The Role of Early Echocardiography in Low‐Risk Patients With Kawasaki Disease

ABSTRACT

Aim

Anecdotally, in some centres in Australia and New Zealand, there is difficulty in obtaining early echocardiograms in patients with Kawasaki disease (KD) considered at low risk of coronary artery (CA) involvement, justified by the argument that the risk of coronary artery aneurysm (CAA) is low and that results will not change management. We sought to determine how often in practice the results of early echocardiography alter management in patients with KD, particularly in those at low risk of medium–giant aneurysms.

Methods

A retrospective chart review of patients with KD at a large paediatric tertiary referral centre was performed. Data collected included demographic features, echocardiography results, and treatments. Analysis of data was stratified by risk of coronary involvement using criteria prospectively defined by the study authors.

Results

Of the 147 included patients, 53 (36.1%) were female and the mean age was 3.4 years (range 0.3–11.4 years). 13 (8.8%) had aneurysms on baseline echocardiogram; 4/86 ‘low‐risk’ (4.7%) and 9/61 (14.7%) ‘high‐risk’ patients (odds ratio 3.55; 95% confidence interval 1.04–12.11). Baseline echocardiogram findings changed management in three ‘low‐risk’ patients who had primary treatment intensification. Seven ‘high‐risk’ patients had changes in management including primary treatment intensification, additional anti‐platelet treatment, and anticoagulation.

Conclusions

These data support recommendations for baseline echocardiography in all patients with KD regardless of perceived risk of CAA. Further research into risk stratification in KD is required.

Summary.

• What is already known on this topic

  • ○
    Kawasaki disease is the most common cause of acquired childhood coronary artery disease in developed regions.
  • ○
    Echocardiography is the primary imaging modality for assessing coronary arteries in Kawasaki disease but is resource intensive.
  • ○
    Tailoring treatment and assessments to individual patients with Kawasaki disease is key to providing high quality and efficient care.
• What this paper adds

  • ○
    Contemporary data on a large multi‐ethnic Australian cohort is presented.
  • ○
    The data presented supports the current recommendations to perform baseline echocardiography in all patients with Kawasaki disease.
  • ○
    This study highlights the importance of further research into risk stratification in Kawasaki disease.

1. Introduction

Kawasaki disease (KD) is an acute childhood vasculitis of an unclear cause that predominantly affects medium‐sized arteries. Cardiovascular sequelae, including coronary artery aneurysms (CAAs), are an important long‐term complication in some patients. Contemporary treatment approaches, including early treatment with intravenous immunoglobulin (IVIG), have significantly reduced the risk of CAA development in children with KD [1]. Minor coronary changes are common early in the acute phase of KD despite treatment [2]. Minor coronary dilation and small aneurysms generally have a favourable prognosis with a high likelihood of eventual regression, in contrast to large and giant aneurysms, which carry a high risk of persistence [3].

Stratifying the risk of CAA development in patients with KD is challenging. Algorithms that reliably predict relevant outcomes such as IVIG resistance and CAA development are lacking, particularly in multi‐ethnic populations [4, 5]. Factors including IVIG resistance, delayed treatment with IVIG, and young age have been associated with increased risk of CAA development but do not accurately predict outcomes [4, 5, 6]. Given the difficulty in anticipating coronary outcomes, current recommendations suggest routine CA assessment by echocardiography for all children with KD [1]. It is recommended that an initial echocardiogram be performed as soon as a diagnosis of KD is suspected to establish a baseline for longitudinal follow‐up of coronary morphology [1].

Initial echocardiography may also affect immediate management in some patients. Current guidelines suggest consideration of additional anti‐platelet and/or anticoagulation therapies for patients with aneurysms that are graded as medium or higher [1, 7, 8]. Patients with significant and evolving CA abnormalities should also have more frequent echocardiography during the acute phase until luminal dimensions stop progressing [1, 8]. Additionally, substantial data indicate that the presence of CA changes in the acute phase of KD should prompt consideration of intensification of primary therapy with agents such as corticosteroids [8, 9]. Indeed, primary intensification of anti‐inflammatory therapy is now recommended in patients with high‐risk findings on baseline echocardiography in the most recent scientific statement on the diagnosis and management of KD from the American Heart Association (AHA) [8].

Echocardiography is safe but resource intensive and requires skilled staff to perform and interpret the scans. Anecdotally, across some centres in Australia and New Zealand, baseline echocardiography is reserved only for patients who are perceived to be at high risk of CAA development. In order to reduce resource utilisation, baseline echocardiography may be omitted in patients with low‐risk characteristics based on historical data from multiple KD patient cohorts, including those who respond well to an initial dose of IVIG.

Prior studies have demonstrated the safety and cost effectiveness of reducing the number of echocardiograms performed in selected low‐risk patients in the later stages of KD follow‐up [10]. To our knowledge, no studies have examined the possibility of selective early echocardiography in low‐risk KD patients in the acute phase. The Royal Children's Hospital (RCH) in Melbourne is the largest tertiary referral children's hospital in Australia and New Zealand. On average, over 40 children per year are treated for KD at RCH, and early echocardiography is routinely performed in all patients. We sought to determine how often at our centre the results of baseline echocardiography altered patient management in children diagnosed with KD, particularly those at low risk of medium‐giant aneurysms. In doing so, we aimed to determine the appropriateness or otherwise of selective rather than routine early echocardiography in low‐risk patients with KD.

2. Materials and Methods

We performed a retrospective chart review including all patients with a discharge diagnosis of KD (including incomplete or atypical Kawasaki disease) from the RCH, Melbourne between May 2016 and May 2020. Patients were excluded if they were not treated with IVIG or received IVIG outside of the RCH.

Patients were stratified by risk of poor coronary outcome based on criteria prospectively defined by the study authors. Given the challenges with defining risk in a retrospective cohort in which not all laboratory parameters relevant to established risk criteria would be available, we assessed simple clinical features that have been shown to correlate with an increased risk of IVIG resistance and/or CAA development in previous studies [4, 5]. For this study, patients were classified as ‘high‐risk’ if they met any of the three following criteria:

• –Less than 12 months of age at time of KD diagnosis
• –Treated with IVIG after 10 days of the onset of KD symptoms
• –Failed to respond to a single dose of IVIG

Patients not meeting all these criteria were classified as ‘low‐risk’.

Baseline demographics, clinical details, laboratory results, echocardiography results, and treatment details were collected. Echocardiogram results were interpreted using the AHA definitions [1] as follows:

• –Small aneurysms = coronary z‐score ≥ 2.5 to < 5
• –Medium aneurysms = coronary z‐score ≥ 5 to < 10 and absolute dimension < 8 mm
• –Large/giant aneurysms = coronary z‐score ≥ 10 or absolute dimension ≥ 8 mm

Odds ratios with 95% confidence intervals were used to compare the ‘low‐risk’ and ‘high‐risk’ groups. Treatment was deemed to have been changed by the results of an early scan if therapies other than IVIg and aspirin were used in the acute phase of management.

3. Results

One hundred and seventy‐eight patients were initially identified. Thirty one were excluded (7 had an alternate final diagnosis, 11 did not receive IVIG, 13 received IVIG at another hospital). Of the 147 included patients, 53 (36.1%) were female and the mean age was 3.4 years (range 0.3–11.4 years) (Table 1). All were treated with low dose aspirin. Ethnicity was documented for 42 (28.6%) patients and was assigned based on family name where it was not documented. Approximately half of the included patients were of Asian ethnicity and a third were of European ethnicity. All but one patient had a baseline echocardiogram and five did not have a follow‐up scan. The median interval from diagnosis to last recorded follow‐up echocardiogram was 8.8 months (range 1–62 months) (Table 2).

TABLE 1.

Demographic and clinical features.

	‘Low‐risk’ (N = 86)	‘High‐risk’ (N = 61)	Total (N = 147)
Female, n (%)	35 (40.7%)	18 (29.5%)	53 (36.1%)
Age in years, mean (range)	3.7 (1.0–11.4)	2.9 (0.3–9.4)	3.4 (0.3–11.4)
Days of fever at treatment, mean (range)	5.9 (1–10)	7.9 (3–22)	6.7 (1–22)
Complete KD, n (%)	56 (65.1%)	47 (77.0%)	103 (70.1%)
Incomplete KD, n (%)	30 (34.9%)	14 (23.0%)	44 (29.9%)
Reason not low‐risk, n (%)
Age< 12 months	NA	14	14 (9.5%)
Treated > 10 days	NA	14	14 (9.5%)
IVIG resistant	NA	43	43 (29.3%)
Treatment, n (%)
Corticosteroids	5 (5.8%)	23 (37.7%)	28 (19.0%)
Second IVIG	0	41 (67.2%)	41 (27.9%)
Biologics	0	0	0
Additional antiplatelet treatment (clopidogrel)	0	1 (1.6%)	1 (0.7%)
Anticoagulation	0	2 (3.3%)	2 (1.4%)

Abbreviations: IVIG, intravenous immunoglobulin; KD, Kawasaki disease; NA, not applicable.

TABLE 2.

Echocardiographic results.

	‘Low‐risk’ (N = 86)	‘High‐risk’ (N = 61)	Total (N = 147)
Echocardiography, mean (range)
Number of scans per patient	2.9 (1–6)	3.1 (1–8)	3.0 (1–8)
Day of baseline scan	6.9 (2–12)	9.0 (3–23)	7.8 (2–23)
Months of follow‐up	7.8 (1–50)	10.1 (1–62)	8.8 (1–62)
Echocardiography result at baseline, n (%)
Not performed	1 (1.2%)	0	1 (0.7%)
No aneurysms	81 (94.2%)	52 (85.2%)	133 (90.5%)
Any aneurysm (ZS ≥ 2.5)	4 (4.7%)	9 (14.8%)	13 (8.8%)
Small (ZS ≥ 2.5 to < 5)	3 (3.5%)	6 (9.8%)	8 (5.4%)
Medium (≥ 5 to < 10 and < 8 mm)	1 (1.2%)	2 (3.3%)	4 (2.7%)
Large/giant (ZS ≥ 10 or > 8 mm)	0	1 (1.6%)	1 (0.7%)
Echocardiography result at any time a , n (%)
No aneurysms	82 (95.3%)	52 (85.2%)	134 (91.2%)
Any aneurysm (ZS ≥ 2.5)	4 (4.7%)	9 (14.8%)	13 (8.8%)
Small (ZS ≥ 2.5 to < 5)	2 (2.3%)	6 (9.8%)	8 (5.4%)
Medium (≥ 5 to < 10 and < 8 mm)	2 (2.3%)	1 (1.6%)	3 (2.0%)
Large/giant (ZS ≥ 10 or > 8 mm)	0	2 (3.3%)	2 (1.4%)
Echocardiography result at last follow‐up, n (%)
No repeat echocardiogram performed	1 (1.2%)	4 (6.6%)	5 (3.4%)
No aneurysms	83 (96.5%)	52 (85.2%)	135 (91.8%)
Any aneurysm (ZS ≥ 2.5)	2 (2.3%)	5 (8.2%)	7 (4.8%)
Small (ZS ≥ 2.5 to < 5)	2 (2.3%)	3 (4.9%)	5 (3.4%)
Medium (≥ 5 to < 10 and < 8 mm)	0	1 (1.6%)	3 (2.0%)
Large/giant (ZS ≥ 10 or > 8 mm)	0	1 (1.6%)	1 (0.7%)

Note: ZS, z‐score.

^a Largest observed aneurysms at any time point.

Eighty six patients were classified as being ‘low‐risk’ and 61 were ‘high‐risk’. The rate of any aneurysm at any time point was 8.8% (n = 13) across the whole cohort; 4.7% (n = 4) in ‘low‐risk’ patients and 14.7% (n = 9) in ‘high‐risk’ patients (odds ratio 3.55; 95% confidence interval 1.04–12.11).

Of the 86 ‘low risk’ patients, four had aneurysmal changes on baseline echocardiogram (three small, one medium). No patients with a normal baseline echocardiogram went on to develop aneurysms through the study period. Two patients with small aneurysms had regression of their aneurysms to within the normal range on follow‐up echocardiography while one had persisting small aneurysms at 30 months of follow‐up. The patient with medium aneurysms had improvement, with small aneurysms noted on their last follow‐up echocardiogram 20 months after diagnosis. Of ‘low‐risk’ patients, one did not have a baseline scan, but their follow‐up echocardiogram was normal at 6 weeks and one patient with a normal baseline echocardiogram did not have a documented follow‐up scan.

Among the ‘low‐risk’ group, three were deemed as having management changed by the initial echocardiogram findings. All were given primary adjunctive corticosteroid treatment due to baseline coronary changes; one with medium aneurysms, one with small aneurysms, and one with prominent coronary arteries (maximum z‐score 2.1) and a trivial pericardial effusion. The patient with moderate aneurysms (maximum z‐score 6.53) did not commence additional anti‐platelet therapy; however, this would be recommended by contemporary management guidelines [8].

Of the 61 ‘high risk’ patients, nine had baseline aneurysmal changes (six small, two medium, one large/giant). As in the ‘low‐risk group’, no patients with a normal baseline echocardiogram went on to develop aneurysms through the study period. Three patients with small aneurysms had regression to normal, and three had persistence of small aneurysms (follow‐up range 13–38 months). One patient with medium changes had initial progression to large/giant aneurysms followed by subsequent regression back to medium aneurysms (19 months follow‐up). The one patient with giant aneurysms had persisting giant changes (29 months follow‐up). One patient with medium aneurysms and three patients with normal baseline echocardiograms did not have a documented follow‐up scan.

Among the ‘high‐risk’ group, seven were assessed as having management changed by the initial echocardiogram findings. Three were given adjunctive primary corticosteroid treatment due to baseline aneurysmal changes. One was given adjunctive primary corticosteroid treatment and had closer echocardiogram surveillance followed by the addition of anticoagulation after progression of their aneurysms. One was anticoagulated due to the presence of large/giant aneurysms at baseline. One patient with a complicated presentation with undifferentiated fever was diagnosed with KD based on echocardiographic findings. One patient with medium aneurysms was commenced on an additional antiplatelet agent.

4. Discussion

This study provides contemporary data on a large Australian cohort of patients with KD. It provides local data regarding the frequency of CA lesions on baseline echocardiography in patients stratified using simple, universally available demographic and clinical information. It also provides data on how often the results of early echocardiography changes patient management in current practice. We found an overall CAA rate of 8.8% at any time point and 4.8% at last follow‐up. These outcomes compare with other Australian KD cohorts, including a 2014 Western Australian study that noted a CAA rate of 6.8% among 353 children [11].

Using a simple risk stratification process, we identified a CAA rate of 4.7% in ‘low‐risk’ patients and 14.7% in ‘high‐risk’ patients. No large or giant aneurysms were noted in our ‘low‐risk’ cohort. Predicting outcomes in KD is difficult, and most predictive scores have been designed primarily to predict IVIG resistance rather than CAA development per se. Commonly used risk scoring systems, such as the Kobyashi, Sano, and Egami scores, perform poorly when predicting CAAs, highlighting the difficulties with such a task [5].

We found that three (3.5%) ‘low‐risk’ and seven (11.5%) ‘high‐risk’ patients had a change in their management based on the result of their baseline echocardiogram. The results of this study reflect local practice but can be generalised using guideline driven standardised care. The 2024 AHA scientific statement on KD management recommends dual anti‐platelet for patients with medium or larger CAAs and anticoagulation for those with large/giant CAAs. The role of primary treatment intensification is more controversial, particularly in patients with small CAAs, however they recommend that this should be considered for any patients with aneurysmal changes in the acute phase [8]. If this guidance was adhered to then early echocardiography would have altered treatment in four (4.7%) ‘low‐risk’ and nine (14.7%) high‐risk patients in our cohort.

Our results need to be interpreted considering other limitations of this study. It has the general limitations of all retrospective research and in particular, some data points such as the laboratory features of traditional risk scores were incomplete, for which reason we devised a simpler clinical algorithm to stratify patient risk. The use of isolated coronary z‐scores is a relatively crude measure that does not appreciate the morphology of changes that may be of significance regarding treatment decisions. CA luminal dimensions converted to body surface area‐adjusted z‐scores are, however, recommended as a reasonable tool to stratify risk of poor outcomes such as myocardial ischaemia [1].

5. Conclusion

In our cohort, 4.7% of ‘low‐risk’ patients had CAAs on baseline echocardiography, the majority of whom had a change in their management as a result. Simple risk stratification based on three universally available demographic and clinical features identified a group of patients with a more‐than‐three‐fold risk of developing CAA compared with ‘low‐risk’ patients. We conclude that routine baseline echocardiography is justified, regardless of perceived risk of CAA, to detect a small but important number of lesions that may have implications for treatment decisions.

Ethics Statement

This project was approved by The Royal Children's Hospital Human Research Ethics Committee (HREC) as a quality improvement project.

Conflicts of Interest

The authors declare no conflicts of interest.

Renton W. D., Carmody J., Mactaggart E., and Akikusa J. D., “The Role of Early Echocardiography in Low‐Risk Patients With Kawasaki Disease,” Journal of Paediatrics and Child Health 61, no. 8 (2025): 1301–1305, 10.1111/jpc.70115.