Recent developments in testicular germ cell tumor research

Recent Developments in TesticularGerm Cell Tumor Research

Gert-Jan M. van de Geijn, Remko Hersmus, and Leendert H. J. Looijenga*

Testicular germ cell tumors of adolescents and adults (TGCTs; the so-called type II variant) are the most frequent malignancies found in Cau-casian males between 20 and 40 years of age. The incidence hasincreased over the last decades. TGCTs are divided into seminomas andnonseminomas, the latter consisting of the subgroups embryonal carci-noma, yolk-sac tumor, teratoma, and choriocarcinoma. The pathogene-sis starts in utero, involving primordial germ cells/gonocytes that areblocked in their differentiation, and develops via the precursor lesioncarcinoma in situ toward invasiveness. TGCTs are totipotent and can beconsidered as stem cell tumors. The developmental capacity of their cellof origin, the primordial germ cells/gonocyte, is demonstrated by thedifferent tumor histologies of the invasive TGCTs. Seminoma representsthe germ cell lineage, and embryonal carcinoma is the undifferentiatedcomponent, being the stem cell population of the nonseminomas. So-matic differentiation is seen in the teratomas (all lineages), whereasyolk-sac tumors and choriocarcinoma represent extra-embryonal differ-entiation. Seminomas are highly sensitive to irradiation and (DNA dam-aging) chemotherapy, whereas most nonseminomatous elements areless susceptible to radiation, although still sensitive to chemotherapy,with the exception of teratoma. To allow early diagnosis and follow up,appropriate markers are mandatory to discriminate between the differ-ent subgroups. In this review, a summary will be given related to sev-eral recent developments in TGCT research, especially selected becauseof their putative clinical impact. Birth Defects Research (Part C)87:96–113, 2009. VC 2009 Wiley-Liss, Inc.

Key words: review; type II (testicular) germ cell tumors; carcinomain situ; OCT3/4; diagnostic markers; primordial germ cells; pathogene-sis; treatment response/resistance; SCF; c-KIT; microRNA

NORMAL GERM CELL

DEVELOPMENT

Germ cells in mammals, whichensure transmission of genetic in-formation to the next generationby production of mature oocytes infemales and spermatozoa in males,are set aside at an early stage dur-ing embryogenesis (Witschi, 1948;Mc et al., 1953; Falin, 1969; Gins-

burg et al., 1990; Donovan, 1998).To fulfil this special task, thesecells, known as primordial germcells (PGCs), have unique and spe-cific characteristics. These includetheir initial capacity to suppressinduction of differentiation, theircommitment to either the male orfemale lineage, and finally thecapacity to generate highly special-

ized daughter cells that, after fer-tilization, regain an activated dif-ferentiation program. This meansthat they can form all embryonicand extra-embryonic tissues,including the germ cell population(Honecker et al., 2006). In otherwords, they are in fact really thetotipotent stem cell population ofthe body, and represent the circleof life.As indicated, PGCs are the stem

cells of gametogenesis, beingoogenesis in female and spermato-genesis in male. They arise earlyduring embryogenesis at the baseof the allantois (in humans in week5–6; Witschi, 1948; Mc et al.,1953; Falin, 1969); in mice at dayE7.5 (Ginsburg et al., 1990; Dono-van, 1998). From here, theymigrate through the embryo viathe hindgut, where they subse-quently exit dorsally and move lat-erally toward both genital ridges,where the gonads will develop(Molyneaux et al., 2001; Moly-neaux et al., 2003; Molyneaux andWylie, 2004). Once the PGCs havereached the gonadal ridges, theyare called gonocytes. During thesestages, the embryonic germ cellsare characterized by severalmarkers, including alkaline phos-phatase, c-KIT, OCT3/4, BLIMP1,VASA, and NANOG. These arerelated to various biological mech-anisms, such as migration (c-KIT),survival (c-KIT and OCT3/4), and

REVIE

W

VC 2009 Wiley-Liss, Inc.

Birth Defects Research (Part C) 87:96–113 (2009)

G. J. M. van de Geijn, R Hersmus, and L. H. J. Looijenga are from Department of Pathology, Erasmus MC, Erasmus UniversityMedical Center, Daniel den Hoed Cancer Center, Josephine Nefkens Institute, Rotterdam, The Netherlands

Grant sponsor: Dutch Cancer Society; Grant number: 2004-3123 (to GJMvdG); Grant sponsor: Erasmus MC (Translational ResearchGrant, 2006 to RH)

*Correspondence to: L. H. J. Looijenga, Department of Pathology, Erasmus University Medical Center, Josephine Nefkens Institute,Room Be 430b, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands. E-mail: [email protected]

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/bdrc.20140

suppression of differentiation(BLIMP1).Depending on the micro-environ-

ment, particularly related to thechromosomal constitution, that is,XX in females and XY in males, thegonocytes will differentiate into ei-ther oogonia or prespermatogonia.In the case of female development,the oogonia will divide multipletimes before they become oocytes,stop proliferating, and enter meio-sis, which starts at 11 to 12 weeksof gestation (Rabinovici and Jaffe,1990). In males, the gonocytes willgradually migrate toward the pe-riphery of the tubules. When theyare in contact with the laminabasalis, they are called presperma-togonia. During the first years afterbirth and until puberty, the pre-spermatogonia change to type Aspermatogonia (Muller and Skak-kebaek, 1983; Chemes, 2001;Berensztein et al., 2002). At theonset of puberty, the sper-matogonia undergo further matu-ration and go into meiosis to pro-duce spermatozoa.

MALIGNANT GERM CELL

DEVELOPMENT

An alternative classification systemfor human germ cell tumors(GCTs), compared to the tradi-tional histological subtypes as rec-ognized by the pathologist, hasrecently been developed and isrecognized by the WHO, (Table 1)(Woodward et al., 2004; Ooster-huis and Looijenga, 2005). The dis-tinction into five entities is basedon various parameters, includingcell of origin, histology, genomicimprinting status, age at and loca-tion of clinical presentation, andchromosomal constitution. Recog-nition of these subtypes allows abetter investigation on identifica-tion of the pathogenetic mecha-nisms involved, and will lead toimprovement of clinical diagnosisand treatment response prediction.Moreover, it will shed light on thevalue of existing in vivo and in vitromodel systems, and eventuallygenerate the optimal test model,both for pathobiological and clinicalstudies.

Type I GCTs are the teratomasand yolk-sac tumors of neonatesand infants that can present in thetestis, ovary, sacral or retroperito-neal region, or other sites in themidline of the body. The type IIGCTs are the seminomatous andnonseminomatous GCTs, the mainfocus of this review (see below).The type III GCTs are the sperma-tocytic seminomas that are typi-cally seen in males more than 50years of age (Woodward et al.,2004). The type IV GCTs are thedermoid cysts that arises fromoogonia or oocytes in the ovary,whereas the type V is the hydati-form mole that develops from anempty egg and a spermatozoa.Interestingly, the stage-specific

properties of germ cells arereflected in their neoplastic deriva-tives. This is for example nicelyillustrated for the (classic) semino-mas (type II TGCT) and spermato-cytic seminomas (type III TGCT)(for review see Oosterhuis andLooijenga, 2005).

TYPE II GCTS

Type II GCTs are interesting fromdifferent points of view, becausethey can be considered true stemcell tumors, capable of differentiat-ing into somatic, extra-embryonal,and even the germ cell lineageitself (Honecker et al., 2006). Inaddition, these tumors are uniquebecause of their overall respon-siveness to DNA damaging ther-apy, although depending on thehistological subtype (Masters andKoberle, 2003). In fact, they showa remarkably high survival rate,even in cases of metastases (seebelow).The seminomatous and nonsemi-

nomatous GCTs are the most fre-quent malignancies in Caucasianadolescents and adults aged 20 to40 years (Adami et al., 1994), ofwhich more than 95% develop inthe testis. The incidence of type IITGCTs, currently 6 to 11 per100,000, has doubled in the last 40years without a clear explanation(Giwercman et al., 1993; Swer-dlow et al., 1998; Richiardi et al.,2004; McGlynn et al., 2005; Shahet al., 2007), although various hy-

pothesis have been generated aswill be discussed later.Seminoma consists of trans-

formed germ cells that closelyresemble the PGCs/gonocytes.Apparently, these tumor cells areblocked in their normal differentia-tion pattern and cannot undergonormal spermatogenesis, but accu-mulate instead. Nonseminomas, incontrast, can be composed of dif-ferent histological elements, whichmirror the pluripotency of the PGC/gonocyte, normally only apparentafter fertilization. Embryonal carci-nomas represent the undifferenti-ated stem cell component, terato-mas represent the somatic differ-entiation, yolk-sac tumors, andchoriocarcinoma are the compo-nents showing extra-embryonaldifferentiation. In addition to theextra-embryonal and somatic dif-ferentiation, germ cell lineage dif-ferentiation in nonseminomas hasbeen reported (Fig. 1) (Honeckeret al., 2006). This demonstratesthat all lines, including the germlineage, can be formed, indicatingthat type II TGCTs are indeed theonly known totipotent solid cancer.Often multiple of these histologiesare present in a single tumor,called a mixed tumor. Even semi-noma- and nonseminoma compo-nents can be intermixed in about10% of the cases. In addition,seminomas can switch to a nonse-minomatous phenotype, a phe-nomenon called reprogramming oractivation of pluri(toti)potency(Looijenga et al., 1999; Oosterhuiset al., 2003).Several model systems are used

to study TGCTs. Type I TGCTs ofthe neonates and infants arereflected in the testicular tumors ofthe 129SV mouse strain (Stevensand Little, 1954; Stevens, 1970;Damjanov et al., 1971; Walt et al.,1993). For type II TGCTs, no ani-mal model has been reported, butrecently, the TCam-2 cell line wasconfirmed to have significant simi-larities to seminoma (Mizuno et al.,1993; Goddard et al., 2007; deJong et al., 2008a; Eckert et al.,2008b). Cell lines that model thenonseminomas have already beenknown for a much longer time(Fogh, 1978; Wang et al., 1980;

RECENT DEVELOPMENTS IN TGCT RESEARCH 97

Birth Defects Research (Part C) 87:96–113, (2009)

TABLE

1.Cla

ssification

ofth

eFiveDiffe

rentTypesofGerm

-Cell

Tum

ors

(Ooste

rhuis

and

Looijenga,2005).SeeTextfo

rFurtherDeta

ils

Type

Anatomicalsite

Phenotype

Age

Originatingce

llGenomic

imprinting

Genotype

Anim

almodel

ITestis/o

vary

/sa

cralre

gion/

retroperito-

neum/

mediastinum/

neck

/midlin

ebra

in/o

therra

resites

(Immatu

re)

tera

toma/y

olk-sac

tumor

Neonatesand

child

ren

EarlyPGC/

gonocy

tes

Bipare

ntal,

partially

era

sed

Diploid

(tera

toma).

Aneuploid

(yolk-sac

tumor):gain

of1q,

12(p

13)and20q,

andloss

of1p,4

and6q

Mouse

tera

toma

IITestis

Seminoma/

nonse

minoma

[15years

(medianage

35and25

years)

PGC/g

onocy

teEra

sed

Aneu

ploid

(1/2

triploid):

gain

of

X,7,8,12p,and

21;loss

ofY,1p,

11,13and18

–

Ovary

Dysg

erm

inoma/

nonse

minoma

[4years

PGC/g

onocy

teEra

sed

Aneuploid

–

Dysg

enetic

gonad

Dysg

erm

inoma/

nonse

minoma

Congenital

PGC/g

onocy

teEra

sed

Diploid/tetraploid

–

Anterior

mediastinum

(thymus)

Seminoma/

nonse

minoma

Adolesc

ents

PGC/g

onocy

teEra

sed

Diploid/tri-

tetraploid

–

Midlin

ebra

in(p

inealgland/

hypoth

alamus)

Germ

inoma/

nonse

minoma

Child

ren

(medianage

13years)

PGC/g

onocy

teEra

sed

Diploid/tri-

tetraploid

–

III

Testis

Sperm

atocy

tic

seminoma

Child

ren/a

dults

Sperm

atogonium/

sperm

atocy

tePartially

complete

patern

al

Aneuploid:gain

of9

Canine

seminoma

IVOvary

Derm

oid

cyst

[50years

Oogonia/o

ocy

tePartially

complete

matern

al

(Near)

diploid,

diploid/tetraploid,

peritriploid

(gain

of

X,7,12,and15)

Mouse

gynogenote

VPlace

nta/u

teru

sHydatiform

mole

Fertile

period

Empty

ovum/

sperm

atozo

aCompletely

patern

al

Diploid

(XX

andXY)

Mouse

andro

genote

98 VAN DE GEIJN ET AL.


Andrews et al., 1984; Teshimaet al., 1988). Type III TGCTs, thespermatocytic seminomas, arereported to resemble canine testic-ular tumors (Looijenga et al.,1994).

THE PRECURSOR LESION

OF TYPE II TGCTS

Both testicular seminomas andnonseminomas arise from a prein-vasive lesion called intratubulargerm cell neoplasia undifferenti-ated or carcinoma in situ (CIS)(Skakkebaek, 1972; Woodwardet al., 2004). It is striking to notethat the incidence of CIS andtype II TGCT is similar, indicatingthat all cases of CIS eventuallymay progress to invasiveness(Giwercman et al., 1991a). CIScells resemble PGCs/gonocytes inmany aspects: both have erasedgenomic imprinting (van Gurpet al., 1994), similar morphology(Rajpert-De Meyts et al., 2003)and express the same immunohis-tochemical markers, such asOCT3/4, PLAP, AP-2c, and c-KIT(Rajpert-De Meyts and Skakke-baek, 1994; Looijenga et al.,2003b; Hoei-Hansen et al., 2004;Honecker et al., 2004; Hoei-Han-sen et al., 2005; Pauls et al., 2005;

Stoop et al., 2005). CIS cells aretherefore assumed to be the malig-nant counterpart of PGCs/gono-cytes. This is in line with epidemio-logical observations, indicatingthat the cells of origin of type IITGCTs are present in a defined pe-riod during embryonal develop-ment. CIS cells are most likely theresult of a delayed or blocked dif-ferentiation of embryonic germcells. CIS typically presents as aseminiferous tubule lacking thecharacteristic multiple-layer con-tent of differentiating spermatogo-nia, spermatocytes, spermatids(Fig. 2A, B). Instead, basallylocated germ cells are seen mostly,under the tight junction of Sertolicells, in contact with each other(Skakkebaek, 1972) (Fig. 2C, D).The intermediate lesion betweenCIS and an invasive TGCT is intra-tubular seminoma or intratubularnonseminoma, being predomi-nantly embryonal carcinoma (Oos-terhuis et al., 2003; Berney et al.,2004; Lau et al., 2007). In thisprocess, CIS cells have become in-dependent from the micro-environ-ment (generated among others bythe Sertoli cells) and fill up thelumen of the seminiferous tubule(Fig. 2E, F; indicated by arrows).This stage is but one step away

from invasiveness, when the cellsspread out of the seminiferoustubule and form a seminoma (Fig.2E, F; bottom half) or embryonalcarcinoma (Fig. 2G, H).

NEWLY REPORTED

MARKERS FOR TGCTS

A significant number of markershas been reported over time thatcan be used to discriminate CIS,seminoma, and embryonal carci-noma. The most common areOCT3/4, c-KIT, PLAP, NANOG,SOX2, AP-2c, and UTF1, amongothers (see Table 2). Several ofthese will be discussed in moredetail here.OCT3/4 is a well-characterized

marker for PGCs. It is positive in allcases of CIS, seminoma, andembryonal carcinoma (Looijengaet al., 2003b) (Fig. 2D, F, and H),confirmed in many subsequentstudies (reviewed in de Jong et al.,2005). There has been a significantamount of reports over the yearsthat OCT3/4 is also expressed innormal adult (stem) cells and nongerm cell-derived cancers. How-ever, recent data indicate thatthese observations are likelyrelated to the use of nonspecificantibodies and primers, the latteralso recognizing pseudogenes(Ledford, 2007; Lengner et al.,2007; Liedtke et al., 2007; Atlasiet al., 2008). OCT3/4 is a tran-scription factor of the family ofoctamer-binding proteins (alsoknown as the POU homeodomainproteins) and is regarded as one ofthe key regulators of pluripotency,(for review see Niwa et al., 2000;de Jong and Looijenga, 2006). Inaddition to OCT3/4, several otherembryonic stem-cell-specific pro-teins are important for maintainingthe so-called ‘‘stemness’’ of pluri-potent cells, such as NANOG andSOX2 (Tanaka et al., 2002; Avilionet al., 2003; Adjaye et al., 2005;Boiani and Scholer, 2005; Boyeret al., 2005; Rajaraman et al.,2005; Yamaguchi et al., 2005;Yates and Chambers, 2005; Playeret al., 2006; Babaie et al., 2007).SOX2 is a member of the SOX pro-tein family, transcription factorsthat regulate development from

Figure 1. Overview of the different histologies from the type II TGCTs. From the pre-cursor lesion carcinoma in situ, either a seminoma or a nonseminoma can develop viathe intermediate stages of intrabulular seminoma (ITSE) or intratubular nonseminoma(ITNS). The different histologies of nonseminomas are displayed in the gray area.Seminomas can also develop into a nonseminoma via so-called ‘‘reprogramming,’’indicated by R. Cells with germ cell characteristics that have been reported within theyolk-sac component of tumors.



the early embryonal stage to dif-ferentiated lineages of specializedcells. SOX proteins are known tocooperate with POU proteins. Thebest characterized SOX-POU coop-eration is that between SOX2 andOCT3/4. SOX2 is not detected in

human germ cells regardless oftheir developmental age, in con-trast to data in mouse embryos (deJong et al., 2008b; Perrett et al.,2008). SOX2 is expressed inembryonal carcinoma, the undif-ferentiated part of nonseminomas,

but it is absent in seminomas,yolk-sac tumors, and normal sper-matogenesis (de Jong et al.,2008b; Perrett et al., 2008). CIScells are indeed negative for SOX2,although SOX2 positive Sertoli cellscan be present in seminiferoustubules lacking germ cells or in thepresence of CIS (de Jong et al.,2008b). Because CIS, seminoma,and embryonal carcinoma are allpositive for OCT3/4, the questionarises if any other SOX familymembers are present in CIS andseminoma to cooperate with OCT3/4. In this context, the different roleof OCT3/4 in PGCs and embryonalstem cells is of interest, being sup-pression of apoptosis and regula-tion of differentiation, respectively(Nichols et al., 1998; Kehler et al.,2004). One study did not identifyany redundant SOX protein inhuman germ cells within the groupB of the SOX gene family (com-prising SOX1, SOX2, SOX3,SOX14, and SOX21) (Perrett et al.,2008). Expression analysis of SOXfamily members in TGCTs revealedthat SOX17 (a member of group Fof the SOX gene family) is specifi-cally expressed in CIS and semi-noma but not in embryonal carci-noma (de Jong et al., 2008b). Inaddition, SOX17 maps to the chro-mosomal region 8p23, which isgained in seminoma (Korkolaet al., 2008). This indicates thatSOX17 is a candidate SOX proteinfor cooperation with OCT3/4 in CISand seminoma. These data alsoillustrate that SOX17 is a newmarker to discriminate CIS andseminoma from embryonal carci-noma. Of interest is that SOX17distinguishes embryonic from adulthematopoietic stem cells (Kimet al., 2007). Current researchfocuses on the processes that mayregulate the differential expressionof SOX2 versus SOX17 and on therole of these SOX proteins in thedifferent histologies of the TGCTsubtypes involved. This can be rel-evant in the context of regulationof apoptosis and differentiation aswell as sensitivity to either irradia-tion and/or chemotherapy.Kristensen et al. (2008) studied

the expression of UTF-1 and REX-1in testes and type II TGCTs. Both

Figure 2. Representative examples of HE (A, C, E, and G) and OCT3/4 immunohisto-chemical stainings (B, D, F, and H) of normal and (pre) malignant testis tissue. (A)and (B): testis with normal spermatogenesis; (C) and (D): CIS; (E) and (F): intratub-ular seminoma (arrows) and seminoma (bottom half of pictures); (G) and (H): embry-onal carcinoma.



TABLE

2.Known

and

New

Im

munohisto

chem

icalMark

ers

ofPote

ntialDia

gnosticValu

eto

Discrim

inate

Sem

inom

ato

usHisto

logie

s,Em

bry

onal

Carc

inom

a,and

Sperm

ato

cyticSem

inom

a.See

Textfo

rAdditio

nalInfo

rmation

Mark

er

NT

CIS

ITSE

SE

EC

SS

Refere

nce

s

OCT3/4

21

11

12

(deJo

ngetal.,2005;Lo

oijengaetal.,2003b)

c-KIT

1a

11

12

1/2

2(R

ajpert-D

eMeyts

andSkakkebaek,1994;

Strohmeyeretal.,1995)

SCF

1a

11

12/1

2(B

okemeyeretal.,1996;Stoopetal.,2008)

PLA

P2

11

11

N.D

.(B

urk

eandMostofi,1988;Slowikowsk

a-H

ilcze

retal.,2001)

NANOG

21

11

12

(Hart

etal.,2005)

AP2-c

21

11

1foca

l2

(Hoei-Hanse

netal.,2004;Pauls

etal.,2005)

M2A

21

11

1foca

lN.D

.(F

rankeetal.,2004;Mark

setal.,1999)

SOX2

2(o

ccasional

Sertolice

lls:1)

2(S

ertolice

lls:1)

22

12

(deJo

ngetal.,2008b;Perrett

etal.,2008)

SOX17

21

11

22

(deJo

ngetal.,2008b)

REX-1

11

11

11

(Kristense

netal.,2008)

UTF-

11

11

11

1(K

ristense

netal.,2008)

BOB1

1(a

lsoSertolice

lls:1)

11

1N.D

.N.D

.(G

ash

aw

etal.,2007)

PROM1

1(spora

dic

sperm

atog.)

11

1N.D

.N.D

.(G

ash

aw

etal.,2007)

MCFD

22

(Sertolice

lls:1)

11

(some)

1(n

earlyall)

N.D

.N.D

.(G

ash

aw

etal.,2007)

aOnly

detectable

onfroze

ntiss

ue,notonform

alin

fixedpara

ffin

embeddedmaterial.

NT,norm

aladulttestis;IT

SE,intratu

bularse

minoma;SE,se

minoma;EC,embry

onalca

rcinoma;SS,sp

erm

atocy

ticse

minoma;N.D

,notdeterm

ined.



are transcription factors expressedin mouse embryonic carcinomacells, regulated by OCT3/4 (Hosleret al., 1989; Okuda et al., 1998),and expressed by spermatogoniain prepubertal and adult testes.UTF-1 and REX-1 are expressed inCIS, seminomas, and nonsemino-mas. UTF-1 is present in all, andREX-1 is expressed in most of thetype III spermatocytic seminomasthat were tested, demonstratingthat these markers are alsoexpressed in the more differenti-ated spermatocytes. Despite thefact that these proteins are associ-ated with pluripotency, theirexpression pattern differs signifi-cantly from that of OCT3/4 andSOX2, the well-characterized pluri-potency proteins, which are nega-tive in type III TGCT.Analysis of expression patterns

in microarray studies revealedadditional markers, MCFD2, BOB1,and PROM1, for seminoma com-pared to normal testis (Gashawet al., 2005). Studies demon-strated indeed increased expres-sion levels of these three proteinsin seminoma cells compared tonormal adult testes (Gashaw et al.,2007). Because all three of thesemarkers are also expressed at lowlevels in normal adult testicular tis-sue, their suitability as practicaladditional diagnostic markersremains to be proven.

SCF-C-KIT SIGNALING

IN TGCT

The SCF c-KIT signaling pathway isimportant for proper germ cell de-velopment. This was first illus-trated by mice carrying naturallyoccurring mutations in the Steel(Sl) and White-spotting (W) locithat involved the ligand-receptorpair SCF and c-KIT, respectively(reviewed in Loveland and Schlatt,1997). In addition to defects in thehematopoietic system and melano-genesis, the number of PGCsdecreases to about 2% of the nor-mal level by day E12.5 in Sl/Slmice and W/W mice. By day E14,PGCs are no longer detectable inSl/Sl mice (Mintz and Russell,1957; McCoshen and McCallion,1975). There are differences in

expression patterns between miceand humans, but SCF and c-KITalso play an important role inhuman spermatogenesis, demon-strated by defects in this pathway(reviewed in Mauduit et al., 1999).c-KIT expression, similar to themarkers PLAP, OCT3/4, and Ki67,disappears after birth (Honeckeret al., 2004). In normal adult sper-matogenesis, c-KIT is downregu-lated when the PGCs/gonocytesmature toward spermatogonia A,and it is not detected in the variousstages of normal spermatogenesis(Rajpert-De Meyts et al., 1996). c-KIT expression can only bedetected in adult testis when a sen-sitive detection method is used onfrozen tissue (Natali et al., 1992;Rajpert-De Meyts and Skakkebaek,1994; Strohmeyer et al., 1995;Bokemeyer et al., 1996; Stoopet al., 2008). In adult testis, SCF isexpressed by Sertoli cells. Themembrane bound form of SCF isthe most efficient in establishingand maintaining germ cells (Dolciet al., 1991; Godin et al., 1991;Matsui et al., 1991; Marziali et al.,1993), while the soluble form ofSCF activates c-KIT on Leydig cellsto induce testosterone production(Yan et al., 2000). CIS and semi-noma express c-KIT on their mem-brane as demonstrated for PGCs.Upon reprogramming of CIS toseminoma or nonseminoma, tumorcells predominantly lose c-KITexpression (Rajpert-De Meyts andSkakkebaek, 1994; Strohmeyeret al., 1995; Biermann et al.,2007). Recently, we reported thatSCF immunohistochemistry is avaluable additional marker to dis-criminate precursor lesions ofTGCTs from cells that are delayedin their maturation. This is an im-portant distinction to make, espe-cially in patients with germ cellmaturation delay, which oftenshow prolonged expression of PGCmarkers (Rajpert-De Meyts et al.,1998; Cools et al., 2005; Coolset al., 2006b). Because in thesespecific cases, detection of OCT3/4was not informative per se, the useof SCF as an additional markerfor malignancy is a valuable diag-nostic addition (Stoop et al.,2008).

Comparative genomic hybridiza-tion experiments revealed specificamplification of region 4q12, cen-tered around c-KIT in TGCTs(McIntyre et al., 2005a). However,in 21% of the seminomas, only thecopy number of c-KIT is increasedin this region and in most cases notthe flanking genes KDR andPDGFRA. This indicates that c-KITcan be specifically involved in TGCTetiology. This is also supported bythe clinical observation that Glee-vec in a single chemotherapy-re-sistant seminoma resulted in com-plete response (Pedersini et al.,2007).An association of a single nucleo-

tide polymorphism (SNP) in c-KITwith low sperm counts in idiopathicmale infertility was reported; how-ever, the links to TGCT develop-ment is still unclear (Galan et al.,2006). As is the case in othermalignancies such as mastocyto-sis, mutations in c-KIT have beenreported in TGCTs. Tian et al.(1999) were the first to reportcodon 816 mutations in a semi-noma and in a mixed ovarian dys-germinoma/yolk-sac tumor (Tianet al., 1999). Exon 17 of c-KIT,which harbors codon 816, encodesfor the kinase domain of the recep-tor. Mutations in this area and inparticular codon 816 mutationslead to constitutive activation of c-KIT and can be a transformingevent in oncogenesis (Furitsuet al., 1993; Moriyama et al.,1996; Boissan et al., 2000). In fol-low up on the report by Tian et al.(1999), c-KIT exon 17 mutationshave also been reported by othersin TGCTs (Sakuma et al., 2003;Kemmer et al., 2004; Nakai et al.,2005), in mediastinal seminomas(Przygodzki et al., 2002), and indysgerminomas (Pauls et al.,2004; Hoei-Hansen et al., 2007b).Our experiments revealed thatcodon 816 mutations were pre-dominantly found in patients withbilateral TGCTs (93%), whereasthey were only detected in 1.3% ofpatients with unilateral disease(Looijenga et al., 2003a). Themutations were only found in CISor in invasive tumor of patients andnever in adjacent normal testicularparenchyma or blood, indicating



that these are very early somaticmutations. We hypothesize thatthese mutations occur early inembryogenesis, before migrationof the PGCs to the genital ridges.We concluded that activatingcodon 816 mutations could predictthe development of bilateralTGCTs. Bilateral TGCTs can presentsynchronously, but most casespresent asynchronously and thelatency period can be up to 20years (Colls et al., 1996). There-fore, analyzing material from thefirst TGCT of a patient for c-KITexon 17 mutations can be used forscreening patients at risk for devel-oping bilateral disease. The pres-ence of codon 816 mutations inbilateral TGCTs was confirmed byindependent reports, although thefrequency at which the mutationswere found varied (Tate et al.,2005; Biermann et al., 2007). Incontrast, others did not detect anincreased incidence of these muta-tions in patients with bilateral dis-ease (Coffey et al., 2008; Sakumaet al., 2008). This discrepancy maybe explained by the different tech-niques that were used to detect themutations. In the original report onthe predictive value of codon, 816mutations for bilateral disease,restriction endonuclease-mediatedselective PCR, and single nucleo-tide primer extension were used,techniques likely to be more sensi-tive than regular PCR combinedwith sequencing (Looijenga et al.,2003a). It is also important to notethat the c-KIT mutations were notalways found in the invasive tumorbut were present in CIS adjacent tothe tumor, indicating that depend-ency on c-KIT may become super-fluous when the CIS has pro-gressed to invasiveness (seeabove). This is in line with reportsthat c-KIT expression is oftendownregulated upon progressionfrom CIS to seminoma (Willmore-Payne et al., 2006; Biermannet al., 2007). Therefore, it is impor-tant to use sensitive detectiontechniques and preferably use tu-mor material containing CIS formutation analysis. c-KIT mutationswere not associated with anincreased risk of TGCTs in familiarcases, but again, a higher propor-

tion of mutations was seen inpatients with bilateral disease(Rapley et al., 2004). The authorssuggest that bilateral disease inthe context of familial TGCT has adifferent pathogenesis than thesporadic bilateral cases and othersomatic and susceptibility genesalso play important roles.In addition to the data on the c-

KIT mutations, it is interesting tospeculate whether any othermutated kinases could play a rolein TGCT as well. A large-scale col-laboration aimed at sequencing theprotein kinase family for mutationsin TGCTs revealed a remarkablylow mutation frequency with only asingle somatic point mutation inthe STK10 gene (Bignell et al.,2006). Therefore, there is noevidence implying any mutatedprotein kinase in the developmentof TGCTs, other than the previousreports on the role of c-KITmutations.

EPIGENETICS OF

TYPE II GCTS

Besides a wealth of informationabout the genomic make up of typeII GCTs, increasing knowledge onthe epigenetic constitution is alsoevolving (Peltomaki, 1991; Koulet al., 2002; Smiraglia et al., 2002;Honorio et al., 2003; Zhang et al.,2005; Kawakami et al., 2006; Lindet al., 2006; Ishii et al., 2007; Lindet al., 2007; Okamoto and Kawa-kami, 2007). The role of epige-netics in normal germ cell develop-ment has recently been reviewed(Biermann and Steger, 2007). Tar-geted as well as genome widestudies demonstrate that the semi-nomas show an overall hypome-thylated DNA status, in contrast tothe various histological types ofnonseminomas. Interestingly, thesupernumerical X chromosomesare inactivated in nonseminomasby methylation (Looijenga et al.,1997). This is, similar to normalembryogenesis, the result of thenon(protein)-coding XIST gene.This unique phenomenon in malesis correlated with hypomethylationof the promoter region, which canbe used as molecular target fortype II GCTs in males (Kawakami

et al., 2003; Kawakami et al.,2004). The difference in methyla-tion status can be demonstratedusing expression profiling for thedifferent forms of the DNA methyl-transferases (DNMT) (Looijenga,2008). DNMT1 is required formaintenance of the methylatedstatus during cell division, and haspreviously been found to be pres-ent in differentiated forms of non-seminomas (Omisanjo et al.,2006), while DNMT3A and DNMT3Bare needed for de novo methyla-tion (Karpf and Matsui, 2005), ashappens during early embryogene-sis. DNMT3L has a role in theestablishment of the pattern ofgenomic imprinting (Oakes et al.,2007). Overall, a specific upregula-tion is observed in the embryonalcarcinomas compared to the semi-nomas. Indeed, this is alsoreflected by immunohistochemistryusing a MC-specific antibody (Looi-jenga, 2008; Netto et al., 2008).Interestingly, a methylation studyof the promoter region of OCT3/4showed that in seminoma andembryonal carcinoma, the pro-moter region is predominantlyhypomethylated, both in in vitrocell lines as well as in in vivotumors (de Jong et al., 2007).Microdissection of embryonal carci-noma cells demonstrated completedemethylation. On differentiationof the embryonal carcinoma cells,OCT3/4 is downregulated in expres-sion, associated with hypermethyl-ation of the promoter region. Thispattern most likely reflects the situa-tion in most genes related to pluripo-tency, showing the same pattern ofexpression as OCT3/4, like NANOG.Histone modification has also

been identified as a significant reg-ulatory element specifying whichgenes will be hypermethylated ondifferentiation from an undifferen-tiated stem cell. This is related tothe histone, H3, methylated at Ly-sine27 (H3K27) by polycomb pro-teins, which is a repressive mark,as well as the active mark, methyl-ated H3K4 (Ohm et al., 2007).Interestingly, this was indeedfound to be the case in cell linesderived from type II GCTs, that is,embryonal carcinoma, in which twoadditional repressive marks are



identified, dimethylated H3K9 andtrimethylated H3K9, both associ-ated with DNA hypermethylation inadult cancers. This fits nicely withthe observed pattern of expressionof the histone de-acetylase (HDAC)in these tumors (Omisanjo et al.,2006). More recently, a relatedstudy investigated the expressionof BLIMP-1 and PRMT-5. Expres-sion of these factors in type IIGCTs mimicks the situation of nor-mal development (Eckert et al.,2008a). These proteins are involvedin the suppression of the somaticdifferentiation program in PGCs/gonocytes, related to dimethylatedhistone H2A and H4 (Ancelin et al.,2006). Knock out of these genesresults in differentiation of mousePGCs (Ohinata et al., 2005; Hayashiet al., 2007). Indeed, these proteinsand epigenetic changes are presentin embryonic germ cells, as well asCIS and seminoma, including therepresentative cell line TCam-2. Asexpected, upon formation of embry-onal carcinoma, these proteins aredownregulated, and the dimethy-lated H2A and H4 are removed.Again, these studies demonstratedthe close relationship between nor-mal embryogenesis and type IIGCTs. It remains a challenge toidentify which of the mechanismsare reflecting normal development,and which are related to the patho-genetic process.

THE ROLE OF

MICRO-RNAS IN TGCT

Over the last years, there has beenan increased interest in the role ofmiRNAs in development and in tu-morigenesis. One of the first indi-cations that miRNAs are also im-portant for germ cell developmentcame from observations of Bern-stein et al. (2003), who demon-strated that knockout mice forDicer, a protein required for thegeneration of mature miRNAs, alsolack PGCs. In addition to this, con-ditional knockout mice for Dicer1 inthe male germ line were infertile.The animals suffered from an earlydecrease in germ cell number andan impaired ability to differentiate,indicating that Dicer1 and miRNAsare important for both survival and

proper differentiation of male germcells (Maatouk et al., 2008). Agenetic screen for novel functionsfor miRNAs revealed that miRNAs372 and 373 can overcome cellcycle arrest mediated by WT TP53(Voorhoeve et al., 2006). In addi-tion, miRNAs 372 and 373 areexpressed in TGCT with WT TP53and in several TGCT cell lines withhigh levels of WT TP53. In contrast,TGCT cell lines with low expressionof WT TP53 or with mutant TP53were negative for these miRNAs.MiRNAs 372 and 373 can bypassthe TP53 checkpoint and allowgrowth of type II TGCT in the pres-ence of WT TP53. Studies withhigh-throughput screening forexpression of miRNAs in type IITGCTs and TGCT-derived cell linesrevealed that unsupervised clus-tering of miRNA expression pat-terns can discriminate TGCT celllines, normal testis, the type IIIspermatocytic seminomas, and thedifferentiated versus the undiffer-entiated type II tumor subgroups(Gillis et al., 2007). These datasupport the model that miRNAs areinvolved in regulating differentia-tion of stem cells and germ cells.Another interesting link on the im-

portance of miRNAs for germ cellsand GCTs came from research onthe Dead end gene (DND1). Untilrecently, DND1 was known to regu-late germ-cell viability and to sup-press the formation of germ celltumors. This was based on work inthe 129 -Ter mouse strain where apremature stop codon in DND1 wasidentified (Matin and Nadeau, 2005;Youngren et al., 2005). These miceare characterized by germ-cell lossand a high frequency of (type I)TGCTs. Studies aimed at identifyingsusceptibility genes additional toDND1Ter in the 129 mouse strainshowed that all variants tested(KitlSlJ, Trp53, Ay, and 129-Chr19MOLF) lead to increased germ celltumor frequency and in particular ofbilateral disease (Lam et al., 2007).Recently Kedde et al. (2007) dem-onstrated that DND1 counteractsmiRNA-mediated destabilization ofmRNAs by binding to mRNAs andprohibiting the association of miR-NAs with their target sites. Thisunderlines the important role of miR-

NAs and regulation of miRNAexpression in germ cell develop-ment. Ongoing mechanistic studiesto reveal ‘‘what leads from deadend’’ (Matin, 2007) showed thatDND1 interacts with the multifunc-tional protein APOBEC3, which isexpressed in testes and adult germcells (Bhattacharya et al., 2008).Linger et al. (2008) focused on therole of DND1 in humans and ana-lyzed the presence of DND1 muta-tions in 263 human TGCTs. In onlyone case was a heterozygous variantfound, which was not seen in any ofthe additional 842 TGCT cases ana-lyzed (Linger et al., 2008). In addi-tion, when analyzing eighteen type ITGCT specifically, this variation wasnot found (Looijenga, 2008). Basedon these observations and in con-trast to its role in the type I TGCTs ofthe 129 background mice, the role ofDND1 in humans remains unclearand may indeed appear to be a deadend. However, we are still at the be-ginning of understanding how miR-NAs are involved in regulating manycellular processes relevant for TGCTs.As an illustration of the ongoinginsight into the functional roles ofmiRNAs, Tay et al. (2008) recentlydemonstrated that, in mouse, inaddition to targeting the 30UTR ofmRNAs, miRNAs against NANOG,OCT3/4, and SOX2 also target thecoding region of these genes, whichaffects embryonic stem cell differen-tiation (Tay et al., 2008). Analysis ofthese miRNAs in human TGCTs (Gilliset al., 2007) did not reveal any differ-ences in expression between the dif-ferent subgroups (unpublishedobservations). This underlines thatthe evolutionary difference in miRNAexpression between mice andhumans is considerable (Plasterk,2006). Further research into thefunctional mechanisms of miRNAsand the role of DND in TGCT are likelyto give more interesting clues.

RISK FACTORS OF

TYPE II GCTS

Risk Factors: Support of aTesticular DysgenesisSyndrome (TDS)

The nature of the risk factors forTGCTs, which are familial predispo-



sition, a history of TGCT, cryp-torchidism, infertility, and variousforms of disorders of sex develop-ment (DSD), support the modelthat the initiating step in thepathogenesis of this cancer occursduring embryonal development(Moller, 1993; Skakkebaek et al.,1998; Jacobsen et al., 2000;Raman et al., 2005; Sonke et al.,2007; Cook et al., 2008). Variousother risk factors have been sug-gested, which await confirmation,although birth weight (both lowand high) seems to be relevant(Michos et al., 2007). Seminoma isfound more frequently in intra-ab-dominal gonads than in scrotalgonads (Ogunbiyi et al., 1996).This also likely explains the prefer-ential occurrence of the seminoma-tous type in the ovary (Tewariet al., 2000). In addition, an earlyage of orchiopexy indeed reducesthe risk for a TGCT (Petterssonet al., 2007; Walsh et al., 2007).This is likely related to the stillongoing maturation of PGC/gono-cyte-like cells to prespermatogonia(see above).So far, it has not been possible to

identify the gene or genes involvedin familial TGCTs (Rapley et al.,2000; Holzik et al., 2004). Overall,the genetic predisposition is diffi-cult to investigate because of thesmall sizes of the affected families,(likely linked to) the relationship tosubfertility, as well as the possiblerole of the (micro)environment.Immigrants from Finland to Swe-den, who have a lower initial riskfor TGCTs, obtain the risk of theSwedish population in their secondgeneration (Hemminki et al.,2002). This demonstrates a signifi-cant effect of the environment onthe incidence in a limited period oftime, which is possibly overruling agenetic component. However,there are strong indications for agenetic component (for review seeKrausz and Looijenga, 2008). It isof interest that most identified riskfactors (are expected to) one wayor another negatively affect matu-ration of embryonic germ cells.This led to the hypothesis of TDS(Skakkebaek et al., 2001; Skakke-baek, 2003; Rajpert-De Meyts,2006; Sonne et al., 2008). This

syndrome integrates various ele-ments, in which the final outcomewill have a negative effect on tes-ticular function, including sub(in)-fertility, cryptorchidism, and/or anincreased risk for TGCTs. The roleof the supportive element, that is,Leydig cells and Sertoli cells, is ofcrucial importance. Indeed, thishas also been concluded based onan independent approach, focusedpredominantly on DSD patientswith an increased risk of type IIGCTs (Hersmus et al., 2008a;Looijenga, 2008). In this context, apossible role of estrogen and anti-androgen function needs specificattention (Looijenga, 2008;Krausz and Looijenga, 2008), forreview.

Risk Factors: Disordersof Sex Development

DSD is a group of developmentalanomalies, previously referred toas intersex, which is defined as acondition of incomplete or disor-dered genital or gonadal develop-ment, leading to a discordancebetween genetic sex (i.e., deter-mined by the chromosomal consti-tution, of the X and Y chromo-somes), gonadal sex (the testicularor ovarian development of thegonad), and phenotypic sex (thephysical appearance of the individ-ual). Recently, a revised classifica-tion system has been proposed,with the aim to reduce uncertain-ties on description (Hughes et al.,2006).As mentioned, DSD patients with

either hypovirilization or gonadaldysgenesis specifically show anincreased risk for the developmentof type II GCTs (for review seeCools et al., 2006a; Looijengaet al., 2007). In dysgeneticgonads, the precursor lesion iscalled a gonadoblastoma. It is mor-phologically different from CIS,and precedes the development ofinvasive tumors (Scully, 1970;Robboy et al., 1982; Savage andLowe, 1990). In DSD cases, theprecursor can be either CIS orgonadoblastoma or a combinationof both (Li et al., 2007b), related tothe level of virilization of thegonad. This can be nicely demon-

strated by the use of immunohisto-chemistry for SOX9 (read-out ofSRY function and Sertoli cell differ-entiation), and FOXL2 (granulosacell differentiation) (Hersmuset al., 2008b). These results indi-cate that CIS and gonadoblastomaare part of a histological contin-uum, in which the developmentinto either a Sertoli cell or a granu-losa cell determines the histologicalcontext of the premalignant cells(Hersmus et al., 2008a), elegantlylinking TDS and DSD.Multiple findings support that at

least a certain amount of testoster-one is needed for the precursorlesion to progress to an invasivedisease. Patients with completeandrogen insensitivity have a sig-nificantly lower risk compared topatients with the partial form ofthis disorder (Cools et al., 2005;Cools et al., 2006a; Hannemaet al., 2006). Most likely, this isrelated to the induction of apopto-sis of germ cells in the testis ofpatients with complete androgeninsensitivity, as observed in Kline-felter patients (Wikstrom et al.,2006). Moreover, complete ab-sence or a very low level of testos-terone also diminishes the risk of atype II GCT. This is nicely illus-trated by patients with hypogona-dotropic hypogonadism, which canpresent with cryptorchid testis, butare not reported to develop TGCTs.

Parameters Related toTumor Risk

In contrast to the link betweenovarian differentiation and FOXL2and testicular differentiation andSOX9 (see above), the correlationbetween the presence of the Ychromosome and testicular devel-opment is less obvious (Coolset al., 2007). In fact, no correlationbetween the Y chromosome andtestis development has been iden-tified in patients with sex chromo-somal mosaicisms, for which noexplanation is available so far.The anatomical position of the

gonad also seems to be signifi-cantly related to the risk of malig-nant transformation. This is in linewith the fact that cryptorchidism isindeed one of the strongest risk



factors for type II TGCTs (Batataet al., 1980; Muller et al., 1984;Giwercman et al., 1987; Abrattet al., 1992).The risk of development of type

II GCTs in DSD patients is directlyrelated to the presence of a specificpart of the Y chromosome, knownas the gonadoblastoma region ofthe Y chromosome (GBY) (Page,1987). This area maps around thecentromeric region, and excludesthe SRY gene as candidate. This issupported by the clinical observa-tion that patients with a transloca-tion of the SRY gene to an X chro-mosome or an autosome, whichresults in 46XX males, have noincreased risk for this type of can-cer. Several candidate genes mapwithin the GBY region, of whichTSPY is one of the most interestingones. It stands for testis-specificprotein on the Y chromosome andis a multicopy gene (Vogel andSchmidtke, 1998). It has similar-ities to the DEK/CAN family of pro-teins, and it interacts with cyclinB1, thereby supposed to beinvolved in cell cycle regulation.Various splice variants have beenreported, which indeed can bepresent in type II TGCTs. The cor-responding protein is present inspermatogonia during normal de-velopment. The level of protein isincreased in CIS and gonadoblas-toma, for which the mechanisticbasis is still unknown (Schniederset al., 1996; Hildenbrand et al.,1999; Lau, 1999; Delbridge et al.,2004; Kersemaekers et al., 2005;Li et al., 2007b). The fact that typeII TGCTs are consistently aneuploidmight be related to this. Theincreased level of TSPY protein isused as a supportive parameter todistinguish a malignant germ cellfrom a germ cell with delayed mat-uration. Upon invasive growth,although the Y chromosome canstill be retained, expression of thegene and, subsequently, the pro-tein, is mostly lost. Therefore, theloss of expression is due to down-regulation. Transfection experi-ments demonstrated that inductionof TSPY in human cells lacking thisprotein results in an increase inproliferation, both in vitro and invivo. In fact, the cells show a

shorter G2 phase of the cell cycle(Oram et al., 2006). Interestingly,a subsequent study shows thatmany of the upregulated genes inthe TSPY transfected cells map tothe short arms of chromosome 12.A correlation between the level ofTSPY and expression of thesegenes, including KRAS2 andNANOG, was only found in the pre-cursor lesion CIS, and not in inva-sive tumors (Li et al., 2007a). Thisobservation nicely fits with thedownregulation of TSPY upon pro-gression of the tumor toward inva-siveness.Mice lack TSPY. Transgenic ani-

mals containing a complete struc-tural human TSPY gene interest-ingly show integration in the Ychromosome, in a tandem repeatorganization, like the organizationin the human genome (Schubertet al., 2003). This is intriguing butunexplained so far. However, noGCTs were identified, not atyounger or older age. In otherwords, the simple overexpressionof TSPY in OCT3/4 positive cells isnot enough to generate a type IIGCT in the mouse.

NONINVASIVE AND

INVASIVE DIAGNOSIS OF

CIS: SEMEN ANALYSIS,

MICROLITHIASIS, AND

TESTICULAR BIOPSIES

Given the similar incidence of CISand type II TGCT, all cases of CISare expected to progress to inva-siveness (Giwercman et al.,1991a). CIS can be cured with low-dose irradiation. This results ininfertility in the affected testis, buthormonal function is preserved inthe majority of cases (Giwercmanet al., 1991b). In contrast, devel-opment of an invasive TGCTrequires (hemi-) castration anddepending on the stage and type ofTGCT, irradiation or chemother-apy. Therefore, early diagnosis andtreatment of CIS can prevent de-velopment of TGCT, consecutivetreatment, and therapy-related se-rious diseases such as secondarymalignancies or cardiovascular dis-eases (van den Belt-Duseboutet al., 2007). As CIS is asymptom-

atic in most cases, it is usually notdetected before development ofinvasiveness. Currently, the onlyway to diagnose CIS is by lookingat the morphology and expressionof markers in a testicular biopsy,taken by a surgically invasive pro-cedure (Schmoll et al., 2004).Because of the low incidence ofTGCT and the invasiveness of theprocedure, with possible sideeffects, a testicular biopsy in thegeneral population is not the pre-ferred screening tool.Over the years, there have been

several reports on noninvasivemethods to detect CIS in semensamples. Detection of CIS cells insemen has been reported firstbased on cytology (Czaplicki et al.,1987; Howard et al., 1989).Giwercman et al. (1988a, 1988b,1988c) demonstrated CIS-like cellsin semen by using immunohisto-chemistry for the marker protein,M2A, or by looking for aneuploidcells in semen. The use of placen-tal-like alkaline phosphatase insemen was reported not to be spe-cific due to cross-reacting germ cellalkaline phosphatase (Brackenburyet al., 1993). Unfortunately, noneof these studies resulted in anassay suitable for clinical screeningpurposes, mainly due to frequentfalse positive or false negative find-ings. Recently, there have beennew attempts on this issue usingimmunohistochemical detection of(some relative new) TGCTmarkers. Hoei-Hansen et al.(2007a) tested the suitability ofimmunohistochemical detection ofseveral fetal germ cell-specificmarkers (AP2c and OCT3/4) in thesemen of 503 men including 294infertile patients and 209 patientswith TGCT or other diseases. AP-2cpositive cells were detected in 50%of the cases with CIS, with similarresults for OCT3/4. Notably, ayoung subfertile man who wasincluded as a control subject wasdetected with CIS by the AP-2c-based method. The authors con-clude that immunohistochemicaldetection of CIS cells in semen hasadditional diagnostic value, but anegative result does not excludethe presence of a TGCT (Hoei-Han-sen et al., 2007a). van Casteren



et al. (2008b) looked at the pres-ence of OCT3/4 positive cells in thesemen of 41 men judged to be atrisk for CIS based on a testicularultrasound and 15 controls. OCT3/4 positive cells were detectable byimmunohistochemistry in semen inthe majority of CIS patients tested.Follow-up studies including largerpatient cohorts are crucial, espe-cially because the number of CIS-only cases (which would be the tar-get patient population for thisscreening tool) tested so far is verylow (1 and 3 cases in the reports byHoei-Hansen and van Casteren,respectively). However, thesereports indicate in principle a pos-sibility to use this technique toscreen for CIS specifically inpatients at risk for TGCT. The non-invasive screening could be of usein selecting patients for surgicalbiopsies, thereby reducingunnecessary surgeries. This willespecially be of interest in popula-tions with an increased risk of de-velopment of a TGCT, such asinfertility, bilateral microlithiasis,and a previous unilateral tumor.Within the subgroup of sub(in)-

fertility, bilateral microlithiasis hasbeen identified as a characteristicof males with a high risk (up to20%) for CIS (De Gouveia Brazaoet al., 2004). This is in accordancewith the high incidence of microcal-cifications in patients with a unilat-eral TGCT, and contralateral CIS(Holm et al., 2003). This findingcan be of value for screening pur-poses. A proposal on how to dealwith testicular microlithiasis in aUrological/Andrological Depart-ment is reported elsewhere (vanCasteren et al., in press).When a unilateral TGCT is

removed, it is standard practice inGermany and Denmark to take abiopsy from the contra lateral tes-tis to screen for the presence ofCIS. This is however not the casein most other countries. Studies byDieckmann et al. (2007a, 2007b)reported significant discordantfindings between double biopsiestaken from contralateral testes inTGCT patients in 5.4% of the caseswhen spermatogenesis was eval-uated and in 31.1% of the caseswhen scored for the presence of

CIS. Discordance was significantlyassociated with poor spermatogen-esis and atrophy when spermato-genesis was the endpoint and dis-cordance was predominantly seenin normal-sized testicles of fertileindividuals when screening for CIS.It was therefore suggested to taketwo-site biopsies for fertility evalu-ation and CIS diagnosis. Doublecontralateral biopsies in selectedpatients with large testes, irregularultrasonography results, or microli-thiasis may yield extra information.Before applying double contralat-eral biopsies to all TGCT patients,additional evaluation is warrantedwith respect to the more invasiveprocedure and the risk of inducingadditional damage (Hoei-Hansenet al., 2007c).Careful evaluation is mandatory

when a testicular biopsy is taken.van Casteren et al. (2008a) and deJong et al. studied material from20 TGCT patients who, before de-velopment of the tumor, hadundergone testicular biopsy whereno malignancy was found. Thebiopsies were reviewed blind by anexperienced pathologist using mor-phology (HE staining) and byimmunohistochemistry for c-KIT,PLAP, and OCT3/4. The use ofimmunohistochemistry over mor-phology alone allowed identifica-tion of four additional cases of CIS,an addition of 20%. This illustrateshow crucial the proper choice ofimmunohistochemical markers iswhen reviewing a testicular biopsyfor the presence of CIS.

TREATMENT SENSITIVITY

AND RESISTANCE

It is proposed that the origin oftype II GCTs also explains theiroverall sensitivity to DNA damag-ing agents (i.e., irradiation and cis-platin-based chemotherapy) (Hongand Stambrook, 2004). This is sup-ported by the fact that it is influ-enced by the histological composi-tion of the tumor: loss of embry-onic features results in induction oftreatment resistance (Masters andKoberle, 2003). Selective eradica-tion of embryonic componentsleaves treatment resistant residualteratoma (Oosterhuis, 1983). In

fact, because of exquisite sensitiv-ity to cisplatin-based chemother-apy, up to 80% of patients withmetastatic disease can be cured bya combination of chemotherapyand tumor resection in case of re-sidual disease (Einhorn, 2002).However, even with modern treat-ment, type II GCTs are a deadlydisease in a minority of cases(approximately 5% of all patients),and the biology underlying treat-ment failure is poorly understood(Mayer et al., 2003). Various find-ings point toward a key role forDNA damage response in theexceptional cure rates in GCTsachieved by DNA damaging ther-apy (Bartkova et al., 2007). Therole of mismatch repair (MMR)deficiency and microsatellite insta-bility (MSI) has previously beeninvestigated, yet results have beencontroversial (Mayer et al., 2002;Olasz et al., 2005; Velasco et al.,2008). Patients with tumors show-ing MSI had a higher rate ofrelapse (Velasco et al., 2004), andcancer specific death was associ-ated with MSI and absent or lowexpression of hMLH1 (Velascoet al., 2008). However, the inci-dence of MSI in the subgroup ofpatients with refractory diseasewas not reported in this analysis.We demonstrated a significantlyhigher incidence of MSI in a smallseries of 11 patients with refrac-tory disease, compared to 100unselected cases of GCT, used as acontrol group (Mayer et al., 2002).In sporadic colorectal cancer, MSIis strongly associated with an acti-vating BRAF mutation, leading to aV600E substitution (Yuen et al.,2002; Deng et al., 2004). In onestudy including 62 stage I TGCTs,BRAF mutation V600E was found inthree of 32 nonseminomas (9%)(Sommerer et al., 2005), althoughthis was not correlated with clinicaldata. Other studies could not dem-onstrate mutated BRAF in GCT tu-mor samples or cell lines (Davieset al., 2002; McIntyre et al.,2005b). Extending our earlier anal-ysis, we investigated a series of 35clinically well documented resistantGCTs and the same control groupof 100 unselected GCTs to clarifythe role of MMR deficiency, MSI,



and BRAF mutation status in treat-ment resistance. The resistanttumors had more MSI affecting twoor more loci than controls. Therewas a significantly higher incidenceof BRAF V600E mutation in resist-ant tumors compared to controls.BRAF mutations highly correlatedwith MSI, and MSI and mutatedBRAF correlated with weak orabsent immunohistochemicalstaining for hMLH1. Low levels orabsence of hMLH1 correlated withpromoter hypermethylation. Thepercentage of lack of either hMLH1or MSH6 was significantly higher inthe resistant GCTs compared tocontrols. This is the first observa-tion of a correlation between agene mutation, BRAF V600E, andcisplatin resistance in type II GCTs(Honecker et al., in press). Theseresults hold promises for the futureuse of targeted therapy in type IIGCTs, as recently reported for asingle metastatic seminoma (Ped-ersini et al., 2007). Multi-kinaseinhibitors targeting BRAF, like sor-afenib, are clinically well estab-lished (Hiles and Kolesar, 2008). Inaddition to improving treatment,these findings should encouragesteps toward screening and moni-toring of treatment in type II GCTsin the future.

CONCLUSIONS AND

FUTURE PERSPECTIVES

TGCTs are a fascinating group ofmalignancies for several reasons.As discussed in this review, theycan show differentiation in all line-ages: somatic, extra-embryonal,and even the germ cell lineage.This indicates that these tumorsrepresent the true stem cell popu-lation of the human body. The‘‘stemness’’ of the cells making upthe TGCTs is the reason for theirsecond unique characteristic, theirexquisite sensitivity to treatment.The less differentiated the tumorcells are, the better they willrespond to DNA damaging therapy,as illustrated by the betterresponses of seminomas versusnonseminomas, and by the fre-quent recurrence of mature terato-mas in residual treatment-resist-ant tumors. For proper diagnosis of

the different histological sub-groups, immunohistochemistry isrequired using a panel of suitablemarkers, including OCT3/4, SOX2,and SOX17. Recent developmentssuch as the discovery or the role ofmiRNAs in oncogenesis alsorevealed highly interesting fea-tures of TGCTs. Specific miRNAswere shown to be involved inbypassing the WT p53 pathway,which is another characteristic ofTGCTs. Further research into therole of miRNAs is likely to givemore useful insights in the biologyof TGCTs, as well as that of stemcells. The fact that the incidence ofTGCTs has been increasing overthe last decades and that the re-markable differences in incidencebetween (adjacent) countries, forwhich no explanations are yetreported, calls for further studies.More insight into the pathogenesisof TGCTs is likely to contribute notonly to better treatment of thesetumors but also to a better under-standing of stem cells and onco-genesis in general.

REFERENCES

Abratt RP, Reddi VB, Sarembock LA.1992. Testicular cancer and cryp-torchidism. Br J Urol 70:656–659.

Adami HO, Bergstrom R, Mohner M,et al. 1994. Testicular cancer in ninenorthern European countries. Int JCancer 59:33–38.

Adjaye J, Huntriss J, Herwig R, et al.2005. Primary differentiation in thehuman blastocyst: comparative mo-lecular portraits of inner cell massand trophectoderm cells. Stem Cells23:1514–1525.

Ancelin K, Lange UC, Hajkova P, et al.2006. Blimp1 associates with Prmt5and directs histone arginine methyla-tion in mouse germ cells. Nat CellBiol 8:623–630.

Andrews PW, Damjanov I, Simon D,et al. 1984. Pluripotent embryonalcarcinoma clones derived from thehuman teratocarcinoma cell lineTera-2. Differentiation in vivo and invitro. Lab Invest 50:147–162.

Atlasi Y, Mowla SJ, Ziaee SA, GokhalePJ, et al. 2008. OCT4 spliced variantsare differentially expressed in humanpluripotent and non-pluripotent cells.Stem Cells 26:3068–3074.

Avilion AA, Nicolis SK, Pevny LH, et al.2003. Multipotent cell lineages in earlymouse development depend on SOX2function. Genes Dev 17:126–140.

Babaie Y, Herwig R, Greber B, et al.2007. Analysis of Oct4-dependent

transcriptional networks regulatingself-renewal and pluripotency inhuman embryonic stem cells. StemCells 25:500–510.

Bartkova J, Rajpert-De Meyts E, Skak-kebaek NE, et al. 2007. DNA damageresponse in human testes and testic-ular germ cell tumours: biology andimplications for therapy. Int J Androl30:282–291; discussion 291.

Batata MA, Whitmore WF Jr, Chu FCH,et al. 1980. Cryptorchidism and tes-ticular cancer. J Urol 124:382–387.

Berensztein EB, Sciara MI, RivarolaMA, Belgorosky A. 2002. Apoptosisand proliferation of human testicularsomatic and germ cells during pre-puberty: high rate of testiculargrowth in newborns mediated bydecreased apoptosis. J Clin Endocri-nol Metab 87:5113–5118.

Berney DM, Lee A, Randle SJ, et al.2004. The frequency of intratubularembryonal carcinoma: implicationsfor the pathogenesis of germ celltumours. Histopathology 45:155–161.

Bernstein E, Kim SY, Carmell MA, et al.2003. Dicer is essential for mousedevelopment. Nat Genet 35:215–217.

Bhattacharya C, Aggarwal S, Kumar M,et al. 2008. Mouse apolipoprotein Bediting complex 3 (APOBEC3) isexpressed in germ cells and interactswith dead-end (DND1). PLoS ONE3:e2315.

Biermann K, Steger K. 2007. Epige-netics in male germ cells. J Androl28:466–480.

Biermann K, Goke F, Nettersheim D,et al. 2007. c-KIT is frequentlymutated in bilateral germ celltumours and down-regulated duringprogression from intratubular germcell neoplasia to seminoma. J Pathol213:311–318.

Bignell G, Smith R, Hunter C, et al.2006. Sequence analysis of the pro-tein kinase gene family in humantesticular germ-cell tumors of adoles-cents and adults. Genes Chromo-somes Cancer 45:42–46.

Boiani M, Scholer HR. 2005. Regula-tory networks in embryo-derived plu-ripotent stem cells. Nat Rev Mol CellBiol 6:872–884.

Boissan M, Feger F, Guillosson JJ,Arock M. 2000. c-Kit and c-kit muta-tions in mastocytosis and other hem-atological diseases. J Leukoc Biol67:135–148.

Bokemeyer C, Kuczyk MA, Dunn T,et al. 1996. Expression of stem-cellfactor and its receptor c-kit protein innormal testicular tissue and malig-nant germ-cell tumours. J CancerRes Clin Oncol 122:301–306.

Boyer LA, Lee TI, Cole MF, et al. 2005.Core transcriptional regulatory cir-cuitry in human embryonic stemcells. Cell 122:947–956.

Brackenbury ET, Hargreave TB, HowardGC, McIntyre MA. 1993. Seminal fluid



analysis and fine-needle aspirationcytology in the diagnosis of carcinomain situ of the testis. Eur Urol 23:123–128.

Burke AP, Mostofi FK. 1988. Placentalalkaline phosphatase immunohisto-chemistry of intratubular malignantgerm cells and associated testiculargerm cell tumors. Hum Pathol 19:663–670.

Chemes HE. 2001. Infancy is not a qui-escent period of testicular develop-ment. Int J Androl 24:2–7.

Coffey J, Linger R, Pugh J, et al. 2008.Somatic KIT mutations occur predom-inantly in seminoma germ cell tumorsand are not predictive of bilateral dis-ease: report of 220 tumors and reviewof literature. Genes ChromosomesCancer 47:34–42.

Colls BM, Harvey VJ, Skelton L, et al.1996. Bilateral germ cell testiculartumors in New Zealand: experiencein Auckland and Christchurch 1978–1994. J Clin Oncol 14:2061–2065.

Cook MB, Graubard BI, Rubertone MV,et al. 2008. Perinatal factors and therisk of testicular germ cell tumors.Int J Cancer 122:2600–2606.

Cools M, van Aerde K, KersemaekersAM, et al. 2005. Morphological andimmunohistochemical differencesbetween gonadal maturation delayand early germ cell neoplasia inpatients with undervirilization syn-dromes. J Clin Endocrinol Metab90:5295–5303.

Cools M, Drop SL, Wolffenbuttel KP,et al. 2006a. Germ cell tumors in theintersex gonad: old paths, new direc-tions, moving frontiers. Endocr Rev27:468–484.

Cools M, Honecker F, Stoop H, et al.2006b. Maturation delay of germcells in fetuses with trisomy 21results in increased risk for the de-velopment of testicular germ celltumors. Hum Pathol 37:101–111.

Cools M, Boter M, van Gurp R, et al.2007. Impact of the Y-containing cellline on histological differentiationpatterns in dysgenetic gonads. ClinEndocrinol 67:184–192.

Czaplicki M, Rojewska J, Pykalo R, Szy-manska K. 1987. Detection of testic-ular neoplasms by cytological exami-nation of seminal fluid. J Urol 138:787–788.

Damjanov I, Solter D, Belicza M, SkrebN. 1971. Teratomas obtained throughextrauterine growth of seven-daymouse embryos. J Natl Cancer Inst46:471–475.

Davies H, Bignell GR, Cox C, et al.2002. Mutations of the BRAF gene inhuman cancer. Nature 417:949–954.

De Gouveia Brazao CA, Pierik FH, Oos-terhuis JW, et al. 2004. Bilateral tes-ticular microlithiasis predicts devel-opment of malignant testicular germcell tumours in subfertile men. J Urol171:158–160.

de Jong J, Looijenga LH. 2006. Stemcell marker OCT3/4 in tumor biology

and germ cell tumor diagnostics: his-tory and future. Crit Rev Oncog12:171–203.

de Jong J, Stoop H, Dohle GR, et al.2005. Diagnostic value of OCT3/4 forpre-invasive and invasive testiculargerm cell tumours. J Pathol206:242–249.

de Jong J, Stoop H, Gillis AJ, et al.2008a. Further characterization ofthe first seminoma cell line TCam-2.Genes Chromosomes Cancer 47:185–196.

de Jong J, Stoop H, Gillis AJ, et al.2008b. Differential expression ofSOX17 and SOX2 in germ cells andstem cells has biological and clinicalimplications. J Pathol 215:21–30.

de Jong J, Weeda S, Gillis AJM, et al.2007. Differential methylation of theOCT3/4 upstream region in primaryhuman testicular germ cell tumors.Oncol Rep 127–132.

Delbridge ML, Longepied G, DepetrisD, et al. 2004. TSPY, the candidategonadoblastoma gene on the humanY chromosome, has a widelyexpressed homologue on the X -implications for Y chromosomeevolution. Chromosome Res 12:345–356.

Deng G, Bell I, Crawley S, et al. 2004.BRAF mutation is frequently presentin sporadic colorectal cancer withmethylated hMLH1, but not in hered-itary nonpolyposis colorectal cancer.Clin Cancer Res 10:191–195.

Dieckmann KP, Kulejewski M, Pichlme-ier U, Loy V. 2007a. Diagnosis ofcontralateral testicular intraepithelialneoplasia (TIN) in patients with tes-ticular germ cell cancer: systematictwo-site biopsies are more sensitivethan a single random biopsy. Eur Urol51:175–183, disc 183–185.

Dieckmann KP, Linke J, Pichlmeier U,et al. 2007b. Spermatogenesis in thecontralateral testis of patients withtesticular germ cell cancer: histologi-cal evaluation of testicular biopsiesand a comparison with healthymales. BJU Int 99:1079–1085.

Dolci S, Williams DE, Ernst MK, et al.1991. Requirement for mast cell growthfactor for primordial germ cell survivalin culture. Nature 352:809–811.

Donovan PJ. 1998. The germ cell–themother of all stem cells. Int J DevBiol 42:1043–1050.

Eckert D, Biermann K, Nettersheim D,et al. 2008a. Expression of BLIMP1/PRMT5 and concurrent histone H2A/H4 arginine 3 dimethylation in fetalgerm cells, CIS/IGCNU and germ celltumors. BMC Dev Biol 8:106.

Eckert D, Nettersheim D, Heukamp LC,et al. 2008b. TCam-2 but not JKT-1cells resemble seminoma in cell cul-ture. Cell Tissue Res 331:529–538.

Einhorn LH. 2002. Curing metastatictesticular cancer. Proc Natl Acad SciUSA 99:4592–4595.

Falin LI. 1969. The development ofgenital glands and the origin of germ

cells in human embryogenesis. ActaAnat (Basel) 72:195–232.

Fogh J. 1978. Cultivation, characteriza-tion, and identification of human tu-mor cells with emphasis on kidney,testis, and bladder tumors. Natl Can-cer Inst Monogr 49:5–9.

Franke FE, Pauls K, Rey R, et al. 2004.Differentiation markers of Sertolicells and germ cells in fetal and earlypostnatal human testis. AnatEmbryol (Berl) 209:169–177.

Furitsu T, Tsujimura T, Tono T, et al.1993. Identification of mutations inthe coding sequence of the proto-oncogene c-kit in a human mast cellleukemia cell line causing ligand-in-dependent activation of c-kit prod-uct. J Clin Invest 92:1736–1744.

Galan JJ, De Felici M, Buch B, et al.2006. Association of genetic markerswithin the KIT and KITLG genes withhuman male infertility. Hum Reprod21:3185–3192.

Gashaw I, Grummer R, Klein-Hitpass L,et al. 2005. Gene signatures of tes-ticular seminoma with emphasis onexpression of ets variant gene 4. CellMol Life Sci 62:2359–2368.

Gashaw I, Dushaj O, Behr R, et al. 2007.Novel germ cell markers characterizetesticular seminoma and fetal testis.Mol Hum Reprod 13:721–727.

Gillis AJ, Stoop HJ, Hersmus R, et al.2007. High-throughput microRNAomeanalysis in human germ cell tumours.J Pathol 213:319–328.

Ginsburg M, Snow MH, McLaren A.1990. Primordial germ cells in themouse embryo during gastrulation.Development 110:521–528.

Giwercman A, Grindsted J, Hansen B,et al. 1987. Testicular cancer risk inboys with maldescended testis: acohort study. J Urol 138:1214–1216.

Giwercman A, Clausen OP, SkakkebaekNE. 1988a. Carcinoma in situ of thetestis: aneuploid cells in semen. BrMed J (Clin Res Ed) 296:1762–1764.

Giwercman A, Marks A, Bailey D, et al.1988b. A monoclonal antibody as amarker for carcinoma in situ germcells of the human adult testis.APMIS 96:667–670.

Giwercman A, Marks A, SkakkebaekNE. 1988c. Carcinoma-in-situ germ-cells exfoliated from seminiferous ep-ithelium into seminal fluid. Lancet1:530.

Giwercman A, Muller J, SkakkebaekNE. 1991a. Prevalence of carcinomain situ and other histopathologicalabnormalities in testes from 399 menwho died suddenly and unexpect-edly. J Urol 145:77–80.

Giwercman A, von der Maase H, Ber-thelsen JG, et al. 1991b. Localizedirradiation of testes with carcinomain situ: effects on Leydig cell functionand eradication of malignant germcells in 20 patients. J Clin EndocrinolMetab 73:596–603.

Giwercman A, Carlsen E, Keiding N,Skakkebaek NE. 1993. Evidence for



increasing incidence of abnormalitiesof the human testis: a review. Envi-ron Health Perspect 101 (Suppl2):65–71.

Goddard NC, McIntyre A, SummersgillB, et al. 2007. KIT and RAS signalingpathways in testicular germ celltumours: new data and a review ofthe literature. Int J Androl 30:337–348; discussion349.

Godin I, Deed R, Cooke J, et al. 1991.Effects of the steel gene product onmouse primordial germ cells in cul-ture. Nature 352:807–809.

Hannema S, Scott I, Rajpert-De MeytsE, et al. 2006. Testicular develop-ment in the complete androgeninsensitivity syndrome. J Pathol 208:518–527.

Hart AH, Hartley L, Parker K, et al.2005. The pluripotency homeoboxgene NANOG is expressed in humangerm cell tumors. Cancer 104:2092–2098.

Hayashi K, de Sousa Lopes SM, SuraniMA. 2007. Germ cell specification inmice. Science 316:394–396.

Hemminki K, Li X, Czene K. 2002. Can-cer risks in first-generation immi-grants to Sweden. Int J Cancer99:218–228.

Hersmus R, de Leeuw H, WolffenbuttelKP, et al. 2008a. New insights intotype II germ cell tumor pathogenesisbased on the studies of patients withvarious forms of disorders of sex de-velopment (DSD). Mol Cell Endocri-nol 291:1–10.

Hersmus R, Kalfa N, de Leeuw B, et al.2008b. FOXL2 and SOX9 as parame-ters of female and male gonadal dif-ferentiation in patients with variousforms of disorders of sex develop-ment (DSD). J Pathol 215:31–38.

Hildenbrand R, Schroder W, Brude E,et al. 1999. Detection of TSPY pro-tein in a unilateral microscopic gona-doblastoma of a Turner mosaicpatient with a Y-derived markerchromosome. J Pathol 189:623–626.

Hiles JJ, Kolesar JM. 2008. Role ofsunitinib and sorafenib in the treat-ment of metastatic renal cell carci-noma. Am J Health Syst Pharm65:123–131.

Hoei-Hansen CE, Nielsen JE, AlmstrupK, et al. 2004. Transcription factorAP-2gamma is a developmentallyregulated marker of testicular carci-noma in situ and germ cell tumors.Clin Cancer Res 10:8521–8530.

Hoei-Hansen CE, Almstrup K, NielsenJE, et al. 2005. Stem cell pluripo-tency factor NANOG is expressed inhuman fetal gonocytes, testicularcarcinoma in situ and germ celltumours. Histopathology 47:48–56.

Hoei-Hansen CE, Carlsen E, JorgensenN, et al. 2007a. Towards a non-inva-sive method for early detection oftesticular neoplasia in semen sam-ples by identification of fetal germcell-specific markers. Hum Reprod22:167–173.

Hoei-Hansen CE, Kraggerud SM, AbelerVM, et al. 2007b. Ovarian dysgermi-nomas are characterised by frequentKIT mutations and abundant expres-sion of pluripotency markers. MolCancer 6:12.

Hoei-Hansen CE, Rajpert-De Meyts E,Daugaard G, et al. 2007c. Does morethan one biopsy of the contralateraltestis in men with a germ cell tumoradd value? Nat Clin Pract Urol 4:652–653.

Holm M, Hoei-Hansen CE, Rajpert-DeMeyts E, Skakkebaek NE. 2003.Increased risk of carcinoma in situ inpatients with testicular germ cellcancer with ultrasonic microlithiasisin the contralateral testicle. J Urol170:1163–1167.

Holzik MF, Rapley EA, Hoekstra HJ,et al. 2004. Genetic predisposition totesticular germ-cell tumours. LancetOncol 5:363–371.

Honecker F, Stoop H, de Krijger RR,et al. 2004. Pathobiological implica-tions of the expression of markers oftesticular carcinoma in situ by fetalgerm cells. J Pathol 203:849–857.

Honecker F, Stoop H, Mayer F, et al.2006. Germ cell lineage differentia-tion in non-seminomatous germ celltumours. J Pathol 208:395–400.

Honecker F, Wermann H, Mayer F,et al. Microsatellite Instability, Mis-match Repair Deficiency, And BRAFMutation in Treatment ResistantGerm Cell Tumors. J Clin Oncol (inpress).

Hong Y, Stambrook PJ. 2004. Restora-tion of an absent G1 arrest and pro-tection from apoptosis in embryonicstem cells after ionizing radiation.Proc Natl Acad Sci USA 101:14443–14448.

Honorio S, Agathanggelou A, WernertN, et al. 2003. Frequent epigeneticinactivation of the RASSF1A tumoursuppressor gene in testiculartumours and distinct methylationprofiles of seminoma and nonsemi-noma testicular germ cell tumours.Oncogene 22:461–466.

Hosler BA, LaRosa GJ, Grippo JF,Gudas LJ. 1989. Expression of REX-1, a gene containing zinc fingermotifs, is rapidly reduced by retinoicacid in F9 teratocarcinoma cells. MolCell Biol 9:5623–5629.

Howard GC, Hargreave TB, McIntyreMA. 1989. Carcinoma in-situ of thetestis diagnosed on semen cytology.Clin Radiol 40:323–324.

Hughes IA, Houk C, Ahmed SF, Lee PA.2006. Consensus statement on man-agement of intersex disorders. ArchDis Child 91:554–563.

Ishii T, Kohu K, Yamada S, et al. 2007.Up-regulation of DNA-methyltrans-ferase 3A expression is associatedwith hypomethylation of intron 25 inhuman testicular germ cell tumors.Tohoku J Exp Med 212:177–190.

Jacobsen R, Bostofte E, Engholm G,et al. 2000. Risk of testicular cancer

in men with abnormal semen charac-teristics: cohort study. BMJ 321:789–792.

Karpf AR, Matsui S. 2005. Genetic dis-ruption of cytosine DNA methyltrans-ferase enzymes induces chromo-somal instability in human cancercells. Cancer Res 65:8635–8639.

Kawakami T, Okamoto K, Sugihara H,et al. 2003. The roles of supernumer-ical X chromosomes and XISTexpression in testicular germ celltumors. J Urol 169:1546–1552.

Kawakami T, Okamoto K, Ogawa O,Okada Y. 2004. XIST unmethylatedDNA fragments in male-derivedplasma as a tumour marker for tes-ticular cancer. Lancet 363:40–42.

Kawakami T, Zhang C, Okada Y, Oka-moto K. 2006. Erasure of methyla-tion imprint at the promoter andCTCF-binding site upstream of H19 inhuman testicular germ cell tumors ofadolescents indicate their fetal germcell origin. Oncogene 25:3225–3236.

Kedde M, Strasser MJ, Boldajipour B,et al. 2007. RNA-binding proteinDnd1 inhibits microRNA access totarget mRNA. Cell 131:1273–1286.

Kehler J, Tolkunova E, Koschorz B,et al. 2004. Oct4 is required for pri-mordial germ cell survival. EMBORep 5:1078–1083.

Kemmer K, Corless CL, Fletcher JA,et al. 2004. KIT mutations are com-mon in testicular seminomas. Am JPathol 164:305–313.

Kersemaekers AM, Honecker F, StoopH, et al. 2005. Identification of germcells at risk for neoplastic transfor-mation in gonadoblastoma. HumPathol 36:512–521.

Kim I, Saunders TL, Morrison SJ. 2007.Sox17 dependence distinguishes thetranscriptional regulation of fetalfrom adult hematopoietic stem cells.Cell 130:470–483.

Korkola JE, Heck S, Olshen AB, et al.2008. In vivo differentiation andgenomic evolution in adult malegerm cell tumors. Genes Chromo-somes Cancer 47:43–55.

Koul S, Houldsworth J, MansukhaniMM, et al. 2002. Characteristic pro-moter hypermethylation signaturesin male germ cell tumors. Mol Cancer1:8.

Krausz C, Looijenga LHJ. Geneticaspects of testicular germ celltumors. Cell Cycle 7:3519–3524.

Kristensen DM, Nielsen JE, SkakkebaekNE, et al. 2008. Presumed pluripo-tency markers UTF-1 and REX-1 areexpressed in human adult testes andgerm cell neoplasms. Hum Reprod23:775–782.

Lam MY, Heaney JD, Youngren KK,et al. 2007. Trans-generational epis-tasis between Dnd1Ter and othermodifier genes controls susceptibilityto testicular germ cell tumors. HumMol Genet 16:2233–2240.

Lau SK, Weiss LM, Chu PG. 2007.Association of intratubular seminoma



and intratubular embryonal carci-noma with invasive testicular germcell tumors. Am J Surg Pathol31:1045–1049.

Lau YF. 1999. Gonadoblastoma, testic-ular and prostate cancers, and theTSPY gene. Am J Hum Genet64:921–927.

Ledford H. 2007. Doubts raised overstem-cell marker. Nature 449:647.

Lengner CJ, Camargo FD, HochedlingerK, et al. 2007. Oct4 expression is notrequired for mouse somatic stem cellself-renewal. Cell Stem Cell 1:403–415.

Li Y, Tabatabai ZL, Lee TL, et al.2007a. The Y-encoded TSPY protein:a significant marker potentially playsa role in the pathogenesis of testicu-lar germ cell tumors. Hum Pathol38:1470–1481.

Li Y, Vilain E, Conte F, et al. 2007b.Testis-specific protein Y-encodedgene is expressed in early and latestages of gonadoblastoma and testic-ular carcinoma in situ. Urol Oncol25:141–146.

Liedtke S, Enczmann J, Waclawczyk S,et al. 2007. Oct4 and its pseudo-genes confuse stem cell research.Cell Stem Cell 1:364–366.

Lind GE, Skotheim RI, Fraga MF, et al.2006. Novel epigenetically deregulatedgenes in testicular cancer includehomeobox genes and SCGB3A1 (HIN-1). J Pathol 210:441–449.

Lind GE, Skotheim RI, Lothe RA. 2007.The epigenome of testicular germcell tumors. APMIS 115:1147–1160.

Linger R, Dudakia D, Huddart R, et al.2008. Analysis of the DND1 gene inmen with sporadic and familial testic-ular germ cell tumors. Genes Chro-mosomes Cancer 47:247–252.

Looijenga LH, Olie RA, van der Gaag I,et al. 1994. Seminomas of the caninetestis. Counterpart of spermatocyticseminoma of men? Lab Invest71:490–496.

Looijenga LH, Gillis AJ, van Gurp RJ,et al. 1997. X inactivation in humantesticular tumors. XIST expressionand androgen receptor methylationstatus. Am J Pathol 151:581–590.

Looijenga LH, de Munnik H, OosterhuisJW. 1999. A molecular model for thedevelopment of germ cell cancer. IntJ Cancer 83:809–814.

Looijenga LH, de Leeuw H, van Oor-schot M, et al. 2003a. Stem cell fac-tor receptor (c-KIT) codon 816 muta-tions predict development of bilateraltesticular germ-cell tumors. CancerRes 63:7674–7678.

Looijenga LH, Stoop H, de Leeuw HP,et al. 2003b. POU5F1 (OCT3/4) iden-tifies cells with pluripotent potentialin human germ cell tumors. CancerRes 63:2244–2250.

Looijenga LH, Hersmus R, OosterhuisJW, et al. 2007. Tumor risk in disor-ders of sex development (DSD). BestPract Res Clin Endocrinol Metab21:480–495.

Looijenga LHJ. 2008. Risk factors andgenetic characterization. In: LagunaMP, Albers P, Richie JP, BokemeyerC, editors. Cancer of the testis.Springer, in press.

Loveland KL, Schlatt S. 1997. Stemcell factor and c-kit in the mamma-lian testis: lessons originating fromMother Nature’s gene knockouts. JEndocrinol 153:337–344.

Maatouk DM, Loveland KL, McManusMT, et al. 2008. Dicer1 is requiredfor differentiation of the mouse malegermline. Biol Reprod 79:696–703.

Marks A, Sutherland DR, Bailey D,et al. 1999. Characterization and dis-tribution of an oncofetal antigen(M2A antigen) expressed on testicu-lar germ cell tumours. Br J Cancer80:569–578.

Marziali G, Lazzaro D, Sorrentino V.1993. Binding of germ cells to mu-tant Sld Sertoli cells is defective andis rescued by expression of thetransmembrane form of the c-kitligand. Dev Biol 157:182–190.

Masters JR, Koberle B. 2003. Curingmetastatic cancer: lessons from tes-ticular germ-cell tumours. Nat RevCancer 3:517–525.

Matin A. 2007. What leads from dead-end? Cell Mol Life Sci 64:1317–1322.

Matin A, Nadeau JH. 2005. Search fortesticular cancer gene hits dead-end.Cell Cycle 4:1136–1138.

Matsui Y, Toksoz D, Nishikawa S, et al.1991. Effect of Steel factor and leu-kaemia inhibitory factor on murineprimordial germ cells in culture. Na-ture 353:750–752.

Mauduit C, Hamamah S, Benahmed M.1999. Stem cell factor/c-kit systemin spermatogenesis. Hum ReprodUpdate 5:535–545.

Mayer F, Gillis AJ, Dinjens W, et al.2002. Microsatellite instability ofgerm cell tumors is associated withresistance to systemic treatment.Cancer Res 62:2758–2760.

Mayer F, Honecker F, Looijenga LH,Bokemeyer C. 2003. Towards anunderstanding of the biological basisof response to cisplatin-based chem-otherapy in germ-cell tumors. AnnOncol 14:825–832.

Mc KD, Hertig AT, Adams EC, DanzigerS. 1953. Histochemical observationson the germ cells of human embryos.Anat Rec 117:201–219.

McCoshen JA, McCallion DJ. 1975. Astudy of the primordial germ cellsduring their migratory phase in Steelmutant mice. Experientia 31:589–590.

McGlynn KA, Devesa SS, Graubard BI,Castle PE. 2005. Increasing incidenceof testicular germ cell tumors amongblack men in the United States. J ClinOncol 23:5757–5761.

McIntyre A, Summersgill B, Grygale-wicz B, et al. 2005a. Amplificationand overexpression of the KIT geneis associated with progression in theseminoma subtype of testicular germ

cell tumors of adolescents andadults. Cancer Res 65:8085–8089.

McIntyre A, Summersgill B, SpendloveHE, et al. 2005b. Activating muta-tions and/or expression levels of ty-rosine kinase receptors GRB7, RAS,and BRAF in testicular germ celltumors. Neoplasia 7:1047–1052.

Michos A, Xue F, Michels KB. 2007.Birth weight and the risk of testicularcancer: a meta-analysis. Int J Cancer121:1123–1131.

Mintz B, Russell ES. 1957. Gene-induced embryological modificationsof primordial germ cells in themouse. J Exp Zool 134:207–237.

Mizuno Y, Gotoh A, Kamidono S, Kita-zawa S. 1993. Establishment andcharacterization of a new human tes-ticular germ cell tumor cell line(TCam-2). Nippon Hinyokika GakkaiZasshi 84:1211–1218.

Moller H. 1993. Clues to the aetiologyof testicular germ cell tumours fromdescriptive epidemiology. Eur Urol23:8–13; discussion 14–15.

Molyneaux K, Wylie C. 2004. Primor-dial germ cell migration. Int J DevBiol 48:537–544.

Molyneaux KA, Stallock J, Schaible K,Wylie C. 2001. Time-lapse analysis ofliving mouse germ cell migration.Dev Biol 240:488–498.

Molyneaux KA, Zinszner H, Kunwar PS,et al. 2003. The chemokine SDF1/CXCL12 and its receptor CXCR4 regu-latemouse germ cell migration and sur-vival. Development 130:4279–4286.

Moriyama Y, Tsujimura T, HashimotoK, et al. 1996. Role of aspartic acid814 in the function and expression ofc-kit receptor tyrosine kinase. J BiolChem 271:3347–3350.

Muller J, Skakkebaek NE. 1983. Quan-tification of germ cells and seminifer-ous tubules by stereological exami-nation of testicles from 50 boys whosuffered from sudden death. Int JAndrol 6:143–156.

Muller J, Skakkebaek NE, Nielsen OH,Graem N. 1984. Cryptorchidism andtestis cancer. Cancer 54:629–634.

Nakai Y, Nonomura N, Oka D, et al.2005. KIT (c-kit oncogene product)pathway is constitutively activated inhuman testicular germ cell tumors.Biochem Biophys Res Commun337:289–296.

Natali PG, Nicotra MR, Sures I, et al.1992. Expression of c-kit receptor innormal and transformed human non-lymphoid tissues. Cancer Res 52:6139–6143.

Netto GJ, Nakai Y, Nakayama M, et al.2008. Global DNA hypomethylationin intratubular germ cell neoplasiaand seminoma, but not in nonsemi-nomatous male germ cell tumors.Mod Pathol 21:1337-1344.

Nichols J, Zevnik B, Anastassiadis K,et al. 1998. Formation of pluripotentstem cells in the mammalian embryodepends on the POU transcriptionfactor Oct4. Cell 95:379–391.



Niwa H, Miyazaki J, Smith AG. 2000.Quantitative expression of Oct-3/4defines differentiation, dedifferentia-tion or self-renewal of ES cells. NatGenet 24:372–376.

Oakes CC, La Salle S, Smiraglia DJ,et al. 2007. A unique configuration ofgenome-wide DNA methylation pat-terns in the testis. Proc Natl Acad SciUSA 104:228–233.

Ogunbiyi JO, Shittu OB, Aghadiuno PU,Lawani J. 1996. Seminoma arising incryptorchid testes in Nigerian males.East Afr Med J 73:129–132.

Ohinata Y, Payer B, O’Carroll D, et al.2005. Blimp1 is a critical determinantof the germ cell lineage in mice. Na-ture 436:207–213.

Ohm JE, McGarvey KM, Yu X, et al.2007. A stem cell-like chromatin pat-tern may predispose tumor suppres-sor genes to DNA hypermethylationand heritable silencing. Nat Genet39:237–242.

Okamoto K, Kawakami T. 2007. Epige-netic profile of testicular germ celltumours. Int J Androl 30:385–392;discussion 392.

Okuda A, Fukushima A, Nishimoto M,et al. 1998. UTF1, a novel transcrip-tional coactivator expressed in pluri-potent embryonic stem cells andextra-embryonic cells. EMBO J 17:2019–2032.

Olasz J, Mandoky L, Geczi L, et al.2005. Influence of hMLH1 methyla-tion, mismatch repair deficiency andmicrosatellite instability on chemore-sistance of testicular germ-celltumors. Anticancer Res 25:4319–4324.

Omisanjo OA, Biermann K, HartmannS, et al. 2006. DNMT1 and HDAC1gene expression in impaired sperma-togenesis and testicular cancer. His-tochem Cell Biol 127:175–181.

Oosterhuis JW. 1983. The metastasisof human teratomas. In: DamjanovI, Knowles B, Solter D, editors. Thehuman teratomas.1st ed. Clifton, NJ:Humana Press. pp.137–171.

Oosterhuis JW, Kersemaekers AM,Jacobsen GK, et al. 2003. Morphol-ogy of testicular parenchyma adja-cent to germ cell tumours. An interimreport. APMIS 111:32–40; discussion41–32.

Oosterhuis JW, Looijenga LH. 2005.Testicular germ-cell tumours in abroader perspective. Nat Rev Cancer5:210–222.

Oram SW, Liu XX, Lee TL, et al. 2006.TSPY potentiates cell proliferationand tumorigenesis by promoting cellcycle progression in HeLa andNIH3T3 cells. BMC Cancer 6:154.

Page DC. 1987. Hypothesis: a Y-chro-mosomal gene causes gonadoblas-toma in dysgenetic gonads. Develop-ment 101(Suppl):151–155.

Pauls K, Wardelmann E, Merkelbach-Bruse S, et al. 2004. c-KIT codon816 mutation in a recurrent and met-astatic dysgerminoma of a 14-year-

old girl: case study. Virchows Arch445:651–654.

Pauls K, Jager R, Weber S, et al. 2005.Transcription factor AP-2gamma, anovel marker of gonocytes and semi-nomatous germ cell tumors. Int JCancer 115:470–477.

Pedersini R, Vattemi E, Mazzoleni G,Graiff C. 2007. Complete responseafter treatment with imatinib in pre-treated disseminated testicular semi-noma with overexpression of c-KIT.Lancet Oncol 8:1039–1040.

Peltomaki P. 1991. DNA methylationchanges in human testicular cancer.Biochim Biophys Acta 1096:187–196.

Perrett RM, Turnpenny L, Eckert JJ,et al. 2008. The early human germcell lineage does not express SOX2during in vivo development or uponin vitro culture. Biol Reprod 78:852–858.

Pettersson A, Richiardi L, NordenskjoldA, et al. 2007. Age at surgery forundescended testis and risk of testic-ular cancer. N Engl J Med 356:1835–1841.

Plasterk RH. 2006. Micro RNAs in ani-mal development. Cell 124:877–881.

Player A, Wang Y, Bhattacharya B,et al. 2006. Comparisons betweentranscriptional regulation and RNAexpression in human embryonic stemcell lines. Stem Cells Dev 15:315–323.

Przygodzki RM, Hubbs AE, Zhao FQ,O’Leary TJ. 2002. Primary mediasti-nal seminomas: evidence of singleand multiple KIT mutations. LabInvest 82:1369–1375.

Rabinovici J, Jaffe RB. 1990. Develop-ment and regulation of growth anddifferentiated function in human andsubhuman primate fetal gonads.Endocr Rev 11:532–557.

Rajaraman R, Rajaraman MM, Rajara-man SR, Guernsey DL. 2005. Neo-sis–a paradigm of self-renewal incancer. Cell Biol Int 29:1084–1097.

Rajpert-De Meyts E. 2006. Develop-mental model for the pathogenesis oftesticular carcinoma in situ: geneticand environmental aspects. HumReprod Update 12:303–323.

Rajpert-De Meyts E, Skakkebaek NE.1994. Expression of the c-kit proteinproduct in carcinoma-in-situ andinvasive testicular germ cell tumours.Int J Androl 17:85–92.

Rajpert-De Meyts E, Kvist M, Skakke-baek NE. 1996. Heterogeneity ofexpression of immunohistochemicaltumour markers in testicular carci-noma in situ: pathogenetic rele-vance. Virchows Arch 428:133–139.

Rajpert-De Meyts E, Jorgensen N,Brondum-Nielsen K, et al. 1998. De-velopmental arrest of germ cells inthe pathogenesis of germ cell neopla-sia. APMIS 106:198–204; discussion204–196.

Rajpert-De Meyts E, Bartkova J, Sam-son M, et al. 2003. The emerging

phenotype of the testicular carci-noma in situ germ cell. APMIS111:267–278; discussion 278–269.

Raman JD, Nobert CF, Goldstein M.2005. Increased incidence of testicu-lar cancer in men presenting withinfertility and abnormal semen analy-sis. J Urol 174:1819–1822; discus-sion1822.

Rapley EA, Crockford GP, Teare D,et al. 2000. Localization to Xq27 of asusceptibility gene for testiculargerm-cell tumours. Nat Genet 24:197–200.

Rapley EA, Hockley S, Warren W, et al.2004. Somatic mutations of KIT infamilial testicular germ cell tumours.Br J Cancer 90:2397–2401.

Richiardi L, Bellocco R, Adami HO,et al. 2004. Testicular cancer inci-dence in eight northern Europeancountries: secular and recent trends.Cancer Epidemiol Biomarkers Prev13:2157–2166.

Robboy SJ, Miller T, Donahoe PK, et al.1982. Dysgenesis of testicular andstreak gonads in the syndrome ofmixed gonadal dysgenesis: perspec-tive derived from a clinicopathologicanalysis of twenty-one cases. HumPathol 13:700–716.

Sakuma Y, Sakurai S, Oguni S, et al.2003. Alterations of the c-kit gene intesticular germ cell tumors. CancerSci 94:486–491.

Sakuma Y, Matsukuma S, Yoshihara M,et al. 2008. Mutations of c-kit genein bilateral testicular germ celltumours in Japan. Cancer Lett259:119–126.

Savage MO, Lowe DG. 1990. Gonadalneoplasia and abnormal sexual dif-ferentiation. Clin Endocrinol (Oxf)32:519–533.

Schmoll HJ, Souchon R, Krege S, et al.2004. European consensus on diag-nosis and treatment of germ cell can-cer: a report of the European GermCell Cancer Consensus Group(EGCCCG). Ann Oncol 15:1377–1399.

Schnieders F, Dork T, Arnemann J, et al.1996. Testis-specific protein, Y-encoded (TSPY) expression in testiculartissues. HumMol Genet 5:1801–1807.

Schubert S, Skawran B, Dechend F,et al. 2003. Generation and charac-terization of a transgenic mouse witha functional human TSPY. BiolReprod 69:968–975.

Scully RE. 1970. Gonadoblastoma: areview of 74 cases. Cancer 25:1340–1356.

Shah MN, Devesa SS, Zhu K, McGlynnKA. 2007. Trends in testicular germcell tumours by ethnic group in theUnited States. Int J Androl 30:206–213; discussion 213–204.

Skakkebaek NE. 1972. Possible carci-noma-in-situ of the testis. Lancet2:516–517.

Skakkebaek NE. 2003. Testicular dys-genesis syndrome. Horm Res 60(Suppl 3):49.



Skakkebaek NE, Rajpert-De Meyts E,Jorgensen N, et al. 1998. Germ cellcancer and disorders of spermatogen-esis: an environmental connection?APMIS 106:3–11; discussion 12.

Skakkebaek NE, Rajpert-De Meyts E,Main KM. 2001. Testicular dysgenesissyndrome: an increasingly commondevelopmental disorder with environ-mental aspects. Hum Reprod16:972–978.

Slowikowska-Hilczer J, Walczak-Jedr-zejowska R, Kula K. 2001. Immuno-histochemical diagnosis of preinva-sive germ cell cancer of the testis.Folia Histochem Cytobiol 39:67–72.

Smiraglia DJ, Szymanska J, KraggerudSM, et al. 2002. Distinct epigeneticphenotypes in seminomatous andnonseminomatous testicular germcell tumors. Oncogene 21:3909–3916.

Sommerer F, Hengge UR, MarkwarthA, et al. 2005. Mutations of BRAFand RAS are rare events in germ celltumours. Int J Cancer 113:329–335.

Sonke GS, Chang S, Strom SS, et al.2007. Prenatal and perinatal risk fac-tors and testicular cancer: a hospi-tal-based case-control study. OncolRes 16:383–387.

Sonne SB, Kristensen DM, NovotnyGW, et al. 2008. Testicular dysgene-sis syndrome and the origin of carci-noma in situ testis. Int J Androl31:275–287.

Stevens LC. 1970. Experimental pro-duction of testicular teratomas inmice of strains 129, A/He, and theirF1 hybrids. J Natl Cancer Inst44:923–929.

Stevens LC, Little CC. 1954. Spontane-ous testicular teratomas in an inbredstrain of mice. Proc Natl Acad SciUSA 40:1080–1087.

Stoop H, Honecker F, Cools M, et al.2005. Differentiation and developmentof human female germ cells during pre-natal gonadogenesis: an immunohisto-chemical study. Hum Reprod 20:1466-1476.

Stoop H, Honecker F, van de Geijn GJ,et al. 2008. Stem cell factor as a noveldiagnostic marker for early malignantgerm cells. J Pathol 216:43–54.

Strohmeyer T, Reese D, Press M, et al.1995. Expression of the c-kit proto-oncogene and its ligand stem cell factor(SCF) in normal and malignant humantesticular tissue. J Urol 153:511–515.

Swerdlow AJ, dos Santos Silva I, Reid A,et al. 1998. Trends in cancer incidenceand mortality in Scotland: descriptionand possible explanations. Br J Cancer77 (Suppl 3):1–54.

Tanaka TS, Kunath T, Kimber WL,et al. 2002. Gene expression profilingof embryo-derived stem cells revealscandidate genes associated with plu-ripotency and lineage specificity. Ge-nome Res 12:1921–1928.

Tate G, Suzuki T, Kishimoto K, MitsuyaT. 2005. A c-KIT codon 816 muta-tion, D816H, in the testicular germ

cell tumor: case report of a Japanesepatient with bilateral testicular semi-nomas. Acta Med Okayama 59:33–36.

Tay Y, Zhang J, Thomson AM, et al.2008. MicroRNAs to Nanog, Oct4 andSox2 coding regions modulate em-bryonic stem cell differentiation. Na-ture 455:1124–1128.

Teshima S, Shimosato Y, Hirohashi S,et al. 1988. Four new human germcell tumor cell lines. Lab Invest59:328–336.

Tewari K, Cappuccini F, Disaia PJ, et al.2000. Malignant germ cell tumors ofthe ovary. Obstet Gynecol 95:128–133.

Tian Q, Frierson HF Jr, Krystal GW, Mos-kaluk CA. 1999. Activating c-kit genemutations in human germ cell tumors.Am J Pathol 154:1643–1647.

van Casteren NJ, de Jong J, Stoop H,et al. 2008a. Evaluation of testicularbiopsies for carcinoma in situ: immu-nohistochemistry is mandatory. Int JAndrol in press.

van Casteren NJ, Stoop H, Dohle GR,et al. 2008b. Noninvasive detectionof testicular carcinoma in situ insemen using OCT3/4. Eur Urol 54:153–158.

van Casteren N, Looijenga LHJ, DohleGR. Testicular microlithiasis and car-cinoma in situ: review and proposedclinical guideline. Int J Androl (inpress).

van den Belt-Dusebout AW, de Wit R,Gietema JA, et al. 2007. Treatment-specific risks of second malignanciesand cardiovascular disease in 5-yearsurvivors of testicular cancer. J ClinOncol 25:4370–4378.

van Gurp RJ, Oosterhuis JW, Kalsche-uer V, et al. 1994. Biallelic ex-pression of the H19 and IGF2genes in human testicular germ celltumors. J Natl Cancer Inst 86:1070–1075.

Velasco A, Riquelme E, Schultz M,et al. 2004. Microsatellite instabilityand loss of heterozygosity have dis-tinct prognostic value for testiculargerm cell tumor recurrence. CancerBiol Ther 3:1152–1158.

Velasco A, Corvalan A, Wistuba II,et al. 2008. Mismatch repair expres-sion in testicular cancer predicts re-currence and survival. Int J Cancer122:1774–1777.

Vogel T, Schmidtke J. 1998. Structureand function of TSPY, the Y-chromo-some gene coding for the "testis-spe-cific protein". Cytogenet Cell Genet80:209–213.

Voorhoeve PM, le Sage C, Schrier M,et al. 2006. A genetic screen impli-cates miRNA-372 and miRNA-373 asoncogenes in testicular germ celltumors. Cell 124:1169–1181.

Walsh TJ, Dall’era MA, Croughan MS,et al. 2007. Prepubertal orchiopexyfor cryptorchidism may be associatedwith lower risk of testicular cancer.J Urol 178:1440–1446.

Walt H, Oosterhuis JW, Stevens LC.1993. Experimental testicular germ celltumorigenesis in mouse strains withand without spontaneous tumours dif-fers from development of germ celltumours of the adult human testis. IntJ Androl 16:267–271.

Wang N, Trend B, Bronson DL, FraleyEE. 1980. Nonrandom abnormalitiesin chromosome 1 in human testicularcancers. Cancer Res 40:796–802.

Wikstrom AM, Hoei-Hansen CE, DunkelL, Rajpert-De Meyts E. 2006. Immu-noexpression of androgen receptorand nine markers of maturation inthe testes of adolescent boys withKlinefelter syndrome: evidence fordegeneration of germ cells at theonset of meiosis. J Clin EndocrinolMetab 92:714–719.

Willmore-Payne C, Holden JA, Chad-wick BE, Layfield LJ. 2006. Detectionof c-kit exons 11- and 17-activatingmutations in testicular seminomas byhigh-resolution melting ampliconanalysis. Mod Pathol 19:1164–1169.

Witschi E. 1948. Migration of the germcells of the human embryos from theyolk sac to the primitive gonadalfolds. Contrib Embryol 209:67–80.

Woodward PJ, Heidenreich A, Looi-jenga LHJ, et al. 2004. Tumours ofthe testis and paratesticular tissue.In: Eble JN, Sauter G, Epstein JI,Sesterhenn IA, editors. World HealthOrganization classification oftumours: Pathology and genetics oftumours of the urinary system andmale genital organs. Lyon: IARCPress. pp. 217–249.

Yamaguchi S, Kimura H, Tada M, et al.2005. Nanog expression in mousegerm cell development. Gene ExprPatterns 5:639–646.

Yan W, Kero J, Huhtaniemi I, ToppariJ. 2000. Stem cell factor functions asa survival factor for mature Leydigcells and a growth factor for precur-sor Leydig cells after ethylene dime-thane sulfonate treatment: implica-tion of a role of the stem cell factor/c-Kit system in Leydig cell. Dev Biol227:169–182.

Yates A, Chambers I. 2005. The homeodo-main protein Nanog and pluripotency inmouse embryonic stem cells. BiochemSoc Trans 33(Part 6):1518–1521.

Youngren KK, Coveney D, Peng X,et al. 2005. The Ter mutation in thedead end gene causes germ cell lossand testicular germ cell tumours. Na-ture 435:360–364.

Yuen ST, Davies H, Chan TL, et al.2002. Similarity of the phenotypicpatterns associated with BRAF andKRAS mutations in colorectal neopla-sia. Cancer Res 62:6451–6455.

Zhang C, Kawakami T, Okada Y, Oka-moto K. 2005. Distinctive epigeneticphenotype of cancer testis antigengenes among seminomatous andnonseminomatous testicular germ-cell tumors. Genes ChromosomesCancer 43:104–112.



Documents

Recent developments in testicular germ cell tumor research