Incomplete (atypical) Kawasaki disease

INTRODUCTION — Kawasaki disease (KD; formerly called mucocutaneous lymph node syndrome) is one of the most common vasculitides of childhood [1]. KD also occurs rarely in adults. It is typically a self-limited condition, with fever and manifestations of acute inflammation lasting for an average of 12 days without therapy [2]. However, complications such as coronary artery (CA) aneurysms may develop and lead to significant morbidity and, rarely, even mortality.

Children suspected of having KD who do not fulfill diagnostic criteria (ie, have fever for five or more days but less than four signs of mucocutaneous inflammation) may have incomplete or atypical KD [1-3]. "Incomplete KD" is the preferred term since these patients do not appear to differ from those with classic KD in any way except that they lack a sufficient number of criteria to fulfill the epidemiologic case definition [2]. Children with incomplete KD are also at risk for cardiovascular sequelae.

The clinical manifestations, diagnosis, and criteria for treatment of incomplete KD are discussed in this review. Unique features in infants and adults are also presented. Other aspects of KD are reviewed separately. (See "Kawasaki disease: Epidemiology and etiology" and "Kawasaki disease: Clinical features and diagnosis" and "Kawasaki disease: Initial treatment and prognosis" and "Kawasaki disease: Complications" and "Cardiovascular sequelae of Kawasaki disease: Clinical features and evaluation".)

EPIDEMIOLOGY — The incidence of incomplete KD is unknown [4,5]. In a retrospective report of 242 Japanese children with KD treated at a single center over a nine-year period, 10 percent of patients were diagnosed with incomplete KD [4]. The incidence appears to be greater in infants younger than six months of age [5,6]. This was illustrated in a retrospective review of 44 children with KD. Five of 11 infants (45 percent) had incomplete disease compared with 4 of 33 (12 percent) older children [5]. KD occurs rarely in adults and presents in an incomplete form more often than in children [7-13].

CLINICAL PRESENTATION OF TYPICAL VERSUS INCOMPLETE KD — Signs and symptoms appear to parallel those in children who fulfill diagnostic criteria for typical disease (table 1) when clinical judgment of reliable observers is used to define incomplete KD. Fever for five or more days is the one common finding. However, fever may be absent or missed in some infants. Patients with incomplete KD have less than four signs of mucocutaneous inflammation. Some infants present only with fever and no other clinical features of KD. (See "Kawasaki disease: Clinical features and diagnosis", section on 'Clinical manifestations' and "Kawasaki disease: Clinical features and diagnosis", section on 'Infants'.)

One report studied 242 patients hospitalized for KD in Japan during a nine-year period and found that 25 (10 percent) ultimately failed to meet diagnostic criteria [4]. Three criteria were met in 17 of the 25 patients (68 percent), and 7 (28 percent) met only two criteria. Only one patient ultimately developed transient dilatation of a coronary artery (CA). In this review, comparison of physical findings among patients with typical and incomplete KD revealed the following:

●Cervical lymphadenopathy was the cardinal manifestation most often absent in children with either complete or incomplete KD. Adenopathy was missing in up to 90 percent of children with incomplete disease versus 40 to 50 percent of those who met criteria for KD.

●Rash was not present in 50 percent of children with incomplete disease compared with 7 to 10 percent of children with typical KD.

●Peripheral extremity changes were absent in approximately 40 percent of incomplete KD cases. In comparison, only 15 percent of those with typical KD failed to develop palmar erythema, dorsal edema, or periungual desquamation.

●Mucous membrane changes were most characteristic of KD and were present in more than 90 percent of children with either typical or incomplete disease.

EVALUATION OF INCOMPLETE KD — The diagnosis of incomplete KD is problematic because the correct diagnosis rests upon clinical judgment and supportive laboratory findings but remains uncertain unless the child develops coronary artery (CA) abnormalities. However, the goal of adding this category of patients was to help providers identify children at risk of developing CA abnormalities who would benefit from treatment for KD regardless of whether they meet diagnostic criteria. As such, the American Heart Association (AHA) and the American Academy of Pediatrics (AAP) have created an algorithm for the evaluation of incomplete KD that uses laboratory studies and echocardiography to aid in the diagnosis of patients whose clinical presentation is consistent with KD but who do not meet diagnostic criteria (table 1) [14]. An adaptation of the algorithm is shown here (algorithm 1). Patients who do not initially meet diagnostic criteria for typical or incomplete KD should be reevaluated if they have persistent fever. At any point in the evaluation, consultation with an expert is recommended if assistance is required or the diagnosis is in question. Involvement of an expert minimizes the tendency to characterize too many children as having incomplete KD when they have other conditions [15]. (See "Kawasaki disease: Clinical features and diagnosis", section on 'Diagnosis' and "Kawasaki disease: Clinical features and diagnosis", section on 'Laboratory findings'.)

Incomplete KD should be suspected and laboratory evaluation performed in patients less than six months of age with unexplained fever for seven or more days, even if they have no clinical findings of KD, and in patients of any age with unexplained fever for five or more days and only two or three clinical criteria (table 1). Other illnesses associated with fever, rash, and similar laboratory abnormalities must be carefully excluded before the label of incomplete KD is applied. Children with a variety of inflammatory and infectious conditions, including adenovirus, atypical measles, sarcoid, polyarteritis nodosa, and systemic juvenile idiopathic arthritis, have been mislabeled as having incomplete KD prior to establishment of the correct diagnosis. The main diagnoses to exclude in adults are drug hypersensitivity reactions and toxic shock syndrome (table 2) [7]. (See "Kawasaki disease: Clinical features and diagnosis", section on 'Differential diagnosis'.)

Laboratory tests — The AHA/AAP-recommended laboratory evaluation includes the following tests:

●Acute-phase reactants (C-reactive protein [CRP] or erythrocyte sedimentation rate [ESR])

●Complete blood count (CBC) with differential white blood cell (WBC) count

●Urinalysis (U/A), preferably clean catch; avoid bladder tap

●Serum alanine aminotransferase level

●Serum albumin

Laboratory findings suggestive of KD include the following:

●Elevated acute-phase reactants (CRP ≥3 mg/dL [≥30 mg/L] or ESR ≥40 mm/hour)

●WBC count ≥15,000/microL

●Normocytic, normochromic anemia for age (table 3)

●Platelet cell count ≥450,000/microL after seven days of illness

●Non-neutrophilic (sterile) pyuria due to urethritis in KD (≥10 WBCs/high-power field)

●Serum alanine aminotransferase level >50 units/L

●Serum albumin ≤3 g/dL

Echocardiography — The presence of cardiac abnormalities detected by echocardiography, although not included in the diagnostic criteria for KD, provides support in ambiguous (incomplete) cases of KD (algorithm 1). Echocardiographic abnormalities suggestive of KD include CA aneurysms (unusual before 10 days of disease), findings consistent with coronary arteritis (eg, CA Z-score >2.5 and lack of tapering of the CAs), decreased left ventricular contractility, mild valvular regurgitation (primarily mitral valve), and pericardial effusion. None of these findings are pathognomonic for KD, but significantly dilated CAs are unusual in other conditions. Mild CA dilatation is sometimes observed in children with other febrile illnesses. A prospective study of echocardiograms in febrile children found that conditions other than KD may be associated with some degree of CA dilatation [16]. In one study, only children with KD had CA Z-scores >2.5, suggesting that this cutoff is a reliable marker of KD [17]. (See "Cardiovascular sequelae of Kawasaki disease: Clinical features and evaluation", section on 'Echocardiography' and "Kawasaki disease: Clinical features and diagnosis", section on 'Diagnosis'.)

The recommendation for echocardiography depends upon the clinical course and initial laboratory findings. Echocardiography is helpful when KD is suspected in a child who does not meet classical criteria for KD but who has the following features:

●Elevation of CRP and/or ESR and less than three supplemental abnormal laboratory findings. KD is less likely if the echocardiogram is normal, particularly if the fever abates and the CA Z-score is <2.0. Repeat echocardiography and consultation with a clinician who has expertise in KD are recommended if fever persists.

●Periungual desquamation after resolution of the fever in someone not meeting epidemiologic criteria for KD (picture 1).

CRITERIA FOR TREATMENT — The American Heart Association (AHA) and the American Academy of Pediatrics (AAP) criteria to diagnose incomplete KD and initiate treatment include clinical features compatible with KD (algorithm 1). Most commonly, such children have fever for at least five days, two or three additional clinical criteria, and elevated C-reactive protein (CRP) and/or erythrocyte sedimentation rate (ESR) (table 1). Intravenous immune globulin (IVIG) therapy should not be withheld when KD is the most likely diagnosis, even without any additional manifestations of disease. (See "Kawasaki disease: Initial treatment and prognosis", section on 'Intravenous immune globulin'.)

Features increasing the likelihood that such a child has incomplete KD that would benefit from IVIG therapy include any of the following:

●Three or more abnormal supplemental laboratory findings (see 'Laboratory tests' above)

●Abnormal echocardiography (see 'Echocardiography' above)

●Fever for seven or more days, even without supportive clinical or laboratory findings, in a child under six months of age who is felt to be more likely to have KD than any alternative diagnosis

This determination is best made by a clinician experienced in the diagnosis and management of KD so as to avoid misdiagnosis and unnecessary treatment. The above recommendations are based upon clinical criteria and reflect the consensus of experts in the field, although they are not based upon firm data and have not been validated. Ultimately, they reflect a bias in favor of treatment in uncertain cases that has developed as a result of the availability of an effective and safe therapy. As such, children with a variety of other febrile conditions may end up receiving IVIG treatment using these guidelines. However, the presence of a concurrent infection does not exclude the diagnosis of KD, as up to 33 percent of children with KD have a concurrent infection [18]. Treatment of KD is discussed in greater detail separately. (See "Kawasaki disease: Initial treatment and prognosis".)

Treating children at risk even though the diagnosis is uncertain should result in fewer children with incomplete KD subsequently developing coronary artery (CA) aneurysms. This was demonstrated in a study that evaluated the performance of the 2004 AHA/AAP criteria in 195 children at four centers, 58 of whom developed CA aneurysms despite failing to meet criteria for KD [19]. Fifty-three of these children had incomplete KD and would have received IVIG at presentation according to the 2004 AHA/AAP guidelines [2]. Two of the remaining five patients would have received IVIG after further monitoring. Overall, application of the 2004 AHA/AAP guidelines would have led to IVIG treatment of 97 percent of the children who developed aneurysms. The corresponding number of children without KD who would have received unnecessary IVIG treatment was not calculated. In any event, as long as the disadvantages of undertreating appear to outweigh the disadvantages of overtreating, some children who have a condition that mimics KD will receive unnecessary IVIG. Thus, continued consideration of alternative diagnoses in "atypical cases" is essential, particularly in children who fail to respond or respond only incompletely to IVIG. (See "Refractory Kawasaki disease".)

PROGNOSIS — Earlier reports suggested a poor prognosis for infants and children with incomplete KD [5,20,21]. One review cited 41 percent mortality, although only children with coronary artery (CA) aneurysms were included in this series [20]. However, subsequent studies have indicated that outcomes in children with incomplete KD are comparable with those of children with complete disease [22]. Nonetheless, children with incomplete features of KD are more likely to be diagnosed later in the course of the illness, and prolonged fever is the strongest predictor of CA aneurysm formation. Thus, children will be best served if KD is at least considered in any case of unexplained fever lasting five or more days. In fact, the goal of the American Heart Association (AHA) and the American Academy of Pediatrics (AAP) guidelines cited above is to remove precisely this obstacle to early diagnosis and treatment. (See 'Evaluation of incomplete KD' above and 'Criteria for treatment' above.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Kawasaki disease".)

SUMMARY

●Rationale for timely diagnosis and treatment – Kawasaki disease (KD) is typically a self-limited condition, but cardiac complications can lead to significant morbidity and mortality. Timely diagnosis and initiation of appropriate therapy minimize cardiac sequelae and improve clinical outcome. (See 'Introduction' above and "Cardiovascular sequelae of Kawasaki disease: Clinical features and evaluation".)

●Diagnostic criteria for KD – KD is a systemic illness characterized by fever, conjunctivitis, mucositis, rash, extremity changes, and cervical lymphadenopathy. These findings are the basis for the diagnostic criteria for KD (table 1). Patients who lack a sufficient number of findings to fulfill the classic criteria may have incomplete KD (algorithm 1). (See 'Clinical presentation of typical versus incomplete KD' above and "Kawasaki disease: Clinical features and diagnosis", section on 'Clinical manifestations'.)

●Laboratory and cardiac studies in ambiguous cases – No laboratory or cardiac studies are included among the diagnostic criteria for typical KD, but certain findings characteristic of KD may support the diagnosis in ambiguous cases (algorithm 1). (See 'Evaluation of incomplete KD' above and "Kawasaki disease: Clinical features and diagnosis", section on 'Laboratory findings' and "Cardiovascular sequelae of Kawasaki disease: Clinical features and evaluation", section on 'Echocardiography'.)

●Criteria to initiate treatment – The criteria to diagnose incomplete KD and initiate treatment in a child in whom an experienced clinician believes that KD is the probable diagnosis include elevated C-reactive protein (CRP) and/or erythrocyte sedimentation rate (ESR) and three or more abnormal supplemental laboratory findings or abnormal echocardiography (algorithm 1). (See 'Criteria for treatment' above.)

●Prognosis – Children with incomplete KD whose diagnosis is delayed are more likely to develop coronary artery (CA) abnormalities. Thus, KD should be in the differential diagnosis of any child with unexplained fever lasting five or more days. (See 'Prognosis' above.)