Case Nos. 2013-1047+ (Prost, O’Malley, Taranto)
Avoiding specification-bound claim constructions (through careful claim drafting or luck of the draw on judges) may just set one up for a different failure. As illustrated by Synthes v. Spinal Kinetics, achieving a plain meaning claim construction that exceeds disclosed examples may still trigger invalidation for lack of written description, even in a mechanical setting. The case thus provides general lessons for complying with § 112(a) and specific cautions when trying to capture competitive scope with broadening amendments in prosecution.
Synthes owned US Patent 7,429,270 directed to implants that could replace intervertebral discs, the cartilaginous shock absorbers between vertebrae. Synthes asserted the ‘270 patent against Spinal Kinetics (called “SK” in the opinion) for manufacturing M6-C and M6-L discs, alleging they infringed the following claim added during prosecution of the ‘270 patent:
29. An intervertebral implant . . . comprising:
a first substantially rigid bone contacting plate. . .;
a second substantially rigid bone contacting plate . . .;
a third plate operatively coupled to the first bone contacting plate, the third plate including a plurality of openings;
a fourth plate operatively coupled to the second bone contacting plate, the fourth plate including a plurality of openings;
a central part . . . between the third and fourth plates . . . including a flexible core and a fiber system, wherein . . . the fiber system at least partially surrounds the core, and is at least partially received within the plurality of openings formed in the third and fourth plates so that the fiber system is joined to the third and fourth plates; and
an elastic sheathing body at least partially surrounding the fiber system and the core, and connected to the third and fourth plates.
The above-bolded openings in the plates were key in this case. Synthes argued for a plain language construction of the opening terms that included any openings, over SK’s arguments for a narrow construction that limited the term to the specification. The broader interpretation ensured infringement by SK’s devices (shown left) that threaded their core fibers through circuitous openings entirely within the top and bottom plates. The district court gave Synthes their requested plain language construction.
The problem, however, was that the ‘270 patent disclosed only one type of “opening,” and it wasn’t enclosed channels like in SK’s plates. Instead, the ‘270 patent disclosed only grooves – open-faced indentations – across diameters of its sandwiching plates (shown below right). The core fibers extended around the plates in the grooves without being fully surrounded by the plate. The jury found, and the district court upheld, that the post-filing addition of “openings” in claim 29 lacked written description when construed to cover any opening, including SK’s slots. Synthes appealed.
A majority of the Federal Circuit affirmed. Written description is a question of fact, and a jury verdict on such an issue must be upheld if supported by substantial evidence that reached the clear and convincing threshold. Turning to the principles of written description in particular, the majority initially noted that there was nothing per se wrong about broadening claims in prosecution:
While broadening claims during prosecution to capture a competitor’s products is not improper, the written description must support the broadened claims. See Liebel-Flarsheim Co. v. Medrad, Inc., 358 F.3d 898, 909 n.2 (Fed. Cir. 2004) (“[I]t is not improper for an applicant to broaden his claims during prosecution in order to encompass a competitor’s products, as long as the disclosure supports the broadened claims.”)
The majority then went through the evidence, which included the content of the ‘270 patent and expert testimony on whether the skilled person would have understood the applicants to have possession of intervertebral implants with any type of opening for core fiber attachment like in claim 29 as construed. The drawings and written description did state that fibers could be attached to the plates by “various means” and explicitly treated grooves as merely an “example” of these means. Conversely, the ‘270 disclosed no other examples besides grooves, and the majority noted the constant refrain of “grooves” in the written description of implant structure and form.
With “grooves” as the only disclosed species of openings, the majority looked at the expert testimony to determine if this failed to support a genus claim to all openings. Strikingly, much of the expert testimony and its analysis looked almost like an enablement discussion vis-à-vis the allegedly infringing device. For example, in addition to testifying that openings anywhere were not suggested by the ‘270 patent disclosure, an expert in the field noted that there were “significant biomechanical property differences between using peripheral grooves and interior slots.” The disclosed grooves had the effect of not allowing the fibers to be positioned differently from the center of the implant, opposed to other opening positions. Another expert stated that “the process of moving from peripheral grooves to internal slots [was] not a simple substitution, but a careful and time-consuming task” that took months of iterative prototyping. The slots in the accused SK device had to be very carefully chosen and sized to meet the functional demands of intervertebral structures. The majority found that the jury was properly able to credit this (unrebutted) testimony of important design considerations and effort in opening selection to conclude that a skilled person would not understand that a disclosure of peripheral grooves would disclose any and all openings on the plates.
SK further argued that, because the claims were to a mechanical invention, there was sufficient predictability in openings that the lone disclosed groove species should support the claim to all openings. While the majority agreed that predictability can lessen written description burdens for a genus, mechanical predictability not a panacea:
SK is correct that a “disclosure of a species may be sufficient written description support for a later claimed genus including that species.” Bilstad v. Wakalopulos, 386 F.3d 1116, 1124 (Fed. Cir. 2004) (emphasis added). But, as we stated in Bilstad:
[i]f the difference between members of [a species] is such that [a] person skilled in the art would not readily discern that other [species] of the genus would perform similarly to the disclosed members, i.e., if the art is unpredictable, then disclosure of more species is necessary to adequately show possession of the entire genus.
Id. at 1125. In other words, predictability is a factual issue judged on a case-by-case basis. Here, SK presented its case to the jury, and the jury inferred that, in the field of intervertebral implants, the disclosure of peripheral grooves does not adequately demonstrate possession of the entire genus of possible openings. Because the jury’s verdict is supported by substantial evidence, we must defer to that finding.
. . . while we did state in Bilstad that the mechanical field was “fairly predictable,” we did not hold that all inventions that may be characterized as “mechanical” allow claiming a genus based on disclosure of a single species.
Perhaps recognizing that this case fell within the margins where reasonable people could differ as to sufficiency of disclosure, the majority reemphasized that the jury was presented with enough evidence to conclude no possession here, regardless of predictability, affirmative species disclosure, and exemplary phrasing:
As we noted in Ariad, there are no “bright-line rules governing, for example, the number of species that must be disclosed to describe a genus claim, as this number necessarily changes with each invention, and it changes with progress in a field.” Ariad, 598 F.3d at 1351. Indeed, factual inquiries will, at times, create confounding results. But, whatever inconsistencies may appear “to exist in the application of the law, those inconsistencies rest not with the legal standard but with the different facts and arguments presented to the courts.” Id. at 1352. That is precisely the situation here.
Judge Taranto dissented, noting that this was not a case where written description was commonly lacking, like where an unpredictable or functional biological term is claimed, or where genus terms lacked any species support in an original disclosure. Rather, this was an instance of a later-added claim being in some ways broader than the disclosure. In such an instance, the written description analysis should look at how the claims were broader and ascertain only whether the identified exceedance impacts the purpose of the claimed invention:
for a challenger to prove insufficiency of the written description to support the claim language, the challenger must identify the respect in which the claim language differs from the disclosed embodiments. At a minimum, the challenger must then demonstrate that, in the eyes of a relevant skilled artisan, that particular difference has a material effect on whether the product or process would achieve the aims of the claims at issue, with materiality of the effect not the same as non-obviousness but related to predictability (this case requiring no further definition of that relation). See Ariad, 598 F.3d at 1351, 1352; Bilstad v. Wakalopulos, 386 F.3d 1116, 1124-25 (Fed. Cir. 2004) (inquiry into whether skilled artisans “would not readily discern that other members of the genus would perform similarly to the disclosed members, i.e., if the art is unpredictable”); In re Curtis, 354 F.3d 1347, 1355, 1358 (Fed. Cir. 2004) (inquiry into whether skilled artisan “could not predict the operability in the invention of any species other than the one disclosed”; focus on “unpredictability in performance”); Rasmussen, 650 F.2d at 1215. (inquiry into whether difference from described embodiment is “unimportant” to operability).
Interestingly, Judge Taranto pointed out that there is always some discretion/uncertainty in implementing any teaching, whether it is making a disclosed example embodiment or practicing a claim. If there is too much absolute uncertainty in either, then there is an enablement problem. But the written description requirement should look only at the relative discretion/uncertainty between practicing the disclosure and claim:
What is critical for present written-description purposes is this: if there is materially the same range of implementation judgments for the described embodiments and the broader claim—such as, here, how many openings/grooves to have, their shape, how close to the center and far from the periphery they would locate the fibers passing through—the need for [implementation] judgments is irrelevant to the written-description question. What matters is only the particular difference between the narrower embodiments and broader claims.
According to the dissent, this comparison is sufficient to meet the policy goal of the written description requirement – preventing patentees from usurping inventive activity by others, i.e., claiming undisclosed inventions that also have undisclosed (or undiscovered) utility.
Under the dissent’s approach, none of the expert testimony could reasonably be interpreted as clear and convincing evidence of lack of possession of the genus of all openings. The slots in the accused device were nothing more than the disclosed grooves with an extra, closing side. And this modification did not change the functional role of the grooves disclosed in the ‘270 patent – allowing the fibers to hold the plates and core together. The expert testimony’s failure to specifically contrast functionality between grooves and slots and statements on general difficulties in settling on an internal slot design with sufficient plate strength did not amount to the required showing of change in functionality of the disclosed grooves and claimed openings. Only the testimony of “a significant difference . . . in biomechanical properties” came close to this showing, but statement was still insufficiently specific or supported by corroboration to amount to substantial evidence. The dissent would have reversed the jury’s invalidity findings and remanded for a new infringement analysis.
Invalidity of claims 29-31 for lack of written description rejection affirmed. The majority also affirmed the district court’s denial of fees under § 285.
Lessons: The obvious, easy take-away here is to disclose as many specific examples as possible. Disclosure of single species should be avoided where possible; instead aim for an open-ended and varied list of multiple species that are consistent with a genus.
The more difficult practice is to try disclose and list examples with as much granularity as possible. Using the ‘270 patent as an example, the patent poorly disclosed attachment of the core to the plates with two examples, fibers in grooves and adhesives. In hindsight, this should have been drafted as a first-level, coarse disclosure of just “attachment” with a list including specific examples of mechanical and chemical attaching. The disclosure should have followed with an additional, finer disclosure of multiple types of chemical attachments (adhesives, chemical fusing, etc.) and multiple types of mechanical attachments (fibers surrounding in grooves, fibers going through plates, etc.) Of course it’s easy to give these examples now seeing the competitor’s art and the court’s opinion, but it reemphasizes the general lesson about considering how to drill-down to various levels of specificity with open-ended lists of examples at each tier in a disclosure.
I think the case also shows the importance of constructing claims from an essential elements standpoint, both in revising claims to ensnare a competitor and to achieve maximum claim scope ab initio. Through knowledge of the field, a prior art search, and/or prosecution, applicants usually get a feel for what is the patentable feature and how it needs to be recited for patentability. Consider omitting elements that are outside such features, not only because it is permitted and achieves general breadth, but because it avoids pinpointing written description issues. Here, for example, the ‘270 patentee presumably included the “openings” language only because the competitor used openings in product. But the ‘270 claims were allowable for an entirely different reason – the flexible core having the fibers and sheathing on it. Had the applicant instead presented claims detailing only the core with these fibers and sheathing (and possibly the surrounding plates for reference) as originally disclosed, without reciting openings or even any details of attachment, this may have preserved infringement and validity without getting into the written description issues of how the fibers passed through the plates. Put another way, if you find your claim linguistically drifting from your disclosure in order to capture infringement, at least consider cutting out the stretched term if allowance can be based on a separate and still infringed element.
Lastly, there are some litigation lessons here. The familiar “be careful what you ask for” applies during claim construction where a broad construction can give rise to enablement and written description challenges. Sometimes it’s safer to go with a narrower construction that requires a bit more complex factual argument for literal infringement or might get favorable doctrine of equivalents treatment. Here, I could see a reasonable argument for how the accused devices still included a “groove” with the internal slot. Also, the “groove” and “openings” terms were not argued during prosecution, so there was no DoE estoppel on what looked like very similar function-way-result elements. Also, don’t leave expert testimony on factual issues unrebutted. Here the majority had no trouble crediting the jury’s determination on written description as reasonable when the jury heard only from experts explaining why written description was lacking.
Comment: I think the dissent got this one right. While the procedural posture following the jury verdict gave the majority some leverage in affirming, I agree with the dissent that the type of evidence offered did not convincingly go to written description. If anything, the expert testimony seemed to show how the SK devices were an improvement over the ‘270 patent, not a design around it. The accused devices included the flexible core point of novelty of the ‘270 patent claims and used the same fiber attachment functionality as the disclosed grooves, with possibly additional/improved wear functionality in the type of attachment chosen. Thus, SK could have probably gotten their own patent on their specific type of improved fiber-plate attachment, and infringement could have been more equitably avoided through cross-licensing the improved design.
If the claim 29 had been an original claim, it would have likely survived a written description challenge because its original presentation constituted its own description. In this hypothetical, SK may have instead attacked the claim as failing enablement with respect to their infringing implants. I wonder if SK might have still carried the day on this argument, based on how similar their written description testimony was to argument about lack of enablement.
Finally, I believe this is Judge Taranto’s first dissent in a precedential patent case. A glimpse of things to come?