Fresenius Kabi Deutschland GmbH, 61346 Bad Homburg
`
`European Patent Office
`
`80298 Munich
`
`GERMANY
`
`Fresenius Kabi
`Deutschland GmbH
`Patent Department
`
`Office
`Borkenberg 14
`61440 Oberursel
`Germany
`Postal Address
`Else-Kroner-StraBe 1
`61352 Bad Homburg
`Germany
`T +49 (0) 6172 686
`
`-4926
`-7389
`-2889
`F +49 (0) 6172 608-39-0234
`patents@fresenius-kabi.com
`www. fresenius-kabi. D5
`
`Oberursel, 20/08/2018
`
`Opposition against European Patent EP 1 687 019 B1
`
`Title: Propylene glycol-containing peptide formulations which are optimal for production
`and for use in injection devices
`
`Proprietor: NOVO NORDISK A/S
`
`Our Ref.: FKE18087
`
`1. General matters
`
`The patent is opposed in its entirety. As will be shown below, none of claims 1-17
`meets the requirements of the EPC. The patent is opposed on the grounds of Article
`100(a) and (c) EPC in conjunction with Articles 54, 56, and 123(2) EPC.
`
`The opposition fee (EUR 785,00) is to be debited from our account no. 28002213.
`
`Full revocation of the patent is requested. Should the Opposition Division be minded to
`maintain the patent in any form, oral proceedings under Article 116 EPC are requested.
`
`2.
`
`List of documents
`
`Reference is made to the following documents:
`
`D1 WO 03/002136 A2 - published: 9 January 2003
`
`D2 WO 2005/046716 A1 - published: 26 May 2005
`
`Novo Nordisk A/S Ex. 2006, P. 1
`Fresenius Kabi v. Novo Nordisk
`IPR2022-00657
`
`

`

`Opposition against EP 1 687 019 - Facts and Arguments
`
`Page 2/10
`
`D3
`
`D4
`
`Pharmaceuticals and Medical Devices Agency Japan: Victoza Subcutaneous
`Injection 18 mg - Report on the Deliberation Results. Evaluation and
`Licensing Division, Pharmaceutical and Food Safety Bureau, Ministry of
`Health, Labour and Welfare - published: 2009
`
`Powell et al.: Parenteral Peptide Formulations: Chemical and Physical
`Properties of Native Luteinizing Hormone-Releasing Hormone (LHRH) and
`Hydrophobic Analogues in Aqueous Solution. Pharmaceutical Research, Vol.
`8, No. 10, 1991
`
`DS WO 03/033671 A2 - published: 24 April 2003
`
`D6 WO 95/22560 A1 - published: 1995
`
`D7
`
`Board of Appeal Decision T 0235/97 - 3.3.2, 10 January 2002
`
`3. Added subject-matter, Article 123(2) EPC
`
`The subject-matter of claim 7 is not found in the application as filed. Claim 1
`refers to a pharmaceutical formulation comprising the peptide Arg 34
`, Lys 26(NE-(y(cid:173)
`Glu(N0-hexadecanoyl)))-GLP-1(7-37), wherein claim 7 further requires that "said
`
`peptide consists of Arg34, Lys26(NE-(y-G/u(N°-hexadecanoy/)))-GLP-1(7-37)". It is
`not apparent how this would further qualify the subject-matter of claim 1, which
`already specifies the exact same peptide. In any event such combination of
`features and terms is nowhere found in the original disclosure.
`
`In addition, the subject-matter of claim 10 extends beyond the content of the
`application as filed, which is silent on a pH range of 8.0 to 8.3.
`
`For these reasons, the requirements of Article 123(2) EPC are not met.
`
`4.
`
`Lack of novelty, Article 54 EPC
`
`4.1 Claim 1 lacks novelty over D1
`
`D1 discloses pharmaceutical formulations of GLP-1 compounds and methods for
`preparation thereof (see abstract). Claim 1 of D1 relates to a pharmaceutical
`formulation comprising a GLP-1 compound and a buffer, the formulation having a
`pH from 7.0 to 10. According to claim 25 of D1, the GLP-1 compound is Arg 34
`,
`Lys 26(NE-(y-Glu(N°-hexadecanoyl)))-GLP-1(7-37). The formulation of D1 further
`comprises an isotonic agent which is present in a concentration from 1 mg/ml to
`50 mg/ml (claims 13 and 14; p19, lines 10-11), which falls within the range of 1
`to 100 mg/ml of claim 1 of the opposed patent.
`
`The isotonic agent may be propylene glycol (D1, p18, lines 34-35). While
`propylene glycol is disclosed as part of a list of possible isotonic agents (D1, p18,
`line 34 - p19, line 4), a selection from a single list of specifically disclosed
`elements does not confer novelty (GL G-VI 8(i)). In addition, D1 notes that "each
`
`Novo Nordisk A/S Ex. 2006, P. 2
`Fresenius Kabi v. Novo Nordisk
`IPR2022-00657
`
`

`

`Opposition against EP 1 687 019 - Facts and Arguments
`
`Page 3/10
`
`one of these specific isotonic agents constitutes an alternative embodiment of the
`invention" (p19, lines 7-8).
`
`Thus, D1 discloses all
`features of claim 1 of the opposed patent, which
`consequently lacks novelty.
`
`4.2 Claim 1 lacks novelty over D2
`
`Document D2 (WO 2005/046716) has a priority date of 13 November 2003 and a
`filing date of 12 November 2004. It was published on 26 May 2005. It has been
`supplied to the European Patent Office in one of its official languages and the
`national fee provided for in Article 22(1) or Article 39(1) PCT has been paid. D2
`thus constitutes state of the art pursuant to Articles 54(3) EPC and 54( 4) EPC
`1973.
`
`In the examination phase of the application leading to the opposed patent, the
`patentee argued that the priority claim of D2 was invalid as far as the disclosure
`of propylene glycol was concerned, since its priority application (Danish patent
`application PA 2003 01689) did not mention propylene glycol. According to the
`applicant, D2 thus had an effective date of 12 November 2004, i.e. its filing date.
`
`However, the priority claim of the opposed patent is also invalid pursuant to
`Article 87( 4) EPC.
`
`As discussed above in section 4.1, D1 discloses the same subject-matter as
`claimed in the opposed patent. Importantly, D1 is an earlier patent application of
`the present patentee which precedes the filing of the priority application of the
`opposed patent. D1 has a filing date of 27 June 2002, whereas the priority
`application of the opposed patent has a filing date of 20 November 2003.
`Consequently, the priority application of the opposed patent is not the first
`application of the patentee for this subject-matter in the sense of Article 87(4)
`EPC. The priority claim is therefore invalid. The effective date of all claims is the
`filing date of the application leading to the opposed patent, i.e. 18 November
`2004, which is later than the filing date of D2 (12 November 2014). Therefore,
`D2 constitutes prior art under Article 54(3) EPC for all claims.
`
`D2 discloses soluble pharmaceutical composition for parenteral administration
`with a pH of 7-9, comprising Arg 34
`, Lys 26(NE-(y-Glu(N°-hexadecanoyl)))-GLP-1(7-
`37) and propylene glycol
`(claims 1, 2, 28, 56-58). Examples of such
`(Arg 34
`, Lys 26(NE-(y-Glu(N°(cid:173)
`formulations comprise
`liraglutide
`pharmaceutical
`hexadecanoyl)))-GLP-1(7-37)) and 14 mg/ml propylene glycol, wherein
`the
`formulations have a pH of 7.7 (D2, p3, lines 15-26, p21-22, Examples 2-4). Such
`compositions fall within the respective ranges of claim 1 of the opposed patent
`and are therefore novelty-destroying for the same.
`
`Novo Nordisk A/S Ex. 2006, P. 3
`Fresenius Kabi v. Novo Nordisk
`IPR2022-00657
`
`

`

`Opposition against EP 1 687 019 - Facts and Arguments
`
`Page 4/10
`
`4.3 Dependent claims
`
`Claim 2
`for parenteral
`is suitable
`formulation
`the
`that
`requires
`further
`administration performed by subcutaneous, intramuscular or intravenous injection
`by means of a syringe, optionally a pen-like syringe. This requirement is also met
`by the respective compositions of D1 (see p16, lines 22-25) and D2 (see p17,
`lines 23-25).
`
`Claims 3, 4 and 5 further require that the concentration of propylene glycol is
`from 1-50, 5-25, and 8-16 mg/ml, respectively. The range of claim 3 is disclosed
`in D1 (claims 13 and 14; p19, lines 10-11). In addition, D1 discloses the range of
`claim 5 (p19, lines 12-13), which is also novelty-destroying for the range of claim
`4. Furthermore, the value of 14 mg/ml propylene glycol as disclosed in D2 (see
`above) also falls within the respective ranges of claims 3, 4 and 5.
`
`Claim 6 requires that the pH of the formulation is 7.0 to 9.5. The same range is
`disclosed in D1 (p17, lines 22-23). Also, the compositions of D2 have a pH of 7.7
`(p3, lines 15-26, p21-22, Examples 2-4), which falls within the range of claim 6.
`
`, Lys 26(NE-(y-Glu(N°(cid:173)
`the peptide consists of Arg 34
`that
`requires
`Claim 7
`hexadecanoyl)))-GLP-1(7-37). As discussed above, it is not apparent how this
`would further limit the subject-matter of claim 1. In any event, the compositions
`of D1 and D2 comprise a peptide consisting of Arg 34
`, Lys 26(NE-(y-Glu(N°(cid:173)
`hexadecanoyl)))-GLP-1(7-37); see above.
`
`Claims 8-10 recite pH
`ranges of 7.0 to 8.3, 7.3 to 8.3, and 8.0 to 8.3,
`respectively. D1 teaches a preferred pH range of 7.5 to 8.0 (p17, lines 24-25),
`which falls within the respective ranges of claims 8 and 9. Also, the upper limit of
`pH 8.0 in D1 is novelty destroying for the range of claim 10. Having a pH of 7.7,
`the compositions of D2 fall within the ranges of claims 8 and 9 (see above). Also,
`claim 3 of D2 discloses a pH range of 7.0 to 8.0, the upper limit of which falls
`within the range of claim 10.
`
`Claim 11 further requires that the formulation comprises a preservative. The
`same subject-matter is found in D1 (p18, lines 16-17; claim 11) and D2 (claim
`1).
`
`Claim 12 further requires that the preservative is present in a concentration from
`0.1 mg/ml to 20 mg/ml. D1 discloses the same range for its preservative (see
`p18, lines 23-24; claim 12). The compositions of D2 contain 40 mM phenol (p3,
`(see
`lines 15-26). With a molecular weight of phenol of 94.11 g/mol
`https://en.wikipedia.org/wiki/Phenol), this corresponds to a phenol concentration
`of about 3.8 mg/ml (40*94.11/1000), which also falls within the range of claim
`12.
`
`Claim 13 further requires that the preservative is phenol. The same subject(cid:173)
`matter is disclosed in D1 (p18, lines 21-22) and D2 (p3, lines 15-26; p5, lines 5-
`6).
`
`Claim 14 further requires that the formulation comprises a buffer. According to
`
`Novo Nordisk A/S Ex. 2006, P. 4
`Fresenius Kabi v. Novo Nordisk
`IPR2022-00657
`
`

`

`Opposition against EP 1 687 019 - Facts and Arguments
`
`Page 5/10
`
`claim 15, the buffer is selected from the group consisting of glycylglycine, sodium
`dihydrogen phosphate, disodium hydrogen phosphate, sodium phosphate or
`mixtures thereof. According to claim 16, the buffer is disodium phosphate
`dihydrate. The compositions of D1 also comprise a buffer (see claim 1). The
`buffers of claims 15 and 16 are also disclosed in D1 (p17, lines 27-33). The same
`applies to D2 (p3, lines 15-26; p14, lines 21-23).
`
`Claim 17 further requires that the concentration of liraglutide is from 0.1 mg/ml
`to 50 mg/ml, or from 0.1 mg/ml to 10 mg/ml. Claim 1 of D1 requires that the
`GLP-1 compound is present in a concentration of 0.1-100 mg/ml, the lower limit
`of which falls within both ranges of claim 17. In addition, claim 9 of D1 discloses
`the same ranges of 0.1-50 and 0.1-10 mg/ml. The example compositions of D2
`contain 1.2 mM liraglutide (p3, lines 15-26). With a molar mass of 3751.2 g/mol
`(https://en.wikipedia.org/wiki/Liraglutide),
`this corresponds
`to a
`liraglutide
`concentration of about 4.5 mg/ml (1.2*3751.2/1000), which falls into the range
`of claim 17. D2 also discloses a concentration range for its peptide of 1-25 mg/ml
`(see claim 30), which also falls within in the range of claim 17.
`
`In summary, all dependent claims lack novelty over D1 and D2, respectively.
`
`5.
`
`Lack of inventive step, Article 56 EPC
`
`Irrespective of the foregoing, it is submitted that none of the claims involves an
`inventive step.
`
`5.1 No technical effect over the whole range of the claim
`
`Claim 1 of the opposed patent relates
`formulations of
`to pharmaceutical
`liraglutide comprising 1-100 mg/ml of propylene glycol, at a pH of 7-10. The
`alleged technical effects associated with these features are specified in paragraph
`[0004] of the opposed patent, as reproduced below (emphasis added):
`
`[0004] The present inventors have discovered that peptide formulations containing propylene glycol at certain con(cid:173)
`centrations exhibit reduced deposits 111 production equipment and in the final product and also exhibit reduced clogging
`of injection devices, The present compositions may be formulated with any peptide and are also physically and chemically
`stable thus rendering them shelf-stable and suitable for invasive (eg. injection, subcutaneous injection, intramuscular,
`intraveneous or infusion) as well as non-invasive (eg nasal, oral, pulmonary, transdermal or transmucosal e.g. buccal)
`means of administration.
`
`When examining the experimental data allegedly supporting the reduction in the
`formation of deposits, clogging and/or
`impurities,
`it is apparent that all
`experiments were conducted with compositions comprising several non-claimed
`components. All tested formulations contain a preservative (5.5 mg/ml phenol)
`and a buffer (1.42 mg/ml disodium hydrogen phosphate, dihydrate); see Example
`1, paragraph [0044]; Example 2, paragraph [0057]; Example 3, paragraph
`[0060].
`
`Novo Nordisk A/S Ex. 2006, P. 5
`Fresenius Kabi v. Novo Nordisk
`IPR2022-00657
`
`

`

`Opposition against EP 1 687 019 - Facts and Arguments
`
`Page 6/10
`
`Quite in line with the above observation, the experimental data submitted by the
`patentee in the examination phase likewise exclusively examines formulations
`containing 5.5 mg/ml phenol and 1.42 mg/ml disodium hydrogen phosphate
`buffer. It thus appears that these features are essential for obtaining the alleged
`effect. Their absence from claim 1 suggests that the alleged effect, if any, is not
`obtained over the entire scope of the claims, i.e. in the absence of 5.5 mg/ml
`phenol preservative and 1.42 mg/ml disodium hydrogen phosphate buffer.
`
`Similarly, all experiments in the opposed patent, or submitted by the patentee
`subsequent to filing, examine liraglutide formulations with propylene glycol in a
`narrow concentration range of 13. 7-14.0 mg/ml; see Experiment 1, Table 1 and
`Table 2; Experiment 2, paragraph [0057]; Experiment 3, paragraph [0060].
`According to the opposed patent, this concentration of propylene glycol provides
`isotonic solution; see paragraph [0046], as reproduced below, emphasis
`an
`added. It is therefore not apparent how the broader range of claim 1, which
`isotonic
`extends to up to 100 mg/ml propylene glycol could result in an
`formulation. Likewise, it is not plausible that the alleged effect, if any, extends to
`this broad range.
`
`[00461 An isotonic solution. has an osmolartty of around 0.286 osmollL As can be seen from Table 2 three of the
`formulations (PEG 400, sucrose and xylitol) are more than 20% from being isotonic (0.229-0.343 osmol/I), however for
`these kind of experiments the osmolarity is not expected to influence the results. though, the tonicity of the formulations
`should be adjusted in future experiments.
`
`Table 2. The measured osmolarity of the formulations
`
`Formulation no.
`
`Isotonic agent
`
`Osmolarlty
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`Glucose monohvdrate (38.0 mg/ml)
`
`Laktose monohydrate (65.0 mg/ml)
`
`Maltose (67.2 mg/ml)
`
`Glycine (15.1 mg/ml)
`
`Polyethylenglykol 400 (77.5 mg/ml)
`
`L-arginin(24.6 mg/ml)
`
`Myo-lnositol (35.2 mg/ml)
`
`Propylene glycol (13.7 mgfml)
`
`0.315
`
`0.283
`
`0.306
`
`0.286
`
`0.370
`
`0.318
`
`0.285
`
`0.268
`
`the same
`liraglutide product appears to contain
`the patentee's own
`Also,
`components, at the same concentrations; see D3, p14, third paragraph, emphasis
`added:
`
`2.A.(2).1) Description and composition of the drug product
`The proposed commercial formulation is a clear colourless aqueous solution for injection containing
`6.0 mg/mL of liraglutide as the active pharmaceutical ingredient (API). Other than the API. it contains,
`per ml. a buffering agent (1.42 mg disodium phosphate dihydrate), a preservative (55 mg phenol), an
`isotonic agent (14.0 mg propylene glycol), pH adjusters (2 mol/L hydrochloric acid. 2 mol/L sodium
`is used
`hydroxide), and solvent
`(water for
`
`injection). An overage of t,o of phenol
`
`This further supports the idea that these features are essential in obtaining a
`stable and pharmaceutically acceptable composition. Similar considerations apply
`
`Novo Nordisk A/S Ex. 2006, P. 6
`Fresenius Kabi v. Novo Nordisk
`IPR2022-00657
`
`

`

`Opposition against EP 1 687 019 - Facts and Arguments
`
`Page 7/10
`
`to the pH of the formulation, which is close to pH 8 in all examples (pH 7.9 in
`Examples 1 and 2, see paragraphs [0044], [0057], and pH 8.15 in Example 3, see
`paragraph [0060]). It is not plausible that any technical effect would extend to
`the entire claimed range of pH 7-10. In fact, post-published document D3, which
`relates to the patentee's liraglutide product, suggests that a pH slightly above 8 is
`critical for physical and in-use stability; see p18, fifth paragraph:
`
`A pH of■ was chosen for formulations in early development stage. The pH was changed to 8.15 to
`ensure physical stability and in-use stability at l 0c for patient convenience. In local tolerance studies
`
`For these reasons, it is submitted that any technical effect or benefit relied upon
`by the patentee cannot be present over the entire scope of claim 1.
`
`If the inventive step of a claimed invention is based on a given technical effect,
`the latter should be achievable over the whole area claimed (Case Law 8th ed.,
`I.D.9.8.3). When defining a technical problem for the purposes of the problem(cid:173)
`solution approach, an effect cannot be relied upon if the promised result is not
`achieved throughout the entire range covered by the claimed subject matter.
`Consequently, the technical problem needs to be redefined in a less ambitious
`way (Case Law 8th ed.,
`I.D.4.4.2).
`In this case,
`the problem has to be
`reformulated as the provision of an alternative formulation of liraglutide (see GL
`G-VII 5.2).
`
`On a related note, and by the patentee's own admission, propylene glycol was not
`(only) selected on the basis of an alleged technical benefit, but due to the
`economic/regulatory rationale that "use of propylene glycol would no[t] require
`that further toxicity studies be tested" (see opposed patent, p9, line 32).
`
`5.2 All claims lack inventive step over D1 as closest prior art
`
`To the extent claim 1 should be considered novel, despite the above observations,
`D1 is a promising starting point for a development leading to the claimed subject(cid:173)
`matter. D1 relates to the same field of pharmaceutical formulations of GLP-1
`compounds, in particular liraglutide (see abstract; p4, lines 13-14; claim 1). In
`addition, D1 relates to the same purpose or objective of providing pharmaceutical
`formulations with improved stability (see title; p3, lines 17-20). Just like the
`opposed patent, D1 is also concerned with avoiding aggregation, precipitation and
`adsorption of the protein/peptide to surfaces; see p2, lines 12-20, reproduced
`below, emphasis added:
`
`Novo Nordisk A/S Ex. 2006, P. 7
`Fresenius Kabi v. Novo Nordisk
`IPR2022-00657
`
`

`

`Opposition against EP 1 687 019 - Facts and Arguments
`
`Page 8/10
`
`It is an important technical challenge to ensure prolonged stability during storage
`(shelf life) of many protein based drug products due to the inherent !ability of macromole(cid:173)
`cules. Hence, protein• are aenaltlVe to bo4h chemical and physical degradation unlike many
`small molecufet. Chemical degradation involves covalent bonds, such as hydrolysis, racemi(cid:173)
`zation, oxidation or crosslinking. Physical degradation involves conformational changes rela(cid:173)
`tive to the native structure, which includes loss of higher order structure, aggregation. pl"eclpi•
`tatton or adsorption to eurfacea. GlP-1 ia. known to be prone to instability due to aggregation.
`Both degradation pathways may ultimately lead to loss of biological activity of the protein
`drug.
`
`As discussed above, there is no distinguishing feature of any of claims 1-17 over
`D1. During examination, the patentee argued that the choice of propylene glycol
`would be non-obvious over the cited prior art. However, propylene glycol is
`already disclosed in D1 (p18, line 35). Even if this were not the case, claim 1
`would lack inventive step over D1.
`
`In accordance with the objective technical problem of providing an alternative
`formulation of liraglutide, as discussed above in section 5.1, the skilled person
`would have been aware of document D4. D4 relates to parenteral peptide
`formulations in aqueous solution (see title). According to D4, the addition of
`propylene glycol suppresses protein gelation and destabilizes liquid crystals, i.e.
`minimizes aggregation of the protein LHRH and
`its analogues nafarelin and
`detirelix (see abstract). At higher concentrations, aqueous
`formulations of
`nafarelin or detirelix undergo peptide aggregation and solution gelation resulting
`in compromised physical stability; p1258, introduction, first paragraph. This is the
`same problem the opposed patent purports to solve.
`
`D4 concludes as follows; see p1262, last paragraph, emphasis added:
`
`LHRH has little effect on chemical stability. In contrast, the
`hydrophobic nature of certain LHRH decapeptidcs such as
`nafarelin and detirelix gives rise to unusual solution dynam(cid:173)
`ics, including peptide aareaation and liquid crystal forma(cid:173)
`tion. These liquid crystals eventually result in solution gela(cid:173)
`tion, with the ramification of compromised formulation ele(cid:173)
`gance and utility. We have demonstrated herein that a
`pharmaceutically acceptable cosolvent, propylene glycol.
`raises the minimum concentration for peptide aggregation,
`and thus, many of the potential problems associated with
`aaareaate formation may be minimized or eradicated by co(cid:173)
`solvent addition.
`
`It would therefore have been obvious to add propylene glycol to the aqueous
`formulation of D1 for obtaining an alternative formulation of liraglutide. Even if
`any aggregation-related and/or stability-related technical effect would have to be
`considered, such combination would have been obvious because D4 teaches that
`these exact benefits are obtained by using propylene glycol.
`
`Novo Nordisk A/S Ex. 2006, P. 8
`Fresenius Kabi v. Novo Nordisk
`IPR2022-00657
`
`

`

`Opposition against EP 1 687 019 - Facts and Arguments
`
`Page 9/10
`
`Alternatively, the skilled person, starting from D1, would have been aware of
`document D5, which relates to GLP-1 peptide mimics that mimic the biological
`activity of the native GLP-1 peptide (abstract). According to D5, propylene glycol
`is a suitable carrier for parenteral solutions of such peptides (p89, lines 26-29).
`Such parenteral composition may contain the active ingredient, propylene glycol
`and water (p91, lines 4-7). D5 also mentions liraglutide for use in combination
`with the disclosed GLP-1 mimics (NN-2211, p77, line 32). It would therefore have
`been obvious to choose propylene glycol for obtaining an alternative parenteral
`formulation of liraglutide.
`
`In another alternative, the skilled person, starting from D1, would have consulted
`document D6, which discloses pharmaceutical formulations of the protein ciliary
`neurotrophic factor (CNTF) suitable for intrathecal or parenteral administration
`(abstract; p13, lines 28-31). According to D6, CNTF in solution tends to be
`unstable due to precipitation or adsorption to surface areas of storage containers
`and dispensing devices (p3, lines 16-32). Therefore, propylene glycol is added to
`prevent precipitation of CNTF in solution (p11, lines 10-12). Propylene glycol was
`also found to stabilize CNTF against precipitation from agitation while having little
`effect on thermal stability (p22, lines 32-35). Consequently, the skilled person
`would have chosen propylene glycol for obtaining an alternative, and even more
`stable, parenteral formulation of liraglutide based on the combined teachings of
`D1 and D6 without using inventive skill.
`
`For these reasons, claim 1 does not involve an inventive step when starting from
`D1 as closest prior art. The same reasoning applies to each of the dependent
`claims since all their features are likewise found in D1 (see section 4.3 above).
`
`5.3 Reasonable expectation of success for using propylene glycol
`
`During the examination proceedings leading to the opposed patent, the patentee
`has argued that propylene glycol is superior to other known isotonic agents in
`terms of providing stable and soluble liraglutide formulations. As discussed above,
`any such effect does not plausibly apply over the entire range of claim 1. Even if
`it did, the claims would lack inventive step in view of a reasonable expectation of
`success of the skilled person in using propylene glycol to that effect.
`
`According to established case law, a given course of action is considered obvious
`if the skilled person would have carried it out in expectation of some improvement
`or advantage (Case Law 8th ed., I.D.7.1). Thus, a finding of obviousness does not
`presuppose that the results are clearly predictable; a reasonable expectation of
`success
`is sufficient (T 149/93). Absolute certainty
`is not required
`for a
`reasonable expectation of success (Case Law 8th ed., I.D.7.1).
`
`Board of Appeal Decision T 235/97, submitted herein as document D7, concerns a
`case quite similar to the present one. The patent in question related to a
`pharmaceutical composition comprising an aqueous solution of erythropoietin
`(EPO) and a cyclodextrin derivative for stabilizing EPO. The closest prior art
`
`Novo Nordisk A/S Ex. 2006, P. 9
`Fresenius Kabi v. Novo Nordisk
`IPR2022-00657
`
`

`

`Opposition against EP 1 687 019 - Facts and Arguments
`
`Page 10/10
`
`disclosed aqueous EPO solutions stabilized by polyethylene glycol, protein, sugar,
`amino acid, inorganic salt organic salt or sulfur-containing reducing agents (see
`reasons 4.4, second paragraph). Thus, the objective technical problem was the
`provision of improved EPO formulations with an overall better long-term stability
`in aqueous solution (see reasons 4.4, last paragraph).
`
`According to the Board, the skilled person would take into account prior art
`relating in general to stabilizers and/or solubilizers for parenteral formulations
`containing particular proteins and/or peptides (D7, reasons 4.9, first paragraph).
`A secondary reference, document (7), disclosed the use of modified cyclodextrin
`as stabilizers and/or solubilizers in parenteral formulations of various drugs, in
`particular proteins and peptides (reasons 4.9, second paragraph). Although the
`secondary document related to inhibiting protein aggregation of insulin, which is
`not a glycosylated protein like EPO, the Board found that the claimed subject(cid:173)
`matter was obvious in view of a reasonable expectation of success in testing the
`same approach with EPO (reasons 4.10, third paragraph).
`
`is similar to the present situation
`This
`in which the secondary reference,
`document D4, teaches that propylene glycol minimizes aggregation of the
`proteins LHRH, nafarelin and detirelix. Although GLP-1 analogues or liraglutide are
`not mentioned, D4 does suggest a broader application of propylene glycol to that
`effect (p1261, left column):
`
`fracture replicates of emulsions and creams (20). Peptide
`aggregation in other peptides and proteins has also been re-
`and insulin (22).
`ported, for example, for pentagastrin
`
`And p1262, conclusions:
`
`and utility. We have demonstrated herein that a
`pharmaceutically acceptable cosolvent, propylene glycol,
`the minimum concentration for peptide aggregation,
`and thus, many of the potential problems associated with
`...... ,.,,.n<>T,,. formation may minimized or eradicated by co-
`solvent addition.
`
`In summary, the skilled person would have reasonably expected similar effects as
`described in D4 when including propylene glycol in a formulation of liraglutide.
`This is all the more so since, in the present case and unlike T 235/97, the closest
`prior art D1
`itself already teaches the use propylene glycol (see above).
`Therefore, no inventive step is required to arrive at this subject-matter.
`
`Dr. Ulrike Rink
`
`European Patent Attorney
`
`Association 370
`
`Novo Nordisk A/S Ex. 2006, P. 10
`Fresenius Kabi v. Novo Nordisk
`IPR2022-00657
`
`