1

Introduction

As the population ages, the incidence of ischemic heart disease and atrial fibrillation (AF) increases. The prevalence of atrial fibrillation is noted to be 9% in those aged 80 years and above . Percutaneous coronary intervention (PCI) procedures are commonly being performed in this cohort of patients who are on oral anticoagulant therapy (OAT) for atrial fibrillation. There are a number of other indications for chronic anticoagulation, including, but not limited to, deep venous thrombosis (DVT) and pulmonary embolism (PE), primary pulmonary hypertension, hypercoaguable states and prosthetic heart valves .

In the past, antithrombotic options were limited when clopidogrel was the predominant antiplatelet agent in cases of PCI needing dual antiplatelet therapy (DAPT) and warfarin the only oral anticoagulant. Today, with the FDA approval of alternative antiplatelet (prasugrel, ticagrelor) and anticoagulant agents (dabigatran, rivaroxaban, apixiban) and the absence of larger randomized controlled studies the decision of combination antiplatelet and anticoagulant therapy is complex.

It is known that 5–10% of patients undergoing PCI have an indication for OAT . In patients with an indication for chronic anticoagulation who need PCI, one has to balance the risk of bleeding, embolic stroke and/or embolism and major adverse cardiovascular events (MACE) including stent thrombosis. In this article we review the available evidence for optimal treatment of this high-risk patient subset.

2

Estimating risk of stroke and thromboembolism

By far, the most common condition requiring long-term anticoagulation for prevention of stroke and systemic embolism is atrial fibrillation. Stroke risk can be reliably estimated using the CHADS2 score which has been used over the past decade for guiding therapy ( Table 1a ). Patients with a score of 0 have a stroke rate of 1.9% per year, 2.8% per year for score of 1, 12.5% per year for score of 5 and 18.2% per year for score of 6 . CHA2DS2-VASc2, a more recently validated and robust predictive instrument, adds three features to CHADS2, i.e. age 65–74, female sex, and vascular disease (defined as prior myocardial infarction, peripheral artery disease, or aortic plaque) Table 1b . In CHA2DS2-VASc, age ≥ 75 is given a weight of 2 points, equal to a prior stroke or transient ischemic attack. A CHA2DS2-VASc score of 0 identifies a “truly low risk” population with a thromboembolic risk of < 1%, and there is unanimous agreement that these patients do not need anticoagulation .

Under the CHADS2 schema, the recommendation is aspirin for a score of 0, aspirin or OAT for a score of 1, and OAT for a score of ≥ 2. For CHADS2VASc2, a score of 0 does not require OAT. Scores of > 2 have clinical benefit in favor of long term OAT . The magnitude of benefit in terms of risk reduction increased with anticoagulation therapy as the risk scores increase .

2

Estimating risk of stroke and thromboembolism

By far, the most common condition requiring long-term anticoagulation for prevention of stroke and systemic embolism is atrial fibrillation. Stroke risk can be reliably estimated using the CHADS2 score which has been used over the past decade for guiding therapy ( Table 1a ). Patients with a score of 0 have a stroke rate of 1.9% per year, 2.8% per year for score of 1, 12.5% per year for score of 5 and 18.2% per year for score of 6 . CHA2DS2-VASc2, a more recently validated and robust predictive instrument, adds three features to CHADS2, i.e. age 65–74, female sex, and vascular disease (defined as prior myocardial infarction, peripheral artery disease, or aortic plaque) Table 1b . In CHA2DS2-VASc, age ≥ 75 is given a weight of 2 points, equal to a prior stroke or transient ischemic attack. A CHA2DS2-VASc score of 0 identifies a “truly low risk” population with a thromboembolic risk of < 1%, and there is unanimous agreement that these patients do not need anticoagulation .

Under the CHADS2 schema, the recommendation is aspirin for a score of 0, aspirin or OAT for a score of 1, and OAT for a score of ≥ 2. For CHADS2VASc2, a score of 0 does not require OAT. Scores of > 2 have clinical benefit in favor of long term OAT . The magnitude of benefit in terms of risk reduction increased with anticoagulation therapy as the risk scores increase .

3

Anticoagulated patient requiring coronary revascularization: scope of the problem

Up to 30% of patients on warfarin for atrial fibrillation have the concomitant diagnosis of coronary artery disease (CAD) and are thus potential candidates for PCI . Additionally, approximately 5–10% of patients undergoing PCI also present with an indication for (OAT) .

Historically, for patients undergoing PCI, the primary focus has been on preventing subsequent coronary events and bleeding rather than anticoagulation for stroke prophylaxis, and consequently patients receive dual antiplatelet therapy only (DAPT) . In a recently published German study , 49% of patients in atrial fibrillation were discharged on triple therapy following acute coronary syndrome (ACS) and revascularization while in the CRUSADE registry (1648 non ST elevation myocardial infarction (NSTEMI) patients > 65 years old with atrial fibrillation) only 27% patients were discharged on triple therapy . In a recent National Cardiovascular Data Registry’s Acute Coronary Treatment and Intervention Outcomes Network Registry (NCDR) analysis evaluating over 69000 patients with acute MI (2008–2009), there were 4947 patients with atrial fibrillation; the combination of OAT plus DAPT (TOAT: Triple oral antithrombotic therapy) was prescribed in only 14.6% patients .

4

Risk of bleeding vs. risk of stroke

Whereas OAT is associated with a significant risk reduction of major stroke and thromboembolism in patients with atrial fibrillation (relative risk reduction 64%), it comes at a cost of increased bleeding . This risk is increased with adjunct antiplatelet therapy and is highest for patients on triple therapy (OAT + DAPT = TOAT). Absolute rates of stroke and bleeding are based on clinical trials conducted > 15 years ago that suggested a 2-fold increase in intra-cranial hemorrhage (ICH) with warfarin . However, contemporary studies have shown ICH risk to be similar to that seen in aspirin-treated patients with AF .

Several predictive tools have been proposed and validated for estimating the risk of bleeding, of which the HAS BLED score ( Table 2 ) has been most popular . It has been recommended by the European Society of Cardiology (ESC) to assess bleeding risk in patients with AF . Three or more of the following criteria define a high bleeding risk: hypertension, abnormal liver or renal function, history of stroke, bleeding history, labile INRs (International normalized ratio), elderly (> 65 years), anti-platelet drug or alcohol use. The HAS-BLED scores range from 0 to 9, with scores of ≥ 3 indicating high risk of bleeding, for which caution and regular review of the patient are recommended .

Stroke and bleeding risks are, in fact, closely related. Many thromboembolic risk factors have also been identified as risk factors for bleeding, and thus, risk factors for OAT-associated bleeding, are often also indications for OAT use in patients with atrial fibrillation .

In one of the largest studies to date 118,606 patients were studied. A 69.9% surviving AF hospitalization had at least 1 prescription filled for warfarin, aspirin, or clopidogrel after discharge . During a mean follow up of 3.3 years, 13,573 patients (11.4%) experienced non-fatal or fatal bleeding. The highest crude incidence rate for bleeding was observed for clopidogrel plus warfarin (13.9% per patient–year) and TOAT (15.7% per patient–year). The majority of bleeding complications occurred within the first month of institution of therapy . However, the longer the patient is on triple therapy the more the risk of bleeding. Rubboli et al calculated the pooled incidence of bleeding complications as reported in 12 studies. The number of study participants in the studies varied widely, and the pooled incidence of all bleeding at the end of 30 days on TOAT was found to be 4.6% (39/852 patients), the pooled incidence of all bleeding at the end of 6 months was 19/137 (13.9%) and the pooled incidence of bleeding at the 12 months was 10.3% (64/622 patients) .

5

Choice of stent in the anticoagulated patient

While DES reduces target vessel revascularization (TVR) risk compared to bare metal stents (BMS) in the general population, there is limited evidence of the comparative safety and efficacy in the atrial fibrillation population. Most studies were small and the number of events low, the populations were heterogeneous with respect to the type of antithrombotic regimen at discharge and it is unclear if the population groups were evenly matched with respect to bleeding and stent thrombosis risk. BMS was most often deployed in acute coronary syndromes, and DES was used in lesions at high restenosis risk .

Juan M. Ruiz-Nodar et al analyzed 604 patients with AF (atrial fibrillation) on OAT treated with PCI over a period of 7 years, 328 patients were treated with BMS (54.3%) and 276 with at least one DES (45.7%). The incidence density of adverse cardiovascular outcomes (myocardial infarction, target vessel failure) as well as mortality was similar. The secondary safety endpoints (thrombosis and embolism) were also similar between the groups. There was a higher incidence of major bleeding in the DES group (2.26 vs. 1.19 per 10 000 days of exposure, p = 0.03). The rate of definitive and probable thrombosis was similar in both DES and BMS groups (0.43 vs. 0.06/10 000 days of exposure, p = 0.09) with 60% of them occurring late or very late. The necessity of TVR was almost two fold greater in the BMS arm, albeit statistically insignificant (5.8% vs. 10.4%; p = 0.13). The rate of death at 3 years was also similar (26.8% BMS vs. 28.0% DES, p = 0.74). There was evidence of net clinical harm in the DES group, related to the use of TOAT, largely related to excess bleeding (2.26 vs. 1.19 per 10 000 days of exposure, p = 0.03) .

In a study with 127 patients all of whom were discharged on TOAT, dual antiplatelet therapy was administrated in all patients for at least 1 month after bare-metal stent (BMS) and 6 to 12 months after drug eluting stent (DES) implantation. There was a significant difference in the rate of target vessel revascularization in favor of DES (14.1% vs. 26.8%, p = 0.041) but no significant differences between DES and BMS were observed in the incidence of major bleeding (5.6% vs. 3.6%, p = 0.6), minor bleeding (1.4% vs. 3.6%, p = 0.57), and adverse cardiovascular outcomes (19.7% vs. 28.6%, p = 0.20) .

Furthermore there are consequences for bleeding that go beyond the actual bleed, and it has been shown that even minor bleeding such as bruising has been associated with clopidogrel cessation which is in turn associated with a risk of stent thrombosis and cardiovascular mortality .

Based on available evidence, routine DES use cannot be justified in most patients with atrial fibrillation and should be carefully considered for the patient with high risk anatomy such as stenting the left main, the left anterior descending artery (LAD) or in the setting of multivessel PCI in a patient at a high risk of restenosis such as diabetes or end stage kidney disease .

6

Importance of OAT in AF patients undergoing PCI

Several studies have reported that the incidence of stroke is higher with DAPT alone compared with OAT with antiplatelet/s in patients with AF who undergo PCI . In the ACTIVE W trial, 6706 patients were randomly assigned to DAPT or OAC with warfarin. Warfarin significantly lowered the annual rate of stroke, non-CNS systemic embolus, myocardial infarction, or vascular death (primary endpoint) compared to DAPT (3.9% vs. 5.6%; relative risk 0.69, 95% CI 0.57–0.85; p = 0.0003) .

Karjalainen and colleagues studied 446 patients on various therapeutic regimens following PCI, patients were broadly differentiated based on whether or not they were receiving warfarin at discharge. The outcomes in patients on warfarin were compared to those not on OAT. Patients who were on DAPT, compared to patients on TOAT, had a significantly higher rate of stroke (8.8 versus 2.8 percent) at 12 months .

7

Risk of stent thrombosis

Early randomized trials have conclusively demonstrated that DAPT is superior to the combination of aspirin and warfarin in the prevention of stent thrombosis . In the pivotal trial by Leon et al. the incidence of primary end point of all clinical events reflecting stent thrombosis was the lowest in the aspirin and ticlopidine group when compared to the aspirin with or without warfarin. (0.5% vs. 2.7% vs. 3.6% respectively, p = 0.001 for the comparison of all three groups) . Furthermore, premature DAPT discontinuation has been the most consistent and powerful predictor of stent thrombosis, associated with a 5- to 36-fold risk increase. Stent thrombosis is associated with considerable morbidity and a high case fatality rate . Duration of DAPT following BMS placement is 1 month and 1 year after DES placement according to the American College of Cardiology (ACC)/American Heart association (AHA) guidelines . There has been growing evidence to suggest that this duration can be limited to 6 months with the use of second generation DES .

Second generation and more recently, third generation DES have been widely adopted as the stent of choice given their lower TVR rate of < 5%. DES are commonly being used for the management of CAD in a wide variety of situations including but not limited to ST-elevation myocardial infarction (STEMI), non ST segment elevation myocardial infarction (NSTEMI), left main disease, chronic total occlusions, long lesions, bifurcations lesions, in stent restenosis, in patients with diabetes and multivessel disease, indeed one study found that the off label use of DES is as high as 60% in western countries . There is some evidence of limiting dual antiplatelet therapy (DAPT) duration to six months with the second generation stents although the current American Heart association (AHA) recommendation remains at 12 months . Patients with stable CAD or low risk acute coronary syndromes who underwent PCI with Endeavour Zotarolimus Stents were randomized to DAPT with aspirin and clopidogrel for 3 or 12 months in the OPTIMIZE trial. There was no difference in the rate of the primary composite end point of net adverse clinical and cerebral events (death (any cause), myocardial infarction, cerebral vascular accident, and major bleeding at 12-month clinical follow-up post-index procedure). In the 3 and 12 month groups the rates of major bleeding and stent thrombosis were 0.2 and 0.5 percent (hazard ratio 0.5, 95% CI 0.12–1.99) and 0.3 and 0.1 percent, (hazard ratio 3.97, 95% CI 0.44–35.49) respectively .

There was no major difference in the combined end point of cardiac death, myocardial infarction (MI), or ischemia-driven revascularization (primary outcome) between the 6 and the 12 month groups in the EXCELLENT trial. The rates of primary outcome were 4.8% in the 6-month DAPT group and 4.3% in the 12-month DAPT group. Stent thrombosis (a secondary end point) was more common in the six-month group (0.9 versus 0.1 percent, p = 0.10) . The rates of stent thrombosis were 0.2 and 0.3 percent, respectively in the 3 and 12 month groups in the RESET trial . The current AHA recommendation of DAPT for 12 months likely errs on the safe side in view of the dire consequences of stent thrombosis, which may cause a clinical myocardial infarction (MI) in up to 70% of patients and death in up to 20% depending on the stent and other circumstances .

Results of ongoing randomized trials such as Randomized, Open Label Trial of 6 Months Versus 12 Months DAPT After Drug-Eluting Stent in STEMI (DAPT-STEMI) and the NIPPON trial comparing 6 and 18 months of DAPT after Nobori stent implantation will help clarify the issue further ( https://clinicaltrials.gov/ ).