close

Вход

Забыли?

вход по аккаунту

?

967

код для вставкиСкачать
The Prostate 43:11–19 (2000)
Patterns of Differentiation and Proliferation in
Intraductal Carcinoma of the Prostate:
Significance for Cancer Progression
Ronald J. Cohen,1* John E. McNeal,2 and Tina Baillie3
1
Urological Research Centre, University of Western Australia,
Perth, Western Australia, Australia
2
Department of Urology, Stanford University Medical Center, Stanford, California
3
Department of Pathology, University of Auckland, Auckland, New Zealand
BACKGROUND. Cribriform prostatic intraepithelial neoplasia (C-PIN) identifies a unique
histological pattern: dysplastic cells line ductal/acinar walls but also span across gland lumens. C-PIN is distinct from other forms of dysplasia; it is seldom seen except within invasive
cancer, it is more frequent in larger/higher-grade cancers; and it appears to contribute independently to aggressive behavior. Hence, C-PIN may represent a separate, more aggressive
entity: intraductal carcinoma of the prostate (IDC-P). Here, support for that distinction stems
from a histologic/biologic subclassification of IDC-P, whose elements are linked to features of
invasive cancer.
METHODS. Histologic criteria were tested against 26 radical prostatectomies, using immunostains for prostate-specific antigen, MUC-2, androgen receptor (differentiation), and Ki-67
(proliferation). Invasive cancer grade, stage, and follow-up were compared.
RESULTS. Architecture of the central (luminal) cell compartment defined three subclasses of
IDC-P: A (trabecular), B (cribriform), and C (solid/comedo), which represented progressive
dedifferentiation with a reciprocal increase in proliferation. The IDC-P subpattern correlated
with cancer stage, grade, and clinical course.
CONCLUSIONS. IDC-P is a separate entity, distinct from PIN; cancers associated with IDC-P
are more aggressive than those associated with only PIN. It comprises a spectrum of histological patterns which appear to be determined in concert with invasive cancer, whose prognosis it worsens. Prostate 43:11–19, 2000. © 2000 Wiley-Liss, Inc.
KEY WORDS:
classification; intraductal prostate cancer; proliferation; differentiation
INTRODUCTION
Prostatic ducts and acini whose lumens are filled by
malignant-appearing glandular epithelial cells represent a well-recognized histologic pattern [1–3], whose
precise histologic criteria and biologic significance
have been variously defined. In the Gleason grading
system [4], this histologic pattern was classified with
invasive cancers and was designated a cribriform subpattern of grade 3 carcinoma. More recently, the identification of basal cells surrounding cribriform nests
has stimulated the reclassification [5] of this lesion as
cribriform prostatic intraepithelial neoplasia (cribriform PIN). Though referring to the same histologic
© 2000 Wiley-Liss, Inc.
pattern, the implications for biologic behavior between these two definitions are vastly different.
PIN or duct-acinar dysplasia refers to alterations of
the epithelial layer lining preexisting benign lumens
by histologic features resembling carcinoma. These include cell proliferation with nuclear multilayering,
nuclear enlargement with nonuniformity of size and
Grant sponsor: Richard M. Lucas Foundation; Grant sponsor: Research Committee, Sir Charles Gairdner Hospital.
*Correspondence to: Dr. R.J. Cohen, Urological Research Centre,
Level 2, M Block, QE II Medical Centre, Nedlands, Western Australia 6009, Australia. E-mail: ronnie@urc.urc.uwa.edu.au
Received 5 April 1999; Accepted 3 November 1999
12
Cohen et al.
contour, hyperchromasia, and enlarged nucleoli. Dysplasia is probably the main precursor lesion for invasive cancer in the prostate, and hence it is often recognized in prostate glands that have not yet
developed invasive carcinoma.
Unique among dysplasia variants, cribriform PIN
has atypical cells completely spanning the lumen and
often filling it [5]. In order to shed light on disparities
in histological classification, we recently studied this
lesion in a multivariate morphologic analysis of 136
serially blocked radical prostatectomy specimens [6].
We found other histologic patterns of severe dysplasia
(PIN) to be extremely common outside the boundaries
of invasive cancer areas, while the cribriform pattern
(and other lumen-spanning patterns) was rare except
inside the boundaries of invasive carcinoma. In addition, duct-spanning lesions were very uncommon in
small tumors, but showed a linear increase in frequency parallel to the volume of invasive carcinoma,
and they reached a peak of 40% in carcinomas over 4
cm3. It was concluded that the above features were
unlikely for a cancer precursor, but they were quite
consistent with the emergence of cribriform duct involvement within an already established invasive carcinoma.
The frequency of intraductal, duct-spanning lesions
also correlated highly with percent of high-grade
(grades 4/5) invasive cancer, but multivariate analysis
indicated an additional independent role for such luminal cell masses in predicting poor prognosis [6–8].
Further, the close resemblance of these duct-spanning
lesions with the histologic and biologic features of the
concurrent invasive cancer was inconsistent with the
features of dysplasia (PIN); accordingly, they were
designated as a separate entity, intraductal carcinoma
[6] of the prostate (IDC-P).
Most of the above findings have been confirmed by
two subsequent studies [7,8]. However, these authors
chose to retain the “cribriform PIN” terminology, despite the declaration in one paper [7] that this “represents intraductal spread of carcinoma within preexisting ducts and acini and should not be categorized as
PIN.”
As a common entity and one closely related to invasive cancer progression, intraductal carcinoma of
the prostate (IDC-P) warrants further investigation.
Accordingly, in this current study we investigated the
different histologic patterns of IDC-P and thoroughly
describe their features. With immunohistochemical
staining for markers of differentiation and proliferation, we have been able to place these along a histologic continuum, which correlates with biologic aggressiveness.
MATERIALS AND METHODS
Tissues
A series of 26 prostatic adenocarcinomas with areas
of intraductal carcinoma (IDC-P) was randomly selected from two previous published series [6,9]. The
first 19 cases were collected from prostatectomies performed at the Stanford University Medical Center, and
each was represented by two selected slides containing the largest IDC-P areas. In all 19 cases, the entire
prostate was blocked and each 3-mm transverse slice
was divided into quarters for ease of processing.
Seven prostatectomies from the Department of Pathology at Auckland University were added and represented by serial whole-mount sections. From both centers, cases had been excluded which showed poor
fixation or had a history of hormonal or irradiation
therapy, and all tumors were peripheral-zone cancers.
All specimens were fixed whole for 18–23 hr in 10%
buffered formalin. Constant circulation of formalin
around the specimen as well as perfusion through the
urethra, using a small pump, assured adequate and
prompt fixation. Fixed prostates were serially blocked
at 3-mm intervals and transferred to 70% ethanol.
There was no further contact with formalin, and all
tissues were processed through solvents to paraffin
wax.
Immunohistochemistry
Eight 5-␮m sections were cut from blocks of all 26
cases and subjected to the following stains: hematoxylin and eosin, anti-prostate-specific antigen (PSA) (ERPR8, 1:100; PASE/4LJ, 1:100, Dako,Glostrup, Denmark), anti-Ki-67 (clone MIB1, 1:100, Immunotech SA,
France), anti-high molecular weight cytokeratin
(HMWCK) (34␤E12,1:50, Dako), anti-androgen receptor (AR) (NCL-ARp(AR70), 1:200, Novocastra Laboratories, UK), MUC-1 (MUSE 11, undiluted [10]), and
MUC-2 (CCP58, 1:1,000 [11]). Heat antigen retrieval
was used in the immunostaining for HMWCK, Ki-67,
and AR.
Intraductal-acinar cancer areas were identified by
previously published criteria [6]. Distinction from
high-grade dysplasia (prostatic intraepithelial neoplasia) required the presence of a “central cell compartment” of epithelium completely spanning the gland
lumen. The intraductal-acinar pattern almost always
comprised one or two well-demarcated areas within
which most contiguous glandular lumens were filled
with cancer of relatively uniform histology.
The immunohistochemical staining was graded according to the intensity of the stain as well as the
Intraductal Carcinoma of the Prostate
Fig. 1. Association of basal cells (arrows) with the epithelial
stromal boundary of cribriform (pattern B) carcinoma growing in
ducts (HMWCK immunostain, ×40).
number of positive cells. Intense positivity in almost
all tumour cells was designated “+++”; 30–50% of
stained tumor cells intensely positive or greater than
50% but with less intense staining was designated
“++”; intense staining of more than 5% but less than
30% of cells or less intense staining of more than 10%
but less than 50% of cells was designated “+”. Lesser
degrees of staining were designated “±”, and complete absence of staining was noted as “−”. McNemar
chi-square analysis was used to compare the different
staining patterns.
For the invasive component of all cancers, the Gleason grade [4] and Gleason score (sum) were determined. Total cancer volume (invasive and IDC-P) was
determined for all 26 prostatectomy cancers [6,9], as
previously reported.
13
Fig. 2. Pattern A intraductal carcinoma (IDC-P) with a “trabecular” architecture, showing two cytologic compartments. The perimeter compartment associated with pleomorphic cells containing macronucleoli can be distinguished from the bland central cells
with small nuclei (×240).
RESULTS
with, the periglandular stroma had prominent malignant features which were indistinguishable from those
of the glandular lining cells of severe dysplasia. This
population was identical in all 36 cases. This area of
histologic equivalence to dysplasia was designated the
perimeter compartment.
We designated as the central compartment all the remaining luminal epithelial cells, which comprised the
transluminal mass and which lay beyond the thickness of the typical dysplastic glandular lining. It was
this compartment alone which showed great variation
in histologic appearance between cases, and these differences were the basis for defining different histologic patterns of IDC-P. In almost all cases, the central
compartment was histologically different from the perimeter compartment, and a sharp line of demarcation
was usually visible, facilitating their distinction.
Histologic Characterization of Intraductal
Carcinoma (IDC-P)
Histologic Patterns of IDC-P
Using previous criteria [6], 26 peripheral zone carcinomas at 4–10 cm3 volume had foci of contiguous
IDC-P ducts and acini which comprised more than
10% of the estimated total cancer volume. In IDC-P
foci of all cases, there were frequent areas mimicking
normal branching duct-acinar prostatic architecture.
Among the 26 cases, 20 (79%) demonstrated a complete basal-cell layer surrounding the intraductal tumor masses (Fig. 1), while in 6 cases the basal-cell
layer was less complete, with spaces and gaps between immunostained areas.
Detailed cytologic examination of epithelial cells of
all glandular IDC-P foci showed that the ring of secretory epithelial cells which was close to, or in contact
The histologic features of the central compartment
in all cases showed one of three architectural patterns.
In pattern A, empty luminal space was the larger component centrally (Fig. 2). The epithelium exhibited a
trabecular pattern; narrow strands, often only two cell
layers thick, spanned the lumen without stromal support, intersecting randomly to form a lacework of very
orderly structure. Sometimes coalescence of strands in
the very center of the lumen created a small cell mass.
The central compartment nuclei were small and uniform, enhancing the orderly appearance, and their cytoplasm was often pale; thus, demarcation between
compartments was clear.
In pattern B, the classic cribriform pattern, multiple
14
Cohen et al.
Fig. 3. Pattern B intraductal carcinoma (IDC-P). The tumor expands existing ducts and acini, and has a cribriform architectural
arrangement (×48).
small, round to elongated lumens perforated a central
compartment cell mass which typically comprised
more than half of the luminal space (Fig. 3). The component microlumens varied considerably between
cases in their size and frequency. Central-compartment nuclei were usually somewhat smaller and more
bland than perimeter-compartment nuclei, with frequent suggestion of a line of demarcation. A small
focus of coagulation necrosis was found at the exact
center of an occasional gland.
Pattern C was a solid pattern throughout the central
compartment (Fig. 4), without clear differentiation between this central area and the perimeter cells. In a
subgroup of this pattern, prominent coagulation necrosis (Fig. 5) obliterated the cells of most of or the
entire central compartment (comedonecrosis), and
both compartments tended to show an even greater
degree of nuclear enlargement and pleomorphism.
Most cases demonstrated a single pattern throughout. Occasional IDC-P showed areas of blending between two patterns, a predominant pattern (>90%)
and a minor pattern (<10%), which were always adjacent categories of the above listing, suggesting evolution along a continuum.
Fig. 4. Pattern C intraductal carcinoma (IDC-P) with a solid
architectural arrangement. The biphasic nature is less obvious with
the loss of the central compartment (×48).
Fig. 5. “Comedo”-like necrosis, identified most frequently in
solid pattern C IDC-P (×60).
Cytoplasmic Differentiation of IDC-P Patterns
by Immunostaining
PSA staining of “+” level or higher in pattern A was
limited to the central compartment in 4 cases (Fig. 6)
and occupied both compartments in the remaining 2
cases, one of which was the only “+++” case among 26
cases (Fig. 7). In pattern B, 4 of 15 cases had no PSA
staining, but in the remaining 8 cases with “+” level or
greater, staining was limited to the central compart-
Fig. 6. Immunostaining of IDC-P with anti-prostate-specific antigen. Luminal staining is noted, while the pleomorphic perimetercell compartment demonstrates little immunoreactivity (×480).
Intraductal Carcinoma of the Prostate
15
Fig. 7. PSA and androgen receptors detected in the two compartments (central and perimeter) in the three patterns (A–C) of intraductal
carcinoma. PSA is mainly confined to the central compartment in all types, while androgen receptors are mostly identified in the perimeter
except in a proportion of less differentiated tumors, where their expression is more widespread.
Fig. 8. MUC-2 immunostaining (arrows) demonstrates positive
staining focally in the central cell compartment of IDC-P (×60).
ment. Likewise, in pattern C, 2 of 5 cases showed no
staining, but the remaining 3 were centrally stained.
Overall, the difference between central and perimeter
staining was highly significant (P < 0.001).
MUC-2, which stains intestinal goblet-cell mucus,
was “+” or higher in 83% of pattern A cases, 80% of
pattern B, and none of pattern C. Positive staining
was selectively localized to the central compartment,
with little or no perimeter stain (Fig. 8). Positively
stained cells were often adjacent to pools of extracellular hematoxiphilic mucin. MUC-1 was negative in
all cases.
Androgen receptor was regarded as a differentiation antigen because of its exclusive presence in nonproliferative secretory cells in benign epithelium. In
pattern A IDC-P, it was localized to the perimeter
compartment in 100% of 6 cases (Fig. 7). This compartment is the analogue of the entire secretory lining of dysplastic (PIN) or normal epithelium. Hence,
there was no shift in location (Fig. 9). In pattern B, 5
of 15 cancers had no staining, 4 stained only peripherally, and the remaining 6 were nearly equally
expressed in both compartments. In pattern C, only
1 of 5 cases showed “+” staining in the central compartment, and it was equally stained in the perimeter.
One case showed no stain and one was “±”. The
remaining 2 were exclusively identified in the perimeter compartment. Strong staining was found in
50% of pattern A and 20% of patterns B and C. Overall perimeter staining for the androgen receptor was
significantly different from the central staining (P <
0.01).
16
Cohen et al.
Fig. 9. Androgen receptor immunostaining confirms nuclear
positivity in the perimeter cell compartment of IDC-P (×120).
Fig. 10. MIB-1 immunostaining identifies many nuclei of cells in
cycle, which are restricted to the perimeter compartment in pattern A IDC-P (×120).
Proliferation in IDC-P Patterns by Immunostaining
Immunostaining for MIB-1 was positive in many
nuclei of patterns A–C among the 26 prostatectomies.
In nearly all cases, frequency and intensity of staining
were comparable to the nuclei of adjacent invasive
carcinoma and much higher than in benign glands.
Distinctively, in patterns A and B alike, proliferationpositive nuclei were almost completely confined to the
secretory cells of the perimeter compartment (Fig. 10).
By contrast, in pattern C prostatectomies, MIB-1positive nuclei were scattered randomly throughout
the entire luminal masses without regard for the central vs. perimeter compartments.
Correlations Between IDC-P Pattern and Invasive
Cancer or Cancer Stage
The frequency of IDC-P pattern C among patients
selected for prostatectomy (metastases not demonstrable clinically) was 20%, while 70% of patients with
known metastases had IDC-P pattern C. Although the
numbers are small and selection bias cannot be ruled
out, a relationship between late stage and high IDC-P
pattern is suggested.
Prostatectomy cases were split into two sets according to whether Gleason sum in the invasive element
was 7 or less as compared to those with a sum of 8 or
more (Table I). The number of cases within these sets
and their coexistent IDC-P patterns (A–C) were analyzed in a chi-square table which demonstrated a significant association (P < 0.01). Follow-up data were
obtained on 17 of the Stanford prostatectomy cases
and all of the Auckland cases (total of 24 of 26 cases).
None of the 4 pattern A Stanford patients and only 1
of the 2 pattern A Auckland cases had progressed
(biochemically) by 3 years [9]. Of the 13 Stanford pat-
tern B and C cases, 54% had relapsed, while all Auckland pattern B and C (3 and 2 cases, respectively) had
clinical or biochemical evidence of tumor progression.
DISCUSSION
In a series of 26 prostatectomy samples with adenocarcinoma, all showing prominent sectors of intraductal, duct-spanning epithelial proliferation, we distinguished three duct-acinar histologic subpatterns.
All three met the recently presented criteria for intraductal carcinoma (IDC-P), which has been proposed
as one pathway of aggressive progression of invasive
cancer [6].
The IDC-P central cell compartment (lumen spanning
cell plug) was the histological basis for distinguishing
between subpatterns: it showed progressive loss of
differentiation from pattern A through B to C. Cells in
this compartment manifested a progressive loss of
both number and aggregate area of microlumens
within a duct or acinus. In close parallel, the centralcompartment cells developed larger, more malignantappearing nuclei. In pattern C, central cells reached a
level of anaplasia equivalent to the perimeter cells,
and there were no longer two distinguishable compartments.
The perimeter cell compartment (cells in close proximity to the stroma) showed the same anaplastic features in all three subpatterns: these anaplastic features
were indistinguishable from those of high-grade dysplasia (PIN), from which it has been proposed that
IDC-P evolves [6].
Morphological evolution of IDC-P from dysplasia is
further supported by the similar patterns of immunostaining for the androgen receptor [12], which in both
Intraductal Carcinoma of the Prostate
17
TABLE I. Tumor Parameters of 26 Prostate Cancers With IDC-P
Gleason grade
Case (n = 26)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Major
Minor
IDC-P
subtype
% IDC-P
Tumor volume
(cm3)a
Serum PSA
(␮g/l)
3
4
3
4
3
3
4
3
3
4
5
3
3
3
3
3
4
3
4
4
5
4
3
3
4
4
3
3
3
3
4
3
4
3
4
5
4
3
4
3
4
4
4
4
4
4
4
5
4
4
5
5
A
A
A
A
A
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
C
C
C
C
C
11–25
11–25
50–75
11–25
50–75
11–25
50–75
75–100
50–75
50–75
50–75
50–75
50–75
50–75
50–75
25–50
50–75
11–25
50–75
50–75
50–75
50–75
11–25
50–75
75–100
50–75
7.0
4.0
22.0
5.0
2.0
9.0
6.0
5.0
10.0
7.0
4.5
6.0
9.0
5.0
6.0
6.0
5.0
6.0
8.0
8.0
3.5
7.0
5.0
6.0
8.5
35.0
14.3
22.5
11.4
3.2
5.5
15.0
8.3
11.6
13.2
8.9
36.3
19.8
13.2
27.9
15.0
14.3
14.7
9.4
7.5
17.0
9.0
25.9
7.8
6.0
12.6
8.0
a
Tumor volumes were rounded to the nearest 0.5 cm3.
lesions identifies a population of proliferating secretory cells. Conversely, in the normal prostate gland,
androgen receptors are identified only in secretory
cells that are nonproliferating, which contrasts with
dysplasia [12]. Similarly in IDC-P, androgen-receptor
staining is restricted to the perimeter (proliferative)
compartment in pattern A, while in pattern B and
more particularly pattern C, loss of distinction between the central and perimeter compartments parallels a loss of segregation of androgen-receptor staining, which we interpret as a loss of differentiation.
Immunohistochemical staining for differentiation
and proliferation markers also supports the concept
that patterns A–C represent a continuum of increasing
malignant potential which is assessed by loss of differentiation. The central compartment in pattern A
displayed the cytoplasmic differentiation markers
PSA and MUC-2, while proliferation was absent, a
phenotype consistent with the benign appearance of
these cells. All proliferation was confined to the perimeter compartment, where it was largely displaced
from the basal cells to the secretory cells, as is described in dysplasia [12]. In the central compartment,
proliferation increased with advance in pattern reciprocal to decreasing differentiation, and in pattern C,
proliferation in the central compartment was identical
with the perimeter.
We propose that the differences in histologic pattern and immunostaining in the central compartment
are coordinated markers of the biologic capacity of the
perimeter compartment to generate differentiated
progeny. By contrast, in dysplasia, progeny do not
survive in the lumen, presumably because they have
not yet acquired the capacity to survive without stromal contact. Thus, in IDC-P the acquisition of the ability to survive among epithelial cells which are bound
only to each other without stromal contact reflects an
abrupt increase in malignant potential, drawing similar parallels to the distinction between invasive grade
4 cancer and the common type of grade 3 carcinoma.
Our findings are of practical clinical significance:
patient prognosis worsens in parallel with advancing
pattern of IDC-P. This finding confirms that of a recent
larger study of 108 cases [8], patterns corresponding to
our pattern A plus B showed a 35% progression rate,
while tumors corresponding to our pattern C pro-
18
Cohen et al.
TABLE II. Features of IDC-P-Containing Cancers Which Differ From Those of Carcinomas Without IDC-P*
Rubin et al. series [7]
Cancer volume >4 cm3
Invasive carcinoma, Gleason score ⱖ7
Invasive carcinoma, Gleason score ⱖ8
Frequency of progression
Wilcox et al. series [8]
IDC-P+ (%)
IDC-P− (%)
IDC-P+ (%)
IDC-P− (%)
60
80
17
41
54
30
61
14
29
49
6
12
*IDC-P+, IDC-P-containing cancers; IDC-P−, carcinomas without IDC-P.
gressed in more than 80% of cases. This follows observations that IDC-P unsorted for pattern has an
overall independent effect on prognosis [6], and progression frequency is fourfold higher [7,8] when compared to cancers without IDC-P (Table II).
With IDC-P, the frequency of Gleason grade 4/5
cancer is reportedly twice that of other prostate cancers [7,8]. This partly explains the adverse influence of
IDC-P overall (Table II). However, we have also found
that the frequency of Gleason grades 4 and 5 also correlated with advancing pattern (C > B > A). A third
adverse factor is cancer volume; cancers with IDC-P
were four times more likely to exceed 4 cc than cancers
without IDC-P elements [6–8]. Thus there is good evidence that the relationship between cancer volume,
grade, and IDC-P play an important role in aggressive
malignant behavior.
By contrast, dysplasia (PIN) without the cribriform
pattern stands apart from IDC-P as a lesion that is
unable to influence volume, grade, or progression of
the invasive carcinoma. The recent references [7,8] to
intraductal carcinoma as “cribriform PIN” contradict
the caution given by both sets of authors that this
lesion must be considered separate from PIN, since it
influences the clinical course of disease. We propose
that this lesion should be classified as a distinct biological and histological entity and termed intraductal
carcinoma of the prostate (IDC-P).
In the Gleason grading system [4], patterns A and B
of IDC-P were called grade 3 (invasive) cancer, cribriform subpattern. Pattern C was called grade 4 if it was
composed of solid cell masses, and it was called grade
5 comedocarcinoma if it showed prominent central
comedonecrosis. These entities should be recognized
and called by the names that fit their true biological
and clinical significance. For diagnostic purposes, only
in Gleason grade 5 carcinoma does it seem not to matter whether the lesion is intraductal or invasive [8].
CONCLUSIONS
Confirming recent studies, we found evidence for
the decreasing cure rate and increasing stage along the
histologic pattern continuum of IDC-P. The pattern
also appeared to correlate with histologic grade of the
invasive component. The close kinship between suggested subpatterns of IDC-P and grades of invasive
cancer supports the concept that IDC-P evolves synchronously with invasive high-grade elements.
Consequently, an attempt should be made to eliminate disparities in terminology because of its adverse
affects on accurate diagnosis. The identification of a
single or several isolated cribriform ducts on needle
biopsy has significant clinical implications and should
stimulate multiple rebiopsies, with the conviction that
invasive and probably high-grade/large-volume cancer will be found.
ACKNOWLEDGMENTS
We thank Dr. Amanda Segal (Consultant Pathologist, PathCentre, Perth, Western Australia) for her
contribution, and Mr. Mario D’Antuono for biostatistical advice. We also acknowledge the Richard M. Lucas Foundation’s financial support to J.E.M, and Abbott Australasia Pty Ltd. for their financial support.
REFERENCES
1. Dube V, Farrow G, Green L. Prostatic carcinoma of ductal origin. Cancer 1973;32:402–409.
2. Kovi J, Jackson M, Heshmat M. Ductal spread in prostatic carcinoma. Cancer 1985;56:1566–1573.
3. McNeal J, Reese J, Redwine E, Freiha F, Stamey T. Cribriform
adenocarcinoma of the prostate. Cancer 1986;58:1714–1719.
4. Gleason DF. Histologic grading and clinical staging of prostatic
carcinoma (the Veterans Administration Cooperative Urological
Research Group). In: Tannenbaum M, editor. Urologic pathology: the prostate. Philadelphia: Lea and Febiger; 1977. p 171–
174.
5. Bostwick D, Amin M, Dundore P, Marsh W, Schultz D. Architectural patterns of high-grade prostatic intraepithelial neoplasia. Hum Pathol 1993;24:298–310.
6. McNeal JE, Yemoto CEM. Spread of adenocarcinoma within
ducts and acini: morphologic and clinical correlations. Am J
Surg Pathol 1996;20:802–814.
7. Rubin M, de La Taille A, Bagiella E, Olsson C, O’Toole K. Cribriform carcinoma of the prostate and cribriform intraepithelial
Intraductal Carcinoma of the Prostate
neoplasia: incidence and clinical implications. Am J Surg Pathol
1998;22:840–848.
8. Wilcox G, Soh S, Chakraborty S, Scardino P, Wheeler T. Patterns
of high-grade prostatic intraepithelial neoplasia associated with
clinically aggressive prostate cancer. Hum Pathol 1998;29:1119–
1123.
9. Cohen R, Chan W, Edgar S, Robinson E, Dodd N, Hoscek S,
Mundy I. Prediction of pathologic stage and clinical outcome in
prostate cancer; an improved pre-operative sequential model
incorporating biopsy determined intraductal carcinoma. Br J
Urol 1998;81:413–418.
10. Ohc Y, Hinoda Y, Irimura T, Imai K, Yechi A. Expression of
sulphated carbohydrate chains detected by monoclonal anti-
19
body 91,9H in human gastric cancer tissues. Jpn J Cancer Res
1994;85:400–408.
11. Blank M, Klussmann E, Kruger-Krasagakes S, Schmitt-Graff A,
Stolte M, Bornhoeft G, Stein H, Xing P, McKenzie I, Verstijnen C,
Riecken EO, Hanski C. Expression of MUC-2 mucin in colorectal
adenomas and carcinomas of different histological types. Int J
Cancer 1994;59:301–306.
12. Leav I, McNeal JE, Kwan PW-L, Komminoth P, Merk FB. Androgen receptor expression in prostatic dysplasia (prostatic intraepithelial neoplasia) in the human prostate. An immunohistochemical and in situ hybridization study. Prostate 1996;29:
137–145.
Документ
Категория
Без категории
Просмотров
2
Размер файла
1 905 Кб
Теги
967
1/--страниц
Пожаловаться на содержимое документа