Hepatoid adenocarcinoma (HA) is a rare neoplasm, which has a striking morphologic similarity to hepatocarcinoma. It has been described in organs such as stomach, lung, and pancreas,1–3 but there are reports describing the tumor in other organs such as esophagus, colon,4 pancreas,5 papilla of Vater, ovaries,6–10 uterus, cervix, urinary bladder,11–13 and renal pelvis.14 The histologic features are consistent in all anatomic sites: polygonal atypical cells with eosinophilic and granular cytoplasm arranged in solid nests and/or in trabecular sheets. In half of the cases, HA shows immunoreactivity to α-fetoprotein (AFP) and polyclonal carcinoembryonic antigen (CEA) (canalicular pattern); besides, serum levels of AFP can be elevated in many cases.1,3,5,11,13
This tumor is mostly observed in older patients and is characterized by a very aggressive biological behavior. HA may be clinically and histologically indistinguishable from hepatocarcinoma.
HA originates in the epithelium of endodermal origin and its localization in the urothelium is exceptionally rare with only few cases reported in the literature.11–14
Here, we report the case of an old female patient affected by HA of the ureter with ovarian, small bowel, and hepatic involvement. We discuss the clinical aspects, the morphologic features, and the immunoistochemical staining useful for differential diagnosis.
An 80-year-old woman presenting gross hematuria performed an ultrasound scan that showed a right pelvic mass involving the ureter, the ovary, and the small intestine. An earlier pelvic scan, 1 year before, resulted normal. In addition, the scan identified multiple lesions involving the liver and an enlarged mesointestinal lymph node. Preoperative colonscopy and gastroscopy showed no abnormalities. Radical nephroureterectomy with resection of the mass involving the ureter, the ovary, and small bowel was performed. Moreover, a hepatic biopsy was obtained and internal iliac lymph nodes were removed. There was no peritoneal carcinosis.
Pathology and Immunohistochemistry
Specimens were fixed in 10% buffered formalin, embedded in paraffin, sectioned, and stained with hematoxylin-eosin stain. Immunohistochemical stains were performed with the following markers: AFP, cytokeratin (CK) 7, CK-high molecular weight (34βE12), CK 8/18 (CAM5.2), CK19, CK-low molecular weight (35βH11), epithelial membrane antigen, endothelial trascription factor 3 (GATA-3), tumor protein p63 (p63), CD15, carcinoembryonic antigen, policlonal (CEAp), CD10, CK20, S100, HMB45, actin-smooth muscle (1A4 actin), vimentin, human chorionic gonadotropin β subunit (β), caudal type homeobox transcription factor 2 (CDX-2), chromogranin, synaptophysin, estrogen receptor protein (ERP), progesterone receptor protein (PRP), CA125, and α-inhibin. The antibodies used are detailed in Table 1. Appropriate positive and negative controls were included. The expression of the antibodies used was scored as 0 in the case of <5% tumor cells positive; 1+ in the case of 5% to 25% of tumor cells positive; 2+ in the case of 26% to 50% of tumor cells positive; 3+ in the case >50% of tumor cells positive.
Macroscopic examination showed a solid and firm white mass of 4×3×2 cm involving the right ureter and ovary, which infiltrated small bowel by extrinseco. Kidney was hydronephrotic and measured 7×4.5×4 cm. Microscopically, the tumor was characterized by tumor cells arranged in an organoid pattern with solid nests, separated by a desmoplastic stroma. There was also evidence of areas constituted by trabeculae or cords of large-medium cells with eosinophilic cytoplasm and round-oval nuclei (Fig. 1A). The cells appeared polygonal, resembling hepatocytes, with clear granular cytoplasm and a large number of prominent nucleoli. There were evident mitoses and nuclear pleomorphism. We did not observe bile pigment. The lesion involved the wall of the right ureter, the right ovary, the small bowel by extrinseco, and the liver. There were metastatic internal iliac lymph nodes. In the ureter, we also observed areas with in situ HA (Fig. 1B). In the ovary, which was rather small (diameter 2.5 cm), we observed that there was a multinodular pattern of growth, with nodules protruding from the surface of the ovary and there were tumor cells in the vessels of the hilus ovarii. The perirenal stroma was also involved by the neoplasm. In the hepatic specimen, tumor cells were adjacent to non-neoplastic hepatocytes arranged in lobular architecture with no cirrhosis (Fig. 2A). No other nodules or chronic active hepatitis were observed. Reticulum staining showed normal liver architecture.
Tumor cells in the different part of the pelvic mass and the hepatic metastasis stained positively for AFP (3+) (Figs. 1C, D), CK19 (3+) (Fig. 2B), p63 (2+) (Figs. 3A, 4A), GATA-3 (2+) (Figs. 3B, 4B), CD15 (2+) (Figs. 3C, 4C), 34βE12 (3+) (Figs. 3D, 4D), CK7 (2+), CAM5.2 (3+), 35βH11 (3+), and epithelial membrane antigen (2+). There was no immunoreactivity for CEAp, CD10, CK20, S100, HMB45, 1A4 actin, vimentin, βHCG, CDX-2, chromogranin, synaptophysin, ERP, PRP, CA125, or α-inhibin. Combining the clinicopathologic and morphologic finding with the immunohistochemical results, the diagnosis of HA of the ureter with metastasis of ovary, small bowel, and liver was made on the basis of the positive staining of AFP and CK19 and of the positive staining of several markers associated with urothelial differentiation including CD15, 34βE 12, and the recently characterized marker GATA-3,15 and the negative staining of CD10, CK20, S100, HMB45, 1A4 actin, vimentin, βHCG, CDX-2, chromogranin, synaptophysin, ERP, PRP, CA125, and α-inhibin.
The term hepatoid has been used to designate a series of tumors that can be grouped in 2 categories11,12: germinal tumor (as yolk sac tumor) with focal areas of hepatoid differentiation and true HAs characterized by formation of cords of polygonal cells or evidence of biliary pigment production, that is strict morphologic criteria of hepatoid differentiation.
The term HA should be strictly reserved to tumors with hepatoid morphologic features and AFP secretion. AFP is a fetal serum protein synthesized by fetal liver, yolk sac, and gastrointestinal tract; the production of which commonly decreases after birth, but it can be higher than normal in patients with hepatocarcinoma or germ cell tumors.16 AFP can be increased in conditions that present similarity with hepatocarcinoma, as in the case of HA. However, as elevated circulating levels of AFP do not always imply its production by an hepatoid neoplasm, HAs are not always accompanied by high levels of circulating AFP, although the cytoplasmatic AFP staining is always evident in tumor cells.12 This was observed in this case. In fact, there was intense positive stain for AFP, whereas AFP serum levels of the patient resulted normal.
HAs are generally aggressive tumors. In literature, a liver involvement is described in some HA of the stomach,1–3 and in only 1 HA of the rectum4 and the lung.3 We report a first case of HA of urothelial origin with hepatic metastasis. For the fact that HAs can mimic a primitive tumor of the liver, mainly hepatocarcinoma, it is mandatory to make an accurate differential diagnosis. The differential diagnosis between HA and hepatocarcinoma can be very difficult, but it can be supported by a careful evaluation of clinicopathologic data. In this case, the finding of multiple neoplastic hepatic nodules in the absence of cirrhosis or chronic active hepatitis, together with the negativity of viral markers, strongly favours the diagnosis of liver metastasis from HA. In addition, in this case, both primary tumors and liver metastasis showed a strong positive immunostaining for CK19, and this result is in contrast with the diagnosis of hepatocarcinoma. In fact, it has been reported that most HAs are immunoreactive for CK19, whereas only a minority of hepatocarcinomas express CK19.3,6
The neoplastic involvement of the ovary needs a further analysis: it is necessary to exclude a primary gonadic HA, and to rule out the possibility of a yolk sac tumor mimicking HA.
The histologic findings do not also match with the diagnosis of a primary hepatoid tumor of the ovary; in fact, in the few cases reported of primitive ovarian carcinoma, areas with serous carcinoma and psammomatous bodies have been observed; moreover, in this case, we observed a typical pattern of metastasis to the ovary, such as a small organ, multinodular pattern of growth, and tumor cells in the ilo's vessels.6–10
In contrast, yolk sac tumor involves patients much younger than those affected by HA: most of these tumors occur in patients <30 years old and all female patients are in reproductive age, whereas HA occurs in older patients. Furthermore, it is possible to observe in the yolk sac tumor glandular structures which are absent in the HA. Features supporting a diagnosis of HA are the presence of giant and bizarre cells with nuclei of variable size and shape, which are absent in the yolk sac tumor. As CK 19 has been reported to be positive in a case of hepatoid yolk sac tumor, CK staining does not seem to be useful in the differential diagnosis. Therefore, the age, the pattern of spreading, and above all, the absence of germ cells (mostly teratomatous elements), excluded the diagnosis of yolk sac tumor in this case.6–10
We were also able to show the origin of this neoplasm. The presence of areas in the epithelium of the ureter showing in situ HA, which shows an AFP positive staining, associated to areas with invasive tumor associated to the positive immunostaining of markers associated urothelial differentiation including CD15, 34βE12, and the recently characterized marker GATA-3, strongly suggest an urothelial origin of the tumor.
The histogenesis of HA remains a matter of debate. Embryologic endodermal differentiation may explain the presence of hepatoid areas in cases which occur in the digestive tract or are associated with yolk sac or other germ cell tumors.11,12,14 HAs involving other organs such as lung, urinary bladder, or renal pelvis are difficult to reconcile with this hypothesis. Some investigators suggest an origin of this neoplasm from a multipotent cell or a differentiation of tumor cells in an endodermal direction.11,13 However, all these theories needs still to be substantiated.
The extreme poor prognosis of this kind of neoplasm is related to the advanced stage at the diagnosis and to the extensive vascular infiltration. Therefore, it is very important to formulate a correct diagnosis of primary HA to help the clinician in the appropriate treatment of these patients.
In conclusion, we present a rare case of HA of the ureter showing the secondary involvement of liver and ovary. We made the diagnosis after combining in a careful analysis morphologic features and a correct combination of immunohistochemical staining.
1. Lin C-H, Hsu C-C, Chang H-C, et al. Hepatoid adenocarcinoma of the stomach with liver metastasis mimicking hepatocellular carcinoma: a case report Cases J.. 2009;2:6317
2. Kumashiro Y, Yao T, Aishima S, et al. Hepatoid adenocarcinoma of the stomach: histogenesis and progression in association with intestinal phenotype Hum Pathol.. 2007;38:857–863
3. Terraciano LM, Glatz K, Mhawech P, et al. Hepatoid adenocarcinoma with liver metastasis mimicking hepatocellular carcinoma Am J Surg Pathol.. 2003;27:1302–1312
4. Borgonovo G, Razzetta F, Assalino M, et al. Rectal hepatoid carcinoma with liver metastases in a patient affected by ulcerative colitis Hepatobiliary Pancreat Dis Int.. 2008;7:539–543
5. Hameed O, Xu H, Saddeghi S, et al. Hepatoid carcinoma of the pancreas: a case report and literature review of a heterogeneous group of tumors Am J Surg Pathol.. 2007;31:146–152
6. Tochigi N, Kishimoto T, Supriatna Y, et al. Hepatoid carcinoma of the ovary: a report of three cases admixed with a common surface epithelial carcinoma Int J Gynecol Pathol.. 2003;22:266–271
7. Ishikura H, Scully RE. Hepatoid carcinoma of the ovary: a newly described tumor Cancer.. 1987;60:2775–2784
8. Scurry JP, Brown RW, Jobling T. Combined ovarian serous papillary and hepatoid carcinoma Gynecol Oncol.. 1996;63:138–142
9. Young RH, Gersell DJ, Clement PB, et al. Hepatocellular carcinoma metastatic to the ovary: a report of three cases discovered during life with discussion of the differential diagnosis of hepatoid tumors of the ovary Hum Pathol.. 1992;23:574–580
10. Yigit S, Uyaroglu MA, Kus Z, et al. Hepatoid carcinoma of the ovary: immunoistochemical finding of one case and literature review Int J Gynecol Cancer.. 2006;16:1439–1441
11. Lopez-Beltran A, Luque RJ, Quintero A, et al. Hepatoid adenocarcinoma of the urinary bladder Virchows Arch.. 2003;442:381–387
12. Sinard J, Macleary L Jr, Melamed J. Hepatoid adenocarcinoma of the urinary bladder. Unusual localization or a newly recognized tumor type Cancer.. 1994;73:1919–1925
13. Burgues O, Ferrer J, Navarro S, et al. Hepatoid adenocarcinoma of the urinary bladder: an unusual neoplasm Virchows Arch.. 1999;435:71–75
14. Ishikura H, Ishiguro T, Enatsu C, et al. Hepatoid adenocarcinoma of the renal pelvis producing alpha-fetoprotein of hepatic type and bile pigment Cancer.. 1991;67:3051–3056
15. Higgins JPT, Kaygusuz G, Wang L, et al. Placental S-100 (S100P) and GATA3: markers for transitional epithelium and urothelial carcinoma discovered by complementary DNA microarray Am J Surg Pathol.. 2007;31:673–680
16. Lu C-H, Chou W-C, Hung Y-S, et al. Alpha-fetoprotein-producing transitional cell carcinoma of the urinary bladder: a case report Int Urol Nephrol.. 2009;41:567–570
© 2011 Lippincott Williams & Wilkins, Inc.