Article Text

Download PDFPDF

Coronary artery aneurysms are more severe in infants than in older children with Kawasaki disease
  1. Scott A Cameron1,
  2. Michael Carr1,
  3. Elfriede Pahl1,
  4. Nicole DeMarais2,
  5. Stanford T Shulman3,
  6. Anne H Rowley3
  1. 1 Division of Cardiology, Ann & Robert H Lurie Children’s Hospital of Chicago, Chicago, Illinois, USA
  2. 2 Chicago Medical School, North Chicago, Illinois, USA
  3. 3 Division of Infectious Diseases, Ann & Robert H Lurie Children’s Hospital of Chicago, Chicago, Illinois, USA
  1. Correspondence to Dr Anne H Rowley, Division of Infectious Diseases, Ann & Robert H Lurie Children’s Hospital of Chicago, Chicago, IL 60611, USA; arowley{at}


Objective We aimed to compare the severity of coronary artery abnormalities in Kawasaki disease between infants and older children.

Methods We retrospectively reviewed and compared coronary artery dilation and aneurysm severity in infants <1 year of age with Kawasaki disease at our centre over a 10-year period with that observed in children ≥1 year of age in the Pediatric Heart Network Trial of Pulse Steroid Therapy in Kawasaki Disease. Coronary artery abnormalities were defined by z-scores according to American Heart Association guidelines.

Results Of the 93 infants identified during the study period, 80 were treated with intravenous gamma globulin within the first 10 days of illness and were included for comparison to 170 children ≥1 year of age treated in the same time frame from the Pediatric Heart Network public database. The mean maximum z-score was significantly higher in infants compared with older children (3.37 vs 2.07, p<0.001). A higher incidence of medium and giant aneurysms was observed in infants compared with children ≥1 year of age (11% vs 3% for medium aneurysms, p=0.015; 8% vs <1% for giant aneurysms, p=0.005).

Conclusions Infants with Kawasaki disease have more severe coronary artery dilation compared with older children, and a higher prevalence of medium and giant aneurysms. Because adverse outcomes are closely linked to the maximal coronary artery diameter in Kawasaki disease, patients diagnosed as infants require very close long-term monitoring for cardiac complications.

  • cardiology
  • infectious diseases
  • vascular disease

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known on this topic?

  • Kawasaki disease is the most common cause of acquired heart disease in children in the developed world.

  • Infants with Kawasaki disease have a higher risk for developing coronary artery aneurysms than other age groups.

  • Long-term outcomes in Kawasaki disease are related to maximal coronary artery diameter.

What this study adds?

  • Infants with Kawasaki disease have a higher prevalence of coronary artery aneurysms than older children, and have more severe coronary artery aneurysms.

  • Infants with significant coronary artery involvement require close long-term monitoring for late complications of Kawasaki disease.

  • Trials of medications aimed at decreasing coronary artery aneurysm severity in infants with Kawasaki disease appear indicated.


Kawasaki disease (KD) is the leading cause of acquired paediatric heart disease in developed countries and its incidence is highest in infants between 9 and 11 months of age.1 2 Although an increased frequency of coronary artery aneurysms (CAA) in young infants with KD has been described,3–5 the severity of these abnormalities has not been closely examined. Our clinical experience has suggested that CAA in infants with KD is often quite severe, with a subgroup requiring anticoagulation therapy in addition to antiplatelet therapy. Therefore, we performed this retrospective review focusing on the severity of coronary artery abnormalities in infants with KD.


Medical records for infants diagnosed with KD at the Ann & Robert H Lurie Children’s Hospital of Chicago (formerly Children’s Memorial Hospital) between 1 June 2005 and 10 November 2015 were reviewed. The Pediatric Heart Network (PHN) Trial of Pulse Steroid Therapy in Kawasaki Disease public database was analysed for comparison.

Infants from our centre were included if they had a diagnosis of KD and were less than 1 year of age. The diagnosis of KD was confirmed in all patients by experienced paediatric infectious disease physicians based on American Heart Association (AHA) criteria.2 Patients were excluded if they were diagnosed at an outside institution, transferred care or were lost to follow-up in the acute period, or had an alternative diagnosis established. Inclusion and exclusion criteria for the PHN Trial of Pulse Steroid in Kawasaki Disease have been published.6

All patients diagnosed with KD at our institution have echocardiograms performed at the time of diagnosis and at both 1–2 weeks and 4–6 weeks after diagnosis. More frequent studies are performed when there are significant coronary artery abnormalities or for clinical concerns. Follow-up echocardiograms are obtained per AHA recommendations based on risk category. Body surface area (BSA) is calculated at the time of each echocardiogram in EchoIMS by the Haycock method. BSA-based z-scores for the proximal left anterior descending coronary artery (pLAD) and right main coronary artery (RMCA) were reported for each study based on the methods described by McCrindle et al.7 Coronary artery abnormalities and aneurysms were categorised according to the z-scores published in the AHA guidelines.2

All patients diagnosed with KD at our centre are treated with a single dose of intravenous immune globulin (IVIG) at 2 g/kg and are started on aspirin at the time of diagnosis. A second dose of IVIG is administered to IVIG non-responders as per AHA guidelines.2 Steroids (intravenous methylprednisolone or oral prednisolone), and less commonly infliximab, are administered to IVIG non-responders per physician discretion. Our centre cares for about 60 new cases of KD per year.

The following data were abstracted from medical records:

  • Demographics: age, gender, ethnicity.

  • The presence or absence of the five classic clinical features (extremity changes, rash, conjunctival injection, oral changes and cervical lymphadenopathy).2

  • Baseline laboratory data: complete blood count, albumin, alanine aminotransferase (ALT), erythrocyte sedimentation rate (ESR), C-reactive protein (CRP).

Coronary artery z-scores for the RMCA and pLAD were obtained by retrospective review of echocardiography reports. Reports were also reviewed for the presence or absence of pericardial effusion (more than trivial), decreased left ventricular systolic function (ejection fraction <55%, or shortening fraction z-score <−2), left ventricular dilation (left ventricular end-diastolic dimension z-score >2), mitral regurgitation (more than trivial) and aortic insufficiency.

Statistical analysis

Descriptive data were reported as means with SDs, medians with IQRs, or frequencies as appropriate. An unpaired t-test was used to determine whether the mean Zmax for infants from our centre with KD was significantly different from the highest mean Zmax reported for children ≥1 year of age in the PHN trial.6 For this comparison, we excluded infants from our centre who were treated after day 10 of illness, as all patients in the PHN trial were treated within 10 days. The χ2 test was used for univariate comparison of nominal variables. Spearman correlation coefficient was used to evaluate the relationship between age and Zmax. P values <0.05 were considered statistically significant. REDCap8 was used for data collection and SPSS version 25 for statistical analysis.


A total of 104 infants with KD were diagnosed at our centre during the study period. Three patients were excluded after alternative diagnoses were determined: one each with systemic onset juvenile rheumatoid arthritis, tumour necrosis factor receptor-associated periodic syndrome and dilation of a coronary ostium secondary to abnormal flow across a stenotic bicuspid aortic valve. Eight patients were excluded for insufficient data: two initially diagnosed elsewhere, one transfer to an outside institution during the acute phase and five who were lost to follow-up in the acute phase.

Ninety-three infants were included for analysis. The median patient age at diagnosis was 7 months (range 1–11 months). The median follow-up duration was 13 months (range 1 month to 9.5 years). There were 170 children ≥1 year of age with a median age of 3.5 years (range 1–12.3 years) in the PHN trial included for comparison. Demographic, clinical, laboratory and clinical outcome data are shown in table 1. There was a higher percentage of Hispanic patients in the infant group compared with older children (36% vs 16%, p<0.001). The diagnosis of KD was incomplete in a higher percentage of infants compared with children ≥1 year of age (38% vs 3%, p<0.001). Cervical lymphadenopathy was the least commonly observed classic feature in infants and older children, but was less commonly observed in infants than in older children (15% vs 56%, p<0.001).

Table 1

Demographic and clinical characteristics

Severity of coronary artery abnormalities in infants with KD diagnosed and treated within the first 10 days of illness when compared with children >1 year of age in the PHN study (all PHN enrolled patients were treated within the first 10 days of illness)

Overall, 38/80 (48%) of infants diagnosed and treated in the first 10 days of illness had a Zmax>2.5 for the pLAD and/or proximal right coronary artery. The mean maximum coronary artery z-score in the first 5 weeks of illness in infants in our cohort who were diagnosed in the first 10 days of illness was 3.37, compared with 2.07 for the children ≥1 year of age in the PHN study (p<0.001; table 2).9 A higher incidence of medium and giant aneurysms was observed in infants compared with children ≥1 year of age (11% vs 3% for medium aneurysms, p=0.015; 8% vs <1% for giant aneurysms, p=0.005; table 3). CAAs were present in 26 of 38 (68%) of infants <6 months of age and 22/55 (40%) of infants ≥6 months of age (p=0.007).

Table 2

Maximum coronary artery z-score

Table 3

Comparison of aneurysm severity

Further analyses of the infant cohort

Severity of coronary artery abnormalities in infants with KD diagnosed and treated after day 10 of illness

Delayed diagnosis was associated with an increased risk of CAA formation in the RMCA and pLAD (z-score ≥2.5) and a higher mean Zmax in the RMCA or pLAD in the first 5 weeks of illness (8.14 vs 3.37, p=0.02; figure 1, online supplementary table 1), as previously reported in many studies. Importantly, giant CAA (z-score ≥10) was observed in 5/13 (38%) infants diagnosed and treated after day 10 of illness (online supplementary table 1). Differences in other echocardiographic features in infants treated before and after the 10th day of illness were not statistically significant (table 4).

Figure 1

Severity of coronary artery dilation in infants with Kawasaki disease. Although treatment of Kawasaki disease after day 10 of fever onset was associated with a higher mean Zmax in the right main coronary artery (RMCA) or proximal left anterior descending coronary artery (pLAD) in the first 5 weeks of illness (8.14 vs 3.37, p=0.02), many infants treated within the first 10 days still had significant coronary artery disease, including large/giant aneurysms (z-score >10). IVIG, intravenous immune globulin.

Table 4

Echocardiographic findings in infants with KD

Severity of coronary artery abnormalities in infants by age

There was a statistically significant relationship between age and Zmax, with the younger patients having a higher Zmax (r=−0.177, p=0.029). The mean Zmax at any point after diagnosis for patients <6 months of age was 6.00 compared with a mean Zmax of 2.93 for patients aged 6 months to 1 year (p=0.003, online supplementary figure 1).

Incomplete and delayed diagnosis of KD in infants by age

Overall, 41% of infants had incomplete KD (including infants treated beyond day 10 of illness). The percentage of patients with a delayed diagnosis who were <6 months of age was similar to that for those 6 months to 1 year of age (13% vs 16%, NS). Overall, 60% of infants presenting with incomplete KD developed CAAs, while 45% of infants with classic KD developed CAAs (NS).

Anticoagulation and dual antiplatelet therapy in infants with KD

Eight of 93 (9%) infants received either warfarin or heparin in addition to antiplatelet therapy, and eight received clopidogrel in addition to aspirin, according to AHA guidelines.2

CAAs by gender

A higher percentage of boys (32/52, 61.5%) developed CAA compared with girls (16/41, 39%; p=0.031). The mean Zmax at any point after diagnosis in boys was 4.86 compared with 3.35 for girls (p=0.095).

IVIG resistance

The overall IVIG resistance rate as defined by the administration of more than 1 dose of IVIG was 16%, with 15/93 patients receiving more than 1 dose of IVIG and similar rates of resistance in infants <6 months of age and 6 months to 1 year of age.

Increase in Zmax after 5 weeks

There were four infants with giant aneurysms who had progressive dilation of the coronary arteries after week 5 of illness. Zmax was observed in these four patients between 6 weeks and 14 months after diagnosis. Serum CRP levels normalised between 14 and 36 days after fever onset in this subset of patients. Two of the four patients with progressive dilation received IVIG within 10 days of fever onset (one of these demonstrated IVIG resistance) and treatment was delayed past day 10 of illness in the other two patients (neither of these had IVIG resistance). All four patients had CAAs present on their initial echocardiograms.

Infants with coronary artery dilation only

The AHA classifies ‘dilation only’ as any z-score of 2.0 to <2.5, or, if a z-score <2.0, a decrease in z-score ≥1.0 during follow-up.2 Eleven of 93 (12%) had a Zmax between 2.0 and <2.5, and 13/93 (14%) had a Zmax<2.0 with a decrease in z-score ≥1.0.

Laboratory data

Compared with children ≥1 year of age, infants had higher white cell and platelet counts; and lower haemoglobin, haematocrit and albumin levels. Statistically significant differences in ALT, CRP and ESR levels were not identified.


Our study demonstrates that infants with KD have a higher severity of coronary artery dilation compared with children ≥1 year of age. It has been clearly demonstrated that risk of adverse clinical outcomes from CAAs in KD is directly related to the maximal CAA diameter during the acute phase.10 11 In our study, 13% (12/93) of infants developed medium CAAs (z-score ≥5 to <10) and 12% (11/93) developed giant CAAs (z-score >10), indicating that these patients are at significant risk for long-term adverse outcomes. Although regression of luminal diameter occurred over time in this group of patients, this should not be perceived as being equivalent to normalisation of the coronary arteries. The appearance of aneurysm regression or normalisation of lumen diameter by echocardiography can be the result of layered thrombus and/or proliferation of modified smooth muscle cells that can result in adverse outcomes.12 A prior study showed a notable skewing of mortality from KD towards those who were infants at the time of the acute illness.13 In light of the high severity of CAAs in infants with KD, this age group may benefit from adjunctive primary therapies for KD. Prior studies have highlighted the increased risk of coronary artery abnormalities in infants <6 months of age.3 In our series, 68% of infants <6 months of age developed coronary aneurysms. However, in our study, infants 6 months to 12 months also had a high prevalence of coronary aneurysms (40%).

Coronary artery diameter continued to increase in size after week 5 of illness in 4% of patients, and these patients may be at even higher risk of worse long-term outcome. Progression of coronary artery diameter at least 2 months after onset has been associated with an increased probability of persistence of medium-sized aneurysms, and an increased risk of myocardial ischaemia for patients with giant aneurysms.14

The severity of coronary artery involvement in infants is not explained by delayed treatment alone. Young infants have a higher rate of delayed treatment than other age groups,15 16 but younger age alone is a risk factor for the development of CAA independent of delayed treatment.3 17 In our series, despite a high frequency of incomplete KD (41%), treatment delay was infrequent (14%). Regardless, improved early recognition of KD is needed,15 which is underscored by the strikingly high prevalence (38%) of giant aneurysms in infants who were treated after day 10 of illness.

Systemic anticoagulation is indicated for giant coronary aneurysms to mitigate the risk of coronary thrombosis.2 In our study, 9% of infants were treated with either enoxaparin or warfarin. Systemic anticoagulation is difficult to manage in infants and carries a risk of bleeding, further complicating management and prognosis.18 Serial follow-up into early and middle adulthood will be required to determine the absolute risk of adverse outcomes.19 In the short term, one of the patients in our cohort with a giant aneurysm developed non-occlusive coronary artery thrombosis diagnosed by echocardiography and confirmed by CT angiography. This patient was treated with dual-antiplatelet therapy in addition to systemic anticoagulation and did not develop clinical evidence of ischaemia. Although our study did not address costs of medical care for infants with KD, the high cost of acute management and long-term follow-up for this population has been reported.20

Our study has several limitations. It is limited by being a single-centre experience in an area with a low Asian population. However, the peak age of KD diagnosis in Asian children is also in infancy,1 and Asian children may be at higher risk of developing CAA than non-Asian children.21 We compared our single-centre z-scores in infants with those determined in a prospective multicentre study in the USA.6 However, our centre has decades of experience with performing high-quality echocardiograms in children with KD and standard techniques of measurements and z-score calculations were used.22 23 In infants with KD, small differences in coronary artery measurements can lead to large differences in z-scores. As there is no specific test for KD, the true incidence of KD is unknown, which may affect the statistics reported in this article. Despite these limitations, we believe our study provides important information on the severity of coronary artery abnormalities in infants with KD.

Our findings have implications for the counselling of families of infants with KD. While a prior study reported low rates of adverse cardiovascular events through age 21 in patients with KD, nearly 80% of the patients in that study never had detectable coronary abnormalities and these patients would be expected to have excellent outcomes.24 When advising parents of children with KD with coronary artery abnormalities regarding outcomes, education should be tailored to the severity of the abnormalities in that child, as classified in the recent AHA guideline on KD.2 Physicians and parents should be aware of the high-risk nature of the diagnosis of KD as an infant and the need for prompt diagnosis and treatment, as well as consistent and long-term cardiology follow-up in those who develop coronary artery abnormalities.


David Russell provided technical support. REDCap is supported at the Feinberg School of Medicine by the Northwestern University Clinical and Translational Science (NUCATS) Institute. Research reported in this publication was supported, in part, by the National Institute of Health’s National Center for Advancing Translation Sciences (grant number UL1TR001422). The NIH/NHLBI Pediatric Heart Network Kawasaki Disease data set was used in preparation of this work. Data were downloaded from on 1 September 2018.



  • Contributors AHR conceptualised the study, drafted the initial manuscript, and reviewed and revised the manuscript. MC, EP and STS aided in conceptualisation, and reviewed and revised the manuscript. SAC aided in conceptualisation, designed the study, drafted the initial manuscript, designed the data collection instruments, collected data, carried out the statistical analyses, and reviewed and revised the manuscript. ND aided in data collection and manuscript review and editing. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

  • Funding This study was supported by NIAMS R21AR068041, NCATS UL1TR001422, Max Goldenberg Foundation, and the Center for Kawasaki Disease at the Ann & Robert H Lurie Children’s Hospital of Chicago.

  • Disclaimer The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

  • Competing interests None declared.

  • Patient consent Not required.

  • Ethics approval Institutional Review Board of Ann & Robert H Luire Children’s Hospital of Chicago.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement The Pediatric Heart Network data used in this study are from a public database that is freely available on the internet at:

  • Collaborators David Russell.