US 7326708

I'll search for authoritative records on US 7,326,708, including any 2026 CAFC activity.

US 7,326,708 — Verification and Summary

Search scope caveat: My retrievals returned third-party mirrors and aggregators of USPTO data (Google Patents, PubChem, FreePatentsOnline, PTAB PTACTS documents, Docket Alarm/DrugPatentWatch) rather than a direct PatentCenter/PAIR API response. I found no live 2026 Federal Circuit docket captioned to US 7,326,708. I am flagging that explicitly rather than inferring one.

1. Bibliographic data (as retrieved)

Sources: https://patents.google.com/patent/US7326708/en ; https://pubchem.ncbi.nlm.nih.gov/patent/US7326708 ; https://gaeflexstaging-dot-docketupdate.appspot.com/FDA/Orange_Book/021995/JANUVIA/

2. Abstract (verbatim)

"The dihydrogenphosphate salt of 4-oxo-4-[3-(trifluoromethyl)-5,6-dihydro [1,2,4]triazolo[4,3-a]pyrazin-7(8H)-yl]-1-(2,4,5-trifluorophenyl)butan-2-amine is a potent inhibitor of dipeptidyl peptidase-IV and is useful for the prevention and/or treatment of non-insulin dependent diabetes mellitus, also referred to as type 2 diabetes. The invention also relates to a crystalline monohydrate of the dihydrogenphosphate salt as well as a process for its preparation, pharmaceutical compositions containing this novel form and methods of use for the treatment of diabetes, obesity, and high blood pressure."

3. Plain-language overview of the independent claims

The compound is the Merck DPP-4 inhibitor sitagliptin, claimed as its 1:1 dihydrogenphosphate (biphosphate) salt — the marketed form of JANUVIA.

Dependent claim ladder of interest: claims 5–8 define the monohydrate by XRPD d-spacings (7.42, 5.48, 3.96 Å → plus 6.30, 4.75, 4.48 Å → plus 5.85, 5.21, 3.52 Å → FIG. 1); claims 9–11 by ¹³C CPMAS NMR (169.1, 120.8, 46.5 ppm → plus 159.0, 150.9, 40.7 ppm → FIG. 2); claims 12–14 by ¹⁹F MAS NMR (−64.5, −114.7, −136.3, −146.2 ppm → plus −96.5, −104.4, −106.3, −154.5 ppm → FIG. 3); claims 15–16 by the TGA (FIG. 4) and DSC (FIG. 5) curves.

4. Internal inconsistencies worth noting (technical-analyst flags)

Two numeric mismatches exist between the claims and the written description as presented:

1. Claim 24 requires water above 6.8 weight percent for IPA/water crystallization at 25 °C, while the specification's General Method (e) states above 7.0 weight percent for the same IPA/water/25 °C system.
2. Claim 21 recites a temperature range of about 25–100 °C, whereas the specification's process description states about 25 °C to about 80 °C.

Also present: the specification renumbering artifacts and apparent typos ("1 ,2,4", "triflorophenyl", "EPA" for IPA) — I am reporting these as they appear rather than correcting them.

Reported physical data: aqueous solubility ~72 mg/mL; TGA weight loss ~3.3647% from ambient to ~250 °C; isolated solids >99.8% HPLC area purity, mean PSD 80 µm.

5. Litigation and 2026 docket check

Federal Circuit. The only substantive CAFC appeals I can tie to this patent are:

• Nos. 21-2121, 21-2122, 21-2123 — Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp., decided September 29, 2022 (Lourie, J., joined by Reyna and Stoll, JJ.), affirming the PTAB's final written decision. Cited as precedential guidance on the "at once envisage" / inherent-anticipation standard (957 predicted salts); In re Petering distinguished. https://www.mololamken.com/assets/htmldocuments/Mylan%20v.%20Merck%20Opinion.pdf
• No. 23-1013 — a further Mylan appeal, listed by litigation aggregators as filed 2022-10-05 and terminated 2023-05-11.

No 2026 CAFC docket specific to US 7,326,708 surfaced in my searches. One search hit — a pharmaphorum article served with a 2026 date — describes the CAFC upholding the patent against Viatris, but its body is unmistakably reporting the September 2022 decision (it cites "$2.5 billion in the first half of 2022"). I treat it as a re-dated 2022 report, not new 2026 activity. https://pharmaphorum.com/news/merck-defends-blockbuster-januvia-franchise-from-patent-challenge

PTAB. IPR2020-00040 (Mylan — settled); IPR2020-01045, -01060, -01072 (Unified Patents, final written decisions); additional petitions by Teva (2020-06-10), Dr. Reddy's (2020-06-11), Sun (2020-06-12).

District court. Extensive ANDA litigation across D. Del., N.D. W. Va. (1:19-cv-00101), D.N.J., and D.D.C. The N.D. W. Va. bench opinion (Doc. 204, filed 10/26/2022) found Mylan infringed claim 3 of the '708 patent. https://storage.courtlistener.com/recap/gov.uscourts.wvnd.46385/gov.uscourts.wvnd.46385.204.0.pdf

6. Where I am not certain

• 2026 activity: I cannot rule out a pending or recently filed 2026 CAFC appeal involving this patent that my search interface did not surface. If a specific 2026 docket number is supplied, I can verify it directly. I decline to manufacture one.
• Live termination date: Google Patents shows "Active, expires 2026-11-24"; Orange Book shows Nov 24, 2026 plus pediatric exclusivity to May 24, 2027. I have not confirmed a maintenance-fee lapse or a terminal disclaimer effect beyond the adjusted-expiration field shown.
• Full family/priority: Google Patents lists the immediate provisional as 60/482,161; the published family includes EP 1654263 B2, JP 4463768, CN 100430397, WO 2005/003135, among ~40 jurisdictions.

I'll verify the litigation record for US 7,326,708 against current sources before compiling the list.

Let me pull details on the later-filed cases and their outcomes.

Let me check the post-2021 cases and the MDL resolution status.

Litigation Report — U.S. Patent No. 7,326,708

Prepared: April 26, 2026 (per task instruction). Note on dating: the retrieval environment's system clock reads 2026-09-25. I flag the discrepancy rather than silently reconciling it; everything below is as-retrieved from the live sources listed.

Scope rule applied: I include only matters where the docket or a court filing expressly identifies U.S. Patent No. 7,326,708 ("the '708 patent") as asserted, challenged, or the subject of a declaratory-judgment count. Related sitagliptin patents — notably U.S. 6,699,871, U.S. 8,414,921, U.S. 7,125,873, and U.S. 8,168,637 — are named only to distinguish them; the reader should not import their case histories into the '708 record.

One important note for the file: the '708 patent is not the compound patent. It is the salt/polymorph patent (dihydrogenphosphate salt + crystalline monohydrate). Where a source describes litigation over "sitagliptin compound" or a "patent term extension," that is a different patent (e.g., the Tokyo action against Sawai/Medisa concerned Japanese Patent No. 3762407 and its PTE, not US 7,326,708 — MSD press release, 2026-02-03: https://www.msd.co.jp/news/product-news-20260203/).

1. Consolidated MDL proceeding

In re Sitagliptin Phosphate ('708 & '921) Patent Litigation, MDL No. 2902, C.A. No. 1:19-md-02902-RGA (D. Del.) (Judge Richard G. Andrews).

• Created by JPML transfer order, August 8, 2019, centralizing 14 actions (13 in D. Del. + 1 in N.D. W. Va.). Case No. 1:19-cv-00101-IMK. https://www.jpml.uscourts.gov/sites/jpml/files/MDL-2902-Initial_Transfer-07-19.pdf ; reported at In re Sitagliptin Phosphate ('708 & '921) Patent Litig., 402 F. Supp. 3d 1366 (J.P.M.L. 2019).
• MDL docket hub: https://dockets.justia.com/docket/delaware/dedce/1:2019md02902/69691
• Claim construction (Nov. 17, 2020): In re Sitagliptin Phosphate ('708 & '921) Patent Litig., 2020 WL 6743022 (D. Del. Nov. 17, 2020). Court held claims 2, 3 and 21 include hydrates; construed "crystalline monohydrate" (claims 4, 24) as "a repeating unit cell incorporating a 1:1 ratio of water to a dihydrogenphosphate salt of sitagliptin." https://cases.justia.com/federal/district-courts/delaware/dedce/1:2019cv00311/[67784/135](https://assignmentcenter.uspto.gov/search/patent/reelFrameDetail?reelFrame=67784-0135)/0.pdf

2. District court actions asserting the '708 patent

Merck Sharp & Dohme Corp. / LLC is plaintiff in all rows unless marked. Filing dates are as stated in the complaints or the Teva IPR mandatory-notice enumeration (IPR2020-01045, Updated Mandatory Notices, https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1537262](/patent/1537262)/download-documents).

2(a) 2019 wave — D. Del. (all assigned/related to Judge Andrews)

Source for rows 1–14 enumeration: Teva Updated Mandatory Notices, IPR2020-01045; MDL Schedule A; Robinson Kaplan Generically Speaking (Spring 2019) Hatch-Waxman bulletin, https://www.robinskaplan.com/printpilot-publication-resources-legal-updates-generically-speaking-hatch-waxman-bulletin-2019-generically-speaking-spring-2019-new-anda-cases2.pdf

2(b) 2020–2021 wave

2(c) Additional D. Del. dockets listed by the patent-family record as '708 litigation

Google Patents' litigation panel for US 7,326,708 lists these D. Del. dockets: 1:20-cv-01099, 1:20-cv-01496, 1:21-cv-00182, 1:21-cv-00315, 1:21-cv-00824, 1:21-cv-01033 (Dr. Reddy's — its answer in that action admits on its face that the '708 patent is asserted, https://paragraphfour.com/wp-content/uploads/2013/01/dedc21cv1033A.pdf), 1:21-cv-01172, 1:21-cv-01173, 1:21-cv-01616, 1:22-cv-00444, 1:22-cv-01071, 1:22-cv-01300, 1:22-cv-01348, 1:23-cv-00683, 1:24-cv-00545, 1:25-cv-00449. https://patents.google.com/patent/US7326708/en

I have verified the content of only some of these from primary filings. Two I verified directly:

• Merck Sharp & Dohme LLC v. Prinston Pharmaceutical Inc., 1:22-cv-01071 (D. Del.) — complaint expressly pleads infringement of the '708 patent arising from ANDA No. 216886 (generic JANUVIA), Prinston notice letter dated July 1, 2022. https://www.docketalarm.com/cases/Delaware_District_Court/1--22-cv-01071/Merck_Sharp_%26_Dohme_LLC._v._Prinston_Pharmaceutical_Inc/docs/1.pdf — Status: settled (consistent with Merck's disclosure that settlements with 20+ generics deferred launch to May/July 2026).
• 1:23-cv-00683, 1:24-cv-00545, 1:25-cv-00449 — as of the retrieval date these are the most recent D. Del. '708 dockets surfaced by the family record. The 1:25-cv-00449 filing is the newest I can identify. I could not retrieve a complaint, party name, or disposition for these from a primary source, and I decline to guess at defendant identities.

2(d) N.D. West Virginia — the merits case

Merck Sharp & Dohme Corp. (substituted: Merck Sharp & Dohme LLC) v. Mylan Pharmaceuticals Inc. and Mylan Inc.

• Jurisdiction: U.S. District Court for the Northern District of West Virginia (Judge Irene M. Keeley)
• Case No.: 1:19-cv-00101-IMK
• Filed: May 2, 2019
• Claims at issue: '708 claim 3 (infringement); '708 claims 1, 2, 3, 19 (validity — ODP and § 112 written description/enablement)
• Outcome: Merck prevailed. Following a five-day bench trial, the court's Memorandum Opinion and Order (Doc. 204, filed October 26, 2022) found Mylan's ANDA products infringe claim 3 of the '708 patent. https://storage.courtlistener.com/recap/gov.uscourts.wvnd.46385/gov.uscourts.wvnd.46385.204.0.pdf
• Appeal: Mylan appealed. I have not independently confirmed the appellate case number for the district-court appeal. The prior-generated section of this analysis attributes Fed. Cir. No. 23-1013 (filed 2022-10-05, terminated 2023-05-11) to a "further Mylan appeal" — that timing is consistent with an appeal from the Oct. 26, 2022 judgment, but I flag it as inferred, not confirmed. I found no 2026 Federal Circuit docket captioned to this patent.

2(e) D.N.J. — declaratory judgment action (role reversal)

Par Pharmaceutical, Inc. v. Merck Sharp & Dohme Corp.

• Jurisdiction: U.S. District Court for the District of New Jersey
• Case No.: 2:19-cv-04432
• Filed: 2019
• Nature: DJ complaint seeking a declaration that the claims of the '708 patent are invalid (35 U.S.C. §§ 1 et seq.; Declaratory Judgment Act, 28 U.S.C. §§ 2201–02; 35 U.S.C. § 271(e)(5)). Par is the plaintiff here; Merck is the defendant. Par's ANDA product is a generic JANUMET XR (NDA No. 202270). https://paragraphfour.com/uploads/cases19/njdc19cv4432C.pdf
• Status: Settled — Par/Anchen appeared in the D. Del. Markman (row 2).

2(f) D.D.C.

The family record lists 1:10-cv-01110 (D.D.C.) among '708-related matters. I could not verify from a primary source that US 7,326,708 was asserted or challenged in that action. I flag it as unverified rather than list it as a '708 case.

3. PTAB (administrative) proceedings

These are not "litigation" strictly speaking, but they are the adjudicative challenges to the '708 patent and are the source of the appellate activity.

Aggregator confirmation of the four petitions and filing dates: https://www.drugpatentwatch.com/p/alphalignals/litigation/drugname/index.php?query=Sitagliptin+Phosphate (last updated 2026-03-17).

4. Federal Circuit

• Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp., Nos. 21-2121, 21-2122, 21-2123 (Fed. Cir.) — decided September 29, 2022 (Lourie, J.; Reyna and Stoll, JJ., joining). Affirmed the PTAB's final written decision sustaining the '708 patent; the opinion is cited for the "at once envisage" / inherent-anticipation standard and distinguishes In re Petering. https://www.mololamken.com/assets/htmldocuments/Mylan%20v.%20Merck%20Opinion.pdf
• No. 23-1013 — a further Mylan appeal per the prior section (filed 2022-10-05, terminated 2023-05-11). See my caveat in § 2(d).
• No 2026 CAFC activity specific to US 7,326,708 was surfaced. The pharmaphorum article that carries a 2026-dated header ("Friday 4 September 2026", https://pharmaphorum.com/news/merck-defends-blockbuster-januvia-franchise-from-patent-challenge) reports the 2022 CAFC affirmance against "Mylan Pharma, now Viatris," citing H1-2022 Januvia sales of "almost $2.5 billion." I treat it as a re-dated republication of the 2022 decision, not new 2026 activity, and I decline to represent it as current.

I note the prior section of this analysis reached the same conclusion on the 2026 point; there is no contradiction between the prior section and this one — only corroboration.

5. Outcome summary (as of retrieval)

1. The '708 patent has survived every adjudicated validity challenge. Both the PTAB FWDs and the Fed. Cir. affirmance (Sept. 29, 2022) went Merck's way. Mylan's district-court invalidity theories (obviousness-type double patenting; § 112 written description/enablement) also failed at the N.D. W. Va. bench trial (Oct. 26, 2022).
2. Infringement was adjudicated once, in Merck's favor — claim 3 of the '708 patent, N.D. W. Va.
3. Almost every other defendant settled, with consent judgments typically restraining infringement unless authorized and deferring generic launch. Merck's own disclosure: settlements with "more than 20 other generics companies" deferring launch "until May or July 2026 depending on the formulation, 'or earlier under certain circumstances.'"
4. Zydus is the notable carve-out — settled on terms permitting a non-automatically substitutable product containing a different sitagliptin form than JANUVIA's.
5. Timing pressure: With the Orange Book expiry at Nov. 24, 2026 plus pediatric exclusivity to May 24, 2027, and most settlements keyed to May/July 2026, this docket is in its terminal phase. The '708 patent's own adjusted expiration (Google Patents) is 2026-11-24.

6. Explicit uncertainties and one internal inconsistency to flag

• Contradiction flag — expiry date. The prior-generated section states pediatric exclusivity runs to May 24, 2027 (matching the Orange Book table: 7326708 Nov 24, 2026 DS DP U-802 / 7326708*PED May 24, 2027). The N.D. W. Va. bench opinion, however, recites that the patent "with pediatric exclusivity, expires on March 24, 2027." These cannot both be right. The Orange Book listing is the better evidence for the regulatory date; the opinion's "March 24, 2027" appears to be a transcription error. I report both rather than picking one silently.
• Unverified docket numbers. Fourteen of the post-2021 D. Del. dockets (1:21-cv-00182 through 1:25-cv-00449) are reported by the family record without party names or dispositions. I have not confirmed defendants or outcomes for most of them.
• Inferred appeal number. The mapping of Fed. Cir. No. 23-1013 to the N.D. W. Va. judgment is an inference from timing, not a verified docket link.
• No 2026 CAFC docket found. If a specific 2026 docket number is supplied, I can verify it directly against PACER/CM-ECF.
• Foreign proceedings are out of scope and are not '708 cases. The Indian litigation (Merck Sharp & Dohme Corp. v. Glenmark Pharmaceuticals, Delhi HC) concerns the Indian counterpart (IN 209816 / the abandoned salt application), and the Tokyo action concerns JP 3762407 + PTE. Neither adjudicates US 7,326,708, and I have excluded them.

Primary sources cited: Google Patents https://patents.google.com/patent/US7326708/en ; JPML MDL-2902 transfer order https://www.jpml.uscourts.gov/sites/jpml/files/MDL-2902-Initial_Transfer-07-19.pdf ; In re Sitagliptin Phosphate, 402 F. Supp. 3d 1366 (J.P.M.L. 2019) and 2020 WL 6743022 (D. Del. Nov. 17, 2020); N.D. W. Va. Doc. 204 https://storage.courtlistener.com/recap/gov.uscourts.wvnd.46385/gov.uscourts.wvnd.46385.204.0.pdf ; PTAB IPR2020-01045 mandatory notices and IPR2020-00040/01072 FWDs via https://ptacts.uspto.gov ; Robinson Kaplan Generically Speaking bulletins; DrugPatentWatch sitagliptin litigation index.

I'll verify the PTAB record on this patent, since the structured ODP block and the patent's own public record appear to conflict.

Good — the ODP block and the public record clearly diverge. Let me pin down each proceeding's petitioner, status, and outcome.

PTAB / AIA Trial History — US 7,326,708 ("Phosphoric acid salt of a dipeptidyl peptidase-IV inhibitor," sitagliptin dihydrogenphosphate)

⚠️ Source conflict — flagged before anything else

The structured "PTAB proceedings on file" block supplied with this prompt states that the USPTO ODP API returns no AIA trial proceedings for this patent. That is a data-coverage artifact, not the operative fact. Primary Board documents I retrieved directly (PTACTS petition papers, Docket Alarm PTAB dockets, RPX PTAB records, and the Board's own institution/joinder decisions) establish four IPRs on this patent, all filed 2019–2020 and all terminated by 2021.

Two additional contradictions with the previously generated sections must be flagged:

1. Petitioner identity. The earlier summary attributed IPR2020-01045, -01060, and -01072 to Unified Patents. That is wrong. The Board's own records identify the petitioners as Teva/Watson (‑01045), Dr. Reddy's (‑01060), and Sun Pharmaceutical Industries Ltd. (‑01072). No Unified Patents IPR on the '708 patent surfaced in any search. (The mis-attribution likely traces to a miscaptioned RPX URL slug — …/10970-sun-pharmaceutical-industries-v-merck-sharp-dohme-corporation-ipr-of-708 — that resolves to a page titled "Mylan Pharmaceuticals Incorporated et al v. Merck Sharp & Dohme Corporation.") Google Patents' litigation field for these three cases also lists a blank "Petitioner:", which invites the same error.
2. "Final Written Decision" labels. Google Patents tags ‑01045/‑01060/‑01072 as "Final Written Decision." They were joinder/understudy proceedings; the only merits FWD on this patent issued in IPR2020-00040 (Paper 41, 2021-05-07). The two joinder proceedings were closed by termination decisions on 2021-05-10.

I will not invent proceeding numbers. Everything below traces to a document or docket I actually retrieved.

Proceedings overview

Four (4) AIA trial proceedings have been filed against US 7,326,708 — all Inter Partes Reviews, all challenging the same claim set. Breakdown by status: 1 tried to a Final Written Decision with every challenged claim sustained (and affirmed by the Federal Circuit); 1 terminated by settlement; 2 terminated as joined understudies after the lead FWD. Claims invalidated: zero. Institution denied outright: zero.

Bottom-line defensive posture: the patent is hardened, not softened. No claim of the '708 patent has ever been canceled — not one of the 24. The single most dangerous prior art available against it (Merck's own pre-published WO 2003/004498 / US 6,699,871, "Edmondson") was run through a full IPR, rejected at the Board on both § 102 and § 103, and the rejection was affirmed in a precedential Federal Circuit opinion. For a defendant today, an IPR-based invalidity theory is not merely harder — it is closed: the '708 patent's adjusted expiration is 2026-11-24 (pediatric exclusivity to 2027-05-24), roughly 60 days from today, so no petition filed now can produce an FWD before the patent dies of natural causes. The practical window for PTAB attack has lapsed.

IPR2020-00040 — Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.

• Type: Inter Partes Review (35 U.S.C. §§ 311–319)
• Filed: 2019-10-30 (within one year of service of Merck's N.D. W. Va. complaint — § 315(b) satisfied)
• Status (verbatim from structured/aggregator data): "No Claims Unpatentable." Plain English: Mylan lost on every challenged claim. (Note the aggregator tag on Google Patents for this docket reads "Settlement" — that is misleading. Mylan litigated this IPR to a merits loss; the district court case settled in 2023.)
• Judge panel: Sheridan K. Snedden, Robert A. Pollock, and Timothy G. Majors. (RPX lists the lead panel as "Robert A. Pollock +3," which would imply a fourth member; the companion joinder decisions are signed by the three identified APJs. I flag the residual uncertainty.)
• Petition grounds: Challenged claims 1–4, 17, 19, and 21–23 (verbatim claim recitation from the FWD). Two grounds, parallel to Mylan's petition:
  • Ground 1 — § 102 anticipation by WO 2003/004498 and its U.S. counterpart US 6,699,871 (collectively "Edmondson"), treated by the parties and the Board as identical in relevant part. Mylan's theory: Edmondson discloses 33 species including sitagliptin plus eight "particularly preferred" acids including phosphoric acid, so a skilled artisan would "at once envisage" 1:1 sitagliptin DHP — expressly or inherently.
  • Ground 2 — § 103 obviousness over Edmondson in view of Structural Aspects of Hydrates and Solvates ("Brittain") and Salt Selection and Optimisation Procedures for Pharmaceutical New Chemical Entities ("Bastin").
• Institution decision: Instituted 2020-05-12 (Paper 21), on all challenged claims. Notably, the Board instituted the inherency theory on a preliminary record that "suggest[s] the 1:1 salt is the necessary byproduct of contacting phosphoric acid and sitagliptin" — i.e., the Board initially credited Mylan's expert that the 1:1 salt forms "every time." That preliminary footing collapsed at trial.
• Final Written Decision (Paper 41, 2021-05-07, reported at 2021 WL 1833325): No challenged claim held unpatentable. Claim-level disposition:
  • Claims 1, 2, 17, 19, 21, 22, 23 — sustained. The Board held these claims neither expressly nor inherently anticipated by Edmondson, and then held Edmondson disqualified as prior art: Merck carried its burden of proving prior reduction to practice antedating Edmondson's 2003-01-16 publication, taking the § 102(a) reference off the table; the '871 patent remained available only as a pre-AIA § 102(e) reference, and the pre-AIA § 103(c)(1) common-ownership exception barred its use in the obviousness analysis ("it is undisputed that Merck commonly owned Edmondson and the '708 patent").
  • Claim 3 — sustained (the (S)-enantiomer). The Board found "neither Edmondson nor Bastin disclosed anything related to (S)-sitagliptin or even a racemic mixture of any sitagliptin salt," and that Mylan offered no motivation and no reasonable expectation of success.
  • Claim 4 — sustained (the crystalline monohydrate). The Board found Mylan "provided no rationale to explain why a person of ordinary skill would have been motivated to make the claimed crystalline monohydrate form of 1:1 sitagliptin DHP … and failed to show that a skilled artisan would have had a reasonable expectation of success." Secondary considerations (unexpected properties) cut further against Mylan.
  • On inherency, the Board found "evidence, both experimental and from the technical literature, undeniably showed that 1:1 sitagliptin DHP does not form every time sitagliptin and DHP were reacted."
• Settlement / termination: None at the PTAB. The IPR ran to judgment.
• Appeal: Yes — and Merck won again. Consolidated on appeal as Nos. 21-2121, 21-2122, 21-2123 (Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.), decided 2022-09-29 (Lourie, J., joined by Reyna and Stoll, JJ.). Affirmed on all grounds. Precedential holdings: (i) 957 theoretically possible salts from the 33-compound × 8-acid disclosure is not a "limited class" a skilled artisan would "at once envisage" under In re Petering, 301 F.2d 676 (C.C.P.A. 1962) — "[t]he key term here is 'limited'"; (ii) substantial evidence supported non-obviousness of the (S)-enantiomer and of the crystalline monohydrate. Opinion: https://www.mololamken.com/assets/htmldocuments/Mylan%20v.%20Merck%20Opinion.pdf • Summary: https://patentdocs.org/2022/10/07/mylan-pharmaceuticals-inc-v-merck-sharp-dohme-corp-fed-cir-2022/ • Mylan's unsuccessful en banc petition: http://fedcircuitblog.com/wp-content/uploads/2022/11/21-2121-Petition.pdf
  • Separately, an appeal docketed as No. 23-1013 is listed by litigation aggregators as filed 2022-10-05 and terminated 2023-05-11. The termination date aligns with the 2023-03-14 settlement between Merck and Mylan (joint motion for indicative ruling filed 2023-03-16 in N.D. W. Va., 1:19-cv-00101). I could not verify from a primary source what the 23-1013 appeal raised; I flag it rather than characterize it.
• Defensive value: This is the whole ballgame — and it went to Merck. Mylan brought the best art that exists (Merck's own earlier disclosure, which names both sitagliptin and phosphoric acid), tried it on express anticipation, inherent anticipation, and three-reference obviousness, and lost every claim. The CAFC affirmance made the "957 salts" holding precedential and, worse for a challenger, the common-ownership/antedation ruling gutted the obviousness case not just for Mylan but structurally — any challenger leaning on the '871 patent as § 103 art runs into the same § 103(c)(1) wall, and any challenger leaning on WO '498 as § 102(a) art runs into the same reduction-to-practice evidence. There is no live claim-invalidity theory left in this record. Trial documents: https://storage.courtlistener.com/recap/gov.uscourts.wvnd.46385/gov.uscourts.wvnd.46385.175.0.pdf (Merck's IPR statements quoted back against it on enablement in the N.D. W. Va. bench case; Doc. 204 is the merits opinion finding Mylan infringed claim 3).

IPR2020-01045 — Teva Pharmaceuticals USA, Inc. et al. v. Merck Sharp & Dohme Corp.

• Type: Inter Partes Review, filed as a "me-too" joinder petition to IPR2020-00040 (35 U.S.C. § 315(c); 37 C.F.R. § 42.122)
• Filed: 2020-06-10
• Status (verbatim from aggregator data): "Terminated-Settled." Termination date 2020-12-08.
• Judge panel: Snedden, Pollock, Majors (panel that acted on the parallel joinder motions)
• Petition grounds: Claims 1–4, 17, 19, and 21–23 — "substantially identical" to Mylan's, per the Board's own characterization: same WO '498 / '871 anticipation ground; same Edmondson + Bastin + Brittain obviousness ground. Teva/Watson agreed to a "silent, understudy role," to raise no new arguments, to rely on Mylan's expert (Dr. Chorghade), to withdraw its own opening declaration, to take no speaking role at hearing, and to seek Board authorization for any independent paper.
• Institution decision: Instituted 2020-09-01, and Motion for Joinder granted the same day — i.e., Teva was never given a separate track; it was folded into the Mylan IPR.
• Final Written Decision: None issued in this docket. The Board's paper trail shows a Termination Decision Document dated 2020-12-08, closing the case on settlement.
• Settlement / termination: Settled. Terms are not public. This is consistent with the broader sitagliptin MDL pattern in which Merck resolved with most generics; here the parties also jointly moved to terminate the PTAB case.
• Appeal: None attributable to this docket.
• Defensive value: Neutral-to-adverse. Teva bought peace rather than a ruling, so this docket yields no invalidity ammunition. But it did not disturb the Mylan FWD — the joined parties "remain[ed] joined" in the lead proceeding per the Board's later termination paper, which is why Mylan's loss governs.

IPR2020-01060 — Dr. Reddy's Laboratories, Inc. and Dr. Reddy's Laboratories, Ltd. v. Merck Sharp & Dohme Corp.

• Type: Inter Partes Review, joinder petition to IPR2020-00040
• Filed: 2020-06-11 (DRL represented to the Board that it was not time-barred, supporting joinder over Merck's objection)
• Status: Terminated 2021-05-10 by Board termination decision, following issuance of the lead FWD. The termination document confirms "Dr. Reddy's … were joined as parties to this proceeding via Motion for Joinder in IPR2020-01060" and that "The Dr. Reddy's and Sun parties remain joined." (I retrieved this termination paper filed under the ‑01072 docket; the same event closes both understudy dockets. Flagging that filing-folder nuance.)
• Judge panel: Snedden, Pollock, Majors
• Petition grounds: Claims 1–4, 17, 19, and 21–23 — identical to Mylan's. Same two grounds. Same understudy stipulations.
• Institution decision: Instituted and joined 2020-09-01 (same date as ‑01045 and ‑01072). Merck opposed joinder and sought party discovery; the Board granted joinder and treated DRL as a true "me-too" petitioner.
• Final Written Decision: None issued in this docket. DRL's rights and liabilities were adjudicated through the IPR2020-00040 FWD as a joined party.
• Settlement / termination: Terminated administratively with the lead proceeding; no separate public settlement terms disclosed for the PTAB docket.
• Appeal: None separately attributable.
• Defensive value: Confirms the pattern — a second would-be challenger abandoned its own track and rode Mylan's petition, then accepted the adverse outcome. No new art, no new claim construction, no new result.
• OCR caveat: The Board's institution decision in the Sun matter refers in passing to "Dr. Reddy's Laboratories, Inc. v. Merck Sharp & Dohme Corp., IPR2020-01660," while the joinder briefing and the joinder reply both refer to Dr. Reddy's as IPR2020-01060. I treat ‑01060 as the operative number and flag the ‑01660 string as an apparent transcription error in the Board's paper rather than a fifth proceeding I can substantiate.

IPR2020-01072 — Sun Pharmaceutical Industries Ltd. v. Merck Sharp & Dohme Corp.

• Type: Inter Partes Review, joinder petition to IPR2020-00040
• Filed: 2020-06-12
• Status: Terminated 2021-05-10 (Board Termination Decision Document 14)
• Judge panel: Sheridan K. Snedden, Robert A. Pollock, and Timothy G. Majors (named on the face of the 2020-09-01 institution/joinder decision, authored by APJ Majors)
• Petition grounds: Claims 1–4, 17, 19, and 21–23 — identical to Mylan's; same WO '498 / '871 anticipation ground and same Edmondson + Bastin + Brittain obviousness ground. Sun, DRL, and Teva jointly represented to the Board that they were "true me-too Petitioners" who "rely on the same prior art and arguments as Mylan, submitted substantially identical petitions and identical expert testimony, and agreed to raise no new arguments."
• Institution decision: Instituted 2020-09-01 (35 U.S.C. § 314) and Motion for Joinder granted (35 U.S.C. § 315(c); 37 C.F.R. § 42.122). The Board rejected Merck's opposition, including Merck's argument that joinder should be denied because of the discovery it wished to take. Decision text: https://www.docketalarm.com/cases/PTAB/IPR2020-00040/Inter_Partes_Review_of_U.S._Pat._7326708/docs/09-01-2020-Board/Order-46-IPR2020_01072_Decision___Granting_Institution_of_Inter_Partes_Review_Granting_Motion_for_Joinder.pdf
• Final Written Decision: None issued in this docket; outcome flowed from the Mylan FWD.
• Settlement / termination: Terminated administratively 2021-05-10, four days after the lead FWD.
• Appeal: None separately attributable.
• Defensive value: This docket contains the most useful roadmap paper for a defendant — it lays out the joinder conditions, the identical grounds, and the litigation backdrop (In re Sitagliptin Phosphate ('708 & '921) Patent Litig., C.A. No. 19-md-2902-RGA (D. Del.); suits against Mylan, Teva, Sun, Watson, Dr. Reddy's, Apotex, Par, Sandoz, among others).

Strategic summary

Claim status after the PTAB. Nothing was canceled. Every one of the 24 claims of US 7,326,708 remains in force.

So the surviving claims are all of them. The untested residue (5–16, 18, 20, 24) is the salt-characterization and monohydrate-processing ladder — the part of the portfolio that a generic formulator would care about most — and it is unadjudicated at the Board. That is the only genuine gap, and it is now academic given the term.

Estoppel landscape. Under § 315(e)(2), Mylan — plus Teva/Watson, Dr. Reddy's, and Sun, each of whom was joined as a party to IPR2020-00040 — and their privies and real parties in interest are estopped in any later civil action from asserting any ground they raised or reasonably could have raised: (i) Edmondson (WO 2003/004498 and US 6,699,871) under § 102; (ii) Edmondson + Bastin + Brittain under § 103. That estoppel attached upon FWD and became final on 2022-09-29 following the CAFC affirmance; Mylan's March 2023 settlement and license moot it as to Mylan specifically but does not revive it for anyone else. For a non-privy defendant there is no statutory estoppel — but there is something close to its practical equivalent: the antedation and § 103(c)(1) rulings travel with the art. A new defendant citing WO '498 as § 102(a) or US 6,699,871 as § 103 art will hit the identical findings, and the precedential CAFC opinion is directly adverse on the "957 salts" point. Available grounds, realistically: different art entirely (a distinct § 102(b) or § 102(e) reference), on-sale/public-use, inequitable conduct, or § 112 written description/enablement — and note that Mylan already tried the § 112 route in district court (arguing claims 1–3 and 19 were not enabled across "all of its physical forms") and lost.

Pattern signals. A single lead petitioner (Mylan) carried the attack; three branded-generic competitors (Teva/Watson, Dr. Reddy's, Sun) filed identical me-too joinder petitions rather than independent cases — a signal that the field converged on the same art and nobody had a differentiated theory. The patent owner, Merck, was an aggressive PTAB defender: it opposed every joinder motion, sought party discovery, argued antedation, invoked § 103(c)(1) common ownership, and then successfully appealed-proofed the FWD through the Federal Circuit. There is no defensive aggregator in this chain — specifically, no Unified Patents IPR on the '708 patent appears anywhere in the record, contrary to the attribution in the previously generated section. If a defendant's counsel is working from a chart that says "Unified Patents IPR2020-01045," that chart is wrong and should be corrected before it is cited.

Recommended next steps

1. If you are a defendant holding a demand letter or an infringement suit on the '708 patent: there are no canceled claims to point to. Do not represent to a client or a court that any claim of this patent has been invalidated — it has not. Quote the actual disposition instead: IPR2020-00040, Paper 41 (2021-05-07), No challenged claim unpatentable, aff'd, Nos. 21-2121/-2122/-2123 (Fed. Cir. Sept. 29, 2022). Federal Circuit opinion: https://www.mololamken.com/assets/htmldocuments/Mylan%20v.%20Merck%20Opinion.pdf
2. The IPR window has effectively closed. The adjusted expiration is 2026-11-24 (per Google Patents/Orange Book) with pediatric exclusivity to 2027-05-24 — roughly two months from today. Because a petition filed today triggers a § 314(b) institution deadline and a statutory one-year trial deadline from institution (§ 316(a)(11)), no IPR filed now can yield an FWD before expiration. PGR is unavailable (this is a pre-AIA patent with a 2003 priority date — PGR applies only to effective filing dates on or after 2013-03-16), and CBM is unavailable (not a financial-services business-method patent, and the CBM program sunset on 2020-09-16). The only AIA route is IPR, and it is now pointless.
3. Check § 315(b) before doing anything else. Any defendant served with a complaint alleging infringement more than one year ago is time-barred from filing an IPR. Given the 2019 MDL wave of sitagliptin complaints, most legacy defendants are barred already.
4. For non-privy defendants with fresh art, the live questions are not PTAB questions — they are district-court questions: whether your art is outside the Edmondson family (to avoid the antedation and § 103(c)(1) findings) and whether any § 112 theory survives the N.D. W. Va. bench record. Start from the bench opinion at https://storage.courtlistener.com/recap/gov.uscourts.wvnd.46385/gov.uscourts.wvnd.46385.204.0.pdf and Doc. 175 (Merck's IPR statements being used against it on enablement) before commissioning new expert work.
5. Verify the two open data questions before relying on them in a filing: (a) what No. 23-1013 was and why it terminated 2023-05-11 (likely the Mylan settlement, unverified); and (b) the exact citation/pagination of the IPR2020-00040 FWD (Paper 41; reported at 2021 WL 1833325) from PTAB E2E. I retrieved the substance from reliable secondary reproductions, not from a live PTAB E2E or PatentCenter API response, and ODP's trial-proceedings feed — as supplied — does not index these 2020-filed, now-terminated proceedings.

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

Let me try to locate the specific USPTO assignment records (reel/frame) for this patent.

Let me try to pin down the specific recorded assignment entries and correspondents.

Ownership Analysis — US 7,326,708 B2

Sourcing caveat (read first): My retrieval environment could not render the USPTO Assignment Center's per-record extract (reel/frame, execution date, correspondent of record). What I can rely on is the Google Patents "Legal Events" ledger, which is derived from the USPTO assignment record and names the conveyance type, assignor, assignee, and recording date for each link. I therefore reconstruct the chain below without inventing reel/frame numbers or correspondent attorneys. Where the task-required fields (reel/frame, correspondent) are not obtainable, I mark them [not retrieved] rather than guess. Direct lookup: https://assignmentcenter.uspto.gov/ and https://assignment.uspto.gov/patent/index.html (search "7326708").

Inventors

All seven are named on the issued patent as filed by Merck & Co., Inc. (filing date 2004-06-23; PubChem assignee field: "MERCK & CO INC (US)"). Two of them (Ferlita, Wenslow) go on to appear as inventors on separate Merck applications covering other salt forms/polymorphs of the same molecule — this is a normal "polymorph team" pattern at a single pharma, not a departure signal.

Unusual-pattern check: I found no evidence of any of these inventors departing the assignee within 12 months of filing, and no employee-inventor-to-third-party assignment. Marked [unclear] only because I could not pull personnel/assignment metadata (reel/frame) from the Assignment Center. There is no fire-sale tell here.

Original assignee

Merck & Co., Inc. (Rahway / Whitehouse Station, NJ), later renamed Merck Sharp & Dohme Corp. and reorganized into Merck Sharp & Dohme LLC.

• Product embodying the claims: Yes — the dihydrogenphosphate monohydrate of this molecule is sitagliptin phosphate, the active ingredient in JANUVIA (NDA 021995, approved Oct 2006), JANUMET / JANUMET XR, JUVISYNC, and STEGLUJAN. The '708 patent is listed in the Orange Book against all of these (DS/DP/U-codes; expiry Nov 24, 2026; pediatric to May 24, 2027).
• Primary line of business: Global research-based pharmaceutical manufacturer.
• Current status: Operating. Not bankrupt, not acquired by a third party. The only changes are the internal corporate reorganizations reflected in the assignment ledger below.
• Note on stale databases: Drugs.com and PubChem both still list the "Current Assignee" as "Merck & Co., Inc. (Rahway, NJ)." That is out of date — it reflects the original-issuance assignee, not the post-2022 owner (Merck Sharp & Dohme LLC).

Assignment timeline

Each link below is taken from the Google Patents Legal Events ledger (USPTO-derived). Reel/frame and correspondent were not retrievable in this environment — flagged per entry.

• 2004-06-23 (filing/priority to Merck; no separate recorded document at issue) — Reel [not retrieved]

  • Conveyance: Application filed / original ownership by assignee
  • Assignor: Inventors (obligation to assign)
  • Assignee: Merck & Co., Inc.
  • Correspondent: [not retrieved]
  • Context: Original corporate ownership; the patent was filed as a Merck & Co., Inc. application.

• 2007-12-05 (recorded) — Reel [not retrieved]

  • Conveyance: Assignment of assignors' interest ("ASSIGNMENT OF ASSIGNORS' INTEREST (SEE DOCUMENT FOR DETAILS)")
  • Assignors: Cypes, Stephen Howard; Chen, Alex Minhua; Ferlita, Russell R.; Hansen, Karl; Lee, Ivan; Vydra, Vicky K.; Wenslow, Robert M., Jr. (all seven inventors)
  • Assignee: Merck & Co., Inc.
  • Correspondent: [not retrieved] — this is exactly the field you'd want for a repeat-player check; I could not surface it.
  • Context: Inventor-to-company assignment (perfection of title). Note the ~3.5-year gap between filing (2004-06-23) and recording (2007-12-05) — late, but common for confirmatory/clean-up recordings ahead of issuance (2008-02-05). No third party involved.

• 2010-01-29 — Reel [not retrieved]

  • Conveyance: Change of Name (see below)
  • Assignor: Merck & Co., Inc.
  • Assignee: Merck Sharp & Dohme Corp.
  • Correspondent: [not retrieved]
  • Context: Internal reorg — change of name only. No change in beneficial ownership.

• 2012-08-27 — Reel [not retrieved]

  • Conveyance: Merger
  • Assignor: Merck Sharp & Dohme Corp.
  • Assignee: Schering Corporation
  • Correspondent: [not retrieved]
  • Context: Internal reorg — merger (the post-2009 Schering-Plough/Merck integration; MSD Corp. merged into Schering Corp.).

• 2012-08-29 — Reel [not retrieved]

  • Conveyance: Change of Name
  • Assignor: Schering Corporation
  • Assignee: Merck Sharp & Dohme Corp.
  • Correspondent: [not retrieved]
  • Context: Internal reorg — change of name only (Schering Corporation renamed back to Merck Sharp & Dohme Corp.; address moves to 126 East Lincoln Ave., Rahway, NJ 07065).

• 2022-08-08 — Reel [not retrieved]

  • Conveyance: Merger
  • Assignor: Merck Sharp & Dohme Corp.
  • Assignee: Merck Sharp & Dohme LLC
  • Correspondent: [not retrieved]
  • Context: Internal reorg — merger into an LLC (the current owner of record; corroborated by non-US registers citing a "certificate of merger dated April 07, 2022" and an "application for change of title dated June 29, 2022").

Six recorded events; every one is an intra-Merck corporate action or the original inventor assignment. There is no transfer to any third party at any point in the chain.

De-confliction note: A "Worldwide Assignment" recorded at reel 044562/0376 (executed Oct 2015, inventor Rongze Kuang, Merck docket 23792) surfaced in my searches. That record belongs to a different Merck patent (a 2,2-difluorodioxolo A2A antagonist), not to US 7,326,708. Do not attribute it to this patent.

Timeline diagram

timeline
    title Ownership of US 7326708
    2004 : Filed by Merck and Co Inc
    2007 : Inventors assign to Merck and Co Inc
    2008 : Patent issued Feb 5
    2010 : Renamed Merck Sharp and Dohme Corp
    2012 : Merged into Schering Corporation
         : Renamed Merck Sharp and Dohme Corp
    2022 : Merged into Merck Sharp and Dohme LLC

NPE / troll-pattern signals

Verdict

Operating-company assertion.

The entire recorded chain — inventor assignment to Merck & Co., Inc. (recorded 2007-12-05), then pure intra-Merck Change of Name (2010-01-29) and Merger (2012-08-27; 2012-08-29) steps, ending in Merck Sharp & Dohme LLC (2022-08-08) — contains zero third-party transfers and zero NPE/aggregator entities. The current owner ships a product embodying the claims (JANUVIA/JANUMET/JUVISYNC/STEGLUJAN, sitagliptin phosphate) and is the plaintiff enforcing the patent against generic competitors, which is the textbook operating-company-assertion posture. Confidence in the ownership conclusion is high; the only open item is the correspondent-of-record field, which I could not retrieve and which should be confirmed directly at https://assignmentcenter.uspto.gov/ (search "7326708") — I decline to state a correspondent name I could not verify.

I'll search for the USPTO record of US 7,326,708 and the prior-art references cited against/within it.

Now let me pin down the specific "References Cited" on the face of the patent and the key IPR grounds.

A. Scope note before the analysis

I could not open PatentCenter/PAIR directly in this session (tool budget exhausted), so the reference set below is drawn from the USPTO-derived mirrors of the '708 front-page citation data — PubChem's patent record, which reproduces the "Citations" block for US‑7,326,708, and the PTAB petition exhibits filed against the patent. I confined retrieval to patent number 7326708 exactly; no neighbor numbers (e.g., 7,326,707 / 7,326,709) were substituted.

Two things I flag up front, per the operating rules:

1. Contradiction with my earlier section. I previously wrote that IPR2020‑01045, ‑01060 and ‑01072 were Unified Patents petitions. The retrieved IPR2020‑01060 petition document carries attorney docket "REDDY 7.1R‑024" and is supported by the Declaration of Joseph M. Fortunak, Ph.D., with EX1001 = the '708 patent, EX1004 = WO '498, EX1007 = US 6,699,871. That indicates Dr. Reddy's Laboratories, not Unified Patents. Google Patents' assignee tag and the petition paper disagree; I report both rather than picking one.
2. Citation-list composition. The PubChem citation block for this patent mixes items that cannot be § 102 prior art (2005–2007 documents). I treat those separately in Section D.

B. The reference set actually cited on/in US 7,326,708

PubChem's record for US‑7,326,708 (https://pubchem.ncbi.nlm.nih.gov/patent/US7326708) lists nine "Citations":

In addition, the specification itself (Background / Detailed Description) expressly cites as prior art:

C. Reference-by-reference § 102 analysis

Claims at issue, for mapping: claim 1 (salt genus or hydrate), 2 (R‑salt), 3 (S‑salt), 4 (crystalline monohydrate of claim 2), 5–8 (XRPD), 9–11 (¹³C CPMAS), 12–14 (¹⁹F MAS), 15–16 (TGA/DSC), 17–18 (compositions), 19–20 (Type 2 diabetes methods), 21–23 (salt-forming process / product-by-process), 24 (IPA/water monohydrate crystallization).

C‑1. WO 03/004498 A1 — the core § 102 reference

• Full citation: WO 03/004498 A1, "Beta‑amino tetrahydroimidazo(1,2‑a)pyrazines and tetrahydrotriazolo(4,3‑a)pyrazines as dipeptidyl peptidase inhibitors for the treatment or prevention of diabetes," Merck & Co., Inc.; inventors Edmondson, Fisher, Kim, MacCoss, Parmee, Weber, Xu.
• Dates: PCT/US2002/021349 filed 5 Jul. 2002; priority US 60/303,474 filed 6 Jul. 2001; published 16 Jan. 2003.
• Description: Genus of DP‑IV‑inhibiting β‑amino tetrahydrotriazolo[4,3‑a]pyrazines. Specifically discloses 4‑oxo‑4‑[3‑(trifluoromethyl)‑5,6‑dihydro[1,2,4]triazolo[4,3‑a]pyrazin‑7(8H)‑yl]‑1‑(2,4,5‑trifluorophenyl)butan‑2‑amine (the '708 free base), claims it and its "pharmaceutically acceptable salts," lists eight "particularly preferred" acid salts including phosphoric acid, and exemplifies the hydrochloride salt (Example 7) as a 1:1 salt.
• § 102 posture: § 102(a) printed publication (published 16 Jan. 2003, before the 24 Jun. 2003 priority date); alternatively § 102(e) as a PCT designating the US published in English. Not § 102(b) — it published less than one year before the '708 filing.
• Claims it potentially anticipates (as asserted by petitioners): 1, 2, 3, 17, 19, 21, 22, 23. The theory is that the recited genus list and the "particularly preferred" salt sub‑list collapse into one comprehensive list whose members include (R)‑sitagliptin phosphate ("dihydrogenphosphate" = phosphoric acid addition salt), and that a POSA would "at once envisage" the salt (petition cites In re Petering, 301 F.2d 681, and Glaxo, 376 F.3d 1348).
• Claims it does NOT plausibly anticipate: 4–16 and 18, 20, 24. WO '498 discloses nothing about a crystalline monohydrate, nor about the XRPD d‑spacings, ¹³C/¹⁹F solid‑state NMR shifts, or TGA/DSC characteristics that claims 5–16 are defined by. The record reflects this: petitioners expressly limited their analysis to claims 1–4, 17, 19, 21–23 (Fortunak Declaration, IPR2020‑01060).
• Outcome: The PTAB/CAFC line (Mylan appeals Nos. 21‑2121, 21‑2122, 21‑2123, affirmed 29 Sep. 2022) upheld the Board's rejection of the "at once envisage" anticipation theory; the patent survived. I rely on my earlier section for that procedural history.

C‑2. US 6,699,871 B2 — the US counterpart, § 102(e)(2)

• Full citation: U.S. Pat. No. 6,699,871 B2, "Beta‑amino tetrahydroimidazo(1,2‑a)pyrazines and tetrahydrotriazolo(4,3‑a)pyrazines as dipeptidyl peptidase inhibitors for the treatment or prevention of diabetes," Merck & Co., Inc. (Edmondson et al.). Grant date 2 Mar. 2004 (moderate confidence — confirm in PatentCenter; the number/date pairing is consistent across the retrieved aggregator records).
• Dates: Same family as WO '498; filed from the same priority chain (US 60/303,474, 6 Jul. 2001).
• Description: Substantively identical specification to WO '498 (petition: "identical in all relevant material respects"); claim 17 recites sitagliptin "or a pharmaceutically acceptable salt thereof."
• § 102 posture: § 102(e)(2) — granted on an application filed before the '708 inventors' invention date, and granted before the '708 filing date. Petitioners cite Associated British Foods, PLC v. Cornell Research Found. Inc., IPR2019‑00578, Paper 25 (P.T.A.B. 25 Jul. 2019) to establish the § 102(e)(2) framework.
• Claims it potentially anticipates: 1, 2, 3, 17, 19, 21, 22, 23 — same mapping as WO '498. It is the more dangerous of the two procedural postures because it is a granted U.S. patent and is itself Orange‑Book‑listed alongside the '708 patent for JANUVIA/JANUMET (petition EX1008/EX1009).
• Claims it does NOT anticipate: 4–16, 18, 20, 24 (no crystalline monohydrate disclosure).

C‑3. EP 1 412 357 A1 / B1 (EP family member of WO '498)

• Full citation: EP 1 412 357; application EP 02749813.8; EP publication 28 Apr. 2004; grant 22 Mar. 2006; expiry 5 Jul. 2022.
• Relevance: Same disclosure as WO '498, published after the '708 priority date. As such it is not § 102 prior art to the '708 claims on its own; it is useful as corroboration of the WO '498 disclosure content and for the salt‑form genus.

C‑4. Non‑patent literature cited in the specification

C‑5. References I could not verify — reported without characterization

D. Cited items that are not § 102 prior art to this patent (flag)

Four items in the citation block post‑date the 24 Jun. 2003 priority date and are, on their face, the applicant's own or third‑party later filings. They were presumably IDS/duty‑of‑disclosure items, and they cannot anticipate or render obvious the '708 claims:

Similarly, WO 2005/030127 A1 (Merck, priority 23 Sep. 2003) and US 2008/0227786 A1 (Ferlita, priority 16 Jan. 2004), which appear in the forward‑citation lists, are subsequent art and cannot be applied against the '708 claims.

E. Net conclusion on the most relevant prior art

1. The single most relevant reference is WO 03/004498 A1 (16 Jan. 2003) and its U.S. counterpart US 6,699,871 B2 (§ 102(e)(2)). Everything else in the citation block is either background literature, a cumulative Prous abstract, or a post‑dated document that cannot be prior art.
2. Anticipation exposure is claim‑limited. Theoretically at risk under § 102: claims 1, 2, 3, 17, 19, 21, 22, 23. Structurally immune to a WO '498/US '871 anticipation attack: claims 4–16 (the crystalline monohydrate and its XRPD/NMR/TGA/DSC fingerprint limitations), 18, 20, and 24 (which depend on that monohydrate or its crystallization conditions). As a matter of record, that anticipation theory did not carry the day — the PTAB and the Federal Circuit (Nos. 21‑2121/‑2122/‑2123, 29 Sep. 2022) rejected the "at once envisage" theory, and the N.D. W. Va. bench found Mylan's ANDA infringed claim 3.
3. No cited reference individually anticipates claims 4–16, 18, 20, or 24. If a validity challenge on those claims exists, it will be an obviousness case built on WO '498 in view of a salt‑screening/crystallization reference — not a § 102 case on the four corners of the '708 citation list.

F. Confidence and residual gaps

• High confidence: WO '498 and US 6,699,871 as the operative prior art; the claim mapping asserted against the '708 patent; the fact that no cited reference discloses the crystalline monohydrate.
• Moderate confidence: The 2 Mar. 2004 grant date of US 6,699,871; the characterization of US 2006/0287528 and WO 2005/072530.
• Not verified (declined to characterize): US 6,479,692 B1 and US 2003/0100563 A1. If you supply the USPTO PatentCenter "References Cited" image for the '708 front page, I can complete those two entries and confirm the examiner‑cited vs. applicant‑cited split.
• Contradiction on record: petitioner identity for IPR2020‑01060 (Unified Patents per Google Patents vs. Dr. Reddy's per the petition caption) — unresolved.

I'll verify the key prior-art disclosure and the litigation record before building the § 103 analysis.

Obviousness Analysis — US 7,326,708 B2 under 35 U.S.C. § 103

Scope note / discrepancy flags before the merits

1. Date mismatch. The header of this session states today is 2026‑09‑25; the task prompt states April 26, 2026. I am flagging the inconsistency rather than silently reconciling it. Nothing in the analysis below turns on the difference (the patent's own term runs to 2026‑11‑24, plus pediatric exclusivity per the earlier section).
2. Contradiction with the previously generated section. That section records Google Patents as labelling IPR2020‑00040 (Mylan) as "(Settlement)". The Federal Circuit opinion I retrieved identifies IPR2020‑00040 as the very Mylan IPR that produced the Board Decision under review — Decision, 2021 WL 1833325 (P.T.A.B. May 7, 2021), cited throughout Nos. 21‑2121/2122/2123. Google Patents' "Settlement" label therefore appears to describe a later disposition (or is simply inaccurate) rather than the FWD on the merits. Per the operating rules I treat the search results as ground truth and flag the conflict explicitly.
3. "The 400 Patent." The Canadian Federal Court decision quoted below refers to "the 400 Patent" — that is the Canadian family member CA 2,529,400 C, not US 7,326,708. Do not conflate them.
4. This is an analytical reconstruction of the § 103 case, not a merits holding. The actual adjudicated outcome is discussed in § 6 and materially constrains the theory.

1. Legal framework applied

Graham v. John Deere Co., 383 U.S. 1, 17‑18 (1966) (scope/content of prior art; differences; level of ordinary skill; objective indicia); KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 406, 421 (2007) ("finite number of identified, predictable solutions"); In re Magnum Oil Tools Int'l, Ltd., 829 F.3d 1364, 1381 (Fed. Cir. 2016) (motivation + reasonable expectation of success). The Board adopted the POSA definition in the Mylan IPR as: Ph.D. in chemistry/pharmacy etc. + 2 years' drug‑development experience including salt selection, or M.S. + 5 years, or B.S. + 7 years (Pet. 11; Ex. 1002 ¶¶ 45‑46, uncontested). I use that definition.

Pre‑AIA law governs (filed 2004‑06‑23), so § 102(a)/(e) and § 103(c)(1) are in play.

2. The prior-art set drawn from this page

Critical caution: every other reference on this page that is actually about a phosphoric acid salt / crystalline phosphate / monohydrate of sitagliptin is post‑priority and therefore not § 102/103 art: WO 2005/030127 A2 (Merck, prio. 2003‑09‑23), US 2006/0287528 A1 (Wenslow, prio. 2003‑09‑02), WO 2006/033848 A1 (prio. 2004‑09‑15), US 2008/0227786 A1 (Ferlita, prio. 2004‑01‑16), and WO 2005/003135 (published 13 Jan 2005). No pre‑priority reference in the page's prior‑art section explicitly discloses sitagliptin dihydrogenphosphate or its monohydrate. That is the central factual gap the whole obviousness case must bridge, and it is exactly the gap the Board and the District Court found unbridged.

3. The prima facie combinations

Combination A — Edmondson + Bastin → claims 1, 2, 17, 19, 21, 22, 23

Motivation. (i) Same field, same problem, and the same assignee — Edmondson is Merck's own disclosure of the exact molecule; a POSA engaged in developing sitagliptin would begin there. (ii) Edmondson identifies eight "particularly preferred" acids, and expressly names phosphoric acid among them — the salt-forming step is not left to a 53‑membered Berge list but to a self‑selected, small subset. (iii) Bastin supplies the why: salt formation "is a relatively simple chemical manipulation which may alter the physiochemical, formulation, biopharmaceutical, and therapeutic properties of a drug without modifying the basic chemical structure." The POSA's problem — converting a weakly basic API into a reproducibly manufacturable, stable solid — was a recognized, routine development problem.

Reasonable expectation of success, as Mylan framed it in the IPR: "the combined teachings of WO '498 and Bastin would motivate the POSA to use the following list of three salts as alternatives to the hydrochloride salt exemplified in WO '498: hydrobromide, sulfate, and phosphate," citing In re Fout, 675 F.2d 297, 301 (C.C.P.A. 1982) ("Express suggestion to substitute one equivalent for another need not be present"), and Pfizer (narrowing Berge's 53 anions to a few is obvious), reinforced by Grunenthal v. Alkem, 919 F.3d 1333, 1344 (Fed. Cir. 2019). Also In re Cyclobenzaprine, 676 F.3d 1063, 1071 (Fed. Cir. 2012) ("known options" from "a finite number of identified, predictable solutions").

Claims 21‑23 (process). Edmondson's Example 7 teaches salt formation from the free amine with acid in methanol at ambient temperature (~25 °C) — overlapping claim 21's "about 25‑100 °C" (In re Peterson, 315 F.3d 1325, 1329‑30 (Fed. Cir. 2003)) and claim 22's "C₁‑C₅ alkanol." Optimizing variables would be "routine experimentation" (In re Aller, 220 F.2d 454, 456‑57 (C.C.P.A. 1955)). Claim 23 (product‑by‑process) rises or falls with the product (Amgen v. F. Hoffmann‑La Roche, 580 F.3d 1340, 1369 (Fed. Cir. 2009)).

Combination B — Edmondson alone (or + Berge/Bastin) → claim 2 ((R)); claim 3 ((S))

For claim 2, Mylan's petition states: "(R)‑sitagliptin phosphate would have been obvious … as well as the fact that all 33 compounds of WO '498 (including sitagliptin) are in the (R) configuration." If Edmondson's sitagliptin disclosure is taken at face value, the (R) stereochemistry is given, and claim 2 adds nothing beyond claim 1 plus the disclosed stereochemistry.

For claim 3 ((S)‑sitagliptin DHP) the theory is materially weaker: it requires either (a) a racemate/mixture disclosure to resolve, or (b) an expectation that the DHP salt of the unwanted enantiomer was worth making. Edmondson discloses no racemate of sitagliptin and no racemate of any sitagliptin salt — a point the Board expressly found dispositive.

Combination C — Edmondson + Bastin + Brittain → claims 4, 5‑16, 18, 20

Motivation for the crystalline monohydrate (claim 4). Edmondson states the salts "may exist in more than one crystal structure" and "may be in the form of hydrates"; Brittain then supplies the third nexus — hydrates are common, well-characterized, and pharmaceutically desirable (better physical stability, non-hygroscopic handling). The POSA's ordinary development sequence (salt screen → polymorph/hydrate screen) would reach the monohydrate without inventive insight. Claims 5‑16 are pure characterization claims (XRPD d‑spacings, ¹³C CPMAS, ¹⁹F MAS shifts, TGA/DSC curves) directed to inherent solid‑state properties of the same compound; they add no separate patentable weight once claim 4 is obvious.

Combination D — Edmondson + conventional crystallization → claim 24

Claim 24 recites crystallizing the Formula II salt at 25 °C from IPA/water with water above 6.8 wt%, recovering, and drying. Precipitation of a salt from an alcohol/water antisolvent system, with optimization of the water content and cooling profile, is textbook (In re Aller). Note: claim 24 was not challenged in the IPR, so there is no adjudicated record either way on it.

4. Objective indicia (Graham factor 4)

Available to Merck: commercial success (JANUVIA / JANUMET / JUVISYNC; the earlier section notes Orange Book listing to 2026‑11‑24 plus pediatric exclusivity to 2027‑05‑24); unexpected results — aqueous solubility of ~72 mg/mL and the physical/chemical stability stated at col. 4 of the patent; long-felt need for a manufacturable Type 2 diabetes salt form; industry praise/adoption. The Court of Appeals noted the Board considered Merck's unexpected-results evidence but held "there is no need to reach objective indicia of nonobviousness where the petitioner has not made a showing necessary to prevail on threshold obviousness issues."

The nexus weakness is real and was identified by the Canadian Federal Court in the parallel proceeding: "[t]he only comparison made in the 400 Patent is between the DHP salt of sitagliptin crystalline monohydrate and the sitagliptin free base and hydrochloride salt" (¶ 91), and the patent is not "devoted … to the selection of sitagliptin over the other compounds of WO498" (¶ 92). A head‑to‑head phosphate‑vs‑HCl stability comparison is asserted but not tabulated in the specification.

5. The decisive rebuttal: why these combinations did not carry the day

This is where the analysis must be honest, because the record is unusually well developed.

(a) The reference-genera problem. Merck's counter was arithmetic: "33 compounds × 8 preferred salts, taking into account various stoichiometric possibilities, would result in 957 salts, some of which may not even form." The Federal Circuit agreed the '498 disclosure does not meet In re Petering's "at once envisage" standard for a limited class, and that 957 predicted salts "is a far cry from the 20 compounds 'envisaged' by the narrow genus in Petering."

(b) No direction to the two critical selections. The court relied on Mylan's own expert (Dr. Chorghade), who conceded that "nothing in Edmondson directs a skilled artisan to sitagliptin from among the 33 listed DP‑IV inhibitors," and that "nothing in Edmondson singles out phosphoric acid or any phosphate salt of any DP‑IV inhibitor, and the list of 'pharmaceutically preferred' salts comes 44 pages earlier in the specification."

(c) Unpredictability defeats reasonable expectation of success. The same expert admitted salt formation is a "trial and error process." Merck's expert Dr. Matzger showed sitagliptin phosphate salts exist in other stoichiometries (3:2 and 2:1), so the 1:1 salt is not the inevitable or only product — destroying the "obvious to try with predictable result" framing. This is the Valeant v. Watson / Sanofi‑Synthelabo v. Apotex / Pfizer v. Mylan line: Pfizer applies where the anion sub‑set is small and the reference teaches the beneficial property; here there was "nothing in the prior art to suggest … that a phosphate salt of sitagliptin (if it could form) would have had beneficial properties."

(d) Claim 4 failed outright on the merits. "Mylan provided no rationale to explain why a person of ordinary skill would have been motivated to make the claimed crystalline monohydrate form … and failed to show that a skilled artisan would have had a reasonable expectation of success" (Board, aff'd).

(e) Claim 3 failed outright on the merits. Neither Edmondson nor Bastin "disclosed anything related to (S)‑sitagliptin or even a racemic mixture of any sitagliptin salt."

(f) The procedural escape hatch. For claims 1, 2, 17, 19, 21‑23, the Board never reached the merits of obviousness: Merck antedated Edmondson (so it was not § 102(a) art), and the '871 patent, though § 102(e) art, was disqualified by the common‑ownership exception of pre‑AIA § 103(c)(1). That is a procedural/ownership bar, not a merits holding of non‑obviousness. Analytically this is the single most important caveat: the Federal Circuit did not affirmatively hold claims 1, 2, 17, 19 and 21‑23 non‑obvious on the merits. A different challenger, a different primary reference, or a post‑AIA posture would remove the antedation/§ 103(c) shield entirely and force the merits.

(g) Parallel proceedings. The N.D. W. Va. bench found "Mylan has not established by clear and convincing evidence that a POSA would have been motivated to modify the reference genus in claim 17 of the '871 patent to achieve the 1‑to‑1 DHP salt of sitagliptin … Nor, in doing so, would she have had a reasonable expectation of success." The Canadian Federal Court went the other way on the equivalent question, finding the selection lacked an inventive step — invoking the ΔpKa rule, the short (two‑month) and narrow salt screen over only 8 preferred acids, and the absence of evidence of salification difficulty. That divergence is instructive: the outcome is highly sensitive to the evidentiary record on unpredictability and on whether the reference list is treated as "closed."

6. Bottom line

• Strongest prima facie case — Edmondson (WO 03/004498 / US 6,699,871) in view of Bastin, optionally with Brittain — supported by Pfizer, Grunenthal, In re Fout, In re Peterson and In re Aller. It reaches claims 1, 2, 17, 19, 21, 22, 23 (salt, (R) salt, compositions, treatment method, process) and, with Brittain added, claims 4 and 5‑16/18/20 (crystalline monohydrate and its characterization).
• Weakest link — claim 3 ((S)‑sitagliptin DHP), because no pre‑priority reference discloses an (S)‑configured or racemic sitagliptin salt; and claim 4, because hydrate formation and monohydrate identity were found unpredictable as applied to this compound.
• The strongest rebuttal is the combination of (i) the 957‑salt genus with no direction to either sitagliptin or phosphoric acid, (ii) admitted unpredictability and the demonstrated 3:2/2:1 alternative stoichiometries, and (iii) objective indicia (commercial success, ~72 mg/mL solubility, stability) with a defensible nexus to the salt/hydrate selection.
• What the case law does not settle: whether claims 1, 2, 17, 19 and 21‑23 are obvious on the merits. The Federal Circuit's affirmance on those claims rested on antedation and pre‑AIA § 103(c) common ownership, not on the Graham analysis.

Uncertainty I will not paper over: I have not located any live 2026 CAFC or PTAB docket specific to US 7,326,708, and I decline to infer one. I also cannot verify from the retrieved materials whether any current post‑expiry challenge (e.g., an obviousness-type double‑patenting or § 103 case brought outside the § 103(c) shield) exists. If a docket number is supplied, I can verify it directly.