Stephen Dean Sharp & David John Hughes v GE Healthcare Bio-Sciences AB
[2014] APO 10
•19 February 2014
IP AUSTRALIA
AUSTRALIAN PATENT OFFICE
Stephen Dean Sharp & David John Hughes v GE Healthcare Bio-Sciences AB [2014] APO 10
Patent Application: 2005317279
Title:Purification of immunoglobulins
Patent Applicant: GE Healthcare Bio-Sciences AB
Opponent: Stephen Dean Sharp & David John Hughes
Delegate: Dr B. Akhurst
Decision Date: 19 February 2014
Hearing Date: 28 November 2013, in Canberra
Catchwords: PATENTS - section 59 opposition - no evidence in answer to the opposition or submissions from the applicant - inventive step - claims found to lack an inventive step in light of a prior art document cited in the application - manner of manufacture - not established that the claimed combination is obvious on the face of the specification - useful - on a proper construction the claimed invention is useful - clarity - the claims provide a workable standard suitable to the intended use - lack of clarity not established - fair basis -the claims need not be limited to the preferred embodiments - lack of fair basis not established.
Representation: Patent applicant: Did not attend the hearing
Opponent: Dr Peter Brown, Griffith Hack, Sydney
IP AUSTRALIA
AUSTRALIAN PATENT OFFICE
Patent Application: 2005317279
Title:Purification of immunoglobulins
Patent Applicant: GE Healthcare Bio-Sciences AB
Date of Decision: 19 February 2014
DECISION
The opposition is successful. The invention of claims 1-2, 4-5 and 7-17 lacks an inventive step.
I believe it is possible to overcome the deficiencies I have identified by amendment of the specification. I therefore allow GE Healthcare Bio-Sciences AB two (2) months from the date of this decision to propose such amendments.
Costs are awarded according to Schedule 8 against GE Healthcare Bio-Sciences AB.
REASONS FOR DECISION
Background
Patent application 2005317279 was filed by GE Healthcare Bio-Sciences AB via the PCT on 12 December 2005 claiming priority from application SE 0403057-3 filed on 14 December 2004 and US 60/638,316 filed on 22 December 2004. The application entered national phase on 30 May 2007 and after examination was advertised as accepted on 24 February 2011.
Stephen Dean Sharp & David John Hughes filed a notice of opposition to grant of a patent on 24 May 2011 and served the statement of grounds and particulars on 24 August 2011. Evidence in support was completed on 24 April 2013 and consisted of a declaration by Robert L. Raison (the RLR declaration) with exhibits RLR-1 to RLR-11, and declarations by librarians Ngaine Ann Pettit-Young and Fiona Grigg.
No evidence was filed in answer to the opposition. On 30 August 2013 the applicant advised that it would not attend the hearing and that it wished the opposition to be determined on the basis of the evidence already before the Patent Office. The opponent was heard in Canberra on 28 November 2013.
Grounds of opposition
At the hearing, the grounds of opposition pressed by the opponent were manner of manufacture, inventive step, usefulness, clarity and fair basis.
Onus of Proof
The request for examination in this case was filed on 15 August 2007. Therefore, the substantive amendments to the Patents Act brought about by the Intellectual Property Laws Amendment (Raising the Bar) Act 2012, including subsection 60(3A) which allows the Commissioner to refuse a patent application if satisfied on the balance of probabilities that a ground of opposition has been made out, do not apply to the present application. Instead, the onus of proof in this opposition proceeding lies with the opponent, who must establish that it is clear that a valid patent cannot be granted (F.Hoffman-La Roche AG v New England Biolabs Inc [2000] FCA 283 at [29], [67]; 50 IPR 305; Commissioner of Patents v Sherman [2008] FCAFC 182 at [18], [22]; 79 IPR 426)
For the same reason, the substantive amendments to grounds of opposition that came into effect on 15 April 2013 do not apply to the present opposition. Instead the law that applies to the grounds of opposition in this case is that which stood immediately prior to 15 April 2013.
The subject matter of the specification
The specification is titled “Purification of immunoglobulins”. On pages 1-2 of the specification immunoglobulins, which are also known as antibodies, are naturally produced in the body in response to infection and immunisation, and are a key element in the immune response. The biological activity of antibodies is today exploited in a range of different applications in the diagnostic, therapeutic and health care sectors. Complementing protein expression and production strategies, protocols for purifying antibodies are designed to obtain highly pure antibodies in a simple but cost-efficient manner.
On page 1 of the specification, the invention relates to the field of antibody preparation and more specifically to a separation matrix for isolation and/or separation of antibodies. The invention also encompasses a chromatography column that comprises the separation matrix, a method of isolating antibodies using the separation matrix and a multistep process for large-scale purification of antibodies from a crude feed.
On pages 2-3, the specification summarises traditional antibody purification methods and associated disadvantages, and discloses the “unique and powerful role” of affinity chromatography in separating biomolecules. On page 3, affinity chromatography involves specifically and reversibly adsorbing (binding) a biomolecule of interest to a ligand immobilised to an insoluble support. A sample containing the biomolecule of interest is applied under conditions that favour its binding. Unbound substances are washed away and the biomolecule recovered by changing the experimental conditions to those that favour its desorption (release) from the ligand. Affinity chromatography is said to have a concentrating effect enabling convenient processing of large sample volumes. Protein A, particularly the recombinant form, is described as a popular ligand for isolation of immunoglobulins due to its ease of use and the high purity obtained (page 3, line 23 to page 4 line 6 of the specification).
Examples 1 and 2 describe the production of a separation matrix comprising agarose particles coupled to the ligand Protein A and demonstrates its improved dynamic binding capacity in comparison to a commercially available matrix. The dynamic binding capacity of an affinity matrix is broadly its capacity to bind antibody; it provides a good indication of the suitability of a matrix for large-scale operation, where an increase greatly improves process economy (page 11 of the specification).
Example 3 describes determination by gel filtration of the gel phase distribution coefficient (Kav) for dextran standards in the Example 1 matrix. From this information the Kav for a dextran of 110 kDa is calculated. The Kav value describes the fraction of the particle volume available to a molecule of a particular size (page 19, lines 28-30).
The claims and their construction
The principles to be applied in construing a patent specification are well settled in law (Flexible Steel Lacing Company v Beltreco Ltd [2000] FCA 890 at [70] - [81]; (2001) 49 IPR 331 at 347 [70] - [81]; Pfizer Overseas Pharmaceuticals v Eli Lilly and Company [2005] FCAFC 224 at [247] - [250]; 68 IPR 1 at 52-54).
The specification was accepted with 18 claims, reproduced below. Claim 1 is the only independent claim.
Claim 1
1. A separation matrix comprising porous carbohydrate particles to which antibody-binding protein ligands have been immobilised, wherein the ligand density ranges from 5.0 to 10 mg/ml and the median particle diameter ranges from 65 to 84 μm, and wherein the gel phase distribution coefficient expressed as Kav for a dextran of size 110 kDa ranges from 0.65 to 0.85.
There are several terms in claim 1 that require careful construction, which I have addressed below.
Antibody-binding protein ligands and ligand density
In view of the definitions of the terms “ligand” and “antibody-binding protein” on pages 6-7 of the specification, I construe the term “antibody-binding protein ligand” to mean a protein capable of binding target antibodies by any mechanism. The ligand density expressed in mg/ml refers to the milligram weight of protein ligand attached to each millilitre volume of separation matrix (RLR at [4.11]).
Median particle diameter (d50v)
The opposed specification on page 10 states that the particles of the present separation matrix are polydisperse, i.e. they exist in the form of a suspension of particles having a range of sizes (The Oxford English Dictionary). The specification teaches that the size of polydisperse particles can be defined by way of “the median particle diameter of the cumulative distribution frequency (d50v)”. The term “median particle diameter” refers to the diameter at the middle of the distribution of particle diameters (RLR at [4.41(c)]). Claim 1 requires this diameter to be between 65 to 84 μm.
Gel phase distribution coefficient (Kav)
The evidence establishes that in the process of gel filtration, a sample containing molecules to be separated is applied to the top of the chromatography column and allowed to flow through. Those molecules that are small enough to enter the pores of the matrix particles become temporarily trapped and as a consequence their passage through the column is slowed. Molecules that are too large to enter the pores of the particles remain outside and move relatively freely through the column in the interstitial fluid (void) volume between the particles. Where they are not bound by affinity ligand, molecules of the same size flow through the matrix at the same rate and elute from (exit) the column at approximately the same time. The volume at which the molecule elutes from the column is called the elution or retention volume, which is related to the Kav for that molecule in a matrix. (RLR at [2.35]-[2.36]).
The Kav is a column independent variable that reflects the fraction (proportion) of the particle volume that is available to a molecule of a particular size (the specification at page 7, lines 13-18 and page 19 lines 26-30; consistent with RLR at [2.36]). On page 7 of the specification, the Kav of a particular molecule is calculated according to the formula:
Kav = (VR-V0)/VC-V0)
where:
VC is the is the geometric volume of the column (represented in grey in the diagram below):
V0 is the interstitial void volume of the column, i.e. the volume of fluid in the column that lies outside the matrix particles (RLR at [2.35], represented in grey in the diagram below):
VR is the retention volume for a molecule of a given size.
Claim 1 requires the matrix to have a Kav for a dextran of size 110 kDa within the range 0.65 to 0.85.
Dependent claims 2-18
2. A separation matrix according to any one of the preceding claims, wherein the ligand density ranges from 5.5 to 9.0 mg/ml.
3. A separation matrix according to any one of the preceding claims, wherein the median particle diameter is about 75 μm.
4. A separation matrix according to any one of the preceding claims, wherein the particles comprise a cross-linked polysaccharide material.
5. A separation matrix according to claim 4, wherein the polysaccharides have been allylated before gelation.
6. A separation matrix according to any one of the preceding claims, which provides a dynamic binding capacity above 40 mg antibody/ml separation matrix at a residence time of 2.4 minutes.
7. A separation matrix according to any one of the preceding claims, wherein the ligands comprise Fc-binding protein.
8. A separation matrix according to claim 7, wherein the Fc-binding protein is Protein A.
9. A separation matrix according to any one of the preceding claims, wherein the ligands comprise a monomer, dimer or multimer of Protein A domains.
10. A method of purification of antibodies by affinity chromatography, which comprises contacting a process feed with a separation matrix according to any one of the preceding claims to adsorb antibodies, optionally washing the antibodies adsorbed to the separation matrix, adding an eluent that releases the antibodies from the separation matrix and recovering the antibodies from the eluate.
11.A method according to claim 10, wherein the process feed comprises fermentation broth.
12. A method according to claim 10 or 11, wherein the antibodies are monoclonal antibodies.
13.A method according to any one of claims 10 to 12, wherein the separation matrix is present in a chromatography column through which the process feed and eluent are passed.
14. A method according to claim 13, wherein the antibodies are released from the separation matrix by adding an eluent [sic] to prepare the separation matrix for re-use.
15. A multi-step process for the purification of antibodies, which comprises a capture step according to any one of claims 10 to 13 followed by one or more steps for intermediate purification and/or polishing of the antibodies.
16. A process according to claim 15, wherein the capture step is followed by hydrophobic interaction and/or ion exchange chromatography.
17. A process according to claim 15, wherein the capture step is followed by multi-modal anion or cation exchange chromatography.
18. A separation matrix according to claim 1 substantially as hereinbefore described with reference to the drawing and/or examples.
Claim 18 limits the separation matrix of claim 1 to comprising agarose particles with recombinant Protein A immobilised at a ligand density of 7.3 mg/ml, a median particle diameter of 80 µm and a Kav of 0.69 for a dextran size of 110 kDa (Example 1, Example 3 page 19 line 30) and/or a dynamic binding capacity as presented graphically in Figure 1. Figure 1 discloses an increase in dynamic binding capacity over 10 minutes residence time in the exemplified matrix including, at 2.4 minutes residence time, approximately 50 mg bound antibody/ml separation matrix (page 15, para 1 and Figure 1).
Inventive step
The opponent has cited the following document against all claims for inventive step purposes.
US 6602990 B1 (Berg, H.) 5 August 2003 (US’990)
Consideration of inventive step in this case requires comparison of the claimed invention with the information disclosed by this document. I will first address to the disclosure of US’990 relevant to the features of opposed claim 1, and then the question of whether the opposed claims involve an inventive step in the light of that disclosure.
The US’990 disclosure relevant to claim 1
‘A separation matrix comprising porous carbohydrate particles …’
US’990 is titled “Process for the production of a porous cross-linked polysaccharide gel and its use as a gel filtration media and in chromatography”. It discloses a process for the production of a porous polysaccharide gel for gel filtration and chromatographic applications including affinity chromatography (US’990 column 1; RLR at [3.4]-[3.6]). Agarose gels are exemplified (Example 1; RLR at [4.41(a)]). Thus, US’990 describes a separation matrix comprising porous carbohydrate particles.
‘… to which antibody-binding protein ligands have been immobilised, wherein the ligand density ranges from 5.0 to 10 mg/ml…’
US’990 does not exemplify a matrix with antibody-binding protein ligands immobilised to the particles. However, it explicitly states that that the gel “according to the invention” can be used, after modification, in different types of affinity chromatography (column 4, lines 57-58). Relevant groups with specific affinity include “IgG-binding protein (Protein A, G, L etc.)” (column 4, line 66-67). US’990 does not disclose or suggest any numerical value for the density of the affinity ligands.
‘… and the median particle diameter ranges between 65-84 mm …’
The opposed specification on page 10 discloses that the size of polydisperse particles can be defined by way of the parameter d50v, described as the median particle diameter of the cumulative distribution frequency. In contrast, US’990 defines d50v as the mean particle size distribution.
Professor Raison at [4.41(c)] addresses this discordancy as follows:
“Tables 1a, 2a and 3a of US 6,602,990 set out the mean particle size distribution (d50v) of the gels designated Examples 1 to 8 in the column marked “d50v”. I note that page 10, lines 15 to 17, of the specification for AU2005317279 indicates the term “median particle diameter” as used in the specification refers to the median particle diameter of the cumulative volume distribution (d50v). I understand the term “median particle diameter” as used in the specification for AU2005317279 to mean the diameter which is in the middle of the distribution of particle diameters. In the context of the specification, I consider the term “median particle diameter” to have the same meaning as the term “mean particle distribution” referred to in US 6,602,990.”
Clearly the mean and the median values in a distribution can vary independently depending on the nature of a distribution (e.g. Gaussian or otherwise) and I do not understand Professor Raison to be suggesting otherwise. Rather, I understand from his evidence that insofar as both documents refer to d50v, in each case they refer to the same parameter, albeit that the opposed application describes it as the median particle diameter and US’990 as the mean particle distribution.
The opposed specification on page 10, describes d50v as a “commonly used way of defining particle size within this field”. I would expect a commonly used term to have a consistent meaning in the art and Professor Raison’s conclusion appears reasonable in the circumstances. On this basis, I accept that the d50v values disclosed in US’990 are equivalent to, and directly comparable to, the median particle diameter as claimed in the opposed specification.
The opponent relied in particular on the exemplified matrices in US’990 designated Examples 7b, 8c and 8d. Table 3a of US’990 provides the d50v of the Example 7b and 8d matrices as 84 mm and 87.5 mm, respectively. That of Example 8c is not disclosed. While Professor Raison speculates at [4.41(c)] that the median particle diameter of the Example 8c matrix would be in the claimed range, he ultimately concludes that this Example only “possibly” represents a separation matrix comprising a porous carbohydrate particle having a median particle diameter which ranges from 65-84 mm.
I conclude that only the Example 7b matrix of US’990 has a particle diameter falling within the range specified by claim 1.
‘… and wherein the gel phase distribution coefficient expressed as Kav for a dextran size of 110 kDa ranges from 0.65 to 0.85 …’
US’990 provides the Kav of proteins of different sizes in the exemplified matrices. In contrast, the opposed specification at Example 3 determines the Kav of dextran standards and claims the matrix in terms of the Kav for a dextran of 110 kDa.
Professor Raison at [2.37] - [2.38] explains that the Kav for a protein is proportional to the log of the molecular weight (MW) of the protein. A plot of Kav against MW for a series of proteins of varying MW can be used to determine the Kav of a protein having a known MW for a particular separation matrix. However, can the protein standardisation data be used to determine the Kav for a dextran falling within the same molecular weight range?
Professor Raison’s evidence on this point is clear and unambiguous. He does not support the opponent’s submission that the protein standards can be used to accurately calculate the Kav of dextrans. Nevertheless, he states that protein Kav data can provide an approximate estimation of the Kav of dextrans, and on this basis Professor Raison concludes that Examples 7b, 8c and 8d would have a Kav for a dextran of size 110 kDa that falls within the range 0.65 to 0.85 (RLR at 4.41(d)).
In the absence of evidence to the contrary, I accept that US’990 in Examples 7b, 8c and 8d discloses matrices with a Kav as specified in opposed claim 1.
In summary, the evidence establishes that US’990 discloses a separation matrix comprising porous carbohydrate particles to which antibody binding ligands may be immobilised. In Example 7b, US’990 discloses a matrix with a median particle diameter of between 65 and 84 mm and a Kav for a 110 kDa dextran within the range 0.65 and 0.85, as required by claim 1.
Consideration of inventive step
Subsection 7(2) provides that an invention is taken to involve an inventive step when compared to the prior art base unless it would have been obvious to a person skilled in the relevant art in the light of common general knowledge within Australia before the priority date, whether considered alone or together with information made publicly available anywhere in the world. “Obvious” means “very plain” and a “scintilla of inventiveness” is all that is required (Lockwood Security Products Pty Ltd v Doric Products Pty Ltd (No 2) [2007] HCA 21 at [51] - [52]; (2007) 72 IPR 447 at 461 [51] - [52]).
Section 7(3) provides that a document is prior art for the purposes of inventive step if the person skilled in the art could, before the priority date of the relevant claim, be reasonably expected to have ascertained, understood, and regarded the document as relevant.
The test for obviousness was provided by Justice Aicken in Wellcome Foundation Ltd v VR Laboratories (Aust) Pty Ltd [1981] HCA 12 at [45]; 148 CLR 262 at 286 as follows:
“The test is whether the hypothetical addressee faced with the same problem would have taken as a matter of routine whatever steps might have led from the prior art to the invention, whether they be the steps of the inventor or not.”
More recently, in Aktiebolaget Hässle v Alphapharm Pty Ltd [2002] HCA 59 at [53]; 212 CLR 411 at [53] the High Court approved the approach taken in Olin Mathieson Chemical Corporation v Biorex Laboratories Ltd [1970] RPC 157 at 187 in which Graham J had posed the reformulated Cripp’s question:
“Would the notional research group at the relevant date, in all the circumstances, … directly be led as a matter of course to try [the claimed combination of integers] in the expectation that it might well produce a [useful or desired result]?” (Emphasis in original)
The usual approach to obviousness is the problem-solution approach. Once the problem has been formulated, and the common general knowledge or prior art base have been determined, the question of whether the claimed solution is obvious must be addressed.
The problem
The description summarises the problems associated with traditional chromatographic separation techniques when used for antibody purification and the relative advantages of affinity chromatography for this purpose. A number of prior art separation matrices, including affinity matrices, are identified on pages 4-5. On page 5, a need is identified for alternative separation matrices for purification of antibodies or antibody constructs, which observe the demands of purity, safety, potency and cost effectiveness”. I conclude that the problem addressed by the application is the provision of such a matrix.
The applicant’s solution is the claimed separation matrix. What must be determined is whether the separation matrix of claims 1-9 and 18 would have been obvious in light of US’990 at the priority date of the application and if so, whether the use of that matrix in the methods and processes of claims 10-17 would also have been obvious.
The person skilled in the art
I accept the opponent’s submission that the hypothetical person skilled in the art in this case encompasses immunologists and other persons with expertise in the preparation of separation matrices and the use of such matrices in antibody isolation. It is clear from Professor Raison’s curriculum vitae and evidence that he has substantial experience before the priority date in this field and as such he is a person skilled in the relevant art.
Ascertained, understood and regarded as relevant
The opponent’s expert, Professor Raison, states that since 1976, he has kept himself informed of developments in Australia and overseas in the relevant field of art by reading journal articles and attending conferences in the field of immunology and by frequent discussions with his colleagues (RLR at [2.20], [20.22]). Since 1994, he routinely reviewed the Current Contents online, Medline and Pubmed databases for information in the relevant art (RLR at [2.18]-[2.21]). Regarding patent documents, since before 14 December 2004, Professor Raison was aware of patents both as an intellectual property right and as a source of information (RLR at [2.22]).
Prima facie, the person skilled in the art seeking information in order to solve the problem identified above, could be reasonably expected to have ascertained US’990. Having done so, it is reasonable to conclude that the skilled person would understand the information in US’990 and regard it as relevant.
The opponent’s case for lack of inventive step
The opponent’s case is essentially that since US’990 discloses a separation matrix comprising porous carbohydrate particles having the median particle diameter and Kav as specified in claim 1, the only step taken over the matter disclosed in US’990 (the use of that matrix in affinity chromatography at the specified ligand density) had been carried out before the priority date for many years by Professor Raison and others on commercially available matrices. For this reason, the opponent submitted that based on the common general knowledge in the art and the information in US’990 the person skilled in the art would directly be led as a matter of course to the claimed invention, with a reasonable expectation that it would produce an affinity matrix useful for the isolation of antibodies.
The matrix claims
Claim 1
I have found above that US’990 discloses a separation matrix comprising porous carbohydrate particles and identifies its suitability for affinity chromatography. Further, in Example 7b, US’990 discloses a separation matrix with a median particle diameter and a gel phase distribution coefficient expressed as Kav for a dextran size of 110 kDa within the ranges specified by claim 1.
The matrix of claim 1 will lack an inventive step if the person skilled in the art, faced with the problem of providing alternative matrices for the purification of antibodies or antibody constructs and having read US’990, would directly be led as a matter of course to try the matrix of Example 7b with antibody-binding ligands immobilised at a ligand density between 5-10 mg/ml, with a reasonable expectation that it might well produce the desired result.
Professor Raison’s evidence at [4.42] is that faced with the problem before the priority date, the matrix of Example 7b is one of three he would have been “immediately drawn to” from those disclosed in US’990 due to their larger pore size. He provides the basis for this choice:
“Examples 7b, 8c and 8d … represent gels that I would use as a base matrix for preparing an affinity matrix for isolation of antibodies. Having the larger pore size would, in my opinion, allow the antibodies to access the internal pore space to thereby increase capture of antibody.”
Professor Raison further states that it would have been a standard routine matter for his laboratory to couple antibody-binding ligands to the matrix at a density in the range 5-10 mg/ml since his laboratory regularly obtained densities within this range when coupling affinity ligand to alternative matrices (RLR at [4.43], supported by [4.41(b)]).
I accept that the evidence establishes that the person skilled in the art, in all the circumstances, would directly be led as a matter of course to try the matrix of Example 7b (among others) with antibody-binding ligands immobilised at a ligand density between 5-10 mg/ml, with a reasonable expectation that it might well produce the desired result. It follows that Claim 1 lacks an inventive step.
Claim 2
Professor Raison at [4.48] believes that ligand densities falling within the range 5-10 mg/ml would often fall within the range 5.5-9.0 mg/ml. For the same reason as above, I find that the ligand density of claim 2 also lacks an inventive step.
Claim 3
The opponent submitted that claim 3, which requires a median particle diameter (d50v) of “about 75 mm”, lacks an inventive step because this value is an arbitrary value with no advantage. Professor Raison notes at [4.28] that this particle size is not exemplified and it is not clear to him why this value is claimed.
Of the matrices that Professor Raison identified as most relevant in US’990, the d50v values provided for two examples are 84 mm (Example 7b) and 87.5mm (Example 8d), which are well above 75 mm, and outside the reasonable range of “about” 75 mm. In the absence of any evidence or reasoned explanation to support a conclusion that that the person skilled in the art would directly be led as a matter of course to try particles with the lower median particle diameter, the opponent has not established that the matrix of claim 3 lacks an inventive step.
Claims 4-5
The Example 7b matrix is prepared according to the method of Example 1 of US’990 (column 10, lines 59-62). Example 1 produces cross-linked, agarose gel particles that are allylated before gelation (column 5 lines 19-56 and RLR at [4.49]). Since the features of claims 4-5 are present in the Example 7b matrix, insofar as these claims are dependent on claims 1-2, they necessarily lack an inventive step.
Claim 6
The opponent submitted that the dynamic binding capacity specified in claim 6 would have been an inherent feature of a separation matrix according to claim 1. If that were the case, claim 6 would be redundant. Furthermore, the opponent has identified no expert evidence or provided any reasoned explanation to support its submission. Consequently, the opponent has not established that claim 6 lacks an inventive step.
Claims 7-9
These claims specify the antibody-binding protein ligand as comprising an Fc-binding protein, in particular Protein A or a monomer, dimer or multimer of Protein A domains. I construe the latter options in claim 9 to encompass Protein A itself since this must necessarily comprise one or more of the unspecified Protein A domains as claimed.
Professor Raison states that before the priority date it would have been a routine step for him to have coupled Protein A to the matrix of US’990 [4.43]. Although this evidence was given in the context of ligand density, I accept that the evidence establishes that the person skilled would have been directly led as a matter of course to try Protein A as the affinity ligand on the Example 7b matrix with a reasonable expectation that it would produce a useful product, in view of the following additional facts:
· US’990 explicitly identifies the IgG-binding proteins A, G and L, among others) as suitable affinity ligands (column 4, lines 66-67);
· Fc-binding proteins, particularly Protein A were well known and regularly used in the preparation of affinity matrices by Professor Raison and others working in the field of antibody isolation well before the priority date (RLR at [2.8], [2.11]-[2.13], [4.50]); and
· Protein A was a commonly used affinity ligand for antibody isolation before the priority date (RLR at [2.42]).
I find that claims 7-9 do not involve an inventive step.
Claim 18
The opponent’s submissions include that claim 18 lacks an inventive step, but I have no evidence before me or any more detailed submissions on which I could make this finding. The opponent has not established that claim 18 lacks an inventive step.
The method and process claims
Claims 10-17
The evidence establishes that the method and process steps identified in claims 10-17 were all standard steps in antibody isolation before the priority date (RLR at [4.35] - [4.36], [4.51]).
Since I have found the separation matrix of claims 1-2, 4-5 and 7-9 to lack an inventive step, I can only conclude that faced with the problem I have identified above, the person skilled in the art would as a matter of routine have used the matrices of these claims according to established procedures, and thereby arrive at the invention of dependent claims 10-17.
In conclusion, I find that the invention of claims 1-2, 4-5 and 7-17 does not involve an inventive step.
Manner of Manufacture
Section 18(1)(a) requires that an invention must be a manner of manufacture within the meaning of section 6 of the Statute of Monopolies. Manner of manufacture is assessed by asking whether the claimed invention lacks the necessary quality of inventiveness on the face of the specification (NV Philips Gloeilampenfabrieken v Mirabella International Pty Ltd [1995] HCA 15 at [9]; (1995) 183 CLR 655).
The opponent submitted that the separation matrix of claims 1-2, 4-8 and 18, and the method and process of claims 10-17, does not pass this threshold test. The opponent’s case for lack of manner of manufacture is that since the preparation of the claimed separation matrix is described in terms of a previously published method (that disclosed in US 6602990) and the application of principles essentially identified as forming part of the common general knowledge in the art, it naturally follows that the end result is obvious on the face of the specification.
I accept the opponent’s summary of the disclosure. However, to make its case under this ground, the opponent must establish that the combination of integers that make up the claimed invention are obvious on the face of the specification (Welch Perrin & Co Pty Ltd v Worrel [1961] HCA 91 at [19]; (1961) 106 CLR 588 at 611; Minnesota Mining & Manufacturing Co v Beiersdorf (Australia) Ltd [1980] HCA 9 at [12]; (1980) 144 CLR 253 at 266 applied).
The specification discloses the preparation of a separation matrix falling within the scope of claim 1 (page 8, lines 15-19 and Example 1) and demonstrates that it has improved properties compared to a prior art product (page 15, para 1 and Figure 1). It is not apparent on the face of the specification that this matrix is obvious, nor that the methods and processes for its use are obvious.
The opponent has not established that the opposed claims do not define a manner of manufacture.
Useful
Section 18(1)(c) requires that a claimed invention is useful. The test for utility was provided by the Full Court of the Federal Court as follows:
“If the claimed invention does what it is intended by the patentee to do and the end result obtained is itself useful, the invention is useful within the meaning of s 18(1)(c) … . As to the first aspect, the invention as claimed must attain the result promised by the patentee …” (Ranbaxy Australia Pty Ltd v Warner-Lambert Co LLC [2008] FCAFC 82 at [141]; 77 IPR 449)
“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.” (H Lundbeck A/S v Alphapharm Pty Ltd [2009] FCAFC 70 at [81], [217]; 81 IPR 228)
At the hearing, the opponent pressed its case for inutility in respect of claim 10 and dependent claims 11-17. The opponent submitted that the terms “process feed” and “fermentation feed” in claims 10-11 should be construed broadly to encompass feeds that do not contain antibodies, the use of which would not result in the purification of antibodies.
I do not agree with the opponent’s construction of claims 10 and 11 for the reasons that follow.
The specification, including the claims, is not be read in the abstract but construed with common sense in the light of the common general knowledge in the art as at the priority date (Kimberly-Clark Australia Pty Ltd v Arico Trading International Pty Ltd (Kimberly-Clark) [2001] HCA 8 at [25]; 207 CLR 1).Absurd constructions should be avoided (Austal Ships Sales Pty Ltd v Stena Rederi Aktiebolag [2008] FCAFC 121 at [14]; (2008) 77 IPR 229). Furthermore, the Full Court of the Federal Court has recognised the danger in considering the integers of a claim individually and in isolation, which could lead to a literal rather than a purposive construction (Pfizer Overseas Pharmaceuticals v Eli Lilly and Company [2005] FCAFC 224 at [250]; 68 IPR 1; referring to the proposition by Lindgren J in Nesbit Evans Group Australia Pty Ltd v Impro Ltd [1997] FCA 1092; (1997) 39 IPR 56, based on Catnic Components Ltd v Hill & Smith Ltd [1982] RPC 183 at 243).
Claim 10 defines a method of purifying antibodies comprising contacting a process feed with a separation matrix to which antibody-binding protein ligands have been immobilised to adsorb antibodies. Claim 11 requires the process feed of claim 10 to comprise fermentation broth.
Construing the terms “process feed and “fermentation broth”, in the context of antibody purification and in the light of the common general knowledge in the art, these feeds must contain antibodies. To construe the claims to include feeds containing no antibody would be absurd.
This ground of opposition fails.
Section 40 issues
Section 40(3) of the Patents Act requires that the claim or claims in a patent specification must be clear and succinct and fairly based on the matter described in the specification.
Clarity
A patent operates as a public instrument, which must define a monopoly in such a way that it is not reasonably capable of being misunderstood (Welch Perrin & Co Pty Ltd v Worrel [1961] HCA 91 at [7]; (1961) 106 CLR 588). A claim is lacking in clarity if a third party could not ascertain whether an act would fall within the scope of the claim (Monsanto Co v Commissioner of Patents (1974) 48 ALJR 59 at 60). However, a lack of precise definition in claims is not fatal to their validity so long as they provide a workable standard suitable to the intended use (Minnesota Mining & Manufacturing Co v Beiersdorf (Aust) Ltd [1980] HCA 9 at [46]; (1980) 144 CLR 253 at 274).
The opponent submitted that in the absence of any precise definition of the term “ligand density” or a specified procedure for determining this parameter, the person skilled in the art could not ascertain the precise extent of the monopoly conferred by claims 1-2 and the dependent claims. The opponent relied on Professor Raison’s statement at [4.12] that he is not able to determine what is meant by the term “ligand density” in the context of the specification, because:
“… the ligand density as an amount of ligand per volume of matrix is an arbitrary value which depends largely on how the volume of the matrix is determined. Whilst I, and others in my field of research, refer to ligand density in general terms as mg/ml, absent specific context of how the ligand density, and, in particular. the volume of the matrix, is determined, the term is ambiguous. In my opinion, ligand density in mg/ml is an inaccurate and indeterminate measure of the amount of ligand present in a matrix.”
Before the priority date (and now) there were multiple techniques for determining ligand density of a matrix, but no accepted standard technique (RLR at [4.20]-[4.21], [4.26]). In particular, different hydration states of the gel matrix can markedly alter the gel volume, having greatest impact on the ligand density value (RLR at [4.20]). The choice from multiple alternative techniques for determining the amount of ligand bound to the matrix introduces further variability (RLR at [4.21]). Any given matrix may have a ligand density falling within the claimed range using one method, but outside this range with an alternative method (RLR at [4.26]). In order to obtain consistent and comparable values for ligand density in Professor Raison’s laboratory before 14 December 2004, a single method was used to monitor batches of Protein A-coupled gel (RLR at [4.13]).
I understand from Professor Raison’s evidence that he fully understands the plain meaning of the term “ligand density”, but he also appreciates its inherent limitations. In the absence of specific information on the methods used to determine (i) the amount of ligand and (ii) the volume of the gel, he considers the scope of the term as it used in the specification difficult to construe. However, this alone would not render the claim unclear if a meaning can be given to the claim such that it would provide a workable standard suitable to its intended use.
Professor Raison acknowledged in the paragraph reproduced above that he and others in his field referred to ligand density in general terms as mg/ml (RLR at [4.12]). In addition, there are documents in evidence that refer to ligand density in mg/ml, without providing any information on the method used to determine the amount of ligand or the state of the matrix when measuring its volume. These are:
· Exhibit RLR-6: an extract from “Affinity Chromatography, Principles Methods” (1988) by Pharmacia LKB Technology. On page 16, in the section relating to methods for coupling ligands to particular gel types, there is a paragraph headed “Ligand concentration dependence” which stresses the need to optimise ligand concentration for efficient adsorption. For protein ligands the recommendation is made to couple 5-10 mg protein per ml of gel.
· Exhibit RLR-11: Hermanson et al. “Immobilized Affinity Ligand Techniques” Academic Press at page 91 states that the maximum ligand density achieved for Protein A is over 30 mg/ml gel, and for human IgG over 20 mg/ml.
The evidence establishes that there are potential difficulties in expressing ligand density in mg/ml without specifying the means for determining ligand concentration or the hydration state of the gel (RLR at [4.12]). However, the use of the term in the specification is consistent with its use in other texts in this field, which presumably the person skilled in the art is expected to interpret appropriately. Since it appears acceptable in the art to express ligand density in mg/ml without qualification, I conclude that this must be a workable approach.
Turning to the specification, since it does not disclose any method for determining the ligand density of the matrix, it follows that the reader may assume that it may be measured in any reasonable manner (Colin Leslie Young v Arthur Yates & Co Ltd and W. Neudorff GmbH [2001] APO 68 at [2b]).
The value obtained for ligand density depends largely on the hydration state of the gel when matrix volume is measured (RLR at [4.12], [4.20]). Professor Raison identifies various strategies for determining matrix volume, including: settling the gel either under gravity or by running buffer through the matrix under a particular flow rate; packing it by centrifugation of a matrix slurry; or drying the gel under vacuum (RLR at [4.20]). Prima facie, these strategies all appear reasonable.
In order to ascertain whether any given matrix would fall within the scope of the claims the person skilled in the art may measure the ligand density of a matrix by any reasonable method known in the art. If the value obtained falls outside the claimed range, then the matrix does not fall within the scope of the claim.
For the reasons provided above, I consider that claims 1 and 2 provide a workable standard by which the person skilled in the art could determine whether any given matrix would fall within the scope of the claims.
I find the claims are clear.
Fair Basis
Section 40(3) requires that the claim or claims in a patent specification must be fairly based on the matter described in the specification.
Regarding the consideration of fair basis, the High Court in Lockwood Security Products Pty Ltd v Doric Products Pty Ltd (Lockwood) [2004] HCA 58 at [69]; (2004) 217 CLR 274 at 300 approved the words of Gummow J in Rehm Pty Ltd v Websters Security Systems (International) Pty Ltd (1988) 81 ALR 79 at 95:
“… the question is whether there is a real and reasonably clear disclosure in the body of the specification of what is then claimed, so that the alleged invention as claimed is broadly, that is to say in a general sense, described in the body of the specification.”
In Olin Corporation v Super Cartridge Co Pty Ltd [(1977) 180 CLR 236 at 240], Barwick CJ said:
“The question whether the claim is fairly based is not to be resolved, in my opinion, by considering whether a monopoly in the product would be an undue reward for the disclosure. Rather, the question is a narrow one, namely whether the claim to the product … as expressed, travels beyond the matter disclosed in the specification.”
The opponent submitted that claims 1, 10 and 11 (and dependent claims) lack fair basis.
Claim 1
The opponent contended that opposed claim 1 lacks fair basis in that it refers to “porous carbohydrate particles” and to “antibody-binding protein ligands” whereas the only separation matrix exemplified is porous agarose particles to which are bound Protein A. As a result, the opponent argued that claim 1 travels beyond the invention described and exemplified in the specification, which does not provide a real and reasonably clear disclosure of the invention of the subject matter of claim 1.
However, the requirement for a “real and reasonably clear disclosure” does not limit disclosures to the preferred embodiments. Furthermore, “[t]he circumstance that something is a requirement for the best method of performing an invention does not make it necessarily a requirement for all claims”. (Lockwood at [69]).
Under the heading “Detailed description of the present invention” the opposed application provides consistory statements consistent with the features of claim 1. Regarding the term “porous carbohydrate particles” the specification on page 9 states:
“The present base matrix may comprise porous particles made from any material within the specified values of the gel phase distribution coefficient, which provides the substantial improvement of dynamic binding capacity (DBC) described herein.”
“Advantageous” embodiments of porous carbohydrate particles are identified in the same paragraph, including commercially available products:
“… the particles of the support matrix are made from a cross-linked carbohydrate material such as agarose, agar, cellulose, dextran, chitosan, konjac, carrageenan, gellan, alginate etc, which are easily prepared according to standard methods, such as inverse suspension gelation … . In one embodiment, the carbohydrate material is highly cross-linked agarose, such as Sepharose™ (Amersham Biosciences, Uppsala, Sweden)”
In the passages reproduced above, the body of the specification describes the particles in terms of a functional requirement and identifies the preferred carbohydrates. As such it provides a real and reasonably clear disclosure of porous carbohydrate particles as claimed.
Regarding the term “antibody-binding protein ligands”, on page 8 of the specification these include Protein A, G and/or the k-binding protein L (page 8, lines 21-22, 29-30). Protein A is a preferred embodiment, particularly the recombinant form (page 8, lines 22-25). “Protein A” is broadly defined as defined as Protein A or IgG-binding functional equivalents thereof such as one or more Protein A domains (page 8, lines 25 to page 8A). Alternatively, the ligands are antibody-binding peptides (page 9, line 10). In these passages of the description, the antibody-binding ligands are described in broad and partly generic terms providing a real and reasonably clear disclosure of this feature.
Claims 10 and 11
Regarding the fair basing of claims 10 and 11, the opponent submitted that no “process feed” or “fermentation broth” is exemplified and the expressions are so broad as to encompass feeds that do not contain antibody. On this basis, the opponent contended that the specification contains no real and reasonably clear disclosure of the invention of these claims. However, I have found above that claims 10 and 11 do not encompass feeds without antibody. Furthermore, the description refers at page 1 to the invention encompassing purification of antibodies from a crude feed. Similar statements regarding crude or process feeds are made on pages 6, 12 and 14. It follows that the opponent’s opposition on this basis cannot succeed.
Consideration of fair basing requires “a comparison between the matter described in the specification and the claim which defines the scope of the monopoly” (Kimberly-Clarke at [15]; 207 CLR 1). In this case, the subject matter of claims 1, 10 and 11 is broadly described in the body of the specification. There is nothing in the specification that is inconsistent with the invention as claimed.
The opponent has not established that claims 1, 10 and 11 lack fair basis.
CONCLUSION
The opponent has succeeded in its opposition. Claims 1-2, 4-5 and 7-17 lack an inventive step.
It is possible to overcome the deficiencies I have identified by amendment of the specification. I therefore allow GE Healthcare Bio-Sciences AB two (2) months from the date of this decision to propose such amendments.
COSTS
It is normal in actions before the Commissioner for costs to follow the event. The opponent has been successful in its opposition, and I see no reason to depart from the normal award of costs. I award costs according to Schedule 8 against the applicant.
Dr B Akhurst
Delegate of the Commissioner of Patents
0
19
0