Stent thrombosis (ST) is the sudden occlusion of a stented coronary artery due to thrombus formation. The clinical consequences of ST are frequently catastrophic – most episodes occur within two weeks of stent implantation – and include death in 20–48% or major myocardial infarction (MI) in 60–70% of cases.1–3 When bare-metal stents (BMS) were first introduced, ST was a common complication. Technical improvements, such as the use of adequately sized balloons and high-pressure deployment, helped to substantially decrease thrombosis rates.3,4 In addition, a regimen of dual-antiplatelet therapy (aspirin plus a thienopyridine) was found to further reduce the incidence of stent thrombosis.3 Dual-antiplatelet therapy is prescribed for four weeks after BMS implantation, the time required for stent endothelialisation to take place. Thrombosis that occurs more than 30 days after the implantation of a BMS is rare. This complication first appeared when intracoronary radiation (brachytherapy) was introduced as a treatment for in-stent restenosis.5,6 Intra-coronary radiation was found to delay re-endothelialisation in animals, thus predisposing to late ST. Prolonging the duration of dual-antiplatelet therapy to six to 12 months largely resolved this problem.5,6 Drug-eluting stents (DES) have an antiproliferative effect, similar to brachytherapy, that delays stent re-endothelialisation. For this reason, DES trials have mandated dual-antiplatelet therapy for two to six months, as well as the indefinite use of aspirin.7
Drug-eluting Stent Thrombogenicity
Stent platforms, polymer coating, eluted drug and other factors may contribute to DES thrombogenicity.8 First, stent design may influence the degree of platelet activation after coronary stent deployment. Second, both the polymer coating and the medication with which it is impregnated may influence the propensity for thrombosis.8 Which of these factors is the pathological basis of DES thrombosis? Virmani et al. proposed initially that a hypersensitivity reaction could contribute to the development of this event by describing a case of fatal acute MI and cardiac rupture as a result of late thrombosis of a Cypher stent deployed 18 months previously.9 The hypersensitivity reaction could be caused by the metallic polymer or sirolimus stent. There is a likely spectrum of allergic responses to DES in sensitive patients, varying from benign reactions to excessive inflammation with medial destruction, stent malapposition and aneurysm formation with late in-stent thrombosis. In addition, despite animal data showing complete coverage at 30 days, DES strut endothelialisation is reduced in humans with coronary artery disease (CAD).10,11
Classification of Stent Thrombosis
Stent thrombosis can be categorised according to the timing of occurrence into early (EST, <30 days) and late (LST, >30 days). The latter includes very late ST (VLST, >1 year). EST can be further categorised into intra-procedural (IPST, during the procedure) and subacute (SAT, after the end of the procedure to 30 days).12 In another, more conventional, classification, EST is categorised to acute (<24 hours) and SAT (between 24 hours and 30 days).
Intra-procedural Stent Thrombosis
IPST is a rare event, with the exception of specific settings such as acute MI, thrombus-containing lesions and dissections, and is relatively ‘benign’ compared with post-procedural ST (early and late). Chieffo et al. reported the occurrence of IPST in five patients (0.7%) during elective implantation of sirolimus-eluting stent (SES) in 670 patients in the absence of acute MI, thrombus-containing lesions and residual peri-stent dissections.13 None of the patients with IPST were pre-treated with glycoprotein (GP) IIb/IIIa inhibitors. Only total stent length per vessel was associated with the occurrence of IPST. It is worth noting that IPST did not occur in patients who were treated with elective GP IIb/IIIa inhibitors. This complication has not been reported in the core and pivotal DES trials.
Post-procedural Stent Thrombosis (Early and Late)
Although both the definition and mechanism for detection vary between reported series, the recorded incidence of ST in the modern era of stent deployment varies from a low 0.4% with intravascular ultrasound (IVUS) guidance to a high 2.8% after multivessel stenting. A meta-analysis of 11 randomised trials (5,013 patients) showed no evidence that the short- to medium-term safety profiles of SES or paclitaxel-eluting stents (PES) differed from those of BMS.14 In another meta-analysis of 10 randomised studies (5,030 patients), the risk of thrombosis after DES versus BMS was compared and the relationship between the rate of DES thrombosis and stent length was evaluated.15 The authors found that DES do not increase the risk of stent thrombosis (both early and late), at least under appropriate antiplatelet therapy. In another meta-analysis of eight trials (total of 13 study arms) in 3,817 patients with CAD who were randomised to either PES or BMS, the authors reported that standard-dose PES do not increase the hazard of ST compared with BMS.16
However, these trials were not powered to detect or exclude an effect of DES on rare events such as ST. In addition, the inclusion criteria for these trials were relatively restrictive, mostly excluding high-risk patients. In a small single-centre registry of 652 patients treated with SES stents, Jeremias et al. reported a rate of approximately 1% at a median follow-up of 100 days.17 Subsequently, Ong et al. reported an incidence of 0.35% (95% confidence interval (CI) 0.17–0.72) of late angiographic stent thrombosis events in an unselected DES population (2,006 patients, SES = 1,017, PES = 989) from a single-centre registry.18 In another study, Iakovou et al. reported the results from a prospective registry including a total of 2,229 consecutive patients who underwent successful SES (1,062 patients) or PES (1,167 patients) implantation.12 At nine-month follow-up, 29 patients (1.3%) had ST (nine (0.8%) with SES and 20 (1.7%) with PES; p=0.09), which is a somewhat higher incidence than that reported in the randomised studies. Fourteen patients had SAT (0.6%) and 15 patients had LST (0.7%). Of these 29 patients, 13 died (case fatality rate 45%). There were subsequent studies scrutinising the matter of ST after DES implantation, which in turn spurred growing concern over a persistent, small rise in the incidence of LST. While the incidence of ST during the first year after DES implantation is no different from the figures for BMS (approximately 0.7% for straightforward lesions and up to 2% for long lesions in vessels smaller than 2.5mm), meta-analyses of published randomised trials up to four years post-stenting for SES and PES have identified a small signal of LST beyond the first year after stent implantation, with an annual risk of around 0.2% throughout the follow-up period.19,20 More recently, at the September 2006 European Society of Cardiology (ESC) Meeting, there were two reports indicating that mortality and late thrombotic events were increased in patients receiving DES compared with BMS at three to four years post-implantation. The first study was presented by Camenzind et al., and was a meta-analysis of all available data concerning company-supported randomised, double-blind clinical trials comparing BMS controls with SES:
- Randomized Study with the Sirolimus-eluting Velocity Balloon-Expandable Stent (RAVEL); and
- SIRolImUS-coated Bx Velocity balloon-expandable stent trial (SIRIUS, E-SIRIUS and C-SIRIUS).
The BMS controls were also applied to PES (TAXUS™ II, IV, V and VI, Boston Scientific). In 870 patients treated with SES, the rate of serious adverse events (death or transmural (Q-wave) MI) was 6.3% at four years compared with 3.9% in 870 BMS-treated patients (p=0.03), a 38% increase; and in 1,685 patients treated with TAXUS stents, the rate of death or MI was 2.6 versus 2.3% in 1,675 BMS-treated patients over the same time period (p=0.68), a 16% increase.21
Moreover, the five-year follow-up of RAVEL reported a higher rate of death and MI with SES, and a recent meta-analysis of 17 randomised DES studies found that there was an association between SES implantation and increased non-cardiac mortality. These deaths were found to arise from cancer (n=15), lung disease (n=6) or stroke (n=5). The increase in cancer rate was attributed by the authors to a probable rapid impairment of the immune system. However, it is most likely that the majority of patients with cancer had stopped clopidogrel during treatment.22 At the Transcatheter Cardiovascular Therapeutics Meeting 2006 in Washington, DC, the Cardiovascular Research Foundation (CRF) presented its independent analysis of the Cypher and TAXUS stent clinical trial patient-level data, which showed a small increase in VLST with a numerical decrease in death or MI compared with BMS at four years for both stents (see Figure 1). Largely in response to the ESC reports and associated patient safety concerns, on 7 and 8 December 2006 the US Food and Drug Administration (FDA) convened a special Center for Devices and Radiological Health Expert Advisory Committee – the Circulatory System Devices Panel – to examine these concerns. At this meeting, DES manufacturers and other experts presented detailed information on DES safety. The opinions of that panel include the following general concepts:
- while the long-term follow-up has demonstrated slight numerical excess in VLST with DES compared with BMS, there is no evidence of excess death or MI compared with BMS in patients treated for ‘on-label’ indications;
- for patients treated beyond the current label indications (i.e. higher-risk patients with more complex lesions and co-morbitidies currently being treated ‘off-label’), the Panel concluded that the existing data thus far do not allow adequate determination of a risk–benefit profile, and that continued study of these patient populations is needed; and
- additionally, physicians should reinforce the importance of dual-antiplatelet therapy and consider the current AHA/ACC/SCAI 2005 PCI guidelines that recommend prolongation of dual-antiplatelet therapy beyond the current three- to six-month label, to 12 months, in patients who tolerate the regimen well.
Predictors of Stent Thrombosis after Drug-eluting Stents
Patient-, procedure- and lesion-related factors have been recognised lately as independent predictors of stent thrombosis:
- premature antiplatelet therapy discontinuation;
- renal failure;
- bifurcation lesions;
- low ejection fraction;
- acute coronary syndrome;
- stent underexpansion and malaposition; and
- stent length.12,15,17,23,24
The incidence of ST in the study by Iakovou et al. according to selected patient characteristics is shown in Figure 2.12 While bifurcational treatment was accompanied with a higher than expected incidence of ST, there were no significant differences between the one- and two-stent techniques. This finding was highlighted in two subsequently published reviews regarding the contemporary treatment of bifurcations.25,26
Stent Thrombosis and Antiplatelet Therapy
In the randomised trials, DES have mandated dual-antiplatelet therapy for two to six months, as well as the indefinite use of aspirin.14,15,27–30 In these trials, the rates of both early and late ST were found to be similar between the DES and BMS groups. The finding that in these trials and their meta-analyses DES implantation was not associated with an increased risk of LST may reflect the fact that the vast majority of patients continued aspirin indefinitely. The interruption of antiplatelet therapy was also reported to have preceded several cases of LST with DES.31,32 In the study by Iakovou et al., the most important predictor of LST was the premature cessation of antiplatelet therapy.12 Therefore, patients may not be as compliant with antiplatelet therapy in actual practice as they are in clinical trials, potentially leading to a higher number of events. In some case reports, the most common reason for stopping antiplatelet therapy was impending non-cardiac surgery. However, most noteworthy is the very late occurrence of the ST cases, ranging from 11 to 21 months after DES implantation.31,32 Thus, it is unclear when it is safe to stop antiplatelet therapy.
It must be noted that topics such as platelet and clopidogrel resistance, with new emerging concepts such as the high post-treatment platelet reactivity and incomplete P2Y12 receptor inhibition, must be taken into account when considering the impact of antiplatelet therapy on stent thrombosis.33–37 Recently, a number of reports, using various definitions, have dichotomised patients who are treated with clopidogrel into a minority of ‘non-responders’ and a majority of ‘responders’.38
Clinical trials are needed to define whether hyporesponders to clopidogrel are at increased risk of thrombotic events, and whether hyper-responders are at increased risk of bleeding. If so, the individualisation of antiplatelet therapy, including clopidogrel dosing, may be possible in the future, but will require the ability to easily and reproducibly measure responsiveness by a method that has been proved to be predictive of clinical events.
Early and late ST following the implantation of DES is rare, but is associated with serious consequences. Thus, it is important for physicians to recognise the risk of LST. The use of DES is currently the optimal way to prevent restenosis; however, DES are associated with a rare risk of VLST (after one year), with consequences such as MI and death. Discontinuation of antiplatelet therapy is the most important predictor of LST with DES. Therefore, two factors to consider before implanting a DES are:
- whether the patient will need to undergo surgery requiring the interruption of antiplatelet therapy; and
- whether the patient is likely to be compliant with antiplatelet therapy.
In situations that may lead to the interruption of antiplatelet therapy, the risk of restenosis (or need for DES) should be seriously weighed against the risk of LST. Ôûá