ORIGINAL ARTICLE

Correspondence:

Jan Wolter Oosterhuis, Department of Pathology,

University Medical Center Rotterdam, PO Box

2040, 3000 CA, Rotterdam, The Netherlands.

E-mail: j.w.oosterhuis@erasmusmc.nl

Keywords:

teratoma, yolk sac tumor, carcinoma in situ testis,

CIS testis, isochromosome 12p, aneuploidy, FISH,

Johnsen score, testis conserving surgery

Received: 21-Jul-2014

Revised: 17-Oct-2014

Accepted: 18-Oct-2014

doi: 10.1111/andr.305

Pediatric germ cell tumorspresenting beyond childhood?

1J. W. Oosterhuis, 1J.A. Stoop, 1M. A. Rijlaarsdam, 1K. Biermann,2V.T.H.B.M. Smit, 1R. Hersmus and 1L. H. J. Looijenga1Department of Pathology, Laboratory for Experimental Patho-Oncology, Erasmus MC CancerInstitute, Erasmus University Medical Center Rotterdam, Rotterdam, and 2Department of Pathology,Leiden University Medical Center, Leiden, The Netherlands

SUMMARYFour cases are reported meeting the criteria of a pediatric (i.e., Type I) testicular germ cell tumor (TGCT), apart from the age of pre-

sentation, which is beyond childhood. The tumors encompass the full spectrum of histologies of pediatric TGCT: teratoma, yolk sac

tumor, and various combinations of the two, and lack intratubular germ cell neoplasia/carcinoma in situ in the adjacent paren-

chyma. The neoplasms are (near)diploid, and lack gain of 12p, typical for seminomas and non-seminomas of the testis of adolescents

and adults (i.e., Type II). It is proposed that these neoplasms are therefore late appearing pediatric (Type I) TGCT. The present report

broadens the concept of earlier reported benign teratomas of the post-pubertal testis to the full spectrum of pediatric TGCT. The pos-

sible wide age range of pediatric TGCT, demonstrated in this study, lends credence to the concept that TGCT should according to

their pathogenesis be classified into the previously proposed types. This classification is clinically relevant, because Type I mature

teratomas are benign tumors, which are candidates for testis conserving surgery, as opposed to Type II mature teratomas, which

have to be treated as Type II (malignant) non-seminomas.

INTRODUCTIONWe have proposed a classification of germ cell tumors (GCT)

based on their pathogenesis, taking into consideration a number

of parameters: clinical presentation, histology, (cyto)genetics,

and imprinting status related to the cell of origin of the neo-

plasm (Oosterhuis & Looijenga, 2005).

According to this classification testicular GCT (TGCT) present-

ing in the testis of adolescents and adults are virtually always

malignant Type II tumors: the seminomas and non-seminomas

that have carcinoma in situ (CIS), also known as intratubular

germ cell neoplasia unclassified, as common precursor. Paren-

chyma adjacent to the tumor virtually always contains CIS, the

histological evidence that the tumor is indeed a Type II GCT.

Chromosomally they are characterized by a total peri-triploid

DNA content and overrepresentation of the short arm of chro-

mosome 12, most often as one or more copies of isochromo-

some 12p [i(12p)]. The testis is the most common localization of

Type II GCTs. In contrast, Type I GCTs are most often localized

in extra-gonadal localizations, and usually occur at the pediatric

age, before age six (Stang et al., 2012). Histologically, the Type I

GCTs are composed of either teratoma or yolk sac tumor or the

combination of the two, but lack CIS. Chromosomally, these

tumors are (near)diploid without overrepresentation of 12p. Tes-

ticular Type I GCTs are rare (Oosterhuis & Looijenga, 2005; Stang

et al., 2012).

In the past, we have encountered several cases of TGCTs with

the histology of either teratoma and/or yolk sac tumor, present-

ing in adolescents and adults and lacking CIS in the adjacent

parenchyma, verified by immunohistochemistry. We have

revised four cases of this specific type, encountered since 2009,

and studied overrepresentation of chromosome 12 and 12p

using fluorescent in situ hybridization (FISH) on representative

tissue sections. This study aims to further define GCTs in the

post-pubertal testis that do not meet the criteria of Type II

GCTs.

MATERIALS ANDMETHODSAll cases were treated elsewhere and submitted for second

opinion of the pathological findings. The patients are males

without signs of abnormal sex differentiation. Patients 1, 2

70 Andrology, 2015, 3, 70–77 © 2014 American Society of Andrology and European Academy of Andrology

ISSN: 2047-2919 ANDROLOGY

and 4 presented with a unilateral, and patient 3 with a bilat-

eral testicular mass. Patient 2 has Down syndrome with men-

tal retardation. The diagnostic workup and treatment of the

patients was carried out according to the most recent Euro-

pean guidelines (Beyer et al., 2013). The most relevant clinical

findings related to the testicular tumor are presented in

Table 1.

Pathological examination

All original H and E stained slides were available, as well as

all immunohistochemical stainings carried out in the referring

pathology laboratories. All formalin-fixed and paraffin-embed-

ded (FFPE) tissue blocks, containing tumor tissue and adjacent

parenchyma, were made available from the four cases for addi-

tional immunohistochemistry and FISH. Number of tissue

blocks per case: case 1: 5 tumor only (T), 7 tumor and testicu-

lar parenchyma (T/P), 9 parenchyma only (P); case 2: 6 T/P, 1

P; case 3: left testis: 1 T, 7 T/P, right testis: 1 T, 5 T/P; case 4: 8

T, 8 T/P. Histological classification of the tumors is based on

the WHO classification (Woodward et al., 2004). Microscopic

examination of the parenchyma adjacent to the tumor was car-

ried out as previously described (Oosterhuis et al., 2003),

including scoring of spermatogenesis (Johnsen, 1970). The

slides were independently scored for CIS by three reviewers

(JWO, JAS, and KB).

Immunohistochemistry

All blocks were stained for OCT3/4 and PLAP to demonstrate

CIS, seminoma and embryonal carcinoma (EC). These are well

established, robust markers for CIS cells, respectively, expressed

in the nucleus and the cell membrane (Looijenga et al., 2003;

Oosterhuis et al., 2011). All blocks containing tumor were addi-

tionally stained for Alpha-Fetoprotein (AFP) and Human Chorio-

nic Gonadotropin (HCG) to show yolk sac and trophoblastic

differentiation, respectively. In addition, in case 2 selected

blocks were stained for CD30, SOX2, SOX17, (de Jong et al.,

2008) and TSPY (Kaprova-Pleskacova et al., 2014) to further

characterize the OCT3/4 positive cells in the tumor, and in case

4 all blocks were stained for desmin to demonstrate smooth

muscle cells.

For immunohistochemistry, 3 lm sections were incubated

with the primary antibody, followed by biotinylated secondary

antibodies for 30 min and a biotinylated streptavidin horse-rad-

ish peroxidase or alkaline phosphatase coupled complex. The

antibodies and conditions used are indicated in Table 2. All sta-

inings were counterstained with hematoxylin.

Fluorescent in situ hybridization

Slides of 5 lm thickness were deparaffinized and heated

under pressure of up to 0.9 bar in 0.01 M SodiumCitrate pH 6.0.

Slides were digested using 0.01% pepsin (Sigma Aldrich, St.

Louis, MO, USA) in 0.02 M HCl at 37 °C, with an optimal diges-

tion time of 2 min. Slides were rinsed, dehydrated, and the

probes dissolved in hybridization mixture were applied. Probes

for chromosome 12 centromere (pa12H8) (Looijenga et al.,

1990) and chromosome 12p (BAC 876C13) were used, labeled

with digoxigenin-11-dUTP and biotin-16-dUTP (Roche Diagnos-

tics, Mannheim, Germany) using a nick-translation kit (Gibco

BRL, Paisley, UK). After denaturation (80 °C for 10 min), hybrid-

ization for 48 h (37 °C) and washing steps, probes were

visualized using Cy3-conjugated avidin (1 : 100; Jackson Immu-

noResearch, West Grove, PA, USA) and Sheep-anti-dig FITC

(1 : 50; Roche Diagnostics) and analyzed using a LSM700 fluo-

rescent microscope (Zeiss, Sliedrecht, The Netherlands).

RESULTSThe results of histological examination, including immunohis-

tochemistry and FISH for the centromere and the short arm of

chromosome 12 are summarized in Table 3, illustrated in the

Figures, and presented per case hereunder.

Table 1 Clinical data

Case Age Clinical presentation Serum AFP and HCG Treatment Follow up

1 16 Tumor in left testis, size 8 cm; stage 1, low risk Negative Orchiectomy NED at 64 months

2 23 Down syndrome; tumor in inguinal right testis,

size 6 cm; stage 1, low risk

AFP elevated 205 U/L

HCG negative

Orchiectomy and

wait and see

NED at 55 months

3 19 Tumor in right testis, size 2 cm; tumor in left testis,

size 1.5 cm; not staged

Negative Bilateral orchiectomy NED at 17 months

4 23 Tumor in right testis, size 7 cm; stage 1, low risk Negative Orchiectomy NED at 7 months

Table 2 Antibodies (source) and detection method used for immunohistochemistry

Antibody for Company Clone, code Pre-

treatment

Dilutions, incubation time

and temperature

Secondary antibody (1 : 200)

(biotinylated)

Visualization

OCT3/4 Santa Cruz c-10, sc-5279 HIAR 1 : 250, overnight 4 °C Rabbit anti mouse (dako E413) ABCplx-hrp2 Vector PK6100

PLAP Dako 8A9

M7191

HIAR 1 : 100, 32 min, 37 °C Amplification Kit

Ventana 760-080

Ultra View Universal DAB

detection kit

Ventana 760-500

CD30 Dako Ber-H2, M0832 HIAR 1 : 250, overnight 4 °C Rabbit anti mouse (dako E413) ABCplx-hrp2 Vector PK6100

SOX2 R & D 245610, AF2018 HIAR 1 : 250, overnight 4 °C Horse anti goat (Vector BA9500) ABCplx-AP Vector AK5000

SOX17 Neuromics Gt15094 HIAR 1 : 1500, overnight 4 °C Horse anti goat (Vector BA9500) ABCplx-AP Vector AK5000

TSPY Gift from C. Lau NA HIAR 1 : 3000, overnight 4 °C Swine anti rabbit (Dako E431) ABCplx-AP2 Vector AK5000

AFP Dako A008 None 1 : 600, overnight 4 °C Swine anti rabbit (Dako E431) ABCplx-AP2 Vector AK5000

HCG Dako A231 None 1 : 10 000, overnight 4 °C Swine anti rabbit (Dako E431) ABCplx-AP2 Vector AK5000

Desmin Monosan D33, mon3001 HIAR 1 : 40, 30 min, room

temperature

Rabbit anti mouse (dako E413) ABCplx-AP2 Vector PK6100

© 2014 American Society of Andrology and European Academy of Andrology Andrology, 2015, 3, 70–77 71

PEDIATRIC GERM CELL TUMORS ANDROLOGY

Case 1

A 16-year-old, adolescent man presenting with a mass in his

left testis. Serum tumor markers AFP and HCG were normal. The

patient underwent orchiectomy. On gross examination the testis

contained a circumscribed tumor with a largest dimension of

8 cm with multiple cysts on cut surface. It bordered on testicular

parenchyma and was confined to the testis. Histologically the

tumor consisted of mature teratoma only with cystic spaces of

varying dimensions, lined by various types of epithelium, most

often intestinal types of epithelium with among others goblet

cells, and less frequently respiratory and squamous epithelium.

Focally there was cyto-nuclear atypia to the degree of mild dys-

plasia; mitotic figures were rare. The wall of the cysts often con-

tained concentric layers of smooth muscle cells. Throughout the

tumor lymphocytic infiltrates were present. The tumor was im-

munohistochemically negative for OCT3/4, AFP and HCG. The

seminiferous tubules showed impaired spermatogenesis (John-

sen score about 4, with occasional tubules with round

spermatids) with quite frequently spermatogonia with enlarged

and hyperchromatic or multiple nuclei. There was no CIS, con-

firmed by the negative results of the staining for OCT3/4 and

PLAP. The parenchyma was devoid of inflammatory cells. There

was no scar as can be seen in the case of a burnt out GCT. The

number of Leydig cells was not increased. FISH showed two cen-

tromeres of chromosomes 12 in the tumor cells, and no excess

of 12p. The patient was staged as stage 1, low risk. No further

treatment was given. At 64 months follow up there is no evi-

dence of disease (NED) (Fig. 1).

Case 2

A 23-year-old man with Down syndrome presenting with a

mass in an inguinal testis in his right groin. Serum AFP was ele-

vated (205 U/L), serum HCG was normal. The inguinal right tes-

tis was resected. On gross examination the testis contained a

tumor with a largest dimension of 6 cm. On cut surface it was

circumscribed, solid and glistening with necrotic areas. The

Table 3 Histological examination, immunohistochemistry, and in situ hybridization

Case Histology IHC OCT3/4 IHC AFP ISH ploidy ISH excess 12p

1 Mature teratoma; no CIS; Johnsen score about 4 Negative Negative Diploid Negative

2 Mainly YST; immature teratoma; no CIS;

Johnsen score about 3

Few positive stem cells Positive Mainly diploid, focal excess #12 Negative

3 Bilateral multicystic mature teratoma; no CIS;

Johnsen score about 2

Negative Negative Diploid Negative

4 >99% mature teratoma; microscopic focus of YST;

no CIS; Johnsen score about 2

Negative Focally positive Mainly diploid, focal excess #12 Negative

(A) (B)

(C) (D)

Figure 1 Case 1. (A) Mature teratoma, the only component of the tumor; shown: cystic space lined by intestinal epithelium with focally mild dysplasia

(upper right corner); cyst wall contains smooth muscle cells (pink strands in the lower left corner). HE, 509. (B) Testis parenchyma with seminiferous

tubules; blocked spermatogenesis with predominantly spermatogonia, and some spermatocytes present; no lymphocytic infiltrates. HE, 1009. (C) Seminif-

erous tubules with spermatogonia with enlarged, hyperchromatic nuclei, and multiple nuclei, not resembling CIS. HE, 2009. (D) Tumor, FISH, 1009. Red

indicates the centromere of chromosome 12, green indicates the short arm of chromosome 12. The overall pattern is diploid with two copies of chromo-

some 12, and no overrepresentation of 12p. The green signal of 12p is rather weak. 9 indicates the magnification, HE indicates hematoxylin–eosin staining.

72 Andrology, 2015, 3, 70–77 © 2014 American Society of Andrology and European Academy of Andrology

J. W. Oosterhuis et al. ANDROLOGY

tumor was partly surrounded by testicular parenchyma and con-

fined to the testis. Histological examination confirmed the pres-

ence of necrosis. The morphologically intact tumor tissue

consisted mainly of YST, positive for AFP as detected immuno-

histochemically. The YST was mostly of the enteric pattern. The

immature teratoma component was highly cellular with areas of

neuro-endocrine differentiation. There was ‘organoid’ pattern-

ing and no invasive growth of this component, making a small

blue round cell-secondary non-germ cell malignancy unlikely. In

the immature teratoma component sporadically small groups of

cells with nuclear expression of OCT3/4 were identified. Neither

by morphology, nor by antigen expression, these cells met the

criteria of EC cells, the stem cell component of Type II non-

seminomas. In fact, they lacked expression of CD30 and SOX2,

and expressed SOX17 in the nucleus, and PLAP in a membra-

nous fashion. Although SOX2 was expressed in some cells of the

immature teratoma component, these were not the cells

expressing OCT3/4. CD30 was entirely negative in the tumor.

SOX17 was rather abundant in the YS-component. PLAP was

only expressed in the OCT3/4 positive cells in the immature ter-

atoma component. Throughout the tumor dense infiltrates of

lymphocytes and plasma cells were identified. These spilled over

into the parenchyma, most often in a perivascular arrangement

and also involving some seminiferous tubules. No scar of a burnt

out GCT was found. With a Johnsen score about 3, spermatogen-

esis was severely impaired, spermatocytes being the highest

degree of maturation. There was no CIS, confirmed by the nega-

tive result of the staining of the parenchyma for OCT3/4 and

PLAP. The number of Leydig cells was not increased. The tumor

was also microscopically confined to the testis, and bordering

the rete testis without infiltrating it. Angio-invasive growth could

not be demonstrated. FISH demonstrated two copies of chromo-

some 12 in most tumor cells, while a few cells showed extra cop-

ies of chromosome 12, however, excess of 12p (as

isochromosome 12p or otherwise) could not be demonstrated.

The patient was staged as stage 1, low risk. No further treatment

was given. There is no evidence of disease after 55 months fol-

low up (Fig. 2).

Case 3

A 19-year-old, adolescent man presented with a mass in his

left and right testis. Serum tumor markers AFP and HCG were

normal. He underwent bilateral orchiectomy. On gross exami-

nation the right testis contained a circumscribed tumor with a

largest dimension of 2 cm. The tumor in the left testis was

1.5 cm largest dimension. On cut surface the tumors appeared

as multilocular cysts filled with whitish keratinous material.

They were surrounded by testicular parenchyma, and confined

to the testis. Histologically the cysts were lined by epidermoid

epithelium with small patches of different types of epithelium:

for example, intestinal type epithelium with goblet cells. There

was no cyto-nuclear atypia and no mitotic activity. The content

of the cysts consisted of squamous material. In both testes the

parenchyma showed severe tubular fibrosis and atrophy (John-

sen score about 2). However, on both sides there were also spo-

radic tubules with completely mature spermatogenesis. There

was no CIS, confirmed by the negative result of the staining for

OCT3/4 and PLAP. Occasional microliths and one or two

microscopic Sertoli cell only nodules were found on both sides.

There was extensive nodular Leydig cell hyperplasia and

testicular angiopathy (Oosterhuis et al., 2003). The parenchyma

was devoid of lymphocytes, and there were no scars. FISH dem-

onstrated two copies of chromosome 12 in the tumor cells, and

no excess of 12p. The patient was not clinically staged. He got

androgen substitution, and had no further tumor related treat-

ment. There is no evidence of disease after 17 months follow

up (Fig. 3).

Case 4

A 23-year-old man presented with a mass in his right testis.

Serum tumor markers AFP and HCG were normal. The patient

underwent orchiectomy. On gross examination the testis con-

tained a circumscribed tumor confined to the testis with a larg-

est dimension of 7 cm with on cut surface multiple small,

glistening cysts, and areas of hemorrhage and necrosis. There

was a rim of testicular parenchyma left. Histology confirmed

hemorrhage and necrosis. The intact tumor tissue consisted of

mature teratoma only, with cystic spaces of varying dimensions,

partly lined by clear cuboidal cells containing glycogen, partly by

taller columnar cells containing mucin. The walls of some of the

cysts contained layers of smooth muscle cells, staining positive

for desmin. Focally there was some cyto-nuclear atypia and

mitotic activity in the epithelium, to the degree of mild dyspla-

sia. AFP expression was found in one small cyst, consistent with

microscopic YST. The seminiferous tubules showed severely

impaired spermatogenesis (Johnsen score about 2, and sperma-

tocytes being the highest degree of maturation), and lacked CIS,

confirmed by the negative result of the staining for OCT3/4 and

PLAP. There were sporadic lymphocytes in the parenchyma, not

infiltrating the tubules. Scar as in a burnt out GCT was not

detected. ISH demonstrated two copies of chromosome 12 in

virtually all tumor cells. Sporadic cells showed extra copies of

chromosome 12, however, no excess of 12p was identified. The

patient was staged as stage 1, low risk. No further treatment was

given. There is no evidence of disease after 7 months follow up

(Fig. 4).

DISCUSSIONOn first impression one is inclined to classify these four

tumors as (malignant, Type II) non-seminomas of the post-

pubertal testis, based on clinical presentation and histology:

two as pure teratomas, and two as mixed non-seminomas with

different combinations of teratoma and YST. What is highly

unusual is the absence of CIS in the concomitant parenchyma.

In our experience, when the orchiectomy specimen is ade-

quately sampled, as is the case in the present four cases

(between 5 and 16 tissue blocks containing parenchyma per

orchiectomy specimen), it will be found in all cases of non-

seminoma, when it is looked for with the support of immuno-

histochemistry (Oosterhuis et al., 2003; Stoop et al., 2011). In

seminoma, CIS may be lacking because of the lymphocytic host

response directed against the tumor and the concomitant CIS,

which may result in the scar of a burnt out tumor and com-

pletely fibrosed seminiferous tubules, invariably associated with

lymphocytic infiltrates. The likelihood of not finding CIS

increases with age, suggesting that the duration of the host

response is an important factor. In patients with a GCT com-

bining seminoma and non-seminoma, who are younger than

patients with a pure seminoma (i.e., median age 30 as opposed

to 35 years), the tubules with CIS are less affected (Oosterhuis

© 2014 American Society of Andrology and European Academy of Andrology Andrology, 2015, 3, 70–77 73

PEDIATRIC GERM CELL TUMORS ANDROLOGY

et al., 2003). This in spite of the presence of a similar host

response in the parenchyma of pure seminomas and combined

tumors (Oosterhuis et al., 2003). The parenchyma in the present

cases does show pathological changes, however, different from

those associated with regression of CIS in seminomas and com-

bined tumors. Fibrosis and lymphocytes are virtually lacking in

the cases 1, 3, and 4. Only in case 2 there was an inflammatory

infiltrate in the parenchyma, but with a highly unusual compo-

sition (predominantly plasma cells). The few tubules affected

by these infiltrates, although still patent and containing germ

cells, lacked CIS. In all cases spermatogenesis was rather uni-

formly impaired, whereas in parenchyma adjacent to non-

(A) (B)

(C) (D)

(E) (F)

(G) (H)

Figure 2 Case 2. (A) Yolk sac tumor component mainly of the enteric type. HE, 509. (B) Immature teratoma component, with neuro-endocrine areas. HE,

2009. (C) Immature teratoma, approximately the same area as in B with nuclear expression of OCT3/4 in some cells. OCT3/4 immunohistochemical stain-

ing, 2009. (D) Immature teratoma, same area as in C, negative for CD30. CD30 immunohistochemical staining, 2009. (E) Immature teratoma, same area

as in C. SOX2 is expressed in cells different cells from those expressing OCT3/4. SOX2 immunohistochemical staining, 4009. (F) Testis parenchyma, semi-

niferous tubules with impaired spermatogenesis with few spermatocytes as highest degree of maturation. HE, 1009. (G) Lymphocytes and plasma cells infil-

trating a seminiferous tubule lacking CIS; similar infiltrates as in the tumor (not shown). HE, 1009. (H) Tumor, FISH, 639. Red indicates the centromere of

chromosome 12, green indicates the short arm of chromosome 12. The overall pattern is diploid with two copies of chromosome 12, and no overrepresen-

tation of 12p. 9 indicates the magnification, HE indicates hematoxylin–eosin staining.

74 Andrology, 2015, 3, 70–77 © 2014 American Society of Andrology and European Academy of Andrology

J. W. Oosterhuis et al. ANDROLOGY

seminomas there typically is a dichotomy between tubules

affected by CIS, lacking lymphocytic infiltration, and tubules

with well-developed spermatogenesis.

The character of the parenchymal changes and the absence

of CIS in these four adequately sampled cases made us doubt

the diagnosis of non-seminoma of the post-pubertal testis, or,

(A) (B)

(C) (D)

Figure 3 Case 3. The histology of tumors and parenchyma is identical in the right and left testis. (A) Mature teratoma, the only component of the tumor,

mainly consisting of cysts lined by keratinizing squamous epithelium with patches of epithelium with goblet cells. HE, 509. (B) Detail of epithelium with

goblet cells, HE, 4009. (C) Testis parenchyma with atrophic tubules and rare tubules with spermatogenesis; nodular Leydig cell hyperplasia, HE, 509. (D)

Tumor, FISH, 639. Red indicates the centromere of chromosome 12, green indicates the short arm of chromosome 12. The overall pattern is diploid with

two copies of chromosome 12, and no overrepresentation of 12p. 9 indicates the magnification, HE indicates hematoxylin–eosin staining.

(A) (B)

(C) (D)

Figure 4 Case 4. (A) Tumor consisting of mature teratoma only; cysts lined by cuboidal clear cells, HE, 509. (B) Testis parenchyma; tubular atrophy, HE,

509. (C) Focus of YST, AFP, 4009. (D) Tumor, FISH, 639. Red indicates the centromere of chromosome 12, green indicates the short arm of chromosome

12. The overall pattern is diploid with two copies of chromosome 12, and no overrepresentation of 12p. 9 indicates the magnification, HE indicates hema-

toxylin–eosin staining.

© 2014 American Society of Andrology and European Academy of Andrology Andrology, 2015, 3, 70–77 75

PEDIATRIC GERM CELL TUMORS ANDROLOGY

as we classify it, a Type II non-seminoma of which CIS is the

precursor lesion (Skakkebæk, 1972). We then realized that the

histology of the tumors is compatible with pediatric GCTs: tera-

toma, YST, or the combination of the two. Seminoma, EC, and

trophoblastic differentiation (i.e., choriocarcinoma or tropho-

blastic giant cells) are not encountered in pediatric GCTs. The

OCT3/4-postive cells in the immature teratoma component of

case 2 do not meet the criteria of EC cells, as they did not, in

addition to OCT3/4, express CD30 and SOX2, but expressed

SOX17 in the nucleus, and PLAP in a membranous fashion.

The negativity of these cells for TSPY excludes germ cell differ-

entiation in the tumor (Honecker et al., 2006) (data not shown).

We have encountered similar cells in immature teratomas of

the ovary (unpublished observations), possibly representing

(committed) stem cells in this tumor type. SOX2, which has a

wider distribution that just EC cells (de Jong et al., 2008) was

focally expressed in other cells in the immature teratoma

component.

In summary, the histology of the tumors and concomitant

parenchyma is compatible with pediatric GCTs, or, as we classify

it, Type I GCTs of the testis. This prompted us to study the ploidy

and number of copies of 12 and 12p using FISH. The results are

supportive of our histological diagnosis. In general there were

two copies of chromosome 12 per tumor cell and there was no

excess of 12p demonstrated. Excess of 12p is a consistent finding

in all tumor cells of Type II GCTs. We conclude that except for

the age of the patient these tumors meet the criteria of a Type I

GCT, and are therefore best classified as late presenting Type I

GCT.

The two pure teratomas presented here are different from the

benign teratomas of the post-pubertal testis recently published

by Zhang et al. (2013) and Semjen et al. (2014). Those teratomas

lack atypia and mitotic activity, and the parenchyma, apart from

a rim bordering on the tumor, contains tubules with normal

spermatogenesis. Regarding our cases in this respect: the exten-

sive changes in the parenchyma of our cases 1, 2, and 4 may be

because of compression by the large sized tumors, and in case 2

also because of the inguinal position of the testis. The atrophy of

the parenchyma in case 3 is unusual. The bilateral presentation

and histopathology are compatible with late changes in severe

mumps orchitis. This condition is not associated with testicular

tumors (Davis et al., 2010). Zhang et al. (2013) suggest that the

benign teratomas and dermoid cysts of the post-pubertal testis

could in fact be late appearing pediatric teratomas. Along the

same line of reasoning we suggest that epidermoid cysts of the

testis, if they lack CIS in the adjacent parenchyma, might repre-

sent pediatric monophyletic teratomas.

The present report broadens the concept of the benign terato-

mas of the post-pubertal testis (Zhang et al., 2013) to the full

spectrum pediatric TGCT. The wide age range of pediatric type

GCT, demonstrated in this study, lends credence to the concept

that TGCT should be classified according to their pathogenesis

into Type I (teratoma and/or yolk sac tumor, lacking 12p rear-

rangements), Type II (seminoma and non-seminoma character-

ized by 12p rearrangements), and Type III (spermatocytic

seminoma, characterized by gain of chromosome 9) [not dis-

cussed here (Looijenga et al., 2006)], rather than based on age of

presentation.

This classification is clinically relevant, because Type I mature

teratomas are benign tumors, which need no systemic treatment

and follow up, and are good candidates for testis conserving sur-

gery, as opposed to Type II mature teratomas, which have to be

treated as Type II non-seminomas. Testis sparing surgery would

have led to a better outcome in particular for case 3, preventing

lifelong dependence on androgen substitution. From the practi-

cal point of view: applying direct staining for alkaline phospha-

tase on a frozen section of parenchyma clearly demonstrates CIS

and allows the distinction between a Type II and a Type I GCT

during surgery (Stoop et al., 2011). When no CIS is demonstrated

the treatment of choice is enucleation of the lesion followed by a

frozen section of it to exclude a tumor that necessitates

orchiectomy.

Thus, Type I tumors also in adults may display the full spec-

trum of histological types of embryo-derived tumors in the

mouse, which, in our view, are a model of Type I GCTs (van

Berlo et al., 1990; Oosterhuis et al., 1993; Oosterhuis & Looij-

enga, 2005; Looijenga & Oosterhuis, 2013). Like their murine

counterparts, Type I tumors of the testis probably originate

from a germ cell precursor that is directly reprogrammed to

pluripotency, without an intermediate stage of neoplastic pri-

mordial germ cells (i.e., CIS), as is the case in Type II GCTs. In

this respect the observation of frequent atypical spermatogonia

in case 1 is intriguing. These atypical cells differ from CIS cells

in that they have a very heterogeneous morphology, and do

not express OCT3/4 or PLAP. Morphologically similar germ

cells were noticed in the testis of 129-mice developing terato-

mas (Walt et al., 1993). In this context the observation of the

role of dmrt1 in the pathogenesis of mouse teratomas is at

least intriguing, although full understanding will require addi-

tional studies (Krentz et al., 2009). It is conceivable that late

presenting Type I GCTs originate in the neonatal period, like

their infantile counterparts, and only become clinically mani-

fest at adolescent or adult age. However, it cannot be excluded

that these tumors originate by a similar pathogenetic mecha-

nism at a later age. Finally, the microscopic characteristics of

testicular dysgenesis (Skakkebaek et al., 2001) in both testicles

in case 3, assuming that they are not because of reactive

changes in the context of mumps orchitis, might indicate over-

lap in the pathogenesis of Type II and Type I GCTs of the

post-pubertal testis. Testicular dysgenesis might, perhaps, not

only predispose for delayed maturation and neoplastic trans-

formation of gonocytes during embryonal development, but

also for disturbance of maturation and reprogramming of other

germ cell precursors at a later age, even in the post-pubertal

testis.

ACKNOWLEDGMENTSProf. Dr. Albert J. H. Suurmeijer is kindly acknowledged for

supplying case three. Prof. Dr. Chris Lau kindly supplied the

TSPY antibody. MAR is supported by a Translational Grant, Eras-

mus MC. The funders had no role in study design, data collec-

tion and analysis, decision to publish, or preparation of the

manuscript.

AUTHOR CONTRIBUTIONSJWO, JAS, RH, and LHJL conceived and designed the experi-

ments. JAS and RH performed the experiments. JWO, LHJL, JAS,

RH, and KB analyzed the data. VTHBM and MAR contributed

reagents/materials/analysis tools. JWO, JAS, MAR, VTHBM, RH,

and LHJL contributed to the writing of the manuscript.

76 Andrology, 2015, 3, 70–77 © 2014 American Society of Andrology and European Academy of Andrology

J. W. Oosterhuis et al. ANDROLOGY

REFERENCESvan Berlo RJ, Oosterhuis JW, Schrijnemakers E, Schoots CJ, de Jong B &

Damjanov I. (1990) Yolk-sac carcinoma develops spontaneously as a

late occurrence in slow-growing teratoid tumors produced from

transplanted 7-day mouse embryos. Int J Cancer 45, 153–155.

Beyer J, Albers P, Altena R, Aparicio J, Bokemeyer C, Busch J, Cathomas

R, Cavallin-Stahl E, Clarke NW, Classen J, Cohn-Cedermark G, Dahl

AA, Daugaard G, De Giorgi U, De Santis M, De Wit M, De Wit R,

Dieckmann KP, Fenner M, Fizazi K, Flechon A, Fossa SD, Germa Lluch

JR, Gietema JA, Gillessen S, Giwercman A, Hartmann JT, Heidenreich

A, Hentrich M, Honecker F, Horwich A, Huddart RA, Kliesch S,

Kollmannsberger C, Krege S, Laguna MP, Looijenga LH, Lorch A, Lotz

JP, Mayer F, Necchi A, Nicolai N, Nuver J, Oechsle K, Oldenburg J,

Oosterhuis JW, Powles T, Rajpert-De Meyts E, Rick O, Rosti G, Salvioni

R, Schrader M, Schweyer S, Sedlmayer F, Sohaib A, Souchon R,

Tandstad T, Winter C & Wittekind C. (2013) Maintaining success,

reducing treatment burden, focusing on survivorship: highlights from

the third European consensus conference on diagnosis and treatment

of germ-cell cancer. Ann Oncol 24, 878–888.

Davis NF, McGuire BB, Mahon JA, Smyth AE, O’Malley KJ & Fitzpatrick

JM. (2010) The increasing incidence of mumps orchitis: a

comprehensive review. BJU Int 105, 1060–1065.

Honecker F, Stoop H, Mayer F, Bokemeyer C, Castrillon DH, Chris Lau

YF, Looijenga LHJ & Oosterhuis JW. (2006) Germ cell lineage

differentiation in nonseminomatous germ cell tumors. J Pathol 208,

395–400.

Johnsen SG. (1970) Testicular biopsy score count – a method for

registration of spermatogenesis in human testes: normal values and

results in 335 hypogonadal males. Hormones 1, 2–25.

de Jong J, Stoop H, Gillis AJ, van Gurp RJ, van de Geijn GJ, Boer M,

Hersmus R, Saunders PT, Anderson RA, Oosterhuis JW & Looijenga

LH. (2008) Differential expression of SOX17 and SOX2 in germ cells

and stem cells has biological and clinical implications. J Pathol 215,

21–30.

Kaprova-Pleskacova J, Stoop H, Bruggenwirth H, Cools M, Wolffenbuttel

KP, Drop SL, Snajderova M, Lebl J, Oosterhuis JW & Looijenga LH.

(2013) Complete androgen insensitivity syndrome: factors influencing

gonadal histology including germ cell pathology. Mod Pathol 27,

721–730.

Krentz AD, Murphy MW, Kim S, Cook MS, Capel B, Zhu R, Matin A,

Sarver AL, Parker KL, Griswold MD, Looijenga LH, Bardwell VJ &

Zarkower D. (2009) The DM domain protein DMRT1 is a

dose-sensitive regulator of fetal germ cell proliferation and

pluripotency. Proc Natl Acad Sci USA 106, 22323–22328.

Looijenga LH & Oosterhuis JW. (2013) Pathobiology of germ cell tumors –

applying the gossip test! Int J Dev Biol 57, 289–298.

Looijenga LH, Smit VT, Wessels JW, Mollevanger P, Oosterhuis JW,

Cornelisse CJ & Devilee P. (1990) Localization and polymorphism of a

chromosome 12-specific alpha satellite DNA sequence. Cytogenet Cell

Genet 53, 216–218.

Looijenga LH, Stoop H, de Leeuw HP, de Gouveia Brazao CA, Gillis AJ,

van Roozendaal KE, van Zoelen EJ, Weber RF, Wolffenbuttel KP, van

Dekken H, Honecker F, Bokemeyer C, Perlman EJ, Schneider DT,

Kononen J, Sauter G & Oosterhuis JW. (2003) POU5F1 (OCT3/4)

identifies cells with pluripotent potential in human germ cell tumors.

Cancer Res 63, 2244–2250.

Looijenga LH, Hersmus R, Gillis AJ, Pfundt R, Stoop HJ, van Gurp RJ,

Veltman J, Beverloo HB, van Drunen E, van Kessel AG, Pera RR,

Schneider DT, Summersgill B, Shipley J, McIntyre A, van der Spek P,

Schoenmakers E & Oosterhuis JW. (2006) Genomic and expression

profiling of human spermatocytic seminomas: primary spermatocyte

as tumorigenic precursor and DMRT1 as candidate chromosome 9

gene. Cancer Res 66, 290–302.

Oosterhuis JW & Looijenga LH. (2005) Testicular germ-cell tumours in a

broader perspective. Nat Rev Cancer 5, 210–222.

Oosterhuis JW, Van Berlo RJ, De Jong B, Dam A, Buist J, Tamminga RYJ &

Zwierstra RP. (1993) Sacral teratoma with late recurrence of yolk sac

tumor: human counterpart of embryo- or yolk sac-derived teratoma? J

Urol Pathol 1, 257–267.

Oosterhuis JW, Kersemaekers AM, Jacobsen GK, Timmer A, Steyerberg

EW, Molier M, Van Weeren PC, Stoop H & Looijenga LH. (2003)

Morphology of testicular parenchyma adjacent to germ cell tumours.

An interim report. APMIS 111, 32–40; discussion 41-32.

Oosterhuis JW, Stoop H, Dohle G, Boellaard W, van Casteren N,

Wolffenbuttel K & Looijenga LH. (2011) A pathologist’s view on the

testis biopsy. Int J Androl 34, e14–e19; discussion e20.

Semjen D, Kalman E, Tornoczky T & Szuhai K. (2014) Further evidence of

the existence of benign teratomas of the postpubertal testis. Am J Surg

Pathol 38, 580–581.

Skakkebæk NE. (1972) Possible carcinoma-in-situ of the testis. Lancet 2,

516–517.

Skakkebaek NE, Rajpert-De Meyts E & Main KM. (2001) Testicular

dysgenesis syndrome: an increasingly common developmental

disorder with environmental aspects. Hum Reprod 16, 972–978.

Stang A, Trabert B, Wentzensen N, Cook MB, Rusner C, Oosterhuis JW &

McGlynn KA. (2012) Gonadal and extragonadal germ cell tumours in

the United States, 1973–2007. Int J Androl 35, 616–625.

Stoop H, Kirkels W, Dohle GR, Gillis AJ, den Bakker MA, Biermann

K, Oosterhuis W & Looijenga LH. (2011) Diagnosis of testicular

carcinoma in situ ‘(intratubular and microinvasive)’ seminoma and

embryonal carcinoma using direct enzymatic alkaline phosphatase

reactivity on frozen histological sections. Histopathology 58, 440–

446.

Walt H, Oosterhuis JW & Stevens LC. (1993) Experimental testicular germ

cell tumorigenesis in mouse strains with and without spontaneous

tumours differs from development of germ cell tumours of the adult

human testis. Int J Androl 16, 267–271.

Woodward PJ, Heidenreich A, Looijenga LHJ, Oosterhuis JW, McLeod DG

& Moller H. (2004) Testicular germ cell tumors. In: World Health

Organization Classification of Tumours. Pathology and Genetics of the

Urinary System and Male Genital Organs (eds JN Eble, G Sauter, JI

Epstein & IA Sesterhann), pp. 217–278. IARC Press, Lyon.

Zhang C, Berney DM, Hirsch MS, Cheng L & Ulbright TM. (2013)

Evidence supporting the existence of benign teratomas of the

postpubertal testis: a clinical, histopathologic, and molecular genetic

analysis of 25 cases. Am J Surg Pathol 37, 827–835.

© 2014 American Society of Andrology and European Academy of Andrology Andrology, 2015, 3, 70–77 77

PEDIATRIC GERM CELL TUMORS ANDROLOGY