US 10093648

I'll search for authoritative information on US 10093648, including USPTO and CAFC docket information.

Let me search for the specific CAFC docket numbers and litigation history associated with this patent.

Let me check the specific CAFC docket 21-2335 and any current 2026 appellate activity.

US Patent 10,093,648 — Summary

Note on identifiers: I interpreted 10093648 literally as US 10,093,648 B1. Searches for that exact number returned consistent, corroborating results; I did not substitute any similar-looking number.

1. Bibliographic data

The molecule is pomalidomide (POMALYST®), the active ingredient in Celgene's multiple-myeloma product. The claimed solid form is the hemihydrate (~0.5 mol water per mol compound), distinguished from the dihydrate (US 10,093,647) and monohydrate (US 10,093,649) patents filed the same day.

(Source: https://patents.google.com/patent/US10093648/en)

2. Abstract

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

3. Plain-language overview of the independent claim

Claim 1 (the only independent claim):

"Crystalline 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione hemihydrate, having an X-ray powder diffraction pattern comprising peaks at 12.0, 17.2, and 25.6 degrees 2θ±0.2 degrees 2θ."

In plain terms, claim 1 is a composition-of-matter claim covering a specific crystalline (polymorphic) hydrated form of pomalidomide — specifically the hemihydrate — identified by three signature X-ray powder diffraction (XRPD) peaks. It contains no method or formulation step; it claims the solid form itself.

Dependent claims (2–10) narrow it by added characterization:

• Claim 2 — XRPD pattern further comprising peaks at 13.9, 16.7, and 24.3 °2θ ±0.2.
• Claim 3 — XRPD pattern corresponding to FIG. 1.
• Claim 4 — DSC thermogram with an endotherm maximum at about 317 °C.
• Claim 5 — DSC thermogram corresponding to FIG. 2.
• Claim 6 — about 3.2% water by mass.
• Claim 7 — TGA showing weight loss between about 2.8% and about 4.3% when heated from about 30 °C to about 225 °C.
• Claims 8–9 — TGA thermogram corresponding to FIG. 3 / FIG. 4, respectively.
• Claim 10 — IR spectrum corresponding to FIG. 5.

4. Patent family (all share the 2017-09-22 priority date)

• US 10,093,648 B1 — app. 15/849,324 (this patent)
• US 10,487,069 B2 — app. 16/124,057, filed 2018-09-06 (continuation)
• US 10,829,472 B2 — app. 16/670,801, filed 2019-10-31 (continuation)
• US 11,866,417 B2 — app. 17/061,473, filed 2020-10-01 (continuation; published as US 2021/0024488 A1)

5. Litigation and CAFC activity

District court (D.N.J.): Google Patents' litigation metadata lists this patent/family in numerous New Jersey cases, including Celgene v. Aurobindo (2:19-cv-05799) and seven other 2019 filings (2:19-cv-05797, -05802, -05804, -05806, -08758, -09737) plus 2:21-cv-02111. The '648 was asserted in Hatch-Waxman suits (e.g., against Aurobindo, Breckenridge) as the hemihydrate-form patent.

Federal Circuit (verified):

• No. 2021-1154, Celgene Corp. v. Mylan Pharmaceuticals Inc. — Fed. Cir., Nov. 5, 2021 (Prost, Chen, Hughes). Affirmed dismissal of the '647/'648/'649 infringement action for improper venue (28 U.S.C. § 1400(b)) and for failure to state a claim as to Mylan N.V. This appears to be a venue-based disposition, not a merits ruling on the '648 claims.
• No. 21-2335 — listed in Google Patents' litigation data as a Federal Circuit case in this family, but I could not verify its subject matter or outcome from the sources retrieved.

Regarding "CAFC 2026 dockets": I could not find any 2026 Federal Circuit docket, opinion, or pending appeal specifically for US 10,093,648. The identifiable appellate activity is from 2021. I flag this as an explicit uncertainty rather than assert that none exists — my searches did not surface a 2026 docket tied to this patent number.

Related antitrust litigation (currently active, cites this patent): The '648 is discussed at length in the Pomalyst/pomalidomide antitrust complaints — e.g., The Cigna Group v. Celgene Corp. (D.N.J., filed on or about 2025-06-24, 1:25-cv-05237-ER) and an earlier 2023 complaint (1:23-cv-07871-ER) — which allege the crystalline-form patents were "sham" filings used to delay generic entry, culminating in settlements providing for generic entry in Q1 2026. One complaint notes the '648 was "Not OB Listed" (i.e., not Orange-Book listed), while other complaints treat it as a patents-in-suit — a discrepancy I have not resolved.

Validity/characterization challenges raised in that litigation (as alleged, not adjudicated): TGA data reported in the '648 specification (up to 4.3% weight loss) exceeds the theoretical maximum for a pomalidomide hemihydrate (3.2%), which opposing parties argued calls the claimed composition into question; enablement, written description, and indefiniteness challenges were also previewed. Celgene withdrew infringement claims as to many claims of the crystalline-form patents on October 12, 2020, and settled with ANDA filers shortly thereafter.

(Sources: https://www.robinskaplan.com/newsroom/insights/celgene-v-mylan-pharms ; https://today.westlaw.com/Document/Ie8f1687f407c11ec9f24ec7b211d8087/View/FullText.html ; https://business.cch.com/ald/TheCignaGroupvCelgeneCorporationComplaint.pdf ; https://www.hbsslaw.com/sites/default/files/case-downloads/pomalyst-pomalidomide/2023-09-07-complaint.pdf ; https://www.docketalarm.com/cases/New_Jersey_District_Court/2--19-cv-05799/)

6. Uncertainties

• I have no authoritative 2026 CAFC docket for this patent; the last verifiable Federal Circuit case is No. 2021-1154 (venue, affirmed). Treat any statement about a live 2026 appeal as unverified.
• I could not confirm the subject matter or disposition of CAFC No. 21-2335.
• Legal status per Google Patents is "Active" with anticipated expiration 2037-12-20, but that is a database assumption, not a legal conclusion; given the extensive 2020–2021 settlements and the antitrust challenges, the practical enforceability picture may differ.
• Whether the '648 is Orange-Book listed is inconsistent across the complaints I retrieved.

I'll research litigation involving US patent 10093648. Let me run several targeted searches.

Let me search for more specific details on the cases identified in the Google Patents litigation records.

Let me get details on the remaining cases and the Federal Circuit appeals.

Litigation Involving US Patent 10,093,648 ("the '648 patent")

Important framing

US 10,093,648 B1 — Crystalline 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione hemihydrate, compositions and methods of use thereof (Celgene Corp.; inventor Jerry Lee Atwood; filed Dec. 20, 2017; issued Oct. 9, 2018) is one of Celgene's three "crystal form" pomalidomide patents — the '647 (dihydrate), '648 (hemihydrate), and '649 (monohydrate) patents. It was asserted only in the second wave of Hatch‑Waxman/ANDA litigation over generic Pomalyst, because it did not issue until October 2018 (after Celgene's original 2017 suits against the ANDA filers). Its claim 1 covers the hemihydrate characterized by XRPD peaks at 12.0, 17.2, and 25.6 °2θ ± 0.2.

Note: The litigation records list cases where the '648 was among the patents-in-suit. Some case numbers below appear in the patent's recorded litigation data but I could not independently confirm the defendant within my search limit; I flag those explicitly rather than guess.

U.S. District Court, District of New Jersey (ANDA / Hatch-Waxman suits asserting the '648 patent)

The Google Patents/Unified Patents litigation record for US 10,093,648 B1 lists the following D.N.J. cases. Most are assigned to Judge Esther Salas (ES) with Magistrate Judge Michael A. Hammer (MAH).

U.S. Court of Appeals for the Federal Circuit

Related / follow-on antitrust litigation that references the '648 patent

These are not infringement suits asserting the '648 patent, but they directly attack the Pomalyst patent family (including the '647/'648/'649 crystal-form patents) as fraudulently obtained and/or the suits on them as "sham" litigation:

• The Cigna Group v. Celgene Corp. — No. 1:25-cv-05237 (E.D.N.Y., filed June 24, 2025) — Sherman Act § 2 / Clayton Act monopolization claims challenging among others the crystalline-form patents; Celgene/BMS moved to dismiss.
• In re Revlimid and Thalomid (Pomalyst) Purchaser Antitrust Litigation — No. 2:19-cv-07532-ES-MAH (D.N.J.) — consolidated class/insurer/MSP purchaser actions.
• Securities/antitrust class actions in S.D.N.Y. (e.g., 1:23-cv-07871-ER; 1:25-cv-… ) addressing the same patent family; courts have twice dismissed "sham litigation" and Walker Process theories as to certain patents.

Summary of outcome

• No case against US 10,093,648 has gone to a merits trial verdict that I can confirm. The D.N.J. ANDA cases asserting the '648 patent were consolidated, litigated through claim construction (June 16, 2020 order) and invalidity/enablement/indefiniteness challenges, and were resolved by settlement allowing generic pomalidomide entry in/around Q1 2026.
• The one appellate decision clearly tied to this case family is the Federal Circuit's affirmance in Celgene Corp. v. Mylan Pharmaceuticals Inc. (venue), which arose from 2:19-cv-05802, one of the '648 cases.

Caveats / limits of this answer

• Per your instruction, all case numbers and patent numbers are reproduced literally; I did not normalize or "correct" them.
• The litigation list is drawn from the patent's own Google Patents/Unified Patents litigation record (which cites New Jersey District Court and Federal Circuit dockets) plus PACER/CourtListener and secondary sources. The party identities for 2:19-cv-05804, 2:19-cv-09737, and 2:21-cv-02111, and the parties/disposition of CAFC 21-2335, could not be independently confirmed before I exhausted my search steps; I have flagged these as unconfirmed rather than assert them.
• For authoritative, current docket-level detail on any single case, verify directly in PACER (D.N.J.) or the Unified Patents litigation portal case list (https://portal.unifiedpatents.com/litigation/caselist).

I'll start by verifying whether any AIA trial proceedings exist for this patent, since the structured ODP block reports none.

Let me run targeted searches on the crystal-form patent family and any PTAB filings.

Let me check specifically for any AIA trial petitions, including by defensive aggregators, against this patent family.

Proceedings overview

There are zero AIA trial proceedings on file for US 10093648 — the USPTO ODP structured block reports no IPR/PGR/CBM, and my independent web search surfaced no PTAB petition, institution decision, or Final Written Decision naming this patent (or its '647 dihydrate / '649 monohydrate siblings). Bottom line for a defendant: the patent has never been tested at the PTAB, so no claim has been canceled and no petitioner is estopped — the claims are alive and untested at the Board, and an IPR is procedurally available to you unless you are already time-barred under § 315(b). This is materially different from "hardened" (the patent has never survived an IPR) but also from "dead" (nothing has been invalidated). The only invalidity pressure on this patent to date has come from district court Hatch-Waxman litigation that settled before any merits judgment, not from the Board.

No AIA trial proceedings to itemize

Because there is no proceeding to model, the usual per-proceeding template would require fabrication. I am not filling it in. Instead, here is the relevant non-PTAB context, clearly labeled as not an AIA trial:

District court Hatch-Waxman litigation (not PTAB). Celgene (now BMS) asserted the '648 patent in the Pomalyst ANDA campaign. Per the structured block, the New Jersey cases include 2:19-cv-05806, 2:19-cv-05804, 2:19-cv-05802, 2:19-cv-05799, 2:19-cv-05797, 2:19-cv-08758, and 2:21-cv-02111, with Federal Circuit appeals 21-1154 and 21-2335. Those are district-court actions (Hatch-Waxman § 271(e)(2) venue and related), not AIA trials. I could confirm 21-1154 is Celgene Corp. v. Mylan Pharmaceuticals Inc., Fed. Cir. Nov. 5, 2021, affirming dismissal for improper venue under § 1400(b) (see Robins Kaplan case note and CourtListener docket 21-1154). I could not confirm the subject matter of 21-2335 and will not assume it is a PTAB appeal.

What happened on validity in the district court. The generic defendants did mount the substantive challenge — anticipation by Celgene's own earlier-marketed Pomalyst API, enablement, written description, indefiniteness, and internal-consistency attacks on the TGA water-content data (the '648 reports up to 4.3% weight loss against a theoretical hemihydrate maximum of ~3.2%). These are recited in the purchaser/antitrust complaints (e.g., Cigna Group v. Celgene complaint, ¶¶ 326–333). However, every one of these cases settled in Fall 2020–2021 (Celgene–Natco, –Teva, –Aurobindo, etc.), before trial or any invalidity judgment. So there is no holding — from the PTAB or a district court — that any claim of the '648 is invalid.

Strategic summary

Claim status. All ten claims of US 10093648 are UNTESTED. Claim 1 is the sole independent claim ("Crystalline 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione hemihydrate, having an X-ray powder diffraction pattern comprising peaks at 12.0, 17.2, and 25.6 degrees 2θ ± 0.2 degrees 2θ."); claims 2–10 are dependents adding XRPD peaks (12.0/17.2/25.6+13.9/16.7/24.3), FIG.-based characterization, DSC endotherm ~317 °C, water content, TGA weight loss, and IR. No claim has been canceled, disclaimed, or held unpatentable in any PTAB or court proceeding I can verify. The family's continuations (US 10,487,069, US 10,829,472, US 11,866,417) are separate patents with their own (also untested) claims.

Estoppel landscape. Because there is no IPR and no FWD, there is no § 315(e)(2) estoppel against anyone. Any party not yet time-barred can still file an IPR on the '648. The practical gates are:

• § 315(b) one-year bar — a petition must be filed within one year of service of a complaint alleging infringement of the '648. Defendants in the 2019/2021 D.N.J. waves are long past that bar for those patents; a newly-served defendant is not.
• § 325(d) and Fintiv-type discretionary denial — with the ANDA litigation largely settled, the parallel-litigation overlay that once made discretionary denial likely has largely evaporated, which arguably improves the prospects of a fresh petition.

Pattern signals. No petitioner has filed even one IPR on this patent, let alone multiple. There is no defensive aggregator in the chain for the '648 — Unified Patents lists the patent in its portal but I found no Unified-filed IPR. Note for accuracy: the well-known Coalition for Affordable Drugs VI LLC v. Celgene, IPR2015-01092 (which invalidated Celgene's '501 REMS patent) concerned a different patent and should not be conflated with the '648. The absence of any IPR here even during a nine-defendant ANDA war is the notable signal: generics litigated validity in court (cheaper, bundled with the ANDA case) and then settled, rather than going to the Board.

Recommended next steps

1. If you have just been served with a '648 infringement complaint, calendar the § 315(b) bar immediately — you have one year from service to file an IPR. That is your only hard deadline and it is running now.
2. Build the IPR around the strongest district-court theories, not new ones. The record the generics developed (but never got to judgment) is the roadmap: (a) § 102 anticipation by Celgene's own pre-2017-marketed Pomalyst API as prior art; (b) § 112 enablement / written description / indefiniteness; (c) § 103 over the overlapping XRPD peaks of the '647/'649 sibling applications and the public polymorph literature. Because there is no FWD to quote, you cannot point to a prior PTAB disposition — you will be the first mover, which cuts both ways (no adverse estoppel, but also no precedent to lean on).
3. Where the earlier ANDA settlements are on your docket, check for estoppel/issue-preclusion gaps — those settlements were confidential as to terms (entry dates were kept secret for ~18 months per the purchaser complaints), and a settlement is not an adjudication, so it confers no invalidity finding in your favor.
4. If the patent is being asserted against you at the ITC or in a new D.N.J. campaign, note the expiry backdrop: anticipated expiration is 2037-12-20 (20 years from the 2017-12-20 filing), so there is a long runway and the patent owner has every incentive to keep the '648 alive.

Confidence note: I am confident in the negative finding — both the ODP structured block and my web searches returned no AIA trial proceeding for US 10093648. I am not able to confirm the subject matter of Fed. Cir. 21-2335, and I flag it as an unverified non-PTAB appeal rather than assert it was a PTAB review.

I'll research the assignment record and ownership history for US 10,093,648. Let me run several searches in parallel.

Let me dig deeper into the assignment record specifics, the inventor's affiliation, and the Celgene–BMS merger's effect on patent ownership.

Inventors

Jerry Lee Atwood — sole named inventor, listed address Columbia, MO.

• Employer at filing: not determinable from the patent front page, but the record indicates he was an outside inventor rather than a Celgene employee. The Cigna v. Celgene antitrust complaint states plainly: "The named inventor is Jerry Atwood, who subsequently assigned the patents to Celgene." (Complaint ¶312 n.148, https://business.cch.com/ald/TheCignaGroupvCelgeneCorporationComplaint.pdf). That is consistent with the USPTO assignment of record, in which Atwood is the assignor conveying rights to Celgene — i.e., he did not file as an employee-inventor under an automatic assignment obligation.
• The Columbia, MO address is consistent with the University of Missouri (Columbia), where a Jerry L. Atwood is a well-known supramolecular/crystal-engineering chemist. I could not verify his employer at the filing date within this session, so treat the university affiliation as likely but unconfirmed.
• Atwood is the common sole inventor across the whole 2017 hydrate trio — US 10,093,647 (dihydrate), US 10,093,648 (hemihydrate, this patent), and US 10,093,649 (monohydrate) — all filed December 20, 2017, all with the same Sept. 22, 2017 priority window (Provisional 62/562,302 here; 62/562,280 for the '649).

Unusual-pattern note: The signal to watch here is not inventor departure (there is only one inventor and no evidence of departure), but the outside/academic-inventor → operating-company assignment structure. An academic crystal-engineering specialist assigning solid-form patents to a large pharma is a normal sponsored-research/collaboration pattern, not an NPE pattern — but it does mean the entire ownership chain rests on that one executed assignment.

Original assignee

Celgene Corporation, Summit, NJ (Delaware corporation).

• Product embodying the claims: Yes. Pomalidomide is the active ingredient in POMALYST®, approved by FDA in 2013 (in combination with dexamethasone) for relapsed/refractory multiple myeloma. The hemihydrate is a crystalline solid form of that same active ingredient, so the original assignee indisputably commercialized the molecule.
• Primary line of business: Branded biopharmaceuticals (oncology/hematology; also REVLIMID®, OTEZLA®, ABRAXANE®).
• Current status: Operating, but no longer independent. Bristol-Myers Squibb completed its acquisition of Celgene on November 20, 2019, and Celgene became a wholly-owned BMS subsidiary (BMS press release, 2019-11-20; Celgene 8-K dated 2019-11-20, capedge.com/filing/816284/0001104659-19-065939). Celgene was not dissolved in bankruptcy and is not an NPE. Today the family sits within the BMS/Celgene oncology portfolio.

Assignment timeline

Important sourcing caveat: The authoritative patent text provided exposes the fact and date of the recorded assignment but does not print the reel/frame number, and my USPTO Assignment Center queries were cut off before I could retrieve the reel/frame directly. I am therefore not inventing reel/frame identifiers. The chain below is built from (a) the Google Patents legal-events record embedded in the authoritative patent text, (b) the USPTO assignment abstract quoted there, and (c) SEC/news records for the corporate events. Verify the reel/frame at the USPTO Assignment Center (https://assignmentcenter.uspto.gov/) by searching patent number 10093648.

1. 2018-01-29 (recorded) — Reel/frame not retrieved

• Conveyance: Assignment (Assignment of Assignors' Interest)
• Assignor: ATWOOD, JERRY LEE
• Assignee: CELGENE CORPORATION
• Correspondent: Not stated in the record I retrieved. Separately, the firm of record for prosecution of this family is Jones Day (New York, NY) — FreePatentsOnline lists "Attorney, Agent, or Firm: Jones Day (New York, NY, US)" for the sibling continuation publication US 2020/0140408. I cannot confirm Jones Day was the correspondent on the assignment recording itself, so I flag rather than assert it.
• Context: Acquisition of a single inventor's solid-form invention by the operating pharma sponsor — the only assignment in the chain; the patent remained under the original assignee thereafter.
• Execution date: Not exposed in the retrieved record. The Google Patents event says only "2018-01-29 Assigned to CELGENE CORPORATION… Assignors: ATWOOD, JERRY LEE."

No further recorded assignments to the patent appear in the legal-events data I retrieved. In particular, no assignment of US 10,093,648 from Celgene Corporation to Bristol-Myers Squibb Company is reflected in the events I could verify. That is consistent with the deal structure: Celgene survived the merger as a wholly-owned BMS subsidiary, so the patent can remain titled in "Celgene Corporation" without a reassignment. Whether BMS nonetheless recorded a merger/change-of-name instrument is unclear from the sources I retrieved — flagging as an open item, not a finding of "no transfer."

Chain-length finding: This is a one-hop chain (inventor → operating company). There is no cascade of LLC transfers, no licensing-only assignee, and no post-issuance sale. For an NPE/ownership analysis, the interesting activity on this patent is the litigation, not the assignment record.

Timeline diagram

timeline
    title Ownership of US 10093648
    2017 : Provisional 62562302 filed
         : Non-provisional filed by Celgene
    2018 : Atwood assigns to Celgene
         : Patent issues as US 10093648
    2019 : Celgene acquired by Bristol Myers Squibb
    2020 : Infringement claims withdrawn

NPE / troll-pattern signals

Cross-reference worth noting (context, not a signal): Pomalidomide's crystalline-form patents have been a subject of antitrust scrutiny, not NPE scrutiny. The Cigna v. Celgene and related complaints allege the '647/'648/'649 were "sham" filings used to extend Pomalyst exclusivity, and flag that the '648 was "Not OB Listed" on the Orange Book (Complaint, patent table at ¶312). They also attack the chemistry (TGA water loss up to 4.3% vs. a 3.2% theoretical maximum for a hemihydrate, ¶329). Those are patentability/antitrust theories against an operating-company patentee — they are the inverse of an NPE narrative.

Verdict

Operating-company assertion.

The ownership chain is a single hop — inventor Jerry Lee Atwood assigned US 10,093,648 to Celgene Corporation (recorded 2018-01-29, per the Google Patents legal-events/assignment abstract); Celgene was thereafter acquired by Bristol-Myers Squibb on 2019-11-20 and became a wholly-owned subsidiary, with no intervening shell-entity or licensing-only assignee appearing anywhere in the record. The patentee is a product-shipping operating company (POMALYST®/pomalidomide) that sued actual generic competitors (Celgene v. Aurobindo, D.N.J. 2:19-cv-05799 and companion 2019 Hatch-Waxman cases). None of the eight NPE signals tests positive; the recurring controversy over this patent is antitrust/validity (post-2017 crystalline-form filings allegedly used to delay generic entry), which is a competitor-dispute posture, not a troll posture.

Verification links:

• USPTO Assignment Center (search patent 10093648): https://assignmentcenter.uspto.gov/
• Legacy assignment search: https://assignment.uspto.gov/patent/index.html
• Google Patents legal events for this patent: https://patents.google.com/patent/US10093648/en

Explicit uncertainties I am not papering over:

1. Reel/frame numbers were not retrievable in this session — I have deliberately left them blank rather than fabricate them.
2. Execution date of the Atwood→Celgene assignment (as distinct from the 2018-01-29 recording date) is not exposed in the sources retrieved.
3. Assignment correspondent on the recording is unconfirmed; Jones Day is the prosecution firm of record for the family but I have not tied them to the assignment filing.
4. Whether BMS recorded a separate merger/change-of-name instrument transferring title from Celgene Corporation to Bristol-Myers Squibb is unclear; the sources I retrieved show the corporate acquisition but not a corresponding patent-office assignment.
5. The '648's Orange Book listing status is inconsistently described across complaints ("Not OB Listed" in one, patents-in-suit in others), which I could not resolve here.

I'll research the prior art for US 10,093,648 by examining its cited references. Let me run several targeted searches.

I found a critical lead: EP2815749A1 apparently discloses crystalline pomalidomide with XRPD peaks at 12.1, 17.4, 25.6. Let me verify the family relationships and dates of the key references.

I've gathered substantial corroborating material. Note: my tool access for a few follow-up queries was cut off after hitting the step limit, so I'll flag any residual uncertainty explicitly rather than fill gaps by inference.

Prior-Art Analysis — US 10,093,648 B1

Patent under analysis: US 10,093,648 B1, "Crystalline 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione hemihydrate, compositions and methods of use thereof" (Celgene; inventor Jerry L. Atwood).
Effective filing / priority date: September 22, 2017 (provisional 62/562,302); actual filing Dec. 20, 2017.
Cited references reviewed: the 45-item "Citations" list printed on the face of the patent (plus the two "Cited By" items).
Sole independent claim: claim 1 — crystalline pomalidomide hemihydrate having an XRPD pattern with peaks at 12.0, 17.2 and 25.6 °2θ ± 0.2. Claims 2–10 are dependent, adding (2) peaks 13.9/16.7/24.3; (3) FIG. 1 XRPD; (4) DSC endotherm ≈317 °C; (5) FIG. 2 DSC; (6) ≈3.2 % water; (7) TGA loss 2.8–4.3 % (30–225 °C); (8)/(9) FIG. 3/FIG. 4 TGA; (10) FIG. 5 IR.

⚠️ Correction / flag against the earlier summary: the earlier section stated the '648 claims the hemihydrate is "distinguished from the dihydrate (US 10,093,647) and monohydrate (US 10,093,649)". That is consistent with what I retrieved. However, the retrieved Cigna complaint gives the '647 dihydrate priority as 5/26/2017 (not 9/22/2017). That does not affect the '648 analysis (priority 9/22/2017), but the family-priority statement in the earlier section was over-generalized.

1. Legal framework and the date that matters

• § 102(a)(1) art = patented / printed publication / public use / on sale / otherwise available before Sept. 22, 2017.
• § 102(a)(2) art = subject matter "effectively filed" before that date but published/issued later, and only if it is a U.S. patent, a U.S. application publication, or a PCT application designating the U.S. (post-AIA 102(d) "effectively filed" analysis).
• § 102 anticipation requires a single reference disclosing every claim element. For claim 1 that means: (i) the pomalidomide molecule, (ii) a crystalline form, (iii) characterized as the hemihydrate, and (iv) XRPD peaks at 12.0/17.2/25.6 ± 0.2.

Most of the 45 citations are process, formulation, REMS/health-IT, or compound patents that do not reach a crystalline hydrate. Only a handful are solid-form relevant. I have organized them into relevance tiers.

2. Tier 1 — Strongest § 102 candidate

WO 2013/126326 A1 — Solid forms of 4-amino-2-(2,6-dioxopiperidine-3-yl)isoindoline-1,3-dione, compositions and methods of use thereof (Celgene Corporation)

• Publication date: Aug. 29, 2013 | Priority: Feb. 21, 2012 (US prov. 61/601,497) → § 102(a)(1) (printed publication >4 years before the '648 priority).
• Disclosure: Solid forms of pomalidomide, including an amorphous form and a crystalline Form A (characterized by Celgene as "anhydrous"), with full XRPD/DSC/TGA/IR figures and single-crystal data. Critically, the generic disclosure expressly recites "a hydrated form of pomalidomide, including, but not limited to, a hemihydrate, a monohydrate, a dihydrate, a trihydrate, and the like." The related European publication EP 2 815 749 A1 describes "a crystalline pomalidomide … characterized by XRPD peaks located at the following approximate positions: 12.1, 17.4, and 25.6 degrees 2θ."
• § 102 assessment — claim 1 (and 2–3): This is the closest single-reference hit. The three reference peaks (12.1/17.4/25.6) fall within ±0.2° of the claimed 12.0/17.2/25.6, and the reference expressly names the hemihydrate as an embodiment. If Form A is in fact hydrated (as the ANDA defendants argued — see § 6), the reference discloses every element of claim 1 in a single document and anticipates claim 1. This is precisely the position the generics took ("Celgene's Pomalyst product and the API … marketed since 2013—four years prior to the earliest priority date … are thus prior art"). The only genuine dispute is whether the reference's crystalline species is the hemihydrate or a truly anhydrous form; if the latter, anticipation of the hemihydrate is defeated and the reference shifts to a leading § 103 reference.
• Note: the '648 specification itself concedes "An amorphous solid and one crystalline form (anhydrous) have been described in WO 2013/126326," confirming WO '326 was of record and squarely on point.

3. Tier 2 — Same-compound art (discloses the pomalidomide molecule / its preparation; not the crystalline hemihydrate)

None of these recites the claimed hemihydrate's XRPD signature, so none anticipates claim 1; all are § 103/background art.

4. Tier 3 — Analogous polymorph / solid-form art (different active molecule) → § 103 only

5. Tier 4 — Same-compound hydrate/coformer/solid-dispersion art (different solid form)

6. Tier 5 — Citations with no bearing on the claimed solid form

These appear only to satisfy formal/background requisites (compound salts/solvates, REMS/drug-distribution software, health-IT, or unrelated engineering):

• US 4,551,177 A (National Starch, 1985-11-05) — compressible starches as tablet binders
• US 5,385,901 A (Rockefeller, 1995-01-31) — treating abnormal TNFα concentrations
• WO 1996/013790 A1 (1996-05-09) — prescription management system
• US 5,594,637 A (Base Ten, 1997-01-14) — medical risk assessment system
• US 5,593,696 A (McNeil-PPC, 1997-01-14) — famotidine/sucralfate composition
• US 5,619,991 A (Lucent, 1997-04-15) — delivery of medical services via electronic data
• US 5,712,291 A (Children's Medical Center, 1998-01-27) — angiogenesis inhibition
• US 5,731,325 A (Andrulis, 1998-03-24) — melanomas with thalidomide
• WO 1998/013783 A1 (Azron, 1998-04-02) — electronic medical records system
• US 5,832,449 A (Cunningham, 1998-11-03) — pharmaceutical trial-product dispensing
• WO 1999/010829 A1 (Deka, 1999-03-04) — physician order-entry system
• US 5,882,656 A (Merck, 1999-03-16) — bisphosphonate dry-mix formulation
• US 5,974,203 A (Canon, 1999-10-26) — pattern-recognition image communication
• US 6,045,501 A (Celgene, 2000-04-04) — foetal-exposure drug delivery (Thalomid REMS)
• US 6,063,026 A (Carbon Based, 2000-05-16) — medical diagnostic analysis system
• WO 2000/051053 A1 (Gemini Genomics, 2000-08-31) — clinical/diagnostic database
• US 6,131,090 A (Pitney Bowes, 2000-10-10) — controlled access to recorded media
• US 6,202,923 B1 (Innovation Associates, 2001-03-20) — automated pharmacy
• US 6,315,720 B1 (Celgene, 2001-11-13) — drug-delivery/avoid-adverse-side-effect method
• US 2002/0054899 A1 (Zeldis, 2002-05-09) — atherosclerosis/restenosis treatment
• WO 2002/043720 A2 (Sloan-Kettering, 2002-06-06) — temozolomide + thalidomide
• WO 2002/059106 A1 (Celgene, 2002-08-01) — isoindole-imide compounds
• WO 2002/064083 A2 (Children's Med. Ctr., 2002-08-22) — synthesis of 3-amino-thalidomide
• US 6,896,399 B2 (Trinity Industrial, 2005-05-24) — coating-material feeding apparatus
• US 7,189,740 B2 (Celgene, 2007-03-13) — lenalidomide for MDS
• US 2007/0155791 A1 (Zeldis, 2007-07-05) — cutaneous lupus treatment
• US 2008/0051431 A1 (Verhelle, 2008-02-28) — IMiD combination therapy
• US 7,393,862 B2 (Celgene, 2008-07-01) — lenalidomide for leukemias
• US 7,629,360 B2 (Celgene, 2009-12-08) — cachexia / GVHD treatment
• US 7,709,031 B2 (LSU Board of Supervisors, 2010-05-04) — plant-extract angiogenic agents
• US 7,968,569 B2 (Celgene, 2011-06-28) — lenalidomide for multiple myeloma

§ 102 effect: none anticipate any claim — they disclose neither the pomalidomide hemihydrate nor its XRPD signature.

"Cited By" items

• US 10,487,069 B2 (Celgene) — the '648's own continuation; not prior art.
• US 11,053,211 B2 (Hetero Research Foundation, 2021-07-06; priority 2015-10-01; published as US 2017/0088537 A1 on 2017-03-30) — "Process for pomalidomide," disclosing pomalidomide crystalline Form I. Because US 2017/0088537 published before Sept. 22, 2017, it is potential § 102(a)(1) art; however it claims a different crystalline form (Form I) and I could not retrieve its XRPD peak list, so I cannot assess whether it overlaps the claimed hemihydrate peaks. Treat as a § 103 / possible § 102(a)(1) reference pending confirmation of Form I's XRPD.

7. Summary — where § 102 actually bites

8. Caveats and unresolved points

• The single decisive factual question for § 102 is whether WO 2013/126326's disclosed crystalline form is (or inherently is) the hemihydrate. The generic ANDA defendants' Supplemental Invalidity Contentions (June 12, 2020, Celgene v. Aurobindo, D.N.J. 2:19-cv-05799, ECF 62-2) advanced exactly this theory; Celgene characterized WO '326 as "anhydrous." I have not independently resolved the anhydrous vs. hydrate character of Form A.
• I confirmed the family/priority relationships of US 10,093,648 and the points of law above from the patent text and the complaint excerpts retrieved; I could not retrieve the full text and XRPD peak list of WO 2013/126326's Form A or US 2017/0088537's Form I within the tool budget, so the exact peak-by-peak comparison is based on the EP 2 815 749 A1 "12.1, 17.4, 25.6 °2θ" recitation and should be verified against the primary documents.
• No reference in the citation list discloses the '648's dependent-claim data (≈317 °C DSC endotherm; 2.8–4.3 % TGA; FIG. 1–5 patterns); those dependents stand or fall with claim 1's independent-form disclosure.
• I did not auto-correct or substitute any patent number; US 10,093,648 was interpreted literally and all identifiers above are reproduced as retrieved.

I'll research the key prior art references cited on the page before building the obviousness analysis.

Let me gather more detail on the key WO 2013/126326 reference and the pomalidomide process patent.

Obviousness Analysis of US 10,093,648 B1 Under 35 U.S.C. § 103

Scope note. This analyzes claim 1 (the sole independent claim) and the dependent claims that merely recite characterizing data (XRPD/FIG. 1, DSC/FIG. 2, % water, TGA/FIGS. 3–4, IR/FIG. 5). It builds on the prior summary rather than repeating the bibliographic record. The critical-date is the Sept. 22, 2017 provisional priority date, so prior art is everything publicly available before that date. All references below predate it.

1. The invention, stripped to its essence

Claim 1 covers a known chemical compound — pomalidomide — in a particular crystalline hydration state (hemihydrate), defined by three signature XRPD peaks (12.0, 17.2, 25.6 °2θ ±0.2). The dependent claims add only analytical characterization of the same solid form: additional XRPD peaks, a DSC endotherm at ~317 °C, a stated water content (~3.2%), TGA weight-loss ranges, and spectra/thermograms matching the figures. No claim recites a method of making, a formulation, a dosage, or a new therapeutic use.

That framing matters to § 103: the claim is a composition-of-matter claim to a new solid form of an old compound, and its limitations are essentially a fingerprint of the crystal. Under settled law, the recited XRPD/DSC/TGA numbers confer no independent patentability if the solid form itself would have been obvious — the peak positions and thermal events are simply the inherent properties of that form. See In re Best, 562 F.2d 1252 (CCPA 1977) (inherency; product claims obvious where an obvious process necessarily yields the claimed product); In re Merck & Co. (property/characterization recitations do not avoid obviousness of the underlying product).

2. The prior-art arsenal on the face of the patent

The "Citations" section of the page identifies the following references, all of which predate the critical date and are available under § 102(a)(1)/(a)(2):

Two additional facts are themselves prior art and are effectively admitted in the '648 Background:

• Pomalidomide's structure "has been known since at least the 1960s" and "An amorphous solid and one crystalline form (anhydrous) have been described in WO 2013/126326." This is a specification admission of the known compound and known solid forms.
• POMALYST® was approved and marketed (with dexamethasone) in 2013 — four years before the critical date. Its API was therefore a public, prior-art material.

3. The legal framework that governs solid-form obviousness

1. Structural similarity + shared utility ⇒ prima facie obviousness. A claimed compound structurally close to a known one with the same utility is presumptively obvious. In re Papesch, 315 F.2d 381 (CCPA 1963); In re Dillon, 919 F.2d 688 (Fed. Cir. 1990) (en banc). Here the "compound" is identical to the prior-art molecule — only its crystal lattice differs — so the structural-similarity calculus is at its strongest.
2. KSR "obvious to try." Where the prior art identifies a finite number of identified, predictable solutions and a POSA has a reasonable expectation of success, the variation is obvious. KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 421 (2007). For a hydrate, the "finite" set is small and predictable: anhydrate, hemihydrate, monohydrate, dihydrate — selected by water activity in the crystallization solvent.
3. Characterization data are not a shield. Reciting XRPD peaks, DSC maxima, % water, and TGA curves does not make an otherwise-obvious crystal form patentable; those are the inherent properties of the form.
4. Counterweight. The patentee will invoke the line of authority requiring a specific teaching or suggestion to the claimed form, arguing crystal forms are unpredictable. Aventis Pharma Deutschland GmbH v. Lupin Ltd., 499 F.3d 1293 (Fed. Cir. 2007). And it may argue unexpected properties. Both are addressed in § 6.

4. Combinations that render claim 1 obvious

Combination A (primary): WO 2013/126326 + US 7,465,800 (+ Caira)

Why the combination is motivated:

• WO 2013/126326 and US 7,465,800 are commonly owned by Celgene and directed to solid forms of the same class of immunomodulatory imide drugs — the same field, same problem (a stable crystalline form of a pharmaceutically active glutarimide). Common ownership and common problem supply strong, articulable motivation to combine.
• WO 2013/126326 disclosed only the amorphous and one anhydrous crystalline form of pomalidomide. A POSA looking to improve the physical/chemical stability and manufacturability of that molecule — precisely the rationale the '648 itself recites — would naturally look to the hydrate family, because hydrates are known to be more thermally and physically robust than anhydrates for many oral drugs.
• US 7,465,800 supplies the specific template. It teaches that the closest structural analogue, lenalidomide — same isoindoline-dione/glutarimide pharmacophore, differing only in the 1-oxo vs. 1,3-dioxo ring — forms a crystalline hemihydrate (Form B) that is pharmaceutically acceptable and is claimed as such (claim 10: ~0.46–0.59 mol water). That is a direct, concrete teaching that the hemihydrate end-member of this family is a real, obtainable, useful form.
• WO 2013/126326's DVS isotherms (Figs. 23–24) further signal that pomalidomide takes up/releases water — i.e., it is a hydrate-former — motivating the POSA to prepare and characterize its hydrates.
• Caira closes the loop by teaching that systematic polymorph/hydrate screening (crystallizing from a matrix of solvent/antisolvent/water systems, then characterizing by XRPD/DSC/TGA) is routine and expected for a crystalline drug candidate.

Why there is a reasonable expectation of success (KSR): the reference art discloses the routine technique (aqueous/organic recrystallization) and the specific target stoichiometry (hemihydrate), and the number of hydrate possibilities is small and governed by well-understood water-activity principles. The '648's own Examples confirm this: every successful preparation is a plain recrystallization of anhydrous pomalidomide from an aqueous-organic solvent — 12:1 or 6:1 THF/water at 80 °C/50 °C; 1:1 ethanol/water at 80 °C; 1:1, 4:5, and 3:1 1,4-dioxane/water; 2:4:2 ethanol/THF/water; 5:1 ethanol/THF. These are textbook hydrate-screening conditions, not an inventive departure.

Combination B: US 5,635,517 + WO 2013/126326 + US 7,465,800

The '517 patent supplies the compound, its synthesis, and its utility — establishing that the pomalidomide molecule and its therapeutic use were fully known. Layering WO 2013/126326 (known solid forms) and the '800 patent (hemihydrate analogue) onto that baseline yields the claim. Because the '517 compound and the claimed hemihydrate have identical utilities, the structural-similarity presumption (Papesch) applies with maximum force and is not rebutted by any new activity.

Combination C: Marketed POMALYST® API + WO 2013/126326 + US 7,465,800

The generic defendants in the underlying Hatch-Waxman litigation expressly 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." (Celgene v. Aurobindo, 19-cv-05799, Revised Joint Proposed Discovery Plan, Aug. 19, 2019.) Combining that marketed, characterized material with the '326 and '800 teachings provides yet another route to the claim: it is the known drug, screened for its hydrate forms by known methods.

Combination D: WO 2013/126326 + WO 2014/160690 + US 7,994,327 + WO 2013/012485

Where the patentee emphasizes unpredictability, the record cuts the other way: Celgene's own WO 2014/160690 (pomalidomide cocrystals) and Amplio's WO 2013/012485 (lenalidomide cocrystals), plus the '327 process patent, show that by the critical date the field had a mature, routine toolkit for generating and characterizing many solid forms of these exact molecules. Combined with the '326 and '800 teachings, they establish that arriving at any given hydrate (including the hemihydrate) required only routine experimentation within a predictable design space.

5. Claim-by-claim mapping

No dependent claim adds a method step, a formulation, a dose, or a property (e.g., a demonstrated stability or bioavailability advantage) that could supply patentable weight. Cf. In re Merck (recited physical properties do not create patentability).

6. The patentee's best rebuttals — and why they are contestable

To be balanced, the following arguments would be pressed, and a fact-finder could credit them:

1. "Polymorphs are unpredictable; there is no specific teaching of this form." Aventis v. Lupin teaches that where the prior art does not teach the specific polymorph and the disclosed process does not inherently yield it, a new form can be nonobvious. Here, however, US 7,465,800 does not merely suggest hydrates generally — it claims a hemihydrate of the closest analogue, which is far more concrete than the art in Aventis. That distinction weakens the rebuttal.
2. Unexpected properties. If the patentee could show the hemihydrate has a surprising stability, solubility, or bioavailability advantage not predictable from the '326/‘800 disclosures, that could rebut the prima facie case. The '648, however, reports no comparative or unexpected-property data — only characterization. The patent therefore lacks the evidentiary hook this argument requires.
3. Enablement/written-description attacks (not § 103). Generic defendants argued the TGA data are internally inconsistent: for the hemihydrate the maximum theoretical water loss is 3.2%, yet the '648 specification reports losses up to 4.3% (and the sibling monohydrate patent up to 7.4%). (The Cigna complaint, ¶¶ 328–331; Celgene v. Aurobindo discovery plan.) These are primarily § 112 attacks, but they are double-edged for § 103: they undermine the patentee's ability to rely on the reported properties to prove the form's existence or its unexpected character, and they reinforce that the "form" may be a trivially obtained partial hydrate rather than anything inventive.
4. Procedural reality. Celgene withdrew infringement claims as to many claims of the crystalline-form patents on October 12, 2020 and settled with ANDA filers, and the patents face antitrust "sham" allegations — conduct that at least suggests the patentee did not view these claims as robust enough to test on the merits. (This is litigation context, not a § 103 holding.)

7. Bottom line

Claim 1, and dependent claims 2–10, are strongly vulnerable to a § 103 challenge. The claimed subject matter is a new crystalline hydration state of a compound that was public for decades, differing from the closest prior art only in the amount of water in the lattice. The primary combination —

WO 2013/126326 (known amorphous and anhydrous solid forms of pomalidomide) + US 7,465,800 (crystalline hemihydrate of the closest analogue, lenalidomide) + Caira (routine nature of polymorph/hydrate screening) —

supplied (i) the compound, (ii) the specific hemihydrate target and its pharmaceutical desirability, and (iii) the routine means (aqueous/organic recrystallization) and a small, predictable set of hydrate end-points. Under KSR's "obvious to try" standard, a POSA would have had both motivation and a reasonable expectation of success to make and characterize the hemihydrate. The recited XRPD peaks, DSC maximum, water percentage, TGA curves, and IR spectrum are inherent properties of that form and add no patentable weight. Combinations B–D reinforce the case from the compound-art, marketed-product, and routine-screening directions, respectively.

Caveats and uncertainties I will not paper over:

• I have not located a final, on-the-merits adjudication of the '648's validity; the identified Federal Circuit dispositions were venue-based (No. 2021-1154), and the antitrust "sham" allegations concern objective baselessness, not a § 103 ruling.
• An Aventis-style nonobviousness outcome remains theoretically available if a specific, credible, unexpected-property showing were made — but no such showing appears in the '648 specification, and the TGA inconsistencies cut against the patentee.
• My description of US 7,465,800's Form B as a hemihydrate rests on the Markman record (claim 10 reciting ~0.46–0.59 mol water; the parties' "hemihydrate" constructions). I did not independently retrieve the '800 patent's full text; if its Form B disclosure proves narrower than described, the strength of Combining Reference A would diminish accordingly, and I flag that as unverified.

(Primary sources: https://patents.google.com/patent/US10093648/en ; EP2815749A1 / WO 2013/126326 text at https://patentimages.storage.googleapis.com/23/95/1b/466400a4b394c4/EP2815749A1.pdf ; CourtListener Markman briefs (D.N.J. 2:10-cv-05197) at https://storage.courtlistener.com/recap/gov.uscourts.njd.[247596](/patent/247596).249.0.pdf and .../252.0.pdf ; The Cigna Group v. Celgene Corp. complaint at https://business.cch.com/ald/TheCignaGroupvCelgeneCorporationComplaint.pdf ; Celgene v. Aurobindo 19-cv-05799 docket materials at https://www.docketalarm.com/cases/New_Jersey_District_Court/2--19-cv-05799/.)