Disruptive innovation has defined the history of the medical device industry. Novel devices and procedures have replaced established medical practices in various medical markets for decades. For example, the first artificial knees and hips were developed in the 1930s, the first implantable pacemaker in 1958, and the first implantable defibrillator in 1980. Each innovation ushered in new levels of improved patient outcome and reduced patient risk, raising the standard of healthcare at that time. The defining characteristic for each major market and clinical success involved innovative product design that focused on meeting underserved patient and market needs.
Although neurostimulation was born several decades ago, the industry has reached an inflection point that is ripe for disruptive innovation. Disruptive innovation refers to creating new markets by applying a different set of values, which often occurs by redefining user needs, product strategy, and business models.1 This is a more comprehensive and sustainable approach to product design compared with generating disruptive technology, a term that refers to a single technology or product that creates market advantage. Companies that have an edge in the market usually do so by creating an unexpected new market space or by increasing marketshare. Manufacturers of neurostimulation technologies can achieve disruptive innovation via designing better user interfaces, incorporating effective device-to-device communication, and finding fast and efficient methods to get products to market.
This article discusses the state of the neurostimulation field from market, clinical, and technology perspectives as well as product development opportunities and how disruptive innovation can generate improved clinical value for patients and the market.
|Figure 1. Applications for the main therapies could expand in the next few years.|
Clinical Indications and the Market
Neurostimulation devices are active implants that help manage conditions such as epilepsy, Parkinson’s disease, depression, and chronic pain. They contain sophisticated electronics packaged within a hermetically sealed titanium case. The devices apply low-level energy to the nervous system to block nerve signals between the brain and peripheral tissue or organs.
Neurostimulation has historically been a last resort for patient treatment. Pharmaceutical or physical therapy was traditionally the first line of defense, with neurosurgery being the next best option for treating refractory diseases—the incidence of which is relatively high for neural diseases. For example, 30–40% of epilepsy patients continue to have seizures while receiving the maximum dose of medicine.2
There are currently four publicly held companies with established revenue from neurostimulators—Medtronic Inc., Boston Scientific Corp., St. Jude Medical Inc., and Cyberonics Inc. Each company has a specific market niche, with little overlap. As these resource-rich companies expand their platform technology, competition should greatly increase. There are more than 20 start-up companies developing neurostimulators to manage a variety of diseases. These companies are financially backed by venture capital firms and funding from the development wings of the aforementioned publicly held companies. Their products are at various stages of development, and many are expected to enter clinical trials within the next five years.
There are three main modalities of implantable neurostimulation therapy—vagus nerve stimulation (VNS), spinal cord stimulation (SCS), and deep brain stimulation (DBS) (see Figure 1).
VNS. A VNS product manufactured by Cyberonics Inc. was first approved for managing epilepsy in 1997. The vagus nerve, which controls the state of various organs, extends from the brainstem to the neck and chest, and into the abdomen. The general surgical technique for managing epilepsy involves implanting a pulse generator and lead subcutaneously in the neck, with helical contact electrodes wrapped around the left vagus nerve. However, VNS is not yet the gold standard for epilepsy. It does not replace surgery for medical refractory epilepsy, but it is considered a viable alternative therapy.3 VNS is also being explored for managing other diseases, particularly those related to organs and muscle function connected to the vagus nerve. For example, EnteroMedics is developing a device that helps regulate stomach and pancreas activity to help manage obesity.
DBS. DBS is currently indicated for managing Parkinson’s disease, essential tremor, and dystonia. Although DBS involves inserting an electrode into the thalamus of the brain, this therapy has a significant advantage over destructive neurosurgery that resects brain tissue. To date, Medtronic has the only FDA-approved DBS device. However, by stimulating different parts of the brain, DBS is being explored for multiple new clinical indications. Neuropace Inc. is developing a brain stimulator for managing epilepsy, and IntElect Medical is creating a device that will help patients recover from stroke and traumatic brain injury. These new indications could have good reimbursement potential, largely due to the poor efficacy of drug therapy and high risks of neurosurgery.
SCS. SCS is mainly used for chronic pain management when physical therapy and drug therapy are ineffective. It is based on the gate control theory, which postulates that nonpainful stimuli can inhibit the awareness of painful stimuli.4 A pulse generator is placed subcutaneously in the lower back with electrodes that target nerve fibers in the dorsal column of the spinal cord. The pulse generator applies up to 25 mA of continuous current, which is a higher energy output than VNS or DBS. As a result of the higher energy output, SCS devices have resorted to using batteries within the pulse generator that are recharged via an external power source. Once implanted, the pulse generator is programmed by a physician, and patients can modulate their therapy with a handheld interface. Both Boston Scientific and St. Jude Medical have FDA-approved SCS devices.
|Figure 2. New devices should integrate more advanced technology.|
It should be noted that neurostimulation therapies do not necessarily treat neural diseases. They manage a disease’s symptoms with the intent to minimize its effects on patients’ lives. Each therapy modality poses different challenges and requires different technology and system architecture (see the sidebar, “Neurostimulation System Architecture”). While VNS, SCS, and DBS are the main therapy modes of neurostimulation, other modes are being developed for various clinical indications. Start-up companies will help expand the benefit that neurostimulators can provide to patients. For example, CVRx Inc.has developed a device that helps improve blood flow by stimulating baroreceptors on the carotid artery, combating hypertension and heart failure. Apnex Medical Inc. is developing a system that manages sleep apnea by stimulating throat muscles during sleep.
For neurostimulators, efficacy remains relatively low at 30–60%. In contrast, conversion efficacy for implantable defibrillators is typically in the low- to mid- 90% range.5 There are several reasons why clinical efficacy has been low for neurostimulators. The industry as a whole is still defining a successful neurostimulation outcome, and this makes it difficult to design a device that delivers optimal clinical value. Neurostimulation covers a broad spectrum of applications, such as motor control, and managing pain depression, obesity, and incontinence. With such varied clinical fields, each application has its own definition of success, which has yet to be fully characterized.
In addition, the actual mechanism for neurostimulation is not well understood, and the mechanism for VNS is largely unknown. This black hole makes it challenging to understand how therapy can achieve the optimal clinical outcome. Some therapy parameters are programmable, such as signal current, frequency, pulse width, and on-off time between pulses. Thus, postimplant, the physician can alter the parameters to optimize device performance for patients. However, the optimization process is slow and uncertain—it sometimes takes up to a year to achieve clinical improvement.6
Since abnormal neural behavior is not yet well characterized, companies have taken a conservative approach to delivering therapy. Most neurostimulation devices operate by applying energy to a specific area of the body indefinitely, without regard to whether the abnormal neural behavior is occurring. This has historically worked, especially in managing chronic pain, where abnormal neural behavior is continuous and somewhat predictable. However, this tactic uses unnecessary device power when managing diseases in which symptoms are intermittent or vary in severity, such as epilepsy.
Neurostimulation devices that are developed within the next five years should be able to sense the onset of a neurological episode, use an algorithm to determine if therapy is appropriate and optimized, and then deliver therapy. These closed-loop devices will depend on accurate sensing, which in turn will depend on a variety of factors such as a well-defined database of brain activity, electrode and sensor placement and fidelity, and smart algorithms for signal processing (see Figure 2). These improvements, coupled with advances in power management, telemetry communication and patient-centric design, should translate into improved device efficacy.
Sensing, Detection, and Therapy. A database of brain waves within the neurostimulation industry has yet to be broadly disseminated and characterized. NeuroVista Corp. is tackling this issue by developing an implantable device that senses and detects neural activity to predict the onset of seizures in epilepsy patients. This is the first major commercial attempt at signal characterization for neural pathologies. As this sensing technology is further developed and evaluated, it may provide the critical input for therapy algorithms that will result in truly disruptive innovation.
However, implementation of such technology is challenging. Device architecture must facilitate a neurostimulator’s ability to sense, detect, and deliver therapy. Microelectronic design—the brains behind the device—is one of the key architecture elements to achieve this device intelligence. The microelectronics must meet a variety of power, speed, processing, and memory requirements, all with strict quality and reliability performance.
Analog and digital application-specific integrated circuit (ASIC) development is also invaluable, because it allows a company to customize platform architecture and develop a robust product design. However, ASIC development can take months to complete and upwards of $1 million in resource time and materials. To significantly reduce business risk, companies can conduct emulation on their ASIC chips. Emulation enables the ASIC chip to be simulated and tested with real-time monitoring of performance. Ensuring that the ASIC chip and microelectronics are providing the best possible architecture for software and firmware will lead to improved signal sensing, detection, and therapy.
Electrode Design and Placement. Historically, electrode designs have not been optimized for neurological anatomy. Initial electrode designs were transferred from cardiac pacing and defibrillation. Designs have since evolved to include multinode electrode arrays, allowing for flexibility in programming. Further innovation is needed in future electrodes for the following three reasons:
- Improved electrode designs can provide more directionality of applied therapy.
- If the electrode can minimize localized scar tissue, the sensitivity of signal sensing can improve.
- The electrode mechanical design should facilitate simple and more accurate insertion, because electrode placement is vital for sensing and delivering therapy.
It still remains a challenge to accurately visualize and affix electrodes during surgery, even though it is easy to visualize electrodes postoperatively. This poses a problem, because it is difficult to change the electrode position after surgery. Fortunately, clinical data show low complication rates with electrode placement, and in poor placement cases, reprogramming the directionality of the electrode usually solves the problem.7 As neurostimulation indications expand, particularly within DBS, there will likely be clinical scenarios for which more accurate electrode placement is required.
|Figure 3. Current neurostimulation devices consist of four subsystems.|
Power Management. Neurostimulators typically work by applying continuous low-level currents. The current applied for VNS and DBS ranges from 1.0 to 3.5 mA, and battery life is typically 8–12 years. For SCS, a current of 20–25 mA is needed, necessitating a larger power source and overall device volume to house a bigger battery.8 As the miniaturization trend for implanted devices has grown, companies developing SCS devices have opted to use rechargeable power systems. As a result, SCS systems have an internal battery that is recharged by an external power source, usually by inductive power transmission. This architecture achieves miniaturization of the implant, at the trade-off of requiring patients to recharge their life-changing batteries on a frequent basis. Although battery charging has been a necessary arrangement, there are two opportunities with neurostimulation power sources to potentially eliminate rechargeable batteries and thus improve the quality of life for patients.
First, battery component technology must continue to improve by incorporating high-energy density cells into device designs. This process can be complicated due to the rigorous level of required quality control, reliability testing, and design validation. In the future, integrating energy scavenging systems that harvest wasted energy from the human body might provide supplemental power to neurostimulators. For example, Zarlink Semiconductor partnered with Perpetuum Ltd. and other companies to create the Self-Energizing Implantable Medical Microsystem (SIMM) project. In 2008, the SIMM consortium clinically tested an energy-scavenging system that partially powered an implanted cardiac pacemaker. The device harvests energy from cardiac output via a catheter that uses pressure gradients within the chambers of the heart to drive a generator. Although this technology could benefit neurostimulator power management, technical challenges remain. The cost and benefit of having an additional implant system leveraging cardiac output or some other energy source should be assessed.
Second, power management must be a focused practice within neurostimulation development. Devices can minimize power consumption by employing advanced sleep and wake algorithms in which the device operates nascent with minimal power and wakes up when therapy must be applied. This requires the electrical architecture to support such an algorithm, as well as accurate sensing capabilities. Improved electrode design and placement will also improve power consumption, because it will require less energy to achieve the same clinical effect as an inaccurately placed electrode.
Telemetry and Data Networks. For implanted neurostimulation devices, telemetry enables physicians to glean basic information, such as identification and battery status, as well as more-detailed information, such as neural activity history. The physician can also program the pulse generator and improve its performance on a periodic basis. Telemetry is a significant area for disruptive innovation because it enables patients to interact with their own neurostimulator to control their own therapy.
In the coming years, telemetry could have a crucial role in improving the overall efficacy for neurostimulation patients. Neural activity is not well characterized in a comprehensive and centralized manner, and future efficacy will be based on the quality of existing databases. Recording neural activity during therapy could provide valuable information for improved device development. Telemetry could benefit scientific understanding and future product improvement by enabling data transmission from multiple-unit recordings across multiple electrodes, acting like an implanted electric map, to central locations.
Patient-Centric Design and Control. Individualized therapy is the holy grail in neurostimulation. Many patients with the same disease present very different symptoms, both in severity and progression, which complicates product design. Although platform technology enables efficient design and manufacturing, it might not produce the best clinical value for highly individualized patients. It is also not financially viable for some companies, especially start-ups, to focus on individualizing therapy.
Crucial decisions concerning system architecture must be made early in the design process. Ideally, platform technology should focus on meeting requirements that are broadly compatible with various applications, while allowing customization via surgical procedure, physician programming, or patient interface, to help optimize the product performance for improving clinical value.
Implantable device operation has usually been isolated from the patients in order to lift their burden of worrying about therapy and pursue normal lives. As patients become more aware of their own therapy, they’re advocating for more information and control. Patient control over a complex medical device is a challenging situation, because medical and technology expertise is required to program and operate the device. However, opportunities for product design that incorporate the patient into the device therapy experience remain.
Neurostimulation devices, specifically SCS, are beginning to incorporate patient interface devices. Boston Scientific’s Precision Plus system pioneered this concept, allowing patients to augment or change the location of pain management therapy via a joystick controller. St. Jude Medical also has a remote control that lets the patient adjust the strength and location of preprogrammed therapy.
These are the first steps to addressing patient needs. Future product innovation can improve clinical benefit by characterizing patient needs via comprehensive user studies. User studies can be the critical design step in ensuring clinical outcome and quality of life are improved, which ultimately will translate into strong market proliferation.
A Quick Market Launch
The opportunity for disruptive innovation exists not only in advancing device technology, but also within the process by which companies accelerate their devices to market. As the device design matures, the critical steps of design transfer, manufacturing process development, and manufacturing ramp up that allow a company to conduct clinical trials and gain market penetration. For large companies with established infrastructure, these phases are generally well managed. However, for start-up companies, the effort, cost, and time to undergo these phases of work are often underestimated, which can result in substantial schedule delays that affect an intended exit strategy and cash flow. Such delays are particularly exacerbated in the highly competitive and underpenetrated neurostimulation market. Thus, the competitive success of neurostimulation start-ups will largely depend on their ability to effectively shift from a research company to a design and manufacturing company (see Figure 4).
Planning for Exit. Early-stage neurostimulation companies, like most medical device start-ups, are founded with the exit strategy of being acquired, licensing out intellectual property, or creating an initial public offering. Tight funding, coupled with increased financial scrutiny, means start-up companies developing Class III devices will rarely be acquired unless good clinical trial data can be demonstrated. The company should demonstrate market adoption via a steady and growing revenue stream—potential parent companies want to know their investment risk is sufficiently reduced before acquisition.
|Figure 4. It can be a challenge for start-up neurostimulation companies to make the move from research to design and manufacturing.|
As a result, the product development process of neurostimulation start-ups must be conducted with the appropriate exit strategy in mind. These start-ups must be prepared to fund themselves longer to demonstrate greater device performance, quality, and reduced risk. This must be achieved while creating a robust quality system, and building appropriate facilities to house equipment, lab space, materials, and resources. All of these requirements cost money, and in light of tight funding, the entire product development process must be more efficient, from concept generation through manufacturing ramp.
Design Transfer and Manufacturing Process Development. Design transfer and process development and validation are routinely phases of product development that are not managed efficiently. Design transfer doesn’t simply refer to the singular time point of handing all design files to a manufacturing team. It is a phased process that involves various levels of analysis for design for manufacturing and assembly. Before design transfer occurs, cleanrooms must be built out, manufacturing processes mapped, capital equipment purchased and installed, and process steps defined. Suppliers and manufacturing partners must be selected and audited, and a consistent component supply chain must be developed. As the design is transferred, manufacturing processes must undergo rigorous performance mapping via operational qualifications to significantly reduce the project risk prior to performance qualifications. Design verification testing should be carefully planned, with time allotted for practice runs to comprehensively evaluate the device system and test methods.
Improving Clinical Value
The rising cost of health care will place greater emphasis on developing devices that offer improved clinical value. Although CE mark and FDA approval are considered milestones for product development, approval from CMS is often the polarizing milestone that can either usher in market success or failure. This situation is magnified for expensive neurostimulation implants. For example, the average cost of a DBS device ranges from $25,000 to $30,000, although reimbursement is typically half the cost. To expand reimbursement coverage, neurostimulation companies must actively engage CMS. This requires effort on multiple fronts. First, more standardized practices for defining clinical efficacy should be developed for various neurostimulation modalities to consistently benchmark devices and therapy performance. Second, neurostimulation companies should incorporate the design of clinical trials early during product planning. Trials should not only demonstrate clinical performance, but also establish the device’s overall benefit and cost compared with the gold standard for the targeted indication. Neurostimulation companies that embrace this concept in their trial design can improve the reimbursement coverage for their products.
Established companies have grown by innovating early, extending product lines to increase market share, and establishing sales force and distribution. When liquidity was greater, the acquisition of small companies integrated novel products into established sales distribution. For example, St. Jude Medical’s acquisition of Advanced Neuromodulation Systems in 2005 catalyzed its neuromodulation business. Now, established companies have the opportunity to evolve their business model. Large companies have the resources and finances to design products and clinical trials that focus on the cost and benefit of a device. Their large research teams can become engines for innovation. At the same time, these companies should avoid early commoditization. A focus on design improvements that increases clinical value in eyes of CMS could be far more fruitful than extending product lines.
For start-up companies, the typical model has been to generate venture capital buzz, take a fast track to clinical trials, get acquired, and then move on to the next venture. However, for the neurostimulation market, there is ample opportunity for disruptive innovation. Start-ups that create platform technology tailored for individualized therapy will increase their staying power and valuation. This will require smart product planning and efficient product development and manufacturing.
In the drive toward achieving better clinical value, neurostimulation companies must refocus their market research, product strategy and pipeline, device development, and technology implementation. Key technology innovations with sensing and detection of neurological pathologies and algorithm development will help improve device efficacy. Redefining user needs through human factors engineering and leveraging telemetry technology will enable patients to gain more control over their disease.
Innovation reaches beyond device design. It presents opportunities for more efficient design transfer and manufacturing ramp up, which enables neurostimulation companies to rapidly hit the market. This is particularly important for early-stage companies, as they compete to enter a largely underpenetrated market. Ultimately, neurostimulation companies that push to improve clinical value will emerge as market leaders and will benefit patients and the healthcare system as a whole.
1.CM Christensen, The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail (Boston: Harvard Business Press, 1997).
2. S Wiebe et al., “A Randomized, Controlled Trial of Surgery for Temporal-Lobe Epilepsy,” New England Journal of Medicine 345, no. 5 (August 2, 2001): 311–318.
3. AA Cohen-Gadol et al., “Neurostimulation Therapy for Epilepsy: Current Modalities and Future Directions,” Mayo Clinic Proceedings 78, no. 2 (February 2003): 238–248.
4. R Melzack and PD Wall, “Pain Mechanisms: A New Theory,” Science 150, no. 3699 (November 19, 1965):971–979.5. M Gold et al., “Comparison of Defibrillation Efficacy and Survival Associated With Right Versus Left Pectoral Placement for Implantable Defibrillators,” American Journal of Cardiology 100, no. 2 (2007): 243–246.
6. JL Vitek, “Deep Brain Stimulation: How Does It Work?,” Cleveland Clinic Journal of Medicine 75, (March 2008): S59–65.
7. A Foletti, A Durrer, and E Buchser, “Neurostimulation Technology for the Treatment of Chronic Pain: A Focus on Spinal Cord Stimulation,” Expert Review of Medical Devices 4, no. 2 (March 2007): 201–214.
8. D Panescu, “Emerging Technologies. Implantable Neurostimulation Devices,” IEEE Engineering in Medicine and Biology Magazine 27, no. 5 (September-October 2008): 100–113.
Rahul Sathe is senior mechanical engineer at Cambridge Consultants (Cambridge, MA).