Human testicular (non)seminomatous germ cell tumours: the clinical implications of recent pathobiological insights.

Human germ cell tumours (GCTs) comprise several types of neoplasias with different pathogeneses and clinical behaviours. A classification into five subtypes has been proposed. Here, the so-called type II testicular GCTs (TGCTs), ie the seminomas and non-seminomas, will be reviewed with emphasis on pathogenesis and clinical implications. Various risk factors have been identified that define subpopulations of men who are amenable to early diagnosis. TGCTs are omnipotent, able to generate all differentiation lineages, both embryonic and extra-embryonic, as well as the germ cell lineage itself. The precursor lesion, composed of primordial germ cells/gonocytes, is referred to as carcinoma in situ of the testis (CIS) and gonadoblastoma of the dysgenetic gonad. These pre-malignant cells retain embryonic characteristics, which probably explains the unique responsiveness of the derived tumours to DNA-damaging agents. Development of CIS and gonadoblastoma is crucially dependent on the micro-environment created by Sertoli cells in the testis, and granulosa cells in the dysgenetic gonad. OCT3/4 has high sensitivity and specificity for CIS/gonadoblastoma, seminoma, and embryonal carcinoma, and is useful for the detection of CIS cells in semen, thus a promising tool for non-invasive screening. Overdiagnosis of CIS due to germ cell maturation delay can be avoided using immunohistochemical detection of stem cell factor (SCF). Immunohistochemistry is helpful in making the distinction between seminoma and embryonal carcinoma, especially SOX17 and SOX2. The different non-seminomatous histological elements can be recognized using various markers, such as AFP and hCG, while others need confirmation. The value of micro-satellite instability as well as BRAF mutations in predicting treatment resistance needs validation in prospective trials. The availability of representative cell lines, both for seminoma and for embryonal carcinoma, allows mechanistic studies into the initiation and progression of this disease.

FOXL2 and SOX9 as parameters of female and male gonadal differentiation in patients with various forms of disorders of sex development (DSD).

The transcription factors SOX9 and FOXL2 are required for male and female mammalian gonadal development. We have used specific antibodies to investigate the role of these key proteins in disorders of sex development (DSD), specifically inter-sex states. In normal gonads, SOX9 was found to be restricted to the presence of (pre-)Sertoli cells, while FOXL2 was found in granulosa cells, and in stromal cells interpreted as early ovarian stroma. Both proteins were found within a single patient, when testicular and ovarian development was present; and within the same gonad, when both differentiation lineages were identified, as in ovotesticular DSD (ie hermaphrodite). Especially SOX9 was informative to support the presence of early testicular development (ie seminiferous tubules), expected based on morphological criteria only. In a limited number of DSD cases, FOXL2 was found within reasonably well-developed seminiferous tubules, but double staining demonstrated that it was never strongly co-expressed with SOX9 in the same cell. All seminiferous tubules containing carcinoma in situ (CIS), the malignant counterpart of a primordial germ cell, ie the precursor of type II germ cell tumours of the testis, seminomas and non-seminomas, showed the presence of SOX9 and not FOXL2. In contrast, gonadoblastomas (GBs), the precursor of the same type of cancer, in a dysgenetic gonad, showed expression of FOXL2 and no, or only very low, SOX9 expression. These findings indicate that gonadal differentiation, ie testicular or ovarian, determines the morphology of the precursor of type II germ cell tumours, CIS or GB, respectively. We show that in DSD patients, the formation of either ovarian or/and testicular development can be visualized using FOXL2 and SOX9 expression, respectively. In addition, it initiates a novel way to study the role of the supportive cells in the development of either CIS or GB.

Ovarian dysgerminomas are characterised by frequent KIT mutations and abundant expression of pluripotency markers.

BACKGROUND: Ovarian germ cell tumours (OGCTs) typically arise in young females and their pathogenesis remains poorly understood. We investigated the origin of malignant OGCTs and underlying molecular events in the development of the various histological subtypes of this neoplasia. RESULTS: We examined in situ expression of stem cell-related (NANOG, OCT-3/4, KIT, AP-2gamma) and germ cell-specific proteins (MAGE-A4, NY-ESO-1, TSPY) using a tissue microarray consisting of 60 OGCT tissue samples and eight ovarian small cell carcinoma samples. Developmental pattern of expression of NANOG, TSPY, NY-ESO-1 and MAGE-A4 was determined in foetal ovaries (gestational weeks 13-40). The molecular genetic part of our study included search for the presence of Y-chromosome material by fluorescence in situ hybridisation (FISH), and mutational analysis of the KIT oncogene (exon 17, codon 816), which is often mutated in testicular GCTs, in a subset of tumour DNA samples. We detected a high expression of transcription factors related to the embryonic stem cell-like pluripotency and undifferentiated state in OGCTs, but not in small cell carcinomas, supporting the view that the latter do not arise from a germ cell progenitor. Bilateral OGCTs expressed more stem cell markers than unilateral cases. However, KIT was mutated in 5/13 unilateral dysgerminomas, whereas all bilateral dysgerminomas (n = 4) and all other histological types (n = 22) showed a wild type sequence. Furthermore, tissue from five phenotypic female patients harbouring combined dysgerminoma/gonadoblastoma expressed TSPY and contained Y-chromosome material as confirmed by FISH. CONCLUSION: This study provides new data supporting two distinct but overlapping pathways in OGCT development; one involving spontaneous KIT mutation(s) leading to increased survival and proliferation of undifferentiated oogonia, the other related to presence of Y chromosome material and ensuing gonadal dysgenesis in phenotypic females.

Gonadoblastoma arising in undifferentiated gonadal tissue within dysgenetic gonads.

PURPOSE: The purpose of the study was to define the histological origin of gonadoblastomas, allowing the identification of high-risk patients. EXPERIMENTAL DESIGN: Sixty paraffin-embedded gonadectomy or gonadal biopsy samples of 43 patients with gonadal dysgenesis were selected from our archives. We studied the morphology and immunohistochemical properties of the germ cells in 40 samples without neoplastic transformation and compared these findings with the morphological and immunohistochemical characteristics of 20 samples containing gonadoblastoma/dysgerminoma. RESULTS: The overall incidence of germ cell tumors in our patient series was 35%. In dysgenetic gonads without germ cell neoplasia, besides the presence of areas with testicular and/or ovarian differentiation, areas of undifferentiated gonadal tissue were identified in 13 of 40 samples (32.5%). A subpopulation of germ cells within these undifferentiated areas stained positive for octamer binding transcription factor (OCT)3/4, the stem cell factor receptor, placental-like alkaline phosphatase, and testis-specific protein-Y encoded. Gonadoblastoma germ cells display identical staining results. Moreover, in gonads containing gonadoblastoma, adjacent to this lesion, areas of undifferentiated gonadal tissue with identical immunohistochemical characteristics were identified in 10 of 20 samples (50%). No adjacent tissue was available in five cases, whereas in the five remaining cases, it consisted of streak tissue. In three cases, an accumulation of OCT3/4-positive germ cells in the proximity of the malignant lesions was found, suggesting clonal expansion and final organization into gonadoblastoma nests. CONCLUSIONS: Based on these observations, we hypothesize that gonadoblastomas originate from surviving OCT3/4-positive germ cells in areas of undifferentiated gonadal tissue within the dysgenetic gonad. Supportive evidence was obtained that carcinoma in situ arises in regions with testicular differentiation.

Utility of immunostaining for S-100 protein subunits in gonadal sex cord-stromal tumors, with emphasis on the large-cell calcifying Sertoli cell tumor of the testis.

This study concerns the immunohistochemical localization of S-100 alpha, S-100 beta, and whole brain S-100 (wbS-100) in testicular large-cell calcifying Sertoli cell tumor (LCCSCT). We examined 8 LCCSCTs (7 benign and 1 malignant), 6 Sertoli cell tumors not otherwise specified (SCTs-NOS), 6 Leydig cell tumors (LCTs), 5 ovarian Sertoli-Leydig cell tumors (SLCTs), and 7 gonadoblastomas (GBLs). The 8 LCCSCTs showed immunoreactivity for S-100 alpha, S-100 beta, and wbS-100. Five of the 6 LCTs and the Leydig cell components in the ovarian SLCTs stained positively for S-100 alpha and wbS-100 but were negative for S-100 beta. SCTs-NOS and the Sertoli cell components in the SLCTs occasionally showed focal and weak/moderate positivity for S-100 alpha, S-100 beta, and wbS-100. Sex cord cells of the GBLs were positive for S-100 beta and wbS-100 and negative for S-100 alpha. Germ cell elements of the GBLs were negative for S-100 alpha, S-100 beta, and wbS-100. In nonneoplastic testicular parenchyma adjacent to the above-mentioned tumors, there was S-100 alpha reactivity in Leydig cells, rete testis, and a few Sertoli cells. S-100 beta reactivity was seen in a few Sertoli cells, Schwann cells, and some endothelial cells. WbS-100 reactivity was present in Leydig cells, a few Sertoli cells, rete testis, Schwann cells, and some endothelial cells. The results indicate that S-100 alpha and S-100 beta can potentially be used as immunohistochemical markers for LCCSCT, especially when differentiating it from LCT, which may mimic LCCSCT on routine histopathology. Although the biological significance of both S-100 subunits expression in LCCSCT remains unknown, these notable calcium-binding proteins may be associated with the characteristic calcification in LCCSCT through regulation of calcium levels in the tumor cells.

Expression of a candidate gene for the gonadoblastoma locus in gonadoblastoma and testicular seminoma.

The gonadoblastoma locus on the Y chromosome (GBY) predisposes the dysgenetic gonads of XY females to develop in situ tumors. It has been mapped to a critical interval on the short arm and adjacent centromeric region on the Y chromosome. Currently there are five functional genes identified on the GBY critical region, thereby providing likely candidates for this cancer predisposition locus. To evaluate the candidacy of one of these five genes, testis-specific protein Y-encoded (TSPY), as the gene for GBY, expression patterns of TSPY in four gonadoblastoma from three patients were analyzed by immunohistochemistry using a TSPY specific antibody. Results from this study showed that TSPY was preferentially expressed in tumor germ cells of all gonadoblastoma specimens. Additional study on two cases of testicular seminoma demonstrated that TSPY was also abundantly expressed in all stages of these germ cell tumors. The present observations suggest that TSPY may either be involved in the oncogenesis of or be a useful marker for both types of germ cell tumors.

Gonadoblastoma: immunohistochemical localization of Müllerian-inhibiting substance, inhibin, WT-1, and p53.

Gonadoblastomas are rare tumors composed of both germ cells and sex-cord cells. In this study, we investigated two such tumors for the expression of the Wilms' tumor gene (WT1), which has a key role in urogenital development, and the expression of markers associated with sex-cord differentiation, i.e., Müllerian-inhibiting substance (MIS) and inhibin (I). We also studied p53 expression. Archival, paraffin-embedded tissue from two patients with gonadoblastoma, one bilateral and both with concurrent germinomatous areas, were evaluated immunohistochemically with antibodies directed against I, MIS, WT1, and p53. I was noted in the sex-cord component of the gonadoblastoma and not in germinoma cells. MIS was noted in both, although it was more strongly expressed in the sex-cord component. p53 expression was noted in the germ cells of the gonadoblastoma, as well as in the frankly germinomatous areas, whereas WT1 was found only in the sex-cord region and not at all in the germ cells of either of our two cases. These findings lead us to propose that WT1 and I are present in the initiation of gonadoblastomas, but are lost with the progression of these tumors. Similarly, MIS may be involved in its tumorigenesis. The expression of p53 seems to support the concept that gonadoblastoma represents in situ germ cell neoplasia with malignant potential. Additional studies, however, are needed to validate this concept.

Heterogeneity of gonadoblastoma germ cells: similarities with immature germ cells, spermatogonia and testicular carcinoma in situ cells.

Gonadoblastoma is defined as a neoplasm containing nests of germ cells and cells resembling Sertoli cells or granulosa cells. Gonadoblastomas arise almost exclusively in dysgenetic gonads. They are associated with an increased risk of developing germ cell tumours. Testicular germ cell tumours in adults are preceded by carcinoma in situ cells, which are characterized by their morphology, by their immunohistochemical expression of placental-like alkaline phosphatase, the proto oncogene c-kit and/or epitopes for the monoclonal antibodies M2A, 43-9F and TRA-1-60, and by their aneuploid DNA content. In order to elucidate if gonadoblastomas are in situ neoplasms from the beginning, showing similarities with carcinoma in situ cells in otherwise normal testes, we investigated the germ cells in gonadoblastomas for their expression of the immunohistochemical markers of carcinoma in situ cells from six patients aged 8 1/2 months to 20 years and 4 months. In addition, the DNA content of the germ cells from five of the six patients was also determined by densitometric measurement on Feulgen stained specimens. The germ cell populations were heterogeneous both within the same patient and between the patients. Expression of the testicular carcinoma in situ markers was detected in specimens from all the patients and germ cells with an aneuploid DNA distribution pattern in accordance with testicular carcinoma in situ cells were detected. However, apparently normal immature germ cells were also present in four of the patients of whom two also had germ cells with a morphology similar to normal spermatogonia. Thus, gonadoblastoma is most likely an in situ germ cell neoplasia from the beginning. It seems probable that the germ cell tumours associated with gonadoblastomas originate from the carcinoma in situ cells inside the gonadoblastoma. Our findings of carcinoma in situ cells in gonadoblastomas from children support the theory that the cells arose prenatally.

Histogenetic consideration of ovarian sex cord-stromal tumors analyzed by expression pattern of cytokeratins, vimentin, and laminin. Correlation studies with human gonads.

A total of 30 sex cord-stromal tumors including 9 adult type and 5 juvenile type granulosa cell tumors (GCTs), 4 Sertoli-Leydig cell tumors (SLTs), 1 gynandroblastoma, 5 thecomas, 2 fibromas and 3 sclerosing stromal tumors were immunohistochemically evaluated by means of cytokeratins of different molecular weight, vimentin and laminin with regard to the histogenesis of these tumors and to the embryogenesis of the sex cord and stroma of developing gonads. For comparison, 7 embryonic gonads, 9 fetal and 9 adult ovaries, 14 fetal and 5 postnatal testes, and 1 gonadoblastoma were also examined. The coelomic epithelium of all gonads were positive for both cytokeratins (CAM 5.2 and AE1) and vimentin. In fetal ovaries, the granulosa cells of primordial follicles express low molecular weight cytokeratins only and those cells of more maturing follicles did not express any cytokeratin or vimentin. In adult ovaries, the granulosa cells of primordial follicles coexpressed low molecular weight cytokeratins and vimentin, but those cells of more maturing follicles expressed vimentin only. In fetal testes before 20 weeks gestational age, the Sertoli and Leydig cells did not express any cytokeratins and vimentin. After that time, both cells expressed vimentin only throughout life. The rete ovarii and rete testis from fetal to adult life coexpressed both low molecular weight cytokeratins and vimentin. The rete ovarii in all ages and rete testis in prenatal and childhood ages were surrounded by the laminin-positive basement membrane, however, the rete testis in adult were not. In neoplasia, the GCTs, thecomas, fibromas, and sclerosing stromal tumors expressed vimentin only.(ABSTRACT TRUNCATED AT 250 WORDS)