Drug-Eluting Stents “Deliver Heartburn”
How Do We Spell Relief Going Forward?
“Breakthrough” technologies may produce rare or unexpected performance issues in postmarket use, especially when rapid market penetration into large patient populations outpaces the development of clinical knowledge. Although high-profile meetings or news media coverage may help draw attention to such issues, ultimately it is careful scientific consideration of safety and effectiveness issues, with transparent public access, that will help physicians and patients make decisions about how and when these technologies are best used in the practice of medicine. To this end, a 2-day special advisory panel on drug-eluting stent (DES) thrombosis was convened in December 2006 by the US Food and Drug Administration (FDA), as has recently been summarized by Laskey et al1 and by Farb and Boam.2
Since the first safety alerts in 2003 to 2004 raised concerns that DES might be associated with an increase in rare but often catastrophic thrombotic events, public, professional, and regulatory attention has been focused on delineating the basis for this concern.3 In the spring of 2006, new reports of increased stent thrombosis rates with DES at the European Society of Cardiology in Barcelona led to a cascade of published reports and opinions4–7 that were characterized in The Wall Street Journal headline as “Coated Stents Deliver Heartburn.”8 The article did not specify whether the reference was to heartburn in already-stented patients, cardiologists, DES manufacturers, the FDA, or Wall Street itself. In fact, the issues surrounding DES touch on all of the above.
The FDA panel’s proceedings served a key purpose by publicly reviewing available information and concluding that DES “on-label” use is both safe and effective relative to bare metal stents (BMS), while pointing clinicians to the American College of Cardiology/American Heart Association/Society for Cardiovascular Angiography and Interventions guidelines for dual-antiplatelet (DAP) therapy out to 1 year in patients with low risk of bleeding.9 The panel also represented a uniquely international effort to focus on this topic. In addition, however, the panel crystallized many key questions that could not be clearly answered about higher event rates with off-label use and optimal duration of antiplatelet therapy.1,2
In the aftermath of the panel, FDA, clinicians, and the industry have continued a dialogue on how best to interpret and apply the panel recommendations. Industry, research organizations, and academics are to be applauded for collaborative efforts to continue to collect and provide unbiased access to new and extended patient-level data, and several leading peer-reviewed journals have expedited publication to facilitate dissemination of these findings. For example, 7 articles on DES outcomes were included in the March 8, 2007, issue of The New England Journal of Medicine, including 5 original peer-reviewed scientific reports of analyses that included larger patient numbers with longer follow-up periods than previously available.10–14 Like the panel proceedings themselves, however, the nobility and utility of these efforts and the collaborations they represent also clarify the limitations imposed by ongoing retrospective analyses. These reports originate from studies designed on the basis of the best insights of the past—in many cases, insights that industry, academics, and regulatory experts shared at that time. It is clear that these studies simply do not contain all the information that is best suited to the differently informed needs of the present and future. For instance, very few data are presented in any of these reports on actual compliance with, or interruption of, DAP therapy or inclusion of many more complex “real-world” patients in randomized cohorts. Thus, although this issue of The New England Journal of Medicine represents a great deal of effort to revisit DES safety as an area of widespread interest, it also highlights some of the potential hazards of retrospective data mining, with 4 of these 5 reports including many of the same clinical trials, mixed together and reported differently. When enthusiasm for generating new analyses risks promoting confusion rather than clarity, it is time to recognize that how much we can learn from looking backward can reach its limits. At such a point, we should more clearly focus on how to address current concerns going forward.
Such a “going-forward” focus is the primary agenda for clinicians, academic researchers, regulatory authorities, and industry in 2007. Objectives for the future can be considered from 3 perspectives: (1) best care of patients who already have permanently implanted approved DES platforms, (2) best application of currently approved DES platforms in patients yet to be treated, and (3) best approaches to safety and effectiveness evaluation of new, investigational DES platforms.
No medical device is entirely safe, and the use of a medical device is routinely based on a clinical balance between its effectiveness and its safety-related risks. What constitutes acceptable equipoise with DES—how much less restenosis or target vessel revascularization might balance how much more stent thrombosis, or whether revascularization with stents is desirable compared with alternatives such as medical therapy or bypass surgery15—is fundamentally a decision for practitioners, professional societies, and patients. Accurately determining the incidence of adverse clinical events, however, especially in the case of rare but catastrophic occurrences such as stent thrombosis, is a challenge and a concern for all constituencies, including regulatory authorities.
For current and immediate-future patient care using already-approved DES platforms, the FDA will begin with a careful reappraisal of DES labeling to ensure that the information provided to physicians and patients is as up to date, accurate, and informative as possible. Oversight of medical product labeling is one of the FDA’s strongest tools in ensuring that the information provided to users is data driven and comprehensive. Labeling also forms the basis for the agency’s evaluation and enforcement of promotion and advertising activities. Labeling can and should be updated to reflect new information that can be reviewed and verified.
Fundamental to the evaluation of new DES platforms is recognizing how difficult it is, from both a clinical and a regulatory perspective, to characterize rare adverse events associated with DES thrombosis. Such events often emerge for the first time in the postmarket environment, after the device has been used on a much larger and more diverse patient population than in the initial clinical studies. Additional complexities range from the consistency and quality of manufacture and the technique of clinical operators to the quality and utility of patient data that are collected. Stent thrombosis also may be affected by the biological characteristics of individual patients, the implantation techniques used, and the structure of the DES, including the polymer, the drug, and the metal stent platform. Late events also are complicated by the fact that, over time, it is more difficult to determine whether adverse outcomes result from stent thromboses per se or are related to the progression of underlying coronary disease or the comorbidities with which it is associated.
Addressing these complexities will be the challenge for all new DES evaluations that seek to determine how new DES platforms compare, in both safety and effectiveness, to existing therapeutic alternatives. DES platforms are complex and quickly iterating. They are classified as class III medical devices requiring (1) successful completion of a premarket approval application before marketing; (2) submission of annual reports, adverse event reports, and data from required postmarket studies after marketing; and (3) approved premarket approval supplements for significant changes to the device or its labeling. The FDA’s “total product lifecycle” approach to regulation of medical devices reflects the FDA’s expectation that, as new data and knowledge are acquired for a single device or type of device, this knowledge will be applied to future research and development and to updated regulatory considerations. Previous data requirements for bench, animal, and clinical studies for DES approval were developed using knowledge extrapolated from our experience with BMS supplemented with pharmacological data from systemic applications of the drug eluted. Future developments in this field will need to incorporate further what we are learning today about these combination products.
In addition, the FDA will reassess the kinds of data needed to ensure the safety and effectiveness of DES, both before they are marketed and after they are in widespread use. This does not mean that studies currently underway will be instantaneously rendered obsolete or that future studies will be held to a different regulatory threshold, but it does mean that relevant new questions must be addressed in a manner consistent with sound scientific principles and appropriate regulatory models. Careful consideration must be given to the need to better define the risks of stent thrombosis and the benefits of reduced restenosis in the populations in whom the devices are used over the time period needed to accurately assess these events. These issues can be addressed by judicious refinement of premarket and postmarket studies with the ultimate goal of designing a comprehensive clinical program that balances patient safety with product availability.
Comprehensive premarket approval evaluations generally require the use of randomized clinical trial (RCT) designs. RCTs of DES will require careful scientific and statistical consideration based on current knowledge. For instance, stent thrombosis comparisons between BMS and DES in early pivotal study designs censored patients after reintervention for restenosis. More recent reevaluations have noted that this analysis plan may bias perception of BMS safety (because of higher restenosis rates, with more censored patients) unfavorably relative to DES. Some experts have suggested that elimination of such censorship actually would produce a more theoretically rigorous “intent-to-treat” analysis plan.12
Analyses of currently available data also should inform the selection, definition, and power calculation for primary safety and effectiveness end points in DES RCTs. The use of the Academic Research Consortium consensus end point definitions related to DES outcomes16 for the FDA panel presentations and most publications since represents an important lesson learned that can be leveraged going forward. The adoption of a single set of consensus definitions reflecting possible, probable, and definite stent thrombosis has been quite useful, even with the realization that limitations of these definitions include the variability of sensitivity/specificity, depending on how they are applied. Justifiable concerns about applying them to previously concluded studies through retrospective readjudication of events are not an issue for prospective planning of new studies going forward. Despite their limitations, the panel made clear the utility of defining clinical end points consistently across a wide range of clinical trials and patient reports. For rare or unexpected postmarket safety events, use of consistent definitions may be important both for detecting a safety signal and for mechanistically evaluating its source—or, as had recently been described, for appreciating the smoke compared with actually seeing what is burning.7
Selection of appropriate primary efficacy end points is another key area for DES evaluations. The use of surrogate measures such as angiographic late loss may be less sufficient in light of current awareness of complex interactions of device features, procedural features, and patient characteristics. More recent analyses of all death and myocardial infarction (including subsets of cardiac/noncardiac death and Q-wave/non–Q-wave myocardial infarction13,14), combined with landmark analyses to help understand whether certain time periods are more vulnerable for particular treatment cohorts, have provided helpful information in retrospective analyses14 and could be considered in future trial designs. Angiographic and intravascular ultrasound measures may be more useful as biomarkers of vessel response to DES implantation than as surrogate markers for DES efficacy.
Another lesson learned from previous DES studies is the influence of study design on reintervention rates, in particular, through use and timing of protocol-mandated angiography that encourages “occulostenotic” interventions. Although for existing studies the independence of such evaluations may not be possible, for future studies, the completion of clinical evaluations before protocol angiography or the separation of angiographic follow-up studies from clinical studies may be useful.
Retrospective comparisons with historical BMS data have been important for informing product labeling and clinical decisions; however, they may become less meaningful as practice patterns evolve. Pivotal DES RCTs have already shifted from superiority designs compared with BMS to active-control, “head-to-head” DES noninferiority studies. It is notable that in the recent studies in The New England Journal of Medicine, not 1 study, even the Swedish national registry, could provide substantive data comparing any 2 DES platforms.10–14 Enabling well-designed clinical trials to keep pace with expanding clinical practice will require discipline and commitment on the part of industry, government agencies, and DES users. Depending on the labeling issues sought in future DES RCTs, either approved BMS or DES platforms might serve as control groups, and superiority, noninferiority, or integrated designs may be appropriate. Confidence intervals used to define such analysis plans will require robust statistical justification based on updated current information.
Over time and as DES products evolve to further stages of the product lifecycle, evaluation of risks and benefits may ultimately be able to leverage larger databases or longitudinal analyses to lower the burden of RCTs or to answer specific DES safety and performance questions without RCTs per se. However, the informative use of registries or larger longitudinal data sets will need the development of formal analytic methods to translate that data into useful information for key clinical decision making and/or reasonable product evaluation.
Another important aspect of DES studies going forward is the recognition that the safety behavior of these devices is tied directly to the obligatory use of an adjunctive pharmacological regimen of DAP therapy with aspirin and thienopyridines such as ticlopidine or clopidogrel. Although this relationship early after stent implantation is well established, the optimal duration or utility of more extended DAP therapy is less clear and may well be different for different DES platforms. Not a single study analyzed in any of the reports in the March issue of The New England Journal of Medicine provided per-patient detailed DAP therapy information because these data generally were not collected. Going forward, patient education and careful documentation of compliance, need for interruption, and patient care strategies during interruption are now recognizable as important areas of interest.
On a more general level, as medical products become more complex and innovative, we can expect an ever-greater number to combine drugs and devices. This means that the FDA must enhance its own intramural collaboration to ensure that these combination products are evaluated smoothly and without duplication of effort. With the passage of the Food and Drug Modernization Act of 1997, the Office of Combination Products was created specifically to provide the administrative oversight for this very circumstance. The interaction of DES safety and DAP therapy steps beyond current intramural structure at the FDA because the safety of the device is driven in part by an obligatory drug regimen that is not actually part of the device per se. For these instances, the FDA will explore how information on the safety and effectiveness of the drug regimen can be incorporated more effectively into the evaluation of the device. Under its Critical Path Initiative, the agency, partnering with a group such as the Duke Clinical Research Institute, can engage in bringing together representatives from various device and drug companies, along with interested parties from the academic community, to develop the clinical trial models needed to answer the scientific questions and to address simultaneously the business and regulatory requirements that will ensue. The FDA also will explore ways in which international collaboration might be expanded to better incorporate data from other countries.
Finally, we now have the opportunity to leverage significantly from the many lessons learned and clarifications resulting from the combined efforts at the December 2006 FDA panel. It was crucial when evaluating the benefits and risks of DES for the panel to distinguish between on-label use of the device (in patients who meet the criteria specified in the product labeling) and off-label use (in patients who do not meet these criteria). This distinction was particularly important in view of the widespread off-label use of DES. Not only are thrombosis rates different in on-label compared with off-label patient populations, but the quantity and quality of data available for these 2 groups differ as well. Thus, in on-label use patients, concurrently randomized cohorts treated with BMS provided robust comparisons from pivotal studies. The use of these studies to characterize rare safety events by compiling larger cohorts through meta-analyses also clarified the criticality of using patient-level data, which supported different conclusions than previous, publicly reported meta-analyses constructed from summary tables. More recent publication of even longer-term follow-up pooling of patient-level data from on-label use studies adds support to the panel conclusion that on-label use is both safe and effective relative to BMS.12,14
Simultaneous with this reassurance, however, is the clarification of how limited the scope of such analyses is for understanding the much larger, real-world use of DES in more complex, off-label patients. That off-label patients have higher event rates overall might represent a population of “higher risk, higher benefit” from DES use compared with other treatment options or might represent a more vulnerable cohort at greater safety risk than on-label patients. The absence of definitive data today highlights the importance of acquiring such data in the future. For any individual DES program, at least 2 broad strategies could be identified. If more complex patients were incorporated into premarket pivotal DES studies, it would both add knowledge of investigational device performance in these patients versus concomitantly randomized controls and serve to expand the definition of on-label patients to better match real-world postmarket practice. Furthermore, such an “enriched” patient study cohort would increase the density of clinical end points, reducing the burden for the size of RCTs.17 The inclusion of more complex patients in premarket pivotal RCTs brings challenges, however, to both statistical analysis plans and other clinical trial operations.18 Another possibility is to develop more systematic approaches to data collection in the postmarket environment. In this setting, both randomized and nonrandomized studies could have potential utility for understanding specific patient populations, regional practice patterns, or other key information that might support extension of indications, enhance the quality and accuracy of safety information related to DES use in real practice, or both.
The range of issues pertinent to new study designs touched on above is summarized in the Table. Thoughtful, program-specific considerations along these lines going forward will help to answer important questions such as whether differences exist between various DES configurations that may provide advantages for specific target populations and lesion types, how different DES models will perform over time, and to what degree DES safety is related to other obligatory medical therapy. The application of these insights may provide more accurate estimates of event rates for new DES designs released for clinical use. It should also be recognized, however, that in the application of these concepts to specific product lines, we are likely to generate new questions and to leave some old questions unresolved. Nonetheless, these directions may well facilitate clarification and, to some degree, redefinition of how to establish reasonable assurance of safety and effectiveness of new DES therapies.
The FDA will continue to work toward a better understanding of the pathophysiology of coronary artery disease, the biological responses to DES and other treatment modalities, and the complex design, manufacturing, and usage issues associated with DES technology. Using tools such as the Critical Path Initiative and the Medical Device Surveillance Network program, the FDA will collaborate with its academic and industry partners to advance the science on which effective regulatory strategies must be based. In the face of limitations of preclinical and animal models, the advance of an “organ donor”–like approach to retrieval of autopsy DES specimens in both noncardiac and cardiac deaths, as is currently being discussed in conjunction with the FDA Critical Path Cardiac Safety initiative and the National Institutes of Health, could provide unique and critical histopathological understanding of DES implantation sites.
Clarifying the risks and benefits of DES and using them in the safest and most effective way possible cannot be accomplished by any single group. Professional societies, physicians, and the industry all have roles to play by conducting effective clinical studies, using devices wisely, reporting adverse events promptly, sharing clinical findings, and developing criteria for appropriate use. In addition, it will require both collaboration and creativity to review and to redefine the landscape of new DES evaluation to encourage the development of new therapeutics while ensuring that, over millions of human exposures, they are both safe and effective. If we all work together to meet this responsibility, it will spell relief from DES heartburn not only looking backward but also going forward.
Source of Funding
This article was supported by funding from the Duke Clinical Research Institute.
Dr Krucoff consults for and receives significant grant support from Cordis Corp/Johnson & Johnson, Conor Medsystems, Boston Scientific, Abbot, Terumo, and Medtronic. The other authors report no conflicts.
Laskey WK, Yancy CW, Maisel WH. Thrombosis in coronary drug-eluting stents: report from the meeting of the Circulatory System Medical Devices Advisory Panel of the Food and Drug Administration Center for Devices and Radiologic Health, December 7–8, 2006. Circulation. 1007; 115: 2352–2357.
Pfisterer M, Brunner-La Rocca HP, Buser PT, Rickenbacher P, Hunziker P, Mueller C, Jeger R, Bader F, Osswald S, Kaiser C, for the BASKET-LATE Investigators. Late clinical events after clopidogrel discontinuation may limit the benefit of drug-eluting stents: an observational study of drug-eluting versus bare-metal stents. J Am Coll Cardiol. 2006; 48: 2584–2591.
Nordmann AJ, Briel M, Bucher HC. Long-term mortality in randomised controlled trials comparing drug-eluting versus bare metal stents in coronary artery disease: a meta-analysis. Eur Heart J. 2006; 27: 2784–2814.
Wijns WC, Krucoff MW. Increased mortality after implantation of first generation drug-eluting stents: seeing the smoke, where is the fire? Eur Heart J. 2006; 27: 2737–2739.
Johnson A Coated stents deliver heartburn. The Wall Street Journal. November 11, 2006.
Smith SC Jr, Feldman TE, Hirshfeld JW Jr, Jacobs AK, Kern MJ, King SB 3rd, Morrison DA, O’Neill WW, Schaff HV, Whitlow PL, Williams DO, Antman EM, Adams CD, Anderson JL, Faxon DP, Fuster V, Halperin JL, Hiratzka LF, Hunt SA, Nishimura R, Ornato JP, Page RL, Riegel B, for the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and American College of Cardiology/American Heart Association/Society for Cardiovascular Angiography and Interventions Writing Committee to Update the 2001 Guidelines for Percutaneous Coronary Intervention. ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/SCAI Writing Committee to Update the 2001 Guidelines for Percutaneous Coronary Intervention). Circulation. 2006; 113: 156–175.
Weintraub WS, Boden WE, O’Rourke RA, Teo KK, Hartigan PM, Maron DJ, Kostuk WJ, Knudtson M, Dada M, Casperson P, Harris CL, Chaitman BR, Shaw L, Gosselin G, Nawaz S, Title LM, Gau G, Blaustein AS, Booth DC, Bates ER, Spertus JA, Berman DS, Mancini GB, Weintraub WS, for the COURAGE Trial Research Group. Optimal medical therapy with or without PCI for stable coronary disease. N Engl J Med. 2007; 356: 1572–1574.
Cutlip DE, Windecker S, Mehran R, Boam A, Cohen DJ, van Es G, Steg PG, Morel M, Mauri L, Vranckx P, McFadden E, Lansky A, Hamon M, Krucoff MW, Serruys PW, on behalf of the Academic Research Consortium. Clinical end points in coronary stent trials: a case for standardized definitions. Circulation. 2007; 115: 2344–2351.
Wang TY, Hasselblad V, Petersen JL, Wijns W, Parhizgar A, Kereiakes DJ, Krucoff MW. The Cobalt Chromium Stent With Antiproliferative for Restenosis II (COSTAR II) trial study design: advancing the active-control evaluation of second generation “breakthrough” technologies. Am Heart J. 2007; 153: 743–748.