European Heart Journal Advance Access originally published online on November 15, 2006
European Heart Journal 2006 27(23):2737-2739; doi:10.1093/eurheartj/ehl378
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Increased mortality after implantation of first generation drug-eluting stents: seeing the smoke, where is the fire?
1 Cardiovascular Center Aalst, OLV Hospital, Aalst, Belgium
2 Duke University Medical Center, CV Devices Unit Director, Duke Clinical Research Institute, Durham, NC, USA
* Corresponding author. Tel: +32 53 72 44 39; fax: +32 53 72 41 85. E-mail address: william.wijns{at}olvz-aalst.be
This editorial refers to ‘Mortality in randomized controlled trials comparing drug-eluting vs. bare metal stents in coronary artery disease: a meta-analysis’
by A.J. Nordmann et al., on page 2784 and ‘Cause of death with bare metal and sirolimus-eluting stents’; by D.R. Holmes et al., on page 2815
During the World Congress of Cardiology 2006 in Barcelona, concerns regarding the long-term safety of drug-eluting stents (DES) have been voiced based on meta-analyses including numbers large enough to relate statistical significance to low incidence events such as late-stent thrombosis and mortality in percutaneous coronary intervention (PCI) patients. Since then, these potential safety issues have commanded widespread attention from the medical community, health care providers, regulatory bodies, the press, and the public.
Interpretation of compiled data sets, formally as meta-analyses or pooled with any statistical methods, is fraught with complexities, in particular in trying to understand relevance to daily clinical practice. Across variations in study inclusion criteria, adjunctive therapies and endpoint definitions, DES studies are notable for such complexity. Nonetheless, it is essential that large and relevant data are published and the two analyses of mortality outcome in patients receiving DES reported back-to-back in this issue of the Journal bear consideration in the larger question of long-term DES safety.
Executive summary of the meta-analysis by Nordmann
Nordmann et al.1 have performed a detailed meta-analysis of 17 trials including 8221 patients of which 1514 were randomized to receive sirolimus-eluting stents (SES), 2822 paclitaxel-eluting stents (PES), and 3885 bare metal stents. Eight trials that included a randomization arm for SES were included. Duration of follow-up was up to 1 year for all trials, 2 years for 12 trials, 3 for nine, and 4 for two trials. The quality of the included trials was judged to be very good. There was a trend towards an increased risk for total mortality in patients treated with DES, compared with bare metal stents. At 2 years, the odds ratios (OR) for non-cardiac mortality were significantly higher in patients receiving DES (1.72, 95% CI 1.01–2.94). The use of SES was associated with increased non-cardiac mortality at 2 (2.74, 95% CI 1.22–6.13) and 3 (2.04, 95% CI 1.00–4.15) years, but not PES. No significant difference in early- or late-stent thrombosis was reported. The authors conclude that DES do not reduce total or cardiac mortality while implantation of SES is associated with an increase in non-cardiac death due to cancer in particular. They are calling for full disclosure of the causes of death during long-term safety evaluation of these devices.
This study comes from a renowned Clinical Epidemiology Institute and the authors have previously published meta-analyses of balloon and stented angioplasty.2,3 The methodology of their analysis is robust and data collection was made exhaustive by the inclusion of multiple data sources. Studies using either non-polymeric paclitaxel delivery or the currently available PES were analyzed collectively, which is debatable. Total and non-cardiac mortality rates for the TAXUS programme were provided directly by the manufacturer, since these data were not systematically included in the relevant publications. Importantly, detailed data on the duration of dual antiplatelet therapy were not available. As another major limitation, the authors mention that none of the trials reported on the differential loss of follow-up beyond 1 year.
Executive summary of the study by Holmes
The companion manuscript4 denies any significant association between the use of SES and mortality of any kind. The analysis pertains to 1748 patients enrolled in RAVEL, SIRIUS, E-SIRIUS, and C-SIRIUS, of which 878 were randomized to receive SES. Reported follow-up was at 3 years for all four trials, with a mean duration of 2.6±0.6 years. Cardiac mortality was not different between SES and control bare metal stents (P=0.55). There was a trend for increased non-cardiac mortality with SES (2.85%) vs. bare metal stents (1.61%, P=0.10, non-significant). As a result, total mortality was not significantly different (36 events with SES vs. 28 events with bare metal stents, P=0.37). No significant difference in early- or late-stent thrombosis was reported. Stent type was not identified as a predictor of death at any time by multiple Cox proportional hazards regression model. The authors suggest that contemporary PCI results in a shift in the cause of death which now becomes predominantly non-cardiac, instead of cardiac as would be expected in western populations. They see no possible relation between any observed cause of non-cardiac death and the use of SES, other than the effect of an uncontrolled confounder.
This study is co-authored by a most respected senior interventional cardiologist, Dr D. Holmes, together with each principal investigator of the major randomized trials that have compared SES with bare metal stents, and representatives of the Cordis J&J company, including Dr D. Donohue who has been directing the clinical evaluation programme for the SES device. From the report it is unclear whether the data were pooled and analysed by the company or by independent experts. The actual duration of dual antiplatelet therapy and the differential loss of follow-up at 3 years are not given, although the overall follow-up rate at 2 years was 96.4%. Adjudication of individual mortality cases was repeated independently from case narratives, but it is not mentioned whether this resulted in changes in cardiac vs. non-cardiac cause of death, as compared with the initial publications. From Table 5, it appears that half of the deaths in the bare metal stents group are of unknown cause, i.e. assumed to be cardiac vs. only 17% in the SES group.
In both studies, mortality figures were calculated using the initial patient cohort as the denominator, with potential to underestimate absolute mortality rates. Interestingly, both analyses are suggestive of a trend for increased rates of non-cardiac mortality with SES, with interpretations of this finding ranging from a direct systemic effect of the small amount of drug delivered to an artefact of chance. The correlation remains essentially unexplained, although with the many potential interactions between systemic effects of locally active polymer and drug and adjunctive therapies and their discontinuation, vigilance to overall mortality is clearly warranted. As mentioned earlier, both analyses concur that stent thrombosis rates are similar between DES and bare metal stents.
The analysis by Holmes4 is restricted to four trials, yielding smaller numbers. When comparing the same four SES trials, it is quite disturbing to note that the number of reported mortality events differ between the two analyses: total mortality was 41 for SES and 27 for bare metal stents in the manuscript by Nordmann.1 Figures are lower for SES (n=36) and higher for bare metal stents (n=28) in the manuscript by Holmes.4 As a result, the difference in total mortality between SES and bare metal stents decreases from 14 to 8, and not surprisingly given the small numbers, statistical significance is not reached. Even with a ‘hard’ endpoint such as mortality, with evidence constructed from both published data and meta-analyses, these discrepancies are both worrisome and illustrative of the subtle influence of redefinition and re-adjudication processes across complex data sets.
Take home message for the clinician
The need to combine modestly sized studies to achieve statistical power for insight into rare safety events brings with it both the ability to detect the signal and the complexity of interpreting through the assumptions intrinsic to the pooling processes themselves. In essence, like smoke, such meta-analyses and combined studies as these from Nordmann and Holmes both alert us to the likelihood of fire and, simultaneously, obscure the view of the source of the fire itself. Mortality, both cardiac and non-cardiac, in patients undergoing PCI with stenting is the most important and appropriate safety outcome to follow. With composite endpoints, such as are used in smaller pivotal DES studies, small effects on mortality can be masked by the profound effect on other outcomes such as restenosis and reduced need for reintervention.
In current practice, the main safety concern with DES relates to the rare syndrome of unexpected late-stent thrombosis,5 which may present variously as acute myocardial infarction, sudden death, or both.6 This problem of late thrombosis was not well defined in initial DES trials, with small patient numbers, restrictive definitions requiring angiographic confirmation of intra-stent thrombosis or occlusion, and follow-up of 9–12 months—all worrisome for the potential to underestimate the problem. Presently, the clinician's primary concern relates to any increased rate of late-stent thrombosis with DES, as compared with bare metal stents. Unfortunately, this major concern was not directly addressed by either of the present studies.1,4
DES can be qualified entirely as breakthrough technologies, addressing the hitherto unmet need of angiographic and clinical in-stent restenosis with a 60–70% treatment effect across all studied patients and lesion subsets compared with bare metal stents. With rapid early adoption and up to 50% ‘off-label’ use, however, the potential of these combination drug-device products to generate unanticipated safety events is all the more likely. While too rare to be unequivocally detected in randomized pivotal studies of below 2000 patients, when several million DES are implanted worldwide, safety events such as late-stent thrombosis have great importance to the practice of interventional cardiology.
With more smoke currently than any well defined fire, further sharing of all currently available information with carefully defined statistical methods, consistent definitions, and longer term follow-up are absolutely mandatory. Going forward, randomized studies of new investigational DES platforms should amplify key data such as compliance with- and duration of- dual anti-platelet therapy as well as long-term follow-up with reporting of all fatalities, both cardiac and non-cardiac. Whenever possible, retrieval of necropsy specimens should be emphasized for their great value. Registries in the post-market ‘real world’ environment, provide information that is different in quality from concomitant randomization studies, but still quite valuable when systematically done as such registries can collect experience with much larger number of patients.
Clinicians in practice should carefully consider patient selection for DES implantation relative to the inclusion/exclusion criteria that have actually been studied and published in peer-review journals, reviewed by regulatory agencies, or both. Attention to optimal stent implantation technique and ensuring that patients are thoroughly familiar with the importance of compliance with dual antiplatelet therapy should also be considered fundamental to good practice.
At this time, based on all reports and data available, there is enough smoke for caution, careful data review and synopsis, but not panic or hysteria. The agenda of both clinician and scientist must be to penetrate the smoke and define the fire, not to let it run wild or to run wildly away from it.
Conflict of interest: No equity holdings or major conflicts. Paid scientific consulting (all under $10 000) and research grants (no personal revenue) from: Boston Scientific, Cordis/J&J, Medtronic, Terumo, Conor Medsystems, Abbott.
Footnotes
The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.
doi:10.1093/eurheartj/ehl385 
References
- Nordmann AJ, Briel M, Bucher HC. (2006) Mortality in randomized controlled trials comparing drug-eluting versus bare metal stents in coronary artery disease: a meta-analysis. Eur Heart J 27:2784–2814 First published on October 4, 2006, doi:10.1093/eurheartj/ehl282.
[Abstract/Free Full Text] - Nordmann AJ, Hengstler P, Leimenstoll BM, Harr T, Young J, Bucher HC. (2004) Clinical outcomes of stents versus balloon angioplasty in non-acute coronary artery disease: a meta-analysis of randomised controlled trials. Eur Heart J 25:69–80.
[Abstract/Free Full Text] - Nordmann AJ, Hengstler P, Harr T, Young J, Bucher HC. (2004) Clinical outcomes of primary stenting versus balloon angioplasty in patients with myocardial infarction: a meta-analysis of randomised controlled trials. Am J Med 116:253–262.[CrossRef][Web of Science][Medline]
- Holmes DR, Moses JW, Schofer J, Morice M-C, Schampaert E, Leon MB. (2006) Cause of death with bare metal and sirolimus-eluting stents. Eur Heart J 27:2815–2822 First published on November 15, 2006, doi:10.1093/eurheartj/ehl385.
[Abstract/Free Full Text] - Mc Fadden EP, Stabile E, Regar E, Cheneau E, Ong AT, Kinnaird T, Suddath WO, Weissman NJ, Torguson R, Kent KM, Pichard AD, Satler LF, Waksman R, Serruys PW. (2004) Late thrombosis in drug-eluting coronary stents after discontinuation of antiplatelet therapy. Lancet 364:1519–1521.[CrossRef][Web of Science][Medline]
- Kuchulakanti PK, Chu WW, Torguson R, Ohlmann P, Rha SW, Clavijo LC, Kim SW, Bui A, Gevorkian N, Xue Z, Smith K, Fournadjieva J, Suddath WO, Satler LF, Pichard AD, Kent KM, Waksman R. (2006) Correlates and long-term outcomes of angiographically proven stent thrombosis with sirolimus- and paclitaxel-eluting stents. Circulation 113:1108–1113.
CiteULike 
Connotea 
Del.icio.us What's this?
Related articles in EHJ:
- Mortality in randomized controlled trials comparing drug-eluting vs. bare metal stents in coronary artery disease: a meta-analysis
- Alain Joel Nordmann, Matthias Briel, and Heiner Claudins Bucher
EHJ 2006 27: 2784-2814.[Abstract] [FREE Full Text] - Cause of death with bare metal and sirolimus-eluting stents
- David R. Holmes, Jr, Jeffrey W. Moses, Joachim Schofer, Marie-Claude Morice, Erick Schampaert, and Martin B. Leon
EHJ 2006 27: 2815-2822.[Abstract] [FREE Full Text]
This article has been cited by other articles:
|
|
 |
|
  M. T. Roe, A. Y. Chen, C. P. Cannon, S. Rao, J. Rumsfeld, D. J. Magid, R. Brindis, L. W. Klein, W. B. Gibler, E. M. Ohman, et al. Temporal Changes in the Use of Drug-Eluting Stents for Patients With Non-ST-Segment-Elevation Myocardial Infarction Undergoing Percutaneous Coronary Intervention From 2006 to 2008: Results From the Can Rapid risk stratification of Unstable angina patients Supress ADverse outcomes with Early implementation of the ACC/AHA guidelines (CRUSADE) and Acute Coronary Treatment and Intervention Outcomes Network-Get With The Guidelines (ACTION-GWTG) Registries Circ Cardiovasc Qual Outcomes, September 1, 2009; 2(5): 414 - 420. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D.-W. Park, S.-C. Yun, S.-W. Lee, Y.-H. Kim, C. W. Lee, M.-K. Hong, S.-S. Cheong, J.-J. Kim, S.-W. Park, and S.-J. Park Stent Thrombosis, Clinical Events, and Influence of Prolonged Clopidogrel Use After Placement of Drug-Eluting Stent: Data From an Observational Cohort Study of Drug-Eluting Versus Bare-Metal Stents J. Am. Coll. Cardiol. Intv., October 1, 2008; 1(5): 494 - 503. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N Melikian and W Wijns Drug-eluting stents: a critique Heart, February 1, 2008; 94(2): 145 - 152. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Schomig, A. Dibra, S. Windecker, J. Mehilli, J. Suarez de Lezo, C. Kaiser, S.-J. Park, J.-J. Goy, J.-H. Lee, E. Di Lorenzo, et al. A Meta-Analysis of 16 Randomized Trials of Sirolimus-Eluting Stents Versus Paclitaxel-Eluting Stents in Patients With Coronary Artery Disease J. Am. Coll. Cardiol., October 2, 2007; 50(14): 1373 - 1380. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kaul, P. K. Shah, and G. A. Diamond As Time Goes By: Current Status and Future Directions in the Controversy Over Stenting J. Am. Coll. Cardiol., July 10, 2007; 50(2): 128 - 137. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. W. Krucoff, A. Boam, and D. G. Schultz Drug-Eluting Stents "Deliver Heartburn": How Do We Spell Relief Going Forward? Circulation, June 12, 2007; 115(23): 2990 - 2994. [Full Text] [PDF]
|
|
|
|
 |
|
 J. Chikwe, T. Athanasiou, P. Collins, and J. R. Pepper Safety and efficacy concerns regarding elective coronary artery surgery in patients with prior coronary stents Eur. Heart J., May 2, 2007; 28(10): 1265 - 1265. [Full Text] [PDF]
|
|
|
|
|
|
G. G.L. Biondi-Zoccai, A. Abbate, M. Valgimigli, and P. Agostoni Keeping a high standard in quantitative analyses, meta-analyses, and systematic reviews Eur. Heart J., February 8, 2007; (2007) ehl502v1. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||