The pharmaceutical industry loves a good narrative about convergence. Drugs are meeting devices. Biologics are meeting sensors. Software is meeting syringes. The story has been told at every industry conference for a decade, and for once, the reality has caught up with the hype. Drug/device combination products are no longer a specialty niche — they are rapidly becoming the primary format for next-generation therapeutics, from prefilled autoinjectors for biologic drugs to AI-driven closed-loop insulin delivery systems.
The U.S. market for drug-device combination products was valued at over $46 billion in 2023 and is forecast to exceed $78 billion by 2030 [1]. That growth is not the result of incremental improvements to old products. It reflects a structural shift: the most commercially important therapies in oncology, immunology, diabetes, and rare disease increasingly cannot be administered without a sophisticated delivery system that is as proprietary as the drug itself.
That shift has profound consequences for patent strategy. When your product is a prefilled syringe for a monoclonal antibody, you are not just protecting a molecule — you are protecting an integrated system whose commercial value depends equally on the drug, the device, and the interaction between the two. Patent law, written largely with simpler inventions in mind, has struggled to keep up. Regulatory frameworks in the U.S. and Europe approach the very definition of ‘combination product’ from completely different philosophical angles, creating parallel tracks that a global company must run simultaneously.
This article is a working guide to that complexity. It covers what combination products actually are under U.S. FDA and EU EMA rules, what makes them uniquely difficult to patent, how to build a portfolio that holds up in litigation, what the major court battles have actually taught us, how filing globally requires three different strategic narratives from the same base document, and where the next wave of patenting challenges is coming from. Competitive intelligence tools like DrugPatentWatch sit at the center of any serious strategy in this space, and their role will be examined directly.
If you have read the standard regulatory primer on combination products, this is not that. The goal here is not to describe the rules — it is to explain what those rules mean for a company trying to build durable market exclusivity around a product that costs hundreds of millions of dollars to develop.
Section 1: The Definitional Problem — What Is a Combination Product, and Why Does the Answer Depend on Who Is Asking
The FDA’s Categorical Framework
The U.S. FDA defines ‘combination product’ under 21 CFR 3.2(e) and sorts products into three formal categories [2]. This categorical approach reflects a regulatory philosophy built around classification first, review pathway second. The agency wants to know what your product is before it decides who should review it.
Single-Entity Products
A single-entity combination product is one where two or more regulated components — a drug and a device, a biologic and a device, or all three — are physically, chemically, or otherwise combined into one unit [3]. A drug-eluting stent is the textbook example: the drug is coated directly onto the metal scaffold, and you cannot separate the two in any clinically meaningful sense. Prefilled syringes, transdermal patches, and inhalers where the drug and device are presented as a unit all fall into this category.
The clinical logic is straightforward. A drug-eluting stent does not function as a drug delivery product if you remove the stent, and it does not function as a cardiovascular device if you remove the drug coating. The regulatory argument follows from that clinical inseparability.
Co-Packaged Products
Co-packaged products involve two or more separately produced components that are sold together for a unified purpose. The components are not physically integrated, but the package is a single commercial unit. A reconstitution kit containing a vial of lyophilized drug and a prefilled syringe of sterile water is a classic example. First-aid kits combining bandages and antibiotic ointment are another, though most of those predate the formal regulatory framework [4].
This category creates a distinct set of patent challenges. The individual components may both be separately patented, but the commercial value of the combination — the convenience, the dosing accuracy, the reduced compounding errors — may itself be patentable. The question of whether the co-packaging creates a protectable combination invention is one that drug companies have increasingly explored as lifecycle management tools.
Cross-Labeled Products
Cross-labeled products are the most conceptually distinct. The drug and the device are sold separately, but each is labeled for use only with the other specific product. A photosensitizing drug that can only be activated by a particular laser system is the canonical example. The drug label references the device, and vice versa. Neither is therapeutically complete without the other [4].
From a patent perspective, cross-labeled products raise some of the most interesting enforcement questions, particularly around divided infringement. The manufacturer of the drug and the manufacturer of the device may be different companies. A patent on the combined therapeutic system — drug plus device, used according to a specific method — can quickly become unenforceable if no single entity is performing all the claimed steps. We will return to this at length.
The EU’s Functional Approach
The European regulatory framework, built primarily around the Medical Device Regulation (MDR) (EU) 2017/745, takes a different path [5]. The EU does not offer a single formal definition of ‘combination product.’ Instead, it asks what the product does — specifically, what its principal mode of action is — and uses the answer to assign a regulatory pathway.
The Principal vs. Ancillary Distinction
For integral drug-device combination products regulated in Europe, the central question is whether the medicinal substance contributes the principal therapeutic action or plays merely an ancillary role in supporting the device’s function [6].
If the drug is principal — meaning it is the primary source of the therapeutic effect — the product is regulated as a medicinal product under Directive 2001/83/EC. The device component must still meet the General Safety and Performance Requirements (GSPRs) of the MDR, but the overall regulatory pathway runs through medicines law.
If the drug is ancillary — supporting a device’s primary mechanical or physical function — the product is regulated as a medical device and requires CE marking. Drug-eluting stents are the most important commercial example of this classification in Europe. The stent’s primary action is mechanical (holding the artery open); the anti-proliferative drug coating is considered ancillary to that mechanical action. The same product is therefore device-led in Europe in a way that the FDA’s PMOA determination might or might not agree with, depending on the specific clinical argument a company makes.
This divergence is not trivial. A company that classifies a product as drug-led in the U.S. — routing it through CDER — may find that the identical product is classified as device-led in Europe, requiring a completely different technical documentation package and engaging a different set of reviewers. The regulatory strategy cannot be monolithic.
Non-Integral Products in the EU
Non-integral products cover situations where a device and a drug are intended to be used together but are not physically combined. These products must each obtain regulatory authorization separately: the device obtains CE marking, and the drug obtains marketing authorization through the normal medicines pathway. A prefilled nasal spray for a vaccine is an example where the vaccine and the spray device, if produced and sold separately, would each require independent authorization [6].
This structure is administratively cleaner in some ways — each component has its own regulatory identity — but it creates commercial complexity. If you want to market the combination as a unified product with a single label and a single commercial presentation, you may need to negotiate between two separate regulatory processes and two potentially different timelines.
Why the U.S.-EU Definitional Difference Actually Matters
The divergence between the FDA’s ‘What is it?’ question and the EMA’s ‘What does it do?’ question is more than philosophical. It has direct operational consequences that propagate through the entire development program.
Clinical trial design, for instance, is shaped by the lead regulatory framework. A drug-led combination product reviewed by CDER will face a different set of primary endpoint expectations than a device-led product reviewed by CDRH. The biocompatibility testing requirements under ISO 10993 are device-specific requirements that CDRH evaluates with different depth and focus than CDER would apply to the same data. A global development program must generate evidence capable of answering both agencies’ questions, which means the evidence-generation plan must be designed from the beginning with regulatory requirements in its architecture, not retrofitted after the fact.
Patent claim drafting, which we will get to in detail, is also affected. The EU’s functional emphasis on principal action provides a conceptual framework for how to argue the inventive step of a combination product at the European Patent Office (EPO). The USPTO examiner, working within a post-KSR framework, is asking a structurally different question about why the combination is non-obvious. A patent specification must support both arguments from its original filing date.
Section 2: Regulatory Pathways — The Two Front Doors and How You Choose the Right One
The U.S. System: PMOA, the OCP, and the Lead Center
The FDA’s Office of Combination Products (OCP), established by the Medical Device User Fee and Modernization Act of 2002, does not review marketing applications [7]. Its job is narrower and, in a practical sense, more consequential: it determines which FDA center — CDER, CDRH, or CBER — will take primary responsibility for reviewing a combination product. That determination flows from the product’s Primary Mode of Action (PMOA), defined as the single mode of action expected to make the greatest contribution to the overall intended therapeutic effects [3].
How the Three Centers Divide the Work
The PMOA determination routes the product to its review home:
Drug-dominant products go to CDER, following an NDA or BLA pathway. A metered-dose inhaler is a standard example — the primary therapeutic effect comes from the inhaled corticosteroid or bronchodilator, not from the aerosol mechanics. The device component is real and technically complex, but the drug is why the patient gets better.
Device-dominant products go to CDRH, following a 510(k), De Novo, or PMA pathway. Drug-eluting stents are the standard example here. The coronary scaffold holds the artery open; the anti-proliferative drug prevents restenosis. The FDA has consistently treated the stent’s mechanical action as primary.
Biologic-dominant products go to CBER, typically requiring a BLA. A nasal spray delivering a live-attenuated influenza vaccine is a combination product with the biologic leading.
What happens when the PMOA is genuinely ambiguous? The FDA’s regulations provide a tiebreaker: the product goes to whichever center has the most expertise in evaluating the most significant safety and effectiveness questions raised by the product [3]. This is a judgment call, and it is one of the reasons the formal and informal designation processes matter so much.
The RFD and Pre-RFD: Getting Binding Clarity Before You Need It
The Request for Designation (RFD) is a formal submission to the OCP that produces a binding determination on a product’s classification and lead center assignment [8]. Filing an RFD gives a company certainty — the kind of certainty that is worth its weight in development budget when you are planning a multi-year clinical program that depends on knowing whether you are filing an NDA or a PMA.
The Pre-Request for Designation (Pre-RFD) is the informal version. It produces non-binding feedback from the OCP on how the agency is likely to classify a product [8]. The Pre-RFD is the tool to use when you want to test your PMOA argument before committing to it in a formal submission. The OCP’s response in a Pre-RFD will often signal which aspects of your PMOA argument are working and which are not. That feedback is strategically valuable at a stage when the development program is still flexible enough to respond to it.
Companies that skip both processes and simply choose a regulatory pathway based on their own internal PMOA determination are taking an unnecessary risk. An incorrect pathway choice can result in a complete resubmission, with the associated time and cost implications.
The Single Application Model and Intercenter Review
The FDA’s preferred approach for combination products is a single marketing application, submitted to the lead center, which then coordinates an intercenter consultative review [9]. This means that even when CDER is the lead center for a drug/device combination product, it formally requests that CDRH review the device-specific data within the NDA — engineering specifications, risk analysis, biocompatibility, and human factors studies.
This structure is administratively rational, but it has a practical consequence that many sponsors underestimate. A deficiency in the device data package — inadequate human factors documentation, an incomplete biocompatibility assessment, missing risk management file elements — can result in a Complete Response Letter (CRL) for the entire application. The drug data may be flawless; the device data gap stops the whole thing. This is why CDRH’s expectations must be understood and met regardless of which center is leading the review.
The EU System: Notified Bodies and the MDR Bottleneck
The European system’s bifurcated structure creates a different kind of operational challenge. For integral combination products that are regulated as medicinal products (i.e., the drug action is principal), Article 117 of the MDR introduced a requirement that has substantially changed the regulatory timeline for these products: the Marketing Authorisation Application (MAA) is not considered complete without a Notified Body Opinion (NBOp) on the device component [5].
The NBOp Requirement and Its Practical Weight
A Notified Body is a private, EU-designated third-party organization that assesses medical device conformity. For drug-led integral combination products, the sponsor must engage a Notified Body to review the device constituent against the General Safety and Performance Requirements (GSPRs) of MDR Annex I and issue a formal opinion [6]. This opinion must accompany the MAA.
The NBOp is not a rubber stamp. Preparing the technical file for the device component to satisfy a Notified Body requires the same level of rigor as a standalone medical device submission. The GSPR checklist alone requires detailed justification and supporting documentation for each applicable requirement — design and development controls, risk management per ISO 14971, biocompatibility per ISO 10993, usability engineering per IEC 62366, clinical evaluation, and more [10].
The operational risk is timing. Notified Bodies are private organizations with their own submission backlogs and review timelines. They are not bound by the EMA’s clock. A sponsor who engages a Notified Body late in the development process, or who underestimates the documentation burden, may find that the NBOp is not ready when the MAA needs to be submitted. This is not a hypothetical risk — the transition to the MDR from the legacy Medical Device Directive has created capacity constraints at several major Notified Bodies that have had measurable effects on product approval timelines.
Compare this to the FDA’s intercenter model. The CDRH review of device data within an NDA is conducted by a government center that operates within the FDA’s own administrative framework. The sponsor’s primary point of contact is the lead center; the intercenter consultation is the FDA’s internal process to manage. In Europe, the sponsor must simultaneously manage two external relationships — the Notified Body and the competent authority or EMA — with interdependent deliverables.
Side-by-Side: The Two Systems Compared
Feature
U.S. FDA
EU EMA / Competent Authorities
Core question
What is the product?
What does the product do?
Key legal text
21 CFR Part 3; FD&C Act §503(g)
MDR (EU) 2017/745; Directive 2001/83/EC
Classification factor
Primary Mode of Action (PMOA)
Principal vs. ancillary action of the medicinal substance
Lead reviewer
Single lead center (CDER, CDRH, or CBER)
Notified Body (device) + competent authority/EMA (drug)
Formal pre-submission tool
Pre-RFD (non-binding) / RFD (binding)
Scientific Advice (EMA); Notified Body pre-submission interaction
Key bottleneck
PMOA determination; intercenter data requirements
Notified Body Opinion (NBOp) availability and timeline
Application model
Single application; lead center coordinates
Two parallel processes with interdependent outputs
Section 3: The Three Patent Traps That End Combination Product Programs
Trap 1: Obviousness in a Post-KSR World
What KSR Actually Changed
Before 2007, the dominant standard for proving patent obviousness in the U.S. required a patent examiner or challenger to find an explicit teaching, suggestion, or motivation (TSM) in the prior art to combine known elements. This framework gave innovators significant room to patent combinations of known components, because the absence of a specific prior art suggestion to combine them was often sufficient to defeat an obviousness rejection.
KSR International Co. v. Teleflex Inc. changed that [11]. The Supreme Court unanimously rejected the rigid TSM test and replaced it with a flexible, expansive approach grounded in common sense. The Court held that combining familiar elements according to known methods is likely obvious when it does no more than yield predictable results. A creative person having ordinary skill in the art would see the benefit of combining elements to solve known problems, especially when there are a finite number of identified, predictable solutions. Pursuing one of those options can be ‘obvious to try’ — and therefore unpatentable — even without a specific prior art reference teaching the combination.
For combination products, the implications are severe. A prefilled syringe containing a known biologic drug delivered by a known injection device, combined for the obvious purpose of patient convenience, is almost certainly obvious after KSR. A metered-dose inhaler delivering a known corticosteroid through a known inhaler device is similarly vulnerable. The KSR framework shifts the patentability question from ‘Is there a prior art reference specifically teaching this combination?’ to ‘Is the result of this combination something unexpected?’
Unexpected Synergy: The Escape Route
The most powerful defense against an obviousness rejection for a combination product is evidence of unexpected synergistic effects. The combination must produce a result that is not merely the sum of the components’ individual contributions — it must produce something that a skilled artisan would not have predicted based on the prior art.
Synergy can take several forms. The device material might unexpectedly improve the long-term stability of a sensitive biologic drug — for example, a specific polymer formulation in the barrel of an autoinjector preventing protein aggregation that would occur in a standard glass syringe. The delivery mechanism might produce unexpected improvements in pharmacokinetics — for instance, a novel nozzle geometry in an inhaler creating a particle size distribution that achieves deeper lung deposition than prior art devices, resulting in measurably better clinical outcomes at lower doses. A combination of drug and device might unexpectedly reduce a class of adverse events compared to either component alone.
Two things matter here. First, the synergy must be real and it must be quantifiable. Qualitative assertions that the combination ‘works better’ will not satisfy a patent examiner operating in the KSR framework. Comparative data, ideally from well-controlled experiments, is required. Second, the synergy must be unexpected — meaning a skilled artisan, reading the prior art, would not have predicted it. If the prior art already teaches that a particular polymer is compatible with biologics, or that a particular particle size improves lung deposition, the effect is no longer unexpected regardless of how well you have quantified it in your own testing.
Long-Felt Need and Teaching Away
Two secondary arguments support non-obviousness claims for combination products. Long-felt but unresolved need is persuasive when an industry problem has persisted for years despite multiple failed attempts at solution. If patients have demonstrably struggled with dosing errors using available injection devices, and prior art records show that the problem was recognized and studied without a satisfactory solution, a new combination product that measurably solves the problem carries a stronger non-obviousness argument.
Teaching away is available when the prior art actively discourages the approach the inventor took. If the prior art teaches that a certain polymer material is incompatible with a particular drug class — that it causes degradation, adsorption, or aggregation — and the inventor discovers a way to use that material successfully, the prior art ‘teaches away’ from the invention, supporting patentability. This argument requires direct evidence from the prior art, not just an absence of prior art teaching the combination.
Trap 2: Disclosure Failures — Written Description and Enablement
Non-obviousness is not the only hurdle. Even a genuinely inventive combination product fails the patent system if it is not properly disclosed. The U.S. patent statute at 35 U.S.C. §112 imposes two distinct requirements: written description and enablement.
The Written Description Requirement
Written description requires that the patent specification demonstrate, to a person having ordinary skill in the art (PHOSITA), that the inventor was in possession of the claimed invention at the time of filing [12]. The specification must describe the invention with enough particularity that it is clear the inventor actually invented what is being claimed — not merely conceived of a broad idea and filed on speculation.
For combination products, written description problems typically arise at the boundaries of claim scope. A claim to ‘a prefilled autoinjector for delivering a biologic drug’ is broad. The written description must support that breadth. If the specification provides detailed description of a single specific autoinjector design for a single specific monoclonal antibody, a claim covering all autoinjectors and all biologics may fail for insufficient written description. The specification must describe representative examples across the claimed scope with enough specificity to demonstrate actual possession of the full invention.
The Enablement Requirement
Enablement requires that the specification teach a PHOSITA how to make and use the full scope of the claimed invention without undue experimentation [13]. The standard is not that every detail must be spelled out — a skilled person can apply routine knowledge to fill ordinary gaps. The standard is that the gap-filling does not require undue experimentation: a substantial research program, a high rate of failure, unpredictable variables, or a requirement for inventive work beyond what a skilled artisan would routinely undertake.
For a combination product, enablement failures typically occur when the claimed scope exceeds what the specification actually teaches. If you claim a system for delivering any biologic drug from any autoinjector, but your specification only provides detailed data for two specific antibody formulations in two specific device configurations, you may face an argument that producing the full claimed range would require undue experimentation. The viscosity range, aggregation propensity, and stability profiles of different biologics vary enormously. The mechanical requirements of an autoinjector change substantially with different drug volumes and viscosities. Demonstrating enablement across the full scope of a broad combination product claim is a real challenge.
The Critical Connection Between Disclosure and Obviousness
There is an important strategic link between these disclosure requirements and the non-obviousness argument. If your strategy for overcoming obviousness is to cite unexpected synergistic effects — say, a specific polymer coating that prevents protein aggregation — then your patent specification must describe that coating in enough detail to enable a skilled artisan to make it. You cannot simultaneously argue to the patent examiner that the effect is unexpected and therefore non-obvious, and then fail to disclose how to achieve the effect. The unexpected result that justifies the patent is also the result that the public must be able to reproduce when the patent expires. That is the bargain. Failing to disclose it adequately breaks the bargain and renders the patent invalid.
This connection has a practical consequence for how R&D and patent filing timelines should be coordinated. The data generated to prove unexpected synergy for the non-obviousness argument must be available, understood, and accurately disclosed before the patent is filed. Filing early with preliminary data and planning to supplement later is risky precisely because the disclosure requirements at the EPO and JPO — as we will discuss — are even stricter than in the U.S. about what can be added after the filing date.
Trap 3: Divided Infringement
The Single-Entity Rule and Why It Matters for Combination Products
Direct patent infringement in the U.S. under 35 U.S.C. §271(a) requires that a single entity perform every step of a claimed method or make, use, or sell every element of a claimed system [14]. This single-entity rule creates an enforcement gap that is particularly dangerous for modern combination products, especially those involving patient interaction or digital health components.
Consider a patent claim for a method of managing Type 1 diabetes using a smart insulin pen system:
The pen’s sensor reads continuous glucose data from a separate monitoring device.
An algorithm in the pen’s companion app calculates a recommended insulin dose.
The app displays the recommendation to the patient.
The patient decides to accept the recommendation and administers the dose via the pen.
Who infringes this claim? The device manufacturer makes the pen and the app that perform steps 1, 2, and 3. But the patient performs step 4 — a step that the manufacturer has no legal control over. Under the single-entity rule, the manufacturer does not perform all four steps and is therefore not a direct infringer. The patient is not commercially motivated to enforce rights against. The patent claim describes a genuine, valuable innovation, and it is essentially unenforceable.
Method Claims vs. System Claims: The Strategic Fix
The primary tool for avoiding divided infringement is to shift claiming strategy from method claims to system or apparatus claims. Instead of claiming the method of performing the therapeutic steps, claim the integrated system that enables those steps.
The same insulin pen system, claimed as an apparatus: ‘A diabetes management system comprising: a glucose sensor configured to receive blood glucose measurements; a processor configured to calculate a recommended insulin dose based on the received measurements; a display configured to present the recommended dose; and a drug delivery mechanism configured to administer a selected dose of insulin.’ The manufacturer who makes and sells this integrated system directly infringes the apparatus claim. There is no divided infringement problem because a single entity — the manufacturer — has made the complete claimed system.
This reframing is not always straightforward. Some of the most valuable aspects of a combination product therapy are the steps that occur between the patient and the device during actual use — the feedback loop, the adherence behavior, the clinical monitoring. These are inherently multi-party activities that are difficult to capture in apparatus claims without losing the specificity that makes the claim valuable. Skilled claim drafting in this space requires understanding both the technical architecture of the product and the behavioral patterns of its use.
The Akamai Doctrine and Its Limits
Courts have recognized a narrow exception to the single-entity rule through the doctrine of joint infringement, articulated in Akamai Technologies v. Limelight Networks [15]. Under this doctrine, a party can be liable for direct infringement even without performing all the claimed steps if it directs or controls the performance of the remaining steps by another party. The test requires either a contractual obligation to perform the steps or a principal-agent relationship.
For combination products, the Akamai doctrine rarely provides meaningful relief. The relationship between a device manufacturer and the patients who use the device is neither contractual in the required sense nor a principal-agent relationship. Patients choose how to use a device; the manufacturer cannot legally compel them to perform specific acts in a specific sequence. The Akamai exception is available in scenarios where a company contracts out specific process steps — a situation that arises more commonly in software patents than in the medical device-drug combination space.
The practical lesson is that the divided infringement problem is not something to solve in litigation. It must be solved in claim drafting, before the patent application is filed. Building a patent portfolio where the core claims are directed to apparatus and system claims held by the manufacturer, with method claims as supplementary protection for the complete user interaction, is the only reliable structural solution.
Section 4: Building the Picket Fence — A Patent Portfolio That Actually Holds
Why Single-Patent Protection Fails
A single core patent on a drug molecule may be sufficient for a simple small-molecule drug sold as a tablet. For a drug/device combination product, a single patent is not a defense — it is a target. Competitors will test every claim in that patent, probe for prior art, challenge the inventive step, and search for design-around opportunities. The answer is not to write a better single patent. It is to make single-patent challenges irrelevant by building a portfolio that forces a competitor to defeat multiple, overlapping layers of protection simultaneously.
This is the ‘picket fence’ or ‘patent thicket’ strategy [16]. A well-constructed picket fence for a next-generation autoinjector delivering a biologic drug might include dozens of patents covering different aspects of the product, all of which must be cleared before a generic manufacturer can bring a competing product to market.
Layer 1: The Drug
The foundation is the drug itself. Core patents on the active pharmaceutical ingredient, its crystalline forms (polymorphs), specific formulations with excipients that improve stability or bioavailability, synthesis methods, and metabolites all belong in this layer [17]. For a biologic, the core patents may cover the antibody sequence, the expression system, the purification process, and specific formulation parameters that are necessary for storage stability.
Formulation patents are particularly important for combination products because the drug formulation is often the interface between the pharmaceutical and the device. A biologic that must have viscosity below a certain threshold to flow through an autoinjector’s needle, or that requires specific pH and ionic strength to remain stable in a polymer container, will have formulation parameters that are both patent-eligible and directly relevant to the device design. These formulation patents sit at the intersection of the drug and device layers and are often the most difficult for a competitor to work around.
Layer 2: The Device — Mechanical, Design, and Human Factors
Device patents are where the combination product strategy diverges most sharply from standard pharmaceutical IP management. Many pharma companies historically treated the delivery device as a commodity — necessary packaging, not a source of IP value. That approach is commercially catastrophic when you are trying to maintain market exclusivity after a drug patent expires.
Utility patents should cover every novel mechanical element of the device: spring mechanisms, needle insertion geometry, dose-setting and dose-delivery systems, safety shields, sterility-maintenance features, and any material choices that produce non-obvious results. Human factors design — the ergonomic and interface features that improve usability for patients with limited dexterity, visual impairments, or needle anxiety — can be patented when those design choices are novel and non-obvious solutions to documented usability problems.
Design patents protect the ornamental appearance of the device independently of its function [18]. They are faster to obtain, cheaper to maintain, and harder for competitors to design around than utility patents on mechanical function. A distinctive device shape, button placement, or visual indicator pattern can be the subject of a design patent that prevents a competitor from closely copying the product’s appearance even after every utility patent has expired or been invalidated.
Layer 3: The Drug-Device Interaction
This is the layer that combination product patents must build that a standalone drug or standalone device company never needs. Patents directed specifically to the synergistic interaction between the drug and the device are potentially the most valuable and the most difficult to obtain.
A claim like ‘an autoinjector comprising [specific polymer materials] wherein the combination reduces aggregation of [drug class] by at least [X]% over [timeframe] compared to glass-siliconized containers’ is valuable because it captures the unexpected synergistic benefit, it is directly tied to measurable clinical outcome (drug stability affecting safety and efficacy), and it is difficult to design around without either changing the drug formulation or finding an alternative material with equivalent properties.
These interaction claims require the most rigorous scientific data to support — exactly the kind of comparative, quantitative data that also supports the non-obviousness argument during prosecution. R&D should generate this data with both purposes in mind. Testing the drug in three different container systems, across a range of storage conditions, with rigorous aggregation and potency assays, produces data that simultaneously demonstrates non-obviousness to the patent examiner and provides the scientific basis for an interaction claim.
Layer 4: Software, Algorithms, and Data
For connected combination products — smart pens, connected inhalers, wearable injectors with monitoring capability — the software layer is increasingly the most commercially important layer in the portfolio. Patents on the algorithms that process sensor data, the machine learning models that personalize dosing recommendations, the user interface designs that improve patient engagement, and the data transmission protocols that securely move health data to the cloud are all available and valuable [19].
The practical challenges here are twofold. First, software patent eligibility in the U.S. under 35 U.S.C. §101 remains uncertain after Alice Corp. v. CLS Bank International [20]. Abstract ideas implemented on a generic computer are not patent-eligible; the claims must be directed to a specific, concrete technical improvement in the software itself, not merely the use of a computer to perform a business method. Carefully drafted claims that emphasize the specific technical problem being solved and the specific technical architecture of the solution are more likely to survive §101 scrutiny than broadly drafted claims on the concept of ‘using an algorithm to optimize dosing.’
Second, the divided infringement problem is at its worst in this layer. Method claims covering the steps of collecting sensor data, processing it through an algorithm, and transmitting a recommendation to a patient almost necessarily involve multiple parties. System claims and apparatus claims must do the heavy lifting.
Layer 5: Manufacturing Processes
Process patents on novel manufacturing and assembly methods are an underutilized layer in many company portfolios. Novel aseptic fill-finish processes, sterilization methods for device components that contact the drug, and assembly techniques that maintain drug stability are all potentially patentable. Process patents have a distinct enforcement advantage: a competitor who uses the same manufacturing process infringes the patent even if their product is physically different. And because manufacturing processes are not visible in the final product, they are more difficult for a competitor to assess during design-around analysis.
Aligning the Picket Fence with Regulatory Exclusivity
Patent protection runs on one clock; regulatory exclusivity runs on another. Effective lifecycle management requires understanding both and designing a strategy that creates overlapping protection across the full commercial life of the product [21].
FDA regulatory exclusivities for drug/device combination products include New Chemical Entity (NCE) exclusivity (5 years from approval, blocking generic ANDAs for the full period), biologic exclusivity (12 years from approval), Orphan Drug Exclusivity (7 years), and Clinical Investigation Exclusivity (3 years for products requiring new clinical studies for approval) [22]. These exclusivities operate independently of patents — they prevent FDA from approving a competing product regardless of patent status.
The strategic objective is to ensure that when regulatory exclusivity expires, patent protection continues — and that as core drug patents near expiration, device and combination patents extend the effective monopoly. A biologic in an autoinjector with 12-year biologic exclusivity and a core antibody patent expiring at year 15 leaves a 3-year gap if no other protection exists. Device patents filed in year 5 or 6 of the product’s life — covering meaningful improvements to the second-generation device — can extend into years 20-25, creating a commercial hurdle for biosimilar manufacturers who must develop non-infringing delivery systems before they can compete.
This strategy, sometimes called ‘evergreening,’ is legitimate when the device improvements are genuine and independently patentable [23]. It becomes legally and commercially dangerous when the device patents are on trivial modifications filed solely to extend exclusivity without providing any patient benefit — the threshold at which a regulator or antitrust authority begins to take notice.
Section 5: Competitive Intelligence — Reading the Patent Landscape
DrugPatentWatch and the Business Intelligence Layer
Patent data, read strategically, is a live map of your industry’s technological movements. A filing by a competitor today may be a product on the market in seven years. Monitoring those filings — systematically, analytically, and in context — is a competitive intelligence function that most pharma companies under-invest in until a generic manufacturer files a Paragraph IV challenge against their most important product.
DrugPatentWatch provides the kind of structured, searchable patent and regulatory data that enables this analysis for drug and combination product portfolios [24]. Analysts can use the platform to trace the full patent landscape around a branded combination product — identifying every patent listed in the Orange Book, mapping expiration dates, tracking inter partes review (IPR) filings against those patents, and monitoring Paragraph IV certification filings by generic manufacturers. For biosimilars and biologic combination products, the platform’s tracking of FDA approval actions and biologic exclusivity periods adds a critical second data stream.
Mapping Competitor Portfolios
A systematic analysis of a competitor’s patent portfolio tells you several things that no press release will. The density of filing in a particular technology area reveals where the company has placed its R&D bets. A sudden cluster of filings on connected device technology, sensor integration, or companion app architecture signals a product pipeline development that may be two to five years from market. The claim scope of those filings tells you how broadly the company believes it can protect those innovations — and where the boundaries are that your own R&D might work around.
Patent ‘white space’ analysis — identifying technology areas within the combination product space where no company has established significant protection — is a legitimate tool for directing R&D investment. If a specific class of wearable injector designs for subcutaneous delivery of high-viscosity drugs has limited patent coverage, that represents both a product development opportunity and a signal that the technical challenges in that area may not yet have been solved.
Predicting Generic Entry
For a branded combination product, the most commercially important question in any long-range financial model is when generic competition will begin. That question requires tracking at least five variables simultaneously: the expiration dates of all Orange Book-listed patents, the patent term extension (PTE) status of those patents under Hatch-Waxman, the presence and outcome of any Paragraph IV challenges, the status of any FDA citizen petitions that might delay generic approval, and the expiration of applicable regulatory exclusivity periods.
A generic manufacturer filing a Paragraph IV certification is declaring that at least one of the listed patents is invalid or will not be infringed by the generic product. That filing triggers a 30-month stay of FDA approval and, frequently, patent litigation. Monitoring these certifications through tools like DrugPatentWatch provides advance warning of a generic entry timeline that is otherwise difficult to predict from public information alone.
For biologic combination products, the analysis is more complex. Biosimilar applicants follow the Biologics Price Competition and Innovation Act (BPCIA) pathway, which includes its own patent dance procedures and a 12-year exclusivity period for the reference product. Tracking biosimilar applications, interchangeability designations, and the BPCIA patent exchange process requires a specialized data infrastructure that generic-focused intelligence platforms have built out in recent years.
Section 6: Lessons from the Courtroom
Boston Scientific v. TissueGen: The Value of a Specific Inventive Narrative
In February 2023, a federal jury awarded TissueGen and the University of Texas $42 million in unpaid royalties against Boston Scientific for infringement of a patent covering a drug-releasing biodegradable fiber implant [25]. The patent was the product of research conducted in the late 1990s by Dr. Kevin Nelson at a time when the medical device industry was focused primarily on simple drug coatings applied to metal stents.
Nelson’s invention was structurally different from standard drug-eluting stents. Rather than coating a permanent metal scaffold with a drug, his research proposed using biodegradable polymer fibers that could release a therapeutic agent and then be absorbed by the body over time, actively facilitating arterial healing rather than simply preventing restenosis with a permanent implant. The conceptual distinction — from coating a permanent device to building an absorbable drug-releasing structure — was the inventive nucleus that the patent was built around.
Boston Scientific’s Synergy stents, which feature an ultrathin biodegradable polymer coating designed to be absorbed after drug elution, were found to infringe this patent. The jury’s verdict reflects the strength of a patent claim built around a specific, functionally superior mechanism rather than the general concept of combining a drug with a cardiovascular device.
The post-KSR lesson from this case is precise. The ‘obvious to try’ argument — that drug-eluting stents were an obvious concept, and that biodegradable polymers were a known material class — did not defeat the patent because the patent was not claiming the general concept. It was claiming a specific structural approach that produced a measurably different clinical outcome: tissue healing facilitated by absorbable polymer fibers, not just restenosis prevention via permanent drug coating. The patent told a specific inventive story, backed by the scientific distinction between its approach and the prior art’s approach.
The commercial implication is direct: combination product patents must be built around specific, demonstrable technical distinctions from the prior art. Broad genus claims covering all biodegradable polymer drug-delivery devices, or all drug-eluting cardiovascular implants, would have been far easier to challenge. The specificity that sometimes seems to limit a patent’s commercial scope is often what makes it defensible in litigation.
The EpiPen Saga: When Patent Strategy Triggers Antitrust
The EpiPen litigation — In re EpiPen (Epinephrine Injection, USP) Marketing, Sales Practices and Antitrust Litigation — produced two major settlements: $345 million from Pfizer in 2021 and $264 million from Mylan in 2022, for a combined total of $609 million [26, 27]. Neither company admitted wrongdoing. Together, the settlements represent one of the most expensive lessons in modern pharmaceutical antitrust law.
The Alleged Patent and Regulatory Tactics
The EpiPen autoinjector’s price rose by over 600% between 2007 and 2016 [28]. Class action and antitrust plaintiffs alleged that this price trajectory was not the result of market forces but of a coordinated scheme to eliminate competition using multiple legal mechanisms simultaneously.
The patent-related allegations focused on what plaintiffs described as a ‘patent thicket’ of overlapping patents covering incremental modifications to the EpiPen device, combined with infringement lawsuits filed against every generic entrant — Teva and Sandoz among them — followed by settlements in which generic manufacturers agreed to delay market entry. These settlements, characterized by plaintiffs as ‘reverse payments’ or ‘pay-for-delay’ arrangements, are the subject of established antitrust analysis following FTC v. Actavis [29].
The regulatory process allegations were perhaps the most distinctive element. Plaintiffs alleged that Mylan filed a Citizen Petition with the FDA shortly before a Teva generic was expected to receive approval. The petition raised design concerns about the generic device. Courts reviewing the antitrust complaint found sufficient allegations that the petition was not a legitimate safety concern but a tactic designed to exploit the regulatory process as a delay mechanism — what courts call ‘sham litigation’ extended to regulatory filings [30].
The Commercial Strategy Allegations
Beyond the patent and regulatory allegations, plaintiffs described a set of commercial practices that, taken together, were alleged to constitute an unlawful scheme to maintain monopoly. In 2011, Mylan eliminated single-unit EpiPen sales in the U.S. and moved exclusively to a two-pack format. Plaintiffs characterized this as tying — forcing consumers to purchase a second, often unnecessary, device — effectively doubling the price without a compelling medical justification [30]. Mylan also offered substantial rebates to pharmacy benefit managers conditioned on preferential formulary placement, and its EpiPen4Schools program signed schools to exclusivity arrangements that blocked competing products from school purchasing channels.
The Antitrust Tripwire Every IP Team Must Know
The EpiPen case is not primarily a patent case. Each individual action — filing a patent on a device improvement, offering commercial rebates, changing product packaging, filing a citizen petition — might be legally defensible in isolation. The plaintiffs’ theory, and the theory that produced $609 million in settlements, was that these actions were part of a coordinated scheme whose purpose was not to improve the product or compete on the merits, but to eliminate competition unlawfully.
This is the antitrust tripwire. IP strategy does not operate in a vacuum. A company with a dominant market position that builds a patent thicket, uses those patents to file serial infringement lawsuits followed by pay-for-delay settlements, manipulates product packaging to foreclose competition, and abuses regulatory processes to delay generic approval is not simply being aggressive in protecting its IP. It is potentially committing antitrust violations for which the entire IP strategy becomes evidence of unlawful conduct.
The practical lesson for business leaders is that an antitrust review of the combined patent and commercial strategy is not optional for a market-leading combination product. The question is not just ‘Can we legally obtain this patent and enforce it?’ The question is also ‘How does this patent enforcement action look in the context of our overall market behavior?’ Getting that second question wrong costs a minimum of $609 million.
Section 7: Filing Globally Without Filing Stupidly — USPTO, EPO, and JPO
Three Different Inventive Step Standards From the Same Specification
The three most commercially important patent jurisdictions for pharmaceutical and medical device companies — the U.S. (USPTO), Europe (EPO), and Japan (JPO) — each evaluate inventive step through a different analytical framework. A patent application that convincingly argues non-obviousness at the USPTO may require a substantially different narrative to succeed at the EPO, and a third narrative still for the JPO. The same underlying data must be framed three different ways.
The USPTO: Flexible Post-KSR Analysis
As discussed in Section 3, the USPTO applies a flexible, holistic approach after KSR. Any rational reason to combine prior art elements — predictable results, known substitutions, obvious-to-try logic — can support an obviousness rejection. The applicant’s strongest response is unexpected results data, long-felt need, or teaching away from the prior art. The USPTO examiner considers the invention as a whole and applies common-sense reasoning.
Arguments at the USPTO for a combination product should lead with the data. Comparative results showing that the combination performs measurably better than the prior art on outcomes a skilled artisan would have cared about, presented as part of the prosecution history, create a robust record for later validity litigation. Examiners are willing to consider declarations by inventors and expert witnesses providing technical context for why results are unexpected.
The EPO: The Problem-Solution Approach
The EPO’s problem-solution approach is formalistic and structured [31]. The examiner identifies the closest prior art — typically the single prior art reference that most closely resembles the claimed invention — and uses it as the starting point. The examiner then determines the ‘objective technical problem’: what specific technical effect or advantage does the claimed invention achieve relative to the closest prior art? Finally, the examiner asks whether a skilled person, starting from the closest prior art and faced with the objective technical problem, would arrive at the claimed invention using common general knowledge.
For a combination product patent at the EPO, the applicant’s strategic task is to define the objective technical problem narrowly enough that the claimed invention is the non-obvious solution. If the closest prior art is a drug-eluting stent with a non-biodegradable polymer coating, the objective technical problem might be framed as ‘how to reduce the risk of delayed hypersensitivity to a permanent polymer implant.’ The claimed biodegradable fiber approach is then presented as a solution to that specific problem — a solution that was not obvious given the prior art’s teaching.
EPO practitioners must be careful about two things. First, the objective technical problem cannot be formulated in a way that already suggests the solution. ‘How to create a biodegradable drug-delivery implant’ essentially answers itself. The problem must be formulated based on the effect achieved, not the means used. Second, the closest prior art is fixed as the starting point; the applicant cannot avoid an obviousness argument by pointing to a different reference as a ‘better’ starting point than the one the examiner chose.
The JPO: Advantageous Effects as a Key Factor
The JPO places considerable weight on ‘advantageous effects’ — demonstrated improvements over the prior art — in its inventive step analysis [32]. An invention that shows a different type of effect from the prior art, or an effect that is significantly and unexpectedly greater in magnitude, carries a strong presumption of inventive step at the JPO.
This makes the JPO particularly receptive to combination product patents that can demonstrate quantitative superiority: bioavailability improvements measured with statistical significance, aggregation reduction expressed as percentage change over time, dosing accuracy measured against predefined user error thresholds. The JPO’s examination guidelines explicitly recognize that marked unexpected effects support a finding of inventive step even when the combination itself might seem obvious in principle [32].
Japanese filings for combination products benefit from including the most detailed comparative data early — in the original application or in a response to an office action that addresses the JPO’s expectation of substantiated effect claims. Post-filing data can sometimes confirm effects that were made plausible in the original filing, but a Japanese application that contains no comparative data and relies entirely on post-filing experimental results is on weak ground.
The Post-Filing Data Problem
The most operationally consequential difference among the three jurisdictions is their approach to data submitted after the original filing date. The USPTO allows an applicant to submit new experimental data during prosecution to support non-obviousness and enablement arguments. This flexibility has been extensively exploited in the U.S.: companies file early, with preliminary data supporting the claimed invention, and use the prosecution period to generate and submit the definitive experiments.
The EPO applies a much stricter ‘plausibility’ doctrine [33]. Post-filing data can be used to confirm an effect that was already made plausible in the original application, but it cannot rescue an application that lacks any credible basis for the claimed invention as filed. The Technical Boards of Appeal have been explicit: if the original application provides no basis for expecting the claimed technical effect, post-filing data demonstrating the effect does not establish inventive step because the skilled person would not have recognized the effect at the filing date.
The JPO takes a similar position. The sufficiency of the original disclosure is assessed based on what was actually disclosed at filing. Post-filing data has limited corrective power.
This creates a direct conflict between standard U.S. provisional application practice — file early with preliminary data, use the 12-month priority period to generate additional support — and the requirements of a viable global filing strategy. A PCT application entering national phase in Europe and Japan based only on preliminary data may encounter EPO and JPO rejections that post-filing data cannot fix. The optimal global strategy often means waiting until comprehensive, comparative data is available before filing the PCT application, using the provisional period in the U.S. to generate that data, and then filing internationally with a specification that can independently support inventive step in all three jurisdictions without reliance on post-filing supplements.
One Specification, Three Narratives
The practical solution to this three-jurisdiction challenge is a sufficiently rich and detailed original specification that contains enough technical disclosure, comparative data, and mechanistic explanation to support all three strategic arguments without amendment.
The specification should:
Include comparative data against the closest prior art compound or device, with quantitative results and statistical analysis — this satisfies the JPO’s advantageous effects requirement and provides the EPO’s plausibility requirement.
Describe the mechanism of the synergistic effect, not just the effect itself — both the EPO and the USPTO want to understand why the combination achieves the unexpected result.
Include sufficient breadth of examples to support the claimed scope without requiring undue experimentation — this addresses enablement at the USPTO and sufficiency of disclosure at the EPO and JPO.
Frame the objective technical problem in the description in a way that sets up the EPO’s problem-solution analysis without being so narrow that it unnecessarily limits claim scope.
Writing this specification requires a collaboration between scientists who can produce and interpret the comparative data, patent attorneys who understand the three jurisdictions’ examination frameworks, and regulatory scientists who can situate the technical claims in the context of the expected clinical use and regulatory pathway.
Section 8: The Next Generation — Digital Health, AI, and Personalized Medicine
Software as a Medical Device and Digital Therapeutics
The regulatory category of Software as a Medical Device (SaMD) is expanding rapidly and is creating the next wave of patenting challenges for combination products. A therapeutic that pairs a standard drug with an FDA-cleared app using cognitive behavioral therapy algorithms to improve adherence and outcomes is a combination product in regulatory terms [34]. An electroceutical that combines a neuromodulatory implant with AI algorithms controlling stimulation parameters based on real-time neural recording is an even more integrated combination.
These products face two distinct patent challenges that compound the difficulties already discussed. The first is §101 patent eligibility. Software claims in the U.S. must survive the Alice framework’s two-step test: the claim must not be directed to an abstract idea, and even if it is, it must contain an inventive concept that amounts to significantly more than the abstract idea itself [20]. A claim on ‘using a processor to calculate an insulin dose’ is almost certainly directed to an abstract idea with no inventive concept. A claim on a specific neural network architecture trained on a specific dataset to predict hypoglycemic episodes with a specific clinical sensitivity and specificity threshold, implemented on a specific hardware configuration integrated with a glucose sensor, is more defensible — but requires a level of technical specification that many patent applicants are not yet generating.
The second challenge is divided infringement, discussed in Section 3, which becomes the default scenario for any therapy involving a patient, a device, a smartphone, and a cloud server. Method claims on the therapeutic use of a digital therapeutic system are almost impossible to enforce against a single entity. The portfolio must be built on apparatus and system claims wherever possible.
Companion Diagnostics and Biomarker-Stratified Therapies
The growth of precision oncology has created a large and commercially important class of cross-labeled combination products: targeted therapies approved only for patients whose tumors express a specific biomarker, paired with a companion diagnostic test used to identify eligible patients. Pembrolizumab and its PD-L1 companion diagnostics are among the most prominent commercial examples.
The patent strategy for these products must cover both components of the cross-labeled pair. Patents on the biomarker itself, the diagnostic assay technology, the algorithm for interpreting assay results, and the clinical cutoffs that determine patient eligibility are all potentially valuable. So are patents on the method of treating the specific biomarker-defined patient population with the drug — use patents that are narrower than a broad treatment claim but potentially more defensible against obviousness arguments because the biomarker-defined population may not have been an obvious therapeutic target.
The regulatory exclusivity strategy for companion diagnostic combinations is also distinctive. The companion diagnostic, regulated as a medical device, does not carry drug-type regulatory exclusivity. If the diagnostic is third-party manufactured under license, the competitive dynamics are different from a single-company combination product. Ensuring robust patent coverage on the diagnostic methodology and the biomarker-drug relationship is particularly important in this structure.
3D-Printed and Patient-Specific Products
The combination product space is moving toward truly personalized products — drug-loaded implants designed from patient imaging data and manufactured on-demand. A patient-specific, 3D-printed bone scaffold impregnated with antibiotics and growth factors chosen for that patient’s infection profile is not a future concept; it is in active development in academic medical centers and early-stage companies.
The patent strategy for these products requires a fundamental rethinking of claim scope. The blockbuster patent model assumes one product for a large, relatively homogeneous patient population. Patient-specific products, by definition, are different for every patient. This does not eliminate patent value — patents on the manufacturing process, the design algorithm, the drug-loading methodology, and the material science of the scaffold can all protect the commercial technology while covering every patient-specific instance produced using that technology.
Section 9: Practical Synthesis — Integrating Patent and Regulatory Strategy from Day One
The Pre-IND, Pre-RFD Window Is the Most Important Strategic Moment
The most consequential decisions in a combination product’s lifecycle are made before the first clinical study begins. The PMOA determination, the lead center assignment, the choice of regulatory pathway, and the initial patent filing strategy are all set during the pre-IND period. Errors made at this stage propagate through the entire program with increasing cost.
A company developing a novel biologic delivered via a new autoinjector platform should be conducting regulatory strategy and patent strategy in parallel, with explicit coordination, from the early feasibility stage. The regulatory team’s PMOA analysis should inform the patent claim drafting: if the therapeutic effect is primarily attributable to the biologic molecule, the core combination product claims should emphasize the drug-device interaction from the drug’s perspective. If the device enables a therapeutic approach that was previously impossible — as is the case with large-volume subcutaneous wearable injectors for drugs that previously required intravenous infusion — the enabling device function is the inventive story that both the regulatory and patent filings should be built around.
Human Factors as a Patent Asset
Human factors engineering, which the FDA requires for most combination products under guidance documents that reference IEC 62366 and FDA’s own human factors guidance, is commonly treated as a regulatory compliance exercise [35]. It is also a patent opportunity that most companies miss.
The data generated during formative and summative human factors studies — use error rates, close call rates, residual risks, design modifications made to address identified hazards — constitutes a documented record of problems that prior art devices failed to solve. If your device design modifications in response to human factors findings produce statistically significant reductions in use errors compared to prior art devices, that data is exactly the kind of ‘long-felt need / unexpected solution’ evidence that supports a non-obviousness argument. The patent application on the human-factors-optimized design should be drafted at the same time the summative study results are analyzed, not as an afterthought after FDA approval.
Coordination Between R&D, Regulatory, and IP Functions
The organizational structure that fails combination product companies is the one that runs R&D, regulatory affairs, and IP as separate functions with minimal interaction until late in the development cycle. R&D generates data to satisfy its own objectives; regulatory uses that data to build submissions; IP files patents based on what has already been invented and characterized. In this structure, patents are often filed after the core inventive work has already been published in regulatory submissions or scientific papers, and the regulatory data strategy is not optimized to generate the comparative evidence needed for patent prosecution.
The structure that works treats patent prosecution as a consumer of R&D output, just as regulatory submission is a consumer. Patent claim scope is discussed in the earliest feasibility stage, when experimental design can still be modified to generate the comparative data that will be needed to support non-obviousness. Regulatory pathway decisions and patent strategy are coordinated explicitly, because the PMOA determination affects which technical arguments are most important to develop. Human factors findings, formulation development data, and device engineering results are all evaluated for patent value at the time they are generated, not years later.
Key Takeaways
The regulatory philosophy determines the strategic framework. The FDA asks what a combination product is (categorical); the EMA asks what it does (functional). These different philosophies produce different data requirements, different review pathways, and different technical narratives. A global program must be designed to answer both questions from the outset.
PMOA determines your U.S. fate; the NBOp determines your European timeline. In the U.S., the Primary Mode of Action determination assigns your product to a lead center that governs the entire review. File an RFD or Pre-RFD early — the cost of certainty is far lower than the cost of a wrong-pathway submission. In Europe, the Notified Body Opinion for the device component is an independent, mandatory deliverable that can delay the entire MAA if not managed proactively as a parallel workstream.
Post-KSR, non-obviousness requires data, not just narrative. Combining a known drug with a known device is presumptively obvious. The inventive step must be demonstrated through unexpected synergistic effects, quantified by comparative experiments against the prior art. This data must be generated during R&D and disclosed in the patent specification — it is both the non-obviousness argument and the enablement foundation.
A picket fence is not optional. For any combination product with meaningful commercial value, a single patent is a single point of failure. The portfolio must cover the drug, the device (mechanical utility patents and design patents), the drug-device interaction, methods of use, manufacturing processes, and — for connected products — software and algorithms. Multiple overlapping patents requiring a competitor to defeat all of them simultaneously is the only structure that provides durable market protection.
System claims prevent the divided infringement trap. For any connected device where the patient performs steps in the therapeutic method, method claims are largely unenforceable against a single party. Apparatus and system claims, directed to the integrated device a manufacturer makes and sells, are the enforceable foundation of the combination product portfolio.
Global filing requires one rich specification, not three different applications. The EPO’s plausibility doctrine and the JPO’s emphasis on demonstrated advantageous effects both require that the original specification contain robust comparative data. U.S. provisional practice of ‘file early, supplement later’ is dangerous in a global filing strategy. Wait until the data is solid, then file internationally.
The antitrust dimension is a hard constraint on IP strategy. An aggressive patent and commercial strategy that, viewed in combination, forecloses competition unlawfully — through patent thickets, pay-for-delay settlements, regulatory process abuse, and exclusionary commercial agreements — exposes a company to antitrust liability that can exceed half a billion dollars. The EpiPen litigation is the reference case. Every IP strategy for a market-leading combination product requires antitrust review of the combined commercial and IP posture, not just the individual components.
Frequently Asked Questions
1. Our biologic drug is the clear therapeutic driver in our combination product. Does that mean we should focus all our IP resources on the drug and treat the device as a secondary priority?
No, and companies that take this approach typically regret it within five years of launch. The device is your primary lifecycle management tool. The biologic exclusivity period runs for 12 years from FDA approval, but as biosimilar applicants prepare to enter the market, they must also design a delivery system. Device patents filed during the product’s commercial life — covering second-generation autoinjector improvements, human-factors-optimized designs, and connected device features — can create barriers to biosimilar market entry that extend years beyond the biologic exclusivity expiration. A biosimilar manufacturer who can make your drug’s active ingredient still cannot launch if every commercially viable delivery device is covered by an enforceable device patent. That leverage disappears if you treated the device as a commodity.
2. We are preparing to file a global patent application on a novel drug-device interaction we have discovered. How early should we file, given the tension between securing a priority date and generating enough data?
This requires an honest assessment of what you have and what the EPO’s plausibility standard requires. If you have preliminary data showing a trend — say, reduced aggregation of your biologic in a novel polymer container, based on a 3-month accelerated stability study — you likely have enough for a U.S. provisional that establishes a priority date. Use the 12-month priority period to run the full 24-month stability study, generate head-to-head comparisons against the closest prior art container system, and characterize the mechanism of the effect. Then file the full PCT application with that complete data package. If you file PCT immediately based only on the preliminary data, you risk EPO and JPO rejections that post-filing data cannot fix. The priority date is less valuable than it appears if the application fails examination in your two largest ex-U.S. markets.
3. We have designed a combination product that involves a patient using a connected app to interact with a drug delivery pump. Our lawyers want to focus on method claims. Is that right?
The method claims are not worthless, but they should not be the core of your enforcement strategy. Method claims for a system involving a patient, a device, and a cloud server almost certainly trigger divided infringement problems — no single entity performs all the steps. Your primary commercial protection should come from apparatus and system claims directed to the integrated product. ‘A drug delivery system comprising: a pump; a sensor array; a processor executing instructions to analyze sensor output and determine a delivery protocol; and a wireless communication module transmitting delivery data to a remote server’ is an apparatus claim that a manufacturer infringes by making and selling the system. The method of using that system to treat a patient can be claimed separately as contributory or induced infringement if the method claims are drafted carefully, but do not build your enforcement strategy on them as standalone claims.
4. We discovered our combination product effect entirely through routine formulation development, not through a designed experiment to identify synergy. Can we still claim unexpected results for non-obviousness?
Yes, but you need to characterize and document the effect rigorously before filing. The fact that a discovery occurred incidentally during routine work does not preclude patenting it — unexpected discoveries are common in pharmaceutical development. What matters is that the effect is real, demonstrable in comparative experiments against the prior art, and actually unexpected based on what a skilled artisan would have predicted from the prior art. Generate the comparative data: test the prior art combination, test your combination, and measure the relevant endpoints side by side. If your combination genuinely outperforms the prior art in a way that the prior art’s teachings did not predict, that is your non-obviousness story regardless of how you found it.
5. We are entering a co-development partnership where we are providing the drug and our partner is providing the device. The combination product is new and valuable. Who owns the IP on the combination product itself?
This is the question that partnership lawyers are most likely to underestimate and that courts see most frequently after a partnership breaks down. The answer depends entirely on what you negotiate before the collaboration begins — and you need to negotiate it in writing with specificity before any joint development work starts.
Key questions to resolve: Who owns background IP each party brings to the partnership, and what license rights does each party have to use the other’s background IP? Who owns foreground IP created jointly during the collaboration — specifically, the patent on the novel combination product, its unique drug-device interaction, and any improvements developed jointly? Is joint ownership acceptable, or does one party acquire full ownership in exchange for royalty obligations? What happens to each party’s rights in the jointly developed IP if the collaboration terminates early, and can either party independently license the joint IP to a third party? These questions sound structural and lawyerly, but the answers determine who controls the most commercially valuable asset that emerges from the partnership. The party who controls the combination product patent controls the market. Discover that ambiguity in discovery during litigation and you have already lost.
References
[1] Grand View Research. (2024, March). U.S. Drug Device Combination Products Market Size, Share & Trends Analysis Report. Grand View Research. https://www.grandviewresearch.com/industry-analysis/us-drug-device-combination-products-market-report
[2] U.S. Food and Drug Administration. (n.d.). Combination product definition combination product types. https://www.fda.gov/combination-products/about-combination-products/combination-product-definition-combination-product-types
[3] Premier Research. (n.d.). Do you have a combination product and is it drug-led or device-led? Early regulatory insights. Premier Research. https://premier-research.com/perspectives/do-you-have-a-combination-product-and-is-it-drug-led-or-device-led-early-regulatory-insights/
[4] U.S. Food and Drug Administration. (n.d.). Frequently asked questions about combination products. https://www.fda.gov/combination-products/about-combination-products/frequently-asked-questions-about-combination-products
[5] BSI Group. (n.d.). An overview of drug-device combination products in the EU. BSI Compliance Navigator. https://compliancenavigator.bsigroup.com/en/medicaldeviceblog/an-overview-of-drug-device-combination-products-in-the-eu/
[6] Medidee. (2023, January 26). Combination products: Similarities and differences of EU and US regulations. https://medidee.com/2023/01/26/article-combination-products-similarities-and-differences-of-eu-and-us-regulations/
[7] U.S. Food and Drug Administration. (n.d.). About combination products. https://www.fda.gov/combination-products/about-combination-products
[9] U.S. Food and Drug Administration. (2022, January). Principles of premarket pathways for combination products. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/principles-premarket-pathways-combination-products
[10] European Medicines Agency. (2019, May 29). Draft guideline on quality requirements for drug-device combinations (First version). https://www.ema.europa.eu/en/documents/scientific-guideline/draft-guideline-quality-requirements-drug-device-combinations-first-version_en.pdf
[11] KSR International Co. v. Teleflex Inc., 550 U.S. 398 (2007).
[12] U.S. Patent and Trademark Office. (n.d.). MPEP 2163: Written description requirement. https://www.uspto.gov/web/offices/pac/mpep/s2163.html
[13] U.S. Patent and Trademark Office. (n.d.). MPEP 2164: The enablement requirement. https://www.uspto.gov/web/offices/pac/mpep/s2164.html
[14] Sufficient Description. (2022, August). Divided infringement and problems of attribution. http://www.sufficientdescription.com/2022/08/divided-infringement-and-problems-of.html
[15] Akamai Technologies, Inc. v. Limelight Networks, Inc., 797 F.3d 1020 (Fed. Cir. 2015).
[16] Medical Design & Outsourcing. (2018, January 25). The 5 biggest issues in patenting combo products. https://www.medicaldesignandoutsourcing.com/5-biggest-issues-patenting-combo-products/
[17] PatentPC. (n.d.). Biopharmaceuticals: The patent implications of combination therapies. https://patentpc.com/blog/patent-implications-of-combination-therapies
[18] Kesselheim, A. S., et al. (2016). Is patent ‘evergreening’ restricting access to medicine/device combination products? PLOS ONE, 11(2), e0148939. https://doi.org/10.1371/journal.pone.0148939
[19] National Institutes of Health. (2023, October). Regulatory knowledge guide for combination products. https://seed.nih.gov/sites/default/files/2024-04/Regulatory-Knowledge-Guide-for-Combination-Products.pdf
[20] Alice Corp. Pty. Ltd. v. CLS Bank International, 573 U.S. 208 (2014).
[21] Congressional Research Service. (2024, January 30). The role of patents and regulatory exclusivities in drug pricing (Report No. R46679). https://www.congress.gov/crs-product/R46679
[22] U.S. Food and Drug Administration. (n.d.). Small business assistance: Frequently asked questions on new drug product exclusivity. https://www.fda.gov/drugs/cder-small-business-industry-assistance-sbia/small-business-assistance-frequently-asked-questions-new-drug-product-exclusivity
[23] Feldman, R., & Wang, A. (2025). Patent thickets and product hops: Challenges and opportunities for legislative reform. Journal of Law, Medicine & Ethics.
[24] DrugPatentWatch. (2024, November 4). Patents for combination products: Challenges and opportunities. https://www.drugpatentwatch.com/blog/patents-combination-products-challenges-opportunities/
[25] MassDevice. (2023, February 2). Boston Scientific hit with $42M verdict in drug-eluting stents IP case. https://www.massdevice.com/boston-scientific-hit-with-42m-verdict-in-drug-eluting-stents-ip-case/
[26] Minding Your Business Litigation Blog. (2022, July 29). $264 million settlement in EpiPen price gouging litigation. https://www.mindingyourbusinesslitigation.com/2022/07/264-million-settlement-in-epipen-price-gouging-litigation/
[27] Proskauer Rose LLP. (2022, July 29). $264 million settlement in EpiPen price gouging litigation. https://www.proskauer.com/blog/264-million-settlement-in-epipen-price-gouging-litigation
[28] Seven Pillars Institute. (2017, September 14). Mylan’s EpiPen pricing scandal. https://sevenpillarsinstitute.org/mylans-epipen-pricing-scandal/
[29] FTC v. Actavis, Inc., 570 U.S. 136 (2013).
[30] California Lawyers Association. (n.d.). EpiPen court finds tying arrangement involving identical products and rejects Noerr-Pennington defense in face of sham citizen petition claims. https://calawyers.org/antitrust-unfair-competition-law/epipen-court-finds-tying-arrangement-involving-identical-products-and-rejects-noerr-pennington-defense-in-face-of-sham-citizen-petition-claims/
[31] European Patent Office. (n.d.). Comparative study on hypothetical/real cases: Inventive step/non-obviousness. https://link.epo.org/trilateral/Comparative_Study_on_Cases_Inventive_Step.pdf
[32] Institute of Intellectual Property. (2015). Comparative analysis of inventive step/nonobviousness standard and case study thereof. https://www.iip.or.jp/e/summary/pdf/detail2015/e27_summary_Chou.pdf
[34] Frontiers in Medicine. (2022, July 21). The regulatory challenges of innovative customized combination products. Frontiers in Medicine, 9, 821094. https://doi.org/10.3389/fmed.2022.821094
[35] U.S. Food and Drug Administration. (n.d.). Applying human factors and usability engineering to medical devices: Guidance for industry and FDA staff. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/applying-human-factors-and-usability-engineering-medical-devices
‘The global market for combination products is expected to reach $177.7 billion by 2026, growing at a CAGR of 9.2% from 2021 to 2026.’
— MarketsandMarkets Report, as cited by DrugPatentWatch
