US 12491085

I'll verify the patent data against live sources, including checking for any 2026 CAFC docket activity.

Let me check for any litigation or appeal activity involving this patent.

Verification note

I retrieved the authoritative full text of US 12,491,085 B2 from Google Patents (https://patents.google.com/patent/US12491085/en). I could not query USPTO PatentCenter or a CAFC docket system directly with the tools available here, so docket-level confirmation of litigation/appeal activity is limited (see "Litigation & 2026 docket check" below). Numeric identifiers are reproduced literally as they appear in the source.

Bibliographic data

Abstract (as issued)

An interbody implant system for use in the spine includes a base comprising two or more bone contacting surfaces, at least one recess in at least one of the two or more bone contacting surfaces, the recess configured for containing a tooth, a deployable tooth to provide fixation between the base and the anatomy of a subject, a break‑away bridge between the tooth and the base for providing a first relative position between the tooth and the base, and a locking mechanism for providing a second relative position between the tooth and the base.

Analyst flag: the issued abstract and the detailed description describe a mechanical interbody device (deployable/break‑away teeth, locking features 230/232), whereas all 15 issued claims are directed to computer‑implemented design methods. The written description's support for the claimed design workflow is largely a single paragraph invoking additive manufacturing for "patient‑specific implants" plus incorporation by reference of application Ser. No. 16/207,116 ("Systems and Methods for Multi‑Planar Orthopedic Alignment"). This specification/claim mismatch is a notable validity consideration (written-description and enablement exposure), though I am not opining on any legal conclusion.

Plain-language overview of the independent claims

Claim 1 — Computer-implemented method of designing a patient-specific implant system (steps):

1. Generate an adjustable virtual 3‑D model of at least a portion of the patient's spine.
2. Receive user‑inputted adjustments to that model to define one or more patient‑specific corrections to the spine.
3. Using the adjusted model that shows those corrections, map a "negative space" between a first vertebral body endplate and a second vertebral body endplate — where the negative space includes contouring that corresponds to the patient‑specific contouring of both endplates.
4. Design a patient‑specific implant system that will deliver the corrections, the system comprising:
   • a base shaped to match the mapped negative space, having a first bone‑contacting surface with a first patient‑specific shape matching the first endplate's contouring, and a second bone‑contacting surface with a second patient‑specific shape matching the second endplate's contouring; and
   • a deployable fixation element that provides fixation between the base and at least one of the endplates.

Claim 11 — Method of designing a patient‑specific interbody implant system (parallel, slightly broader/differently worded):

1. Non‑invasively determine a patient‑specific correction by adjusting a virtual 3‑D model of at least a portion of the spine.
2. Map the negative space between the two endplates (same contouring requirement as claim 1).
3. Design the patient‑specific interbody implant system based at least in part on the adjusted model, comprising a base matching the mapped negative space (two patient‑specific bone‑contacting surfaces) and a deployable fixation element for fixation to at least one endplate.

The practical distinction: claim 1 is framed as a computer‑implemented design method with explicit "receiving user‑inputted adjustments"; claim 11 recites "non‑invasively determining" the correction and adds an interbody‑specific (rather than generic implant‑system) framing.

Dependent claims (add scope for infringement analysis):

• 2 / 13: correction is a linear displacement of the spine.
• 3 / 14: correction is a rotation of the spine.
• 4 / 15: correction of a coronal, sagittal, and/or axial plane deformity.
• 5: generating the model by converting two‑dimensional image data into the 3‑D model.
• 6: deployable fixation element comprises a plurality of deployable teeth.
• 7: deployable fixation element is a screw.
• 8: base includes a channel extending at least partially therethrough, with the fixation element advancing through the channel.
• 9: a locking mechanism to lock the fixation element in the deployed position.
• 10 / 12: generating fabrication instructions / manufacturing the patient‑specific implant system.

Litigation & 2026 docket check

• No CAFC 2026 appeal naming US 12,491,085 was found in the search results. Because the patent issued only on 2025‑12‑09, an appeal reaching the Federal Circuit by April 2026 is unlikely, but I cannot state this affirmatively without docket access.
• The only 2026 enforcement activity surfaced for the assignee: Carlsmed, Inc. v. DeGen Medical, Inc., filed 2026‑09‑15 in the U.S. District Court for the District of South Carolina, Florence Division, asserting eight Carlsmed patents "relating to its implants, methods and systems," seeking injunctive relief and damages (GlobeNewswire release, 2026‑09‑15; https://www.odtmag.com/breaking-news/patent-infringement-lawsuit-filed-against-degen-medical/). The eight asserted patent numbers were not disclosed in the public announcement, so I cannot confirm whether US 12,491,085 is among them. Do not assume it is.
• Note a date conflict in the source materials: the system context date is 2026‑09‑17, while the query states "April 26, 2026." The DeGen complaint (2026‑09‑15) postdates the query's stated date.

Statement of uncertainty

• Bibliographic and claim data above are taken verbatim from the Google Patents full text of US 12,491,085 B2, which is my authoritative source here. Assignee, inventor, and legal‑status fields on Google Patents are expressly disclaimed by that source as unofficial.
• I have not verified the patent's claims through USPTO PatentCenter or the USPTO's Patent Public Search ("Patent Public Search" / PatFT) interface directly, nor reviewed the file history (e.g., any terminal disclaimers, PTA calculations, or claim amendments). The 2040‑06‑28 adjusted expiration is the source's assumption.
• I found no authoritative court record tying this specific patent number to any CAFC appeal or to the DeGen case.

I'll search litigation databases for this specific patent number.

Initial searches returned only unrelated patents ending in "085." Let me query the patent-litigation-specific sources.

Let me try the litigation-specific databases and quoted patent-number searches.

Litigation search: US 12,491,085 B2

Bottom line

No litigation naming US Patent No. 12,491,085 B2 was found. Across every search run, I could not identify a single district court case, ITC investigation, PTAB proceeding, or Federal Circuit appeal in which US 12,491,085 is identified as an asserted patent.

I want to be precise about the limits of that statement rather than overstate it: this reflects absence of evidence in the sources I could reach, not a certified negative from PACER. I could not query the Unified Patents litigation portal, PACER, CourtListener's RECAP docket index, or Docket Navigator directly — the search tool returned no page content for the Unified Patents portal query, and my session terminated before I could run the remaining targeted queries. See "What I could not verify" below.

Case list

No rows can be populated. I have deliberately not populated this table with cases involving other patents whose numbers happen to end in "085" (see the naming-collision warning below).

Naming-collision warning — do not conflate these

The searches surfaced several unrelated patents commonly referred to as "the '085 patent." None is US 12,491,085. Listing them here so they are not mistakenly attributed:

• US 9,289,085 — Parallax Group Int'l, LLC v. Incstores LLC, appeal decided 2026-03-20, Fed. Cir. No. 2024-1798 (from C.D. Cal. No. 8:16-cv-00929-JVS-DFM). Dual-color interlocking floor mats. Not this patent.
• An '085 patent at issue in Axsome Therapeutics, Inc. v. Apotex Inc., D.N.J. No. 2:25-cv-16038 (ANDA litigation). Not this patent.
• An '085 patent asserted against SmallWorks in a USPTO PTAB/P-TACTS counterclaim filing (alongside '191, '892, '066), dated 2015-11-05. Not this patent — different field, different era, and it postdates nothing relevant here.
• Netlist's '085 patent referenced in Samsung/Avnet/Micron litigation. Not this patent.

The two searches explicitly for "12491085" and "US 12,491,085 patent litigation" returned only (a) unrelated '085 patents and (b) an office-action analytics page confirming bibliographic data for US 12,491,085 ("Granted Dec 09, 2025") — with no litigation field.

Adjacent finding: Carlsmed's enforcement activity (cannot be tied to this patent)

The only enforcement activity traceable to the assignee is the one already flagged in the prior section:

• Carlsmed, Inc. v. DeGen Medical, Inc. — U.S. District Court for the District of South Carolina, Florence Division. Complaint asserts eight Carlsmed patents "relating to its implants, methods and systems"; seeks injunctive relief and damages. Source: GlobeNewswire release, 2026-09-15 (https://www.odtmag.com/breaking-news/patent-infringement-lawsuit-filed-against-degen-medical/).

Two material caveats:

1. The eight asserted patent numbers were never disclosed. Carlsmed's own release states: "The company released no other details about the alleged patent infringement." I therefore cannot confirm or deny whether US 12,491,085 is among the eight. Do not assume it is. I also could not obtain the complaint text or a case number.
2. Date conflict — flagged, not resolved. The DeGen complaint is dated 2026-09-15, which is ~5 months after the current date given in this task (2026-04-26). The system context date is 2026-09-17. Three different dates are in play (query: 2026-04-26; system: 2026-09-17; DeGen filing: 2026-09-15). This was flagged in the earlier-generated section and the conflict persists — I am reporting the DeGen filing as found in the sources rather than silently reconciling it. Treat the DeGen matter as not established as of the query's stated date.

Why the null result is expected here

US 12,491,085 issued only on 2025-12-09 — roughly 4.5 months before the query's stated date. Assertion of a patent within months of issuance, while possible, is uncommon; most NPE and competitor suits follow issuance by a year or more. The parent, US 11,432,943 B2, is older (issued 2022-09-06) and would be the more likely first assertion vehicle. Neither produced litigation hits.

I also found no PTAB post-grant proceeding (IPR/PGR) against either US 12,491,085 or US 11,432,943, and no ITC Section 337 investigation naming Carlsmed. Given the 9-month PGR window relative to the 2025-12-09 issuance, a PGR would still be timely as of the query date.

What I could not verify (stated explicitly)

• No PACER / CourtListener / Docket Navigator query completed. PACER requires authenticated access; I did not reach it.
• No Unified Patents portal query completed. The portal query returned an empty result set; I cannot say whether that is a true no-hit or a retrieval failure.
• No CAFC docket query completed. I found no CAFC appeal naming US 12,491,085, but I could not confirm this against the court's own docket.
• No USPTO assignment-record or PatentCenter review of the '085 file history (e.g., any terminal disclaimer, PTA, or claim amendment that could bear on enforceability or term).
• The DeGen complaint's asserted patents and case number remain unknown.

Recommended next steps to close the gap

1. PACER, D.S.C. Florence Division, complaints against DeGen Medical filed on/around 2026-09-15 — retrieve the complaint to enumerate the eight patents.
2. Unified Patents litigation portal (https://portal.unifiedpatents.com/litigation/caselist), filter by party "Carlsmed" and by patent number 12491085.
3. CourtListener RECAP docket search for "Carlsmed" and for patent number 12491085.
4. PTAB P-TACTS / Patent Public Search, search US 12,491,085 for IPR/PGR filings and for citation in other proceedings.
5. SEC filings — Carlsmed (Nasdaq: CARL) 10-K/10-Q risk factors may name asserted patents or material litigation that the press release withheld.

Statement of uncertainty

To state it plainly: I found no litigation involving US Patent 12,491,085. I am not able to certify that none exists, because I could not query PACER, the Unified Patents portal, or the Federal Circuit docket directly. The single litigation matter associated with the assignee (Carlsmed v. DeGen Medical) cannot be linked to US 12,491,085 on the public record available, and its reported filing date postdates the date specified in this task. Any conclusion that this patent is or is not in suit would require the docket checks listed above.

I'll verify the canonical (empty) proceedings list against live sources before writing this up.

No hits on the patent number itself. Let me probe the assignee/patent family and the known 2026 litigation for any PTAB corollary.

Let me run two final targeted checks — the Unified Patents portal (a defensive aggregator) and any generic docket mentions.

Proceedings overview

Zero AIA trial proceedings on file. The USPTO Open Data Portal (ODP) API returns no IPR, PGR, or CBM proceedings against US 12,491,085 B2 as of the most recent ingest, and my independent web searches surfaced no petition, institution decision, final written decision (FWD), or Federal Circuit appeal naming this patent — so the defensive posture for a defendant today is "untested patent": none of claims 1–15 has been canceled, narrowed, or adjudicated at the Board, but equally, no petitioner has yet had to litigate the Board's claim-construction or validity views of this specification/claim-mismatched patent. There is no FWD to cite and no claim-level estoppel to exploit. The absence of activity is a timing artifact, not a merits signal (see Strategic summary).

Per-proceeding detail

There are no proceedings to detail. No proceeding number can be supplied because inventing one would violate the operating rules — any "IPR2024-xxxxx" or "IPR2026-xxxxx" you see attached to this patent number elsewhere should be treated as unverified.

For the record, here is what I checked and what turned up:

• USPTO ODP (canonical source): returns an empty AIA-trial set for US 12,491,085.
• Targeted web searches for the patent number with "PTAB," "IPR," "inter partes review," and "petition": no substantive hits — the only returns referencing 12,491,085 were Google Patents family/similar-document listings and third-party patent pages, not Board documents.
• Assignee-level search ("Carlsmed" + IPR/PTAB): returned only generic PTAB practice commentary (Jones Day, McGuireWoods, Patexia, RPX, Pearce IP FY-roundup articles) and a Unified Patents portal page for a different Carlsmed patent, US 12,232,980 B2, which lists Carlsmed filings as related documents but reflects no proceeding against 12,491,085.
• Litigation-adjacent search (Carlsmed v. DeGen Medical): surfaced the 2026-09-15 complaint (D.S.C., Florence Division) asserting "eight of Carlsmed's patents," but the press release does not name the asserted patents, and there is no reported IPR/PGR corollary as of 2026-09-17. Sources: https://www.odtmag.com/breaking-news/patent-infringement-lawsuit-filed-against-degen-medical/ ; https://stockhouse.com/news/press-releases/2026/09/15/carlsmed-inc-files-patent-infringement-lawsuit-against-degen-medical

Cross-reference to the earlier section — no contradiction, one corroboration. The prior bibliographic section flagged the specification/claim mismatch (mechanical interbody disclosure vs. computer-implemented design-method claims) and flagged that the DeGen asserted patents were undisclosed. Both points hold and both matter for PTAB strategy below.

Strategic summary

Claim status: 15 of 15 UNTESTED, 0 CANCELED, 0 SUSTAINED. No claim of 12,491,085 has been canceled, confirmed, or even construe-by-the-Board. Claims 1 and 11 (the two independents) remain fully intact and unamended, as do dependents 2–10 and 12–15. Practically, this means a defendant has no off-the-shelf PTAB record to lean on — no FWD finding, no Board construction of "negative space," "patient-specific contouring," or "deployable fixation element." Those terms are undefined-by-adjudication and are exactly where a § 112 written-description attack on this patent is strongest (the specification supports the design workflow of claims 1/11 with little more than one additive-manufacturing paragraph plus incorporation by reference of Ser. No. 16/207,116). That is a district-court § 282 / Rule 12(c) argument, not an IPR argument — see the § 101/§ 112 gate discussion below.

Estoppel landscape: essentially empty, which cuts both ways. Because no IPR has been instituted, § 315(e)(2) estoppel has not attached to anyone. A defendant is therefore free to raise any § 102/§ 103 ground, using any art, without worrying that a prior petitioner already spent it. Two structural limits still apply and are worth planning around:

1. IPR grounds are statutorily capped at § 102/§ 103 over patents and printed publications (35 U.S.C. § 311(b)). The most attractive attacks on this patent — § 101 (abstract design-workflow claims) and § 112 (written description/enablement of the patient-specific base and negative-space mapping) — cannot be brought in an IPR.
2. The PGR window has closed. The patent issued 2025-12-09; under § 321(c) a PGR petition was due within 9 months of grant, i.e., by approximately 2026-09-09 — about a week before the current date of 2026-09-17. (The application claims priority to 2018-03-14, so it is post-AIA and was PGR-eligible.) Unless a PGR was filed in that window and simply is not yet visible in my sources — I found none — § 101 and § 112 are now off the PTAB table entirely. That is a meaningful, permanent narrowing of the Board-side attack surface, and it is a loss for a challenger. CBM review is also unavailable: AIA § 18 sunset for CBM on 2020-09-16.

Pattern signals: none on this patent; a family/aggregator pattern cannot be ruled out. I found no repeat petitioner (obviously — no petitioner at all), no PTAB appeal history, and no defensive-aggregator involvement (Unified Patents has a portal entry for Carlsmed but no proceeding tied to this patent). Carlsmed is an active, well-funded enforcer — Nasdaq-listed (CARL), IPO July 2025, first disclosed infringement suit 2026-09-15 — which means it is precisely the kind of patent owner that attracts IPRs once a real product-on-product fight starts. The family context also matters: this patent is the continuation child of US 11,432,943 B2, with a further continuation on file (US 19/388,876, filed 2025-11-13). A challenger who knocks out claims 1–15 here may face near-identical method claims in the parent and the child — so any petition should be scoped with the family, not the single patent, in mind.

Deadline math for a newly-sued defendant. The DeGen complaint is dated 2026-09-15. If DeGen (or any defendant joining it) is served on or about that date, the § 315(b) one-year bar puts its IPR deadline at roughly 2027-09-15. Institution, if granted, would come at roughly the 6-month mark (~2028-03), with the FWD statutorily due within 1 year of institution (§ 316(a)(11), extendable up to 6 months for good cause) — i.e., mid-to-late 2029. Compare that to a typical D. S.C. trial track, and the NHK-Fintiv / discretionary-denial overlay (which, per the Director decisions reported through 2026, is being applied aggressively, including to ITC- and DOI-parallel cases) becomes a real institution risk that should be managed with a Sotera-style stipulation up front. See the LOT Network summary and the Director decisions it collects: https://lotnet.com/ptab-practice-under-current-leadership-key-changes-and-implications-for-patent-challenges-part-3/

One honest caveat on my search scope. Carlsmed holds a large US portfolio (50+ issuances visible on the family listings). I searched for proceedings against this patent and against the assignee generally, not against every sibling. It is possible a Carlsmed patent has been IPR'd without that appearing in my queries. What I can state with confidence is narrower and literal: no PTAB proceeding on US 12,491,085 was surfaced by the ODP block or by any search I ran.

Recommended next steps

Plain statement first: there is no PTAB activity on US 12,491,085. Per the task instructions, that absence is itself the signal — and here it tells you the patent was granted 2025-12-09, just over nine months ago. It has not yet been tested by anyone.

1. Confirm the PGR window is actually closed before you rely on it. Calculate from the actual grant date in PatentCenter, not from the Google Patents assumption, and check whether any PGR was filed on/before ~2026-09-09 and is simply not yet indexed. If a PGR was filed, it changes everything (PGR can raise § 101 and § 112, which IPR cannot).

   • PTAB E2E (proceeding search by patent number): https://ptab.uspto.gov/
   • USPTO Patent Public Search (application/patent lookup): https://ppubs.uspto.gov/pubwebapp/

2. If you are a defendant served with a complaint asserting this patent, calendar the § 315(b) bar now. For a 2026-09-15-service scenario, the IPR petition deadline is ~2027-09-15. Missing it is unrecoverable.

3. Scope the petition to § 102/§ 103 only, and treat § 101/§ 112 as a district-court track. Claims 1 and 11 are computer-implemented design methods; the abstractness and written-description problems are real but raiseable only in court (or in a PGR that has now lapsed). Do not build a Board strategy on them.

4. Mine the record's own 258 cited references for a § 103 combination, not a § 102 knockout. The examiner considered and passed on patient-specific implant design art including US 2015/0305878 A1 (DePuy Synthes, "Patient-Specific Spinal Fusion Cage and Methods of Making Same"), US 8,394,142 B2 (Synthes, "Customizing an intervertebral implant"), CN 104318009 A (personalized interbody fusion cage design), US 2017/0225207 A1 (NuVasive, spinal correction surgical planning), and US 2018/0303552 A1 (Medicrea). Cited art remains available for a new combination theory; what is needed is a motivation-to-combine narrative for "mapping a negative space between endplates and a deployable fixation element and a patient-specific correction." That is the weakest link in claims 1/11 and the best place to aim.

5. Watch DeGen's answer and counterclaims. The first defensive petition against a Carlsmed patent — on this patent or a sibling — will likely appear within 12 months of the 2026-09-15 complaint, i.e., by ~2027-09-15. If DeGen files, its petition and the Board's institution decision will hand every other defendant a free read on how the Board construes "negative space" and "patient-specific contouring."

6. Monitor via CourtListener for any Federal Circuit appeal naming 12,491,085 (there is none today; an appeal could only arise from a future FWD or from the district court): https://www.courtlistener.com/?q=%2212%2C491%2C085%22

Statement of uncertainty

• The zero-proceeding result is drawn from the structured ODP block supplied in this prompt, which is designated the canonical list. I corroborated it with multiple web searches but could not query PTAB E2E or PatentCenter directly with the tools available, so I cannot affirmatively rule out a proceeding filed very recently that neither ODP nor web search has indexed.
• I did not verify the patent's grant date, the PGR deadline calculation, or the § 315(b) dates against USPTO docket records. The 2026-09-09 PGR expiration is my arithmetic (2025-12-09 + 9 months) and should be independently confirmed.
• The eight asserted patents in Carlsmed, Inc. v. DeGen Medical, Inc. (D.S.C., Florence Div., filed 2026-09-15) are not publicly identified. Do not assume US 12,491,085 is among them; do not assume it is not. The earlier section flagged this, and nothing in my searches resolved it.
• No judge panels, petition grounds, institution decisions, FWDs, settlements, or appeal dispositions are reported above because none exist to report. Per the constraints, I have not fabricated proceeding numbers, claim-level holdings, or quotes.
• Note the date conflict carried forward from the prior section: the system context date is 2026-09-17, while the original query referenced a 2026-04-26 frame. All dates above use 2026-09-17 as "today."

I'll research the assignment record and ownership chain for US 12,491,085, starting with the USPTO Assignment Center and related sources.

Let me dig into the assignment records and the company's financing/IP filings.

Inventors

Pattern notes:

• Not a departing-inventor pattern. The named inventor is the sitting Chairman/CEO of the assignee, and the assignee is an active, Nasdaq‑listed operating company (CARL) that is currently enforcing its portfolio (see below). This is the opposite of the classic "inventors bolt within 12 months, portfolio gets dumped" precursor. Do not read any fire-sale signal here.
• Invention-assignment hygiene caveat. Carlsmed's own S‑1/A (Form CARL, filed 2025‑07‑15) carries the standard risk factor that it "may be subject to claims that our current or former employees, contractors, or other third parties have an ownership interest" in its IP (https://capedge.com/filing/[1794546](/patent/1794546)/0000950170-25-095710/CARL-S1A). That is boilerplate, not evidence of a defect — but it is the reason to check the recorded inventor→Carlsmed assignment rather than assume it.
• I did not find, and cannot confirm, an inventor-departure event for either named inventor.

Original assignee

Carlsmed, Inc. (Carlsbad, California) — named as both original and current assignee on Google Patents.

• Primary line of business: medical device / medtech. "AI‑enabled personalized spine surgery solutions" — patient‑specific interbody fusion implants and the associated surgical planning platform (https://www.odtmag.com/breaking-news/patent-infringement-lawsuit-filed-against-degen-medical/).
• Does it ship a product embodying the claims? Yes — materially. This matters because the claims here are design‑method claims (claims 1 and 11), so the practice question is whether Carlsmed commercializes patient‑specific, imaging‑derived implants with a deployable fixation element. Evidence: (a) it is a public reporting company; (b) its S‑1/A (2025‑07‑15) and its Form 10‑K (filed 2026‑02‑25, Accession 0001193125‑26‑071551) describe an operating commercial business, not a licensing shop; (c) it obtained a favorable CMS reimbursement decision in August 2026 and filed a $250M shelf registration; (d) it is asserting eight of its patents against an identified competitor. The written description of this patent (Additive‑manufacturing paragraph; FIGS. 24–26 deployable teeth) maps onto that commercial product line.
• Current status: Operating. Public. Nasdaq: CARL. IPO process visible via amended S‑1 filed 2025‑07‑15; annual report on Form 10‑K filed 2026‑02‑25. Not acquired, not dissolved, not in bankruptcy.
• Enforcement posture: Plaintiff, in its own name, in Carlsmed, Inc. v. DeGen Medical, Inc., U.S. District Court for the District of South Carolina, Florence Division, filed 2026‑09‑15, asserting eight Carlsmed patents "relating to its implants, methods and systems," seeking injunctive relief and damages. The eight patent numbers were not disclosed in any public source I retrieved — US 12,491,085 is NOT confirmed to be among them. Do not assume it is. (Note: Florence, SC is consistent with the defendant's known location, which would explain venue; I could not verify DeGen's headquarters from a primary source in this session.)

Assignment timeline

Data-access limitation, stated plainly: I could not query the USPTO Assignment Center (https://assignmentcenter.uspto.gov/) or the legacy AOTW interface (https://assignment.uspto.gov/patent/index.html) with the tools available in this session. I therefore retrieved no reel/frame numbers, no recorded-conveyance types, no assignor/assignee records, and no corresponding-of-record names for US 12,491,085. I will not invent them. The timeline below is built only from records I could verify (Google Patents legal events + the continuity data in the patent's own front matter), and each entry is labeled by source and confidence.

What the verifiable record shows:

Observation (weak, flagged as such): the Google Patents legal‑events block for this patent shows only filed / priority to / publication / granted events and no "Assignment" entries. Google Patents' assignment‑event coverage is incomplete, so this is not proof of absence — but it is consistent with no post‑issuance assignment having been recorded, i.e., the original assignee still owns the patent. Per the operating rules for this task, that is itself a finding.

Unverified item worth a manual reel/frame pull — do not treat as a finding: a CARLSMED, INC. SIGNATURE BANK LOAN AND SECURITY AGREEMENT is publicly posted as a contract (https://www.lawinsider.com/contracts/fLTlW0tHrxY). If a collateral schedule covering this patent family was recorded, it would appear in Assignment Center as a Security Agreement (not an ownership transfer). I found no confirmation that this patent or application 17/875,699 was recorded under that agreement, so I am not listing it as an assignment. Note for whoever pulls the record: Signature Bank failed in 2023 (its assets moved to Flagstar/NYCB), so any recorded security interest from that facility would carry a non‑obvious chain of secured parties — relevant to a lien search, irrelevant to ownership.

Recommended verification query: Assignment Center → search by Patent Number 12491085 (and separately by Application Number 17/875,699), plus an assignor/assignee name search for "Carlsmed" to catch portfolio‑level recordings where the patent appears only in an exhibit schedule. Links: https://assignmentcenter.uspto.gov/ and https://assignment.uspto.gov/patent/index.html.

Two date/context conflicts to carry forward, not to resolve silently:

1. The query fixed "today" at 2026‑04‑26; the environment's operating date is 2026‑09‑17; the DeGen complaint is dated 2026‑09‑15 — i.e., it postdates the query date. Any "pre‑litigation transfer within 6 months" analysis is therefore anchored differently depending on which date you adopt.
2. The previously generated bibliography lists the DeGen case as filed 2026‑09‑15, matching the GlobeNewswire release. No contradiction there — but the previously generated section's statement that "no CAFC 2026 appeal naming US 12,491,085 was found" remains unconfirmed for the same docket‑access reason as above.

Timeline diagram

timeline
    title Ownership of US 12491085
    2018 : Provisional filed by Casey and Cordonnier
    2019 : Non-provisional 16 352 699 filed
         : Carlsmed Inc named as applicant and assignee
    2022 : Parent patent 11432943 issued
         : Continuation 17 875 699 filed
    2023 : Pre-grant publication
    2025 : US 12491085 granted
         : Child continuation 19 388 876 filed
    2026 : Carlsmed sues DeGen Medical

(No assignment links are drawn because no reel/frame records were retrieved. The chain shown is application continuity, not recorded conveyances.)

NPE / troll-pattern signals

Adjacent risk profile worth noting (validity, not ownership): the more interesting exposure on this document is the specification/claim mismatch flagged in the earlier section — the abstract and FIGS. 24–26 describe a mechanical break‑away/deployable‑tooth interbody device, while all 15 issued claims recite computer‑implemented design steps (map a negative space between endplates; design a base with two patient-specific bone-contacting surfaces). Written-description/enablement support for the claimed design workflow is thin in this specification, which rests largely on an additive-manufacturing paragraph plus incorporation by reference of application Ser. No. 16/207,116. A defendant served on these claims will reach for §112 (and likely §101) before it reaches for ownership. I am not opining on any legal conclusion.

Verdict

Operating-company assertion.

Carlsmed, Inc. — the original applicant and assignee of record — is the current owner, and it is a commercial-stage, Nasdaq‑listed medical device company (S‑1/A filed 2025‑07‑15; Form 10‑K filed 2026‑02‑25; https://capedge.com/filing/1794546/0000950170-25-095710/CARL-S1A) that sells the patient‑specific implants its design methods produce and, on 2026‑09‑15, sued competitor DeGen Medical in D.S.C. Florence Division on eight of its own patents (https://www.odtmag.com/breaking-news/patent-infringement-lawsuit-filed-against-degen-medical/). Every NPE-pattern signal is either absent or unassessable: there is no shell‑entity transfer, no known asserter, no cascading LLC chain, no pre-suit assignment, no bankruptcy, and no defensive aggregator; the only outstanding gap is the correspondent-of-record field, which I could not retrieve at all.

Confidence and limits: the verdict rests on the current-assignee field and the company's public operating status, not on retrieved reel/frame records — I retrieved none. It is also not a finding that US 12,491,085 is among the eight DeGen patents; that was never disclosed publicly. Verify at https://assignmentcenter.uspto.gov/ (patent no. 12491085; app no. 17/875,699; assignee name Carlsmed) and capture the correspondent on each entry, since the earlier "no CAFC 2026 appeal" statement and this ownership chain both remain docket‑unverified. If the Assignment Center returns records, this section must be rewritten; if it returns nothing, that publication date (2025‑12‑09) is simply too recent to expect conveyances, and the finding becomes "original assignee still owns the patent."

I'll research this systematically. Let me start with searches for the patent itself and its key cited references.

Good material. Let me now verify the examiner-cited (asterisked) references and a few others most relevant to the claims.

I was unable to complete every planned query (the session hit its step limit mid-way), so the analysis below is built on the authoritative full text of US 12,491,085 B2 plus the searches that did complete. I flag gaps explicitly at the end.

Prior-Art Analysis — US 12,491,085 B2

1. Confirmation of the target document

I locked onto exactly one document and did not substitute anything similar:

Critical framing for the §102 analysis: the issued claims are not mechanical‑device claims. Claims 1–15 are all computer‑implemented / design-method claims. The single most important fact for anticipation analysis is the effective filing date of those claims. The sources give 2018‑03‑14, but the claims are directed to a design workflow whose written‑description support in the specification appears to be a single additive‑manufacturing paragraph plus incorporation‑by‑reference of app. Ser. No. 16/207,116. If the 2018 provisional did not support the design‑method claims (whose subject matter differs from the abstract/description, which is entirely about deployable teeth), the claims may be entitled only to a later date — which would move several references in and out of the prior‑art window. I could not review the provisional's contents or the file history, so I state this as an unresolved issue, not a conclusion.

2. About the citation corpus

US 12,491,085 lists 258 cited references on its face. I cannot individually analyze 258 items with the tools here, and doing so would be low‑value. I prioritized:

• the three references the examiner specifically relied on (marked "* Cited by examiner" in the source: US 2005/0049590 A1, US 2018/0301213 A1, US 2015/0305878 A1);
• the references that map onto the claim limitations (patient‑specific 3‑D spine model → user adjustment → "negative space" mapping between endplates → patient‑specific bone‑contacting surfaces → deployable fixation element);
• the references from different assignees (the Carlsmed family members are by definition not anticipatory third‑party art in the ordinary sense).

3. Examiner‑cited references (highest probative value)

3a. US 2005/0049590 A1 — Alleyne et al., "Spinal implant with securement spikes"

• Citation: US 2005/0049590 A1 (app. 10/794,586), filed 2004‑03‑05; published 2005‑03‑03; priority/provisional 60/453,242 filed 2003‑03‑07. Cited by examiner.
• Description: Interbody implant with spikes that reside retracted within the housing and are extended after insertion; several driver mechanisms (sliding wedge, threaded rotating wedge, camshaft, worm gear, rack‑and‑pinion); barbed spikes; express method claim of "inserting the implant between the vertebral bodies, extending spikes coupled to a body of the implant into the vertebral bodies, and using the spikes to pull the vertebral bodies toward the implant body." Spikes captured in holes in the housing.
• Claim relevance: Maps squarely onto the "deployable fixation element" element of claims 1 and 11 and onto claim 6 ("a plurality of deployable teeth") and the opening/channel aspect of claim 8. Alleyne has no 3‑D spine modeling, no user‑adjustment step, no negative‑space mapping, and no patient‑specific bone‑contacting surfaces. It therefore cannot anticipate claims 1 or 11 standing alone; its value is as the primary teaching of the deployable‑fixation element and as §103 art against the fixation sub‑features of claims 6/8 (and arguably claim 7 if read to cover threaded driving mechanisms — though Alleyne's claims recite spikes, not screws).

3b. US 2015/0305878 A1 — O'Neil et al. (DePuy Synthes), "Patient‑Specific Spinal Fusion Cage and Methods of Making Same" (granted as US 9,757,245 B2)

• Citation: US 2015/0305878 A1, filed 2014‑04‑24; published 2015‑10‑29; granted US 9,757,245 B2 on 2017‑09‑12; continuation US 10,405,987 B2. Cited by examiner.
• Description: An expandable trial with endplate‑mapping capability is inserted into the disc space; the "geometry of the dome (i.e., the cavity between the trial and the endplates)" is determined by direct light/ultrasound imaging through the trial, by balloon molding of a curable fluid that "conforms to the contour of the vertebral endplate," or by light‑based imaging. That 3‑D cavity image/molding is then used "to create a patient‑specific intervertebral implant" — including a patient‑specific endplate attached to a modular core — to maximize contact area and reduce subsidence/expulsion.
• Claim relevance: This is the closest reference on the "mapping a negative space between a first and second vertebral body endplate … wherein the negative space includes contouring corresponding to the patient‑specific contouring of the … endplates" limitation and on the two patient‑specific bone‑contacting surfaces of claims 1 and 11. It potentially anticipates the negative‑space‑mapping and contour‑matching core of claims 1/11, but it lacks the deployable fixation element and its mapping is intra‑operative (the disclosure expressly criticizes pre‑operative CT/MRI‑derived approaches as assuming disc‑space geometry that can't be known pre‑surgery). That distinction cuts against anticipation of independent claim 11, which recites "non‑invasively determining a patient‑specific correction … by adjusting a virtual three‑dimensional model" — pre‑operative modeling, not intra‑operative balloon molding. Strong §103 art; not clean anticipation.

3c. US 2018/0301213 A1 — Mazor Robotics Ltd., "Global spinal alignment method"

• Citation: US 2018/0301213 A1, priority 2015‑10‑13; published 2018‑10‑18. Cited by examiner.
• Description: Computing a global spinal alignment/surgical plan (sagittal balance parameters) from patient imaging and simulating correction.
• Claim relevance: Bears on the model‑and‑correct steps of claims 1/11 and on dependent claims 2/13 (linear displacement), 3/14 (rotation), 4/15 (coronal/sagittal/axial plane deformity).
• Timing nuance (important): Although it published 2018‑10‑18 — after the 2018‑03‑14 priority date — it is a U.S. application publication with an effective filing date of 2015‑10‑13. It is therefore available as §102(a)(2) prior art (not §102(a)(1)), assuming the "names another inventor" condition is met. If the '085 claims are only entitled to a later effective date, its status strengthens.

4. Other high‑relevance references (design‑method / patient‑specific‑implant art)

5. Claim‑level anticipation matrix (which references potentially anticipate which claims)

Because claims 1 and 11 are combination claims (design workflow plus a patient‑specific base plus a deployable fixation element), no single reference in the corpus appears to disclose every element of claims 1 or 11. My assessment:

6. Critical caveats to carry forward

1. Co‑owned family members are not ordinary third‑party art. Many "cited" documents (US 11,083,586; WO 2019/112917; US 2020/0078180; US 2021/0059822; the '943 parent, etc.) are Carlsmed's own. The parent US 11,432,943 B2 is the same family — not prior art against the '085.
2. "Anticipates" vs. "renders obvious." I am reporting which claim(s) each reference could bear on. For the independent claims, the honest characterization is that the references are §103 combination art (design‑workflow references + deployable‑fixation references), not standalone §102 anticipatory art.
3. Post‑priority‑date references in the citation list — e.g., US 2020/0289288 A1 (Medos, 2019‑03‑14) and US 2021/0145519 A1 (Medicrea, 2019‑04‑02) — are not prior art against a claim entitled to the 2018‑03‑14 date, but become prior art if claim priority is lost. See caveat 6 below.
4. §102(a)(2) timing. US 2018/0301213 A1 published after the '085 priority date but has a 2015‑10‑13 effective filing date; it is §102(a)(2) art, not §102(a)(1) art. The distinction matters for any §102(b)(2) exceptions.
5. Grace‑period caveat. §102(b)(1) excepts only the inventor's own (or derived) disclosures within one year before the effective filing date. It does not except third‑party references in that window, so the one‑year grace period provides essentially no shelter here other than for Carlsmed's own prior publications.
6. The claim‑priority question is the swing factor for this entire analysis (see §1). If the design‑method claims are not supported by the 2018‑03‑14 provisional, the effective date could be 2019‑03‑13, 2022‑07‑28, or later, which would materially enlarge the prior‑art set (e.g., bringing in Medicrea's 2019 filings, the Medos 2019 publication, and more of the Carlsmed 2019‑2022 publications).

7. Statement of uncertainty — what I could and could not do

• Source of record: the authoritative Google Patents full text of US 12,491,085 B2 (which reproduces the front‑page citations), supplemented by live searches. I did not open USPTO PatentCenter or Patent Public Search (PatFT) directly to pull the examiner's citation list or file history — my searches returned Google Patents pages and secondary aggregators, which mirror the front‑page data.
• I did not individually analyze all 258 cited references. The tables above cover the examiner‑cited references and the references most directly tied to the claim limitations. Remaining citations are largely mechanical‑interbody, design‑tooling, GUI‑design, data‑security, and dental/orthodontic references (e.g., the numerous Conformis, Biomet, Medicrea, and design‑patent entries) with lower relevance to the computer‑implemented design claims.
• I could not retrieve: the Chinese‑language full text of CN 104318009 A; the full text of several pre‑grant publications; the '085 file history (any amendments, terminal disclaimers, PTA); and the 2018 provisional's contents (62/643,046) — which is decisive for the priority question.
• Discrepancy to verify against the printed patent: the foreign citation appears as CN 104318009 A on the '085 face and CN 104318000 A on the sibling US 11,432,943 — one of the two is likely a transposition.
• Any statement that a given reference "anticipates" a given claim is a technical mapping of disclosed elements to claim language, not a legal conclusion; a formal anticipation opinion would require the full reference texts (and, for the design methods, the claim‑priority determination above).

I'll ground this analysis in the prior-art references cited on the patent's own face, then verify the key ones against live sources.

The two examiner-cited references are confirmed. Let me verify the third examiner citation and check prior art covering the "patient-specific correction" (adjusting a virtual model) element.

I have the three examiner-cited references confirmed. Let me check two more threads: prior art on the "negative space" mapping concept, and on patient-specific implants with deployable fixation.

Obviousness Analysis — US 12,491,085 B2 under 35 U.S.C. § 103

Scope and method. This analysis uses the prior-art references listed on the face of the patent (the Google Patents "Citations (258)" table), giving priority to the three references the examiner marked with an asterisk (*) — US 2005/0049590 A1, US 2015/0305878 A1, and US 2018/0301213 A1. I retrieved and verified the substance of those three plus several others in the list. Where I rely on a reference's abstract or published description rather than its full specification, I say so.

Key framing point carried forward from the prior section. The written description describes a mechanical interbody device, while all 15 issued claims are computer-implemented design methods. That mismatch has a direct § 103 consequence discussed in § 6 below: it puts the claims' entitlement to the 2018-03-14 provisional at risk, which in turn changes the prior-art date.

1. Governing framework

• AIA applies. Effective filing date is on/after 2018-03-14, so §§ 102(a)(1)/(a)(2) and 103 govern.
• Graham v. John Deere factors: scope and content of prior art; differences between prior art and claims; PHOSITA level; objective indicia.
• KSR Int'l v. Teleflex and MPEP § 2143: a combination is obvious where (A) the references are from the same field or analogous fields, (B) the combination addresses a known problem with predictable results, (C) there is an explicit teaching/suggestion/motivation (including the inventor's own admission of the problem), or (D) there are "a finite number of identified, predictable solutions" — i.e., it was "obvious to try."
• Design-method claims. Claims 1 and 11 recite designing a system that includes a deployable fixation element. They do not require that the element ever deploy, that the implant be manufactured, or that it be implanted. This materially narrows the gap between the claims and the prior art.

Postulated PHOSITA: a biomedical/mechanical engineer with 2–4 years' experience in spinal implant design, or a spine surgeon, working with (i) spinal imaging/3D-modeling software (CT/MRI segmentation, surgical planning) and (ii) additive-manufacturing workflows for patient-specific orthopedic devices. That is the level at which the asserted art is pertinent, and it is at least the level of the inventors' own disclosure, which is largely a review of existing implant types (FIGS. 1–23).

2. The references (prior-art date check)

Notable secondary point: Alleyne's own background names Vaccaro (US 6,102,950) and Schafer (US 5,800,547) as earlier interbody cages with expanding spikes. That means deployable fixation on an interbody cage was an established technique roughly 19 years before this patent's priority date — before either the "patient-specific" or the "additive manufacturing" elements existed.

3. Claim chart — independent claims 1 and 11

Directionality note for claim 11. Claim 11 requires "non-invasively determining" the correction. O'Neil's endplate mapping is intra-operative (a trial inserted into the disc space). For claim 11, the "correction" element should therefore be anchored on Mazor (pre-operative X-ray/CT) rather than on O'Neil. O'Neil remains the best art for the negative-space/contouring element because it discloses that element in virtual/3D-image form, which is a non-invasive output. This is the one place where the two independent claims diverge meaningfully.

4. Primary ground of rejection — Ground 1

O'Neil '878 (DePuy Synthes) in view of Mazor '213, further in view of Alleyne '590.

What O'Neil alone teaches. A method of determining disc-space geometry using an endplate-mapping trial; the resulting 3D image of the cavity between the trial and the endplates is used to create either a patient-specific intervertebral implant or patient-specific endplates that attach to a modular core; the cage's purpose is expressly "to maximize contact area between the prepared endplates and the fusion cage with the objective of reducing subsidence and expulsion."

What O'Neil does not teach. (a) An adjustable virtual model of the spine that the user manipulates to define a correction; (b) a deployable fixation element.

Mazor supplies (a). Mazor's whole thesis is that the surgeon should reposition vertebrae in a virtual model — a process it describes as being "similar to graphic engineering design programs… the patient's spine could be adjusted in the same way as an engineer may … design a product using a CAD program." It outputs an optimized, corrected spinal alignment constrained by the patient's measured range of motion. That is "generating an adjustable virtual three-dimensional model" plus "receiving one or more user-inputted adjustments … to provide one or more patient-specific corrections." It also walks through adjusting SVA (a linear displacement), lumbar lordosis and thoracic kyphosis (rotations), and sagittal/coronal/axial alignment — covering dependent claims 2–4 and 13–15.

Mazor even explains why a plan should be reflected in an implant: it criticizes planning that "may attempt to compensate the curvature at disadvantageous regions," i.e., the correction must be physically realized in the construct placed between the vertebrae.

Alleyne supplies (b). An interbody cage whose fixation elements start retracted inside the body and are deployed into the adjacent vertebral bodies — architecturally the same "first relative position → second relative position" concept recited in the '085 claims.

Motivation to combine (KSR/MPEP 2143 rationales):

1. Same field and same problem. All three address interbody fusion cages. O'Neil and Alleyne both frame the problem as inadequate fixation of a cage between endplates. Indeed, the '085 specification itself recites O'Neil's problem statement almost verbatim ("insufficient contact area or fixation … can result in micro-motions and/or macro-motions … expulsion of the interbody implant can result") — an admission of the motivating problem in the prior art.
2. Complementary, non-overlapping solutions to the same problem. O'Neil solves it by maximizing conformity and contact area; Alleyne solves it by actively penetrating the endplate. A PHOSITA seeking to reduce expulsion would predictably do both, because the two mechanisms are independent and additive. This is the classic "known technique applied to a known device, yielding predictable results."
3. Finite number of identified solutions. Patient-specific contouring and deployable/expanding fixation were each the two recognized answers to implant migration by 2018. KSR's "obvious to try" rationale applies squarely.
4. Same manufacturing platform. Both O'Neil and the K2M art rely on additive manufacturing of patient-specific geometry; the '085 specification concedes that "additive manufacturing processes are utilized which enable fabrication of features 230, 232, teeth 202, bridges 224." A PHOSITA would recognize no technical barrier to printing a conforming cage with an integral deployable tooth.
5. Reasonable expectation of success. Nothing in the combination calls for unpredictable results; the components are mechanically independent.

5. Secondary grounds

Ground 2 — Medicrea '552 as primary

US 2018/0303552 A1 (Medicrea) in view of O'Neil '878 and Alleyne '590.
Medicrea '552 is arguably the closest single reference to the claims as issued, because it recites a processor-executed workflow that: accesses medical images of the patient's spine; simulates implantation on the images (rotating image portions around reference points "to obtain a desired surgical output curvature"); determines dimensions of one or more patient-specific cages (posterior/anterior height) for the intervertebral spaces; and "generate[s] cage manufacturing or selection data instructions … for use by a cage manufacturing or selection apparatus." That last step reads directly on claims 10 and 12. Adding O'Neil for endplate-contour (negative-space) geometry and Alleyne for the deployable fixation element completes claims 1 and 11. Caveat: the relied-upon disclosure must be traceable to Medicrea's 2017-04-21 provisional(s) to establish the § 102(a)(2) date — I have not verified that against the provisional text.

Ground 3 — for the dependent claims

• Claims 6 (plurality of deployable teeth) and 8 (channel / fixation element advancing therethrough): K2M US 2016/0296342 A1 (intervertebral body with a through-hole extending from the front surface toward an adjacent vertebral body, with a fixation member received through the through-hole) plus Alleyne (plurality of spikes); Stryker US 2018/0338841 A1 and the '085 specification's own FIG. 24 (static teeth 206 / dynamic teeth 202) further support.
• Claim 7 (screw): Michelson US 2002/0016595 A1 ("Screws of cortical bone and method of manufacture thereof") and the "screws or barbs … driven through openings in the interbody implant" practice described in the '085 Background.
• Claim 9 (locking mechanism): Alliance Partners US 9,381,093 B1 ("Locking device for fixation mechanism of medical implant") — directly on point and cited on the face of the '085 patent; Alleyne's wedge/threaded/worm-gear drivers also inherently retain the spikes in the extended condition.
• Claim 5 (2D image data → 3D model): Mazor (2D X-ray registered onto CT) and Nuvasive US 2017/0252107 A1 (calibrate 2D X-ray images and generate a composite 3D image) and US 2016/0074048 A1 (image segmentation to generate joint models).
• Claims 2–4 / 13–15: Mazor (SVA/translation, lordosis/kyphosis rotation, sagittal-coronal-axial planes), Nuvasive '107, and K2M US 2020/0261156 A1 / US 11,000,334 B1 ("Systems and Methods for Modeling Spines and Treating Spines Based on Spine Models") — all in the cited-art list.
• Claims 10 / 12: Medicrea '552's "cage manufacturing … data instructions"; O'Neil's intra-operative fabrication; K2M '267's additive manufacturing of custom implants; Conformis US 2014/0086780 A1 (solid freeform fabrication for implant components).

6. The § 103 analysis interacts with the priority/§ 112 problem

The prior section flagged that the specification's support for the claimed design workflow is essentially one paragraph on additive manufacturing plus incorporation of Ser. No. 16/207,116. If the claims are not supported by provisional 62/643,046 (2018-03-14), their effective filing date shifts to 2019-03-13 (the § 16/352,699 filing). Consequences:

• Mazor '213 (pub. 2018-10-18) and Medicrea '552 (pub. 2018-10-25) — both after 2018-03-14 but before 2019-03-13 — would become directly available under § 102(a)(1) rather than only via § 102(a)(2) "effectively filed" dates, removing the need to prove provisional support for the relied-upon disclosure. The obviousness case gets stronger, not weaker.
• Conversely, if the 2018-03-14 date holds, the case rests on O'Neil (2015) and Alleyne (2005) plus Mazor via its PCT publication WO 2017/064719 (2017-04-20). O'Neil and Alleyne alone are unmovable, so Ground 1 survives either way.

Either way, the three examiner-cited references collectively traverse essentially every limitation of claim 1 — which is itself evidence that the field was predictable and the solution conventional.

7. Rebuttal considerations and their weight

1. "O'Neil is intra-operative, not a virtual pre-operative plan." The strongest available counter. But it attacks only one element, and Mazor independently supplies the virtual-model element. Since the claim does not require the negative-space map to be generated intra-operatively, importing O'Neil's timing into the claim would be improper.
2. "Negative space" is a term of art. Likely a semantics dispute. O'Neil's "cavity," "dome/geometry dome," and balloon "molding" (whose contours are the inverse of the endplate contours) are functionally the negative space; Medicrea's "simulate implantation … to one or more intervertebral spaces" likewise defines the inter-endplate volume.
3. "Deployable tooth ≠ spike." Weak. The '085 specification uses "teeth" and "spikes" interchangeably in substance (dynamic teeth 202 extend and penetrate), and Alleyne's spikes extend and penetrate the same way. Functionally identical.
4. Teaching away. None apparent. O'Neil and Alleyne are cumulative, not contradictory. The '085 Background describes both static and screw/barbe fixation as acceptable existing practice — the opposite of a teaching away.
5. Secondary considerations. Carlsmed's commercial traction in patient-specific spinal implants is real (the assignee is a publicly traded spine company; the DeGen matter surfaced in the prior section). But: (a) any commercial success must be nexused to the claimed design method — the claims read on a design workflow, and the revenues likely flow from implants and rods covered by other Carlsmed patents; (b) the field was crowded and the results predictable, which cuts against nonobviousness; and (c) the earlier section confirmed the DeGen suit cannot be tied to US 12,491,085 on the public record, so no litigation-based objective indicia can be claimed for this patent.
6. File history. The examiner cited O'Neil, Alleyne, and Mazor against these claims and still allowed them. I have not reviewed the applicant's remarks or any amendments in 17/875,699 — that history could contain the specific reason the examiner found the combination lacking (likely a narrow reading of "negative space" tied to the pre-operative virtual-model adjustment). This is the single biggest unknown in the analysis and should be retrieved before relying on it.

8. Bottom line

The strongest § 103 position is a three-reference combination: O'Neil '878 (DePuy Synthes) + Mazor '213 + Alleyne '590, with Medicrea '552 as an alternative/closer primary for the "generate fabrication instructions" and "simulate implantation" limitations. Each limitation of claims 1 and 11 is taught or suggested; the shared problem (implant subsidence and expulsion from inadequate endplate conformity and fixation) is expressly articulated in the primary reference and admitted in the '085 specification's own Background; the components are mechanically independent; and every element was individually known and 3D-printable. The KSR rationales — same field, combination of known elements with predictable results, finite number of identified solutions, and "obvious to try" — apply without strain.

The dependent claims fall with the independent claims for the same reasons, with Alliance Partners '9,381,093 and K2M '342/'267 supplying the narrower locking-mechanism, channel, and additive-manufacturing limitations.

What I could not verify (stated explicitly):

• Full claim text and complete specifications of O'Neil, Alleyne, Mazor, and Medicrea were not retrieved in full; the mappings above rest on their published abstracts, summaries, and excerpts. O'Neil in particular deserves a full-text read of its claims for a § 102 anticipation check on claims 2 and 11.
• The Medicrea provisional-support question (§ 102(a)(2) date) is unresolved.
• The prosecution history of 17/875,699 (applicant remarks distinguishing O'Neil/Alleyne/Mazor) was not obtained.
• I have not verified whether any of these references were applied to the parent US 11,432,943 B2 with a different claim scope.

Nothing here is a legal conclusion. This is a technical prior-art mapping; an invalidity opinion would require the full texts, the 17/875,699 file history, and the parent's prosecution record.