Since 2015, when the U.S. Court of Appeals for the Federal Circuit started to issue decisions in appeals from the Patent Trial and Appeal Board, the affirmance rate of the PTAB's decisions has hovered around 70% to 80%.
The rate of the court reversing or vacating in full or in part a PTAB decision has remained low. Through the end of 2023, the Federal Circuit affirmed every issue in about 74% of PTAB appeals, including inter partes review, covered business method and post-grant review appeals, but not ex parte appeals, while the reversals and vacaturs, including partial affirmance decisions, amounted to fewer than 23%.1
Appellants were even less successful in 2024.2 We identified 219 Federal Circuit decisions for PTAB appeals: 184 (84%) were affirmed in full, 12 (5.5%) were reversed, at least in part, and 23 (10.5%) were vacated, at least in part, and remanded. This article reviews those decisions and identifies the types of issues that were successfully appealed.
Methodology
Using Docket Navigator, we identified 368 PTAB appeals, including ex parte appeals, terminated in 2024. Only 219 of these appeals were litigated to decision because 149 of them were dismissed, usually voluntarily. Of the 219 appeals decided, 184 were fully affirmed,3 12 were reversed, at least in part, and 23 were vacated, at least in part.
This does not mean, however, that the court issued 219 opinions in PTAB appeals. Many of these appeals were consolidated. So, while the court reversed, at least in part, in 12 appeals, it issued just six opinions because of consolidation.
For example, the written opinion in Cytiva BioProcess R&D AB v. JSR Corp., involved 10 appeals from six final written decisions addressing dozens of claims in three patents.4
The court affirmed most of the PTAB decisions on appeal but reversed on four claims of two patents addressed in four appeals. In our win-loss calculation, we counted the 10-appeal opinion in Cytiva as six affirmances and four reversals. Other consolidations were similarly counted.5
Also, in successful appeals, from the appellants' point of view, the appellant sometimes wins on more than one issue.
For instance, in Netflix Inc. v. Avago Technologies International Sales Private Ltd.,6 the court faulted two aspects of the PTAB's decision: The PTAB misconstrued the claims and failed to consider an argument the patent challenger properly raised.
We counted Netflix once in our challenger win column, but it supplies two data points for our issue‑identification tallies — one for the claim construction issue and one for whether a party raised a new theory. As a result, the number of win-loss judgments (19 total) is less than the number of issues identified (24 total).
Outcome Breakdown
Overall, there were eight Federal Circuit decisions in patentees' favor, and 11 decisions in challengers' favor.
Issue Breakdown
Observations
Successful appellants raise fewer issues.
In most of the successful appeals, the appellant raised, and ultimately the court addressed, fewer than three issues.
Indeed, in about 63% of the appellant-win opinions analyzed here, including for the consolidated appeals, the court focused its discussion on just one or two key issues.
The maximum number of issues the court addressed is four, and only in two opinions. These results confirm the wisdom that cases should not be retried on appeal. Pick your best one or two issues and focus on those. Rarely will the fifth-best argument win if the first four did not.
There appears to be no pro- or anti-patentee bias for PTAB appeals.
We found a total of 19 opinions for the 35 appellant victories we identified.7 In these 19 opinions, the patentee prevailed eight times on least one issue (42%), and the patent challenger prevailed 11 times on at least one issue (57%).
Because it is a small sample size, we do not believe the lower patentee success record should be viewed as evidence of an anti-patentee bias. Flipping a coin 19 times will result in eight or fewer heads about a third of the time.8
Appellants in 2024 succeeded more often on high-deference issues than on de-novo-review issues.
One surprising trend we saw was that more appellants won on issues that are subject to deferential review than where review is without deference.9
For example, in one appeal, the appellant won by arguing that the PTAB refused to consider its reply brief arguments that addressed the patentee's new claim-construction arguments.10 The court reviewed this procedural argument for an abuse of discretion, one of the more deferential appellate review standards.11
Also, 12 decisions determined that the PTAB failed to support its findings with substantial evidence, another deferential standard.
These 12 substantial-evidence cases fell into three subcategories: (1) whether every claim limitation was in the prior art (seven appeals total, four for the patentee,12 three for the challenger);13 (2) whether there was a motivation to combine or modify references (four appeals, one for the patentee,14 three for the challenger),15 or (3) whether the specification supported the claim (one decision in the patentee's favor).16
On issues reviewed without deference, however, appellants succeeded only 11 times. Claim construction errors accounted for only seven successful appeals (two for the patentee,17 five for the patent challenger).18
Statutory construction and application of law issues accounted for three appellate victories (one for the patentee19 and two for the patent challenger).20
In addition, in Netflix, the court reviewed de novo whether the patent challenger improperly raised a new theory or argument in the petitioner reply brief, finding in favor of the patent challenger.21
Conclusion
Strategically focusing on fewer, stronger issues continues to be a winning appeal strategy.
And, although identifying issues that will be reviewed without deference is also a time-honored appellate strategy, 2024's results demonstrate that appellants should also scrutinize the PTAB's factual findings and procedural decisions. That includes, for instance, whether the PTAB failed to address arguments raised below or provided adequate reasoning for its conclusion.
Obviously, the likelihood of prevailing in an appeal cannot be predicted based on the success of other appeals having similar issues, but the aggregate outcomes summarized here might help parties decide whether to appeal a close case and, if so, decide which issue to take to appeal.
Originally published by Law360.
Footnotes
3. This includes 87 appeals affirmed with a written opinion and 97 summary affirmances under Rule 36.
4. Cytiva BioProcess R&D AB v. JSR Corp.
, 122 F.4th 876 (Fed. Cir. 2024) (reversals in Case Nos.
23-2239, 23-2252, 23-2253, 23-2255).
5. This methodology differs from other analyses of the Federal Circuit's PTAB appeals for 2024. For instance, in their recent study of 2024 appeals, our friends at Perkins Coie reported a total of 153 PTAB appeals. See Dan Bagatell, "Fed. Circ. Patent Decisions In 2024: An Empirical Review" Law 360 (Jan. 8, 2024) (https://www.law360.com/articles/2280249/fed-circ-patent-decisions-in-2024-an-empirical-review). Unlike us, they counted an opinion resolving multiple appeals as a single decision. Id. So although our numbers differ, our results are consistent.
6. Netflix Inc. v. Avago Techs. Int'l Sales Pte.
Ltd. , Nos. 22-1936, 2024 WL 3507601 (Fed. Cir. July 23,
2024).
7. We counted reverse-in-part or vacate-in-part judgments as appellant victories. We recognize that sometimes the part of the decision below that was affirmed can be more important than the part(s) reversed. But we discuss only the numbers here, not the merits or significance of the outcomes.
8. See https://www.omnicalculator.com/statistics/coin-flip-probability.
9. Our analysis here does not extend to affirmances, so we do not have any data on the relative rates of success for deferential- and de-novo-review issues.
10. Apple Inc. v. Omni MedSci Inc. , 2024 WL 3084509, at *6 (Fed. Cir. June 21,
2024).
11. Id. at *3 ("We review for abuse of discretion the Board's decision to disregard a reply argument as exceeding the proper scope.").
12. Paul Hartmann AG v. Attends Healthcare Prods.
Inc. , No. 2022-1724, 2024 WL 911880 (Fed. Cir. Mar. 4, 2024);
Pfizer Inc. v. Sanofi Pasteur Inc. , 94 F.4th 1341, 1353 (Fed. Cir. 2024); Koninklijke
Philips N.V. v. Quectel Wireless Sols. Co. , No. 2023-1896, 2024 WL 3983189, at *3-4
(Fed. Cir. Aug. 29, 2024); Regents of Univ. of California v.
Satco Prods. Inc. , No. 2023-1356, 2024 WL 4972639, at *4
(Fed. Cir. Dec. 4, 2024).
13. Tennant Co. v. Oxygenator Water Techs. Inc.
, No. 2022-2304, 2024 WL 3507694 (Fed. Cir. July 23, 2024)
(holding the Board failed to address an inherent-anticipation
argument); ZyXEL Commc'ns Corp. v. UNM Rainforest
Innovations , 107 F.4th 1368, 1384 (Fed. Cir. 2024); Weber Inc. v.
Provisur Techs. Inc. , 92 F.4th 1059, 1072 (Fed. Cir. 2024).
14. Virtek Vision Int'l ULC v. Assembly Guidance
Sys. Inc. , 97 F.4th 882, 888 (Fed. Cir. 2024).
15. Cisco Sys., Inc. v. K.Mizra LLC , No. 2022-2290, 2024 WL 3841809, at *4 & *6 (Fed.
Cir. Aug. 16, 2024); Cytiva, 122 F.4th at 890-91; Palo Alto
Networks, Inc. v. Centripetal Networks LLC , 122 F.4th 1378 (Fed. Cir. 2024).
16. Rai Strategic Holdings Inc. v. Philip Morris
Prods. S.A. , 92 F.4th 1085, 1092 (Fed. Cir. 2024).
17. CoolIT Sys. Inc. v. Vidal , No. 2022-1221, 2024 WL 981364, at *6 (Fed. Cir.
Mar. 7, 2024); IOENGINE LLC v. Ingenico Inc. , 100 F.4th 1395, 1404–05 (Fed. Cir. 2024) (finding
that the Board improperly gave certain claim limitations no
patentable weight).
18. Google LLC v. EcoFactor Inc. , 92 F.4th 1049, 1057-58 (Fed. Cir. 2024); Roku Inc.
v. Universal Elecs. Inc. , No. 2023-1019, 2024 WL 3042701, at *5 (Fed. Cir. June
18, 2024); Cytiva, 122 F.4th at 887-88; Netflix, 2024 WL 3507601,
at *5; Weber, 92 F.4th at 1070.
19. IOENGINE, 100 F.4th at 1404-05 (holding that the Board misapplied "printed matter" doctrine in construing the claim terms at issue).
20. Weber, 92 F.4th at 1067 (addressing whether product
manuals at issue constituted "printed publications" under
§102); LKQ Corp. v. GM Glob. Tech. Operations LLC
, 102 F.4th 1280 (Fed. Cir. 2024) (addressing obviousness
standard for design patents).
21. Netflix, 2024 WL 3507601, at *7 & *7 n.10 (distinguishing the review standard for newness and that for responsiveness).
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
