The determination of estrogen (ER) and progesterone receptor

The Utility of Estrogen Receptor and Progesterone Receptor Immunohistochemistry in the Distinction of Metastatic Breast Carcinoma From Other Tumors in the Liver Jason W. Nash, DO; Carl Morrison, MD; Wendy L. Frankel, MD Context. The distinction of metastatic breast adenocarcinoma (MBA) to the liver from hepatocellular carcinoma (HCC), cholangiocarcinoma (CC), and metastatic adenocarcinoma from other sites may require the use of immunohistochemistry. The use of antibodies directed against estrogen receptor (ER) and progesterone receptor (PR) has been suggested to help make this distinction. Objective. To examine the utility of ER and PR immunohistochemistry in the distinction of MBA from HCC, CC, and other metastatic adenocarcinomas in the liver. Methods. Ninety-two previously characterized hepatic neoplasms were identified, including HCC (n 14), CC (n 16), and metastatic tumors from breast (n 17), colorectal (n 14), pancreatic (n 15), and esophageal/gastric (n 16) origins. For all cases of metastatic tumor, the primary tumor was reviewed to verify the diagnosis. All tumors were graded as well, moderately, or poorly differentiated. Estrogen receptor and PR immunohistochemical staining was performed on all cases and evaluated by 2 pathologists. Results. Immunoreactivity for ER was identified only in MBA, with 6 (35%) of 17 cases positive. Positive immunoreactivity for PR was not restricted to MBA, but was seen in HCC, CC, esophageal/gastric, and pancreatic metastases. Positive immunostaining with PR was nearly as frequent in poorly differentiated carcinomas of nonbreast origin (3/16 cases, 19%) as in poorly differentiated breast carcinomas (2/8 cases, 25%). Conclusion. Progesterone receptor exhibited poor specificity and sensitivity for the distinction of MBA from HCC, CC, and other metastatic adenocarcinomas. Estrogen receptor exhibited poor sensitivity for MBA, although the specificity was good. The finding that PR positivity was present with a similar frequency in poorly differentiated tumors of breast and nonbreast origin limits the usefulness of this marker. Therefore, ER and PR staining have limited utility in the distinction of MBA from HCC, CC, and other metastatic adenocarcinomas in the liver. (Arch Pathol Lab Med. 2003;127:1591 1595) The determination of estrogen (ER) and progesterone receptor (PR) status has become part of the standard workup for breast carcinoma. Information regarding steroid receptor status is used therapeutically and prognostically. Some pathologists in clinical practice have attempted to use immunohistochemistry for ER and PR to determine whether breast is the site of origin for metastatic adenocarcinoma to the liver. Estrogen receptors have been documented in nonneoplastic tissues other than breast, including liver, pancreas, ovary, stomach, skin, and colon. 1 4 Many neoplasms have been shown to express ERs, including breast carcinoma; hepatocellular carcinoma (HCC); cerebral meningioma; non small cell lung carcinoma; papillary thyroid carcinoma; pancreatic, colonic, and gastric adenocarcinoma; and skin adnexal tumors. 4 14 Progesterone receptors have also been reported in tissues other than breast. 6 9,13,15 Accepted for publication July 24, 2003. From the Department of Pathology, The Ohio State University, Columbus. Reprints: Wendy L. Frankel, MD, Department of Pathology, E-401 Doan Hall, 410 W Tenth Ave, Columbus, OH 43210-1228 (e-mail: frankel-1@medctr.osu.edu). Given the wide distribution of ER and PR in a variety of neoplastic and nonneoplastic tissues, the utility of these markers may be limited in distinguishing metastatic breast adenocarcinoma (MBA) from tumors of nonbreast origin in the liver. We examined the immunoreactivity with ER and PR of HCC and metastatic adenocarcinomas to the liver from various sites, including breast, to evaluate the usefulness of these stains in determining the site of origin for tumors in the liver. MATERIALS AND METHODS The archival files of the Department of Pathology, The Ohio State University, Columbus, were searched, and 92 previously characterized hepatic neoplasms were identified, including HCC (n 14), cholangiocarcinoma (CC) (n 16), and metastatic adenocarcinomas from breast (n 17), colorectal (n 14), pancreatic (n 15), and esophageal/gastric (n 16) origins. All breast adenocarcinomas were of infiltrating ductal type. Consecutive cases were selected in a retrospective fashion from 1987 to 2000. For all cases of metastatic tumor, the primary tumor was reviewed to verify the original diagnosis. Immunoperoxidase staining was performed on formalin-fixed, paraffin-embedded tissue cut at 4 m and placed on positively charged slides. Slides were then placed in a 60 C oven for 1 hour, cooled, and deparaffinized and rehydrated through xylenes and Arch Pathol Lab Med Vol 127, December 2003 ER/PR Immunohistochemistry in Metastatic Breast Cancer Nash et al 1591

Table 1. Estrogen Receptor (ER) and Progesterone Receptor (PR) Immunoreactivity in Tumors in the Liver Tumor Type ER, No. (%) PR, No. (%) Breast (n 17) 6 (35) 5 (29) Hepatocellular carcinoma (n 14) 0 1 (7) Biliary (n 16) 0 5 (31) Esophageal/gastric (n 16) 0 2 (13) Colorectal (n 14) 0 0 Pancreatic (n 15) 0 1 (6) graded ethanol solutions to water. All slides were quenched for 5 minutes in a 3% hydrogen peroxide solution in methanol to block for endogenous peroxidase. Antigen retrieval was performed by a heat method in which the specimens were placed in a citric acid solution (Target Retrieval Solution, ph 6.1; Dako Cytomation, Carpinteria, Calif), for 30 minutes at 94 C using a vegetable steamer. Slides were then placed on a Dako Autostainer for use with immunohistochemistry and stained with the ER (clone 1D5, 1:50, Dako Cytomation A/S, Glostrup, Denmark) and PR (clone 1A6:1:10, Dako Cytomation, Carpinteria, Calif) antibodies. The detection system used was a labeled streptavidinbiotin complex. This method is based on the consecutive application of (1) a primary antibody against the antigen to be localized, (2) a biotinylated linking antibody, (3) enzyme-conjugated streptavidin, and (4) substrate chromogen (diaminobenzidine). Tissues were avidin and biotin blocked prior to the application of the biotinylated secondary reagent. Slides were then counterstained in Richard Allen hematoxylin, dehydrated through graded ethanol solutions, and cover-slipped. The positive control was a known ER- and PR-positive infiltrating ductal breast adenocarcinoma in the liver. For the negative control, nonimmune mouse immunoglobulin G1 was used in place of the primary antibody. The positive and negative controls stained appropriately. Nuclear staining for ER and PR was qualitatively recorded as positive or negative. Greater than 1% nuclear staining was considered positive. Additionally, the tumors were evaluated on hematoxylin-eosin stained sections, and each was graded as well, moderately, or poorly differentiated. All slides were reviewed by 2 pathologists. RESULTS Table 1 shows ER and PR immunoreactivity data for all the tumors studied. Estrogen receptor immunoreactivity was seen exclusively in MBA, but only 6 (35%) of 17 cases of MBA demonstrated immunoreactivity with ER. Figure 1 shows an MBA that was positive for ER and negative for PR. Metastatic breast adenocarcinoma, as well as CC, HCC, and metastatic adenocarcinoma from esophagus, stomach, and pancreas, showed immunoreactivity with PR in some cases. Figure 2 shows a poorly differentiated gastric adenocarcinoma that was positive with PR and negative with ER. When subdivided according to degree of differentiation, positive staining with PR was nearly as frequent in poorly differentiated metastatic adenocarcinomas (3/16, 19%) as in poorly differentiated MBAs (2/ 8, 25%; Table 2). COMMENT In most cases, the site of origin for tumors in the liver can be determined by clinical history and radiographic findings. In some instances, the pathologist may be called on to help determine the primary site for metastatic tumors in the liver. The determination of site of origin based on morphologic findings can be difficult in some cases, particularly with small biopsies. Metastatic breast adenocarcinoma commonly involves the liver, and the histologic distinction of breast adenocarcinoma from other metastatic tumors in the liver can be challenging when the patient has no known history of breast adenocarcinoma, or when a previously diagnosed breast cancer is not available for comparison. Estrogen receptor and PR immunohistochemistry has been used by some in clinical practice to establish the diagnosis of MBA to the liver. Positive staining with either or both steroid receptor markers may be interpreted by some as consistent with breast origin. Estrogen and progesterone receptors have been documented in numerous locations throughout the body, including breast, ovary, uterus, colon, esophagus, stomach, liver, pancreas, skin, pituitary, and bone. 2,3,6 9,13,15 Receptors in these sites serve a variety of physiologic functions. Estrogen receptors and PRs have also been documented in many neoplasms from various locations, including breast, liver, pancreas, colon, ovary, skin adnexa, and uterus. 5 14,16 Recently ERs have been subdivided into and subtypes. 2,17 The function of the subtype is unclear, with most of the physiologic effects of estrogen being attributed to its activity at the receptor. Currently, the immunohistochemical stains used are specific for the subtype. The presence of physiologic hormone receptors in multiple anatomic sites, as well as their expression in different neoplasms, compounds the difficulty of relying on their presence or absence (as determined by immunohistochemical methods) in metastatic lesions to delineate tumor site of origin. In breast cancer, the ER/PR status of the tumor is useful for both prognosis and therapy, with more chemotherapeutic options available to patients with hormone receptor positive tumors. 18,19 Breast adenocarcinoma has been shown to be positive for ER in 24% to 63% of cases and positive for PR in 9% to 37% of cases. 14,20,21 Breast adenocarcinoma may demonstrate immunophenotypic variability in its expression of ER and PR, with differences dependent on histologic grade, histologic subtype, antibody clone applied, and immunohistochemical techniques used. These factors limit the sensitivity of these markers for excluding MBA in cases of unknown primary site. We investigated the utility of ER and PR immunohistochemical staining in tumors in the liver, including HCC, CC, and metastatic adenocarcinoma from several sites, such as breast, esophagus/stomach, colorectum, and pancreas. Our series of cases showed excellent specificity of ER immunostaining for MBA when compared to extramammary metastatic carcinomas in the liver and HCC regardless of the degree of differentiation. However, ER exhibited poor sensitivity for MBA, staining only 35% of these cases. Progesterone receptor staining was neither sensitive nor specific, and decorated metastatic lesions of both breast and nonbreast origin. The determination of tumor site of origin is the most challenging in poorly differentiated tumors; therefore, immunohistochemistry tends to be used most often in these cases. In our series of cases, PR staining was seen in poorly differentiated tumors of breast, colorectal, esophageal/ gastric, and hepatocellular origin. In the HCC cases, none of the well-differentiated to moderately differentiated tumors showed PR staining, while 25% of the poorly differentiated cases showed PR positivity. Nearly 50% of the MBAs we examined were poorly differentiated, and of these, only 25% showed PR immunoreactivity. Progesterone receptor staining was nearly as frequent in poorly dif- 1592 Arch Pathol Lab Med Vol 127, December 2003 ER/PR Immunohistochemistry in Metastatic Breast Cancer Nash et al

Figure 1. A, Moderately differentiated breast adenocarcinoma metastatic to the liver (hematoxylin-eosin, original magnification 100). B, Nuclear immunopositivity for estrogen receptor in metastatic breast adenocarcinoma (original magnification 200). C, Metastatic breast adenocarcinoma showing negativity for progesterone receptor (original magnification 200). Figure 2. A, Poorly differentiated gastric adenocarcinoma metastatic to the liver (hematoxylin-eosin, original magnification 100). B, Poorly differentiated gastric adenocarcinoma showing nuclear positivity for progesterone receptor in the majority of cells (original magnification 200). C, Poorly differentiated gastric adenocarcinoma negative for estrogen receptor (original magnification 200). Arch Pathol Lab Med Vol 127, December 2003 ER/PR Immunohistochemistry in Metastatic Breast Cancer Nash et al 1593

Table 2. Progesterone Receptor (PR) Immunostaining in Poorly Differentiated Tumors in the Liver Tumor Type Poorly Differentiated Tumors, No. (%) PR Staining in Poorly Differentiated Tumors, No. (%) Breast (n 17) 8 (47) 2 (25) Hepatocellular carcinoma (n 14) 4 (29) 1 (25) Biliary (n 16) 2 (13) 1 (50) Esophageal/gastric (n 16) 4 (25) 1 (25) Colorectal (n 14) 1 (7) 0 Pancreatic (n 15) 5 (33) 0 ferentiated nonbreast carcinomas (3/16, 19%) as it was in poorly differentiated MBAs (2/8, 25%). The lack of specificity with PR in poorly differentiated tumors severely limits the utility of PR as a marker of MBA, particularly for those poorly differentiated cases in which this distinction is most important. Some investigators have concluded that ER and PR staining in MBA is fairly specific, yet lacks sensitivity. 14,20 23 However, steroid receptor positivity has been documented in a number of other nonmammary tumors. As many as 28% (30/108) of gastric carcinomas in one study were immunoreactive with ER. 12 In the same study, it was noted that the predominant grade of gastric carcinoma staining for ER was poorly differentiated. 12 Colorectal adenocarcinoma has also been shown to mark with ER in up to 46% of cases. 10,15 Estrogen receptor and PR immunopositivity in colorectal adenocarcinomas has been shown to positively correlate with higher stage disease. 15 In our study, we did not demonstrate ER positivity in any of the 16 esophageal/gastric or 14 colorectal carcinomas. This is in contrast to the findings of some earlier studies, which showed 28% and up to 46% ER staining in gastric and colorectal carcinomas, respectively. 10,12,15 Differences in methodology likely explain some of the apparent discrepancies in staining results. Discrepancies may be encountered with the use of ER and PR immunohistochemistry owing to differences in tissue processing/preparation, antigen retrieval techniques, and the antibody clone applied. Some variability may be encountered with different histologic grades of tumors. One report used radiolabeled receptor-ligand assays to quantify ER cytosolic and/or nuclear positivity, 10 while another used a different anti-er antibody clone (H222) than was used in our study. 12 In a study of the expression of ER immunopositivity of primary lung adenocarcinomas, ER was noted to be present in many lung adenocarcinomas when using the 6F11 (Ventana Medical Systems, Tucson, Ariz) clone, while no ER staining was noted using the 1D5 (Dako) clone. 24 One study that used methods and antibody clones that were apparently similar to ours noted some ER positivity in colorectal carcinomas, whereas we did not. 15 This discrepancy could be due to differences in tumor grade. The variability in the findings with different antibody clones and tumor grades highlights the potential hazards in the reliance on a single antibody such as ER to be highly specific for MBA. Immunohistochemistry for ER and PR has been demonstrated to be helpful in distinguishing the site of tumor origin when used together with other markers. Estrogen and progesterone receptors were shown to be helpful when added to a panel of immunomarkers, including thyroid transcription factor-1 for the distinction of breast versus lung adenocarcinoma in pleural- or pericardial-derived effusion specimens. 25 The utility of ER and PR has been documented when used in a larger panel of immunomarkers, including gross cystic disease fluid protein-15 (GCDFP-15), S100 protein, carcinoembryonic antigen, placental alkaline phosphatase, CD15, epithelial membrane antigen, and cytokeratins (AE1/AE3, MAK-6, CAM 5.2) for distinguishing MBA from primary pulmonary neoplasms in a study of 30 cases of primary and metastatic adenocarcinoma in the lung. 26 For the differentiation of metastatic adenocarcinoma to the brain, ER and GCDFP- 15 have been shown to be specific, but not sensitive for breast carcinoma, while PR demonstrated good sensitivity and lacked specificity. 23 In a study comparing MBA and other metastatic adenocarcinomas of unknown primary sites, immunostaining for GCDFP-15 in conjunction with ER and/or PR positivity allowed greater sensitivity (83%) and specificity (93%) for adenocarcinomas of the breast versus bronchogenic, pancreatic, colonic, gastric, renal, and ovarian adenocarcinomas than ER and/or PR alone. 20 Immunohistochemistry can be very useful to help determine the site of origin in some liver tumors when panels of immunomarkers are used. However, results should be interpreted cautiously and in conjunction with clinical and radiographic data. Not all MBAs in the liver display immunoreactivity with ER and/or PR, and adenocarcinomas of extramammary origin can show ER and PR positivity. The distinction of poorly differentiated tumors in the liver creates the greatest diagnostic challenge, and immunoreactivity with ER and/or PR should not be relied on in these cases to distinguish site of origin for metastatic tumors to the liver. References 1. Molteni A, Bahu RM, Battifora HA, et al. Estradiol receptor assays in normal and neoplastic tissues: a possible diagnostic acid for tumor differentiation. Ann Clin Lab Sci. 1979;9:103 108. 2. Enmark E, Gustafsson JA. Oestrogen receptors: an overview. J Intern Med. 1999;246:133 138. 3. Thornton MJ. The biological actions of estrogens on skin. Exp Dermatol. 2002;11:487 502. 4. Kruijver FP, Balesar R, Espila AM, Unmehopa UA, Swaab DF. Estrogen receptor-alpha distribution in the human hypothalamus in relation to sex and endocrine status. J Comp Neurol. 2002;454:115 139. 5. Nagasue N, Yu L, Yukaya H, Kohno H, Nakamura T. Androgen and oestrogen receptors in hepatocellular carcinoma and surrounding liver parenchyma: impact on intrahepatic recurrence after hepatic resection. Br J Surg. 1995;82:542 547. 6. Kvasha MS, Zakhartsev LM. Estrogen and progesterone receptors in cerebral meningioma and their clinical value. Klin Khir. 2002;8:50 52. 7. Lewy-Trenda I. Estrogen and progesterone receptors in neoplastic and nonneoplastic thyroid lesions. Pol J Pathol. 2002;53:67 72. 8. Radzikowska E, Langfort R, Giedronowicz D. Estrogen and progesterone receptors in non small cell lung cancer patients. Ann Thorac Cardiovasc Surg. 2002;8:69 73. 9. Yeh TS, Jan YY, Chiu CT, et al. Characterization of oestrogen receptor, progesterone receptor, trefoil factor 1, and epidermal growth factor and its receptor in pancreatic cystic neoplasms and pancreatic ductal adenocarcinoma. Gut. 2002;51:712 716. 10. Francavilla A, Di Leo A, Polimeno L, et al. Nuclear and cytosolic estrogen receptors in human colon carcinoma and in surrounding noncancerous colonic tissue. Gastroenterology. 1987;93:1301 1306. 11. Linsalata M, Messa C, Russo F, Cavallini A, Di Leo A. Estrogen receptors and polyamine levels in human gastric carcinoma. Scand J Gastroenterol. 1994; 29:67 70. 12. Yokozaki H, Takekura N, Takanashi A, Tabuchi J, Haruta R, Tahara E. Estrogen receptors in gastric adenocarcinoma: a retrospective immunohistochemical analysis. Virchows Arch A Pathol Anat Histopathol. 1988;413:297 302. 13. Daniel SJ, Nader R, Kost K, Huttner I. Facial sweat gland carcinoma metastasizing to neck nodes: a diagnostic and therapeutic challenge. Arch Otolaryngol Head Neck Surg. 2001;127:1495 1498. 1594 Arch Pathol Lab Med Vol 127, December 2003 ER/PR Immunohistochemistry in Metastatic Breast Cancer Nash et al

14. Busam KJ, Tan LK, Granter SR, et al. Epidermal growth factor, estrogen, and progesterone receptor expression in primary sweat gland carcinomas and primary and metastatic mammary carcinomas. Mod Pathol. 1999;12:786 793. 15. Kaklamanos IG, Bathe OF, Franceschi D, et al. Expression of receptors for estrogen and progesterone in malignant colonic mucosa as a prognostic factor for patient survival. J Surg Oncol. 1999;72:225 229. 16. Dahmoun M, Backstrom T, Boman K, Cajander S. Apoptosis, proliferation, and hormone receptors in endometrial carcinoma: results depending on methods of analysis. Int J Oncol. 2003;22:115 122. 17. Harris HA, Katzenellenbogen JA, Katzenellenbogen BS. Characterization of the biological roles of the estrogen receptors, ERalpha and ERbeta, in estrogen target tissues in vivo through the use of an ERalpha-selective ligand. Endocrinology. 2002;143:4172 4177. 18. Bejar J, Sabo E, Eldar S, Lev M, Misselevich I, Boss JH. The prognostic significance of the semiquantitatively determined estrogen receptor content of breast carcinomas: a clinicopathological study. Pathol Res Pract. 2002;198:455 460. 19. Lamy PJ, Pujol P, Thezenas S, et al. Progesterone receptor quantification as a strong prognostic determinant in postmenopausal breast cancer women under tamoxifen therapy. Breast Cancer Res Treat. 2002;76:65 71. 20. Kaufmann O, Deidesheimer T, Muehlenberg M, Deicke P, Dietel M. Immunohistochemical differentiation of metastatic breast carcinomas from metastatic adenocarcinomas of other common primary sites. Histopathology. 1996; 29:233 240. 21. Kaufmann O, Kother S, Dietel M. Use of antibodies against estrogen and progesterone receptors to identify metastatic breast and ovarian carcinomas by conventional immunohistochemical and tyramide signal amplification methods. Mod Pathol. 1998;11:357 363. 22. Brown RW, Campagna LB, Dunn JK, Cagle PT. Immunohistochemical identification of tumor markers in metastatic adenocarcinoma: a diagnostic adjunct in the determination of primary site. Am J Clin Pathol. 1997;107:12 19. 23. Perry A, Parisi JE, Kurtin PJ. Metastatic adenocarcinoma to the brain: an immunohistochemical approach. Hum Pathol. 1997;28:938 943. 24. Dabbs DJ, Landreneau RJ, Liu Y, et al. Detection of estrogen receptor by immunohistochemistry in pulmonary adenocarcinoma. Ann Thorac Surg. 2002; 73:403 405. 25. Lee BH, Hecht JL, Pinkus JL, Pinkus GS. WT1, estrogen receptor, and progesterone receptor as markers for breast or ovarian primary sites in metastatic adenocarcinoma to body fluids. Am J Clin Pathol. 2002;117:745 750. 26. Raab SS, Berg LC, Swanson PE, Wick MR. Adenocarcinoma in the lung in patients with breast cancer: a prospective analysis of the discriminatory value of immunohistology. Am J Clin Pathol. 1993;100:27 35. Arch Pathol Lab Med Vol 127, December 2003 ER/PR Immunohistochemistry in Metastatic Breast Cancer Nash et al 1595