Under 35 U.S.C. § 103, a claim is not patentable if the “differences between the claimed invention and prior art are such that the claimed invention as a whole would have been obvious” to a person of ordinary skill in the art, before the effective filing date of the claimed invention. The U.S. Supreme Court set forth half a century ago a four-prong test to determine obviousness: (i) the scope and content of prior art, (ii) differences between claimed subject matter and prior art, (iii) the level of ordinary skill in the art, and (iv) objective evidence of nonobviousness, such as long-felt but unsolved need, failure of others, commercial success, unexpected results, and skepticism. Graham v. John Deere Co., 383 U.S. 1, 17–18 (1966).
To establish obviousness, a challenger must show that “a [person of ordinary skill in the art] would have had reason to combine the teaching of the prior art references to achieve the claimed invention, and that the [person of ordinary skill in the art]would have had a reasonable expectation of success from doing so.” Par Pharm. Inc. v. TWI Pharms. Inc., 773 F.3d 1186, 1193 (Fed. Cir. 2014). However, the Supreme Court, in addition to reaffirming the obviousness analytical framework set forth in Graham, approved reliance on an “obvious to try” rationale when (1) there was evidence of a design need or market pressure to solve a problem and a finite number of identified, predictable solutions; and (2) a person of ordinary skill in the art would have anticipated success. KSR International Co. v. Teleflex Inc., 550 U.S. 398, 421 (U.S. 2007).
This article analyzes obviousness issues related to antibody patents at the Patent Trials and Appeal Board (PTAB) and in federal courts. We review several cases categorized by type of claims in search of trends on what types of antibody-related claims are more likely to survive an obviousness challenge.
Antibodies Claimed by Structure
An antibody defined by its structure is commonly claimed by its six complementarity determining regions (CDRs) (claimed as a genus), or, more narrowly, by its two variable regions (also claimed as a genus), or even by its heavy and light chain sequences (claimed as species). An exemplary antibody from US Patent No. 7,985,842 is shown below:
An antibody or fragment thereof, comprising a heavy chain VR and a light chain VR, wherein said heavy chain VR comprises heavy chain CDRs comprising the amino acid sequences of SEQ ID NOS: 1, 54 and 3, and wherein said light chain VR comprises light chain CDRs comprising the amino acid sequences of SEQ ID NOS: 4, 5 and 6.
It is relatively straightforward to patent antibodies defined by their structures at the United States Patent and Trademark Office (USPTO), provided a minimal number of amino acid residues for each of the six CDRs is recited in the claim.
In contrast, at the European Patent Office (EPO), a claim of this type is often rejected for lack of inventive step if an antibody to the same target already existed in the prior art, unless the claimed antibody has some unexpected property. The EPO’s position is that, if an antibody to the same target has already been described in the prior art, then a new antibody that is made through routine experimentation and does not show any unexpected properties is simply an alternative without special effect and thus is obvious. The type of properties that can be considered unexpected includes epitope specificity, new or improved biological activity, stability in solution, reduced treatment-related toxicity, and improvement in affinity. Any unexpected beneficial properties, as objective evidence, would also be helpful to establish nonobviousness if such claims are challenged in the United States.
Functional Antibody Claims
Functional claims to a genus of antibodies typically follow the form of “an antibody with functional property X” without reciting any antibody sequence information. A common type of such a claim is “an antibody that competes with antibody X for binding to antigen Y…..” Such a broad claim, covering nearly any antibody in a group of competing antibody products that are directed to the same target, could go a long way towards protecting the patent holder’s market share. However, functional claims may face obviousness challenges if it is considered routine to create many different kinds of antibodies against a certain target and some of the antibodies may be desirable. Consequently, as the industry capability to synthesize more complex and effective antibody therapies improves and the body of prior art increases, such functional claims will become increasingly vulnerable to obviousness challenges. This type of claims may also be vulnerable to attacks for lack of written description, as recently seen in the case of Amgen v. Sanofi, 872 F.3d 1367 (Fed. Cir. 2017).
For example, one of AbbVie’s patents covering briakinumab (then in clinical trials but later withdrawn), and allegedly covering Centocor’s Stelara®, was invalidated for obviousness by the district court. Abbott GMBH v. Centocor Ortho Biotech, Inc., 971 F. Supp.2d 171 (D. Mass. 2013). A representative claim at issue is below:
An isolated human antibody, or antigen-binding portion thereof that binds to human IL-12 and dissociates from human IL-12 with a K4 of 1×10-10 M or less and a koff rate constant of 1×10-3 s-1 or less, as determined by surface plasmon resonance.
In holding the above claim obvious, the court relied on the “obvious to try” rationale and concluded that “there was clear and convincing evidence of a need to create a human, neutralizing, high-affinity antibody to IL-12. A person of ordinary skill in the art at the time knew that the overproduction of IL-12 was causing diseases, and that an antibody that neutralized IL-12 could be therapeutic.” Abbott, 971 F.Supp.2d at 185. “There was also a small number of identifiable solutions being used at the time to create antibodies to human antigens – namely phage display and transgenic mice.” Id. Moreover, there was “market pressure” to create human antibodies, because they were more likely to be successfully marketed. Id. at 186.
The second arm of the “obvious to try” rationale was met because the court found that there was “clear and convincing evidence that a person of ordinary skill in the art at the time could both anticipate success and actually achieve it.” Id. In this regard, the court “place[d] particular emphasis on the functional nature of the claimed result,” concluding that “[i]f any one method of achieving any single embodiment would have caused a person of ordinary skill in the art to anticipate success, that is sufficient to render the invention obvious.” Id. There was evidence that the use of transgenic mice would produce such antibodies. Id. This case was appealed to the Federal Circuit and was decided on other grounds on appeal, without any opinion on the obviousness issue of the claims. Abbott GMBH v. Centocor Ortho Biotech, Inc., 759 F.3d 1285 (Fed. Cir. 2014).
Antibody-Drug Conjugates (ADC)
With only four ADCs in the US market so far (Mylotarg®, Adcetris®, Besponsa®, and Kadcyla®), there has not been sufficient litigation in this field to draw significant trends. However, there have been approximately 60 IND submissions for ADC approval between 2013 and 2017 alone and thus more litigation is expected. The Kadcyla®’s patents have been challenged at the PTAB but, as discussed below, they were not held unpatentable.
The portfolio covering Herceptin® and its ADC Kadcyla® includes more than 100 U.S. patents. Phigenix, a Non-Practicing Entity (NPE), challenged Immunogen’s claim to “[a]n immunoconjugate comprising an anti-ErbB2 antibody conjugated to a maytansinoid, wherein the antibody is huMAb4D5-8” in an IPR against U.S. Pat. 8,337,856 (“’856 patent”). Phigenix, Inc. v. Immunogen, Inc., IPR2014-00676, Paper 39 (P.T.A.B. Oct. 27, 2015). The references asserted in this IPR overlapped with the ones raised in IPR2014-00842, which was denied. IPR2014-00842 is discussed, infra.
The prior art (Chari 1992) taught immunoconjugates comprising an anti-ErbB2 mouse monoclonal antibody, TA.1, chemically coupled to the maytansinoid toxin, DM1, using SPDP or SMCC as a linker. The HERCEPTIN® Label describes the use of huMAB4D5-8 (i.e., HERCEPTIN®) for the treatment of patients with metastatic breast cancer, as well as the combination of HERCEPTIN® with a pharmaceutically acceptable carrier.
Based on these teachings, Phigenix argued that the claimed immunoconjugate would have been obvious because at the time the ’856 patent was filed there was motivation to substitute the mouse monoclonal TA.1 antibody in the immunoconjugate of Chari 1992 with the humanized mAb huMAB4D5-8 to produce the claimed immunoconjugates. According to Phigenix, motivation to modify existed because it was known that (1) humanized antibodies were preferred over their mouse-derived counterparts for clinical applications, (2) huMAb4D5-8 selectively bound to HER2 with high affinity and had been approved to treat breast cancer, and (3) clinical studies had indicated that huMAb4D5-8 worked well in combination with microtubule-directed chemotherapy. Phigenix also presented various arguments in support of a reasonable expectation of success including (1) huMAb4D5-8 was more effective in treating breast cancer in combination with paclitaxel, (2) Chari’s immunoconjugates targeted the same cells as huMAb4D5-8; and (3) an immunoconjugate containing a humanized antibody was less immunogenic and thus more effective in humans than one containing a mouse antibody.
The PTAB instituted the IPR, finding that Phigenix had sufficiently made out a prima face case of obviousness and that the record was not sufficiently complete enough to weigh objective evidence of nonobviousness in determining whether to proceed with a trial. Phigenix, Inc. v. Immunogen, Inc., IPR2014-00676, Paper 11 (P.T.A.B. Oct. 29, 2014).
However, after examining further briefings and evidence from the parties, the PTAB concluded in the Final Written Decision that the instituted claims were not shown by a preponderance of the evidence to be unpatentable. Phigenix, IPR2014-00676, Paper 39, at 21-22 (P.T.A.B. Oct. 27, 2015). The PTAB held that Immunogen provided persuasive evidence that, at the time of the invention, a POSITA would have expected HERCEPTIN®-maytansinoid immunoconjugates to be unacceptably toxic. Id. at 16. In contrast, Phigenix had not established by a preponderance of the evidence that the general statements in Chari 1992 and the HERCEPTIN® Label, in view of the evidence from Immunogen, would have motivated the skilled artisan to substitute TA.1 with HERCEPTIN®. Id. at 21. Immunogen also presented evidence of difficulty and unpredictability when preparing “any antibody-toxin immunoconjugate,” which the PTAB found persuasive and weighed against the existence of a reasonable expectation of success in the claimed invention. Id. In addition, Immunogen’s objective evidence of nonobviousness supported the patentability of certain claims. Id. at 23-26.
An appeal to the Federal Circuit was dismissed after the Federal Circuit held that Phigenix did not have standing under Art. III of the Constitution: “Phigenix has not offered sufficient proof establishing that it has suffered an injury in fact, it lacks standing to bring suit in federal court.” Phigenix Inc. v. Immunogen, 845 F.3d 1168, 1169 (Fed. Cir. 2017). Phigenix “is not engaged in any activity that would give rise to a possible infringement suit.” Id. at 1176, quoting Consumer Watchdog v. Wisconsin Alumni Research Foundation, 753 F.3d 1258, 1262 (Fed. Cir. 2014).
Antibody formulation claims are the second most frequently challenged type of antibody claim (the most frequently challenged are method of treatment claims, discussed below). Source: Finnegan study of 106 antibody-related IPR petitions filed. At least 13 IPRs have been filed against claims to antibody formulations. Among these IPRs, one was instituted, two were dismissed, and 10 were denied (grant rate 9% (1/11)). Id. The large majority of the IPR petitions were denied because the petitioner failed to establish “a reasonable likelihood that the petitioner would prevail.” In the one instituted IPR relating to antibody formulations, all of the claims survived the challenge.
Therefore, formulation claims so far have had a high survival rate both in terms of challenges being denied and surviving a decision on the merits. Although formulation claims may face routine optimization arguments, it is also known that there is high unpredictability in the art. For example, small changes in antibody formulations can have unexpected effects on protein aggregation, viscosity, and other factors, rendering it unpredictable and thus non-obvious to make certain antibody formulations.
For example, Amgen challenged a formulation claim for obviousness in Amgen, Inc. v. AbbVie Biotechnology Ltd., IPR2015-01514:
Claim 1. A stable liquid aqueous pharmaceutical formulation comprising
(a) a human IgG1 anti-human Tumor Necrosis Factor alpha (TNF?) antibody, or an antigen-binding portion thereof, at a concentration of 20 to 150 mg/ml,
(b) a tonicity agent,
(c) a surfactant, and
(d) a buffer system having a pH of 4.0 to 8.0,
wherein the antibody comprises the light chain variable region and the heavy chain variable region of D2E7.
Amgen argued that optimizing different parameters of antibody formulations was generally known in the art. AbbVie argued high unpredictability in formulating stable liquid antibody compositions, especially for subcutaneous administration. AbbVie pointed out that the prior art formulations asserted were either liquid or lyophilized.
The PTAB concluded that Amgen had not provided sufficient showing that a POSITA “would have had a reasonable expectation of success in arriving at the formulation of stable, liquid pharmaceutical compositions comprising antibodies at a concentration of 20 to 150 mg/ml.” Amgen, Inc. v. AbbVie Biotechnology Ltd., IPR2015-01514, Paper 9, at 14 (P.T.A.B. Jan. 14, 2015). No reference showed a liquid antibody product with the claimed concentration range. Id. In addition, references simply showing lists of untested potential targets and possible combinations were not specific enough to be a ground for obviousness. Id. at 18. The PTAB was convinced of the high unpredictability in the art, and denied the IPR petition. Id.
Methods of Treatment
Method of treatment (MOT) claims are by far the most popular target of IPR petitioners. Source: Finnegan study of 106 antibody-related IPR petitions filed. The IPR petition grant rate on MOT claims is 61% (19 petitions granted, 12 denied), Id. generally in line with the USPTO-reported grant rate of 63% (327/520) for bio/pharm petitions. But, in all 6 of the Final Written Decisions to date, all instituted claims were held unpatentable. So far, if the PTAB instituted an IPR on a MOT antibody claim, petitioners are likely successful at challenging those claims.
The discovery that specific subsets of patient populations respond particularly well to a given treatment, which is often the result of post-approval studies, has led to claims to methods of treatment directed to those populations that appear to most benefit from the antibody treatment. This type of claim is expected to become more frequent with the continuous growth of personalized medicine. It appears that those claims may be patentable at least when the applicant can show that, prior to the applicant’s discovery, there was no motivation to select that particular population for treatment.
In another case related to Kadcyla®, Phigenix, Inc. v. Genentech, Inc., IPR2014-00842, Phigenix challenged Genentech’s patent, 7,575,748 (“the ’748 patent”), directed to methods for treating a tumor comprising administering an immunoconjugate, on six obviousness grounds.
Independent claim 1 of the ’748 patent recites a specific antibody-toxin drug conjugate administered to a particular patient population as shown below:
1. A method for the treatment of a tumor in a mammal, comprising the steps of
(i) identifying said tumor as being characterized by overexpression of an ErbB2 receptor and as being a tumor that does not respond, or responds poorly, to treatment with an anti-ErbB antibody, and
(ii) intravenously administering to the mammal a therapeutically effective amount of a conjugate of a humanized antibody huMab 4D5-8 covalently linked via a thioether linking group with a maytansinoid DM1 having the structure…
at a dose of between about 0.2 mg/kg and about 10 mg/kg (antibody-maytansinoid conjugate weight/body weight) and
at a frequency of dosing selected from the group of dosing frequencies consisting of bolus, less than about 1 time per week, one time per week, two times per week, more than two times per week, and continuous infusion,
whereby said tumor characterized by overexpression of an ErbB2 receptor and that does not respond, or responds poorly, to treatment with an anti-ErbB antibody, is treated.
The prior art, which overlapped with that of IPR2014-00676 (U.S. Patent No. 8,337,856) discussed above, was directed to the ADC itself and disclosed immunoconjugates comprising an anti-ErbB2 mouse monoclonal antibody linked to a maytansinoid toxin that showed powerful desirable cytotoxicity. The HERCEPTIN® label mirrors the claim language and describes the use of huMAB4D5-8 (i.e., HERCEPTIN®) for the treatment of a breast cancer in humans who “do not respond, or responds poorly, to treatment with an anti-ErbB antibody.”
The PTAB agreed with Phigenix that “the HERCEPTIN® Label teaches or suggests that certain patients failed to respond to HERCEPTIN®, and that HERCEPTIN® increased the effectiveness of chemotherapy, such as paclitaxel, in patients who responded to HERCEPTIN® and chemotherapy.” Phigenix, Inc. v. Genentech, Inc., IPR2014-00842, Paper 10, at 15 (P.T.A.B. Dec. 9, 2014). But the PTAB held that was insufficient to show that it would have been obvious to a POSITA “whether or how to treat these patients.” Id. at 13 (emphasis in original). The PTAB found no motivation to treat the claimed population with the ADC. In particular, Phigenix did not sufficiently show that a POSITA would understand from the prior art to treat patients, which are unresponsive to HERCEPTIN® alone, with a HERCEPTIN® conjugate. Id. at 16. Nothing in the prior art suggested that patients unresponsive to HERCEPTIN® would respond to the conjugate and thus there was also no reasonable expectation of success in reaching the claimed invention.
The PTAB was also persuaded by Genentech’s evidence that an ordinary artisan “would have thought it likely that HERCEPTIN® resistance occurred because obstacles prevented the antibody from binding to its target or other events that prevented HERCEPTIN® from working as expected.” Id. at 17. Phigenix’s IPR petition was denied. Id.
Conclusions: Patenting Antibodies
This article analyzed a few cases in which antibody-related claims were challenged for obviousness. These cases are highly fact-specific, but a few broad observations may be made. To our knowledge, no claims directed to an antibody claimed purely by structure have been challenged, except for those of an ADC (KADCYLA®), where both the antibody part and the drug part of the conjugate were previously known. An antibody broadly claimed by function may be considered obvious “[i]f any one method of achieving any single embodiment [of such claim] would have caused a person of ordinary skill in the art to anticipate success.” Abbott GMBH, 971 F. Supp.2d at 186. Objective indicia of nonobviousness have been helpful, but only where there was a clear nexus between the objective evidence and the claims.
“Routine optimization” is a typical ground of obviousness from challengers. High unpredictability in the art can be a strong argument for patent owners. General assertions of “routine optimization” are not sufficient to substitute for showing how and why a POSITA would have been motivated to modify or combine the references with a reasonable expectation of success.