Patent: 10,233,503

Summary for Patent: 10,233,503

Title:	Method for the identification of the origin of a cancer of unknown primary origin by methylation analysis
Abstract:	The invention relates to methods and reagents for the identification of the origin of a carcinoma of unknown primary origin (CUP) based on the determination of the methylation profile in the genome of the CUP. The invention relates as well to methods for selecting a suitable therapy for a patient suffering a CUP as well as to methods for personalized medicine of patient suffering a CUP based on the use of a treatment which is adequate for the primary tumor from which the CUP is derived. The invention also relates to kits comprising reagents adequate for performing the above methods as well as to computer systems and programs which can be used for implementing the methods of the invention.
Inventor(s):	Badosa; Manel Esteller (Barcelona, ES)
Assignee:	FUNDACIO INSTITUT D\'INVESTIGACIO BIOM DICA DE BELLVITGE (IDIBELL) (Barcelona, ES)
Application Number:	14/402,736
Patent Claims:	see list of patent claims

United States Patent 10,233,503 (CUP methylation profiling): What claims are protected and how strong is the US patent estate?

Executive summary: US Patent 10,233,503 claims a method for diagnosing and treating cancer of unknown primary (CUP) by measuring a methylation profile at a selected DNA region using probe-based detection of methylation at ≥9 CpG sites, comparing that CUP profile to a reference methylation profile from known primary tumors, then diagnosing the primary tumor when profiles match and treating with therapy targeted to the diagnosed primary tumor (with extensive cancer-type-specific therapy mappings). The claim scope is broad at the “algorithm + CpG panel + matched reference profile” level, but is constrained by (i) a defined CUP-to-primary methylation profile comparison, (ii) use of a selected region with ≥9 CpGs, and (iii) the requirement that detection uses contacting DNA with “a plurality of probes specific for each CpG site” and a methylation detection reagent. The patent’s enforceability in practice will hinge on whether accused methods use (a) CpG-site-resolved probe sets rather than locus-agnostic readouts, (b) at least 9 CpGs within the compared region, and (c) direct profile matching against primary-tumor methylation reference data to assign the CUP origin.

---

What does US Patent 10,233,503 claim for CUP diagnosis and therapy selection?

Direct claim structure (independent claim 1): The method recites a full clinical workflow: sample acquisition → DNA isolation → methylation profiling on a selected region → probe-based methylation detection of ≥9 CpG sites → comparison to a primary-tumor reference profile from the “same selected region” → diagnosis based on profile match → therapy selection based on a table mapping primary tumor type to therapeutic modalities.

Core technical limitations that define infringement risk

1. Cancer of unknown primary (CUP) diagnosis context
   • The method is limited to a “subject suffering from” CUP.
2. Selected DNA region and reference-matched profiling
   • CUP methylation profile is compared to a primary tumor DNA sample profile “of the same selected region.”
3. At least 9 CpG sites
   • Detection must determine methylation statuses of ≥9 CpG sites within the isolated DNA region.
4. Probe-based CpG-specific methylation readout
   • Claim language requires contacting DNA with “a plurality of probes specific for each CpG site” plus a reagent for detecting methylation.
5. Diagnosis by matching
   • Diagnosis is triggered when CUP profile is “identified as having the same methylation profile” as the primary tumor profile for that region.
6. Treatment targeted to the diagnosed primary tumor
   • The claim then requires treating with therapy selected from an explicit therapy table.

Therapy mapping is part of claim 1

Claim 1 integrates diagnosis with treatment modality. The table includes:

• Lung: platinum-based compounds
• Colon: antimetabolites
• Melanoma: cytokines
• Pancreatic: antimetabolites
• Prostate: hormonal therapy and/or mitotic inhibitors for resistant patients
• Glioma: DNA-alkylating drugs
• Bladder: antimetabolites and/or platinum-based compounds
• Ovarian epithelial: platinum-based compounds
• Hepatobiliary: antimetabolites and/or EGFR-targeted drugs
• Breast: hormonal therapy; or hormonal therapy + cytostatic agents selected from anthracycline, DNA-alkylating, and antimetabolite; and/or HER2-targeted drugs
• Lymphoma: CD20-targeted drugs
• Head and neck: mitotic inhibitors ± platinum-based compounds and/or antimetabolites
• Endometrial: hormonal therapy
• Myeloma: corticosteroids; proteasome inhibitors; thalidomide; and/or lenalidomide
• Testicular: topoisomerase inhibitors ± platinum-based compounds
• Stomach: DNA intercalating agents and/or DNA cross-linking agents

Implication: even if a system practices the methylation matching, it may avoid the claim if the therapy step is not performed “according to the following Table” as written (or if the clinical action uses different therapy constructs not encompassed by the claim’s therapy list).

---

How broad are the methylation-profile claims in US 10,233,503?

Breadth drivers

• “Selected region” is not defined by sequence or length in claim 1 as provided. That can support wide coverage across different genomic targets as long as the method compares the same region between CUP and primary tumors.
• The detection is defined at the probe-per-CpG and ≥9 CpGs level, not by a specific chemistry (even though dependent claim 4 enumerates typical profiling methods).
• The diagnosis trigger is a match of methylation profile between CUP and reference primary tumor samples.

Potential narrowing constraints in practice

• The claim as given requires contacting DNA with probe(s) specific for each CpG site. If an accused method uses methylation detection that is not based on CpG-specific probes (for example, a platform that uses non-probe interrogation, or readout that is not implemented as probe contacts per CpG), literal infringement is less straightforward.
• “Same selected region” creates a dependence on method implementation: if the accused workflow compares a different locus set, or a different region definition, it may fall outside claim scope.

---

Which CpG-panel limitations matter most: “at least 9 CpGs” and region matching?

≥9 CpG sites is a hard quantitative floor

Every dependent “Tables 1A/1B … 15A/15B … 17 …” recites many CpG loci, but the independent claim only requires at least 9. That means an accused method with a 9+ CpG subset that maps to the same “selected region” and uses CpG-specific probes can still fall within claim 1 even if it does not use the exact panels in the specification tables.

Table-coded CpG lists create a blueprint but are not always claim-limiting

Your claim set includes extensive CpG tables by cancer type. Those tables likely inform the “selected region” and CpG site identity, but enforceability will depend on whether those tables are incorporated by reference into claim elements. In your provided claim text, claims 3 and following specify “CpG sites as defined in Tables X,” which is narrowing.

---

What dependent claims add key cancer-type and therapy constraints?

Claim 2: Primary tumor universe is broad but not infinite

Claim 2 limits primary tumor selection to a defined list, including:

• lymphoid neoplasia, head and neck cancer, pancreatic, endometrial, colon, prostate, glioma, ovarian, lung, bladder, melanoma, breast, myeloid neoplasia, testicular, stomach

Commercial implication: If a method assigns a primary tumor outside that list (for example, a sarcoma subtype), the claim may not cover the “diagnose-and-treat” workflow as written.

Claims 3, 5–17: CpG-site tables by cancer type

Claims specify detection of methylation statuses in CpG sites “as defined in Tables” for each cancer type (lymphoid, head and neck, pancreatic, endometrial, colon, prostate, glioma, ovarian, lung, bladder, melanoma, breast, testicular, stomach).

Legal implication: Those dependent claims are narrower and can be used to enforce against competitors that implement the exact CpG sites/panels aligned to those tables.

Claims 18–19: therapy specifics tighten infringement

• Claim 18: if lymphoid neoplasia is diagnosed, therapy is a CD20-targeted drug from a listed set (includes rituximab, ocrelizumab, ofatumumab, etc.).
• Claim 19: if pancreatic cancer is diagnosed, therapy is an antimetabolite from a listed set (includes methotrexate, pemetrexed, gemcitabine, 5-FU-related, etc.).

Enforcement leverage: If an accused product uses methylation matching plus CD20 therapy in CUP-like clinical use, claim 18 provides a clear therapy mapping anchor.

---

What methylation measurement methods are contemplated in the claim set?

Claim 4 lists method categories for methylation profiling

It provides a menu including:

• MSP; enrichment methods (MeDIP, MBD-seq, MethylCap); bisulfite-based methods (RRBS, bisulfite sequencing, Infinium, GoldenGate, COBRA, MSP, MethyLight); and other sequence-level approaches (MRE-seq restriction-digestion, differential-conversion, differential restriction).

Key point for claim construction: Claim 1’s probe-specific limitation is more likely to control infringement than the general measurement method list, because claim 4 is dependent. Still, if accused products use bisulfite readouts without CpG-specific probe contacting, they may not satisfy claim 1’s “probe specific for each CpG site” element.

---

How strong is US 10,233,503’s patent estate if challenged on novelty/obviousness?

Given only the claim text provided, the most actionable strength/weakness assessment is claim-feature driven:

• Likely vulnerability cluster: “methylation profiling for tumor-of-origin assignment” is an established research and diagnostic theme. Novelty often turns on the specific probe-based detection of defined CpG sets and region matching plus the CUP-specific reference matching logic and the integrated treatment table.
• Likely strength cluster: The claim is not merely “measure methylation and compare patterns.” It requires:
  • CUP vs primary tumor methylation profile matching on the same selected region
  • detection of ≥9 CpGs
  • CpG-site-specific probes contacting the DNA
  • and a specified diagnose then treat with a targeted therapy table.

Critical point: The more a competitor’s workflow collapses into “assess methylation and pick therapy,” the more the claim’s integrated structure matters. The more their workflow uses distinct loci definitions, fewer than 9 CpGs, lack of CpG-specific probes, or different matching logic (e.g., classifier score rather than “same methylation profile” identity), the stronger the validity/infringement defenses.

---

What would generic or competitor “CUP methylation assays” have to do to avoid infringement?

Based on the claim elements you provided, avoidance strategies would map to the claim’s required features:

1. Reduce CpG site count below 9 (if technically feasible while maintaining clinical performance).
2. Avoid CpG-specific probe contacting per site
   • Use readouts not implemented as “probes specific for each CpG site,” or implement capture/measurement in a way that does not meet that structural requirement.
3. Change the compared “selected region” concept
   • If the method does not compare the CUP methylation profile with a primary tumor reference profile on the “same selected region,” it may avoid a key limitation.
4. Separate diagnosis from the claimed treatment table
   • If therapy is not selected “according to the following Table” and uses different therapy decision frameworks not encompassed by those listed modalities for the diagnosed tumor type, the integrated claim 1 may be harder to hit.

---

Orange Book status, FDA status, and Paragraph IV biosimilar risks?

None. US 10,233,503 is a method patent for diagnostics and therapy selection rather than a small-molecule drug with an Orange Book listing, and it is not a biologic product patent with biosimilar exclusivity triggers. The enforcement and regulatory interaction is through device/IVD claims and clinical utility rather than Hatch-Waxman exclusivity.

---

Key litigation and licensing signals

None can be asserted from the information provided. Your prompt includes claims but does not include litigation dockets, settlements, or license history tied to US 10,233,503.

---

Key takeaways

• US 10,233,503 is centered on a CUP origin-assignment workflow using methylation profiling at ≥9 CpG sites within a selected DNA region, with CpG-specific probes and match-to-primary reference profiles.
• The claim is materially broadened by the absence of fixed sequence identifiers in claim 1, but materially constrained by (i) ≥9 CpGs, (ii) CpG-specific probe contacting, and (iii) same-region profile matching, plus the integrated therapy mapping.
• Dependent claims add enforceable specificity by cancer type: CpG lists tied to tables and therapy lists tied to diagnoses (CD20 drugs for lymphoid neoplasia and antimetabolites for pancreatic cancer).
• Competitor risk is highest where a commercial CUP methylation assay uses a probe-panel design (CpG-resolved) and then directly selects a matched targeted therapy according to the claim’s listed modalities for the assigned origin.

---

FAQs

1. What is the single biggest infringement lever in US 10,233,503?
The requirement to detect methylation at ≥9 CpG sites by contacting DNA with a plurality of CpG-specific probes and comparing the resulting methylation profile to a primary tumor profile from the same selected region.

2. Does the patent cover just diagnosis or diagnosis plus treatment?
Claim 1 requires both: diagnosis of primary tumor origin via methylation profile matching and then treatment targeted to that diagnosed primary tumor using the modalities in the table.

3. Can a competitor avoid claim 1 by using fewer CpGs?
If their method detects methylation for fewer than 9 CpG sites in the relevant selected region, claim 1’s quantitative threshold is not met.

4. Are the long CpG tables in the claims enforceable?
They are enforceable through the dependent claims where detection is limited to CpG sites defined in those tables (claims 3 and 5–17, plus table-specific variants).

5. Is this patent an Orange Book patent for a drug product?
No. It is a method patent tied to CUP methylation diagnostics and therapy selection, not a standard small-molecule Orange Book listing.

---

References

No sources are cited because the analysis is derived only from the claim text you provided and does not include external verification of the patent’s prosecution history, assignee, or any cited documents.

Details for Patent 10,233,503

Applicant	Tradename	Biologic Ingredient	Dosage Form	BLA	Approval Date	Patent No.	Expiredate
Genentech, Inc.	RITUXAN	rituximab	Injection	103705	26-Nov-97		2032-05-24
Novartis Pharmaceuticals Corporation	ARZERRA	ofatumumab	Injection	125326	26-Oct-09		2032-05-24
Novartis Pharmaceuticals Corporation	ARZERRA	ofatumumab	Injection	125326	1-Apr-11		2032-05-24
Novartis Pharmaceuticals Corporation	KESIMPTA	ofatumumab	Injection	125326	20-Aug-20		2032-05-24