US 8404717

I'll run targeted searches for this specific patent number.

Let me check the claim set and any recent 2026 litigation/appeal activity.

Let me do a couple more targeted checks for any 2026 appellate activity and the PTAB outcome.

Summary — US Patent 8,404,717 (US 8,404,717 B2)

Note on sourcing: The full patent text supplied to me (from Google Patents, fetched 2026‑10‑01) contains the front matter, abstract, and specification but the text cut off before the claims. Claim language below is drawn from the PTAB petition in IPR2018‑01507 (which quotes claim 1 verbatim) and from DrugPatentWatch's claim listing — two independent sources that agree. Treat the claim wording as well‑corroborated but verify against the printed patent if precision matters.

Bibliographic data

Priority chain (per the cross‑reference section): provisional 60/418,468 (Oct. 15, 2002) → 10/411,649 (Apr. 11, 2003, now US 7,189,740) → 11/654,550 (Jan. 16, 2007, now US 7,393,863) → 11/985,032 (Nov. 12, 2007, now US 7,863,297) → 12/777,765 (May 11, 2010, now US 8,404,716) → 13/070,761 (this patent, filed Mar. 24, 2011).

Abstract (as granted)

"Methods of treating, preventing and/or managing myelodysplastic syndromes are disclosed. Specific methods encompass the administrations of 3-(4-amino-1-oxo-1,3-dihydro-isoindol-2-yl)-piperidin-2,6-dione in combination with 5-azacytidine."

Independent claim 1 — plain language

"1. A method of treating a patient having transfusion dependent anemia due to low to intermediate‑1‑risk myelodysplastic syndrome, which comprises administering to said patient in need thereof about 5 to about 25 mg per day of 3‑(4‑amino‑1‑oxo‑1,3‑dihydro‑isoindol‑2‑yl)‑piperidine‑2,6‑dione having the formula: [structure] or a pharmaceutically acceptable salt, solvate or stereoisomer thereof."

In plain terms, claim 1 covers treating a specific MDS sub‑population — patients whose anemia is transfusion‑dependent and whose disease is low‑ or intermediate‑1‑risk by IPSS — by giving lenalidomide alone, orally or otherwise, at 5–25 mg/day (or a salt/solvate/stereoisomer).

Important nuance / apparent inconsistency: The abstract and specification are drafted around combination therapy with 5‑azacytidine, but the issued independent claim is a lenalidomide monotherapy claim with no azacitidine limitation. The azacitidine combination claims appear to reside in the sibling patent US 8,404,716 (titled "…with a combination therapy using lenalidomide and azacitidine"). This divergence between the abstract and the granted claim scope is intrinsic to the document and is worth flagging.

Dependent claims (2–10)

All depend from claim 1 and narrow only the dose and form:

• 5 mg/day (claim 2); 5 mg as an oral capsule (claim 3)
• 10 mg/day (claim 4); 10 mg as an oral capsule (claim 5)
• 15 mg/day (claim 6); 15 mg as an oral capsule (claim 7)
• 25 mg/day (claim 8); 25 mg as an oral capsule (claim 9)
• Claim 10: the compound of claims 3, 5, 7 or 9 in free‑base form.

So the claim set is effectively one genus (5–25 mg/day lenalidomide for TDA in low/Int‑1‑risk MDS) plus four dose‑specific species and dosage‑form/form‑of‑matter narrowings.

Prosecution history (as reflected in PTAB record)

• Preliminary Amendment filed Mar. 24, 2011 (Patent Owner Ex. 2007)
• Office Action May 9, 2012 (Ex. 2009); response Sept. 10, 2012 (Ex. 2010)
• Notice of Allowance Jan. 17, 2013 (Ex. 2011)

PTAB / litigation activity found

• IPR2018‑01507 — [Dr. Reddy's Laboratories, Inc.](/litigations/by-plaintiff/Dr.%20Reddy's%20Laboratories%2C%20Inc.) v. Celgene Corp., filed Aug. 3, 2018, on US 8,404,717 (application 13/070,761). Google Patents' litigation record labels this proceeding "Not Instituted – Merits." Sibling petitions by the same petitioner targeted US 9,056,120 (IPR2018‑01504) and US 7,189,740 (IPR2018‑01509).
• District court (D.N.J. unless noted): 2:10‑cv‑05197 (Natco/Arrow/Watson); 2:17‑cv‑05314 (Dr. Reddy's); 2:17‑cv‑06842 (Lotus/Alvogen); 2:18‑cv‑11518; 2:19‑cv‑06999; 2:19‑cv‑14731; 2:19‑cv‑15449; 2:20‑cv‑07759; 2:20‑cv‑14389; 2:21‑cv‑10398; 2:21‑cv‑11261; 2:21‑cv‑12927; and N.D. W. Va. 1:20‑cv‑00003 (Mylan).
• The '717 patent also appears in the In re Revlimid and Thalomid Purchaser Antitrust Litigation (D.N.J. 2:19‑cv‑07532) and in a 2026 Bristol‑Myers Squibb/Celgene antitrust complaint in D.N.J. as one of the listed Revlimid method‑of‑use patents.

CAFC 2026 docket check — explicit uncertainty

I searched for Federal Circuit activity in 2026 tied to 8404717 / 8,404,717 and found no 2026 CAFC appeal or docket entry involving this patent. What surfaced instead were unrelated 2026 Federal Circuit materials (a Rule 40(c) en banc petition in the Woodway/LifeCORE treadmill case, 25‑1431, and commentary on Google's mandamus petition regarding VirtaMove's US 7,519,814). I cannot treat a negative search as authoritative: my keyword search is not equivalent to a complete Docket Alarm / PACER / CAFC docket query, and the patent expired in 2023, which makes a 2026 appeal unlikely but not impossible. If you need a definitive docket-level confirmation, the reliable path is a direct Docket Alarm or PACER/CM-ECF query by patent number and by the party names (Celgene Corp. / Bristol‑Myers Squibb Co.).

Bottom line: US 8,404,717 is Jerome Zeldis's Celgene method‑of‑use patent for lenalidomide monotherapy at 5–25 mg/day in transfusion‑dependent, low/Int‑1‑risk MDS. Filed 2011 as a continuation, issued Mar. 26, 2013, expired (fee‑related) with an anticipated expiration of Apr. 11, 2023. It was asserted broadly in the Revlimid Paragraph IV litigation wave, was the subject of Dr. Reddy's IPR2018‑01507 (not instituted on the merits), and I found no live 2026 appellate proceeding for it.

I'll research the litigation history for US patent 8404717. Let me search multiple sources.

Let me search for the specific case numbers associated with this patent.

Litigation Involving U.S. Patent No. 8,404,717

Bottom line: U.S. 8,404,717 (the "'717 patent," "Methods of treating myelodysplastic syndromes using lenalidomide," Celgene Corp.) has been asserted in multiple Hatch‑Waxman (ANDA) patent‑infringement suits brought by Celgene against generic lenalidomide manufacturers, all in the District of New Jersey, plus one PTAB inter partes review. Every district-court action I could confirm resolved by settlement/consent judgment — I found no judgment on the merits of the '717 patent itself in the district courts. Below is what the sources support, with an explicit flag where I could not confirm a mapping.

How the '717 patent entered litigation

The '717 patent issued Jan. 17, 2013 (application 13/070,761, filed Mar. 24, 2011; priority chain back to U.S. Provisional 60/418,468, filed Oct. 15, 2002). Celgene caused it to be listed in the Orange Book for REVLIMID on April 10, 2013 and first asserted it by Fifth Amended Complaint in the Natco case on May 6, 2013. (Sources: multiple complaints in the Revlimid antitrust MDL, quoting Celgene's litigations.)

District Court Cases (D.N.J. unless noted)

PTAB Proceeding

• IPR2018-01507 — Dr. Reddy's Laboratories, Inc. et al. v. Celgene Corp. (P.T.A.B.), filed Aug. 3, 2018, challenging the '717 patent. Status: Institution Denied (merits) — institution decision dated Feb. 11, 2019; refund notices filed 2019. Related Dr. Reddy's IPRs on sibling Revlimid patents include IPR2018-01504 and IPR2018-01509.

Additional cases listed on the patent's litigation docket (parties not fully confirmed)

The Google Patents "Family has litigation" panel for US 8,404,717 also lists these United States proceedings, which I could not individually map to party names within the sources retrieved:

• D.N.J.: 2:19-cv-06999, 2:19-cv-14731, 2:19-cv-15449, 2:20-cv-07759, 2:20-cv-14389, 2:21-cv-10398, 2:21-cv-11261, 2:21-cv-12927, 2:21-cv-20459
• West Virginia Northern District Court: 1:20-cv-00003
• A "first worldwide family litigation" entry (Darts‑IP)
• The patent is also flagged in the IPR2018-01507 docket ("Not Instituted – Merits")

Given the naming pattern of the confirmed New Jersey cases, these unconfirmed D.N.J. numbers are consistent with additional Celgene v. [generic manufacturer] ANDA suits on Revlimid, but I am not asserting specific plaintiff/defendant pairings for them because I did not verify them.

Related (non-'717) antitrust litigation that discusses the '717 patent

The '717 patent is extensively discussed — though not asserted as the patent‑in‑suit — in the Revlimid antitrust litigation challenging Celgene's patent‑enforcement and settlement conduct:

• In re Revlimid & Thalomid Purchaser Antitrust Litigation, 2:19-cv-07532 (D.N.J.) — purchaser/insurer class actions alleging sham litigation and pay‑for‑delay.
• Cigna Corp. v. Celgene Corp. / Bristol‑Myers Squibb, D.N.J. (amended complaint May 27, 2022).
• Mayo Clinic v. Celgene Corp., 3:23-cv-05144 (N.D. Cal.) — complaint filed Oct. 6, 2023.
• Various later antitrust complaints (e.g., BMS/Celgene antitrust complaint, D.N.J., 2026).

In those pleadings, plaintiffs allege Celgene's '717 and '740 "method‑of‑treatment" patents (both expiring April 11, 2023) were asserted as part of a "sham litigation" strategy. This is context, not a separate '717 infringement action.

Caveats

1. No merits judgment on the '717 patent: All confirmed district-court cases ended in settlement or consent judgment; the sole PTAB challenge (IPR2018-01507) was denied institution. This is significant, because Celgene never had the validity or infringement of the '717 patent adjudicated.
2. Case-number mapping: I confirmed party/outcome data for 2:10-cv-05197, 2:17-cv-05314, 2:17-cv-06842, and 2:18-cv-11518, and confirmed the existence of the Hetero, Cipla, and Dr. Reddy's disputes. For the remaining D.N.J. and W.D. Va. case numbers listed on the patent page, I could not verify the parties within the retrieved sources and have flagged rather than guessed.
3. Where live sources conflict with my recollections, I have followed the sources (e.g., the Fifth Amended Complaint date of May 6, 2013 for the '717 assertion against Natco).

If you want, I can attempt targeted lookups for the specific unidentified docket numbers (2:19-cv-14731, 2:20-cv-14389, 1:20-cv-00003, 2:21-cv-10398, etc.) on PACER/CourtListener to nail down the parties.

I'll start with the canonical structured data and then verify/fill in the one PTAB reference visible in the patent record.

Now let me verify whether any other IPRs exist on this patent, and get the denial reasoning.

Let me pull the institution decision's reasoning and check for any later-filed IPR on this patent.

Proceedings overview

Total AIA trial proceedings on file for US 8,404,717: one (1) — IPR2018-01507, which was terminated at the institution stage. Breakdown: 0 active, 0 claims invalidated, 0 claims sustained on the merits, 0 settled, 1 institution denied. No Final Written Decision has ever issued on any claim of this patent. Defensive bottom line: the patent is un-narrowed — claims 1–10 were never canceled, never amended, and never even placed on trial — but the patent itself expired on 2023-04-11 and Google Patents records it as "Expired – Fee Related," so the practical posture today is that a defendant faces an expired, unadjudicated method patent rather than a hardened one. This is the opposite of the "claims 1–5 canceled" scenario: there is no claim-level invalidity ruling you can point to, and no FWD to quote.

Data-integrity flag (required): The structured "PTAB proceedings on file" block states the USPTO ODP API returns no AIA trial proceedings for this patent. That is wrong/incomplete. The patent's own Google Patents record lists "PTAB case IPR2018-01507 filed (Not Instituted – Merits)" — https://patents.google.com/patent/US8404717/en — and the proceeding is independently confirmed below. Treat the ODP zero-count as a coverage gap, not as "no PTAB activity."

IPR2018-01507 — [Dr. Reddy's Laboratories, Inc.](/litigations/by-plaintiff/Dr.%20Reddy's%20Laboratories%2C%20Inc.) v. Celgene Corporation

• Type: Inter Partes Review (35 U.S.C. §§ 311–319; pre-AIA § 103 applies to the challenged claims)
• Filed: 2018-08-03 (accorded filing date 2018-08-03)
• Status: "Institution Denied" (per Patexia/IPVerse case records); Google Patents renders the same outcome as "Not Instituted - Merits." Plain English: the Board never instituted a trial — no FWD, no cancellation, no amendment.
• Judge panel: Christopher G. Paulraj, Grace Karaffa Obermann, and Tina E. Hulse. Obermann is listed as the authoring judge. Tech Center 1600, Art Unit 1618.
• Petition grounds (claims 1–10, all challenged):
  • Ground 1 (§ 103, pre-AIA): claims 1–10 obvious over List 2001 in view of the '230 patent (U.S. Pat. No. 6,281,230, Muller et al.) and Celgene Press Releases dated 5/8/2001 and 8/28/2001.
  • Ground 2 (§ 103, pre-AIA): claims 1–10 obvious over Thomas 2000a in view of the '230 patent and the same two Celgene press releases.
  • No § 102 anticipation ground and no § 112 ground was asserted. Petitioner relied on the declaration of Dr. Levin (Ex. 1003). Source: Petition for Inter Partes Review, IPR2018-01507, https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1516997](/patent/1516997)/download-documents (Petition and Patent Owner Preliminary Response).
• Institution decision: DENIED — 2019-02-11 (Paper 7). The Board held Petitioner failed to make the threshold showing required for institution, because it did not establish that the two Celgene press releases on which both grounds depended were publicly accessible printed publications under § 311(b)/§ 102. The panel opened by confirming no claim term needed construction (Petitioner and Patent Owner agreed the terms carry their plain and ordinary meaning), then went straight to the public-accessibility question under Kyocera v. ITC, Panduit, and Dynamic Drinkware, citing Frontier Therapeutics (IPR2016-00649), Instradent USA v. Nobel Biocare (IPR2015-01786), Symantec v. Columbia (IPR2015-00371), and Temporal Power for the proposition that a petitioner must make a threshold showing of public accessibility before the Board reaches the merits. The Board's reasoning at page 10 of Paper 7 is cited in later PTAB briefing for the rule that a document's copyright notice and the publisher's general reputation do not cure a failure to prove public accessibility. (I could not retrieve the complete verbatim text of Paper 7 — only the excerpted portions above — so I am not characterizing any additional grounds for the denial beyond what the excerpts support.)
  • Patent Owner (Celgene, represented by Jones Day / Quinn Emanuel / in-house) had also argued discretionary denial under § 314(a) (petition filed late relative to the advanced D.N.J. litigation; institution would upset the Hatch-Waxman balance), denial for lack of specificity under § 42.104(b), and denial under § 325(d) (substantially the same art and arguments already before the Office during prosecution of the '740, '717, and '120 patents). Those arguments were briefed; the public-accessibility defect is the ground the panel is publicly cited for reaching.
  • Confirmatory source: Jones Day engagement summary — "the PTAB denied institution of inter partes review for all three patents, finding that Dr. Reddy's Laboratories, Inc. had not shown a reasonable likelihood of prevailing in its unpatentability challenges" — https://www.jonesday.com/en/practices/experience/2019/02/celgene-successfully-blocks-institution-of-eminter
• Final Written Decision: None. Because institution was denied, there is no FWD and therefore no claim-level verdict — claims 1–10 all remain as issued. Do not let anyone tell you a claim of the '717 patent was canceled; the record says otherwise. For reference, the claims that were challenged (and left standing) are:
  • Claim 1 (independent): "A method of treating a patient having transfusion dependent anemia due to low to intermediate-1-risk myelodysplastic syndrome, which comprises administering to said patient in need thereof about 5 to about 25 mg per day of 3-(4-amino-1-oxo-1,3-dihydro-isoindol-2-yl)-piperidine-2,6-dione having the formula: or a pharmaceutically acceptable salt, solvate or stereoisomer thereof."
  • Claims 2–9 (dependent): dose-specific (5, 10, 15, 25 mg/day) and capsule-form limitations.
  • Claim 10 (dependent on claim 3, 5, 7 or 9): free-base limitation.
• Settlement / termination: None. This proceeding ended by Board denial of institution, not by settlement or adverse judgment. No termination-on-settlement on the docket.
• Appeal: No Federal Circuit appeal found, and none is available as of right — a denial of institution is not appealable under 35 U.S.C. § 314(d) (Cuozzo Speed Techs. v. Lee, 579 U.S. 261 (2016)). I found no CAFC docket or CourtListener entry for this proceeding.
• Defensive value: Limited but not zero. Because no FWD issued, no statutory estoppel under § 315(e)(2) attached — the art Dr. Reddy's used (List 2001, Thomas 2000a, the '230 patent, the Celgene press releases) is not estopped, and a different defendant could in principle re-run those theories. But the denial turned on an evidentiary defect (public accessibility of press releases) rather than the merits of the obviousness case, so the Board never blessed the '717 claims as nonobvious; a future petitioner who plugs that gap with proper dissemination declarations is not foreclosed by the decision. Practically, the patent expired 2023-04-11, so the denial's main value is defensive history: it rebuts any "Celgene's MDS patents were PTAB-hardened" narrative as to this patent.

Related-family proceedings (NOT on US 8,404,717 — flagged for context only)

Same petitioner, same day, same family and same art theory: IPR2018-01504 (U.S. Pat. No. 9,056,120) and IPR2018-01509 (U.S. Pat. No. 7,189,740), both also denied institution on 2019-02-11 — see Celgene's Patent Owner Mandatory Notices in IPR2018-01507, which states: "The '717 patent … is not at issue in any other inter partes review or inter partes reexamination proceedings," and identifying the '-01504 and '-01509 petitions against the '120 and '740 patents. Do not attribute any '+740 / '120 outcome to the '717 patent. Jones Day's summary lumps the three denials together: "IPR Nos. 2018-01504, -01507, -01509 (PTAB)."

Strategic summary

Claim status: nothing was canceled, nothing narrowed, nothing tested. All ten claims of US 8,404,717 — independent claim 1 plus dependent claims 2–10 — survived IPR2018-01507 untouched because the Board never instituted. There are no CANCELED claims and no "SUSTAINED" claims in the § 318(a) sense; the correct label is UNTESTED. Any representation that a PTAB panel "upheld" or "confirmed" the validity of claims 1–10 overstates the record — an institution denial is a threshold determination that the petitioner failed to meet its § 314(a) burden, not a merits adjudication. (Celgene's own litigation-side brief in the D.N.J. antitrust matter does characterize the result as "in all five proceedings the PTAB ruled in Celgene's favor" — https://storage.courtlistener.com/recap/gov.uscourts.njd.[399896](/patent/399896)/gov.uscourts.njd.399896.104.1.pdf — but that is a party's advocacy characterization spanning multiple patents and should be treated as such.)

Estoppel landscape: essentially empty, which cuts both ways. Because no Final Written Decision issued, no § 315(e)(1)/§ 315(e)(2) estoppel attached to Dr. Reddy's or its privies with respect to this patent. A defendant is therefore not blocked by estoppel from raising List 2001, Thomas 2000a, the '230 patent, or Celgene's 2001 press releases in district court — though the press-release theory carries the same public-accessibility vulnerability the Board identified at Paper 7 p.10 (a publisher's reputation and a copyright notice prove nothing about dissemination). Since the '717 patent expired 2023-04-11, the practical question is not which grounds remain available but whether any damages period still exists at all.

Pattern signals. (1) Same petitioner, three petitions, one day: Dr. Reddy's filed IPR2018-01504, -01507, and -01509 on 2018-08-03 against the '120, '717, and '740 patents respectively — a coordinated generic-defendant attack on the whole '740 patent family, all three denied on 2019-02-11. (2) No defensive aggregator visible here: the petitioner of record is Dr. Reddy's Laboratories, Inc. (a commercial ANDA filer), not Unified Patents; Unified's involvement with this patent appears only as a litigation-data aggregator on the Google Patents page, not as an IPR petitioner. (3) Patent owner appeals: none — there was nothing to appeal, since Celgene won the institution decision. (4) Parallel litigation: the '717 patent was asserted in Celgene v. Dr. Reddy's Labs., No. 2:17-cv-05314 (D.N.J., filed 2017-07-20) and Celgene v. Lotus Pharm., No. 2:17-cv-06842 (D.N.J., filed 2017-09-06), and previously in the long-running Celgene v. Natco, No. 2:10-cv-05197 (D.N.J., filed 2010-10-08) — that earlier case settled (Natco/Celgene settlement announced 2019-03-30 per Taiwanese disclosure reporting), which is a district-court settlement, not a PTAB one. The several 2:18-, 2:19-, 2:20-, and 2:21-cv D.N.J. matters listed on the Google Patents page are unrelated antitrust/pay-for-delay-style suits against Celgene/BMS, not infringement actions on the '717 patent.

Recommended next steps

1. If you're a defendant and a demand letter cites US 8,404,717: lead with expiration, not invalidity. The patent expired 2023-04-11 (20 years from the 2003-04-11 filing of parent application 10/411,649), and Google Patents lists the legal status as "Expired – Fee Related." There is no FWD to cite, so your strongest position is temporal: no enforceable term remains for any post-2023 activity. Verify current fee/expiration status directly in USPTO Patent Center for US 13/070,761 before relying on it in correspondence.
2. Do not over-claim the PTAB win as an invalidity finding. If you plan to argue the claims are invalid, you cannot rest on IPR2018-01507. The only thing it establishes is that Dr. Reddy's failed a threshold evidentiary showing — and the Board never reached obviousness. A district-court § 103 case on List 2001/Thomas 2000a/'230 is not estopped and is available, but you'd be litigating it fresh.
3. Check the family-wide picture before asserting a "hardened patent" or "invalid patent" theory. The '717 patent's three sibling method patents ('740, '120) and the compound patent family were all attacked; the same-day trio of D.R.L. petitions (‑01504, ‑01507, ‑01509) all failed at institution on 2019-02-11. If your matter involves the '740/'120 patents too, pull those institution decisions separately — they are separate proceedings with separate records and this memo does not cover them.
4. No upcoming PTAB milestones to track. There are no active proceedings, so there is no institution deadline, no oral hearing, and no statutory 1-year FWD deadline running against this patent. The only live question is whether any assertion survives the 2023 expiration.
5. Primary-source verification pointers. The IPR2018-01507 file (Petition, Patent Owner Preliminary Response filed 2018-11-30, Mandatory Notices, Power of Attorney, and Paper 7 dated 2019-02-11) is available via USPTO PTAB E2E — https://ptacts.uspto.gov/ptacts/ — with the petition materials at https://ptacts.uspto.gov/ptacts/public-informations/petitions/1516997/download-documents. A third-party mirror of the full document set is at https://www.docketalarm.com/cases/PTAB/IPR2018-01507/Inter_Partes_Review_of_U.S._Pat._8404717/. Case metadata confirming "Institution Denied," the 2019-02-11 decision date, and claims 1–10 challenged: https://services.patexia.com/lawsuits/Dr-Reddy%27s-Laboratories-Inc-et-al-v-Celgene-Corporation-id-[116423](/patent/116423) and https://ipverse.greyb.com/ptab-web/cases/case-details/IPR2018-01507.

I'll research the assignment record and prosecution/ownership history for US 8404717.

Found the key recorded assignment. Let me dig deeper on the correspondent of record and any later transfers.

Ownership & Assignment Analysis — US Patent 8,404,717 B2

Patent: US 8,404,717 B2 — "Methods of treating myelodysplastic syndromes using lenalidomide"
Application: 13/070,761 (filed 2011-03-24) · Issued: 2013-03-26 · Priority: 2002-10-15 (prov. 60/418,468) → 2003-04-11 (parent 10/411,649)
Google Patents: https://patents.google.com/patent/US8404717/en
Assignment Center lookup: https://assignmentcenter.uspto.gov/ (search by patent number 8404717) — also https://assignment.uspto.gov/patent/index.html

Inventors

Pattern note: No "inventor exodus" pattern. This is a single inventor who was a senior employee of the assignee at filing, and who signed an employee invention assignment the same day the application was filed. The '717 patent is a downstream continuation in a family Celgene had prosecuted since 2002–2003 (chain: 60/418,468 → 10/411,649 → 11/654,550 → 11/985,032 → 12/777,765 → 13/070,761, per the patent's own cross-reference section). The same-day assignment is a routine practice of recording inventor→company title each time a new continuation is filed, not a departure/fire-sale indicator.

Original assignee

Celgene Corporation, a Delaware corporation, 86 Morris Avenue, Summit, New Jersey 07901 (as recited verbatim in Celgene's Patent Owner Mandatory Notice in IPR2018‑01507).

• Business: Brand pharmaceutical manufacturer (hematology/oncology; immunomodulators). Celgene is not a licensing vehicle.
• Product embodying the claims — yes. Revlimid® (lenalidomide) capsules, NDA 021880, approved 2005‑12‑27. US 8,404,717 was listed in the FDA Orange Book against Revlimid with use code U‑1215 (method of treating transfusion-dependent anemia due to low-/intermediate‑1‑risk MDS), with a listed expiration of 2023‑04‑11 (41st Annual Orange Book). Celgene also listed the '717 patent in the Orange Book during the Natco litigation.
• Assertion history: Celgene (plaintiff) asserted the '717 patent against ANDA generic filers — its actual competitors — in D.N.J. and N.D. W. Va. (see timeline). Dr. Reddy's Laboratories petitioned for IPR against it, and institution was denied on 2019‑02‑11 (IPR2018‑01507).
• Current status: Acquired. Bristol‑Myers Squibb completed its acquisition of Celgene on 2019‑11‑20; Celgene continues as a wholly owned BMS subsidiary. The '717 patent reached its 20‑year term on 2023‑04‑11 and is now of no commercial effect; Google Patents records its status as "Expired – Fee Related." Celgene and BMS are currently defendants in In re Revlimid & Thalomid Purchaser Antitrust Litigation (D.N.J. 3:23‑cv‑05144, filed 2023‑10‑06), in which plaintiffs (incl. Mayo Clinic) allege sham/reverse-payment settlements tied to these same lenalidomide patents.

Assignment timeline

The USPTO Assignment Center record for US 8,404,717 contains exactly one recorded assignment (this is corroborated by the single "reassignment" event on Google Patents Legal Events, and by the reel/frame cited in Celgene's PTAB filing).

• Executed 2011‑03‑24 / recorded 2011‑03‑24 — Reel 026014 / Frame 0756
  • Conveyance: Assignment (Assignment of Assignors' Interest)
  • Assignor: Jerome B. Zeldis (inventor)
  • Assignee: Celgene Corporation (Delaware; 86 Morris Avenue, Summit, NJ 07901)
  • Correspondent: Not verifiable from the sources available to me. The IPR2018‑01507 mandatory notice cites only "Reel 026014, Frame 0756" and does not name the recording correspondent, and the Assignment Center entry itself is not reproduced in any indexed secondary source I could reach. I will not guess an attorney or firm name. (Flagging as an open item: the correspondent field should be pulled directly from the Assignment Center record — note that large-cap pharma inventor assignments of this type are typically filed by in-house IP paralegals or the prosecution firm of record, not by an NPE-side filing attorney; with only one link in the chain the "repeat correspondent" tell cannot be evaluated in any event.)
  • Context: Routine employee/inventor title transfer to the operating company, recorded the same day the continuation application was filed — not an acquisition, securitization, or transfer-to-asserter.

No other post-issuance assignments recorded. In particular: there is no recorded assignment to Bristol‑Myers Squibb, to a Celgene IP-holding subsidiary, or to any LLC. The BMS/Celgene combination was a stock merger, so Celgene Corporation remains the record owner of the patent (title does not flow through a recorded patent assignment in a share purchase / merger of that type). Whether an assignment "correcting" ownership or a merger conveyance was recorded after 2019 should be confirmed in Assignment Center — I found none in the indexed record.

Timeline diagram

timeline
    title Ownership of US 8404717
    2002 : Priority application filed
    2003 : Parent application filed
    2011 : Continuation filed by Zeldis
         : Assigned to Celgene Corp reel 026014
    2013 : Patent issued
         : Listed in Orange Book for Revlimid
    2017 : Celgene sues generic ANDA filers
    2018 : Dr Reddys files IPR2018-01507
    2019 : IPR institution denied
         : Celgene acquired by BMS
    2023 : Patent term expires Apr 11

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. The only recorded conveyance is inventor → Celgene Corporation, a large operating brand manufacturer (Reel 026014/0756, 2011‑03‑24). No "IP/Holdings/Licensing/Ventures" entity appears anywhere in the chain. Celgene's address of record is a corporate headquarters (Summit, NJ), not a registered-agent service.
2. Known asserter in the chain — Not present. Celgene does not appear on the Acacia / Marathon / IV / Wi‑LAN / Conversant / Vringo / Pendrell / Spangenberg-type assertion lists; it is an operating innovator. Adjacent fact worth noting (not an ownership finding): Coalition for Affordable Drugs VI LLC (a Spangenberg-affiliated entity) petitioned for IPR against a different Celgene patent (IPR2015‑01092, re the '501 REMS patent) — that petitioner is adverse to Celgene and appears nowhere in the '717 ownership chain.
3. Repeat correspondent across the chain — Not present / not assessable. With a single recorded link there is no recurrence to detect, and the correspondent of record could not be retrieved (see above). Marked as an open verification item, not a finding.
4. Cascading transfers — Not present. One assignment in the entire 21-year life of the family; no chained LLCs, no shared correspondent addresses, no sub‑24‑month transfer run.
5. Pre-litigation transfer — Not present. The assignment (2011‑03‑24) is contemporaneous with the application filing and predates any assertion of the '717 patent by roughly six years (the '717 patent first surfaced in the Revlimid ANDA litigation/Orange Book listings around 2013). No "clean-up" transfer was needed before assertion because Celgene's title was already recorded.
6. Bankruptcy fire-sale — Not present. Celgene was never a debtor; it was acquired in a ~$74B all-stock merger by BMS (closed 2019‑11‑20). No Chapter 7/11 sale of this patent exists.
7. Privateering — Not present. Celgene asserted the patent itself, in its own name, against generic competitors; there is no transfer to a proxy litigant.
8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN.

Litigation docket (assignee as plaintiff, from Google Patents legal-events + D.N.J. filings): 2:10‑cv‑05197 (Celgene v. Natco, Oct 2010); 2:17‑cv‑05314 (Dr. Reddy's); 2:17‑cv‑06842 (Lotus/Alvogen); 2:18‑cv‑11518; 2:19‑cv‑06999; 2:19‑cv‑14731; 2:19‑cv‑15449; 1:20‑cv‑00003 (N.D. W. Va., Celgene v. Mylan); 2:20‑cv‑07759; 2:20‑cv‑14389; 2:21‑cv‑10398; 2:21‑cv‑11261; 2:21‑cv‑12927; 2:21‑cv‑20459. Every one of these is Celgene as plaintiff against generic/pharmaceutical defendants — the inverse of an NPE pattern. Note also that the '717 patent was asserted alongside the '740, '569, '800, '217 and other Revlimid-family patents in Dr. Reddy's / Lotus / Sun / Cipla / Mylan ANDA suits.

Verdict

Operating-company assertion.

Justification: the sole recorded assignment (Reel 026014 / Frame 0756, recorded 2011‑03‑24) is a same-day inventor-to-employer conveyance from Jerome B. Zeldis to Celgene Corporation, and no further assignment, LLC transfer, or defensive-aggregator acquisition is recorded. Celgene is an operating brand manufacturer that ships the embodiment (Revlimid® / lenalidomide, NDA 021880, Orange Book use code U‑1215, expiration 2023‑04‑11) and used the patent to sue actual generic competitors (Dr. Reddy's, Lotus/Alvogen, Natco/Teva, Sun, Cipla, Mylan) in D.N.J. and N.D. W. Va., with Dr. Reddy's IPR2018‑01507 institution denied 2019‑02‑11. The patent has since reached term expiry (2023‑04‑11) and Google Patents lists it as "Expired – Fee Related," so it is no longer an active assertion vehicle — but at no point in its life did its ownership chain exhibit an NPE pattern.

Open verification items: (1) the Assignment Center record's correspondent field for Reel 026014/0756, which I could not retrieve from publicly indexed sources and deliberately did not guess; and (2) confirmation whether any post‑2019 merger-related recording exists tying the patent to Bristol‑Myers Squibb. Both require a direct query at https://assignmentcenter.uspto.gov/ for patent 8404717.

I'll search for the USPTO record and the cited prior art references for this specific patent.

Let me dig into the specific patent citations listed on the face of US 8,404,717 and the IPR record.

Let me pull the actual front-page "References Cited" list for US 8,404,717.

Prior-Art Analysis — US 8,404,717 (Celgene, Zeldis)

Sourcing caveat (read first)

Two things about this task need to be stated up front, because they affect what follows:

1. I could not retrieve the complete front‑page "References Cited" table for US 8,404,717. The authoritative full text supplied to me (Google Patents, fetched 2026‑10‑01) is truncated after §5.1 of the specification — it contains neither the claims nor the front‑page U.S. Patent Documents / Foreign Patent Documents / Other Publications tables. I will not fabricate an examiner-citation list. Instead I give you (a) every reference I can affirmatively document as cited in the specification I was given, and (b) the art that was actually asserted against this patent in prosecution and in IPR2018‑01507, which is the operative prior-art record for this patent. Where I cannot verify a detail, I say so.

2. The date in your instruction (April 26, 2026) differs from the system date (October 1, 2026). My searches were run against the live index; nothing I found is date-sensitive to that gap, but flagging it per the "interpret identifiers literally / don't auto-correct" rule.

I also did not find any 2026 Federal Circuit or district-court activity that changes the prior-art picture (consistent with the earlier section), so no contradiction with the previously generated section arises — except one item I must flag, below.

1. The specific patent, confirmed

Sources: https://patents.google.com/patent/US8404717/en ; https://pubchem.ncbi.nlm.nih.gov/patent/US-[8404717](/patent/8404717)-B2

Flag / apparent inconsistency with the earlier section: your earlier summary correctly notes that the abstract and specification are drafted around lenalidomide + 5‑azacytidine combination therapy, while granted claim 1 is a lenalidomide monotherapy claim (no azacitidine limitation). This materially changes the prior-art relevance map: all of the azacitidine-specific art cited in the specification is irrelevant to §102 analysis of claim 1, and the references that actually matter are the lenalidomide-in-MDS ones. I build on that point below.

2. What the record shows about the cited art

The only complete, verifiable prior-art record I could obtain for this patent is the IPR2018‑01507 petition and its exhibit list ([Dr. Reddy's Laboratories, Inc.](/litigations/by-plaintiff/Dr.%20Reddy's%20Laboratories%2C%20Inc.) v. Celgene Corp.), which sets out the art Celgene itself and the examiner treated as the closest art, plus the Patent Owner Response (Nov. 30, 2018).

• Petition / exhibit list: https://www.docketalarm.com/cases/[PTAB](/ptab)/IPR2018-01507/Inter_Partes_Review_of_U.S._Pat._8404717/docs/08-03-2018-Petitioner/Petition-2-717_Petition.pdf
• Patent Owner Response: https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1516997](/patent/1516997)/download-documents?artifactId=S_yvN43vhaUkD8VcHz2rr9hVxfo53nmfu2eNa_A5GXzfbFRtB83tlro
• Exhibit list (sibling IPR2018‑01509, Celgene exhibits incl. '717 file history): https://gaeflexstaging-dot-docketupdate.appspot.com/cases/PTAB/IPR2018-01509/Dr._Reddy%27s_Laboratories_Inc._v._Celgene_Corporation/

Outcome (from the record): IPR2018‑01507 was not instituted on the merits (Google Patents litigation record labels it "Not Instituted – Merits").

3. Reference-by-reference analysis

I have split the citations into three groups, because the §102 answer is different for each.

Group A — Patent documents cited in the specification (compound/background art)

These are the "patent citations" reproduced in the specification I was given. Several are Celgene's own compound patents.

Group B — Non‑patent literature cited in the specification

The specification's "Background" section cites, among others: Merck Manual 953–954 (17th ed. 1999); List et al., J. Clin. Oncol. 8:1424 (1990); Bennett et al., Ann. Intern. Med. 103(4):620‑625 (1985); Besa, Med. Clin. North Am. 76(3):599‑617 (1992); Harris et al., J. Clin. Oncol. 17(12):3835‑3849 (1999); Greenberg et al., Blood 89:2079‑2088 (1997) (the IPSS); Epstein & Slease, Surg. Ann. 17:125 (1985); Dexter (1987, 1989); Metcalf (1985); Golde & Gasson (1988); Tabbara & Robinson (1991); Ogawa (1989); Stanley (1976); Schrader (1981); Moore (1980, 1991); Kurland (1979); Handman & Burgess (1979); Vadas (1983 ×2); Weibart (1986); Schuster (1990); Besa (1990); Hellstrom (1990); Bowen (1991); Koch, Prog. Med. Chem. 22:165‑242 (1985); Moller et al., J. Immunol. 159:5157‑5161 (1997); Vasiliauskas et al., Gastroenterology 117:1278‑1287 (1999); Ehrenpreis et al., Gastroenterology 117:1271‑1277 (1999); D'Amato et al., PNAS 91:4082‑4085 (1994); Singhal et al., N. Engl. J. Med. 341(21):1565‑1571 (1999); Marx et al., Proc. Am. Soc. Clin. Oncol. 18:454a (1999); Costa et al., Blood 92(10 suppl. 1):235b (1998); McCann, Drug Topics 41‑42 (1999‑06‑21); Kropff, Blood 96(11 pt.1):168a (2000); Munshi et al., Blood 94(10 pt.1):578a (1999); Penichet & Morrison, J. Immunol. Methods 248:91‑101 (2001); Emens et al., Curr. Opin. Mol. Ther. 3(1):77‑84 (2001); Corral et al., Ann. Rheum. Dis. 58(Suppl I):I107‑I113 (1999); Marriott et al., Expert Opin. Biol. Ther. 1(4):1‑8 (2001); Muller et al., J. Med. Chem. 39(17):3238‑3240 (1996); Muller et al., Bioorg. Med. Chem. Lett. 8:2669‑2674 (1998); Kornblith et al., J. Clin. Oncol. 20(10):2441‑2452 (2002) and Silverman et al., J. Clin. Oncol. 20(10):2429‑2440 (2002) (azacitidine in MDS); Deeg et al., Leukemia 16:162‑164 (2002) (etanercept in MDS); and a large "Other Publications" list reproduced by PubChem (https://pubchem.ncbi.nlm.nih.gov/patent/US-8404717-B2).

§102 bottom line for Group B: None of these anticipates any of claims 1–10. They are mechanistic, epidemiological, IPSS/first-line-therapy background. Notably, Kornblith 2002 and Silverman 2002 (azacitidine) have no §102 relevance to claim 1, which contains no cytidine-analog limitation. Deeg 2002 (etanercept) is actually exculpatory art — it evidences that TNF‑α blockade failed in MDS (this became Celgene's "teach-away" argument in the IPR).

Group C — Art actually asserted against the '717 claims (IPR2018‑01507)

This is where the substantive prior-art fight was, and it is the closest thing to "the most relevant prior art" for this patent.

4. Direct answer to "which claim(s) does each reference potentially anticipate under §102"

No cited reference — patent or non‑patent — anticipates claims 1–10, singly or in combination, under 35 U.S.C. § 102. Claim 1 requires the simultaneous presence of four elements:

1. the specific compound 3‑(4‑amino‑1‑oxo‑1,3‑dihydro‑isoindol‑2‑yl)‑piperidine‑2,6‑dione (or salt/solvate/stereoisomer);
2. administration at about 5 to about 25 mg per day;
3. to a patient having transfusion‑dependent anemia;
4. due to low‑ to intermediate‑1‑risk MDS.

Mapping the record:

• The Celgene compound patents (A1, A2, A3, A10) supply element 1 only.
• The Celgene press releases (C4) supply, at most, elements 1–2 (up to 25 mg/day, in MM/solid tumours).
• List 2001 (C1) and Thomas 2000a (C2) supply, at most, background on the MDS indication (element 4, generically) via thalidomide — not lenalidomide.
• Raza 2001 (C5) supplies element 3–4 background, but for thalidomide, not the claimed compound.
• No reference supplies all four, and no reference supplies elements 1 + 2 + 3 + 4 to the specific low/Int‑1‑risk TDA sub‑population.

Accordingly, the asserted grounds in IPR2018‑01507 were §103 obviousness combinations (Ground 1: List 2001 in view of the '230 patent and the press releases; Ground 2: Thomas 2000a in view of the '230 patent and the press releases) — not §102 anticipation. The Board did not institute. And the azacitidine-specific citations (A11, plus Kornblith 2002 and Silverman 2002) are outside claim 1 entirely because claim 1 has no cytidine-analog limitation.

If you are looking for anti‑anticipation pressure points rather than anticipation proper, the only citations with genuine §102-adjacent bite are:

• C4 (Celgene's own May 8, 2001 press release) — closest to elements 1–2, but fails on the MDS/TDA limitation and on public-accessibility proof; and
• A13 (WO 01/87307) — an earlier Zeldis/Celgene MDS disclosure, which is the one reference of record that could theoretically be §102(e)/§102(b) art against the method concept; I could not verify its content and am not asserting that it anticipates anything.

5. Honest limitations of this analysis

• I did not retrieve the literal front‑page "References Cited" tables of US 8,404,717; the list above is drawn from the specification text I was given plus the IPR/prosecution record. Treat the Group A/B lists as complete for what the specification cites, but not certified as coextensive with the examiner's citation list.
• I did not retrieve the 2012‑05‑09 Office Action, which per PO applied the "same rationale [as the IPR] but different art." The identity of the examiner's §102/§103 references is an open item.
• WO 01/87307's title, disclosure and dates are unverified and I have not analyzed it.
• No 2026 appellate activity for US 8,404,717 surfaced; as previously noted, a negative search is not a substitute for a PACER/Docket Alarm query.

If you want the certified citation list, the authoritative sources are the USPTO PatentCenter "References Cited" tab for application 13/070,761, the printed patent front page, and the 2012‑05‑09 Office Action — none of which are fully reproduced in the material available to me here.

I'll ground the obviousness analysis by verifying a few key reference facts that the prior-art section flagged as open or unverified — especially WO 01/87307, the List 2001 ASH paper, Raza 2001, and the '230 patent's actual disclosure.

The searches surfaced a pivotal new fact: the '717 patent's own face cites WO 01/87307 (Nov. 2001), and the N.J. antitrust complaint alleges the PTO found the parent '740 claims obvious over "Raza (August 2001) in view of WO '307." Let me pin down WO '307's disclosure and the '717 claim text before analyzing.

Obviousness Analysis — U.S. Patent No. 8,404,717 Under 35 U.S.C. § 103

Built on the prior-art section above. I do not re-derive the reference list; I use it, plus four new primary-source facts that materially sharpen the § 103 picture and that surfaced only in this round of verification. Those new facts are flagged in § 1 because one of them changes the analysis.

1. Four verification updates from this session (read first)

These refine — and in one case partially contradict a caveat in the earlier prior-art section.

(a) WO 01/87307 is no longer a mystery reference — it is the examiner's own parent-case reference.
The earlier prior-art section listed WO 01/87307 as "unverified … I could not independently verify this document's disclosure. Open item; do not treat as analyzed." That open item can now be partially closed:

• Confirmed: WO 01/87307 has an international publication date of November 22, 2001 and is cited on the face of both the '717 patent and its parent US 7,189,740 (the "WO 01/87307 A2 * 11/2001" entry, the asterisk marking examiner citation). It also appears on the faces of US 8,034,831 and US 9,283,215 in the same family. Source: US 8,404,717 face; US 7,189,740 face (https://patentimages.storage.googleapis.com/5c/0c/f3/6a7c014c5d9051/US7189740.pdf).
• Alleged in litigation: the Cigna/Hagens Berman amended complaint (Consolidated, D.N.J., filed 2022-05-27) alleges the PTO "rejected the claims [of the '740] … as obvious over a newly-cited prior art reference, Raza (August 2001, Blood), in view of WO 01/87307 (international publication date of November 22, 2001)," and that the claims were also "rejected … as anticipated or obvious over WO 01/87307 and for double patenting over U.S. Application No. 10/438213." Source: https://www.hbsslaw.com/sites/default/files/case-downloads/revlimid/2022-05-27-amended-complaint.pdf and the mirrored N.J. complaint at https://storage.courtlistener.com/recap/gov.uscourts.njd.[498914](/patent/498914)/gov.uscourts.njd.498914.1.0_1.pdf.
• Still unverified: WO '307's actual disclosure text. I could not retrieve its specification, so I do not assert what it teaches. The date and citation status are now solid; the content is not.

(b) The '717 patent issued subject to a terminal disclaimer (face: "*This patent is subject to a terminal disclaimer"). Its parent '740 likewise issued "subject to terminal disclaimer as to U.S. Patent Application No. 10/438,213." This matters for § 103/OTDP strategy (§ 8, below).

(c) The parallel IPR on sibling '120 patent contains an express, quotable motivation-to-combine theory against the same family, same art, same day as the '717 petition. That is the single most useful piece of new evidence for this analysis and I quote it in § 5. Source: Petition, IPR2018-01504 (U.S. Pat. 9,056,120), https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1516995](/patent/1516995)/download-documents.

(d) Raza 2001's author list includes "J. Zeldis." The Raza reference that both the examiner and Dr. Reddy's relied on is co-authored by the '717 patent's sole inventor. This creates a § 102(b) statutory-bar problem (an inventor's own publication more than one year before filing is a bar and cannot be sworn behind), which I discuss at § 3.

2. Governing law and the analytical frame

• Pre-AIA § 103 applies. The '717 claims priority to 2002–2003 and the Board itself stated the '120/'717/'740 claims "are to be reviewed under pre-AIA § 103." The AIA's § 102(a)(1)/(a)(2) and the "common ownership" carve-outs do not apply.
• Standard: Graham v. John Deere (scope/content/differences/objective evidence) as reshaped by KSR Int'l v. Teleflex, 550 U.S. 398 (2007) — predictable combinations, "obvious to try" where the number of identified solutions is small, and "a court must ask whether the improvement is more than the predictable use of prior art elements according to their established functions."
• Claim structure: one independent claim (claim 1) plus nine narrowing dependents (doses 5/10/15/25 mg/day; capsule form; free base). Claims 2–10 stand or fall with claim 1 on the § 103 merits, because they add only routine dosage-form/dose-amount limitations.
• Critical divergence from the spec: as flagged in the earlier review, the abstract and specification are written around lenalidomide + 5-azacytidine combination therapy, but granted claim 1 is monotherapy. Therefore every azacitidine-specific reference in the record — US Ser. No. 11/849,958, Kornblith 2002, Silverman 2002, and the azacitidine dosing paragraphs — is irrelevant to § 103 analysis of claims 1–10. That is not a small point: it removes the patent's own "unexpected synergy" story from the claim 1 case and forces the obviousness inquiry onto the lenalidomide-in-MDS art alone.

3. The effective-filing-date question is outcome-determinative — and it cuts against the patent

Claim 1's four elements:

Whether each reference qualifies as prior art depends entirely on the date, so the presumptive critical dates matter:

Two consequences:

1. Raza 2001 is a § 102(b) statutory bar. Published 2001-08-15 — 14 months before the 2002-10-15 provisional and 20 months before the 2003-04-11 parent filing. A § 102(b) reference cannot be overcome by a Rule 131/132 declaration of earlier conception (In re Kathawala; MPEP 715). This is a tension with the litigation's account: the Cigna complaint alleges the PTO rejected the '740 over "Raza (August 2001) in view of WO 01/87307" and that Celgene beat it with Zeldis's conception declarations. If the rejection truly rested on Raza as § 102(b) art, the declarations could not lawfully have overcome it. Either the examiner's rejection was under § 102(a)/(e) only (with Raza as evidence of the state of the art and/or WO '307 as the § 102(e) reference), or the pleading over-simplifies the ground. I cannot resolve this from the sources available; I flag it rather than assert it. What is certain is that the examiner made a Raza-plus-WO-'307 rejection against the parent, and the '717 rides on the same chain.
2. If the '717 is denied the provisional's benefit (a real risk: the provisional must describe the 5–25 mg/day, TDA/low-Int-1 limitations), the critical date slides to April 11, 2002, which only enlarges the body of prior art (WO '307 and List 2001 become squarely prior art) and does not rescue Raza, '230, Thomas or the press releases. The patent's effective-date posture is therefore a one-way ratchet toward more prior art.

4. The prior-art clusters, mapped to claim 1's elements

The key structural insight: no single reference supplies all four elements, but each element is supplied by a § 102(b) or § 102(a) reference, and the references are all in the same anti-angiogenic/anti-TNFα MDS literature that the inventor himself was publishing in. That is the classic KSR "combination of familiar elements according to known methods" fact pattern.

5. Combination theories, with motivation articulated

Theory A (principal) — Raza 2001 + '230/'517 + clinical dose disclosures

The combination: Raza 2001 teaches that thalidomide, a glutarimide immunomodulator, produces hematologic improvement and transfusion independence in MDS, with the responses concentrated in patients without excess blasts (low/Int-1 risk). The '230 (and '517) patent teaches the specific compound lenalidomide as a member of the same thalidomide-derived class. The Celgene press releases and MM phase I disclosures teach that lenalidomide is orally administered in humans at 5–50 mg/day, including "up to 25 mg per day," with acceptable tolerability.

Why a POSA would combine them (the motivation prong), and this is exactly how Petitioner framed it in the parallel '120 IPR:

"Prior to the priority date, a skilled artisan was aware of … lenalidomide and its position as the 'next generation,' more potent analog of thalidomide; lenalidomide's and thalidomide's successful clinical administrations in myelodysplastic syndromes (MDS), including MDS with resulting transfusion-dependent anemia (TDA) … and lenalidomide's repeated, successful clinical administrations at 5-50 and 'up to 25' mg/day. … lenalidomide's successes had motivated searches for additional therapeutic applications, and would have been motivated to administer lenalidomide for MDS/TDA. Similarly, it would have been obvious to 'upgrade' to lenalidomide for conditions in which thalidomide had already shown promise, such as MDS/TDA … Thalidomide suffered from recognized drawbacks, including its toxicity profile and known teratogenicity. Lenalidomide, with increased potency, decreased toxicity/side effects, and action on the same therapeutic target, helped remedy thalidomide's known drawbacks."
— Petition, IPR2018-01504 (Petitioner's stated motivation; a party's position, not a tribunal finding), https://ptacts.uspto.gov/ptacts/informations/petitions/1516995/download-documents

That articulation maps onto four independent KSR rationales:

1. Simple substitution of a known element for its better-known analogue to obtain predictable results. Thalidomide → its 4-amino-glutarimide analog, sharing TNF-α inhibition, anti-angiogenic activity, and T-cell co-stimulation (Corral 1999, cited on the '717 face, expressly describes "two distinct classes of thalidomide analogues that are potent inhibitors of TNF-alpha").
2. Known technique (dose-ranging) applied to a known drug. The 5–25 mg/day window sits inside the 5–50 mg/day range already administered to humans; selecting a sub-range from a disclosed range is presumptively obvious (In re Peterson, 315 F.3d 1325; In re Woodruff).
3. "Obvious to try" with a small number of predictable solutions. Once thalidomide's MDS activity was reported (Raza 2001; Zorat 2001; Musto 2002) and lenalidomide was known as the more potent, better-tolerated analog, the pool of candidate IMiDs for the same indication was small and the expected result predictable.
4. Same field, same problem, same mechanism. Both references address anemia in MDS and the same molecular target (TNF-α / angiogenesis). KSR makes this the paradigm case for combination.

Theory B — The Dr. Reddy's List 2001 theory (Ground 1 of the actual IPR)

List 2001 (ASH Education Program Book) in view of the '230 patent and the Celgene press releases. Same motivation as Theory A, with List 2001 supplying the MDS/angiogenesis teaching instead of Raza. Practical weakness: List 2001 was presented 2001-12-07/11, after the 10/15/2001 critical date — so it is at best § 102(a)/(e) art that can be sworn behind, and Celgene disputed its prior-art status entirely. This is the weaker of the two primary grounds.

Theory C — The Dr. Reddy's Thomas theory (Ground 2)

Thomas 2000a in view of the '230 patent and press releases. Material weakness: Thomas 2000a describes a SelCID (CDC-801) — a different compound class — entering an MDS trial, and reportedly characterizes certain thalidomide analogs as among the least potent TNF-α inhibitors. Celgene's response (that this points away from lenalidomide) is colorable. Theory A, using Raza, is stronger because Raza is about the same glutarimide class.

Theory D — The PTO's own parent-case combination: Raza + WO 01/87307

This is the combination the examiner actually used, per the pleadings: Raza (Aug. 2001) in view of WO 01/87307 rejected the parent '740 claims. If that rejection is genuine — and the face-citation of WO '307 on the '740 and '717 confirms the reference is of record — then the § 103 case against the genus (lenalidomide for MDS) was already made by the Office in 2006. Claim 1 of the '717 is a narrowing continuation that adds (i) the 5–25 mg/day dose and (ii) the TDA/low-Int-1 subpopulation. The obviousness question therefore reduces to whether those two added limitations are themselves nonobvious — and on the record above, the doses are disclosed in the clinical art and the subpopulation is taught by Raza itself. Caveat: the "PTO found obvious" statement is a party allegation in an antitrust complaint; I have not seen the Office Actions, and I flag the § 102(b)/swear-behind inconsistency noted in § 3(b). Treat Theory D as the strongest theory of record but not as an adjudicated finding.

6. Claim-by-claim obviousness exposure

Note the asymmetry: the genus claim (1) is the most defensible, because it requires all four elements simultaneously — while the species claims toward 25 mg and the capsule/free-base claims are the most exposed, because each adds only a routine optimization.

7. The nonobviousness case (the patent's best answers) — and how much weight each carries

A rigorous § 103 opinion must state the strongest counter-case, because two of these are genuinely strong.

(i) No reference discloses lenalidomide for MDS. True, and this is the classic "new use of a known compound" scenario. But under pre-AIA § 103 it is not dispositive: where the prior art suggests both the compound and the indication, efficacy in the claimed indication does not confer patentability absent unexpected results (In re Kao line of reasoning; Pfizer v. Apotex). Celgene's whole case must rest on unexpected results.

(ii) Teach-away #1 — thalidomide analogues cause cytopenias, the cardinal MDS sign. Celgene assembled a large body of art (Musto 2002; Zorat 2001; Cortelezzi 2000; Bincoletto 1998; Gupta 1999; Zang 2001; Rosenfeld 2000) for the proposition that thalidomide-class drugs suppress marrow and therefore would not be pursued in a cytopenic population. This is corroborated by the agents' actual toxicity profile — later literature reports lenalidomide's dose-limiting adverse events as neutropenia (~65%) and thrombocytopenia (~74%). This is the strongest nonobviousness argument.

(iii) Teach-away #2 — direct TNF-α blockade failed in MDS. Deeg 2002 (etanercept, Leukemia 16:162-164), Maciejewski 2002 and Rosenfeld 2002 report that soluble TNF-receptor constructs did not produce meaningful hematologic benefit in MDS. If the asserted motivation is lenalidomide's anti-TNF-α activity, the failure of the most direct TNF-α blockade in the same disease undercuts the motivation prong. Also strong — and, notably, the '717 patent's own specification concedes the point by listing etanercept among optional agents while the art shows its failure as monotherapy.

(iv) Unexpected results / objective indicia. The patent family's specification reports a 16-patient study yielding 9 erythroid responses, all in low/Int-1-risk patients, with 5q31-33 cytogenetic remissions. The published List 2005 NEJM data (56% TI in del(5q); ~38% in non-del(5q)) are the strongest objective evidence — but they post-date the October 2002 priority date and so can only be used as subsequent evidence of an inherent, unexpected property (the Knoll v. Teva / Kollman line). They are not prior art and cannot be used against the patent; they are the patentee's ammunition.

(v) Honest counterweight against the patent on the subpopulation. Raza 2001 expressly reports that responders were the patients with lower pre-therapy blasts and the reference's own conclusion is that thalidomide "is effective … especially those who present without excess blasts." So element E4 (low/Int-1 risk) is arguably taught, not selected. Similarly, transfusion dependence was the measured endpoint of the Raza trial itself. The two elements the patentee might most want to rely on for nonobviousness are the two the primary reference most clearly supplies. That is the crux of why claim 1 is vulnerable despite arguments (ii) and (iii).

Net assessment of the counter-case: Teach-away arguments (ii) and (iii) are real and legally cognizable, and a fact-finder could find them persuasive. But neither squarely rebuts the Raza + '230 + dose combination, because Raza is thalidomide-class art (not a "different mechanism") and because the Federal Circuit demands that a teach-away be more than a general preference against a class. The objective-indicia case is parked on post-priority evidence. On a clean record, I would expect claim 1 to be a close, genuinely contestable § 103 case, with the dependent species claims materially weaker.

8. Why the actual IPR is not the measure of the § 103 case

This is the most important practical caveat:

• IPR2018-01507 was denied at institution (2019-02-11) because Petitioner failed the threshold showing that the Celgene press releases were publicly accessible printed publications — not because the Board blessed the claims as nonobvious. No Final Written Decision exists. Source: Paper 7; Jones Day engagement summary, https://www.jonesday.com/en/practices/experience/2019/02/celgene-successfully-blocks-institution-of-eminter.
• The obvious fix is to substitute better-proven art for the press releases. The dose element (E2) can be supported by § 102(b)-quality or independent art: the ASH phase I disclosures (Zangari et al., Blood 98(11):775a, Abstr. #3226; Richardson et al.), or the '230 patent's own dosing guidance, rather than the press releases whose dissemination is contestable. A § 103 theory that does not depend on the press releases is not estopped (no FWD, so no § 315(e) estoppel) and is not precluded by the denial.
• But the patent is expired (2023-04-11) and the practical posture today is an expired, unadjudicated method patent — the invalidity question is worth litigating only if a live damages tail exists.

Secondary attack worth noting: every patent in this chain issued subject to terminal disclaimers ('717 face; '740 disclaimed over App. 10/438,213). Where a challenged claim is terminally disclaimed to a parent, the ordinary tactical consequence is that the continuation's claims must be patentably distinct from the parent's or they rise and fall together. If claim 1 of the '740 patent is vulnerable over Raza + WO '307 (as the pleadings allege), a defensive theory that the '717's 5–25 mg/TDA claims are an obvious variant of the '740's broader MDS claims (i.e., OTDP) is worth pleading in the alternative.

9. Bottom line

10. Caveats and open items

1. The "PTO found the '740 obvious over Raza + WO '307" statement is a party allegation in an antitrust complaint (Cigna/BMS side), not an adjudicated finding; I have not retrieved the 2006 Office Action or the Zeldis Rule 131/132 declarations. I also flagged the § 102(b)/swear-behind inconsistency in that account (§ 3).
2. WO 01/87307's disclosure remains unverified. I confirmed its Nov. 22, 2001 publication date and its citation of record, but not what it teaches. Do not attribute a specific disclosure to it without pulling the document.
3. The press releases' public accessibility is unresolved and was the basis for the institution denial; any § 103 theory should be built without relying on them.
4. No 2026 appellate or PTAB activity for US 8,404,717 surfaced, and the patent expired 2023-04-11; a live § 103 challenge has practical value only if a damages period survives.
5. Note the date discrepancy carried forward from the prior-art section: the task states April 26, 2026; the system date is October 1, 2026. Nothing in this analysis is date-sensitive to that gap.

If you want, I can build a formal element-by-element claim chart (claim 1, element E1–E4, mapped reference-by-reference with pinpoint citations), or a substitute-ground petition skeleton that replaces the press releases with the ASH phase I disclosures to cure the defect that defeated IPR2018-01507.