Long-term Consequences of Kawasaki Disease

Kawasaki disease is an acute febrile illness with mucosal inflammation, skin rash, and cervical lymphadenopathy recognized most often in children younger than 4 years of age.¹² This is an acute vasculitis syndrome of unknown etiology that mainly affects small and medium-sized arteries, particularly the coronary artery. The long-term clinical issues in KD are concerned with the coronary artery lesions that develop aneurysmal formation, thrombotic occlusion, progression to ischemic heart disease, and premature atherosclerosis.³⁴⁵ In Japan as well as in North America, KD is presently a leading cause of acquired heart disease in children.⁶ It has been 25 years since KD was first reported, and a substantial number of KD patients have grown to adulthood. The coronary artery sequelae of these early patients with KD have become part of the pool of adult ischemic heart disease.⁷ However, the long-term consequences and the natural history of the cardiovascular sequelae in KD remain uncertain. In the present report, we present our comprehensive long-term follow-up study to clarify the natural history of KD patients who have been managed in the pre-IVGG treatment era.

Methods

Patients

Between January 1973 and December 1983, we examined 598 patients with acute KD. These included 383 boys and 215 girls from 1 month to 9 years old at onset (mean, 1 year 3 months). All patients were treated with 30 mg/kg of aspirin during the acute stage of the illness.⁸ No high-dose IVGG was given because all patients suffered from KD before the first report of this treatment.⁹ All patients except 4 were examined for coronary artery lesions by CAG performed within 3 months of onset. The 4 patients who refused CAG had not demonstrated any cardiac symptoms on physical examination, chest X-ray, or ECG and were excluded from our study. One hundred and forty-six patients (24.6%) were diagnosed as having coronary aneurysms; their age ranged from 1 month to 9 years (mean, 1.4 years). These 146 patients with coronary aneurysms received 5 mg/kg of aspirin after the acute stage of the illness while their coronary artery lesions were persisting.¹⁰

Follow-up

All patients except the drop-out cases have been followed up from their acute stage of KD for 10 to 21 years (mean, 13.6 years). During the first 5 years of their illness, almost all the patients were followed up; however, during the next 5 years, some, particularly those with normal coronary artery findings, dropped out from the follow-up study. Eventually, 203 patients (34.2%) dropped out from this long-term follow-up study. These 203 patients included 190 (42.4%) of the 448 with normal coronary artery findings, 10 (14.3%) of the 70 with regressed coronary aneurysms, and 3 (8.6%) of the 35 with persistent coronary aneurysms. A total of 133 (91%) of the 146 with coronary aneurysms in the acute stage have been followed up for more than 10 years at our clinic at intervals from 1 month to 1 or 2 years, depending on the severity of their lesions. Patients with small or medium-sized coronary aneurysms or with regressed aneurysms have been regularly examined every 6 months. Patients with giant aneurysms and with stenotic lesions in the coronary artery have been examined every 1 to 3 months at our clinic. They have received either a combination of 5 mg/kg aspirin and 3 mg/kg dipyridamole or 5 mg/kg aspirin and 0.1 to 0.2 mg/kg warfarin sodium per day. Patients with no coronary artery lesions were examined every year for the first 3 years and then every 2 or 3 years thereafter. In these patients, 5 mg/kg aspirin was administered for 2 months after onset and was then discontinued.

Cardiac Evaluations

Coronary Angiography

The presence of coronary artery lesions in acute KD was determined by an initial CAG performed at 22 days to 3 months (mean, 34 days) after onset in all but 4 of the 598 patients.¹¹ Their age at the initial CAG ranged from 1 month to 9 years (mean, 1 year 2 months). A 6F or 7F sheath was inserted percutaneously into the femoral artery, and a 5F or 6F Judkins-Kato catheter for pediatric coronary angiography (Cook Co) was advanced through the sheath up to the ostium of the right or left coronary artery by use of the Judkins technique. Cineangiography in the left and right coronary arteries was performed with manual injection of iohexol into the coronary artery. Then, the right anterior oblique 30° projection, the posterior anterior projection, the left anterior oblique 60° projection for CAG, and a left ventriculogram were taken by use of a cineangiography system (model KXO-2050, Toshiba Corp). In the 15 patients <3 months of age for whom selective CAG was difficult to perform, cineaortography was recorded by injection of dye into the ascending aorta just above the aortic valve.

A second CAG was performed at 1.7±2.1 years after the initial CAG for the 146 patients previously found to have coronary aneurysms.¹¹ A third CAG was performed for 64 patients including 58 with persistent coronary aneurysms with or without stenosis and 6 with regression of coronary aneurysms at the second CAG; this CAG was performed at 6.8±3.0 years after onset. A fourth CAG was performed at 9.6±2.0 years after onset in 29 patients who had abnormal findings at the third CAG. For 17 patients, a fifth CAG was performed at 15.2 years after onset. Thirty-four patients received follow-up CAG after 10 to 21 years of illness; this included 25 patients with persistent coronary aneurysms with or without stenotic lesions, 7 with regressed aneurysms, and 2 with normal coronary findings at the first CAG but who later complained of chest pain during the follow-up. These 2 patients demonstrated normal findings on the follow-up CAG.

CAG findings were classified according to the “Guidelines for Coronary Lesions in KD” reported by the Research Committee on Kawasaki Disease.¹²¹³¹⁴ Dilated lesions were divided into three groups according to their internal diameter as mild dilatation or small aneurysms (<1.5× normal, or <5 mm), moderately sized aneurysms (between 1.5 and 4× normal, or between 5 and 8 mm), or giant aneurysms (4 or more times normal, or >8 mm). Stenotic lesions were divided into three groups as occlusion (100% obstruction), segmental stenosis, or localized stenosis.

Assessment of Systemic Artery Involvement

During the first 6 years (1973 through 1978), 217 consecutive patients were examined regularly by use of aortography to determine any systemic artery involvement, and 3 (1.4%) were found to have systemic artery aneurysms. These 3 patients also had giant coronary aneurysms. Only patients with coronary aneurysms developed systemic artery aneurysms. On the basis of these observations, in 1979, we began to limit our examination for systemic artery involvement to those with a coronary aneurysm shown by test injection of dye into the systemic artery; if abnormal findings were discovered, we performed cineangiography of the involved systemic artery. During the next 5 years (between 1979 and 1983), systemic artery aneurysms were found in 10 (2.6%) of 377 patients.

Two-dimensional Echocardiography

We adopted two-dimensional echocardiography for evaluation of coronary aneurysms beginning in 1979⁵ using an Aloka SSD-870 with a 5-MHz transducer. Serial echocardiograms were taken at least three times during the acute stage of the illness: on admission, at ≈10 days of illness, and at 2 or 3 weeks of illness. The two-dimensional echocardiographic examination included display of the main trunk, anterior descending and circumflex branches of the left coronary artery, and the main and distal segments of the right coronary artery by standard manipulation, as described previously.¹⁵ The coronary artery morphology in echocardiography was evaluated in the same manner as in coronary angiography. The echo density in the coronary arterial wall or in the region of an aneurysm was investigated further for calcification and changes in tissue characterization compared with a normal portion of the arterial wall by the method of changing the attenuation of the echo gain. Left ventricular function and the presence of pericardial effusion were also investigated. Valvular regurgitation and the function of the cardiac valves were determined by use of Doppler echocardiography. Repeat two-dimensional echocardiographic studies were performed at follow-up.

Other Cardiac Evaluations

Dipyridamole stress ²⁰¹Tl myocardial scintigraphy or an exercise stress test by the Bruce protocol usually was performed during follow-up for patients >4 years of age with giant coronary aneurysms or stenotic lesions in the coronary arteries.¹⁶ If patients demonstrated findings that suggested myocardial ischemia, a follow-up examination was performed several months to several years later, depending on the patient's condition.

Results

Serial Coronary Angiography

Outcome of Coronary Aneurysms

The outcome of coronary aneurysms and the clinical course of the 594 patients during our 10 to 21 years of follow-up are shown in Fig 1. In the acute stage, 146 patients (24.6%) were diagnosed as having coronary aneurysms. At the second CAG 6 to 18 months later, 72 (49.3%) of these patients showed completely normal findings and demonstrated regression of the aneurysms. Fig 2A shows the time and incidence of regression, and Fig 2B shows progression to stenotic lesions in these 146 patients. Overall, 80 (54.8%) of these patients, or 13.5% of all 594 KD patients, demonstrated regression in the coronary aneurysms during follow-up. All 80 had coronary aneurysms that were small or moderate in size. No giant coronary aneurysms showed regression. Regression in coronary aneurysms was recognized within 2 years from onset in 90% of the 80 patients with regressed coronary aneurysms. Seven of these 80 patients with regressed coronary aneurysms received a follow-up CAG within 10 to 21 years of illness, which demonstrated no progression to stenotic lesion in the coronary artery in all 7 cases. Fig 3 shows a representative case.

Overall, 28 patients (4.7% of all KD patients), including 5 who died, showed progression to coronary artery stenosis. Stenotic lesions were recognized within 2 years from onset in about half of these 28 patients. However, in the others, the stenosis developed gradually over several years. The longest time from onset to stenosis was 17 years.

Giant coronary aneurysms were found at the first CAG in 26 patients (4.4% overall; 20 boys and 6 girls), 12 of which progressed to stenosis or complete obstruction of the coronary artery. Myocardial infarction occurred in 8 of these 12 patients, and 4 of these 8 died. The other 14 patients showed persistent coronary aneurysms without significant stenosis in the coronary artery over a follow-up period of 10 to 21 years. Fig 4 shows CAGs of a patient in whom stenotic lesions developed 11 years after disease onset.

After the introduction of two-dimensional echocardiography in 1979, we recognized mild transient dilatation in the coronary artery that lasted 7 to 10 days during the acute stage in about 40% of those patients examined. In these patients, CAG was performed just after the acute stage and demonstrated normal angiographic findings. In the present study, such patients were managed as having a normal coronary artery and did not receive further CAG.

In those with no coronary lesions and those with regressed aneurysms, regular examinations that included the exercise stress test and ²⁰¹Tl myocardial scintigraphy found no abnormal cardiac signs or ischemic findings. However, one patient with aortic regurgitation and a normal coronary artery revealed ischemic findings on ²⁰¹Tl myocardial scintigraphy. Half of the patients with regressed coronary aneurysms continued to demonstrate a dense or bright echo in the coronary artery wall on two-dimensional echocardiography.

Systemic Artery Involvement

Thirteen patients (2.2%), including 10 boys and 3 girls, had systemic artery aneurysms. Their age at onset of KD ranged from 3 months to 2.3 years (mean, 0.8±0.6 years). The lesions in the systemic artery appeared in the axillary artery (11 patients), common iliac artery (9 patients), renal artery (5 patients), subclavian artery (4 patients), internal iliac artery (7 patients), superior mesenteric artery (1 patient), internal thoracic artery (1 patient), and femoral artery (1 patient). Each of these patients also had multiple giant coronary aneurysms. Two of these patients subsequently died of acute myocardial infarction. Eight (62%) of these 13 showed natural regression in the systemic artery lesions similar to that frequently seen in the coronary aneurysms. A few systemic artery aneurysms progressed to obstructive lesions, particularly in the iliac artery. However, no ischemic symptoms developed.

Clinical Course

Patients With No Coronary Artery Lesion

A total of 448 patients were found to have a normal coronary artery in the initial CAG. One hundred and ninety of these (31.9% of the overall patients studied) dropped out from a long-term follow-up, and the other 258 (43.4% of the overall patients studied) have been followed up for more than 10 years. During this long-term follow-up, all 258 patients showed no cardiac symptoms or abnormal findings in routine cardiac examinations that included chest x-ray, ECG, and two-dimensional echocardiography. In 10 patients with a normal coronary artery and in 54 with regressed coronary aneurysms, the thallium myocardial scintigraphy and the exercise stress test were performed during follow-up and demonstrated normal findings.

Myocardial Infarction and Fatal Cases

There were 11 patients (1.9% overall), including 9 boys and 2 girls, who developed myocardial infarction. Each of these 11 patients had severe stenotic lesions in two or three vessels of the coronary artery. Eight (73%) of these 11 patients previously had giant coronary aneurysms, and the other 3 had moderately sized aneurysms on the first CAG. Four (36.3%) of these patients had myocardial infarction within 1 year and an additional 5 (45.4%) within 3 years after onset. Five (45.4%) of the 11 patients had asymptomatic myocardial infarction, and 1 of these died of recurrent myocardial infarction. The other 6 patients demonstrated symptomatic myocardial infarction. Four of these 6 developed recurrent myocardial infarction and subsequently died. Asymptomatic myocardial infarction was diagnosed by the appearance of abnormal deep Q waves, a significant obstruction in the coronary arteries on CAG, and abnormal thallium scintigraphy. Overall, 5 patients (0.8%) died at 3 months to 7 years after onset (mean, 4.3±1.6 years).

Valvular Heart Disease

Six patients (1.0% overall) showed mitral valve regurgitation detected by auscultation of the heart and Doppler echocardiography in the acute or subacute stage of KD (7 to 35 days of illness; mean, 20 days). In three patients, the mitral regurgitations spontaneously disappeared between 1 month and 7 years from onset, whereas in two others, it progressed to myocardial infarction and resulted in death. The remaining patient has continued to have persistent mitral regurgitation for 13 years, probably due to papillary muscle dysfunction of the mitral valve. Aortic valve regurgitation developed in one patient (0.17%) at 7 months after onset and progressed for 12 years. This patient developed progressive heart failure and received aortic valve replacement with a St Jude medical valve at 13 years after onset.

Coronary Intervention and Bypass Surgery

The seven patients (1.2%) who had giant aneurysms with severe stenosis of the coronary arteries underwent coronary artery bypass surgery. Two of these patients previously had a myocardial infarction. A saphenous vein graft was used in two patients, an internal thoracic artery graft in five, and a right gastroepiploic artery graft in three. In all seven patients, the patency of the bypass grafts was confirmed by follow-up CAG, and their quality of life after surgery has been favorable.

Coronary thrombolytic treatment¹⁷ by direct infusion of urokinase or tissue plasminogen activator was performed in 16 patients, most of whom had giant coronary aneurysms. Four of these patients received coronary thrombolytic treatment within 6 hours of acute myocardial infarction; of the remaining 12 patients, 4 received thrombolytic therapy because of previous myocardial infarction, 3 for angina pectoris, and 5 for massive thrombus formation in giant coronary aneurysms without ischemic findings. In all 4 patients with acute myocardial infarction, 2 of the 3 with angina pectoris, and all 5 with massive thrombus formation, the thrombolytic treatment was effective, as confirmed by two-dimensional echocardiography or CAG.

In two patients, balloon angioplasty was performed for severe stenotic lesions in the coronary artery, one at 12 and the other at 15 years after onset of KD. Both were examined by use of intravascular ultrasound imaging. One patient was treated successfully with an ASUKA balloon catheter 4 mm in diameter. However, in the other patient, the severity of the stenosis with circumferential calcification caused the balloon angioplasty to fail, and a rotablator was used, which resulted in excellent revascularization.

Discussion

Long-term Prognosis of Patients With a Normal Coronary Artery

Four hundred and forty-eight (75.4%) of the 594 patients demonstrated a normal coronary artery at the first CAG. Although a considerable number (190, or 42.4%) of these patients dropped out from the follow-up, we were able to follow up 258 patients for 10 to 21 years using routine ECGs, two-dimensional echocardiography, chest x-ray, an optional exercise stress test, and thallium myocardial scintigraphy. None of these 258 patients demonstrated any cardiac symptoms or ischemic findings during the follow-up. Paridon et al¹⁸ reported a high incidence of a myocardial perfusion defect on exercise stress SPECT, even in patients with a normal coronary artery. However, we experienced only 1 case of a perfusion defect on the dipyridamole stress ²⁰¹Tl SPECT among 30 patients with normal coronary artery angiograms. Our single case also had aortic regurgitation.¹⁶ We think patients with a normal coronary artery may have had no morbidity in childhood, provided their coronary artery lesions were correctly evaluated by CAG or two-dimensional echocardiography during the acute stage of KD. It is uncertain whether it is necessary to follow up such patients with a normal coronary artery over the long term. However, we currently continue to examine these patients every 3 to 5 years.

The natural history or long-term pathological sequelae of patients with transient dilatation of the coronary artery is uncertain at the present time. When serial two-dimensional echocardiography was adopted in 1979, it became evident that transient mild dilatation appeared during the acute stage in about 40% of KD patients. In these patients, CAG was performed just after the acute stage and demonstrated normal angiographic findings. Therefore, in the present study, those with transient dilatation were included in the group with a normal coronary artery. The patients with transient dilatation were followed up for more than 10 years and demonstrated no cardiac symptoms or ischemic findings. Although these findings suggest that transient dilatation may present no clinical problem, these patients should be followed up over the long term to screen for coronary artery disease or development of premature atherosclerosis in adulthood.

Follow-up of Coronary Aneurysms

The long-term natural history of coronary aneurysms is a particularly important issue in KD. A total of 146 (24.6%) patients with acute KD had coronary aneurysms. Approximately 50% of those in our series who had coronary aneurysms have shown regression on follow-up angiography. This is a characteristic phenomenon in Kawasaki vasculitis, first described by us in 1975.⁴¹¹ Regression is likely to occur within 1 or 2 years after onset, and it is unlikely to occur more than several years after onset. All patients with regression demonstrated no cardiac symptoms and no ischemic findings on ECG, exercise stress test, or thallium myocardial scintigraphy and no change in left ventricular function during the long-term follow-up. Whether regression may reverse and eventually develop into stenosis is uncertain at the present time. Suzuki et al¹⁹ mentioned that this was possible in the later stage of illness. In our series, 7 of the 70 patients with regressed coronary aneurysms had another CAG more than 10 years after onset, and this CAG demonstrated normal findings, with no stenosis and no irregular arterial wall. Our experience indicates that a stenotic lesion does not develop, at least within 1 or 2 decades, in patients who have demonstrated aneurysmal regression.

However, in our series, the regressed coronary aneurysms demonstrated a marked thickening of the intima with or without calcification that bore a striking resemblance to early atherosclerotic lesions.²⁰ Recently, our studies that used intravascular ultrasound imaging have demonstrated similar findings.²¹ Also, we have found that the coronary artery wall at the site of the regressed aneurysm had poor distensibility and was stiff on examination with intracoronary isosorbide dinitrate.²² Therefore, the sequelae of Kawasaki vasculitis may include a long-term coronary risk factor even in patients with regressed aneurysms, and atherosclerotic lesions might develop in adulthood. These patients should be followed up carefully and educated to avoid other atherosclerogenic risk factors, such as smoking, obesity, hypertension, and hypercholesterolemia.

Eight of 11 patients with myocardial infarction had giant coronary aneurysms. We examined a total of 26 patients with giant coronary aneurysms, of whom 12 developed stenotic lesions and none showed regression. These findings suggest that giant coronary aneurysms have a strong potential for development to ischemic heart disease. Formerly, we had used low-dose aspirin in combination with dipyridamole for patients with giant coronary aneurysms. However, we could not prevent thrombotic formation. Currently, we use low-dose aspirin in combination with warfarin, although it is uncertain at the present time whether this regimen can prevent thrombotic formation.

Recently, high-dose λ-globulin has been used frequently in the treatment of KD,⁹²³ and the incidence of coronary aneurysms and cardiovascular sequelae has declined. However, a long-term follow-up study on patients who have received high-dose γ-globulin has not yet been undertaken, and this should be done.

Systemic Artery Involvement

Axillary artery aneurysms were common lesions among those with systemic artery involvement and almost always were associated with coronary aneurysms. When axillary artery aneurysms are palpated on physical examination in the acute stage, the presence of coronary aneurysms, primarily giant aneurysms, is presumed. Systemic artery aneurysms also had a tendency to regress, as did coronary aneurysms. Stenotic lesions in the systemic arteries were likely to develop in the iliac arteries, but such lesions did not demonstrate any ischemic findings. There have been some reports of digital gangrene,²⁴ but we have had no such patients. We had seven patients with renal artery aneurysms that subsequently regressed and did not progress to renovascular hypertension. The prognosis for systemic artery involvement seems to be favorable.

Valvular Heart Disease

It has been reported²⁵²⁶ that mitral or aortic regurgitation has developed during the clinical course of KD. However, little information is available on its natural course in the long term.²⁷ In our study, six patients (1% of all KD patients) had mitral valve regurgitation, which appeared primarily during the acute or subacute stage. In three patients, mitral regurgitation disappeared within 1 month to 7 years from onset. However, two of these patients later had a myocardial infarction and died. The cause of this condition may be valvulitis or papillary muscle dysfunction. In one patient, mitral valve regurgitation had persisted for 13 years, and the cause may have been of ischemic origin. Mitral regurgitation due to valvulitis may disappear in the long term. Aortic regurgitation appears in rare instances, and we have experienced only one case, which persisted and progressed over 12 years. Valvular heart disease may develop in the long term in some instances.

Issues in Adult Cardiology

It has now been >25 years since Kawasaki first described KD in 1967, and a certain number of the early patients have now grown to adulthood. We⁷ reported previously that coronary artery sequelae of early KD patients have led to ischemic heart disease in young adults. In our series, 70 of the 594 patients are now over 20 years old, and 5 of these have stenotic coronary artery lesions. These patients had coronary aneurysms with stenosis and calcification. Two patients have already had coronary bypass surgery. The coronary artery sequelae of KD may become an important cause of ischemic heart disease in young adults, and adult cardiologists should be aware of this condition and determine any history of childhood KD when they examine a patient with ischemic heart disease, particularly one <40 years old. Another potential problem is that the sequelae of Kawasaki vasculitis may develop into premature atherosclerosis in adulthood. KD was originally a disease in infants and young children, and its cardiovascular lesions are now an important cause of acquired heart disease in children. However, advanced problems of long-term sequelae appear to be developing and encroaching on the field of adult cardiology. Additional long-term follow-up studies are continuing and are essential.

Selected Abbreviations and Acronyms

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Footnotes