MYLAN’S HERCEPTIN IPR (BIOSIMILAR INTER PARTES REVIEW)

Mylan Pharmaceuticals Inc. (Mylan) has had a busy year filing IPRs, including two IPR petitions (IPR2016-01694 and IPR2016-01693) against Genentech Inc.’s patent (U.S. Patent 6407213 (‘213 Patent)) that covers the drug Herceptin, which is a humanized mouse monoclonal antibody.  Mouse monoclonal antibodies can possess a desirable therapeutic effect, but patients receiving mouse antibodies experience a human anti-mouse antibody (HAMA) immunogenicity response.  To neutralize the HAMA response, the constant region and the variable framework regions flanking the Complementarity Determining Regions (CDRs) are replaced with human sequences, leaving only the CDRs of the mouse.  This extensive humanization, however, has the drawback of changing the conformation of the mouse CDRs, making the antibody less effective.  To maintain the effectiveness of the antibody, Genentech’s patent claims substituting some of the amino acids from the human flanking regions back to the mouse sequence. 

Genentech’s ‘213 Patent claims a “humanized antibody” with amino acid substitution in the Framework Regions (FR) flanking the CDRs:

1. A humanized antibody variable domain comprising non-human Complementarity Determining Region (CDR) amino acid residues which bind an antigen incorporated into a human antibody variable domain, and further comprising a Framework Region (FR) amino acid substitution at a site selected from the group consisting of: 4L, 38L, 43L, 44L, 58L, 62L, 65L, 66L, 67L, 68L, 69L, 73L, 85L, 98L, 2H, 4H, 36H, 39H, 43H, 45H, 69H, 70H, 74H, and 92H, utilizing the numbering system set forth in Kabat. 

30. An antibody which binds p185.sup.HER2 and comprises a humanized antibody variable domain, wherein the humanized antibody variable domain comprises non-human Complementarity Determining Region (CDR) amino acid residues which bind p185.sup.HER2 incorporated into a human antibody variable domain, and further comprises a Framework Region (FR) amino acid substitution at a site selected from the group consisting of: 4L, 38L, 43L, 44L, 46L, 58L, 62L, 65L, 66L, 67L, 68L, 69L, 73L, 85L, 98L, 2H, 4H, 36H, 39H, 43H, 45H, 69H, 70H, 74H, 75H, 76H, 78H and 92H, utilizing the numbering system set forth in Kabat

60. The antibody of claim 30 wherein the residue at site 78H has been substituted.

Claims 30 and 60 specify that the antibody is an antibody that binds to p185^HER2, which is the target for Genentech’s Herceptin.  These claims also have a substitution at 78H (absent in claim 1), which Mylan admits is present in Herceptin.

Mylan argues that claim 1 of the patent is anticipated by Kurrle (EP0403156), since “Kurle substituted several corresponding murine amino acids for human framework residues under Kabat’s numbering system, including 4L and 69H, as found in claim 1.”  Mylan further argues that claim 1 is anticipated since “one of ordinary skill in the art at the time of the ’213 patent . . . would have readily understood that Queen 1990 [WO199007861] explicitly taught the substitution of framework sites immediately adjacent to CDRs… Using the numbering system set forth by Kabat 1987, the ‘immediately adjacent’ framework residues to CDRs as taught by Queen 1990 and recited in claim 1 include 98L and 36H.”

Kurle and Queen 1990 do not seem to make the substitution at position 78H, which is present in Herceptin, forcing Mylan to rely on an obviousness argument for claims limited to the substitution at position 78H: “Given published accounts regarding other monoclonal antibody humanization efforts, and the strength of 4D5 as a clinical target, the logical and necessary next step would have been to humanize 4D5.. . .  Following the detailed roadmap of Queen 1990, a POSITA would have looked to Chothia & Lesk and identified FR positions that could interact with or influence CDR conformation, and antigen binding and specificity, including residue 78H.” Mylan’s IPR petitions further state that the prior art taught a number of substitutions, but the list of the positions in the prior art that Mylan provides do not include position 78H: “From Queen 1989 or Queen 1990, together with known antibody structures available in the PDB database, a POSITA would have recognized that claimed framework positions 4L, 58L, 66L, 67L, 73L, 98L, 2H, 36H, 45H and 69H were readily identifiable as residues that: 1) are adjacent to CDRs; or 2) contact CDRs.”  The issue to be decided by the PTAB (Patent Trial and Appeal Board) is whether a general teaching in the prior art to humanize an antibody will be sufficient to make obvious substitutions at specific positions not taught in the prior art.

Biosimilar IPR petitions also need to address any secondary considerations since failing to do so can result in the denial of the petition.  IPR2015-01792, Paper No. 14, at 18.  Mylan argued that any secondary considerations do not preclude obviousness: “First, none of the heavy chain residues cited in claim 1 are even modified in Herceptin, and only 1 of the 13 heavy chain residues (78H) cited in claims 30, 62 and 63 is modified in Herceptin. . . . Even claim 30, which recites that the antibody binds p185^HER2 , is exceptionally broad, not being limited to any specific anti-p185^HER2 antibody.”  

The PTAB has not yet ruled on the petition.