When do biologic drug patents expire and when will biosimilars launch?

United States Patent 10,398,781: What Claims Cover and How Tight the Landscape Is

United States Patent 10,398,781 claims insulin conjugates where insulin is linked to one or more carbohydrate ligands selected from AEG, AEM, AEBM, AETM, and beta-aminoethyl-N-acetylglucosamine (AEGA). The core novelty in claim scope is not just the ligand selection but the combination of (i) an “indirect” conjugation via a “conjugate framework” and (ii) defined insulin conjugation sites (A1, B1, Lys(B29), Lys(B3)) and (iii) branched/multivalent architectures with multiple ligands on one framework and multiple conjugation points.

The practical consequence for freedom-to-operate (FTO) and infringement risk is that the claims are drafted broad on “framework” and “branching,” yet narrow on the specific ligand list and the insulin attachment residues/positions. The landscape risk is concentrated in other insulin conjugates that use the same carbohydrate ligands (or close equivalents) and in any patents that teach branched multivalent “framework” constructs that indirectly link insulin to those ligands.

What Is Claimed in US 10,398,781 (Claim-by-Claim Scope Map)

1. What ligands are inside the claim boundary?

Claim 1 limits ligands to five carbohydrate amines:

AEG (aminoethylglucose)
AEM (aminoethylmannose)
AEBM (aminoethylbimannose)
AETM (aminoethyltrimannose)
AEGA (β-aminoethyl-N-acetylglucosamine)

This list is the first gating item. If a product uses a different targeting moiety (different sugar, different linker chemistry, or different terminal derivatization that falls outside these named ligands), claim 1 (and all dependent claims anchored to claim 1) generally do not read.

2. What does “indirect conjugation via a conjugate framework” do to claim reach?

Claim 1 requires the ligand(s) are “indirectly conjugated” to insulin “via a conjugate framework.” That language typically covers designs where:

insulin is linked to a scaffold,
scaffold displays ligand handles at one or more positions, and
the ligand does not directly attach to insulin’s functional group without the framework intervention.

From a design-around standpoint, the key is whether an accused product can argue ligand attachment is direct (no “framework” between) or that the scaffold is not a “conjugate framework” as construed by the patent.

3. Where on insulin must the attachment occur?

Claim 2 constrains insulin conjugation to at least one of the following residues:

A1 amino acid residue
B1 amino acid residue
epsilon-amino group of Lys(B29)
epsilon-amino group of Lys(B3)

Dependent claim strategy then expands to dual and multiple frameworks, but claim 2 is a major narrowing lens. If a competitor attaches through another residue (or through a glyco- or side-chain route outside these residues), claim 2 and all dependent claims requiring those sites are vulnerable.

4. How do the multivalent and branched claims expand coverage?

A cluster of claims focuses on multi-ligand and branched presentation:

Claim 3: insulin conjugated to two or more separate ligands
Claim 4: two or more ligands conjugated to a single conjugate framework also conjugated to insulin
Claim 5: ligands and insulin each located on separate branch(es) of a single branched framework
Claim 6: insulin and ligands at termini of separate branches of the branched framework
Claim 7: two or more ligands attached via two or more different conjugation points
Claims 8 to 12: specify combinations using insulin through defined residues plus two separate branched frameworks, each bearing two ligands at branch termini
Claims 13 to 17: specify that the “two or more separate ligands” are the listed ligand(s) (AEG, AEM, AEBM, AETM, AEGA) within claim 3

5. Which insulin products are explicitly named?

Claims 18 to 20 list insulin identities:

Human/porcine/bovine insulin (claim 18)
Insulin lispro, insulin aspart, insulin glulisine (claim 19)
Insulin glargine or insulin detemir (claim 20)

These are explicit “insulin molecules selected from” groups. A competitor using another insulin analog not listed would likely avoid those dependent claims. Independent claim 1 still covers “an insulin molecule” conjugate, but dependent claims add listed insulin types.

Critical Claim Construction Risks (Where Infringement and Validity Pivot)

1. “Conjugate framework” and “indirectly conjugated” drive both infringement and validity

The phrase “indirectly conjugated … via a conjugate framework” can be attacked as vague or overbroad if the specification does not define a consistent framework architecture that maps to prior art multicomponent conjugates. In practice, however, patents like this often survive because “framework” is implemented through a scaffold or linker that functions as a connector between insulin and ligand.

Infringement consequence: accused products must likely show that ligand attachment occurs through an intermediate scaffold rather than a direct single-step functional group coupling to insulin’s targeted residues.

Validity consequence: broad “framework” drafting can pull in prior art combinations where insulin is linked to a scaffold that then displays carbohydrate ligands. If the prior art already taught the same ligand list and a scaffold-based multivalent architecture, claim novelty could narrow to the insulin attachment residue locations and branching topology.

2. Insulin residue limitation is the main defensive feature

Claim 2 restricts to A1, B1, Lys(B29), Lys(B3). These are specific and not interchangeable with random conjugation sites.

Design-around strategies (high level):

attach to other residues,
attach via terminal chemistry that avoids those exact lysine epsilon-amino or A1/B1 positions,
or use a conjugation method that changes which insulin functional group is linked.

Infringement consequence: claim coverage becomes fragile if the claim chart cannot establish the attachment chemistry maps to those exact residues.

3. Branching topology claims are “structure-heavy” and can be easier to avoid

Claims 4 to 6 and 10 to 12 require:

one framework vs multiple frameworks,
branched frameworks,
ligands and insulin placed on separate branches,
and in claim 6, that they are on termini.

Such structure claims can be designed around by:

using linear linkers,
using clustered ligands on a single chain without branching,
or positioning insulin and ligands on the same branch.

Validity consequence: branching and multivalency are common in carbohydrate conjugate technologies. If prior art taught multivalent branched carbohydrate scaffolds, then the patent may stand or fall based on whether the prior art also matched the insulin residue limitation and the specific ligand list.

4. Multi-ligand via “different conjugation points” narrows technical overlap

Claim 7 adds “two or more different conjugation points.” This can prevent literal reading if a competitor uses identical conjugation points for all ligand attachments or if the architecture is multivalent but assembled via repeated identical attachment positions.

Where the Landscape Risk Likely Clusters (Based on Claim Dial Settings)

Because the claims are constrained to a specific ligand set and a specific insulin attachment residue set, the strongest landscape overlap tends to come from patents in: 1) insulin conjugates with carbohydrate targeting ligands (AEG/AEM/AEBM/AETM/AEGA),
2) scaffold-based or “framework” linking to connect insulin to carbohydrate ligands, and
3) branched multivalent carbohydrate architectures.

The highest risk scenario for infringement/FTO arises when a prior art patent already teaches all of these simultaneously:

the same five ligand types (or close naming-matching equivalents),
indirect scaffold presentation, and
attachment to insulin at the same specified residues.

A second tier risk scenario arises when prior art matches:

ligand set + branched multivalent scaffolds, but differs in insulin attachment site; in that case, claim 2 and dependent residue-specific claims become the key differentiator.

Criticality of Dependent Claim Ladder (How Coverage Expands)

The claims build from a general conjugate to increasingly structured architectures:

Claim 1 sets ligand universe and framework requirement.
Claim 2 sets insulin residue attachment options.
Claims 3 to 7 add multivalent ligand requirements, including a single framework vs branched frameworks and multi-point attachment.
Claims 8 to 12 add specific insulin attachment patterning (A1 and Lys(B29) combinations), plus dual frameworks with ligands on separate branches and termini.
Claims 13 to 17 lock the multiligand selection to one or more named ligands.
Claims 18 to 20 lock the insulin identity to named products.

This means an accused product can avoid dependent claims by modifying:

ligand identity (outside the five named ligands),
number of ligands (single ligand rather than two or more),
framework structure (linear rather than branched, or branching not consistent with claims),
conjugation residue (not A1/B1/Lys(B29)/Lys(B3)),
insulin identity (use an analog not listed).

Business-Grade Infringement Tests (What Must Be Shown)

To establish literal infringement against independent claim 1 and dependent claims, the allegation must cover:

Ligand identity
The product must use one or more ligands in the specified set: AEG, AEM, AEBM, AETM, AEGA.
Indirection and framework
The ligand must be “indirectly conjugated” to insulin via a “conjugate framework,” not direct coupling.
Insulin attachment residues (for dependent claims that require claim 2)
The insulin attachment must involve A1 or B1 or Lys(B29) epsilon-amino or Lys(B3) epsilon-amino.
Multivalency and branching (for dependent claims)
For claims 3 to 7 and 10 to 12: the architecture must show multiple separate ligands, plus single vs branched frameworks, plus positioning at branch termini.
Insulin identity (for claims 18 to 20)
If the insulin type is not in the named group, only claim 1 and residue-agnostic coverage might be relevant.

Validity Pressure Points (Where Prior Art Typically Attacks)

A critical view of patentability for this kind of claim stack focuses on:

Carbohydrate ligand novelty vs scaffold novelty
If AEG/AEM/AEBM/AETM/AEGA insulin conjugates were already known, the main novelty must come from how the insulin is connected and how branching/multivalency is engineered.
Insulin site conjugation as the differentiator
If prior art already used those same residues (A1, B1, Lys(B29), Lys(B3)) for conjugation, the claim likely faces stronger obviousness attacks.
Branched frameworks and multivalent conjugates as a known architecture class
If branched multivalent carbohydrate presentation with insulin is known, the claim’s structural limitations can still be obvious if they are routine variations.
Overbreadth in “framework” language
If the specification supports only a narrow set of frameworks but the claims cover broader scaffold forms, that gap can create enablement or written description issues. Whether that plays out depends on the specification, which is not provided here.

Key Takeaways

US 10,398,781 claims insulin conjugates that combine a specific ligand set (AEG/AEM/AEBM/AETM/AEGA) with scaffold-based indirect conjugation and, in dependent claims, specific insulin attachment residues (A1, B1, Lys(B29), Lys(B3)).
The strongest infringement risk is for products that meet all of: ligand identity, indirect framework linking, multivalent display, and branching/termini architecture.
The tightest design-around lever is the insulin attachment site. If conjugation does not involve the claim-listed residues, much of the dependent claim ladder falls away.
The next most important lever is architecture (branched vs linear, single framework vs multiple frameworks, ligands on separate branches/termini).
Any landscape overlap that already teaches AEG/AEM/AEBM/AETM/AEGA insulin conjugates with multivalent branched frameworks is likely the main competitive threat; overlap that differs only by insulin attachment site tends to narrow relevance to dependent claims.

FAQs

1. What does “indirectly conjugated … via a conjugate framework” typically cover?

It generally covers architectures where the ligand is attached through an intervening scaffold rather than a direct bond to insulin’s functional group, implying multi-component assembly in which insulin and ligand are linked through a connecting framework.

2. Which single claim element is most critical for design-around?

The insulin attachment positions in claim 2: A1, B1, Lys(B29) epsilon-amino, Lys(B3) epsilon-amino. If the product attaches through other residues, dependent claim coverage drops substantially.

3. Do the claims require a specific insulin manufacturing form?

The patent depends on the insulin “molecule” identity in dependent claims (human/porcine/bovine; lispro/aspart/glulisine; glargine/detemir), but claim 1 uses broader “insulin molecule” language.

4. Does multivalency matter for claim coverage?

Yes. Claim 3 and dependent branching claims require two or more separate ligands. A single-ligand conjugate may avoid those dependent claims.

5. Are branched frameworks a strict requirement?

For specific dependent claims (4 to 6 and 10 to 12), yes. Those claims require branched topology and branch placement at termini, which can be engineered around by using linear linkers or non-matching branching placement.

References

[1] United States Patent 10,398,781.

Title:	Conjugate based systems for controlled drug delivery
Abstract:	Conjugates which comprise a drug and a ligand which includes a first saccharide; wherein the conjugate is characterized in that, when the conjugate is administered to a mammal, at least one pharmacokinetic or pharmacodynamic property of the conjugate is sensitive to serum concentration of a second saccharide. Exemplary conjugates and sustained release formulations are provided in addition to methods of use and preparation.
Inventor(s):	Zion; Todd C. (Marblehead, MA), Lancaster; Thomas M. (Wenham, MA)
Assignee:	SMARTCELLS, INC. (Kenilworth, NJ)
Application Number:	15/680,262
Patent Claims:	see list of patent claims
Patent landscape, scope, and claims summary:	United States Patent 10,398,781: What Claims Cover and How Tight the Landscape Is United States Patent 10,398,781 claims insulin conjugates where insulin is linked to one or more carbohydrate ligands selected from AEG, AEM, AEBM, AETM, and beta-aminoethyl-N-acetylglucosamine (AEGA). The core novelty in claim scope is not just the ligand selection but the combination of (i) an “indirect” conjugation via a “conjugate framework” and (ii) defined insulin conjugation sites (A1, B1, Lys(B29), Lys(B3)) and (iii) branched/multivalent architectures with multiple ligands on one framework and multiple conjugation points. The practical consequence for freedom-to-operate (FTO) and infringement risk is that the claims are drafted broad on “framework” and “branching,” yet narrow on the specific ligand list and the insulin attachment residues/positions. The landscape risk is concentrated in other insulin conjugates that use the same carbohydrate ligands (or close equivalents) and in any patents that teach branched multivalent “framework” constructs that indirectly link insulin to those ligands. What Is Claimed in US 10,398,781 (Claim-by-Claim Scope Map) 1. What ligands are inside the claim boundary? Claim 1 limits ligands to five carbohydrate amines: AEG (aminoethylglucose) AEM (aminoethylmannose) AEBM (aminoethylbimannose) AETM (aminoethyltrimannose) AEGA (β-aminoethyl-N-acetylglucosamine) This list is the first gating item. If a product uses a different targeting moiety (different sugar, different linker chemistry, or different terminal derivatization that falls outside these named ligands), claim 1 (and all dependent claims anchored to claim 1) generally do not read. 2. What does “indirect conjugation via a conjugate framework” do to claim reach? Claim 1 requires the ligand(s) are “indirectly conjugated” to insulin “via a conjugate framework.” That language typically covers designs where: insulin is linked to a scaffold, scaffold displays ligand handles at one or more positions, and the ligand does not directly attach to insulin’s functional group without the framework intervention. From a design-around standpoint, the key is whether an accused product can argue ligand attachment is direct (no “framework” between) or that the scaffold is not a “conjugate framework” as construed by the patent. 3. Where on insulin must the attachment occur? Claim 2 constrains insulin conjugation to at least one of the following residues: A1 amino acid residue B1 amino acid residue epsilon-amino group of Lys(B29) epsilon-amino group of Lys(B3) Dependent claim strategy then expands to dual and multiple frameworks, but claim 2 is a major narrowing lens. If a competitor attaches through another residue (or through a glyco- or side-chain route outside these residues), claim 2 and all dependent claims requiring those sites are vulnerable. 4. How do the multivalent and branched claims expand coverage? A cluster of claims focuses on multi-ligand and branched presentation: Claim 3: insulin conjugated to two or more separate ligands Claim 4: two or more ligands conjugated to a single conjugate framework also conjugated to insulin Claim 5: ligands and insulin each located on separate branch(es) of a single branched framework Claim 6: insulin and ligands at termini of separate branches of the branched framework Claim 7: two or more ligands attached via two or more different conjugation points Claims 8 to 12: specify combinations using insulin through defined residues plus two separate branched frameworks, each bearing two ligands at branch termini Claims 13 to 17: specify that the “two or more separate ligands” are the listed ligand(s) (AEG, AEM, AEBM, AETM, AEGA) within claim 3 5. Which insulin products are explicitly named? Claims 18 to 20 list insulin identities: Human/porcine/bovine insulin (claim 18) Insulin lispro, insulin aspart, insulin glulisine (claim 19) Insulin glargine or insulin detemir (claim 20) These are explicit “insulin molecules selected from” groups. A competitor using another insulin analog not listed would likely avoid those dependent claims. Independent claim 1 still covers “an insulin molecule” conjugate, but dependent claims add listed insulin types. Critical Claim Construction Risks (Where Infringement and Validity Pivot) 1. “Conjugate framework” and “indirectly conjugated” drive both infringement and validity The phrase “indirectly conjugated … via a conjugate framework” can be attacked as vague or overbroad if the specification does not define a consistent framework architecture that maps to prior art multicomponent conjugates. In practice, however, patents like this often survive because “framework” is implemented through a scaffold or linker that functions as a connector between insulin and ligand. Infringement consequence: accused products must likely show that ligand attachment occurs through an intermediate scaffold rather than a direct single-step functional group coupling to insulin’s targeted residues. Validity consequence: broad “framework” drafting can pull in prior art combinations where insulin is linked to a scaffold that then displays carbohydrate ligands. If the prior art already taught the same ligand list and a scaffold-based multivalent architecture, claim novelty could narrow to the insulin attachment residue locations and branching topology. 2. Insulin residue limitation is the main defensive feature Claim 2 restricts to A1, B1, Lys(B29), Lys(B3). These are specific and not interchangeable with random conjugation sites. Design-around strategies (high level): attach to other residues, attach via terminal chemistry that avoids those exact lysine epsilon-amino or A1/B1 positions, or use a conjugation method that changes which insulin functional group is linked. Infringement consequence: claim coverage becomes fragile if the claim chart cannot establish the attachment chemistry maps to those exact residues. 3. Branching topology claims are “structure-heavy” and can be easier to avoid Claims 4 to 6 and 10 to 12 require: one framework vs multiple frameworks, branched frameworks, ligands and insulin placed on separate branches, and in claim 6, that they are on termini. Such structure claims can be designed around by: using linear linkers, using clustered ligands on a single chain without branching, or positioning insulin and ligands on the same branch. Validity consequence: branching and multivalency are common in carbohydrate conjugate technologies. If prior art taught multivalent branched carbohydrate scaffolds, then the patent may stand or fall based on whether the prior art also matched the insulin residue limitation and the specific ligand list. 4. Multi-ligand via “different conjugation points” narrows technical overlap Claim 7 adds “two or more different conjugation points.” This can prevent literal reading if a competitor uses identical conjugation points for all ligand attachments or if the architecture is multivalent but assembled via repeated identical attachment positions. Where the Landscape Risk Likely Clusters (Based on Claim Dial Settings) Because the claims are constrained to a specific ligand set and a specific insulin attachment residue set, the strongest landscape overlap tends to come from patents in: 1) insulin conjugates with carbohydrate targeting ligands (AEG/AEM/AEBM/AETM/AEGA), 2) scaffold-based or “framework” linking to connect insulin to carbohydrate ligands, and 3) branched multivalent carbohydrate architectures. The highest risk scenario for infringement/FTO arises when a prior art patent already teaches all of these simultaneously: the same five ligand types (or close naming-matching equivalents), indirect scaffold presentation, and attachment to insulin at the same specified residues. A second tier risk scenario arises when prior art matches: ligand set + branched multivalent scaffolds, but differs in insulin attachment site; in that case, claim 2 and dependent residue-specific claims become the key differentiator. Criticality of Dependent Claim Ladder (How Coverage Expands) The claims build from a general conjugate to increasingly structured architectures: Claim 1 sets ligand universe and framework requirement. Claim 2 sets insulin residue attachment options. Claims 3 to 7 add multivalent ligand requirements, including a single framework vs branched frameworks and multi-point attachment. Claims 8 to 12 add specific insulin attachment patterning (A1 and Lys(B29) combinations), plus dual frameworks with ligands on separate branches and termini. Claims 13 to 17 lock the multiligand selection to one or more named ligands. Claims 18 to 20 lock the insulin identity to named products. This means an accused product can avoid dependent claims by modifying: ligand identity (outside the five named ligands), number of ligands (single ligand rather than two or more), framework structure (linear rather than branched, or branching not consistent with claims), conjugation residue (not A1/B1/Lys(B29)/Lys(B3)), insulin identity (use an analog not listed). Business-Grade Infringement Tests (What Must Be Shown) To establish literal infringement against independent claim 1 and dependent claims, the allegation must cover: Ligand identity The product must use one or more ligands in the specified set: AEG, AEM, AEBM, AETM, AEGA. Indirection and framework The ligand must be “indirectly conjugated” to insulin via a “conjugate framework,” not direct coupling. Insulin attachment residues (for dependent claims that require claim 2) The insulin attachment must involve A1 or B1 or Lys(B29) epsilon-amino or Lys(B3) epsilon-amino. Multivalency and branching (for dependent claims) For claims 3 to 7 and 10 to 12: the architecture must show multiple separate ligands, plus single vs branched frameworks, plus positioning at branch termini. Insulin identity (for claims 18 to 20) If the insulin type is not in the named group, only claim 1 and residue-agnostic coverage might be relevant. Validity Pressure Points (Where Prior Art Typically Attacks) A critical view of patentability for this kind of claim stack focuses on: Carbohydrate ligand novelty vs scaffold novelty If AEG/AEM/AEBM/AETM/AEGA insulin conjugates were already known, the main novelty must come from how the insulin is connected and how branching/multivalency is engineered. Insulin site conjugation as the differentiator If prior art already used those same residues (A1, B1, Lys(B29), Lys(B3)) for conjugation, the claim likely faces stronger obviousness attacks. Branched frameworks and multivalent conjugates as a known architecture class If branched multivalent carbohydrate presentation with insulin is known, the claim’s structural limitations can still be obvious if they are routine variations. Overbreadth in “framework” language If the specification supports only a narrow set of frameworks but the claims cover broader scaffold forms, that gap can create enablement or written description issues. Whether that plays out depends on the specification, which is not provided here. Key Takeaways US 10,398,781 claims insulin conjugates that combine a specific ligand set (AEG/AEM/AEBM/AETM/AEGA) with scaffold-based indirect conjugation and, in dependent claims, specific insulin attachment residues (A1, B1, Lys(B29), Lys(B3)). The strongest infringement risk is for products that meet all of: ligand identity, indirect framework linking, multivalent display, and branching/termini architecture. The tightest design-around lever is the insulin attachment site. If conjugation does not involve the claim-listed residues, much of the dependent claim ladder falls away. The next most important lever is architecture (branched vs linear, single framework vs multiple frameworks, ligands on separate branches/termini). Any landscape overlap that already teaches AEG/AEM/AEBM/AETM/AEGA insulin conjugates with multivalent branched frameworks is likely the main competitive threat; overlap that differs only by insulin attachment site tends to narrow relevance to dependent claims. FAQs 1. What does “indirectly conjugated … via a conjugate framework” typically cover? It generally covers architectures where the ligand is attached through an intervening scaffold rather than a direct bond to insulin’s functional group, implying multi-component assembly in which insulin and ligand are linked through a connecting framework. 2. Which single claim element is most critical for design-around? The insulin attachment positions in claim 2: A1, B1, Lys(B29) epsilon-amino, Lys(B3) epsilon-amino. If the product attaches through other residues, dependent claim coverage drops substantially. 3. Do the claims require a specific insulin manufacturing form? The patent depends on the insulin “molecule” identity in dependent claims (human/porcine/bovine; lispro/aspart/glulisine; glargine/detemir), but claim 1 uses broader “insulin molecule” language. 4. Does multivalency matter for claim coverage? Yes. Claim 3 and dependent branching claims require two or more separate ligands. A single-ligand conjugate may avoid those dependent claims. 5. Are branched frameworks a strict requirement? For specific dependent claims (4 to 6 and 10 to 12), yes. Those claims require branched topology and branch placement at termini, which can be engineered around by using linear linkers or non-matching branching placement. References [1] United States Patent 10,398,781. More… ↓ ⤷ Start Trial

Applicant	Tradename	Biologic Ingredient	Dosage Form	BLA	Approval Date	Patent No.	Expiredate
Eli Lilly And Company	HUMALOG	insulin lispro	Injection	020563	June 14, 1996	⤷ Start Trial	2037-08-18
Eli Lilly And Company	HUMALOG	insulin lispro	Injection	020563	August 06, 1998	⤷ Start Trial	2037-08-18
Eli Lilly And Company	HUMALOG	insulin lispro	Injection	020563	September 06, 2007	⤷ Start Trial	2037-08-18
Eli Lilly And Company	HUMALOG	insulin lispro	Injection	020563	June 06, 2017	⤷ Start Trial	2037-08-18
Eli Lilly And Company	HUMALOG	insulin lispro	Injection	020563	November 15, 2019	⤷ Start Trial	2037-08-18
>Applicant	>Tradename	>Biologic Ingredient	>Dosage Form	>BLA	>Approval Date	>Patent No.	>Expiredate

Patent: 10,398,781

Summary for Patent: 10,398,781