US 10093647

I'll search for authoritative information on this specific patent number and check current litigation status.

US Patent 10,093,647 B1 — Analysis

Sourcing note: I do not have direct, live access to USPTO PatentCenter or PACER/CAFC docket query tools in this session. The summary below is grounded in (a) the authoritative full patent text supplied for this patent, and (b) secondary public sources (court filings, Google Patents, Justia) returned by web search. Where I could not verify something directly, I flag it. I did not find any 2026 Court of Appeals for the Federal Circuit docket specifically naming U.S. Patent 10,093,647; the Federal Circuit activity found (21-1154, 21-2335) dates to 2021.

Bibliographic Data

The compound is pomalidomide, the active ingredient in POMALYST®.

Abstract (as issued)

"Provided herein is a crystalline 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione dihydrate. Pharmaceutical compositions comprising the crystalline 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione dihydrate are also disclosed."

Independent Claim — Plain-Language Overview

Despite the broad "compositions and methods" framing of the title, the patent has exactly one independent claim, and it is directed to a crystalline solid form. All other claims depend from it.

Claim 1 — A crystalline dihydrate of pomalidomide (i.e., the pomalidomide molecule co-crystallized with approximately two molecules of water per molecule of drug), characterized by an X-ray powder diffraction (XRPD) pattern that includes peaks at 13.9, 16.6, and 25.5 degrees 2θ (± 0.2°). In plain terms: this is a defined water-containing crystal form of pomalidomide identified by three signature XRPD peaks.

Representative dependent claims (all narrow the characterization of the same crystal form):

• Claim 2 — Adds three more required XRPD peaks (11.9, 16.9, 28.2 °2θ), for six total.
• Claims 3–4 — Define the form by reference to the XRPD pattern in FIG. 1 and FIG. 2, respectively.
• Claims 5–8 — Define the form by DSC thermogram: endotherm maximum at about 308 °C (claim 5) or about 309 °C (claim 6), or by correspondence to FIG. 3 (claim 7) or FIG. 4 (claim 8).
• Claim 9 — Defines the form by water content: about 11.6% water by mass.
• Claim 10 — TGA weight loss of between about 10.1% and about 10.4% when heated from about 30 °C to about 225 °C.
• Claim 11 — TGA thermogram corresponding to FIG. 5.
• Claim 12 — IR spectrum corresponding to FIG. 6.

Important scoping note: The specification discusses pharmaceutical compositions, dosage forms (0.1–1,000 mg/day ranges, capsule excipients, etc.), and methods of treating multiple myeloma, but these are not captured in the granted claims. The claims are limited to the crystalline dihydrate per se and its analytical fingerprints. There are no method-of-treatment or composition claims in this patent.

Preparation (Example): Pomalidomide Form A (anhydrate) slurried in 4:1 1,4-dioxane/water, dissolved at 70 °C, evaporated under vacuum at 90 °C, triturated with water, centrifuged, and vacuum dried. Karl-Fischer titration gave 10.5% water. Alternative solvent systems that worked included 1:1:1 acetone/water/isopropanol and 4:1 THF/water.

Litigation / Post-Grant Context

This patent is one of a set of three "crystal form" (polymorph) patents — 10,093,647 (dihydrate), 10,093,648 (hemihydrate), and 10,093,649 (monohydrate) — all listing Atwood and all issued on the same date (2018-10-09). Multiple D.N.J. suits were filed against generic ANDA filers (e.g., case nos. 2:19-cv-05797, 05799, 05802, 05804, 05806, 08758, 09737, and 2:21-cv-02111), with Federal Circuit appeals 21-1154 and 21-2335 in 2021. Subsequent antitrust litigation (e.g., the CenterWell and Cigna complaints against Celgene/BMS) characterizes the polymorph patents as having been asserted in allegedly baseless "sham" litigation and raises Walker Process theories of fraudulently obtained patents; the reverse-payment settlement narrative points to generic pomalidomide entry no earlier than Q1 2026.

Caution on a common mix-up: Do not confuse this patent with U.S. 9,993,467 ("the '467 patent"), which is a formulation patent (oral dosage form / capsule with pomalidomide, starch, and mannitol) — a different patent with an easily conflated number. The patent discussed here is 10,093,647 (the crystal-form dihydrate patent).

Where I Have Lower Confidence

• Live USPTO status and 2026 CAFC docket: I could not directly query PatentCenter or PACER. Google Patents lists the patent as "Active" with anticipated expiration 2037-12-20, but I cannot independently confirm current maintenance-fee status or any pending 2026 appeal specifically involving this patent number.
• Exact litigation disposition of the '647 patent itself (as opposed to the related polymorph patent family): the public filings I retrieved discuss the polymorph patents collectively rather than resolving each individually. I did not find an authoritative final judgment or IPR decision specific to 10,093,647.
• Any statements about invalidity, fraud, or sham litigation above are allegations drawn from third-party complaints, not adjudicated findings.

I'll search for litigation involving US patent 10093647. Let me run several targeted searches.

Let me dig deeper into the specific cases and the Unified Patents litigation records.

Let me confirm the specific case numbers and their outcomes.

Litigation Involving U.S. Patent No. 10,093,647 ("the '647 patent")

Note on the identifier (strict-rule compliance): I interpreted the ID literally. "US10093647" corresponds to U.S. Patent No. 10,093,647 B1, titled "Crystalline 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione dihydrate, compositions and methods of use thereof," issued October 9, 2018, to inventor Jerry Lee Atwood, assigned to Celgene Corporation (later a Bristol Myers Squibb subsidiary). It is one of a three-patent "polymorph"/crystal-form family: the '647 (dihydrate), '648 (hemihydrate), and '649 (monohydrate). I did not return results for any other number.

Bottom line: The '647 patent was asserted by Celgene in a wave of Hatch-Waxman infringement suits (all in D.N.J.) beginning February 2019, and it appears in subsequent antitrust litigation as part of the alleged Pomalyst monopoly scheme. I found no separate ITC action or non-pharma assertion.

A. Hatch-Waxman patent infringement cases asserting the '647 patent

All filed by Celgene Corporation as plaintiff in the U.S. District Court for the District of New Jersey (D.N.J.), assigned to Judge Esther Salas (Magistrate: Michael A. Hammer). The complaints alleged infringement of the '647, '648, and '649 patents under 35 U.S.C. § 271(e)(2) arising from the defendants' Pomalyst® (pomalidomide) ANDAs.

Related docket also linked to this family: 2:19-cv-05799 (Celgene v. Aurobindo) and 2:21-cv-02111 appear in the patent's litigation record but I could not confirm within my search budget which specific patents each asserted.

Key caveat: The '647 patent is not Orange Book–listed (per the antitrust complaints, the '647/'648/'649 were "Not OB Listed"). Consequently, suits on these patents did not trigger a new 30-month FDA stay; the cases were litigated against the backdrop of the earlier 2017 formulation/method-of-use actions.

There is no publicly reported bench or jury verdict of infringement or invalidity as to the '647 patent — the D.N.J. cases were resolved by dismissal (venue), consolidation, or the global 2020–2021 settlement wave in which Celgene/BMS settled with the generics (Natco/Breckenridge, Teva, Aurobindo, Apotex, Mylan, Hetero, Alvogen/Synthon) on terms delaying U.S. generic pomalidomide entry until Q1 2026.

B. Federal Circuit appeals

C. Antitrust / consumer-protection litigation referencing the '647 patent (not infringement of the '647 itself)

These are not patent-infringement suits on the '647 patent, but the '647 patent is pleaded as part of the alleged scheme to delay generic pomalidomide:

• In re Pomalyst (pomalidomide) Antitrust Litigation / class actions — S.D.N.Y. — Beginning September 2023, entities filed putative class actions against Celgene, Bristol Myers Squibb, and individuals alleging antitrust, consumer-protection, and unjust-enrichment claims arising from obtaining and litigating the Pomalyst patents (including the '647 family). Per BMS's Q1 2025 Form 10-Q: in March 2025 the court dismissed the complaints against Celgene, BMS, and the named individuals; plaintiffs sought leave to amend.
• CenterWell Pharmacy, Inc. v. Celgene Corp. & Bristol Myers Squibb Co. — S.D.N.Y. (docket referenced as No. 1:23-cv-07871). Monopolization claims under Section 2 of the Sherman Act citing the '647/'648/'649 as part of the alleged pay-for-delay scheme.
• The Cigna Group v. Celgene Corp. & Bristol Myers Squibb Co. — D.N.J., No. 1:25-cv-05237-ER, filed June 24, 2025. Sherman Act § 2 / Clayton Act claims; the complaint extensively discusses the '647 patent's prosecution and alleged invalidity.
• A D.N.J. antitrust complaint (docket referenced as nj d. 498914) similarly frames the '647 family within the alleged monopoly scheme against Revlimid/Thalomid competitors.

Confidence and caveats

• High confidence: Celgene asserted the '647 patent in D.N.J. Hatch-Waxman cases in Feb.–Apr. 2019; the Mylan case number 2:19-cv-05802, filed Feb. 14, 2019, dismissed for venue and affirmed in Fed. Cir. 21-1154 (Nov. 5, 2021); Hetero 2:19-cv-05797; Breckenridge/Natco 2:19-cv-05804; Synthon 2:19-cv-09737 (filed Apr. 12, 2019); and the global 2026 entry settlements.
• Medium confidence / not independently re-verified in this session: the exact mapping of 2:19-cv-05806 → Apotex, 2:19-cv-08758 → Teva, and the captions/dispositions of 2:19-cv-05799, 2:21-cv-02111, and Fed. Cir. 21-2335.
• Not found: any ITC § 337 investigation, any judgment holding the '647 patent valid/infringed or invalid, and any litigation outside the U.S. Pomalyst/Revlimid context.

If you need docket-level verification (PACER/RECAP) of the specific case numbers flagged above, I can pull those next.

I'll verify the structured "no PTAB activity" data point against live sources before writing.

Proceedings overview

Zero. As of 2026-10-01, there are no AIA trial proceedings on file against U.S. 10,093,647 — no IPRs, PGRs, or CBMs; therefore no active trial, no claims invalidated, no claims sustained, no settlement, and no institution denial to report. The structured USPTO ODP block in this prompt returns an empty proceeding list, and my independent web checks are consistent with that (see "Verification trail" below).

Bottom-line defensive posture: this is the opposite of a hardened patent and the opposite of a paper tiger — it is an untested patent. No Article III tribunal or the PTAB has ever adjudicated the validity of claim 1. If a demand letter cites 10,093,647, you cannot point to a canceled claim as sanction-bait; but you also have a completely clean evidentiary slate, no petitioner estoppel working against you, and (critically) the patent's strongest invalidity theories are non-IPR-eligible and must be run in district court.

Proceedings

None. There is no proceeding number, panel, institution decision, FWD, or appeal to report for US 10,093,647. I am not going to manufacture one.

Verification trail (2026-10-01)

What the search did surface (context, not '647 proceedings):

• Celgene's PTAB exposure has been on unrelated patents: the REMS patents U.S. 6,045,501 and 6,315,720 (Coalition for Affordable Drugs VI LLC IPRs, instituted 2015-10-27, final written decisions 2016-10-26 holding those patents invalid on obviousness), and the ABRAXANE® patents U.S. 8,138,229 / 7,923,536 / 7,820,788 / 8,260 (Actavis IPR petitions filed 2017-04-04). Neither cluster involves the pomalidomide crystal-form patents.
• The pomalidomide generic fight was fought exclusively in D.N.J. (case nos. 2:19-cv-05797, 05799, 05802, 05804, 05806, 08758, 09737, 19-15343, 2:21-cv-02111), resolved by consent judgments and settlements rather than PTAB outcomes. Example: the Dr. Reddy's consent judgment in Civ. No. 19-15343 / 21-2111 (signed 2022-01-28) enjoins DRL's ANDA No. 213234 on the patents-in-suit (which list includes 10,093,647) with all claims, counterclaims, affirmative defenses, and demands dismissed with prejudice — but that is a private settlement disposition, not an invalidity holding. (consent judgment)
• Federal Circuit 21-1154 (Celgene Corp. v. Mylan Pharms., Inc., 2021 WL 5143311, Fed. Cir. Nov. 5, 2021, Prost/Chen/Hughes) is the only appellate ruling I can tie to the '647. It is a venue and failure-to-state-a-claim decision arising from the '647/'648/'649 infringement action: the court held that under § 271(e)(2) the ANDA submission itself is the infringing act (receipt of the paragraph IV notice letter in New Jersey is not), that Mylan had no "regular and established place of business" in New Jersey, and that Mylan N.V. was improperly named. Affirmed. It teaches you nothing about the '647's validity, but it is a useful venue roadmap. (CourtListener docket 21-1154)
• 21-2335 is the pendant Federal Circuit appeal (venue/coordination issues in the related Mylan action); like 21-1154 it is not an IPR appeal.

Caveat: ODP ingest can lag live filings. If you need a belt-and-suspenders confirmation, run a direct docket search on PTAB E2E / the USPTO Patent Trial and Appeal Board endpoint for the patent number (PTAB E2E, PTAB Decisions). I flag this as a residual uncertainty, not a finding of hidden proceedings.

Strategic summary

Claim status: everything is UNTESTED. All 12 claims of 10,093,647 stand as issued. Claim 1 (the sole independent claim, the dihydrate defined by XRPD peaks at 13.9, 16.6, 25.5 °2θ) has never been canceled, narrowed, or construed by the PTAB. Claims 2–12 are unadjudicated dependents. Zero proceedings means zero claim-level outcome data — do not let anyone tell you a claim is "dead."

Estoppel landscape: clean, but watch the clock. Because no IPR was ever instituted, there is no § 315(e)(2) estoppel binding anyone. That cuts both ways:

• No estoppel against you — you can raise any ground in district court that a petitioner could have raised.
• But the § 315(b) one-year bar likely forecloses a fresh IPR. Section 315(b) bars an IPR petition filed more than one year after the petitioner (or real party in interest/privy) is served with a complaint alleging infringement of the patent. The '647 was asserted against Mylan, Hetero, Natco/Breckenridge, Apotex (2019-02-14), Teva (2019-03-19), and Synthon/Alvogen (2019-04-12). Any of those parties is long time-barred. Only a new party that has never been served — e.g., a fresh ANDA filer or a current commercial actor receiving a first assertion now — could file an IPR, and it would have to do so within one year of service.
• The best prior art is not IPR-eligible anyway. The core invalidity theory articulated in the antitrust complaints is a catch-22: Pomalyst and its API (the dihydrate form) were marketed and publicly used since 2013, and the generics' earlier-in-time ANDA products allegedly already contained the claimed form — meaning if the claim is infringed, it is anticipated/obvious over the prior public use and sale. Public use and on-sale art cannot be raised in an IPR (35 U.S.C. § 311(b) limits IPRs to "prior art consisting of patents or printed publications"). Likewise the enablement/written-description/indefiniteness attacks (§ 112) are historically off-limits in IPR. So the theory that the generics themselves characterized as their strongest is structurally unavailable at the PTAB and must go to an Article III court under § 282. Secondary reference points for an IPR-only case would be printed publications: WO 2013/126326 (Celgene's own prior amorphous/anhydrous disclosure), the POMALYST® label (rev. June 2016), and the XRPD/TGA literature cited on the '647 face.
• PGR is time-barred (9-month post-issuance window from 2018-10-09 closed in 2019) and CBM is unavailable — this is a crystal-form chemistry patent, not a "financial product or service" patent, and the CBM program's sunset for new petitions has passed.

Pattern signals: No petitioner has ever filed against this patent, so no serial-petitioner pattern exists. The patent owner (Celgene → Bristol Myers Squibb) has not had to defend a single PTAB challenge on the '647, but it has litigated this patent heavily in D.N.J. and has run the appeal gauntlet twice (21-1154, 21-2335) on venue issues. No defensive aggregator (e.g., Unified Patents) appears anywhere in the chain — consistent with a small-molecule ANDA fight run by would-be generic competitors, not a PAE target.

Family durability warning: even a successful IPR canceling claim 1 would not clear the field. The '647 is the first of a continuation chain — 10,494,361 B2 (issued 2019-12-03), 10,781,199 B2 (issued 2020-04-30), and 11,518,753 B2 (issued 2022-11-29, same specification, same FIG. 1/2 XRPD fingerprints, family expiration 2037-12-20) — plus the sibling hemihydrate ('648) and monohydrate ('649) patents, each with its own independent claim. Treat the assertion risk as family-wide.

Recommended next steps

1. If you are a defendant being asserted today: do not assume a PTAB win is available. (a) If you or a privy were served in the 2019 D.N.J. wave, you are time-barred under § 315(b). (b) Even if a fresh IPR were available, the winning theories (public use/sale since 2013, § 112 enablement/indefiniteness due to the TGA-over-water-content discrepancy: dihydrate theoretical max 11.64% vs. Celgene's own reported 10.1–10.5% values) are not IPR-eligible. Build the case in district court.

2. Pull the record from the settled D.N.J. actions. Although settled (with prejudice as to those parties), the pleadings are public and contain the generics' developed invalidity contentions against the crystal-form patents — see the June 12, 2020 amended invalidity contentions excerpted in the antitrust complaints, and the June 16, 2020 Markman order in Celgene v. Hetero, No. 2:17-cv-03387, 2020 WL 3249117 (D.N.J.). These filings are the best free roadmap to the '647 attack.

3. Mind the estoppel asymmetry. Because no IPR was instituted, there is no § 315(e)(2) bar on anyone; but any petitioner you do engage on the family will be estopped as to grounds raised or reasonably raisable. There is no prior petitioner to piggyback on.

4. Monitor, don't wait. With zero PTAB activity and the original ANDA defendants all settled/time-barred, the realistic future trigger for a '647 IPR is a new entrant served with a complaint less than one year ago. If you are that entrant, preserve § 315(b) dates now.

5. Use the 21-1154 venue holding offensively if you are sued in New Jersey. Celgene Corp. v. Mylan Pharms., Inc., 2021 WL 5143311 (Fed. Cir. 2021), aff'g dismissal, holds that the § 271(e)(2) infringing act is the ANDA submission, not the notice letter sent to the branded company's home state — a motion to dismiss for improper venue may be viable for non-N.J. ANDA filers. (CourtListener 21-1154, summary)

Final caution: the statements above about fraud, sham litigation, and TGA discrepancies are allegations from third-party antitrust complaints (the CenterWell, Cigna, and Hagens Berman/Louisiana Health Service complaints), not adjudicated findings. I found no PTAB proceeding, FWD, or IPR appeal touching 10,093,647 — only the unrelated Celgene PTAB history noted above.

I'll research the assignment record for this patent using the available sources.

Let me dig into the assignment record specifics and the inventor's affiliation.

Let me pin down the specific reel/frame and check for any post-issuance transfers.

Ownership & Assignment Chain — US 10,093,647 B1

Sourcing note: I could not directly query USPTO Assignment Center — the live search endpoint (assignmentcenter.uspto.gov) is not reachable as a structured data source in this session, and I did not retrieve a copy of the recorded assignment document itself. The reconstruction below is grounded in (a) the authoritative patent text supplied, (b) the Google Patents "Legal Events" block mirroring USPTO assignment data (which shows a single inventor→Celgene recordation), and (c) the Cigna/CenterWell and HBSS antitrust complaints, which describe the ownership facts in narrative form and cite the assignment. I do not have the reel/frame numbers or the correspondent of record — see the explicit gap flagged in the timeline. Nothing below is inferred from entity naming alone.

Inventors

Pattern note (present, but not a fire-sale tell): The Cigna complaint (n.148) states plainly: "The named inventor is Jerry Atwood, who subsequently assigned the patents to Celgene. To avoid confusion, the applicant for the crystal form patents is referred to here simply as 'Celgene.'" This is the classic captured-external-expert pattern for polymorph patents: an academic crystal-engineering specialist is the named inventor of record while the operating company is the applicant ((71) Celgene Corporation, Summit, NJ). There is therefore no "inventor departed the assignee within 12 months" signal to assess — Atwood was never inside Celgene. The inventor's mailing address on the sibling filings (Columbia, MO) is consistent with an independent/consulting relationship rather than employment.

Original assignee

• Entity: Celgene Corporation, 86 Morris Avenue, Summit, New Jersey 07901 (Delaware corporation).
• Primary line of business: Branded pharmaceutical manufacturer. It shipped a product embodying the claims — pomalidomide is marketed as POMALYST®, FDA-approved (2013) with dexamethasone for relapsed/refractory multiple myeloma, and later for Kaposi sarcoma and other indications. The crystalline dihydrate is a solid form of that active ingredient.
• Current status: Operating, but no longer independent. In November 2019 Celgene was acquired by Bristol Myers Squibb and became a wholly-owned subsidiary (Cigna complaint ¶9: "In 2019, Celgene Corporation was acquired by, and became a wholly owned subsidiary of, Bristol Myers."). Celgene continues to hold and assert its patent estate; BMS sells POMALYST® today. Celgene is not dissolved and is not in bankruptcy.

Assignment timeline

Recorded assignments: one (1)

2017-??-?? (executed, date not disclosed in sources retrieved) / recorded 2018-01-29
    Reel/Frame: NOT RETRIEVED — cannot be stated without fabrication
    Conveyance: ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS)
    Assignor:    ATWOOD, JERRY LEE  (sole inventor)
    Assignee:    CELGENE CORPORATION, 86 Morris Avenue, Summit, NJ 07901
    Correspondent: NOT RETRIEVED — see gap note below
    Context: Original inventor-to-manufacturer assignment at
             prosecution. Executed during pendency of application
             15/849,285 (filed 2017-12-20); issue followed 2018-10-09.

Gap note (explicit, not inferred): Google Patents' legal-events pane for US 10,093,647 shows exactly one "Assigned to CELGENE CORPORATION … Assignors: ATWOOD, JERRY LEE" entry with the 2018-01-29 recordation date. It does not surface reel/frame or the correspondent (attorney of record who filed the recording). Without the reel/frame I cannot name the correspondent for this patent, and I will not guess. For comparison, Celgene's assignee-side correspondent of record on other patents has historically been its outside prosecution firm — e.g. the '720 patent assignment recorded 2001-01-05 at Reel 11469, Frame 882, and the '717 patent assignment recorded 2011-03-24 at Reel 026014, Frame 0756. No correspondent on this patent family has been confirmed by me, so I make no recurrence claim.

No post-issuance transfers recorded

I found no recorded assignment of US 10,093,647 after the 2018-01-29 inventor→Celgene recording. In particular:

• No assignment to any LLC, NPE, or special-purpose entity.
• No recorded assignment tied to the Bristol Myers Squibb acquisition of Celgene (2019). This is expected: the transaction was a stock/merger acquisition, so the patent title simply stayed with Celgene Corporation as a surviving/wholly-owned subsidiary; a separate patent-by-patent recordation is not required and I did not find one for this patent.
• The later family members — US 10,494,361 (continuation, appl. 16/124,066), US 10,781,199 (appl. 16/670,358), and US 11,518,753 (appl. 17/016,095) — are all listed as Celgene-family filings with no assignee changes surfaced.

If Assignment Center returns additional records I could not see (e.g. a merger/name-change recordation for the BMS transaction), the practical ownership outcome is unchanged: the chain terminates at Celgene/BMS.

Timeline diagram

timeline
    title Ownership of US 10093647
    2017 : Provisional filed 26 May
         : Application filed 20 Dec
    2018 : Inventor assigns to Celgene recorded 29 Jan
         : Patent issued 9 Oct
    2019 : Celgene acquired by Bristol Myers Squibb
         : Celgene remains patent owner
    2020 : Continuation still under Celgene
         : No NPE or LLC transfer recorded

NPE / troll-pattern signals

Verdict

Operating-company assertion

The chain consists of a single inventor→manufacturer assignment — Jerry Lee Atwood to Celgene Corporation, recorded 2018-01-29 (per the Google Patents legal-events record mirroring USPTO; reel/frame not retrieved). The assignee is a branded pharmaceutical manufacturer that ships POMALYST® embodying the claimed crystalline dihydrate and has asserted the patent itself against generic ANDA competitors (D.N.J. 2:19-cv-05797 et al.; Fed. Cir. 21-1154, 21-2335), so the assertion is operating-company vs. competitor, not NPE vs. operating company. No shell entity, known asserter, cascading transfer, privateering, or defensive-aggregator link appears anywhere in the record. Separately, third-party antitrust complaints (Cigna/CenterWell; HBSS) allege this crystal-form patent was procured and asserted to extend Celgene's monopoly and raise Walker Process / sham-litigation theories — a distinct incumbent-patent-abuse allegation against an operating company, which is not an NPE/troll-pattern finding.

Verify at: USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search patent 10,093,647; or https://assignment.uspto.gov/patent/index.html). Confirm the single 2018-01-29 Celgene recordation and check for any post-2019 merger recordation; if Assignment Center shows only the inventor→Celgene link, the original assignee still owns the patent.

Lower-confidence items: (i) exact execution date of the Atwood assignment (only the 2018-01-29 recordation date is visible in my sources); (ii) reel/frame numbers; (iii) correspondent of record; (iv) whether any name-change/merger recordation for the BMS acquisition of Celgene exists for this patent. I did not fabricate any of these.

I'll research this patent's prior art. Let me start with several targeted searches on the USPTO database and the key cited references.

Let me gather details on the key cited references that bear on patentability.

Prior-Art Analysis — U.S. Patent No. 10,093,647 B1

Sourcing note (strict-ID compliance): I searched for and confirmed the literal identifier 10093647 → U.S. Patent No. 10,093,647 B1, "Crystalline 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione dihydrate, compositions and methods of use thereof," Jerry Lee Atwood, Celgene Corporation, filed 2017-12-20, issued 2018-10-09, priority U.S. Provisional 62/511,878 (2017-05-26). I did not substitute any similar number. Because I could not directly query USPTO PatentCenter in this session, the reference list below is taken from the authoritative full patent text supplied for this patent (the "Citations (45)" and "Patent Citations (82)" tables), which mirror the USPTO/Google Patents record.

Framework for the § 102 analysis. All 12 claims are drawn to a crystalline solid form — the pomalidomide dihydrate — characterized by stated XRPD peaks (claims 1–4), DSC endotherms (5–8), water content (9), TGA weight loss (10–11), and IR (12). For a reference to anticipate under 35 U.S.C. § 102 it must disclose, in a single reference, every element of the claim — i.e., the same crystalline dihydrate form with the same identifying characteristics (or a form that inherently possesses them). A reference that discloses only the compound pomalidomide, a synthetic process, or a formulation, without the crystalline dihydrate, cannot anticipate claims 1–12, however relevant it may be under § 103. I flag each reference accordingly.

1. Most relevant prior art (the references that actually bear on the dihydrate claims)

2. Analogous polymorph art (methodology references; examiner-cited)

3. Synthesis / compound references cited as background (no § 102 effect on the dihydrate claims)

These disclose only the compound or routes to it; each fails at least the "crystalline dihydrate" element of claim 1 and therefore does not anticipate any of claims 1–12:

• US 6,335,349 B1 — Celgene, substituted 2-(2,6-dioxopiperidin-3-yl)isoindolines (filed 1996-07-24; issued 2002-01-01).
• US 6,316,471 B1 — Celgene, isoindolines/methods of use (issued 2001-11-13).
• US 6,476,052 B1 — Celgene, isoindolines, method of use and pharmaceutical compositions (issued 2002-11-05).
• US 7,041,680 B2 — Celgene, (R) and (S) isomers of substituted 2-(2,6-dioxopiperidin-3-yl)phthalimides and 1-oxoisoindolines (issued 2006-05-09).
• US 7,709,502 B2 — Celgene, substituted 2-(2,6-dioxopiperidin-3-yl)-phthalimides and 1-oxoisoindolines (issued 2010-05-04).
• US 4,551,177 A — National Starch, "Compressible starches as binders for tablets or capsules" (filed 1984-04-23; issued 1985-11-05) — tablet-excipient art; irrelevant under § 102 to a crystalline-form claim (it appears only in the general-formulation discussion).
• US 11,053,211 B2 (Hetero Research Foundation), "Process for pomalidomide" — appears as a "Cited By" (third-party) document, not prior art on the face of the '647 patent; it post-dates the '647 filing and is a process reference.

4. The most serious § 102 attack is not a printed reference — it is the POMALYST® product/API public use and sale

The generic ANDA defendants articulated the strongest anticipation theory, which I note because it is the practically decisive § 102 issue:

"Celgene's Pomalyst product and the API Celgene uses to prepare its Pomalyst product have been marketed by Celgene since 2013 — four years prior to the earliest priority date of the hydrate-patents-in-suit — and are thus prior art to the hydrate-patents-in-suit." (Celgene v. Aurobindo, 19-cv-05799, Revised Joint Proposed Discovery Plan, ECF No. 37, at 15.)

Under § 102(a)(1) (public use / on sale), if the marketed pomalidomide product or the API inherently contained the claimed crystalline dihydrate, the claims would be anticipated regardless of any printed reference. This "inherent anticipation" theory, together with the arguments that (i) the three sibling applications' XRPD patterns (the '647 dihydrate, '648 hemihydrate, '649 monohydrate) "substantially overlap" and (ii) the TGA water-loss data are internally inconsistent with the claimed stoichiometries, forms the core of the third-party invalidity case (see The Cigna Group v. Celgene Corp., D.N.J. No. 1:25-cv-05237-ER, ¶¶ 320–335). These are allegations, not adjudicated findings.

⚠️ Note: the '647 patent is reported as not Orange Book–listed, so suits on it did not trigger a 30-month FDA stay.

5. Non-substantive citations on the '647 record

A substantial fraction of the 45/82 "Citations" are boilerplate medical-informatics / REMS references unrelated to solid-form chemistry and have no § 102 bearing on claims 1–12. Examples: US 5,594,637 (medical risk assessment), WO 96/13790 (prescription management system), US 5,619,991 (delivery of medical services via electronic data), WO 98/13783 (electronic medical records), US 6,133,090 (controlled access to portable recording media), US 6,202,923 (automated pharmacy), US 5,832,449 / 6,055,507 (pharmaceutical trial product management), US 6,063,026 (medical diagnostic analysis), and the Celgene REMS family (US 6,045,501; 6,561,976; 6,908,432; 8,155,853; 8,204,763; 8,589,188; 8,626,531; 7,959,566). I list them as a group because individually none discloses a crystalline pomalidomide dihydrate or any solid-form characteristic.

Bottom line

1. No cited printed reference anticipates claims 1–12. Every one of the 12 claims requires the crystalline pomalidomide dihydrate with specified XRPD/DSC/TGA/IR characteristics, and none of the cited references — compound, process, formulation, or analogous-compound art — discloses that specific crystalline form.
2. WO 2013/126326 A1 (Celgene, 2013-08-29) is the most relevant reference — the closest prior art and the leading § 103 (obviousness) reference, disclosing pomalidomide's amorphous and crystalline anhydrous forms (and only a generic recitation of "dihydrate").
3. US 5,635,517 (the pomalidomide compound) and the synthesis patents (US 6,335,349; 6,316,471; 6,476,052; 7,041,680; 7,709,502; 7,994,327) supply the molecule and methods but disclose no solid form.
4. US 7,465,800 / WO 2013/012485 / US 6,627,646 are analogous polymorph art usable for a § 103 obviousness case, not for anticipation.
5. The strongest § 102(a)(1) theory is inherent anticipation by the POMALYST® product/API (on sale/publicly used since 2013), as pressed by the generic ANDA filers and recited in the antitrust complaints.

Confidence: High on the identity of the references and their dates as listed in the authoritative patent text and mirrored on Google Patents. Medium on whether any embodiment of the WO 2013/126326 family actually depicts the '647 dihydrate XRPD (I could not retrieve the full figure-level data). The invalidity theories above are third-party allegations, not adjudicated holdings. If you want docket-level or figure-level verification (e.g., a side-by-side of the WO '326 XRPD peaks against the '647's 13.9/16.6/25.5 °2θ), I can pull that next.

I'll ground this analysis in the prior art cited on the patent's own page and verify key reference details where I can.

Obviousness Analysis — U.S. Patent No. 10,093,647 B1 ("the '647 patent")

Scope note. This analysis is built from the prior-art references listed in the "Citations"/prior-art section of the '647 patent page as supplied, cross-checked against the publicly filed papers retrieved by search. I did not recover the complete D.N.J. invalidity contentions (e.g., the December 17, 2017 statement of invalidity contentions, or the June 12, 2020 proposed amended contentions), so the combinations below are constructed from the cited references themselves plus the methodological teachings they contain — not from the actual service copies. Statements about what particular litigants argued are attributed to the pleadings, not presented as adjudicated findings. Identifier rule honored: "US10093647" = U.S. Patent No. 10,093,647 B1; "the '467 patent" (U.S. 9,993,467) is a different patent and is not at issue here.

1. Legal framework

A § 103 challenge proceeds under the Graham v. John Deere factors: scope/content of the prior art, differences between the prior art and the claims, the level of ordinary skill, and secondary considerations. KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), permits a combination where the improvement was "obvious to try" — i.e., where there are "a finite number of identified, predictable solutions."

Three doctrines drive the analysis of a polymorph/hydrate product-by-property claim like claim 1:

1. In re Papesch, 315 F.2d 381 (CCPA 1963) — a compound and all of its properties are inseparable; discovering a new property of a known compound does not make the compound patentable. Applied here, the XRPD/DSC/TGA/IR "fingerprints" are properties of the same chemical entity (pomalidomide dihydrate), not an independent inventive contribution.
2. In re Best, 562 F.2d 1252 (CCPA 1977) — where the claimed product appears identical to a prior-art product, the applicant bears the burden to show a difference. If the prior-art aqueous-crystallization route inherently produced this crystalline form, the claim is anticipated; if it merely made the form obvious to obtain, § 103 applies.
3. Rationale for hydrates/solvates. The Board and courts have repeatedly treated crystalline hydrate/solvate forms of known APIs as obvious where the only difference from the known form is water/solvent of crystallization and no unexpected property is shown. (Cf. In re Cruciferous Sprout Litig. on obvious physical forms generally.)

2. Scope and content of the prior art (from the '647 citation list)

Two dozen additional citations on the page are directed to safety/REMS, formulation, and method-of-use matters and are not material to solid-form obviousness.

3. Level of ordinary skill

A POSA here is a formulation/process chemist or solid-state pharmaceutical scientist with an advanced degree and several years' experience in polymorphism, hydrate/solvate screening, and crystallization, routinely using XRPD, DSC, TGA, and Karl-Fischer titration. Selecting aqueous-organic solvent systems to obtain crystalline hydrates is squarely within ordinary skill.

4. The differences between the prior art and claim 1

Claim 1 is a product-by-property claim:

"Crystalline 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione dihydrate, having an X-ray powder diffraction pattern comprising peaks at 13.9, 16.6, and 25.5 degrees 2θ ± 0.2 degrees 2θ."

The only differences over the art are:

• (a) the hydration state — "dihydrate" (two waters of crystallization per molecule), versus the known anhydrous Form A and amorphous form; and
• (b) the characterization data — the three signature XRPD peaks (and, in the dependents, DSC maxima at ~308 °C/~309 °C, ~11.6% water, TGA 10.1–10.4%, and the FIG. 1–6 fingerprints).

Nothing in claim 1 recites a manufacturing step, a property advantage, or any structure other than the molecule plus waters of crystallization. As the Cigna complaint summarizes the family: "Each of Celgene's patents has a single independent claim, claiming a specific chemical composition and a specific crystalline form identified … by an x-ray powder diffraction pattern ('XRPD')." (Cigna Compl. ¶ 314, https://business.cch.com/ald/TheCignaGroupvCelgeneCorporationComplaint.pdf)

5. Combinations that render the claims obvious

Combination A — WO 2013/126326 alone (single-reference / "product of a known process")

WO 2013/126326 discloses pomalidomide, its crystalline Form A and amorphous form, and — critically — general methods of making solvates and hydrates of pomalidomide, including from water-bearing solvent systems. The '647 Example uses precisely such a system: Form A slurried in 4:1 1,4-dioxane/water, dissolved at 70 °C, evaporated, triturated in water. Additional working examples use 4:1 THF/water and 1:1:1 acetone/water/IPA.

Why obvious: A POSA following the primary reference's own aqueous-crystallization teachings, in order to make a form of pomalidomide incorporating water, would necessarily obtain a pomalidomide hydrate. Because the product of that process is a crystalline hydrate, and hydrates of a monoacidic/base API occur in only a finite, predictable set of stoichiometries (hemi-, mono-, di-), reaching the dihydrate is an "obvious to try" step with a reasonable expectation of success under KSR. Under In re Best, if the reference's aqueous route inherently yields the same crystalline form, the claim is anticipated outright; if not, it is at minimum an obvious product.

Combination B — WO 2013/126326 + US 7,465,800 (lenalidomide polymorphs) [± WO 2013/012485]

Motivation to combine: Lenalidomide is the immediately adjacent analog (identical isoindoline-1,3-dione + 2,6-dioxopiperidine pharmacophore; it differs by the fused-ring amino/oxo pattern and is the compound from which the pomalidomide program emerged). US 7,465,800 expressly claims/describes polymorphic forms of lenalidomide, and WO 2013/012485 discloses further crystalline forms of lenalidomide. A POSA optimizing pomalidomide solid form would look to what was done for the closest marketed analog and expect the same behavior — that the compound can be crystallized in multiple forms, including hydrated forms accessible from aqueous solvent.

Why obvious: The art supplies both (i) the compound to be screened (WO 2013/126326) and (ii) the express reason and template to hydrate/polymorph-screen the analog class (US 7,465,800; WO 2013/012485). Combined, they teach the claimed dihydrate as the predictable product of routine aqueous crystallization, with no unexpected property needed or shown.

Combination C — WO 2013/126326 + WO 2014/160690 / US 9,695,146 (pomalidomide coformer forms)

Motivation to combine: The coformer reference (same assignee, same molecule) discloses pomalidomide solids in which water is incorporated into the crystal lattice ("a hydrated cocrystal"), and expressly discusses the compound's capacity to form H-bonded/hydrated lattices and lists solvent systems including water and aqueous organics.

Why obvious: A reference that already places water in the pomalidomide lattice and describes aqueous crystallization supplies a direct lead to the binary pomalidomide–water system. Obtaining the specific dihydrate from that lead would be a matter of routine solvent/water-ratio and temperature optimization.

Combination D — WO 2013/126326 + US 6,627,646 (norastemizole polymorphs)

Motivation to combine: US 6,627,646 is a general, enabling teaching that APIs commonly exist as polymorphs/solvates obtainable by choosing crystallization conditions and that discovering such forms is standard practice.

Why obvious: It provides the "why would I bother screening?" rationale absent from the pomalidomide-specific art — establishing that, as of the priority date, hydrate/polymorph screening of a known API was a conventional, routine exercise.

Combination E — the commercial product as prior art ("publicly available" API)

The '647 priority date is 2017-05-26, four years after POMALYST® (and its API) entered public use/on sale in 2013. The generic ANDA defendants argued exactly this: "Celgene's Pomalyst product and the API Celgene uses to prepare its Pomalyst product have been marketed by Celgene since 2013—four years prior to the earliest priority date of the hydrate-patents-in-suit—and are thus prior art" (quoted in Cigna Compl. ¶ 326, citing Celgene v. Aurobindo, 19-cv-05799, ECF No. 37 at 15). If the marketed API was itself a pomalidomide hydrate, the '647 form would be the same product (anticipation/§ 102) or an obvious variant of it (§ 103).

6. Claim chart — claim 1 and why each limitation is met or inherent

Notably, the dependent claims do not recite any property that distinguishes this form over the anhydrous Form A, the amorphous form, or the co-crystal forms — there is no comparative dissolution, stability, bioavailability, or hygroscopicity limitation anywhere in the claims.

7. Motivation and reasonable expectation of success (why the combination is not hindsight)

1. Same molecule, same assignee. The primary and coformer references are Celgene's own pomalidomide solid-form applications — the POSA working on this exact molecule would have them in hand.
2. Known aqueous processing. Both WO 2013/126326 and the '647 specification itself list the standard toolkit (slurry recrystallization, solvent recrystallization, antisolvent addition, exposure to water, vacuum drying) and standard water/organic systems.
3. Finite, predictable candidate set. For a small-molecule API, hydrate stoichiometries and solvent/water ratios are a bounded experimental space, satisfying KSR's "finite number of identified, predictable solutions."
4. Reason to try. Crystalline hydrates are sought for manufacture/tableting/flowability/stability benefits — precisely the "physical properties" the '647 specification itself recites as the rationale for the invention.
5. Reasonable expectation of success. There is no unpredictability argument tied to claim 1: the patent discloses no data showing the dihydrate possesses unexpected properties relative to the known anhydrous/amorphous forms.

Secondary considerations. The '647 specification contains no comparative data, no unexpected-results showing, and no commercial-success evidence with a nexus to this crystalline form (the commercial success, if any, belongs to POMALYST® generally, whose protecting patents are the method-of-use and formulation patents). Absent nexus, secondary considerations do not rebut the prima facie case.

8. Where the obviousness case is weaker (honest counterweight)

• No single reference discloses the dihydrate form per se. The primary reference discloses the anhydrous Form A and amorphous form, so the case depends on combining generic hydrate-forming teachings rather than on an express disclosure of the claimed XRPD.
• Hydrate stoichiometry is not perfectly predictable. The '647 specification's own Table 2 lists dozens of failed experiments that did not yield the dihydrate — a patentee could argue this shows unpredictability and cuts against obviousness. The counter is that KSR / In re Kubin make "obvious to try" sufficient in an otherwise routine arts-space, and the specification's own working examples (dioxane/water, THF/water, acetone/water/IPA) show the result was readily reached.
• The claims survived prosecution and the PTO issued them without a § 103 rejection on these references — a fact entitled to some, but not controlling, weight.
• Discrepant TGA data. Third parties allege the TGA figures (10.1–10.4% loss vs. the theoretical ~11.6% water content of a true dihydrate) actually undermine the claims' internal consistency (Cigna Compl. ¶¶ 328–331). That is a § 112 (enablement/written-description/indefiniteness) argument more than a § 103 argument, but it bears on whether the claimed form is even adequately described.

9. Conclusion

On the cited record, the strongest § 103 attack is:

WO 2013/126326 (pomalidomide solid forms; aqueous crystallization) in view of US 7,465,800 and/or WO 2013/012485 (analog lenalidomide polymorphs), optionally further in view of WO 2014/160690 / US 9,695,146 (pomalidomide hydrated/co-crystal forms) and US 6,627,646 (polymorph screening as routine).

A POSA, seeking a manufacturable crystalline form of the known pomalidomide molecule, would have been motivated to crystallize it from water-bearing solvent systems, would have expected one or more hydrates to result, and would have arrived at a crystalline dihydrate with a reasonable expectation of success. Because claim 1 recites only the known molecule plus waters of crystallization and the inherent analytical fingerprint of that form — and because In re Papesch and In re Best treat such properties as inseparable from the composition — claims 1–12 are, on this record, likely unpatentable as obvious. Claim 1 is the only independent claim, so a finding against it would take the whole patent with it.

Caveats. I could not obtain the actual invalidity contentions or any final merits ruling (the D.N.J. cases ended in venue dismissal, consolidation, and the 2020–2021 settlement wave, not a judgment of invalidity). No IPR specific to the '647 patent was located. This is a technical § 103 assessment on the cited references, not a legal opinion or an adjudicated result.

Key sources: '647 full text (https://patents.google.com/patent/US10093647/en); WO 2013/126326 (https://patentimages.storage.googleapis.com/7c/a0/dc/1a7f4460db9550/WO2013126326A1.pdf); WO 2014/160690 / US 9,695,146 (http://patentimages.storage.googleapis.com/eb/f7/2f/c0c639b349057c/US9695146.pdf); Celgene v. Mylan, Nos. 2021-1154, 2021 WL 5143311 (Fed. Cir. Nov. 5, 2021) (https://www.robinskaplan.com/newsroom/insights/celgene-v-mylan-pharms); The Cigna Group v. Celgene, D.N.J. 1:25-cv-05237-ER Compl. (https://business.cch.com/ald/TheCignaGroupvCelgeneCorporationComplaint.pdf).