Residual risk for secondary ischemic events in patients with atherothrombotic disease: Opportunity for future improvements in patient care

Abstract

Atherothrombotic disease is highly prevalent in Western countries and is associated with morbidity, mortality, and a significant economic burden. The primary pathophysiological mechanism of acute ischemic events in patients with atherothrombotic disease is complex but involves thrombotic occlusion in response to rupture or erosion of atherosclerotic lesions. Current treatments for long-term secondary prevention in patients with established atherothrombotic disease, such as those with prior myocardial infarction, ischemic stroke/transient ischemic attack, or symptomatic peripheral artery disease, include therapies aimed at preventing rupture/erosion of atherosclerotic lesions (life-style modification and blood pressure reduction, in addition to statins and angiotensin II-active agents) and thrombus formation (primarily antiplatelet agents, such as aspirin, thienopyridines (clopidogrel, prasugrel, ticlopidine), and, to a lesser degree, anticoagulants). Despite the proven benefits and broad use of these therapies, the long-term rates of mortality and recurrent ischemic events remain high. This residual risk can be attributed to the fact that atherothrombosis continues in the presence of current treatments; because these agents each inhibit relatively specific pathways, atherosclerosis, thrombus formation, and other processes may progress. These considerations suggest that novel therapies with a different mechanism of action may provide additional reductions in morbidity and mortality beyond those observed with current agents.

Key words::

• Antiplatelets
• antithrombotics
• atherothrombosis
• ischemic stroke
• myocardial ischemia
• peripheral artery disease
• residual risk
• statin

Key messages

• Despite the clinical benefit achieved with life-style modifications, blood pressure reduction, statins, angiotensin II-active agents, and antiplatelet agents, the residual risk for recurrent acute events in patients with established atherothrombotic disease remains substantial. Observational registry data confirm the high residual risk of patients with documented atherothrombotic disease and highlight the need for more effective therapies.

• Residual risk for thrombotic events is likely related to progression of atherosclerosis in the presence of statins and angiotensin II-active agents, lack of more complete inhibition of platelet activation in the presence of aspirin and clopidogrel, and other factors yet to be identified.

• Current antiplatelet agents target thromboxane A₂ and adenosine diphosphate and do not interfere with a multitude of other platelet activation pathways that may contribute to thrombotic events. These agents are also associated with bleeding risk.

• These considerations underscore a critical need for novel therapies that provide more comprehensive platelet inhibition when used with current oral antiplatelet therapy for greater protection against thrombotic events, preferably without an incremental bleeding risk.

Abbreviations
ACAT	=	acyl-coenzyme A:cholesterol acyltransferase
ACC	=	American College of Cardiology
ACE	=	angiotensin-converting enzyme
ACS	=	acute coronary syndromes
ADP	=	adenosine diphosphate
AHA	=	American Heart Association
ARB	=	angiotensin receptor blocker
BP	=	blood pressure
CAC	=	coronary artery calcification
CAD	=	coronary artery disease
CETP	=	cholesteryl ester transfer protein
CHD	=	coronary heart disease
CRP	=	C-reactive protein
CV	=	cardiovascular
DTI	=	direct thrombin inhibitor
ESRD	=	end-stage renal disease
FXa	=	factor Xa
GP	=	glycoprotein
HF	=	heart failure
HPS	=	Heart Protection Study
INR	=	International normalized ratio
LDL-C	=	low-density lipoprotein cholesterol
Lp-PLA2	=	lipoprotein-associated PLA2
LV	=	left ventricular
LVH	=	left ventricular hypertrophy
MI	=	myocardial infarction
PAD	=	peripheral artery disease
PLA2	=	phospholipase A2
PWV	=	pulse wave velocity
SMCs	=	smooth muscle cells
sPLA2	=	soluble PLA2
TIA	=	transient ischemic attack
TLRs	=	Toll-like receptors
TxA2	=	thromboxane A2

Introduction

Atherothrombotic disease is the leading cause of morbidity and mortality in the United States (1) and the Western world. For patients with atherothrombotic disease, such as those with a history of myocardial infarction (MI), ischemic stroke/transient ischemic attack (TIA), or peripheral artery disease (PAD), several pharmacological options are available, in addition to long-term life-style modification and blood pressure (BP) control. These include statins and angiotensin II-active agents, which reduce the likelihood of atherosclerotic plaque progression and rupture/erosion to prevent events, and antithrombotic agents, which inhibit platelet activation that may lead to occlusive thrombus-related events. While chronic life-style modification and BP control in conjunction with the use of these agents have been associated with a significant reduction in adverse events, the long-term residual risk for recurrent ischemia-related events remains substantial (1–6). The purpose of this review is to summarize recent epidemiological data for atherothrombotic disease risk, the pathophysiology of atherothrombosis, and the available pharmacological treatment options (including results of key clinical trials), which are useful when added to life-style modification and BP reduction for secondary prevention of atherothrombotic disease. This information should provide a useful framework for discussion of novel therapies aimed at reducing the residual risk.

Morbidity, mortality, and cost of atherothrombotic disease in the United States

Atherothrombotic disease is highly prevalent and is associated with significant morbidity, mortality, and economic consequences. The estimated prevalence of MI in the United States is 8,100,000 (7,8), and the annual incidence of MI in 2008 is estimated at 920,000 (600,000 new and 320,000 recurrent cases) (8). In 2004, over 150,000 Americans died of MI (8). Among individuals aged ≥40 years with a first MI, 33% of men and 43% of women are estimated to die within 5 years. MI is also associated with significant morbidity: among individuals aged 40–69 years with a first MI, an estimated 4% of men and 6% of women will develop a stroke, and 7% of men and 12% of women will develop heart failure (HF).

The estimated prevalence of stroke in the United States is 5,800,000 (7–9), and the annual incidence of stroke in 2008 is estimated at 780,000 (600,000 new and 180,000 recurrent cases) (7–9). Ischemic stroke accounts for 87% of all strokes (8). In 2004, over 150,000 Americans died of stroke, and about one half of men and women with a first stroke will die within 5 years. Furthermore, 15%–30% of stroke survivors are permanently disabled (8). Approximately 13% of men and 22% of women between 40 and 69 years of age who have experienced stroke will have a recurrent stroke. Approximately 15% of all strokes are preceded by a TIA, and almost one half of those with TIAs will have stroke, MI, or vascular death in the next 10 years (8).

The estimated prevalence of PAD in the United States is 8,000,000 (10,11), and the annual mortality rate for patients with lower-extremity PAD estimated from epidemiological studies is 4%–6% (12–16). A prospective study estimated that patients with large-vessel PAD are at 3-fold higher risk of all-cause mortality, a >6-fold higher risk for fatal coronary artery disease (CAD), and a 6-fold higher risk for cardiovascular (CV) death (12). These patients have a 40% higher risk of stroke compared with those without PAD (17,18). In addition, patients with either asymptomatic or symptomatic PAD have impaired quality of life due to a progressive loss of lower-extremity function over time (13,19).

The American Heart Association (AHA) estimated direct and indirect costs of all cardiovascular disease (CVD) for 2008 to be $448.5 billion (8). Of the total cost, the estimated direct and indirect cost of coronary heart disease (CHD) is $156.4 billion, and the estimated direct and indirect cost of stroke is $65.5 billion. These costs are anticipated to rise as our population ages and the epidemics of obesity, metabolic syndrome, and diabetes continue.

Pathobiology of atherothrombosis and rationale for medical therapy

Atherosclerotic plaque development

Atherosclerosis develops within the intima of large-and medium-sized arteries and can be triggered by behavioral, environmental, biochemical, or genetic factors (20). Studies in animal models and humans document that various factors are associated with dysfunction of endothelial cells, vascular smooth muscle cells (SMCs), and their precursor cells, which are important for understanding the limitations of vascular repair. These functional alterations lead to initial structural lesions (fatty streaks) and then raised lesions (plaques) that are relatively small and do not result in significant narrowing of the blood vessel lumen (20,21). The disease begins very early in life, as fatty streaks have been identified in the aortae of aborted fetuses and children <2 years old dying of accidental causes (22–24). These stable, asymptomatic, and often undetectable plaques may progress to unstable, high-risk, disruption-prone lesions as a result of multiple local and systemic factors (20). A developing non-occlusive but unstable plaque is composed of inflammatory cells, including leukocytes, SMCs, platelet aggregates, inflammatory cytokines, growth factors, and lipid-laden foam cells (20,25). An expanding lipid core of unstable lesions is separated from the arterial lumen by a fibrous cap, which is composed primarily of SMCs and collagen (20,21,25). Rupture or erosion of these lesions exposes circulating blood to a highly thrombogenic environment, ultimately leading to occlusion of the lumen by platelet-rich thrombi. Formation of platelet-rich thrombi in response to rupture or erosion of atherosclerotic plaques leads to partial or complete obstruction of blood flow in the affected arteries, resulting in insufficient oxygen supply (ischemia) that manifests clinically as an MI, unstable angina, stroke/TIA, or symptomatic PAD (20,26).

Thrombosis

Platelets play a critical role in atherothrombotic disease, as they are the primary constituent of occlusive thrombi at the sites of ruptured or eroded plaques. Collagen and von Willebrand factor in the extracellular matrix of plaques facilitate initial platelet recruitment (27,28). Soluble mediators (e.g. thromboxane A₂ (TxA₂) and adenosine diphosphate (ADP)) released from the initial adherent platelets in response to contact with collagen promote additional platelet recruitment (29). Thrombin, generated by tissue factor in ruptured or eroded plaques, facilitates formation of fibrin-rich clots that stabilize the protective platelet monolayer. Adherent platelets stabilized by fibrin-rich clots help prevent blood loss in response to vascular trauma and are also known as hemostatic plugs; under pathological conditions, however, multiple factors released from adherent platelets and/or generated at sites of plaque rupture or erosion stimulate subsequent platelet activation and aggregation (20,21,25). Platelets can be activated by multiple factors, including thrombin, TxA₂, ADP, collagen, serotonin, and others, which collectively amplify and sustain platelet activation (25). A key feature of platelet activation is conversion of glycoprotein (GP) IIb/IIIa receptors on the platelet surface into an active form that binds fibrinogen and leads to platelet aggregation via cross-linking of GP IIb/IIIa receptors on adjacent platelets by fibrinogen bridges (30). Multiple platelet aggregation reactions ultimately result in generation of occlusive platelet-rich thrombi that are likely responsible for many of the acute clinical manifestations of atherothrombotic disease (e.g. ST-elevation MI, non-ST-elevation MI, unstable angina, and ischemic stroke/TIA). Fibrin-rich clots generated by thrombin may also contribute to occlusive thrombosis.

Inflammation

Platelets interact with the vascular endothelium and link inflammation, thrombosis, and atherogenesis (31,32). Platelet participation in inflammation contributes to the development of atherosclerotic lesions and atherothrombosis, which is characterized by interactions between platelets, endothelial cells, and leukocytes. For example, activated platelets release CD40 ligand, promoting endothelial inflammation via production or recruitment of reactive oxygen species, adhesion molecules, chemokines, and tissue factor (31). In addition, secretion of reactive oxygen species by activated platelets enhances platelet recruitment to a growing thrombus (33 and induces enhanced lipid peroxidation of cell membrane phospholipids and circulating low-density lipoprotein (LDL). Furthermore, platelets express multiple Toll-like receptors (TLRs), which are mediators of innate immune response (34). TLR2 and TLR4 are involved in the pathogenesis of CAD, establishing a link between the progression of coronary atherosclerosis and the immune response to endogenously generated inflammatory ligands (35).

Taking into account the complex pathobiology of atherothrombotic disease, the goals of pharmacological therapy for prevention of acute ischemic events are to: 1) slow progression of atherosclerosis (reduce development of new lesions vulnerable to rupture or erosion) with life-style modification and statins (and possibly angiotensin II-active agents); 2) reduce the risk of rupture or erosion of existing atherosclerotic lesions (with a statin or with angiotensin II-active agents); and 3) inhibit thrombotic occlusion, primarily with antiplatelet therapy (e.g. aspirin and ADP receptor antagonists) and, to a lesser degree, with anticoagulants (21,25). Inflammation may represent a pathophysiological mechanism for thrombosis that may furnish novel therapeutic approaches.

Secondary prevention of atherothrombotic events: evidence from clinical trials

Life-style modification

The large international INTERHEART study (36 identified nine risk factors that collectively accounted for >90% of acute MI risk. Risk predictors included life-style factors (such as smoking, low fruit and vegetable consumption, low exercise levels, lack of alcohol consumption), comorbidities (hypertension, diabetes, abdominal obesity, abnormal lipid profiles), as well as psycho-social factors (36).

Multiple randomized trials have shown that life-style modifications can result in weight loss, improved fitness, and significant reduction in BP, each of which is important in reducing risk for atherothrombotic disease. The combination of group life-style modification counseling and pharmacotherapy resulted in approximately double the average weight loss of life-style modification or pharmacotherapy alone (6). The benefits of a Mediterranean-style diet (defined as a diet rich in fruits, vegetables, bread, other forms of cereal, potatoes, beans, nuts, and seeds, with olive oil as an important fat source; low-to-moderate consumption of dairy products, fish, and poultry; little red meat; and low-to-moderate wine consumption) have also been well documented (4). Available evidence from these studies supports life-styles that adopt strategies to lose weight, stop cigarette smoking, engage in regular moderate exercise and relaxation, and regularly consume light-to-moderate alcohol and (to a lesser extent) fatty fish in order to reduce platelet reactivity and coagulability, and to promote fibrinolysis. The overall effects can be expected to translate into an improved CV prognosis or other beneficial clinical outcome in healthy individuals and those with CV risk factors or established CHD (5). Nevertheless, a high residual risk for secondary ischemic events, despite life-style modifications, is evident from the results of recent clinical practice registries, which justifies the need for adding pharmacotherapy.

Lipid-lowering therapy

Multiple trials have documented the benefits of statins for secondary prevention of CHD (Table I) (37–48). Early trials with simvastatin (37) and pravastatin (38,47) demonstrated 25% to 35% reductions in low-density lipoprotein cholesterol (LDL-C), which were accompanied by reductions in morbidity and mortality (Table I).

Table I. Primary outcomes in selected randomized secondary prevention trials with statins.

Trial name	Treatment groups (mean duration of treatment)	Study population (n)	Primary outcome	Primary outcome rate (relative risk reduction, %; P-value)
4S (37)	Placebo versus Simvastatin 20 mg qd; (5.4 y)	Acute MI or angina pectoris (n=4,444)	Total mortality	Placebo 11.5%, Simvastatin 8.2% (RRR 30%; P=0.0003)
CARE (47)	Placebo versus Pravastatin 40 mg qd; (5 y)	MI (n=4,159)	CAD death or non-fatal MI	Placebo 13.2%, Pravastatin 10.2% (RRR 24%; P=0.003)
LIPID (38)	Placebo versus Pravastatin 40 mg qd; (6.1 y)	MI/UA pectoris (n=9,014)	CAD death or non-fatal MI	Placebo 8.3%, Pravastatin 6.4% (RRR 24%; P<0.001)
PROVE-IT (40)	Atorvastatin 80 mg qd versus Pravastatin 40 mg qd; (2 y)	ACS (MI with or without ST-segment or UA) (n=4,162)	Composite death from any cause, MI, UA requiring hospitalization, revascularization, stroke	Atorvastatin 22.4%, Pravastatin 26.3% (RRR 16%; P=0.005)
TNT (44,48)	Atorvastatin 80 mg qd versus Atorvastatin 10 mg qd; (4.9 y)	Stable CHD (n=10,001)	CHD death, non-fatal MI, resuscitation after cardiac arrest, fatal or non-fatal stroke	Atorvastatin 80 mg 8.7%, Atorvastatin 10 mg 10.9% (RRR 22%; P<0.001)
A-Z (41)	Placebo 4 mo followed by Simvastatin 20 mg qd versus Simvastatin 40 mg qd for 1 mo followed by 80 mg qd; (2 y)	ACS (n=4,500)	CV death, non-fatal MI, readmission for ACS, stroke	Placebo + Simvastatin 20 mg qd 16.7%, Simvastatin 40 mg qd/80 mg qd 14.4% (P=NS)
IDEAL (45)	Atorvastatin 80 mg qd versus Simvastatin 20 mg qd; (4.8 y)	MI (n=8,888)	Coronary death, non-fatal confirmed acute MI, cardiac arrest with resuscitation	Atorvastatin 80 mg 10.4%, Simvastatin 20 mg 9.3% (P=NS)
SPARCL (39)	Placebo versus Atorvastatin 80 mg qd; (4.9 y)	Stroke or TIA (n=4,731)	Fatal or non-fatal stroke	Placebo 13.1%, Atorvastatin 11.2% (RRR 16%; P=0.03)
Primary outcomes in other randomized trials with statins:
CORONA (43)	Rosuvastatin 10 mg versus Placebo	Ischemic systolic HF; aged ≥ 60 y (n=5,011)	CV death, non-fatal MI, or non-fatal stroke	Rosuvastatin 11.4%, Placebo 12.3% (P=NS)
GISSI-HF (42)	Rosuvastatin 10 mg versus Placebo	Chronic HF; aged ≥ 18 y (n=4,574)	Time to death	Rosuvastatin 29%, Placebo 28% (P=NS)
JUPITER (46)^a	Rosuvastatin 10 mg versus Placebo	LDL-C < 130 mg/dL; CRP ≥ 2 mg/L (n=17,802)	MI, stroke, arterial revascularization, hospitalization for UA, or CV death	Rosuvastatin 0.77^b Placebo 1.36^b (P<0.00001)

^aPrimary prevention trial.

^bRates per 100 person-years.

ACS = acute coronary syndromes; CAD = coronary artery disease; CHD = coronary heart disease; CRP = C-reactive protein; CV = cardiovascular; HF = heart failure; LDL-C = low-density lipoprotein cholesterol; MI = myocardial infarction; NS = not significant; qd = once daily; TIA = transient ischemic attack; UA = unstable angina.

Recent trials have assessed the benefit of more aggressive lipid-lowering for secondary prevention in patients with CHD (Table I). The PROVE-IT trial (40) demonstrated a 16% reduction in the primary composite outcome with atorvastatin 80 mg versus pravastatin 40 mg once daily (qd) in patients with acute coronary syndromes (ACS) (Table I). The TNT trial (44,48) demonstrated a direct correlation between a decrease in LDL-C and a reduction in secondary events in subjects with stable CHD (Table I). Phase Z of the A-Z trial (41) showed no significant difference between an early and intensive simvastatin regimen versus a delayed and less intensive simvastatin regimen in patients with ACS, possibly due to low event rates and high rates of non-compliance (Table I). The IDEAL trial (45) compared low-dose simvastatin versus high-dose atorvastatin in patients with prior MI. After 4 years, intensive LDL-C lowering did not significantly reduce CV or all-cause mortality (Table I), potentially due to a smaller than expected difference in LDL-C levels between the groups over the study duration (25% decrease in LDL-C versus ≈35% in earlier studies), a shorter follow-up duration than anticipated, and an increase in high-density lipoprotein cholesterol levels with simvastatin (which may compensate for LDL-C reductions with atorvastatin).

The benefits of long-term statin therapy in secondary prevention were also demonstrated by the UK Heart Protection Study (HPS), which included patients with prior MI, cerebrovascular disease, or PAD (3). In addition to demonstrating a reduction in the primary outcome of all-cause mortality (Figure 1A), simvastatin also reduced rates of first major vascular event in the total population, as well as in patients with prior MI, cerebrovascular disease, or peripheral vascular disease (Figure 1B). Simvastatin also reduced the relative risk of total stroke and ischemic stroke by 25% and 30%, respectively (3).

Figure 1. Event rates in patient subgroups allocated simvastatin versus placebo in the Heart Protection Study (HPS) (3). A: The rate of the primary end-point of all-cause mortality was significantly lower in patients allocated simvastatin (12.9%) versus placebo (14.7%) in the total study population in the HPS (12% relative risk reduction, 88% residual risk; P=0.003). B: The rate of the first major vascular event was significantly lower in patients receiving simvastatin (19.8%) than in those receiving placebo (25.2%) in the total study population (21% relative risk reduction, 79% residual risk; P<0.001), as well as in patients with prior myocardial infarction (MI) (23.5% versus 29.4%; 20% relative risk reduction, 80% residual risk; P<0.001), patients with cerebrovascular disease (18.7% versus 23.6%; 21% relative risk reduction, 79% residual risk; P<0.001), and patients with peripheral vascular disease (PVD) (24.7% versus 30.5%; 19% relative risk reduction, 81% residual risk; P<0.001). Green arrows indicate relative risk reduction between treatment arms. Red arrows indicate residual risk for recurrent events.

Beyond secondary prevention, large observational studies and small prospective trials of statins have suggested improved cardiac function and decreased cardiac death in patients with HF (49–51). However, large prospective randomized trials have not confirmed these observations. Rosuvastatin did not demonstrate a reduction in mortality in patients with ischemic HF in CORONA or in patients with chronic HF in GISSI-HF (Table I) (42,43). These results suggest that once HF is established, statins may not be able to improve the underlying disease. Any potential benefit of statins may be offset by any detrimental effects of lowering LDL-C in this patient population.

Statins are recommended for patients with ischemic stroke (9), although there is somewhat less evidence to support their benefits for secondary prevention of stroke. While simvastatin reduced the risk of stroke and TIA in the HPS (3), it did not reduce the incidence of stroke in patients with pre-existing cerebrovascular disease (3). The SPARCL trial (39), however, demonstrated a benefit for high-dose atorvastatin (Table I). Importantly, a small increase in hemorrhagic strokes was observed with atorvastatin. A considerably delayed initiation of statin therapy in the HPS (patients were enrolled 4.3 years after the index event) compared with immediate statin initiation in SPARCL may be one potential reason for the difference in study outcomes. In addition to timing, the use of different statins may also have contributed to the observed differences in the outcomes between the two trials.

Few studies have examined the benefits of statins in patients with PAD. The HPS (3) indicated that simvastatin reduced the relative risk of secondary events in the subgroup of patients with PAD by 25%. Statins are recommended for all patients with PAD to achieve target LDL-C levels of <100 mg/dL, whereas in patients with lower-extremity PAD who are at high risk for thrombotic events the LDL-C goal is <70 mg/dL (13).

The beneficial effects of statins on atherothrombotic disease have been confirmed by several recent meta-analyses, including those in patients with ACS (52), stroke (53), and PAD (54). The JUPITER trial suggests that statins may also have clinical benefit in lower-risk individuals, as mentioned above.

Angiotensin II-active agents

Early trials of angiotensin-converting enzyme (ACE) inhibitors in patients with HF suggested a reduction in non-fatal MIs (55), and subsequent studies evaluated their efficacy in secondary and primary prevention among patients without HF or left ventricular (LV) dysfunction. The ACE inhibitor quinapril did not demonstrate a benefit in reduction of ischemic events in patients with ischemic heart disease and preserved LV function in the QUIET trial (56). In contrast, the HOPE trial demonstrated a 22% relative risk reduction in CV death, MI, or stroke with ramipril 10 mg over 5 years (P<0.001). The beneficial effect of ramipril was evident in both the secondary and primary prevention cohorts (57). Similarly, in the EUROPA trial, treatment with perindopril was associated with a 20% relative risk reduction in CV death, MI, or cardiac arrest (P=0.0003) during the 4.2-year study period, with consistent benefits among patients with or without prior MI or revascularization (58). However, the PEACE trial did not demonstrate a beneficial effect of trandolapril, possibly due to a lower-risk profile of the study population, a high rate of coronary revascularizations prior to enrollment, and more intensive management of risk factors (59). A meta-analysis of QUIET, HOPE, EUROPA, and PEACE (pooled n=31,555) found that ACE inhibition significantly reduced the risk of all-cause mortality, CV mortality, acute MI, and stroke (P<0.001 for each), supporting the benefit of ACE inhibition for secondary prevention in patients with preserved LV function (60).

In TRANSCEND, the angiotensin receptor blocker (ARB) telmisartan did not reduce CV death, MI, stroke, or hospitalizations in patients with CVD or diabetes with end-stage organ damage who were intolerant to ACE inhibitors (61). A modest reduction in the secondary outcome of CV death, MI, or stroke was observed. The ONTARGET trial demonstrated that the combination of the ACE inhibitor ramipril and the ARB telmisartan provided no incremental benefit over either therapy alone in patients with vascular disease or high-risk diabetes, while resulting in a higher rate of adverse events, particularly renal dysfunction (62). Among patients with ischemic stroke, telmisartan initiated soon after ischemic stroke and continued for 2.5 years did not significantly lower the rate of recurrent stroke or major CV events in the PROFESS trial (63). These findings are in contrast to the PROGRESS trial, which demonstrated reduced risk of recurrent stroke and other CV events in patients randomized to receive active treatment with the ACE inhibitor perindopril plus the diuretic indapamide (64). (Flexibility regarding the use of combination therapy or perindopril monotherapy was allowed in patients allocated to receive active treatment for whom the treating physician was concerned about intensive blood pressure lowering.)

Collectively, the results of these trials suggest the potential benefit of angiotensin II-active agents may be attenuated in patients who are treated with other evidence-based therapies because intensive pharmacological management may reduce residual risk for recurrent ischemic events. Current AHA/American College of Cardiology (ACC) guide-lines for secondary prevention recommend ACE inhibitors for all patients with LV ejection fraction ≤40% and for those with hypertension, diabetes, or chronic kidney disease. ACE inhibitors may be considered for all other patients (65).

Glycemic control

The findings from ADVANCE, ACCORD, and UKPDS have contributed to the understanding of the utility of intensive glycemic control in the prevention of vascular disease. The benefit of intensive glycemic control (66), but not BP control (67), on microvascular disease, renal and metabolic complications, MI, and death was evident in UKPDS during 10 years of post-trial monitoring. Both UKPDS and ADVANCE (68) demonstrated that tighter glucose control prevented microvascular but not macrovascular complications during the period of randomized comparison. Neither trial showed evidence of an increase in mortality, a finding that was observed with intensive glucose control in ACCORD (69). All three trials similarly observed that good or nearly normal glycemic control does not reduce major CV events in the short term.

The UKPDS reports suggest that in patients with type 2 diabetes, intensive glucose control may have a ‘legacy effect’, i.e. clinical benefit remains long after the cessation of therapy, whereas the same benefit was not observed with tight BP control. However, the increase in death rates observed in ACCORD indicates that it is not simply a matter of ‘lower is better’ for glucose control. Taken together, these studies indicate that it is not appropriate to manage risk factors and conditions individually. The results reinforce the importance of utilizing a collective approach to risk management and to maintenance of good glycemic control, not only for the prevention of diabetic complications but also for protection against the development of CVD in the long term.

β-Adrenergic system blockers

The benefit of β-blockers in ischemia and hypertension has been demonstrated in several trials, and subsequent studies showed benefit in chronic HF, despite prior contraindication. β-Blockers have been considered valuable for secondary prevention in patients after unstable angina and MI. However, the CV benefit of β-blockers has been challenged by subsequent randomized trials and meta-analyses. The COMMIT trial, which separately assessed the efficacy and safety of early β-blocker therapy (intravenous then oral metoprolol) and antiplatelet therapy (adding clopidogrel to aspirin) in suspected acute MI, found that allocation to metoprolol did not reduce the rate of death, reinfarction, or cardiac arrest, but significantly increased the risk of cardiogenic shock (P<0.00001) (70). A meta-analysis of nine trials of patients with hypertension found that heart rate reduction (as achieved with β-blocker therapy) was significantly associated with greater risk of all-cause mortality, CV mortality, MI, and HF (all P<0.0001) (72). Furthermore, a meta-analysis of 33 trials of patients having non-cardiac surgery found that perioperative β-blocker therapy did not reduce the risk of all-cause mortality, CV mortality, or HF, but decreased the risk of MI. However, the risks of stroke and perioperative bradycardia and hypotension were significantly increased (71). The disparity between older and more recent trials assessing the CV benefit of β-blockers could stem from the evolution of treatment intensity during this time frame. As patients are now treated with multiple agents, the benefit of β-blockers is not as manifest. Currently, the strongest evidence for the clinical benefit of β-blockers is in patients with HF (73), and AHA/ACC guide-lines strongly recommend the use of these agents before discharge for secondary prevention in those patients with compensated HF or LV systolic dysfunction (74).

Anticoagulants

The anticoagulant warfarin has been widely used for prophylaxis and treatment of deep vein thrombosis and pulmonary embolism, as well as for prevention of ischemic stroke in patients with atrial fibrillation. However, warfarin is rarely used for secondary prevention in patients with atherothrombosis, because clinical trials have not demonstrated a clear benefit over aspirin alone. Furthermore, warfarin is associated with bleeding risk, numerous drug and food interactions, and requires frequent blood testing for monitoring.

The CARS trial (75) indicated that the combination of a low fixed dose of warfarin (1 mg or 3 mg) with aspirin 80 mg resulted in a higher rate of spontaneous major bleeding versus aspirin 160 mg alone, but was not associated with a reduction in ischemic events in patients with prior MI. However, the doses of warfarin used in CARS were not anticoagulant in most subjects, which limits the interpretation of these findings. The CHAMP trial (76) similarly compared the efficacy of aspirin 81 mg plus warfarin (INR (International normalized ratio) 1.5 to 2.1) versus aspirin 162 mg monotherapy in patients with prior MI, demonstrating no difference in ischemic events and significantly higher rates of both major and minor bleeding with combination therapy. Warfarin did not demonstrate prevention of recurrent ischemic events in patients with prior ischemic stroke as monotherapy at an INR of 1.4–2.8 versus aspirin in the WARSS trial (77) or in patients with symptomatic PAD as an add-on to aspirin at an INR of 2.0–3.0 in the WAVE trial (78). Hemorrhagic stroke and bleeding risk were significantly higher in the warfarin-plus-aspirin arm in WAVE. The SPIRIT trial, which evaluated a more intensive anticoagulation regimen (INR 3.0–4.5) versus aspirin in patients with TIA or minor stroke, was halted early because of an excess of major bleeding (particularly intracranial bleeding) in patients receiving warfarin (79). In contrast to these findings, a meta-analysis of ten smaller, mostly European trials (80) in patients with recent ACS showed that the combination of warfarin plus aspirin resulted in reduced rates of MI, ischemic stroke, and revascularization versus aspirin alone. However, major bleeding risk was 2.5-fold higher with warfarin plus aspirin versus aspirin alone (80).

The clinical experience with warfarin has prompted the development of several novel anticoagulants that aim to improve upon its risk/benefit profile (reviewed in (81)). The coagulation pathway provides many potential therapeutic targets, including factor Xa (FXa) and thrombin. Numerous oral, direct FXa inhibitors are in various stages of clinical development, including rivaroxaban, apixaban, LY517717, YM150, DU-176b, and betrixaban. Dabigatran is the only oral direct thrombin inhibitor (DTI) in late-stage development. On-going trials will assess whether these agents can mitigate bleeding risk without the need for frequent monitoring or dose adjustment.

Antiplatelet agents

Secondary prevention with antiplatelet therapy has been associated with significant reductions in ischemic events among patients with documented atherothrombotic disease. A comprehensive meta-analysis of 287 studies in patients with prior MI, prior stroke/TIA, or PAD (2) confirmed a 25% relative risk reduction in secondary events (non-fatal MI, non-fatal stroke, vascular death) in patients treated with antiplatelet therapy (either aspirin alone or the combination of aspirin with another antiplatelet agent). However, regular use of aspirin ≤300 mg qd has been associated with a 2-fold increased risk of gastrointestinal bleeding (82).

The platelet ADP P2Y₁₂ receptor antagonist clopidogrel has also demonstrated beneficial effects in secondary prevention. The CAPRIE trial (1) evaluated clopidogrel 75 mg versus aspirin 325 mg in patients with established atherothrombotic disease (MI, ischemic stroke, or objectively established PAD). Vascular death, MI, or ischemic stroke occurred in 5.8% of patients receiving aspirin and 5.3% of those receiving clopidogrel (P=0.043). Subgroup analysis showed a relative risk reduction of 23.8% (P=0.003) with clopidogrel among patients with PAD, whereas no significant difference between clopidogrel and aspirin was apparent in patients with prior ischemic stroke or MI. Bleeding rates were not significantly different between the two treatment arms. The CHARISMA trial (83) compared the efficacy of clopidogrel plus aspirin versus aspirin monotherapy in patients who had either CVD, cerebrovascular disease, or PAD (secondary prevention cohort) or multiple risk factors (primary prevention cohort). There was no significant difference in the incidence of CV death, MI, or stroke between the treatment arms (Figure 2) (83). Severe bleeding was not different between treatment arms, although moderate bleeding was higher in the clopidogrel-plus-aspirin arm (P=0.004). Subgroup analyses indicated that treatment with clopidogrel plus aspirin appeared more effective than aspirin alone in the secondaryprevention cohort (6.9% versus 7.9%, respectively; P=0.046), whereas a trend toward harm was evident in the primary prevention cohort (6.6% versus 5.5%, respectively; P=0.20) (Figure 2). The primary outcome was reduced from 8.8% with aspirin to 7.3% with aspirin plus clopidogrel (P=0.01) in the ‘CAPRIE-like’ group of patients with prior ischemic events (Figure 2) (84). Results from post hoc analyses should be interpreted with caution and require additional studies for confirmation of findings.

Figure 2. Primary end-point rate in the CHARISMA trial (83,84). No significant difference was obtained in the primary end-point rate (composite of myocardial infarction, stroke, or death from cardiovascular causes) between placebo+aspirin versus clopidogrel+aspirin groups in the overall study population and in the primary prevention cohort (83). Significant reductions in favor of clopidogrel+aspirin were evident in the secondary prevention cohort (83) and in the CAPRIE-like population (84). Green arrows indicate relative risk reduction between treatment arms. Red arrows indicate residual risk for recurrent events.

Other trials have documented the benefit of antiplatelet therapy in patients with prior stroke or TIA. Early trials of the phosphodiesterase inhibitor dipyridamole, alone or in combination with aspirin, gave conflicting results (85–87). The ESPS-2 trial demonstrated a relative 18% reduction in stroke with aspirin alone, 16% with extended-release dipyridamole alone, and 37% with the combination, suggesting an additive effect (88). However, several criticisms of this trial, including low aspirin dose, high patient withdrawal rate, and low compliance rate, have led to reservations about its results. The ESPRIT trial demonstrated that the combination of extended-release dipyridamole (ERDP) plus aspirin significantly reduced the relative risk for death from all vascular causes, non-fatal stroke, non-fatal MI, or major bleeding complications by 20% versus aspirin alone (89). Dual antiplatelet therapy with aspirin and clopidogrel did not provide significant clinical benefit compared with clopidogrel alone in patients with recent ischemic stroke or TIA in the MATCH trial. The addition of aspirin to clopidogrel caused a significantly greater frequency of life-threatening and major bleeding compared with clopidogrel alone (88). The antiplatelet comparison within the PROFESS trial compared the benefit of aspirin plus ERDP versus clopidogrel in patients with recent stroke followed for 2.5 years (90). Similar rates of recurrent stroke were observed in the two treatment arms (8.8% versus 9.0%). Higher rates of hemorrhagic events were observed in patients receiving aspirin plus ERDP versus clopidogrel (4.1% versus 3.6%), including intracranial hemorrhage (1.4% versus 1.0%). Most recently, the ACTIVE trial evaluated the benefit of aspirin plus clopidogrel in patients with atrial fibrillation who were not candidates for vitamin K antagonist therapy (91). Dual antiplatelet therapy was associated with a significant 11% reduction in the relative risk of major vascular events versus aspirin alone, which was mainly attributed to a reduction in the risk of stroke (2.4% versus 3.3%; relative risk 0.72; P<0.001). Major bleeding occurred more frequently with clopidogrel plus aspirin versus aspirin alone (2.0% versus 1.3%; P<0.001). Current AHA/ACC guide-lines for stroke prevention recommend aspirin, the combination of aspirin and extended-release dipyridamole, or clopidogrel for patients with ischemic stroke or TIA, with the combination of aspirin and extended-release dipyridamole suggested over aspirin alone (9).

Residual risk

Despite the unambiguous clinical benefit achieved with life-style modification, BP control, statins, angiotensin II-active agents, and antiplatelet agents, the residual risk for recurrent acute events in patients with established atherothrombotic disease remains substantial (Figure 3) (2,52). This residual risk is likely related to progression of atherosclerosis in the presence of statins and angiotensin II-active agents, lack of more complete inhibition of platelet activation in the presence of aspirin and clopidogrel, and other factors yet to be identified.

Figure 3. Residual atherothrombotic risk: rates of cardiovascular (CV) death and myocardial infarction (MI) persisting with active treatments from meta-analyses of trials of secondary prevention with statins (52), angiotensin-converting enzyme (ACE) inhibitors (60), β-blockers (71), antiplatelets (2), and warfarin (80). Treatment and comparator arms consisted of the following: high-dose versus standard-dose statins; ACE inhibitor versus placebo; β-blocker versus placebo or other antihypertensive agents; antiplatelet therapy versus control; and warfarin+aspirin versus aspirin. *Event rate is shown for coronary death/MI for statin meta-analysis (individual rates for MI not available). OR = odds ratio.

It is currently unknown how much additional reduction of this residual risk could be achieved with improved approaches to limit atherosclerosis progression or platelet-mediated thrombosis. Recent attempts to further limit atherosclerosis progression with novel approaches, such as inhibition of cholesteryl ester transfer protein (CETP) (92,93), inhibition of acyl-coenzyme A:cholesterol acyltransferase (ACAT) (94), and inhibition of intestinal cholesterol absorption (95), have generally been unsuccessful. Further attempts at CETP inhibition with the novel agent anacetrapib have shown favorable changes in lipid profiles in patients with dyslipidemia, without safety concerns in early clinical trials (96), and a phase 3 trial is on-going (Clinicaltrials.gov identifier: NCT00685776). The addition of niacin to a statin has been shown to slow the progression of atherosclerosis (97), and an on-going trial is comparing the effects of niacin versus a cholesterol absorption inhibitor, ezetimibe, on progression of atherosclerosis in patients receiving a statin (Clinicaltrials.gov identifier: NCT00687076). However, a recent study with niacin/laropiprant designed to assess the effect on atherosclerosis progression was prematurely terminated (Clinicaltrials.gov identifier: NCT00384293). It should be noted that the studies that evaluated the effect of statins on progression of atherosclerosis have produced equivocal results (92,93,95,98,99). Nevertheless, novel lipid-modifying approaches have so far not been demonstrated to further reduce adverse CV events beyond statin alone (100), and a large on-going trial is evaluating the effect of adding ezetimibe to a statin on CV outcomes (Clinicaltrials.gov identifier: NCT00202878). The effect of adding niacin to a statin in secondary prevention has not been evaluated to date. It is also interesting to speculate that improving therapeutic regimens to include treating the inflammatory component of CV disease may also improve patient outcomes, as mentioned above.

Response variability (also referred to as platelet resistance) to antiplatelet therapy is considered to be a potentially important limitation of antiplatelet therapy and may contribute to residual risk for thrombotic events. Although a standardized definition and methodology for measurement of low responsiveness to antiplatelet therapy has not been established, sufficient evidence supports the concept that persistence of enhanced platelet reactivity despite the use of aspirin (101) and/or clopidogrel (102) is associated with adverse clinical outcomes. A number of factors may influence response to antiplatelet therapy, including drug dose or absorption, patient compliance, and genetic polymorphisms (102). Recent analyses suggest that genetic polymorphisms of the cytochrome P (CYP) 450 enzymes, which are required for the conversion of clopidogrel into an active metabolite, can significantly modulate individual response to clopidogrel and are important determinants of prognosis (103–105). Whether new antiplatelet agents exhibit less variability or no variability in platelet responsiveness and the association with clinical outcomes will be interesting to note as these agents progress through clinical development.

Novel approaches to limit platelet-mediated thrombosis also offer promise. Ticagrelor (AZD6140) is a third-generation P2Y₁₂ receptor antagonist that is currently in phase 3 trials (Clinicaltrials.gov identifiers NCT00385138, NCT00305162, and NCT00391872). Phase 3 trials of cangrelor, a novel P2Y₁₂ receptor antagonist, were recently halted due to a lack of efficacy in an interim analysis. Prasugrel is approved in Europe and the US. Prasugrel, which demonstrated increased efficacy and bleeding versus clopidogrel in TRITON, also exhibits less variability in platelet responsiveness (106). Current antiplatelet agents and those in development inhibit the TxA₂ or ADP platelet activation pathways but do not interfere with a multitude of other platelet activation pathways that may contribute to thrombotic events. In addition to significant residual risk for ischemic events, dual antiplatelet therapy is associated with increased bleeding. These considerations underscore the critical need for novel therapies that provide more comprehensive platelet inhibition for greater protection against thrombotic events, preferably without an incremental bleeding risk.

Surrogate markers

Available data from epidemiological studies indicate that traditional risk factors do not fully explain the residual risk for thrombotic events or treatment responses (107). There is a need for the development and clinical implementation of new markers of atherothrombotic disease which are intended to predict future risk of thrombotic events, particularly in high-risk patients. For example, measurement of urinary 11-dehydro-thromboxane B₂, a stable metabolite of TxA₂ can be used to assess the degree of inhibition of TxA₂ production by aspirin (and hence aspirin resistance) (108). An analysis of HOPE revealed that increasing quartiles of urinary 11-dehydro-thromboxane B₂ were associated with increased risk of MI, stroke, or CV death (odds ratio 1.8; 95% confidence interval 1.2–2.9; P<0.009 for first versus fourth quartile), suggesting the predictive value of this metabolite (109).

Vascular calcification has also been shown to have predictive value. The presence and extent of coronary artery calcification (CAC) is associated with atherosclerotic plaque burden (110). Elevated CAC scores have been shown to predict future cardiac events in both asymptomatic (111–115) and symptomatic (116,117) individuals. Current guidelines state that asymptomatic individuals at intermediate risk for CHD ‘may be reasonable candidates’ for CAC testing, whereas in symptomatic patients, exclusion of CAC score may be an effective risk stratification tool before invasive procedures or hospital admission (118).

Aortic stiffness is associated with changes in the arterial wall resulting in increased systolic BP and may also reflect coronary artery lesions (119). As measured by aortic pulse wave velocity (PWV), aortic stiffness has predictive value for all-cause and CV mortality, MI, and stroke in a wide spectrum of patients, including those with end-stage renal disease (ESRD) (120), diabetes (121), hypertension (122,123), and healthy individuals (124–126). Whether a reduction in PWV independently predicts a reduction in CV events has been suggested in patients with ESRD (127) but remains to be established in patients at lower CV risk. Left ventricular hypertrophy (LVH) is an additional predictor of CV events which is associated with hypertension severity. Increasing LVH is associated with increases in risk of stroke, CHD, and heart failure, and reducing LVH with antihypertensive medications lowers CV risk (128). However, different classes of antihypertensive drugs reduce LVH to varying degrees which may not correlate with reduction in BP (129).

Chronic inflammation plays a key role in plaque instability and subsequent occlusive thrombosis. Elevated levels of inflammatory mediators have predictive value for vascular events (reviewed in Libby et al. (130), and several of these molecules are under clinical investigation as therapeutic targets. For example, human A₂ phospholipases (PLA₂), such as lipoprotein-associated PLA₂ (Lp-PLA₂), independently predict risk of CHD and ischemic stroke (131). Several novel inhibitors of Lp-PLA₂ and secretory PLA₂ (sPLA₂) are under development (132,133). The oral agent darapladib showed a reduction in atheroma growth in patients with documented coronary disease receiving standard of care (133). Of note, darapladib did not lower LDL-C (133), whereas an 8% reduction in LDL-C was noted with the sPLA₂ inhibitor A-002 (132), suggesting disparate effects on LDL clearance. STABILITY, a large on-going phase 3 trial, is assessing the benefit of darapladib on CV events in patients with CHD (134), while A-002 is in phase 2 testing in combination with atorvastatin (135). C-reactive protein (CRP) is arguably the most promising inflammatory biomarker based on results of multiple population-based studies showing that CRP levels predict future CV events, diabetes, and hypertension in patients with and without CVD (reviewed by Ridker (136). These studies have suggested that the relative predictive power of CRP is at least as strong as other established risk factors such as LDL-C, BP, and smoking (136). The results of the recent JUPITER trial, which demonstrated the benefit of rosuvastatin in patients without hyperlipidemia but with elevated CRP, support the notion that CRP levels may be useful in stratifying patient risk and suggest that statin treatment may be beneficial in otherwise healthy individuals (46). Current guide-lines recommend the optional use of CRP measurement in patients at intermediate risk (i.e. 10%–20% risk of CHD over 10 years) to help guide considerations for further evaluation or therapy (137).

Registry analyses document unmet needs in current clinical practice

The global REACH registry offered a unique opportunity to asses the prevalence of risk factors, the use of various medications, and clinical outcomes around the world, including North America (138). Patients eligible for enrollment included those with documented atherothrombotic disease (CAD, cerebrovascular disease, or PAD) as well as those with ≥3 risk factors for atherothrombosis. Over 68,000 patients in 44 countries were followed for 1 year from 2003 to 2004. North American participants (n=25,999), of whom 75% had symptomatic atherothrombotic disease, were characterized by a high prevalence of hypertension (86.4%), hypercholesterolemia (82.7%), and high body mass index (≥25; 77.8%), as well as current or former smoking (58.1%) and diabetes (50.6%). The vast majority of North American patients (88.4%) were managed by either general/family practitioners or internists, and only 7.4% were managed by cardiologists. Statins and aspirin were used in 77.1% and 71.4% of North American patients, respectively, whereas only a minority (14.2%) received oral anticoagulants. Use of dual antiplatelet therapy in North American patients was also low (14.2%), suggesting opportunities for improvement in patient care. In the global cohort, the combined incidence of CV death, MI, stroke, or hospitalization for an atherothrombotic event at 1 year was almost 3-fold higher in patients with established atherothrombotic disease than in those with multiple risk factors (14.4% versus 5.3%, respectively). In the North American cohort that included 75% of patients with established atherothrombosis, the 1-year composite rate of CV death, MI, stroke, or hospitalization for atherothrombosis was 11.6%. Furthermore, the annual incidence of CV death, MI, or stroke exceeded the anticipated 3% event rate in all regions including North America (3.7%). These observational data confirm the high residual risk of patients with documented atherothrombotic disease reported in clinical trials and highlight the need for more effective therapies.

Conclusion

Although death rates due to CVD continue to decline, atherothrombotic disease remains a major public health issue; its prevalence can be expected to rise as our population ages and the incidence of diabetes/metabolic syndrome continues to increase. Life-style modifications, BP control, statins, angiotensin II-active agents, and antiplatelet therapy—current cornerstones of secondary prevention—target key pathophysiological processes responsible for acute ischemic events, namely atherosclerotic plaque rupture/erosion and platelet-mediated thrombosis. Clinical trials with pharmacological agents added to life-style modification have demonstrated significant reductions in morbidity and mortality over those achieved with attempts at life-style modification alone. Despite these proven benefits, residual risk for secondary ischemic events remains substantial, as thrombotic events continue to occur even in the presence oflife-style modification and pharmacotherapy. Therefore, there is a need for novel agents, perhaps those directed at other pathways, that can further reduce ischemic events by providing more comprehensive inhibition of platelet-mediated thrombosis, without increasing the risk of bleeding. In addition, the approach to care will likely require a risk/benefit assessment in high-risk patients in order to balance safety with preventing potentially fatal outcomes.

Acknowledgements

Dr Pepine reports serving as a consultant for Schering-Plough. Dr Pepine did not receive payment for the development of this manuscript, and no outside sources were involved in the development, review, or final approval of its content.

Dr Pepine would like to thank Gina Fusaro, PhD, and Joshua Barbach, MA, for providing editorial assistance during the preparation of this manuscript. This assistance was funded by Schering-Plough.