Efficacy of immunoglobulin plus prednisolone for prevention of coronary artery abnormalities in severe Kawasaki disease (RAISE study)

The use of steroid therapy in Kawasaki disease (KD) has a long and complicated history. A recent well-designed randomised controlled trial (RCT) provides strong evidence that the addition of a prolonged course of steroids to the standard treatment of intravenous immunoglobulin (IVIG) and aspirin significantly improves coronary artery (CA) outcomes in Japanese children at high risk of IVIG non-response.[1] Of 2014 KD patients, only the 248 (12%) at high risk of IVIG failure were included, based on the authors' severity score. The incidence of CA abnormalities during the study period was 4/121 (3%) in the steroid treatment group and 28/121 (23%) in the standard treatment group (P < 0.0001); number needed to treat (NNT) to prevent CA abnormalities was five. At week 4, the incidence of CA abnormalities was 4/120 (3%) in the steroid treatment group and 15/120 (13%) in the standard treatment group (P = 0.014); NNT was 10.

These are impressive results. Other findings also suggest that steroids may be of benefit in those at highest risk of IVIG non-response. A RCT of adjunctive primary steroids in an unselected KD population in the USA failed to show any benefit on CA outcomes (albeit with a shorter and different steroid regimen to the Japanese study).[2] However, a post hoc subgroup analysis of those who subsequently required IVIG retreatment suggested a significant beneficial effect on CA outcomes.[2]

Caution is warranted in generalising the RAISE study findings as there are important differences between the study population and KD patients outside Japan.[3] Crucially, Japanese risk-scoring systems, including that used in the RAISE study, have reasonable specificity but unacceptably low sensitivity in non-Japanese populations.[4] There are other potentially important differences in the RAISE study: treatment was commenced at a median of day 4, which is earlier than is usual in Australia and other non-Japanese populations. Finally, the steroid regimen is partly intravenous and relatively prolonged, potentially increasing inpatient costs.

The development of severity scores predictive of IVIG non-response in non-Japanese patients would allow further studies of adjunctive regimens, including steroids, in those with the most severe KD.

Reviewers: Katherine Chen, Katherine.chen@rch.org.au; David Burgner, david.burgner@mcri.edu.au; Nigel Curtis, nigel.curtis@rch.org.au