Efficacy between low and high dose aspirin for the initial treatment of Kawasaki disease: Current evidence based on a meta-analysis

Abstract

Background: Kawasaki disease (KD) is now the leading cause of acquired heart disease in children in developed countries. Intravenous immunoglobulin (IVIG) and aspirin were considered as the standard initial treatment of KD for decades. However, the optimal dose of aspirin has remained controversial. In recent years, many studies compared the efficacy of low-dose with high-dose aspirin in the acute phase of KD, but the results have not always been consistent. Therefore, we performed this meta-analysis to evaluate the efficacy of low-dose aspirin compared with high-dose for the initial treatment of KD.

Methods: Studies related to aspirin therapy for KD were selected from PubMed, EMBASE, the Cochrane Central Register of Controlled Trials, China National Knowledge Infrastructure, and Google scholar through Mar 25th, 2019. Data were analyzed using STATA Version 15.1. Additionally, publication bias and sensitivity analysis were also performed by STATA version 15.1.

Results: Six studies were included in our analysis of the rate of coronary artery lesion (CAL), five reports for IVIG-resistant KD (rKD), and four for the duration of fever and hospitalization. However, no significant differences were found between low-dose and high-dose aspirin groups in the incidence of CAL (risk ratio (RR), 0.85; 95%CI (0.63, 1.14); P = 0.28), the risk of rKD (RR, 1.39; 95%CI (1.00, 1.93); P = 0.05), and duration of fever and hospitalization (the mean standard deviation (SMD), 0.03; 95%CI (-0.16, 0.22); P = 0.78).

Conclusion: Low-dose aspirin (3-5 mg·kg-1·d-1) may be as effective as the use of high-dose aspirin (≥30 mg·kg-1·d-1) for the initial treatment of KD. Further well-designed randomized clinical trials are needed to evaluate the efficacy of low-dose aspirin for the initial treatment of KD.

Fig 1. PRISMA flow chart of studies inclusion and exclusion.

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med. 6(7): e1000097. doi:10.1371/journal.pmed.1000097. For more information, visit www.prisma-statement.org.

Fig 2. Forest plot for comparison of the incidence of coronary artery lesion identified in the meta-analysis of six trials using random-effect model.

Only the first author of each study is given. Test for overall effect, z = 1.09, P = 0.28; test for heterogeneity, I² = 53.1%, P = 0.058. RR, risk ratio. CI, confidence interval.

Fig 3. Egger’s publication bias plots for the assessment of potential publication bias.

Each dot represents each study in the meta-analysis. Asymmetry of the dot distribution between regression lines indicates potential publication bias. (A) For the risk measurement of CAL, the Egger’s plot did not show marked asymmetry, P = 0.601, t = -0.57, 95% CI (-3.22, 2.12), (B) For the risk analysis of rKD, the Egger’s plot did not show significant asymmetry, P = 0.950, t = -0.07, 95% CI (-4.87, 4.66), (C) For the evaluation of days of fever or hospital, the Egger’s plot also did not show significant asymmetry, P = 0.390, t = -1.09, 95%CI (-10.37, 6.19). This Egger’s plot indicates no publication bias with a P value >0.05. RR, risk ratio. CI, confidence interval.

Fig 4. Forest plot for comparison of the rate of IVIG-resistant KD in the meta-analysis of five trials using random-effect model.

Only the first author of each study is given. Test for overall effect, z = 1.98, P = 0.05; test for heterogeneity, I² = 65.6%, P = 0.020. RR, risk ratio. CI, confidence interval.

Fig 5. Forest plot for comparison of days of fever or hospital in the meta-analysis of four trials using random-effect model.

Only the first author of each study is given. SMD, the mean standard deviation. Test for overall effect, z = 0.27, P = 0.78; test for heterogeneity, I² = 76.2%, P = 0.006. RR, risk ratio. CI, confidence interval.

Fig 6. The meta-regression of the enrolled studies.

(A) For the regions of studies, the meta-regression did not detect it is a dramatic impact on the homogeneity of the enrolled studies, P = 0.220, t = 1.45, 95%CI (0.87, 1.55). (B) For diagnostic criteria for KD, the meta-regression did not find it is a dramatic impact on the homogeneity of the enrolled studies, P = 0.542, t = 0.67, 95%CI (0.43, 4.02). (C) For specific dosages of high-dose aspirin groups the meta-regression did not detect it is a dramatic impact on the homogeneity of the enrolled studies, P = 0.713, t = -0.39, 95%CI (0.35, 2.19). The meta-regression could determine the correlation between the potential factors and the existing heterogeneities. When a significant difference was discovered, the factor should have a dramatic impact on the homogeneity of the enrolled studies with a P value >0.05. rr, risk ratio. CI, confidence interval.

Fig 7. Sensitivity analysis of the individual trials on the results.

(A) For the incidence analysis of CAL, (B) For the risk analysis of rKD, (C) For the evaluation of days of fever or hospital. Not any single study was detected to incur undue weight in the analysis.