Acute coronary syndromes (ACS) and acute pulmonary embolism (APE) are leading causes of hospital admission and mortality. Accurate risk stratification is important for both conditions to guide management and the use of therapies that lead to improved outcomes.
Acute Coronary Syndromes
Despite a steady decline in developed countries, ischaemic heart disease is still the leading cause of death worldwide and by 2030 is projected to cause 12.6 million deaths annually (14.2% of deaths worldwide).1 In the US alone, ACS account for >1.4 million hospitalisations per year, with up to one-third of patients presenting with acute ST-segment-elevation myocardial infarction (STEMI).2 New definitions for acute coronary syndrome and MI reflect more sensitive tests for diagnosis and a greater understanding of the underlying pathophysiology, including acute disruption of an atherosclerotic plaque with associated platelet aggregation leading to partial or complete thrombotic occlusion of a coronary artery and consequent myocardial necrosis.3–5 Across the spectrum of ACS, mortality is highest in acute STEMI, with 30-day mortality rates of 6% reported in large trials6 and up to 11% in community, non-trial settings.7 Acute non-STEMI is associated with slightly lower 30-day mortality rates of 5.7 and 7.4% in trial and community settings, respectively.6,7 In contrast, 30-day mortality in unstable angina (without elevated cardiac injury markers) is significantly lower, estimated at 1.7–2.4%.6,7 Early risk stratification is strongly recommended in guidelines for ACS, to guide optimal treatment strategies.3,8
Clinical Predictors of Risk in Acute Coronary Syndromes
A number of clinical indices predict mortality in ACS. Age is a strong predictor of short- and long-term mortality, with in-patient mortality rates as low as 1% in subjects ≤65 years of age compared with 10% in those ≥85 years of age.9–14 Acute heart failure occurs in up to 15% of ACS presentations and is associated with three- to four-fold higher in-hospital and long-term mortality.10,13–16 The highest risk in non- STEMI is seen in subjects with Killip class III or IV (overt pulmonary oedema or cardiogenic shock), where six-month mortality approaches 25%.17 Lesser degrees of haemodynamic compromise – including lower systolic blood pressure or resting tachycardia at presentation – also predict higher mortality.9,10,13,17
The presence of renal dysfunction is a powerful independent predictor of both shortand long-term outcome, and an antecedent history of conventional risk factors such as hypertension or diabetes also predicts a higher likelihood of heart failure and other adverse clinical events.10,14,18–20
Electrocardiography and Risk in Acute Coronary Syndromes
Electrocardiography (ECG) provides powerful risk stratification in the setting of ACS. In patients with STEMI, the extent of ST-segment elevation, evidence of prior infarction or QRS prolongation are all independent predictors of increased mortality.21,22 In patients without STEMI, the presence of ST-segment depression (≥1mm), particularly in the anterior leads, predicts increased risk of mortality or re-infarction and is associated with benefit from an early invasive approach.18,23–26
Biomarkers and Risk in Acute Coronary Syndromes
Due to their sensitivity and specificity for detecting even minor cardiac injury, cardiac troponins I and T have become established as the preferred biochemical markers for the diagnosis of ACS.27 The troponins also provide independent risk stratification across the spectrum of ACS from unstable angina to STEMI.16,28–31 In studies such as the Global Utilisation of Strategies to Open occluded arteries (GUSTO) IIa trial, an elevated troponin level was associated with a more than two-fold higher 30-day mortality compared with those with a negative troponin.31 Data from community hospital settings suggest that mortality in troponin-positive subjects may be closer to eight-fold higher than for those with negative troponin.29 There is an almost linear relationship between troponin level and risk of mortality or re-infarction, with stepwise increases in mortality rates observed for increasing levels of troponin.28,30–32
Elevated troponin levels play an important role in guiding therapy in ACS. Treatment with low-molecular-weight heparin was associated with lower rates of death, MI or need for urgent re-vascularisation in troponin-positive but not in troponin-negative subjects.33,34
Benefit from glycoprotein (GP) IIb/IIIa inhibitors as an adjunct to either medical therapy or an invasive strategy for ACS is limited to subjects with elevated troponin.35–37 Findings from the Fast Revascularisation during Instability in Coronary artery disease (FRISC) II and Treat Angina with aggrastat and determine Cost of Therapy with Invasive or Conservative Strategy – Thrombolysis In MI (TACTICS–TIMI) 18 trials demonstrated that the benefit of an early invasive strategy for ACS was seen in subjects with elevated troponin, but not in those with normal troponin.38–41
B-type Natriuretic Peptide
B-type natriuretic peptide (BNP) and its amino terminal fragment (NT-proBNP) are predominantly synthesised in the ventricular myocardium and released into the circulation in response to myocardial stretch resulting from pressure or volume overload. Secretion of both peptides is increased during myocardial ischaemia or infarction and circulating plasma levels reflect the extent of myocardial damage, correlating inversely with left ventricle (LV) ejection fraction.39,42,43 Elevated BNP or NT-proBNP levels in ACS are among the most powerful markers of outcome in ACS and are associated with increased risk of mortality or heart-failure events independent of clinical risk factors, LV ejection fraction and troponin level.18,42–44 Elevation of NT-proBNP levels may also identify subjects who will benefit from an early invasive strategy.38
High-sensitivity C-reactive Protein
Inflammatory mechanisms play an important role in atherosclerosis and may contribute to plaque disruption associated with ACS.5 C-reactive protein (CRP) is a non-specific acute-phase reactant; whether it has a pathogenic role in ACS is unclear, but recent studies suggest that plasma high-sensitivity CRP (hsCRP) levels are elevated in ACS and are associated with adverse outcome.45,46 In the acute setting there have been mixed reports, but data from TIMI 11A and GUSTO IV ACS demonstrated that hsCRP levels above 15.5 and 9.62mg/l, respectively, at presentation were associated with increased mortality.45,46 The data in terms of long-term prognosis are more consistent with data from GUSTO-IV ACS, TIMI IIIB and FRISC II, demonstrating that hsCRP levels predict mortality up to four years, independently of troponin levels.45–47
As persistently elevated hsCRP levels are associated with increased risk of vascular events,18 there may be a role for targeting statin therapy to achieve lower hsCRP levels. Statins have been shown to lower hsCRP independent of actions on low-density lipoprotein, and in the Pravastatin or Atorvastatin Evaluation and Infection Therapy (PROVE IT)-TIMI 22 study patients who achieved lower hsCRP levels experienced the lowest event rates.48
As the cardiac troponins, the BNPs, CRP and other markers provide independent prognostic information use of a multimarker approach may provide incremental and more sophisticated risk stratification after ACS. Patients with levels of troponin, NT-proBNP or CRP in the highest quartile have the highest risk of mortality at one year (see Figure 1). In the Orbofiban in Patients with Unstable coronary Syndromes – Thrombolysis In MI (OPUS–TIMI) 16 study, when patients were categorised on the basis of the number of elevated biomarkers at presentation, there was a near doubling of the mortality risk at 30 days and 10 months for each additional biomarker (troponin I, BNP or CRP) that was elevated.49 Multimarker strategies may potentially guide therapy in ACS. Data from the GUSTO-IV study demonstrated that the benefit from an early invasive strategy was greatest in subjects with above median levels of both NT-proBNP and troponin T (TnT), whereas for subjects with below median levels of both peptides who underwent an early invasive strategy there was increased mortality at one year (see Figure 2).38,39
A number of novel risk markers have been identified that reflect pathophysiological pathways in ACS. Increased levels of markers of inflammation such as CD-40 ligand, interleukin (IL)-6 and myeloperoxidase (MPO) are associated with an increased risk of mortality, MI and heart failure after ACS.38,50–54 In the case of MPO, an enzyme released by activated neutrophils plays a major role in inflammatory responses: its levels are elevated in plasma and in infarcted myocardium after ACS, and elevated levels are associated with increased mortality risk.51 Plasma levels of signal peptide-CUB-1 (SCUBE-1) – a novel marker of platelet activation – are also raised in ACS and are associated with increased risk.55 While newer biomarkers are the focus of ongoing study, their role in the management of ACS, at least for the moment remains unclear.19
Cardiac Imaging and Risk in Acute Coronary Syndromes
Whether assessed by echocardiography, radionuclide ventriculography or cardiac magnetic resonance imaging, indices of LV size and function are important determinants of outcome after ACS. In particular, LV end-systolic volume and LV ejection fraction are powerful independent predictors of mortality and subsequent heart failure.14,56–59 The severity of diastolic abnormality provides incremental risk prediction over ejection fraction alone, with a restrictive transmitral filling pattern or an elevated ratio of early transmitral filling velocity to early annular myocardial relaxation velocity (E/E’) predicting higher risk.60 Right ventricle (RV) dysfunction or increasing severity of mitral regurgitation also predict increased risk.61,62 At coronary angiography, greater extent and severity of coronary stenoses also predict mortality or risk of further MI.63
Functional Testing in Acute Coronary Syndromes
Functional testing is useful for assessing the burden of inducible ischaemia, with greater sensitivity and specificity when performed in conjunction with radionuclide imaging or echocardiography. Evidence of inducible ischaemia, particularly at low workload, is associated with increased risk of MI or death following ACS.64 In this context, re-vascularisation may lower the risk of adverse events, while absence of inducible ischaemia may identify a group of subjects for whom conservative therapy may be appropriate.3,65
Risk Scores in Acute Coronary Syndromes
There are several validated risk scores available for ACS that combine the most powerful clinical, ECG and biochemical indices at the time of presentation into simple algorithms, which can be used for risk stratification in the bedside clinical setting. Of these, the Thrombolysis in MI (TIMI), Global Registry of Acute Coronary Events (GRACE) and Predictors of outcome in patients with ACS without persistent ST-segment elevation (PURSUIT) risk scores are the best known and validated (see Table 1).9,10,66,67 These scores are similar in many respects, although the PURSUIT score does not include a marker of myocardial injury. A more complex GRACE score, which includes more variables, can be applied at discharge.
For each risk score, when patients are categorised by the number of variables present, there is an incremental increase in risk of mortality with each added risk variable.9,10,66,67 All three risk scores provide similar independent risk stratification for events in hospital or at one year and in the case of the GRACE score up to four years.68–70 Utility may depend on the context, but in one comparison prediction of events in hospital or at one year was slightly superior for the GRACE and PURSUIT risk scores than for the TIMI risk score.68–70 Risk scores may also guide therapy decisions in ACS. For example, the greatest benefit from an early invasive strategy has been demonstrated in subjects with higher TIMI risk scores.71
Risk stratification plays an important role in identifying high-risk subjects and guiding therapy in ACS. Integration of clinical, imaging and biochemical indices provides the most sophisticated risk assessment. Decision-making regarding treatment strategies, such as an early invasive approach in non-STEMI, can be guided by early risk assessment including troponin levels, with the adjunctive use of NT-proBNP/BNP or functional testing in some settings.
Acute Pulmonary Embolism
APE is a relatively common condition that carries significant mortality. The true incidence of venous thromboembolism (VTE) is uncertain, but has been estimated in hospitalised patients at 70–113 cases per 100,000 population per year, with 28–41% of cases due to PE.72 In community settings, incidences of APE of six per 10,000 population per year are estimated.73 In autopsy studies, VTE has been identified in up to 25% of cases, with APE defined as the cause of death in up to 13% of cases.74
The overall mortality for APE is reported at 15.3% in three months, due largely to recurrent APE or to underlying causes including cancer.75 Mortality rates are highest in patients with massive APE and clinical shock. For patients with ‘sub-massive’ or ‘non-massive’ APE there is a spectrum of risk that can be stratified according to the degree of haemodynamic compromise, the extent of RV dysfunction and evidence of myocardial injury (see Table 2).74 Careful risk stratification is vital to identify high-risk patients and potentially guide therapy decisions (see Table 3).
Management of PE is directed at accurate early diagnosis, attention to predisposing causes and prompt anticoagulation, frequently initially with low-molecular-weight heparin. Current guidelines reserve thrombolysis or embolectomy for patients with massive pulmonary embolus and clinical shock.74–77
Clinical Risk Markers in Pulmonary Embolism
A variety of clinical indices predict outcome in acute pulmonary embolus. Older age (>70 years) and a previous history of cancer, congestive heart failure or chronic obstructive pulmonary disease are associated with increased mortality from PE. Haemodynamic status at time of presentation is the most powerful marker of mortality risk (see Table 3).74,78 In-hospital and 30-day mortality is greatest in patients presenting with massive APE and shock, with in-hospital mortality rates in excess of 25% and 30-day mortality as high as 58% reported in this context.74,75,78 In subjects without shock but with hypotension, defined as a systolic blood pressure <90mmHg or a fall of >40mmHg for more than 15 minutes, in-hospital mortality rates are lower than for shock, but comprise about 15%.77 Within the group of patients with sub-massive or non-massive PE, risk can be further sub-stratified by assessing markers of RV dysfunction and myocardial strain or injury.74
Imaging and Risk in Pulmonary Embolism
The presence of RV enlargement or dysfunction is a powerful predictor of adverse outcome in acute pulmonary embolus and should therefore be assessed carefully, particularly in hypotensive patients.74
Echocardiography allows assessment of RV size and function. Up to 25% of subjects presenting with APE have RV abnormalities on echocardiography.74 Regardless of systemic blood pressure, the presence of RV abnormality – defined as RV dilatation, RV hypokinesis, elevated peak tricuspid regurgitation velocity (>3m/s) or right-to-left shunt through a patent foramen ovale – is associated with a near doubling in mortality risk.74,79–82 When these findings are absent and RV size and function are normal, early mortality rates are low (<1%).74
Computed Tomography Pulmonary Angiogram
In addition to facilitating diagnosis of acute PE, multidetector computed tomography pulmonary angiogram (CTPA) also enables an assessment of the RV.74 While assessment of RV systolic function is not feasible, estimates of RV size from CTPA correlate strongly with echocardiographic measurements. The presence of a dilated RV on CTPA at the time of diagnosis of APE is independently associated with increased mortality.83,84 A ratio of RV to LV area >0.9 is associated with a more than four-fold increase in mortality risk, while a value <0.9 has a negative predictive value >95% for mortality.85,86 Additional findings, such as flattening of the interventricular septum, dilatation of the vena cava or azygous vein and reflux of contrast into the inferior vena cava, may also be identified; although these correlate with increased RV pressures, they do not appear to provide additional risk stratification.74
Biomarkers and Risk in Pulmonary Embolism
Elevation of biochemical markers that reflect myocardial strain or injury is associated with increased mortality from acute PE, even in normotensive subjects. Elevated plasma BNP and NT-proBNP levels are associated with RV dilatation and dysfunction in the setting of acute PE.87,88 Low levels of BNP or NT-proBNP have a high negative predictive value, essentially ruling out RV dysfunction. Kruger et al. demonstrated that a BNP <90pg/ml ruled out RV dysfunction in subjects with acute PE.87 Similarly, patients with acute APE and NT-proBNP levels <1,000pg/ml are unlikely to have RV dysfunction or elevation of cardiac troponin levels.88 BNP or NT-proBNP levels could therefore be used to identify patients who are more likely to need cardiac imaging to evaluate RV size and function. More importantly, elevated BNP/NT-proBNP levels are powerful predictors of adverse clinical outcomes in acute PE.89 Although no cut-off points have been established in guidelines for estimation of risk in everyday clinical practice, low levels of BNP or NT-proBNP exhibit very high negative predictive value for mortality,90–92 with BNP levels <500pg/ml associated with very low mortality risk in normotensive patients.90
Elevation of the cardiac troponins I and T can be seen in pulmonary embolus and may reflect RV infarction or micro-infarction. Troponin elevation is more common in massive or sub-massive APE or in the setting of RV dysfunction.74,93 The presence of an elevated troponin is associated with an increased risk of developing shock and an increased 30-day mortality risk, independent of haemodynamic status at presentation and RV function on cardiac imaging.88,94–96 Even in haemodynamically stable patients, elevation of cardiac troponin is associated with a five-fold increase in mortality from PE.93 Similar to BNP and NT-proBNP, normal troponin levels at admission are associated with low likelihood of clinical deterioration or early mortality.
Use of multiple biomarkers in combination with cardiac imaging provides incremental risk stratification in the setting of acute PE.96,97 The risk of dying or clinical deterioration is higher in subjects with an elevated biomarker and evidence of RV dysfunction on imaging than in those with abnormality of only one marker.90,93,98 Conversely, the risk of clinical deterioration or mortality due to APE is lowest in patients with normal or low levels of both BNP and troponin and with normal RV function.74
ACS and acute pulmonary embolus are common diagnoses that are associated with high rates of mortality and adverse clinical outcome. Careful assessment of clinical and haemodynamic status in combination with cardiac imaging and quantitation of biomarkers such as the cardiac troponins and the B-type natriuretic peptides can allow more accurate status risk stratification and may guide more optimal therapy.