`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`MYLAN PHARMACEUTICALS INC.
`Petitioner
`
`v.
`
`3M COMPANY et al.
`Patent Owner
`
`Case IPR2015-02002
`Patent 6,743,413
`
`DECLARATION OF RICHARD N. DALBY, PH. D.
`
`1
`
`2011
`3M COMPANY ____
`Mylan Pharmaceuticals Inc. v. 3M Company
`IPR2015-02002
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`I, Richard N. Dalby, declare as follows:
`
`I. Qualifications
`
`1)
`
`In 1983, I graduated with Honors with a B. Pharm. from Nottingham
`
`University School of Pharmacy.
`
`2)
`
`In 1988, I earned a Ph. D. in Pharmaceutical Sciences from the
`
`University of Kentucky, College of Pharmacy. My research mentor was Dr. Peter
`
`Byron, a leading research scientist in respiratory drug delivery systems, including
`
`MDIs. My research involved aerosol formulations.
`
`3)
`
`From October 1988 to August 1989, I worked as a Development
`
`Scientist for Fisons Pharmaceuticals (now Sanofi Aventis). In that role, I was
`
`responsible for development of new aerosol formulations based on using both
`
`chlorofluorocarbons (CFCs) and hydrofluoroalkanes (HFAs) as propellants.
`
`4)
`
`From December 1989 to August 1992, I was a Research Assistant
`
`Professor at Medical College of Virginia /Virginia Commonwealth University in
`
`the Department of Pharmacy and Pharmaceutics. My research was directed to
`
`looking for alternatives to CFCs as propellants in aerosol formulations.
`
`5)
`
`In 1992, I joined the Department of Pharmaceutical Sciences,
`
`University of Maryland as an Assistant Professor, becoming a full Professor in
`
`2003.
`
`2
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`6)
`
`Beginning in 1992, I have provided consulting services, workshops
`
`and seminars, primarily to pharmaceutical and biotechnology companies in the
`
`field of pharmaceutical aerosol formulations.
`
`7)
`
`I have authored over 200 papers, abstracts, or book chapters related to
`
`pharmaceutical aerosol formulations. In 2002, I was appointed Deans
`
`Distinguished Educator by the University of Maryland, School of Pharmacy. In
`
`2004, I was elected a Fellow of the American Association of Pharmaceutical
`
`Scientists.
`
`8)
`
`I have served as a consultant and as a member of the FDA Advisory
`
`Committee on Orally Inhaled and Nasal Drug Products, and as a Special Reviewer
`
`for the NIH on the Development of Novel Drug and Gene Delivery Systems and
`
`Devices.
`
`II. Person of Ordinary Skill
`
`9)
`
`In my opinion, a person of ordinary skill in the pharmaceutical aerosol
`
`formulation field, in or about December 1991, would have been someone with
`
`either: (a) at least a bachelor’s degree related to pharmaceutical sciences, or other
`
`related studies such as chemistry, chemical engineering, or colloidal science, along
`
`with corresponding 3-5 years of experience in the formulation and testing of
`
`pharmaceutical aerosol products and/or devices; or (b) an advanced degree
`
`(Masters or Ph.D.) in the same areas of academic study with about 1-2 years of
`
`3
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`corresponding experience in the field of pharmaceutical aerosol products. The
`
`same level of skill would apply in May 1992. My opinions expressed herein would
`
`be the same if the relevant date were May 1992.
`
`III. Background on Pharmaceutical Aerosol Formulations
`
`10) Many drugs, especially those for treating respiratory and nasal
`
`disorders, are administered through the mouth or nose. Johnson ‘123 patent,
`
`Exhibit 2001, col. 1, lines 11-12. Inhalation is the most widely used route for
`
`delivering bronchodilating drugs and steroids to the airways of asthmatic patents.
`
`Purewal ‘183 patent, Exhibit 2004, col. 1, lines 15-17.
`
`11) A metered-dose inhaler (MDI) is a handheld device designed to
`
`deliver a specific amount of medication in aerosol form. A MDI consists of a
`
`liquefied propellant-pressurized canister sealed with a metering valve that is
`
`attached to a plastic actuator incorporating a mouthpiece. During patient use the
`
`canister is oriented such that the value is pointed down. To use a MDI, the patient
`
`presses on the base of the canister to discharge a metered dose from the valve.
`
`This dose exits the mouthpiece and is inhaled into the lungs.
`
`12) As of December 1991, there were basically two types of MDI
`
`formulations: those in which the drug is dissolved (“solution formulations”) and
`
`those in which substantially all of the drug is suspended as fine particles
`
`4
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`(“suspension formulations”). Exhibit 2005 at 2. Suspension formulations were the
`
`predominant type of MDI formulation as of December 1991. Exhibit 2005 at 2.
`
`13) Suspension formulations containing only drug and propellant are
`
`physically unstable. Exhibit 1015 at 339. Aggregated, flocculated, or discrete
`
`drug particles can float to the surface of the propellant or sink to the bottom of the
`
`canister (assuming the canister to be in the “valve-up” orientation). Whether a
`
`particle sinks or floats depends on its density relative to the liquefied propellant.
`
`Generally, the larger the aggregate, the floccule, or the discrete particle the more
`
`quickly it will either float to the top or sink to the bottom. Exhibit 2005 at 26.
`
`Sufficient suspension stability is necessary to achieve a viable therapeutic
`
`formulation.
`
`14) As of December 1991, the most common propellants in MDIs were
`
`chlorofluorocarbons, known as CFCs. Exhibit 2003 at 1. At this time, CFCs were
`
`generally recognized to contribute to the depletion of the ozone layer, which led
`
`many countries to begin limiting, or even banning, their use. Exhibit 2003 at 1.
`
`15) The Montreal Protocol, an international treaty signed in 1987, called
`
`for a series of decreasing limits on CFC use and production. Because of the
`
`impending phase-out of CFCs, pharmaceutical formulators were seeking
`
`alternative, non-CFC propellants suitable for use in MDIs. One class of
`
`propellants considered as a replacement for CFCs were hydrofluoroalkanes or
`
`5
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`HFAs. Exhibit 2003. One molecule in that class was l,l,l,2-tetrafluoroethane,
`
`known as HFA 134a or Propellant 134a.
`
`16)
`
` The transition away from the use of CFCs in MDIs was not expected
`
`to be an easy one due to a need to demonstrate appropriate physical, chemical, and
`
`toxicological properties in the corresponding HFA-based MDIs. In 1994, CFC-
`
`based MDI formulations were granted essential use exemptions from the Montreal
`
`Protocol through 1997, recognizing that it would be a complicated task to replace
`
`CFCs as the propellant. Exhibit 2008, Sheila D’Souza, The Montreal Protocol and
`
`Essential Use Exemptions, J. Aerosol Med., 8 (Suppl 1): S13–S17 (1995).
`
`17)
`
`In December 1991, the goal of pharmaceutical formulators seeking to
`
`develop HFA-based suspension formulations was to re-create the appearance and
`
`performance of CFC suspension formulations known to be safe and effective for
`
`therapeutic use. The most straightforward approach they took was to substitute
`
`HFA 134a in place of a CFC propellant or blend of CFC propellants. Formulators
`
`soon recognized, however, that simply replacing the CFC or CFC blend with HFA
`
`134a would not re-create the appearance and performance of therapeutically
`
`effective CFC-based suspension formulations. Exhibit 2009, Richard N. Dalby et
`
`al., CFC Propellant Substitution: P-134a as a Potential Replacement for P-12 in
`
`MDIs, Pharm. Tech., March 1990 at 26. My 1990 article in Pharmaceutical
`
`Technology listed some of the significant differences in properties between HFA
`
`6
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`134a and CFCs that were expected to make the simple “drop-in” approach
`
`infeasible. Exhibit 2009 at 32-33.
`
`18) One important difference between CFCs and HFA 134a is that HFA
`
`134a is a poor solvent for the surfactants that had been used in CFC-based
`
`formulations up to December 1991. Exhibit 2009 at 32. Surfactants serve many
`
`purposes in suspension formulations, including preventing the aggregation of drug
`
`particles and lubricating the valve. For example, an Aerosol Age article I authored
`
`in 1990 stated that “adequate valve lubrication and reduction of inter-particulate
`
`aggregation must be provided by surfactants incorporated into the [suspension]
`
`formulation.” Exhibit 2005 at 89.
`
`19) As of December 1991, formulators did not view suspension
`
`formulations containing only HFA 134a propellant and drug as satisfactory for
`
`therapeutic use. Suspension formulations containing only HFA 134a and drug did
`
`not recreate the appearance and performance of CFC-based formulations that were
`
`known to be safe and effective for therapeutic use. Following shaking, the CFC-
`
`based formulations were substantially free of large aggregates and visually
`
`homogenous. The formulations containing only HFA 134a and drug were not.
`
`Johnson ‘123 patent, Exhibit 2001, col. 6, lines 1-40.
`
`20) Physical stability describes how long drug particles in a formulation
`
`remain suspended after shaking. Various degrees of physical stability can be seen
`
`7
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`in Figure 2 of my 1990 Aerosol Age article, Exhibit 2005, where I photographed
`
`the appearance of three CFC-based formulations at intervals over an hour. The top
`
`formulation, labelled “1” on the canister and “A” underneath the photos, started
`
`out as a homogenous suspension, but within 2 minutes the drug particles had
`
`flocculated and floated to the surface. Likewise, the formulation labelled “2” on
`
`the bottle and “B” underneath the photos also quickly flocculated and rose to the
`
`top of the canister. In contrast, the bottom formulation, labelled “4” on the canister
`
`and “C” underneath the photos, was more physically stable and maintained a
`
`homogenous suspension for over 5 minutes.
`
`21) As of December 1991, formulators believed that suspension
`
`formulations that flocculated quickly, resulting in drug inhomogeneity, were
`
`unsatisfactory for treating patients. Formulators believed that if a formulation
`
`flocculated quickly, then the MDI would not reliably deliver the target dose of
`
`medication to the patient. Suspensions in which the particles quickly floated to the
`
`top or sunk to the bottom displayed a non-uniform concentration of medication
`
`throughout the liquefied formulation. For example, if the flocculated drug particles
`
`had risen to the top of the liquefied formulation, then the drug concentration would
`
`be higher at the top then at the bottom. Exhibit 2005 at 28. The inverse would
`
`obviously be true if the drug particles had sunk to the bottom. Formulators
`
`believed that this presented an obstacle to achieving reproducible dosing, because
`
`8
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`the valve of the MDI extracts a specific volume of formulation from the bottom,
`
`and it is the product of the valve volume and the concentration of drug in the
`
`formulation (in the vicinity of the valve) which determines the nominally emitted
`
`dose.
`
`22) Prior to releasing a dose from the metering valve, a patient is required
`
`to shake a MDI to re-establish the homogeneity of drug particles within the
`
`suspension formulation after a period of non-use. In CFC-based suspension
`
`formulations that were known to be safe and effective for therapeutic use, the drug
`
`was readily re-dispersed by shaking, allowing the target dose of drug to be reliably
`
`sprayed into the lungs of the patient.
`
`23) As of December 1991, a formulator believed that adequate
`
`redispersibility only guaranteed accurate dosing if the suspension homogeneity
`
`were maintained for an appropriate period of time. This can be seen in Figure 3 of
`
`Exhibit 2005, which compares the percent of the target dose delivered by two
`
`formulations described in Figure 2: the unstable formulation “A” and the stable
`
`formulation “C”. The x-axis of Figure 3 is referred to as the “Shake – Fire
`
`Interval,” or the length of time between shaking and discharge of a dose. As
`
`Figure 3 shows, both the stable and the unstable formulations delivered 100% of
`
`the target dose immediately after shaking, because both showed adequate
`
`redispersibility. As more time elapsed after shaking, the unstable formulation “A”
`
`9
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`quickly delivered less than the target dose as the drug particles flocculated and
`
`floated to the surface. The stable formulation “C”, however, continued to deliver
`
`the full target dose, even 16 seconds after shaking, because the formulation existed
`
`as a uniform, homogenous suspension. For these reasons, formulators in
`
`December 1991 were trying to develop uniform, homogenous suspensions that
`
`were stable for an extended period of time following shaking. As shown by
`
`Exhibit 2005, suspension formulations that showed significant flocculation within
`
`a minute after shaking were not thought to display accurate and reproducible
`
`dosing.
`
`24)
`
` As of December 1991, formulators believed that suspension
`
`formulations containing only HFA 134a and drug did not re-create the appearance
`
`and performance of CFC-based formulations known to be safe and effective for
`
`therapeutic use, because formulations containing only HFA 134a and drug
`
`flocculated quickly. The Johnson ’123 patent, Exhibit 2001, reports that a control
`
`formulation containing only albuterol and HFA 134a displayed significant
`
`flocculation in less than 2 seconds. Exhibit 2001 at col. 6, lines 1-40 (Table 1).
`
`The Johnson ‘123 patent deems that level of performance unsatisfactory, saying
`
`that “[i]f significant flocculation occurs, that is, a cognizable coarseness and/or
`
`clear region can be observed, in less than about 15 sec., the suspension is deemed
`
`not stable enough for a practical aerosol inhalation drug formulation.” Exhibit
`
`10
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`2001 at col. 5, lines 49-53. Similarly, the Purewal ‘183 patent states that “use of
`
`[HFA] 134a and drug as a binary mixture … does not provide formulations having
`
`suitable properties for use with pressurised inhalers.” Exhibit 2004, col. 2, lines
`
`58-62. Both the Johnson ‘123 patent and the Purewal ‘183 patent reflect the belief
`
`at the time that a suspension formulation containing only HFA 134a and the drug
`
`would be unsatisfactory for therapeutic use.
`
`25)
`
`In December 1991, formulators believed that suspension formulations
`
`containing only HFA 134a and drug were unsatisfactory for therapeutic use. In
`
`response to this problem, formulators took many different approaches to try to re-
`
`create the appearance and performance of CFC-based formulations known to be
`
`safe and effective for therapeutic use. One approach was to try surfactants
`
`different than the surfactants conventionally used in CFC-based formulations. This
`
`is the approach taken by the Johnson ‘123 patent, which described how “[i]t has
`
`now been found that [HFA] 134a-soluble surfactants, especially soluble
`
`perfluorinated surfactants, effectively improve the stability of micronized
`
`inhalation drug suspensions in [HFA] 134a.” Exhibit 2001 at col. 2, lines 54-57.
`
`According to the Johnson ‘123 patent, using new surfactants that were soluble in
`
`HFA 134a made it possible to re-create the appearance and performance of CFC-
`
`based formulations known to be safe and effective for therapeutic use (Exhibit
`
`2001 at col. 3, lines 1-7 (emphasis added)):
`
`11
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`Thus it is now possible with the present invention to prepare
`
`aerosol formulations of inhalation drugs with [HFA] 134a which
`
`have sufficient stability for the purposes of this invention to deliver
`
`the active drug in the desired way as presently marketed MDI's,
`
`but without the environmental problems associated with CFC's.
`
`26) A second approach, taken in WO 1991/004011 (the ‘011 PCT
`
`application), was to pre-coat conventional surfactants that were insoluble in HFA
`
`134a onto the drug particles. Exhibit 1007 at page 3, lines 17-28.
`
`27) A third approach illustrated by the Purewal ‘183 patent, Exhibit 2004,
`
`added to HFA 134a a co-solvent (or “adjuvant”) that could dissolve existing
`
`surfactants. Exhibit 2004 at col. 3, lines 1-5 (“The addition of a compound of
`
`higher polarity than Propellant 134a to Propellant 134a provides a mixture in
`
`which increased amounts of surfactant may be dissolved compared to their
`
`solubility in Propellant 134a alone.”). As the Purewal ‘183 patent shows, adding a
`
`co-solvent or adjuvant “provides a propellant system which has comparable
`
`properties to those of propellant systems based on CFC’s.” Exhibit 2004 at col. 2,
`
`lines 26-29. Once again, the goal of this approach was to re-create the properties
`
`of CFC-based formulations known to be safe and effective for therapeutic use.
`
`28) A fourth approach was to try different blends of propellants, including
`
`blends of CFC and HFA 134a. I described such an approach in 1990, writing that
`
`“the ability of any proposed blend to dissolve a sufficient quantity of surfactant is
`
`12
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`of crucial importance.” Exhibit 2009 at 33. My goal in pursuing this approach
`
`was to try to develop a formulation that exhibited similar appearance and
`
`performance to purely CFC-based formulations. I considered it an advantage that,
`
`in pursuing the approach, “one sees strong similarities to conventional propellant
`
`blends” of CFCs. Exhibit 2009 at 33.
`
`29) The inventors of the ‘413 patent recognized that HFA 134a-based
`
`formulations need not necessarily re-create the appearance and performance of
`
`CFC-based formulations known to be satisfactory for therapeutic use. The
`
`inventors of the ‘413 patent recognized that satisfactory suspension formulations
`
`did not necessarily need to maintain stability for 15 seconds or longer, as
`
`formulators had previously believed. For example, in October 1990 I reported a
`
`study I performed where I tracked visual homogeneity over a full hour as a metric
`
`of formulation suitability. Exhibit 2005 at 23. Similarly, the Johnson ‘123 patent
`
`evaluated flocculation over 4 minutes as a metric of suitability for use. Exhibit
`
`2001 at Table 1. In contrast, the ‘413 patent concluded that “the drug substances
`
`… can be formulated in HFA 134a alone” after observing them to be visually
`
`acceptable for only 10 and 30 seconds after shaking. ‘413 patent, col. 7, lines 34-
`
`51. The realization that suspension formulations which achieved visual
`
`homogeneity for only a short period of time could reliably delivered reproducible
`
`doses surprised formulators in December 1991.
`
`13
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`IV. The Surfactant-Free Controls of the ‘011 PCT Application Would Not
`Motivate a Formulator to Use Suspension Formulations Containing
`Only HFA 134a and Drug to Treat Patients.
`
`30)
`
`I have read Dr. Smyth's Declaration, Exhibit 1006. While I
`
`respectfully disagree, I understand Dr. Smyth to have concluded, in view of the
`
`'011 PCT application's teaching that the concentration of surfactant should be kept
`
`to a minimum, and given the performance of Formulations 7 and 8, that an
`
`ordinary artisan would understand that the surfactant-free controls of Formulations
`
`7 and 8 of Example 1 would be suitable for therapeutic use. In my opinion, the
`
`performance of Formulations 7 and 8 relative to their respective surfactant-free
`
`controls did not provide a reasonable basis, as of December 1991, for concluding
`
`that a suspension formulation containing only HFA 134a and drug was satisfactory
`
`for therapeutic use.
`
`31) A formulator would have read the ‘011 PCT application as consistent
`
`with the state of the art in 1991. At that time it was known that suspension
`
`formulations, also known as dispersions, were physically unstable systems that
`
`required additional components to render the formulations satisfactory for
`
`therapeutic use. Purewal ‘183 patent, Exhibit 2004, col. 2, lines 58-62.
`
`32) The ‘011 PCT application makes clear that some amount of surfactant
`
`should be included to obtain stable dispersions of drug particles. It describes
`
`suitable concentrations of surfactant as ranging from 0.001 to 20% surfactant by
`
`14
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`weight of the drug particle to be suspended. Exhibit 1007 at page 7, lines 30-33.
`
`When specifically discussing formulations suitable for “medicinal purposes” the
`
`‘011 PCT application teaches a range of 0.001 to 3% surfactant by weight. Exhibit
`
`1007 at page 7. This statement reaffirms that suspension formulations containing
`
`only propellant and drug were not considered satisfactory for therapeutic use. In
`
`satisfactory formulations, the ‘011 PCT application teaches that at least 0.001 %
`
`surfactant by weight should be pre-coated onto the drug particles.
`
`33) The ‘011 PCT application states that pre-coating the drug with a
`
`surfactant leads to stable dispersions of the drug. Exhibit 1007, page 3, lines 9-28:
`
`It has been found that non-perfluorinated surfactants, which have
`
`previously been used as dispensing agents for powdered
`
`medicaments in propellants in which the non-perfluorinated
`
`surfactant is soluble, may be used to form stable dispersions of
`
`powdered medicament in propellants in which the non-
`
`perfluorinated surfactant is insoluble provided the medicament is
`
`pre-coated with the surfactant prior to dispensing in the propellant.
`
`This result is particularly surprising in view of the fact that the
`
`same stable dispersions cannot be achieved by simple admixture of
`
`the surfactant, propellant and medicament.
`
`34)
`
` The ‘011 PCT application instructs that the surfactant should be pre-
`
`coated onto the drug particles after which the pre-coated drug is added to the
`
`liquefied propellant. Exhibit 1007 at page 10, line 34 to page 11, line 19. The pre-
`
`15
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`coating step was an attempt to overcome the fact that conventional non-
`
`perfluorinated surfactants were insoluble in HFA 134a. Exhibit 1007 at page 2,
`
`lines 26-38.
`
`35) The ‘011 PCT application repeatedly states that pre-coating the
`
`surfactant on the drug particles results in stable dispersions of the suspended drug.
`
`For example, the ‘011 PCT application says that “[i]t has been found that non-
`
`perfluorinated surfactants … may be used to form stable dispersions of powdered
`
`medicament in propellants.” Exhibit 1007, at page 3, lines 17-22. Similarly, the
`
`‘011 PCT application states that its invention is an “aerosol composition
`
`comprising at least 0.001% by weight of a finely-divided solid medicament coated
`
`with a non-perfluorinated surface-active dispersing agent which constitutes from
`
`0.001 to 20% by weight of the coated solid medicament.” Exhibit 1007 at page 3,
`
`lines 9-14.
`
`36) The statement in the ‘011 PCT application that the concentration of
`
`surfactant should be “kept at a minimum” actually emphasizes that some
`
`meaningful amount of surfactant should be pre-coated onto the drug particles. The
`
`‘011 PCT application states that, “For best results, the concentration of surface-
`
`active agent is kept at a minimum as it may tend to increase the droplet size and the
`
`tendency for particle agglomeration.” Exhibit 1007 at page 7, lines 36-39. This
`
`statement reflects what was known in 1991 and is consistent with the formulation
`
`16
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`principle of including only the amount of excipient necessary (and specifically
`
`avoiding the addition of excessive amounts). Additionally, excessive amounts of
`
`surfactants could also give a formulation an unpleasant taste. Exhibit 2010,
`
`Richard N. Dalby et al., Medical Devices for the Delivery of Therapeutic Aerosols
`
`to the Lungs, in INHALATION AEROSOLS 441, 445 (1996). The ‘011 PCT
`
`application teaches that some amount of surfactant is necessary to achieve a stable
`
`dispersion and sets the minimal amount as 0.001 %. The ‘011 PCT application
`
`does not say or even suggest that surfactants could be eliminated altogether.
`
`37) Example 1 of the ‘011 PCT application describes the performance of
`
`Formulations 1-8 using only a single metric, the Drug Deposition Potential (DDP).
`
`Exhibit 1007 at 12-14. The DDP does not capture all the important considerations
`
`in developing satisfactory MDI formulations. DDP is not a metric that is
`
`frequently used in this field.
`
`38) The DDP compares drug deposition on the interior wall of an MDI
`
`canister against an arbitrarily chosen value. In the ‘011 PCT application that value
`
`is the amount of drug stuck to the canister wall in the absence of surfactant. As
`
`stated in the ‘011 PCT application, the DDP is used for the “quantification of the
`
`improvement” of how much drug stuck to the interior canister walls after pre-
`
`coating the drug with surfactant compared to the surfactant-free controls. Exhibit
`
`1007 at 12. Based on how the ‘011 PCT application defines and uses DDP, it
`
`17
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`would be an unwarranted stretch to use the DDP metric to make conclusions about
`
`whether surfactant-free controls would be satisfactory for therapeutic use.
`
`39) A formulator reading the '011 PCT application would understand that
`
`the surfactant-free controls serve only to establish an arbitrary baseline against
`
`which to measure the improvement brought about by pre-coating the drug particles
`
`with surfactant. Exhibit 1007 at page 12, lines 6-12. The ‘011 PCT application
`
`uses the surfactant-free controls to judge the test formulations. The ‘011 PCT
`
`application does not use the test formulations to judge the surfactant-free controls.
`
`A formulator would not understand the term “higher performing control.” Rather
`
`a formulator would expect the control to represent an arbitrary value against which
`
`the performance of test formulations would be evaluated.
`
`40) For example if the DDP were above 1.0—meaning that more drug
`
`sticks to the canister walls in the test formulation than in the surfactant-free
`
`control—that would support the conclusion that the test formulation is low
`
`performing and not suitable for therapeutic use. In this scenario, a formulator
`
`would not describe the surfactant-free controls as high performing. A formulator
`
`would describe the test formulation as low performing, because the test
`
`formulation could not even reach the arbitrary baseline set by the surfactant-free
`
`control. Therefore, even a DDP greater than 1.0 would not have led a formulator
`
`to believe that the surfactant-free controls were satisfactory for therapeutic use.
`
`18
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`41) Dr. Smyth states that a person of ordinary skill would interpret the
`
`DDPs of Formulations 7 and 8 in Example 1 of the ‘011 PCT application with
`
`reference to an 18% coefficient of variation which he calculates from data in
`
`Example 2. Smyth Decl., Exhibit 1006, ¶ 41-42. I respectfully disagree. Example
`
`2 only provides data from one formulation (same drug, surfactant, and surfactant
`
`concentration) prepared by two methods, each made in triplicate. ‘011 PCT
`
`application, Exhibit 1007, page 15. A formulator would not apply a measure of
`
`statistical variance derived from one formulation prepared by two different
`
`methods to the more diverse series of formulations in Example 1, which include
`
`different drugs, surfactants, and surfactant concentrations. It is not reasonable to
`
`apply a statistical variance calculated from that very limited data set to a much
`
`more diverse data set without a stronger statistical justification than Dr. Smyth was
`
`able to provide. Based on what the ‘011 PCT application discloses, the 0.92 DDP
`
`of Formulation 7 and the 0.85 DDP of Formulation 8 both demonstrate that
`
`surfactant-containing formulations were more stable than their surfactant-free
`
`controls. A formulator would have accepted this finding on its face.
`
`42) Example 2 of the ‘011 PCT application demonstrates the value of the
`
`pre-coating step. Exhibit 1007 at page 15. It attempts to demonstrate how adding
`
`the surfactant via a pre-coating step is superior to simply adding surfactant directly
`
`to the propellant. Example 2 does not suggest that surfactants could be omitted,
`
`19
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`but it does suggest that surfactant should be added in a particular way. In fact,
`
`Example 2 reaffirms that the surfactant-free controls are used only to establish an
`
`arbitrarily chosen baseline. Example 2 disparages those formulations displaying a
`
`DDP greater than 1.0 instead of saying that the surfactant-free controls are high
`
`performing formulations. Exhibit 1007 at page 15 (“The formulations of the
`
`invention cannot be arrived at by simply admixing ingredients and agitating the
`
`mixture in a conventional manner.”).
`
`V. The ’333 PCT Application Would Not Have Led a Formulator to Use
`Suspension Formulations Containing Only HFA 134a and Drug to
`Treat Patients.
`
`43) As of December 1991, a formulator would have read the ‘333 PCT
`
`application with the background understanding that suspension formulations
`
`containing only HFA 134a and drug were not satisfactory for therapeutic use. The
`
`prior art taught that suspension formulations containing only HFA 134a and drug
`
`did not re-create the appearance and performance of CFC formulations, see above
`
`at ¶ 24. For example, the Purewal ‘183 patent (Exhibit 2004) states (col. 2, lines
`
`58-62; emphasis added):
`
`It has been found that the use of Propellant 134a and drug as a
`
`binary mixture or in combination with a conventional surfactant
`
`such as sorbitan trioleate does not provide formulations having
`
`suitable properties for use with pressurised inhalers.
`
`20
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`44) Similarly, the Johnson ‘123 patent tests a control formulation having a
`
`drug suspended only in HFA 134a, but concludes that the formulation was not
`
`satisfactory for treating patients because it was not sufficiently stable. Exhibit
`
`2001 at Table 1 & col. 5, lines 49-53 (emphasis added):
`
`If significant flocculation occurs, that is, a cognizable coarseness
`
`and/or clear region can be observed, in less than about 15 sec., the
`
`suspension is deemed not stable enough for a practical aerosol
`
`inhalation drug formulation.
`
`45)
`
`In my opinion the Johnson ‘123 patent and Purewal ‘183 patent are
`
`representative of the state of the art as of December 1991. Both patents were filed
`
`prior to December 1991. In view of the state of the art, a formulator would not
`
`have read the ‘333 PCT application as suggesting that a suspension formulation
`
`containing only HFA 134a and drug would have been sufficiently stable for
`
`therapeutic use in treating patients.
`
`46) Neither the Mylan Petition nor Dr. Smyth’s Declaration addresses the
`
`consensus in the prior art that suspension formulations containing only HFA 134a
`
`and drug were unsatisfactory for therapeutic use. Instead, both argue that the
`
`claims are obvious (Smyth Decl., Exhibit 1006, ¶¶ 194-195) by relying solely on
`
`two passages in the ‘333 PCT application, which read (Exhibit 1011 at page 2 &
`
`6):
`
`21
`
`
`
`Case IPR2015-02002
`Attorney Docket No. 26368-0021IP1
`
`The formulations used in the invention contain fentanyl or a
`
`derivative thereof either in solution or suspension in the aerosol
`
`propellant system, optionally in the presence of a cosolvent. The
`
`solvent for fentanyl will generally be present in an amount in the
`
`range 5 to 25% by weight of the composition. The compositions
`
`may additionally comprise one or more surface active agents, for
`
`example oleic acids, complex esters and ester-ethers, e.g., sorbitan
`
`trioleate, Span 85, lecithins such as Epikuron 200, and fluorinated
`
`surfactants. The weight ratio of surface active agent to fentanyl is
`
`generally in the range 1 : 100 to 10 : 1.
`
`
`
` ….
`
`The perfluorinated surfactant constitutes at least 0.001% and
`
`generally up to 50%, usually up to 20%, desira