BACKGROUND: Differentiating Kawasaki disease (KD) from cervical lymphadenitis (CL) is clinically difficult but essential given that treatment and outcome differ significantly. Research on differentiation between KD and CL using ultrasound (US) and computed tomography (CT) is limited. The purpose of this study was to identify cervical US and CT findings that may differentiate KD from CL. METHODS: We retrospectively reviewed cervical US of 25 KD patients and 25 CL patients, and CT of 14 KD patients, and 14 CL patients. Two radiologists analyzed specific imaging features on US (lymph node size, shape, echogenicity, margins, laterality, necrosis, and presence of normal hilum) and on CT (size and location of enlarged nodes, laterality, perinodal infiltration, and retropharyngeal edema). RESULTS: On US, patients with KD more frequently had lymph nodes with a "cluster of grapes" appearance (KD vs CL: 64% vs 32%, P < 0.05) and less frequently had poorly circumscribed margins (0% vs 36%, P < 0.01), necrosis (0% vs 32%, P < 0.01), or non-visualization of the hilum (4% vs 36%, P < 0.01). On CT, KD patients more frequently had retropharyngeal edema (100% vs 29%, P < 0.001) and less frequently had level 4 lymphadenopathy (14% vs 79%, P < 0.01) than CL patients. CONCLUSIONS: Ultrasound is mainly useful for excluding purulent lymphadenopathy while CT is a useful diagnostic tool for differentiating KD from CL, especially in patients with incomplete KD, who present with prominent cervical lymphadenopathy and other equivocal principal findings.
BACKGROUND: Differentiating Kawasaki disease (KD) from cervical lymphadenitis (CL) is clinically difficult but essential given that treatment and outcome differ significantly. Research on differentiation between KD and CL using ultrasound (US) and computed tomography (CT) is limited. The purpose of this study was to identify cervical US and CT findings that may differentiate KD from CL. METHODS: We retrospectively reviewed cervical US of 25 KDpatients and 25 CL patients, and CT of 14 KDpatients, and 14 CL patients. Two radiologists analyzed specific imaging features on US (lymph node size, shape, echogenicity, margins, laterality, necrosis, and presence of normal hilum) and on CT (size and location of enlarged nodes, laterality, perinodal infiltration, and retropharyngeal edema). RESULTS: On US, patients with KD more frequently had lymph nodes with a "cluster of grapes" appearance (KD vs CL: 64% vs 32%, P < 0.05) and less frequently had poorly circumscribed margins (0% vs 36%, P < 0.01), necrosis (0% vs 32%, P < 0.01), or non-visualization of the hilum (4% vs 36%, P < 0.01). On CT, KDpatients more frequently had retropharyngeal edema (100% vs 29%, P < 0.001) and less frequently had level 4 lymphadenopathy (14% vs 79%, P < 0.01) than CL patients. CONCLUSIONS: Ultrasound is mainly useful for excluding purulent lymphadenopathy while CT is a useful diagnostic tool for differentiating KD from CL, especially in patients with incomplete KD, who present with prominent cervical lymphadenopathy and other equivocal principal findings.
Authors: Ji Eun Park; Young Jin Ryu; Ji Young Kim; Young Hoon Kim; Ji Young Park; Hyunju Lee; Hyoung Soo Choi Journal: Eur Radiol Date: 2020-03-18 Impact factor: 5.315
Authors: Sooyeon Lim; Na Young Lee; Seung Beom Han; Dae Chul Jeong; Jin Han Kang Journal: Clin Exp Otorhinolaryngol Date: 2019-10-11 Impact factor: 3.372