FY 2007 Regulatory Support Activities
Research conducted in the Office of Science and Engineering Laboratories supports the regulatory activities of the Agency, primarily as follows:
- Strategically manage research with the aim of providing a scientifically sound basis for responding to current needs and anticipating future regulatory challenges, and
- Provide technical consults in support of the Center’s pre-market, post-market, and compliance activities.
Both activities are coordinated within OSEL in an effective manner so as to best meet the Center’s regulatory science needs. Laboratory research is the cornerstone upon which the Office provides the regulatory support function. (The research is described in subsequent sections.) It is largely based on investigations related to the mechanistic understanding of device performance or test procedures to enable the Center and device manufacturers to gain an improved understanding of issues related to safety and efficacy. In general, although the research is directed toward issues identified at the pre-market approval level, the reality is that the research has the largest impact on the post-market end of the Center’s business because most often the research is anticipatory in terms of potential issues of medical devices identified at the pre-market level.
The regulatory support function of the Office is provided through consults supporting both pre-market decisions and post-market actions using expertise developed in the laboratory. A consult is a request for expert advice or information of a specific nature, where it is perceived that expertise is more discipline-related than medical device-related. Consults provide information which contribute to sound regulatory decisions. They may be based on acknowledged scientific/engineering principles or on independent data generated in OSEL laboratories.
The following provides a consolidated picture of the breadth of OSEL consults in 2007:
Number of consults to pre-market issues: 1494
Number of consults to post-market issues: 257
Number of activities related to standards 405
The information provided by a consult is used in some of the following ways:
- evaluate a pre-market submission (IDE, HDE, PMA, 510(k));
- support a compliance action (regulatory case support/development, Health Hazard Evaluation, Health Risk Assessments, etc.);
- assist a scientific collaboration;
- answer a consumer inquiry;
- provide opinions on guidance documents;
- provide revisions to one pagers for the new device approval page; and
- assist health hazard evaluation/health risk assessments or in device determinations/classifications.
For many post-market as well as pre-market regulatory issues, the reviews and investigations conducted by OSEL independently assess the claims made by manufacturers or other parties concerning safety or effectiveness. In other cases, OSEL reviews may assess the adequacy of a design, a failure investigation, a production process, or a quality process employed by the manufacturer. These reviews and analyses rely on in-house expertise and are often augmented by expertise solicited from colleagues in academia, other government laboratories, or even other industry sectors. OSEL laboratory investigations may be undertaken in instances where the veracity of a performance claim needs to be independently verified by testing, or when the claimant lacks the resources to conduct the investigation. Specifically, OSEL provides analytical support to post-market regulatory activities in a variety of ways:
- Provide scientific and engineering reviews and analyses;
- Conduct laboratory investigations of product performance;
- Participate in inspections of medical device establishments;
- Conduct forensic reviews and investigations;
- Identify device safety and performance issues;
- Provide training to FDA and industry; and
- Contribute to Center-wide teams on issues identification as well as science-based analysis of post-market device performance.
Developing standards and measurements are also significant products of this office. OSEL continues to provide innovative solutions to public health problems by constructing generic techniques that lead to the creation of national and international standards that will enhance product safety and effectiveness. OSEL staff actively participate in developing standards at the national and international levels by performing research to establish standard procedures and by managing, developing, and supporting standards used for regulatory assessments.
Office of Science and Engineering Laboratories 2007 Highlights
Active Materials (Division of Chemistry and Materials Science)
Approximately 18 months ago, the Division of Chemistry and Materials Science began a study of the behavior of silver nanoparticles. This was driven both by the appearance of these materials in recently approved medical devices as well as the growing awareness of the need to better understand nanotechnology in general as it applies to medical devices. Scientists in the Division of Chemistry and Materials Science devised a two-pronged approach, using both experimental and computational approaches to access the ability of silver nanoparticles to act as a source of silver ions, a biocidal agent. To accomplish the experimental work, infrastructure was developed to prepare silver nanoparticles in-house so as to have full control of their surface chemistry. An analytical model was developed to account for the electrochemical processes responsible for the dissolution of the silver nanoparticles.
One of the early outcomes of this work has been the realization that the release rate of silver from the nanoparticles is strongly affected by the substrates they are deposited on. Initial theoretical work indicates that nanoparticles above a diameter of 20nm behave in a predictable manner, but below this size silver ions release at a much accelerated rate. Both findings have impact of efficacy and safety. In addition to impacting new devices employing silver nanotechnology, the National Toxicology Program will use initial data and study protocols developed by DCMS and the Division of Biology to define their research directions.
Biological Risk Assessment (Division of Biology)
The Department of Defense (DOD) awarded a grant to the Air Force Research Laboratories (AFRL) in partnership with the Food and Drug Administration (FDA) Centers for Devices and Radiological Health (CDRH) and Biologics Evaluation and Research (CBER); Center for Disease Control (CDC)- National Institute of Occupational Safety and Health (NIOSH); University of Florida (UF); North Carolina State University (NC State U) and University of Nebraska (NU). The specific aim of this project is to develop strategies for reusing single-use, disposable respirators in the event of a shortage.
The technical effort expended on this project provides a comprehensive approach to address the goal of developing and validating methods for decontaminating respirators. It is not certain that all of the decontamination technologies will be effective; but the strategy of trying many technologies attenuates the overall risk of failure, and the effort will validate some effective protocols. If many protocols are validated, first responders may select those protocols that best fit their needs. As a necessary condition to validating biological decontamination protocols, this effort will first develop and validate a standard biological aerosol test method (BATM). The BATM will both enable evaluation of the biological efficacy of decontamination technologies for this effort and provide a standard method to evaluate other antimicrobial/bio-decontamination methods. The BATM here focuses on viruses, but can readily be adapted to test with spores and vegetative bacteria. This work will result in the optimization and validation of some effective protocols and technical information that will be provided to the appropriate standards subcommittee for issuance as a new standard.
Electromagnetic and Wireless Technologies (Division of Physics)
Electromagnetic interference emitted by iPods
Recently, malfunctioning of a cardiac pacemaker electromagnetic, caused by electromagnetic interference (EMI) by fields emitted by personal portable music players (iPods) was highly publicized around the world. A clinical study of one patient was performed by a group not associated with the FDA. Two types of interference were observed when clinicians placed a pacemaker programming head and an iPod adjacent to the patient’s implanted pacemaker. The authors concluded that “Warning labels may be needed to avoid close contact between pacemakers and iPods.” EMI experts in OSEL‘s Electromagnetics And Wireless Laboratory were highly skeptical of this report. OSEL performed an in-vitro study to evaluate these claims of EMI and presented the findings of “no-effects” in a paper submitted and accepted for publication in a peer-reviewed journal.
OSEL performed in-vitro evaluations of the low frequency magnetic field emissions from various models of the Apple, Inc. iPod music player. OSEL measured magnetic field emissions with a three-coil sensor (diameter of 3.5 cm) placed within 1 cm of the surface of the player. Highly localized fields were observed (only existing in a one-square cm area). OSEL also measured the voltages induced inside an “instrumented-can” pacemaker with two standard unipolar leads. Each iPod was placed in the air, 2.7 cm above the pacemaker case. The pacemaker case and leads were placed in a saline-filled torso simulator per pacemaker electromagnetic compatibility standard ANSI/AAMI PC69:2000. Voltages inside the can were measured. Emissions were strongest (≈ 0.2 μT pp) near a few localized points on the cases of the two iPods with hard drives. Emissions consisted of 100 kHz sinusoidal signal with lower frequency (20 msec wide) pulsed amplitude modulation. Voltages induced in the iPods were below the noise level of laboratory instruments (0.5 mV pp in the 0 – 1 kHz band or 2 mV pp in the 0 – 5 MHz bandwidth.
Laboratory measurements of the magnitude and the spatial distribution of low frequency magnetic flux density emissions by four different models of iPod portable music players. Levels of less than 0.2 μT exist very close (1 cm) from the case. The measured voltages induced inside an “instrumented-can” pacemaker were below the noise level of our instruments. Based on the observations of our in-vitro study we concluded that no interference effects can occur in pacemakers exposed to the iPod devices OSEL tested. A recent clinical study of a group of pacemaker patients was performed and confirmed that no effects on the proper performance of pacemakers could be induced by iPod music players.
Preclinical and clinical endpoints for cochlear implant safety
Safety testing of electrical stimulation from cochlear implants has been streamlined in order to get new stimulation paradigms implemented quickly. The streamlined safety criteria have both preclinical and clinical components. The preclinical electrical stimulation parameters are for electrode charge density (<100 micro-coulomb/square cm) and net DC (<1 micro-amp). When new stimulation paradigms are proposed in IDEs, the clinical testing endpoints are for stability in electrical threshold for auditory perception, maximum comfortable listening level, dynamic range, and electrode impedance. These streamlined clinical endpoints were validated with long-term, post-approval studies. One recent example of the application of the streamlined clinical testing was for the implementation of high frequency stimulation, which better matches the sound envelope. The stimulation parameters met the basic charge density and DC criteria, but little was known about the safety of high frequency. The streamlined clinical testing of these devices was consistent with safety. After approval on this basis, there have been no reports of stimulation-related safety problems in properly functioning cochlear implants. This success came about through collaborative efforts between OSEL’s Division of Physics and ENT Branch of the Office of Device Evaluation.
Electromagnetic and Wireless Technologies (Division of Physics)
Electromagnetic compatibility (EMC) and radio frequency identification (RFID):
The OSEL Division of Physics has a long history of discovering electromagnetic (EM) sources that interfere with medical devices, and RFID is no exception. RFID is an automatic identification method that can read a tag from several meters away and does not require line of sight as barcodes do. To achieve this, an RFID reader transmits radio frequency (RF) energy to a tag. This tag then uses RF energy to respond with its unique identifying information. FDA is promoting RFID technology to track and trace drugs through the supply chain to help mitigate counterfeit drugs.
A study was performed to test the effects RFID readers had on pacemakers and implantable cardiac defibrillators (ICDs). Collaborating with industry through AAMI standards, scientists in the Division of Physics adapted test methods to evaluate the EMC between pacemakers/ICDs and RFID readers. A total of 18 pacemakers and 19 ICDs from 5 of the leading pacemaker and ICD manufacturers were tested for immunity of RF emissions generated by 7 RFID readers. Reactions of the pacemakers and ICDs ranged from non-clinically significant events to the potentially harmful inappropriate tachyarrhythmia detection and delivery of therapy or complete inhibition of cardiac pacing. The research conducted is entirely proactive. There have been no incident reports of pacemakers or ICDs being adversely affected by RFID to this date.
The information collected in the aforementioned study has been documented and analyzed by FDA and industry and will be used to better current EMC standards for pacemakers and ICDs. These standards assist CDRH reviewers in pre-market determinations of devices. In addition, the information and experiences collected on RFID technology will assist CDRH in determining pre-market approval of RFID medical devices such as sponge counting systems and the human implanted VeriChip.
Fluid Dynamics (Division of Solid and Fluid Mechanics)
Evaluating the Use of In-Vitro Bench and Computational Techniques for Improving Cardiovascular Regulatory Decisions
This year OSEL scientists proposed a research project titled “Standardization of Computational Fluid Dynamic (CFD) Techniques used to Evaluate Performance and Blood Damage Safety in Medical Devices” that successfully competed for funding as a Critical Path Initiative project. This co-operative effort between FDA and several academic laboratories will refine the details of using computational techniques to model blood flow through medical devices to enable device manufacturers to optimize their designs to minimize damage to patients’ blood exposed to such devices; this reduces the need for time consuming and expensive animal experiments. This multi-year project has two phases. The first phase compares independently derived results on an idealized model flow path. The second phase applies these techniques to a more realistic ventricular assist device flow path.
The core group of the FDA laboratory and three university laboratories has designed and built an initial physical device model, have implemented first pass computer simulations, and have measured flow patterns through the model. A web site for enrolling perhaps a few dozen additional university participants and for sharing data among the collaborators should go live in early 2008. The primary goal of the project is to develop information necessary for a guidance document on proper validation of computational flow modeling of medical devices, which will include mesh design and refinement, model verification, and parameter sensitivity testing.
Imaging Analysis (Division of Imaging and Applied Mathematics)
Full Field Digital Mammography Guidance Development
Digital mammography came to the Center’s attention in the early 1990s. At that time the clinical impact of the new technology was unknown. Attempts to clear this device through the pre-market notification path, known as a 510(k), failed due to significant variability in reader performance. Since then about five systems have been approved for marketing through the PMA process. During the intervening years many papers have been published on digital mammography, and our knowledge of these devices has improved to the point where PMAs are no longer necessary to assure safety and effectiveness. This view was ratified by a May 2006 Radiological Devices panel meeting, and a project to reclassify digital mammography devices to class 2 was initiated at that time. To accomplish this, a working group drawn from the Center’s Office of Device Evaluation, Office of Communication, Education, and Radiation Programs, and Office of Science and Engineering Laboratories was formed. The group moved swiftly to draft a special control guidance for Digital Mammography devices, a 510(k) guidance for Digital Mammography Accessories, a notice of availability, and a proposed rulemaking document.
OSEL research played a unique role in the path to this reclassification action. Research performed in the Division of Imaging and Applied Mathematics has contributed significantly to the professional consensus in the imaging community regarding the appropriate physical measurements necessary to characterize the technical efficacy of digital x-ray detectors. The expertise accumulated over years of investigation on full-field digital detectors and x-ray imaging physics enabled OSEL scientists to take the lead in developing clear technical requirements and measurement methods to assure the safety and efficacy of new digital mammographic systems.
Imaging Diagnostics (Division of Imaging and Applied Mathematics)
Bone sonomety reclassification
OSEL has played a lead role in efforts to reclassify bone sonometry, which is an ultrasound-based technology for screening for osteoporosis. Reclassification from class 3 to class 2 will greatly decrease the regulatory burden for manufacturers. The reclassification will be accompanied by the publication of a special controls guidance document, which will provide manufacturers with explicit recommended protocols for establishing safety and efficacy of bone sonometers. The development of this guidance has been facilitated by years of laboratory bench work and clinical trials conducted within OSEL, which have enabled OSEL scientists to test basic measurement procedures before recommending them to manufacturers. We expect this reclassification to result in 1) new devices getting to market more quickly and 2) an increase of access to new diagnostic devices for those at risk for osteoporosis.
Software Forensics (Division of Electrical and Software Engineering)
Explaining software forensics
Forensic is an adjective, relating to or dealing with the application of scientific knowledge to legal problems. A forensic laboratory is one devoted to the application of science in the investigation of crimes or other legal matters. The focus is on understanding the root cause of software failures in medical devices with an eye toward public health protection. It is remotely possible that our findings might lead to a criminal prosecution; it is far more likely, however, that our efforts will lead to a voluntary recall or other regulatory remedy. In some cases, our investigations have eliminated the software as the "culprit" in an adverse event.
All CDRH research laboratories, including the software laboratory, have conducted these kinds of forensic investigations for many years. In the past, when a software defect was suspected in an adverse event, and the manufacturer would not or could not confirm the root cause of the failure, CDRH engineers would audit the software design by hand--a laborious and difficult process. But this was not the main focus of the laboratory. Our principal interest was --- and remains --- developing analytical methods and tools for assuring the safety and security of software. In other words, the ultimate objective of the research is to enable software developers to “get it right” the first time. As we have worked with academic researchers toward this goal, we gradually realized that some of the software development tools we were studying could be applied in forensic investigations to speed up the analysis of suspect software and detect latent design errors. Our first attempts to use these software design tools in a forensic investigation were highly successful. This early success led to a structured effort to develop a software forensic capability.
Limitations in assessing software problems
Computers keep growing exponentially in terms of processing power and memory capacity per unit cost. As a consequence, medical device software is becoming more complex with each passing year. For many medical devices, software now determines much of the product's functionality and performance. We routinely see medical devices containing more than 100,000 lines of code. In addition, there are devices exchanging data over computer networks and being controlled by software running on another computer. The management of complexity is a huge challenge for medical device designers and regulators alike.
With many early software-controlled medical devices, it was reasonable to rely on a clinician to intervene before a software error caused harm. As computers have become more powerful, many health care protocols have been automated to the point that competent medical intervention is no longer viable as a control measure. The combination of increased complexity and decreased user oversight can be deadly.
Collaborations and partnerships
We have on-going research collaborations with several software experts and agencies. As it happens, our software laboratory is at the forefront when it comes to investigating the performance of embedded software. The FBI, for example, devotes a lot of effort to detecting accounting fraud, pornography, cybersecurity threats, and online terrorism activities; but embedded software is simply not high on their list of “bad guys.” NASA and the Department of Defense (DoD), by contrast, have a great deal of interest in embedded software and, like FDA, are focusing primarily on defect prevention. We are participating in an interagency committee, led by the National Science Foundation, devoted to coordinating Federal research efforts in this area.
Continued growth of the software forensics laboratory and relevance to post-market issues
Our forensic capability is still in the early stages of development. We are evaluating a number of commercially available tools and some that are still the subject of academic research. Each tool has different strengths and weaknesses with respect to finding various kinds of design defects. We expect to publish some of our findings when we have gained a little more experience.
Given the current state of the art, the tools are still very expensive and require highly skilled analysts to use them successfully. Most of them produce a fairly lengthy list of questionable situations, which then need to be reviewed manually. In one recent case, analysis of 100,000 lines of code identified about 180 questionable constructs. Only two of those turned out to be real design issues. Nevertheless, those 180 areas of interest represented only a tiny fraction of the total body of code, so you can see how effective the tools were in zeroing in on the problems. Additionally, many of the 180 red flags were examples of poor coding practice, so the manufacturer gained a great deal of insight into ways to increase their defensive posture by utilizing more robust coding practices in the future.
We do not expect to reach the point where we would use these tools routinely in either pre-marketor post-marketsituations. Only cases that present an imminent public health threat warrant the level of effort required to do the analysis. Over the long term, we believe that these tools will progress to the point that they will routinely be used by software developers.
The Software Forensics Laboratory has exposed several software design errors that were linked to adverse events, and at least one latent error that had the potential to cause an injury. Each of these cases resulted in appropriate corrective actions. In two other cases, the software received a clean bill of health.
Some of these tools are being employed in other industry sectors, notably in mission-critical applications like automotive control systems and aeronautics. We are aware of a few medical device companies that have begun to apply the tools and analytical methods. In an ideal world, some of these early adopters might “report out” their experience in the medical device trade press. Unfortunately, we believe that those who have traversed the learning curve may view this as a competitive advantage and thus be reluctant to share their experience.
Solid Mechanics (Division of Solid and Fluid Mechanics)
Fatigue Testing of PMMA Bone Cement
PMMA (polymethyl methacrylate) bone cement continues to be an essential material in joint replacement surgery to fix metal and plastic prosthetic devices to living bone. In recent years, surgeons have begun using PMMA bone cement (also referred to as acrylic bone cement) to treat pathological fracture including osteoporotic vertebral body compression fractures (VBCFs) in two new applications (vertebroplasty, and kyphoplasty). Although fatigue failure has been identified as a clinical failure mode, no standard fatigue test method existed until very recently. The lack of a reliable standard test method makes comparison of fatigue results from different regulatory submissions and in the published literature difficult or impossible and keeps open the possibility that an inappropriate formulation could be used clinically.
Standards groups have been working for many years to develop standard fatigue test methods for acrylic bone cement. ASTM International recently published a standard test method for fatigue testing of acrylic bone cement, F2118, and the Material Test Methods subcommittee conducted round robin testing to establish precision and bias data for this method. OSEL engineers participated in the round robin testing. The first round of testing showed large variations between laboratories and within some laboratories. The participants identified some areas in which the test method could be improved. In particular, we determined that the sample preparation procedure needs to be investigated and specified more precisely in the standard. We have begun a study to systematically investigate the effects of specimen fabrication techniques on the mechanical properties and fatigue life of acrylic bone cement. The data on the effects of specimen fabrication techniques will allow us to define an optimal method of sample preparation that can be included in a revision to F2118.
Ultrasonics (Division of Solid and Fluid Mechanics)
Development and Characterization of a Blood Mimicking Fluid for High Intensity Focused Ultrasound
A blood-mimicking fluid (BMF) has been developed for the acoustic and thermal characterization of high intensity focused ultrasound (HIFU) ablation devices. This fluid, when combined with the tissue-mimicking material previously developed here, can be incorporated into instrumented test phantoms to allow meaningful evaluations of new HIFU devices and indications involving or affected by flowing blood. The BMF is based on a degassed and de-ionized water solution dispersed with low-density polyethylene micro-spheres, nylon particles, gellan gum and glycerol. A broad range of physical parameters, including attenuation coefficient, speed of sound, acoustical impedance, viscosity and thermal conductivity and diffusivity were characterized as a function of temperature (20 ºC to 70 ºC). The nonlinear parameter B/A and backscatter coefficient were also measured at room temperature. Importantly, the attenuation coefficient is linearly proportional to the frequency (2 MHz – 8 MHz) with a slope of about 0.2 dB/cm-MHz in the 20 ºC to 70 ºC range as in the case of human blood. Furthermore, sound speed and blood-like backscattering also indicate the usefulness of the BMF for ultrasound flow imaging and ultrasound-guided HIFU applications. Most of the other temperature-dependent physical parameters are also close to the reported values in human blood. These properties, along with the ability to withstand temperature increases above 70 ºC without significant damage, make it a unique HIFU research tool. This reusable, nontoxic BMF is appropriate for developing standardized exposimetry techniques, validating numerical models, and determining the safety and efficacy of HIFU ablation devices.