Page 1 of 12
`
`YEDA EXHIBIT NO. 2046
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`The glatiramoid class of immunomodulator drugs
`
`neurodegenerative disorders [17-21] and preliminary data
`from small studies in RRMS patients [22,23] suggest that
`daily treatment with GA may have neuroprotective and/or
`neurogenerative effects.
`Thus, GA was the first – and until recently was the only –
`member of the glatiramoids. Glatiramoids are a family
`of synthetic copolymer mixtures comprising the four amino
`acids, L-glutamic acid, L-alanine, L-lysine and L-tyrosine,
`in a defined molar ratio. GA has a unique mechanism of
`action, which although not completely explained, has
`demonstrated effects on several different components of the
`immune system [24-28]. GA is an antigen-based therapy; that
`is, a GA-specific immune response is the sine qua non of GA
`efficacy [28]. Furthermore, GA is unique in that the active
`epitopes (amino acid sequences associated with clinical effi-
`cacy) within the polypeptide mixture cannot be completely
`identified or characterized using state-of-the-art multidimen-
`sional separation techniques. Therefore, no two glatiramoids
`can ever be proved identical; however, it is possible to dis-
`tinguish among members of the glatiramoid class using ana-
`lytical, biological and immunological methods.
`A second glatiramoid, protiramer, was recently developed
`by Teva Pharmaceutical Industries and has been tested in
`two small Phase II clinical trials in RRMS patients [29]. Pro-
`tiramer is produced by making slight changes to the GA
`manufacturing process. Protiramer has a higher molecular
`mass (MM) distribution than GA and was synthesized to
`determine whether
`the
`increased
`immunoreactivity of
`higher MM peptides could improve efficacy and/or decrease
`dosing frequency (early GA studies used a higher MM
`formulation than the now marketed formulation [5]). Addition-
`ally, Sigma-Aldrich Co. manufactures a glatiramoid called
`Poly(Ala:Glu:Lys:Tyr) (Alanine:glutamic acid:lysine:tyrosine)
`that is described as having an effect similar to GA in EAE;
`this glatiramoid is not recommended for use in humans [30].
`More glatiramoids may become available as other manufacturers
`develop their own copolymer mixtures.
`As described below, preclinical experience with protiramer
`demonstrates that glatiramoids that are quite similar to each
`other, but not identical, cannot be presumed to have com-
`parable safety and efficacy profiles. Even slight differences in
`MM distribution or in the primary, secondary or tertiary
`polypeptide structure of different glatiramoids can signifi-
`cantly alter their pharmacologic activity, as was illustrated by
`the results of preclinical studies of protiramer (described
`below). This review describes experience with GA, the best-
`studied member of this growing therapeutic class, and the
`relevance of GA findings as they relate to important safety
`and efficacy considerations for new glatiramoid mixtures
`in development.
`
`2. Chemical characterization
`
`Glatiramoids are complex polypeptide mixtures that share a
`specific molecular formula. GA is a glatiramoid prepared
`
`from N-carboxy-α-amino acid anhydrides (monomers)
`with diethylamine as the polymerization initiator. The
`bifunctional amino acids are protected (the δ-NH2 of lysine
`is protected by a trifluoroacetyl group and the γ–COOH of
`glutamic acid is protected by a benzyl group); therefore, the
`polymerization occurs through the growth of linear chains
`from monomers, with no crosslinking between the polymer
`chains. Polymerization is followed by polymer cleavage and
`deprotection. The amino acid sequences and the size of the
`resultant polypeptides are dependent on factors such as the
`relative reactivity of the activated amino acid monomers and
`reaction conditions such as temperature and duration of
`cleavage process. As a result, the sequences of polypeptides
`in GA, although not uniform, are not entirely random
`and are highly reproducible under strictly controlled
`reaction conditions.
`Glatiramoids are characterized by the molecular formula
`below, in which X represents an anion (e.g., acetate or any
`other pharmaceutically acceptable salt). The superscripts rep-
`resent the relative molar ratios of amino acids and the sub-
`script, n, relates to the polymeric chain length, and m is the
`molar quantity of counter-ions.
`(L-Glu13 – 15, L-Ala39 – 46, L-Tyr8.6 – 10, L-Lys30 - 37)n mX.
`GA is composed of Glu, Ala, Tyr and Lys in an approximate
`molar ratio of 0.14:0.43:0.09:0.34 and the average MM of
`GA is 5000 – 9000 daltons (Da) [31]. Most of the polymers
`and copolymers of amino acids in GA have an MM distri-
`bution of ∼ 2500 – 20,000 Da. The glatiramoid, protiramer
`(formerly known as TV-5010), which is produced by
`making slight changes to the GA manufacturing process
`(e.g., temperature, reaction time) following the polymeri-
`zation reaction has the same molar ratio of amino acids
`as GA and the average MM is 13,500 – 18,500 Da.
`Sigma reagent Poly(Ala:Glu:Lys:Tyr) has a molar ratio of
`0.14:0:0.42:0.07:0.36 of Glu, Ala, Tyr and Lys, respectively,
`and the average MM is 10,000 – 20,000 Da [30].
`GA and other glatiramoids contain an almost incalculably
`large number of amino acid sequences (> 1036 possible
`theoretical sequences in GA). It is at present impossible to
`isolate and identify active amino acid sequences (i.e., those
`acting as epitopes), even using the most technologically
`sophisticated multidimensional separation techniques. The
`consistency of polypeptide sequences within GA is depen-
`dent on a well-controlled proprietary manufacturing process.
`Therefore, no
`two glatiramoid mixtures prepared by
`different manufacturers can be shown to be ‘identical’ and
`new glatiramoids must be considered distinct members
`of the class. There are means by which to differentiate
`glatiramoid mixtures. Members of this class can be distin-
`guished by the following characteristics: MM distribution
`profile, peptide mapping by capillary electrophoresis profile,
`certain nonrandom and reproducible patterns in amino acid
`sequences, secondary and tertiary structures, specific hydrophobic
`interactions owing to unique charge dispersion, characteristic
`ratio between molecules with C-terminal carboxylates
`
`658
`
`Expert Opin. Pharmacother. (2009) 10(4)
`
`Page 2 of 12
`
`YEDA EXHIBIT NO. 2046
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`Varkony, Weinstein, Klinger, Sterling, Cooperman, Komlosh, Ladkani & Schwartz
`
`Protiramer
`
`GA
`
`0.00
`
`5.00
`
`10.00
`
`15.00
`
`20.00
`25.00
`30.00
`Retention time, min
`
`35.00
`
`40.00
`
`45.00
`
`50.00
`
`Sigma poly (Ala:Glu:Lys:Tyr)
`
`GA
`
`A.
`
`mAU
`
`B.
`
`mAU
`
`0.00
`
`5.00
`
`10.00
`
`15.00
`
`20.00
`25.00
`30.00
`Retention time, min
`
`35.00
`
`40.00
`
`45.00
`
`50.00
`
`Figure 1. Molecular mass distribution by gel permeation chromatography of: (A) GA and protiramer; (B) GA and Sigma
`Poly(Ala:Glu:Lys:Tyr).
`Ala:Glu:Lys:Tyr: Alanine:glutamic acid:lysine:tyrosine; GA: Glatiramer acetate.
`
`and C-terminal diethylamides, and proteolytic enzymatic
`digestion profile.
`For GA, the MM distribution profile, based on the
`separation of polypeptides according to size, is determined
`using a gel permeation column calibrated using a set of
`sequence-defined, well-characterized proprietary
`linear
`polypeptide markers selected based on certain nonrandom
`patterns of amino acid sequences. There is some overlap
`in MM distribution between GA and protiramer (Figure 1A),
`and between GA
`and Sigma Poly(Ala:Glu:Lys:Tyr)
`(Figure 1B). Polypeptide mapping using capillary electrophore-
`sis separation of polypeptide fragments obtained after
`
`digestion with trypsin and mapping based on the proteolytic
`hydrolysis by carboxypeptidase P followed by separation
`of the fragments by reverse-phase HPLC are methods
`of discerning sequence differences among GA structures
`and those of other glatiramoids (Figures 2 and 3).
`The sequence of amino acids (primary structure) of
`the polymer obtained at the first stage of the synthesis
`in a bulk solution is governed mainly by the homopolymer-
`ization rate constants of each of the activated amino
`acids (monomers) present and by reaction conditions
`(e.g., temperature and concentration). The size of the
`GA mixture components and the nature of the terminal
`
`
`
`Expert Opin. Pharmacother. (2009) 10(4)
`
`659
`
`Page 3 of 12
`
`YEDA EXHIBIT NO. 2046
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`The glatiramoid class of immunomodulator drugs
`
`17.883
`
`16.079
`
`14.396
`
`18.226
`
`16.767
`
`16.350
`15.479
`
`21.317
`
`19.638
`
`19.933
`
`19.383
`
`24.92524.417
`23.27522.926
`22.508
`
`24.071
`
`22.737
`21.633
`20.983
`
`20.279
`
`31.992
`
`30.388
`29.954
`29.358
`
`27.929
`26.958
`
`25.467
`
`17.208
`
`16.996
`16.571
`
`17.421
`
`19.013
`18.746
`18504
`
`13.229
`13.067
`12.246
`
`15.104
`
`14.804
`
`11.967
`
`14.046
`13.504
`
`12.854
`
`12.496
`
`30
`
`25
`
`20
`
`15
`
`10
`
`5
`
`0
`
`-5
`
`mAU
`
`0
`
`5
`
`10
`
`15
`
`20
`
`25
`
`30
`
`35
`
`Min
`
`Figure 2. Typical electropherogram of GA fragments after exposure to trypsin.
`
`amino acids are dependent on the acetolytic cleavage
`conditions.
`Among tests to explain the primary structure are spectro-
`scopic techniques (Fourier transform infrared, ultraviolet, proton
`and carbon13 NMR) and enzymatic hydrolysis followed by
`chromatographic separation of the fragments to demonstrate
`the characteristic composition of the obtained mixture.
`Another test of the primary polypeptide structure is Edman
`degradation, a step-wise sequential hydrolysis of amino acids
`starting from the N-terminal end of the polypeptide. In this
`method, the characteristic sequence of amino acids in the
`polypeptide chain at the N-terminal end is determined by
`step-by-step cleavage of amino terminal residues without
`disrupting other polypeptide bonds. The GA polypeptide
`mixture exhibits a consistent and characteristic average order of
`amino acids in the N-terminal region. Additionally, GA has
`a certain fixed ratio of molecules with C-terminal carboxylic
`acids to those with C-terminal diethylamide (originating
`from the polymerization reaction initiator, diethylamine).
`Information on the secondary structure of GA can be
`obtained by circular dichroism measurements showing that
`GA possesses relatively stable secondary structures with sub-
`stantial α-helical content. These results were supported by
`evaluating the denaturation energy of GA drug substance
`(by measuring circular dichroism at different temperatures) and
`demonstrating the presence of a specific absorbance by second
`derivative Fourier transform infrared that is characteristic of
`α-helical structures.
`
`Information on GA tertiary structure can be obtained by
`comparing the size of the GA molecules before and after
`denaturation with guanidine HCl and by measuring the
`migration time on a gel permeation column, expressed in
`Kav (the smaller the Kav value, the larger the molecule size).
`Glatiramer acetate possesses a small degree of tertiary structure.
`Together, these tests indicate GA contains certain nonrandom
`sequences, is characterized by partial α-helical structure and
`has a small degree of tertiary structure.
`It is known that proteins tend to form quaternary structures
`resulting in formation of high MM aggregates. Although
`polypeptides are less likely to aggregate, their presence is
`monitored in GA when it is produced and in stability studies.
`Quantitation of stable high MM species indicates that levels
`are typically quite low.
`
`3. Mechanisms of action
`
`After extensive study in laboratories worldwide focusing on
`the mechanism of action of GA, the active epitopes in the
`GA mixture and their specific effects on the immune system
`are still not fully understood. Preliminary data suggest
`protiramer has a similar, but not identical, mechanism of
`action to that of GA.
`
`3.1 Experience with GA
`Mechanisms that are thought to contribute to GA effects
`include:
`i) high affinity binding
`to MHC class II
`
`660
`
`Expert Opin. Pharmacother. (2009) 10(4)
`
`Page 4 of 12
`
`YEDA EXHIBIT NO. 2046
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`Varkony, Weinstein, Klinger, Sterling, Cooperman, Komlosh, Ladkani & Schwartz
`
`AU
`
`0.001
`
`0.000
`
`AU
`
`0.001
`
`0.000
`
`AU
`
`0.001
`
`0.000
`
`AU
`
`0.001
`
`0.000
`
`AU
`
`AU
`
`0.001
`
`0.000
`
`0.001
`
`0.000
`
`GA 538532
`
`GA 538533
`
`GA 538534
`
`GA 538535
`
`GA 538542
`
`Sigma poly
`
`*
`
`*
`
`*
`
`*
`
`*
`
`*
`
`*
`
`4.00
`
`6.00
`
`8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00
`Min
`
`Figure 3. Comparative reverse-phase HPLC chromatograms of GA (the top five chromatograms show different batches of GA
`drug product) and Sigma Poly(Ala:Glu:Lys:Tyr) (bottom chromatogram) proteolytic digests by carboxypeptidase P.
`Ala:Glu:Lys:Tyr: Alanine:glutamic acid:lysine:tyrosine; GA: Glatiramer acetate.
`
`molecules on antigen presenting cells (APCs) and competition
`with MBP at the APC level for binding to MHC; ii) inhibition
`of MBP-specific T-cell activation through competition with
`MBP/MHC complexes for the T-cell receptor; iii) induction
`and activation of glatiramer acetate-reactive T cells and a
`shift from a type-1 T helper (TH1) phenotype, which
`tends to promote inflammation, to a type-2 T helper (TH2)
`phenotype, which typically promotes an anti-inflammatory
`environment;
`iv) preferential migration of GA TH2
`cells into the CNS leading to decreased local inflammation
`through ‘bystander suppression’; and v) neuroprotection
`and axonal protection related to GA-stimulated secretion
`of brain-derived neurotrophic factor, an important factor
`for neuronal
`survival, neurotransmitter
`release
`and
`dendritic growth [24-28].
`Researchers continue to investigate and discover novel
`mechanisms of GA activity. Recently, scientists at the
`
`Weizmann Institute demonstrated that GA treatment interferes
`with demyelination directly at the myelin and stimulates
`remyelination in an EAE model [32]. These effects were
`attributed not only to reduced inflammation, but also to a
`GA effect on the proliferation, differentiation and survival
`of oligodendrocyte progenitor cells and their recruitment to
`injury sites, thereby enhancing repair in situ.
`A substantial fraction of the therapeutic GA dose is
`hydrolyzed locally at the site of injection [31,33]. GA interacts
`with peripheral blood lymphocytes locally at the site of injec-
`tion, and the immune response is secondarily manifested as a
`systemic distribution of activated GA-specific T cells. T cells
`produced in the periphery cross the blood–brain barrier and
`accumulate in the CNS [27,28]. Thus, systemic distribution of
`the drug is irrelevant to effects following s.c. administration and
`systemic concentrations of GA or its metabolites are not
`indicative of drug activity or exposure to the immune system.
`
`
`
`Expert Opin. Pharmacother. (2009) 10(4)
`
`661
`
`Page 5 of 12
`
`YEDA EXHIBIT NO. 2046
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`The glatiramoid class of immunomodulator drugs
`
`GA affects immune cells in an antigen-specific way; that
`is, GA administered subcutaneously daily over many years
`works as an antigen-based vaccine [28]. On repeated expo-
`sure, GA-specific T cells shift from a TH1 to a TH2 pheno-
`type. After several months of treatment, GA-reactive T-cell
`proliferation declines [33]. Despite the decline, during long-
`term treatment there is no decrease in magnitude of crossre-
`activity between GA-reactive T cells and MBP, and the
`cytokines released by GA-reactive T cells remain TH2-
`biased [34,35]. This effect is sustained in patients taking daily
`GA injections for > 6 – 9 years [34]. Because T cells respon-
`sive to myelin antigen epitopes and GA-reactive T cells seem
`to represent the same or overlapping T-cell populations,
`long-term chronic administration of GA may restore immu-
`nologic tolerance in MS patients by sustained deletion of, or
`anergy induced toward, myelin-antigen-specific T cells [34,35].
`
`3.2 Considerations for new glatiramoids
`Because the active epitopes of GA and other glatiramoids
`have not been identified, it is impossible to predict whether
`they have comparable pharmacologic and immunological
`activity. Moreover, unlike most conventional drugs, glati-
`ramoids are not amenable to typical pharmacokinetic profil-
`ing, eliminating this approach to establishing comparability
`between two glatiramoids. Several approaches have been
`used to determine the bioavailability of GA with little
`success. The use of radiotracers in animal models was unin-
`formative owing to the extremely rapid breakdown of GA
`after s.c. injection and attempts to assess bioavailability by
`urinary measurements were not effective for the same reason.
`More
`importantly, preclinical studies suggest that the
`GA-induced T-cell mediated immune response, not systemic
`concentrations of GA or its metabolites, is associated with
`drug efficacy. The immunomodulating activity of GA can be
`adoptively transferred to recipient mice by GA-specific
`T cells and not by the drug or by the serum [36]. Therefore,
`measuring concentrations of a new glatiramoid or its
`metabolites in the systemic circulation would not be indicative
`of drug activity at the site of action (i.e., the CNS).
`T-cell receptors (TCR) respond to antigenic portions of
`peptides in a characteristic way and modifications to the
`amino acids in the peptide, for example, a substitution or
`deletion, create an altered peptide ligand (APL). Thus,
`differences in amino acid sequences of the polypeptides of
`different glatiramoids alter the interaction at the TCR. APLs
`can affect the repertoire and specificity of T cells induced,
`which could influence the strength and the nature of the
`immunological response [37]. As noted above, GA activity
`depends on antigen (epitope) presentation by APCs to
`T cells followed by T-cell activation. T-helper cells bind to
`complexes of short contiguous amino acid sequences (about
`13 – 17 amino acids) of the antigen protein bound to
`MHC class II molecules present on the surface of the APCs.
`The TCR interacts with both the MHC molecule and the
`polypeptide fragment. Thus, the repertoire of activated T-cell
`
`clones following exposure to the polypeptide antigen and
`the nature of the T-cell response are driven by the specific
`set of short contiguous amino acid sequences generated fol-
`lowing antigen presentation by the APCs. The TCR can be
`exquisitely sensitive to changes in the structure presented by
`an APL. Studies have shown that interactions with APLs can
`result in dramatically different phenotypes of induced T
`cells and mutation of even a single amino acid is sufficient
`to alter the T-cell response from strong killing to no response
`at all [37-43]. In a study investigating the basis by which a
`TCR can discriminate between two polypeptides differing
`only at a single MHC anchor residue, it was shown that the
`residue substitution did not significantly alter binding of the
`polypeptide to the MHC class II molecule; however, it
`reduced the specific T-cell response ∼ 1000-fold [39].
`
`4. Immunogenicity
`
`Glatiramoids affect both cellular and humoral immunity
`and there are consistent pharmacodynamic effects associated
`with their administration [28,29,33]. In addition to stimu-
`lating peripheral blood
`lymphocytes, both GA and
`protiramer induce the production of antibodies in MS
`patients. Anti-drug antibodies should be characterized by
`their ability to neutralize drug efficacy, cause serious adverse
`events or bind to endogenous proteins that are crossreactive
`with the drug.
`
`4.1 Experience with GA
`GA induces the formation of circulating anti-GA-specific
`antibodies in all treated animals and patients [33,44-46]. These
`antibody levels peak between 3 and 6 months of treatment
`and then gradually decline (Figure 4A) [33]. GA-reactive anti-
`bodies are mainly of the IgG class. These antibodies tend to
`shift from an IgG1 toward IgG2 and IgG4 isotypes, which
`correlates with the GA-mediated shift in T-cell phenotype
`from a TH1 to a TH2 milieu [33,44,45].
`Anti-GA antibodies do not seem to be neutralizing.
`Several preclinical and clinical studies have evaluated the
`effect of GA-reactive antibodies on biological activity and
`clinical efficacy [33,46]. In vivo and in vitro studies using
`serum samples from patients with the highest titers of GA-
`reactive antibodies indicate that these antibodies do not
`interfere with: i) the ability of GA to block EAE induction
`in mice in vivo; ii) activation of GA-specific T cells in vitro;
`iii) the ability of GA to inhibit the activation of MBP-spe-
`cific T-cell lines in vitro; or iv) binding of GA polypeptides
`to MHC class II molecules in vitro [46]. Additionally, in the
`35-month pivotal trial, clinical benefits of GA were apparent
`early and maintained throughout the treatment period,
`regardless of changes in GA-reactive antibody levels in
`RRMS patients [8].
`The safety implications of GA-reactive antibodies in MS
`patients receiving chronic treatment have also been evaluated.
`So far, no correlation has been found between anti-GA
`
`662
`
`Expert Opin. Pharmacother. (2009) 10(4)
`
`Page 6 of 12
`
`YEDA EXHIBIT NO. 2046
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`Page 7 of 12
`
`YEDA EXHIBIT NO. 2046
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`The glatiramoid class of immunomodulator drugs
`
`5. Preclinical and clinical experience
`
`5.1 Experience with GA
`The literature is replete with preclinical data and results of
`clinical efficacy trials of GA in RRMS patients. A detailed
`discussion is beyond the scope of this article and the reader
`is referred to primary publications of these data [1-16]. Briefly,
`in preclinical studies, GA prevented onset of EAE or amelio-
`rated the symptoms of existing EAE effects irrespective of
`animal species, disease type or encephalitogen used for EAE
`induction [1,2]. The most frequent adverse effects in preclini-
`cal studies were injection site reactions, including mild
`fibrosis. In RRMS patients, extensive clinical data demon-
`strate the beneficial effects and safety of GA on the clinical
`indices of MS – relapse rate and progression of disability –
`as measured by the Expanded Disability Status Scale [7-10,15,16].
`Supportive evidence for the anti-inflammatory and neuro-
`protective effects of GA was demonstrated by conventional
`and more advanced MRI techniques. MRI scans of the brain
`show that GA treatment reduces the number of enhancing
`lesions, decreases lesion load, inhibits new lesions from
`developing into permanent ‘black holes’ (areas of severe and
`permanent tissue damage) and reduces brain atrophy [11-14].
`Finally, studies show that GA efficacy and safety are
`sustained with long-term continuous use (> 12 years) [16]. In
`MS patients, the most frequent adverse events observed with
`GA therapy are injection site reactions and symptoms asso-
`ciated with an immediate, transient post-injection reaction,
`which may include vasodilation, chest pain, palpitation,
`tachycardia or dyspnea [7-10,16,31].
`
`5.2 Experience with protiramer
`Protiramer was tested in preclinical safety evaluations in
`monkeys, rats and swine in studies ranging in duration from
`6 days to 52 weeks. Repeated injections (twice weekly) in
`monkeys and rats resulted in severe injection-site lesions
`with disseminated necrosis and inflammation of dermal
`structures, including muscle, nerves and blood vessels.
`Extensive fibrosis was observed in all treated animals.
`Although GA is administered more frequently than protiramer,
`only mild and well-tolerated inflammation with slight
`fibrosis has been seen with GA at comparable doses.
`On chronic administration of protiramer in Sprague-Dawley
`rats (0, 2.5, 40 and 300 mg/kg s.c. protiramer twice weekly
`for 26 weeks), several treatment-related mortalities occurred
`in rats treated at the middle (40 mg/kg) and high (300 mg/kg)
`protiramer dose levels (300 mg/kg is ∼ 70-fold greater than
`the estimated concentration of protiramer used in human
`clinical trials[47]). Local reactions to treatment (e.g., induration,
`erythema, hematoma) were dose-dependent. Some hematol-
`ogy and serum clinical chemistry parameters were affected,
`including reductions in prothrombin time/activated partial
`thromboplastin time and red blood cell counts, increased
`platelet and neutrophil counts, decreased total protein and
`albumin concentrations, increased globulin concentration,
`
`increased total cholesterol levels, and slightly decreased
`serum sodium and creatinine levels. There was a marked
`increase in protein levels in the urine of animals treated with
`40 and 300 mg/kg. Treatment-related changes seen in histo-
`pathology examinations were confined to the injections sites,
`liver and kidneys of animals treated with 2.5, 40 and 300
`mg/kg. The incidence and severity of lesions at the injection
`site were dose-related and consisted of a thick fibrotic layer
`associated with necrosis in the deep dermis. Liver lesions
`consisted mainly of bridging fibrosis with bile duct cell pro-
`liferation and lymphoid cell infiltration in the periportal
`area, leading to restricted vascular perfusion of the liver.
`Kidney lesions indicated progressive nephropathy composed
`of fibrosis, lymphoid cell infiltration, tubular basophilia and
`tubular dilatation.
`A 52-week study was conducted to determine the toxicity
`and immunotoxicity of protiramer in the cynomolgus monkey.
`Protiramer was administered to monkeys at dose levels of 0, 2,
`10 and 60 mg/kg twice weekly. Two deaths occurred in the
`highest (60 mg/kg) protiramer dose group at 24 weeks and
`at 40 weeks of treatment. In both cases, pathological examina-
`tion revealed fibrosis, lymphoid and eosinophilic infiltrates, as
`well as s.c. and/or vascular necrosis at the injection site.
`Although these signs were considered to be factors contributing
`to death, other causes of death cannot be excluded.
`Again, these toxicity signs were not observed in chronic
`toxicity studies of GA in rats and monkeys.
`Importantly, the serious toxic effects of protiramer only
`became apparent after > 3 months of chronic administration;
`short-term (3 month) toxicity studies in rats and monkeys were
`completed successfully, with no serious adverse effects detected.
`Therefore, longer preclinical toxicity testing is warranted to
`ensure the safety of new glatiramoid drugs for chronic use.
`Based on generally favorable results of the short-term
`toxicity studies, and before serious toxic effects in preclinical
`testing were observed, protiramer was approved for testing
`in two small, short-term clinical studies with RRMS
`patients [29]. One study evaluated a 15 mg once-weekly dose
`and the other evaluated a 30 mg once-weekly dose; both
`studies comprised a 10-week pretreatment phase followed by
`a 36-week treatment phase. Once-weekly s.c. protiramer
`injections were apparently well tolerated. The most common
`adverse events in both studies were injection site reactions
`(erythema, pain, induration) and transient immediate post-
`injection site reactions. MRI outcomes showed that treat-
`ment with a 15 mg/week protiramer dose was suboptimal,
`but 30 mg/week protiramer significantly reduced gadolinium
`(Gd)-enhancing and T2-weighted lesions compared with
`pretreatment values. However, study results should be
`weighed cautiously because there was a large reduction in
`Gd-enhancing and T2-weighted lesions in these patients
`during the 10-week pretreatment period. Moreover, MRI
`changes do not necessarily predict clinical effects. For
`example, increasing the dose of GA from 20 to 40 mg daily
`in RRMS patients showed a trend for better efficacy with
`
`664
`
`Expert Opin. Pharmacother. (2009) 10(4)
`
`Page 8 of 12
`
`YEDA EXHIBIT NO. 2046
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`Varkony, Weinstein, Klinger, Sterling, Cooperman, Komlosh, Ladkani & Schwartz
`
`the higher GA dose, as indicated by a borderline significant
`reduction in the number of Gd enhancing lesions on
`T1-weighted MRI, but the reduction was not accompanied by
`a clinically meaningful reduction in relapse rate [50].
`Three patients (two in the 15 mg/week and one in the
`30 mg/week) had anti-protiramer IgE antibodies that were
`slightly above the limit of detection; none of these patients
`developed a hypersensitivity reaction [29]. As noted above,
`the typical (IgG) antibody profile with protiramer was dif-
`ferent from that observed with GA (Figure 4A and B).
`Whether continued treatment with protiramer would have
`led to other adverse effects, or whether altered immunogenic
`responses to the drug would have eventually compromised
`clinical efficacy is unknown. Drug development of protiramer
`has been terminated owing to the serious adverse events
`observed in chronic toxicity studies.
`
`5.3 Considerations for new glatiramoids
`As was done with GA and protiramer, and should be done for
`any new chemical entity, a careful prospective development
`program for new glatiramoids is required from preclinical
`toxicology and pharmacology studies, through well-con-
`trolled preclinical and clinical trials, using sensitive and validated
`procedures. MRI outcomes are not validated or accepted surro-
`gate endpoints for clinical efficacy [51]; therefore, glatiramoid
`pivotal studies should measure clinical end points (relapse rate
`and disability progression). Glatiramoids require long-term
`chronic administration and clinical studies must have sufficient
`power and duration to adequately assess the safety, efficacy
`and immunogenicity of each new member of the class.
`Experience with natalizumab and reports of progressive mul-
`tifocal leukoencephalopathy emphasize the need for longer-term
`safety data for immunomodulatory therapies [52,53]. MRI is
`a useful tool for MS diagnosis and is widely used in Phase II
`proof of concept studies; however, monitoring only CNS lesions
`using conventional MRI is neither indicative of complete
`disease burden nor predictive of clinical events, and, therefore,
`clinical outcomes are preferred [54,55].
`Another important consideration for new glatiramoids is
`the risk of potentially detrimental immunological consequences
`when introducing a new glatiramoid with altered epitopic
`sequences to patients previously treated with GA. Similarly,
`the likelihood of a potentially dangerous immunologic reaction
`in a patient switching to GA from another glatiramoid is
`unknown. Therefore, a crossover clinical study, in which
`patients who have been treated with GA are switched to the
`new glatiramoid mixture under evaluation, and patients
`exposed to the glatiramoid are subsequently treated with GA,
`may be necessary to adequately address this safety concern.
`
`6. Conclusion
`
`Glatiramoids are quite complex mixtures of copolymers with
`significant and varied effects on the human immune system.
`The novel immunomodulatory mechanisms of action of
`
`glatiramoids and their potential to prevent neuronal damage
`and promote neuroregeneration make these medications
`highly attractive for future drug development for the treatment
`of MS and perhaps other neurodegenerative disorders.
`GA, the first and best-studied glatiramoid, has been shown
`to reduce relapse rate, delay progression of disability and
`ameliorate MRI indices of disease in RRMS patients.
`Although GA polypeptide sequences are not entirely random,
`it is impossible to identify all of the active epitopes associ-
`ated with clinical efficacy. The consistent safety and efficacy
`of GA are dependent on its well-controlled proprietary man-
`ufacturing process. Likewise, new glatiramoids must be
`manufactured using a well-controlled process and have
`proven stability, and a battery of sensitive and validated
`quality assurance tests are necessary to ensure drug consistency
`and safety.
`Differences among glatiramoids are detectable using a
`variety of analytical, biological and immunological methods.
`It is impossible to predict whether or which of these differ-
`ences will produce unwanted pharmacologic effects. As
`experience with protiramer illustrates, even small differences
`in MM and in the immunological properties among glati-
`ramoid mixtures may have significant safety and efficacy
`implications. The benefit:risk ratio of new glatiramoids must
`be established in well-controlled preclinical, clinical and
`immunological studies.
`
`7. Expert opinion
`
`MS is a life-altering progressive irreversible disease for which
`there is no guarantee that once neurological damage is
`sustained, damage can be repaired or reversed. Given that an
`ineffective or unsafe product can lead to irrevocable neuro-
`logic and axonal damage, the manufacture, chemical compo-
`sition and clinical activity of any glatiramoid to be used for
`MS (or any other progressive CNS disorder, for that matter