Gary B Cox v MacroGenics, Inc.

[2019] APO 13

IP AUSTRALIA

AUSTRALIAN PATENT OFFICE

Gary B Cox v MacroGenics, Inc. [2019] APO 13

Patent Application:                2012259162

Title:Deimmunized serum-binding domains and their use for extending serum half-life

Patent Applicant:                   MacroGenics, Inc.

Opponent:  Gary B Cox

Delegate:  Dr S. J. Smith

Decision Date:  27 March 2019

Hearing Date:  28 November 2018, in Canberra

Catchwords:  PATENTS – opposition to the grant of a patent – grounds of clear enough and complete enough disclosure, support and utility – principle of general application – plausibility – undue burden – contribution to the art – specification does not comply with s 40(2)(a) or s 40(3) – lack of utility not established – costs awarded

Representation:  Patent attorney for the applicant: Dr David Loch of Spruson & Ferguson

Counsel for the opponent: Mr Andrew Fox

Patent attorney for the opponent: Mr Craig Humphris of Wrays

IP AUSTRALIA

AUSTRALIAN PATENT OFFICE

Patent Application:                2012259162

Title:Deimmunized serum-binding domains and their use for extending serum half-life

Patent Applicant:                   MacroGenics, Inc.

Date of Decision:                   27 March 2019

DECISION

The specification does not comply with section 40(2)(a) or section 40(3). MacroGenics, Inc. is allowed two months from the date of this decision to file amendments to overcome these deficiencies.

I award costs according to Schedule 8 against MacroGenics, Inc.

REASONS FOR DECISION

Background

1. Patent application 2012259162 was filed by MacroGenics, Inc. (the Applicant) on 16 May 2012 under the provisions of the Patent Cooperation Treaty.  The application claims priority from US 61/488,725 which was filed on 21 May 2011. 

2. Acceptance of the application was advertised on 3 November 2016.  Gary B Cox (the Opponent) filed a notice of opposition under section 59 of the Patents Act 1990 (the Act) on 3 February 2017.

3. The applicant proposed post-acceptance amendments to the specification on 6 November 2017 which were allowed unopposed on 23 February 2018 and this decision is in relation to the specification as amended.

4. The statement of grounds and particulars was filed on 2 May 2017, a first amendment to the statement was allowed on 16 August 2017 and a further amendment allowed on 13 April 2018, following the amendment to the specification.  The statement as amended identifies three grounds of opposition: clear enough and complete enough disclosure, support and utility.  I note that by filing an early amendment to the claims and subsequently to the statement of grounds and particulars the parties have assisted in focusing the opposition on the matters now in dispute.

   Evidence

5. The evidence filed during the evidentiary periods is summarised in the table below:

Evidence	Declarant	Exhibits	Declaration Date	Reference
In Support	Peter Hudson	PH-1 to PH-12	4 August 2017	Hudson 1
In Answer	Matthew Baker	MB-1 to MB-10	29 November 2017	Baker 1
In Reply	Peter Hudson	-	7 February 2018	Hudson 2

1. Following completion of the evidentiary periods, the Applicant filed a further declaration by Matthew Baker dated 26 June 2018 (Baker 2) and requested that it be considered under the provisions of regulation 5.23.  I informed the parties that I would consider paragraphs 1-14 of that declaration under regulation 5.23 and provided the opponent with an opportunity to file responding evidence.  The Opponent duly filed on 28 September 2018 a further declaration by Peter Hudson dated 28 September 2018 (Hudson 3).

2. At the hearing the Applicant provided further information for consideration under regulation 5.23, consisting of a report describing the in silico approach to predicting MHC class II binding peptides that was undertaken and the identification of the five p1 anchor positions (of 9mers with high MHC class II binding potential) reported in the specification.

3. The operation of regulation 5.23 was considered in Merial Limited v Bayer Intellectual Property GmbH[1]:

   “I conclude that a decision under regulation 5.23 must have regard to the nature of the information and whether the information is likely, if not certain, to change the outcome of the opposition in a significant way.”

   Subsequent decisions have elaborated on other relevant factors to consider in relation to regulation 5.23,[2] but for the present purposes a consideration of the nature and relevance of the information is sufficient to make a decision. 

   [1] [2015] APO 16 at [24].

   [2] Reflex Instruments Asia Pacific Pty Ltd v Minnovare Limited [2017] APO 8; Intervet International B.V. v E.I. du Pont de Nemours and Company [2017] APO 11.

4. I indicated to the parties at the hearing that my preliminary view was that I would not consider this material under regulation 5.23.  That is also my concluded position.  The information provided was not present in the specification as filed, but as will become apparent, I consider that the conclusion that the p1 residue positions identified in the specification are the only such residues in SEQ ID NO: 304 can be derived from the specification on its face.  Furthermore, while the data underpinning the conclusions in the specification may be of interest, I do not consider that it has the potential to significantly change the outcome of the opposition in the present case where the relevant grounds are based principally on the content of the specification and what it teaches a skilled person.  

   Onus

5. The request for examination of the opposed application was filed on 26 November 2015.  Consequently, the amendments of the Act brought about by the Intellectual Property Laws Amendment (Raising the Bar) Act 2012 (the Raising the Bar Act) apply to the present application. This includes section 60(3A) of the Act, which provides that the Commissioner may refuse an application if satisfied on the balance of probabilities that a ground of opposition exists.  It is the opponent who carries the onus of proof.

   The specification

6. The application is titled “Deimmunized serum-binding domains and their use for extending serum half-life” and the claims are directed to a polypeptide (particularly an antigen-binding molecule) comprising a portion of a deimmunized serum-binding protein, wherein the serum-binding protein portion extends the serum half-life of the polypeptide relative to the polypeptide lacking the deimmunized serum-binding protein portion.  Much of the disclosure relates to the preparation and investigation of diabodies.  The specification includes a sequence listing and ends with 66 figures and 21 claims.

7. Before construing the specification, I note the comments of Middleton J in Eli Lilly and Company Limited v Apotex Pty Ltd[3]:

   “It is well settled that the Court should, from the outset, approach the task of patent construction with a generous measure of common sense. The Court must place itself in the position of a person skilled in the relevant art, being the subject matter of the patent. From this perspective, the patent is to be read as a whole, in the context of the specification and in light of the prevailing common general knowledge and state of the relevant art at the priority date.”

   [3] [2013] FCA 214; 100 IPR 451 at [139].

   The person skilled in the art

8. It is well established that many of the issues in an opposition are answered by reference to the person skilled in the art:

   “He is the person to whom the patent is addressed and who must construe it.  He is the person whose knowledge will determine whether a patent is novel.  He is the person who will judge whether a patent is obvious.”[4]

   [4] Root Quality Pty Ltd v Root Control Technologies Pty Ltd [2000] FCA 980 at [70].

9. The hypothetical skilled person works in the field with which the invention is connected, and is a non-inventive person or team likely to have a practical interest in the subject matter of the invention.[5]  As discussed above, the present specification is directed primarily to antigen-binding molecules and deimmunized serum-binding proteins, and therefore the person skilled in the art will have experience or a practical interest in this field. 

   [5] Root Quality Pty Ltd v Root Control Technologies Pty Ltd [2000] FCA 980 at [70]-[72].

10. Dr Hudson has significant experience in the development of antibodies and antibody-like molecules, including modification of affinity and pharmacokinetic properties.  At the priority date Dr Hudson had experience with pegylation of diabodies to increase blood circulation time.[6]

    [6] Hudson 1 at [28]-[32].

11. Dr Baker has extensive experience in the biopharmaceutical industry in antibody design.  He has developed and applied novel antibody humanisation techniques, and has worked extensively in the area of protein deimmunization.[7]  Dr Baker is also an inventor of the EpiScreen assay, an in vitro T cell response assay.[8]

    [7] Baker 1 at [8].

    [8] Baker 1 at [18]

12. The Applicant submitted that Dr Hudson lacks first-hand practical experience with respect to albumin binding domains and deimmunizing proteins by targeting T cell epitopes.  The Opponent submitted that Dr Baker, as an inventor of the EpiScreen assay, is over-skilled, such that his views do not reflect those of the notional skilled addressee.  I consider that the evidence supports a conclusion that Dr Baker has extensive practical experience with respect to deimmunization and immunogenicity assays relative to Dr Hudson.[9]  It is also clear that Dr Hudson and Dr Baker are both highly skilled and I am satisfied that both declarants have backgrounds that enable them to understand the specification and provide evidence in relation to what a person skilled in the art knew or would have understood at the relevant date.  Where there is conflicting evidence I will decide which evidence should be given greater weight, bearing in mind the parties’ submissions.

    [9] Baker 1 at [8]; exhibit MB-1; Hudson 1 at [31]-[33]; exhibit PH-3; Hudson 3 at [33].

    The invention as described

    Background of the invention

13. The specification indicates that the invention described is directed to a polypeptide comprising a polypeptide portion of a deimmunized serum-binding protein capable of binding to the serum-binding protein, thereby extending the serum half-life of the polypeptide.[10]  The invention is particularly directed to diabody molecules (also termed “dual affinity retargeting reagents” or “DARTS”) that comprise at least two polypeptide chains that associate to form at least two epitope binding sites, recognising the same or different epitopes.[11]  The specification goes on to describe prior art relating to diabodies and effector cell receptors.[12]

    Nature of the invention

    [10] [0003].

    [11] [0004].

    [12] [0005]-[0015].

14. The specification then states that:

    “The present invention particularly relates to covalent diabodies and/or covalent diabody molecules and to their use in the treatment of a variety of diseases and disorders including cancer, autoimmune disorders, allergy disorders and infectious diseases caused by bacteria, fungi or viruses.  Preferably, the diabody of the present invention can bind to two different epitopes on two different cells wherein the first epitope is expressed on a different cell type than the second epitope, such that the diabody can bring the two cells together.

    In detail, the invention provides a polypeptide (and especially an antigen-binding molecule) comprising a polypeptide chain that comprises a portion of a deimmunized serum-binding protein capable of binding to said serum protein; wherein said serum-binding protein portion extends the serum half-life of said polypeptide, relative to the serum half-life of said polypeptide if lacking said portion of said deimmunized serum-binding protein.

    The invention also concerns the uses of (or methods involving the use of) such polypeptide portion of a deimmunized serum-binding protein to extend the serum half-life of a polypeptide covalently linked thereto, the extension of serum half-life being relative to the serum half-life of the antigen-binding molecule if lacking the albumin binding protein.

    The invention additionally concerns the embodiment of the above-described polypeptide or uses, wherein the polypeptide chain comprises an additional portion of a deimmunized serum-binding protein, wherein the portions of the deimmunized serum-binding protein are both capable of binding to the serum protein.

    The invention particularly concerns the embodiment of the above-described polypeptides or uses wherein the polypeptide portion of the deimmunized serum binding protein is an albumin-binding domain (ABD) of a Streptococcal Protein G, or albumin-binding variant thereof, and especially wherein the albumin-binding domain (ABD) of the Streptococcal Protein G is an albumin-binding domain 3 (ABD3) of protein G of Streptococcus strain G148.”[13]

    [13] [0016]-[0020].

15. The specification further discloses a method for increasing the half-life of a molecule comprising an antigen-binding domain of an antibody, wherein the method comprises linking two or more albumin-binding domains to the molecule.[14] 

    [14] [0049].

16. The description of preferred embodiments of the invention is largely directed to the nature of relevant diabodies and conjugation and characterisation thereof.  The specification indicates that the molecules of the invention are useful for the treatment/prevention of diseases, including those in which an effector cell function mediated by FcγR is desired (such as cancer or infectious diseases).  Specifically, by comprising an epitope binding domain that immunospecifcally recognises an FcγR, a diabody may bind a cell surface antigen and an FcγR on an immune effector cell (for example, an NK cell) stimulating an effector function against the cell.[15] 

    [15] [00327].

17. While the process of deimmunization is not described in detail in the specification, before considering the examples it is useful to set out some background information regarding protein deimmunization.  At a general level, this is a process whereby immunogenic sequences which would elicit an immune response, with associated detrimental outcomes, are removed or modified.[16]  Dr Baker explained that a person skilled in the art would understand the following with respect to protein deimmunization at the priority date:

    “the most common practise for deimmunizing complex protein therapeutics comprised an initial in silico analysis using major histocompatibility complex (MHC; i.e., the cell surface proteins that function to bind peptide fragments derived from pathogens and display them on the surface of antigen presenting cells for recognition by the appropriate T cells) class II peptide binding prediction algorithms of the primary sequence to identify potential MHC class II binding 9mer peptides.  The next stage in this process would typically then involve designing mutations to reduce the number of MHC class II binding alleles by targeting the most important pocket residues contained in the 9mer MHC class II binding sequence.  For HLA-DR (Human Leukocyte Antigen – antigen D Related, which accounts for >80% of MHC class II molecules on the surface of an antigen presenting cell (APC)) pocket residues that typically contribute most to binding to the MHC class II binding groove are in order from the greatest to smallest: pocket 1 (P1)>P9>P4=P6>P7.

    It was also known as of the Priority Date that the MHC class II groove is open ended and can accommodate long peptides, typically 25mers, and that there are known T-cell receptor (TCR) contact residues that lie just outside of the core 9mer (e,g., the P1 and P11 positions).  These positions as well as TCR contact residues in the core 9mer (P2 and P5) can also be targeted to reduce TCR binding, if mutation of the MHC class II pocket residues were to destroy the activity of the protein to be deimmunized.

    Thus, key to the process of deimmunization is selection of mutating residues … this step typically required multiple in vitro assay screening stages to empirically confirm that each mutating residue and combinations of mutating residues that reduce MHC class II binding do not deleteriously affect biological activity of the overall molecule.  Finally, an in vitro T cell assay was also typically performed on a small number of lead deimmunized sequences to confirm the lack of T cell stimulation.”[17]

    The Examples

    [16] Hudson 1 at [64]-[65].

    [17] Baker 1 at [9]-[11]. I note that this explanation is consistent with the 2007 publication exhibited to Hudson 1 (exhibit PH-9): Baker, M.P. & Jones, T.D. (2007) ‘Identification and removal of immunogenicity in therapeutic proteins’,  Current Opinion in Drug Discovery & Development, vol. 10, pages 219-227.

18. The description ends with 24 examples.  Many of these are directed to the preparation and properties of DARTs with various binding specificities.  Conjugation of DART molecules to the albumin-binding domain 3 (ABD3) from protein G of Streptococcus strain G148 (SEQ ID NO: 304), which consists of 46 amino acid residues forming a stable three-helix bundle, is exemplified and results in an increase in serum half-life.[18]  The specification indicates that an ABD may be introduced at one or more termini of a DART molecule, with the C-terminus preferred.[19] 

    [18] Example 6.18, [00613]-[00634]; Example 6.20, [00643]-[00649].

    [19] [00615].

19. Of particular interest with respect to the claimed invention are Examples 6.21-6.24. 

20. Example 6.21 describes amino acid modifications introduced into residues 60-79 of the amino acid sequence of ABD3 (SEQ ID NO: 304) and the effect thereof on binding to human serum albumin (HSA).  Table 20, reproduced below, recites five identified p1 anchors, Y60, Y61, L64, I65 and V71.  Dr Baker provided a graphical depiction of the 9mers in SEQ ID NO: 304 based on these P1 anchor positions:[20]

    [20] Baker 1 at [24].

21. Dr Baker explained in relation to this figure:

    “The coloured lines above indicate the position of the 5 core 9mers based on identification of the MHC class II P1 anchor residue.  Since the core 9mers overlap, the mutation strategy of the Opposed Application involved identifying point mutations that could disrupt MHC class II binding to more than one core 9mer, thus limiting the total number of required mutations in the ABD and thereby increasing the probability that the mutated ABD will retain a desirable biological function.  In view of the disclosure of the Opposed Application, this assessment could, in my opinion, have been performed by a person skilled in the art as of the Priority Date.”[21]

    [21] Baker 1 at [25].

22. The means of identification of these anchor positions is not referred to in Example 6.21, but it is subsequently indicated in Example 6.23 that in silico analysis revealed the presence of five such positions:

    “In silico analysis revealed that there are five p1 anchor positions with high MHC class II binding potential in the ABD sequence.  In order to produce deimmunized version of ABD, site directed mutations were performed based on the prediction provided by the in silico analysis.”[22]

    [22] [00675].

23. While the location of this statement, following the examples in which the mutations were investigated, does not assist in understanding the disclosure, I do not accept the Opponent’s submission that the specification does not disclose that the full SEQ ID NO: 304 was characterised in silico for the presence of MHC class II binding peptides.  I consider that the clear inference from this passage is that “the ABD sequence”, rather than a fragment thereof, was subject to in silico analysis, and that the five p1 anchor positions referred to are those identified in Example 6.21.  This appears to be in accord with Dr Baker’s reading of the specification: “the Opposed Application clearly identifies that five core 9mer peptides were identified in the ABD3 sequence of SEQ ID No. 304”[23] and is also consistent with the Applicant’s submission that, with an object to “decrease the immunogenicity of the ABD”,[24] it would be surprising for parts of the sequence to be ignored.[25] 

    [23] Baker 1 at [14].

    [24] [00651].

    [25] Applicant’s written submissions at [87].

1. While I accept that Dr Hudson found it unclear from the specification which residues were postulated to be in the p1 position,[26] I prefer Dr Baker’s evidence with regard to the information derivable from Example 6.21.[27]  His conclusions seem more consistent with reading with a view to making sense of the, admittedly somewhat meandering, specification, and, as noted above, I consider that Dr Baker has more practical experience in relation to protein deimmunization than Dr Hudson.

   [26] Hudson 1 at [95].

   [27] Baker 1 at [23]-[26], [51].

2. In Example 6.21, particular mutations that would be expected to remove or reduce MHC class II binding are described:

   “Peptides from Y60 and Y61 share a common pocket residue at N66 (P6/P7) and also L64 and I65 share a P6/P7 pocket at T70.  Mutation of N66 to D or E in combination with T70D was therefore postulated to have the ability to effectively remove MHC class II binding.  Alternatively, it was postulated that MHC class II binding could be removed by targeting the four p1 anchors Y60, Y61, L64 and I65.”[28]

   [28] [00652].

3. A two round process of mutational analysis was undertaken, with the results of the first round shown in Table 20, and a summary of the results from both rounds shown in Table 23.

4. It is apparent that Table 20 provides information regarding both immunogenicity and HSA binding, noting that HSA binding information relates to the retention of function in the mutant but does not provide any information in relation to its immunogenicity.  Table 23 provides information only in relation to HSA binding.  I note that the effect of single mutations between residues 60 and 79 on HSA binding is reported in Table 19 of the specification, and Table 23 indicates that mutations that individually result in substantial retention of HSA binding do not necessarily result in retention of HSA binding when combined.  For example, the Y61A, I65A and V71A mutations individually result in retention of HSA binding affinity at wildtype level, but the combination of these three mutations leads to minimal binding affinity. 

5. Example 6.21 concludes with reference to figures depicting the albumin binding ability of a number of double and triple mutants of the ABD of SEQ ID NO: 304. 

6. Example 6.22 indicates that ABD variants, particularly variants at V71, exhibit decreased immunogenicity, and reports the testing of three 21mer peptides (consisting of residues 60 to 80 of the ABD peptide) for immunogenic potential using EpiScreen^TM time course T cell assays: wild-type ABD, N66D/T70S/V71A ABD (DSA) and L64A/I65A/V71A ABD (AAA).  Both T cell proliferation and IL-2 secretion (ELISpot assay) were measured.  These assays were based on a cohort of 50 donors whose allotypes indicated that coverage of >80% of world allotypes was achieved and all major HLA-DR allotypes were well represented.  In these assays samples inducing responses of greater than or equal to twice an empirical threshold of a stimulation index (SI) (SI≥2.00) were deemed positive.  A summary of the magnitude of positive T cell proliferation responses is shown in Table 24 and a summary of the magnitude of positive T cell IL-2 secretion responses is shown in Table 25.  The mean SI values reported in the tables correspond to the average of all positive donor responses observed, and include borderline responses (SI≥1.90, p<0.05).[29]

   [29] [00669], [00672].

7. There is an inconsistency in this example.  With respect to the data presented in Table 24 the specification states: “The mean magnitude (SI) of positive (SI≥ 2.00 p< 0.05) T cell proliferation responses against the wt peptide was SI 2.55, which was slightly higher than the two deimmunized peptides (Table 24).”[30]  Clearly, Table 24 indicates that the DSA mutant has a higher mean SI than the wildtype peptide.  Dr Hudson stated that the conclusion drawn in the text was “not appropriate given the data reported”,[31] while Dr Baker considers that there is a typographical error and the text should refer to the % response.[32]  Ultimately, as will become apparent, it is not necessary that I form a concluded view on how this inconsistency in the specification should be resolved.  It is apparent that in both measures reported, the mean SI values for both the wildtype and mutant ABDs are quite similar,[33] but the % response is reduced in both mutant ABDs in both assays.  From these results the specification concludes that the overall immunogenic potential of the wildtype peptide is moderate, while that of the mutants is low:[34]

   “Data from the study of the wt peptide had indicated that the overall potential risk of immunogenicity for the wt peptide was moderate as the combined frequency of proliferation and IL-2 responses was 14% of the study cohort.  However the combined frequency of proliferation and IL-2 responses to the two deimmunized peptides in this study was low with 4% of donors responding against N66D/T70S/V71A and 6% of donors responding against L64A/I65A/V71A.”[35]

   [30] [00669].

   [31] Hudson 1 at [117].

   [32] Baker 1 at [53].

   [33] Baker 1 at [53].

   [34] [00670], [00672]-[00674].

   [35] [00674].

8. With respect to the assays adopted, the specification indicates that:

   “EpiScreen^TM time course T cell assays of multiple commercial biologics have shown a clear correlation between the percentage of donor T cell responses in the EpiScreen^TM assay and the level of immunogenicity (anti-protein therapeutic antibody responses) observed in the clinic. … In general, protein therapeutics that induce >10% positive response in the EpiScreen^TM assay are associated with a significant risk of immunogenicity in the clinic.  In comparison to protein therapeutics tested in EpiScreen^TM assays, the data shows that the deimmunized peptides N66D/T70S/V71A and L64A/I65A/V71A give CD4⁺ T cell responses which fall into the same range as Xolair, Herceptin and Avastin, and would be considered as having a low potential risk of immunogenicity.  Therefore, it is concluded that the deimmunized peptides exhibit a favourable immunogenicity profile in the EpiScreen^TM immunogenicity assay.”[36]

   [36] [00674].

9. Example 6.23 provides an in silico assessment of the AAA and DSA mutants described in the previous example as having decreased immunogenicity:

10. This example goes on to describe the serum half-life of hCD16xhCD32B ABD (AAA) DART and hCD16xhCD32B ABD (DSA) DART.  The DSA mutant results in a similar half-life extension to the wildtype ABD and the AAA mutant gives rise to an increased half-life relative to the unmodified DART.

11. Example 6.24 describes the effect of increased valency of ABD on the half-life of recombinant ABD-DARTs.  Specifically, DART proteins containing multiple ABDs were prepared, either by fusion of ABDs at the C-terminal of both chains or by fusing tandem ABDs at the end of one chain.  The half-life of the DART was extended to a greater extent in the presence of more than one ABD.[37]

    [37] [00678].

    The invention as claimed

12. The correct approach to the construction of claims was discussed by Bennett J in H Lundbeck A/S v Alphapharm Pty Ltd[38]:

    “the words in a claim should be read through the eyes of the skilled addressee in the context in which they appear. … While the claims define the monopoly claimed in the words of the patentee’s choosing, the specification should be read as a whole … It is not permissible to read into a claim an additional integer or limitation to vary or qualify the claim by reference to the body of the specification … terms in the claim which are unclear may be defined or clarified by reference to the body of the specification.”

    [38] [2009] FCAFC 70; 81 IPR 228 at 254, [118]-[120].

13. The entire claim set is reproduced at Annex A.  Claim 1 reads:

    A polypeptide that comprises a portion of a deimmunized albumin-binding protein capable of binding to serum albumin, wherein said deimmunized albumin-binding protein portion is a variant of a wild-type albumin-binding domain (ABD) of a Streptococcal Protein G, said wild-type ABD having the amino acid sequence of SEQ ID NO: 304; wherein said variant ABD has an amino acid sequence that differs from that of SEQ ID NO: 304 in comprising:

    (A)a leucine to alanine variation at ABD position 64, an isoleucine to alanine variation at ABD position 65, and a valine to alanine variation at ABD position 71; wherein said variant ABD positions are respectively positions 24, 25, and 31 of SEQ ID NO: 304; or

    (B)an asparagine to aspartic acid variation at ABD position 66, a threonine to serine substitution at ABD position 70, and a valine to alanine variation at ABD position 71; wherein said variant ABD positions are respectively positions 26, 30, and 31 of SEQ ID NO: 304;

    wherein said deimmunized albumin-binding protein portion extends the serum half-life of said polypeptide, relative to the serum half-life of said polypeptide if lacking said portion of said deimmunized albumin-binding protein.

14. The only other independent claim is claim 15, which reads:

    Use of a polypeptide that is a portion of a deimmunized albumin-binding protein capable of binding to serum albumin to extend the serum half-life of a covalently linked polypeptide, wherein said deimmunized albumin-binding protein portion is a variant of a wild-type albumin-binding domain (ABD) of a Streptococcal Protein G, said wild-type ABD having the amino acid sequence of SEQ ID NO: 304; wherein said variant ABD has an amino acid sequence that differs from that of SEQ ID NO: 304 in comprising:

    (A)a leucine to alanine variation at ABD position 64, an isoleucine to alanine variation at ABD position 65, and a valine to alanine variation at ABD position 71; wherein said variant ABD positions are respectively positions 24, 25, and 31 of SEQ ID NO: 304; or

    (B)an asparagine to aspartic acid variation at ABD position 66, a threonine to serine substitution at ABD position 70, and a valine to alanine variation at ABD position 71; wherein said variant ABD positions are respectively positions 26, 30, and 31 of SEQ ID NO: 304;

    wherein said extension of serum half-life is relative to the serum half-life of said covalently linked polypeptide if lacking said portion of said deimmunized albumin-binding protein.

15. At a high level, I understand the claims to be directed to fusion proteins comprising a broadly defined first polypeptide covalently linked to a variant ABD, such that the serum half-life of the resulting fusion protein is extended relative to that of the protein absent the ABD.  Notably, the claim includes limitations with respect to deimmunization and serum half-life, such that embodiments in which the ABD is not deimmunized or the half-life of the polypeptide is not extended do not fall within the scope of the claims.  In the context of the specification, it is clear that the variant ABD portion of the polypeptide must be located at either the C- or N-terminal of the first polypeptide. 

16. The distinction between the independent claims seems to be subtle.  While typically a “use of” claim is understood as directed to a method or process, in this case, what is defined is the use of an ABD to extend the half-life of a covalently linked polypeptide, i.e. a polypeptide to which the ABD is covalently linked with an attendant increase in half-life.  To my mind this would seem to encompass the polypeptide conjugate per se.  However, given that the claim is drafted as a use claim, despite it not reciting an explicit synthetic or preparative step I think it is reasonable that this claim also be understood to extend to the preparation of the polypeptide conjugate (that is, the use of the ABD to extend the half-life by covalent linkage to the polypeptide).

    Polypeptide

17. The specification does not provide an explicit definition of polypeptide.  However, this is a well-understood term, and I consider that it includes within its scope diabodies and antigen-binding polypeptides, but also any other polypeptide, for example, structural and signalling proteins.  This is consistent with the evidence of both declarants.[39]

    Portion

    [39] Hudson 1 at [78]; Baker 1 at [19].

18. In the context of the claims and the specification I understand portion as indicating that SEQ ID NO: 304 is a portion (albumin-binding domain 3) of an albumin-binding protein, and it is a requirement of the claim that albumin-binding portion of the polypeptide is deimmunized.

    Deimmunized

19. As an initial observation, I note that the use of “deimmunized” in the claim refers only to the albumin-binding protein portion – the polypeptide as a whole is not required to be deimmunized.

20. There is no specific definition of “deimmunized” in the specification, but Dr Hudson and Dr Baker agreed that this would be understood to mean “that amino acid substitutions have been made across a polypeptide chain to diminish significantly the potential of its proteolytically processed peptides to bind to MHC class II”.[40]

    Variant of a wild-type albumin-binding domain (ABD) of a Streptococcal Protein G

    [40] Hudson 1 at [78]; Baker 1 at [17].

21. As noted above, SEQ ID NO: 304 is ABD3 of protein G of Streptococcus strain G148 and consists of 46 amino acid residues that form a stable three-helix bundle.  This domain has broad albumin binding specificity.[41]  It is clear that the claims require that SEQ ID NO: 304 be modified as specified in options (A) and (B) of the claim – it must have the AAA or DSA mutations as discussed previously. 

    [41] Specification, [00615].

22. There is a question with regard to the extent to which a variant ABD may differ from SEQ ID NO: 304 – that is, whether the variant may differ only with respect to these mutations, or if it may have those mutations and additional mutations/deletions/additions, and if so, how much variation of the sequence is encompassed by the claims.  The claim recites that the “variant ABD has an amino acid sequence that differs from that of SEQ ID NO: 304 in comprising” either the AAA or DSA mutations.  The meaning of “comprising” has been variously construed by courts in Australia as either inclusive or exclusive of further components depending on the context of its use.[42] 

    [42] See, e.g., Fresenius Medical Care Australia Pty Limited v Gambro Pty Limited [2005] FCAFC 220; 67 IPR 230 at [64]; General Clutch Corporation v Sbriggs Pty Ltd [1997] FCA 499; 38 IPR 359 at 372-376; Asahi Kasei Kabushiki Kaisha v WR Grace and Co [1991] FCA 530; 22 IPR 491 at 514-515.

23. The parties agreed that an inclusive meaning is appropriate in the context of the present application – that is, the variant may have any number of mutations, provided that it has the AAA or DSA mutations (and satisfies the other requirements of the claim).  I am satisfied that this construction is reasonable.  There are a number of considerations which lead me to this view.  Claims 4 and 5 define the polypeptide of claim 1, wherein the sequence of the variant ABD differs from SEQ ID NO: 304 only by those variations specified in options (A) and (B) of claim 1, respectively.  Typically claims are construed to avoid redundancy and, while this is not an inviolable rule,[43] the presence of claims 4 and 5 to my mind supports construction of claim 1 as inclusive of additional mutations.  Additionally, “comprises” is used on other occasions in the description in a manner consistent with an inclusive meaning[44] (noting that it is also used in a sense which appears to be exclusive), and claim 8 (and claim 19) makes clear that the polypeptide of the preceding claims may include an additional ABD, such that the first use of “comprises” in claim 1 is non-exhaustive. 

    [43] David Kahn Inc. v Conway Steward & Co. Ltd. [1974] RPC 279 at 308; Nichia Corporation v Arrow Electronics Australia Pty Ltd [2019] FCAFC 2 at [41].

    [44] For example, [00147]: “one or more of the polypeptide chains of the diabody comprises an Fc domain”; [00153]: “a polypeptide chain of the invention comprises a hinge domain”.

24. Having determined that the claim should be construed as inclusive of additional mutations, there is nothing in the specification, or in the evidence, that would suggest that a skilled person would understand there to be a limit to the number or nature of additional mutations by which the variant ABD may differ from SEQ ID NO: 304 (beyond the requirement of deimmunization, serum binding ability and the result of extending serum half-life of the polypeptide to which it is conjugated). 

25. I note that this construction is consistent with Dr Hudson’s reading of the claim:

    “as I read Amended Claim 1, it is not limited to a polypeptide comprising a deimmunized ABD portion containing only the AAA or DSA mutations, but includes a polypeptide comprising a deimmunized ABD portion that contains the AAA or DSA mutations as well as additional residue mutations to the 46 amino acid ABD sequence”[45]

    and

    “[t]he ABD domain in Claim 1 is not even limited to a length of 46 amino acids and therefore could encompass mutations that shortened or lengthened the ABD domain.”[46]

    Said deimmunized albumin-binding protein portion extends the serum half-life of said polypeptide

    [45] Hudson 2 at [10].

    [46] Hudson 3 at [23].

26. It is apparent that the extension of the serum half-life of the polypeptide associated with covalent linkage to the albumin-binding protein portion is relative to the half-life when the polypeptide has no conjugated albumin-binding protein portion, and not relative to the half-life when the polypeptide is conjugated to a wild-type albumin-binding protein portion.

    Appended claims

27. The claims appended to claims 1 and 15 define narrower embodiments.  For example, as noted previously claims 4 and 5 define the polypeptide wherein the AAA and DSA mutations are the only differences from SEQ ID NO: 304.  Claim 6 defines the polypeptide wherein the variant ABD is the sequence of SEQ ID NO: 328 or 329.  Claims 6-14 and 16-20 define the polypeptide to which the variant ABD is linked, and define embodiments wherein the polypeptide is an antigen-binding molecule, including a diabody wherein each of the chains comprise the variant ABD.  Claim 21 is directed to a pharmaceutical composition comprising the polypeptide of any one of claims 1-14 and a pharmaceutically acceptable carrier.

    Clear enough and complete enough disclosure

28. Paragraph 40(2)(a) as amended by the Raising the Bar Act requires that the claimed invention be described in a manner which is clear enough and complete enough for the invention to be performed by a person skilled in the relevant art.

29. It is clear that this provision requires that the person skilled in the art must be able to perform the invention across the scope of the claim without undue burden or inventive skill.[47]  In Evolva SA[48] (Evolva) the Deputy Commissioner considered paragraph 40(2)(a), and having drawn guidance from European and UK decisions, adopted the following approach to assessing a clear enough and complete enough disclosure:

    What is the scope of the invention as claimed?

    What does the specification disclose to the skilled person?

    Does the specification provide an enabling disclosure of all the things that fall within the scope of the claims, and in particular:

    (a) Is it plausible that the invention can be worked across the full scope of the claim?

    (b) Can the invention be performed across the full scope of the claim without undue burden?

    [47] Novartis AG v Johnson & Johnson Medical Limited [2010] EWCA Civ 1039 at [74].

    [48] [2017] APO 57 at [45].

30. This approach has subsequently been adopted by delegates of the Commissioner,[49] and while the section 40 provisions as amended by the Raising the Bar Act were recently considered by the Federal Court for the first time in Encompass Corporation v InfoTrack Pty Ltd[50], the detail of the approach to considering clear enough and complete enough disclosure was not at issue.  I will adopt the Deputy Commissioner’s approach to considering this ground.

    [49] See, e.g. Rimfrost AS v Aker BioMarine Antarctic AS [2018] APO 34; Cytec Industries Inc. v Nalco Company [2018] APO 4; Grant Fisher v ToolGen Incorporated [2018] APO 65.

    [50] [2018] FCA 421; 130 IPR 387.

1. The Supreme Court in the UK recently provided further guidance with regard to the plausibility consideration in Warner-Lambert Company LLC v Generics (UK) Limited (t/a Mylan) and ors[51] (Warner-Lambert) considered the plausibility element of clear enough and complete enough disclosure with respect to a second medical use (Swiss style) claim.  In a majority decision, the conclusions of the lower courts with regard to the plausibility threshold were overturned, with Lord Sumption (with whom Lords Reed and Briggs agreed) stating:

   “They [the Court of Appeal] considered that the threshold was not only low, but that the test could be satisfied by a ‘prediction … based on the slimmest of evidence’ or one based on material that was ‘manifestly incomplete’.  Consistently with that approach, they considered (paras 40, 130) that the Board’s observations in SALK laid down no general principle.  I respectfully disagree.  The principle is that the specification must disclose some reason for supposing that the implied assertion of efficacy in the claim is true.  …  The test is relatively undemanding.  But it cannot be deprived of all meaning or reduced, as Floyd LJ’s statement does, to little more than a test of good faith.  Indeed, if the threshold were as low as he suggests, it would be unlikely to serve even the limited purpose that he assigns to it of barring speculative or armchair claims.”[52]

   [51] [2018] UKSC 56.

   [52] Warner-Lambert at [36].

2. He then commented that “[p]lausibility is not a term of art, and its context is inevitably influenced by the legal context” and identified a number of considerations relevant to the context of second medical use claims:

   “First, the proposition that a product is efficacious for the treatment of a given condition must be plausible.  Second, it is not made plausible by a bare assertion to that effect, and the disclosure of a mere possibility that it will work is no better than a bare assertion.  As Lord Hoffman observed in Conor Medsystems Inc v Angiotech Pharmaceuticals Inc [2008] RPC 28, para 28, ‘it is hard to see how the notion that something is worth trying or might have some effect can be described as an invention in respect of which anyone would be entitled to a monopoly’. But, third, the claimed therapeutic effect may well be rendered plausible by a specification showing that something was worth trying for a reason, ie not just because there was an abstract possibility that it would work but because reasonable scientific grounds were disclosed for expecting that it might well work. The disclosure of those grounds marks the difference between a speculation and a contribution to the art. … Fourth, although the disclosure need not definitively prove the assertion that the product works for the designated purpose, there must be something that would cause the skilled person to think that there was a reasonable prospect that the assertion would prove to be true. … Sixth, … the effect on the disease process need not necessarily be demonstrated by experimental data. It can be demonstrated by a priori reasoning. … Seventh, sufficiency is a characteristic of the disclosure, and these matters must appear from the patent.  The disclosure may be supplemented or explained by the common general knowledge of the skilled person.  But it is not enough that the patentee can provide that the product can reasonably be expected to work in the designated use, if the skilled person would not derive this from the teaching of the patent.”[53]

   [53] Warner-Lambert at [37].

3. To my mind, these considerations are all directed in a broad sense to the notion that plausibility is a technical consideration and assertions must be based on a reasonably credible technical or scientific basis derivable from the specification as understood by the skilled person.  While Sumption LJ’s comments were made in the context of second medical use claims, this notion is generally applicable. 

4. I note in particular that, consistent with the observation of the Court of Appeal in Regeneron Pharmaceuticals, Inc v Kymab Limited that “the skilled person may use his or her common general knowledge to supplement the information contained in the specification”[54] (applied in Grant Fisher v ToolGen Incorporated[55]), the requirement that sufficiency is a characteristic of the disclosure is tempered to the extent that the common general knowledge may be used to supplement or explain the disclosure.   

   [54] [2018] EWCA Civ 671 at [209].

   [55] [2018] APO 65.

5. The Explanatory Memorandum directs to the corresponding provisions of UK legislation and the European Patent Convention in relation to the application of section 40(2)(a) as amended by the Raising the Bar Act.[56]  Sumption LJ in Warner-Lambert has drawn from the body of UK and European cases in arriving at his conclusion, and accordingly I consider this judgment highly persuasive with regard to the application of the approach set out in Evolva.  Further, while it is clear that the Supreme Court has adopted a higher threshold with respect to plausibility than that adopted by the lower courts in the UK, I do not view the reasoning as inconsistent with the Deputy Commissioner’s statements in Evolva.[57]  It remains the case that, as set out in Evolva, the consideration is one of technical credibility or believability.  Plausibility may be a low threshold, but it is a threshold nonetheless, and is not satisfied by mere speculation or assertion.

   [56] The Explanatory Memorandum, Intellectual Property Laws Amendment (Raising the Bar) Bill 2011, item 8.

   [57] Evolva at [43].

6. The second limb of the enablement test is whether performing the invention across its full scope would constitute an undue burden.   The concept of an undue burden was discussed in Evolva.  Having considered UK and European authorities the Deputy Commissioner concluded:

   “My understanding of these authorities is that the emphasis in relation to undue burden has been on the nature of the work that is required by the skilled person in view of the guidance in the specification.  To this end, one approach has been to ask whether the skilled person would be required to undertake a ‘research programme’ in order to perform the invention.”[58]

   [58] Evolva at [33].

7. Put another way (as approved by Kitchin J in Eli Lilly & Co v Human Genome Sciences, Inc[59]):

   “… patent specifications need not set out every detail necessary for performance, but can leave the skilled man to use his skill to perform the invention. In so doing he must seek success. He should not be required to carry out any prolonged research, enquiry or experiment. He may need to carry out the ordinary methods of trial and error, which involve no inventive step and generally are necessary in applying the particular discovery to produce a practical result. In each case, it is a question of fact, depending on the nature of the invention, as to whether the steps needed to perform the invention are ordinary steps of trial and error which a skilled man would realise would be necessary and normal to produce a practical result.”

   [59] [2008] EWHC 1903; [2008] RPC 29 at [241].

8. The consideration of what constitutes an undue burden is necessarily dependent upon the nature of the technology, and factors relevant to the consideration include the level of predictability in the art and the level of guidance in the specification.[60]  

   [60] Evolva at [34]-[35].

9. A feature which is defined in broad terms will be allowable if it can be understood to be a principle of general application – which was described by Lord Hoffmann in Kirin-Amgen Inc v Hoechst Marion Roussel Ltd as:

   “an element of the claim which is stated in general terms.  Such a claim is sufficiently enabled if one can reasonably expect the invention to work with anything that falls within the general term”. [61]

   [61] [2005] RPC 169 at [112].

10. However, the requirements of section 40 cannot be avoided merely by the incorporation of a functional limitation restricting the scope of the claims to only those products or elements that have the relevant activity or function. It is still necessary that the invention can be performed across the scope of the claims without an undue burden.[62]  In American Home Products Corporation v Novartis Pharmaceuticals UK Ltd[63] (American Home Products) Aldous LJ made the following observations in relation to enablement in the context of functional limitations to a claim:

    [62] See, e.g., Novartis AG v Johnson & Johnson Medical Limited [2009] EWHC 1671 at [244].

    [63] [2001] RPC 159.

    “There is a difference between on the one hand a specification which requires the skilled person to use his skill and application to perform the invention and, on the other, a specification which requires the skilled person to go to the expense and labour of trying to ascertain whether some product has the required properties. When carrying out the former the skilled person is trying to perform the invention, whereas the latter requires him to go further and to carry out research to ascertain how the invention is to be performed. If the latter is required the specification would appear to be insufficient.

    …

    The judge held that the number of possible derivatives was vast and whether any particular molecule derived from rapamycin would work at all was impossible to predict with certainty.  Many derivatives would not exhibit immunosuppresant activity.  Those which involved small changes to the side chain would be the most likely to work.  Thus the skilled person could make up a list of possibles, with those believed to be the most likely at the top of the list.  Even so, finding appropriate derivatives, if they existed, would involve a systematic and iterative process.  Further, when a derivative which had appropriate activity had been identified, it would be impossible to be certain that it did not exhibit unpredictable defects.  To discover whether it did would require further tests which would take a long time.

    The very uncertainty and unpredictability found by the judge meant that the skilled person was being required to carry out research.  The duty upon the patentee is to provide a description which enables the skilled person to perform the invention, in this case across the breadth of the claim; not to supply a starting point for a research programme. …

    …the specification has to be sufficient to enable the invention to be performed.  There is a difference between research to find out which derivatives work and the application of the teaching in the specification with appropriate skill and tenacity.  In this case the specification tells the skilled man where to start but, upon the construction of claim 1 sought by the patentees, it leaves him to ascertain by research what will work.  Once it is appreciated that a claim which encompasses derivatives has to be sufficient across its breadth, the extent of the research task becomes apparent.  The number of derivatives is vast and the task of ascertaining which will satisfy the functional part of the claim will also be vast and correspondingly burdensome.”[64]

    [64] American Home Products at [40]-[44].

11. I have construed the present claims and set out the relevant disclosure of the specification above. The Opponent’s case in relation to section 40(2)(a) is directed to both aspects of the claimed invention: the polypeptide to which the ABD is conjugated, and the ABD itself. These involve quite separate considerations and I will address them separately. I note at the outset, however, that the Opponent stressed the observations made in Warner-Lambert regarding the requirement that plausibility be derivable from the disclosure of the specification.

    Polypeptide

12. As noted above, polypeptide is a broad term, encompassing diabodies and other antigen-binding proteins, but also proteins with other properties and functions.  In the Opponent’s submission the specification is directed to diabodies, and not polypeptides beyond that class, and the nature and form of a protein is fundamental to whether its serum half-life will be extended by a deimmunized ABD. 

13. I agree with the Opponent that the specification is largely directed to diabodies.  The related art described is directed to diabodies, as are the examples.  However, there are parts of the disclosure that refer to polypeptides more generally, and the claims relate to polypeptides.  The question, to my mind, is whether this broadly put element of the claim is a principle of general application such that it is plausible, having regard to the specification and the common general knowledge, that the invention insofar as it relates to this element could be worked across its full scope.  That is, is it plausible that the variant ABD defined in the claim can be conjugated to any polypeptide and result in extended serum half-life?

14. In Dr Baker’s opinion the answer to that question is yes: 

    “In my opinion I believe that the data presented in which deimmunised ABD’s fused to hCD16 x hCD32b DARTS could not only be extended to DART’s with other specificities, but also to other unrelated molecules, particularly those that display very rapid clearance (i.e., have a very short half life).”[65]

    [65] Baker 1 at [64].

15. Dr Baker does not elaborate on the reasons for this opinion, but, as submitted by the Applicant, statements in contemporaneous documents support a view that use of albumin binding proteins to extend serum half-life was known to be applicable to a range of polypeptides:

    ·In Hopp, J. et al. ‘The effects of affinity and valency of an albumin-binding domain (ABD) on the half-life of a single-chain diabody-ABD fusion protein’, Protein Engineering, Design & Selection 2010, vol. 23, pages 827-824 (Hopp; Exhibit PH-6) the authors “analyzed the effect of affinity and valency of the ABD on the pharmacokinetic properties of a bispecific single-chain diabody”[66] and concluded that “altering the affinity or valency of albumin binding has only minor effects on the half-life of scDb-ABD fusion proteins in mice”.[67]  The article concludes with the observation that “even molecules with medium affinity for albumin are appropriate to prolong half-life of therapeutic proteins”.[68]

    ·In Andersen, J.T. et al. ‘Extending half-life by indirect targeting of the neonatal Fc receptor (FcRn) using a minimal albumin binding domain’, Journal of Biological Chemistry 2011, vol. 286, pages 5234-5241 (Andersen; Exhibit PH-7) the authors explain that “a minimal three-helical albumin-binding molecule with SpG has been widely used as a fusion partner for Fab fragments, single chain diabodies, and Affibody molecules”[69] and that “use of albumin fusion or albumin binding for the purpose of extending the serum half-life of short lived molecules has been extensively described”.[70]  The article explains that the long half-life of albumin can be attributed to the receptor-mediated recycling pathway involving the neonatal Fc receptor (FcRn).[71]

    Fusion of an ABD to either the amino or carboxy terminal of an Affibody is exemplified, and “the ability to bind albumin was preserved for both fusion variants”, although fusion at the carboxy terminal was preferred.[72]  The article goes on to state that “[b]y not interfering with the pH-dependent binding of HSA to shFcRn, ABD meets the fundamental criterion for a carrier molecule to be used for half-life extension based on the FcRn recycling mechanism.”[73]  The conclusion of the article is “that the ABD fusion technology is a widely applicable strategy for extending the half-life and improving the biodistribution of therapeutics via indirect targeting of FcRn.”[74]

    ·US 2010/0273979 A1 (Abrahmsén et al.) 28 October 2010 (US’979; Exhibit PH-12) is directed to engineered polypeptides with an affinity for albumin and exemplifies conjugation of and albumin binding domain to an affibody leading to extended half-life and decreased immunogenicity.[75]  In the background section, this patent application states that:

    “Several strategies have been reported to either covalently couple proteins directly to serum albumins or to a peptide or protein that will allow in vivo association to serum albumins. Examples of the latter approach have been described … The idea is to fuse the bacterially derived, albumin binding peptide/protein to a therapeutically interesting peptide/protein, which has been shown to have rapid clearance in blood.  The thus generated fusion protein binds to serum albumin in vivo, and benefits from its longer half-life, which increases the net half-life of the fused therapeutically interesting peptide/protein.”[76]

    [66] Abstract.

    [67] Page 834.

    [68] Page 834.

    [69] Page 5234.

    [70] Page 5239.

    [71] Pages 5234-5235.

    [72] Page 5239.

    [73] Page 5240.

    [74] Page 5241.

    [75] Example 4.

    [76] [0003].

16. While the focus of these documents is antigen-binding molecules, I understand them to refer to ABD fusion as a strategy for half-life extension of proteins generally.  In line with this, the specification states that “[a]lbumin possesses several small molecule binding sites that permit it to non-covalently bind to other proteins and thereby extend their serum half-lives”[77] and Dr Hudson indicated that:

    “By at least 2000, strategies for protein half-life extension were well understood and widely reported in the literature.  Researchers were at that time, routinely using these strategies to engineer fusion proteins that bypassed renal clearance extending the longevity of many therapeutic proteins.”[78]

    [77] [00616].

    [78] Hudson 1 at [52].

17. He also indicated that the strategy of fusion of an albumin-binding moiety to a therapeutic protein, such that it would bind to serum albumin following administration, was well-established by May 2011 with the most common relying on binding albumin with an ABD from Streptococcal Protein G (SPG).[79]  He stated that “[f]usion of any one of SPG ABD 1-3 with a protein (such as recombinant antibodies or bispecific single-chain diabody) had been shown to prolong the plasma half-life of these molecules.”[80]

    [79] Hudson 1 at [53]-[55].

    [80] Hudson 1 at [56].

18. However, Dr Hudson went on to state, with reference to Hopp and Anderson, that:

    “They taught what I understood, and what I expect other researchers working in the field of antibody engineering and protein science would have commonly and generally understood, about the strategy of using ABD’s for the extension of protein half-life at May 2011.  That is, the strategy was understood but unpredictable and uncertain.”[81]

    [81] Hudson 1 at [60].

19. It seems to me that in context, this statement can most reasonably be understood to relate to a lack of predictability of the outcome of modifications to the ABD on serum half-life, not to the use of ABDs as a half-life extension strategy at a general level, which is consistent with his subsequent observation that:

    “...nothing was predictable or certain in this field of using amino acid substitutions to engineer new or improved properties into protein-based biopharmaceutical molecules.

    While the concept that the half-life of a protein could be improved using ABD’s was well understood by my colleagues and I at May 2011, there was little that could be predicted about the impact that particular changes to a given ABD had on protein half-life.”[82]  

    This is also consistent with Dr Hudson’s apparent view that the half-life extension required by the claims is relative to that attained with the wildtype ABD rather than no ABD.[83] 

    [82] Hudson 1 at [61]-[62].

    [83] See, e.g. Hudson 1 at [133].

20. However, Dr Hudson subsequently stated that:

    “…I would not expect that any polypeptide or even any DART which comprises either the AAA or DSA mutations in the ABD, would exhibit deimmunization and increased serum half-life.  …other DART polypeptides comprise different V domains and will possess unique design features and properties characteristic only for that DART, and therefore, the results for one particular DART cannot necessarily be extrapolated to other DARTs, and certainly not all polypeptides in general. This is a highly unpredictable field. Accordingly, I, and I expect others skilled in the art, cannot extrapolate from the results of one sequence and conclude that all sequences, or at least all other sequence classes, will necessarily display the same effect or outcome. Such an extrapolation would, in my opinion, be scientifically unsound and grossly overreaching having regard to the state of this particularly unpredictable field of science.

    …The addition of an ABD domain to a polypeptide changes its sequence length, and therefore has the capacity to have a significant impact on that polypeptide’s conformation, and therefore greatly influence the activity of that polypeptide in vivo, including potentially inhibiting its therapeutic effect.  This is particularly likely to be a problem with smaller polypeptides.  Accordingly, I, and I expect others skilled in the art, would have to undertake a very significant amount of experimentation … to determine which polypeptides (if any) would exhibit a deimmunized effect and increased serum half-life when comprising an ABD with AAA or DSA mutations.”[84]

    [84] Hudson 2 at [16]-[17].

1. Accordingly, I understand his view to be that it is not plausible that the claimed invention can be worked with any polypeptide, although as noted above, I consider that it is only the ABD that is required to be deimmunized.  Furthermore, to the extent that Dr Hudson’s concerns are directed to the retention of the biological activity of the polypeptide when fused to the ABD, I understand the requirement of the claim to be simply that the serum half-life of the polypeptide is extended.

2. The Opponent submitted that no information about how to conjugate the ABD to the DART or any other protein is provided, referring to Dr Hudson’s evidence:

   “There are no examples of how to fuse the ABD to the protein.  Rather the reader is left to experiment to determine this.  This is not necessarily an easy set of experiments because multiple ABD’s must be made and one need to determine whether the ABD should be fused merely to one end or to both ends of the protein.  The Patent Specification is silent in its teaching about this.  Furthermore, you cannot predict that which is going to work.  Rather, I and I believe other researchers working in the field would need to design a whole series of experiments and carry those experiments out before we could hope to determine what works and what does not work.”[85]

   [85] Hudson 1 at [139].

3. In contrast, Dr Baker did not consider that precise details of the fusion method were necessary:

   “The Opposed Application teaches fusion through the E/K coils and a GS linker, which is and was as of the Priority Date a fairly standard approach.  It is possible that optimised fusion methods would have to be performed for polypeptides that are not DART’s, although this level of additional experimentation is not above the typical level required to apply a new technology to any new molecule.  Additionally, I consider that the Opposed Application extensively teaches fusion molecules being made using recombinant DNA technology (see, e.g., SEQ ID Nos: 307 and 308).”[86]

   [86] Baker 1 at [65].

4. I agree with the Applicant that there are several examples in the specification that disclose fusion of ABDs to DART molecules.[87]  On balance, I am satisfied that the nature of the work required to work the invention in this regard, while likely to be non-trivial, is in the nature of routine experiments with a view to making the invention work, rather than a research program.

   [87] Examples 6.18, 6.20.

5. The statements in contemporaneous documents, albeit not forming part of the common general knowledge themselves, indicate that ABD fusion is a general serum half-life extension strategy, which is consistent with Dr Baker’s evidence and Dr Hudson’s more general statements explaining the basis on which it is expected that ABDs will generally extend serum half-life (by binding to serum albumin).  Accordingly, while the specification itself does not set out a basis for expecting the claimed ABDs to display the desired result with any polypeptide, I am satisfied that a skilled person equipped with the common general knowledge, would find it plausible that fusion of a deimmunized ABD as defined in the claims to any polypeptide would extend the serum half-life of that polypeptide. Put another way, in the context of the present claims, I consider that the “polypeptide” to which the ABD is conjugated is a principle of general application. Indeed, I note that the deimmunization and binding affinity studies in the specification in relation to the ABD are undertaken prior to its conjugation to the diabody, and there is no indication that these steps are undertaken with a view to the polypeptide to which the ABD will ultimately be conjugated (nor have the declarants indicated that any different approach would be expected to be adopted). That is, this aspect of the specification is directed to the ABD and its properties, not its fusion partner. I do not consider that the reference to this broad element in the claims offends s 40(2)(a).

   ABD

6. I have decided above that the claims must possess the specific mutations defined, but may also include any number of other variations – provided that albumin binding and deimmunization is retained.  Notably, to the extent that the variant ABDs extend beyond the specific mutants described, they are partly defined by the parameters of albumin binding and deimmunization, not the technical features by which those parameters will be achieved.  The Opponent’s submissions in relation to the variant ABD of the claims can be separated into two parts.  First, that the specification does not adequately, or plausibly, demonstrate that the specific ABDs – the AAA and DSA variants – disclosed in the specification are deimmunized, and second, that the extension of the claims to encompass any number of mutations additional to those defined in the claims is not enabled. 

   AAA and DSA ABDs

7. I have indicated above the extent of the examples in the specification.  The data presented in the specification as indicative of deimmunization are in silico predictions (at Tables 20 and 26) and EpiScreen assays conducted with 21mer fragments of SEQ ID NO: 304.

8. The Opponent submitted that the in vitro studies (EpiScreen T cell stimulation and ELISpot assays) described in the specification are limited to 21mer fragments and therefore do not establish the immunogenicity of the 46 amino acid ABD, or any other ABD sequence length.  In the Opponent’s submission, because the specification does not indicate whether any MHC class II binding peptides are present outside residues 60-79 of SEQ ID NO: 304, the skilled person does not know whether the full sequence has been screened for such peptides.  I have already construed the specification and concluded, consistent with the Applicant’s submission, that a fair reading of it is that the whole sequence was subject to in silico analysis and all the P1 anchor positions identified in that analysis have been recited in the specification. 

9. Furthermore, the Opponent submitted that the results in the specification do not demonstrate that even the 21mers have been deimmunized.  Tables 24 and 25, presenting EpiScreen assay results, are set out earlier in this decision.  The declarants disagreed with regard to how deimmunization should be assessed when considering the results of these assays.  As discussed previously, the specification reports both the mean magnitude (SI) of T cell proliferation response and the frequency of T cell response (i.e. the % of responders in a given cohort).  The Opponent submitted that where there is inconsistency between these two measures no conclusions can be drawn regarding deimmunization.[88] I understand Dr Hudson to consider the mean SI values to be of importance given that he formed a view that the DSA mutant is potentially more immunogenic than the wild-type peptide based on the mean magnitude SI scores,[89] and considered the differences in the percentage response data to be minimal, and to indicate only that neither the wildtype nor the mutants peptides are very immunogenic.[90]

   [88] Opponent’s written submissions, Appendix 6 at [24].

   [89] Hudson 1 at [117]; Hudson 2 at [31].

   [90] Hudson 1 at [118], [121]; Hudson 2 at [26].

10. In contrast, Dr Baker stated:

    “The most important metric for assessing immunogenicity using the EpiScreen^TM assay is the frequency of T cell response in a given study cohort (i.e., % response) and I consider that this would have been known by the skilled person as at the Priority Date.  No one skilled in the art as of May 2011 would have selected leads from the EpiScreen^TM assay based on mean magnitude (SI).  To this end, it would only take a single donor to respond with a high magnitude response to produce a high immunogenicity prediction, whereas multiple donors may score with lower magnitude and give a low immunogenicity prediction.  There is and was as of the Priority Date no correlation between the magnitude of T cell response in small numbers of donors and immunogenicity in the clinic.”[91]

    [91] Baker 1 at [53].

11. Having regard to this explanation, and the indication in the specification that circumstances where the mean magnitude (SI) is similar but frequency of T cell response is decreased (in relation to both T cell proliferation and IL-2 secretion) are indicative of deimmunization,[92] and also noting that the mean SI value reflects only positive and borderline responses, I prefer Dr Baker’s approach. I also note that as Dr Baker is an inventor of the EpiScreen assay, and has significant experience with the technique,[93] I give his views on interpretation of the results of the assay greater weight than Dr Hudson’s with regard to what a skilled person would understand from the specification.  I do not understand Dr Baker to be totally disregarding the magnitude of the mean SI, but indicating that where there is a low response frequency that is of greater significance given the potential impact of low numbers of high responders on the mean SI magnitude.  This is consistent with the teaching of the specification itself, in which the overall conclusions are based on the percentage response.[94]  In Dr Baker’s view “the difference presented in the Opposed Application between the wild type and the two deimmunized triple mutant ABD’s is reasonably large for a cohort of 50 donors.”[95]  I consider that the data presented supports the conclusion presented in the specification that the immunogenic potential of the AAA and DSA mutants is lower than the wildtype ABD.

    [92] [00670], [00672].

    [93] Baker 1 at [18], [54].

    [94] [00673].

    [95] Baker 1 at [54]. I also note that Dr Baker indicated that “variations in the region of 10% difference between in vitro T cell responses” would be considered statistically significant (Baker 1 at [18]).

12. Finally, the Opponent noted that no embodiment has been tested and shown to exhibit both deimmunization and provision of an extended serum half-life – only 21mers were tested in the EpiScreen assays and only full length ABDs were conjugated to diabodies to demonstrate increased serum half-life.  In this regard, Dr Hudson indicated that “[d]ue to the unpredictability of identify and mutating potential T cell epitopes, I cannot extrapolate results from an isolated 21 amino acid to a 46 amino acid parent polypeptide.”[96]  In contrast, Dr Baker stated:

    “It would have indeed been preferable to include the entire 46 amino acid ABD polypeptides with or without the deimmunizing mutations.  However, the testing of the 21 amino acid sequence containing the mutations identified from the previous in silico analysis show that the potential T cell epitopes identified, were indeed attenuated by certain combinations of the deimmunizing mutations.”[97]

    and concluded:

    “I believe that the Opposed Application sufficiently provides for a deimmunized ABD based on at least SEQ ID NO: 304, where the lead triple mutant molecules presented therein appear to have reduced immunogenicity and enhanced serum half life as required by Claim 1.”[98]

    [96] Hudson 1 at [115].

    [97] Baker 1 at [52].

    [98] Baker 1 at [61].

13. As indicated previously, I am satisfied that the teaching of the specification is that the only MHC class II binding peptides identified from in silico analysis of SEQ ID NO: 304 are located in this region.  While I accept that there are different in silico techniques, and these techniques are not fully reliable,[99] I note that plausibility is a low threshold not requiring definitive proof.  I am satisfied that demonstrating diminished immunogenicity of the region identified in the specification by in silico testing as containing MHC class II binding peptides is sufficient to plausibly demonstrate deimmunization of SEQ ID NO: 304. 

    [99] See, for example, Hudson 1 at [89]; Exhibit PH-8.

14. Accordingly, I am satisfied that the disclosure of the specification is sufficient with respect to the specific variant ABDs exemplified (SEQ ID NOs: 328 and 329) – I consider that the information in the specification makes it plausible that these sequences are deimmunized and result in extended half-life of a protein to which they are conjugated, and there was no suggestion that there is an undue burden in preparing these specific ABDs.  The question remains as to whether the claims are enabled across their full scope insofar as they encompass further variations to the ABD.

    Variant ABDs

15. I have construed the claims previously and concluded that there is no limit on the number and nature of additional mutations by which the variant ABD may differ from SEQ ID NO: 304, subject to the requirements in the claim regarding deimmunization and serum half-life extension.

16. In summary, the Opponent submitted that the claims are not enabled across their full scope with respect to ABD variants because 1) the specification provides no guidance as to what modifications can be made to the length of SEQ ID NO: 304 without disrupting the helical sequence and HSA binding capacity;[100] 2) noting the data in the specification, the effect of combinations of substitutions on HSA binding, and immunogenicity, is unpredictable; and 3) the invention is not directed to a common set of mutations that unites a class, but to “a disparate and diverse set of modified ABD sequences.”[101]  In the Opponent’s submission the specification “presents a gigantic research project”.[102]

    [100] Opponent’s written submissions at [43].

    [101] Opponent’s written submissions at [102].

    [102] Opponent’s written submissions at [114].

17. In response, the Applicant submitted that the specification provides guidance as to particular mutations that adversely impact HSA binding, for example indicating that 61A mutants should be avoided in combinations[103] and concluded that the skilled person “when armed with the CGK, clearly has the knowledge as to which amino acid changes are likely to retain the function of the DSA and AAA mutant ABDs”.[104]  While I accept that the level of skill of the relevant person skilled in the art is high, and there are examples of combinations of mutations to the ABD and their effect on HSA binding in the specification,[105] there is a lack of evidence supporting the Applicant’s contentions in this regard, and I note that the claims encompass mutations outside of the range of residues contemplated in the specification.

    [103] Applicant’s written submissions at [114]. Dr Baker also noted that Y61A “is not worth pursuing” – Baker 1 at [44].

    [104] Applicant’s written submissions at [119].

    [105] Figures 57-62.

18. The Applicant further submitted that conservative substitution of amino acids in proteins was part of the common general knowledge (with the specification referring to the manner in which protein derivatives having conservative substitutions at predicted non-essential amino acid residues can be generated[106]) and that the critical residues of ABDs involved in binding albumin were well known in the art.  In this regard the Applicant pointed to documents in evidence demonstrating various mutations that can be made to the ABD of SEQ ID NO: 304, and to the existence of a mutational study identifying critical residues involved in binding albumin referred to in Hopp.[107] 

    [106] [00201], referring to antibodies or fragments thereof.

    [107] Page 827.

19. While there are documents reporting the effect of various mutations to ABDs in evidence, and discussed previously, it is not apparent that those documents themselves formed part of the common general knowledge, and there is no information of this sort in the specification.  Further, it is not apparent that a skilled person would appreciate what mutations (or combinations of mutations) could be combined with the AAA or DSA mutations.  To the contrary, the evidence that the result of combinations of mutations on HSA binding is unpredictable is uncontradicted:[108] 

    “Amended Claim 1 contemplates single mutations in addition to the triple mutations DSA or AAA or multiple mutations of multiple combinations in addition to the triple mutations DSA or AAA, generating a potentially vast number of possible mutations that could be made to the ABD, and still fall within claim 1.  There is no description of other mutations that could be made in addition to AAA and DSA as recited in the Amended Claim 1 which demonstrated deimmunization and increased serum half-life, nor is there information provided in the description that could guide me on what further mutations could be made.  The Patent Specification is silent in its teaching and I am left to experiment to determine this.  Furthermore, because of this lack of information in the Patent Specification, I would not have the confidence to predict which mutations may work.”[109]

    [108] Baker 1 at [43]; Hudson 2 at [35].

    [109] Hudson 2 at [35].

20. I find the Applicant’s submission that Dr Hudson’s comments in this regard should be afforded little or no weight unpersuasive.  Dr Baker has not contradicted this evidence directed to further mutations of the AAA and DSA ABDs, and it seems consistent with the non-additive effect of combinations of mutations on HSA binding described in the specification.

21. Furthermore, in addition to retention of HSA binding ability, the ABDs of the claims are required to be deimmunized.  Where further variations (substitutions, additions or deletions) are made to the AAA and DSA ABD variants, there is the potential for the introduction of immunogenic regions, both into the region of the protein investigated in the specification, and to the remainder of the ABD variant.[110] 

    [110] Hudson 2 at [13].

22. Both parties have provided evidence as to the nature and cost of the work that would be required to identify/produce further deimmunized and functional ABDs within the scope of the claims.  While there is disagreement as to the specific times and costs associated with the work, the declarants seem to be in broad agreement that the general steps required are:[111]

    ·   in silico screening of the sequence for T cell epitopes to determine which residues could be mutated to produce a deimmunizing effect

    ·   expression and purify ABDs containing potentially deimmunizing mutations

    ·   confirmation of retention of HSA binding affinity in the mutated ABD

    ·   immunogenicity assays on ABDs with retained HSA binding affinity

    ·   testing of half-life of conjugated proteins in vivo

    [111] Hudson 2 at [10]; Baker 2 at [14].

23. While the costs and time associated with experiments required to carry out an invention can provide a context to whether those experiments are in the nature of routine trial and error or not, in this case I do not need to attempt to resolve the differences in these values presented by Dr Baker and Dr Hudson.  It is agreed that the basic approach to identifying further embodiments within the scope of the claims is the known approach for deimmunizing a protein sequence at the priority date.[112]  I do not understand either declarant to suggest that the actual carrying out of this work, in itself, presents particular difficulties.  Nor do I consider that the evidence establishes that it is implausible that ABD variants satisfying the functional requirements of the claim could be identified and used to extend polypeptide half-life.  To my mind, the question is whether the requirement to undertake the work required is an undue burden (noting that a skilled person would not be expected to persist with a particular variant after failure at one of the steps).  Are these the “ordinary steps of trial and error which a skilled man would realise would be necessary and normal to produce a practical result”?  Or is the skilled addressee presented with a research project? 

    [112] Baker 1 at [31]; Hudson 3 at [36].

24. In Dr Hudson’s view it is the latter:

    “I accept that although a pharmaceutical company might be enticed to conduct this research -- with high risk there is high reward -- I disagree that this research is anything near ‘routine’ research or research that is trial and error.  The high risk and lack of certainty associated with the research makes the entire project highly speculative.

    While a person skilled in the art would be familiar with particular techniques and assays that must be performed to investigate possible mutations and polypeptides that fall within the scope of the claims, familiarity with the techniques does not, in my experience, make the experiments any easier, cheaper, or produce a predictable outcome. The research that is required is an expensive and time-consuming process that is far from routine, particularly so because the skilled person would not have any ability to predict what may or may not work.”[113]

    [113] Hudson 3 at [35]-[36].

1. In contrast, Dr Baker stated that:

   “such experimentation, as described in the specification … would, in my experience, form part of the normal analysis routinely performed by pharmaceutical companies during their development of a deimmunized binding domain, and be reasonable for the skilled person to carry out, despite their cost, as part of normal trial and error.”[114]

   [114] Baker 2 at [9].

2. However, Dr Baker also considered that following the process set out in the specification yields a novel and inventive result, which, while given in a different context, in my view suggests a level of uncertainty and unpredictability in following that process:

   “…I highlight Dr Hudson’s comment that: ‘[t]here would be a very real prospect of not being able to select a polypeptide which would be deimmunized and still retain HSA binding capacity to increase the serum half-life of the polypeptide’.  Given that the Opposed Specification has largely followed the methodology set out by Dr Hudson to perform the present invention and the Applicant has subsequently arrived at the two triple mutant ABD molecules recited in Claim 1 as amended that are deimmunized whilst also retain the ability to bind serum albumin, this in my opinion makes the outcome of the presently claimed invention novel and inventive.”[115]

   [115] Baker 1 at [56].

3. This is also consistent with his statement that:

   “…the addition of the Y61A mutation to the other two mutations 66S and 70S abolishes binding of the mutant ABD to HSA, whereas the Y61A mutation alone has no effect on such binding, as shown in Table 20.  In my experience such results upon the combination of mutations is not unusual and highlights the need to take a systematic approach to deimmunizing a protein by first mutating single anchor positions then testing combinations of mutations in a stepwise fashion”.[116]

   and Dr Hudson’s indication that:

   “This particular field of science is not that predictable.  I and I believe other researchers working in the field would need to design a whole series of experiments and carry those experiments out before we could hope to determine what works and what does not work.  These experiments would involve testing many hundreds of permutations of mutations to understand what works, and therefore would be very expensive and time consuming.”[117]

   [116] Baker 1 at [43].

   [117] Hudson 2 at [35].

4. Accordingly, while the Applicant submitted that “additional mutations do not amount to an undue burden in the context of a technology where it is routine to produce, express and screen large numbers of polypeptides”,[118] the evidence supports a conclusion that at each step of the “routine” approach to protein deimmunization there is the potential for failure,[119] and neither the declarants nor the specification indicate any manner of predicting what is and is not likely to work. As noted above, the Applicant’s submission that the embodiments in the specification guide the skilled person as to what additional mutations in the region of the identified T cell epitopes might be made is not supported by the evidence, and this guidance is not obvious on the face of the specification. Instead, it is clear from the specification that combining mutations is unpredictable, at least with respect to HSA binding, and the Applicant has not identified anything in the specification which guides a person as to how to proceed when confronted with failures.

   [118] Applicant’s written submissions at [146].

   [119] For example, Hudson 3 at [30], Baker at [47].

5. Having regard to the disclosure of the specification and the evidence in the present opposition,[120] on balance I consider that the nature of the experimentation required to carry out the invention across its full scope is more in the nature of “research to find out which derivatives work”, akin to the scenario considered by Aldous LJ in American Home Products, than application of the teaching of the specification with a view to working the invention.  The number of possible variants, taking into consideration mutations leading to sequences of varying length and composition, is large, and the evidence indicates that the work required to identify variants with the desired properties, particularly HSA binding ability, while systematic and iterative, is unpredictable, with limited guidance in the specification.  Accordingly, in my view the work required to work the invention across the scope of the claims constitutes an undue burden, being in the nature of a research project, and it follows that the specification does not provide a clear and complete enough disclosure of the claimed invention insofar as it relates to variants of ABDs with unlimited further substitutions.

   [120] In particular, Hudson 2 at [35]-[36], Baker 1 at [43], Hudson 3 at [35]-[36].

6. I will note that the present situation is to my mind quite different from the circumstances contemplated by the Deputy Commissioner in Evolva, where the applicant pointed, persuasively, to the requirement for 90% sequence homology, the premise of the invention being the use of a non-native UGT for mogroside synthesis, and the knowledge of the skilled addressee as to which amino acid changes were likely to result in retention of function.[121]  In such circumstances, the nature of the work was not considered to constitute a research program.

   [121] Evolva at [67]-[68].

   Support

7. Subsection 40(3) as amended by the Raising the Bar Act requires that the claims must be supported by matter disclosed in the specification.  The requirement of support can be summarised as requiring that the scope of the claims “should correspond to the technical contribution to the art”.[122]

   [122] Fuel Oils/EXXON (T409/91) [1994] OJ EPO 653 at 659.

8. The requirement of support has been more fully explained as follows:

   “in other words it is the definition of the invention in the claims that needs support. In the Board’s judgement, this requirement reflects the general legal principle that the extent of the patent monopoly, as defined by the claims, should correspond to the technical contribution to the art in order for it to be supported, or justified. This means that the definitions in the claims should essentially correspond to the scope of the invention as disclosed in the description. In other words, as was stated in Decision T 26/81, the claims should not extend to subject-matter which, after reading the description, would still not be at the disposal of the person skilled in the art. Consequently, a technical feature which is described and highlighted in the description as being an essential feature of the invention, must also be a part of the independent claim or claims defining this invention”.[123] (emphasis in the original, citations omitted)

   [123] Fuel Oils/EXXON (T409/91) [1994] OJ EPO 653 at 659-660.

9. To determine whether the requirements of support are satisfied the following steps were set out in CSR Building Products Limited v United States Gypsum Company[124]:

   i. construe the claims to determine the scope of the invention as claimed,

   ii. construe the description to determine the technical contribution to the art, and
   

   [124] [2015] APO 72 at [115].

   iii. decide whether the claims are supported by the technical contribution to the art.

10. There is substantial overlap in the submissions made by both parties in relation to section 40(3) and section 40(2)(a) which it is not necessary to repeat. In relation to this ground the Applicant submitted:

    “With respect to the concept of what constitutes an Applicant’s technical contribution to the art, the Delegate in CSR cited a UK authority to highlight that the invention’s technical contribution to the art is related to its inventive concept or non-obvious novelty.

    To this end, what is new and nonobvious in respect of the present invention is an ABD that includes either of two specific variations (i.e., AAA and DSA) that not only deimmunize the ABD, but also retain its ability to bind HSA and extend the serum half-life of polypeptides bound thereto.”[125]

    [125] Applicant’s written submissions at [134]-[135].

11. It is true that for a simple product claim, the technical contribution may be equated with non-obvious novelty.[126]  The claims in question, however, are not simple product claims.  They are claims to a large number of functionally defined products.  Support requires “the description to be the base which can fairly entitle the patentee to a monopoly of the width claimed.”[127]  This means that, as quoted above, “the claims should not extend to subject-matter which, after reading the description, would still not be at the disposal of the person skilled in the art.”[128]

    [126] Generics (UK) Ltd v H Lundbeck A/S [2009] UKHL 12; [2009] RPC 12 at [95].

    [127] Schering Biotech Corp.’s Application [1993] RPC 249 at 252.

    [128] Fuel Oils/EXXON (T409/91) [1994] OJ EPO 653 at 659.

12. It follows from my reasons above in relation to section 40(2)(a) that I consider the technical contribution to the art includes the deimmunized ABDs disclosed and their use as fusion partners to extend the half-life of polypeptides to which they are conjugated. Equally, insofar as the claims encompass any number of unspecified mutations to the ABD I consider that such subject matter is not enabled for the reasons given in relation to section 40(2)(a), and it therefore cannot be viewed as forming part of the technical contribution to the art – it is not at the disposal of the skilled addressee. There is a question as to whether the technical contribution to the art extends beyond the specific ABDs disclosed to include some variants. On balance, in the absence of a clear indication in the specification of a broader contribution and given the lack of evidence that a skilled person would understand the contribution more broadly, I conclude that it does not.

13. The claims extend beyond the technical contribution to the art and lack support.

    Utility

14. Paragraph 18(1)(c) of the Act requires that the claimed invention be useful.  Utility was considered by the Full Court of the Federal Court in H Lundbeck A/S v Alphapharm Pty Ltd with Emmett J stating:

    “A claim is bad if it covers means that will not produce the desired result, even if a skilled person would know which means to avoid.  That is to say, everything that is within the scope of a claim must be useful, otherwise the claim will fail for inutility.”[129]

    [129] [2009] FCAFC 70; 81 IPR 228 at [81].

15. Section 7A of the Act provides that for an invention to be considered useful the specification must disclose a specific, substantial and credible use for the invention.  This requirement is additional to the requirement that the claimed invention must achieve the promised benefit.[130] Section 7A was introduced in the Raising the Bar Act and is elaborated on in the Explanatory Memorandum[131] (with references omitted):

    [130] See, e.g., Artcraft Urban Group Pty Ltd v Streetworx Pty Ltd [2016] FCAFC 29 at [120]-[121].

    [131] Explanatory Memorandum, Intellectual Property Laws Amendment (Raising the Bar) Bill 2011 (Cth), pages 44-45.

    “The intent is that specific, substantial and credible be given the same meaning as is currently given by the US courts and the United States Patent and Trade Mark Office (USPTO).

    Currently, the US courts interpret the terms as follows:

    ·‘specific’ means a use specific to the subject matter claimed and can ‘provide a well-defined and particular benefit to the public.’  

    ·‘substantial’ means the claimed invention does not require further research to identify or reasonably confirm a ‘real world use’. ‘An application must show that an invention is useful to the public as disclosed in its current form, not that it prove useful at some future date after further research’.

    ·an asserted use will be ‘credible’ ‘unless there is evidence that the invention is inoperative (i.e. does not operate to produce the results claimed by the patent application) or there is reason to doubt the objective truth of the statements in the specification.’

    The relevant principles are set out in more detail by the USPTO.

    The amendment will strengthen the test for usefulness and prevent the speculative claiming of inventions that would require further experimental effort before they could be put into practice.”

16. The Opponent submitted that the evidence in the specification does not support a conclusion that the ABDs have been deimmunized, similar to the submissions made in relation to disclosure.  The Applicant responded that there was no evidence to support the Opponent’s submission in this regard.

17. I have decided above that the evidence in the specification is sufficient to support a conclusion that the specific AAA and DSA ABDs (i.e. SEQ ID NOs: 328 and 329) are deimmunized and extend the serum half-life of a polypeptide to which they are conjugated.  The question then is whether the claims extend to subject matter which does not achieve the promised benefit or have a specific, substantial or credible use. 

18. I have observed previously that the claims are drafted such that the variant ABD of the claim is deimmunized and does extend the serum half-life of a polypeptide to which it is conjugated.  While drafting claims to exclude non-functional embodiments does not avoid a lack of enablement, this conclusion does not follow with respect to the ground of lack of utility.  It seems to me that given the manner in which the claims are drafted there is nothing included within the scope of the claims that will not achieve the promised benefit, and given my conclusions regarding the specific exemplified variants, I am satisfied that the specification discloses a specific, substantial and credible use for the invention.  That the scope of the claims is such that it encompasses embodiments that are not enabled does not, in my view, provide a basis for a finding of lack of utility. 

19. The Opponent has not established that the claims lack utility.

    Conclusion

20. The opposition is successful. The specification does not comply with section 40(2)(a) or section 40(3).

21. This finding may be overcome by amendment and I will provide the Applicant with a period of time in which to file suitable amendments.

    Costs

22. It is usual in matters before the Commissioner that costs follow the event.  In its written submissions the Applicant highlighted the lengthy nature of the Opponent’s summary of submissions and submitted that the Opponent ought not to be awarded costs with regard to those submissions irrespective of the outcome of the opposition.  While I agree that the Opponent’s summary of submissions was unusually lengthy, I do not agree that this is behaviour that warrants a variation in the award of costs.  The Opponent has been successful in its opposition and I will award costs according to Schedule 8 against the Applicant.
    
    

    Dr S. J. Smith
    Delegate of the Commissioner of Patents

    Annex A: The claims

    1. A polypeptide that comprises a portion of a deimmunized albumin-binding protein capable of binding to serum albumin, wherein said deimmunized albumin-binding protein portion is a variant of a wild-type albumin-binding domain (ABD) of a Streptococcal Protein G, said wild-type ABD having the amino acid sequence of SEQ ID NO: 304; wherein said variant ABD has an amino acid sequence that differs from that of SEQ ID NO: 304 in comprising:

    (A) a leucine to alanine variation at ABD position 64, an isoleucine to alanine variation at ABD position 65, and a valine to alanine variation at ABD position 71; wherein said variant ABD positions are respectively positions 24, 25, and 31 of SEQ ID NO: 304; or

    (B) an asparagine to aspartic acid variation at ABD position 66, a threonine to serine substitution at ABD position 70, and a valine to alanine variation at ABD position 71; wherein said variant ABD positions are respectively positions 26, 30, and 31 of SEQ ID NO: 304;

    wherein said deimmunized albumin-binding protein portion extends the serum half-life of said polypeptide, relative to the serum half-life of said polypeptide if lacking said portion of said deimmunized albumin-binding protein.

    2. The polypeptide of claim 1, wherein the amino acid sequence of said variant ABD comprises said leucine to alanine variation at ABD position 64, said isoleucine to alanine variation at ABD position 65, and said valine to alanine variation at ABD position 71, wherein said variant ABD positions are respectively positions 24, 25, and 31 of SEQ ID NO: 304.

    3. The polypeptide of claim 1, wherein the amino acid sequence of said variant ABD comprises said asparagine to aspartic acid variation at ABD position 66, said threonine to serine substitution at ABD position 70, and said valine to alanine variation at ABD position 71, wherein said variant ABD positions are respectively positions 26, 30, and 31 of SEQ ID NO: 304.

    4. The polypeptide of claim 1, wherein the amino acid sequence of said variant ABD differs from that of SEQ ID NO:304 only by said leucine to alanine variation at ABD position 64, said isoleucine to alanine variation at ABD position 65, and said valine to alanine variation at ABD position 71, wherein said ABD positions 64, 65 and 71 are respectively positions 24, 25 and 31 of SEQ ID NO:304.

    5. The polypeptide of claim 1, wherein the amino acid sequence of said variant ABD differs from that of SEQ ID NO:304 only by said asparagine to aspartic acid variation at ABD position 66, said threonine to serine variation at ABD position 70, and said valine to alanine variation at ABD position 71, wherein said ABD positions 66, 70 and 71 are respectively positions 26, 30 and 31 of SEQ ID NO:304.

    6. The polypeptide of claim 1, wherein the amino acid sequence of said variant ABD is the sequence of SEQ ID NO:328, or SEQ ID NO:329.

    7. The polypeptide of any one of claims 1-6, wherein said polypeptide comprises an antigen-binding molecule.

    8. The polypeptide of claim 7, wherein said antigen-binding molecule is linked to an additional portion of a deimmunized albumin-binding protein, wherein said portions of said deimmunized albumin-binding protein are both capable of binding to said albumin.

    9. The polypeptide of claim 7, wherein said antigen-binding molecule is a diabody composed of at least a first and a second polypeptide chain which interact with one another to form two antigen-binding sites, wherein at least one of said polypeptide chains comprises said portion of said deimmunized albumin-binding protein that is capable of binding to said serum albumin.

    10. The polypeptide of claim 9, wherein both of said first and said second polypeptide chains comprise said portion of said deimmunized albumin-binding protein capable of binding to said serum albumin.

    11. The polypeptide of any one of claims 9-10, wherein said first and said second polypeptide chains are covalently linked to one another.

    12. The polypeptide of any one of claims 9-11, wherein said diabody binds to:

    (A)      the Natural Killer Group 2D (NKG2D) receptor or the T-cell receptor (TCR); and

    (B)      a tumor-associated antigen.

    13. The polypeptide claim 7, wherein said antigen is a breast cancer antigen, an ovarian cancer antigen, a prostate cancer antigen, a cervical cancer antigen, a pancreatic carcinoma antigen, a lung cancer antigen, a bladder cancer antigen, a colon cancer antigen, a testicular cancer antigen, a glioblastoma cancer antigen, an antigen associated with a B cell malignancy, an antigen associated with multiple myeloma, an antigen associated with non-Hodgkin's lymphoma, or an antigen associated with chronic lymphocytic leukemia.

    14. The polypeptide of any one of claims 9-12, wherein said diabody has an antigen-binding site that binds to a breast cancer antigen, an ovarian cancer antigen, a prostate cancer antigen, a cervical cancer antigen, a pancreatic carcinoma antigen, a lung cancer antigen, a bladder cancer antigen, a colon cancer antigen, a testicular cancer antigen, a glioblastoma cancer antigen, an antigen associated with a B cell malignancy, an antigen associated with multiple myeloma, an antigen associated with non-Hodgkin's lymphoma, or an antigen associated with chronic lymphocytic leukemia.

    15. Use of a polypeptide that is a portion of a deimmunized albumin-binding protein capable of binding to serum albumin to extend the serum half-life of a covalently linked polypeptide, wherein said deimmunized albumin-binding protein portion is a variant of a wild-type albumin-binding domain (ABD) of a Streptococcal Protein G, said wild-type ABD having the amino acid sequence of SEQ ID NO: 304; wherein said variant ABD has an amino acid sequence that differs from that of SEQ ID NO: 304 in comprising:

    (A) a leucine to alanine variation at ABD position 64, an isoleucine to alanine variation at ABD position 65, and a valine to alanine variation at ABD position 71; wherein said variant ABD positions are respectively positions 24, 25, and 31 of SEQ ID NO: 304; or

    (B) an asparagine to aspartic acid variation at ABD position 66, a threonine to serine substitution at ABD position 70, and a valine to alanine variation at ABD position 71; wherein said variant ABD positions are respectively positions 26, 30, and 31 of SEQ ID NO: 304,

    wherein said extension of serum half-life is relative to the serum half-life of said covalently linked polypeptide if lacking said portion of said deimmunized albumin-binding protein.

    16. The use of claim 15, wherein said covalently-linked polypeptide is a polypeptide of an antigen-binding molecule.

    17. The use of claim 16, wherein said antigen-binding molecule is a diabody composed of at least a first and a second polypeptide chain which interact with one another to form two antigen-binding sites, wherein at least one of said polypeptide chains comprises said polypeptide that comprises said portion of said deimmunized albumin-binding protein that is capable of binding to said serum albumin.

    18. The use of claim 17, wherein each of said first and said second polypeptide chains comprise one of said polypeptide that comprises said portion of said deimmunized albumin-binding protein capable of binding to said serum albumin.

    19. The use of claim 17, wherein said polypeptide chain that comprises said portion of said deimmunized albumin-binding protein that is capable of binding to said serum albumin additionally comprises a second of such portion of said deimmunized albumin-binding protein capable of binding to said serum albumin.

    20. The use of any one of claims 17-19, wherein said first and said second polypeptide chains are covalently linked to one another.

    21. A pharmaceutical composition comprising the polypeptide of any one of claims 1-14, and a pharmaceutically acceptable carrier.