Scantibodies Laboratory, Inc. v. Immutopics, Inc. (Fed. Cir. 2010)

    By Andrew Williams --

Last week, in Scantibodies Laboratory, Inc. v. Immutopics, Inc., the Federal Circuit affirmed a claim construction decision and corresponding summary judgment of non-infringement rendered by the U.S. District Court for the Central District of California.  The patent at issue was U.S. Patent No. 6,689,566 ("the '566 patent"), "Methods, Kits, and Antibodies for Detecting Parathyroid Hormone."  Claim 5, the only claim addressed on appeal, reads (with the claim terms at issue underlined):

5.  A method for measuring an amount of whole parathyroid hormone in a sample comprising:

    a)  adding to a sample a labeled antibody or antibody fragment specific for an initial peptide sequence of whole parathyroid hormone wherein said initial peptide sequence consists of VAL-SER-GLU-ILE-GLN-LEU-MET (SEQ ID NO:3), and wherein at least four amino acids in said initial peptide sequence are part of a reactive portion to said labeled antibody;

    b)  allowing said labeled antibody to bind to whole parathyroid hormone present, thereby forming a complex; and

    c)  measuring the amount of said labeled complex to measure the amount of whole parathyroid hormone in said sample while not detecting an interfering non-(1-84) parathyroid hormone fragment.

The District Court construed these terms to mean:

• "specific for" = having a measurable affinity for and detectable binding to an epitope having at least four amine acids of the seven in SEQ ID No. 3 (VAL-SER-GLU-ILE-GLN-LEU-MET)

• "not detecting an interfering non-(1-84) parathyroid hormone fragment" = having no detectable binding to an interfering non-(1-84) parathyroid hormone fragment

Because Scantibodies conceded that the accused Immutopics antibody did not meet the later limitation, the District Court granted summary judgment of non-infringement.

The application that issued as the '566 patent was filed January 14, 1999.  The application described antibodies, methods of use, and kits for detecting whole or non-fragmented parathyroid hormone ("w-PTH") in a biological sample.  Measuring serum levels of PTH in patients is important for a large number of diseases, including familial hypocalciuric, osteoporosis, and Paget's bone disease.  However, measuring biologically active levels of PTH in humans has been challenging.  One problem is that PTH circulates at extremely low levels.  In addition, these circulating levels of protein are heterogeneous due to the large number of PTH fragments, which leads to interference when attempting to measure the biologically active form ("the (1-84) PTH fragment").  In order to overcome this fragment interference, two-site immunoradiometric assays had previously been developed, which allow for the measurement of intact PTH ("I-PTH").  A measurement of I-PTH, however, includes w-PTH and a large PTH fragment cleaved at amino acids 5 to 8.  In fact, patients with hyperparathyroid or renal failure have significant concentrations of these large, non-whole PTH fragments, which can give inaccurate measurements of the biologically active form.  To overcome this problem, the Scantibodies scientists discovered methods for detecting w-PTH in a biological sample while not detecting this non-(1-84) large PTH fragment component of I-PTH.  They also developed antibodies for use with such methods.  The Scantibodies method used an antibody "specific for" at least four amino acids of seven amino acids disclosed from the N-terminal end of the protein.  Use of such an antibody was important, because it would not react with the interfering non-(1-84) parathyroid hormone fragment.

The originally filed claims in the application that lead to the '566 patent did not contain the limitation "not detecting an interfering non-(1-84) parathyroid hormone," even though the necessity of not detecting this fragment was described in the specification.  During prosecution, however, the examiner cited to several pieces of art in both novelty (§ 102) and obviousness (§ 103) rejections.  The most significant art described the methodology of the measurement of I-PTH.  In order to obtain allowable claims, therefore, the applicant amended the claims to include the "not detecting" limitation, because apparently all of the prior art methods couldn't differentiate w-PTH from this interfering non-(1-84) parathyroid hormone.

Scantibodies filed the present lawsuit on October 26, 2004.  Subsequently, both parties filed reexamination requests, which eventually resulted in a reexamination certificate that altered the antibody and kit claims, but apparently did not change the method of use claims.  This may be the reason that the parties focused on claim 5 during the appeal.  The District Court determined that it needed to construe the claims before it could issue a non-infringement summary judgment order.  And, after two claim construction orders, the Court did just that.  Scantibodies appealed.

The Federal Circuit recognized that the District Court's construction of "not detecting" to mean "no detectable binding" created a difficult, although not insurmountable, hurdle for finding infringement.  Apparently, Scantibodies argued that such a definition would exclude all antibodies, because it was impossible not to have some cross-reactivity.  However, the specification of the '566 patent apparently touted the fact that no cross-reactivity could be achieved.  Moreover, the Federal Circuit pointed out that this construction was based on another limitation that Scantibodies did not challenge ("does not specifically bind to an interfering non-(1-84) parathyroid hormone fragment").  The Court did not give much weight to the testimony of one of the patentees, nor did the Court find it relevant that Scantibodies' product literature defined "no cross-reactivity" as "no significant cross-reactivity."  Scantibodies argued that no PTH assay can absolutely detect PTH without detectable cross-reactivity.  However, the Court pointed out a 2001 reference from Gao et al., which included the inventors as authors, disclosing that an N-terminal PTH antibody that bound to the first few amino acids, and did not have detect the (7-84) PTH fragment at concentrations of 10,000 pg/ml.  The Federal Circuit used this as proof that it was possible to meet the limitations of the claim as construed by the District Court.

Finally, the Federal Circuit noted that the error should be assigned to the patent drafters, because if they had wanted "not detecting" to mean something else, they could have either used different claim language, or provided a different definition in the specification.  This analysis ignores the reality that this claim limitation was required to overcome prior art cited during prosecution.  Moreover, it ignores the fact that the specification highlighted the criticality of not detecting an interfering non-(1-84) parathyroid hormone fragment.  In fact, it is possible that if the claims had issued without the "not detecting" limitation, Scantibodies could have had a written description problem.  Therefore, the Federal Circuit is correct in pointing out that claim drafters need to be careful in choosing the language of claim limitations, but claim drafters also need to mindful of the invention that is actually disclosed in the specification.

Scantibodies Laboratory, Inc. v. Immutopics, Inc. (Fed. Cir. 2010)
Nonprecedential disposition
Panel: Chief Judge Michel and Circuit Judges Plager and Moore
Opinion by Chief Judge Michel

Image of parathyroid hormone (above) by Emw, from the Wikipedia Commons under the Creative Commons license.

Comments

[James Demers]

It's impossible to overstate the value of asking yourself, "How could I design around this claim?" when drafting or amending a claim. It's also impossible, sometimes, to get it right under the constraints of amendment and reexam practice. In this case, "not detecting" was an excessively narrowing limitation -- but the two-page description of the invention provided no more moderate language.

[Andrew Williams]

Dear James,

I completely agree with you. Drafting patent applications and claims requires one to walk a fine-line between adequately describing the invention on one hand, including providing the most expansive list of alternatives as the invention will permit in order to avoid both as-of-yet unknown prior art and future design arounds, while on the other hand avoiding excessively long specifications. It is no wonder that application drafting is one of the most challenging aspects of intellectual property law, and possibly fraught with the most danger. As you indicate, "not detecting" was excessively narrowing, but I am not sure from reading the patent that the invention itself would have tolerated a more moderate alternative. Thank you for your comment.

Andrew