficial fair.

I think it makes the analysis more transparent. And yeah, it seems like a reasonable thing to do to use data that is relevant to the specific market wherever it's possible.

Right? And we'll see and get Yes.

And, and we there. I'm going to suggest to you that there are a few and I think you may have dealt with this in your examination in chief but that there are a couple of aspects to this particular market that make it make using the real world data particularly attractive. And the first one I will suggest to you is that as the market the Canadian Zopa market was, in fact, genericized by Apotex within a month of when it would have been genericized in the bud for world so in my I would suggest to you that that's a helpful fact in this case, to make the real world experience even more useful. Is that fair?

Yes. Although to be a little picky, it's more like five weeks but not quite within a month. So yeah.

Well, but to what we were just talking about we don't know exactly when the date the first sale was. So it's, it might be three weeks, it might be four weeks, but it's it might be five weeks, but but the reality is it's not five years later or something like that, like in some other cases. Yeah.

Not five years later, and so I'm just working with the assumptions that I was given.

Yeah, yeah. But you've worked in other in we talked about some of the other cases you've worked in and in those cases, we you were looking at differences in in yours between the but for worlds you were trying to model and the genericized ation that took place in the real world, right. Yes. And here. We're not talking about years. We're talking about weeks. Yes. Yeah. And that's helpful here. That's in part why you were able to resolve issues one and two so easily correct. Yes. And you You used the real world data for both one and two, and you have no dispute for both of your issues. One and two, the size of the total market and the size of the generic portion of the total market. And you don't have any concerns about having done that right. That's right. And Dr. Gluten doors did the same thing.

Yes. So I think the difference between us is very small in those two.

And so I'm going to turn now to your so that that deals with

H and I or at least the yes Chennai moving forward to your Section J

just in case you were concerned I was going to spend half an hour on each section I skip those two and like a minute so we're on to Section J of your report, stage three page 10. Let me know if you're there. On there. Okay, great. And I want to start with if we go over the page, I want to start with what you

what you describe as

in this section, which is the largest section of your of your first report at your estimate one your your

know. And so

if you're over the page on page 11 of your report at paragraph 46. You're the first approach your approach one is the suggestion you're making is to simply flip the position of Apotex in pharmacy science from from the real world into the but for World correct

with appropriate adjustments. Yes.

But the first part of that is that you you generate your estimates one by putting pharmacists in the position that Apotex was in during the proxy period, correct?

Yes, as the essentially suggesting in that case that the prime issue is who's first and and not which companies which?

Right. And of course, we you'll agree that Apotex and pharmacists are two different companies Correct.

I feel I'm obliged to agree with you there.

You can agree even if you don't feel obliged, even if it's just correct. But yes, okay. And and you'll also agree that Apotex is a larger company than pharma science. You use that in your report? Yes. And Apotex has been the the largest Generic Pharmaceutical company in Canada for quite some time. Yes, yes. They were the number one generic in December 2014. When they got their NOC for their results in a product. Yes, I believe so. And and through the 2015 to 2018 time period, Apotex is still the number one generic in the Canadian market.

You know what I haven't actually seen the date on that. But I mean, quickly, Tebow is getting closer after the introduction

of but you so then you'll accept though that whether Apotex was number one or number two, they throughout the relevant time period of 2015 to 2018. Apotex is substantially larger than pharma science, correct.

That's my belief. Yeah.

And you incorporated that to some degree in your opinion, the difference in size, correct, yes.

And so if we go forward through, at your opinion, your report to paragraph 48 That's what you're saying in the first sentence of paragraph 48 of your report is that Apotex was a larger firm than pharma science and might have benefited because of its size, correct? Yes. And you say in the next sentence, and you're referring here to your iGEM paper, I think that larger firms systematically obtain larger shares of generic sales. Yes, right. And you talked about that with Mr. stainsby this morning that this idea that sales and other markets meant that larger firms entering into a new market with smaller firms tended to get more market share. And that's both intuitive and actually what you were able to determine from your more systematic approach in the iGEM paper, correct?

Yes, I mean, having said that, there actually some issues with that and one of the things is that some small generic firms actually only sell into Quebec men, so they have naturally a limited size in terms of what we would have measured in that paper and also therefore, just by construction, a smaller possible share of the market. They get 0% outside of Quebec. We didn't try to account for that in the paper. And I don't know how important it is. I might somewhat upward bias the estimate of the size of what the impact of the firms the generic firm size is on its ability to capture generic share in any

given market. So if I understand what you mean, it might affect the percentage that you would apply if you were applying the IgM model to a new market, but it will affect the general on

a larger share. Is that right? That's fair.

And you talked a bit about differences that you had with Dr. Gruden doors of opinion. And I was happy to hear you describe Dr. Gnosis as a friend that's that's a plus. It's good to have friends on both sides sometimes.

And just so I understand. You don't

see his approach of using pharma science to model pharma science. As inherently improper, correct.

Not as inherently improper, I think, but I I find it problematic because it imposes so much of the zolpidem outcome on what happened in the what would have happened in the but for world so we know that, as I said earlier in my testimony and in chief, I'm in the actual zolpidem market, the first generic entrant actually really dominated the market and we don't know the reason for why

that was. Well, no, what you actually said was that and it's maybe Symantec, but I don't think it's Symantec is that in the real world. Apotex dominated the generic market and yes, they were also the first entrant but they're still Apotex, right? Yes. And so what I'm suggesting to you is that it wasn't inherently improper for Dr. Gruden doors to keep pharma science as pharma science and increase their share as opposed to it can considering that the company stays the same focusing on that aspect as opposed to focusing on who was first or second.

Yeah, I could imagine. I mean, it's a it's a, it's an approach to take you to addressing this and in some ways, it's really not very different from my method one. Essentially taking the data from the actual market and then adjusting my case either for firm identity plus plus to incumbency effect or and document

for entry timing right. So

similar and so if I were to say that what he did was improper, I'd have to say that

what I did was improper too. And you're not saying either of those things just so we're clear. I'm not saying either of those things. Okay, great.

So let let's, let's go to the iGEM paper for a second. It should be you should have it. I mean, obviously you have it, but it's also attached to your report at exhibit three

jet paper.

If you have the documents by FC number, I think it's FC 38.

Thank you there,

okay. And was that I have that in my notes. So it was that right is FC 38 the Newton's first law paper? Yes. Okay, good. And we've been calling it the IgM paper and you're fine with that.

I'm propose that as the name of yours, so yes, perfect.

And this was a new and Dr. Guten Dorst and a PhD student. of yours at USC. Correct? That's right.

I'm allowed to call it u of c being an alumnus of that August institution. So

now I see why we're getting along so well.

I was looking at I left just before you join the university and I spent most of my time there in the basement of science BS so I don't think we would have run into each other but and I for your sake, I hope you've never been to the basement of science B. It's not a wasn't done anyways, a particular lovely part of the university anyways, so the idea paper explores determinants of generic market share in the Canadian market, is that right? Yes. And two of the determinants that you examined in the IGF paper were entry timing. And the size of the generic manufacturers that right?

Yes, those are two. And if we look at

that and we were just talking about this, but you address it a couple of times in the paper at at the bottom of page two a one of the paper which is the very first page of the IGF paper, very last sentence at the bottom of the first page and over to the top of the second page of the article, which is page 202 of the journal you you and your co authors refer to the fact that firms are referred to you make the statement that firms with larger product portfolios can reduce pharmacies transaction costs, by offering one stop shopping and can offer bundle discounts to drive volume. You see that? Yes. And that's a that's a fair statement. You still believe that to be the case? Yes. And provide another explanation of or potential explanation for the manufacturer or size determinant over at page 215. of the article.

We'll go forward to that.

Thank you, I'm there.

And I'm

at page 215.

That's sort of below table five. Look below the bottom of Table five. You have a sentence there that starts actually at the bottom of page 213. But the relevant portion is the latter part that says manufacture manufacturers size continues to be important in determining market share even after controlling for entry order. This is not surprising we expect that large firms will be able to exploit their position and existing relationships to increase sales, right? Yes. And that's a knot not only something again, that's intuitively appealing, but it's something that

just sits in the paper.

That's why we make that statement.

So I That's correct, then

yes, it's based on the analysis in the paper, right. Or to be more precise, it's an inference that we draw based on the statistical analysis that we do in the paper.

Right.

But, yes, but the as I understood one of the purposes of

sorry, Mr. stainsby. Are you trying to make an objection? Sorry, you're muted.

We can't hear you. Mr. stainsby.

You're muted. Just Oh, he

says he knows he can't unmute himself. Perhaps.

I'm not speaking are you doing just as in can you hear me? This is Jonathan drawl the speaking? Yes, I

can hear you.

I just got an email from Mr. stainsby that he's his computer is frozen. He just needs a break. Maybe he can just restart we take minutes. Yes, certainly can do that. Okay, thank you.

Well, that explains it. How long do you wish should we should we break for 215 or something just as soon as that makes sense.

Obviously, this shouldn't take more than about five minutes I wouldn't think but it

will be back at 212 then.

It is adjourned until Thank you recording stopped Hear me again, Mr. stainsby. Can see you though I think we're good. Perfect next month. Sorry about that.

I did I don't know what happened but I could do anything

happens recording in progress

okay, you're fully with is Mr. stainsby

Yeah, I believe so. If Mr. Scotto would be so kind as to orient me to where he is in this paper so I can catch up.

Yeah, we were on page 215. But I'm going to be leaving page 215 in a second.

And, in

fact, where I want to go now.

Dr. Hollis is too.

203 of the IGF paper

the third should be the third page of the PDF. Hopefully.

I'm there. And

in that first full paragraph on on page two or three of the IGF paper your pricing and and rebates and and the fact that there's gonna be lots of competition, but that hasn't resulted in lower prices. You see that

and or just the

little bit past the second half of that first full paragraph on page 203. There's a line that starts with the word compete. It's about over 13 lines down the sentence I'm going to but that's the line that it starts at the word compete. It's above the 63%. Yes. It's that it's the sentence that starts in the middle of the compete line on page 203 That I to look at and in that sentence you say for many years. Provinces set the cap at 63% of the brand price and by that you mean the cap for what the generic could charge for its product, correct? Yes. And then in brackets 20% of the brand price if there was just one generic on the market, and that that's in the past, but that you're you're making the point there that that was in the past. I'm reading that correctly. And then you go on to say since these prices were well above generic firm's marginal costs rebates were substantial. According to the Bureau 2007 report, rebates averaged in the range of 40% of the invoice price. So there was significant variation with some high volume products having rebates more than 80% of the list. Price. See that? Yes.

And that

that information you took from a Competition Bureau report in 2007. Is that right?

That's what it says. That given that density, we can assume that that's right.

I believe so. Yeah. Fair enough.

So if we, if I understand what what information you're providing there, it's that there was a cap at one point of 63 to 70% of the brand price. For the generic and that gave generics enough room to give rebates averaging 40% of the invoice price. Is that right? Yes. And so did were you provided with any pricing information for the zolpidem products in this litigation.

Aside from the IMS data, which shows both volumes and revenues, no.

And did you calculate out what what the what the prices were at any point in your analysis?

I'm afraid that I didn't

Yeah, I didn't see it and I'm not I'm not blaming you but the we can we pull up a an FC document. I don't think the prices are in dispute here.

Sorry. Maybe I should just correct my testimony. If I could. I'm just not 100% sure that the data that I was provided had revenue data normally it does.

So now we've got an issue it appears that some of my colleagues have vanished from the ball there's still here but

I'm a I'm a I'm a slow convert to zoom trials but it's starting to happen appeal. Why do we need some time for your team to get back up there?

They're all back on Sunday. They all came I don't know how it's possible that they all got kicked out and I didn't given we're all on the same Wi Fi but our network. Anyways, they're back so nevermind. Okay, so what what I was going to ask you about Dr. Hollis whether you were aware that that the thing in the context of Sablon ox versus the ginger put in prices whether you were aware that the generic is open and prices were at 87% of the brand price?

No, but I don't find it surprising either.

Okay, and so then at 87% of the brand price, if at 63 to 70%. There was room for lots of rebates, you'd agree that 87% There'd be room for lots of rebates as well is that fair?

So in the general case, that we described paper there's lots of room for rebates, I don't know the costs of production of this. So in the absence of that information, I can't really agree with you that there's lots of room for rebates for this product. I don't know.

Okay, fair enough.

Right because let me just see if I understand the basis for that. I hope I do. You have a price. That is the the price the innovator has been selling their product for for however, many years they've been on. Or it's the price at which they're selling at the time the generic comes on anyways, maybe they've changed over time, but there's a price for the product when the when the when the market gets genericized right. And then the nature of generic products is that their product is at less expensive by some percentage. Yes, the list price times the brand will lower its price to match but right but didn't mean to interrupt you go ahead.

Yeah, so sometimes the brand will lower its price to match so it's not necessarily the case that the generic price will be lower.

That's a rule up.

Right, but in the scenario you're talking about, both of them have lowered their price from the price it was at at the moment of generalization. Yes. And so what I'm saying is we start with the the innovator price at the moment of genericized ation, and then the generic price is almost always lower in this case. We see it is in fact lower by some percentage. Yes. And then at the bottom end, we have the so we've got a generic price. Now that's some some amount lower than the innovator price. And then at the bottom, we have the cost of production. That's an expense. Yes. And so the delta between the generic price and the cost of production is composed of profit and rebates.

Yeah, I mean, maybe not necessarily profit Exactly. Since you have to pay for the costs of running the company. But yeah, in economics, we call them quasi rents. So perhaps a new word for enrichment.

Oh, I understand. I understand what you mean. And I My point was that the high that that all other things being equal, the higher the percentage, the higher the generic price, the more room they're going to have for rebates.

Yes, although generally, the higher the price the higher the costs as well, or at least if you have high costs of production, you're typically going to have a higher price. So yeah, holding costs constant. And I'm sure that I agree with you that if you have a higher generic price that there is more room, potentially to pay rebates right.

And so the looking at your IgM paper at your exhibit three, as I understand one of the one of the objectives of the paper was to update an earlier paper that you had written, yes. And the purpose of both of those papers was to collect as much data about what actually happened in the chimeric Canadian generic market, and then test certain hypotheses about why things were happening the way they were.

Yeah, I mean, the academic publication processes may be a bit more, a bit less. Ideally, you know, collecting all the information that we possibly could and often more about, what would be an interesting publication that we could get published.

And this one's a bit of both.

Well, I mean, we have the data that we had. So I mean, I we tried to add some interesting questions here about what were the what were the reasons for the success of first generics?

Right. But I

Yeah, and I haven't seen another paper certainly on the Canadian pharmaceutical market that has more data that is more systematic than this one.

Are you aware of one

and the Competition Bureau report from 2007 just had access to more data than we do.

Right. That's a bit older, though. Than the data in yours. I mean, the data in years one up to 2011. Right. So yes,

yeah. But

just to go back for a second part of the purpose of writing the paper from a from an academic standpoint, not just trying to get things published was to, to put some systematic rigor to looking at the questions you looked at in this paper, right? Yes, yeah. So not just gut feelings or anecdotal experiences of one market or another but collect a set of data about a series of markets and then create a, an actual regression model that might be able to explain behavior,

rather than to interpret about what was happening right now.

And is that it? Briefly just for a second at paragraph 53 of your first report.

Then that section we were we were in Section J paragraph

30. You're first looking at it

and not going to the first couple of sets to talk about. You talk about something we've already talked about, which is not really understanding why Apotex had such a commanding share. But it's the third sentence I'm interested in there, where you say typical outcome would be the one that is based on a larger set of data. Do you see that? Yes. And by the larger set of data, you're referring to the fact that the IGN paper looks at 3740 markets, is that right? Whatever it is, yes. Yeah.

And so it's not

it using that model is helpful because it by looking at so many model markets and and analyzing them systematically, you're surrounding off the edges of the anecdotal experiences of one drug in one market versus one another. Drug in another one. Is that fair?

I think that is fair.

And that's why you that that that rigor is in part why you used it as the source of the IGF paper as the source of your second sentence. Is that right?

That's correct.

The first set of minutes based on the Zopa market at the relevant time, a second with adjustments and the second set of estimates in putting that information into your your model from IGN.

Yes, using the relevant timing and market size effects into the idea

paper. And now if we go digest paper at

page 207 of that paper and there there's a heading data. Yes. And you spoke about it briefly with Mr. stains with this morning but I wanted to take you to it because I think you remember exactly but here you say we obtained correlate dollar dollar and unit sales data from IMS Health Canada. From the start of the first quarter of 2001. Two the end of the first quarter of 2011. Had first generic entry between 2003 and 2007. Correct yes.

And so if we go.

I lost my place so if we look just above the heading data. Yes, you in the second half of the paragraph just above data on page 207. The IgM paper is a couple of sentences that start with to test this. We divided our sample of drugs into three types, drugs for chronic conditions, drugs for acute conditions and antidepressants and antipsychotics. Those are the three types.

Yes.

And if we look at table one, which is on the next page of the IGF paper at page 208 And we look at the bottom row of table one on page two. Wait, we see that there were 37 markets that you looked at in this paper. Is that right? Yes. And of the 37 markets, if we total up the numbers above it appears that there were 11 markets where there were two or more first entrants. That right 37 minus 26. Yes. That is that if we can all do I think even on the fly 37 minus 26. And add to our more first entrance. Meaning two or more firms by the by the preferred definition, meaning two or more firms that received their notice of compliance in the same month, correct? Yes. And the 26 markets where there was a not more than one first entrant where markets were the the first company to receive its NOC and the second company to generics. The first generic to receive it's no see and the second generic to receive it's no see. We're not in the same month. Yes, right. And you explain that in the paper. We go back one page to the bottom of 207. So we're right in the same part of this paper, but information straddles this. Underneath the data section we were just looking at is a section titled entry timing indicators. Do you see that on 207? Yes, and so here you say? We utilized two methods to determine entry timing. Our preferred measure is the date that a product was granted a notice of compliance and see, which indicates Health Canada's market authorization, correct? Yes, that's what it says. And that's in fact how you generated the models. Yeah, by using that measure, yes. And then the next sentence says, in this case, all products that and I think we just talked about this in this case, all products that received a NOC in the same month are considered as having entered at the same time, right? Yes. And so that's the

marks we were just looking at. Yes.

And then you you go on to say since NLCS for authorized generics, are sometimes granted long before the product ever comes to market. We also used an alternative measure of entry timing. This measure uses the first quarter with positive sales as the time of entry for a product that's correct. Yes. And then you say our results are robust to this alternative measure of entry timing, correct? Yes. And by robust to the alternative measure of entry timing, you mean that you you ran a second set of analyses and the results were consistent with the preferred measure?

That's right, at least directionally they were showing roughly this similar results,

right. And you

think this same concept comes up a couple other times in the paper?

If we go over to page 210 of the job,

under the heading the evolution of market shares, from there you're describing figure one at the beginning of that section. But I'm curious about the parenthetical in the middle of that paragraph. And where it says

recall that quote, unquote, time

is measured from the quarter of second entry, except for first entrants, in which case, time starts in the first quarter with at least two generic products in the market. And so here's another variable where the quarter of sales is and have some significance, is that right?

Yes.

And in part Y, in terms of the sales and related issues, you're referring repeatedly in the paper to it by quarter is because that's all you had was quarterly data right? On sale so that's correct. On yes on say so. So for no C's, we looked at, you know, see data, you can just go on the internet and find that when companies got there no C's going back decades, right? Yes. And so you you are perhaps your your grad student at the time somebody could have gone to or perhaps did go to the NOC database and and look up when all the NOC is for granted. Is that right? Well,

it's so yes, yeah.

For sales data, especially across 40 markets, for what in some cases were many, many generics, you relied on the IMS data that you were provided. Yes, and that data was a decade's worth. But only by quarter.

Yes

and so.

So if we go back to page 207, and that entry timing indicators paragraph

at least we're in the document. We're not jumping around too much. It's only

no I appreciate it. And I appreciate your giving me the time to orient myself and find the location.

Thank you. Yeah, no, no problem. And so we're at the bottom of 207 100 entry timing indicators. And so given that Apotex sees its NOC in December of 2014, and pharma science in the bud for World receives its NOC in January of 2015. Were you to have had these molecules in your list of molecules at when when doing the regression analysis for this paper. Apotex would have been the first entrant and pharma science would have been the second entrant Correct.

Indeed, we would have missed categorize the sort of relative entry timing. Well,

and I'm going to quibble with you there Dr. Hollis. I would say that my understanding of these sorts of scientific endeavors is that one generates a hypothesis and a set of characteristics a Priore, ie in advance and then applies them objectively and that here applying objectively, Farmer science and Apotex Apotex would be the first entrant in pharma science would have been the second set, right.

If we were to have basically taken the

data that we had

for zolpidem and then transformed it into quarterly sales data and obtain the NOC dates and then put it into this paper, then that's exactly what we would have done. That wouldn't really have been consistent with the goal of my investigation in the report, which is to determine what this paper how this paper informs the hypothetical world in which I'm assuming that farmer signs actually does enter first. And you'll notice in this entry timing and the indicators discussion, midway through that paragraph at the bottom of page 207. We note that there's the possibility of MIS categorization of NOC s. And we we recognize that problem so we never came at this problem and saying, you know, NOC date is what really matters. Clearly what we're interested in was when does the firm actually begin to do sell into the market and to compete?

Well, that may be doctor but what you had was NOC date. And quarterly sales data right. In the paper, yes. Yeah. When when generating the models that result in the percentages and the regressions that you have in the paper, right? Yes. And those those percentages, and and one of the models is what you applied to that the circumstances here, right.

Yeah, I mean, though, if you're to sort of, take me to this, the, the challenges that if some NOC dates are miscategorized, for example, suppose that in the data that we add, sometimes the first generic entrant that we categorize as a first year attention wasn't really first at all. Right, then we thought that we would have ended up with is that we would have underestimated the impact of being the first generic into the market. And you can see why that's the case if just stick with me, suppose there's a normally a whatever 30% effect from being the first generic and then we miscategorized some of them. As it happens, then what's going to happen is that that will result in firms that we think are first generics actually ending up with a low market share, and firms that appear to be second generics ending up with a high market share. When we run that regression, we're going to end up basically under estimating the effect of being first because we've mis categorized some of those second generics as first ones. And so if anything, right, the, the possibility of MIS categorization which could have arisen in our original analysis, in which this case actually highlights as the possibility just would result in are providing too low an estimate of what the impact of first generic entry

well, but that only holds Dr. Hollis if it's true that one generic launching its product, say, one month ahead of another generic but within the same quarter is going to achieve the so called first entrant effect. But that that is a conclusion I put to you can't draw from this paper because you didn't have that data.

On the contrary, we had NOC data which in general is going to be reasonably well correlated with the entry timing.

Right, but But my point is, you don't know. Because you didn't have the monthly data. Whether the order of the NLCS match the order of the date of for sale. Correct.

For this Yeah, I think that's actually fair that we did not have the ability to go back and check at least within the three month categories. Certainly it would be true across quarters. For example, suppose the firm entered within NOC in March and then another firm entered in April then we'd actually get different quarters in that data and different NOC months. So for at least a third of the time, we should be able to distinguish when there's a a one month difference difference.

But in a circumstance like what we have in this case, where the NLCS are a month apart, but the data for sale we I think we can all agree is sometime in the same

quarter the first quarter of 54 world

you wouldn't have been able to distinguish between the two companies correct.

Now if we use this data, we would have missed categorize them. So yeah, we, I think right because we would have listed Apotex as entering in the previous month.

Well, what no I so I'm going to take issue with your use of the word miscategorized. I would suggest that if the definition of first entrant is based on the date of the month of the NOC and that's what your preferred definition was in the paper right month of NOC that was what we had to work with. So so in the IGF paper and for the models that are in the IGF paper, the preferred determinant was for what you would call a first entrant for the purposes of the paper was month of NOC Yes, so that might not match what you think of as a first entrant outside this paper, but for the purposes of this paper, and these analyses, that's what you did. Right.

I mean, having said that, we recognized right that NLCS do not correspond necessarily with sales which is why we had to make this note about authorized generics as well as looking at the first quarter with positive sales. So I think we recognize that the the idea is does the firm actually arrive on the market and begin selling product? And perhaps, that's not explicitly said, although I think we get pretty close with recognizing that no sees may not in fact, relate to market entry in a effective

way. Well, to be fair, but you're talking about in whether the but I understood from the entry timing indicators paragraph at the end of 207. Dairy, or corroborative analysis being done at the quarterly data was with respect to authorized generics that are granted long before the product comes to market.

Yep, that's what that's referring to.

So in that sense, we're talking about not a minority, few weeks, but months and months that the innovator is getting ready. Well in advance.

Those are the ones that we were able to identify, right,

because you couldn't identify within the same quarter in terms of when the data for sale was right.

That's correct. We didn't have access to that data, although in many cases, it might run across quarters. So I mean, as a kind of an exercise here. If if the hypothetical entry dates, instead of being January in February, had been March and April, we wouldn't be having this conversation because none of it would make sense in this context. So the only way that we can have this conversation is because it happens that those two months are in the same quarter, but they aren't necessarily generally going to be in the same quarter.

Well, so I'm going to quibble with you there to Dr. Hollis. I would suggest that that so I'm going to suggest to you again that by if we strictly approach apply the preferred measure in the IGF paper to the facts of this case. Apotex is the first entrant and pharma Sciences The second entrant, yes. Yeah. But and you'll see from Dr. Gruden doors opinion that he looked at that and looked at the fact that the two were December 31 and January 2 at and I think, hopefully you will agree that it's reasonable to treat those two dates, more or less the same given when they happen in the year

I think those those two dates are. Yes, reasonably close. So much so that especially given the difficulty of actually making substantial sales on January the first that you wouldn't think would really matter.

Yeah, or even having people in the office on January. First to get the fat I think in those days Health, Health Canada stuff and stuff. So you know that it may well be that an LLC that arises in December 31 sits on the fax machine until January 2 Anyways. Right?

Yeah. I mean, I've heard about other cases in which Apotex actually had its, you know, trucks ready to go. So I don't know. I mean, I think the the point here is that we're really if they, if you can make sales by merely having an NOC and if you can gain market share by clearly having an NOC but no supply then that would be the best possible case for a generic company because they have no cost of manufacturing. In the in the real world if you actually want to enter a market, you have to have product. Right? That's That's why you know, when I'm thinking about brokers, first, I'm thinking about the firm that's actually able to ship product to, to the distributors and to the to the retailers and hospitals. And here the information that we can get from the IgM paper really relates to who is it they say, you know, what, and not the effect of barely having an NOC even though that was the best that we had the best proxy was able to enter the market, but it's just a proxy. It's not about whether you can make sales and begin to capture market share.

Yet, but

but we're not in a position here. You'll agree with me where Apotex Scott and NOC but had no product it had product whether you want to say that the product was available in in January of 2018, or at the end of February of 2015. It's not a situation where they got an NOC and it sat for months and months with no product, right.

So the the assumption I've been given is that they didn't start shipping until February the 13th.

That's still only a matter of a few weeks. Plus, and it's still in the same quarter Correct? Yeah, I can't disagree with it. Right. And just to go back to, to the paper, the models that you the regression models that you generated, were based on the definitions in the paper, correct? Yes. And while you might have wanted to be able to distinguish at between companies that got there no C's in a, you know, company A and company to launch in reverse order. If it happened within the same quarter, you couldn't distinguish between them, right?

That's right. I mean, it could have happened within the same three month period and we would have been able to distinguish but not necessarily within the same calendar quarter.

Correct. As well and even no sees within the same month. You know, one company A gets their NOC on March 1 and company B gets their NOC on March 31. And by by this paper, they got their NOC in the same month, so they would be considered dual first entrance, even though that NOC those indices were 30 days

apart. Yes, yes.

And similarly, march 31, and April 1, and their one would be considered to be first entrance and the second would be considered to be second entrance, right. Yes. And the purpose, I had understood that one of the purposes of looking at this across 40 markets was as I said earlier, to kind of smooth out the edges of these things. Sometimes they would go in one direction, sometimes they would go in the other but if you do a systematic review, hopefully it comes out in the wash.

Exactly. And on average firms that get their NOC is earlier than other firms enter before those other firms.

Well, that's not the vision you're able to draw from this paper. That's just a statement in general.

That's a statement I'm making in general. And

as I understand it, it's a statement you're making in general that is why symbols you to apply a different measure of timing which is date of for sale, to the models in the IGF paper than the definitions that are actually in it. That's fair enough.

Let's just garden if this is a good time for a break, I'm in need of a cup of coffee.

Again,

I won't take that personally for everyone. salutely happy to take a break now. Thank you.

Let's take 15 minutes. Thank you thank

gerund until 307308 Recording Recording is resumed

Thank you very much. Okay,

hello, fair that that break was fortuitously I'm going to move to a completely different topic at this point. And it's still in your report. And not that far from where we're we're at paragraph 49 of your first report. Dr. Hollis. That's at 12 Yes. And then great. Thank you. And so this is where you have begin a discussion of one of the adjustments that you made as a result of noting a supply shortage of April zolpidem, do you see that? Yes. And you know, that there was a supply shortage of April Zopa them in the months February to April 2017. And you attach a document substantiating that right. And then you indicate that there's an effect in May of 2017 in the iqvia data, yes. And so in your, if I understand that correctly, your conclusion there or your connection that you're drawing is that one month after the supply shortage starts is when the effect of that supply shortage was seeing in the iqvia data?

At least one month after it's being reported?

Yes. I see. Right. To Health Canada. Yes.

And, and so you then logged on below to, as I believe you said, interpolate the data to adjust for that. Yes. And were you aware that pharma science had a reported shortage for some of its PMS is OpenAM tablets during the proxy period?

So know what time

Avila we can we can pull it up? Yep, we will pull up FC 294.

And you see I think at the bottom, there's a

URL

WW drug shortages. candidate.ca is that the website for which the document that you attach to your report? It is and that that's a website that you're familiar with? Yes. And so if we, if we then same part of the page, just draw our eyes up three to five rows there from there's an actual start date, estimated end date and actual end date. Do you see those? Yes. And and the way this information is reported to Health Canada, as the as I understand it, and help me with this, if I'm misunder stood is that the manufacturers have to report when a shortage is happening tell Health Canada when they think it will be over and then update it as to when it is actually over? Is that right?

The last few charts mean I'm not certain about but I believe that is

and so that in terms of actual study, made an end date and actual end date are all between Well, the start date of this shortage, which if you look up on the page are the five milligram PMS oppo T product. The actual start date is in February of 2018. And the actual end date appears to be in May of 2018. Set right? Yes. And do you have any reason to doubt the information in this? No. And so, but you weren't aware of this shortage? Is this information wasn't provided, preparing your report? No. And so in your view, should this be taken into account in terms of calculating farmer sciences? Estimated losses the fact that they had a shortage during the first 23 months of their sales?

Yes, I probably also look to see whether there was a corresponding dip in their sales as recorded by IMS.

Right Well, in this case actually we even have their their ex factory sale. So you, but your point being that you'd want to look to see if there was an impact on sales, period, whatever the source of that information fair. Yes. Yeah. My point being you looked at the iqvia data for Apotex because that's all you had here. You have both the IQ via data and the X factory. So performance science, right? Yes, yeah. But if you saw some kind of impact, given that there was a shortage in the first 23 months of pharma sciences sales you if you saw an impact, you would want to incorporate that into your analysis is that fair? That would be reasonable.

Yes. Okay.

Can we make that in an exhibit please that document and

FC two nine

Yes, we can is that the one that was earlier for identification?

I believe it may have been.

Okay, so I think that's exhibit 128. Is that correct?

I think the earlier one might have been the Apotex shortage. I'm not sure though.

Or perhaps maybe 128. FC 294 Is the exhibit to be identified it later.

Yeah, they Apotex. One is in doctor offices. Reports. Okay. Just on there.

But 128.

Now, I want you to an issue you addressed in in your second report, revision report. We did go there earlier. I don't know if you still have it. Right Okay great.

Okay, let me know if I'm ready. Okay. Great. So that if any, any any evaluating the effect of the OSI designation, in, in this case, you conducted an analysis of data from relating to, to to other drugs at the 2010 time period. So that right

and you can look at 50s

I'm looking, I'm looking at that. Yes.

Okay. And paragraph. If you're at that page, you say, you requested counsel to provide OSI dates for three comparable drugs that success identified that you identified, which are granted RFI status and how did you identify these drugs as comparable?

Because they were all Oh, if I listed drugs, so I want to drugs that we're going to have in Ontario only off formulary interchangeability rather than being listed on the formulary.

Okay, but there are I mean, we look at the I looked at the formulae, there are other drugs that have OSI designation. That's true.

And in this case, I think it might be fair to say that it's lenient sample I have used these drugs before. Oh, like it was in pre previous

analysis for a different different case you mean? Yes. So you are we're already aware that the mantiene and finasteride had a phi designation? Yes. And as members all right, I would end up using Ultimately though, yes, yeah. For this case. I mean, yeah, right. Yeah, but you're right. But there are there are other that have been granted OSI designation that other times that is still leaving now. That's a thing, right?

It's still a thing. Right?

So you could have it wasn't that there was something about mantiene or finasteride that were that were particularly analogous to zolpidem or there's open a market.

I think that's fair. Okay. And so

we turn over the page in your revision report for to page five. Yeah, and it looks to me as though the the math part of the analysis here is that paragraphs 19 to 21 Yes. And, and here as I understand it, you compare first the Ontario market, month to month on either side, by designation? Yes, yes. And then you compare that increase to the that increase, if any seen in the rest of Canada? Yes. And for finasteride, the increase in Ontario only was 58%. And the increase in the rest of Canada was 10%. Yes, and so you subtract 10 from 58 You get the 48% increase for that you attribute to Ontario. You've explained it pretty well math you know it's this math anyways, not so bad. And then you do the same thing for mantiene increase in anti seen in Ontario month to month and then compare that to rest of

Canada. It

that that is a similar process was what Dr. Group doors did with respect to the zolpidem data Correct? He looked at it month to month and then Ontario versus rest of Canada. Correct.

He's actually comparing the the generic market share in zolpidem in Ontario, to the generic market share in other provinces in each month. But in those other provinces there's no nation

right but there's no bonafide designation in the other provinces for finasteride or romantic either is there?

That's right, the approach that the benefit that doing this, the thing that it buys you is that you can actually see what happens when Oh foi designation happens to a drug. So you're I'm tracing across Ontario and then I'm controlling for the essentially the rate of growth which is happening anyway in other provinces. So there I'm able to basically identify the the RFI effect in Ontario for these two drugs. And then I use the average for zolpidem. The way that Dr Gruden does does in contrast is to say in Ontario, here was the average generic market share for zolpidem. And then I'm going to compare their provinces you know, various number of ways and some which simply don't have off formulary interchangeability as a thing. And then I'm going to assume that that's the effect of RFI designation. And unfortunately, it doesn't really allow you, I think, to draw that conclusion since the other provinces, the generic rate and the other provinces does not reflect some non Oh ePHI situation. You know, it depends on the province, but in many provinces, it's actually reflecting the rate at which generics are being substituted by pharmacists, because that is what they do. So he's putting the RFI rate in zolpidem in generic market share in Ontario with RFI is taking that and then subtracting the average generic market share in other provinces, where in fact RFI is not required for substitution. So it's not really the case that he's finding an RFI effect in Ontario is finding the difference between Ontario market share with RFI and the market share in other provinces where RFI is not a thing. And if I

look at your paragraph 20 of your reply report, you say in the mantiene we observe a 27% increase in generic market share in Ontario in January 2010 while the average increase in other provinces is only 3%. attributable increase in market share from the OSI designation is there for about 24% Isn't that the exact same stuff that you're describing? Dr. Gertz having done no, let

me try again. So what we're seeing segmenting, and what I'm analyzing is by the change in Ontario when MFI is given room doors does not able to identify what's the change in Ontario because he never observes the change. So instead, he just has to subtract the average generic market share in other provinces from the generic market share in Ontario given Vi has been designated. So they're about objections. So I guess that the same in that sense, but you're not subtracting the same thing. And what I'm doing is identifying the specific OSI effect that's you know, and Dr. Gruden doors does identifying what happens if an Ontario yellow fi and then you subtract other provinces, some of which already have active substitution by pharmacists?

Right but the the other provinces in your rest of Canada might also are in the exact same boat. They have active substitution by their by pharmacists to

know that they absolutely do and that's why I'm I'm tracking that relative change over time. So I use the other provinces to help identify what's happening on average. And so, you know, if any other provinces I'm seeing an increase of 3%, you know, month to month, well, probably that was going to happen in Ontario as well even without an RFI designation. So the fact of the OSI designation as a track to that, right. So I could have just like, I could have just said oh foi moved, resulted whatever in big strides. I saw go from 3% to 61%. So that says 58%. I don't want to do that, because I figured there's going to be natural growth anyway. So I'm able to control for that natural growth by using the other provinces.

Right,

but for finasteride and man teen, which might or might not be

comparable to SOPA.

Yeah, that's a very fair point. They're not they're not zolpidem

right.

Okay, so then I will move to your reply. Report or working in chronological order here sort of and and

that is, what is the FC number for that remind me FC p 353. I hope Yeah. Okay, you're already ahead of me. And so if you can go to the end of that report, starting at paragraph 47. On page it's got a 19 at the bottom, but I don't know how you're not sure what the PDF is. It's a heading as errors in the group doors report in

paragraph 47. Yes. And I'm

just started to do preface the comments in this report. I even got Dr. Dos have worked together quite a bit. It would appear.

Yes. And we continue to add Yeah, we

look at

your CV and it it appeared as though the two of you had a had collaborated virtually had published anyways almost a year together and multiple times a year is that fair?

I mean, the publication's so yeah, but we're we're in pretty good contact and, yeah, we've worked together on projects. So supervise each other students. And generally, collaborating in a number of ways over the years.

Yeah, and that, you know, we had looked at your CV and, obviously, we've spoken to Dr. Gruden Dorsen. So if you go over your, your page, two, paragraph 48

and in paragraph 48

You're addressing Dr. Gruden doors approach to his analysis, and you refer to his statement in paragraph 48. First in the second sentence says a faulty claim. And then in the last sentence says twice you refer to it as a false claim. That just struck us as an awfully strong language in respect of a colleague and friend is that what do you mean by false claim?

Incorrect okay.

I certainly don't intend to that there's any intention, you know, to

thing on his part. Right and

your your report to be fair is based on your analysis, including the question of who's the first entrant or whether there's entrance and all of that is based in in large measure on assumptions provided to you? Correct?

Yes.

assumptions about when is especially assumptions but for a world assumptions on when farmer science would have received its NOC and when it would have made its first sale, right? Yes. And they have to be assumptions because they did actually happen in the real world, right?

Again, I must agree with you right fair.

We can take five minutes justice in all that collaborate with my colleagues and but I think our policy may were either done or we're near done. So

gestion just assume that we take a little longer so that I can do the same, Mr. Scogan his colleagues are in the same room as him and mine or not, it's sometimes a little more difficult to corral them. That way I could gin up any anything following? Not that I'm aware of anything but just if we can take a little longer.

That's no problem. How much time do you think you may need? Mr. stainsby?

Why don't we take 15 minutes Mr. Scogan is almost finished. I can't imagine I would be more than very short if if there's anything. So

we'll take 15 minutes and then adjourn until 345. Thank you very much.

Thank you. For your Recordings. Stoughton. Yes, for him ready. Recording and process is now reserved. Thank you.

Anything further Mr. Scogan?

Know we do not. Thank you, Dr. Hollis. So much. Appreciate it.

Thank you, Mr.

stainsby.

For from the plaintiff in re examination and further from the plaintiff for this case of close the case.

Before you do that, thank you very much Dr. Hollis. And enjoy your weekend.

Thank you so much.

I Yeah.

I understood this morning it exchanged by by this morning's exchange but the issues had been was just gotten.

Yeah, yes. I had noted that the exhibit whatever it was that went in on the first day or something like that is we're not doing anything to adjust that or make any further noises about that.

Okay, so the exhibit three will then be taken as having been read in any housekeeping issues.

Not that I'm aware of and I'm sure if any minor comes up that comes to our mind, I'm sure Mr. Scogan that'll let us reopen and build in the first bit of Monday. I know I'm well enough to he's not gonna be he's not gonna be a dog in the manger.

I don't know. Oh,

given given the obvious cooperate. Yeah,

exactly. Now, I'm not aware of any other housekeeping issues either. And so if I haven't closed his case, the one thing we do want to know is we've, you know, been paying attention to some degree has been going along. And we don't intend to call Mr. Lavoie on Monday morning. We don't need the things that he was going to address that's been covered by my friends witnesses, either indirect or in cross. We are still going to be calling Dr. Khaleesi. And he's scheduled for half an hour right? At this point. We've paid him down as well. We'll be one minute but it'll be less than the half an hour. That's a lot it so just for my friends, information and for the courts information. I don't know how to be in cross with Dr. cabooses. But we should be well into Dr. Gruden doors down by by lunch.

Okay. Well, enjoy the weekend. Council. And we'll see you Monday at 930. Thank you justice.

Thank you is adjourned until Monday, November. Second at 930 Recording