Twenty-Five-Year Outcome of Pediatric Coronary Artery Bypass Surgery for Kawasaki Disease
Background— The long-term outcome of pediatric coronary artery bypass for patients with severe inflammatory coronary sequelae secondary to Kawasaki disease is unknown.
Methods and Results— One hundred fourteen children and adolescents ranging in age from 1 to 19 (median, 10) years at operation were followed up for as long as 25 years with a median of 19 years. The number of distal anastomoses was 1.7±0.8 per patient, and the internal thoracic artery was used in all but 3, most frequently for left anterior descending artery lesions. Saphenous vein grafts were used in 24 patients, mostly for non-left anterior descending artery lesions. Patients underwent multiple angiograms to evaluate their coronary and graft status. There was no operative or hospital mortality. Both 20- and 25-year survival rates were 95% (95% confidence interval [CI], 88 to 98). Five deaths occurred, all cardiac in origin. Cardiac event-free rates at 20 and 25 years were 67% and 60% (95% CI, 46 to 72), respectively. Percutaneous coronary intervention and reoperation were the most common events. Overall, the 20-year graft patency rate was 87% (95% CI, 78 to 93) for internal thoracic artery grafts (n=154) and 44% (95% CI, 26 to 61) for saphenous vein grafts (n=30) (P<0.001), and the rate for non-left anterior descending artery lesions was also significantly better for arterial grafts (87% [95% CI, 73 to 94]; n=59) than for saphenous vein grafts (42% [95% CI, 23 to 60]; n=27) (P=0.002). Eighty-eight patients (77%) remain on medications, but all 109 survivors are presently symptom free in their daily activities.
Conclusions— Although the 25-year survival was excellent after pediatric coronary bypass for Kawasaki disease, the event-free rate declined progressively. This reality mandated continued follow-up. Reinterventions successfully managed most cardiac events. An internal thoracic artery graft was the most favorable for children.
Received December 4, 2008; accepted April 21, 2009.
The long-term outcome for children and adolescents who underwent coronary bypass surgery for severe coronary sequelae of Kawasaki disease (KD) is unknown. We reported the first pediatric coronary artery bypass with the use of a saphenous vein graft (SVG) for this disease in 19761 and the first with the use of the internal thoracic artery (ITA) in 1985.2 Since then, we have observed 114 children and adolescents who underwent bypass surgery at Nara Medical College or the National Cardiovascular Center with a 100% follow-up rate up to a maximum of 27 years. Clearly, after 20 years, all pediatric patients have reached adulthood, in which atherosclerotic disease may involve the coronary artery and graft.3–5 We describe the long-term outcome of these 114 patients after their surgery for KD coronary complications, presently the world’s most common cause of pediatric coronary artery disease.6
Editorial see p 6
Clinical Perspective on p 68
One hundred fourteen children and adolescents aged <20 years underwent a coronary artery bypass operation for inflammatory coronary obstructive lesions in association with aneurysms as sequelae of KD, on the basis of our surgical indications,7,8 which have not changed significantly for the last 25 years. Their age at operation varied from 1 to 19 years, with a median age of 10 years; 86 were male and 28 female. There were 57 patients aged <10 years and 57 patients aged ≥10 years. Female patients constituted 25% of all subjects, and gender distribution did not differ among the patients aged <10 or ≥10 years (female gender, 19% versus 30%; P=NS). The age at onset of acute symptoms strongly suggestive of KD varied from 0 to 15 years, with a median of 1 year, with the highest incidence (45%) during the first year of life. Thirty-six patients (32%) had had a previous myocardial infarct (MI) with abnormal Q waves at the time of operation, including anteroseptal MI in 12, inferior in 18, anterolateral in 4, and inferolateral in 2 patients, as confirmed by ECG. A preoperative left ventricular ejection fraction (EF) of <0.5 was identified in 16 patients (14%) and was always associated with a preceding MI. Coronary artery involvement, including both obstructive and aneurysmal lesions, was 2.1±0.6 vessels per patient, ranging from 1 to 3. Left main trunk lesions were present in 52 patients (46%) in whom aneurysms predominated over obstructive lesions, the latter being found in only 1 patient. Obstructive lesions were located primarily at the distal outlet portion of the aneurysm and much less often in its proximal inlet portion.8 None of the patients had a percutaneous coronary intervention (PCI) before operation.
The time interval from acute KD to operation also varied significantly, from <1 year to 19 years, with a median of 6 years after the acute illness. After surgery, follow-up was 100% in these patients and was >20 years in 36 patients (32%) and 15 to 20 years in 69 (61%). The median follow-up duration was 19 years. Multiple selective angiographic studies were conducted in all patients, generally before operation, within 1 month after operation, at 1 year, at 5 to 10 years, at 10 to 20 years, at >20 years, and in the event of cardiac episodes. Determination of graft patency depended on these studies. Cardiac events included deaths from any cause, acute MI, an episode of syncope, angina pectoris, ventricular tachyarrhythmias, and requirement for further intervention (PCI or reoperation). Deaths included a traffic accident because of suspected syncope during motorcycle riding. Graft failure itself was not considered a cardiac event. Demographic data of these patients are summarized in Table 1.
Indications for surgery in KD have been reported and debated,7,8 but there has been little change for the last 25 years. Briefly, our indications were limited to angiographically significant obstructive coronary lesions, usually stenosis ≥75% and preferably ≥90%, with the presence of clinical signs of myocardial ischemia or positive signs of ischemia by ECG or scintigraphic examination with either exercise or pharmacological interventions such as dipyridamole. Surgical treatment was strongly recommended for children with a previous MI because prognosis after MI was reported to be unfavorable.9
The presence of a large aneurysm, particularly in the left main trunk, was certainly a difficult issue for surgery, but when coronary arteries adjacent to aneurysms were not associated with angiographically significant stenosis, surgery has not been indicated in our series because of a high possibility of arterial graft regression or failure (string sign). Medical anticoagulation therapy was continued with frequent follow-up. We elected not to ligate or interrupt the proximal native coronary artery segment to ensure maximal flow through the arterial graft to prevent graft regression because we were afraid of hypoperfusion in the immediate postoperative period due to limited flow through the narrow ITA of pediatric patients.
The ITA was used as a pedicle (in situ) graft in all patients. The SVG was harvested primarily from the thigh. Surgery was performed under cardiopulmonary bypass and cardioplegic arrest in all patients with the use of either crystalloid or blood cardioplegic solution. ITA-coronary anastomosis was performed with the use of magnifying glasses (×4.0), surgical microscopes (×8.0), or both and with the use of 8-0 monofilament continuous and some interrupted sutures. For 114 patients, a mean of 1.7±0.8 grafts per patient was anastomosed: 1 graft for 49 patients (43%), 2 grafts for 53 (46%), 3 grafts for 9 (8%), and 4 or 5 grafts for 3 (3%). Two patients had a concomitant mitral valve repair or replacement with a mechanical prosthesis for coexisting severe mitral regurgitation.10
One hundred eleven patients (97%) had an ITA graft, and 90 patients (79%) had only arterial grafts (right or left ITA, right gastroepiploic artery [GEA], or both). Bilateral ITAs were used in 43 patients.11 Twenty-four patients (21%) had a SVG, primarily in association with an ITA in 21 patients.
A total of 198 grafts were placed, with 111 left ITAs, 43 right ITAs, 30 SVGs, and 14 GEAs for 120 left anterior descending (LAD) or diagonal arteries, 57 right coronary arteries (RCA) or posterior descending arteries, and 21 left circumflex arteries (LCX). Target vessels for each graft are shown in Tables 2 and 3⇓. In this series, PCI was not attempted before the operation because we believe that surgery is safer for complicated coronary lesions in children.
Quantitative values are expressed as mean±SD or median with range. The time-related survival, event-free, and graft patency rates were calculated by the Kaplan-Meier method with 95% confidence interval (CI). Differences of graft patency rates were assessed by log-rank test, and differences were considered statistically significant at P<0.05. Different grafts within the same patient were treated as separate entities because graft patency has been proven to depend on graft materials and target vessels within a single patient.12 We used the Cox proportional hazards model to assess the association between variables and the incidence of cardiac events. Variables included gender, age at KD onset, age at operation, presence or absence of previous MI, completeness of revascularization (the number of grafts minus the number of lesions; −2 to +2, as shown in Table 3), left main trunk involvement, presence of SVG, graft patency, and left ventricular EF ≥0.5 or <0.5. Statistical analyses were performed with SAS statistical software, version 9.1 (SAS Institute, Inc, Cary, NC).
Long-Term Survival Rates
There were no operative or hospital deaths, but 5 late deaths occurred (4%), 4 males and 1 female, from 3 months to 12 years after the operation. Demographic data of the 5 deaths are shown in Table 4. Death was sudden in 3 patients, and 1 death was due to cardiac allograft vasculopathy after heart transplantation for severe left ventricular failure unresponsive to bypass surgery. All 5 had a history of previous MI resulting in a reduced left ventricular EF of ≤0.5 in addition to frequent ventricular arrhythmias. ITAs were used in all 5 late death patients, and patency of the graft had been confirmed before death. Among survivors, no one needed a heart transplant. The 10-, 20-, and 25-year survival rates were 98% (95% CI, 93 to 100), 95%, and 95% (95% CI, 88 to 98), respectively (Figure 1).
Cardiac Event-Free Rate
Cardiac events occurred 36 times in 33 patients (29%), including the aforementioned deaths, nonfatal acute MI in 1 patient, angina pectoris in 1, syncope in 1, ventricular tachyarrhythmias in 2, and indications for postoperative PCI or redo bypass operation in 26 patients (23%). Two patients with ventricular tachyarrhythmias received an implantable cardiac defibrillator. PCI was performed in 17 patients (15%) and reoperation in 9 (8%). Three patients experienced >1 event. The 5-, 10-, 20-, and 25-year cardiac event-free rates were 87%, 81%, 67% (95% CI, 55 to 76), and 60% (95% CI, 46 to 72), respectively (Figure 2).
Indications for Postoperative Reinterventions (PCI or Reoperation)
Indication for further intervention was the most frequent cardiac event. When significant stenosis ≥75% or obstruction of the graft and/or the native coronary artery was angiographically identified, exercise ECG, echocardiography, and myocardial scintigraphy were repeated as needed. Even in the absence of subjective symptoms, PCI or redo operation was undertaken to correct anatomic lesions when studies confirmed the presence of stenosis ≥75%, ischemia, or regional contraction abnormality of the left ventricle related to obstructive coronary lesions.
Of the 26 patients who underwent reinterventions, 17 underwent PCI 9±6 years after the operation (aged 17±9 years). Six patients had a percutaneous transluminal coronary rotational ablation: 3 for original coronary lesions because of graft failure and 3 for new obstructive coronary lesions that had been considered insignificant at the first operation and were left untreated. Ten patients had percutaneous transluminal balloon angioplasty for ITA-coronary artery anastomotic stenosis, and 1 patient in adulthood had a stent inserted for SVG midportion stenosis with atherosclerosis. Nine patients had a coronary reoperation at 22±8 years of age, 13±8 years after the first operation. The indications for reintervention, postoperative term at reintervention, and their success rates are shown in Table 5. When the cardiac event-free rate from reintervention (PCI and reoperation) was analyzed separately from other cardiac events, the reintervention-free rate at 20 years was 75% (95% CI, 64 to 84), and that at 25 years was 68% (95% CI, 52 to 79) (Figure 3). The early postoperative indication for reintervention was mainly percutaneous transluminal balloon angioplasty for ITA-coronary artery anastomotic stenosis that occurred 1.8±3.6 years after the operation. Newly developed coronary lesions (13.0±6.5 years) and SVG atherosclerosis (15 years) were late indications for PCI, stenting, and reoperation. In the midterm, graft obstruction due to graft shrinkage, intimal hyperplasia, and thrombosis was the most common reason for reinterventions. The success rate for reinterventions was acceptable. The right or left ITAs together with the GEA or radial artery were used for all reoperations to ensure total arterial grafting. None of the patients who needed additional treatment by either PCI or reoperation died. Analysis of the risk factor variables for cardiac events or postoperative reinterventions by the Cox proportional hazards model showed no significant associations.
Postoperative Left Ventricular Function
Postoperative left ventricular EF was measured in 72 patients (63%). The mean postoperative EF was 0.55±0.08. A depressed EF depended on a history of preoperative MI, and the EF at rest was essentially unchanged after surgery. All 5 subjects who died had an EF of ≤0.5 due to previous MI.
Medications After Surgery
Only 26 patients (23%) were medication free at their last follow-up. The remaining 88 patients (77%), including the patients who died, were receiving various combinations of 1 to 6 medications including warfarin, aspirin, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, calcium antagonists, diuretics, nitrates, statins, and β-blockers. All 88 patients were on aspirin, as recommended by the guidelines.6,13 Those with a giant left main trunk aneurysm not supported by 2 patent grafts for the LAD and LCX (n=21) were kept on both warfarin and aspirin, maintaining an international normalized ratio ≈2.0.6 Aspirin was given at 5 mg/kg daily, and the maximum dose was 100 mg per day.
Generally, medication-free patients were those with good left ventricular function and single-vessel disease with a patent ITA graft 3 to 5 years after the operation. Other patients with additional nonstenotic coronary aneurysms, SVG, or mild to moderately depressed left ventricular function with or without ventricular arrhythmias were placed on medications selected from those listed.
Postoperative Activity Restrictions
Although the cardiac event-free rate at 20 years after the operation was 67% and that at 25 years was only 60% (95% CI, 46 to 72), all 109 surviving patients were symptom free at their last assessment. In accord with the level of exercise advised by their cardiologist, 45 (41%) of the survivors were presently permitted all physical activities, including long-distance swimming and/or running, although all of them had been strictly prohibited from exercise or sport activities before the operation. Forty-two patients (39%) were allowed unlimited daily activities but were prohibited from long-distance swimming and/or running. Ten patients (9%) were prohibited from strenuous exercise, and 12 (11%) were recommended not to participate in athletic events. Six female patients had uneventful delivery of babies during their follow-up.
Long-Term Graft Patency
As demonstrated in Figure 4, the 20-year graft patency for 154 ITAs as a whole was 87% (95% CI, 78 to 93), and that for 30 SVGs was 44% (95% CI, 26 to 61). This difference was highly significant at P<0.001. The 5-year patency for the GEA was 86% (n=14), similar to the ITA, although the observation period was short for this graft.
Twenty-year patency of the ITA was 91% for the LAD (n=109), 100% for the LCX (n=10), and 84% for the RCA (n=31). In contrast, 20-year patency of the SVG was 67% for the LAD (n=3), 69% for the LCX (n=11), and 25% for the RCA (n=12). Because the ITA was used primarily for the LAD, and the SVG was used mainly for the RCA and LCX, graft patency rates for the non-LAD vessels are also compared in Figure 5. When we analyzed the patency of the ITA plus GEA and SVG for the non-LAD vessels, a significant patency difference between the arterial graft and SVG was again apparent (P=0.002). The arterial graft patency rate for the non-LAD vessels was 87% (95% CI, 73 to 94) (n=59) at 20 years after operation. In contrast, the SVG patency for the same target vessels was only 42% (95% CI, 23 to 60) (n=27). The ITA patency rate was similar for both the LAD and the non-LAD vessels. In Figure 6, differences in graft patency according to the age at operation (≥10 or <10 years) are shown. The long-term patency for the ITA was not significantly different regardless of the patient’s age (≥10 or <10 years) (93% [95% CI, 83 to 97] versus 86% [95% CI, 74 to 93]; P=0.163). In contrast, SVG patency was significantly lower for the patients aged <10 years (58% [95% CI, 32 to 78] versus 25% [95% CI, 6 to 51]; P=0.004).
Characteristics of the Graft
The 20-year patency of the SVG was as low as 44%, but 14 SVGs have remained patent with the aid of medications, PCI, and/or stenting at the time of last angiographic evaluation. However, most of the patent SVGs showed various degrees of degenerative changes (n=10; 71%),14 such as irregularity of the graft wall and caliber with a maximum/minimum diameter ratio >1.5 (n=6), shortening or growth failure of the SVG resulting in traction or kinking of the target coronary artery that jeopardized distal runoff flow (n=3),15 and apparent development of SVG atherosclerosis (n=1) before 40 years of age. On the other hand, the wall characteristics of the ITA were very smooth and appropriately size matched with the adjacent coronary artery and its flow situation. ITA graft failure due to flow competition (string phenomenon) reversed and the graft became patent in 4 patients when their native coronary lesions progressed over nearly 10 years.16 This reverse remodeling of the ITA graft increased the ITA patency rate by ≈3%. This phenomenon did not occur with SVGs. Growth of the ITA graft was also apparent in these patients.17,18
Survival for >20 years of children and adolescents after a bypass operation for severe coronary artery involvement secondary to KD is encouraging. In many patients, the coronary circulation was maintained by the graft, and in some, both coronary orifices were totally obstructed and coronary blood supply was completely dependent on enlarged ITA grafts.8,11 Less satisfactorily, the cardiac event-free curve slowly but progressively declined, being 67% after 20 years and only 60% (95% CI, 46 to 72) at 25 years (Figure 2). Among the cardiac events, the need for reintervention was the most common. The early decline of the reintervention-free curve (Figure 3) reflects ITA-coronary artery anastomotic stenosis necessitating percutaneous transluminal balloon angioplasty with excellent patency thereafter. The late decline at ≈15 to 20 years depended primarily on the development of new lesions, and SVG atherosclerosis and midterm reinterventions were attributed mainly to graft closure, as shown in Table 5. Analysis of variables by the Cox proportional hazards model, however, failed to demonstrate significant contributing factors among the variables we evaluated to the occurrence of cardiac events or to the need for reinterventions, probably because of the small sample size and the unpredictable fate of the remaining coronary lesions after surgery.
The number of coronary arteries involved, including both obstructive and aneurysmal lesions, was 2.1±0.6 lesions per patient, ranging from 1 to 3. In addition, left main trunk disease involved aneurysm formation in the majority of patients. The number of bypass grafts was 1.7±0.8 per patient. The difference in the number of diseased vessels and number of grafts was attributed to the presence of aneurysms without obstructive lesions that did not require a bypass graft at the first operation. However, additional interventions will be needed in the future because the remaining coronary aneurysmal lesions may evolve into new obstructive lesions with time. For now, all patients with patent aneurysms with no distal support by bypass grafts are kept on a combination of warfarin and aspirin to decrease the possibility of MI or sudden death.19
The ITA is the most important vessel for coronary revascularization in growing children and adolescents who will be entering the age of atherosclerosis in 20 to 30 years. Its importance is emphasized by evidence that it is apparently spared from atherosclerosis,20 and adult patients revascularized with this graft survive longer than those revascularized with a SVG.12 The importance of the ITA graft should be even greater in children7,8,21,22 because of its ability to grow in accordance with the somatic growth of the patient.17,18 In addition, ITA involvement in KD is extremely rare, quite different from the coronary artery. The effect of ITA grafts on survival may depend not only on graft longevity but also on the function of the ITA endothelium. Endothelial function of the coronary artery affected by KD is probably impaired and depressed3–5; thus, the supplemental endothelial function from the less affected ITA, such as production of nitric oxide and other bioactive products, may enhance and possibly reverse the depressed endothelial function of the diseased coronary artery.19,23–25
In contrast, SVG implants in childhood showed a high incidence of closure and degenerative changes. The first pediatric patient who underwent a bypass operation in 1975 with 2 autologous SVGs had patent vein grafts at 1 month1; unfortunately, both SVGs were totally occluded within a couple of years with life-saving collateral developments. Evaluation of the angiographic graft morphology in this study suggests that vein grafts may progressively fail in the future. Among 14 patent SVGs, 10 (71%) showed degenerative changes. Thus, whenever feasible, we believe that total arterial bypass grafting provides a greater likelihood of an event-free future for children with KD, and, in fact, all of our reoperations for KD employed arterial grafts such as the remaining ITA, GEA, and/or the radial artery, anticipating longer patency and better quality of graft status.
In 199426 and 2004,27 we reported the Japanese national survey of operations for KD, in which detailed analysis of cardiac events was difficult because data collection by mailed questionnaire was limited. However, the survey showed that ITA graft patency was less favorable in patients aged <9 to 12 years, most probably because of technical difficulties in small children. In our present study, as shown in Figure 6, the difference in ITA graft patency according to the patient’s age at operation (<10 or ≥10 years) no longer exists (P=0.163); however, SVG graft patency was significantly worse (P=0.004) in children aged <10 years compared with older children. The ITA graft tended to close within a couple of years after surgery because of technical factors or flow competition in the coronary artery with a less stenotic lesion. However, ITA graft failure due to string phenomena may reverse, and the graft may function well when native coronary lesions progress after many years.16,25 Since 1994, with more aggressive correction of anastomotic stenosis by simple balloon dilation, ITA graft patency has significantly improved, up to 94% in our institute.28 We recommend this strategy for children with an ITA bypass graft. Once good flow of the ITA graft is confirmed, closure of the ITA graft seldom occurred during the next 20 years.
The presence of giant coronary aneurysms with no apparent obstructive element creates difficulties when surgery is considered. In our present series, none of such patients underwent either resection of the aneurysm or coronary ligation with distal grafting, partly because coronary aneurysms due to KD never rupture except in the acute phase of the disease and partly because surgery on the aneurysm creates difficulties in reconstructing branch vessels and raises some ethical issues about intentional progression of the grade of coronary stenosis by ligation, which may enhance the risk of MI when an ITA bypass graft fails to function well or is too narrow to carry sufficient blood flow for myocardial recovery after anoxic arrest during operation.8 Recently, partial resection of the noncalcific anterior aneurysmal wall for size reduction of the aneurysm was reported in 9 patients.29 Vortex blood flow was reduced, but long-term clinical benefits remain uncertain, with high occlusion and ineffective rates (56%) after the procedure.29
Although postoperative data on left ventricular function were available in only 72 patients (63%), the left ventricular EF appeared to be an important determinant of outcome in these patients. The postoperative EF averaged 0.55, and that for the 5 deaths was ≤0.5. Poor left ventricular function tended to be associated with sudden death and ventricular arrhythmias, which occurred in 4 of the deaths in this series. Although the EF did not improve after surgery, left ventricular function during exercise improved after successful bypass operations.30 In fact, it was found that 80% of survivors were enjoying unlimited daily activities, including sports.
All of the patients in this series are now adults with increased risk of atherosclerotic changes that may significantly modify future cardiac status.4,5 Prolonged observation is mandatory for these patients because of the progressive decline in the cardiac event-free curve demonstrated in this study. By employing early additional therapeutic procedures such as PCI and redo operations, the patients’ survival and quality of life were favorably influenced during 25 postoperative years. All of the patients who underwent additional procedures are asymptomatic at present.
Study Limitations and Conclusions
This is an observational study with the usual drawback of possible bias in the selection of patients for surgery, but surgical indications have not changed significantly since the first operation for KD coronary sequelae. During the 25-year follow-up, there has been a decreasing trend for small children to undergo operation because of progress in medical care, including early γ-globulin administration,6,13 and improved PCI is now more frequently attempted before surgical treatments.31 Although less frequent than before, patients with severe complicated disease still occur. Eight thousand new KD patients and 20 to 30 operations per year are currently reported in Japan.13 The indications for and long-term effects of PCI and coronary stenting in growing children remain uncertain, and comprehensive reports on PCI for KD are lacking. Despite its limitations, this study shows that patient survival with surgery is excellent and that the increase in cardiac events with time can be successfully managed with reinterventions. In addition, a continued benefit of the ITA graft on clinical outcome up to 25 years was demonstrated. Pediatric coronary bypass surgery with the use of the ITA should be an established treatment for severe coronary disease due to KD.
Kitamura S, Kawashima Y, Fujita T, Mori T, Oyama C. Aortocoronary bypass grafting in a child with coronary artery obstruction due to mucocutaneous lymph node syndrome: report of a case. Circulation. 1976; 53: 1035–1040.
Dhillon R, Clarkson P, Donald AE, Powe AJ, Nash M, Novelli V, Dillon MJ, Deanfield JE. Endothelial dysfunction late after Kawasaki disease. Circulation. 1996; 94: 2103–2106.
Newburger JW, Takahashi M, Gerber MA, Gewitz MH, Tani LY, Burns JC, Shulman ST, Bolger AF, Ferrieri P, Baltimore RS, Wilson WR, Baddour LM, Levison ME, Pallasch TJ, Falace DA, Taubert KA, Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease, Council on Cardiovascular Disease in the Young, American Heart Association. Diagnosis, treatment, and long-term management of Kawasaki disease: a statement for health professionals from the Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease, Council on Cardiovascular Disease in the Young, American Heart Association. Pediatrics. 2004; 114: 1708–1733.
Kato H, Sugimura T, Akagi T, Sato N, Hashino K, Maeno Y, Kazue T, Eto G, Yamakawa R. Long-term consequences of Kawasaki disease: a 10- to 21-year follow-up study of 594 patients. Circulation. 1996; 94: 1379–1385.
Kitamura S. Invited commentary to “Saphenous vein graft growth 13 years after coronary bypass in a child with Kawasaki disease” by El-Khouri HM, Danilowicz DA, Slovis AJ, Colvin SB, Artman M. Ann Thorac Surg. 1998;65:1128–1130.
Tsuda E, Fujita H, Yagihara T, Yamada O, Echigo S, Kitamura S. Competition between native flow and graft flow after coronary artery bypass grafting: impact on indications for coronary artery bypass grafting for localized stenosis with giant aneurysms due to Kawasaki disease. Pediatr Cardiol. 2008; 29: 266–270.
Kitamura S, Seki T, Kawachi K, Morita R, Kawata T, Mizuguchi K, Kobayashi S, Fukutomi M, Nishii T, Kobayashi H, Oyama C. Excellent patency and growth potential of internal mammary artery grafts in pediatric coronary artery bypass surgery: new evidence for a “live” conduit. Circulation. 1988; 78 (suppl I): 129–139.
Abe M, Suzuki N, Katsube Y, Fukazawa R, Kamisago M, Ohkubo T, Ikegami E, Watanabe M, Hajikano M, Watanabe M, Ochi M, Ogawa S. Evaluation of down-sizing operation for coronary giant aneurysm after Kawasaki disease. In: Program and Abstracts of the Ninth International Kawasaki Disease Symposium, Taipei, Taiwan, April 10–12, 2008. Abstract 034.
Kawachi K, Kitamura S, Seki T, Morita R, Kawata T, Hasegawa J, Kameda Y. Hemodynamics and coronary blood flow during exercise after coronary artery bypass grafting with internal mammary arteries in children with Kawasaki disease. Circulation. 1991; 84: 618–624.
Ishii M, Ueno T, Ikeda H, Iemura M, Sugimura T, Furui J, Sugahara Y, Muta H, Akagi T, Nomura Y, Homma T, Yokoi H, Nobuyoshi M, Matsuishi T, Kato H. Sequential follow-up results of catheter intervention for coronary artery lesions after Kawasaki disease: quantitative coronary artery angiography and intravascular ultrasound imaging study. Circulation. 2002; 105: 3004–3010.
Nearly 25 years have passed since the first operation for severe inflammatory coronary artery pathology due to Kawasaki disease. We present a comprehensive long-term follow-up study of 114 such patients. There were no operative or hospital deaths, but 5 late deaths, mostly sudden, occurred during the follow-up. Nonsurvivors were characterized by previous myocardial infarction, frequent ventricular arrhythmias, and a left ventricular ejection fraction of ≤0.5. Survival after coronary bypass operation in childhood and adolescence was excellent, with a 20- to 25-year rate of 95% (95% confidence interval, 88 to 98). However, the cardiac event-free rate declined to 67% at 20 years and 60% (95% confidence interval, 46 to 72) at 25 years. Postoperative interventions (percutaneous coronary intervention or reoperation) were most common as a result of graft anastomotic stenosis (early event), graft closure (midterm), new obstructive lesions, and vein graft atherosclerosis (late events). The increase in cardiac events with time can be successfully managed with reinterventions. Internal thoracic artery grafts had significantly better 20-year patency (87%; 95% confidence interval, 78 to 93) than vein grafts (44%; 95% confidence interval, 26 to 61) (P<0.001) regardless of the patient’s age at operation or target coronary vessels. All 109 survivors are symptom free, and 80% of them are performing unlimited daily activities, although 77% are still receiving 1 to 6 medications, including aspirin in all and warfarin in 21 patients with a patent aneurysm not supported by distal grafts. Pediatric coronary bypass surgery with the use of the internal thoracic artery graft is an established effective treatment for children and adolescents with complicated coronary lesions due to Kawasaki disease.