| | Expression of Fas and FasL in human serous ovarian epithelial tumors☆Accepted 26 September 2002. Abstract The expression of Fas and FasL was studied in 86 patients with benign, borderline, and malignant serous ovarian lesions. Four normal ovaries, and monolayer epithelial cultures from a human fetal ovary, a borderline, and a serous adenocarcinoma were used for comparison. Expression of Fas and FasL was studied immunohistochemically and flowcytometrically. Fas was expressed in all 90 lesions; FasL in 57 lesions, including 2 normal ovaries. Fas expression was significantly increased in borderline tumors compared with benign (P = 0.005, t = −2.94) or malignant serous tumors (P = 0.0001, t = 4.15). FasL expression was significantly increased in malignant tumors compared with benign (P = 0.039, t = −2.10) and borderline tumors (P = 0.0016, t = −3.33). Flow cytometry showed a range of Fas expression in short-term cultures isolated from normal, borderline, and malignant ovarian serous tissue; in the few samples studied, FasL was not expressed. Expression in three serous ovarian cell lines was similar. Fas and FasL expression differed throughout the spectrum of ovarian lesions. FasL expression was increased in malignant tumors, and Fas expression was increased in borderline tumors. Changes in Fas/FasL expression in ovarian surface epithelium might play a functional role in the biology of ovarian tumors. HUM PATHOL 34:74-79. Copyright 2003, Elsevier Science (USA). All rights reserved.
Disturbances of apoptotic processes, caused by mutations of the controlling genes, play a central role in the pathogenesis of human disease.1 Cytotoxic drugs used in therapy of malignant tumors cause apoptosis (programmed cell death) and necrosis in target cells.2, 3 The best-characterized “cell death” receptor with respect to its signal transduction pathway is the Fas (Apo-1/CD95) protein. This M 45,000 type I transmembrane cell surface receptor of the tumor necrosis factor (TNF)/nerve growth factor superfamily mediates apoptosis when triggered by antibodies or its ligand, FasL. This ligand FasL/CD95L, a type II transmembrane molecule of M 40,000, is expressed on cell membranes or in soluble form and belongs to a coevolved family of proteins of the TNF family.4, 5, 6
Fas is expressed by various cells of lymphoid and nonlymphoid origin, including those of normal follicles in the ovaries.7 In the process of follicular atresia, apoptosis eliminates thecal and granulosa cells that compose the follicle.8, 9 It also has been shown that ovarian surface epithelial cells of dispersed mouse corpora lutea undergo programmed cell death before ovulation. Rapid repair of the surface cells occurs after ovulation.10
FasL expression in normal tissue is largely restricted to T lymphocytes and macrophage lineages. The Fas/FasL system's primary function is thought to be maintenance of homeostasis in the immune system.11 Additionally, cytotoxic T lymphocytes (CTLs) and natural killer (NK) cells induce apoptosis in their targets such as tumor cells via Fas/FasL engagement.4, 12, 13 Recent reports have shown that many tumors are able to down-regulate Fas expression, thereby avoiding FasL-mediated apoptosis by CTLs and NK cells. In addition, many tumors up-regulate the expression of FasL, which binds to the Fas present on CTLs, triggering apoptosis of CTLs and preventing CTL recruitment in human tumors.14, 15 It is clear that Fas and FasL expression may be important determinants of tumor behavior in vivo.
Of the malignant epithelial ovarian tumors, which account for >90% of all ovarian cancers, the serous papillary variants form the largest subgroup. The biological spectrum of ovarian epithelial tumors includes benign, proliferating (borderline), and malignant neoplasms.16, 17
In the present study, expression of Fas and FasL in normal ovarian epithelium and within the biological spectrum of benign through frankly malignant serous lesions was examined by immunohistochemistry in 90 patients. Fas and FasL expression was also studied by flow cytometry in cell cultures of epithelial cells from normal, borderline, and malignant ovarian tissue and serous carcinoma cell lines.
Materials and methods  Ovarian tissue Fixed and paraffin-embedded tissue from 86 patients with ovarian serous surface epithelial tumors and 4 patients with normal ovaries was obtained from files of the Department of Anatomical Pathology, Royal Prince Alfred Hospital, Sydney, Australia. Thirty patients had benign tumors, 22 patients had histologically proven borderline tumors, and 34 patients had invasive cancer. Primary fixation was in 0.1 M phosphate-buffered (pH 7.0) formalin 10% (3.6% formaldehyde) up to 1985 or formaldehyde/acetic-acid/alcohol after 1985. Sections (4 μm thick) were cut and mounted on silanized glass slides. The surface epithelial tumors were typed and categorized according to the current protocols of the World Health Organization and the International Society of Gynecological Pathologists16 by 2 investigators not involved in the original diagnosis to ensure consistency throughout the study material. Although tumors of borderline malignancy had been clinically divided into low-grade and high-grade variants (according to the severity of architectural and cellular changes, mitotic activity, and nuclear atypia), for this study lesions were combined into a single group, mainly because of the limited number of cases. One slide for each case, considered representative for the lesion as a whole, was selected by agreement. In cases with manifest heterogeneity, the slide with the most atypical area or highest grade of abnormality was chosen. Cell cultures Three serous carcinoma cell lines (JV, JC, and JG)18 were grown in RPMI-1640 (Trace Biosciences Sydney, Australia), supplemented with 20 mmol HEPES, 6 mmol L-glutamine, 20 μg mL-1 garamycin, and 10% fetal bovine serum (FBS) (Trace Biosciences). When near confluent, the cell lines were studied by flow cytometry for Fas and FasL expression. The human lymphoma cell line BJAB (Centenary Institute, Sydney, Australia) was used as a positive control of Fas expression. The cells were grown as suspension cultures in 25-cm2 flasks. Monolayer epithelial cultures were established from fresh tissue samples, selected in theatre by the pathologist to be representative of the lesion. Primary samples of 2 ovarian borderline tumors19 and 2 serous adenocarcinoma were cut into <1-mm-diameter fragments using sterile knifes with regular washings of sterile medium in a laminar flow cabinet. These fragments were seeded into 25-cm2 flasks to establish epithelial cell cultures. Mesothelial cells and fibroblasts were mechanically depleted under visualization through an Olympus JMT inverted microscope (Olympus Corp, Lake Success, NY). The monolayer cultures were harvested after 3 weeks to study Fas and FasL expression. Two fetal tissue samples were obtained with written informed consent and permission of the Ethical Review Board of the host institute and tissue cultures established in a fashion identical to that for neoplastic tissue. Antibodies Antibodies and fluorochrome were purchased from PharMingen (San Diego, CA) for flow cytometry. Monoclonal mouse IgG1 anti-human Fas (DX2) and monoclonal mouse IgG1 anti-human FasL (NOK-1) were used. Isotype-matched biotinylated mouse IgG1 was used as a negative control. Secondary labeling was achieved by use of streptavidin-phycoerythrin. Detection of Fas and FasL in fixed, paraffin-embedded tissues was done using immunostaining kits obtained from Santa Cruz Biotechnology (Santa Cruz, CA). These kits contain prediluted rabbit polyclonal antisera to Fas (C20) (sc 715K; Santa Cruz Biotechnology) and FasL (N20) (sc 834K; Santa Cruz Biotechnology), and a biotin-streptavidin-peroxidase detection system. Negative control sera (sc 715P and sc 834P) were used at the same protein concentrations as the primary antisera. Immunohistochemistry Paraffin sections (4 μ thick) were floated to silanized slides and dried at 65°C for 30 minutes. Before immunostaining, sections were subjected to heat-induced epitope retrieval (HIER). Dewaxed, rehydrated sections were immersed in an EDTA Tris-citrate antigen-retrieval medium, pH 8.0, and boiled in a microwave oven for 10 minutes. After cooling to room temperature, the slides were washed in running water and rinsed in Tris-buffered saline (pH 7.6), and the immunostaining procedure was performed following the manufacturer's instructions. Peroxidase activity was demonstrated using diaminobenzidine, and the sections were counterstained using Harris' hematoxylin. Grading of lesions and analysis of antigen expression The tumors were classified according to the proportion of positive cells in the epithelium of the lesions as follows: ++++, >75% of cells staining positive; +++, 50% to 75% of cells staining positive; ++, 25% to 50% of cells staining positive; +, <25% cells staining positive; ±, <10% of cells staining positive; −, no cells staining positive.20 With blinding to histopathologic classification and original diagnosis and after repeated randomization, all specimens were graded 3 times by 2 independent investigators. Results were recorded for analysis of intraobserver and interobserver reproducibility of classification of lesions into each of the 6 classes used. Flow cytometry Cells from each cell line, grown in 75-cm2 flasks, were trypsinized in 0.05% trypsin/0.02% EDTA and washed twice in phosphate-buffered saline. Then 106 cells were labeled with 20 μL biotin-conjugated primary antibody (1 μg/sample) anti-human Fas and anti-human FasL, at 4°C for 45 minutes. Negative control cells were labeled in the same way with a biotin-conjugated mouse1 IgG1 (as discussed earlier). As a positive control for Fas, 106 BJAB cells were labeled with DX2. After incubation, the pelleted cells were centrifuged through fetal bovine serum and resuspended in a 1 μg/sample with streptavidin-phycoerythrin conjugate for 45 minutes at 4°C. The cells were then centrifuged through fetal bovine serum and resuspended in 300 μL of medium for flow cytometry. Fluorescence was measured using a FACSCalibur Sort (Becton-Dickinson, Mountainview, CA) with an argon laser set at 488 nm; 5×104 cells were analyzed from each labeled sample. Dead cells were excluded from analysis on the basis of low forward scatter (cell size) and low side scatter (membrane configuration) and propidium iodide staining. Data analysis was performed using the CellQuest computer program; the results are presented as histograms or line graphs. Results from individual experiments are presented in histograms as event number versus log-10 fluorescence intensity. For representation as line graphs, mean fluorescence was calculated and used as a measure of relative cell surface molecule expression. Data were normalized against the isotype control fluorescence and amalgamated from at least 3 experiments, +/− the standard error of the mean as an indication of cell surface molecule expression. Statistical analysis For all statistical comparisons of 2 groups, an unpaired, 2-tailed Student t test (assuming equal variances) was used. A P value <0.05 was considered significant. Results Immunohistochemistry Expression of Fas and FasL in normal ovarian epithelial cells and serous ovarian tumors is detailed in Table 1.
Intraobserver and interobserver reproducibility for classification of lesions based on fraction of positive epithelial cells was >95%. No cases were found with disagreement (intraobserver or interobserver) greater than 1 class. In cases where differences occurred, a case was classified according to the dominant class chosen. No differences were noted between lesions fixed in classic buffered formaldehyde or formaldehyde/acetic-acid/alcohol. | | |  | | Fas | FasL | |
|---|
| Tissue/tumour (patient number) | − | ± | + | ++ | +++ | ++++ | − | ± | + | ++ | +++ | ++++ | |
|---|
| Normal (4) | | 2 | | 1 | 1 | | 2 | 1 | 1 | | | | |
| Benign (30) | | 1 | 10 | 14 | 5 | | 10 | 12 | 7 | 1 | | | |
| Borderline (22) | | | 3 | 9 | 7 (a) | 3 | 12 | 7 | 3 | | | | |
| • Low-grade (18) | | | 2 | 8 | 5 | 3 | 9 | 6 | 3 | | | | |
| • High-grade (4) | | | 1 | 1 | 2 | | 3 | 1 | | | | | |
| Malignant (34) | | 1 | 18 | 11 | 4 | | 9 | 7 | 10 | 8 (b) | | | |
| ++++, >75% cells stain, +++, 50% to 75% cells stain; ++, 25% to 50% cells stain; +, < 25% cells stain; ±, staining of some cells (<10%); −, no cells stain. (a) Fas expression in tissue from patients with borderline tumours was significantly different from that in both benign tumours (Student t test, P = 0.005, t = −2.94) and malignant tumors (P = 0.0001, t = 4.15). (b) FasL expression in tissue from patients with malignant tumors was significantly different from that in benign tumors (Student t test, P = 0.039, t = −2.10) and borderline tumors (P = 0.0016, t = −3.33). | | | | |
The fraction of epithelial cells positive for each of the markers differed significantly between individual tumors. All benign tissues expressed Fas varying between cases from + to +++, whereas FasL was expressed in 20 of these specimens (12 patients ±, 7 patients +, 1 patient ++). The study originally included 4 cases of borderline lesions with a greater degree of abnormality. As is evident from Table 1, this group is too small to allow separate interpretation of findings, and thus results were pooled for all lesions of borderline type. The epithelium in all 22 tumors of borderline malignancy expressed Fas (Fig 1), and 10 expressed FasL (7 patients ±, 3 patients +) (Fig 2).
The epithelium of all malignant tumors (34) expressed Fas, whereas epithelium in 25 specimens showed positive staining for FasL, of which 8 stained positive for 25% to 50% of the cells. Epithelium of all 4 normal ovaries expressed Fas, and 2 of 4 expressed FasL. In total, FasL was expressed in 57 of 90 ovarian specimens with 30% of cases showing some cells staining (±), 23% of cases with <25% cells staining (+), and 10% of cases with 25% to 50% cells staining (++). In total, the epithelium of 90 ovaries expressed Fas; in 34% of cases the epithelium stained positive in <25% of the cells (+), in 38% of cases the epithelium stained positive in 25% to 50% of the cells (++), and in 17% of cases the epithelium stained positive in 50% to 75% of the cells (+++). Only 3 borderline ovarian tumors had Fas expression in >75% of the epithelial cells (++++). Statistical analysis of Fas and FasL expression was performed using unpaired, 2-tailed Student t tests (assuming equal variances). Significantly increased Fas expression was found in borderline tumors compared with normal or benign epithelial disease as well as with malignant tumors. FasL expression was significantly increased in malignant tumors (Table 1) compared with the others. The correlation of Fas and FasL from the individual values for expression in each patient in each category (Table 1) was plotted (Fig 3).
The correlation coefficients for the linear fits for the benign, borderline, and malignant groups were 0.407, 0.495, and 0.071, respectively. Thus the pattern of correlation differed in the 3 groups, with increasing Fas expression tending to correlate with increased FasL expression in benign and borderline tumors. However, there is clearly no correlation between Fas and FasL expression in malignant tumors. This is notwithstanding the fact that FasL expression is significantly increased in its own right compared with the other groups. Flow cytometry Cell surface expression of Fas and FasL was measured using a FACSCalibur sort. Cultures of a human fetal ovary and a borderline tumor as well as a malignant tumor all stained for Fas but not for FasL. A similar finding was obtained in the 3 cell lines JV, JC, and JG (Fig 4).
Discussion Apoptosis, caused by physiologic and pathologic stimuli, can be induced by receptor mechanisms. The so-called “death ”receptors that initiate apoptosis are the TNF receptor system and the Fas receptor.1, 21 Fas and FasL expression, considered to reflect receptor activation, were investigated in a spectrum of ovarian tissue from normal epithelium, through nonneoplastic precursor lesions to high-grade malignant serous ovarian tumors. Consistent with the view that apoptosis occurs naturally in most tissues, expression of Fas was found in all grades of serous tumor tissue, as well as in normal ovarian surface epithelium. In our study, Fas expression was significantly greater in borderline serous tumors than in either benign or malignant serous tumors, whereas FasL expression appeared similar in benign and borderline tumors but significantly increased in malignant serous tumors. An earlier study by Leithauser et al7 reported staining of APO-1/Fas in thecal cells of follicles of the ovary, whereas studies on follicular atresia provided strong evidence of the role of Fas in programmed cell death.8, 9 Although Zusman et al22 reported that they did not find (although not an explicit part of the reported study) apoptotic activity in the normal ovary, we in fact found such activity in all of the 4 normal ovaries. In the light of reports of apoptotic activity under explicit nonmalignant conditions in organs and structures as diverse as the eye23 and Sertoli cells,24 and the limited number of cases studied, this apparent discrepancy warrants further study in larger group. Expression of Fas and FasL in benign ovarian tumors at high levels was also found by Zusman et al,22 whereas Munakata et al25 found only 1 patient with some staining for FasL in a group of benign serous tumors. Twenty of 30 specimens of borderline tumors were positive for FasL in our study. However, these findings may not necessarily imply that (higher-grade) malignant neoplasms develop from lower-grade or benign lesions in cases of serous neoplasm of the ovary. At this stage, we have limited our analysis to the epithelial fraction of the lesions. Obviously, apoptosis occurs in the stroma of these lesions. Future studies on this aspect and on its relation (if any) with apoptotic activity of the epithelium are foreseen. Similarly, we have not at this stage included an assessment of the relationship between apoptotic activity in the epithelium and the presence or absence of stromal (T) lymphoid cells. Recently, it has been shown that many tumors down-regulate Fas expression, thereby impairing cellular recognition by CTLs and NK cells. In addition, new evidence has been found that tumors can up-regulate expression of FasL, which binds to the Fas present on CTLs, triggering apoptosis—the so-called “Fas counterattack”.13, 14 In our study, malignant tumors had increased FasL expression, and the relationship to Fas expression differed in these specimens from that in benign or borderline tumors. The dissociation between Fas and FasL expression in malignant tumors (Fig 3) clearly indicates altered Fas/FasL regulation in these cells. Expression of Fas in ovarian cancer cell lines26, 27, 28, 29 and FasL in ovarian tumor cells from ascitic fluid30 have previously been studied by flow cytometry. Rabinowich et al29 showed that ovarian ascitic cells expressing functional FasL were capable of triggering cell death in Fas-expressing T cells. In our study, Fas and FasL expression was measured in serous ovarian cancer cell lines as well as in short-term cultures of normal control and borderline ovarian cells. The short-term cell cultures and the 3 cell lines expressed Fas with different intensities but did not express FasL. It is appreciated that a correlation must exist between the fraction of cells undergoing apoptosis as visualized in our study and the number of final end-stage apoptotic cells (apoptotic bodies). With potential variability of the duration of apoptotic involution and the possibility of apoptotic arrest (in analogy to mitotic arrest), this relationship may be complex, but a study will nevertheless be informative. As such, this is considered within the priorities of our research group. In conclusion, our study has demonstrated changes in Fas/FasL expression in ovarian tumors, which may have relevance to tumor behavior. These features may modulate immune surveillance and tumor recognition by the cellular immune response, and they warrant further study. Acknowledgements We thank Anna Doubrovsky for her assistance with the statistical analysis. References 1.
1
Kam PCA, Ferch NI.
Apoptosis: Mechanisms and clinical implications (review article).
Anaesthesia. 2000;55:1081–1093. MEDLINE |
CrossRef
2.
2
Barry MA, Behnke CA, Eastman A.
Activation of programmed cell death (apoptosis) by cisplatin, other anticancer drugs, toxins and hyperthermia.
Biochem Pharmacol. 1990;40:2353–2362. MEDLINE |
CrossRef
3.
3
Hickman JA, Potten CS, Merritt AJ, et al.
Apoptosis and cancer chemotherapy.
Philos Trans R Soc Lond B Biol Sci. 1994;345:319–325. MEDLINE |
CrossRef
4.
4
Nagata S, Golstein P.
The Fas death factor.
Science. 1995;267:1449–1456. MEDLINE 5.
5
Suda T, Takahashi S, Golstein P, et al.
Molecular cloning and expression of the Fas ligand, a novel member of the tumor necrosis factor family.
Cell. 1993;75:1169–1178. MEDLINE |
CrossRef
6.
6
Itoh N, Yonehara S, Ishii A, et al.
The polypeptide encoded by cDNA for human cell surface antigen Fas can mediate apoptosis.
Cell. 1991;66:233–243. MEDLINE |
CrossRef
7.
7
Leithauser F, Dhein J, Mechtersheimer G, et al.
Constitutive and induced expression of APO-1 a new member of the nerve growth factor/tumor necrosis factor receptor superfamily, in normal and neoplastic cells.
Lab Invest. 1993;69:415–429. MEDLINE 8.
8
Quirk SM, Cowan RG, Joshi S, et al.
Fas antigen-mediated apoptosis in human granulosa/luteal cells.
Biol Reprod. 1995;52:279–287. MEDLINE |
CrossRef
9.
9
Foghi A, Ravandi A, Teerds KJ, et al.
Fas-induced apoptosis in rat thecal/interstitial cells signals through the sphingomyelin-ceramide pathway.
Endocrinology. 1998;139:2041–2047. MEDLINE |
CrossRef
10.
10
Quirk SM, Cowan RG, Huber SH.
Fas antigen–mediated apoptosis of ovarian surface epithelial cells.
Endocrinology. 1997;138:4558–4566. MEDLINE |
CrossRef
11.
11
Nagata S.
Apoptosis by death factor.
Cell. 1997;88:355–365. MEDLINE |
CrossRef
12.
12
Peters ME, Scaffidi C, Medema JP, et al.
The death receptors.
In:
Kumar S editors.
Apoptosis, Problems and Diseases. Heidelberg, Germany: Springer; 1998;. 13.
13
O'Connell J, Bennett MW, O'Sullivan GC, et al.
The Fas counterattack: Cancer as a site of immune privilege.
Immunol Today. 1999;20:46–52. MEDLINE |
CrossRef
14.
14
O'Connell J, O'Sullivan GC, Collins JK, et al.
The Fas counterattack: Fas-mediated T cell killing by colon cancer cells expressing Fas ligand.
J Exp Med. 1996;184:1075–1082. MEDLINE |
CrossRef
15.
15
O'Connell J, Bennett MW, O'Sullivan GC, et al.
Fas counterattack: The best form of tumor defense?.
Nat Med. 1999;5:267–268. MEDLINE |
CrossRef
16.
16
Russell P, Farnsworth A.
Surgical Pathology of the Ovaries.
ed 2. New York, NY: Churchill Livingstone; 1997;. 17.
17
Scully RE, Young RH, Clement PB.
Tumors of the ovary, maldeveloped gonads, fallopian tube, and broad ligament.
In:
Scully RE, Young RH, Clemente PB editor.
Atlas of Tumor Pathology, Third Series. Washington, DC: Armed Forces Institute of Pathology; 1998;. 18.
18
van Haaften-Day C, Russell P, Rugg C, et al.
Flow cytometric and morphological studies of ovarian carcinoma cell lines and xenografts.
Cancer Res. 1983;43:3725–3731. MEDLINE 19.
19
van Haaften-Day C, Russell P, Davies S, et al.
An in vitro study of ovarian atypical proliferating (borderline) serous tumors.
Int J Gynecol Cancer. 1992;2:41–48. 20.
20
van Haaften-Day C, Russell P, Boyer CM, et al.
Expression of cell regulatory van proteins in ovarian borderline tumors.
Cancer. 1996;77:2092–2098. 21.
21
Rudin CM, Thompson CB.
Apoptosis and disease: Regulation and clinical relevance of programmed cell death.
Ann Rev Med. 1997;48:267–281. MEDLINE |
CrossRef
22.
22
Zusman I, Gurevich P, Gurevich E, et al.
The immune system, apoptosis and apoptosis-related proteins in human ovarian tumours (a review).
Int J Oncol. 2001;18:965–972. MEDLINE 23.
23
Griffith TS, Brunner T, Fletscher SM, et al.
Fas ligand-induced apoptosis as a mechanism of immune privilege.
Science. 1995;270:1189–1192. MEDLINE 24.
24
Bellgrau D, Gold D, Selawry H, et al.
A role for CD95 ligand in preventing graft rejection.
Nature. 1995;377:630–632. MEDLINE |
CrossRef
25.
25
Munakata S, Enomoto T, Tsujimoto M, et al.
Expressions of Fas ligand and other apoptosia-related genes and their prognostic significance in epithelial ovarian neoplasms.
Br J Cancer. 2000;82:1446–1452. MEDLINE 26.
26
Wakahara Y, Nawa A, Okamoto T, et al.
Combination effect of anti-Fas antibody and chemotherapeutic drugs in ovarian cancer cells in vitro.
Oncology. 1997;54:48–54. 27.
27
Cai Z, Stancou R, Korner M, et al.
Impairment of Fas-antigen expression in adriamycin-resistant but not TNF-resistant MCF7 tumour cells.
Int J Cancer. 1996;68:535–546. MEDLINE |
CrossRef
28.
28
Uslu R, Jewett A, Bonavida B.
Sensitization of human ovarian tumor cells by subtoxic CDDP to anti-Fas antibody–mediated cytotoxicity and apoptosis.
Gynecol Oncol. 1996;62:282–291. MEDLINE |
CrossRef
29.
29
Baldwin RL, Tran H, Karlan BY.
Primary ovarian cancer cultures are resistant to Fas-mediated apoptosis.
Gynecol Oncol. 1999;74:265–271. MEDLINE |
CrossRef
30.
30
Rabinowich H, Reichert TE, Kashii Y, et al.
Lymphocyte apoptosis induced by Fas ligand–expressing ovarian carcinoma cells.
J Clin Invest. 1998;101:2579–2588. MEDLINE |
CrossRef
Department of Cancer Medicine and Pathology, University of Sydney, New South Wales, Australia and the Department of Anatomical Pathology, Royal Prince Alfred Hospital, New South Wales, Australia ☆ Address correspondence and reprint requests to Dr. Caroline van Haaften-Day, LCPL Laboratory, Rooseveltstraat 4C, 2321 BM, Leiden, The Netherlands. PII: S0046-8177(02)29707-5 doi:10.1053/hupa.2003.7 © 2003 Elsevier Science (USA). All rights reserved. | |
|
FULL TEXT29707-5/journalimage?img=PIIS0046817702297075.gr1.lrg.gif&fig=&kwhquery=&issn=0046-8177&ishighres=false&allhighres=false&free=yes','journalimage');)
29707-5/journalimage?img=PIIS0046817702297075.gr2.lrg.gif&fig=&kwhquery=&issn=0046-8177&ishighres=false&allhighres=false&free=yes','journalimage');)
29707-5/journalimage?img=PIIS0046817702297075.gr3.lrg.gif&fig=fig3&kwhquery=&issn=0046-8177&ishighres=false&allhighres=false&free=yes','journalimage');)
29707-5/journalimage?img=PIIS0046817702297075.gr4.lrg.gif&fig=&kwhquery=&issn=0046-8177&ishighres=false&allhighres=false&free=yes','journalimage');)