Comparisons of Baseline Demographics, Clinical Presentation, and Long-Term Outcome Among Patients With Early, Late, and Very Late Stent Thrombosis of Sirolimus-Eluting Stents
Observations From the Registry of Stent Thrombosis for Review and Reevaluation (RESTART)
Background— Stent thrombosis (ST) after sirolimus-eluting stent implantation has not yet been adequately characterized, mainly because of its low incidence.
Methods and Results— The Registry of Stent Thrombosis for Review and Reevaluation (RESTART) is a Japanese nationwide registry of sirolimus-eluting stent–associated ST comprising 611 patients with definite ST (early [within 30 days; EST], 322 patients; late [between 31 and 365 days; LST], 105 patients; and very late [>1 year; VLST], 184 patients). Baseline demographics, clinical presentation, and long-term outcome of sirolimus-eluting stent–associated ST were compared among patients with EST, LST, and VLST. Baseline demographics were significantly different according to the timing of ST. Characteristic demographic factors for LST/VLST versus EST identified by multivariable model were hemodialysis, end-stage renal disease not on hemodialysis, absence of circumflex target, target of chronic total occlusion, prior percutaneous coronary intervention, and age <65 years. For LST versus VLST, they were hemodialysis, heart failure, insulin-dependent diabetes mellitus, and low body mass index. Patients with LST had a significantly higher rate of Thrombolysis in Myocardial Infarction grade 2/3 flow (36%) at the time of ST than those with EST (13%) (P<0.0001) and VLST (17%; P<0.0001). Mortality rate at 1 year after ST was significantly lower in patients with VLST (10.5%) compared with those with EST (22.4%; P=0.003) or LST (23.5%; P=0.009).
Conclusion— ST timing–dependent differences in baseline demographic features, Thrombolysis in Myocardial Infarction flow grade, and mortality rate suggest possible differences in the predominant pathophysiological mechanisms of ST according to timing after sirolimus-eluting stent implantation.
Received August 28, 2009; accepted May 11, 2010.
Although stent thrombosis (ST) is a dreaded complication of percutaneous coronary intervention (PCI) with drug-eluting stents (DES), ST occurring late after DES implantation has not yet been adequately characterized, mainly because of its low incidence. Previous studies reporting risk factors of ST were hampered by the small number of events analyzed.1–4 Furthermore, although the incidence and timing of ST occurring between 30 days and 1 year (late ST [LST]) were reported to be similar after both bare metal stent (BMS) and DES implantation, ST beyond 1 year after stent implantation (very LST [VLST]) was reported to occur more frequently after DES implantation than after BMS implantation.5 The underlying mechanisms might be different between LST and VLST. However, no previous study has evaluated the differences between LST and VLST in terms of baseline characteristics, clinical presentation, and long-term outcome. The present study was undertaken to further characterize ST according to the timing of ST in a large number of patients with ST after sirolimus-eluting stent (SES) implantation in real-world clinical practice in Japan.
Clinical Perspective on p 61
The Registry of Stent Thrombosis for Review and Reevaluation (RESTART) is a Japanese nationwide registry of patients with ST after SES implantation. After the approval of SES in Japan in May 2004, the Japanese Ministry of Health, Labor, and Welfare mandated the SES manufacturing company (Cordis Cardiology Japan, Johnson and Johnson, Tokyo, Japan) to undertake a comprehensive surveillance of ST after SES implantation. The company contacted all PCI centers in Japan every 2 weeks and collected information on the patients with ST after SES implantation. RESTART, as an investigator-initiated study, was designed to adjudicate whether those patients who were reported to have ST after SES implantation did in fact experience definite ST after SES implantation and to collect additional information on those patients. The participating centers were also encouraged to enroll in the RESTART study those patients who had ST but who thus far had not been reported to the company. Adjudication of ST, which was conducted by a physician (T. Kimura) and clinical research coordinators (Hitomi Sasae and Naoko Okamoto) in the data management center (Kyoto University Hospital, Department of Cardiology, Kyoto, Japan), was based on the angiographic findings and signs and symptoms of acute coronary syndrome reported by the site investigators. Logical inconsistencies were resolved by inquiries to the site investigators.
Angiographic and intravascular ultrasound (IVUS) substudies were also conducted, and for patients enrolled in the substudies, coronary angiograms and/or IVUS images both at baseline and at the time of ST were analyzed by the angiographic (Cardiocore, Tokyo, Japan) and IVUS (Cardiovascular Core Analysis Laboratory, Stanford University, Stanford, Calif) core laboratories. The study protocol was approved by the ethics committees of the 2 centers with which the 2 principal investigators are affiliated (Kyoto University Hospital and Teikyo University Hospital). Because this study was conducted retrospectively, written informed consent was waived according to the guidelines for epidemiological studies issued by the Ministry of Health, Labor, and Welfare of Japan.
ST was defined according to the Academic Research Consortium (ARC) definition.6 Only patients with ARC-definite ST were enrolled in RESTART. ST was judged by the site investigators to have occurred if Thrombolysis in Myocardial Infarction (TIMI) flow was grade 0, 1 with occlusion originating in the peri-stent region, 2, or 3 in the presence of angiographic evidence of thrombus associated with signs and symptoms of acute coronary syndrome. ST was categorized according to the timing of occurrence of ST as early ST (within 30 days [EST]), LST (between 31 and 365 days), and VLST (>1 year). Myocardial infarction was adjudicated according to the definition in the Arterial Revascularization Therapy Study.7
Site investigators evaluated status of antiplatelet therapy (APT) at the time of ST by asking the patients and/or their relatives about their compliance with APT. If the patient had discontinued aspirin and/or thienopyridine, the date of discontinuation was recorded in the case report form. If a patient had restarted APT after temporary discontinuation, the patient was included in the group of patients without discontinuation of APT at the time of ST.
ST related to surgical procedures included not only ST occurring after surgical procedures but also ST that occurred while the patient stopped APT before the scheduled surgical procedures. Surgical procedures were defined as any procedure requiring general or local anesthesia except for percutaneous endovascular procedures. Endoscopic therapeutic procedures were included as surgical procedures.
Categorical variables were compared by use of the χ2 test. Continuous variables were expressed as mean±SD unless otherwise indicated. Continuous variables were compared by use of the Student t test or Wilcoxon rank-sum test on the basis of the distribution.
Baseline demographics, clinical presentation of ST and APT at the time of ST, and treatment and outcome of ST were compared according to the timing of ST. Multivariable logistic regression models were constructed using those baseline demographic variables that were significantly different by univariate analysis in the comparison of LST/VLST versus EST and LST versus VLST, respectively. The nonparsimonious multivariable models were refitted so that they retained only those variables that had nominally significant results. The selected models were confirmed by forward and backward procedures. Characteristic demographic factors in each comparison were expressed as odds ratios and their 95% confidence intervals.
Cumulative incidences of recurrent ST and death after the index ST events were estimated by the Kaplan-Meier method, and differences were assessed with the log-rank test. In the comparison of mortality after ST according to the timing of ST, differences in baseline characteristics at the time of the index SES implantation were adjusted by use of a Cox proportional-hazard model.
All analyses were conducted by a physician (T. Kimura) and an independent statistician (T.M.) using SAS software version 9.1 (SAS Institute, Inc, Cary, NC), and all reported P values are 2-sided. Values of P<0.05 were regarded as statistically significant.
Among the 1335 Japanese PCI centers invited, 543 centers agreed to participate in RESTART (see the Appendix in the online-only Data Supplement). From May 2004 to June 2008, 491 128 PCI procedures, including 294 212 procedures with SES, were performed in the participating centers. The study patient flow is shown in Figure 1. The study population for the present analysis consisted of 611 patients with ARC-definite ST reported until June 30, 2008. Except for 1 patient with pathological confirmation only, 610 patients had angiographic confirmation of ST.
There were 322 patients with EST (52 acute ST within 24 hours and 270 subacute ST between 1 and 30 days), 105 patients with LST, and 184 patients with VLST. ST events occurred most commonly during the first 4 weeks, particularly during the first week after SES implantation (Figure 2). During hospitalization for SES implantation, aspirin and thienopyridine were administered in 97% and 97% (ticlopidine, 89%; clopidogrel, 8%) of patients, respectively.
Baseline demographic features were significantly different between the 2 groups of patients with EST and LST/VLST (Table 1). Characteristic demographic factors for LST/VLST compared with EST identified by multivariable logistic model were hemodialysis, end-stage renal disease not on hemodialysis, absence of left circumflex coronary artery target lesion, target lesion of chronic total occlusion, any previously performed PCI, and age <65 years (Table 2). Among 315 patients enrolled in the angiographic substudy, residual dissection was seen significantly more frequently in patients with EST (n=169) compared with those with LST/VLST (n=144) (17% versus 4.2%; P=0.0002; see Table I in the online-only Data Supplement).
Baseline demographic features were also significantly different between the 2 groups of patients with LST and VLST (Table 3). Characteristic demographic factors for LST compared with VLST were hemodialysis, history of heart failure, insulin-dependent diabetes mellitus, and low body mass index (Table 4). The prevalence of hemodialysis in patients with LST was as high as 29%. An angiographic substudy demonstrated that calcification was significantly less prevalent in patients with VLST (n=59) compared with those with LST (n=85) (22% versus 42%; P=0.01; Table II in the online-only Data Supplement).
Clinical Presentation of ST
Overall, clinical presentation of ST was ST-segment elevation acute coronary syndrome in 69%, non–ST-segment elevation acute coronary syndrome in 23%, and cardiac arrest or ventricular fibrillation in 8% of patients, a proportions pattern that did not significantly differ with the timing of ST (Table 5). Multivessel ST was present in 3.4% of patients, which accounted for 17.4% of the 121 ST patients who underwent multivessel stenting. Initial TIMI flow grades at the time of ST were significantly different according to the timing of ST. Patients with LST had a markedly higher rate of TIMI grade 2 or 3 flow compared with patients in both the EST and VLST groups.
Surgical Procedure and Status of APT Before ST
Overall, ST was related to surgical procedures in 4.6% of patients (Figure 3). APT before surgical procedures included none in 68%, aspirin alone in 14%, dual APT in 14%, and thienopyridine alone in 4% of patients (Table 5). Among 20 patients with ST after surgical procedures, dual APT was not stopped before surgery in 4 patients, and no other patients restarted dual APT before ST events.
The status of APT at the time of ST was significantly different according to the timing of ST (Table 5). In patients with EST, 76% of patients were on dual APT at the time of ST. Although the proportion of patients on dual APT decreased with time, 52% of patients with LST and 21% of patients with VLST had ST while they were taking both aspirin and thienopyridine.
Discontinuation of APT before ST events was reported in 216 patients (36%; discontinuation of both thienopyridine and aspirin, 77 patients; discontinuation of thienopyridine only, 133 patients; and discontinuation of aspirin only, 6 patients), with a median interval between discontinuation and ST of 121 days (interquartile range, 14 to 513 days). The median interval from discontinuation of APT to the onset of ST was significantly shorter after discontinuation of both aspirin and thienopyridine than after discontinuation of thienopyridine only (13 days [interquartile range, 6 to 61 days] versus 314 days [interquartile range, 79 to 711 days]; P<0.0001). However, among patients with ST after discontinuation of both thienopyridine and aspirin, 62 patients (81%) had ST beyond 5 days after discontinuation (Figure 4).
Treatments and Outcomes of ST
Treatments of ST were generally similar according to the timing of ST (Table 6). PCI was performed in 97% of patients with frequent use of thrombus aspiration. Additional stent placement was performed in 36% of patients. Final TIMI grade 3 flow after PCI was achieved in 84% of patients. Emergency coronary artery bypass graft surgery was performed in only 1.2% of patients.
Myocardial infarction was the most common clinical sequel of ST, occurring in 89% of patients. During follow-up, cumulative incidence of recurrent ST at 1 year after the index ST was 4.6% (Figure 5A). Cumulative incidence of death after the index ST events was significantly lower in patients with VLST compared with patients with EST or LST (Figure 5B). However, after adjustment for baseline demographics, differences were of borderline significance for EST versus VLST (hazard ratio, 2.09; 95% confidence interval, 1.07 to 4.07; P=0.03) and not significant for LST versus VLST (hazard ratio, 1.29; 95% confidence interval, 0.61 to 2.74; P=0.5). Cumulative incidence of death in patients with multivessel ST was 39% at 1 year after the index ST events.
The main finding of the present study analyzing the largest number of patients with SES-associated ST ever reported was that baseline characteristics, TIMI flow grade at the time of ST, and mortality rate after ST were markedly different according to the timing of ST after SES implantation.
Early ST within 30 days has been relatively well characterized compared with ST occurring beyond 30 days. Risk factors reported for EST included acute coronary syndrome, suboptimal APT, and procedural factors such as residual dissection and underexpansion of stents, which were mostly consistent across studies.1–4 Several recent studies assessed risk factors for late ST beyond 30 days.1–4 In the Estudio Español Sobre Trombosis De Stents Farmacoactivos(ESTROFA) registry, ST-elevation myocardial infarction, stenting of the left anterior descending coronary artery, and long stents were identified as risk factors for late ST beyond 30 days, whereas in the Dutch ST registry, undersizing, malignancy, proximal lesion, peripheral artery disease, diabetes mellitus, bifurcation, long stents, and younger age were reported to be risk factors for late ST.2,3 The apparent inconsistency for the reported risk factors across studies might partly be explained by the fact that no previous study discriminated between LST and VLST. In the present study, we demonstrated marked differences in the baseline characteristics between the patients with LST and VLST, suggesting that the predominant mechanisms also might be different between LST and VLST. From these findings, distinguishing between LST and VLST in future investigations of ST seems to be crucial.
In the earlier pathological studies of BMS, nearly complete endothelialization was reported to be present by 3 to 4 months after BMS implantation.8 A postmortem human pathological study reported that in all 14 patients with late ST, including VLST of DES, delayed arterial healing was found to be a cardinal risk factor and was the only pathological risk factor in 3 patients (21%).9 However, if the predominant mechanisms of LST were to be related only to delayed arterial healing, it would be difficult to explain why the rate of ST at 1 year was similar between SES- and BMS-treated lesions despite the fact that the latter were reported to heal better pathologically.5,9 It is noteworthy that hemodialysis and insulin-dependent diabetes mellitus were highly prevalent in patients with LST. Although restenosis was dramatically reduced by SES, restenosis is most commonly seen during the first year of SES implantation. Hemodialysis and insulin-dependent diabetes mellitus were reported to be very strong risk factors for restenosis of SES.10 In addition, patients with LST had a markedly higher rate of TIMI grade 2 or 3 flow at the time of ST compared with those in both the EST and VLST groups, and it is possible that patients with a very aggressive restenosis process were judged as having ST. Alternatively, there might be some causal relationship between neointimal hyperplasia and ST within the first year. Significant neointimal proliferation might be prone to ultimate thrombus formation. In a postmortem human pathological study, 2 cases of DES LST and 2 cases of BMS LST were reported to be causally related to in-stent restenosis with superimposed thrombus.9 The hypothesis of a causal relationship between neointimal hyperplasia and ST could partially explain why the rate of ST at 1 year was not lower after BMS implantation than after SES implantation.5,9
The mechanisms of VLST are currently very poorly understood. Delayed arterial healing had also been incriminated as the predominant cause of VLST. However, in a serial analysis using optical coherence tomography, additional neointimal coverage was observed between 6 and 12 months after SES implantation.11 In addition, a case report of human pathological evaluation of an SES-treated lesion showed complete endothelialization at 11 months after implantation.12 Despite these observations suggesting improved healing with time after DES implantation, the slopes of the cumulative incidence curve of VLST were constant (0.3% to 0.6%) up to 3 years.1,4 Therefore, some mechanisms other than delayed arterial healing must be operative in the pathogenesis of VLST. Cook et al13 compared the IVUS findings in 13 patients who had VLST with those findings in 144 patients undergoing routine IVUS examination at 8 months after DES implantation. Incomplete stent apposition was reported to be more frequent (77% versus 12%) with a larger incomplete stent apposition area in patients with VLST compared with control subjects. Stented segments in patients with VLST revealed more marked positive remodeling compared with control subjects. The same group of investigators extended the observation to another 10 patients with VLST by both IVUS and histological examination of the thrombi aspirated, demonstrating that VLST was associated with histopathological signs of inflammation and IVUS evidence of vessel remodeling.14 They also demonstrated that eosinophilic infiltrates were more common in thrombi harvested from patients with VLST compared with other causes of myocardial infarction. These observations were consistent with chronic inflammation (hypersensitivity) reported in several autopsy cases after DES implantation, suggesting profound chronic inflammation as one of the most dominant mechanisms of VLST.9
Furthermore, the de novo atherosclerosis, defined as lipid-laden foamy macrophage infiltrates within the neointima above the stent that did not communicate with the underlying atherosclerotic plaque, was reported to be seen earlier and more frequently in lesions treated with DES compared with those treated with BMS.15 Those de novo atherosclerotic plaques might be susceptible to VLST.
It is intriguing that diabetic and elderly patients were less common in the group of patients with VLST, a finding consistent with reports from the Bern/Rotterdam cohort and from the ESTROFA registry.1,2 Together with the finding that angiographic calcification was significantly less common in lesions with VLST, this indicates that very advanced atherosclerosis such as that often seen in diabetic and elderly patients might be less susceptible to the inflammatory reactions suspected as the dominant mechanism of VLST.9,14 Alternatively, fibrotic maturation of the relatively thick neointima after DES implantation in diabetic patients might protect the thrombogenic constituents of the plaque from being exposed to the bloodstream in a fashion similar to the plaque-sealing effect of BMS.
Relative to the status of APT at the time of ST, one third of the patients had ST after discontinuation of APT. Although the proportion of patients on aspirin monotherapy increased with time after SES implantation, long intervals between discontinuation of thienopyridine only and ST cast doubt on the causal link between discontinuation and ST, particularly in cases of VLST. Relatively short intervals between discontinuation of both thienopyridine and aspirin and ST suggest a causal relationship between discontinuation and ST. These observations are consistent with our prior report suggesting an increased risk of ST in patients stopping both aspirin and thienopyridine but not in patients stopping thienopyridine only.4
Finally, it is noteworthy that multivessel ST was reported in 3.4% of patients, which accounted for 17.4% of the 121 ST patients undergoing multivessel stenting. This observation suggests that the pathological processes prone to ST may not be limited to 1 site in patients undergoing multivessel stenting. In addition, ST in 1 lesion might be the trigger of ST in another lesion through several mechanisms, including activation of platelets and the sympathetic nervous system. Considering the dismal prognosis of multivessel ST and the potential involvement of platelet activation in this situation, stringent compliance and adherence to APT seem to be particularly important in patients undergoing multivessel stenting.
The present study has several important limitations. First, although this study included the largest number of patients with SES-associated ST ever reported, we did not have data on control patients without ST. Therefore, we could not evaluate the risk factors of ST according to the timing after SES implantation. Second, we did not use angiographic core laboratory evaluation for the presence of thrombus and TIMI flow at the time of ST. Therefore, adjudication of patients with TIMI grade 2 or 3 flow as having ST was dependent on the interpretation of the site investigators. However, among 313 patients included in the angiographic substudy, 301 patients (96%) were confirmed by the angiographic core laboratory as having ARC-definite ST with angiographic evidence of thrombus. In addition, TIMI flow grade reported by the site investigators was confirmed to be concordant with the evaluation by the angiographic core laboratory in 91% of cases. Third, considering that 224 patients were reported spontaneously by the referring centers to RESTART, the accuracy of the system monitoring ST by the government and the SES manufacturing company seemed to be far from complete. In addition, we should admit that we could not guarantee consecutive enrollment of ST patients by all the participating centers. ST events in patients who had lost contact with the referring hospital were likely not to be reported. Furthermore, the mean number of ST cases per center was only 1.24. Among 543 participating centers, 279 centers did not report any ST cases. Therefore, it is possible that ST cases might have been underreported by some participating centers. Finally, although we used the ARC definition to classify the timing of ST, we are not sure whether the ARC classification of ST is the best way to account for the different pathophysiological mechanisms that may be operative at various times after stent implantation.
ST timing–dependent differences in baseline demographic features, TIMI flow grade, and mortality rate suggest possible differences in the predominant pathophysiological mechanisms of ST according to timing after SES implantation.
We appreciate the efforts of the investigators in the 543 participating centers and of the clinical research coordinators supporting the study. We also appreciate the secretarial support by Hiromi Yoshida, Megumi Hirose, and Mai Fujino.
Source of Funding
This work was supported by Cordis Cardiology Japan, Johnson and Johnson.
Dr Kimura is an advisory board member for, speaker for, and recipient of research grants from Cordis Cardiology Japan, Johnson and Johnson. Dr Kozuma is a recipient of honoraria from Cordis Cardiology Japan, Johnson and Johnson. Drs Aizawa and Isshiki are advisory board members for and recipients of honoraria from Cordis Cardiology Japan, Johnson and Johnson. Drs Miyazaki and Yamaguchi are advisory board members for Cordis Cardiology Japan, Johnson and Johnson. E. Hiyoshi and E. Nishimura are full-time employees of Cordis Cardiology Japan, Johnson and Johnson. The other authors report no conflicts.
Daemen J, Wenaweser P, Tsuchida K, Abrecht L, Vaina S, Morger C, Kukreja N, Jüni P, Sianos G, Hellige G, van Domburg RT, Hess OM, Boersma E, Meier B, Windecker S, Serruys PW. Early and late coronary stent thrombosis of sirolimus-eluting and paclitaxel-eluting stents in routine clinical practice: data from a large two-institutional cohort study. Lancet. 2007; 369: 667–678.
de la Torre-Hernandez JM, Alfonso F, Hernandez F, Elizaga J, Sanmartin M, Pinar E, Lazano I, Vazquez JM, Bosta J, de Prado AP, Hernandez JM, Sanchis J, Ruiz Nodar JM, Gomez-Jaume A, Larman M, Diarte JA, Rodriguez-Collado J, Rumoroso JR, Lopez-Minguez JR, Mauri J, for the ESTROFA Study Group. Drug-eluting stent thrombosis: results from the multi-center Spanish registry ESTROFA. J Am Coll Cardiol. 2008; 51: 986–990.
Kimura T, Morimoto T, Nakagawa Y, Tamura T, Kadota K, Yasumoto H, Nishikawa H, Hiasa Y, Muramatsu T, Meguro T, Inoue N, Honda H, Hayashi Y, Miyazaki S, Oshima S, Honda T, Shiode N, Namura M, Sone T, Nobuyoshi M, Kita T, Mitsudo K, for the j-Cypher Registry Investigators. Antiplatelet therapy and stent thrombosis after sirolimus-eluting stent implantation. Circulation. 2009; 119: 987–995.
Cutlip DE, Windecker S, Mehran R, Boam A, Cohen DJ, van Es GA, Steg PG, Morel M, Mauri L, Vranckx P, McFadden E, Lansky A, Hamon M, Krucoff MW, Serruys PW, for the Academic Research Consortium. Clinical end points in coronary stent trials: a case for standardized definitions. Circulation. 2007; 115: 2344–2351.
Serruys PW, Unger F, Sousa JE, Jatene A, Bonnier HJ, Schonberger JP, Buller N, Bonser R, van den Brand MJ, van Herwerden LA, Morel MA, van Hout BA. Comparison of coronary-artery bypass surgery and stenting for the treatment of multi-vessel disease. N Engl J Med. 2001; 344: 1117–1124.
Ikari Y, Kotani J, Kozuma K, Kyo E, Nakamura M, Yokoi H, for the J-PMS Study Group. Assessment of sirolimus-eluting coronary stent implantation with aspirin plus low dose ticlopidine administration: one year result from the Cypher Stent Japan Post-Marketing Surveillance Registry (J-PMS). Circ J. 2009; 73: 1038–1044.
Cook S, Wenaweser P, Togni M, Billinger M, Morger C, Seiler C, Vogel R, Hess O, Meier B, Windecker S. Incomplete stent apposition and very late stent thrombosis after drug-eluting stent implantation. Circulation. 2007; 115: 2426–2434.
Cook S, Ladich E, Nakazawa G, Eshtehardi P, Neidhart M, Vogel R, Togni M, Wenaweser P, Billinger M, Seiler C, Gay S, Meier B, Pichler WJ, Juni P, Virmani R, Windecker S. Correlation of intravascular ultrasound findings with histopathological analysis of thrombus aspirates in patients with very late drug-eluting stent thrombosis. Circulation. 2009; 120: 391–399.
Stent thrombosis (ST) after sirolimus-eluting stent implantation has not yet been adequately characterized, mainly because of its low incidence. The Registry of Stent Thrombosis for Review and Reevaluation (RESTART) is a Japanese nationwide registry of sirolimus-eluting stent–associated ST comprising 611 patients with definite ST (early [within 30 days; EST], 322 patients; late ]between 31 and 365 days; LST], 105 patients; and very late [>1 year; VLST], 184 patients). Baseline demographics, clinical presentation, and long-term outcome of sirolimus-eluting stent–associated ST were compared among patients with EST, LST, and VLST. Baseline demographics were significantly different according to the timing of ST. Characteristic demographic factors for LST/VLST versus EST identified by multivariable model were renal failure, absence of circumflex target, target of chronic total occlusion, prior percutaneous coronary intervention, and age <65 years. For LST versus VLST, they were hemodialysis, heart failure, insulin-dependent diabetes mellitus, and low body mass index. Patients with LST had a significantly higher rate of Thrombolysis in Myocardial Infarction grade 2/3 flow (36%) at the time of ST than those with EST (13%; P<0.0001) and VLST (17%; P<0.0001). Mortality rate at 1 year after ST was significantly lower in patients with VLST (10.5%) compared with those with EST (22.4%; P=0.003) or LST (23.5%; P=0.009). From these findings, distinguishing between LST and VLST in future investigations of ST seems to be crucial. ST timing–dependent differences in baseline demographic features, Thrombolysis in Myocardial Infarction flow grade, and mortality rate suggest possible differences in the predominant pathophysiological mechanisms of ST according to timing after sirolimus-eluting stent implantation.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.109.903955/DC1.