Introduction As interventional technologies have evolved over the years, the use of intravascular stents to revascularize stenotic carotid arteries and reduce the incidence of stroke has been the topic of heated discussion within the clinical community. While the U.S. Food and Drug Administration (FDA) approved the first carotid stent system in the United States in 2004,1 this decision served only to inflame this debate, rather than quench it. There are several facets to this difference in opinion, and while many of them are not unique to carotid stenting, it is indeed rare to find so many of them present in a single technology. Given the high profile of this technology and the significance of its potential benefits (and shortcomings), it may be helpful to identify the key parties that have helped shape the current environment for carotid stenting and delineate the roles they play and the challenges they face. Medical Device Regulation and Marketing The regulation of the marketing of carotid stents and associated embolic protection systems falls under the purview of FDA’s Center for Devices and Radiological Health (CDRH). The CDRH mission is to promote and protect the health of the public by ensuring the safety and effectiveness of medical devices and the safety of radiological products. Acceptance of a Premarket Approval (PMA) application for a carotid stent is a data-driven decision based on a demonstration that the device is safe and effective for the indicated patient population.2 For embolic protection systems, the device can be cleared for marketing via a Premarket Notification [510(k)] application, if the available data show that the device is substantially equivalent in design, safety, and performance to another embolic protection system currently marketed in the United States.3 Current FDA policy requires that for a carotid embolic protection device to be cleared, clinical data must have been gathered using an FDA-approved carotid stent. As of April 2006, there are two FDA-approved carotid stent systems (both only indicated for use in patients at high risk for morbidity or mortality from the surgical alternative, carotid endarterectomy1) and three cleared, filter-based embolic protection systems indicated for use in the carotid arteries. For both types of devices, CDRH bases its decisions on both preclinical data gathered using appropriate benchtop and in vivo models, and on data from clinical studies typically conducted under an Investigational Device Exemption4 (IDE; Figure 1). For both approved carotid stent systems, the FDA determined that the results of preclinical and clinical evaluation provided valid scientific evidence and reasonable assurance that the devices are safe and effective when used in accordance with their approved labeling to treat both symptomatic and asymptomatic patients. As part of their decision-making process, the FDA considers what types of data are appropriate for the indicated patient population, keeping in mind the risks to the patient and the potential benefit offered by the treatment. For carotid stenting, one example of this approach is that if a sponsor wanted to conduct a study to determine the safety and effectiveness of carotid stenting in low surgical risk patients, CDRH would expect to see clinical data gathered from a randomized, controlled clinical study comparing stenting and endarterectomy outcomes. However, for patients at high risk for carotid endarterectomy, such a randomization might be inappropriate, and thus CDRH would accept a different study design. If new questions related to device safety and effectiveness are raised during PMA review, CDRH may also receive input from the appropriate device advisory panel, which consists of external physicians, statisticians, and other experts in the general device area. While the recommendations of the advisory panel are taken into consideration during the review process, these recommendations are not binding to the FDA. CDRH regulation of medical devices does not end after the marketing of the device commences. As part of its commitment to the total product life cycle, CDRH conducts surveillance of the post-market performance of the device through diverse mechanisms such as adverse event reporting via the Medical Device Reporting (MDR) system, as well as post-approval studies conducted as conditions of PMA approval. Post-approval studies can be required for a number of reasons, including detection of rare adverse events that the clinical study was not powered to detect, evaluation of long-term device safety and performance, assessment of the adequacy of the device training program, and as a mechanism for gathering “real-world” clinical data from patient and investigator populations that may not have been included in the clinical study cohort, but may be more representative of the actual recipients and operators of the device. CDRH has considered such studies appropriate for carotid stent systems, especially given the relatively recent introduction of this technology to the marketplace. Coverage and Reimbursement Like the FDA, the Centers for Medicare and Medicaid Services (CMS) is an agency with the Department of Health and Human Services. While the two agencies do communicate, they are separate entities that make independent decisions. The CMS mission is to assure healthcare security for older Americans, those with disabilities and the disadvantaged by paying for certain types of medical assistance. CMS national or local coverage decisions are predicated on a determination that the prospective item or service is reasonable and necessary for CMS beneficiaries, in much the same way that FDA bases its decision-making process for a prospective device on a determination of safety and effectiveness for the indicated patient population. Coverage can be obtained in three ways: by statute, by the local coverage process, or by the national coverage process. The coverage determination for carotid stenting was made at the national level, and the national coverage process is described in the following paragraphs. There are five general steps to receive Medicare reimbursement. The first step is FDA approval of the device. The second step is the determination of the reimbursement benefit category for the device. Benefit categories are established by Congress, and CMS assigns the most appropriate benefit category or categories to the device. If no category is appropriate, then no mechanism exists for Medicare to reimburse for that device. The third step is the national coverage determination (NCD). In the final two steps, a payment code is assigned and the amount of payment is established for the device. A request for coverage can come from a party external to CMS or can be prompted from within CMS. An NCD can be requested by any person or organization and for any reason. Some of the more common reasons for an external request include a current national non-coverage policy or substantial variation in local coverage policy. CMS may initiate an NCD for a number of reasons, such as publication in the medical literature of an important new study for a device, if a technological advance with a potential major or economic impact occurs, or if there are serious concerns about inappropriate use. CMS encourages frequent communication with the NCD requestor during the NCD process. The time from acceptance of the national coverage request to the announcement of the coverage decision takes typically nine months, as mandated in the Medicare Modernization Act of 2003 (Figure 2). During this nine-month period, CMS conducts a technology assessment, posts a proposed decision memorandum (DM) to its website, collects public comments on the proposed DM, and then posts a final DM. If CMS decides to have an external technology assessment performed or if a Medicare Coverage Advisory Committee is held, or both, then an additional three months is added to the NCD timeline. The NCD is effective the day the final DM is posted. The end result of the NCD process is one of three coverage decisions: coverage, non-coverage or coverage with restrictions (e.g., for specific populations; for specific providers/facilities; with evidence development). As part of the NCD process, CMS evaluates whether a particular item or service is covered nationally under title XVIII of the Social Security Act, including whether the item or service is reasonable and necessary for the diagnosis or treatment of illness or injury, or will improve the functioning of a malformed body member. This evaluation includes review of appropriate outcomes data, such as whether the product provides improved, equivalent or complementary net health outcomes in the Medicare population as compared to alternative treatments or diagnostics already covered by the program, using the principles of evidence-based medicine. CMS also requires sufficient evidence to conclude that the device is generalizable to the Medicare population, and to the general provider community. To merit coverage, the application that accompanies a request for coverage must be supported by clinical studies and other appropriate evidence that answer the technology assessment questions posed by CMS without possessing significant biases that would call into question the validity of the results reported. However, CMS does not necessarily require submission of definitive information or particular types of studies. For some new medical products, large-scale trials are not feasible due to the small number of patients with the target indication, ethical concerns, or the nature of the product. For these reasons, CMS relies on a variety of trial designs and also considers, when appropriate, case reports, expert opinion and standards of medical practice in making its decisions. Ultimately, the requestor must provide a rationale for how the evidence demonstrates a medical benefit for the Medicare population. For carotid stenting, CMS currently reimburses treatment of patients at high risk for carotid endarterectomy, provided they meet one of the following conditions5: 1. They have symptomatic carotid artery stenosis (? 70%); 2. They are participating in an IDE clinical trial and are symptomatic (? 50% stenosis) or asymptomatic (? 80% stenosis); or 3. They are participating in an FDA-mandated post-approval study and are being treated according to the approved device indications. As a result of these qualifications, some patients could be treated according to the FDA-approved device indications and yet not be eligible for CMS reimbursement. This situation illustrates the fact that CMS and FDA decisions do not have to completely overlap. A similar situation occurred in 2003 when CMS decided not to cover implantable cardioverter defibrillators for MADIT II criteria, after the FDA had already approved a device for these indications.6 CMS coverage has since been expanded to include the MADIT II indications, plus others.7 Such divergent determinations can occur because of the separate yet equally valuable missions of CMS and the FDA, and simply because complex clinical data can produce a multitude of separate conclusions (in much the same way that the clinical community has reached several different conclusions regarding the proper indications for carotid stenting). The independence of the FDA and CMS is further illustrated by the fact that CMS may choose to nationally cover or non-cover both labeled and off-label device indications. Impact on (and of) Medical Practice The introduction of carotid stent systems to the United States market in 2004 provided practicing clinicians with a new FDA-approved tool with which to treat carotid stenosis, complementing the currently available surgical and medical management options. Of course, a large number of physicians had already been treating carotid lesions with stents not approved for use in the carotid vasculature for many years. Since the FDA only regulates medical device marketing and not the practice of medicine, such off-label use is not a violation of FDA regulations. The treating physicians, not the FDA or the device manufacturers, are in the best position to know what treatment is optimal for their individual patients. The FDA does, however, forbid any marketing or promotion of devices as being safe and effective for an unapproved use. CMS also does not seek to regulate or otherwise influence medical practice, but their coverage decisions may have this unintended effect. As over 50 million Americans receive their health care benefits through CMS, the availability of the various treatment options for these patients may be affected by CMS reimbursement policies. Although many of the patients who have received unapproved carotid stents have likely benefited from off-label use, this approach is not without far-reaching negative effects. For example, widespread off-label use may have contributed to the perceived delay in U.S. carotid stent approval (compared to the years-old availability of carotid stent systems outside the United States) because with other stent systems available, there was little impetus to conduct the well-organized and controlled clinical studies needed for FDA approval. Likewise, because these patients were mainly treated outside of clinical studies that utilized well-defined protocols for data collection, treatment of these patients did not contribute much to the body of clinical evidence supporting the safety and effectiveness of these devices. This type of conundrum also applies to other types of emerging therapies, such as stenting of the arteries of the lower extremities. Another serious issue associated with off-label use is that patients are treated with devices for which there is no evidence of clinical benefit, and in many cases there are not even preclinical data suggesting acceptable device performance. Because the treating physician is often unaware of the extent (or lack thereof) of available information, off-label use can put the prospective patient at risk for device-related adverse effects without their knowledge or permission. As a result, extensive off-label use can have powerful consequences in both the short and long term. A separate clinical issue is the effect of physician specialization as regards one’s position on and acceptance of this technology. While interventionalists were clearly excited about the newly available carotid stent systems, it is understandable that many non-interventional clinicians did not welcome their arrival. For every interventional cardiologist that expounds the benefits of minimally invasive carotid stenting with embolic protection, there is a vascular surgeon reminding us that carotid endarterectomy can be at least as safe and effective for many patients, if not better. Such differences in opinion are not new to the clinical community, and they contribute greatly to the ongoing discourse regarding the exact place of carotid stenting within the hierarchy of treatment options. Current Challenges With the myriad roles that the governmental agencies play in providing access to and coverage of emerging medical technologies, it is unsurprising that many clinicians feel that the government exerts too much control over medical practice. For carotid stenting, this opinion may be rooted in the belief that the data requirements for FDA approval/clearance of stents and protection devices are unduly burdensome, and that CMS has effectively (albeit unintentionally) mandated the standard of care for their beneficiaries by limiting reimbursement to a segment of the indicated patient population. Conversely, other physicians may feel that the agencies do not provide enough control of medical practice, and that unrestrained rollout of this technology to an unprepared clinical population could result in disaster. While one’s position on this issue may, in part, depend on how one’s own practice will be affected by the availability and acceptance of carotid stenting, the fact that FDA and CMS decisions do have a tangible effect on device access and usage cannot be argued. What may be less apparent is the rationale behind these decisions. FDA and CMS decisions are data-driven, and as such, affirmative agency approval and coverage decisions require the availability of persuasive data. Data quality is at least as important as quantity, and thus a large collection of disparate and unfocused clinical data sets is often less convincing than a single, focused and well-controlled clinical trial. In the absence of completely persuasive data, one potential outcome is that approval/coverage is only granted for indications for which such data are available. This situation represents the status quo for carotid stenting. One obvious pathway to expanded approval/coverage is the collection of more robust clinical data from studies designed to evaluate additional specific indications. In this manner, approval/ coverage may be expanded as additional clinical data are collected to identify patient populations that may benefit from carotid stenting. However, this approach assumes that we know or can determine which populations might derive the most benefit from carotid stenting, which may not currently be the case. For example, the indications for the currently approved carotid stent systems require a patient to be at high risk for adverse events from carotid endarterectomy, and while surgical treatment may not be the best option for these patients, it is not implicit that interventional treatment offers a lower risk of complication or failure. One solution would be to identify the factors that make a patient less likely to experience complications with carotid stenting as compared to carotid endarterectomy, and compare the safety and effectiveness of stenting and surgery in these patients. It is important to appreciate that FDA approval and CMS coverage are not static processes that are immutable once an initial decision is made. Approval of additional indications for carotid stenting is likely once relevant and convincing data become available, and reimbursement of stenting in expanded patient populations may follow suit. The level of control exerted by the FDA and CMS minimize the likelihood of regulatory disaster, especially for nascent technologies. Ongoing Clinical Study Issues Marketing approval for carotid stents indicated for a high surgical risk population currently can be granted based on the availability of convincing registry data instead of clinical data from a randomized, controlled trial. In formulating this policy, the FDA considered input from the members of the Circulatory Systems Advisory Panel as well as its internal experts.8 Many of these high surgical risk registry studies have experienced fairly rapid enrollment. However, the growth of these registries coincides with a decrease in enrollment in randomized studies designed to evaluate the safety and effectiveness of carotid stenting in multiple patient populations, including high surgical risk, low surgical risk, or “all-comer” groups. This correlation may be due to the rising number of potentially competing studies currently ongoing. Another contribution to these trends may be from a patient population with increased access to clinical data and promotional materials. There may be a growing perception among these patients that carotid stenting is superior to surgical treatment of carotid stenosis, and these patients may be less willing to participate in any study in which surgery is a potential treatment option. Because randomized studies can result in clinical data sets and statistical analyses not available in single-arm studies, it is important that these randomized studies continue unhindered. Registry studies may be well suited for areas other than acquisition of data, specifically, to support a marketing application. For example, collection of post-market registry data will enhance our current understanding of the durability of carotid stenting and could influence future FDA approval and CMS coverage decisions. In addition, registry data may be used to address scientific hypotheses not typically addressed in premarket studies, such as identification of patient populations that would derive the most benefit from carotid stenting. Conclusion: Where are We Headed? While no single group is likely to be completely satisfied with the current regulatory environment for carotid stenting, this universal uneasiness may be a sign that the main interested parties, from the regulatory authorities to the end user, have handled the associated challenges passably, if not optimally. With so much at stake, sudden shifts in the regulatory and clinical climate may upset the delicate balance between bureaucratic oversight and individual practice. Having weathered the storm of the initial agency decisions on carotid stenting and lived to tell the tale, we can move forward with increased confidence and a better sense of everyone’s expectations and responsibilities.
SEARCH ALL OF VDM