Efficacy of noninvasive cardiac imaging tests in diagnosis and management of stable coronary artery disease

Abstract

The aim of this review was to discuss the current literature regarding the utility of noninvasive imaging in diagnosis and management of stable coronary artery disease (CAD) including recent data from large randomized trials assessing diagnosis and prognosis. Current guidelines recommend revascularization in patients with refractory angina and in those with potential prognostic benefit. Appropriate risk stratification through noninvasive assessment is important in ensuring patients are not exposed to unnecessary invasive coronary angiograms. The past 20 years have seen an unprecedented expansion in noninvasive imaging modalities for the assessment of stable CAD, with cardiovascular magnetic resonance and computed tomography complementing established techniques such as myocardial perfusion imaging, echocardiography and exercise electrocardiogram. In this review, we examine the current state-of-the-art in noninvasive imaging to provide an up-to-date analysis of current investigation and management options.

Introduction

Cardiovascular disease (CVD) is the leading cause of death worldwide, and despite advances in diagnosis and treatment, coronary artery disease (CAD) remains the number 1 cardiovascular (CV) cause of mortality and morbidity.¹ Given the increasing burden of CAD worldwide, diagnostic techniques for identification of CAD are particularly important. In routine clinical practice, the definitive diagnosis of significant CAD is typically made during invasive coronary angiography (ICA), with significant disease being historically defined as ≥50% stenosis in the left main stem or ≥70% stenosis in other coronary arteries.² The past 20 years have seen a rapid expansion in the number of different noninvasive imaging modalities used for the assessment of stable CAD. In addition to established noninvasive techniques such as the exercise electrocardiogram (exECG), myocardial nuclear perfusion imaging (MPI – most commonly single-photon emission computed tomography [SPECT]) and stress echocardiography, newer techniques such as CV magnetic resonance (CMR) and computed tomography (CT) are likely to have a significant clinical utility in the future. The aim of this review was to discuss the current literature on the efficacy of noninvasive imaging tests compared to ECG and MPI in the diagnosis of stable CAD.

Current recommendations for diagnostic testing in stable CAD

Current European and American guidelines for the assessment of stable CAD^2,3 recommend a common approach. First, an assessment of the pretest probability (PTP) of CAD is made on the basis of clinical history, examination and basic tests such as the ECG. Following this assessment of PTP, a decision on further investigation should be taken. In patients with high PTP (>85%), no further noninvasive testing is required to make the diagnosis of CAD, and ICA can be performed for risk stratification. Conversely, in patients at low risk (PTP < 15%), the likelihood of CAD is so low that consideration should be directed toward non-CAD causes of chest pain. If CAD still needs to be excluded in patients with low PTP, CT coronary angiography (CTCA) has a high negative predictive value and is useful in this group.

Further noninvasive testing for CAD is particularly required in the intermediate-risk group (PTP 15–85%). This group comprises the majority of patients referred to the clinic for evaluation of chest pain, and therefore, imaging plays a key role in risk stratification and management. The main aim of noninvasive testing is to select patients likely to benefit from ICA and revascularization. The Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial demonstrated that although percutaneous coronary intervention (PCI) in addition to optimal medical therapy (OMT) reduced frequency of angina compared to OMT alone, there was no reduction in death, nonfatal myocardial infarction [MI] or hospitalization for acute coronary syndrome (ACS).⁴ This trial, along with evidence from the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI-2D) trial and from meta-analyses,^5–7 has led to the general guideline recommendation that invasive angiography revascularization with PCI should only be performed in patients whose symptoms are refractory to OMT, or in those in whom there is likely to be prognostically significant disease. In addition, however, ICA can also be performed for “softer” reasons, in particular, to provide diagnostic clarity and reassurance.

In light of these studies, the role of noninvasive imaging in guiding selection of patients for ICA has become increasingly important. We now discuss each of these noninvasive techniques in turn. Table 1 summarizes all discussed modalities.

Table 1.

Summary of non-invasive imaging modalities for assessment of suspected stable angina

Modality	Advantages	Disadvantages	Diagnostic accuracy	Ionizing radiation dose
exECG	• Cheap • Easy to perform • Requires little post-processing/analysis • Extensive long-term data	•Relatively poor diagnostic accuracy •Limited by patient’s physical capacity	Sensitivity: 60% Specificity: 76%9	Nil
MPS/SPECT	• Extensive long-term prognostic data • FFR (current invasive gold standard) initially validated against SPECT • Can be performed with exercise, vasodilator or dobutamine stress • Offers functional information • PET/CT has potential to offer anatomical information in addition to functional ischemia assessment	• Radiation • Difficulty in the assessment of balanced ischemia • False positives in LBBB due to partial volume effects	Per vessel47 Sensitivity: 61% (vs. FFR) Specificity: 84% (vs. FFR) Per patient47 Sensitivity: 74% Specificity: 79% (vs. FFR)	7.2 mSv75
Stress echo	• Nonionizing • Extensive long-term data • Relatively fast to perform • Information on LV function and valves • Viability assessment possible with dobutamine	• Exercise limited by patient’s physical capacity • Despite use of contrast views may still be suboptimal • False positives with LBBB • Dobutamine not as physiological as exercise	Per patient47 Sensitivity: 69% Specificity: 84% (vs. FFR)	Nil
CMR	• Nonionizing • Gold standard for noninvasive assessment of ventricle mass and function • Offers viability assessment with the use of LGE imaging • Offers assessment of valves and extra-cardiac structures • Able to assess the heart in any plane	• Expensive • Expertise not yet widely available in all centers • Unable to perform in patients with claustrophobia or renal impairment (eGFR <30 mL/min) • Image quality may be degraded by arrhythmia/tachycardia, breath holding • Relatively time-consuming • MR angiography not yet at the required resolution to offer coronary angiography	Per vessel47 Sensitivity: 87% Specificity: 91% (vs. FFR) Per patient47 Sensitivity: 89% Specificity: 87% (vs. FFR)	Nil
CTCA	• Excellent anatomical detail • Particularly useful as a rule-out test due to high sensitivity and low false-negative rate • Able to offer some information on function and valves • Readily available on most modern CT scanners with minimal upgrading • Also offers extra-cardiac assessment such as “triple rule-out” test	• Radiation (although if optimal is equivalent invasive coronary angiogram) • No ventricular functional assessment unless retrospective gating is used with a consequent increase in radiation dose • Techniques to provide functional assessment of ischemia are not yet ready for routine clinical use (CT perfusion, FFR-CT) • Image quality very dependent on optimization of patient factors (heart rate/rhythm, breath holding) • Caution in patients with renal impairment	Per vessel76 Sensitivity: 91% Specificity: 58% (vs. FFR) Per patient76 Sensitivity: 90% Specificity: 39% (vs. FFR)	1–3 mSv56,78

Note: Median ionizing radiation dose for invasive coronary angiography is 7.4 mSv.⁷⁸

Abbreviations: CMR, cardiovascular magnetic resonance; CT, computed tomography; CTCA, CT coronary angiography; exECG, exercise echocardiography; eGFR, estimated glomerular filtration rate; FFR, fractional flow reserve; LBBB, left bundle branch block; LGE, late gadolinium enhancement; LV, left ventricular; MPS, myocardial perfusion scintigraphy; MR, magnetic resonance; PET, positron emission tomography; SPECT, single-photon emission computed tomography.

exECG

The exECG is one of the most widely studied and utilized techniques for the identification of ischemia. Exercise can be performed using either treadmill or exercise bicycle (supine or upright). The most commonly conducted technique for the exECG is the Bruce protocol.⁸ This involves exercise increased incrementally at 3-minute intervals until the patient reaches a desired target (most commonly heart rate of 85% of predicted heart rate for age) or until the patient can no longer continue. Continuous 12-lead ECG monitoring and noninvasive blood pressure are measured throughout. Typical end points suggestive of the presence of ischemia include ST segment elevation, ST segment depression (downsloping or flat and >2 mm), development of ventricular arrhythmias or reproduction of angina symptoms. exECG has the advantage of being cheap and relatively easy to perform and interpret as well as providing rapidly available clinically useful information which can be used to direct further investigation or reassure patients.

One of the main limitations of exECG is the presence of resting ECG abnormalities such as left bundle branch block (LBBB). Lack of physical fitness also may provide inconclusive results. In addition, there is the potential for induction of ventricular arrhythmias.

exECG performs moderately for the diagnosis of significant CAD. A meta-analysis of 34 studies including more than 3,000 patients by Banerjee et al⁸ reported that a negative test still leads to a diagnosis of CAD in 37% of men and 18% of women. exECG also has a significant amount of prognostic data, with a negative, low-risk test conferring a very good prognosis.^10,11 In addition to the assessment of ischemia, functional capacity on exercise testing also provides important prognostic information, with patients able to exercise to 10 metabolic equivalents (METs) having an extremely low risk of major adverse CV events independent of the amount of ischemia.^12–14

Despite these limitations and its downgrading in recent guidelines, exECG continues to be used in many centers due to its easy availability and clinicians’ confidence with its performance. It is also still important in other settings such as in exercise-induced arrhythmias.

Nuclear imaging

Although encompassing a number of techniques, the most commonly used nuclear imaging technique is SPECT. This technique can be performed using either exercise or pharmacological methods (using vasodilators such as dipyridamole) and relies upon the uptake of radioactive tracers into the myocardium at rest and stress. The most commonly used tracers are thallium and technetium. The heart is scanned using a gamma camera to assess the perfusion of the tracer into the myocardium. A positive test for inducible ischemia is obtained when there is a relative reduction in uptake of the tracer at stress due to the reduction in blood flow to the ischemic territory compared to normal tissue which becomes hyperemic. SPECT is also able to provide information on viability as areas of previous MI which are nonviable will display a fixed perfusion defect with a reduction in tracer uptake at both rest and stress. SPECT is also able to provide an estimate of left ventricular function.

One limitation of SPECT is in the assessment of “balanced ischemia.”¹⁵ Because interpretation of the study relies on comparison of ischemic areas to normal areas, in patients (such as those with significant triple vessel disease) where there is widespread ischemia throughout the myocardium, there is the possibility of incorrectly reporting a scan as showing no inducible ischemia when in fact all areas are relatively underperfused during stress, thus causing a false-negative study.

A further limitation is caused by the relatively high radiation dose, which most recently has been reported as around 7 mSv.¹⁶ Furthermore, many of the commonly used tracers take some time to leave the myocardium. Given this, many protocols involve the initial acquisition of stress images, with rest images taken at a later interval (depending on the tracer this can be up to 1 week later), thus meaning that results can be delayed as well as the study involving two visits for the patient. SPECT is also affected by LBBB. Because the septum is dyskinetic and contracts at a slightly different time to the lateral wall, the septum may not thicken at the time of image acquisition. Because the septum appears thinner, tracer uptake in the area may appear less, giving the impression of left anterior descending artery (LAD) territory ischemia. This is less of an issue using vasodilators rather than exercise however.¹⁷

Despite these limitations, there is a wealth of diagnostic and prognostic information on the use of SPECT for the assessment of stable CAD.

One important study on the prognosis of stable CAD patients using SPECT was the COURAGE nuclear substudy.¹⁸ This was an observational substudy of the COURAGE trial in which 314 patients underwent SPECT imaging at baseline and at 6–18 months post randomization. The study found that patients who underwent PCI + OMT had a greater reduction in ischemia than those who had OMT alone. There was also a trend to improved outcomes, with patients who had no residual ischemia having no clinical events. This substudy was underpowered for clinical events, however, and so the question of ischemia burden and using PCI to improve prognosis is still uncertain.

Another nuclear imaging modality that has recently been studied for use in stable CAD is positron emission tomography (PET). One particular advantage of this technique is that it gives the ability to measure absolute myocardial blood flow, thus avoiding problems with balanced ischemia.¹⁹ Coronary flow reserve is then calculated from the difference in blood flow in stress and rest.²⁰ In addition, it can be combined with CT to provide both functional and anatomical assessment of CAD. The most common PET tracers are rubidium-82 and ¹³N-labeled ammonia, although various other tracers can be used. A meta-analysis by Mc Ardle et al²⁰ reported an increased diagnostic accuracy compared to SPECT for diagnosis in patients with suspected CAD, with a sensitivity of 90% and specificity of 95%. There have also been some recent data on prognosis using PET, with several studies reporting a significant increase in adverse events in patients with reduced coronary flow reserve.^22–24 PET using rubidium-82 has been shown to provide incremental prognostic value above clinical variables in a series of 1,432 patients undergoing scans, with an increasing burden of ischemia being associated with worse outcome.²⁵

Despite these benefits, PET remains predominantly a research tool, with availability and expertise the predominant limitations to more widespread clinical use.

Stress echocardiography

Alongside SPECT, stress echocardiography is the most commonly used technique for the assessment of stable CAD. It can be performed using either exercise (treadmill or bicycle) or using pharmacological stress with dobutamine. Bicycle exercise has the advantage of allowing acquisition of images at peak exercise if performed while supine, whereas cycling in the upright position or on a treadmill requires the patient to be repositioned to an optimal position to obtain adequate echo images. Continuous 12-lead ECG monitoring is usually also performed. Of course, echocardiography remains the “workhorse” of cardiology, providing rapid assessment of ventricular volumes, function and assessment of valves. During stress echo, in ischemic segments, the wall motion abnormalities develop at peak exercise which are not present at rest (with wall motion abnormalities present at rest which do not improve on exercise suggestive of nonviable tissue). In addition, hibernating myocardium is associated with an improvement in wall motion on exercise. Exercise echocardiograph also provides a physiologically accurate assessment of the effects of exercise on cardiac function. This may be particularly important in the setting of associated dyspnea, which may be caused by exercise-induced or exacerbated mitral regurgitation, left ventricular outflow tract obstruction or pulmonary hypertension.

Dobutamine stress echocardiography (DSE) can also be performed. This has the advantage of being able to be used in patients unable to exercise. Unfortunately, some of the physiological assessment benefits (such as in exercise-induced pulmonary hypertension) are more difficult to interpret when dobutamine is used. Typical protocols involve the use of “high-dose” dobutamine (as opposed to “low dose” in patients with aortic stenosis), with increasing increments of dobutamine up to a usual maximum of 40 μg/kg/min, with the additional use of atropine if target heart rate is not reached. One advantage of DSE is the assessment of viability, which is more validated than with exercise echo. This relies on a bimodal change in wall motion, with ischemic but viable segments of myocardium typically initially improving in function at low doses of dobutamine (10–20 μg/kg/min) before function worsens again at high doses. This is because at low doses dobutamine initially acts as an inotropic agent, improving myocardial contractility, whereas at high doses its chronotropic and vasodilator effects are more predominant. In contrast, in nonviable tissue, wall motion remains impaired at all doses.²⁶

Stress echo performs reasonably for the diagnosis of CAD, with all techniques performing similarly. Sensitivities are 85%, 80% and 78% and specificities are 77%, 86% and 91% for exercise, dobutamine and dipyridamole stress, respectively.²⁷ While these values are not too different to SPECT,²⁸ they are significantly better than exECG.²⁹ In terms of prognosis, a normal stress echo has been associated with a mortality rate of around 1% per year in several large observational studies.^30–33

Of course, there are limitations with stress echocardiography. First, as with resting echocardiography, scan quality can be limited by the patient body habitus and inability to obtain suitable windows. This difficulty is often exacerbated at the high heart rates caused by exercise. Endocardial border definition can be improved by the addition of left ventricular contrast, and many centers use this routinely. Contrast can also be used to provide an assessment of myocardial perfusion, although this is not typically used in standard clinical practice.^34–36 Nevertheless, in some patients, despite considerable time and effort, adequate images are not available. A further limitation with the use of dobutamine is the potential for ventricular arrhythmias; thus, DSE is usually performed with a clinician and resuscitation facilities available.

Another limitation with stress echocardiography is that analysis is often qualitative, with a reviewer making a judgment call as to the presence and severity of any wall motion abnormality. Quantitative measures of wall motion using speckle tracking have been tested; however, their utility is still to be confirmed, particularly in the posterior circulation. Overall, speckle tracking does not appear to improve diagnostic accuracy sufficiently to mandate its incorporation into routine clinical practice.^37,38

Three-dimensional (3D) echocardiography has some theoretical benefits, including faster acquisition (ability to obtain multiple planes in one scan) and more accurate assessment of ventricular volumes and function. 3D echo may be more accurate than 2D, particularly in the LAD territory.^39,40 Again, there are no large trials to strongly support its routine use, and it still remains predominantly a technique used in specialist centers.

CMR

CMR has become increasingly utilized in clinical cardiology over the past 2 decades. CMR has a particular advantage in that it is able to offer noninvasive imaging in any plane without any limitations due to lungs or ribs and avoiding nonionizing radiation. The most useful technique for the assessment of stable CAD is stress CMR. Although there have been reports of exercise stress,^41,42 stress CMR is traditionally performed using pharmacological stress. Both vasodilator (typically using adenosine or regadenoson) and dobutamine stress have been used. Vasodilator methods involve the administration of intravenous gadolinium contrast during vasodilator infusion, with the presence of perfusion defects (seen as subendocardial hypointense areas) in ischemic territories during peak hyperemia, typically defined as a 15 mmHg drop in systolic blood pressure and an increase in heart rate. Dobutamine stress CMR, on the other hand, uses a similar principle to that of DSE, relying on the demonstration of wall motion abnormalities in the presence of ischemia.⁴³ Imaging of perfusion is commonly performed in three short-axis planes to obtain a 16-segment model, while wall motion abnormalities using dobutamine can also be imaged in long-axis planes.

In routine clinical practice, the assessment of perfusion defects by CMR is most commonly qualitative, with visual assessment by the operator of the presence and extent of ischemia. More recently, several groups have conducted research on quantitative measures of perfusion. These are either semi-quantitative, using the difference in signal intensity between areas of the myocardium, or fully quantitative, measuring absolute blood flow using mathematical modeling.⁴⁴ The use of qualitative measures appears to improve diagnostic accuracy; however, it does require further post-processing which makes its routine clinical use less attractive currently.⁴⁵

CMR stress perfusion appears to have better diagnostic accuracy than stress echo and SPECT with a meta-analysis by Jaarsma et al⁴⁶ reporting a sensitivity of 89% and specificity of 76% compared to ICA with visual assessment and a more recent meta-analysis by Takx et al⁴⁷ reporting a sensitivity of 89% and specificity of 87% at the per-patient level compared to fractional flow reserve (FFR). There have also been a number of randomized controlled trials performed comparing CMR and SPECT. The Magnetic Resonance Imaging for Myocardial Perfusion Assessment in Coronary artery disease Trial (MR-IMPACT) trial compared adenosine stress CMR vs. SPECT in 234 patients in 18 centers. The authors found that CMR performed better than SPECT when an optimal gadolinium dose (0.1 mmol/kg) was used, suggesting at least that it was a viable alternative to SPECT.⁴⁸

The CMR and single-photon emission CT for the diagnosis of coronary heart disease (CE-MARC trial) by Greenwood et al⁴⁸ was a trial conducted in two UK centers to determine the diagnostic accuracy of CMR (including adenosine stress, magnetic resonance [MR] angiography and late gadolinium enhancement [LGE] imaging at 1.5 T) vs. SPECT in the diagnosis of stable angina with a gold standard of ICA. In this trial, 752 patients judged to be at intermediate risk of CAD were randomized to CMR or SPECT. The authors again found that CMR had superior diagnostic accuracy to SPECT (CMR: sensitivity 86.5%, specificity 83.4%; SPECT: sensitivity 66.5%, specificity 82.6%), although they did find that MR angiography did not add any significant diagnostic benefit. These results were replicated in the multicenter, multi-vendor MR-IMPACT II, which again showed that adenosine stress CMR was superior to SPECT.⁵⁰

While these studies suggested that CMR was at least as good, if not better than SPECT, there was still some doubt as to whether the cost and expertise required were worth investing in. The CE-MARC 2 trial, which randomized patients to a CMR-guided strategy, SPECT or UK National Institute for Health and Care Excellence guideline recommendation found that CMR was equivalent to SPECT and that both strategies reduced unnecessary angiography, the primary outcome of the trial.⁵¹

One important factor in these large randomized trials was the use of ICA as the “gold standard.” As discussed earlier, FFR assessment of ischemia is now accepted as the optimal measure of coronary lesion significance. The MR-INFORM trial was a multicenter trial comparing adenosine stress CMR at 1.5 T vs. FFR in patients with suspected stable angina.⁵² The study showed that CMR was non-inferior to FFR, with both conferring ≤4% risk of major adverse cardiovascular events within 1 year, providing some prognostic data. In addition, only 37 out of 454 patients (8.1%) in the CMR-guided arm had a negative invasive angiogram (48.8% had a negative CMR and did not proceed to invasive testing). The full results of the MR-INFORM study are yet to be reported; however, these preliminary results suggest that adenosine stress CMR could well be a useful technique in the setting of stable angina, with some strong prognostic data also.

Beyond the improvement in imaging quality compared to echocardiography,⁵³ CMR has several other advantages. First, CMR is currently accepted as the noninvasive gold standard for the assessment of cardiac structure and function. This allows clearer assessment of wall motion abnormalities as well as providing further diagnostic information. Second, the use of gadolinium contrast also allows the assessment of myocardial fibrosis and scarring. Imaging performed 10–15 minutes after intravenous contrast injection, known as LGE imaging, can demonstrate areas of myocardial infarct scar. The presence of an infarct-type pattern can help in the differentiation between ischemic and non-ischemic cardiomyopathy, while the transmural extent of the infarct scar can be used to determine myocardial viability.⁵⁴

Nevertheless, CMR has some limitations. The technique itself is relatively time-consuming, expensive and not readily available in all centers. In addition, not all patients are able to tolerate the scan, either due to claustrophobia or inability to breath hold adequately. Gadolinium contrast should be used with caution and only when strictly necessary in patients with an estimated glomerular filtration rate of <30 mL/min due to the risk of nephrogenic systemic fibrosis. Finally, image quality may be degraded by arrhythmia, although techniques such as real-time imaging can be used to mitigate this.

Cardiac CT (CCT)

Although CCT is a noninvasive imaging technique, it differs from all of the previously discussed techniques as it is a predominantly anatomical technique. CCT commonly refers to two different modalities – calcium scoring and CTCA. Both are carried out using ECG gating to minimize cardiac motion artifact, with images usually acquired in diastole where the heart is at its most still. In addition, breath holding is used to minimize respiratory motion artifact.

CT calcium scoring is performed without contrast. With modern 64-slice scanners, imaging is usually acquired in one breath hold. The noncontrast study is analyzed using rapid post-processing software to measure the amount of coronary calcium. The most commonly used scoring method is the Agatston score which takes into account the size and radiographic density of the plaque to provide a score which reflects the coronary calcified plaque burden.⁵⁵ The presence of coronary artery calcification does reflect the burden of CV risk; however, not all significant obstructive coronary lesions are calcified; in contrast, heavily calcified coronary arteries can often be non-obstructed, hence calcium scoring alone cannot be used as a diagnostic tool in stable angina. Indeed, calcium scoring has fallen out of favor in recent guidelines.²

Currently, CTCA is the preferred method of CCT imaging. Current-generation 64-slice CT scanners have reduced concerns about radiation dose – mean radiation dose for one CCT is 1–2 mSv with appropriate scan optimization.⁵⁶ Attaining these doses requires the patients to have slow (<60 bpm), regular heart rates. This can be achieved by oral and intravenous beta-blockade if necessary (calcium channel blockers and ivabradine can also be used). A bolus of contrast is injected intravenously, and using ECG gating, a coronary angiogram is obtained. CTCA provides excellent anatomical detail and has shown excellent diagnostic accuracy in comparison to ICA. CTCA is recommended for use in patients with low-to-intermediate risk of CAD. This is predominantly because of its high sensitivity (95–99%) and negative predictive value (97–99%) but low specificity (64–83%).^57,58

Beyond its low specificity, ionizing radiation dose and scan preparation, CTCA does have some other limitations. Scan accuracy is reduced in the presence of heavy calcification, which can result in “blooming” artifact, which can make it difficult to interpret the study.⁵⁹ In addition, CTCA is not as accurate in the diagnosis of in-stent restenosis, particularly in stents <3 mm.⁶⁰ Because of this, CTCA is currently only recommended in low–intermediate-risk patients in the current guidelines, predominantly as a “rule-out” technique. Finally, CTCA is an anatomical technique, providing no functional information on stenosis severity. In the context of what is known about ICA, this does pose an issue. The Fractional Flow Reserve vs. Angiography for Guiding Percutaneous Coronary Intervention (FAME) trial demonstrated that use of FFR, an invasive measure of ischemia, leads to superior outcomes compared to visual stenosis assessment for guiding intervention.⁶¹ This, in combination with the COURAGE trial, has led to an acceptance within cardiology that revascularization in stable angina should preferentially be performed in functionally significant lesions only.⁶²

Outcome data are also now emerging. The recent Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) trial was designed to compare outcomes in patients with CT vs. functional testing.⁶³ In this trial of 10,003 intermediate-risk patients, CTCA was associated with significantly fewer invasive angiograms with no significant CAD; however, patients in the CTCA arm were more likely to undergo ICA within 90 days. Clinical outcomes were not different between the two cohorts. The SCOT-HEART trial, which randomized 4,146 patients with suspected angina to CTCA or standard care, revealed that CTCA changed management in many cases; however, it did not have any impact on readmission rates.⁶⁴ There was also a reduction in MI with CTCA (26 vs. 42, p = 0.053).⁶⁵ Several observational studies have also reported the prognostic significance of CAD identified on CTCA.^66–68

Recently, two more novel techniques have been used to add functional assessment to CTCA and overcome this limitation. The first is CT perfusion imaging. This uses a similar principle to CMR perfusion, relying on the passage of (iodinated rather than gadolinium) contrast from the myocardial blood pool into the myocardium. Areas of infarction or ischemia appear hypointense compared to normal myocardium. Again, imaging can be performed at rest and stress to determine the functional significance of lesions.⁶⁹ CT perfusion requires optimal timing, as continuous imaging is not performed to reduce radiation dose, unlike in CMR. Diagnostic accuracy appears to be reasonable; however, due to limited availability and radiation dose concerns it has not yet been adopted into routine clinical practice.⁴⁷

Another promising technique that is under investigation is CT fractional flow reserve (CTFFR). This technique uses computational fluid dynamics to provide a prediction of the invasive FFR. One advantage of this technique is that it requires no further imaging beyond a standard CTCA and also does not involve administration of adenosine to induce hyperemia. In the most studied technique, CT scans are sent to a CTFFR company, which provides an FFR in all major vessels at any point in the coronary tree within 24 hours. The methods for measuring CTFFR are however proprietary and somewhat time-consuming, requiring intense computational fluid dynamics and computer modeling. In addition, there may be some reservations regarding sending studies to a separate company. Several companies however are working on CTFFR software which will be available on workstations to all clinicians. CTFFR has shown good reproducibility and diagnostic accuracy in several randomized trials. The Diagnosis of Ischemia-Causing Stenoses Obtained Via Noninvasive Fractional Flow Reserve (DISCOVER-FLOW) trial was the first proof of concept trial using a CTFFR in patients with known or suspected angina, reporting a diagnostic accuracy of 84.3% per vessel, and 87.4% per patient, with invasive FFR as the gold standard. Importantly, these were significantly improved compared to standard CTCA (58.5% and 61.2%, respectively).⁷⁰ These promising results were also replicated in two further trials.^71,72 The PLATFORM trial used CTFFR as part of a strategy comparing CTCA with standard care and found that CTCA led to reduced referrals for ICA and also had similar clinical outcomes at 1 year and lower cost than usual care.^73,74 Further data are required however for this technique to reach routine clinical use.

Conclusion

Overall, a recent meta-analysis by Takx et al⁴⁷ suggested that in comparison to FFR, CMR perfusion, PET and CT perfusion were better techniques for diagnosing CAD at both vessel and patient level than SPECT and stress echocardiography. While SPECT and stress echo are both relatively cheap and widely available, as expertise is gained and more prognostic data become available, CMR, CT and PET may become more attractive for routine clinical practice. CT is quick and particularly useful in ruling out obstructive CAD, particularly in patients with low PTP; however, CMR and PET provide more information on myocardial structure and function. Figures 1 and 2 show a proposed management algorithm for the investigation of stable angina. This of course would need to be tailored dependent on the patient and the expertise available in the hospital. The ongoing ISCHEMIA trial (NCT01471522), which in contrast to both COURAGE and BARI-2D, will randomize patients to OMT or an invasive strategy prior to ICA, but after noninvasive imaging has been performed will also provide further insights into the role of noninvasive imaging in stable CAD. While none of these techniques is likely to be the sole one used, in correctly identified appropriate patients, they will undoubtedly provide useful information with which to guide management.

Figure 1.

Proposed investigation algorithm for patients with suspected angina focusing on patients with low or high PTP of CAD.

Abbreviations: CAD, coronary artery disease; CTCA, computed tomography coronary angiography; ECG, echocardiography; ICA, invasive coronary angiography; PTP, pretest probability.

Figure 2.

Proposed investigation algorithm for patients with suspected angina and intermediate probability of CAD in an “ideal” hospital with all modalities available.

Abbreviations: CAD, coronary artery disease; CTCA, computed tomography coronary angiography; CTFFR, computed tomography fractional flow reserve; CMR, cardiovascular magnetic resonance; ECG, electrocardiogram; exECG, exercise ECG; SPECT, single-photon emission computed tomography.