Kawasaki disease (KD) is a self-limited vasculitis of unknown etiology typically characterized by persistent fever for at least 5 days, polymorphous rash, nonpurulent conjunctivitis, inflammatory changes in the lips or oral cavity, edema of the hands and feet and/or redness of the palms and soles and cervical lymphadenopathy of > 1.5 cm. Later in the convalescent stage periungual desquamation is pathognomonic. To confirm the diagnosis of KD, 5 of the 6 criteria must be documented or, if fewer criteria are present, coronary artery aneurysms must be demonstrated. Retrospective evaluation of patients diagnosed with KD revealed that younger children, especially infants younger than 1 year of age, present more often with an incomplete form not meeting all of the above criteria. 1,2 Unusual symptoms that could be attributed to other febrile disorders can be misleading and delay the correct diagnosis and adequate treatment. In the cases presented here, the pulmonary symptoms were the most prominent features of the illness. The patients were initially diagnosed as having pneumonia but were later identified to have KD.
An 8-week-old girl was admitted to the hospital because of persistent fever in the previous 5 days, increasing cough and tachypnea. Physical examination revealed a body temperature of 39°C, respiration rate 70/min, coarse breath sounds, no abnormalities on cardiac auscultation and abdominal examination and no conjunctival injection, rash or lymphadenopathy.
Laboratory studies showed an elevated C-reactive protein (CRP) (36 mg/dL), leukocytosis (22,300/μL), mild anemia (9.5 g/dL), 753,000 thrombocytes/μL and elevated liver enzymes [alanine aminotransferase, 54 units/L (normal, <21); γ-glutamyl transpeptidase, 137 units/L (normal, <13)]. Microbial cultures (blood, pharyngeal swab) were negative for pathogenic specimens; in serologic tests no IgM antibodies could be found for Mycoplasma, Chlamydia, cytomegalovirus, Epstein-Barr virus, influenza, parainfluenza, adenovirus and enterovirus.
Chest radiograph showed interstitial pneumonia and antibiotic therapy was started with flucloxacillin and cefotaxime for 10 days and erythromycin for 14 days (Fig. 1). The patient’s clinical condition and radiologic examination improved, but the platelet count increased to 1,300,000 platelets/μL on day 13. Acetylsalicylic acid (ASA) was started with 2 mg/kg. The infant was afebrile on day 15 (body temperature constantly <37.5°C) and was discharged from the hospital on day 24 in good clinical recovery; however, the CRP and leukocyte counts remained elevated (11.4 mg/dL and 21,000/μL, respectively).
The patient was readmitted to the hospital on day 36 with irritability, vomiting and apathy, but no fever; laboratory tests again showed an increased CRP (19.2 mg/dL), with erythrocyte sedimentation rate 82 mm/h, marked leukocytosis (32,200/μL), thrombocytosis (1,200,000/μL) and anemia (8.6 g/dL). Physical examination revealed tachypnea (60/min), tachycardia (170/min), gallop rhythm, elevated blood pressure (mean arterial pressure, 90 mm Hg) and hepatomegaly. The electrocardiogram showed S-T elevations in V4–V6, compatible with an acute anteroseptal infarction, also reflected by an elevated creatine kinase (310 units/L; normal, <122 units/L) with a creatine kinasemuscle-brain isoenzyme portion >6% (29 units/L). Echocardiography showed dilation of the left ventricle with diminished contractility and dyskinesia of the septum. Additionally, a dilated aortic root and multiple aneurysms of the left coronary artery (maximum diameter, 7 mm) were observed. The febrile illness was reclassified as atypical Kawasaki disease, with the anteroseptal infarction probably due to thrombosis of the aneurysms. Because the infant again had elevated CRP and leukocyte count, a relapse of KD was diagnosed. She received systemic thrombolytic treatment with recombinant tissue plasminogen activator for 4 days and standard therapy for KD with intravenous immunoglobulin (2 g/kg) and ASA (100 mg/kg). Coronary angiography on day 75 showed normal coronary arteries and a slightly widened ascending aorta; echocardiographic findings at that time revealed a hypokinetic area at the left ventricular apex and a mild aortic valve insufficiency with good overall heart function.
A 14-month-old girl developed a rash and fever up to 39.8°C and was admitted to the hospital on day 4 of illness because of clinical deterioration with irritability. Physical examination revealed tachypnea with ronchus, a maculopapular exanthema over the trunk, cervical lymphadenopathy, a markedly reddened pharynx, rhinitis, cough and bilateral conjunctival injection. Abdominal examination and cardiac auscultation were unremarkable. Laboratory studies showed a moderately elevated CRP (4.59 mg/dL) and normal white blood cell count (9570/μL), 351,000 platelets/μl, mild anemia (10.8 g/dL), slightly elevated liver enzymes [aspartate aminotransferase, 33 units/L (normal, <23); alanine aminotransferase, 48 units/L (normal, <21); γ-glutamyl transpeptidase, 16 units/L (normal, <13)]. Two pharyngeal Strep-A tests were negative. Initial echocardiography showed no abnormalities. The chest roentgenogram was compatible with bronchopneumonia, and antibiotic therapy with cefuroxime and erythromycin was started. The infant’s condition improved gradually, but complete defervescence occurred only on the ninth day of illness. This was accompanied by a fall of CRP (1.14 mg/dL) while the thrombocyte count rose to 456,000/μL. The child was discharged from the hospital but readmitted 12 days after initial symptoms with periungual desquamation typical of Kawasaki disease. At that time, the child still had a reddened pharynx and raspy breathing, but no fever; CRP was 0.89 mg/dL, platelet count had increased to 523,000/μL. Intravenous immunoglobulin (2 g/kg) was administered in combination with 25 mg of ASA per day. An echocardiogram on day 13 was normal, as were follow-up echocardiograms 1, 2, 4 and 6 months after the acute stage of the disease.
Retrospectively the girl fulfilled all criteria of typical Kawasaki disease in the course of the illness; it is remarkable that the pulmonary symptoms were most prominent and the CRP level was only moderately elevated, with a normal white blood cell count.
Our cases indicate that prominent pulmonary signs and symptoms may be misinterpreted as pneumonia in children with KD. Only recently, 2 reports about pulmonary complications associated with KD have been published: Voynow et al 3 describe a 6-year-old girl who developed dyspnea with lowered O2 saturation, bibasilar interstitial lung disease and pleural effusions on radiography during the course of KD. Chronic interstitial pneumonitis with infiltrates of mononuclear inflammatory cells and focal organizing pneumonia without evidence of vasculitis was established by biopsy. Uziel et al 4 report 2 children diagnosed with “unresolving pneumonia” caused by a febrile illness with lobar consolidation shown by chest radiograph which subsequently were recognized as KD. In classic KD with all clinical criteria fulfilled, pneumonia seems to be rare or underestimated as an additional significant clinical finding. 5 However, the prevalence of respiratory signs such as tachypnea, dyspnea, cough or coryza in the context of KD is described to be higher in infants younger than 6 months of age (6 of 8 patients) 1 and children with atypical KD (8 of 17 patients). 6
In a retrospective study, Umezawa et al 7 found pulmonary involvement on chest radiographs in Kawasaki disease in 14.7% of 129 KD patients. This feature was correlated with a more severe course of the disease, measured by clinical and laboratory findings. In pathologic examinations of pulmonary tissue in acute fatal KD, inflammatory infiltrates were found in the lung parenchyma as well as peribronchially located in >80% of cases. 8 Recently Freeman et al 9 have reported on 3 children with KD who developed pulmonary nodules in addition to coronary artery aneurysms. Immunohistochemical studies confirmed the inflammatory nature of these infiltrates, which were composed of lymphocytes, macrophages and plasma cells. Further studies revealed the presence of many plasma cells in inflammatory infiltrates in coronary arteries and pulmonary tissue of fatal KD patients with IgA as the predominant immunoglobulin isotype. 10 Additionally it was shown that this IgA immune response was oligoclonal, suggesting an antigen-driven B cell activation at a mucosal site. 11 Epidemiologic case-control studies on KD outbreaks in Rochester, NY, Massachusetts and Hawaii revealed an association of KD with a history of antecedent respiratory tract infection. 12,13 Taken together, epidemiologic, immunologic and clinical findings suggest a respiratory portal of entry of the etiologic agent of KD. Many efforts have been made to find a specific pathogen as the cause of KD, but no consistent results were obtained either for bacterial (Mycoplasma, Chlamydia, mycobacteria) or viral (Epstein-Barr virus, cytomegalovirus, parvovirus B19, TT virus, herpes virus 8, hepatitis G virus) specimens or superantigens. 14
In the 2 cases described here, the pulmonary symptoms were misleading because of their predominance in the course of the disease; additionally the abnormalities of the chest radiographs were compatible with pneumonia as the cause of illness. In the absence of other typical features, the first case is atypical or incomplete Kawasaki disease, although the laboratory studies showed a marked and prolonged inflammatory process. Retrospectively the second case fulfilled the diagnostic criteria for KD, but moderately elevated CRP and normal white blood counts are unusual for the systemic inflammatory response normally seen in classic KD. The correct diagnosis was made only after characteristic complications or late manifestations of Kawasaki disease occurred.
In conclusion pulmonary symptoms and chest radiographs with changes compatible with pneumonia should be reevaluated in the course of a febrile illness. When there is a prolonged inflammatory reaction and no infectious agent identified or when other physical findings such as rash, lymphadenopathy or conjunctival injection occur, Kawasaki disease should be taken into consideration.
1. Burns JC, Wiggins JW Jr, Toews WH, et al. Clinical spectrum of Kawasaki disease
in infants younger than 6 months of age. J Pediatr
2. Rosenfeld EA, Corydon KE, Shulman ST. Kawasaki disease
in infants less than one year of age. J Pediatr
3. Voynow JA, Schanberg L, Sporn T, Kredich D. Pulmonary complications associated with Kawasaki disease
. J Pediatr
4. Uziel Y, Hashkes PJ, Kassem E, Gottesman G, Wolach B. “Unresolving pneumonia” as the main manifestation of atypical Kawasaki disease
. Arch Dis Child
5. Tizard EJ, Suzuki A, Levin M, Dillon MJ. Clinical aspects of 100 patients with Kawasaki disease
. Arch Dis Child
6. Levy M, Koren G. Atypical Kawasaki disease
: analysis of clinical presentation and diagnostic clues. Pediatr Infect Dis J
7. Umezawa T, Saji T, Matsuo N, Odagiri K. Chest x-ray findings in the acute phase of Kawasaki disease
. Pediatr Radiol
8. Amano S, Hazama F, Kubagawa H, Tasaka K, Haebara H, Hamashima Y. General pathology of Kawasaki disease
: on the morphological alterations corresponding to the clinical manifestations. Acta Pathol Jpn
9. Freeman AF, Crawford SE, Finn LS, et al. Inflammatory pulmonary nodules in Kawasaki disease
. Pediatr Pulmonol
10. Rowley AH, Shulman ST, Mask CA, et al. IgA plasma cell infiltration of proximal respiratory tract, pancreas, kidney, and coronary artery in acute Kawasaki disease
. J Infect Dis
11. Rowley AH, Shulman ST, Spike BT, Mask CA, Baker SC. Oligoclonal IgA response in the vascular wall in acute Kawasaki disease
. J Immunol
12. Bell DM, Brink EW, Nitzkin JL, et al. Kawasaki syndrome: description of two outbreaks in the United States. N Engl J Med
13. Dean AG, Melish ME, Hicks R, Palumbo NE. An epidemic of Kawasaki syndrome in Hawaii. J Pediatr
14. Barron KS. Kawasaki disease
: etiology, pathogenesis, and treatment. Cleve Clin J Med
. 2002;69(suppl 2):SII69–SII78.