Late mortality rates following percutaneous coronary intervention (PCI) among dialysis patients markedly exceed those in patients without end-stage renal disease,1,2 and data preceding the approval and widespread use of drug-eluting stents indicate that PCI compared with surgical revascularization is associated with higher mortality rates in these patients.3 Though drug-eluting stents have been shown to reduce the risk of restenosis in populations consisting predominantly of patients with normal or mildly impaired renal function,4–6 the safety and clinical efficacy of these devices in dialysis patients are unknown. We examined the clinical outcomes of dialysis patients undergoing coronary stenting, either with or without drug-eluting stents, in recruitment waves 3 and 4 of the National Heart Lung and Blood Institute (NHLBI) Dynamic Registry. Methods Patient population and protocol. Details of the design and function of the NHLBI Dynamic Registry have been described in detail previously.7 Briefly, the Dynamic Registry is a prospective cohort that enrolls and follows consecutive patients undergoing PCI at medical centers across the United States and Canada. Enrollment “waves” occur approximately every 18 months and each wave includes approximately 2,000 patients. This analysis included patients enrolled in wave 3 (October 2001 to March 2002; n = 2,047) and wave 4 (February and May of 2004; n = 2,112), and who had complete data regarding current dialysis status. Baseline demographic, clinical and procedural features and the rates of major adverse cardiovascular events (MACE) during hospitalization were recorded. Study coordinators at each site contact patients at 1 year to obtain vital status and information regarding cardiovascular events or procedures that occurred during the year following the index PCI. Additionally, the vital status of dialysis patients who were lost to follow up (n = 10; nine of whom received at least 1 stent) was ascertained using the Social Security Administration’s Death Master File (URL: http://www.ntis.gov/products/ssa-dmf.asp). Each center received institutional approval, and data were analyzed at the University of Pittsburgh (Pittsburgh, Pennsylvania). Patients on chronic dialysis therapy at the time of enrollment who had greater than or equal to 1 stent(s) implanted during the index PCI procedure and who provided informed consent, comprise the primary study population (n = 74). Endpoints, data collection and statistical analysis. The primary endpoint was a composite of MACE at 1 year, defined as all-cause death, myocardial infarction (MI) and repeat target vessel revascularization (the need for repeat PCI or coronary artery bypass surgery). The components of the primary endpoint have been defined previously.7 Patients undergoing dialysis were stratified by stent type (drug-eluting stent vs. bare-metal stent) and continuous data were summarized as means ± 1 SD and compared via the Wilcoxon rank sum test. Categorical variables, presented as percentages, were compared using Chi-square tests or Fisher’s exact tests. One-year event rates were calculated using the Kaplan-Meier approach, and unadjusted comparisons of survival curves were performed using the log-rank test. Adjusted survival curves were plotted and compared using inverse probability weights.8 Cox proportional hazards modeling was used to estimate 1-year hazard ratios for MACE and mortality in relation to stent type. Covariate adjustment was also perfomed such that clinical, demographic and procedural variables were entered individually into models that also included the indicator for stent type. Confounding variables, as well as those only associated with the outcome, were then assessed in a forward stepwise manner to determine the final model. The proportionality assumption was assessed and met for all Cox proportional-hazards models. Hazard ratios with their corresponding 95% confidence intervals are reported. A sensitivity analysis was also performed for outcomes. For patients either lost to follow up or who did not consent to be contacted, 1-year event rates were calculated using Kaplan-Meier methodology, assuming that these patients all had an adverse event and then again assuming that none of them had experienced the event. Event rates were compared using the log-rank test. For all analyses, a two-sided p n = 4,143), 88 patients (2.1%) were on regular dialysis therapy at the time of enrollment. Of these, 74 patients (83.7%) underwent coronary stent implantation: 41 patients received bare-metal stents only and 33 patients were treated with drug-eluting stents, either alone or in combination with a bare-metal stent (2 patients treated with drug-eluting stents also had a single bare-metal stent implanted during the index procedure). Baseline characteristics. The clinical and procedural features of patients stratified by the type of stent used are shown in Table 1. Compared to patients treated with bare-metal stents only, those receiving a drug-eluting stent were more likely to undergo PCI for a recent acute MI, more frequently had hyperlipidemia or a history of stroke. Treated lesions in patients receiving at least 1 drug-eluting stent were more often considered complex angiographically. Other baseline and procedural characteristics of study patients treated with bare-metal versus drug-eluting stents did not differ significantly. Clinical outcomes. As shown in Figure 1, all-cause mortality and MACE rates were significantly higher among dialysis patients compared with patients not on dialysis treatment at the time of enrollment. In-hospital and 1-year clinical outcomes among dialysis patients treated with drug-eluting stents versus without are presented in Table 2. Nonfatal MACE were infrequent during hospitalization, independent of the type of stent used. These, as well as in-hospital mortality, did not differ significantly between both groups. However, with a marked increase in event rates from discharge to 1-year follow up, the composite MACE endpoint occurred significantly less frequently in patients treated with versus without drug-eluting stents. There was also a trend toward lower all-cause mortality among the patients treated with drug-eluting stents. Unadjusted 1-year rates of freedom from all-cause mortality and the composite MACE endpoint among patients treated with drug-eluting compared with bare-metal stents are depicted in Figure 2. The 1-year rates of repeat revascularization procedures and nonfatal MI did not differ significantly in patients treated with versus without drug-eluting stents. Multivariate analysis and covariate-adjusted survival. Covariate-adjusted rates of freedom from all-cause mortality and the composite MACE endpoint at 1 year are shown in Figure 3. For both endpoints, event-free survival was enhanced in patients treated with drug-eluting stents compared with those treated with bare-metal stents. Multivariable correlates of all-cause mortality and the composite MACE endpoint at 1 year are shown in Table 3. After adjustment for age, factors independently associated with MACE included bare-metal stent use, history of cancer at study entry and treatment of an occluded vein graft. While not significant at the 0.05 level, bare-metal stent use was associated with higher 1-year mortality after adjustment for age and a history of cancer at the time of enrollment (p = 0.06). MACE rates based on the sensitivity analysis did not differ substantially from the observed event rates. One-year MACE rates, assuming that all patients without follow-up contact experienced an event, were 61.1% in the bare-metal stent group and 30.3% in the drug-eluting stent group (p = 0.012), and assuming the opposite, the event rates were 48.9% in the bare-metal versus 24.2% among the drug-eluting stent patients (p = 0.037). The true observed event rates were 57.3% (bare-metal stents) versus 25.5% (drug-eluting stents) (p = 0.01). The crude risk ratio for all-cause mortality with drug-eluting stents versus bare-metal stents was 2.01. However, the assumption that 1 patient in the bare-metal stent group who died, in fact lived, resulted in a crude risk ratio of 1.88. The assumption that 1 patient in the drug-eluting stent group who died, in fact lived, resulted in a crude risk ratio 1.72. Finally, the assumption that both situations existed simultaneously resulted in a crude risk ratio of 1.61, indicating that the association between drug-eluting stent versus bare-metal stent use and mortality must be interpreted with circumspection given the study’s small sample size. Discussion The present study of high-risk dialysis patients undergoing PCI suggests that the use of drug-eluting compared with bare-metal stents is associated with lower MACE rates at 1 year. PCI studies preceding the approval of drug-eluting stents have consistently reported a significantly higher risk of late mortality among dialysis patients compared with patients not treated by dialysis.2,9 Consistent with these reports and others,3 1-year mortality rates among dialysis patients in the Dynamic Registry exceeded 30%, a nine-fold higher compared with patients not on dialysis. Though stenting might provide some prognostic advantage over balloon angioplasty, the preponderance of data preceding drug-eluting stents indicates that surgical revascularization compared with PCI is associated with lower long-term mortality in these high-risk patients.3 Nonetheless, any potential long-term advantages of bypass surgery over PCI in dialysis patients are limited by the invasiveness of this method and the higher risk for periprocedural complications and in-hospital mortality.10,11 With respect to nonfatal MACE, the data on restenosis in dialysis patients are inconsistent. Some reports have noted the risk of angiographic restenosis following balloon angioplasty to be higher in dialysis patients compared with those not on dialysis.12,13 Angiographic follow-up data from stent studies are limited, though with regard to clinical restenosis, most studies conducted in the bare-metal stent era have reported similar rates of repeat revascularization procedures in dialysis patients compared with patients without end-stage renal disease.2,14 Given the high attrition rates among dialysis patients and referral bias against repeat diagnostic and interventional procedures perceived as high-risk in this patient population, as well as the discrepancy between the studies with angiographic versus clinical endpoints, it is uncertain whether bare-metal stents afford a clinically important reduction in the risk of restenosis over balloon angioplasty, as suggested by one small study.15 In patients with mild-to-moderate renal failure, the benefit of drug-eluting stents in reducing the risk of clinical and angiographic restenosis has been shown to be independent of renal function and of similar magnitude to that in patients without renal failure, though the risk of mortality is unaffected by these devices.16 The association between drug-eluting stent use and adverse outcomes in patients with end-stage renal disease (excluded from the pivotal randomized drug-eluting versus bare-metal stent trials4,5) has hitherto not been adequately studied. Previous studies on the efficacy and safety of drug-eluting stents in dialysis patients are limited. A recent study comparing outcomes among patients with chronic renal failure undergoing either drug-eluting or bare-metal stent implantation included 29 and 62 dialysis patients, respectively.17 In that study, in which outcomes in dialysis patients were not analyzed separately, drug-eluting compared with bare-metal stent use was associated with lower rates of repeat revascularization procedures at 6-month follow up, but without a difference in mortality. To our knowledge, this is the first study to compare 1-year clinical outcomes among dialysis patients undergoing coronary stent implantation either with or without drug-eluting devices. We identified a significant reduction in 1-year MACE rates in patients treated with drug-eluting stents compared with patients who received bare-metal stents only, a difference that persisted after adjustment for important confounders. In the present study, 1-year mortality in patients treated with bare-metal stents was well in excess of 30%, a finding consistent with the largest study of coronary revascularization in dialysis patients prior to drug-eluting stents,3 with mortality rates approximately 55% lower in patients treated with drug-eluting stents. A trend indicating lower 1-year covariate adjusted mortality rates with drug-eluting stents was found, although these data must be interpreted cautiously due to the small sample size. Mechanisms underlying a potential beneficial impact of drug-eluting versus bare-metal stents on event-free survival in dialysis patients are speculative. Considering that the majority of fatal cardiac events in dialysis patients are due to sudden death and acute MI,18 and given some data indicating high restenosis rates after bare-metal stenting in this population,19 it is possible that higher target vessel patency rates afforded by drug-eluting stents results in less “jeopardized myocardium”,20,21 and thus a lower tendency to fatal outcomes in the setting of acute ischemia and/or arrhythmic episodes. Composite 1-year MACE rates, which were driven primarily but not solely by mortality, were lower following drug-eluting than following bare-metal stent implantation. While not statistically significant, it should be noted that the 1-year repeat revascularization rate was two-thirds lower in patients undergoing drug-eluting stenting compared to patients in the bare-metal stent group. Considering the small study size and the consistent reduction of restenosis by drug-eluting stents in patients without end-stage renal disease,4,5,16 it is plausible that the failure to detect a significant reduction in clinical restenosis with drug-eluting stents in the present study reflects a type II error (i.e., low power). Though lacking statistical significance, it is also noteworthy that new nonfatal MI events following discharge after the index PCI procedure were more frequent among patients receiving drug-eluting stents than in those treated with bare-metal stents. Whether these data reflect an initial transitory increase in the risk of nonfatal MI rates associated with drug-eluting stent implantation, followed by a late protective effect by these devices, could not be determined by the present small study. Study limitations. The present study is limited by its small sample size. As such, its data are suggestive and should be interpreted with some caution. Treatment allocation within the registry was not randomized, and thus the observed survival benefit with drug-eluting compared to bare-metal stents must be considered hypothesis-generating rather than definitive. The study’s limited power to detect statistically meaningful differences in rates of nonfatal MI and repeat revascularization procedures has been acknowledged. Also, the small sample size did not permit a comparison of outcomes between the 2 drug-eluting stents approved for clinical use during the time of the study (the polymer-based, slow-release, paclitaxel-eluting stent Taxus® [Boston Scientific Corp., Natick, Massachusetts] and the polymer-based, sirolimus-eluting stent Cypher™ [Cordis Corp., Miami, Florida]). Lastly, our findings are not generalizable to drug-eluting stent systems not represented in the study. In conclusion, drug-eluting versus bare-metal stent implantation was associated with better MACE-free survival in dialysis patients undergoing PCI. Given the extremely high risk of cardiac mortality in dialysis patients, as well as prior bare-metal stent data indicating higher mortality after PCI versus bypass surgery, further research including randomized trials comparing drug-eluting stent implantation with surgical revascularization are needed in this population.
References 1. U.S. Renal Data System, USRDS 2005 Annual Data Report: Atlas of End-Stage Renal Disease in the United States. National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, Maryland, 2005, pp. 180‚Äì181. 2. Best PJ, Lennon R, Ting HH, et al. The impact of renal insufficiency on clinical outcomes in patients undergoing percutaneous coronary interventions. J Am Coll Cardiol 2002;39:1113‚Äì1139. 3. Herzog CA, Ma JZ, Collins AJ. Comparative survival of dialysis patients in the United States after coronary angioplasty, coronary artery stenting, and coronary artery bypass surgery and impact of diabetes. Circulation 2002;106:2207‚Äì2211. 4. Moses JW, Leon MB, Popma JJ, et al. Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2003;349:1315‚Äì1323. 5. Stone GW, Ellis SG, Cox DA, et al. A polymer-based, paclitaxel-eluting stent in patients with coronary artery disease. N Engl J Med 2004;350:221‚Äì231. 6. Stone GW, Ellis SG, Cannon L, et al. Comparison of a polymer-based paclitaxel-eluting stent with a bare metal stent in patients with complex coronary artery disease: A randomized controlled trial. JAMA 2005;294:1215‚Äì1223. 7. Laskey WK, Williams DO, Vlachos HA, et al. Changes in the practice of percutaneous coronary intervention: A comparison of enrollment waves in the National Heart, Lung, and Blood Institute (NHLBI) Dynamic Registry. Am J Cardiol 2001;87:964‚Äì969. 8 Cole SR, Hernan MA. Adjusted survival curves with inverse probability weights. Comput Methods Programs Biomed 2004;75:45‚Äì49. 9. Gruberg L, Dangas G, Mehran R, et al. Clinical outcome following percutaneous coronary interventions in patients with chronic renal failure. Catheter Cardiovasc Interv 2002;55:66‚Äì72. 10. Franga DL, Kratz JM, Crumbley AJ, et al. Early and long-term results of coronary artery bypass grafting in dialysis patients. Ann Thorac Surg 2000;70:813‚Äì818. 11. Liu JY, Birkmeyer NJ, Sanders JH, et al. Risks of morbidity and mortality in dialysis patients undergoing coronary artery bypass surgery. Northern New England Cardiovascular Disease Study Group. Circulation 2000;102:2973‚Äì2977. 12. Kahn JK, Rutherford BD, McConahay DR, et al. Short- and long-term outcome of percutaneous transluminal coronary angioplasty in chronic dialysis patients. Am Heart J 1990;119:484‚Äì489. 13. Schoebel FC, Gradaus F, Ivens K, et al. Restenosis after elective coronary balloon angioplasty in patients with end stage renal disease: A case-control study using quantitative coronary angiography. Heart 1997;78:337‚Äì342. 14. Le Feuvre C, Dambrin G, Helft G, et al. Comparison of clinical outcome following coronary stenting or balloon angioplasty in dialysis versus non-dialysis patients. Am J Cardiol 2000;85:1365‚Äì1368. 15. Hase H, Nakamura M, Joki N, et al. Independent predictors of restenosis after percutaneous coronary revascularization in haemodialysis patients. Nephrol Dial Transplant 2001;16:2372‚Äì2377. 16. Halkin A, Mehran R, Casey CW, et al. Impact of moderate renal insufficiency on restenosis and adverse clinical events after paclitaxel-eluting and bare metal stent implantation: Results from the TAXUS-IV Trial. Am Heart J 2005;150:1163‚Äì1170. 17. Kuchulakanti PK, Torguson R, Chu WW, et al. Impact of chronic renal insufficiency on clinical outcomes in patients undergoing percutaneous coronary intervention with sirolimus-eluting stents versus bare metal stents. Am J Cardiol 2006;97:792‚Äì797. 18. Herzog CA. Is there something special about ischemic heart disease in patients undergoing dialysis? Am Heart J 2004;147:942‚Äì944. 19. Azar RR, Prpic R, Ho KK, et al. Impact of end-stage renal disease on clinical and angiographic outcomes after coronary stenting. Am J Cardiol 2000;86:485‚Äì489. 20. Graham MM, Faris PD, Ghali WA, et al. Validation of three myocardial jeopardy scores in a population-based cardiac catheterization cohort. Am Heart J 2001;142:254‚Äì261. 21. Alderman EL, Kip KE, Whitlow PL, et al. Native coronary disease progression exceeds failed revascularization as cause of angina after five years in the Bypass Angioplasty Revascularization Investigation (BARI). J Am Coll Cardiol 2004;44:766‚Äì774.