US 11402381

I'll search for authoritative information on US patent 11402381.

Let me check for any CAFC/litigation activity and additional claim details.

Let me try to retrieve the full claim set and confirm there's no CAFC 2026 activity.

Summary of US Patent 11,402,381

I searched for this specific number and its variants. Here is what authoritative sources (the Google Patents full text you supplied, the USPTO/patentimages PDF, Justia, FreePatentsOnline, and Patent Leaderboard) show. I found no CAFC docket entry or 2026 appellate activity for this patent; see the caveat at the end.

Bibliographic data

Identifier note: the granted PDF renders the fourth-named inventor as "Birgit Wehml," which appears to be an OCR artifact of "Birgit Wehnl" (the spelling used on Google Patents and Justia). I am reporting both spellings literally rather than silently correcting.

Abstract (as issued)

"The present disclosure relates to a method of identifying an individual having non-small cell lung carcinoma as to be treated by chemotherapy based on marker molecules cytokeratin-19 fragments (CYFRA 21-1) and carcinoembryonic antigen (CEA) as well as the use of the marker molecules for the identification of an individual to be treated by chemotherapy."

Overview of the independent claim

The patent has one independent claim (claim 1); claims 2–8 are dependent (e.g., claim 3 depends on claim 2, and claim 8 depends on claim 1, per the Justia text). In plain language, claim 1 covers a method of identifying that a human patient diagnosed with non-small cell lung carcinoma, squamous cell carcinoma subtype (SCC-NSCLC), stage I or II should receive chemotherapy after surgical resection of that cancer. The steps are:

• (a) Obtain a sample from such a human individual;
• (b) Measure, using a specific binding agent for each, the amount/concentration of CYFRA 21-1 protein and CEA protein in the sample;
• (c) Compute a combined value by weighted calculation of the two measured marker amounts;
• (d) Compare that combined value to a cut-off value established in a reference population also diagnosed with SCC-NSCLC stage I or II;
• (e) Identify the individual as one to be treated by chemotherapy following resection when an increased combined value is detected relative to the reference population; and
• (f) Treat the individual (the claim text as retrieved is truncated at this step, so the exact wording of the final "treating" limitation is not fully verified — see caveat).

The dependent claims add limitations such as blood-sample types (serum/plasma/whole blood), the timing of sample collection relative to surgery, the use of the weighted/logarithmic combination formula, and the requirement that the binding agents be antibodies (claim 8).

Underlying concept (from the specification): SCC-NSCLC stage I/II patients can be split into high- vs. low-relapse-risk groups using a weighted combination of CYFRA 21-1 and CEA (in Example 2, C = 0.459·log₂(CYFRA 21-1) + 0.017·log₂(CEA), with a median cut-off of 0.960). The specification reports that only the high-risk subgroup benefits from adjuvant chemotherapy, whereas the T-stage-based standard of care does not discriminate responders, and that the markers are not predictive in the adenocarcinoma subtype.

Caveats / uncertainty

• CAFC 2026 dockets: I could not locate any Federal Circuit docket, appeal, IPR, or district-court litigation involving US 11,402,381. If the user has a specific docket number, it could be checked directly; my searches returned nothing linking this patent to 2026 appellate activity. Treat "no 2026 CAFC activity found" as a negative search result, not proof of absence.
• Full claim text: The authoritative text provided truncates claim 1 at step (f) ("treating t…") and does not include the verbatim text of claims 2–8. My descriptions of the dependent claims come from secondary sources (Justia) and should be re-verified against the granted claim set.
• Legal status ("Active," expiration 2036-12-22) is Google Patents' automated statement and explicitly flagged as an assumption, not a legal conclusion.

I'll search for litigation involving US patent 11402381 across patent litigation databases and court records.

Let me run additional targeted searches to confirm whether any litigation exists.

Let me do a couple more targeted checks against litigation databases and PTAB to be thorough.

Bottom Line

No litigation involving U.S. Patent No. 11,402,381 is known. I searched patent‑litigation sources (Unified Patents‑style litigation coverage via web search, CourtListener/RECAP, Docket Alarm/UniCourt snippets, and general docket searches) using the exact identifier 11402381 and no case — district court, ITC, or Federal Circuit — was found that asserts this patent.

What the Patent Is (for identification)

• Patent: US 11,402,381 B2 — "Methods of identifying an individual to be treated by chemotherapy based on cytokeratin‑19 fragments (CYFRA 21‑1) and carcinoembryonic antigen (CEA) marker molecules and related uses"
• Granted: August 2, 2022
• Current assignee: Roche Diagnostics Operations, Inc.
• Application: US 15/952,731; priority date October 23, 2015
• Subject matter: in‑vitro diagnostic/method claims (measuring CYFRA 21‑1 and CEA via specific binding agents, computing a weighted combined value, comparing to a cut‑off, to identify SCC‑NSCLC stage I/II patients for chemotherapy)

Search Results — Distinguishing Look‑Alikes (Important)

Several results contained similar‑looking but different identifiers that should not be confused with this patent:

1. IPR2024‑01180 and IPR2024‑01181 — These are inter partes review petition numbers, not patent numbers. They concern U.S. Patent Nos. 9,198,076 and 10,749,700, asserted by Headwater Research LLC v. Motorola Mobility LLC and Lenovo (United States) Inc., No. 4:23‑cv‑04496‑JST (N.D. Cal.). The coincidence of "01181" with "11402381" is superficial; neither IPR involves US 11,402,381.
2. Roche Diagnostics v. Meso Scale Diagnostics, LLC, No. 1:17‑cv‑00189 (D. Del. 2020) — Unrelated; concerns electrochemiluminescence licensing, not this patent.
3. Roche Diagnostics v. LifeScan (D. Del., filed 2007) and Roche Diagnostics v. Polymer Technology Systems (S.D. Ind., filed 2003) — Older Roche patent suits, unrelated to this patent.
4. Symbology Innovations, LLC v. Roche Diagnostics GMBH Ltd., No. 2:23‑cv‑00630 (E.D. Tex., filed Dec. 27, 2023) — Roche is a defendant here; this is unrelated to US 11,402,381.

None of the above involves the '381 patent. Note that US patent numbering has not yet reached the 11,402,xxx range in a way that collides with the IPR number 2024‑01181; the similarity is coincidental.

Confidence and Caveats

• High confidence that no reported/known litigation exists, based on the searches performed. The patent is a diagnostic method patent granted in 2022, and method‑of‑diagnosis claims of this type have faced §101 eligibility challenges (cf. Roche Molecular Systems v. Cepheid, Fed. Cir. 2018) — which may partly explain why it does not appear to have been asserted.
• I cannot rule out an unreported, sealed, or very recently filed complaint that is not yet indexed in the free databases searched. There is also no indication of any PTAB proceeding (IPR/PGR) challenging the '381 patent.
• No outcome, jurisdiction, case number, or filing date can be provided for a case that does not exist. I have not fabricated any.

If you need certainty, the authoritative next steps would be direct queries to PACER, the Unified Patents litigation portal, Docket Alarm, and USPTO PTAB (Patent Trial and Appeal Board) using the exact number 11,402,381 — I am flagging these as the recommended verification channels rather than asserting results I could not retrieve.

Proceedings overview

Total AIA trial proceedings on file for US 11,402,381: 0. The structured "PTAB proceedings on file" block (USPTO Open Data Portal, most recent ingest) returns no IPR, PGR, or CBM petitions against this patent, and targeted web searching surfaced no PTAB docket numbers, institution decisions, or Final Written Decisions referencing US 11,402,381 (11,402,381). Bottom line for a defendant: the patent is untested at the PTAB — there is no IPR estoppel, no invalidated claim, and no FWD to lean on, but also no petitioner-side roadmap. Any defense is currently a from-scratch invalidity case, not a "the claims are already dead" case.

Caveat on sourcing: the absence of proceedings is drawn from the canonical ODP list plus negative web-search results. It is not the same as a guarantee. Confirm directly in PTAB E2E (https://ptab.uspto.gov) and on the patent's Google Patents "Litigation"/"Trials" tab before relying on it in a brief.

Proceedings

None to report. No proceeding number exists to quote, and I will not invent one. For completeness, the fields you asked me to populate would be empty across the board:

• No petitioner/applicant identified in any AIA trial against this patent.
• No institution decision, FWD, settlement/termination, or Federal Circuit appeal on this patent.
• No PTAB final written decision, and therefore no canceled, sustained, or untested-at-PTAB claims to report. All claims stand as issued and unadjudicated.

Strategic summary

Claim status. None of the claims of US 11,402,381 have been canceled or narrowed by the PTAB — the patent issued 2022-08-02 and remains Active (Google Patents legal status; adjusted expiration 2036-12-22). The patent carries 8 claims (per the claims listing), with claim 1 as the method claim directed to measuring CYFRA 21‑1 and CEA with specific binding agents, forming a weighted combined value, comparing to a SCC‑NSCLC stage I/II reference-population cut-off, identifying the patient as to be treated by chemotherapy following resection, and the final step of treating. All 8 claims are UNTESTED in an AIA trial. Because all claims are untested, there is no PTAB-based narrowing to exploit; invalidity must be proven in district court or in a first-filed IPR.

Estoppel landscape. With zero proceedings, § 315(e)(2) estoppel does not attach to anyone. If you file an IPR, your own estoppel will be created by that filing, not by a predecessor — so you would be the first mover and would own the consequences. There is no earlier petitioner whose prior-art grounds were already litigated, meaning the full universe of prior art (including the references cited on the face of the patent, e.g. Muley 2008/2010/2013 TMI work, Hanagiri 2011, Blankenberg 2008, Kulpa 2002, Moro 1995) remains available absent your own filing. Note the § 315(b) one-year clock runs from service of an infringement complaint on you — if you have been served, that clock, not the PTAB docket, is your binding deadline.

Pattern signals. No repeat-petitioner pattern (no petitioner at all). No indication of a defensive aggregator such as Unified Patents in the chain. The patent owner (Roche Diagnostics Operations, Inc., with Roche Diagnostics GmbH / Roche Diagnostics International AG as upstream assignors) has not had to defend claims at the Board, so there is no track record of aggressive PTAB appellate practice to read. This is a commercially held, relatively young diagnostic-method patent (priority 2015-10-23) that appears never to have drawn a validity challenge in an AIA forum.

Recommended next steps

1. Verify the null result before relying on it. Pull the patent in PTAB E2E (https://ptab.uspto.gov/patent/applicationsearch or the Trials search) and check the Google Patents Litigation / Trials panels for US 11,402,381. Because proceedings can be filed and not yet appear in an ODP ingest, a fresh E2E check is the control on the "no activity" conclusion.
2. If you are a defendant already served, treat the § 315(b) one-year window as your governing deadline; there is no existing IPR to piggyback on, so a petition would be a standalone first challenge. Budget for a full IPR (institution decision ~6 months from filing; statutory FWD deadline 12 months from institution under § 316(a)(11)).
3. If no proceedings exist and you are evaluating risk, the absence is itself a signal in the opposite direction from a "hardened" patent: well-asserted patents typically attract IPRs within 1–3 years of assertion. A patent with zero PTAB history is more consistent with (a) it has not been widely asserted, or (b) assertions have settled before petitioning — either way, the validity record is blank and untested.
4. Build the invalidity theory around the printed prior art already of record. The patent's own IDS cites the closest prognostic work (Muley TMI publications, Hanagiri, Blankenberg, Kulpa, Moro), suggesting § 103 combinations of those references with routine immunoassay knowledge are the natural first-line grounds — again, none of these have been tested or estopped.

Do not rely on any proceeding number for this patent — none exists in the sources checked. If a demand letter or complaint cites a "PTAB outcome" for US 11,402,381, that citation should be treated as unverified until you confirm it in PTAB E2E.

I'll research the assignment chain for US 11,402,381 before writing up the analysis.

Let me dig deeper into the specific assignment records and the correspondent of record.

Note on sources and a material limitation

The USPTO Assignment Center / Patent Assignment Search interface is a JavaScript application that I could not query directly in this session, and the public mirrors (Google Patents "Legal Events," FreePatentsOnline, Justia, Patent Leaderboard) expose the assignment events and dates but not the reel/frame numbers or the correspondent of record. I am therefore reporting the chain from the Google Patents legal-events record (which mirrors the USPTO assignment data) and explicitly flagging every field I could not verify rather than inventing reel/frame or correspondent values. Fabricating reel/frame numbers would be worse than a documented gap.

Inventors

The employer attributions are inferred from how the inventors were grouped on the recorded assignment instruments (see timeline), not from independent employment records. Two structural patterns are worth noting:

• Two-institution inventorship. Six inventors sit with Roche entities; two (Herth, Muley) sit with Thoraxklinik Heidelberg gGmbH, a German university chest clinic. This corresponds to the clinical/cohort-data contribution behind Example 2 (the 88-patient SCC-NSCLC series), so the Thoraxklinik assignment was a necessary link in the chain to consolidate title in Roche.
• No mass-departure signal. I could not verify "all inventors departed the original assignee within 12 months" — but the observable counter-signal is that Roche itself filed the continuation (Apr 13, 2018) and later obtained the patent, so there was no orphaned application to sell. Dayyani's later move to academia is a single-inventor move, not a portfolio fire-sale indicator.

Original assignee

Roche Diagnostics Operations, Inc. (9115 Hague Road, Indianapolis, IN 46250) is the assignee on the issued patent, having acquired the U.S. rights from Roche Diagnostics GmbH and Roche Diagnostics International AG.

• Product embodiment: Yes. Roche Diagnostics markets the Elecsys® CYFRA 21-1 and Elecsys® CEA electrochemiluminescence immunoassays — the specification names these exact products (CYFRA 21-1 product no. 11820966160; CEA material no. 11731629160) as the preferred measurement tools. Roche is thus an operating IVD company that commercializes assays within the claim scope.
• Primary line of business: In-vitro diagnostics (immunoassay analyzers and reagent kits); Roche Diagnostics Operations is a U.S. operating/holding subsidiary of F. Hoffmann-La Roche AG (Basel, CH).
• Current status: Operating, no bankruptcy, no dissolution. Roche Diagnostics is the global #1 IVD franchise; this patent sits inside a large, live corporate portfolio.

Assignment timeline

All recorded U.S. assignment events surfaced for US 11,402,381 share a recording date of 2018-05-09, i.e. they were filed as a single batch shortly after the Apr 13, 2018 U.S. filing. Conveyance type is Assignment for all six. Reel/frame and correspondent were not retrievable in this session (see limitation note above).

• 2018-05-09 (recorded) — Reel/Frame not verified (not retrievable in this session)

  • Conveyance: Assignment
  • Assignor: Felix Herth; Thomas Muley
  • Assignee: Thoraxklinik Heidelberg gGmbH
  • Correspondent: not retrievable — cannot confirm whether it recurs elsewhere without the reel/frame record
  • Context: Tochter/institutional consolidation — clinic inventors assign to their own institutional employer, the first link needed before Roche can take clean title.

• 2018-05-09 (recorded) — Reel/Frame not verified

  • Conveyance: Assignment
  • Assignor: Birgit Wehnl; Ying He; Julia Riedlinger; Vinzent Rolny; Friedemann Krause
  • Assignee: Roche Diagnostics GmbH
  • Correspondent: not retrievable
  • Context: Internal corporate — Roche-employed inventors assign to their operating employer (Roche Diagnostics GmbH, Mannheim).

• 2018-05-09 (recorded) — Reel/Frame not verified

  • Conveyance: Assignment
  • Assignor: Thoraxklinik Heidelberg gGmbH
  • Assignee: Roche Diagnostics GmbH
  • Correspondent: not retrievable
  • Context: Internal consolidation — clinic's rights pass to the Roche operating entity, completing the institutional half of the chain.

• 2018-05-09 (recorded) — Reel/Frame not verified

  • Conveyance: Assignment
  • Assignor: Farshid Dayyani; Achim Escherich
  • Assignee: Roche Diagnostics International AG (Rotkreuz, CH)
  • Correspondent: not retrievable
  • Context: Internal corporate — the Swiss-entity inventors assign to their Roche employer.

• 2018-05-09 (recorded) — Reel/Frame not verified

  • Conveyance: Assignment
  • Assignor: Roche Diagnostics GmbH
  • Assignee: Roche Diagnostics Operations, Inc.
  • Correspondent: not retrievable
  • Context: Internal corporate — German operating entity transfers U.S. rights to the U.S. group company.

• 2018-05-09 (recorded) — Reel/Frame not verified

  • Conveyance: Assignment
  • Assignor: Roche Diagnostics International AG
  • Assignee: Roche Diagnostics Operations, Inc.
  • Correspondent: not retrievable
  • Context: Internal corporate — Swiss entity transfers U.S. rights to the U.S. group company, completing consolidation of full U.S. title in Roche Diagnostics Operations, Inc.

No post-issuance assignment is recorded. The last link is the 2018 consolidation into Roche Diagnostics Operations, Inc.; the patent issued (Aug 2, 2022) in the name of that same entity. That is itself a finding: no transfer to any third party, licensing vehicle, or collateral agent occurred after grant.

Timeline diagram

timeline
    title Ownership of US 11402381
    2015 : EP priority application filed
    2016 : PCT application filed
    2018 : US continuation filed
         : Inventors assign to Roche entities
         : Thoraxklinik assigns to Roche GmbH
         : Roche GmbH and RD International assign to RD Operations
    2022 : Patent granted to Roche Diagnostics Operations

NPE / troll-pattern signals

1. Shell-entity transfer — not present. No assignee in the chain is a licensing-only or holding LLC. Every assignee is a named Roche operating entity (Roche Diagnostics GmbH, Roche Diagnostics International AG, Roche Diagnostics Operations, Inc.) or a German university clinic (Thoraxklinik Heidelberg gGmbH). No "IP / Licensing / Ventures" suffix; no registered-agent address; no single-purpose Delaware/Texas LLC appears in the record.

2. Known asserter in the chain — not present. No assignee matches any public NPE list (Acacia, Marathon, IV, IPNav, Wi-LAN/Conversant, Vringo, Pendrell, Round Rock, etc.). Roche Diagnostics Operations, Inc. is an operating IVD manufacturer, not an NPE.

3. Repeat correspondent across the chain — unclear / cannot be assessed. This is the one signal I most wanted to test, and it is precisely the field the retrievable sources omit. The correspondent of record for all six 2018-05-09 recordings is not visible in Google Patents legal events or the secondary mirrors. Because the batch shares one recording date and one assignee family, it is likely a single Roche in-house attorney or a single outside firm handled all six — but I have no reel/frame or correspondent text to cite, so I will not assert recurrence. Verification action: pull the six 2018-05-09 records in Assignment Center and compare the "Correspondent" field.

4. Cascading transfers — not present (as an NPE signal). There are multiple assignments, but they are all recorded on a single day (2018-05-09) as one coordinated title-cleanup batch, all flowing toward a single ultimate assignee (Roche Diagnostics Operations, Inc.). This is the classic multi-entity operating-company consolidation pattern (foreign + domestic + university collaborators → U.S. group company), not chained shell transfers. No chain of LLCs; no 24-month cascade.

5. Pre-litigation transfer — not present. No infringement suit naming this patent exists (confirmed in the prior litigation analysis, and re-checked here). With no suit, there is no pre-suit transfer within 6 months to assess.

6. Bankruptcy fire-sale — not present. Roche Diagnostics has no bankruptcy history; the chain shows no trustee, receiver, or §363 sale.

7. Privateering — not present. No transfer to a third-party NPE asserting on Roche's behalf. Title terminates at Roche's own operating subsidiary.

8. Defensive aggregator (anti-NPE) — not present. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. Roche retains direct ownership.

Verdict

Operating-company assertion. Of the menu options, this is the only substantive fit: the chain terminates at Roche Diagnostics Operations, Inc., which (a) ships products embodying the claims — the Elecsys® CYFRA 21-1 and Elecsys® CEA assays named in the specification — and (b) holds title directly with no intervening NPE. Caveat on the label: there is no recorded assertion — no district-court, ITC, or PTAB proceeding naming this patent was found — so the accurate description is "operating-company owned, currently non-asserting." I did not select any NPE verdict because zero of the eight signals are affirmatively present; the single most probative signal (repeat correspondent) is unclear solely because the correspondent field is not exposed by the retrievable data sources, not because of any evidence of an NPE pattern. Every recorded transfer is an internal Roche/title-consolidation assignment dated 2018-05-09, i.e. the opposite of a shell cascade.

Verification link: https://assignmentcenter.uspto.gov/ (search patent number 11402381) and https://assignment.uspto.gov/patent/index.html — retrieve the six 2018-05-09 records to capture the missing reel/frame and correspondent fields, which are the only gaps in this report.

Contradiction flag: None found against the previously generated sections. The prior summaries reported the same assignee (Roche Diagnostics Operations, Inc.), the same inventors (with the "Wehml/Wehnl" OCR discrepancy already noted), and the same absence of litigation; this ownership analysis is consistent with all of that.

I'll research the prior-art citations for US 11,402,381, starting with the patent's citation records.

Let me retrieve the full citation lists from the patent's Google Patents and FreePatentsOnline records.

I have the citation lists. Let me now pull details on the specific patent references cited.

Prior Art Assessment — US 11,402,381 B2

Scope and a necessary framing caveat

You asked me to identify "the most relevant prior art for US 11,402,381" and to analyze "each patent citation for 11402381" for § 102 anticipation. Two framing points must be stated up front, because they materially affect how the § 102 column should be read:

1. These are references cited on the face of the patent. The list below is the "References Cited" front-page record of US 11,402,381 B2 (retrieved from the granted patent's citation record via Justia, FreePatentsOnline, and Google Patents). By definition, the claims issued over these references. Therefore none of them literally anticipates claim 1 — if one did, the patent would not have granted. Where I use the word "anticipates," I mean "could be argued to read on, and was probably cited against," not "invalidates." The honest § 102 answer is: no cited reference discloses every element of claim 1 arranged as claimed.

2. Identifier rule observed. I have interpreted every alphanumeric identifier literally and have not corrected any number. Where results were contradictory or sparse, I say so rather than filling gaps.

I could retrieve the citation inventory from three aggregators (Justia, FPO, Google Patents) but was unable, within this session, to independently pull the hand-signed USPTO front page (PatentCenter/patentimages) to reconcile the exact inventory character-for-character. Treat the inventory below as high-confidence but not court-certified.

A. Patent documents cited (the "patent citations" you asked about)

B. Non-patent literature cited (often the more probative art here)

The examiner/applicant cited a dense set of journal literature; for a diagnostic-marker case, this is where the real § 102/§ 103 fight lives.

C. § 102 bottom line

No cited reference, alone, anticipates any claim of US 11,402,381.

Claim 1 is a specific combination requiring all of:

• (a) a human diagnosed with SCC-NSCLC stage I or II;
• (b) measuring CYFRA 21-1 protein and CEA protein with specific binding agents for each;
• (c) a combined value by weighted calculation of the two;
• (d) comparison to a cut-off of combined values established in a reference population also diagnosed with SCC-NSCLC stage I or II;
• (e) identification that the patient is to be treated with chemotherapy following resection when the combined value is increased; and
• (f) the treatment step.

Mapping that against the citations:

• CYFRA 21-1 + CEA combined, prognostic → Muley 2010/2003, Blankenberg 2008, Hanagiri 2011, Moro 1995, Ebert 1994. But these are prognostic, not predictive of chemotherapy benefit, and none discloses the weighted value with a cut-off established in an SCC-stage-I/II reference population.
• Adjuvant-chemo selection problem → Muley 2013 abstract, WO 2012/160177, Morgensztern 2015. But Muley 2013 is an abstract without the claimed weighting/cut-off structure, and WO 2012/160177 uses gene expression, not CYFRA 21-1/CEA protein.
• Assay/binding-agent mechanics → Bodenmüller 1994, Mantovani 1994, Paul 1993, Bendig 1995, US 2005/0244973, US 2006/0155200, US 2015/0038365, US 5,283,173, US 5,468,614. These bear on claims 1(b)/8, i.e., § 112 support and obviousness, never anticipation.
• Multi-marker panel with algorithm → US 2010/0179067 (Patz; CART on a different panel, CEA but no CYFRA 21-1, detection not treatment selection).

The realistic invalidity theory is § 103, not § 102 — e.g., Muley 2010 (TMI) or Blankenberg 2008 in view of Kulpa 2002 (SCC) or Edge 2010 (staging) — and that theory was overcome during prosecution, since the claims issued in 2022. The patent's own specification anticipates and pre-empts the TMI line of attack by expressly distinguishing itself from TMI-only methods (spec: "the innovative finding… is the focus of SCC histology, the addition of T stage and applicability to both stage I and II, all of which are different from the methods described above and based on the tumor marker index, only").

D. Most-relevant-prior-art ranking (my assessment)

1. Muley T. et al., Tumor Biology 2010;31(Suppl.1):S87 (TMI abstract) — closest on substance; defeated by the SCC-stage-I/II-specific weighted cut-off and predictive (treatment-selection) steps.
2. Blankenberg F. et al., Tumor Biology 2008;29:272-277 — preoperative CYFRA 21-1 + CEA score in stage I NSCLC; same gaps.
3. WO 2012/160177 (Novartis) — closest on the clinical problem (selecting early-stage NSCLC patients for adjuvant therapy), but the wrong analyte class (gene expression).
4. US 2010/0179067 (Patz) — closest U.S. patent document on a serum-marker + algorithm panel, but omits CYFRA 21-1 and is detection-oriented.
5. Muley Th. et al., 2013 Hamburg Symposium Abstract 50 — directly frames the adjuvant-chemo-qualification question.
6. Kulpa J. et al., Clin. Chem. 2002;48(11):1931-1937 — supplies the SCC-histology link (useful as a § 103 secondary reference).

E. Caveats and verification steps I could not complete

• The ISR categories are the key missing item. The Nov 22, 2016 ISR for PCT/EP2016/075313 is cited on the patent's face; its X/Y/A designations would tell you exactly which reference the examiner deemed "particularly relevant." I could not retrieve those category assignments in this session. If you need the authoritative § 102/§ 103 slope, obtain that ISR (and any EPO counterpart search report for EP 15 003 039.3).
• Unverified disclosures. I could not separately verify the subject matter of US 2005/0244973, US 2006/0155200, US 2012/0076725, US 2015/0038365, EP 2071337, EP 2908133, or JP 5221825 within this session. My characterizations of those are marked accordingly and should be re-checked against each document's text before being used in any opinion of validity.
• WO 2011/140234 identification is provisional. The title I surfaced ("Methods for predicting sensitivity to treatment with a targeted tyrosine kinase inhibitor") comes from a Google Patents page that lists US 11,402,381 B2 among citing documents; I could not confirm within this session that this is the same document as the front-page citation. Verify before relying on it.
• No front-page reconciliation. I did not reconcile the aggregator lists (Justia/FPO/Google Patents) against the signed USPTO front page. The three aggregators agreed on the U.S. and foreign patent-document lists, which raises confidence, but the authoritative record is the granted patent itself (or PatentCenter).
• Post-priority, non-citable art. For completeness: the inventors' own Muley T. et al., Lung Cancer 2018;120:46-53, and WO 2017/068097 (referenced in that paper as the pending patent containing the data), disclose substantively the same subject matter but post-date the Oct 23, 2015 priority and therefore are not § 102 prior art against this patent. Do not treat them as such.

One explicit contradiction I am flagging between sources I encountered: Justia's citation list renders the "Other references" first entry as "Mantovani (European Journal of Cancer, vol. 30A, No. 3, p. 363-369, 1994)" with no title, and the FPO page shows the same terse form. I have reported it literally; I could not confirm its title or contents, so I have not assigned it a § 102 theory.

I'll research the prior art references cited on the face of US 11,402,381 to ground an obviousness analysis.

Obviousness Analysis — US 11,402,381 B2 under 35 U.S.C. § 103

Analyst note on sources. This analysis is built on the prior art that appears on the face of US 11,402,381 (the "Referenced Cited" list at Justia and the NPL citations in the Google Patents / patentimages record, which I re-fetched and verified) and on the published abstracts/full texts of those references retrieved live. Where a statement rests on a specific reference, I cite it. Where I am inferring rather than reading, I say so. I flag one point at which a reference on the face of the patent cuts against my analysis (Blankenburg 2008), and I do not hide it.

1. Effective filing date and the prior-art window

• Earliest priority: 2015‑10‑23 (EP 15003039.3); PCT filed 2016‑10‑21 (PCT/EP2016/075313); U.S. filed 2018‑04‑13.
• Assuming the priority claim is valid, the critical date for § 102/§ 103 purposes is on or about 2015‑10‑23.
• Every reference I rely on below published before 2015‑10‑23, and most are on the face of the patent itself — which materially strengthens a § 103 attack because they are the applicant's own cited art (a finding of "applicant appreciated the references" undercuts a "non-analogous art" or "teaching away" defense).

One family member is not prior art. WO 2017/068097 (published 2017‑04‑27) is the PCT publication of this application. The inventors' own 2018 paper (Muley, Rolny, He, Wehnl, Escherich, Dayyani et al., Lung Cancer, "…as a potential predictor of benefit from adjuvant chemotherapy in early stage squamous cell carcinoma of the lung (SCC)", S0169500218303040) post-dates priority, and it expressly states "Patent WO2017/068097 containing data included in this manuscript is currently pending." Neither the WO publication nor the 2018 paper may be used as § 102/§ 103 art against the '381 claims. Do not cite them as the primary reference; cite them only as evidence of what the inventors themselves viewed as the incremental contribution.

2. The claim to be attacked (literal text, claim 1)

Claim 1 is a method of identifying that a human diagnosed with SCC‑NSCLC stage I or II is to be treated by chemotherapy following resection, comprising:

Dependent claims, all of which are narrower but individually weak (see § 6), add: pre-therapy/pre-surgery sample timing (2); sample 21 days-to-immediately-before resection (3); reference population excludes adenocarcinoma (4); serum/plasma/whole blood (5); cut‑off = median (6); specific chemo drugs (7); binding agents are antibodies (8).

Critical scope observation: the specification's stated "innovative finding" includes "the addition of T stage," but claim 1 does not recite T stage. The claim is therefore broader than the asserted distinguishing feature the patentee will point to. That matters to § 103 — the T‑stage feature cannot be relied on to save a claim that omits it.

3. Level of ordinary skill (POSA)

A POSA here is a team or individual with an M.D. or Ph.D. in oncology/pulmonology or a related biomedical field, plus several years' experience with serum tumor‑marker immunoassays (CYFRA 21‑1, CEA) and clinical survival analysis (Kaplan–Meier, Cox proportional hazards). The patent itself treats the statistical methods (Cox regression, Kaplan–Meier) as conventional — it cites Cox 1972, Breslow 1975, Kaplan & Meier 1958, and Rich 2010 as background. That self‑characterization is useful: the patentee cannot later claim that applying Cox regression to derive weights and a cut‑off was beyond ordinary skill.

4. The prior art of record (verified)

Core prognostic/treatment-selection art (the heart of any § 103 case):

1. Muley, Dienemann & Ebert (2004), Anticancer Res 24:1953‑1956 — "CYFRA 21‑1 and CEA are independent prognostic factors in 153 operated stage I NSCLC patients." Both markers independently prognostic; combining increases detection of worse‑outcome patients; divides patients into three risk groups by TMI; concludes elevated markers identify patients "who may benefit from more aggressive treatment approaches," while the lowest‑risk group "probably does not need further treatment." (URL: read.qxmd.com/read/15274383)
2. Muley, Fetz, Dienemann, Hoffmann, Herth, Meister & Ebert (2008), Lung Cancer 60(3):408‑415 — "Tumor volume and tumor marker index based on CYFRA 21‑1 and CEA are strong prognostic factors in operated early stage NSCLC." 261 completely resected stages I and II NSCLC patients; TMI = geometric mean of normalized CYFRA 21‑1 and CEA values; a cut‑off of TMI ≤ 0.54 vs > 0.54 separates risk groups; explicit conclusion: "elevated levels of TMI and V(path) have a strong negative prognostic impact on survival in operated early stage of NSCLC. These patients might be considered for adjuvant chemotherapy." (URL: pubmed.ncbi.nlm.nih.gov/18083270)
3. Muley et al. (2010), Tumor Biology 31(Suppl 1):S87 — "Tumor Marker Index (TMI) Based on CYFRA 21‑1 and CEA Identifies P‑Stage I NSCLC Patients with High Risk of Recurrence, Potential Candidates for Adjuvant Chemotherapy?" (title as listed on the face of the '381 patent)
4. Muley et al. (2013), Int. J. Cancer Res 23(6B), Abstract 50 — "Do Postoperatively Elevated CYFRA 21‑1 or CEA Values Qualify for Adjuvant Chemotherapy in Completely Resected Stage I NSCLC Patients?" (as listed on the face of the '381 patent)
5. Tomita, Shimizu, Ayabe, Yonei & Onitsuka (2010), Anticancer Res 30(7):3099‑3102 — "Prognostic significance of tumour marker index based on preoperative CEA and CYFRA 21‑1 in non‑small cell lung cancer." Uses preoperative CEA + CYFRA 21‑1 combined as TMI with a cut‑off of 1.0; concludes TMI is an independent prognostic determinant and states: "it can be hypothesized that adjuvant chemotherapies may be useful for patients with high TMI… patients with high TMI could represent a reasonable study population for an adjuvant therapy trial," while noting "there is no evidence that adjuvant therapy would be useful in patients with a TMI greater than 1.0, but this may be a question for future studies." (URL: ar.iiarjournals.org/content/anticanres/30/7/3099.full.pdf)
6. Hanagiri et al. (2011), Lung Cancer 74(1):112‑117 — "Preoperative CYFRA 21‑1 and CEA as prognostic factors in patients with stage I non‑small cell lung cancer." Preoperative serum drawn within 30 days prior to surgery in 341 completely resected stage I NSCLC patients. Reports histology‑specific positive rates: CYFRA 21‑1 76.8% in squamous vs. 33.3% in adenocarcinoma; CEA 26.8% in squamous vs. 23.8% in adenocarcinoma. Concludes a high preoperative CYFRA 21‑1 is an independent stage‑I prognostic factor and — critically — "Further clinical studies will be necessary to evaluate the efficacy of adjuvant therapy for the patients selected according to this criterion." (URL: pubmed.ncbi.nlm.nih.gov/21397974)
7. Kulpa, Wójcik, Reinfuss & Kołodziejski (2002), Clin. Chem. 48(11):1931‑1937 — evaluates CEA, SCC antigen, CYFRA 21‑1 and NSE specifically "in squamous cell lung cancer patients." Establishes the squamous‑selective behavior of CYFRA 21‑1 in the exact histology claim 1 targets. (on the face of the '381 patent)
8. Blankenburg, Hatz, Nagel, Ankerst, Reinmiedl, Gruber (2008), Tumor Biol. 29(4):272‑277 — "Preoperative CYFRA 21‑1 and CEA as prognostic factors in patients with stage I NSCLC: external validation of a prognostic score." (See the caveat in § 7 — a secondary source reports this validation as negative for TMI.)
9. Muley, Dienemann & Ebert (2003), Anticancer Res 23:4085‑4095 — "Increased CYFRA 21‑1 and CEA levels are negative predictors of outcome in p‑stage I NSCLC." (on the face of the '381 patent)
10. Moro et al. (1995), Lung Cancer 13:169‑176 — "CEA, CYFRA21‑1 and SCC in non‑small cell lung cancer." Earliest of the combined‑marker line.
11. Wieskopf et al. (1995), Chest 108:163‑169 and van der Gaast et al. (1994), Br. J. Cancer 69:525‑528 — CYFRA 21‑1 sensitivity/specificity and prognostic role in NSCLC.

Assay/measurement art (for the "specific binding agent" element):

1. Bodenmüller et al. (1994), Int. J. Biol. Markers 9(2):75‑81 — development of the two‑monoclonal‑antibody CYFRA 21‑1 assay, including the antibodies KS 19.1 and BM 19.21.
2. Nisselbaum et al. (1988), Clin. Chem. 34:761‑764 — comparison of Roche RIA/EIA, Hybritech and Abbott EIA methods for measuring CEA. (cited within Hanagiri's reference list).
3. Roche Elecsys® CYFRA 21‑1 (product 11820966160) and Elecsys® CEA (product 11731629160) — commercially available ECLIA sandwiches using specific monoclonal antibodies, as the specification itself concedes.

Treatment art (for the "treating" step and claim 7):

1. Winton et al. (2005), N. Engl. J. Med. 352:2589‑2597 (JBR.10) and Douillard et al. (2006), Lancet Oncol. 7:719‑727 (ANITA) — adjuvant cisplatin + vinorelbine after resection in stage IB–IIIA NSCLC. These and the LACE pooled analysis define the routine adjuvant regimens recited in claim 7.
2. Morgensztern et al. (2015), J. Clin. Oncol. 33(15 Suppl): Abstr 7520 — "Adjuvant chemotherapy for patients with T2N0M0 NSCLC" — the patent's own cited statement of the clinical‑need problem (T‑stage‑based selection is imperfect).

5. Obviousness combinations

Combination A (primary) — Muley 2008 + Tomita 2010 + Muley 2010/2013 abstracts + Hanagiri 2011

What is taught. Muley 2008 discloses, in the same clinical population the claim recites (completely resected stage I and II NSCLC), a combined value built from CYFRA 21‑1 and CEA (the TMI), compared against a cut‑off (0.54) established in the study cohort, to stratify patients into high‑ and intermediate‑risk groups who "might be considered for adjuvant chemotherapy." Tomita 2010 independently replicates the TMI approach with preoperative serum and a 1.0 cut‑off and expressly hypothesizes that adjuvant chemo may be useful for high‑TMI patients, calling them "a reasonable study population for an adjuvant therapy trial." Muley's 2010 and 2013 abstracts are, on their face, titled around whether TMI/high markers designate "potential candidates for adjuvant chemotherapy." Hanagiri 2011 supplies the preoperative‑sampling protocol (within 30 days before surgery) and the explicit call to test "the efficacy of adjuvant therapy for the patients selected according to this criterion."

Why a POSA would combine them. These are not disparate fields; they are the same research program reported serially (Muley 2004 → 2008 → 2010 → 2013) plus independent confirmations (Tomita 2010; Hanagiri 2011). The references are analogous art, address the identical problem (which resected early‑stage NSCLC patients need adjuvant therapy), and the later references expressly build on and cite the earlier ones. The Supreme Court's KSR rationale — "if a technique has been used to improve one device, and a person of ordinary skill in the art would recognize that it would improve similar devices in the same way, using the technique is obvious" — applies almost mechanically.

The § 103 gap the patentee will point to. Combination A yields a prognostic (risk) selection tool for all NSCLC; the claim recites a predictive selection for SCC‑NSCLC specifically. Two sub‑arguments bridge that gap:

1. SCC focus is suggested by Kulpa 2002 and Hanagiri 2011. CYFRA 21‑1's heightened sensitivity in squamous histology was known (Kulpa 2002 studied squamous‑cell patients only; Hanagiri 2011 reported a 76.8% CYFRA 21‑1 positive rate in squamous vs. 33.3% in adenocarcinoma). Routine optimization of a known biomarker test to the histology in which the marker performs best is precisely the "obvious to try" territory of KSR and In re Aller.
2. The predictive/prognostic distinction collapses here. Tomita 2010 already hypothesized the predictive use ("adjuvant chemotherapies may be useful for patients with high TMI"), and Hanagiri 2011 called for exactly the study. When the prior art expressly identifies the next experiment ("this may be a question for future studies"), running that experiment is obvious unless the art taught away or the results were unexpected — and the results (high‑risk SCC benefit) merely confirmed the hypothesis; they did not reverse it.

Combination B — A + assay art (Bodenmüller 1994; Nisselbaum 1988; Elecsys)

Step (b) requires specific binding agents for each marker. This was entirely conventional by 2015: Bodenmüller 1994 discloses the two monoclonal antibodies (KS 19.1, BM 19.21) that define the CYFRA 21‑1 epitopes, and CEA sandwich immunoassays were long established (Nisselbaum 1988). The patent specification itself concedes that "monoclonal antibodies to the markers (Cyfra21‑1 and CEA) [are] used in quantitative … immunoassay" and that Elecsys® kits are the preferred, off‑the‑shelf implementation. Combining a known clinical index with a known, commercially available assay for each of its two components is the paradigm of a predictable combination; the assay does not change the principle of operation of the index and vice versa.

Motivation: the commercial availability and routine use of the Elecsys® CYFRA 21‑1 and CEA assays (and the ARCHITECT/ADVIA equivalents recited in the patent) supplies both the motivation and the reasonable expectation of success. There is no new chemistry in the claims.

Combination C — A + B + adjuvant‑therapy art (Winton 2005/ANITA) for step (f) and claim 7

Once a patient is identified as "to be treated by chemotherapy following resection," the actual treatment step uses drugs and regimens that were standard of care in resected NSCLC (cisplatin/carboplatin ± vinorelbine, docetaxel, etoposide, pemetrexed). Winton 2005 and Douillard 2006 supply these. Claim 7's drug list is a verbatim recitation of the then‑current adjuvant formulary; selecting a known drug to treat a patient identified by a known index is obvious. The specification's own text ("Presently preferred adjuvant chemotherapies include cisplatin or carboplatin, often in combination vinorelbine, docetaxel, etoposid or permetrexed") is essentially an admission of conventionality.

Combination D — for claim 6 (median cut‑off) and claim 4 (SCC‑only reference population)

• Claim 6 (cut‑off = median of reference population): Muley 2008 and Tomita 2010 both derive population‑based cut‑offs; the use of a median is a routine statistical choice with no unexpected effect. The patent gives no data showing a median is superior to the 0.54 or 1.0 cut‑offs already known.
• Claim 4 (reference population excludes adenocarcinoma): this is a direct consequence of the SCC‑restriction in claim 1, supported by Kulpa 2002 and Hanagiri 2011's histology data.

6. Claim‑by‑claim § 103 map

7. Required caveats, teaching‑away evidence, and the patentee's counterarguments

Because a § 103 opinion that ignores the other side is not useful, here is the opposing case:

1. Teaching‑away / conflicting evidence — Blankenburg 2008. A secondary review (Semantic Scholar PDF, "Tumor marker CEA and CYFRA21‑1 has been reported as poor prognostic factors in NSCLC," Fig. and text) states: "In contrast, Blankenburg et al. reported that TMI was not associated with a worse outcome in stage I NSCLC." A single negative external‑validation study is not generally "teaching away" (it does not criticize or discourage the combination; it reports a failure to replicate), and In re Gurley / DePuy Spine require the reference to discourage the solution, not merely fail to confirm it. But the patentee will argue that Muley 2008, Tomita 2010 and Blankenburg 2008 together created an unresolved controversy, so the claimed result was not predictable. Expect this to be a live dispute.

2. The "unexpected results" argument. The patentee's strongest non‑obviousness position is that the marker combination is predictive in SCC‑NSCLC but not in adenocarcinoma (spec FIGS. 2B vs. 3B; the inventors' 2018 study). If the patentee can show (a) a difference in kind, not degree, and (b) that the art taught the opposite or provided no expectation of a histology‑dependent split, it could rebut a prima facie case under In re Soni/WBIP. But note: the art already distinguished histology (Hanagiri 2011's 76.8% vs. 33.3% CYFRA positivity by histology; Kulpa 2002 studied squamous patients specifically). An examiner or district court is likely to treat "the SCC subgroup responds" as a predictable refinement, not a surprise.

3. Small‑sample / statistical fragility. The specification's own working example rests on 88 patients with 21 treated. Expect the patentee to argue the result was not reasonably predictable from small, single‑center, retrospective data. That cuts both ways — small n also weakens the patent's own showing of unexpected results.

4. Secondary considerations (unaddressed in the record). I found no evidence of commercial success, licensing, copying, industry praise, or long‑felt‑but‑unsolved need specifically tied to the '381 claims. The absence of a robust secondary‑considerations record makes the prima facie § 103 case harder to rebut. (I did not locate a licensing/royalty nexus; treat this as a negative search result, not proof of absence.)

5. § 101 overlay (context, not part of this task). Method‑of‑diagnosis claims of this shape have repeatedly run into Mayo/Alice at the Federal Circuit (cf. Roche Molecular Systems v. Cepheid, 2018, noted in the earlier litigation section of this file). The same claim‑drafting that leaves claim 1 without a T‑stage limitation — broadening it over the art — simultaneously weakens any § 101 eligibility argument. This is worth flagging as a strategic interaction, not as a § 103 ground.

8. Bottom line

• A prima facie § 103 case exists and is reasonably strong. The primary combination — Muley 2008 (TMI of CYFRA 21‑1 + CEA in resected stage I/II NSCLC, with a cut‑off, explicitly "considered for adjuvant chemotherapy") + Tomita 2010 (preoperative TMI, express hypothesis that high‑TMI patients are candidates for adjuvant chemo) + Hanagiri 2011 (preoperative CYFRA/CEA, histology data, express call to test adjuvant therapy in marker‑selected patients) — discloses every substantive step of claim 1 except the SCC‑specific restriction, and Kulpa 2002 + Hanagiri 2011 supply the motivation to apply the known CYFRA 21‑1‑based index in the squamous histology.
• The primary open issue is the "SCC‑only / not predictive in adenocarcinoma" limitation. That is the feature the patentee will defend. It is bridgeable with the histology‑specificity art (Kulpa 2002; Hanagiri 2011) under KSR's "obvious to try" and routine‑optimization principles, but it is not automatic, and the Blankenburg non‑replication controversy gives the patentee a genuine (if modest) rebuttal hook.
• Every dependent claim is weak. Claims 2–8 recite sampling timing, sample type, a median cut‑off, a list of known adjuvant drugs, and antibodies — each squarely met by the references above and each a routine design choice.
• Do not use the inventors' own WO 2017/068097 or the 2018 Lung Cancer paper as prior art. They are not, and citing them will undermine the credibility of the § 103 case.

Recommended next steps to firm this up: pull the full texts (not just abstracts) of Muley 2008, Tomita 2010, Hanagiri 2011, and Kulpa 2002 to quote their precise claim‑element disclosures verbatim; confirm the content (not just title) of the Muley 2010 and 2013 abstracts, which are currently known to me only by their titles as listed on the patent; and verify Blankenburg 2008's actual conclusion to determine whether it truly failed to validate TMI (which would materially affect the teaching‑away analysis).

Confidence: moderate‑to‑high on the reference disclosures (abstracts verified live); moderate on the ultimate § 103 conclusion, which turns on the SCC‑specificity limitation and the disputed Blankenburg replication.