US Patent 6,602,503 (Anti-alpha4 VCAM-1/VLA-4 humanized antibody): Claim-by-claim strength, design-around risk, and landscape position

What does US 6,602,503 actually claim?

US 6,602,503 is an antibody-centric patent that locks down a specific humanized anti-α4 (VLA-4 integrin / CD49d, plus VCAM-1/VLA-4 pathway context) format through (1) CDR grafting coordinates (Kabat positions), (2) explicit amino-acid constraints on selected framework positions, (3) explicit sequence identifiers (SEQ ID numbers) tying back to particular murine and human frameworks, and (4) full molecule variants, nucleic acids, vectors, expression systems, and use claims.

The claim set is unusually broad in coverage of formats (antibody; fragments; DNA; vectors; host cells; production processes; therapeutic/diagnostic compositions; treatment methods), but unusually narrow in sequence space because it repeatedly nails down:

Heavy chain CDRs: Kabat CDR1 31-35, CDR2 50-65, CDR3 95-102 from a mouse anti-α4 antibody.
Heavy chain framework residues: non-human residues at Kabat positions 27-30, specifically Phe27, Asn28, Ile29, Lys30 (SEQ ID NO: 107).
Light chain CDRs: Kabat CDR1 24-34, CDR2 50-56, CDR3 89-97 from the mouse anti-α4 antibody.
Light chain frameworks: derived from the REI monoclonal antibody (SEQ ID NO: 100) with explicit framework sequence constraints including Val104, Glu105, Lys107 and at least one non-human framework residue.
Optional heavy-chain additional framework exceptions: non-human residues at Kabat position 75 and/or additional positions listed in dependent claims.
Optional light-chain additional framework exceptions: non-human residues at Kabat positions 60 and 67.
Explicit murine CDR source: HP1/2 murine monoclonal antibody is named as the CDR donor (SEQ ID NO: 6 and SEQ ID NO: 10).
Named production ecosystem: an ATCC deposit is claimed: NSO cell line ATCC CRL 11175.
Named functional constraint: binding affinity range relative to HP1/2, defined by half-maximal binding constant, 20% to 100% of HP1/2.

The claims also cover assembled formats:

Full IgG-like bivalent format is explicitly enumerated in multiple places: two full-length heavy chains and two full-length light chains (claims 5, 18, 31, 37, 43, 49, 81, 87, 96, 98, 100, 102, 108, 110, 114, 116, 118, 120, 123, 125, 127, 129, 132, etc., depending on claim family).

Core independent claim coverage map

Independent claim anchors (as provided) include:

Claim 1 (antibody molecule with heavy+light constraints and specific framework position rules)
Claim 7 (DNA encoding the heavy chain with those constraints)
Claim 10 (DNA encoding the light chain with those constraints)
Claim 14 (DNA encoding antibody molecule with both heavy+light constraints)
Claim 20/24/27/33 (vector and expression vector equivalents)
Claim 39 (host cell transformed with first vector for heavy and second vector for light)
Claim 51/54/57/61/65/71 (processes for production, with the key constraints incorporated)
Claim 77/83 (therapeutic/diagnostic compositions that incorporate the antibody constraint set)
Claim 89 (treatment method using the specific antibody constraint set)
Claim 95 and later (humanized variants specified by allowed variable heavy and light variable region SEQ IDs)

How is the claim language structured for enforceability? (Strength and failure points)

The enforceability profile depends on whether an accused product lands inside the claim’s sequence-anchored architecture or can argue around it through:

different CDR sequences,
different framework backbones,
different heavy-light pairing architecture,
different Kabat position mapping assumptions,
different binding affinity characterization, and/or
different fragment class (Fab vs full IgG vs other formats).

1) Sequence-anchoring is strong

Claim 1’s heavy chain and light chain constraints repeatedly tie to:

Exact Kabat position ranges
Exact amino-acid identity at framework 27-30 (heavy)
Exact amino-acid identity at key light framework positions 104, 105, 107
A named light framework source (REI monoclonal; SEQ ID NO: 100)
A named murine CDR source (HP1/2; SEQ ID NO: 6 and SEQ ID NO: 10)

This structure is difficult to evade with “conventional” humanization variations. Many design-around attempts in antibody humanization change framework residues widely; this patent forces specific residues in defined Kabat coordinates.

2) “At least one” language broadens, but does not fully decouple

Claim 1 and its relatives use “at least one heavy chain” and “at least one light chain,” and later claims specify a two-heavy/two-light format. For infringement analysis, this means:

a maker using heterodimers or engineered subunit ratios might still fall under “at least one” if the subunits match the required sequences and CDRs/framework residue constraints.
but the requirement that the molecule is an “anti-α4-binding” humanized recombinant antibody with those constraints still binds the product identity.

3) Affinity range is a functional escape hatch

Claims 138 and 139 impose an affinity constraint:

binding affinity as defined by half maximal binding constant 20% to 100% of HP1/2

If a competitor can show their product’s binding is outside that window (below 20% or above 100% relative definition), it may create a non-literal design-around argument. In practice, “relative to half-maximal binding constant of HP1/2” is measurable, but relative measurement disputes can still trigger evidentiary fights.

4) Fragment coverage is incomplete

Claim 146 states: “.alpha.4-binding fragment is an Fab fragment.” That suggests at least some fragment types are encompassed. But the claim set as provided does not show a full panoply of engineered fragments (e.g., scFv, VHH, diabodies, multispecifics) with their own explicit structural definitions. A clean design-around could target formats not clearly covered by Fab or not meeting “antibody heavy chain + light chain” definitions.

What is the likely technology intent behind the claimed framework selections?

The patent uses a common humanization approach:

Take murine anti-α4 CDRs (HP1/2)
Graft them onto human framework regions
Preserve a small subset of murine framework positions thought to maintain affinity or proper CDR-HCDR/LCDR structure

Two explicit “must-match” framework patches stand out:

Heavy chain framework 27-30: fixed to Phe-Asn-Ile-Lys (SEQ ID NO: 107).
Light chain REI framework anchors: fixed motif includes Val104-Glu105-Lys107 and at least one additional non-human framework residue; optional additional LCDR framework residues are specified at 60 and 67.

This is the kind of design that can yield better affinity retention than swapping in a fully human consensus. The downside is it makes the patent easier to map against sequences because the allowed residues are not purely discretionary.

Where do the claims create predictable infringement hotspots?

Hotspot A: the specific CDR donor (HP1/2)

Multiple claims state that non-human CDRs are derived from HP1/2 murine monoclonal with SEQ IDs tied to CDRs. If a competitor keeps HP1/2 CDR sequences but changes frameworks, they still risk capture if the framework constraints are met.

Hotspot B: the heavy chain framework patch (27-30)

A maker can change most frameworks and still preserve CDR structure, but this patent hard-codes 27-30 for the heavy chain. Any competitor that preserves those residues to maintain affinity is likely inside.

Hotspot C: light chain framework anchors (REI + Val104-Glu105-Lys107)

The light chain is constrained by a named framework source (REI; SEQ ID NO: 100) plus explicit residues at 104/105/107. That combination is a strong discriminator.

Hotspot D: nucleic acid and expression system claims

The patent includes:

DNA encoding heavy and light chains
full antibody DNA constructs
vectors and expression vectors
host cell and transformation claims
processes for producing the antibody

This creates a layered enforcement path. Even if a product is distributed as a purified antibody, infringement can also be argued via manufacturing steps or by proving the existence of the claimed sequences in plasmids/constructs.

How broad are the later SEQ ID “humanized variants” claims?

Later claims (95 onward) specify allowed variable region combinations using SEQ ID enumerations:

Humanized heavy variable region allowed among SEQ ID NO: 43, 47, 51, 55
Humanized light variable region allowed among SEQ ID NO: 31, 63, 67
Then dependent claims constrain to the full two heavy + two light format.

Claim 107 gives a concrete paired combination:

Heavy variable region: V_H-AS (SEQ ID NO: 55)
Light variable region: VK2-SVMDY (SEQ ID NO: 63)

So the patent does not only cover the “CDR+framework patchwork” defined in Claim 1. It also covers a second layer that identifies particular “final humanized” variable regions by SEQ ID numbers.

That tends to reduce ambiguity in infringement because the SEQ ID variants become effectively sequence-licensed embodiments.

What is the likely competitive landscape pressure point: who else is playing in anti-α4 humanized antibodies?

In the anti-VLA-4/VCAM-1 space, the market has seen multiple antibodies (including widely known commercial products and developer pipelines). The key IP risk is not “anti-α4 antibodies exist,” but whether competitors’ humanized anti-α4 antibodies:

share HP1/2 CDR sequences,
use the REI light framework with the same anchored residues, and
maintain the specific heavy framework residues 27-30.

Without external claim sets and sequence mapping from other patents, the only defensible statement is structural: US 6,602,503 is anchored to named murine and human framework components plus fixed Kabat-residue identities. That is a narrow target for generic entry-by-humanization.

Design-around strategies: what can realistically avoid this patent’s claim scope?

Below are the main routes a competitor can use to reduce risk, based strictly on the claim text you provided.

1) Change the murine CDR donor (avoid HP1/2 CDR sequences)

Because the claims repeatedly identify the non-human CDRs as derived from HP1/2 (SEQ ID NO: 6 and SEQ ID NO: 10), using different CDR sequences reduces likelihood of literal infringement.

2) Keep the antigen-binding intent but alter the required framework residues

If the accused antibody does not have:

heavy framework Phe27-Asn28-Ile29-Lys30, or
light framework anchored residues including Val104-Glu105-Lys107 in the REI framework context, then it exits the explicit residue constraints of Claim 1 and its dependent claim set.

3) Change affinity positioning outside 20% to 100% relative window

Claim 138/139 define binding relative to HP1/2 by a half-maximal binding constant window. If a competitor can show binding is outside that window, it can attempt a non-infringement argument, depending on the evidence and how the “as defined” relationship is established in litigation.

4) Avoid formats not clearly captured

If a competitor uses a non-Fab fragment format or a non-standard antibody architecture, it can try to argue the fragment is not within the claim’s intended structural classification. The patent expressly mentions Fab (Claim 146), but it does not define other fragment types in what you provided.

5) Manufacturing route claims create additional coverage, but also additional proof burdens

The patent covers production processes with transfection of host cells using specific vectors and then culturing. A competitor can attempt:

different expression platforms,
different assembly pathways,
or different sequence constructs. But if they still use the claimed antibody sequences, DNA/vector/host claims are hard to avoid.

Critical reading of claim breadth: where the patent may overreach and where it is likely clean

Where the claims are “clean”

The claims provide explicit residue constraints and anchored framework sources.
They tie both antibody sequences and encoding sequences.
They cover both composition and method of use (inflammation associated with VCAM-1/VLA-4 pathway).

Where the claims are exposed to non-infringement arguments

The heavy and light framework constraints are strict. Competitors can redesign frameworks and still keep anti-α4 function, reducing capture probability if they do not preserve the exact anchored residues.
The binding affinity window introduces a measurable limit.

Where the claims could face validity pressure (conceptual, based on style)

Without the file history, prior art list, or prosecution record, only the claim structure can be judged. The claim set reads like a late-stage humanized embodiment with multiple dependent refinements and explicit SEQ ID enumerations. That often helps novelty and clarity versus purely functional “humanized antibody” claims, but it can also mean the breadth is capped tightly to specific embodiments. The analysis of validity (anticipation/obviousness) requires prior art sequences and earlier antibody humanization disclosures, which are not provided here.

US 6,602,503 claim set as an enforceability “ladder”

The patent effectively builds from: 1) Antibody identity (Claim 1 + dependent variants) 2) Genetic encoding (Claims 7, 10, 14) 3) Vectors and expression (Claims 20, 24, 27, 33) 4) Host cell / production process (Claims 39, 51, 54, 57, 61, 65, 71) 5) Commercial delivery formats (therapeutic and diagnostic compositions) 6) Clinical indication use (VCAM-1/VLA-4 inflammation) 7) Final humanized variable region sequence variants (Claim 95 onward by SEQ ID combinations) 8) Deposited cell line (Claim 121)

That ladder makes it difficult for a competitor to avoid “indirect” infringement by hiding behind formulation changes, since the DNA and manufacturing claims create additional attack surfaces.

Landscape implications for R&D and investment

1) If you are developing a humanized anti-α4 antibody

The primary decision is whether your lead candidate:

uses HP1/2 CDR sequences, and
uses the REI light framework with the anchored residues,
and maintains the heavy framework patch at 27-30.

If yes, your freedom-to-operate risk is elevated because the claims are highly specific. If no, the risk can drop sharply because the patent’s discriminators are sequence- and residue-based rather than generic “humanized anti-α4 antibodies.”

2) If you are evaluating licensing or entry timing

US 6,602,503’s strength sits in sequence anchoring and the inclusion of production-related claims. That reduces the practical utility of “wait for expiry” unless you model remaining patent term and any extensions or related family members, none of which are provided here.

3) If you are building an enforcement or clearance plan

Your analysis should focus on:

exact alignment to Kabat numbering ranges claimed for CDRs,
verification of heavy framework 27-30 residues (SEQ ID NO: 107),
verification of light framework residue identities including Val104-Glu105-Lys107 (REI; SEQ ID NO: 100 context),
matching of variable region SEQ IDs (43/47/51/55 heavy; 31/63/67 light; plus the specific pair in Claim 107),
and measurement placement if affinity window is a contested issue.

Key Takeaways

US 6,602,503 is highly sequence-anchored: it hard-codes specific Kabat frameworks (heavy 27-30; light framework anchors within REI) in addition to requiring HP1/2-derived CDRs.
Freedom-to-operate risk is not triggered by “anti-α4” alone. It is triggered when an accused antibody matches the specific CDR and framework constraints or the enumerated humanized variable SEQ ID combinations.
The patent’s enforceability ladder is strong because it spans antibody, DNA, vectors, host cells, production processes, and therapeutic/diagnostic uses.
Design-around is feasible if your candidate changes the HP1/2 CDR set and/or breaks the exact framework residue requirements; affinity window constraints add another potential non-infringement axis.
The strongest practical discriminator for clearance is residue-level matching, not just functional binding to α4.

FAQs

1) Does the patent claim any anti-α4 humanized antibody, or only specific engineered embodiments?
It claims specific embodiments defined by CDR position ranges plus explicit framework residue identities (including heavy 27-30 = Phe-Asn-Ile-Lys; light REI framework with Val104-Glu105-Lys107), and by SEQ ID-defined humanized variable regions.

2) Are the nucleic-acid and manufacturing claims likely to matter if we only sell the purified antibody?
Yes. The patent claims DNA, vectors, host cells, and production processes that encode or produce the defined antibody molecules, creating manufacturing-route and sequence-construct infringement angles.

3) What is the biggest design-around lever?
Breaking the framework residue requirements (heavy 27-30 and light REI anchored residues) and/or using non-HP1/2 CDR sequences.

4) Does binding affinity play a role beyond sequence?
Yes. Claims 138 and 139 impose an affinity window relative to HP1/2 by half-maximal binding constant (20% to 100%).

5) Is the antibody format limited to full IgG or does it include fragments?
The claims repeatedly include full-length two-heavy/two-light formats, and Claim 146 mentions Fab fragments. Other fragment architectures are not explicitly defined in what is provided.

References

[1] US Patent 6,602,503, “Humanized recombinant anti-α4 antibody” (claims excerpt provided in prompt).

Title:	Recombinant anti-VLA-4 antibody molecules
Abstract:	The present invention disclosed recombinant anti-VLA-4 antibody molecules, including humanized recombinant anti-VLA-4 antibody molecules. These antibodies are useful in the treatment of specific and non-specific inflammation, including asthma and inflammatory bowel disease. In addition, the humanized recombinant anti-VLA-4 antibodies disclosed can be useful in methods of diagnosing and localizing sites of inflammation.
Inventor(s):	Lobb; Roy R. (Westwood, MA), Carr; Frank J. (Balmedie, GB), Tempest; Philip R. (Royston, GB)
Assignee:	Biogen, Inc. (Cambridge, MA)
Application Number:	08/454,899
Patent Claims:	see list of patent claims
Patent landscape, scope, and claims summary:	US Patent 6,602,503 (Anti-alpha4 VCAM-1/VLA-4 humanized antibody): Claim-by-claim strength, design-around risk, and landscape position What does US 6,602,503 actually claim? US 6,602,503 is an antibody-centric patent that locks down a specific humanized anti-α4 (VLA-4 integrin / CD49d, plus VCAM-1/VLA-4 pathway context) format through (1) CDR grafting coordinates (Kabat positions), (2) explicit amino-acid constraints on selected framework positions, (3) explicit sequence identifiers (SEQ ID numbers) tying back to particular murine and human frameworks, and (4) full molecule variants, nucleic acids, vectors, expression systems, and use claims. The claim set is unusually broad in coverage of formats (antibody; fragments; DNA; vectors; host cells; production processes; therapeutic/diagnostic compositions; treatment methods), but unusually narrow in sequence space because it repeatedly nails down: Heavy chain CDRs: Kabat CDR1 31-35, CDR2 50-65, CDR3 95-102 from a mouse anti-α4 antibody. Heavy chain framework residues: non-human residues at Kabat positions 27-30, specifically Phe27, Asn28, Ile29, Lys30 (SEQ ID NO: 107). Light chain CDRs: Kabat CDR1 24-34, CDR2 50-56, CDR3 89-97 from the mouse anti-α4 antibody. Light chain frameworks: derived from the REI monoclonal antibody (SEQ ID NO: 100) with explicit framework sequence constraints including Val104, Glu105, Lys107 and at least one non-human framework residue. Optional heavy-chain additional framework exceptions: non-human residues at Kabat position 75 and/or additional positions listed in dependent claims. Optional light-chain additional framework exceptions: non-human residues at Kabat positions 60 and 67. Explicit murine CDR source: HP1/2 murine monoclonal antibody is named as the CDR donor (SEQ ID NO: 6 and SEQ ID NO: 10). Named production ecosystem: an ATCC deposit is claimed: NSO cell line ATCC CRL 11175. Named functional constraint: binding affinity range relative to HP1/2, defined by half-maximal binding constant, 20% to 100% of HP1/2. The claims also cover assembled formats: Full IgG-like bivalent format is explicitly enumerated in multiple places: two full-length heavy chains and two full-length light chains (claims 5, 18, 31, 37, 43, 49, 81, 87, 96, 98, 100, 102, 108, 110, 114, 116, 118, 120, 123, 125, 127, 129, 132, etc., depending on claim family). Core independent claim coverage map Independent claim anchors (as provided) include: Claim 1 (antibody molecule with heavy+light constraints and specific framework position rules) Claim 7 (DNA encoding the heavy chain with those constraints) Claim 10 (DNA encoding the light chain with those constraints) Claim 14 (DNA encoding antibody molecule with both heavy+light constraints) Claim 20/24/27/33 (vector and expression vector equivalents) Claim 39 (host cell transformed with first vector for heavy and second vector for light) Claim 51/54/57/61/65/71 (processes for production, with the key constraints incorporated) Claim 77/83 (therapeutic/diagnostic compositions that incorporate the antibody constraint set) Claim 89 (treatment method using the specific antibody constraint set) Claim 95 and later (humanized variants specified by allowed variable heavy and light variable region SEQ IDs) How is the claim language structured for enforceability? (Strength and failure points) The enforceability profile depends on whether an accused product lands inside the claim’s sequence-anchored architecture or can argue around it through: different CDR sequences, different framework backbones, different heavy-light pairing architecture, different Kabat position mapping assumptions, different binding affinity characterization, and/or different fragment class (Fab vs full IgG vs other formats). 1) Sequence-anchoring is strong Claim 1’s heavy chain and light chain constraints repeatedly tie to: Exact Kabat position ranges Exact amino-acid identity at framework 27-30 (heavy) Exact amino-acid identity at key light framework positions 104, 105, 107 A named light framework source (REI monoclonal; SEQ ID NO: 100) A named murine CDR source (HP1/2; SEQ ID NO: 6 and SEQ ID NO: 10) This structure is difficult to evade with “conventional” humanization variations. Many design-around attempts in antibody humanization change framework residues widely; this patent forces specific residues in defined Kabat coordinates. 2) “At least one” language broadens, but does not fully decouple Claim 1 and its relatives use “at least one heavy chain” and “at least one light chain,” and later claims specify a two-heavy/two-light format. For infringement analysis, this means: a maker using heterodimers or engineered subunit ratios might still fall under “at least one” if the subunits match the required sequences and CDRs/framework residue constraints. but the requirement that the molecule is an “anti-α4-binding” humanized recombinant antibody with those constraints still binds the product identity. 3) Affinity range is a functional escape hatch Claims 138 and 139 impose an affinity constraint: binding affinity as defined by half maximal binding constant 20% to 100% of HP1/2 If a competitor can show their product’s binding is outside that window (below 20% or above 100% relative definition), it may create a non-literal design-around argument. In practice, “relative to half-maximal binding constant of HP1/2” is measurable, but relative measurement disputes can still trigger evidentiary fights. 4) Fragment coverage is incomplete Claim 146 states: “.alpha.4-binding fragment is an Fab fragment.” That suggests at least some fragment types are encompassed. But the claim set as provided does not show a full panoply of engineered fragments (e.g., scFv, VHH, diabodies, multispecifics) with their own explicit structural definitions. A clean design-around could target formats not clearly covered by Fab or not meeting “antibody heavy chain + light chain” definitions. What is the likely technology intent behind the claimed framework selections? The patent uses a common humanization approach: Take murine anti-α4 CDRs (HP1/2) Graft them onto human framework regions Preserve a small subset of murine framework positions thought to maintain affinity or proper CDR-HCDR/LCDR structure Two explicit “must-match” framework patches stand out: Heavy chain framework 27-30: fixed to Phe-Asn-Ile-Lys (SEQ ID NO: 107). Light chain REI framework anchors: fixed motif includes Val104-Glu105-Lys107 and at least one additional non-human framework residue; optional additional LCDR framework residues are specified at 60 and 67. This is the kind of design that can yield better affinity retention than swapping in a fully human consensus. The downside is it makes the patent easier to map against sequences because the allowed residues are not purely discretionary. Where do the claims create predictable infringement hotspots? Hotspot A: the specific CDR donor (HP1/2) Multiple claims state that non-human CDRs are derived from HP1/2 murine monoclonal with SEQ IDs tied to CDRs. If a competitor keeps HP1/2 CDR sequences but changes frameworks, they still risk capture if the framework constraints are met. Hotspot B: the heavy chain framework patch (27-30) A maker can change most frameworks and still preserve CDR structure, but this patent hard-codes 27-30 for the heavy chain. Any competitor that preserves those residues to maintain affinity is likely inside. Hotspot C: light chain framework anchors (REI + Val104-Glu105-Lys107) The light chain is constrained by a named framework source (REI; SEQ ID NO: 100) plus explicit residues at 104/105/107. That combination is a strong discriminator. Hotspot D: nucleic acid and expression system claims The patent includes: DNA encoding heavy and light chains full antibody DNA constructs vectors and expression vectors host cell and transformation claims processes for producing the antibody This creates a layered enforcement path. Even if a product is distributed as a purified antibody, infringement can also be argued via manufacturing steps or by proving the existence of the claimed sequences in plasmids/constructs. How broad are the later SEQ ID “humanized variants” claims? Later claims (95 onward) specify allowed variable region combinations using SEQ ID enumerations: Humanized heavy variable region allowed among SEQ ID NO: 43, 47, 51, 55 Humanized light variable region allowed among SEQ ID NO: 31, 63, 67 Then dependent claims constrain to the full two heavy + two light format. Claim 107 gives a concrete paired combination: Heavy variable region: V_H-AS (SEQ ID NO: 55) Light variable region: VK2-SVMDY (SEQ ID NO: 63) So the patent does not only cover the “CDR+framework patchwork” defined in Claim 1. It also covers a second layer that identifies particular “final humanized” variable regions by SEQ ID numbers. That tends to reduce ambiguity in infringement because the SEQ ID variants become effectively sequence-licensed embodiments. What is the likely competitive landscape pressure point: who else is playing in anti-α4 humanized antibodies? In the anti-VLA-4/VCAM-1 space, the market has seen multiple antibodies (including widely known commercial products and developer pipelines). The key IP risk is not “anti-α4 antibodies exist,” but whether competitors’ humanized anti-α4 antibodies: share HP1/2 CDR sequences, use the REI light framework with the same anchored residues, and maintain the specific heavy framework residues 27-30. Without external claim sets and sequence mapping from other patents, the only defensible statement is structural: US 6,602,503 is anchored to named murine and human framework components plus fixed Kabat-residue identities. That is a narrow target for generic entry-by-humanization. Design-around strategies: what can realistically avoid this patent’s claim scope? Below are the main routes a competitor can use to reduce risk, based strictly on the claim text you provided. 1) Change the murine CDR donor (avoid HP1/2 CDR sequences) Because the claims repeatedly identify the non-human CDRs as derived from HP1/2 (SEQ ID NO: 6 and SEQ ID NO: 10), using different CDR sequences reduces likelihood of literal infringement. 2) Keep the antigen-binding intent but alter the required framework residues If the accused antibody does not have: heavy framework Phe27-Asn28-Ile29-Lys30, or light framework anchored residues including Val104-Glu105-Lys107 in the REI framework context, then it exits the explicit residue constraints of Claim 1 and its dependent claim set. 3) Change affinity positioning outside 20% to 100% relative window Claim 138/139 define binding relative to HP1/2 by a half-maximal binding constant window. If a competitor can show binding is outside that window, it can attempt a non-infringement argument, depending on the evidence and how the “as defined” relationship is established in litigation. 4) Avoid formats not clearly captured If a competitor uses a non-Fab fragment format or a non-standard antibody architecture, it can try to argue the fragment is not within the claim’s intended structural classification. The patent expressly mentions Fab (Claim 146), but it does not define other fragment types in what you provided. 5) Manufacturing route claims create additional coverage, but also additional proof burdens The patent covers production processes with transfection of host cells using specific vectors and then culturing. A competitor can attempt: different expression platforms, different assembly pathways, or different sequence constructs. But if they still use the claimed antibody sequences, DNA/vector/host claims are hard to avoid. Critical reading of claim breadth: where the patent may overreach and where it is likely clean Where the claims are “clean” The claims provide explicit residue constraints and anchored framework sources. They tie both antibody sequences and encoding sequences. They cover both composition and method of use (inflammation associated with VCAM-1/VLA-4 pathway). Where the claims are exposed to non-infringement arguments The heavy and light framework constraints are strict. Competitors can redesign frameworks and still keep anti-α4 function, reducing capture probability if they do not preserve the exact anchored residues. The binding affinity window introduces a measurable limit. Where the claims could face validity pressure (conceptual, based on style) Without the file history, prior art list, or prosecution record, only the claim structure can be judged. The claim set reads like a late-stage humanized embodiment with multiple dependent refinements and explicit SEQ ID enumerations. That often helps novelty and clarity versus purely functional “humanized antibody” claims, but it can also mean the breadth is capped tightly to specific embodiments. The analysis of validity (anticipation/obviousness) requires prior art sequences and earlier antibody humanization disclosures, which are not provided here. US 6,602,503 claim set as an enforceability “ladder” The patent effectively builds from: 1) Antibody identity (Claim 1 + dependent variants) 2) Genetic encoding (Claims 7, 10, 14) 3) Vectors and expression (Claims 20, 24, 27, 33) 4) Host cell / production process (Claims 39, 51, 54, 57, 61, 65, 71) 5) Commercial delivery formats (therapeutic and diagnostic compositions) 6) Clinical indication use (VCAM-1/VLA-4 inflammation) 7) Final humanized variable region sequence variants (Claim 95 onward by SEQ ID combinations) 8) Deposited cell line (Claim 121) That ladder makes it difficult for a competitor to avoid “indirect” infringement by hiding behind formulation changes, since the DNA and manufacturing claims create additional attack surfaces. Landscape implications for R&D and investment 1) If you are developing a humanized anti-α4 antibody The primary decision is whether your lead candidate: uses HP1/2 CDR sequences, and uses the REI light framework with the anchored residues, and maintains the heavy framework patch at 27-30. If yes, your freedom-to-operate risk is elevated because the claims are highly specific. If no, the risk can drop sharply because the patent’s discriminators are sequence- and residue-based rather than generic “humanized anti-α4 antibodies.” 2) If you are evaluating licensing or entry timing US 6,602,503’s strength sits in sequence anchoring and the inclusion of production-related claims. That reduces the practical utility of “wait for expiry” unless you model remaining patent term and any extensions or related family members, none of which are provided here. 3) If you are building an enforcement or clearance plan Your analysis should focus on: exact alignment to Kabat numbering ranges claimed for CDRs, verification of heavy framework 27-30 residues (SEQ ID NO: 107), verification of light framework residue identities including Val104-Glu105-Lys107 (REI; SEQ ID NO: 100 context), matching of variable region SEQ IDs (43/47/51/55 heavy; 31/63/67 light; plus the specific pair in Claim 107), and measurement placement if affinity window is a contested issue. Key Takeaways US 6,602,503 is highly sequence-anchored: it hard-codes specific Kabat frameworks (heavy 27-30; light framework anchors within REI) in addition to requiring HP1/2-derived CDRs. Freedom-to-operate risk is not triggered by “anti-α4” alone. It is triggered when an accused antibody matches the specific CDR and framework constraints or the enumerated humanized variable SEQ ID combinations. The patent’s enforceability ladder is strong because it spans antibody, DNA, vectors, host cells, production processes, and therapeutic/diagnostic uses. Design-around is feasible if your candidate changes the HP1/2 CDR set and/or breaks the exact framework residue requirements; affinity window constraints add another potential non-infringement axis. The strongest practical discriminator for clearance is residue-level matching, not just functional binding to α4. FAQs 1) Does the patent claim any anti-α4 humanized antibody, or only specific engineered embodiments? It claims specific embodiments defined by CDR position ranges plus explicit framework residue identities (including heavy 27-30 = Phe-Asn-Ile-Lys; light REI framework with Val104-Glu105-Lys107), and by SEQ ID-defined humanized variable regions. 2) Are the nucleic-acid and manufacturing claims likely to matter if we only sell the purified antibody? Yes. The patent claims DNA, vectors, host cells, and production processes that encode or produce the defined antibody molecules, creating manufacturing-route and sequence-construct infringement angles. 3) What is the biggest design-around lever? Breaking the framework residue requirements (heavy 27-30 and light REI anchored residues) and/or using non-HP1/2 CDR sequences. 4) Does binding affinity play a role beyond sequence? Yes. Claims 138 and 139 impose an affinity window relative to HP1/2 by half-maximal binding constant (20% to 100%). 5) Is the antibody format limited to full IgG or does it include fragments? The claims repeatedly include full-length two-heavy/two-light formats, and Claim 146 mentions Fab fragments. Other fragment architectures are not explicitly defined in what is provided. References [1] US Patent 6,602,503, “Humanized recombinant anti-α4 antibody” (claims excerpt provided in prompt). More… ↓ ⤷ Start Trial

Applicant	Tradename	Biologic Ingredient	Dosage Form	BLA	Approval Date	Patent No.	Expiredate
Genentech, Inc.	AVASTIN	bevacizumab	Injection	125085	February 26, 2004	⤷ Start Trial	2015-05-31
Biogen Inc.	TYSABRI	natalizumab	Injection	125104	November 23, 2004	⤷ Start Trial	2015-05-31
>Applicant	>Tradename	>Biologic Ingredient	>Dosage Form	>BLA	>Approval Date	>Patent No.	>Expiredate

Country	Patent Number	Estimated Expiration
World Intellectual Property Organization (WIPO)	9416094	⤷ Start Trial
United States of America	8226950	⤷ Start Trial
United States of America	7829092	⤷ Start Trial
United States of America	7482003	⤷ Start Trial
>Country	>Patent Number	>Estimated Expiration

Patent: 6,602,503

Summary for Patent: 6,602,503