N-TERMINAL PRO-BNP CORRELATES WITH MYOCARDIAL INVOLVEMENT IN PATIENTS WITH KAWASAKI DISEASE

Abstract

N-terminal pro-BNP (NT-proBNP) has been recognized as a marker of Kawasaki disease (KD). This is based on the hypothesis that cardiac inflammation during acute KD causes myocardial wall stress. Parameters of myocardial involvement determined by electrocardiogram (PR interval, QT dispersion, QTc interval, R-T axis) and by echocardiogram (systole and diastole) were correlated with serum level of NT-proBNP in the acute (1 week), sub-acute (2-3 months) and chronic (6 months to 1 year) phases of KD. KD patients were compared to a febrile group. KD patients were further subdivided into 2 groups according to the levels of NT-proBNP, KD-1 with normal NT-proBNP (NT-proBNP Z-score < 2), KD-2 with elevated NT-proBNP (Z-score ≥ 2). There were a total of 56 subjects, 14 controls, 19 KD-1 and 23 KD-2 patients. Age was similar between groups (Control vs. KD, 3.8 ± 4.3 vs. 3.3 ± 2.3 years-old, p=0.609). There was a significantly reduced shortening fraction in KD patients in the acute phase, more intensely seen in KD-2 patients as witnessed by a diminished shortening fraction Z-score (-0.5 ± 1.5 in KD-1 vs. -1.6 ± 1.5 in KD-2; p=0.025). There was also a lower ejection fraction in KD patients compared to C (61.9 ± 6.5 C vs. 57.4 ± 7.5 % KD, p=0.049). There was a longer QTc interval in KD patients vs. C (412.3 ± 21.0 mS vs. 390.6 ± 14.6 mS), respectively. In contrast, there were no significant differences for left ventricular mass index (p=0.935) or LV end-diastolic diameter (p=0.565). Likewise, there were no significant differences for the PR interval (p=0.344), QT dispersion (p=0.288) or R-T axis (p=0.577). Finally, there were no differences between groups regarding diastolic function parameters. Further analysis showed a significant correlation between coronary artery (CA) involvement (CA z-score ≥ 2) and the likelihood of lower LV ejection fraction (p=0.049) and higher NT-proBNP z-score (p=0.043), but no correlation with normalized LV shortening fraction (p=0.16) or QTc (p=0.14). In acute KD, there is a reduction in shortening fraction and ejection fraction, to a higher extent in cases with elevated NT-proBNP (Z-score ≥ 2.0). This also correlates with CA involvement (CA z-score ≥ 2.0). These differences disappear in the sub-acute phase. On the other hand, a lengthening of QTc interval in the acute phase, irrespective of NT-proBNP status, resolves after 2-3 months. KD patients with elevated zNT-proBNP may warrant specific myocardial follow-up.