Primary germ cell tumors of the central nervous system (CNS) sometimes pose diagnostic difficulty. In this study we analyzed the diagnostic utility of a novel marker, SALL4, in 77 such tumors (59 pure and 18 mixed) consisting of the following tumors/tumor components: 49 germinomas, 7 embryonal carcinomas, 27 yolk sac tumors, 3 choriocarcinomas, and 14 teratomas. We also stained SALL4 in 99 primary non-germ cell tumors to test SALL4 specificity. We compared SALL4 with OCT4 in all germ cell tumors and compared SALL4 with a-fetoprotein and glypican-3 in all yolk sac tumors. The staining was semiquantitatively scored as 0 (no staining), 1 þ (o ¼ 30%), 2 þ (31-60%), 3 þ (61-90%), and 4 þ (490%). Strong SALL4 staining was observed in all 49 germinomas (4 þ in 48, 3 þ in 1), 7 embryonal carcinomas (all 4 þ ), and 27 yolk sac tumors (1 þ in 1, 2 þ in 2, 3 þ in 7, 4 þ in 17). SALL4 staining, 1 þ weak to focally strong, was observed in 2 of 3 choriocarcinomas (in mononucleated trophoblasts) and in 9 of 14 teratomas (in primitive neuroepithelium and teratomatous glands). All germinomas and embryonal carcinomas showed strong OCT4 staining (4 þ in all except 1 germinoma with 3 þ ), whereas other germ cell tumors were negative. Out of 27 yolk sac tumors, 26 showed positive a-fetoprotein staining (1 þ in 9, 2 þ in 7, 3 þ in 5, and 4 þ in 5). All yolk sac tumors showed positive glypican-3 staining (1 þ in 6, 2 þ in 6, 3 þ in 7, and 4 þ in 8). The mean percentage of yolk sac tumor cells stained was 84% with SALL4, 45% with a-fetoprotein, and 63% with glypican-3 (Po0.01). No non-germ cell tumors showed SALL4 staining. Our results indicate that SALL4 is a novel sensitive diagnostic marker for primary germ cell tumors of the CNS with high specificity. SALL4 is a more sensitive marker than a-fetoprotein and glypican-3 for yolk sac tumors.

LIN28 has been shown to have an important role in primordial germ cell development and malignant transformation of germ cells in mouse. In this study, we examined the immunohistochemical profile of LIN28 in 131 primary human extragonadal germ cell tumors (central nervous system (CNS) 76, mediastinum 17, sacrococcygeal region 30, pelvis 3, vagina 2, liver 1, omentum 1, and retroperitoneum 1), including the following tumors and/or components: 57 seminomas/germinomas, 10 embryonal carcinomas, 74 yolk sac tumors, 6 choriocarcinomas, 15 mature, and 13 immature teratomas. We compared LIN28 with SALL4 to assess its diagnostic value. To determine its specificity, we examined LIN28 in 406 extragonadal-non-germ cell tumors (103 carcinomas, 91 sarcomas, 9 melanomas, 12 mesotheliomas, 83 lymphomas, 9 plasmacytomas, 82 CNS tumors, and 17 thymic epithelial tumors). The staining was semi-quantitatively scored as 0 (no cell stained), 1 þ (0-30%), 2 þ (31-60%), 3 þ (61-90%), and 4 þ (490%). LIN28 staining was seen in all seminomas/germinomas (3 þ in 1 and 4 þ in 56), embryonal carcinomas (4 þ in all 10), and yolk sac tumors (3 þ in 3 and 4 þ in 71). Variable LIN28 staining was seen in 5 of 6 choriocarcinomas (1 þ to 4 þ ), 8 of 13 immature teratomas (1 þ to 2 þ in immature elements), and in 1 of 15 mature teratomas (1 þ ). Only 11 of 406 non-germ cell tumors showed 1 þ LIN28 staining. Therefore, LIN28 is a sensitive (100% sensitivity) marker for primary extragonadal seminomas/germinomas, embryonal carcinomas, and yolk sac tumors with high specificity. Compared with SALL4, LIN28 demonstrated a similar level of diagnostic sensitivity for seminomas/ germinomas and embryonal carcinomas. For primary extragonadal yolk sac tumors, although SALL4 stained all tumors (1 þ in 1, 2 þ in 2, 3 þ in 10, and 4 þ in 61), LIN28 stained more tumor cells (mean 95 vs 90%, P ¼ 0.03) and was therefore more sensitive. For primary extragonadal yolk sac tumors, combining LIN28 and SALL4 can achieve a higher diagnostic sensitivity than either alone.

No abstract