2
CONSTRUING THE PATENT CLAIM
Words matter. Words in a patent claim matter more.
At the heart of every patent infringement suit is the question of whether a patent claim encompasses an allegedly infringing activity. Equally important is the question of whether such a claim, even if infringed, is nevertheless invalid.
Answering these questions requires first determining what the patent claim actually encompasses. Only then can a court determine whether the accused conduct falls within its scope. Determining what a claim encompasses must also precede determining whether it is invalid; for example, whether previously published information destroys its novelty or renders it obvious. Chapters 9 and 10 cover patent infringement and invalidity defenses, respectively.
To determine what a patent claim covers, a court must construe it. This process is known as claim construction. Claim construction is the province of judges and attorneys. This sets claim construction apart from other areas of patent law, such as patent prosecution (covered in chapter 5). Patentees, inventors, corporate officers, investors, and other non-attorneys may have a vested interest in a given patent and may play an active role in obtaining it. However, they typically have little, if any, involvement in the claim construction process.
Having said that, this judicial process is substantive. And its outcome is often dispositive of a patent infringement suit and, hence, determinative of the asserted patent’s commercial worth. It is for these reasons that this chapter briefly presents the topic of patent claim construction.
Construing a claim is an exercise in determining what its language means. It is an evidence-based and imperfect process of understanding the meaning of each claim term in the context of the claim as a whole. In the United States, federal courts accomplish this task through what is known as a Markman hearing, named after the seminal 1996 U.S. Supreme Court decision in Markman v. Westview.
In a Markman hearing, a court may consider the claim’s grammar or punctuation (e.g., the significance of a comma or semicolon), the words used in the claim (e.g., the effect of preceding a certain noun with “a” instead of “the”), and other claims in the patent, all with respect to their effect on the possible meanings of the claim being construed. Another key focus of judicial inquiry is the patent’s specification (i.e., the rest of the patent). Of particular importance are definitions, examples, and other explanatory language in the specification that sheds light on a claim’s scope.
Beyond considering the four corners of a patent document itself, a court can also consider its prosecution history, which includes all written correspondence exchanged between the patentee and the Patent Office in order to obtain the patent. A patent’s prosecution history can provide a wealth of information about what a claim encompasses. It can also reveal what a claim does not encompass. Indeed, as we will see in chapters 5 and 9, patent prosecution often involves a patent applicant’s establishing, in considerable detail, how a proposed claim does not encompass that which was already known before the patent application was filed. As a related matter, the prosecution histories of kindred U.S. and counterpart foreign patents can also inform a court’s claim construction.
Ideally, this intrinsic evidence is sufficient for a court to construe a patent’s claims. When it is not, extrinsic evidence, such as dictionary definitions, scientific publications, and expert testimony, may also come into play. The rules governing claim construction—particularly which types of evidence may or must be considered and under what circumstances—are evolving and vary from court to court.
Suffice it to say that virtually any claim term, and hence virtually any claim, may be construed in surprisingly diverse ways. This reality adds to the unpredictability of litigation and the unpredictability of ascribing a particular value to a patent.
The following example underscores the importance and unpredictability of claim construction.
EXAMPLE 2.1
Company X owns a U.S. patent to a certain composition of matter. Claim 1 of the patent provides “a composition comprising (i) a pharmaceutically acceptable carrier, and (ii) an antibody that specifically binds to an epitope comprising the amino acid sequence his-arg-gly-gly-pro, wherein the antibody is labeled with a toxic moiety.”
Company Y sells an allegedly infringing product in the United States. This product is a cancer imaging agent containing a humanized Fab fragment that binds to X-his-arg-gly-gly-pro-X (wherein X = 20 amino acid residues) at an affinity of 1 × 10−9 M, binds to X-his-arg-gly-val-pro-X at an affinity of 2 × 10−9 M, and contains a radioactive atom permitting diagnostic imaging at a safe dose.
Does Company Y’s product infringe claim 1 of Company X’s patent? That is, does claim 1 cover the product by virtue of the product’s falling within the claim’s scope?
To answer this question, a court must first construe the claim by determining the meaning of each relevant claim term. In litigation, determining which terms to construe is based on the facts specific to the claim in question. For the purpose of this example, however, we will assume that these terms include antibody, specifically binds to, epitope comprising, and toxic moiety.
Before discussing these terms individually, there is something worth noting. Each term, by itself and in the abstract, is basic to anyone having even a modest level of training in biotechnology. At first blush, these terms would appear to need little, if any, explanation were they to appear in a scientific publication or textbook. When they appear in a patent claim, however, this is seldom the case.
Our first claim term is antibody. What does antibody mean as it is used in this claim? There are many possibilities depending on the relevant facts. For example, must the antibody be of a particular type, such as IgG or IgM? Must the antibody be intact, in that it contains all of the chains and regions that a naturally occurring antibody of that type would contain? Or, is the antibody merely a fragment of an intact antibody? If so, must that fragment possess certain functions like the ability to bind an antigen (e.g., a Fab fragment) or immune cell (e.g., an Fc fragment)? Must, or may, the antibody possess a special physical feature, such as being linear, chimeric, or humanized?
What does the term specifically binds to mean, with respect to both the antibody that the term modifies and the epitope to which the antibody binds? Without knowing more about the rest of the patent or any other available evidence, at least three possible interpretations of this term exist. The first is that the antibody must bind to the designated epitope more strongly than to any other epitope. The second is that the antibody must bind to the designated epitope, but not to any other epitope. The third is that the antibody must merely bind to the designated epitope more strongly than it binds nonspecifically to another surface, such as the side of a Petri dish.
What does epitope comprising mean with respect to the five-residue amino acid chain recited in claim 1? For instance, an epitope is generally understood to be the part of an antigen recognized by a particular antibody. Absent additional clarity from the patent document or other permissible evidence, it is still unclear what an epitope comprising the five-residue sequence is. Must the antibody come into direct contact with all five residues? Must it come into direct contact with at least one residue? Or, can the epitope comprise the five-residue sequence without the antibody’s directly contacting any of the residues, as might be the case, for example, when the five residues are part of, and help stabilize, a twenty-residue epitope having other residues that directly contact the antibody?
Finally, the term toxic moiety must be construed in this example, despite its apparent simplicity. The many relevant inquiries might best begin with determining what toxic means. That is, to what must the moiety be harmful? Must it harm a patient, a tumor, an individual cell, an organelle, or DNA? At what dose must the moiety be harmful? To what degree must the moiety be harmful? Must it kill a cell, or must it merely inhibit a biological function? If the latter is true, must it permanently inhibit the biological function or merely do so transiently?
Once again, in this example, a court must construe claim 1 of Company X’s patent to determine whether Company Y’s product infringes it. If the court construes each of the four terms in Company X’s favor, infringement can be found, and Company X can prevail. If not, the court cannot rule that Company Y’s product infringes claim 1, and Company Y will prevail.
More specifically, a court would have to construe the four terms as follows in order for Company X to prevail.
First, the court would have to construe the term antibody as including at least a humanized Fab fragment. Second, the court would have to construe the phrase specifically binds, with respect to an epitope comprising his-arg-gly-gly-pro, as including at least binding to X-his-arg-gly-gly-pro-X at 1 × 10−9 M and to X-his-arg-gly-val-pro-X at 2 × 10−9 M. Third, the court would have to construe the phrase epitope comprising, with respect to his-arg-gly-gly-pro, as including at least the X-his-arg-gly-gly-pro-X sequence recognized by the Fab. Finally, the court would have to construe the term toxic moiety as including at least a radioactive atom permitting imaging at a safe dose.
If any of the four claim terms were not construed in this way, the result would be a finding of non-infringement in favor of Company Y. So, for example, Company Y would prevail if the court were to construe antibody as an intact antibody, but not an antibody fragment. Chapter 9 discusses patent infringement in more detail.
FIGURE 2.1 These four Venn diagrams illustrate selected claim term constructions described in example 2.1. The upper-left diagram shows how Company Y’s product (Y) satisfies the claim term antibody when that term is broadly construed to include at least a humanized Fab fragment. The upper-right diagram shows how the product fails to satisfy that term when it is narrowly construed to include an intact antibody but not a Fab fragment. The lower-left diagram shows how the product satisfies the claim term toxic moiety when that term is broadly construed to include at least a radioactive atom permitting imaging at a safe dose. The lower-right diagram shows how the product fails to satisfy that term when it is narrowly construed to exclude the radioactive atom at that dose. If a court were to narrowly construe even one of claim 1’s four terms discussed in example 2.1, the product would not satisfy all of the claim’s elements. In that case, the product would not infringe claim 1.