Secondary Logo

Journal Logo

Original Articles

Expression of Pax8 as a Useful Marker in Distinguishing Ovarian Carcinomas From Mammary Carcinomas

Nonaka, Daisuke MD*; Chiriboga, Luis PhD*; Soslow, Robert A. MD

Author Information
The American Journal of Surgical Pathology: October 2008 - Volume 32 - Issue 10 - p 1566-1571
doi: 10.1097/PAS.0b013e31816d71ad
  • Free


Ovarian metastasis is common in patients with breast cancer,27 and can be seen in 22% to 40% of prophylactic oophorectomy specimens from patients with advanced breast cancer.16,18,20,31 Ovarian metastasis is rarely found before a diagnosis of breast tumor, and such a case may clinically mimic a primary ovarian tumor.48 Additionally, patients with breast carcinomas have an increased risk of developing second malignant tumors, among which ovarian carcinoma is one of the most common.8 The frequency increases with patients who are BRCA mutation carriers.23 Conversely ovarian carcinomas can rarely metastasize to axillary lymph nodes and even mammary glands.9,32 Both ovarian and breast carcinomas share similar morphologic and immunohistochemical features. In problematic cases, a battery of immunohistochemistry stains are useful for a determination of the primary site.41 Pax8 is a crucial transcription factor for organogenesis of the thyroid gland, kidney, and Müllerian system, and it also regulates Wilms tumor suppressor gene (WT1) expression.36 A recent DNA microarray study reported Pax8 expression to be one of the best discriminatory markers between ovarian and breast carcinomas.34 We applied Pax8 immunostain on a large number of both ovarian and breast carcinomas to investigate the diagnostic utility.


The following cases of ovarian surface epithelial carcinomas and invasive breast carcinomas were studied: 124 cases of ovarian surface epithelial carcinomas including 84 serous papillary carcinomas, 18 endometrioid carcinomas, 10 clear cell carcinomas, and 12 mucinous carcinomas; 243 cases of invasive breast carcinomas including 178 ductal carcinomas and 65 lobular carcinomas. Among the 178 cases of invasive ductal carcinomas, there were 16 cases of grade 1, 109 cases of grade 2, and 53 cases of grade 3. Histologic diagnosis was based on recent classifications.38,39

Formalin-fixed, paraffin-embedded tissues of all the above-mentioned cases were used for tissue microarray construction. Tissue microarrays were assembled using a Manual Tissue Arrayer I (Beecher Instruments, Sun Prairie, WI). A representative area of each tumor was identified on the conventional sections, and 1 to 3 cylinders per tissue were arrayed using a punch biopsy (needles with a diameter of 0.6 mm).

Immunohistochemical studies were performed on all the above-mentioned tissues using anti-Pax8 (rabbit polyclonal, Protein Tech Group Inc, Chicago, IL) and WT1 (mouse monoclonal 6F-H2, Dako, Carpinteria, CA).

The sections were deparaffinized in xylene (3 changes), rehydrated through graded alcohols (3 changes 100% ethanol, 3 changes 95% ethanol), and rinsed in distilled water. Heat-induced epitope retrieval was performed in a 1200-Watt microwave oven at 90% power in 10 mM citrate buffer of pH 6.0 for 10 minutes. Sections were allowed to cool for 30 minutes and then rinsed in distilled water. Antibody incubations and detection were carried out at 37°C on a NEXes instrument (Ventana Medical Systems Tucson, AZ) using Ventana's reagent buffer and detection kits unless otherwise noted. Endogenous peroxidase activity was blocked with hydrogen peroxide. Antibodies against Pax8 and WT1 were diluted 1:25 and 1:100, respectively. These antibodies were incubated overnight at room temperature, and were detected with Ventana's biotinylated goat antimouse secondary. After secondary antibody application, streptavidin-horseradish-peroxidase conjugate was applied. The complex was visualized with 3,3′ diaminobenzidene and enhanced with copper sulfate. Slides were washed in distilled water, counterstained with hematoxylin, dehydrated, and mounted with permanent media. Appropriate positive and negative controls were included with the study sections.

The extent of nuclear staining was graded as follows: 1+, 1% to 25%; 2+, 25% to 50%; 3+, 50% to 75%; 4+, ≥75%.


The results of the immunohistochemistry stains are summarized in Table 1. Pax8 was expressed in 108 of 124 ovarian carcinomas (87.1%) and 107 of 112 nonmucinous ovarian carcinomas (95.5%), whereas WT1 was expressed in 78 of 124 ovarian carcinomas (62.9%), and 78 of 112 nonmucinous ovarian carcinomas (69.6%).

Immunohistochemistry Results for Ovarian Carcinomas and Mammary Carcinomas

In ovarian serous papillary carcinoma (Fig. 1A), Pax8 nuclear expression was found in 81 of 84 cases (96.4%), with 3+ and 4+ in 79 cases (94%) (Fig. 1B), whereas WT1 expression was seen in 73 of 84 cases (86.9%), with 3+ and 4+ in 59 cases (70.2%) (Fig. 1C). All 3 cases of Pax8-negative serous papillary carcinoma were also negative for WT1. On the other hand, 11 WT1 negative serous papillary carcinomas comprised 3 Pax8-negative tumors, and 8 Pax8-positive tumors with 3+ in 2 cases and 4+ in 6 cases.

Ovarian carcinomas. Serous papillary carcinoma: A, Hematoxylin and eosin (H&E) stain. B, Diffuse Pax8 nuclear reaction. C, Diffuse WT1 nuclear staining. Endometrioid carcinoma: D, H&E stain. E, positive Pax8 reaction. F, Negative WT1 reaction. Clear cell carcinoma: G, H&E stain. H, Positive Pax8 reaction. I, Negative WT1 reaction. Mucinous carcinoma: J, H&E stain. K, Negative Pax8 reaction. L, Negative WT1 reaction.

The sensitivity of Pax8 was much higher in endometrioid carcinoma (Fig. 1E) and clear cell carcinoma of the ovary (Fig. 1H) with 88.9% and 100%, respectively than that of WT1 (Figs. 1F, I) with 27.8%, and 0%, respectively. In clear cell carcinoma (Fig. 1H), almost all tumor cells uniformly expressed Pax8 whereas WT1 expression was completely negative (Fig. 1I). In endometrioid carcinoma (Fig. 1D), Pax8 expression was more variable, ranging from merely focal to diffuse (Fig. 1E) so was WT1 (Fig. 1F). In mucinous carcinoma (Fig. 1J) both Pax8 and WT1 were negative (Fig. 1K, L) except for 1 case showing Pax8 (2+), WT1 (−) pattern. In mammary carcinomas, Pax8 was negative in all the cases of both invasive ductal and lobular carcinomas whereas WT1 was positive in 5 of 243 invasive carcinomas (2.1%), all of which were of ductal type (Fig. 2). Among the 5 WT1-positive invasive ductal carcinomas, 2 cases were grade 3 and 3 cases were grade 2.

Invasive ductal carcinoma of the mammary gland showing focal WT1 nuclear reaction.


Metastatic tumors to the ovaries represent 6% to 7% of ovarian neoplasms,43 and among the metastatic tumors of nongynecologic origin, the tumors of the mammary gland are common and account for 8.5%, second to the tumors of the colon and appendix.27 Metastasis of breast carcinoma to the ovaries can be seen in 22% to 40% of prophylactic oophorectomy specimens from patients with advanced breast cancer.5,16,18,20,31 Lobular carcinomas spread to the ovary more frequently than ductal carcinoma, however, given that ductal carcinoma is more common, overall frequency of metastasis is 3 to 6 times higher in ductal carcinoma.7,11

On the other hand, ovarian carcinomas can rarely metastasize to the mammary glands and axillary lymph nodes.32 The most common histotype is serous carcinoma,32 but metastases of other categories of ovarian tumors such as endometrioid carcinoma, germ cell tumor, and sex-cord stromal tumors, have also been reported.6,12,14,22 Those patients with such metastasis usually have a high-stage disease at initial presentation. The metastatic tumor can be found concurrently or after a period of time with an average of 30 months after the presentation of ovarian carcinoma, and can present as either single or multiple lesions. Cases of a single metastasis or a concurrent ovarian tumor with metastasis can clinically simulate a carcinoma of mammary gland origin.32

Moreover, carcinomas of the ovary and mammary gland have similar morphologic features, and particularly serous carcinoma of the ovary, which is more prone to metastasis, may be histopathologically difficult to distinguish from micropapillary carcinoma of the mammary gland.32 In patients with BRCA1 mutations, who are at a greatly increased risk of developing both breast and ovarian carcinomas, the breast and ovarian carcinomas are both typically characterized by high architectural and nuclear grade with frequent mitotic activity.45 Both ovarian nonmucinous surface epithelial carcinoma and breast carcinoma also have similar immunoprofiles. Both show CK7(+)/CK20(−) pattern.44 Both tumor types show estrogen receptor and progesterone receptor expression.4,46 Mammaglobin, a secretory protein, is expressed in 55% of breast carcinomas whereas it is expressed in 3% (1 of 40 cases) of ovarian serous carcinoma and 39% (23 of 59 cases) of uterine endometrioid adenocarcinoma.2 The finding presumably indicates that mammaglobin is more sensitive marker for breast carcinoma than gross cystic disease fluid protein (GCDFP)-15, but that it is not as specific for breast carcinomas.2 CA125 is expressed in a majority (up to 90%) of ovarian carcinomas and a subset (between 10% and 30%) of breast carcinomas.41 Its expression in ovarian carcinomas is typically strong and diffuse whereas it is only focal or weak in breast carcinomas. Although CA125 is useful in this context, it can be expressed in a variety of carcinomas including cholangiocarcinoma and pancreatic duct carcinoma,19 and malignant mesothelioma.1

In a majority of cases, a distinction of ovarian carcinoma from metastatic breast carcinoma is not difficult by histopathologic examination when the material of the primary breast carcinoma is available for comparison. However, when a history of primary breast carcinoma is unknown, when the previous material is not available for review, when the sample of the ovarian lesion is small, or when the tumor is morphologically poorly differentiated and nondescript, the distinction between the 2 tumors can be challenging. In such circumstances, a battery of immunohistochemical stains has been used to determine the origin of the tumor.17,26,41 Among those markers investigated, WT1 and GCDFP-15 (BRST-2) were reported as being the most useful markers for ovarian and breast carcinomas, respectively.17,26,41 GCDFP-15 is negative in ovarian carcinomas.41 However, GCDFP-15 shows a low sensitivity (23% to 45%) for breast carcinomas,2,10,33 and more than half of breast carcinomas do not express this marker. WT1, detected by either monoclonal 6F-H2 or polyclonal C-19, is a sensitive marker for ovarian carcinoma with a sensitivity of 76% to 83%,30,41,47 particularly ovarian serous carcinoma, with a sensitivity of 93% to 94%,13,41 but it can be expressed in a subset of breast carcinomas, with the reported frequency of 2% by using monoclonal 6F-H2,28 and 6.8% by using polyclonal C-19.13 In our series, 5 of 178 invasive breast carcinomas (2.8%) showed variable WT1 nuclear expression. Indeed, it has been proposed that WT1 protein expression and high-level WT1 mRNA expression may be associated with the aggressive behavior in the breast carcinomas,25,42 and WT1-positive breast carcinomas tend to be morphologically poorly differentiated, and large size, estrogen receptor negative and Her-2 positive.37 It has been reported that HER2/neu oncogene increases WT1 expression.42 WT1 expression is generally negative in endometrioid, clear cell, and mucinous types of ovarian carcinomas.13 These findings indicate that WT1 is not completely reliable for distinguishing between ovarian carcinomas of serous papillary and endometrioid types, and breast carcinomas.

Pax8 is a crucial transcription factor for thyroid gland organogenesis along with thyroid transcription factor (TTF)-1 and TTF-2 (FoxE1). It is also expressed in the metanephros, midhindbrain boundary region, and Müllerian duct.24 In adults, Pax8 is expressed in nonciliated mucosal cells of the fallopian tubes and simple ovarian epithelial inclusion cysts, but not in the surface epithelial cells of the ovary.3,29 This distribution pattern of Pax8 expression in the embryonic and mature stages of Müllerian system is reminiscent of that of Pax2, which is expressed by 67% of ovarian serous carcinoma.40 These expression patterns of Pax8 and Pax2 suggest that nonmucinous surface epithelial carcinomas, particularly, serous, endometrioid, and clear cell types, originate from Müllerian epithelium, for example, fallopian tube mucosa, endosalpingiosis, Müllerian metaplasia, or endometriosis, rather than coelomic mesothelium, the latter of which covers the ovarian surface and has been believed to be the origin of ovarian surface epithelial tumors.15,21,35,39

Recently, we described the utility of Pax8 in the diagnosis of thyroid gland tumors.29 In the study, we found Pax8 expression in ovarian serous carcinoma and clear cell carcinoma of the kidney, which are the tumors derived from the Müllerian duct and metanephros, respectively.29 Pax8 expression has been scarcely investigated in ovarian cancers. Bowen et al,3 reported, with a small number of cases, Pax8 expression in 79% (15/19 cases), 64% (9/14), 100% (2/2), and 11% (2/18) of serous, endometrioid, clear cell, and mucinous carcinomas, respectively. Our results are at variance with the previous report on the subject. The reason for this disagreement may be attributed to a difference in antibodies given that the Pax8 antibody used in the current series was not used in the previous study.3 The reliability of our marker was described and demonstrated on a variety of normal tissues and common malignant neoplasms in our previous study.29

A recent study compared gene expression patterns between breast and ovarian carcinomas by DNA microarray technique, and found that Pax8, along with mesothelin and ephrin-B1 (EFNB1), was more highly expressed among 551 genes examined in the ovarian carcinomas, and, therefore, proposed that these 3 were the best discriminators between the 2 carcinomas.34 We validated this information by performing immunohistochemistry on a variety of histotypes of ovarian carcinomas and breast carcinomas with a large number of cases. The sensitivity of Pax8 in nonmucinous ovarian surface epithelial carcinomas was higher than that of WT1 in the current study, and seems to be higher than that of Pax2, which was described in the literature.40

In summary, it is clear that Pax8 serves as an important marker for discriminating ovarian nonmucinous surface epithelial carcinomas from breast carcinomas, with greater sensitivity and specificity than WT1. It is particularly useful for the diagnosis of clear cell and endometrioid types where WT1 expression is generally negative or only focal.


1. Attanoos RL, Webb R, Dojcinov SD, et al. Value of mesothelial and epithelial antibodies in distinguishing diffuse peritoneal mesothelioma in females from serous papillary carcinoma of the ovary and peritoneum. Histopathology. 2002;40:237–244.
2. Bhargava R, Beriwal S, Dabbs DJ. Mammaglobin vs. GCDFP-15: an immunohistologic validation survey for sensitivity and specificity. Am J Clin Pathol. 2007;127:103–113.
3. Bowen NJ, Logani S, Dickerson EB, et al. Emerging roles for PAX8 in ovarian cancer and endosalpingeal development. Gynecol Oncol. 2007;104:331–337.
4. Comin CE, Saieva C, Messerini L. h-caldesmon, calretinin, estrogen receptor, and Ber-EP4: a useful combination of immunohistochemical markers for differentiating epithelioid peritoneal mesothelioma from serous papillary carcinoma of the ovary. Am J Surg Pathol. 2007;31:1139–1148.
5. Curtin JP, Barakat RR, Hoskins WJ. Ovarian disease in women with breast cancer. Obstet Gynecol. 1994;84:449–452.
6. Fishman A, Kim HS, Girtanner RE, et al. Solitary breast metastasis as first manifestation of ovarian carcinoid tumor. Gynecol Oncol. 1994;54:222–226.
7. Fleischhacker RH, Young RH, Scully RE. Breast carcinoma metastatic to the ovary: a study of 76 cases. Mod Pathol. 1994;7:88A.
8. Galper S, Gelman R, Recht A, et al. Second nonbreast malignancies after conservative surgery and radiation therapy for early-stage breast cancer. Int J Radiat Oncol Biol Phys. 2002;52:406–414.
9. Gokaslan H, Yoruk P, Pekin T, et al. Bilateral metastatic breast cancer as the first manifestation of ovarian cancer: case report. Eur J Gynaecol Oncol. 2005;26:336–338.
10. Han JH, Kang Y, Shin HC, et al. Mammaglobin expression in lymph nodes is an important marker of metastatic breast carcinoma. Arch Pathol Lab Med. 2003;127:1330–1334.
11. Harris M, Howell A, Chrissohou M, et al. A comparison of the metastatic pattern of infiltrating lobular carcinoma and infiltrating duct carcinoma of the breast. Br J Cancer. 1984;50:23–30.
12. Harwood TR. Metastatic carcinoma to the breast. JAMA. 1971;218:97.
13. Hwang H, Quenneville L, Yaziji H, et al. Wilms tumor gene product: sensitive and contextually specific marker of serous carcinomas of ovarian surface epithelial origin. Appl Immunohistochem Mol Morphol. 2004;12:122–126.
14. Kattan J, Droz JP, Charpentier P, et al. Ovarian dysgerminoma metastatic to the breast. Gynecol Oncol. 1992;46:104–106.
15. Kindelberger DW, Lee Y, Miron A, et al. Intraepithelial carcinoma of the fimbria and pelvic serous carcinoma: evidence for a causal relationship. Am J Surg Pathol. 2007;31:161–169.
16. Lecca U, Medda F, Marcello C, et al. Ovarian metastasis in breast cancer. Eur J Gynaecol Oncol. 1980;1:168–174.
17. Lee BH, Hecht JL, Pinkus JL, et al. 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.
18. Lee YT, Hori JM. Significance of ovarian metastasis in therapeutic oophorectomy for advanced breast cancer. Cancer. 1971;27:1374–1378.
19. Loy TS, Quesenberry JT, Sharp SC. Distribution of CA 125 in adenocarcinomas. An immunohistochemical study of 481 cases. Am J Clin Pathol. 1992;98:175–179.
20. Lumb G, Mackenzie DH. The incidence of metastases in adrenal glands and ovaries removed for carcinoma of the breast. Cancer. 1959;12:521–526.
21. Marquez RT, Baggerly KA, Patterson AP, et al. Patterns of gene expression in different histotypes of epithelial ovarian cancer correlate with those in normal fallopian tube, endometrium, and colon. Clin Cancer Res. 2005;11:6116–6126.
22. McIntosh IH, Hooper AA, Millis RR, et al. Metastatic carcinoma within the breast. Clin Oncol. 1976;2:393–401.
23. Metcalfe KA, Lynch HT, Ghadirian P, et al. The risk of ovarian cancer after breast cancer in BRCA1 and BRCA2 carriers. Gynecol Oncol. 2005;96:222–226.
24. Mittag J, Winterhager E, Bauer K, et al. Congenital hypothyroid female pax8-deficient mice are infertile despite thyroid hormone replacement therapy. Endocrinology. 2007;148:719–725.
25. Miyoshi Y, Ando A, Egawa C, et al. High expression of Wilms' tumor suppressor gene predicts poor prognosis in breast cancer patients. Clin Cancer Res. 2002;8:1167–1171.
26. Monteagudo C, Merino MJ, LaPorte N, et al. Value of gross cystic disease fluid protein-15 in distinguishing metastatic breast carcinomas among poorly differentiated neoplasms involving the ovary. Hum Pathol. 1991;22:368–372.
27. Moore RG, Chung M, Granai CO, et al. Incidence of metastasis to the ovaries from nongenital tract primary tumors. Gynecol Oncol. 2004;93:87–91.
28. Nakatsuka S, Oji Y, Horiuchi T, et al. Immunohistochemical detection of WT1 protein in a variety of cancer cells. Mod Pathol. 2006;19:804–814.
29. Nonaka D, Tang Y, Chiriboga L, et al. Diagnostic utility of thyroid transcription factors Pax8 and TTF-2 (FoxE1) in thyroid epithelial neoplasms. Mod Pathol. 2008;21:192–200.
30. Ordonez NG. Value of thyroid transcription factor-1, E-cadherin, BG8, WT1, and CD44S immunostaining in distinguishing epithelial pleural mesothelioma from pulmonary and nonpulmonary adenocarcinoma. Am J Surg Pathol. 2000;24:598–606.
31. Osborne MP, Pitts RM. Therapeutic oophorectomy for advanced breast cancer. The significance of metastases to the ovary and of ovarian cortical stromal hyperplasia. Cancer. 1961;14:126–130.
32. Recine MA, Deavers MT, Middleton LP, et al. Serous carcinoma of the ovary and peritoneum with metastases to the breast and axillary lymph nodes: a potential pitfall. Am J Surg Pathol. 2004;28:1646–1651.
33. Satoh F, Umemura S, Osamura RY. Immunohistochemical analysis of GCDFP-15 and GCDFP-24 in mammary and non-mammary tissue. Breast Cancer. 2000;7:49–55.
34. Schaner ME, Ross DT, Ciaravino G, et al. Gene expression patterns in ovarian carcinomas. Mol Biol Cell. 2003;14:4376–4386.
35. Scully RE, Young RH, Clement PM. Surface epithelial-stromal tumors, serous tumors. In: Scully RE, Young RH, Clement PM, eds. Tumors of the Ovary, Maldeveloped Gonads, Fallopian Tube, and Broad Ligament. Washington, D.C.: Armed Forces Institute of Pathology; 1998:51–79.
36. Siehl JM, Thiel E, Heufelder K, et al. Possible regulation of Wilms' tumour gene 1 (WT1) expression by the paired box genes PAX2 and PAX8 and by the haematopoietic transcription factor GATA-1 in human acute myeloid leukaemias. Br J Haematol. 2003;123:235–242.
37. Silberstein GB, Van Horn K, Strickland P, et al. Altered expression of the WT1 wilms tumor suppressor gene in human breast cancer. Proc Natl Acad Sci USA. 1997;94:8132–8137.
38. Tavassoli FA, Devilee P. Tumours of the breast. In: Tavassoli FA, Devilee P, eds. Tumours of the Breast and Female Genital Organs. Lyon: IARC Press; 2003a:9–112.
39. Tavassoli FA, Devilee P. Tumours of the ovary and peritoneum. In: Tavassoli FA, Devilee P, eds. Tumours of the Breast and Female Genital Organs. Lyon: IARC Press; 2003b:113–202.
40. Tong GX, Chiriboga L, Hamele-Bena D, et al. Expression of PAX2 in papillary serous carcinoma of the ovary: immunohistochemical evidence of fallopian tube or secondary Mullerian system origin? Mod Pathol. 2007;20:856–863.
41. Tornos C, Soslow R, Chen S, et al. Expression of WT1, CA 125, and GCDFP-15 as useful markers in the differential diagnosis of primary ovarian carcinomas versus metastatic breast cancer to the ovary. Am J Surg Pathol. 2005;29:1482–1489.
42. Tuna M, Chavez-Reyes A, Tari AM. HER2/neu increases the expression of Wilms' Tumor 1 (WT1) protein to stimulate S-phase proliferation and inhibit apoptosis in breast cancer cells. Oncogene. 2005;24:1648–1652.
43. Ulbright TM, Roth LM, Stehman FB. Secondary ovarian neoplasia. A clinicopathologic study of 35 cases. Cancer. 1984;53:1164–1174.
44. Wang N, Zee S, Zarbo R, et al. Coordinate expression of cytokeratins 7 and 20 defines unique subsets of carcinomas. Applied Immunohistochemistry. 1995;3:99–107.
45. Werness BA, Ramus SJ, Whittemore AS, et al. Histopathology of familial ovarian tumors in women from families with and without germline BRCA1 mutations. Hum Pathol. 2000;31:1420–1424.
46. Wong KK, Lu KH, Malpica A, et al. Significantly greater expression of ER, PR, and ECAD in advanced-stage low-grade ovarian serous carcinoma as revealed by immunohistochemical analysis. Int J Gynecol Pathol. 2007;26:404–409.
47. Yamamoto S, Tsuda H, Kita T, et al. Clinicopathological significance of WT1 expression in ovarian cancer: a possible accelerator of tumor progression in serous adenocarcinoma. Virchows Arch. 2007;451:27–35.
48. Young RH, Carey RW, Robboy SJ. Breast carcinoma masquerading as primary ovarian neoplasm. Cancer. 1981;48:210–212.

Pax8; WT1; ovarian carcinoma; breast carcinoma; ovary; mammary gland

© 2008 Lippincott Williams & Wilkins, Inc.