Industries
Solutions
Capabilities
About
Insights
The drug delivery device landscape is rich with brilliant ideas that never quite made it to patients or were launched but achieved limited commercial success. Some failed quietly in development. Others launched with fanfare only to be withdrawn. A few still linger in limbo – technically alive but commercially dormant. Together, they form a fascinating catalogue of ambition, engineering ingenuity, patient focus, regulatory friction, and market reality.
Below is a tour through the major categories of devices that have struggled to achieve their potential. We cover why they struggled, and what their stories reveal about the future of drug delivery.
This article was written in the spirit of learning from common challenges and educating the next generation of drug delivery pioneers. The teams behind these innovations often brought a wealth of talent and expertise. Most of the device categories discussed below have protagonists that currently are working hard to bring their devices to market or increase their market share if on the market. This article does not take a view on their merits or future chances of success. Rather, it is an opportunity to learn from experiences where others have struggled so that, together with our peers in the drug delivery devices ecosystem, we can maximize the success of innovation so that patients have devices that are increasingly effective, usable, safe, and economical.
1. Needle-free Injectors: A Perennial Promise That Rarely Sticks
Needle-free injection has been a recurring dream for decades: painless, fast, sharps-free, and potentially self-administered. Yet relatively few devices have made it to market. Needle-free technology has achieved most of its success in the vaccine market where HCPs provide the administration.
Challenges
· Inconsistent dose delivery – especially with viscous biologics. Dose volume can be controlled, but jet geometry and penetration depth are strongly dependent on the formulation viscosity. Most drug manufacturing systems control the concentration of active pharmaceutical ingredients or biologic. Viscosity can vary significantly because the curve of viscosity versus concentration tends to be steep and non-linear at high concentrations.
· Skin variability — age, hydration, and anatomy all affect jet penetration.
· Pain perception — needle-free does not mean pain free, and the pain studies have not convinced the market to switch to needle-free drug delivery.
· Complexity and cost — devices were often bulky, expensive, required special training, and/or manufacturing processes could be more complex.
· Non-standard containers – most implementations require drug containers that are different from standard pharmaceutical cartridges and syringes. These bring risks in drug stability, filling, and device verification such as dose accuracy and free-fall tests.
· Regulatory caution—cross contamination concerns after early multiuse jet injectors.
Notable examples
· Comfort-In is a spring-based jet injector kit available in 4 countries plus the European Union.
· Injex30 is approved for use in the US, EU, and China with a special focus on injecting anesthetics for dentistry.
· PharmaJet Stratis/Tropis: used in vaccination settings but has not yet achieved significant injector market share compared to needle-based devices.
· Biojector: widely used in the 1990s and 2000s, eventually discontinued.
· Zogenix DosePro: approved for sumatriptan; commercial uptake limited; discontinued product due to commercial reasons.
· Merck Group explored a needle-free injector for Rebif called RebiJet in the mid-2000s which was based on the Intraject technology. It never reached commercial launch, probably due to the shift of focus to the RebiSmart reusable electromechanical injector and the Rebidose single-use autoinjector. The technology went on to become Crossject’s Zeneo which is still in development.
· Portal Instruments developed Prime, an electromechanical needle-free injector which is not commercialized yet. The company has now adopted the core technology for a needle-based reusable electromechanical autoinjector using pre-filled syringes.
Needle-free injectors are technologically impressive but have struggled to achieve the commercial success of needle-based injectors.
2. Transdermal Patches: The Rise, Plateau, and Quiet Retreat
Transdermal delivery had its golden era with patches developed to deliver medicines such as:
• Nicotine replacement therapy,
• Pain management e.g. fentanyl or morphine,
• Contraceptives,
• Hormone replacement therapy e.g., estrogen or testosterone,
• Cardiovascular treatment e.g., nitroglycerin, or
• Motion sickness reduction e.g. scopolamine.
However, the next generation of “active” or “enhanced” patches often failed to reach the market.
Challenges
Skin is a formidable barrier, especially for large molecules.
· Microneedle patches faced challenges with drug load, repeatable pharmacokinetics, manufacturing, stability, usability, skin reactions and regulatory hurdles.
· Thermal ablation patches never scaled beyond early trials.
Notable examples
· Exelon Patch was approved in 2025 for the treatment of mild to moderate Alzheimer’s disease. Developed by Noven Pharmaceuticals, the patch delivers the drug, rivastigmine, through the skin, offering a more convenient alternative to existing oral medications
· Gylden Pharma’s Titanium Microneedle Platform is a titanium microneedle patch developed by ALZA, spun out as Macroflux which became Zosano. Zosano had a partnering deal with Novo Nordisk (2014-2015) on insulin delivery but ultimately filed for bankruptcy in 2022 when the FDA requested more bioequivalence data. Gylden Pharma is pending regulatory approval for using the technology to deliver zolmitriptan.
· PassPort uses an applicator to create microscopic porous channels in the skin before a patch is applied. The company’s lead preclinical candidate is a CNS therapeutic in development in the US via the 505(b) (2) pathway and they released interim results from a Phase 1 trial in 2024.
· Various iontophoretic patches have been launched. These drive drug through the skin using electric fields. One example, Ionsys for delivering fentanyl, was launched in the European Union and United States for postoperative pain in 2006 but was recalled from the EU in 2008 due to a corrosion issue that could cause unintended fentanyl release and was withdrawn in 2009. It was relaunched in some markets in 2015 then voluntarily withdrawn from the US market in 2017 for commercial reasons. Most current iontophoretic patches are used for physiotherapy. They tend to be medical devices rather than combination products, and the drugs they are used for include hydrocortisone, acetic acid, dexamethasone, and lidocaine. Examples are IontoPatch and InotoGo.
· Sonophoretic patches such as SonoPrep from Echo Therapeutics use ultrasound to disrupt the stratum corneum and enable drug to penetrate the skin. SonoPrep was approved in 2004, but Echo pivoted towards its sonophoresis-enabled glucose monitoring platform and the company shut down in 2016.
The transdermal patch field continues to attract interest such as several active investigations for intradermal vaccine delivery.
3. Gas Powered Injectors: Elegant Engineering, Limited Adoption
Gas powered autoinjectors use compressed gas (often CO₂, nitrogen or argon) to drive a plunger. They offer smoother injection profiles than springs and the ability to inject viscous biologics in a reasonable injection time.
Challenges
· Cost and complexity—gas cannister add manufacturing and regulatory burden.
· Temperature sensitivity—gas pressure varies with ambient conditions.
· Bulky form factors—gas-powered injectors can be larger compared to spring-based devices once piercers, valves, and seals are included. However, gas power is typically used in applications that require more force or energy than can be provided by a spring so comparisons must be fair.
· Needs clear clinical advantage to justify premium pricing.
Examples that made it to market
· Auvi-Q was developed to deliver epinephrine for anaphylactic shock. It was launched in 2012 but recalled in 2015 due to the risk of inaccurate dose delivery. It was relaunched in 2017, and the same platform is used to deliver naloxone in the Rapid Opioid Countermeasure System.
· The iJect / BioJect family mentioned above.
Examples that were cancelled
· Ovaleap’s gas powered injector prototypes were never commercialized.
· Bespak Syrina and Viscala were never commercialized.
Examples that are in development
· Windgap DualFlo, OneMix, and MultiMix.
· CrossJect Zeneo
· AltaViz AltaVISC
· Congruence Autoinjector
· SMC’s Bios platform
Spring-based autoinjectors are difficult to beat on simplicity and cost for drug formulations with modest dose volume and viscosity. Other actuator technologies such as motors tend to compete with gas power where large dose volumes or high viscosities need to be delivered.
4. Connected Devices: A Wave of Smart Tech That Has Not Yet Found Its Market
The mid-2010s saw a surge of enthusiasm for “smart injectors,” Bluetooth-enabled autoinjectors, smart inhalers, adherence trackers, and cloud connected drug delivery ecosystems.
Challenges
· Low patient demand—most patients don’t want or need connectivity.
· Reimbursement gaps—payers rarely fund digital addons.
· Cybersecurity and privacy concerns.
· Device complexity—more components, more failure modes.
· Timing with respect to drug lifecycles—hard to justify connected device development at the time of drug development, or multiyear digital investments.
Pharma had a difficult time looking beyond the drug product, struggled where to place the responsibility in the organization, and how to define Return-On-Investment.
Notable examples
· Merck Group’s easypod was launched in 2007 and its second generation introduced connectivity with easypod Connect in 2011. It continues to be on the market, now in its third generation.
· Connected insulin pens have found market success with an estimated value of USD 905 million and an estimated Compound Annual Growth Rate of 11.96%. Biocorp’s Mallaya and Innovation Zed’s InsulCheck DOSE™ are connected caps for a variety of disposable insulin pens. The Common Sensing GoCap insulin pen add-on became part of the Bigfoot Unity diabetes management system, now owned by Abbott. Several other connected pen caps have come and gone, such as Timesulin.
· Propeller Health and Adherium developed smart add-ons for inhalers but had limited uptake. Propeller was acquired by Resmed in 2018 then shut down. Adherium’s Hailie platform is commercialized in the USA with shipments of around 1400 devices per quarter.
· Teva launched the connected inhaler Digihaler for ProAir, ArmonAir, and AirDuo starting in 2019 but withdrew the devices from the market in April 2024.
· Bayer markets the BETACONNECT connected electronic autoinjector for Betaseron to treat multiple sclerosis.
· There are other connected autoinjector pre-market programs such as SHL’s Elexy and autoinjector add-ons such as Ypsomed’s SmartPilot, which has 510(k) clearance.
· Lilly is going to discontinue the Tempo™ Personalized Diabetes Management Platform at the end of 2026.
The lesson: connectivity must solve a real problem, not just add a feature. The value proposition (business case) needs to be convincing, and the benefits to patients and clinicians must be compelling if they are being asked to perform any actions related to connectivity.
5. Solid Dose and Powder Injection Systems: High Hopes, Hard Realities
Solid dose injectors and powder jet systems promised needle-free delivery of vaccines and biologics using microprojectiles or high velocity powder streams.
Challenges
· Dose variability—especially with powder dispersal. For example, one study on excised human skin membranes showed that only 50% of the powder formulation left the device and the ejected powder had significant comminution [fragmentation].
· Complex device mechanics.
· Regulatory uncertainty—novel routes raise safety questions.
· Cold chain and stability issues—ironically, many systems did not eliminate cold chain needs.
Notable examples on the market
· Solid dose implants for ophthalmic use such as Iluvien, Ozurdex, and Durysta. These are solid doses but are delivered via a needle so they are distinct from the needle-free solid dose injector such as aVaxMed and PowderJect.
Notable examples in development
· aVaxMed: Spring-driven solid dose injector. Formerly Glide Pharma then Enesi Pharma.
Notable examples that have ceased development
• PowderJect: the technology injected a stream of dry powder particles. It was discontinued before reaching the market, probably due the changing competitive environment: in the late 1990s disposable syringes improved, needlestick protection systems became more widely adopted, autoinjectors became common, and needle-based vaccine administration remained inexpensive.
Lessons from the experiences of solid dose injectors include:
• Solid doses are a compelling option when extended release is required, for example in ophthalmic implants.
• For other applications such as vaccines, the value proposition must be convincing versus needle-based liquid injections, especially in the modern era where cold chain distribution and needlestick protection have improved greatly.
6. Inhaled Biologics: The Exubera Effect and Its Aftermath
Inhaled insulin was once heralded as the future of diabetes care. If inhaled insulin had been very successful, then it would have paved the way for many more inhaled systemic biologics. Instead, it became a cautionary tale.
Challenges
· Large devices (notably for Exubera, less so in other cases such as Afrezza).
· Pulmonary safety concerns—long term effects of inhaled proteins remain uncertain.
· Variable absorption—affected by lung function, technique, and comorbidities.
· Cost and reimbursement challenges.
· Competition from improved injectable pens and pumps.
Notable examples
· Exubera (Pfizer): launched 2006, withdrawn 2007.
· Afrezza (MannKind): available on the market but not yet a blockbuster at USD 75m sales per year.
· Lilly-Alkermes AIR insulin system which reached Phase 2 clinical trials.
· Novo Nordisk-Aradigm AERx iDMS (liquid-based insulin system) reached Phase 3 clinical trials. The liquid strips and complex electronic device would have led to high cost-of-goods.
· Inhaled GLP1 and other biologics—multiple programs discontinued.
The lungs are a promising but unforgiving route. Dry powder inhalers have limits around dose mass and drug suitability for micronization. Pressurized metered dose inhalers present formulation challenges with approved propellants. Nebulizers do not have these drawbacks and can be used in Phase 1 and 2 clinical trials onwards, but a successful platform needs to support easy adaptability to the given drug formulation and viscosity.
7. On-body Injectors: A Category With Successes And High Potential but Several Casualties
On-body injectors (OBIs, also called “patch pumps”, “On Body Delivery Systems”, or “bolus injectors”) are designed for self-administration of large volume biologics, ideally negating the need for patient visits to infusion centers.
This category of devices has seen some notable successes and shows that new device categories are possible if the clinical need is present, the technology is not too risky, and the regulatory path is like other successful device formats.
Challenges
· Skin adhesion or reaction issues.
· User anxiety—patients uncomfortable wearing a device for more than a few minutes.
· Complexity and cost—more components than prefilled syringes or autoinjectors.
· Manufacturing challenges—sterility, assembly, and reliability.
· Drug device integration issues—especially with viscous or sensitive biologics.
· Technology unfamiliarity and patient adoption perception with stakeholders.
Notable examples of devices which have reached the market
· West Pharmaceutical Services SmartDose:
o Launched then discontinued in the 3.5 mL variant for Amgen’s Repatha. Patients were advised to switch from using one on-body injector to three autoinjectors.
o Launched for AbbVie’s Skyrizi in 2020. AbbVie acquired manufacturing and supply rights to SmartDose 3.5 mL in January 2026
o Approved in 2022 but never launched for Alexion’s Ultomiris (ravulizumab).
o The 10 mL variant was approved in the USA in October 2022 for scPharmaceuticals Furoscix to treat congestive heart failure. It delivers 80 mg of Furoscix over 5 hours.
· Neulasta Onpro (Insulet). At the time of writing, more than 1 million patients have used Neulasta Onpro.
· EMPAVELI Enfuse (Enable Injections) approved by FDA in 2023.
· UDENYCA Sorrel (LTS Lohmann) was launched in 2024.
· Lasix Onu SensAir (Gerresheimer) was approved by FDA in 2025.
Notable examples of devices in development
· Prefilled single use on-body injectors such as BD’s Libertas and Ypsomed’s YpsoDose.
· Part-reusable motor-driven on-body injectors such as Stevanato Group’s Vertiva and Nemera’s Symbioze.
· Subcuject’s osmosis-driven on-body injector.
Several on-body injectors have been discontinued such Ratio Inc.’s hydrogel-powered injector or Bespak’s gas-powered Lapas.
8. Nasal Delivery Of Biologics: Can Be Fast and Simple, But Variable
Nasal delivery has been successful for small molecules (e.g., naloxone, sumatriptan, epinephrine) but attempts to expand into biologics and vaccines have been rocky.
Challenges
· Mucociliary clearance—drug is swept away quickly.
· Limited absorption of large molecules.
· Irritation and tolerability issues.
· Device variability—spray plume geometry, delivered dose, and droplet size distribution depend on pump pressure (therefore typically the force provided by the user), nozzle tolerances, and formulation properties (viscosity and surface tension for liquids, particle morphology and electrotribology for dry powders).
· Notable examples
· Nasal insulin programs—multiple discontinued.
· Nasal vaccines— many candidates failed to show adequate immunogenicity, but FluMist (influenza virus subtypes A and B) got FDA approval first in 2012 and in 2024 for self or caregiver administration
· Baqsimi (glucagon) delivered by Aptar’s Unidose powder device.
· NOSA and Hogne TrigeNas nose plugs in development.
Nasal delivery remains attractive but technically unforgiving.
9. Why So Many Device Categories Struggle
Across categories, the same patterns recur:
· Biology beats engineering
The human body is not a passive receptacle. Skin, lungs, nasal mucosa, and subcutaneous tissue all impose constraints that clever engineering cannot always overcome.
· Simplicity wins
· Patients, clinicians, and payers gravitate toward devices that are:
· Small
· Intuitive, minimal use steps
· Reliable
· Low technology complexity, risk and cost
· Anything more complex must deliver a clear clinical benefit.
· Integration is hard
Drug device combination products require:
· Timing for market pull and technology push must be right: the pharmaceutical company needs the product just at the right time
· Stability compatibility
· Regulatory coordination, preferred technology has been already approved by a regulatory agency
· Manufacturing harmonization e.g. using a standard primary container enables reuse of filling technology, and provides potential as platform technology
· Speed to market to stay ahead
Many programs fail at these interfaces.
The market is ruthless
Even technically successful devices can die if:
· Reimbursement is weak
· Training burden is high
· Competitors launch simpler alternatives
· The target population might be small
Adoption takes time
Successful launches of products by startup companies tend to take more than 10 years. For example, Kaleo (formerly Intelliject) was founded in 2000 and launched Auvi-Q in 2012, Sensile Medical (now part of Gerresheimer) was founded in 2004 and launched the D-MINE pump in 2019, and Enable Injections was founded in 2010 and launched enFuse in 2023.
10. What We Can Learn About the Future
Despite the challenges, innovation and evolution in drug delivery devices continue at pace. The next generation of devices will likely succeed if they:
· Solve a real clinical or workflow problem
· Reduce burden, not add features
· Integrate seamlessly into existing care pathways
· Demonstrate clear economic value
· Apply to diverse patient populations
· Get the right fit and timing for market pull and technology push
· Provide advantage versus competitor devices
· Address why similar technologies struggled
Reference lists available on request.
Written by: Toam Oakley, VP Strategic Partnership, Sanner and Ulrich Bruggeman, Takeda.
Tom Oakley, Vice-President Strategic Partnerships at Sanner Group (previously Springboard), leads engineering and scientific teams developing new injection devices, pumps and inhalers.
Ulrich Bruggemann, Lead, Advanced Drug Delivery Technical Development, at Takeda Pharmaceuticals, has been working on the development of medical device and diagnostics technologies and products for over 30 years in the US and Germany. Mr Bruggemann is an employee of Takeda Development Center Americas, Inc, a subsidiary of Takeda Pharmaceutical Inc and owns stock.