SALL4 Is a Novel Sensitive and Specific Marker for Metastatic Germ Cell Tumors, With Particular Utility in Detection of Metastatic Yolk Sac Tumors

Transcription

1 Original Article SALL4 Is a Novel Sensitive and Specific Marker for Germ Cell Tumors, With Particular Utility in Detection of Yolk Sac Tumors Dengfeng Cao, MD, PhD 1 ; Peter A. Humphrey, MD, PhD 1 ; and Robert W. Allan, MD 2 BACKGROUND: The correct diagnosis of metastatic germ cell tumors is critical, because these tumors can be effectively treated and are even cured with modern therapy. Their histopathologic diagnosis can be challenging without immunohistochemical markers, which currently have limitations. SALL4 is a novel stem cell marker essential to maintain pluripotency and self-renewal of embryonic stem cells. In the current study, the authors investigated the utility of SALL4 as a potential diagnostic marker for metastatic germ cell tumors. METHODS: Ninety metastatic germ cell tumors from testis, ovary, and extragonadal sites were stained with a monoclonal SALL4 antibody. In addition, 170 metastatic nongerm cell malignancies, including 158 carcinomas (6 head and neck, 8 thyroid, 12 lung, 8 breast, 7 hepatocellular, 3 cholangiocarcinomas, 2 ampullary, 10 pancreatic, 18 gastric, 15 esophageal, 10 renal cell, 10 urothelial, 12 prostatic, 18 ovarian, 6 uterine, and 13 colonic) and 12 melanomas, were also stained to test SALL4 specificity. RESULTS: All 22 seminomas, 7 dysgerminomas, 22 embryonal carcinomas, and 14 of 15 yolk sac tumors displayed strong and diffuse SALL positivity in >90% of tumor cells (80% of tumor cells were strongly positive in the remaining yolk sac tumor). Five of 7 choriocarcinomas and 9 of 18 teratomas were also variably positive for SALL4. In contrast, only 10 (esophageal, gastric, and colonic adenocarcinomas) of 170 metastatic somatic tumors demonstrated focally weak SALL4 reactivity (<25% tumor cells). CONCLUSIONS: SALL4 is a novel sensitive and highly specific marker for metastatic germ cell tumors, and is particularly useful for detecting metastatic yolk sac tumors. Cancer 2009;115: VC 2009 American Cancer Society. KEY WORDS: metastatic germ cell tumors, yolk sac tumor, seminoma, dysgerminoma, embryonal carcinoma, SALL4, diagnostic marker. germ cell tumors are relatively uncommon. The majority are from the testis and, less commonly, from the ovary. Rarely, extragonadal germ cell tumors also metastasize. 1,2 In contrast to metastatic somatic malignancies, most metastatic germ cell tumors can be effectively treated and even cured with Corresponding author: Dengfeng Cao, MD, PhD, the Lauren V. Ackerman Laboratory of Surgical Pathology, Division of Anatomic and Molecular Pathology, Washington University School of Medicine, 660 S Euclid Avenue, Campus Box 8118, St. Louis, MO 63110; Fax: (314) ; dcao@path.wustl.edu 1 The Lauren V. Ackerman Laboratory of Surgical Pathology, Division of Anatomic and Molecular Pathology, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri; 2 Department of Pathology, Immunology, and Laboratory Medicine, University of Florida, Gainesville, Florida Received: October 6, 2008; Revised: November 20, 2008; Accepted: November 26, 2008 Published online: April 13, 2009 VC 2009 American Cancer Society DOI: /cncr.24308, Cancer June 15, 2009

2 Diagnostic Marker for Germ Cell Tumors/Cao et al modern therapy. 3-5 Therefore the accurate pathologic diagnosis of metastatic germ cell tumors and distinction from metastatic nongerm cell tumors is critical. However, germ cell tumors often exhibit diverse morphologic features, and the correct pathologic diagnosis of metastatic germ cell tumors can be challenging for multiple reasons. First, because of their rarity, metastatic germ cell tumors may be not considered in the initial differential diagnosis of the pathologist. Second, it is well known that primary testicular germ cell tumors (most frequently seminoma and less commonly choriocarcinoma) can undergo spontaneous regression, with the patient presenting with metastatic disease. 6,7 Occasionally, metastatic disease is identified even before the detection of the primary testicular tumor. 8 Third, the diagnostic difficulty in metastatic germ cell tumors is further complicated by small biopsy, extensive necrosis, crush effects, altered morphology after treatment, and tumor cells obscured by inflammatory cells. 9 To overcome these difficulties and to facilitate the diagnosis, immunohistochemical markers are often needed. Several immunohistochemical markers have been used to facilitate the diagnosis of metastatic germ cell tumors. Earlier markers included placental-like alkaline phosphatase (PLAP), CD30, C-KIT, and a-fetoprotein (AFP). However, only 50% to 60% of metastatic germ cell tumors are positive for PLAP. 10,11 Many types of nongerm cell tumors have also been reported to be positive for PLAP. 12,13 CD30 is positive in embryonal carcinoma, but it is also expressed in anaplastic lymphoma, Hodgkin lymphoma, carcinoma, and melanoma. 14,15 In addition, CD30 expression is lost in approximately two thirds of embryonal carcinomas after chemotherapy. 16 C-KIT labels seminoma and dysgerminoma, but many other nongerm cell tumors also demonstrate C-KIT expression, limiting its diagnostic utility in metastatic germ cell tumors. 17,18 AFP is considered a yolk sac tumor marker, but it lacks sensitivity and specificity The AFP staining in yolk sac tumors is often patchy ,22 In addition, AFP may be lost in metastases even if the primary tumor is positive for AFP. 23 Furthermore, many other types of nongerm cell tumors, such as hepatocellular carcinoma and ovarian clear cell carcinoma, can be positive for AFP. 21 All these data indicate that these earlier markers lack adequate sensitivity and/or specificity for metastatic germ cell tumors. More recently, the novel stem cell markers OCT4, NANOG, and SOX2 have emerged as more sensitive and specific immunohistochemical markers for detecting metastatic testicular germ cell tumors. OCT4 and NANOG label metastatic testicular seminoma and embryonal carcinoma. 11,24,25 SOX2 is positive in metastatic testicular embryonal carcinomas and teratomas. 25 Although OCT4, NANOG, and SOX2 are useful for detecting metastatic seminomas and embryonal carcinomas, they are not useful in detecting metastatic yolk sac tumors because they are not positive in yolk sac tumors. OCT4, NANOG, and SOX2 are stem cell markers essential to the maintenance of the pluripotency and self-renewal of embryonic stem cells Recent studies have shown that the maintenance of pluripotency and self-renewal of embryonic stem cells is controlled by a regulatory circuit that includes not only OCT4, NANOG, and SOX2 but also SALL4. 26,29,30 In this network, SALL4 appears to control transcription of OCT4. 29 The relation between SALL4 and OCT4, NANOG, and SOX2 suggests that SALL4 might be also a marker for metastatic germ cell tumors. The goal of the current study was to investigate the potential utility of SALL4 as a diagnostic marker in a large series of 90 metastatic germ cell tumors from the testis (n ¼ 73), ovary (n ¼ 13), and extragonadal sites (n ¼ 4). To test SALL4 specificity in metastatic germ cell tumors, we included 170 metastatic nongerm cell tumors from various organs. We also compared the staining results of SALL4 staining with those of OCT4 in these metastatic germ cell tumors. MATERIALS AND METHODS Case Selection Permission to perform this study was obtained from the institutional review boards of Washington University and the University of Florida. The surgical pathology files of Washington University Medical Center and the University of Florida were searched for metastatic germ cell tumors from the testis, ovary, and extragonadal sites. We identified 73 metastatic testicular germ cell tumors (67 metastases, 6 mixed with 2 components each), including 22 seminomas, 20 embryonal carcinomas, 8 yolk sac tumors, 5 choriocarcinomas, and 18 teratomas; 13 metastatic ovarian germ cell tumors (7 dysgerminomas, 1 Cancer June 15,

3 Original Article Table 1. SALL4 Expression in Seminomas, Dysgerminomas, Embryonic Carcinomas, and Teratomas Tumors SALL4 Staining seminomas (n¼22) (100%) dysgerminomas (n¼7) (100%) embryonal carcinomas (n¼21) (100%) teratomas (n¼18) 9 (50%) 9 (50%) embryonal carcinoma, 4 yolk sac tumors, and 1 choriocarcinoma); and 4 metastatic germ cell tumors from extragonadal sites (3 yolk sac tumors from the liver and sacrum and 1 choriocarcinoma from the mediastinum). The metastatic sites included lymph nodes (43 metastases, including 32 retroperitoneal, 6 inguinal, 3 cervical/supraclavicular, 1 mediastinal, and 1 pelvic), lung (8 metastases), retroperitoneum (8 metastases), brain (6 metastases), pelvis and omentum (6 metastases), soft tissue (5 metastases, including 2 in the groin, 1 in the neck, 1 in the abdominal wall, and 1 in the thigh), liver (2 metastases), bowel (2 metastases), mediastinum (1 metastasis), stomach (1 metastasis), bladder (1 metastasis), and bone (1 metastasis). To test specificity of SALL4, we included 170 metastatic somatic tumors from various sites including 6 head and neck squamous cell carcinomas, 8 thyroid carcinomas, 12 lung carcinomas, 8 breast carcinomas, 7 hepatocellular carcinomas, 3 cholangiocarcinomas, 2 ampullary adenocarcinomas, 10 pancreatic adenocarcinomas, 18 gastric adenocarcinomas, 15 esophageal carcinomas, 10 renal cell carcinomas, 10 urothelial carcinomas, 12 prostatic adenocarcinomas, 18 ovarian carcinomas, 6 uterine carcinomas, 13 colonic adenocarcinomas, and 12 melanomas. Immunohistochemical Staining and Evaluation For each case, 1 to 2 formalin-fixed and paraffin-embedded tissue blocks containing the metastatic germ cell tumor were retrieved. Four-micrometer unstained slides were generated from these paraffin blocks for immunohistochemical staining with a monoclonal antibody to SALL4 (Clone 6E3, dilution 1:100; Abnova Corporation, Taipei, Taiwan) and a monoclonal antibody to OCT4 (SC-5279, dilution 1:100; Santa Cruz Biotechnology, Inc, Santa Cruz, Calif). The immunohistochemical staining was performed with a Benchmark XT automatic immunostainer (Ventana Medical Systems, Tucson, Ariz). Appropriate positive and negative controls were included for each run. Only nuclear staining was considered positive. The staining intensity was scored as weak, moderate, or strong. The percentage of tumor cells that labeled with SALL4 was scored as 0 (no tumor cell stain), 1þ (0%-30%), 2þ (31%-60%), 3þ (61%-90%), and 4þ (>90%). For yolk sac tumors, we also performed pan cytokeratin, epithelial membrane protein (EMA), AFP, and PLAP (all prediluted antibodies from Ventana Medical Systems) immunostains. The percentage of cells stained with these markers (cytoplasmic and/or membranous) was scored using the same criteria as described above. Statistical Analysis The Fisher exact test was used to compare the staining results of SALL4 with those of OCT4. A P value <.05 was considered statistically significant. RESULTS Seminomas, Dysgerminomas, and Embryonal Carcinomas In this study, we tested 22 metastatic seminomas, 7 metastatic dysgerminomas, and 21 metastatic embryonal carcinomas (20 from the testis and 1 from the ovary) for SALL4 expression by immunohistochemical stain. All 50 metastatic tumors demonstrated 4þ (>90% tumor cells) strong nuclear staining of SALL4 in neoplastic cells (Table 1) (Figs. 1 and 2). Forty-three of these 50 cases (86%) 2642 Cancer June 15, 2009

4 Diagnostic Marker for Germ Cell Tumors/Cao et al FIGURE 1. Immunohistochemical staining of SALL4 is shown in metastatic dysgerminoma and seminoma in lymph node (A1-A2, dysgerminoma; B1-B2, seminoma), omentum (C1-C2, seminoma), and bladder (D1-D2, seminoma). Greater than 95% of tumor cells were strongly positive for SALL4 (A3 to D3). The tumor cells were also found to be strongly reactive for OCT4 (A4 to D4). H&E indicates hematoxylin and eosin. FIGURE 2. Immunohistochemical staining of SALL4 is shown in metastatic embryonal carcinoma in lymph node (A1 and A2), brain (B1 and B2), and soft tissue (C1 and C2). SALL4 was found to be strongly positive in >95% of tumor cells (A3 to C3). These tumor cells were also strongly positive for OCT4 (A4 to C4). H&E indicates hematoxylin and eosin. Cancer June 15,

5 Original Article FIGURE 3. Immunohistochemical staining of SALL4 is shown in metastatic choriocarcinoma and teratoma. In metastatic choriocarcinoma (A and C), SALL4 stained the mononucleated trophoblastic cells but not syncytiotrophoblastic cells (B and D). In approximately half of the metastatic testicular teratomas, some of the teratomatous glands (E, G, and I) were focally positive for SALL4 (F, H, and J) but, rarely, as many as 60% to 70% of epithelial cells in a given gland (G) could be stained with SALL4 (H). Other mature teratomatous elements, such as smooth muscle (E and G) and brain tissue (I), were found to be negative for SALL4 (F, H, and J). Immature elements in metastatic teratoma (K) may demonstrate focal strong SALL4 staining (L). H&E indicates hematoxylin and eosin. demonstrated >98% tumor cells stained with SALL4, and in the remaining 7 cases, the percentage of tumor cells stained with SALL4 varied between 95% and 98%. Teratomas This study included 18 metastatic teratomas (all from testis); 2 contained both mature and immature elements, and the remaining 16 contained only mature elements. Four of these 18 metastatic teratomas were also associated with another germ cell tumor component (1 with yolk sac tumor and 3 with embryonal carcinoma). Of the 16 teratomas with only mature elements, 8 demonstrated weak to moderate 1þ SALL4 staining in teratomatous glands (Table 1) (Fig. 3). The number of glandular epithelial cells staining with SALL4 ranged from 2% to 20% from case to case (Fig. 3). The majority of the SALL4-positive glands demonstrated focal weak to moderate staining of epithelial elements (<50%); only rarely were 60% to 70% of epithelial cells of a single gland found to be moderately positive for SALL4 staining (Fig. 3). Other mature elements, such as brain tissue, choroid plexus, squamous epithelium, cartilage, smooth muscle, and mesenchyme/stroma, were all negative for SALL4 staining (Fig. 3). The mature elements in the 2 teratomas that also contained immature elements were negative for SALL4 staining; immature elements included low-grade immature mesenchyme in 1 case and primitive neuroectodermal tissue in 1 case. The former was negative for SALL4 staining. The latter occupied more than a 4 field, and thus qualified as a primitive neuroectodermal tumor (PNET). Approximately 10% of the tumor cells in this PNET demonstrated SALL4 staining; the majority of SALL4-positive cells demonstrated strong staining, and few cells demonstrated weak staining (Fig. 3). Choriocarcinomas Seven metastatic nongestational choriocarcinomas were included in this study (5 from the testis, 1 from the ovary, and 1 from the mediastinum). Five of these 7 (71%) cases demonstrated SALL4 staining in the mononucleated trophoblastic cells (1þ in 3 cases and 3þ in 2 cases); syncytiotrophoblastic cells were negative (Table 2). The percentage of mononucleated trophoblastic cells that were stained for SALL4 varied from 10% to 80% (mean ¼ 40%); 2644 Cancer June 15, 2009

6 Diagnostic Marker for Germ Cell Tumors/Cao et al Table 2. Immunohistochemical Stain of SALL4 in Nongestational Choriocarcinomas Case No. Sex Primary Site Site Mononucleated Trophoblast Syncytiotrophoblast 1 Male Testis Lung 1þ (10% cells) moderate intensity 0 2 Male Testis Jejunum 1þ (25%-30% cells) moderate to strong intensity 0 3 Male Testis Lung 0 (negative) 0 4 Male Testis Lung 3þ (60%-65% cells) weak to strong intensity 0 5 Male Testis Brain 0 (negative) 0 6 Female Ovary Stomach and 3þ (80% cells) strong intensity 0 lymph nodes 7 Female Mediastinum Lung 1þ (25%-30% cells) moderate to strong intensity 0 Table 3. Immunohistochemical Stain of SALL4, Pan Cytokeratin, EMA, AFP, and PLAP in 15 Yolk Sac Tumors Case No. Age/Sex Primary Site Site SALL4 Pan Cytokeratin EMA AFP PLAP 1 34 y/m Testis Retroperitoneum 4þ (98%)* 4þ (100%) 1þ (<1%) 1þ (5%) 1þ (1%) 2 33 y/m Testis Retroperitoneum 3þ (80%) 4þ (100%) 0 3þ (75%-80%) y/m Testis Lymph node 4þ (>90%) 4þ (100%) 1þ (1%) 1þ (15%-20%) y/m Testis Perirenal soft tissue 4þ (>95%) 4þ (100%) 1þ (0.5%) 1þ (5%-10%) 1þ (1%) 5 34 y/m Testis Liver 4þ (95%) 3þ (70%) y/m Testis Groin 4þ (>90%) 4þ (90%) 0 1þ (5%-10%) 1þ (1%) 7 33 y/m Testis Brain 4þ (>95%) 4þ (90%) 1þ (2%-3%) 2þ (40%) y/m Testis Brain 4þ (>95%) 4þ (95%-98%) 0 1þ (15%-20%) mo/f Liver Lung 4þ (>95%) 4þ (>95%) /F Sacrum Lymph node, 4þ (>95%) 4þ (>95%) 0 1þ (0.5%) 2þ (40%-50%) soft tissue 11 2/M Sacrum Lung 4þ (>95%) 4þ (>95%) 0 1þ (25%) /F Ovary Omentum 4þ (>95%) 4þ (>95%) 0 1þ (2%) 3þ (80%) 13 30/F Ovary Omentum 4þ (>90%) 4þ (>90%) 0 1þ (20%-25%) 3þ (70%) 14 23/F Ovary Colon 4þ (>95%) 4þ (>95%) 0 1þ (1%-2%) /F Ovary Pelvis 4þ (95%) 4þ (>90%) 0 2þ (40%-50%) 0 EMA indicates epithelial membrane protein; AFP, a-fetoprotein; PLAP, placental-like alkaline phosphatase; M, male; F, female. * (xx%) ¼ xx% tumor cells positive. the staining intensity varied from weak to strong (Table 2) (Fig. 3). Yolk Sac Tumors We identified 15 metastatic yolk sac tumors (8 from the testis, 4 from the ovary, 2 from the sacrum, and 1 from the liver) (Table 3). SALL4 staining was strongly 4þ (>90% tumor cells) positive in 14 cases and was strongly 3þ (approximately 80% tumor cells stained) positive in 1 case (Fig. 4). For comparison, immunostains for pan cytokeratin, EMA, AFP, and PLAP were performed in these cases. All metastatic yolk sac tumors were strongly (4þ in 14, 3þ in 1) positive for pan cytokeratin (Fig. 4). The majority of the yolk sac tumors were negative for EMA, except in 4 cases in which each contained approximately 0.5% to 3% (1þ) EMA-positive tumor cells. AFP was positive in 13 of 15 cases (1þ in 10 cases, 2þ in 2 cases, and 3þ in 1 case), and the percentage of AFP-positive cells in these 13 cases ranged from <1% to between 70% and 80% (mean, 15%-18%) (Fig. 4). Two cases were negative for AFP staining (Fig. 4). The AFP staining was patchy in most cases. No metastatic yolk sac tumor demonstrated 4þ AFP staining. Only 6 of these 15 (40%) yolk sac tumors were positive for PLAP staining: 1þ in 3 tumors, 2þ in 1 tumor, and 3þ in 2 tumors. Comparison of SALL4 and OCT4 in Detecting Germ Cell Tumors We compared SALL4 staining with that of OCT4 in all these metastatic germ cell tumors. The results are Cancer June 15,

7 Original Article FIGURE 4. Immunohistochemical staining of SALL4 is shown in a metastatic yolk sac tumor in the brain (A1 and A2), demonstrating diffuse pan cytokeratin (CK) positivity (A3) but negative epithelial membrane protein (EMA) staining (A4). a-fetoprotein (AFP) stained the majority of the tumor cells (A5), whereas the tumor cells were negative for placental-like alkaline phosphatase (PLAP) (A6) and OCT4 (A7). In contrast, >95% of tumor cells were found to be strongly positive for SALL4 (A8). Occasionally, metastatic yolk sac tumor (B1-B2 to lymph node, C1-C2 to lung) demonstrated only focal (B3) or even negative (C3) AFP staining, but the tumor cells demonstrated strong and diffuse SALL4 staining (B4 and C4). H&E indicates hematoxylin and eosin. Table 4. Comparison of Immunohistochemical Staining of SALL4 and OCT4 in Germ Cell Tumors Tumor Type seminomas (n¼22) dysgerminomas (n¼7) embryonal carcinomas (n¼21) yolk sac tumors (n¼15) teratomas (n¼18) choriocarcinomas (n¼7) SALL4 Positive OCT4 Positive P 22/22 (100%) 22/22 (100%) /7 (100%) 7/7 (100%) /21 (100%) 21/21 (100%) /15 (100%) 0/15 (0%) < /18 (50%) 0/18 (0%) /7 (71%) 0/7 (0%).02 summarized in Table 4. All 22 metastatic seminomas were positive for OCT4, but 2 of them demonstrated only 2þ variable weak and strong staining (31%-60% tumor cells), and 4 demonstrated 3þ staining, in which there was only focal strong intensity staining and the majority of the tumor was weak (61%-90% tumor cells). Of 21 metastatic embryonal carcinomas, 2 demonstrated 3þ strong staining, and 19 demonstrated 4þ strong staining. All 7 metastatic dysgerminomas demonstrated 4þ strong OCT4 staining. OCT 4 staining was not found to be present in any of the 15 metastatic yolk sac tumors, 7 metastatic nongestational choriocarcinomas, or 18 metastatic teratomas. Nongerm Cell Tumors From Various Sites Among the 170 cases of metastatic somatic tumors from various sites, only 6 of 18 metastatic gastric adenocarcinomas, 3 of 15 metastatic esophageal adenocarcinomas, and 1 of 13 colonic adenocarcinomas demonstrated 1þ weak nuclear SALL4 staining (Table 5). Of the 6 SALL4-positive gastric adenocarcinomas, the percentage of tumor cells stained with SALL4 ranged from 0.5% to 25% (average, 12%). The percentage of SALL4-positive tumor cells 2646 Cancer June 15, 2009

8 Diagnostic Marker for Germ Cell Tumors/Cao et al Table 5. Immunohistochemical Stain of SALL4 in Non germ Cell Tumors Tumor Type No. of Cases SALL4-positive Cases Head and neck squamous 6 0 cell carcinoma Thyroid carcinoma 7 0 Lung carcinoma 12 0 Breast carcinoma 8 0 Esophageal adenocarcinoma 15 3 (all 1þ weak) Gastric adenocarcinoma 18 6 (all 1þ weak) Pancreatic adenocarcinoma 10 0 Ampullary adenocarcinoma 2 0 Hepatocellular carcinoma 7 0 Cholangiocarcinoma 3 0 Colonic adenocarcinoma 13 1 (1þ weak) Renal cell carcinoma 10 0 Urothelial carcinoma 10 0 Ovarian carcinoma 18 0 Uterine adenocarcinoma 6 0 Prostatic adenocarcinoma 12 0 Melanoma 12 0 Total /170 (6.0%) (all 1þ weak) in 3 esophageal adenocarcinomas ranged from 5% to 25% (mean 12%). Approximately 10% of tumor cells were weakly positive for SALL4 in 1 metastatic colonic adenocarcinoma. All 10 metastatic carcinomas with focal weak staining of SALL4 demonstrated strong EMA staining in >50% of the tumor cells. Other metastatic somatic malignancies, including head and neck squamous cell carcinomas, thyroid carcinomas, lung carcinomas, breast carcinomas, hepatocellular carcinomas, cholangiocarcinomas, ampullary adenocarcinomas, pancreatic adenocarcinomas, renal cell carcinomas, urothelial carcinomas, prostatic adenocarcinomas, ovarian carcinomas, uterine carcinomas, colonic adenocarcinomas, and melanomas, were all negative for SALL4. DISCUSSION SALL4 is a zinc finger transcription factor and homologous to the Drosophila spalt (sal) gene. 31 In Drosophila, sal acts as a region-specific homeotic gene involved in the specification of head and tail regions during embryonal development. 32 In mice, Sall4 is essential to early embryogenesis, and homozygous mutant mice exhibit early embryonic lethality. 33,34 In humans, SALL4 is located on chromosome 20q As in other species, SALL4 is essential to human development, and mutations in SALL4 lead to acro renal-ocular and Okihiro syndromes. 33,35 In human embryonic stem cells, SALL4 is essential to maintain embryonal stem cell pluripotency and self-renewal by forming a regulatory network with OCT4, NANOG, and SOX2. 26,28-30,36,37 In this study, SALL4 was found to be strongly positive in >90% of tumor cells in all metastatic seminomas, dysgerminomas, embryonal carcinomas, and 14 of 15 yolk sac tumors. Recently, we also found that SALL4 is strongly positive in these types of tumors in the testis and ovary (unpublished data). Strong expression of SALL4 in these germ cell tumors suggests that SALL4 might play some role in pathogenesis of germ cell tumors, but to our knowledge the underlying mechanism is unknown. To the best of our knowledge, the current study is the first to link SALL4 to germ cell tumors. In contrast to the diffuse and strong positive SALL4 staining reported in all metastatic seminomas, dysgerminomas, embryonal carcinomas, and yolk sac tumors, only focal weak SALL4 staining was observed in rare metastatic carcinomas among the 170 metastatic somatic malignancies. These findings indicate that SALL4 is a novel sensitive and highly specific marker for these metastatic germ cell tumors. These findings also demonstrate important diagnostic utilities. First, SALL4 can be used to confirm the germ cell origin of a metastatic tumor. This is important because an estimated 3% to 5% of newly diagnosed cancers present with unknown primary sites. 38 somatic malignancies and germ cell tumors are managed differently and vary in their prognosis. The majority of germ cell tumors can be effectively treated and even cured with modern therapy. 3-5 Therefore, it is critical to distinguish metastatic germ cell tumors from nongerm cell tumors. Because of morphologic overlapping, it is highly desirable to have a marker that is sensitive and specific, and that can be used in routine pathology practice. 39,40 Earlier markers such as PLAP, AFP, C-KIT, and CD30 lack adequate sensitivity and specificity for metastatic germ cell tumors Recently, novel stem cell markers such as OCT4, NANOG, and SOX2 have emerged as more sensitive and specific markers for metastatic germ cell tumors (seminoma and embryonal carcinoma), but to our knowledge they do not label any metastatic yolk sac tumors 9,11,24,25,41 ; this is a significant deficiency because these tumors can mimic many other tumor types. In the current study, all metastatic seminomas, dysgerminomas, Cancer June 15,

9 Original Article FIGURE 5. Diagnostic utility of SALL4 in a metastatic yolk sac tumor is shown in limited material. In this core needle biopsy sample, metastatic yolk sac tumor was present in the omentum (A) and demonstrated glandular and papillary structures with eosinophilic and clear cytoplasm (B). The tumor cells demonstrated diffuse positivity for pan cytokeratin (CK) (C), and rare tumor cells were also positive for epithelial membrane protein (EMA) (D). a-fetoprotein (AFP) (E) labeled only rare tumor cells. Placentallike alkaline phosphatase (PLAP) (F) and OCT4 (G) stains were negative. In contrast, SALL4 stain was found to be strongly positive in >95% of tumor cells (H and I). H&E indicates hematoxylin and eosin. embryonal carcinomas, and yolk sac tumors were found to be strongly positive for SALL4 in >90% of tumor cells; only 1 case of yolk sac tumor had 80% tumors cells found to be strongly positive for SALL4. In this sense, SALL4 demonstrated 100% sensitivity in detecting metastatic seminoma, dysgerminoma, embryonal carcinoma, and yolk sac tumor. It is highly likely that a metastatic tumor that is negative for SALL4 staining is not a seminoma, dysgerminoma, embryonal carcinoma, or yolk sac tumor (the negative predictive value is 100%). However, because in a few of these cases <100% of tumor cells stain for SALL4, the slim possibility still exists that an extremely small biopsy comprised of only a few cells might harbor the few negative cells in an otherwise SALL4-positive tumor. With regard to specificity, although a few metastatic carcinomas (10 of 170 in this study) demonstrated focal weak staining, SALL4 demonstrated 100% specificity for metastatic seminoma, dysgerminoma, embryonal carcinoma, and yolk sac tumor in this study if strong and diffuse (at least 50% tumor cells positive) staining was used as an indicator of positivity. Further studies including more tumors from various organs are warranted to address the sensitivity and specificity of SALL4. Compared with OCT4, NANOG, and SOX2, 9,11,24,25 SALL4 is most likely a better marker in detection of metastatic germ cell tumors, because it not only labels seminoma/dysgerminoma and embryonal carcinoma, but also yolk sac tumor and choriocarcinoma. Even in metastatic seminoma and embryonal carcinoma, SALL4 stains more tumor cells than OCT4. OCT4, NANOG, and SOX2 do not label metastatic yolk sac tumor or choriocarcinoma. 9,11,24,25 Second and most importantly, SALL4 is a novel diagnostic marker for metastatic yolk sac tumors from the testis, ovary, and extragonadal sites. Yolk sac tumor notoriously displays multiple histologic patterns, and thus can mimic or be mimicked by many other types of tumors To our knowledge to date, no specific immunohistochemical marker has been identified for metastatic yolk 2648 Cancer June 15, 2009

10 Diagnostic Marker for Germ Cell Tumors/Cao et al sac tumors. Earlier diagnostic markers for yolk sac tumors included PLAP and AFP. PLAP is neither highly sensitive nor specific for yolk sac tumor or other germ cell tumors. 10,11 Many nongerm cell tumors have been reported to be immunohistochemically positive for PLAP. 12,13,46 In this study, only 6 of 15 (40%) metastatic yolk sac tumors were found to be positive for PLAP staining, which is similar to other studies. 10,13,19 Three of 6 PLAP-positive cases in the current study demonstrated only 1% to 2% cells stained with this marker. Although AFP is considered a useful marker for yolk sac tumor, its diagnostic utility is also limited by its low sensitivity. 19,20 In addition, its staining is characteristically focal and patchy, and therefore may be negative in limited biopsy specimens, as noted in the current study (Fig. 5). 19,20,22 Furthermore, AFP may be lost in metastatic foci even if the primary tumor is positive for AFP. 23 In our study, although 13 of 15 metastatic yolk sac tumors were found to be positive for AFP, the average percentage of cells staining with AFP was only 15% to 20%, and 3 of them demonstrated only 1% to 2% of cells staining. Furthermore, AFP lacks specificity for yolk sac tumor because AFP positivity has been reported in other types of tumors, such as ovarian serous 47 and clear cell carcinomas, 21 pancreatic adenocarcinoma, 48 and hepatocellular carcinoma. 49 In contrast to AFP and PLAP, the results of the current study indicated that SALL4 was strongly positive in >90% of tumor cells in 14 of 15 metastatic yolk sac tumors and strongly positive in 80% of tumor cells in the remaining case (100% sensitivity). It also demonstrated high specificity. With the exception of a few metastatic carcinomas demonstrating focal weak staining, the remaining nongerm cell malignancies were entirely negative for SALL4. Therefore, SALL4 appears to be a better marker than AFP and PLAP for diagnosing metastatic yolk sac tumor, especially in limited biopsy specimens (Fig. 5). In the ovary, yolk sac tumors are often negative for EMA and CK7, helping to distinguish them from clear cell carcinoma. 50 However, negative staining for EMA and CK7 is not specific or confirmatory for yolk sac tumor, and this pattern can be observed in many types of tumors, especially in metastatic foci. In the literature, the only other types of malignancies reported to be positive for SALL4 are precursor B-cell lymphoblastic lymphoma and acute myeloid leukemia. 51,52 Morphologically, metastatic yolk sac tumor or other germ cell tumors should be easily distinguished from these entities. In difficult cases, an immunohistochemical panel including antibodies toward CD10, CD19, CD79a, and TDT for precursor B lymphoblastic lymphoma and myeloperoxidase, leukocyte common antigen (CD45), CD43, and CD34 for myeloid leukemia should help in making the distinction from germ cell tumor. In summary, we investigated the expression of the novel stem cell marker SALL4 in a large series of 90 metastatic germ cell tumors from the testis, ovary, and extragonadal sites. SALL4 was found to be strongly positive in all metastatic seminomas, dysgerminomas, embryonal carcinomas, and yolk sac tumors. In contrast, only rare metastatic nongerm cell tumors demonstrated focal weak SALL4 staining. The results indicate that SALL4 is a novel sensitive and specific marker for metastatic germ cell tumors and is particularly useful for detecting metastatic yolk sac tumors. Conflict of Interest Disclosures The authors made no disclosures. References 1. Nakatani T, Sugimura K, Naganuma T, et al. urinary bladder tumor from extragonadal germ cell tumor: a case report. Oncol Rep. 2002;9: Satake I, Tari K, Nakagomi K, et al. Extragonadal germ cell tumor of retroperitoneal origin: report of 2 cases. Jpn J Clin Oncol. 1994;24: Einhorn L. Chemotherapeutic and surgical strategies for germ cell tumors. Chest Surg Clin N Am. 2002;12: Gershenson DM. Management of ovarian germ cell tumors. J Clin Oncol. 2007;25: Pectasides D, Pectasides E, Kassanos D. Germ cell tumors of the ovary. Cancer Treat Rev. 2008;34: Azzopardi JG, Hoffbrand AV. Regression in testicular seminomas with viable metastases. J Clin Pathol. 1965;18: Balzer BL, Ulbright TM. Spontaneous regression of testicular germ cell tumors: an analysis of 42 cases. Am J Surg Pathol. 2006;30: Ulbright TM. Testis risk and prognostic factors. The pathologist s perspective. Urol Clin North Am. 1999;26: Sung MT, MacLennan GT, Cheng L. Retroperitoneal seminoma in limited biopsies: morphologic criteria and Cancer June 15,

11 Original Article immunohistochemical findings in 30 cases. Am J Surg Pathol. 2006;30: Weissbach L, Bussar-Maatz R, Mann K. The value of tumor markers in testicular seminomas. Results of a prospective multicenter study. Eur Urol. 1997;32: Iczkowski KA, Butler SL, Shanks JH, et al. Trials of new germ cell immunohistochemical stains in 93 extragonadal and metastatic germ cell tumors. Hum Pathol. 2008;39: Hamilton-Dutoit SJ, Lou H, Pallesen G. The expression of placental alkaline phosphatase (PLAP) and PLAP-like enzymes in normal and neoplastic human tissues. An immunohistological survey using monoclonal antibodies. APMIS. 1990;98: Wick MR, Swanson PE, Manivel JC. Placental-like alkaline phosphatase reactivity in human tumors: an immunohistochemical study of 520 cases. Hum Pathol. 1987;18: Kneile JR, Tan G, Suster S, Wakely PE Jr. Expression of CD30 (Ber-H2) in nasopharyngeal carcinoma, undifferentiated type and lymphoepithelioma-like carcinoma. A comparison study with anaplastic large cell lymphoma. Histopathology. 2006;48: Polski JM, Janney CG. Ber-H2 (CD30) immunohistochemical staining in malignant melanoma. Mod Pathol. 1999;12: Berney DM, Shamash J, Pieroni K, Oliver RT. Loss of CD30 expression in metastatic embryonal carcinoma: the effects of chemotherapy? Histopathology. 2001;39: Rajpert-De Meyts E, Skakkebaek NE. Expression of the c- kit protein product in carcinoma-in-situ and invasive testicular germ cell tumours. Int J Androl. 1994;17: Sever M, Jones TD, Roth LM, et al. Expression of CD117 (c-kit) receptor in dysgerminoma of the ovary: diagnostic and therapeutic implications. Mod Pathol. 2005;18: Niehans GA, Manivel JC, Copland GT, Scheithauer BW, Wick MR. Immunohistochemistry of germ cell and trophoblastic neoplasms. Cancer. 1988;62: Eglen DE, Ulbright TM. The differential diagnosis of yolk sac tumor and seminoma. Usefulness of cytokeratin, alphafetoprotein, and alpha-1-antitrypsin immunoperoxidase reactions. Am J Clin Pathol. 1987;88: Esheba GE, Pate LL, Longacre TA. Oncofetal protein glypican-3 distinguishes yolk sac tumor from clear cell carcinoma of the ovary. Am J Surg Pathol. 2008;32: Jacobsen GK, Norgaard-Pedersen B. Placental alkaline phosphatase in testicular germ cell tumours and in carcinoma-in-situ of the testis. An immunohistochemical study. Acta Pathol Microbiol Immunol Scand [A]. 1984;92: Mostofi FK. Histological change ostensibly induced by therapy in the metastasis of germ cell tumors of testis. Prog Clin Biol Res. 1985;203: Cheng L. Establishing a germ cell origin for metastatic tumors using OCT4 immunohistochemistry. Cancer. 2004; 101: Santagata S, Ligon KL, Hornick JL. Embryonic stem cell transcription factor signatures in the diagnosis of primary and metastatic germ cell tumors. Am J Surg Pathol. 2007;31: Zhou Q, Chipperfield H, Melton DA, Wong WH. A gene regulatory network in mouse embryonic stem cells. Proc Natl Acad Sci U S A. 2007;104: Wang J, Rao S, Chu J, et al. A protein interaction network for pluripotency of embryonic stem cells. Nature. 2006;444: Loh YH, Wu Q, Chew JL, et al. The Oct4 and Nanog transcription network regulates pluripotency in mouse embryonic stem cells. Nat Genet. 2006;38: Zhang J, Tam WL, Tong GQ, et al. Sall4 modulates embryonic stem cell pluripotency and early embryonic development by the transcriptional regulation of Pou5f1. Nat Cell Biol. 2006;8: Wong CC, Gaspar-Maia A, Ramalho-Santos M, Reijo Pera RA. High-efficiency stem cell fusion-mediated assay reveals Sall4 as an enhancer of reprogramming. PLoS ONE. 2008;3:e Reuter D, Schuh R, Jackle H. The homeotic gene spalt (sal) evolved during Drosophila speciation. Proc Natl Acad Sci U S A. 1989;86: Jurgens G. Head and tail development of the Drosophila embryo involves spalt, a novel homeotic gene. EMBO J. 1988;7: Sakaki-Yumoto M, Kobayashi C, Sato A, et al. The murine homolog of SALL4, a causative gene in Okihiro syndrome, is essential for embryonic stem cell proliferation, and cooperates with Sall1 in anorectal, heart, brain and kidney development. Development. 2006;133: Warren M, Wang W, Spiden S, et al. A Sall4 mutant mouse model useful for studying the role of Sall4 in early embryonic development and organogenesis. Genesis. 2007; 45: Kohlhase J, Heinrich M, Schubert L, et al. Okihiro syndrome is caused by SALL4 mutations. Hum Mol Genet. 2002;11: Li SS, Liu YH, Tseng CN, Chung TL, Lee TY, Singh S. Characterization and gene expression profiling of 5 new human embryonic stem cell lines derived in Taiwan. Stem Cells Dev. 2006;15: Wu Q, Chen X, Zhang J, et al. Sall4 interacts with Nanog and co-occupies Nanog genomic sites in embryonic stem cells. J Biol Chem. 2006;281: Haineswoth JD, Greco FA. Management of patients with cancer of unknown primary site. Oncology. 2000;14: Kirsten F, Chi CH, Leary JA, Ng AB, Hedley DW, Tattersall MH. adeno or undifferentiated carcinoma from an unknown primary site natural history and 2650 Cancer June 15, 2009

12 Diagnostic Marker for Germ Cell Tumors/Cao et al guidelines for identification of treatable subsets. Q J Med. 1987;62: Hillen HF. Unknown primary tumours. Postgrad Med J. 2000;76: Jones TD, Ulbright TM, Eble JN, Baldridge LA, Cheng L. OCT4 staining in testicular tumors: a sensitive and specific marker for seminoma and embryonal carcinoma. Am J Surg Pathol. 2004;28: Kurman RJ, Norris HJ. Endodermal sinus tumor of the ovary: a clinical and pathologic analysis of 71 cases. Cancer. 1976;38: Ulbright TM. Germ cell tumors of the gonads: a selective review emphasizing problems in differential diagnosis, newly appreciated, and controversial issues. Mod Pathol. 2005;18(suppl 2):S61-S Clement PB, Young RH, Scully RE. Endometrioid-like variant of ovarian yolk sac tumor. A clinicopathological analysis of 8 cases. Am J Surg Pathol. 1987;11: Prat J, Bhan AK, Dickersin GR, Robboy SJ, Scully RE. Hepatoid yolk sac tumor of the ovary (endodermal sinus tumor with hepatoid differentiation): a light microscopic, ultrastructural and immunohistochemical study of 7 cases. Cancer. 1982;50: Saad RS, Landreneau RJ, Liu Y, Silverman JF. Utility of immunohistochemistry in separating thymic neoplasms from germ cell tumors and metastatic lung cancer involving the anterior mediastinum. Appl Immunohistochem Mol Morphol. 2003;11: Higuchi Y, Kouno T, Teshima H, et al. Serous papillary cystadenocarcinoma associated with alpha-fetoprotein production. Arch Pathol Lab Med. 1984;108: Nojima T, Kojima T, Kato H, Sato T, Koito K, Nagashima K. Alpha-fetoprotein-producing acinar cell carcinoma of the pancreas. Hum Pathol. 1992;23: Sato K, Tanaka M, Kusaba T, Fukuda H, Tanikawa K. Immunohistochemical demonstration of alpha-fetoprotein in small hepatocellular carcinoma. Oncol Rep. 1998;5: Ramalingam P, Malpica A, Silva EG, Gershenson DM, Liu JL, Deavers MT. The use of cytokeratin 7 and EMA in differentiating ovarian yolk sac tumors from endometrioid and clear cell carcinomas. Am J Surg Pathol. 2004;28: Cui W, Kong NR, Ma Y, Amin HM, Lai R, Chai L. Differential expression of the novel oncogene, SALL4, in lymphoma, plasma cell myeloma, and acute lymphoblastic leukemia. Mod Pathol. 2006;19: Yang J, Chai L, Gao C, et al. SALL4 is a key regulator of survival and apoptosis in human leukemic cells. Blood. 2008;112: Cancer June 15,