Claim-construction used to be the bread-and-butter of Federal Circuit decisions. That standard was of course upset by the onslaught of eligibility decisions from district courts and BRI obviousness cases out of the PTAB.
The patents at issue here all relate to circuit board construction – and dealing with the problem of delamination. The claimed solution is uses “a unique surface structure . . . comprised of teeth that are preferably angled or hooked like fangs or canine teeth to enable one layer to mechanically grip a second layer.” The figure below from the patent provides an example of how the layers might fit together.
The toothy layers can be formed in various ways — etching
Step 6, as discussed subsequently in greater detail, involves the etching cavities, veins, openings, or gaps in the applied dielectric material 8, or more particularly an outermost surface thereof, to accommodate the teeth. One technique for forming the teeth is somewhat similar to what has been known as the swell and etch or desmear process, except that contrary to all known teachings in the prior art, in effect, a “double desmear process” is utilized.
During prosecution, the patentee also explained that the invention uses “two separate swell and etch steps” to form the teeth. However, the claims themselves do not expressly require the double pass process.
In the first decision in this case, the district court considered the claims, specification, and prosecution history and found that the claims should be interpreted to require a double pass — even though not expressly required by the claims. Because the accused infringers only used a single-pass, non-infringement was then an easy finding.
On appeal, the Federal Circuit has rejected the narrow claim construction. As is its process, the court began with the claim language — finding that ” at least based on the plain language, the claims are not limited to a repeated desmear process.” Moving then to the specification, the court found no “clear and unmistakable disclaimer” of the claim scope covering a single-pass method. For the court here, the deciding factor appears to be the patent attorney’s care in drafting the specification. Here, the specification calls double desmear “one technique” that the invention “can be carried out” and “for example” rather than calling the approach “the invention.”
This approach here is properly seen as a patent drafting trick rather than an improved disclosure or invention. We know this because, although the specification called its double-pass approach “one technique” it did not describe any other technique — except for the prior art single-pass technique that the specification bemoaned as inferior.
Still, the Federal Circuit sided with the patentee and found that the words of ambiguity sufficient to avoid a claim scope narrowing.
Heeding the warning in Phillips to keep in mind that a goal of the specification is to provide a best mode to make and use an invention, phrases such as “one technique,” “can be carried out,” and “a way” indicate that using Probelec XB 7081 is only one method for making the invention and does not automatically lead to finding a clear disavowal of claim scope. We have also “expressly rejected the contention that if a patent describes only a single embodiment, the claims of the patent must be construed as being limited to that embodiment.” Id. . . . Additionally, distinguishing the double desmear process as “contrary to” or “in stark contrast” with the single desmear process, which again appears within the context of disclosures of the preferred embodiment, are not clear and unmistakable limiting statements. We have held that “[m]ere criticism of a particular embodiment . . . is not sufficient to rise to the level of clear disavowal.”
I’ll note that this decision is an example of the rhetorical approach of disassembly. After the appellee pointed to ten different reasons for limiting claim scope, the court walked through each one individually — and concluding that each offered “mere” subtleties and not a “clear disclaimer.” The court did not consider the impact of the whole. This is an easy criticism of the court’s decision, but the reality is that the court is now in a tough position. I would have also criticized the decision if it had gone the other way (which it could have based upon an alternative statement of the case and reading of precedent). The bigger problem here is that we need to step back and repeatedly ask – how can we build up this system in ways that truly encourage beneficial innovation and do not simply encourage extensive lawyering. This case is also a clear failure of the examiner who was told by the patentee that a particular aspect of the innovation was important, but then allowed the patent to issue without requiring that it be expressed in the claims.
Athena Diagnostics, Inc. v. Mayo Collaborative Services, LLC (Fed. Cir. 2019)
There is a lot to unpack in this decision, and so this is just a small discussion of an important Federal Circuit opinion – DC
Mayo is involved in another diagnostic method patent dispute — and again has come out on-top with a finding that the asserted patent claims are ineligible under 35 U.S.C. 101. That district court finding has now been affirmed on appeal, although subject to a strong dissent from Judge Newman.
The named inventors on Athena’s patent were working on a subset of Myasthenia gravis (MG) patients who did not exhibit the usual acetylcholine receptor antibodies. The researchers discovered that these patients were instead generating excess antibodies to a muscle-specific tyrosine kinase (MuSK) that the body uses in neuromuscular junctions. So, the key discovery here was the relationship between MuSK autoantibody production and MG.
The patent at issue is not directed a this relationship per se, but rather a method for diagnosing MuSK related disorders by looking for those MuSK autoantibodies. The particular claims at issue include a basic diagnostic method involving: mixing a person’s “bodily fluid” with a labeled antigen to the MuSK antibodies; immunoprecipitating any MuSK complexes from the fluid; and then looking for the label in the precipitate. Some claims indicate that the label is a radioactive iodine isotope.
Conventional Techniques: As we get into the eligibility analysis, it makes sense to note here that the claimed method is basically the one that any biochem PhD researcher would have come up with after learning about the importance of the MuSK autoantibody — i.e., it probably would be obvious if the relationship between MuSk and MG were in the prior art. However, the claimed method does involve creating non-naturally-occurring labeled antigens and then causing a chemical reaction that does not naturally-occur.
Under Mayo v. Prometheus, 566 U.S. 66 (2012) a law of nature may not be patented. In that case, the provided an example of a law of nature — the correlation between the a blood metabolite and the appropriate treatment dosage. The court explained explain that it is proper to have a claim “directed to” a law of nature, so long as the claim also includes “something more” such as an inventive practical application. In other words: “Laws of nature are not patentable, but applications of such laws may be patentable.” Athena.
Here, the Federal Circuit agreed with the lower court that the claims at issue are directed to a law of nature: “the correlation between the presence of naturally-occurring MuSK autoantibodies in bodily fluid and MuSK-related neurological diseases like MG.” This correlation exists in nature, even if only recently discovered by humans, and “there can thus be no dispute that it is an ineligible natural law.”
Here, although the claims include “certain concrete steps,” the court still determined that they were – as a whole – directed to the abstract idea “because the claimed advance was only in the discovery of a natural law, and . . . the additional recited steps only apply conventional techniques to detect that natural law.” To support this conclusion, the court cited to portions of the specification explaining that “the actual steps [are] known per se in the art” or are “standard techniques in the art.”
Regarding the required creation and use of a man-made molecule. The court held that “the use of a man-made molecule is not decisive if it amounts to only a routine step in a conventional method for observing a natural law.” I will note here that the molecule itself would likely be eligible — to the extent they were created by the researchers.
In its analysis, the court also distinguishes its recent decision in Vanda, which found a method of treatment based upon genotype to be eligible.
We consider it important at this point to note the difference between the claims before us here, which recite a natural law and conventional means for detecting it, and applications of natural laws, which are patent-eligible. See Vanda Pharm. Inc. v. West-Ward Pharm. Int’l Ltd., 887 F.3d 1117, 1133–36 (Fed. Cir. 2018) (holding that method of treatment by administering drug at certain dosage ranges based on a patient’s genotype was not directed to a natural law). Claiming a natural cause of an ailment and well-known means of observing it is not eligible for patent because such a claim in effect only encompasses the natural law itself. But claiming a new treatment for an ailment, albeit using a natural law, is not claiming the natural law.
Within step-one of the Mayo/Alice analysis the court basically answered step-two as well — does the claim include “something more.” Here, the court appears to have added the notion that step-one should focus only on the claim as a whole while step-two includes an element-by-element analysis. Regardless, the court found that the application steps were all simply recitations of steps known in the art.
Because the specification defines the individual immunoprecipitation and iodination steps and the overall radioimmunoassay as conventional techniques, the claims fail to provide an inventive concept. . . . [A]pplying standard techniques in a standard way to observe a natural law does not provide an inventive concept.
Although conventionality can now be seen as an issue of fact, the court found that it was effectively admitted within the specification.
Judge Lourie penned the majority opinion and was joined by Judge Stoll. Judge Newman wrote in dissent. Judge Newman argued that the claims should be seen at face value:
The claims … preparation of the new radioactive entities and their chemical reactions to detect autoantibodies to the protein muscle-specific tyrosine kinase (MuSK).
This, according to Judge Newman is “not a law of nature.”
Note here Footnote 4 from the majority opinion — agreeing with the dissent that “the public interest is poorly served by adding disincentive to the development of new diagnostic methods” and lamenting their inability to act. “Our precedent leaves no room for a different outcome here.”
President Trump’s State of the Union Speech included only a small focus on intellectual property rights — and only in the international sphere:
We are now making it clear to China that after years of targeting our industries, and stealing our intellectual property, the theft of American jobs and wealth has come to an end. . . .
Our new U.S.-Mexico-Canada Agreement — or USMCA — will replace NAFTA and deliver for American workers: bringing back our manufacturing jobs, expanding American agriculture, protecting intellectual property, and ensuring that more cars are proudly stamped with four beautiful words: made in the USA.
This appeal stems from a pending E.D.Tex. infringement lawsuit filed in 2017. SEVEN Networks, LLC v. Google LLC, 2:17-CV-00442 (E.D. Tex. Aug. 22, 2017). In the case, SEVEN accused Google of infringing ten of its patents:
8,078,158, titled “Provisioning Applications for a Mobile Device”
8,811,952, titled “Mobile Device Power Management in Data Synchronization Over a Mobile Network With or Without a Trigger Notification”
9,325,600, titled “Offloading Application Traffic to a Shared Communication Channel for Signal Optimization in a Wireless Network for Traffic Utilizing Proprietary and Non-Proprietary Protocols”
9,351,254, titled “Method for Power Saving in Mobile Devices by Optimizing Wakelocks”
9,386,433 titled “System and Method for Providing a Network Service in a Distributed Fashion to a Mobile Device”
9,444,812, titled “Systems and Methods for Authenticating a Service”
9,516,127, titled “Intelligent Alarm Manipulator and Resource Tracker”
9,553,816, titled “Optimizing Mobile Network Traffic Coordination Across Multiple Applications Running on a Mobile Device,”
Google filed for dismissal under TC Heartland — arguing that it did not reside in E.D. Texas, and had no “regular and established place of business” within the district as required for proper venue under 28 U.S.C. § 1400(b).
However, the district court found venue to be proper. The crux of the conclusion was based upon the fact that Google owns and uses computer servers within the district — “edge nodes” — that Google operate to quickly deliver content to its users within the district. Note here that, although Google owns the servers, it rents the rack space within various ISP server hubs.
The Federal Circuit then denied Google’s writ of mandamus in a non-precedential 2018 order — noting in dicta that the district court decision was fact-specific and reasoned:
Google owns and wholly controls the servers, located in the district under specific contracts with ISPs—which may not even tighten a screw without Google’s instruction. The location of the servers in the district serves Google’s business interests—by serving interests of Google’s ISP customers (e.g., in saving transport costs), Google’s end-user customers (e.g., in quick delivery of content), or both—as confirmed by Google’s advertising of its Edge Network and the Edge Nodes for efficient content delivery; and upon installation of the servers, Google places its inventory in (loads its content onto) those servers, and when Google’s end-user customers ask Google for that inventory, Google can and often does fulfill those requests from those local servers. And the server placement contracts provide Google very strong control to
keep the servers at their locations once they are installed.
Rather than ruling on the merits, the appellate panel court denied the writ — holding that mandamus is not warranted in this case. Rather, the majority approach is to wait until the conclusion of the trial and, if Google loses, it can then appeal the improper venue issue.
The mandamus denial was issued in a per curiam order by Judges Dyk and Taranto. Judge Reyna wrote in dissent — arguing that “Google’s petition presents fundamental issues concerning the application of § 1400(b) that have far reaching implications and on which district courts have disagreed.” Google persisted and filed a petition for en banc rehearing of its mandamus petition. That request has now also been denied, with the inclusion of another Judge Reyna dissent — this time precedential and joined by Judges Newman and Lourie.
The question poised before the court is whether Google’s servers … which have no physical interaction with Google employees or customers and are installed by third-parties in the facilities of third-party [ISPs] located in the Eastern District of Texas, constitute a regular and established place of business under 35 U.S.C. 1400(b) and this court’s decision in Cray.
In his dissenting opinion, Judge Reyna argues (1) this is an important question that should be addressed on mandamus; and (2) the answer is that servers likely do not count as a “regular and established place of business.”
[E]xclusive ownership and control over the servers may be insufficient under Cray. See Cray, 871 F.3d at 1363 (“Relevant considerations include whether the defendant owns or leases the place, or exercises other attributes of possession or control over the place.” (emphases added)). It is undisputed that no Google employee has ever visited the places where the servers are installed. Nor do those facilities resemble one of the many Google offices in other venues that would satisfy § 1400(b) under a straightforward application of the statute. See id. at 1364 (“A further consideration for this requirement might be the nature and activity of the alleged place of business of the defendant in the district in comparison with that of other places of business of the defendant in other venues.”). … For many [internet focused] companies, the reasoning of the district court’s holding could essentially reestablish nationwide venue, in conflict with TC Heartland, by standing for the proposition that owning and controlling computer hardware involved in some aspect of company business (e.g., transmitting data) alone is sufficient.
Here, Judge Reyna is not deciding the question, but suggesting that the answer is likely that venue is improper.
Back on the ground – the case has been proceeding throughout the entire mandamus process. Trial was set for January 2018. However, just before trial the parties asked for a stay of the case based upon an “agreement in principle” to settle the case. A full settlement is now likely within a few days.
Mylan Pharms. Inc. v. Research Corp. Techs., Inc. (Fed. Cir. 2019)
This post begins with an obviousness discussion, then moves to IPR standing, and finally talks about SAS.
Obviousness: In Mylan, the Federal Circuit affirms a PTAB decision siding with the patentee and holding that RCT’s pharmaceutical compound patent claims were not proven unpatentable (i.e., confirming patentability).
The patent at issue is directed toward an enantiomeric isomer (R configuration) of lacosamide. The dependent claims at issue are directed to the use of the drug for anticonvulsant effective (treating epilepsy symptoms).
As with most Inter Partes Review (IPR) decisions, the focus is obviousness. The listed inventor – Harold Kohn – is also the author of the lead prior art reference.
Harold Kohn et al., Preparation and Anticonvulsant Activity of a Series of Functionalized α-Heteroatom-Substituted Amino Acids, 34 J. Medicinal Chemistry 2444 (1991); J.A. 2404–12.
The article identifies a particular compound as the “most active” as a potential anticonvulsant whose structure is shown below.
Although the Board decided that there was sufficient similarity to institute the IPR, it ultimately ruled that the modification of the 1991 compound to create lacosamide was beyond ordinary skill in the art.
On appeal, the Federal Circuit affirmed — holding that even if the 1991 compound was an appropriate lead compound the challenger “failed to meet their burden to establish a motivation to modify.” Here, the challenger relied upon an organic chemistry drug design text for the principle that a developer would be motivated to replace an amine in the 1991 compound with a methylene link to form lacosamide.
According to Appellants, a person of skill in the art would then have been motivated to modify compound 3l by replacing the amine of its methoxyamino group with a methylene link to yield a more stable, synthetically accessible, pharmaceutically common and acceptable moiety.
On appeal, however, the Federal Circuit rejected this approach — holding that the PTAB decision was based upon substantial evidence. In particular, the Board looked at contemporaneous evidence that such a modification would have reduced therapeutic activity.
= = = = =
The case includes an important discussion regarding standing to appeal and the one-year bar under 5 U.S.C. § 319. Here, Argentum filed the IPR petition in November 2015. Argentum had not ever been sued under the patent and arguably would not have Article III standing to challenge the patent in court. However, the IPR statute allows anyone to petition.
After institution, Mylan, Brekenridge, and Alembic all filed their own IPR petitions as well as petitions for joinder under 35 U.S.C. § 315(c).
(c) Joinder.— If the Director institutes an inter partes review, the Director, in his or her discretion, may join as a party to that inter partes review any person who properly files a petition under section 311 that the Director, after receiving a preliminary response under section 313 or the expiration of the time for filing such a response, determines warrants the institution of an inter partes review under section 314.
The joinder petitions were granted. One quirk here is that the Mylan, Brekenridge, and Alembic IPR petitions would not have been granted on their own since they were all time-barred under 35 U.S.C. § 315(b). Joinder, however, is a different matter — since the statute expressly indicates that “The time limitation set forth in [§ 315(b)] shall not apply to a request for joinder under subsection (c).”
In the appeal, the original petitioner Argentum dropped-out and did not appeal — and arguably could not have appealed. Only the newcomers, Mylan, Brekenridge, and Alembic, appealed.
On appeal, RCT called foul — on the challengers approach of cobbling together rights.
Argentum able to file IPR, but no right to appeal.
Mylan, et al., not able to file IPR on its own, but standing to appeal.
In the end though, the Federal Circuit found the approach proper under the statute. Once Mylan was joined as a petitioner, it had a statutory right to appeal.
= = = = =
Final point to discuss here involves SAS Institute, Inc. v. Iancu, 138 S. Ct. 1348 (2018). In that case, the Supreme Court held that the PTAB must either institute as to either (1) all challenged claims or (2) no claims. The Federal Circuit has expanded that holding to now include a requirement of instituting as to either (1) all grounds for challenge or (2) no grounds at all.
Here, the PTAB instituted as to all challenged claims, but only some grounds for challenge. In oral arguments (rebuttal), Mylan’s counsel requested, in the “alternative,” that the case be remanded for consideration of all grounds. On appeal, the Federal Circuit held that the right to remand had been waived. In particular, applicant waited more than 6-months until the rebuttal portion of oral arguments to request the remand. “Given the circumstances in this case, we find that Appellants have waived their request for remand.”
Chad Gilles at BigPatentData reports on the first handful of PTAB decisions considering eligibility under the 2019 Revised Subject Matter Eligibility Guidance. The outcome — 101 Rejection Affirmed in five of six cases. Gilles writes “If your claim can still be lumped into the very broad ‘certain methods of organizing human activity’ the new guidance may not be of much help.”
This image is a bit hard to read, but it shows the cross-border invention data mentioned in yesterday’s post — focusing on which nations tend to have collaborative inventors with which other nations. US is central here, but there are other clusters as well.
U.S. patents list the country of residence of the inventors at the time of patent issue — although I imagine that many applicants do not update residence during application pendency. The chart below looks at the first-listed inventor for recently issued patents and then reports the average number of inventors per patent for each country. An interesting caveat to this report is that about 8% of U.S. patents involve cross-border inventions. These are multi-inventor patents with at least two of the inventors residing in different countries.
Duncan Parking Techs., Inc. v. IPS Group, Inc. (Fed. Cir. 2019) and IPS Group, Inc. v. Duncan Solutions Inc. (Fed. Cir. 2019)
This decision involves parallel appeals from (1) a district court siding with the accused infringer on noninfringement grounds; and (2) a PTAB IPR decision siding with the patentee confirming patentability. On appeal, the accused infringer won on several grounds, but it appears that at least some of the claims will be found both infringed and valid.
In this post, I focus on the PTAB decision and self-harming prior art.
0=
The two patents in suit are related to single-space, credit-card enabled, solar-powered parking meters. U.S. Patents 8,595,054 and 7,854,310. Of the two, the ‘054 patent has the earlier effective filing date (December 2006) while the ‘310 patent’s associated application was not filed until 14 months later (February 2008). The two patents are both owned by IPS and have one overlapping inventor (King). However the patents also have several non-overlapping inventors. Relevant for this case, Schwarz is a listed inventor of the ‘054 patent but not the ‘310 patent.
Once the ‘054 patent published in 2009, it became prior art under pre-AIA 35 U.S.C. 102(e) with a prior art date of December 2006. 102(e) offers two additional tricks for our situation here: First, a published application or patent only counts as prior art if filed “by another”; and second, 102(e) prior art cannot be used for obviousness purposes if the purported prior art is owned by the same entity as the application/patent being challenged.
By Another: This gets a little bit tricky, but the baseline law is that a prior art application is “by another” if invented by someone not listed in the later application being challenged. Here, the purported prior art (the ‘054 patent) lists King and Schwarz as inventors — that counts as a different since Schwarz is not listed on the ‘310 patent. However, the court allows parties to get into the weeds on this point and ask — who actually invented the particular aspects of the disclosure being used. Here, the patentee argued that King (the overlapping inventor) is the one who created the particular disclosure. On appeal, the Federal Circuit agreed with the legal analysis, but disagreed on the facts:
If Schwarz is a joint inventor of the anticipating disclosure, then it is “by another” for the purposes of § 102(e). . . . Because the [later] ’310 patent claims clearly include elements previously disclosed in Figure 8 [of the prior ‘054 patent], the question is whether Schwarz conceived those elements as they were disclosed in the ’054 patent. . . .
While IPS argues that Schwarz’s role consisted of “drawing a schematic,” the record shows that creating the block diagram was not a minor task. Schwarz had to coordinate with D+I as they designed a device that could
physically fit within an existing parking meter housing. As a result of Schwarz’s contribution, the ’054 patent discloses a parking meter device with a detailed electrical system, a challenge King described as one of “two big hurdles” to the invention. Thus, Schwarz’s contribution, measured against the dimension
of the full invention, was significant.
Okay, based upon the Federal Circuit’s conclusion that Schwarz is an inventor of that disclosure in the ‘054 patent and not listed on the ‘310 patent, his contribution is “by another” and thus can count as 102(e) prior art.
Same Owner: The same-owner caveat on 102(e) is quite powerful. However, pre-AIA it only applied to limit the obviousness doctrine. Here, the challenge is anticipation and so the reference counts as prior art.
The result here is that the earlier application counts as prior art — as such, Federal Circuit reversed the PTAB decision and remanded for a finding that the claims should be cancelled.
= = = =
Note here that these details have been substantially changed by the America Invents Act. Under the new law, the same-company exception applies for both obviousness and anticipation. Thus, this case would have a different outcome.
In addition, the “by another” language was removed and replaced with “names another inventor.” I don’t believe that this second language actually changes the law.
= = = =
The result here is that the ‘310 patent is invalid, but the ‘054 patent remains strong and at least some of its claims appear infringed.
On February 26, the USPTO will be holding a 3-hour online training focused on the 2019 Revised Subject Matter Eligibility Guidelines. [Free – Sign-Up Here] This will basically parallel examiner training on the same subject matter and will be led by USPTO trainers. The Office has limited space, but has scheduled potential additional training on February 27 and 28. Sign-up by February 3, 2019.
Hikma Pharmaceuticals USA Inc., v. Vanda Pharmaceuticals Inc. (Supreme Court 2019)
I wrote earlier about the Federal Circuit’s decision on Vanda’s drug dosage claims that have two basic steps:
Determining whether a patient is likely a poor metabolizer of the drug iloperidone based upon DNA analysis (i.e., if the patient has has the CYP2D6 genotype)
Administering iloperidone at a lower dose to predicted poor metabolizers in order to reduce the risk of “QTc prolongation” for poor metabolizers.
In the case, iloperidone was already known as a drug treatment and that some folks were poor metabolizers in a way that created the risk. For poor metabolizers, the lower dose is still effective (since the drug stays in the body longer). A remaining problem though was how to predict who should get the low dosage. The core discovery here is that a genetic difference explains the risk, and the inventors made that discovery and then implemented it with its administration steps. The question then is whether this approach is patent eligible.
I previously explained my position in the case as follows:
(1) Vanda’s claims should be patent eligible [on policy grounds]; (2) but the claims are not patent eligible under Mayo Collaborative Services v. Prometheus Laboratories, Inc., 566 U.S. 66 (2012) and Alice Corp. Pty. v. CLS Bank Int’l, 134 S. Ct. 2347 (2014).
The Federal Circuit disagreed with my second conclusion and in a 2-1 decision held the claims eligible. Patent challenger Hikma (formerly West-Ward) has petitioned the Supreme Court review and that petition is supported by two strong amicus briefs:
Hikma Pharmaceuticals USA Inc., et al. v. Vanda Pharmaceuticals Inc.
All three briefs say basically same thing — The Federal Circuit’s Vanda decision conflicts with Mayo and Flook. “The Federal Circuit’s decision effectively overturns this court’s precedents, thwarts the proper development of patent eligibility law, and will lead to countless improperly issued patents.”
The Law Professor Brief was filed by Stanford’s IP Clinic – although neither Mark Lemley nor Lisa Larrimore Ouellette signed-on. Professors Josh Sarnoff (DePaul) and Katherine Strandburg (NYU) substantially drafted the brief. Top pharma-patent litigator Douglass Hochstetler (KelleyDrye) filed the AAM brief.
I recently posted a chart showing a short time series showing the number of inventors per patent. The basic conclusion from that post was in the headline: Average Number of Inventors per Patent Continues Steady Rise. Here, I have updated and added to those charts. The first chart below reaches back to 1976 — showing that the average number of inventors per patent has been steadily increasing over all of the past 40 years. The trend has been amazingly steady and linear, and should serve as good fodder for economic analysis. My simple linear model suggests that we’ll top the 3.0 inventors-per-patent threshold by 2021 (both the mean and median).
The second and third charts (below) provides a bit more detail and looks to answer some questions posted by James Daily of WashU regarding the underlying distributions. The basic conclusion there is that the proportion of one-inventor patents has dropped steadily — and can be linked with a steady rise in patents listing three or more inventors. During this time, the proportion of two-inventor patents has remained steady at about 25%. The charts also demonstrate that the growth in 3+ inventor patents is itself skewed: For example, the proportion of patents with 3 inventors doubled during the 43 year span while the proportion of patents with 7+ inventors saw a nine-fold increase.
Although the movement was already fully underway, recognize that the America Invents Act (AIA) helps foster this as well by (1) allowing patents to be filed by the applicant rather than the inventors; and (2) weakening the penalty for errors in listing of inventorship. That change was by design and part of what Dir. Kappos notion that major inventions these days are quite often major team projects, and that the patent law should foster and encourage those teams projects at least as much as it encourages individual innovation.
The new HUAWEI indictments make include some amount of comedy. The basic claim centers on an anthropomorphic robot named “Tappy.” Tappy is owned by T-Mobile, but HUAWEI was allowed to use Tappy to test phones it was making for T-Mobile (subject to a non-disclosure agreement). HUAWEI then started developing its own robot — with the much less friendly name of xDeviceRobot. HUAWEI tried to buy Tappy, but was refused and so figured out a way to infuse xDeviceRobot with some of Tappy’s features. That infusion of ideas is what los federales consider criminal theft of a trade secret in violation of the Federal Economic Espionage Act.
I’ll note here that Tappy the Robot is also a street performer with its (his?) own FaceBook page. I might suggest a dance-off to settle the dispute.
It turns out that HUAWEI perhaps didn’t take all the steps necessary for criminal trade secret theft. But, the criminal prosecutors were able to indict the company on conspiracy to steal trade secrets and attempted trade secret theft; as well as wire fraud and obstruction of justice.
This past fall at the Administering Patent Law symposium at Iowa Law, Professor Colleen Chien presented an argument in favor of more intentional experimentation by administrative agencies such, as the USPTO to test policy concepts and proposed several possibilities. Below, Professor Chien describes one such project, based on the Merges/Crouch proposal for deferring patentable subject matter analyses in prosecution. Read the accompanying PatentlyO Law Journal article here:
Colleen V. Chien, Piloting Applicant-Initiated 101 Deferral Through A Randomized Controlled Trial, 2019 Patently-O Patent Law Journal 1. (2019.Chien.DeferringPSM)
Guest post by Colleen V. Chien, Justin D’Atri Visiting Professor of Business Law, Columbia University School of Law; Professor, Santa Clara University Law School
“The journey of a thousand miles begins with a single step.” – Lao Tzu
Responding to frustration with the Supreme Court’s patentable subject matter (PSM) decisions, the Federal Circuit has issued clarifying decisions, the USPTO has released new guidance regarding applying Section 101 (and 112), and Senators Coons and Tillis have been holding roundtables. 101 appeals and rejections, including pre-abandonment rejections, have risen following Alice and Mayo, particularly within impacted technology areas. (Figs 1A and 1B, details here and here). To support tracking the impact of its own guidance and related developments, the USPTO should release updated versions of office action data (the last release was in 2017).
In this post, I propose a complementary approach that aims to conserve agency and applicant resources even while policymakers work to clarify the law. It builds on the idea of deferring 101 subject matter until other issues are exhausted first proposed by Robert Merges and Dennis Crouch, but with some important modifications. Deferral would be at applicant’s option, not mandatory, and the USPTO should roll out this intervention through a controlled trial with randomization, as the agency has done previously, to determine the practice’s effectiveness.
Experimenting with Ordering
The case for deferring subject matter builds on the insight that 101 is rarely the single dispositive issue – among office actions prior to abandonment, less than 15% of TC36BM and medical diagnostic applications and 2% applications overall are “only 101.” Yet 101 is controversial – the rate of ex parte decisions addressing 101 has also shot up, from less than 10% to over 80% in 2018 in medical diagnostic and software technology areas, and to 26% overall. (Fig 1B)[1]
Deferring 101 would borrow from the Supreme Court doctrine of avoidance, which allows the Court to “resolve[] cases on non-constitutional grounds whenever possible,” in order to conserve court resources and legitimacy. Other federal agencies have also used deferral, for example, in the mid to late 2000s, when the application of immigration law to same-sex couples in some cases required family separation. To avoid this harsh consequence, Department of Homeland Security prosecutors administratively closed some cases, immigration judges granted continuances for unusually long periods in hopes that the law would change, and US Customs and Immigration enforcement officials granted requests for deferred action. That is to say, they avoided the law by deferring its application.
Avoiding PSM issues would allow cases to resolve, through allowance or abandonment, on less controversial, non-101 grounds.[2] (The USPTO’s new guidelines implement this logic to a degree, shifting focus towards 112 and away from 101). But making 101 deferral optional, at applicant’s discretion, would preserve the benefits of compact prosecution and the freedom applicants enjoy to select a fast or slow track for each application.
The Advantages of Experimentation
Trying out applicant-initiated 101 deferral through a rigorous pilot has several advantages. First, unlike changing the law or its application which requires all to adjust, only applicants dissatisfied with the status quo would see a change. 85% of office actions don’t even include a PSM rejection I’ve found previously.
Second, implementing 101 deferral as a pilot with randomization, like the USPTO did when it conducted the post-trademark registration proof-of-use pilot, will support effective policy-making through evaluation, iteration, and refinement.[3] Design (intent-to-treat), ethical (through consent) and methodological (power, randomization, etc.) concerns and details are discussed in the accompanying piece.Rigorously tracking the impact of PSM on patent prosecution over the course of a pilot can also provide much-needed empirical support to 101 policy-making in general.
Finally, my proposal preserves the benefits of the status quo, including the little-discussed incentive prosecutors now have to add details to their patent applications and claims, a good thing. In work with students that I presented at the Federal Trade Commission’s recent hearings building on an earlier analysis in IP Watchdog by Will Gvoth, Rocky Bernsden, and Peter Glaser from Harrity & Harrity LLP, I observed that there has been a “flight to quality” among patent complaints and applications. Carrying out a differences-in-differences analysis, we found that specification length and counts of words and unique words in the first claim have grown among software applications relative to others following the Alice decision. Deferring 101 would preserve this previously unexplored impact of the “Mayo-Alice effect” on software patent drafting. If 101 were eviscerated by Congress, so too could the incentive to be more concrete in describing and claiming inventions.
Conclusion
If successful, the treatment would result in the diminished presence of 101 subject matter issues within ex parte appeals and pre abandonment rejections and, potentially, resolution time, e.g. closer to pre-Mayo or Alice levels. Applicant and prosecutor satisfaction with the process, changes made to what gets filed, timelines, and such factors would also be worth tracking.
Read it here: Colleen V. Chien, Piloting Applicant-Initiated 101 Deferral Through A Randomized Controlled Trial, 2019 Patently-O Patent Law Journal 1. (2019.Chien.DeferringPSM)
With thanks to Santa Clara Law University students Jiu-Ying Wu, Nicholas Halkowski, Marvin Mercado, and Saumya Sinha and to William Gvoth, Peter Glaser, and Rocky Bernsden of Harrity LLP for assistance with data, and Hans Sauer and Jonathan Probell for their comments on earlier drafts.
[1] Even though 101 patentable subject matter issues appear in a quarter of 2018 appeals decisions, this share is still small compared to the share of decisions that mention, for example, “102” (, “103”, or “112” issues though note that these numbers are inflated because they do not include case specific limitations that can weed out false positives based on incidental mentions (such as “ 10/282102”)
[2] For example, if an invention is incurably anticipated, obvious or unsupported, the case will have been resolved without considering subject matter. Likewise, if the claims have been re-formulated responsive to non-101 rejections and mooted any subject matter defects in the process, PSM will not factor into the application’s outcome.
[3] In that pilot, the USPTO randomly selected 500 registrations to participate in the initial program to assess the accuracy and integrity of the trademark register. Though companies selected to participate in the pilot had to comply with additional regulatory requirements, randomization ensured that the pilot’s findings were representative and its burdens, fairly distributed. See, 77 FR 30197, 30198 (explaining that randomness in the pilot was necessary to “ensure that the resulting assessment is not skewed by consideration of registrations with particular criteria, and that implementation of the rules does not create an unfair burden on specific types of trademark owners”). The results of the test and related USPTO reports, outreach and deliberations were used to expand and make the program permanent.
Prior Patently-O Patent L.J. Articles:
David A. Boundy, Agency Bad Guidance Practices at the Patent and Trademark Office: a Billion Dollar Problem, 2018 Patently-O Patent Law Journal 20. (Boundy.2018.BadGuidance)
Paul M. Janicke, Patent Venue: Half Christmas Pie, And Half Crow, 2017 Patently-O Patent Law Journal 13. (Janicke.2017.ChristmasPie.pdf)
Paul M. Janicke, The Imminent Outpouring from the Eastern District of Texas, 2017 Patently-O Patent Law Journal 1 (2017) (Janicke.2017.Venue)
Mark A. Lemley, Erik Oliver, Kent Richardson, James Yoon, & Michael Costa, Patent Purchases and Litigation Outcomes, 2016 Patently-O Patent Law Journal 15 (Lemley.2016.PatentMarket)
Bernard Chao and Amy Mapes, An Early Look at Mayo’s Impact on Personalized Medicine, 2016 Patently-O Patent Law Journal 10 (Chao.2016.PersonalizedMedicine)
James E. Daily, An Empirical Analysis of Some Proponents and Opponents of Patent Reform, 2016 Patently-O Patent Law Journal 1. (Daily.2016.Professors)
Tristan Gray–Le Coz and Charles Duan, Apply It to the USPTO: Review of the Implementation of Alice v. CLS Bank in Patent Examination, 2014 Patently-O Patent Law Journal 1. (GrayLeCozDuan)
Robert L. Stoll, Maintaining Post-Grant Review Estoppel in the America Invents Act: A Call for Legislative Restraint, 2012 Patently-O Patent Law Journal 1 (Stoll.2012.estoppel.pdf)
Paul Morgan, The Ambiguity in Section 102(a)(1) of the Leahy-Smith America Invents Act, 2011 Patently-O Patent Law Journal 29. (Morgan.2011.AIAAmbiguities)
Joshua D. Sarnoff, Derivation and Prior Art Problems with the New Patent Act, 2011 Patently-O Patent Law Journal 12 (sarnoff.2011.derivation.pdf)
Bernard Chao, Not So Confidential: A Call for Restraint in Sealing Court Records, 2011 Patently-O Patent Patent Law Journal 6 (chao.sealedrecords.pdf)
Benjamin Levi and Rodney R. Sweetland, The Federal Trade Commission’s (FTC) Recommendations to the International Trade Commission (ITC): Unsound, Unmeasured, and Unauthoritative, 2011 Patently-O Patent Law Journal 1 (levi.ftcunsound.pdf)
Kevin Emerson Collins, An Initial Comment on King Pharmaceuticals: The Printed Matter Doctrine as a Structural Doctrine and Its Implications for Prometheus Laboratories, 2010 Patently-O Patent Law Journal 111 (Collins.KingPharma.pdf)
Robert A. Matthews, Jr., When Multiple Plaintiffs/Relators Sue for the Same Act of Patent False Marking, 2010 Patently-O Patent Law Journal 95 (matthews.falsemarking.pdf)
Kristen Osenga, The Patent Office’s Fast Track Will Not Take Us in the Right Direction, 2010 Patently-O Patent L.J. 89 (Osenga.pdf)
Michael Kasdan and Joseph Casino, Federal Courts Closely Scrutinizing and Slashing Patent Damage Awards, 2010 Patently-O Patent L.J. 24 (Kasdan.Casino.Damages)
Edward Reines and Nathan Greenblatt, Interlocutory Appeals of Claim Construction in the Patent Reform Act of 2009, Part II, 2010 Patently‐O Patent L.J. 7 (2010) (Reines.2010)
Gregory P. Landis & Loria B. Yeadon, Selecting the Next Nominee for the Federal Circuit: Patently Obvious to Consider Diversity, 2010 Patently-O Patent L.J. 1 (2010) (Nominee Diversity)
Paul Cole, Patentability of Computer Software As Such, 2008 Patently-O Patent L.J. 1. (Cole.pdf)
John F. Duffy, The Death of Google’s Patents, 2008 Patently O-Pat. L.J. ___ (googlepatents101.pdf)
Mark R. Patterson, Reestablishing the Doctrine of Patent Exhaustion, 2007 Patently-O Patent L.J. 38
Arti K. Rai, The GSK Case: An Administrative Perspective, 2007 Patently-O Patent L.J. 36
Joshua D. Sarnoff, BIO v. DC and the New Need to Eliminate Federal Patent Law Preemption of State and Local Price and Product Regulation, 2007 Patently-O Patent L.J. 30 (Download Sarnoff.BIO.pdf)
John F. Duffy, Are Administrative Patent Judges Unconstitutional?, 2007 Patently-O Patent L.J. 21. (Duffy.BPAI.pdf)
Joseph Casino and Michael Kasdan, In re Seagate Technology: Willfulness and Waiver, a Summary and a Proposal, 2007 Patently-O Patent L.J. 1 (Casino-Seagate)
In re Ikeda Food Research Co., Ltd. (Fed. Cir. 2019)
Ikeda is attempting to patent a glucose sensor using a particular enzyme (“a soluble flavin compound-binding glucose dehydrogenase”) that promotes glucose oxidizing. U.S. Application No. 12/851,668. One U.S. patent is already issued in the family as well as Japanese , Chinese, and European patents. In this case, however, the examiner rejected the claims as obvious and the PTAB affirmed. The Federal Circuit has now also affirmed that finding.
The case is most interesting for its inherency finding. Here, the claims at issue particularly require low-maltose activity (“enzymatic activity to maltose in the enzymatic reaction layer is 5% or less relative to the enzymatic activity to glucose.”) The prior art did not expressly teach that limitation, but the PTAB found it inherently disclosed. In particular, the cited prior art used the same enzyme preparation as the patentee, and the PTAB concluded that it would inherently exhibit the same low-level maltose activity.
Teachings that are inherent to a disclosure work as prior art — both for anticipation and obviousness. However, the court applies fairly strict requirements before it will assume the existence of unstated elements. “[T]he limitation at issue necessarily must be present[] or the natural result of the combination of elements explicitly disclosed by the prior art.” PAR Pharm., Inc. v. TWI Pharm., Inc., 773 F.3d 1186 (Fed. Cir. 2014).
In reviewing the evidence, the Federal Circuit found that the PTAB had reasonably concluded that when the enzyme was used in the prior art that it had the identical activity as claimed by Ikeda. In particular, the PTO properly made a prima facie case of inherency that was not sufficiently rebutted by the patent applicant.
The one good argument Ikeda made here involved a separate reference (Tsuji) that also used the identical enzyme and ended up with high-maltose activity. The courts, however, distinguished Tsuji because there was evidence of sample contamination in that case and no evidence that other non-related samples were contaminated.
The approach here follows Southwire Co. v. Cerro Wire LLC, 870 F.3d 1306 (Fed. Cir. 2017), where the court found it reasonable to assume that functional limitations were met if all of the other process limitations are met. “[T]here is no indication that
the limitation is anything other than mere quantification of the results of a known process.” Id.
In the absence of any evidence that the claimed [low activity] would have been unexpected in light of [Senior’s] disclosure, there is no indication that the limitation is anything other than mere quantification of the results of a known process.
= = = = =
One unusual aspect of this case is that the parties are not fully named. Although Ikeda is apparently title owner, there is a second unnamed party that is “the Real Party in interest.” In its certificate of interest, Ikeda explains that it “shares a confidential business relationship with this entity.” Judge O’Malley granted Ikeda’s unopposed motion to keep this info secret.
Before the PTAB, Ikeda apparently did not reveal the darkhorse real-party in interest. In its decision, the PTAB explains “According to Appellants, the real party in interest is Ikeda Food Research Co. Ltd.”
Google, LLC v. Oracle America (Supreme Court 2019)
This new petition from Google asks important questions about the role of copyright in protecting software. Questions presented:
Whether copyright protection extends to a software interface.
Whether, as the jury found, petitioner’s use of a software interface in the context of creating a new computer program constitutes fair use.
Rather than licensing Java, Google rewrote the programming language for Android App development in a way that mimics Java. No problem there so long as done without copying. However, rather than renaming all the function calls, Google decided to copy the names of all the function calls. I previously explained:
As an example, Google used the Java method header “Math.max(a,b).” . . . In Android’s API, Google copied a set of 37 different Java “packages” (such as Math) that each contain many classes and method calls (such as “max()”). Overall, Google copied the header structure for more than six-thousand methods.
Google’s reason for copying the naming structure was a form of free-riding. There were already many thousand Java programmers and Google was looking for a way to get them to develop Apps for Android. In its shortest form, Google explains the case as follows:
As is relevant here, software interfaces are lines of computer code that allow developers to operate prewritten libraries of code used to perform particular tasks. Since the earliest days of software development, developers have used interfaces to access essential tools for building new computer programs. Contravening that longstanding practice, the Federal Circuit in this case held both that a software interface is copyrightable and that petitioner’s use of a software interface in a new computer program cannot constitute fair use as a matter of law.
After reading the Federal Circuit decision, I wrote that the case is “likely heading to the Supreme Court.” Although I believe that the case has a very good shot – one difficulty is that it involves a decision by the Federal Circuit applying Ninth Circuit law — it effectively holds no weight and can be simply rejected by the next Ninth Circuit panel addressing the same issues.
Willis Electric (Taiwan) owns several patents on artificial Christmas trees. In 2015, Willis sued Macao-based Polygroup for infringement. Polygroup turned-around and filed several inter partes review (IPR) petitions that were granted. In its final decisions, the PTAB sided with the patentee – finding that Polygroup had not proven the claims unpatentable. On appeal, the Federal Circuit has rejected the PTAB’s claim construction. Below, I look at just one of the cases — this one involving “modular artificial tree.”
Claim Construction: The PTAB has recently begun performing “standard” claim construction during inter partes review proceedings – rather than the broadest reasonable interpretation. This case, however falls under the old regime.
The claim term at issue in the preamble phrase – “modular artificial tree”
A modular artificial tree, comprising: . . .
Even applying the broadest-reasonable-interpretation, the PTAB found the phrase limiting — and construed it to require “a tree constructed of modular portions, each modular portion being a separate tree section” with pre-attached branches. The cited prior art references all had branches that separately attached to the trunk — and thus did not provide a prior art teaching of the pre-attached branchs.
On appeal, the Federal Circuit rejected this construction — finding that it “does not represent the broadest reasonable interpretation of ‘modular artificial tree.'” None of the claims require branches, and the proffered expert testimony regarding the term’s meaning was “conclusory” and “unsupported” by corroborating evidence.
Relying on the testimony of Willis’s expert, Dr. Brown, the Board concluded that a modular tree is “a ‘distinct’ type of artificial tree with tree sections that come ‘with branches pre-attached to the trunk.’” . . .
Dr. Brown’s statement is a conclusory and unsupported assertion, and we find it to be inconsistent with the intrinsic record. So, his testimony does not provide substantial evidentiary support for the Board’s conclusion that a modular tree is one with “branches pre-attached to the trunk.”
Instead, the Federal Circuit found that a more natural broadest-reasonable construction of modular artificial tree is as “an artificial tree with elements capable of being easily joined or arranged with other parts or units.”
On remand, the Board will consider whether this change alters the non-obviousness conclusion.
In a split decision, the Federal Circuit has sided with the patentee Dr. Barry — affirming the induced infringement and $17 million damages judgments as well as several determinations regarding the public-use and on-sale statutory bars. The patents here are directed to spine surgery methods that involve large screws drilled into the spine. U.S. Patent Nos. 7,670,358 and 8,361,121.
The dissenting opinion by Chief Judge Prost offers a concise explanation of a primary issue in the case:
The facts are simple. More than one year before filing for the ’358 patent, Dr. Barry successfully performed his claimed surgical method on three different patients, charging each his normal fee. Dr. Barry’s method was thus prima facie “on sale” or in “public use” before the critical date under 35 U.S.C. § 102(b) [pre-AIA].
The majority added to these facts — that Dr. Barry was not satisfied that his method worked well enough until follow-up testing that occurred after the third surgery.
Here, the patented invention is a method — and so cannot really be “on sale.” In addition, Barry’s pre-critical-date performance of the method was within the surgical rule where HIPAA and other privacy rules ensure that the method was not publicly used. However, one conflict was the extent that Barry was commercializing the the invention — with the idea that commercialization counts as a public use. In a footnote, the Court concludes that he did not “exploit” his invention prior to the critical date since he charged the same for his new surgery has he did for the old.
Though earning his normal fees from the three surgeries [likely $$$$$], Dr. Barry did not “exploit” his invention as a means to attract the three patients for those surgeries or to charge more because he used his new technique. The jury could find that he would have gotten the same business, and earned the same fee, even if he had not planned to use or used the inventive process.
There are a number of major problems with this conclusion from the court — likely enough to fill a law review article.
In the end, for the majority, the case turned on the question of whether the invention was “ready for patenting” at the time when it was used for the surgeries. The majority found that it was not yet ready for patenting – and thus that the uses did not serve as prior art. Even though the invention had been reduced to practice used in practice, it was not ready for patenting because Dr. Barry didn’t really know if it worked to create a stable improvement in spine function. “[T]here is substantial evidence that Dr. Barry’s invention was not ready for patenting until January 2004 because the final follow-up from the October surgery was reasonably needed for the determination that the invention worked for its intended purpose.” The dissent again disagrees — arguing accurately that “Dr. Barry could have obtained a patent before the critical date.”
Note here that in my original short explanation of this decision, I indicated that the invention “had been reduced to practice.” According to the majority, the invention was not reduced to practice because that requires a determination (by the inventor) “that the invention would work for its intended purpose.” In re Omeprazole Patent Litig., 536 F.3d 1361 (Fed. Cir. 2008).
