PURDUE PHARMACEUTICAL PRODUCTS L.P. et al v. ACTAVIS ELIZABETH LLC
Filing
423
(REDACTED) OPINION. Signed by Judge Jose L. Linares on 3/27/15. (sr, )
NOT FOR PUBLICATION
UNITED STATES DISTRICT COURT
DISTRICT OF NEW JERSEY
PURDUE PHARMACEUTICAL PRODUCTS
L.P. et al.,
Plaintiffs,
Civil Action No. 12-5311 (JLL) (JAD)
OPINION
v.
ACTAVIS ELIZABETH LLC, et al.,
Defendants.
LINARES, District Judge.
This case involves the issues of infringement and validity of three patents covering
Plaintiffs’ product Intermezzo®. Intermezzo® is a drug manufactured for the treatment of
insomnia when middle-of-the-night awakening is followed by difficulty returning to sleep. After
careful consideration of the evidence presented at a bench trial held December 1 -15, 2014, the
Court finds as follows: As to the ’131 patent, Defendants have met their burden of proving this
patent is invalid as obvious, but failed to prove that the claim element “without residual sedative
effects” is invalid as indefinite. Plaintiffs have proved by a preponderance of the evidence that the
asserted claims of the ’131 patent are infringed by all Defendants. As to the ’628 patent, Plaintiffs
have failed to meet their burden of proving infringement as to Defendants, DRL and Actavis.
Plaintiffs, however, have met their burden of proving that Novel infringes the ’628 patent. While
Defendants have failed to prove this patent is invalid as anticipated, the ’628 patent is invalid as
obvious. As to the ’809 patent, Plaintiffs have met their burden of proving infringement as to
Defendants, DRL and Novel. Defendants, on the other hand, have proved by clear and convincing
evidence that the ’809 patent is obvious and therefore invalid. This Opinion articulates the basis
for each of these conclusions.
INTRODUCTION
This is an infringement action brought by Plaintiffs relating to the patents covering
Intermezzo®. This action was commenced as a result of Defendants each filing an Abbreviated
New Drug Application (“ANDA”) pursuant to the Hatch-Waxman Act, seeking FDA approval to
sell a generic version of Intermezzo® prior to the expiration of the relevant patents. This Court
has subject matter jurisdiction over this action under 28 U.S.C. §§ 1331 and 1338(a). No party
contests personal jurisdiction or venue for the purposes of this civil action.
BACKGROUND 1
I. The Parties
Plaintiffs, Purdue Pharma L.P., and Purdue Pharmaceutical Products L.P., are the current
holders of New Drug Application No. 022328, for sublingual tablets containing 1.75 mg and 3.5
mg of zolpidem tartrate.
These Plaintiffs market the approved drug under the tradename
Intermezzo®. Plaintiff, Transcept Pharmaceuticals, Inc., (“Transcept”), is the owner of the
relevant patents where Purdue Pharma L.P., and Purdue Pharmaceutical Products L.P. 2 are
exclusive licensees under these patents. Plaintiffs offer that Transcept is currently known as
“Paratek Pharmaceuticals, Inc.” (See e.g. Pls.’ Proposed Findings of Fact, (“PFOF”) ¶42). The
1
The facts set forth herein are the Court’s findings of facts which are based on the Court’s
observations and credibility determinations of the witnesses who testified and a thorough review
of all the evidence admitted at trial.
2
Transcept Pharmaceuticals, Inc., Purdue Pharma L.P., and Purdue Pharmaceutical Products
L.P, will collectively be referred to as “Plaintiffs.”
2
Court, however, for purposes of this Opinion, refers to the named assignee of the patents-in-suit
as “Transcept.”
While there are five Defendants in this action, two of said Defendants, Par Pharmaceutical,
Inc. (hereinafter “Par”), and TWi Pharmaceuticals, Inc. (hereinafter “TWi”), entered into
stipulations wherein both agreed to be bound by the outcome of the trial without them actually
participating. That is, TWi “agree[s] to be bound by the Final Judgment … including any related
injunctions, of the District Court in the Intermezzo Action resolving all claims and counterclaims
of infringement, validity and enforceability of the ’131 patent following litigation on the merits.”
(Stipulation, ECF No. 332, ¶4). Similarly, Par, “agree[s] to be bound by the Final Judgment …
including any related injunctions, of the District Court in the Intermezzo Action resolving all
claims and counterclaims of infringement, validity and enforceability of the ’131 and ’809 patents
following litigation on the merits.” (Stipulation, ECF No. 323, ¶3). Both TWi and Par stipulated
to personal jurisdiction for purposes of this action. The remaining Defendants in this action are
Actavis Elizabeth, LLC (hereinafter “Actavis”), Dr. Reddy’s Laboratories, Inc., and Dr. Reddy’s
Laboratories, LTD, (hereinafter collectively “Dr. Reddy’s” or “DRL”), and Novel Laboratories,
Inc., (hereinafter “Novel”).
II. Intermezzo and the Patents-in-Suit
Intermezzo® is a drug manufactured for the treatment of insomnia when middle-of-thenight (or “MOTN”) awakening is followed by difficulty returning to sleep. (PTX-225 at 1).
Intermezzo® is intended for use only if the patient has four hours or more remaining before the
planned time of waking. Intermezzo® is in the form of a tablet that is placed under the tongue to
disintegrate. This formulation is for transmucosal absorption. The three relevant patents at issue
3
covering Intermezzo are U.S. Patent No. 8,242,131 (the “’131 patent”), U.S. Patent No. 7,682,628
(the “’628 patent”), and U.S. Patent No. 8,525,809 (the “’809 patent”). Nikhilesh N. Singh is the
sole named inventor of each of the three relevant patents, except regarding the ’131 patent only,
where Sathasivan Indiran Pather is named as a co-inventor.
The ’809 patent is a patent indicated for the treatment of MOTN insomnia. The ’131 patent
is directed to a method of treating MOTN insomnia where the ’628 patent is directed to a method
of treating insomnia. The ’131 Patent is entitled “Methods of Treating Middle-of-the-Night
Insomnia” and was issued by the PTO on August 14, 2012.
The ’628 Patent is entitled
“Compositions for Delivering Hypnotic Agents Across the Oral Mucosa and Methods of Use
Thereof” and was issued by the United States Patent and Trademark Office (“PTO”) on March 23,
2010. The ’809 Patent is entitled “Compositions for Treating Insomnia” and was issued by the
PTO on August 28, 2012. The priority dates are: 1) for the ’628 Patent, February 17, 2004; 2) for
the ’131 Patent, May 25, 2005; and 3) for the ’809 Patent, May 25, 2005.
III. The Claims at Issue and Relevant Markman Construction
A. Claims of the ’131 Patent
Plaintiffs claim all Defendants will induce infringement of Claims 8, 10, 18 and 19 of the
’131 patent. Because Claims 1 and 12 of the ’131 patent are independent claims, and therefore
Claims 8, 10, 18, and 19 depend therefrom, a detailed analysis of Claims 1 and 12 is applicable to
the infringement and validity inquiries. Claims 1 and 12 are distinct insofar as Claim 1 applies to
non-elderly patients and Claim 12 applies to elderly patients. Defendants argue that that the claims
of the ’131 patent are obvious. Defendants also argue that limitation found in Claims 1 and 12,
namely, “without residual sedative effects” is indefinite. These claims state the following:
4
Claim 1: A method of treating middle-of-the night insomnia
in a non-elderly patient without prophylactically
administering zolpidem, comprising: dosing the patient with
a pharmaceutical composition comprising about 0.5 to about
4.75 mg of zolpidem hemitartrate or a molar equivalent
amount of a pharmaceutically acceptable form of zolpidem,
wherein the pharmaceutical composition is substantially free
of other hypnotic agents, wherein the patient awakens from
sleep and desires to resume sleep for less than 5 hours,
wherein the step of dosing the pharmaceutical composition
is performed after the patient awakens from sleep, and
wherein the pharmaceutical composition permits the patient
to awaken at a time about four hours after dosing without
residual sedative effects.
Claim 12: A method of treating middle-of-the-night
insomnia in an elderly patient without prophylactically
administering zolpidem, comprising dosing the patient with
a pharmaceutical composition comprising about 1.5 to 2.5
mg of zolpidem hemitartrate or a molar equivalent amount
of a pharmaceutically acceptable form of zolpidem, wherein
the pharmaceutical composition is substantially free of other
hypnotic agents, wherein the patient awakens from sleep,
and desires to resume sleep for less than 5 hours, wherein the
step of dosing the pharmaceutical composition is performed
after the patient awakens from sleep, and wherein the
pharmaceutical composition permits the patient to awaken at
a time about four hours after dosing without residual
sedative effects.
This Court construed “without residual sedative effects” to mean “with no or minimal subjective
feelings of sedation, as evaluated by: (a) testing acceptably in at least one test exploring
psychomotor performance, attention, information processing, and memory used by those of skill
in the art; and/or (b) demonstrating plasma levels of zolpidem, at an appropriate time point, below
about 20 ng/ml.” (Opinion, ECF No. 185 at 5-7). A list of appropriate psychomotor performance,
attention, information processing, and memory used by those of skill in the art are exemplified in
the patent specification.
5
Claim 8 depends from Claim 1 and is substantially identical to such except it refers to the
3.5mg dose. (JTX 1, Claim 8). Claim 10 of the ’131 Patent recites: “The method of claim 8,
wherein the pharmaceutical composition provides delivery of zolpidem across the patient’s oral
mucosa.” (JTX 1, Claim 10). Claim 18 depends from Claim 12 and is substantially identical to
such except it refers to the 1.75mg dose. Claim 19 of the ’131 Patent recites: “The method of
Claim 18, wherein the pharmaceutical composition provides delivery of zolpidem across the
patient’s oral mucosa.”
B. Claims of the ’628 Patent
Plaintiffs claim Defendants will induce infringement of Claims 9, 16 and 17 of the ’628
patent. The ’628 patent contains an independent Claim 1 which states the following:
Claim 1: A method for treating insomnia, comprising the
steps of: administering a solid pharmaceutical composition
comprising zolpidem or a pharmaceutically acceptable salt
thereof to a subject prone to insomnia, the pharmaceutical
composition further comprising a buffer, wherein the buffer
raises the pH of saliva to a pH of about 7.8 or greater,
wherein zolpidem is absorbed across a permeable membrane
of the subject’s oral mucosa, and wherein at least 75% of the
solid pharmaceutical composition dissolves within 10
minutes or less within an oral cavity following
administration.
Claim 9 of the ’628 Patent recites as follows: “The method of claim 1, wherein the buffer comprises
a carbonate buffer and a bicarbonate buffer.” (JTX 003, Claim 9). Claim 16 of the ’628 Patent
recites as follows: “The method of claim 1, wherein the zolpidem or pharmaceutically acceptable
salt thereof is in an amount from about 1 mg to about 5 mg.” (JTX 003, Claim 16). Claim 17 of
the ’628 Patent recites: “The method of claim 1, wherein the zolpidem or pharmaceutically
acceptable salt thereof is in an amount from about 2 mg to about 5 mg.” (JTX 003, Claim 17).
6
C. Claims of the ’809 Patent
Claim 1 of the ’809 Patent, upon which asserted Claims 11, 17, and 18 depend, recites as
follows:
A solid unit dosage composition for the treatment of MOTN
insomnia, said composition comprising an effective amount of
zolpidem or a salt thereof, formulated for delivery of zolpidem
across a subject’s oral mucosa, wherein said effective amount is an
amount of less than 1.30 x 10-5 moles of zolpidem, and between
about 25 ng/mL and about 50 ng/mL within 20 minutes of
administration, when evaluated in an appropriate patient population.
Claim 11 of the ’809 Patent recites: “The solid unit dosage composition of claim 1, wherein the
zolpidem is delivered across at least one of the sublingual or buccal mucosa.” Claim 17 of the ’809
Patent recites: “The solid unit dosage composition as in any of claims 1–6, 15, or 16, containing
about 1.75 mg of zolpidem hemitartrate.” Claim 18 of the ’809 Patent recites: “the solid unit dosage
composition as in any of claims 1–6, 15, or 16, containing about 3.5 mg of zolpidem hemitartrate.”
Claim 12 of the ’809 Patent, upon which asserted Claim 22 depends, recites:
A solid unit dosage composition for the treatment of MOTN
insomnia, said composition comprising an effective amount of
zolpidem or a salt thereof and at least one buffering agent,
formulated for delivery of zolpidem across a subject’s oral mucosa,
wherein said effective amount is 0.5 to 4.75mg of zolpidem
hemitartrate, and is an amount sufficient to produce a plasma
concentration between about 25 ng/mL and about 50 ng/mL within
20 minutes of administration, when evaluated in an appropriate
patient population.
Claim 22 of the ’809 Patent recites: “The solid unit dosage composition of claim 12, further
comprising a binary buffer system that raises the pH of said subject’s saliva to a pH greater than
about 8.5, irrespective of the starting pH of saliva.”
“Binary buffer system” was construed by the Court to mean “a system used to maintain
and/or achieve an approximate pH range comprising at least one proton-donating component and
at least one proton accepting component.” (Opinion, ECF No. 185 at 26). This element is similar
7
to the “buffer” element of the ’628 patent except it is somewhat more specific as it refers to a
“system.”
IV. Procedural History
The original Complaint in this action was filed on August 23, 2012 against Defendant,
Actavis. By February 26, 2014, all actions brought by Plaintiffs relevant to Intermezzo® and
patent infringement were consolidated. Each of the Defendants represented to the Court that it
will not launch its ANDA product prior to March 31, 2015. The consolidated action against
Actavis, DRL, and Novel was heard by the Court at a bench trial on December 1–15, 2014.
V. MOTN Insomnia
Insomnia is a common malady that occurs in approximately one third of the adult
population. (See JTX 016). The term insomnia is used to describe all conditions related to the
patient’s perception of inadequate or non-restful sleep. (Tr. 7.136:17-21 (Winkelman)). 3 Insomnia
possesses three elements of difficulty: falling asleep, staying asleep, or waking up too early in the
morning. (Tr. 7.133:16-21 (Winkelman)). Prior to 2005, the method of treating all types of
insomnia was primarily through prophylactic administration to prevent insomnia rather than on
the “as needed” basis described in the ’131 patent. Middle-of-the-night insomnia was construed
by the Court to be a “condition wherein a subject, after falling asleep, awakens and has difficulty
returning to sleep.” (ECF No. 92 at 2). In sum, those who suffer from MOTN insomnia suffer
from frequent nocturnal awakenings.
3
“Tr.,” refers to the bench trial transcript.
8
While prophylactic administration made sense for those who had trouble initially falling
asleep, patients suffering from MOTN insomnia irregularly were in some cases medicating
themselves unnecessarily because whether they would wake in the middle-of-the-night on that
particular night was unpredictable. The treatment for MOTN insomnia (as opposed to other forms
of insomnia) also presented an obstacle as the ability to get a person back to sleep in the middle of
the night rather than before bed, was assumed at one point to be at least slightly more difficult
because a person’s drive to sleep is lessened.
Indeed, overall propensity or drive to sleep depends on the interaction between two
biological processes: the homeostatic sleep drive and circadian drive. (Tr. 10.32:10–10.33:7
Czeisler)). Homeostatic sleep drive increases with every waking hour, but when a person falls
asleep, homeostatic sleep drive declines, such that by the middle of the night, much of homeostatic
sleep drive is dissipated. (Tr. 10:35:4–24 (Czeisler)). On the other hand, the circadian drive is
governed by a biological clock that sends out a signal to wake or be alert during the day and a
signal to quiet the drive to wake at night. (Tr. 10.37:16–10.39:19 (Czeisler)). The circadian drive
promotes waking during the day (as the circadian signal for alertness increases) and promotes sleep
(or the absence of waking) through the night (as the circadian signal for alertness decreases). (Tr.
10.37:16–10.39:19, 10.44:19–10.45:14 (Czeisler)).
But these processes must be reconciled with the average person who sleep for eight straight
hours. In other words, what keeps them asleep when their homeostatic sleep drive is decreasing
by the hour? Ultimately the conclusion inevitably drawn is that the circadian signal keeps someone
asleep. 4 In any event, to skilled artisans, the interaction of these processes results in a person’s
4
Indeed, this was the only plausible explanation provided to the Court for how a person sleeps
stays asleep. (Tr. 7.133:10-11).
9
overall urge—or drive to sleep—to be the greatest at bedtime (as opposed to the middle of the
night).
VI. Ambien®/Zolpidem
Physicians and psychiatrists were routinely prescribing the drug Ambien® to treat
insomnia, prior to the filing of the patents-in-suit. (Tr. 2.93:13-20 (Ocalssen)). The active
ingredient in Ambien® is the same as that in Intermezzo®, zolpidem tartrate. As of 2004,
Ambien® was commercially available and prescribed in two dosages, a 10 mg tablet for nonelderly patients and a 5 mg tablet for the elderly. (JTX 41 at DRL0013528). Ambien®’s dosing
information indicates that the elderly are dosed with half the normal dose as they may be especially
sensitive to the effects of zolpidem tartrate. This is consistent with the general understanding that
elderly people are more sensitive to the side effects of medications and metabolize drugs
differently (usually more slowly) causing medications to reside in the body at high levels for
longer. (Tr. 7.126:2-12 Winkelman)).
Ambien® was an incredibly successful drug commercially, which stimulated the medical
research community to extensively investigate and review zolpidem. Undeniably, by 2004 the
prevailing view amongst medical professionals was that zolpidem was well tolerated and posed
minimal risk of abuse and dependence at the therapeutic doses of 5 mg and 10 mg. Ambien®,
through its label, was indicated for prophylactic administration. While it unclear how common the
practice was, at least some doctors were prescribing fractional (half of one) Ambien® to treat
MOTN insomnia prior to 2004. (Tr. at 2.78:7-21 (Oclassen), 7.145:19-7.147:10, 7.149:237.150:23 (Winkelman)). This is likely because Ambien® was known by 2004 to be a safe/effective
10
sleep drug but more importantly, have a short duration of action of about 4 hours. (Tr. 2.75:14-16
(Oclassen), 6.161:11-21 (Michniak-Kohn), 7.143:16-7.144:10 (Winkelman)).
VII. Zolpidem v. Zaleplon
As a general proposition medication will be gone from the body in three half-lives. (Tr.
1.161:21-1.62:9 (Kryger)).
While zolpidem was known to have a rapid onset and short half-life
of about two to three hours, another hypnotic, zaleplon, was known to have an even shorter halflife of just one hour. Zaleplon is the active ingredient in the sleeping pill Sonata®, which around
2001, was a “comparatively new medication.” (JTX 033 at 116). Pertinent articles and studies
prior to 2004 were predomininantly limited to analyzing zolpidem and zaleplon, therefore
suggesting that these two hypnotics were the only current effective options to treat insomnia.
While on the surface it may seem as though zaleplon’s ultra short half life could be ideal for
treatment in the middle of the night to avoid morning residual sedative effects, zaleplon proved
unsatisfactory as the risk of waking up again was identified. (Tr. 1.161:21-1.162:9 (Kryger)).
Particularly when analyzed according to the indication of Intermezzo®, which targeted patients
who woke up in the middle of the night with at least four hours of sleep remaining, Sonata®,
zaleplon, was too short acting, where a patient could not count on getting four hours of sleep. (Tr.
7.156:22-24 (Winkelman); see also Teitelbaum (PTX 033: “[M]ost FMS patients I have treated
with Sonata have not found it to be helpful. I think Ambien is better.” (emphasis in original)).
VIII. Dosage Strength of Hypnotics
Given the prophylactic nature of administration of hypnotics such as zolpidem and zaleplon
prior to 2004, the dosage strengths of such were usually 5mg for an elderly patient and 10mg for
non-elderly. The primary goal of hypnotics, as with many drugs, was to find the lowest effective
11
dose to prevent overmedicating without compromising the dose’s efficacy. Similarly, for safety
reasons, any dose of a hypnotic should leave a patient without a “hangover” in the morning,
meaning they are free from residual sedative effects. For example, in the morning many patients
would wake up and then drive to work. Driving while experiencing residual sedative effects is
both hazardous to the patient and the general public. This is why one method for determining if
residual sedative effects are present after administration of a hypnotic, is in fact a driving test. (See
Vermeeren study). One well-known solution for sleep experts and drug formulators to combat
side effects including residual sedative effects, was to lower the amount of the hypnotic in a given
dose. (See e.g. (Tr. 7.156:7-15 (Winkelman): Teitelbaum is suggesting that Ambien is the go to
drug, but if a patient was too hungover “you could…reduce the dose.”
(Tr. 7.156:7-15
(Winkelman)). Thus, the overarching objective in treating MOTN insomnia was to strike a balance
between lowering doses to avoid residual sedative effects, while maintaining a dosage strength
that was effective for four hours.
IX. Transmucosal Delivery
Most medications are formulated in an oral, swallow form.
However, routes of
administration of a drug are changed when taking into account the indication of the treatment (for
example, a drug indicated for the treatment of MOTN insomnia) and the amount of active
ingredient. When taking an oral swallow pill, at least some portion of the drug will not get
absorbed and in turn, goes out through the system. More specifically, when a drug endures a “first
pass effect” it goes to the liver where some of it gets broken down before making its way to
systemic circulation. (Tr. 7.84:2-25 (Winkelman)). Importantly, systemic circulation provides the
12
delivery of the drug to the brain. (Id.). Thus, the primary benefits of systemic administration in
general, are that a drug will work more quickly and therefore potentially wear off more quickly.
While there are a number of methods of systemic administration, such as intravenous, skin
permeation and inhalation, many are not practical for specific treatments. One more sensible and
universal method involves delivery through the body’s mucosal surfaces—in the mouth or nose—
where there are many blood vessels close to the surface for a drug to enter directly. Within the
mouth, drugs can therefore be delivered across the sublingual or buccal mucosa by “transmucosal
delivery.” While this method was well known prior to 2004, it was noted that not all drugs can or
should be delivered this way. This is for the same reason that not all drugs—for example those
that must be delivered in voluminous quantities to be efficacious—are suitable for delivery by an
oral swallow tablet, either because it is an impossible route or is less beneficial.
Fortunately, for many drugs whose properties are well known (e.g. pharmacokinetics) and
have been extensively researched and published by the medical community, formulations for
transmucosal delivery became more identifiable. Pharmacokinetic behavior of a drug is in some
cases an accurate parameter for determining if a drug is suitable for transmucosal delivery. By
way of example, in broad terms, drugs that are considered “lipophilic” will pass more easily
through membranes, a determination that can be made when the logP value is known or published.
(See e.g. Tr. 6.156: 9-11: Zolpidem’s published logP value is 2.42.). Further, a well-known theory
called the Henderson-Hasselbalch principle explains that a drug can be made more lipophilic by
changing the pH, using for instance, a buffer. (Tr. 5.63:7-17 (Singh)). While zolpidem contains
apparent properties that are suitable for transmucosal delivery, prior to 2004 there was not a
developed formulation for zolpidem in sublingual doses.
13
X. Trial Witnesses
The following witnesses either appeared or had their recorded deposition admitted as
evidence and played at the bench trial. 5
A. Plaintiffs’ Witnesses
1. Meir Kryger, M.D.
The Court accepted Dr. Kryger as an expert in sleep medicine and the clinical research and
treatment of sleep disorders, including insomnia. (Tr. 1.136:4–16). Dr. Kryger is a Professor in
the Department of Internal Medicine at Yale University in New Haven, Connecticut. (Tr. 1.124:7–
10; PTX 11 at 1). He is also the Director of the Clinical Sleep Fellowship Program at the Yale
Program of Sleep Medicine, and a practicing physician in the VA Connecticut Health System,
specializing in sleep medicine. (Tr. 1.124:7–10, 1.128:12–21; PTX 11 at 1). Dr. Kryger has
experience with Ambien® in clinical trials, and opined on its properties and how it was prescribed
by physicians, dating from before its U.S. approval in 1992. (Tr. 10.189:12–10.190:9). Dr. Kryger
was presented by Plaintiffs and testified regarding the understanding in the art concerning
insomnia and MOTN insomnia and appropriate treatments of the condition at the time of the
inventions claimed in the patents-in-suit. Dr. Kryger also offered opinions on invalidity (objective
indicia of non-obviousness) and infringement.
5
The Court has omitted some witnesses from this section having found their testimony either
redundant or irrelevant based upon the Court’s findings. The Court did however, consider all of
the testimony at trial to make such findings.
14
2. Charles Czeisler, Ph.D., M.D.
The Court accepted Dr. Czeisler as an expert in sleep and sleep disorders. (Tr. 10.24:8–15
(Czeisler)). Dr. Czeisler is the Baldino Professor of Sleep Medicine and the Director of the
Division of Sleep Medicine at the Harvard Medical School, and Chief of the Division of Sleep
Medicine in the Department of Medicine at the Brigham and Women’s Hospital in Boston,
Massachusetts. (PTX 13 at 1, 3). Dr. Czeisler opined on the physiology of the human circadian
timing system and its relationship to the sleep-wake cycle.
3. David Drover, M.D., M.Sc
The Court also accepted Dr. Drover as an expert in clinical pharmacology. (Drover Tr.
2.133:21–2.134:5). Dr. Drover is a Professor of Anesthesia at Stanford University, where he has
been teaching since 1995. (PTX 14 at 1, 7; Drover Tr. 2.128:1–2). He conducts research within
the field of clinical pharmacology and has been involved in more than 50 clinical studies, some
involving the hypnotics zaleplon and zolpidem. (Drover Tr. 2.130:18–19, 2.131:11––2.133:1). Dr.
Drover testified about the pharmacokinetic properties of different dosage forms, zolpidem
formulations known at the time of the inventions and those formulations claimed in the patentsin-suit.
4. James Polli, Ph.D.
The Court accepted Dr. Polli as an expert in pharmaceutics, pharmaceutical formulation,
and drug delivery. (Tr. 3.39:14–20). Dr. Polli is a Professor of Pharmaceutical Sciences and the
Ralph F. Shangraw Endowed Chair in Industrial Pharmacy and Pharmaceutics at the University of
Maryland School of Pharmacy. (PTX 15 at 1; Tr. 3.30:18–22, 3.32:8–12). Dr. Polli gave expert
testimony about drug formulation, delivery, and absorption, including of course, zolpidem.
15
5. Glenn Oclassen
Mr. Oclassen was a co-founder of Transcept. (Tr. 2.57:13–21.) During the development of
Intermezzo, and until 2014, Mr. Oclassen was President and CEO of Transcept. (Tr. 2.57:7–12).
6. Thomas Roth, Ph.D.
Dr. Roth was a consultant to Plaintiff, Transcept during the time that it developed
Intermezzo® and submitted a declaration during the prosecution of the ’131 Patent.
(Tr.
10.237:18–20). From 1978 to 2014, Dr. Roth was the Director of the Sleep Disorders and Research
Center at Henry Ford Health Systems. (PTX 536 at 1; Tr. 10.238:22–25). Dr. Roth was deposed
by Defendants in this matter on April 15, 2014.
7. James Garegnani
Mr. Garegnani’s deposition testimony was admitted at trial. At the time of his deposition,
Mr. Garegnani was Director of Product Development for Novel. Mr. Garegnani was Novel’s Rule
30(b)(6) designee on topics related to the development of Novel’s ANDA Product, the formulation
of Novel’s ANDA Product, and any validity analysis of prior art conducted by Novel. (Tr. 2.216:9–
12).
8. Alfred Liang, Ph.D.
Mr. Liang’s deposition testimony was admitted at trial. At the time of his deposition, Dr.
Liang was a Director of Product Development for Actavis. (Tr. 4.77:10–11). Dr. Liang was
Actavis’ 30(b)(6) designee on topics related to Actavis’ ANDA, the research and development
16
leading to Actavis’ ANDA Product, and testing that Actavis did on its products and Intermezzo.
(Tr. 4.77:12–15).
9. Kranthi Kumar Gorlamari
Mr. Gorlamari is a former employee of Novel, who Plaintiffs deposed in his personal
capacity. (Tr. 3.6:18–21). He was a formulation scientist who worked on the development of
Novel’s ANDA Product. (Tr. 3.7:14–3.8:6).
10. Narayanan Badri Viswanathan, Ph.D.
Dr. Viswanathan’s deposition testimony was admitted at trial.
At the time of his
deposition, Dr. Viswanathan was the senior director of formulations for DRL. (Tr. 4.55:1–2
(Viswanathan)). Dr. Viswanathan was DRL’s Rule 30(b)(6) designee on topics related to the
content of DRL’s ANDA, the research and development leading to DRL’s ANDA Product, the
product’s formulation, and DRL’s knowledge of Intermezzo and the patents-in-suit. (Tr. 4.55:2–
6).
B. Defendants’ Witnesses
1. Bozena Michniak-Kohn, Ph.D.
Dr. Michniak-Kohn has a Ph.D. in pharmacology has practiced as a pharmacist. As a
pharmacist and as a teacher, she consulted prescribing and labeling information for drugs and has
formulated drugs for transmucosal delivery.
Dr. Michniak-Kohn’s current position is full
professor with tenure in pharmaceutics at the Ernest Mario School of Pharmacy, at Rutgers, the
17
State University of New Jersey. (Tr. at 6.19:23-6.20:1). The Court qualified Dr. Michniak-Kohn
as an expert in pharmaceutical sciences, pharmacy, pharmacology and formulation science. (Tr.
at 6.30:11-16).
2. John Winkelman, Ph.D., M.D.
Dr. Winkelman is a currently a practicing physician and Chief of the Sleep Disorders
Clinical Research Program at Massachusetts General Hospital. Dr. Winkelman received his Ph.D.
in Psychobiology from Harvard University in Cambridge, Massachusetts in 1983 and a medical
degree from Harvard Medical School in 1987. Dr. Winkelman was proffered and accepted as an
expert in sleep science, sleep medicine, and the treatment of sleep disorders. (Tr. at 7.81:17-24).
3. Umesh Banakar, Ph.D.
Dr. Banakar is an independent consultant and advisor to pharmaceutical companies and
governmental agencies for the development and evaluation of pharmaceutical formulations. (DTX
3021 at 1-2). He received his Ph.D. in Pharmaceutical Technology from Duquesne University in
Pittsburgh, Pennsylvania and completed his post-doctorate research relating to Advances in
Controlled Release Technology at the Massachusetts Institute of Technology in Boston,
Massachusetts in 1989. Dr. Banakar has assisted in formulating about 10 to 15 sublingual tablets,
including a zolpidem sublingual tablet commercially available outside of the United States. As a
result, Dr. Banakar was proffered and accepted by the Court as an expert in the field of
pharmaceutical formulations. (Tr. 4.108:15-23).
18
4. Jason McConville, Ph.D.
Dr. McConville is an associate professor of pharmaceutics at the College of Pharmacy at
the University of New Mexico in Albuquerque, New Mexico and has over 20 years of experience
in the field of drug formulation and delivery and in the field of transmucosal drug delivery. He
received of Bachelor of Science with Honors in Applied Chemistry from Coventry University in
Coventry, United Kingdom in 1994. (Tr. at (DTX 1000)). Dr. McConville’s current area of
research is transmucosal drug delivery, which incorporates all transmucosal delivery such as oral
and lung transmucosal delivery. (Tr. 4.160:18-21 ). Dr. McConville was proffered and accepted as
an expert in the field of drug formulation and delivery, and an expert in the field of transmucosal
drug delivery. (Tr. 4.161:3-9).
5. Ann Kraft
Ann Kraft, appearing in her capacity as a Fed. R. Civ P. 30(b)(6) corporate designee and
in her individual capacity, was deposed by Defendants in this matter on March 11, 2014 deposition.
At the time of her deposition, portions of which were presented at trial via designation, Ms. Kraft
was the Executive Director of Licensing and Business Development for Plaintiff, Purdue Pharma
LP.
6. Margaret Moline, Ph.D.
Dr. Moline, appearing in her capacity as a Fed. R. Civ P. 30(b)(6) corporate designee and
in her individual capacity, was deposed by Defendants in this matter on March 13th, 2014
deposition. At the time of her deposition, portions of which were presented at trial via designation,
Dr. Moline was a Director in the Medical Research department at Purdue Pharma LP.
19
7. Nilesh Parikh, Ph.D.
Dr. Parikh submitted on behalf of Plaintiff, Transcept, a declaration to the U.S. Patent and
Trademark Office during the prosecution of the patent application that would eventually issue as
the ’628 patent. (JTX 8; Tr. at 8.231:3-5 (Parikh)). Dr. Parikh was deposed by Defendants in this
matter on March 31, 2014.
8. Nikilesh Singh, Ph.D.
Dr. Singh is the named inventor on each of the patents-in-suit. He is also the founder of
Plaintiff Transcept. At the time of his deposition, Dr. Singh was the Senior Vice President and
Chief Scientific Officer for Plaintiff Transcept. (Tr. 5.61:17-18). Dr. Singh, appearing in her
capacity as a Fed. R. Civ P. 30(b)(6) corporate designee and in his individual capacity, was deposed
by Defendants in this matter on March 25, 2014 deposition.
LEGAL ANALYSIS
Plaintiffs assert claims of the ’131 and ’628 patent against all Defendants. Plaintiffs assert
the claims of the ’809 patent against Novel and DRL. Defendants argue that each of the patentsin-suit is invalid as obvious. Defendants also claim the ’628 patent is invalid as anticipated and
the ’131 patent element “without residual sedative effects” is invalid as indefinite. While it is
Plaintiffs’ burden to prove infringement by a preponderance of the evidence, because all patents
“shall be presumed valid,” 35 U.S.C. § 282, the “burden is on the party asserting invalidity [here,
Defendants,] to prove it with facts supported by clear and convincing evidence.” Linear Tech
Corp. v. Int’l Trade Comm’n, 566 F.3d 1049, 1066 (Fed. Cir. 2009) (internal citations omitted).
20
I. Infringement
A patent is infringed when a person “without authority makes, uses, offers to sell, or sells
any patented invention, within the United States, or imports into the United States any patented
invention during the term of the patent ....” 35 U.S.C. § 271(a). Determining infringement requires
a two step inquiry. Step one requires a court to construe the disputed terms of the patent at issue
and step two requires a court to compare the accused products with the properly construed claims
of the patent. Step one is a question of law; step two is a question of fact. Markman v. Westview
Instruments, Inc., 52 F.3d 967, 979–81 (Fed.Cir.1995). To prove literal infringement, the patentee
must show that the accused device contains every limitation in the asserted claims. Dolly, Inc. v.
Spalding & Evenflo Cos., 16 F.3d 394, 397 (Fed.Cir.1994). If even one limitation is missing or
not met as claimed, there is no literal infringement. Mas-Hamilton Grp. v. LaGard, Inc., 156 F.3d
1206, 1211 (Fed. Cir. 1998)(internal citations omitted).
In Hatch-Waxman litigation, infringement cases are filed before the alleged infringing
product is sold. Consequently, the infringement analysis is based on an assumed or hypothetical
set of facts. Under 35 U.S.C. 271(e)(2):
It shall be an act of infringement to submit . . . an [ANDA] for a
drug claimed in a patent or the use of which is claimed in a patent,
. . . if the purpose of such submission is to obtain approval . . . to
engage in the commercial manufacture, use, or sale of a drug or
veterinary biological product claimed in a patent or the use of
which is claimed in a patent before the expiration of such patent.
Thus, the filing of an ANDA seeking approval for an indication claimed in a patent constitutes
infringement under Section 271(e)(2). As summarized in Harman, Patents and the Federal Circuit
494 n. 161 (9th ed. 2009): “The inquiry under § 271(e)(2) is a standard infringement test. The only
difference is that the allegedly infringing drug has not yet been marketed and therefore the question
of infringement must focus on what the ANDA applicant will likely market if its application is
21
approved.” See also Warner–Lambert Co. v. Apotex Corp., 316 F.3d 1348, 1366 (Fed.Cir.2003)
(“The proper inquiry under § 271(e)(2)(A) is whether, if a particular drug were put on the market,
it would infringe the relevant patent.”). Further, “[t]his hypothetical inquiry is properly grounded
in the ANDA application and the extensive materials typically submitted in its support.” Id. at
1248 (quotation omitted). Therefore, it is proper for this Court to consider the ANDA itself,
materials submitted by the ANDA applicant in support of its ANDA, and any other pertinent
evidence. Id. at 1248–49.
Similarly, the sale of a product specifically labeled for use in a patented method constitutes
inducement to infringe that patent. See Astrazeneca LP v. Apotex, Inc., 633 F.3d 1042, 1060
(Fed.Cir.2010) (finding intent to induce infringement based on the product label authorizing the
patented use, which “would inevitably lead some consumers to practice the claimed method.”
(emphasis added)).
Indeed, “the substantive determination whether actual infringement or
inducement will take place is determined by traditional patent infringement analysis, just the same
as it is in other infringement suits, including those in a non-ANDA context, the only difference
being that the inquiries now are hypothetical because the allegedly infringing product has not yet
been marketed.” Warner-Lambert Co. v. Apotex Corp., 316 F.3d 1348, 1365-66 (Fed. Cir.
2003). This is because “pharmaceutical companies do not generally treat diseases; rather, they sell
drugs to wholesalers or pharmacists, who in turn sell the drugs to patients possessing prescriptions
from physicians. Pharmaceutical companies also occasionally give samples of drugs to doctors
and hospitals. In none of these cases, however, does the company itself treat the disease.” Id. at
1363. With this framework in mind, the Court analyzes infringement of the elements of each
asserted claim and patent.
22
A. ’628 Patent
There are two outstanding issues regarding infringement of the ’628 patent.
First,
Defendants, DRL and Actavis (only), argue that their ANDA products do not infringe Claims 16
and 17 of the ’628 patent because their products do not contain a “buffer.” Second, all Defendants
claim that Plaintiffs have failed to meet their burden of proving Defendants’ respective ANDA
products infringe the in vivo limitations recited in Claim 1 of the ’628 patent. The Court finds
Plaintiffs have failed to meet their burden with reference to the former, but have done so
successfully regarding the latter.
1. “Buffer”
Claim 1 of the ’628 patent recites “a solid pharmaceutical composition comprising
zolpidem or a pharmaceutically acceptable salt thereof . . . , the pharmaceutical composition further
comprising a buffer . . . .” (JTX 003 at Claim 1). Claims 2-17 of the ’628 patent, which each
depend directly or indirectly upon Claim 1, incorporate by reference all limitations of Claim 1.
(Id. at Claims 12-17). The term “buffer,” as used in Claim 1 of the ’628 Patent, means “a buffer
system of two or more buffering agents.” (Claim Construction Order, ECF No. 186, at 2). The
term "buffering agent," is construed to mean "a proton-donating component or proton-accepting
component used to maintain and/or achieve an approximate pH range.” (Id.). It is undisputed
that Novel’s ANDA product contains a buffer system of two or more buffering agents and
therefore, this analysis shall apply to Defendants, Actavis and DRL, alone.
The first buffering agent in Actavis’ ANDA product is
a pH-adjusting agent,
as admitted by Actavis’ formulation expert Dr. Banakar. (Tr. 4.136:1-3 (Banakar)). Plaintiffs
assert that Actavis infringes Claims 16 and 17 of the ’628 patent, because Actavis’ ANDA products
23
contain
in addition to tartaric acid, as a second buffering agent. (See Tr. at 3.107:17-
3.108:25, 4.12:10-24, 4.15:16-23 (Polli)). Similarly, DRL’s first buffering agent is
, which is listed on its ANDA product as a pH-modifier. While DRL appeared to dispute at
trial whether
is in fact a buffering agent, DRL had previously characterized
as a buffering agent in its non-infringement contentions. (Tr. 5.34:15–5.36:21
(McConville)). This admission taken in conjunction with DRL’s senior director of formulations,
Dr. Viswanathan’s testimony, stating that
is used in DRL’s ANDA Product to
“increase [pH] beyond 8.2 . . . and ensure that it doesn’t come below that 8.2,” confirms that
is used to achieve and maintain an alkaline pH range.
(Tr. 4.65:1–7
(Viswanathan); see also Tr. 4.60:17–20). As construed, a “buffering agent” performs this function
and therefore the Court concludes
is in fact a buffering agent. Plaintiffs assert
that DRL infringes Claims 16 and 17 of the ’628 patent, because DRL’s ANDA products contain
and tartrate as buffering agents used to achieve and maintain pH to facilitate
the transmucosal delivery of zolpidem. (PFOF ¶579). However, based upon the evidence and for
the reasons that follow, the Court finds that Plaintiffs have failed to prove by a preponderance of
the evidence that Actavis’ and DRL’s ANDA products contain a second buffering agent, neither
as tartaric acid nor tartrate.
a. Tartrate cannot exist as a single chemical entity and zolpidem tartrate is not a buffering
agent.
At trial, Defendants’ expert, Dr. McConville, testified in great detail about the structure of
zolpidem tartrate and why tartrate could not be removed from its bond with zolpidem to then
become a single buffering agent. The Court finds this reasoning persuasive. That is, zolpidem
tartrate is a salt containing a stoichiometric amount of two molecules of zolpidem cation per one
24
tartrate anion. (Tr. 4.189:4-4.190:13 (McConville)). Dr. McConville explained that a salt, such
as zolpidem tartrate, is a neutral compound consisting of a positively charged cation and a
negatively charged anion where the cation and anion are ionically bound (or attracted to one
another), forming one of the strongest bonds in chemistry. (Tr. 4.170:10-23 (McConville)). It was
also explained by Dr. Banakar, that tartrate cannot exist as a single chemical entity because it is a
charged negative compound, and thus no portion of zolpidem can separate from the tartrate in the
mouth. (Tr. 4.124:21-24 (Banakar)).
Dr. Banakar, in his testimony, explained further to the Court how a buffer works and why
zolpidem tartrate could not provide assistance to this function and therefore, is not a “buffering
agent.” That is, in order for a buffer to work, there has to be a conjugate acid for the base. (Tr.
4.125:12-23. (Banakar)). A “proton-donating component” is an acid and a “proton-accepting
component” is a base. (Tr. 4.168:14-19 (McConville)). Dr. Banakar then explained zolpidem
tartrate is a neutral compound (neither acidic or basic) and thus neutral compounds are neither
proton-donating nor proton-accepting as required by the construction of “buffering agent.” (Tr.
4.120:4-7 (Banakar)). In sum, because the Court is persuaded by the evidence that zolpidem
tartrate is neutral, and thus unable to form a buffer with
for example, a base, the tartrate
cannot be characterized as a buffering agent.
b. Zolpidem Tartrate in Actavis’ and DRL’s ANDA products is a single compound which
does not contain tartaric acid.
At trial, Plaintiffs presented evidence through their expert, Dr. Polli, that because the
tartrate component of zolpidem tartrate comes from tartaric acid, it donates a proton to the
compound and therefore is proton-donating under the Court’s construction of “buffering agent.”
(Tr. 3.73:133-3.74:12 (Polli)). Dr. Polli supported this proposition by pointing to the USAN, the
25
committee that gives drugs name in the United States, who characterizes tartaric acid as a buffering
agent in pharmaceutics. (Tr. 3.72:21–3.73:12 (Polli); PTX 450 at PLSEXP404). However,
Defendants do not challenge whether or not tartaric acid may in some cases act as a buffering
agent, but rather whether tartaric acid is present at all in Defendants’ ANDA products. Indeed, the
Court finds there was sufficient evidence at trial to refute Plaintiffs’ assertion that tartaric acid
exists in DRL or Actavis’ ANDA products.
Dr. McConville’s testimony exemplified that there is no tartaric acid in either DRL’s or
Actavis’ ANDA products through his explanation of zolpidem tartrate and polymorphism. (Tr.
4.186:6-7 (McConville); DTX 1001). Polymorphism exists when a compound or material exists
in distinct crystalline forms or types, known as “polymorphs,” each of which can be can be
distinguished from other polymorphs using known methods such as powder X-Ray Diffraction.
(Tr. 4.185:25-4.186:5; 5.14:18-5.17:6 (McConville); DTX 1001). According to Dr. McConville,
analytical studies using X-ray diffraction and thermal analysis of
of zolpidem
tartrate—the polymorph used in Actavis’ and DRL’s ANDA products—demonstrate that zolpidem
tartrate does not contain tartaric acid. (See Tr. 4.185:19-4, 186:14, 5.17:1-5.18:3, 5.21:6-16
(McConville); PTX 55). Plaintiffs have not credibly refuted this analysis and therefore the Court
is unconvinced by a preponderance of the credible evidence, that tartaric acid is present.
Consequently, tartrate, in the Court’s view, is not characterized as the second buffering agent in
either DRL or Actavis’ ANDA products.
c. The language of Claim 1 indicates that the active pharmaceutical ingredient and “buffer”
are intended be two distinct and separate components of the claimed solid pharmaceutical
composition.
Claim 1 of the ’628 patent provides a “method for treating insomnia, comprising the steps
of: administering a solid pharmaceutical composition comprising zolpidem or a pharmaceutically
26
acceptable salt thereof to a subject prone to insomnia, the pharmaceutical composition further
comprising a buffer.” After a review of the evidence at trial, the Court finds that the claim
delineates that in addition to zolpidem tartrate, there must be two buffering agents in an ANDA
product for it to infringe the ’628 patent, therefore eliminating the possibility that tartrate can be
categorized as a buffering agent in the composition.
The phrase “further comprising” signals that these claimed elements (“zolpidem or a
pharmaceutically acceptable salt thereof,” on the one hand, and “a buffer” on the other) are distinct
components of the solid pharmaceutical composition. See HTC Corp. v. IP Com GmbH & Co.,
667 F.3d 1270, 1275 (Fed. Cir. 2012) (“[F]urther comprising” signals something “additional.”);
Remediation Prods., Inc. v. Adventus Ams., Inc., 3:07-cv-153-RJC, 2009 WL 57456 (W.D.N.C.
Jan. 7, 2009) (“The construction of the phrase ‘further comprising’ includes additional recited
elements.”). Dr. Banakar opines that this reading—where one ingredient cannot serve more than
one purpose in the same composition—aligns with the FDA and regulatory agencies around the
world requiring “that a single ingredient serve one function in any given composition.” (Tr.
4.128:20-4.130:18 (Banakar)). Indeed, Intermezzo® which encompasses the ’628 patent, further
verifies individualized functions as Intermezzo® explicitly contains zolpidem tartrate in addition
to two buffering agents (sodium carbonate and sodium bicarbonate). (See e.g. Tr.4.123:4-7
(Banakar)). Even if the Court had construed tartrate and/or tartaric acid to be a buffering agent,
Plaintiffs have not established by a preponderance of the credible evidence that tartrate, a part of
the pharmaceutically acceptable salt, was intended to serve dual functions in the claim. The Court
therefore concludes that each listed function of the ingredients in the ANDA products are the sole
functions. Thus, the single function of zolpidem tartrate in Actavis’ and DRL’s ANDA products
is as listed, the active pharmaceutical ingredient. The single function of
27
in Actavis’
ANDA products is therefore a pH-adjusting agent. And finally, the single function of
in DRL’s ANDA products is as listed, a pH modifier. DRL and Actavis’ ANDA products
each contain just one buffering agent.
For the reasons set forth above, the Court determines that substantial evidence supports
that DRL and Actavis’ ANDA products each contain only one buffering agent and that Plaintiffs
have not come forth with sufficient credible evidence to establish infringement of the ’628 patent
by a preponderance of the evidence. Having concluded that Defendants, DRL and Actavis ANDA
products, if sold, would not infringe Claim 1 of the ’628 patent, the Court further concludes that
these ANDA products would also not infringe Claims 16 and 17, which depend from Claim 1.
2. The in vivo claim limitations of the ’628 patent. 6
Defendants assert that Plaintiffs have failed to prove by a preponderance of the evidence
that two limitations of the ’628 patent are infringed by Defendants’ ANDA products. Defendants
therefore maintain that Claims 9, 16 and 17 of the ’628 patent are not infringed. Defendants take
issue with the limitation that the buffer “raises the pH of saliva to a pH of about 7.8 of greater”
and requirement that the solid pharmaceutical composition “dissolves within about 10 minutes or
less within the oral cavity following administration.” (JTX-002 at Claim 1). Defendants’
argument, in sum, provides that Plaintiffs cannot rely on in vitro data to establish infringement of
the in vivo claim limitations articulated above.
For this proposition, Defendants rely almost
exclusively on Alza Corp. v. Mylan Laboratories, Inc. 388 F.Supp. 2d 717, 725 (N.D. W. Va.
2005) aff’d 464 F.3d 1286 (Fed Cir. 2006). As set forth below, the Court does not agree with
6
The Court hereby incorporates this Section into its later analysis of the in vivo limitations of the
’809 patent.
28
Defendants’ interpretation of Alza Corp., as it relates to the facts of the case at bar and finds
sufficient evidence of infringement of the in vivo limitations of the ’628 patent. Defendants 7,
therefore, are found to have infringed Claims 9, 16 and 17 of the ’628 patent in this regard.
a. Alza Corp. v. Mylan Laboratories, Inc.
The Alza Corp., litigation arose from Defendants’ filings of ANDAs for once-daily,
controlled-release oxybutynin formulations. Alza Corp. v. Mylan Labs., Inc., 464 F.3d 1286, 1288
(Fed. Cir. 2006). Oxybutynin is a drug used to treat urinary incontinence. Once-a-day dosing
provides the usual benefits of convenience, steady-dosing, and in addition, possibly reduced
absorption of a metabolite that leads to side-effects. Id. Claim 2 of the '355 patent was at issue
and stated:
A sustained-release oxybutynin formulation for oral administration
to a patient in need of treatment for urge incontinence comprising a
therapeutic dose of an oxybutynin selected from the group
consisting of oxybutynin and its pharmaceutically acceptable salt
that delivers from 0 to 1 mg in 0 to 4 hours, from 1 mg to 2.5 mg in
0 to 8 hours, from 2.75 to 4.25 mg in 0 to 14 hours, and 3.75 mg to
5 mg in 0 to 24 hours for treating urge incontinence in the patient.
Id. at 1289 (emphasis in original). The district court construed the '355 patent claims in its
Markman Order, construing the word “deliver” to refer to the rate of in vivo release in the
gastrointestinal (“GI”) tract. Id.
At trial, Plaintiff, Alza Corp., did not present direct evidence that one Defendants' ANDA
formulation released drug in the GI tract at the rates claimed by the '355 patent. Id. However, it
7
Defendants, DRL and Actavis, do not infringe the ’628 patent as they do not contain “buffer,”
but the Court completes its infringement analysis regarding the ’628 patent for purposes of
Defendant, Novel (as well as TWi and Par if applicable).
29
did offer two other types of evidence: 1) the rate at which the generic product released oxybutynin
in an in vitro dissolution apparatus; and 2) the rate at which the ANDA product resulted in the
accumulation of oxybutynin in the bloodstream. Ultimately, the district court found that Alza had
failed to meet its burden of proof on infringement, stating:
Alza cannot rely exclusively on in vitro test results to prove
infringement of in vivo release rates. See Amgen, Inc. v. Chugai
Pharm. Co., Ltd., 927 F.2d 1200 (Fed.Cir.1991) (holding that “the
district court erred in accepting the in vitro data as support for claim
containing what has been found to be an in vivo limitation”). Indeed,
without reliable in vivo data comparing the release rates of the
accused product against the claimed ranges of the '355 patent, there
can be no finding of infringement—either literally or under the
doctrine of equivalents.
Alza Corp. v. Mylan Labs., Inc., 388 F. Supp. 2d 717, 725 (N.D.W. Va. 2005). On September 6,
2006, the Federal Circuit affirmed the ruling of the district court, but made clear why, in the Alza
case, the in vitro data was insufficient. The Federal Circuit explained: “The critical deficiency in
the evidence presented by Alza was not that it was ‘indirect’ rather than ‘direct,’ but rather that it
failed to credibly link these pieces of evidence with the relevant pharmacokinetic parameter—the
rate of in vivo dissolution in the GI tract.” Alza Corp. v. Mylan Labs., Inc., 464 F.3d 1286, 1296
(Fed. Cir. 2006) (emphasis added). That is, because “the obtained in vitro dissolution rates vary
widely with the choice of experimental parameters,” the Federal Circuit found that “Alza's
evidence of in vitro dissolution rates [was] irrelevant absent evidence demonstrating that the in
vitro system is a good model of actual in vivo behavior.” Id. at 1297.
Defendants in the present case, read the Alza Corp., litigation to stand for the proposition
that in vivo claim limitations can only be infringed upon conduction of in vivo testing. The Court
disagrees. In vitro testing is suitable to prove in vivo claim limitations if there is a credible link
between the in vitro data and in vivo data. In other words, without affront to the Federal Circuit
30
holding in Alza Corp., infringement can be found based on in vitro data where the evidence
demonstrates that the in vitro system adequately modeled the results that would be derived from
in vivo conditions. Allergan, Inc. v. Watson Labs., Inc.-Florida, 869 F. Supp. 2d 456, 500 (D.
Del.) aff'd, 470 F. App'x 903 (Fed. Cir. 2012). Because the Court finds (and explains below) that
at trial, Plaintiffs presented evidence of in vivo limitations via Defendants' ANDA data and in vitro
dissolution data which was then considered in combination with the inventor’s prosecution
declaration, the known information of dissolution rates and in vivo data from Intermezzo®,
Plaintiffs have met their burden by linking the in vitro and in vivo data. Accordingly, Defendants
are found to infringe the in vivo limitations of the ’628 patent. 8
b. Plaintiffs’ evidence, in vitro and otherwise, of infringement.
i. “[R]aises the pH of saliva to a pH of about 7.8 of greater.”
Defendants argue, in the main, that Plaintiffs have failed to show their products meet the
“raises the pH of saliva to a pH of about 7.8 of greater” element of the asserted claims because the
dissolution testing results (i.e. the in vitro data) does not sufficiently model in vivo conditions of
the mouth.
Defendants cite a number of propositions in support of this contention. First,
Defendants assert that simulated saliva does not adequately mimic saliva and therefore pH must
be measured in natural human saliva. (Tr. 2.231:4-2:232:11 (Garegnani), 6.41:6-14 (MichniakKohn)).
Second, Defendants attempt to discredit dissolution test results by reasoning that
depending on which simulated saliva recipe one chooses, pH measurements will fluctuate,
sometimes dramatically. (DTX 291 at 1109; Tr. 6.61:5-11 (Michniak-Kohn)). Lastly, Defendants
point the Court to the fact that Plaintiffs did not conduct any in vivo testing on Defendants’ ANDA
8
Such an analysis also applies equally to the ’809 patent where appropriate.
31
products. After a thorough consideration of the totality of the evidence presented, the Court finds
Defendants’ position unpersuasive and disagrees with the contention that Plaintiffs, via Dr. Polli’s
testimony, have failed to show that it is more likely than not that Defendants' proposed products
“raise[] the pH of saliva to a pH of about 7.8 of greater.”
At trial, Plaintiffs presented evidence that each Defendants’ product would raise the pH of
saliva to a pH of about 7.8 of greater. First, Plaintiffs offered evidence that each Defendant has
represented to the FDA the specific effect each product has on the pH of saliva. DRL informs
FDA that it’s “
” (PTX 98R at DRL0001294; Tr. 3.66:4–23 (Polli)). Actavis’ ANDA
specifies that “
” (PTX
55 at ACT-ZOL-0000275; Tr. 3.103:10–22 (Polli)). Novel’s ANDA tells FDA that “[t]he amounts
of sodium bicarbonate and sodium carbonate [in Novel’s product] were also challenged and
optimized based on the drug product target pH,” which it indicated to be “a pH above 9.” (PTX
138 at NOVZ0007909; Tr. 3.58:7–25, 3:59:17–3.60:6 (Polli)).
Next, Plaintiffs offered
Defendants’ ANDAs that report in vitro pH testing data to substantiate their statements to the FDA
about the pH levels. The pH achieved in this in vitro testing (above 9.5 for all Defendants) far
surpasses the minimum bar of at least 7.8 for this required pH level limitation. Specifically,
Novel’s testing produced a pH of 9.56 in simulated saliva; DRL’s testing produced pH values as
high as
in simulated saliva; and Actavis’ testing produced a pH of
in deionized water. (Tr.
3.58:7–3.60:19; 3.91:6–3.94:16, 3.109:21–3.111:4 (Polli)). This data, when properly linked to the
in vivo limitation, fully supports a finding of infringement.
32
Plaintiffs also provided the Court with the patent inventor’s prosecution declaration. Said
declaration expressly instructs that in vitro measurements in simulated saliva (Novel, DRL) or
deionized water (Actavis) are appropriate to establish the in vivo pH claim element. Dr. Singh,
the inventor, used pH testing in deionized water and simulated saliva to successfully persuade the
USPTO that the prior art did not “raise[] the pH of saliva to a pH of about 7.8 or greater.” (JTX 6
at TRANSIZ00059306–14; id. at TRANSIZ00059338–41) (See also Pls.’ FOF ¶¶ 461–62, 569–
70, 600–01). Dr. Polli explained that this is indicative of how the patent office was “inspired by
the experiments … such the simulated saliva allows one to differentiate whether a formulation is
within that claim limitation or outside.” (Tr. 3.89:19-22 (Polli)). Indeed, this is further confirmed
by Mr. Gorlamari, Novel’s former formulation scientist that developed its ANDA product, who
testified that “we can’t get the actual saliva, so, usually, we typically get a simulated saliva,” and
that, based on his experience in the pharmaceutical industry, the simulated saliva Novel used
“should be equivalent to the saliva” in a subject’s mouth for purposes of pH testing. (Tr. 3:11:8–
3.12:4 (Gorlamari)). Mr. Garegnani, Novel’s corporate 30(b)(6) witness on the development and
formulation of Novel’s ANDA products, also acknowledged that simulated saliva, including the
volume Novel used to run its pH testing, was used to approximate actual saliva in the mouth. (Tr.
2.232:22–2.233:13 (Garegnani)). This is precisely the breed of credible “link” that the Alza Corp
litigation approved. Here, the in vitro testing is authenticated by numerous sources to be “a good
model of actual in vivo behavior.” Alza Corp. v. Mylan Labs., Inc., 464 F.3d 1286, 1297 (Fed.
Cir. 2006).
With respect to simulated saliva criticisms of Defendants’ expert, Dr. Michniak-Kohn, who
claimed that there are too many different formulations that can be used with differing results, even
she acknowledged that actual saliva varies person to person and can even vary for the same person
33
at different times of the day. (Tr. 6.56:3–24, 7.25:11–16 (Michniak-Kohn)). Therefore, differing
formulations are actually representative of differing conditions in the mouth from person to person.
Admittedly, saliva is 99% water, and simulated salivas “have compositions[] which are more or
less the same as that of natural saliva.” (Tr. 7.25:11–7.26:1, 7.24:4–10 (Michniak-Kohn), DTX
291). As a result, the Court grants this critique only limited credibility. However, of more practical
importance, it would be difficult to imagine an in vivo test to gain the data Defendants request.
Throughout trial, Defendants have persistently criticized Plaintiffs for failing to test Defendants’
respective ANDA products. However, the Court will not overlook the implications of such
unapproved testing. Indeed, it would be highly unethical to have testing performed on humans
(i.e. in vivo) for unapproved products particularly for the limited purpose of patent litigation as
Defendants’ appear to suggest. Allergan, Inc. v. Watson Labs., Inc.–Fla., 869 F. Supp. 2d 456,
500 (D. Del. 2012) (“[T]esting defendants’ unapproved products in live human subjects is neither
feasible nor ethical.”). 9 For both this reason and those articulated above, the in vitro data is
credibly and sufficiently linked to other credible evidence presented to prove infringement by a
preponderance of the evidence for the claim limitation “raises the pH of saliva to a pH of about
7.8 of greater.”
ii. “[W]herein the solid pharmaceutical composition dissolves within about 10 minutes or less
within the oral cavity following administration.”
The majority of Defendants’ arguments related to the claim limitation “wherein the solid
9
See also Zenith Labs. Inc. v. Bristol-Myers Squibb Co., No. 91-3423, 1992 WL 340761, at *18
(D.N.J. Aug. 6, 1992) (“[I]n vivo experimentation could not be justified under medical ethics
constraints merely to prove patent infringement.”), rev’d on other grounds, 19 F.3d 1418 (Fed.
Cir. 1994).
34
pharmaceutical composition dissolves within about 10 minutes or less within the oral cavity
following administration” revolves around the element “oral cavity.” Defendants essentially argue
that dissolution of the composition, for purposes of infringement, must be measured in the oral
cavity. Defendants note the differences between USP disintegration tests and conditions within
the oral cavity in support of this proposition. The Court finds however, as set forth below, that
based on the credible evidence presented at trial, Plaintiffs have met their burden of proving
infringement of this “oral cavity” in vivo element of the ’628 patent.
Plaintiffs provided the Court with each of Defendants’ ANDA’s USP test results, which all
reflect a showing of at least
dissolution within 10 minutes or less. (See PTX 121, PTX 67,
PTX 102). This is undisputed. To link these results to the oral cavity, Plaintiffs offered evidence
that each of Defendants’ ANDA products indicate that they were designed for
all of which are qualities a POSA would anticipate for dissolution
in the oral cavity. (See PTX 121, PTX 55, PTX 97). Additionally, Plaintiffs provided that both
the ’628 patent itself, as well as the standard practice in the pharmaceutical industry, specify that
in vitro USP dissolution data is appropriate to measure in vivo dissolution. On cross-examination,
Dr. Polli explained how those of skill in the art rely on USP in vitro testing as a surrogate for in
vivo testing, stating:
I don’t know of a way to measure dissolution in the mouth. I don’t
know anybody that does. That would be extremely unusual. And
then when I read the patent, I see what I would have expected, USP
dissolution testing, so it kind of confirms what I would have
expected anyway.
(Tr. 3.151:8–24 (Polli)). The specification of the ’628 Patent further corroborates Dr. Polli’s
opinion and defines in vitro USP tests as suitable for determining the extent to which a solid dosage
35
form (e.g., Defendants’ ANDA Products) dissolves in a patient’s mouth (i.e., in vivo). Under the
claimed inventions:
The terms “disintegration” and “dissolution” are used
interchangeably to refer to the reduction of a solid dosage form of
the present invention to a liquid form. More particularly, a complete
disintegration or dissolution of a solid dosage form refers to less
than about 25% by weight of the solid dosage form remaining in the
mouth following an appropriate time period, e.g., 5 minutes or less,
after administration. . . . Suitable methods known in the art for
determining the dissolution profile of a solid dosage form include,
e.g., USP dissolution tests such as USP <711> Apparatus 1 or USP
<711> Apparatus 2.
(JTX 3 at 6:38–51 (emphases added)).
While Defendants appear to take issue with the volume (
) and paddle speed (
) parameters used in their own tests, Dr. Polli testified on cross-examination that these
parameters are commonly used to make sure that the tests are not only related to in vivo
performance but are also reproducible, a point that this Court finds credible. (Tr. 3.149:7–3.150:13
(Polli)). Dr. Polli explained that these tests are “intended to mimic what goes on in the mouth.”
(Tr. 3.148:6–12 (Polli)).
Indeed, Defendants conducted in vitro USP tests using the same
parameters, including the approximate pH
and temperature
of the oral cavity. (PTX
102).
(Tr. 3.96:23–3.99:1
(Polli); See e.g. PTX 99). The preponderance of the evidence thus weighs in favor of infringement
of the in vivo claim limitation “wherein the solid pharmaceutical composition dissolves within
about 10 minutes or less within the oral cavity following administration.”
36
3. Remaining Uncontested Infringement Evidence
Beyond the elements discussed supra or within this Court’s analysis of the ’809 patent, and
for purposes of completeness, the Court addresses the undisputed elements of the claims asserted
against Novel, and finds that Plaintiffs have proven infringement by a preponderance of the
evidence. First, with reference to independent Claim 1 of the ’628 patent, administering Novel’s
ANDA Product according to their proposed labeling comprises a method of treating insomnia.
(Kryger Tr. 1.245:19–1.246:8; PTX 50 at NOVZ00007869). Further, Novel’s ANDA Product
according to this label includes administering a solid pharmaceutical composition comprising
zolpidem or a pharmaceutically acceptable salt thereof to a subject prone to insomnia under Claim
1 of the ’628 Patent. (Kryger Tr. 1.246:9–1.247:17; PTX 50 at NOVZ00007869). Zolpidem is
also absorbed across a permeable membrane of the subject’s oral mucosa under Claim 1 of the
’628 Patent after administration according to Novel’s proposed labeling. (Tr. 2.155:13–2.176:4
(Drover)). Thus, the remaining elements of Claim 1 are found to be infringing.
Novel’s ANDA Product contains zolpidem or a pharmaceutically acceptable salt thereof in
an amount from about 1 mg to about 5 mg, under Claim 16 of the ’628 Patent, and in an amount
from about 2 mg to about 5 mg, under Claim 17 of the ’628 Patent. (Kryger Tr. 1.247:15–1.249:16;
Polli Tr. 3.63:2–12; PTX 50 at NOVZ00007869). Novel did not present any evidence to the
contrary to dispute this. Sodium bicarbonate and sodium carbonate in Novel’s ANDA product
also constitute a “buffer” under Claim 1 of the ’628 Patent for the same reasons that these
components constitute a “binary buffer system” under Claim 22 of the ’809 Patent (see analysis
below). (Tr. 3.117:6–3.117:21 (Polli)). That is, because the sodium bicarbonate and sodium
carbonate satisfy the “binary buffer system” element of the ’809 patent, they necessarily satisfy
the broader “buffer” element. Moreover, the ’628 Patent explicitly identifies buffers using sodium
37
bicarbonate and sodium carbonate, like those in Novel’s ANDA product, as preferred
embodiments of the invention. (Tr. 3.117:22–3.118:17 (Polli)). As sodium carbonate is a
“carbonate buffer,” and sodium bicarbonate, a “bicarbonate buffer,” Novel’s ANDA product
indisputably contains the “carbonate buffer” and “bicarbonate buffer” required under the
additional asserted claim, Claim 9, of the ’628 Patent. (Tr. 3.119:10–18; PTX 138 (Polli)). Novel
therefore infringes all remaining claims and elements of the ’628 patent.
B. ’809 Patent
Plaintiffs assert that Claims 11, 17, and 18 of the ’809 patent are infringed by Novel and
DRL only. Additionally, Plaintiffs claim that Novel infringes Claim 22 of the ’809 patent. This
Court previously granted Plaintiffs’ Rule 52 motion that Novel infringes Claims 11, 17, and 18,
therefore isolating this Court’s infringement analysis to Claim 22. (ECF No. 366 at 4).
Similarly, DRL challenges its infringement of the ’809 patent only as to the “appropriate patient
population” element of (independent) Claim 1. 10
1. Novel Infringes Claim 22 of the ’809 Patent.
Novel contested infringement of Claim 22 of the ’809 patent only with respect to the “pH
of said subject’s saliva” limitation. The Court incorporates by reference Section A., 2., b., i., of
this Opinion (finding “the in vitro data is credibly linked to other evidence to prove infringement
10
For reference, Claim 1 of the ’809 patent recites: A solid unit dosage composition for the
treatment of MOTN insomnia, said composition comprising an effective amount of zolpidem or a
salt thereof, formulated for delivery of zolpidem across a subject’s oral mucosa, wherein said
effective amount is an amount of less than 1.30 x 10-5 moles of zolpidem, and between about 25
ng/mL and about 50 ng/mL within 20 minutes of administration, when evaluated in an appropriate
patient population. (JTX 002)(emphasis added).
38
by a preponderance of the evidence for the claim limitation [‘]raises the pH of saliva to a pH of
about 7.8 of greater.[’’’]). For purposes of completeness, the Court provides additional relevant
evidence to Novel only, which demonstrates Plaintiffs have proved Novel infringes Claim 22 of
the ’809 patent.
The face of the ’809 patent indicates that the preferred embodiment of the invention
includes a binary buffer system comprising sodium bicarbonate and sodium carbonate—the same
system as in Novel’s ANDA Product. (Tr. 3.50:18–3.52:4 (Polli); JTX 2 at 28:37–39). Sodium
carbonate is a proton-accepting component and sodium bicarbonate is a proton donating
component. (Tr. 3.51:6-17 (Polli)). Further, Dr. Polli opines that the weight ratio of sodium
carbonate and sodium carbonate in Novel’s ANDA product—2.3 to 1—falls within the preferred
ranges taught in the ’809 Patent to raise the pH of saliva above about 8.5. (Tr. 3.52:5–3.54:3; JTX
2 at 27:37–57, 28:46–48 (Polli)). Additionally Dr. Polli points to the pKa values of carbonate and
sodium bicarbonate to confirm the teachings of the ’809 of achieving and maintaining a pH range
above 8.5. (Tr. 3.55:21–3.56:13, 3:57:18–3.58:6 (Polli)). Finally, it is undisputed that Novel
conducted pH testing in simulated saliva, and reported in its ANDA that the amounts of sodium
carbonate and sodium bicarbonate in its final ANDA Product achieved a pH of 9.56. (Tr. 3.60:7–
23; Stip. Facts ¶¶ 215–22 (Polli)). For these reasons and the corresponding findings articulated in
Section A., 2., b., i., of this Opinion, Novel is found to infringe Claim 22 of the ’809 patent as its
ANDA product contains a system used to “maintain and/or achieve an approximate pH range
comprising at least one proton-donating component and at least one proton accepting component,”
as required by the construction of Claim 22’s “binary buffer system.”
39
2. DRL Infringes the “appropriate patient population” element of the ’809 patent.
Plaintiffs assert that DRL infringes Claims 11, 17 and 18 of the ’809 patent where Claim
11 depends from Claim 1 and Claims 17 and 18 are multiple dependent claims. DRL contests
infringement as to Claim 1 of the ’809 patent only. Independent Claim 1 of the ’809 patent states
in relevant part:
A solid unit dosage composition for the treatment of MOTN
insomnia, said composition comprising an effective amount of
zolpidem or salt thereof … sufficient to produce a plasma
concentration between 25 ng/ml and about 50 ng/ml within 20
minutes of administration, when evaluated in an appropriate patient
population.
(JTX 002) (emphasis added). DRL argues that Plaintiffs failed to prove their ANDA product
infringes this patent because DRL’s ANDA product does not meet the required plasma
concentration at 20 minutes when evaluated in an “appropriate patient population.” DRL first
opposed this element during the summary judgment phase of litigation. That is, the “appropriate
patient population” element was not disclosed through DRL’s non-infringement contentions,
which only disputed Claim 5 of the ’809 patent. (Opinion, ECF No. 325 at 26). Pursuant to
Local Patent Rule 3.7, leave to amend infringement contentions may be granted “by order of the
Court upon a timely application and showing of good cause.” DRL never sought leave to amend
its non-infringement contentions to reflect the arguments it made regarding the “appropriate
patient population” element. Therefore, the Court ordered that DRL would “not be permitted to
offer new evidence of non-infringement regarding the claim limitation ‘appropriate patient
population,’ but shall only rebut the sufficiency of Plaintiffs’ evidence of infringement of this
claim.” (Order, ECF No. 331 at n. 1). With this in mind, the Court evaluates the sufficiency of
Plaintiffs’ evidence and finds DRL has infringed this element.
40
At trial Plaintiffs pointed the Court to DRL’s bioequivalence study and the expert
testimony of Dr. Drover for the majority of its evidence relating to DRL’s ANDA product
meeting the required plasma concentration at 20 minutes when evaluated in an “appropriate
patient population.” DRL commissioned a clinical trial to determine whether its ANDA Product
is bioequivalent to Intermezzo®. (PTX 101). In the clinical trial, which DRL submitted to FDA
as part of its ANDA filing, DRL’s 3.5 mg ANDA Product was administered to 56 subjects. (Id.
at DRL0003800; Tr. 2.185:13–23 (Drover)). Blood was taken from each subject at multiple
points in time, including at 20 minutes after administration where each of the blood samples was
then analyzed to determine the subject’s zolpidem plasma concentration. (PTX 101 at
DRL0003791). The clinical study report includes all of the raw data from the study, as well as
the statistical analysis of that data. According to the report, the average plasma concentration at
20 minutes after administration was 34.46 ng/mL, which is within the claimed range of “between
about 25 ng/mL and about 50 ng/mL.” (PTX 101 at DRL0004169; Tr. 2.187:1–21 (Drover)).
The ’809 Patent defines an “appropriate patient population” to include “a patient
population used for a clinical study.” (JTX 2 at 10:1–5). Dr. Drover explained that the study
population used in DRL’s bioequivalence study, a clinical trial, was typical of that used in other
clinical trials. (Tr. 2.185:13–2.186:25 (Drover)). This study population consisted of 13 men and
43 women, ages 20 to 65 who were chosen on the basis of various inclusion and exclusion criteria
specified in the study report. (PTX 101 at DRL0003784–86). The study subjects were chosen on
the basis of (among other things) their age, weight, and ability to metabolize zolpidem—all of
which the ’809 Patent expressly identifies as being relevant factors in constructing an appropriate
patient population. (JTX 002 at 10:1–5; PTX 101 at DRL0003784–86; Tr. 2.185:21–2.186:4
41
(Drover)). Dr. Drover further explained that these criteria were also typical of those used in clinical
trials.
There is no evidence before the Court that the ’809 Patent requires that men and women
be segregated into separate patient populations as DRL would lead it to believe. (Tr. 2.191:18–21
(Drover)). The ’809 patent lists a number of factors that can be considered in assembling an
appropriate patient population, including “age, weight, the number of hours of time in bed
remaining, and/or the ability of a subject to metabolize zolpidem.” (JTX 2 at 10:1–5). Notably,
the patent does not include gender as one of the exemplary inclusion or exclusion criteria, and thus
indicating that a mixed-gender population is appropriate. (Id.). Dr. Drover explained that even
though the study population for DRL’s bioequivalence study included both men and women, it
was still conducted in an “appropriate patient population,” because even with women included, the
inclusion and exclusion criteria used for the study were typical and similar to those previously
used in published zolpidem studies. (Tr. 2.190:18–25 (Drover)). In light of the aforementioned,
and particularly highlighting the guidance of the ’809 patent itself, DRL’s contention that its
bioequivalence study was not conducted on an “appropriate patient population” because it included
both men and women is unconvincing to the Court.
Similarly, the fact that DRL’s bioequivalence study included one elderly subject does not
overcome the conclusion that the study was performed on an appropriate patient population and
therefore preclude DRL from infringement of this element. (Tr. 2.188:7–10 (Drover)). The
reported 20-minute concentration for the one elderly subject was close to the study population’s
average. (Tr. 2.189:24–2.190:4 (Drover)). Thus, the elderly subject did not alter Dr. Drover’s
opinion that the study was performed in an “appropriate patient population” because the reported
concentration did not meaningfully change the average 20 minute plasma concentration, even
42
though the subject, because she was elderly, is expected to exhibit higher plasma concentrations
than a non-elderly subject. (Tr. 2.188:7–10, 2.189:4–2.190:7 (Drover)). Even assuming arguendo
that the elderly and non-elderly should be considered separate patient populations, Dr. Drover
calculated that, if the one elderly subject were removed from the study, the average plasma
concentration 20-minutes after administration would be 34.3 ng/mL, which is still between the
delineated range of “about 25 ng/mL and about 50 ng/mL,” according to the ’809 patent. (Tr.
2.188:7–10, 2.189:4–2.190:7 (Drover)). DRL’s bioequivalence study, when taken in conjunction
with Dr. Drover’s testimony, establishes that DRL’s tablets contain an effective amount of
zolpidem that is “sufficient to produce a plasma concentration between about 25 ng/mL and about
50 ng/mL within 20 minutes of administration, when evaluated in an appropriate patient
population.”
3. Remaining Uncontested Infringement Evidence
Although Defendant, DRL, did not dispute infringement as to the remaining elements of
Claim 1 nor Claims 11, 17, 18, the Court addresses whether Plaintiffs have met there burden of
proving infringement. Likewise, Claim 22 of the ’809 patent was asserted against Novel and found
by the Court to be infringed. However, because asserted Claim 22 depends from independent
Claim 12 of the ’809 patent. The Court must determine if Claim 12 is also infringed. The findings
below establish that Plaintiffs have met their burden of proving these Claims and elements are
infringed by DRL and Novel by a preponderance of the evidence.
43
a. DRL Infringes Claim 1 of the ’809 Patent
Asserted Claims 11, 17, and 18 depend from independent Claim 1. (JTX 002, Claims 1,
11, 17, 18). DRL’s proposed label establishes the majority of the relevant claim elements. The
first element of claim 1, “[a] solid unit dosage composition for the treatment of MOTN insomnia”
is found in the “Indication and Usage” section of the label, stating that DRL’s ANDA product
comprises “tablets . . . indicated for use as needed for the treatment of insomnia when a middleof-the-night awakening is followed by difficulty returning to sleep.” (PTX 48 at DRL0000783; Tr.
1.237:10–1.238:4 (Kryger)). A “tablet” is “a solid unit dosage composition,” and the indication
shows that the tablets will be used “for the treatment of MOTN insomnia.” (Tr. 1.237:10–1.238:4
(Kryger)). DRL’s proposed label also establishes that its ANDA Product satisfies the second
element of Claim 1, requiring that “said composition compris[e] an effective amount of zolpidem
or a salt thereof[.]” (JTX 002, Claim 1). The Court construed “effective amount of zolpidem” to
mean “amount of zolpidem that is capable of achieving a therapeutic effect in a subject in need
thereof.” (Order, ECF No. 186 at 2). Dr. Kryger explained that a POSA would understand the
“therapeutic effect” to be efficacy in treating MOTN insomnia. (Tr. 1.238:23–1.239:3 (Kryger)).
DRL’s proposed label reports the results of a sleep laboratory study: “Doses of 3.5 mg and 1.75
mg zolpidem tartrate [i.e., DRL’s ANDA Product] significantly decreased both objective and
subjective sleep latency after a scheduled middle-of-the-night awakening as compared to placebo,”
and the results of an outpatient study: “Subjective (patient-estimated) time to fall back to sleep
after middle-of-the-night awakening was significantly shorter for zolpidem tartrate 3.5 mg [i.e.,
DRL’s ANDA product] compared to placebo.” ((PTX 48 at DRL0000795-96; Tr. 1.239:8–1.240:9
(Kryger)). It was therefore demonstrated at trial that a POSA would read this clinical data to
44
demonstrate that DRL’s ANDA Product has an effective amount of zolpidem to achieve a
“therapeutic effect.” (Tr. 1.240:10–13 (Kryger)).
The third element of Claim 1, requiring that “said composition . . . [be] formulated for
delivery of zolpidem across a subject’s oral mucosa,” was shown in DRL’s proposed label and
formulation design choices. A “Guidance for Industry” from FDA indicates that the intended site
of absorption for a “sublingual tablet” is the oral cavity. (PTX 206 at PLSEXP0000256.) In its
proposed label, DRL refers to its ANDA product as a “sublingual tablet,” showing that, under
FDA’s understanding, DRL’s ANDA product is formulated for delivery across the oral mucosa.
(PTX 48 at DRL0000781; Tr. 3.42:4–3.44:11 (Polli)). Moreover, the dosing instructions in DRL’s
proposed label tell a patient to place the product under their tongue, allow it to break apart
completely, and then swallow, which also cooreberates transmucosal delivery as this instruction
facilitates oral absorption. (PTX 48 at DRL0000806; Tr. 3.44:12–24 (Polli)). Finally, DRL’s
ANDA product satisfies the fourth element of claim 1, “wherein said effective amount is an amount
of less than 1.30 x 10-5 moles of zolpidem,” as demonstrated, again, by Dr. Kryger. (Tr. 1.240:17–
1.241:13 (Kryger)). DRL’s ANDA seeks approval for two dosage forms, one containing 3.5 mg
of zolpidem tartrate and the other containing 1.75 mg of zolpidem tartrate. (Stip. Facts ¶ 140; Tr.
1.176:22–1.177:3 (Kryger); PTX 48 at DRL0000781). Dr. Kryger explained that moles and
milligrams are two different ways of measuring the amount of zolpidem in the tablet, where 1.30
x 10–5 moles of zolpidem is equal to 4.975 mg. (Tr. 1.241:1–6 (Kryger)). Thus, both dose
strengths of DRL’s ANDA product contain an amount of zolpidem less than 4.975 mg, and Claim
1 is infringed.
45
b. DRL Infringes Claims 11, 16 and 17 of the ’809 Patent
DRL’s ANDA product satisfies the additional element of Claim 11, “wherein the zolpidem
is delivered across at least one of the sublingual or buccal mucosa,” through its proposed label and
bioequivalence study. Specifically, four pieces of evidence regarding DRL’s ANDA product—
the dosing instructions (directing a patient to put the product under the tongue and allow it to
“disintegrate completely”), higher early plasma concentrations than Ambien® (avoid first pass
effect), shorter lag time than Ambien®, and higher Cmax than Ambien®—prove that the zolpidem
in DRL’s ANDA product is delivered across the sublingual mucosa. (Tr. 2.2.169:8–17 (Drover)).
This evidence was not rebutted.
Regarding Claims 17 and 18, DRL’s ANDA establishes that its ANDA product satisfies
the elements of Claims 17 and 18: “containing about 1.75 mg of zolpidem hemitartrate” (JTX 2,
claim 17), and “containing about 3.5 mg of zolpidem hemitartrate.” (JTX 2, Claim 18). Zolpidem
hemitartrate is another name for zolpidem tartrate. (Tr. 4.24:10–13 (Polli); see also Tr. 5.13:17–
20 (McConville)). DRL’s ANDA seeks approval for two dosage forms, one containing 3.5 mg of
zolpidem tartrate and the other containing 1.75 mg of zolpidem tartrate. (Stip. Facts ¶ 140; PTX
48 at DRL0000781). The former dosage practices the additional element of Claim 18, and the
latter practices the additional element of Claim 17. (Tr. 1.241:7–13 (Kryger)).
c. Novel Infringes Independent Claim 12 of the ’809 Patent
Novel’s proposed label establishes that its ANDA Product satisfies the first and second
elements of Claim 12 for the same reasons DRL infringes Claim 1, as the proposed labels are
identical in this regard. (PTX 50 at NOVZ0007870; Tr. 1.237:10–1.238:4, 1.240:10-13(Kryger)).
46
Novel’s ANDA product also satisfies the third element of Claim 12 requiring that “said
composition . . . [be] formulated for delivery of zolpidem across a subject’s oral mucosa,” as Novel
stipulated that its ANDA product is formulated for delivery across the oral mucosa. (Stip. Facts ¶
234). The fourth element of Claim 12 requires that “said effective amount is 0.5 to 4.75 mg of
zolpidem hemitartrate.” (JTX 2, Claim 12). Novel’s ANDA infringes this Claim as it seeks
approval for two dosage forms, one containing 3.5 mg of zolpidem tartrate and the other containing
1.75 mg of zolpidem tartrate, both of which are between 0.5 and 4.75 mg. (PTX 50 at
NOVZ0007869; Tr. 1.243:10–13 (Kryger)).
Finally, the last element of Claim 12 is established by Novel’s bioequivalence study,
requiring that “said effective amount . . . is an amount sufficient to produce a plasma concentration
between about 25 ng/mL and about 50 ng/mL within 20 minutes of administration, when evaluated
in an appropriate patient population.” Novel commissioned a clinical trial to determine whether
its ANDA Product is bioequivalent to Intermezzo®, using an “appropriate patient population,” as
defined in the ’809 Patent. (PTX 139; Tr. 2.181:13–19 (Drover)). In the study, which Novel
submitted to FDA as part of its ANDA filing, Novel’s 3.5 mg ANDA Product was administered
to 36 subjects and blood was then taken from each subject at different points in time, including at
20 minutes after administration. (Tr. 2.181:20–2.182:20 (Drover); PTX 139 at NOVZ0017682).
Each of the blood samples was analyzed to determine the subject’s zolpidem plasma concentration.
(Id.). According to the clinical study report, the average plasma concentration at 20 minutes after
administration was 29.268 ng/mL, which is within the claimed range of “between about 25 ng/mL
and about 50 ng/mL.” (PTX 139 at NOVZ0017704; Tr. 2.181:20–2.182:20 (Drover)).
Consequently, Novel’s 1.75 mg tablets also contains an effective amount of zolpidem because the
pharmacokinetics of zolpidem are linear where a dose reduced by half for the elderly simply
47
reduces the plasma concentrations by half. (Tr. 1.151:17–25 (Drover); see also Tr. 6.112:1–4
(Michniak-Kohn)). Accordingly, an elderly patient that takes Novel’s 1.75 mg tablet will have
approximately the same blood concentrations as a non-elderly patient taking the 3.5 mg tablet. (Tr.
1.183:19–24 (Drover); see also Tr. 6.116:6–10 (Michniak-Kohn)).
C. ’131 Patent
Plaintiffs claim all Defendants will induce infringement of Claims 8, 10, 18 and 19 of the
’131 patent. Claim 1 of the ’131 patent is an independent claim, and therefore Claims 8, 10, 18,
and 19 depend therefrom. The only element at issue is the “without residual sedative effects”
limitation highlighted below:
Claim 1: A method of treating middle-of-the night insomnia
in a non-elderly patient without prophylactically
administering zolpidem, comprising: dosing the patient with
a pharmaceutical composition … wherein the
pharmaceutical composition permits the patient to awaken at
a time about four hours after dosing without residual
sedative effects.
This Court construed “without residual sedative effects” to mean “with no or minimal subjective
feelings of sedation, as evaluated by: (a) testing acceptably in at least one test exploring
psychomotor performance, attention, information processing, and memory used by those of skill
in the art; and/or (b) demonstrating plasma levels of zolpidem, at an appropriate time point, below
about 20 ng/ml.” (Opinion, ECF No. 185 at 5-7). It is undisputed that the accused products of
Novel, Actavis, and DRL, when tested at four hours after administration, all give zolpidem plasma
levels above 20 ng/ml, and therefore Plaintiffs cannot show that “without residual effects” in
infringed under prong (b) of this Court’s construction.
Specifically, the mean plasma
concentrations of zolpidem at four hours after the administration of Defendants’ ANDA products
48
were: Actavis’ ANDA product yielded
Novel’s ANDA product yielded 25.4037
ng/ml, and DRL’s ANDA product yielded
(PTX 90, PTX 139, PTX 101). Thus the
Court’s inquiry is limited to whether Plaintiffs have proven, by a preponderance of the evidence,
that Defendants’ ANDA products would test acceptably in at least one test set forth in part (a) of
the Court’s construction of the “without residual sedative effects” limitation.
1. Vermeeren Driving Study
Each of Defendants’ proposed labels includes a section entitled “Driving Study” that
reports on the results of a driving performance test conducting on Intermezzo. (Tr. at 1.193:19–22
(Kryger); PTX 46 at ACT-ZOL-0000215; PTX 48 at DRL0000796; PTX 50 at NOVZ0007884.)
This is particularly significant as each of Defendants’ proposed labels state that “[w]hen you wake
up in the morning, be sure that at least 4 hours have passed since you have taken Zolpidem Tartrate
Sublingual Tablet and you feel fully awake before driving.” (PTX 50 at NOVZ0007894; Tr.
1.218:4–1.219:2 (Kryger)). At the outset, the Court notes that Defendants’ argument that the
driving study should be discredited simply because it was not conducted on Defendants’ ANDA
products is, again, dismissed. From a cumulative standpoint, Plaintiffs have met their burden of
proving infringement of this element. To credit this conclusion, the Court points specifically to
the following: 1) Defendants’ ANDA products’ bioequivalence data 11; 2) the fact that each ANDA
product contains the same amount of the active pharmaceutical ingredient (and undisputedly the
only hypnotic agent) zolpidem tartrate; and 3) expert testimony that Defendants’ ANDA products
11
While true that bioequivalence alone, does not prove patent infringement, when viewed in light
of the totality of the evidence, the Court concludes Defendants have infringed the ’131 patent. See
Alza, 388 F. Supp. 2d at 722.
49
are expected to produce no residual sedative effects at four hours after dosing and behave just as
Intermezzo® did in the driving study.
According to the ’131 Patent and confirmed by Dr. Kryger, a driving performance test is
an accepted and reliable test in the art for evaluating residual sedative effects. (JTX 001 at 6:59–
60; Tr. at 1.193:15–1.193:18 (Kryger)).
Under the heading “Driving Study,” Defendants’
proposed labels each describe the study as:
A randomized, double-blind, placebo-controlled, active-control, singlecenter, four period, crossover study in 40 healthy subjects was conducted
to evaluate the effects of middle-of the-night administration of Zolpidem
Tartrate Sublingual Tablets on next-morning driving performance. The
four randomized treatments included Zolpidem Tartrate Sublingual
Tablets 3.5 mg four hours before driving. . .
(Tr. 1.194:3–15 (Kryger); PTX 50 at NOVZ0007884).
A double-blind study means that “the
person and the experimenter do not know what treatment the person is on, whether they are on
placebo or whether they are on a medication.” (Tr. 1.194:19–1.195:6 (Kryger)). The results of the
driving study showed that when driving began 4 hours after taking Intermezzo®, “statistically
significant impairment was not found.”
(Tr. 1.197:24–1.198:18 (Kryger); PTX 50 at
NOVZ0007884). Dr. Kryger explained that a POSA would understand the word “impairment” in
this context to equate to residual sedative effects. (Tr. 1.198:12–18 (Kryger)). Thus, the labels’
statement that there was no statistically significant impairment at 4 hours after dosing means that
there were no residual sedative effects. (Id.).
Defendants, through their expert Dr. Winkelman, challenge the driving study as an
acceptable measurement of sedation. First, Defendants claim that Vermeeren did not measure
sedation at a time of about four hours after dosing because the protocol required that the test
patients were awake and alert 45 minutes prior to starting the driving study, and occurred over the
50
course of 4 to 5 hours after administration. Second, Defendants argue that driving studies in the
prior art literature that measured residual sedative effects did so based on the “statistically
significantly different from placebo” standard rather than the symmetry analysis of subjects who
change from their own SDLP (standard deviation of lateral position) as in Vermeeren. Many of
these arguments overlap those made by Defendants relating to their indefiniteness argument of the
same claim element. Because the Court finds a more detailed discussion is appropriate in the
indefiniteness context, suffice it to say that with reference to Section II., C., 2., c., of this Opinion,
the Court concludes that a driving performance test under these conditions is considered an
accepted and reliable test in the art for evaluating residual sedative effects and the symmetry
analysis used in Vermeeren is a persuasive measurement tool.
2. DSST
The Digit Symbol Substitution Test (DSST) is an accepted test of psychomotor
performance in the art and one of the tests listed in column 6 of the ’131 Patent. Plaintiffs point
the Court to two articles by Roth to support a finding that the Defendants’ ANDA products, which
contain the same active ingredient at the same dosages used in those studies, would also test
acceptably in a psychomotor performance test at 4 hours after dosing. A 2007 article by Roth et
al., published in the journal Human Psychopharmacology, found that both 3.5 mg and 1.75 mg
zolpidem led to no residual sedative effects at 4 hours after dosing, as measured by the DSST. (Tr.
1.221:23–1.222:19 (Kryger); PTX 258 at PIZ00315144). In addition, a 2008 article by Roth et al.
published in the journal Sleep described administering 3.5 mg and 1.75 mg doses of zolpidem to
subjects and found that neither dose led to residual sedative effects at 4 hours after dosing. (Tr.
1.225:2–9 (Kryger); PTX 264 at JNTDEF0000559). Dr. Kryger opined that even when blood
51
plasma levels after subjects were given a 3.5mg dose or 1.75 mg dose of zolpidem are above 20
ng/ml
a composition can still pass DSST.
Specifically, Dr. Kryger explained: “What we are looking at here is the DSST data. So even though
the level was above 20, even though the level was above 20 at four hours, there was no abnormality
with the DSST.” (Tr. 1.227:18-23 (Kryger)). This evidence supports a finding that Defendants’
ANDA products will test acceptably in the DSST psychomotor performance test and therefore
infringe the ’131 patent limitation “without residual sedative effects.”
3. Defendants’ Proposed Labelling
Defendants’ proposed labels are virtually identical to that of Intermezzo®’s but for small
changes such as replacing the word “Intermezzo” with another word describing the particular
Defendant’s product. Thus, the labeling Defendants submitted to the FDA for ANDA approval
encourage infringement. The Parties’ experts agree that there are no differences among the
Defendants’ labels that are relevant to the infringement analysis. (Tr. 1.175:11–1.176:2 (Kryger);
Tr. 7.106:4–9, 8.175:16–21(Winkelman)). By providing instructions for use that when followed
would lead to infringement, each Defendant would induce infringement under 35 U.S.C. § 271(b).
See, e.g., Eli Lilly & Co. v. Actavis Elizabeth LLC, 435 F. App’x 917, 926 (Fed. Cir. 2011) (“We
have long held that the sale of a product specifically labeled for use in a patented method
constitutes inducement to infringe that patent, and usually is also contributory infringement.”);
AstraZeneca LP v. Apotex, Inc., 633 F.3d 1042, 1060 (Fed. Cir. 2010) (“In the context of specific
intent, . . . [t]he pertinent question is whether the proposed label instructs users to perform the
patented method. If so, the proposed label may provide evidence of [an] affirmative intent to
induce infringement.”).
52
If Defendants truly believed their products would cause residual sedative effects, they
could have pursued a non-infringing label. Defendants’ proposed labels include statements
indicating that it is safe for a patient to perform tasks requiring daytime awareness at 4 hours after
dosing, which ultimately the Court finds to describe a lack of residual sedative effects. (Tr.
1.215:3–1.219:2 (Kryger)). Defendants’ proposed labels also state: “Limitations of Use: Zolpidem
Sublingual Tablet is not indicated for the treatment of middle-of-the-night insomnia when the
patient has fewer than 4 hours of bedtime remaining before the planned time of waking.” (PTX
50 at NOVZ0007870). Dr. Kryger explained that this statement implies that the patient can take
the drug with 4 or more hours of time in bed remaining and wake up without residual side effects.
(Tr. 1.215:11–1.216:7 (Kryger)). The Court too sees no alternative reading of the label. In further
support of this point, Defendants’ proposed labels also include a dosing time chart that tells patients
when they would need to take the drug depending on when they have to get up. The dosing time
chart tells patients that they must take the drug at least 4 hours before waking. (PTX 50 at
NOVZ0007891–92; Tr. 1.216:25–1.218:3 (Kryger)).
As a matter of common sense, because Defendants’ proposed labels instruct “[w]hen you
wake up in the morning, be sure that at least 4 hours have passed since you have taken Zolpidem
Tartrate Sublingual Tablet and you feel fully awake before driving,” to argue that Defendants’
ANDA products would produce residual sedative effects four hours after dosing would be
juxtaposed to their labels’ very warnings. This is evidence of infringement as well. To clarify,
the Court is not suggesting that FDA regulations and patent laws can or cannot overlap. Production
of a non-infringing product may be unsafe and compliance with FDA regulations may
induce infringement. However, the two can be reconciled when appropriate. For instance, as
indicated previously, having found that DRL’s and Actavis’ products do not contain “buffer,”
53
innovation beyond the ’628 patent was promoted while safety was maintained. Such is not the
case with the ’131 patent and the limitation of no residual sedative effects. Should the Court find
Defendants’ products are likely to yield, by a preponderance of the evidence, residual sedative
effects, the Court would also be finding that Defendants’ proposed labels are inaccurate. The
Federal Circuit also addressed this tension in concluding “[b]ecause drug manufacturers are bound
by strict statutory provisions to sell only those products that comport with the ANDA's description
of the drug, an ANDA specification defining a proposed generic drug in a manner that directly
addresses the issue of infringement will control the infringement inquiry.” Bayer AG v. Elan
Pharm. Research Corp., 212 F.3d 1241, 1248 (Fed. Cir. 2000). Indeed, the aforementioned
statements in Defendants’ proposed labels indicate to patients that they will be free of residual
sedative effects four hours after dosing therefore infringing the method delineated in the ’131
patent.
4. Plasma Levels and Effects on Residual Sedative Effects
Defendants propose that
a POSA would
conclude that residual sedative effects are likely to be present. However, while higher plasma
concentrations can be indicative of residual sedative effects in some cases, Dr. Kryger explains
that the problem with using the number 20 is that this number 20 “is really a safe harbor, and it is
-- values below 20 would be considered a zero chance of having residual effect from that treatment,
and values above 20, we don't know necessarily what they are.” (Tr. 1.226:2-8 (Kryger)). In fact,
Intermezzo® is an example of a product that when showing plasma levels of zolpidem above 20
54
nanograms per milliliter, is free from residual sedative effects at the appropriate time. (Tr.
1.226:24-1.227:2 (Kryger)). Additionally, to determine infringement, the Court’s inquiry is
directly related to the claim construction of the term. Here, the Court’s claim construction is clear
in stating that residual sedative effects can be evaluated by testing acceptably in one psychomotor
test or demonstrating plasma levels of zolpidem below 20ng/ml. (Opinion, ECF No. 185 at 5-7)
(emphasis added). Thus, this construction uses the word “or” to allow for more than one method
of testing for residual sedative effects for the precise reasons articulated by Dr. Kryger, and a
finding of infringement follows.
5. Remaining Uncontested Infringement Evidence
With regard to the ’131 patent, Defendants only contested the “without residual sedative
effects” limitation discussed above. However, the remaining elements of Claims 1 and 12, as well
as asserted Claims 8, 10, 18 and 19 are also found to be infringed by Defendants. The Court first
reiterates that all experts agree that there are no differences among Defendants’ labels that are
relevant to the infringement analysis. (Tr. 1.17511-1.76:2 (Kryger); 7.106:4-9, 8.175:16-21
(Winkelman)). With this is mind, Plaintiffs used Novel’s label as representative of all Defendants’
labels for purposes of the infringement inquiry.
Claims 10 and 19 require “delivery of zolpidem across the patient’s oral mucosa.” (JTX
001 at Claims 10 and 19). These claims are infringed for the same reasons articulated with regards
to the ’809 patent. (See Section I., B., 3., b., of this Opinion: “Specifically, four pieces of evidence
regarding DRL’s ANDA product—the dosing instructions (directing a patient to put the product
under the tongue and allow it to “disintegrate completely”), higher early plasma concentrations
than Ambien (avoid first pass effect), shorter lag time than Ambien, and higher Cmax than
Ambien—prove that the zolpidem in DRL’s ANDA product is delivered across the sublingual
55
mucosa. (citing Tr. 2.2.169:8–17 (Drover)). Defendants’ proposed labels include an indication
and usage for treating insomnia “when middle of the night awakening is followed by difficulty
returning to sleep,” therefore satisfying a method from treating MOTN insomnia in Claims 1 and
12. (See e.g. PTX 50 at NOVZ0007870). These labels also indicate usage for non-elderly patients
(required by Claims 8 and 10) as well as elderly patients (required by Claims 18 and 19) “without
prophylactically administering zolpidem,” which is required by Claims 1 and 12. (Tr. 1.179:2022, 1.80:10-17 (Kryger)). The quantities of zolpidem hemitatrate (required as a range in Claims 1
and 12, and 3.5 mg and 1.75 mg in Claims 8 and 18) are infringed for the same reasons expressed
regarding the ’809 patent and Defendants’ labels. The only hypnotic agent (required by Claims 1
and 12) in Defendants’ ANDA products is zolpidem tartrate. (Stip Facts ¶¶ 111, 116; Tr. 1.181:251.182:7 (Kryger)). Finally, Defendants’ labels describe a patient desiring to resume sleep “for less
than 5 hours,” therefore satisfying all elements of Claims 1 and 12. Consequently, each of the
asserted claims of the ’131 patent are infringed by Defendants.
II. Patent Invalidity
A. Obviousness
“An obviousness analysis measures the difference between the claimed invention and the
prior art to determine whether ‘the subject matter as a whole would have been obvious at the time
the invention was made’ to a person having ordinary skill in the art.” Unigene Labs., Inc. v. Apotex,
Inc., 655 F.3d 1352, 1360 (Fed. Cir. 2011). Obviousness is a question of law based on underlying
factual findings. Honeywell Int’l, Inc. v. United States, 609 F.3d 1292, 1297 (Fed. Cir. 2010).
“The factual underpinnings, often referred to as the Graham factors, include: 1) the scope and
content of the prior art; 2) the level of ordinary skill in the art; 3) the differences between the
56
claimed invention and the prior art; and 4) evidence of secondary factors, also known as objective
indicia of nonobviousness.” Id. at 1360.
“Obviousness requires more than a mere showing that the prior art includes separate
references covering each separate limitation in a claim under examination. Rather, obviousness
requires the additional showing that a person of ordinary skill at the time of the invention would
have selected and combined those prior art elements in the normal course of research and
development to yield the claimed invention.” Unigene, 655 F.3d at 1360. Moreover, the party
challenging validity must show that a person of ordinary skill in the art “would have been
motivated to combine the teachings of the prior art references to achieve the claimed invention,
and . . . would have had a reasonable expectation of success in doing so.” Procter & Gamble v.
Teva Pharm., 566 F.3d 989, 994 (Fed. Cir. 2009) (quotation omitted). A claimed invention may,
however, be obvious even when the prior art does not teach each claim limitation, so long as the
record contains some reason that would cause one of skill in the art to modify the prior art to obtain
the claimed invention. Beckson Marine, Inc. v. NFM, Inc., 292 F.3d 718, 728 (Fed. Cir. 2002). A
finding of obviousness cannot, however, be based on “the hindsight combination of components
selectively culled from the prior art to fit the parameters of the patented invention.” Crown
Operations Int'l, Ltd. v. Solutia, Inc., 289 F.3d 1367, 1376 (Fed. Cir. 2002) (quoting ATD Corp. v.
Lydall, Inc., 159 F.3d 534, 546 (Fed. Cir. 1998)).
“A person of ordinary skill at the time of the invention interprets the prior art using common
sense and appropriate perspective.” Unigene, 655 F.3d at 1361; see generally KSR Int’l Co. v.
Teleflex Inc., 550 U.S. 398, 420-421 (2007) (“A person of ordinary skill is also a person of
ordinary creativity, not an automaton.”). In the same vein, although an analysis of the teaching,
suggestion, or motivation to combine elements from different prior art references is helpful, this
57
Court’s obviousness analysis requires an “expansive and flexible approach.” Kinetic Concepts,
Inc. v. Smith & Nephew, Inc., 688 F.3d 1342, 1360 (Fed. Cir. 2012).
Finally, Defendants, as the patent challengers, must prove obviousness by clear and
convincing evidence. Tokai Corp. v. Easton Enterprises, Inc., 632 F.3d 1358, 1367 (Fed. Cir.
2011). Clear and convincing evidence is a higher burden of proof than preponderance of the
evidence. See Colorado v. New Mexico, 467 U.S. 310, 316 (1984). To be clear and convincing,
evidence must “place[ ] in the factfinder ‘an abiding conviction that the truth of [the] factual
contentions are highly probable.’ ” Procter & Gamble, 566 F.3d at 994 (quotation omitted).
Defendants assert that the claimed invention is invalid for obviousness because the claims
of the patents-in-suit would have been obvious to a POSA inasmuch as the scope and content of
the prior art teaches all claimed elements. At trial, Defendants, “like all those who seek to prove
claims obvious, was required to show that ‘the differences between the claimed invention and the
prior art are such that the claimed invention as a whole would have been obvious before the
effective filing date of the claimed invention to a person having ordinary skill in the art to which
the claimed invention pertains.’” Galderma Labs., L.P. v. Tolmar, Inc., 737 F.3d 731, 737 (Fed.
Cir. 2013) (citing 35 U.S.C. § 103).
Plaintiffs’ main response to Defendants’ obviousness challenge is that the claimed
invention was not obvious because the prior art does not disclose: 1) efficacious low doses of
zolpidem; 2) non-prophylactic dosing; and 3) transmucosal delivery of zolpidem. Plaintiffs also
argue that a POSA would not have been motivated to combine the prior art because some
references teach away from the claimed invention, nor would a POSA have had a reasonable
expectation of success with low doses or transmucosal delivery of zolpidem. Finally, Plaintiffs
ask the Court to consider the objective indicia of nonobviousness: licensing, industry
58
acquiescence, long-felt need and skepticism. While Plaintiffs as the party defending the patentsin-suit may offer evidence of secondary considerations of nonobviousness, these may not, by
themselves, overcome a strong prima facie case of obviousness. Wyers v. Master Lock Co., 616
F.3d 1231, 1246 (Fed. Cir. 2010). For the reasons set forth below, the Court finds Defendants
have proved a case of obviousness by clear and convincing evidence even after consideration of
Plaintiffs’ purported objective indicia of nonobviousness.
1. Person of Ordinary Skill in the Art
A person of ordinary skill in the art in the technology of the subject matter of the ’131
patent is a person working in the field of sleep therapeutics and has: a least a Ph.D. in clinical
psychology having at least one year of research experience in the field; or an MD having at least
one year of clinical experience in the field. (Tr. 7.90:14-7.91:3 (Winkelman)). While Plaintiffs
provided their own definition, both Parties’ experts agree that none of their opinions concerning
the ’131 patent would change depending on which party’s definition of a POSA is adopted.
On the other hand, a person of ordinary skill in the art to whom the ’809 and ’628 patent
would be directed would have at least a bachelor’s degree, and more likely a Master’s or Ph.D.
degree in pharmacy or a related science, and most likely several years of experience formulating
active pharmaceutical ingredients, including some experience in transmucosal delivery. If this
person of ordinary skill had a Bachelor’s or Master’s in pharmacy, or any other related subject,
such a person would typically have more than five years of experience formulating active
pharmaceutical ingredients. If they already had a Ph.D. in pharmacy, they would typically have
fewer years of experience. If a person of ordinary skill in the art did not have actual experience
with the developing transmucosal dosage forms, that person would at least have a deep knowledge
59
of the related scientific literature on the topic and be able to understand that scientific literature.
(Tr. 6.31:1-24 (Michniak-Kohn)). While the Court employs Defendants’ proposed definition,
Plaintiffs’ formulation expert agrees that his opinions on these patents would not alter depending
upon which definition of a POSA was adopted. (Tr. 9.186:6-13 (Polli)).
2. Scope and Content of the Prior Art
In conducting the obviousness analysis, this Court views the claimed invention in light of
the art that existed at the time the invention was made. See 35 U.S.C. § 103(a); Uniroyal, Inc. v.
Rudkin-Wiley Corp., 837 F.2d 1044, 1050–51 (Fed. Cir. 1988). “Prior art has been defined as
follows: ‘[t]he existing state of knowledge in a particular art at the time an invention is made. It
includes the issued patents * * *, publications, and all other knowledge deemed to be common
thereto such as trade skills, trade practices, and the like,’ ” available a year or more before the
patent filing date.” Trio Process Corp. v. L. Goldstein's Sons, Inc., 461 F.2d 66, 69 n. 3 (3d Cir.
1972) (quoting A. Smith, PATENT LAW, CASES, COMMENTS AND MATERIALS 2 (1964)).
As previously stated, the asserted ’131 and ’809 patents are analyzed according to the prior
art as of May 2005, while the ’628 patent is compared to the prior art as of February 2004. The
Court notes that none of the prior art described below is applicable to just one patent but not the
others by virtue of its published date. The Court considers all the teachings in the prior art in the
obviousness determination, “including that which might lead away from the claimed invention.”
In re Dow Chem. Co., 837 F.2d 469, 473 (Fed. Cir. 1988).
It was well known by February 2004 that zolpidem was suitable for treating insomnia.
Indeed, by that point in time, Ambien® (active ingredient being zolpidem tartrate) was the most
popular sedative hypnotic for treating insomnia. The Ambien® label indicates it is used to “treat
60
different types of sleep problems” including “waking up often during the night.” (JTX 41). Thus,
before the priority dates of the patents-in-suit, a POSA would have been motivated to develop a
treatment for MOTN insomnia that would be better than or as good as Ambien®. The Parties do
not dispute that, given Ambien®’s success, a POSA aiming to treat MOTN insomnia specifically
(or in the case of the ’628 patent, insomnia generally) would have carefully considered the
disclosures of Ambien® before or in conjunction with research of any additional prior art.
a. Ambien®
Ambien®, comprising the single hypnotic agent, zolpidem, was commercially released in
1992 and approved by the FDA “for the short-term treatment of insomnia.” (JTX 041 at
DRL0013526). Ambien® was the most successful drug for treating insomnia in 2004 where
physicians prescribed Ambien® to 24 million patients. (Tr. 8.21:4-17 (Winkelman)). At the time
of the invention, the recommended dose of Ambien® was 10 mg for non-elderly adults, and 5 mg
for elderly adults. The Ambien® label indicates it is also effective to treat MOTN insomnia stating
Ambien® “is used to treat different types of sleep problems, such as: trouble falling asleep; waking
up too early in the morning; and waking up often during the night.” (JTX 041 at DRL0013528)
(emphasis added).
While it is undisputed that Ambien® was indicated for MOTN insomnia, Ambien®’s label
instructed patients to take Ambien® at bedtime, therefore treating MOTN insomnia through
prophylactic dosing. Specifically, Ambien®’s label stated: “Do not take Ambien unless you are
able to get a full night’s sleep before you must be active again.” (JTX 029 at 3195). Further, the
Ambien® label explains “Daytime drowsiness is best avoided by taking the lowest dose possible
that will still help you sleep at night. Your doctor will work with you to find the dose of Ambien
that is best for you.” To Dr. Winkelman, this established the two main goals of a sleep aid:
61
effective to help the sleep problem but avoid residual sedative effects. (Tr. 7.164:3-24
(Winkelman) quoting JTX 041). Dr. Winkelman’s assessment is both reasonable and credible to
this Court.
As a general premise, Ambien® and the patents-in-suit, collectively, differ in three
overarching ways. It is no surprise that these three differences are also the limitations Plaintiffs
argue are either absent from the prior art or were taught away from by the prior art. First, Ambien®
is an oral swallow tablet while the asserted claims of the patents-in-suit each provide delivery of
zolpidem across the patient’s oral mucosa. (See ’131 Patent, JTX 001 at Claims 10 and 19; ’809
Patent, JTX 002 at Claims 1, 11, and 12; ’628 Patent, JTX 003 at Claim 1). Next, Ambien® is
available in higher doses of 10 mg (for non-elderly) and 5 mg (for elderly) while the patents-insuit disclose either the range of dose being 0.5mg to 5.0 mg of zolpidem, (See ’131 Patent, JTX
001 at Claims 1 and 12; ’809 Patent, JTX 002 at Claims 1 and 12; ’628 Patent, JTX 003 at Claims
16 and 17), or 3.5 mg (for non-elderly) and 1.75 mg (for elderly). (See ’131 Patent, JTX 001 at
Claims 8 and 18; ’809 Patent, JTX 002 at Claims 17 and 18). Finally, Ambien® is indicated for
prophylactic administration while the ’131 patent is a method to be used at the time of need, not
prophylactically administering zolpidem. (See JTX001 at Claims 1 and 12). The Court now turns
to a determination of whether it was obvious to treat MOTN insomnia by delivering zolpidem
transmucosally, in low doses, and non-prophylactically. Based on the totality of the evidence
presented, the Court is of the abiding conviction that in light of the prior art, these differences
between Ambien® and the patents-in-suit were obvious to a POSA.
b. Prior Art Relevant to Transmucosal Delivery: Tauber and Pinney
Formulation for transmucosal delivery dates back to the 1902 treatment of angina. (Tr.
6.83:14-23 (Michniak-Kohn)). To formulate drugs for delivery across the oral mucosa, a POSA
62
would predict how much drug is available in its un-ionized form for a particular pH because “it is
the un-ionized … that will actually cross membranes.” (Tr. 6.84:12-15 (Michniak-Kohn)). In
1917, the Henderson-Hasselbach equation was established to aid in this function. Dr MichniakKohn described the Henderson-Hasselbach equation as “Chemistry 101,” alerting the Court it was
well within the knowledge of a POSA. (Tr. 6.84:22-25 (Michniak-Kohn)). This equation was
used to provide a known pH range for un-ionized zolpidem based on a known p/k/a. (Tr. 6.161:210 (Michniak-Kohn)). Thus, the Court is convinced that once a POSA had the p/k/a, they could
then conclude that most of the zolpidem will be unionized at pHs about above 7.8. (Id.).
In the same vein, the Court is also convinced that the efficiency of absorption of drugs in
the oral cavity was explained by the prior art Beckett (1967), which disclosed raising the pH to
promote absorption. (Tr. 6.85:8-20 (Michniak-Kohn)). Furthermore, the Parties do not dispute
that sublingual tablets in general, were established well before 2003. (Tr. 6.73:22-24 (MichniakKohn)).
However, the first disclosure of transmucosal delivery in relation to the treatment of
insomnia, came in 1984.
The 1984 Tauber study (“Plasma Levels of Lormetazepam After
Sublingual and Oral Administration of 1 mg to Humans”) measured plasma levels after oral
administration of sublingual 1 mg lormetazepam (“sleeping wafer”). The results showed, on
average, an earlier rise in the lormetazepam levels after sublingual administration as compared to
oral, specifically, the sublingual dosage produced statistically higher levels between 7.5 and 25
minutes than the oral tablet. Therefore, Tauber ultimately concluded that “[i]t is anticipated that
subligual administration of the new formulation will lead to 40-50% reduction of sleep latency.”
(Id. at 1587, 1591).
These findings aligned with the two requirements— according to Tauber—that from a
pharmacokinetic point of view, a modern hypnotic should fulfill. These are the following: 1) “the
63
plasma levels of the active ingredient should increase immediately after administration to
guarantee that the patient will fall asleep;” and 2) “after induction of sleep the plasma level of
pharmacologically active substances should decay rapidly in order to reduce the possibility of
hangover effects, drug accumulation and possible late interactions, e.g. with alcohol.” (DTX 066
at 1587-1588). From a practical standpoint, Tauber described the clinical advantages of sublingual
administration of hypnotics including: 1) convenient administration as “no glass of water is
necessary;” 2) easy dissolution without leaving behind excess undissolved material; 3) rapid
absorption through the oral mucosa, (avoiding the first-pass metabolism in the liver) resulting in
“prompt onset.” (Id. at 1596). These advantages to transmucosal delivery in the hypnotics context,
continue to date. (See also Zhang (2002), JTX 038: “Oral transmucosal technology offers an
alternative means for administering drugs. It allows more rapid absorption into the bloodstream
than is possible with oral administration to the gastrointestinal tract. Oral transmucosal
administration is noninvasive, nontechnical, and convenient for patients;” see also Tr. 9.70:11-21:
(Moline), Agreeing with the conclusions of Zhang.)).
The prior art Pinney is a patent application dated November 29, 2001. The invention
disclosed in Pinney, “Chewing Gums, Lozenges, Candies, Tablets, Liquids, and Sprays for
Efficient Delivery of Medications and Dietary Supplements,” is summarized as the following:
A transmucosal delivery system according to the invention
comprises a carrier suitable for oral administration. A buffer is
dispersed within the carrier, and there is sufficient buffer to achieve
a predetermined pH within the oral cavity of a user. An active
ingredient is dispersed within the carrier, at least a portion of the
active ingredient being unionized at the predetermined pH for
transmucosal absorption within the oral cavity.
(DTX 062 at 4). Pinney confirms one advantage in Tauber, namely, that this system avoids the
“first pass effect” through the liver of swallowed tablets which can lead to only a small fraction of
64
the amount of the active ingredient entering the bloodstream. (Id. at 1-2). Put differently, a higher
bioavailability of active ingredients may be achieved by transmucosal delivery than by oral
ingestion. (See Tr. 6.136:17-22 (Michniak-Kohn) Explaining that Pinney says you have to take
into account the drug’s bioavailability).
Pinney then goes on to describe ideal characteristics of active ingredients (acidic) and
buffers (suggesting citric acid) but warns that “under pH conditions in the mouth (pH 6.0 to pH
7.0), many of the useful compounds would be highly ionized and would not be efficiently absorbed
into the bloodstream by that transmucosal route.” (Id. at 4, 6). Pinney denotes a pH of 7-10 in
mouth conditions for “efficient absorption of most active ingredients.” (Id. at 7). Specifically
relevant, Pinney explains that “tablets” as in the patents-in-suit, are dosage delivery systems for
medicants that are placed in the mouth or under the tongue for rapid dissolution of active
ingredients and absorption through epithelial, where the “dissolution times” should be “preferably
in the range of 5-15 minutes.” (Id. at 8, 11). While Pinney lists zolpidem as a medicant suitable
to transmucosal delivery, it does not specifically disclose how to formulate zolpidem nor is it
targeted towards a method for treating insomnia. (Id. at 13).
c. Prior Art Relevant to Low Doses
At the time of the inventions, the lowest recommended dose of Ambien® and therefore
zolpidem, was 10 mg for non-elderly adults, and 5 mg for elderly adults. Ambien® undoubtedly
taught a POSA that the elderly should receive half the dose of zolpidem than a non-elderly patient.
(See also Tr. 7.186:12-14: “[I]n the elderly, we need to lower the dose;” citing Olubodun (2003),
JTX 026). However, these doses was for a full night’s sleep (not half a night’s sleep), taken at
bedtime. At bedtime a person’s drive to sleep is at its peak, as opposed to the middle of the night,
after some sleep has occurred. This distinction results from the interaction of two biological
65
processes, the circadian drive and homeostatic sleep drive. 12 The Borbély model (“Figure 4”),
published in Borbély et al., “A two Process Model of Sleep Regulation,” describes this interaction
and its effect on a person’s overall propensity to sleep. (DTX 211 at Fig. 4) (emphasis added).
According to the Borbély model a person’s drive for sleep—taking into account the homeostatic
sleep drive and the circadian drive—is much greater at bedtime than in the middle of the night and
therefore, it may be more difficult for a person to return to sleep.
As Ambien® was only prescribed for prophylactic dosing, one prior art reference
suggested a way to combat MOTN insomnia even when one’s sleep drive is lower. That is, Jacob
Teitelbaum, M.D., published the book “From Fatigued To Fantastic,” in 2001 which included a
chapter where zolpidem was explained as a prescription medication to aid in “A Good Night’s
Sleep.” (JTX 33 at 105, 115). Teitelbaum explained:
I like Ambien [Zolpidem] because it is short-acting (that is, less likely
to leave you hungover)… Because it is short-acting, it may not keep
you asleep all the way through the night but will likely give you four
to six hours of good, solid sleep as a foundation. The normal dosage
is one-half to one 10 milligram tablet, taken at bedtime. If you wake
up in the middle of the night you can take an extra one-half to one
tablet (leave it by your bedside with a glass of water) and any sedation
is usually worn off by the time you are ready to wake up in the
morning. One-half tablet is usually enough for the middle of the
night.
(JTX 33 at 115-116) (emphasis added). Thus, Teitelbaum suggests taking a total of 15 mg of
zolpidem, 10 mg at bedtime and 5 mg in the middle of the night. However, Teitelbaum’s further
guidance that “[o]ne-half tablet is usually enough for the middle of the night,” certainly suggests
that 5 mg is effective even when overall propensity to sleep is decreased. This is not surprising
considering that the 1995 reference, Roth et al., found 7.5 mg of zolpidem to be effective to treat
12
Explained in “Background, V.,” of this Opinion.
66
transient insomnia. (JTX 30).
Transient insomnia is “occasional episodes of acute sleep disturbance.” (JTX 30 at 246).
Roth et al., “Zolpidem in the Treatment of Transient Insomnia: A Double-Blind, Randomized
Comparison With Placebo,” used a testing model known as the “first-night effect” to examine the
effects of zolpidem (5, 7.5, 10, 15 and 20 mg) on transient insomnia in a large subject population.
The “first-night effect” models transient insomnia in healthy subjects because subjects, on their
first night in a sleep laboratory, will sleep less well. (See Tr. 10.96:19-10.97:9 (Czeisler)).
Subjects were dosed with zolpidem or placebo at bedtime and then awakened 8 hours later to
perform various tests. Roth only conducted statistical analysis of the 7.5 mg and 10 mg doses, but
provided the resulting data related to sleep latency inclusive of the 5 mg dose in the table below:
TABLE 2 (JTX 30 AT 248) (OMITTED)
(JTX 30, “Table 2” at 248). Roth concluded that 7.5 mg and 10 mg doses of zolpidem were
effective in treatment of transient insomnia. While Roth did not draw any conclusions related to
the low 5 mg dose, in 1996, the Walsh article analyzed Roth’s data stating: “Disregarding dose,
zolpidem was highly effective in shortening latency to persist sleep and in reducing the number of
67
nighttime awakenings and time spent awake after sleep onset.
These effects were highly
significant in the groups that had received 7.5 of 10 mg zolpidem, but not in the 5 mg group
(although numerical trends were evident at this dose.” (PTX 282 at 130). Dr. Winkelman,
however, concluded that the 5 mg dose in Roth, if a statistical analysis had been done, it would
have been “statistically significant,” and therefore, effective.
(Tr. 7.221:18-7.222:10
(Winkelman)). Given the Roth (see “Table 2” above) data shows that 5 mg of zolpidem was shorter
by 8 to 9 minutes than placebo in getting people to sleep, the Court agrees. (Tr. 7.220:18-20
(Winkelman)).
While intuitively a POSA may conclude from the Borbély model and Roth that higher doses
may be needed in the middle of the night, this goal of efficacy must always be counterbalanced
with the requirement of no residual effects in the morning. (See Tr. 7.145:11-19 (Winkelman)
Part of being a physician is that with any drug “you always want the lowest effective dose,”
balancing a dose that is effective but also not giving a dose too high which results in side effects.).
With this in mind, the relevant prior art as a whole, 13 including Roth, which Plaintiffs opine would
lead a POSA away from low doses of zolpidem, was unconvincing to the Court. Thus, the Court
finds that a POSA would undoubtedly attempt to find the lowest effective dose of zolpidem to
avoid potential residual sedative effectives. Coincidently, the prior art reference Merlotti, set out
to do just that.
Merlotti is titled: “The Dose Effects of Zolpidem on the Sleep of Healthy Normals.”
(DTX 063). In 1989, Merlotti performed a dose-ranging study for zolpidem in non-elderly, healthy
13
A few redundant, unpersuasive, or disconnected references have been omitted from this
Section for sake of brevity. The Court however, considered all the prior art admitted in evidence
and explained by the experts at trial, prior to concluding.
68
(without insomnia) subjects. Before bed, subjects received zolpidem (2.5, 5.0, 7.5, 10.0 or 20.0
mg) or placebo. On the third night of each treatment, subjects always received placebo. Merlotti
specifically sets out to address “what is the lowest dose of zolpidem that consistently produces
hypnotic activity in normal volunteers?” (DTX-063 at 1). The study analyzed two measurements
of sleep induction: 1) “wake before sleep,” which is minutes of wake before persistent sleep; and
2) “latency to persistent sleep,” which is minutes from the beginning of the recording to the start
of the first 10 consecutive sleep minutes. (DTX-063 at 11).
As Dr. Winkelman explained, Merlotti’s results showed the 5 mg dose of zolpidem was
“statistically superior in getting people to sleep than placebo.” (Tr. 7.170:18-21 (Winkelman)).
Merlotti ultimately concluded that zolpidem is hypnotically active at doses lower than previously
tested including the 5.0 mg dose. (In contrast see Vogel (1988): “[F]indings indicate that zolpidem
was an efficacious hypnotic in the treatment of transient insomnia. Efficacy, defined as significant
difference from placebo, usually occurred at doses of 7.5 mg and above. The drug improved both
sleep latency and sleep maintenance. Its effect on sleep maintenance occurred only during the first
4 hours of bedtime.” (JTX 036 at 67)).
Although the Merlotti study was done at bedtime, when sleep drive is at its highest, the
prior art reference Kim tackled the issue of whether doses of hypnotics are hindered by less sleep
drive. The Kim reference is titled: “Dose and Time Dependent Discrimination of Daytime
Sleepiness Measured by Multiple Sleep Latency Test (MSLT), Psychomotor Performances Tests
(PPT), and Stanford Sleepiness Scale (SSS) after a Single AM Administration of a Sedative
Hypnotic Drug.” (DTX 197).
Kim explained that MSLT, PPT, and SSS index physiological,
manifest, and introspective factors of sleepiness, but assessing these tests at peak drug effect after
nighttime administration is confounded by the subjects’ natural circadian drowsiness. Thus, in the
69
Kim study, MSLT, PPT, and SSS tests were performed to understand the dose and time dependent
influence of zolpidem after AM administration in well-rested healthy volunteers. Zolpidem was
administered in either a 5 mg or 10 mg doses (some received placebo) in the morning. The results
showed that a significant decrease in sleep latency on the first four tests was found with zolpidem
at both doses, so even the 5 mg dose of zolpidem was able to get subjects to sleep in statistically
significantly shorter time than placebo. From these findings, the authors concluded: “In wellrested healthy volunteers, AM administration of zolpidem produced sedation as demonstrated by
changes in physiological, manifest, and introspective measures of daytime sleepiness.” (Id. at
JNTDEF0006712).
Finally, because the patents-in-suit all differ from the aforementioned method of
administration—Teitelbaum, Roth and Merlotti all observed zolpidem in its oral (swallow) form
of administration—it is important to determine if a prior art reference would lead a POSA to
understand an effective dose of systemic administration, specifically transmucosal administration.
Defendants argue that this alternative route would be noted by a POSA to achieve the most rapid
onset possible in the middle of the night and also avoiding residual sedative effects. Defendants
assert that the change in administration can be reconciled by the prior art, Patat. (See Tr. 7.204:28 (Winkelman): Based on Patat, “[i]f you give systematic administration of zolpidem, you are
going to get a more rapid appearance of indicators of sleep,” as subligual is also one kind of
systemic delivery). The Patat reference studied pharmacodynamics and pharmacokinetics of
zolpidem after daytime administration both orally (swallow) and intravenous (IV). While Patat
did not study transmucosal delivery, it did measure IV which is also systemic. Specifically, the
Patat reference measured EEG (electroencephalogram) brain waves and results using the Stanford
sleepiness scale (SSS). Appropriately, this 1993 prior art was titled “EEG profile of intravenous
70
zolpidem in healthy volunteers.” (JTX 028).
As the method employed, subjects were given zolpidem in the morning (5 mg, 10 mg, or
20 mg) either by mouth to swallow (yielding 70% bioavailability) or intravenously (yielding 100%
bioavailability) and required to stay awake throughout the study. Patat explained, “delta activity
appeared rapidly 10 minutes after IV administration of zolpidem and between 20 and 45 minutes
after oral administration of zolpidem.” Delta waves are the slow waves that are present at the
deepest stage of sleep (or delta activity). (Tr. 7.202:24-7.203:3 (Winkelman)). Thus, their slowing
of the EEG showed significant sleepiness more quickly through IV administration. Additionally,
the results of the SSS (describing how sleepy a person feels) show that at four hours after
administration, the 5 mg dosage was no different from placebo indicating no expectation of
residual sedative effects. In sum, the authors concluded that “EEG changes and scores of SSS
were in good correlation with what has been observed with insomniac patients. Zolpidem has a
rapid onset and a short duration of action, whatever the route and the dose.”
(Id. at
JNTDEF0004144).
d. Prior Art Relevant to Non-Prophylactic Dosing
As previously stated, Ambien® was indicated for prophylactic administration.
Prophylactical administration in this context refers to taking Ambien® every single night at
bedtime “whether or not you are going to have insomnia that night.” (Tr. 10.180:1-10 (Kryger)).
Alternatively, non-prophylactic administration is pro re nata or “as-needed” dosing.
(Tr.
10.180:11-19 (Kryger)). Prior to the filing of the patents-in-suit, the prevailing approach for
treating insomnia was prophylactic administration. (See e.g. Teitelbaum, JTX 033 at 115-116:
“normal dosage is one-half to one 10 milligram tablet, taken at bedtime.” (emphasis added)).
Further, even though Ambien® was indicated for treatment of MOTN insomnia, it only instructed
71
use at bedtime, not in the middle of the night. However, the disadvantages of prophylactic
treatment and the need for flexibility in treating MOTN insomnia were well established in the prior
art literature (see below: Doghramji, Danjou, Hindmarch) and clearly within the knowledge of a
POSA. (See e.g. (Tr. 7.143:7-12 (Winkelman): “[F]or people who only wake up in the middle of
the night sometimes, you’re giving them medication when they don’t need it, because how can
you predict when you’re going to wake up in the middle of the night.”). But it remained clear that
“as-needed” administration was in constant tension with lingering residual sedative effects.
i. Danjou Reference
In 1999, Danjou purported to compare the duration of residual hypnotic and sedative
effects of zaleplon with those of zolpidem (and placebo) following nocturnal administration at
various times before morning awakening. The study used a 10 mg dose of zaleplon as well as a
10 mg dose of zolpidem, each administered orally (or swallowed). The subjects were then “gently”
woken up—mimicking MOTN insomnia— at various times during the time and residual sedative
effects were measured using the psychomotor performance and memory tests: digital symbol
substitution test (DSST), critical flicker fusion (CFF) threshold, choice reaction time (CRT),
memory test (word list), and Sternberg memory scanning. The results showed that no residual
effects were demonstrated after zaleplon 10mg was administered as little as 2 hours before waking.
Zolpidem 10 mg however, showed significant residual effects on DSST and memory after
administration up to 5 hours before waking.
Residual effects were also shown using CFF
threshold and Sternberg memory scanning after administration up to 4 hours before waking.
Danjou stated that the lack of residual sedative effects for zaleplon results are consistent with its
pharmacokinetic profile featuring rapid absorption, distribution and clearance. Therefore, the
72
conclusion was that zaleplon at the 10 mg dose (recommended) dose, would “seem to provide
physicians with a hypnotic free of residual effects at least in normal volunteers.” (JTX 015 at
JNTDEF0003922).
ii. Hindmarch Reference
Similar to Danjou, Hindmarch was a study in 2001 where the objective was to assess
residual effects of zaleplon and zolpidem after a middle of the night administration. Subjects
received placebo, 10 mg or 20 mg of zolpidem, or 10 mg or 20 mg of zaleplon. The results showed
that zaleplon 10 mg had no or minimal residual effects when administered in the middle of the
night as little as one hour before waking. Zolpidem 10 mg produced significant detrimental
residual effects in various tests when administered 3-5 hours before waking with the exception of
the CFF test. The conclusion hypothesized that the lack of clinically significant residual effects
with zaleplon may be explained by its unique pharmacokinetic profile of rapid elimination halflife, “providing some advantages over existing treatments[] for the management of insomnia and
sleep disturbance… even when administered in the middle of the night.” (PTX 256 at 166).
iii. Doghramji Reference
The 2000 reference Doghramji is titled “The Need for Flexibility in Dosing Hypnotic
Agents.” (JTX 016). This prior art was not a clinical study but rather an article specifically
targeting MOTN insomnia.
Doghramji describes non-prophylactic dosing, stating “[t]he
intermittent occurrence of most insomnia suggests that treatment is best accomplished by using
hypnotics on an ‘as needed’ basis.”
(JTX 016 at JNTDEF0000168).
This is because
“[e]pidemiologic studies suggest that insomnia does occur on a regular basis in most people.” (Id.
73
at JNTDEF0000168). Indeed, Doghramji touts the various advantages to “as needed dosing” to
treat insomnia including: 1) optimal management while using the lowest amount of a drug; 2)
providing a patient with a “sense of control” which prevents insomnia from being a “significant
problem;” and 3) potentially reducing anticipatory anxiety prior to sleep. (Id. at JNTDEF0000170)
(See also Scharf (2001): “hypnotic therapy administered prior to bedtime, as is currently
recommended for most compounds, is not appropriate for all insomnia patients” (PTX 477 at 20))
Doghramji was limited to analyzing three hypnotics for flexible administration: triazolam,
zolpidem, and zaleplon. Triazolam was dismissed from the onset because it is an agent with
“drawbacks that are not ideal from treatment of insomnia.” (Id. at JNTDEF0000169). Ultimately,
it was determined that zaleplon is best suited for MOTN administration because of its half-life of
one hour (while admitting that zolpidem has a rapid onset and short half-life). The article focuses
on two trials conducted on zaleplon, (including Danjou), where a lack of residual sedative effects
followed MOTN zaleplon administration. On the basis of these two tests, Doghramji concluded
that “zaleplon appears to be suited for flexible use on an as-needed basis.” (JTX 016 at
JNTDEF0000171) (See also Scharf: “clinical trial data presented suggest zaleplon may represent
an important breakthrough…Clinicians are now able to focus on a specific sleep disturbance by
prescribing a medication that can be administered on intermittent nights only when symptoms
occur—at bedtime or during the night—so long as 4 hours remain prior to a scheduled awakening.”
(PTX 477 at 23)).
3. Differences between the Prior Art and Claimed Invention
While it is clear that the prior art certainly fills in many of the prominent gaps between
Ambien® and the patents-in-suit, some differences remain outstanding. First, while Pinney and
Tauber teach transmucosal delivery of hypnotics, they do not teach how (or if) zolpidem can be
74
delivered this way. Next, while the Court is persuaded by clear and convincing evidence that
Merlotti certainly concludes that lower doses of zolpidem such as 5 mg are effective, but the
patents-in-suit specify even lower dose ranges and specific doses for non-elderly (3.5 mg) and
elderly (1.75 mg). (See e.g. ’131 Patent, JTX 001 at Claims 1 and 12; ’809 Patent, JTX 002 at
Claims 1 and 12; ’628 Patent, JTX 003 at Claims 16 and 17). Finally, while Doghramji, Danjou,
and Hindmarch teach that MOTN insomnia is best treated non-prophylactically, each of these
references ultimately concludes that zaleplon—not zolpidem—is best suited for MOTN insomnia.
In addition to pointing out these differences, Plaintiffs argue that there was no reasonable
expectation of success in either transmucosal delivery of zolpidem, claiming this formulation is
“unpredictable,” or low effective doses, as a POSA, with knowledge of the Borbély model, would
have assumed the dose must be increased when sleep drive is lower in the middle of the night.
Plaintiffs also assert that claim-by-claim, a few additional elements— relevant to the ’809 and
’628 patent exclusively — are lacking in the prior art. The Court addresses each issue in turn.
a. Transmucosal Delivery of Zolpidem was Obvious to a POSA.
The Court finds that a POSA would have been motivated to achieve the most rapid action
possible when formulating a drug to be taken in the middle of the night, as any delay in onset
necessarily results in less sleep. When formulating a hypnotic, Tauber clearly explained that rapid
onset is one of the main goals, stating, “the plasma levels of the active ingredient should increase
immediately after administration to guarantee that the patient will fall asleep.” (DTX 066 15871588). Pinney taught a POSA that transmucosal delivery would achieve the drug in the
bloodstream within minutes of application (yielding high initial plasma levels), rather than
approximately 30 minutes with conventional oral (swallow). (Tr. 6.139:16-20 (Michniak-Kohn)).
75
Indeed, Dr. Michniak-Kohn agrees that oral (swallow) administration results in the relief of
symptoms being “substantially delayed,” which would, in this Court’s view, alert a POSA that
transmucosal is faster. (Tr. 6.137:1-25 (Michniak-Kohn)). This evidence is clear and convincing
to the Court.
At trial, Plaintiffs presented examples of different medicants—with the remarkable
exclusion of zolpidem—which did not produce initially higher plasma concentrations in the
sublingual dosage form. (See e.g. Tr. 9.110-9.112 (Drover) “[E]rgoloid mesylate[]… oral
formulation generated a higher plasma concentration earlier than the sublingual formulation.”).
The Court finds this evidence unpersuasive as it fails to draw a parallel to zolpidem even though
zolpidem tartrate’s pharmacodynamics and pharmacokinetics were well understood, published and
established by 2003. (Tr. 6.94:16-21 (Michniak-Kohn)). What is, however, most persuasive to
the Court were the properties of zolpidem, as explained by Dr. Michniak-Kohn, that would indicate
to a POSA that it can be delivered transmucosally. These include: 1)logP; 2) p/k/a; 3) solubility;
and molecular weight. (Tr. 6.155:14-6.158:4 (Michniak-Kohn)).
The LogP, at 2.42, tells a POSA that zolpidem is lipophilic which means it passes more
easily through membranes. (Tr. 6.156:9-23 (Michniak-Kohn)). P/k/a values of zolpidem were 6.9
and 6.16 which, when plugged into the Henderson-Hasselbach equation, calculate a value that
predicts how much drug is available in its un-ionized form for a particular pH to cross membranes.
(Tr. 6.157:1-23, 6.84:12-15 (Michniak-Kohn)). Zolpidem’s molecular weight is 307.4 grams per
mole which is a “suitable size for passing through mucosal membranes.”
(Tr. 6.158:4-12
(Michniak-Kohn)). Plaintiffs did not rebut this evidence. However, as it pertains to the final
property, solubility, Plaintiffs suggest that a POSA would have not had a reasonable expectation
of success in dissolving zolpidem tartrate in the mouth, which is required for delivery across the
76
oral mucosa. Defendants respond by pointing to the Material Safety and Database Sheet (MSDS)
which shows that zolpidem solubility is “water solubility 23 mg per ml in water at 20 degrees
centigrade.” (DTX 302). Furthermore, Dr. Michniak-Kohn explains that there is at least 1
milliliter in the mouth and the mouth is warmer than 20 degrees centigrade so there “wouldn’t be
a problem to dissolve it.” (Tr. 6.150:20-151:11 (Michniak-Kohn)). The Court finds these
deductions rational and credible despite Plaintiffs’ expert, Dr. Polli’s vague criticism that Dr.
Michniak-Kohn fails to take into account that zolpidem’s solubility is going to be “pH dependent.”
(Tr. 9.1201:15-19 (Polli)).
In response to Plaintiffs’ final objections to a finding that transmucosal delivery of
zolpidem was reasonably expected by a POSA to be successful, the Court notes that the Federal
Circuit has made clear that “obviousness cannot be avoided simply by a showing of some degree
of unpredictability in the art so long as there was a reasonable probability of success.” Pfizer, Inc.
v. Apotex, Inc., 480 F.3d 1348, 1364 (Fed. Cir. 2007). For this reason, Dr. Drover’s claims that:
1) reformulating zolpidem for sublingual delivery is not obvious because the change in
bioavailability is “not easy to predict;” (Tr. 9.137:3-12 (Drover)); and 2) the formulation is not
obvious because transmucosal products are “not always successful,” (Tr. 9.78:20-25 (Drover), are
properly rejected by the Court in light of the aforementioned evidence which the Court finds clear
and convincing. (Tr. 9.78:20-25 (Drover). 14 Consequently, the Court holds that a POSA would
have a reasonable expectation of success in formulating zolpidem for transmucosal delivery.
14
“Obviousness does not require absolute predictability of success,” but rather, requires “a
reasonable expectation of success.” See Medichem, S.A. v. Rolabo, S.L., 437 F.3d 1157, 1165
(Fed.Cir.2006) (quoting In re O'Farrell, 853 F.2d 894, 903–04 (Fed.Cir.1988)).
77
b. A POSA would have Anticipated a Reasonable Expectation of Success with Low Doses of
Zolpidem Based Upon Predictable Calculations.
As an initial premise, the Court finds that in light of Danjou and Hindmarch, a POSA
seeking to treat MOTN insomnia non-prophylactically with zolpidem, would have been motivated
to lower the 10 mg dose to eliminate the residual sedative effects. (See e.g. (Tr. 7.198:1-3
(Winkelman) Hindmarch shows that if you want to give a dose in the middle of the night, “10 [mg]
is too much…you should lower the dose.” (Tr. 7.198:1-3 (Winkelman)). Further, as Merlotti’s 5
mg dose of zolpidem (or a half dose of Ambien®) was shown to be effective, a POSA would begin
their hunt for the lowest effective dose with 5 mg. (See also Tr. 7.146:14-15: Referring to
Ambien® doses, Dr. Winkelman explained “full night, 10 milligrams. Half night, 5 milligrams. I
mean, it is very simple math.”). Plaintiffs however, are correct in their contention that there is no
single prior art reference before the Court teaching a POSA that 3.5 mg and 1.75 mg doses are
effective. In this context, the Court must therefore determine if these lower doses were obvious to
a POSA. The Court finds that they were.
i. Lowering the Dose from 5.0mg to 3.5mg and 2.5mg to 1.75mg
In conjunction with the prior art references Merlotti and Patat, Defendants cite to a POSA’s
knowledge of dose optimization of zolpidem tartrate based on zolpidem’s known physical and
chemical properties, including linear pharmacokinetics and available dose information. (See e.g.
Tr. at 8.23:5-8.25:10, 8.26:4-7 8.16:12-8.17:1, 8.25:15-8.26:10). In sum, a transmucosal dose of
3.5 mg of zolpidem is appropriately compared to a 5 mg oral dose of zolpidem because the
bioavailability of zolpidem is 70%, and 70% of 5 mg is equal to 3.5 mg. (Tr. 6.112:15-6.113:4
(Michniak-Kohn)). Dr. Michniak-Kohn explained that from Patat, you would use the 70%
78
bioavailability figure of zolpidem and multiply it by the 5 mg dose (published in Merlotti) and it
would yield 3.5 mg of zolpidem for non-elderly and then 70% multiplied by half of the 5 mg dose
(2.5 mg) would yield 1.75 mg for elderly. (Tr. 6.129:7-21, 6.130:16-21 (Michniak-Kohn)).
Thereafter, she concluded, and this Court agrees, that a POSA would have a “reasonable
expectation that th[ese] [calculations] would succeed” in translating the known low dose from oral
to transmucosal while maintaining efficacy. (Tr. 6.132:3-8 (Michniak-Kohn)). Indeed, even the
inventor of the patents-in-suit referred to this as “a simple calculation.” (Tr. 5.73:20-5.74:2 (Singh)
(emphasis added)). 15
Plaintiffs, nevertheless, argue that these calculations are misplaced as the bioavailability
for sublingual zolpidem is actually around 75%, not 100% as calculated by Dr. Michniak-Kohn
and Dr. Winkelman. The Court’s conclusion is unfettered by this argument. The proper inquiry
before the Court is not whether the calculations are correct, but rather whether they would have
been obvious to apply and therefore yield the lowest doses. Dr. Winkelman explained that
sublingual doses put the drug on the blood vessels and it gets absorbed right into the bloodstream
so “[i]t is not the same as intravenous but it is quite close.” (Tr. 7.204:9-12, 7.208:14-17
(Winkelman)). Knowing that intravenous would yield 100% bioavailability, it is credible and
reasonable to find that it would be obvious to a POSA to use 100% to yield a predicted dose for
sublingual form. Therefore, the Court is convinced that the lower doses of zolpidem are obvious.
15
The Court notes that in this context, Dr. Singh’s statement is taken into consideration to establish
that the calculations were well within the knowledge of a POSA. The Court does not refer to Dr.
Singh’s statement without the proper anti-hindsight perspective.
79
ii. Reasonable Expectation of Success
Plaintiffs argue that a POSA would assume low doses of zolpidem would not be successful
in effectively putting one back to sleep in the middle of the night. Plaintiffs point the Court to
Teitelbaum and the Borbély model for this premise claiming a POSA would expect that a subject
would need more, not less zolpidem in the middle of the night because their sleep drive is less.
Plaintiffs’ expert, Dr. Czeisler concludes that a POSA “would have expected to use certainly at
least the same dose [of 10 mg or 5 mg zolpidem], if not a higher dose [of zolpidem], in the middle
of the night because of the decrease in homeostatic sleep drive and the muting of the circadian
system.” (Tr. 10.83:10-23 (Czeisler)). Teitelbaum seems to suggest something similar in stating:
“the normal dosage is one-half to one 10 milligram tablet, taken at bedtime. If you wake up in the
middle of the night you can take an extra one-half to one tablet (leave it by your bedside with a
glass of water) and any sedation is usually worn off by the time you are ready to wake up in the
morning.” (PTX 033). However, Teitelbaum concludes this suggestion with, “one-half tablet is
usually enough for the middle of the night.” (Id.). This runs contrary to Dr. Czeisler’s suggestion
that one would need higher or equal to a 10 mg dose in the MOTN and is plainly stated in the prior
art.
Finally, the Kim reference tested zolpidem in the morning, when one’s sleep drive would
be at its very least and concluded that 5 mg zolpidem was able to get patients to sleep statistically
significantly shorter than placebo, therefore demonstrating its efficacy regardless of sleep drive.
The Court therefore is persuaded by Dr. Winkelman’s “common sense” suggestion that if a subject
“wanted to sleep for half a night, four hours, [they] would take half of the dose that [they] would
for a full night.” (Tr. 7.146:10-13 (Winkelman)). In sum, there is, before this Court, clear and
convincing evidence that there was a reasonable expectation of success in treating MOTN
80
insomnia by taking less amounts of zolpidem in the middle of the night.
c. Doghramji, Danjou, and Hindmarch Do Not Teach Away from Zolpidem at the Doses
Claimed.
Where, as here, the claim limitations are found in a combination of prior art references,
this Court, as the factfinder, must determine “[w]hat the prior art teaches, whether it teaches away
from the claimed invention, and whether it motivates a combination of teachings from different
references.” In re Fulton, 391 F.3d 1195, 1199–1200 (Fed.Cir.2004). While the Court is cognizant
that as a “useful general rule,” references that teach away cannot serve to create a prima facie case
of obviousness, a POSA seeking to treat MOTN insomnia by a better means than Ambien® would
inevitably use non-prophylactic dosing. In re Gurley, 27 F.3d 551, 553, 31 USPQ2d 1130
(Fed.Cir.1994). This is because MOTN insomnia as a “condition wherein a subject, after falling
asleep, awakens and has difficulty returning to sleep,” is so entangled with treatment “as-needed”
as it is impossible to know, until the middle of the night, whether the insomnia will occur. (ECF
No. 92 at 2). This principle, in and of itself, is convincing to the Court that non-prophylactic
administration was obvious. Nevertheless, the Court also concludes that the relevant prior art
references do not teach away from zolpidem.
The Court agrees with Defendants that Doghramji, Danjou, and Hindmarch all promote
the benefits of taking a hypnotic agent on an “as needed” basis. The studies in Danjou and
Hindmarch each dosed a subject in the MOTN, representative of non-prophylactic administration
at bedtime. Specifically, Danjou is directed to “nocturnal administration” and Hindmarch to
“middle of the night administration.” (JTX 015 at 367, PTX 256 at 159). Hindmarch also
references “[p]atients having sleep maintenance problems or difficulties falling asleep, especially
after being awakened during the night.” (PTX 256 at 160). Lastly, Doghramji explains that
81
hypnotic agents are typically given “prophylactically prior to going to bed,” but now the
“availability of a new hypnotic agent with a short half-life” (although, referencing zaleplon)
suggests a patient may be able to take the agent during the night. (JTX 016 at JNTDEF0000171172). However, as Plaintiffs correctly point out, the ultimate conclusion of each of these references
was that zaleplon was better suited for MOTN administration due to the lingering residual sedative
effects produced with zolpidem.
A reference “teaches away” when it “suggests that the line of development flowing from
the reference's disclosure is unlikely to be productive of the result sought by the applicant.”
Medichem, S.A. v. Rolabo, S.L., 437 F.3d 1157, 1165 (Fed.Cir.2006) (quoting In re Gurley, 27
F.3d 551, 553 (Fed.Cir.1994)). “A reference may be said to teach away when a person of ordinary
skill, upon reading the reference, would be discouraged from following the path set out in the
reference, or would be led in a direction divergent from the path that was taken by the applicant.”
In re Gurley, 27 F.3d 551, 553 (Fed.Cir.1994). Whether a prior art reference teaches away from
the claimed invention is a question of fact. Para–Ordnance Mfg., Inc. v. SGS Imps. Int'l, Inc., 73
F.3d 1085, 1088 (Fed.Cir.1995).
The Court however, concludes that a POSA would not be deterred from these findings.
The key basis for this conclusion is that Danjou, Doghramji, and Hindmarch, each tested the
higher 10 mg dose of zolpidem for MOTN administration and their findings and subsequent
recommendations were based solely on residual sedative effects. Even Plaintiffs’ expert admits
that a POSA would know that a lower dose of zolpidem would decrease the time that hypnotic
effects would occur. (Tr. 10.54:19-24 (Czeisler)). Certainly, a POSA would not disregard the
touted benefits of non-prophylactic dosing to achieve their ultimate goal simply because the final
conclusion of a reference chooses zaleplon at a dose more than double that of the claimed
82
invention. (See also Tr. 6.108:2-9 (Michniak-Kohn) Doghramji leads a formulator to “decrease
the dose” as they knew they wanted the drug to wear off in four hours.).
Additionally, both
Parties’ expert agree there are specific disadvantages to zaleplon, and Plaintiffs’ experts have
failed to offer persuasive evidence that a POSA would consider zaleplon over zolpidem given
Ambien®’s success and known efficacy. (See Tr. at 1.161:21-1.162:9 (Kryger), Tr. 7.156:18157:6 (Winkelman)). Viewed against the backdrop of the totality of collective teachings of the
prior art and the common knowledge of a POSA that reducing the dose would reduce residual
sedative effects, Danjou, Doghramji, and Hindmarch do not teach away from zolpidem in a
manner that would deter a person of ordinary from combining these references with the low doses
articulated in Merlotti.
d. Buffer Claims of the ’628 and ’809 Patents
The Court refers to the “Buffer Claims,” in its obviousness analysis with reference to: 1)
Claim 1 of the ’628 patent’s limitation “wherein the buffer raises the pH of saliva to a pH of about
7.8 or greater;” 2) Claim 9 of the ’628 patent identifying “the buffer comprises a carbonate buffer
and bicarbonate buffer;” and 3) Claim 22 of the ’809 patent requiring the composition to contain
a “binary buffer system.” Plaintiffs assert that these Claims are not obvious for interrelated
reasons. Plaintiffs oppose a finding that Claim 1 of the ’628 patent is obvious because of the
requirement that the pH be elevated to above 7.8, which they maintain, was not established by
clear and convincing evidence.
With reference to Claim 9, Plaintiffs argue that while Pinney
identifies 13 individual buffering agents, it fails to disclose or render obvious zolpidem combined
with a “carbonate buffer and bicarbonate buffer,” as required by the Claim.
Similarly, pursuant to Claim 22 of the ’809 patent, the composition must contain a “binary
83
buffer system that raises the pH of said subject’s saliva to a pH greater than about 8.5, irrespective
of the starting pH of saliva.” The Court construed a “binary buffer system,” in Claim 22 of the
’809 patent to mean “a system used to maintain and/or achieve an approximate pH range
comprising at least one proton-donating component and at least one proton accepting component.”
(Opinion, ECF No. 185 at 26). Plaintiffs purport that is was not obvious to “raise[] the pH of said
subject’s saliva to a pH greater than about 8.5.” The Court does not agree with Plaintiffs for the
following reasons.
Pinney expressly teaches a specific pH range in the mouth—between 7 and 10—for
efficient oral administration through the oral mucosa. (Tr. 6.139:21-6.140:15 (Michniak-Kohn)).
Thus, the pH values of 7.8 and 8.5 are well within Pinney’s claimed ranges. The Federal Circuit
has made clear that if the relevant comparison between disputed claim limitations and the prior art
pertains to a range of overlapping values, “we and our predecessor court have consistently held
that even a slight overlap in range establishes a prima facie case of obviousness.” In re Peterson,
315 F.3d 1325, 1329 (Fed.Cir.2003). This Court follows suit.
Additionally, evidence before the Court established that the carbonate/bicarbonate buffer
in Claim 9 of the ’628 patent (or binary buffer for purpose of the ’809 patent) was a well-known
buffer as of 2003 and how to make a buffer to raise the oral cavity to a desired pH when
administering a transmucosal drug was well within the knowledge of a POSA would know. (Tr.
6.144:6-21 (Michniak-Kohn)). Beckett for example, teaches a binary buffer used to raise the pH
(while Beckett uses phosphate buffer, bicarbonates are very common in use). (Tr. 6.145:14-24
(Michniak-Kohn)). Pinney also teaches a formulator to use a pH-raising agent in the oral mucosa
for transmucosal delivery. (Tr. 6.133:1-11 (Michniak-Kohn)).
Even the inventor, Dr. Singh,
cooreberates that Dr. Michniak-Kohn’s understanding was within the knowledge of a POSA. He
84
explains that changing the pH by way of a buffer for transmucosal delivery is “basic chemistry.”
(Tr. 5.64:10-18 (Singh)). Moreover, the Court declines to ignore the Pinney disclosure of this
specific buffer among just 13 options. While the Federal Circuit has predicated a finding of
nonobviousness on a sheer number of variable combinations, it did so in the face of a prior art
disclosure of a “potentially infinite genus.” In re Baird, 16 F.3d 380, 382 (Fed.Cir.1994) (quoting
In re Jones, 958 F.2d 347, 350 (Fed.Cir.1992)). The case at bar does not remotely approach an
infinite genus, as it is quantifiable in just 13. The Court therefore finds the Buffer Claims were
obvious in view of the above.
e. ’809 Patent: Outstanding Claim Elements of Claims 1, 12
The composition claimed in the ’809 patent must contain, among other things, an effective
amount of zolpidem “sufficient to produce a plasma concentration between about 25 ng/mL and
about 50 ng/mL within 20 minutes of administration when evaluated in an appropriate patient
population.” (JTX 002 at Claims 1 and 12). Plaintiffs argue that Dr. Michniak-Kohn’s assertion
that the 20-minute plasma concentrations are “inherent” in the doses (3.5 mg and 1.75 mg) is both
incorrect and unsupported by data. Plaintiffs scold Dr. Michniak-Kohn for failing to consider that
pharmacokinetic parameters will depend on the specifics of the formulation, not just the dose of
the drug. (Tr. 9.140:13-22 (Drover)). Remarkably, however, Dr. Drover subsequently admits that
he is not a formulation expert. (Tr. 9.140:20 (Drover)).
Regardless, a closer look at Dr. Michniak-Kohn’s testimony reveals that based upon the
linear pharmacokinetics, presented in a demonstrative graph in fact used by Dr. Drover, she opined
that a POSA could easily predict offset blood concentrations for sublingual doses because when
you “half the dose… that means half the plasma concentration at the same point.” (Tr. 6.111:15-
85
19 (Michniak-Kohn)). This is contrary to Plaintiffs’ blanket assertion that the only evidence before
the Court was inherency in the dose. (See also Tr. 6.112:1-7: Predictions of offset blood
concentrations can be done with zolpidem because a POSA the pharmacokinetics of zolpidem
known by 2003 were “linear enough.” (Michniak-Kohn)).
Dr. Michniak-Kohn further points to
plasma concentrations for Ambien®, the Merlotti reference and Patat reference, to explain that a
POSA would take into account the dose as well as the bioavailability data. (See e.g. Tr. 7:46:117 (Michniak-Kohn)). Specifically, Dr. Michniak-Kohn explains how a POSA would understand
that Patat reported the blood concentration of 20 nanograms for a male as the onset and offset
threshold for sedation of zolpidem (referring to the concentration at which zolpidem begins to start
sedating or stop sedating a patient). (Tr. 6.87:16-20 (Michniak-Kohn). Indeed, the plain text of
Patat states:
[R]eturn to baseline also occurred at concentrations ranging from 20
(5 mg PO or IV) to 75 ng.ml (20 mg PO). As the EEG effects are
very rapid, whatever the dose or the route of administration, it can
be suggested that the threshold concentration of zolpidem was
already attained.
(JTX 028 at JNTDEF0004150). Indeed, 20 nanograms, viewed in light of Patat as the lowest
concentration cited for onset/offset sedation—with a total range of 20-75 ng/ml—the claimed
range of between “about 25ng/ml and about 50ng/ml” naturally flows and is disclosed. See e.g.,
In re Woodruff, 919 F.2d 1575, 1578 (Fed. Cir. 1990) (concluding that a claimed invention was
rendered obvious by a prior art reference whose disclosed range (“about 1–5%” carbon monoxide)
abutted the claimed range (“more than 5% to about 25%” carbon monoxide)). This Court therefore
finds Dr. Michniak-Kohn’s reading of Patat—and determination that it renders this element of
Claim 1 obvious—to be both credible and convincing.
It is important to also bolster this finding with a case where the Federal Circuit has even
86
previously upheld such an “inherent property” on similar facts. In Santarus, Inc. v. Par Pharm.,
Inc., 694 F.3d 1344 (Fed. Cir. 2012) the court identified that “an obvious formulation cannot
become nonobvious simply by administering it to a patient and claiming the resulting serum
concentrations.” Id. at 1351; see also In re Kao 639 F.3d 1057, 1070 (Fed. Cir. 2011) (“To hold
otherwise would allow any formulation—no matter how obvious—to become patentable merely
by testing and claiming an inherent property.”). Distinctly though, neither party disputed that the
blood serum concentrations claimed in Santarus were expected in light of the dosages disclosed
in the prior art. Id. However, here, the Court finds that in light of the dosages delineated in the
’809 patent and Patat, the concentrations in the ’809 patent are disclosed, if not inherent, and
therefore obvious.
4. Motivation to Combine
As stated, MOTN insomnia was previously being treated the same way as other types of
insomnia, prophylactically. The uniqueness inherent in MOTN insomnia however, is that a person
will not know at bedtime—when prophylactic doing occurs—if they are going to experience it.
Such prophylactic dosing therefore leads to overmedication and drug dependence. (See e.g. Tr.
1.52:9-24 (Kryger)).
Thus, the problem in the context of the patents-in-suit, was to develop a
method and composition for treating MOTN insomnia exclusively while avoiding the downfalls
of prophylactic administration.
The Court finds that a POSA would have been motivated to combine Ambien® with the
prior art references to solve this problem. The record at trial clearly established that a skilled
person seeking to formulate a drug to treat MOTN insomnia had 4 well-known goals: 1) administer
the drug on an as-needed basis (upon MOTN wakening); 2) employ an active ingredient that was
known to deliver rapid onset of action to get one back to sleep as quickly as possible; 3) use the
87
lowest effective dose; and 4) avoid residual sedative effects upon awakening. (See e.g. Tr. at
8.11:6-22; 8.12:4-8.13:7 (Winkelman)). In order to achieve the aforementioned goals, a POSA, at
the time of the invention—armed with the knowledge of the prior art—would use low-dose
zolpidem (Merlotti and Ambien®) administered in the middle of the night (See e.g. Danjou and
Doghramji) in a formulation that is delivered across the oral mucosa (Pinney and Tauber). (Tr. at
8.25:13-8.26:12 (Winkelman)). This methodology suffices to establish a motivation to combine
as such motivation does not have to be explicitly stated in the prior art, and can be supported by
testimony of an expert witness regarding knowledge of a POSA at the time of invention. Alza
Corp. v. Mylan Labs., Inc., 464 F.3d 1286, 1294 (Fed. Cir. 2006).
However, Plaintiffs’ further claim that a POSA would not be motivated to combine low
doses of zolpidem with a transmucosal delivery system because zolpidem was already known to
have rapid action. Contrary to this assertion, the record indicates a number of benefits to sublingual
administration, particularly in the hypnotics context. To name a few, Tauber found a 40-50%
decrease in sleep latency and Pinney found transmucosal delivery would achieve the drug in the
bloodstream within minutes of application, rather than approximately 30 minutes with oral. (See
also Tr. 7.142:1 (Winkelman) After middle of the night awakening, the “clock is ticking” to fall
back asleep.). Viewed as a whole, a POSA would have been motivated to make a version of
Ambien® that could be used solely for MOTN insomnia. Organically, the combination of the
pertinent prior art references did just that.
5. Prima Facie Case of Obviousness
As articulated above, the Court finds that Defendants have presented clear and convincing
evidence that the asserted claims of each of the patents-in-suit are obvious and these patents as a
88
whole are obvious. Giving the elderly half the dose of the non-elderly was taught by Ambien®
and substantiated by the knowledge of a POSA. The claims relating to transmucosal delivery are
obvious in light of Pinney, Tauber, the Henderson-Hasselbach equation, and the known (and
widely published) properties of zolpidem. Thus, the relevant elements of Claims 10 and 19 of the
’131 patent, Claims 1, 11, and 12 of the ’809 patent, and Claim 1 of the ’628 patent are obvious.
The claim ranges and specific low doses of zolpidem are obvious in view of Merlotti, Patat and a
POSA’s knowledge of dose optimization of zolpidem tartrate, (including linear pharmacokinetics
and Dr. Michniak-Kohn’s and Dr. Winkelman’s calculations). Subsequently, these elements of
Claims 1, 8 and 18 of the ’131 patent, Claims 1, 18 and 17 of the ’809 patent, and Claims 16 and
17 of the ’628 patent are obvious.
The Court is cognizant that a claimed invention may be obvious even when the prior art
does not teach each claim limitation, so long as the record contains some reason that would cause
one of skill in the art to modify the prior art to obtain the claimed invention. Beckson Marine, Inc.
v. NFM, Inc., 292 F.3d 718, 728 (Fed. Cir. 2002). The record has convinced the Court of just that,
grounded in the clear goals of targeting the specific insomnia occurring only in the middle of the
night. Even so, the remaining elements and claims specific to each patent are also obvious. Nonprophylactic dosing in Claims 1 and 12 of the ’131 patent are obvious based upon Doghramji,
Danjou and Hindmarch, as well as the nature of MOTN insomnia. The Buffer Claims, in light of
Pinney and Beckett, render Claim 22 of the ’809 patent and Claims 1 and 9 of the ’628 patent,
invalid as obvious. Finally, the remaining element of Claim 1 of the ’809 patent, namely, plasma
concentration of 25 ng/ml to 50 ng/ml within 20 minutes, is invalid in light of Patat.
89
6. Secondary Considerations
With Defendants having met their burden to establish a prima facie case of obviousness,
the Court will go on to consider the fourth Graham factor: facts regarding objective indicia of
nonobviousness. It is well-settled that “all evidence relevant to obviousness or nonobviousness be
considered, and be considered collectively.” In re Cyclobenzaprine Hydrochloride, 676 F.3d at
1078. As they can give “light to the circumstances surrounding the origin of the subject matter
sought to be patented,” Graham, 383 U.S. at 17–18, 86 S.Ct. 684, objective considerations serve
as a check against hindsight bias and “ ‘may often be the most probative and cogent evidence in
the record.’ ” In re Cyclobenzaprine Hydrochloride, 676 F.3d at 1075–76, 1079 (quoting
Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530, 1538–39 (Fed.Cir.1983)). Indeed, these
considerations can have the force of “ ‘establish[ing] that an invention appearing to have been
obvious in light of the prior art was not.’ ” Id. at 1075–76 (quoting Stratoflex, 713 F.2d at 1538–
39). The Court will now consider each of the objective considerations raised by the Parties.
a. Licensing, Industry Aquiescence, Commercial Success
Plaintiffs rely on the license deal between Purdue and Transcept as well as other licensing
“offers” and “interest” to support a finding of nonobviousness. (PFOF ¶¶ 650-655). Primarily,
the Court notes that in accordance with In re GPAC Inc., 57 F.3d 1573 (Fed.Cir.1995), licenses
“may constitute evidence of nonobviousness; however, only little weight can be attributed to such
evidence if the patentee does not demonstrate a nexus between the merits of the invention and the
licenses of record.” Id. at 1580 (internal quotations and citations omitted). Plaintiffs have failed
to convince the Court of such nexus. Moreover, whatever little significance the licenses may have
is clearly outweighed by the strong evidence of obviousness found in the prior art. See Brown &
Williamson Tobacco Corp. v. Philip Morris Inc., 229 F.3d 1120, 1131 (Fed.Cir.2000).
90
When further viewed in conjunction with Intermezzo®’s lack of industry praise and lack
of commercial success, the Court is not inclined to give one licensing deal much weight. Indeed,
“’the mere existence of ... licenses is insufficient to overcome the conclusion of obviousness' when
there is a strong prima facie case of obviousness.” Iron Grip Barbell Co. v. USA Sports, Inc., 392
F.3d 1317, 1324 (Fed.Cir.2004) (citing SIBIA Neurosciences, Inc. v. Cadus Pharm. Corp., 225
F.3d 1349, 1358 (Fed.Cir.2000). Actual sales for Intermezzo® are under $10 million while the
market projections suggested sales of Intermezzo® in 2015 of $495 million. (Tr. 2 .113:2-3
(Oclassen), 9.15:10-23 (Kraft)). This striking disparity is significant as the Federal Circuit has
noted that commercial success “ ‘is usually shown by significant sales in a relevant market.’ ” J.T.
Eaton & Co. v. Atlantic Paste & Glue Co., 106 F.3d 1563, 1571 (Fed.Cir.1997). Needless to say,
Intermezzo® is far from reaching a significant sales mark. Plaintiffs have also failed to point to
credible evidence of industry praise, reassuring the Court that this factor, as a whole, weighs in
favor of obviousness.
b. Long-Felt Need and Failure of Others
“Long-felt need is closely related to the failure of others. Evidence is particularly probative
of obviousness when it demonstrates both that a demand existed for the patented invention, and
that others tried but failed to satisfy that demand.”
In re Cyclobenzaprine Hydrochloride
Extended-Release Capsule Patent Litig., 676 F.3d 1063, 1082 -1083 (Fed. Cir. 2012). Failure of
others “to find a solution to the problem which the patent[ ] in question purport[s] to solve” is
evidence of nonobviousness. Symbol Techs., Inc. v. Opticon, Inc., 935 F.2d 1569, 1578
(Fed.Cir.1991) (internal quotation marks and citation omitted). The problem the patents-in-suit
sought to solve is a targeted treatment for MOTN insomnia only, where treatment can be taken “as
needed.” At the outset, the Court notes that Plaintiffs have provided no evidence of failure of
91
others to solve this problem and therefore rely solely on their theory there was a long-felt and
unmet need for treating MOTN insomnia.
Prior to the filing of the patents-in-suit, Ambien® was used to treat MOTN insomnia
prophylactically. The Court admittedly observed a number of downsides to prophylactic treatment
at trial, including overmedication. (Tr. 1.52:9-24 (Kryger)). Furthermore, the prior art references
such as Doghramji (and Scharf), clearly articulated a need for flexibility to treat MOTN insomnia
on an “as-needed” basis. (See e.g. JTX 016). However, while Plaintiffs point the Court to these
references, Doghramji was published in 2000 and Scharf in 2001. This is just four years prior to
the filing of the patents-in-suit. Thus, the Court concludes that the intervening time between the
prior art’s teaching of the “as needed” treatment and the eventual preparation of a successful
composition, is hardly “long-felt.” (See Ecolochem, Inc. v. S. Cal. Edison Co., The length of the
intervening time between the publication dates of the prior art and the claimed invention can also
qualify as an objective indicator of nonobviousness. 227 F.3d 1361, 1376–77 (Fed.Cir.2000)).
c. Skepticism
“General skepticism of those in the art ... is also ‘relevant and persuasive’ evidence of
nonobviousness.” Monarch Knitting Mach. Corp. v. Sulzer Morat GmbH, 139 F.3d 877, 885
(Fed.Cir.1998) (internal quotations and citation omitted). This is so because “[p]roceeding
contrary to the accepted wisdom is ... strong evidence of unobviousness.” Ruiz v. A.B. Chance
Co., 234 F.3d 654, 668 (Fed.Cir.2000) (internal quotation marks and citation omitted). In support
of their skepticism argument, Plaintiffs direct the Court to: 1) the prior art which indicated
zolpidem for treatment of MOTN insomnia would likely lead to residual sedative effects; (See e.g.
Danjou JTX 015); and 2) the maker of Ambien®, Sanofi’s, licensing discussions with Trancept
expressing “very substantial skepticism” according to Mr. Oclassen, Trancept’s former CEO. (Tr.
92
2.69:23-2.70:1 (Oclassen)). The Court finds neither of these compelling.
As previously articulated, the prior art references indicating that zolpidem would produce
residual sedative effects when administered in the middle of the night were only directed at the
high doses of 10 mg (for non-elderly) and 5 mg (for elderly). No reference convinced the Court
that zolpidem at lower doses would result in residual sedative effects, or that a POSA would
believe so. Further, Mr. Oclassen’s statement of “substantial skepticism” is rebutted by the record.
(See e.g. Tr. at 2.219:1-14 (Garegnani) Explaining that the technology associated with making
Intermezzo® “seemed very straightforward and kind of [] easy...;” Tr. 2.75:17-20 (Oclassen). For
instance, it was known prior to 2004 that avoiding the first-pass effect by going from an oral
swallow administration to transmucosal administration would increase bioavailability and allow
for lower doses. (See Pinney). The Court is therefore unconvinced that the literature or testimony
predating the filing of the patents-in-suit should be credited for a finding of skepticism.
7. Conclusion of Obviousness
For the reasons set forth above, the Court concludes that Defendants have made a prima
facie showing that the asserted claims of the patents-in-suit would have been obvious in view of
the prior art, the clear motivation to combine the references, and a reasonable expectation of
success in doing so. The Court also finds that the Plaintiffs’ evidence of secondary considerations
is inadequate to raise any doubt as to the obviousness of these claims. The objective indicia
presented really lent more evidence towards obviousness and thus most certainly did not carry
sufficient weight to override a determination of obviousness based on primary considerations.
Each patent-in-suit, when viewed as a whole, it therefore invalid.
93
B. Anticipation
Defendants argue that Claim 9 of the ’628 patent should be invalidated as anticipated by
the prior art reference Pinney. Claim 9 of the ’628 patent recites as follows: “The method of claim
1, wherein the buffer comprises a carbonate buffer and a bicarbonate buffer.” (JTX 003, Claim 9).
Claim 9 is also dependent on independent Claim 1 which states the following:
Claim 1: A method for treating insomnia, comprising the
steps of: administering a solid pharmaceutical composition
comprising zolpidem or a pharmaceutically acceptable salt
thereof to a subject prone to insomnia, the pharmaceutical
composition further comprising a buffer, wherein the buffer
raises the pH of saliva to a pH of about 7.8 or greater,
wherein zolpidem is absorbed across a permeable membrane
of the subject’s oral mucosa, and wherein at least 75% of the
solid pharmaceutical composition dissolves within 10
minutes or less within an oral cavity following
administration.
(JTX 003, Claim 1). Pinney was published in 2001 and is undisputedly prior art to the patentsin-suit. Pinney is titled “Chewing gums, lozenges, candies, tablets, liquids, and sprays for
efficient delivery of medications and dietary supplements.” (DTX 062).
1. Legal Standard
Pursuant to 35 U.S.C. §§ 102, a claimed invention is “anticipated,” and is therefore not
novel if it “was known or used by others in this country, or patented or described in a printed
publication in this or a foreign country, before the invention thereof by the applicant” or “was
patented or described in a printed publication in this or a foreign country or in public use or on
sale in this country, more than one year prior to the date of the application for patent in the United
States.” 35 U.S.C. §§ 102(a)-(b). “A patent is invalid for anticipation if a single prior art reference
94
discloses each and every limitation of the claimed invention,” and “a prior art reference may
anticipate without disclosing a feature of the claimed invention if that missing characteristic is
necessarily present, or inherent, in the single anticipating reference.” Schering Corp. v. Geneva
Pharm., 339 F.3d 1373, 1377 (Fed.Cir.2003) (internal citation omitted). In order to demonstrate
anticipation, the proponent must show “that the four corners of a single, prior art document
describe every element of the claimed invention.” Xerox Corp. v. 3Com Corp.,, 458 F.3d 1310
1322 (Fed. Cir. 2006)(quoting Advanced Display Sys., Inc. v. Kent State Univ., 212 F.3d 1272,
1282 (Fed. Cir. 2000)). Anticipation is a question of fact, Sanofi–Synthelabo v. Apotex, Inc., 550
F.3d 1075, 1082 (Fed.Cir.2008) (internal citation omitted), that must be established at trial by clear
and convincing evidence. Purdue Pharma L.P. v. Boehringer Ingelheim GmbH, 237 F.3d 1359,
1365 (Fed. Cir. 2001).
2. Pinney Fails to Anticipate Each and Every Limitation of Independent Claim 1.
Defendants argue that Pinney enables the method of treatment claimed in the ’628 patent.
This is significant because a claimed invention cannot be anticipated by a prior art reference if the
allegedly anticipatory disclosures cited as prior art are not enabled. Amgen Inc. v. Hoechst Marion
Roussel, Inc., 314 F.3d 1313, 1354 (Fed. Cir. 2003). However, Defendants argument that the
Pinney reference is anticipatory ignores a few key differences between the ’628 patent and Pinney.
Therefore, the Court concludes that Plaintiffs have demonstrated that the asserted claims survive
the validity challenge posed by the Pinney reference.
While Plaintiffs concede that some of the elements of the ’628 patent are in fact disclosed
and therefore anticipated by Pinney, Plaintiffs persuasively argue that two elements are missing.
First, Plaintiffs claim Pinney fails to indicate that that transmucosal absorption of zolpidem is even
possible, let alone disclosing how to accomplish this function. Next, Plaintiffs contend that Pinney
95
fails to disclose any relation to “insomnia” and therefore does not disclose “methods of treating
insomnia” in a subject “prone to insomnia,” but rather, one would need to supplement Pinney with
another reference, such as the Ambien® label, to arrive at the claimed invention. Because
anticipation requires a more stringent finding than obviousness—by limiting the inquiry to one
prior art reference—the Court agrees with Plaintiffs on these points in the limited context of
anticipation. 16
Plaintiffs argue that because the pharmacokinetics of transmucosal formulations are
unpredictable to a POSA viewing Pinney, and therefore it fails anticipate Claim 1 of the ’628
patent as it does not teach how zolpidem can be formulated for transmucosal absorption. This
Court agrees. It cannot be disputed that zolpidem fails to appear as the forefront of Pinney.
Zolpidem is merely mentioned in a long list of potential active ingredients for a composition.
(DTX 062 at 17). Thus, while true that Pinney mentions zolpidem, it does not present any findings
or guidance on how or if zolpidem can be absorbed transmucosally.
Dr. Drover explained that bioavailability is key to proper development of this type of
transmucosal formulation, although he claims an increase in bioavailability will not always be
achieved when switching from an oral swallow tablet to a transmucosal formulation. (Tr. 9.102:5–
9.104:6, 9.106:16–9.110:3, 9.122:13–9.123:17 (Drover)). Dr. Michniak-Kohn agrees in part,
stating Pinney tells a formulator to take into account the drug’s bioavailability for transmucosal
delivery. (Tr. 6.136:17-22 (Michniak-Kohn)). However, the only known bioavailability of
16
See e.g. In re Fracalossi, 681 F.2d 792, 794 (CCPA 1982); Although a claimed invention can
be obvious but not anticipated, it “cannot have been anticipated and not have been obvious.”
(emphasis added).
96
zolpidem comes from another reference, Patat, which is impermissible for a finding of
anticipation.
Similarly, Plaintiffs’ formulation expert Dr. Polli explained that a POSA would not
understand Pinney to teach that each of the 160 listed active ingredients can be delivered
transmucosally, as acetaminophen, a medicant listed in Pinney, is therapeutically effective at a
dose of hundreds of milligrams, far too large a dose to be considered a candidate for oral
transmucosal delivery. (DTX 062 at JNTDEF0004166; Tr. 9.193:22–9.194:2 (Polli)). The Court
is constrained to agree based on the legal standard for anticipation. This is because Defendants’
expert, Dr. Michniak-Kohn admits the need for additional information outside of Pinney to
conclude zolpidem can be formulated for transmucosal delivery as Pinney provides “no data about
zolpidem.” (Tr. 9.198:19 (Polli); See Tr. 6.155:14-6.158:4 (Michniak-Kohn: Properties of
zolpidem that would indicate to a POSA that it can be delivered transmucosally include: 1) logP;
2) p/k/a; 3) solubility; and molecular weight.).
In the same vein, Pinney cautions that “many active ingredients display chemical properties
that prevent transmucosal absorption,” yet, as Dr. Michniak-Kohn agreed, Pinney does not identify
which of the actives display such chemical properties. (DTX 062 at JNTDEF0004155). The only
formulation specifically disclosed in Pinney is a chewing gum for delivery of nicotine. (DTX 62
at JNTDEF0004167–69; Polli Tr. 9.195:18–20.)
The Federal Circuit explained in In re
Omeprazole Patent Litigation, 483 F.3d 1364 (Fed.Cir.2007) that “anticipation by inherent
disclosure is appropriate only when the reference discloses prior art that must necessarily include
the unstated limitation, [or the reference] cannot inherently anticipate the claims.” 483 F.3d 1364,
1378 (Fed.Cir.2007). This Court therefore agrees with Plaintiffs that transmucosal absorption of
97
zolpidem is not inherent in Pinney for Defendants have failed to convince the Court that this
limitation would necessarily be recognized.
Pinney is also devoid of explicit instructions to treat insomnia. Indeed, during her analysis
of obviousness, Dr. Michniak-Kohn admits that when designing the dosage for a formulation of a
sedative hypnotic, the general method she would use would begin with reading up on the
indication. (Tr. 6.81:6-22 (Michniak-Kohn)). Thus, a formulator would first need to be directed
to insomnia literature before Pinney, as Pinney fails to give a POSA the indication of the ’628
patent. With an established need to consult sources other than Pinney to find “each and every
element as set forth in the claim[s],” no finding of anticipation shall issue. Verdegaal Bros., Inc.
v. Union Oil Co., 814 F.2d 628, 631 (Fed.Cir.1987).
In sum, Pinney does not anticipate the ’628 patent because the prior art method did not
dictate that zolpidem could be absorbed transmucosally nor the indications of insomnia. Here,
Plaintiffs are claiming a method that consists of a new way of using a previously known process
of delivery. While Pinney discloses the transmucosal delivery process, it fails to clearly indicate
this process to treat insomnia or for delivery of zolpidem. The ’628 patent required a POSA to
exercise a combining of other prior art references to formulate zolpidem to absorption
transmucosally and thus, for anticipation, “will not be denied the merit of patentability.” Quoting
Ansonia Brass & Copper Co. v. Elec. Supply Co., 144 U.S. 11, 18, 12 S.Ct. 601, 36 L.Ed. 327
(1892).
C. Indefiniteness
Defendants claim “without residual sedative effects” is an indefinite claim term, therefore
rendering the ’131 patent invalid. As previously indicated, each of the claims of the ’131 patent
98
asserted by Plaintiffs contain the limitation “without residual sedative effects,” construed by this
Court to mean “with no or minimal subjective feelings of sedation, as evaluated by: (a) testing
acceptably in at least one test exploring psychomotor performance, attention, information
processing, and memory used by those of skill in the art (hereinafter “Part A”); and/or (b)
demonstrating plasma levels of zolpidem, at an appropriate time point, below about 20 ng/ml,”
(hereinafter “Part B”). (Opinion, ECF No. 185 at 5-7). The ’131 patent lists the following
psychomotor performance, attention, information processing, and memory tests (Part A tests):
a Sleep Latency Test (SLT), a Visual Analog Test (VAT), a Digit
Symbol Substitution Test (DSST), a Symbol Copying Test (SCT), a
Critical Flicker Fusion threshold test (CFF), a Simple Reaction time
test (visual or auditory; SRT), a Choice Reaction Time test (CRT),
a Word Learning Test (WLT), a Critical Tracking Test (CTT), a
Divided Attention Test (DAT), a digit or letter cancellation test,
sleep staging through polysomnographic (PSG) measurements,
Continuous Performance Task test (CPT), Multiple Sleep Latency
Test (MSLT), a Rapid Visual Information Processing test (RVIP), a
mental calculation test, a body sway test, a driving performance test,
and others.
(JTX 3 at 6:45-60). According to Defendants, at the zolpidem doses claimed in the ’131 patent,
the presence or absence of infringement will depend on which of the various psychomotor
performance, attention, information processing, and memory tests are administered. Defendants
therefore conclude that because these tests will prove “outcome-determinative” of the infringement
inquiry, the claim term is invalid as indefinite. This Court is not convinced of same.
1. Legal Standard
35 U.S.C. § 112, ¶ 2 requires that the specification of a patent “conclude with one or more
claims particularly pointing out and distinctly claiming the subject matter which the applicant
regards as his invention.” The U.S. Supreme Court has held that courts should hold a claim to be
99
indefinite and therefore, invalid, “if its claims, read in light of the specification delineating the
patent, and the prosecution history, fail to inform, with reasonable certainty, those skilled in the
art about the scope of the invention.” Nautilus, Inc. v. Biosig Instruments, No. 13-369, 2014 U.S.
LEXIS 3818, at *6 (June 2, 2014).
Because claims delineate the patentee's right to exclude, the patent statute requires that the
scope of the claims be sufficiently definite to inform the public of the bounds of the protected
invention, i.e., what subject matter is covered by the exclusive rights of the patent. Halliburton
Energy Servs., Inc. v. M-I LLC, 514 F.3d 1244, 1249 (Fed. Cir. 2008). Otherwise, competitors
cannot avoid infringement, defeating the public notice function of patent claims.
Athletic
Alternatives, Inc. v. Prince Mfg., Inc., 73 F.3d 1573, 1581 (Fed.Cir.1996) (“[T]he primary purpose
of the requirement is ‘to guard against unreasonable advantages to the patentee and disadvantages
to others arising from uncertainty as to their [respective] rights.’ ”) (quoting Gen. Elec. Co. v.
Wabash Appliance Corp., 304 U.S. 364, (1938)). In other words,
[a] patent holder should know what he owns, and the public should
know what he does not. For this reason, the patent laws require
inventors to describe their work in “full, clear, concise, and exact
terms,” 35 U.S.C. § 112, as part of the delicate balance the law
attempts to maintain between inventors, who rely on the promise of
the law to bring the invention forth, and the public, which should be
encouraged to pursue innovations, creations, and new ideas beyond
the inventor's exclusive rights.
Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722, 731, 122 S.Ct. 1831, 152
L.Ed.2d 944 (2002).
The focus of indefiniteness rests on the meaning that claim terms would have to one of
ordinary skill in the art. Energizer Holdings, Inc. v. Int'l Trade Comm'n, 435 F.3d 1366, 1370
(Fed.Cir.2006). However, “[e]ven if a claim term's definition can be reduced to words, it is still
indefinite if a person of ordinary skill in the art cannot translate the definition into meaningfully
100
precise claim scope.” Halliburton Energy Servs., Inc. v. M–I LLC, 514 F.3d 1244, 1251
(Fed.Cir.2008). Claims that are “insolubly ambiguous” are indefinite and therefore invalid. Id. at
1250 (quoting Datamize, LLC v. Plumtree Software, Inc., 417 F.3d 1342, 1347 (Fed.Cir.2005)).
An issued patent is presumed valid and, therefore, invalidity must be proven by clear and
convincing evidence. 35 U.S.C. § 282; Metabolite Labs., Inc. v. Lab. Corp. of Am. Holdings, 370
F.3d 1354, 1365 (Fed.Cir.2004).
2. “Without Residual Sedative Effects” is Not Indefinite.
Defendants present the Court with four somewhat interwoven arguments for indefiniteness.
Each argument pertains to the potential for divergent results as to whether or not there are residual
sedative effects at 4 hours after dosing.
This conclusion, Defendants claim, is “outcome
determinative” of the infringement inquiry under Honeywell v. International Trade Commission,
341 F.3d 1332 (Fed.Cir.2003) (hereinafter “Honeywell”) and the claim term must be deemed
indefinite. Specifically, Defendants argue: 1) that the Part A and Part B results may conflict; 2)
there are “limitless” methods of testing for residual sedative effects under Part A; 3) the Vermeeren
driving study exemplifies a test where results conflicted depending on the statistical method
employed; and 4) the Danjou reference evidences divergent results between two Part A tests for
the same dose of zolpidem. After Defendants’ arguments and their factual underpinnings are
analyzed by the Court, a review of the holding in Honeywell and its distinguishing facts is
appropriate. Ultimately, the Court finds the evidence Defendants cite to for these propositions
does not meet the clear and convincing standard.
101
a. Part A vs. Part B
Defendants first argue that the same formulation can test acceptably in a psychomotor
performance test (claim construction Part A) and therefore infringe this limitation, while, at the
same time, produce a blood plasma concentration (showing the amount of drug in the blood) higher
than 20 ng/ml (claim construction Part B), therefore evidencing non-infringement. However, as
previously articulated by the Court in its summary judgment Opinion on the same topic, this
contradiction simply leads to a finding of infringement of the claim as the Court intentionally
construed the claim with use of the conjunctive term “and/or,” to allow for such. (emphasis added).
In line with this construction, the evidence produced at trial convinced the Court that 20
ng/ml is not a bright line test for residual sedative effects, but rather a safe harbor. Figures 1, 3,
and 4 of the ’131 patent itself depict the results of a DSST test and zolpidem blood levels following
administration of a 3.5 mg dose of zolpidem and show that at four hours the blood level
concentration of zolpidem was greater than 20 ng/mL but the results of the DSST test had returned
to normal. Dr. Kryger explained that a return to baseline levels of impairment is possible despite
an elevated blood level of zolpidem because the amount of impairment will depend on the level of
zolpidem in the brain, not in the blood, due to the “blood-brain barrier.” (emphasis added) (Tr.
1.227:3–1.229:22 (Kryger)). Indeed, Figure 4 of the patent shows that the change in DSST score
had returned to baseline at four hours even though zolpidem blood levels remained above 20
ng/mL, therefore demonstrating that impairment will disappear even though zolpidem may remain
in the blood. With this in mind, Defendants’ indefiniteness argument must be confined to only
Part A; psychomotor performance, attention, information processing, and memory tests.
102
b. Part A Does Not Delineate Limitless Methods
The ’131 patent lists many psychomotor performance, attention, information processing,
and memory tests and concludes by stating “and others.” Defendants take issue with this phrase,
claiming it establishes that there are “limitless” tests for infringement and thus, the term must be
rendered invalid as indefinite. (Defs.’ Proposed Findings of Fact (“DFOF”) ¶ 478). Defendants
cite to Dr. Winkelman’s testimony specifying other “Part A” tests such as the Stanford Sleepiness
Scale (as used in the Patat reference) and the Go/No Go Test. (Tr. at 7.97:1-17 (Winkelman); JTX
028 at 139). However, having a wide number of tests is not the standard to render a claim indefinite
as “[b]readth is not indefiniteness.” In re Gardner, 57 427 F.2d 786, 788 (1970). In any event,
Defendants have not offered evidence that either of these tests, the Stanford Sleepiness Scale nor
the Go/No Go Test, would produce divergent results from any of the other aforementioned tests.
For this reason, Defendants’ argument fails.
However, the crux of Defendants’ position for indefiniteness rests on the notion that
because “testing acceptably in at least one” of the Part A tests is sufficient to demonstrate that the
patient will awaken at four hours after dosing “without residual sedative effects,” a zolpidem
composition may still infringe this limitation, despite failing one or more of the Part A tests, so
long as the “possibility” exists of passing “at least one test” among the limitless set of Part A tests
for residual sedative effects. (DFOF ¶ 479).
In support, Defendants point the Court to the
Vermeeren driving study and the Danjou reference at trial, claiming that in each of these, the tests
exploring psychomotor performance, attention, information processing, and memory used by those
of skill in the art, produced divergent results. While the Court finds the evidence at trial failed to
demonstrate this proposition by clear and convincing evidence, the Court takes each of the
references in turn.
103
c. Vermeeren Driving Study
As previously explained, a study called Vermeeren was conducted on Intermezzo® which,
after MOTN administration, analyzed the driving performance of subjects to gage residual sedative
effects that occurred the morning after. (PTX 252). It is undisputed that this type of driving study
is one of the Part A tests used to measure residual sedative effects as set forth in the ’131 patent
itself. At trial, Plaintiffs presented the Vermeeren study to show that 4 hours after dosing, subjects
were free from residual sedative effects. On the other hand, Defendants contend that the driving
study in fact produced divergent results, evidencing the indefiniteness of the claim term “without
residual sedative effects.” More specifically, Defendants, through their expert Dr. Winkelman,
purported that Vermeeren demonstrates both the existence and the absence of residual sedative
effects, depending on the statistical measurement employed.
The two standards of measurement at issue are as follows: 1) determining whether the
standard deviation of lateral position (“SDLP”) (i.e. weaving) was statistically significantly
different from placebo (i.e. there were residual sedative effects); and 2) driving impairment based
on a McNemar symmetry analysis. Both Parties agree that the McNemar symmetry analysis
conducted in the driving study demonstrates a lack of residual sedative effects. This means that
the claim limitation “without residual sedative effects” is met, or, alternatively, would be infringed.
However, Defendants argue that the SDLP data of the same study does show residual sedative
effects, and therefore there are divergent results. Upon thorough review of Vermeeren as well as
the testimony of Plaintiffs’ relevant expert Dr. Kryger and Defendants’ relevant expert Dr.
Winkelman, the Court cannot agree with Defendants.
Vermeeren states the following:
Results showed that when ZST (Intermezzo) was taken 4 h before
driving, there was no statistically significant difference in the
104
proportions of impaired and improved drivers. The mean SDLP at
that time was significantly higher than PBO, but the overall increase
was small (0.83 cm), and the 95% CI was well below the 2.5 cm
threshold for impairment (95% CI, 0.1-1.15cm) … Overall, the data
support that driving at least 4 h after taking ZST 3.5 mg, consistent
with labeling instructions, does not negatively affect driving
performance.
(PTX 252 at 494). To rebut the clear finding of Vermeeren that there were no residual sedative
effects 4 hours after dosing, Defendants presented the testimony of Dr. Winkleman who explained
the McNemar symmetry analysis is not the accepted standard in the industry and the SDLP
measurement raw data showed statistically significant difference between Intermezzo and placebo
at 4 hours. (See e.g. Tr. 7.115:18-22 (Winkelman)). Dr. Winkelman therefore concluded that
Vermeeren showed Intermezzo® did not test acceptably in SDLP. (Tr. 7.109:1-7 (Winkelman)).
However, the conclusion of Vermeeren found the opposite. Vermeeren used the mean SDLP data
by applying a threshold of impairment of 2.5 cm (described as the standard in the art) and found
no impairment. According to Dr. Kryger, Vermeeren’s conclusions are correct because the raw
SDLP data is not determinative of clinically meaningful impairment in the patient population. The
Court finds no reason not to analogize “impairment” with “residual sedative effects” in the driving
study. All in all, Defendants’ proposed discrepancy is not an inherent measurability problem, but
rather a dispute between experts as to whether the measurements of Vermeeren were correctly
performed, which certainly does not amount to indefiniteness.
d. Danjou Reference
At trial, Defendants used the Danjou reference in an attempt to illustrate the point that
depending on the chosen Part A test, the same dosage amount will result in two divergent results.
To Defendants, these contrary results, if true, indicated the presence of residual sedative effects
105
(non-infringing) while also indicating an absence of residual sedative effects (infringing).
Specifically, in Danjou, 10 mg oral zolpidem “tested acceptably” in the DSST test four hours after
administration, but the same dosage amount did not “test acceptably” in the Critical Flicker Fusion
and the Choice Reaction Tests. (JTX 015 at Figs. 1, 2, 3). At the summary judgment stage, this
Court previously dismissed the use of Danjou as demonstrating outcome-determinative results
because Danjou did not test residual sedative effects for the low doses of zolpidem at issue (3.5
mg and 1.75 mg), but rather 10 mg dose. Because Defendants have failed to link the data in Danjou
to the low doses of the ’131 patent, the Court, again, rejects Defendants’ argument as unpersuasive.
The claim limitation “without residual sedative effects” of the ’131 patent is not invalid for
indefiniteness as Defendants have failed to meet their burden of proving invalidity, in this regard,
by clear and convincing evidence.
e. Honeywell v. International Trade Commission
The Court distinguishes Honeywell for purposes of completeness. Honeywell involved a
patent disclosing “a process for production of a particular multifilament polyester product called
polyethylene terephthalate (“PET”) yarn” used as a reinforcement for automobile tires. 341 F.3d
at 1334. All claims in the patent at issue in that case “require[d] that the yarn produced by the
claimed process fall within a specified . . . [melting point elevation] at some point during the
process.” Id. at 1335. The dispute in the case “focused on the method of measuring one claimed
feature—the melting point elevation (“MPE”).” Id. Although there were four methods for
preparing PET yarn that were well known to persons of ordinary skill in the art, “neither the claims,
the written description [of the patent at issue], nor the prosecution history reference[d] any of the
four sample preparation methods that can be used to measure the MPE.” Id. at 1339. In Honeywell,
106
the court noted that depending upon which method was used, “the calculated MPE for a given
sample can vary greatly.” Id. at 1336. With this in mind, the court held that the claims containing
the disputed term “melting point elevation” were “insolubly ambiguous, and hence indefinite”
because “the claims, the written description, and the prosecution history fail[ed] to give . . . any
guidance as to what one of ordinary skill in the art would interpret the claim to require.” Id. at
1340.
Contrary to the facts of this case, in Honeywell there was evidence that the method of
preparation and testing was critical to the measurement, and that only one of the four methods
produced a measurement within the claimed range; whereby the court concluded that the claims
were “insolubly ambiguous, and hence indefinite.” Id. at 1340. Here, the only credible and
pertinent evidence before the Court showed consistent results. The Court will not appease
Defendants and find indefiniteness based on a hypothetical possibility for inconsistent results.
Such is far from the clear and convincing standard. As the Federal Circuit has previously held,
“there is the potential for inconsistent results even within the same method of measurement, but
that surely does not render a claim indefinite.” Takeda Pharm. Co. v. Zydus Pharm. USA, Inc.,
743 F.3d 1359, 1367 (Fed. Cir. 2014). Finally, in Honeywell it was shown that persons in the field
of polymer chemistry understood that polymer melting point determinations vary significantly
with the method used, rendering the claims “insolubly ambiguous.” In contrast, it was not credibly
disputed that persons in the field of the '131 patent would fail to understand how to measure
residual sedative effects by the Part A tests. Honeywell is therefore distinguishable.
107
CONCLUSION
After a careful consideration of all the evidence presented at trial and for the reasons stated
above, the Court concludes that Defendants have met their burden of proving the ’131, ’809, and
’628 patents are invalid as obvious by clear and convincing evidence. The Court further finds that
Defendants have failed to prove that the claim element “without residual sedative effects” of the
’131 patent is invalid as indefinite. Defendants have also failed to prove the ’628 patent is invalid
as anticipated. The Court, however, finds that Plaintiffs have proved by a preponderance of the
evidence that the asserted claims of the ’131 patent are infringed by all Defendants, but also finds
that Plaintiffs have failed to meet their burden of proving infringement of the ’628 patent as to
Defendants DRL and Actavis only. Novel is found to infringe the ’628 patent. As to the ’809
patent, Plaintiffs have met their burden of proving infringement as to Defendant, DRL and
Defendant, Novel.
This Court’s Opinion will be filed under temporary seal. The Opinion will be unsealed on
Monday, April 20, 2015 unless an appropriate motion to seal same (pursuant to Local Civil Rule
5.3(c)) is filed by either of the Parties by April 17, 2015.
An appropriate Order accompanies this Opinion. Counsel are hereby directed to submit a
proposed form of judgment consistent with this Opinion.
s/ Jose L. Linares
Jose L. Linares
United States District Judge
Date: March 27, 2015
108
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?