Patent: 8,889,834

Summary for Patent: 8,889,834

Title:	Antigen binding proteins to proprotein convertase subtilisin kexin type 9 (PCSK9)
Abstract:	Antigen binding proteins that interact with Proprotein Convertase Subtilisin Kexin Type 9 (PCSK9) are described. Methods of treating hypercholesterolemia and other disorders by administering a pharmaceutically effective amount of an antigen binding protein to PCSK9 are described. Methods of detecting the amount of PCSK9 in a sample using an antigen binding protein to PCSK9 are described.
Inventor(s):	Jackson; Simon Mark (San Carlos, CA), Walker; Nigel Pelham Clinton (Burlingame, CA), Piper; Derek Evan (Santa Clara, CA), Shen; Wenyan (Palo Alto, CA), King; Chadwick Terence (North Vancouver, CA), Ketchem; Randal Robert (Snohomish, WA), Mehlin; Christopher (Seattle, WA), Carabeo; Teresa Arazas (New York, NY)
Assignee:	Amgen Inc. (Thousand Oaks, CA)
Application Number:	14/260,985
Patent Claims:	see list of patent claims

United States Patent 8,889,834 (PCSK9 Monoclonal Antibodies): Claim Validity Signals and Competitive Patent Landscape

US 8,889,834 is directed to monoclonal antibodies that bind PCSK9 at epitopes defined by residue sets drawn from SEQ ID NO: 1 and that block PCSK9 binding to the LDL receptor (LDLR). The claims tightly constrain: (i) the epitope composition (at least three or at least four specified residues), (ii) antibody identity (human; neutralizing), and (iii) binding performance (a KD threshold and pH condition in dependent claims). Several dependent claim limitations anchor the patent to specific sequence embodiments (e.g., light chain SEQ ID NO: 157; competition with an antibody defined by heavy-chain variable region SEQ ID NO: 49 and light-chain variable region SEQ ID NO: 23).

What do the claims actually cover?

How broad is independent claim 1?

Claim 1 covers an isolated monoclonal antibody that:

• binds an epitope of PCSK9 comprising at least three of the following amino acid residues of SEQ ID NO: 1: 207, 208, 162, 164, 167, 132, 351, 390, 413, 123, 129, 311, 313, or 337
• and blocks PCSK9 binding to LDLR.

Key breadth mechanics:

• Epitope definition uses residue inclusion logic: “at least three of” a defined list. This creates a combinatorial range of possible epitopes.
• No full sequence requirement appears in claim 1 beyond the residue list and LDLR blocking function.
• No antibody sequence requirement appears in claim 1, so coverage is functional to an extent (epitope + blocking).

Practical effect:

• The claim is broad in antibody identity but narrow in antigen binding site characterization. It is a classic “epitope-based functional” construct: broad across antibody formats, narrow on where on PCSK9 the binding must occur.

What do dependent claims narrow?

Claims 2–20 add layered constraints that reduce scope to specific human/neutralizing antibodies, specific variable/light-chain sequences, kinetic thresholds, and competitive-binding relationships.

Human and neutralizing limitations

• Claim 2: antibody is human.
• Claim 4: antibody is neutralizing.
• Claim 11–20 largely parallel claims 1–10 with “at least four residues” and additional specificity.

Specific light-chain sequence anchor

• Claim 3 and Claim 5/13: antibody comprises a light chain with amino acid sequence SEQ ID NO: 157.

This is a strong coverage anchor. Infringers can avoid literal coverage by using a different light chain sequence, even if the epitope and LDLR-blocking functional behavior match.

Affinity and assay-condition constraints

• Claim 6: KD ≤ 5 × 10^-9 M (5 nM).
• Claim 7: KD measured at pH < 7.4.
• Claims 15–16 repeat the KD threshold and pH condition.

These constraints are high-precision. If a competitor’s antibody binds with KD above the threshold or under a different pH condition, literal infringement risk changes. If prosecution history estopped broader interpretation of the KD term (not provided here), equivalents could be constrained.

Epitope multiplicity escalation

• Claim 1: at least three residues.
• Claim 19: at least four residues (same residue universe).
• Claim 20: epitope contains at least one of a subset: 162, 164, 167, 207, 208.

This set of claim variants implies the patentee sought coverage around both:

• a minimal epitope footprint (3 residues)
• and a tighter footprint (4 residues)
• plus a separate subset claim tied to a partial overlap region.

Functional vs structural epitope labels

• Claim 9: epitope is a functional epitope.
• Claim 10: epitope is a structural epitope.

Those labels can be used to argue different scientific definitions of “functional” and “structural” epitopes. Without the specification text, the risk is that these terms could be construed as either:

• scientific categories that do not limit binding beyond the residue pattern, or
• meaningful limitations that require particular evidence (e.g., competition with LDLR, mutation impact, conformational dependence). The residue list already limits epitope composition, but these dependent claims add extra interpretive hooks.

Competition with a specific prior-art-like antibody embodiment

• Claim 17 and Claim 18: antibody “competes for binding to PCSK9” with an antibody comprising:
  • heavy chain variable region SEQ ID NO: 49
  • light chain variable region SEQ ID NO: 23

This is an important landscape signal: the patentee identified a specific binding competitor/benchmark and tied claim scope to antibodies that compete with it. That can be used in infringement/validity arguments to show overlapping binding sites (or lack thereof).

Critical reading of key claim terms (where challenges tend to cluster)

What are the most vulnerability points in claim scope?

1. Epitope residue list definition

   • The claim defines the epitope through selected residue positions on PCSK9 (SEQ ID NO: 1).
   • Validity challenges often target whether those residues provide a sufficiently definite and enablement-consistent description of the binding site across all embodiments. If the specification does not provide a clear mapping between the residue set and binding mode, the claims can be attacked for indefiniteness or inadequate enablement.

2. “At least three of” residue inclusion

   • This logic creates potentially many epitopes that include the same residues in various combinations, while allowing other residues to vary.
   • If the patent does not demonstrate that antibody binding across all such combinations yields LDLR blocking, the functional coupling (blocks LDLR binding) can become a target for non-enabled breadth arguments.

3. Functional epitope vs structural epitope

   • These terms can invite construction disputes because “functional” and “structural” are not always rigidly defined unless the specification uses a specific test or definition.
   • If claim language depends on post-hoc scientific classification rather than a test with clear thresholds, it can raise definiteness risk.

4. KD and pH dependence

   • KD is an assay-dependent parameter. The claim locks the pH condition to “pH < 7.4,” which reduces ambiguity compared to a generic KD, but does not define the assay method, buffer composition, temperature, or measurement platform (unless in the specification).
   • Enforcement becomes fact- and method-intensive.

5. Specific sequence dependence (SEQ ID NO: 157, SEQ ID NO: 49/23)

   • Sequence-anchored limitations are harder to design around than residue-based constraints, but also easier to invalidate if the underlying sequence embodiments were already disclosed elsewhere.
   • If those sequences were published in earlier patents or public databases, novelty and obviousness risks increase.

Is there redundancy you can exploit?

Yes. Claims 1 and 11 are substantively the same structure; 1 focuses on “at least three residues,” 11 uses similar phrasing; 19 expands to “at least four residues”; 20 uses a subset of residues and the human neutralizing framing.

This redundancy can strengthen enforceability via multiple claim paths, but it can also:

• broaden internal coverage logic enough to invite validity scrutiny on enablement across ranges,
• and make it easier to find non-infringing designs if a competitor avoids a specific residue footprint or avoids LDLR blocking.

The patent landscape implications (strategic positioning)

How does this patent fit the PCSK9 antibody field?

PCSK9 antagonism by monoclonal antibodies is well established in the US market. The claim structure indicates the patentee is targeting antibodies that bind a PCSK9 region proximal to LDLR interaction sites, expressed via a specific residue set from SEQ ID NO: 1, and requiring LDLR binding blockade.

This claim design suggests a landscape where:

• many competitors can be forced into design-around by changing epitope footprint
• or by altering antibody variable sequences such that the specific dependent claims (SEQ ID NO: 157; competition with SEQ ID NO: 49/23) do not read on their molecules.

What does the competition limitation signal about overlap risk?

Claims 17–18 incorporate a specific competitive relationship:

• The accused antibody must compete with an antibody defined by heavy-chain variable SEQ ID NO: 49 and light-chain variable SEQ ID NO: 23.

This shifts the inquiry from “does it bind the same residue set?” to “does it share a binding site / overlap enough to compete.” In litigation, that is often measured by competition assays. For freedom-to-operate, this creates a practical test:

• if a candidate antibody does not compete with that benchmark antibody, it may avoid these dependent claims even if it binds within the residue set.

From a portfolio strategy view, this limitation can compress the infringement surface to a smaller subset of binding modes.

What does the KD threshold strategy imply?

Claims 6–8 and 15–16 embed a quantitative performance threshold (KD ≤ 5 nM, measured at pH < 7.4). That tends to:

• reduce risk that weak binders outside the patentee’s disclosed performance read on the patent,
• but also increase the likelihood that enforcement depends on precise experimental characterization.

From a patentability view, if the specification demonstrates these KD values for the claimed antibodies, the KD limitations can help defend against prior-art antibodies with weaker affinities.

Claim-by-claim coverage map (what a competitor must do to avoid)

Which elements are “hard to avoid” vs “easy to vary”?

Harder to avoid (high constraint):

• Epitope residue footprint (at least 3 or at least 4 of the listed residues)
• LDLR blocking (functional requirement)
• Human and/or neutralizing (dependent claims)

Easier to vary (design-around levers):

• Antibody light chain sequence (SEQ ID NO: 157) for dependent claims 3/5/13/6 line
• Assay-specific KD value and pH measurement conditions (for dependent claims 6–8 and 15–16)
• Competition behavior relative to the antibody defined by SEQ ID NO: 49 and SEQ ID NO: 23 (for dependent claims 17–18)
• For claim 20: residue subset coverage (presence of at least one of 162/164/167/207/208) still allows exclusion by altering the binding footprint away from that subset (while still possibly binding other residues in claim 1/11 if pursued)

How coverage changes between “3 residues” and “4 residues”?

• Claim 1/11: at least three residues from the list.
• Claim 19: at least four residues from the list.

A competitor that binds a smaller footprint that includes exactly three residues (or includes three but not four) could avoid claim 19 while still potentially intersecting claim 1/11 if LDLR blocking is met.

Likely litigation and prosecution dynamics

What would be the strongest infringement theories?

For infringement of claim 1/11/20, an asserted antibody must show:

1. binding to PCSK9 at an epitope that includes the required residue subset (3 or 1+subset)
2. LDLR binding blockage.

Enforcement-ready evidence typically includes:

• epitope mapping (e.g., alanine scanning, peptide mapping, hydrogen-deuterium exchange, structural modeling with residue involvement)
• functional LDLR competition or inhibition assays showing PCSK9-LDLR blockade.

What would be the strongest defenses?

For claim 1/11:

• shift the epitope footprint such that the antibody binds PCSK9 but does not include enough of the listed residues
• or does not block PCSK9 binding to LDLR under the claimed conditions.

For dependent claims:

• avoid “human” status (less likely for therapeutics in this space)
• avoid neutralization classification (more likely to be contested)
• change light chain sequence away from SEQ ID NO: 157
• show KD above the threshold under the claimed pH conditions
• show non-competition vs the benchmark antibody defined by SEQ ID NO: 49/23.

Key takeaways

Key Takeaways

• Core scope is an epitope-and-function construct: PCSK9 antibodies binding epitopes defined by specific residue sets from SEQ ID NO: 1 and that block PCSK9-LDLR binding.
• Breadth vs constraint splits cleanly: independent claims use “at least three residues” and LDLR blocking, while dependent claims add human, neutralizing, specific light-chain sequence (SEQ ID NO: 157), KD ≤ 5 nM at pH < 7.4, and competition versus a specifically defined antibody (SEQ ID NO: 49/23).
• Design-around most plausibly targets the residue footprint size/composition (3 vs 4 residues), the SEQ ID NO: 157 light chain dependence, and the competitive binding requirement to the SEQ ID NO: 49/23 antibody.
• Enforcement risk concentrates on epitope mapping methodology and assay-defined LDLR blockade and KD/pH measurement conditions.

FAQs

1) What is the central claim limitation that defines the antibody epitope?

The epitope must include at least three (claim 1/11) or at least four (claim 19) of a specified set of PCSK9 residues from SEQ ID NO: 1, including positions such as 207/208 and 162/164/167.

2) Does the patent require a specific antibody sequence in the broadest claim?

No. Claim 1 is epitope- and function-defined and does not require the antibody to contain a particular variable or light-chain sequence. Sequence dependence appears in dependent claims (e.g., SEQ ID NO: 157).

3) Which dependent claims could be avoided by changing the antibody’s variable regions?

Claim 3/5/13 depend on the light chain containing SEQ ID NO: 157. Changing the light chain sequence can avoid those dependent claims even if the epitope and LDLR blockade requirements still hold.

4) How does the KD limitation affect freedom-to-operate for higher-affinity antibodies?

If a competitor meets KD ≤ 5 nM at pH < 7.4, it stays within the performance constraint of the dependent claims. If KD is higher or measured under different conditions, literal read can fall away, shifting the analysis toward epitope/function claims.

5) What does the “competes for binding” language add to the landscape?

Claims 17–18 require competition with an antibody defined by heavy chain variable SEQ ID NO: 49 and light chain variable SEQ ID NO: 23, tightening infringement to antibodies that bind an overlapping or functionally competing site.

References

No sources were provided in the prompt for the text, specification details, prosecution history, or the broader US patent family. Therefore, no citations can be reliably produced.

Details for Patent 8,889,834

Applicant	Tradename	Biologic Ingredient	Dosage Form	BLA	Approval Date	Patent No.	Expiredate
Amgen Inc.	REPATHA	evolocumab	Injection	125522	August 27, 2015	8,889,834	2034-04-24
Amgen Inc.	REPATHA	evolocumab	Injection	125522	July 08, 2016	8,889,834	2034-04-24

International Patent Family for US Patent 8,889,834

Country	Patent Number	Estimated Expiration
World Intellectual Property Organization (WIPO)	2009026558
United States of America	9920134
United States of America	9493576
United States of America	9056915