`Ophthalmology (E.M.F.), and the Depart-
`ment ofNeuro|ogy and the Center for Im-
`munology (M.K.R.), University of Texas
`Southwestern Medical Center at Dallas,
`Dallas; and the Division of Neuropatho|-
`ogy,
`the Department of Pathology and
`Neurology, Albert Einstein College ofMed-
`icine, Bronx. N.Y. (C.S.R.).Address reprint
`requests to Dr. Frohman at the Depart-
`ment of Neurology, University of Texas
`Southwestern Medical Center at Dallas,
`5323 Harry Hines Blvd., Dallas, TX 75390,
`or at e||iot.frohman@utso uthweste rn.edu.
`
`N Engl] Med 2006;354:942-55.
`Copyright © 2006 Massachusetts Medical Society.
`
`The NEW ENGLAND JOURNAL of MEDICINE
`
`I
`
`REVIEW ARTICLE
`
`MEDICAL PROGRESS
`
`I
`
`Multiple Sclerosis — The Plaque
`and Its Pathogenesis
`
`Elliot M. Frohman, M.D., Ph.D., Michael K. Racke, M.D.,
`and Cedric S. Raine, Ph.D., D.Sc.
`
`UBSTANTIAL ADVANCES HAVE OCCURRED IN THE UNDERSTANDING or
`
`some of the central mechanisms underlying the inflammation, demyelination,
`and neurodegeneration that occur in multiple sclerosis since the topic was
`last reviewed in the journal.‘ Accordingly, the available clinical strategies for the
`management of the disease have widened (Table 1).‘ However, the treatment options
`for the disease are most effective during the relapsing—remitting phase (relapsing-
`remitting multiple sclerosis), which is characterized by clinical exacerbations, in-
`flammation, and evidence of plaques within the brain and spinal cord on magnetic
`resonance imaging (MRI). Less understood are factors that promote the transition
`from relapsing—remitting multiple sclerosis to treatment-resistant secondary pro-
`gressive multiple sclerosis. Evidence now suggests that neurodegenerative mecha-
`nisms within the disease plaques constitute the pathologic substrate for the latter
`disabling phase.“ Effector mechanisms that underlie the relapsing inflammatory
`and the progressive neurodegenerative phases of multiple sclerosis appear to be
`distinctly different.
`This review focuses on the current knowledge of the pathogenesis of the inflam-
`matory and neurodegenerative elements of the multiple sclerosis plaque.
`
`EVOLUTION OF THE MULTIPLE SCLEROSIS PLAQUE
`
`A central mission in multiple sclerosis research has been to determine the sequence
`of events underlying the development of the inflammatory plaque. It is generally held
`that this histopathological hallmark originates from a breach in the integrity ofthe
`blood—brain barrier in a person who is genetically predisposed to the disease. One
`hypothesis suggests that some forms of systemic infection may cause the up-regula-
`tion ofadhesion molecules on the endothelium of the brain and spinal cord, allowing
`leukocytes to home to and traverse vessel walls to enter the normally immunologi-
`cally privileged central nervous system. If lymphocytes programmed to recognize
`myelin antigen exist within the cell infiltrate, they may trigger a cascade of events
`resulting in the formation of an acute inflammatory, demyelinating lesion.‘ These
`lesions typically develop in white matter, where the primary targets are the myelin
`sheath and the myelinating cell, the oligodendrocyte (Fig. 1). However, gray-matter
`lesions, in which the primary target is also myelin, are known to occur.’
`
`CELLS INVOLVED IN THE PATHOGENESIS OF THE MULTIPLE
`
`SCLEROSIS PLAQUE
`
`T CELLS
`
`Studies of animal models demonstrating that autoreactive T cells (CD4+ or CD8+)
`can result in inflammatory demyelination of the central nervous system support the
`
`942
`
`N ENGLJ MED 354:9 WWW.NEjM.ORG MARCH 2, 2oo6
`
`AMNEAL
`The New England Journal of Medicine
`Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on November 12, 2014. For personal use only. No other uses without permission.
`Copyright © 2006 Massachusetts Medical Society. All rights reserved.
`
`EXHIBIT NO. 1039 Page 1
`
` AMNEAL
`
`
`
`medical progress
`
`Table 1. Treatment Options for Multiple Sclerosis.*
`
`Status
`
`Treatment
`
`Suggested Mechanism of Action
`
`Approved by the
`Food and Drug
`Administration
`
`Interferon beta
`
`Inhibits adhesion
`Inhibits synthesis and transport of MMPs
`Blocks antigen presen tation
`
`Glatiramer acetate
`
`Increases regulatory T cells
`Suppresses inflammatory cytokines
`Blocks antigen presen tation
`
`Mitoxantrone
`
`Reduces Th1 cytokines
`Eliminates lymphocytes
`
`Uses and Range
`of Effects
`
`Forms of Multiple
` Sclerosis Affected
`
`Treatment of relapses
`Slows progression
`Reduces lesions seen on
`MRI and brain atrophy
`Potential cognitive benefit
`
`Treatment of relapses
`Reduces lesions seen
`on MRI
`
`Treatment of relapses
`Reduces lesions seen
`on MRI
`Slows progression
`
`Relapsing
`
`Relapsing–remitting
`
`Relapsing–remitting
`Secondary progressive
`Progressive relapsing
`
`Possible adjunctive
`therapy
`
`Corticosteroids (in-
`travenous or oral
`formulations)
`
`Inhibit synthesis and transport of MMPs
`Alter cytokine profile
`Reduce CNS edema
`
`Treatment and prevention
`of relapses
`
`Relapsing
`
`Azathioprine
`
`Methotrexate
`
`Inhibits purine synthesis, affecting B
`cells, T cells, and macrophages
`
`Treatment of relapses
`Slows progression
`
`Relapsing–remitting
`Secondary progressive
`
`Acts as folate antagonist, affecting
`DNA synthesis in immune cells
`
`Slows progression
`
`Secondary progressive
`
`Plasma exchange
`
`Removes deleterious antibodies
`
`Treatment of relapse
`
`Intravenous immune
`globulin
`
`Has antiidiotypic effects
`Blocks Fc receptors
`Alters cytokine profile
`
`Treatment and prevention
`of relapses
`
`Relapsing
`
`Relapsing
`
`* This table is adapted from Goodin et al.2 MMPs denotes matrix metalloproteinases, MRI magnetic resonance imaging, Th1 type 1 helper
`T cells, and CNS central nervous system. Natalizumab had been approved by the FDA for treatment of multiple sclerosis but was withdrawn
`from the market in February 2005, to allow assessment of the risk of progressive multifocal leukoencephalopathy.
`
`theory that multiple sclerosis is an immune-
` mediated disorder involving one or more antigens
`located in the myelin of the central nervous sys-
`tem.8-10 Patients with multiple sclerosis and healthy
`persons appear to have similar numbers of T cells
`in peripheral blood that react to myelin. Never-
`theless, these two groups have substantial quali-
`tative differences in responses mediated by cir-
`culating mononuclear-cell populations (B cells,
`T cells, and macrophages). Myelin-reactive T cells
`from patients with multiple sclerosis exhibit a
`memory or activated phenotype, whereas these
`same antigen-specific cells in healthy persons
`appear to have a naive phenotype.11,12 Marked
`differences in the cytokines secreted and the spe-
`cific chemokine receptors expressed suggest that
`myelin-reactive T cells from patients with multiple
`sclerosis are relatively more inflammatory.13,14
`Further, myelin-specific CD8+ T cells appear to be
`more abundant in patients with relapsing multi-
`ple sclerosis than in healthy persons or in those
`with secondary progressive disease.13,15
`Perhaps the most convincing evidence that
`myelin-reactive T cells lead to inflammatory de-
`
`myelination came from a clinical trial in which
`an altered peptide ligand was used as a putative
`disease-modifying treatment in patients with
`multiple sclerosis.16 In this study, either clinical
`exacerbations or an increase in disease activity,
`as measured by MRI, unexpectedly developed in
`several patients treated with the ligand (a peptide
`developed to stimulate autoreactive T cells and
`render them inactive). These changes coincided
`with marked increases in T cells responding to a
`specific component of myelin basic protein (sig-
`nifying immune-cell activation rather than inac-
`tivation). In contrast, in another study, a lower
`dose of this peptide ligand actually reduced evi-
`dence of disease activity on MRI.17 This treatment
`strategy is currently being studied in a phase 2
`clinical trial.
`The cytokine-producing phenotype of myelin-
`specific T cells determines the ability of these
`cells to cause inflammation in the central ner-
`vous system.13 Organ-specific autoimmune dis-
`eases such as multiple sclerosis are thought to
`be mediated by type 1 helper T cells (Th1) that
`produce interferon-γ.9 Abundant data also sug-
`
`n engl j med 354;9 www.nejm.org march 2, 2006
`
`943
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on November 12, 2014. For personal use only. No other uses without permission.
`
` Copyright © 2006 Massachusetts Medical Society. All rights reserved.
`
` EXHIBIT NO. 1039 Page 2
`
` AMNEAL
`
`
`
`The NEW ENGLAND IOURNAL of MEDICINE
`
`interleu-
`of immunoregulatory cytokines (e.g.,
`kin-4 and interleukin-5). Myelin-reactive T cells
`from patients with multiple sclerosis produce cy-
`tokines more consistent with a Th1-mediated re-
`
`sponse, whereas myelin-reactive T cells from
`healthy persons are more likely to produce cyto-
`kines that characterize aTh2-mediated response.“
`Cytokines such as interleukin-12 and type 1 inter-
`ferons such as interferon-[3 can activate the tran-
`scription factor Stat-4 in human T cells, thus caus-
`ing the cells to differentiate into pathogenic Th1
`lymphocytes.” Interferon beta, which has been
`used to treat patients with multiple sclerosis (‘Table
`1), was thought to cause a shift from a Th1-medi-
`ated to a Th2-mediated response.” However, mi-
`croarray studies indicated that a number of genes
`in patients with multiple sclerosis that are up-
`regulated by this cytokine are associated with
`differentiation into Th1 rather than Th2 lympho-
`cytes, suggesting that such a shift may not be the
`mechanism of action of interferon beta.“
`
`Certain members of the interleukin-12 family
`of proteins probably have a role in the regulation
`of T-cell responses that have potential relevance
`to multiple sclerosis.” In experimental models of
`inflammatory demyelination, such as experimen-
`tal autoimmune encephalomyelitis in mice, the
`likelihood of disease development depends on
`which interleukins are functional. For example,
`experimental autoimmune encephalomyelitis did
`not develop in mice deficient in both interleukin-12
`and interleukin-23, but severe disease developed
`in animals with a deficiency of interleukin-12
`alone. Other studies indicate that interleukin-23
`
`probably has an essential role in brain inflamma-
`tion.” For instance, interleukin-23-deficient mice
`are resistant to experimental autoimmune en-
`cephalomyelitis but have a normal Th1 response.“
`Such studies in mice may be directly relevant to
`patients with multiple sclerosis. Interleukin-23
`causes T cells to produce interleukin-17, which
`some investigators believe is the chief determi-
`nant of brain inflammation, rather than inter-
`feron-y. Recent microarray studies of lesions of
`multiple sclerosis from patients demonstrated in-
`creased expression of interleukin-17, suggesting
`that it may be an important factor in the devel-
`opment of inflammatory demyelination.” Stud-
`ies of experimental autoimmune encephalomy-
`elitis in mice have recently shown that the T-bet
`and Stat-4 (necessary for Th1 differentiation)
`transcription factors are important in the differ-
`entiation of autoimmune T cells.2°‘23 Studies in
`
`Figure 1. Cross Section of white-Matter Lesions Target-
`ing the Myelin Sheath and Oligodendrocytes.
`In Panel A. light microscopy reveals myelin sheaths
`(dark blue rings) around axons in a cross section of
`myelinated white matter (toluidine blue). Two darker-
`staining oligodendrocytes, the cells that make and
`maintain myelin, lie to the right of center (arrow). In
`Panel B, an electron micrograph reveals the myelin
`sheath in cross section to be a spirally wrapped mem-
`brane beginning in the lower left as an outer (o|igoden-
`droglial) “tongue" of cytoplasm and spiraling counter-
`clockwise to terminate at an inner tongue inside the
`myelin sheath to the right of the axon. Microtubules
`and neurofilaments can be seen cut in cross section
`
`within the axoplasm.
`
`gest that inflammatory immune responses or
`delayed hypersensitivity responses are primarily
`mediated by inflammatory Th1 cells, which pro-
`duce lymphotoxin and interferon-'y, but
`little
`interleukin-4.“ Alternatively, CD4+ type 2 helper
`T cells (Th2) represent an antiinflammatory popu-
`lation oflymphocytes that produce large amounts
`
`944
`
`N man.) MED 354:9 WWW.NEjM.ORG MARCH 2, 2oo6
`
`AMNEAL
`The New England Journal of Medicine
`Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on November 12, 2014. For personal use only. No other uses without permission.
`Copyright O 2006 Massachusetts Medical Society. All rights reserved.
`
`EXHIBIT NO. 1039 Page 3
`
` AMNEAL
`
`
`
`MEDICAL PROGRESS
`
`humans are now needed to determine whether
`
`similar transcriptional programs determine the
`mechanism underlying the pathogenic potential
`of rnyelin-reactive T cells in multiple sclerosis.
`
`B CELLS
`
`It has long been recognized that intrathecal syn-
`thesis ofimmunoglobulins is increased in patients
`with multiple sclerosis, as evidenced by the pres-
`ence of oligoclonal bands on agarose-gel electro-
`phoresis and an increased IgG index or synthesis
`rate. Many studies have suggested that these an-
`tibodies recognize myelin antigens, but only
`recently has it become possible to characterize
`the antibody response on a molecular level in the
`cerebrospinal fluid ofpatients with multiple scle-
`rosis. Perhaps not surprisingly, the demonstration
`in the cerebrospinal fluid of B-cell proliferation
`and increased mutations in B-cell receptors, a pro-
`cess called somatic hypermutation, suggest that
`a B-cell response to a specific antigen is occur-
`ring in the central nervous system, whereas cor-
`responding clones are absent from the peripheral
`circulation.1°»3° Examination of these B-cell clones
`
`also indicated that some B cells had undergone
`a process called receptor revision, or editing, in
`which these cells recognize the body’s misguided
`capability to manufacture autoantibodies and sub-
`
`sequently remove this capacity.” Investigations
`are now under way to determine which specific
`central nervous system antigens are recognized by
`the autoantibodies generated by clonally expand-
`ed B cells in patients with multiple sclerosis. The
`observed overexpression ofimmunoglobulin genes
`and Fc receptors in lesions of this disease sug-
`gests that targeting the B-cell component of the
`immune response (e.g., with rituximab) may rep-
`resent an attractive therapeutic strategy (Table 2).
`
`OTHER IMMUNE CELLS
`
`It is likely that still other types ofcells play a role
`in the pathogenesis of multiple sclerosis. For ex-
`ample, regulatory cells, such as CD4+ICD25+ and
`CD8+ regulatory T cells, appear to be deficient in
`patients with this disease.”-33 Glatiramer acetate,
`a treatment for multiple sclerosis that may in-
`crease the numbers of these regulatory cells, may
`provide a means of reconstituting tolerance to
`self-antigens (Table 1).
`
`DISEASE INITIATION
`AND PATHOGENESIS
`
`There is substantial evidence to support the hy-
`pothesis that genetics has an important role in a
`person’s susceptibility to multiple sclerosis, prob-
`
`Table 2. Neuroprotective and Rstorative Strategies for Multiple Sclerosis.*
`
`Strategy
`
`Rationale or Mechanism
`
`Preliminary Observations
`
`Combinations of approved agents
`
`Targets multiple injury mechanisms
`
`Rituximab
`
`Depletes B cells
`
`Chemokine-receptor antagonists
`
`Reduce entry oflymphocytes into CNS
`
`Evidence of reduced activity on MRI
`Reduced relapses
`
`Clinical trials under way
`
`Clinical trials under way
`
`Riluzole
`
`Phenytoin and flecainide
`
`Blocks N-methyl-D-aspartate and sodium channels Reduced spinal-cord atrophy
`Reduced number ofhypointense lesions
`on T,-weighted MRI
`Neuroprotective in animals
`Clinical trials under way
`
`Block sodium channels
`
`Blockers of neurite outgrowth inhibitor
`
`Promote axonal sprouting
`
`Studies in animals under way
`
`Blodters of NG2, LINGO-1, Notch, andjagged
`
`Promote oligodendrocyte differentiation
`
`Studies in animals under way
`
`Activation of oligodendrite transcription
`factor 1
`Stem cells
`
`Promotes oligodend rocyte differentiation
`
`Under development
`
`Initiate myelin repair
`
`Established efficacy in animal models
`Early trials in humans under way
`
`Growth factors
`
`Antiapoptosis factors
`
`Promote neuronal survival
`
`Under development
`
`Promote survival of neurons and oligodendrocytes
`
`Studies in animals under way
`
`* CNS denotes central nervous system, NG2 neuron-glia antigen 2, and LINGO-1 leucine-rich—repeat and immunoglobulin-domain—contain-
`ing neurite outgrowth inhibitor receptor—interacting protein 1.
`
`N ENGLJ MED 354:9 WWW.NE}M.ORG MARCH 2, 2006
`
`945
`
`AMNEAL
`The New England Journal of Medicine
`Downloaded fiom nejm.org at UC SHARED JOURNAL COLLECTION on November 12, 2014. For personal use only. No other uses without permission.
`Copyright © 2006 Massachusetts Medical Society. All rights reserved.
`
`EXHIBIT NO. 1039 Page 4
`
` AMNEAL
`
`
`
`T h e n e w e n g l a n d j o u r n a l o f m e d i c i n e
`
`ably in conjunction with environmental factors.
`Although some investigators argue for a direct
`causal link between various infectious agents and
`this disorder, such agents may merely provide the
`appropriate milieu for the development of an auto-
`reactive immune response directed against central
`nervous system myelin. Recent work in experi-
`mental autoimmune encephalomyelitis has fo-
`cused on pathogens that can stimulate toll-like
`receptors, highly conserved receptors that rec-
`ognize pathogen-associated molecular patterns.
`These patterns are important for the initiation of
`disease and the production of interleukins, spe-
`cifically interleukin-12 and interleukin-23, which
`lead to the differentiation of T cells into autore-
`active effectors.34,35 Infectious agents may also
`have a role in the central mechanism that culmi-
`nates in the interaction between T cells and the
`cerebrovascular endothelium by up-regulating ad-
`hesion molecules important for immune-cell re-
`cruitment into the central nervous system.36
`Studies of experimental autoimmune enceph-
`alomyelitis in mice demonstrated the importance
`of T-cell expression of a family of cell-surface re-
`ceptors (the integrins) that promote adhesion and
`transport mechanisms. Such studies led to the
`development of a therapeutic antagonist of inte-
`grin, natalizumab, a monoclonal antibody specifi-
`cally against α4 integrin.37 This agent significantly
`reduced both clinical relapses and the formation
`of gadolinium-enhancing lesions in patients with
`multiple sclerosis.38 Despite its early promise, the
`development of progressive multifocal leukoen-
`cephalopathy in a few patients receiving natali-
`zumab in combination with interferon, or with
`azathioprine and infliximab, result ed in its with-
`drawal from the market and a halting of all clinical
`trials in February 2005; whether it will return to
`the market is unknown as of January 2006.39-41
`These observations underscore the principle that
`strategies interfering with the recruitment of leu-
`kocytes in the pathogenesis of multiple sclerosis
`may also interfere with routine immunosurveil-
`lance functions of the central nervous system.
`Several additional targets for potential study
`and therapeutic intervention have been identified
`with the use of microarray techniques. For in-
`stance, this approach led to the discovery that
`osteopontin was overexpressed in multiple scle-
`rosis lesions and subsequently to the finding that
`it has an important role in the progression of ex-
`perimental autoimmune encephalomyelitis.42
`
`Pathogenesis of Multiple Sclerosis
`Lesions Revisited
`
`In the light of the current consensus that the
`pathogenesis of the lesions of multiple sclerosis
`is heterogeneous, it is not surprising that no sin-
`gle predominant mechanism for this disease has
`emerged. Indeed, with a condition that includes
`fulminant as well as chronic forms with such a
`wide-ranging phenotype, multiple pathogenetic
`mechanisms have been proposed.43 In fact, the
`pattern of the lesions appears to be totally unpre-
`dictable; both acute and chronic cases have old
`as well as new lesions, illustrating the dynamic
`nature of the disease process. Regardless of this
`innate variability, the end-point chronic silent le-
`sion (without active inflammation) is a constant
`and pathognomonic feature of multiple sclerosis.
`
`Neuropathology
`The histologic features of lesions of acute multi-
`ple sclerosis (Fig. 2A, 2B, and 2C) include an in-
`distinct margin, hypercellularity, intense peri-
`vascular infiltration by small lymphocytes (Fig.
`2D and 2E), parenchymal edema, loss of myelin
`and oligodendrocytes, widespread axonal dam-
`age (Fig. 2F and 2G), plasma cells, myelin-laden
`macrophages, hypertrophic astrocytes, and little
`or no astroglial scarring. Demyelination in acute
`lesions may be due to an antimyelin antibody–
`mediated phenomenon in which normal lamellar
`myelin is transformed into vesicular networks
`(Fig. 2H and 2I), coated with antimyelin oligo-
`dendrocyte glycoprotein or antimyelin basic pro-
`tein immunoglobulin, and phagocytosed in the
`presence of complement by local macrophages.44
`Remyelination is occasionally seen.
`Lesions of chronic active multiple sclerosis
`display a sharp edge; along the edge are perivas-
`cular cuffs of infiltrating cells, lipid-laden and
`myelin-laden macrophages, hypertrophic astro-
`cytes, and some degenerating axons, and demye-
`lination is occurring (Fig. 3A). In contrast to acute
`lesions, demyelination in chronic active lesions is
`associated with the deposition of immunoglobu-
`lin and the dissolution of myelin into droplets,
`which undergo phagocytosis once they become
`attached to macrophages.45,46 An increase in the
`number of oligodendrocytes and some remyelin-
`ation are not uncommon in chronic lesions. The
`centers of such lesions are hypocellular and con-
`tain naked axons embedded in a matrix of scar-
`ring (fibrous) astrocytes, lipid-laden macrophages,
`
`946
`
`n engl j med 354;9 www.nejm.org march 2, 2006
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on November 12, 2014. For personal use only. No other uses without permission.
`
` Copyright © 2006 Massachusetts Medical Society. All rights reserved.
`
` EXHIBIT NO. 1039 Page 5
`
` AMNEAL
`
`
`
`MEDICAL PROGRESS
`
`Figure 2. The Active Lesion of Multiple Sclerosis in Human Tissue (Panels A, B, C, D, F, G, H, and I) and Mouse Tissue (Panel E).
`In Panel A, lesions are present in the cerebral hemisphere in the frontal (upper) and parieto-occipital (lower) periventricular white mat-
`ter. The image on the left shows a proton-density image obtained by MRI, the center image shows a gross specimen from the same lev-
`el, and the image on the right shows a section from the same level stained with Luxol fast blue to reveal myelin and the demyelinated
`lesions. The margins of the lesions are rather irregular and indistinct, suggestive of ongoing activity, and some demyelinated areas are
`pale blue, consistent with the presence of remyelination. In Panel B, a section of temporal lobe stained with Luxol fast blue for myelin
`reveals a periventricular plaque extending to a cup-shaped zone of dernyelination beneath the sulcus to the lower right. Some areas
`around the margins of the lesion stain pale blue (e.g., adjacent to the affected sulcus), probably indicative of remyelination. V denotes
`ventricle. In Panel C (Luxol fast blue), a section of an acute lesion has an indistinct margin and numerous perivascular infiltrates
`(arrows). In Panel D (toluidine blue), a blood vessel, with red cells in the lumen of an acute lesion, is ringed by small lymphocytes —
`probably T cells; the surrounding parenchyma of the central nervous system has undergone demyelination. In Panel E, an electron
`micrograph from an actively dernyelinating lesion in the spinal cord of a mouse with acute experimental autoimmune encephalitis, an
`animal model of multiple sclerosis, shows a small lymphocyte, possibly a T cell, within the blood-vessel lumen (BV), adherent to the
`vascular endothelium, and another cell has almost traversed the endothelium through a gap (arrow), to enter the Vircl1ow—Robin space,
`and another to the right is within the perivascular space. The perivascular cuff of cells is separated from the central nervous system
`parenchyma below (myelinated nerve fibers are seen) by a layer of astroglial cells known as the glia limitations. In Panel F (toluidine blue
`stain), an area within an acute plaque is completely demyelinated and contains numerous transected, damaged axons (arrows) that form
`spheroids. In Panel G, an electron micrograph shows axonal spheroids; each is filled with accumulations of mitochondria, dense bodies.
`and neurofilaments. In Panel H (toluidine blue stain), a longitudinal section of the myelin sheath surrounding an axon at the margin of
`an acute lesion displays vacuolation and vesiculation; macrophages are present next to the degenerating myelin (arrows) and are filled
`with droplets of myelin. In Panel I, an electron micrograph shows an axon (A) within an acute lesion surrounded by a myelin sheath that
`has been transformed into a vesicular network. the result of interactions with antimyelin antibodies.’
`
`N ENGLJ MED 354;9 WWW.NEJM.ORG MARCH 2. 2006
`
`947
`
`AMNEAL
`The New England Journal of Medicine
`Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on November 12, 2014. For personal use only. No other uses without permission.
`Copyright 0 2006 Masachusetts Medical Society. All rights reserved
`
`EXHIBIT NO. 1039 Page 6
`
` AMNEAL
`
`
`
`The NEW ENGLAND JOURNAL of MEDICINE
`
`\-.
`
`l’ ".
`
`'x:,§;',‘(>'.-»
`
`*
`
`Figure 3. The Chronic Lesion of Multiple Sclerosis in Humans.
`In Panel A, an electron micrograph of a chronic active lesion shows a myelinated fiber undergoing demyelination.
`The arrow shows myelin droplets on the macrophage surface being internalized by the cell. The fiber is invested
`by a microglial cell, which is engaged in the phagocytosis of myelin droplets as they are divested from the myelin
`sheath. The end product of this process is shown in Panel B (toluidine blue stain).An area from a chronic silent gli-
`otic lesion is made up of astroglial scar tissue, in which intact demyelinated axons (light profiles) are embedded;
`mitochrondria can be seen within the axons; the smaller nuclei belong to microglial cells, but no oligodendrocytes
`are present. In Panel C, an electron micrograph with a field similar to that in Panel B shows large-diameter demye-
`linated axons (A) within the glial scar; an astroglial-cell body is at the upper right. In Panel D (toluidine blue stain),
`a biopsy specimen from a patient with secondary progressive multiple sclerosis shows an area of remyelination
`(shadow plaque) in which the myelin sheaths of many axons are disproportionately thin and oligodendrocytes (OL)
`are overabundant. These cells are probably oligodendroglial precursor cells recently recruited into the lesion. In
`Panel E, an electron micrograph shows remyelination; the myelin sheaths are thin in comparison to the diameters
`ofthe axons, and two oligodendrocytes are evident (OL). In Panel F (Luxol fast blue and periodic acid—SchifT),
`there is an abrupt transition at the edge of the chronic multiple sclerosis lesion. The myelin internodes (blue) termi-
`nate sharply at the demyelinated plaque. oligodendrocytes are present (arrows) up to the edge of the lesion. but
`not within the lesion. Rod cells (microglia) are lined up along the boundary. A denotes axon, and As astrocyte.
`
`948
`
`N moi.) MED 354:9 www.N:)M.onc MARCH 1, 2006
`
`AMNEAL
`The New England Journal of Medicine
`Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on November 12, 2014. For personal use only. No other uses without permission.
`Copyright 0 2006 Massachusetts Medical Society. All rights reserved
`
`EXHIBIT NO. 1039 Page 7
`
` AMNEAL
`
`
`
`medical progress
`
`a few infiltrating leukocytes, and virtually no
`oligodendrocytes. Lesions of chronic silent dis-
`ease display sharp edges, astroglial scar tissue
`(Fig. 3B and 3C), a reduced number of demyelin-
`ated axons, macrophages, and vessels with thick-
`ened (hyalinized) walls around which occasional
`leukocytes are seen; these lesions contain few or
`no oligodendrocytes.43
`
`Cl assification of Lesions —
`S tages or T ypes?
`
`The current classification schemes for the lesions
`of multiple sclerosis are under study, principally
`stimulated by a series of insightful categorizations
`by Lucchinetti and colleagues.47-49 The most quot-
`ed categorization from these investigators grouped
`active lesions into four types, I through IV, on the
`basis of the pathogenesis of the lesions.47 All four
`types of lesions are characterized by T-cell and
`macrophage-dominated inflammation. Type I is
`characterized by demyelination and macrophage-
`related products (e.g., tumor necrosis factor α).
`Type II is characterized by the presence of immu-
`noglobulin and complement. Type III lacks im-
`munoglobulin and complement, yet it shows early
`loss of myelin-associated glycoprotein and no re-
`myelination; the demyelination in type III has been
`attributed to oligodendrocyte dysfunction. Type IV
`is distinguished by apoptosis of oligodendrocytes
`through DNA fragmentation. Although this ap-
`proach has merit, concern has been expressed re-
`garding the details of the cases on which it is
`based and whether the central nervous system–
`biopsy specimens examined came from patients
`with typical multiple sclerosis. Furthermore, Bar-
`nett and Prineas50 have recently demonstrated that
`lesions from a given patient can, in fact, contain
`features of more than one category of lesions.
`This important observation underscores the fact
`that the development of a cogent and accurate
`lesion-classification scheme remains a work in
`progress.48,49
`
`A xonal Dysfunction
`and Channelopathy
`
`The oligodendrocyte–myelin–axon unit represents
`a unique structural and functional specialization
`within the central nervous system. Myelin not
`only increases the cross-sectional diameter of the
`nerve axon (Fig. 1A), which increases conduction
`
`velocity, but also contributes protectively and tro-
`phically to the health of the axon.51 Disruption of
`these relationships appears at the earliest stages
`of multiple sclerosis.3-5
`Axonal injury in multiple sclerosis appears to
`be partly explained by demyelination and the
`proliferation of abnormal expression of sodium
`channels localized within the membrane (Fig. 4).52
`In an attempt to reestablish normal conduction,
`there is an increased entry of sodium, slowing of
`nerve conduction, and potentially, even conduc-
`tion block. These processes appear to be followed
`by reversal of the sodium–calcium exchanger (i.e.,
`sodium efflux and calcium influx), which can
`trigger intracellular cascades of calcium-medi-
`ated injury, ultimately leading to neuronal degen-
`eration (Fig. 4). This hypothesis has been sup-
`ported by recent evidence that sodium-channel
`blockers such as flecainide and phenytoin pre-
`serve axons in mice with experimental auto-
`immune encephalomyelitis, thereby preserving
`physiologic function, as compared with that in
`untreated animals (Table 2).53,54
`It has been known since the time of Charcot
`that axonal injury is a pathological feature of the
`multiple sclerosis plaque. This important charac-
`teristic of the composition of the lesion has been
`long deemphasized, since the main pathological
`features appeared to be related to myelin. An
`analysis by Trapp et al.3 rekindled interest in the
`topic.4 Pathological changes in axons, detectable
`early in the process on the basis of the accumu-
`lation of amyloid-precursor protein,55 appear to
`be due to inflammation and are important in the
`evolution of the lesion (Fig. 2F and 2G); they are
`most conspicuous during the acute and progres-
`sive stages. Cumulative loss of axons, as a result
`of inflammatory demyelination and ultimately,
`transection, correlates with irreversible disability.
`The number of axons continues to dwindle
`throughout the course of the disease, and some
`old lesions display an axonal loss of more than
`80 percent.56
`Factors that have been associated with axonal
`damage include cytokines, nitric oxide, proteases,
`superoxides, CD8+ T cells, and glutamate excito-
`toxicity.57 Furthermore, such damage may even
`extend to nerve cells, mediated by microglia
`through cholesterol-breakdown products.58 In
`addition, MRI studies have correlated axonal
`loss with a reduction of N-acetyl aspartate (a
`substrate within axons) and hypointensity (gray
`
`n engl j med 354;9 www.nejm.org march 2, 2006
`
`949
`
`The New England Journal of Medicine
`
`Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on November 12, 2014. For pe