by Dennis Crouch

The Solicitor General’s new brief in Amgen v. Sanofi shrugs off the case–identifying petitioner’s arguments as meritless:

Petitioners contend that the court of appeals erred by treating enablement as a question of law and by examining the full scope of the claims in assessing whether they are fully enabled. Those arguments lack merit and further review is not warranted.

Gov’t CVSG Brief in Amgen Inc. v. Sanofi, 21-757 (Supreme Court 2022).

The petition argued that genus claims–especially those functionally claimed–are being asked to comply with a heightened enablement standard.  The Gov’t agreed that at times it may be difficult to enable an entire genus, but not because of any heightened standard.  Rather, “broad claims naturally require more extensive enablement” since the patent’s disclosure “must be commensurate with the scope of its claims. . . . When, as here, a patent claims an entire genus based on its function, the patent must enable that entire genus.” The brief cites to Consolidated Electric Light Co. v. McKeesport Light Co., 159 U.S. 465 (1895) and Holland Furniture Co. v. Perkins Glue Co., 277 U.S. 245 (1928) as representative.  In those cases, the Supreme Court used to enforce patents whose scope encompassed, but far exceeded, the examples given in the specification.  Of course, the Brief’s bold statements fail to contend with the reality that enabling the “entire genus” will effectively curtail genus claims, and the resulting policy shifts.  I expect that a Supreme Court decision in the the case would not go as far as suggested by the Gov’t Brief, but the middle ground might not be enough to flip the result here.

The Gov’t Brief also explains enablement as a mixed question of fact and law.  Although the jury sided with the patentee, the district court judge rejected the verdict on JMOL  — holding that the claims lacked enablement.  That JMOL determination was affirmed on appeal.  The Government brief does not delve deeply into the issues raised but simply states:

Petitioners complain about the formulation the court of appeals used in articulating the standard of review, but they do not identify any practical implications flowing from that disagreement.

Id.

Usually, the SG brief is the most important amicus brief at predicting certiorari.  However, that might no longer be true for today’s 6-3 conservative majority hearing remarks from a liberal administration.

by Dennis Crouch

Novartis Pharms Corp. v. Accord Healthcare, Inc. (Fed. Cir. 2022)

Without any fanfare or dissent, the Federal Circuit has denied the Novartis en banc petition.  The court used a questionable procedure flip its own prior decision by replacing Judge O’Malley with a more defendant-friendly Judge Hughes for the rehearing.

The merits decision appears to further tighten-up on the written description requirement — especially with regard to ‘negative’ claim limitations.  The basic holding is that the written description must either expressly or inherently disclose the invention.  Novartis Pharm. Corp. v. Accord Healthcare, Inc., 38 F.4th 1013 (Fed. Cir. 2022) (rehearing majority decision).  Here, ‘inherently’ is a term of art in patent law and has much stricter definition than its cousin ‘impliedly.’  Inherency in patent law indicates unstated certainty.

The patent at issue in Novartis claims a drug treatment method.  Sometimes in drug treatment, you start a patient off with a high ‘loading dose’ to get the blood-levels up to an operational state.  The patent application does not discuss a loading dose one way or the other.  During prosecution, the patentee added a no-loading-dose negative limitation: “a daily dosage of 0.5 mg, absent an immediately preceding loading dose regimen.”

Someone skilled in the art might read the specification as implying that no loading dose was necessary. But, that same person would have to admit that the specification could be interpreted in alternative ways–and that the absence of a loading dose was not necessarily inherent in the disclosures.  Since the majority required inherency, the claim lacked written description support.

by Dennis Crouch

The PTAB is the busiest patent court in the country.  The Board cancels lots of patent claims, but also regularly sides with patentees in the Final Written Decision (at least as to some claims).  Post-IPR estoppel attaches to the cleared-claims and prevents the petitioner (or privies) from later challenging the validity of those claims on grounds actually raised in the IPR as well grounds the petitioner “reasonably could have raised during the inter partes review.”  35 U.S.C. 315(e).  In its 2022 Caltech decision, the Federal Circuit overturned its prior Shaw precedent and found that the estoppel broadly applies to all claims challenged in an IPR and all grounds “which reasonably could have been asserted” against the petitioned claims. California Inst. of Tech. v. Broadcom Ltd., 25 F.4th 976, 991 (Fed. Cir. 2022) (“Caltech”).

More recently, the court was faced with unique partial-institution circumstances in Click-to-Call Techs. LP v. Ingenio, Inc., 45 F.4th 1363 (Fed. Cir. 2022), and again the court concluded that broad estoppel is appropriate.  The court particularly concluded that estoppel applies even to non-instituted claims.  In a new petition for en banc rehearing, the accused infringer asks the following question:

Whether 35 U.S.C. § 315(e)’s IPR estoppel provision applies only to claims addressed in the final written decision, as consistent with the holdings in Shaw, and Intuitive Surgical, and whether that interpretation remains correct after SAS Institute, Inc. v. Iancu, 138 S. Ct. 1348 (2018).

Ingenio En Banc Petition.

You’ll recognize Click-to-Call  from the 2020 Supreme Court decision finding institution decisions ordinarily not judicially reviewable. Thryv, Inc. v. Click-to-Call Technologies, LP, 140 S. Ct. 1367 (2020) (Thryv is a DBA name of Ingenio).

The unique circumstances mentioned above are that the PTAB only partially-instituted the IPR.  In particular, the USPTO granted institution of most of the challenged claims of Click-to-Call’s US5818836 based upon the “Dezonno” reference, but denied institution with-regard-to claim 27 and also refused to conduct an IPR relying upon the Freeman reference (the only ground raised in the IPR for challenging claim 27).  All of the instituted claims were eventually found unpatentable and cancelled in the IPR–but that left claim 27 still standing.  Partial institution was eventually ruled improper by the Supreme Court in SAS Institute, Inc. v. Iancu, 138 S. Ct. 1348 (2018).  Since the IPR appeal was still pending, post-SAS remand would have been available to re-institute and pick-up claim 27, but neither party asked the PTAB to take-up that task.  The patent challenger had already won on all-but-one claim before the PTAB, and re-institution might have required a new trial on all of the challenged claims.  The patentee likely saw the PTAB as an unfriendly forum and had no interest in filing a motion to the effect of “yes, may I have another lump please.”   Note that Ingenio contests this procedual idea — arguing that dismissal order by the Supreme Court ended the case without remand.

Although I identified this as a “unique circumstance,” partial institution was a regular practice for several years prior to SAS and so there are hundreds of patents in a similar position.  Point being, this issue is goes beyond the parties here and will have a larger impact, if only transitionary post-SAS.

Once the IPR concluded, district court litigation restarted, focusing on claim 27.  The district court sided with defendant Ingenio on summary judgment. In particular, the court found claim 27 was invalid based upon Dezonno, the same prior art reference used by the PTAB to cancel all the other claims.  Click-to-call argued estoppel, but that argument was rejected.  On appeal, the Federal Circuit reversed, holding that “the district court erred in not applying IPR estoppel under 35 U.S.C. § 315(e)(2) to claim 27 based on Dezonno.”

Ingenio has argued that estoppel applies only to invalidity challenges to claims that were subject to the IPR final written decision.  The statute is not crystal clear on this point–even more so now that we recognize that partial institution was improper.  This is a case where ask about the correct statutory interpretation in a situation where we now know that the agency acted improperly.

But, lets look at the statute.  315(e) estoppel applies on a claim-by-claim basis as follows:

The petitioner in an inter partes review of a claim in a patent … that results in a final written decision … may not assert [in patent litigation] that the claim is invalid on any ground that the petitioner raised or reasonably could have raised during that inter partes review.

Id.  I have been struggling through this statute.  In my reading, the most straightforward construction sides with the Federal Circuit in this situation. It requires that estopped claims be part of the IPR and that the IPR result in a final written decision, but does not require that all of the estopped claims be adjudged in the final written decision.

The en banc petition also suggests that the court may want to reconsider its decision in Caltech.  In particular, the three-judge Caltech panel finding broad estoppel overturned the prior Federal Circuit panel decision in Shaw.  Ordinary rules of  appellate precedent prohibit one panel from overturning the decision of a prior panel. But, the Caltech judges found that the Supreme Court’s intervening decision in SAS so abrogated the reasoning of Shaw as to open the door for panel action. The petition  argues that was improper:

The Caltech panel incorrectly relied on SAS to overturn Shaw. SAS did not itself overturn Shaw, and in fact analyzed a different statute entirely (35 U.S.C. § 318(a) vs. § 315(e)(2)). Nonetheless, the Panel held that, because Shaw’s reasoning rested on the assumption that partial institutions were permissible, SAS’s elimination of partial institutions required Shaw to be overruled. But that is incorrect. Shaw was not limited to the particular facts of its case, but instead held that the plain language of § 315(e )(2) prohibits IPR estoppel from attaching to grounds that were not part of the IPR, which began at institution. SAS did not undermine or challenge the validity of Shaw as a textual construction of § 315(e)(2).

Petition.  If you recall, the Caltech decision is also still pending on petition for writ of certiorari before the Supreme Court.  Thus, this case may provide an opportunity for the Federal Circuit to preemptively speak to the high court.

Question presented in Thaler v. Vidal:

"Whether an artificial intelligence can be an inventor for purposes of patent law, which implicates the most fundamental aspects of patent law, namely, the nature of inventorship and therefore whether AI discoveries can be patented."

— Dennis Crouch (@patentlyo) September 20, 2022

If you are thinking about this from the corporate level, check out my new book chapter: Dennis Crouch, Legal Fictions and the Corporation as an Inventive Artificial Intelligence, Forthcoming in Research Handbooks on Intellectual Property and Artificial Intelligence (R. Abbott, ed). A draft version is available online here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4081569 .

by Dennis Crouch

In re Google, 22-1611 (Fed. Cir. 2022) (non-precedential order)

In Arthrex, the Supreme Court rewrote the Patent Act, charging the USPTO Director with authority to review final written decisions stemming from inter partes and post grant review proceedings (IPR/PGR).  One open question is the Director’s role in mill run patent applications that have been rejected by the Board. There are two particular actions that Directors have taken in recent decades:

1. Create a stacked rehearing panel that overturns the prior PTAB panel decision. This approach is typically taken when the Director wants to make a precedential statement on some matter of law or procedure.
2. If the party appeals, admit to the Federal Circuit that the PTAB erred and seek remand to reconsider the PTAB panel decision.  This second path is the approach taken here for Google.

Google’s pending Application No 15/487,516 claims video processing method using adaptive composite intra-prediction.  In its decision, the Board construed three key  claim limitations as “conditional limitations.”  The Board generally gives no patentable weight to conditional method steps in the anticipation/obviousness analysis.  See Ex Parte Schulhauser, 2016 WL 6277792 (PTAB Apr. 28, 2016) (precedential).  Google filed its notice of appeal to the Federal Circuit and must have had some discussions with the Solicitors office because the agency filed its request for remand even before Google filed an opening brief.

I have included the claim below, but the basic setup is that the claim includes the following stylized step:

A method comprising …

in response to a determination that a first block is available, generate a pixel value …

The basic question here is whether this claim limitation requires any action.  The PTAB read it as merely conditional — effectively replacing it with the following:

A method comprising …

if the system determines that a first block is available, then generate a pixel value …

The Board explained its position that the claims did not recite a step of “determining” that the first block exists and so it is not required by the claims.

[W]e understand the phrase “in response to a determination that” a condition exists to be equivalent to a recitation of “if” that condition exists because claim 1 does not affirmatively recite a step of determining that the first prediction pixel is available prior to reciting the step that is performed in response to—i.e., when or if—such a condition exists.

PTAB Appeal 2020-005221.  The PTAB then applied Schulhauser to conclude that “the Examiner need not present evidence of the anticipation of any of the disputed conditional method steps, because they are not required to be performed under the broadest reasonable interpretation of the method steps recited in representative independent claim 1.”  Google sought reconsideration, but the Board stuck to its original decision.

Now before the Federal Circuit, the Director has concluded that the Board erred.  Through the SG, the Director submitted an interesting reason for its determination–that a close reading of the specification shows that the first determination step is present in “every embodiment.”  Thus, according to the Director, the PTAB’s reading of the claims as conditional limitations “would be inconsistent with the specification.”

Upon review, and under the specific facts of this case, the Director acknowledges that the Board erred in designating L1 and L2 as conditional limitations governed by Schulhauser because such a reading would be inconsistent with the specification, which teaches that for every embodiment a first prediction pixel is available and a determination of whether a second prediction pixel is available is made.

Dir Motion to Dismiss.  This statement from the Director appears somewhat contrary to the usual approach of Broadest Reasonable Interpretation (BRI) which avoids importing limitations into the claims simply because they are present in each disclosed embodiment.

Google did not object to the dismissal and the Federal Circuit has ordered a remand.

One important caveat to this case — the application itself appears to still be unpublished. Thus, the information that I’m reporting comes only from public documents filed with the Federal Circuit.  I don’t have the application itself or any of the briefs filed with the PTAB.  I do have the challenged claim:

A method comprising:

generating, by a processor in response to instructions stored on a non-transitory computer readable medium, a decoded current block by decoding an encoded current block, wherein decoding the encoded current block includes adaptive composite intra-prediction, and wherein adaptive composite intra prediction includes:

in response to a determination that a first prediction pixel from a first block immediately adjacent to a first edge of the encoded current block is available for predicting a current pixel of the encoded current block:

determining whether a second prediction pixel from a second block immediately adjacent to a second edge of the encoded current block is available for predicting the current pixel, wherein the second edge is opposite the first edge; and

in response to a determination that the second prediction pixel is available, generating a prediction value for the current pixel based on at least one of the first prediction pixel or the second prediction pixel;

generating a reconstructed pixel corresponding to the current pixel based on the prediction value; and

including the reconstructed pixel in the decoded current block; and

outputting or storing the decoded current block.

Google NOA with PTAB Decisions Attached.

Schulhauser involved a heart-monitoring method claim with the following conditional step:

triggering an alarm state if the electrocardiac signal data is not within the threshold electrocardiac criteria;

Ex parte Schulhauser, APPEAL 2013-007847, 2016 WL 6277792 (Patent Tr. & App. Bd. Apr. 28, 2016) (precedential).  The panel concluded that step in conditional form can be ignored for anticipation purposes.  The Board relied particularly on the Federal Circuit’s non-precedential decision in Cyber settle, Inc. v. Nat’l Arbitration Forum, Inc., 243 Fed. Appx. 603, 607 (Fed. Cir. 2007) where the court explained that: “If the condition for performing a contingent step is not satisfied, the performance recited by the step need not be carried out in order for the claimed method to be performed.”  Id.  Schulhauser distinguished this method-claim from  system claims that include the same conditional functionality since the system “still requires structure for performing the function should the condition occur.” Id.  The MPEP now reflects this same approach:

The broadest reasonable interpretation of a method (or process) claim having contingent limitations requires only those steps that must be performed and does not include steps that are not required to be performed because the condition(s) precedent are not met. For example, assume a method claim requires step A if a first condition happens and step B if a second condition happens. If the claimed invention may be practiced without either the first or second condition happening, then neither step A or B is required by the broadest reasonable interpretation of the claim. If the claimed invention requires the first condition to occur, then the broadest reasonable interpretation of the claim requires step A. If the claimed invention requires both the first and second conditions to occur, then the broadest reasonable interpretation of the claim requires both steps A and B.

MPEP 2111.04 (II).  The Board has relied upon Schulhauser in hundreds of decisions.  However, there appears to be some ongoing debate within the PTAB about the scope of the case, with a number of cases having dissents on this point.  See., e.g., Ex Parte Botman, APPEAL 2021-004052, 2022 WL 4093710 (Patent Tr. & App. Bd. Sept. 2, 2022); Ex Parte Gopalan, APPEAL 2017-007009, 2018 WL 2386111 (Patent Tr. & App. Bd. May 21, 2018); Ex Parte Erhart, APPEAL 2019-004505, 2021 WL 195811 (Patent Tr. & App. Bd. Jan. 8, 2021).

In the appeal, Google was represented by Jonathan Tietz and Andy Dufresne (Perkins).  The per curiam order was issued by Judges Lourie, Chen, and Stark.

by Dennis Crouch

Back in June, I wrote about the Federal Circuit’s wild decision in Centripetal Networks, Inc. v. Cisco Sys., Inc., 38 F.4th 1025 (Fed. Cir. 2022). Dennis Crouch, Judicial Recusal Order Saves Cisco $2.75 Billion, Patently-O (June 23, 2022). The patentee has now petitioned the Supreme Court on a the question of whether the statute demands judicial recusal:

Whether placing stock in a blind trust satisfies 28 U.S.C. §455(f) and, if not, whether placing trivial amounts of stock in a blind trust, in lieu of selling it outright, constitutes harmless error under Liljeberg v. Health Services Acquisition Corp., 486 U.S. 847 (1988).

[Centripetal Petition]

Back in 2018, Centripetal sued Cisco, alleging infringement the claims of eleven cybersecurity related patents.  Some of the claims were cancelled in IPRs, but others made it through the gauntlet.  After staying the case for a year, the district court Judge Henry C. Morgan picked up the case again in September 2019 and began moving forward with claims from five patents no longer subject to IPR proceedings.  Both parties waived their right to a jury trial and and the district court held a 22-day bench trial over a six-week period and a subsequent damages evidentiary hearing.  In his verdict, the district court generally sided with the patentee–finding claims from four of the five patents to be valid and infringed.  The judge awarded a 10% royalty on infringing products–totaling $755 million in damages.  In addition, the court found that Cisco’s infringement was willful and enhanced the damage by 2.5x–upping the award to $1.9 billion for back damages.  The court also required a running royalty that likely takes the damages up  to above $3 billion by now.  This is apparently the largest patent-damage award in U.S. history.

On appeal, the Federal Circuit did not concern itself with the merits of the patent case.  Rather, the Federal Circuit vacated the trial determinations upon finding that the Judge Morgan should have recused himself from the entire case. To be clear, there was no indication that Judge Morgan was biased nor was there any legitimate claim questioning his impartiality–the problem is that Judge’s wife held $5k in Cisco stock.

The general rule for judicial recusal was codified by Congress in 28 U.S.C. § 455(a):

[A Judge] shall disqualify himself in any proceeding in which his impartiality might reasonably be questioned.

Id. This standards-based statute is followed by a series of express rules, including the following on ownership:

He shall also disqualify himself in the following circumstances: . . . He knows that he, individually or as a fiduciary, or his spouse … has a financial interest … in a party to the proceeding.

28 U.S.C. § 455(b).  The statute also provides an avenue for curing this situation under § 455(f).  In particular, the statute states that “disqualification is not required” in a situation where the judge has devoted “substantial judicial time” so long as the judge (or their family member) “divests himself or herself of the interest that provides the grounds for the disqualification.” 28 U.S.C. § 455(f).  The big question in this case is (1) whether Judge Morgan properly “divest[ed] himself” and, (2) if not, was it a “harmless error.”

The bench trial ended in June 2020.  On August 11, Judge Morgan was in the process of writing writing-up has decision when he became aware that his spouse owned 100 shares of Cisco stock (valued at $4,687.99).  He recognized at that point that he should get her to sell the shares immediately, but there was a big problem. Judge Morgan was just about to issue a multi-billion-dollar judgment against Cisco that would likely cause the stock to drop.  He had already decided most of the issues and written 130 pages of the eventual 167 page opinion.  It would look really bad if he sold the stock with that insider knowledge and, according to Judge Morgan “undermine the purpose of section 455.”  So, Judge Morgan instead placed the stock into a blind trust, with instructions for the trustee using his own decision-making to eventually sell of the Cisco stock.  Judge Morgan was supposed to be notified once the assets had been sold-off.  And, apparently the trust still held Cisco stock at the time he issued his final judgment.

Placing stock in a trust did shift the ownership: A trust is legally owned by the trustee rather than his spouse. However, § 455(b)/(f) are not tied to ownership, but rather a “financial interest.” And, it appears that the spouse was still the beneficiary of the trust.

On appeal, the Federal Circuit held that the statute required divestiture, and transferring title and control to a trustee was insufficient.  Rather, the stock needs to be either sold or donated — that “is the only cure envisioned under § 455(f).” Centripetal Networks, Inc. v. Cisco Sys., Inc., 38 F.4th 1025 (Fed. Cir. 2022).

Pause for a note on impartiality: A person’s ownership of Cisco stock would tend to make them want the stock price to go up. When Judge Morgan raised the issue, Centripetal announced that it had no problem with the Judge’s ownership; But Cisco complained.  Why did Cisco complain–because it had lost its pretrial motions and it  strongly suspected that the case was going to be decided in Centripetal’s favor.  I.e., its reason for complaint had nothing to do with partiality or bias, but only as a technical rule to allow it to escape from judgment and seek a new, potentially more favorable judge.

Is it Harmless Error: Despite this situation where the actual judgment goes against the apparent bias potential, the Federal Circuit rejected the argument of “harmless error.”  In a number of prior cases, the courts have held that “mandatory recusal does not require mandatory vacatur.”  In other words, the fact that a judge was supposed to have recused himself under the rule does not automatically require an appellate court to vacate that judgment.  The basic harmless error test outlined by the Supreme Court in Liljeberg asks three questions:

1. What is the risk of injustice to the parties in the particular case?
2. What is the risk that denial of relief here will result in injustice in other cases?
3. What is the risk that denial of relief here will undermine the public’s confidence in the judicial process?

In reviewing these elements, the Federal Circuit found them to all favor vacatur in this situation.  To be clear though, the Federal Circuit’s section explaining risk of injustice to the parties reads like six pages of total BS.  The only line that sticks out as interesting is the court’s speculation that in a case like this “there is a substantial risk that [the Judge] might bend over backwards to rule against that party to try to prove that there is no bias.” Id.

In its petition for writ of certiorari, Centripetal argues that the Federal Circuit “fundamentally misconstrued and misapplied both §455(f) and harmless-error doctrine” with the result of inflicting harsh consequences against the patentee.  Particular concerns:

1. The court focused on the word “divest” in § 455(f) and giving it a powerful meaning while “ignoring the specific context addressed” by the statute.  The court here suggests that what Judge Morgan should-have-done is sell-off the stock and create the exact appearance of impropriety that the statute endeavors to avoid.  “That is plainly not what Congress intended by expressly providing an option for judges to avoid recusal when minor financial interests are discovered after substantial judicial resources have been invested in a case.”
2. Regarding harmless error, we can recognize that both parties agreed that the judge was not biased or partial nor did he operate with any malice or indifference.  Centripetal conjects that the Federal Circuit’s “zero-tolerance posture” was “motivated by recent public scrutiny of judicial stock ownership” rather than the specifics of this case.

[Petition].  Before the Federal Circuit, Cisco had argued that the statute demands vacatur of all judicial decisions starting from the August 11 date when Judge Morgan learned of his disqualifying financial interest. “Any other outcome would send the undesirable message that this Court does not take the disqualification statute seriously and would undermine public confidence in the judiciary.”

= = =

The Federal Circuit’s decision was authored by Judge Dyk and joined by Judges Taranto and Cunningham. Former US Solicitor General and Scalia Clerk Paul Clement is representing the patentee in the petition. Paul Andre (Kramer Levin) handled both the trial and the appeal is also on the petition.  Bill Lee (Wilmer) represented Cisco before the district court and the Federal Circuit and is likely to handle the Supreme Court briefing if a brief is requested by the Court.

by Dennis Crouch

Sawstop v. Vidal, — F.4th — (Fed. Cir. 2022)

This is a Patent Term Adjustment (PTA) case.  Back before 1995, patents were awarded a term of 17-years following issuance.  As part of an international negotiation and with some anti-submarine policy justifications, the US switched over to a 20-year term that starts counting days as of the application’s filing date.  The 20-year term from filing date was seen as roughly equivalent to 17-years from issuance since most patent applications take about three years to issue.  In the end, this kept overall patent term roughly the same. But, it also allowed for some major variability. Some patent applicants were concerned that delays by the USPTO due to backlogs or unreasonable examination practices might effectively cut into the patent term.  Congress heard those concerns and created the PTA system that adds extra days to the patent term when those delays exist. 35 U.S.C. § 154(b). Unfortunately, the PTA statute is not a model of clarity and has been the subject of numerous court battles over various quirky elements.

The provision at issue in Sawstop has to do with PTA added for PTAB appeals and district court challenges. Under the law, added term is triggered by the following:

(iii) appellate review by the [PTAB] or by a Federal court in a case in which the patent was issued under a decision in the review reversing an adverse determination of patentability,

35 U.S.C. § 154(b)(1)(C)(iii).

I have written previously about Sawstop technology that covers a safety feature for power saws “that instantly stops the saw blade upon contact with flesh.”  The appeal here is a consolidated appeal regarding two Sawstop patents.  Sawstop is struggling because its patents are all expiring and competitors are hot-on-its-heels.

In the first case (‘476), the applicant had appealed an obviousness rejection by the examiner.  The PTAB rejected the examiner’s decision, but simultaneously issued a new ground of rejection on obviousness. Thus, the ultimate conclusion was that the rejection was affirmed, albeit for different reasons.  On remand, the claims eventually issued, but only after an RCE (which cut-off further PTA).  After the patent issued, the owner asked for PTA associated with the PTAB appeal, but the PTO refused as did the district court. 

In the appeal, the Federal Circuit has also sided against Sawstop.  The rule quoted above is triggered by “a decision in the review reversing an adverse determination of patentability.” On appeal, the Federal Circuit found no reversal by the PTAB — although it did “cast aside the examiner’s basis for rejecting claim 11,” the Board simultaneously established reasoning for rejecting that same claim.  The result was that the claim remained rejected.  The Federal Circuit here focused on the substantive outcome — the patentee’s appeal was not successful and therefore there was no reversal.

The Federal Circuit also addressed a second element of the test — that the “patent was issued under a decision in the [appellate] review.”  The court has interpreted that requirement as mandating that at least one claim considered by the PTAB have found its way into the issued patent without significant or substantive post-appeal amendments. Here, the patentee had amended the claims after the PTAB decision in order to achieve allowance.  Thus, no Type-C PTA.

The second case (‘796) is a bit more complicated, the applicant had appealed rejections of claims 1 and 2 for anticipation a non-statutory double patenting (provisional).  The Board issued a partial-reversal–holding that claim 1 was no good, but claim 2 was patentable. But Sawstop really wanted claim 1 and so filed a civil action with the DC District Court–and won that case.  The one caveat is that the civil action focused only on anticipation and did not address non-statutory double patenting.  Back at the Office, the PTAB maintained its double-patenting rejection and Sawstop was given a choice: (1) drop claim 1 or (2) file a terminal disclaimer.  Sawstop chose the latter, and cancelled claim 1.  But again, filed an RCE with several further amendments to other claims and the specification before eventual allowance.

The USPTO awarded PTA for the successful PTAB appeal, but refused any PTA for delay in getting the district court decision. On appeal, the Federal Circuit affirmed this analysis again on two grounds.  First, although the district court did reverse the anticipation claim, it did not decide that claim 1 was “patentable.”  As such, the court did not “revers[e] an adverse determination of patentability.”  Sawstop had argued that – at the time – the OTDP rejection was only provisional and thus should not count.  The Federal Circuit might have bought this argument if Sawstop also explained that it was somehow barred from including the issue in its Section 145 Civil Action, but Sawstop did not provide such an explanation.  The second reason for agreeing with the PTO is that Claim 1 (the Subject of the District Court civil action) never issued in the patent.  And, as discussed above, the court has interpreted the rule to only apply when the claim at issue in the ‘appeal’ ends up in the issued patent.  Thus, no Type-C PTA for this one either.

Throughout this process, the Federal Circuit repeatedly justified its ruling as a plain language interpretation of the rule.  I’ll note, however, that the court does not do so with the “appellate review” portion of the statute.  A Section 145 civil action is not an appellate review of the PTAB decision, but rather separate action.  See Kappos v. Hyatt, 566 U.S. 431 (2012). But, the court appears to have no trouble seeing it as an appeal in this context.

= = =

Judge Linn wrote the decision here joined by Judges Newman and Chen.  Sawstop was represented by Jared Newton and his team from Quinn Emanuel as well as David Fanning, SawStop’s inhouse counsel. Hugham Chan the E.D. Va. US Attorney’s Office argued for the Gov’t.  Chan & Newton both graduated from GWU Law in 2010 & 2011 respectively. Chan is a former clerk of Judge Chen; Newton is a former patent examiner.

Hologic v. Minerva, 19-2054 (Fed. Cir. 2022) (en banc petition)

The Supreme Court’s 2021 decision retained assignor estoppel as a cognizable patent law doctrine that can bar a former patent owner from later challenging the patent’s validity.  Although the doctrine was retained, the Court also concluded that the Federal Circuit had applied the doctrine “too expansively.” Like other estoppels, assignor estoppel is equitable in nature and so courts should consider any “unfairness in its assertion” and particularly only apply the doctrine when the assignor has made either “explicit nor implicit representations in conflict with an invalidity defense.” Minerva Surgical, Inc. v. Hologic, Inc., 210 L. Ed. 2d 689 (June 29, 2021).

On remand, the Federal Circuit found Minerva still estopped, despite the narrowed doctrinal scope.  Hologic, Inc. v. Minerva Surgical, Inc., 44 F.4th 1358 (Fed. Cir. 2022); Dennis Crouch, Narrowed Assignor Estoppel Still Bars Minerva Surgical from Challenging Validity, Patently-O (Aug 15, 2022).    Now, Minerva has petitioned the Federal Circuit for an en banc rehearing on the following question:

For purposes of determining whether assignor estoppel applies, how should a court determine an inventor’s representation of patent scope in a never-issued claim in the original patent application?

Minerva Petition.

The basic setup here is that back in the 1990s the inventor (Csaba Truckai) filed his patent application for an endometrial ablation tool that operates as a “moisture transport system.”  The embodiments in the application all use a “moisture permeable” applicator head. The application also the disclosed prior art moisture impermeable approach, disparaged that approach that led to a “problem of steam and liquid buildup at the ablation site.”  The patent document is clearly directed toward the permeable portion as a key aspect of the invention.  But, one claim in the original application (Claim 31) was directed to other aspects of the invention and did not expressly require use of the permeable member.  And, although Truckai argued that the specification’s description should limit the claim scope to inherently include the permeability clause, the Federal Circuit disagreed and instead concluded that “nothing in the intrinsic record supports limiting claim 31” to require a moisture permeable material.  Truckai also prosecuted claim 31 for some time – overcome an anticipation rejection.   Eventually though the claim was dropped by the patentee following a restriction requirement.

The Truckai application was filed in 1998 and that same year Truckai signed an assignment of rights to start-up company Novacept.  Later Novacept was purchased for $300m by Hologic.  In this process, Truckai filed another assignment in 2004 warranting it had “no present knowledge from which it could reasonably conclude” that the assigned IP rights were invalid or unenforceable.”  By that time though, claim 31 had been withdrawn following a restriction requirement.

Now at his next startup (Minerva), Truckai conceived of another device.  The new device relied upon a moisture impermeable head.  Truckai apparently showed the device to Hologic as part of a potential sales agreement.  Rather than buying, Hologic filed a continuation of that 1998 application with new scope expanded out to cover Truckai’s new impermeable membrane.

Once the patent issued, Hologic sued Minerva for patent infringement.  Minerva argued that the new claims were invalid for lack of enablement and written description — that the expanded claim scope improperly covered moisture impermeable heads that were not disclosed in the original specification; Hologic argues that Minerva’s invalidity challenge is barred by assignor estoppel.

The Federal Circuit initially sided with the patentee, but that judgment was vacated by the Supreme Court.  On remand, the Federal Circuit considered the case history and the Supreme Court guidance, and eventually maintained with its original conclusion barring the invalidity defenses under the doctrine of assignor estoppel.  After construing that original Claim 31, the court concluded that it was broad enough to encompass both permeable and impermeable devices; and therefore that Truckai’s actions at least an implicit representation that he was transferring patent rights covering impermeable devices.  Ergo, assignor estoppel prohibits him from now claiming that such expansive claims are invalid.

The new petition for rehearing argues that the Federal Circuit got this all-wrong by applying the principles of claim construction as it did.

The difference between a primary focus on the inventor/assignor’s representations (as the Supreme Court directed) and the mechanics of claim construction (as the Panel purported to apply) is more than semantic.

With claim construction, we know that the use of “comprising” creates an open-ended scope.  Thus, the absence of the permeable clause in claim 31 allows or claim scope that includes impermeable heads.  But, this quirk of claim construction is different from concluding that Truckai had made affirmative statements that his invention covers impermeable heads.  The defendant argues, “silence is not a representation.”  Of course, the Supreme Court made clear that the representations can be “implicit.”

In the lawsuit, Truckai wants to argue that the claims covering impermeable heads are invalid for failure of disclosure.  Assignor estoppel will apply if Truckai made some representation that his assigned patent rights cover impermeable heads.  In its petition, Truckai argues that the court should take special precaution in judging  claims submitted in a patent application that never-issued.  Because those original applied-for claims are “inchoate rights of indefinite scope,” the court should take  extra measures to read meaning from the specification:

But if the task is to  determine what the inventor represented in a never-issued claim, then the fact that the specification declares moisture trapped at the ablation site as the problem the invention solves represents that the device could not work if it included moisture impermeable material. An inventor did not represent that his invention is something his own description of the invention makes clear would not work.

Id.

The general rule for en banc petitions is that no response need be filed unless requested by the Court.  However, I expect that such a request will be forthcoming on this interesting issue.

by Dennis Crouch

The level of estoppel was a sticking point a decade ago in the lead-up to the America Invents Act (AIA).  Patentee’s were concerned with repeat harassment by accused infringers; Accused infringers were concerned that IPRs would be ineffective and that they would then lose their right to challenge the patent in court. We now know that IPRs are very effective, but sometimes patent challengers would still like a second bite.  That is the situation in the pending petition for certiorari in Apple Inc. v. California Institute of Technology, 22-203 (U.S.).  Apple lost its IPRs against CalTech and now wants to bring additional invalidity challenges in court that it could have raised in its IPR petitions.

In the 2011 AIA legislation, Congress created two forms of petitioner estoppel, each of which attach only if the IPR petition is granted and the PTAB reaches a Final Written Decision, and only with respect the claims actually adjudged in the IPR final written decision.

315(e)(1) Estoppel of Further Proceedings Before the USPTO: The petitioner [or privy/RPI] may not request or maintain a proceeding before the Office with respect to that claim on any ground that the petitioner raised or reasonably could have raised during that inter partes review.

315(e)(2) Estoppel of Invalidity Arguments in Court: The petitioner [or privy/RPI] may not assert either … that the claim is invalid on any ground that the petitioner raised or reasonably could have raised during that inter partes review.

35 U.S.C. 315(e). As you can see, both provisions create estoppel barring a petitioner from later raising “any ground that the petitioner raised or reasonably could have raised during that inter partes review.”  The question in

Apple v. CalTech: CalTech sued Apple and Broadcom for infringing its patents covering wireless communication chip logic. Apple responded with several IPR petitions. The petitions were granted initiating the IPR, but the USPTO eventually sided with patentee in its final written description. Apple is thus prohibited from re-challenging the validity of the already challenged claims “any ground that the petitioner raised or reasonably could have raised during that inter partes review.”

Back in district court, Apple attempted to raise additional invalidity challenges, but the district court barred those defenses. The district court concluded that Apple could reasonably could have included those as grounds in its IPR petition and thus barred under 315(e)(2). CalTech eventually won a $1 billion jury verdict.  On appeal, the Federal Circuit vacated the verdict on other grounds, but affirmed on the issue of estoppel. The Federal Circuit particularly held that estoppel extends to “all grounds not stated in the petition but which reasonably could have been asserted against the claims included in the [IPR] petition.” California Inst. of Tech. v. Broadcom Ltd., 25 F.4th 976 (Fed. Cir. 2022), overruling Shaw Industries Group, Inc. v. Automated Creel Sys., Inc., 817 F.3d 1293 (Fed. Cir. 2016).

Apple’s new petition to the Supreme Court identifies what it sees as a fundamental flaw in the Federal Circuit’s textual analysis. The Federal Circuit concluded that Apple was estopped from later raising grounds that it reasonably could have raised in the inter partes review petition. But, the statute as written creates a slightly different test — asking what “reasonably could have raised during that inter partes review.” The following is Apple’s question presented to the Supreme Court:

Whether the Federal Circuit erroneously extended IPR estoppel under 35 U.S.C. § 315(e)(2) to all grounds that reasonably could have been raised in the petition filed before an inter partes review is instituted, even though the text of the statute applies estoppel only to grounds that “reasonably could have [been] raised during that inter partes review.”

Apple question presented.

The IPR process begins with an IPR petition. 35 U.S.C. § 311. The statute charges the USPTO Director the role of instituting the IPR, although that duty had been delegated tot he PTAB. 35 U.S.C. § 314.  After institution, the PTAB conducts a trial and issues its final written description.  Apple’s basic argument is that the statute’s use of “during” excludes potential actions that could have taken place prior to institution.  To wit, Apple recognizes that it reasonably could have made additional arguments in its IPR filings, but those took place prior to the IPR; And, once the IPR petition is granted, the challenger no longer has any opportunity to raise any other grounds to challenge the patents.  In this scenario, the statute is practically limited to cover only grounds actually raised.  Apple explains:

When “an inter partes review of a claim in a patent … results in a final written decision,” a petitioner is estopped from asserting in district court litigation only “any ground that the petitioner raised or reasonably could have raised during that inter partes review.” 35 U.S.C. § 315(e)(2) (emphasis added). By expanding statutory estoppel to cover any ground that could have been raised in a petition before an inter partes review even begins, the Federal Circuit effectively rewrote the statute. This Court’s review is urgently needed to restore Section 315(e) to its proper scope.

Apple Petition.

Apple’s petition relies heavily on the work of Cleveland-Marshall Prof. Christa Laser and her 2018 Florida Law Review article.  Christa Laser, The Scope of IPR Estoppel: A Statutory, Historical, and Normative Analysis, 70 Fla. L. Rev. 1127 (2018).  Laser suggested that the Supreme Court reinterpret the statute and “should hold that IPR estoppel applies narrowly only to grounds actually raised and instituted.” Id.  Baltimore Prof. Greg Dolin penned a responsive article concluding that Laser’s approach is too narrow. Gregory Dolin, Estopping Patent Harassment: A Response to Christa J. Laser, 70 Fla. L. Rev. F. 136 (2019). From my view, Dolin and the Federal Circuit have the better view on this one over Apple and Laser.  We’ll see this fall whether the Supreme Court agrees.

by Dennis Crouch

Inaction is generally defined as “the lack of action.”  The two terms are antonyms.  Still, courts regularly find that–at times–inaction constitutes action.  Most of us learned in criminal law about crimes of omission where the actus reus is the “failure to perform a legal duty even when one has the capacity to do so.”  The Federal Circuit performed this analysis in its recent decision in Gilbert Hyatt v. USPTO, 21-2324, — F.4th — (Fed. Cir. Sept 8, 2022).

Hyatt’s microprocessor patent applications have been pending since the 1990s and claim priority back much earlier filings.  All this was prior to the 1995 implementation of the Uruguay Round Agreements Act that started measuring patent term from the filing date. Under the old rules that govern Hyatt’s applications, a patent is given force for 17 years following issuance.  For the most part, these applications are also unpublished and the file histories are kept secret.

Hyatt’s appeal here relates to his pending U.S. Patent Application No. 08/435,938.  Back in 2013 the application included about 200 pending claims. However, the USPTO forced Hyatt to select eight of those claims for examination as part of the PTO’s Hyatt-Management-Efforts.  The examiner then rejected all eight of those claims.  Hyatt then significantly amended the pending claims.  According to the Federal Circuit, the amendments replaced all of the words of one of the claims except for the “A” starting the claim and the word “comprising.”  The examiner then responded with a restriction requirement.

For most patent applicants, restriction requirements are just slightly annoying and add some cost to the clients. But, for the most part the restriction/divisional don’t substantially reduce the rights. And, in fact at times they can be beneficial because they avoid obviousness-type double patenting problems.

For Hyatt, the restriction requirement was very problematic.  If he filed a divisional application to the remaining claims, he would lose his pre-GATT status.  The result for the divisional application would be zero-days of patent term.  Hyatt filed a civil action that the PTO’s restriction requirement violated the Administrative Procedure Act (APA).  In particular, Hyatt pointed to 37 C.F.R. § 1.129 as prohibiting the restriction practice. The district court sided with the USPTO as has the Federal Circuit on appeal.

Section 129(b) generally prohibits the USPTO from issuing a restriction requirement in applications that has a priority claim from at least three years before the the June 1995 cut-off.  Hyatt’s case meets that three year timeline.  The provision also includes the following exception: A restriction requirement is allowed if: “(ii) The examiner has not made a requirement for restriction in the present or parent application prior to April 8, 1995, due to actions by the applicant.”   The fight here is on the question of whether the PTO’s failure to issue a restriction requirement prior to April 1995 was “due to actions by the applicant.”

Due to Action by the Applicant: The USPTO’s explains that it didn’t issue a restriction requirement pre-1995 because, back then the application’s claims were all directed to the same invention.  Hyatt’s amendments 20-years later necessitated the restriction. In this scenario, the USPTO explained that Hyatt’s pre-1995 action was that he “failed to disclose claims to a separate invention”  pre-1995.

On appeal, the Federal Circuit agreed that Hyatt’s pre-1995 actions could be characterized as the action of “withholding” the future claims, and it is easy imagine the causation chain–the PTO’s failure to issue a restriction pre-1995 was due to Hyatt’s withholding of the future claims.  Since around 2001, the MPEP explanation of how the rule works included an example that appears to exactly match Hyatt’s situation:

Examples of what constitute “actions by the applicant under 37 CFR 1.129(b) are: … (C) applicant disclosed a plurality of independent and distinct inventions in the present or parent application, but delayed presenting claims to more than one of the disclosed independent and distinct inventions in the present or parent application such that no restriction requirement could be made prior to April 8, 1995.

MPEP 803.03 (8th Ed, 2001).  Here, Hyatt disclosed several different inventions “delayed presenting claims” until much later.  The court notes, as it always does, that the MPEP does not carry force of law. Still, the manual does at least put everyone on notice.

Affirmed — the restriction requirement stays.

by Dennis Crouch

The Federal Circuit’s recent decision in Arendi S.A.R.L. v. LG Electronics Inc., 21-1967, — F.4th — (Fed. Cir. Sept 7, 2022), upholds Judge Stark’s discovery penalty against the patentee Arendi with the result that the trial moves forward against only one product of the hundreds that were accused.

Arendi sued LG in 2012 for infringing its US7917843.  As required by Delaware local rules, Arendi provided a set of infringement contentions to LG.  The local rules require the patentee to first provide a list of accused products and patents alleged being infringed.  Subsequently, the patentee must disclose a set of initial claim charts mapping showing how each accused product maps to each asserted claim.

In Delaware, these local rules are seen as discovery requirements that operate alongside other mandatory discovery found in FRCP 26.  See D. Del. Default Standard for Discovery § 4.  The local rule states that the required disclosures here are “initial” and that the parties “shall be permitted to supplement.”

In its disclosures, Arendi first provided a list of 200+ infringing products.  But then when it came time to provide the claim charts, Arendi addressed only one product (the LG Rebel 4) and labelled it as “exemplary.”  Later, Arendi and LG agreed on
eight representative products to stand-in for all accused products. The Rebel 4 was one of these eight. Despite the agreement on exemplary models, LG twice warned Arendi that it had not yet provided the claim charts as required by the local discovery rules. Arendi did not respond to those warnings.

Following the close of fact discovery, Arendi provided an expert report that spelled out its infringement contentions for five additional representative models. At that point, LG moved to strike those portions of the expert report based upon Arendi’s failure to timely disclose its infringement contentions as required by the local rule. Judge Stark was in charge (during his tenure as a district court judge) and granted the motion — holding that Arendi had “failed to fulfil its discovery obligations” and thus had lost its right to provide such disclosures at that point.  In particular, the district court also noted the prejudicial impact since: “LG understood Arendi was accusing only the Rebel 4.”

At that point, Arendi simply filed a new lawsuit, still in the District of Delaware specifically alleging infringement of the excluded products from the expert report.  That new case was also routed to Judge Stark who dismissed it with prejudice under the duplicative-litigation doctrine. On appeal, the Federal Circuit has affirmed.

Improper duplicative litigation typically involves a plaintiff bringing two different lawsuits that align along four axes: (1) same claims (2) against the same defendant (2) in the same court (4) at the same time. Walton v. Eaton Corp., 563 F.2d 66 (3d Cir. 1977) (en banc).  In that scenario, the district court is given discretion to dismiss one of the cases as part of its “general power to administer its docket.”

On appeal, Arendi argued that its two cases did not involve the same claims once the district court granted LG’s motion to strike. In general, a patentee’s infringement assertion against one product is a different claim than its assertion of the same patent against a different product.  On appeal though the Federal Circuit found that Arendi had mischaractered the district court’s order.

The court did not grant LG’s motion to strike parts of Arendi’s infringement expert report because Arendi failed to sufficiently accuse the non-Rebel 4 products. The court granted the motion to strike because Arendi “failed to fulfill its discovery obligations” with respect to those products, so Arendi’s infringement allegations in its expert report were procedurally untimely.

Slip Op.  The distinction here is technical, but that is civil procedure.  In Arendi I, the district court did not did not dismiss Arendi’s claims against non-Rebel products. Rather, the court but rather excluded the evidence necessary for Arendi to actually win on those claims.  The difference is important because it meant that the same claims were still pending in both cases at the same time. “Left with the simple and obvious fact that the non-Rebel 4 products accused in Arendi II are identical to products accused in Arendi I, we determine that the district court did not err in dismissing the Arendi II complaint as improperly duplicative.” Id.

The Arendi I court indicated that it will consider allowing the amendment depending upon the outcome of this appeal. Of course, by now Judge Stark has moved to the Federal Circuit and the case is currently not assigned to any judge.

= = = =

Opinion by Judge Prost
Judges: Prost, Chen, Stoll
Kemper Diehl represented the plaintiff on appeal along with his team from Susman Godfrey; Andrew Schwentker from Fish & Richardson argued for the defendant.