Astrazeneca AB, et al v. Apotex Corporation, et al
Filing
240
OPINION AND ORDER. Following trial, Astra has carried its burden of showing that it is entitled to damages in the amount of $76,021,994.50 plus pre-judgment interest. The parties shall present a proposed judgment by December 6, 2013. (Signed by Judge Denise L. Cote on 11/26/2013) (gr)
UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF NEW YORK
-------------------------------------- X
:
ASTRAZENECA AB, et al.,
:
:
Plaintiffs,
:
:
-v:
:
APOTEX CORPORATION, et al.,
:
:
Defendants.
:
:
-------------------------------------- X
APPEARANCES:
For plaintiffs
John W. Treece
David C. Giardina
Courtney A. Rosen
Angelo J. Suozzi
Zachary A. Madonia
Sidley Austin LLP
1 S. Dearborn St.
Chicago, IL 60603
Joshua E. Anderson
Sidley Austin LLP
555 W. Fifth St.
Los Angeles, CA 90012
Paul J. Zegger
Sidley Austin LLP
1501 K Street NW
Washington, DC 20005
For defendants
Robert B. Breisblatt
Brian J. Sodikoff
Stephen P. Benson
Dennis C. Lee
Katten Muchin Rosenman LLP
525 W. Monroe St.
01 Civ. 9351 (DLC)
OPINION & ORDER
Chicago, IL 60661
David A. Crichlow
Rebecca Kinburn
Katten Muchin Rosenman LLP
575 Madison Ave.
New York, NY 10022
TABLE OF CONTENTS
PROCEDURAL HISTORY ............................................ 4
BACKGROUND .................................................... 7
I. Omeprazole .............................................. 10
II. The First Wave Litigation .............................. 12
III. The PPI market in 2003 ................................ 14
A. Generic launches ...................................... 14
B. Astra’s Nexium® strategy .............................. 19
C. Third Party Payers .................................... 24
D. Price Comparison of Nexium® and Omeprazole: December
2002 to November 2003 ..................................... 30
E. Status of Market in November 2003 ..................... 38
IV. Atra’s Licensing Practices as of November 2003 ......... 43
A. Licenses for other PPIs ............................... 44
B. Prilosec OTC® ......................................... 46
V. Apotex’s Alternative Formulations ....................... 52
A. Changes to Apotex’s formulation ....................... 57
B. Copying other formulations ............................ 60
C. Microtablet formulation ............................... 62
VI. The Book of Wisdom: Post-entry Information ............ 64
A. Nexium® ............................................... 65
B. Prilosec OTC® ......................................... 69
C. Apotex’s Sales ........................................ 71
D. Settlements ........................................... 73
1. Andrx offer.......................................... 73
2. Teva settlement...................................... 75
E. Infringement Litigation ............................... 77
DISCUSSION ................................................... 79
I. The Isolated Value of the Subcoating .................... 82
II. The Pediatric Exclusivity Period ....................... 85
III. Standing .............................................. 90
IV. The Reasonable Royalty ................................ 101
A. Apotex’s position .................................... 103
1. Apotex’s profits.................................... 103
2
2. Apotex’s ability to avoid infringement..............
B. Astra’s position .....................................
C. Two remaining factors: Factors 2 and 14 .............
CONCLUSION ..................................................
105
115
126
127
DENISE COTE, District Judge:
This is the last of several patent infringement actions
consolidated before this Court concerning generic production of
omeprazole, commonly known by its brand name, Prilosec®.
Plaintiffs’ Astrazeneca AB, Aktiebolaget Hassle, KBI-E Inc., KBI
Inc., and Astrazeneca LP (collectively, “Astra”) patent on the
omeprazole molecule expired in 2001, but several patents
covering the formulation of the drug, including the patents at
issue in this case, did not expire until 2007.
Beginning in
1997, anticipating the expiration of the molecule patent, eight
generic drug manufacturers, including defendants Apotex Corp.,
Apotex, Inc., and TorPharm, Inc. (collectively, “Apotex”), filed
Abbreviated New Drug Applications (“ANDAs”) with the FDA,
seeking permission to manufacture and sell omeprazole.
Infringement litigation ensued.
Apotex, Canada’s largest generic pharmaceutical company,
began selling its generic omeprazole in November 2003, during
the pendency of the litigation, and continued selling until
2007, when it was found to infringe Astra’s patents.
3
The only
remaining issue is the measure of damages to which Astra is
entitled for over three years of infringing sales.
The parties
have agreed that these damages are to be based on a reasonable
royalty for the use made of the patents, which this Court must
set by imagining a successful hypothetical licensing negotiation
between Astra and Apotex in November 2003, on the eve of
Apotex’s launch.
PROCEDURAL HISTORY
Fact and expert discovery in this action concluded on
August 16, 2013.
The parties’ Joint Pretrial Order, proposed
findings of fact and conclusions of law, and pretrial memoranda
were submitted on September 13.
At the time the trial was
scheduled, the parties agreed that a bench trial would resolve
Astra’s outstanding damages claim.
With the parties’ consent,
the trial was conducted in accordance with the Court’s customary
practices for non-jury proceedings, which includes taking direct
testimony from witnesses under a party’s control through
affidavits submitted with the pretrial order.
The parties also
served with the Joint Pretrial Order copies of all exhibits and
deposition testimony that they intended to offer as evidence in
chief at trial.
4
At trial, Astra called two fact witnesses and five experts.
Astra’s fact witnesses were Kenneth E. Graham, Sr. (“Graham”),
an Astra employee for 21 years who held the title of Brand
Leader for Nexium®, among others; and Mark Uhle (“Uhle”), CFO
for AstraZeneca Pharmaceuticals LP.
Astra’s expert witnesses
were Dr. Martyn Davies (“Davies”), Professor of Biomedical
Surface Chemistry at the University of Nottingham in the United
Kingdom; Dr. David T. Lin (“Lin”), formerly a Team Leader in the
FDA’s Division of Reproductive and Urologic Drug Products;
Robert Navarro (“Navarro”), Clinical Professor in the Department
of Pharmaceutical Outcomes & Policy at the University of Florida
College of Pharmacy; Dr. Gordon Rausser (“Rausser”), Robert
Gordon Sproul Distinguished Professor at the University of
California, Berkeley; and Dr. Christine S. Meyer (“Meyer”), Vice
President at National Economic Research Associates, Inc.
Affidavits submitted by Astra constituted the direct testimony
of its fact and expert witnesses.
Each of these witnesses
appeared at trial and was cross-examined.
Astra also offered excerpts from the depositions of Dr.
Bernard Sherman (“Sherman”), Chairman and CEO of Apotex, Inc.;
Beth Hamilton (“Hamilton”), National Sales Director for Apotex
Corp.; Gordon Fahner (“Fahner”), Vice President of Business
5
Operations and Finance at Apotex, Inc.; and Tammy McIntire
(“McIntire”), president of Apotex Corp.
Apotex offered counter-
designations as to Sherman, Hamilton, Fahner, and McIntire.
During the presentation of its defense, Apotex presented
affidavits constituting the direct testimony of two fact
witnesses and five experts.
Apotex’s fact witnesses were Dr.
David Beach (“Beach”), formerly president of TorPharm, a
division of Apotex Pharmaceuticals and Bernice Tao (“Tao”),
formerly Director of Global Regulatory Operations at Apotex,
Inc.
Apotex’s expert witnesses were Dr. Stephen W.
Schondelmeyer (“Schondelmeyer”), Professor of Pharmaceutical
Management and Economics at the University of Minnesota; Roy
Weinstein (“Weinstein”), Managing Director at Micronomics; Dr.
Jeffrey Leitzinger (“Leitzinger”), Managing Director at Econ One
Research, Inc.; Gordon R. Johnston (“Johnston”), formerly Deputy
Director of the FDA’s Office of Generic Drugs; and Dr. Cory
Berkland (“Berkland”), Professor of Pharmaceutical Chemistry and
of Petroleum Engineering at the University of Kansas.
Affidavits submitted by Apotex constituted the direct testimony
of its fact and expert witnesses.
Each of these witnesses
appeared at trial and was cross-examined.
6
Apotex also offered excerpts from the depositions of Lisa
Schoenberg (“Schoenberg”), Vice President of Sales and Marketing
for AstraZeneca Pharmaceuticals LP; and Jeanette Tremayne
(“Tremayne”), Director of Alliance Management Finance for Merck
Sharpe & Dohme, the parent of defendant KBI Inc.
Astra offered
counter designations as to both Schoenberg and Tremayne.
The nonjury trial was held in this action from November 4
to 19, 2013.
This Opinion presents the Court’s findings of fact
and conclusions of law.
The findings of fact appear principally
in the following Background section, but also appear in the
remaining sections of the Opinion.
This Opinion concludes that
Astra is entitled to a reasonable royalty for Apotex’s
infringement of the Patents in the amount of 50% of Apotex’s
profits on its infringing sales or $76,021,994.50, plus prejudgment interest.
BACKGROUND
The evidence at trial has emphasized several key areas of
dispute between the parties that would have been crucial in a
November 2003 licensing negotiation.
First, Apotex’s evaluation
of the value of a license to sell omeprazole would have been
largely shaped by the condition of the market in November 2003
7
for the class of drugs in which omeprazole belonged.
was called proton pump inhibitors or PPIs.
This class
PPIs address
problems associated with gastric distress related to the
production of gastric acid.
Entry into the PPI market, the
evidence has shown, remained an extremely valuable opportunity
in November 2003, despite the presence of several branded PPIs,
three other generic manufacturers of omeprazole, an over-thecounter PPI branded as Prilosec OTC®, and the increasing
importance of Astra’s next generation branded PPI, Nexium®.
Second, Apotex would have been acutely aware, as it approached
the negotiating table in November 2003, of the difficulties it
faced if it attempted to develop a new, non-infringing
alternative formulation.
While one other generic manufacturer
had already shown that it was indeed possible to commercialize
omeprazole without infringing Astra’s formulation patents,
Apotex did not have a non-infringing formulation ready in
November 2003 and would have had little confidence that it could
develop one without substantial delay.
Since earlier generic
products have a significant advantage over later entrants,
Apotex would have been keenly motivated to avoid that delay of
entry.
8
Astra too would have had a core set of concerns as it
approached a hypothetical negotiation.
While its patent on the
omeprazole molecule had expired and it was in the midst of
executing a strategy to promote Nexium®, Astra still earned
substantial profits on sales of Prilosec®.
Moreover, the price
of generic omeprazole had not been substantially eroded by the
entry of three generic omeprazole products in the months leading
up to November 2003, and Astra would have been concerned that
Apotex, with a license in hand, would have cut prices to gain
market share, further eating into Astra’s profits on Prilosec®.
Of perhaps even more concern to Astra was the impact the Apotex
sales might have had on the growth of Nexium®.
Astra offered
aggressive rebates on Nexium® to ensure that its cost to Third
Party Payers (“TPPs”), such as pharmacy benefit managers
(“PBMs”), did not rise relative to its principal competitors,
which included not only the other branded PPIs but also generic
omeprazole.
It believed that without a competitive price,
Nexium® would fall out of favor with the various entities that
help determine how commonly various drugs are dispensed, and at
what cost to consumers.
Astra also did not have a practice of
licensing its branded products to generic manufacturers, and
would have wanted to maximize its profit on any license it
9
granted to Apotex.
These would have been the key motivations
and concerns of the parties as they approached a November 2003
negotiation, and they have been the key areas of dispute at
trial.
I.
Omeprazole
Omeprazole is in a class of drugs called proton pump
inhibitors or PPIs.
PPIs, which are indicated for many
gastrointestinal conditions, work by reducing the production of
gastric acid.
Omeprazole proved to be a difficult drug to
formulate, chiefly because it degrades in acidic environments
like the stomach and yet is most effective when it passes
through the stomach and is absorbed in the small intestine.
In
re Omeprazole Patent Litig., 222 F. Supp. 2d 423, 434-35
(S.D.N.Y. 2002) (“Omeprazole I”).
Astra’s scientists also found
that omeprazole had stability issues, and had difficulty
developing a formulation with an adequate shelf-life.
Id.
Astra ultimately solved these problems with a formulation
that combined three key elements: (1) a core combining the
active pharmaceutical ingredient with an Alkaline Reacting
Compound or “ARC”, which provided the necessary stability; (2)
an outer, enteric coating that protected the drug from the
acidic environment of the stomach but degraded in the intestine,
10
and (3) a water soluble inert subcoating to separate the core
from the enteric coating.
Id. at 393.
These elements are
combined in small spheres, which are inserted into capsules of
10, 20, and 40 milligram strength.
In 1987, Astra applied for patents on this formulation, and
received the patents-in-suit (the “Patents”).
‘230 patents claim these three elements.
Both the ‘505 and
The ‘505 patent
includes the formulation and the omeprazole compound; the other
patent covers a class of benzimidazole compounds, including
omeprazole, and their salts, rather than just omeprazole itself.
Id. at 445.
When Astra received FDA approval to sell omeprazole in the
United States in 1989 it was the first PPI on the market.
Astra
considered omeprazole, sold under the brand name Prilosec®, 1 the
“new global standard in the short and long term treatment of
acid related diseases,” and doctors similarly regarded it as the
gold standard among PPIs until Astra’s next PPI, Nexium®,
entered the market in 2003.
Prilosec® was a blockbuster drug,
with extraordinary sales success.
As of 2001, Prilosec® was the
“most-prescribed anti-secretory product in the U.S. and in the
1
For a brief time, Prilosec® was first sold under the brand name
Losec®.
11
world, and . . . the largest prescription pharmaceutical product
ever in terms of sales.”
II.
The First Wave Litigation
Anticipating the expiration of Astra's patent on the
omeprazole molecule, which occurred on October 5, 2001, generic
drug companies began filing ANDAs for omeprazole in 1998.
These
were accompanied by “Paragraph IV” certifications challenging
the validity of the ‘505 and ‘230 patents and/or asserting noninfringement.
Between May 1998 and April 2001, Astra filed
lawsuits against eight generic drug companies, and the lawsuits
were assigned to the Honorable Barbara S. Jones of this
district. 2
Judge Jones divided the defendants into two groups.
The First Wave defendants were Andrx Pharmaceuticals, Inc.
(“Andrx”); Genpharm Inc. (“Genpharm”); Kremers Urban Development
Co. and Schwarz Pharma, Inc. (collectively, “KUDCo”); and
2
Judge Jones resigned from the bench in January 2013 and the
patent litigation that remained before her relating to the '505
and '230 patents was assigned to this Court. At a conference on
February 6 and in an Order of February 8, a schedule was set to
bring the remaining litigation to a conclusion. Since that
conference, the two other cases, Astra Aktiebolag v. Andrx
Pharmaceuticals, Inc., No. 99 Civ. 9887 (DLC), and Aai Pharma,
Inc. v. Kremers Urban Development Co., No. 02 Civ. 9628 (DLC),
have been resolved, both on the eve of trial. Thus, this
Opinion will resolve before the District Court the last of this
suite of cases.
12
Cheminor Drugs, Ltd., Reddy-Cheminor, Inc., and Schein
Pharmaceutical, Inc. (collectively, “Cheminor”).
The Second
Wave defendants were Impax Laboratories, Inc. (“Impax”); Mylan
Laboratories Inc., Mylan Pharmaceuticals Inc., Esteve Quimica,
S.A., and Laboratorios Dr. Esteve, S.A. (collectively, “Mylan”);
Lek Pharmaceuticals d.d. and Lek Services, Inc. (collectively,
“Lek”), and Apotex.
On October 16, 2002, Judge Jones issued an Opinion finding
that the Patents were not invalid and that three of the First
Wave Defendants -- all except KUDCo -- infringed the Patents.
Of particular importance to Apotex, Judge Jones held that the
Andrx product infringed the Patents based on the presence in the
Andrx product of an inert subcoating that formed in situ rather
than being applied.
Omeprazole I, 222 F. Supp. 2d at 528.
More
specifically, Judge Jones concluded based on testimony by
Astra’s expert that Andrx’s formulation led to the creation, in
situ and thanks to a chemical reaction, of a microscopically
thin salt layer between the core and the enteric coating, and
that this layer was a “subcoating” within the meaning of the
Patents, since it was inert, water-soluble, and formed an
isolating barrier between the core and the enteric coating.
13
Id.
at 538-41. 3
Later, during the Second Wave litigation, Judge
Jones would conclude that Apotex’s product also infringed the
Patents due in part to the in situ formation of a subcoating
underneath the enteric coating.
III.
The PPI market in 2003
A.
Generic launches
Three generics launched between the issuance of Judge
Jones’s First Wave Opinion in October 2002 and Apotex’s launch
in November 2003: KUDCo, Mylan, and Lek.
KUDCo, whose
formulation had been found to be non-infringing, id. at 433,
launched its omeprazole product in December 2002.
KUDCo was the
first generic on the market, and enjoyed a 180 day period of
exclusivity in that role. 4
Despite this advantage, however,
3
When it affirmed Judge Jones’s Second Wave Opinion in 2008, the
Federal Circuit described the function of the inert subcoating
in Astra’s formulation of omeprazole. While the enteric coating
protects omeprazole from gastric acid in the stomach, the court
noted, it contains acidic compounds that can cause the
omeprazole in the drug core to decompose while the drug is
stored. In re Omeprazole Patent Litig., 536 F.3d 1361, 1365
(Fed. Cir. 2008) (“Omeprazole V”). To increase storage
stability, the ARC is added to the drug core, but the ARC can
compromise the enteric coating. This problem was solved with
the addition of an inert subcoating that rapidly disintegrates
in water, increasing storage stability while also preventing
omeprazole from degrading in the stomach. Id.
4
KUDCo had not been the first to file an ANDA. It purchased its
right to sell during the exclusivity period from its competitors
14
KUDCo did not have the manufacturing capacity to supply the full
needs of the market immediately. 5
For this and the reasons that
commonly apply to generics selling during the exclusivity
period, KUDCo kept the price of its omeprazole product high. 6
Lek and Mylan, on the other hand, were in the same position
as Apotex.
They were Second Wave defendants and the court had
not yet ruled on Astra’s infringement claims against them.
Nevertheless, Lek and Mylan made the decision to launch their
products in August 2003, knowing they were “at risk” of later
being held to infringe.
Facing the uncertainty of potential
liability to Astra, they did not cut prices aggressively. 7
As
both Dr. Rausser and Dr. Schondelmeyer explained at trial, the
Andrx and Genpharm, which held the right jointly.
5
In a December 9, 2002 press release, KUDCo reported that it had
the capacity to supply about 50% of the omeprazole market. On
April 1, 2003, KUDCo reported that it was able to supply about
70% of the market and that it was divesting of certain products
to “accelerate redirection of manufacturing resources to . . .
omeprazole.”
6
As an internal Astra email in January 2003 noted, “almost all
managed care plans are unimpressed with the price point of
KUDCo’s generic. This has caused some to wait to make changes
to their formulary until additional generics become available
and price point drops.”
7
In her Second Wave Opinion, Judge Jones eventually concluded
that Mylan and Lek did not infringe the Patents, while Apotex
did. In re Omeprazole Patent Litig., 490 F. Supp. 2d 381, 390
(S.D.N.Y. 2007) (“Omeprazole III”).
15
uncertainty about infringement that accompanies an at-risk
launch makes an at-risk generic drug manufacturer hesitant to
cut prices.
The parties vigorously dispute the degree to which these
three prior launches impacted the market.
Astra suggests that
all three kept their prices high relative to Prilosec® and that
Apotex, if it had obtained a license from Astra, would have
transformed the generic market by launching at a significantly
lower price.
Apotex argues that prices for generic omeprazole
were already beginning to erode, and that the competition among
the three generics would have continued to erode prices even
without the entry of a licensed fourth product.
The record suggests that both are right to a degree.
The
prices of generic omeprazole had declined, but not
significantly.
An Astra review of the market in October 2003
noted that with Mylan and Lek having launched and KUDCo no
longer “restrained” by problems with its production capacity,
“price pressures” were increasing and “Managed Care formularies
are in a state of flux.”
The report went on to conclude,
however, that thus far, the “high price of generics has limited
therapeutic substitution.”
Similarly, an Astra email dated
October 29, 2003 noted that “price competition in the generic
16
market has increased and both Mylan and KUDCo are offering
steeper discounts,” but then observed that “this has not caused
a change in [managed care] behavior.”
The entry of generic competition for Prilosec®, however,
had a profound impact on the sale of Prilosec®.
The market
share of Prilosec®, as measured by the total number of
prescriptions written, had been steadily declining since at
least 2001, when Nexium® was introduced.
It dropped
precipitously, however, in late 2002 as KUDCo’s generic
omeprazole product came on the market, before settling back into
a steady decline in 2003.
The following graph from Astra’s quarterly performance
review in April 2004 illustrates the shifting market shares of
Prilosec®, branded PPIs and generic omeprazole from March 2001
through early 2004.
While Prilosec® begins with the highest
market share in the prescription PPI market, it has the lowest
by the end of the period.
Conversely, Nexium® steadily gains
market share, ending the period as the product with the second
largest share, trailing Prevacid®, a branded PPI competitor.
Generic omeprazole, which entered the market in late 2002,
17
quickly took market share from Prilosec®, rising to nearly 15%
market share for a period of time. 8
Taken as a whole, the evidence regarding the omeprazole
market as of November 2003 indicates that generic drug prices
remained relatively and uncharacteristically high.
This was due
to the fact that only one generic drug company was operating
freely and without the threat of litigation hanging over it.
8
A
“TRx EU” stands for “total prescriptions -- extended units” and
includes the mail-order market.
18
fourth generic, entering the market with a license in hand,
would therefore have had a golden opportunity to take
significant market share away from both other generic
manufacturers and perhaps even branded PPIs by launching at a
lower price.
B.
Astra’s Nexium® strategy
Anticipating the expiration of its patent on the omeprazole
molecule and an inevitable decline in its sales of Prilosec®,
Astra had long been planning the introduction of a new branded
drug that it hoped would take Prilosec®’s place as the gold
standard in the PPI market.
This drug became Nexium®, which
launched in March 2001, about six months before the expiration
of the omeprazole molecule patent.
In 1998, Astra summarized its strategy as the following.
Predicting that Prilosec® could lose up to 80% of its sales
volume within six months of the introduction of generic
omeprazole, it hoped to extend the period of market exclusivity
for Prilosec® through the strategic use of patents and to
transfer the maximum amount of Prilosec® business to a new
patent-protected product that would be introduced before the
patent protection for omeprazole expired.
19
It wanted that new
product to be differentiated from Prilosec® as a superior drug.
That new product was Nexium®.
Astra identified several challenges in implementing this
strategy.
They included the need to prove that Nexium® was
clinically superior to omeprazole and the need to convert the
Prilosec® business to Nexium® before the entrance of generic
competitors for the Prilosec® business.
Astra achieved the first of these goals.
Astra was
ultimately able to emphasize Nexium's clinical advantages over
omeprazole as part of a campaign to encourage physicians and
patients to see Nexium® as the most effective PPI on the market.
The FDA-approved label for Nexium® indicated it for “treatment
of Gastroesophageal Reflux Disease (GERD),” including healing
erosive esophagitis (for which it was indicated for one or
possibly two courses of short-term treatment of 4 to 8 weeks),
maintenance of healing of erosive esophagitis, and treatment of
symptoms associated with GERD, including heartburn.
The label
reports that Nexium® 40mg is more effective and faster at
achieving sustained symptom resolution for patients with erosive
esophagitis than omeprazole 20mg.
At its launch, Astra announced that
Nexium™ 40mg demonstrates higher healing rates in
erosive esophgitis than Prilosec® 20mg (the approved
20
dose for this indication). In three studies, Nexium™
40mg consistently demonstrated higher healing rates
than Prilosec® 20mg and two of these studies were
statistically significant. Healing rates achieved for
Nexium™ 20mg were between those achieved with Nexium™
40mg and Prilosec® 20mg.
In the 1990s, Astra developed another tactic as well to
help encourage the transfer of Prilosec® patients to Nexium®.
Astra planned to help Nexium® gain PPI market share by
introducing an over the counter version of omeprazole called
Prilosec OTC®.
In 2001, Astra predicted that PBMs would quickly implement
a maximum allowable cost (or “MAC”) after two to three generic
omeprazole products had entered the market.
A MAC limits the
amount that a pharmacy will be reimbursed for a drug; it is tied
to the price of a low-priced generic version of the drug, as
further described below.
Astra expected a MAC to be implemented
even sooner where the branded drug is a blockbuster, as was true
in the case of Prilosec®.
As a result, Astra anticipated that
any patients who remained loyal to Prilosec® would incur
substantial out-of-pocket costs in meeting the co-pays imposed
by their insurance plan.
Astra expected approximately half of
the PBMs to adopt mandatory generic substitution with the advent
of a single generic, and for most PBMs to do so as soon as the
second or third entrant appeared.
21
Thus, as of July 2001, Astra
expected the “full effects of genericization of omeprazole to
have occurred in the fall of 2003.”
This expectation was based
on research suggesting that approximately 40% of HMO enrollees
would be impacted by mandatory generic substitution during
KUDCo’s exclusivity period, while 100% would be impacted after
the expiration of KUDCo’s exclusivity period, starting seven
months after patent expiration. 9
Based on this expectation,
Astra timed the introduction of Prilosec OTC® for the Fall of
2003.
Astra believed before the launch of Prilosec OTC® that the
over-the-counter (“OTC”) and prescription markets were
fundamentally different.
In an internal analysis dated July 9,
2003, Astra examined the launch of Claritin as an over-thecounter product and concluded that the OTC and prescription
markets were “distinct,” in part because physicians tend not to
push OTC products, and OTC products tend not to appear on the
drug formularies of insurance providers.
Nevertheless, Astra
believed that Prilosec OTC® could help promote Nexium®, because
Prilosec OTC® would be viewed as similar to prescription
9
The document reporting these findings refers to Andrx’s
exclusivity period, because Andrx and Genpharm had not yet sold
their exclusivity to KUDCo.
22
omeprazole.
In other words, Astra believed that patients who
failed on Prilosec OTC® would proceed straight to Nexium®,
because it was “the only PPI proven better than Prilosec.”
Astra thus thought that it could “position OTC as a reason to
discourage prescribing of omeprazole and encourage prescribing
of the superior brand, Nexium®.”
The parties hotly debated at trial the degree to which the
appearance of Prilosec OTC® actually affected the prescription
PPI market.
The evidence shows that the advent of Prilosec OTC®
caused a substantial drop in the market share of generic
omeprazole and a further drop in the market share of Prilosec®. 10
Indeed, Prilosec’s net sales dropped from roughly $67 million in
September 2003 to roughly $28 million in October 2003 and $17
million in November 2003.
On the other hand, the presence of
Prilosec OTC® in the marketplace did not have any effect on
omeprazole pricing, because the systems through which
prescription and OTC drugs are paid for are largely separate, as
will be explained in more detail below.
10
As of November 2003,
Consistent with its strategy of transferring Prilosec®
business to Nexium, when it launched Nexium Astra stopped
marketing Prilosec® to doctors and consumers and stopped giving
free samples. Astra continued, however, to support the
wholesale price of Prilosec® through a rebate program in an
effort to keep it on insurance company formularies.
23
Astra’s overall plan regarding the transfer of the Prilosec®
business to Nexium® appeared to be working.
Nexium’s percentage
of the prescription PPI market continued to grow after the
introduction of Prilosec OTC®.
C.
Third Party Payers
Despite Astra’s strategy of promoting Nexium® as clinically
superior to other PPIs on the market, it nevertheless faced
significant price pressure as it promoted the new drug.
This
was in part due to the nature of prescription drug pricing and
the ways in which various entities other than the manufacturer
and the patient influence which drugs are sold, and at what
prices.
Some background is useful in understanding the complex
process by which drug prices are set.
In the decades following
the popularization of health insurance after the Second World
War, insurers had little role in the dispensing of prescription
drugs.
During this period, a doctor would prescribe a drug to a
patient, the patient would take the prescription to a pharmacy,
the pharmacy would fill the prescription as written, and the
patient would then submit the bill to her insurer, which would
pay the bill, often minus a 20% patient portion or co-pay.
24
This
system gave little room for insurers to affect the cost of
drugs.
By the early 1990s, this had changed, as managed care
organizations began to aggressively manage prescription drug
costs.
Managing such costs has two central components.
First,
the organizations seek to control or at least influence the
choice of drugs made by doctors, patients and even pharmacies,
and second, using that control as leverage, they extract
discounts for prescription drugs from drug manufacturers.
One methodology that third party payers or TPPs have
adopted for managing their reimbursement of prescription drug
costs is the creation of a “formulary,” or list of covered
drugs.
During the time at issue in this litigation, formularies
often placed drugs into three tiers.
Tier I was generally
reserved for generic drugs and required the patient to pay
nothing or the smallest co-pay.
Tier II usually contained the
preferred branded drug and required a higher co-pay by the
patient.
Tier III usually included non-preferred branded drugs
and required the highest co-pay.
certain drugs altogether.
25
“Closed” formularies excluded
Managed care health plans and PBMs 11 often use formularies
to encourage doctors and pharmacists to provide a patient with a
generic in place of the brand name drug or to prefer one branded
drug over another.
Health plans and PBMs can exert influence
over physicians by contractually requiring them to adhere to the
formulary policies, by educating them about the drugs on the
formulary, or by requiring pharmacists in a network to call
physicians to ask them to alter prescriptions that are
inconsistent with formulary policies.
Health plans and PBMs may
also have contractual arrangements with network pharmacies that
reimburse them only for the less expensive generic on the
formulary even if the pharmacy dispenses a branded drug.
When there is widespread acceptance of a generic drug and
confidence in its supply, health plans and PBMs may impose a MAC
or maximum allowable cost.
When a MAC is imposed, the pharmacy
will only receive reimbursement for a drug at the MAC price,
regardless of the pharmacy’s actual acquisition cost for the
11
PBMs provide administrative and management services to deliver
a prescription drug benefit. They may be part of a managed care
organization or a separate firm. By 1999, the top seven PBMs
estimated that they covered 234 million “lives”. As of 2006,
roughly two-thirds of outpatient prescriptions were covered by
commercial third party prescription programs, with the remainder
being covered by public programs such as Medicaid and Medicare,
and a smaller fraction covered by patients paying cash.
26
product.
A market is considered “genericized” at that point;
prices often fall by about 80% and the sales volume of the
branded drug collapses.
The MAC price may be chosen from the
federal Medicare program’s price for the drug or some other data
source.
Similar pressure is placed on a patient’s choice of a
drug therapy when the patient is required to pay a higher co-pay
when she obtains a non-preferred drug.
When negotiating prices
with a pharmaceutical manufacturer, a health plan or PBM can
threaten to exclude a drug it considers overpriced from a closed
formulary or to move the drug to a less preferred position on a
formulary in order to secure a discount.
As a general matter, the more costly a drug is to a third
party payer, the more likely the payer is to develop a strategy
to reduce costs.
If it does take steps that disadvantage the
drug, those steps may lead to a reduction in the sales of the
more costly drug compared to lower cost drugs in the same class
of pharmaceuticals.
As a result, branded drugs often have an
incentive to offer rebates to TPPs, either to ensure placement
on Tier II of a formulary or to ensure that they are not
excluded from a formulary altogether.
These rebates can vary
from 3% to 25% or more off the drug’s WAC or wholesale
27
acquisition cost. 12
If the branded drug offers favorable
rebates, it may be cheaper from the TPP’s perspective than a
generic equivalent, in which case the TPP is less likely to take
adverse action against the branded drug.
There are other ways a TPP can disadvantage a branded drug,
beyond simply changing its formulary status.
A “step therapy”
or “step edit” requirement mandates that patients first try a
certain preferred drug, like generic omeprazole, before doctors
can prescribe a disfavored, more expensive drug, like Nexium®.
If step therapy is imposed, the doctor must determine that the
patient is not responding adequately to the preferred drug
before the patient can move on to the disfavored drug.
Another
tactic, known as a “prior authorization requirement,” mandates
that doctors receive authorization from the TPP before they
prescribe the disfavored drug to a particular patient.
Generic versions of drug products are typically cheaper for
most third party payers, and TPPs generally attempt to drive a
culture of generic substitution across a formulary.
Even if a
generic is not cheaper than a branded drug for a third party
payer, it may be more profitable for pharmacies to dispense the
12
The WAC is a publicly available list price for drugs published
by the industry.
28
generic version.
Since TPPs cannot require pharmacies to
dispense a generic from a particular manufacturer, pharmacies
decide which manufacturer’s generic will be dispensed.
Pharmacists customarily stock a single generic product.
Pharmacies are thus able to negotiate significant discounts off
the published average wholesale price or AWP, 13 and in some cases
these may be as steep as an 80% discount.
These financial realities have a significant impact on the
pharmaceutical market.
Generic drugs are filling a growing
share of outpatient prescriptions.
They rose from roughly 30%
of the prescriptions filled in 1990 to over 70% in 2010.
Astra's own files provide an example of how formulary policies
can drive sales.
In 2001, Astra tried to understand how the
entry of generics had led to the rapid erosion of the sales of
another branded drug -- Prozac®.
This study emphasized the role
of pharmacists in driving the “aggressive conversion” to the
generic drug.
The research identified the most influential
factor as the restrictions placed by managed care institutions
13
The AWP during the relevant time period was typically 20 to
25% higher than the WAC for the same drug. A pharmacy was
usually reimbursed for branded drugs at a contracted percentage
discount from the AWP, such as 15%, and given a small dispensing
fee, such as $1.50. The amount of the co-pay would also be
subtracted from the reimbursement since the pharmacy would
collect the co-pay directly from the patient.
29
on pharmacy practices, either by requiring a higher co-pay for
the brand, refusing to include the brand on their formularies,
or imposing step-therapy. 14
D.
Price Comparison of Nexium® and Omeprazole:
December
2002 to November 2003
This system by which TPPs influenced drug prices had a
significant effect on the drugs at issue here.
Most importantly
for Astra, the emergence of generic omeprazole placed increasing
pressure on Astra to pay larger rebates to maintain a favorable
formulary position for Nexium®.
These rebates kept the
effective price of Nexium® comparable to that of omeprazole.
Indeed, there is evidence that the effective cost of Nexium®
therapy was, perhaps remarkably since Nexium® was the most
recently patented PPI, lower than the cost of omeprazole therapy
during the period from December 2002 through November 2003, at
least from the perspective of certain major TPPs.
This was due
to several factors, which included the relatively small
difference between the wholesale prices of the Nexium® and
14
The other key findings related to the high volume of branded
Prozac® (which made it a logical target for cost containment),
the perceived expensiveness of the branded drug, the greater
profit margin available to pharmacists from dispensing the
generic, and the unprecedented media attention given to the
launch of the generic drug.
30
omeprazole products, the manufacturer rebates Astra provided for
Nexium®, and the fewer number of pills necessary for a
successful course of treatment when Nexium® is prescribed.
Each
of these factors merits some further discussion.
The difference or spread between the wholesale price of
Nexium® and generic omeprazole in the interval between December
2002 (when the first generic as introduced) and November 2003
(when the hypothetical negotiation occurred) appears to have
been roughly a dollar or less.
The spread between the wholesale
acquisition cost or WAC, which is the price wholesalers pay to
pharmaceutical manufacturers like Astra, for these two products
ranged during that time frame from $0.66 to $1.05.
Similarly, the spread during this time frame of the average
wholesale price or AWP of these two products was relatively
small.
This is not surprising since during that time the AWP
was typically equal to 120% or 125% of the WAC.
Neither the WAC nor the AWP, however, necessarily provides
a complete measure of the relative prices between Nexium® and
generic omeprazole, since Astra paid significant discounts and
rebates to TPPs that had a significant impact on the actual
prices they paid.
Nexium® rebate data for 13 major third party
payers shows an average rebate of 14% of their gross sales in
31
the fourth quarter of 2002, rising to 17% in the fourth quarter
of 2003. 15
Another feature of any price comparison is the need to
consider the cost of a successful course of treatment.
One tool
that is commonly used for such a comparison is the Daily Average
Consumption or DACON.
On average, as of November 2003,
omeprazole patients consumed 0.12 pills more per day than
patients using Nexium®.
At this rate, if a third party payer’s
cost per pill of Nexium® were 13.9% higher than for omeprazole,
the true cost of treatment would be equal for the two drugs.
A study by Astra’s expert Dr. Rausser of pharmacy log data
from sixteen pharmacies in California and fourteen in
Massachusetts for all prescription PPI purchases demonstrated
that the effective cost of Nexium® therapy for those TPPs to
whom Astra offered rebates was often less than omeprazole
therapy during the period between December 2002 and November
15
This analysis does not include rebates generic omeprazole
manufacturers may have been giving to mail order pharmacies.
The data available to Dr. Rausser, who conducted this analysis,
on these rebates included only the “deepest available rebates,”
which did not allow him to determine what effect they might have
on average prices of omeprazole. Furthermore, the mail order
market accounted for only 13% of all U.S. drug sales in 2003 and
only 8.5% of Nexium sales from January to October 2003.
32
2003. 16
When DACON and rebate data were factored in, the median
third party payment for Nexium® was lower than that for
omeprazole in 78% of the cases in California, 17 and 89% in
Massachusetts. 18
Combining the data from the two states, there
were 149 instances in which matching sales appeared in the same
month for the same third party payer, which allows a direct
comparison of that TPP’s cost of Nexium® with omeprazole.
In
125 of those instances, or 84%, the median effective third party
payment for Nexium® was lower than that for omeprazole.
For
twelve out of the thirteen third party payers, Nexium® was less
16
Dr. Rausser drew these conclusions from a study of pharmacy
logs obtained by subpoenas issued during state court litigation
in California and Massachusetts in which Astra was responding to
class action claims that consumers and others had been misled
into believing that Nexium was superior to Prilosec® and were
injured by having to pay an unjustified price premium for
Nexium. At that time, Dr. Rausser used the data to support his
opinion that a large portion of the proposed class suffered no
economic harm because Nexium was a less expensive therapy.
17
The 78% figure reflects 61 out of 78 cases, that is, data for
the nine third party payers to the California pharmacies for
each month in a twelve month period between December 2002 and
November 2003 for which there was sufficient data.
18
The 89% figure reflects 63 out of 71 cases, that is, data for
the nine third party payers to the Massachusetts pharmacies for
each month in a twelve month period between December 2002 and
November 2003 for which there was sufficient data.
33
expensive than omeprazole in at least half of the months for
which there was sufficient data to make a comparison.
While Astra would not have had access to the pharmacy log
data Dr. Rausser used in November 2003, his study corroborates
Astra's own contemporaneous internal analysis.
A July 9, 2003
Astra analysis indicates that the net cost to a TPP of a 30 day
Nexium® prescription was over $18 less than the net cost of a
generic omeprazole prescription, i.e., $73.74 versus $91.90.
Astra attributes this differential to the closeness in the list
prices of the products and the co-pay differentials and rebates
provided by Astra.
The following table from Astra’s report
presents this data.
34
Apotex has challenged the reliability of the California and
Massachusetts data drawn from Dr. Rausser’s study of data from
thirty pharmacies.
Most of the thirty pharmacies which supplied
the data were parts of large chains or a mass merchandiser:
Wal-Mart, Longs and CVS in California; Wal-Mart, CVS and Target
in Massachusetts.
The logs typically include the date of the
transaction, the name and dosage of the drug, the number of
pills sold, any co-pay paid for insured transactions, and the
name of and the amount paid by any third party payer.
Dr.
Rausser calculated the third party payment per pill by dividing
the dollar amount of the recorded third party payment by the
quantity of pills sold.
Dr. Rausser next identified the twenty largest third party
payers responsible for the largest share of Nexium® sales and
was able to obtain rebate data for thirteen of the twenty.
These thirteen TPPs were significant participants in the
market. 19
They represented 82% of the insured omeprazole
19
Six of these thirteen represented 67% of the pharmacy benefit
manager or PBM market share in 2003. Six others were large
national PBMs or health plans, such as CIGNA and United
Healthcare. The thirteenth was a large regional health plan.
35
transactions, and over 95% of the Nexium® rebates paid by
Astra. 20
Apotex complains that the data is not drawn from a
It comes from only two states, 21
statistically reliable sample.
comes from only large volume buyers of Nexium®, and then only
those to which Astra paid rebates, and omits the mail order
segment of the market.
The data also has gaps, including a lack
of data from months for which no data existed for both drugs.
With these and other similar arguments, Apotex argues that Dr.
Rauser's conclusion -- that for many major third party payers
during the period from December 2002 through November 2003 the
cost of generic omeprazole therapy was greater than the cost of
Nexium® therapy -- is “invalid”.
While Apotex has correctly pointed out that there are
limitations in the pharmacy log data, and that there is a need
20
Dr. Rausser’s analysis compared the price of 20 mg omeprazole
with 40 mg Nexium, which he explained was proper because the
price of Nexium to both TPPs and patients was the same for both
20 and 40 mg doses. Dr. Rausser observed that there was much
more data in the pharmacy logs for 40 mg Nexium than for 20,
while the opposite was true for omeprazole.
21
Mr. Weinstein notes that the prices of omeprazole differ
significantly between the two states, underscoring the
unreliability of using data from just two states to make
generalizations about a regional or nationwide market.
36
to be careful in drawing lessons from it, it has not shown that
the data is unreliable or, to use its phrase, invalid.
In part
because it is so costly and time-consuming to procure and
analyze, it is unusual to have access to this rich vein of
information about third party payer and pharmacy pricing of the
relevant drugs during the relevant time period.
Astra has
succeeded in showing that this pharmacy data, particularly when
combined with other trial evidence, supports a conclusion that
for many TPPs during the period in which the hypothetical
negotiation was occurring, the cost of Nexium® therapy remained
quite close to that of treatment with generic omeprazole.
It must be remembered, however, that the proximity in price
to generic omeprazole was only one component in Astra’s
overarching strategy for promoting Nexium® and maintaining its
favorable formulary position.
Astra continued to position
Nexium® as a clinically-superior, next generation drug relative
to all of the other PPIs on the market, including the other
branded PPIs.
In a mid-2003 Nexium® Strategic Plan document,
Astra described its pricing strategy for Nexium® as follows.
The pricing strategy for Nexium has been one of
premium price due to its clinical differentiation and
heritage within the PPI class. The product has been
strategically priced at a discount to Prevacid and
branded Prilosec to help add incentive for conversion
from other PPIs once the clinical rationale for
37
utilizing the product has been established. Nexium
continues to rank third in listed WAC price with
Aciphex, generic omeprazole and Protonix rounding out
the class. [Astra] has continued to take nominal
price increases to keep in step with the competition.
Unlike Protonix, Aciphex and at times Prevacid,
[Astra] has not chosen to deep discount the product to
Managed Care.
Indeed, from the point of view of the consumer, Nexium® was
more expensive than generic omeprazole in 2003, reflecting its
higher co-payment requirements.
A majority of consumers paid
$10 or less for generic omeprazole and $20 or more for Nexium®.
This reflects the fact generic omeprazole had a preferred
formulary position.
Nevertheless, Astra’s official rebate
policies support Dr. Rausser’s overarching hypothesis that
Nexium® rebates were to some degree responsive to the price of
generic omeprazole.
Following the entry of generic omeprazole,
Astra increased its rebates on Nexium® and included the generic
product in calculations of market share for rebates that were
determined on that basis.
E.
Status of Market in November 2003
As of the date of the hypothetical negotiation of a license
for Apotex to sell generic omeprazole using the formulation
covered by the Patents, the PPI market was a huge, lucrative,
and competitive market.
As Dr. Sherman, Apotex’s CEO, noted in
a memo to Dr. Beach, Prilosec® was the “largest selling drug in
38
the U.S., with U.S. annual sales approaching $2 billion.”
In
planning for the expiration of the omeprazole molecule patent in
2001, Astra itself noted that “Prilosec® is the most attractive
market for generics ever.”
In projections dated June 18, 2003,
Apotex estimated that the prescription omeprazole market would
be worth a total of roughly $2.6 billion over the course the
year, and $1.4 billion in 2004.
Because the other branded PPIs
had launched much later than Prilosec®, generic omeprazole was
the only generic PPI on the market and would be for years.
As of November 2003, Astra had two branded PPI products,
Prilosec® and Nexium®, introduced in 1989 and 2001,
respectively.
Three other branded PPIs were also successful
market participants and generic omeprazole had been on the
market for about a year.
Most recently, Prilosec OTC® had
entered the scene.
As of November 2003, pharmacies would have felt free to
dispense generic omeprazole instead of Prilosec® when filling
Prilosec® prescriptions, since generic omeprazole was an ABrated substitute for Prilosec®. 22
22
Prilosec®’s share of the
AB-rated drugs are FDA-approved bioequivalents of the branded
drug; they have been judged by the FDA to be therapeutically
equivalent. The FDA rated the generic omeprazole manufactured
by KUDCo, Mylan and Lek as AB-rated for Prilosec® even though by
39
market had dropped precipitously with the advent of generic
omeprazole, but Astra continued to support Prilosec® sales
through a rebate program and Prilosec® continued to provide a
significant stream of income to Astra.
In 2003, Astra sold $1.3
billion worth of Prilosec® with $437 million in net sales
adjustments (which includes rebates), leading to $865 million in
net sales.
In that same year, Astra sold $3.2 billion worth of
Nexium®, with $770 million in net sales adjustments and $2.5
billion in net sales.
Although three generic pharmaceutical manufacturers had
begun selling omeprazole, generic prices remained high.
No MAC
had been imposed by a health plan or PBM, a phenomenon which
would have led to a crash in omeprazole prices. 23
This led Astra
to conclude that it had not yet seen significant price erosion
and had not yet felt the full effects of cheap generic
omeprazole.
It was of course unclear how much longer these
three generics would maintain their high prices.
Nonetheless,
2007 it had been determined that none of these formulations
relied upon the patented technology.
23
There was no evidence at trial that a MAC was ever imposed for
omeprazole, reflecting perhaps that the litigation over many of
the generic omeprazole products took many years to resolve.
Decision makers are wary of imposing steep price constraints
when there may be a sudden deficit in supply.
40
Astra had every reason to expect that the launch of a fourth
generic, particularly for a licensed product, would swiftly
accelerate the decline in omeprazole prices and lead to the
destruction of the remaining Prilosec® market.
The launch of Prilosec OTC® in September 2003 was having a
significant impact on the overall PPI market as well.
As will
be discussed in more detail below, Astra calculated internally
that Prilosec OTC® had taken roughly 30% of the generic
omeprazole market.
Indeed, Astra documents show that the total
number of omeprazole prescriptions, which declined sharply just
after Prilosec OTC®’s launch, did not begin to recover until the
Fall of 2004, around the time Teva -- the fifth generic
pharmaceutical manufacturer to enter the market -- launched its
generic omeprazole.
An Astra “Prilosec OTC Update” dated October 21, 2003
reported that TPPs were beginning to take action against
omeprazole in light of Prilosec OTC®.
These actions included,
in certain cases, discontinuing reimbursements of prescribed
omeprazole, sending OTC coupons to PPI users, and moving PPIs to
the third tier of formularies.
While the principal impact of
Prilosec OTC® was on Prilosec® itself and generic omeprazole, it
also had an impact on the entire PPI market.
41
An outside report
later commissioned by Astra calculated that Prilosec OTC®
resulted in a 3% decline in Nexium® sales.
The Astra document
cited earlier in this paragraph anticipated that Prilosec OTC®,
combined with multiple generics, would affect Nexium® by
reducing gross sales by $55 million in 2003 and $398 million in
2004.
Finally, as of November 2003, Astra remained committed to
an expensive rebate program to support Nexium® prices, in the
hope that this price support would preserve Nexium’s favorable
formulary status and avoid implementation of step therapy or
other adverse actions.
Since formulary decisions are made in
the context of an entire class of drugs, the prices, supply and
therapeutic profile of each drug in the class are in play.
Astra’s rebate strategy was therefore aimed directly or
indirectly at all of its PPI competitors and the interrelated
and evolving positions of those competitors regarding supply,
price, and any other competitive advantage.
But, since the only
new competitor in the prescription PPI market in 2003 was
generic omeprazole, a principal consideration in Astra’s
adjustment of its rebate strategy was the impact that the entry
of generic omeprazole was having on drug formulary practices and
pharmacies.
42
Thus, as of November 2003, any new entrant that altered
that competitive landscape in a material way would impose upon
Astra the burden of increasing its effort to compete for
favorable formulary treatment.
Astra’s primary tools in that
effort would have been its hefty rebate program and its ongoing
effort at drug differentiation.
IV.
Atra’s Licensing Practices as of November 2003
There is no evidence that Astra would have initiated or
encouraged licensing discussions with Apotex.
Astra had been
defending the Patents for more than five years, and would
continue to do so for another ten years.
It had no active
program for licensing any of its patented prescription drugs to
generic pharmaceutical companies or history of doing so.
Astra traditionally markets its brand name prescription
drug products to doctors and third party payers by explaining
that it is the sole source of the product.
Its general policy
is not to license others to use its inventions to sell products
that will be sold in direct competition with its patented
products.
While Astra's patent for omeprazole had expired in
2001, its formulation patents for Prilosec® did not expire until
2007.
43
In any hypothetical negotiation of a license to a generic
manufacturer, Astra would have considered the opportunity in
terms of its expected net present value.
In other words, in
making such judgments it discounts future cash flows to what
they would be worth presently, taking into account the
probability of receiving the future cash flows.
Astra would
also consider the anticipated cannibalization of the revenues
from the branded product and other Astra products.
If the
expected net present value of a license is above zero, Astra's
financial team recognizes that it is up to management to
determine whether that value is sufficient in light of the
potential impacts of the license, the expected return on
investment and other financial metrics, as well as the license’s
overall alignment with corporate strategy.
A.
Licenses for other PPIs
Apotex has emphasized, however, that Astra did enter
licensing agreements that encompassed the Patents, and that it
did license a compound that used the valuable Prilosec®
trademark and that competed directly with Prilosec®.
Thus,
while Astra never licensed the Patents to a company selling
generic prescription omeprazole products, between 1994 and 1998,
Astra had entered into four agreements that involved the
44
Patents.
The most significant of these for the purposes of
these proceedings -- the license for Prilosec OTC® -- will be
discussed last.
In 1994, Astra entered into an agreement with Takeda
Chemical Industries, Ltd. (“Takeda”), the maker of the PPI
Prevacid®, that included the licensing of the ‘230 patent.
The
agreement resolved legal proceedings in a number of countries in
which the two companies had accused each other of infringing the
other’s PPI patents and provided cross-licenses to those
patents.
Takeda was required to pay Astra a small royalty for
its U.S. sales -- 2.5% -- which was cut to 1% after the
expiration of the patent on the omeprazole molecule.
The
agreement did not permit Takeda to sell omeprazole.
In 1996, Astra entered into an agreement with Byk Gulden
Lomberg Chemische Fabrik GmbH (“Byk”), which had developed the
PPI Protonix®.
The agreement resolved infringement actions
Astra had brought against Byk in several countries, as well as
proceedings Byk had brought seeking to invalidate Astra patents.
The agreement granted Byk a royalty-free worldwide license under
certain patents, including the patents-in-suit, to make, use,
and sell Protonix®.
Byk agreed to drop all proceedings in which
45
the validity of an Astra patent was called into question.
The
agreement did not permit Byk to sell omeprazole.
In 1996, Astra also entered into an agreement with Eisai
Co. Ltd. (“Eisai”), which had developed the PPI Aciphex®.
Under
the agreement, both parties agreed to drop challenges to each
others’ patents and exchanged cross-licenses allowing the use of
their patents (including, in Astra’s case, the patents-in-suit)
in connection with the manufacture and sale of their respective
PPIs.
Eisai agreed to pay Astra a royalty of .1% on its U.S.
net sales of Aciphex®.
Eisai was not, however, permitted to use
the Patents to make or sell omeprazole.
B.
Prilosec OTC®
Prilosec OTC® was launched in September 2003, two months
before the hypothetical negotiation, pursuant to a six year old
licensing agreement between Astra and Procter & Gamble (“P&G”).
The parties hotly contest whether the licensing terms for this
product provide an appropriate benchmark for the hypothetical
license between Astra and Apotex.
The product Prilosec OTC® arose from a November 20, 1997
agreement between Astra and P&G, an agreement that was amended
in August of 1999.
Their agreement established a framework for
the cooperative development, Astra’s supply, and P&G’s marketing
46
of Prilosec OTC®.
Under the agreement, P&G paid Astra a base
royalty of 7% of net sales for the first twenty years, to be
increased based on sales volume to a royalty rate as high as 40%
in the first three years after launch and 20% thereafter.
Astra
expected these sales targets to be met and they were, meaning
that P&G indeed paid Astra 40% on a portion of its net sales on
Prilosec OTC® for three years.
For example, in October 2003,
Astra reported that P&G had not changed its expectations for the
first year sales, which stood at between $200 to $400 million.
P&G also paid up-front milestone payments of $56 million
and invested hundreds of millions of dollars to develop and
market the product, although Astra was also required to shoulder
certain development costs.
In addition, Astra allowed its
highly valuable trademarked Prilosec® brand name to be used for
Prilosec OTC®, and P&G was required to use its “best efforts” to
market and sell Prilosec OTC®.
Astra was also required to
supply the drug to P&G essentially at cost.
At trial, Uhle
explained that the royalty payments made by P&G ended up
amounting to a blended rate of approximately 20% of its net
sales over the course of the first three years Prilosec OTC® was
on the market, or 23% of net sales if the upfront milestone
payments are included.
47
As its name suggests, Prilosec OTC® does not require a
prescription.
In contrast to the active ingredient in
Prilosec®, omeprazole, the active ingredient in Prilosec OTC® is
omeprazole magnesium.
Prilosec OTC® is indicated for 14 days of
treatment of heartburn, which is an approved indication not
shared by the other PPIs discussed here.
Nexium® and Prilosec®
were both indicated for four to eight weeks’ treatment of GERD.
Prilosec OTC® is only available in a 20mg formulation, while
Prilosec® is available in 10, 20 and 40mg formulations and
Nexium® in 20 and 40mg formulations.
The broader strategic underpinnings of Prilosec OTC® from
Astra’s perspective and the eventual fate of Prilosec OTC® and
its effects on the PPI market are of crucial importance to the
parties for two reasons.
First, Astra’s expectations about
Prilosec OTC® help determine how comparable its license with P&G
was to the license at issue in this case.
In other words, if
Astra expected Prilosec OTC® to directly compete with
prescription Prilosec®, or with Nexium®, then the terms of its
license are more helpful in analyzing a license with Apotex.
Second, Astra’s strategy for Prilosec OTC® says a lot about how
it evaluated the prescription omeprazole market in the mid to
late 1990s and how it expected that market to fare over the
48
coming years.
Thus, even independent of the precise royalty
rate Astra negotiated with P&G, Astra’s plans for Prilosec OTC®
help shed light on what it would have viewed as the risks of
granting a license to Apotex.
As already noted, within the pharmaceutical industry the
OTC and prescription markets are usually viewed as distinct
consumer markets.
The industry assumes that physicians tend
“not to push” OTC products, and it was rare in 2003 for health
plans and PBMs to provide insurance reimbursement for OTC
products. 24
Prilosec OTC® was priced at $0.71 per pill, which
made it a more expensive treatment option for most consumers
than generic omeprazole, which was often available to insured
patients at a co-pay of $10 or less for a month’s prescription. 25
24
Navarro only learned of one plan that provided coverage of
Prilosec OTC® as of late 2003. Moreover, the pharmacy log data
analyzed by Dr. Rausser for the period between the launch of
Prilosec OTC® and November 11, 2003 shows 3,975 PPI purchases of
Nexium reimbursed by a third party payer but none covering
Prilosec OTC®.
25
Apotex argues that the course of treatment price comparison
should factor in the cost to a patient of the doctor’s visit
that would be necessary to obtain the prescription for generic
omeprazole. For any individual, however, there may be a host of
factors that lead to a doctor’s visit and among all patients
there is a large range in the cost of making such a visit.
There is no simple mathematical formula that can be applied
here, and Apotex has not provided one.
49
As a consequence, Astra initially thought that Prilosec OTC®
would target “self-treaters” who pay cash for their medicine and
that it would compete against other OTC heartburn products
called histamine antagonists.
Nevertheless, Astra felt that Prilosec OTC® could help it
promote Nexium®.
More specifically, Astra’s plan was to
“position OTC omeprazole as a reason to discourage prescribing
of omeprazole and encourage prescribing of the superior brand,
NEXIUM.” 26 “If Prilosec is available OTC,” Astra observed in a
2002 Product Strategic Plan for Prilosec®, after a person has
used Prilosec OTC® and failed to get the relief she needs,
“physicians will want to write Nexium more because it is the
only PPI proven better than Prilosec.”
As of 2002, of course,
Prilosec® was still providing “significant revenue” to Astra and
was viewed as “the launch platform for Nexium.”
Astra therefore expected Prilosec OTC® to cause significant
erosion of the prescription omeprazole market, and it positioned
the product accordingly.
Prilosec OTC® was marketed to both
26
An internal “long-term vision statement for the
gastrointestinal area at Astra” drafted in October 1998
indicated that Prilosec® could lose up to 80% of its sales
volume within the first three to six months of generic
introduction. The document went on to analyze Astra’s strategy
of converting from Prilosec® to Prilosec OTC® and Nexium.
50
consumers and physicians as an equivalent to omeprazole.
Indeed, Prilosec OTC®’s label explained that 20.6mg of Prilosec
OTC® was “equivalent to” 20mg of omeprazole.
As Graham
explained at trial, these efforts were part of Astra’s broader
strategy of cannibalizing Prilosec® to promote Nexium®.
In October 2003, about a month after launch, Astra
estimated that Prilosec OTC® had captured 30% of the omeprazole
prescription market, including both generic omeprazole and
Prilosec®.
Nexium® and Protonix® had gained market share in the
prescription PPI market as a result, and generics had increased
their price competition with each other.
There had been little
or no change, however, in the treatment of PPIs on formularies
and by PBMs generally.
Astra viewed these developments as “more
or less as expected.”
These market dynamics help place the Prilosec OTC® license
in proper perspective.
Astra’s strategy, conceived in the
1990s, at a time when it thought that the omeprazole market
would be “genericized” during 2003 and that Prilosec® would be
having great difficulty remaining competitive with generic
omeprazole, was to introduce Prilosec OTC® as a potent
competitor to generic prescription omeprazole and as a driver of
Nexium® sales.
For reasons already discussed, however, the
51
price of generic omeprazole had remained unexpectedly and
unusually high, and Prilosec® had retained a footing in the
market.
Astra was not surprised, therefore, in the Fall of 2003
to see Prilosec OTC® take market share from both Prilosec® and
prescription omeprazole.
Like the entry of generic omeprazole,
the entry of Prilosec OTC® further eroded Prilosec®’s volume.
Astra remained optimistic because it regarded Nexium® as the
future of its business in the PPI market, and Prilosec OTC® was
expected to help encourage the growth of Nexium®, not hinder it.
This is the environment in which Astra would have been
negotiating a license with Aptoex.
The advent of a licensed
generic omeprazole would have been a very different phenomenon
than the introduction of Prilosec OTC®.
While Prilosec OTC® had
not caused any substantial change in TPP behavior, Astra would
have expected a licensed generic omeprazole to use price to take
market share from its three generic predecessors.
Any material
decrease in the price of generic omeprazole would cause all TPPs
to reconsider their strategies for the PPI market and would
likely place pressure on Nexium®, causing Astra to increase its
rebate program in a way that Prilosec OTC® did not.
V.
Apotex’s Alternative Formulations
52
Another key factor in any November 2003 licensing
negotiation would have been the amount of potential revenue
Apotex stood to lose by walking away from the negotiating table.
Because Apotex did not have a non-infringing alternative
formulation ready for a November 2003 launch, any decision to
forgo a license would have necessitated a return to the drawing
board, delay, and uncertainty.
Just how much of its profits
over three and a half years of sales Apotex would have
sacrificed by attempting to develop an alternative formulation
is thus a key issue.
As it prepared to enter the market in 2003, Apotex
developed a series of projections of the market share it
expected to gain and the profits it expected to earn.
In June
2003, Apotex projected its market share as 5% at launch, 15% in
2004, and 25% from 2005 to 2007, for a total of roughly $581
million in sales over its first five years on the market.
In
November 2003, Apotex prepared another set of projections
covering the period through the end of its fiscal year, June
2004.
Apotex forecast that it would capture 7% of the generic
market over that time, leading to roughly $27 million in profits
at a 92.5% profit margin.
53
Apotex argues that it could have avoided infringement by
changing its omeprazole product in three ways.
It could have
made certain changes to its pellet formulation that would have
rendered it non-infringing, copied other non-infringing generic
formulations, or developed a microtablet formulation.
The parties debate how feasible any of these approaches are
and how long it would have taken Apotex to successfully
manufacture alternative formulations and receive FDA approval
for them.
Before engaging in detail with their arguments in
this regard, it is worth noting that Apotex began working on a
formulation of omeprazole as early as 1996, and by 2003, seven
years later, had developed only an infringing product.
Second,
in defending this action, including in particular this damages
phase of the litigation, Apotex has not chosen to demonstrate
the ease with which it could have developed a non-infringing,
bioequivalent, and stable version of omeprazole by actually
manufacturing and testing such a formulation.
If, as it argues,
the development of a non-infringing formulation would have been
a relatively easy and swiftly accomplished task, its failure of
proof in this regard is telling.
Finally, it is notable that
although Genpharm, Andrx, and Cheminor were found to infringe
the Patents in the First Wave Opinion in October 2002, not one
54
of them developed and launched a non-infringing alternative
formulation in 2003, or indeed at any time before the expiration
of the Patents in 2007.
Both Astra and Apotex agree that Apotex would have had to
await FDA approval before marketing the revised product.
The
parties also agree that changes to the pellet formulation would
have required Apotex to submit a PAS or Prior Approval
Supplement.
See 21 C.F.R. § 314.70(b) (requiring supplemental
submission and prior approval for “major changes,” including
“changes in the qualitative or quantitative formulation of the
drug product, including inactive ingredients”).
Dr. Lin also
suggested that the creation of a microtablet might have required
submission of an entirely new ANDA rather than a PAS.
Apotex would have assumed in the Fall of 2003 that it would
take as much as a year to receive FDA approval of a PAS and as
much as two years to receive approval of an ANDA.
PASs were and
are reviewed on a “first in, first out” basis, as are ANDAs.
Moreover, PASs and ANDAs are assigned to the same set of
reviewers.
There is no publicly-available data concerning the
median time for review of PASs in the 2003 time period, but
using the data on approval of ANDAs, Dr. Lin estimates that
Apotex “could potentially have faced” a delay of seventeen
55
months before approval of a PAS.
Johnston opines, quite
optimistically, that the “reasonable expectation” for the
approval time for a well-documented PAS in the 2003 to 2004
timeframe was about six months.
One of Apotex’s employees, Tao, testified that between
December 2002 and November 2003, Apotex submitted “at least”
seven PASs to the FDA, but Apotex has only provided documents
showing how long it took for two of them to receive approval.
Those two took 7.5 and 10.5 months to receive FDA approval. 27
Tao testified that during 2003 Apotex projected internally that
the approval of a PAS would take roughly eight months.
Apotex
agrees with Astra that the median time from submission of an
original ANDA to approval was seventeen months in 2003.
Astra has shown that Apotex would have had no reasonable
confidence in the Fall of 2003 that it could find a noninfringing formulation for omeprazole that would obtain FDA
27
These two PASs involved metformin, a drug used in the
treatment of diabetes. As a result, they would have been
assigned to a different team than a PAS for omeprazole. One of
the supplements was submitted on March 31, 2003 and approved on
February 18, 2004, about 10.5 months later. In the interim,
Apotex received two deficiency notices, the second of which it
received about nine months after the initial submission. The
second supplement was filed on March 15, 2004. It was approved
on November 4, 2004, which is approximately 7.5 months later.
56
approval, much less one that could be developed, tested, and
approved without years of effort.
In essence, the evidence at
trial showed that each of its proposed alternatives would have
entailed significant delay and uncertainty, and would have cost
Apotex a substantial portion if not all of its profits.
Of
course, even if it had found a viable, non-infringing
formulation, before filing its submission with the FDA, Apotex
would need to complete the testing necessary to create a PAS and
ANDA.
The necessary stability tests for FDA approval required
three months of data.
A.
Changes to Apotex’s formulation
Apotex contends that it could have avoided infringement of
the Patents by making one of three changes to its pellet
formulation.
First, it could have avoided the use of an ARC,
one of the claimed elements of Astra’s patented formulation, by
removing the magnesium hydroxide from its product core.
Second,
it could have prevented the in situ formation of an inert
subcoat by changing the binder it used in the core.
Third, it
could have prevented the formation of the subcoat by changing
the coating it applied.
It is unlikely, however, that Apotex
would have believed in November 2003 that any of these changes
would have been successful or easy for it to pursue.
57
First, as to the ARC, Apotex argues that the magnesium
hydroxide constitutes only 2% of the core of the Apotex product
and could have been easily replaced by simply adding more of the
inert filler, i.e., the mannitol, which would have saved time in
the manufacture of the core.
Magnesium hydroxide, however,
protects omeprazole from degrading in an acidic environment,
such as the highly acidic aqueous environment in the human
stomach.
The combination of omeprazole and this particular ARC
in the core of the product, therefore, helps to stabilize the
compound.
In its development process, Apotex considered formulations
with no ARC in the product but chose not to pursue that course.
It also evaluated several different alkaline compounds over the
course of at least a year before selecting magnesium hydroxide
as the best candidate.
The Apotex formulation of its at risk
generic omeprazole, therefore, like the Astra formulation, uses
magnesium hydroxide as an excipient.
Similarly, several other pharmaceutical companies have used
an ARC in the composition of the core to protect the drug’s
active ingredient from degradation due to acidic conditions.
Without protecting the active ingredient from such degradation,
a manufacturer also risked discoloration of the product, which
58
could interfere with acceptance in the marketplace.
Moreover,
Apotex’s decision not to remove the ARC of magnesium hydroxide
from its product following the opinion rendered in the First
Wave litigation is further evidence that it believed the ARC to
be a necessary component of its product.
Astra has shown,
therefore, that Apotex would have had no reasonable expectation
that removing the ARC from its formulation would have been
either an easy or necessarily successful avenue to producing a
non-infringing product that could be approved and marketed.
Apotex argues that other non-infringing formulations were able
to avoid the use of an ARC in the core of their products.
For
the reasons discussed below, none of those would have presented
an available alternative.
Apotex suggests that it could have avoided the formation of
an in situ subcoating by altering its formulation in two
principal ways.
It could have changed the binder in the pellet
cores from Povidone (“PVP”) to hydroxypropyl cellulose (“HPC”)
or changed the polymer in the enteric coating from methylacrylic
acid copolymer Type C to hydroxypropyl methycellulose phthalate.
Astra’s expert Dr. Davies explained the numerous technical
difficulties and uncertainties associated with these changes.
For instance, while Apotex suggests that it could have changed
59
its binder from PVP to HPC, the latter is much less hard and
more friable than the former, which could necessitate other
changes to the manufacturing process.
Changing the binder is
also capable of changing the resulting drug’s dissolution
profile.
Changing the enteric coating too is capable of
affecting dissolution and stability, and the new coating Apotex
proposes would have required the use of an organic solvent,
something that Apotex had regarded as “highly undersirable” and
a “last resort.”
B.
Copying other formulations
Apotex next argues that it could have copied the
formulations of other, non-infringing generic omeprazole
manufacturers.
As an initial matter, it is worth remembering
that in November 2003, only KUDCo had been held to be noninfringing.
The KUDCo formulation did not contain an ARC in its
core and therefore did not infringe.
2d at 551.
Omeprazole I, 222 F. Supp.
KUDCo also, however, has patents on its formulation,
and Apotex has not explained how it could copy that formulation
without infringing those patents.
Lek and Mylan were, like Apotex, Second Wave defendants,
and so had launched at risk in 2003.
There would have been no
reason for Apotex to have regarded their formulations as non-
60
infringing in November of that year, or to have taken steps to
copy them.
Apotex suggests that because it is assumed to have
known in November 2003 that it infringed the Patents under the
hypothetical negotiation framework, it should also be assumed to
have known that Mylan and Lek did not infringe.
cited any law in support of this view.
Apotex has not
The hypothetical
negotiation framework does not treat the parties as having
knowledge of all events between the negotiation and the finding
of infringement simply because it requires them to assume that
the Patents are valid and infringed.
At any rate, even assuming that Apotex would have wanted to
copy the formulations of Mylan and Lek, there is no indication
that it could have done so.
The Mylan/Esteve formulation uses
different excipients and a significantly different manufacturing
process for its product.
formulation.
It also holds patents on its
Because of the very different approach taken by
Mylan/Esteve, the adoption of this route would have required
Apotex to engage in very substantial research and development,
at considerable time and expense.
Lek was able to avoid the use of an ARC by excluding the
use of water in its formulation.
Apotex used water in both the
granulation and coating steps for its generic omeprazole.
61
The
conversion to a profoundly different manufacturing process would
have entailed a significant and expensive investment and a
lengthy delay in development and testing.
Moreover, like KUDCo
and Mylan, Lek holds patents related to its formulation.
C.
Microtablet formulation
Apotex argues that it could have pursued a third route to a
non-infringing product: the creation of a microtablet.
But,
Astra has shown that Apotex would have had absolutely no
confidence that it could succeed following this path either.
It
had tried mightily and without success for almost two years to
make an omeprazole microtablet and had abandoned those efforts
as futile.
Aware that Astra’s patent on omeprazole was due to expire
in April 2001, Apotex began as early as 1996 to attempt to
create an omeprazole product.
Apotex first tried to formulate
omeprazole core pellets, but quickly placed that project on hold
to pursue the development of mini-tablets for placement in hard
gelatin capsules.
Dr. Sherman, CEO of Apotex, wrote to Dr. Beach, head of
Torpharm, on August 1, 1997, to highlight the potential value of
omeprazole to the company and emphasize the need for haste in
developing a formulation.
Dr. Sherman recited that omeprazole
62
is “the largest selling drug in the U.S., with U.S. annual sales
approaching $2 billion.”
He listed the patents-at-issue, noting
that the omeprazole patent expired in 2001, but that the
formulation patents did not expire until 2007.
As he observed,
those patents covered the use of an alkaline agent in the core
and a subcoat under the enteric coat to improve stability.
With
the approach that Dr. Sherman recommended, he was confident that
Apotex “will not infringe.”
Dr. Sherman wanted to make tablets,
four of which would fit in a capsule.
He noted that, in
contrast, “the pellets are difficult to make. . . . As
omeprazole is unstable at neutral or acid pH, an alkaline
compound has to be included to avoid degradation in processing.
This makes it difficult to avoid the formulation patents.”
Over the course of the next roughly seventeen months, Dr.
Sherman sent numerous memos to Dr. Beach listing the problems
with the microtablet formulation and urging speed in developing
solutions.
In a memo dated July 24, 1998, Dr. Sherman explained
that a portion of the tablets were not disintegrating, and that
the explanations might be either too much coating or not enough
coating.
Finally, on December 18, 1998, Dr. Sherman instructed
Dr. Beach to abandon the microtablet formulation, noting the
persistent problems with stability and bioequivalence.
63
As this history demonstrates, Apotex tried mightily, with
the specific goal of avoiding infringement of the ‘505 and ‘230
patents, to develop a microtablet formulation.
months, it concluded that it had failed.
After seventeen
There is no evidence
to suggest that Apotex would have believed that this effort
would be any easier or more successful in November 2003.
These anticipated delays would have translated into lost
profits for Apotex as it scrambled to enter the market as
quickly as possible.
Using Apotex’s projections of how rapidly
it would be able to gain a certain share of the generic
omeprazole market, Dr. Meyer calculated the percentage of its
profits Apotex would have sacrificed with each year of delay.
Starting with Apotex’s June 2003 projections of the market share
it expected to capture, Dr. Meyer determined that a one year
delay would have cost Apotex 31.6% of its profits, a two year
delay would have cost 59.2% of its profits, and a three year
delay would have cost 84.5% of its profits.
The likelihood that
Apotex would have faced at least a two year delay if it
attempted to develop a non-infringing formulation thus does much
to shed light on the value to Apotex of a license.
VI.
The Book of Wisdom:
Post-entry Information
64
A.
Nexium®
When Apotex entered the generic omeprazole market in
November 2003, it launched its product “at risk” of a later
finding of infringement and without the freedom to shape a
pricing strategy that a license would have given it.
Despite
that distinction, it is helpful to look at the impact of
Apotex’s launch on Astra’s Nexium® strategy.
Data from various
sources show that Astra offered increasing rebates on Nexium® in
the years after November 2003.
Astra’s internal manufacturer
rebate data shows the average manufacturer rebate rates for the
13 TPPs selected by Dr. Rausser steadily increasing from 2004 to
2006.
Astra’s financial data also shows increasing rebate
expenses as a percentage of gross sales.
Nevertheless, the decreasing price of generic omeprazole
was only one factor among several influencing Nexium’s market
share and Astra’s need to offer rebates to maintain its
position.
Indeed, isolating exactly why Astra offered
increasing rebates on Nexium® is not possible.
Internal
documents suggest that Astra may have been motivated by pricing
pressures from generic omeprazole, other branded PPIs, and even
Prilosec OTC®.
65
An April 2004 Astra strategy document analyzed Nexium’s
“access relative to the competition” and listed as the
competition only Protonix® and Prevacid®.
The same document
described the “changes in the PPI market for managed care” as
including multiple generics, but also noted that branded
competitors were pricing more aggressively and that Protonix®
was “on the rise.”
In August 2004, Astra conducted a workshop regarding its
strategy with Nexium®.
As of that time, PPIs were the second
largest drug class as measured by sales in the United States.
Although the sales of Prilosec OTC® had been generally “flat”
since its launch, Astra noted that PPI market growth had
“stopped” at the time of Prilosec OTC®’s launch.
As of that
time, Prevacid®, Nexium® and Protonix® each had a market share
of between 29 to 24%, in that descending order.
In contrast,
the omeprazole share of the prescription PPI market was less
than 10% and Prilosec® was less than 3%.
Astra also noted that
the PPI class was “extremely price competitive,” and prepared a
chart showing the “WAC less rebate” for the various drugs in the
market.
This chart showed that Nexium® had the highest price
($2.68), with Prevacid® ($1.94), Prilosec® ($1.90), omeprazole
66
($1.89), and Aciphex® ($1.80) roughly even.
Protonix® ($1.22)
and Prilosec OTC® ($0.71) were substantially cheaper. 28
Astra’s August 2004 strategy document also showed that
there was a strong negative trend in formulary coverage without
restrictions for all brand name PPIs, including Nexium®,
starting in the fall of 2002.
Nonetheless, Nexium® was on the
formulary at the large PBMs, who control most of the “covered
28
By January 2005, Astra calculated that the WAC with the
deepest rebate included in the per pill price resulted in a pill
of Nexium® costing $2.65, omeprazole $1.17, and Prilosec OTC®
$0.71.
67
lives” in the market.
Indeed, the national PBMs generated over
half of Nexium’s total volume.
In planning for the future,
Astra noted that the resolution of the generic litigation could
change the landscape.
An Astra document prepared in September 2004 noted that
while Nexium® continued to “experience wide access and
reimbursement on MC formularies,” the “PPI class remains very
competitive and the introduction of Prilosec OTC® and generic
omeprazole have increased the competitive pressure.”
The
document went on to examine Nexium’s formulary status for twelve
national TPPs that together with mail order represented more
than 82% of the market.
For these major accounts, Astra noted,
“Nexium’s formulary status has remained largely unchanged
throughout the launch of generic omeprazole and Prilosec OTC®.”
The document later analyzed “managed market trends” and noted
that many TPPs were encouraging patients to use Prilosec OTC®
“as first line therapy in a means to reduce branded PPI Rx
costs.”
The document also noted that while “NEXIUM is in
similar position to other branded PPIs in terms of restricted
access,” “the formulary position for NEXIUM in key accounts has
not changed!
NEXIUM.”
40% of covered lives have UNRESTRICTED access to
(Emphasis in original.)
68
One report commissioned by Astra and prepared in October
2004 evaluated how much volume Nexium® had gained from various
competitors.
It reported that Nexium® had gained 3% of its
volume from Prilosec® and 0% from omeprazole while it lost 2.8%
of its volume to Prilosec OTC®.
In a January 2005 Nexium® performance review, Astra noted
that Nexium® had enjoyed excellent growth in market share over
the course of 2004 and that rebate rates had been “lower than
expected.”
Again there was no mention of generic omeprazole
being a factor in Nexium’s rebate strategy, or of Apotex in
particular.
In sum, the post November 2003 evidence supports the idea
that Astra offered increasing rebates on Nexium® to compete in
the PPI market, and that the price of generic omeprazole was one
factor among several influencing Nexium’s competitive position.
It is therefore reasonable to suggest that Astra’s Nexium®
strategy would have made it disinclined to grant Apotex a
license, with which Apotex might have cut prices much more
drastically.
B.
Prilosec OTC®
A series of reports commissioned by Astra and prepared by a
consulting firm called Ipsos analyzed the performance and impact
69
of Prilosec OTC® in the months and years following its launch.
Ipsos’s “Prilosec OTC® Year 1 Review,” prepared in October 2004,
generally characterized Prilosec OTC® as behaving like a
prescription product and noted that it would therefore “continue
to threaten Rx somewhat more than OTC brands.”
A chart
contained in the report shows that omeprazole lost 16% of its
volume to Prilosec OTC®, Prilosec® lost 12%, and Nexium® lost
3%.
Ipsos prepared another report, entitled “Prilosec OTC 20-
Month Review,” in June 2005.
Again Ipsos noted that Prilosec
OTC® was behaving like an Rx product, and again it observed
large losses in volume by Prilosec® (14%) and omeprazole (7%) to
Prilosec OTC®.
Nexium® was reported to have lost 1% to Prilosec
OTC®.
Astra’s internal analysis confirmed that Prilosec OTC® had
a significant impact on the total volume of omeprazole
prescriptions.
A chart prepared by Astra shows the total number
of omeprazole prescriptions falling from roughly 130,000 to
70,000 in the months following OTC’s launch.
Astra’s internal
documents show that it also believed that Prilosec OTC® had
taken volume from Nexium®.
In its 4th quarter 2004 performance
review for Nexium®, dated January 11, 2005, Astra noted that
70
while Nexium® continued to grow in volume, Prilosec OTC® had
slowed its growth by about 4%.
These figures were seen despite the fact that the out of
pocket cost to consumers of Prilosec OTC® was higher than the
cost of omeprazole and Prilosec®.
This was due largely to
Prilosec OTC®’s lack of coverage by TPPs.
Dr. Rausser
calculated that in the interval between November 12, 2003, and
the end of 2004, there were 673 instances in which a third party
payer contributed to the cost of Prilosec OTC®, which
represented only 2.5% of PPI purchases.
plan sponsors covered Prilosec OTC®.
In 2004, only 6% of
A series of reports
produced by Takeda contained similar numbers.
They indicated
that 4% of plan sponsors covered Prilosec OTC® in 2004, 6%
covered it in 2005, and 9% covered it in 2006.
C.
Apotex’s Sales
Although Apotex launched its product at a lower WAC price
than Mylan, Lek, and KUDCo, it did not cut prices aggressively,
launching at a WAC price of $2.65 compared with Lek’s $2.69.
Nevertheless, omeprazole was one of Apotex's largest selling
products in the United States. 29
For the twelve months ending
29
Dr. Sherman explained that “for some reason the competitors
decided not to lower prices to keep us out, so we [Apotex] were
71
March 31, 2007, omeprazole generated the second largest amount
of net revenue for Apotex of all of its products sold in the
United States.
During that same period, omeprazole capsules produced the
highest gross margin for the company.
Apotex initially
surpassed its projection that it would earn a 92% margin on
omeprazole, earning margins of 96.2% and 95.8% in November and
December 2003, respectively.
profit margins declined:
Eventually, however, Apotex’s
in 2004 it earned a margin of 88.5%
and in 2005 it earned a margin of 68.3%.
Overall, Dr. Meyer
calculated that Apotex sold $201,791,249 worth of omeprazole in
the U.S. from November 2003 to October 2007, for a total profit
of $152,043,989, or a 75.3% margin.
The effect of Apotex’s entry on generic omeprazole prices
was hotly disputed at trial.
A chart prepared from the pharmacy
log data collected by Dr. Rausser shows a downward trend in
prices beginning roughly a month before Apotex’s entry, and
continuing at essentially the same rate after Apotex’s launch.
In the period from Apotex’s launch until 2006, the price of
omeprazole dropped from $3 to $2.50.
Dr. Rausser testified that
able to take a larger market share than we expected with less
price reduction than we expected.”
72
this data showed a more serious downward adjustment in prices
after Apotex’s launch, but the evidence at trial generally
showed that it was not possible to isolate Apotex’s contribution
to omeprazole’s price decline after November 2003.
Of course,
this price data does little to shed light on how Apotex would
have priced its omeprazole had it been a licensed entrant.
D.
Settlements
There are two additional sets of negotiations that occurred
after November 2003 that the parties have discussed as potential
benchmarks for the reasonable royalty between Astra and Apotex.
One is the 2005 settlement offer made by Andrx; the other is the
2010 settlement with Teva.
1.
Andrx offer
On May 11, 2005, Andrx sought a license to sell its generic
omeprazole from July 1 until the expiration of the relevant
patents in 2007.
Andrx and Astra had been engaged in ANDA
litigation since May of 1998, and as of May 2005 Andrx had
already been found to infringe and had seen this finding upheld
on appeal before the Federal Circuit.
All that remained in the
litigation was the determination of damages, which Astra sought
for Andrx’s commercial manufacture of omeprazole in the period
73
before October 2002, when Judge Jones held that Andrx was
infringing the Patents.
Andrx sought the end of all litigation surrounding the
Patents.
In exchange, it would purchase a license from Astra
for 70% of its profits from the sale of the 40mg strength
omeprazole.
As an alternative, Andrx offered to license the 10,
20, and 40mg strengths, in which case it would pay 70% of its
profits on the 40mg strength, but only 50% of its profits on the
10 and 20mg strengths.
At that time, Andrx was the exclusive
first-filer for the 40mg dosage, and therefore would have been
able, after it had obtained all of the necessary approvals, to
launch its product as the first and exclusive authorized generic
version of Prilosec® at that dosage for 180 days.
In exchange,
Andrx proposed that Astra dismiss its outstanding claim for
damages in connection with Andrx’s infringement, as well as any
claim for fees and costs in connection with the litigation.
An
undated draft license agreement, which Andrx provided to Astra,
included a minimum royalty payment for the first 180 days of $25
million.
Apotex points out that the negotiations concerned not just
the Patents but also two other patents: the '281 and '905
patents.
Apotex also observes that Andrx did not offer to
74
license only the 10 and 20mg strengths, and suggests that
without the ability to sell the 40mg strength, the value of a
license to only the 10mg and 20mg omeprazole would have been
lower than even the 50% rate in Andrx’s proposal.
On the other
hand, Andrx’s letter was only an opening offer, and it is
reasonable to assume that it would have been willing to agree to
a higher royalty had Astra been interested in negotiating.
The
record contains, however, no evidence of any interest on the
part of Astra in Andrx’s proposal.
Uhle, who testified
regarding Astra’s licensing practices, said that he had not
heard about the Andrx offer until he was made aware of it in
preparing for this litigation.
At any rate, the offer was not
accepted.
2.
Teva settlement
In January 2010, Astra reached a settlement with Teva and
Impax to resolve Astra’s pending claims against them for
infringement of the Patents, including claims for damages based
on Teva/Impax’s at-risk launch of their generic omeprazole in
September 2004.
Teva and Impax had entered into a Strategic Alliance
Agreement in 2001.
Impax manufactured the omeprazole capsules
and Teva marketed them.
Pursuant to their agreement, Teva paid
75
Impax the cost of manufacture plus 35% of its defined profit
from the sale of omeprazole for the first 18 months after
launch, 40% for the next six months, and 50% thereafter.
Apotex
emphasizes that this royalty agreement was reached as a
component of a strategic alliance between Teva and Impax.
Among
other things, Teva provided a loan for construction of a new
facility and purchased Impax stock.
Teva was one of the Second Wave defendants, and Judge Jones
found in May 2007, less than three years after its at-risk
launch, that Teva’s omeprazole capsules infringed the Patents.
490 F. Supp. 2d at 499.
2008.
That decision was affirmed in June
In re Omeprazole Patent Litig., 281 Fed. Appx. 974 (Fed.
Cir. 2008) (“Omeprazole IV”).
Acknowledging the validity of Astra’s patents, Teva/Impax
settled Astra’s outstanding damages claim for a lump-sum payment
of less than $10 million.
The parties here dispute how best to
translate this lump-sum payment into a royalty rate that could
be used as a benchmark in this litigation.
Astra’s expert Dr.
Meyer calculated various royalty rates implied by the total
amount of the settlement.
First, Dr. Meyer calculated the
amount as a percentage of Teva’s net sales, based on both
publicly available sales transaction data, and on confidential
76
reports Teva and Impax used to report such figures to each
other.
More importantly for this case, Dr. Meyer also
calculated how the settlement amount would translate into a
percentage of profits, this time using only the Teva/Impax
reports.
This analysis showed that the settlement represented
54% of Teva’s profits on its infringing sales. 30
E.
Infringement Litigation
On October 6, 2003, the FDA granted final approval of the
10 and 20 milligram strengths of Apotex’s product, and tentative
approval of the 40 milligram product.
F.Supp.2d at 471.
Omeprazole III, 490
On November 12, 2003, Apotex started to sell
its 10 and 20 milligram doses in the United States.
Id.
On May 31, 2007, Judge Jones found that Apotex’s 10, 20 and
40 milligram ANDA omeprazole products infringed claims 1, 5, 6
and 10 of the ‘505 patent and claims 1, 6, 7 and 13 of the ‘230
patent.
Id. at 486.
She also held that its filing of the ANDAs
constituted acts of infringement, and that its manufacture, sale
and offering for sale of FDA-approved 10 and 20 milligram
generic omeprazole directly infringed the Patents.
30
Id.
Dr. Meyer also calculated a percentage based on the combined
profits of both Teva and Impax, but explained that this figure
is less relevant, because Teva was responsible for making the
lump sum payment to Astra.
77
On June 14, Judge Jones also concluded that a period of
pediatric exclusivity applied to the Patents and therefore ruled
that the effective date of approval for Apotex’s and other
generic products “shall be no earlier than October 20, 2007”.
In re Omeprazole Patent Litig., No. 01 Civ. 9351 (BSJ) (S.D.N.Y.
June 14, 2007) (Judgment).
On June 28, the FDA revoked its
final approval of Apotex’s ANDA until at least October 20, 2007.
In response, Apotex sued the FDA, seeking to enjoin its
revocation of Apotex’s approval.
Apotex Inc. v. U.S. Food and
Drug Admin., 508 F. Supp. 2d 78, 80 (D.D.C. 2007).
On September
17, 2007, the District Court in that district denied the motion
for injunctive relief and for a stay pending appeal.
Id. at 89.
In rendering its decision on the Apotex application for an
injunction, the court held that once a court issues a ruling
establishing pediatric exclusivity, “the FDA had no authority to
issue final approval” of an ANDA prior to the expiration of the
patents.
Id. at 84.
The court noted that a decision denying
Astra the pediatric exclusivity period would “deprive it of the
statutorily-awarded benefit of its financial investment” in
conducting pediatric studies on the drug.
Id. at 88.
“[A]bsent
an exclusivity period, not only would [Apotex] be able to
distribute its product, but other generic Prilosec manufacturers
78
& marketers would similarly be able to flood the market.
The
erosion of Astra’s statutory right is a significant harm.”
Id.
Thus, Apotex did not regain approval from the FDA to resume
marketing its product until October 22, 2007.
In 2008, the Federal Circuit affirmed Judge Jones’s
decisions in two separate opinions.
In the first, the Federal
Circuit affirmed Judge Jones’s findings of non-infringement as
to Mylan.
In re Omeprazole Patent Litig., 281 Fed. Appx. 974
(Fed. Cir. 2008).
In the second opinion, the Federal Circuit
affirmed Judge Jones’s findings of infringement as to Apotex and
Teva/Impax.
In re Omeprazole Patent Litig., 536 F.3d 1361 (Fed.
Cir. 2008).
Its decision included a finding that the district
court had jurisdiction to provide relief under Section
271(e)(4)(A) to reset the effective date of the ANDA to reflect
Astra’s six-month pediatric exclusivity period despite the
expiration of the patents.
Id. at 1381.
DISCUSSION
The Drug Price Competition and Patent Term Restoration Act
of 1984, Pub. L. No. 98-417, 98 Stat. 1585, commonly known as
the Hatch-Waxman Act, allows a generic manufacturer of an
already approved brand-name drug to obtain expedited approval to
79
market that drug by filing an Abbreviated New Drug Application,
or “ANDA.”
(2013).
See F.T.C. v. Actavis, Inc., 133 S. Ct. 2223, 2228
The filing of an ANDA pursuant to 35 U.S.C.
§ 355(j)(2)(A)(vii)(IV)(“Paragraph IV certification”) qualifies
as an act of infringement, and allows the generic manufacturer
and the brand-name patentee to litigate issues of infringement
before the generic is approved.
Actavis, 133 S. Ct. at 2228.
Indeed, the FDA is generally forbidden from approving the
generic manufacturer’s ANDA until either 30 months have elapsed
or a court has found that the underlying patent is invalid or
will not be infringed.
21 U.S.C. § 355(j)(5)(B)(iii); Caraco
Pharmaceutical Labs., Ltd. v. Novo Nordisk A/S, 132 S. Ct. 1670,
1677 (2012).
A generic that is the first to file an ANDA under this
subdivision enjoys a period of 180 days of exclusivity from the
first commercial marketing of its drug, a period that represents
“the vast majority of potential profits for a generic drug
manufacturer.”
Actavis, 133 S. Ct. at 2229 (citation omitted).
The Hatch-Waxman Act provides that, for an infringement action
based on the filing of an ANDA, “damages or other monetary
relief may be awarded . . . only if there has been commercial
manufacture, use, offer to sell, or sale within the United
80
States or importation into the United States of an approved
drug.”
35 U.S.C. § 271(e)(4)(C).
The Patent Act generally
provides that “upon finding for the claimant the court shall
award the claimant damages adequate to compensate for the
infringement, but in no event less than a reasonable royalty for
the use made of the invention by the infringer.”
35 U.S.C.
§ 284.
When the damages sought are based on a “reasonable
royalty,” that royalty is derived “from a hypothetical
negotiation between the patentee and the infringer when the
infringement began.”
ResQNet.com, Inc. v. Lansa, Inc., 594 F.3d
860, 868 (Fed. Cir. 2010).
The aim of the hypothetical
negotiation is to determine what royalty payment scheme willing
parties would have executed if infringement had not occurred.
Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301, 1325 (Fed.
Cir. 2009).
The royalty is to be determined by considering the
fifteen so-called Georgia-Pacific factors.
Dow Chemical Co. v.
Mee Industries, Inc., 341 F.3d 1370, 1382 (Fed. Cir. 2003)
(citing Georgia-Pacific Corp. v. United States Plywood Corp.,
318 F. Supp. 1116, 1120 (S.D.N.Y. 1970)).
Before moving to an analysis of the Georgia-Pacific factors
as applied by the parties’ experts and a determination of the
81
reasonable royalty that applies here, three legal arguments made
by Apotex will be addressed.
First, Apotex suggests that the
royalty should be based on the value of the infringing aspect of
its formulation, that is, the inert subcoating that formed in
situ between the active core of its pellets and the pellet’s
enteric coating, rather than the value of the omeprazole
capsules that were sold to consumers.
Second, Apotex argues
that it would be improper for the royalty to include the sixmonth “pediatric exclusivity period” that ran from April 30 to
October 20, 2007.
Third, Apotex argues that certain of the
plaintiff companies lack standing.
I.
The Isolated Value of the Subcoating
Apotex argues that Astra’s royalty calculations are flawed
because they are based on the value of the omeprazole capsules
as a finished product rather than on the isolated value of the
inert subcoating that was found to infringe Astra’s Patents.
Apotex suggests that the value of this subcoating was “minimal.”
Where a product, typically an electronic product, is
composed of many different components, royalties for
infringement are awarded “based not on the entire product, but
instead on the smallest salable patent-practicing unit.”
LaserDynamics, Inc. v. Quanta Computer, Inc., 694 F.3d 51, 67
82
(Fed. Cir. 2012) (citation omitted).
Under a narrow exception
known as the “entire market value” rule, however, a patentee may
assess damages based on the market value of the entire product
“where the patented feature creates the basis for customer
demand or substantially creates the value of the component
parts.”
Uniloc USA, Inc. v. Microsoft Corp., 632 F.3d 1292,
1318 (Fed. Cir. 2011) (citation omitted).
In these
circumstances, a patentee must either “separate or apportion the
defendant’s profits and the patentee’s damages between the
patented feature and the unpatented features” or establish that
“the entire value of the whole machine, as a marketable article,
is properly and legally attributable to the patented feature.”
Id. (citing Garretson v. Clark, 111 U.S. 120, 121 (1884)).
Animating the entire market value rule are two goals:
“determining the correct . . . value of the patented invention,
when it is but one part or feature among many, and ascertaining
what the parties would have agreed to in the context of a patent
license negotiation.”
Lucent, 580 F.3d at 1337.
It is clear that the royalty in this case is properly
calculated based on the value of Apotex’s omeprazole capsule and
not, as Apotex suggests, the “negligible” value of the
subcoating itself.
As an initial matter, there is little reason
83
to import these rules for multi-component products like machines
into the generic pharmaceutical context.
Notably, Apotex has
cited no precedent for doing so, and has submitted no evidence
suggesting that the infringing subcoating is itself a “salable
patent-practicing unit,” LaserDynamics, 694 F.3d at 67, of its
omeprazole product.
It would therefore be improper to attempt
to isolate its value as distinct from the value of the
marketable capsules in determining a reasonable royalty.
Even under the entire market value rule, while the
subcoating did not of course create customer demand for
omeprazole, it did “substantially create[] the value,” Uniloc,
632 F.3d at 1318, of the drug as it was formulated by Apotex.
While Apotex argues that the coating “had no noticeable
benefit,” this contention is belied by factual findings in the
court’s prior opinions.
More specifically, several features of
omeprazole made it notoriously difficult to formulate.
Omeprazole is most effective when absorbed by the small
intestine and yet is highly susceptible to degradation in the
acidic environment of the stomach, meaning that scientists had
to develop a formulation that would allow the drug to pass
through the stomach and be absorbed by the small intestine, all
the while ensuring adequate shelf life in a drug that is
84
sensitive to heat, moisture, organic solvents, and light.
Omeprazole I, 222 F. Supp. 2d at 433-37.
After years of effort,
Astra scientists determined that a water soluble subcoat helped
solve many of these problems and allowed them to formulate a
commercially viable drug.
Id. at 437.
The court’s factual
account thus demonstrates that the subcoat was a crucial aspect
of the process embodied in the ‘505 patent, and that Astra’s
prior formulations, which lacked a subcoat, were not
commercially viable.
II.
Id.
The Pediatric Exclusivity Period
Apotex next argues that it would be legally improper for
the reasonable royalty to include payments made for the six
month “pediatric exclusivity” period that followed the
expiration of the patents-in-suit.
Under the Food and Drug
Administration Modernization Act of 1997, Pub. L. No. 105-115,
111 Stat. 2296, the FDA may request, “after consultation with
. . . the holder of an approved application for a drug,” that a
patent holder perform pediatric studies of a drug.
§ 355a.
21 U.S.C.
If those studies are performed, the period during which
the FDA is barred from approving an ANDA filed by another drug
manufacturer is extended by six months.
(c).
85
21 U.S.C. § 355a(b)-
The effect of this “pediatric exclusivity period” on
litigation under the Hatch-Waxman Act is somewhat complicated.
When a generic drug manufacturer files an ANDA with a Paragraph
IV certification, the patent holder may file suit to litigate
issues of validity and infringement.
§ 355(j)(5)(B)(iii).
21 U.S.C.
If the patent holder proves infringement,
the district court must set the ANDA’s effective date no earlier
than the date the patent expires.
35 U.S.C. § 271(e)(4)(A).
What this means in practice is that if the court finds that the
generic’s ANDA infringes the patent before the FDA approves the
ANDA, then the FDA may not approve the ANDA (and the generic may
not sell its drug) until after the patent expires.
In re
Omeprazole Patent Litig., 536 F.3d 1361, 1367 (Fed. Cir. 2008).
If the FDA has already approved the ANDA, however, the court’s
order results in a retroactive change in the ANDA’s effective
date.
Id. at 1367-68.
As noted above, the six month pediatric exclusivity period
extends the time during which the FDA is barred from approving
an ANDA.
21 U.S.C. § 355a(b)(1)(B)(ii).
Accordingly, although
the Patents expired on April 30, 2007, the court in this case
set the effective date for the Apotex ANDA six months later, at
October 20, 2007.
The FDA then revoked its earlier approval of
86
Apotex’s ANDA, and re-approved it on October 22, 2007.
The
court’s rulings in this respect were affirmed by the Federal
Circuit, which squarely rejected the argument that “the district
court may not grant relief relating to the period of market
exclusivity after a patent has expired.”
Omeprazole V, 536 F.3d
at 1368.
Apotex does not appear to dispute that the effect of this
statutory scheme is to bar the sale of a generic until after the
expiration of the six-month pediatric exclusivity period.
See
Janssen Pharmaceutica, N.V. v. Apotex, Inc., 540 F.3d 1353, 1362
n.7 (Fed. Cir. 2008).
Naturally, then, a license acquired in
2003 would have had to include both the right to sell omeprazole
during the original term of the Patents and during Astra’s
pediatric exclusivity period.
Apotex nonetheless argues that it would be impermissible to
include royalty payments for the pediatric exclusivity period in
any damages calculation, because a patent holder may not collect
royalty payments beyond a patent’s expiration date.
Apotex
takes this premise from Brulotte v. Thys, 379 U.S. 29 (1964), in
which the Court held that “a patentee’s use of a royalty
agreement that projects beyond the expiration date of the patent
is unlawful per se.”
Id. at 32.
87
But Brulotte, which predated
the legislation concerning the pediatric exclusivity period, has
been roundly criticized for years, and does not actually support
such a sweeping generalization.
Brulotte’s reasoning was
grounded in a fear that the legislative scheme governing patents
would be corrupted if a patent holder could use a license “to
project its monopoly beyond the patent period,” thereby
subjecting “the free market visualized for the post-expiration
period . . . to monopoly influences that have no proper place
there.”
Id. at 32-33. 31
This concern is simply not present here, since Astra is not
attempting to extend Apotex’s royalty payments beyond the
statutory exclusivity period.
Charging Apotex for sales made
during the pediatric exclusivity period does not impair “the
free market visualized for the post-expiration period,” id. at
32, since by creating the pediatric exclusivity period Congress
obviously determined that there should be no “free market”
during those six months.
31
This reasoning has been condemned as “lack[ing] economic or
logical sense.” Zila, Inc. v. Tinnell, 502 F.3d 1014, 1019 n.4
(9th Cir. 2007) (citation omitted); see also Scheiber v. Dolby
Labs., Inc., 293 F.3d 1014, 1017 (7th Cir. 2002) (describing
Brulotte’s reasoning as “not true,” because “charging royalties
beyond the term of the patent does not lengthen the patentee’s
monopoly; it merely alters the timing of royalty payments”).
88
Apotex argues that the pediatric exclusivity period is not
technically an extension of the patent term, meaning that
royalty payments made during this period are impermissible under
Brulotte.
This problem is adequately addressed by Astra’s
expert, who distinguished between “royalty” payments made during
the patent term and “waiver” payments made in exchange for
Astra’s waiver of the pediatric exclusivity period.
Apotex’s
assertion that these labels are merely semantic rings hollow,
since its own argument relies on drawing an essentially
artificial distinction between an extension of the patent term
and the statute’s grant of “market exclusivity,” both of which
have the same effect in terms of Apotex’s ability to sell its
infringing product.
Apotex also suggests in the alternative that the royalty
rate should be lower during the pediatric exclusivity period,
since even without a license it would have been able to
manufacture (but not sell) omeprazole during this time.
Dr.
Meyer explained at trial that the ability to manufacture alone
would not have affected Apotex’s willingness to pay for a
license for the ability to sell, and without a license during
the pediatric exclusivity period its profits would have
suffered.
Apotex offers no compelling reason why it would have
89
been less eager to obtain a license to sell the infringing
formulation during the pediatric exclusivity period than during
the period before the expiration of the patents.
Indeed, any
six month interruption in its ability to sell its generic
product would be expected to have far reaching consequences.
It
is thus entirely proper for the damages calculation in this case
to include payments made for Apotex’s right to sell omeprazole
during the pediatric exclusivity period.
III.
Standing
Article III, Section 2 of the United States Constitution
limits the jurisdiction of the federal courts to actual cases
and controversies.
See Sprint Comm. Co., L.P., v. APCC Servs.,
Inc., 554 U.S. 269, 273 (2008).
One component of this
limitation is the requirement that a plaintiff suing in federal
court have standing to sue.
(2007).
Lance v. Coffman, 549 U.S. 437, 439
To establish standing, a plaintiff must allege
“personal injury fairly traceable to the defendant’s allegedly
unlawful conduct and likely to be redressed by the requested
relief.”
Hein v. Freedom From Religion Foundation, Inc., 551
U.S. 587, 598 (2007) (citation omitted).
“Since the patent
statutes give rise to the right to sue others for patent
infringement, they also define the nature and source of the
90
infringement claim and determine the party that is entitled to
judicial relief.”
Morrow v. Microsoft Corp., 499 F.3d 1332,
1339 (Fed. Cir. 2007).
The Patent Act entitles a “patentee” to bring a “civil
action for infringement of his patent.”
35 U.S.C. § 281.
A
“patentee” can be both the person to whom the patent was issued
and “successors in title to the patentee.”
35 U.S.C. § 100(d).
The Federal Circuit has described
three general categories of plaintiffs encountered
when analyzing the constitutional standing issue in
patent infringement suits: those that can sue in their
own name alone; those that can sue as long as the
patent owner is joined in the suit; and those that
cannot even participate as a party to an infringement
suit.
Morrow, 499 F.3d at 1339.
The first category, plaintiffs who
can sue in their own name alone, includes both the patentee
itself and any entity to whom the patentee has transferred “all
substantial rights to the patent.”
omitted).
Id. at 1340 (citation
The second category, plaintiffs who can sue as long
as the patent owner is joined in the suit, includes those who
“hold exclusionary rights and interests created by the patent
statutes, but not all substantial rights to the patent.”
Among this group are “exclusive licensees,” who hold
“exclusionary rights” by virtue of the fact that they may
91
Id.
“prevent others from practicing the invention.”
Id.
Finally,
those who lack exclusionary rights do not meet the injury in
fact requirement and therefore lack constitutional standing.
Id.; see also Sicom Sys. Ltd. v. Agilent Techs. Inc., 427 F.3d
971, 976 (Fed. Cir. 2005) (“A nonexclusive license confers no
constitutional standing on the licensee to bring suit or even to
join a suit with the patentee because a nonexclusive licensee
suffers no legal injury from infringement.”).
The Federal Circuit has explained that to qualify as an
exclusive licensee for standing purposes, “a party must have
received, not only the right to practice the invention within a
given territory, but also the patentee’s express or implied
promise that others shall be excluded from practicing the
invention within that territory as well.”
Spine Solutions, Inc.
v. Medtronic Sofamor Danek USA, Inc., 620 F.3d 1305, 1317 (Fed.
Cir. 2010) (citation omitted).
Notably, an exclusive licensee
need not be the only party with the ability to license the
patent.
WiAV Solutions LLC v. Motorola, Inc., 631 F.3d 1257,
1266 (Fed. Cir. 2010).
Rather, “a licensee is an exclusive
licensee of a patent if it holds any of the exclusionary rights
that accompany a patent.”
Id.
The exclusive licensee’s
standing, however, is “coterminous” with the exclusionary rights
92
it has been granted, meaning that it “may have standing to sue
some parties and not others.”
Id.
In sum, “if an exclusive
licensee has the right to exclude others from practicing a
patent, and a party accused of infringement does not possess,
and is incapable of obtaining, a license of those rights from
any other party, the exclusive licensee’s exclusionary right is
violated.”
Id. at 1266-67.
At the outset, Astra makes two arguments as to why the
issue of standing should not even be addressed at this time.
First, Astra argues that Apotex should have raised any objection
to Astra’s standing sooner and suggests that by failing to do so
Apotex has waived the issue.
It is well settled, however, that
“standing is jurisdictional and not subject to waiver.”
Spine
Solutions, 620 F.3d at 1319 (citation omitted).
Second, Astra argues that its standing has been established
as “law of the case.”
While no court in this litigation has
previously mentioned the issue of standing, Astra argues that
because standing is jurisdictional, this court and the Federal
Circuit must have concluded implicitly that all the plaintiffs
had standing, and that that conclusion should not be revisited
now.
The law of the case doctrine, however, “is a discretionary
rule of practice and generally does not limit a court’s power to
93
reconsider an issue.”
592 (2d Cir. 1991).
Liona Corp. v. PCH Assocs., 949 F.2d 585,
In light of the court’s “independent
obligation to examine [its] own jurisdiction,” United States v.
Hays, 515 U.S. 737, 742 (1995), it would be unwise to avoid the
question of standing by means of the law of the case doctrine
where no court in this litigation has explicitly addressed the
issue.
See Wyoming v. Oklahoma, 502 U.S. 437, 462 (1992)
(Scalia, J., dissenting) (“[L]aw of the case principles . . .
have never to my knowledge been applied to jurisdictional issues
raised (or reraised) before final judgment.”).
The Court
therefore turns to the standing of the five plaintiffs.
Of those five plaintiffs, Apotex contends that four lack
standing.
Apotex concedes that Hassle owns the patents-in-suit
and therefore has standing.
AstraZeneca AB, a Swedish company,
owns Hassle in its entirety.
AstraZeneca LP is a U.S. company
that is majority owned by AstraZeneca AB.
KBI and KBI-E are
U.S. entities that Astra describes as “affiliated” with
AstraZeneca AB, but that are not owned by any of the other
plaintiffs. 32
32
In its pretrial briefing, Astra indicates that it previously
offered to withdraw the claims of KBI and AstraZeneca LP, and
now makes no argument in its briefing as to why those two
entities have standing.
94
The relationships between these various parties vis-à-vis
the Patents are shaped by a series of agreements.
AstraZeneca
AB, the sole owner of Hassle, entered into an agreement with
Hassle in 1985 called the Commission Company Agreement.
The
Commission Company Agreement, a one-page document, gives
AstraZeneca AB the right to control the business of Hassle, but
does not mention the patents or intellectual property in
particular.
The conduct of AstraZeneca AB over the years
indicates that the parties to the Commission Company Agreement
understood that AstraZeneca AB’s power to control Hassle
included the power to control the use of the Patents.
At
various times, AstraZeneca AB has granted licenses to the
patents-in-suit to other companies without Hassle’s involvement,
while Hassle has never granted a license to another party
without AstraZeneca AB’s involvement.
Furthermore, Hassle has
allowed AstraZeneca AB to join it as a plaintiff in numerous
litigations, including the First and Second Waves of this
litigation.
In 1998, the various plaintiff entities entered into a
series of agreements the parties refer to as the “Project Nobel”
agreements, which restructured the ownership and operations of
Astra Merck Inc., a now-defunct joint venture between
95
AstraZeneca AB and Merck.
In one of these agreements, the
“Amended and Restated License and Option Agreement” (the “Option
Agreement”), AstraZeneca AB granted to KBI “an exclusive license
. . . to make, have made, use and sell” omeprazole products.
The Option Agreement also gave KBI various rights regarding
infringement litigation, including the right to initiate an
action for infringement, either jointly with AstraZeneca AB or
on its own, and the right to participate in decisions regarding
settlement.
KBI then assigned its rights under the Option
Agreement to KBI-E, via the “Assignment and Assumption of
Amended and Restated License and Option Agreement” (the
“Assignment”).
In the “Limited Sublicense Agreement,” KBI-E
then sublicensed back to KBI the non-exclusive right to make
omeprazole products, for the limited purpose of supplying those
products to AstraZeneca LP under another agreement called the
“Supply Agreement.”
KBI-E sold the exclusive right to
distribute omeprazole and other products to AstraZeneca LP for a
one-time fee.
AstraZeneca AB has standing as an exclusive licensee of the
patents-in-suit.
Although the Commission Company Agreement
between it and Hassle did not mention the Patents explicitly, or
indeed any other patent, an exclusive license may be granted
96
implicitly.
See Spine Solutions, 620 F.3d at 1317. 33
By
obtaining complete control of Hassle, AstraZeneca AB gained
control over the Patents, as demonstrated by the parties’
behavior in the years following the execution of the Commission
Company Agreement.
For instance, AstraZeneca AB granted several
licenses to the patents-in-suit before the Project Nobel
agreements were executed without the participation of Hassle,
while Hassle never granted licenses without the involvement of
AstraZeneca AB.
AstraZeneca AB also participated in litigation
to enforce the Patents.
This behavior indicates that the
Commission Company Agreement implicitly gave AstraZeneca AB the
right to exclude others from practicing the patents-in-suit and
that AstraZeneca AB was therefore an implied exclusive licensee.
See Rite-Hite Corp. v. Kelley Co., Inc., 56 F.3d 1538, 1552
(Fed. Cir. 1995) (en banc); cf. WiAV Solutions, 631 F.3d at 1266
(“[I]f the patentee allows others to practice the patent in the
licensee’s territory, then the licensee is not an implied
33
Apotex argues, citing Abraxis Bioscience, Inc. v. Navinta LLC,
625 F.3d 1359 (Fed. Cir. 2010), that AstraZeneca AB lacks
standing because “[c]ommon corporate structure does not overcome
the requirement that even between a parent and a subsidiary, an
appropriate written assignment is necessary to transfer legal
title from one to the other.” Id. at 1366. AstraZeneca AB’s
standing, however, is as an exclusive licensee, not an assignee.
97
exclusive licensee.” (citation omitted)).
AstraZeneca AB thus
has standing to sue as long as the patent owner, Hassle, is
joined in the suit.
Morrow, 499 F.3d at 1339.
The standing of KBI and KBI-E is more clear-cut.
As noted
above, the Option Agreement transferred AstraZeneca AB’s rights
in the patents to KBI, which then assigned those rights to KBIE.
KBI-E was thus left with an “exclusive license . . . to
make, have made, use and sell” omeprazole products, as well as
the right to engage in litigation if its exclusive rights were
infringed.
KBI-E was thus an exclusive licensee, and therefore
has standing to join Hassle as a plaintiff in this suit.
on the other hand, did not retain its exclusive rights.
KBI,
While
KBI-E transferred back to KBI the non-exclusive right to
manufacture omeprazole in the Limited Sublicense Agreement, the
non-exclusive nature of the right KBI received deprives it of
standing.
KBI-E thus has standing, while KBI does not.
Apotex makes two arguments as to why KBI-E lacks standing,
neither of which has merit.
First, Apotex argues that because
AstraZeneca AB did not receive its rights to the Patents in
writing, it could not have transferred those rights to KBI.
As
explained above, however, an exclusive license need not been in
writing, and indeed AstraZeneca AB’s transfer to KBI supports
98
its own status as an exclusive licensee.
After all, without an
exclusive license it would not have been able to give KBI its
own exclusive license.
Second, Apotex argues that the
agreements between KBI and KBI-E left neither with an exclusive
license.
This argument is belied by the agreements themselves,
which indicate that KBI assigned its exclusive license to KBI-E,
which then transferred back to KBI only a non-exclusive right to
manufacture, retaining for itself the exclusive rights it
received under the Assignment.
AstraZeneca LP, on the other hand, bought from KBI-E the
“sole and exclusive right to promote, distribute, market and
sell” omeprazole products.
This transaction made AstraZeneca LP
an exclusive licensee with regard to the distribution of
omeprazole, while KBI-E retained an exclusive license to
manufacture omeprazole (which it sublicensed to KBI on a nonexclusive basis for the sole purpose of supplying AstraZeneca
LP).
AstraZeneca LP therefore gained the right to exclude
others from distribution of omeprazole and has standing to sue
Apotex insofar as Apotex distributed its infringing product.
Weinar v. Rollform Inc., 744 F.2d 797, 806-07 (Fed. Cir. 1984)
(holder of “exclusive right to sell as sole distributor in the
United States” had standing as an exclusive licensee); see also
99
WiAV Solutions, 631 F.3d at 1266 (“[A]n exclusive licensee lacks
standing to sue a party who has the ability to obtain such a
license from another party with the right to grant it.”).
In arguing that AstraZeneca LP lacks standing, Apotex
relies on Rite-Hite Corp. v. Kelley Co., 56 F.3d at 1553, but
that case did not abrogate the holding of Weinar that “a
licensee with the exclusive right to sell in the entire United
States . . . shared the property rights represented by a patent”
and therefore had standing.
Id. (citation omitted).
Rather,
the court in Rite-Hite concluded that the plaintiffs whose
standing was being challenged were at best “non-exclusive
licensees by implication.”
Because the rights granted in
Id.
the Distribution Agreement were explicitly declared to be “sole
and exclusive,” AstraZeneca LP has standing.
Of the five
plaintiffs, therefore, four have standing to bring the claims in
this litigation.
Only KBI lacks standing.
One final observation on the question of standing is
appropriate.
Apotex argues, more broadly, that its standing
arguments will impact the calculation of a reasonable royalty in
this case by affecting the interests that can be considered in
the hypothetical negotiation.
In other words, Apotex suggests
that if AstraZeneca LP (which earned the bulk of the profits
100
from sales of Prilosec) lacks standing, then the profits it
earned should not factor into the reasonable royalty
determination.
This argument misunderstands the nature of
reasonable royalty damages.
In a hypothetical negotiation, a patent owner can be
expected to account for the profits its affiliates and licensees
earn on its patents.
See Union Carbide Chemicals & Plastics
Tech. Corp. v. Shell Oil Co., 425 F.3d 1366, 1378 (Fed. Cir.
2005), overruled on other grounds, 576 F.3d 1348 (Fed. Cir.
2009) (“[T]he holding company would not enter any negotiation
without considering the competitive position of its corporate
parent.”).
Therefore, even if Apotex is correct that
AstraZeneca LP lacks standing, the reasonable royalty analysis
would still account for its economic position.
IV.
The Reasonable Royalty
The reasonable royalty inquiry begins with the Georgia-
Pacific factors.
Dow Chemical, 341 F.3d at 1382.
These are:
(1) royalties the patentee has received for licensing
the patent to others;
(2) rates paid by the licensee for the use of
comparable patents;
(3) the nature and scope of the license (exclusive or
non-exclusive, restricted or nonrestricted by
territory or product type);
(4) any established policies or marketing programs by
the licensor to maintain its patent monopoly by not
licensing to others to use the invention or granting
101
licenses under special conditions to maintain the
monopoly;
(5) the commercial relationship between the licensor
and licensee, such as whether they are competitors;
(6) the effect of selling the patented specialty in
promoting sales of other products of the licensee;
(7) the duration of the patent license term;
(8) the established profitability of the product made
under the patent, including its commercial success and
current popularity;
(9) the utility and advantages of the patent property
over the old modes or devices;
(10) the nature of the patented invention and benefits
to those who have used the invention;
(11) the extent to which the infringer has used the
invention and the value of that use;
(12) the portion of profit or of the selling price
that may be customary in that particular business to
allow for the use of the invention or analogous
inventions;
(13) the portion of the realizable profit that should
be credited to the invention as opposed to its nonpatented elements;
(14) the opinion testimony of qualified experts;
(15) the results of a hypothetical negotiation between
the licensor and licensee.
Whitserve, LLC v. Computer Packages, Inc., 694 F.3d 10, 27 n.11
(Fed. Cir. 2012).
In this case, it is useful to approach this analysis by
analyzing the basic negotiating positions of each party rather
than examining each of the Georgia-Pacific factors separately.
Such an analysis necessarily takes into account almost all the
Georgia-Pacific factors.
addressed above:
Two of the factors have already been
the duration of the patent term (factor 7)
will extend through the pediatric exclusivity period; the entire
102
capsule as opposed to non-patented elements (factor 13) is the
appropriate measure for assessing realizable profit.
Two
remaining factors, factors 2 and 14, will be addressed at the
end of the discussion below.
A.
Apotex’s position
Apotex’s negotiating position would have been shaped by one
basic concern.
Because Apotex expected to (and did) make
substantial profits from its sales of omeprazole, it would have
been willing to pay a large share of those profits for the right
to use the patents in 2003.
In deciding how large a share of
its profits it would pay, however, Apotex would have carefully
evaluated the cost and delay associated with developing and
obtaining approval for the sale of an alternative formulation
that did not infringe the Patents.
1.
Apotex’s profits
Apotex expected to, and did, earn substantial profits from
selling omeprazole between November 2003 and October 2007, the
time period covered by the hypothetical license.
While it
experienced a gross margin ranging from 31% to 48% for most of
the generic products it sold in the United States, it estimated
that its gross margin on generic omeprazole would easily double
that.
It also predicted that it would quickly gain a sizeable
103
foothold in the generic market and therefore enjoy massive
sales.
Those expectations were reasonable, and Apotex achieved
sales and growth roughly consistent with them.
During any negotiation for a license, of course, Apotex
would have been relying on even rosier expectations.
With a
license from Astra, its patent litigation would end and Apotex
would not have to act with the caution in pricing its generic
product that is customary for “at risk” entrants into the
generic market.
It would have been able to maximize its profit
even if that meant deep discounting.
Only one of the three
predecessor generic manufacturers in the PPI market -- KUDCo -would have been able to compete with that same freedom in
decision-making.
And, Apotex would have looked ahead to several years of
extraordinary profits.
Omeprazole was the only PPI that would
have a generic version during the period covered by the Patents.
Apotex also would have expected to enjoy increased sales of its
other generic products as a result of its launch of omeprazole - the only generic PPI molecule in one of the most widely-used
class of drugs in the United States. 34
34
As Apotex explained in
The sixth Georgia-Pacific factor is “the effect of selling the
patented specialty in promoting sales of other products of the
104
opposing Astra’s requested injunction in 2007, sales of
omeprazole help Apotex sell other products, since “customers
tend to expand upon their current base of sales from current
suppliers.”
In other litigation, an Apotex witness explained
that “the presence of a blockbuster in the product line is
perhaps the single most effective way to increase sales across
all product lines.”
Astra does not advance any definite figure
for the increased sales Apotex would have expected its other
products to enjoy thanks to omeprazole, and Astra does not
include this consideration in the profit calculations that
support its reasonable royalty figure.
Nevertheless, this
effect does lend support for the idea that Apotex would have
been willing to pay a substantial royalty for the right to sell
omeprazole.
2.
Apotex’s ability to avoid infringement
In the hypothetical negotiation, Apotex would have
understood that its formulation infringed the Patents.
As of
that time, Apotex would not have had any confidence that it
could create a non-infringing product, much less do so on a
licensee.”
Whitserve, 694 F.3d at 27 n.11.
105
timetable that would permit it to become a strong player in the
generic PPI market.
Time was not on Apotex’s side.
KUDCo had entered the
market about a year earlier, and two other generic manufacturers
had launched their products three months earlier.
Thus, Apotex
was already faced with the hurdle of being the fourth generic in
a market in which every additional generic manufacturer would
face stiffer competition.
Generally, pharmacies only have one
generic version of a drug on hand.
Each generic version of a
drug is competing with every other generic to win that
placement.
A late entrant thus has the burden of displacing
established commercial relationships to get that business.
In postponing entry into this lucrative market while
engaged in an effort to develop a non-infringing product, Apotex
risked losing even the fourth pole position.
Other non-
infringing generic omeprazole products might enter in the
interim, or other generic products might enter at risk.
With
extended delay, therefore, Apotex risked being effectively
closed out of the market even if it succeeded in finding a noninfringing formulation.
During the negotiation, Apotex would have been quite
pessimistic about its chances of developing a non-infringing
106
formulation.
It had spent close to two years trying to develop
microtablets to place in a capsule because it thought that that
would be a clear path toward a non-infringing product.
Apotex
abandoned that path because it could not produce a stable
bioequivalent formulation.
It had tried to work at an intense
pace to be the first ANDA filer in this extraordinarily
lucrative market, and it missed that mark by three years.
Genpharm filed in December 1997; Apotex filed the eighth ANDA in
December 2000.
And Apotex knew not just from its own experience, but also
from the experience of others that finding a non-infringing
formulation was tough.
Three of the four defendants in the
First Wave litigation were found to be infringers, one of them
because its product, like Apotex’s product, had an inert
Thus, even a product that was
subcoating that formed in situ.
not designed to have a subcoating ran the risk of infringement
if a subcoating formed in situ through the chemical reactions of
the ingredients.
As discussed in detail above, Apotex suggests that it could
have avoided infringement by making any one of three changes to
its formulation:
modifying its pellets, copying a competitor’s
non-infringing formulation, or using a microtablet formulation.
107
Apotex has not shown that these approaches were available to it
or would result in a non-infringing product.
See Spectralytics
Inc. v. Cordis Corp., 649 F.3d 1336, 1346 (Fed. Cir. 2011).
As for Apotex’s proposals for tinkering with the
ingredients in its pellets, it is pure speculation whether any
of its various proposals would create a stable, bioequivalent
product that was non-infringing.
Apotex has never asked one of
its many experts to try to create the revised formulation, much
less to create and test it.
See SynQor, Inc. v. Artesyn Techs.,
Inc., 709 F.3d 1365, 1382 (Fed. Cir. 2013) (where an alleged
substitute is not on the market, “the accused infringer has the
burden to overcome the inference that the substitute was not
‘available’”) (citation omitted).
There is a reason that Apotex chose the ingredients that it
did for its pellets following six years of research and testing.
Those ingredients created a successful product.
This is no easy
task given the challenges of working with the omeprazole
molecule and delivering it sufficiently intact to the part of
the body in which it is most effective.
By the end of the trial, Apotex had largely abandoned its
argument that it could have altered the infringing formulation
108
successfully. 35
In any event, during the trial it described
three potential changes to its formulation.
remove magnesium hydroxide from the core.
The first was to
But as Dr. Sherman
admitted, that ingredient helped stabilize the core and had been
the best stabilizer that Apotex could find after testing
alternatives.
The next was to avoid the formation of a
subcoating by replacing the binding agent PVP with HPC.
HPC is
an inferior binder, however, and Apotex did not show that its
use would eliminate the formation of a subcoat in situ.
Finally, Apotex suggests that it could have explored a change to
the enteric coating of the pellet.
But, again, it has not shown
that its proposed change would not result in the formation of an
in situ subcoating, and its proposal is that the coating it has
acknowledged to be superior -– Eudragit L30D –- be replaced by
an inferior product that would require a solvent other than
water to be used in the manufacturing process.
For several good
reasons, Apotex had already rejected, however, the use of
organic solvents in the manufacturing process.
Even if one of these proposals could have resulted in the
creation of an alternative, non-infringing formation that would
35
In its summation, Apotex only discussed one of the three
proposed modifications to the pellets.
109
receive FDA approval as a stable, bioequivalent form of
omeprazole, of course that creation, testing, and approval would
have taken time.
First, Apotex would have expected that it
would take it at least a year to identify a successful new, noninfringing formulation.
After all, after six years of work it
had been unable to arrive at such a solution.
Interestingly,
despite determinations in October 2002 that the generic
omeprazole formulations of three manufacturers infringed the
Patents, there is no evidence that those manufacturers -Genpharm, Andrx, and Cheminor -- ever looked for or found a noninfringing alternative.
The three month stability tests and the
FDA approval process for any new Apotex product would easily
consume another year, assuming again that all went smoothly.
At
its most optimistic, Apotex would thus have expected a delay of
at least two years before it would be able to begin selling an
approved product.
Apotex would not have been sanguine about its
ability to tinker with its formulation of the pellets as an
alternative route to entering the generic market; as Dr. Meyer
showed, a two-year delay in entering the market would have cost
Apotex 59.2% of its profits.
Apotex also argues that it could have copied the
formulations of other generic manufacturers, but the only
110
generic formulation which a court had found was non-infringing
as of November 2003 was the KUDCo formulation.
That was not
available to Apotex to copy, as it was protected by KUDCo
patents. 36
Rite-Hite, 56 F.3d at 1548 (product protected by
patent is not “available”).
Finally, Apotex has argued that it could have revived its
work on the omeprazole tablet and used that clearly noninfringing formulation to enter the market.
But, as explained
above, despite its best efforts Apotex was unable to solve the
problems associated with its tablet, and it has provided no
reason to believe that it would have been any more successful in
that work in 2003 or 2004 than it had been in the 1990s.
Apotex argues that analyzing its negotiating position by
reference to the cost and delay associated with developing a
non-infringing alternative formulation is improper.
Doing so,
Apotex says, leads to a royalty rate that is unfairly based on
the “hold-up” value of patents rather than their actual economic
advantage over alternative, non-infringing approaches.
Apotex
here relies on a 2011 FTC Report, which notes that “[a]
36
Apotex suggests that it could have copied the Lek and Mylan
formulations since they were also non-infringing. But as of
November 2003, Apotex did not know those two formulations would
be found in 2007 to be non-infringing formulations. In any
event, they were also protected by patents.
111
reasonable royalty damages award that is based on high switching
costs, rather than the ex ante value of the patented technology
compared to alternatives, overcompensates the patentee.”
Federal Trade Comm’n, The Evolving IP Marketplace: Aligning
Patent Notice and Remedies with Competition 190 (2011),
available at http://www.ftc.gov/os/2011/03/110307
patentreport.pdf.
Apotex therefore suggests that in the
“hypothetical world” in which the negotiation occurs it should
be assumed that Apotex would have begun developing a noninfringing alternative formulation in 2000, when it filed its
ANDA.
Apotex’s argument contradicts settled law.
The hypothetical negotiation approach to determining a
reasonable royalty posits that a negotiation occurred at a
particular time, which in this case has been stipulated to be
November 2003.
As the Federal Circuit explains, “[t]he
hypothetical negotiation requires the court to envision the
terms of a license agreement reached as the result of a supposed
meeting between the patentee and the infringer at the time the
infringement began.”
supplied).
Rite-Hite, 56 F.3d at 1554 (emphasis
This framework is incompatible with Apotex’s
suggestion that it should be assumed to have been pursuing noninfringing alternatives as early as 2000.
112
Hanson v. Alpine Valley Ski Resort, Inc., 718 F.2d 1075
(Fed. Cir. 1983), is on point.
In that case, the court rejected
the argument that a royalty was unreasonable because it exceeded
the cost of a non-infringing alternative.
Id. at 1081.
In
doing so, the court observed that
Alpine could have avoided infringement, and paying
royalties therefor, by purchasing non infringing
machines . . . . It chose, however, to purchase and
use [the] infringing machines. Having followed that
course, it cannot invalidate an otherwise reasonable
royalty on the claim that by hindsight it would have
been better off if it had purchased the non-infringing
. . . machines.
Id. at 1081-82.
The hypothetical negotiation is hypothetical in the sense
that the negotiation itself is imaginary, not in that it allows
the parties to construct an entirely imaginary world that
ignores the facts as they existed at the date of infringement.
Those facts show that Apotex did not have a non-infringing
alternative formulation ready and waiting.
That this was the
situation in which Apotex found itself in November 2003 is one
of the most salient features of the negotiating dynamic in this
case and may not now be ignored.
The cases on which Apotex relies, SK Hynix Inc. v. Rambus
Inc., No. C-00-20905 (RMW), 2013 WL 1915865, at *19-20 (N.D.
Cal. May 8, 2013), and Microsoft Corp. v. Motorola, Inc., No. 10
113
Civ. 1823 (JLR), slip op. (W.D. Wash. April 25, 2013), are not
to the contrary.
Those cases deal with the special situation in
which a technical standard is set for an industry that puts one
patent holder “in a position to ‘hold up’ industry participants
from implementing the standard.”
Qualcomm Inc. v. Broadcom
Corp., 548 F.3d 1004, 1010 (Fed. Cir. 2008).
In such a
situation, the organizations setting the technical standards
often “require participants to disclose and/or give up
[intellectual property rights] covering a standard.”
Id.
Patent holders in such situations also typically agree to
license their patents on a “reasonable and non-discriminatory”
basis, a standard that courts will enforce by, for instance,
allowing other parties to practice a patent as long as they are
willing to pay a reasonable and non-discriminatory royalty.
Microsoft Corp. v. Motorola, Inc., 696 F.3d 872, 876-77 (9th
Cir. 2012) (citation omitted).
These considerations are simply
inapplicable here, as the Patents do not cover a standard
technology.
While the Georgia-Pacific factors do include a
consideration of “the utility and advantages of the patent
property over the old modes or devices,” Whitserve, 694 F.3d at
27 n.11, they do not set the economic utility of the patent
114
property in the abstract as an upper limit on the royalty
figure.
Furthermore, taking into account the costs and risks
Apotex would incur by switching formulations in November 2003
does take into account the patented technology’s advantages over
non-infringing alternatives.
Notably, Apotex began its efforts to formulate omeprazole
by working on a micro-tablet formulation that would not have
infringed the patents-in-suit.
As Dr. Beach testified, and as
numerous internal Apotex documents show, Apotex switched to the
infringing pellet formulation after its microtablet formulation
encountered repeated obstacles and delays.
microtablet formulation was a failure.
In short, Apotex’s
This history
demonstrates that the patented formulation did have “utility and
advantages” for Apotex.
Consideration of this Georgia-Pacific
factor does not require ignoring the time and money Apotex
invested in the patented formulation, or the time and money it
would have spent in any effort to develop an alternative.
B.
Astra’s position
In contrast to Apotex’s eagerness to obtain a license from
Astra, Astra would have had no desire in November 2003, or
indeed at any time, to issue a license to Apotex.
Rather, it
would have believed that any license granted to a generic
115
omeprazole supplier would be likely to alter the dynamic in the
prescription PPI market in a way that would damage Astra
financially.
Since it is necessary, however, to assume that
both parties were willing to enter into a license, Astra has
shown through overwhelming evidence that it was in the driver’s
seat in the negotiations and would have required Apotex to pay a
hefty portion of its profits for a license.
To begin with, Astra had no program to license its
Patents in connection with generic prescription omeprazole, and
had no interest in doing so.
Astra’s position in a hypothetical
licensing negotiation with Apotex would have been shaped by two
primary concerns: first, ensuring that the license would
adequately compensate it for any economic harm it would suffer
as a result of Apotex’s entry into the market, and second,
maximizing the licensing fee it could expect to receive from
Apotex, based on its understanding of how much Apotex would be
willing to pay.
Astra would have believed in November 2003 that the entry
of a licensed generic omeprazole might lead to the complete
conversion of the prescription omeprazole market to generic
products through the imposition of MACs.
If this happened, it
would have had a significant impact on two Astra products,
116
Prilosec® and Nexium®.
Prilosec® would probably have lost any
remaining positions on formularies and Nexium® would have been
confronted with fiercer competition for formulary placement and
other favorable treatment.
If Apotex had entered the market with a license, then there
would have been two generic omeprazole manufacturers who could
engage in price competition without any constraint imposed by
litigation risk:
KUDCo and Apotex.
The Lek and Mylan products
had been launched at risk in August, and their competition with
each other and with KUDCo had brought down the price of generic
omeprazole, but by a relatively modest amount given the historic
experience of price declines when generic products can compete
without restriction with each other.
With a license, Apotex
would have been expected to change that market dynamic.
Moreover, the Lek and Mylan products had a three month head
start on Apotex in the market.
While Apotex would have wanted
to keep its price as high as possible, it would confront the
classic interaction of supply and demand.
Would its profits
increase with a substantially lower price because it would
capture enough market share to offset the price decline?
Astra would therefore have had to consider with care
Apotex’s pricing strategy.
Astra’s fears in this regard would
117
have been well founded.
According to a study of IMS retail
sales data from 1999 to 2004 performed by the FDA’s Center for
Drug Evaluation and Research, a second unconstrained entrant
typically brings a current generic price down from 94% of the
brand’s price to 52% of the brand’s price.
After four generic
entrants, the generic price is often 39% of the brand’s original
price.
As it turned out, when Apotex entered the market without
a license it priced its product to obtain roughly a 30% market
share by 2005.
When the price of a generic product is low enough and there
is assured supply, it is customary for insurers to impose a MAC.
If this severe limitation on the reimbursement of omeprazole
products had been put in place, then Astra would have had to
consider the tsunami that would engulf the entire prescription
PPI market.
In such a situation, Astra would probably have been faced
with something like the following two choices for Prilosec®.
It
could either turn Prilosec® into a generic (an unpalatable
option for a company like Astra) and compete with other generic
products, or it could sell Prilosec® as a very low-priced
branded product to those few consumers who were willing to pay
118
cash.
After all, once a MAC is imposed, it would be unrealistic
to expect Prilosec® to remain on any formulary.
With the elimination of Prilosec® from the market, Astra
would have lost access to one of its strategies for
transitioning patients to Nexium®.
Moreover, with the
imposition of a MAC for omeprazole, Astra could expect that the
competition among branded PPIs would become ever fiercer.
They
would be jockeying with greater intensity to be the one
preferred branded PPI on Tier II, and to avoid imposition of
step therapy and other adverse actions.
To succeed, they would
have to make a substantially increased commitment to their
rebate program so that they remained as attractive as possible
to TPPs.
With a MAC imposed, Nexium® would have lost any
chance, even through a generous rebate program, to remain price
competitive with the generic PPI.
While Astra would have had to consider whether a MAC would
be imposed on generic omeprazole as a result of Apotex’s entry
into the market with a license, it would also have considered
less dramatic impacts from that entry.
Under any scenario,
however, Astra would have reasonably expected that Apotex’s
licensed entry would make it substantially more difficult to
support Prilosec®’s position on drug formularies and that the
119
reduced prices for generics would have increased the demands on
Astra to provide rebate support for Nexium®.
Thus, Astra would have tried to calculate whether the
expected income stream from an Apotex license would offset lost
sales from Prilosec® and the increase in financial support
required to keep Nexium®’s market share stable or to improve it.
While Prilosec®’s market share had plummeted by late 2003, it
still had roughly 5% of the PPI market and generated net sales
of $865 million in 2003 and $361 million in 2004.
Astra’s
rebate program for Nexium® consumed roughly 17% of its gross
sales by the end of 2003 and 28% of its gross sales by the end
of 2004.
These are the financial parameters in which Astra was
operating.
In this context, it is reasonable to conclude that Astra
would not have licensed Apotex for anything less than 50% of
Apotex’s profits.
Apotex earned profits on omeprazole of
$152,043,989 between November 2003 and October 2007.
Using a
50% royalty, this results in a fee of $76,021,994.50.
While there are no perfect benchmarks for this licensing
fee, a 50% licensing fee fits comfortably within the range of
negotiations that occurred in connection with the patents-in-
120
suit.
These negotiations included actual licenses and offers to
settle or settlements of litigation.
Astra received a step royalty ascending from 7% to 40%,
with a blended royalty rate of 20% to 23%, from P&G for the sale
of Prilosec OTC®.
Unlike Apotex’s omeprazole, this is a product
that Astra expected to benefit its product line.
From the
1990s, Astra had envisioned Prilosec OTC® as a product that
would take sales away from generic omeprazole and encourage
consumers to try Nexium® if Prilosec OTC® proved unsuccessful.
Moreover, while Astra provided the product at cost to P&G, this
required no great investment, and P&G bore the entire burden of
using its best efforts to market the product.
Thus, Astra
received a handsome royalty for a product that was an essential
part of its long term PPI strategy.
While Apotex has emphasized
that the base royalty of 7%, P&G and Astra expected that the
sales for the first three years would surpass the threshold that
triggered the 40% rate, and those expectations were met.
Andrx offered in 2005 to pay Astra 70% of its profits to
enter the market in that year as the sixth generic.
The 70%
figure was for the sale of 40mg generic omeprazole, a product
not yet sold on the market.
It offered Astra 50% of its profits
on 10mg and 20mg generic omeprazole with a $25 million minimum
121
royalty for the 180 days following its first commercial sale.
Even though these royalties were offered to settle litigation,
including the claim that Andrx had infringed the Patents through
prior manufacture of commercial quantities of omeprazole, they
serve as a marker of the value of licensing rights.
And of course, the settlement agreement between Teva and
Astra in 2010 resulted in a payment to Astra of the equivalent
of 54% of Teva’s profits from its infringing sales.
Moreover,
back in 2003, when Teva exercised its option to add omeprazole
to its strategic alliance agreement with Impax, it agreed to pay
Impax on a sliding scale from 35% to 50% of its profits after
the two year mark.
Each of these markers provides some support for the choice
of 50% at the reasonable licensing fee for Apotex.
In contrast,
Apotex has not shown that any of the cross-licensing agreements
or other settlements that did not involve the sale of omeprazole
provide a relevant marker.
Its contention that Astra would have
agreed to a licensing fee no greater than 7% must be dismissed
out of hand.
In addition to those already addressed, Apotex makes
essentially four arguments why the 50% licensing fee is too
high.
It contends it would have been unwilling to pay that
122
much; that any license would have been non-exclusive and
therefore of limited value; that greater weight must be given to
the impact of Prilosec OTC® on Astra’s other products; and that
its forecasts from before its launch did not accurately reflect
the costs attributable to omeprazole and therefore provided an
artificially high profit margin.
In summation, Apotex asserted bluntly that generic drug
manufacturers simply do not pay licensing fees of 50% of
profits.
It was unable to point to any evidentiary support for
that contention, and in fact the record contains evidence that
is directly contrary.
Apotex’s products at the time typically
achieved a gross margin of 31 to 48%, or an average across all
products of 40%.
Apotex projected a gross margin of over 92%
from its sale of unlicensed omeprazole.
Based on those
estimates, a 50% royalty would have left it with an anticipated
gross margin of 46%, close to the top of this range.
And, as
the rough benchmarks described above show, generic drug
manufacturers are willing indeed to pay fees in the range of 50%
of their profits or higher.
As for Apotex’s argument that its license would not have
been exclusive and therefore would have been worth far less than
50% of its profits, this is again belied by the various other
123
benchmark licenses discussed at trial.
Notably, Andrx offered
to buy a license from Astra for 70% of its profits on 40mg
omeprazole, for which it would have been the exclusive generic
supplier, and 50% of its profits on the 10 and 20mg doses, for
which it would have been the sixth generic supplier.
Teva,
which was never an exclusive supplier of omeprazole, settled its
claims with Astra for what amounts to 54% of its profits.
And,
without any license (let alone an exclusive one), Apotex
anticipated achieving gross margins of 92%.
Apotex emphasized at trial that Astra introduced one of the
major PPI competitors when it introduced Prilosec OTC®, and that
that launch occurred in September 2003, just months before the
hypothetical negotiation.
Apotex is quite right that Prilosec
OTC® had a major impact on the PPI market.
While it had very
minimal effect on Nexium®, it dealt a blow to omeprazole’s
market share from which took about a year to recover.
But,
Apotex failed to show at trial that this impact on sales to
consumers actually had any significant impact whatsoever on
formulary position or the price competition within the
prescription PPI market.
After all, Prilosec® did not have to
compete for placement on the shelves behind the pharmacy
counter; it was sold in the drugstore aisles and competed with
124
other OTC products for shelf space there.
Very few formularies
included Prilosec OTC® in 2003 and this remained true for a long
time thereafter.
As a consequence, generic prescription
omeprazole and branded prescription PPIs were competing with
each other and not with Prilosec OTC® through rebates and
product differentiation to avoid adverse actions by formularies
and for favorable formulary placement.
The existence of
Prilosec OTC® in the marketplace simply would not have altered
in any fundamental way the calculus between the parties as they
negotiated a license.
Their basic concern would have been how
the entry Apotex’s licensed generic product would alter the
field of play in the prescription market, and specifically, the
policies of TPPs.
Apotex’s final argument, that its projections in the months
leading up to its launch did not contain meaningful cost
estimates and therefore overstated the profit margin it expected
to earn, is similarly unavailing.
At the outset, there is no
evidence in the record to support the theory that the cost
estimates in these internal projections were not intended to be
accurate.
Apotex is of course correct that its actual profit
margin fell short of its projections; while it projected a 92%
profit margin, it ended up earning a roughly 75% margin on its
125
infringing sales.
Even if Apotex paid 50% of its profits out of
this lower margin, it is still left with a profit margin of 36%,
which is solidly in the range of 31 to 48% margins it typically
earned on its products at the time.
C.
Two remaining factors:
Factors 2 and 14
Two Georgia-Pacific factors remain to be addressed.
Factor
2 requires consideration of the rates paid by the licensee for
the use of comparable patents.
Just as Astra had no practice of
licensing generic drug products that competed directly with its
branded products, Apotex had no practice of entering the generic
market via a licensing agreement with the branded drug.
Apotex
has not placed any particular emphasis, however, on this issue
at trial.
Since the analysis conducted above gives careful
consideration to Apotex’s general experience in selling generic
products, including the gross margin it customarily achieves
through those sales, the absence of a practice of negotiating
licenses does not alter the conclusion that it would have
readily paid a licensing fee of 50% here.
Factor 14 requires the Court to examine with care the
opinion testimony of qualified experts.
here.
That has been done
Where their opinions were well supported by the evidence
they proved helpful to the analysis of the issues.
126
On many
background facts, such as the operation of the prescription
market or the procedures followed by the FDA in reviewing a PAS,
the parties’ experts largely agreed.
On occasion, the experts
presented by Apotex did not provide well supported testimony or
provided testimony that was substantially undercut by cross
examination.
When that occurred, their opinions were accorded
less weight.
This Opinion has carefully considered all of the testimony,
including that given by lay and expert witnesses, and has found
substantial support in their testimony and the trial exhibits
for the findings herein.
Astra has shown not just by a
preponderance of the evidence but convincingly that the
hypothetical licensing fee to which Astra and Apotex would have
agreed would have been at least 50% of the Apotex gross margin
from its sales of omeprazole.
Applying that rate to the
infringing sales made by Apotex, Apotex owes $76,021,994.50 plus
pre-judgment interest.
CONCLUSION
Following trial, Astra has carried its burden of showing
that it is entitled to damages in the amount of $76,021,994.50
127
plus pre-judgment interest.
The parties shall present a
proposed judgment by December 6, 2013.
SO ORDERED:
Dated:
New York, New York
November 26, 2013
__________________________________
DENISE COTE
United States District Judge
128
Disclaimer: Justia Dockets & Filings provides public litigation records from the federal appellate and district courts. These filings and docket sheets should not be considered findings of fact or liability, nor do they necessarily reflect the view of Justia.
Why Is My Information Online?