Pediatrics · Medical Literature
Randomized Trial of Adjunctive Prednisolone for Kawasaki Disease.
Abstract
The effect of adjunctive glucocorticoids in the primary treatment of Kawasaki disease in unselected patients remains unknown. In this multicenter, open-label, randomized, controlled trial in China, we assigned participants with newly diagnosed Kawasaki disease in a 1:1 ratio to receive prednisolone plus standard treatment or standard treatment alone. The primary outcome was the occurrence of coronary-artery lesions at 1 month after illness onset. Prespecified key secondary outcomes, for which analyses were not controlled for multiplicity, included receipt of rescue therapy, duration of fever, change in the C-reactive protein (CRP) level, and changes in coronary-artery z scores. A total of 3208 participants underwent randomization, with coronary-artery lesions detected at baseline in 870 of 3184 participants (27.3%). At 1 month, coronary-artery lesions were detected in 16.0% of the participants receiving prednisolone plus standard treatment and in 13.8% of those receiving standard treatment alone (adjusted risk difference, 1.1 percentage points; 95% confidence interval, -1.0 to 3.4; P = 0.31). Rescue therapy was used in 4.6% of the participants receiving prednisolone plus standard therapy and in 10.1% of those receiving standard treatment alone; the median duration of fever was 8.4 hours and 13.2 hours, respectively, and the reductions in the C-reactive protein level at 72 hours were 67.5 mg per liter and 59.8 mg per liter. Decreases in coronary-artery z scores were similar in the two groups. At 3 months, the incidence of coronary-artery lesions was 12.6% with prednisolone plus standard therapy and 10.5% with standard treatment alone; the percentage of participants with progression of coronary-artery lesions was 28.6% and 28.9%, respectively, and the incidence of medium-to-giant coronary-artery aneurysms was 1.9% and 1.1%. The overall incidence of adverse events did not differ significantly between the two groups. The addition of prednisolone to standard primary treatment for Kawasaki disease did not reduce the incidence of coronary-artery lesions at 1 month after illness onset. (Funded by the Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences and the National Natural Science Foundation of China; ClinicalTrials.gov number, NCT04078568.).
Clinical relevance
In unselected children with newly diagnosed Kawasaki disease, does the addition of prednisolone to standard primary therapy (IVIG plus aspirin) reduce the incidence of coronary-artery lesions at one month after illness onset compared with standard therapy alone? __SNAPSHOT__ {"studyDesign":"Phase 3 multicenter, open-label, randomized controlled trial (1:1 allocation)","setting":"28 pediatric centers in China (Chinese Kawasaki Disease Collaboration Network); led by Children's Hospital of Fudan University, National Children's Medical Center, Shanghai","sampleSize":"n=3,208 randomized (prednisolone + standard treatment: n=1,604; standard treatment alone: n=1,604); coronary-artery lesion status at baseline evaluable in 3,184 participants","population":"Children ≥1 month of age meeting American Heart Association (AHA) criteria for newly diagnosed Kawasaki disease; mean age approximately 2.7 years; baseline coronary-artery lesions present in 870/3,184 participants (27.3%)","intervention":"Adjunctive prednisolone added to standard therapy: intravenous methylprednisolone 1.6 mg/kg/day until afebrile for 3 days, followed by oral prednisolone 2 mg/kg/day (maximum 60 mg/day) until CRP normalization, then tapered over 15 days; combined with standard IVIG (2 g/kg; maximum 60 g) and oral aspirin (30 mg/kg/day in three divided doses, reduced after fever resolution and CRP normalization, maintained for ≥6 weeks)","comparator":"Standard therapy alone: IVIG 2 g/kg (maximum 60 g) plus oral aspirin 30 mg/kg/day in three divided doses, with aspirin reduced and maintained ≥6 weeks after fever resolution and CRP normalization","primaryOutcome":"Occurrence of coronary-artery lesions (defined as a Z-score ≥2 in any coronary segment per 2017 AHA criteria) at 1 month after illness onset, assessed by two-dimensional echocardiography","followUp":"Assessments at enrollment, 2 weeks, 1 month, and 3 months after illness onset; echocardiographic follow-up planned through 12 months, but 6-month and 1-year data were not available in this analysis","fundingSource":"Government (Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences; National Natural Science Foundation of China); no industry funding reported; authors declared no relevant financial disclosures"}
Key findings
PRIMARY OUTCOME: Coronary-artery lesions at 1 month were detected in 16.0% of participants in the prednisolone group versus 13.8% in the standard-treatment group (adjusted risk difference [ARD] +1.1 percentage points; 95% CI, −1.0 to 3.4; P=0.31). The trial therefore failed to demonstrate superiority of adjunctive prednisolone for the primary endpoint; the direction of the point estimate numerically favored standard treatment alone, though the confidence interval crossed zero and included small potential benefits or harms in either direction. Absolute risk difference NNT/NNH: not calculable as a benefit; if the point estimate were taken at face value, prednisolone would be associated with a marginally higher (not lower) coronary-artery lesion rate (+1.1 pp, calculated). At 3 months, coronary-artery lesions were present in 12.6% (prednisolone) versus 10.5% (standard treatment), again numerically favoring standard treatment; progression of coronary-artery lesions was 28.6% versus 28.9%, respectively — essentially identical between groups. Medium-to-giant coronary aneurysms (Z-score ≥5) at 3 months occurred in 1.9% (prednisolone) versus 1.0% (standard treatment) (risk ratio 0.98; 95% CI, 0.62 to 2.06 — not statistically significant, and confidence interval wide). KEY SECONDARY OUTCOMES (not corrected for multiplicity): Rescue therapy was required in 4.6% of the prednisolone group versus 10.1% of the standard-treatment group (risk ratio 0.45; 95% CI, 0.34 to 0.59) — a statistically significant and potentially clinically meaningful reduction of approximately 55% in relative terms (ARD −5.5 pp, calculated; NNT ≈18, calculated). However, notably, among participants who did ultimately require rescue therapy, coronary-artery lesions at 1 month were present in 45.2% (prednisolone group) versus 25.6% (control group) (ARD +13.5 pp; 95% CI, 2.7 to 24.3), raising concern that glucocorticoid masking of fever may have delayed recognition of IVIG failure in some prednisolone-treated patients. Median fever duration after IVIG initiation was shorter in the prednisolone group (8.4 hours vs. 13.2 hours; HR 0.67; 95% CI, 0.61 to 0.73). CRP reduction at 72 hours post-IVIG was greater in the prednisolone group (−67.5 vs. −59.8 mg/L; mean difference −7.6 mg/L; 95% CI, −10.9 to −4.3). SAFETY: Adverse events occurred in 2.1% (prednisolone) versus 2.8% (standard treatment) of participants (P=0.17); no deaths were recorded during the trial. The adverse event profile was similar between groups.
Practice implications
This large phase 3 RCT (n=3,208) provides the most definitive evidence to date that adjunctive prednisolone added to standard primary therapy (IVIG plus aspirin) does NOT reduce coronary-artery lesion incidence in unselected children with Kawasaki disease. The null finding for the primary endpoint is robust given the sample size and the direction of the point estimate (slightly favoring control). The trial DOES establish that prednisolone reduces the need for rescue IVIG therapy, shortens fever duration, and accelerates CRP decline — findings consistent with enhanced suppression of systemic inflammation — but these secondary gains did not translate into coronary protection. A critical and clinically important signal is that among the subset who still required rescue therapy despite prednisolone, coronary-artery lesion rates were substantially and significantly higher (45.2% vs. 25.6%), suggesting that steroid-induced masking of fever may delay identification of IVIG non-responders and result in a worse coronary outcome in that subgroup. This finding warrants heightened echocardiographic surveillance in prednisolone-treated patients who appear to defervesce but may harbor ongoing coronary inflammation. The population studied is unselected children with KD in China, a population with a higher baseline rate of coronary-artery lesions than Western cohorts; generalizability to other ethnic populations and to IVIG-resistant or high-risk subgroups requires caution. The trial was not powered to evaluate outcomes in pre-specified subgroups (e.g., high-risk for IVIG resistance), and current guidelines already recommend glucocorticoids in high-risk patients — a recommendation that this trial neither supports nor refutes for that specific subpopulation. Practice should not change for unselected KD patients on the basis of this trial: routine adjunctive prednisolone is not supported. Ongoing 6-month and 12-month follow-up data are needed before the full safety and coronary outcome picture is complete. Only the abstract was used as the source for this summary; full-text review is recommended for detailed subgroup analyses, rescue therapy protocols, and echocardiographic methodology. __LIMITATIONS__ Open-label design: inability to blind treating clinicians or echocardiographers introduces ascertainment and performance bias, and is particularly problematic given fever masking by glucocorticoids,Echocardiographic assessments performed locally at 28 sites rather than through a centralized core laboratory, raising concerns about inter-site measurement variability and Z-score consistency,Trial enrollment fell below pre-specified targets during the COVID-19 pandemic, with an increased dropout rate, potentially compromising statistical power — particularly for subgroup analyses,The primary analysis included some participants who did not fully meet eligibility criteria (per Medscape report), which may introduce contamination bias,Secondary outcomes were not corrected for multiplicity, increasing the risk of false-positive findings among the secondary results,Subgroup and exploratory analyses are acknowledged to be underpowered, precluding reliable conclusions about high-risk or IVIG-resistant subgroups,Follow-up data at 6 months and 12 months were not available for this analysis; longer-term coronary outcomes (e.g., aneurysm regression, late cardiac events) remain unknown,Single-country (China) multicenter design limits generalizability to non-Asian populations who may have different KD phenotypes, baseline coronary lesion rates, and IVIG response rates,The concerning signal of higher coronary-artery lesion rates in prednisolone-treated rescue therapy recipients (45.2% vs. 25.6%) is derived from a small subgroup and was not a prespecified primary analysis; it should be hypothesis-generating rather than definitive,No placebo control arm; open-label design makes behavioral and management differences between groups (beyond the allocated treatment) difficult to exclude __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ In this large phase 3 open-label RCT of 3,208 children with Kawasaki disease conducted across 28 Chinese centers, adjunctive prednisolone added to standard IVIG and aspirin therapy did not reduce coronary-artery lesion incidence at 1 month (16.0% vs. 13.8%; ARD +1.1 pp; 95% CI, −1.0 to 3.4; P=0.31), despite significantly reducing rescue therapy use and accelerating fever and CRP resolution. Routine addition of prednisolone to primary KD therapy in unselected patients is not supported by this evidence, and clinicians should be alert to the possibility that glucocorticoid-induced fever suppression may mask IVIG treatment failure and delay recognition of progressive coronary injury in a subset of patients.