Stress echocardiography in paediatrics

Stress echocardiography is most commonly used in adult patients with possible atherosclerotic coronary artery disease to determine the need for intervention either by surgery or by percutaneous coronary angioplasty. In paediatrics, early onset acquired coronary heart disease is uncommon but there are several congenital and acquired conditions in which stress imaging may play an important role in management. Choice of stressor, either exercise or pharmacological, depends on several variables, including the specific disease in question and the goal of the study, age, and ability to perform the exercise protocol as well as institutional resources, preferences, and experience. According to a 2006 American Heart Association Council on Cardiovascular Disease in the Young survey of paediatric cardiology training programmes, stress echocardiography was performed in only 57% of institutions responding to the survey.Reference Paridon, Alpert and Boas¹ Training and experience in paediatric stress echocardiography are, therefore, not universal, and large studies focussing on the accuracy and outcomes for specific diseases are not available; however, from survey data and the available literature, stress testing in the paediatric population appears to be safe and has very few complications. Other stress imaging modalities include exercise single-photon emission CT, as well as pharmacological positron-emission tomography or MRI, although echocardiography has distinct advantages over these techniques, including safety, because of the lack of ionising radiation, simplicity, accessibility, and cost, which make it the modality of choice in most cases.

Indications for stress imaging

There are no commonly agreed upon indications or appropriate use criteria for stress echocardiography in paediatrics; however, there are three general categories of information that can be obtained: detection of stress-induced wall motion abnormalities due to ischaemia from coronary abnormalities, changes in Doppler gradients, and global systolic function enhancement associated with contractile reserve. The latter two categories will not be addressed in this study, although they may account for a substantial percentage of studies carried out in the paediatric laboratory and include detection of exercise-related mid-left ventricular cavity gradients in hypertrophic cardiomyopathy, assessment of stress-related pulmonary hypertension by tricuspid regurgitation jet Doppler, especially in symptomatic patients with left-sided obstructive disease not otherwise meeting the criteria for intervention, assessment of mitral or aortic stenosis gradients, especially in the setting of symptoms or ventricular dysfunction, and evaluation of global function after recovery from myocarditis or cardiotoxic chemotherapy. Regarding the detection of coronary stenosis by stress echocardiography compared with stress electrocardiography alone, there is a cascade of ischaemic events in the at-risk myocardium exposed to stress that follow a distinct timeline, beginning with metabolic changes, followed by detectable flow perturbations, wall motion abnormalities, electrocardiography changes, and finally chest pain.Reference Kimball² Although positron-emission tomography and single-photon emission CT may detect early changes, they lack specificity for clinically important stenosis, with >50–70% luminal narrowing, whereas electrocardiography stress alone lacks the sensitivity to detect ischaemia that wall motion assessment affords. Thus, stress echocardiography affords an optimum compromise of sensitivity and specificity due to its target event in the ischaemic cascade and gathers electrocardiography data as well as clinical assessment of symptoms.

Assessment of regional wall motion

The most common diagnoses with potential for stress-induced regional wall motion abnormalities in the paediatric laboratory include Kawasaki disease, post-transplant coronary vasculopathy, and post-operative coronary stenosis after arterial switch operation or anomalous coronary origin repair.Reference Kimball, Witt and Daniels³ Pharmacological stress is most commonly induced with dobutamine and atropine, although adenosine and dipyridimole are also used in some laboratories. Although dobutamine simulates exercise by increasing heart rate, contractility, and thus myocardial oxygen consumption, adenosine and dipyridimole detect ischaemia through differential dilatation of the coronary vessels. In dobutamine studies, a baseline echocardiography is first performed to image the left ventricular wall motion at rest during at least the systolic portion of the cardiac cycle, from the apical four-chamber and two-chamber views as well as from the parasternal long-axis and short-axis views. Additional imaging from the apical long-axis and parasternal short-axis views at the base and apex in addition to mid-papillary muscle level may be particularly helpful when windows are limited from the standard views. Dobutamine by continuous infusion is then started at a low dose, usually 5 mcg/kg/minute, and increased at 3- to 5-minute intervals until target heart rate is achieved or limiting symptoms, rhythm abnormalities, or signs of definite ischaemia occur. At each stage, additional echocardiography images, along with the 12-lead electrocardiography and vital signs, are acquired. If the target heart rate is not achieved despite 40–50 mcg/kg/minute of dobutamine, the infusion is continued and atropine boluses are administered every minute until either peak heart rate occurs or the maximum cumulative dose of atropine is administered. Once the target heart rate is achieved – typically 85% of the predicted maximum heart rate for age – the infusion is continued until peak imaging is completed, and then the infusion is discontinued with one final set of images acquired in recovery. A distinct advantage of pharmacological over treadmill exercise stress is the ability to sustain peak heart rate until imaging is complete and the lack of patient movement while images are acquired, which may often lead to better sensitivity for detection of coronary lesions, especially those that result in wall motion abnormalities at higher levels of stress only. The test can be carried out on patients of all ages, including infants, and does not require the ability to run, bicycle, or cooperate with an exercise protocol. In addition, as imaging takes place during each successive stage, changes in wall motion between low and high dose indicative of a hibernating but viable myocardium can be detected, which may inform decisions about re-vascularisation. Nevertheless, an intravenous line must be placed, there is a slightly higher risk of complications such as arrhythmia, and in most young children some form of sedation will be required, preferably a form that does not negatively affect heart rate and contractility, such as ketamine.

For older children who are able to cooperate, exercise stress echocardiography is preferred over pharmacological stress echocardiography because of the simpler protocol and the similarity to actual exercise experienced by the patient in daily life, including cardiorespiratory interactions, higher systolic blood pressures and overall metabolic demands, which may be particularly useful in the assessment of symptoms related to exertion. Treadmill or bicycle, either upright or supine, protocols can be used with advantages and disadvantages to both. In general, treadmill exercise is more natural for most patients and may lead to higher workloads achieved, with possible improved sensitivities compared with a bicycle, although this advantage may be somewhat offset by the ability to scan patients during peak exercise when using the supine bicycle protocol. The treadmill exercise stress echocardiography protocol involves acquisition of resting images, then exercising to near exhaustion or development of concerning symptoms or signs of ischaemia or rhythm abnormalities on continuous electrocardiography monitoring. Immediately following peak exercise, the patient must move quickly from the treadmill to the supine position for post-stress imaging. As heart rate usually falls rapidly in children and adolescents after exercise, optimum target heart rate is usually >90% of the predicted maximum, and post-stress imaging should be completed within the initial 30 seconds following exercise, always starting with the views containing the wall segments of highest clinical interest.

Although not currently approved by the FDA for use in children, ultrasound contrast use has been reported in the paediatric age group and can be particularly helpful in both pharmacological and exercise stress echocardiographies when imaging windows are suboptimal.Reference Zilberman, Witt and Kimball^4–Reference Mulvagh, Rakowski and Vannan⁶ Intravenous contrast can be administered at baseline and during stress to enhance left ventricular opacification and to improve wall motion visualisation, extending the usefulness of echocardiography as the primary stress imaging modality in children.

For both pharmacological and exercise protocols, stored DICOM images can be reviewed in a pre- and post-stress layout with the playback speed of the stress images reduced to play synchronously with baseline images for the ease of visually detecting qualitative differences in wall segment motion. Wall segments that have normal systolic thickening and shortening at rest, but appear hypokinetic compared with other segments with stress, are presumed to have reduced coronary blood supply (Fig 1). Subtle abnormalities can be difficult to detect and require experience with multiple studies to optimise reader accuracy. As this experience can be difficult to acquire in most paediatric laboratories, spending time in busy adult laboratories and consulting with adult echocardiography colleagues regarding interpretation difficulties can speed up the learning curve of paediatric cardiologists. In addition, dedicating a core group of sonographers to gain experience in the rapid acquisition and interpretation of these images is useful. Clinical interpretation of stress echocardiography remains primarily a matter of careful visual inspection of each segment and assignment of a score for increasingly abnormal motion from normal to hypokinetic to akinetic to dyskinetic. Although there are ongoing attempts to automate and quantitate segmental motion comparisons, none has yet replaced the visual inspection method in most clinical settings. Speckle tracking methods of quantifying myocardial segment displacement, velocity, and strain changes during systole may offer a more objective and less visually dependent method of analysis; however, the current methodology used is problematic at the higher heart rates required for stress studies.

Figure 1

Stress-induced apical septal wall motion abnormality.

Sensitivity, specificity, and prognostic value

In adults, the reported average sensitivity for stress echocardiography is 80–88% with specificity of 83–86% for the detection of coronary artery stenosis resulting in at least a 50% luminal diameter reduction, based on data pooled from multiple studies and thousands of patients as summarised in the 1998 American Society of Echocardiography document on stress echocardiography, updated in 2007.Reference Armstrong, Pellikka, Ryan, Crouse and Zoghbi^7, Reference Pellikka, Nagueh, Elhendy, Kuehl and Sawada⁸ This is higher compared with stress electrocardiography alone, which has a sensitivity and specificity of 52 and 71%, respectively. Studies comparing stress echocardiography with single-photon emission CT have generally shown similar to slightly higher sensitivity for single-photon emission CT with lower specificity (84 and 77%, respectively) compared with echocardiography.

False negatives most commonly result from failure to achieve 85–90% of the predicted maximum heart rate for age, or in the case of the treadmill protocol imaging during the recovery phase when heart rate is lower than the peak achieved. Single vessel disease is also more subject to missed wall motion abnormality. False positives can result from single segment abnormal motion, with increased specificity resulting from requiring at least two contiguous segments to be abnormal. Global dysfunction can be seen in the setting of hypertensive response to stress and in dilated cardiomyopathy that is not due to coronary abnormalities. Left bundle branch block and other post-operative changes can result in wall motion abnormalities, although these should be present on baseline imaging, and even if they appear to worsen with stress they are not necessarily indicative of ischaemia, particularly if they do not follow a coronary distribution pattern.

In adults, based on a total of 9000 patients, a normal stress echocardiogram predicts an annual event risk of 0.4–0.9%, and risk based on a positive test can be further classified into intermediate (1–3%) or high (>10%) depending on specific clinical features.Reference Sicari, Nihoyannopoulos and Evangelista⁹ According to Bayes theorem, calculating the positive and negative predictive values of a test require knowing, in addition to the sensitivity and specificity, the pre-test probability that the patient being studied actually has the disease, in this case coronary luminal stenosis >50%. Thus, in adults with very high pre-test probability of coronary artery disease, such as a 65-year-old man with typical angina (>90% pre-test probability of disease), a negative result from stress imaging would more likely be a false negative than a true negative, and in a 30-year-old woman with non-anginal chest pain (<5% pre-test probability of disease) a positive stress test is more likely to be a false positive.Reference Douglas, Khandheria and Stainback¹⁰ A pre-test probability of disease in the 20–80% range is considered to be optimum for the use of stress echocardiography to assist in clinical decision making for adult patients with possible coronary heart disease.

Kimball et alReference Kimball, Witt and Daniels³ reported on 74 dobutamine stress echocardiography studies carried out on 46 children with a variety of diagnoses including post-operative arterial switch, Kawasaki disease, and heart transplantation. Worsening stress-induced wall motion abnormalities were seen in eight patients. In the 34 patients who also had angiography, 94% had concordance between studies; five of the eight patients with a positive stress test developed cardiac events during the follow-up period of 8–27 months, including a patient with heart transplant arteriopathy who required a second transplantation, three patients who developed angina, and one who required coronary bypass. In addition, 36 of 37 patients with a normal stress test remained event free, with one undergoing bypass despite the lack of symptoms and extensive collateralisation of a nearly occluded left coronary artery. Noto et alReference Noto, Kamiyama and Karasawa¹¹ recently published 15-year follow-up data of 58 patients with Kawasaki disease who had dobutamine stress echocardiography results, 36 with and 22 without coronary involvement. The severity of stress-induced wall motion abnormality at the initial stress testing was the only independent predictor of major adverse cardiac events. Dipchand et alReference Dipchand, Bharat, Manlhiot, Safi, Lobach and McCrindle¹² published dobutamine stress echocardiography results of 102 paediatric heart transplant patients and found low sensitivity (35%) with high specificity (94%) for detection of coronary heart disease. On the other hand, more recently, using exercise stress echocardiography, Chen et al,Reference Chen, Abernathey and Lunze¹³ reported higher sensitivity (88.9%) with excellent specificity (91.9%) for detecting cardiac allograph vasculopathy in 47 children after heart transplantation. Using dobutamine stress echocardiography, Hui et alReference Hui, Chau, Leung, Chiu and Cheung¹⁴ found new regional wall motion abnormalities in 23 of 31 patients following arterial switch for transposition of the great arteries. Stress single-photon emission CT in 22 patients showed reversible perfusion defects in 17 of them.

Stress echocardiography in patients with anomalous aortic origin of the coronary artery

There have been no published systematic, prospective studies of pre or post-operative stress imaging in patients with anomalous aortic origin of the coronary artery. Romp et alReference Romp, Herlong and Landolfo¹⁵ reported nine patients with anomalous aortic origin of the coronary artery – seven with left and two with right coronary arteries arising anomalously – between 1995 and 2001; eight patients underwent post-operative exercise stress echocardiography, all of which were negative. Osaki et alReference Osaki, McCrindle, Van Arsdell and Dipchand¹⁶ used dobutamine stress echocardiography to evaluate 17 of 31 unrepaired patients – 13 with left and 18 with right coronary arteries arising anomalously – seen from 1994 to 2006, with a median age of 6.2 years. None showed stress-induced wall motion abnormalities, including four patients with a history of syncope with exercise. Mumtaz et alReference Mumtaz, Lorber, Arruda, Pettersson and Mavroudis¹⁷ reported 22 patients – seven with left and 15 with right coronary arteries arising anomalously – operated on between 1998 and 2008. Post-operative single-photon emission CT imaging was carried out in five patients, exercise stress echocardiography in four patients, and exercise stress electrocardiography in one patient. All the stress tests were negative. Wittlieb-Weber et alReference Wittlieb-Weber, Paridon, Gaynor, Spray, Weber and Brothers¹⁸ extended an earlier study from the Children’s Hospital of Philadelphia to report medium-term outcomes in a series of 24 patients after repair of anomalous aortic origin of the coronary artery, with a median age at follow-up of 18 years and a median follow-up from repair of 63 months. 54% of the patients had symptoms post-operatively, mainly chest pain, and half of them had the same symptoms pre-operatively; seven patients underwent post-operative exercise stress echocardiography including two patients with exertional syncope and one with chest pain. The only positive stress test was in an asymptomatic patient with repair of anomalous left coronary artery origin.

To summarise the available literature, <20 patients with unrepaired anomalous aortic origin of the coronary artery and stress echocardiogram data have been reported, all with negative stress test results including some carried out in patients with symptoms. Post-operative stress echocardiography has been reported in 22 patients, and only one of them was positive.

As large, disease-specific studies in children with coronary stenosis – particularly with both stress echocardiography and angiography data – are not available, sensitivities and specificities of stress testing in children are not clear, and information on the prognostic value of a positive or a negative test is limited. In children, even by assuming sensitivities and specificities similar to adult studies for equivalent disease processes, careful attention must be paid to the pre-test probability of significant coronary stenosis in order to understand the meaning of a positive or a negative test. Children with chest pain and no other evidence of coronary heart disease are very unlikely to have a true positive result, and thus stress imaging is probably not helpful. Relatively rare coronary anomalies such as anomalous aortic origin are often discovered incidentally in asymptomatic children by echocardiography, and risk assessment would be desirable to help in decision making regarding activity restrictions and possibly surgical intervention. In this case, stress imaging in the unrepaired child is not being used to make the diagnosis of anomalous coronary origin, but rather to assess whether or not a stress test can provoke ischaemia given this known abnormality, which is then assumed to be related to the risk of sudden death. In fact, the test is asking whether this congenital coronary anomaly, when subjected to the stress of running on the treadmill or riding a bicycle, creates the equivalent obstruction to coronary flow for this patient of at least a 50% luminal narrowing due to atherosclerosis; however, the pathological mechanisms of ischaemia in these anomalies are poorly understood, likely multifactorial, and are not necessarily similar to adult-onset coronary atherosclerotic disease. Therefore, stress imaging by current protocols may have lower sensitivity and specificity for distinguishing potentially life-threatening from more benign anatomical arrangements. Likewise, as the number of patients with these anomalies who develop symptoms or experience sudden death is lower, and by some estimates much lower than the total number with the anomaly,Reference Peñalver, Mosca, Weitz and Phoon¹⁹ the pre-test probability of exercise testing resulting in coronary narrowing of >50% may be below the optimum range for detection by stress echocardiography, particularly for anomalous aortic origin of the right coronary artery. For example, let us assume a sensitivity of 80% and a specificity of 86% for stress echocardiography being able to detect an anomalous coronary configuration significantly pathological enough to result in >50% luminal stenosis. Following the reasoning of Peñalver et al, let us also assume that only 6.3% of patients with anomalous left coronary artery origin and 0.2% of those with anomalous right coronary artery origin are at risk of dying. For simplicity, we will further assume that the risk of death is the same as the pre-test probability of having >50% luminal stenosis. The positive predictive value of stress echocardiography would then be 28% for anomalous left coronary artery origin and 1% for anomalous right coronary artery origin, making a positive result much more likely to be a false positive than a true positive, although by the same logic a negative result would most likely be a true negative. Thus, given the lack of evidence regarding the actual sensitivity and specificity of stress echocardiography for anomalous coronary origin, and assuming that the pre-test probability of having a positive test is low to very low given the above, understanding the result of the test and using this to inform clinical decision making is problematic. Nevertheless, a common justification for pre-operative testing is that a negative result can be used as reassurance that surgery and possibly even activity restriction is not required, at least in asymptomatic patients with anomalous right coronary arteries. From the limited available data, even when symptoms possibly due to coronary ischaemia are present, stress imaging may be negative, and, in fact, post-operative chest pain similar to pre-operative symptoms has been reported in some patients following what otherwise appears to be a successful repair, again with negative post-operative stress imaging.Reference Wittlieb-Weber, Paridon, Gaynor, Spray, Weber and Brothers¹⁸

Post-operative stress imaging following repair of an anomalous coronary artery has similar rationale to that following arterial switch operation in that after a repair involving moving the coronary artery or unroofing the origin, either the repair itself may be problematic or with time, changes in geometry or occlusion of the orifice with scar tissue may occur. Typically, these problems would be associated with symptoms such as exertional chest pain, the evaluation of which should include stress echocardiography. In asymptomatic adults, routine risk assessment in the first 5 years following coronary bypass surgery is considered an inappropriate use of stress echocardiography.Reference Douglas, Khandheria and Stainback¹⁰ It is unclear whether there is a strong rationale for adopting a different standard of appropriateness for stress imaging in asymptomatic children following coronary artery surgery; however, it could be argued that in the growing child, and in particular for surgical repairs performed less frequently than coronary bypass surgery in adults, prudent surveillance might include periodic stress imaging, especially for children and young adults engaged in competitive sports. Once again, however, the actual sensitivity, specificity, and pre-test probability are unknown; thus, interpretation of the results and use for clinical decision making are not straightforward.

Deciding which test is the best

To decide which test is the most appropriate, it helps to have a clear idea of what the most important concern is that has to be addressed and how likely it is that the condition exists in the patient. If a coronary artery abnormality is being investigated, it has to be decided whether the functional assessment of wall motion (echocardiography), perfusion (single-photon emission CT), or anatomy (CT or angiography) is most important or would ischaemic changes on stress electrocardiography be sufficient (Fig 2). In some cases, more than one test will be required to fully understand the problem and plan for intervention. The patient’s age, level of cooperation and coordination, and the ability to exercise will help determine whether pharmacological testing will be necessary. Imaging and Doppler studies may be needed if additional haemodynamic information with stress such as valve gradients are important; however, before obtaining any test, a mental review of how the information will be used to guide clinical decision making in an evidence-based manner if possible is recommended.

Figure 2

Algorithm for deciding which test is the best. CAD=coronary artery disease; DSE=dobutamine stress echocardiography; ECG=electrocardiography; ESE=exercise stress echocardiography; LBBB=left bundle branch block; OHT = orthotopic heart transplant; SPECT=single-photon emission CT; WPW=Wolff–Parkinson–White syndrome.

Conclusions and recommendations

Stress echocardiography is a safe and useful diagnostic test in the paediatric echocardiography laboratory and should be a part of any full service programme; however, due to the unique diagnoses and lower case volume in paediatrics, experience takes longer to acquire compared with adult stress echocardiography. Regarding the role of stress echocardiography in the pre-operative evaluation of patients with anomalous aortic origin of the coronary artery, evidence is lacking and for many reasons it may be incorrect to simply apply predictive data from adult coronary atherosclerotic disease to inform the interpretation of pre-operative stress tests in this situation. In patients with anomalous left coronary artery, especially if symptomatic, the risk of mortality may be sufficiently high that a negative stress test should not dissuade the prudent physician from recommending surgery or activity restriction; however, for anomalous right coronary artery origin a negative test may be reassuring and may add some justification for following a conservative course of action, especially in the asymptomatic patient. Post-operative stress echocardiography should be used to evaluate symptoms possibly caused by problems with the repair, and annual or biannual re-evaluation may provide reassurance to asymptomatic patients engaged in competitive sports. These recommendations are for the clinical management of patients given currently available evidence. Multicentre, prospective studies are needed to fully understand the proper role and interpretation of stress imaging in the management of these rare congenital coronary anomalies.