Kawasaki disease (KD) is an acute, self-limited vasculitis characterised by fever, bilateral non-exudative conjunctivitis, erythema of the lips and oral mucosa, changes in the peripheral extremities, rash, and cervical lymphadenopathy, observed most commonly in infants and children.1 The cause of the illness remains unknown, and 15–25% of untreated children develop coronary artery dilatations or aneurysms.2 Intravenous immunoglobulin (IVIG) is now well recognised as the standard treatment for KD,3 and coronary complications have become less common. However, 10–20% of patients do not respond to the initial treatment with IVIG, and the incidence of coronary artery abnormalities in those patients is still high.4 5 IVIG-resistant patients are usually treated with additional IVIG,6 but other treatments are steroids,5 7^–9 ulinastatin10 (which is a human trypsin inhibitor used in Japanese patients mainly), plasma exchange,11 immunosuppressants,12 and, recently, tumour necrosis factor-α antagonist (inflixmab).13 The treatment regimen for IVIG-resistant patients varies between institutions, and the best option has not yet been established. Steroid treatment for KD was once avoided because of an early study by Kato et al14 which reported that steroids may increase the incidence of coronary artery abnormalities. However, since then, there have been several reports suggesting that steroids reduce fever and prevent coronary complications without any apparent adverse effects.5 7^–9 15 16 Therefore, steroid therapy for KD has been reconsidered, and use of steroids for IVIG-resistant cases is increasingly reported. Steroid is now recognised as an additional treatment for KD.17^–21 In this study, we attempted to establish a protocol for the treatment of IVIG-resistant KD at our hospital and affiliated institutions, by investigating the use of intravenous methylprednisolone (IVMP) pulse therapy as an additional treatment and analysing its usefulness.

SUBJECTS AND METHODS

Subjects

The subjects were 411 children (243 boys and 168 girls), 1–141 months of age (mean 28 months), who presented with KD to Juntendo University Hospital or one of its seven affiliated medical institutes in Japan and were treated with a single dose of IVIG along with 30 mg/kg/day of aspirin between May 2003 and March 2006. The diagnosis of KD was based on the criteria of the Japanese Kawasaki Disease Research Committee.22 Patients who did not receive IVIG were excluded.

Treatment protocols

All patients were given a single dose of IVIG (2 g/kg) over 12–24 h and aspirin (30 mg/kg/day) (box 1). Aspirin was tapered to 10 mg/kg/day after alleviation of fever and to 5 mg/kg/day subsequent to normalised serum C-reactive protein (CRP) concentration. We defined IVIG resistance as persistent or recurrent fever 36 h after the initial IVIG treatment. Per protocol, IVIG-resistant patients received IVMP (methylprednisolone 30 mg/kg/day over 2 h for 3 consecutive days with heparin infusion 10–20 U/kg/h) as second-line therapy. Prednisolone (1 mg/kg/day) was administered and tapered over 7 days as aftercare. Treatment with additional IVIG was performed only for patients whose family refused megadose steroid therapy. We then retrospectively compared the effectiveness of these treatments, the rate of coronary lesions, and adverse effects between patients who received IVMP (IVMP group) and patients who were treated with additional IVIG (1–2 g/kg) (additional-IVIG group).

We defined IVMP resistance/non-responder as the persistence of fever during IVMP administration or recurrence after IVMP of KD-associated fever along with symptoms and inflammatory reactions that required subsequent treatment. Additional-IVIG resistance/non-response was defined as persistent or recurrent fever within 36 h of additional-IVIG treatment. We administered a third-line therapy to patients who did not respond to the second-line therapy (IVMP or additional IVIG). However, it was not fixed and depended on the circumstances of each case; it was either further additional IVIG or IVMP.

Coronary artery aneurysm (CAA)

CAAs were assessed by two-dimensional echocardiography performed once or twice a week during the acute phase of KD. CAA was defined as an internal lumen diameter of more than 3 mm for a child younger than 5 years old or more than 4 mm for those 5 years old and above. A diameter of more than 8 mm was considered to be a giant aneurysm. A coronary artery abnormality regressing within 30 disease days was defined as a transient dilatation.

Statistical analysis

The Mann–Whitney U test was used to analyse comparisons between the two groups. Gender was compared using the χ² test. Data are expressed as mean (SD), and p<0.05 was considered significant.

RESULTS

Treatment results and CAAs

Of the 411 patients with KD treated with a single dose of IVIG, 348 (86.7%) responded to IVIG, and 63 (13.3%) were non-responders (fig 1). Of the latter, 44 were given IVMP, and 19 were given additional IVIG as second-line therapy. There was no significant difference between the two groups with respect to baseline characteristics (age, sex, white blood cell count, CRP and Harada’s score23) on admission (table 1). In the IVMP group, 34 patients (77%) responded to IVMP and 10 (23%) did not. In the additional-IVIG group, fever was alleviated by additional IVIG in 12 patients (63%) and not in seven (37%). As fever had resolved after day 1 of IVMP treatment in all patients, non-responders were defined as those whose fever recurred after 4–11 days (mean 6.4 days). We administered third-line therapy to those who did not respond to the second-line treatment: eight of the 10 IVMP-resistant patients received IVMP, and two received IVIG. Four of the seven patients who were resistant to additional IVIG received IVMP, and three were treated again with IVIG.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Study protocol and treatment results. All patients with Kawasaki disease (KD) were given a single dose of intravenous immunoglobulin (IVIG) (2 g/kg with aspirin 30 mg/kg/day) as the initial treatment. IVIG resistance was defined as persistence or recurrence of KD-associated fever 24 h after IVIG administration. Steroid pulse therapy (intravenous methylprednisolone (IVMP) 30 mg/kg/day for 3 days) with heparin infusion 10–20 U/kg/h was given to IVIG-resistant patients, unless the family refused the megadose steroid therapy, in which case the patients were given additional IVIG (1–2 g/kg). CAA, coronary artery aneurysm.

Eight of the 411 patients (1.9%) developed CAA, one of whom responded to the initial IVIG treatment. None of the initial IVIG non-responders who responded to the second-line treatment (IVMP or additional IVIG) developed CAA. However, five of the 10 patients (50%) in the IVMP-resistant group developed CAA, three of whom had giant aneurysms, which amounted to 11% of all the cases treated with IVMP. Two of the seven patients (29%) in the additional-IVIG-resistant group developed CAA, none of which was a giant aneurysm, which amounted to 10% of the cases treated with additional IVIG.

Comparison between the IVMP-resistant group and the additional-IVIG-resistant group

When the 10 IVMP-resistant and seven additional-IVIG-resistant patients were compared, CRP and Harada’s score on admission were significantly higher in the IVMP-resistant group than in the additional-IVIG-resistant group (table 1). There was no significant difference in the disease day on which initial IVIG therapy was given, or that of the second-line therapies, between the two groups (table 2). However, the third-line therapy was administered significantly earlier in the additional-IVIG-resistant group than in the IVMP-resistant group. Furthermore, fever was alleviated significantly faster in the additional-IVIG-resistant group than in the IVMP-resistant group.

Adverse effects of IVMP

Of the 48 patients given IVMP (including four IVIG non-responders who subsequently received IVMP), hypertension (SBP⩾130 mmHg) occurred in five (10%), hypothermia (⩽35°C) in three (6%), and sinus bradycardia (<60/min) in three (6%) (table 3). All symptoms were temporary and resolved without any treatment. Rectal bleeding and embolism were not observed. Transient fibular nerve paralysis was seen in a 6-year-old girl 6 days after administration of IVMP on disease day 8 as a rare adverse effect. Causal association with IVMP is uncertain.

DISCUSSION

The effectiveness of IVMP was compared with that of additional IVIG as second-line therapy for KD in this non-randomised study. The decision to use IVMP or additional IVIG was left up to each family. Families were given a full explanation of the protocol along with the beneficial and potential adverse effects of IVMP and that of other treatment options. The families were able to choose whether to use IVMP or additional IVIG on the basis of information provided.

In this study, fever was alleviated rapidly after IVMP administration in all 44 IVIG-resistant patients. Thirty-four of the patients (77%) recovered without recurrence of KD and did not develop CAA. Our findings suggest that early IVMP treatment in IVIG-resistant patients is as effective as treatment with additional IVIG as second-line therapy. However, some patients did not respond to IVMP or additional IVIG, all of whom developed CAAs. Although the result was not statistically significant, five of the 10 IVMP-resistant patients developed CAA, indicating a tendency to develop CAA in these patients. One explanation may be that the condition of the IVMP-resistant patients was more serious than that of those in the additional-IVIG-resistant group, as reflected in the significantly higher CRP and Harada’s score. In fact, there is a possibility that IVMP was not effective in these patients because the symptoms were more severe. Another finding of this study is that, although fever resolved rapidly after IVMP administration in all patients, 10 patients who did not respond to IVMP experienced recurrent fever after a time lapse of 1 week on average, resulting in the subsequent treatments (the third-line therapy) being delayed. In comparison, the additional-IVIG-resistant group received the necessary treatment earlier because of the persistence or immediate recurrence of fever after the additional treatment with IVIG. The risks and choices of subsequent therapies for IVMP-resistant patients need to be carefully evaluated for management of IVMP-resistant cases in the early stages. In addition, increasing the amount, or extending the duration, of prednisolone administration as aftercare to high-risk patients may contribute to the reduction in KD relapse.

Adverse effects requiring treatment were not seen in this study, suggesting that IVMP administration to patients with KD is safer than traditionally believed. However, several adverse effects of IVMP have been reported in the past,24 including a case report of a ruptured coronary aneurism after administration of IVMP.25 Before conducting this study, we also encountered an IVMP-treated KD case in which gastrointestinal tract bleeding that required blood transfusion occurred. Further studies on adverse effects are clearly needed. Some believe that steroids may inhibit the production of cytokines and delay reconstruction of damaged vessels.9 25 26 Therefore, more studies on the long-term prognosis of CAA lesions in patients with KD who are having IVMP treatment should be carried out. We administered heparin with IVMP to prevent potential hypercoagulation and subsequent thrombosis,5 which is one of the major adverse effects of high-dose methylprednisolone. Although this procedure may be unfamiliar in some countries, it is commonly performed in Japan.24

Several issues with respect to steroid therapy for intractable KD still need to be assessed. The dosage and timing of initiation are important considerations for inhibiting inflammation at an early stage. Administering steroids in combination with IVIG during the initial stages of treatment seems to be reasonable if the standard steroid dosage is used. In a multicentre, prospective, randomised, non-blinded trial that compared an IVIG-plus-prednisolone (2 mg/kg/day) regimen with an IVIG-alone regimen as initial therapy for KD, Inoue et al20 reported fewer coronary artery abnormalities for 1 month, along with a shorter duration of fever and a rapid decrease in CRP, in the IVIG-plus-prednisolone regimen. Although the length of prednisolone administration is yet to be determined, they tapered prednisolone over 15 days after CRP normalised. IVMP, which rapidly inhibits inflammation, may be desirable for IVIG-resistant patients. However, no studies have so far shown whether a 3-day or a 1-day course of IVMP is more effective. In addition, the length of administration during aftercare is uncertain. Newburger et al21 reported a multicentre, randomised, double-blind, placebo-controlled trial in which IVIG and aspirin were administered with or without the additional single dose of IVMP (30 mg/kg, without aftercare). The initial period of hospitalisation was shortened, and laboratory findings, such as erythrocyte sedimentation rate and CRP, 1 week after treatment were significantly lower in the group with IVMP. However, there was no significant difference in the incidence of coronary artery abnormalities and the length of hospital stay. Although IVMP suppresses inflammation immediately, the short half-life of IVMP, 2–3 h,27 28 may result in recurrent inflammation in some cases. The important point appears to be the duration of aftercare using prednisolone after IVMP. Aftercare in the present study consisted of tapering of prednisolone over 7 days; it might have been better to continue it for longer. As mentioned above, we need further evaluation, in randomised controlled trials, of the safety of IVMP therapy, not only in the acute phase but also in the long term.

Treatment with IVMP is less expensive than treatment with additional IVIG.9 Therefore, its effectiveness is worth re-evaluation. If we are able to predict IVMP-resistant patients and therefore decide subsequent treatments earlier or reduce the number of IVMP-resistant patients through improvements in aftercare, IVMP should feature more prominently in treatment of KD in the future.