Case Reports 305 J Vet Diagn Invest 22:305 309 (2010) Lipid-rich carcinoma in the mammary gland of a Djungarian hamster (Phodopus sungorus) Hisashi Yoshimura, Norie Kimura, Rei Nakahira, Masaki Michishita, Kozo Ohkusu-Tsukada, Kimimasa Takahashi 1 Abstract. A lipid-rich carcinoma of the mammary gland was diagnosed in a female Djungarian hamster (Phodopus sungorus), which was kept as an indoor pet. The animal underwent surgery for a primary tumor arising in the mammary gland at the age of 16 months, and also for a recurrent tumor 6 months later. Histologically, the primary neoplasm was composed of 2 different cell populations: nonvacuolated glandular neoplastic cells with moderate atypia, and vacuolated neoplastic cells with marked atypia. Transition from the nonvacuolated glandular cells to the vacuolated cells was frequently seen. The recurrent neoplasm was composed predominantly of vacuolated neoplastic cells that often invaded the surrounding soft tissue. The cytoplasmic vacuoles contained neutral lipids, as confirmed by oil red O and Nile blue staining. The vacuolated neoplastic cells were immunopositive for cytokeratin and negative for vimentin, a-smooth muscle actin, p63, estrogen receptor a, and androgen receptor. Presumably, this high-grade, lipid-rich mammary carcinoma had developed from a low-grade mammary adenocarcinoma. Key words: Djungarian hamster; immunohistochemistry; lipid-rich carcinoma; mammary tumor; Phodopus sungorus. <!?show "fnote_aff1"$^!"content-markup(./author-grp[1]/aff./author-grp[1]/dept-list)> In human medicine, lipid-rich carcinoma is a rare variant of invasive breast carcinoma in which tumor cells contain abundant cytoplasmic neutral lipids. 4,16 The morphological features of this neoplasm were first described in a human patient in 1963. 1 However, because of the great rarity of this neoplasm, many aspects of its pathology remain to be elucidated. In the veterinary literature, lipid-rich mammary carcinoma has been recognized as a rare neoplasm in 12 female dogs 5,11 13 and 1 female cat. 6 The Djungarian hamster (Phodopus sungorus), also known as the Dzhungarian hamster, Siberian dwarf hamster, or Russian dwarf hamster, is a popular, indoor small pet in Japan. This species is also used as a laboratory animal, although it is used less often than the Syrian hamster (Mesocricetus auratus). In Djungarian hamsters, mammary tumor is one of the most common spontaneous neoplasms, 3,7,8,14,18 and androgen-dependent atypical fibromas are likewise known to arise in the skin with aging. 2 A female Djungarian hamster, which weighed 52 g at the age of 16 months and was kept as an indoor pet, developed a subcutaneous mass at the site of the fourth mammary gland on the left side. At 9 months of age, this hamster had undergone surgery to excise a neoplasm in the first left mammary gland. This neoplasm had been diagnosed as an adenocarcinoma. The present mass in the fourth left From the Department of Veterinary Pathology, Faculty of Veterinary Medicine, Nippon Veterinary & Life Science University, Tokyo, Japan. 1 Corresponding Author: Kimimasa Takahashi, Department of Veterinary Pathology, Faculty of Veterinary Medicine, Nippon Veterinary and Life Science University, 1-7-1, Kyonan-cho, Musashino, Tokyo 180-8602, Japan. kimimasa@nvlu.ac.jp mammary gland (the primary tumor) was surgically excised. It measured 2.0 cm 3 1.5 cm 3 1.5 cm and weighed 1.7 g. When the animal was 22 months old and weighed 53 g, the neoplasm in the fourth left mammary gland recurred and was surgically excised again. This mass measured 2.2 cm 3 2.0 cm 3 1.9 cm and weighed 3.9 g. Further follow-up clinical data on the hamster could not be obtained. On gross examination, both the primary and the recurrent neoplasms were brown-white to gray-white and were partly cystic on the cut surface. After fixation in 10% neutral buffered formalin, the tissues were processed routinely, embedded in paraffin, serially sectioned at 4-mm thickness, and stained with hematoxylin and eosin, alcian blue (AB) at ph 2.5, periodic acid Schiff (PAS) with or without diastase digestion, and mucicarmine. Formalin-fixed tissues of the primary and recurrent neoplasms also were frozen, sectioned at 6-mm thickness, and stained with oil red O and Nile blue. Replicate paraffin tissue sections were immunostained using a labeled streptavidin biotin method. Aspects of the primary antibodies (source and dilution), tissue section pretreatment, and positive control tissues used for immunohistochemical analyses are presented in Table 1. Histological examination of the primary and recurrent tumors revealed that both neoplasms were unencapsulated and composed of 2 distinct cell populations. The first population consisted of nonvacuolated, cuboidal to columnar, neoplastic epithelial cells with a moderate amount of eosinophilic cytoplasm. These cells were arranged in tubulopapillary and cystic growth patterns. These neoplastic cells exhibited moderate nuclear atypia and low mitotic activity. The second population was composed of vacuolated neoplastic polyhedral cells arranged in solid sheets
306 Case Reports Table 1. Primary antibody characteristics, tissue section pretreatment, and positive control tissues used in the current study. Antibody* Source Dilution Pretreatment Positive control tissue{ Cytokeratin (AE1/AE3) Dako a 1:200 121uC for 10 min in Vimentin (V9) Dako a 1:100 121uC for 10 min in Alpha smooth muscle actin (1A4) Dako a 1:400 121uC for 10 min in p63 (4A4) NeoMarkers b 1:200 121uC for 10 min in Estrogen receptor alpha (1D5) Dako a 1:50 105uC for 30 min in Androgen receptor (rabbit polyclonal) Abcam c Prediluted 121uC for 10 min in * Clone appears in parentheses. { Obtained from healthy Djungarian hamsters. Ovary Testis and nests or vacuolated cuboidal cells forming tubular and cystic structures. Cellular nuclei had an atypical appearance, including anisokaryosis, nuclear irregularity, and prominent nucleoli. Individual nuclei were often compressed and displaced to the periphery of the foamy or vacuolated cytoplasm. Mitotic figures were rare. Half of the primary neoplasm consisted of variably sized lobules of nonvacuolated neoplastic cells separated by thin to thick fibrous septa (Fig. 1). The other half of the tumor consisted predominantly of vacuolated neoplastic cells that often invaded the surrounding fibrous tissue. At the boundary between these 2 areas, a transition from nonvacuolated to vacuolated tumor cells was observed (Fig. 2). In contrast, the recurrent tumor was composed predominantly of vacuolated neoplastic cells admixed with a few nonvacuolated neoplastic cells. Compactly arranged neoplastic cells were divided into solid nests by thin fibrous septa, but occasional tubular structures also were observed (Figs. 3, 4). The vacuolated neoplastic cells often invaded the surrounding soft tissue. In the recurrent tumor, the cytoplasmic vacuoles within the neoplastic cells were often larger than those in the primary tumor. In tumor cell nests, cells at the periphery had a few small vacuoles, whereas cells in the central area contained larger vacuoles and often had a tendency to degenerate. In some areas, tubular and cystic structures were lined by foamy cuboidal epithelial cells with apical cytoplasmic blebs, suggesting apocrine secretion. The lumens contained a few desquamated epithelial cells, macrophages, and neutrophils. The stroma contained abundant foamy macrophages, a few multinucleated giant cells, and cholesterin crystal deposits. The foamy or vacuolated cytoplasm of the neoplastic cells stained positively with oil red O (Fig. 5), whereas AB and mucicarmine stainings were negative. The cytoplasm also stained red with Nile blue dye, indicating the presence of neutral lipids. The cytoplasmic vacuoles of the neoplastic cells were unreactive with PAS staining; however, the cytoplasm of nonvacuolated glandular neoplastic cells sometimes contained a few PAS-positive, diastase-resistant granules. Both nonvacuolated and vacuolated neoplastic cells were positive for cytokeratin (Fig. 6) and negative for vimentin, a-smooth muscle actin, p63, estrogen receptor a (ERa), and androgen receptor. According to the World Health Organization criteria, lipid-rich breast carcinoma in humans is defined as a breast carcinoma in which approximately 90% of the neoplastic cells contain abundant cytoplasmic neutral lipids. 4 Half of the primary tumor in the current case consisted of nonvacuolated neoplastic cells. This portion of the neoplasm had a similar histologic appearance to tubulopapillary carcinoma of a low grade of malignancy that has been described previously in Djungarian hamsters. 8 However, the other half of the primary and recurrent neoplasms consisted predominantly of vacuolated tumor cells that fulfilled the diagnostic criteria for lipid-rich carcinoma. Therefore, the primary and recurrent neoplasms were diagnosed as lipid-rich carcinoma arising in a tubulopapillary carcinoma and lipid-rich carcinoma, respectively. Although mucin-producing carcinomas, glycogenrich clear cell carcinomas, and secretory carcinomas may contain cytoplasmic vacuoles in the neoplastic cells, they tend to stain with AB, PAS, and/or mucicarmine. 4 In addition to lipid-rich carcinomas, sebaceous carcinomas of Figure 1. ; primary neoplasm, Djungarian hamster (Phodopus sungorus). The nonvacuolated neoplastic cells are arranged in tubular and cystic patterns. Bar 5 100 mm. Figure 2. ; primary neoplasm, Djungarian hamster (Phodopus sungorus). This area of the neoplasm is composed of nonvacuolated glandular (G) and vacuolated (V) neoplastic cells. The arrow indicates a transition from nonvacuolated glandular to vacuolated cells. Bar 5 100 mm. R
Case Reports 307 Figure 3. ; recurrent neoplasm, Djungarian hamster (Phodopus sungorus). Foamy, vacuolated neoplastic cells are arranged in solid sheets, whereas a small number of nonvacuolated glandular neoplastic cells form tubular structures. Bar 5 100 mm. Figure 4. ; recurrent neoplasm, Djungarian hamster (Phodopus sungorus). Higher magnification of Figure 3. A tubular structure is lined with both nonvacuolated cuboidal cells and foamy to vacuolated cells. Bar 5 30 mm. Figure 5. ; recurrent neoplasm, Djungarian hamster (Phodopus sungorus). The vacuoles of the tumor cells are oil red O positive, indicating the presence of lipid. Mayer hematoxylin counterstain. Bar 5 50 mm. Figure 6. ; recurrent neoplasm, Djungarian hamster. The vacuolated neoplastic cells are positive for cytokeratin. Immunohistochemistry, diaminobenzidine chromogen, Mayer hematoxylin counterstain. Bar 5 50 mm.
308 Case Reports the mammary gland contain cells that are positive for oil red O and negative for AB, PAS, and mucicarmine. However, sebaceous carcinoma differs from lipid-rich carcinoma in that there is a second cell population of smaller, oval to spindle, reserve cells. These reserve cells contain a small amount of eosinophilic cytoplasm and lack vacuoles. 4 Human lipid-rich carcinoma has a more aggressive clinical course and worse prognosis than other types of breast cancer. 15,17 In dogs, lipid-rich carcinoma has clinical features of inflammatory mammary carcinoma 12 and is associated with recurrence and metastases. 5 In the present case, the invasiveness and atypia of the neoplastic cells were greater than those of mammary tubulopapillary carcinomas in the Djungarian hamster, 8 and the neoplasm recurred. Regarding the aggressiveness of lipid-rich cancer cells, it is hypothesized that the incorporation of excess free fatty acids onto the neoplastic cell membrane induces membrane phase separation, reduces cell cell contact, increases surface adhesion, and promotes tissue invasion. 9 Human lipid-rich carcinomas are considered to be poorly differentiated neoplasms that do not display a glandular or papillary growth pattern. 17 Although the vacuolated cells of the present neoplasm were sometimes arranged in a tubular or cystic growth pattern, lipid-rich carcinomas in the dog and cat are also composed of tubular or tubuloacinar structures formed by lipid-rich neoplastic cells. 6,13 A recent study showed that apocrine secretion by the neoplastic luminal epithelial cells was observed in all 12 reported cases of mammary neoplasms (adenomas and/or carcinomas) in Djungarian hamsters. 8 The recurrent tumor in the present case had multifocal tubular or cystic structures lined by foamy, neutral, lipid-rich neoplastic cells suggesting apocrine secretion. These areas were similar to those of a peculiar type of human lipid-rich carcinoma described previously. 19 Generally, the absence of ERa in human and canine mammary carcinomas is related to an unfavorable prognosis. 4,11 No ERa expression was detected in any of 49 human patients with lipid-rich carcinoma. 17 In addition, all 10 lipid-rich carcinomas reported in dogs also lacked ERa expression, 5,12 as did the neoplasm in the Djungarian hamster of the current study. Androgen receptor expression in mammary tumors of Djungarian hamsters is associated with apocrine differentiation of the neoplastic cells, 8 although this was not observed in the present neoplasm. As indicated by the results of immunohistochemistry for a- smooth muscle actin and p63, myoepithelial cells were not detected among the vacuolated tumor cells of the primary and recurrent lipid-rich carcinomas in the current study. A case report of a presumed lipid-rich carcinoma in the breast of a human male patient also has been published. 10 Mammary tumors also may occur in both male and female Djungarian hamsters, 7,8 although the present lipid-rich carcinoma occurred in a female. All reported lipid-rich carcinomas in the dog and cat also have been described solely in females. 5,6,12,13 Little is known about the histogenesis of lipid-rich carcinoma. Most human lipid-rich carcinomas are characterized histologically as grade 3 invasive carcinomas, which may be associated with in situ lobular or ductal carcinoma. 1,4,15 In the primary mammary neoplasm in this hamster, the boundary between the nonvacuolated glandular and vacuolated neoplastic cells exhibited a transition favoring the formation of vacuolated cells. This indicates that the lipid-rich carcinoma in this hamster may have arisen from a preexisting mammary adenocarcinoma of low grade of malignancy. Acknowledgements. This work was supported in part by the Academic Frontier Project for Private Universities: matching fund subsidy (2005 2009) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT). Sources and manufacturers a. Dako Denmark A/S, Glostrup, Denmark. b. NeoMarkers Inc., Fremont, CA. c. Abcam Inc., Cambridge, MA References 1. Aboumrad MH, Horn RC Jr, Fine G: 1963, Lipid-secreting mammary carcinoma. Report of a case associated with Paget s disease of the nipple. Cancer 16:521 525. 2. Baba Y, Takahashi K, Nakamura S: 2003, Androgen-dependent atypical fibromas spontaneously arising in the skin of Djungarian hamsters (Phodopus sungorus). Comp Med 53:527 531. 3. Cooper JE, Knowler C, Pearson AJ: 1991, Tumours in Russian hamsters (Phodopus sungorus). Vet Rec 128:335 336. 4. Ellis IO, Schnitt SJ, Sastre-Garau X, et al.: 2003, Invasive breast carcinoma. In: Pathology & genetics: tumours of the breast and female genital organs, World Health Organization classification of tumours series, ed. Tavassoli FA, Devilee P, pp. 13 59. IARC Press, Lyon, France. 5. Espinosa de los Monteros A, Hellmén E, Ramírez GA, et al.: 2003, Lipid-rich carcinomas of the mammary gland in seven dogs: clinicopathologic and immunohistochemical features. Vet Pathol 40:718 723. 6. Kamstock DA, Fredrickson R, Ehrhart EJ: 2005, Lipid-rich carcinoma of the mammary gland in a cat. Vet Pathol 42:360 362. 7. Kondo H, Onuma M, Shibuya H, Sato T: 2008, Spontaneous tumors in domestic hamsters. Vet Pathol 45:674 680. 8. Kondo H, Onuma M, Shibuya H, Sato T: 2009, Morphological and immunohistochemical studies of spontaneous mammary tumours in Siberian hamsters (Phodopus sungorus). J Comp Pathol 140:127 131. 9. Le TT, Huff TB, Cheng JX: 2009, Coherent anti-stokes Raman scattering imaging of lipids in cancer metastasis. BMC Cancer 9:42. 10. Mazzella FM, Sieber SC, Braza F: 1995, Ductal carcinoma of male breast with prominent lipid-rich component. Pathology 27:280 283. 11. Misdorp W, Else RW, Hellmen E, et al.: 1999, Histological classification of mammary tumors of the dogs and the cats. Armed Forces Institute of Pathology and the World Health Organization Collaborating Center for Worldwide Reference on Comparative Oncology, Washington, DC. 12. Peña L, Perez-Alenza MD, Rodriguez-Bertos A, Nieto A: 2003, Canine inflammatory mammary carcinoma: histopathology, immunohistochemistry and clinical implications of 21 cases. Breast Cancer Res Treat 78:141 148. 13. Pérez-Martínez C, García-Iglesias MJ, Durán-Navarrete AJ, et al.: 2005, Histopathological and immunohistochemical
Case Reports 309 characteristics of two canine lipid-rich mammary carcinomas. J Vet Med A Physiol Pathol Clin Med 52:61 66. 14. Pogosianz HE: 1975, Djungarian hamster a suitable tool for cancer research and cytogenetic studies. J Natl Cancer Inst 54:659 664. 15. Ramos CV, Taylor HB: 1974, Lipid-rich carcinoma of the breast. A clinicopathologic analysis of 13 examples. Cancer 33:812 819. 16. Rosen PP, Oberman HA: 1992. Atlas of tumor pathology, 3rd series, fascicle 7: tumors of the mammary gland. Armed Forces Institute of Pathology, Washington, DC. 17. Shi P, Wang M, Zhang Q, Sun J: 2008, Lipid-rich carcinoma of the breast. A clinicopathological study of 49 cases. Tumori 94:342 346. 18. Sokova OI, Prigozhina EL, Pogosiants EE: 1975, [Characteristics of the reproduction of Djungarian hamsters in captivity and spontaneous tumors]. Vopr Onkol 21:32 37. In Russian. Abstract in English. 19. van Bogaert LJ, Maldague P: 1977, Histologic variants of lipid-secreting carcinoma of the breast. Virchows Arch A Pathol Anat Histol 375:345 353.