close

Вход

Забыли?

вход по аккаунту

?

772

код для вставкиСкачать
57
CANCER
CYTOPATHOLOGY
Effusion Cytology of Malignant Melanoma
A Morphologic and Immunocytochemical Analysis Including Application of the
MART-1 Antibody
Michael W. Beaty, M.D.1
Patricia Fetsch, M.T., A.S.C.P.1
Anna Maria Wilder, C.T., A.S.C.P.1
Francesco Marincola, M.D.2
Andrea Abati, M.D.1
BACKGROUND. Malignant effusions are complications of metastatic malignant melanoma (MM). Differential diagnosis often involves distinguishing MM from adenocarcinoma and reactive mesothelial cells. Descriptions in the literature of the morphologic and immunocytochemical (IM) staining characteristics of MM in effusions
are sparse. A combination of morphology and immunocytochemistry should yield
the most accurate diagnostic results. The MART-1 antigen, a transmembrane pro-
1
Cytopathology Section, National Institutes of
Health, National Cancer Institute, Bethesda,
Maryland.
2
Surgery Branch, National Institutes of Health,
National Cancer Institute, Bethesda, Maryland.
tein, is specifically expressed in melanocytes and MM. A recently developed monoclonal antibody to the MART-1 antigen may represent a useful marker for the
identification of MM in effusions.
METHODS. The authors conducted a retrospective review of 32 effusion samples
diagnosed as MM. The review consisted of morphologic and IM analyses of the
effusion samples with antibodies to MART 1, HMB45, S-100, and cytokeratins (AE1/
AE3).IM stains were performed on cell block or cytospin material, depending on
availability. In the morphologic review, emphasis was placed on Diff Quik– stained
material, due to its enhanced cytoplasmic volume and detail.
RESULTS. Predominant cytologic features noted were lack of cellular cohesion (in
100% of cases), large eccentric nuclei with prominent nucleoli (in 100%), multinucleation (in 84%), variable cytoplasmic vacuolization (in 75%), pigment (in 72%),
and cell-in-cell engulfment (in 47%). All immunoreactive cases with sufficient material stained with at least one of the markers used. Tumor cells were positive with
IM stains to MART-1 in 78% of cases, HMB45 in 81%, and S-100 in 81%. Coexpression of MART-1, HMB45, and S-100 was noted in 63% of cases. Of cases that
showed expression for only 1 of the 3 antigens, the MART-1 was positive in 1
case, and HMB45 and S-100 were positive in 2 cases each. Three cases showed
immunoreactivity for cytokeratins in the melanoma cells.
CONCLUSIONS. The diagnosis of MM in effusions can be made reliably through a
combination of morphologic and IM features. Differential diagnosis often involves distinguishing MM from adenocarcinoma or reactive mesothelial cells. Cytoplasmic vacuolization, multinucleation, prominent nucleoli, and cell-in-cell engulfment are cytologic
features common to all three. The lack of IM staining for cytokeratins alone cannot
reliably distinguish MM; 11% of cases showed positive staining with this antibody in
the melanoma cells. The use of a panel of antibodies increases the accuracy of diagnosing MM. In this study, MART-1 proved a useful adjunct to the HMB45/S-100/cytokeratin
panel for the diagnosis of MM in effusions, staining 78% of the immunoreactive cases,
with positivity in 1 case that was negative for HMB45 and S-100. Cancer (Cancer
Address for reprints: Andrea Abati, Cytopathology Section, NIH/NCI Building 10, Room 2A19,
10 Center Drive, Bethesda, MD 20892.
Received November 8, 1996; revision received
December 20, 1996; accepted December 23,
1996.
Cytopathol) 1997;81:57–63. q 1997 American Cancer Society.
KEYWORDS: melanoma, effusion, immunocytochemistry, MART-1 antibody, triage.
M
alignant effusions associated with metastatic melanoma (MM)
may exhibit a variable cytologic appearance.1 – 5 The diagnosis
q 1997 American Cancer Society
/ 7302$$935d
01-25-97 07:41:47
ccyta
W: Can Cyto
58
CANCER (CANCER CYTOPATHOLOGY) February 25, 1997 / Volume 81 / Number 1
is facilitated by the positive identification of cytoplasmic melanin pigment. In the absence of obvious
pigment, adjunctive studies are necessary to confirm
the diagnosis.1,6 Immunocytochemical analyses using
melanoma-associated antigen HMB45 and S-100 protein are the current ‘‘gold standard’’ in this regard.6 –9
The identification of MART-1, a specifically expressed, transmembrane, melanocytic and melanoma
protein recognized by tumor infiltrating T-lymphocytes, has led to the development of a monoclonal
antibody at our institution.10,11 Aside from its presence
in normal melanocytes, retinal tissue, and malignant
melanoma, MART-1 RNA has not been detected in
other normal tissues.12,13 Ongoing immunohistochemical studies in our laboratory of normal tissue and various neoplasms in the histologic differential diagnosis
of malignant melanoma have supported this finding,
indicating that this antibody has a staining distribution similar to HMB45.14,15 This antibody has been
shown to be an important marker for the identification
of MM in FNA material, particularly in patients best
suited to antigen specific vaccine therapies.10 Preliminary studies have reported MART 1 positivity in 92%
of fine-needle aspirations performed on documented
malignant melanomas.16 In response to these initial
findings supporting the usefulness of MART-1, we attempted to extend its application to effusions containing MM.
Our objective was to examine in detail the cytomorphologic and immunocytochemical characteristics of MM in effusions. We employed a panel of antibodies (including HMB45, S-100, cytokeratins, and
MART-1) that were optimal for discriminating MM
from a mesothelial or epithelial etiology, and we correlated the staining patterns.
MATERIALS AND METHODS
Thirty-two malignant peritoneal and pleural effusions
from patients with a documented diagnosis of metastatic MM, diagnosed in the Cytopathology Section of
NIH between 1986 and 1995, comprised this study.
The cases consisted of air-dried cytospins and smears
stained with Diff-Quikt (DQ) (Baxter, McGaw Park,
IL); EtOH-fixed, Papanicolaou-stained smears; and hematoxylin and eosin –stained cell block material.
In our laboratory, all fluid samples are triaged to
optimize specimen preparation. Thus, as part of the
workup of the cases involved in this study, bloody and/
or inflammatory specimens were concentrated and
lysed with ACK lysing buffer (Biofluids Inc., Rockville,
MD) or lymphoprepped (Accu-Prep, AN-551, Accurate
Chemical and Scientific Corp., Westbury, NY) to enhance the concentration of malignant cells and de-
/ 7302$$935d
01-25-97 07:41:47
crease the background blood and/or inflammatory elements.
The following general cytologic features were examined: degree of cellular cohesion of malignant cells,
percentage of malignant cells in the effusion, and
background cellular components. The sizes of the malignant cells were compared with those of the reactive
mesothelial cells in each specimen. We also examined
nuclear features, including size, shape, and placement
within the cytoplasm, as well as the presence of
nucleoli, cytoplasmic intranuclear pseudoinclusions,
mitoses, and multinucleation. The following cytoplasmic characteristics were noted: cell-in-cell engulfment; cytoplasmic volume and staining pattern;
vacuolization; and pigmentation, which we defined as
fine granules staining brown to navy blue on Diff-Quik
stains. Fontana-Masson staining was performed in selected cases to confirm the presence of melanin.
Of the 32 cases reviewed, 30 cases were also studied with immunocytochemistry. Material used for
these studies included paraffin embedded cell blocks
(26 cases, 5 micron sections on charged slides, Fisher
Scientific, Pittsburgh, PA) and cytospins (4 cases).
Antibodies used for the immunocytochemical
analysis were HMB45 (1:10 dilution, BioGenex Laboratories, San Ramon, CA), S-100 protein (1:1dilution, Biogenex Laboratories), AE1/AE3 (1:100 dilution, Boehringer Mannheim, Indianapolis, IN), and MART-1
(clone M2-7C10, supernatant 1:10 dilution). The secondary antibody used was affinity-purified, biotin-labeled goat antimouse immunoglobulin (Ig)G (1:1000
dilution, Kirkegaard and Perry Laboratories, Inc.,
Gaithersberg, MD). Immunocytochemical studies
were performed using a modified avidin-biotin peroxidase technique (Vector Laboratories, Burlingame CA),
with 3-3’ diaminobenzidine (Sigma Chemical Co., St.
Louis, MO) used as the chromogen, and a hematoxylin
counterstain. A red chromogen, 3-amino-9 ethylcarbazole, was used to aid in interpretation in one case due
to heavy cytoplasmic pigmentation. Slides designated
for AE1/AE3 analysis were subject to a 1% trypsin digestion prior to immunostaining.
All paraffin embedded samples requiring staining
with MART-1 were subjected to microwave pretreatment.
The microwave-based antigen retrieval consisted of placing the deparaffinized slides in 1500mL 10mM citrate
buffer, pH 6.0, and microwaving in an 800-watt oven for
40 minutes. The slides were then allowed to cool for 30
minutes and rinsed in distilled water. After cooling, a
modified avidin-biotin procedure was followed.
For a negative control, we used purified myeloma
protein, mouse IgG1 kappa (Organon Teknika Corp., Durham, NC).
ccyta
W: Can Cyto
Effusion Cytology of Malignant Melanoma/Beaty et al.
59
TABLE 1
Morphologic Features of 32 Eff
Cellular features
Cellularity
Discohesion
Clusters (rare)
Background cells
Mesothelial cells
Histiocytes
Blood
Lymphocytes
Neutrophils
Nuclei
Eccentric nuclei
Prominent nucleoli
Multinucleation
Pseudoinclusions
Mitoses (rare)
Cytoplasm
Vacuolization
Small
Mixed
Large, signet-ring
Pigment
Diff-Quikt
Fontana-Masson
Cell-in-cell engulfment
Biphasic staining
High N:C ratio
Present
Absent
% of cases
32
7
0
25
100%
22%
32
8
22
13
14
0
24
10
19
18
100%
25%
69%
41%
44%
32
32
27
8
5
0
0
5
24
27
100%
100%
84%
25%
16%
24
21
2
1
23
22
23
15
4
1
8
11
30
31
9
10
9
17
28
31
75%
66%
6%
3%
72%
69%
72%
47%
12.5%
3
FIGURE 1. The malignant cells shown exhibit cell-in-cell engulfment, an
unexpected feature present in 47% of the cases in this study (Papanicolaou
stain, 11000).
N:C ratio: nuclear to cytoplasmic ratio.
RESULTS
We studied 32 cases of MM diagnosed in pleural effusions (23 cases) and peritoneal effusions (9 cases) from
27 patients (10 female, 17 male) who ranged in age
from 15 to 66 years.
FIGURE 2.
Multinucleated tumor cells such as those shown were a
consistent feature, present in 84 % of the cases in this study. (Diff-Quikt,
11000).
Cytomorphology
In the morphologic review, emphasis was placed on
DQ-stained material, due to its enhanced cytoplasmic
volume and detail (Table 1). The malignant cells were
largely discohesive, with cluster formation present in
only seven cases. The background cellularity consisted
predominantly of reactive mesothelial cells and scattered histiocytes. A majority of the samples contained
blood and various inflammatory components. An unexpected feature was the presence of cell-in-cell engulfment, which was seen in 15 cases (Fig. 1).
MM cells were round to oval in shape and showed
an extreme variability in size, ranging from one to five
times the size of background mesothelial cells. Cytoplasm was invariably abundant and usually vacuolated: 21 cases had a fine, evenly distributed vacuolization; 2 cases showed both large and small vacuoles;
/ 7302$$935d
01-25-97 07:41:47
and 1 case had very large, ‘‘signet ring’’ – type cytoplasmic vacuolization. Eight cases showed no vacuoles. In 4 cases, the DQ-stained tumor cells exhibited
a biphasic staining pattern, that is, an inner, darker,
basophilic staining with an outer rim of magenta cytoplasm. The morphology of the MM cells in these cases
was almost indistinguishable from that of the background reactive mesothelial cell population.
The presence of cytoplasmic melanin pigment was
variable. In only two cases was the pigment a prominent and diagnostic feature. The remaining pigmented
cases showed variable and patchy pigmentation
within the neoplastic cells and associated histiocytes.
Melanin pigment appeared as fine, black-to-dark-blue
cytoplasmic granules in DQ-stained material. Fon-
ccyta
W: Can Cyto
60
CANCER (CANCER CYTOPATHOLOGY) February 25, 1997 / Volume 81 / Number 1
TABLE 2
Review of Immunocytochemical Stains
No. of cases
Antibody
Positive
Negative
Total
% of cases
MART-1
HMB45
S-100
Cytokeratin
Single Expression
MART-1
HMB45
S100
Cytokeratin
Coexpression
(MART-1 /
HMB45 /
S-100)
Nonimmunoreactive
No tumor cells in
sample
analyzed
21
22
22
3
5
1
2
2
0
6
5
5
24
—
4
3
3
0
27
27
27
27
27
5
5
5
5
78%
81%
81%
11%
19%
20%
40%
40%
0%
17
—
—
—
27
2
63%
7%
—
—
1
3%
tana-Masson stain, used to confirm the DQ interpretation of pigment, identified one additional pigmented
case originally interpreted as amelanotic with DQ.
The nuclei of the malignant cells were large, round
to oval in shape, eccentrically placed within the cell,
and contained smooth, sharply-demarcated membrane contours. Only rare malignant cells showed central nuclei and/or irregular nuclear contours. An increase in the nuclear to cytoplasmic (N:C) ratio of the
malignant population was not noted, as the amount
of the cytoplasm appeared to be proportional to the
nuclear size in most instances. A high N:C ratio was a
prominent feature in only one case.
In general, the nuclei were hyperchromatic, containing single or multiple, prominent to macro ( at
least 1/3 the diameter of the nucleus) nucleoli. The
nucleoli exhibited irregular shapes and contours.
Multinucleation was a consistent feature, present in
27 cases (Fig. 2). Cytoplasmic intranuclear pseudoinclusions were present in only eight cases. Notably, only
rare mitoses were identified in 5 of the 32 cases examined; 27 cases showed no mitotic activity.
Immunocytochemistry
Thirty cases had sufficient material for IM (Table 2).
Two cases were considered nonimmunoreactive when
no cells, including background mesothelial cells, were
positive for any of the four antibodies. In addition,
one case contained no tumor cells within the cell block
material, although background mesothelial cells
stained appropriately for cytokeratins. Thus, 27 cases
/ 7302$$935d
01-25-97 07:41:47
were considered appropriate for IM analysis. MM cells
were positive for the MART-1 antigen in 21 (78%) of the
27 interpretable cases, showing a cytoplasmic staining
pattern; 15 of these 21 cases (71%) showed strong cytoplasmic staining. Immunostaining for S-100 protein
and HMB45 was positive in 22 (81%) of the 27 interpretable cases. Strong cytoplasmic staining was observed in 19 (86%) and 17 (77%) of the cases that were
positive for S-100 and HMB45, respectively. Cytokeratin positivity in the tumor cells was present in 3 cases,
whereas mesothelial cells stained for cytokeratins in
all 28 immunoreactive samples.
Single antibody expression for only 1 of the melanoma markers (MART-1, HMB45, and S-100) was present in a total of 5 cases; HMB45 and S-100 positivity
was observed in two cases each. One case, which was
nonpigmented but diagnostic for MM on morphologic
grounds, stained positively for MART-1 alone. Coexpression of the three antibodies was observed in 17
cases (63%) (Fig. 3). Thus, the MART-1 antibody displayed a staining distribution similar to that of S-100
and HMB45.
DISCUSSION
We analyzed 32 malignant effusions from patients
known to have MM. In general, due to the enhanced
cytoplasmic volume secondary to air-drying, we preferred to perform diagnostic interpretations on DQstained cytospins. Triaging specimens appropriately
greatly increased the relative tumor cellularity, particularly in paucicellular specimens.
The differential diagnosis of MM in an effusion
includes distinguishing the malignant cells from reactive mesothelial cells as well as from adenocarcinoma.
Classic discohesive, large, multinucleated cells with
prominent nucleoli, intranuclear cytoplasmic pseudoinclusions, and abundant melanin pigment generally
do not pose a diagnostic challenge. However, these
cellular features are represented in a minority of
cases.1,5 Indeed, many of the cellular features consistently present in the cases we studied, including discohesion (100%), multinucleation (84%), prominent
nucleoli (100%), and cytoplasmic vacuolization (75%),
can be identified in adenocarcinomas and malignant
mesotheliomas.17– 19 It is noteworthy that cell-in-cell
engulfment, typically considered a morphologic indication of malignancy in mesothelial and epithelial malignant effusions, was present in 47% of the cases we
reviewed.19
The morphology of malignant cells can also resemble normal reactive mesothelial cells. In 4 (12.5%)
of the cases we studied, the malignant cells were virtually cytologically indistinguishable from the background reactive mesothelial cell population.
ccyta
W: Can Cyto
Effusion Cytology of Malignant Melanoma/Beaty et al.
FIGURE 3.
61
Immunocytochemical stains on cell block sections of the
same case show strong reactivity and coexpression of three melanoma
antibodies: (A) HMB45, (B) S-100, and (C) MART-1 (Hematoxylin and
diaminobenzidine, 11000).
Immunocytochemical studies are imperative for narrowing the diagnostic possibilities when specimens that
show overlapping cytologic features are being examined,
particularly when no pigment is shown to be present with
DQ or Fontana-Masson stains, as frequently occurs in
metastases from melanomas.5 The use of a panel of antibodies is a well-established practice in the diagnosis of
MM.7 Although the use of immunohistochemistry in cytologic specimens has previously been questioned, in our
laboratory it is routinely—and effectively—employed for
making specific cytologic diagnoses. Recent studies support this contention.20,21
We analyzed immunoreactivity by using a panel of
antibodies limited to cytokeratins AE1/AE3, HMB45, S100, and a new melanoma marker, MART-1. Vimentin,
while positive in melanomas, is also positive in mesotheliomas and many adenocarcinomas; therefore, this antibody was not chosen for use in our panel.22,23 In general,
immunocytochemical interpretation was difficult in cases
with very few tumor cells. However, when specimens
were triaged appropriately, in most instances we were
able to increase the concentration of tumor cells. We also
observed an improvement in the quality of the immunocytochemical stains. This improvement was presumably
the result of the removal of inflammatory cells that may
/ 7302$$935d
01-25-97 07:41:47
have taken up stains nonspecifically, giving rise to high
background staining and potentially making interpretation difficult. Of the 30 cases stained, 3 cases were either
nonimmunoreactive or contained no tumor cells within
the paraffin embedded cell block material.
Initial studies have shown that the MART-1 antibody
gives strong staining in fresh cytospin and frozen section
material.10 In paraffin embedded material, staining for
MART-1 is enhanced by the use of microwave antigen
retrieval.10,16 We observed a 78% positivity for the MART1 antigen. Staining was highly specific for the malignant
cells, with no staining of the background cells in any
positive case. MART-1 staining was generally as strong as
that of HMB45 and S-100. In cases where cytospin material was used, however, MART-1 showed more intense
staining, corroborating earlier findings.10 The preliminary
data, including application of this antibody in other normal and neoplastic processes, indicate that MART- 1 is a
highly sensitive and specific marker for MM, having staining characteristics similar to those of HMB45.10,14,15 The
results of this study suggest that the MART-1 antibody is
highly specific, with a sensitivity approximating that of
HMB45 and S-100 to MM cells in effusions.
The presence of cytoplasmic melanin pigment is generally thought to be diagnostic. It is easily identified on
ccyta
W: Can Cyto
62
CANCER (CANCER CYTOPATHOLOGY) February 25, 1997 / Volume 81 / Number 1
DQ stains, its appearance ranging from varying amounts
of fine navy blue cytoplasmic stippling to obvious dark
pigment granules, and it is confirmed with Fontana-Masson stains.In our study, 9 cases (28%) showed no evidence
of pigmentation. IM staining patterns in the pigmented
(P) and nonpigmented (NP) melanomas showed some
variability. The percentage of cases that stained with
HMB45 was the same (78%) for both P and NP melanomas, whereas a higher percentage of pigmented cases
showed staining for S-100 (89% P vs. 71% NP) and MART1 (74% P vs. 67% NP).
Overall, the antibodies stained every specimen containing immunoreactive tumor cells and confirmed the
diagnosis of malignant melanoma in each case, including
a case originally diagnosed as ‘‘poorly differentiated tumor’’ and two of four cases originally diagnosed as ‘‘suspicious for involvement by melanoma.’’ Due to the ambiguous morphologic features of these cases, IM stains were
helpful in making a definitive diagnosis. In the two cases
in which IM was not available, the diagnosis of MM was
made based on morphologic review by two of the authors
(M.W.B., A.A.).
We used the cytokeratin antibody to help differentiate tumor cells morphologically indistinguishable from
reactive mesothelial cells and/or adenocarcinoma. However, cytokeratin-positive malignant cells were present in
3 of 27 cases (11%). Two of the three cases showed very
weak staining. In these cases, the presence of staining
for the other markers was helpful in distinguishing the
cytokeratin positive cells from the reactive mesothelial
cells. Cytokeratin has a reported positivity rate in MM
ranging from 1% to 27%, depending on the fixation of the
specimen. Generally, frozen and ethanol fixed specimens
have a higher positivity than formalin fixed materials.8,24
In summary, the cytomorphologic characteristics of
MM in effusions vary and resemble other malignancies
or even benign reactive processes. Melanin pigment is
absent in many cases, thus making immunocytochemical
analysis extremely important in confirming the diagnosis.
We support the use of a panel of antibodies, which should
include the cytokeratins, HMB45, and S-100 on cell block
material.
Our experience suggests that the MART-1 antibody
may be a useful adjunct to the above panel, particularly
when employed on fresh cytospins or paraffin embedded
material after microwave antigen retrieval. Current studies are now underway to explore more fully the utility of
the MART-1 antibody in malignant melanomas in patients chosen for antigen specific vaccine therapies.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
REFERENCES
1.
2.
Hajdu SI, Savino A. Cytologic diagnosis of malignant melanoma. Acta Cytol 1973;17:320 – 7.
Nakleh RE, Wick MR, Rocamora A, Swansen PE, Dehner LP.
/ 7302$$935d
01-25-97 07:41:47
18.
ccyta
Morphologic diversity in malignant melanomas. Am J Clin
Pathol 1990; 93:731 –40.
Niemann TH, Thomas PA. Melanoma with signet-ring cells
in a peritoneal effusion. Diagn Cytopathol 1995; 12:241 –4.
Sears D, Hajdu SI. The cytologic diagnosis of malignant neoplasms in pleural and peritoneal effusions. Acta Cytol
1987; 31:85– 97.
Walts AE. Malignant melanoma in effusions: a source of
false-negative cytodiagnosis. Diagn Cytopathol 1986; 2:150–
3.
Filho AL, de Carvalho LV, da Cunha Santos G, Oyafuso MS,
Lombardo V, Bortolan J, et al. Cytologic diagnosis of melanoma in serous effusions: a morphologic and immunocytochemical study. Acta Cytol 1995;39:481 – 4.
Angeli S, Koelma I, Fleuren GJ, Van Steenis GJ. Malignant
melanoma in fine needle aspirates and effusions: an immunocytochemical study using monoclonal antibodies. Acta
Cytol 1988;32:707 – 12.
Miettinen M, Franssila K. Immunohistochemical spectrum
of malignant melanoma: the common presence of keratins.
Lab Invest 1989; 61:623 – 8.
Ordoñez NG, Sneige N, Hickey RC, Brooks TE. Use of monoclonal antibody HMB-45 in the cytologic diagnosis of melanoma. Acta Cytol 1988;32:684 – 8.
Marincola FM, Hijazi YM, Fetsch P, Salgaller ML, Rivoltini
L, Cormier J, et al. Analysis of expression of the melanomaassociated antigens MART-1 and gp100 in metastatic melanoma cell lines and in in situ lesions. J Immunother Emphasis Tumor Immunol 1996; 19:192 – 205.
Kawakami Y, Eliyahu S, Sakaguchi K, Robbins PF, Rivoltini
L, Yanelli JR, et al. Identification of the immunodominant
peptides of the MART-1 human melanoma antigen recognized by the majority of HLA-A2-restricted tumor infiltrating
lymphocytes. J Exp Med 1994; 180:347– 52.
Kawakami Y, Eliyahu S, Delgado CH, Robbins PF, Rivoltini
L, Topalian SL, et al. Cloning of the gene coding for a shared
human melanoma antigen recognized by autologous T-cells
infiltrating into tumor. Proc Natl Acad Sci USA 1994;
91:3515– 9.
Kawakami Y, Eliyahu S, Delgado CH, Robbins PF, Sakaguchi
K, Appella E, et al. Identification of a human melanoma
antigen recognized by tumor infiltrating lymphocytes associated with in-vivo tumor rejection. Proc Natl Acad Sci USA
1994; 91:6458–62.
Fetsch PA, Kleiner D, Marincola F, Abati A. Analysis of melanoma-associated antigen, MART-1, in normal tissues and in
selected non-melanomatous neoplasms [abstract presented
at the American and Canadian Academy of Pathology,
March 1997, Orlando, Florida.].
Fetsch PA, Fetsch J, Travis W, Batts K, Abati A. Comparison of
melanoma-associated membrane antigen, MART-1, to HMB45: further evidence to support a common lineage for angiomyolipoma, lymphangiomyomatosis, and clear cell ‘‘sugar tumor’’
[abstract presented at the American and Canadian Academy of
Pathology, March 1997, Orlando, Florida].
Fetsch PA, Cormier J, Hijazi YM. Immunocytochemical detection of MART-1 in fresh and paraffin-embedded malignant melanomas. J Immunother Emphasis Tumor Immunol. In press.
Bottles K, Reznicek MJ, Holly EA, Ahn DK, Layfield LJ, Cohen
MB. Cytologic criteria used to diagnosis adenocarcinoma in
pleural effusions. Mod Pathol 1991; 4:677 – 81.
Spriggs AI, Grunze H. An unusual cytologic presentation of mesothelioma in serous effusions. Acta Cytol 1983;27:288–92.
W: Can Cyto
Effusion Cytology of Malignant Melanoma/Beaty et al.
19. Stevens MW, Leong ASY, Fazzalari NL, Dowling KD, Henderson
DW. Cytopathology of malignant mesothelioma: a stepwise logistic regression analysis. Diagn Cytopathol 1991;8:333–41.
20. Shield PW, Perkins G, Wright RG. Immuocytochemical staining of cytologic specimens: how helpful is it? Am J Clin Pathol 1996; 105:157– 62.
21. Leong ASY. Immunostaining of cytologic specimens [editorial]. Am J Clin Pathol 1996;105:139 –40.
22. Motoyama T, Watanabe T, Okazaki E, Tanaka N, Watanabe
H. Immunohistochemical properties of malignant mesothe-
/ 7302$$935d
01-25-97 07:41:47
63
lioma cells in histologic and cytologic specimens. Acta Cytol
1995; 39:164 – 70.
23. Duggan MA, Masters CB, Alexander F. Immunohistochemical differentiation of malignant mesothelioma, mesothelial
hyperplasia, and metastatic adencarcinoma in serous effusions, utilizing carcinoembryonic antigen, keratin, and vimentin. Acta Cytol 1987; 31:807 – 14.
24. Banks ER, Jansen JF, Oberle E, Davey DD. Cytokeratin positivity in fine needle aspirates of melanomas and sarcomas.
Diagn Cytopathol 1995; 12:230 –3.
ccyta
W: Can Cyto
Документ
Категория
Без категории
Просмотров
2
Размер файла
316 Кб
Теги
772
1/--страниц
Пожаловаться на содержимое документа