ª Springer Science+Business Media New York 2014

Abdominal Imaging

Abdom Imaging (2014) DOI: 10.1007/s00261-014-0104-2

CT findings of primary clear cell carcinoma of liver: with analysis of 19 cases and review of the literature Haiyan Wang,1 Bingyi Tan,2 Bin Zhao,1 Gary Gong,3 Zhuodong Xu1 1

Shandong Medical Imaging Research Institute, Shandong University, Jingwu Road No.324, Jinan, People’s Republic of China 250021 2 Shandong Provincial Hospital, Jingwu Road No.324, Jinan 250021, People’s Republic of China 3 Department of Radiology, Johns Hopkins Hospital, 4940 Eastern Avenue, Baltimore, MD 21224, USA

Abstract Purpose: To analyze CT characteristics of primary clear cell carcinoma of the liver (PCCCL) and improve the current understanding and diagnose accuracy of the tumor. Methods: Pre- and post-contrast CT images of 19 patients with pathology proven PCCCL were retrospectively analyzed. The clinical data and CT findings as well as relevant literature reports were reviewed. Results: Thirteen patients were tested positive for HBsAg, and two patients were positive for HCVAb. The serum alpha-fetoprotein (AFP) levels of most tumors (14/19) were £20 ng/ml with 14 cases were associated with liver cirrhosis. All lesions were solitary intraparenchymal mass lesions which have well-defined boarders. On pre-contrast CT scans, 15 lesions appeared as hypo-attenuation and four lesions appeared as isointensity to the adjacent liver parenchyma. On postcontrast CT scans, 16 lesions showed avid enhancement on the hepatic arterial phase, of which 6 lesions were hypo-attenuation, and 10 lesions remained slightly hyper-attenuation or iso-attenuation on the portal venous phase images. Three lesions showed only mild enhancement on the hepatic arterial phase and hypoattenuation on the portal venous phase. All lesions demonstrated hypointensity on the equilibrium phase. There are 12 lesions showed pseudocapsules. None of patients showed signs of portal vein thrombosis. There was no distal metastasis except only one patient had lymph node metastasis. Conclusion: The characteristics of CT imaging of PCCCL, such as tend to form pseudocapsules and less

Correspondence to: Bin Zhao; email: [email protected]

involved with vascular invasion, could be useful in differentiating from common type hepatocellular carcinoma (CHCC). Some CT imaging characteristics of PCCCL are similar to CHCC, such as prone to occur in patients with liver cirrhosis and early enhancement pattern on the hepatic arterial phase as well as hypoattenuation on the equilibrium phase. Those features could be useful in differentiating PCCCL from other liver tumors, such as hemangioma and hepatic metastases. Key words: Liver neoplasm—Primary clear cell carcinoma—Computed tomography—Hepatocellular carcinoma—Pseudocapsules

Primary clear cell carcinoma of liver (PCCCL) is an uncommon subgroup of hepatocellular carcinoma (HCC) in which a large number of cells show clear cytoplasm that does not stain with hematoxylin and eosin. PCCCL is generally diagnosed when the tumor contains more than 50% clear cells [1–3]. The frequency is between 7.5% and 12.5% of all HCC reported in the literature [4]. PCCCL has been characterized by a higher rate of HCV infection, and low-grade malignancy, which makes it a unique entity, differs from common type hepatocellular carcinoma (CHCC) [5–7]. It is generally reported that PCCCL has a favorable prognosis compared with that of CHCC [5, 8, 9]. The prognosis of patients in the PCCCL group was related to clear cell ratio, preoperative liver function, liver cirrhosis, HCV infection, pseudocapsule formation, vascular invasion, and multiple tumor occurrences [10].

H. Wang et al.: CT findings of primary clear cell carcinoma of liver

As we know, the CT imaging features of PCCCL have not been fully reported. [4, 11]. In this study, we retrospectively analyzed the CT imaging features of 19 cases with PCCCL in order to investigate the CT diagnostic factors and improve the understanding and diagnosis of the PCCCL.

thrombosis, and lymph node metastasis. The size of each lesion was estimated as the maximum diameter measured on the axial CT images where the lesions were best revealed.

Materials and Methods

Differences in patient age and tumor size were assessed with an independent samples Student t test. Statistical analysis was performed using the SPSS 13.0 software package (SPSS Inc., Chicago, IL, USA).

Study population Institutional research ethics review board approval was obtained for this retrospective study, and informed consent was waived. Patient anonymity was maintained. Nineteen cases with pathologically proven PCCCL in our hospital between January 2009 and December 2012 were enrolled in this study. Eighteen patients underwent tumor resection, and 1 patient underwent biopsy. All pathologic specimens were reviewed by two pathologists with concurred diagnoses of PCCCL. Those with clear cells in less than 50% of the lesion were considered insufficient to be classified as PCCCL in our study.

CT technique All patients underwent dynamic CT examination (12 patients performed on Siemens sensation cardiac 64, Siemens Healthcare, Germany; 7 patients on dual source flash 128-slice MDCT system; Siemens; Siemens Healthcare, Germany). Scan parameters were as follows: section thickness: 1.5 mm; image interval: 1.0 mm; pitch: 0.6. All CT acquisitions were started at the top of the liver and proceed in a craniocaudal direction during a single breath-hold on deep inspiration. A bolus of 1.5 ml/kg of contrast medium (Schering Ultravist, Iopromide, 350 mg I/ml, Berlin, Germany) was injected intravenously at rate of 3.5 ml/s by using a dual-head power injector (Stellant; Medrad, Indianola, PA), followed by a half volume of the total contrast medium saline flush. All patients underwent CT scans included the acquisition of unenhanced images of the entire liver followed by acquisition of triple-phase contrast-materialenhanced images during the hepatic arterial phase, portal venous phase, and equilibrium phase which were obtained at 25, 70, and 120 s, respectively, after the of injection of contrast agent.

Image interpretation All the original films were reviewed by two radiologists with 12 and 16 years experiences in liver imaging, respectively. They reached a final consensus after their independent evaluation. The clinical data and the following CT findings were reviewed: the presence of liver cirrhosis, tumor location, number, size, density, pattern of contrast enhancement relative to the surrounding liver parenchyma, pseudocapsule presence, portal vein

Statistical analysis

Results Clinical findings Age of the patients ranged from 22 to 76 years (mean = 50.4 years). There were more men in this group (14/19) with a male-to-female ratio of 2.8:1. Ten cases presented with right-sided upper abdominal pain or discomfort, fatigue, and anorexia. Nine cases were incidentally discovered to have liver mass during abdominal imaging as part of routine physical examination. Thirteen cases were tested positive for HBsAg, 2 cases were positive for hepatitis C virus antibody, and 4 cases were negative for both HBsAg and HCVAb. The serum alphafetoprotein (AFP) levels of most cases (14/19) were £20 ng/ml (the upper limit of the normal in our hospital.).

CT findings In our study of 19 cases, all the lesions were solitary mass; 12 lesions were located in right lobe and seven lesions in left lobe. Fourteen cases were associated with liver cirrhosis. The tumor sizes ranged from 2.6 to 17.1 cm (mean = 7.0 cm) with 14 lesions presented greater than 5.0 cm and three lesions less than 3.0 cm at the time of discovery. All lesions were well circumscribed as intraparenchymal solid masses. On pre-contrast CT scans, 15 lesions appeared as hypo-attenuation and four lesions appeared as iso-attenuation when compared to the adjacent normal appearing liver parenchyma. Heterogeneous irregular focal hypo-attenuation regions were found in 15 lesions. On post-contrast CT scans, 16 lesions demonstrated avidly enhancement on the early hepatic arterial phase images. In those 16 lesions, six were hypo-attenuation, and 10 were slightly hyperattenuation to iso-attenuation on the portal venous phase. Three lesions showed only mild post-contrast enhancement on the hepatic arterial phase and hypoattenuation on the portal venous phase. All of those 19 lesions were hypo-attenuation compared with normal liver parenchyma on the equilibrium phase. (Figs. 1, 2, 3). Twelve lesions had pseudocapsules. The pseudocapsules demonstrated as low attenuation haloes on precontrast scans and rim enhancement after contrast

H. Wang et al.: CT findings of primary clear cell carcinoma of liver

Fig. 1. Primary clear cell carcinoma of the liver in a 31-yearold woman. A On pre-contrast computed tomography scan, the mass shows slight hypo-attenuation; B at hepatic arterial phase, the mass shows early obvious enhancement; C at

portal venous phase, the mass presents hypo-attenuation; D at the equilibrium phase, the mass presents hypo-attenuation; E pathologically, the mass is mainly composed of clear cells (HE, 9200).

administration around the lesions (Fig. 4). There was no portal vein thrombosis or vascular invasion demonstrated. There was no distal metastasis demonstrated at the time of diagnosis, with only one patient had regional lymph node metastasis..

Discussion PCCCL is a rare and specific subgroup of primary HCC. Some authors diagnose PCCCL when the tumor contains more than 30% clear cells on pathology specimen [7].

H. Wang et al.: CT findings of primary clear cell carcinoma of liver

Fig. 2. Primary clear cell carcinoma of the liver in a 76-yearold man. A On pre-contrast computed tomography scan, the mass shows slight hypo-attenuation; B at hepatic arterial phase, the mass shows early obvious enhancement; C at

portal venous phase, the mass presents iso-attenuation; D at the equilibrium phase, the mass presents hypo-attenuation; E pathologically, the mass is mainly composed of clear cells (HE, 9200).

H. Wang et al.: CT findings of primary clear cell carcinoma of liver

Fig. 3. Primary clear cell carcinoma of the liver in a 57-yearold woman. A On pre-contrast computed tomography scan, the mass shows hypo-attenuation; B at hepatic arterial phase, the mass shows slightly enhancement; C at portal venous

phase, the mass presents hypo-attenuation with rim enhancement (arrows); D at the equilibrium phase, the mass presents hypo-attenuation; E pathologically, the mass is mainly composed of clear cells (HE, 9200).

H. Wang et al.: CT findings of primary clear cell carcinoma of liver

Fig.4. Primary clear cell carcinoma of the liver in a 42-yearold man. A On pre-contrast computed tomography scan, the mass shows slight hypo-attenuation with a hypo-attenuation halo (arrows); B at hepatic arterial phase, the mass shows early enhancement with slightly rim enhancement (arrows); C

at portal venous phase, the mass presents hypo-attenuation with rim enhancement (arrows); D at the equilibrium phase, the mass presents hypo-attenuation with rim enhancement (arrows); E pathologically, the mass shows a pseudocapsule (arrows) (HE, 9100).

H. Wang et al.: CT findings of primary clear cell carcinoma of liver

Most pathologists diagnosed PCCCL when the proportion of clear cells is >50% [1–3, 5]. Tumors with clear cells ranging from 90% to 100% are extremely rare [7, 12]. We used the diagnostic criteria generally accepted by pathologists in China to diagnose PCCCL only when it contained more than 50% of tumor cells are clear cell type [13]. The hypothesis for the mechanism of clear cell dominance is that the clear cell change and fatty change may be caused by metabolic derangement within the tumor nodule or by impaired blood supply related to inadequate vascular development in the tumor nodule during the early stage [7]. This could explain, in part, the relative less aggressiveness of this tumor. However, the complete explanation of the clear cell development in this tumor is not known. The notable clinicopathological features of the patients in the PCCCL group include a higher rate of history of HCV infection, pseudocapsule formation, and a lower rate of vascular invasion [5, 11]. Previous studies have characterized PCCCL by a higher rate of HCV infection [5–7] as well. Emile et al. reported 40% (4 of 10) PCCCL patients with positive HCV antibody [6], and Liu and coworkers [13] also found seven PCCCL cases in nine HCV-associated HCC patients (77.8%). Liu et al. [5] reported 14% (6 of 43) patients with positive HCV antibody in PCCCL patients and only 1.2% positive of HCV antibody (7 of 597) in the CHCC . This difference in positive HCV antibody was statistical significant (P < 0.05) [5]. In our series, two patients (10.5%) were positive for HCV infection. The ratio of HCV infection was close to that of Liu et al. Pseudocapsule formation is an important gross pathologic feature of PCCCL. Previous studies have reported that PCCCL were more prone to form pseudocapsules when compared with CHCC [5, 11, 14]. Murakata et al. [14] reported pseudocapsules formation in 90% of PCCCL patients (9 of 10 cases). Liu et al. reported pseudocapsules formation in 88.4% (38 of 43) in the PCCCL group, which was significantly higher than the CHCC group (68%) [5]. Liu et al. [11] revealed that PCCCL was more associated with pseudocapsules formation with a rate of 75.0% (15 of 20) compared with CHCC (49.6%). There were 12 lesions (63.2%) had pseudocapsules formation in our investigation, slightly lower than that of Liu et al. [5, 11] PCCCL have lower rate in vascular invasion than in CHCC. There were no patients with vascular invasion seen in our group. The clinical symptoms of PCCCL are similar to those of conventional HCC [5]. The most common symptoms in our study were abdominal pain or discomfort, fatigue, and anorexia. Some of the lesions were found incidentally. Buchanan et al. reported that the ratio of the male and female of PCCCL was 1.6:1, indicating a high prevalence in females compared with CHCC [2]. The male-to-female ratio (2.8:1) in our group was very close to that of Liu et al. [5] (2.9:1). Liu et al. [5] found a

relatively higher incidence in female patients, but there was no statistically significant difference compared with CHCC group. No significant difference in AFP-positive rate was observed between the PCCCL and CHCC groups [5]. However, in our series, the serum AFP levels of most patient (14/19) were £20 ng/ml. The reason of the difference was unclear. Previous studies have reported that the prevalence of PCCCL was higher in patients with liver cirrhosis [6, 7]. However, other authors reported that PCCCL have lower ratio of patients associated with liver cirrhosis by compared that of PCCCL group with CHCC group [5]. There were 14 patients with liver cirrhosis in our group. The association of liver cirrhosis is 73.7% with PCCCL in our group, similar to the result of Liu et al (76.7%) and is lower than incidence in the CHCC group (86.1%) [5]. PCCCL had a better prognosis than CHCC, and surgical resection is an effective way to achieve favorable outcomes for patients with PCCCL [5, 15, 16]. Computed tomography (CT) is an important imaging modality for the detection and characterization of liver tumors, except for few case reports [2, 11]. The imaging features of PCCCL have rarely been reported in the English literature [4, 11, 17]. Takahashi et al. [4] described CT, MR and angiographic findings of PCCCL in a woman with a non-cirrhotic liver. Liu et al. [11] retrospectively analyzed the CT and MRI appearances of 20 patients with PCCCL with comparison to CHCC findings. Yan et al. [17] retrospectively analyzed the imaging findings of seven cases of PCCCL as well. Their studies revealed also that the imaging characteristics of PCCCL were similar to those of CHCC, and PCCCL was more prone than CHCC to form pseudocapsules [17]. The typical enhancement pattern of small HCCs of 2 cm or less on dynamic contrast scanning showed early enhancement at hepatic arterial phase and rapid contrast medium washout at portal venous phase. Large HCCs may have a number of characteristic features, such as mosaic or heterogenous pattern [18]. In our series, 6 lesions showed typical ‘‘wash-in and wash-out’’ enhancement pattern, indicating that the tumor has rich blood supply. However, 10 lesions showed early enhancement on hepatic arterial phases, sustained enhancement on portal venous phases, and slow washout at equilibrium phases. The cases demonstrating such enhancement pattern were bigger lesions in our group and were the majority of the masses. This pattern could represent one of the characteristic features of PCCCL but further confirmation with larger sample populations is needed. Three lesions showed only mild enhancement on the hepatic arterial phase and hypo-attenuation on the portal venous and equilibrium phase, indicating that those three tumors probably had impaired or otherwise altered vascular structures. Therefore, variations of contrast enhancement of this tumor exist. On pre-contrast CT images, pseudocapsules formation presented as a low-

H. Wang et al.: CT findings of primary clear cell carcinoma of liver

attenuating thin rim surrounding the mass. On enhancement CT scans, the pseudocapsules showed linear enhancement, which were best seen on portal venous phase images. This study has several limitations. First, our patient population is relatively small with limited statistical power, and a bigger patient population is needed in further research. Second, we did not do a comprehensive comparative examination between PCCCL and CHCC groups. Such study could provide possibly more or better differentiating features between PCCCL and CHCC, given the biological behaviors of those two cancers are likely different. If this observation could be proven to be true, biopsy of those lesions to obtain tissue diagnosis maybe needed to guide better clinical management. In conclusion, we analyzed CT findings of pathology proven PCCCLs in our 19-case series as well as findings in the reported literature. Although some of the CT imaging characteristics of PCCCL are similar to that of CHCC, such as higher prevalence in patients with liver cirrhosis, pre-contrast hypo-attenuation, and early enhancement pattern on the hepatic arterial phase. Images features of their higher incidence of pseudocapsules formation, apparent less vascular involvement, and hypo-attenuation on delayed images could be the differentiating characters from other liver masses such as CHCC, hemangioma, and hepatic metastasis. References 1. Wu PC, Lai CL, Lam KC, et al. (1983) Clear cell carcinoma of liver. An ultrastructural study. Cancer 52:504–507 2. Buchanan TF Jr, Huvos AG (1974) Clear-cell carcinoma of the liver. A clinicopathologic study of 13 patients. Am J Clin Pathol 61:529–539 3. Ji SP, Li Q, Dong H (2010) Therapy and prognostic features of primary clear cell carcinoma of the liver. World J Gastroenterol 16:764–769

4. Takahashi A, Saito H, Kanno Y, et al. (2008) Case of clear-cell hepatocellular carcinoma that developed in the normal liver of a middle-aged woman. World J Gastroenterol 14:129–131 5. Liu Z, Ma W, Li H, et al. (2008) Clinicopathological and prognostic features of primary clear cell carcinoma of the liver. Hepatol Res 38:291–299 6. Emile JF, Lemoine A, Azoulay D, et al. (2001) Histological, genomic and clinical heterogeneity of clear cell hepatocellular carcinoma. Histopathology 38:225–231 7. Yang SH, Watanabe J, Nakashima O, et al. (1996) Clinicopathologic study on clear cell hepatocellular carcinoma. Pathol Int 46:503–509 8. Li T, Fan J, Qin L-X, et al. (2011) Risk factors, prognosis, and management of early and late intrahepatic recurrence after resection of primary clear cell carcinoma of the liver. Ann Surg Oncol 18:1955–1963 9. McDermott WV, Cady B, Georgi B, et al. (1989) Primary cancer of the liver. Evaluation, treatment, and prognosis. Arch Surg 124:552– 554 10. Ji S-P, Li Q, Dong H (2010) Therapy and prognostic features of primary clear cell carcinoma of the liver. World J Gastroenterol 16(6):764–769 11. Liu Q-Y, Li H-G, Gao M, et al. (2011) Primary clear cell carcinoma in the liver: CT and MRI findings. World J Gastroenterol 17(7):946–952 12. Pecorella I, Ciardi A, Aiello E, et al. (1994) Clear cell hepatocellular carcinoma treated with liver transplantation. Pathologica 86:307– 310 13. Zhang YH, Liu FS, Liu TH, et al. (2001) Pathological classification of Chinese tumor. Beijing: Scientific Literature and Technology Press, p 229. 14. Murakata LA, Ishak KG, Nzeako UC, et al. (2000) Clear cell carcinoma of the liver: a comparative immunohistochemical study with renal clear cell carcinoma. Mod Pathol 13:874–881 15. Lao XM, Zhang YQ, Jin X, et al. (2006) Primary clear cell carcinoma of liver: clinicopathologic features and surgical results of 18 cases. Hepatogastroenterology 53:128–132 16. Ye XP, Li LQ, Peng T, et al. (2010) Diagnosis and treatment of primary clear cell carcinoma of the liver. Zhonghua Zhong Liu Za Zhi 32:64–66 17. Tan Y, Xiao E-h (2013) Rare hepatic malignant tumors: dynamic CT, MRI, and clinicopathologic features: with analysis of 54 cases and review of the literature. Abdom Imaging 38(3):511–526 18. Jeong YY, Yim NY, Kang HK (2005) Hepatocellular carcinoma in the cirrhotic liver with helical CT and MRI: imaging spectrum and pitfalls of cirrhosis-related nodules. Am J Roentgenol 185:1024– 1032

CT findings of primary clear cell carcinoma of liver: with analysis of 19 cases and review of the literature.

To analyze CT characteristics of primary clear cell carcinoma of the liver (PCCCL) and improve the current understanding and diagnose accuracy of the ...
11MB Sizes 1 Downloads 3 Views