Clinical Cancer Research Vol. 11, 5722-5729, August 15, 2005
© 2005 American Association for Cancer Research
Identification of Two Molecular Groups of Seminomas by Using Expression and Tissue Microarrays
Matthias D. Hofer1,2,
Tara J. Browne1,2,
Le He1,
Rolf I. Skotheim3,
Ragnhild A. Lothe3 and
Mark A. Rubin1,2
Authors' Affiliations: 1 Department of Pathology, Brigham and Women's Hospital; 2 Harvard Medical School, Boston, Massachusetts and 3 Department of Genetics, Institute for Cancer Research, The Norwegian Radium Hospital, Oslo, Norway
Requests for reprints: Mark A. Rubin, Brigham and Women's Hospital/Harvard Medical School, 221 Longwood Ave., EBRC 422A, Boston, MA 02115. Phone: 617-525-6747; Fax: 617-264-5169; E-mail: marubin{at}partners.org.
 |
Abstract
|
|---|
Highly effective tailored clinical management of testicular germ cell tumors is based on the identification of two major histologic subtypes: seminomatous and nonseminomatous germ cell tumors. Expression array analysis of these two histologic subtypes using hierarchical clustering reveals two tumor groups, one composed solely of seminomas and the other containing embryonal carcinomas and seminomas. Supervised analysis between these groups identified 55 significantly dysregulated genes (false discovery rate = 2.3). The genes with the highest overexpression in the first group compared with the second included SLC43A1 (POV1), NET-7, IGF2, and JUP; down-regulated genes included GRB7, PFKP, and CDC6. In situ hybridization of SLC43A1 mRNA showed significantly increased signal intensity in the seminomas. At the protein level, expression of the immunohistochemical markers cytokeratins (pan-cytokeratin staining), placental-like alkaline phosphatase, anti-cytokeratin clone 5.2, CD30, anion exchanger 1/3, junction plakoglobulin (JUP), and POU domain, class 5, transcription factor 1 (octomer-binding transcription factor 3/4) was significantly different between seminomas and embryonal tumors. Hierarchical clustering based on a refined protein expression profile identified two groups, the first consisting solely of seminomas the other of seminomas and embryonal carcinomas. No histomorphologic differences were observed between the two seminoma groups such as the presence or absence of lymphocytes or extent of stromal elements. In summary, using independent methodologies and samples, we have identified two groups of seminomas. One group of seminomas has a molecular profile similar to embryonal carcinoma. The findings in the current study may help explain aberrant immunoprofiles seen with some seminomas.
Testicular germ cell tumors (TGCT) are the most common solid tumor in young men aged between 20 and 35 years and over the last several decades their incidence has increased (1, 2). Seminomas account for 
50% of all TGCT. Both seminomas and nonseminomatous tumors are believed to arise from carcinoma in situ of the testis (3) and currently there are two models that describe their further development. One model suggests that seminomas and nonseminomatous tumors are derived independently from carcinoma in situ, whereas the other model suggests a continuum with seminomas progressing into nonseminomatous tumors (3, 4). The presence of an isochromosome 12p or gain of 12p sequences are most likely found in all germ cell tumors (5). In addition, previous genomic studies have showed many of the same chromosomal aberrations among seminomas and nonseminomatous tumor, which are typically in the hypertriploid and hypotriploid range, respectively (6). Mutations or promoter methylation have been shown for some genes to be preferentially found in either seminomas or nonseminomatous carcinomas (6).
In most cases, the distinction between seminoma and embryonal carcinoma for example can be rendered by gross examination of the tumor and light microscopy by the experienced pathologist. Illustration of TGCT including both seminoma and embryonal carcinomas is presented in Fig. 1. The histomorphology plays a critical role in determining treatment. Through a series of well-controlled clinical trials in the 1970s and 1980s, it was recognized that orchiectomy and adjunct therapy was the most effective way to cure men of TGCTs. Seminomas are highly sensitive to external beam radiation, which has become the therapy of choice in both clinical stage I and metastatic seminomas. Platinum-based chemotherapy regimes however may be used as an alternative, especially in higher-stage tumors (7). Because only a small percentage (10-20%) of clinical stage I seminomas harbor occult retroperitoneal lymph node metastases (8, 9), surveillance-only protocols are now being used as an alternative treatment strategy for these early tumors. Likewise, the primary treatment for stage I nonseminomatous tumor is controversial and includes retroperitoneal lymph node dissection and chemotherapy, whereas metastatic nonseminomatous tumor is primarily treated with chemotherapeutic agents (7). The relapse rate for stage I seminomas treated with radiotherapy is 1% to 2% after 18 months and in stage II seminoma 5% to 15% after 2 years. The reported relapse rates for patients managed by surveillance only ranges from 15% to 19% (10–14). The relapse rate for nonseminomatous tumor is 10% to 12% for stage I disease treated with retroperitoneal lymph node dissection and 3% after 2 years for those treated with chemotherapy. The prognosis of stage II nonseminomatous tumor is better than that of stage II seminomas with a recurrence rate of only 6% for those treated by retroperitoneal lymph node dissection and chemotherapy.
View larger version (91K):
[in this window]
[in a new window]
|
Fig. 1. A, gross picture of a multilobulated testicular tumor showing a homogenous cream-colored, fleshy cut surface consistent with a seminoma. B-C, medium and high-power magnification of a seminoma exhibiting an interstitial and diffuse-type growth pattern of tumor cells. The cells have a central nucleus with vesicular chromatin, prominent nucleolus, and moderate amount of clear to eosinophilic cytoplasm with distinct cytoplasmic borders. Admixed are numerous lymphocytes. D, gross picture of a testicular tumor exhibiting a hemorrhagic, necrotic, and heterogeneous cut surface. E-F, medium and high-power magnification of an embryonal carcinoma exhibiting a nodular growth pattern intermingled with areas of necrosis. The cells are crowded and polygonal in shape, with large irregular nuclei, vesicular and clumped chromatin, single or multiple nucleoli, and moderate amounts of basophilic cytoplasm with ill-defined borders.
|
|
Despite the effectiveness of radiotherapy in stage I seminomas, clinicians are now rethinking the standard treatment strategies for early-stage disease especially because the risk of second malignancies in post-radiation patients is a real concern, particularly given the early age of diagnosis in many of these patients (15–17). Travis et al. conducted a large population based study that included 28,000 testicular cancers and found that the actual risk for developing a second malignancy increased over time from diagnosis and was 18.2% at 25 years (17). This raises the question whether current treatment strategies despite being extremely effective may be too aggressive. However, given the increased recurrence rate of seminomas in stage II disease compared with nonseminomatous tumors, radiation may not be the optimal treatment for all seminomas.
The current study examines cDNA expression array and protein profiling to help gain insight into these clinical observations.
|
Materials and Methods
|
|---|
Microarray analysis. Recently, members of this current study (R.I.S and R.A.L) published a cDNA microarray study on testicular germ cell tumors (eight pure seminomas and four embryonal carcinomas), which had also been studied previously by comparative genomic hybridization (4). Hierarchical clustering was done in the initial analysis (unsupervised analysis). In this current study, we did a supervised analysis to determine genes that were significantly differentially expressed between seminomas and embryonal carcinomas using significance analysis of microarrays (SAM). The primary analysis of the cDNA expression data was done with the Genepix software (Axon Instruments, Union City, CA). Cluster analysis (Stanford University, Palo Alto, CA) and generation of figures with TreeView was done using software developed by Eisen et al. (18). The SAM software program (Stanford University) was used to assign a score to each gene based on changes in gene expression relative to the SD of repeated measurements (19). For genes with scores greater than an adjustable threshold, SAM uses permutations of the repeated measurements to estimate the percentage of genes identified by chance referred to as the false discovery rate. Cluster, Treeview, and SAM software can be obtained at http://www.dnachip.org (Stanford University). Additional statistical analysis was done using SPSS (Chicago, IL).
Tissue microarray construction and analysis. Forty-five germ cell tumors including 22 pure seminomas and 23 embryonal carcinomas (13 pure embryonal carcinomas and 10 predominantly embryonal carcinomas) were identified from the files of the Brigham and Women's Hospital in accordance with the Institutional Review Board guidelines. All H&E-stained slides on the cases were reviewed to confirm the original diagnosis. The demographics were recorded and archival formalin-fixed paraffin-embedded tissue from orchiectomy specimens was used to construct a tissue microarray composed of replicate 0.6-mm cores extracted from the best representative areas of each of the 45 TGCT. Care was taken to avoid areas of necrosis. An H&E-stained microarray section was reviewed by pathologists to confirm that the sample was representative of the original tumor.
Immunohistochemistry. A panel of standard and novel germ cell tumor biomarkers was evaluated by immunohistochemistry using an automated image analysis system (ACISII, Chromavision, San Juan Capistrano, CA). Antigen retrieval was done on the following antibodies by protease digestion for 10 minutes in a water bath at 37°C at the following dilutions: cytokeratin clone 5.2 (1:50), anion exchanger 1/3 (1:200), and pan-cytokeratin (1:700). Antigen retrieval for Basal Cell Cocktail [p63 (1:200) plus 34ßE12 (1:250), at a ratio 1:1] and CD30 was done by placing the sections in a citrate buffer solution (pH 8.0) for 30 minutes, whereas junction plakoglobulin (1:50) were placed in an EDTA solution before microwaving. Antigen retrieval was not required for epithelial membrane antigen (1:200), c-kit (1:250), or placental-like alkaline phosphatase (1:250). For POU domain, class 5, transcription factor 1, antigen retrieval was done with a 25-minute incubation in citrate steam. Antibodies were obtained from following companies: DAKO Cytomation, Carpinteria, CA (anion exchanger 1/3, 34ßE12, CD30, placental-like alkaline phosphatase, epithelial membrane antigen, c-kit); Becton Dickinson, Palo Alto, CA (cytokeratin clone 5.2, junction plakoglobulin); Neomarkers, Fremont, CA (p63); and Santa Cruz Biotechnology, Santa Cruz, CA (POU domain, class 5, transcription factor 1). Sections were then processed with Envision+ horseradish peroxidase method (DAKO Cytomation). Cluster analysis (Stanford University) and generation of figures with TreeView was done in a similar manner as that used for the cDNA expression array data. An expression score was used taking into account both scoring intensity and percentage of positive cells.
In situ hybridization. The probe for SLC43A1 [solute carrier family 43, member 1, formerly prostate cancer overexpressed gene 1 (POV1)] was obtained by amplifying a 650-bp fragment with SLC43A1 sequence specific primers from a Burkitt lymphoma cDNA library (Becton Dickinson) using PCR. DIG-labeled riboprobes were prepared using an in vitro transcription kit (Roche Molecular Biochemicals, Mannheim, Germany). The efficiency of transcription and incorporation of digoxigenin-UTP to the riboprobes were evaluated by 5% acrylamide-urea gel electrophoresis and dot blot on nylon membrane, respectively (DIG Nucleic Acid Detection Kit-nitroblue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate, Roche Molecular Biochemicals). Automated ISH was done on the Discovery System (Ventana Medical Systems, Tucson, AZ). After deparaffinization, slides were fixed in 4% paraformaldehyde for 10 minutes and digested with proteinase K (10 µg/mL) for 10 minutes at 37°C. DIG-labeled riboprobes were diluted at 1:100 in hybridization solution in hybridization solution (RyboHybe, Ventana Medical Systems). The hybridization was done for 6 hours at 65°C. After hybridization, slides were washed in 0.1x SSC twice at 75°C for 6 minutes each time. The detection was done using biotinilated anti-digoxigenin antibody (Biogenex, San Ramon, CA) followed by streptavidin-alkaline-phosphatase conjugate and visualized by nitroblue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate substrate reaction (Ventana BlueMap Detection Kit, Ventana Medical Systems). Slides were counterstained by nuclear fast red (Vector Laboratories, Inc., Burlingame, CA), dehydrated, and coverslipped for microscopic analysis. Evaluation for the in situ hybridization was done by the study pathologist (T.J.B.) using a four-tier scoring system with intensity ranging from 1 (absent) to 4 (strong).
|
Results
|
|---|
Supervised analysis of expression array data reveals distinct groups of seminomas. Members of our group (R.I.S. and R.A.L.) have previously published cDNA expression array analysis of 14 TGCTs that included eight pure seminomas and six nonseminomatous tumors (four embryonal carcinomas and two immature teratomas) as well as one carcinoma in situ and three benign samples (4). The cDNA microarray contained 636 genes and expressed sequence tags from chromosome 17 as well as 512 additional genes located throughout the genome. Chromosome 17 was targeted due to the fact that over half of germ cell tumors analyzed showed amplification in two distinct regions on chromosome 17. Unsupervised analysis using hierarchical cluster analysis grouped normal tissue, carcinoma in situ, seminomas, immature teratomas, and embryonal carcinomas in separate groups. We reanalyzed this data set using a supervised approach with SAM and identified 55 genes that were significantly up-regulated and down-regulated between seminomas and embryonal carcinomas with a false discovery rate of 2.3. Genes with a 
2-fold up-regulation or down-regulation are presented in Table 1. Up-regulated genes included the SLC43A1 gene (formerly POV1) with a >9-fold up-regulation and the IGF2. Down-regulation was subtle and genes included GRB7 with a 0.2-fold expression as well as PFKP (0.2-fold expressed). The major goal of this analytic exercise was to identify genes that could be used to distinguish embryonal carcinoma from seminoma. However, when we clustered these cases using the selected genes from the supervised analysis using TreeView, one unexpected result was that four of eight seminomas clustered with the embryonal carcinomas, whereas the remaining four seminomas clustered with each other (Fig. 2A).
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 2. Identification of two classes of seminomas. A, using only the significantly dysregulated genes determined from SAM analysis, we did hierarchical clustering. All genes represented on the microarray are shown to illustrate the differences in gene expression between the clusters. Two groups of testicular cancer were identified, one consisting solely of seminomas, whereas the other one contains seminomatous and embryonal carcinomas. The dendrogram from the clustered cDNA expression array data demonstrates a group of pure seminomas (group 1, Sem 5-8) and a mixed group of embryonal carcinoma (EC 1-4) and a subgroup of seminomas (group 2, Sem 1-4). Group 2 seminomas are more closely related to the embryonal carcinomas than the group 1 seminomas. The zoomed section represents some of the genes dysregulated between seminomas and embryonal carcinomas regardless of exact level of up-regulation or down-regulation. For a complete list, please refer to Table 1. B, in situ hybridization of SLC43A1 mRNA in embryonal carcinomas and seminomas. SLC43A1 is strongly expressed in seminomas and nondetectable in embryonal carcinomas.
|
|
In situ hybridization supports cDNA microarray results. To validate the results of our expression array analysis, we analyzed the expression level of SLC43A1 (POV1), the most significantly dysregulated gene between seminomas and nonseminomatous tumors, using in situ hybridization instead of immunohistochemistry as no antibody is currently available. Hybridization was done on a tissue microarray containing seminomatous and embryonal tumors. The staining intensity was scored by the study pathologist (T.J.B.) in a four-tier system and showed absent staining in embryonal carcinomas and strong expression in seminomas as the representative tissue cores of embryonal carcinoma and seminoma depicted in Fig. 2B show. SLC43A1 mRNA was significantly up-regulated in seminomas compared with embryonal carcinomas (P = 0.002, 
2 test), supporting the cDNA microarray results.
Demographics of testicular germ cell tumor cases represented on the tissue microarray. To analyze expression of a panel of known and novel biomarkers of TGCT, we constructed a high-density tissue microarray containing 22 pure seminomas and 23 embryonal carcinomas. The mean age of patients suffering from pure seminomas was 36.64 years (range, 23-56 years) and was significantly higher than that of patients with embryonal carcinomas, who had a mean age of 28.92 years (range, 19-41 years; P = 0.004). The mean age of patients with seminomatous-mixed tumors was the lowest with 27.7 years (range, 18-47). The mean size of the tumors was not significantly different among those three groups ranging between 1.2 and 6 cm (mean, 3.359 cm) for seminomas, 1 to 4.5 cm (mean, 2.4 cm) for embryonal cancers, and 2 to 6.2 cm (mean, 3.43 cm) for seminomatous-mixed tumors.
Immunohistochemical biomarkers separate seminomas from nonseminomatous tumors. Seminoma and embryonal carcinomas are rare tumors compared with those commonly evaluated by most practicing pathologists, and a distinct panel of biomarkers that distinguish reliably the two major groups is clinically needed.
Therefore, to identify molecular markers that reliably distinguish seminomas from nonseminomatous tumors, we used a panel of antibodies directed against several epithelial markers and previously described markers of TGCT. The immunohistochemical panel results based on tumor type are presented in Table 2 and Fig. 3. Intensities of immunohistochemical staining were determined using a semiautomated microscope system. The intensities of the immunostaining of cytokeratins (as determined by pan-cytokeratin staining), placental-like alkaline phosphatase, the anti-cytokeratin clone 5.2 antibody, antibodies directed against the cell epitope CD30, the human anion exchanger 1/3, junction plakoglobulin (
-catenin), and POU domain, class 5, transcription factor 1 (formerly octomer-binding transcription factor 3/4) were significantly different between seminomas and embryonal tumors. Different staining of borderline significance was observed for c-kit, a marker clinically often used to separate between seminomas and embyronal carcinomas, whereas no significant difference was found for staining of the epithelial membrane antigen, and a basal cell cocktail (consisting of anti-p63 plus anti-34ßE12 at a 1:1 ratio).
View larger version (128K):
[in this window]
[in a new window]
|
Fig. 3. Immunohistochemical detection of CD30, JUP, PLAP, and AE1/3 expression in embryonal carcinomas, seminomatous-mixed tumors and seminomas in tissue cores of the TMA. For comparison purposes, also the H&E-stain of representative tissue cores are shown. Expression levels of these markers significantly differ between embryonal carcinomas and seminomas, seminomatous-mixed tumors show a staining profile that overlapped with embryonal carcinoma. Greatest differences in staining intensities between the tumor groups are shown which are not representative of the quantitative means. For exact differences in staining intensities among these three groups, please refer to Table 2.
|
|
Multiplex biomarker analysis reveal two groups of seminomas. The protein expression levels of all the immunohistochemical biomarkers were determined using a semi automated image analysis system that provides a continuous intensity score. Therefore, similar to the approach used with cDNA expression array data, we did hierarchical clustering (Fig. 4). Two of 22 seminomas clustered with embryonal cancers suggesting two subgroups of seminomas, one of which on a molecular level is closely related to embryonal carcinoma. There was no difference in size, age, stage, or clinical presentation preoperatively and postoperatively between these two cases and other seminomas.
View larger version (38K):
[in this window]
[in a new window]
|
Fig. 4. Analyzing expression of standard and novel immunohistochemistry markers of germ cell tumors with an automated microscope system and consecutive clustering of the expression levels, we again identified two groups of germ cell tumors, one consisting solely of seminomas and the other one of seminomas and embryonal tumors.
|
|
|
Discussion
|
|---|
TGCTs have been divided into two main groups, seminomas and nonseminomatous tumors, based primarily on the observed different therapeutic responses based on these histomorphologic differences. The histologic distinction is perhaps one of the best examples where the pathologic classification dictates response to treatment. In the current study, we used independent methodologies and samples to do a molecular profile to complement the histologic classification of TGCTs. Two groups of seminomas emerged from this analysis that were otherwise similar by microscopic examination. One group consisted of seminomas alone, whereas the other group consisted of seminomas, which exhibited molecular and immunophenotypic characteristics similar to embryonal carcinomas. This is the first report describing a subgroup of seminomas, which seem to show some molecular and immunophenotypic characteristics similar to nonseminomatous tumors. This observation might explain why most seminomas can be reliably distinguished using light microscopy and immunohistochemistry, but a small subset represents a clinical challenge for the pathologist. The distinction between seminomas and nonseminomatous tumor is obviously critical given the different therapies.
In the current study, we used SAM to determine significantly dysregulated genes between seminomas and NSGT. To validate this result, we used in situ hybridization for the most significantly dysregulated gene that best distinguished these two groups. In a second step, we used the panel of genes and did a hierarchical cluster analysis as described by Eisen et al. (18) to identify whether detection of expression of these genes by immunohistochemistry could provide information as to type of TGCT (i.e., seminoma or nonseminomatous tumors). A panel of immunohistochemical markers previously reported to distinguish seminomas from nonseminomatous tumors (20–24) was used to classify theses tumors. Protein expression was measured on a continuous scale (range, 0-255) with an automated image analysis system, minimizing subjective categorical analysis (e.g., negative, weak, moderate, and strong) and intraobserver variation. In this study, 7 of 10 markers showed significant differences between two classes. Clustering the results obtained from the immunochemistry profile, we identified a subgroup of seminomas that showed closer relationship to nonseminomatous tumors than to seminomas supporting the cDNA expression array results. Of note, at the protein level the fraction of seminomas clustering with nonseminomatous tumors is smaller than at the expression level. Because the treatment of TGCTs is so effective, the true natural history of seminomas is uncertain and it is well likely, that fewer seminomas are related to nonseminomatous tumors than suggested by the expression analysis. To clarify this, a validation using larger clinical trial cohorts would be needed.
Based on light microscopy and histologic evaluation, most TGCTs can be categorized into seminomas and nonseminomatous tumors. However, in the current study, using two separate approaches, we were able to identify the molecular fingerprints of a subset of seminomas, which was more closely related to nonseminomatous tumors than to seminomas. Given the increasing interest in using surveillance for stage I seminomas, these data would suggest that there may be a role for the development of a molecular profiles to identify higher-risk patients. This is of particular concern in patients with TGCTs where the vast majority of patients are younger than 35 years and may live long enough to experience the development of secondary malignancies (15–17, 25, 26). Validation of the panel of biomarkers presented in this study will require identifying patients with seminomas that failed on clinical trials using standard seminoma treatment protocols and we are currently in the process of obtaining such samples. The goal will be to improve our ability to determine treatment response. However, we also recognize that due to the overwhelming success of current treatment it may be difficult to improve on the standard histomorphologic approach. The current protocols were established through a series of clinical trials and any biomarker panel will need to provide a high level of confidence supporting alterations in treatment protocols. Perhaps a limited panel could be included in future trial protocols to evaluate this concept on a prospective basis.
|
Footnotes
|
|---|
Grant support: U.S. Department of Defense grant PC030214 (M.D. Hofer and M.A. Rubin) and National Cancer Institute grant CA97063 (M.A. Rubin).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: M.D. Hofer and T.J. Browne contributed equally to this work.
Received 3/ 9/05;
revised 5/17/05;
accepted 5/24/05.
|
References
|
|---|
- Bergstrom R, Adami HO, Mohner M, et al. Increase in testicular cancer incidence in six European countries: a birth cohort phenomenon. J Natl Cancer Inst 1996;88:727–33.[Abstract/Free Full Text]
- Devesa SS, Blot WJ, Stone BJ, Miller BA, Tarone RE, Fraumeni JF, Jr. Recent cancer trends in the United States. J Natl Cancer Inst 1995;87:175–82.[Abstract/Free Full Text]
- Dieckmann KP, Skakkebaek NE. Carcinoma in situ of the testis: review of biological and clinical features. Int J Cancer 1999;83:815–22.[CrossRef][Medline]
- Skotheim RI, Monni O, Mousses S, et al. New insights into testicular germ cell tumorigenesis from gene expression profiling. Cancer Res 2002;62:2359–64.[Abstract/Free Full Text]
- Atkin NB, Baker MC. i(12p): specific chromosomal marker in seminoma and malignant teratoma of the testis? Cancer Genet Cytogenet 1983;10:199–204.[CrossRef][Medline]
- Frigyesi A, Gisselsson D, Hansen GB, Soller M, Mitelman F, Hoglund M. A model for karyotypic evolution in testicular germ cell tumors. Genes Chromosomes Cancer 2004;40:172–8.[CrossRef][Medline]
- Laguna MP, Pizzocaro G, Klepp O, Algaba F, Kisbenedek L, Leiva O. EAU guidelines on testicular cancer. Eur Urol 2001;40:102–10.[CrossRef][Medline]
- Peckham MJ, Hamilton CR, Horwich A, Hendry WF. Surveillance after orchiectomy for stage I seminoma of the testis. Br J Urol 1987;59:343–7.[Medline]
- Maier JG, Sulak MH, Mittemeyer BT. Seminoma of the testis: analysis of treatment success and failure. Am J Roentgenol Radium Ther Nucl Med 1968;102:596–602.[Medline]
- Warde P, Specht L, Horwich A, et al. Prognostic factors for relapse in stage I seminoma managed by surveillance: a pooled analysis. J Clin Oncol 2002;20:4448–52.[Abstract/Free Full Text]
- Warde P, Gospodarowicz MK, Banerjee D, et al. Prognostic factors for relapse in stage I testicular seminoma treated with surveillance. J Urol 1997;157:1705–9. Discussion 1709–10.[CrossRef][Medline]
- Warde PR, Gospodarowicz MK, Goodman PJ, et al. Results of a policy of surveillance in stage I testicular seminoma. Int J Radiat Oncol Biol Phys 1993;27:11–5.[Medline]
- von der Maase H, Specht L, Jacobsen GK, et al. Surveillance following orchidectomy for stage I seminoma of the testis. Eur J Cancer 1993;29A:1931–4.
- Horwich A, Alsanjari N, A'Hern R, Nicholls J, Dearnaley DP, Fisher C. Surveillance following orchidectomy for stage I testicular seminoma. Br J Cancer 1992;65:775–8.[Medline]
- Travis LB. Therapy-associated solid tumors. Acta Oncol 2002;41:323–33.[CrossRef][Medline]
- Travis LB, Andersson M, Gospodarowicz M, et al. Treatment-associated leukemia following testicular cancer. J Natl Cancer Inst 2000;92:1165–71.[Abstract/Free Full Text]
- Travis LB, Curtis RE, Storm H, et al. Risk of second malignant neoplasms among long-term survivors of testicular cancer. J Natl Cancer Inst 1997;89:1429–39.[Abstract/Free Full Text]
- Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 1998;95:14863–8.[Abstract/Free Full Text]
- Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 2001;98:5116–21.[Abstract/Free Full Text]
- Cheville JC, Rao S, Iczkowski KA, Lohse CM, Pankratz VS. Cytokeratin expression in seminoma of the human testis. Am J Clin Pathol 2000;113:583–8.[CrossRef][Medline]
- Kemmer K, Corless CL, Fletcher JA, et al. KIT mutations are common in testicular seminomas. Am J Pathol 2004;164:305–13.[Abstract/Free Full Text]
- Madani A, Kemmer K, Sweeney C, et al. Expression of KIT and epidermal growth factor receptor in chemotherapy refractory non-seminomatous germ-cell tumors. Ann Oncol 2003;14:873–80.[Abstract/Free Full Text]
- Okada K, Katagiri T, Tsunoda T, et al. Analysis of gene-expression profiles in testicular seminomas using a genome-wide cDNA microarray. Int J Oncol 2003;23:1615–35.[Medline]
- Skotheim RI, Abeler VM, Nesland JM, et al. Candidate genes for testicular cancer evaluated by in situ protein expression analyses on tissue microarrays. Neoplasia 2003;5:397–404.[Medline]
- Swerdlow AJ, Barber JA, Hudson GV, et al. Risk of second malignancy after Hodgkin's disease in a collaborative British cohort: the relation to age at treatment. J Clin Oncol 2000;18:498–509.[Abstract/Free Full Text]
- Bokemeyer C, Schmoll HJ. Treatment of testicular cancer and the development of secondary malignancies. J Clin Oncol 1995;13:283–92.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
L. H. J. LOOIJENGA, AD. J. M. GILLIS, H. J. STOOP, R. HERSMUS, and J. W. OOSTERHUIS
Chromosomes and Expression in Human Testicular Germ-Cell Tumors: Insight into Their Cell of Origin and Pathogenesis
Ann. N.Y. Acad. Sci.,
December 1, 2007;
1120(1):
187 - 214.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Materials and Methods
