Hoffmann-La Roche, Inc. v. Promega Corp.

Case Overview & Background

Parties. The plaintiffs-appellants were Hoffmann-La Roche, Inc. and its wholly owned U.S. subsidiary Roche Molecular Systems, Inc. (collectively "Roche") — an operating pharmaceutical/diagnostics company, not an NPE or university. Roche's title came by acquisition: Cetus Corporation, the Emeryville, California biotech that developed PCR, licensed U.S. Patent No. 4,889,818 to Promega in June 1990 and then assigned the patent to Roche in December 1991 as part of Roche's ~$300 million purchase of the Cetus PCR/Taq portfolio. Hoffmann-La Roche, Inc. v. Promega Corp., 323 F.3d 1354, 1358–59 (Fed. Cir. 2003) (WorldLII; Justia). The defendant-appellee, Promega Corporation of Madison, Wisconsin, is a life-science reagents and molecular-biology tools supplier — here simultaneously a licensee of the patent and a competing Taq supplier. The dispute is therefore a licensor/licensee-and-competitor fight between two operating companies, not an NPE assertion.

Technology and patent at issue. The case is about Taq DNA polymerase, the thermostable enzyme from Thermus aquaticus that made the polymerase chain reaction practical: because Taq survives the high-temperature strand-separation step, it need be added only once per reaction, rather than every cycle. The single patent on appeal, the '818 patent ("Purified Thermostable Enzyme"), issued Dec. 26, 1989 to Cetus and has only three claims: claim 1 claims purified Taq polymerase defined structurally by its migration on a denaturing polyacrylamide gel (faster than phosphorylase B, slower than bovine serum albumin) and an estimated molecular weight of 86,000–90,000 daltons; claim 2 claims the polymerase isolated from T. aquaticus; and claim 3 claims the polymerase isolated from a recombinant organism transformed with a Taq-expression vector. Id. at 1358–59 & n.1. The accused activity was Promega's sale of Taq polymerase products (native and recombinant) allegedly outside/after breach of its 1990 license — Roche sued for patent infringement and breach of contract, and Promega counterclaimed that the '818 patent was unenforceable for inequitable conduct, which quickly became the whole case. Id. at 1359; Promega press release, Dec. 7, 1999 (archive).

Procedural posture. The infringement/contract action was filed in 1992 and docketed in the U.S. District Court for the Northern District of California as No. C-93-1748 VRW, before Judge Vaughn R. Walker in San Francisco — the venue where Cetus and Roche Molecular Systems were based. The district court granted partial summary judgment in August 1996 that four material misrepresentations had been made during prosecution; after a four-week bench trial in February 1999, Judge Walker held on December 7, 1999 that all claims of the '818 patent were unenforceable, finding eight instances of intentional withholding or distortion of information from the PTO. See Hoffmann-La Roche, Inc. v. Promega Corp., No. C-93-1748 VRW, 1999 WL 1797330 (N.D. Cal. Dec. 7, 1999); 323 F.3d at 1359. On appeal (No. 00-1372, argued to a panel of Judges Newman, Archer (Sr.), and Bryson), the Federal Circuit affirmed in part — upholding the findings as to the "Example VI" specification misrepresentations (an experiment written in the past tense that was never run as described) and the comparative fidelity/template-dependence statements — reversed in part on the molecular-weight-related omissions (including the undisclosed "Stoffel experiment" and size-exclusion chromatography evidence), and remanded for the district court to perform the equitable weighing of whether the surviving misconduct warranted unenforceability. 323 F.3d at 1359–73. Judge Newman dissented in full, warning that the majority "revive[d] that misbegotten era" of inequitable conduct charges and would unleash a "plague" of groundless attacks. Id. at 1374–81 (Newman, J., dissenting). On remand the district court entered judgment of unenforceability, effectively ending the patent.

Why the case matters. (1) Commercial stakes: the '818 patent was a foundational PCR patent; contemporaneous reporting put global Taq sales at roughly $200 million annually, with Roche and Perkin-Elmer collecting royalties of 15–17 cents per unit on a reagent used in essentially every molecular-biology lab (Nature 402, 709 (1999); Nat. Biotechnol. 18, 7 (2000), DOI). (2) Doctrinal significance: it is a standard citation on "prophetic" versus performed examples and on inferring deceptive intent from verb tense and inventor Rule 56 declarations; the Newman dissent became a frequently cited critique in the run-up to Therasense, Inc. v. Becton, Dickinson & Co., 649 F.3d 1276 (Fed. Cir. 2011) (en banc), which tightened the materiality/intent standards. (3) Downstream litigation and parallel proceedings: there was no IPR/PGR linkage — the case predates the AIA and the PTAB; the parallel challenges were the EPO revocation of Roche's counterpart native-Taq patent in 2000, German and Australian actions, and a follow-on U.S. antitrust MDL in D.D.C. asserting Walker Process fraud claims by direct purchasers of Taq (Judge Henry H. Kennedy, Jr.), which settled for $33 million in 2008. Roche's later follow-on effort to extend unenforceability to its related PCR patents ('195 and '202) failed on the "immediate and necessary relation" element. Notably, the Taq enzyme resurfaced at the Supreme Court in Life Technologies Corp. v. Promega Corp., 580 U.S. 140 (2017), but that case concerned § 271(f)(1) and the Tautz reissue patent (with Taq polymerase as the single U.S.-supplied kit component) — it did not involve the '818 patent.

Sourcing caveats. I could not access PACER; the district court docket number, the 1992 filing date, and the scope of Promega's license come from the Federal Circuit opinion, the 1999 WL citation as reproduced in secondary literature, and contemporaneous press/Nature reporting rather than from docket sheets. The February 1999 trial date, the "eight instances" finding, and the December 7, 1999 ruling are well corroborated. Less well sourced: the precise terms of the parties' 1990 license and the specific accused product lines/sales volumes, which the 2003 opinion does not discuss because the appeal was confined to the inequitable conduct ruling.

I'll research this case thoroughly. Let me start with several parallel searches.

Good foundation. Now let me dig into the district court docket, the remand, and any parallel proceedings.

Hoffmann-La Roche, Inc. v. Promega Corp., No. 00-1372 (Fed. Cir.) — Chronology of Key Developments

Scope note: This was a 1990s-era patent case, so several categories you asked about simply do not exist here. There was no jury, no Markman/claim-construction proceeding, and no PTAB IPR/PGR (the AIA trial regimes did not begin until 2012-09-16, nine years after this appeal ended). The case was litigated and resolved entirely on an equitable defense — inequitable conduct — before Judge Vaughn R. Walker in the N.D. Cal. I flag below where sources are thin.

---

1. The patent and the parties' pre-suit relationship

• 1989-12-26 — U.S. Patent No. 4,889,818 ("Purified Thermostable Enzyme") issues to Cetus Corp., covering purified Thermus aquaticus (Taq) DNA polymerase, the enzyme central to PCR. Inventors: David Gelfand and Susanne Stoffel. The patent has only three claims.
• June 1990 — Cetus licenses the '818 patent to Promega (a non-PCR-use license).
• 1991 — Cetus sells the '818 patent and related PCR rights to Hoffmann-La Roche / Roche Molecular Systems (~$300 million).

Source: Hoffmann-La Roche, Inc. v. Promega Corp., 323 F.3d 1354, 1357–58 (Fed. Cir. 2003) ("Roche II") (CourtListener; Justia).

2. Filing and initial pleadings (1992)

• 1992 — Roche sues Promega in the U.S. District Court for the Northern District of California (San Francisco; Judge Walker), alleging patent infringement and breach of contract — the dispute arose from Promega's refusal to accept a renegotiated license with materially higher royalties. Promega counterclaimed, asserting among other things that the '818 patent was unenforceable for inequitable conduct and invalid. That counterclaim "soon became the focus of the litigation."
• The exact 1992 filing date and the N.D. Cal. docket number are not confirmed in the sources I could access; contemporaneous press accounts only say "1992." (Roche II, 323 F.3d at 1358; Nature Biotechnology, "Roche's Taq patent invalid"; UNCTAD case summary).

Why this matters: the case was never a conventional infringement damages case. From 1992 forward, Roche's infringement theory was held hostage to the enforceability of a patent prosecuted by a predecessor (Cetus) whose lab notebooks and internal memos became the central evidence.

3. Pre-trial motions of substance

• August 1996 — On summary judgment, the district court held that the inventors had made four material misrepresentations during prosecution of the '818 patent. This was a partial merits ruling on inequitable conduct — highly unusual to resolve on SJ, and the Federal Circuit later reversed part of it. (Roche II, 323 F.3d at 1358.)
• No motion to dismiss, transfer, or stay is reported in the sources I accessed. Notably, there was no stay pending reexamination, even though Roche later asserted that the PTO had reexamined the '818 patent and recertified the claims without amendment (see §8 below).
• Roche's own characterization of the procedural history includes a January 2000 hearing request on a motion directed at the adverse judgment (described in trade press as a "motion to appeal"/certification step); the precise relief sought is not well documented in accessible sources (Patent Litigation roundup, Feb. 2000).

4. Claim construction

None. The case never reached a Markman proceeding. The '818 patent's three claims were not construed, because the defense that decided the case (inequitable conduct) does not turn on claim scope. The Federal Circuit's opinion addresses claim language only incidentally (e.g., claim 1's recitation of an 86,000–90,000 dalton molecular weight) in assessing materiality of prosecution arguments.

5. Discovery milestones with strategic significance

Discovery dates are not published, but the record reveals the pivotal materials:

• The "Stoffel experiment" — an undisclosed Cetus purification experiment in which a 62 kD Taq fragment failed to bind a phosphocellulose column. The district court found withholding it was a material omission; the Federal Circuit reversed on both materiality and intent because the experiment used 80 mM salt versus Chien's 10 mM.
• Size-exclusion chromatography materials — Stoffel purification data, a Drummond sizing experiment, hydrophobicity data, and an internal Raymond memorandum noting Taq "migrates differently on [Z]orbax or other sizing columns." These underlay the molecular-weight inequitable conduct theory.
• Cetus/Roche laboratory notebooks and prep records (Prep 3 and Prep 4) — used to prove that patent Example VI was never actually run as written.
• Deposition/trial admissions — Gelfand conceded: "Susanne Stoffel and I never carried out Example 6 from start to finish as laid out." Stoffel's testimony was similar. This single admission effectively decided the Example VI issue.
• Expert discovery — Promega's experts Drs. Mosbaugh, Linn, Burgess, and Kunkel; Roche's Dr. Chamberlin. Dr. Kunkel's testimony (a scientist who had worked with Gelfand) supplied the intent evidence on the fidelity/template-dependence misstatements, and the Federal Circuit relied on it heavily — the focus of Judge Newman's dissent.

Source: Roche II, 323 F.3d at 1360–72; CourtListener opinion text; J. Marshall Rev. Intell. Prop. L. case note.

6. Trial and the district court's liability ruling

• February 1999 — Four-week bench trial before Judge Walker (equitable issues tried to the court; no jury).
• 1999-12-07 — Judge Walker rules the '818 patent unenforceable for inequitable conduct, finding eight separate material misrepresentations/omissions made with intent to deceive the PTO, spanning three categories: (1) molecular weight differences vs. the Chien/Kaledin prior art; (2) the Example VI misrepresentation; and (3) comparative fidelity/template-dependence statements. Citation: Hoffmann-La Roche, Inc. v. Promega Corp., 1999 WL 1797330 (N.D. Cal. Dec. 7, 1999) ("Roche I"). The court described the conduct in terms approaching "scientific misconduct."
• Practical effect: Promega was held not liable for infringement, and Roche's royalty stream on native Taq was placed in jeopardy; press reports at the time noted other suppliers began selling unlicensed Taq after the ruling.

Sources: Roche I, 1999 WL 1797330; Nature, "Roche's Taq patent 'obtained by deceit'"; Genome Biology, "Promega wins Taq patent round" (2003-04-01).

• Post-trial motions: No JMOL/new-trial practice applied (court-tried equitable issues). The court instead proceeded to a remedies phase, including Promega's fee request. In the later remand decision the court noted Promega's attorney-fee request was premature because significant claims remained; I found no record of a fee award.

7. Appeal to the Federal Circuit — No. 00-1372

• Appeal docketed 2000 (No. 00-1372), from the unenforceability order.
• 2003-03-31 — Decision: affirmed in part, reversed in part, and remanded. 323 F.3d 1354, 66 U.S.P.Q.2d 1385, 2003 WL 1634054. Judge Bryson wrote for the majority; Judge Newman dissented.
  • Reversed: the entire molecular-weight category (Stoffel experiment materiality and intent; size-exclusion chromatography/hydrophobicity/Raymond memo). Of the eight predicate findings, Nos. 1 and 6 were reversed.
  • Affirmed: the Example VI misrepresentation (specification written in past tense describing an experiment never performed as written; best-mode inclusion does not license suggesting the best mode was performed) and the fidelity/template-dependence misrepresentations (material even though purity/fidelity were not claim limitations).
  • Remanded: for the district court to exercise discretion on the second step — "whether under all the circumstances, the incidents of inequitable conduct we have sustained are such as to justify the sanction of rendering the '818 patent unenforceable." 323 F.3d at 1372.
  • Judge Newman's dissent: the majority "finds misrepresentation in correct science, infers malevolence from verb tense, and grounds intent to deceive on personal slurs by a hostile witness," and "revives that misbegotten era" of inequitable-conduct "plague."
• 2003-05-28 — The Association of Patent Law Firms filed an amicus brief urging rehearing en banc, targeting the majority's materiality and intent holdings on Example VI (brief). The disposition of that petition is not confirmed in the sources I accessed; no en banc opinion issued.

8. Remand and the final enforceability ruling

• 2004-05-13 — On remand, Judge Walker again held the '818 patent unenforceable, rejecting Roche's argument that the two surviving findings reflected only "meager" materiality and intent. Hoffmann-La Roche, Inc. v. Promega Corp., 319 F. Supp. 2d 1011 (N.D. Cal. 2004), 2004 WL 1192709. Key holdings:
  • Materiality is not a "but for" standard; the surviving purity/fidelity misstatements were material even if molecular weight was the lead argument.
  • The court then applied unclean hands and dismissed Roche's 1st–4th causes of action (breach of contract) — reasoning that the license of the '818 patent was the primary consideration, that "without the '818 patent, there would have been no license agreement," and that a prospective licensee is ordinarily unable to discover prosecution misconduct pre-license. It also dismissed the 8th cause of action.
  • The court denied dismissal of the

I'll research the docket and opinions for this appeal to identify counsel of record for the plaintiff.

I have the opinion's counsel block. Now let me verify the Federal Circuit docket and dig into the district court counsel of record and each attorney's background.

Bottom line

The published opinion supplies a complete counsel block for the plaintiffs‑appellants — Hoffmann‑La Roche, Inc. and Roche Molecular Systems, Inc. (collectively "Roche"). The plaintiff side was represented by a two‑firm team: Pennie & Edmonds LLP (New York) as lead/appellate counsel and Orrick, Herrington & Sutcliffe LLP (San Francisco) as supporting/of‑counsel counsel. See Hoffmann‑La Roche, Inc. v. Promega Corp., 323 F.3d 1354, 1356–57 (Fed. Cir. 2003) (counsel block), available at Justia and FindLaw (same counsel block reproduced).

Plaintiff‑side counsel of record

Important caveats and gaps

1. Source of the roster is the opinion, not PACER. The Federal Circuit's CM/ECF docket for No. 00‑1372 was not retrievable in this review, and the N.D. Cal. docket (No. C‑93‑1748 VRW, Judge Vaughn R. Walker) was not pulled. The counsel block printed at 323 F.3d at 1356–57 is the authoritative public record of appearances at the briefing/argument stage. A live docket could show additional appearances, withdrawals, or in‑house counsel not reflected there.
2. No in‑house Roche attorney is identified. I found no public record naming Roche Molecular Systems in‑house counsel as counsel of record on this appeal. If one appeared, I could not verify it — I am not inventing one.
3. District‑court appearances are only inferable. The presence of Misrock (Pennie & Edmonds) and Caulfield (Orrick) as of‑counsel on the appeal, plus the firms' San Francisco/New York footprints, is consistent with Pennie & Edmonds and Orrick having handled the N.D. Cal. trial, but the trial‑level appearance list was not verified. The remand decisions are reported at 1999 WL 1797330 (N.D. Cal. Dec. 7, 1999) and 319 F. Supp. 2d 1011 (N.D. Cal. 2004).
4. Firm continuity. Pennie & Edmonds, the New York IP boutique that led Roche's appellate briefing, subsequently disbanded (its lawyers dispersed to other firms, including Orrick and Jones Day). This is general background rather than something I verified in a docket source; it explains why the same attorneys appear at different firms in later records.
5. Caption note. The docket caption on appeal named both Roche entities as plaintiffs‑appellants — Hoffmann‑La Roche, Inc., and Roche Molecular Systems, Inc. v. Promega Corp. — even though the case metadata lists only Hoffmann‑La Roche, Inc. For purposes of a counsel table, the relevant plaintiff‑side party is the Roche group.
6. Disposition nuance. The opinion's own judgment line reads "VACATED and REMANDED," 323 F.3d at ¶ 72, reflecting affirmance of two of three inequitable‑conduct categories, reversal of the molecular‑weight findings, and vacatur of the unenforceability order — slightly different phrasing from the "affirmed in part and reversed in part" status label, though substantively consistent.

Not represented here: any counsel for defendant‑appellee Promega (James R. Troupis of Michael Best & Friedrich, Madison, WI, argued; with Michael E. Husmann, J. Donald Best, John C. Scheller, Karla M. Davis; of counsel Peter G. Carroll and Kamrin T. MacKnight of Medlen & Carroll, San Francisco, and Russell L. Johnson of Skjerven, Morrill, MacPherson, Franklin & Friel, San Jose). That side is omitted because you asked only for plaintiff‑side counsel, but the same counsel block is the source.

I'll research the docket, opinions, and counsel records for this Federal Circuit appeal.

The opinion's counsel block gives me a strong starting point. Let me verify against additional sources, including the Justia/CourtListener records and the district court docket.

The opinion's official counsel block is now confirmed by multiple reporters. Let me verify roles and backgrounds for each named attorney, and check the district court docket line.

Counsel of Record — Defendant-Appellee Promega Corporation

Bottom line: Promega's appellate team is fully identified in the published opinion's official counsel block (323 F.3d 1354, 1357), which is reproduced verbatim by FindLaw, Justia, and Casetext/vLex. James R. Troupis of Michael Best & Friedrich (Madison, WI) was lead/arguing counsel, supported by four Michael Best colleagues on the brief and three "of counsel" attorneys from two California firms. No in-house Promega attorney appears in the counsel block.

Source for the list

The Federal Circuit's official caption block reads: "James R. Troupis, Michael Best & Friedrich LLP, of Madison, WI, argued for defendant-appellee. With him on the brief were Michael E. Husmann, J. Donald Best, John C. Scheller, and Karla M. Davis. Of counsel on the brief were Peter G. Carroll and Kamrin T. MacKnight, Medlen & Carroll LLP, of San Francisco, CA; and Russell L. Johnson, Skjerven, Morrill, MacPherson, Franklin & Friel LLP, of San Jose, CA." — 323 F.3d 1354, 1357 (FindLaw; Justia; vLex/Casetext).

Notes on role classification

• "Argued for defendant-appellee" is the court's own designation of lead appellate counsel — that is Troupis.
• "With him on the brief" denotes the core briefing team (Michael Best); the court does not rank among them.
• "Of counsel on the brief" covers the California IP counsel. Neither Medlen & Carroll nor Skjerven Morrill served as "local counsel" in the formal sense (the Federal Circuit has no local-counsel requirement); functionally they were Promega's Northern California trial counsel and their appellate support was of-counsel. Russell L. Johnson's separate appearance line in the N.D. Cal. remand orders confirms the trial-court role.
• In-house: no Promega in-house lawyer is named in the opinion's counsel block, and I found no in-house appearance. (The general counsel quoted in contemporaneous Nature coverage, Melinda Griffith, was general counsel of Roche Molecular Systems — i.e., the appellant side — not Promega.)

Sourcing caveats — what I could not verify

• The Federal Circuit docket sheet for No. 00-1372 is not freely retrievable. Pre-2004 Federal Circuit dockets are not available through RECAP/CourtListener, and I could not pull the PACER docket in this session. The counsel list above is therefore grounded in the published opinion's counsel block, which is the authoritative public record of who appeared on the appeal, rather than in the docket entries themselves. I cannot report individual docket entry numbers for appearances or the date each attorney entered.
• Nothing indicates sealing on the appellate counsel designations; the inequitable-conduct record was litigated publicly (bench trial Feb. 1999; findings at 1999 WL 1797330; remand decision at 319 F. Supp. 2d 1011 (N.D. Cal. 2004)). I did not, however, review the N.D. Cal. docket (No. C 93-1748 VRW) in full, so any additional district-court-only counsel — including any Promega in-house lawyer who entered an appearance at trial but not on appeal — may exist and is not captured here.
• Biographical details for Husmann, Best, Scheller, and Davis are thin in the sources I could reach. I confirmed Husmann's Michael Best affiliation as of 1999–2000 with a Milwaukee office address in the Best Lawyers directory; the opinion gives only "Madison, WI" at the firm level, so I have flagged that discrepancy rather than silently resolving it. I have deliberately not supplied practice-area descriptions or case lists for these four that I could not source.
• Troupis's later career is well documented (22 years at Michael Best, departure June 2010 to found Troupis Law Office LLC in Middleton; the Wisconsin Law Journal coverage of the $52M Promega–Life Technologies verdict, later overturned, with Troupis as Promega's counsel) — see wislawjournal.com. His role in this appeal as lead is confirmed by the opinion; his pre-2003 patent docket beyond this case is not something I verified.

If you need the district-court appearance list (including any trial-only counsel and entry dates), that requires a PACER pull of N.D. Cal. No. C 93-1748 VRW — say the word and I'll frame the specific docket items to request.