Are Arbutus’ patents indefinite?
Zachary Silbersher
I’ve previously written about Arbutus’ patent disputes with Moderna and Pfizer. In Arbutus’ case against Pfizer, the court recently conducting a claim construction hearing. Before that hearing, Pfizer argued that many of Arbutus’ patents are “indefinite.” The court held it was premature for Pfizer to make this argument, but Pfizer nonetheless established that when the case reaches summary judgment, it will rely upon the doctrine of indefiniteness to prevail against Arbutus. What is indefiniteness in patent law? Are Arbutus’ patents indefinite? How viable is Pfizer’s “indefiniteness” defense?
Arbutus has claimed that Pfizer’s Covid vaccine infringes five of Arbutus’ patents. The five patents can be split into two groups. The first group includes U.S. Patent Nos. 9,504,651 (“the ’651 patent”); 11,298,320 (“the ’320 patent”); and 11,318,098 (“the ’098 patent). The second group includes U.S. Patent Nos. 8,492,359 (“the ’359 patent”) and 11,141,378 (“the ’378 patent”).
Very generally, the patents claim lipid vesicles, or apparatuses for producing lipid vesicles, that include four ingredients: a cationic lipid, phospholipid, cholesterol and a PEG-lipid. In many instances, the patents also require specific molar percentage ranges for different ingredients, e.g., “wherein the phospholipid consists of from 3 mol % to 15 mol % of the total lipid present in the particle.” (‘378 patent). In other instances, the patents require that at least a certain percentage (such as 70% or 80%) of the mRNA in the formulation is “fully encapsulated in the lipid vesicles”.
Pfizer has argued that specific terms used in the patents are “indefinite.” This is another way of arguing that the patents themselves are “indefinite.” If Pfizer can prove that the patents are indefinite, then that would invalidate the patents. If Pfizer invalidates the patents, then it will avoid all liability to Arbutus in this suit.
In almost every patent lawsuit, the accused infringer (in this case, Pfizer) typically argues, as a defense to liability, that the asserted patents are invalid. Typically, that means arguing that the purported inventions in the patents are not novel, but rather obvious. That normally requires showing that the purported inventions in the patents were already taught in the prior art, or scientific literature preceding the patent’s filing date.
Yet, obviousness is not the only way to invalidate a patent. Patents can also be invalidated for reasons that have nothing to do with the prior art. This includes a lack of written description, lack of enablement, improper inventorship, inequitable conduct and indefiniteness, among others. Compared to other grounds for proving invalidity, indefiniteness is relatively obscure, and it doesn’t occur nearly as frequently. Arbutus previously defeated numerous obviousness challenges brought by Moderna in a series of IPRs, and that has crippled both Moderna’s as well as Pfizer’s ability to defend against Arbutus with an obviousness defense to the patents. (This may partially explain Pfizer’s resort to indefiniteness in this case.)
But what is indefiniteness? The Supreme Court has held that a patent is indefinite if it “fail[s] to inform, with reasonable certainty, those skilled in the art about the scope of the invention.” Nautilus, Inc. v. Biosig Instruments, Inc., 572 U.S. 898, 912 (2014). Another way to think about indefiniteness is from the perspective of a company interested in designing around a patent. If your competitor has a patent, and you wish to design your own product that avoids infringement of your competitor’s patent, then you need to be able to locate with precision the line between infringement and non-infringement. However, if the patent doesn’t let you do that, then it’s indefinite.
Indefiniteness typically arises when a patent requires a certain measurement occur within a specific range or above or below a certain threshold. Anyone wishing to design around that patent will need to know how to take that measurement. That does not mean that any patent requiring a measurement is necessarily indefinite just because the patent does not specify how to take the measurement. Rather, indefiniteness will typically exist where multiple measurement techniques exist in the art, some yielding infringement, while others yielding non-infringement, but the patent does not teach which measurement technique to use. In that case, any seeking to design around the patent will be stifled from discerning whether they will or will not infringe.
For instance, in one case, a claim requiring a specific “molecular weight of about 5 to 9 kilodaltons” was indefinite because among three different measurement techniques (Mp, Mn, Mw), each yielded a different molecular weight, some infringing, some not, and the patent was not clear about which technique to use. See Teva Pharm. USA, Inc. v. Sandoz, Inc., 723 F.3d 1363 (Fed. Cir. 2013). In another case, a patent claim required a specific range of “melting point elevation” (MPE) for a prepared “polyethylene terephthalate” (PTE), but among the methods available to prepare PTE (coil method, cut method, restrained method, ball method), each yielded different MPE results, some within the claimed range, others not, and the patent did not teach which PTE preparation method to use. See Honeywell Intern., Inc. v. Int'l Trade, 341 F.3d 1332 (Fed. Cir. 2003).
In Arbutus’ case, Pfizer has argued that at least the following claim terms are indefinite: “cationic lipid,” “fully encapsulated” (‘651 patent, ‘359 patent, ‘378 patent), and “consisting essentially of” (‘378 patent). The term “cationic lipid” is recited in all five asserted patents, and technically could alone doom Arbutus’ case against Pfizer if its indefiniteness defense is successful. The other terms would only doom the specific patents in which those terms are recited.
The parties have agreed on the construction of “cationic lipid” to be “any of a number of lipid species which carry a net positive charge at a selected pH.” Pfizer has argued that this term is indefinite “because a person of ordinary skill would not have reasonable certainty over what constitutes a ‘selected pH.’” Case No. 23-01876, Dkt. 66-1 at 3 n.2 (D. Del. Apr. 25, 2024). The ‘651 patent includes a definition of the term “cationic lipid,” which reads: “The term ‘cationic lipid’ refers to any of a number of lipid species which carry a net positive charge at a selective pH, such as physiological pH.” (‘651 patent col. 5:4-6). The patent then provides several examples of “cationic lipids,” such as DODAC, DOTMA, DDAB, DOTAP, DC-Chol, among many others.
Pfizer has yet to brief this issue for the court, and we cannot know for certain how Pfizer will frame their argument that the term, “cationic lipid,” is indefinite. That said, the argument will presumably go something like this: a cationic lipid is something that carries a net positive charge “at a selective pH.” That suggests that, at one pH, it may have a net positive charge, but at a different pH, it may not. The patent’s specification, however, provides no guidance on which pH should be “selected.” For this reason, a company wishing to design around Arbutus’ patents would not know how to do so because it will not know which pH to “select.”
With respect to the “fully encapsulated” term, Pfizer is likely to argue that this requires a measurement, but the patents are silent on which measurement technique to use. Pfizer has already explained to the court when making mRNA therapeutic products, such as vaccines, with lipid vesicles, the process is not precise. Rather, “[s]ome nucleic acids remain free-floating within the mixture of particles, some stick to the exterior of a particle, some may protrude from particles, and some may be entirely within the particle.” There exist methods for measuring the extent to which the nucleic acids are encapsulated within the vesicle. A measure of this is sometimes referred to as encapsulation efficiency. See Case No. 23-cv-01876, Dkt. 83 at 5 (D.N.J).
According to Pfizer, encapsulation itself could refer to different things: “among other things, the measurement of the percentage of the starting nucleic acids that are encapsulated, the measurement of the amount of nucleic acids in the final mixture that are inside a particle versus outside (either before or after unencapsulated nucleic acids are removed), or the ratio of lipids to nucleic acid in the final mixture.” Pfizer also claims that there exist different measurement techniques that can be used to measure encapsulation efficiency, such as chromatography, nucleic acid degradation, or fluorescent dyes. See Case No. 23-cv-01876, Dkt. 83 at 5 (D.N.J).
Again, Pfizer has yet to brief its indefiniteness arguments, but its prior statements to the court already reveal where its argument is likely heading. Pfizer will likely claim that each of the techniques for measuring encapsulation efficiency may yield different results. And because the patents specific ranges that must be “fully encapsulated,” one technique may yield infringement, whereas another may not. If that’s all true, then as discussed above, that’s the typical hallmark of demonstrating a patent is invalid for being indefinite.
On the other hand, Pfizer may not be able to prove all of the facts necessary to show indefiniteness. For instance, in the case of “fully encapsulated,” are there really multiple techniques for measuring encapsulation efficiency? Is one technique regarded as the gold standard whereas others have been disavowed by the literature? Does the patent really not teach that one specific technique should be used? These are factual questions, which will most certainly require expert opinion. And just like Pfizer, Arbutus will certainly marshal as much evidence as possible to answer these questions in its favor.
Importantly, there is some overlap between the patents asserted by Arbutus against both Pfizer and Moderna. In advance of the claim construction hearing in the Moderna case, Moderna did not stake out as much territory over indefiniteness as Pfizer has done. Yet, that will not necessarily preclude Moderna from making similar indefiniteness arguments. And to the extent either party prevails on indefiniteness, that holding may potentially read through to the other case.