UNITED STATES PATENT AND TRADEMARK OFFICE
`___________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`___________________
`
`NOVEN PHARMACEUTICALS, INC.,
`Petitioner
`
`v.
`
`NOVARTIS AG AND LTS LOHMANN THERAPIE-SYSTEME AG,
`Patent Owners
`
`___________________
`
`Inter Partes Review No.: IPR2014-00550
`
`U.S. Patent No. 6,335,031
`
`REPLY DECLARATION OF CHRISTIAN SCHÖNEICH, PH.D.
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`1 of 50
`
`

`

`TABLE OF CONTENTS
`
`Page
`
`B.
`
`C.
`
`QUALIFICATIONS.......................................................................................1
`I.
`INFORMATION CONSIDERED..................................................................1
`II.
`III. REPLY TO DR. KLIBANOV’S DECLARATION ......................................1
`A.
`Dr. Klibanov’s Understanding of a POSA Is Inconsistent with
`the Clear Teachings in the Art .............................................................1
`Dr. Klibanov Misrepresents the State of the Art..................................3
`(1.) The Structural Features Affecting Bond Strength and
`Susceptibility to Oxidation Would Have Been Known to
`the POSA in 1998 ......................................................................3
`(2.) Dr. Klibanov’s Discussion of the Many Types of
`Physical and Chemical Degradation is Misleading and
`Unscientific..............................................................................10
`A POSA Would Have Predicted that Rivastigmine Would Be
`Susceptible to Oxidative Degradation................................................12
`(1.) Dr. Klibanov’s Definition of “Susceptibility to
`Oxidation” Fails.......................................................................12
`(2.) A POSA Would Have Predicted Susceptibility to
`Oxidation Based on the Molecule’s Chemical Structure.........14
`a.
`The prior art demonstrates that a POSA would
`examine a drug’s chemical structure and
`reasonably predict the type of degradation
`(including oxidation) .....................................................14
`Dr. Klibanov’s application of functional group
`chemistry contradicts his opinion that a POSA
`cannot make predictions based on the presence of
`functional groups...........................................................19
`(3.) Dr. Klibanov Confuses the Mechanism of Oxidative
`Degradation with a Compound’s Susceptibility to
`Oxidation..................................................................................20
`(4.) A POSA Would Predict Susceptibility to Oxidation
`Without Identifying a Rate-Limiting Step...............................25
`
`b.
`
`i
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`2 of 50
`
`

`

`TABLE OF CONTENTS
`(continued)
`
`Page
`
`(5.) Commercial Formulations Without an Antioxidant
`Would Not Teach a POSA that the Drug Molecule Was
`Stable........................................................................................27
`(6.) Dr. Klibanov Relies on Drug Molecules that do Not
`Contain the Combination of Structural Elements that
`Cause Rivastigmine to be Particularly Susceptible to
`Oxidative Degradation.............................................................34
`(7.) The Similarities Between the Structures of Nicotine and
`Rivastigmine and Nicotine’s Known Susceptibility to
`Oxidation Would Have Supported a POSA’s
`Understanding that Rivastigmine Is Susceptible to
`Oxidation..................................................................................37
`a.
`None of the distinctions between the structures of
`rivastigmine and nicotine raised by Dr. Klibanov
`make a difference ..........................................................37
`Linnell teaches that nicotine is susceptible to
`oxidation at the benzylic carbon....................................41
`(8.) Dextromethorphan Was Known to Be Susceptible to
`Oxidation..................................................................................43
`(9.) Comparisons in the Prior Art to Physostigmine Would
`Not Teach a POSA that Rivastigmine Is Oxidatively
`Stable........................................................................................45
`
`b.
`
`ii
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`3 of 50
`
`

`

`I, Christian Schöneich, Ph.D., declare and state as follows:
`
`I.
`
`QUALIFICATIONS
`
`1.
`
`For a discussion of my qualification and credentials, I refer to my
`
`curriculum vitae (Ex. 1023) and my April 2, 2014 declaration (Ex. 1011), which
`
`also provides a list of matters in which I have testified over the last four years, and
`
`my compensation.
`
`II.
`
`INFORMATION CONSIDERED
`
`2.
`
`I have reviewed the Declaration of Dr. Klibanov (Ex. 2012) and the
`
`documents cited in that report. Dr. Klibanov makes numerous statements in his
`
`declaration that are misleading and/or unscientific. I address these statements
`
`below.
`
`3.
`
`In forming my opinions, I have relied upon my accumulated
`
`scientific knowledge and experience. I have reviewed the documents cited in my
`
`April 2014 declaration (Ex. 1011), including the documents listed in paragraph 9
`
`of that declaration. I have also reviewed the documents cited in this declaration.
`
`III. REPLY TO DR. KLIBANOV’S DECLARATION
`
`A.
`
`4.
`
`Dr. Klibanov’s Understanding of a POSA Is Inconsistent with the
`Clear Teachings in the Art
`
`Dr. Klibanov states that a POSA could not make any predictions
`
`about the physical or chemical properties of a compound based on its structure:
`
`1
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`4 of 50
`
`

`

`I [Dr. Klibanov] disagree that a POSA would be able to
`make predictions about the physical or chemical
`properties of a compound based on its chemical structure.
`
`(Ex. 2012 at ¶ 25.) This statement is incorrect. Ordinarily-skilled artisans in 1998
`
`routinely made predictions about the physical/chemical properties of compounds
`
`based on chemical structure. (Ex. 1038 at 3.)
`
`5.
`
`As I described in my opening report (see, e.g., Ex. 1011 ¶¶ 14-46 in
`
`particular ¶¶ 32-35) and discuss below (see ¶¶ 7-15), a POSA could also make
`
`reasoned predictions about the strength of particular chemical bonds in a drug
`
`molecule and the susceptibility of the molecule to degradation, including oxidative
`
`degradation. A POSA was instructed by the prior art to assess a molecule’s
`
`chemical structure and make such determinations during pharmaceutical
`
`formulation development. (Ex. 2020 at 110; Ex. 2014 at 181, in particular see ¶¶
`
`23-25 below.)
`
`6.
`
`Indeed, Dr. Klibanov confirms the predictive value of chemical
`
`structure analysis. In his declaration, Dr. Klibanov states that a POSA could
`
`predict a molecule’s susceptibility to hydrolysis based on whether it contained a
`
`monomethyl or a dialkyl carbamate functional group. Dr. Klibanov states that
`
`monomethyl carbamates in general were known to degrade by hydrolysis (Ex.
`
`2012 ¶ 82) and “dialkyl carbamates were hydrolytically stable” (Ex. 2012 ¶ 86).
`
`These statements are inconsistent with the above statement by Dr. Klibanov that a
`2
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`5 of 50
`
`

`

`POSA could not make predictions about the physical/chemical properties of a
`
`compound because susceptibility to hydrolysis is a chemical property of a
`
`compound.
`
`B.
`
`Dr. Klibanov Misrepresents the State of the Art
`
`(1.) The Structural Features Affecting Bond Strength and
`Susceptibility to Oxidation Would Have Been Known to the
`POSA in 1998
`
`7.
`
`All principles presented in the “Background” section of my April
`
`2014 declaration (Ex. 1011 ¶¶ 14-46) were general organic chemistry principles
`
`known to the POSA in 1998. Dr. Klibanov, however, stated that my “opinions”
`
`concerning the predictability of rivastigmine’s susceptibility to oxidative
`
`degradation based on its chemical structure were “not background or state of the
`
`art as of January 12, 1998.” (Ex. 2012 ¶ 32.) I disagree. The information
`
`provided in the Background section of my April 2014 declaration concerning the
`
`basis for a POSA’s understanding that rivastigmine would be susceptible to
`
`oxidation under pharmaceutically relevant conditions would have been common
`
`knowledge to a POSA in 1998, and was documented in textbooks like Carey &
`
`Sundberg (Ex. 1007).
`
`8.
`
`A POSA in 1998 understood that the tertiary and benzylic nature of
`
`a carbon-hydrogen bond had real implications on the bond’s strength, and was not
`
`a mere “theoretical” issue as characterized by Dr. Klibanov (see e.g., Ex. 2012 ¶¶
`
`3
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`6 of 50
`
`

`

`122, 124, FN 15). It was well-known that carbon-hydrogen bonds to tertiary
`
`carbons and to benzylic carbons are substantially weaker because the radical that is
`
`formed on the carbon at those positions (by breaking the bond to hydrogen) is
`
`stabilized by the adjacent groups. (Ex. 1007at 679.) Carey & Sundberg (Ex.
`
`1007), a standard organic chemistry textbook, provides a table with the carbon-
`
`hydrogen bond strengths (i.e., the bond dissociation energy) of tertiary carbon-
`
`hydrogen bonds and benzylic carbon-hydrogen bonds. (Id. at 683.)
`
`9.
`
`As can be seen from the above table, increasing the number of alkyl
`
`substituents (e.g., -CH3) bonded to a carbon was known to reduce the strength of
`4
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`7 of 50
`
`

`

`the bond between that carbon and hydrogen (C–H) as follows:
`
`Chemical Context
`
`Methane
`CH3–H
`C-H bond to a primary C
`CH3CH2–H
`C-H bond to a secondary C
`(CH3)2CH–H
`C-H bond to a tertiary C
`(CH3)3C–H
`
`(Id. at 679, 683.)
`
`Relative Radical
`Stability
`Least stable
`
`Most stable
`
`Bond Dissociation
`Energy
`104 kcal/mol
`(highest bond strength)
`98 kcal/mol
`
`94.5 kcal/mol
`
`91 kcal/mol
`(lowest bond strength)
`
`10.
`
`An aromatic ring was also known to reduce the bond strength of an
`
`adjacent carbon-hydrogen bond because the aromatic ring stabilizes the resulting
`
`radical at the adjacent carbon by a phenomenon known as electron delocalization
`
`(or resonance). (Id. at 679-680.) Replacing one of the hydrogen atoms in methane
`
`(CH4) with a benzene aromatic ring (also called a “phenyl” ring) significantly
`reduces the carbon-hydrogen bond strength from 104 kcal/mol to 85 kcal/mol. (Id.
`
`at 683, compare the bond dissociation energies for methane (CH3–H) and toluene
`(PhCH2–H) in Table 12.4 (shown above) .)1
`
`
`1 A POSA would understand that “Ph” is a phenyl (i.e., benzene) aromatic ring so
`
`that the carbon-hydrogen bond is a benzylic carbon-hydrogen bond. (See ¶¶ 27-31
`
`of my April 2014 declaration for a detailed description of benzylic positions.)
`
`5
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`8 of 50
`
`

`

`11.
`
`Indeed, Carey and Sundberg confirms that a POSA would have
`
`known that benzylic and tertiary positions are “especially susceptible” to
`
`oxidation:
`
`Substrates that are relatively electron-rich or that provide
`particularly stable radicals are the most easily oxidized.
`Benzylic, allylic, and tertiary positions are especially
`susceptible to oxidation.
`
`(Ex. 1007at 693, emphasis added.)
`
`12.
`
`The effect of a tertiary amine on the strength of an adjacent carbon-
`
`hydrogen bond was likewise not “theoretical” to the POSA in 1998. A POSA
`
`would have known that a tertiary amine reduces the carbon-hydrogen bond
`
`strength at an immediately adjacent carbon by stabilizing the radical at the carbon.
`
`Carey & Sundberg confirms this effect as well, stating that “the stabilizing role of
`
`other functional groups can also be described in resonance terms” and providing
`
`the example of a tertiary amine group, further stating “dimethylamino [groups]
`
`have a stabilizing effect on a radical intermediate at an adjacent carbon.” (Ex.
`
`1007at 680.)
`
`13.
`
`Of particular importance here is that a POSA would have known that
`
`when two or more of the functional groups that render a carbon-hydrogen bond
`
`weak and susceptible to oxidation are adjacent to the same carbon-hydrogen bond,
`
`that position’s susceptibility to oxidation is substantially increased. Again, this
`
`6
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`9 of 50
`
`

`

`consideration was known and predictable and not “theoretical” as Dr. Klibanov
`
`contends. Carey & Sundberg, for example, describe the enhanced reactivity
`
`towards oxygen that occurs when the carbon-hydrogen bond at a benzylic position
`
`is also a tertiary position. Specifically, Carey & Sundberg provides the following
`
`table:
`
`(Ex. 1007at 693.) The relative reactivity of the carbon-hydrogen bond at a
`
`benzylic position that is also a tertiary carbon (as found in rivastigmine) is assigned
`
`the highest relative value of “1.0” for reactivity towards oxygen in the above table.
`
`The relative reactivity of the benzylic carbon-hydrogen bond at a benzylic position
`
`that is not also a tertiary position is 0.015. In other words, a carbon-hydrogen bond
`
`at a benzylic position that is also a tertiary position is 67 times more reactive to
`
`oxygen than a carbon-hydrogen bond to a benzylic position alone.2
`
`
`2 The known enhanced reactivity toward oxygen that occurs when multiple
`
`structural features are adjacent to the same carbon-hydrogen bond is particularly
`
`7
`
`Continued. . .
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`10 of 50
`
`

`

`14.
`
`A POSA in 1998 would not have considered the application of such
`
`well-known chemical principles to be “theoretical.” Rather, a POSA would have
`
`understood that the chemical principles discussed above and in my April 2014
`
`declaration were, in fact, background and state of the art. A POSA would have
`
`examined the structure of rivastigmine during preformulation development. By
`
`applying these well-known chemical principles to rivastigmine, a POSA would
`
`have immediately recognized that rivastigmine contains these structural features
`
`making a carbon-hydrogen bond particularly susceptible to oxidation: (i) the
`
`carbon-hydrogen bond (in red below) is immediately adjacent to an aromatic ring
`
`(in blue below); (ii) the carbon-hydrogen bond is also immediately adjacent to a
`
`tertiary amine (in green below); and (iii) the carbon-hydrogen bond is also
`
`immediately adjacent to an additional carbon substituent (-CH3, in purple below)
`making the red carbon a tertiary carbon.
`
`
`important. As I discuss below (see ¶¶ 49-51), Dr. Klibanov does not consider this
`
`effect when he discusses compounds that have a tertiary amine that is not
`
`immediately adjacent to the benzylic position.
`
`8
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`11 of 50
`
`

`

`Dr. Klibanov does not dispute the technical conclusion that the carbon-hydrogen
`
`bond in rivastigmine (in red above) would be weakened by virtue of its immediate
`
`proximity to these three features.
`
`15.
`
`Each of these three features was known by the ordinarily skilled
`
`artisan to reduce the bond strength between an immediately adjacent carbon and a
`
`hydrogen bound to that carbon because each stabilizes the resulting radical formed
`
`on that adjacent carbon when the carbon-hydrogen bond is broken. Further, the
`
`presence of all three features immediately adjacent to the same carbon-hydrogen
`
`bond would have been recognized by a POSA to render that bond particularly
`
`9
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`12 of 50
`
`

`

`susceptible to oxidation. Thus a POSA would have considered the determination
`
`of whether a drug like rivastigmine would be reasonably susceptible to oxidative
`
`degradation under pharmaceutically relevant conditions due to the presence of a
`
`readily identifiable and relatively weak covalent bond to be predictable and subject
`
`to well-known principles of organic chemistry. These principles would have
`
`provided the basis for a POSA to reasonably expect that rivastigmine is susceptible
`
`to oxidative degradation and accordingly observing oxidative degradation would
`
`not have been surprising to a POSA in 1998.
`
`(2.) Dr. Klibanov’s Discussion of the Many Types of Physical
`and Chemical Degradation is Misleading and Unscientific
`
`16.
`
`Dr. Klibanov provides a long list of chemical and physical
`
`degradative mechanisms and states that a POSA could not predict if an active
`
`pharmaceutical ingredient (API) would undergo any of the mechanisms:
`
`[whether] and to what extent an API undergoes any of
`these types of degradation under pharmaceutically
`relevant conditions in general would not be reasonably
`predicted in advance.
`
`(Ex. 2012 ¶ 33.) I disagree with this statement for two reasons. First, a POSA
`
`would have been able to make predictions regarding the susceptibility of an API to
`
`the types of degradation listed based on the chemical structure of the molecule,
`
`including predictions concerning oxidation as I address in greater detail below (see
`
`¶¶ 22-25). Dr. Klibanov’s statement that a POSA could not predict whether an
`10
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`13 of 50
`
`

`

`API will undergo “any of these types of degradation [including hydrolysis] under
`
`pharmaceutically relevant conditions in general” (Ex. 2012 ¶ 33) also conflicts
`
`with his later opinion that a POSA would have been able to reasonably predict that
`
`molecules containing a monomethyl carbamate are susceptible to hydrolysis. (Ex.
`
`2012 ¶ 82.)
`
`17.
`
`Second, although there are a variety of possible degradative
`
`mechanisms, the two of greatest concern to the POSA (who is undertaking the
`
`formulation of a drug) are oxidation and hydrolysis. (See, e.g., Ex. 2020 at 110)
`
`(“Chemically the most frequently encountered destructive processes are hydrolysis
`
`and oxidation”). Moreover, a POSA would have recognized that many of the types
`
`of degradation listed by Dr. Klibanov would not apply to a pharmaceutical
`
`formulation containing rivastigmine. For example, Dr. Klibanov lists solvolysis
`
`(including hydrolysis), dehydration, and decarboxylation among “the many
`
`different types of chemical and physical degradation that an active pharmaceutical
`
`ingredient (“API”) potentially may undergo.” (Ex. 2012 ¶ 33.) However, a POSA
`
`would have immediately recognized from the structure of rivastigmine that it does
`
`not contain functional groups that would be susceptible to hydrolysis, dehydration
`
`or to decarboxylation. Dr. Klibanov even admits that a POSA would understand
`
`that the carbamate group of rivastigmine is not susceptible to hydrolysis because
`
`the carbamate is a dialkyl carbamate. (Ex. 2012 ¶ 86.)
`
`11
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`14 of 50
`
`

`

`C.
`
`A POSA Would Have Predicted that Rivastigmine Would Be
`Susceptible to Oxidative Degradation
`
`(1.) Dr. Klibanov’s Definition of “Susceptibility to Oxidation”
`Fails
`
`18.
`
`Dr. Klibanov states that “[e]ssentially all organic compounds
`
`undergo oxidation under sufficiently harsh conditions (e.g., burning) regardless of
`
`their structure.” (Ex. 2012 ¶ 123.) This statement misses the point. As I explained
`
`in my earlier declaration, a POSA is an individual or group of individuals working
`
`to develop pharmaceutical formulations. (Ex. 1011 ¶ 11 (fn 1), ¶ 13.) The POSA
`
`would be concerned with susceptibility to degradation under pharmaceutically-
`
`relevant conditions, as addressed in my April declaration. As such, the POSA
`
`would not determine stability of drug molecules in the context of the “sufficiently
`
`harsh conditions (e.g., burning)” raised by Dr. Klibanov.
`
`19.
`
`I disagree with Patent Owner’s characterization of the term
`
`“susceptible” as used by me to refer “only to the theoretical potential of a
`
`compound to oxidatively degrade.” (Paper 25 at 16.) Patent Owner’s statement
`
`that I defined “susceptible” to mean “potential” (id.) is incomplete and takes my
`
`testimony from the district court case out of context. In fact, I also testified in that
`
`case that by susceptibility, I mean “likelihood.” (Ex. 1025 at 93:11-21, 115:3-22.)
`
`This testimony is consistent with my use of the term in my April 2014 declaration
`
`(Ex. 1011). For example, when I discussed that benzylic C-H bonds were known
`
`12
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`15 of 50
`
`

`

`to be “susceptible to oxidative degradation,” I stated that
`
`In other words, an ordinarily-skilled artisan would have
`known that the presence of a benzylic C–H bond in a
`drug molecule would likely dispose the drug to hydrogen
`abstraction and radical formation at that position.
`
`(Ex. 1011 ¶ 34). Further, in my previous testimony, I clearly stated that a POSA
`
`would have understood that rivastigmine, due to the presence of three adjacent
`
`functional groups, would be “particularly susceptible” and “prone to oxidation”
`
`because it contains an “especially weak,” and “readily-cleaved C-H bond.” (See,
`
`e.g., Ex. 1025 at 48:2-49:13; Ex. 1011 ¶¶ 12, 55.) Clearly my prior testimony does
`
`not support Patent Owner’s attempt to reinterpret my opinions by changing the
`
`meaning of susceptible.
`
`20.
`
`I also disagree with Dr. Klibanov’s characterization of the predicted
`
`susceptibility of rivastigmine to oxidation as “theoretical” (see, e.g., Ex. 2012 ¶¶
`
`122, 124) or as “speculation” (see, e.g., Ex. 2012 ¶ 132). As I discussed above (see
`
`¶¶ 7-15) and in my April 2014 declaration (see Ex. 1011, ¶¶ 14-46), the structural
`
`features that cause some carbon-hydrogen bonds to be weaker and prone to
`
`oxidation were known to the POSA and are not (and were not in 1998)
`
`“theoretical.”
`
`21.
`
`The table of bond dissociation energies in Carey & Sundberg (Ex.
`
`1007 at 683, Table 12.4) does not “confirm that essentially all organic molecules
`
`13
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`16 of 50
`
`

`

`(i.e., carbon-containing molecules) undergo degradation under sufficiently harsh
`
`conditions” as Dr. Klibanov contends. (Ex. 2012 ¶ 128, emphasis in original.) To
`
`the contrary, this table is presented by Carey & Sundberg to illustrate the
`
`“substituent effect on bond dissociation energies” in order to demonstrate that
`
`certain substituents cause an adjacent carbon-hydrogen bond to be weaker and
`
`therefore more susceptible to hydrogen abstraction (i.e., oxidation) by a free
`
`radical initiator. (Ex. 1007at 683; see also id. at 679.)
`
`(2.) A POSA Would Have Predicted Susceptibility to Oxidation
`Based on the Molecule’s Chemical Structure
`
`a.
`
`The prior art demonstrates that a POSA would
`examine a drug’s chemical structure and reasonably
`predict the type of degradation (including oxidation)
`
`22.
`
`Dr. Klibanov’s statements that a POSA could not reasonably predict
`
`whether a drug compound would be susceptible to degradation in general (Ex.
`
`2012 ¶ 33), or specifically to oxidation (Ex. 2012 ¶ 119) are incorrect. Dr.
`
`Klibanov contends that a “POSA would not have been able to reasonably predict
`
`whether rivastigmine would undergo oxidative degradation under pharmaceutically
`
`relevant conditions” “because of . . . the influence of the structure of the molecule
`
`as a whole on stability.” (Ex. 2012 ¶ 122, see also ¶¶ 32, 120, 132, 153, 156 and fn
`
`16). I respectfully find Dr. Klibanov’s opinion to be unscientific and misleading
`
`for at least two reasons: (i) the presence of certain functional groups in a
`
`molecule’s chemical structure was routinely used by POSAs to predict chemical
`
`14
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`17 of 50
`
`

`

`reactivity including oxidation; and (ii) Dr. Klibanov’s opinions on the
`
`susceptibility of monomethyl carbamates in general to hydrolysis violates his
`
`“whole molecule determines stability” notion and demonstrates that a POSA would
`
`make predictions on stability based on the presence of certain functional groups.
`
`In any event, I expressly considered the whole molecule (as a POSA would) when
`
`providing my opinions in my April 2014 declaration and previous testimony (Ex.
`
`1011 ¶¶ 11-12, 53-61; Ex. 1025 at 73:17-24, 81:7-82:12.) Dr. Klibanov does not
`
`opine that I overlooked any portion of the rivastigmine molecule that would have
`
`resulted in a different analysis.
`
`23.
`
`Art cited by Dr. Klibanov confirms that the POSA in 1998 would
`
`have examined a drug’s chemical structure during preformulation and could have
`
`reasonably predicted susceptibility to degradation processes including oxidation by
`
`applying known chemical concepts. Pharmaceutical Dosage Forms and Drug
`
`Delivery Systems (Ex. 2020), in a section entitled “Drug Stability,” states:
`
`Initial investigation begins through knowledge of the
`drug's chemical structure which allows the
`preformulation scientist to anticipate the possible
`degradation reactions.
`
`(Ex. 2020 at 110).
`
`24.
`
`Dr. Klibanov’s argument that degradation reactions are
`
`unpredictable because “the structure of the molecule as a whole (and not just the
`
`15
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`18 of 50
`
`

`

`presence of certain functional groups) determines the stability” (Ex 2012 at ¶ 120)
`
`attempts to propose that organic chemistry is unpredictable and that the POSA
`
`cannot make any predictions regarding the chemical reactivity of organic
`
`molecules. This proposition is incorrect. The identification of functional groups
`
`and the ability to predict reactivity based on functional group properties is a
`
`foundation of organic chemistry. Morrison & Boyd define functional groups
`
`stating:
`
`The atom or group of atoms that defines the structure of a
`particular family of organic compounds and, at the same
`time, determines their properties is called the functional
`group.
`
`(Ex. 1038 at 167, underline added, bold in original.) Other texts confirm that a
`
`POSA would have understood that the presence of functional groups within a
`
`molecule determine, in large part, the reactivity of the molecule:
`
`A study of functional groups is especially profitable
`because the reactions of a functional group tend to be
`about the same, regardless of the nature of the rest of
`the molecule.”
`
`(J.E. Leffler, A Short Course of Organic Chemistry, Ex. 1047 at 46, emphasis
`
`added.)
`
`25.
`
`Modern Pharmaceutics, cited by Dr. Klibanov, confirms that a
`
`POSA would have predicted a drug molecule’s susceptibility to degradation under
`16
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`19 of 50
`
`

`

`pharmaceutically relevant conditions based on the presence of particular functional
`
`groups:
`
`A cognizance of reactions of particular functional
`groups is important if one is to gain a broad view of drug
`degradation. It is a difficult task to recall degradative
`pathways of all commonly used drugs. Yet, through the
`application of functional group chemistry, it is possible
`to anticipate the potential mode(s) of degradation that
`drug molecules will likely undergo. In the following
`discussion, therefore, degradative routes are
`demonstrated by calling attention to the reactive
`functional groups present in drug molecules.
`
`(Ex. 2014 at 181, emphasis added). This chapter from Modern Pharmaceutics also
`
`provides two tables with some examples of functional groups that when present
`
`cause the drug to be “subject to” hydrolysis or oxidation (Ex. 2014 at 182, 183).
`
`These texts demonstrate that a POSA would, in fact, have made reasonable
`
`predictions of a drug molecule’s susceptibility to degradation under
`
`pharmaceutically relevant conditions by examining the molecule’s structure and
`
`applying known chemical concepts. Moreover, such determinations of
`
`susceptibility would not have required testing as Dr. Klibanov contends (see, e.g.,
`
`2012 ¶ 125).
`
`26.
`
`The fact that the extent of oxidative degradation depends on the
`
`17
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`20 of 50
`
`

`

`formulation does not mean that a POSA cannot predict whether rivastigmine will
`
`undergo oxidation under pharmaceutically relevant conditions as Dr. Klibanov
`
`contends (see, e.g., 2012 ¶ 125). A POSA would have understood that the
`
`oxidative reactions are influenced by environmental factors like the presence of
`
`initiators that cause oxidation. “Many autoxidation reactions are initiated by trace
`
`amounts of impurities, such as metal ions or hydroperoxides.” (Ex. 2014 at 183.)
`
`A formulator can attempt to limit the amount of initiators by careful selection of
`
`excipients. Likewise, the formulator can attempt to limit oxidation by excluding
`
`oxygen. However, excluding initiators and oxygen as a means of preventing
`
`oxidation can be problematic because often small amounts are all that is needed to
`
`create oxidation issues. As Connors explains:
`
`One of the major problems encountered in dealing with
`oxidation reactions is that some reactants such as oxygen
`or metal ion need not be present in more than trace
`quantities to produce significant stability problems.
`
`(Ex. 2021 at 80.) In addition, a formulation may address oxidation issues by
`
`adding an antioxidant. However, no matter the formulation or the steps taken to
`
`minimize oxidation, the susceptibility of a drug to oxidation is an inherent property
`
`of the drug based on its chemical structure. In other words, a POSA would
`
`understand that rivastigmine is particularly susceptible to oxidation due to its
`
`chemical structure. A POSA would understand that the amount of degradation
`
`18
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`21 of 50
`
`

`

`observed will depend upon the formulation including the presence of initiators and
`
`whether steps were taken to address the oxidation like adding an antioxidant.
`
`b.
`
`Dr. Klibanov’s application of functional group
`chemistry contradicts his opinion that a POSA cannot
`make predictions based on the presence of functional
`groups
`
`27.
`
`Dr. Klibanov contradicts his “whole molecule” notion, when he
`
`applies standard functional group chemistry in his declaration. Dr. Klibanov cites
`
`to no art supporting his opinion that “the structure of the molecule as a whole (and
`
`not just the presence of certain functional groups) determines the stability.” (Ex.
`
`2012 ¶ 120.) Dr. Klibanov’s only support for this proposition is a single sentence
`
`from the ’548 patent (Ex. 2037) stating “physostigmine free base is a particularly
`
`labile compound because its two basic tertiary amine groups facilitate hydrolysis of
`
`its [carbamate] group.” (Ex. 2012 ¶ 83 and FN16 citing the ’548 patent, Ex. 2037
`
`at 3:51-54.)
`
`28.
`
`Even though this particular example relates to hydrolysis of the
`
`carbamate group of physostigmine, Dr. Klibanov does not even apply his
`
`“molecule as a whole” notion when considering the stability of carbamates. To the
`
`contrary, Dr. Klibanov states that monomethyl carbamates in general were known
`
`to degrade by hydrolysis (Ex. 2012 ¶ 82) and “dialkyl carbamates were
`
`hydrolytically stable” (Ex. 2012 ¶ 86) without any consideration of the remaining
`
`chemical structure. In other words, Dr. Klibanov applies functional group
`19
`
`Noven Exhibit 1032
`Noven v. Novartis and LTS Lohmann, IPR2014-00550
`22 of 50
`
`

`

`chemistry when discussing hydrolysis of carbamates even though the only example
`
`he provides to support his “whole molecule” notion involves the hydrolysis of
`
`carbamates. Such inconsistency reveals the flaws in Dr. Klibanov’s opinions.
`
`29.
`
`It should also be noted that a POSA would have understood that
`
`there is no long distance effect of the amines in physostigmine on the hydrolysis of
`
`the carbamate in the same molecule. Rather, a POSA would have understood that
`
`the two basic tertiary amine groups in physostigmine facilitate hydrolysis of a
`
`carbamate group on other molecules of physostigmine. This effect would not have
`
`been surprising or unpredictable to a POSA. To the contrary, a POSA would have
`
`understood that the presence of the basic tertiary amines in one molecule would
`
`facilitate hydrolysis of the carbamate group on other molecules by the same
`
`mechanism as the “OH– -catalyzed deprotonation of (i.e., removal of the H from)
`
`the carbamate’s NH group” as described by Dr. Klibanov. (Ex. 2012 ¶ 92.) In
`
`other words, a POSA would understand and expect other basic groups (like the
`
`amines on other molecules of physostigmine) to catalyze the deprotonation of the
`
`carbamate just like