Challenges of Proving Inventorship: Corroboration of All Inventive Facts

by Dennis Crouch

Medtronic, Inc. v. Teleflex Innovations S.A.R.L., — F.4th — (Fed. Cir. 2023)

The case involving Medtronic and Teleflex centered on five patents related to a coaxial guide catheter used in interventional cardiology procedures. These patents, US Patents 8,048,032, RE45,380, RE45,776, RE45,760, and RE47,379, cover inventions devised to offer an “enhanced backup support” in contrast to using a guide catheter individually.

In an attempt to invalidate these patents, Medtronic launched a succession of Inter Partes Review (IPR) petitions. Despite the initial approval of the petitions, the Board eventually ruled in favor of Teleflex, the patent holder. The pivotal question on appeal was whether U.S. Patent 7,736,355 (“Itou”) could be regarded as prior art.

In a split decision, the Federal Circuit confirmed the Board’s decision, establishing that the first-to-invent law from before the America Invents Act (AIA) allowed Teleflex to antedate the prior art by demonstrating prior invention. The case underlines a situation where the transition to the AIA has resulted in the loss of certain rights for patent holders. Under the AIA’s provisions, Itou would have been considered prior art, therefore preventing Teleflex’s patenting. The also case stands out as Itou was clearly the first inventor, but this evidence was not taken into account due both to arcane pre-AIA rules and due to the IPR process focusing only on the published patent documents.

Teleflex’s patents claim a priority filing date of May 3, 2006, based on a purported conception date in early 2005. Itou’s priority application was filed in Japan in September 2004, and a year later, the US application was filed on September 23, 2005. Despite Itou pioneering the invention in Japan and filing the Japanese patent application prior to Teleflex’s invention, Teleflex arguably demonstrated an invention date that preceded Itou’s US filing date, which was critical for the court’s decision.

A crucial regulation in this context is pre-AIA 102(e), which introduces a form of ‘secret prior art’ for patent applications that are unpublished as of the patentee’s invention date. Under this regulation, a third party’s US or PCT patent application becomes prior art once published. However, the quirk of 102(e) as compared to pre-AIA 102(a) lies in the fact that these applications are treated as prior art from their filing date, not their publication date. An important subtlety is that a foreign patent filings, like Itou’s Japanese application, do not qualify as prior art under 102(e). In re Hilmer, 149 USPQ 480 (CCPA 1966).

Proving Prior Invention: The concept of ‘invention date’ is crucial in pre-AIA patent cases, and typically, this date is assumed to be the filing date of the patent application. The patentee can then furnish evidence to establish an earlier date of invention, which usually involves a sworn declaration from the inventor, corroborated by additional evidence. Perfect Surgical Techs., Inc. v. Olympus Am., Inc., 841 F.3d 1004 (Fed. Cir. 2016).

Both the Patent Trial and Appeal Board (PTAB) and the majority judges (Judge Lourie and Chief Judge Moore) concurred that Teleflex had provided adequate evidence to substantiate an earlier invention date. Judge Dyk, however, wrote in dissent.

It’s important to understand that the process of invention involves two steps – conception and reduction to practice. The majority and dissent agreed that Teleflex had provided evidence sufficient to proven both conception and the creation of a prototype before the critical date. However, reduction to practice also necessitates evidence demonstrating the suitability of the invention for its intended purpose. For complex inventions like the one in this case, the courts can require testing to establish this suitability prong. Steinberg v. Seitz, 517 F.2d 1359 (C.C.P.A. 1975). While the majority deemed the evidence of testing adequate, the dissent argued it was insufficient — with the debate focusing on the extent of corroboration required.

The majority observed, “[B]oth inventors Howard Root and Gregg Sutton testified regarding testing performed on a prototype of the claimed invention. These tests, although more qualitative than quantitative, were sufficient to confirm that the prototype would work for its intended purpose — providing increased backup support as compared with a guide catheter alone.” The judges concurred with the Board’s view that the tests did not require a direct comparison or a quantitative analysis to prove improvement in function. They also noted expert testimony suggesting “qualitative testing would have been sufficient.”

However, Judge Dyk, dissenting, questioned the lack of corroborative evidence supporting the prior testing. He cited to Cooper v. Goldfarb, 154 F.3d 1321 (Fed. Cir. 1998), to assert that “[i]n order to establish an actual reduction to practice, an inventor’s testimony must be corroborated by independent evidence.” Judge Dyk particularly concluded that every “inventive fact” should be corroborated with evidence beyond the inventors’ testimonies.

The corroborating evidence in this case included a sales presentation and two non-inventors’ testimonies. However, Judge Dyk believed that this evidence did not provide any information regarding proof of operation for the intended purpose. Specifically, there was “no specific description of what tests were performed or the results of the tests … [or] when the tests were performed and whether they were performed before the critical date.” Dissent.

The majority responded that not every detail of reduction-to-practice needs to be independently corroborated, advocating for a rule of reason. Despite agreeing with this general principle, Judge Dyk insisted that the lack of corroboratory evidence regarding testing for the intended use couldn’t be overlooked.

Judge Dyk, in his dissent, noted the lack of internal documents from Teleflex corroborating any testing during the crucial period in 2005. “Common sense, and Teleflex’s own testimony [showing record keeping in this area], suggest that these documents would exist if testing had occurred.” Dissent. He argued requiring such documents did not created an unreasonable standard for inventors seeking to enforce a patent claim. Judge Dyk’s dissent concludes that “I think that Itou has been shown to be prior art to the patents at issue and therefore could support a determination of anticipation or obviousness.”

However, the majority, led by Judge Lourie, disagreed with the dissent and upheld the PTAB’s decision in favor of Teleflex.

The attorneys for the appellants (Medtronic) were led by Tasha Bahal from Wilmer Cutler. Representing Teleflex were attorneys from Carlson Caspers, led by Derek Vandenburg.

11 thoughts on “Challenges of Proving Inventorship: Corroboration of All Inventive Facts

  1. 4

    The case underlines a situation where the transition to the AIA has resulted in the loss of certain rights for patent holders.

    You might have to explain this a bit better or rephrase it, because this case seems to be a run-of-the-mill “swearing behind the reference” dispute under pre-AIA that doesn’t involve the AIA transition causing a loss of rights for anyone. It isn’t a loss of rights for a pre-AIA patent to have more avenues than an AIA patent for overcoming potential prior art, since the AIA status of a patent doesn’t change.

    1. 4.1

      I think Dennis refers to the fact that after the AIA, an inventor / patent owner can no longer swear behind a reference?

    2. 4.2

      He cites In re Hilmer, so there is also an angle about extending secret prior art to non-US apps.

  2. 3

    A couple comments: taking the summary above at face value, it seems like Dyk has the better position here. The decision at the Board level (affirmed by the CAFC) may have made all the difference in the final (?) result.

    Also kudos to Congress for eliminating this messy “investigation” in post-AIA proceedings.

  3. 2

    Hmmm . . . anyone know what % of precedentials contain a dissent?

    Relatedly, do dissenting judges go along with designating opinions precedential?

    1. 2.1

      The court’s Internal Operating Procedure 10.6 addresses the second question about dissenting judges and precedential opinions.
      link to

      It says this:

      “An election to issue a precedential opinion shall be by a majority of the panel, except that, when the decision includes a dissenting opinion, the dissenting judge may elect to have the entire opinion issued as precedential notwithstanding the majority’s vote. These election rights may be made at any time before issuance of an opinion.”

      1. 2.1.1

        Thanks dcl. Interesting. Three questions come to mind:

        1. How often does the dissenting judge agree to designate precedential?

        2. Has is ever happened that a dissenting judge and just one of the majority judges (assuming a three judge panel) want precedential?

        3. Are such stats keep / publicly available?


          These are all interesting questions. I don’t know the answers, except that I’m reasonably sure the answer to #3 is no.

          I would guess that the answers to #1 and #2 are that precedential vs. non-precedential is usually done by consensus rather than an outright vote, with other panel members deferring either to the author’s preference or any strong views by other panel members. There probably isn’t a vote count; it’s probably more like one judge says “I think this should be precedential,” and the other two say “sure, whatever.”

          Discussions about whether to make an opinion precedential all happen behind closed doors. The Federal Circuit hears oral argument in almost every case in which there’s counsel on both sides and no one’s waived argument. Right after a morning of arguments, the judges on the panel meet to figure out (a) whether they agree on the result, (b) whether to decide the case with a Rule 36 disposition, non-precedential opinion, or precedential opinion, and (c) who writes the opinion if there is one. (IOP#8.1) After that, they go their separate ways, and the assigned author goes to work on the opinion.

          Finally, I think it’s relatively rare for a split decision not to be published. Under IOP 10.6, the dissenting judge can unilaterally decide the opinion should be published. And if there’s a dissent, the majority judges are less likely to think it’s a run-of-the-mill case.

  4. 1

    Curious as to this being precedential, given the limited shelf-life of PRE-AIA circumstances.

Comments are closed.