Efficacy of Diagnostic Testing of Suspected Coronary Artery Disease: A Contemporary Review

Coronary artery disease (CAD) affects over 20 million Americans, and its spectrum of impact leads to an estimated 7 million deaths, as well as the loss of 129 million disability-adjusted life years, worldwide, each year [1‒3]. Meanwhile, chest pain accounts for almost 14 million office and emergency room visits each year in the USA [4, 5], and over 10 million noninvasive cardiac tests are ordered to evaluate for CAD [6‒8]. CAD develops as atherosclerotic plaque forms in the coronary arteries. These plaques can eventually limit blood flow to myocardial tissue resulting in ischemia and are at risk for acute plaque rupture and acute coronary syndrome [9‒11]. While chest pain may represent a wide array of underlying diseases, given the high morbidity and mortality associated with CAD, an ischemic cardiac etiology must always be considered. Early diagnosis and treatment of CAD can improve patient outcomes through guiding risk factor modification and treatment modalities including medical and invasive approaches [12]. While discovering coronary disease early can allow for treatments which can yield great benefit [13], many tests are equivocal and can be associated with additional risk and unnecessary cost [14]. This leaves the question; if there is some concern for stable CAD, who should we test, and which modality should be used?

A patient presenting with acute, ongoing chest pain should be sent to the emergency department to rule out life-threatening pathology, but further evaluation and management of acute chest pain will not be detailed here. For patients with intermittent chest pain, evaluation for underlying CAD may be indicated. The latest clinical practice guidelines for the evaluation of stable CAD from the European Society of Cardiology (ESC) and American Heart Association/American College of Cardiology (AHA/ACC) are summarized in Table 1. Patients with symptoms consistent with stable CAD may be evaluated with a multitude of screening methods, from simple electrocardiograms (EKG) to more advanced modalities, such as myocardial perfusion imaging (MPI) and coronary computed tomography angiography (CCTA) [15]. Both the 2019 ESC and 2021 AHA/ACC guidelines for the evaluation and diagnosis of chest pain advise using a patient-centered and risk-based method, emphasizing the pretest probability of CAD to determine the benefit of further testing [16]. As shown in Figure 1, patients may be risk stratified using the history and physical, their demographics, and comorbidities, as well as by assessing for other risk factors such as tobacco use and family history of CAD [17‒19]. Risk stratification has gone through many iterations including the original Diamond and Forrester model to describe chest pain symptoms. Patients having “cardiac chest pain” experience substernal chest pain which is exacerbated by exertion and relieved with rest or nitrates [17]. This constellation of symptoms is associated with a higher pretest probability of CAD. Additionally, older age, a strong family history, tobacco use, and comorbidities such as hypertension, hyperlipidemia, and diabetes increase the pretest probability of CAD as seen in the Genders model used in the 2013 ESC guidelines [19]. However, it is worth noting that more recent studies indicate these historical scoring models overpredict CAD prevalence and, therefore, pretest probability [20, 21]. Due to this overestimation, the 2019 ESC and 2021 AHA/ACC guidelines use an updated model including a pooled analysis of recent studies to more accurately predict these important metrics [18].

After determining a patient’s pretest probability, physicians must determine the next best steps for screening. In the latest AHA/ACC guidelines, the direct pathway to invasive coronary angiography (ICA) for those with a high pretest probability has been removed, and now the recommendation is for an anatomic, or functional evaluation before possible ICA in the workup of patients with suspected stable CAD [16]. The ESC guidelines, however, continue to suggest that those in the highest risk categories warrant direct ICA [22]. While those with low pretest probability (≤15% in ESC and AHA/ACC guidelines) generally warrant no further cardiac testing, both AHA/ACC and ESC guidelines recommend obtaining additional data points such as coronary artery calcium (CAC) scores, resting EKG, and resting echocardiogram to augment the risk stratification and pretest probability to aid in decision-making. For patients with intermediate to high levels of risk (15–85% pretest probability), screening is recommended with numerous additional options to choose from [16, 22].

Testing modalities may diagnose CAD via direct visualization of the coronary arteries (anatomic testing) or by demonstrating the pathological effects of flow-limiting lesions on cardiac function (functional testing). Anatomic testing includes CCTA and ICA. These modalities allow for more accurate estimation of the degree of coronary occlusion. Nevertheless, these tests may often require augmentation with additional hemodynamic assessments such as fractional flow reserve (FFR) to better assess for significant ischemia from the visualized disease [23]. Under normal conditions, coronary arteries can vasodilate to accommodate increased blood flow during exertion. When an obstructive lesion is present, exertion can cause transient ischemia due to a relative inability to augment coronary blood flow secondary to flow limitation [24]. Functional testing includes exercise EKG, stress echocardiogram, and nuclear stress testing and uses this concept to detect dynamic blood flow-limiting plaques due to their impact on downstream cardiac function [25]. Functional testing cannot visualize lesions directly and, therefore, may not adequately report subclinical CAD [26]. Physicians must balance patient characteristics, the cost of testing, and the strengths and weaknesses of each modality to choose the optimal test for each patient. It is also important to consider the institutional variation in available modalities and their competency with each test. This comprehensive review aimed to describe the available CAD testing modalities (summarized in Table 2), detail their risks and benefits, and describe when each should be considered in the evaluation of a patient with suspected CAD.

The EKG was the first diagnostic test utilized in the evaluation of ischemic CAD and still has a role in evaluating patients, both in the acute and chronic setting. In a patient experiencing acute chest pain, ST-segment elevations or depressions can aid with the diagnosis of acute coronary syndrome. In the absence of acute chest pain, changes such as Q waves, a left bundle branch block, or T wave changes can be indicative of past coronary events or other underlying cardiac disease and can aid in risk stratification. One meta-analysis found that when resting EKG changes were added to existing risk models, there was a minor improvement in AUC/C statistics of 0.001–0.05 [27]. While the resting EKG is easy to obtain, has no contraindications, and can be helpful in risk stratification, it does not offer significant benefit in the diagnosis of CAD. The sensitivity has been shown to be as low as 23% with specificity of 87% [28]. Furthermore, the AHA/ACC and ESC guidelines do not list this as an appropriate method for ruling out underlying CAD (class 3; LOE C) but rather recommend its widespread use to aid in screening and risk stratification (ESC class 1; LOE C, AHA/ACC class I; LOE B) [16, 22].

The resting echocardiogram (echo) is an important tool in the assessment of individuals with chest pain. It offers an evaluation for structural defects and acute pathologies, such as aortic dissection, pericardial effusion, and acute valvular abnormalities [16]. Traditional resting echocardiography can also provide a direct assessment of the left and right ventricular function and evaluate for regional wall motion abnormalities (RWMA), which can be invaluable for prompting further workup to diagnose CAD. Despite these benefits, it alone does little to reliably diagnose stable CAD [29]. However, advanced imaging techniques, such as resting 2-dimensional speckle tracking echocardiography, which assesses for global longitudinal strain, have been shown to be an independent predictor of CAD, with global longitudinal strain > −18% conferring a 91.1% sensitivity and 63% specificity for the detection of significant CAD. Furthermore, regional longitudinal strain can provide information which can assist in localizing the individual epicardial coronary artery which might be affected [30, 31]. These techniques have also been shown to provide added accuracy to exercise testing, which suggests they are an effective adjunct modality for the diagnosis of CAD [30, 31]. Current AHA/ACC guidelines recommend a bedside resting echocardiogram for patients with stable chest pain to assist with risk stratification (class 1; LOE B). This is especially important in high-risk individuals who have evidence of electrocardiographic Q waves, heart failure symptoms, complex ventricular arrhythmias, or a murmur [16]. The ESC guidelines also recommend a resting echocardiogram as a part of the initial diagnostic workup of patients with suspected CAD (class 1; LOE B) in order to assist with risk stratification and to exclude alternative causes of angina, identify RWMA suggestive of CAD, and measure systolic function [32]. No guidelines currently promote the use of resting echocardiography as the exclusive testing modality for the diagnosis of stable CAD, although future studies utilizing advanced imaging techniques may change this paradigm.

CAC scoring is a noninvasive modality for assessing calcium deposition in coronary arteries, which is a highly specific feature of coronary atherosclerosis [33]. A typical CAC score can be obtained within 10–15 min via computed tomography (CT) chest, only subjecting patients to approximately 1 mSv of radiation, and does not require the use of contrast agents [33]. The outcome of CAC testing is a numerical value (the Agatston score), calculated based on the volume and density of the calcifications. This score, combined with other risk factors, can be used to make decisions regarding the initiation of medications, such as statins or aspirin [29]. In the setting of stable chest pain and suspected CAD, a nonzero CAC score has a sensitivity of 98% and specificity of 40% for the detection of significant CAD on ICA [34]. The strength of CAC in the setting of suspected CAD comes from its high negative predictive value. In a 10,000 person contemporary registry based study of patients with suspected CAD, among those with a CAC score of 0, only 1.8% and 1.0% of individuals had obstructive CAD on CCTA and ICA, respectively, with a negative predictive value of 98.2% [35]. Individuals with CAC score of 0 have a <5% prevalence of obstructive CAD and very low risk of death or nonfatal myocardial infarction (MI) (<1% annual risk) [32]. Patients with a low pretest probability of CAD and CAC scores of 0 may not require additional diagnostic testing [16]. Limitations of CAC include reduced utility in individuals <45 years as in those patients, 17.7% of individuals with a CAC of 0 had obstructive stenoses, suggesting that CAC score alone is insufficient to rule out CAD in this population [35]. This diagnostic uncertainty is possibly because CAC cannot exclude stenosis caused by noncalcified plaques, which may be more prevalent in younger populations [36]. The AHA/ACC guidelines recommend that patients with stable chest pain and no prior history of CAD, who are categorized as low risk, may undergo CAC score as a first-line test to exclude calcified plaque and identify individuals with a low likelihood of obstructive CAD (class 2; LOE B) [16]. The ESC guidelines suggest that CAC scoring may be useful in improving risk stratification but fail to provide specific guidelines or recommendations for the use of CAC scoring [32].

ICA is the gold standard for diagnosis of CAD and can also offer therapeutic benefit for patients with suspected stable CAD [37]. ICA is accomplished through the selective injection of radiocontrast dye into the coronary ostia, under fluoroscopy, to evaluate for possible obstruction [38]. Orthogonal views allow for a 3D direct visualization of the coronary anatomy; therefore, ICA offers excellent anatomic evaluation for CAD [16]. During ICA, the anatomic evaluation may be augmented with functional measures such as FFR to evaluate lesions for significance. FFR is derived by dividing pressure distal to a lesion by pressure proximal to a lesion. These pressure measurements are obtained in hyperemia with pressure transducing catheters by giving adenosine and allow physicians to determine if a lesion is significantly limiting coronary blood flow and leading to ischemia. ICA continues to be the gold standard given its unique ability to diagnose significant CAD through both an anatomic (≥70% obstruction or ≥50% obstruction in left main coronary artery) and functional evaluation (FFR ≤0.80 and/or instantaneous wave-free ratio <0.89) [32, 38, 39].

As ICA has traditionally been considered the gold standard for diagnosing CAD, its sensitivity and specificity are not reported as with noninvasive tests because it is used as the reference against which other tests are compared [40, 41]. While ICA has great diagnostic and therapeutic utility, it may also be the most disadvantageous diagnostic modality due to high periprocedural risk of complications. The incidence of major complications during ICA ranges from 0.2% to 3.2% and includes pathology such as access site hematomas, arteriovenous fistulas, distal limb ischemia, coronary dissection, retroperitoneal bleeding, and cerebrovascular accident such as stroke [42‒44]. Furthermore, ICA is expensive, and exposes the patient to a significant radiation (4–10 mSv) and contrast load [43]. Despite its stellar performance, these risks lead physicians to consider other methods of diagnosis prior to ICA. Contraindications to ICA include inability to tolerate contrast due to allergy or kidney disease, and peripheral vascular disease limiting ability to pass diagnostic catheters. Current ESC guidelines for the diagnosis of suspected stable CAD recommend ICA in patients who have received other testing which proved equivocal (class 2; LOE A) or who have refractory symptoms with minimal exertion despite maximal medical therapy (class I; LOE B) [32]. The AHA/ACC guidelines recommend ICA (along with CCTA) in patients with previously negative stress testing that have stable chest pain and persistently high clinical suspicion of CAD (class 2; LOE B) [16].

CCTA is a noninvasive imaging modality used to anatomically assess both the extent and severity of coronary artery lesions, both nonobstructive and obstructive [16]. CCTA testing uses between 3 and 5 mSv of radiation and requires the use of contrast agents [45]. It boasts an overall sensitivity of 95–97%, and specificity of 78–79%, for the detection of anatomic CAD [46, 47]. Large meta-analyses have shown that when compared to exercise EKG and MPI stress testing, CCTA used for individuals with stable chest pain is associated with a reduced rate of MIs and an increased incidence of coronary revascularization. Furthermore, the Scottish Computed Tomography of the HEART Trial (SCOT-HEART) showed a reduction in cardiovascular deaths when CCTA was added to routine testing modalities such as MPI and exercise EKG [48, 49]. In patients with stable chest pain who were referred for ICA, CCTA allowed 77% of patients to avoid receiving ICA and lead to a reduction in revascularization and stroke [50]. Given these trials, CCTA is increasingly being recognized as a possible “gatekeeper” to prevent unnecessary invasive testing [34]. Additionally, the Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) trial showed that a CCTA which showed no stenosis or plaques yielded a 3-year CAD event rate of only 0.9%, compared to 2.1%, for a negative stress test [30]. Given this predictive value, CCTA can aid in management, as current guidelines suggest further testing with ICA when a CCTA shows >50% stenosis of the left main coronary artery, or severe stenosis (>70%) in the left anterior descending, circumflex, and right coronary artery [16]. Limitations to CCTA involve the spatial resolution of current CT scanners, which limits the detection of plaques to only those greater than 0.5–1.0 mm [51]. Image quality is also affected by the presence of stents due to artifact, and extensive calcification can lead to an overestimation of stenosis severity and thus lead to false-positive results [33]. CCTA alone does not provide a functional assessment of the consequences of stenosis and, as such, cannot directly evaluate if a stenotic lesion is the cause of clinically significant ischemia [52]. CCTA is contraindicated in patients who cannot tolerate contrast, have uncontrolled tachyarrhythmia, or are unable to receive medications such as beta blockers to assist with image acquisition [16]. AHA/ACC guidelines support using CCTA in intermediate-high-risk patients with stable chest pain for the diagnosis of CAD, further risk stratification, and to guide treatment decisions (class 1; LOE A) [16]. CCTA is also recommended in individuals who have intermediate-high-risk with stable chest pain and an abnormal or inconclusive exercise EKG or stress imaging test (class 2a level B), and in stable chest pain with other negative testing (class 2b; LOE C) [16]. ESC guidelines recommend using CCTA as the initial diagnostic test for CAD in symptomatic patients where obstructive CAD cannot be ruled out with clinical judgment alone (class 1; LOE B) [53].

While CCTA has historically failed to provide functional information regarding visualized lesions, CT fractional flow reserve (CT-FFR) aims to offer a solution to this problem. CT-FFR utilizes CT imaging, in conjunction with machine learning and computational fluid dynamics, to provide noninvasive pressure gradients throughout the coronary vessels [54]. This mimics invasive FFR as detailed previously, providing a functional assessment of blood flow and ischemia distal to stenotic lesions [54]. CT-FFR sensitivity ranges from 85% to 90% with a specificity between 82% and 86% and has been shown to be superior to CCTA alone in the identification of functionally significant, ischemia-inducing lesions [53, 55]. CT-FFR correlates well with ICA FFR for FFR_CT values below 0.60 (86.4%), and above 0.80 (87.3%), with worse diagnostic accuracy (<80%) for FFR_CT values between 0.60 and 0.80 [55]. CT-FFR also has a strong prognostic value, especially for individuals with intermediate risk. Amongst patients with stable CAD, FFR_CT > 0.80 was associated with a significant reduction in all-cause mortality and MI at 12 months when compared to those with FFR_CT < 0.80, with an increasing risk for all-cause mortality and MI at 12 months with each reduction in FFR_CT of 0.10 [56]. Individuals with a stenosis between 30 and 70% on CCTA and an FFR_CT > 0.80 had a similar incidence of all-cause death, MI, hospitalization for unstable angina, and unplanned revascularization at 24 months (3.9%), when compared to individuals with CCTA stenosis <30% [57]. In this same study, an FFR_CT < 0.80 was associated with a cumulative incidence of 9.4% for these same outcomes, suggesting an FFR_CT < 0.80 could represent a trigger to prompt further invasive testing [57]. CCTA combined with CT-FFR has been shown to reduce revascularization procedures without an increased risk of major adverse cardiovascular events (MACE) at 1 year [58]. For individuals with extensive coronary lesions (left main or three-vessel disease), CCTA augmented by CT-FFR was found to be non-inferior to ICA both for decision-making, as well as for the identification of targets for revascularization [59]. Limitations of CT-FFR are like that of CCTA, including suboptimal imaging in the presence of heavy calcifications or motion artifacts, the need for intravenous contrast, radiation exposure, and reduced availability [60]. CT-FFR has the same contraindications as in CCTA. The AHA/ACC guidelines recommend CT-FFR for intermediate-high-risk individuals with stable chest pain and 40–99% stenosis on a proximal or middle coronary segment on CCTA (class 2a; LOE B). In this population, CT-FFR can assist with the diagnosis of vessel-specific ischemia and guide decisions on revascularization [16]. The ESC guidelines do not provide specific guidance on the use of CT-FFR for the diagnosis and management of stable CAD.

The exercise EKG has long been a staple in the testing for CAD, but its role may be changing due to newer, more sensitive, and specific methods of testing. Exercise EKG testing offers a functional evaluation for CAD using increasing levels of physical activity and a continuous 12-lead EKG [61]. The patient walks on a treadmill or pedals a stationary bike at increasingly strenuous levels to reach 85% of their age-predicted maximal heart rate and induce dynamic ischemia [62, 63]. The test is positive if a patient experiences ST depressions of >1 mm in 2 or more contiguous leads, high arrhythmia burden, hypotension, or symptoms of chest pain [64, 65]. Exercise EKG offers a sensitivity of 58–68% and a specificity of 62–77% [47, 64] for the detection of anatomically significant CAD. While some studies have showed similar results between exercise EKG testing and other modalities, most trials show that exercise EKG has a lower diagnostic yield than other tests [49, 66‒69]. Despite this diagnostic inferiority, it is one of the cheapest modalities, is often available to be performed the same day and in-office, and it allows the patient to avoid exposure to radiation and contrast. This modality does have significant limitations and cannot be performed in patients who have baseline EKG changes such as a LBBB or ST depressions, decompensated heart failure, severe aortic stenosis, or who are unable to exercise [70]. The AHA/ACC guidelines advocate that exercise EKG can be used in patients with no known CAD and a low pretest probability (≤15%) to aid in risk stratification (class 2; LOE A), and while it can be used for the diagnosis of CAD, other methods such as CCTA or MPI should be considered first [37]. The ESC guidelines are less confident with exercise EKG and do not recommend it as a sole diagnostic method (class 3; LOE C), preferring to utilize it only when other modalities are not available, or if a physician wishes to evaluate a patient’s exercise capacity [22].

Exercise stress echocardiography involves obtaining two dimensional echocardiographic images both at rest and stress, secondary to exercise or pharmacologic agents. Exercise stress is conducted via pedaling a stationary bike or walking on a treadmill, similar to exercise stressors discussed in the exercise EKG section [70]. These images are compared synchronously side by side, and the stressed images are evaluated for new or worsening RWMA, a failure to augment contractility in response to stress, or for changes in global LV function [70, 71]. Pharmacological stress echocardiography substitutes exercise with a medication to stress the heart, either via coronary vasodilation, accomplished with medications such as adenosine, regadenoson, and dipyridamole, or through the use of inotropic agents, such as dobutamine, to increase contractility and myocardial oxygen demand [33, 71, 72]. The segments of the myocardium are assessed separately, and the risk, as well as the prognostic value, is held in characteristics such as the number of abnormal segments, percentage of total myocardium affected, and poor response in multiple coronary artery territories, among others [71]. If adjacent segments or a coronary territory cannot be adequately differentiated, the use of ultrasound-enhancing agents, such as contrast, can be helpful [16, 71]. Failure to escalate to the target heart rate, generally to 85% of max heart rate, as predicted based on age, can significantly reduce the sensitivity of these studies, and may warrant the addition of other agents such as atropine or the use of another testing modality altogether [72].

Exercise stress echocardiography has a sensitivity of 70–85%, and a specificity of 77–89%, with pharmacologic stress echo having a slightly higher sensitivity of 85–90%, with similar specificity of 79–90% [33]. Exercise stress can be preferred over pharmacological stress, as it represents a true physiologic response and provides important information on parameters, such as the time to symptoms, blood pressure response, and capacity to exercise. Given these benefits, pharmacologic stress is often used preferentially only in patients who cannot exercise [71]. Stress echocardiography is utilized in intermediate to high-risk patients, especially in individuals >65 years of age, as the pretest probability for obstruction and ischemia is higher in this cohort [16]. Stress echocardiography also may be more sensitive for the detection of clinically significant CAD with left main disease, or multivessel CAD, when compared to perfusion imaging, due to its ability to identify balanced multivessel ischemia [71]. The negative predictive value for stress echocardiography is high, with a large meta-analysis showing that, with a recent negative stress echo, the annual event rate for MI and cardiac death was 1.66% [73]. Limitations to testing with exercise stress echocardiography include exercise intolerance (due to immobility, severe aortic stenosis, left ventricular outflow tract obstruction, severe arrythmia), severe hypertension, technician skill, and poor left ventricular endocardial visualization due to anatomical constraints such as obesity or chronic obstructive pulmonary disease [33]. Pharmacologic stress with vasodilatory testing is contraindicated in patients with reactive airway disease, or hypotension, and inotropic agents are contraindicated in similar settings as exercise echocardiography as mentioned previously [72]. AHA/ACC guidelines recommend stress echocardiography for patients with intermediate-high risk with stable chest pain and no known CAD to diagnose ischemia and estimate risk for MACE (class 1; LOE B) [16]. Similarly, ESC guidelines recommend noninvasive functional testing (including stress echo) as the initial test for CAD in symptomatic patients for whom obstructive CAD cannot be ruled out by clinical assessment alone (class 1; LOE B) [32].

MPI includes single positron emission computed tomography (SPECT), and positron emission tomography (PET) and offers a functional evaluation for CAD. These modalities measure myocardial perfusion both at rest and under stress utilizing a radioactive tracer uptake as a surrogate for myocardial blood flow. If obstructive and hemodynamically significant coronary disease is present, the test will show decreased blood flow to the region of the heart supplied by the narrowed artery. If that area has normal blood supply at rest, there is a possibility that an obstructive coronary lesion is present [74, 75]. The stressor in the case of SPECT testing may come in the form of exercise (similar to exercise EKG), or secondary to a medication such as adenosine, dipyridamole, or regadenoson, while PET scans may only be pharmacologically stressed. The diagnostic yield is similar in exercise and pharmacologic stress modalities, but exercise testing can be procedurally difficult; therefore, patient characteristics and institutional practice should drive the choice of stressor [76].

SPECT offers a sensitivity of 85–88%, as well as a specificity of 70–85% for anatomically significant CAD. PET offers an improved sensitivity of 90–93% and a similar specificity of 81–88% for anatomically significant CAD [47, 77, 78]. Aside from the possible diagnostic benefit, PET is superior to SPECT in that it allows for higher resolution and accuracy [79], is not impacted as significantly by higher body mass index or breast tissue [80], offers reduced radiation dosage [81], and can quantify myocardial blood flow and flow reserve [82]. Myocardial flow reserve quantifies coronary blood flow to ensure that the post-stress perfusion is not globally reduced, which assists in ruling out severe multivessel “balanced ischemia” [78, 83, 84]. Despite these advantages, SPECT is still widely used and is often the more readily available and cheaper option at many institutions. Despite these benefits, these testing modalities are expensive and expose patients to elevated doses of radiation (12–14 mSv for SPECT, 3–4 mSv for PET [81]) when compared to other modalities. Furthermore, not all patients can undergo SPECT/PET, with contraindications including patients who are experiencing arrhythmia, hypotension, or have an inability to tolerate exercise or pharmacologic stress agents [37]. Both methods of testing offer highly sensitive and specific functional testing for CAD and may be preferred in patients with intermediate-high pretest probability who are unable to undergo CCTA. AHA/ACC guidelines suggest using stress PET or SPECT as a co-first line modality (along with CCTA) to diagnose CAD in intermediate-high-risk patients (class 1; LOE B). It is, however, recommended to favor PET over SPECT if both are available options (class 2; LOE A) [37]. The ESC guidelines agree with the AHA/ACC and offer a recommendation (class 1; LOE B) for SPECT/PET as co-first line diagnostic modalities in patients with a concern for stable CAD [22].

Cardiac magnetic resonance imaging (CMR) is a noninvasive tool that uses magnetic fields and radiofrequency waves to detect and localize myocardial ischemia and infarction while also assessing for myocardial viability [16]. CMR detects ischemia by assessing myocardial perfusion after the injection of a gadolinium-based contrast. Myocardial perfusion CMR allows for 3D whole heart visualization and high spatial resolution, highlighting areas with reduced or delayed contrast uptake [85]. Stress CMR is completed via pharmacological vasodilators, such as adenosine, dipyridamole, or regadenosine to induce coronary vasodilation and augment ischemia in flow-limiting lesions. The stress perfusion images are compared with resting perfusion images to help distinguish true perfusion defects from scar and artifact [85]. Stress CMR yields an 81–90% sensitivity, and 80–86% specificity, for the detection of functionally significant CAD, and a sensitivity of 89%, and specificity of 87%, for the detection of anatomically significant CAD [22, 86]. Contraindications to CMR include retained metal [87], severe renal dysfunction (defined as eGFR <30 mL/min/m²), and pregnancy [37]. Patient-related barriers include significant obesity and possible claustrophobia, which occurs in up to 15% of patients undergoing CMR [88]. Considerations include a high cost and lower relative institutional availability. CMR is beneficial in that it does not expose patients to ionizing radiation, it can assess viability of myocardial tissue for reperfusion, and it can evaluate for microvascular dysfunction [89‒92]. The ESC guidelines recommend stress CMR for the diagnosis of CAD in symptomatic patients where obstructive CAD cannot be excluded by clinical assessment alone (class 1; LOE B) or when CCTA is nondiagnostic for CAD (class 1; LOE B). It is also recommended for detecting suspected CAD when stress echocardiogram with contrast is inconclusive (class 2; LOE B) [22, 93, 94]. The current AHA/ACC guidelines endorse stress CMR for evaluating intermediate-high-risk patients with stable chest pain and no known CAD for diagnosis of CAD and estimating risk of MACE (class 1; LOE B) [16].

The AHA/ACC and ESC guidelines recommend pretest risk stratification and the choice of testing modality, based on individual patient characteristics. These characteristics can be related to a patient’s ability to tolerate a testing modality, or participate with exercise, but also can be related to underlying comorbidities. The pathogenesis of CAD is related to, and exacerbated by, other conditions such as peripheral artery disease (PAD), which can increase the likelihood of concomitant CAD. Furthermore, risk factors such as increased CAC scores, Lipoprotein(a) (Lp[a]), and elevated polygenetic risk scores can impact the diagnostic pathways. These patient-specific considerations can impact their pretest probability which, in turn, can alter the diagnostic pathway. In addition to risk stratification, patient comorbidities may impact the technical feasibility of certain diagnostic modalities and guide physicians to choose one method of testing over another due to contraindications.

PAD and CAD share a similar underlying pathogenesis and risk factors as they both develop secondary to atherosclerosis [95, 96]. While PAD does not directly cause CAD, it is more likely for patients who do have PAD to also have or develop CAD [97]. Despite this increased risk, the 2016 AHA/ACC and 2017 ESC guidelines for PAD do not recommend asymptomatic testing for CAD, as it has not been shown to improve clinical outcomes [98, 99]. Additionally, there is no evidence to suggest that the presence of PAD should alter the diagnostic pathway, outside of guiding a pretest risk stratification. The choice of testing modality can be impacted, however, by contraindications secondary to severe underlying PAD. PAD can limit exercise tolerance through claudication and decrease the yield of exercise EKG or MPI [70]. Furthermore, severe PAD may complicate ICA if the peripheral vasculature is too narrow or altered to pass diagnostic catheters through, but it does not represent a true contraindication. Given these considerations, the approach to testing for CAD in a patient with known PAD must be nuanced to recognize that the presence of PAD confers a higher pretest probability of CAD and may bring contraindications or limitations to specific testing modalities.

Calcific aortic valve stenosis (CAS) is the most prevalent valvular disease in the world and often occurs concurrently with CAD due to shared risk factors [100]. Not only have studies suggested that degenerative AS is highly prevalent in patients with CAD (40–75%) but they have also shown that the advancement of aortic valve calcification coincides with a simultaneous increase in the total volume of coronary calcified plaque [101, 102]. Due to this shared pathophysiology and relationship, it has been suggested that CAS could be used as a marker for CAD and should be used for assistance in risk stratification in patients with suspected CAD [100, 103]. The diagnostic workup for suspected CAD in patients with CAS may be impacted by burden of calcification and hemodynamic significance of the AS limiting yield and ability to tolerate diagnostic procedures. Exercise or inotropic testing may be contraindicated in severe symptomatic AS and may be used cautiously in severe asymptomatic AS. This is because this population may experience a limitation in exercise tolerance and even syncope or cardiac arrest during exertion [40, 70, 104]. Additionally, functional testing may have reduced diagnostic yield as AS can increase LV afterload and myocardial oxygen demand [105]. Anatomic testing with CCTA may also be limited, as an extensive coronary calcification burden may decrease the diagnostic yield and limit a true understanding of luminal stenosis [106]. Despite these considerations, for most patients, the diagnostic approach for CAD in CAS is not different than the general pathway as discussed previously in this paper [16, 32]. If an evaluation for CAD is being completed in a patient who is being considered for a CAS surgical or catheter-based intervention, the AHA/ACC and ESC guidelines advocate for the use of ICA [107].

Another set of conditions which may confer an increased risk of developing CAD include chronic inflammatory diseases such as systemic lupus erythematosus, human immunodeficiency virus, and inflammatory bowel disease. These conditions are associated with an increased risk of CAD due to increased systemic inflammation, proatherogenic lipid profiles, immune dysregulation, and side effects of treatment [108, 109]. These conditions should impact patient risk stratification, but the data still support an evaluation with typical noninvasive testing modalities to capture early CAD, identify high-risk plaque morphology, and define plaque at baseline [108, 109]. As with the pathways mentioned above, pretest risk stratification and patient-specific contraindications to a testing modality should drive the choice of test. Despite an increased risk for CAD in such patients, there has been no indication for asymptomatic screening in studies, thus far, for these patients. There have been a few smaller trials indicating a benefit for asymptomatic CAC screening for assistance in risk estimation to help clinicians with determining when medical therapy for CAD could be beneficial [110].

Apolipoprotein(a), derived from the LPA gene, binds to cholesterol-rich LDL to produce Lp(a) [111]. Meta-analyses of prospective observational studies demonstrate that an elevated serum Lp(a) concentration is linked to a dose-dependent increased risk for CAD [112]. Additionally, Mendelian randomization analyses have supplied robust evidence suggesting a causal relationship between Lp(a) levels and CAD risk [113‒115]. The AHA/ACC guidelines on Primary Prevention of Cardiovascular Disease recognize elevated Lp(a) as a risk-enhancing factor at levels ≥50 mg/dL or ≥125 nmol/L. Despite the clearly increased risk, current AHA/ACC and ESC guidelines do not offer specific diagnostic testing strategy for suspected CAD and recommend using patient-specific factors and risk stratification methods to choose the best noninvasive modality for each patient [16, 32]. Furthermore, there has been no indication that an increased Lp(a) should prompt asymptomatic screening, but some studies have suggested using CAC scoring to further assist with risk stratification.

Polygenic risk scores (PRS) provide a numerical representation that combines the influences of numerous small-effect genetic variants identified in genome-wide association studies [116]. Studies have shown that individuals with a higher PRS have an increased incidence and prevalence of CAD, with one meta-analysis of almost 1,000,000 patients showing that a 1 standard deviation increase in PRS was associated with a 67% increase in CAD [116‒118]. Despite this promise, PRS are not routinely used in screening patients for CAD and are primarily used as a research tool for now [117]. Due to this, current clinical risk estimators and diagnostic frameworks do not account for an increase in PRS [117]. The ESC recognizes that genetic risk scores may improve risk prediction separately from conventional risk factors; however, the clinical utility of PRS in diagnostic testing for suspected CAD is still under investigation, and more evidence is needed to establish its role in clinical practice [32, 119, 120]. Furthermore, the current AHA/ACC guidelines do not directly comment on PRS [16]. Given these considerations, patients for whom a PRS is available and elevated should be considered to have a higher risk of developing CAD, but this should not solely guide choice of testing modality or prompt asymptomatic screening.

While many of the above factors confer an increased risk for developing CAD, CAC offers a direct measure which has been well verified in determining the presence of CAD. An increase in calcification often occurs concurrently with plaque formation, eventually forming larger deposits, detectable by CT [121]. CAC uses this principle to predict coronary atherosclerosis burden based on the presence of calcification, with higher scores indicating an elevated cardiovascular disease risk [122]. CAC scores, when available, should be used along other factors to help determine pretest probability and the appropriate next steps in testing [34]. Obtaining a CAC score may also be warranted when patients have intermediate cardiovascular risk to rule out calcified plaques and to avoid further testing if the CAC score is low [16]. Furthermore, CAC has been suggested in some asymptomatic populations with low-intermediate cardiovascular risk, but the latest guidelines do not support the use of routine asymptomatic screening, even in these populations. When a CAC score is obtained, a score of 0 is associated with a low prevalence of obstructive CAD (<5%) despite its inability to test for noncalcified plaque, and should warrant no further testing for most patients [32, 36, 123]. On the other hand, extremely elevated CAC conveys a high level of calcification which increases the patient’s risk and may limit the utility of CCTA in determining true CAD burden. Given these considerations, the AHA/ACC guidelines support adding CAC to other testing and using it to help risk stratify patients with low perceived risk, but do not provide specific guidance regarding further testing given a certain CAC [16]. The ESC guidelines suggest not using CAC as a sole screening modality but make no recommendations regarding its use in other settings [32, 124].

The proposed diagnostic strategy (shown in Fig. 2) includes an initial risk stratification for pretest probability of CAD based on symptoms and available clinical data. Patients with a pretest probability less than 5% should receive no further testing, while patients with a high pretest probability should be considered for direct ICA. In patients with a pretest probability between 5 and 15%, a CAC score and/or exercise EKG can be obtained to further risk stratify patients to low-risk or intermediate-high-risk. Intermediate-high-risk patients should be tested with CCTA (preferred) versus PET/SPECT based on their individual patient characteristics and institutional availability. If this testing is equivocal or positive, medical management should be initiated, risk reduction counseling should be performed, and further invasive angiography should be considered if the patient would be a candidate for revascularization.

The incidence of CAD is expected to increase 31% over the next 35 years; therefore, the decision regarding who to screen and how to do so will continue to arise for physicians [125]. While our understanding of the inflammatory pathogenesis of atherosclerosis and the methods for diagnosing CAD have improved drastically, there are many questions that remain unanswered [9, 95]. Recent trials such as International Study of Comparative Health Effectiveness With Medical and Invasive Approaches (ISCHEMIA), and a placebo-controlled trial of percutaneous coronary intervention for the relief of stable angina (ORBITA-1 + 2), have shown no mortality benefit in revascularizing obstructive lesions in stable CAD, and show that optimal medical therapy is as effective in reducing symptom burden [126‒128]. These studies call into question whether a diagnostic strategy evaluating for obstruction, or for ischemia alone, is sufficient. Pathologies such as ischemia with no obstructive coronary arteries, secondary to microvascular ischemia or vasospastic disease, as well as nonobstructive CAD, are not always well tested using traditional noninvasive methods, but may be responsible for a significant portion of patients’ symptoms [129, 130]. Given these shortcomings, we must move toward a future where CAD testing assesses overall risk, not just evaluating for obstructive or functional CAD burden. While anatomical testing can assess for subclinical CAD, functional testing can uncover pathologies such as microvascular disease and ischemia with no obstructive coronary arteries [131, 132]. Given this synergy, it is increasingly likely that the future of noninvasive CAD testing will combine these into a single test, such as CCTA with CT-FFR, which will better define the significance of a patient’s CAD.

Even when faced with an anatomically and functionally tested lesion, it can still be difficult to predict which patients will have cardiac events or persistent symptom burden. Several new technologies are being investigated which will offer clinicians better predictive efficacy in determining the true risk of a patient with CAD. Autopsy has shown that high-risk atherosclerotic plaques often have a thin fibrous cap, are lipid rich, and have a necrotic core [133, 134]. Invasive catheter-based techniques such as intravascular ultrasound and optical coherence tomography have been shown to offer some insight into plaque composition. Vessel dimensions, plaque morphology, calcium composition, and more may be determined which can aid in risk prediction but continue to require an invasive acquisition. New innovations in imaging have allowed for plaque composition to be assessed noninvasively. CT findings such as plaque cores with lower attenuation, spotty calcification, and remodeling, have been correlated with autopsy high-risk plaque features [135, 136]. Furthermore, there are efforts to quantify intra-arterial inflammation using epicardial adiposity [137]. Imaging findings consistent with high-risk plaque composition and intravascular inflammation have both been associated with an increased risk of cardiac events [138‒141]. Additionally, biomarker-based assays have been used to evaluate the blood for subtle clues of future CAD risk [142]. The field of genomics is also being studied to evaluate for a genetic predisposition and trends in CAD risk [143, 144]. Machine learning offers another avenue for improving risk prediction. Artificial intelligence is being used in imaging, as well as in risk assessment modeling, to better predict those who are more likely to have future ischemic cardiac events [145‒148]. As the cost of these screening methods decreases and risk prediction tools improve, our timeline for CAD screening may evolve to include widespread asymptomatic testing akin to breast or colon cancer screening. This future of CAD testing involves utilizing a comprehensive approach which emphasizes the risk of a patient’s overall atherosclerotic burden and plaque composition to better predict their risk of future cardiac events. This will allow for the early implementation of medical therapy and risk factor reduction to reduce a patient’s overall long-term risk of morbidity and mortality and improve their clinical outcomes.

The authors have no conflicts of interest to declare.

This review article was produced without financial support or sponsorship of any kind.

E.W., N.N., J.B., and S.C.: writing – original draft preparation; N.N., M.K., and C.K.: visualization; A.R., R.S., R.K., M.K., and C.K.: writing – reviewing and editing; and C.K.: supervision.