Cardiac Events and the Maximum Diameter of Coronary Artery Aneurysms in Kawasaki Disease

ARTICLE IN PRESS THE JOURNAL OF PEDIATRICS • www.jpeds.com

ORIGINAL ARTICLES

Cardiac Events and the Maximum Diameter of Coronary Artery Aneurysms in Kawasaki Disease Etsuko Tsuda, MD, PhD, Nobuyuki Tsujii, MD, and Yosuke Hayama, MD Objectives To clarify the occurrence of cardiac events based on the maximal diameter of the maximal coronary artery aneurysm (CAA) in Kawasaki disease (KD). Study design Two hundred fourteen patients (160 male and 54 female) who had had at least 1 CAA in the selective coronary angiogram less than 100 days after the onset of KD were studied. We measured the maximal CAA diameters in the major branches of the initial coronary angiograms. Death, myocardial infarction and coronary artery revascularization were included as cardiac events in this study. We divided the patients into three groups based on the maximal CAA diameter (large ≥8.0 mm; medium ≥6.0 mm and <8.0 mm; small <6.0 mm). Further, we also analyzed the cardiac events based on laterality of maximal CAA (bilateral, unilateral) and body surface area (BSA). Results Cardiac events occurred in 44 patients (21%). For BSA < 0.50 m2, the 30-year cardiac event-free survival in the large and medium groups was 66% (n = 38, 95% CI, 49-80) and 62% (n = 27, 95% CI, 38-81), respectively. For BSA ≥ 0.50 m2, that in large group was 54% (n = 58, 95% CI, 40-67). There were no cardiac events in the medium group for BSA ≥0.50 m2 (n = 36) and the small group (n = 56). In the large analyzed group, the 30year cardiac event-free survival in the bilateral and unilateral groups was 40% (n = 48, 95% CI, 27-55) and 78% (n = 48, 95% CI, 63-89), respectively (P < .0001). Conclusions The group with the highest risk of cardiac events was the patient group with the maximal CAA diameter ≥6.0 mm with BSA < 0.50 m2 and the maximal CAA diameter ≥8.0 mm with BSA ≥ 0.50 m2. At 30 years after the onset of KD, cardiac event-free survival was about 60%. Given the high rate of cardiac events in this patient population, life-long cardiovascular surveillance is advised. (J Pediatr 2017;■■:■■-■■).

K

awasaki disease (KD) is an acute febrile disease that can lead to coronary artery lesions (CALs) in about 15% of affected children treated without intravenous immunoglobulin before the 1990s and about 5% of patients with intravenous immunoglobulin IVIG since 2000.1,2 CALs caused by KD lead to cardiac events such as acute myocardial infarction (MI), and they are a cause of acquired ischemic heart disease in children. Children with KD between the 1970s and 1990s who developed CAL are now young adults. However, there are few long-term reports about the occurrence of cardiac events after the onset of KD.3,4 We investigated the occurrence of cardiac events based on the maximal diameters of coronary artery aneurysms (CAAs) in the initial coronary angiograms (CAGs) obtained immediately after the onset of KD.

Methods There were 579 patients with CAL who had previously undergone CAGs between 1978 and 2011 in our institution. Selective CAG by cardiac catheterization was the only method for the precise diagnosis of CAL until 1990. In the early 1980s, most of the patients with KD had undergone CAG for the diagnosis of CAL in our institution. Between 1990 and 2011, the patients who were diagnosed with CAA by 2-dimensional echocardiography had undergone CAG. The diagnosis of CAA has recently been improved by the application of noninvasive methods such as 2-dimensional echocardiography, computed tomography angiography, and magnetic resonance angiography. Therefore, CAG for the diagnosis of CAA within 2-3 months after the onset of acute KD was not done after 2011 in our hospital. Among the 579 patients, there were 214 patients who had had at least 1 coronary artery aneurysm in the initial CAGs less than 100 days after the onset of KD. For this study, the final diagnosis of KD and CAA was based on diagnostic guidelines by the Japanese Circulation Society.5,6 The age at onset of acute KD ranged from

BSA CAAs CABG CAGs CALs KD MI PCI

Body surface area Coronary artery aneurysms Coronary artery bypass grafting Coronary angiogram Coronary artery lesions Kawasaki disease Myocardial infarction Percutaneous coronary intervention

From the Department of Pediatric Cardiology, National Cerebral and Cardiovascular Center, Osaka, Japan The authors declare no conflicts of interest. 0022-3476/$ - see front matter. © 2017 Elsevier Inc. All rights reserved. http://dx.doi.org10.1016/j.jpeds.2017.05.055

1 FLA 5.4.0 DTD ■ YMPD9227_proof ■ June 26, 2017

THE JOURNAL OF PEDIATRICS • www.jpeds.com

Volume ■■

Figure 2. Coronary artery angiograms. A 4-month-old boy with acute KD (upper). In the initial CAG done 71 days after the onset of KD, the maximum diameter at the segment 1 of the right coronary artery was 6.5 mm (left). The maximum diameter at the segment 6 of the left coronary artery was 7.5 mm (right). The patient was included in the bilateral medium group of BSA < 0.50 m2. A 2-year-old girl with acute KD (lower). In the initial CAG done 60 days after the onset of KD, the maximum diameter of the right coronary artery was 6.8 mm (left). The maximum diameter at the segment of the left anterior descending artery and the segment 11 of the left circumflex were 3.2 mm and 3.6 mm (right). The patient was included in the unilateral medium group of BSA ≥ 0.50 m2.

2 months to 13 years with the median of 23 months. Initial CAG was performed from 20 to 99 days with the median of 59 days, and body surface area (BSA) at the initial CAG ranged from 0.31 to 1.63 with a median of 0.52 m2. The follow-up period ranged from 2.5 months to 37.8 years with the median of 16.8 years. The ethics committee of our institution approved this retrospective study. No extramural funding was used to support this work. The diagnosis of CAA was determined by 2 pediatric cardiologists. Two pediatric cardiologists had measured the maximal CAA diameters in the major branches from the initial CAGs. The major branches included the right coronary artery, the left anterior descending artery, and the left circumflex. If an aneurysm was present at the bifurcation of the left coronary artery, the diameter of the aneurysm at the left coronary artery was also measured. We previously described how to measure coronary arteries and the intra- and interobserver accuracy. 7,8 We divided the patients into 3 groups determined by the maximal CAA diameter in each patient (large ≥8.0 mm; medium ≥6.0 mm and <8.0 mm; and small <6.0 mm) (Figure 1; available at www.jpeds.com). Cardiac events included cardiac death, MI, and coronary artery revascularization (coronary artery bypass grafting [CABG] and percutaneous coronary intervention [PCI]). PCI included percutaneous transluminal coronary balloon angioplasty and percutaneous transluminal coronary rota-

tional ablation. We determined cardiac events for the respective groups from medical records. First, we analyzed the survival, MI-free survival, coronary artery revascularization-free survival, and survival for total cardiac events in the 3 groups based on the maximal CAA diameter. Second, we determined total cardiac event-free survival in the groups based on the maximal CAA diameter and BSA in the initial CAGs. BSA was calculated using the Heycock formula. One group had BSA less than 0.50 m2, and the other group had BSA greater than or equal to 0.50 m2. Third, we determined cardiac event-survival in the groups based on the maximal CAA diameter and the laterality of maximal CAA in the initial CAGs. They were divided into groups depending on the laterality of the maximal CAA, which determined each group (Figures 1 and 2). Finally, we also determined the cardiac eventfree survival in the groups based on the maximal CAA diameter, BSA, and the laterality of maximal CAA in the initial CAGs. Statistical Analyses Statistical analysis was performed using JMP v 10 (SAS Institute Inc, Cary, North Carolina). Measurements are expressed as mean ± SD. The Turkey-Kramer test was used to test for differences of variables among the groups. Survival, MI-free survival, coronary artery revascularization-free survival, and total cardiac event-free survival were analyzed by the KaplanMeier method with 95% CIs. Differences were assessed by the log-rank test. A P value of less than .05 was considered statistically significant.

2

Tsuda, Tsujii, and Hayama FLA 5.4.0 DTD ■ YMPD9227_proof ■ June 26, 2017

■■ 2017

ORIGINAL ARTICLES Results

The number of patients per decade of the initial CAG was 16 patients for 1978-1979; 113 patients for 1980-1989; 55 patients for 1990-1999; and 30 patients for 2000-2011. The number with large, medium, and small CAAs were 96, 62, and 56, respectively. The distribution in BSA < 0.50 m2 was large, 38; medium 27; small 30; and those in BSA ≥ 0.50 m2 was large, 58; medium 35; and small, 26; respectively. Clinical information, medication, and the duration of medication use are shown (Table I). In this study, intravenous immunoglobulin for acute KD had been used since 1985. However, there was no change of aspirin use in the late period. Coumadin had been added in patients with the maximal CAA diameter ≥8.0 mm since the 1990s. Various indices of infection in patients showed no significant change between the groups and no differences in the initial day of treatment of KD and the treatment given. The follow-up periods in the large, medium, and small groups were 20 ± 10 (mean ± SD), 18 ± 10, and 12 ± 7 years, respectively. Cardiac events occurred in 44 patients (21%) (Table II; available at www.jpeds.com). Death, MI, and coronary artery revascularizations occurred in 5, 20, and 36 patients, respectively. Sudden death caused by acute MI occurred in 1 patient; sudden deaths in the late period after previous MI occurred in 4 patients. MI occurred in 20 patients. Twelve patients (60%) among the 20 patients after MI had undergone CABG. CABG was performed in 32 patients and PCI in 6 patients. Thirtyseven patients belonged to the large group and 7 patients were in the medium group for BSA < 0.50 m2 (Table II). Furthermore, 31 patients (70%) had bilateral maximal CAA. There were no cardiac events in the small group. The 30-year survival in the large and medium groups was 94% (n = 96, 95% CI, 84-98) and 96% (n = 62, 77-99), re-

spectively (P = .484). The 30-year MI-free survival in the large and medium groups was 78% (n = 96, 95% CI, 70-87) and 96% (n = 62, 86-99), respectively (P < .0001) (Figure 3, left upper). The 30-year coronary artery revascularization-free survival in the large and medium groups was 63% (n = 96, 95% CI, 5173) and 88% (n = 62, 75-95), respectively (P < .0001) (Figure 3, right upper). The 30-year cardiac event-free survival in large and medium groups were 59% (n = 96, 95% CI, 48-69) and 85% (n = 62, 72-93), respectively (P < .0001) (Figure 3, lower). The larger the maximal CAA diameter, the greater was the occurrence of cardiac events (P < .0001). In the large group, the 30-year cardiac event-free survival for BSA <0.50 m2 and BSA ≥ 0.50 m2 was 66% (n = 38, 95% CI, 49-80) and 54% (n = 58, 40-67), respectively (Figure 4, left upper). The 30-year cardiac event-free survival of the medium group for BSA < 0.50 m 2 was 62% (n = 27, 38-81). For BSA < 0.50 m2, there was no significant difference in events between the large and medium groups. There were no cardiac events in the medium group for BSA ≥ 0.50 m2. In the large group, the 30-year cardiac event-free survival for the bilateral and unilateral groups was 40% (n = 48, 95% CI, 27-55) and 78% (n = 48, 63-89), respectively (P < .0001) (Figure 4, right upper). In the large group, the 30-year cardiac event-free survival for the bilateral group was significantly lower than that in the unilateral group. In the medium group, the 30-year cardiac event-free rates for bilateral and unilateral groups were 76% (n = 18, 95% CI, 45-93) and 88% (n = 44, 72-96), respectively. There was no significant difference between bilateral and unilateral in the medium group. For BSA <0.50 m2, the 25-year cardiac event-free survival in the bilateral large, unilateral large, bilateral medium, and unilateral medium groups was 54% (n = 22, 95% CI 33-73), 85% (n = 16, 55-96), 55% (n = 8, 20-86), and 66% (n = 19, 36-87), respectively (Figure 4, lower). For BSA ≥ 0.50 m2, the

Table I. Patient characteristics in the respective groups Groups Patient (n) Male KD typical Age at KD (mo) Fever (d) WBC (m3) CRP (mg/dL) Acute treatment None Late treatment* Unknown Patient (n) Aspirin IVIG Steroid Initial CAG (d) Follow-up (y) Treatment in the late period Antiplatelet Coumadin Interval of medication (y)

Large

Medium

Small

96 71 (74%) 86 (90%) 41 ± 3 15 ± 8 (n = 89) 18 630 ± 6047 (n = 47) 17 ± 9 (n = 47)

62 46 (74%) 60 (97%) 31 ± 4 14 ± 5 (n = 58) 18 141 ± 4599 (n = 30) 16 ± 7 (n = 33)

56 43 (77%) 55 (98%) 22 ± 4 13 ± 4 (n = 55) 20 076 ± 5431 (n = 36) 16 ± 6 (n = 37)

0 7 5 84 72 (86%) 49 (58%) 23 (27%) 58 ± 19 20 ± 10

2 2 3 55 48 (87%) 24 (44%) 17 (31%) 58 ± 19 18 ± 10

2 2 3 49 44 (90%) 15 (30%) 8 (16%) 58 ± 20 12 ± 7

96 (100%) 31 (32%) 18 ± 10

62 (100%) 7 (11%) 13 ± 11

55 (98%) 9 (16%) 2±3

CRP, C-reactive protein; IVIG, intravenous immunoglobulin; WBC, white blood cell. *Treatment was given after the 10th day when CAA were diagnosed.

Cardiac Events and the Maximum Diameter of Coronary Artery Aneurysms in Kawasaki Disease FLA 5.4.0 DTD ■ YMPD9227_proof ■ June 26, 2017

3

THE JOURNAL OF PEDIATRICS • www.jpeds.com

Volume ■■

Figure 3. MI-free rate, coronary artery revascularization free rate, and free rate for total cardiac events. MI-free rate (left upper); coronary artery revascularization free rate (right upper); and total cardiac events (lower).

Figure 4. Free rate for total cardiac events in the respective groups. Free rate for total cardiac events based on BSA and the maximum CAA diameter (left upper). Free rate for total cardiac events based on the maximal CAA diameter and the laterality of maximal CAA (right upper). Free rate for total cardiac events based on BSA, the maximal CAA diameter, and the laterality of maximal CAA (lower). 4

Tsuda, Tsujii, and Hayama FLA 5.4.0 DTD ■ YMPD9227_proof ■ June 26, 2017

■■ 2017

ORIGINAL ARTICLES

25-year free survival in the bilateral large and unilateral large groups was 30% (n = 26, 15-50) and 75% (n = 32, 55-88), respectively. The 25-year cardiac event-free survival rates in the respective groups are shown in Table III (available at www.jpeds.com).

Discussion The most important cardiac event influencing the outcome in patients with CAL after onset of KD is MI. MI is likely to occur within 1 year after onset of KD, and myocardial involvement caused by MI can lead to sudden death in the late period.9 Antithrombotic therapy with coumadin and antiplatelet therapy is one means of reducing the occurrence of acute MI.4,10,11 This study was retrospective, and the number of patients receiving antithrombotic therapy was small. Therefore, the frequency of MI in the medium and large groups might be lower in the future, based on the impact of antithrombotic therapy. Cardiac eventfree survival in the large group was lower in the first 10 years after the onset of KD. In addition to thrombotic complications, patients with KD may also develop coronary artery stenosis because of myofibroblastic proliferation, which can progress gradually over many years.7 It is considered that the reduction in cardiac event-free survival after the first 15 years of onset of acute KD might depend on progression of localized stenosis in the medium group with BSA <0.50 m2. In the classification by the laterality of the maximal CAA, the worst group was the large bilateral group in patients with BSA ≥ 0.50 m2. The larger the diameter of the maximal CAA, the more extensive are the coexisting CAAs.12 The larger the CAA, the greater the occurrence of cardiac events. We have reported that the outcome in patients with bilateral giant aneurysms was worse than that in patients with unilateral giant aneurysms.4 The combination of antithrombotic therapy and careful patient follow-up with coronary artery revascularization as indicated is the most likely strategy to achieve optimal patient outcomes at present.12,13 In this study, there were no cardiac events in the small group, although the interval of follow-up was significantly shorter compared with the other groups. Longer follow-up periods for patients with KD with small aneurysms are required to determine outcomes.14 This study included patients who underwent CAG from 1978 to the present. Because the treatment for acute KD and CAL has improved over time, this study has some limitations. We could not base our analysis on z scores in CAG. Larger maximal CAA diameters, especially when combined with bilateral lesions, result in high frequency of cardiac events. At 30 years after the onset of KD, cardiac event-free survival was about 60% in the patients with maximal CAA diameter ≥6.0 mm for BSA < 0.50 m2 and maximal CAA diameter

≥8.0 mm for BSA ≥ 0.50 m2. Given the high rate of cardiac events in this patient population, life-long cardiovascular surveillance is advised. ■ We thank Professor Peter Olley and Dr Setsuko Olley for their English language consultation. We also thank Koko Asakura for statistic consultation. Submitted for publication Feb 1, 2017; last revision received May 2, 2017; accepted May 19, 2017 Reprint requests: Etsuko Tsuda, MD, PhD, Department of Pediatric Cardiology, National Cerebral and Cardiovascular Center, 5-7-1 Fujishirodai, Suita, Osaka, Japan. E-mail: [email protected]

References 1. Kato H, Koike S, Yamamoto M, Ito Y, Yano E. Coronary aneurysms in infants and young children with acute febrile mucocutaneous lymph node syndrome. J Pediatr 1975;86:892-8. 2. Nakamura Y, Yashiro M, Uehara R, Watanabe M, Yanagawa H. Epidemiologic features of Kawasaki disease in Japan: results of the 2009-2010 nationwide survey. J Epidemiol 2012;22:216-21. 3. Suda K, Iemura M, Nishiono H, Teramachi Y, Koteda Y, Kishimoto S, et al. Long-term prognosis of patients with Kawasaki disease complicated by giant coronary aneurysms: a single-institution experience. Circulation 2011;123:1836-42. 4. Tsuda E, Hamaoka K, Suzuki H, Sakazaki H, Murakami Y, Nakagawa M, et al. A survey of the 3-decade outcome for patients with giant aneurysms caused by Kawasaki disease. Am Heart J 2014;167:249-58. 5. Research Committee on Kawasaki Disease. Report of subcommittee on standardization of diagnostic criteria and reporting of coronary artery lesion in Kawasaki disease. Tokyo, Japan: Ministry of Health and Welfare 1984. 6. JCS Joint Working Group. Guidelines for diagnosis and management of cardiovascular sequelae in Kawasaki disease (JCS 2013). Digest version. Circ J 2014;78:2521-62. 7. Tsuda E, Kamiya T, Ono Y, Kimura K, Kurosaki K, Echigo S. Incidence of stenotic lesions predicted by acute phase changes in coronary arterial diameter during Kawasaki disease. Pediatr Cardiol 2005;26:73-9. 8. Tsujii N, Tsuda E, Kanzaki S, Kurosaki K. Measurements of coronary artery aneurysms due to Kawasaki disease by dual-source computed tomography (DSCT). Pediatr Cardiol 2016;37:442-7. 9. Tsuda E, Hirata T, Matsuo O, Abe T, Sugiyama H, Yamada O. The 30year outcome for patients after myocardial infarction due to coronary artery lesions caused by Kawasaki disease. Pediatr Cardiol 2011;32:176-82. 10. Suda K, Kudo Y, Higaki T, Nomura Y, Miura M, Matsumura M, et al. Muticenter and retrospective case study of warfarin and aspirin combination therapy in patients with giant coronary aneurysms caused by Kawasaki disease. Circ J 2009;73:1319-23. 11. Su D, Wang K, Qin S, Pang Y. Safety and efficacy of warfarin plus aspirin combination therapy for giant coronary artery aneurysm secondary to Kawasaki disease. A meta-analysis. Cardiology 2014;129:55-64. 12. Tsuda E, Tsujii N, Kimura K, Suzuki A. Distribution of Kawasaki disease coronary artery aneurysms and the relationship to coronary artery diameter. Pediatr Cardiol 2017;38:932-40. 13. Tsuda E. Coronary artery bypass grafting for coronary artery stenosis caused by Kawasaki disease. Expert Rev Cardiovasc Ther 2009;7:533-9. 14. Tsujii N, Tsuda E, Kanzaki S, Ishiduka J, Nakashima K, Kurosaki K. Late wall thickening and calcification after Kawasaki disease. J Pediatr 2017;181:167-71.

Cardiac Events and the Maximum Diameter of Coronary Artery Aneurysms in Kawasaki Disease FLA 5.4.0 DTD ■ YMPD9227_proof ■ June 26, 2017

5

THE JOURNAL OF PEDIATRICS • www.jpeds.com

Volume ■■

Figure 1. Classification based on the maximal diameter and the laterality of the maximum CAA. Classification based on the maximal CAA diameter (upper). Large group (left side); medium group (middle); and small group (right side). Classification based on the maximal diameter and the laterality of the maximum CAA (lower). We classified between either bilateral or and unilateral groups, adopting the laterality of the maximum CAA to decide the respective group. The left side case is included in the bilateral large group, and the middle case in the unilateral medium group. The right side is in the unilateral small group.

Table II. Cardiac events in the respective group BSA < 0.50 m2

0.50 m2 ≤ BSA

Large Bilateral

Large Unilateral

Medium Bilateral

Medium Unilateral

Large Bilateral

Large Unilateral

22

16

8 1

19

26 3 1 8

32 1 2

2

2

6

3 1

2 (13%)

3 (38%)

1 2 1 4 (21%)

18 (69%)

7 (22%)

Patient (n) MI, death MI MI, CABG MI, CABG, PCI CABG PCI PCI, CABG

3 1 5 1 10 (45%)

Total 5 3 11 1 19 4 1 44 (21%)

Table III. The 25-year cardiac event-free survival in the respective groups BSA <0.50 m2 Groups Large (%) Medium (%) Small (%)

0.50 m2 ≤ BSA

Bilateral

Unilateral

Bilateral

Unilateral

54 55 100

85 66 100

30 100 100

75 100 100

5.e1

Tsuda, Tsujii, and Hayama FLA 5.4.0 DTD ■ YMPD9227_proof ■ June 26, 2017