Pfizer Inc v. Genentech Inc

1. Case Identification

• Case #: IPR2018-00331
• Patent #: 9,249,218
• Filed: December 18, 2017
• Petitioner(s): Pfizer, Inc.
• Patent Owner(s): Genentech, Inc.
• Challenged Claims: 1, 5-7

2. Patent Overview

• Title: PROTEIN PURIFICATION
• Brief Description: The ’218 patent describes therapeutic compositions of the anti-HER2 antibody humMAb4D5-8 (trastuzumab) containing a mixture of the antibody and its acidic variants. The claims require that the total amount of acidic variants be less than about 25%, and that the predominant acidic variants are those deamidated at asparagine residue 30 (Asn30) in the light chain.

3. Grounds for Unpatentability

Ground 1: Anticipation of Claims 1 and 5-7 under 35 U.S.C. §102(b) over Andya

• Prior Art Relied Upon: Andya (WO 1997/04801).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Andya explicitly and inherently disclosed every limitation of the challenged claims. Andya described lyophilized and reconstituted aqueous formulations of the humMAb4D5-8 antibody for therapeutic use. It taught that the major degradation route for this antibody in solution is deamidation, specifically at "30Asn of light chain." Petitioner's analysis of stability data in Andya's Figure 5 showed 82% native protein at time zero, which necessarily means the maximum amount of non-native variants (including acidic variants) was 18%, well below the claimed threshold of "less than about 25%." Because Andya taught deamidation at Asn30 was the "major degradation route" and disclosed no other acidic variants, it was argued to be the "predominant" variant.
  • Key Aspects: Petitioner supported its anticipation argument with an inherency contention. An expert declaration (Dr. Buick) was submitted to show that producing the humMAb4D5-8 antibody according to methods known in the prior art, such as those used to generate the material for Andya's studies, necessarily and inevitably results in a composition with the claimed characteristics, including the specific deamidation at Asn30 as the predominant acidic variant.

Ground 2: Obviousness of Claims 1 and 5-7 under 35 U.S.C. §103 over Waterside and POSA General Knowledge

• Prior Art Relied Upon: Waterside (a 1996 conference presentation) and the general knowledge of a Person of Ordinary Skill in the Art (POSA).

• Core Argument for this Ground:

  • Prior Art Mapping: Petitioner asserted that Waterside disclosed cation-exchange chromatograms for the same antibody (identified as rhuMAb HER2), clearly showing the main antibody peak and several smaller "charged variants." Peaks 1 and 2 were identified as "more acidic" variants. Waterside explicitly identified the variant in peak 1 as containing a deamidation at amino acid position 30 (Asn30 to Asp30). Petitioner calculated from the chromatograms that the total percentage of these acidic variants was between 9.5% and 16.4%, which is less than the claimed 25% limit. Because peak 1 (the Asn30 deamidated variant) was the largest acidic variant peak, it was argued to be the predominant one.
  • Motivation to Combine: A POSA would be motivated to prepare a composition with low levels of acidic variants as taught by Waterside. Waterside disclosed that the deamidated Asn30 variant (peak 1) retained only 82% of the specific activity of the parent antibody. This loss of activity would have motivated a POSA to reduce the amount of this variant in a final therapeutic product to ensure maximum potency and product consistency, a common goal in biopharmaceutical development.
  • Expectation of Success: A POSA would have had a high expectation of success in achieving the claimed composition. Waterside itself demonstrated the use of cation-exchange chromatography (CEX) to separate these variants. CEX was the well-known method of choice for separating proteins based on charge differences, and a POSA could have readily used or optimized standard CEX techniques to reduce acidic variants to the levels disclosed in Waterside and claimed in the patent.

• Additional Grounds: Petitioner asserted additional obviousness challenges against claims 1 and 5-7 based on Andya in view of POSA general knowledge, and Harris (a 1995 journal article) in view of POSA general knowledge. These grounds relied on similar arguments that the references disclosed the claimed antibody with acidic variants below 25% and identified Asn30 deamidation, and that a POSA would have been motivated to reduce such variants to improve product quality.

4. Key Claim Construction Positions

• "therapeutic composition": Petitioner argued this term should be construed to mean "a composition appropriate for administration in a therapeutic treatment regimen." It contended the term does not require a specific dosage amount, a particular manufacturing process or scale, or proven therapeutic effectiveness, as these limitations are not in the claim language.
• "humMAb4D5-8": Petitioner asserted this term should be construed as synonymous with terms used in the prior art, such as "rhuMAb HER2" and "humAb4D5-8." This position was based on explicit statements in the ’218 patent specification acknowledging these terms refer to the same antibody, as well as binding admissions made by the patent owner during the prosecution of a parent application.

5. Relief Requested

• Petitioner requested institution of an inter partes review and cancellation of claims 1 and 5-7 of the ’218 patent as unpatentable.