«Abstract The high investments necessary to bring an antibody therapeutic to the market require a sound patent strategy. Although compound protection ...»
IP Issues of Therapeutic Antibodies
The high investments necessary to bring an antibody therapeutic to the
market require a sound patent strategy. Although compound protection provides
the broadest scope of protection, other ways of follow-up protection should be
considered by innovators to achieve as long protection as possible. Further, in case
a theoretical antibody against a given target is already prior art, innovators should
be aware of methods to create compound protection for second or higher generation antibodies.
Á IP Á Antibodies Á Compound Á Protection Á Biosimilars Keywords Patent 1 Introduction Therapeutic antibodies are the fastest growing group of protein therapeutics. With a limited set of underlying technologies, drugs for a wide area of indications, including cancer, autoimmunity, neurodegeneration, and infections, can be generated.
Table 1 shows the therapeutical antibodies which have, in 2010, achieved global sales of one billion USD or more (data taken from company information).1 Note The phrase ‘‘key patent’’ refers to only one member of a patent family that exists for the product. INN, international non-proprietary name.
U. Storz (&) Michalski Huettermann & Partner Patent Attorneys, Neuer Zollhof 2, 40221 Duesseldorf, Germany e-mail: email@example.com U. Storz et al., Intellectual Property Issues, SpringerBriefs in Biotech Patents, DOI: 10.1007/978-3-642-29526-3_1, Ó The Author(s) 2012 Table 1 Therapeutic mAbs which have, in 2010, achieved global sales C1 billion USD Antibody Brand name Company Key indication Target Key IP right US Key IP Gobal sales right EP in 2010 (billion USD) Inﬂiximab Remicade J&J Rheumatoid Arthritis TNFa 8.0 Bevacizumab Avastin Roche Colon cancer VEGF-A US7060269 EP0666868 6.8 EP1167384 EP1325932 Rituximab Rituxan Roche Non Hodgkin Lmyphoma CD20 US5736137 EP0669836 6.7 Adalimumab Humira Abbott Rheumatoid Arthritis TNFa US6090382 EP0929578 6.5 US6509015 Trastuzumab Herceptin Roche Breast Can
As discussed earlier in this book series,2 the development of a new drug is a costly and time-consuming matter. According to a study performed by the Tufts Center for the Study of Drug Development, the estimated average costs of developing a new Biologic are 1.2 billion USD,3 while development times are slightly longer than those reported for small molecule drugs.4 In order to recover the expenses invested into research and development of new biopharmaceutics, particularly into a new antibody, patents are an indispensable tool, as they provide an exclusive right with respect to the protected subject matter.
Only the patentee, or his licensee, is thus allowed to exploit the invention commercially, e.g., by marketing the protected antibody.
Antibodies are proteins and, as such, chemical compounds. For this reason, antibody patents are subject to similar principles as patents related to small molecular drugs, although some differences apply, particularly with respect to inventive step.5 The basic principles for protecting antibody compounds will be discussed in the following. Further, additional ways to create follow-up protection for antibody therapeutics are discussed.
2 Compound Protection
Compound protection is probably the most important protection antibody companies can rely on, as it provides an exclusive right to offer and sell the respective antibody on different markets. Furthermore, while patents protecting a particular technology expire after, roughly, two decades, it remains still possible to achieve compound protection for a new antibody even after expiry of the respective method patents used to generate, or produce, the said new antibody.
Different possibilities exist to specify an antibody in a patent. As a general rule, the speciﬁcation of an antibody by its target provides the broadest scope of protection, but can only be achieved at a very early stage, e.g., when the target has been discovered and ‘‘deorphaned’’, i.e., attributed with a physiological role. A patent protecting second or higher generation antibodies against the said target has, usually, a narrower scope, because more technical details (e.g., structural or functional) are required in the patent claim to make it novel over the prior art. In even later stages, combinations comprising an antibody against the said target plus one or more other compounds, or dosage regimen or new medical uses of the said antibody, can be claimed, yet patents of this type have an even narrower scope of
protection, allowing, e.g., off-label use by personal prescription. In the following, the different types of protection will be discussed brieﬂy. In Table 2, examples are given for each type.
2.1 Speciﬁcation by Target In case a new protein has been discovered and a therapeutic use thereof has been disclosed, both the European Patent Ofﬁce (EPO) and the United States Patent and Trademark Ofﬁce (USPTO) grant claims related to a theoretical antibody against the said protein,6,7 even if the applicant has not produced a real antibody, or provides no data or enablement related to such antibody. Claims of such type have, obviously, the broadest scope, as they encompass all future antibodies against the said target put into practice later on (i.e., during the 20 years after ﬁling of the target patent). The ofﬁces’ rationale is that the provision of a novel protein X enables a skilled artisan to produce an antibody against said protein. Therefore, it is considered a fair reward for the applicant of protein X to be granted a claim related to an antibody against the protein. Once granted, the scope of protection of such claim extends to next generation antibodies against protein X as well. This means that somebody who provides a well-deﬁned antibody against protein X will be, in his right to practice, dependent on the assignee of the ﬁrst-generation patent, despite the fact that the said assignee has never provided a ‘‘real’’ antibody, and although he himself might as well be awarded a Patent on his antibody.
While cellular signaling processes are today well understood, new potential targets for antibody therapy are still being discovered. Today, about 100 such targets are addressed by approved biopharmaceuticals,8 but the spectrum of soluble proteins or membrane receptors yet undiscovered that represent potential therapeutic targets should be much higher. Although the evaluation of a new target and the subsequent development of a respective antibody are costly endeavors, recent advancements in antibody technology may accelerate the validation of new targets, in particular those relevant to cancer, autoimmune diseases, infectious diseases, and neurodegenerative diseases. Patents which merely claim any conceivable antibody against a new target will thus be a frequent sight even in the future.
However, with respect to the currently established targets (see Table 1), the respective patents are about to expire, or have already expired. In case a company develops a second antibody against a target addressed by these antibodies, other strategies are necessary to protect such second generation antibodies.
2.2 Speciﬁcation by Target-Independent Functional Properties Antibodies can have functional properties which are target-independent. The development of such a new functional property can thus give rise to a patent the scope of which extends to all antibodies, irrespective of the target they bind, which have such property. This can, for example, apply to increased effector function by sequence engineering (e.g., US7863419 by Biogen) or post-translational glycoengineering of the Fc region (e.g., US7214775 by Biowa, or EP1071700 by Glycart), to increased serum half-life by Fc glycoengineering (e.g., US7361740 by Protein Design Labs), or to increased antibody-dependent cell cytotoxicity (ADCC) by using a novel expression system which, as such, creates an N-glycan structure which is essentially fucose-free (e.g., WO2011107520 by Cilian AG).
2.3 Speciﬁcation by Epitope
Another way to create patent protection for a second generation antibody is to claim a speciﬁty against a given epitope, or subdomain, of a target, in case the said epitope has not yet been described as clinically relevant. This approach makes sense in case the target as such has already been described, particularly when blocking only a speciﬁc epitope instead of the whole target may yield at least a theoretical beneﬁt. However, the scope of a patent claiming an antibody against a speciﬁc epitope of a given target will not encompass antibodies against other epitopes of the same target.
This has been, for example, decided by a US District Court9 recently in a case where Genentech and Biogen sued GlaxoSmithKline (GSK) and Genmab for infringement of a patent protecting Rituxan, namely by GSK’s Arzerra. Both Arzerra and Rituxan target CD20; however, Arzerra binds an epitope of the latter different from Rituxan, and with a different afﬁnity. US Patent 7682612 claims the treatment of Chronic Lymphatic Leukemia by administration of an anti-CD20 antibody, and is thus not per se restricted to a particular epitope thereof. However, in order to overcome an ofﬁce objection related to lack of enablement, Biogen has, during the patent prosecution, stated that the term ‘‘anti-CD20 antibody’’ shall mean antibodies having similar afﬁnity and speciﬁty as Rituxan. The respective court construed the patent claims as being restricted to anti-CD20 antibodies having similar afﬁnity and speciﬁty as Rituxan and thus concluded that Arzerra would not fall under the scope of the said patent. Although the claim language as such was not restricted to Rituxan, the court construed the claim in such way because of the applicant’s statements made in the prosecution history.
2.4 Speciﬁcation by Target-Dependent Functional Properties Another way to create patent protection for a second generation antibody is to specify the latter through target-dependent functional properties, e.g., binding afﬁnity against a given target, or competitive binding. The former is often done by claiming a minimum afﬁnity to a target. In such case, all later antibodies having even better afﬁnity will fall under the scope of protection of such patents, even if they have no material relationship to the antibody which has been provided by the patentee. Existing patents with these claims are a real threat to competitors, particularly to those specializing in antibody optimization (‘‘Biobetters’’).
2.5 Speciﬁcation by Sequence
Yet another way to create patent protection for a second or higher generation antibody is to specify a sequence thereof. Claimed sequences are commonly speciﬁed in such a way that, besides the mere sequence, a certain similarity interval (e.g., 85%) is comprised as well. In antibody claims, this makes little sense as the speciﬁty of a given antibody is highly dependent on its sequence.
Therefore, second or higher generation antibody claims are commonly drafted in such form that a DNA or AA sequence is claimed (e.g., SEQ ID No 1), sometimes together with possible variations (e.g., R112T). The scope of protection is thus clearly deﬁned, yet quite narrow. Competitors which replace one of the claimed residues by a residue which is not claimed thus no longer fall under the literal scope of the patent, although the antibody may retain its function despite such modiﬁcation.
Most legal systems provide doctrines of equivalents to anticipate situations as discussed above. As a rule of thumb, German judges10 tend to provide a broader scope of equivalence than UK judges,11 although attempts have been made under the European Patent Convention to establish a uniformal deﬁnition of the term ‘‘equivalent’’.12 There is, however, no court decision in the US or in Europe which has yet deﬁned the scope of equivalence for biosequence claims. This means that it is uncertain how far a competitor must amend the claimed sequence to make sure not to be sued for equivalent infringement.
It is yet noteworthy that, after the ‘‘Festo’’ decision issued by the U.S. Supreme Court,13 legal action related to equivalent infringements can no longer be enforced Three-step approach, as applied in the BGH decisions ‘‘Kunststoffrohrteil’’, ‘‘Schneidmesser I’’, ‘‘Schneidmesser II’’, ‘‘Custodiol I’’, ‘‘Custodiol II’’, GRUR 2002, 511–531.
‘‘Catnic test’’ as applied in Kirin-Amgen, Inc. v. Hoechst Marion Roussel Ltd. (2004) UKHL 46 (2004-10-21).
Article 2 of the Protocol on the Interpretation of Article 69 EPC.
Festo Corp. versus Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722 (2002).
8 U. Storz in the US if, during patent prosecution, the scope of the patent has been narrowed in such a way that the alleged infringement is no longer covered by the literal scope of protection (so-called ‘‘prosecution history estoppel’’).
The effect of this ruling on antibody sequence claims which are narrowed down during prosecution (e.g., from a sequence claim reciting ‘‘amino acid Seq. ID No.
1 and any sequence which has 85% identity to the former’’ to a claim which is restricted to the mere Seq. ID No. 1) has not made its way into case law yet, but it is to be expected that, in such case, competitors can easily circumvent the scope of protection by amending a single amino acid residue only. This requires that applicants draft their patent claims with caution, while competitors should always have a look at the patent prosecution history.
Because of the legal uncertainties with respect to the equivalence problem, applicants should restrict the length of the claimed sequence to the bare minimum needed to meet the novelty/inventive step requirement. From that perspective, it is preferable to only specify e.g. one or more CDRs in the claims, instead of a full variable domain, let alone a full heavy or light chain, because for a competitor it is much more difﬁcult to replace one or more amino acids in the CDRs than, say, in the Fc region, at least in case he does not want to compromise the binding behavior of his antibody.