Setting the Context
USE OF SCIENCE-BASED ASSESSMENT IN THE DEVELOPMENT OF STANDARDS
John T. Watson
National Heart, Lung, and Blood Institute
National Institutes of Health
Science-based testing uses science, engineering, and technology to assess a product’s capability of reliably performing its intended clinical function for a given lifetime and predicted quality of life. This type of testing also assesses the rate of occurrence of adverse events under the specified operating conditions and in the context of the patient’s condition. Because scientific knowledge is incomplete and imperfect, the results of science-based testing are imperfect. Assessment based on such testing is nonetheless an essential component of research, regulatory processes, payment decisions, and commercialization decisions. Despite its imperfections, science-based testing provides valuable information when market approval judgments are being made by expert panels.
Science-based assessment should be used as a component of guidelines, not to create standards. Guiding principles should be based on patient safety and benefit, research on science-based assessment methods, clinically relevant testing, and objective measures such as quality of life, adverse events, and patient function. Assessment methods should stimulate, rather than inhibit, innovation. The peer review process can be used effectively here.
An implementation strategy for science-based assessment should open federal lines of communication, minimize duplication of requirements, and use peer review for guidance and approvals. Such a strategy would emphasize safety in premarket conditional approvals, monitor clinical outcomes for decisions on retention of postmarket approval, and include federal agency support for research on science-based testing methods. Federal agencies could jointly support 100 exploratory research grants of $150,000 each over 3 years using the new NIH study sections. In addition, a pilot study could be undertaken where one or two similar products were selected to undergo a parallel review process. A steering committee could monitor how these products went through existing approval processes and determine how these processes could be improved. Finally, a 2-year reliability study could be implemented to determine the postmarket performance of these products.
SETTING THE CONTEXT: CLINICIAN
Renu Virmani
Department of Cardiovascular Pathology
Armed Forces Institute of Pathology
An implanted medical device must treat the targeted disease, and, above all else, must do no harm. Clinicians must be trained to look at the disease first, and then the device. Before any device is tested, it is essential to understand the biology of the disease process as well as how a normal organ would react to insertion of the device.
For example, research has been undertaken to determine the consequences of inserting metal stents into atherosclerotic coronary arteries. Work done on animal models (pigs and rabbits) has focused on: (1) determining the type of injury to the vessel wall that is caused by inserting a metal stent following balloon expansion; (2) determining the type of vascular reaction that a foreign body might induce; and (3) determining how a normal vessel wall would react to the placement of a balloon-expandable metal stent. The results of these tests indicate that thrombosis, inflammation, and injury are important determinants of neointimal growth and restenosis.
In addition, retrieved devices have yielded data on human healing following insertion of a balloon-expandable stainless steel stent. These data indicate that: (1) stent strut inflammation is influenced by medial disruption and is associated with restenosis; (2) healing is much slower in human atherosclerotic arteries than in normal animal coronary arteries; (3) the extent of injury is a strong determinant of restenosis; (4) inflammation and fibrin deposition are strong predictors of restenosis, with inflammation related to the extent of injury and type of atherosclerotic plaque; and (5) stent strut penetration of the necrotic core correlates with greater neointimal formation.
Drug-eluting stents may contain cytostatic or cytotoxic agents; there have been both successful and unsuccessful applications of such devices. Studies indicate that the use of stents coated with chondroitin sulfate and gelatin (CSG) containing varying concentrations of paclitaxel (between 1.5 and 42.0 µg) results in smaller neointimal thickness at 28 days postdeployment, although medial necrosis and persistent fibrin deposition occur at higher doses. But the benefit of smaller neointima at 28 days is lost at 90 days.
A registry of 15 patients was created to determine the effects of the QUADS-QP2 (7-hexanoyltaxol) stent implanted in humans. Angiographic restenosis was present in 13.3 percent of these cases after 6 months and in 61.5 percent after 12 months. The mechanisms of this restenosis were toxicity from the high drug dose and reaction to the plastic sleeve, along
with persistent fibrin and smooth muscle cell infiltration in atherectomy specimens retrieved from a few patients.
For these reasons, greater understanding of drug-eluting stents is needed and can come only from examination of retrieved devices at autopsy or surgery (although animal models are an important means of understanding the consequences of device insertion for a normal vessel). Device retrieval enables researchers to see what harm has been done by insertion of the device into humans and thereby enables design improvements. Stents must be designed to do less damage and, when damage is unavoidable, to do damage over a longer, rather than a shorter, period of time. Impurities in the stent, as well as in the polymer carriers used in drug-eluting stents, can have a tremendous impact on the patient’s reaction to the therapy.
SCIENCE-BASED TESTING: BALANCING RISK AND REWARD
Paul Citron
Medtronic, Inc.
In the context of a new medical technology, there is a perception that science-based testing is an implicitly good thing. The existence of a panoply of scientifically grounded tests undertaken to prove the safety and effectiveness of a medical technology comports with societal expectations and is perceived to be a means of protecting patients’ interests. The existence of such tests provides a sense of comfort due to the perception that risk has been minimized.
A closer examination, however, suggests that science-based testing can have a negative impact on the innovation process in the field of medical technology. This is especially true when such tests are required without adequate consideration of relevance and when it comes to breakthroughs in medical technology. In some cases, therefore, requirements for science-based testing can be contrary to the interests of seriously ill patients who are inadequately treated by available methods and who might benefit from promising new technologies.
Requirements for science-based testing can be overly burdensome and can lead to rules-driven, rather than outcome-driven, processes. Once in place, these requirements may be difficult to eliminate, even if they are not relevant or are no longer relevant. Because of the considerable financial investment required to validate any new technology, requirements for science-based testing may restrict the early obsolescence of existing technologies that have been bypassed by new knowledge and cause significant
delays in the time it takes for new technologies to become available for seriously ill patients. In some instances, requirements for science-based testing can stop the pursuit of life-saving innovations. In a worst-case scenario, science-based testing requirements could become prescriptive and inflexible and could strongly inhibit, if not eliminate, the use of sound clinical judgment.
It is illuminating to examine how several major medical breakthroughs—kidney dialysis, prosthetic heart valves, the transistorized external cardiac pacemaker, and the tined pacing lead—reached clinical practice. In each case, although researchers appreciated the need for science-based testing prior to patient use, clinical judgment overrode the science in decision making. These reasoned judgments served the interests of patients and society by enabling the timely introduction of innovative treatments for diseases that were previously either untreatable or ineffectively treated. In addition, these breakthroughs provided the technological foundation for clinically significant spin-off innovations that are now part of the therapeutic armamentarium.
In summary, although researchers generally appreciate the importance of science-based testing, calculated risk taking based on clinical judgment is often an integral part of innovation. A certain degree of empiricism plays an important role in breakthroughs. In the context of life-threatening, poorly treated diseases, an approach that responsibly minimizes the time from concept to clinic ultimately favors patients and society. For breakthrough innovations addressing unmet clinical needs, a considered balance between a rigorous and relatively inflexible science-based testing approach and judgment-based empiricism may facilitate the more rapid introduction of safe and effective technologies. As technologies become more mature, the importance of science-based testing increases to help ensure favorable performance comparisons, quality, and consistency.
BIOFILMS AND MEDICAL DEVICES
William Costerton
Center for Biofilm Engineering
Montana State University
Biofilms are formed when bacteria attach to surfaces and aggregate in a hydrated polymeric matrix of their own synthesis. Many persistent and chronic bacterial infections—including periodontitis, otitis media, and cystic fibrosis pneumonia—are caused by the formation of these sessile communities and their inherent resistance to antimicrobial agents. New diseases can
become manifest when these bacterial biofilms form on the inert surfaces of biomedical devices; for example, nosocomial infections can occur around sutures and exit sites, catheters, vascular grafts, orthopedic devices, and other implanted materials and devices. Because biofilms are particularly resistant to treatment, removal of the infected device is often necessary.
New analytical tools and cross-disciplinary studies have recently advanced understanding of the basic biology of biofilms. It is now understood that most bacteria cause pathogenesis only when combined with an inert surface, such as a medical device, or in an individual with compromised health. On medical devices, bacteria can attach specifically to different surfaces or coaggregate with multiple other bacteria to form a dense bacterial plaque. Bacterial biofilms consist of microcolonies on a surface, where the bacteria have developed into organized communities with functional heterogeneity.
Biofilms are characterized by a protected mode of growth that allows survival in a hostile environment. The structures that form in biofilms contain channels in which nutrients can circulate. Cells in different regions of a biofilm exhibit different patterns of gene expression. Biofilms grow slowly, in one or more locations, and biofilm infections are often slow to produce overt symptoms. While sessile bacterial cells release antigens and stimulate the production of antibodies, the antibodies are not effective in killing bacteria in biofilms and may cause damage to surrounding tissues. Even in individuals with excellent cellular and humoral immune reactions, biofilm infections are rarely resolved by the host defense mechanisms. Antibiotic therapy typically reverses the symptoms caused by planktonic cells released from the biofilm but fails to kill the biofilm itself.
One mechanism of biofilm resistance to antimicrobial agents is the failure of the agent to penetrate the full depth of the biofilm. The polymeric substances that constitute the biofilm matrix retard diffusion and establish a formidable penetration barrier. Antimicrobial oxidants, such as the products of oxidative burst from phagocytic cells, penetrate poorly into the biofilm matrix before being deactivated. Cells within the biofilm may also become less susceptible to toxic substances through reduced metabolic rates. Finally, cells within the biofilm may develop a specific phenotype that protects them from metabolic attack.
Recent advances in genomics and in the sequencing of microbial genomes have provided clues to the genetic mechanisms associated with biofilm development. Bacteria can undergo programmed events that ensure biofilm formation and colony survival, much like the programmed events that white blood cells undergo when summoned to a site of injury. Detailed studies of Pseudomonas aeruginosa, for example, reveal that different genes are involved in the processes of adhesion to a solid surface, formation of microcolonies on the surface, and finally differentiation of microcolonies into
polysaccharide-encased mature biofilms. Cells within the biofilm communicate with each other through the release of soluble factors in a process of quorum sensing that is akin to the process of signal transduction in eukaryotic cells.
A number of important questions remain, however, regarding biofilm development, particularly on biomedical devices. Are the mechanisms of attachment and colony formation the same regardless of the characteristics of the surface? What pathways are used in quorum sensing and is biofilm formation prohibited if any of them are blocked? Do all bacteria communicate in quorum sensing or is this specific to Pseudomonas aeruginosa?
An understanding of the underlying biology of biofilm formation can provide the information needed to begin development of more effective modalities and treatments for medical devices. Such treatments could include specific signal inhibitors or drug delivery mechanisms, as well as combined therapies to target both sessile bacteria through specific signaling and planktonic bacteria through antibiotic/metabolic attack. In order to avoid failure and removal of implants, we must leverage our arsenal of micro-scopic, physical, chemical, and molecular techniques to answer these and other questions and to develop effective therapies for biofilm formation on medical devices.
TESTING FOR SAFETY AND EFFICACY: AN ETHICIST’S PERSPECTIVE
Leonard J. Weber
University of Detroit Mercy
The perspectives in this presentation are offered in an effort to promote ethical best practices, rather than being focused on what must be done to comply with regulations or to avoid wrongdoing. Health care ethics is about clinical care and scientific/professional behavior, but it is also about business practices and decision making in health-related industries. The two issues discussed are how to determine acceptable risk in developing new technologies for clinical application, and how conflicts of interest are handled in clinical trials.
Regarding risk, one of the major implications of ethical analysis today is that the acceptable level of risk should be determined largely by assessing the impact of the new technology on health care quality and cost. A greater risk is more acceptable in a lower-cost treatment than in a higher-cost treatment and in a more beneficial treatment (both in terms of the effect on
the quality of life and compared with other options) than in a less beneficial treatment.
Regarding conflicts of interest in clinical research, there has recently been a growing recognition that the scientific integrity of clinical testing is sometimes threatened by interests that are antagonistic to professional or ethical responsibility and that are substantial enough that they might reasonably affect judgments or actions. These conflicts are frequently not recognized—in fact, are often denied—and inadequate attention is given to the need to prevent or manage them. Best practices that reduce the risk of conflicts of interest include systematic attention to the design of the clinical testing processes and to the financial arrangements involved, with the goal of protecting objectivity. In addition, independent oversight is required.