Attachment 4 Court Reporting - Sample PTAB Transcript 6Dec23.pdf

trials@uspto.gov Case IPR2015-01117, Paper No. 52 Case IPR2015-01127, Paper No. 48 571-272-7822 September 9, 2016

UNITED STATES PATENT AND TRADEMARK OFFICE

BEFORE THE PATENT TRIAL AND APPEAL BOARD

PAR PHARMACEUTICAL, INC.,

Petitioner, v.

HORIZON THERAPEUTICS, INC.,

Patent Owner.

Case IPR2015-01117, Patent 8,642,012 B2 Case IPR2015-01127, Patent 8,404,215 B1

Held: July 26, 2016

BEFORE: TONI R. SCHEINER, DEBORAH KATZ, and GRACE KARAFFA OBERMANN, Administrative Patent Judges.

The above-entitled matter came on for hearing on Tuesday, July 26, 2016, commencing at 10:02 a.m., at the U.S. Patent and Trademark Office, 600 Dulany Street, Alexandria, Virginia.

APPEARANCES:

ON BEHALF OF THE PETITIONER:

ELIZABETH J. HOLLAND, ESQ.

CYNTHIA LAMBERT HARDMAN, ESQ.

Goodwin Procter LLP The New York Times Building 620 Eighth Avenue New York, New York 10018 and

DAVID H. SILVERSTEIN, M.S., J.D.

AZIZ BURGY, ESQ.

Axinn, Veltrop & Harkrider LLP 114 West 47th Street New York, New York 10036

ON BEHALF OF PATENT OWNER:

ROBERT F. GREEN, ESQ.

EMER SIMIC, ESQ.

Green Griffith 455 N. Cityfront Plaza Drive, Suite 3100 Chicago, Illinois 60611

P R O C E E D I N G S 1

- - - - - 2

JUDGE OBERMANN: Please be seated. 3

Good morning and welcome to the PTAB. This is a 4 consolidated trial hearing in IPR2015-01117 and IPR2015-01127. 5

Petitioner is Par Pharmaceutical, Inc., and Patent Owner is 6

Horizon Therapeutics, Inc. I'm Judge Obermann, Judge Scheiner 7 is on my right, and Judge Katz is on my left. 8

Before we begin, I would just like to remind the parties 9 that the hearing is open to the public, and a full transcript of it 10 will become part of the public record. When you refer to an 11 exhibit on the screen, please state the slide, the exhibit, or the 12 page number to which you are referring for the record. That's 13 important for clarity in the transcript. 14

As you know from our order of July 18th, each party 15 has one hour, in total, to present their argument. Because 16

Petitioner has the burden to show unpatentability of the original 17 claims, there's no motion to amend here, Petitioner will proceed 18 first, followed by Patent Owner. Petitioner's counsel may reserve 19 rebuttal time and divide the hour of time among the cases as it 20 wishes. 21

Now, in order to keep this case focused on the merits, I 22 am going to ask counsel not to interrupt the other side to make 23 objections. Any objections should be discussed during your own 24 allotted argument time. 25

I have two preliminary matters that I wanted to discuss. 1

The first, we sent the parties an email yesterday regarding paper 2

47 in IPR '117 and paper 43 in IPR '127. Those are identified in 3 our filing system as Petitioner's Reply in Support of the Motion to 4

Exclude, but the actual papers appear to be duplicate copies of 5

Petitioner's Supplemental Reply to Patent Owner's Corrected 6

Response. 7

So, as we explained in our email yesterday, Petitioner, 8 you may address that issue in your opening argument if you like, 9 and, Patent Owner, you may respond during your allotted hour of 10 time to the extent that the Petitioner has argued it in their case in 11 chief. 12

The second thing I'd like to do, I want to let you know 13 that the Panel has considered the joint objections to 14 demonstratives that were filed in both cases. We've determined 15 that both parties may use their demonstrative exhibits today. We 16 overrule Patent Owner's objections based on the timing of 17 service. In that regard, we understand that Petitioner did file what 18 they purported to be final exhibits the day before our order 19 issued, but our order does trump the timing in this case. So, we 20 will let the Petitioner use their demonstratives as served five days 21 before. 22

We take under advisement Petitioner's concerns that the 23

Patent Owner's demonstratives include new evidence. We 24 routinely disregard new evidence that's inserted into 25 demonstratives. They're not evidence in the trial, and we will 1 focus only on content that has been made of record during the 2 trial. We're adept at making that distinction. So, for that reason, 3 we are not going to exclude any exhibits. 4

We will also take under advisement Petitioner's second 5 objection to Patent Owner's demonstratives in IPR '127, going to 6 the content of slide 19. We will take that objection into account 7 when we prepare our final written decision, but, again, Patent 8

Owner is free to use slide 19 during the hearing. 9

Do we have any questions from either side on either of 10 those? 11

MS. HOLLAND: No, Your Honor. 12

JUDGE OBERMANN: Okay, thank you. With that, 13 let's take a moment for counsel introductions. Who will be 14 presenting argument for Petitioner today? 15

MS. HOLLAND: Good morning, Your Honor. 16

Elizabeth Holland of Goodwin Procter, LLP. We are actually 17 counsel for Petitioner Lupin in the joint IPR, but I will be arguing 18 today on behalf of both Lupin and Par. 19

JUDGE OBERMANN: Okay. And it's Ms. Holland? 20

MS. HOLLAND: Yes. 21

JUDGE OBERMANN: Thank you very much. 22

And for Patent Owner? 23

MS. SIMIC: Emer Simic from Green Griffith, 24 representing Horizon Therapeutics, Inc. 25

JUDGE OBERMANN: Thank you very much. 1

Petitioner, would you like to reserve any time for 2 rebuttal? 3

MS. HOLLAND: Yes. I would like to reserve 30 4 minutes for rebuttal. 5

JUDGE OBERMANN: Okay. What I am going to do, 6 then, is I'm going to allow you to approach the Bench, and for 7 both counsel, your time will start when you begin speaking. I am 8 going to do is put your time behind me on the clock. So, I will 9 set the clock right now for 30 minutes for your opening argument, 10

Petitioner. 11

MS. HOLLAND: Thank you. 12

JUDGE OBERMANN: Okay. When you begin 13 speaking, I will start the clock, Ms. Holland. Thank you. 14

MS. HOLLAND: Thank you. 15

Good morning. The patents at issue in these two IPRs 16 relate to what was admittedly a very standard treatment as of 17

2008, the use of nitrogen-scavenging drugs to treat urea cycle 18 disorders, which I will be referring to as UCDs. There's no 19 dispute that the general methods claimed in these patents -- in 20 other words, the use of urinary PAGN to select dosages for 21 nitrogen-scavenging drugs as claimed in the '012 and the use of 22 blood ammonia levels as claimed in the '215 patent -- that those 23 were known generally in the prior art as of the priority dates here. 24

What Patent Owner claims are really observations that 1 the Patent Owner claims to have made based on particular data 2 sets from particular patients that were tested in clinical trials. 3

Petitioner's IPRs here are rooted in the testimony of Dr. Neal 4

Sondheimer, who is an expert in the treatment of UCDs. 5

Dr. Sondheimer explains the prior art here from the perspective of 6 a person of ordinary skill in the art, which, of course, is what is 7 required under the statute and under Federal Circuit law. It's 8 telling in this case that Patent Owner did not submit any expert 9 evidence, even though there was an invitation to do so in the 10 institution decisions in this case. 11

I'm going to start by talking about the '012 patent, so if 12 we can go to slide 2, please, and let's go to slide 3. Thank you. 13

Slide 3 has independent claim 1 of the '012 patent. The '012 14 patent is Exhibit 1001. And there is really no dispute here that 15 determining a target UPAGN output was known in the art; 16 calculating an effective initial dosage of a PAA prodrug selected 17 from HPN-100 or PBA or pharmaceutically acceptable salt of 18

PBA was also known in the art; and under subpart (c), 19 administering the effective dosage of PAA prodrug to the patient 20 was also known in the art. That's all conceded in this case. 21

The only dispute really relates to the wherein clause you 22 see on the screen in independent claim 1, which says that the 23 effective dosage of the PAA prodrug is calculated based on a 24 mean conversion of PAA prodrug to urinary PAGN of about 60 25 percent. So, the first issue here is, what does "about 60 percent" 1 mean. 2

Can we go to slide 10, please. As Petitioners have 3 argued in their papers, the broadest reasonable interpretation of 4

"about 60 percent" in this case has to be at least 53 to 67 percent. 5

Now, there's no dispute that the upper limit here has to be at least 6

67 percent. That was conceded by Patent Owner in their papers, 7 and, of course, they had to concede that because the Scharschmidt 8 declaration that was submitted during prosecution had the mean 9 conversion rate of 67 percent, and based on the information in 10 that declaration, the claim of "about 60 percent" was allowed. 11

So, certainly, it has to go up to at least 67 percent, but 12 this is, of course, a claim term that says "about," which is a 13 plus/minus term. If there are seven percentage points above 60, 14 there need to be seven percentage points below 60. So, at a 15 minimum, the claim has to range from 53 to 67 percent. 16

Now, why do I say "at a minimum"? Because when 17 you look at the patent, what the Patent Owner observed here in 18 these clinical trials and what they attempt to claim in the patent is 19 actually a lot broader than 53 to 67 percent. So, for example, if 20 we look at the specification -- can we go to Exhibit 1001 at 21 lines -- page 14, lines 28 to 29, please. I'm sorry, I wanted to look 22 at the patent, Exhibit 1001 at page 14, lines 28 to 29. It's column 23

14, lines 28 to 29. I apologize. Sorry for the slight interruption 24 here. 25

Let me just read it while we wait for that to get on the 1 screen. What Patent Owner says in column 14 is it has been 2 found that between 50 and 85 percent of HPN-100 is converted 3 into urinary PAGN, typically about 60 percent to about 75 4 percent, and they refer to this as the conversion efficiency for 5

HPN-100 in UCD patients. So, the actual observations that are 6 discussed in the specification of the patent, based on what they've 7 seen in clinical trials with UCD patients, ranges from 50 to 85 8 percent. So, there is really quite a large range, quite a lot of 9 intervariability among patients in terms of the percentage 10 conversion of PAA prodrugs to PAGN in the urine. 11

If we go to column 30 -- again, this is Exhibit 1001 -- 12 and we look at Example 2, what the patent says is that the mean 13 percentage conversion of administered PAA into urinary PAGN 14 began was about 75 percent. And can we blow up the table at the 15 bottom of the page in the lower right-hand corner, please? So, 16 this is the table at the bottom of column 30 of the '012 patent, 17

Exhibit 1001. If you look at the line for mean range molar 18 percent of dose ammonia scavenge, what you will see there is 19 quite a large range of observed rates of conversion from PAA 20 prodrug to PAGN, ranging in absolute terms from 58.2 to 85, but, 21 of course, with very large standard deviations. So, again, we 22 believe that the broadest reasonable construction here has to be at 23 least 53 to 67 percent, but actually, based on the specification, 24 can be broader than that. 25

Now, I'd like to start with Brusilow Exhibit '91. Could 1 we go to slide 16, please. Brusilow '91 discusses conversion of 2

PAA prodrug to PAGN in a patient with three different dosages, 3 and what you can see clearly from Brusilow '91, Table 1, is that 4 there was incomplete conversion of the prodrug to PAGN. You 5 can see in period one, it was 83 percent; period two, 90 percent; 6 period three, 80 percent. So, there is no question, based on the 7 testing that was actually done in Brusilow '91, Exhibit 1012, that 8 there was incomplete conversion of -- into PAGN in this patient. 9

A person of ordinary skill in the art looking at Brusilow 10

'91, as explained by Dr. Sondheimer in his declaration, would see 11 that there's incomplete conversion here. This is, however, based 12 on one patient. The person of ordinary skill in the art is going to 13 want to look at the prior art as a whole to determine how much 14 variability is there really in these determinations of conversion 15 into PAGN, and what the person of ordinary skill in the art finds 16 is the Sherwin '19 reference. 17

If we can go to slide 19, please. Slide 19 is from the 18

Sherwin '19 reference, Exhibit 1016, at page 116. It's Table 1. 19

And, again, what we see is a range of conversion percentages into 20 urinary PAGN, and you can see it ranges from 49.65 to 67.67. 21

So, a person of ordinary skill in the art, looking at Sherwin '19, 22 sees about 60 percent conversion within the meaning of the 23 claims. 24

Dr. Sondheimer explained that a POSA would actually 1 understand the percentages in Sherwin to be a little low because 2 the subject wasn't dosed throughout the day; and in addition, this 3 is a healthy subject, and the person of ordinary skill in the art 4 would expect the conversion percentage to be a bit higher in a 5

UCD patient. 6

But the person of ordinary skill in the art would know, 7 based on the Shiple reference, that basically the metabolism in a 8 healthy person and in a UCD patient is the same, that the prodrug 9 is going to be converted into PAGN in the same way. So, while 10 there may be small differences in the amount of the conversion, 11 generally the pathway is the same. 12

Can we go to slide 25, please. When a person of 13 ordinary skill in the art went to the prior art to look at the urinary 14

PAGN conversion percentages, they would also see the 2002 15

Comte reference, and this is -- on slide 25 is an excerpt from 16

Comte, Exhibit 1025 at page 589. And what you see here, again, 17 is very similar to what we saw in Sherwin in terms of the 18 percentage conversion. What we see here is 53.4, plus or minus 19

4.5 percent conversion. 20

So, the person of ordinary skill in the art really sees a 21 range in the prior art, sees a lot of variability, does see in 22

Brusilow '91 the 80 -- approximately 80 to 90 percent; sees in 23 other references approximately 53 percent, in Comte as well as in 24

Sherwin '19; and understands that the prior art discloses about 60 25 percent as one range in the prior art for the PAGN conversion 1 percentage. 2

In addition, the person of ordinary skill in the art would 3 know about the Simell reference, and this is Exhibit 1005. If we 4 could put that up on the screen, please, at page 1121, and can we 5 bring up the highlighted portion, please. This, again, is from the 6

Simell reference. This reference tested five children who actually 7 had UCD disorder. So, we're not talking about healthy patients in 8 this reference; we're talking about UCD patients. 9

And, again, what we find, very similar to what we saw 10 in Sherwin '19, very similar to what we saw in Comte, we see 54 11 percent of the PAA was excreted as urinary PAGN. So, the prior 12 art really is coming to the consensus -- if you want to use that 13 term, that's a term that Patent Owner used in their papers -- but if 14 you really want to look at the prior art, the consensus more so is 15 that the conversion rate was about 60 percent within the meaning 16 of the claims. 17

Now, if we can put up on the screen the chart that's on 18 pages 6 to 7 of the reply, of Petitioner's reply, and that's paper 19 number 30. What you can see here is really a wide -- just putting 20 in one place the wide variety of percentages converted to 21

UPAGN that we've seen in the prior art. Can we make that a little 22 larger, please? 23

So, starting with Brusilow '91, we see, as I've discussed 24 already, 83 percent, 90 percent, 80 percent. In Sherwin '19, 25 again, we have this variability. We go 53 percent, 68 percent -- 1 can we continue to the next page? Thanks -- 50 percent, 52 2 percent, 51 percent. Comte has the 54 percent. And Simell, as 3 well, has the 54 percent. 4

So, based on the prior art as a whole, we see every 5 element of the claims of the patent, the '012 patent. We see, 6 based on Brusilow '91, administration to a UCD patient. We see 7 conversion rates falling within the claims in many references, 8

Sherwin '19, Comte, and Simell. 9

I'm going to move on now to the '215 patent, unless 10 there are any questions. So, turning to the '215 patent, unlike for 11

'012 patent, Patent Owner did give a detailed response on the 12 substantive issues in this case in its -- that had been raised in the 13 petition. After considering the Patent Owner's arguments and the 14 evidence, the Board found that there was a reasonable likelihood 15 that the Petitioner would prevail in this case, but the full trial 16 record here is very, very similar to what it was at that stage. 17

As I mentioned earlier, Patent Owner hasn't submitted 18 any expert evidence to rebut anything that Dr. Sondheimer has 19 said. So, Petitioners respectfully submit that the Patent Owner 20 here has given the Board really no reason to deviate or to come to 21 a different decision than the one that was reached at the 22 institution stage. 23

Can we go to slide 48, please. Slide 48 is independent 24 claim 1 of the '215 patent, and these elements are all in the prior 25 art. The Fernandes reference that's discussed in our papers 1 teaches measuring a fasting blood ammonia level. It teaches 2 comparing the fasting -- I'm sorry, it teaches measuring of blood 3 ammonia level, it teaches comparing the blood ammonia level to 4 the upper limit of normal, and it teaches administering the 5 adjusted dosage of the nitrogen-scavenging drug where the 6 dosage is greater than the initial dosage if the blood ammonia 7 level is greater than half the upper limit of normal. There's really 8 no way to read that reference and not see all those claim 9 elements. The one thing that is not explicit in that reference is to 10 look for fasting blood ammonia levels, but that was clearly in the 11 prior art as well. 12

If we turn first -- let's look first a little bit more deeply 13 at the Fernandes reference. Can we go to slide 62, please. Thank 14 you. So, slide 62 is Figure 17.2 from the Fernandes reference, 15

Exhibit 1011, and this is a flow chart that's provided in Fernandes 16 to teach persons of ordinary skill in the art how to manage 17 treatment of patients with UCD disorders. 18

And as Dr. Sondheimer stated, clearly the central 19 decision factor here is plasma ammonia. You can see that at the 20 top of the decision tree. It's the first thing that the clinician -- the 21 person of ordinary skill in the art has to look at, and as 22

Dr. Sondheimer says, this Figure 17.2 in Fernandes really does 23 give a very realistic picture of how physicians manage patients 24 with UCD disorders. 25

JUDGE KATZ: Can I stop you and ask you a question 1 about the figure? So, the plasma ammonia level is a decision, but 2 it seems like you can get to treatment whether you have either 3 less than 80 -- whatever it is -- micromoles, molar, or more than 4

80. How do you explain that that makes a decision, then? 5

MS. HOLLAND: Clearly on the chart, the first thing 6 that the physician looks at is the plasma ammonia level. You are 7 correct that there are two decision branches there that could go in 8 either direction based on the 80, but clearly one of the decision 9 branches is based on greater than 80, which is greater than the 10 upper limit of normal, which falls within the claim. 11

And if you go down -- maybe it's easiest if we look at 12 slide 66 to show exactly that particular branch I'm talking about. 13

If we go to slide 66, you see the first is a decision point. You see 14 one possibility of that first point is greater than 80 micromoles 15 per liter, and if you follow that down, one of the outcomes there 16 is to increase the level of medicine, of nitrogen-scavenging drug. 17

So, while there are different treatment methods within 18 this one figure, one of them is one that falls within the claims of 19 the patent, and, you know, there can be others on this chart, but 20 the question is, is there -- does this chart teach one way of getting 21 to the claims in the patents, and it does. 22

Can we go to slide 67, please. So, what we've just 23 looked at is this method of adjusting dosages based on the 24 upper -- above half the upper limit of normal, and as I mentioned 25 earlier, the one thing that's not explicit in Fernandes is this fasting 1 blood level, but that's also very clear in the prior art, and as 2

Dr. Sondheimer explains, very, very clearly what the person of 3 ordinary skill in the art understands based on the prior art and 4 based on their clinical practice. 5

So, this is from the Blau reference, Exhibit 1006, and as 6

Dr. Sondheimer explained, this is a companion reference to the 7

Fernandes reference, and this goes towards laboratory diagnosis 8 of metabolic diseases. It's a physician's guide not to necessarily 9 what happens in the clinic, but what happens in the laboratory. 10

And what Blau says is when you are testing for 11 ammonia, the precondition is that it's at least four hours after the 12 end of the last meal or stopping intravenous AA supply from a 13 central vein or artery; in other words, fasting blood levels. 14

And this is -- as we've seen in many of the references 15 here, it's -- persons of ordinary skill in the art understand that they 16 need to monitor these blood ammonia levels for the patient 17 consistently, and when they take a specimen from a patient, to 18 measure the blood ammonia level. When they do it, they do it in 19 a fasting state. 20

And if you think about it, that makes a lot of sense, as 21

Dr. Sondheimer explained. It's the optimum way to determine 22 ammonia level because you don't have the variations in ammonia 23 level that happen if you take a sample after a meal, for example. 24

And if you want to be consistent, the way to do that is to take 1 fasting blood ammonia levels. 2

Can we go to slide 74. Thank you. This is another 3 prior art reference that we discussed in our papers, the '859 4 publication. It's Exhibit 1008. And, again, this is a method that's 5 set forth in the '859 publication for identifying a suitable starting 6 dose or a dosage for a UCD patient and adjusting that dosage. 7

And if you look at paragraph 98 in Exhibit 1008, the 8

'859 publication, what you see is that it clearly states "optionally 9 measuring blood ammonia to determine if the initial dosage is 10 sufficient to control blood ammonia levels, or to establish a 11 suitable average ammonia level." So, a clear teaching here. It's 12 one method of practicing what's in the '859 patent, one teaching 13 of the '859 patent, is to measure blood ammonia levels to 14 determine if the initial dosage is sufficient. 15

Can we now go to slide 73, please. In the '859 16 publication, what we see is the upper limit of normal described as 17 between 26 and 36 micromoles for the subjects, and then it says 18 generally below about 40 micromoles. And what we see in the 19 prior art is, you know, similar to what we see in the '012 patent in 20 terms of conversion percentages, there's variability here. So, in 21 other words, if you're looking at a neonate, the upper limit of 22 normal is going to be higher. It's going to be maybe in the range 23 of 80 micromoles, which is what we saw in Fernandes. If you're 24 looking at adults or older children, you are going to see 25 something lower, but that is known and standard to the person of 1 ordinary skill in the art, as Dr. Sondheimer explains. 2

So, again, based on the '859 publication, Fernandes, 3

Blau, we contend here that these claims, the claims of the '215 4 patent, are obvious and that they should be cancelled. 5

Are there any questions or else I will reserve the rest of 6 my time? 7

JUDGE OBERMANN: Thank you very much. 8

MS. HOLLAND: Thank you. 9

JUDGE OBERMANN: So, I am going to put five 10 minutes to your rebuttal time, if that's all right. 11

MS. HOLLAND: Thank you very much. 12

JUDGE OBERMANN: When you are ready, I will start 13 your time. I am just going to put the clock up to 60 minutes for 14 you. 15

(Pause in the proceedings.) 16

MS. SIMIC: May it please the Board. 17

Prior to the inventions that are claimed in the '012 18 patent, it is undisputed that physicians were making dosing 19 recommendations for patients with urea cycle disorders assuming 20 complete or 100 conversion of PAA to PAGN. Now, the inventor 21 of the '012 patent discovered, upon analyzing a set of clinical data 22 from urea cycle disorder patients, that, in fact, conversion of PAA 23 to PAGN was, in fact, incomplete and as low as about 60 percent. 24

If I can turn to slide 2. So, the inventor of the '012 1 patent, Bruce Scharschmidt, translated his discovery into a 2 method of improving dosing in urea cycle disorder patients that 3 focused on calculating an effective dose based on that mean 4 conversion of PAA prodrug to PAGN of about 60 percent. 5

Claim 1 deals with the situation where you're 6 calculating an initial dose, whereas claim 8 deals with the 7 situation where you are calculating an adjusted dose, and under 8 those circumstances, you administer an initial dose, you measure 9 the amount of PAGN being excreted by the patient, and then you 10 calculate an effective dose based on the amount excreted by the 11 patient and bearing in mind the mean conversion data that was 12 already known. 13

So, there are three primary points I'd like to make in 14 response to Petitioner's arguments. The first is -- and turning to 15 slide 6. Turning to slide 6, Brusilow '91 actually teaches away 16 from the claimed inventions. The express conclusion in Brusilow 17 is that conversion of PAA to PAGN is, in fact, complete or nearly 18 complete in the patients that he studied. 19

Second of all, with respect to claim 8, Brusilow 20 specifically did not teach or suggest calculating an effective 21 dosage based on the urinary PAGN that was excreted by the 22 patient. Instead, he simply predetermined the doses and 23 measured how much PAGN came out. 24

And finally, a person of ordinary skill in the art would 1 not have combined the Brusilow '91 disclosure with Sherwin '19 2 for a number of reasons. First of all, Brusilow taught away from 3 a combination with an additional reference concerning 4 incomplete conversion because it taught the conversion was 5 complete; but second of all, Sherwin '19 was, in fact, discredited 6 in a later paper by the same author who found that the method of 7 extraction of PAGN from the urine was, in fact, unreliable and 8 led to incomplete -- or, rather, led to the conclusion of incomplete 9 conversion. 10

Now, the secondary references that counsel mentioned, 11

Comte and Simell, do not cure the deficiencies of Brusilow '91 in 12 that respect. They don't contain any relevant disclosure of overall 13

PAA to PAGN conversion. And I'll note for the record that the 14

Simell reference is, in fact, a reference that was not the subject of 15 the institution decision. It has appeared in Petitioner's reply. 16

Patent Owner has not had an opportunity to respond to the 17 argument based on Simell, and we would object to the use of that 18 reference in this IPR. 19

Now, moving on to my first point, which is really the 20 key to this issue, Brusilow '91 clearly teaches that conversion of 21

PAA to PAGN was complete. His conclusion is based on the 22 data in his article. The data is not limited to Table 1, which I can 23 bring up, but, in fact, was based on the overall data that was 24 collected for the individual patient. 25

So, just to give the Board a little background on the 1 experiment, Brusilow administered 10, 12, and 14 grams of a 2

PAA prodrug -- two different drugs, in fact -- to a seven-year-old 3 boy and then collected the amount of -- collected urine over 24 4 hours as best they could for that patient and measured how much 5

PAGN essentially appeared and was recovered from the urine. 6

So, with respect to the recovery that was found in the urine that 7 was collected, they determined that between 83 and 90 percent 8 had been recovered. 9

However, that's not the whole story. The reason 10

Brusilow concluded that conversion was nearly complete in the 11 patient was based upon the fact that no unconverted drug was 12 found in the urine. So, less than 1 percent of the dose was found 13 unconverted in the urine, and there was no accumulation in the 14 overnight plasma for the patient. So, that led Brusilow to his 15 conclusion, which was that conversion was, in fact, complete or 16 substantially -- or significantly complete, up to 100 percent, in 17 that patient. 18

So, counsel's argument that Table 1 shows incomplete 19 conversion is, in fact, incorrect, and it conflicts with the express 20 conclusions in Brusilow '91, which -- if I just may refer to slide 21

7 -- Brusilow concludes that phenylbutyrate appears to be 22 completely oxidized to phenylacetate and that phenylacetate is 23 completely or nearly so conjugated with glutamine, and that 24 complete conjugation occurs may be further adduced by the 25 insignificant amount of unchanged drug or their esters in the 1 urine. 2

Moving to my second point, which relates to claim 8, 3 and that's step (c) of claim 8, with respect to Petitioner's argument 4 that there was somehow a dosage adjustment or a determination 5 of an effective dosage based on the actual urinary PAGN excreted 6 by the patient, that simply makes no sense based on how the 7 publication is written and essentially reads an interpretation of the 8 publication that has no support. 9

First of all, the idea that Brusilow averaged the 10 conversion from period one to period two to come up with an 11 adjusted dose for period three simply conflicts with the -- with the 12 data and the language in Table 1. 13

Now, Petitioner's expert asserts that in paragraph 65 of 14 his declaration, Exhibit 1002, that Brusilow calculated the 15

14-gram dose in period three based on the average conversion of 16

PAA to PAGN observed in periods one and two. This simply 17 makes no sense because for each period where the PAA prodrug 18 was administered to the patient, Brusilow predicted 100 percent 19 conversion, and that was, in fact, admitted by Petitioner's expert 20 at deposition. He admitted that for each period, Brusilow's 21 express prediction was 100 percent conversion. 22

JUDGE OBERMANN: Why don't we need testimony 23 on that from you? 24

MS. SIMIC: Well, we believe the record is clear on this 1 point, and Petitioners appear to be using their expert to read into 2 the text something that just simply doesn't exist and isn't 3 corroborated elsewhere. The conclusion is clear that conversion 4 was complete, and the conclusion also is clear that -- rather, the 5 data is clear that he predicted 100 percent conversion in each 6 instance. 7

JUDGE OBERMANN: Okay. 8

JUDGE SCHEINER: Excuse me. You are basing that 9 on your contention that he predetermined all three of those doses 10 before beginning the trial? 11

MS. SIMIC: That's right, and that's how it's stated in 12 the article itself. He simply states that he administered 10, 12, 13 and 14 grams, round number doses, to the patient in order to 14 measure or to study the stoichiometry of the reactions, so simply 15 administering increasing doses and then determining how much 16 he could recover from the urine to determine a percentage 17 recovery. That's all that Dr. Brusilow states. 18

So, to adopt Petitioner's interpretation simply asks 19 someone to -- asks the Board to read between the lines to adopt 20 an interpretation that Brusilow himself -- the only -- the only 21 assertion you can make is that he was conceding from the reader 22 how he was calculating his dose in group three, which just makes 23 no sense. 24

JUDGE OBERMANN: So, in other words, your 1 position is that their expert testimony is not supported by the best 2 objective evidence, which is the document itself? 3

MS. SIMIC: That's correct. 4

JUDGE OBERMANN: All right. 5

MS. SIMIC: Additionally, the publications postdating 6

Brusilow '91 support Patent Owner's position on this point. So, if 7 we turn to slide 12, the 2003 FDA-approved prescribing 8 information for Buphenyl, which is a PAA prodrug also known as 9 sodium phenylbutyrate, specifically assumed 80 to 100 percent -- 10 so near complete conversion of PAA to PAGN -- in dosing 11 patients. So, physicians were essentially being instructed to dose 12 patients based on an assumption of complete conversion. 13

Likewise, the 2001 Urea Cycle Disorder Conference 14 statement, which is a series of recommendations based on thought 15 leaders in the urea cycle disorder community, they make no 16 mention of monitoring PAGN for patients or calculating a dosage 17 based on incomplete conversion in a urea cycle disorder patient, 18 because that's simply not what Brusilow '91 taught. 19

Now, moving on to my third point, a person of ordinary 20 skill in the art simply would not have combined Brusilow '91 with 21

Sherwin '19. Brusilow '91 concludes that conversion is complete 22 or nearly complete. So, it clearly teaches away from the claims 23 and refutes any purported motivation to combine. 24

However, even if a person of ordinary skill in the art 1 was to seek out additional information about conversion of PAA 2 to PAGN, Patent Owner -- rather, Petitioner provides no 3 explanation as to why they would have sought out a reference 4 dating from 90 years before concerning healthy patients in order 5 to determine what the conversion percentage was. And, indeed, 6 not only does Petitioner say that they would have sought out that 7 reference, they, in fact, suggest that the person of ordinary skill in 8 the art would have substituted the data, so the data from the 1919 9 reference, for the data that Brusilow himself obtained with 10 respect to a urea cycle disorder patient in 1991. 11

Moreover, Sherwin '19 was actually discredited by a 12 later publication by the same author, a publication that we refer to 13 as Sherwin '33, which is Exhibit 2027. Sherwin '33 sought to 14 explore further why Sherwin '19 recovered so little PAGN from 15 the urine, and they, in fact, repeated the experiments that Sherwin 16

'19 performed and determined that the experiment itself, the 17 method of extraction, was responsible for the incomplete 18 conversion and not incomplete conjugation. 19

Furthermore, Sherwin '33 actually concludes, based on 20 additional experiments, that conversion of PAA to PAGN is 21 complete. This goes to Patent Owner's point throughout its 22 briefing that the consensus in the art was that PAA conversion to 23

PAGN was, indeed, complete. Petitioner provides no explanation 24 as to why its expert does not address the Sherwin '33 reference, 25 even though its expert admitted at deposition that Sherwin '33 1 would have been one of the references that a person of ordinary 2 skill in the art would have considered. 3

Petitioner's argument in response to Patent Owner's 4 position that Sherwin '19 was discredited focuses on superficial 5 details, simply that Sherwin '33 didn't specifically cite to Sherwin 6

'19, but it's clear from the face of the articles that Sherwin '33, ten 7 years after the original experiments, went out and tried to figure 8 out what was going on and how could they account for the 9 incomplete conversion that they found ten years previous. I noted 10 earlier that the same author is the co-author on both papers and 11 the same university or laboratory was being used for both studies. 12

Now, in Sherwin '33, the author specifically sought to 13 examine whether it was the method of analysis that accounted for 14 the low recovery, and to do so he performed a recovery 15 experiment. So, what he did was he dissolved a known amount 16 of PAGN in urine and then used essentially the same method as 17 what Sherwin '19 used to recover that PAGN from the urine, and 18 he discovered -- and -- and the method that was used is not 19 specifically described in Sherwin '19. He refers to "the usual 20 way" in footnote 5. 21

That refers to Exhibit 29, which is the Hawk reference, 22 and at slide 16 here you'll see that the Hawk reference, just like 23

Sherwin '19, used a method of evaporation from a water bath. In 24 fact, Hawk even refers to the Sherwin '19 method for the second 25 half of the recovery experiment for any isolated PAGN from its 1 urea product. So, Sherwin '33 performs the experiment again and 2 discovers that he only recovers 47 percent of PAGN from the 3 urine. 4

So, based on that recovery experiment, he concludes 5 that the apparently incomplete conjugation of phenylacetic acid to 6

PAGN that he found in his earlier experiment was not due to 7 conjugation with other subjects but instead due to hydrolysis of 8 phenylacetylglutamine upon evaporation in the water bath. 9

Now, Sherwin '33 also performed some additional 10 experiments that demonstrated that conversion of PAA to PAGN 11 was, in fact, complete. He actually concluded in that '33 12 experiment that about 95 percent of PAA had been detoxicated or 13 conjugated with glutamine. 14

Further support for Patent Owner's position regarding 15

Sherwin '19 being discredited can be found in later references. 16

Later references in the art do not cite to Sherwin '19 for 17 conversion of PAA to PAGN. Instead, they cite to Sherwin '33. 18

As you'll see on slide 19, I have an excerpt from the '647 patent 19 on the left-hand side of the page. The '647 patent was the patent 20 for which Dr. Saul Brusilow was the inventor, and Dr. Brusilow, 21 who is the author of all these papers we're talking about -- 22

Brusilow '91 and Simell -- Dr. Brusilow actually pioneered the 23 use of PAA prodrugs in the eighties, and he was the first to claim 24 the use of nitrogen-scavenging agents to treat urea cycle 1 disorders. 2

And in the background of his specification in the '647 3 patent, he specifically calls out the Sherwin '33 reference. He 4 refers to it as Ambrose, et al., and states that they reported they 5 found 98 percent of already administered phenylacetic acid was 6 excreted in the urine at phenylacetylglutamine. So, here we have 7 the expert in the field specifically citing to Sherwin '33, not to 8

Sherwin '19. And similarly, Brusilow, in his 1991 paper, begins 9 his experiment assuming complete conversion of PAA to PAGN, 10 and predicts complete conversion prior to beginning his 11 experiments. 12

Petitioner's assertion that Comte or Kasumov -- which 13 are actually essentially the same disclosure with respect to PAA 14 to PAGN conversion, they both describe the same experiment -- 15 their assertion that somehow these references provide a relevant 16 disclosure about overall PAA to PAGN conversion is simply 17 incorrect and by their own standards doesn't meet the relevant 18 criteria. So, for example, in Comte, the subjects, which were 19 healthy subjects, were dosed with only 15 to 25 percent of a 20 therapeutic dose. 21

Now, if I may briefly put on the ELMO -- let's see -- oh, 22 it's up there, okay. So, just referring to Petitioner's slide 21, as I 23 mentioned, Comte only describes or, rather, doses a single dose 24 of 15 to 25 percent of a therapeutic dose, which is actually 25 amounted to about a five-gram dose for the patients studied, and 1

Petitioner's expert, Dr. Sondheimer, specifically states that a 2 clinically relevant dose, one that we should consider for 3 determining what a mean conversion is, would be ten grams or 4 higher, and he uses that metric to extract data from the Sherwin 5

'19 reference and combine it supposedly with the Brusilow '91 6 reference. So, by Petitioner's own expert's admission, the dose in 7

Kasumov would not be relevant in determining whether 8 conversion of PAA to PAGN or determining, rather, the overall 9 conversion of PAA to PAGN. 10

Moreover, the other element that Comte lacks in this 11 regard is -- pardon me for the interruption. 12

The other relevant data -- rather, the other relevant thing 13 to take away from Comte is that they actually didn't recover urine 14 over 24 hours, and Brusilow '91 did collect urine over 24 hours 15 for his patient, and they, in fact, only collected urine for eight 16 hours. And Petitioner's expert at his deposition similarly agreed 17 that with respect to the Comte data, you can't make any 18 conclusions about overall PAA to PAGN conversion. 19

JUDGE OBERMANN: Excuse me, I do have one 20 question. I heard you discuss the Hawk reference, 2029, and it 21 sounded to me that you were discussing its content and its 22 implications as prior art. What happens if we determine that we 23 can't consider Hawk, Exhibit 2029, for that purpose? What does 24 that do to your case as far as tying Exhibit 2027 to Exhibit 1016? 25

MS. SIMIC: Well, the Hawk reference, as I mentioned 1 earlier, is specifically referred to in footnote 5 of Exhibit 2027, 2 the Sherwin '33 reference. So, Sherwin '33 tells the reader that 3

I'm using this method to do a recovery experiment to determine 4 whether PAGN is complete. 5

However, even if the Board was not to consider that 6 reference, the conclusion from Sherwin '33 is clear. He seeks to 7 determine whether his previous method was correct. He goes and 8 performs a recovery experiment with PAA to P -- rather, with 9

PAGN in urine and discovers that evaporation in the water bath, 10 which he expressly states was the method in Sherwin '19, that that 11 evaporation step is what results in incomplete conversion. 12

So, even if the Board doesn't consider Exhibit 2029, the 13

Board will still have the knowledge, based on Exhibit 2027, that, 14 indeed, the evaporation step, which is the same in Sherwin '19 15 and Sherwin '33, that step is what's responsible for the incomplete 16 conversion. And so the -- his conclusions still stand in the 17 absence of the Hawk reference. 18

JUDGE OBERMANN: Okay, thank you. 19

MS. SIMIC: So, I believe my last point simply with 20 respect to Comte is that it doesn't provide any recovery over 24 21 hours for urine, which is the relevant recovery according to 22

Brusilow for PAA to PAGN recovery. 23

Now, I'd also like to refer to one of Petitioner's slides 1 that related to the Comte reference, because I think that slide is 2 particularly misleading. There we go. 3

So, I'm referring here to slide 25 of Petitioner's deck. 4

So, Petitioner referred to slide 25 and stated that this slide 5 supports a data point in the prior art of PAA to PAGN conversion 6 of about 53.4 percent. This statement is completely misleading. 7

The 53.4 percent is the total number of metabolites or, rather, the 8 total percentage of metabolites that were identified by the author 9 after the study. So, that included PAGN, which is the subject of 10 the claims, but also a different metabolite, PBGN. Comte only 11 concluded that about 32 percent of the -- of conversion actually 12 was to PAGN and roughly 20 percent to PBGN. So, Comte does 13 not make any conclusions that have conversion of PAA to PAGN 14 at around 54 percent. 15

Nevertheless, the conclusion in Comte is simply only 16 good for the proposition that under incomplete collection times 17 and with a nontherapeutic dose, they obtained 32.6 percent 18 recovery of PAGN. So, it's our position that that disclosure is not 19 relevant with respect to making any conclusions about overall 20 conversion of PAA to PAGN in urea cycle disorder patients. 21

Now, Petitioner brought up the reference -- rather, 22 relied on the Simell reference. Simell, which we mentioned our 23 objection to earlier, is simply not within the scope of the 24 instituted IPR, and Petitioner's reliance on that reference in its 25 reply is not as rebuttal to Patent Owner's statement. That's not the 1 case. In fact, in their reply at page 6 to 7, they specifically list 2

Simell as supposed instituted prior art and list the recovery 3 percentage that was disclosed in that reference. 4

So, it's our position that they're relying on it 5 affirmatively and asking the Board to rely on that data point in 6 reaching its conclusions, which is inappropriate given that Patent 7

Owner has had no response to respond -- excuse me, no 8 opportunity to respond to that argument because it appeared for 9 the first time in the reply. Nevertheless, it's Patent Owner's 10 position that Simell also does not provide any relevant 11 conclusions about PAA to PAGN conversion. I think, 12 unfortunately, it's kind of hard to see, but the Simell reference 13 actually shared a co-author in Dr. Brusilow, which we have 14 mentioned repeatedly today, and data from 1986, so five years 15 before his 1991 experiment, which is Petitioner's primary prior art 16 reference. 17

Now, Simell is a unique experiment, and it differed 18 significantly from the Brusilow 1991 disclosure. Simell involves 19 five Finnish urea -- I'm sorry, lysinuric intolerance patients who 20 essentially don't have a typical urea cycle disorder but, rather, 21 have the difficulty retaining amino acids, and the deficiency in 22 amino acids for those patients results in a -- what's the word? -- 23 in a difficulty in operation of the urea cycle because essentially 24 they don't have the right building blocks to process and to make 25 urea. So, those patients sometimes can have episodes of 1 hyperammonemia but are typically not treated with PPA prodrugs 2 or nitrogen scavengers; instead, just a simple amino acid 3 supplement along with protein restriction works with those 4 patients. So, they are not typical urea cycle disorder patients. 5

Now, in addition, the experiment in Simell also was not 6 typical. The experiment involved artificial induction of 7 hyperammonemia in patients, alongside co-infusion of 8 nitrogen-scavenging medication, intravenous infusion. So, you 9 have a single dose of nitrogen scavenger being co-infused at the 10 same time as the hyperammonemia is being artificially induced. 11

Now, based on that study, they did measure urinary 12

PAGN. It was collected in the urine from those patients over 24 13 hours, but it's very unclear -- but it's unclear from the Simell 14 reference what, in fact, the overall recovery had been for those 15 patients. So, you'll see on page 120 of Simell, which is Exhibit 16

1005, that, in fact, the range of recovery for the patients was 17 between 15 and 110 percent, and it appears that they're 18 concluding that there's 40 percent actual recovery over 24 hours. 19

Now, in other parts of the paper, they mention a 51 20 percent and a 54 percent number, but there doesn't seem to be any 21 other support in the paper for that number. Also, the 40 percent 22 number appears to have excluded one patient who had markedly 23 higher PAGN conversion that exceeded the dose. So, again, this 24 is an incomplete data set. 25

The conclusions in Simell are also informative in this 1 regard because the conclusions lead us to believe that the 2 artificially induced hyperammonemia, alongside a single IV dose, 3 actually influenced the results in this case, and so the authors are 4 very careful to couch their conclusions in saying, you know, this 5 is what we got with this individual, unique experiment. We likely 6 would have had a different response if we had done something 7 else. So, let me just direct the Board to those conclusions. 8

Here we see on page 1121, the author says that 9 induction of hyperammonemia by protein intake, as opposed to 10 the artificial alanine infusion in this instance, probably leads to 11 more even distribution of nitrogen in metabolic compartments 12 and, consequently, may be associated with better responses to 13 treatment with the ACAT (phonetic) agents. So, what they're 14 saying there is essentially you would have gotten likely more 15 conversion, more response, had the hyperammonemia been 16 naturally occurring based on intake of protein rather than infusion 17 of alanine. 18

Similarly, they also conclude that because they only 19 dosed a single dose, that that also affected the experiments. They 20 say during prolonged administration, the percentages are 21 probably higher, and, again, referring to the fact that it was just a 22 single dose here. Dr. Sondheimer also makes the point that -- 23 based on Sherwin that dosing and frequency of dosing also 24 matters. 25

So, at slide 20 of Petitioner's demonstratives today, they 1 have a callout that specifically comes from Sherwin '19 which 2 they rely on in combination with Brusilow that says, it's probable 3 that more of the phenylacetylglutamine would have appeared in 4 the urine after each dose of the acid, had the acid been ingested at 5 regular intervals covering a period of 10 or 12 hours. 6

So, what we can take away from the Simell experiment 7 is merely that under the conditions that were studied, which were 8 unique and differed significantly from the Brusilow conditions, 9 that conversion in those instances appeared to range between 15 10 and 110 percent for those patients, but we can't take any overall 11 conclusions for the broader urea cycle disorder population from 12 that study. 13

And I'd also just like to mention that Brusilow was a 14 co-author on the Simell paper, and five years later, when he wrote 15 his Brusilow '91 paper, which is a primary prior art reference 16 here, he doesn't refer or even cite to Simell. He doesn't even cite 17 to that data or indicate that it has any relevance on his overall 18 conclusions about conversion of PAA to PAGN in urea cycle 19 disorder patients. 20

So, if I may, I'd like to turn to the '215 patent. Prior to 21 the '215 patent and prior to 2011, the prior art deemed treatment 22 of urea cycle disorder patients to be satisfactory and their blood 23 ammonia levels to be at safe levels if they were below 80 24 micromoles per liter, and that's specifically referenced in 1

Fernandes, which is Petitioner's primary prior art reference. 2

Eighty micromoles per liter is a level much higher than 3 the upper limit of normal; however, patients who were deemed to 4 have satisfactory ammonia levels still suffered from incidences of 5 unpredictable hyperammonemia or acute hyperammonemia 6 where their blood ammonia levels would go to dangerously high 7 levels, which could result in neurological damage, coma, and 8 even death in these patients. 9

So,…