Objective Patients with Kawasaki disease (KD) with coronary artery enlargement at diagnosis are at the highest risk for persistent coronary artery aneurysms (CAAs) and may benefit from primary adjunctive anti-inflammatory therapy beyond intravenous immunoglobulin (IVIG). We evaluate the effect of primary adjunctive corticosteroid therapy on outcomes in patients with CAA at diagnosis.
Design Single-centre, retrospective review.
Patients Patients with KD diagnosed within 10 days of fever onset and with baseline CA z-score ≥2.5.
Interventions Primary treatment with IVIG (n=162) versus IVIG plus corticosteroids (n=48).
Main outcome measures Treatment resistance (persistent fever >36 hours after initial treatment), CAA regression rate.
Results Of the 92 patients with KD who received corticosteroids at our institution from 2012 to 2019, 48 met the inclusion criteria for primary adjunctive therapy. The corticosteroid group was younger and had larger baseline CAAs compared with historical controls. Demographics and laboratory values were otherwise similar between groups. The corticosteroid group had a less treatment resistance (4% vs 30%, p=0.003) and a greater improvement in C reactive protein. After adjusting for baseline CA z-score, age and baseline bilateral versus unilateral CAA, the corticosteroid group had a higher odds of (OR 2.77 (1.04, 7.42), p=0.042) and a shorter time to CAA regression (HR 1.94 (1.27, 2.96), p=0.002).
Conclusion Primary adjunctive corticosteroid therapy is associated with decreased initial treatment resistance, greater improvement in inflammatory markers and higher likelihood of CAA regression in patients who have CAA at diagnosis. Multi-centre, randomised controlled trials are needed to confirm the benefits of corticosteroids in patients with CAA at diagnosis and to compare corticosteroids with other adjunctive therapies.
Statistics from Altmetric.com
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?
Primary adjunctive therapy improves coronary artery and clinical outcomes in high-risk Japanese children with high-risk Kawasaki disease (KD) based on clinical risk scores.
Clinical risk scores are not sensitive or specific in the American population.
Coronary artery aneurysm (CAA) at the time of diagnosis is the strongest risk factor in the American population for persistent aneurysm.
What this study adds?
Use of primary adjunctive corticosteroids in patients with KD with CAA at the time of diagnosis is associated with higher rate of CAA regression.
Use of primary adjunctive corticosteroids in patients with KD with CAA at the time of diagnosis is associated with lower rate of treatment resistance.
Kawasaki disease (KD) is an acute vasculitis that preferentially affects medium-sized arteries, particularly the coronary arteries (CAs)1 2 with involvement ranging from transient, mild dilatation to giant coronary artery aneurysms (CAAs). Patients with giant CAAs are at risk for cardiac events including stenosis, thrombosis and myocardial infarction.3–5 Intravenous immunoglobulin (IVIG) is well established as first-line therapy and reduces CAA incidence ~5-fold,6 but a significant burden of CA disease remains. Indeed, using the American Heart Association (AHA) z-score criteria for CAA, ~20%–25% of children and 50% of infants develop CAA despite timely IVIG.7 To further reduce CAA incidence, several adjunctive anti-inflammatory therapies including corticosteroids, infliximab, etanercept and ciclosporin have been explored.8–12 8–10 12–14 Use of corticosteroids has long been controversial, but there is growing evidence supporting their use in select populations. The strongest data for primary adjunctive corticosteroid therapy comes from the Randomised Controlled Trial to Assess Immunoglobulin Plus Steroid Efficacy (RAISE) which demonstrated that in high-risk Japanese children (defined by Kobayashi score), IVIG plus corticosteroids led to improved outcomes compared with IVIG alone.9 Subsequent studies and meta-analyses have also shown that corticosteroids improve outcomes, particularly in Japanese children at high risk based on clinical scores.13 15–17
Despite persuasive evidence for the efficacy of corticosteroids in high-risk Japanese children, the generalisability of the RAISE study results to the non-Japanese population is problematic due in part to low sensitivity and specificity of clinical scores in the non-Japanese population.1 18–20 Moreover, the RAISE trial intended to exclude patients with baseline CAA, which has been shown to be the most powerful predictor of persistent CAA in the US population.9
In response to the compelling Japanese data on corticosteroid efficacy, our KD treatment group adopted a protocol for treatment of high-risk patients with KD that administers IVIG plus a primary adjunctive corticosteroid regimen similar to that used in the RAISE trial. In this study, we describe the outcomes in US patients treated with IVIG plus primary adjunctive corticosteroids and compare their outcomes with those of similar historical patients who did not receive primary adjunctive corticosteroid therapy, with a particular interest in patients who received corticosteroids for the indication of CAA at diagnosis.
In this single-centre retrospective study, we reviewed all patients with KD managed at Boston Children’s Hospital from January 2000 through May 2019. Our centre adopted the use of RAISE-dose corticosteroid regimen in 2012.9 The RAISE trial used the Kobayashi clinical score as selection criteria for corticosteroids. Due to inadequate sensitivity and specificity of scoring systems in the US population,1 18–20 our centre has used the following criteria as indications for primary adjunctive corticosteroids since 2012: CAA at the time of diagnosis (defined as z-score >2.5 for proximal right coronary artery (RCA) or left anterior descending coronary artery (LAD)), age <6 months, presentation with KD shock syndrome or KD associated with macrophage activation syndrome. All patients with baseline CA z-score >2.5 who initially presented to our centre within 10 days of fever were treated with primary adjunctive steroids according to the RAISE trial dosing. Prior to 2012, RAISE-dose corticosteroids were not used at our centre and there was wide practice variation in indications for and choice of anti-inflammatory therapy beyond IVIG in patients with KD with high-risk features.
In the descriptive section of the study, we included all patients who received modified RAISE-dose corticosteroids regardless of indication or timing of corticosteroid administration. For the analysis comparing those receiving primary adjunctive corticosteroid therapy with historical controls who did not receive primary adjunctive corticosteroids, corticosteroid patients were included if they (1) received IVIG within 10 days of fever onset, (2) received primary adjunctive corticosteroids and (3) had CAA on echocardiogram at diagnosis (CA z-score ≥2.5 for proximal LAD or RCA). For the historical control group, patients were included if they (1) received IVIG within 10 days of fevers onset, (2) did not receive primary adjunctive corticosteroids and (3) had CAA on echocardiogram at diagnosis. Second episodes of KD were excluded.
Demographic, clinical, laboratory and cardiac imaging data were collected from medical records. Laboratory values at diagnosis, 1 week post-IVIG and 4 weeks post-IVIG were collected. Echocardiograms were obtained at diagnosis and then every 2–3 days until CA size was stable. Outpatient echocardiograms were routinely obtained at 1 and at ~4 weeks post-IVIG, and more frequently at clinician’s discretion. Echo data included measurements and z-scores for left main coronary artery (LMCA), proximal and distal LAD, circumflex, proximal and distal RCA, and posterior descending CA. Using the reported measurements and body surface area, z-scores for the proximal LAD and proximal RCA were calculated using the Boston z-score formulas.21 22 We defined maximum RCA or LAD z-score (zMax) as the higher value between RCA and LAD z-score on each echocardiogram. LMCA z-score was not used for inclusion due to previously reported variability in LMCA anatomy and measurement.23–25 Analyses included CA size and z-scores at initial and latest echocardiograms, as well as the highest zMax on any echocardiogram over follow-up.
We defined and classified CAA according to AHA guidelines as small (z-score=2.5–4.9), moderate (z-score=5–9.9) and large/giant (z-score ≥10).1 CAA regression was defined as normalisation of the internal lumen diameter for all CA segments (z-score <2). CAA progression was defined as an increase in LAD or RCA z-score of ≥2 units from initial echo to maximal size. Resistance to initial treatment was defined as persistent fever >36 hours after completion of initial IVIG. In the primary adjunctive corticosteroid group, several patients received a second dose of IVIG due to baseline characteristics (baseline CAA with z-score >4 and/or KD shock syndrome) and not due to persistent fever.
All patients received IVIG (2 g/kg) and medium-dose aspirin (30–50 mg/kg/day) until defervescence, followed by low-dose aspirin (3–5 mg/kg/day). IVIG treatment was considered delayed if given >10 days after fever onset. Patients were classified as receiving primary adjunctive corticosteroids if corticosteroids were initiated within 24 hours of initiation of the first IVIG and as secondary/rescue corticosteroids if initiated ≥24 hours after initial IVIG. Corticosteroids were started as 1 mg/kg every 12 hours of intravenous methylprednisolone and transitioned to 1 mg/kg twice daily of oral prednisone at hospital discharge. Once C reactive protein (CRP) was <1 mg/dL, corticosteroids were tapered over 2–4 weeks with the dose cut in half every 5 days. Repeat IVIG and rescue anti-inflammatory medications were given at the discretion of the treating team. Our treatment protocol recommends a second dose of IVIG for all patients presenting with CA z-score ≥4. Patients with giant CAA (z-score ≥10 or absolute dimension >8 mm) were anticoagulated in accordance with AHA guidelines.1
Demographic and clinical data are presented as count (per cent) or median with (IQR) or (range). Patient characteristics and outcomes were compared for the two treatment groups using Fisher’s exact test or Wilcoxon rank sum test. Differences in time to CAA regression were displayed using Kaplan-Meier curves and evaluated using the log-rank test. Multivariable linear regression was used to compare maximum z-score over follow-up and change in z-score between initial and final follow-up echocardiograms between groups, adjusting for age, baseline CAA z-score and bilateral versus unilateral CAA. Logistic regression was used to compare occurrence of CAA regression over 2 months, and Cox regression to evaluate time to CAA regression, adjusting for the same baseline characteristics.
Of 433 patients with KD treated during the acute phase of illness at our centre from 2012 to May 2019, 92 (21%) patients were treated with RAISE-dose corticosteroids, of whom 70 received primary adjunctive corticosteroids and 22 received secondary/rescue corticosteroids (figure 1). Baseline characteristics, timing and indications for corticosteroids are described in table 1. Most patients (74%) in the primary adjunctive corticosteroid group had CAA at diagnosis as an indication for corticosteroids. The most common indications for rescue/secondary corticosteroid treatment were new or worsening CAA and IVIG resistance. In the primary adjunctive corticosteroid cohort, progression in CAA size after treatment was infrequent and directly proportional to the size of CAA at diagnosis. Among the 14 patients with normal initial CA size who were treated with primary adjunctive corticosteroids, none developed CAA. Of those presenting with small CAA, progression in CA size was infrequent (2/39, 5%), whereas CAA progression was more common in those with moderate CAA (2/12, 17%) or large/giant CAA (2/5, 40%) at diagnosis. Aneurysm regression rate was high (81%) in the primary adjunctive corticosteroid cohort and even higher when limiting analyses to patients treated ≤day 10 of illness and with baseline CAAs that were small or moderate (91%).
We compared outcomes of 48 patients who were treated with primary adjunctive corticosteroids within 10 days of fever onset and whose indication for corticosteroids was CAA at diagnosis (figure 1) with a similar historical control group (n=162) who did not receive primary adjunctive corticosteroids. Compared with historical patients, corticosteroid patients were younger and more likely to have ≥moderate CAA at diagnosis (27% vs 12%, p=0.016) (table 2). Demographic, laboratory and baseline echocardiographic values were otherwise similar between groups. The number of patients receiving rescue therapy due to worsening CAA and/or persistent fever was similar between groups. The primary adjunctive corticosteroid patients were more likely to receive a second dose of IVIG in accordance with our treatment protocol, which recommended a second dose of IVIG for all patients presenting with CA z-score >4.
Resistance to initial treatment was lower in the primary adjunctive corticosteroid group compared with historical controls (4% vs 30%, p<0.001) (table 3). On 1 week post-IVIG laboratories, CRP was lower and haematocrit was higher in the corticosteroid group. White blood cell count and absolute neutrophil count were higher at 1 week in the corticosteroid group, and ~25% developed significant leukocytosis (white blood cell count >30 000 cells/µL), which resolved after steroids were discontinued. Laboratory values at 4 weeks post-IVIG did not differ between groups.
Maximal CA z-score over follow-up and the percentage of patients experiencing progression in CAA size after baseline echo did not differ significantly between groups (table 3). However, the primary adjunctive corticosteroid group had a larger decline in CAA z-score over follow-up. After adjusting for baseline CAA z-score, age and bilateral CAA, the corticosteroid group was more likely to have CAA regression and had shorter time to CAA regression compared with historical controls (table 4). There were no treatment-related serious adverse events and no major adverse cardiac events in either group with median follow-up of 4.2 years in steroid group.
The 2017 AHA KD Guidelines recommend consideration of primary adjunctive, anti-inflammatory therapy for patients at high risk for CAA.1 However, lack of evidence regarding which agents to use as well as difficulty in identifying high-risk children in non-Japanese populations prevented the guidelines from making specific recommendations, leading to wide practice variation in adjunctive therapy. In this single-centre study, primary adjunctive corticosteroid therapy in patients with CAA at diagnosis was associated with lower treatment resistance rates and higher likelihood of CAA regression when controlling for baseline characteristics. Our findings as well as the increasing evidence of efficacy of steroids in high-risk Japanese children9 13 15 16 26 suggest that RAISE-dose corticosteroids should be considered for primary adjunctive therapy in patients with KD with CAA at diagnosis.
Corticosteroids have long been controversial in KD, particularly in patients with CAA. Corticosteroids were used as the initial therapy prior to the first report of IVIG efficacy, with the earliest reports suggesting a high rate of cardiac events and a potential risk of CA rupture.27 Although these early studies were observational and prone to bias, concerns about steroid use in patients with CAA have lingered.28 In this study, we demonstrate that a regimen of RAISE-dose corticosteroids plus IVIG is safe in patients with CAA at diagnosis. No patients had CA ruptures, adverse medication-related events or adverse cardiac events over medium-term follow-up. Moreover, the use of primary adjunctive corticosteroids was associated with improvement in some, but not all, CA outcomes. Compared with historical controls, corticosteroid therapy was associated with a higher likelihood of CAA regression to normal internal lumen diameter when adjusting for baseline characteristics. This finding helps alleviate previously raised concerns that corticosteroids may impair vascular remodelling and prevent CAA regression.28 While CAA regression rate improved with corticosteroid therapy, maximal CA z-scores and CAA progression rates were similar between groups. It is important to note that ~80% of patients in both groups had their largest z-score at diagnosis, indicating that the majority of patients respond well to IVIG when given within 10 days.
In addition to being safe and leading to an improvement in rate of CAA regression, we found that primary adjunctive corticosteroid therapy was associated with a lower rate of initial treatment resistance and faster improvement in the inflammatory markers. By 1 week after IVIG administration, the primary adjunctive corticosteroid group had a lower CRP and higher haematocrit than historical controls.
Our findings on CA outcomes are largely concordant with the landmark RAISE trial9 as well as the post-RAISE observational study13 evaluating adjunctive corticosteroids in KD. Although we identified improved CA outcomes with use of primary adjunctive corticosteroids in this study, the benefit in our study is less marked than the RAISE trial. This is likely related to several important differences between our patient population and the RAISE trial patients. Most notably, the RAISE trial excluded patients with CAA at diagnosis, which was our primary inclusion criterion. Another key difference is the high-risk criteria used: Kobayashi score in RAISE trial and CAA at diagnosis in our analysis. The primary outcome assessed in the RAISE study was incidence of CAA in acute period, while all patients in our study had baseline CAA based on inclusion criteria, and thus we assessed CAA regression over follow-up. Other notable differences include patient ethnicity and later day of illness at treatment in our study (median day of illness at initial treatment: 7 vs 4 days). Later day of illness at initial treatment as well as use of AHA z-score criteria likely account for the higher rate of baseline CAA in this study compared with prior Japanese trials.
Data evaluating adjunctive primary therapy in patients presenting with CAA are limited. A retrospective study in high-risk North American patients with KD suggested that primary adjunctive therapy with either infliximab or corticosteroids is associated with a lower rate of subsequent CAA enlargement compared with IVIG alone.14 A retrospective study evaluating primary adjunctive therapy with infliximab in patients with baseline CAA showed no difference in CA outcomes between groups.12 A North American trial showed that primary adjunctive therapy with etanercept did not improve treatment resistance rates, but may improve CA outcomes in patients with baseline CAA.10 No trials to date have tested whether primary adjunctive therapy is superior to IVIG alone in patients with baseline CAA or compared the efficacy of different adjunctive therapies.
There are several important limitations to our findings. As a single-centre retrospective study, the treatment groups differed in some baseline features including era of treatment and baseline CAA size. Notably, despite having larger baseline CAA, the primary adjunctive corticosteroid group had better CA outcomes than historical controls. The majority of patients in both groups had small CAA which limited our ability to evaluate CA outcomes in patients with ≥moderate CAA. Our institutional KD treatment guideline recommends a second dose of IVIG for all patients with CAA z-score >4 regardless of response to initial therapy and accordingly we cannot exclude a potential benefit of second IVIG in these patients.
In conclusion, we found that primary adjunctive corticosteroids are associated with decreased initial treatment resistance, faster resolution of inflammation and higher likelihood of CAA regression in patients with KD with CAA at diagnosis. Randomised controlled trials are needed to confirm the benefits of corticosteroid therapy in high-risk KD and to compare corticosteroids with other adjunctive therapies.
Correction notice This paper has been corrected since it was published online. The last author's surname was spelled incorrectly.
Contributors KGF, KG, AB, MBS, RS, AD, TG, SDF, JWN: substantial contributions to the conception of the work, data acquisition, analysis, and interpretation; drafting and revising the work for important intellectual content, final approval of the manuscript submitted.
Funding This work was supported by The McCance Family Foundation and Gordon and Marilyn Macklin Foundation.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval Institutional Review Board approval was obtained (IRB #M06-09-0436).
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data will be made available on reasonable request.