Upload
yukie
View
212
Download
0
Embed Size (px)
Citation preview
Immunohistochemical and UltrastructuralFeatures of Neuroendocrine Differentiated
Carcinomas of the Prostate: AnImmunoelectron Microscopic Study
Daisaku Hirano, MD
Department of Urology, Nihon
University School of Medicine,
Tokyo, Japan
Toyoharu Jike
Electron Microscope Division,
Nihon University School of
Medicine, Tokyo, Japan
Yasuhiro Okada, MD,
Sadatsugu Minei, MD, Shuji
Sugimoto, MD, Kenya
Yamaguchi, MD, Tetsuo
Yoshikawa, MD, Takahiko
Hachiya, MD, Toshio Yoshida,
MD, and Yukie Takimoto, MD
Department of Urology, Nihon
University School of Medicine,
Tokyo, Japan
ABSTRACT The purpose of this study was to further define the immunohis-
tochemical and ultrastructural characteristics of neuroendocrine (NE) differ-
entiated prostatic carcinomas. Seventy-seven specimens were obtained from
prostatic carcinoma tumors during prostatectomy, transurethral resection of
prostate or biopsy in 77 prostate cancer patients, and analyzed by immuno-
histochemical staining for chromogranin A (CgA). Nine of these tumors were
also studied by elctron microscopy and 4 were examined by pre-embedding
immunoelectron microscopy. CgA-stained cells were detected in 36 tumors
(47%). Clinically advanced tumors or tumors with higher histological grades
were associated with increased NE differentiation. Three of the tumors stud-
ied by electron microscopy contained cells showing unequivocal NE differ-
entiation revealed by the presence of neurosecretory granules, while the
poorly NE-differentiated malignant cells contained pleomorphic granules,
which were lysosomal-like rather than NE-type granules. Immunoelectron
microscopy demonstrated the presence of CgA immunoreactivity on the
pleomorphic granules in the poorly differentiated malignant glands. This
study suggests that NE-differentiated malignant cells in prostate cancer
tissues may induce aggressive behavior in adjacent proliferating neoplastic
cells via a paracrine mechanism.
KEYWORDS immunoelectron microscopy, immunohistochemistry, neuroen-
docrine differentiation, prostate cancer, ultrastructure
Small-cell prostate carcinomas or carcinoid-like tumors represent only 1–2%
of all prostate malignancies and are characterized as pure neuroendocrine
(NE) tumors [1]. However, NE-differentiated cells are also detected immuno-
histochemically in the normal prostate, in benign prostatic hyperplasia, and
in common prostatic adenocarcinoma, and approximately 10% of common
prostatic carcinomas have extensive and multifocal NE features [2, 3]. These
NE-differentiated prostatic carcinomas are similar to small cell carcinomas of
Received 16 February 2005;accepted 21 February 2005.
Address correspondence to DaisakuHirano, MD, Department of Urology,Nihon University School of Medicine,30-1 Ooyaguchi Kamimachi Itabashi-ku, Tokyo, 173-88610, Japan. E-mail:[email protected]
Ultrastructural Pathology, 29:367–375, 2005Copyright # Taylor & Francis Inc.ISSN: 0191-3123 print=1521-0758 onlineDOI: 10.1080/019131290945718
367
Ultr
astr
uct P
atho
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
SUN
Y S
tate
Uni
vers
ity o
f N
ew Y
ork
at S
tony
Bro
ok o
n 10
/31/
14Fo
r pe
rson
al u
se o
nly.
the lung, with comparable immunoreactivity and the
presence of neurosecretory granules on electron
microscopy [2, 3], and comprise a highly specialized
cell population containing neurosecretory granules
rich in various peptide hormones and biogenic
amines, such as chromogranin A (CgA) [3], serotonin
[4], bombesin [5], vascular endothelial growth factors
[6], and somatostatin [7]. di Sant’ Agnese and Cockett
[2] suggested that these regulatory peptides, which
may act through lumencrine, endocrine, paracrine,
and autocrine mechanisms, are probably involved
in normal growth, differentiation, and secretory func-
tions of the gland. However, the precise biological
behavior of NE-differentiated carcinomas in the com-
mon prostate cancer tissues is not understood.
Therefore, we have further studied the clinicopatho-
logical features of the NE-differentiated prostatic
carcinomas in the common prostate cancer, by
immunostaining for a specific NE marker, CgA, and
by assessment of their ultrastructural characteristics.
In addition, we used immunoelectron microscopy
to confirm the presence of substantial quantities of
CgA in the neurosecretory granules of NE-differen-
tiated prostatic carcinoma cells.
MATERIALS AND METHODS
Patients and Specimens
Specimens of prostatic carcinoma from 77 patients
were studied, of which 72 were obtained from rad-
ical prostatectomy, 1 during transurethral resection
(TUR) of prostate, and 4 during prostate biopsy.
Patient characteristics are given in Table 1. Clinical
stage and pathological grade of prostate cancer were
definedaccording to theWhitmore-Jewett classification
[8] and the WHO grading system [9], respectively.
All patients gave informed consent that their pros-
tatic specimens from surgical extirpation or biopsy
could be used in this study. All specimens were
evaluated for NE differentiation by immunohisto-
chemical staining for CgA. Nine specimens were
further studied by electron microscopy and 4 speci-
mens were studies by immunoelectron microscopy.
Immunohistochemical Analysis
Each prostate specimen obtained at radical prosta-
tectomy, TUR, or biopsy was fixed in 10% buffered
formalin by conventional methods. After fixation,
the gland obtained from radical prostatectomy was
step-sectioned at 5-mm intervals perpendicular to
the urethra and embedded in paraffin. The most cen-
tral parts of the radical prostatectomy and TUR speci-
mens with representative tumor growth, as indicated
by hematoxylin and eosin staining, were chosen for
immunohistochemistry. Each 4-mm section was
deparaffinized and incubated in 1% H2O2 in meth-
anol for 20min at room temperature to block
endogenous peroxidase activity. The sections were
then washed in phosphate-buffered saline (PBS),
pH 7.4, preincubated with 10% bovine serum albu-
min for 20min at 37�C, and incubated for 2 h at room
temperature with rabbit anti-human CgA antibody
(Dako, Carpinteria, CA, USA) diluted 1:250. After
washes in PBS, the sections were incubated with
avidin–biotin conjugated mouse anti-rabbit immuno-
globulins (Dako) for 30min at room temperature.
After washes in PBS, the section was incubated with
0.03% 3,30-diaminobenzidine tetrahydrochloride
(Wako, Osaka, Japan) in buffer containing hydrogen
peroxide. Nuclei were counterstained with hematox-
ylin, and the sections were dehydrated and cover-
slipped.
Positive controls included NE cells from adjacent
benign prostatic tissues and NE cells from pancreatic
tissues. For negative controls, appropriately diluted
nonimmune serum was substituted for anti-CgA
antibody.
Tumors were evaluated for positive CgA staining
in all neoplastic lesions using a gridded eyepiece at
200�magnification, and scored as follows: 0 ¼ no
immunoreactive tumor cells; þ1 ¼ <10% immunore-
active neoplastic cells; þ2 ¼ 10–20% immunore-
active tumor cells; þ3 ¼>20% immunoreactive
neoplastic cells.
TABLE 1 Patient Characteristics
Median age (range) (years) 69 (50–78)
Clinical stage (%)
T1b, T1c 22 (29)
T2 26 (34)
T3 24 (31)
Distant metastases 5 (6)
Histological grade (%)
Wel 16 (21)
Mod 40 (52)
Por 21 (27)
D. Hirano et al. 368
Ultr
astr
uct P
atho
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
SUN
Y S
tate
Uni
vers
ity o
f N
ew Y
ork
at S
tony
Bro
ok o
n 10
/31/
14Fo
r pe
rson
al u
se o
nly.
Electron Microscopy
Portions of the specimens obtained by surgical
extirpation or biopsy were immediately cut into 1-
mm blocks and placed in 2% glutaraldehyde in
0.2M cacodylate buffer (pH 7.4) for 2 h. The blocks
were postfixed in 1% buffered osmium tetroxide
for 1 h, dehydrated in graded alcohol, and embedded
in Epon 812. Semithin sections were cut and
observed by light microscopy to locate relevant areas
for electron microscopic examination. This sections
were cut with a diamond knife, using a LEICA Ultra-
tome UCT, stained with uranyl acetate and lead
citrate, and examined using an electron microscope
model 1200EX-II, JEOL, Tokyo, Japan.
Pre-embedding Immunoelectron
Microscopy
Portions of specimens obtained at surgery were
immediately placed in 4% paraformaldehyde, cut
into 50-mm-thick sections on a vibratome, and fixed
overnight. The sections were incubated in 3%
H2O2 in methanol for 30min at room temperature
to block endogenous peroxidase activity, and for
1 h with Dako Protein Block to prevent nonspecific
staining. The sections were incubated overnight with
rabbit anti-human CgA antibody diluted 1:250,
washed, and incubated with EnVision labeled
polymer-HRP anti-rabbit IgG (Dako) for 2 h. Immuno-
reactivity was visualized with 0.03% 3,30-diamino-
benzidine tetrahydrochloride as above. The sections
were then placed in 2% glutaraldehyde for 2 h, post-
fixed in 1% osmium tetroxide for 1 h, dehydrated in
graded alcohol, flatly embedded in Epon 812, and
mounted on glass slides.
Immunoreactive prostatic carcinoma cells were
located by light microscopy and relevant sections
were attached to conventional Epon 812 block, cut
with a diamond knife, and examined without uranyl
acetate and lead citrate by electron microscopy as
above.
Statistical Analysis
Associations between clinicopathological charac-
teristics and NE differentiation were analyzed for
statistical significance using the Mann-Whitney U test
and Kruskal-Wallis rank test with StatView software
(Abacus Concepts, CA, USA). All calculated p
values < .05 were considered statistically significant.
RESULTS
Immunohistochemical Analysis
Examination of the morphological features of
immunoreactive neoplastic cells indicated that in
well-differentiated tumors, NE differentiation was
most highly concentrated in scattered CgA-stained
areas or occasionally in individual CgA-positive cells;
in moderately differentiated tumors, including those
with a cribriform pattern, NE differentiation was
mostly found in focal or clustered NE neoplastic
cells; while in poorly differentiated tumors it was
mostly associated with diffuse NE cells (Figure 1).
CgA-positive tumor cells were found in 36 of the
77 specimens (47% of the patients). The staining
scores for immunoreactive cancer cells were þ1,
þ2, and þ3 in 28, 6, and 2 specimens, respectively.
CgA staining of the tumor cells was not observed in
the remaining 41 specimens (53% of the patients).
Analysis of the correlation between NE differen-
tiation and clinical stage indicated that the median
staining scores were 0 in disease that was organ-
confined clinically (stage B or less), and þ1 in
advanced disease (stage C or more). The CgA
staining scores were significantly higher in the
advanced tumors ( p < .003) (Figure 2). Analysis of
the relationship between NE differentiation and
histological grades indicated that the median staining
scores were 0 in well to moderately differentiated
tumors, and þ1 in poorly differentiated tumors.
The CgA staining scores were significantly higher
in tumors with the highest histological grade
( p < .04) (Figure 3).
Electron Microscopy
Electron microscopy revealed cells with unequivo-
cal NE differentiation in two radical prostatectomy
and one TUR specimens. The malignant NE cells
contained numerous neurosecretory granules such
as a variable number of dense-core-bound granules.
In the well-differentiated malignant glands, the api-
cal cytoplasm of the NE cells contained relatively
small and uniform dense core granules ranging from
200 to 600 nm. Some granules excreting into the
lumen were observed (Figure 4). Other organelles
369 Neuroendocrine-differentiated Carcinomas of the Prostate
Ultr
astr
uct P
atho
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
SUN
Y S
tate
Uni
vers
ity o
f N
ew Y
ork
at S
tony
Bro
ok o
n 10
/31/
14Fo
r pe
rson
al u
se o
nly.
such as the endoplasmic reticulum, Golgi apparatus,
and mitochondria were scattered in the cytoplasm. In
the cribriform pattern of the moderately differen-
tiated malignant glands, NE malignant cells clustered,
as observed by light microscopy, and had relatively
pleomorphic neurosecretory granules ranging from
120 to 770 nm (Figure 5). In the poorly differentiated
malignant glands, NE neoplastic cells contained
pleomorphic granules ranging from 20 to 600 nm;
these were lysosomal-like rather than NE-type
granules (Figure 6). Some of these cells extended
processes to adjacent carcinoma cells.
Immunoelectron Microscopy
CgA-immunoreactive prostate cancer cells, as seen
on light microscopy of semithin sections, were
detected in 1 of the 4 tumors analyzed by immuno-
electron microscopy. The immunoreactive malignant
cells were observed in the poorly differentiated
FIGURE 2 Relationship between NE differentiation and clinical
stage.
FIGURE 3 Relationship between NE differentiation and histo-
logical grade.
FIGURE 1 Immunohistochemical analysis of CgA staining in neoplastic glands (A) Well-differentiated tumors. NE cells are scattered in
the glands; immunostaining score: þ1. (B) Moderately differentiated tumors. NE cells are focal or clustered; immunostaining score: +2. (C)
Poorly differentiated tumors. Diffuse NE cells are seen in the glands; immunostaining score: þ3.
D. Hirano et al. 370
Ultr
astr
uct P
atho
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
SUN
Y S
tate
Uni
vers
ity o
f N
ew Y
ork
at S
tony
Bro
ok o
n 10
/31/
14Fo
r pe
rson
al u
se o
nly.
glands, and were characterized by the presence of
pleomorphic CgA-positive granules, round or elon-
gated in shape and of various sizes (Figure 7). The
morphological features of these immunoreactive
granules in the poorly differentiated glands were
similar to those in the poorly differentiated NE malig-
nant cells examined by conventional electron
microscopy.
DISCUSSION
NE cells, histologically identified by immunostain-
ing for serotonin and chromogranin [4, 10] or by sil-
ver precipitation reactions [11], are present in small
cell carcinomas or carcinoid tumors [1]. Although
small cell carcinomas or carcinoid tumors of the
prostate are extremely rare, NE differentiation in
prostatic adenocarcinoma is more frequent [2, 3]. In
the current study, none of the tumors had the
histological appearance specific to small cell NE
carcinomas or carcinoid tumors observed upon
hematoxylin and eosin staining, and the histology
spectrum for the carcinomas with NE differentiation
ranged from well-differentiated microacinar, through
cribriform, to poorly differentiated, as shown by
immunohistochemical analysis.
Pruneri et al. [12] have suggested that NE differen-
tiation was not associated with clinical stage, but was
related to a high Gleason score, while Speights et al.
[13] reported that NE differentiation and the prolifer-
ation index in high-grade=high-stage disease were
increased compared to those in low-grade=low-stage
disease. In this study, immunohistochemical analysis
indicated that NE-differentiation status correlates
with both high stage and high grade of disease.
As we [15] and Abrahamsson [14] have reported,
there is an association between the clinical features of
NE differentiation in malignant glands and hormone-
refractory prostate cancer. These clinicopathological
features together with our immunohistochemical
FIGURE 4 Electron micrograph of NE neoplastic cells in well-differentiated tumors showing numerous neurosecretory granules (arrow).
371 Neuroendocrine-differentiated Carcinomas of the Prostate
Ultr
astr
uct P
atho
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
SUN
Y S
tate
Uni
vers
ity o
f N
ew Y
ork
at S
tony
Bro
ok o
n 10
/31/
14Fo
r pe
rson
al u
se o
nly.
data suggest that NE differentiation in the malignant
prostatic glands is associated with aggressive beha-
vior and poor prognosis.
In conventional electron microscopy studies, NE
cells can be identified by the presence of classic
dense core granules in the cytoplasm [16, 17]. In this
study, malignant NE-differentiated cells were
identified in only 3 (33%) of the 9 specimens evalu-
ated by electron microscopy; 2 of these NE-positive
specimenswereobtainedfrompatientswithhormone-
refractory prostate cancer following androgen-
deprivation therapy. Weaver et al. studied the
characteristics of the granules in NE-differentiated
cells [17]. They found that the NE-differentiated
benign epithelial cells in the prostate contained elec-
tron-dense, lysosomal-like or exocrine-like granules
of variable size (460–690 nm) that filled the supra-
nuclear cytoplasm, while cells of prostatic carcino-
mas with Paneth cell-like change had NE-type
granules ranging from 240 to 480 nm. In contrast,
di Sant’ Agnese and de Mesy Jensen [16] pointed out
that cells from several NE-differentiated carcinomas
had an increased number of pleomorphic granules
when compared with benign prostatic NE cells. Heitz
andWegmann [18] described Paneth cells in an adeno-
carcinoma of stomach; these cells, histochemically
similar to NE-differentiated cells in prostatic carcino-
mas, had supranuclear homogeneous electron-dense
granules, an extensive Golgi complex, parallel arrays
of paranuclear rough endoplasmic reticulum, and
short microvilli. In the current study, NE-differen-
tiated cells in the well-differentiated malignant
glands possessed relatively homogeneous electron-
dense granules located in the supranuclear region,
some of which excreted into the lumen; however,
the other organelles were not developed. In the
poorly differentiated glands, NE-differentiated carci-
noma cells contained pleomorphic and lysosomal-
like granules of varying sizes, rather than NE-type
granules, and we demonstrated that the pleomorphic
FIGURE 5 Electron micrograph of NE neoplastic cells in moderately differentiated tumors showing relatively pleomorphic secretory
granules in clustered NE neoplastic cells.
D. Hirano et al. 372
Ultr
astr
uct P
atho
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
SUN
Y S
tate
Uni
vers
ity o
f N
ew Y
ork
at S
tony
Bro
ok o
n 10
/31/
14Fo
r pe
rson
al u
se o
nly.
or lysosomal-like granules observed by conventional
electron microscopy stained for CgA in the immunoe-
lectron microscopy analysis.
To our knowledge, no immunoelectron micro-
scopy study of NE-differentiated carcinomas
of the prostate has been published, while there have
been several such reports for carcinomas of other
organs, such as colon [19] and breast [20], using post-
embedding immunoelectron microscopy with a col-
loidal gold-labeled antibody. The NE tumors arising
from these organs had a relatively higher proportion
of NE-differentiated neoplastic cells. This study sug-
gests that pre-embedding proedures are appropriate
for identification of NE-differentiated prostatic carci-
nomas by immunoelectron microscopy; thus, among
the cancer cells that could be evaluated by light
immunohistochemical microscopy, over 10% CgA-
stained malignant cells were found in only 8 of the
77 specimens (10.3% of the cases) as well as the
extremely sparseness of NE-differentiated neoplastic
cells in the conventional electron microscopic
observation.
With respect to the biological characteristics of
prostatic carcinomas with marked NE features,
Bonkhoff et al. [3] documented that CgA-positive
tumor cells lacked the proliferation-associated MIB-
1 and Ki67 antigens that identify cycling cells in G1,
S, and M phases of the cell cycle, and exocrine cells
with increased proliferative activity were detected
surrounding these NE tumor cells. Endocrine cells
in prostate cancer produce a variety of regulatory
peptides with growth-promoting activities in vitro,
including serotonin, bombesin, and parathyroid
hormone-related peptide [4, 5, 21], while NE neoplas-
tic cells lack androgen receptors [22]. In addition,
recent studies [23, 24] show that prostatic NE cells
express the new anti-apoptosis protein, survivin,
and that prostate cancer cells with NE features escape
programmed cell death. These biological features,
together with our immunohistochemical and electron
FIGURE 6 Electron micrograph of NE neoplastic cells in poorly differentiated tumors showing lysosomal-like rather than neurosecre-
tory, pleomorphic granules of various sizes.
373 Neuroendocrine-differentiated Carcinomas of the Prostate
Ultr
astr
uct P
atho
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
SUN
Y S
tate
Uni
vers
ity o
f N
ew Y
ork
at S
tony
Bro
ok o
n 10
/31/
14Fo
r pe
rson
al u
se o
nly.
microscopy findings, suggest that NE neoplastic cells
act on the proliferation of surrounding cancer cells in
malignant tissues through the NE granules.
In conclusion, this study showed that the NE-
differentiation status in common prostate cancer
tissues was significantly increased in high-stage or
high-grade disease, and that poorly differentiated
NE malignant cells contained extremely pleomorphic
granules. In addition, we confirmed the presence of
CgA-immunoreactive deposits in the pleomorphic
granules by immunoelectron microscopy. These
results strongly suggest that NE-differentiated malig-
nant cells in prostate cancer tissues may induce
aggressive behavior in adjacent proliferating neo-
plastic cells via a paracrine mechanism.
REFERENCES
1. di Sant’ Agnese PA. Neuroendocrine differentiation in carcinoma ofthe prostate: Diagnostic, prognostic and therapeutic implications.Cancer. 1992;70:254–258.
2. di Sant’ Agnese PA, Cockett AT. Neuroendocrine differentiation inprostatic malignancy. Cancer. 1996;78:357–361.
3. Bonkhoff H, Wernert N, Dhom G, Remberger K. Relation of neuro-endocrine–paracrine cells to proliferation in normal, hyperplasticneoplastic human prostate. Prostate. 1991;19:91–98.
4. Abrahamsson PA, Wadstrom LB, Alumets J, Falkmer S, Grimelius L.Peptide hormone and serotonin-immunoreactive cells in carcinomaof the prostate. Pathol Res Pract. 1987;182:298–307.
5. Aprikian A, Han K, Guy L, Landry F, Begin LR, Chevalier S. Neuroen-docrine differentiation and bombesin=gastrin-releasing peptide fam-ily of neuropeptides in the progress of human prostate cancer.Prostate Suppl. 1998;82:738–743.
6. Haper ME, Glynne-jones E, Goddard L, Thurston VJ, Griffiths K. Vas-cular endothelial growth factor (VGEF) expression in prostatic tumorsand its relationship to neuroendocrine cells. Br J Cancer.1996;74:910–916.
7. di Sant Agnese PA, de Mesy Jensen K. Somatostatin and somatosta-tin-like immunoreactive endocrine–paracrine cells in the human pros-tate gland. Arch Pathol Lab Med. 1984;108:693–696.
8. Jewett HJ. The present status of radical prostatectomy for stage Aand B prostatic cancer. Urol Clin North Am. 1975;2:105–124.
9. Mostofi FK. Problems of grading carcinoma of prostate. Semin Oncol.1976;3:161–169.
10. Wilson BS, Lioyd RV. Detection of chromogranin in neuroendocrinecells with a monoclonal antibody. Am J Pathol. 1984;115:458–468.
11. Aguirre P, Scully RE, Wolfe HJ, Delellis RA. Endometrial carcinomawith argyrophil cells: A histochemical and immunohistochemicalanalysis. Hum Pathol. 1984;15:210–217.
FIGURE 7 Immunoelectron micrograph of NE neoplastic cells showing CgA-immunoreactive deposits in pleomorphic granules.
D. Hirano et al. 374
Ultr
astr
uct P
atho
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
SUN
Y S
tate
Uni
vers
ity o
f N
ew Y
ork
at S
tony
Bro
ok o
n 10
/31/
14Fo
r pe
rson
al u
se o
nly.
12. Pruneri G, Galli S, Rossi RS, et al. Chromogranin A and B secretogra-nin II in prostatic adenocarcinomas: Neuroendocrine expression inpatients untreated and treated with androgen deprivation therapy.Prostate. 1998;34:113–120.
13. Speights VO, Cohen MK, Riggs MW, Coffield KS, Keegan G, Arber DA.Neuroendocrine stains and proliferative indices of prostatic adenocarci-nomas in transurethral resection samples. Br J Urol. 1997;80:281–286.
14. Abrahamsson PA. Neuroendocrine differentiation and hormonerefractory prostate cancer. Prostate Suppl. 1996;6:3–8.
15. Hirano D, Okada Y, Minei S, Takimoto Y, Nemoto N. Neuroendocrinedifferentiation in hormone refractory prostate cancer followingandrogen deprivation therapy. Eur Urol. 2004;45:586–592.
16. di Sant’ Agnese PA, de Mesy Jensen K. Neuroendocrine differen-tiation in prostatic carcinoma. Hum Pathol. 1987;18:849–856.
17. Weaver MG, Abdul-Karim FW, Strigley J, Bostwick DG, Ro JY, AyalaAG. Paneth cell-like change of the prostate gland: A histological,immunohistochemical, and electron microscopic study. Am J SurgPathol. 1992;16:62–68.
18. Heitz PU, Wegmann W. Identification of neoplastic paneth cells in anadenocarcinoma of the stomach using lysozyme as a marker, andelectron microscopy. Virchows Arch A. 1980;386:107–116.
19. Seretis E, Gavrill A, Agnantis N, Golematis V, Voloudakis-Baltatzis IE.Comparative study of serotonin and bombesin in adenocarcinomasand neuroendocrine tumors of the colon. Ultrastruct Pathol.2001;25:445–454.
20. Monaghan P, Roberts JDB. Immunocytochemical evidence for neu-roendocrine differentiation in human breast carcinomas. J Pathol.1985;147:281–289.
21. Iwamura M, Wu G, Abrahamsson PA, di Sant’ Agnese PA,Cockett ATK, Deftos LJ. Parathyroid hormone-related protein isexpressed by prostatic neuroendocrine cells. Urology.1994;43:667–674.
22. Krijnen JLM, Janssen PJA, Ruizevel de Winter JA, van Krimpen H,Shroder FH, van der Kwast TH. Do neuroendocrine cells in humanprostate cancer express androgen receptor? Histochemistry.1993;100:393–398.
23. Xing N, Qian J, Bostwick D, Bergstralh E, Young CYF. Neuroendo-crine cells in human prostate over-express the antiapoptosis proteinsurviving. Prostate. 2001;48:7–15.
24. Fixemer T, Remberger K, Bonkhoff H. Apoptosis resistance of neu-roendocrine phenotype in prostatic adenocarcinoma. Prostate.2002;53:118–123.
375 Neuroendocrine-differentiated Carcinomas of the Prostate
Ultr
astr
uct P
atho
l Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
SUN
Y S
tate
Uni
vers
ity o
f N
ew Y
ork
at S
tony
Bro
ok o
n 10
/31/
14Fo
r pe
rson
al u
se o
nly.