Biliary adenofibromas are rare, primary hepatic tumors with potential for malignant transformation. Previous literature has described tumors ranging in size from 1.8 to 16 cm, with patient ages ranging from 23 to 83 years, and no characteristic symptoms or laboratory findings [2]. Pathologically, they are characterized by microcystic and tubule-acinar glandular structures lined by non-mucin-secreting biliary epithelium and supported by a fibroblastic stromal scaffolding [1]. The suggested pre-malignant changes included papillary growths within the cysts lined by an eosinophilic epithelium with apocrine-like changes, secretory snouts, and vesicular nuclei with prominent nucleoli. Carcinoma arising from biliary adenofibroma is a conventional adenocarcinoma in most cases [10, 12, 13], as could be seen in our case. Till date, 28 cases of biliary adenofibroma have been reported in English literature [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24]; a total of 14 cases have been associated with malignant transformation [2,3,4,5,6,7,8,9,10,11,12,13, 16, 18]. However, the frequency of malignant transformation is unknown due to the lack of large studies.

On ultrasound, two cases presented as a hypoechoic mass [2, 23], while the other four cases were mostly a hyperechoic mass [3, 4, 19, 20]. In the present case, the portion of biliary adenofibroma showed high echogenicity with a clear boundary and the portion of cholangiocarcinoma showed low echogenicity with an irregular margin. Microscopically, the hyperechogenic lesion was composed of abundant microcysts. From this finding, we hypothesized that the high echogenicity might reflect multiple ultrasound reflections from numerous tiny cysts inside the tumor like pancreatic serous cyst neoplasm [25]. Various cyst sizes composing the biliary adenofibroma might have resulted in varied echogenicities in previous reports.

On CT, biliary adenofibroma is reported as a solid-cystic mass with a well-defined margin, lobulated shape, and internal septa [2]. On MR imaging, the serous cystic component shows high signal intensity on T2-weighted images and low signal intensity on T1-weighted images, with multiple septation. Solid components and septa are likely to show low signal intensity on T2-weighted images reflecting high content of fibrous stroma. Unlike intraductal papillary neoplasm of the bile duct, communication or dilation of bile duct is not usually seen. Unlike mucinous cyst neoplasm, intracystic hemorrhage and calcifications are not common [2]. These are consistent with the present case. In addition, our case showed water-like high signal intensity on heavily T2-weighted imaging; these areas showed contrast enhancement, like pancreatic serous cystic neoplasm [25].

On enhanced CT and MR imaging, biliary adenofibroma has been reported to show a gradual delayed enhancement pattern reflecting abundant stroma [2], which is not consistent with our case showing a heterogenous early enhancement and delayed wash-out on CT. Pathologically, in the present case, the density of glandular tissue, expansion of glandular lumen, and fibrous stroma were not uniform. In some areas, tubule-acinar glandular structures were abundant, and glandular expansion and fibrous stroma were sparse; these areas may have shown early enhancement due to abundant blood flow and wash-out due to sparsity of fibrous tissue. As no atypia was seen in these areas, the pattern of enhancement might have been influenced by the proportion of each tissue rather than malignancy.

For malignant transformation, previous research has suggested the following imaging findings as specific features for malignant biliary adenofibroma and not found in benign lesions: (1) multiple lesions, (2) restricted diffusion on MR imaging, (3) obscure margin, (4) enhancement in the arterial phase followed by wash-out in the venous phase [2]. In the present case, diffusion restriction and obscure margin were observed in the malignant component and not in the benign component, which is in accordance with the previous report. However, the malignant component showed peripheral early enhancement with delayed centripetal filling, which is like conventional intrahepatic cholangiocarcinoma, and benign component showed heterogeneous early enhancement with delayed wash-out. As noted in our case, the malignant transformation generally shows conventional cholangiocarcinoma pathology; hence, the delayed centripetal enhancement may have depended on the amount of fibrous stroma of adenocarcinoma.

On FDG-PET/CT, high accumulation was seen in the malignant component and not in the benign component, thereby, clearly distinguishing between the benign and malignant component. The FDG accumulation of the malignant component was compatible with the typical findings of cholangiocarcinoma [26]. To the best of our knowledge, this is the only case showing FDG-PET/CT manifestation of biliary adenofibroma. Hence, further investigation is needed to confirm the usefulness of FDG-PET/CT in the stratification of the risk of malignant transformation.