Target Lesion Revascularization after Bare-Metal or Drug-Eluting Stents: Clinical Presentations and Outcomes

   ABSTRACT: Objective. We sought to examine the clinical presentations and subsequent clinical outcomes of patients undergoing target lesion revascularization (TLR) after either bare-metal stent (BMS) or drug-eluting stent (DES) placement. Background. The widely held notion that BMS TLR is benign has recently been challenged. While DES substantially reduce TLR, little is known about the clinical syndromes accompanying DES TLR and the long-term clinical outcomes after TLR. Methods. The clinical syndrome at the time of hospitalization when TLR was performed and subsequent clinical outcomes after TLR were assessed in 1,147 BMS patients and 1,246 DES patients who were followed for 3 years. Patients were considered to have TLR when repeat target lesion PCI was required including those with myocardial infarction (MI) and stent thrombosis. Results. At 3 years, the overall incidence of TLR was higher after BMS compared to DES 98/1,147 (9.2%) vs. 56/1,246 (4.5%); p Conclusions. The clinical presentation at the time of TLR is not always a benign clinical event and identifies a subgroup of stent-treated patients at high risk for non-fatal MI or death in the 3 years following the index percutaneous coronary intervention, independent of stent type. J INVASIVE CARDIOL 2010;22:266–270 Key words: revascularization, mortality, stents    In-stent restenosis occurs in 20–30% of cases following bare- metal stent (BMS) use.^1,2 Clinically, restenosis had been viewed as a benign entity, with infrequent cases of myocardial infarction (MI) as the presenting event.^3,4 Recent studies, however, have challenged this notion, observing acute coronary syndromes including unstable angina in up to 50–70% and MI in up to 19% of patients at the time of presentation of restenosis.^5–9 Drug- eluting stents (DES) have substantially reduced the need for repeat revascularization compared to BMS.^10,11 However, only limited contemporary data are available evaluating the type of clinical syndrome other than stent thrombosis at the time of presentation resulting in target lesion revascularization (TLR).^12,13 Moreover, the subsequent fate of patients undergoing TLR has not been fully characterized. Accordingly, we evaluated the clinical presentations and subsequent non-fatal MI or death outcomes of “real-world” consecutive patients with BMS and DES requiring TLR within 3 years of their index percutaneous coronary intervention (PCI).

Methods

   Patients at our institution (WFUBMC) undergoing PCI from April 2002 to April 2005 were eligible for inclusion in the study. A total of 1,147 patients underwent coronary artery stenting with BMS between April 2002 and April 2003 (prior to FDA approval of DES in the United States) and were available for clinical follow up, including 105 (9.2%) patients who underwent TLR after their index procedure. 1,246 consecutive patients received DES between February 2004 and April 2005 (after DES were fully available and had replaced BMS as the stent of choice) and were available for clinical follow up, including 56 (4.5%) patients who underwent TLR after their index PCI.¹⁴ TLR was defined as the need for repeat target lesion PCI and included patients who presented with MI and stent thrombosis. Patients were not excluded from the study for any reason. The study was approved by the institutional review board of WFUBMC. These patients were included in previous reports from our aggregate PCI experience.^14–16    PCI was performed according to standard techniques. Anti- coagulation during PCI was accomplished with unfractionated heparin or bivalirudin per standard protocol. Patients received glycoprotein IIb/IIIa receptor inhibition according to usual protocol with abciximab or eptifibatide at the discretion of the in- terventionist.¹⁴ All patients were treated with aspirin (81–325 mg a day) prior to PCI and indefinitely thereafter. Patients also received clopidogrel (300–600 mg as a loading dose given prior to or immediately after the procedure, followed by 75 mg/day). Clopidogrel was given for a minimum of 1 month in BMS- treated patients, for a minimum of 3 months for sirolimus-eluting-stent-treated patients, and for a minimum of 6 months for paclitaxel-eluting-stent-treated patients. Additional clopidogrel use was at the discretion of the physician responsible for the clinical care of the patient.    Prior to hospital discharge, patient and procedural data and hospital outcomes were entered into the WFUBMC Cardiovascular Information Services Database. Collection of data and outcomes measures conformed to the American College of Cardiology National Cardiovascular Database Registry definitions for cardiovascular data.¹⁷ Clinical follow up at 3 years for the entire group of BMS and DES patients has been previously de- scribed.¹⁴ For the patients with TLR, additional follow up was collected up to 1 year after the TLR. Patients who reported undergoing TLR at an outside institution were sent a form authorizing the release of medical records. If the patient consented, medical records from the other institution were collected to determine if a TLR actually occurred.    Clinical outcomes were defined per recent guidelines.¹⁸ Stent thrombosis was defined as presentation with acute coronary syndrome (ACS) and definite angiographic or pathologic evidence of stent thrombosis. Non-fatal MI was defined as ischemic symptoms and an elevation of CK-MB or troponin I above the upper limit of normal, with or without ST elevation or development of Q waves.    Statistical methods. Baseline and repeat revascularization characteristics were compared by TLR status and stent type using chi-square testing or Fisher’s exact test where appropriate for categorical variables or the Wilcoxon rank sum test for continuous variables. The unadjusted hazards of non-fatal MI or death outcomes were evaluated by stent type by comparing Kaplan-Meier estimates of cumulative incidence using the log-rank test. For the analysis of cumulative 3-year non-fatal MI or death by TLR status and stent type, the cumulative incidence of non- fatal MI or death was estimated from the time of index PCI out to 3 years, including those events occurring at the time of presentation for TLR. For the landmark analysis of non-fatal MI or death occurring after TLR, the cumulative incidence of non-fatal MI or death was estimated only from those events occurring after discharge from the hospital for the TLR out to 1 year post TLR. Multivariable adjusted analysis of independent predictors of TLR was performed using Cox proportional hazards modeling. Variable retention in the final model involved a combination of backward selection of statistically significant (p Results    Characteristics at presentation for TLR. Over the 3-year follow-up period, TLR occurred in 98/1,147 (9.2%) BMS compared to 56/1,246 (4.5%) DES; p Clinical outcomes. The cumulative hazard of non-fatal MI or death up to 3 years following index stenting was much higher in those undergoing TLR compared to those not undergoing TLR; hazard ratio (HR) 2.65 (2.00–3.52), independent of stent type (Figure 1). We performed a secondary landmark analysis of events only occurring after TLR (Figure 2) which indicated that the hazard of non-fatal MI or death was lower after DES TLR compared to BMS TLR (HR 0.46 [0.16–1.39]), due to a numer- ically higher number of cardiac deaths in the BMS subgroup. However, the number of patients with MI or death after TLR was very small (BMS = 16, DES = 4).

Discussion

   In this retrospective observational study, patients initially treated with DES underwent TLR far less frequently than those initially treated with BMS over a 3- year study period. However, the clinical syndromes at the time of presentation resulting in TLR were similar. The clinical presentation at the time of TLR was not always benign, with an MI in up to one- third of the patients, regardless of stent type. Over the course of 3 years, patients with TLR had a three-fold increase in the hazard of non-fatal MI or death compared to those without TLR, independent of stent type. These findings suggest that in contemporary practice, TLR is often accompanied by a MI, and identifies patients at particularly high risk for subsequent non-fatal MI or death.    In-stent restenosis (ISR) after BMS has generally been viewed as a benign clinical entity, with a very low incidence of MI.^3,4 However, recent data reexamining the clinical presentations associated with BMS restenosis have observed a MI in up to 19% of patients.^5–9 Bainey et al reported 70% of patients with clinical ISR requiring repeat angiography presented with ACS.⁵ Likewise, in a report comparing BMS to angioplasty, Walters et al reported 68% of patients with BMS ISR presented with ACS.⁹ Similarly, in our study population, BMS were associated with ACS in the majority of patients returning for clinically driven TLR. The reasons for the apparent differences in clinical syndromes at the time of BMS restenosis in these studies are not clear. Potential explanations include reporting biases with respect to cardiac biomarkers, a lower threshold in contemporary practice for admitting patients with a first episode of recurrent chest pain and more aggressive index revascularization in higher- risk patients in more contemporary studies. Utilization of dual- antiplatelet therapy may have been suboptimal. However, a similar proportion of patients with TLR of both stent types were taking DAPT at the time of the TLR (50%). Moreover, the DAPT usage at 2 years exceeded the recommendations for DAPT for DES at the time the study was conducted. Thus, it is unlikely that differences in DAPT utilization accounted for the study observations. Further studies will need to be performed to more clearly elucidate the reasons responsible for these apparent differences.    There is also the possibility that the events leading to a TLR might represent a manifestation of delayed thrombosis, and not just intimal proliferation. Stent thrombosis, acute or otherwise, with complete vessel obstruction is readily identifiable both clinically and at the time of urgent cardiac catheterization. In- complete obstruction within a stent not accompanied by either clinical or angiographic features of stent thrombosis has traditionally been attributed to intimal proliferation, but not stent thrombosis. However, analogous to incomplete versus complete flow obstruction due to thrombus in N-STEMI versus STEMI, it is conceivable that the infarctions we observed at the time of TLR arose from a form of incomplete stent thrombosis. There are numerous studies documenting incomplete endothelialization of DES,^19,20 and pathologic studies noting thrombin formation at DES stents.²¹ Considerably less data are available on BMS. However, scattered reports from angioscopic^22,23 and pathologic observations^24,25 suggest that incomplete stent endothelialization may occur with BMS, but at a much lower frequency. Since incomplete endothelialization is believed to serve as a nidus for thrombus formation within stents, it is plausible that the necessary substrate for thrombosis may be present in a small percentage of patients after BMS placement. Further observations in BMS-treated patients will be necessary to address this possibility.    The relative proportion of TLR manifesting as stent throm- bosis observed in this study merits further discussion. Traditionally, TLR of coronary stents has been synonymous with the PCI procedure performed to treat restenosis due to intimal hyperplasia or recoil within the stent. This has been supported by a recent position paper describing ARC definitions of clinical endpoints.¹⁸ Including stent thrombosis as a measure of TLR as defined in this study results in the apparently paradoxical observation that the proportion of DES TLR manifesting as stent thrombosis is higher than that observed with BMS, despite similar overall rates of stent thrombosis for DES and BMS at 3 years (1.8% DES, 1.7% BMS; p = 0.837).¹⁴ The explanation for this apparent dichotomy is that stent thrombosis rates were similar for the two stent types, while there was a significantly increased rate of non-stent thrombosis TLR with BMS compared to DES. This resulted in a proportionately greater incidence of DES stent thrombosis manifesting as TLR than for BMS. Inclusion of stent thrombosis as a form of TLR may seem somewhat novel. However, we believe it adds value to the evaluation of late stent events since it represents a biologic response to stent use and results in a repeat PCI within the previously placed stent. Further studies will help elucidate if this is an important metric in the evaluation of events late after DES use.    There are few studies of routine clinical practice evaluating the clinical outcomes of the subgroup of patients under- going TLR after both BMS and DES.^12,13,26 The subgroup undergoing TLR had a substantially higher risk of subsequent non-fatal MI or death than those who did not undergo TLR in the 3 years after the index PCI. Our study did not allow evaluation of the potential mechanism of the apparent post-TLR DES benefit. However, index PCI of saphenous vein grafts (SVGs) was associated with an increased risk of TLR and is known to be associated with worse long-term outcomes than with diseased native coronaries. In our TLR cohort, 10% of BMS and 18% of DES TLR procedures involved one or more SVGs (p = 0.185), with post-TLR landmark mortality rates of 33% with BMS and 0% with DES (p = 0.245). Some of the increased rate of adverse outcomes up to 3 years can be explained by the incidence of MI accompanying the presentation at the time of the TLR. However, subsequent non-fatal MI events occurred up to 1 year post TLR, particularly in those with BMS, independent of MI events at the time of presentation for TLR. Thus, unmeasured confounders may have also influenced the observed out- comes and further evaluation is merited.    Study limitations. Our study was limited in size (105 BMS and 56 DES patients). However, these numbers represent the total number of patients who required TLR from one entire year of stent therapy at one institution and thus reflect what is seen in routine clinical practice. Observational studies such as this are potentially subject to selection bias due to unequal follow up. However, we had similar rates of follow up to 3 years by stent type (96% BMS, 91% DES). Thus, selection bias resulting from failure to detect TLR occurring within 3 years was unlikely. The DES cohort may have benefited from advances in medical treatment relative to a historical, rather than contemporaneous, BMS cohort. The BMS cohort was comprised of patients who underwent index PCI in the year prior to DES approval, while the DES cohort was comprised of patients who underwent index PCI more than 1 year after DES gained FDA approval for use. However, use of historical BMS controls avoided the potential selection bias of stent choice when both stents were available, and similarities in dual-antiplatelet therapy at the time of presentation for TLR are reassuring that potential bias resulting from changes in medical treatment practices was minimal. Finally, the observations from a single center such as this may not be generalizable to other practices or institutions. However, the incidence of TLR for both the BMS and DES cohorts are consistent with rates observed in other observational and registry studies of outcomes of these stents in routine clinical practice.²⁷

Conclusion

   Over 3 years, patients with DES underwent TLR less often compared to those with BMS, although the clinical syndromes at the time of presentation resulting in TLR for the two stent types were similar. The clinical presentation at the time of TLR, however, was not always benign, with an MI including stent thrombosis occurring in up to one-third of the patients, regard- less of stent type. Over the course of 3 years, patients with TLR had a three-fold increase in the hazard of non-fatal MI or death compared to those without TLR, independent of stent type. These findings suggest that in contemporary practice, TLR is often accompanied by a MI and identifies patients at particularly high risk for subsequent non-fatal MI or death.

_________________________________________________ From Wake Forest University School of Medicine, Section of Cardiology, Winston-Salem, North Carolina. The authors report no conflicts of interest regarding the content herein. Manuscript submitted November 10, 2009, provisional acceptance given January 20, 2010, final version accepted March 18, 2010. Address for correspondence: Robert J. Applegate, MD, Section of Cardiology, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27157-1045. E-mail: bapplega@wfubmc.edu

1. Serruys PW, DeJaegere P, Kiemeneij F, et al. A comparison of balloon expandable stent implantation with balloon angioplasty in patients with coronary artery dis- ease. N Engl J Med 1994;331:489-495.
2. Fischman DL, Leon MB, Baim D, et al. A randomized comparison of coronary stent placement and balloon angioplasty in the treatment of coronary artery dis- ease. N Engl J Med 1994;331:496–501.
3. Greenberg D, Bakhai A, Cohen DJ. Can we afford to eliminate restenosis? Can we afford not to? J Am Coll Cardiol 2004;43:513–518.
4. Steinberg DH, Slottow TLP, Buch AN, et al. Impact of in-stent restenosis on death and myocardial infarction. Am J Cardiol 2007;100:1109–1113.
5. Bainey KR, Norris CM, Graham MM, et al. Clinical in-stent restenosis with bare metal stents: Is it truly a benign phenomenon? Internatl J Card 2008;128, 378–382.
6. Celik T, Iyisoy A, Yuksel C, et al. The clinical presentation of restenosis after bare metal stent implantation: Behind a facade of benignity. Internatl J Card 2009;135:401–403.
7. Chen MS, John JM, Chew DP, et al. Bare metal stent restenosis is not a benign clin- ical entity. Am Heart J 2006;151:1260–1264.
8. Nayak AK, Kawamura A, Nesto RW, et al. Myocardial infarction as a presentation of clinical in-stent restenosis. Circulation 2006;70:1026–1029.
9. Walters DL, Harding SA, Walsh CR, et al. Acute coronary syndrome is a common clinical presentation of in-stent restenosis. Am J Cardiol 2002;89:491–494.
10. 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;349:1315–1323.
11. Stone GW, Ellis SG, Cox DA, et al. A polymer-based, paclitaxel-eluting stent in pa- tients with coronary artery disease. N Engl J Med 2004;350:221–231.
12. Chacko R, Mulhearn M, Novack V, et al. Impact of target lesion and nontarget lesion cardiac events on 5-year clinical outcomes after sirolimus-eluting or bare-metal stent- ing. J Am Coll Cardiol Intv 2009;2:498–503.
13. Leon MB, Allocco DJ, Dawkins KD, Baim DS. Late clinical events after drug-eluting stents: The interplay between stent-related and natural history-driven events. J Am Coll Cardiol Intv 2009;2:504–512.
14. Applegate RJ, Sacrinty MT, Kutcher MA, et al. 3-year comparison of drug-eluting stents versus bare metal stents. J Am Coll Cardiol Intv 2009;2:231–239.
15. Applegate RJ, Sacrinty MT, Kutcher MA, et al. "Off-Label" stent therapy: 2-year com- parison of drug-eluting versus bare-metal stents. J Am Coll Cardiol 2008;51:607–614.
16. Applegate RJ, Sacrinty MT, Kutcher MA, et al. Comparison of drug-eluting versus bare metal stents on later frequency of acute myocardial infarction and death. Am J Cardiol 2007;99:333–338.
17. American College of Cardiology-National Cardiovascular Data Registry Module Ver- sion 3.0. 2004. Available at: http://www.accncdr.com/webNCDR/NCDRDocu- ments/datadictdefsonlyv30.pdf. American College of Cardiology. 6-1-2006.
18. Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in coronary stent trials: A case for standardized definitions. Circulation 2007;15: 2344–2351.
19. Ako J, Morino Y, Honda Y, et al. Late incomplete stent apposition after sirolimus- eluting stent implantation: A serial intravascular ultrasound analysis. J Am Coll Car- diol 2005;46:1002–1005.
20. Kotani J, Awata M, Nanto S, et al. Incomplete neointimal coverage of sirolimus-elut- ing stents: Angioscopic findings. J Am Coll Cardiol 2006;47:2108–2111.
21. Joner M, Finn AV, Farb A, et al. Pathology of drug-eluting stents in humans: Delayed healing and late thrombotic risk. J Am Coll Cardiol 2006;48:193–202.
22. Ichikawa M, Mishima M. Neointimal regression-induced incomplete coverage of a bare-metal stent in the left main trunk: Serial angiographic and angioscopic evidence obtained by 5-year follow-up. Catheter Cardiovasc Interv 2009;73:787–790.
23. Mudra H, Regar E, Klauss V, et al. Serial follow-up after optimized ultrasound-guided deployment of Palmaz-Schatz stents. In-stent neointimal proliferation without sig- nificant reference segment response. Circulation 1997;95:363–370.
24. Farb A, Burke AP, Kolodgie FD, Virmani R. Pathological mechanisms of fatal late coronary stent thrombosis in humans. Circulation 2003;108:1701–1706.
25. Hasegawa K, Tamai H, Kyo E, et al. Histopathological findings of new in-stent le- sions developed beyond five years. Catheter Cardiovasc Interv 2006;68:554–558.
26. Doyle B, Rihal CS, O’Sullivan CJ, et al. Outcomes of stent thrombosis and restenosis during extended follow-up of patients treated with bare-metal coronary stents. Circulation 2007;116:2391–2398.
27. Brodie BR, Stuckey T, Downey W, et al. Outcomes and complications with off- label use of drug-eluting stents: Results from the STENT (Strategic Transcatheter Evaluation of New Therapies) group. J Am Coll Cardiol Intv 2008;1:405–414. Er- ratum in J Am Coll Cardiol Intv 2008;1:601.