US 10093649

I'll search for current information on this patent number, including any 2026 CAFC docket activity.

Let me try more targeted searches for this specific patent and its litigation history.

Let me do a final targeted check for any 2026 CAFC activity specifically on this patent.

US Patent 10,093,649 — Summary

I searched for this specific number and related identifiers. Below is what I can confirm from authoritative sources (the patent document itself, USPTO/Google Patents bibliographic data, and court filings), with uncertainty flagged where applicable.

Bibliographic Data

Abstract

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

Plain-Language Overview of the Claims

The patent has a single independent claim:

• Claim 1 — A crystalline monohydrate form of pomalidomide (the compound whose chemical name is in the title), defined by its X-ray powder diffraction (XRPD) "fingerprint": peaks at 11.8, 17.1, and 24.2 degrees two-theta (2θ) ±0.2°. In plain terms, it claims the solid crystalline form of pomalidomide that contains roughly one molecule of water per molecule of drug, identified by three signature XRPD peaks.

• Claims 2–9 are dependent claims adding further characterization of that same crystalline form:

  • Claim 2 — adds three more XRPD peaks (13.9, 16.5, 25.7 degrees 2θ).
  • Claim 3 — XRPD pattern corresponding to FIG. 1.
  • Claim 4 — DSC thermogram with an endotherm maximum at about 312 °C.
  • Claim 5 — DSC thermogram corresponding to FIG. 2.
  • Claim 6 — about 6.2% water by mass.
  • Claim 7 — TGA weight loss of about 4.9%–7.4% when heated from ~30 °C to ~225 °C.
  • Claim 8 — TGA thermogram corresponding to FIG. 3.
  • Claim 9 — IR spectrum corresponding to FIG. 4.

Note: The title and summary mention "compositions and methods of use," but the granted claims are directed only to the crystalline monohydrate solid form — no composition or method-of-treatment claims were issued.

Litigation Context

• This patent is one of three related "crystal form" patents by the same inventor (the '647 dihydrate, '648 hemihydrate, and '649 monohydrate), all filed December 20, 2017 and issued October 9, 2018.
• It was asserted in numerous ANDA suits in the District of New Jersey (e.g., 2:19-cv-05797, -05799, -05802, -05804, -05806, -08758, -09737; 2:21-cv-02111).
• Continuations in the same family include US 10,590,103 B2, US 10,919,873 B2, and US 12,065,423 B2.

On the CAFC 2026 Docket Search — Important Uncertainty

I was unable to confirm any 2026 Federal Circuit docket activity involving patent 10,093,649. The only CAFC matters I could verify connected to this patent family are:

• Celgene Corp. v. Mylan Pharms., Inc., No. 21-1154 (Fed. Cir. Nov. 5, 2021) — a venue/pleading appeal affirming dismissal, not a merits decision on the '649 patent.
• No. 21-2335 (Federal Circuit, 2021).

I also found ongoing antitrust litigation (e.g., class actions by Cigna and Blue Cross/Blue Shield filed 2023–2025 alleging sham litigation and reverse-payment settlements around the Pomalyst patents, including the '649), but these are district-court matters, not CAFC appeals.

Caveat: My search tools did not surface a CAFC 2026 docket entry specifically for 10,093,649 (or for the '649 shorthand) even though I searched for it. I cannot rule out a recent or non-public docket entry, but I have no authoritative source confirming one. Any statement that this patent is on the 2026 CAFC docket should be independently verified against the Federal Circuit's official docket (PACER/CAFC CM-ECF) rather than assumed.

I'll research litigation involving US patent 10093649 across patent litigation databases.

Let me search for the specific cases and the '649 patent litigation details.

Let me get details on the specific co-pending cases and appeals.

Litigation involving U.S. Patent No. 10,093,649 ("the '649 patent")

Important framing note: US 10,093,649 B1 (the "monohydrate" crystal-form patent) is one of three Celgene crystal-form (polymorph) patents issued October 9, 2018 — the '647 (dihydrate), '648 (hemihydrate), and '649 (monohydrate) patents. These three were asserted together as a family in the same Hatch‑Waxman ANDA litigations, so most dockets below list the '649 patent alongside the '647 and '648 patents. This answer lists the cases that Google Patents associates with the '649 patent and that I could corroborate in litigation databases. Where I could not verify a defendant, I say so rather than guess.

Cases linked to the '649 patent

The patent's Google Patents record ("Family has litigation") lists the following matters directly tied to this patent:

Confirmed contextual facts

• All four U.S. district cases filed Feb. 14, 2019 (‑05797, ‑05799, ‑05802, ‑05804, ‑05806) were filed in the District of New Jersey as § 271 ANDA infringement actions and were consolidated before Judge Esther Salas (M.J. Michael A. Hammer). A docket order (Oct. 28, 2020) references "2:19‑cv‑08758‑ES‑MAH; 2:19‑cv‑05797‑ES‑MAH; 2:19‑cv‑05799‑ES‑MAH; 2:19‑cv‑05802‑ES‑MAH…," confirming the consolidation.
• The '649 patent is expressly referenced in the Hetero complaint (2:19‑cv‑05797), attached as Exhibit C, and patents‑in‑suit included U.S. Patent Nos. 10,093,647, 10,093,648, and 10,093,649.
• Ultimate resolution: These crystal‑form suits were not tried to judgment. Consistent with Celgene/BMS reporting and the antitrust complaints (Cigna, CenterWell, Humana/insurer litigation in S.D.N.Y. and D.N.J.), the pomalidomide litigation was resolved through settlements in late 2020–2021 (e.g., the Celgene–Natco/Breckenridge agreement ~Nov. 2020, Celgene–Teva ~Mar. 2021, Celgene–Aurobindo ~spring 2021), with agreed generic entry delayed until early 2026.
• The '649 patent was also swept into antitrust follow‑on litigation (e.g., In re Pomalyst/Pomalidomide antitrust actions and the Cigna/CenterWell complaints) as one of the three "crystal form" patents alleged to be part of an overall scheme.

What I could not confirm

• The defendants in 2:19-cv-05802, 2:19-cv-05804, 2:19-cv-09737, and 2:21-cv-02111 and the appellant/appellee details in Fed. Cir. 21-2335. Based on the Jones Day description of the POMALYST Hatch‑Waxman program, the universe of generic defendants in this period included Teva, Apotex, Hetero, Aurobindo, Mylan, Breckenridge, Natco, and Alvogen/Synthon, so the unidentified dockets most likely correspond to one or more of those parties — but I will not assert that as fact without confirmation.
• Precise termination dates and dispositions for each individual docket (as opposed to the consolidated group) were not all retrievable.

Sources consulted

• Google Patents record for US 10,093,649 B1 (the patent's own litigation links) — https://patents.google.com/patent/[US10093649B1](/patent/US10093649B1)/en
• CourtListener / RECAP dockets (e.g., Celgene Corp. v. Apotex Inc., 2:19‑cv‑05806; Celgene Corp. v. Hetero Labs Ltd., 2:19‑cv‑05797; Celgene Corp. v. Mylan Pharms. Inc., Fed. Cir. 21‑1154)
• Docket Alarm (2:19‑cv‑05797 case details)
• DrugPatentWatch litigation records (2:19‑cv‑05797; 2:17‑cv‑06842)
• Unified Patents litigation portal (links listed on the patent record)
• Jones Day experience page, "Celgene asserts its POMALYST® patents against multiple generic companies under Hatch‑Waxman Act"
• Antitrust complaints (Cigna/The Cigna Group v. Celgene/BMS, S.D.N.Y.; CenterWell v. Celgene/BMS) referencing the '647/'648/'649 patents
• Bristol Myers Squibb 10‑Q / annual report disclosures re: Pomalyst antitrust litigation

Bottom line: Yes — U.S. Patent No. 10,093,649 has been litigated. It was asserted by Celgene in a series of consolidated Hatch‑Waxman ANDA suits in the District of New Jersey (Spring 2019 through 2021) against generic pomalidomide manufacturers, with two related Federal Circuit appeals (21‑1154 and 21‑2335). Those disputes were resolved by settlement rather than judgment, and the patent was later implicated in follow‑on antitrust litigation. I could not independently confirm the defendants for four of the listed dockets, and I have flagged those rather than guessing.

I'll verify the "no proceedings on file" default against public sources before writing this up.

Let me check the patent-specific PTAB databases and any IPR filings around the crystalline-form patents.

Proceedings overview

Total AIA trial proceedings on US 10,093,649: zero. The USPTO Open Data Portal (ODP) structured block returned no IPR, PGR, or CBM proceedings on this patent as of the most recent ingest, and my independent web checks (DrugPatentWatch's drug-level PTAB table for pomalidomide, last updated 2026-03-25; PTAB/Unified Patents dockets; and the litigation record) surface no petition, institution decision, FWD, or appeal naming this patent — so the breakdown is 0 active / 0 invalidated / 0 sustained / 0 settled / 0 institution-denied. Defensive posture: the patent is entirely un-hardened but also entirely un-tested at the PTAB — claim 1 has never survived (or been subjected to) an AIA challenge, so a defendant gets no free "claims are already dead" argument, but also faces no adverse PTAB precedent, no PTO claim constructions, and no § 315(e)(2) estoppel wall limiting what it can argue.

I want to flag one apparent contradiction with the record rather than paper over it: the patent has been heavily asserted since 2019 (seven-plus D.N.J. ANDA cases plus a 2021 CAFC appeal), which is exactly the profile of a patent that normally attracts IPRs. Its absence from the PTAB is a real signal, and I explain the likely reason below. I found no evidence of a defensive aggregator (Unified Patents, RPX, etc.) ever filing on this patent or its '647/'648 siblings.

Per-proceeding detail

There are no proceedings to detail. Per the operating rule "Do not invent proceeding numbers," I am not creating placeholder entries. For transparency, here is what I checked and what I found instead:

What actually exists in the pomalidomide PTAB space (NOT this patent)

• IPR2015-01092, IPR2015-01096, IPR2015-01102, IPR2015-01103 — Coalition for Affordable Drugs VI LLC v. Celgene Corp. These are the only pomalidomide-related AIA trials I can locate. They target US 5,635,517 and US 6,045,501 — the compound and distribution patents, not the crystalline-form patents. All four were terminated 2016-10-26 (Paper Nos. 73/73/75/76), consistent with the Coalition's standard practice of settling/abandoning after institution. None of these proceedings addressed US 10,093,649, its claims, or its XRPD-based claim format, and none produced claim-level estoppel touching this patent. Sources: DrugPatentWatch pomalidomide PTAB table; the IPR papers are cited at D.N.J. 2:22-cv-06440, Doc. 60 at 47.
• No IPR/PGR/CBM exists for US 10,093,647 (dihydrate), US 10,093,648 (hemihydrate), or US 10,093,649 (monohydrate), nor for the family continuations US 10,590,103, US 10,919,873, or US 12,065,423. I verified this by searching patent numbers, the "crystal form patents" framing, and the Celgene/Atwood inventor combination.

Where this patent was actually challenged: the district court, not the PTAB

Because the client-facing question is "what happened," the material events are all Article III:

• D.N.J. ANDA litigation (2019–2021): the '649 was asserted in, e.g., Celgene v. Teva et al. (2:19-cv-08758), Celgene v. Aurobindo (2:19-cv-05799), 2:19-cv-05797, -05802, -05804, -05806, -09737, and 2:21-cv-02111. All resolved by settlement before trial; there is no merits validity judgment on the '649.
• Invalidity theories developed but never PTAB-tested. The generics' contentions — as recited in the antitrust complaints now pending — attacked all three crystal-form patents on (a) public use / on-sale / § 102(b) prior art (Pomalyst and its API were marketed from 2013, years before the 2017 priority date); (b) substantial XRPD peak overlap among the '647/'648/'649 patterns; (c) § 112 problems because claimed TGA weight losses (e.g., 7.4% for the monohydrate) exceed the theoretical maximum for a monohydrate (6.18–6.19% water); and (d) indefiniteness / written description / enablement. Sources: Cigna v. Celgene complaint, ¶¶ 325–333; BCBS/antitrust complaint. None of this was ever filed at the PTAB.
• 2020-10-12 — withdrawal: the parties filed an amended joint claim construction and prehearing statement withdrawing Celgene's infringement claims as to many of the claims of the crystalline-form patents, followed within a month by settlements. Source.
• CAFC appeals (2021): Nos. 21-1154 (Celgene v. Mylan, venue/pleading) and 21-2335. These are appeals from the D.N.J. infringement litigation — not Rule 1295 appeals from any PTAB FWD, because there is no FWD to appeal. (Consistent with the previously-generated bibliographic section; I reaffirm the earlier "no 2026 CAFC activity found" caveat.)

Strategic summary

Claim status. There is no PTAB-imposed claim status to report. All nine claims of US 10,093,649 — claim 1 (XRPD peaks at 11.8, 17.1, 24.2 °2θ ±0.2) and dependent claims 2–9 — are UNTESTED, not CANCELED and not SUSTAINED. Every claim that issued on 2018-10-09 is still in force, with an anticipated expiration of 2037-12-20. Contrast this with the lenalidomide Form A crystal patents ('219/'598/'357), which were narrowed/fenced in by the 2014 D.N.J. Markman ruling — the pomalidomide crystal patents never got a comparable construction because the case settled first.

Estoppel landscape. Because no IPR/PGR was filed by anyone, § 315(e)(2) estoppel is a non-issue — there is no petitioner, no privy chain, and no "raised or reasonably could have raised" bar constraining what any defendant may assert. There is also no § 325(e) estoppel and no statutory 1-year § 315(b) problem for a new petitioner who has not been served with a complaint more than one year ago. The entire prior-art universe remains open: the 2013-onward Pomalyst product/API public-use and on-sale art, the '647/'648/'649 cross-family XRPD overlap, the internal TGA inconsistency, and secondary references such as WO 2013/126326 (the anhydrous Form A disclosure Celgene itself admits) are all still available as IPR grounds. This is a rare, entirely clean slate.

Pattern signals — and the likely reason there is nothing on file. The same generics fought this patent for two years and then settled; no repeat petitioner, no aggregator, and no patent-owner appeal activity at the PTAB exist for the '649. The most plausible explanation is structural, not strategic weakness: the '649 was asserted in Hatch-Waxman ANDA suits, and the defendants were served with complaints in 2019–2021. Under 35 U.S.C. § 315(b), a petitioner is time-barred one year after service — so the natural IPR window for the actual ANDA defendants closed around 2020–2022, at precisely the moment they were settling instead. That is a timing artifact, not a validity finding. It also explains why the antitrust plaintiffs' complaints describe invalidity theories that were briefed in district court invalidity contentions but never adjudicated: Celgene's settlement strategy (alleged by plaintiffs to be deliberate "sham litigation" designed to avoid judicial scrutiny) removed the cases from decision before either the district court or the PTAB could rule.

Net for a defendant today: you face an untested but structurally vulnerable patent. The '649's best invalidity theories have never been run through an adversarial adjudication, so nothing binds you — but nothing has softened the patent either, and it lives until 2037. Also note the nuance that at least one crystal-form sibling (the '647) is described in the antitrust complaints as "Not OB Listed," meaning Orange Book listing is itself worth verifying for the '649 before assuming the patent can support a Hatch-Waxman 30-month stay against your ANDA.

Recommended next steps

1. Treat "no PTAB activity" as the finding, not a gap in research. There is no FWD to link to and no disposition to quote for this patent. If you need a citation for the negative result, the cleanest public reference is the DrugPatentWatch pomalidomide PTAB table (no '649 entry) plus the USPTO PTAB E2E / Patent Trial and Appeal Board Center search on patent number 10093649.
2. Run a fresh § 315(b) and real-party-in-interest check before filing. If your client was served with a complaint asserting the '649 more than one year ago, the IPR door may be closed; if you have not been served, or are a new/independent party (e.g., a different ANDA filer, a privy-free manufacturer, or a defensive aggregator), you are not barred and can file on the full ground set with zero estoppel risk. Consider a PGR only for non-prior-art grounds — irrelevant here given the 2018 issue date, so IPR under §§ 102/103 is the only viable vehicle.
3. Build the petition around the strongest ground the district court never got to decide: the § 102(b) public-use / on-sale / printed-publication attack (Pomalyst and its API sold from 2013; the '649 priority is 2017-09-22), plus the § 112 TGA-impossibility theory (claimed ~7.4% weight loss for a nominal monohydrate whose theoretical water content is ~6.19%). The TGA argument is unusually strong IPR fare because it attacks the patent's own specification as internally inconsistent, and the XRPD-overlap argument lets you argue the three sibling forms are not distinct inventions.
4. Verify Orange Book listing and litigation status. Confirm whether the '649 is currently OB-listed for Pomalyst (the antitrust complaints state the '647 is not) and pull the current docket in the D.N.J. and S.D.N.Y. antitrust/patent matters before relying on any settlement scope. Settlements in ANDA cases are frequently confidential and may include licenses that affect your freedom to operate even where the patent is unadjudicated — say so, and don't assume a settlement equals a validity concession.
5. If you represent the patent owner (BMS/Celgene), the flip side is identical: an untested patent with a hostile literature and an active antitrust narrative is exposed to a first-ever IPR at any time until 2037. Consider whether a defensive reissue/reexam or a validity-focused expert record is warranted before any new assertion campaign — the absence of any PTAB history means there is no favorable IPR precedent to point to.

Confidence and caveats: I am highly confident there is no AIA trial proceeding on US 10,093,649. I could not independently re-verify the ODP ingest beyond the structured block provided in the prompt, and my web searches were subject to the tool-step limit; I did not find any petitioner, judge panel, institution decision, FWD, or PTAB appeal to report, and I have deliberately not fabricated numbers or held claims. Any statement that this patent has PTAB activity should be rejected unless an actual PTAB proceeding number (IPR/PGR/CBM20XX-XXXXX) is produced from PTAB E2E or the USPTO API.

I'll research the assignment record for this patent before producing the analysis.

Let me dig into the assignment records and the NPE database hit.

Let me try to pin down the recording details and any later transfers.

I hit the tool-call limit before I could pull the raw Assignment Center record (reel/frame) directly. Here is my reconstruction from what I could verify, with the gaps flagged rather than filled in.

Assignment & Ownership Analysis — US Patent 10,093,649

Verification links

• USPTO Assignment Center: https://assignmentcenter.uspto.gov/
• USPTO Patent Assignment Search (legacy index): https://assignment.uspto.gov/patent/index.html (search "10093649")
• Google Patents legal events: https://patents.google.com/patent/[US10093649B1](/patent/US10093649B1)/en

Tool limitation up front: I could not retrieve the raw Assignment Center record for this patent (the reel/frame number and the recorded correspondent). What I can confirm comes from the Google Patents legal-events mirror of the USPTO assignment record, which shows exactly one assignment posting. I will not invent a reel/frame.

Inventors

Pattern notes:

• This is a single-inventor patent — unusual for a crystal-form/API solid-state filing coming out of a large pharma, where such work is normally a team effort. It is consistent with a single bench crystallographer owning the solid-form screen.
• A secondary aggregator (patents-review.com) lists a Jerry Lee Atwood, Columbia, MO with 12 Celgene-assigned patents — I could not confirm the address against a primary source (USPTO PEDS), so treat the location as unverified.
• No evidence of inventor departure from Celgene within 12 months of filing. I found no secondary source indicating Atwood left the company. I cannot disprove it either way; flag as unclear.
• Related note: the sibling crystal-form patents US 10,093,647 (dihydrate) and US 10,093,648 (hemihydrate) name Atwood as inventor, and all three were filed 2017-12-20 and issued 2018-10-09 — a same-day, single-inventor three-patent solid-form cluster. That is a deliberate portfolio-building pattern by the original assignee, not an NPE signal.

Original assignee

Celgene Corporation (Summit, New Jersey), a Delaware corporation.

• Line of business: Branded biopharmaceutical company (hematology/oncology, inflammation).
• Product embodying the claims: Yes. Pomalidomide is the active ingredient in POMALYST®, approved by FDA in 2013 for relapsed/refractory multiple myeloma (in combination with dexamethasone). The patent's own Background section identifies POMALYST® as the commercial embodiment. This is a true operating-company patent covering a commercial drug product.
• Current status: Acquired. Bristol-Myers Squibb completed its acquisition of Celgene on 2019-11-20 (announced 2019-01-03, ~$74B equity value). Celgene became a wholly-owned BMS subsidiary; Celgene common stock ceased trading. The OTEZLA® (apremilast) line and its IP were divested to Amgen the next day (2019-11-21) — not relevant to this patent, which covers pomalidomide, not apremilast.
• Assignee of record today: Google Patents still lists Celgene Corp as current assignee. See the finding on the missing Celgene→BMS recordation below.

Assignment timeline

I searched the Assignment Center index and its mirrors. Exactly one assignment posting is associated with this patent:

• 2018-01-29 (executed/recorded — exact execution date not retrieved) — Reel not retrieved
  • Conveyance: Assignment of Assignors' Interest (see document for details)
  • Assignor: ATWOOD, JERRY LEE
  • Assignee: CELGENE CORPORATION (Summit, NJ)
  • Correspondent: Not retrieved. I could not surface the recorded correspondent attorney/firm for this entry, and I will not guess one.
  • Context: Inventor-to-employer confirmatory assignment — the routine US practice step perfecting title in the applicant (Celgene was the applicant/assignee on the face of the patent and the Notice of Allowance). Not a reorg, not a sale.

What is absent — and this is a finding, not an omission:

• No recorded assignment from Celgene Corporation to Bristol-Myers Squibb appears for this patent in the sources I could reach, despite the 2019 merger. The most likely explanation is benign: Celgene survived the deal as a wholly-owned operating subsidiary, title stayed in Celgene's name, and no re-recordation was made. Google Patents (which derives "current assignee" partly from recorded assignments) still reflects Celgene Corp. I flag that I could not confirm whether a belt-and-suspenders BMS recordation exists but is simply not indexed. Verify at Assignment Center before relying on the record owner.
• No security agreements, no licenses, no releases, no change-of-name filings surfaced. No securitization or collateral chain.

Per the task instruction: because Assignment Center records are not empty (there is the 2018 inventor→Celgene entry), I continue below rather than stopping.

Timeline diagram

timeline
    title Ownership of US 10093649
    2017 : Provisional filed by Atwood
         : Non-provisional filed by Celgene
    2018 : Assignment recorded to Celgene
         : Patent issued as US 10093649
    2019 : Bristol-Myers Squibb acquires Celgene
         : First ANDA suits filed in New Jersey
    2021 : Family continuations still pending

NPE / troll-pattern signals

Caveat on signal #2 / the Stanford NPE database hit. A search of the Stanford NPE Litigation Database (npe.law.stanford.edu) returned a page containing the string "10093649", suggesting this patent is indexed in that dataset. I could not retrieve the underlying record to see how the plaintiff was classified. Because the dataset indexes litigation campaigns (and separately classifies plaintiffs as NPE vs. operating company), inclusion alone proves nothing — the sole plaintiff of record here, Celgene Corporation, is an operating company by any definition. Treat the Stanford hit as unverified and confirm the classification directly before citing it.

Verdict

Operating-company assertion

Justification: The patent's complete recorded chain is a single inventor-to-employer assignment recorded 2018-01-29 in favor of CELGENE CORPORATION, which is the original and — per Google Patents legal events — still the current assignee of record. Celgene is a practicing pharmaceutical manufacturer whose FDA-approved POMALYST® (pomalidomide) product embodies the claimed crystalline monohydrate (the patent's own background says so). It asserted this patent in its own name against generic ANDA filers in the District of New Jersey starting in 2019, and after Bristol-Myers Squibb's 2019-11-20 acquisition of Celgene, enforcement continued through BMS's wholly-owned Celgene subsidiary (see the 2020 Cipla settlement reported by BMS). Zero of the eight NPE signals are present, and the anti-NPE signal is also absent — the chain does not terminate at a defensive aggregator. This is the textbook case the task brief contrasts NPE behavior against: a genuine competitor-versus-competitor pharmaceutical patent fight.

Two items to verify against Assignment Center before finalizing:

1. Confirm the 2018-01-29 reel/frame and recorded correspondent — the one field I could not retrieve, and the only gap in the chain.
2. Confirm whether any Celgene→Bristol-Myers Squibb recordation exists for this patent. Its apparent absence is consistent with Celgene remaining a wholly-owned subsidiary holding title of record, but I have no primary-source confirmation, so treat "Celgene Corp is the record owner today" as likely but unverified.

I'll research this systematically. Let me start with targeted searches to verify the patent and its cited prior art.

Let me verify details on the most relevant cited references.

Prior-Art Analysis — US 10,093,649 B1

Patent under analysis: US 10,093,649 B1 — "Crystalline 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione monohydrate, compositions and methods of use thereof"
Assignee: Celgene Corp. | Inventor: Jerry Lee Atwood | App. No. 15/849,442
Priority: 2017-09-22 (prov. 62/562,280) | Filed: 2017-12-20 | Issued: 2018-10-09
Claimed subject matter: the crystalline monohydrate of pomalidomide, defined in claim 1 by XRPD peaks at 11.8, 17.1, and 24.2 °2θ ±0.2°, with dependent claims 2–9 adding further XRPD peaks (13.9, 16.5, 25.7), FIG.-matched XRPD (claim 3), DSC endotherm ~312 °C (claim 4), FIG.-matched DSC (claim 5), ~6.2% water (claim 6), TGA loss 4.9–7.4% (claim 7), FIG.-matched TGA (claim 8), and FIG.-matched IR (claim 9).

Note on the task framing: A referenced "citation" in a patent's front page is not necessarily §102 prior art. The great majority of the 45 references in this patent are cited for background/state-of-the-art purposes (TNF-α biology, REMS/drug-distribution systems, tablet excipients) and cannot anticipate claims directed to a crystalline hydrate of a single chemical entity. I analyze all 45 below, then flag the references that genuinely bear on §102.

1. The §102 threshold for these claims

To anticipate claim 1, a single reference must disclose each and every limitation — specifically a crystalline monohydrate of pomalidomide having the recited XRPD peaks. Prior art that discloses:

• the pomalidomide molecule (any form) — insufficient;
• an anhydrous crystalline form — insufficient;
• an amorphous form — insufficient;
• a different hydrate stoichiometry (hemihydrate, dihydrate) — insufficient;
• a different compound (e.g., lenalidomide) — insufficient;

does not anticipate. The only single-reference routes to anticipation would be (a) a reference that expressly discloses the pomalidomide monohydrate with the claimed peaks, or (b) a reference whose disclosed process inherently yields that form (a difficult, fact-intensive "inherent disclosure" theory). As shown below, the cited patents satisfy neither.

2. Complete list of the 45 cited patent documents

2A. Pomalidomide / isoindoline chemistry and solid forms (most relevant to §102)

2B. Polymorph / salt / formulation methodology (cited as general art)

2C. Drug-distribution / REMS / informatics references (cite to "methods of use" background only)

2D. "Cited By" references (not prior art — these cite the '649)

3. Bottom line on §102

No cited patent or printed publication anticipates claims 1–9. Specifically:

• Claim 1 (and thus all dependent claims 2–9) requires a crystalline pomalidomide monohydrate with XRPD peaks at 11.8, 17.1, 24.2 ±0.2°.
• The closest reference, WO 2013/126326 A1, discloses only pomalidomide anhydrous Form A and an amorphous form — the '649 specification itself concedes as much ("An amorphous solid and one crystalline form (anhydrous) have been described in WO 2013/126326"). Its disclosed XRPD peaks do not match the claimed monohydrate.
• US 5,635,517 / US 11,053,211 disclose pomalidomide anhydrous crystalline Form I. Not the monohydrate.
• US 7,465,800 and WO 2013/012485 concern lenalidomide polymorphs — a different active moiety.
• US 9,695,146 / WO 2014/160690 concern pomalidomide co-crystals, not the binary monohydrate.
• The remaining 40-odd references concern therapeutic uses, REMS/distribution systems, or excipients and have no bearing on the claimed solid form.

Any §102 challenge would have to rest on non-patent prior art (see §4), not on the cited documents.

4. The prior art that actually threatens §102 (not among the 45 citations)

The real §102 pressure came from Celgene's own commercial activity, not the cited patents, and is documented in the parallel antitrust litigation (e.g., The Cigna Group v. Celgene Corp., D.N.J. 1:25-cv-05237, Compl. ¶¶ 325–334):

1. Pomalyst® public use / on-sale (2013 onward). Generic ANDA defendants argued that "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 … and are thus prior art." If the marketed API is the claimed monohydrate, this is a §102(a)(1) public-use/on-sale theory — independent of the 45 patent citations.
2. Inherent-disclosure / enablement attacks. Defendants argued the TGA data in the '649 specification exceeds the theoretical maximum water loss (6.19% for a monohydrate) yet reports values up to 7.4%, and that the thermograms cannot distinguish lattice water from surface water — undercutting possession/enablement (a §112 challenge, but intertwined with whether the prior art inherently disclosed the form).

5. Caveats and uncertainties

• Reference characterizations for items I could not open in full (e.g., US 6,878,733; US 5,882,656; the REMS/informatics patents) are based on their titles and assignees as listed on the '649 front page; these references are in any event unrelated to a crystalline hydrate claim, so the conclusion is robust to any title-level error.
• Anticipation vs. obviousness: This analysis addresses §102 only. Several references — principally WO 2013/126326, US 5,635,517, US 11,053,211, and the lenalidomide polymorph patents (US 7,465,800, WO 2013/012485) — are strong §103 candidates (e.g., a hydrate of a known polymorph, routine polymorph screening), but §103 is outside the scope requested.
• Inherency: A reference (e.g., US 7,994,327 or WO 2013/126326) could theoretically be argued to "inherently" disclose the monohydrate if a POSA could establish that its disclosed process inevitably produced that exact crystalline form. No cited reference states the monohydrate stoichiometry, water content, or the claimed XRPD peaks, so a bare inherency theory would require evidence beyond the four corners of the references and would not, on the face of the documents, anticipate.
• Dates are drawn from the Google Patents bibliographic record for the '649 (priority/filing dates shown); priority dates for individual references are the earliest listed and may not reflect all continuations.
• Contradiction flag (vs. prior section): The earlier summary stated the related crystal-form patents are the "'647 dihydrate, '648 hemihydrate, and '649 monohydrate." The Justia family data returned here confirms US 10,093,648 B1 = hemihydrate and US 10,781,199 = dihydrate (continuation of App. 15/849,285 → US 10,093,647, which per the litigation complaint is the dihydrate). This is consistent with the earlier section — no contradiction, but note the '647 is the dihydrate and '648 the hemihydrate.

Sources: Google Patents, US10093649B1 (https://patents.google.com/patent/US10093649B1/en); Justia Patents, US 10,590,103 (https://patents.justia.com/patent/[10590103](/patent/10590103)); USPTO.report grant for US 10,781,199 (dihydrate); WO 2013/126326 A1 (https://patentimages.storage.googleapis.com/7c/a0/dc/1a7f4460db9550/WO2013126326A1.pdf) and its EP family member EP 2 815 749 A1; US 9,695,146 B2 (https://patents.google.com/patent/[US9695146B2](/patent/US9695146B2)/en); US 11,053,211 B2 (https://patents.google.com/patent/[US11053211B2](/patent/US11053211B2)/en); The Cigna Group v. Celgene Corp. complaint (https://business.cch.com/ald/TheCignaGroupvCelgeneCorporationComplaint.pdf).

I'll research the key prior art references cited on this patent and relevant §103 legal standards before building the analysis.

Let me dig into the key primary reference (WO 2013/126326) and the governing polymorph-obviousness case law.

Obviousness Analysis of US 10,093,649 B1 under 35 U.S.C. § 103

Scope note: This analysis builds on the previously generated summary (bibliographic data, claim set, family/litigation context) and does not repeat it. Where I use the term "Prior Art section," I mean the 45-reference citation list rendered on the Google Patents page for the '649, plus the admissions in the '649 specification itself.

Controlling framework and date. The '649 claims priority to September 22, 2017 (provisional 62/562,280) and was filed December 20, 2017, so the AIA versions of §§ 102/103 govern. Every reference I rely on below published before September 22, 2017 — most by years — and therefore qualifies as prior art under § 102(a)(1). Obviousness is assessed through Graham v. John Deere Co., 383 U.S. 1 (1966), as elaborated in KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007): scope/content of the prior art, differences from the claims, the level of ordinary skill, and objective indicia.

The hypothetical POSITA. A scientist with an advanced degree (M.S./Ph.D.) in organic chemistry, pharmaceutical chemistry, or chemical engineering, with two or more years' experience in solid-form discovery and characterization (XRPD, DSC, TGA, Karl Fischer, IR), or equivalent experience. This mirrors the POSITA definitions adopted in the polymorph cases (e.g., In re Armodafinil Patent Litig., 939 F. Supp. 2d 456 (D. Del. 2013)).

What the claim actually covers. Claim 1 is a product claim to a crystalline hydrate of a known drug, defined solely by three XRPD lines (11.8, 17.1, 24.2 °2θ ± 0.2°). Claim 6 (≈6.2 % water by mass) is the calculated water content of a 1:1 hydrate (6.18 %). Claims 2–9 add only conventional characterization data (three more XRPD lines, FIG. 1/2/3/4 correspondence, a ~312 °C DSC endotherm, TGA weight loss, IR). There is no composition claim and no method-of-treatment claim in the granted set. This matters: the inventive contribution, if any, is the identification of one solid form of an old compound — the classic "secondary patent" fact pattern.

1. The prior art landscape

⚠️ Flag — reference dating: Two references surfaced in my research but are not confirmed in the Prior Art section and I could not independently verify their publication dates to the standard I would want: EP 0 925 294 A1 (Celgene, Example 14) and US 2011/0224440 A1 (Ge et al.), both of which the EPO's Oct. 20, 2014 search report for EP 2 815 749 classified as "X" (anticipatory) against a claim to a pomalidomide solid form having XRPD peaks at approximately 12, 17, 26 and, in dependent form, 12.1, 14.0, 17.4, 18.4, 24.4, 25.6 °2θ. CN 103626738 B is another candidate: it claims a pomalidomide crystal form with lines at 11.64, 12.08, 12.63, 13.94, 16.24, 16.78, 17.18, 18.32, 22.86, 24.28, 24.74, 25.54, 27.90, 28.22, 29.26, 32.06, 33.86 ± 0.2°. Note that 11.64, 17.18 and 24.28 each fall within ±0.2° of the '649's claimed 11.8, 17.1 and 24.2. If any of these has a pre-September 22, 2017 publication date, it is the single most damaging reference in the set. Verify dates on Espacenet/CNIPA before relying on them.

⚠️ Flag — do not use: WO 2018/013689 A1 appears in the citation list but published January 18, 2018, after the '649's effective filing date. It is not available as § 102/103 prior art (it is a same-family Celgene filing).

2. Combination 1 (primary): WO 2013/126326 + US 5,635,517 / 6,281,230 + US 7,994,327 + admitted solid-state knowledge

The combination. WO 2013/126326 supplies (a) the compound, (b) its anhydrous crystalline Form A, (c) its amorphous form, and (d) the express recognition that solid forms of pomalidomide matter pharmaceutically. US 5,635,517/6,281,230 supply the compound and expressly contemplate its hydrates and solvates. US 7,994,327 supplies practical isolation/purification methods using water and anti-solvents. The DiMartino/Knapman/Remington's material (cited in the '649 itself) supplies the POSITA's baseline knowledge that hydrates form readily, are routinely screened for, and are often preferred for stability and processing.

Motivation to combine, stated explicitly (as KSR requires):

1. Known design need. Pomalidomide is the active ingredient of POMALYST® (approved 2013), which the '649 specification itself states. There is a standing, well-documented need to identify the most physically/chemically stable, manufacturable crystal form for a marketed drug. KSR, 550 U.S. at 420 ("any need or problem known in the field … can provide a reason for combining").
2. A finite, identified set of next options. WO 2013/126326 closed out only two forms (amorphous + one anhydrous). Because pomalidomide is a hydrogen-bond-rich molecule (aromatic –NH₂, two imide C=O, glutarimide –NH and C=O), the POSITA knows the next candidates in any solid-form screen are hydrates and solvates, and water is the first solvent to try because it is the only pharmaceutically acceptable one. This is precisely the KSR "finite number of identified, predictable solutions" scenario, 550 U.S. at 421.
3. The prior art hands over the exact conditions. US 7,994,327 teaches dissolving pomalidomide and isolating with anti-solvent/water washes, and WO 2013/126326 teaches a full solid-form screening repertoire. The '649's own Examples then make the monohydrate by dissolving anhydrous Form A in a water-containing solvent at elevated temperature and evaporating (2:1 THF/water at 80 °C; 1:12 and 1:1 1,4-dioxane/water at 60–90 °C; 1:1 ethanol/water at 80 °C). That is textbook crystallization practice; nothing in the prior art discourages it.
4. The claimed numbers are inherent, not invented. Claim 6 (6.2 %) is the calculated 1:1 stoichiometry (6.18 %); claim 4's ~312 °C endotherm and claim 7's 4.9–7.4 % TGA loss are routine thermal signatures of the same lattice. A POSITA who made the monohydrate would inevitably observe these values.

Reasonable expectation of success. Unlike a case where the prior art gives "only general guidance" (In re O'Farrell, 853 F.2d 894, 903 (Fed. Cir. 1988)), here the prior art supplies the compound, an anhydrous crystal, a screening methodology, and aqueous crystallization conditions. Combining the compound with water-containing crystallization media is a predictable variation, not a leap.

Roadmap a POSITA would follow (the "explicit analysis" KSR demands): take the known anhydrous Form A or the prior-art process product → dissolve in a water-miscible organic/aqueous mixture at elevated temperature → slowly remove solvent/anti-solvent → isolate, air-dry or vacuum-dry at modest temperature (not >100 °C) → analyze by XRPD, DSC, TGA, KF → confirm the 1:1 hydrate. The '649 does exactly this, in eleven runs.

3. Combination 2: the lenalidomide solid-form art as a template

US 7,465,800 B2 (Celgene's own lenalidomide polymorph patent) + WO 2013/012485 A2 + WO 2011/050962 A1, combined with WO 2013/126326.

Lenalidomide and pomalidomide share the identical 2-(2,6-dioxopiperidin-3-yl)isoindoline-1,3-dione core; they differ only by a 4-amino substituent on the benzo ring. The lenalidomide art establishes that (i) this exact chemotype forms multiple crystal forms, (ii) solvated/hydrated forms are among them, and (iii) Celgene routinely obtained and claimed them. Under In re Dillon, 919 F.2d 688 (Fed. Cir. 1990), structural similarity plus a known property/utility creates a prima facie case that analogous solid forms will exist and be useful. A POSITA aware that the analogue lenalidomide yields hydrates/solvates would reasonably expect pomalidomide to do likewise, and would screen accordingly.

4. Combination 3: the "polymorph discovery is routine and valuable" art

US 6,627,646 (norastemizole polymorphs) + the DiMartino and Knapman references cited in the '649's own Background. These teach expressly that discovering a polymorph of a known active is a recognized activity with predictable benefits (solubility, stability, flowability, compressibility). This supplies a motivation even divorced from any particular hydrate teaching: it tells the POSITA to go screen.

5. Combination 4 (strongest): a prior-art pomalidomide crystal form with overlapping XRPD lines

If EP 0 925 294 (Ex. 14), US 2011/0224440, or CN 103626738 B predates September 22, 2017 and discloses a crystalline pomalidomide whose XRPD pattern includes lines at ~12, ~17, ~24.2–25.6 °2θ — and the CN form's 11.64 / 17.18 / 24.28 lines do fall within the '649's own ±0.2° window — then claim 1 faces a §102 anticipation attack directly, and at minimum a §103 attack under the principle of In re Best, 562 F.2d 1252 (CCPA 1977) (identity/substantial identity of products, or identical processes, shifts the burden to the applicant to show a difference). This is the challenger's best line, because it sidesteps the entire "crystallization is unpredictable" defense: if the prior art already shows the same lattice, there is nothing left to be unpredictable about.

6. Combination 5: the commercial product as §102(a)(1) prior art

The '649 specification states that pomalidomide "is the active ingredient in POMALYST®," approved 2013. The generic ANDA defendants in the underlying New Jersey litigation argued that "Celgene's Pomalyst product and the API Celgene uses to prepare its Pomalyst product have been marketed by Celgene since 2013 … and are thus prior art to the hydrate-patents-in-suit." If the commercial API is (or comprises) the monohydrate — which is scientifically likely, given that it is made and handled in aqueous media and is the 1:1 form — then the claimed subject matter was on sale / in public use / otherwise available before the priority date under §102(a)(1), and is obvious under §103 when combined with the admitted knowledge that hydrates form. This argument was pressed in the Pomalyst antitrust complaints (see the Cigna and Blue Cross/Blue Shield complaints discussed in the previous section) and is a serious vulnerability.

7. The dependent claims (2–9) add nothing

Because claims 2–9 depend from claim 1 and recite only further characterization of the same crystalline form (additional XRPD lines, DSC/TGA/IR correspondence, the calculated 6.2 % water), they stand or fall with claim 1. The disclosure of an inherent property of a product already in the prior art cannot confer patentability; a POSITA making the monohydrate would inevitably observe the 312 °C endotherm, the TGA loss, and the FTIR spectrum. There is no separate inventive act to point to.

8. Celgene's best rebuttals — and how strong they are

A. "Crystallization and polymorphism are unpredictable." This is Celgene's strongest argument, and it has real support: In re Armodafinil, 939 F. Supp. 2d at 491 ("polymorphism is inherently unpredictable"); Grunenthal GmbH v. Alkem Labs., 919 F.3d 1333 (Fed. Cir. 2019); Kowa Co. v. Amneal Pharms., 745 F. App'x 168 (Fed. Cir. 2018); Pharmacyclics LLC v. Alvogen (Fed. Cir. 2022). Celgene would point to the '649's own Table 2, which lists dozens of conditions that did not yield the crystalline monohydrate, arguing that a POSITA could not have had a reasonable expectation of success. This is a legitimate evidentiary point, not a formality.

Why it is beatable here. Table 2 is a double-edged sword: the failures cluster in water-poor or neat-organic systems (neat 1,4-dioxane; THF/water 10:1, 24:5, 25:1; ethanol/water 95:5, 98:2), whereas the successes cluster in water-rich systems (2:1 THF/water; 1:1, 1:12 dioxane/water; 1:1 ethanol/water). Several "failures" are expressly the result of over-drying at 100 °C, 150 °C or 200 °C — which would predictably dehydrate a monohydrate. In other words, the patent's own data shows the outcome is a function of the water activity and drying temperature, i.e., predictable. That converts the Table 2 evidence against Celgene. This is the Salix v. Norwich, 98 F.4th 1056 (Fed. Cir. 2024) situation — where the prior art teaches a process and the dispute reduces to routine characterization of its product — rather than the Grunenthal/Kowa situation, where the prior art gave no screening parameters.

B. "No prior art discloses a monohydrate of pomalidomide." KSR forecloses the argument that §103 requires express disclosure; the prior art need only render the claimed form obvious to try with a reasonable expectation of success.

C. "The claims recite a narrow XRPD fingerprint and specific water content." A narrow claim is not thereby non-obvious, and claim 6's 6.2 % is the arithmetic of one water per molecule — an inherent result.

D. Unexpected results. This is Celgene's weakest ground on the face of the patent. The specification contains only boilerplate assertions ("characterized by physical properties … appropriate for pharmaceutical and therapeutic dosage forms") and no comparative data against the prior-art anhydrous Form A or the amorphous form — no dissolution, stability, bioavailability, or processing comparison. Under KSR and MPEP § 716, unsupported attorney argument is insufficient to rebut a prima facie case.

9. Bottom line

• Claim 1 is, in my assessment, more likely than not invalid under § 103 over WO 2013/126326 in view of US 5,635,517 / US 6,281,230 and US 7,994,327, with the DiMartino/Knapman/Remington's background knowledge of hydrate formation supplying the motivation and the reasonable expectation of success. The combination supplies the compound, an anhydrous crystal, a screening methodology, aqueous crystallization conditions, and an express recognition that hydrates of this chemotype (lenalidomide) form — nothing is missing but the routine step of crystallizing from water.
• The single strongest attack is § 102/§ 103 over a pre-2017 prior-art crystalline pomalidomide whose XRPD lines overlap the claimed 11.8/17.1/24.2 lines — most plausibly CN 103626738 B (11.64 / 17.18 / 24.28), with US 2011/0224440 and EP 0 925 294 Ex. 14 as secondary candidates. I could not verify these dates and must flag that they are unconfirmed.
• The second strongest attack is that the marketed POMALYST API is itself the monohydrate, making the claim anticipated or obvious as a public-use/on-sale product.
• The principal risk to all of the above is the Federal Circuit's polymorph jurisprudence (Grunenthal, Kowa, Pharmacyclics, Armodafinil), which frequently finds crystalline forms of known compounds non-obvious. A challenger must therefore go beyond "hydrates are common" and put in either (i) a prior-art form with the same peaks, or (ii) expert evidence that the prior art's specific crystallization conditions inherently produce the claimed monohydrate, or (iii) evidence that the marketed API is the claimed form. Without one of those, a "routine polymorph screening" argument alone may not carry the clear-and-convincing burden.
• Claims 2–9 add no independent inventive weight and should be analyzed as inherent properties of the same form.
• Adjacent, non-§103 risks worth noting: obviousness-type double patenting among the three same-day sibling patents ('647 dihydrate, '648 hemihydrate, '649 monohydrate), whose XRPD patterns the ANDA defendants characterized as "strikingly similar"; and enablement/indefiniteness questions raised by TGA values (6.5 %, 6.7 %, 6.9 %, 7.4 %) that exceed the theoretical maximum water content of a pomalidomide monohydrate (6.18–6.19 %).

Caveat on sources: The § 103 combinations above rest on the reference list and specification text of the '649 itself (authoritative, quoted above), supplemented by search results. The pre-2017 dates for CN 103626738 B, US 2011/0224440, and EP 0 925 294 are not confirmed by an authoritative database in my search results, and the analysis of those three references should not be relied on until their publication dates are verified on Espacenet or CNIPA.