US 9447401

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

Let me dig deeper into the claim set and any 2026 litigation activity.

Let me pull the actual claim text and confirm the claim structure.

US 9,447,401 B2 — Analyst Summary

Source of record: I relied first on the full patent text you supplied (Google Patents, fetched 2026‑09‑24) and corroborated bibliographic/claim details against DrugPatentWatch and PTAB/Google Patents documents surfaced in search. Where I could not retrieve verbatim text, I flag it.

---

1. Bibliographic data

Abstract (verbatim): "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."

---

2. Plain-language overview

PH20 is a neutral‑active human hyaluronidase — an enzyme that degrades hyaluronan (hyaluronic acid, "HA"), a major component of the interstitial/connective‑tissue barrier. Degrading HA transiently increases tissue permeability, which is why hyaluronidases are used as "spreading" agents to help other drugs (e.g., antibodies, insulin) disperse and be absorbed — the basis of Halozyme's ENHANZE/rHuPH20 platform.

The patent discloses a systematic structure/function study of human PH20 in which roughly every residue of the soluble catalytic domain (the 447‑residue SEQ ID NO:3 backbone) was individually substituted with about 15 different amino acids (~6,000–7,000 variants), and each variant was assayed for hyaluronidase activity and for stability under stressing conditions (heat, agitation, low/no salt, and protein‑denaturing excipients such as phenolic preservatives). The specification thus provides a large map of which substitutions increase activity/stability and which abolish activity. The '401 claims a genus of such modified PH20 polypeptides that are stabilized against a denaturing condition — especially an antimicrobial preservative.

---

3. Claim overview

Note on completeness: I was able to retrieve the full text of dependent claims 5–22 (via DrugPatentWatch), but not verbatim text of independent claim 1 or of any independent method/composition claims beyond the dependent set. The description below is therefore grounded in the specification language and in the retrieved dependent claims; treat the exact wording of claim 1 as not independently verified.

Claim 1 (independent — modified PH20 polypeptide): In substance, a modified PH20 polypeptide containing at least one amino acid replacement (relative to a reference/unmodified PH20, with positions referenced to SEQ ID NO:3 by alignment) that confers increased stability under a protein‑denaturing condition — the condition being, e.g., elevated temperature, agitation, low/no salt, or a denaturing excipient such as a phenolic preservative. Increased stability is measured as greater retained hyaluronidase activity than the corresponding unmodified PH20 under the same conditions.

Representative dependent claims (retrieved, verbatim substance):

• 5 — ≥15% of hyaluronidase activity retained in the presence of the phenolic preservative(s) for at least 4 hours vs. absence of preservative, same conditions otherwise.
• 6–7 — Stable in an anti‑microbially effective amount of one or more phenolic preservatives; amount is about 0.05%–0.6% (w/v) total.
• 8–9 — Phenolic preservative is phenol, m‑cresol, benzyl alcohol, and/or a paraben; particularly m‑cresol, phenol, or m‑cresol + phenol.
• 10 — Adds at least one replacement at a specified large group of positions (58, 10, 12, 20, 22, 26, 34, 36, 46, 50, 52, 68, 70, 74, 82, 83, 84, 86, 97, 127, 131, 138, 142, 143, 144, 166, 169, 174, 193, 195, 196, 205, 206, 213, 219, 234, 237, 238, 240, 249, 261, 267, 277, 279, 291, 309, 310, 314, 315, 317, 318, 347, 367, 375, 376, 399, 401, 407, 416, 419, 421, 431, 433, 439, 440, 443, 445 — referenced to SEQ ID NO:3).
• 11 — Specific replacement residues at those positions (e.g., G at 10; K at 12; S at 20; T at 22; M at 26; W at 34; C/K/R/N/Y/P/H at 58; P at 68; V at 83; P at 204; …).
• 14 — A polypeptide with ≥98% identity to SEQ ID NO:49 and replacement with P at position 204.
• 15–18 — Purified/isolated; further modified by glycosylation, sialation, albumination, farnesylation, carboxylation, hydroxylation, phosphorylation; glycosylated with N‑acetylglucosamine at ≥3 Asn residues (particularly N200, N333, N358 of SEQ ID NO:3).
• 19–21 — Conjugates: polymer (dextran or PEG), or a multimerization domain, toxin, detectable label or drug.
• 22 — Pharmaceutical composition comprising the modified PH20 polypeptide of claim 1 (this appears as an independent composition claim).

Method claims: The specification contains substantial methods subject matter — notably a method for identifying/selecting a modified hyaluronan‑degrading enzyme that is stable under a denaturing condition (by comparing activity with vs. without the denaturant and requiring retained activity, e.g., ≥5%/≥40%), and iterative/evolutionary versions of that method. I could not verify whether the '401 issued with such method claims or with additional independent claims beyond the set above. The related family member US 10,865,400 B2 (a division of the '731 application) issued with 18 claims and includes method‑type claims, so an independent method claim in the '401 is plausible but unconfirmed by my retrieval. Flag as uncertain.

---

4. Litigation / CAFC 2026 status for this specific patent

• No Federal Circuit 2026 docket for US 9,447,401 was found in my searches. The 2026 CAFC activity and PTAB activity I found concerns later family members, not the '401 itself:
  • Halozyme, Inc. v. Merck Sharp & Dohme LLC, No. 2:25‑cv‑03179 (D.N.J.) — the related patents identified are 10,865,400; 11,041,149; 11,066,656; 11,952,600; 12,018,298; 12,037,618; 12,049,652; 12,054,758; 12,091,692; 12,104,185; 12,152,262; 12,077,791; 12,195,773; and 12,264,345. US 9,447,401 is not listed among the asserted patents.
  • PTAB PGRs brought by Merck (e.g., PGR2025‑00004, ‑00006, ‑00009, ‑00017, ‑00039, IPR2026‑00312) target later patents in the same family (e.g., 11,952,600; 12,110,520; 12,018,298; 12,152,262). In 2026 the PTAB issued several final written decisions holding such claims unpatentable under 35 U.S.C. §112 for lack of written description and enablement, reasoning that the broad "modified PH20" genera are not adequately described/enabled. Those decisions construe "modified PH20 polypeptide" to require hyaluronidase activity.
  • In those proceedings US 9,447,401 is cited as a prior‑art/related patent and submitted as an exhibit (e.g., EX1025/EX1026, file‑history materials), and it is the root of the priority chain: the '401 application (13/694,731) is the common ancestor identified in the '520 patent's continuity statement.

Bottom line: As of the information available to me, US 9,447,401 is the family‑root patent for Halozyme's modified‑PH20 estate; it is not itself a defendant‑side target in the 2026 CAFC appeals or the enumerated PGRs/New Jersey case I could locate. I cannot state with certainty that no appeal naming the '401 exists — my search coverage is not exhaustive — but nothing in the results indicates one. If you need a definitive litigation check, a PACER/CAFC docket query by patent number would be required.

I'll research litigation involving US patent 9447401. Let me search multiple sources.

The initial results returned several "740" patents that are different patents (e.g., 8,944,740) — I will exclude those. Let me search specifically for this patent's owner and this number.

Let me search for litigation specifically naming the '401 patent, including PTAB proceedings and the related family.

Litigation Involving U.S. Patent No. 9,447,401

Bottom line up front

Based on my searches, I could not identify any court case or PTAB proceeding that specifically names or asserts U.S. Patent No. 9,447,401 ("the '401 patent") as a patent-in-suit. I want to be explicit about the confidence level here: my searches surfaced the patent's family litigation extensively, but not the '401 patent itself as an asserted patent. If litigation naming the '401 patent exists, I did not find it.

Two important caveats about my process:

1. Search results repeatedly used the shorthand "the '740 patent" to mean U.S. Patent No. 8,944,740 — a completely different patent. Per your strict rule, I did not auto-correct or merge these. One such hit was Moskowitz Family LLC v. Globus Medical Inc., No. 6:19-cv-672 (W.D. Tex., filed Nov. 20, 2019, transferred to E.D. Pa. as 2:20-cv-03271), and another was a case involving CIG v. Sand Revolution over "8,944,740." These involve 8,944,740, not 9,447,401, and are excluded.
2. Google Patents' page for US9447401 carries a "Family has litigation — First worldwide family litigation filed" indicator (with a Darts-ip link, family=47595052). That confirms litigation exists somewhere in this patent family, but the indicator does not identify which family member was asserted. It is not evidence that the '401 patent itself was asserted.

What the '401 patent is

U.S. Patent No. 9,447,401, "PH20 polypeptide variants, formulations and uses thereof," Halozyme, Inc., inventors Ge Wei, H. Michael Shepard, Qiping Zhao, Robert James Connor; Application No. 13/694,731, filed Dec. 28, 2012; priority Dec. 30, 2011; issued Sept. 20, 2016; assignee Halozyme. This is the foundational patent in Halozyme's "MDASE" modified-PH20 hyaluronidase family.

Family litigation (context only — these assert other patent numbers)

Because the '401 patent appears to be the parent of the family being litigated, here is the related activity. None of the case numbers below is reported to assert the '401 patent.

U.S. district court

• Halozyme, Inc. v. Merck Sharp & Dohme Corp., D.N.J. (New Jersey), Case No. 2:25-cv-03179, complaint filed April 24, 2025. Plaintiff: Halozyme, Inc. Defendant: Merck Sharp & Dohme Corp. The complaint asserts 15 patents — U.S. Nos. 10,865,400; 11,041,149; 11,066,656; 11,952,600; 12,018,298; 12,037,618; 12,049,652; 12,054,758; 12,077,791; 12,091,692; 12,104,185; 12,110,520; 12,152,262; 12,195,773; and 12,264,345 — all relating to berahyaluronidase alfa in SC Keytruda (Keytruda Qlex). Status: pending. Notably, the '401 patent is not on that list, though U.S. 10,865,400 appears to be a continuation claiming priority to the '401 application.

PTAB (United States)

• A series of post-grant reviews filed by Merck Sharp & Dohme LLC against Halozyme: PGR2025-00003, -00004, -00006, -00009, -00017, -00024, -00030, -00033, -00042, -00046, -00050, -00052, -00053, and -00087 (the last against U.S. 12,371,685). Reported outcomes: FWD finding all challenged claims of U.S. 11,952,600 unpatentable (PGR2025-00003, May 12, 2026); FWD finding all challenged claims of U.S. 12,152,262 unpatentable (PGR2025-00006, May 18, 2026); FWD finding claims of U.S. 12,110,520 unpatentable (PGR2025-00017, Sept. 1, 2026); institution denied for PGR2025-00087 (May 15, 2026). Grounds were principally §112 written description/enablement, not obviousness.
• IPR2026-00312, -00313, -00314 (Merck v. Halozyme), filed Mar. 22, 2026, challenging U.S. 10,865,400; 11,041,149; and 11,066,656. Reported as instituted.
• IPR2026-00176 (Halozyme v. Alteogen), filed Dec. 9, 2025, challenging Alteogen's U.S. 12,221,638; institution denied May 15, 2026.

United Kingdom

• Merck Sharp & Dohme (UK) Ltd v. Halozyme Inc., Patents Court (High Court), Case No. HP-2025-000041, filed Aug. 1, 2025. Concerns European patents, including EP(UK) 3 130 347 and EP 2 797 622. Reported decisions: [2026] EWHC 121 (Pat) (Richards J, Jan. 21, 2026) and [2026] EWHC 1838 (Pat) (Mellor J, July 20, 2026). Halozyme reportedly agreed to revocation of EP 622 during the action. Halozyme (Quinn Emanuel) acts for the defendant/counterclaimant; MSD (Hogan Lovells) for the claimant.

Germany

• Munich Regional Court granted Halozyme a preliminary injunction on Dec. 4, 2025 against MSD's distribution of Keytruda SC in Germany, based on EP 2 797 622. MSD reportedly filed a nullity action in the German Federal Patent Court (Bundespatentgericht) in August 2025.

What I could not confirm

• No case (district court, CAFC, or PTAB) specifically listing U.S. 9,447,401 as an asserted or challenged patent.
• One USPTO PTACTS document I retrieved (petition 1557766) contains the string "12/28/2012 PAT 9447401 which claims benefit of 61/796,208 … and claims benefit of 61/631,313," but this appears to be a priority/family listing inside a PGR/IPR exhibit, not an assertion of the patent. I would not treat it as litigation evidence.
• I did not obtain direct PACER or Unified Patents caselist results for this specific number (search step limits), so a comprehensive docket-level confirmation is not something I can claim.

If you need certainty, the confirmation path would be: (a) the Darts-ip family record linked from the Google Patents page (family 47595052); (b) PACER / Docket Alarm queries on "9,447,401" as a patent-in-suit; and (c) the PTAB API/PTACTS for any proceeding identifying the '401 patent as a challenged patent.

I'll start by verifying the structured data against public sources.

Let me dig deeper on proceedings specific to the '401 patent.

Now let me verify no IPR exists on the '401 patent specifically, and check the related litigation.

PTAB / AIA Trial Proceedings — US 9,447,401 B2

"PH20 polypeptide variants, formulations and uses thereof"
Assignee: Halozyme, Inc. (current: Halozyme Inc / Halozyme Therapeutics Inc) · App. 13/694,731, filed 2012-12-28 · Granted 2016-09-20 · Priority 2011-12-30 · Google Patents status: Active, adjusted expiration 2034-03-02 (US9447401B2)

---

Proceedings overview

The USPTO Open Data Portal reports zero AIA trial proceedings on file for US 9,447,401 — no IPRs, no PGRs, no CBMs, in any status (active, instituted, FWD, settled, or institution-denied) — which means there is no claim-level PTAB outcome to report for this patent and no § 315(e)(2)/§ 325(e)(2) estoppel running against any challenger; a defendant today faces a completely untested claim set that has never been through an AIA trial.

That "no activity" result is consistent with everything I could find publicly. It is not an artifact of a quiet patent: the '401 sits at the root of one of the most heavily attacked biotech families of 2025–2026, and multiple searches specifically keyed to the '401 number surfaced no petition, institution decision, or final written decision naming it as the challenged patent. In the one family proceeding where I found the '401 named repeatedly, it appears as Petitioner's exhibit EX1025 — i.e., as prior-art/background evidence about Halozyme's own priority chain — not as the subject patent (e.g., Merck's petition referencing "[t]he '731 Application … issued as U.S. Patent No. 9,447,401 (EX1025)").

Residual uncertainty, stated plainly: I could not complete a live PTAB E2E docket pull in this session, and the ODP block is only as current as its last ingest. Prior to relying on "no IPR," confirm independently (see Recommended next steps).

Why the '401 is structurally IPR-only

This is the single most important defensive takeaway on this patent. The theory that has been invalidating Halozyme's later PH20 patents — Merck's § 112(a) written description/enablement attack — is not available at the PTAB against the '401, because the '401 predates PGR eligibility and IPR is limited to § 102/§ 103. A challenger must litigate § 112(a) on the '401 in district court (or, in theory, at the ITC), where there is no 1-year statutory trial clock and no Phillips-with-a-PTAB-twist claim construction framework.

---

Proceedings on US 9,447,401

None. There are no per-proceeding entries to report, because the canonical structured list is empty and no web search surfaced a proceeding against this patent. Per the constraints of this task, I will not manufacture proceeding numbers, panels, FWDs, or dispositions.

What I can state affirmatively:

• No inter partes review of the '401 was identified.
• No post-grant review of the '401 was identified (and none is legally available — see table above).
• No CBM of the '401 was identified.
• No institution decision, FWD, adverse judgment, or settlement naming the '401 as subject patent was identified.
• No Federal Circuit appeal of any '401 PTAB decision exists, because there is no such decision.

---

Family context — proceedings that do NOT involve the '401 (flagged for accuracy)

These are real, verified proceedings against other, later-issued Halozyme patents in the same priority family (all descendants of App. 13/694,731). They are the source of the "Halozyme PH20 patents are being invalidated" narrative a defendant will encounter, and they matter as signals — but none of them cancels, narrows, or estoppes anything about the '401's claims.

Grounds in the Merck PGRs were § 112(a) written description, § 112(a) enablement, § 112(b) indefiniteness, and § 103 over the '429 Patent + Chao + POSA knowledge. The Board's institution rationale was characterized by Banner Witcoff's PTAB Highlights as: "Structural talk is not enough. Show us the goods." — i.e., institution where Halozyme's "modified PH20 polypeptide" genus was defined by vague structural recitations with no roadmap to make or test the full scope. (Banner Witcoff, PTAB Highlights; PGR2025-00004 FWD; PGR2025-00017 Institution Decision)

Also relevant: Google Patents flags the family as having litigation ("First worldwide family litigation filed," Darts-IP family 47595052 → https://patents.darts-ip.com/?family=47595052). Reporting describes a Halozyme suit in the District of New Jersey against Merck over Merck's subcutaneous pembrolizumab (Keytruda Qlex / ALT-B4), with trial projected around May 2030, and Merck's petitions noting the accused product is an unapproved biologic shielded by 35 U.S.C. § 271(e)(1). I could not confirm from the sources retrieved whether the '401 itself is asserted in that case — my docket-level search on the asserted-patent list was cut off. Treat that as an open verification item, not a finding.

---

Strategic summary

Claim status on this patent: everything is UNTESTED. No claim of the '401 has been canceled, disclaimed, confirmed, or construed by the PTAB. The five-plus unpatentability FWDs and the ~66 claims Halozyme statutorily disclaimed across nine patents all involve sibling/child patents (the '600, '298, '262, '035, '520, and higher-numbered continuations), not the '401. Anyone who tells you "the Halozyme PH20 claims are dead" is describing a different patent number — check the number on the face of the demand letter or complaint.

Estoppel landscape: empty. Because no AIA trial was ever instituted on the '401, no § 315(e)(2) or § 325(e)(2) estoppel attaches to any ground against this patent. Merck's PGRs are on different patents, so its estoppel does not run to the '401 either. Practically, that means a defendant/supplier/distributor facing the '401 has the full § 102 and § 103 universe available for an IPR — including art the Board has already credited in the sibling PGRs (the '429 Patent and Chao were the substantive obviousness backbone; note that within an IPR on the '401 those references would be § 103 art only, not § 112 evidence). It also means no petitioner has yet consumed the PTAB option on the '401, so there is no "second-bite" or § 325(d) baggage.

The real attack surface is § 112(a), and it is venue-limited. The Board's repeated holding across the siblings — that the common disclosure does not describe or enable the claimed genus of modified PH20 polypeptides — logically reaches the '401, which contains "generally the same" disclosure and shares the same 2012-12-28 specification and 2011-12-30 priority chain (Merck asserted exactly this linkage: the '731 Application "'alters several passages of the provisional disclosures, adds new examples and tested mutants,'" and does not support later claims). But a PTAB petitioner cannot raise it against the '401. For a defendant, that cuts both ways: a strong invalidity story that cannot be deployed in IPR must be litigated in court (or via ITC/declaratory judgment), and there is no PTAB-driven stay to shortcut it.

Priority-date leverage. With a 2011-12-30 priority claim and a pre-AIA § 102/§ 103 analysis, art after 2011-12-30 is out. Conversely, if a defendant can prove the '401's claims lack § 112(a) support in the 2011-12-30 provisionals (Merck's theory for the children), the effective date shifts and post-2011 art opens up — that is the highest-value discovery target in the file.

Pattern signals. One petitioner (Merck Sharp & Dohme LLC) is running a coordinated, multi-patent PGR campaign against the family, with Halozyme responding by statutory disclaimer of large claim blocks rather than defense — a signal that the Patent Owner itself does not regard much of the genus claims as defensible, and which the Board and Merck used against it ("If Halozyme actually believed the merits of Merck's § 112(a) challenges were weak, it would not have disclaimed all these claims"). Halozyme, for its part, has gone offensive with its own IPR (against Alteogen's US 12,221,638) and lost at institution. No defensive aggregator (Unified Patents, RPX, etc.) appears anywhere in this chain; the challenger is a commercial adversary, not a crowd-funded validity clearinghouse. The Board's "structural talk is not enough" framing, plus the sister-patent FWDs, make this family a known-quantity target for any future petitioner — the playbook is written; only the patent number changes.

---

Recommended next steps

1. Confirm the null result directly before relying on it. Search USPTO PTAB E2E (https://ptacts.uspto.gov/ptabweb/) by patent number 9447401 and by "Halozyme" as patent owner, and check the PTAB Decisions/statistics portal. The ODP block is authoritative-as-of-ingest; PTAB E2E is authoritative-as-of-now.
2. Check the '401 file wrapper for ex parte reexamination. Reexam requests do not appear in AIA-trial data and would be invisible to both the ODP block and PTAB E2E. Pull the '401 wrapper via USPTO Patent Center (https://patentcenter.uspto.gov/) and confirm no § 302/§ 311 reexam order is outstanding. A reexam would raise different estoppel/claim-scope issues than an IPR.
3. Confirm whether the '401 is in the New Jersey litigation. Pull the Halozyme v. Merck docket via CourtListener RECAP (https://www.courtlistener.com/?q=%229447401%22) or PACER and list the asserted patents. If the '401 is asserted, the § 112(a) story from the sibling PGRs becomes directly usable in that court; if it is not asserted, that itself tells you whether the Patent Owner views the '401 as its strongest or weakest family member.
4. Monitor the family FWDs for CAFC appeals. Multiple adverse FWDs against sibling patents (PGR2025-00003, -00004, -00006, -00009, -00017) were decided in the May–September 2026 window and are within the 63-day appeal clock. A consolidated Federal Circuit appeal on the § 112(a) written-description holding would produce the first appellate precedent on the PH20 genus — and would be `informative for the '401's § 112 exposure even though the '401 is not on appeal. No appeal on the '401 can exist, since there is no '401 PTAB decision.
5. If you are a defendant: do not assume the sibling-patent FWDs invalidate the '401. Verify the patent number in any assertion first. If the demand letter cites 9,447,401, you are dealing with an untested, un-estopped claim set whose priority and disclosure are nonetheless identical to patents the Board has repeatedly found unsupported — the invalidity case exists, but it must be built in district court, not at the PTAB, and it must be built on § 112(a) plus pre-2011-12-30 prior art, not on the Merck PGR record as such.

Nothing in this report should be read as a finding that the '401's claims are invalid, narrowed, or canceled. They are simply untested.

I'll research the assignment record for US 9,447,401, starting with the USPTO assignment data and cross-referencing legal events.

Let me try to locate the actual assignment records with reel/frame numbers.

Assignment Record — US 9,447,401 B2

Data-access caveat (read first). I was unable to retrieve the raw USPTO Assignment Center records (reel/frame, execution dates, and especially the correspondent of record) for this patent within this session — the Assignment Center search pages were not retrievable through my search tooling, and no third-party mirror surfaced a "Patent Assignment Abstract of Title" for 13/694,731 / 9,447,401. What follows is reconstructed from the legal-events record embedded in the authoritative Google Patents text supplied for this patent, which does surface three recorded assignment events. Fields I could not verify (reel/frame numbers, execution dates, correspondent) are explicitly marked not retrieved rather than estimated. Per the operating rules, I will not fabricate them.

---

Inventors

All four named inventors, with residence of record as listed on the application (per the PTAB-filed application transmittal surfaced in search):

• All four inventors assigned to the Halozyme corporate group (see timeline). This is the ordinary employee-inventor pattern for a corporate research program — three of four are San Diego–area residents, consistent with Halozyme's Sorrento Valley HQ.
• Unusual-pattern check: not present. There is no evidence any inventor departed the assignee within 12 months of filing, and no inventor-held residual rights that were later sold off. The assignment to the employer was recorded ~2.5 months after filing (2013-03-08), i.e., normal prosecution-housekeeping timing, not a pre-sale cleanup.
• Minor data inconsistency to flag: Shepard's residence is rendered "Eugene, OR" in one family document and "Springfield, OR" in another (both surfaced from PTAB filings for related family members). These are adjacent Oregon cities; likely a change of address between filings or a transcription variance. Not material to ownership.

---

Original assignee

Halozyme, Inc., San Diego, CA (per the issued patent's assignee data). Google Patents additionally lists Halozyme Therapeutics, Inc. among "current assignees."

• Corporate structure (important for reading the timeline). Halozyme is a two-tier group. Per the company's own 10-K definition: "References to 'Halozyme,' 'the Company,' 'we,' 'us' and 'our' refer to Halozyme Therapeutics, Inc. and its wholly owned subsidiary, Halozyme, Inc., and Halozyme, Inc.'s wholly owned subsidiary, Halozyme Holdings Ltd." So Halozyme Therapeutics, Inc. (NASDAQ: HALO) is the public parent; Halozyme, Inc. is its wholly owned operating subsidiary. The 2013 recordings below move rights within this group — not out of it.
• Primary line of business: clinical-stage-to-commercial biopharmaceutical company; developer of the rHuPH20 / ENHANZE® drug-delivery platform (recombinant human PH20 hyaluronidase used to enable subcutaneous administration of co-administered biologics).
• Product embodying the technology: Yes — Halozyme commercializes Hylenex® and XYOSTED® and licenses ENHANZE® (rHuPH20) to Roche, Takeda, J&J/Janssen, BMS, Pfizer, Eli Lilly, argenx, ViiV, Chugai, and others. Caveat: the commercial enzyme is wild-type recombinant human PH20; the '401 claims modified (variant) PH20 polypeptides, so whether the shipping product literally embodies the '401 claims is not established from the record — the patent reads as an upstream variant/portfolio patent covering the platform rather than the marketed enzyme per se.
• Current status: Operating, publicly traded (NASDAQ: HALO), headquartered San Diego with offices in Ewing NJ, Minnetonka MN, and Boston MA. Not acquired, dissolved, or in bankruptcy. (Recent 2026 press: interim-CFO appointment of David Ramsay, effective 2026-03-23.)

---

Assignment timeline

Reconstructed from the Google Patents legal-events record supplied in the authoritative text. All three events below carry the same recorded date of 2013-03-08 (i.e., all recorded together), which is ~10 weeks after the 2012-12-28 filing.

Reel/Frame: not retrieved. I could not obtain the reel/frame identifiers or the correspondent of record for any of the three records; the Google Patents legal-events feed does not expose those fields.

1. Executed: not retrieved / Recorded 2013-03-08 — Reel NNNNNN/NNNN (not retrieved)

   • Conveyance: Assignment of assignors' interest (see document for details)
   • Assignor: HALOZYME THERAPEUTICS, INC.
   • Assignee: HALOZYME, INC.
   • Correspondent: not retrieved
   • Context: Internal corporate reorganization — downward transfer of rights from the public parent to its wholly owned operating subsidiary; both entities are in the same Halozyme group. Not a transfer to an unaffiliated entity.

2. Executed: not retrieved / Recorded 2013-03-08 — Reel NNNNNN/NNNN (not retrieved)

   • Conveyance: Assignment of assignors' interest
   • Assignor: SHEPARD, H. MICHAEL (inventor)
   • Assignee: HALOZYME, INC.
   • Correspondent: not retrieved
   • Context: Inventor-to-employer assignment (individual link).

3. Executed: not retrieved / Recorded 2013-03-08 — Reel NNNNNN/NNNN (not retrieved)

   • Conveyance: Assignment of assignors' interest
   • Assignor: CONNOR, ROBERT JAMES; WEI, GE; ZHAO, QIPING (inventors)
   • Assignee: HALOZYME THERAPEUTICS, INC.
   • Correspondent: not retrieved
   • Context: Inventor-to-employer assignment (group link) — the three San Diego–based inventors assigned to the parent entity.

No post-issuance assignments are recorded. The legal-events record shows no assignment activity after 2013-03-08 — nothing at issue (2016-09-20), nothing during the 2025–2026 litigation campaign. The chain therefore terminates inside the original Halozyme corporate group. That continuity is itself a finding: there has been no outbound transfer, no securitization lien recording, and no assignment to a licensing vehicle.

(Anomaly worth flagging, not a contradiction of the above: the three records show inventors splitting between two Halozyme entities — Shepard → Halozyme, Inc.; Connor/Wei/Zhao → Halozyme Therapeutics, Inc. — with a single entity-to-entity record moving Halozyme Therapeutics → Halozyme, Inc. This is consistent with a consolidated intra-group title cleanup recorded on one date. I cannot confirm the directional reading of each record without the underlying assignment instruments, which were not retrievable.)

---

Timeline diagram

timeline
    title Ownership of US 9447401
    2011 : Priority provisionals filed
    2012 : Application 13/694,731 filed
    2013 : Inventor assignments recorded
         : Halozyme Therapeutics to Halozyme Inc
    2016 : Patent US 9447401 issues
    2025 : Halozyme sues Merck over family

---

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. The only recorded assignees are Halozyme, Inc. and Halozyme Therapeutics, Inc. (both recorded 2013-03-08). No "IP / Patents / Licensing / Holdings / Ventures" suffix entity appears; no registered-agent-service address is in the chain; both assignees are the actual operating pharmaceutical company and its parent. The Halozyme Therapeutics → Halozyme, Inc. record is an intra-group move, not a move to a licensing-only shell.

2. Known asserter in the chain — NOT PRESENT. Neither current nor prior assignee matches any public NPE roster (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Round Rock, etc.). The assignee is a NASDAQ-listed operating biopharma (HALO).

3. Repeat correspondent across the chain — UNCLEAR / INSUFFICIENT DATA. The correspondent of record for the 2013-03-08 recordings was not retrievable, so I cannot determine whether a single attorney/firm handled all three records, and I will not infer one. Note for the file: Halozyme's prosecution correspondence for this family is handled by Morgan, Lewis & Bockius LLP (PH), 2222 Market Street, Philadelphia, PA 19103 (per family application transmittal documents) — but that is prosecution counsel, and prosecution counsel is not evidence of NPE-style assignment handling. No finding.

4. Cascading transfers — NOT PRESENT. Three assignments were recorded contemporaneously (all 2013-03-08), not cascading over a <24-month window, and all flow to/from the same two related Halozyme entities. No chained third-party LLCs, no shared-agent address pattern.

5. Pre-litigation transfer — NOT PRESENT. The assignments are dated 2013-03-08; the first family litigation I could identify (Halozyme, Inc. v. Merck Sharp & Dohme LLC, No. 2:25-cv-03179, D.N.J.) was filed in 2025 — roughly twelve years later. There is no assignment within 6 months preceding suit. (Indeed, per the previously generated section, the '401 itself is not enumerated among the asserted NJ patents; the family members asserted are the later continuations/divisions.)

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 proceeding; Halozyme is an active, profitable operating company (FY2023 revenue ≈ $829M; operating margin ≈ 40%).

7. Privateering — NOT PRESENT. There is no transfer to a third-party NPE asserting on Halozyme's behalf. Halozyme asserts as the patent owner itself, against a commercial counterparty (Merck), which is direct operating-company enforcement, not privateering. (The counter-posture — Merck/PTAB challenging the later family claims under §112 in PGR2025-00004/-00006/-00009/-00017/-00039 and IPR2026-00312 — further confirms this is an operating-company-vs-operating-company dispute.)

8. Defensive aggregator — NOT PRESENT. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. It terminates at Halozyme, Inc. (operating subsidiary).

---

Verdict

Operating-company assertion.

Justification: the entire recorded chain consists of three assignments all recorded on 2013-03-08 and all confined to the Halozyme corporate group (inventors → Halozyme entities; Halozyme Therapeutics, Inc. → its wholly owned operating subsidiary Halozyme, Inc.), with no post-issuance transfers of any kind through the 2025–2026 enforcement campaign. The current owner is a NASDAQ-listed operating biopharmaceutical company (NASDAQ: HALO) that commercializes rHuPH20/ENHANZE® and is suing a commercial competitor (Merck) over the same patent family — the antithesis of the shell-LLC/NPE pattern. Two caveats: (i) I could not verify reel/frame numbers or the correspondent of record, so signal 3 is genuinely unresolved; and (ii) the marketed enzyme is wild-type rHuPH20, so literal embodiment of the '401 variant claims by a shipping product is not established.

Verify at: USPTO Assignment Center — search by patent number 9447401 or application 13/694,731 at https://assignmentcenter.uspto.gov/ (mirror: https://assignment.uspto.gov/patent/index.html). Obtaining the reel/frame and correspondent fields there would close out signal 3, the one open item.

I'll search for the specific patent and its cited prior art references.

The search results are noisy (some concern a different patent, 8,944,740 — a sand-handling patent, which I will not conflate with 9,447,401). Let me pull the actual "References Cited" section for 9,447,401.

I need the actual front-page "References Cited" list. Let me try to retrieve the patent PDF and the Google Patents citation table.

Important scoping note before the analysis

Two things must be flagged up front, because the search environment is contaminated and the strict rule about literal identifiers matters here.

1. Do not conflate US 9,447,401 with US 8,944,740. Multiple search results returned the sand-handling patent US 8,944,740 B2 (Sand Revolution II LLC v. Continental Intermodal Group, IPR2019-01393, Forsyth/Haskins/Blackman/Grotte). That is a completely unrelated patent that happens to share the digit string "8,944,740." None of that material is prior art to, or about, US 9,447,401. I have excluded it.

2. What I was able to retrieve. I could not retrieve the verbatim front-page "(56) References Cited" block of US 9,447,401 B2 in the searches available to me, and I hit the tool step limit before fully resolving it. Rather than reconstruct that list from memory, I am going to give you (a) exactly what the searches did return, (b) the references that are demonstrably material to § 102 on this family, and (c) an honest statement of what remains unverified. I am not going to fabricate a citations list.

---

1. Positive identification of the target

This is the family-root patent of Halozyme's "MDASE" modified‑PH20 estate. Confirmed via the Google Patents family table: 13/694,731 → division/continuation chain 15/226,489 (US 10,865,400) → 16/824,572 (US 11,041,149) → 16/912,590 (US 11,066,656) → 17/327,586 (US 12,037,618), etc. This matters for § 102 because it means the '401 is prior art to its own descendants, not the reverse.

---

2. What the searches actually returned on citations

(a) The '401's own front page — not retrieved verbatim. I could not confirm the exact (56) block for the '401.

(b) A same-specification sibling's front page — partially retrieved. From the Google Patents PDF of US 11,066,656 B2 (app. 16/912,590, a division of 13/694,731 — same specification), the "(56) References Cited / U.S. Patent Documents" list begins with a large block of older formulation, PEG, packaging, and expression art, including:

3,536,809 A (10/1970, Applezweig) 424/28; 3,598,123 A (8/1971, Zaffaroni); 3,630,200 A (12/1971, Higuchi); 3,710,795 A (1/1973, Higuchi); 3,845,770 A (11/1974, Theeuwes); 3,916,899 A (11/1975, Theeuwes); 4,002,531 A (1/1977, Royer) 435/188; 4,008,719 A (2/1977, Theeuwes); 4,044,126 A (8/1977, Cook); 4,179,337 A (12/1979, Davis) 435/181; 4,364,923 A (12/1982, Cook); 4,414,209 A (11/1983, Cook); 4,769,027 A (9/1988, Baker) 424/493; 4,952,496 A (8/1990, Studier) 435/91.41; 4,980,286 A (12/1990, Morgan) 435/371; 5,033,252 A (7/1991, Carter); 5,052,558 A (10/1991, Carter); 5,059,595 A (10/1991, Le Grazie); 5,073,543 A (12/1991, Marshall) 514/21; 5,120,548 A (6/1992, McClelland); 5,122,614 A (6/1992, Zalipsky) 548/520; 5,323,907 A (6/1994, Kalvelage); 5,324,844 A …"

Caveat: this list is from the '656 (a 2020-filed division), not the '401. Because the specifications are the same and the older patents are boilerplate applicant-cited background art, it is likely the '401 front page carries the same or a subset of this block — but I am flagging this as an inference, not a verified quotation of the '401.

(c) Non-patent literature (retrieved from a related family/EPO document's "references cited in the description"):

• Bookbinder, Hofer, Haller, Zepeda, Keller, Lim, Edgington, Shepard, Patton, Frost, "A recombinant human enzyme for enhanced interstitial transport of therapeutics," J Control Release 2006, 114:230‑241
• Borders & Raftery, "Purification and Partial Characterization of Testicular Hyaluronidase," J Biol Chem 1968, 243:3756‑3762
• Chao, Muthukumar, Herzberg, "Structure of human hyaluronidase‑1, a hyaluronan hydrolyzing enzyme involved in tumor growth and angiogenesis," Biochemistry 2007, 46:6911‑6920
• Frost, "Recombinant human hyaluronidase (rHuPH20): an enabling platform…," Expert Opin Drug Deliv 2007, 4:427‑440
• Hofinger, Bernhardt, Buschauer, "Kinetics of Hyal‑1 and PH‑20 hyaluronidases…," Glycobiology 2007, 17:963‑971
• Thomas, Yocum, Flament, "The INFUSE‑Morphine IIB Study…," J Pain Symptom Manag 2009, 38:673‑682
• Chen et al., J Biosci Bioeng 2016, 122:673‑678 — post‑dates the '401's 2011‑12‑30 priority; cannot be § 102 art against the '401.
• Kreidieh et al., World J Clin Oncol 2016, 7:87‑97 — also post‑priority; not § 102 art against the '401.

(d) The references that are actually litigated as prior art against this family (from the 2025 PTAB PGR institution decisions — these are the petitioner's art, not necessarily '401 front-page citations, but they are the most probative § 102/§ 103 references in the ecosystem):

• US 7,767,429 ("the '429 Patent") — cited by Merck as EX1005 in PGR2025‑00003 and the related PGRs; the Halozyme soluble‑PH20 (rHuPH20 / sHASEGP) patent.
• Chao et al. 2007 (HYAL‑1 crystal structure) — EX1006.
• Arming et al. (PH‑20 site‑directed mutagenesis) — EX1011; teaches D111N (~3% activity), E113Q and E249Q (no detectable activity).
• Zhang et al. (Hyal‑1 point mutants) — EX1010.

---

3. Reference‑by‑reference § 102 assessment

Legal frame: a reference anticipates under 35 U.S.C. § 102 only if a single reference discloses every limitation, arranged as in the claim. For a claimed genus, a prior‑art species within the genus anticipates; a prior‑art genus generally does not anticipate a claimed species. Front‑page citations include applicant‑cited background art that is not anticipatory.

Bottom line of the § 102 pass: on the record I could retrieve, no cited reference anticipates the independent claims of the '401. Every candidate reference fails for at least one of three reasons: (i) it does not disclose PH20 at all (the formulation/PEG/packaging block); (ii) it discloses only the unmodified soluble PH20 ('429, Bookbinder, Frost, Thomas/Flament); or (iii) it discloses a different enzyme or a substitution that the claims expressly exclude (Chao, Arming, Zhang). The cited art is essentially § 103 (obviousness) and background/enablement art.

---

4. The self‑evident § 102 carve‑out in the patent text (grounded in the specification you supplied)

The '401 specification contains an express proviso that is directly a § 102 exercise. For a modified PH20 having SEQ ID NO:3, 7, 69, or 72:

• Single‑replacement claims exclude: V12A, N47A, D111N, E113Q, N131A, R176G, N200A, N219A, E249Q, R252T, N333A, N358A (positions by reference to SEQ ID NO:3).
• Two‑replacement claims exclude: P13A/L464W, N47A/N131A, N47A/N219A, N131A/N219A, N333A/N358A.

These are precisely the known N‑linked glycosylation‑site substitutions (N47, N131, N200, N219, N333, N358) and the catalytic‑residue substitutions (D111N, E113Q, E249Q) that appear in the prior art — including Arming et al. (D111N, E113Q, E249Q) and the glycosylation‑site art. That proviso is a strong indication that the applicant was designing around specific § 102/§ 103 art, and it is the most productive line for an invalidity analyst: any prior‑art reference disclosing those exact single or double substitutions cannot anticipate, and vice versa.

I also note the specification states the modified polypeptide must not consist of any of SEQ ID NOS: 3, 6‑66, 69‑72, 856‑861, 869, or 870 — i.e., the unmodified/native sequences are positively excluded, which forecloses an anticipation theory built solely on a native‑PH20 reference such as the '429.

---

5. Limitations of this answer / how to close the gap

I want to be explicit rather than pad the list:

1. The verbatim (56) block of US 9,447,401 B2 was not retrieved. The table above is built from (a) a same‑specification sibling's front page, (b) family NPL lists, and (c) the references actually asserted in the 2025 PGRs. Treat the dates and citations I quoted directly as reliable and the proposition that the '401 front page contains exactly this set as unverified.
2. Front‑page citations are not the test for invalidity. The probative § 102 art for a PH20‑variant patent is typically (i) the applicant's own earlier Halozyme filings (US 7,767,429 and the priority provisionals 61/631,313 and 61/796,208), and (ii) PH20 mutagenesis literature (Arming et al.; the N‑glycosylation‑site work). Those are the references to pull in full text.
3. To get the definitive list, pull the '401's front page and its full file history from USPTO PatentCenter / Global Dossier (application 13/694,731) and read the examiner's cited references and the IDS transmittals — that distinguishes examiner‑cited art (probative) from applicant‑cited background art. A PACER/CAFC docket check by patent number would also confirm the litigation posture (my earlier summary found no 2026 CAFC appeal naming the '401, and that conclusion stands).

Consistency check with the earlier section: nothing here contradicts the prior summary. The prior summary's flag that method claims in the '401 were unverified remains open — my searches surfaced method/composition claims in descendant patents (e.g., position 313/320/324 substitutions and contraceptive/use claims in US 11,952,600 / 12,110,520 / 12,091,692), which are not the '401's claims and should not be attributed to it.

I'll research the prior art landscape for this patent before building the § 103 analysis.

Let me pin down the issued claim text and the patent's own cited prior art list.

Obviousness Analysis — U.S. Patent No. 9,447,401 B2

"PH20 Polypeptide Variants, Formulations and Uses Thereof" (Halozyme, Inc.)

35 U.S.C. § 103 / Graham v. John Deere / KSR Int'l v. Teleflex

---

0. Two threshold corrections to the earlier-generated sections

Before the § 103 analysis, two record items must be corrected — both materially affect the analysis.

(a) The issued claim 1 is NOT the broad genus described in the earlier summary. A third-party claim database renders the granted first claim as:

"1. A modified PH20 polypeptide that exhibits increased stability in the presence of a phenolic preservative, comprising an amino acid replacement in an unmodified PH20 polypeptide, wherein: the amino acid replacement confers the increased stability; the amino acid replacement is replacement with proline at a position corresponding to position 204 … with reference to … SEQ ID NO:3; corresponding amino acid positions are identified by alignment of the modified PH20 polypeptide with … SEQ ID NO:3; the modified PH20 polypeptide has at least 85% sequence identity to the sequence of amino acids set forth in SEQ ID NO:3 or 7; and stability is increased compared to the unmodified polypeptide without the amino acid replacement, wherein stability is measured by resistance to denaturation in the presence of a phenolic preservative."
— RPX/Insight patent record for US 9,447,401 B2, https://insight.rpxcorp.com/patent/[US9447401B2](/patent/US9447401B2)

If accurate, the '401 is not the "genus" patent at all — it is the F204P (Phe204→Pro) phenophile patent. This is corroborated by the family pattern revealed in the 2026 PTAB record: each sibling claims a different single position identified in the same screening campaign — the '590 patent claims a modification at position 307, the '520 patent at position 320, the '652 and '692 patents at position 313/324, etc. (PGR2025‑00017 FWD; PGR2025‑00033 institution decision; PGR2025‑00046 institution decision). The '401 fits that pattern at position 204.

Caveat / flag: the authoritative Google Patents text supplied to me did not include the claims, and the earlier analyst characterized claim 1 as a broad "at least one replacement" genus (retrieving only dependents 5–22). I cannot reconcile those two accounts and cannot verify the claim verbatim. Everything below proceeds on the F204P reading, and I flag where the analysis changes if claim 1 is in fact broader.

(b) "The Prior Art section of this page" did not load. The fetched Google Patents text contains only the prior-art keyword field ("position corresponding; polypeptide; agent; modified; seq") and no reference list. I therefore built the art base from (i) the references that actually appear in the 2024–2026 PTAB/CAFC record on this family (where the '401 is the root of the priority chain and its disclosure is "the common disclosure"), and (ii) the '401 specification's own admissions about the state of the art. If a front-page reference list exists that I did not see, it should be reconciled against § 4 below.

---

1. The claim as construed

Key structural note: element D makes the polypeptide backbone extraordinarily broad (any PH20 with ≥85 % identity to SEQ ID NO:3 or 7, any species), while element C pins the mutation to one residue. This asymmetry is important: § 103 is fought over element C; § 112 is fought over element D (and, per the 2026 FWDs, that is exactly where the family is failing).

---

2. Legal framework applied

• Graham factors: scope/content of prior art; differences; PHOSITA level; secondary considerations.
• KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007): a combination need not be taught expressly; "obvious to try," "design incentives," "finite number of identified, predictable solutions," and "ordinary creativity" count. But KSR does not eliminate the need for a rational underpinning and a reasonable expectation of success.
• In re Boesch, 617 F.2d 272, 276 (CCPA 1980) — "discovery of an optimum value of a result effective variable … is ordinarily within the skill of the art" (directly relevant to "which residue to mutate").
• In re Kubin, 561 F.3d 1351 (Fed. Cir. 2009) — "obvious to try" applies with force in the predictable‑screen context, but the Federal Circuit has repeatedly refused to apply it where the art is unpredictable and success is not reasonably expected.
• In re Alonso, 545 F.3d 1015 (Fed. Cir. 2008); Noelle v. Lederman, 355 F.3d 1343 (Fed. Cir. 2004); Ariad Pharms. v. Eli Lilly, 598 F.3d 1353 (Fed. Cir. 2010) — unpredictability in biotechnology cuts against obviousness as well as description.

---

3. Person having ordinary skill in the art (PHOSITA)

A Ph.D. (or M.S. + several years) in biochemistry, molecular biology, protein engineering or pharmaceutical sciences, with working knowledge of (i) recombinant expression of glycosylated human enzymes in CHO cells, (ii) site‑directed mutagenesis and site‑saturation/directed‑evolution library screening, (iii) hyaluronidase activity assays (microturbidity/USP), and (iv) protein formulation science, including the destabilizing effect of phenolic preservatives in multidose injectables. This is essentially the skill set the '401 Examples themselves apply (epPCR/site‑saturation → CHO‑S expression → microturbidity screen → m‑cresol stress).

---

4. The prior art available as of 2011‑12‑30 / 2012‑12‑28

The '401 specification's own admissions are themselves prior‑art evidence. The specification states: "PH20 hyaluronidase, such as rHuPH20, rapidly loses activity in the presence of preservatives, likely due to unfolding of the protein and subsequent aggregate formation"; "following incubation with 0.4 % m‑cresol for 4 hours, PH20 retains only about 10 % of its activity"; and the m‑cresol‑induced drop in unfolding temperature "from 44 °C to 24 °C." These admissions establish that the problem to be solved was known and that the art already understood the mechanism (hydrophobic phenolic compounds binding nonspecifically and perturbing structural integrity).

---

5. Element‑by‑element mapping, and where the case turns

Therefore every § 103 theory stands or falls on whether the art made F204P (a) suggested, or (b) the predictable output of a routine screen.

---

6. Combination A — the "generic variant + known formulation problem" theory

'429 (Frost) + Frost 2007 + the preservative‑formulation art (Maa/Hsu), optionally + WO 2010/077297

Motivation to combine: Strong and easy to articulate.

• The '429 patent expressly discloses soluble PH20₁₋₄₄₇ with a single substitution ("sHASEPGs") and claims pharmaceutical compositions thereof, alone or co‑administered with another agent, administered subcutaneously.
• Frost 2007 frames rHuPH20 as a co‑administration platform for SC delivery — the commercial driver toward high‑concentration, multidose, preservative‑containing presentations.
• The formulation art (and the '401's own background) establishes that phenolic preservatives are de rigueur for multidose injectables and are known to destabilize proteins, PH20 included.

Rational underpinning: "A POSA seeking a stable multidose PH20 formulation would (i) select a preservative from the known set, (ii) recognize that it destabilizes PH20, and (iii) screen PH20 variants for preservative resistance rather than abandon preservatives."

Weakness: Every step is generic. The '429 teaches that variants can be made; it does not teach which to make, and the claim's phenotype ("increased stability in a phenolic preservative") is not disclosed anywhere in the art as a property even being sought in PH20 before the inventors' screen. This is the In re Kubin problem: knowing that you may mutate a protein is not a suggestion to make a particular mutation with a particular result.

---

7. Combination B — the "rational design / rational mutagenesis" theory

Chao 2007 (+ homology model of PH20) + '429 / WO 2010/077297 + preservative‑formulation art

Motivation: Chao 2007 supplies the HYAL1 catalytic‑domain structure, from which PH20's fold can be modelled by homology (they are ~40 % identical and share the (α/β)₈ TIM‑barrel core). If position 204 is a solvent‑exposed hydrophobic residue, a POSA aiming to reduce non‑specific binding of hydrophobic phenol/m‑cresol could be motivated to reduce surface hydrophobicity.

Why this theory is materially weaker than it first appears:

1. Substituting Phe with Pro is counter‑intuitive as a stabilizer. Proline is a helix‑breaker and constrains φ; introducing Pro typically destabilizes secondary structure. Nothing in Chao 2007 or the '429 patent suggests Pro as the substitution of choice at a hydrophobic surface residue — the conventional playbook is Phe→Ala/Ser/Thr/Tyr.
2. The PH20 three‑dimensional structure was not known. In the Halozyme litigation record, an expert (Dr. Zalipsky) stated that "the three dimensional structure of rHuPH20 was unknown at the time of filing, and to the best of my knowledge, is still unknown" (Opening Expert Report, Halozyme litigation, E.D. Va. 1:16‑cv‑01580, Dkt. 96‑2, https://storage.courtlistener.com/recap/gov.uscourts.vaed.[356568](/patent/356568).96.2.pdf). Without a structure, "position 204 is a surface hydrophobe whose mutation reduces phenol binding" is a post‑hoc rationale, not an articulated pre‑invention one.
3. The art teaches that PH20 activity is fragile. Arming 1997 showed that even conservative changes (E113Q, E249Q, D111N) abolish or devastate activity; Zhang 2001 showed the same position‑specificity for HYAL1. The PTAB institution decisions cite these very references for the proposition that "even conservative mutations may significantly impact the PH20 polypeptide hyaluronidase function" (PGR2025‑00033 Inst. Dec.; PGR2025‑00046 Inst. Dec.).

---

8. Combination C — the "systematic scan" / KSR "finite number of predictable solutions" theory (the strongest theory)

'429 + Frost 2007 + Arming 1997 + Zhang 2001 + preservative‑formulation art, with the argument that the skilled artisan, faced with a known denaturation problem, would systematically substitute each residue of the 447‑residue catalytic domain (or at least the surface residues) and screen for retained activity in m‑cresol.

Motivation/incentives that support this:

• The '429 patent already taught engineered single‑substitution PH20 variants and CHO expression, so "make one substitution, test it" was squarely within the art.
• The phenotype to be optimised (activity in preservative) is a result‑effective variable discoverable by routine assay (microturbidity), and In re Boesch makes optimising such a variable "ordinarily within the skill of the art."
• KSR instructs that where there are "a finite number of identified, predictable solutions," a POSA has good reason to pursue the known options.
• The '401 inventors did exactly this — an exhaustive single‑site scan of ~6,000–7,000 variants (~409 in the reported stability rescreen, per the '590 petition record) — which Petitioners in the sibling PGRs characterised as an ordinary "trial and error approach." If the claimed F204P emerged from a routine, exhaustive screen, that is textbook obvious‑to‑try.

Why even this theory is not a clean win for a challenger:

• A "finite number of predictable solutions" must exist. Here the solution space is 447 positions × ~19 residues ≈ 10⁴–10⁵ single variants — large, though tractable — and, decisively, the art provided no predictor of which variants would retain activity, much less which would gain preservative resistance. The '401's own data show that enormous numbers of substitutions are loss‑of‑function (they are claimed as "inactive mutants"; Tables 5 and 10). Screening a library in which the vast majority of members fail is the antithesis of a "predictable" solution set.
• Petitioner‑side experts conceded in the sibling PGRs that the specification's own stability data are "largely meaningless"/lacking "insight" (PGR2025‑00024 petition). Paradoxically, that concession helps the patentee on § 103: if the data reveal no pattern, then a POSA could not have predicted F204P's phenotype.
• The magnitude of the effect is far outside the noise: wild‑type PH20 retains ~10 % activity after 4 h in 0.4 % m‑cresol, whereas the '401 reports F204P retained "approximately equal" activity with and without m‑cresol, and calls it "highly resistant." That is a ~10‑fold difference against a screen whose positive control varies by up to ~97 % across runs (per the '590 petition record).

---

9. A fourth theory I would not press: anticipation / single‑reference obviousness

If any pre‑2011 reference discloses proline at position 204 of PH20 — e.g., in a sequence listing of a Halozyme application in the '429 / US 2004/0268425 / WO 2010/077297 chain — the analysis collapses into a § 102 or single‑reference § 103 case, and the "unexpected results" defence weakens considerably. I could not identify such a disclosure, and I could not retrieve the '401's own front‑page reference list (see § 0(b)). This is the single highest‑value investigative step remaining.

---

10. Counterarguments a patentee will (and should) run

1. No suggestion of position 204. None of '429, Frost 2007, Chao 2007, Arming 1997, Zhang 2001 or WO 2010/077297 identifies residue 204 or proline substitution as a route to preservative resistance. Under In re Kubin, "obvious to try" fails where the art gives no direction as to which of a large number of candidates will work.
2. Unexpected results. The ~10‑fold improvement in m‑cresol resistance from a single Phe→Pro change — a change ordinarily expected to destabilise — is strong evidence of non‑obviousness (In re Soni; In re Chupp).
3. Teaching away from Pro. The helix‑breaking nature of proline and the Arming/Zhang data showing activity is destroyed by conservative substitutions in this fold argue against a reasonable expectation of an active and more stable Pro variant.
4. No structural basis for rational design. rHuPH20's three‑dimensional structure was unavailable (Zalipsky report), so a homology‑modelling rationale for position 204 is post‑hoc.
5. Secondary considerations (if provable): long‑felt need for multifdose hyaluronidase formulations; failure of others to solve it via excipient strategies (the '401 spec itself describes hyaluronidase inhibitors as stabilizers, i.e., a workaround the inventors rejected); commercial success of the resulting products; copying/licensing (the siblings are being litigated for infringement against Keytruda Qlex, D.N.J. 2:25‑cv‑03179); and, importantly, Petitioner's own conduct — Merck has attacked this family exclusively on § 112 written description/enablement and has not obtained an obviousness win, which is weak but not irrelevant evidence that no clean § 103 combination existed in the art.

---

11. Litigation cross‑check (per the earlier sections, extended)

• US 9,447,401 is not among the patents asserted in Halozyme, Inc. v. Merck Sharp & Dohme LLC, D.N.J. No. 2:25‑cv‑03179, nor among the patents challenged in the enumerated PGRs (PGR2025‑00003…‑00087) or IPR2026‑00312/‑00313/‑00314. No CAFC docket naming the '401 was found. So the § 103 question is academic as to current disputes, but the '401 remains (i) the family root in the priority chain and (ii) a cited exhibit in those proceedings.
• Ground alignment matters: The 2026 FWDs invalidating siblings (e.g., PGR2025‑00003, ‑00006, ‑00017 at https://biologicshq.com/wp-content/uploads/2026/09/PTAB-PGR2025-00017-Merck-v.-Halozyme-Keytruda-Qlex-Final-Written-Decision-Unpatentable.pdf) rest on § 112, not § 103. That is a meaningful signal: sophisticated Petitioners with full access to the specification chose the written‑description/enablement attack over obviousness — consistent with my conclusion below that the obviousness case is the weaker of the two attacks on this family.

---

12. Bottom line

My assessment: On the record I can assemble, F204P is not rendered obvious by the prior art with the clear‑and‑convincing clarity a § 103 challenge requires. The art establishes (i) the PH20 backbone, (ii) that PH20 variants can be made, and (iii) that phenolic preservatives destabilise proteins and PH20 in particular — but it nowhere identifies position 204, nowhere suggests proline substitution at a hydrophobic residue as a stabilisation strategy (indeed it teaches the opposite as a general matter), and supplies no predictor that any given substitution would retain full activity. The best challenger's theory is a KSR "routine optimisation of a result‑effective variable" argument, and it is undercut by the very unpredictability the Petitioners elsewhere used successfully under § 112: the art and the specification alike show that almost all substitutions destroy PH20 activity, so a "finite number of predictable solutions" cannot fairly be said to exist.

The claim's real vulnerability is § 112, not § 103 — specifically the combination of (a) a single disclosed species (F204P) and (b) an ≥ 85 %‑identity backbone spanning SEQ ID NO:3 or 7 across all PH20 orthologues. That is precisely the attack that has succeeded against the '401's children in 2026.

---

13. What I could not verify — and what would change the answer

1. The verbatim text of granted claim 1 and the full claim set. I used a third‑party rendering (RPX) that conflicts with the earlier analyst's description; verify against USPTO PatentCenter / the printed patent. If claim 1 is genuinely the broad "at least one replacement at a position selected from among 1, 11, 12, 14, 20 … 443" genus, the § 103 analysis changes substantially in the challenger's favour — a broad genus of thousands of variants with a functional "increased stability" limitation is far more vulnerable under KSR/In re Boesch.
2. The '401's own "References Cited" list (the "Prior Art section of this page" that did not load). Any U.S. patent or printed publication on that list disclosing a proline at position 204 (or a PH20 variant with increased preservative resistance) would convert this into a § 102 or single‑reference § 103 case.
3. The '401's prosecution history — specifically whether claim 1 was narrowed to F204P over art cited on the merits, and whether applicant relied on unexpected results. That would show what the Examiner already considered.
4. Whether any pre‑2011 Halozyme application's sequence listing discloses F204P (a very real possibility given Halozyme's parallel filings). This is the highest‑leverage remaining search.