US 10993914

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

The first pass didn't surface information specific to 10993914. Let me refine.

Let me check specifically for any 2026 Federal Circuit activity involving this patent.

US Patent 10,993,914 — "Stabilizing camptothecin pharmaceutical compositions"

Important scoping note: I searched for the literal identifier 10,993,914 / US10993914B2. I did not find any Federal Circuit (CAFC) 2026 docket or opinion in which this patent is the appealed subject. The only 2026 appellate-adjacent activity I could find in this family relates to a different patent (U.S. 11,344,552) at the PTAB. Details and caveats below.

---

1. Bibliographic data (authoritative from the patent text and cross-checked)

Sources: https://patents.google.com/patent/US10993914/en ; https://patents.justia.com/patent/10993914

2. Abstract (as published)

"Irinotecan phospholipid liposomes with improved storage stability are provided, with related methods of treatment and manufacture. The irinotecan liposomes can have reduced formation of lyso-phosphatidylcholine (lyso-PC) during storage, and prior to administration to a patient."

Source: https://www.drugpatentwatch.com/p/patent/10993914 (the abstract text was not separately reproduced in the fetched Google Patents body text I was given).

Commercial relevance: the patent is listed in the Orange Book against ONIVYDE (irinotecan liposome injection), NDA 207793-001, listed expiry 2036-10-15. Same source also records ~37 family members across 19 countries.

3. Plain-language overview of the independent claim(s)

I have verbatim claim text only for independent claim 1 and dependent claims 2–11 (the secondary source truncated the remainder). I could not verify the full 22-ish claim set, including whether a separate independent process/method claim issues. One source classifies the patent's claims as "Composition; Process," which suggests at least one independent process claim exists, but I cannot confirm its number or wording — treat that as unverified.

Claim 1 (independent — composition, defined in part by its manufacturing process):
A storage-stabilized liposomal irinotecan composition containing irinotecan sucrose octasulfate (SOS), cholesterol, and one or more phospholipids, where:

• the irinotecan-to-phospholipid ratio is 500 g ±10% irinotecan moiety per mole of total phospholipids;
• the composition is stabilized so that it has less than 20 mol% lyso-PC (relative to total phospholipids) after 180 days / 6 months of storage at 2–8 °C; and
• it is obtained by a process of: (a) forming liposomes of DSPC + cholesterol + methoxy-terminated PEG-distearoylphosphatidylethanolamine encapsulating a substituted ammonium salt of SOS at 0.4–0.5 M sulfate; (b) incubating those liposomes with an aqueous irinotecan solution to load irinotecan as irinotecan sucrose octasulfate inside the liposomes; and (c) adjusting the pH of the preparation to about 7.25–7.50.

In plain terms: a refrigerated liquid liposomal-irinotecan product, made in a specified three-step way (make "empty" SOS liposomes → load irinotecan → bring pH up to ~7.3–7.5), defined by the practical outcome that it does not degrade into more than 20 mol% lyso-PC within ~6 months in the fridge.

Dependent claims 2–11 (verified verbatim through 10, partial at 11): unilamellar liposomes (2); ~4.3 mg/mL irinotecan free base (3); <20 mol% lyso-PC out to 9–12 months (4); isotonic HEPES buffer at 2–20 mM (5); NaCl at 130–160 mM (6); irinotecan in a gelated/precipitated state as the SOS salt (7); the PEG-lipid is MPEG-2000-DSPE (8); DSPC 6.1–7.5 mg/mL, cholesterol 2–2.4 mg/mL, MPEG-2000-DSPE 0.11–0.13 mg/mL (9); the specific 6.81 / 2.22 / 0.12 mg/mL levels (10); storage temperature about 4 °C (11).

Source for claim text: https://www.drugpatentwatch.com/p/patent-claims/10993914

Technical concepts the specification relies on (useful context, not claim language): a "Stability Ratio" (SR) = (g irinotecan moiety per mol phospholipid) ÷ (molar sulfate-group concentration in the trapping-agent solution), with SR > ~950 (preferably > 992) taught as predictive of lower lyso-PC formation; and an "ER" (irinotecan/sulfate gram-equivalent ratio) of >0.9 (e.g., 0.9–1.1). The specification reports a comparator at SR ≈ 724 / pH 6.5 generating >30–35 mol% lyso-PC, versus SR ≈ 990–1047 / pH 7.25–7.5 generating far less.

4. Litigation / PTAB / appellate status (as of this search)

• District court (D.N.J., not CAFC): Ipsen Biopharmaceuticals Inc. et al. v. Conjupro Biotherapeutics Inc. et al., No. 1:24-cv-04991 (RMB)(MJS), filed 2024-04-15, later consolidated with 1:24-cv-08723 and 1:25-cv-13647. 10,993,914 is expressly named in Conjupro's Paragraph IV notice letter and pleaded as one of the "Initial Patents-In-Suit." Defendants/related parties include CSPC Pharmaceutical Group, CSPC Ouyi, Conjupro Biotherapeutics, and Apotex. Listed expiry for '914 in that case: 10/15/2036. Sources: https://paragraphfour.com/wp-content/uploads/2024/04/njdc24cv8723A.pdf ; https://paragraphfour.com/wp-content/uploads/2024/04/njdc25cv13647A.pdf ; https://paragraphfour.com/members/parryashford/activity/11735/
• PTAB (2026 activity in the same family, but a different patent): IPR2025-00505 (CSPC/CSPC Ouyi/Conjupro/Apotex v. Ipsen) challenges claims 1–15 of U.S. 11,344,552; Apotex's IPR2025-01531 was joined to it (decision entered 2026-03-12), with oral argument scheduled 2026-05-19. These proceedings do not concern 10,993,914. Sources: https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557313](/patent/1557313)/download-documents (Ipsen's April 6, 2026 oral-argument request); NYIPLA March 2026 PTAB committee materials (https://www.nyipla.org/images/nyipla/Programs/2026/March17PTABMeeting/March%20PTAB%20Committee%20Meeting_.pdf).
• CAFC 2026: No appeal involving 10,993,914 was located. I searched for Ipsen/Onivyde Federal Circuit activity in 2026 and found none tied to this patent number.

5. Explicit uncertainties

1. I could not retrieve the complete, official claim set for 10,993,914 (post-claim-11 text was truncated in the available source). The number of independent claims and the existence/wording of any independent process claim are unverified.
2. The abstract quoted is drawn from a secondary database (drugpatentwatch); I did not see an abstract block in the fetched Google Patents body text, so I cannot independently verify it character-for-character.
3. The "anticipated expiration 2036-10-15" is a computed estimate from Google Patents' legal-status metadata, not a legal conclusion, and would be affected by any PTE/PTA or terminal disclaimer.
4. Assignee history: Google Patents records Merrimack → Ipsen reassignments recorded 2021-03-22 (after the 2017 Merrimack–Ipsen asset sale), consistent with the inventor list originating at Merrimack.
5. The PTAB list of "IPR2026-00083/-00084/-00085/-00093" etc. that appeared in the 2026 PTAB materials could not be mapped to 10,993,914; do not assume any of those target this patent.

I'll research litigation involving US 10993914 using the sources you specified.

Let me dig deeper into the specific cases and look for any additional defendants.

Let me verify whether the '914 patent was affirmatively asserted or only subject to counterclaims, and check for any other defendants or IPR proceedings.

Bottom line

I found one litigation matter in which US Patent 10,993,914 ("the '914 patent") is implicated, plus two consolidated companion cases. All are in the U.S. District Court for the District of New Jersey. No Federal Circuit appeals and no IPRs directed at the '914 patent itself were found.

Important caveat before the details: the '914 patent appears in this matter through the defendant's declaratory-judgment counterclaims (triggered by a Paragraph IV certification), rather than as a patent Ipsen affirmatively asserted. This distinction is documented in the pleadings themselves. I explain it below and flag where sources disagree.

---

Case details

1. Ipsen Biopharmaceuticals, Inc. et al. v. Conjupro Biotherapeutics, Inc. et al. — D.N.J. 1:24-cv-04991 (RMB)(MJS) (lead consolidated case)

Patents affirmatively asserted in the April 15, 2024 complaint: 8,329,213; 9,339,497; 9,364,473; 9,452,162; 9,492,442; 9,717,724; 10,980,795; 11,369,597. The '914 patent is not in that list.

How the '914 patent enters the case: Conjupro's Paragraph IV Notice Letter (received by Ipsen "at the latest by March 7, 2024") certified that, among others, U.S. Patent No. 10,993,914 ("the '914 Patent") was invalid, unenforceable, and/or not infringed by Conjupro's NDA product. Conjupro/CSPC then filed counterclaims seeking declaratory judgments of non-infringement and/or invalidity of the '914 patent (along with the '867 and '360 patents), expressly alleging that Ipsen had not sued on those three patents. See the counterclaim text: "On April 15, 2024, Counterclaim-Defendants filed an infringement action ... alleging infringement of the '213 patent, the '497 patent, the '473 patent, the '162 patent, the '442 patent, the '724 patent, the '795 patent, and the '597 patent. Upon information and belief, Counterclaim-Defendants have not sued Counterclaim-Plaintiffs for infringement of the '867 patent, the '360 patent, or the '914 patent."

Source conflict noted: Paragraph Four Associates lists 10,993,914 (expiry 10/15/2036) among the "Litigated Patents" for this case, whereas DrugPatentWatch's litigation page for 1:24-cv-04991 does not list '914 among the asserted patents. Reading the pleadings literally, the '914 patent is in the case via the DJ counterclaims, not via Ipsen's affirmative counts.

2. Ipsen Biopharmaceuticals, Inc. et al. v. Conjupro Biotherapeutics, Inc. et al. — D.N.J. 1:24-cv-08723 (RMB)(MJS)

3. Ipsen Biopharmaceuticals, Inc. et al. v. Conjupro Biotherapeutics, Inc. et al. — D.N.J. 1:25-cv-13647 (RMB)(MJS)

---

Related PTAB proceedings (none on the '914 patent)

• IPR2025-00505 — Conjupro/CSPC (real parties in interest: CSPC Pharmaceutical Group Limited, CSPC Ouyi Pharmaceutical Co., Ltd., Conjupro Biotherapeutics, Inc.), filed January 17, 2025, challenging U.S. Patent 11,344,552 — not the '914 patent. Instituted for trial; oral argument reported scheduled May 19, 2026.
• IPR2025-01531 — Apotex Inc., filed September 12, 2025, also on U.S. Patent 11,344,552; joined with the above.
• I found no IPR, PGR, or CBM petition directed at US 10,993,914.

---

No CAFC / PACER appellate activity found

No appeal to the U.S. Court of Appeals for the Federal Circuit involving the '914 patent was identified. The case remains at the district-court level.

---

Context confirming the patent's commercial significance

• US 10,993,914 B2 is owned by Ipsen Biopharm Ltd., titled "Stabilizing camptothecin pharmaceutical compositions," application 16/567,902, granted May 4, 2021, with an anticipated expiration of October 15, 2036 (Google Patents).
• It is listed in the Orange Book for ONIVYDE® (irinotecan liposome injection), NDA 207793, applicant Ipsen, with a patented use; DrugPatentWatch lists it as protecting Onivyde.
• The Google Patents record itself carries the flag "Family has litigation — First worldwide family litigation filed" (Darts-ip family 57219018), consistent with the New Jersey matter above.

---

What I could not confirm (stated explicitly rather than guessed)

1. Whether Ipsen ever affirmatively added the '914 patent to the consolidated action (e.g., by amended complaint). The sources I retrieved show it only as a counterclaim patent; I did not find a count of infringement on the '914 patent.
2. Any other defendants or jurisdictions (e.g., other generic/505(b)(2) filers such as Apotex or Cipla, the latter appearing only in an RPI dispute in the '552 IPR, not as a defendant on the '914 patent).
3. Any merits outcome on the '914 patent — the consolidated case is reported as still in discovery, and any invalidity/non-infringement adjudication of the '914 patent has not occurred.
4. A trial or case-management schedule specific to the '914 patent.

Caveat on dates: Some retrieved pages carried "last updated" stamps later than the date you gave me (e.g., August 2026). I report the case status as reflected in those records; docket status in the lead case (1:24-cv-04991) should be verified directly via PACER for the current position.

---

Key sources

• Paragraph Four Associates activity record (case name, court, case number, filing date, judge, litigated patents incl. 10,993,914): https://paragraphfour.com/members/parryashford/activity/11735/
• Conjupro/CSPC counterclaims identifying the '914 patent and the Paragraph IV Notice Letter: https://paragraphfour.com/wp-content/uploads/2024/04/njdc25cv13647A.pdf and https://paragraphfour.com/wp-content/uploads/2024/04/njdc24cv8723A.pdf
• Justia docket for 1:24-cv-08723 (parties, filing date, Oct. 18, 2024 consolidation order): https://dockets.justia.com/docket/new-jersey/njdce/1:2024cv08723/[553645](/patent/553645)
• DrugPatentWatch litigation page for Ipsen v. Conjupro (1:24-cv-04991) and patent page for 10,993,914: https://www.drugpatentwatch.com/p/patent/10993914
• Google Patents record for US10993914B2 (assignee, expiry, family-litigation flag): https://patents.google.com/patent/US10993914/en
• PTAB institution decision in IPR2025-00505 ('552 patent; RPIs): https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557313](/patent/1557313)/download-documents

Summary: Litigation touching US 10,993,914 is limited to the consolidated D.N.J. action Ipsen Biopharmaceuticals, Inc. et al. v. Conjupro Biotherapeutics, Inc., CSPC Pharmaceutical Group Limited, and CSPC Ouyi Pharmaceutical Co., Ltd., No. 1:24-cv-04991 (RMB)(MJS) (filed April 15, 2024; still in discovery), where the '914 patent is pleaded as a declaratory-judgment counterclaim patent arising from Conjupro's Paragraph IV certification — not as an affirmatively asserted patent by Ipsen.

I'll verify the structured data against live sources before reporting.

Let me dig further on whether any AIA trial specifically names the '914 patent.

Let me confirm whether any petition or appeal directly names the '914 patent, and check the related-patent IPR landscape.

Proceedings overview

Total AIA trial proceedings on US 10,993,914: zero. The structured "PTAB proceedings on file" block from the USPTO Open Data Portal returns no AIA trial proceedings, and my targeted searches (patent number, patentee, petitioner names, IPR2024/IPR2025/IPR2026 prefixes) surfaced no IPR, PGR, or CBM naming this patent — so the breakdown is 0 active / 0 invalidated / 0 sustained / 0 settled / 0 institution-denied. Bottom line for a defendant: the '914 has never been tested at the Board. No claim of US 10,993,914 is canceled, no claim has been adjudicated, and no § 315(e)(2) estoppel attaches to any party with respect to this patent. The patent is untested, not hardened — the two are very different postures, and the art being developed in the pending IPRs against its sibling patent (US 11,344,552) is a public roadmap, even though estoppel from those proceedings is patent-specific and does not reach the '914.

Important framing caveat: several Onivyde® family challenges exist, and Ipsen has actively litigated this family — but on other patents. I flag those below expressly as not proceedings on US 10,993,914. Two facts anchor the count:

• Google Patents lists no PTAB "AIA trial" linkage for the '914; its only litigation tag is the family-level Darts-IP litigation entry (family 57219018) — the "Family has litigation" flag on the '914 Google Patents page reflects district-court litigation, not PTAB activity.
• In the D.N.J. litigation, Conjupro's own counterclaims state that Ipsen did not sue Conjupro on the '914 (or the '867 or '360 patents) in the April 15, 2024 complaint, even though the '914 was named in Conjupro's Paragraph IV notice letter and is the subject of Conjupro's declaratory-judgment counterclaims. (Conjupro counterclaims, via PTAB E2E document viewer)

---

No proceedings to report for US 10,993,914

There is nothing to summarize at the proceeding level. There is no petition number, no institution decision, no Final Written Decision, no termination, and no Federal Circuit appeal on this patent. I will not manufacture one. If you need a negative for a diligence file: as of the most recent ODP ingest and as of my searches, the Board has never considered US 10,993,914.

---

Adjacent proceedings (NOT on US 10,993,914 — context only)

These are the closest live challenges in the same family. They name US 11,344,552 ("Method of Treating Pancreatic Cancer," NALIRIFOX regimen), a different patent in the Onivyde family, not the '914.

IPR2025-00505 — CSPC Pharmaceutical Group Ltd., CSPC Ouyi Pharmaceutical Co., Ltd., and Conjupro Biotherapeutics, Inc. v. Ipsen Biopharm Ltd.

• Type: Inter Partes Review
• Filed: 2025-01-17
• Status: Trial instituted (per public reporting, institution granted 2025-08-14)
• Patent challenged: US 11,344,552 — not the '914
• Judge panel: Not public in the sources I retrieved; I will not guess the APJs.
• Petition grounds: Ground 1 — § 103 over Conroy (FOLFIRINOX), Conroy Protocol, Conroy Appendix, Mahaseth, in combination with Bayever, Saif, Ko, and Cantore, against claims 1, 3-6, and 8-14. Ground 2 — § 103 over the Ground 1 art plus Masi and Ginocchi, against claims 2, 7, and 15. (grounds summary, IPR2025-00505 docket record)
• Institution decision: Instituted for trial; oral argument reportedly scheduled for 2026-05-19. Ipsen had filed a Request for Discretionary Denial arguing § 325(d) (Conroy and Bayever were the same lead references overcome during prosecution) and Director's March 26, 2025 Memorandum considerations. (Ipsen request, PTAB E2E)
• Final Written Decision: None issued as of the public record I retrieved; trial-stage, not decided.
• Settlement / termination: None reported.
• Appeal: None reported.
• Defensive value for the '914: None directly. An adverse FWD on the '552 would not cancel any '914 claim. But it would (a) publicize the strongest obviousness theory against the NALIRIFOX/irinotecan-liposome family, and (b) give you a tested prior-art record to copy. Estoppel under § 315(e)(2) runs "with respect to that patent," so it would not estop CSPC/Conjupro from asserting the same art against the '914 in court.

IPR2025-01531 — Apotex Inc. v. Ipsen Biopharm Ltd.

• Type: Inter Partes Review
• Filed: 2025-09-12
• Status: Joined to IPR2025-00505 and instituted for trial (per public reporting)
• Patent challenged: US 11,344,552 — not the '914
• Judge panel / FWD / appeal: None public in my sources; joined proceedings typically adopt the host proceeding's schedule.

---

Strategic summary

Claim status on the '914: everything is UNTESTED. No independent claim, no dependent claim, and no claim term of US 10,993,914 has been construed or adjudicated by the Board. I am deliberately not listing "surviving claims," because a claim can only "survive" if it was challenged; reporting any claim list as narrowed would be false. The patent issued 2021-05-04 with its original claim set intact and stands unreviewed. For reference on family scope, the '914 is an ancestor of the later-issued US 12,059,497 and US 12,364,691 (both dated off the '914's 2015-10-16 priority chain and the 2021-03-22 continuation filing), and it is Orange-Book listed against Onivyde (NDA 207793) with an anticipated expiration of 2036-10-15.

Estoppel landscape — the field is wide open for the '914. Section 315(e)(2) estoppel is patent-specific, and because no IPR/PGR was ever instituted on the '914, no petitioner is estopped as to any '914 ground. A defendant today can raise any § 102 or § 103 ground supported by patents and printed publications, subject only to the ordinary constraints: § 315(b)'s one-year bar runs from service of a complaint alleging infringement of that patent — and, critically here, Conjupro's counterclaims state that Ipsen never asserted the '914 against it, so a § 315(b) clock on the '914 may not have started at all for that defendant (and a declaratory-judgment counterclaim does not itself trigger § 315(b)). Note also that the PGR window for the '914 closed nine months after issuance (≈2022-02-04), so § 112 and § 101 challenges are no longer available at the Board; any AIA attack now is IPR-only, limited to §§ 102/103 on patents and printed publications. And be realistic about discretionary denial: the Board's current practice (bifurcated institution procedure; "settled expectations" considerations raised against older patents; parallel-litigation stipulation requirements) makes institution materially harder than it was in the 2015-2024 era, and the '914 is now a 2021 patent in active district-court litigation.

Pattern signals. No serial-petitioner pattern on the '914 exists because no petition exists. The same petitioner group (CSPC/CSPC Ouyi/Conjupro) did attack the sibling '552, and Apotex joined that fight — indicating generic challengers are concentrating on the method-of-treatment patents rather than the liposomal composition patents like the '914. Ipsen is litigating aggressively on the district-court side (three D.N.J. cases: 1:24-cv-04991, 1:24-cv-08723 over the '497, and 1:2025-cv-13647 over the '691), and it fought the '552 petition on § 325(d) grounds. There is no defensive aggregator (e.g., Unified Patents) in the chain on the '914. The absence of any IPR against a patent that has been Orange-Book listed since 2015 and asserted in Paragraph IV litigation is a genuine signal: challengers appear to have judged the composition claims harder to invalidate on printed publications than the method claims.

Recommended next steps

• Do not tell a client this patent is "hardened." Say plainly: no PTAB proceeding has ever been filed against US 10,993,914; there is no IPR2025/IPR2026-style safety net here, and no claim has been canceled. Any representation to the contrary would be wrong.
• If you are a defendant, plan for an IPR you would have to file yourself. Verify the § 315(b) posture for your entity: the bar attaches only on service of a complaint alleging infringement of the '914 specifically. If you have not been sued on the '914 (as was the case for Conjupro, per its own counterclaims), no bar has yet run — but if you are a privy of a party that was served on the '914, the clock is running.
• Mine the '552 IPR record for free art. The Conroy / Bayever / Saif / Ko / Cantore combination and the expert declarations from IPR2025-00505 are public at PTAB E2E — but remember the '914 is a composition/stability patent (lyso-PC formation, irinotecan:SOS gram-equivalent ratio, pH > 6.5), so method-of-treatment art aimed at the '552 will not map onto the '914 claims. Expect to need art on liposomal irinotecan formulation stability, DSPC hydrolysis, and loading gradients. The '914's own prosecution history and the specification's own admissions about pH 6.0-6.5 minimizing PC hydrolysis may be productive starting points.
• Track the '552 trial as a bellwether. If IPR2025-00505 (instituted 2025-08-14, oral argument reported for 2026-05-19) produces an FWD invalidating '552 claims, it will not touch the '914, but it tells you whether the Board is receptive to Onivyde-family obviousness theories. FWD is due within 12 months of institution under § 316(a)(11).
• Check the parallel litigation docket for the '914's actual status. Because the '914 was named in Conjupro's Paragraph IV notice and DJ counterclaims but (per those counterclaims) not in Ipsen's infringement counts, confirm from the D.N.J. docket whether any amended complaint has since added the '914 — that would start a fresh § 315(b) clock and change the analysis.

Sources: US10993914B2 on Google Patents · PTAB E2E PTACTS public documents · IPR2025-00505 grounds summary · Ipsen v. Conjupro docket summary (paragraphfour.com) · D.N.J. 1:24-cv-08723 complaint analysis · DrugPatentWatch entry for 10,993,914

Confidence note: I have high confidence there are no AIA trial proceedings on US 10,993,914. I have lower confidence in the exact institution date (2025-08-14) and oral-argument date (2026-05-19) for IPR2025-00505, which come from secondary reporting rather than the Board's own papers, and I could not retrieve the APJ panels or a Final Written Decision for that proceeding — treat those as unverified.

I'll research the assignment record for US 10993914, starting with the assignment chain and corporate context.

Let me dig into the assignment records specifics and litigation/LinkedIn-level detail.

I hit the tool-call limit before I could pull the raw reel/frame entries from USPTO Assignment Center. Below I flag exactly what is documented versus what I could not retrieve, and I do not invent reel/frame numbers.

Ownership reconstruction — US 10,993,914 B2

Retrieval note (read first): I could not open USPTO Assignment Center / the assignment search API in this session, so no reel/frame numbers, execution dates, or correspondent-of-record entries were directly retrieved. Everything below on the assignment chain comes from the Google Patents "Legal Events" record for US 10,993,914 (which mirrors the same assignment feeds) plus SEC/press filings for corporate context. Where a field is unknown I say so rather than fill it in.

Inventors

Named on the face of the patent (8): Daryl C. Drummond, Dmitri B. Kirpotin, Mark Eamon Hayes, Charles Noble, Kevin Kesper, Antoine M. Awad, Douglas J. Moore, Andrew J. O'Brien.

• Employer at filing: Merrimack Pharmaceuticals, Inc. (Cambridge, MA). The 2015-10-16 priority filing was made by Merrimack, and the 2021-03-22 Legal Event "Assigned to MERRIMACK PHARMACEUTICALS, INC." lists all eight inventors as assignors — direct evidence that each was under an obligation to assign to Merrimack. Drummond and Kirpotin are long-standing nano-liposomal scientists of the Hermes Biosciences → Merrimack lineage (background, not an assignment-record fact).
• Unusual pattern — yes, but explainable: every inventor's obligation ran to an entity that had already ceased to exist as an operating company. Merrimack transferred 100+ employees to Ipsen at the April 2017 closing, and Merrimack itself terminated its entire workforce by 2019-07-12 ("The Company does not have any employees," Merrimack 10-K/10-Q; confirmed in Merrimack's 2020 10-K). So the inventor group did not "depart ahead of a fire sale" in the troll-pattern sense — they were moved to the acquirer as part of an asset purchase, and the residue of Merrimack became a milestone-collection shell.

Original assignee

• On the face of the patent: Ipsen Biopharm Ltd (this is the assignee-of-record that Google Patents reports as both "original" and "current" assignee).
• Applicant at the time of the priority filing (2015-10-16): Merrimack Pharmaceuticals, Inc.
• Primary line of business: Ipsen Biopharm Ltd is the UK specialty-pharma subsidiary of Ipsen S.A. (Euronext Paris: IPN; OTC: IPSEY), a commercial-stage pharmaceutical group (oncology, neuroscience, endocrinology).
• Did it ship a product embodying the claims? Yes. The patent is listed as covering ONIVYDE® (irinotecan liposome injection) — Ipsen's approved product (US FDA approval Oct 2015; patents listed against it include US 10,993,914, expiry Oct 15, 2036, per GreyB/Pharsight drug–patent listing). Ipsen markets ONIVYDE in the US and holds the ex-US rights formerly licensed to Shire (now Servier).
• Merrimack's status: still a registered public company (Nasdaq: MACK) but non-operating — no employees since 2019-07-12, purely a contingent-milestone vehicle; it received the $225M first-line milestone on 2024-03-27 and resolved to dissolve and distribute it. It never filed bankruptcy (a bankruptcy filing was a stated contingency in the 2017 proxy only).

Assignment timeline

Documented links (from Google Patents Legal Events; reel/frame not retrieved, so omitted rather than guessed):

• Executed date not shown / recorded 2021-03-22 — Reel not retrieved

  • Conveyance: Assignment (inventor → company)
  • Assignor: Drummond, Hayes, Kesper, Kirpotin, Moore, Noble, O'Brien, Awad (8 inventors)
  • Assignee: MERRIMACK PHARMACEUTICALS, INC.
  • Correspondent: not retrieved — no basis to flag a repeat correspondent.
  • Context: Confirmatory/clean-up inventor assignment recorded immediately before issuance (patent granted 2021-05-04).

• Executed date not shown / recorded 2021-03-22 — Reel not retrieved

  • Conveyance: Assignment
  • Assignor: MERRIMACK PHARMACEUTICALS, INC.
  • Assignee: IPSEN BIOPHARM LTD.
  • Correspondent: not retrieved.
  • Context: Corporate asset acquisition (the April 2017 Ipsen/Merrimack Purchase and Sale Agreement, $575M upfront + up to $450M milestones), recorded late — i.e., title clean-up pinning the continuation application to the Ipsen patent estate before grant.

No other recorded post-issuance assignments appear in the Legal Events feed. Google Patents also shows assignments to Merrimack/Ipsen in 2021-03-22 only; the later 2024-06-07 and 2021-03-22 "priority" entries are family/priority links (e.g., US 17/208,042 → US 12,059,497; US 18/737,163), not ownership changes. Current owner: Ipsen Biopharm Ltd.

Caveat: Dechert LLP is identified in Ipsen's 2017 press materials as Ipsen's M&A counsel on the Merrimack transaction — that is deal counsel, not the assignment correspondent of record, so it is not a finding for Signal 3.

Timeline diagram

timeline
    title Ownership of US 10993914
    2015 : Priority filing by Merrimack
    2017 : Ipsen buys Onivyde assets from Merrimack
    2019 : Continuation filed
         : Merrimack ends all employment
    2021 : Inventors assign to Merrimack
         : Merrimack assigns to Ipsen Biopharm
         : Patent issues to Ipsen
    2024 : Ipsen pays 225M milestone to Merrimack

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The chain runs inventors → Merrimack → Ipsen Biopharm Ltd, a named operating subsidiary of a listed pharma group. No "IP/Holdings/Ventures" entity, no registered-agent service address, no single-member LLC appears anywhere in the recorded events (both 2021-03-22 entries). The one shell-like entity in the story is the seller (Merrimack post-wind-down), not an asserter.
2. Known asserter in the chain — not present. Neither Merrimack nor Ipsen appears on the Acacia / Marathon / IV / IPNav / Wi-LAN / Conversant / Vringo / Pendrell / Round Rock / Spangenberg-type lists. Ipsen is a commercial manufacturer enforcing an FDA-approved product, the opposite profile.
3. Repeat correspondent across the chain — unclear / unverifiable. The correspondent-of-record fields were not retrievable in this session. Both 2021-03-22 entries would need to be inspected for a common attorney; without reel/frame I cannot state recurrence, and a single appearance would not be a finding anyway.
4. Cascading transfers — not present. Two recorded conveyances, both on 2021-03-22, transferring a single application up a two-step chain (inventors → Merrimack → Ipsen). No string of unrelated assignees, no <24-month LLC-to-LLC flipping.
5. Pre-litigation transfer — not present as a setup signal. The transfer to Ipsen reflects a 2017 corporate asset purchase, recorded in 2021 only for title clean-up; it predates any assertion and was not arranged within 6 months of a suit. (Google Patents flags "Family has litigation" for family 57219018 via its Darts-ip link, i.e., there is family litigation — but I could not retrieve the case caption/date, so I cannot tie it to a transfer window.)
6. Bankruptcy fire-sale — not present. Merrimack never filed Chapter 7/11 (a filing was only a disclosed contingency in the 2017 proxy; see SEC PREM14A language, "we may have to consider other strategic alternatives, including filing for bankruptcy"). The 2017 Ipsen deal was a solvent, stockholder-approved asset sale, and later disposals (MM-121/MM-111 → 14ner Oncology, 2019; preclinical nano-liposome assets → Celator, 2020-03-27) were out-of-court wind-down sales, not bankruptcy-auction sales. The Kodak/Nortel/Polaroid fact pattern is absent.
7. Privateering — not present. Ipsen did not acquire the patent to assert on a third party's behalf; it acquired the commercial business — product, manufacturing, supply chain, and 100+ employees — and sells ONIVYDE itself.
8. Defensive aggregator — not present. The chain terminates at an operating pharma manufacturer, not at RPX / AST / LOT / Unified Patents / OIN.

Verdict

Operating-company assertion.

The recorded chain is a clean two-link corporate transfer — eight inventors to Merrimack Pharmaceuticals, Inc. and then Merrimack to Ipsen Biopharm Ltd, both recorded 2021-03-22 (one day of clean-up immediately before the 2021-05-04 grant) — landing title in the manufacturer of ONIVYDE®, the irinotecan liposome product whose claims this patent covers. None of the eight NPE signals has evidentiary support: no shell assignee, no listed asserter, no cascading LLC chain, no bankruptcy sale, and no defensive-aggregator endpoint. The only "shell" in the record is the seller (Merrimack) after it discharged all employees on 2019-07-12 — not the party asserting the patent.

Verification links: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search "10993914"); Google Patents Legal Events — https://patents.google.com/patent/US10993914/en (assignments tab). The reel/frame and correspondent fields must be read off those pages; I could not retrieve them here, and I have not guessed them.

I'll search for the patent and its citation records.

Let me pull the citation records directly.

Prior-Art Analysis — US Patent 10,993,914 B2

Patent under analysis: US 10,993,914 B2, "Stabilizing camptothecin pharmaceutical compositions" (app. 16/567,902; filed 2019‑09‑11; priority 2015‑10‑16; granted 2021‑05‑04; Ipsen Biopharm Ltd.).

---

0. Scoping note — what I could and could not retrieve (read this first)

Strict-identifier compliance: every search was run on the literal string 10993914 / US10993914 / US10993914B2. I did not substitute near-neighbours (e.g., 10,993,941; 10,993,914-as-Siemens-part-number; 10993914 as an Elsevier DOI fragment, which appeared as a spurious hit and was discarded).

Critical gap. The authoritative Google Patents full text supplied in this task does not contain the "Citations" / "References Cited" block, nor the claims block — the fetched body begins at the Classification/Definitions sections and the machine extraction stripped both. My searches for the face-of-patent "U.S. Patent Documents" list (Justia, FPO, patentimages OCR of the front page) returned only partial non-patent-literature excerpts and similar-document tables from other patents, not a verified front-page citation list for '914.

Consequently, I separate what follows into three evidentiary tiers and label each reference with a confidence marker:

Any statement below that a given document is an "examiner-cited reference" would be an inference, not a verified fact, and I have not made it.

---

1. Tier A — Patent documents cited in the '914 specification (incorporated by reference)

These are the only patent citations I can verify with certainty, because the specification text names them. All are incorporated by reference "in its entirety."

§ 102 analysis — Tier A

---

2. Tier B — Non-patent literature on the '914 reference list (as reported by Justia's "References Cited")

The Justia record for 10,993,914 exposes a long NPL list. The items below are those I could read directly from retrieved excerpts. Note the date problem: many are post-priority (2016–2017) and were submitted as IDS/acknowledged art, not applied as prior art.

---

3. Tier C — Family / sibling documents (NOT prior art — do not cite as §102 art)

These surfaced repeatedly in searches and are sometimes mistaken for prior art:

---

4. Synthesis — which claim(s), if any, are at §102 risk

I will be blunt about the limits of the record I obtained.

1. No verified anticipation across the references I could retrieve. Every Tier A/B reference is missing at least one element of every independent claim as characterized in the authoritative text (Embodiments 1–22) and in the earlier claim-1 reading: (i) the DSPC/cholesterol/MPEG‑2000‑DSPE 3:2:0.015 liposome; (ii) irinotecan sucrosofate encapsulated therein; (iii) 500 g ±10 % irinotecan moiety per mmol PL (or ~471 g, or ~433 g with ≤100 ppm TEA per Embodiments 20–21); (iv) pH 7.00–7.50 after manufacture; and (v) the functional limitation of <20 mol % lyso‑PC relative to total liposome phospholipid at 6 months (Embodiments 1–4, 9–10) or ≤1 mol % pre-storage (Embodiment 9).

2. The functional lyso‑PC limitation is the real §102 shield. Under In re Schreiber-style analysis, a product-by-process/functional claim can only be anticipated by a reference disclosing the product itself with all its properties; the SDNY/D.N.J. practice — and the specification's own comparator data (Sample 12, SR ≈ 724, pH 6.5 → >30 mol % lyso‑PC at 3 months; >35 mol % at 9 months) — establishes that the prior commercial process produced a different product. A reference practising Sample‑12-type manufacture is therefore outside the claims, not anticipatory of them.

3. The 2015‑10‑16 priority date is doing heavy lifting. The ONIVYDE FDA approval/label (B7, Oct. 22, 2015) and the ASCO 2015 nal‑IRI presentations (B4) sit within days of the priority date. If any claim of '914 is found to lack written-description support for the alkaline‑pH or ER/lyso‑PC features as of 2015‑10‑16, the effective date slips and the ONIVYDE 2015 label itself becomes the most dangerous single reference — because it discloses the commercial irinotecan‑SOS liposome (6.81 / 2.22 / 0.12 mg/mL lipid levels, 4.3 mg/mL irinotecan) that the '914 dependent-claim/embodiment set essentially re-describes. This is the vulnerability to test, and I cannot resolve it on the citations record available.

4. Where the true §102/§103 fight will be: the D.N.J. action Ipsen Biopharmaceuticals Inc. v. Conjupro Biotherapeutics Inc., No. 1:24‑cv‑04991 (consolidated with 1:24‑cv‑08723 and 1:25‑cv‑13647), in which '914 is named in Conjupro's Paragraph IV notice and pleaded as an Initial Patent‑In‑Suit (listed expiry 10/15/2036). The prior-art grounds actually litigated there — not the examiner-cited list — will define the operative §102/§103 record for this patent. I did not retrieve the invalidity contentions and cannot state what art the defendants assert.

---

5. Contradictions / corrections against the previously generated sections

• Flagged — claim text source. The earlier summary states it has "verbatim claim text only for independent claim 1 and dependent claims 2–11" drawn from drugpatentwatch. The authoritative full text supplied here contains no claims section at all (Google Patents extraction omitted it), and the numbered items it does contain are Embodiments 1–22, which are not claims. I am therefore treating the earlier claim-1/claims-2–11 wording as unverified against the authoritative source, and I have mapped my §102 analysis to the Embodiments (which I can quote) rather than to claim numbers (which I cannot).
• New datum not in the earlier summary. The authoritative text (Embodiment 20/Tier-A excerpt) confirms a claim/embodiment limitation on residual triethylammonium: "not more than about 100 ppm of triethylammonium encapsulated within the phospholipid liposome," and Embodiment 21 recites 30–100 ppm TEA or DEA. A supplementary source rendered this as "about 433 g irinotecan moiety/mol phospholipid and not more than about 100 ppm triethylammonium." No Tier A/B reference retrieved discloses a TEA-content limitation, so this narrows rather than broadens the §102 exposure.
• No contradiction found on bibliographic facts (filing 2019‑09‑11; priority 2015‑10‑16; grant 2021‑05‑04; assignee Ipsen Biopharm Ltd.; examiner Isaac Shomer).

---

6. Explicit uncertainties

1. The face-of-patent "References Cited" list for US 10,993,914 was not obtainable in this search session. The §102 mappings above cover the references I could verify (Tier A: spec-cited patents; Tier B: partial NPL list) plus clearly-labelled inferences (Tier C). Any reference in the examiner's actual citation list that I have not named is unanalysed.
2. Titles and exact issue dates for A1–A4 are not verified from a primary USPTO/Espacenet record in this session; only the numbers are verified, because the specification names them.
3. No PCT/WO publication number for this family was confirmed. EP 3 337 478 B1 appeared in search-adjacent family tables that also displayed US10993914B2 and US12059497B2; the mapping of EP 3 337 478 to this family is plausible but unverified, so I have not used it as a citation.
4. I did not obtain the full 10,993,914 claim set, so I cannot confirm the number or wording of independent process claims; my §102 conclusions are keyed to Embodiments 1–22 and to the earlier (unverified) claim-1 summary.
5. The Justia NPL list is a truncated third-party rendering; several entries (e.g., Han et al. ACTR‑33, the Düzgüneş reference appearing as a page fragment) were not retrievable in full, and their dates are unconfirmed.

Bottom line: On the verifiable record, no cited reference anticipates any claim of US 10,993,914, because every candidate lacks the alkaline-pH processing limitation and/or the lyso‑PC functional limitation; the specification's own Sample‑12 comparator (pH 6.5, SR ≈ 724) affirmatively shows the prior-process product falls outside the claims. The genuine §102 exposure is not in the citation list at all — it is the Oct. 22, 2015 ONIVYDE label sitting six days after the 2015‑10‑16 priority date, and it only bites if the priority claim fails.

I'll build the § 103 analysis on the prior-art record already established, but first I'll run targeted searches to ground the motivation-to-combine reasoning in verifiable references (especially the liposome-stability and SOS-loading art).

Let me ground two more points: the actual claim set of '914, and the specific liposome-stability / loading art that a challenger would use.

Obviousness Analysis — US 10,993,914 B2 under 35 U.S.C. § 103

Patent: US 10,993,914 B2, "Stabilizing camptothecin pharmaceutical compositions" (app. 16/567,902; filed 2019‑09‑11; priority 2015‑10‑16; granted 2021‑05‑04; Ipsen Biopharm Ltd.). Post‑AIA patent → AIA § 102/103 governs; critical date = 2015‑10‑16.

Instruction compliance: every identifier below is used literally. Where a search result contradicts the earlier sections, I flag it in § 9. I build on the Prior Art section's Tiers A/B rather than repeating them.

Important upgrade from the prior-art section: the Prior-Art section could not verify the face-of-patent reference list and treated US 8,147,867 as a low-confidence, spec-cited item (its Tier‑A3). Live retrieval now confirms US 8,147,867 B2 as a fully described, dated, § 102(a)(1) reference that is the ONIVYDE‑enabling patent in the same corporate chain (Hermes Biosciences → Ipsen Biopharm; inventors Hong, Drummond, Kirpotin). That materially changes the § 103 picture: the primary reference is no longer speculative.

---

1. Legal framework and claim-construction predicates

The obviousness case must be built against the granted claim 1, which (per the verified claim text) is:

A storage stabilized liposomal irinotecan composition comprising irinotecan sucrose octasulfate (SOS), cholesterol, and one or more phospholipids with a ratio corresponding to a total of 500 grams ±10% by weight irinotecan moiety per mol total phospholipids, the liposomal irinotecan composition stabilized to have less than 20 mol %, with respect to total phospholipids, of lyso‑phosphatidylcholine (lyso‑PC) after 180 days or 6 months of storage … at 2 to 8 °C, the liposomal irinotecan composition obtained by a process comprising the steps of:
(a) forming liposomes comprising DSPC, cholesterol, and methoxy‑terminated polyethylene glycol‑distearoylphosphatidyl ethanolamine, encapsulating a substituted ammonium salt of SOS with a sulfate concentration of from 0.4 to 0.5 M;
(b) incubating the liposomes with an aqueous solution comprising irinotecan, at conditions effective to load the irinotecan, thereby forming a preparation of liposomes encapsulating irinotecan sucrose octasulfate within the liposomes; and
(c) obtaining the storage stabilized liposomal irinotecan composition by at least one additional step comprising adjusting the pH of the preparation of liposomes to a pH of from about 7.25 to about 7.50.
— Source: https://www.drugpatentwatch.com/p/patent-claims/[10993914](/patent/10993914)

Three doctrinal predicates control:

(i) Product-by-process treatment. Claim 1 is a composition "obtained by" a three-step process. Under In re Thorpe and Abbott Labs v. Sandoz, the process limitations do not by themselves impart patentability if the product itself would have been obvious; but the claimed product characteristics — including the ≤20 mol % lyso‑PC limitation — are structural/chemical properties of the composition and must be weighed. In re Schreiber/In re Nordt cut against the patentee only where a reference discloses the same product with all its properties. Here, as shown below, the art's own product (Sample 12-type, pH 6.5) does not exhibit the claimed property; that is the patentee's best shield and the challenger's central problem.

(ii) Result-effective variable / optimization. KSR Int'l v. Teleflex, 550 U.S. 398 (2007), and In re Applied Materials, In re Boesch, In re Aller make "optimization of a result-effective variable" routine. pH, drug:lipid ratio, and the identity/concentration of the trapping‑agent counterion are all result‑effective variables in the liposome-stability arts. A challenger therefore does not need a reference that names pH 7.25; it needs (a) recognition of the problem, and (b) a reasonable expectation that tuning the variable would produce the recited result. The patentee's counter is In re O'Farrell / In re Gurley teaching away.

(iii) The functional limitation is the battleground. The ≤20 mol % lyso‑PC at 180 days/6 months at 2–8 °C is a numerical stability limit. A challenger will call it a "desired result"; the patentee will call it a "criticality"/unexpected‑result boundary (the 942/992 “Stability Ratio” cliff; the pH 6.5→7.25 jump). See § 6.

---

2. Element-by-element map of the prior-art record

Cross-cutting "problem-recognition" references. Sætern, A. et al., "Camptothecin‑Catalyzed Phospholipid Hydrolysis in Liposomes," Int. J. Pharm. 288(1):73–80 (2005) (appearing on the sibling US 10,456,360 IDS; also a companion Int. J. Pharm. paper, Camptothecin-catalyzed phospholipid hydrolysis in liposomes) teaches that camptothecin drugs themselves catalyze PC hydrolysis — i.e., it frames the exact problem the '914 solves. Ochi 1980 (Chem. Pharm. Bull. 28:638–641) and US 4,990,610/US 5,783,568 supply the SOS chemistry and the polyanionized‑sugar genus. Mui 1993 (Biophys. J. 65:443–453) supplies the trapped‑volume basis (≈1.7 L/mol PL) used to compute the ER/SR numbers.

---

3. Combination 1 — the primary § 103 case (pH optimization)

References: US 8,147,867 B2 (primary) + Grit 1993 (both papers) + WO 94/22430 + Ickenstein 2006 + Sætern 2005 (secondary).

The case as a challenger would frame it:

1. US 8,147,867 gives the complete, enabled liposomal irinotecan‑SOS platform, with substituted‑ammonium (TEA/DEA) counterion, DSPC/cholesterol/PEG‑DSPE lipids, gradient drug loading, ~100–110 nm, 4.3 mg/mL, and refrigerated storage for 6 months. It is the platform the '914's own specification identifies as "generally known in the art … See, e.g., U.S. Pat. No. 8,147,867" — an applicant admission that this is background art.
2. Grit 1993 establishes the mechanism and, critically, that pH is a known, controlling, result‑effective variable for PC hydrolysis: "hydrolysis rates depend on temperature and pH," with lyso‑PC (1‑acyl‑LPC) as the first hydrolysis product.
3. WO 94/22430 expressly frames "a stability‑enhancing pH … at which lipid degradation, including lipid hydrolysis … is inhibited" and gives a working range of 5.5 to 7.5.
4. Sætern 2005 identifies that the encapsulated camptothecin drives PC hydrolysis, giving the POSA both the problem and the incentive to tune formulation pH/ionic environment.
5. Because the POSA is optimizing a known variable over a disclosed range to a known end (minimizing PC hydrolysis), KSR/In re Aller make the resulting pH value presumptively obvious, and the ≤20 mol % lyso‑PC limit follows as the expected consequence of the optimization.

Why this combination is nevertheless weak — and the patentee's answer:

• Teaching away. Grit 1993 (both papers), Ickenstein 2006, and WO 94/22430 all identify ≈ pH 6.5 as the minimum of the hydrolysis‑rate curve, and WO 94/22430 names 6.5 as the most desirable value. The '914 specification itself concedes the art believed "a pH of 6.0 or 6.5 … minimize[s] hydrolysis of phosphatidylcholine" and that "conditions where the pH is above 6.5 tend to increase … the amount of lyso‑PC." The claimed pH window (7.25–7.50) is therefore on the disfavored side of the prior-art optimum. On the specification's own characterization, the observed result "was unexpected in view of prior reports showing that a pH of 6.5 as optimal for minimizing phosphatidylcholine hydrolysis (Grit, M et al …)." That is a textbook In re Gurley / In re O'Farrell teaching‑away posture.
• The art's own product fails the claim. Comparative Sample 12 (pH 6.5, SR ≈ 724, the US 8,147,867/ONIVYDE‑type process) generated >30 mol % lyso‑PC at 3 months and >35 mol % at 9 months — i.e., the closest prior process lands outside the claim. Anticipation is therefore unavailable, and the § 103 "expected result" premise collapses: the POSA following the art's stated preference (6.5) would not have arrived at the claimed composition.
• Claimed narrowness. The granted claim recites 7.25–7.50, which does not even read on the specification's broader "greater than 6.5 / 7.0–7.5" disclosure. A challenger must therefore prove the specific sub‑range obvious, not merely the genus.

Net: Combination 1 is a legally cognizable but likely‑to‑fail § 103 theory on the pH element standing alone. Its strength rises sharply if the challenger can pair it with the synergistic-ratio theory (Combination 2) so that the motivation to move off 6.5 comes from the drug:lipid/trapping‑agent ratio rather than from hydrolysis chemistry.

---

4. Combination 2 — the "loading‑ratio / Stability‑Ratio" case (the stronger theory)

References: US 8,147,867 B2 (primary) + Haran/Barenholz 1993 + Fenske & Cullis 2008 review + US 2005/0106231 A1 + Mui 1993.

Motivation to combine (the challenger's best footing):

• US 8,147,867 itself teaches the empty‑liposome → gradient‑load architecture and treats the drug:lipid ratio and the trapping‑agent concentration as adjustable manufacturing variables.
• The remote‑loading literature makes the direction of optimization explicit: Fenske & Cullis 2008 states that for several drugs "drug retention could be significantly improved by increasing the drug‑to‑lipid ratio," and that sulfate counter‑anions (ammonium sulfate) produce intraliposomal precipitates that "stabilize … accumulation for prolonged storage periods (>6 months)" (Haran/Barenholz 1993). A POSA optimizing the irinotecan‑SOS system therefore had an articulated reason to (i) raise the drug:lipid ratio toward the claimed 500 g ±10 %/mol PL and (ii) choose the lower end of the sulfate concentration (0.4–0.5 M vs. the 0.60–0.65 M used in the comparative process).
• Mui 1993 supplies the trapped‑volume relation (≈1.7 L/mol PL) that makes the resulting irinotecan:sucrosofate (ER) ratio calculable in advance — i.e., the "ER > 0.9" and "SR > 950" constructs are predictable from known quantities, which supports an "obvious to optimize" argument. Hmm — note the caveat: ER is not recited in the granted claim 1 (it appears only in the application's Embodiment 2 and in the sibling US 10,456,360), so this theory maps to claim 1 only through the sulfate‑concentration and 500 g/mol‑PL integers.

Why this theory is stronger than Combination 1:

1. It supplies a motivation independent of the pH question — improving drug retention and reducing residual counterion — and the POSA would then discover that the higher drug:lipid ratio and lower sulfate load also suppress lyso‑PC. That is the classic KSR "finite number of identified, predictable solutions" argument.
2. It explains why the POSA would move to 0.45 M (mid‑point of the claimed 0.4–0.5 M) without any teaching away: the art contains no statement that 0.6–0.65 M is preferred for stability.

Why it still may not carry the day: the linkage the '914 claims is not mere retention — it is a lipid‑hydrolysis effect. Nothing in the retention literature (US 8,147,867; Haran/Barenholz; Fenske) teaches or suggests that the sulfate‑group concentration per se governs PC hydrolysis into lyso‑PC; that is the '914's stated discovery ("the stability of the resulting drug‑loaded liposomes depends on initial concentration of sulfate groups … and the ratio of camptothecin encapsulated to phospholipid"). Absent that link, the "optimization" yields better retention, not necessarily ≤20 mol % lyso‑PC — and the specification's own data show that SR alone (e.g., SR 942 vs. 724 at pH 6.5: 35.7 vs. 35.4 mol %) does not deliver the claimed limit unless pH is also raised.

Practical conclusion: Combinations 1 and 2 are individually vulnerable but jointly dangerous, because the claim requires the combination of high drug:PL ratio, 0.4–0.5 M sulfate and pH 7.25–7.50. A petitioner who can show (a) motivation to raise the ratio (retention art), and (b) motivation to scan pH (WO 94/22430's 5.5–7.5 range), and then (c) argue that the interaction was a predictable "synergy" of two known variables, has the strongest available § 103 theory. The patentee rebuts (c) with the 942/992 discontinuity and the pH 6.5 anomaly (§ 6). Whether that rebuttal succeeds is a battle of experts on "predictability of the interaction."

---

5. Combination 3 — the ONIVYDE‑label theory (contingent on priority)

Reference: ONIVYDE (irinotecan liposome injection) FDA label, rev. Oct. 22, 2015 — 6 days after the 2015‑10‑16 priority date.

Significance: The label discloses the near-verbatim commercial composition the dependent claims recite: "unilamellar lipid bilayer vesicle, approximately 110 nm … irinotecan … gelated or precipitated state as the sucrose octasulfate salt … DSPC 6.81 mg/mL, cholesterol 2.22 mg/mL, and MPEG-2000‑DSPE 0.12 mg/mL … HEPES 4.05 mg/mL … sodium chloride 8.42 mg/mL" and "4.3 mg/mL irinotecan free base."

Why it is not currently available as prior art: it post‑dates the 2015‑10‑16 priority date by six days. It becomes § 102(a)(1) art only if claim 1 is denied the 2015‑10‑16 priority for lack of written‑description/enablement of the alkaline‑pH, high‑ratio, ≤20 mol %‑lyso‑PC subject matter. That is a live but untested vulnerability that the prior‑art section already flagged, and I confirm it here: the '914's own Example 7 reproduces the ONIVYDE label almost verbatim, which a challenger would use to argue that the specification's "invention" is coextensive with the already‑marketed product. If priority slips, ONIVYDE + Grit (pH variable) + Haran/Barenholz (SOS/sulfate salt, 6‑month stability) is a strong combination (though still not an anticipation, because the label is silent on pH and on lyso‑PC).

Bottom line: Combination 3 is the highest-value theory for a challenger but is priority‑dependent and cannot be resolved on the citations record alone.

---

6. Secondary considerations and the patentee's rebuttal

1. Teaching away (strong). All the lipid‑hydrolysis art (Grit, Ickenstein, WO 94/22430) points to pH ≈ 6.5; the invention moves the other way. This is the single most powerful non‑obviousness fact in the record, and it is admitted on the face of the specification.
2. Unexpected results / "criticality" (moderate‑to‑strong). The specification reports a non‑linear response: at pH 6.5, moving SR from 942 to 992 (a 5 % change) cut lyso‑PC ~34 %, whereas moving SR 724→942 (a 30 % change) changed it by only ~1 %. "Criticality" arguments have traction when the prior art discloses the broad range but not the point of inflection (In re Wood's Patent line). A court will ask whether the 942/992 threshold reflects a true physicochemical discontinuity or experimental scatter.
3. The art's own product fails the claim (strong). Sample 12 demonstrates that the closest prior process generates 30–38 mol % lyso‑PC, well outside the claim — evidence both against anticipation and against a "reasonably expected" result.
4. Weak/against‑the‑patentee facts. (a) The patentee's own theory — residual protonated TEA amine drives hydrolysis — makes the problem and the fix (exhaust the amine; raise external pH) mechanistically understandable (KSR "design incentives"); (b) Example 8 shows the pH effect is absent for topotecan (both pH 6.5 and 7.25 gave ~30 mol % at 9 months), undercutting any claim to a general principle and making the irinotecan result look like a system‑specific optimization; (c) the "Stability Ratio" is simply a ratio of two known process quantities, itself an invitation to optimize.

---

7. Dependent claims

Because the granted '914 claim set does not recite the ER ratio (Embodiment 2), the residual‑TEA ppm limits (Embodiments 20–21/TEA), or the "Lyso‑PC Stability Ratio ≥ 990" (Embodiment/Spec ¶[0121]), those narrowing features — which are present in the sibling US 10,456,360 — are not available to the '914 patentee as claim limitations and should not be weighed in this § 103 analysis. Flagged in § 9.

---

8. Bottom line

• No single reference anticipates claim 1 (consistent with the prior-art section): US 8,147,867 lacks the pH 7.25–7.50 step and the ≤20 mol % lyso‑PC limit; Grit/Ickenstein/WO 94/22430 disclose pH control only as a genus pointing to 6.5; the ONIVYDE label post‑dates priority.
• The best § 103 theory is a two‑strand combination: (A) US 8,147,867 (complete liposomal irinotecan‑SOS platform) + (B) the lipid‑hydrolysis pH art (Grit 1993; Ickenstein 2006; WO 94/22430), optionally reinforced by (C) the retention-optimization art (Haran/Barenholz 1993; Fenske & Cullis 2008; Mui 1993) and the problem‑recognition reference Sætern 2005. The motivation is to reduce phospholipid hydrolysis (and thereby extend refrigerated shelf life) of the known liposomal irinotecan product; the expectation is that pH and drug:lipid/sulfate ratios are result‑effective variables that can be optimized by routine experimentation.
• The patentee's defenses are strong on the pH axis — the art teaches away (6.5 optimum) and the closest prior process (Sample 12) lands outside the claim — and moderate on the interaction axis (the 942/992 "criticality" and the unexpected direction of the pH effect).
• The single most dangerous attack is not a printed reference at all: it is a priority/written‑description attack that would convert the Oct. 22, 2015 ONIVYDE label (whose composition the dependent claims essentially recite) into prior art, at which point Combinations 3 becomes materially stronger.

Confidence: high on reference identities/dates (US 8,147,867 B2; Haran/Barenholz 1993; Grit 1993 ×2; Ickenstein 2006; WO 94/22430; Ochi 1980; Mui 1993; Sætern 2005). Medium on the scope of US 2005/0106231 A1 and US 2008/0113016 A1 (verified as existing and on‑point, not read in full). Lower on whether the 942/992 discontinuity and the pH‑effect direction would defeat a well‑built KSR case — that is a fact‑intensive question I cannot resolve from the citations record.

---

9. Contradictions / corrections against the previously generated sections

1. US 8,147,867 upgraded. The Prior Art section listed it as Tier‑A3 with "Low" confidence in its title and date. Live retrieval confirms: US 8,147,867 B2, "Liposomes useful for drug delivery," inventors Hong/Drummond/Kirpotin, filed 2005‑05‑02, issued 2012‑04‑03, priority 2004‑05‑03, assignee Ipsen Biopharm Ltd (originally Hermes Biosciences). It is therefore § 102(a)(1) prior art (patented > 1 year before the 2015 priority), and it is the primary reference, not a background cite. The prior-art section's "Closest of the four, but … no [anticipation]" conclusion on anticipation holds; its confidence rating should be raised.
2. Claim‑text correction. The prior-art section flagged that the authoritative full text contains Embodiments 1–23, not claims, and that the claim set was unverified. It is now verified for the granted '914: claim 1 requires pH 7.25–7.50 (not 7.00–7.50) and omits the ER‑ratio and TEA‑ppm limitations that appear in the application's Embodiment list and in the sibling US 10,456,360. The earlier summary's claim‑1 wording and the prior-art section's Embodiment mapping therefore describe different documents (granted '914 claims vs. the '360/application embodiments). Any § 103 argument relying on ER or TEA ppm must be redirected to US 10,456,360 — those features are not in the '914 claims.
3. Family identity confirmed. Search retrieval shows the '914 specification text is reproduced almost verbatim in US 2018/0110771 A1, US 10,456,360 B2, and WO 2017/066726 A1 (PCT/US2016/057247). These are the same family, not independent prior art — consistent with the prior-art section's Tier‑C treatment, now confirmed.
4. New references not in the prior-art section (added here as § 103 anchors, all pre‑2015): Haran, Cohen, Bar & Barenholz 1993, BBA 1151(2):201–215; Fenske & Cullis 2008 review (drugto‑lipid ratio and liposomal retention); WO 94/22430 (storage‑stable liposome; "stability‑enhancing pH"); US 2005/0106231 A1 (camptothecin remote loading by ammonium‑sulfate gradient); Sætern et al. 2005, Int. J. Pharm. 288(1):73–80 ("Camptothecin‑Catalyzed Phospholipid Hydrolysis in Liposomes"). None of these appeared in the earlier Tier A/B lists; each is flagged here with its source.
5. No contradiction found on the bibliographic spine (filing 2019‑09‑11; priority 2015‑10‑16; grant 2021‑05‑04; assignee Ipsen Biopharm Ltd.; D.N.J. 1:24‑cv‑04991 family litigation).

---

10. Explicit uncertainties

1. The face-of-patent "References Cited" list was still not retrievable; the additions in § 9 come from live retrieval of on‑point art and from the sibling patent's IDS, not from a verified front‑page list. An examiner‑cited reference I have not named remains unanalyzed.
2. US 2005/0106231 A1 and US 2008/0113016 A1 are verified as existing and topically relevant but were not read in full; their disclosure of a "500 g irinotecan/mol PL"‑type ratio is inferred, not confirmed.
3. Sætern 2005 appears to me through the US 10,456,360 IDS listing (title and citation), not through the article itself; its precise teaching (whether it suggests pH or counterion mitigation) is not verified.
4. I did not obtain the defendants' invalidity contentions in the D.N.J. action, so I cannot state which art they actually assert against the '914 — this analysis is a priori, not a report of litigation positions.
5. Whether an en banc/panel‑level priority challenge succeeds (Combination 3) is unresolved on this record and is the pivot on which the strongest § 103 theory turns.