International Journal of Gynecological Pathology 33:425–431, Lippincott Williams & Wilkins, Baltimore r 2014 International Society of Gynecological Pathologists
Original Article
PAX8 Expression in Uterine Malignant Mesodermal Mixed Tumor (Carcinosarcoma) Brittany J. Holmes,
M.D.,
Allen M. Gown, M.D., Russell Vang, and Anna Yemelyanova, M.D.
M.D.,
Brigitte M. Ronnett,
M.D.,
Summary: PAX8 has emerged as a useful immunohistochemical marker for epithelial neoplasms of gynecologic origin. Expression of PAX8 in uterine malignant mesodermal mixed tumors (MMMT, carcinosarcoma) has not been characterized in detail. The goal of this study is to evaluate PAX8 expression in uterine MMMTs, with particular attention to its distribution in specific tumor components. Thirty-seven cases were studied. PAX8 expression was assessed by immunohistochemistry and scored separately in the epithelial and mesenchymal components of the tumors. The extent of staining was scored based on the estimated percentage of positive tumor cells as 1+: 1% to 25%; 2+: 26% to 50%; 3+: 51% to 75%; 4+: 76% to 100%. The epithelial component expressed PAX8 in all but 1 tumor, with 92% of tumors displaying 3+ and 4+ extent of staining. The mesenchymal component lacked PAX8 expression in 27 cases (73%) with variable expression in the remaining 10 cases. In addition, 12 tumors contained undifferentiated areas that were not readily classifiable as carcinoma or sarcoma based on morphologic features. Of these, 8 (67%) were negative for PAX8, whereas 4 (33%) demonstrated variable extent of expression. Thus, PAX8 is expressed in the carcinomatous components of nearly all uterine MMMTs (97%), with expression in sarcomatous and undifferentiated components being less common and less extensive. The uniform, extensive expression in the carcinomatous components makes PAX8 a useful marker for diagnosis of carcinomatous metastases of uterine MMMT at extrauterine sites. Its infrequent expression in the sarcomatous and undifferentiated components limits its utility in identifying sarcoma-predominant metastases as gynecologic in origin. Key Words: PAX8—Malignant mesodermal mixed tumor— Carcinosarcoma—Immunohistochemistry—Uterus.
Malignant mesodermal mixed tumor (MMMT), also termed carcinosarcoma or malignant Mu¨llerian mixed tumor, is an aggressive uterine neoplasm of postmenopausal women. Surgical stage and the depth
of myometrial invasion are the most important prognostic factors (1). The pathologic features of MMMT include distinct but mixed carcinomatous and sarcomatous components. The epithelial component may demonstrate serous or endometrioid differentiation, and the sarcoma is classified as homologous (Mu¨llerian) or heterologous (most commonly rhabdoid or cartilaginous differentiation) (2). MMMTs typically metastasize through lymph-vascular invasion, and the carcinomatous component typically predominates in metastases (1,3,4). Recent studies exploring the pathogenesis of MMMT suggested that they evolve from endometrial carcinoma, with the sarcoma representing dedifferentiation into a
From the Departments of Pathology (B.J.H., R.V., B.M.R., A.Y.); Gynecology and Obstetrics (R.V., B.M.R.), The Johns Hopkins University School of Medicine, Johns Hopkins Hospital, Baltimore, Maryland; and PhenoPath Laboratories (A.M.G), Seattle, Washington. The authors declare no conflict of interest. Address correspondence and reprint requests to Anna Yemelyanova, MD, Department of Pathology, The Johns Hopkins University School of Medicine, The Johns Hopkins Hospital, 401 North Broadway, Weinberg Building, Room 2242, Baltimore, MD 21231. E-mail: [email protected].
425
DOI: 10.1097/PGP.0b013e31829d7705
426
B. J. HOLMES ET AL.
mesenchymal cell type (2,3). The process of epithelialmesenchymal transition appears to progress through specific steps including decreased E-cadherin and changing patterns of micro-RNA expression (2). Molecular studies have confirmed a monoclonal origin of the carcinoma and sarcoma in the majority of MMMTs using a variety of techniques, including X chromosome inactivation, p53 and Kras mutational analysis, and loss of heterozygosity studies (5–8). However, some recent data suggested that the gene expression profile of MMMT may have more in common with uterine sarcoma than with endometrial carcinoma (9,10). A small subset (10%–15%) of MMMTs appears to arise from adenosarcomas with subsequent malignant transformation of the epithelium, leading to a distinct, genuinely biphasic collision tumor (11). Although the pathogenesis of MMMT and its relationship to endometrial carcinoma remains a topic of active investigation, these 2 entities appear to have distinct clinical outcomes. MMMT behaves more aggressively than endometrial carcinoma, with presentation at a more advanced stage and lower cancer-specific survival independent of stage (12). Thus, accurate diagnosis of MMMT remains a priority because of its prognostic significance. PAX8 is a member of the PAX family of pairedbox, DNA-binding transcription factors critical in cell lineage determination (13,14). Specifically, PAX8 plays an important role in the organogenesis of the central nervous system, thyroid, kidney, and Mu¨llerian tract (13–15). Thus, PAX8 is normally expressed in non-neoplastic endometrium, endocervical epithelium, and basal and secretory cells within the fallopian tube epithelium, whereas the normal ovarian stroma is negative (13–15). In addition to renal cell carcinomas, thyroid carcinomas, and thymic neoplasms, nuclear expression of PAX8 has been demonstrated in endometrial serous and endometrioid carcinomas and ovarian serous and other nonmucinous carcinomas (13–15). However, only scattered and inconsistent reports of PAX8 expression in MMMT are reported in the literature (13,16–18). Some studies demonstrated positive PAX8 staining in a proportion of cases (13,16,17), whereas others reported lack of immunoreactivity (15,18). Effusion cytology specimens from 1 endometrial MMMT and 2 ovarian MMMTs were shown to label with antibodies to PAX8 (19). Details regarding which components within these tumors demonstrated PAX8 expression are not provided in most studies. Thus, the goal of this study is to characterize PAX8 expression in uterine MMMTs, Int J Gynecol Pathol Vol. 33, No. 4, July 2014
with particular attention to its distribution in specific tumor components. MATERIALS AND METHODS Case Selection The study was approved by The Johns Hopkins University School of Medicine Institutional Review Board. Uterine MMMT (carcinosarcoma) cases in hysterectomy specimens were retrieved from the surgical pathology archives of The Johns Hopkins Hospital. A total of 37 cases were studied. The patient’s age and tumor stage were collected for each case. The original hematoxylin and eosin-stained slides for all cases were reviewed by 2 pathologists (B.J.H. and A.Y.) with selected cases reviewed by other coauthors to confirm the diagnosis of MMMT. The carcinomatous component was classified as either endometrioid, serous, mixed, or not otherwise specified (NOS) based on established criteria (20). The sarcomatous component was evaluated for the pattern of mesenchymal differentiation and described as chondrosarcomatous, rhabdomyosarcomatous, leiomyosarcomatous (LMS), or NOS. Cases containing poorly differentiated areas that were not readily classifiable as carcinoma or sarcoma based on morphology were identified separately as having an undifferentiated component. Immunohistochemical Analysis of PAX8 Expression Five-micron sections from formalin-fixed, paraffinembedded tissues were deparaffinized and rehydrated. PAX8 expression was assessed using mouse monoclonal BC12 antibody (Biocare Medical, LLC, Concord, CA), employed at a dilution of 1:250. Epitope retrieval was carried out for 8 min in a pressure cooker in a microwave oven using citrate buffer, pH 6. The reaction was developed with the UltraVision Quanto polymer-based detection system (ThermoScientific, Waltham, MA) on Dako Autostainer Plus. Nuclear expression was considered a positive reaction. The epithelial/carcinomatous, mesenchymal/sarcomatous, and undifferentiated components of the tumors were scored individually. The immunolabeling was also assessed in normal tissue (endometrial glands, stroma, and myometrium) adjacent to MMMT when present. The extent of staining was scored based on the estimated percentage of positive tumor cells as 1+: 1% to 25%; 2+: 26% to 50%; 3+: 51% to 75%; 4+: 76% to 100%. The staining intensity was recorded separately on
PAX8 EXPRESSION IN UTERINE MMMT a scale of 0 to 3 as follows: 0—negative; 1—weak; 2—moderate; 3—strong. RESULTS A total of 37 tumors were studied. The patients’ ages ranged from 42 to 91 yr (mean/median: 68/70). The tumor stages were as follows: Stage I—16 cases, Stage II—2 cases, Stage III—13 cases, and Stage IV—6 cases. Of the 37 tumors, the epithelial component displayed serous differentiation in 19 cases; endometrioid in 7 cases; and indeterminate (NOS) or mixed serous and endometrioid in 11 cases. The sarcomatous component demonstrated chondroid differentiation in 6 cases, rhabdoid in 5 cases, LMS in 1 case, and NOS in 23 cases. Two additional cases demonstrated both rhabdoid and chondroid or both LMS and chondroid differentiation. An un-
427
differentiated component (not readily classifiable as carcinoma or sarcoma) was present in 12 cases. The epithelial component of MMMT expressed PAX8 in all but 1 case (97%). Independent of the differentiation pattern, the vast majority of cases (94%) showed 3+ to 4+ extent of expression with moderate to strong intensity. Expression (percentage of positive tumor cells) was highest in the epithelium with serous differentiation (mean/median: 93%/ 100%; range, 30%–100%); followed by indeterminate (NOS) or mixed (mean/median: 80%/90%; range, 0%–100%); with the least expression in cases with endometrioid differentiation (mean/median: 71%/80%; range 30%–90%) (Fig. 1). In 27 cases (73%), the sarcomatous component was completely negative for PAX8 expression. The 10 positive cases showed variable expression with predominantly 1+ to 2+ extent of staining (60% of
FIG. 1. Malignant mesodermal mixed tumor, composed of intimately admixed malignant epithelial and mesenchymal components (A), demonstrates the typical pattern of PAX8 expression with diffuse/strong (4+) expression in the carcinomatous component and complete lack of expression in the sarcomatous component (B). The carcinomatous component displays serous differentiation (C), and the sarcomatous component is high grade, without specific differentiation (D).
Int J Gynecol Pathol Vol. 33, No. 4, July 2014
428
B. J. HOLMES ET AL.
FIG. 2. Malignant mesodermal mixed tumor, composed of high-grade adenocarcinoma with serous features and high-grade sarcoma (not otherwise specified), demonstrates the less common finding of diffuse/strong PAX8 expression in both the carcinomatous (A and B) and sarcomatous components (C and D).
positive cases) (Fig. 1). Overall, the sarcomatous component demonstrated a mean/median percent positive cells of 11%/0% (range, 0%–100%) (Fig. 2).
PAX8 expression was observed in an undifferentiated component in 4 cases, which demonstrated 3+ to 4+ extent of staining in 3 of 4 cases. The
FIG. 3. Undifferentiated component, not readily classifiable as carcinoma or sarcoma (A), lacks PAX8 expression (B).
Int J Gynecol Pathol Vol. 33, No. 4, July 2014
PAX8 EXPRESSION IN UTERINE MMMT
429
reported case of MMMT (19) and labeled 3 out of 4 effusion specimens with metastatic MMMT in another series (16). Prior immunohistochemical characterization of MMMT has demonstrated that the immunolabeling pattern mirrors the tumor’s distinctive morphologic appearance. The epithelial component typically expresses cytokeratins and epithelial membrane antigen. In contrast, the sarcomatous elements label with CD10, CD34, vimentin, or markers of muscle differentiation consistent with the morphologic appearance, such as myoglobin, desmin, or musclespecific actin (2,21,22). Our results demonstrate that PAX8 can be a helpful adjunct to morphologic diagnosis in highlighting the carcinomatous component and confirming the bimorphic nature of the neoplasm. Our findings expand upon prior individual reports of PAX8 expression in MMMT. One study reported 3 of 5 cases of endometrial MMMT focally positive for PAX8 (13); similar studies identified 2 ovarian MMMTs and 1 uterine MMMT positive for PAX8 (16) and a single endometrial MMMT that was negative for PAX8 (18). Another study demonstrated PAX8 staining in 3 of 6 ovarian MMMTs, 2 of 3 uterine MMMTs, and 7 of 9 primary omental MMMTs (17). A single case of ovarian MMMT was negative for PAX8 expression in another report (15). None of these prior studies have described the details of PAX8 immunolabeling in the different components of MMMTs. The results of the present series demonstrate that nearly all MMMTs express PAX8 in the carcinomatous component. Our data are consistent with prior
remaining 8 cases with an undifferentiated component were completely negative for PAX8 in that component (Fig. 3). The immunohistochemical results for PAX8 labeling of the MMMTs are summarized in Table 1. Normal endometrial glands and stroma were present in 10 cases and normal myometrium in 19 cases. Normal uterine tissue adjacent to MMMTs displayed 3+ to 4+ immunolabeling of strong to moderate intensity in all endometrial glands (n = 10), and no expression in endometrial stroma (n = 10) or myometrium (n = 19). A single case contained a leiomyoma, which was also negative for PAX8 expression. DISCUSSION PAX8 has recently become established as a useful immunohistochemical marker for neoplasms of kidney, thyroid, and Mu¨llerian tract origin (13–15). Coupled with the lack of PAX8 expression in breast, pancreaticobiliary, and gastrointestinal tract tumors, this marker is useful as part of an immunohistochemical panel for identifying the primary site of metastatic carcinomas (18). Some studies have demonstrated concordance of PAX8 expression between primary tumors and their metastases for all primary sites, with superior sensitivity over Pax2 for Mu¨llerian tumors (14,17). Similarly, PAX8 has been validated as a consistent marker of Mu¨llerian neoplasms in cytopathologic effusion specimens with greater sensitivity than Pax2 (16,19). PAX8 specifically marked the carcinomatous metastatic cells in 1
TABLE 1. PAX8 expression in 37 cases of MMMT Extent of staining Histologic component Epithelial/carcinomatous Serous Endometrioid Mixed or NOS Mesenchymal/sarcomatous Chondrosarcomatous Rhabdomyosarcomatous Leiomyosarcomatous NOS Undifferentiated
n 37 19 7 11 37* 8 6 2 23 12
0 1 (3%) 0 0 1 (9%) 29 (74%) 6 (75%) 6 (100%) 1 (50%) 16 (70%) 8 (67%)
1+
5 2 1 2 0
0 0 0 0 (13%) (25%) 0 (50%) (9%)
2+ 2 1 1 0 1
(5%) (5%) (14%)
(3%) 0 0 0 1 (4%) 1 (8%)
3+ 4 1 1 2 2
(11%) (5%) (14%) (18%) (5%) 0 0 0 2 (9%) 1 (8%)
4+
Staining intensity (mean)w
(81%) (89%) (71%) (73%) (5%) 0 0 0 2 (9%) 2 (17%)
2.6 2.8 2.0 2.7 2.7 2.5 N/A 2.0 2.8 2.4
30 17 5 8 2
Extent of staining based on percentage positive cells: 0—negative (0%); 1+—1%–25%; 2+—26%–50%; 3+—51%–75%; 4+— 76%–100%. Staining intensity: 0—negative; 1—weak; 2—moderate; 3—strong. *In 1 case, the sarcomatous component contained both rhabdomyosarcomatous and chondrosarcomatous areas, whereas another case showed both leiomyosarcomatous and chondrosarcomatous differentiation. wThe mean staining intensity was calculated only among cases expressing PAX8, excluding negative cases. MMMT indicates malignant mesodermal mixed tumors; NOS, not otherwise specified.
Int J Gynecol Pathol Vol. 33, No. 4, July 2014
430
B. J. HOLMES ET AL.
studies that demonstrated higher expression of PAX8 in serous carcinoma of the endometrium compared with endometrioid carcinoma (23–25). The sarcomatous component lacked PAX8 expression in the majority of cases; the remaining one third of cases predominantly demonstrated focal (1+) immunolabeling. This may explain the reported negative results in previous studies (13,15,17,18). Given the most accepted theory that the majority of MMMTs evolve from high-grade nonendometrioid carcinomas (3,5,26), loss of PAX8 expression in the sarcomatous component likely reflects dedifferentiation of the carcinomatous component. Metastatic MMMT most frequently demonstrates a carcinomatous morphology or a combination of carcinoma and sarcoma, with only rare cases of metastatic sarcoma alone (3,4). Thus, the lack of PAX8 expression in the sarcomatous component may not significantly limit its utility when evaluating metastases of MMMT. In addition to testing PAX8 expression in a variety of carcinomas of different sites, understanding the immunolabeling pattern in non-neoplastic tissues provides a helpful context (14,17). As confirmed by our results, consistent expression in uninvolved endometrial glands serves as an internal control for the adequacy of the immunohistochemical preparation while evaluating uterine neoplasms; uninvolved myometrium and endometrial stroma do not express PAX8. In summary, this study demonstrates that PAX8 is expressed in the carcinomatous components of nearly all uterine MMMTs (97%). This consistent expression in the carcinomatous component makes PAX8 a useful marker for recognizing the gynecologic origin of carcinomatous metastases of MMMT at extrauterine sites. PAX8 expression in sarcomatous and undifferentiated components is less common and can be used as an adjunct to morphologic evaluation to highlight the bimorphic nature of the uterine tumor.
REFERENCES 1. Silverberg SG, Major FJ, Blessing JA, et al. Carcinosarcoma (malignant mixed mesodermal tumor) of the uterus. A Gynecologic Oncology Group pathologic study of 203 cases. Int J Gynecol Pathol 1990;9:1–19. 2. D’Angelo E, Prat J. Pathology of mixed Mullerian tumours. Best Pract Res Clin Obstet Gynaecol 2011;25:705–18. 3. Sreenan JJ, Hart WR. Carcinosarcomas of the female genital tract. A pathologic study of 29 metastatic tumors: further evidence for the dominant role of the epithelial component and the conversion theory of histogenesis. Am J Surg Pathol 1995;19:666–74.
Int J Gynecol Pathol Vol. 33, No. 4, July 2014
4. Bitterman P, Chun B, Kurman RJ. The significance of epithelial differentiation in mixed mesodermal tumors of the uterus. A clinicopathologic and immunohistochemical study. Am J Surg Pathol 1990;14:317–28. 5. Wada H, Enomoto T, Fujita M, et al. Molecular evidence that most but not all carcinosarcomas of the uterus are combination tumors. Cancer Res 1997;57:5379–85. 6. Abeln EC, Smit VT, Wessels JW, et al. Molecular genetic evidence for the conversion hypothesis of the origin of malignant mixed mullerian tumours. J Pathol 1997;183: 424–31. 7. Fujii H, Yoshida M, Gong ZX, et al. Frequent genetic heterogeneity in the clonal evolution of gynecological carcinosarcoma and its influence on phenotypic diversity. Cancer Res 2000;60:114–20. 8. Jin Z, Ogata S, Tamura G, et al. Carcinosarcomas (malignant mullerian mixed tumors) of the uterus and ovary: a genetic study with special reference to histogenesis. Int J Gynecol Pathol 2003;22:368–73. 9. Maxwell GL, Chandramouli GV, Dainty L, et al. Microarray analysis of endometrial carcinomas and mixed mullerian tumors reveals distinct gene expression profiles associated with different histologic types of uterine cancer. Clin Cancer Res 2005;11:4056–66. 10. Chiyoda T, Tsuda H, Tanaka H, et al. Expression profiles of carcinosarcoma of the uterine corpus-are these similar to carcinoma or sarcoma? Genes Chromosomes Cancer 2012;51:229–39. 11. Seidman JD, Chauhan S. Evaluation of the relationship between adenosarcoma and carcinosarcoma and a hypothesis of the histogenesis of uterine sarcomas. Int J Gynecol Pathol 2003;22:75–82. 12. Bansal N, Herzog TJ, Seshan VE, et al. Uterine carcinosarcomas and grade 3 endometrioid cancers: evidence for distinct tumor behavior. Obstet Gynecol 2008;112:64–70. 13. Laury AR, Perets R, Piao H, et al. A comprehensive analysis of PAX8 expression in human epithelial tumors. Am J Surg Pathol 2011;35:816–26. 14. Ozcan A, Shen SS, Hamilton C, et al. PAX 8 expression in non-neoplastic tissues, primary tumors, and metastatic tumors: a comprehensive immunohistochemical study. Mod Pathol 2011;24:751–64. 15. Tacha D, Zhou D, Cheng L. Expression of PAX8 in normal and neoplastic tissues: a comprehensive immunohistochemical study. Appl Immunohistochem Mol Morphol 2011;19:293–9. 16. Tong GX, Devaraj K, Hamele-Bena D, et al. Pax8: a marker for carcinoma of Mullerian origin in serous effusions. Diagn Cytopathol 2011;39:567–74. 17. Ozcan A, Liles N, Coffey D, et al. PAX2 and PAX8 expression in primary and metastatic mullerian epithelial tumors: a comprehensive comparison. Am J Surg Pathol 2011;35:1837–47. 18. Woodard AH, Yu J, Dabbs DJ, et al. NY-BR-1 and PAX8 immunoreactivity in breast, gynecologic tract, and other CK7+ carcinomas: potential use for determining site of origin. Am J Clin Pathol 2011;136:428–35. 19. Wiseman W, Michael CW, Roh MH. Diagnostic utility of PAX8 and PAX2 immunohistochemistry in the identification of metastatic Mullerian carcinoma in effusions. Diagn Cytopathol 2011;39:651–6. 20. Ellenson LH, Ronnett BM, Soslow RA, et al. Endometrial carcinoma. In: Kurman RJ, Ellenson LH, Ronnett BM, eds. Blaustein’s Pathology of the Female Genital Tract.; 6th ed. New York, NY: Springer Science+Business Media, LLC; 2011: 426–30. 21. Geisinger KR, Dabbs DJ, Marshall RB. Malignant mixed mullerian tumors. An ultrastructural and immunohistochemical analysis with histogenetic considerations. Cancer 1987;59:1781–90. 22. George E, Carlos Manivel J, Dehner LP, et al. Malignant mixed mu¨llerian tumors: an immunohistochemical study of 47
PAX8 EXPRESSION IN UTERINE MMMT cases, with histogenetic considerations and clinical correlation. Hum Pathol 1991;22:215–23. 23. Brunner AH, Riss P, Heinze G, et al. Immunoexpression of PAX 8 in endometrial cancer: relation to highgrade carcinoma and p53. Int J Gynecol Pathol 2011;30: 569–75. 24. Mhawech-Fauceglia P, Wang D, Samrao D, et al. Pair-Box (PAX8) protein-positive expression is associated with poor
431
disease outcome in women with endometrial cancer. Br J Cancer 2012;107:370–4. 25. Yemelyanova A, Gown AM, Ronnett BM, et al. Pax8 expression in uterine adenocarcinoma: immunohistochemical analysis of 94 cases. Mod Pathol 2010;23:269A. 26. Lopez-Garcia MA, Palacios J. Pathologic and molecular features of uterine carcinosarcomas. Semin Diagn Pathol 2010; 27:274–86.
Int J Gynecol Pathol Vol. 33, No. 4, July 2014
Original Article
PAX8 Expression in Uterine Malignant Mesodermal Mixed Tumor (Carcinosarcoma) Brittany J. Holmes,
M.D.,
Allen M. Gown, M.D., Russell Vang, and Anna Yemelyanova, M.D.
M.D.,
Brigitte M. Ronnett,
M.D.,
Summary: PAX8 has emerged as a useful immunohistochemical marker for epithelial neoplasms of gynecologic origin. Expression of PAX8 in uterine malignant mesodermal mixed tumors (MMMT, carcinosarcoma) has not been characterized in detail. The goal of this study is to evaluate PAX8 expression in uterine MMMTs, with particular attention to its distribution in specific tumor components. Thirty-seven cases were studied. PAX8 expression was assessed by immunohistochemistry and scored separately in the epithelial and mesenchymal components of the tumors. The extent of staining was scored based on the estimated percentage of positive tumor cells as 1+: 1% to 25%; 2+: 26% to 50%; 3+: 51% to 75%; 4+: 76% to 100%. The epithelial component expressed PAX8 in all but 1 tumor, with 92% of tumors displaying 3+ and 4+ extent of staining. The mesenchymal component lacked PAX8 expression in 27 cases (73%) with variable expression in the remaining 10 cases. In addition, 12 tumors contained undifferentiated areas that were not readily classifiable as carcinoma or sarcoma based on morphologic features. Of these, 8 (67%) were negative for PAX8, whereas 4 (33%) demonstrated variable extent of expression. Thus, PAX8 is expressed in the carcinomatous components of nearly all uterine MMMTs (97%), with expression in sarcomatous and undifferentiated components being less common and less extensive. The uniform, extensive expression in the carcinomatous components makes PAX8 a useful marker for diagnosis of carcinomatous metastases of uterine MMMT at extrauterine sites. Its infrequent expression in the sarcomatous and undifferentiated components limits its utility in identifying sarcoma-predominant metastases as gynecologic in origin. Key Words: PAX8—Malignant mesodermal mixed tumor— Carcinosarcoma—Immunohistochemistry—Uterus.
Malignant mesodermal mixed tumor (MMMT), also termed carcinosarcoma or malignant Mu¨llerian mixed tumor, is an aggressive uterine neoplasm of postmenopausal women. Surgical stage and the depth
of myometrial invasion are the most important prognostic factors (1). The pathologic features of MMMT include distinct but mixed carcinomatous and sarcomatous components. The epithelial component may demonstrate serous or endometrioid differentiation, and the sarcoma is classified as homologous (Mu¨llerian) or heterologous (most commonly rhabdoid or cartilaginous differentiation) (2). MMMTs typically metastasize through lymph-vascular invasion, and the carcinomatous component typically predominates in metastases (1,3,4). Recent studies exploring the pathogenesis of MMMT suggested that they evolve from endometrial carcinoma, with the sarcoma representing dedifferentiation into a
From the Departments of Pathology (B.J.H., R.V., B.M.R., A.Y.); Gynecology and Obstetrics (R.V., B.M.R.), The Johns Hopkins University School of Medicine, Johns Hopkins Hospital, Baltimore, Maryland; and PhenoPath Laboratories (A.M.G), Seattle, Washington. The authors declare no conflict of interest. Address correspondence and reprint requests to Anna Yemelyanova, MD, Department of Pathology, The Johns Hopkins University School of Medicine, The Johns Hopkins Hospital, 401 North Broadway, Weinberg Building, Room 2242, Baltimore, MD 21231. E-mail: [email protected].
425
DOI: 10.1097/PGP.0b013e31829d7705
426
B. J. HOLMES ET AL.
mesenchymal cell type (2,3). The process of epithelialmesenchymal transition appears to progress through specific steps including decreased E-cadherin and changing patterns of micro-RNA expression (2). Molecular studies have confirmed a monoclonal origin of the carcinoma and sarcoma in the majority of MMMTs using a variety of techniques, including X chromosome inactivation, p53 and Kras mutational analysis, and loss of heterozygosity studies (5–8). However, some recent data suggested that the gene expression profile of MMMT may have more in common with uterine sarcoma than with endometrial carcinoma (9,10). A small subset (10%–15%) of MMMTs appears to arise from adenosarcomas with subsequent malignant transformation of the epithelium, leading to a distinct, genuinely biphasic collision tumor (11). Although the pathogenesis of MMMT and its relationship to endometrial carcinoma remains a topic of active investigation, these 2 entities appear to have distinct clinical outcomes. MMMT behaves more aggressively than endometrial carcinoma, with presentation at a more advanced stage and lower cancer-specific survival independent of stage (12). Thus, accurate diagnosis of MMMT remains a priority because of its prognostic significance. PAX8 is a member of the PAX family of pairedbox, DNA-binding transcription factors critical in cell lineage determination (13,14). Specifically, PAX8 plays an important role in the organogenesis of the central nervous system, thyroid, kidney, and Mu¨llerian tract (13–15). Thus, PAX8 is normally expressed in non-neoplastic endometrium, endocervical epithelium, and basal and secretory cells within the fallopian tube epithelium, whereas the normal ovarian stroma is negative (13–15). In addition to renal cell carcinomas, thyroid carcinomas, and thymic neoplasms, nuclear expression of PAX8 has been demonstrated in endometrial serous and endometrioid carcinomas and ovarian serous and other nonmucinous carcinomas (13–15). However, only scattered and inconsistent reports of PAX8 expression in MMMT are reported in the literature (13,16–18). Some studies demonstrated positive PAX8 staining in a proportion of cases (13,16,17), whereas others reported lack of immunoreactivity (15,18). Effusion cytology specimens from 1 endometrial MMMT and 2 ovarian MMMTs were shown to label with antibodies to PAX8 (19). Details regarding which components within these tumors demonstrated PAX8 expression are not provided in most studies. Thus, the goal of this study is to characterize PAX8 expression in uterine MMMTs, Int J Gynecol Pathol Vol. 33, No. 4, July 2014
with particular attention to its distribution in specific tumor components. MATERIALS AND METHODS Case Selection The study was approved by The Johns Hopkins University School of Medicine Institutional Review Board. Uterine MMMT (carcinosarcoma) cases in hysterectomy specimens were retrieved from the surgical pathology archives of The Johns Hopkins Hospital. A total of 37 cases were studied. The patient’s age and tumor stage were collected for each case. The original hematoxylin and eosin-stained slides for all cases were reviewed by 2 pathologists (B.J.H. and A.Y.) with selected cases reviewed by other coauthors to confirm the diagnosis of MMMT. The carcinomatous component was classified as either endometrioid, serous, mixed, or not otherwise specified (NOS) based on established criteria (20). The sarcomatous component was evaluated for the pattern of mesenchymal differentiation and described as chondrosarcomatous, rhabdomyosarcomatous, leiomyosarcomatous (LMS), or NOS. Cases containing poorly differentiated areas that were not readily classifiable as carcinoma or sarcoma based on morphology were identified separately as having an undifferentiated component. Immunohistochemical Analysis of PAX8 Expression Five-micron sections from formalin-fixed, paraffinembedded tissues were deparaffinized and rehydrated. PAX8 expression was assessed using mouse monoclonal BC12 antibody (Biocare Medical, LLC, Concord, CA), employed at a dilution of 1:250. Epitope retrieval was carried out for 8 min in a pressure cooker in a microwave oven using citrate buffer, pH 6. The reaction was developed with the UltraVision Quanto polymer-based detection system (ThermoScientific, Waltham, MA) on Dako Autostainer Plus. Nuclear expression was considered a positive reaction. The epithelial/carcinomatous, mesenchymal/sarcomatous, and undifferentiated components of the tumors were scored individually. The immunolabeling was also assessed in normal tissue (endometrial glands, stroma, and myometrium) adjacent to MMMT when present. The extent of staining was scored based on the estimated percentage of positive tumor cells as 1+: 1% to 25%; 2+: 26% to 50%; 3+: 51% to 75%; 4+: 76% to 100%. The staining intensity was recorded separately on
PAX8 EXPRESSION IN UTERINE MMMT a scale of 0 to 3 as follows: 0—negative; 1—weak; 2—moderate; 3—strong. RESULTS A total of 37 tumors were studied. The patients’ ages ranged from 42 to 91 yr (mean/median: 68/70). The tumor stages were as follows: Stage I—16 cases, Stage II—2 cases, Stage III—13 cases, and Stage IV—6 cases. Of the 37 tumors, the epithelial component displayed serous differentiation in 19 cases; endometrioid in 7 cases; and indeterminate (NOS) or mixed serous and endometrioid in 11 cases. The sarcomatous component demonstrated chondroid differentiation in 6 cases, rhabdoid in 5 cases, LMS in 1 case, and NOS in 23 cases. Two additional cases demonstrated both rhabdoid and chondroid or both LMS and chondroid differentiation. An un-
427
differentiated component (not readily classifiable as carcinoma or sarcoma) was present in 12 cases. The epithelial component of MMMT expressed PAX8 in all but 1 case (97%). Independent of the differentiation pattern, the vast majority of cases (94%) showed 3+ to 4+ extent of expression with moderate to strong intensity. Expression (percentage of positive tumor cells) was highest in the epithelium with serous differentiation (mean/median: 93%/ 100%; range, 30%–100%); followed by indeterminate (NOS) or mixed (mean/median: 80%/90%; range, 0%–100%); with the least expression in cases with endometrioid differentiation (mean/median: 71%/80%; range 30%–90%) (Fig. 1). In 27 cases (73%), the sarcomatous component was completely negative for PAX8 expression. The 10 positive cases showed variable expression with predominantly 1+ to 2+ extent of staining (60% of
FIG. 1. Malignant mesodermal mixed tumor, composed of intimately admixed malignant epithelial and mesenchymal components (A), demonstrates the typical pattern of PAX8 expression with diffuse/strong (4+) expression in the carcinomatous component and complete lack of expression in the sarcomatous component (B). The carcinomatous component displays serous differentiation (C), and the sarcomatous component is high grade, without specific differentiation (D).
Int J Gynecol Pathol Vol. 33, No. 4, July 2014
428
B. J. HOLMES ET AL.
FIG. 2. Malignant mesodermal mixed tumor, composed of high-grade adenocarcinoma with serous features and high-grade sarcoma (not otherwise specified), demonstrates the less common finding of diffuse/strong PAX8 expression in both the carcinomatous (A and B) and sarcomatous components (C and D).
positive cases) (Fig. 1). Overall, the sarcomatous component demonstrated a mean/median percent positive cells of 11%/0% (range, 0%–100%) (Fig. 2).
PAX8 expression was observed in an undifferentiated component in 4 cases, which demonstrated 3+ to 4+ extent of staining in 3 of 4 cases. The
FIG. 3. Undifferentiated component, not readily classifiable as carcinoma or sarcoma (A), lacks PAX8 expression (B).
Int J Gynecol Pathol Vol. 33, No. 4, July 2014
PAX8 EXPRESSION IN UTERINE MMMT
429
reported case of MMMT (19) and labeled 3 out of 4 effusion specimens with metastatic MMMT in another series (16). Prior immunohistochemical characterization of MMMT has demonstrated that the immunolabeling pattern mirrors the tumor’s distinctive morphologic appearance. The epithelial component typically expresses cytokeratins and epithelial membrane antigen. In contrast, the sarcomatous elements label with CD10, CD34, vimentin, or markers of muscle differentiation consistent with the morphologic appearance, such as myoglobin, desmin, or musclespecific actin (2,21,22). Our results demonstrate that PAX8 can be a helpful adjunct to morphologic diagnosis in highlighting the carcinomatous component and confirming the bimorphic nature of the neoplasm. Our findings expand upon prior individual reports of PAX8 expression in MMMT. One study reported 3 of 5 cases of endometrial MMMT focally positive for PAX8 (13); similar studies identified 2 ovarian MMMTs and 1 uterine MMMT positive for PAX8 (16) and a single endometrial MMMT that was negative for PAX8 (18). Another study demonstrated PAX8 staining in 3 of 6 ovarian MMMTs, 2 of 3 uterine MMMTs, and 7 of 9 primary omental MMMTs (17). A single case of ovarian MMMT was negative for PAX8 expression in another report (15). None of these prior studies have described the details of PAX8 immunolabeling in the different components of MMMTs. The results of the present series demonstrate that nearly all MMMTs express PAX8 in the carcinomatous component. Our data are consistent with prior
remaining 8 cases with an undifferentiated component were completely negative for PAX8 in that component (Fig. 3). The immunohistochemical results for PAX8 labeling of the MMMTs are summarized in Table 1. Normal endometrial glands and stroma were present in 10 cases and normal myometrium in 19 cases. Normal uterine tissue adjacent to MMMTs displayed 3+ to 4+ immunolabeling of strong to moderate intensity in all endometrial glands (n = 10), and no expression in endometrial stroma (n = 10) or myometrium (n = 19). A single case contained a leiomyoma, which was also negative for PAX8 expression. DISCUSSION PAX8 has recently become established as a useful immunohistochemical marker for neoplasms of kidney, thyroid, and Mu¨llerian tract origin (13–15). Coupled with the lack of PAX8 expression in breast, pancreaticobiliary, and gastrointestinal tract tumors, this marker is useful as part of an immunohistochemical panel for identifying the primary site of metastatic carcinomas (18). Some studies have demonstrated concordance of PAX8 expression between primary tumors and their metastases for all primary sites, with superior sensitivity over Pax2 for Mu¨llerian tumors (14,17). Similarly, PAX8 has been validated as a consistent marker of Mu¨llerian neoplasms in cytopathologic effusion specimens with greater sensitivity than Pax2 (16,19). PAX8 specifically marked the carcinomatous metastatic cells in 1
TABLE 1. PAX8 expression in 37 cases of MMMT Extent of staining Histologic component Epithelial/carcinomatous Serous Endometrioid Mixed or NOS Mesenchymal/sarcomatous Chondrosarcomatous Rhabdomyosarcomatous Leiomyosarcomatous NOS Undifferentiated
n 37 19 7 11 37* 8 6 2 23 12
0 1 (3%) 0 0 1 (9%) 29 (74%) 6 (75%) 6 (100%) 1 (50%) 16 (70%) 8 (67%)
1+
5 2 1 2 0
0 0 0 0 (13%) (25%) 0 (50%) (9%)
2+ 2 1 1 0 1
(5%) (5%) (14%)
(3%) 0 0 0 1 (4%) 1 (8%)
3+ 4 1 1 2 2
(11%) (5%) (14%) (18%) (5%) 0 0 0 2 (9%) 1 (8%)
4+
Staining intensity (mean)w
(81%) (89%) (71%) (73%) (5%) 0 0 0 2 (9%) 2 (17%)
2.6 2.8 2.0 2.7 2.7 2.5 N/A 2.0 2.8 2.4
30 17 5 8 2
Extent of staining based on percentage positive cells: 0—negative (0%); 1+—1%–25%; 2+—26%–50%; 3+—51%–75%; 4+— 76%–100%. Staining intensity: 0—negative; 1—weak; 2—moderate; 3—strong. *In 1 case, the sarcomatous component contained both rhabdomyosarcomatous and chondrosarcomatous areas, whereas another case showed both leiomyosarcomatous and chondrosarcomatous differentiation. wThe mean staining intensity was calculated only among cases expressing PAX8, excluding negative cases. MMMT indicates malignant mesodermal mixed tumors; NOS, not otherwise specified.
Int J Gynecol Pathol Vol. 33, No. 4, July 2014
430
B. J. HOLMES ET AL.
studies that demonstrated higher expression of PAX8 in serous carcinoma of the endometrium compared with endometrioid carcinoma (23–25). The sarcomatous component lacked PAX8 expression in the majority of cases; the remaining one third of cases predominantly demonstrated focal (1+) immunolabeling. This may explain the reported negative results in previous studies (13,15,17,18). Given the most accepted theory that the majority of MMMTs evolve from high-grade nonendometrioid carcinomas (3,5,26), loss of PAX8 expression in the sarcomatous component likely reflects dedifferentiation of the carcinomatous component. Metastatic MMMT most frequently demonstrates a carcinomatous morphology or a combination of carcinoma and sarcoma, with only rare cases of metastatic sarcoma alone (3,4). Thus, the lack of PAX8 expression in the sarcomatous component may not significantly limit its utility when evaluating metastases of MMMT. In addition to testing PAX8 expression in a variety of carcinomas of different sites, understanding the immunolabeling pattern in non-neoplastic tissues provides a helpful context (14,17). As confirmed by our results, consistent expression in uninvolved endometrial glands serves as an internal control for the adequacy of the immunohistochemical preparation while evaluating uterine neoplasms; uninvolved myometrium and endometrial stroma do not express PAX8. In summary, this study demonstrates that PAX8 is expressed in the carcinomatous components of nearly all uterine MMMTs (97%). This consistent expression in the carcinomatous component makes PAX8 a useful marker for recognizing the gynecologic origin of carcinomatous metastases of MMMT at extrauterine sites. PAX8 expression in sarcomatous and undifferentiated components is less common and can be used as an adjunct to morphologic evaluation to highlight the bimorphic nature of the uterine tumor.
REFERENCES 1. Silverberg SG, Major FJ, Blessing JA, et al. Carcinosarcoma (malignant mixed mesodermal tumor) of the uterus. A Gynecologic Oncology Group pathologic study of 203 cases. Int J Gynecol Pathol 1990;9:1–19. 2. D’Angelo E, Prat J. Pathology of mixed Mullerian tumours. Best Pract Res Clin Obstet Gynaecol 2011;25:705–18. 3. Sreenan JJ, Hart WR. Carcinosarcomas of the female genital tract. A pathologic study of 29 metastatic tumors: further evidence for the dominant role of the epithelial component and the conversion theory of histogenesis. Am J Surg Pathol 1995;19:666–74.
Int J Gynecol Pathol Vol. 33, No. 4, July 2014
4. Bitterman P, Chun B, Kurman RJ. The significance of epithelial differentiation in mixed mesodermal tumors of the uterus. A clinicopathologic and immunohistochemical study. Am J Surg Pathol 1990;14:317–28. 5. Wada H, Enomoto T, Fujita M, et al. Molecular evidence that most but not all carcinosarcomas of the uterus are combination tumors. Cancer Res 1997;57:5379–85. 6. Abeln EC, Smit VT, Wessels JW, et al. Molecular genetic evidence for the conversion hypothesis of the origin of malignant mixed mullerian tumours. J Pathol 1997;183: 424–31. 7. Fujii H, Yoshida M, Gong ZX, et al. Frequent genetic heterogeneity in the clonal evolution of gynecological carcinosarcoma and its influence on phenotypic diversity. Cancer Res 2000;60:114–20. 8. Jin Z, Ogata S, Tamura G, et al. Carcinosarcomas (malignant mullerian mixed tumors) of the uterus and ovary: a genetic study with special reference to histogenesis. Int J Gynecol Pathol 2003;22:368–73. 9. Maxwell GL, Chandramouli GV, Dainty L, et al. Microarray analysis of endometrial carcinomas and mixed mullerian tumors reveals distinct gene expression profiles associated with different histologic types of uterine cancer. Clin Cancer Res 2005;11:4056–66. 10. Chiyoda T, Tsuda H, Tanaka H, et al. Expression profiles of carcinosarcoma of the uterine corpus-are these similar to carcinoma or sarcoma? Genes Chromosomes Cancer 2012;51:229–39. 11. Seidman JD, Chauhan S. Evaluation of the relationship between adenosarcoma and carcinosarcoma and a hypothesis of the histogenesis of uterine sarcomas. Int J Gynecol Pathol 2003;22:75–82. 12. Bansal N, Herzog TJ, Seshan VE, et al. Uterine carcinosarcomas and grade 3 endometrioid cancers: evidence for distinct tumor behavior. Obstet Gynecol 2008;112:64–70. 13. Laury AR, Perets R, Piao H, et al. A comprehensive analysis of PAX8 expression in human epithelial tumors. Am J Surg Pathol 2011;35:816–26. 14. Ozcan A, Shen SS, Hamilton C, et al. PAX 8 expression in non-neoplastic tissues, primary tumors, and metastatic tumors: a comprehensive immunohistochemical study. Mod Pathol 2011;24:751–64. 15. Tacha D, Zhou D, Cheng L. Expression of PAX8 in normal and neoplastic tissues: a comprehensive immunohistochemical study. Appl Immunohistochem Mol Morphol 2011;19:293–9. 16. Tong GX, Devaraj K, Hamele-Bena D, et al. Pax8: a marker for carcinoma of Mullerian origin in serous effusions. Diagn Cytopathol 2011;39:567–74. 17. Ozcan A, Liles N, Coffey D, et al. PAX2 and PAX8 expression in primary and metastatic mullerian epithelial tumors: a comprehensive comparison. Am J Surg Pathol 2011;35:1837–47. 18. Woodard AH, Yu J, Dabbs DJ, et al. NY-BR-1 and PAX8 immunoreactivity in breast, gynecologic tract, and other CK7+ carcinomas: potential use for determining site of origin. Am J Clin Pathol 2011;136:428–35. 19. Wiseman W, Michael CW, Roh MH. Diagnostic utility of PAX8 and PAX2 immunohistochemistry in the identification of metastatic Mullerian carcinoma in effusions. Diagn Cytopathol 2011;39:651–6. 20. Ellenson LH, Ronnett BM, Soslow RA, et al. Endometrial carcinoma. In: Kurman RJ, Ellenson LH, Ronnett BM, eds. Blaustein’s Pathology of the Female Genital Tract.; 6th ed. New York, NY: Springer Science+Business Media, LLC; 2011: 426–30. 21. Geisinger KR, Dabbs DJ, Marshall RB. Malignant mixed mullerian tumors. An ultrastructural and immunohistochemical analysis with histogenetic considerations. Cancer 1987;59:1781–90. 22. George E, Carlos Manivel J, Dehner LP, et al. Malignant mixed mu¨llerian tumors: an immunohistochemical study of 47
PAX8 EXPRESSION IN UTERINE MMMT cases, with histogenetic considerations and clinical correlation. Hum Pathol 1991;22:215–23. 23. Brunner AH, Riss P, Heinze G, et al. Immunoexpression of PAX 8 in endometrial cancer: relation to highgrade carcinoma and p53. Int J Gynecol Pathol 2011;30: 569–75. 24. Mhawech-Fauceglia P, Wang D, Samrao D, et al. Pair-Box (PAX8) protein-positive expression is associated with poor
431
disease outcome in women with endometrial cancer. Br J Cancer 2012;107:370–4. 25. Yemelyanova A, Gown AM, Ronnett BM, et al. Pax8 expression in uterine adenocarcinoma: immunohistochemical analysis of 94 cases. Mod Pathol 2010;23:269A. 26. Lopez-Garcia MA, Palacios J. Pathologic and molecular features of uterine carcinosarcomas. Semin Diagn Pathol 2010; 27:274–86.
Int J Gynecol Pathol Vol. 33, No. 4, July 2014
