Feasibility of Exercise Stress Echocardiography for the Follow-up of Children With Coronary Involvement Secondary to Kawasaki Disease
Background The development of coronary aneurysms as sequelae of Kawasaki disease can result in myocardial ischemia, infarction, and sudden death. Traditionally, these patients have undergone coronary angiography and nuclear stress imaging for risk stratification and follow-up. However, angiography is invasive, and both modalities expose the patient to repeated radiation, which is an important issue in children. The purpose of this study was to determine the feasibility of performing exercise stress echocardiography in children diagnosed with coronary abnormalities secondary to Kawasaki disease.
Methods and Results Treadmill exercise stress echocardiographic studies were performed in 28 children ages 6 to 16 years. All had acute Kawasaki disease 1 to 10 years before study, and coronary artery abnormalities were identified during previous echocardiographic imaging. Patients were exercised using a standard Bruce protocol. Transthoracic echocardiographic images, obtained in the parasternal long, short, apical two- and four-chamber views immediately before and after exercise, were digitized for review and analysis. In baseline studies before exercise, wall motion abnormalities were identified in 2 patients; these segments became normal with exercise. Two patients developed new exercise-induced wall motion abnormalities that corresponded to angiographically defined critical stenosis of the left anterior descending coronary artery. No patients had resting or exercise-induced ECG evidence of ischemia. There were no adverse reactions, and 26 of 28 patients had normal exercise tolerance.
Conclusions Among patients with coronary artery involvement resulting from Kawasaki disease, exercise stress echocardiography is a safe, noninvasive procedure and may identify children with myocardial ischemia that was not detected with ECG stress test alone.
Kawasaki disease is the leading cause of acquired coronary artery disease in children. Myocardial infarction as well as sudden death can occur in patients with significant coronary artery involvement.1 Aneurysms involving the coronary arteries occur in as many as 20% of patients afflicted with Kawasaki disease. These coronary aneurysms may become significantly stenotic, leading to reduced flow that results in myocardial infarction and death in up to 1% of patients with coronary involvement.1 2 The presence of obstructive coronary lesions is best defined by angiography; however, there is considerable controversy regarding indications for coronary angiography in this young population.3 4 Some recommend serial coronary angiography while others perform invasive studies only in patients with giant aneurysms because of the low incidence of stenosis in patients with only minor abnormalities on echocardiography.
Exercise or pharmacologically induced myocardial ischemia results in regional wall motion abnormalities before ECG changes.5 Transthoracic echocardiography can identify these segmental wall motion abnormalities. Stress echocardiography is routinely used in risk stratification of flow-limiting lesions in adults with atherosclerotic coronary artery disease.5 6 7 This study was undertaken to evaluate the feasibility of stress echocardiography for detection of flow-limiting lesions in patients with coronary aneurysms after Kawasaki disease.
Patients were selected from a population of over 450 children diagnosed with Kawasaki disease (1979 to July 1991) by standard clinical criteria8 and referred to The Children’s Memorial Hospital, Chicago, Ill, for evaluation. All 51 patients (11.3%) who had coronary artery abnormalities identified by screening two-dimensional transthoracic echocardiography after presentation with acute symptoms of Kawasaki disease were elibible for study. There were 10 such patients who were excluded because they were too young to perform on a treadmill (age less than 6 years), and 18 patients either refused or were lost to follow-up. The study group consisted of 23 children from this cohort as well as 5 patients from the Children’s Hospital of Pittsburgh who had significant coronary artery involvement documented by coronary angiography and history of Kawasaki disease. The Pittsburgh patients were recruited from a group of 290 Kawasaki patients. There were 32 Pittsburgh patients identified by coronary angiography as having coronary abnormalities initially and 17 had persistent abnormalities on follow-up. Of these 17 patients, 1 died, 9 were lost to follow-up or refused, and 5 agreed to be studied.
There were 22 boys and 6 girls aged 6 to 16 years (median and mean, 10.7). Kawasaki disease was diagnosed 1 to 10.6 years (mean, 7.3; median, 7.6) before stress echocardiographic study. The median age was 2.1 years at the time of presentation with acute Kawasaki disease. Only 4 of the patients in this study had received intravenous gamma globulin therapy at the time of acute illness. All patients reported normal exercise tolerance. All patients were receiving aspirin, 3 dypyridamole, and l warfarin. Informed written consent was obtained from the parent or legal guardian. The stress echocardiographic protocol was approved by the Institutional Review Board of The Children’s Memorial Hospital, Northwestern University.
All transthoracic coronary artery echocardiographic imaging was performed with Hewlett-Packard Sonos 500 or 1000 ultrasound systems using a 7.5- or 5-MHz transducer with short or medium focus. Imaging techniques have been described by several investigators.9 10 11 12 Transthoracic imaging of a patient with a normal left coronary artery and of a patient with a giant aneurysm with thrombus is demonstrated in Fig 1⇓.
Coronary Artery Measurements
The coronary arteries were measured using the standard Hewlett-Packard measurement package. The accuracy of this technique in comparison to angiography has been previously described.9
Patients were subdivided into three groups, based on the severity of residual coronary involvement, as diagnosed from echocardiographic and/or angiographic data. The groups were defined based on severity of residual coronary involvement, with group 3 having the greatest risk of significant coronary stenosis and development of ischemia. Aneurysms or ectasia were defined by previously established criteria.13
Group 1: No Residual Abnormalities
This group had 15 patients with coronary dilation, ectasia, or small aneurysms initially that had resolved entirely on serial follow-up echocardiographic studies. Only 1 patient in this group has undergone coronary angiography.
Group 2: Mild Residual Coronary Abnormalities
This group had 9 patients with dilation or ectasia of a coronary segment but no giant aneurysms. Two of them had coronary angiography.
Group 3: Severe Abnormalities
This group had 4 patients with giant coronary aneurysms (diameter >8 mm by echocardiography) and coronary stenosis by angiography. One of these patients had a coronary bypass graft 2.5 years before stress echo study. All patients in this group have had serial coronary angiography as well as nuclear studies.
Stress Echocardiographic Technique
Baseline echocardiographic images were obtained with patients in the left lateral decubitus position. Images were acquired in standard parasternal long- and short-axis and apical two- and four-chamber views. Images were recorded on half-inch videotape and were digitized on line and stored, using a Freeland cine-view plus system. The images were acquired beginning at the R wave of the QRS. Eight frames are obtained, 50 milliseconds apart at baseline to 33 milliseconds apart at higher heart rates. These images are then displayed in a cineloop format for easy review and interpretation. Immediately after exercise, transthoracic images were obtained in the four views described above, and all images were again recorded on videotape and digitized onto a floppy disk.
The standard treadmill exercise protocol described by Bruce et al14 was used. Individual performance was compared with established normal values for children.15 Heart rate, rhythm with 12 ECG leads (Marquette recording system), and blood pressure were monitored throughout the protocol. Patients exercised to the point of fatigue. Rowland16 reviewed several studies that recommended obtaining a peak heart rate of 200/min when children are exercised on a Bruce protocol. Patients held on to the guardrails transiently to obtain balance and during blood pressure recordings.
Wall Motion Analysis
All exercise echocardiographic images were interpreted by at least three observers with complete consensus in findings (E.P., R.S., F.C.). The observers were not blinded to the coronary status of most of the patients. However, two of them were not involved in the clinical care of the patients. The videotape and digitized images were reviewed at baseline and before exercise. Wall motion abnormalities of the left ventricle were assessed. The left ventricle is divided into 16 segments, according to American Society of Echocardiography recommendations.17 Each segment was graded as being normal, hypokinetic, akinetic, dyskinetic, or aneurysmal. Overall left ventricular cavity size was measured before and after exercise.18 Wall motion was described as normal, hypokinetic, akinetic, or hypercontractile.6 7 Over 1000 studies have been performed in F.C.’s adult echocardiography laboratory, with an interobserver and intraobserver variability of <5% (personal observations/data, F.A. Chaudhry). All postexercise images were completed within 24 to 52 seconds (mean, 35 seconds) at a mean heart rate of >80% of age-predicted maximum heart rate.
Aortography and selective coronary angiography had been performed previously in all patients with giant aneurysms (measuring >8 mm in diameter echocardiographically) and was repeated within 0 to 2 months in all patients with abnormal stress echocardiographic study. The technique has been described previously.3 Serial coronary angiography was performed in 5 patients who were identified as high risk for progression of obstructive lesions (4 patients had giant aneurysms and 1 patient with an abnormal thallium scan wished to exercise competitively). This last patient had a normal coronary angiogram (group 1) and negative exercise stress echocardiogram.
The patients belonging to the three groups were compared using χ2 tests for discrete variables and ANOVA for continuous variables. A value of P<.05 was considered significant.
The mean values and ranges for clinical variables and exercise parameters are reported in Table 1⇓. For all clinical parameters, there were no significant differences between the groups. All tests were safe, as no patients had adverse symptoms such as chest pain, hypotension, near syncope, or skeletal muscle injury. Furthermore, there were no ischemic changes or arrythmia detected electrocardiographically throughout the protocol in any patients.
Exercise tolerance was normal in 26 of 28 patients.15 16 Only 2 patients had exercise capacity less than the tenth percentile for age- and sex-matched peers, and both were morbidly obese. Mean and median peak heart rates were 195 and 197 beats per minute, respectively, for all patients, and were similar when the groups were divided (Table 1⇑). The three groups were similar in all exercise parameters assessed, including the double product.
Despite rapid fall in heart rate in these pediatric patients, all images were acquired within l minute of stopping the treadmill (mean, 35 seconds), and patients were still at 62% to 94% (mean, 80%) of their peak heart rate when imaging was complete, which is a mean of 80% of age predicted maximum (ie, 200 beats per minute).
Wall Motion Analysis
Technically adequate studies were obtained in 27 of 28 patients (96%); 1 patient had a suboptimal window. No patient in group 1 or group 2 demonstrated new exercise-induced wall motion abnormalities. Two patients (both in group 2) had minor resting wall motion abnormalities that improved after exercise and were thought unlikely to be ischemic.
Table 2⇓ describes the findings for group 3 patients. T.C. had baseline hypokinesis and giant aneurysms. She was found to have a critical stenosis of the left anterior descending coronary artery and underwent bypass surgery. She exercised only 8 minutes but had no ECG evidence of ischemia. S.E. had an internal mammary artery graft 3 years before exercise stress echocardiography, and follow-up angiography showed graft patency. He had a normal exercise echocardiogram.
The other patients (E.C. and S.M.) developed new segmental wall motion abnormalities after exercise, consistent with myocardial ischemia. E.C. had serial coronary angiography that demonstrated progression of abnormalities: A large aneurysmal right coronary artery became thrombosed and subsequently showed nearly complete occlusion on cardiac catheterization performed 2 months before the stress echocardiogram. After exercise, the midanteroseptal region became akinetic.
S.M.’s initial angiogram 3 months after acute illness demonstrated giant aneurysms of all coronary arteries. He had been asymptomatic with a normal exercise thallium 2 years previously. Imaging after exercise revealed akinesis of the basal anteroseptum, midanteroseptum, and midseptum, consistent with exercise-induced ischemia (Fig 2⇓). This abnormal study prompted selective coronary angiography (Fig 3⇓) that correlated with the stress echocardiography findings. An exercise sestamibi study performed just before coronary angiography showed a perfusion abnormality at the basilar and apical septum. He was referred for coronary bypass surgery.
We have demonstrated that exercise stress echocardiography is feasible in children. This is the first study to report exercise stress echocardiography in patients with history of Kawasaki disease and coronary abnormalities. Our discussion focuses on the benefits of this technique in following these patients.
Exercise stress echocardiography combines the technique of a standard Bruce protocol to allow quantitation of the child’s exercise endurance with imaging of the left ventricle immediately after exercise to give a functional assessment of coronary flow reserve and hemodynamic consequences of coronary stenosis. One can evaluate global and segmental right ventricular and left ventricular function at rest and immediately after exercise. Exercise echocardiographic studies in adults suggest a high sensitivity and specificity comparable to single-photon emission computed tomography (SPECT) in identifying flow-limiting coronary artery stenosis.7
This study identified 2 patients who had abnormal wall motion immediately after exercise. In both cases, critical stenosis was confirmed on selective coronary angiography; however, neither patient had symptoms or ECG changes to suggest ischemia during exercise. Although exercise treadmill ECG studies have been reported to be normal in Kawasaki disease patients with minimal residual coronary disease,19 20 the sensitivity for detecting coronary ischemia is likely to be poor if adult ECG treadmill data are considered.21 Suzuki and Kamiya4 showed that only half of children with severe stenotic lesions had ECG changes consistent with ischemia on a treadmill study. Thus, ECG stress test alone is not a sufficiently sensitive technique to identify flow-limiting lesions and would have missed these 2 patients. We limited our study group to those patients who had coronary abnormalities identified by echocardiography because they were more likely to have a flow-limiting lesion.
Cross-sectional transthoracic echocardiography is currently used widely in the acute phase of illness to detect coronary artery involvement in patients diagnosed with Kawasaki disease with excellent sensitivity (>97%) and specificity (>97%).11 22 Reported coronary abnormalities include saccular or fusiform dilation and ectasia; however, the most severe coronary involvement in Kawasaki disease is giant aneurysms (>8-mm diameter), which are more likely to develop flow-limiting lesions and/or thrombosis (Fig 1⇑). The majority of coronary lesions improve without leading to thrombi or significant residual stenosis. Children with coronary involvement as a sequelae of Kawasaki disease generally have normal exercise tolerance and are asymptomatic19 20 ; however, it is not known if the coronary flow reserve in these patients is normal. The ideal modality for risk stratification and follow-up has not been defined.
Japanese centers have assessed coronary flow, using stress nuclear perfusion scans.23 24 Thallium SPECT has been the most popular technique. The advantages of exercise echocardiography over thallium exercise study are that it is noninvasive, provides immediate information, and has no radiation exposure or intravenous line. Exercise stress echocardiography can be performed at most institutions rather than only at specialized centers that have the capability to provide nuclear cameras and handle radioactive tracers. The test has minimal risk of injury or side effects, and no patient suffered any adverse effects. Patients with a history of Kawasaki disease with coronary involvement can undergo stress echocardiography serially, with coronary angiography reserved for those patients with large aneurysms, a poor ultrasound window, or an identified wall motion abnormality.
Patients with Kawasaki disease who develop myocardial infarction are usually asymptomatic before the event1 ; thus, it is crucial that patients at risk be identified and followed closely, noninvasively if possible, so that appropriate surgical intervention can be undertaken before infarction or sudden death. This technique also may be useful in serially following children who have been surgically revascularized, as is being developed for adult patients.
The most important limitation of the technique we describe is the rapid return to resting heart rate after exercise, which can occur within l minute in some well-conditioned children. If images are not obtained very rapidly, myocardial redistribution may occur; thus, a significant stenosis may not be detected and lead to a false-negative study. This problem is not encountered with thallium SPECT techniques, in which rapid return to resting heart rate does not affect the study.
A second limitation is that the reviewers of the stress echocardiograms were not blinded to the coronary status of the patients, and the interobserver and intraobserver variability was not addressed in this pediatric population. One of the authors (F.C.) has assessed this in an adult population of 1000 patients (personal observations/data, F.A. Chaudhry).
Another limitation to this study is that the majority of the 15 patients who had normal stress echocardiograms and now normal-appearing coronary arteries by echocardiographic study (group 1) have not undergone cardiac catheterization. Although we believe that they are very unlikely to have any significant flow-limiting coronary stenotic lesions, we cannot conclude that their coronary arteries are entirely normal and have elected to continue low-dose aspirin therapy. We see no indication, however, to restrict these patients from exercise activities.
A comparison study that includes stress echocardiography, stress thallium, and coronary angiography should preferably be done as a multicenter trial to verify the accuracy and utility of this technique in Kawasaki patients.
Exercise stress echocardiography was performed in a pediatric population with history of Kawasaki disease. Despite rapid fall in heart rate after they completed a treadmill protocol, rapid image acquisition allowed an adequate study in children. Two asymptomatic patients with normal stress ECGs developed wall motion abnormalities on postexercise imaging and had severe coronary stenosis on selective coronary angiography. We conclude that exercise stress echocardiography is safe and feasible and may be an important addition to the modalities used to identify children at risk for myocardial ischemia after Kawasaki disease.
The authors wish to thank Kathleen Corydon, RN, and Anne Rowley, MD, for assistance in patient recruitment, Carol Voda-Jones for technical assistance, Sang C. Park, MD, for photographic assistance, and Arlee Frantz for assistance with manuscript preparation.
- Received April 29, 1994.
- Accepted July 31, 1994.
- Copyright © 1995 by American Heart Association
Kato H, Inoue O, Akagi T. Kawasaki disease: cardiac problems and management. Pediatr Rev.. 1988;9:209-217.
Suzuki A, Kamiya T. Visualization of the coronary arterial lesions in Kawasaki disease by coronary angiography. Cardiol Young.. 1991;1:225-233.
Quinones MA, Verani MS, Haichin RM, Mahmarian JJ, Suarez J, Zoghbi WA. Exercise echocardiography versus 201TI single-photon emission computed tomography in evaluation of coronary artery disease. Circulation. 1992;85:1026-1031.
Ettedgui JA, Neches WH, Pahl E. The role of cross-sectional echocardiography in Kawasaki disease. Cardiol Young.. 1991;1:221-224.
Rowland TW. Aerobic exercise testing protocols. In: Pediatric Laboratory Exercise Testing. Champaign, Ill: Human Kinetics Publishers; 1993.
Kondo C, Hiroe M, Nakanishi T, Takao A. Detection of coronary artery stenosis in children with Kawasaki disease: usefulness of pharmacologic stress 201-TL myocardial tomography. Circulation. 1989;80:615-624.