US 12049652

I'll research US 12049652 using the authoritative patent text provided and live searches for litigation/docket status.

Let me get the specific claim text and any 2026 CAFC activity.

Let me find the full claim set (independent claims) of the '652 patent.

US Patent 12,049,652 B2 (= US 12049652)

I searched for the literal identifier 12049652 / 12,049,652. Note on formatting: the unpunctuated grant number US12049652B2 and the comma-formatted citation U.S. Patent No. 12,049,652 ("the '652 patent") used in PTAB/litigation papers refer to the same document — application 18/064,886, filed 2022-12-12, issued 2024-07-30. All sources below are cross-consistent on that application number and issue date, so I treat them as the same patent.

Bibliographic data

Family continuity (per the '652 specification and the PTAB institution decision): a long chain of continuations/divisions out of the '731 application, including US 10,865,400; US 11,041,149; US 11,066,656; and others. The '652 patent issued from a division of Ser. No. 17/327,568 (filed May 21, 2021).

Abstract

Caveat: the full text I was able to retrieve did not include the face-page Abstract verbatim. The specification's opening summary sentence is: "Modified PH20 hyaluronidase polypeptides, including modified polypeptides that exhibit increased stability and/or increased activity, are provided. Also provided are compositions and formulations and uses thereof." I cannot confirm the exact wording of the printed Abstract, so treat this as characterization rather than a verbatim quote.

Independent claims (plain language)

The only independent claim whose text I could verify authoritatively is claim 1, as reproduced in the PTAB's institution decision (quoting Ex. 1001 at 309:18–43):

1. A modified PH20 polypeptide, comprising one or more amino acid modifications in an unmodified PH20 polypeptide, wherein:

• the unmodified PH20 polypeptide consists of the amino acid sequence selected from the group consisting of SEQ ID NOs: 3, 7 and 32–66;
• amino acid modifications are selected from the group consisting of amino acid replacement(s), deletion(s), and/or insertion(s);
• the modified PH20 polypeptide comprises an amino acid replacement at a position corresponding to residue 320, with reference to amino acid positions set forth in SEQ ID NO:3;
• the replacement at the position corresponding to residue 320 is selected from the group consisting of H, K, R, and S;
• corresponding amino acid positions are identified by alignment of the PH20 polypeptide with the polypeptide having the amino acid sequence of SEQ ID NO:3; and
• the modified PH20 polypeptide has at least 91% sequence identity to a polypeptide having the amino acid sequence selected from the group consisting of SEQ ID NOs: 3, 7 and 32–66.

In plain terms, claim 1 covers: a variant of a soluble human PH20 hyaluronidase — the unmodified template being soluble PH20 (SEQ ID NO:3), full-length mature PH20 (SEQ ID NO:7), or any of the C-terminally truncated soluble PH20 forms (SEQ ID NOs 32–66) — in which (i) at least one mutation is present, (ii) the amino acid at position 320 (numbered per SEQ ID NO:3) is swapped to histidine, lysine, arginine or serine, and (iii) the variant still is at least 91% identical to one of those template sequences. Positions in other PH20 sequences are set by alignment to SEQ ID NO:3.

Key interpretive points drawn from the record:

• Claim 1 does not expressly require hyaluronidase activity. The PTAB noted the patent owner's position that "claims 2 and 6–29 do not require any hyaluronidase activity." The '652 specification contemplates both active mutants (>40% residual activity) and inactive mutants usable as contraceptive-vaccine antigens.
• Claim 2 is the same claim narrowed to at least 95% sequence identity.
• Claim 7, per petitioner's mapping in the PGR record, is limited to the single-substitution D320S variant; the other (non-disclaimed) claims encompass at least the single-substitution D320K variant.

Uncertainty on other independent claims: Claims 3–5 and 30–40 were statutorily disclaimed before institution, and I could not retrieve their text. Halozyme's New Jersey complaint represents that the patent "further claims methods of treatment and/or administration using modified PH20" (Count 7, ¶168), which suggests at least one method-type independent claim may have sat in the now-disclaimed claim range. I cannot confirm the number or wording of any independent claim other than claim 1.

Status, litigation and docket information

No Federal Circuit appeal of this patent was found. My searches of CAFC activity returned appeals involving other patents (e.g., 2026-1192, 2025-1487, 26-1033, 26-1884) and nothing identifying US 12,049,652. The Federal Circuit docket images turned up were unrelated cases. I would treat "no CAFC appeal identified for the '652 patent" as my finding, with the caveat that a very recent notice of appeal could exist without being indexed in what I retrieved.

PTAB — instituted post-grant review (this is the live validity proceeding):

• PGR2025-00033, Merck Sharp & Dohme LLC v. Halozyme, Inc. (Patent 12,049,652 B2). Petition filed March 7, 2025; institution granted October 1, 2025 on claims 1, 2 and 6–29 (claims 3–5 and 30–40 disclaimed by the patent owner, leaving those claims out of the trial).
• Grounds instituted: (1) lack of written description under §112; (2) non-enablement under §112; (3) obviousness under §103 over Halozyme's own US 7,767,429 ("the '429 patent") in view of Chao. The PTAB instituted on the §112 grounds and did not find the obviousness ground more likely than not.
• Panel: APJs Mitchell, Hardman, Valek. Petitioner counsel: Sidley Austin (Jeffrey Kushan et al.). Patent owner counsel: Sterne, Kessler, Goldstein & Fox; Halozyme; Quinn Emanuel.
• Oral argument reset to 1:30 p.m. ET, Thursday, July 23, 2026 (amended scheduling order).
• Petitioner's core theory: claims 1–40 cover a genus of between 10⁵⁹ and 10¹¹² distinct polypeptides (one substitution required at position 320 plus up to 41 additional optional substitutions), while the specification exemplifies essentially only single-substitution mutants (Table 8: 6,753 entries; Table 10: 3,380 inactive mutants) and offers only a "make-and-test" research plan for multiply-modified variants. The Board also preliminarily found the claims may not be entitled to pre-March 16, 2013 priority, which is what made the patent PGR-eligible in the first place — a finding that, if it holds, affects the effective filing date and the listed 2032 expiration.
• Related IPRs were instituted September 9, 2026 against three family members: IPR2026-00312 (US 10,865,400), IPR2026-00313 (US 11,041,149), IPR2026-00314 (US 11,066,656).

District court litigation:

• Halozyme, Inc. v. Merck Sharp & Dohme LLC (f/k/a Merck Sharp & Dohme Corp.), No. 2:25-cv-03179 (D.N.J.) — Halozyme asserts the '652 patent (Count 7) and numerous related patents against Merck's subcutaneous pembrolizumab (Keytruda Qlex™ / SC Keytruda, pembrolizumab + berahyaluronidase alfa-pmph, approved September 2025). Halozyme seeks damages and an injunction and alleges induced/contributory infringement via the product label.
• Google Patents' family page also flags first worldwide family litigation (Darts-IP family 47595052) and European activity: a December 2025 Munich Regional Court preliminary injunction against MSD regarding EP 2797622, with invalidation actions filed in Germany, the UK, the Netherlands and France; and a July 2026 District Court of The Hague refusal of a PI for Denmark/Sweden.

Date-caveat: you stated the current date as April 26, 2026. My searches returned materials dated through September 2026 (e.g., a Pearce IP update of September 9, 2026 reporting PTAB final written decisions invalidating four other Halozyme patents and the three September 2026 IPR institutions). I am reporting them because they came back in the search, but flag that they post-date your stated "today" — so treat anything after April 2026 as later-breaking and verify against PTAB/USPTO sources before relying on it.

Sources

• https://patents.google.com/patent/US12049652/en (authoritative full text supplied)
• PTAB institution decision, PGR2025-00033 (Oct 1, 2025): https://biologicshq.com/wp-content/uploads/2025/10/PTAB-PGR2025-00033-Institution-Decision.pdf
• PTAB amended scheduling order, PGR2025-00033 / -00039 (Nov 20, 2025; hearing reset to July 23, 2026): https://www.docketalarm.com/cases/PTAB/PGR2025-00039/.../AMENDED_SCHEDULING_ORDER/
• Merck petition, PGR2025-00033: https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557626](/patent/1557626)/...
• Halozyme complaint Count 7, No. 2:25-cv-03179 (D.N.J.): https://www.statnews.com/wp-content/uploads/2025/06/halozyme-v-merck-re-keytruda-patents.pdf
• PTAB institution coverage: https://biologicshq.com/ptab-grants-institution-of-three-pgrs-related-to-mercks-keytruda-qlex/
• Pearce IP status update (Sept 9, 2026): https://www.pearceip.law/2026/09/09/update-on-msds-challenges-to-us-halozyme-patents-in-sc-keytruda-pembrolizumab-dispute/
• Lexology/PTAB case summary: https://ai-lab.exparte.com/case/ptab/PGR2025-00033/doc/1005

I'll research known litigation involving US 12049652 using litigation-tracking sources.

Let me verify the district court case and confirm whether US 12049652 is asserted there, plus check the PTAB proceeding status.

Let me check the most recent status developments, including the May 2026 report and whether the '652 patent remains asserted.

Litigation Involving U.S. Patent No. 12,049,652

Patent identified literally: U.S. 12,049,652 B2, "PH20 polypeptide variants, formulations and uses thereof" (application 18/064,886), assigned to Halozyme, Inc. / Halozyme Therapeutics, Inc. Priority date 2011‑12‑30; adjusted expiration 2032‑12‑28.

Bottom line: US 12,049,652 (the '652 patent) is involved in two active U.S. proceedings arising from the same dispute: (1) a patent-infringement action in the District of New Jersey, and (2) a post-grant review at the PTAB. It is also within the worldwide family that is being litigated in Europe, though the European actions concern a different (EP) patent. No final merits outcome has been reached as to the '652 patent itself in either U.S. proceeding on the record located.

---

1. District Court Action (patent infringement / declaratory judgment)

Sources: Google Patents family-litigation notice (https://portal.unifiedpatents.com/litigation/New%20Jersey%20District%20Court/case/2%3A25-cv-03179); district-court complaint cited in PTAB exhibit lists ("Complaint for Patent Infringement and Declaratory Judgment of Patent Infringement, Halozyme, Inc. v. Merck Sharp & Dohme Corp., Civil Action No. 2:25-cv-03179-ES (D.N.J.), filed April 24, 2025"); Pearce IP and Lexology case summaries (https://biologicshq.com/ptab-grants-institution-of-three-pgrs-related-to-mercks-keytruda-qlex/ ; https://www.lexology.com/library/detail.aspx?g=af57d058-dd73-4761-abc4-7beea1b33ac1).

---

2. PTAB Post-Grant Review of the '652 Patent

Sources: Google Patents PTAB notice (https://portal.unifiedpatents.com/ptab/case/PGR2025-00033); Unified Patents PTAB case list (https://portal.unifiedpatents.com/ptab/caselist?judges.panel=Michael+A.+Valek); GreyB/IPVerse case detail (https://ipverse.greyb.com/ptab-web/cases/case-details/PGR2025-00033); Venable BiologicsHQ (https://biologicshq.com/ptab-grants-institution-of-three-pgrs-related-to-mercks-keytruda-qlex/); Docket Alarm PTAB docket (https://www.docketalarm.com/cases/PTAB/PGR2025-00033/Merck_Sharp_%26_Dohme_LLC_v._Halozyme_Inc/).

---

3. Related proceedings that do NOT directly involve US 12,049,652 (flagged for completeness)

These are part of the same worldwide dispute but concern different patents — I am listing them only so you do not conflate them with the '652 patent:

• Munich Regional Court preliminary injunction (Germany), December 4, 2025 — Halozyme obtained a PI against MSD's distribution/sale of Keytruda SC™ based on Halozyme's EP 2797622 ('622 patent), not US 12,049,652.
• MSD invalidation actions against EP 2797622 in the German Federal Patent Court and the UK High Court (August 2025), and later the Netherlands and France; District Court of The Hague dismissed a Halozyme PI application in July 2026 (per reporting).
• Other PTAB PGRs/IPRs filed by Merck against Halozyme MDASE patents (e.g., US 11,952,600, US 12,110,520, US 12,104,185, US 12,091,692, US 12,264,345, US 12,060,590, US 12,054,758, etc.). The '600 patent (PGR involving US 11,952,600) was reportedly held unpatentable in a May 12, 2026 FWD — this is a different patent from the '652 patent.

---

4. Summary of current status

• US 12,049,652 is actively litigated in both the New Jersey district court action (2:25-cv-03179-ES) and the instituted PTAB PGR (PGR2025-00033).
• No final merits ruling or judgment specific to the '652 patent has been located. The district case is pending; the PGR was instituted October 1, 2025 and remains pending.
• Ownership on both sides of the dispute: Halozyme, Inc. / Halozyme Therapeutics, Inc. (owner) v. Merck Sharp & Dohme / MSD (defendant-petitioner and accused infringer).

5. Caveats and uncertainties (important)

1. No other litigation involving US 12,049,652 specifically was located beyond the New Jersey action and PGR2025-00033. I did not find any CAFC appeal involving the '652 patent; the Federal Circuit materials returned in searching were unrelated (e.g., a Vivint/SB IP appeal). If an appeal or additional suit exists, I could not confirm it.
2. Date discrepancy in sources: The materials I retrieved include items dated after the current date stated in your instruction (e.g., reports dated May 2026, July 2026, and September 2026). I have reported them as found, with their stated dates, rather than "correcting" them. You should verify the most recent district-court docket (PACER, D.N.J. 2:25-cv-03179) and the PTAB docket for PGR2025-00033 for the authoritative, current posture, including whether the '652 patent has been dropped from the district-court case or whether any stay has been entered.
3. Stay posture: Commentary in the record suggests that a district-court stay is a significant possibility given the co-pending instituted PGRs, but I did not locate a confirmed stay order for the New Jersey case. Treat this as unconfirmed.
4. I could not verify current assignee/counsel changes beyond the PTAB counsels of record listed above.

I'll start by researching the PTAB docket for this patent using the identifiers I have.

Let me dig into the institution decision, panel, and claim-level details.

Now let me check whether a Final Written Decision has issued (the FWD deadline is ~2026-10-01, and today is 2026-09-26).

Proceedings overview

One AIA trial proceeding is on file for US 12,049,652 — PGR2025-00033 (Merck Sharp & Dohme LLC v. Halozyme, Inc.), status "Trial Instituted" / "Pending - Instituted," instituted 2025-10-01, no Final Written Decision yet as of 2026-09-26 (statutory FWD deadline ~2026-10-01). Fourteen of the patent's 40 claims — claims 3–5 and 30–40 — have already been statutorily disclaimed by the patent owner and are gone regardless of the PTAB outcome. Institution was granted only on the § 112(a) written-description and enablement grounds; the § 103 obviousness ground was denied institution. Bottom line for a defendant: the patent is not hardened — it is under active § 112(a) attack with a precedented loss record across its sibling patents, so an assertion letter built on claims 3–5 or 30–40 is worthless, and an assertion built on claims 1–2 or 6–29 rests on claims that three sibling patents in the same family have now lost on written description/enablement.

Data-integrity flag: the structured "PTAB proceedings on file" block (USPTO ODP) returned no AIA trial proceedings. That ingest is stale or incomplete. The Google Patents record for US12049652 itself lists "PTAB case PGR2025-00033 filed (Pending - Instituted)" under "Family has litigation," and the proceeding is independently confirmed by the Board's institution decision, the amended scheduling order, and multiple law-firm/publication trackers. I am reporting the one proceeding as real and the ODP field as unreliable for this patent. Do not confuse PGR2025-00033 with the unrelated IPR2025-00033 (that one is on patent 11,744,802, "Not Instituted - Procedural") — the numbering coincides by coincidence.

---

PGR2025-00033 — Merck Sharp & Dohme LLC v. Halozyme, Inc. (U.S. Patent No. 12,049,652)

• Type: Post-Grant Review (PGR) — full § 282-type grounds, including § 112, are available.
• Filed: 2025-03-07
• Status: "Trial Instituted" (Unified Patents entry: "Pending - Instituted"). Plain English: trial is live; the one-year clock runs, decision pending.
• Judge panel: Per the amended scheduling order (Paper 36, 2025-11-20), the trial panel is Susan L. C. Mitchell (lead), Cynthia M. Hardman, and Michael A. Valek, Administrative Patent Judges. A third-party docket additionally lists Jeffrey N. Fredman on the case; I could not independently confirm whether Fredman sat on the institution panel or is a docket artifact, so treat that name as unverified.
  • Amended scheduling order: https://www.docketalarm.com/cases/PTAB/PGR2025-00039/Merck_Sharp_%26_Dohme_LLC_v._Halozyme_Inc/11-20-2025-Board/Order__Other-36-AMENDED_SCHEDULING_ORDER/
• Petition grounds: Challenge to claims 1–40 (all claims) on three grounds:
  1. § 112(a) written description — claims 1–40. The claims require one replacement at position 320 but permit up to 41 additional substitutions, allegedly capturing between 10⁵⁹ and 10¹¹² distinct polypeptides; the specification discloses only singly-modified mutants plus a prophetic "make-and-test" research plan.
  2. § 112(a) enablement — claims 1–40. Undue (indeed allegedly impossible) experimentation to identify which multiply-modified PH20 polypeptides retain activity.
  3. § 103 obviousness — claims 1–2 and 5–40 over Halozyme's own U.S. Patent No. 7,767,429 in view of Chao (2007, Biochemistry), targeting the single-substitution D320K / D320S mutants.
  • PGR eligibility was itself contested: Merck argued the '652 patent is PGR-eligible because at least one claim lacks § 112(a) support in the pre-March-16-2013 '731 application (the Inguran / Collegium line). Halozyme's contrary arguments on this point did not carry the day at institution.
• Institution decision: Partially instituted — 2025-10-01. The Board instituted on Grounds 1 and 2 (§ 112(a) written description and enablement) but denied institution on Ground 3, finding that Petitioner "had not shown it was more likely than not to establish that the patents are obvious in light of the prior art." The Board also found the evidence supported a broad construction of "modified PH20 polypeptide." Patent Owner had argued discretionary denial (Fintiv); the Board instituted notwithstanding. Note that six sibling Halozyme PGRs had their institution decisions referred to panels by the USPTO Director in August–September 2025 after the Director declined discretionary denial — a signal the Office was not going to let Fintiv bottle up this family.
  • Institution decision (PDF): https://biologicshq.com/wp-content/uploads/2025/10/PTAB-PGR2025-00033-Institution-Decision.pdf
  • Summary of the three-institution batch: https://biologicshq.com/ptab-grants-institution-of-three-pgrs-related-to-mercks-keytruda-qlex/
• Final Written Decision: Not issued as of 2026-09-26. The statutory deadline under 35 U.S.C. § 326(a)(11) is one year from institution, i.e., 2026-10-01, extendable by up to six months for good cause (to as late as ~2027-04-01). Oral hearing was reset to 2026-07-23, 1:30 PM ET at both parties' request. Do not assume any claim-level FWD outcome for the '652 patent — I found no reported FWD, and the claim-level invalidations reported in the press to date are for the sibling patents, not this one.
• Settlement / termination: None. No adverse judgment, no termination, no settlement reported.
• Appeal: None identified. There is no FWD to appeal yet. I found no Federal Circuit appeal involving PGR2025-00033 as of 2026-09-26; a CAFC appeal can only follow the FWD (or an adverse judgment).
• Key pre-institution claim attrition: Halozyme statutorily disclaimed claims 3–5 and 30–40 of the '652 patent before institution, leaving claims 1–2 and 6–29 (26 claims) in effect for the trial. Statutory disclaimer means those 14 claims are treated as though they never existed — they cannot be asserted against anyone, by anyone, ever.
• Parallel, same-parties litigation: Halozyme, Inc. v. Merck Sharp & Dohme Corp., Case No. 2:25-cv-03179 (D.N.J.), filed 2025-04-24, in which Halozyme asserts the '652 patent (among ~15 family patents) against subcutaneous pembrolizumab (Keytruda Qlex™ / berahyaluronidase alfa). The Board identified the D.N.J. case as a related matter and noted the FWD would issue years before a likely trial date.
  • https://portal.unifiedpatents.com/litigation/New%20Jersey%20District%20Court/case/2%3A25-cv-03179
• Defensive value: Halozyme already surrendered claims 3–5 and 30–40 rather than defend them — any demand letter citing those claims is bluffing. The live claims (1–2, 6–29) are under an instituted § 112(a) attack on a written-description/enablement theory that has already succeeded three times in this family, so the realistic defensive posture is strong and improving, but not yet final — a defendant should be prepared to litigate the '652 claims in district court if the FWD does not land before trial.

---

Strategic summary

Claim-level posture on US 12,049,652. Forty claims issued. Claims 3–5 and 30–40 (14 claims) are CANCELED by statutory disclaimer — not merely at risk, but legally erased. Claims 1–2 and 6–29 (26 claims) are INSTITUTED and UNTESTED — challenged under § 112(a) written description and enablement, with the § 103 ground denied institution. No claims of the '652 patent have been sustained on the merits, and none have been invalidated on the merits by the Board. If you are drafting around this patent today, the only claims that matter are 1–2 and 6–29, and the only precedential-level guidance on their validity comes from other patents in the family.

Family-level precedents a defendant can ride. Three sibling Halozyme MDASE patents have already gone down on the same written-description/enablement theory, which makes the '652 proceeding a leading indicator rather than an outlier:

• US 11,952,600 (PGR2025-00003) — FWD 2026-05-12: claims 1–4 and 8–21 unpatentable, after Halozyme disclaimed claims 5–7. https://www.pearceip.law/2026/05/12/msd-invalidates-us-halozyme-patent-in-sc-pembrolizumab-dispute/
• US 12,152,262 (PGR2025-00006) — FWD invalidating the patent, reported May 2026.
• US 12,018,298 (PGR2025-00004) — FWD 2026-06-08: remaining claims unpatentable on written description and enablement. https://www.thebionews.net/news/articleView.html?idxno=25272
  These FWDs are persuasive (not binding) authority in the '652 trial, but they are the same panel, same specification, same theory.

Estoppel landscape. Because no FWD has issued, no estoppel has attached yet to anyone. When/if it does, § 325(e)(2) (the PGR analogue of § 315(e)(2)) bars Merck, its real parties-in-interest, and their privies — not the world — from asserting in a civil action or ITC proceeding any ground raised or that reasonably could have been raised in the PGR. Practical consequences for a different defendant being asserted today:

• You are not estopped by anything. The PGR was Merck's; a non-party retains the full ground set.
• New PGRs are foreclosed by timing, not estoppel. The '652 patent issued 2024-07-30, so the 9-month PGR window (§ 321(c)) expired on or about 2025-04-30. There is no fresh PGR path to raise § 112 in the PTAB.
• IPRs remain available but are § 102/§ 103-only — an IPR cannot carry a written-description or enablement attack. Note that Merck pivoted to exactly this: it filed three IPR petitions on 2026-03-23 (IPR2026-00312 on 10,865,400; IPR2026-00313 on 11,041,149; IPR2026-00314 on 11,066,656) on § 102/§ 103 art after exhausting PGRs — the same playbook is open to you for the '652 patent, subject to § 315(b)'s one-year bar from service of any complaint on you.
• Joinder to PGR2025-00033 is effectively time-barred. Under 37 C.F.R. § 42.122(b), a joinder request must be filed within one month of the institution date — i.e., by roughly 2025-11-01. That window has closed; do not plan on § 325(c) joinder unless the Board grants relief.
• Merck's denied obviousness ground is generally not an estoppel trigger (Shaw Indus. Grp., Inc. v. Automated Creel Sys., Inc., 817 F.3d 1293 (Fed. Cir. 2016)) — subject to the Board's evolving application of that rule — so that '429 + Chao combination stays available to Merck in D.N.J. regardless of how the FWD comes out.

Pattern signals. This is not a defensive-aggregator case — the petitioner is Merck Sharp & Dohme LLC, a large operating company defending its own commercial product (Keytruda Qlex™ with Alteogen's ALT-B4 enzyme), litigating in parallel in D.N.J. (2:25-cv-03179) and in Europe (Munich PI against MSD in December 2025; invalidity actions in Germany, the UK, the Netherlands, and France). Merck filed 14 PGRs across the family between November 2024 and November 2025 (13 instituted, the 14th denied 2026-05-15) plus the three March 2026 IPRs — a coordinated, repeat-player campaign, not a one-shot challenge. Halozyme, for its part, is aggressively defending and enforcing: it fought discretionary denial on Fintiv grounds, and it has been narrowing by disclaimer rather than by amendment across the family (claims 3–5, 16, 31–35 of the '520 patent; claims 3–5 and 30–40 here; claims 5–6 of the '185 patent; claims 5–7 of the '600 patent). There is also parallel UK Patents Court activity on a family member (referred to in the judgment as "EP347"), which I mention as context only — I have not verified its correspondence to the '652 patent. Note: I found no Federal Circuit appeal by Halozyme in this family as of 2026-09-26; the appeal window on the three sibling FWDs is live, and I would not be surprised to see one.

---

Recommended next steps

If you are a defendant facing an assertion of US 12,049,652:

1. Immediately strip the demand letter. If the assertion cites claims 3–5 or 30–40, respond that those claims were statutorily disclaimed by Halozyme before the 2025-10-01 institution decision and are treated as never having existed. They cannot support infringement or damages. Demand the specific claim numbers being asserted, in writing.
2. Calendar the FWD. The statutory FWD deadline is 2026-10-01 (extendable to ~2027-04-01). Watch for the decision; if the Board follows its three sibling FWDs, all remaining claims (1–2, 6–29) are invalidated as unpatentable, which functionally ends the patent. Monitor via PTAB E2E: https://ptacts.uspto.gov/ptacts/ and the Unified Patents docket entry: https://portal.unifiedpatents.com/ptab/caselist?judges.panel=Michael+A.+Valek
3. Build the § 112(a) record now. Because PGR is time-barred for you, your written-description/enablement attack must be made in district court or the ITC. Import the three sibling FWDs (PGR2025-00003, -00004, -00006) as persuasive authority, and engage an expert on the 10⁵⁹–10¹¹² genus point and the singles-only working examples.
4. Preserve a § 102/§ 103 escape hatch. If you intend to use IPR, file before your § 315(b) one-year bar from service of the complaint runs, and understand that you cannot raise § 112 there. A follow-on IPR cannot cure a missed § 112 argument.
5. Quantify the residual value. With claims 3–5 and 30–40 gone and claims 1–2 and 6–29 facing an instituted § 112(a) trial, any settlement/license calculus should value the patent at a substantial discount to face value — and should include a termination-on-FWD clause if the FWD invalidates the remaining claims.

If the FWD has issued by the time you read this and the remaining claims were held unpatentable: quote the disposition and claim list verbatim from the decision and link to the PTAB E2E copy — that is your complete defense. If the FWD sustained claims 1–2 or 6–29, the defensive picture changes materially and you should re-run the obviousness analysis that the Board declined to institute, since the Board's denial of Ground 3 (2025-10-01) is a favorable-on-its-face datapoint for Halozyme on § 103. I am not able to tell you which of those two outcomes occurred as of today, and I will not guess.

Confidence and limits: high confidence on the proceeding number, parties, patent, filing date (2025-03-07), institution date (2025-10-01), the disclaimed claim groups, the three grounds, and the partial institution. Medium confidence on the exact composition of the trial panel (Mitchell/Hardman/Valek confirmed by the Board's own order; Fredman unverified). No FWD located for PGR2025-00033 as of 2026-09-26 — if you need that answer definitively, pull the docket from PTAB E2E or call the Board; do not rely on this summary for a claim-level holding that may have issued in the last few days.

I'll start by pulling the authoritative record and then cross-checking the assignment/litigation trail with live sources.

Let me try to pin down the specific assignment reel/frame records and any correspondent of record.

Inventors

Notes on pattern:

• No depart-and-sell pattern. The opposite is true: the same four inventors are named on the originally-filed 2012 application (US 13/694,731 → US 9,447,401, issued 2016-09-20) and on this 2022 continuation (US 18/064,886). All of them are the inventors of record on the assignment instruments recorded in December 2022 — i.e., they were still executing papers for the assignee roughly a decade after the priority filing. That is a hallmark of an internally-retained corporate portfolio, not a fire-sale.
• Shepard's out-of-state (Oregon) residence address is the only mildly unusual item; it is consistent with a founder-inventor who has relocated, and does not by itself indicate anything about ownership.
• Inventor cities/addresses in the file were taken from the Application Data Sheet for a closely related continuation (US 18/068,327) rather than the 2012 parent; I could not independently verify the 2012-era addresses.

Original assignee

Halozyme, Inc. (originally recorded as Halozyme Therapeutics, Inc.; the two names both appear in the Google Patents current-assignee field and the December 2022 recordings consolidate into Halozyme, Inc.). San Diego, California.

• Line of business: Clinical-stage-turned-commercial biopharmaceutical company. It is an operating company, publicly traded (NASDAQ: HALO), founded 1998 specifically to develop a recombinant human hyaluronidase. Its business is the ENHANZE® drug-delivery platform (recombinant human PH20, rHuPH20) plus a specialty injectables / auto-injector business acquired with Antares Pharma in Q2 2022, which brought in XYOSTED® and EpiPen®-type auto-injector revenue and the Tlando/NeutraGard products.
• Product embodying the claims: Yes. Halozyme commercializes HYLENEX® (recombinant human hyaluronidase), a PH20 product, and licenses rHuPH20/ENHANZE into co-formulated products from Roche, Takeda, J&J (DARZALEX FASPRO®, DARZALEX® SC), BMS, argenx, ViiV and others. The patents in this family (the "MDASE™ portfolio") are described by Halozyme as covering modified human hyaluronidases. Halozyme's stated patent exclusivity for this portfolio is through 2032 ex-US / 2034 in the U.S.
• Current status: Operating, profitable, no bankruptcy. 2024 revenue of ~$1.015 billion reported; ~500 employees.

Assignment timeline

Important limitation on sourcing: I retrieved the assignment events from the Google Patents legal-events record for US 12,049,652, which mirrors the USPTO record. I was not able to retrieve the USPTO Assignment Center reel/frame numbers, the recorded conveyance execution dates, or the correspondent/attorney of record for these entries within this session. I am therefore not reporting reel/frame identifiers, because inventing them would be worse than leaving them blank. Verify at the Assignment Center (search by patent number 12049652) before relying on any of the below for a brief.

Recorded assignment events (event/recording dates per the Google Patents legal-events feed):

• 2022-12-16 (execution date not retrieved) — Reel/Frame not retrieved

  • Conveyance: Assignment (inventor-to-company)
  • Assignor: CONNOR, ROBERT JAMES; WEI, GE; ZHAO, Qiping
  • Assignee: HALOZYME THERAPEUTICS, INC.
  • Correspondent: not retrieved. If a single correspondent recurs on all three of the December 2022 recordings, that would simply be Halozyme's outside patent counsel handling a routine batch recordal; flag it only if the same name also appears as correspondent for an unrelated NPE-family chain.
  • Context: Routine inventor assignment of rights to the research employer.

• 2022-12-16 — Reel/Frame not retrieved

  • Conveyance: Assignment (intra-group)
  • Assignor: HALOZYME THERAPEUTICS, INC.
  • Assignee: HALOZYME, INC.
  • Correspondent: not retrieved.
  • Context: Internal corporate consolidation / entity re-alignment, not a third-party transfer.

• 2022-12-16 — Reel/Frame not retrieved

  • Conveyance: Assignment (inventor-to-company)
  • Assignor: SHEPARD, H. MICHAEL
  • Assignee: HALOZYME, INC.
  • Correspondent: not retrieved.
  • Context: Founder-inventor assignment recorded directly to the ultimate parent entity.

What is not in the chain: there is no assignment to an "IP Holdings," "Licensing," "Ventures," or "Patents" entity; no security agreement; no merger conveyance; no transfer to any third party. The chain terminates at the same operating company that filed the application in 2012.

Supporting litigation/administrative events (for context, not assignment records):

• 2025-03-07 — Merck Sharp & Dohme LLC files PGR2025-00033 against US 12,049,652; instituted 2025-10-01 (Unified Patents lists the patent owner as Halozyme Inc, entity type "Operating Company").
• 2025-04-24 — Halozyme, Inc. v. Merck Sharp & Dohme Corp., No. 2:25-cv-03179-ES (D.N.J.), asserting 15 Halozyme patents including the '652 patent over SC KEYTRUDA® / KEYTRUDA QLEX™.
• Halozyme's U.S. counsel of record is Walsh Pizzi O'Reilly Falanga LLP (local) with pro hac vice counsel; Merck's PGR counsel is Sidley Austin LLP (Jeffrey P. Kushan). These are litigation counsel, which is a different field from the assignment "correspondent of record."

Timeline diagram

timeline
    title Ownership of US 12049652
    2011 : Priority applications filed
    2012 : Parent application filed by Halozyme
    2016 : US 9447401 issued in family
    2022 : Continuation filed by Halozyme
         : Inventors assign rights to Halozyme
         : Rights consolidated into Halozyme Inc
    2024 : US 12049652 issued
    2025 : Halozyme sues Merck in New Jersey
         : Merck PGR instituted at PTAB

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. Every recorded assignee on 2022-12-16 is a Halozyme operating entity (Halozyme Therapeutics, Inc., then Halozyme, Inc.). There is no suffix entity (IP/Holdings/Ventures/Licensing), no registered-agent address, and the assignee ships products (HYLENEX, ENHANZE-licensed products, auto-injectors).

2. Known asserter in the chain — Not present. Neither Halozyme, Inc. nor Halozyme Therapeutics, Inc. appears on any of the referenced NPE lists (Acacia, Marathon, IV, IPNav, Wi-LAN, Conversant/Mosaid, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, etc.). Unified Patents classifies Halozyme's entity type as "Large/Operating Company — Medical."

3. Repeat correspondent across the chain — Unclear / not assessable. This is the one signal I cannot close out: the correspondent (attorney/firm) of record on the three December 2022 recordings was not retrievable in this session. Given that all three recordings are same-day batch recordals by the same applicant, a recurring correspondent is expected and is not probative of NPE activity on its own. Do not score this signal either way without pulling the actual Assignment Center abstracts.

4. Cascading transfers — Not present. There are three records, all on a single day (2022-12-16) and all terminating at the same final assignee. There is no chained series of unrelated LLCs over months, and no common-principal shell pattern.

5. Pre-litigation transfer — Not present. The last recorded assignment (2022-12-16) precedes the first suit naming this patent (2025-04-24, 2:25-cv-03179) by roughly 28 months, far outside the 6-month window that would suggest chain-arranging to set venue or standing. The patent itself only issued 2024-07-30, so it could not have been asserted earlier.

6. Bankruptcy fire-sale — Not present. No Chapter 7/11, no §363 sale, no receiver. Halozyme reported ~$1.015B revenue in 2024 and remains an active NASDAQ issuer; it has bought assets (Antares Pharma, 2022) rather than being liquidated.

7. Privateering — Not present. The patent stayed with the practicing/liquidating operating company that developed the technology; it was not handed to a third-party assertion vehicle to sue Halozyme's competitors. Halozyme is the named plaintiff in its own name.

8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. Note the inverse situation: Unified Patents and RPX-adjacent data sources are tracking Halozyme as the operating-company patent owner being challenged by Merck, which is the mirror image of the defensive-aggregator signal.

Verdict

Operating-company assertion.

Justification: the complete recorded chain is three same-day (2022-12-16) instruments moving rights from inventors Ge Wei, Qiping Zhao and Robert James Connor to Halozyme Therapeutics, Inc., from Halozyme Therapeutics, Inc. to Halozyme, Inc., and from founder-inventor H. Michael Shepard directly to Halozyme, Inc. — with no third-party, no shell suffix, no security interest, and no pre-litigation transfer (the last recordal predates the 2025-04-24 D.N.J. complaint by ~28 months, and the patent did not issue until 2024-07-30). The current assignee is a NASDAQ-listed, revenue-generating San Diego biopharmaceutical company that ships HYLENEX and licenses ENHANZE-incorporated products, and it is asserting this patent in its own name against a competitor (Merck) in Halozyme, Inc. v. Merck Sharp & Dohme Corp., No. 2:25-cv-03179-ES (D.N.J.) — the definition of operating-company assertion rather than NPE activity.

Caveat on confidence: my reel/frame and correspondent data are incomplete, so signals #3 (repeat correspondent) is unresolved. That gap does not change the verdict — the assignee identity, product sales, and the ~28-month gap to litigation are dispositive — but the reel/frame citations should be filled in before any filing.

Verify at:

• USPTO Assignment Center — https://assignmentcenter.uspto.gov/ (search by patent number 12049652)
• Legacy Assignment Search — https://assignment.uspto.gov/patent/index.html
• Patent text and legal events — https://patents.google.com/patent/US12049652/en
• PTAB proceeding — https://portal.unifiedpatents.com/ptab/case/PGR2025-00033 (or https://ipverse.greyb.com/ptab-web/cases/case-details/PGR2025-00033)
• District court docket — https://dockets.justia.com/docket/new-jersey/njdce/2:2025cv03179/[566903](/patent/566903)

I'll research the cited prior art for US 12,049,652 using USPTO and patent-citation sources.

Let me look specifically for the front-page "References Cited" / patent citations list of US 12,049,652 and its family.

Let me pull the front-page references and citation lists from the patent-family sources.

Prior Art Analysis — U.S. Patent No. 12,049,652 B2

0. Verification limitation (read first)

I must be explicit about a boundary on this task: I was not able to retrieve the authoritative front-page "(56) References Cited" list for 12,049,652 itself. The Google Patents page you supplied as the authoritative text contains the description and claims but the fetched text does not include the front-page bibliographic citation block, and my searches returned family-level IDS/file-history reference lists (reproduced inside PTAB exhibits) rather than the '652 patent's own PatentCenter "References Cited" table.

Consequently, I can substantiate the substance and identity of the family's cited art, but I cannot certify that each item below appears verbatim on the '652 front page, nor can I give you the examiner's designated "Desig. ID" codes for the '652 patent specifically. The definitive list must be pulled from USPTO PatentCenter for application 18/064,886 (or the '652 grant PDF, page 1, column 1) at https://patents.google.com/patent/[US12049652B2](/patent/US12049652B2). I flag every item below by how I found it so you can spot-check rather than rely on my synthesis.

---

1. Patent identity (confirmed as found)

Sources: https://portal.unifiedpatents.com/ptab/case/PGR2025-00033 ; https://portal.unifiedpatents.com/patents/patent/US-[11952600](/patent/11952600)-B2 (family/Art-Unit data).

Analytical consequence of the family structure. The '652 patent is a continuation sitting atop a very long chain in the same specification family — 13/694,731 → US 9,447,401 → 15/226,489 → US 10,865,400 → 16/824,572 → US 11,041,149 → 17/327,568 / 17/327,586 → US 11,066,656, US 11,952,600, US 12,018,298, US 12,037,618, US 12,049,652, etc. (family chain recited in the specification text; expirations listed at https://www.drugpatentwatch.com/p/biologics/patent/index.php?query=[12018298](/patent/12018298)). In such continuations the examiner's cited art is ordinarily the family's accumulated IDS art, which is exactly the art reproduced in the PTAB exhibits I retrieved. That is the basis for the tables below.

---

2. Threshold point that governs the entire §102 question

Two facts sharply limit any §102 anticipation theory against the '652 patent:

1. The patent was allowed over this art. The very references below were before the examiner during prosecution of this family, and the claims issued. Anticipation requires a single reference disclosing every limitation — including, for the variant genus claims, the specific residue replacement at the position identified in the claim and the recited identity threshold. None of the listed references is a single-reference anticipation of the variant claims.

2. Institution of PGR2025-00033 did not rest on §102 art. Per the earlier litigation section (which I am instructed to treat as authoritative) and consistent with the PTAB docket, the Board instituted on §112 written description and enablement and found the petitioner's §103 obviousness theory was not shown to be more likely than not. That is an independent, current-record datapoint that the cited art does not even clear the §103 bar — a fortiori, no reference in it was shown to anticipate under §102. Sources: https://biologicshq.com/ptab-grants-institution-of-three-pgrs-related-to-mercks-keytruda-qlex/ ; https://portal.unifiedpatents.com/ptab/case/PGR2025-00033.

So the honest answer to "which claims does each reference potentially anticipate?" is: on the located record, none of these references was shown to anticipate any claim, and the most that can responsibly be said is which claim categories each reference is topically capable of reaching — and, for a handful of them, which claim category the petitioner/patent owner actually litigated.

---

3. Patent references (as located in the family IDS/file-history record)

Provenance note: these were located in IDS/"Other Documents" listings and family cite lists reproduced in PTAB petition exhibits (e.g., https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557959](/patent/1557959)/download-documents?artifactId=kSLrZJw8wL8uaGyx8gVDbYwbTCiWsrkDrfS5OtOS6QW9qnjuf0_cSFI ; https://www.docketalarm.com/cases/PTAB/PGR2025-00009/Merck_Sharp_%26_Dohme_LLC/docs/12-27-2024-Petitioner/Exhibit-1002-2-File_History_of_US_Patent_No_12,123,035_Part_2.pdf) and in the JP/EP family citation tables (https://patents.google.com/patent/JP2021507676A/en). They are predominantly applicant-submitted IDS art unless stated otherwise. Dates are given as found; where a snippet did not carry a date I say so.

3A. U.S. patent documents — the core PH20/hyaluronidase art

3B. Non-patent literature (the genuinely §102-relevant items)

---

4. Which reference maps to which claim category — restored to the patent's own vocabulary

The specification text you supplied enumerates the claim categories that issue in this family ("provided are … modified PH20 polypeptides"; "nucleic acid molecules … vectors … cells"; "methods of producing"; "compositions/pharmaceutical compositions"; "insulin formulations"; "method[s] for identifying or selecting a modified hyaluronan-degrading enzyme"; and treatment uses). Mapping the references to those categories:

• Modified-PH20 polypeptide claims → only Arming 1997, US 10,328,130 / WO 2009/128917 / EP 2,662,090, and WO 2015/003167 are even topically capable. None discloses the specific residue replacement + identity-threshold combination recited; none was shown to anticipate.
• Unmodified/soluble-PH20, nucleic acid, vector, host-cell claims → US 5,721,348, US 5,958,750, US 7,767,429, US 2006/0104968 are the closest, but these features are the antecedent disclosure, and the '652 claims in this family are directed to variants, so they do not read on them.
• Assay/selection-method claims → Bailey & Levine 1993, Bonner & Cantey 1966 (Clin. Chim. Acta 13:746–752) as methodology.
• Insulin/preservative/co-formulation claims → nothing in the located list is directed to hyaluronidase + insulin with a phenolic preservative; the closest art surfaced in a different family's EPO search (EP 3,785,701 file) was US 2018250397 A1, WO 20111012837 A2, Shpilberg & Jackisch (Br. J. Cancer 109:1556–1561, 2013), and Bittner (2018) — but those were cited against a different application, not against the '652 patent, and I flag them as not confirmed as '652 citations.

---

5. The one §102 theory that is actually live against the '652 patent — the family's own disclosures

This deserves separate treatment because it is the only anticipation-style attack that maps cleanly onto the located record:

The petitioner in the parallel PtAB proceedings argued that the family may not be entitled to the 2011-12-30 / 2012-11-01 priority of the '731 application (this is expressly discussed for US 12,110,520 in the institution decision at https://biologicshq.com/wp-content/uploads/2025/10/PTAB-PGR2025-00017-Institution-Decision.pdf: "the '520 patent might not receive the benefit of priority to the earlier filed applications … and … is eligible for post-grant review"). If that reasoning were applied to the '652 patent, then:

• WO 2013/102144 A2 (the family's own PCT, published 2013-07-11) and US 9,447,401 B1 (granted 2016-09-20) and US 10,865,400 B1 (granted 2020-12-15) would all be §102(a)(1) art (patented/published more than one year before the 2022-12-12 filing), disclosing the same modified PH20 variants, the same tables of residue replacements, and the same preservative-stability data as the '652 patent.
• Because §102(b)(2)(C)'s common-ownership exception does not save §102(a)(1) art, family members whose issuance/publication predates the critical date by more than a year would be anticipatory on the '652 patent's own specification.
• I did not locate any institution or FWD in PGR2025-00033 adopting this reasoning for the '652 patent. Treat it as an argument in the record, not an adjudicated outcome.

This is, in my assessment, the highest-value §102 line against the '652 patent — and it comes from the patent's own family, not from any third-party reference.

---

6. Discrepancies and caveats you should carry forward

1. Contradiction with the earlier section (flagged): the earlier litigation section states PGR2025-00033 was instituted on §112 grounds with the §103 theory not shown. My searches are consistent with that (instituted 2025-10-01, status "Pending – Instituted"), but I could not independently retrieve the '652 institution decision text. If the institution decision in fact also instituted on obviousness, the "art cannot even reach §103" conclusion in §2 above would need revision. Verify against the PTAB docket.
2. The earlier section referenced a May 12, 2026 FWD holding US 11,952,600 unpatentable. I did not confirm this in the present search round. It concerns a different patent and should not be attributed to the '652 patent.
3. I could not verify the '652 patent's actual claim language. This materially limits any "which claim is anticipated" statement. The claim-numbering facts I can carry (claims 3–5 and 30–40 statutorily disclaimed before institution) come from the earlier section, and I did not independently confirm them here.
4. Publication dates marked "as found" come from citation-annex tables that occasionally contain typographical errors (e.g., the family annex listing "WO 2013-102144 A2 … 04/07/2013" versus the more usual 2013-07-11 publication). Confirm dates against the primary documents before relying on them for a critical-date calculation.
5. No third-party §102 reference was located that anticipates the '652 claims. The most on-point third-party reference in the cited corpus is Arming et al. 1997, and it is a mutagenesis/characterization paper, not a disclosure of the claimed residue-specific variants.

---

7. What I recommend to close the gap

1. Pull the (56) References Cited block for US 12,049,652 from USPTO PatentCenter (application 18/064,886) or the grant PDF page 1 — that is the only way to be certain which references I listed above are actually on the '652 front page versus inherited family IDS art.
2. Pull the PGR2025-00033 institution decision (due on the docket; a FWD would be due ~2026-10-01) to confirm whether §102 or §103 grounds survived institution.
3. Pull the file history of 18/064,886 to see which references the examiner actually applied in any §102/§103 rejection, and how they were overcome — that is the operative "prior art" answer rather than the raw IDS list.

I'll research the prior art references associated with US 12,049,652, including the references cited during prosecution and rehearsed in the PTAB record.

Let me drill into the specific obviousness theory and prior art applied to the '652 patent's position-320 limitation.

Obviousness Analysis — U.S. Patent No. 12,049,652 B2

("PH20 polypeptide variants, formulations and uses thereof") under 35 U.S.C. § 103

Prepared 2026-04-26. Builds on the previously generated Patent Summary and Litigation Summary; those sections are treated as authoritative and are not repeated. Where this analysis refines or contradicts them, it is flagged.

---

0. Scope note — what "the Prior Art section of this page" actually contains

An important methodological caveat first. The authoritative full text of US12049652B2 supplied to me did not include a Google Patents "Prior Art," "Citations," or "Similar Documents" block. The page text I was given begins at the abstract/summary and runs through the definitions section; it contains no face-page reference list and no "Cited By" table. I therefore reconstructed the prior-art universe the '652 patent actually faces from three verifiable sources, and I say which is which throughout:

1. Admissions inside the '652 specification itself (these are prior art by the patentee's own citation — see the Definitions block supplied: Arming et al. (1997), ten Have et al. (1998), Chowpongpang et al. (2004), Primakoff et al. (1988), Tung et al. (1997), Bordier et al. (1981), Laurent et al. (1992), U.S. Pat. No. 7,767,429 and U.S. Pub. No. 2010/0143457, and — importantly — U.S. Provisional Appl. No. 61/520,962 and U.S. application Ser. Nos. 13/507,263 and 13/507,262).
2. The PTAB record in PGR2025-00033, which identifies the two references Merck actually asserted (EX1005 = U.S. 7,767,429; EX1006 = Chao (2007)) and the references the Examiner previously considered (Stern, Zhang, Arming, the '429 patent, Lin, Morrison).
3. Sibling PGRs (PGR2025-00003, ‑00017, ‑00024, ‑00039, ‑00042, ‑00050, ‑00053) in which Merck ran the same obviousness theory against the same claim architecture.

Date caveat (carried forward and still live): the searches returned materials dated through September 2026, i.e. after the April 26, 2026 date in my instructions. I use them only where they corroborate the April-2026 status, and I flag anything later-breaking.

---

1. The claim that governs the §103 analysis

Claim 1 (quoted by the Board from Ex. 1001 at 309:18–43) requires:

Three consequences drive everything below:

• The minimal infringing/claimed species is a single-substitution variant. Per Dr. Hecht's declaration in PGR2025-00033 (heading reproduced in the docket): "Claims 1‑6, and 8‑40 Encompass a Single-Replacement PH20₁₋₄₄₇ Mutant Where D at Position 320 is Changed to K … While Claim 7 Encompasses a Single-Replacement PH20₁₋₄₄₇ Mutant Where D at Position 320 is Changed to S." This cross-confirms the Patent Summary's statement that claim 7 is limited to D320S and that the other surviving claims encompass D320K. I found no contradiction between sections on this point.
• No activity, no stability, no formulation limitation is recited in claim 1. The Board recorded the patent owner's own position that "claims 2 and 6‑29 do not require any hyaluronidase activity." That removes one common non-obviousness hook (unexpected property), but it also weakens the patent owner's motivation story: a claim that requires only "a pH20 with K at 320" cannot rely on a teaching that a specific substitution produced an improved property.
• The claim is a genus, not a species. Because additional modifications are optional, claims 1–2 also read on D320K flanked by any known prior-art PH20 mutations — e.g., the glycosylation mutants (N200A, N333A, N358A) and catalytic mutants that the patentee expressly carved out of the claim (see §3.8).

---

2. Effective filing date — does it change the analysis?

The '652 patent claims priority to provisional 61/631,313 (2011‑12‑30) and 61/796,208 (2012‑11‑01) through Ser. No. 13/694,731 (2012‑12‑28). The Board preliminarily found the claims may not be entitled to that priority, which is precisely what made the patent PGR-eligible (effective filing date no earlier than the '652's own 2022‑12‑22/12‑12 filing).

This does not rescue the claims from §103. Both asserted references predate even the earliest claimed date:

• U.S. 7,767,429 issued 3 August 2010 — more than 16 months before 30 December 2011.
• Chao (2007), Biochemistry 46:6911–6920 — four years before the earliest priority date.

So even on the patentee's most favourable priority theory, both are prior art under §102(a)(1) (patents/printed publications publicly available before the effective filing date). This is worth stressing because the '429 patent is Halozyme's own patent: common ownership defeats only §102(a)(2)/§102(b)(2)(C) art, not §102(a)(1) art. Because the '429 patent issued in 2010, its printed disclosure is §102(a)(1) prior art against the '652 claims notwithstanding common ownership. (Under pre-AIA §103(c), the same logic applies to §102(e) art of common ownership, but again not to a pre-critical-date printed publication.)

The priority question therefore matters for the expiry (2032‑12‑28 per Google) and for PGR eligibility, but it does not remove either reference from the §103 combination.

---

3. The prior-art references and what each supplies

---

4. Combination A (primary): '429 + Chao (+ routine multiple-sequence alignment) — the theory Merck ran

This mirrors Ground 3 of PGR2025‑00033 exactly, so it is the theory with a real evidentiary record.

Element-by-element:

Motivation to combine (grounded in the record):

• Same field, same problem, same molecule: both references concern engineering human hyaluronidases for the same drug-delivery use ('429's sHASEGP/Enhanze platform; Chao's five-human-hyaluronidase family including PH20).
• The '429 patent is the patentee's own foundational disclosure and invites the modification: it states non-essential-region substitutions do not substantially alter activity and that further substitutions "can be determined empirically."
• The '429 patent is eight years older than the 2011 knowledge baseline and contains no PH20 structure (Park admitted "the structure of human PH20 was not solved [in December 2011]"). A POSA implementing its suggestion would therefore look to the most recent structural work on human hyaluronidases — Chao — which is exactly the KSR "improvement of one device suggests improvement of similar devices in the same way" rationale.
• The technique itself (site-directed mutagenesis + CHO expression + HA turbidity assay) was routine; even the patent owner's declarant Dr. Triggs-Raine conceded "making the modified PH20 polypeptides would have required nothing more than routine molecular biology and protein biochemistry techniques."

Reasonable expectation of success: the '429 patent's express assurance that such substitutions "do not substantially alter biological activity," plus Chao's demonstration that residues varying across homologues are productively variable, plus a PH20 homology model placing D320 outside the catalytic cleft.

Practical significance: the Board instituted only on the §112 grounds and expressly found the §103 ground was not shown to be more likely than not. So this combination, as pleaded, failed at the threshold. Any future §103 attack must cure the defect the Board identified (see §11).

---

5. Combination B: '429 + Chao + the preservative-instability art (motivation boost)

Merck's formulation was criticised by the patent owner precisely because the '429 patent says non-essential substitutions preserve rather than improve activity — so why tinker with a working enzyme? The fix is to add the motivation that the '652 specification itself concedes existed:

• Halozyme's own 61/520,962 / 13/507,262 / 13/507,263, plus Maa & Hsu, Remmele, Fransson, Lam and Singh, establish a recognised, unsolved problem: phenolic preservatives (phenol, m-cresol) aggregate and inactivate proteins, and PH20 "retains only about 10% of its activity" after 4 h in 0.4% m-cresol. Multi-dose, preservative-containing protein formulations (the insulin/CSII-pump context the '652 itself describes) required a preservative-tolerant PH20.
• The patentee's own earlier work had already framed the solution as screening single-substitution libraries for "phenophilic" and "thermophilic" mutants (the '652's own Examples 3 and 5 assays, applied to its Table 9 variant set, and its disclosure that F204P was identified as an m-cresol-resistant hit).

Why this helps: it supplies the why modify that the Board found missing. Why it does not fully solve the problem: claim 1 does not require improved stability, and this combination still does not explain why position 320 rather than any of the ~300+ non-conserved positions. It strengthens motivation but leaves the "which position" gap that the Board found dispositive in the sibling cases.

---

6. Combination C: '429 + Arming + Zhang (± Chao) — essential-residue elimination

Motivation: Arming teaches that D111, E113 and E249 are essential catalytic residues; Zhang teaches that N350 in the C-terminal EGF-like domain is essential while N216 is not; the '429 patent teaches that the conserved cysteines forming disulfides (C25/C316, C189/C203, and the C341–C352, C346–C400, C402–C408, C423–C429 pairs, all recited in the '652 specification's own Definitions) and the conserved glycosylated asparagines are structurally load-bearing.

A POSA seeking a tolerated substitution would therefore avoid the catalytic cleft and the conserved Cys/Asn network and pick a solvent-facing, non-conserved position. Position 320 sits immediately C-terminal to the conserved C316 and in a variable segment — i.e., in the "safe" category that survives the elimination. Chao Fig. 3 and the MSA then confirm the region.

This combination has the advantage of being built almost entirely on art the Examiner already considered (Arming, Zhang, Stern and the '429 patent all appear on the face of the family's IDS), which avoids the §325(d) Advanced Bionics exposure that the patent owner pressed hard in the sibling PGRs. Chao was the one reference Merck relied on that had not been considered, which is why the patent owner argued §325(d) at step one.

---

7. Combination D: Lin + Morrison (+ Wells/Cunningham-style alanine scanning) — the Examiner's own rejection

This is not speculation; it is the rejection actually made during prosecution of the family:

The Examiner cited Lin as disclosing the amino acid sequences of human and monkey PH20 and Morrison as teaching replacement of individual wild-type amino acids with alanine throughout a protein to assess the function of the substituted residue, and reasoned it would have been obvious "to modify the PH20 polypeptide with an alanine anywhere along the protein sequence," and that "it is not inventive to modify a protein with alanine in a polypeptide polymer." The applicant overcame the rejection by amending claim 1 to eliminate alanine (A) as an option at the replacement position, and substituting the H/I/K/M/Q/R/S (or, in '652, H/K/R/S) set.

Two consequences:

1. The Office has already found position-agnostic single substitution of PH20 to be obvious. The '652 claims do not escape that finding by narrowing the identity of the residue; they merely exclude the mechanistically most neutral choice (Ala→Ala scanning).
2. Eliminating alanine while claiming H, K, R and S actually cuts against the patentee. On the patentee's own record, alanine (small, neutral, non-disruptive) was the "obvious" substitution; requiring only charged/polar residues means the claim covers the more disruptive substitutions — which makes it harder, not easier, to argue a teaching away.

---

8. Combination E: naturally occurring homologue + '429 (the strongest, if the sequences line up)

If any PH20 in the sequence listing that is ≥91% identical to SEQ ID NO:3 — e.g. chimpanzee (SEQ ID NO:10 / 869 / 870), gibbon (857), marmoset (859), orangutan (861), all of which are highly similar to human PH20 and are expressly named in the '652 specification — already carries K, R, S or H at the position aligning to 320, then the claimed genus reads on, or is obvious in view of, a naturally occurring protein, and the case is materially stronger. The rationale would be the classic "copy what the homologue does at that position" (phylogenetic/consensus substitution) rationale, and would not require any inference from Chao's Fig. 3 at all.

I could not verify this. I do not have the actual residue at the aligned position 320 for SEQ ID NOs 10/857/859/861, and I flag this expressly as an unverified lead rather than a finding. It is nonetheless the single most valuable thing to check in the file, because it converts a weak "why this position?" argument into a two-reference combination with a species-level expectation of success. (Note also that the '652 specification itself lists SEQ ID NOs 857, 859, 861 and 870 as unmodified templates, which shows the patentee contemplated these homologues as starting points.)

---

9. Claim-by-claim §103 mapping

Note the structural dependency: every dependent claim inherits claim 1's position-320 limitation, so the whole §103 case collapses or stands on claim 1's "why position 320" question. That is why the Board's refusal to institute on §103 is so consequential — and why the dependent claims (17–19, 20–22, 24–29) are, counter-intuitively, the easier ones to attack but are unreachable if claim 1 stands.

---

10. The motivation inquiry, stated precisely

The petition needed four links; I summarise where each is strong and where it fails:

1. Make a single substitution in PH20₁₋₄₄₇. Strong — the '429 patent expressly endorses it.
2. Use Chao for structural insight. Strong — same field, more recent, and the '429 patent itself used a hyaluronidase crystal structure (bee venom) for the same purpose.
3. Identify the region containing position 320 as tolerant. Contested. Depends on equating "non-conserved" with "non-essential." The Board in sibling cases held squarely that "tolerance is not a positive reason to make a substitution" (quoting Knauf Insulation, Inc. v. Rockwool Int'l A/S, 788 F. App'x 728 (Fed. Cir. 2019)).
4. Select K or S specifically. Weak-to-moderate — supplied only by alignment frequency and a homology model.

KSR rationales that support the combination: (i) same field / same problem / same molecule; (ii) "a finite number of identified, predictable solutions"; (iii) routine optimisation of a known, working protein; (iv) express teaching in the primary reference that the modification is tolerated.

---

11. Counterarguments — and why the Board did not institute on §103

I would be misleading you if I did not foreground this. The PTAB instituted PGR2025‑00033 on §112 only. On §103 it found the petitioner had not shown a reasonable likelihood. The reasons, drawn from the institution decision and the sibling decisions, are:

1. No reason to select the species from the genus. "It is not enough, even after KSR, … that a reference includes a broad generic disclosure and a common utility …; there must be some reason to select a species from the genus." Neither the '429 patent nor Chao names or discusses position 320 (parallel holdings for position 307 in the '590 patent and position 309 in the '618 patent).
2. Tolerance ≠ motivation. The Board has repeatedly rejected MSAs as a source of motivation, because alignment data shows what can be substituted, not why one would.
3. The '429 patent cuts the other way. It teaches that non-essential-region substitutions do not alter activity or function. The patent owner's argument — "Petitioner has not provided any reason why POSAs, who have in hand a working enzyme, would tinker with that enzyme with the sole purpose of maintaining the same activity profile" — was persuasive on the preliminary record.
4. Hindsight. Park conceded in deposition that he was "asked by counsel to report [his] conclusions with respect to position 320," and that he could not identify a reason to prefer 320 over "the hundreds, if not thousands, of other substitutions he predicts would have been tolerated." The decisions invoke In re Stepan Co., 868 F.3d 1342, 1346 n.2 (Fed. Cir. 2017); Takeda Chem. Indus. v. Alphapharm, 492 F.3d 1350, 1356 (Fed. Cir. 2007); Amerigen v. UCB, 913 F.3d 1076, 1089 (Fed. Cir. 2019); Belden v. Berk‑Tek, 805 F.3d 1064, 1073 (Fed. Cir. 2015); Kingston Tech. v. North Star (IPR2018‑01784); Intel v. VLSI (IPR2018‑01033).
5. Specific technical objections to the D→K change. The patent owner argued D320 lies within an α-helix, that replacing a negatively charged residue with a positively charged one is non-conservative, and that the ~88 "bounding essential residues" in Chao Fig. 3 make the "region" identification arbitrary.
6. Unpredictability evidence inside the patent. The '652's own tables report 6,753 single substitutions, of which 3,380 (~57%) were inactive and only ~10% increased activity. That is powerful evidence against a "reasonable expectation of success" for single-substitution engineering generally, and it is the patentee's own data.
7. The breadth problem. Petitioner's declarants put the claim-1 genus at 10⁵⁹–10¹¹² distinct polypeptides; the same analysis drove the instituted §112 grounds. Practically, however, that broad-genus evidence is unhelpful on §103: for obviousness the question is whether a species within the claim is obvious, and the answer turns back on the position-320 selection.

Practical implication for future challenges: because the §103 ground was raised in the petition and the Board will issue an FWD on the instituted grounds, there is at least a live question whether §315(e) estoppel will bar Merck (or a privy) from re-running the §103 ground later; and a new petitioner would face General Plastic / §325(d) hurdles. My assessment: a fresh §103 attack on the '652 claims would most plausibly succeed only if it (a) adds a reference that names or squarely identifies position 320 or the corresponding structurally defined region of the Hyal-EGF/catalytic-domain junction, or (b) shows a naturally occurring PH20 with K/R/S/H at the aligned position (Combination E), or (c) supplies a concrete reason to modify (Combination B: the preservative problem) that makes the "why tinker" objection untenable.

---

12. A parallel validity theory worth noting: obviousness-type double patenting

Not a §103 ground strictly, but a §103-type analysis: in the earlier Halozyme litigation over the '171 application, the court concluded all pending claims were unpatentable on obviousness-type double patenting over each of Halozyme's '429, '431 and '081 patents, holding that the rejected claims were "nearly identical to claims 9 and 10 of Halozyme's '429 patent." If the '652 claims require a position-320 substitution that the '429 claims do not recite, ODP is likely a weaker theory here than against the PEGPH20-era claims — but the family's history shows the Office and the courts have repeatedly treated Halozyme's PH20 claim sets as patentably indistinct from one another, and any terminal disclaimer filed against the '429 patent would confirm the parties' own view that the disclosures overlap.

---

13. Bottom line

• Strongest §103 combination: U.S. 7,767,429 + Chao (2007) for the core claims, but as pleaded it was insufficient — the PTAB found no reasonable likelihood on §103 and instituted only on written description and enablement. The fatal gap is not "could the POSA make D320K" (everyone agrees that was routine) but "why would the POSA have picked position 320 out of hundreds of variable positions."
• Most promising ways to close that gap: (i) add the preservative-instability motivation from Halozyme's own 61/520,962 / 13/507,262 / 13/507,263 and from Maa & Hsu, Remmele, Fransson, Lam and Singh; (ii) add Arming and Zhang to supply the essential-vs-permissive map on art already considered by the Examiner (avoiding §325(d) exposure); (iii) verify whether a naturally occurring PH20 in SEQ ID NOs 10/857/859/861 (and 869/870) has K, R, S or H at the position aligning to 320 — if so, the case changes materially.
• Easiest claims if claim 1 falls: 17–19 (glycosylation), 20–22/28 (PEG/signal sequence), 24 and 29 (compositions) — all squarely taught by the '429 patent and Halozyme's own Frost/'380-family disclosures. Claims 6–15 (preservative stability) are strong on motivation once Combination B is added. Claims 3–5 and 30–40 are statutorily disclaimed and moot.
• Confidence: I am high confidence on the claim-1 text, the asserted combination, the reference identities, and the Board's refusal to institute on §103. I am moderate on the full claim set beyond the claims Merck charted (claims 16, 23 and the precise subject matter of 6–15 are reconstructed from the family's parallel claim sets, not from the '652 claim text, which I could not retrieve). I am low confidence, and expressly flag as unverified, on the 320-position residues of the non-human PH20 sequence listing and on the exact Chao page citations beyond those quoted in the PGR record.

Primary sources relied on

• U.S. 12,049,652 B2, full text supplied (https://patents.google.com/patent/US12049652/en)
• PTAB institution decision, PGR2025‑00033 (Oct. 1, 2025): https://biologicshq.com/wp-content/uploads/2025/10/PTAB-PGR2025-00033-Institution-Decision.pdf
• Merck petition, PGR2025‑00033 (Ground 3 mapping; exhibits EX1005 '429, EX1006 Chao, EX1003 Hecht, EX1004 Park, EX2055 Triggs‑Raine): https://ptacts.uspto.gov/ptacts/public-informations/petitions/[1557626](/patent/1557626)/
• Ex Parte PTAB case record for PGR2025‑00033: https://ai-lab.exparte.com/case/ptab/PGR2025-00033/doc/1005
• Hecht declaration docket entry (claim/species mapping D320K vs D320S): https://www.docketalarm.com/cases/PTAB/PGR2025-00009/Merck_Sharp_%26_Dohme_LLC_v._Halozyme_Inc/04-14-2025-Patent_Owner/Exhibit-2036-...
• Sibling institution decisions (parallel holdings on positions 307/309): https://bannerwitcoff.com/wp-content/uploads/2025/11/PGR2025-00042.pdf ; https://biologicshq.com/wp-content/uploads/2025/10/PTAB-PGR2025-00024-Institution-Decision.pdf ; https://biologicshq.com/wp-content/uploads/2025/10/PTAB-PGR2025-00017-Institution-Decision.pdf
• Prosecution rejection over Lin + Morrison: https://ai-lab.exparte.com/documents/ptab/ptacts-uspto/[1557034](/patent/1557034)/.../PGR2025-00006_doc_15_exparte.pdf
• Venable/BiologicsHQ PGR institution summary: https://biologicshq.com/ptab-grants-institution-of-three-pgrs-related-to-mercks-keytruda-qlex/
• District-court decision on ODP over '429/'431/'081: https://valawyersweekly.com/files/2018/07/018-3-314.pdf