`MONCHEN
`
`KDNIG SZYNKA TlLMANN
`VON FIENESSE
`
`
`
`FATENTANWALTE PAHTNEHSCHAFT mbB
`European Patent 03 748 4311-1412] 1 528 933
`
`
`
`LUDWIG-
`MAXJ'MILIANS-
`UN Ivsnsn'Ar
`
`DEPARTMENT PEARMAZXE
`
`PHARMAZEUTIBCHE TECHNOLOGIE UND
`
`BIOPHARMAZIE
`PROF. D‘Ri OERHARD WINTER
`
`
`LMU. Prat Dr. 6. Winter, Burenandmr. 6. Hair: 8, 81877 MLinchen
`
`Prof. Dr. G, Winter
`
`1hr Zeichen, lhre Nachricht vom
`
`Unser Zeichen
`
`Telefon +46 (0)89 2180-77022
`Telefax++9 (0)89 2180—77020
`Gerhard.Wmte1-@lrz.uni-
`nmenchende
`
`Pos tanschnf‘t
`Butenandtstr. 5—13, Hans B
`31377 Munchen
`
`Munchen, .13. 12014
`
`Expert Opinion
`
`I have been asked by Kbnig Szynka Tilmann von Renesse to comment on an experimental
`report produced by Coriolis Pharma GmbH (“Coriolis”) for
`the patent EP 1 528 933
`
`oppositiOn proceedings and on eXperiments filed during prosecution (D30) of the patent.
`
`Personal Background and Experience
`
`After I finished my PhD. under the supervision of Prof. Stricker in the area of transdermal
`
`absorption simulations in 1987 with summa cum laude I worked for several years in the
`pharmaceutical
`industry.
`In 1987 I was
`leading a laboratory for solid dosage form
`development at Merck in Darmstadt before I joined Bochringer Mannheim, later Roche, in
`l988 where I stayed until 1999. During this time my responsibilities were parenteral and
`liquid dosage form development including every development step from preformulation,
`to
`
`stability studies and clinical supplies to scale up, production transfer and dossier submission.
`
`[led a large scale aseptic pilot plant to supply worldwide clinical studies with lyophilisates
`
`and sterile solutions. My special interest was focused on protein and peptide formulations,
`
`freeze drying and parenteral drug delivery systems and I was involved in the development of
`major drug products like EPO and rt—PA.
`
`In 1999
`I was
`appointed
`a
`full professor
`for Pharmaceutical Technology and
`Biopharmaceutics at
`the University of Munich, Center for Drug Research, where I am
`currently working on protein stabilisation, parenteral dosage form technology, novel drying
`technologies, drug delivery systems and colloidal drug formulations.
`
`
`
`Dienrtgebnude
`Butenandutr. 5-13, Hang B
`Baum 31.092
`31377 Mllnchen
`
`Ol‘fentliche Verkehr: mittel
`Bus £66
`U-Bahn Us
`
`Bayerische Landesbank Munchen
`Km. 9* 858 3L2 700 500 00
`USt-ldNr. DE 811 9.05 325
`
`AMGEN INC.
`
`Exhibit 1055
`
`Ex. 1055 - Page 1 of 8
`
`Ex. 1055 - Page 1 of 8
`
`AMGEN INC.
`Exhibit 1055
`
`

`

`
`
`LUDWlG-MAXIMILIANS—UNIVERSITKT MUNCHEN SEITE fl VON 8
`
`I am an inventor on numerous patents (about 50) and author of numerous publications (about
`100) in my research field (see Annex) and I am a member of many scientific societies and
`scientific advisory boards.
`
`Instructions and Summary
`
`I have read and understood the opposed patent EP 1 528 933 and the publication D11 (Barrera
`et al. Ann Rheum Dis 2001, 60:660-669).
`I have also been provided with the experimental
`report describing the stability study performed by Coriolis and the “Test Results” reported in
`
`D30. I have analyzed the Coriolis study, especially the raw data and I have also analyzed the
`“Test Results” provided in document D30. I comment as follows.
`
`In summary, the Coriolis study is a detailed comparison of different storage conditions in
`order to test the storage stability of a formulation falling under the patent (called the “Claim
`
`formulation” in the Coriolis report and a formulation according to Barrera (called the “Citrate
`formulation” in the Coriolis report). I will refer to these terms below.
`
`The outcome of the Coriolis study is a higher storage stability for the Claim formulation
`
`compared to the Citrate formulation. Furthermore,
`
`the Test Results from D30 support the
`
`aggregation
`stability against
`the Claim formulation maintains
`conclusion that
`fragmentation at relatively high protein concentration and at different pH values.
`
`and
`
`I have been asked to answer the questions below which relate to the Coriolis report and the
`D30 Test Results.
`
`1. Please explain the outcome ofthe study performed by Coriolis as well as the test methods applied
`
`The Coriolis comparative stugz
`
`formulation with 20 mg/ml and 25 mg/ml D2E7 antibody
`a
`In the Coriolis report
`(Adaiimumab), mannitol, Polysorbate 80 and citrate/phosphate buffer
`system (“Claim
`formulation”) were compared to formulations of the same D2E7 antibody concentrations in a
`citrate buffer with mannitol (Citrate formulation).
`I
`
`The Coriolis report provides raw data from the comparative stability study of the Claim
`formulation and the Citrate formulation from Barrera et al., each in two concentrations. All
`
`four formulations are defined with all excipients and respective concentrations in the report
`(see page 13). I have looked at the raw data very closely and I will comment in detail below.
`
`a) The applied test methods
`
`The comparative stability tests performed by Coriolis have applied the standard approaches
`and analytical methods that are used in pharmaceutical
`industry in accordance with the
`
`European Pharmacopeia in order to come to a well-founded conclusion on storage stability.
`
`The analytical methods and their purposes are shown below.
`
`Ex. 1055 - Page 2 of 8
`
`Ex. 1055 - Page 2 of 8
`
`

`

`SEITE 8 VON s
`LUDWlG-MAXIMILIANS-UNIVERSITKT MUNCHEN
`
`
`
`
`
`
`
`
`
`
`
`
`
`__
`
`
`
`
`
`
`
`
`
`Circular Dichroism (CD) spectroscopy
`Secondary Structure, Tertiary Structure
`
`Far-UV-CD
`moseopy
`Near—UV—CD
`
`spectroscopy
`
`
`
`
`Micro-Flowl'maging(MFI)
`
`in
`Particles
`visualimtion)
`
`the um-size range
`
`(quantification and
`
`Visual inspection (Ph.Eur.)
`
`Osmolality
`
`Visible particles
`
`Osmolality
`
`”V'V’SS”°“°“°” _
`
`
`
`A combination of the methods shown above is necessary in order to cover the full range of
`qualitative (physical and chemical) and quantitative changes of the antibody during storage.
`Every method has its strengths and weaknesses.
`
`For example the DLS method which has shown a significant difference between the Claim
`
`formulation and the Citrate formulation does not distinguish accurately on the quantitative
`level for small particles if larger particles are present.
`
`HP-SEC is the standard analysis method for aggregation and degradation. However, large
`aggregates may not be detected within the detection range of the HP-SEC method. HP-SEC
`together with the other methods allows a complete picture and identification of chemical and
`physical changes.
`
`Finally, UV—Vis Spectroscopy shows that material has not been affected or lost prior to the
`measurement.
`
`The applied methods can be grouped into different groups according to their different focus.
`
`Group 1: Methods 1 and 2 are used to detect major losses of substance due to whatever reason,
`
`be it a mistake in preparing the sample, or losses due to chemical, physical degradation,
`adsorption, or massive structural losses.
`
`
`
`Ex. 1055 - Page 3 of 8
`
`Ex. 1055 - Page 3 of 8
`
`

`

`SEITE 4r VON a
`LUDWlG—MAXIMILIANS-UNIVERSITA'T MUNCHEN
`W
`
`Group [1: Methods 3—7: are each used to detect and to quantify physical degradation towards
`unfolded, aggregated and finally precipitated protein,
`small
`(nm range)
`and larger
`aggregates/particles (um range) to deliver a complete picture.
`
`Group III: Method 8 is able to detect chemical and physical degradation as it separates and
`detects soluble aggregates as well as fragments of the protein. Separation criterion is the size
`of the degradation product in relation to the size of the native monomeric molecule.
`
`Group IV: Method 9 is used to determine chemical degradation. Such degradation typically
`results
`in protein species that show small differences in their charge/mass or charge
`distribution.
`
`Group V: Methods 10 -ll allow for the determination of structural deteriorations of the
`
`protein. Such changes in structure can be a result of physical or chemical degradation at
`exposure to a denaturing environment.
`
`1. Loss of substance
`
`1. UV-Vis Spectroscopy
`2. HP-S EC
`
`II. Physical stability and aggregation
`
`3. Visual inspection, Visual appearance - color
`4. Clarity (Turbidity)
`5. Light obscuration (LO)
`
`6. Micro-Flow Imaging (MFI)
`
`7. Dynamic light scattering (DLS)
`III, Chemical changes and aggregation
`
`8. High pressure size exclusion chromatography (HP-SEC) —- standard method
`IV. Chemical changes
`9.
`lon~exchange chromatography (CEX/IEX)
`V. Physical or chemical changes:
`
`10. Fluorescence spectroscopy (intrinsic fluorescence)
`
`1 l. Far-UV and Near-UV circular dichroism (CD) spectroscopy
`
`b) The test results of the comparative storage stability tests
`
`I have looked closely at the raw data from the different tests in the Coriolis report. The data
`show that in the different storage stability tests, the D213? antibody in the Claim formulation
`has a better stability regarding aggregation and regarding fragmentation than the same
`antibody in the Citrate formulation.
`
`Specifically, even at 50°C the Claim formulation shows only a very small amount of particle
`formation (visible or subvisible), hardly any aggregation in the Inn particle size range and
`very little monomer loss as can be seen in the HP-SEC results. The raw data show that the
`
`Citrate formulation has higher levels of particle formation, aggregation and fragmentation.
`Consequently, the Claim formulation has a much higher storage stability on the physical and
`on the chemical level.
`
`Ex. 1055 - Page 4 of 8
`
`Ex. 1055 - Page 4 of 8
`
`

`

`
`
`LUDWIG»MAXlMILIANS-UNIVERSITXT MUNCHEN 831112 a VON sm
`
`I will now discuss the data in the Coriolis report according to the type of change that has been
`measured.
`
`Particle formation:
`
`The Citrate formulation shows a high amount of particle formation compared to the .Claim
`formulation. This can be seen in Table l3 (visual
`inspection — visible particles),
`in Figure
`4/Tables 16-17 (Light Obscuration —« subvisible pm size particles) and in Figure 5/Tables 18-
`19 (Micro flow Imaging — subvisible um size particles) ofthe Coriolis report.
`
`It is therefore quite obvious that the Citrate formulation is physically less stable compared to
`the Claim formulation when temperature stress is applied.
`
`The aggregation behavior of the formulations is shown in Tables 20-21 of the Dynamic Light
`scattering test
`(DLS).
`It has to be noted that due to the nature of the method,
`large
`aggregates/particles
`have
`to
`be
`determined
`in
`the
`presence
`of
`small
`particles/molecules/aggregates without separation. This leads to the fact that larger particles
`may dominate the amount and nature of the scattered light and the results calculated by the
`correlation function in the system software may not represent the true nature of the size
`distribution of the sample. In other words, the method can very well and reliably detect that
`the sample has changed and that a significant number of aggregates has been formed, but it
`cannot deliver reliable data on the size and number of the aggregates formed and the size and
`number of the monomers remaining. This is particularly true for samples with a large amount
`of non-monomeric particles and/or samples with very large particles. It is therefore necessary
`to combine this method with other methods covering the lower (HP-SEC) and higher (LO,
`MFI) size range to deliver a comprehensive picture of the quality of the sample as has been
`done by the selection of methods performed by Coriolis.
`
`The aggregation seen in the DLS analysis is not visible in the HP-SEC results, which is most
`likely due to the fact that the aggregates formed are larger than. the exclusion size of the
`column. Larger aggregates/particles are held back in the column, often already at the inlet
`filter preventing the column from massive clogging.
`
`Regarding fragmentation, the high pressure size exclusion chromatography (HP-SEC) results
`in section 3.10 show very little fragmentation for the Claim formulation even in the samples
`stored at 40°C and 50°C (of. Figures 22 and 30) while the Citrate formulation shows complete
`monomer loss (main peak) after one week at 50°C and significant monomer loss at 40°C due
`to fragmentation (of. Figures 23 and 30).
`
`It is therefore obvious that the Citrate formulation shows also more chemical degradation (as
`fragmentation is a clear sign of chemical degradation; covalent bonds are broken which are
`
`linking the fragments together in the native state) compared to the Claim formulation when
`stressed.
`
`Even the samples stored at 25 °C (room temperature) already showed a clear trend towards
`fragmentation for the Citrate formulation.
`
`IEX Chromatograghx
`
`Ex. 1055 - Page 5 of 8
`
`
`
`r...,m.,...ws.,figmm‘w~.___,.,._.,,._..,.,,s......
`
`
`
`
`
`Ex. 1055 - Page 5 of 8
`
`

`

`
`
`LunWIG-MAXIMXLIANS-UerakslTAT MUNCHEN SEITE e VON a
`
`Results are presented in Fig. 33-44. IEX chromatography presents different charge variants of
`the protein drugs. Already in the native state, before storage, a pattern of charge variants is
`present, its retention or change over time and stress is assessed. At 25°C no charge variant
`shifts can be found for the Claim formulation, whereas the Citrate formulation shows a
`
`decrease in the so called “LysO” peak and increases in two other peaks. For higher
`temperature, the situation is more complex. The changes in charge variants are quantitatively
`different in a way that the Citrate formulation shows an overall much higher loss of the native
`peak distribution,
`in addition, the shift goes into a different direction as compared to the
`Claim formulation.
`
`it can be concluded that chemical degradation towards new charge variants is
`in summary,
`less pronounced for the Claim formulation compared to the Citrate formulation.
`
`Studies at elevated temperature are proposed as valuable and necessary in the international
`ICH guidelines, e.g.,
`in guideline QSC, chapter 6.3 “Guideline for Industry Quality of
`Biotechnological Products: Stability Testing of Biotechnological/Biological Products”.
`Stability testing at elevated temperatures is a concept applied in the biopharmaceutical
`industry for many years, a more recent review has been published by Hawe et a1 (Hawe et al.
`J Pharm Sci 101(3), 895-913 (2012).
`
`As a conclusion 1 can say that the Claim formulation shows a significantly higher stability,
`Le. a lower aggregation and lower fragmentation in the storage stability tests.
`
`2. Please exglain the long—term stabiligg results shown in document D30 and compare these wigh the
`outcome at the Coriolis Regort
`,
`
`I was asked to compare the results from the storage stability tests of Coriolis discussed above
`with the respective stability data in the Document D30 Test Results in order to find out
`
`whether the findings by Coriolis are in line with the results in D30 and whether it is possible
`to come to the conclusion that the storage stability that has resulted from the Coriolis tests is
`valid for the claimed concentration range and at different pH values. 1 comment as follows:
`
`In summary, the results in D30 can be interpreted as follows: Stability of the antibody drug
`molecule (D2137) is maintained at different pH values and concentrations in a combination of
`
`citrate and phosphate buffer. Usually the pH value has a strong impact on protein stability. It
`appears however, that a formulation including “phosphate and citrate” compared to “only
`citrate” stabilized the antibody over a broad pH range and to a large extent even without
`Polysorbate 80. Consequently,
`it is not the Polysorbate 80 alone that is responsible for the
`high stability regarding preservation of the monomeric antibody but the overall formulation
`and to a large extent the combination of phosphate and citrate.
`
`In D30, various stability tests have been performed and analytical methods comparable to
`those of the Coriolis report have been applied. Especially the high pressure size exclusion
`chromatography (HP-SEC) data can be compared with the results discussed above for the
`
`Coriolis report, but also the particle related analyses (clarity, visible and subvisible particles).
`
`The HP-SEC results indicate monomer loss due to fragmentation and formation of soluble
`
`aggregates and in D30 this analytical method has been used in storage stability tests for
`
`
`
`ta-rms“urn-swwa-msau—Fw-W.”New.-.“m..........r
`
`Ex. 1055 - Page 6 of 8
`
`Ex. 1055 - Page 6 of 8
`
`

`

`LUDWlG-MAXJM[LIANS-UNIVERSITKT MUNCHEN
`SEITE 'l VON 8
`
`
`different Adalimumab (this term is used in D30 for DZE7) formulations. I will discuss the»
`
`details in the following paragraphs.
`
`Example A, Series 1 is a 24 month 2-8°C long-term stability study of Adalimumab in the
`
`Claim formulation at different concentrations. The concentration range even exceeds the
`claimed range, ranging from 12.5 mg/ml to 158 mg/ml. The particle results shown in tables 3-
`
`5 (clarity, visible, subvisible particles) show little or no instability (e.g. aggregation) for all
`concentrations after 24 months.
`
`The results of the HP-SEC analysis are shown in table 6. All concentrations show extremely
`low monomer loss after 24 months of storage at 2-8°C.
`
`Usually an increase in protein concentration can be expected to lead to an increase in
`aggregation when accelerated or long term storage studies are carried out (which would be
`seen as monomer loss) over time.
`
`Fragmentation can be expected to be less affected by different levels of protein concentration.
`The above results show a surprisingly high antibody stability (low level of aggregation and
`low level of fragmentation) independent of the antibody concentration.
`
`Additional 12 month stability tests are shown in test Series 3 for different pH values (pH 4
`and pH 6) in an antibody concentration range from 0.2 mg/ml to 100 mg/ml. I have asked for
`the details on the used formulation as these were not completely clear from D30.
`I was
`informed that the formulation used in these tests was a 10 mM citrate, 10 mM phosphate
`formulation with no other excipients. Table 8 shows the results: a low monomer loss due to
`aggregation or fragmentation after 12 months at the different concentrations for all the tested
`pH values.
`
`In my View, these results can be interpreted as follows: Stability is maintained at different pH
`values and antibody concentrations in a formulation using a combination of citrate and
`
`phosphate buffers. Usually the pH value has a strong impact on protein stability. In the results
`of test Series 3 it appears that the buffer used (combination of citrate and phosphate buffer)
`stabilized the antibody over a broad pH range.
`
`This is confirmed and extended in Example B, Series 1, Fig. 5a and 5b. For the pH range of
`3-8 Fig. 5b shows the HP~SEC results of a storage study at 5°C and additionally of
`accelerated studies at elevated temperature (40°C and 50°C). The results show that
`the
`monomer loss is minimal from pH 5-8, 4-7 and even 3—8 in the tested formulation is. a
`
`combination of a citrate and a phosphate buffer with 2 mg/ml antibody concentration at 50°C,
`40°C and 5°C, respectively. Although this is a much lower antibody concentration than
`claimed,
`there should be no major impact of this difference in antibody concentration on
`chemical stability such as fragmentation. Consequently, Example B, Series 1 shows that the
`
`advantage of high fragmentation stability can be expected over the claimed pH range.
`
`In Example C Table 20 shows that storage stability over 24 months is maintained to a large
`extent also without Polysorbate 80. It appears, that the text below Tables 19 and 20 has been
`
`confused but the data with and without Polysorbate 80 in Tables 19 and 20 show comparably
`
`"mum-amuuwtmm-iwam.-.r
`
`Ex. 1055 - Page 7 of 8
`
`Ex. 1055 - Page 7 of 8
`
`

`

`
`
`LUDWIG—MAXIMXLIANS-UNIVERSITKT MflNCHEN 5151'“! 9 VON a
`
`high monomer levels after 24 months at 2-8°C and at 25°C after 6 months (slightly lower in
`both cases after 6 months at 40°C). Table 13 shows that there is a higher particle formation
`
`without Polysorbate 80. Consequently, Polysorbate 80 contributes to stability but it is not the
`Polysorbate 80 alone that is responsible for the high stability regarding preservation of the
`monomeric antibody but to a large extent the phosphate/citrate combination.
`
`3. Do the specific experimental results allow a conclusion for concentrations and pH values
`not tested?
`
`The analysis of the different studies shows that the storage stability for the Claim formulation
`
`can be expected at different concentrations and pH values.
`
`the Citrate formulation had a significantly lower storage stability at
`In the Coriolis study,
`concentrations of 20 mg/ml and 25 mg/mert higher concentrations and different pH values I
`
`would expect the same quality of results. In other words, I would expect an increase of pH to
`have a negative effect on the stability of the antibody in all formulations, but I would still
`expect higher stability for the Claim formulation compared to the Citrate formulation as
`explained below.
`
`Aggregation
`
`If the Citrate formulation aggregates at 20 mg/ml and 25 rug/ml as shown in the Coriolis
`
`report, it can be expected that it would also aggregate at 50 rug/ml or 130 mg/ml because a
`higher protein concentration would be expected to lead to an even higher degree of
`aggregation than shown in the Coriolis report. On the other hand the Coriolis experiments
`show that the antibody in the Claim formulation was stable at 20 mg/ml and 25 mg/ml and for
`
`higher concentrations it is shown in D30 that the antibody in the Claim formulation was stable
`at up to 158 mg/ml (cf. Example A, Series 1).
`
`Consequently, I would expect a lower aggregation in the Claim formulation than in the Citrate
`formulation at higher pH values and at higher antibody concentrations because the antibody in
`the Citrate formulation would most likely aggregate even more under these conditions than it
`
`did in the Coriolis study.
`
`I believe the statements I have made to be true. The opinion I have expressed above is my
`
`own professional opinion.
`
`Iv.
`
`Place, Date
`
`marge/131!, 60 {/0
`
`
`
`
`
`
`Ex. 1055 - Page 8 of 8
`
`Ex. 1055 - Page 8 of 8
`
`