Table 1 presents definitions of selected terms used in the recommendations, and Table 2 presents the recommendations with their supporting PICO questions and level of evidence. Figure 1 shows a schematic highlighting the key recommendations for the treatment of KD. Conditional recommendations are indicated as such and are generally conditional due to a lack of high-quality evidence (e.g., multiple randomized controlled trials) supporting the recommendation.

Good Practice Statement: IVIG is the standard-of-care therapy for the initial treatment of KD.

IVIG has been established as the standard-of-care treatment for KD for the last 4 decades due to the significant reduction in the rate of coronary artery aneurysms as well as a reduction in the duration of fever and other symptoms associated with its use (11). Although no comparative trials of IVIG and glucocorticoids have been conducted, glucocorticoids alone were considered inadequate to treat KD based on findings from an early case series (12,13). Thus, IVIG use is a mainstay of KD therapy. The mechanisms of the therapeutic effect of IVIG in KD are still under study (14).

Recommendation: For patients with acute KD who are at high risk of IVIG resistance or developing coronary artery aneurysms, use of IVIG with adjunctive glucocorticoids as initial therapy is conditionally recommended over treatment with IVIG alone.

Risk scores have been developed to identify patients at high risk of IVIG resistance in Japanese populations, but these scores had poor sensitivity and specificity in a multiethnic US population (15–17). A recent study of patients in North America identified the following demographic and clinical characteristics as predictors for the development of coronary artery aneurysms at 2–8 weeks: a Z score in the left anterior descending or right coronary artery of ≥2.0, age <6 months, Asian race, and a C-reactive protein level of ≥13 mg/dl (18). In the US, features that suggest an increased likelihood of morbidity in KD include age <6 months or >9 years at diagnosis (19,20). In patients with giant aneurysms (Z score >10) or multiple aneurysms, regression of the aneurysms (via remodeling) at late stages (21) is less likely, and large aneurysms are associated with the greatest morbidity rate (21,22).

Emerging evidence suggests that adding glucocorticoids to IVIG as primary therapy can decrease the risk of developing coronary artery aneurysms, although the strongest evidence is from a Japanese population (16). There is also emerging evidence that the addition of glucocorticoids can decrease the progression of coronary artery aneurysms in patients with coronary aneurysms at the time of diagnosis (23,24).

Thus, the use of glucocorticoids with IVIG is a treatment option in patients who are at high risk of coronary artery aneurysms. For this recommendation, the Voting Panel defined high-risk features as a Z score of ≥2.5 (25–27) for the left anterior descending or right coronary artery at the time of the initial echocardiography and age <6 months. This definition uses the Z score of 2.5 instead of 2.0, since a score of 2.5 is defined as representing a true aneurysm. We also do not include patients >9 years of age, as only a portion of these patients are believed to be “high-risk.” The Voting Panel acknowledges that further investigation clarifying the features of “high-risk” KD in a US population is needed. Optimal dosing and duration of glucocorticoids are still to be determined and will require further study in a US population, but a typical dosage would be prednisone starting at 2 mg/kg/day (maximum 60 mg/day), with the dose tapered over 15 days. The use of IVIG alone is conditionally recommended when the treating physician is unsure whether a patient is at high risk of developing coronary artery aneurysms, or when additional glucocorticoid exposure may be detrimental to the patient. The use of glucocorticoids is generally not recommended in patients who do not have features suggestive of high risk.

Recommendation: For patients with acute KD who are at high risk of IVIG resistance or developing coronary artery aneurysms, using IVIG with other nonglucocorticoid immunomodulatory immunosuppressive agents as initial therapy is conditionally recommended over treatment with IVIG alone.

For patients with features of high-risk disease, adjunctive therapy with a nonglucocorticoid immunomodulatory agent, such as infliximab, anakinra, or cyclosporine, may improve outcomes in acute disease and may also improve cardiac outcomes (21,23,28,29). Thus, adjunctive therapy with a nonglucocorticoid immunomodulatory agent can be considered in patients with KD who are at high risk of not responding to IVIG. For any patient considered to have features of high-risk disease, a rheumatologist or other clinician with expertise in KD should be consulted before adding an adjunctive agent (glucocorticoids or nonglucocorticoid immunomodulatory agents) to ensure correct diagnosis and appropriate utilization of adjunctive therapies. The Voting Panel also advocated for a stepwise therapeutic algorithm, in which either a glucocorticoid or nonglucocorticoid immunosuppressive agent along with IVIG is initially prescribed in high-risk patients. Currently, there is more evidence supporting the use of glucocorticoids than nonglucocorticoid immunosuppressive agents in this patient population, and further study is warranted to compare the efficacy of glucocorticoid therapy to nonglucocorticoid therapy in this population. There may be specific situations in which glucocorticoids are contraindicated, but the patient exhibits high-risk features that would indicate the need for adjunctive nonglucocorticoid immunomodulatory therapy. Once again, use of IVIG alone is appropriate if the treating physician is unsure whether a patient has high-risk features, as described in the previous recommendation, or if using a nonglucocorticoid immunomodulatory agent presents a greater risk of an adverse effect to the patient.

Recommendation: For patients with incomplete KD, prompt treatment with IVIG at the time of diagnosis is strongly recommended over delaying treatment until day 10 or later.

Incomplete KD is defined as a condition of suspected KD that lacks the sufficient number of features to meet criteria for KD (3,30) (Table 1). Patients with incomplete KD should be evaluated using the algorithm for incomplete KD as outlined in the AHA guidelines for KD (3,30). The performance of this algorithm for identifying patients with incomplete KD has been validated (30). The practice of completing diagnostic testing and initiating treatment by day 10 of fever is not based on mechanisms or studies of disease pathogenesis or treatment outcomes. Instead, this “treatment deadline” was chosen as a research end point in the original studies in KD (31). It is recognized that prompt treatment prevents adverse outcomes, and patients should receive treatment as soon as the diagnosis of incomplete KD is made, rather than waiting until day 10 to see if they meet criteria for complete KD. Although delayed IVIG treatment after 10 days of fever is associated with worse outcomes (32), IVIG treatment should still be administered to patients even if they present after >10 days of fever. Thus, patients who meet the criteria for incomplete KD according to the AHA guidelines should be treated promptly to avoid adverse consequences, since the risk of developing coronary artery dilatation increases each day without treatment. This is a strong recommendation because this intervention is standard of care, and the consequence of delaying treatment may lead to significant coronary artery aneurysms or rupture, which are severe adverse outcomes. In addition, resolution of fever prior to day 10 is not an indication to withhold treatment in patients meeting criteria for incomplete or complete KD. In patients with characteristically elevated levels of acute-phase reactants, treatment is recommended, as these patients remain at high risk of developing adverse outcomes (3,33).

Recommendation: For patients with acute KD and suspected or diagnosed macrophage activation syndrome (MAS), treatment with IVIG for KD and additional agents to treat MAS is strongly recommended.

MAS, a form of secondary hemophagocytic lymphohistiocytosis (HLH), is a potentially underrecognized complication of KD (34,35). Formal diagnostic criteria for MAS in the setting of KD have not been developed. However, drawing on experience with other secondary HLH presentations, MAS may be suspected in KD patients with persistent fever, splenomegaly, elevated ferritin levels, and thrombocytopenia (35). Inadequate treatment of either KD or MAS could result in severe consequences. These include large coronary aneurysms or coronary artery stenosis, leading to death via cardiac infarct or coronary rupture in KD, or death due to multiorgan dysfunction in MAS (36).

Thus, to ensure appropriate therapy, each disease entity should be considered separately with appropriate targeted therapy. KD should be treated with IVIG as the first-line therapy, and MAS should also be treated with appropriate agents for targeting cytokine storm or underlying triggers. Anakinra and glucocorticoids are preferred for treatment in these patients over a primary HLH–directed treatment protocol with cytotoxic agents. Although no head-to-head comparison trials have been published, primary HLH–directed therapy, which has considerably more associated toxicity, may not be warranted in patients without an underlying genetic predisposition to HLH. This is a strong recommendation because the consequence of not appropriately treating KD or MAS may be associated with increased mortality.

Recommendation: For patients with acute KD and persistent fevers after initial treatment with IVIG, a second course of IVIG is conditionally recommended over the use of glucocorticoids.

Findings from 6 studies provide indirect evidence for this recommendation, though no direct comparative studies were available (37–42). For KD patients with persistent fevers after the initial course of IVIG, studies suggest that there is no difference in coronary artery outcomes between repeating the course of IVIG versus a single dose of pulse glucocorticoids (i.e., 30 mg/kg with a maximum of 1 gm) (43). A meta-analysis of 12 studies involving 372 patients whose disease was resistant to treatment with a single course of IVIG failed to demonstrate significant differences in coronary artery outcomes between patients treated with a second dose of IVIG versus glucocorticoids versus infliximab (44). Although current evidence does not clearly indicate the superiority of a second course of IVIG over glucocorticoids, a second course of IVIG in patients who have persistent fever for >36 hours after the first dose is conditionally recommended, as it is the current standard of care. However, as a conditional recommendation, glucocorticoids are a reasonable alternative (e.g., starting at 2 mg/kg/day and tapering over 15 days or a single dose of 20–30 mg/kg). Repeated doses of IVIG may put patients at risk of hemolytic anemia, and the results of ongoing studies may alter this recommendation in the future (45–47). In patients with risk factors for hemolytic anemia with IVIG, such as non–type O blood groups, alternative therapies, such as glucocorticoids or nonglucocorticoid immunomodulatory therapy, should be considered (45,48). Combination therapy (i.e., multiple anticytokine agents) is not recommended for routine care and generally should only be considered in patients with extremely severe disease.

Ungraded Position Statement: For patients with acute KD and persistent fevers after repeated treatment with IVIG, either nonglucocorticoid immunosuppressive therapy or glucocorticoids may be used.

In patients with KD and persistent fevers after 2 doses of IVIG, use of another agent is indicated. Findings from studies of infliximab and cyclosporine for the treatment of refractory KD suggest some potential benefit of these agents (28,29). Head-tohead comparisons between these nonglucocorticoid immunosuppressive agents and glucocorticoids have not been performed. There is no clinical evidence to suggest superiority of either glucocorticoids or nonglucocorticoid immunomodulatory agents, and the Voting Panel was evenly split as to which to recommend. Thus, the Voting Panel believes that use of either class of agents would be appropriate based on the specific clinical scenario. Combination therapy (glucocorticoid with a nonglucocorticoid immunosuppressive agent) can be considered in severe cases, such as rapidly expanding aneurysms or imminently lifethreatening disease.

Recommendation: For patients with acute KD, using aspirin is strongly recommended over no aspirin.

The use of aspirin in patients with KD to reduce inflammation and prevent thrombosis through its antiplatelet effect is considered standard of care. However, the optimal dosage is unclear.

Historically, high dosages of aspirin (80–100 mg/kg/day) were used during the acute phase for antiinflammatory effects, but there is no evidence of benefit with high- versus low-dose aspirin (3–5 mg/kg/day) when considering coronary vascular damage (49–51). The AHA guidelines should be consulted for recommendations regarding anticoagulation in patients with larger aneurysms (3). This is a strong recommendation because aspirin is recognized to be a low-cost intervention with limited toxicity at antiplatelet dosing, and also because there are significant consequences of not inhibiting platelet activity, such as coronary artery thrombosis.

Recommendation: For patients with acute KD with subsequent resolution of fevers, continued daily monitoring is strongly recommended over not monitoring for fevers.

Patients with KD may experience recurrence of disease or treatment-refractory disease heralded by returning fever and other symptoms. In addition, the duration of fever is a predictor of coronary artery aneurysms (52). Thus, patients should be monitored daily for fevers for 1–2 weeks after discharge, with fever defined as an oral temperature in older children and a rectal temperature in infants of >38.0C or an axillary temperature of >37.5C. Parents or guardians should be instructed by the discharging physician on how to take a temperature and told to contact the physician should fever recur. Daily fever monitoring is strongly recommended because it is low cost, without harms, and may identify recurring KD.

Recommendation: For patients with acute KD who have arthritis that persists after IVIG treatment and who do not have coronary artery aneurysms, using nonsteroidal antiinflammatory drugs (NSAIDs) to treat arthritis is conditionally recommended over not using NSAIDS.

Patients with acute and subacute KD can develop arthritis, but treatment with IVIG decreases the incidence of arthritis in this setting from 30% to ~2–12% (53). Arthritis in KD is usually shortlived, lasting between 7 and 21 days (53–55). In general, arthritis can be treated with NSAIDs, but NSAIDs directly inhibit the ability of aspirin to acetylate cyclooxygenase 1, thus decreasing aspirin’s protective effects (56). The impact of NSAID use can be partially mitigated by administering aspirin prior to NSAIDs, but combination therapy is usually not recommended given the potential increased toxicity (56).

Patients without coronary artery aneurysms generally do not require long-term treatment with aspirin. The optimal duration of treatment with aspirin in these patients is not well defined but in practice is generally 6–8 weeks. Patients with coronary artery aneurysms require long-term and potentially indefinite aspirin use (3). For patients without coronary artery aneurysms who do not require long-term use of aspirin and who have arthritis requiring additional therapy, aspirin can be temporarily suspended and a short course (3–4 weeks) of NSAIDs can be used as needed.

Conversely, if indefinite or long-term aspirin use is needed due to coronary artery aneurysms, then either acetaminophen, a short course of glucocorticoids, or nonsystemic NSAID pain management options (e.g., a topical NSAID) can be used. A pediatric hematologist or cardiologist should be consulted to consider an alternative anticoagulant (e.g., clopidogrel) if prolonged use of systemic NSAIDs is required (i.e., >3 weeks), especially in a patient with coronary artery aneurysms.

Recommendation: For children with suspected incomplete KD and fever, obtaining an echocardiogram with coronary artery measurements without delay is strongly recommended over not obtaining an echocardiogram.

Patients with incomplete KD have at least the same or an increased risk of coronary artery lesions compared to patients with classic KD (57). Therefore, in patients in whom incomplete KD is suspected, a diagnosis should be confirmed as soon as possible so that features associated with high risk of coronary artery aneurysms can be assessed, treatment can be initiated without delay, and adverse outcomes can be prevented. Echocardiographic assessment of the absolute dimensions of the coronary artery and body surface area–adjusted Z scores can help establish a diagnosis of incomplete KD and should be obtained promptly when this diagnosis is suspected. The AHA guidelines for KD management provide an algorithm for the evaluation and treatment of suspected incomplete KD (3), which has been validated (30). Echocardiography is strongly recommended in this scenario because it has minimal potential harms and may prevent adverse outcomes by prompting the treatment of incomplete KD.

Recommendation: For children with unexplained shock physiology, obtaining an echocardiogram with coronary artery measurements is strongly recommended.

Unexplained shock can be due to Kawasaki shock syndrome, a recognized manifestation of KD (58). Echocardiograms are often obtained in children with unexplained shock, especially in those with the prodrome of prolonged fever, to evaluate cardiac function and help identify the potential etiology. We recommend that the coronary arteries be included in the echocardiographic examination to evaluate for KD as a potential etiology of shock. This is strongly recommended because echocardiography has minimal potential harms and may prevent catastrophic outcomes by facilitating the diagnosis and prompt treatment of KD.

Recommendation: For children with unexplained MAS, obtaining an echocardiogram with coronary artery measurements is strongly recommended.

As noted above, KD is one of the clinical inflammatory entities that can predispose patients to the development of MAS (59). For patients with unexplained MAS, KD should be considered as a potential underlying etiology. Echocardiography with inclusion of the coronary arteries is one method of identifying KD as a potential treatable underlying condition. This is strongly recommended because echocardiography has minimal potential harms and may prevent adverse outcomes by facilitating the diagnosis and prompt treatment of KD.