US 9504651

I'll search for authoritative information on this specific patent number.

Let me get the claim text and the Federal Circuit docket details.

Let me confirm the full claim set and any 2026 Federal Circuit activity specific to this patent.

US 9,504,651 B2 — Summary (as of April 26, 2026)

Caveat on sourcing: I do not have direct USPTO PatentCenter/PACER credentials, so the below is grounded in (a) the full patent text supplied to me from Google Patents, (b) court filings indexed on CourtListener/archive.org, and (c) Federal Circuit briefing PDFs. Where a figure is my reading rather than an authoritative record, I flag it.

Bibliographic data

Two literal-reading flags I will not "correct":

1. The recorded USPTO assignment data on the Google Patents page lists the assignor as "GIESBRECHT, NOELLE", whereas the patent front page and pleadings name Cory Giesbrecht. I report both as they appear.
2. Google Patents states the adjusted expiration as 2023-07-15. A third-party (Jefferies) research deck in the record refers to the '651 patent expiring "Jul 2028," which conflicts. The patent-face/Google datum is 2023-07-15.

Abstract (verbatim)

"The present invention provides apparatus and processes for producing liposomes. By providing a buffer solution in a first reservoir, and a lipid solution in a second reservoir, continuously diluting the lipid solution with the buffer solution in a mixing chamber produces a liposome. The lipid solution preferably comprises an organic solvent, such as a lower alkanol."

Notable mismatch worth flagging: the abstract and specification are directed to a liposome manufacturing process/apparatus (a T-connector mixing chamber, stepwise ethanol dilution of lipid in alkanol, "SPLP" / stable plasmid-lipid particles, DSPC:Chol:PEG-DSG:DODMA at ~20:45:10:25, DNA:lipid ratios, ultrafiltration, etc.). The granted claims, however, are composition claims to an mRNA-containing lipid vesicle formulation. The title tracks the claims, not the specification.

Independent claim (plain language)

Public pleadings state the '651 patent has one independent claim — claim 1 (First Amended Complaint, SDNY 1:22-cv-02229, ¶69). Claims 13 and 14, though dependent, were separately construed because they recite progressively higher encapsulation thresholds.

Claim 1 — "A lipid vesicle formulation comprising:"

• (a) a plurality of lipid vesicles, where each vesicle contains:
  • a cationic lipid;
  • an amphipathic lipid; and
  • a polyethyleneglycol (PEG)-lipid; and
• (b) messenger RNA (mRNA), wherein at least 70% of the mRNA in the formulation is fully encapsulated in the lipid vesicles.

In plain terms: it is a composition-of-matter claim to a lipid-nanoparticle-style formulation — three required lipid classes plus mRNA payload — where the defining quantitative limitation is that ≥70% of the mRNA is "fully" (not partially) encapsulated. No process steps, apparatus, molar ratios, particle size, or specific lipid identities are required by claim 1 itself.

Representative dependent claims (from the Federal Circuit record):

• Claim 4: each vesicle further comprises a sterol.
• Claim 6: sterol is cholesterol and amphipathic lipid is a phospholipid.
• Claim 7: the phospholipid is selected from a Markush group (phosphatidylcholine, PE, PS, PI, phosphatidic acid, POPC, lyso-PC, lyso-PE, DPPC, DOPC, DSPC, DLPC).
• Claim 9: each lipid vesicle is a "lipid-nucleic acid particle."
• Claim 11: the cationic lipid only carries a positive charge below physiological pH (i.e., ionizable).
• Claim 13: ≥80% fully encapsulated. Claim 14: ~90% fully encapsulated.

Claim construction (D. Del., announced Apr. 4, 2024): "fully encapsulated" was construed as "fully, as distinct from partially, contained inside the lipid vesicles." Moderna had argued indefiniteness, contending the specification's only relevant sentence ("'lipid encapsulated' can refer to ... full encapsulation, partial encapsulation, or both") gives no measurement method and that "fully" cannot mean the proportion encapsulated because proportion is recited separately (70/80/90%).

Litigation and the CAFC 2026 docket

• Arbutus Biopharma Corp. v. Moderna, Inc., Fed. Cir. No. 2026-1581 — appeal from D. Del. 1:22-cv-00252-JDW (Judge Joshua D. Wolson). Google Patents' litigation feed links the family to case 26-1581.
  • Scope of the appeal is narrow: the sole issue is 28 U.S.C. § 1498(a) — whether the government-contractor defense bars Arbutus/Genevant's claims for the vaccine doses Moderna made under the C-100 contract. It is not an appeal of the '651 patent's validity or of claim construction.
  • Timeline: opening brief 5/29/2026; U.S. amicus brief supporting Moderna 6/19/2026; joint appendix 8/21/2026; Arbutus/Genevant responsive brief ~8/2026; the United States moved (unopposed) on 9/1/2026 for leave to argue, seeking 20 min/side.
  • Consent judgment (3/4/2026): Moderna consented to judgment of infringement and no invalidity as to four Genevant/Arbutus patents (the '651 among them), leaving only § 1498 for appeal.
  • Settlement (3/3/2026): $950M upfront (payable Q3 2026) plus up to $1.3B contingent on a favorable Federal Circuit ruling.
• Other '651-related matters on the docket map: Arbutus/Genevant v. United States, Fed. Cl. No. 26-446 (filed 3/19/2026) for the government-employee doses; Pfizer/BioNTech actions (D.N.J. 3:23-cv-01876 and 3:23-cv-04200); D. Del. 1:22-cv-00252; SDNY 1:22-cv-02229. The '651 was also one of six patents asserted in the Moderna case (with 8,058,069; 8,492,359; 8,822,668; 9,364,435; 11,141,378).
• IPR: I found no PTAB inter partes review petition specific to the '651 patent in my searches. The record I retrieved shows Moderna challenged the patent's validity in district court (invalidity contentions, indefiniteness, anticipation/obviousness over Saravolac '171, Semple '278, Zhang '745, Meulien '831, etc.) rather than via IPR on this patent. I cannot rule out an unfound petition — treat that as not confirmed.

Priority-art context

The patent's asserted priority reaches back to the 2002-06-28 provisional, which is well before the mRNA-LNP vaccine era. Moderna's invalidity theory leaned heavily on the process disclosures of the same specification (encapsulation "up to about 90%") and on prior art such as Saravolac '171 and the Semple references teaching 40–80%/≥90% nucleic acid encapsulation and salt/pH optimization. This is the core tension in the case: the specification teaches pDNA encapsulation via an ethanol-dilution process, while claim 1 is a genus-level composition claim reciting mRNA — a point Moderna pressed in its indefiniteness and enablement positions.

Explicit uncertainties

• Total claim count and whether any independent claim other than claim 1 exists. Pleadings state claim 1 is the sole independent claim; the Federal Circuit brief labels claims 1 and 9 as "representative," which does not itself prove independence of claim 9. I did not independently verify the full printed claim set from the patent face.
• Expiration date — reported as 2023-07-15 by Google Patents but contradicted in one secondary source; I have not confirmed via USPTO PatentCenter.
• I did not access USPTO PatentCenter or PACER directly; bibliographic details on the front page (inventors, assignee chain, issue date) are consistently corroborated across Google Patents, Unified Patents, the patent's own assignment record, and multiple federal complaints, so I treat them as reliable.

I'll research the litigation docket for US 9,504,651. Let me start with the litigation records referenced in the patent record and cross-check with other sources.

Key findings so far. Let me dig into each docket and the Federal Circuit appeal.

Let me resolve the remaining docket numbers (S.D.N.Y. 1:22-cv-02229, D.N.J. 3:23-cv-04200) and the Federal Circuit appeal.

Summary

US 9,504,651 ("Lipid compositions for nucleic acid delivery," originally Protiva Biotherapeutics, now Arbutus Biopharma Corp.) has been asserted or put at issue in four U.S. court proceedings — three district court actions and one Federal Circuit appeal. Two are affirmative infringement suits by Arbutus/Genevant; one is a declaratory-judgment action brought by a supplier (Acuitas); one is Moderna's pending § 1498 appeal.

Sources for the docket list: the Google Patents litigation record (https://patents.google.com/patent/US9504651/en), which links to Unified Patents litigation pages for D. Del. 1:22-cv-00252, D.N.J. 3:23-cv-01876, D.N.J. 2:23-cv-01876, D.N.J. 3:23-cv-04200, S.D.N.Y. 1:22-cv-02229, and Fed. Cir. 26-1581.

Case details

1. Arbutus/Genevant v. Moderna — D. Del. 1:22-cv-00252 (the primary '651 case)

• Plaintiffs: Arbutus Biopharma Corporation (US) and Genevant Sciences GmbH (Switzerland) — Genevant is Arbutus's exclusive licensee of the LNP portfolio.
• Defendants: Moderna, Inc. and ModernaTX, Inc.
• Jurisdiction/venue: U.S. District Court for the District of Delaware, Wilmington. Filed February 28, 2022 as 1:22-cv-00252-MSG (Judge Mitchell S. Goldberg, sitting by designation); reassigned in July 2025 to Judge Joshua D. Wolson (E.D. Pa., sitting by designation) after Judge Goldberg retired.
• Patents asserted: U.S. 8,058,069; 8,492,359; 8,822,668; 9,364,435; 9,504,651; and 11,141,378 — asserted against Moderna's Spikevax (mRNA-1273) COVID-19 vaccine, and later against mRESVIA (RSV vaccine).
• '651-specific developments: In the April 3, 2024 claim construction ruling, the court construed the '651 encapsulation limitation ("wherein at least 70% / at least 80% / about 90% of the mRNA in the formulation is fully encapsulated in the lipid vesicles") to mean "wherein at least 70% / at least 80% / about 90% of the mRNA is fully, as distinct from partially, contained inside the lipid vesicles." (Arbutus press release, Apr. 4, 2024, https://investor.arbutusbio.com/news-releases/news-release-details/arbutus-biopharma-announces-claim-construction-ruling-its)
• Summary judgment (Feb. 2, 2026): The court largely rejected Moderna's attempt to shift liability to the Court of Federal Claims under 28 U.S.C. § 1498(a), holding that only doses administered directly to U.S. government employees (~6.2 million doses) were "for the Government"; most doses distributed to the general public were not. The court granted summary judgment for Moderna on prosecution history estoppel (barring doctrine-of-equivalents infringement for the molar-ratio patents, requiring literal infringement) but found genuine disputes of material fact on indefiniteness, leaving validity for the jury. (VitalLaw/IP Law Daily, Feb. 3, 2026, https://www.vitallaw.com/news/patent-d-del-most-covid-19-vaccine-patent-infringement-claims-against-moderna-proceed-government-use-defense-narrowed/ipm01ddf9ae6f2aff4c83bbe21d3762646c71)
• Outcome: A jury trial was set for March 9, 2026. On March 3, 2026 the parties announced a global settlement resolving the U.S. case and all related worldwide Moderna litigation (Spikevax and mRESVIA). Moderna agreed to pay $950 million in Q3 2026, with up to $1.3 billion contingent on the outcome of its § 1498 appeal (potential total $2.25 billion). Moderna consented to entry of a judgment of infringement and of no invalidity on the asserted patents. On March 4, 2026, Judge Wolson entered the consent judgment and order, and the case was terminated. (Bloomberg Law, Mar. 5, 2026, via https://law.illinois.edu/wp-content/uploads/2026/03/SherkowModernaBloomberg.pdf; Kilpatrick client note, https://ktslaw.com/-/media/2026/Kilpatrick-Client-Success-Kilpatricks-Patent-Prosecution-Expertise-Plays-Key-Role-in-Record-Breaking-Covid-Vaccine-Settlement.pdf)
• Note on the number of patents in the final judgment: The Rule 6 report to the Commissioner on the docket lists five patents (US 8,058,069; 8,492,359; 8,822,668; 9,364,435; 9,504,651), while the Kilpatrick client note refers to "four asserted patents." The case opened with six patents in suit, so the asserted set was narrowed during litigation; I could not confirm from the sources I retrieved exactly which subset was in the consent judgment.

2. Arbutus/Genevant v. Pfizer/BioNTech — D.N.J. 3:23-cv-01876 (the live '651 case)

• Plaintiffs: Arbutus Pharma Corp. (also captioned Arbutus Biopharma Corp.) and Genevant Sciences GmbH.
• Defendants: Pfizer Inc. and BioNTech SE (BioNTech is also a counterclaimant).
• Jurisdiction/venue: U.S. District Court for the District of New Jersey (Trenton), Case No. 3:23-cv-01876-ZNQ-TJB (Judge Zahid N. Quraishi; Magistrate Judge Tonianne J. Bongiovanni).
• Filing date: April 2023 — the S.D.N.Y. letter of May 16, 2023 states Arbutus "sued Pfizer and BioNTech in New Jersey" on April 4, 2023 (Acuitas Dkt. 72). Some contemporaneous filings cite the number as 2:23-cv-01876 (the Google Patents record lists both a "2:23-cv-01876" and a "3:23-cv-01876" D.N.J. entry); the operative docket number used by the court is 3:23-cv-01876.
• Patents asserted: U.S. 9,504,651; 8,492,359; 11,141,378; 11,298,320; and 11,318,098 — against Pfizer/BioNTech's COMIRNATY COVID-19 vaccine. Note: Arbutus's September 9, 2025 Form 8-K/Exhibit 99.1 lists the first patent as "9,504,561," which appears to be a transposition of '651; the court's own opinion (Dkt. 246) captions the patents as "9,504,651 ('the '651 Patent'), 8,492,359, 11,141,378, 11,298,320, and 11,318,098." (https://www.sec.gov/Archives/edgar/data/[1447028](/patent/1447028)/000117184325005815/exh_991.htm)
• Key '651 development: The court construed four disputed terms on September 9, 2025: "lipid vesicle," "fully encapsulated," "mol %," and "consisting essentially of."
• Current status: Still pending. An agreed fact-discovery deadline of May 8, 2026 was reported to the court in January 2026, with discovery disputes (privilege log, custodial files of Janet Shaw, motion to amend invalidity contentions) continuing into August 2026. (CourtListener docket 67142175, e.g., Dkt. 323, 329.)
• Related global expansion: In July 2026 Arbutus/Genevant extended the Pfizer/BioNTech campaign globally, adding claims in the Federal Court of Canada and two UPC actions at the The Hague local division (UPC-CFI-0002562/2026 on EP 4 241 767 and UPC-CFI-0002566/2026 on EP 4 495 237). Those assert European family members, not US 9,504,651.

3. Acuitas Therapeutics Inc. v. Genevant Sciences GmbH and Arbutus Biopharma Corp. — S.D.N.Y. 1:22-cv-02229

• Plaintiff: Acuitas Therapeutics Inc. (supplier of the LNP lipids used in Comirnaty).
• Defendants: Genevant Sciences GmbH and Arbutus Biopharma Corp.
• Jurisdiction: U.S. District Court for the Southern District of New York; filed March 18, 2022; initially assigned to Judge Edgardo Ramos, later to Judge Mary Kay Vyskocil (1:22-cv-02229-MKV).
• Nature: Declaratory-judgment action (and, via BioNTech/Pfizer counterclaims filed July 10, 2023) seeking declarations that nine Arbutus patents — including 9,504,651, 8,492,359, and 11,141,378 — are not infringed and/or invalid as applied to Comirnaty. Acuitas brought the case after Arbutus/Genevant sent § 287(a) notice letters to Pfizer and BioNTech; Acuitas alleged potential induced/contributory infringement liability and indemnification obligations to BioNTech. (Docket materials at https://www.courtlistener.com/docket/63169706/acuitas-therapeutics-inc-v-genevant-sciences-gmbh/)
• Outcome: After Arbutus filed the New Jersey action against Pfizer/BioNTech, Acuitas gave notice of voluntary dismissal without prejudice on August 4, 2023 (Dkt. 79). The docket is closed.

4. Arbutus Biopharma Corporation and Genevant Sciences GmbH v. Moderna, Inc. and ModernaTX, Inc. — Fed. Cir. No. 26-1581

• Parties: Moderna, Inc. and ModernaTX, Inc. as Appellants (defendants-appellants); Arbutus and Genevant as Appellees.
• Origin: Appeal from the U.S. District Court for the District of Delaware, Judge Joshua D. Wolson, Case No. 1:22-cv-000252-JDW.
• Issue: Whether Moderna's government-contractor immunity defense under 28 U.S.C. § 1498 limits its liability for vaccine doses supplied under federal contracts. The appeal is preserved as part of the settlement; if the Federal Circuit affirms the district court, Moderna owes up to an additional $1.3 billion within 90 days; if Moderna prevails, no further payment (and any payment would be refunded with interest).
• Status: Pending. Moderna's opening brief (dated May 29, 2026) reproduces representative claims of U.S. 9,504,651 (claim 1: a lipid vesicle formulation comprising lipid vesicles with a cationic lipid, an amphipathic lipid, and a PEG-lipid, plus mRNA wherein at least 70% of the mRNA is fully encapsulated) and of U.S. 9,364,435. (https://fedcircuitblog.com/wp-content/uploads/2026/06/Arbutus-Biopharma-Corp.-v.-Moderna-Inc.-Opening-Brief.pdf)

Listed but unverified

• D.N.J. 3:23-cv-04200. The Google Patents litigation record for US 9,504,651 lists a New Jersey District Court action at case 3:23-cv-04200 (https://portal.unifiedpatents.com/litigation/New%20Jersey%20District%20Court/case/3%3A23-cv-04200). I was unable to verify its parties, filing date, patents, or status before my search budget was exhausted — I am explicitly not asserting what this case is. It may be a related/consolidated matter arising from the same Pfizer/BioNTech dispute (July 2023 vintage), but that is a hypothesis, not a verified fact.
• D.N.J. 2:23-cv-01876. The Google Patents record also lists this number. The only D.N.J. docket I could verify is 3:23-cv-01876; the "2:" variant appears in party correspondence referring to the same Pfizer/BioNTech action (see Acuitas Dkt. 72, describing "No. 2:23-cv-01876-ZNQ (D.N.J.)"). Treat these as the same case unless a separate docket is confirmed.

Related proceedings that do not assert US 9,504,651

• Moderna's IPRs: Moderna challenged U.S. 8,058,069 and 9,364,435 in inter partes reviews (final decisions upheld the '069 claims and upheld some '435 claims; the Federal Circuit affirmed as to '069 on Dec. 1, 2021 and dismissed Moderna's '435 appeal for lack of standing the same day, leaving the surviving '435 claims in force). I did not identify any PTAB proceeding against 9,504,651 itself — I did not run a dedicated PTAB search for '651, so this is a stated limitation of my search rather than a confirmed negative.
• The March 3, 2025 international Moderna suits (Canada T-704-25 on CA2721333; Japan 2025 (Wa) 70079 on JP5475753; Switzerland; UPC 10280/2025 on EP2279254; UPC 10284/2025 on EP4241767) assert foreign counterparts in the same family, not the US '651 patent itself. They are relevant context but are not litigation "involving US patent 9504651" in the strict sense.
• Patent status context: The Google Patents record shows US 9,504,651 as "Expired – Lifetime," adjusted expiration July 15, 2023, which is relevant to the temporal reach of damages.

Caveats

• The record's own docket links are the authoritative starting point; several case details above come from secondary sources (law-firm client alerts, news reports, and party filings) that may be updated or corrected.
• Date discrepancy: Your instructions state the current date is April 26, 2026, but the sources retrieved include documents dated after that date (e.g., the March 4, 2026 consent judgment and Rule 6 report, Moderna's May 29, 2026 appellate brief, and August 2026 discovery letters). I have reported them as found rather than suppressing them; if you need the state of play strictly as of April 26, 2026, the Moderna case had already settled and the Pfizer/BioNTech New Jersey case was in discovery with a May 8, 2026 fact-discovery deadline.

I'll start from the canonical structured block — which reports no AIA trials on file — and then search the web to check for any proceedings the ODP ingest may have missed.

Let me verify whether any AIA trial specifically names the '651 patent, and check the related Arbutus/Moderna IPR cluster.

Let me check the Vaccine Patent PTAB Tracker and search specifically for any petition against the '651 patent by Pfizer/BioNTech or others.

PTAB / AIA trial analysis — U.S. Patent No. 9,504,651 B2

Bottom line up front: I could not identify any AIA trial proceeding (IPR, PGR, or CBM) ever filed against U.S. Patent No. 9,504,651 B2. The structured "PTAB proceedings on file" block (USPTO ODP) reports none, and my independent searching — PTAB/patent databases, the Vaccine Patent PTAB Tracker, Arbutus SEC disclosures, and the parties' Delaware/New Jersey filings — surfaced no petition naming the '651. Everything below about other patents is included only as context, and is labeled as such.

Proceedings overview

Zero AIA trial proceedings on file for US 9,504,651 — 0 active, 0 with claims invalidated, 0 with claims sustained, 0 settled at the Board, 0 institution denials. The '651 has never been before the PTAB. It is not a "hardened" patent in the sense of having survived IPRs, nor is it a "walked-back" patent whose claims are canceled: it is PTAB-untested, meaning there is no Final Written Decision to cite either for or against it, and no § 315(e)(2) estoppel runs against anyone on this patent's claims. The invalidity fight over the '651 — derivation under § 102(f), anticipation/obviousness over Semple and Bischoff-based art, and § 112 enablement/written-description/indefiniteness of "fully encapsulated" and "mRNA" — has lived and died (so far) in district court, not at the Board.

Two structural facts drive that outcome: (i) the principal targets of this patent (Moderna, served 2022-02-28; Pfizer/BioNTech, sued 2023-04-04) burned — or are burning — their § 315(b) one-year windows; and (ii) per the Google Patents legal-status data, the '651 shows an adjusted expiration of 2023-07-15, so it is expired and today supports only backward-looking damages exposure, which materially reduces the ROI of an IPR. Note also that the '651's effective filing date is 2002/2003, so it is ineligible for PGR (post-AIA filing date required) and CBM is unavailable (not a financial-services patent; program sunset for new petitions anyway).

Adjacent proceedings — DIFFERENT PATENTS (context only; none involves the '651)

The following three IPRs are the entire known PTAB footprint of Arbutus's LNP portfolio. None of them names U.S. 9,504,651, none is in the '651's 2002 "Lipid compositions" family, and none creates estoppel on the '651. They are ordered by impact.

IPR (proceeding number not confirmed in sources reviewed; petition filed 2018-02-21) — Moderna Therapeutics, Inc. v. Arbutus Biopharma Corp. (U.S. Patent 9,404,127, "the '127 Patent")

• Type: Inter Partes Review
• Filed: 2018-02-21 (petition)
• Status: Final — all claims held unpatentable as anticipated; affirmed on appeal. (I could not confirm the IPR number from the sources retrieved; I am deliberately not guessing it.)
• Judge panel: Not confirmed in the sources reviewed.
• Petition grounds: § 102 anticipation (and § 103 in the alternative) against all claims.
• Institution decision: Instituted 2018-09-12.
• Final Written Decision: Issued 2019-09-10, holding all claims invalid by reason of anticipatory prior art (Arbutus 10-K disclosure). The decision was thereafter vacated and remanded in light of United States v. Arthrex; after the Supreme Court decided Arthrex on 2021-06-21, the Federal Circuit reinstated the appeal sua sponte, Arbutus waived its Arthrex challenge, and on 2023-04-11 the Federal Circuit affirmed, holding all claims of the '127 patent invalid as anticipated.
• Settlement / termination: None at the Board.
• Appeal: Yes — Federal Circuit affirmed the all-claims-invalid holding on 2023-04-11. (Docket number not confirmed in the sources reviewed.)
• Defensive value (for the sibling patent, not the '651): This is the one Moderna IPR that produced a complete kill. It shows the Board will invalidate Arbutus LNP claims on clean prior art — but those claims recited physicochemical structure ("non-lamellar"/"stable" particle morphology), not the formulation-with-encapsulation subject matter of the '651.

IPR2018-00739 — Moderna Therapeutics, Inc. v. Protiva Biotherapeutics, Inc. / Arbutus Biopharma Corp. (U.S. Patent 9,364,435, "the '435 Patent")

• Type: Inter Partes Review
• Filed: 2018 (exact petition date not confirmed in the sources reviewed)
• Status: Final — mixed: certain challenged claims canceled; claim 7 sustained as patentable.
• Judge panel: Not confirmed in the sources reviewed.
• Petition grounds: § 102 / § 103 over lipid-formulation prior art (the Board's opinion turned on overlapping mol-% ranges and the In re Peterson / In re DuPont presumption of obviousness).
• Institution decision: Instituted (date not confirmed).
• Final Written Decision: FWD dated 2019-09-11 (Paper 51). The Board held claim 7 patentable and, per the Federal Circuit, held certain other claims unpatentable as anticipated. Arbutus's own disclosure describes the decision as "holding certain claims invalid and upholding [claim 7]." I could not verify the complete per-claim list from the sources retrieved, so I am not reciting claim numbers beyond claim 7.
• Settlement / termination: None at the Board.
• Appeal: Yes — ModernaTx, Inc. v. Protiva Biotherapeutics, Inc., Nos. 2020-1184, -1186 (Fed. Cir.). On 2021-12-01 the Federal Circuit dismissed Moderna's appeal for lack of standing (18 F.4th 1352) — the license-based injury was too speculative — while affirming, on Arbutus's cross-appeal, the Board's unpatentability holdings as to the claims Moderna won against. Net effect: the invalid claims are dead; there is no Article III merits review of the claim-7 patentability holding.
• Defensive value: A partial win against a sibling patent. Note the standing trap: an IPR petitioner who is a licensee may win at the Board and still be unable to appeal.

IPR2019-00554 — Moderna Therapeutics, Inc. v. Arbutus Biopharma Corp. (U.S. Patent 8,058,069, "the '069 Patent")

• Type: Inter Partes Review
• Filed: 2019-01-09
• Status: Final — patent owner prevailed; claims 1–22 held not unpatentable.
• Judge panel: Not confirmed in the sources reviewed.
• Petition grounds: § 103 over WO 2005/007196 ("the '196 publication") and U.S. 2006/0134189 ("the '189 publication"), alone and in further combination with Lin and Ahmad; alternatively § 102/§ 103 over U.S. 2006/0240554 ("the '554 publication"). Challenged claims 1–22.
• Institution decision: Instituted (date not confirmed; the Board's FWD in the '739 IPR issued after institution of this proceeding, per the patent owner's brief).
• Final Written Decision: Issued July 2020 (the D. Del. exhibit list identifies a Judgment, IPR2019-00554, Paper 40, dated 2020-07-23). The Board found Moderna failed to show claims 1–22 unpatentable, reasoning that the asserted references "do not anticipate or otherwise render obvious a nucleic acid-lipid particle containing each of the recited lipid components within the claimed ranges, including specifically a phospholipid range of 4–10%," and declining to apply a presumption of obviousness from overlapping ranges.
• Settlement / termination: None at the Board. (Arbutus later dropped the '069 patent from the Delaware action.)
• Appeal: Yes — ModernaTx, Inc. v. Arbutus Biopharma Corp., No. 2020-2329 (Fed. Cir.). On 2021-12-01 the Federal Circuit affirmed the nonobviousness holding, and found standing (unlike in the companion '435 appeal) based on Arbutus's conduct toward Moderna's COVID-19 vaccine.
• Defensive value: Demonstrates the Board's hostility to "overlapping ranges = obvious" arguments against Arbutus's Lipid Formulation claims — a caution for anyone contemplating a § 103 attack built on range overlap, though the '651's claims are differently drafted (encapsulation percentages, not mol-% ranges).

Strategic summary

Claim status of the '651. Nothing has been canceled, confirmed, or even reviewed. The patent is untested at the PTAB in its entirety. For reference, representative claims are claim 1 (a lipid vesicle formulation comprising a cationic lipid, an amphiphilic lipid, a PEG-lipid, and mRNA "wherein at least 70% of the mRNA ... is fully encapsulated") and claim 9 (claim 1 wherein each lipid vesicle is a lipid-nucleic acid particle); dependent claims add ≥80% (claim 13) and about 90% (claim 14) encapsulation. Arbutus has asserted this patent against Moderna (D. Del. 1:22-cv-00252-JDW, filed 2022-02-28) and against Pfizer/BioNTech (D.N.J., filed 2023-04-04), and Acuitas sought a declaratory judgment of invalidity/non-infringement in S.D.N.Y. 1:22-cv-02229 — but nobody petitioned the Board. Per Google Patents, the '651 shows an adjusted expiration of 2023-07-15, so it is expired.

Estoppel landscape. Because no FWD ever issued on the '651, § 315(e)(2) estops no party as to the '651's claims, and § 315(e)(1) estops no party from petitioning the Office on those claims. Conversely, nobody can benefit from an earlier PTAB record either — there is no Board finding to borrow. The estoppel that exists is entirely on sibling patents: Moderna is estopped (as to the '069 and the claims of the '435 and '127 that reached FWD) from re-running those grounds. That spillover is contested, not settled: in the Delaware case, Arbutus argued issue preclusion/estoppel flowing from the '069 FWD into the broader Ratio Patents ('435, '359, '378), while Moderna argued that § 315(e)(2) cannot bar district-court invalidity defenses where no Article III court reviewed the agency decision (true for the '435 claim-7 holding, which was never merits-reviewed because Moderna's appeal was dismissed for standing). For a defendant on the '651, none of that matters: prior-art grounds you wish to raise on the '651 are unencumbered by any PTAB estoppel. The real constraints are statutory timing (§ 315(b)) and the practical one — an IPR judgment on an expired patent buys you only a defense to past damages.

Pattern signals. One petitioner — Moderna — filed three IPRs against three Arbutus patents (the '127, '435, and '069), an early, targeted campaign that produced one complete kill, one partial kill, and one loss. Notably, Moderna never petitioned on the '651 or on any patent in the '651's 2002 "Lipid compositions for nucleic acid delivery" family — its petitions targeted the 2009 "Lipid formulations" family and the '127 patent. The patent owner, Arbutus (with Genevant as exclusive licensee), has been an aggressive defender and appellant, including a successful cross-appeal that locked in the invalidation of certain '435 claims, and a win in the '069 appeal. No defensive aggregator (Unified Patents or similar) appears anywhere in this chain — the only Unified Patents material I found is its third-party patent/litigation database, not a petition. Note that Arbutus also faced EPO oppositions by Moderna and Merck to EP 2 279 254 (upheld as amended on 2024-06-06; on appeal), which is EPO, not PTAB.

Recommended next steps

1. Verify the null result at the source. Confirm zero AIA trials on US 9,504,651 via PTAB E2E (https://ptacts.uspto.gov/ptacts/) and the USPTO Patent Center/ODP for application 14/304,578. Absent a newly indexed petition, treat "no PTAB activity on the '651" as established — and note that the absence is itself a signal for a heavily asserted patent: three sophisticated defendants with every incentive to file (and one that filed three IPRs on the siblings) chose not to take the '651 to the Board, most plausibly because of the § 315(b) window and the patent's 2023 expiry.
2. If you are a newly-served defendant, check your § 315(b) clock first. Pfizer/BioNTech were sued 2023-04-04 and Moderna 2022-02-28, so both are already time-barred on the '651; if you received the complaint within the last 12 months, an IPR is still an open option, and it would be the first-ever PTAB challenge to this patent. Because the patent is expired, frame the IPR as a defense to past damages, not as prospective freedom to operate.
3. If you are a defendant relying on the '651's invalidity story, build it on the district-court record, not the Board. Moderna's Delaware contentions against the '651 are the deepest public roadmap: § 102(f) derivation from Moderna (with priority/§ 102(f) attacks on the 60/392,887 provisional), anticipation and obviousness over the Semple Article and Semple '278 (alone and with Malone 1989), obviousness over Bischoff + Adler-Moore 2002 + Zhang '745, and § 112 attacks on "fully encapsulated" (indefinite, non-enabled) and on the mRNA limitation (written description/enablement). The D. Del. court construed "fully encapsulated" on 2024-04-03 as requiring the mRNA be "fully, as distinct from partially, contained inside the lipid vesicles" — a construction that did not resolve the enablement fight, on which the court (D.I. 743, 2026-02-17) allowed Moderna's expert to testify.
4. Track the live litigation threads, which are where this patent now lives. The Moderna matter settled in 2026 (Moderna paid $950 million in July 2026, with $1.3 billion contingent on the § 1498 government-contractor appeal; Moderna took a global non-exclusive license to SM-102 mRNA vaccine technology, Arbutus/Genevant dropping the suit). The Pfizer/BioNTech D.N.J. case (asserting the '651 plus '359, '378, '320, and '098) remains pending with claim construction heard December 2024, and Arbutus/Genevant expanded internationally in July 2026 (Canada, UPC). The Federal Circuit appeal in the litigation block (No. 26-1581, Arbutus Biopharma Corp. v. Moderna, Inc., Moderna's opening brief dated 2026-05-29, on appeal from D. Del. 1:22-cv-00252-JDW) is a district-court appeal on 28 U.S.C. § 1498, not a PTAB appeal — do not miscite it as a PTAB review.
5. Recheck periodically. If a petition against the '651 ever appears, the operative deadlines would be: institution decision within 6 months of the PO's preliminary response, trial completion within 12 months of institution, and a hard statutory FWD deadline under § 316(a)(11).

Confidence and sourcing caveats. The "zero proceedings" conclusion rests on the ODP block plus the searches described above; a very recently filed, not-yet-indexed petition cannot be excluded. Where I could not verify a proceeding number (the '127 IPR) or a judge panel (all three sibling IPRs), or a complete per-claim list (the '435 FWD), I have said so rather than fill the gap. The '651's expiration date is taken from Google Patents' legal-status data (https://patents.google.com/patent/US9504651/en) and is labeled as an assumption there, not a legal conclusion. Nothing here is legal advice.

I'll research the assignment record and ownership chain for this patent. Let me run several targeted searches.

Let me dig for the actual reel/frame records and the Protiva corporate history.

Let me check the cross-license/Genevant structure and any security interests or additional recorded conveyances.

Assignment & Ownership Reconstruction — US 9,504,651 B2

Sourcing caveat (read first). I do not have authenticated access to USPTO Assignment Center (assignment.uspto.gov / assignmentcenter.uspto.gov), and I could not retrieve the underlying reel/frame numbers or the correspondent-of-record for this patent's recordings through the sources available to me. The full patent text supplied to me contains a Google Patents legal-events feed that lists two post-issuance-facing reassignment events, but that feed does not carry reel/frame or correspondent data. Consistent with the operating rules, I will not fabricate reel/frame numbers or attorney names. Where a field is unobtainable I mark it [NOT RETRIEVED]. The narrative ownership chain below is corroborated across USPTO PTAB filings, SEC filings (Arbutus 10-K/20-F/prospectus exhibits), the patent's own assignment feed, and district-court complaints.

Inventors

All four are Protiva personnel; the invention was assigned by the inventors to Protiva (see timeline). The 2002 priority date and 2003 parent filing mean the inventive work predates Protiva's 2008 acquisition by Inex/TPC (Tekmira), so "employer at filing" = Protiva, not Arbutus.

Literal-reading discrepancies to flag (do not auto-correct):

1. The recorded assignment event in the patent's legal-events feed names the assignors as "GIESBRECHT, NOELLE; JEFFS, LLOYD; MACLACHLAN, IAN; PALMER, LORNE R." — i.e., "Noelle" Giesbrecht, whereas the patent front page and every pleading name Cory Giesbrecht. This is almost certainly a data/OCR artifact in the feed, but I report it as it appears. (Cross-referenced to the earlier summary section, which reached the same flag.)
2. The Unified Patents record for US 9,504,651 additionally surfaces the names "Hirt, Erin" and "Haghighatian, Mina" in an inventor-adjacent field. These names appear on no patent front page or pleading I can find; they are likely a database artifact (possibly prosecution- or assignment-side metadata). [Status: unclear — not a finding.]

Unusual-pattern check — "inventors departing within 12 months": Not present. Maclachlan and the Protiva team remained with the entity through the Tekmira/Arbutus era; Maclachlan is described as a long-tenured Arbutus/Tekmira scientist. There is no evidence of a coordinated inventor exodus preceding a portfolio sale.

Original assignee

Protiva Biotherapeutics, Inc. (British Columbia, Canada). Named as original assignee on the issued patent.

• Primary line of business: preclinical/clinical-stage biotech focused on lipid-nanoparticle (LNP) nucleic-acid delivery — the platform later licensed to Alnylam (patisiran/ONPATTRO) and, ultimately, the technology family asserted against the mRNA vaccine makers.
• Product embodying the claims: Protiva/its successor did not itself market an approved product under the '651 claims. Its LNP platform enabled a commercial product — Alnylam's ONPATTRO (patisiran) — via license, and Arbutus states its LNP platform is the "only clinically-validated LNP delivery technology" and has "enabled an approved therapy." So the technology is practiced by licensees, not shipped by Protiva directly.
• Current status of the original assignee: dissolved as a separate entity. Protiva became a wholly-owned subsidiary of Inex/TPC (Tekmira) on May 30, 2008, and was amalgamated into Arbutus Biopharma Corporation effective January 1, 2018 (PTAB filings: "amalgamated … in January 2018"; Arbutus/SEC: "effective January 1, 2018"). Protiva no longer exists.

Assignment timeline

All entries below are drawn from the patent's legal-events feed and corroborating filings. Reel/frame and correspondent fields could not be retrieved and are marked as such. I found no evidence of any security agreement, release, license, or change-of-name recording against this patent from the sources available — treat "none found" as not confirmed rather than a negative.

1. 2014-06-26 (recorded) — Reel [NOT RETRIEVED]/[NOT RETRIEVED]

• Conveyance: Assignment of Assignors' Interest
• Assignor(s): GIESBRECHT, NOELLE; JEFFS, LLOYD; MACLACHLAN, IAN; PALMER, LORNE R.
• Assignee: Protiva Biotherapeutics, Inc.
• Correspondent: [NOT RETRIEVED]
• Context: Ordinary inventor-to-employer assignment perfecting title in the continuation application (US 14/304,578, filed 2014-06-13) in Protiva's name. Note this occurred while Protiva was already a wholly-owned Tekmira subsidiary; because Protiva remained a separate legal entity until 2018, title correctly stayed in Protiva, not Tekmira.

2. 2018-02-20 (recorded) — Reel [NOT RETRIEVED]/[NOT RETRIEVED]

• Conveyance: Merger (amalgamation)
• Assignor: Protiva Biotherapeutics Inc.
• Assignee: Arbutus Biopharma Corporation
• Correspondent: [NOT RETRIEVED]
• Context: Internal corporate reorganization — statutory amalgamation of the wholly-owned subsidiary into its parent (effective January 1, 2018; recorded February 20, 2018). This is a change-of-ownership-by-operation-of-law, not a sale to a third party.

Events I could NOT confirm and therefore do not assert:

• Any 2003-era inventor→Protiva assignment for the original application (10/611,274). The legal-events feed supplied to me begins at 2014 and does not show it. It may exist as a separate recording; [not confirmed].
• Any recording arising from the 2008 Inex/TPC–Protiva share purchase. That was a stock acquisition of the subsidiary, so no patent-level assignment would be expected — consistent with title remaining in Protiva.

Timeline diagram

timeline
    title Ownership of US 9504651
    2002 : Priority provisional filed
    2003 : Parent application filed by Protiva
    2008 : Protiva becomes Tekmira subsidiary
    2014 : Application filed 13 Jun
         : Inventors assign to Protiva 26 Jun
    2016 : Patent issued 29 Nov
    2018 : Protiva amalgamated into Arbutus
    2022 : Suit v Moderna filed
    2023 : Suit v Pfizer BioNTech filed

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. The only post-inventor transfer is a merger (2018: Protiva → Arbutus Biopharma Corporation), i.e., an internal amalgamation into a publicly traded operating parent (Nasdaq: ABUS). The assignee name is that of an operating biopharmaceutical company, not an "IP/Licensing/Holdings/Ventures" LLC, and there is no registered-agent-service address or single-purpose Delaware/Texas LLC in the chain.

2. Known asserter in the chain — NOT PRESENT. Neither Protiva Biotherapeutics, Inc. nor Arbutus Biopharma Corporation appears on the standard NPE lists (Acacia, Marathon, IV, IPNav, Wi-LAN, Conversant, Vringo, Pendrell, Innovatio, MPHJ, Round Rock, etc.) or as an RPX/Unified "high-frequency plaintiff." Arbutus is a clinical-stage drug developer and licensor that also asserts its patented platform.

3. Repeat correspondent across the chain — UNCLEAR / NOT RETRIEVED. I could not retrieve the correspondent-of-record for either recording, so I cannot assess recurrence. For completeness only (and expressly not a finding): the PTAB filings for sibling patents in this family were handled by Wilson Sonsini Goodrich & Rosati (Customer No. 21971) as litigation counsel for Protiva/Arbutus — litigation counsel is not the assignment correspondent, and WSGR is a large general-practice firm, so per the rules this does not count as a signal.

4. Cascading transfers — NOT PRESENT. Two recorded events, six years apart (2014 assignment; 2018 merger), with no chain of consecutive LLC-to-LLC transfers and no shared-principal shells.

5. Pre-litigation transfer — NOT PRESENT. The ownership consolidations (2008 acquisition; 2018 amalgamation) predate the first suit naming the '651 by roughly four to fourteen years (Moderna suit filed 2022-02/2022-04; Pfizer/BioNTech suit filed 2023-04). No assignment was recorded within six months before the first assertion.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 sale of this patent. Arbutus underwent 2024–2025 restructuring (workforce reductions, HQ exit, R&D wind-down per its 10-K/20-F narrative) but this is not a bankruptcy proceeding, and ownership of the LNP patents was unaffected.

7. Privateering — NOT PRESENT. There is a licensing joint venture — Arbutus licensed its LNP platform to Genevant Sciences (Roivant JV, April 2018), and Genevant co-plaintiffs the mRNA-LNP suits alongside Arbutus. However, Genevant is a product-developing licensee (RNA therapeutics pipeline, sublicensing), Arbutus retains ownership and appears as a co-asserting operating company, and the targets are direct market competitors (Moderna; Pfizer/BioNTech). This is a licensor-licensee co-assertion, not a straw-man NPE funded to attack rivals.

8. Defensive aggregator (anti-NPE) — NOT PRESENT. The chain terminates at Arbutus Biopharma Corporation; the patent has not been transferred to RPX, AST, LOT, Unified, or OIN.

Verdict

Operating-company assertion.

The ownership record shows only an ordinary inventor→employer assignment (recorded 2014-06-26 to Protiva Biotherapeutics, Inc.) followed by an internal subsidiary-into-parent merger (recorded 2018-02-20, Protiva → Arbutus Biopharma Corporation). Every NPE hallmark is absent: no shell-entity transfer, no known-asserter assignee, no cascade of LLCs, no pre-litigation assignment, no bankruptcy sale, no defensive-aggregator terminus. Arbutus is a publicly traded operating biopharma that generated a real, licensed platform (ONPATTRO via Alnylam) and is suing direct competitors (Moderna; Pfizer/BioNTech) — the textbook profile of an operating-company assertion rather than an NPE campaign.

Verification link: https://assignment.uspto.gov/patent/index.html (search "9504651" or app. 14/304,578) — I recommend pulling the two reel/frame records directly to capture the correspondents and confirm there are no additional conveyances (security agreement, release) not visible in the legal-events feed.

Residual uncertainties (explicit)

• Reel/frame numbers and correspondents of record were not retrievable and are the main gap; the timeline's documentary backbone (dates, parties, conveyance types) is corroborated, but the "troll-tell" field the task prioritizes is blank here.
• Total recorded conveyances: I confirmed two; I cannot rule out earlier (2003) or additional recordings.
• "Noelle Giesbrecht" vs. "Cory Giesbrecht" and the "Hirt, Erin"/"Haghighatian, Mina" names remain unexplained feed artifacts — flagged, not resolved.

I'll search for the specific patent and its cited prior art references.

Let me get the specific "References Cited" section from the patent face.

Let me search for the specific examiner-cited references and the patent's citation lists.

Prior Art Analysis — US 9,504,651 B2

Caveat up front (important): I was unable to retrieve the verbatim "References Cited (56)" block from the printed face of US 9,504,651. My search budget was exhausted before I could pull the full examiner/applicant citation list from the patent's front page or the file wrapper (Ex. 63 – 651 FH Excerpts appears on the D. Del. docket but I did not retrieve its contents). What follows is assembled from (a) the specification text supplied to me, (b) the Unified Patents citation record for this exact patent number, and (c) the prior-art references Moderna identified in its invalidity contentions as reported in the litigation record. I flag each source. I have not auto-corrected any number; where secondary sources show a variant (e.g., "9,504,561"), I say so.

1. The patent being searched

Because this is a continuation in an unbroken chain back to the 2002 provisional, the effective § 102 date for the granted subject matter is on or about 2002-06-28 (subject to the priority being supported — a point Moderna contested).

2. References cited in the specification itself

These are the only two documents I can state with high confidence are "citations for 9504651," because they appear verbatim in the Background of the patent text:

Note: both are prior art relative to the 2002 priority date. Their role is chiefly as admitted/background art and as § 103 combinators, not as § 102 anticipants of the granted mRNA composition claims.

3. References in the '651 citation record (Unified Patents "Patent Art" list)

Source: https://portal.unifiedpatents.com/patents/patent/9504651 (the record lists "Patent Art (177)"; I recovered the following leading entries). Interpretive caveat: the list mixes pre-2002 documents (which must be cited art) with later ones, so I cannot confirm that all of these are the printed "References Cited." Treat the pre-2002 items as the likely cited prior art and verify against the face.

4. Prior art Moderna actually asserted (litigation record)

Source: Moderna's invalidity contentions as summarized in the litigation record referenced in the previously generated sections (D. Del. 1:22-cv-00252; Ex. 78 – 2024-06-28 Moderna's Invalidity Contentions). These are the references against which claim 1 was actually tested:

(These entries come to me second-hand via the litigation record, not the patent face; treat the exact citations as needing verification against Moderna's contention exhibit.)

5. § 102 claim-mapping — bottom line

Overall: the cited/listed prior art is dominated by process, cationic-lipid, PEG-liposome, and DNA/oligonucleotide art from 1985–2002. None of the references I identified discloses, in a single reference, the full claim-1 combination of (i) a three-lipid-class vesicle and (ii) mRNA and (iii) ≥70% full encapsulation. The most threatening reference for the mRNA element is WO 2003/059381 (INSERM), whose priority/publication timing relative to 2002-06-28 is the key open question.

6. What I could not verify (stated limitations)

1. The verbatim printed "References Cited (56)" section. I could not retrieve it; the lists in § 3 come from Unified Patents' "Patent Art" compilation and are not confirmed to equal the face citations.
2. Whether the Unified Patents "Patent Art (177)" list is cited-by vs. cites. The pre-2002 entries must be cited art, but the list appears mixed.
3. Exact bibliographic data for Saravolac '171, Semple '278, Zhang '745, Meulien '831 — I have these only as names/numbers from the litigation record, not full citations.
4. The '651 file history (Ex. 63) — not retrieved; the examiner's actual § 102/§ 103 rejections and the references relied on are unconfirmed.
5. NO number auto-correction was performed. Note the conflicting secondary-source citation "US 9,504,561" (Arbutus Sept. 2025 8-K) — I read this literally and do not substitute '651.

If you can supply the printed front-page "References Cited" block or the file wrapper (Ex. 63 – 651 FH Excerpts), I can replace § 3–§ 5 with a verified, line-by-line citation table and a claim-by-claim § 102 chart.

I'll ground the prior-art references before analyzing. Let me verify the key references named in the record and the patent's own "Prior Art" material.

Let me resolve the remaining reference numbers ("Semple '278," "Zhang '745," "Meulien '831," "Wheeler '152") so I don't misidentify them.

Obviousness Analysis — US 9,504,651 B2 under 35 U.S.C. § 103

Note on scope and posture. This builds on the previously generated Patent Summary and Litigation Summary rather than repeating them. Two framing points carried forward:

1. The '651 is a pre-AIA patent. It claims priority through 10/611,274 (filed 2003-06-30) to provisional 60/392,887 (2002-06-28), so §§ 102/103 as they read before the AIA govern, with a PHOSITA as of mid-2002.
2. No court has adjudicated the '651's validity. Moderna's March 3–4, 2026 consent judgment included no invalidity as to the asserted patents, and the pending Federal Circuit appeal (No. 26-1581) is confined to 28 U.S.C. § 1498(a). The §103 question is therefore live principally in the D.N.J. Pfizer/BioNTech action — not in the Moderna appeal.

I. The claim at issue and the PHOSITA

Claim 1 is a composition claim: a lipid vesicle formulation comprising a plurality of vesicles, each containing (i) a cationic lipid, (ii) an amphipathic lipid, and (iii) a PEG-lipid, plus (iv) mRNA, wherein ≥70% of the mRNA is "fully encapsulated." Dependent claims add a sterol (claim 4), cholesterol + phospholipid (claim 6), a Markush list of phospholipids ending in DSPC (claim 7), "lipid-nucleic acid particle" morphology (claim 9), an ionizable cationic lipid charged only below physiological pH (claim 11), and higher encapsulation thresholds of ≥80% (claim 13) and ~90% (claim 14).

Critically — and this is the hinge of the whole analysis — claim 1 recites no process, no apparatus, no molar ratio, no particle size, and no specific lipid identity. The '651's specification is a 2002-vintage T-connector/ethanol-dilution manufacturing disclosure aimed at plasmid DNA ("SPLP"). The claims are broader than, and largely orthogonal to, that disclosure. That mismatch is what makes the prior art so easily mapped onto the claims.

The PHOSITA is a formulation scientist with an advanced degree and several years' experience in liposomal/lipid-nucleic acid delivery, familiar with the cationic-lipid/PEG-lipid/helper-lipid/sterol LNP architecture, with pH-gradient/electrostatic loading, with ethanol-dilution vesicle formation, and with the standard assays for distinguishing encapsulated from surface-associated nucleic acid (gel filtration, dye-binding with and without detergent, nuclease protection).

II. The prior art of record

Grounding note: several references named in the litigation record used short-form citations I could only partially verify. I identify below what I confirmed and flag what I did not.

Contradictions to flag. (a) My earlier Patent Summary listed "Semple '278" as a litigation reference; the number is unresolved and the underlying disclosure appears to be the US 6,858,225 family. (b) I could not verify "Zhang '745" or "Meulien '831" by number; treat those as record citations, not confirmed bibliographic facts. (c) My earlier summary said no IPR against the '651 was confirmed — that remains true, but the '651's own prosecution record (Retrievable at the D. Del. litigation file, https://archive.org/download/gov.uscourts.ded.78146/gov.uscourts.ded.78146.181.2.pdf) contains an examiner obviousness rejection over Saravolac that materially bears on §103 and was not discussed in the earlier sections.

III. The strongest §103 combinations

Combination A (primary) — Saravolac '171, optionally + Yoshioka/Adler-Moore (sterol), + Meulien '831 / Malone 1989 (mRNA)

This is not a hypothetical: it is the combination the '651's own examiner applied. The Office reasoned that "it would have been obvious … to make liposomes having a cationic lipid, a PEG-lipid, a sterol, and a phospholipid … because Saravolac teaches that it was known in the art to make liposomes and lipid-nucleic acid particles from cationic lipids, PEG-lipids, a sterol, and fusogenic lipids which allow for increased encapsulation efficiencies of over 80%, specifically up to about 86%," and that "it is well known in the art that mRNA is formed of nucleic acids" (file history, Office action reproduced at gov.uscourts.ded.78146.181.2). Mapping:

• Cationic lipid / amphipathic lipid / PEG-lipid per vesicle → Saravolac '171 at 13:9–16 (preferred compositions comprise cationic lipids as bilayer-forming/‑stabilizing lipids, fusogenic lipids, and PEG-lipids; DODAC, DOTMA, DOTAP, DC-Chol, DMRIE), 13:22–28, 14:12–21, 15:39–43.
• mRNA → Saravolac's "nucleic acids" includes RNA (9:6–13); Meulien '831 and Malone 1989 supply express mRNA teaching.
• ≥70% fully encapsulated → Saravolac teaches 40–80% encapsulation (2:45–54; 3:6–12; 14:43–45), up to ~86% (Fig. 8), plus removal of substantially all unencapsulated nucleic acid (15:13–18), which the examiner reasoned "would increase the percent of the nucleic acid in the formulation that is within the lipid vesicles beyond 80%."
• Sterol / cholesterol (claims 4, 6) → Saravolac 8:55–57; Yoshioka (cholesterol stabilizes membranes).
• Markush phospholipids (claim 7) → the list is the conventional phospholipid genus repeated verbatim in the '651's own definitions.
• "Lipid-nucleic acid particle" (claim 9) and ionizable cationic lipid (claim 11) → Saravolac's lipid-bilayer-nucleic acid carriers and its DODAP/DODMA-type lipids.

Combination B — Semple family ('591/'225/'153) alone or + Malone 1989/Meulien '831

Semple's disclosures map to claim 1 almost element-for-element and go further on the two limitations Moderna's own expert and the examiner found hardest:

• The 25:45:20:10 DSPC:Chol:DODAP:PEG-CerC14 architecture supplies all three lipid classes.
• Semple's ionizable-lipid teaching ("cationic at pH below the pKa and neutral at pH above," pKa 4–11, DODAP) directly satisfies claim 11.
• Most importantly, the family defines "lipid-encapsulated … particle" as one where ≤50%, preferably ≤10%, of nucleic acid is externally detectable, and US 2005/0255153 claim 105 uses the phrase "fully encapsulated." That is a near-verbatim antecedent for the construed limitation ("fully, as distinct from partially, contained inside the lipid vesicles," D. Del., Apr. 4, 2024). A prior-art composition meeting Semple's ≤10%-external criterion is, on the court's construction, one in which the nucleic acid is fully encapsulated.
• Encapsulation >50%, >60%, and up to ~80% brackets the 70% threshold; the 80% and ~90% thresholds of claims 13–14 are met by Saravolac's ~86% and by routine optimization.

Semple '591 issued 2001-09-11, more than one year before the 2002-06-28 priority date — § 102(b) art. Semple '225 and '153 are § 102(e)-dated to their 1998/2001 parents; whether the '153's added "fully encapsulated" claim language itself is entitled to that benefit is a point a challenger would need to shore up.

Combination C — Wheeler '860 or Knopov WO 01/05373 + Semple family / Saravolac

Wheeler '860 is admitted prior art on the '651's face, and Knopov is admitted and distinguished in the specification. Both disclose lipid-vesicle/nucleic-acid particle technology of the same class. Because claim 1 is a composition claim, the '651's own specification's distinction of Knopov — that Knopov uses a turbulent static mixer while the '651 uses a non-turbulent T-connector — is legally irrelevant to claim 1's patentability. That is an unusually clean §103 argument: the applicant's own characterization of Knopov as a different apparatus still concedes that Knopov's compositions (ethanolic-lipid/citrate-buffer vesicles, DODAP/DSPC/Chol/PEG-ceramide, "fully lipid-encapsulated" oligonucleotide particles) were known.

Combination D — the "obvious to try" case for claims 11, 13, 14

The finite set of known ionizable lipids (DODAP, DODMA, DMDMA — recited in the '651's own specification as "cationic and hav[ing] a positive charge at below physiological pH") makes claim 11 a textbook case of a known element working a known function. Likewise, claims 13–14 recite only higher values of the same result-effective variable, with no recited structural change; absent evidence of unexpected results or "criticality," range-narrowing is ordinarily obvious. See In re Woodruff and In re Peterson; the counterpart is KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), "'a finite number of identified, predictable solutions."

IV. Motivation to combine

The KSR factors line up unusually well for the challenger:

1. Same field, same problem. All four families come from the same Vancouver LNP cluster (Inex/Protiva/UBC), address the same problem — serum-stable, systemically deliverable, high-efficiency lipid-encapsulated nucleic acid particles — and share inventors/assignees (Cullis, Hope, Ansell, Semple appear across the Semple/UBC and Inex families; Protiva acquired the portfolio; Arbutus now holds both sides).
2. Known elements, known functions. Cationic/ionizable lipid (charge pairing with polyanion), helper phospholipid (bilayer formation), cholesterol (membrane stabilization — expressly the examiner's Yoshioka rationale), and PEG-lipid (steric stabilization/aggregation prevention — Saravolac 13:56–65; Semple 18:10–16). This is the classical one-part "LNP," and the art repeatedly framed it as four interchangeable components.
3. Explicit teaching that the cargo is RNA. Semple's own family enumerates antisense, ribozyme and RNA-encoding-a-therapeutic-protein; Saravolac teaches "DNA or RNA"; Malone 1989 encapsulates mRNA. There is no teaching away from mRNA — only the observation that the 2002 specifications were exemplified with pDNA/antisense, which is a disclosure-adequacy (§ 112) point, not an obviousness point.
4. Reasonable expectation of success. The advance over the art is charge-mediated, not sequence- or base-mediated. mRNA is a polyanion; the mechanism by which Saravolac/Semple encapsulate a polyanionic nucleic acid does not depend on whether the polyanion is plasmid DNA, antisense, ribozyme, or mRNA. The expert report relies precisely on this (Dkt. 667-28, ¶ 248: "a POSA would have had a reasonable expectation of success … because of the disclosure of mRNA in Saravolac '171 and Meulien '831['s] teaching that lipid formulations of phospholipids and cholesterol could encapsulate mRNA").
5. Design incentive / no teaching away. The art supplies a finite menu of lipids and a stated desire to raise encapsulation toward and above the 70% level; both Saravolac (removal of free nucleic acid to push encapsulation >80%) and Semple (surface neutralization to strip external nucleic acid) teach the exact technique for reaching the claimed thresholds.

V. Where the obviousness case is weakest

Intellectual honesty requires identifying the counterarguments a patentee would press:

• "Fully encapsulated" ≠ "40–80% encapsulation." The '651's own Rule 132 declaration (Dr. James Heyes, dated 2015-12-07, filed 2015-12-14) reported that following Saravolac's procedure produced only ~50% encapsulation — below claim 1's 70% threshold — and the applicant argued the examiner had not "completely follow[ed] the method" of Saravolac. That is a genuine fact question on the "fully encapsulated" limitation, and the examiner's rubric ("mRNA is made of nucleic acids") is exactly the kind of generic-genus reasoning that appellate courts sometimes reject when the prior art discloses a large genus without direction to the claimed species. The Semple "fully encapsulated" language materially strengthens the challenger here; Saravolac alone is weaker.
• The "fully, as distinct from partially" construction is a quality, not a quantity. A reference that reports 86% total encapsulation does not necessarily report 86% fully encapsulated, since surface-associated nucleic acid can be counted in a high total figure. Post-2024, the challenger should map the 70% limitation to a reference that excludes externally detectable nucleic acid (Semple ≤10% external), not merely to a high total-efficiency number.
• Presumption of validity and prosecution history. The claims issued notwithstanding the Saravolac rejection, so the applicant presumably overcame it (amendment and/or traversal); the Federal Circuit's KSR jurisprudence cuts both ways, and the burden is clear-and-convincing.

VI. Conclusion

On the record available, the strongest § 103 showing is Saravolac '171 in view of a sterol-stabilization reference (Yoshioka / Adler-Moore 2002), further in view of Meulien '831 and Malone 1989 for the mRNA limitation — which is essentially the combination the USPTO examiner already applied against claim 1, with the caveat that the Heyes declaration raised a contested factual issue as to whether Saravolac's method actually reaches ≥70% "fully encapsulated." A second, arguably cleaner, showing is the Semple family (US 6,287,591; 6,858,225; and the US 2005/0255153 "fully encapsulated" claims) alone or in view of Malone 1989, because that family supplies all three lipid classes, the ionizable-lipid limitation, and a definition of encapsulation that aligns with the court's "fully, as distinct from partially" construction. Wheeler '860 and Knopov WO 01/05373 are useful supplements (and admissions on the '651's face), and claims 11, 13 and 14 are vulnerable to a straightforward "obvious to try / optimization of a result-effective variable" attack grounded in the finite list of ionizable lipids recited in the '651's own specification.

Caveat: this is an analyst's technical assessment, not legal advice and not a court holding; the '651 has never been held invalid, and Moderna consented to a judgment of no invalidity as to it.

Primary sources used: the patent text at https://patents.google.com/patent/US9504651/en; the '651 prosecution record (examiner's Saravolac rejection and Heyes Rule 132 declaration) at https://archive.org/download/gov.uscourts.ded.78146/gov.uscourts.ded.78146.181.2.pdf and .181.3.pdf; Moderna's invalidity expert report (Saravolac/Semple claim charts) at https://storage.courtlistener.com/recap/gov.uscourts.ded.78146/gov.uscourts.ded.78146.667.28.pdf; Saravolac US 6,734,171 at https://patentimages.storage.googleapis.com/88/dc/19/bb5284b717f9e7/US6734171.pdf; Semple US 6,858,225 at http://www.alnylam.com/Files/IP/US6858225-Semple.pdf and US 2005/0255153 at https://uspto.report/patent/app/20050255153; Knopov WO 01/05373 at https://portal.unifiedpatents.com/patents/patent/WO-2001005373-A1; Wheeler US 5,478,860 at https://patents.justia.com/patent/[5478860](/patent/5478860).