Langerhans' cell histiocytosis as a cause of periportal abnormal ...

376 M. Kim et al.: MRI of hepatic Langerhans’ cell histiocytosisor nodular. Some hypoechoic nodules looked like targets.We suspect that periportal hypoechogenicity maybe attributed to the infiltration of histiocytes and inflammatorycells based on our biopsy result, whichaccorded with the reports of Hara et al. [4] and Muwakkitet al. [5].MRI findings of the liver in LCH have been notedin two case reports [6, 8]. These cases presented asmass lesions of the liver, which had imaging characteristicsof periportal fibrosis and focal fatty infiltration.In the present cases, MRI showed PASI ratherthan a focal mass. Matsui et al. [12] described periportalabnormal intensity on MRI from patients withcholangitis, obstructive jaundice that was due toedema, inflammatory cell infiltration, and proliferationof bile ductules. On each T1-WI and T2-WI, signalintensity of periportal lesions in our patients was lowerand higher than that of surrounding hepatic parenchyma,respectively. Signal intensity of these lesionswas increased on contrast-enhanced T1-WI with fatsuppression. PASI extended diffusely from the centrallarge portal tracts to peripheral small ones in patients 1and 2, who eventually developed hepatic dysfunction.In addition to the extensive PASI, intrahepatic bile ductdilatation and nodularity due to secondary biliary cirrhosiswere noted in patient 2. This finding is in contrastto PASI confined to the central large portal tract inpatient 3 who has not yet developed hepatic dysfunction.Pathologic examination of this patient showedbile ductular proliferation and periductal histiocyticand lymphocytic infiltration. On follow-up MRI obtained2 years later, PASI was not seen in all sequencesas compared with that of the initial study. We considerthat MRI of the liver in children with disseminatedLCH may be different according to the pathologicalchange, ranging from triaditis to periportal fibrosis andbiliary cirrhosis.In summary, we could find PASI as a common featureon initial MRI of three children with disseminated LCH.As already known, PASI itself is nonspecific. However,in conjunction with clinical and laboratory data, MRI canbe valuable in the estimation of morphologic change ofthe liver in children.ReferencesFig. 3. Case 3. A T1-weighted image (550/17) with fat suppressionshows periportal low signal intensity that is confined to the central portaltracts. B Periportal enhancement is clearly seen on contrast-enhancedT1-weighted image (600/11) with fat suppression. C T1-weighted image(500/8) obtained 2 years later shows no periportal low signal intensity.1. Heyn RM, Hamoudi A, Newton WA. Pretreatment liver biopsy in20 children with histiocytosis X: a clinicopathologic correlation.Med Pediatr Oncol 1990;18:110–1182. Squires RH, Weinberg AG, Zweiner RJ, et al. Langerhans’ cellhistiocytosis presenting with hepatic dysfunction. J Pediatr GastroenterolNutr 1993;16:190–1933. Leblanc A, Hadchouel M, Jehan P, et al. Obstructive jaundice inchildren with histiocytosis X. Gastroenterology 1981;80:134–1394. Hara T, Mizuno Y, Ishii E, et al. Histiocytosis X presenting asmultiple intrahepatic nodules. Acta Haematol Jpn 1988;51:1059–1062