Abstract
Oral anticoagulation (OAC), usually with vitamin K antagonists (such as warfarin, target international normalized ratio [INR] 2–3, or higher with some prosthetic valves) is the proven therapy and recommended treatment for a wide range of patients at risk of thromboembolic events, including atrial fibrillation (AF), mechanical heart valves, deep vein thrombosis, pulmonary embolism and left ventricular thrombi. These conditions are common, and with an increasing average life expectancy, some of these conditions (especially AF) are increasing in prevalence [Lip and Lim, 2007; Kannel et al. 1998]. Furthermore, pathologies such as AF are commonly associated with coronary artery disease: approximately 20–30% of AF patients have concomitant coronary artery disease [Nieuwlaat et al. 2005]. Such patients may present with acute coronary syndrome (ACS) and/or undergo percutaneous coronary intervention (PCI), often with stenting (PCI-S).
In contrast to the pathogenesis of thrombosis in AF, where the clot is mainly fibrin-rich (‘red clot’), the thrombus in ACS (and stent thrombosis) is largely platelet-rich (‘white clot’). Thus, following PCI-S, there is the consequent need for dual antiplatelet therapy (aspirin and a thienopydrine inhibitor, commonly clopidogrel), with the aim of minimizing both early and late stent thrombosis. The duration of dual antiplatelet therapy is also governed by the nature of the stent, i.e. bare metal stents (BMSs) versus drug eluting stents (DESs), and to an extent, whether a second-generation or third-generation DES is used [Rubboli et al. 2008; Rubboli and Verheugt, 2008]. An ACS presentation would also require a longer duration of antiplatelet therapy compared with an elective procedure.
This raises an important management issue in patients who require chronic OAC, who need combination aspirin–clopidogrel therapy following an ACS or PCI-S. Dual antiplatelet therapy is inferior to OAC in reducing thromboembolic strokes in AF patients [Connolly et al. 2006]. Ultimately, there is the need to balance the prevention of thromboembolism, and recurrent cardiac ischaemia post-ACS and/or stent thrombosis on the one hand, and the risk of bleeding on the other [Rubboli and Halperin, 2008; Roldán and Marín, 2008]. The major bleeding sources on such therapy will be haemorrhagic events either from the gastrointestinal tract or intracranially.
Given that conditions such as AF are a growing epidemic, and that new management guidelines place greater emphasis on OAC [Camm et al. 2010], an increasing number of patients on OAC would need to be treated with dual antiplatelet therapy as well, i.e. triple therapy. Despite the growing size of this patient population, there is still a paucity of randomized double-blinded controlled trials that adequately address this clinical dilemma, with conflicting results from the small cohort studies already available. Thus, there is the need to rely on systematic literature reviews and meta-analyses as well as expert consensus recommendations [Lip et al. 2010] to provide a guide to the best practice available.
In the current issue of Therapeutic Advances in Cardiovascular Disease, the meta-analysis by Singh and colleagues addresses this problematic therapeutic question, of the safety and efficacy of triple antithrombotic therapy after percutaneous coronary intervention in patients needing long-term anticoagulation [Singh et al. 2010]. In this paper, they analysed all of the studies where patients had triple therapy with OAC and dual antiplatelet therapy. Not only cardiovascular outcomes were investigated, but bleeding complications (major/life threatening) too, either as a primary or secondary endpoint. It is apparent from this work that relatively few studies actually included bleeding as an endpoint. The meta-analysis concluded that triple therapy is beneficial in reducing the incidence of ischaemic cardiovascular events but was associated with higher incidence of major bleeding. Elsewhere in the literature, the most common adverse event in the AF patient population taking triple therapy was major bleeding (accounting for 12.3%, from a total of 36.6% adverse events [Ruiz-Nodar et al. 2008]. This meta-analysis is therefore complementary to other overviews of the recent evidence in this field [Lip et al. 2010; Rubboli et al. 2008; Rubboli and Verheught, 2008; Rubboli and Di Pasquale, 2007]. A 12.0% yearly incidence of bleeding for triple therapy patients was also demonstrated in the Danish registry of patients following myocardial infarction [Sørensen et al. 2009], where triple therapy posed a threefold increase in risk of bleeding in the subjects with AF [Hansen et al. 2010].
However, the meta-analysis by Singh and colleagues does not analyse whether the risk of bleeding is early, late or indeed constant during the duration of triple therapy [Singh et al. 2010]. Any patient presenting with ACS and/or undergoing PCI-S will have a periprocedural bleeding risk, given the use of various antithrombotic agents and/or interventions. The primary question is whether triple therapy confers a markedly excess risk, over the background bleeding risk, which, after all, is multifactorial. Indeed, major bleeding peri-PCI is more common in ‘high-risk’ subjects, with the potential for more adverse outcomes at follow up [Eikelboom et al. 2006].
If the bleeding risk is mainly early, then choice of concurrent medication peri-PCI (for example, bivalirudin) and glycoprotein IIb/IIIa inhibitors (GPIs) as well as access site choice, play a major role in reducing bleeding complications [Lip et al. 2010; Rubboli et al. 2010; Karjalainen et al. 2008]. Another analysis has indicated a 30-day incidence of major bleeding of 2.2% (95% CI of 0.7–3.7%) with the use of triple antithrombotic therapy in patients with AF and coronary artery stents [Paikin et al. 2010]. However, if bleeding is mainly late or constant, then avoiding triple therapy or reducing the duration of its use is perhaps the key.
Current recommendations attempt to reduce the effect of peri-PCI events/medication on the future risk of bleeding. For example, radial access is being encouraged as well as judicious use of GPIs, bivalirudin and/or low molecular weight heparin (LMWH) depending on patient characteristics. Indeed, triple therapy is an independent predictor of late major bleeding, whilst GPI being independent predictors of early major bleeding [Manzano-Fernández et al. 2008]. On the other hand, it is estimated that the prevalence of major bleeding in triple therapy patients is 2.6–4.6% at 30 days, which increases to 7.4–10.3% at 12 months, indicating that the risk of bleeding increases with a prolonged duration of triple therapy [Camm et al. 2010]. Similarly, an analysis of the patients enrolled in the OASIS Registry, OASIS-2 and SURE studies indicated that there is a fivefold increase in death in high-risk patients who bleed in the first 30 days post PCI-S. In this context, the results and observations of the WAR-STENT registry are awaited [Rubboli et al. 2009].
From a different prospective, several studies are currently being conducted to investigate the benefit of shorter dual antiplatelet therapy post PCI-S, e.g. the OPTIMIZE trial [ClinicalTrials.gov Identifier: NCT01113372]. The ISAR-TRIPLE trial addresses the question of shorter duration of clopidogrel therapy (6 weeks) with long-term aspirin and OAC [ClinicalTrials.gov Identifier: NCT00776633]. The WOEST trial is investigating the efficacy of clopidogrel and oral anticoagulant therapy versus triple therapy [Dewilde and Berg, 2009].
A number of scores have been used to assess risk of bleeding with antithrombotic therapy. The HAS-BLED score reflects these comorbidities (and incorporates the patient’s age [>65], renal and liver function, previous history of hypertension, stroke and bleeding, as well as concomitant drugs/alcohol intake and labile INR control), and is now favoured as a simple, practical schema for bleeding risk stratification, where a score of ≥3 indicates a bleeding risk that merits caution and/or regular review [Pisters et al. 2010]. Even those patients with a low CHADS2 score are still at a high risk of thromboembolic events and thus, triple therapy should even be considered in this patient population [Ruiz-Nodar et al. 2010].
What do the guidelines say? In patients with high bleeding risk, the duration of dual antiplatelet therapy should be minimized by avoiding DESs or limiting their use to the appropriate lesion size and coronary anatomy (e.g. small size vessels, or long segments of disease in diabetics, etc). The role of balloon angioplasty and coronary artery bypass grafting (CABG) should not be forgotten [ClinicalTrials.gov Identifier: NCT01113372; Holmes et al. 2009]. However, when triple therapy is unavoidable then tight control of the INR (e.g. in the range 2–2.5) and high time in therapeutic range, has been shown to be associated with lower risks of bleeding, with an aim is restricting triple therapy to 4–6 weeks [Singh et al. 2010]. The safety of other new OAC therapies in this clinical setting will become an important part of day-to-day practice in the future [Holmes et al. 2009].
Recommended antithrombotic strategies following coronary artery stenting in patients with atrial fibrillation at moderate-to-high thromboembolic risk (in whom oral anticoagulation therapy is required). (Adapted from Camm et al. [2010] and Lip et al. [2010]).
ACS, acute coronary syndrome; AF, atrial fibrillation; INR, international normalized ratio; VKA, vitamin K antagonist.
Gastric protection with a proton pump inhibitor (PPI) should be considered where necessary.
Sirolimus, everolimus and tacrolimus.
Combination of VKA (INR 2.0–3.0) + aspirin ≤100 mg/day (with PPI, if indicated) may be considered as an alternative.
Drug-eluting stents should be avoided as far as possible, but, if used, consideration of more prolonged (3–6 months) triple antithrombotic therapy is necessary.