Efficacy of the Delayed Use of Low-dose Aspirin in Intravenous Immunoglobulin Therapy for Acute-phase Kawasaki Disease

The mainstay of current standard therapy for acute-phase Kawasaki disease (KD) is intravenous immunoglobulin (IVIG) therapy at 2 g/kg. However, the efficacy of combining mediumor high-dose aspirin with IVIG therapy at 2 g/kg has not been fully investigated. Some studies suggested that aspirin may inhibit coronary artery lesion (CAL) prevention in IVIG therapy and that the delayed use of aspirin in IVIG therapy may be beneficial for the suppression of CALs and prevention of coronary artery stenosis in patients with KD. The efficacy of the delayed use of lowdose aspirin in IVIG therapy for acute-phase KD remains unclear. Therefore, this retrospective study aimed to assess the efficacy of the delayed use of low-dose aspirin, when combined with IVIG therapy for acute-phase KD. Data were obtained from 193 KD patients who underwent acute-phase treatment from January 2009 to October 2020 and IVIG therapy at 2 g/kg with the delayed use of aspirin/flurbiprofen. The patients were divided into three groups: (1) low-dose group, in which 40 patients received low-dose aspirin (5 mg/kg/day); (2) medium-dose group, in which 90 patients received medium-dose aspirin (30 mg/kg/day); and (3) flurbiprofen group, in which 63 patients received flurbiprofen (3–5 mg/kg/day). KD patients with liver damage or those present during influenza season underwent flurbiprofen therapy between January 2009 and November 2017. All patients except one received low-dose aspirin after December 2017. The serum albumin level (median 3.40 vs. 3.30 g/dL, P = 0.026) and Egami score (median 1.0 vs. 2.0, P < 0.001) before the initial treatment were significantly different between the medium-dose group and the flurbiprofen group. The rates of initial IVIG therapy resistance (25.0% vs. 18.9% vs. 25.4%, P = 0.790), rescue therapy (17.5% vs. 8.9% vs. 17.5%, P = 0.721), and CALs (5.0% vs. 0.0% vs. 4.8%, P = 0.713) were similar among the low-dose, medium-dose, and flurbiprofen groups. Overall, the efficacy of the delayed use of low-dose aspirin was similar to that of the delayed use of medium-dose aspirin/flurbiprofen in IVIG therapy for acute-phase KD.


I. INTRODUCTION
Kawasaki disease (KD) is a form of acute febrile systemic vasculitis that primarily affects children younger than 5 years [1]. Coronary artery lesions (CALs) are a severe complication of KD [2].
The mainstay of the current standard therapy for acutephase KD is intravenous immunoglobulin (IVIG) therapy at 2 g/kg with the concomitant use of medium-/high-dose aspirin. Moreover, evidence has established its efficacy in CAL suppression [3]. However, the efficacy of combining medium-or high-dose aspirin with IVIG therapy at 2 g/kg has not been fully investigated [4]. A randomized controlled trial on the effectiveness of IVIG monotherapy versus IVIG therapy combined with high-dose aspirin in the acute KD stage is ongoing [5]. Some studies suggested that aspirin may inhibit CAL prevention in IVIG therapy and that the delayed use of aspirin (DUA) for IVIG therapy may be beneficial for the suppression of CALs and prevention of coronary artery stenosis in patients with KD [6]- [9]. Furthermore, a recent study showed favorable medium-term outcomes of CALs in KD patients who underwent IVIG therapy at 2 g/kg with DUA [10].
Moreover, the optimal dose of aspirin for the acute-phase treatment of KD has not been determined yet [11], [12]. Some studies showed that the use of low-dose aspirin (3-5 mg/kg/day) may be as effective as the use of medium-/high-dose aspirin (≥ 30 mg/kg/day) in the initial treatment of KD [13], [14]. The efficacy of the delayed use of lowdose aspirin in IVIG therapy for acute-phase KD remains unclear. Therefore, this retrospective study aimed to assess the efficacy of the delayed use of low-dose aspirin when combined with IVIG therapy for acute-phase KD. Use of Low-dose Aspirin in  Intravenous Immunoglobulin Therapy for Acute-phase  Kawasaki Disease   Toshimasa Nakada II. PARTICIPANTS AND METHODS Our institutional ethics committee approved the study protocol and waived the requirement of patient consent because of the retrospective nature of the study.

Efficacy of the Delayed
Data were obtained from 193 KD patients who underwent acute-phase treatment from January 2009 to October 2020 and IVIG therapy at 2 g/kg with the delayed use of aspirin/flurbiprofen. The patients were divided into three groups: (1) low-dose group, in which 40 patients received low-dose aspirin (5 mg/kg/day); (2) medium-dose group, in which 90 patients received medium-dose aspirin (30 mg/kg/day); and (3) flurbiprofen group, in which 63 patients received flurbiprofen (3-5 mg/kg/day).
A total of six patients with complications including CAL before the initial therapy were excluded.
KD was diagnosed based on the criteria (Japanese, fifth edition) mentioned in the diagnostic guidelines for KD until August 2019 and on the revised criteria (Japanese, sixth edition) from September 2019 [2], [15]. IVIG resistance was defined as a fever that persisted or reappeared 24 h after the first-line treatment [3]. Presentations of KD were considered as relapses when a second episode appeared within 2 months of the first one. [16], [17]. The Egami score, a risk score for predicting IVIG resistance based on clinical findings such as age, days of illness, platelet count, alanine aminotransferase level, and C-reactive protein level, was evaluated before the initial IVIG therapy [18]. The outcomes included the rates of IVIG resistance, rescue therapy, and CALs.

A. Initial Therapy
During the study period, a single IVIG infusion at 2 g/kg/dose was given as initial therapy starting on Day 5 of the illness, whenever possible [10]. Patients without inflammation or complications upon presentation underwent acute-phase therapy without IVIG [10]. Between January 2009 and November 2017, anti-inflammatory drugs (aspirin or flurbiprofen) were initiated within 24 h after the initial IVIG infusion ended [6]. Aspirin and flurbiprofen were started at 30 and 3−5 mg/kg/day, respectively, and decreased to 5−10 and 3 mg/kg/day, respectively, once the patient was afebrile [6]. Each treating physician chose aspirin or flurbiprofen after considering the patient's liver function and the risk of Reye syndrome during influenza season between January 2009 and November 2017. After December 2017, low-dose aspirin (5 mg/kg/day) was begun on Days 8−10 of the illness after the completion of IVIG infusion, including the second course of therapy [10].

B. Rescue Therapy
The decision to use rescue therapies in resistant patients was made 48 to 72 h after the completion of the initial IVIG infusion. Physicians made this decision using comprehensive clinical parameters, including body temperature, major KD symptoms, general condition, and laboratory data [10]. Course 2 of therapy comprised rescue IVIG infusion at 2 g/kg/dose, and course 3 was an ulinastatin infusion, third course of IVIG therapy, or plasma exchange [10].
C. Diagnosis of CAL CALs were diagnosed using echocardiography based on the Japanese criteria reported by Kobayashi et al [19]. CAL was diagnosed if any examination showed an internal lumen diameter of ≥ 3 mm in a patient younger than 5 years or a diameter of ≥ 4 mm in a patient older than 5 years if the internal diameter of a segment was at least 1.5 times larger than that of an adjacent segment or if the lumen appeared irregular. A transient CAL was defined as the disappearance of a CAL within 30 days of the illness.
D. Statistical Analysis Statistical analyses were conducted using Stat Flex Version 6 for Windows (Artech Co., Ltd., Osaka, Japan). Chi-square test, Fisher exact test, Mann-Whitney U test, and Kruskal-Wallis test were used accordingly, with sample size considerations. P < 0.05 was considered statistically significant.

III. RESULTS
The rate of the era of onset after 2018 between the lowdose and medium-dose groups, and that between the lowdose group and the flurbiprofen group were significantly different (Tables I and II). The Egami score of the flurbiprofen group was significantly higher than that of the low-dose and medium-dose groups (Tables II and III). Furthermore, the Egami scores of the low-dose, mediumdose, and flurbiprofen groups were significantly different among the three groups (Table IV). The serum albumin level of the flurbiprofen group was significantly lower than that of the medium-dose group (Table III). The rates of the initial IVIG therapy resistance, rescue therapy, and CALs were similar among the low-dose, medium-dose, and flurbiprofen groups (Tables I, II, III, and IV).

IV. DISCUSSION
The main finding of this study was that the rates of initial IVIG therapy resistance, rescue therapy, and CALs were similar among the low-dose, medium-dose, and flurbiprofen groups (Tables I, II, III, and IV). This finding indicated that the efficacy of the delayed use of low-dose aspirin was similar to that of the delayed use of medium-dose aspirin/flurbiprofen in IVIG therapy for acute-phase KD.
Current evidence based on a meta-analysis demonstrated that treatment combined with low-dose aspirin compared with high-dose aspirin (≥ 30 mg/kg/day) for the acute-phase treatment of KD showed no significant difference in the incidence of CAL, the risk of IVIG therapy resistance, or the length of fever or hospital stay [14]. According to this metaanalysis, the risk ratio of the low-dose aspirin group regarding CAL development was less than 1.0 in five of six studies compared with that of the high-dose aspirin group [13], [14], [20]- [24]. Moreover, one study showed that the adjusted risk difference was statistically significant, favoring the low-dose aspirin group (the low-dose vs. high-dose group: −4.0% vs. −1.9%, P value = 0.024) for medium or larger CAL caused by KD [13]. This was the result when combining the concomitant use of aspirin with IVIG therapy [13].
Two studies from Korea reported the outcomes of CAL in patients who underwent IVIG therapy at 2 g/kg with delayed use of low-dose aspirin for the acute-phase treatment of KD [8], [25]. One study showed that no patients had CALs with an internal diameter of ≥ 6 mm, which indicates a risk of stenotic lesions [8]. Another study showed that the prevalence of CALs in the patients who underwent IVIG therapy at 2 g/kg with the delayed use of low-dose aspirin was similar to the prevalence in those who underwent IVIG therapy at 2 g/kg with the concomitant use of high-dose aspirin (3.9% vs. 7.8%, P = 0.514) [25]. Furthermore, a study using the data of the nationwide survey of KD conducted in Korea showed that the prevalence of CALs in 509 subjects who received low-dose aspirin was significantly lower than the prevalence in 7947 subjects who received medium-/high-dose aspirin during the acute febrile phase [21]. A recent meta-analysis showed that prescribing low-dose or no aspirin in the acute-phase KD might be associated with the decreased incidence of CAL [26].
Regarding the favorable outcomes of CAL in the lowdose group, those findings are consistent with those of the present study (Tables I, II, and IV). In our previous study, the findings suggested that the concomitant use of mediumdose aspirin/flurbiprofen may inhibit CAL prevention in IVIG therapy, whereas the delayed use of those antiinflammatory drugs in initial IVIG therapy may be beneficial for the suppression of CALs in patients with KD [6]. Moreover, recent studies demonstrated the usefulness of IVIG therapy at 2 g/kg with DUA for the prevention of coronary artery stenosis and the favorable medium-term outcomes in patients with KD [9], [10].
A recent report regarding platelet activation characteristics and the effect of antiplatelet therapy for patients with acute KD using a newly automated flow chamber system to evaluate platelet aggregate formation under shear stress condition showed that the patients developed early and unstable platelet aggregates regardless of aspirin use at different dosages and the use of an alternative drug (flurbiprofen) [27], [28]. Furthermore, the characteristics of platelet thrombus formation under a high shear condition in children with KD were not specific compared with those in febrile child controls [27]. Those findings support the effectiveness of KD treatment without aspirin or with low-dose aspirin [27].
Although high-dose aspirin is generally well-tolerated in children, reports of adverse effects, including gastrointestinal bleeding and Reye syndrome, have been documented [29], [30]. Meanwhile, low-dose aspirin has not been associated with Reye syndrome [31], is administered once daily, and may be better tolerated. Given these potential advantages, an investigation into the outcomes of children treated with low-dose aspirin is important [32]. Some studies suggested that high-dose aspirin does not reduce CALs, with a surprising possibility that aspirin might be associated with a higher incidence of CALs [6], [7], [13], [20]- [26], [33]. A recent meta-analysis recommended a low-dose aspirin plus IVIG as the first-line therapy in the initial treatment of KD [12].
Previous studies found that treatment with IVIG alone without aspirin in the acute stage of KD did not affect the response rate of IVIG treatment [34], [35]. Furthermore, the recent meta-analysis indicated that the rates of nonresponders between the low-dose and high-dose aspirin groups were not statistically different [14]. This finding was consistent with the results of the present study: the rates of non-responders were similar among the three groups (Tables  I, II, III, and IV). The decision for rescue therapy was comprehensively made according to clinical parameters, including body temperature, major KD symptoms, general condition, and laboratory data, after initial therapy in the present study [10], [36]. The similar rates of non-responders and rescue therapy among the three groups in this study may also be due to this strategy.
KD relapse is a risk factor for CAL development [17], [37]. A study showed that the rates of KD relapse and CAL development were similar between the patients who were treated with low-dose aspirin and those who were treated with high-dose aspirin [32]. Those findings were consistent with those of the present study: the rates of KD relapse and CAL development were similar among the three groups (Tables I, II, III, and IV).
The use of high-dose aspirin has a risk of salicylateinduced hepatitis and Reye syndrome in acute-phase of KD [11], [29], [38]. Flurbiprofen has been used in KD patients with elevated liver enzyme levels or during an influenza epidemic in Japan, and Reye syndrome was not considered as one of the adverse effects of flurbiprofen [27], [39]. A study on platelet aggregate formation under shear stress conditions supports the treatment for acute-phase KD using flurbiprofen instead of medium-dose aspirin as antiplatelet therapy as well as low-dose aspirin [27]. In fact, the efficacy of flurbiprofen was similar to those of low-/medium-dose aspirin, and no patients had salicylate-induced hepatitis or Reye syndrome in the present study.
An increase in liver enzyme levels is a risk factor for IVIG therapy resistance and CALs [18], [19]. In the present study, KD patients with liver damage or those present during influenza season received flurbiprofen therapy between January 2009 and November 2017. Low serum albumin level indicates severe arteritis in the acute-phase of KD, and is considered a risk factor for IVIG resistance and CAL development [40], [41]. The higher Egami score and lower serum albumin levels before initial therapy in the flurbiprofen group suggested more severe KD patients in this group (Tables II, III, and IV). The lower rate of initial IVIG therapy resistance, rescue therapy, and CALs in the medium-dose group may be due to the different severity of the disease before initial therapy because patients in the medium-dose group and those in the flurbiprofen group received IVIG therapy in the same period (between January 2009 and November 2017).
The limitations of this study include the small sample size and the retrospective study design.
In conclusion, the rates of initial IVIG therapy resistance, rescue therapy, and CALs were similar among the low-dose, medium-dose, and flurbiprofen groups. Furthermore, the efficacy of the delayed use of low-dose aspirin may be similar to that of the delayed use of medium-dose aspirin/flurbiprofen in IVIG therapy at 2 g/kg for acutephase KD.      V. ACKNOWLEDGMENT I would like to thank all of those who were involved in the medical management of the patients included in this study, and Enago (www.enago.jp) for the English language review.