News Letter

News Letter

K&P’sIntellectual Property High Court Decision Report in 2018

December, 2018

Updated 8 JUL 2019

1. Does Antibodies Competing for Binding to an Antigen with the Reference Antibody Satisfy the Support Requirement?

Sanofi v. Amgen Inc., Case No. 2017(Gyo-Ke)10225 (Decision rendered on December 27, 2018)
Sanofi v. Amgen Inc., Case Nos 2017(Gyo-Ke)10226 (Decision rendered on October 10, 2018)

The Patentee, Amgen Inc. (Amgen), obtained several patents relating to an isolated monoclonal antibody against PCSK9. Against two of the Amgen's patents (Japanese patent Nos. 5705288 and 5906333), Sanofi filed invalidation trials with the JPO in 2016. The JPO dismissed the Sanofi's demands in 2016. Sanofi filed appeals against each of the JPO's decisions to the IPHC in 2017.

During the invalidation trials, the granted claims of both patents were corrected. Claim 1 as corrected of Japanese patent No. 5705288 is:

1. An isolated monoclonal antibody which is capable of neutralizing binding of PCSK9 to LDLR, and which competes for binding to PCSK9 with an antibody comprising a heavy chain comprising a heavy chain variable region consisting of the amino acid sequences of SEQ ID NO: 49, and a light chain comprising a light chain variable region consisting of the amino acid sequences of SEQ ID NO: 23.

Claim 1 as corrected of Japanese patent No. 5906333 is:

1. An isolated monoclonal antibody which is capable of neutralizing binding of PCSK9 to LDLR, and which competes for binding to PCSK9 with an antibody comprising a heavy chain comprising a heavy chain variable region consisting of the amino acid sequences of SEQ ID NO: 67 and a light chain comprising a light chain variable region consisting of the amino acid sequences of SEQ ID NO: 12.

One of the main issues in these cases is whether an antibody whose structure is not defined with the amino acid sequence satisfies the support and enablement requirements. As shown above, the claims do not recite the CDRs amino acid sequences, let alone the amino acid sequences of the variable regions. The claims recite only the amino acid sequences of the reference antibodies, and the neutralizing activity. Under the Japanese patent practice, the support requirement is examined based on whether or not the claimed invention is within the extent of the disclosure of the specification to which those skilled in the art would understand that the problem to be solved could be actually solved. What is called "function claims" or "reach-through claims", i.e. claims directed to a chemical compound (or the use of that compound) defined only in functional terms with regard to the technical effect, usually tend not to satisfy the support or enablement requirement.

According to the specification, the inventors obtained 2441 positive hybridomas secreting antibodies against PCSK9, and then identified 85 antibodies that blocked interaction between the PCSK9 and the LDLR greater than 90%. Two of the antibodies, 21B12 and 31H4, are the reference antibodies in Japanese patent Nos. 5705288 and 5906333, respectively.

The epitope binning assay was performed on 32 neutralizing antibodies (including 21B12 and 31H4); 19 antibodies belonged to bin 1 (antibodies competing with 21B12, not 31H4), 7 antibodies belonged to bin 3 (antibodies competing with 31H4, not 21B12), one antibody belonged to bin 2 (an antibody competing with both 21B12 and 31H4), and one antibody belonged to bin 4 (an antibody competing with neither 21B12 nor 31H4). When an alternative set (39 antibodies) of binning experiments was conducted, 19 antibodies belonged to bin 1, three antibodies belonged to bin 2, and ten antibodies belonged to bin 3. Fifteen of the 19 antibodies belonging to bin 1 and seven of the 10 antibodies belonging to bin 3 were confirmed to be neutralizing antibodies. The reference antibodies were confirmed to be capable of neutralizing binding of PCSK9 to LDLR, and thereby reducing serum cholesterol level.

The Plaintiff argued as follows:

(1) Regarding the support requirement, the problem to be solved by the present invention is to provide an antibody which is capable of neutralizing binding of PCSK9 to LDLR. The claims can encompass a large number and a large variety of antibodies having a wide variety of amino acid sequences because the claimed antibody is defined only with the function and property (i.e. neutralizing activity and competition with the reference antibody). In contrast, classifying the resultant antibodies according to the amino acid sequences of CDRs, the specification specifically discloses only two or three groups. Furthermore, competition with the reference antibody is not an index of neutralization. Thus, not all the competitive antibodies can solve the problem of the present invention, i.e. provision of a neutralizing antibody.

(2) Regarding the enablement requirement, as discussed above, classifying the resultant antibodies according to the amino acid sequences of CDRs, the specification specifically discloses only two or three groups. The claims can encompass even as-yet-undiscovered antibodies. Those skilled in the art would definitely have to perform undue trial and error experimentation to obtain all of the claimed antibodies.

The IPHC decided that the enablement and support requirements are satisfied. The reasons are as follows:

Those skilled in the art could have expected to obtain antibodies competing with the reference antibody through the binning assay of the other positive antibodies. In addition, in order to obtain neutralizing antibodies competing for binding to PCSK9 with the reference antibody, in accordance with the disclosure of the specification, those skilled in the art could have produced immunized mice and hybridomas, screened for antibodies which strongly block the binding of PCSK9 to LDLR, and conducted the binning assay of the resultant antibodies.

Regarding (1), it is common technical knowledge that the amino acid sequence is to be determined during or after screening of hybridomas and identification of monoclonal antibodies having specific binding activity. Thus, in order to produce an antibody having certain binding activity, it is not essential to determine the amino acid sequence of the antibody in advance.

Those skilled in the art would have understood that the claimed neutralizing antibody competing with the reference antibody could be produced through immunization of mice, screening of hybridomas, identification of antibodies having the binding activity, and the epitope binning assay as disclosed in the specification, without referring to the amino acid sequence of the claimed antibody.

Even though not all the competitive antibodies are capable of neutralizing the binding of PCSK9 to LDLR, the claimed antibody is clearly defined with the phrase "which is capable of neutralizing binding of PCSK9 to LDLR".

Regarding (2), in order to produce an antibody having certain binding activity, it is not essential to determine the amino acid sequence of the antibody in advance. Those skilled in the art could have obtained the claimed neutralizing antibody competing with the reference antibody through immunization of mice, screening of hybridomas, identification of antibodies having the binding activity, and the epitope binning assay as disclosed in the specification.

Conclusively, the IPHC dismissed the Sanofi’s appeals and upheld the JPO’s decisions.

An appeal to the Supreme Court was filed against this decision, and thus the decision is NOT final and binding.

K&P’s Comments These decisions are the first cases to decide that the support and enablement requirements are satisfied for claims reciting the amino acid sequences of the reference antibody for competition assay, but not the amino acid sequences of the claimed antibody. Some cases with such competitive antibody claims are getting to be allowed in Japan. The number of this type of claim formats is predicted to increase.

These decisions have caused controversy in Japan. Some of learned individuals criticize the decisions. A learned individual, however, commented: In order to have antibody claims granted, in many cases, we have been required to recite at least six CDR sequences in the claims. Such competitive antibody claims will be one of the best ways to get broad protection.

In the present cases, the claims clearly recite the specific property that the claimed antibody has, i.e. "capable of neutralizing binding of PCSK9 to LDLR". More importantly, the specification has enough examples, in particular, enough binning assay results, to satisfy the support and enablement requirements. The IPHC, based on these facts, seems to conclude that the competitive antibody claims satisfy the support and enablement requirements.

Amgen filed an injunction demand against Sanofi based on the above two patents with the Tokyo District Court in 2017, and the Court has granted the injunction.
An appeal to the IPHC was filed against this decision, and thus the decision is NOT final and binding.

(by Naomi YOSHIDA, Patent Attorney)

In December 2018, the IPHC handed down 14 decisions including the above case on patent, and overturned the previous decisions in 4 cases.

In December 2018, the IPHC handed down 5 decisions on trademark, and maintained all the previous decisions.

In December 2018, the IPHC handed down no decisions on industrial design.