ARTICLE
`
`Multiple Sclerosis 2007; 13: 754 762
`
`Clinical response to glatiramer acetate correlates with
`modulation of IFN-g and IL-4 expression in multiple
`sclerosis
`
`RM Valenzuela1, K Costello2, M Chen2, A Said2, KP Johnson2 and S Dhib-Jalbut1
`
`Objective To determine whether glatiramer acetate (GA)-induced lymphoproliferation and IFN-g and
`IL-4 modulation correlate with the clinical response in multiple sclerosis (MS).
`Background GA therapy involves the induction of anti-inflammatory cytokine shifts. However, it is
`not known whether this response correlates with the clinical outcome.
`Methods Thirty-six relapsing-remitting (RR) MS patients were treated with GA for at least two years,
`and classified clinically as GA-responders (GA-R/22) or hypo/non-responders (GA-HR/NR/14).
`Proliferation of peripheral blood mononuclear cells (PBMC) to GA and Tetanus toxoid (TT), as well as
`IL-4 and IFN-g ELISPOT, were performed.
`Findings There was no difference in PBMC proliferation to GA or TT between GA-R and GA-HR/NR
`before and during treatment (P/0.05). The mean number of IFN-g ELISPOTS in unstimulated, TT
`and anti-CD3/CD28-stimulated PBMC was lower among GA-R (unstimulated: GA-R/10.19/6.21
`(n/22) versus GA-HR/NR/17.89/12.7 (n/14), P/0.04; TT-GA-R/12.29/4.06 (n/12) versus
`GA-HR/NR/26.89/21.0 (n/8), P/0.028; anti-CD-3/CD28 GA-R/217.39/140.4 (n/22) versus
`GA-HR/NR/368.59/170.1 (n/14), P/0.006). In contrast, the number of IL-4 ELISPOTS remained
`unchanged in the GA-R group, but was progressively reduced in the GA-HR/NR group during GA
`therapy (GA-HR/NR IL-4: pre-Rx: 599/34 versus 229/11 at 12 months (n/6), P/0.0429). The IL-4/
`IFN-g ratio in anti-CD3/CD28-stimulated PBMC was significantly higher among GA-R compared to
`GA-HR/NR (P/0.0474).
`Interpretation Lymphoproliferation to GA did not differentiate GA-R from GA-HR/NR. However,
`reduced IFN-g expression and stable IL-4 expression in anti-CD3/CD28-stimulated PBMC, and an
`increased IL-4/IFN-g ratio was associated with favorable clinical response. More data are needed to
`validate the prospective use of IL-4/IFN-g expression in PBMC as a biomarker of clinical response to
`GA for individual patients. Multiple Sclerosis 2007; 13: 754 762. http://msj.sagepub.com
`
`Key words: multiple sclerosis; glatiramer acetate; interferon-gamma; Interleukin-4
`
`Introduction
`
`The mechanism of action of glatiramer acetate (GA)
`in the treatment of multiple sclerosis (MS) is not
`fully elucidated. The molecule binds to HLA class II
`molecules associated with MS,
`including DR2
`and DR4, on antigen presenting cells [1 5]. Recent
`studies, both in experimental autoimmune ence-
`[6 8,12] and in humans
`phalomyelitis
`(EAE)
`[7,9 14], indicate that a likely in vivo mechanism
`
`of action of GA involves the induction of immu-
`nomodulatory regulatory T-helper type 2 (Th2)
`cells. The GA-specific T cells may exert
`their
`protective action systemically and by entering the
`CNS compartment, as evidenced by inhibition of
`new magnetic resonance imaging (MRI) gadoli-
`nium-enhancing lesions progressing into black
`holes [15], and the production of anti-inflamma-
`tory cytokines and brain-derived neurotrophic fac-
`tor (BDNF) [16], in response to cross-recognition of
`
`1 UMDNJ-Robert Wood Johnson Medical School, New Brunswick, NJ, USA
`2 School of Medicine, University of Maryland, Baltimore, MD, USA
`Author for correspondence: Suhayl Dhib-Jalbut, MD, Department of Neurology, 97 Paterson Street, Room 205, New
`Brunswick, NJ 08901, USA. E-mail: jalbutsu@umdnj.edu
`Received 24 March 2006; accepted 30 October 2006
`
`– 2007 SAGE Publications
`
`10.1177/1352458506074510
`
`Downloaded from
`
`msj.sagepub.com
`
` by guest on November 18, 2015
`
`Page 1 of 9
`
`YEDA EXHIBIT NO. 2124
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`myelin antigens (bystander suppression) [6,17]. We
`and others have demonstrated that GA treatment in
`MS results in the induction of GA-specific T cells
`with a predominant Th2 phenotype, both in re-
`sponse to GA and crossreactive myelin antigens
`[17,18]. These results strongly suggest that the
`mechanism of action of GA in MS involves the
`induction of crossreactive GA-specific T-cells with a
`predominant Th2 cytokine profile. Importantly,
`the induction of GA-specific Th2 cells was not
`universal
`in MS patients treated with the drug
`[19]. Since clinical response to GA varies among
`treated patients, this raises the possibility that the
`immunological effects of GA may correlate with the
`clinical response, and may distinguish responders
`from hypo or non-responders.
`Recently, Farina et al. [20], reported a triad of
`immune responses to GA that predict treated from
`untreated patients. This triad consisted of: (1) a
`decline in GA-induced proliferation; (2) positive IL-
`4 CD4 cells; and (3) positive IFNg CD8 cells in
`response to high dose of GA (100 mg/mL). Whether
`this triad correlates with the clinical response is
`not certain at present. In this study, we determined
`whether the lymphoproliferative response to GA
`and the GA-induced cytokine shifts in IL-4 and
`IFN-g correlate with the clinical response to GA.
`
`Materials and methods
`
`Subjects
`
`Clinical response to GA
`
`755
`
`Cells
`
`Approximately 60 cc of heparinised blood was
`obtained by venipuncture from each MS patient at
`different time points during treatment, as indicated
`in Table 1. In a subset of patients, blood samples
`were obtained before treatment, and at three, six,
`12 and 18 24 months during treatment. The study
`was approved by the University of Maryland In-
`stitutional Review Board, and informed consent
`was obtained from each patient enrolled in the
`study. Peripheral blood mononuclear cells (PBMC)
`were purified using a Ficoll-Hypaque gradient, as
`described in the supplier’s protocol (ICN Biomedi-
`cals Inc., OH, USA).
`
`Lymphoproliferation assay
`
`One /105 PBMC/well were seeded in a 96-well
`U-bottom microtiter plate together with 1/105
`irradiated autologus PBMC,
`in the presence of
`antigen in triplicate wells (glatiramer acetate at 40
`and 100 mg/mL, Teva Pharmaceutical Industries,
`Ltd., Israel). Tetanus toxoid (TT; 5 mg) obtained
`from Pasteur Merieux Connaught (North York,
`Ont., Canada) was used as a control antigen. After
`cells were cultured for 48 hours, 1 mCi 3H-thymin-
`dine was added. Cells were harvested 18 hours later
`using TOMTEC cell harvester (TOMTEC, Hamden,
`CT). Incorporated radioactivity was measured using
`a liquid Scintillation counter (Wallac MicroBeta
`Trilux, Perkin Elmer, Boston, MA).
`
`Thirty-six consecutive patients, diagnosed with
`relapsing-remitting (RR) MS and treated with GA
`for at least two years, were included in the study.
`Thirty-one patients were female with a mean age of
`44 years, and five were male with a mean age of 49
`years. Patients had at least one relapse during the
`year prior to initiation of GA therapy. All subjects
`were closely followed up at the University of Mary-
`land Center for MS. The clinical characteristics of
`the patients are presented in Table 1. After at least
`two years on GA therapy, patients were classified as
`GA-responders (R) (n/ 22) or hypo/non-responders
`(HR/NR) (n/ 14) based on criteria arbitrarily set for
`this study. We adopted a more conservative criteria
`for patient classification into GA-R and GA-HR/NR
`compared to those recently reported in the litera-
`ture [21,22]. A responder (R) is a patient with an
`annual relapse rate (ARR) B/0.5 and no evidence of
`disease progression, as measured by Expanded
`Disability Status Scale (EDSS). A HR/NR is a patient
`with an ARR /0.5 and/or with progression in the
`EDSS of at least 1 point sustained for six months.
`
`ELISPOT assay
`
`Human IFN-g (Th1 indicator) and IL-4 (Th2 indi-
`cator) were measured by the ELISPOT assay, accord-
`ing to the manufacturer’s protocol (BD Biosciences,
`San Diego, CA, USA). Briefly, ELISPOT plates (PVDF
`plate, Millipore Corp., MA, USA) were coated with
`100 mL/well of 5 mg/mL capture antibody at 48C
`overnight. The plates were then washed with PBS,
`and incubated with 200 mL/well of blocking solu-
`tion (culture medium) for 2 hours. PBMC (2/105
`cells/well for IFN-g and 4/105 cells/well for IL-4
`detection) and antigen (GA 40 and 100 mg/mL, TT 5
`mg/mL, CD3Ab (2.5 mg/mL)/CD28 Ab (1 mg/mL))
`was added and cultured overnight at 378C in a 5%
`CO2 incubator. Patients were tested in triplicate
`wells in response to each antigen. After overnight
`culture, the plates were washed twice with deio-
`nised water, and three times with PBS, then in-
`cubated with 100 mL of 2.0 mg/mL biotinylated
`antibody for 2 hours at room temperature. The
`plates were then washed three times with wash
`buffer (0.05% Tween-20 PBS), and incubated with
`
`http://msj.sagepub.com
`
`Multiple Sclerosis 2007; 13: 754 762
`
`Downloaded from
`
`msj.sagepub.com
`
` by guest on November 18, 2015
`
`Page 2 of 9
`
`YEDA EXHIBIT NO. 2124
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`756
`
`RM Valenzuela et al.
`
`Table 1 Characteristics of MS patients and time points during treatment when blood samples were obtained
`
`Time sample analysed (months)
`
`EDSS
`
`ARR
`
`ID No.
`
`Age/sex
`
`Disease duration (years)
`
`Rx duration (months)
`
`0
`
`3
`
`42
`39
`43
`44
`45
`46
`47
`48
`49
`9
`50
`52
`7
`53
`54
`55
`56
`57
`58
`59
`1
`2
`3
`4
`6
`7a
`8
`10
`11
`12
`14
`17
`18
`21
`22
`24
`
`29/F
`67/F
`36/F
`41/F
`39/F
`40/F
`37/M
`64/F
`46/F
`49/F
`44/F
`51/F
`52/F
`40/M
`46/F
`41/F
`22/F
`53/F
`48/F
`43/F
`32/F
`45/M
`45/F
`32/F
`60/F
`53/F
`53/F
`43/F
`52/F
`40/F
`51/F
`60/M
`29/F
`49/F
`63/M
`36/F
`
`4
`37
`5
`Uncertain
`18
`19
`13
`25
`12
`6
`15
`20
`5
`3
`5
`5
`2
`32
`5
`Uncertain
`5
`6
`23
`37
`5
`Uncertain
`18
`13
`25
`12
`21
`5
`5
`5
`6
`5
`
`35
`40
`55
`27
`60
`24
`56
`27
`24
`24
`45
`38
`38
`24
`30
`25
`24
`57
`40
`41
`36
`24
`24
`26
`48
`24
`29
`36
`24
`24
`24
`24
`24
`24
`27
`24
`
`*
`*
`
`*
`
`*
`
`*
`
`*
`*
`*
`
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`
`6
`
`*
`
`*
`
`*
`
`*
`
`*
`
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`
`12
`
`18 24
`
`Pre-Rx On-Rx
`
`Pre-Rx On-Rx
`
`Patient status
`
`1
`1.5
`0
`1.5
`6.5
`2
`1.5
`2
`2
`2.5
`1.5
`3
`1.5
`0
`2.5
`2.5
`2
`6
`0
`1.5
`1.5
`2
`1.5
`3.5
`7
`3.5
`2
`1.5
`2
`1.5
`1.5
`2.5
`1.5
`2
`2.5
`1.5
`
`0
`1.5
`0
`1
`6.5
`2
`1
`1.5
`1.5
`4
`1.5
`4
`2.5
`1.5
`2.5
`2.0
`0
`7
`1
`1.5
`1.5
`2.5
`1.5
`1.5
`1.5
`7
`2
`2.5
`6
`1
`2.5
`2
`1.5
`2
`2.5
`1.5
`
`1.5
`2
`1.5
`1.5
`3
`2
`1.5
`1
`1.5
`2
`1.5
`1.5
`1
`0.5
`1.5
`0.5
`2
`1
`1
`0.9
`2.5
`0.5
`2
`1.5
`3
`0.5
`1.5
`1
`1.5
`0.5
`1.5
`1
`1.5
`1
`2
`2
`
`0.3
`0
`0.9
`0.4
`0
`0
`0
`0
`0
`1.5
`0
`1.3
`0
`0
`0.4
`0
`0
`0.4
`0.6
`1.5
`0
`0.99
`0
`0.46
`0
`0.49
`0
`0.99
`0
`0
`1.33
`1.2
`0
`0
`0
`0
`
`R
`R
`NR
`R
`R
`R
`R
`R
`R
`NR
`R
`NR
`NR
`NR
`R
`R
`R
`NR
`NR
`NR
`R
`NR
`R
`R
`R
`NR
`R
`NR
`NR
`R
`NR
`NR
`R
`R
`R
`R
`
`*
`
`*
`
`*
`*
`*
`
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`
`*
`
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`*
`
`Rx: treatment; pre-Rx, pre-treatment; post-Rx, post-treatment; ARR, annual relapse rate; R, responder; HR/NR, hypo/non-responder; LV, last visit.
`*Blood sample obtained.
`
`Multiple Sclerosis 2007; 13: 754 762
`
`http://msj.sagepub.com
`
`Downloaded from
`
`msj.sagepub.com
`
` by guest on November 18, 2015
`
`Page 3 of 9
`
`YEDA EXHIBIT NO. 2124
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`100 mL of 1:100 dilution of Avidin-HRP for 1 hour
`at room temperature, and washed three times with
`wash buffer. Finally, 100 mL of AEC (3-amino
`-9-ethyl-carbazole) substrate was added, and the
`reaction was stopped by washing plates with dis-
`tilled water. Spots were counted and analysed by
`CTL Analysers LLC ELISPOT Plate Reader (Cleve-
`land, OH, USA).
`
`Statistical analysis
`
`A software package (Graphpad, PrismTM) was used in
`the statistical analysis. The differences in cytokine
`spots between responder and hypo/non-responder
`were compared using the Student’s t -test and
`Bonferroni’s multiple comparison test. A P value
`B/0.05 was considered significant. Although only
`2/105 cells/well were utilised for the IFN-g ELI-
`SPOT, the values were adjusted to 4/105 cells/well
`for both the IL-4 and IFN-g spots to account for the
`corrected dilution factor.
`
`Results
`
`Classification of clinical responders and hypo/non-
`responders
`
`The patients were classified as GA-responders or
`hypo/non-responders after being treated with GA
`for at least 24 months, based on their annual
`relapse rate and progression in disability while on
`treatment (Table 1). The mean treatment duration
`was 34 months. The GA-responder group had a
`mean ARR of 1.68 and a mean EDSS of 2.31 before
`treatment compared to 0.07 ARR and an EDSS of
`1.70 during treatment. The hypo/non-responder
`group had a mean ARR of 1.10 and a mean EDSS
`of 1.96 before treatment compared to 0.75 ARR and
`an EDSS of 3 during GA therapy. Overall, a 96%
`reduction in the ARR, and a 0.61 point decrease in
`the EDSS occurred in the GA-R, whereas a 31%
`reduction in the ARR, and a 1.14 point increase
`in the EDSS occurred in the GA-HR/NR during
`treatment.
`
`Clinical response to GA
`
`757
`
`anti-CD3/CD28-stimulated PBMC, was significantly
`lower among GA-R compared to GA-HR/NR
`(unstimulated: GA-R/10.19/8.21 (n/ 22) versus
`GA-HR/NR/17.89/12.7 (n/ 14), P/ 0.03; TT-GA-
`R/12.29/4.06 (n/ 12) versus GA-HR/NR/26.89/
`21.0 (n/ 8), P/ 0.028; anti-CD-3/CD28 GA-R/
`217.39/140.4 (n/ 22) versus GA-HR/NR/368.59/
`170.1 (n/ 14), P/ 0.006). The mean number of
`IL-4 spots
`in anti-CD3/CD28-stimulated PBMC
`was not significantly different between GA-R and
`GA-HR/NR (GA-R/92.09/84.9 (n/ 22) versus GA-
`HR/NR/80.89/80.6 (n/ 14), P/ 0.05). The IL-4/
`IFN-g ratio in anti-CD3/CD28-stimulated PBMC
`was significantly higher among GA-R compared to
`GA-HR/NR (0.939/1.02 versus 0.349/0.40, respec-
`tively; P/ 0.0474).
`In 16 of the 36 patients (10 GA-R and six GA-HR/
`NR) pre-treatment and during treatment IL-4 and
`IFN-g ELISPOTS were determined. Cytokine expres-
`sion studies in anti-CD3/CD28-stimulated PBMC
`were performed at serial time points, ranging from
`three to 24 months for this sub-group (Figure 1).
`Mean IL-4 spots did not change significantly during
`treatment compared to pre-treatment levels in the
`GA-R group (Figure 1a). In contrast, mean IL-4 spots
`decreased from pre-treatment levels in the GA-HR/
`NR group beginning at three months of GA ther-
`apy, and the decrease became statistically signifi-
`cant at 12 months and beyond (Figure 1b). The
`mean IFN-g spots steadily decreased during treat-
`ment
`in the GA-R group, beginning at
`three
`months, and this decrease became statistically
`significant at ]/12 months (Figure 1c). Although a
`decrease in IFN-g was observed in the GA-HR/NR
`group, this decrease was delayed (]/12 months),
`and did not reach statistical significance (Figure
`1d). The mean IL-4/IFN-g ratio increased steadily
`during treatment in the GA-R group, and became
`statistically significant at six months and later
`the mean IL-4/IFN-g
`(Figure 1e).
`In contrast,
`ratio remained unchanged at all time points in
`the GA-HR/NR group (Figure 1f).
`
`Cytokine shifts
`
`Correlation of IL-4/IFNg ratio with the clinical
`response
`
`Initially, expression of IFN-g and IL-4 was examined
`by ELISPOT in PBMC obtained ex vivo from GA-
`treated MS patients at random time points during
`treatment
`(Table 2). There was no significant
`difference in cytokine production between GA-R
`and GA-HR/NR in response to GA stimulation
`at 40 and 100 mg/mL. The mean number of
`IFN-g spots in unstimulated TT (5 mg/mL) and
`
`Linear regression analysis using Spearman correla-
`tion showed no significant correlation between the
`number of IL-4 spots and ARR (Figure 2a), while a
`significant positive correlation was observed be-
`tween the number of IFN-g spots and ARR (P/
`0.040) (Figure 2b). A significant negative correla-
`tion between ARR and IL-4/IFN-g ratio was observed
`(P/ 0.041) (Figure 2c).
`
`http://msj.sagepub.com
`
`Multiple Sclerosis 2007; 13: 754 762
`
`Downloaded from
`
`msj.sagepub.com
`
` by guest on November 18, 2015
`
`Page 4 of 9
`
`YEDA EXHIBIT NO. 2124
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`758
`
`RM Valenzuela et al.
`
`Table 2 IL-4 and IFN-g expression by Elispot obtained from MS patients treated with GA
`
`R
`
`HR/NR
`
`R
`
`HR/NR
`
`R
`
`HR/NR
`
`R
`
`HR/NR
`
`No. Ag
`
`GA (40 mg/mL)
`
`GA (100 mg/mL)
`
`TT (5 mg/mL)
`
`CD3/CD28
`(2.5 mg/1 mg/mL)
`
`5.409/3.90
`(n/22)
`9.279/5.54
`(n/12)
`11.319/6.45
`(n/12)
`10.969/5.79
`(n/12)
`92.09/84.95
`(n/22)
`
`4.009/2.56
`(n/14)
`8.069/2.84
`(n/8)
`11.899/5.14
`(n/8)
`8.339/3.23
`(n/8)
`80.89/80.6
`(n/14)
`
`10.19/6.21
`(n/22)
`10.839/7.2
`(n/12)
`14.09/8.04
`(n/12)
`12.29/4.06
`(n/12)
`217.39/140
`(n/22)
`
`17.89/12.7
`(n/14)
`13.09/9.53
`(n/8)
`16.59/9.41
`(n/8)
`26.89/21.0
`(n/8)
`368.59/170
`(n/14)
`
`0.59/0.60
`
`0.259/0.40 0.25 0.20
`
`1.779/2.15
`
`0.6589/0.29 0.73 0.60
`
`0.9679/0.57
`(n/12)
`1.359/0.94
`
`0.829/0.39 0.85 0.74
`
`0.769/0.27 1.12 0.41
`
`0.939/1.02
`
`0.3419/0.4
`
`0.39 0.15
`
`R, GA-responder; HR/NR, GA-hypo/non-responder; n, number of patients.
`*Mean number of spots per 4/105 cells/well.
`The mean IL-4 and IFN-g spots shown in columns 2 and 3 were calculated as follows: first, the mean number of spots during the three
`to 24 month-treatment period was determined for each individual. Subsequently, the average of the mean number of spots for each
`group was calculated. The IL-4/IFN-g ratio shown in column 4 represents the mean of the ratios for each group. Note that the ratio in
`column 4 is not the ratio of IL-4 and IFN-g means shown in column 2 and 3.
`
`Lymphoproliferative response
`
`The ex vivo PBMC proliferative response to GA and
`TT are presented in Figure 3 as DCPM (mean cpm
`in antigen stimulated minus mean cpm in unsti-
`mulated wells) and as stimulation indexes (SI/
`mean CPM in stimulated wells/mean CPM in
`unstimulated wells). Overall, there was an initial
`increase in the proliferative response to GA fol-
`lowed by a decline after six months of treatment,
`which was statistically significant at 12 months
`(P/0.04) and 18 24 months (P/0.01). An initial
`decrease in the proliferative response to TT was
`observed at three months (P/0.007), but this
`response recovered after six months, and was not
`significantly different from pre-treatment values.
`There was no difference in PBMC proliferation to
`GA or TT between GA-R and GA-HR/NR before and
`during treatment at three, six, 12 months or two
`years (P values at all time points were /0.05).
`
`Discussion
`
`Previous studies suggested that a decrease in the
`lymphoproliferative response to GA and changes in
`IL-4 and IFN-g expression are associated with GA
`treatment in MS [13,18,21]. The goal of this study
`was to determine whether the lymphoproliferative
`response to GA and GA-induced cytokine shifts
`correlate with the clinical response. While GA
`treatment initially induces a strong lymphoproli-
`ferative response during the first three months, this
`response declines over time [20,23]. It is believed
`that the residual T-cell response to GA after six
`months of treatment includes predominantly Th2-
`biased cells with significant degeneracy in their
`antigen recognition patterns [24]. These cells cross-
`react with myelin antigens [18], and are believed to
`
`mediate the GA-therapeutic effect through the
`release of anti-inflammatory Th2 cytokines, result-
`ing in bystander suppression and by secretion of
`brain-derived neurotrophic factor
`(BDNF)
`[16].
`However,
`it is not known at the present time
`whether GA-induced cytokine shifts correlate with
`the clinical response to treatment with GA.
`Earlier reports have shown a reduction in PBMC
`proliferation to GA among GA-treated patients
`[20,24,31]. However, these results were not corre-
`lated with patients’ clinical response to the drug.
`Subsequently, Farina et al. [20], and Weder et al.
`[31], examined the proliferative response to GA in
`GA-treated MS patients versus healthy controls
`and untreated MS patients, and found that T cells
`from GA-treated MS patients had a significant
`reduction in stimulation indexes. Although the
`PBMC proliferation data presented by Farina et al.
`showed a trend towards a higher GA-induced
`stimulation index among the GA non-responders,
`the difference between GA-responders and GA-
`non-responders was not statistically significant,
`and failed to differentiate the two groups.
`In
`addition, they found that a strong in vitro response
`for both IL-4 and IFN-g correlates with a successful
`clinical outcome [29]. Our data showed an initial
`increase in both the DCPM and stimulation in-
`dexes within three months of treatment. However,
`a significant reduction was observed in the DCPM
`and stimulation indexes after six months of GA
`therapy, which is consistent with the findings
`reported by Duda et al. [24]. Although our study
`and those of others suggest
`that
`the reduced
`proliferation is specific to GA, the proliferation of
`GA and TT-stimulated PBMCs did not differentiate
`GA-responders
`from the hypo/non-responders.
`This raises the possibility of a mechanism unre-
`lated to proliferation involved in the differential
`
`Multiple Sclerosis 2007; 13: 754 762
`
`http://msj.sagepub.com
`
`Downloaded from
`
`
`
` by guest on November 18, 2015msj.sagepub.com
`
`Page 5 of 9
`
`YEDA EXHIBIT NO. 2124
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`Clinical response to GA
`
`759
`
`GA-HR/NR (n=10)
`
`*
`
`*
`
`Pre-Rx 3mo
`
`6mo 12mo 24mo
`
`on Rx
`
`Pre-Rx
`
`3mo
`
`6mo
`
`12mo
`
`24mo
`
`on Rx
`
`75
`
`50
`
`25
`
`0
`
`b
`
`IL-4 spots per 4x105 cells
`
`600
`
`500
`
`400
`
`300
`
`200
`
`100
`
`0
`
`d
`
`IFN-γ spots per 4x105 cells
`
`f
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`IL-4/IFN-γ Ratio
`
`aG
`
`A-R (n=10)
`
`Pre-Rx 3mo
`
`6mo 12mo 24mo
`on Rx
`
`*
`
`*
`
`Pre-Rx 3mo
`
`6mo
`
`12mo 24mo
`
`on Rx
`
`*
`
`*
`
`*
`
`75
`
`50
`
`25
`
`0
`
`IL-4 spots per 4x105 cells
`
`c
`
`600
`
`500
`
`400
`
`300
`
`200
`
`100
`
`0
`
`IFN-γ spots per 4x105 cells
`
`e
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`0.0
`
`IL-4/IFN-γ Ratio
`
`Pre-Rx 3mo
`
`12mo
`6mo
`on Rx
`
`24mo
`
`0.0
`
`Pre-Rx
`
`3mo
`
`24mo
`
`12mo
`6mo
`on Rx
`Figure 1 Comparison of the mean IL-4 and IFN-g spots among GA-R (a, c) and GA-HR/NR (b, d) with anti-CD3/CD28
`stimulation before treatment and at three, six, 12 and 18 24 months during therapy. A comparison of the mean IL-4/IFN-g
`ratio before and during treatment among clinical responders (e) and non-responders (f) is also shown. Asterisk indicates
`statistically significant difference from pre-treatment (PB/0.05). The IL-4/IFN-g ratio was calculated as follows: first, the mean
`IL-4/IFN-g ratio at each time point during treatment for individual patient was determined. Then the average of the ratios at
`time points ranging from three to 24 months was determined for each patient. The IL-4/IFN-g ratio shown in (e) and
`(f) represent the mean of the ratios for each group.
`
`response to the drug among responders and hypo/
`non-responders.
`The cytokine shifts observed during GA treat-
`ment are possibly one of the drug’s key mechanisms
`of action [6,18,21,24 28]. Our data revealed that
`
`cytokine expression in unstimulated ex vivo PBMC
`did not differentiate the GA-responders from the
`hypo/non-responders, nor did ex vivo PBMCs sti-
`mulated in vitro with GA at concentrations of
`40 and 100 ug/mL, prior to and during treatment.
`
`http://msj.sagepub.com
`
`Multiple Sclerosis 2007; 13: 754 762
`
`Downloaded from
`
`
`
` by guest on November 18, 2015msj.sagepub.com
`
`Page 6 of 9
`
`YEDA EXHIBIT NO. 2124
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`anti-CD3/CD28 which induces a global T-cell res-
`ponse is more likely to reveal differences between
`GA-R and GA-HR/NR, as demonstrated in this study.
`The latter also suggest that GA-reactive T-cells pro-
`duce a systemic bystander effect on Th1 cells, since
`IFN-g expression was significantly reduced in anti-
`CD3/CD28-stimulated PBMC during treatment.
`In a sub-group of patients (GA-R/10 and GA-
`HR/NR/6) pre-treatment and post-treatment IL-4
`and IFN-g expression was compared. The cytokine
`changes observed among GA-R with anti-CD3/
`CD28 stimulation were predominantly a reduction
`in the frequency of IFNg spots without an increase
`in IL-4 expression. Consequently,
`the increase
`observed in the IL-4/IFNg ratio among GA-R was
`mainly driven by the reduction in the occurrence of
`IFN-g spots. It is noteworthy that IL-4 expression
`was reduced during treatment in the GA-HR/NR
`group, and was unchanged in the GA-R group.
`While the mechanism of
`this differential
`IL-4
`response to GA is not clear, it suggests heterogene-
`ity in the patients’ response to GA, which may
`underlie their clinical response to the drug. It also
`suggests that a Th2 skew favors a clinical response
`to GA. Hong et al. recently proposed a mechanism
`of action of GA that involves GA-induced regula-
`tory T-cells (CD4 CD25 FoxP3) [30]. These
`T-reg cells, induced in response to GA therapy, can
`potentially explain the reduction in the number of
`IFN-g positive ELISPOTS. Although we have not
`specifically examined these regulatory cells in our
`samples, it is possible that their induction may
`correlate with the clinical response status of the
`patients. In the absence of activation of IL-4, a Th2
`cytokine, a question arises as to the role of GA-
`specific induction of Th2 cells in the suppression of
`MS, and whether reduction in IFN-g plays a more
`important role in GA therapy. We believe that
`while our findings highlight the importance of
`IFN-g reduction in the clinical response, and the
`association of IL-4 suppression with hypo-respon-
`siveness, it does not preclude a role for GA-reactive
`Th2 cells in mediating the therapeutic effect, since
`GA-reactive Th2 cells have been demonstrated in
`multiple laboratories, and are likely to be present in
`low frequencies, that are undetectable in our assay.
`Since the goal of immunomodulation in MS is to
`prevent relapses and disability, we correlated the
`patients’ annual relapse rate with IL-4 and IFN-g
`expression and the IL-4/IFN-g ratio. The results
`showed no significant correlation between ARR
`and the Th2 cytokine IL-4. However, a positive
`correlation between ARR and the Th1 cytokine IFN-
`g, and a significant inverse correlation between ARR
`and IL-4/IFN-g ratio were observed. These findings
`suggest that modulation of IFN-g and IL-4 expres-
`sion contribute to the beneficial clinical effect of
`this therapy.
`
`760
`
`RM Valenzuela et al.
`
`p>0.05
`
`0
`
`50 100 150 200 250 300 350
`IL-4
`
`p = 0.040
`
`0
`
`100
`
`200
`IFN-γ
`
`300
`
`400
`
`p = 0.041
`
`0.5
`
`0.4
`
`0.3
`
`0.2
`
`0.1
`
`0.0
`
`a
`
`Annula Relapse Rate
`
`b
`
`1.75
`
`1.50
`
`1.25
`
`1.00
`
`0.75
`
`0.50
`
`0.25
`
`0.00
`
`1.75
`
`1.50
`
`1.25
`
`1.00
`
`0.75
`
`0.50
`
`0.25
`
`0.00
`
`Annula Relapse Rate
`
`c
`
`Annula Relapse Rate
`
`0.0
`
`0.5
`
`2.0
`1.5
`1.0
`IL-4/IFN-γ Ratio
`
`2.5
`
`3.0
`
`Figure 2 Linear regression analysis of ARR against (a) IL-4,
`(b) IFN-g and (c) IL-4/IFN-g ratio using Spearman correlation.
`
`However, the mean number of IL-4 and IFN-g spots
`in PBMC stimulated with anti-CD3/CD28 was dif-
`ferent between the two groups. A possible explana-
`tion for the unobtainable differences with GA
`stimulation is that the precursor frequency of
`GA-reactive T-cells is very low following chronic
`treatment, so that GA-stimulation ex vivo may
`not elicit detectable differences in cytokine expres-
`sion between GA-R and GA-HR/NR. Such differences
`may be detectable if one generates GA-reactive T-cell
`lines from GA-treated patients [18,19]. In contrast,
`
`Multiple Sclerosis 2007; 13: 754 762
`
`http://msj.sagepub.com
`
`Downloaded from
`
`msj.sagepub.com
`
` by guest on November 18, 2015
`
`Page 7 of 9
`
`YEDA EXHIBIT NO. 2124
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`Clinical response to GA
`
`761
`
`*
`
`TT
`
`*
`
`GA
`
`*
`
`Pre-Rx
`N=16
`
`3mo
`24
`
`6mo
`21
`
`12mo
`17
`
`24mo
`22
`
`GLATIRAMER ACETATE
`
`GA-R
`
`GA-HR/NR
`
`b
`
`50
`
`40
`
`30
`
`20
`
`10
`
`0
`
`10.0
`
`7.5
`
`5.0
`
`2.5
`
`0.0
`
`STIMULATION INDEX
`
`d
`
`STIMULATION INDEX
`
`a
`
`50000
`
`40000
`
`30000
`
`20000
`
`10000
`
`0
`
`*
`
`TT
`
`*
`
`*
`
`GA
`
`Pre-Rx
`N=16
`
`3mo
`24
`
`6mo
`21
`
`12mo
`17
`
`24mo
`22
`
`GLATIRAMER ACETATE
`
`15000
`
`12000
`
`9000
`
`6000
`
`3000
`
`0
`
`GA-R
`
`GA-HR/NR
`
`∆CPM
`
`c
`
`∆CPM
`
`Pre-Rx 3mo 6mo 12mo 24mo
`
`Pre-Rx 3mo 6mo 12mo 24mo
`
`10
`R
`15
`14
`10
`13
`HR/NR 6
`9
`7
`7
`9
`Figure 3 Proliferative response of PBMC among MS patients expressed as DCPM (a, c) and as stimulation index (b, d) after
`stimulation with TT 5 mg/mL (a, b) and GA 40 mg/mL (a d) before treatment and at different time points during GA treatment.
`(a) and (b) show data for the entire patient group, while (c) and (d) compare the results between clinical responders (n/10)
`and hypo/non-responders (n/6).
`
`R
`15
`10
`HR/NR 6 9
`
`14
`7
`
`10
`7
`
`13
`9
`
`In the European/Canadian GA study group by
`Comi et al.
`[15],
`it was demonstrated that the
`number of enhancing lesions and accumulation of
`new T2 lesions begin to decrease at six months
`following GA treatment. Our findings of cytokine
`changes among GA-R showed a significant increase
`in the IL-4/IFN-g ratio starting six months after
`treatment, which was not observed among the GA-
`HR/NR group. Thus, the time course of the immu-
`nological changes induced by GA-therapy seem to
`parallel the time course of the changes on MRI. These
`observations further support the significance of the
`IL-4/IFN-g changes in the clinical response to GA.
`In summary, new findings have emerged from
`this study: First, the lymphoproliferative response
`to GA did not differentiate the GA-responders from
`the hypo/non-responders. Second, the Th1 cyto-
`kine IFN-g (in anti-CD3/CD28-stimulated PBMC)
`was significantly lower among the GA-responders
`compared to hypo/non-responders during GA ther-
`apy, and was the main driving force towards a
`significant difference in the IL-4/IFN-g ratio ob-
`served between GA-R and GA-HR/NR. Third, a
`reduction in the expression of the Th2 cytokine
`IL-4 seems to be associated with hypo/non-respon-
`ders to GA. Fourth, an increased IL-4/IFN-g ratio
`
`and a decreased IFN-g cytokine response correlate
`with a lower annual relapse rate.
`The mechanism underlying the differential re-
`sponse to GA between the responders and hypo/
`non-responders is not clear, but could involve
`transcriptional regulation of cytokine genes and
`the HLA background of the patients among other
`factors [23]. More data are needed to validate the
`prospective use of IL-4/IFN-g expression in PBMC as
`a biomarker of clinical response to GA for indivi-
`dual patients.
`
`Acknowledgements
`
`This study was supported by grants to Dr Dhib-
`Jalbut from The National Institutes of Health (K24-
`N502082), the Wadsworth Foundation and Teva
`Pharmaceuticals Inc.
`
`References
`
`1. Fridkis-Hareli M, Aharoni R, Teitelbaum D, Arnon
`R, Sela M, Strominger JL. Binding of random copoly-
`mers of three amino acids to class II MHC molecules. Int
`Immunol 1999; 11: 635 41.
`
`http://msj.sagepub.com
`
`Multiple Sclerosis 2007; 13: 754 762
`
`Downloaded from
`
`msj.sagepub.com
`
` by guest on November 18, 2015
`
`Page 8 of 9
`
`YEDA EXHIBIT NO. 2124
`MYLAN PHARM. v YEDA
`IPR2015-00644
`
`

`
`762
`
`RM Valenzuela et al.
`
`2. Fridkis-Hareli M, Neveu JM, Robinson RA, Lane
`WS, Gauthier L, Wucherpfennig KW et al . Binding
`motifs of copolymer 1 to multiple sclerosis and rheuma-
`toid arthritis-associated HLA-DR molecules. J Immunol
`1999; 162: 4697 704.
`Fugger L,
`3. Fridkis-Hareli M, Rosloniec EF,
`Strominger JL. Synthetic amino acid copolymers that
`bind to HLA-DR proteins and inhibit type II collagen-
`reactive T cell clones. Proc Natl Acad Sci USA 1998; 95:
`12528 31.
`4. Fridkis Hareli M, Santabrogio L, Stern JN, Fugger
`L, Brosnan C, Strominger JL. Novel synthetic amino
`acid copolymers that inhibit autoanitigen-specific T cell
`responses and suppresses EAE. J Clin Invest 2002; 109:
`1635 43.
`5. Fridkis-Hareli M, Stern JN, Fugger L, Strominger
`JL. Synthetic peptides that inhibit binding of myelin
`basic protein 85-89 epitope to MS associated HLA-DR2
`molecules and MBP-specific T cell responses. Hum Im-
`munol 2001; 62: 753 63.
`6. Aharoni R, Teitelbaum D, Sela M, Arnon R. By-
`stander suppression of experimental autoimmune ence-
`phalomyelitis by T-cell lines and clones of the Th2 type
`induced by copolymer 1. J Neuroimmunol 1998; 91: 1358.
`7. Aharoni R, Teitalbaum D, Sela M, Amon R. Copo-
`lymer 1 induces T cells of the T-helper type 2 that cross-
`react with myelin basic protein and suppress experimen-
`tal autoimmune encephalomyelitis. Proc Natl Acad Sci
`USA 1997; 94:10821.
`8. Arnon R. The development of Cop 1 (Copaxone), an
`innovative drug for the treatment of multiple sclerosis:
`personal reflections. Immunol Lett 1996; 50: 1 15.
`9. Aharoni R, Teitelbaum D, Leitner O, Meshorer A,
`Sela M, Arnon R. Specific Th2 cells accumulate in the
`central nervous system of mice protected against experi-
`mental autoimmune encephalomyelitis by copolymer-1.
`Proc Natl Acad Sci USA 2000; 97: 11472 77.
`10. Dhib-Jalbut SS. Glatiramer acetate (Copaxone) therapy
`for multiple sclerosis. Pharmacol Therapeut 2003; 245 55.
`11. Gran B, Tranquill LR, Chen M, Bielekova B, Zhou
`W, Dhib-Jalbut SS et al . Mechanisms of immunomo-
`dulation by glatiramer acetate. Neurology 2000; 55:1704
`14.
`12. Johnson KP, Brooks BR, Cohen JA, Ford C, Gold-
`stein J, Lisak RR et al . Copolymer 1 reduces relapse rate
`and improves disability in relapsing-remitting multiple
`sclerosis: results of phase III multicenter, double-blind,
`placebo-controlled trial. Neurology 1995; 45: 105.
`13. Miller AS, Shapiro R, Gershtein A, Kinarty H,
`Rawashdeh S, Honigman S et al . Treatment of multi-
`ple sclerosis with copolymer-1 (Copaxone): implicating
`mechanisms of Th1 to Th2/Th3 immune deviation.
`J Neuroimmunol 1998; 92: 113 21.
`14. Qin Y, Zhang DQ, Prat A, Pouly S, Antel J.
`Characterization of T-cell lines derived from glatiramer
`acetate-treated multiple sclerosis patients. J