US 6090925

I'll verify the patent's identity and check for any litigation/docket activity before summarizing.

Let me pull the claim text and check for any litigation activity specifically tied to this patent.

US 6,090,925 — verification and summary

Search caveat first: I searched for the literal string 6090925 / 6,090,925 and screened results for CAFC 2026 docket activity. Every relevant hit matched that exact number (no near-number contamination like 5,190,925 or 6,360,925 was counted). I did not get a direct hit on a USPTO Patent Center / Public PAIR page or a CAFC docket entry, so my bibliographic data comes from the Google Patents record and the patent's own front matter, corroborated by FreePatentsOnline.

Bibliographic data

Continuity (as recited in the patent itself): a continuation-in-part of Ser. No. 08/206,456, filed Mar. 4, 1994 (now U.S. Pat. No. 5,578,709), which is itself a CIP of Ser. No. 08/028,237, filed Mar. 9, 1993 (now abandoned).

Assignment chain (per Google Patents reassignment records):

• 1996-10-17 — assigned to Middlesex Sciences, Inc. (assignors: Blizzard, Sudhalter, Brown, Di, Scott, Woiszwillo)
• 1998-06-05 — Epic Therapeutics, Inc. (change of name from Middlesex Sciences)
• 1998-07-13 — Epic Therapeutics, Inc. (assignor: Frank J. Riske) — note: Riske appears in this assignment record but is not listed as an inventor on the '925 patent, unlike the sibling patent 5,981,719
• 2003-06-10 — Baxter International Inc. and Baxter Healthcare SA (from Epic Therapeutics, Inc.)

Representative classifications: G01N33/54346 (nanoparticles); A61K9/16, A61K9/1694; C07K1/30, C07K1/32.

Abstract (verbatim)

"Microparticles formed by mixing a macromolecule with a polymer at a pH near the isoelectric point of the macromolecule and incubating the mixture in the presence of an energy source for a predetermined length of time. The microparticles are composed of homogeneously distributed, intertwined macromolecule and polymer. Each microparticle allows aqueous fluids to enter and allows solubilized macromolecule and polymer to exit the microparticle and may be formulated to provide a sustained release of macromolecule and polymer from the interior of the microparticle when placed in an appropriate aqueous medium, such as under physiological conditions. Methods of production and methods of use for research, diagnostics and therapeutics are provided."

Disclosure in one paragraph

The patent teaches a non-emulsion, aqueous process: dissolve a macromolecule (protein, peptide, polysaccharide, nucleic acid, virus particle, or a drug–carrier conjugate) with a soluble linear/branched polymer (dextran, or a PVP/PEG mixture, poloxamer, PVA, Eudragit E100, etc.) at a pH near the macromolecule's isoelectric point (preferably within 3–4 pH units, most preferably 1.5–2), then expose the mixture to an energy source — preferably heat at 37–90 °C for minutes to hours — to drive dehydration/volume exclusion and form microparticles. The resulting particles are said to be substantially spherical, generally <10 µm, with a narrow size distribution, permeable to water, and composed of at least 40% and <100% macromolecule by weight (polymer ≤30%). Release kinetics are tunable via polymer identity/concentration, temperature, time, pH, and macromolecule concentration. Stated objects include avoiding water-in-oil emulsions and organic solvents, avoiding large-bore needles, avoiding the PLGA "burst" effect, and preserving macromolecule bioactivity. Uses span separations, diagnostics, immunoaffinity chromatography, vaccines, gene therapy, and sustained/targeted drug delivery.

Independent claims — I cannot give you authoritative text

Uncertainty flag, stated plainly: the full-text copy supplied to me was truncated inside the Detailed Description (it ends at "Prior to being incorporated in…") and the claims section was not included. My web searches likewise did not return a verbatim claim set (the USPTO full-text / Patent Center pages did not surface). I therefore will not paraphrase independent claims as if quoting them.

What I can say with grounding:

1. The claim set is large and has multiple independent claims. A JP search report in the family cited "US 6 090 925 A … column 13, lines 10–21; example 5" as relevant to claims 1–10, 14–25 (JP2012500118A), indicating at least 25 claims with the gap at 11–13 suggesting a separate independent claim or claims around there.
2. Claim-type inference from the Summary of Invention, which frames the disclosure as "Microparticles, methods of production, and methods of use thereof." That language maps onto the conventional three-way split you'd expect: (a) a process claim for making microparticles by mixing macromolecule + soluble polymer near the macromolecule's pI and applying an energy source; (b) a composition-of-matter claim for the microparticle itself, with the at-least-40%-macromolecule / ≤30%-polymer weight limitations; and (c) a method-of-use claim (e.g., administration for sustained/targeted delivery, or a diagnostic/separations use). This is inference from the specification, not a reading of the claim text — treat it as a hypothesis to be confirmed against the granted claims.

If you need the claims verbatim, the reliable sources are patents.google.com/patent/US6090925A/en, the USPTO PatentCenter full-text for application 08/699,586, or the issued patent PDF at patentimages.storage.googleapis.com/.../US6090925.pdf.

CAFC 2026 dockets — negative result

No CAFC docket activity involving US 6,090,925 was found. My CAFC-targeted search returned only unrelated 2026 appeals (e.g., FullView, Inc. v. Polycom, Inc., 2026-121, order dated April 16, 2026; AudienceScience v. Google; Global Tubing v. Tenaris; Magnolia Medical v. Kurin; Moskowitz Family v. Globus; Ironburg v. Valve). None referenced the '925 patent.

That negative result is consistent with the patent's status: the Google Patents record shows anticipated expiration March 9, 2013 and legal status "Expired – Lifetime." An expired patent cannot support a live infringement appeal, so 2026 CAFC activity on this number would be surprising. Caveat: absence from search results is not proof of absence from the docket, and I did not query the CAFC's PACER-adjacent docket system directly.

Related family members worth noting (same title/family, distinct numbers)

• US 5,981,719 — Epic Therapeutics, Inc.; filed Dec. 14, 1998 (Ser. No. 09/211,018); issued Nov. 9, 1999; inventors include Frank J. Riske in addition to the six named above. Same title.
• US 5,578,709 — the parent named in the '925 cross-reference.
• US 6,268,053 — cited alongside the '925 in family references.
• US 5,525,519 and WO 93/14110 (PCT/US 93-00073, filed Jan. 7, 1993, published Jul. 22, 1993) — cited in the '925 specification for polymer preparation methods.

Note on the record: the '925 patent's own text credits the polymer work to "U.S. Pat. No. 5,525,519 to James E. Woiszwillo" and to WO 93/14110 published "by James E." (name truncated in the OCR of the provided text).

I'll search litigation databases for this specific patent number.

Let me query litigation-specific databases directly.

Litigation search: US 6,090,925

Bottom line: I found no litigation — no district court action, no CAFC appeal, no ITC investigation, and no PTAB post-grant proceeding — in which US 6,090,925 is the patent-in-suit or an asserted/contested patent. I'm reporting that as a negative result with the caveats below, because a negative search result is not the same as proof of absence.

What I searched and what came back

No case was returned with a plaintiff, defendant, jurisdiction, case number, filing date, or outcome — because no such case appeared. I am not going to manufacture a table row for a case I could not find.

Contamination I screened out (so you know these are not your patent)

The search space throws off several decoys that a careless read would map onto "6090925":

• 1:26-cv-00925 — K.Mizra LLC v. Twingate Inc. (D. Del., filed 07/28/2026). This is a case number that happens to end in 0925; the asserted patent is US 8,234,705, not the '925 patent.
• 1:26-cv-00925 — Quartz Auto Technologies LLC v. Lyft, Inc. (W.D. Tex., filed ~04/13/2026). Again a case-number coincidence.
• JP 06 090925 A (Hitachi Ltd.) — a Japanese laid-open publication cited as prior art against an unrelated PCT application. Not the US patent.
• IPR2025-00847 / IPR2026-00130 re "the '609 patent" — this is US 9,318,609 B2 (Marlin Semiconductor v. TSMC / Marlin v. TSMC ITC Inv. No. 337-TA-1443). The "'609" shorthand refers to the last three digits, i.e. a different patent entirely. Do not conflate.
• Endo v. Baxter, N.D. Ill. Case No. 1:25-cv-02365 (filed 03/05/2025) — asserts a '280/'837 epinephrine patent against Baxter International. Baxter is the current assignee of the '925 patent, so this is a Baxter-as-defendant case, but the '925 patent is not asserted.

Why a negative result is plausible here

The patent's own record is consistent with zero litigation:

• Legal status: "Expired – Lifetime." Google Patents lists an anticipated expiration of March 9, 2013 (20 years from the March 9, 1993 priority date). A patent that expired in 2013 cannot be asserted in a 2026 infringement suit, and any appeal involving it would be long resolved.
• Ownership is consolidated with a large, non-practicing-defendant entity (Baxter Healthcare SA / Baxter International Inc., from Epic Therapeutics, Inc. in 2003-06-10), which is the profile of a patent used defensively or as manufacturing know-how rather than as a litigation asset.
• The technology was commercialized as ProMaxx microspheres (with LeuProMaxx as the lead program, per the Epic/Baxter profile surfaced in search) — a licensing/platform business, not a patent-troll posture.

Family notes that could carry litigation (unverified, flagged as leads only)

If you want to broaden the check, the same inventors/specification family may be more litigated than the '925 itself. I did not find litigation on these either, but they are the logical places to look next:

• US 5,981,719 (Epic Therapeutics; same title; adds Frank J. Riske)
• US 5,578,709 (parent, named in the '925 cross-reference)
• US 6,268,053
• EP 1060741 ("Microsphères à libération prolongée" / sustained-release microspheres; applicant Epic Therapeutics, Inc., proprietor Baxter International Inc.; French national part revoked/déchu définitivement 28/02/2017, i.e. a prosecution/maintenance lapse, not litigation)

Limitations you should weigh before relying on this

1. I could not query the primary sources directly. I did not get a live, structured result from https://portal.unifiedpatents.com/litigation/caselist, from PACER, or from the CAFC docket system — I searched the open web for those sources. A subscription litigation database (Docket Navigator, Darts-IP, Lex Machina) is the authoritative check.
2. Absence of search hits ≠ absence from the docket. Older, settled, or voluntarily-dismissed cases (2013 or earlier) are poorly indexed on the open web.
3. No litigated "patent-in-suit" hit does not mean the patent was never mentioned — it appears as a cited reference in later patent families and in specifications (e.g., EP 1418890 B1, RU 2457854 C2), which is citation activity, not litigation.
4. Date inconsistency to flag: the task header says "Current Date: April 26, 2026," but my operating context and the search index show dates out to September 26–27, 2026. I've treated the later date as operative; if the April date is the intended cutoff, some of the 2026-dated material above postdates it.

Recommended next step

To close this out definitively rather than by inference, I'd run the patent number through Docket Navigator or Unified Patents' litigation portal (which indexes by patent number) and pull the USPTO PatentCenter "Litigation Search" / assignment record for application 08/699,586. Those tools key on the patent number itself and will surface any 1993–2013 era case that the open web has not indexed. Based on everything available to me, I'd expect that search to confirm no litigation.

I'll research PTAB activity on this patent, starting with the structured data's default and then probing the web for anything the ODP hasn't indexed.

The initial sweep turned up nothing tied to the '925 number specifically — only same-owner (Baxter) IPRs on different patents. Let me probe deeper to be sure.

Both sweeps returned only noise for the '925 number (a gene-variant coordinate, an EIN, a Greek phone number) plus same-owner Baxter IPRs on different patents. Let me do a final confirmatory check on the sibling family and on Unified Patents-style filers.

Proceedings overview

Zero. The USPTO Open Data Portal reports no AIA trial proceedings on file for US 6,090,925 (no IPR, no PGR, no CBM) — and four independent web sweeps corroborate that negative: every hit on the literal strings 6090925 / 6,090,925 resolved to non-patent noise (a ClinVar genomic coordinate at 20p12.3, a Greek school-district phone number, an IRS EIN 39-6090925), to patent citations of the '925 as prior art, or to PTAB proceedings involving Baxter-owned but entirely different patents. Bottom line for a defendant: the patent has never been tested at the PTAB, so there is no claim-level PTAB outcome to lean on — the defense here is calendar-based, not estoppel-based. US 6,090,925 issued 2000-07-18 and reached its anticipated expiration on 2013-03-09 ("Expired – Lifetime"). A patent that expired over thirteen years ago cannot support a live infringement claim, and no AIA-trial remedy exists or is needed against it.

Proceeding-by-proceeding

There are no proceedings to enumerate. Per the constraint against fabricating proceeding numbers, I will not manufacture an "IPR20xx-xxxxx" heading for this patent.

Near-misses you will find in search results — none of these is US 6,090,925

A diligent search surfaces several PTAB matters that look related because they share a patent owner (Baxter), a petitioner (Ethicon), or a technology label. They are not this patent, and none of their outcomes transfer. I flag them only so a defendant does not mistake them for relevant precedent:

Source for the Ethicon cluster: PTO Litigation Center Report, 2015-03-27 — https://natlawreview.com/node/43923/printable/pdf

Structural reason the cupboard is bare

Three independent timing/eligibility facts explain the zero count; this is why the absence is credible rather than merely under-indexed:

1. PGR was never available. PGR became available 2012-09-16 and applies only to first-inventor-to-file patents. The '925 was granted 2000-07-18 — its nine-month PGR window closed in April 2001, more than a decade before the AIA created the vehicle.
2. CBM is inapplicable. CBM reaches patents covering a financial product or service. Protein/polysaccharide/nucleic-acid microparticles for drug delivery and diagnostics are outside that subject matter.
3. IPR's practical window was ~6 months against a dead patent. IPR became available 2012-09-16. The patent's "Expired – Lifetime" status capped its life at 2013-03-09. That left a six-month sliver, and no rational petitioner spends IPR money to invalidate claims that are about to expire — IPRs of expired patents are permissible but strategically pointless absent a live past-damages dispute.

Strategic summary

Claim status — you cannot get a canceled-claims answer, because no PTAB panel ever parsed these claims. No claim of US 6,090,925 is canceled by any AIA final written decision, and none is invalidated by the Board; conversely, none is sustained (the PTAB never affirmed patentability either). The correct label for every claim is untested at the PTAB. Notably, the full patent text supplied to me truncated inside the Detailed Description before the claims, and my searches did not surface the granted claim set verbatim — so I will not represent which specific claims exist or what they recite (see the prior section's uncertainty flag). What is documented is that an unrelated later Baxter application (PCT/US2009/054504, published as JP2012500118A) cites the '925 as prior art against claims 1–10 and 14–25, i.e., the patent's own disclosure has become a squeeze on later-filed family members — a fact of note, but not a PTAB result.

Estoppel landscape — there is none, and it does not matter. Because no IPR/PGR was ever instituted, § 315(e)(2) estoppel is not triggered against anyone: no petitioner, real party in interest, or privy is barred from raising any § 102 or § 103 ground. In the abstract, a challenger retains the entire universe of printed-publication and patent prior art — including, ironically, the '925's own parent US 5,578,709, its sibling US 5,981,719, the epoxy/PLGA art it distinguished, EP 0 809 110, and specifications like US 5,525,519 and WO 93/14110 that the '925 itself cites. That breadth is academic, however, because the patent's expiry forecloses assertion in the first place.

Pattern signals. There is no PTAB pattern on this number: no repeat petitioner, no serial petitioning, no appeal to the Federal Circuit (confirmed by the prior section's CAFC-negative search, consistent with the 2013 expiry), and no defensive aggregator such as Unified Patents anywhere in the chain. The Baxter/Ethicon 2015 IPR cluster shows only that Baxter was, in that era, a target of a serial petitioner on its polymer-composition portfolio — but that campaign expressly left this patent alone, presumably because it was already expired.

Recommended next steps

• Do not plan an IPR. The patent is expired with a legal status of "Expired – Lifetime" and an anticipated expiration of 2013-03-09. Even the six-year § 286 damages lookback for pre-expiry conduct would have closed in 2019. If a demand letter cites US 6,090,925, the dispositive response is a calendar argument, not a validity challenge.
• If you nonetheless need to confirm the negative yourself, query the PTAB End-to-End system directly: https://ptacts.uspto.gov/ptacts/ (search "6,090,925"), and cross-check CourtListener for any Federal Circuit appeal — https://www.courtlistener.com/. Verify no ex parte reexamination exists, which is the one post-grant route not within the AIA-trial framing and which I did not independently confirm in this pass; flag that as an open item rather than assume it away.
• Absence is the signal here, and it is the strongest possible version of it. The general heuristic — "well-asserted patents eventually attract IPRs" — is confirmed in reverse: a broadly-licensed, Baxter-held microparticle patent that spawned an entire commercial platform (Larry Brown's later EPIC/Baxter microparticle filings) was never IPR'd, because it died at the 20-year term on 2013-03-09 before any challenger had a reason to act. Note the asymmetry versus the Ethicon cluster above: Baxter's later, unexpired polymeric-composition patents drew six IPRs in a single day; this one drew none.
• One honest caveat on my method. My searches did not return a direct PTAB E2E docket page for this number; the "zero proceedings" conclusion rests on the structured ODP data plus a consistently empty, non-contaminated web result set. I found no contradicting evidence. Given the tool budget, treat the ODP block as authoritative and the web check as corroboration, not as an independently docket-verified negative.

I'll search for the assignment records, reel/frame data, and correspondent information before reconstructing the chain.

Let me try to surface reel/frame and correspondent data directly.

Let me make one more attempt to surface reel/frame and correspondent-of-record data, and check the sibling patent's chain.

Data-limitation notice (read first)

Two of the fields you asked me to capture — reel/frame and correspondent of record — are not exposed by any source I could reach. The Google Patents legal-events mirror records recording dates, conveyance type, assignor, and assignee only; it does not carry reel/frame or the recording correspondent. My attempts to pull the Assignment Center abstract-of-title page (assignmentcenter.uspto.gov / assignment.uspto.gov) for 6090925, and to surface reel/frame or correspondent text through general search, returned no such fields. I am therefore not fabricating reel/frame numbers or attorney names. Where a field could not be retrieved, it is marked not retrievable. This also means NPE signal #3 (repeat correspondent) cannot be assessed and is reported as unclear rather than guessed.

A second caveat: the dates below are recording dates, not execution dates. Google Patents' reassignment output gives the recordation date; execution dates are a separate field I could not retrieve. Treat each date as "recorded on."

Inventors

All six named inventors were Massachusetts residents concentrated in the Boston suburbs (Milford, Newton ×2, Winchester, Norwood, West Roxbury), consistent with a single common employer at the time of filing: Middlesex Sciences, Inc., of Norwood, MA — the predecessor name of Epic Therapeutics, Inc. (Source: FreePatentsOnline record for 6,090,925.)

Unusual-pattern check — negative. The "all inventors depart within 12 months" fire-sale tell is not present. The evidence points the other way: Epic operated as a wholly owned Baxter subsidiary in Norwood, MA for years after the 2002 acquisition, retaining its scientific staff. A 2007 industry article identifies Larry Brown as Epic's CTO, Anthony Garramone as Epic president, and Julia Rashba-Step as director of formulation research at "Epic Therapeutics, Inc., a wholly owned subsidiary of Baxter Healthcare Corporation" (drug-dev.com, June 2007 PDF). Founder-inventors staying post-acquisition is the opposite of a pre-fire-sale exodus.

Note on the record: Woiszwillo is separately credited in the '925 specification as the named inventor of the polymer-preparation work in U.S. 5,525,519 and WO 93/14110 (PCT/US 93-00073) — i.e., he is the founder-level inventor anchoring this family.

Original assignee

Middlesex Sciences, Inc. (Norwood, MA), which changed its name to Epic Therapeutics, Inc. The '925 issued on the face in the name of Epic Therapeutics, Inc. (FreePatentsOnline: "Assignee: Epic Therapeutics, Inc. (Norwood, MA)").

• Primary line of business: a privately held drug-delivery / formulation company, not a holding vehicle. Its platform was PROMAXX, a water-based process for making uniform protein microspheres.
• Did it ship a product embodying the claims? It commercialized the technology into a named pipeline asset and was running formulation work for partners. The lead program was LeuProMaxx (sustained-release leuprolide), described as entering Phase III for a 28-day formulation with an 84-day formulation in Phase II; PROMAXX insulin microspheres had been demonstrated in dry-powder and metered-dose inhalers (Pharmaceutical Technology, Feb. 2003). This is a products/services operating company, not an NPE.
• Current status: acquired. Baxter Healthcare Corporation signed a definitive agreement on Nov. 11, 2002, closed in November 2002, for $50–100 million (Baxter/PRNewswire release via Silicon Investor; Chicago Tribune, Nov. 12, 2002; PitchBook — "Epic Therapeutics was acquired by Baxter International"). No bankruptcy.

Assignment timeline

Sourced from the Google Patents legal-events (reassignment) record for US 6,090,925. Execution dates and reel/frame are not retrievable from the sources available to me.

• executed ? / recorded 1996-10-17 — Reel not retrievable/not retrievable

  • Conveyance: Assignment of Assignors' Interest
  • Assignors: Blizzard, Charles D.; Sudhalter, Judith; Brown, Larry R.; Di, Jie; Scott, Terrence L.; Woiszwillo, James E.
  • Assignee: Middlesex Sciences, Inc.
  • Correspondent: not retrievable
  • Context: Initial inventor-to-company assignment of rights (standard employee-invention recordation), one of the two records filed ~two months after the 1996-08-19 filing.

• executed ? / recorded 1998-06-05 — Reel not retrievable/not retrievable

  • Conveyance: Change of Name
  • Assignor: Middlesex Sciences, Inc.
  • Assignee: Epic Therapeutics, Inc.
  • Correspondent: not retrievable
  • Context: Internal reorg / name change only — same legal entity, no change in beneficial ownership. Corroborated by PitchBook: Epic Therapeutics "Formerly Known As Middlesex Sciences."

• executed ? / recorded 1998-07-13 — Reel not retrievable/not retrievable

  • Conveyance: Assignment of Assignors' Interest
  • Assignor: Riske, Frank J.
  • Assignee: Epic Therapeutics, Inc.
  • Correspondent: not retrievable
  • Context: Internal employee/inventor assignment to the company. Riske is not a named inventor on the '925; he is a named inventor on sibling US 5,981,719 (same title). Treat this as a related-application assignment landing in the same family record, not a transfer of the '925 itself.
  • Flag vs. prior section: this matches the earlier-generated note that Riske appears in the '925 assignment record despite not being an inventor on the '925. No contradiction.

• executed ? / recorded 2003-06-10 — Reel not retrievable/not retrievable

  • Conveyance: Assignment of Assignors' Interest
  • Assignor: Epic Therapeutics, Inc.
  • Assignee: Baxter International Inc. and Baxter Healthcare SA (a corporation of Switzerland)
  • Correspondent: not retrievable
  • Context: M&A acquisition — Baxter's purchase of Epic (announced 2002-11-11, closed Nov. 2002); the recording is a routine post-closing title perfection ~7 months after close.
  • Independent corroboration: the European sibling EP 0 688 429 shows a French national-register transfer of total ownership recorded 12/11/2003 (INPI data) — same transaction, same window, different national register.

Finding on completeness: the chain does not end in an anonymous LLC. It ends at Baxter International Inc. / Baxter Healthcare SA, a NYSE-listed operating company, and the patent then expired on 2013-03-09 for failure to reach its 20-year term being asserted — legal status "Expired – Lifetime." A useful cross-check: Baxter later spun its biopharma business out as Baxalta GmbH in 2015, and some Baxter families were assigned to Baxalta (visible in Australian IP records), but the '925 expired in 2013, two years before the Baxalta spinoff, so it would not have been swept into that divestiture.

Timeline diagram

timeline
    title Ownership of US 6090925
    1993 : Priority application filed Mar 9
    1996 : CIP filed Aug 19
         : Inventors assign to Middlesex Sciences
    1998 : Name change to Epic Therapeutics
         : Riske assignment recorded
    2000 : Patent issued Jul 18
    2003 : Epic acquired by Baxter
         : Baxter files title perfection record
    2013 : Patent expires Mar 9

NPE / troll-pattern signals

Verdict

Operating-company assertion.

Forced to pick one, this is the only bucket whose factual predicate holds: the chain runs operating company → operating company throughout (Middlesex Sciences → Epic Therapeutics name change recorded 1998-06-05 → Epic's acquisition by Baxter recorded 2003-06-10), and the current holder, Baxter International Inc./Baxter Healthcare SA, is an operating company that commercialized the claimed technology as PROMAXX / LeuProMaxx. No NPE-typical signal is present: no shell LLC, no listed asserter, no cascading LLC transfers, no bankruptcy sale, no defensive aggregator. The one honest caveat is that the "assertion" prong is undocumented — no litigation naming the '925 was found, and the patent expired 2013-03-09 — so the accurate characterization is operating-company-owned, non-asserting. Note also that signal #3 is unresolvable, because reel/frame and correspondent were not retrievable; if anyone needs that field, pull the abstract of title directly from the USPTO Assignment Center search page for patent 6090925.

I'll research this systematically — first confirming the record, then pulling the patent's cited references and the claim text (which was truncated in my earlier pass).

I have the patent's cited reference lists partially. Let me pull the full front-page reference list and the claim text.

Prior-art analysis — US 6,090,925

0. Scope, and three limits you need before reading the tables

What you asked: pull the USPTO record for US 6,090,925 and analyze each patent citation on the face of that patent — full citation, publication/filing date, brief description, and the claims each could potentially anticipate under 35 U.S.C. § 102.

What I could and could not retrieve. I hit my tool-call ceiling mid-research, so I am reporting exactly what I have and no more:

1. The reference lists I retrieved are partial. FreePatentsOnline's record for US 6,090,925 returned the "US Patent References" list truncated with ellipses and beginning at US 5,851,451 (Dec. 1998). That means the retrieved slice is biased toward the newest examiner citations; the older US citations were cut off. Similarly, the everypatent.com list I retrieved begins mid-way through the JP series at JP 6-316524 (Nov. 1994). Do not treat the tables below as the complete front page.
2. Filing dates for the cited references were not retrieved. I have issue/publication dates (from the front page) but not the underlying US filing dates. This matters enormously for § 102(e) — see § 4.
3. I still do not have the claim text of US 6,090,925 (the same truncation I flagged in the earlier summary section, which ended at "Prior to being incorporated in…"). I therefore will not assign claim numbers to references where the mapping is not independently grounded. Where I give a claim family, I say so.

One decoy to name explicitly, per your "no similar numbers" instruction: OSTI record 6090925 ("Method of cooling and deashing," US 4,474,584, Texaco Development Corp.) is an OSTI ID, not a patent number. It is unrelated. Nothing from it is used below.

1. What "patent citations for 6090925" means here

There are three distinct citation populations, and conflating them is the most common error in this kind of analysis:

Everything in § 2 and § 3 below is population A. The Background art (population B) is treated separately in § 5 because some of it is the most legally significant material for § 102, and it is easy to miss.

2. United States patent documents cited (partial — see Limit 1)

Citations as printed on the face of the patent. Descriptions are inferred from the printed title unless I say otherwise — I did not retrieve abstracts for most of these.

Truncation warning, restated: the list above is the tail of the US citations. Older US references — precisely the ones most likely to be pre-1993 § 102 art — were not returned. Treat this as approximately the most recent third of the US list.

3. Foreign patent documents cited (partial)

Identifier-integrity note (per your strict rule): the GB numbers appear in the source as 02017125, 02145992, etc., and the JP numbers in an 8-digit fused form (407097334). I have reproduced them literally and offered the conventional reading in parentheses. I have not auto-corrected any number. Flagging two possible conflations rather than resolving them: the specification's Background cites "European Patent Application Publication Number 248,531 to Southern Research Institute," while the front page carries GB 2,245,831 (Jan. 1992). Those are different numbers and different decades; I have no basis to treat one as the other.

4. The problem with this citation list that decides the whole § 102 question

This is the most important analytical point in the answer, and it is not obvious from the tables.

US 6,090,925 has a priority date of March 9, 1993 (via Ser. No. 08/028,237), a CIP filing date of August 19, 1996 (Ser. No. 08/699,586), and issued July 18, 2000.

Almost every U.S. citation I retrieved issued in 1998 — i.e., after both candidate effective filing dates. Under pre-AIA § 102, that makes them unusable as § 102(a)/(b) art unless their § 102(e) dates save them. And pre-AIA § 102(e) is keyed to the reference's U.S. filing date, not its issue date:

a patent granted on an application for patent by another filed in the United States before the invention thereof by the applicant (pre-AIA § 102(e)(2)).

So a reference that issued in September 1998 but was filed in 1992 would be § 102(e) prior art against a March 1993 invention date — and a reference filed in 1997 would not be, no matter how early it issued.

I did not retrieve the filing dates. I therefore cannot tell you, for any single 1998-issued citation, whether it is § 102 prior art at all. Anyone who hands you a confident per-reference § 102 table for this patent without those filing dates is guessing. The retrieval step that resolves it is one click per reference — the "Worldwide applications" tab on each reference's Google Patents page, or the reference's PatentCenter continuity data.

Two corollaries worth stating plainly:

• Against claims entitled to the March 9, 1993 date, only art published before March 9, 1993 (§ 102(a)/(b)) or filed in the U.S. before that date (§ 102(e)) can anticipate. That is a narrow set: the GB/NL/WO documents from 1972–1992 and any US patents filed pre-1993.
• Against claims entitled only to the August 19, 1996 CIP date (i.e., claims to matter first added in the CIP), the 1998-issuing references become live § 102(e) candidates if their U.S. filing dates precede the applicant's invention date.

5. Which citations actually matter — and the one I'd put first

5.1 The single most significant citation: WO 93/14110 (PCT/US93/00073)

• Full citation: WO 93/14110 A1, "…," applicant/inventor James E. Woiszwillo; PCT/US93/00073, international filing date January 7, 1993; published July 22, 1993.
• Why it is first: its international filing date of January 7, 1993 precedes the '925's March 9, 1993 priority date by roughly two months. If it designates the United States and was published in English, it is a candidate § 102(e)(1)/(2) reference as of its international filing date — the only citation on this list that can reach the '925's earliest-entitled claims on a filing-date basis. Its U.S. national-phase counterpart is cited in the '925 specification itself as U.S. Pat. No. 5,525,519 (Woiszwillo), which the patent describes as the source of its polymer preparation methods.
• The "by another" wrinkle: § 102(e) requires the reference be "by another." WO 93/14110 / '519 name Woiszwillo alone; the '925 names six inventors (Woiszwillo, Brown, Scott, Di, Sudhalter, Blizzard). Different inventive entities — so the "by another" hurdle is technically cleared. That is precisely what makes this a live § 102(e) reference rather than a mere § 120 benefit/obviousness-type double-patenting issue.
• Claims potentially affected: the process claims (mixing macromolecule with a soluble polymer at a pH near the macromolecule's pI, applying an energy source) and any claim reciting the polymer species/concentration limitations, because the specification itself concedes the polymer work comes from this reference. Caveat: I am identifying the claim family by limitation, not by claim number — I do not have the '925 claim text.

5.2 The coacervation/phase-separation references — closest third-party art on the process claims

5.3 The composition claims (≥40% macromolecule, polymer ≤30%, <10 µm, water-permeable)

5.4 Nucleic-acid / gene-therapy claims

• US 5,783,567 — Microparticles for delivery of nucleic acid (Hedley), Jul. 1998, class 514/44. This is the on-point citation for the '925's DNA-microparticle and gene-therapy subject matter (Examples 8 and FIG. 7 territory). Same § 102(e) filing-date caveat.

5.5 Peptide / sustained-release and administration-route claims

• US 5,776,885 — Sustained and controlled release of water-insoluble polypeptides (Orsolini), Jul. 1998, class 514/2.
• US 5,792,475 — Lymphatic delivery composition (Davis), Aug. 1998.
• US 5,804,212 — Small particle compositions for intranasal drug delivery (Illum), Sep. 1998, class 424/434. Relevant to the '925's particle-size/route claims (the specification's stated objects include avoiding large-bore needles and enabling inhalation delivery).
• US 5,759,551 — Immunogenic LHRH peptide constructs and synthetic universal immune stimulators for vaccines (Ladd), Jun. 1998. Relevant to the LHRH/leuprolide/nafarelin and vaccine claims.
• US 5,753,234 — Single-shot vaccine formulation. Relevant to the vaccine/adjuvant claims.
• RE35,862 — Delivery systems for pharmacological agents encapsulated with proteinoids (Steiner), Jul. 1998, class 424/455. Relevant to protein-matrix microparticle claims.

5.6 Diagnostic / separations claims

• US 5,795,719 — Biotinylated latex microsphere, process for preparation and use as agent for biological detection (Richard), Aug. 1998, class 435/6.
• US 5,776,706 — Polymeric particles with biodegradable gelatin or aminodextran coating (Siiman), Jul. 1998, class 435/7.21 — dextran-coated particles, which touches the '925's dextran embodiment.
• US 5,770,222 — Therapeutic drug delivery systems (Unger), Jun. 1998, class 424/450 (liposomal).

5.7 Background art cited in the specification (population B) — legally the cleanest § 102 material

These are pre-1993 and therefore are not subject to the filing-date problem of § 4. They are the applicant's own characterization of the prior art:

6. Direct answer on anticipation, with the honesty it requires

Can I tell you which claim(s) each reference anticipates under § 102? Not to claim-number precision, and I will not manufacture it. Two independent gaps cause this:

• Gap 1 — no claim text. My copy of US 6,090,925 truncates inside the Detailed Description. I established in the earlier section that the claim set is large (a JP search report for JP 2012500118A cites US 6,090,925 as an X reference against that application's claims 1–10 and 14–25, pinpointing "column 13, lines 10–21; example 5" — implying at least 25 claims with a separate independent claim or claims in the 11–13 gap). That is structure, not text. Assigning claim numbers from it would be fabrication.
• Gap 2 — no filing dates for the cited references. Without them, § 102(e) status is indeterminate for the entire 1998 cohort, which is most of what I retrieved.

What I can give you is the limitation-to-claim-family mapping in § 5, which is grounded in the abstract, the Summary of Invention, and the specification's own recitation of objects. The three claim families the disclosure frames — "Microparticles, methods of production, and methods of use thereof" — map to the citation tiers as follows:

Bottom line on anticipation: on the record as retrieved, the only citation with a clean, date-qualified § 102(e) theory against the earliest-entitled claims is WO 93/14110 / US 5,525,519 — and it is the applicants' own earlier work, which is exactly why it is important that it is not co-extensive in inventive entity with the '925. Nearly everything else on the retrieved front page is a § 103 reference (or a § 102(e) reference only against CIP-date claims), not an anticipation reference. And nothing I retrieved is a § 102(b) reference against the March 9, 1993 date.

7. Corrections and additions to the earlier sections

I am flagging these rather than silently reconciling them:

1. Family membership was under-inclusive in my earlier "Patent summary." I listed US 5,981,719, US 5,578,709, US 6,268,053, US 5,525,519. Search now shows the PROMAXX family as recited in the later art itself — JP 2011500054A names US 5,554,730; 5,578,709; 5,981,719; 6,090,925; and 6,268,053 together as the PROMAXX patents. US 5,554,730 is new. So is US 5,849,884 (Woiszwillo, Dec. 1998), which appears in a later patent's citation table with an unexplained "(withdrawn)" annotation.
2. Corroboration, not contradiction, on the European sibling. I earlier noted EP 1060741 ("Microsphères à libération prolongée," Epic Therapeutics / Baxter, revoked 28/02/2017). The search retrieved EP1060741B1's text, and paragraphs [0045]–[0051] are near-verbatim identical to the '925's pH/pI, pH-adjustment, energy-source, incubation-time, and macromolecule-definition passages. One substantive divergence worth noting: EP 1060741 [0047] recites exposure "to a crosslinking agent and/or an energy source," language absent from the '925. That is a claim-scope evolution within the family, not a § 102 issue for the '925.
3. Litigation cross-check. My earlier litigation section reported no case asserting the '925, and explained why (expired 2013-03-09). The searches this round surfaced IPR2025-00514 (Apotex Inc.), whose Exhibit 1065 is US 2003/0211167 A1 — a document that discusses US 5,981,719 (a family member, not the '925) as prior art and criticizes its heating/crosslinking conditions and sub-0.1 µm particle sizes. This does not change my earlier negative finding on the '925 itself, and I am not representing it as litigation involving the '925. It is a lead worth pulling if you care about the family.
4. Date inconsistency, restated. The task header says current date April 26, 2026; my operating context and index show September 26–27, 2026. I flagged this earlier and remain unable to resolve it. Nothing in this answer turns on it.

8. What I would do next, in order

1. Pull the claim text — patents.google.com/patent/US6090925A/en (scroll to Claims), or the issued PDF at patentimages.storage.googleapis.com/.../US6090925.pdf, or USPTO PatentCenter for application 08/699,586. Without this, no claim-number level § 102 mapping is defensible.
2. Pull the complete front page. The lists in § 2 and § 3 are truncated at the old end — the pre-1993 US citations are the ones most likely to be true § 102 art and they are the ones I am missing. everypatent.com/comp/pat6090925.html and FreePatentsOnline's full page carry them.
3. Pull the U.S. filing date of every 1998-issued citation (Google Patents "Worldwide applications" tab per reference). This single step converts § 4 from an open question into a determinate § 102(e) yes/no per reference.
4. Confirm WO 93/14110's designation of the United States and English-language publication — that is the linchpin of the only clean anticipation theory I can currently support.
5. Retrieve the "Other Publications" list. I saw only one fragment of it on the front page of the '925 PDF — an item beginning "The controlled parenteral delivery of polypeptides and proteins," terminating at pages 259–265. I could not retrieve the author, journal, or year, so I am not citing it as a reference — I am reporting that it exists and is unretrieved. Non-patent literature is often where the actual anticipation reference hides in this art unit.

Again: partial lists, no filing dates, and no claim text. The analysis in § 4 and § 5.1 is the part I would stand behind; the rest is scoped by those three gaps.

I'll ground this in the prior-art references of record for the '925 patent before running the §103 analysis.

Good — I now have the actual front-page references. Let me pull the claim text and a couple of the key cited references.

Obviousness analysis — US 6,090,925 under 35 U.S.C. § 103

0. Sourcing caveats, and one correction to the earlier sections

No "Prior Art" section was supplied to me. The current task says to "use the results from the Prior Art section of this page," but the material I was given contains only the full patent text (which ends mid-Detailed-Description), the earlier generated summary/litigation sections, and the "Prior art keywords" line (microparticle, macromolecule, polymer, microparticles, polymers) from the Google Patents record. I did not receive a prior-art table. Per my operating rules I am stating this explicitly rather than inventing one.

To avoid fabricating, I retrieved and now rely on: (a) the prior art admitted in the '925 specification's own Background; (b) the References Cited list as rendered by a secondary database (everypatent.com/comp/pat6090925.html) — retrieved only partially (the US list from ~5,326,692 forward, plus the foreign list); and (c) family/search-report documents retrieved from patent office records. Where a reference's presence on the '925's own face is unverified, I label it.

Correction to the earlier "Patent summary" section. That section stated that the JP search report in JP2012500118A mapped "claims 1–10, 14–25" against the '925, "indicating at least 25 claims." That is a misattribution. In that search report the "Relevant to claim No." ranges (1–10, 14–25; 13, 26–36) belong to the application being searched — PCT/US2009/054504, which has 36 claims — not to the '925. The '925 was cited as the prior-art document (category X) against that later application, at "column 13, lines 10–21; example 5." I have no verified claim count for the '925. (JP2012500118A search report)

(Date note, carried forward: the task header says April 26, 2026; my environment and search index show September 26–27, 2026. I treat the later date as operative.)

1. The claim scope being tested — proxy, clearly labeled

I still do not have the '925's verbatim claims. For claim-scope structure I use the near-identical sibling US 5,849,884 (same title, same family, Woiszwillo et al.), whose claim text surfaced in search and is a reliable structural proxy — not the '925's text (justia: US 5,849,884):

So the '925's claim set almost certainly runs in three clusters: (i) process (mix macromolecule + soluble polymer at pI, apply energy); (ii) composition/product-by-process with the 40/30 wt-% limitations; (iii) method-of-use (therapeutic delivery, diagnostic/separations). All my conclusions are keyed to those clusters and must be re-verified against the granted text.

2. The threshold issue: critical date (this dominates everything else)

The '925 (Ser. 08/699,586, filed 1996-08-19) is a CIP of Ser. 08/206,456 (filed 1994-03-04, now US 5,578,709), itself a CIP of Ser. 08/028,237 (filed 1993-03-09, abandoned). Pre-AIA §§ 102/103 apply.

• If a claim is fully supported by the 1993 application, the § 102(b) bar date is 1992-03-09 and the intervening mid-1990s microsphere boom is irrelevant.
• If a claim depends on matter first added in the 1994 or 1996 CIP — and the "at least 40% / less than 100% by weight" and "≤30% polymer" limitations, plus the release-kinetics figures (Figs. 1–6) and the DOX/LHRH/nafarelin examples, are prime candidates — then the effective date slides to 1994-03-04 or 1996-08-19, and a large body of art becomes available.

Why this matters so much here: the documents citing this family (per the RU2426590C2 "citing" record) carry priority dates clustered between March 1993 and August 1996 — e.g. US 5,543,158 (Gref, MIT, 1993-07-23), EP 0727984 (1993-11-18), WO 95/13799 and US 5,650,173 (Alkermes, 1993-11-19), PL 319600 (Andaris, 1994-09-29), US 5,665,428 (Macromed, 1995-10-25), US 6,270,795 (MRA, 1995-11-09). Any of these with a § 102(e) filing date before the operative critical date is § 103-only prior art, and § 103(c) cannot disqualify it unless it is commonly owned.

Corollary for the applicant's own art: WO 93/14110 (PCT/US93/00073, filed 1993-01-07, published 1993-07-22) and US 5,525,519 (Woiszwillo) — both cited in the '925 specification itself — are the most dangerous references. Pre-AIA:

• Not § 102(b) if the claims hold a 1993-03-09 date (published after it); but § 102(a)/§ 102(b) if the claims slip to 1994-03-04 or 1996-08-19 (for 1996-08-19, it is comfortably >1 year before filing → § 102(b), which § 103(c) cannot cure).
• A PCT publication is not pre-AIA § 102(e) art for a 1993 filing, so the WO document's status is entirely date-driven.
• The co-pending parent US 5,578,709 issued 1996-11-26, after the '925 filing — not prior art at all; and being the same inventive entity/commonly owned, the other Woiszwillo references (US 5,484,894; 5,554,730; 5,552,519) are § 103(c)-disqualifiable if they are § 102(e)/(f)/(g) art. They are not disqualifiable if they qualify as § 102(a)/(b) printed publications. This is a live, dispositive fork.

I could not retrieve the text of US 5,525,519, so any statement about what it discloses is flagged as unverified.

3. Prior art available for combination

Tier 1 — admitted in the '925 specification (applicant's own characterization, citable as an admission):
US 5,213,812 (Ruiz); US 4,530,840, 4,897,268, 5,075,109, 5,360,610 (Tice); US 5,417,986 (Reid); US 5,102,872 (Singh); US 5,384,133 (Boyes); EP 248,531 (Southern Research Institute) — all PLGA/PLA microspheres by phase separation, solvent evaporation, emulsification, spray drying; US 4,904,479 (Illum, poloxamer 407); US 5,149,543 (Cohen, polyphosphazenes); US 5,422,120 (Kim, lipid vesicles); US 5,525,519 (Woiszwillo) + WO 93/14110.

Tier 2 — verified on the retrieved portion of the '925's References Cited: e.g. US 4,671,954 (Lee — hydrophilic polypeptide microspheres from an aqueous protein dispersion in an organic solution of a high-molecular-weight soluble polymer, then crosslinked); US 5,069,936 (protein microspheres; its own background cites Martodam et al., PNAS 76:2128–2132 (1979) for albumin microspheres by emulsification, Widder et al., Adv. Pharmacol. Chemother. 16:213–271 (1979), and US 4,107,288 (Oppenheim, nanoparticles)); US 5,543,158 (Gref, MIT); US 5,422,120 (Kim); Kossovsky 5,460,830/831, 5,462,750/751 (protein particles).

Tier 3 — field art located by search, presence on the '925 face unverified: EP 0 495 187 A1 (monodisperse sub-micron protein nanomatrices formed by simple mixing, without heating or crosslinking, with stabilizer-ratio control); US 4,147,767 (albumin spherules by heat-insolubilization at 110–180 °C, 5–80 µm, porosity/release tuned by time/temperature; expressly notes heat >110 °C degrades labile drugs and that room-temperature crosslinking was developed to avoid it); Gupta, Gallo, Hung & Perrier, Drug Dev. Ind. Pharm. (1987) and Gupta, Hung & Lam, J. Microencapsulation 6:147–160 (1989) (heat stabilization temperature/time and protein concentration control entrapment, recovery and biphasic release of albumin microspheres); the albumin-microsphere review literature teaching that pH of the albumin solution governs particle size — smallest particles at pH ~5, "attributed to overall change in electric charge of albumin molecules at pH values above or below isoelectric point" (albumin pI ≈ 4.7–5.0); and the classical protein-chemistry teaching that PEG and dextran dehydrate/volume-exclude proteins and that proteins are least soluble at their pI.

4. Reference-by-reference: why no single reference anticipates, and what each supplies

No single reference, on the record I can verify, discloses the full combination of (a) single aqueous phase, (b) soluble dehydrating polymer present in the particle at low wt-%, (c) pH near the macromolecule's pI, (d) heat as the sole energy source, (e) ≥40 wt-% macromolecule. The § 103 case is therefore a combination case.

5. Combinations that would render the claims obvious

Combination A — primary case (process + composition clusters)

Tice '268/'840 + US 4,671,954 + US 4,147,767 (or Gupta 1989) + PEG/dextran volume-exclusion & pI teaching.

• Claim elements: '268/'840 supply the microparticle/depot concept and the injectability motivation; '954 supplies "protein microsphere stabilized by a high-molecular-weight soluble polymer"; '147,767/Gupta supply "heat alone stabilizes/conjoins a protein particle, and temperature/time control the loading and release"; the volume-exclusion/pI teaching supplies "polymer removes water from the macromolecule" and "pH near the isoelectric point."
• Motivation (KSR / MPEP 2143 rationales):

1. Recognized problem in the field. The '925's own Background admits the field knew PLGA microparticles require organic solvents that "could denature proteins" and leave "residual organic solvents [that] could be toxic," cause a "burst effect," and require large-bore needles. Removing the oil/organic phase is an express design incentive — KSR, 550 U.S. at 421 (design incentives/market forces).
2. Known technique to improve a known device in the same way. Polymer-induced precipitation/volume exclusion of proteins at pI was a standard, predictable separation technique (cf. the parent '709's own admission that "in some precipitation processes, such as precipitation with ammonium sulfate … microparticles may also be created as a by-product"). Using the known dehydrating polymer to make the particle rather than to inhibit it is a predictable substitution.
3. "Obvious to try" with a finite, identified set of solutions. PEG, dextran, PVP, poloxamer were the recognized protein-compatible polymers; albumin microsphere art already taught that pH at the protein's pI minimizes solubility and yields the smallest particles.
4. Design choice: elimination of glutaraldehyde. '147,767 itself notes high-temperature insolubilization degrades labile drugs and that crosslinkers were introduced to avoid it; a skilled artisan seeking a mild, non-toxic route had only a few options (heat below denaturation, ionic completion, dehydration).

• Claim clusters reached: process cluster; the ≥40%/<100% and ≤30% weight limitations; the "heat 37–70 °C, 5 min–24 h" species; polymer-genus species.

Combination B — the weight-percentage limitations specifically

Gupta 1989 (25% w/v albumin microspheres, ~4–12% w/w drug) + any of Combination A.
Gupta's heat-stabilized albumin microspheres are overwhelmingly protein by weight. The ≥40%-macromolecule limitation thus reads on the prior art product class; the ≤30%-polymer limitation is merely the arithmetic consequence of the '925's own 2:1 polymer-solution:protein-solution ratio. Under In re Aller and In re Kao, a recited numerical range is obvious absent evidence of criticality; the specification asserts no criticality for the 40% and 30% endpoints — it presents them as preferences ("Preferably, the concentration of polymer is less than 30%"). These two claims are the most vulnerable in the set.

Combination C — method-of-use cluster (therapeutic delivery)

US 5,480,656 / 5,476,663 (Okada; LHRH analogs in sustained-release microspheres) or US 5,389,613 (Labrie, LHRH) + any of Combinations A/B. Okada directly teaches depot LHRH/GnRH analog microspheres for the same indications (prostate cancer, endometriosis) the '925's Examples 8 and 12 address. Administering a sustained-release peptide microparticle is old; substituting the '925's particle is an obvious substitution of a known carrier. The delivery-method claims are, on this record, highly exposed.

Combination D — nucleic-acid particles

Polylysine–DNA/Ca-phosphate complexation and receptor-mediated gene-delivery art (Wu & Wu; the cationic-lipid patent family cited in the '925's own reference list) + a protein carrier particle. The '925's own text explains that nucleic acids "are anions," so a cation such as polylysine "aids greatly in the formation of microparticles" — ionic condensation of DNA with polycations was routine by 1993. Claims to DNA/polylysine microparticles, and to their use for transfection, are strong § 103 candidates. (Note: US 5,783,567 (Hedley, nucleic-acid microparticles) issued 1998-07-28 — after the '925 filing, so it is not prior art against the '925 despite appearing on its face.)

Combination E — surface decoration, targeting, diagnostics

Kossovsky protein-particle patents + US 5,543,158 (Gref: PEG/surface moieties on biodegradable nanoparticles) + the antibody-bead/immunoaffinity art the '925 itself describes in the Background. Attaching antibodies, receptors, phospholipids or polysaccharides to a protein microparticle, and using microparticles in immunoaffinity chromatography or ELISA, is the express subject matter of the admitted prior art. These claims add little patentable weight.

6. Where the § 103 case is weak (the applicant's real defenses)

1. Emulsion/heat teaching away. The entire protein-microsphere art taught a water-in-oil emulsion and stabilization at 100–180 °C or with glutaraldehyde. The '925 forms particles in a single aqueous phase at 37–70 °C with no crosslinker. Under In re Gurley/DePuy Spine, art that "criticizes, discredits or otherwise discourages" the aqueous, low-temperature route is a teaching-away argument — and the '147,767 discussion of drug degradation at >110 °C supports it.
2. Non-crosslinked protein particles were believed unstable. As late as US 6,391,343 (2002), the field was still patenting ways to keep non-crosslinked protein particles "stable against resolubilization." That is evidence of a long-felt need and earlier failure, weighing against obviousness.
3. Comparative release data. Figs. 1–6 are the strongest evidence: a sustained, tunable release without the PLGA burst. If the applicant produced a comparative showing that the closest aqueous art produced an initial burst or rapid resolubilization, that is unexpected-results evidence.
4. Availability of the most dangerous reference is date-dependent. If WO 93/14110 / US 5,525,519 are excluded (same inventive entity, or § 103(c)-disqualified, or the claims hold the 1993-03-09 date), the remaining art all requires an emulsion or an organic solvent, and the obviousness case weakens materially.
5. Nexus problems cut both ways. Commercialization as ProMaxx/LeuProMaxx (Baxter's acquisition of Epic Therapeutics, 2003) supports secondary considerations only if a nexus to the claimed features is shown.

7. Bottom line

• Weakest (most likely invalid) claims: (a) the product/composition claims reciting only the ≥40%/<100% macromolecule and ≤30% polymer weight ranges; (b) dextran and PVP/PEG polymer-species claims; (c) the therapeutic-delivery method claims (Combination C); (d) surface-decoration/targeting and diagnostic-use claims (Combination E). For (a) and (b), a prima facie case under § 103 is made out by Tice/Reid/Ruiz + US 4,671,954 + US 4,147,767/Gupta + the volume-exclusion/pI teaching, with the ranges attacked as non-critical optimizations (In re Aller, In re Kao, KSR).
• Strongest claims: any claim whose scope is tied to the combination of single-aqueous-phase formation at pH ≈ pI with heat as the sole energy source and no emulsion, especially where the specification's release-kinetics data and bioactivity-retention showings carry nexus. Those claims have credible teaching-away and unexpected-results answers.
• Everything turns on two facts I could not establish: (i) the actual claim language of the '925, and (ii) whether the challenged claims are entitled to the 1993-03-09 priority date. If priority slips even to 1994-03-04, WO 93/14110 (published 1993-07-22) and a wave of 1993–1995 filings become available art and the § 103 exposure rises sharply.
• Practical note: Google Patents records the patent as Expired – Lifetime (anticipated expiration 2013-03-09), so a § 103 determination today is retrospective — relevant to past damages, § 285, or an ex parte reexamination (in which the patent owner would be unable to amend claims of the expired patent into enforceable form).

To close the two open facts, the authoritative sources are the granted claims at patents.google.com/patent/US6090925A/en / patentimages US6090925.pdf, the file history (priority/support rejections, § 103 rejections and the IDS) via USPTO PatentCenter for application 08/699,586, and the full front-page References Cited list — I retrieved only a partial rendering of that list and did not verify it against the printed front page.