- Email: [email protected]
Expression of c-kit in uterine carcinosarcoma
Gynecologic Oncology 96 (2005) 210 – 215 www.elsevier.com/locate/ygyno
Expression of c-kit in uterine carcinosarcoma Joseph Menczera,*, Vladimir Kravtsovb, Tally Levya, Esther Bergerb, Marek Glezermana, Ilana Avinoachb a
Department of Obstetrics and Gynecology, Gynecologic Oncology Unit, Edith Wolfson Medical Center, Holon, Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel b Department of Pathology, Edith Wolfson Medical Center, Holon, Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel Received 19 July 2004 Available online 5 November 2004
Abstract Objectives. The use of tyrosine kinase inhibitors (TKIs) has resulted in successful treatment of KIT-positive neoplasms. Carcinosarcoma is a very aggressive neoplasm. Consequently, c-kit expression may have significant clinical implications for this tumor. The purpose of the present study was to assess c-kit expression in the carcinomatous and sarcomatous element of carcinosarcoma to identify if KIT represents a therapeutic target for treatment of this neoplasm. Methods. Immunohistochemical staining for c-KIT was performed on paraffin-embedded tissue blocks of 20 consecutive uterine specimens with carcinosarcoma, 40 with endometrial carcinoma, and 12 with atrophic endometrium. Two pathologists assessed the scoring index of staining. Results. In the stromal element of carcinosarcoma, no immunohistochemically c-KIT stained cells were observed and in the epithelial element, a low scoring index of staining was present. In endometrial endometrioid carcinoma, the scoring index was high in only 15% of the specimens. In the atrophic endometrium, a low scoring index was seen in all specimens. Conclusions. Our results seem to suggest that c-kit probably plays no major role in the pathogenesis of the majority of these tumors while it may be involved in the pathogenesis of some endometrial carcinomas. In view of the multiple molecular targets that are activated by imatinib mesylate, no definitive conclusions can be drawn with regard its effectiveness in carcinosarcomas based on the present study. Further studies of c-kit expression in larger series of carcinosarcomas in order to settle the controversial issues with regard to frequency of expression, prognostic, and clinical value are warranted. D 2004 Elsevier Inc. All rights reserved. Keywords: C-kit expression; Carcinosarcoma; Endometrial carcinoma; Atrophic endometrium
Introduction C-kit is a proto-oncogene that codes for a transmembrane tyrosine kinase receptor KIT (c-kit protooncogene product = CD117/KIT) [1,2]. In humans, the proto-oncogene c-kit is localized on the q11–q12 region of chromosome 4 [3]. * Corresponding author. Gynecologic Oncology Unit, Department of Obstetrics and Gynecology, Edith Wolfson Medical Center, Holon 58100, Israel. Fax: +972 3 634 6466. E-mail address: [email protected] (J. Menczer). 0090-8258/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2004.09.045
Sequencing of the c-kit complementary DNA from human hematological proliferative disorders and gastrointestinal stromal tumor (GIST) cells demonstrated a high frequency of somatic mutations that lead to constitutive activation of KIT. Its activation causes uncontrolled aberrant cellular proliferation and resistance to apoptosis [4–6]. Deregulation of protein kinase activity therefore plays an important role in the pathogenesis of human malignant neoplasms. The use of tyrosine kinase inhibitors such as imatinib mesylate (STI-571; Gleevec) has resulted in successful treatment of KIT-positive hematological neoplasms and GIST [6–
J. Menczer et al. / Gynecologic Oncology 96 (2005) 210–215
11]. Consequently, c-kit expression may have significant clinical implications. C-kit expression has been assessed in a variety of gynecological tumors including endometrial carcinoma [12–14] and carcinosarcoma [15–22]. The results of c-kit expression assessment in carcinosarcoma studies are inconsistent. Uterine carcinosarcomas are very aggressive neoplasm with a poor prognosis due to unsatisfactory treatment. At the time of diagnosis, they are confined to the uterine body in only about 45% of the patients and their median survival is only 21 months [23]. The purpose of the present study was to assess c-kit expression in both elements of carcinosarcoma, that is, in the epithelial carcinomatous and in the stromal sarcomatous element, to identify if KIT represents a therapeutic target for treatment of this neoplasm. Expression of c-kit in endometrial carcinoma and atrophic endometrium was assessed as well for comparison.
Patients and methods Paraffin-embedded tissue blocks of 20 consecutive uterine specimens with carcinosarcoma and 40 uterine specimens with endometrial endometrioid adenocarcinoma diagnosed from 1995 to 2003 were examined, after institutional review board approval. Uterine specimens with a non-neoplastic atrophic endometrium from 12 patients were also examined. All patients with malignancy had surgical staging. Formalin-fixed hematoxylin–eosin stained 6 Am slides from the tissue of the same cases were newly performed and reviewed by two expert pathologists (IA and VK) in order to verify the diagnosis. Clinical data were abstracted from hospital files. Immunostaining Additional 4 Am unstained slides were prepared from each case for c-KIT immunohistochemical staining. The detection of CD117/c-KIT protein in all tissue samples was performed using a rabbit polyclonal antibody (Biocare Medical, CA, USA) in a 1:25 dilution. Immunohistochemistry was performed by deparaffinization of slides in xylene and degraded alcohols 70% and 96%. Immunoperoxidase stains were performed using the modified labeled streptavidine technique and run on an automated system (Ventana Autostainer Nexes, Tucson, AZ, USA), using amino ethyl carbazole as chromogen and amplification kit (Ventana). All sections were counterstained with Mayers hematoxylin. The immunohistochemical cell membrane staining in all specimens was evaluated with microscopy by counting 10 high power fields (400) with a minimum of 1000 cells counted per slide.
211
The two pathologists assessed two parameters in each section: (a)
the proportion of the positively stained cells of the total number of tumor cells counted in the fields examined, ranging from 0% to 100%. (b) the intensity of immunohistochemical staining that was graded subjectively on a 0 to 3 scale, in which 0 reflected no detectable staining and 3 represented very intense staining. The scoring index was then calculated by multiplying the intensity grade of the stained cells by the percentage of the positively stained cells. The score was considered low when the index was equal to 1 or less, and high when it was more than 1. Sections of GIST tissue known to contain c-KIT served as positive controls. Sections from the same samples served as negative controls, but the first antibody was omitted. The pathologists were blinded to the clinical data. Differences in staining intensity interpretation occurred in two cases and were reconciled following appropriate discussion between the two pathologists. Analysis of data was carried out using SPSS statistical analysis software (SPSS Inc., Chicago, IL, USA, 1999). Distribution of categorical variables such as type of cancer, stage and grade were compared using the chi-square or Fischer exact test. Survival was calculated by the Kaplan– Meier analysis and differences in survival by the log-rank method. The median follow up of the patients was 21 months. A P value of b0.05 was considered significant.
Results The mean age of our patients with carcinosarcoma and endometrial carcinoma (65.3 and 64.7, respectively) and the range (35–81 and 32–83, respectively) were similar. The ethnicity of these Israeli Jewish women was also similar, 80.0% and 78.1%, respectively, being of Ashkenazi origin. The mean age of the patients with atrophic endometrium was similar as well (60.2) ranging from 50 to 75. Table 1 presents additional selected clinical characteristics of both tumor study groups. The proportion of patients with stage I was statistically significantly higher in endometrial carcinoma than in carcinosarcoma (80% vs. 50%, P = 0.02). Although a higher proportion of carcinosarcoma than endometrial carcinoma patients had advanced grade tumors, the difference was statistically not significant. Most (65%) of the carcinosarcomas were of the homologous type and all endometrial carcinomas were of the endometrioid type. The estimated 5-year survival in carcinosarcoma was statistically significantly lower than in endometrial carcinoma (44.9% vs. 87.0; P = 0.004). Results of the immunohistochemical staining for c-KIT in both elements of carcinosarcoma, in endometrial carcinoma
212
J. Menczer et al. / Gynecologic Oncology 96 (2005) 210–215
Table 1 Selected characteristics of carcinosarcoma and endometrial carcinoma patients Characteristic Total Stage I II–IV Type Homologous Heterologous Unspecified Grade 1 2,3 Unspecified 5 year survival
Carcinosarcoma
Endometrial carcinoma
No.
%
No.
20
100.0
40
100.0
10 10
50.0 50.0
33 7
82.5 17.5
12 7 1
60.0 35.0 5.0
– – –
– – –
– 14 6 –
– 70.0 30.0 44.9
20 20 – –
50.0 50.0 – 87.0
%
and non-neoplastic atrophic endometrial tissue are presented in Table 2. In the stromal element of carcinosarcoma, no immunohistochemical c-KIT stained cells were observed. In the epithelial element of carcinosarcoma, c-KIT staining was present in only a minority of cells when either the z10% or the z30% stained cells criterion was used for positivity. In none of the specimens was a high intensity of staining (grades 2–3) or a high scoring index (score N1) observed. In endometrial carcinoma, a high proportion of cells stained for c-KIT with a high grade of intensity. The scoring index was high in 15% of the specimens. The difference between the scoring index in the epithelial element of carcinosarcoma and endometrial carcinoma was statistically not significant ( P = 0.16). In the atrophic endometrium low intensity staining (grades 0–1) was found in a large proportion (83.3%) of the specimens but the scoring index was low in all of them. In the control GIST tissue, high intensity (grade 2–3) in 100% of the cells was observed and the scoring index was high (score 2.7) as well.
Discussion Our data indicate a low proportion of c- KIT immunohistochemically stained cells and a low scoring index in the epithelial element and lack of expression in the sarcomatous element of carcinosarcoma. A high scoring index was found in only a small proportion of endometrial carcinomas. We are not aware of any other study that compared c-KIT expression in carcinosarcoma to its expression in endometrial carcinoma. Previous studies of carcinosarcoma comprise a small number of specimens and the results are inconsistent [15–22]. In one of these studies [16], all the tumors stained, in two of them [17,18], no immunohistochemically positive staining has been found and in the others positive staining ranged between 43 and 60%. As pointed out by Lucas et al. [24] with regard to desmoid fibromatosis, the discordant results in the previous studies of carcinosarcoma may be explained by the fact that different companies have supplied the antibodies used and that the immunohistochemical staining method was not uniform. The definition of positivity varied as well (Table 3). Since thresholds as low as 10% of tumor cells staining have been used for the definition of positivity by others [18,21,25,26], we have presented the results when this criterion was used as well. Our staining assessment and scoring technique, namely taking into account intensity of staining as well as the percentage of stained cells, has been previously used by others [27,28] and seems more appropriate. The reliability of our staining method is supported by the high percentage of stained cells, high intensity of staining, and high scoring index (score 2.7) observed in the control GIST tissue. Inconsistent results of c-KIT staining have been reported in uterine leiomyosarcomas as well [25,29]. They have been attributed to the above-mentioned factors as well as to possible differences between populations [25]. Expression of c-kit in carcinosarcoma according to ethnicity has been
Table 2 Expression of c-kit in carcinosarcoma and endometrial carcinoma C-kit expression Carcinosarcoma component Epithelial
Total % stained z 10% z 30% Intensity 0 1 2 3 Scoring index* Low High
Endometrial carcinoma
Stromal
Atrophic endometrium
No.
%
No.
%
No.
%
No.
%
20 5 4
100.0 25.0 20.0
20 – –
100.0 – –
40 28 16
100.0 70.0 40.0
12 9 4
100.0 75.0 33.3
13 7 0 0
65.0 35.0 0.0 0.0
– – – –
– – – –
10 9 14 7
25.0 22.5 35.0 17.5
2 8 2 0
16.7 66.6 16.7 0.0
20 0
100.0 0.0
– –
– –
34 6
85.0 15.0
12 0
100.0 0.0
* Scoring index: low = V 1; high N 1.
J. Menczer et al. / Gynecologic Oncology 96 (2005) 210–215
213
Table 3 Immunohistochemical staining results of c-kit expression assessment in previous studies Author
No. of cases
Antibody supplier
Positivity assessment
Positive No. (%)
Winter WE [15] Rushing RS [16] Klein WM [17] Ramondetta LM [18] Huh WK [19]* Constantinescu M [20]* Sawada M [21] Leath CA [22]
21 14 6 16 5 25 16 5
Dako Corp. Santa Cruz Biotechnology Dako Corp. Dako Corp. Santa Cruz Biotechnology Not specified Dako Corp. Santa Cruz Biotechnology
z 30% Scoring index % stained z 10%; Intensity Intensity categories Score z 10%; Intensity Score
9 14 0 0 5 25 8 3
(43) (100) (0) (0) (100) (100) (50) (60)
* Abstract only.
examined in only one study [15]. No statistically significant difference between white and black patients has been observed. Our patients consisted of Israeli Jewish women mainly of Ashkenazi origin. The current general notion is that carcinosarcomas are of monoclonal origin and that the sarcomatous element is the result of metaplasia [30]. Results of previous studies of similar p53 and p27 expression in the two elements of carcinosarcoma are in line with this notion [27,31–33]. In line with this concept are also other studies of carcinosarcoma that have found similar c-kit expression in the sarcomatous and epithelial elements [15,16,19,20]. The results of our study, namely a low scoring index in the epithelial element and complete lack of expression in the sarcomatous element of carcinosarcoma do not quite contradict this concept but are intriguing. Similar to our findings in endometrial carcinoma, some level of c-kit expression has been reported in a high proportion of this neoplasm in other studies, ranging from 44% to 100% [12–14,34] of the cases. In benign active endometrium, c-kit expression has been found in 93% and 79% cells in the proliferative and secretory phase, respectively. A possible relationship between estrogen and c-kit expression has therefore been suggested [12]. The practical lack of intense and high score c-KIT staining in atrophic endometrium in our study seem to support this possibility. It has been reported that in endometrial carcinoma, cKIT positivity is an adverse prognostic factor [13]. As expected, our endometrial carcinoma patients had a significantly higher 5-year survival rate than the carcinosarcoma patients. The significantly more favorable stage distribution of endometrial carcinoma has certainly contributed to this finding. With regard to c-kit expression, the better outcome in endometrial carcinoma patients was found although 40% stained even when the z30% criterion was used. Granted, a high scoring index in these patients was found in only 15% of the specimens. On the other hand, the carcinosarcoma patient had a low survival rate although they had a low frequency of stained cells and a low scoring index of c-kit expression. The difference in the proportion of high scoring index in carcinosarcoma and endometrial carcinoma did not reach significance probably due to the small number of patients. Our data set of both
malignancies was also too small for meaningful statistical analysis of the correlation between c-kit expression and prognostic factors such as stage, grade, and survival. Interestingly, conflicting results with regard to c-kit expression and prognosis have been also reported for ovarian carcinoma. One study found the lack of expression of c-kit to be associated with poor prognosis [35], while another study, found lack of c-kit expression in low grade and its presence in 26% of high-grade serous ovarian tumors [36]. Similar conflicting results have been reported for GIST as well [37,38]. Our results seem to suggest that c-kit probably plays no major role in the pathogenesis of the majority of carcinosarcomas while it may be involved in the pathogenesis of some endometrial endometrioid carcinomas. In addition, the immunohistochemically staining results of our study of c-kit expression in carcinosarcomas can be interpreted to imply that the current commercially available tyrosine kinase inhibitor imatinib mesylate (Gleevec) may be of questionable therapeutic value in these highly aggressive and lethal neoplasms. Rushing et al. [16] recently reported that, although all of their 14 carcinosarcomas expressed c-KIT, no KIT-activating mutations in exons 11 or 17 of c-kit were found in any of their tumors. They therefore concluded that these tumors are not candidates for treatment with imatinib mesylate since it is effective mainly in tumors with KIT-activating mutations [10,11]. Yet, as pointed out by these authors themselves, other mechanisms for KIT activation may occur, c-kit may contain other mutations, it might be activated by non-mutational mechanisms [10] and other TKIs may be identified. Moreover, imatinib mesylate inhibits not only c-KIT but also several other tyrosine kinases not assayed in the present study. In view of the multiple molecular targets that are activated by imatinib mesylate, it seems questionable to predict its lack of effectiveness in carcinosarcomas based on the present study. Furthermore, negative immunohistochemical staining does not necessarily mean that c-kit expression is truly negative. Other modalities of testing can be applied. Obviously, immunohistochemistry does not address kinase activation or the presence of c-kit mutations. There is also as yet no clearly defined level at which tumors must express cKIT to be subjects for effective TKI treatment.
214
J. Menczer et al. / Gynecologic Oncology 96 (2005) 210–215
Further studies of c-kit expression in larger series of carcinosarcomas in order to elucidate these issues and to settle the controversies with regard to its frequency of expression and its prognostic and therapeutic value are therefore warranted.
References [1] Yarden Y, Kuang WJ, Yang-Feng T, Coussens L, Munemitsu S, Dull TJ, et al. Human proto-oncogene kit, a new cell surface-receptor tyrosine kinase for unidentified ligand. EMBO J 1987;6:3341 – 51. [2] Williams DE, Eisenman J, Baird A, Rauch C, Van Ness K, March CJ, et al. Identification of a ligand for the c-kit proto-oncogene. Cell 1990;63:167 – 74. [3] d’Auriol L, Mattei MG, Andre C, Galibert F. Localization of the human c-kit protooncogene on the q11–q12 region of chromosome 4. Hum Genet 1988;78:374 – 6. [4] Nakata Y, Kimura A, Katoh O, Kawaishi K, Hyodo H, Abe K, et al. C-kit point mutation of extracellular domain in patients with myeloproliferative disorders. Br J Haematol 1995;91:6610 – 30. [5] Rubin BP, Singer S, Tsao C, Duensing A, Lux ML, Ruiz R, et al. KIT activation is a ubiquitous feature of gastrointestinal stromal tumors. Cancer Res 2001;61(22):8118 – 21. [6] Demetri GD. Identification and treatment of chemoresistant inoperable or metastatic GIST: experience with the selective tyrosine kinase inhibitor imatinibe mesylate (STI571). Eur J Cancer 2002;38(Suppl 5):S52–9. [7] Druker BJ, Sawyers CL, Kantarjian H, Resta DJ, Reese SF, Ford JM, et al. Activity of a specific inhibitor of the BCR ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N Engl J Med 2001;344:1038 – 42. [8] DeMatteo RP. The GIST of targeted cancer therapy: a tumor (gastrointestinal stromal tumor), a mutated gene (c-kit), and a molecular inhibitor (STI571). Ann Surg Oncol 2002;9:831 – 9. [9] Demetri GD. Targeting c-kit mutations in solid tumors: scientific rationale and novel therapeutic options. Semin Oncol 2001;28(Suppl 17):19 – 26. [10] Heinrich MC, Rubin BP, Longley BJ, Fletcher JA. Biology and genetic aspects of gastrointestinal stromal tumors: KIT activation and cytogenetic alterations. Hum Pathol 2002;33:484 – 95. [11] Heinrich MC, Blanke CD, Druker BJ, Corless CL. Inhibition of KIT tyrosine kinase activity: a novel molecular approach to treatment of KIT-positive malignancies. J Clin Oncol 2002;20:1692 – 703. [12] Elmore LW, Domson K, Moore JR, Kornstein M, Burks RT. Expression of c-kit (CD117) in benign and malignant human endometrial epithelium. Arch Pathol Lab Med 2001;125:146 – 51. [13] Scobie JV, Acs G, Bandera CA, Blank SV, Wheeler JE, Pasha TL, et al. C-kit immunoreactivity in endometrial adenocarcinomas and its clinicopathological significance. Int J Gynecol Pathol 2003;22: 149 – 55. [14] Alkushi A, Irving J, Hsu F, Dupuis B, Liu CL, Rijn M, et al. Immunoprofile of cervical and endometrial adenocarcinomas using a tissue microarray. Virchows Arch 2003;442:271 – 7. [15] Winter III WE, Seidman JD, Krivak TC, Chauhan S, Carlson JW, Rose GS, et al. Clinicopathological analysis of c-kit expression in carcinosarcomas and leiomyosarcomas of the uterine corpus. Gynecol Oncol 2003;91:3 – 8. [16] Rushing RS, Shajahan S, Chendil D, Wilder JL, Pulliam J, Lee EY, et al. Uterine sarcomas express KIT protein but lack mutation(s) in exon 11 or 17 of c-KIT. Gynecol Oncol 2003;91:9 – 14. [17] Klein WM, Kurman RJ. Lack of expression of c-kit protein (CD117) in mesenchymal tumors of the uterus and ovary. Int J Gynecol Pathol 2003;22:181 – 4.
[18] Ramondetta LM, Burke TW, Jhingran A, Schmandt R, Bevers MW, Wolf JK, et al. A phase II trial of cisplatin, ifosfamide, and mesna in patients with advanced or recurrent uterine malignant mixed Mullerian tumors with evaluation of potential molecular targets. Gynecol Oncol 2003;90:529 – 6. [19] Huh WK, Conner MG, Straughn Jr JM, Leath CA, Barnes MN, Alvarez RD, et al. Immunohistochemical evaluation of the c Kit protooncogene in sarcomas of the uterus. Gynecol Oncol 2002;84:531 [abstract]. [20] Constantinescu M, Turbat-Herrera EA, Nordberg ML, et al. CD117 and CD34 expression by immunohistochemistry in uterine mixed Mullerian tumors. Lab Invest 2003;83:186A – 7A [abstract]. [21] Sawada M, Tsuda H, Kimura M, Okamoto S, Kita T, Kasamatsu T, et al. Different expression patterns of KIT, EGFR, and HER-2 (c erbB-2) oncoproteins between epithelial and mesenchymal components in uterine carcinosarcoma. Cancer Sci 2003;94:986 – 91. [22] Leath III CA, Straughn Jr JM, Conner MG, Barnes III MN, Alvarez EE, Partridge EE, et al. Immunohistochemical evaluation of the c-kit proto-oncogene in sarcomas of the uterus: a case series. J Reprod Med 2004;49:71 – 5. [23] Gadducci A, Sartori E, Landoni F, Zola P, Maggino T, Cosio S, et al. The prognostic relevance of histological type in uterine sarcomas: a Cooperation Task Force (CTF) multivariate analysis of 249 cases. Eur J Gynecol Oncol 2002;23:295 – 9. [24] Lucas DR, al-Abbadi M, Tabaczka P, Hamre MR, Weaver DW, Mott MJ. c-Kit expression in desmoid fibromatosis. Comparative immunohistochemical evaluation of two commercial antibodies. Am J Clin Pathol 2003;119:339 – 45. [25] Wang L, Felix JC, Lee JL, Tan PY, Tourgeman DE, O’Meara AT, et al. The proto-oncogene c-kit is expressed in leiomyosarcomas of the uterus. Gynecol Oncol 2003;90:402 – 6. [26] Miettinen M, Sobin LH, Sarlomo-Rikala M. Immunohistochemical spectrum of GISTs at different sites and their differential diagnosis with reference to CD117 (KIT). Mod Pathol 2000;13:1134 – 42. [27] Abargel A, Avinoach I, Kravtsov V, Boaz M, Glezerman M, Menczer J. Expression of p27 and p53: comparative analysis of uterine carcinosarcoma and endometrial carcinoma. Int J Gynecol Cancer 2004;14:354 – 9. [28] Bamberger A, Riethdorf L, Milde-Langosch K, Bamberger CM, Thuneke I, Erdmann I, et al. Strongly reduced expression of the cell cycle inhibitor in endometrial neoplasia. Virchows Arch 1999;434: 423 – 8. [29] Raspollini MR, Villanucci A, Amunni G, Paglierani M, Taddei A, Taddei GL. C-kit expression in uterine leiomyosarcomas: an immunocytochemical study of 29 cases of malignant smooth muscle tumors of the uterus. J Chemother 2003;15:81 – 4. [30] McCluggage WG. Uterine carcinosarcomas (malignant mixed Mullerian tumors) are metaplastic carcinomas. Int J Gynecol Cancer 2002;12:687 – 90. [31] Kounelis S, Jones MW, Papadaki H, Bakker A, Swalsky P, Finkelstein SD. Carcinosarcoma (malignant mixed Mullerian tumors) of the genital tract: comparative molecular analysis of epithelial and mesenchymal components. Hum Pathol 1998;29:82 – 7. [32] Guarino M, Giordano F, Pallotti F, Polizzoti G, Tricomi P, Cristofori E. Malignant mixed Mullerian tumor of the uterus. Feature favoring its origin from a common cell clone and an epithelial to mesenchymal transformation of histogenesis. Tumori 1998;84:391 – 7. [33] Szukala SA, Marks JR, Burchette JL, Elbendary JL, Krigman HR. Coexpression of p53 by epithelial and stromal elements in carcinosarcoma of the female genital tract: an immunohistochemical study of 19 cases. In J Gynecol Cancer 1999;9:131 – 6. [34] Arber DA, Tamayo R, Weiss LM. Paraffin section detection of the ckit gene product (CD117) in human tissues: value in the diagnosis of mast cell disorders. Hum Pathol 1998;29:498 – 504. [35] Tonary AM, Macdonald EA, Faught W, Senterman MK, Vanderhyden BC. Lack of expression of c-KIT in ovarian cancers is associated with poor prognosis. Int J Cancer 2000;89:242 – 50.
J. Menczer et al. / Gynecologic Oncology 96 (2005) 210–215 [36] Schmandt RE, Broaddus R, Lu KH, et al. Expression of c-ABL, cKIT, and platelet derived growth factor receptor-beta in ovarian serous carcinoma and normal ovarian surface epithelium. Cancer 2003;98:758 – 64. [37] Taniguchi M, Nishida T, Hirota S, Isozaki K, Ito T, Nomura T, et al.
215
Effect of c-kit mutation on prognosis of gastrointestinal stromal tumors. Cancer Res 1999;59:4297 – 300. [38] Feng F, Liu XH, Xie Q, Liu WQ, Bai CG, Ma DL. Expression and mutation of c-kit gene in gastrointestinal stromal tumors. World J Gastroenterol 2003;9:2548 – 51.
Expression of c-kit in uterine carcinosarcoma Joseph Menczera,*, Vladimir Kravtsovb, Tally Levya, Esther Bergerb, Marek Glezermana, Ilana Avinoachb a
Department of Obstetrics and Gynecology, Gynecologic Oncology Unit, Edith Wolfson Medical Center, Holon, Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel b Department of Pathology, Edith Wolfson Medical Center, Holon, Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel Received 19 July 2004 Available online 5 November 2004
Abstract Objectives. The use of tyrosine kinase inhibitors (TKIs) has resulted in successful treatment of KIT-positive neoplasms. Carcinosarcoma is a very aggressive neoplasm. Consequently, c-kit expression may have significant clinical implications for this tumor. The purpose of the present study was to assess c-kit expression in the carcinomatous and sarcomatous element of carcinosarcoma to identify if KIT represents a therapeutic target for treatment of this neoplasm. Methods. Immunohistochemical staining for c-KIT was performed on paraffin-embedded tissue blocks of 20 consecutive uterine specimens with carcinosarcoma, 40 with endometrial carcinoma, and 12 with atrophic endometrium. Two pathologists assessed the scoring index of staining. Results. In the stromal element of carcinosarcoma, no immunohistochemically c-KIT stained cells were observed and in the epithelial element, a low scoring index of staining was present. In endometrial endometrioid carcinoma, the scoring index was high in only 15% of the specimens. In the atrophic endometrium, a low scoring index was seen in all specimens. Conclusions. Our results seem to suggest that c-kit probably plays no major role in the pathogenesis of the majority of these tumors while it may be involved in the pathogenesis of some endometrial carcinomas. In view of the multiple molecular targets that are activated by imatinib mesylate, no definitive conclusions can be drawn with regard its effectiveness in carcinosarcomas based on the present study. Further studies of c-kit expression in larger series of carcinosarcomas in order to settle the controversial issues with regard to frequency of expression, prognostic, and clinical value are warranted. D 2004 Elsevier Inc. All rights reserved. Keywords: C-kit expression; Carcinosarcoma; Endometrial carcinoma; Atrophic endometrium
Introduction C-kit is a proto-oncogene that codes for a transmembrane tyrosine kinase receptor KIT (c-kit protooncogene product = CD117/KIT) [1,2]. In humans, the proto-oncogene c-kit is localized on the q11–q12 region of chromosome 4 [3]. * Corresponding author. Gynecologic Oncology Unit, Department of Obstetrics and Gynecology, Edith Wolfson Medical Center, Holon 58100, Israel. Fax: +972 3 634 6466. E-mail address: [email protected] (J. Menczer). 0090-8258/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2004.09.045
Sequencing of the c-kit complementary DNA from human hematological proliferative disorders and gastrointestinal stromal tumor (GIST) cells demonstrated a high frequency of somatic mutations that lead to constitutive activation of KIT. Its activation causes uncontrolled aberrant cellular proliferation and resistance to apoptosis [4–6]. Deregulation of protein kinase activity therefore plays an important role in the pathogenesis of human malignant neoplasms. The use of tyrosine kinase inhibitors such as imatinib mesylate (STI-571; Gleevec) has resulted in successful treatment of KIT-positive hematological neoplasms and GIST [6–
J. Menczer et al. / Gynecologic Oncology 96 (2005) 210–215
11]. Consequently, c-kit expression may have significant clinical implications. C-kit expression has been assessed in a variety of gynecological tumors including endometrial carcinoma [12–14] and carcinosarcoma [15–22]. The results of c-kit expression assessment in carcinosarcoma studies are inconsistent. Uterine carcinosarcomas are very aggressive neoplasm with a poor prognosis due to unsatisfactory treatment. At the time of diagnosis, they are confined to the uterine body in only about 45% of the patients and their median survival is only 21 months [23]. The purpose of the present study was to assess c-kit expression in both elements of carcinosarcoma, that is, in the epithelial carcinomatous and in the stromal sarcomatous element, to identify if KIT represents a therapeutic target for treatment of this neoplasm. Expression of c-kit in endometrial carcinoma and atrophic endometrium was assessed as well for comparison.
Patients and methods Paraffin-embedded tissue blocks of 20 consecutive uterine specimens with carcinosarcoma and 40 uterine specimens with endometrial endometrioid adenocarcinoma diagnosed from 1995 to 2003 were examined, after institutional review board approval. Uterine specimens with a non-neoplastic atrophic endometrium from 12 patients were also examined. All patients with malignancy had surgical staging. Formalin-fixed hematoxylin–eosin stained 6 Am slides from the tissue of the same cases were newly performed and reviewed by two expert pathologists (IA and VK) in order to verify the diagnosis. Clinical data were abstracted from hospital files. Immunostaining Additional 4 Am unstained slides were prepared from each case for c-KIT immunohistochemical staining. The detection of CD117/c-KIT protein in all tissue samples was performed using a rabbit polyclonal antibody (Biocare Medical, CA, USA) in a 1:25 dilution. Immunohistochemistry was performed by deparaffinization of slides in xylene and degraded alcohols 70% and 96%. Immunoperoxidase stains were performed using the modified labeled streptavidine technique and run on an automated system (Ventana Autostainer Nexes, Tucson, AZ, USA), using amino ethyl carbazole as chromogen and amplification kit (Ventana). All sections were counterstained with Mayers hematoxylin. The immunohistochemical cell membrane staining in all specimens was evaluated with microscopy by counting 10 high power fields (400) with a minimum of 1000 cells counted per slide.
211
The two pathologists assessed two parameters in each section: (a)
the proportion of the positively stained cells of the total number of tumor cells counted in the fields examined, ranging from 0% to 100%. (b) the intensity of immunohistochemical staining that was graded subjectively on a 0 to 3 scale, in which 0 reflected no detectable staining and 3 represented very intense staining. The scoring index was then calculated by multiplying the intensity grade of the stained cells by the percentage of the positively stained cells. The score was considered low when the index was equal to 1 or less, and high when it was more than 1. Sections of GIST tissue known to contain c-KIT served as positive controls. Sections from the same samples served as negative controls, but the first antibody was omitted. The pathologists were blinded to the clinical data. Differences in staining intensity interpretation occurred in two cases and were reconciled following appropriate discussion between the two pathologists. Analysis of data was carried out using SPSS statistical analysis software (SPSS Inc., Chicago, IL, USA, 1999). Distribution of categorical variables such as type of cancer, stage and grade were compared using the chi-square or Fischer exact test. Survival was calculated by the Kaplan– Meier analysis and differences in survival by the log-rank method. The median follow up of the patients was 21 months. A P value of b0.05 was considered significant.
Results The mean age of our patients with carcinosarcoma and endometrial carcinoma (65.3 and 64.7, respectively) and the range (35–81 and 32–83, respectively) were similar. The ethnicity of these Israeli Jewish women was also similar, 80.0% and 78.1%, respectively, being of Ashkenazi origin. The mean age of the patients with atrophic endometrium was similar as well (60.2) ranging from 50 to 75. Table 1 presents additional selected clinical characteristics of both tumor study groups. The proportion of patients with stage I was statistically significantly higher in endometrial carcinoma than in carcinosarcoma (80% vs. 50%, P = 0.02). Although a higher proportion of carcinosarcoma than endometrial carcinoma patients had advanced grade tumors, the difference was statistically not significant. Most (65%) of the carcinosarcomas were of the homologous type and all endometrial carcinomas were of the endometrioid type. The estimated 5-year survival in carcinosarcoma was statistically significantly lower than in endometrial carcinoma (44.9% vs. 87.0; P = 0.004). Results of the immunohistochemical staining for c-KIT in both elements of carcinosarcoma, in endometrial carcinoma
212
J. Menczer et al. / Gynecologic Oncology 96 (2005) 210–215
Table 1 Selected characteristics of carcinosarcoma and endometrial carcinoma patients Characteristic Total Stage I II–IV Type Homologous Heterologous Unspecified Grade 1 2,3 Unspecified 5 year survival
Carcinosarcoma
Endometrial carcinoma
No.
%
No.
20
100.0
40
100.0
10 10
50.0 50.0
33 7
82.5 17.5
12 7 1
60.0 35.0 5.0
– – –
– – –
– 14 6 –
– 70.0 30.0 44.9
20 20 – –
50.0 50.0 – 87.0
%
and non-neoplastic atrophic endometrial tissue are presented in Table 2. In the stromal element of carcinosarcoma, no immunohistochemical c-KIT stained cells were observed. In the epithelial element of carcinosarcoma, c-KIT staining was present in only a minority of cells when either the z10% or the z30% stained cells criterion was used for positivity. In none of the specimens was a high intensity of staining (grades 2–3) or a high scoring index (score N1) observed. In endometrial carcinoma, a high proportion of cells stained for c-KIT with a high grade of intensity. The scoring index was high in 15% of the specimens. The difference between the scoring index in the epithelial element of carcinosarcoma and endometrial carcinoma was statistically not significant ( P = 0.16). In the atrophic endometrium low intensity staining (grades 0–1) was found in a large proportion (83.3%) of the specimens but the scoring index was low in all of them. In the control GIST tissue, high intensity (grade 2–3) in 100% of the cells was observed and the scoring index was high (score 2.7) as well.
Discussion Our data indicate a low proportion of c- KIT immunohistochemically stained cells and a low scoring index in the epithelial element and lack of expression in the sarcomatous element of carcinosarcoma. A high scoring index was found in only a small proportion of endometrial carcinomas. We are not aware of any other study that compared c-KIT expression in carcinosarcoma to its expression in endometrial carcinoma. Previous studies of carcinosarcoma comprise a small number of specimens and the results are inconsistent [15–22]. In one of these studies [16], all the tumors stained, in two of them [17,18], no immunohistochemically positive staining has been found and in the others positive staining ranged between 43 and 60%. As pointed out by Lucas et al. [24] with regard to desmoid fibromatosis, the discordant results in the previous studies of carcinosarcoma may be explained by the fact that different companies have supplied the antibodies used and that the immunohistochemical staining method was not uniform. The definition of positivity varied as well (Table 3). Since thresholds as low as 10% of tumor cells staining have been used for the definition of positivity by others [18,21,25,26], we have presented the results when this criterion was used as well. Our staining assessment and scoring technique, namely taking into account intensity of staining as well as the percentage of stained cells, has been previously used by others [27,28] and seems more appropriate. The reliability of our staining method is supported by the high percentage of stained cells, high intensity of staining, and high scoring index (score 2.7) observed in the control GIST tissue. Inconsistent results of c-KIT staining have been reported in uterine leiomyosarcomas as well [25,29]. They have been attributed to the above-mentioned factors as well as to possible differences between populations [25]. Expression of c-kit in carcinosarcoma according to ethnicity has been
Table 2 Expression of c-kit in carcinosarcoma and endometrial carcinoma C-kit expression Carcinosarcoma component Epithelial
Total % stained z 10% z 30% Intensity 0 1 2 3 Scoring index* Low High
Endometrial carcinoma
Stromal
Atrophic endometrium
No.
%
No.
%
No.
%
No.
%
20 5 4
100.0 25.0 20.0
20 – –
100.0 – –
40 28 16
100.0 70.0 40.0
12 9 4
100.0 75.0 33.3
13 7 0 0
65.0 35.0 0.0 0.0
– – – –
– – – –
10 9 14 7
25.0 22.5 35.0 17.5
2 8 2 0
16.7 66.6 16.7 0.0
20 0
100.0 0.0
– –
– –
34 6
85.0 15.0
12 0
100.0 0.0
* Scoring index: low = V 1; high N 1.
J. Menczer et al. / Gynecologic Oncology 96 (2005) 210–215
213
Table 3 Immunohistochemical staining results of c-kit expression assessment in previous studies Author
No. of cases
Antibody supplier
Positivity assessment
Positive No. (%)
Winter WE [15] Rushing RS [16] Klein WM [17] Ramondetta LM [18] Huh WK [19]* Constantinescu M [20]* Sawada M [21] Leath CA [22]
21 14 6 16 5 25 16 5
Dako Corp. Santa Cruz Biotechnology Dako Corp. Dako Corp. Santa Cruz Biotechnology Not specified Dako Corp. Santa Cruz Biotechnology
z 30% Scoring index % stained z 10%; Intensity Intensity categories Score z 10%; Intensity Score
9 14 0 0 5 25 8 3
(43) (100) (0) (0) (100) (100) (50) (60)
* Abstract only.
examined in only one study [15]. No statistically significant difference between white and black patients has been observed. Our patients consisted of Israeli Jewish women mainly of Ashkenazi origin. The current general notion is that carcinosarcomas are of monoclonal origin and that the sarcomatous element is the result of metaplasia [30]. Results of previous studies of similar p53 and p27 expression in the two elements of carcinosarcoma are in line with this notion [27,31–33]. In line with this concept are also other studies of carcinosarcoma that have found similar c-kit expression in the sarcomatous and epithelial elements [15,16,19,20]. The results of our study, namely a low scoring index in the epithelial element and complete lack of expression in the sarcomatous element of carcinosarcoma do not quite contradict this concept but are intriguing. Similar to our findings in endometrial carcinoma, some level of c-kit expression has been reported in a high proportion of this neoplasm in other studies, ranging from 44% to 100% [12–14,34] of the cases. In benign active endometrium, c-kit expression has been found in 93% and 79% cells in the proliferative and secretory phase, respectively. A possible relationship between estrogen and c-kit expression has therefore been suggested [12]. The practical lack of intense and high score c-KIT staining in atrophic endometrium in our study seem to support this possibility. It has been reported that in endometrial carcinoma, cKIT positivity is an adverse prognostic factor [13]. As expected, our endometrial carcinoma patients had a significantly higher 5-year survival rate than the carcinosarcoma patients. The significantly more favorable stage distribution of endometrial carcinoma has certainly contributed to this finding. With regard to c-kit expression, the better outcome in endometrial carcinoma patients was found although 40% stained even when the z30% criterion was used. Granted, a high scoring index in these patients was found in only 15% of the specimens. On the other hand, the carcinosarcoma patient had a low survival rate although they had a low frequency of stained cells and a low scoring index of c-kit expression. The difference in the proportion of high scoring index in carcinosarcoma and endometrial carcinoma did not reach significance probably due to the small number of patients. Our data set of both
malignancies was also too small for meaningful statistical analysis of the correlation between c-kit expression and prognostic factors such as stage, grade, and survival. Interestingly, conflicting results with regard to c-kit expression and prognosis have been also reported for ovarian carcinoma. One study found the lack of expression of c-kit to be associated with poor prognosis [35], while another study, found lack of c-kit expression in low grade and its presence in 26% of high-grade serous ovarian tumors [36]. Similar conflicting results have been reported for GIST as well [37,38]. Our results seem to suggest that c-kit probably plays no major role in the pathogenesis of the majority of carcinosarcomas while it may be involved in the pathogenesis of some endometrial endometrioid carcinomas. In addition, the immunohistochemically staining results of our study of c-kit expression in carcinosarcomas can be interpreted to imply that the current commercially available tyrosine kinase inhibitor imatinib mesylate (Gleevec) may be of questionable therapeutic value in these highly aggressive and lethal neoplasms. Rushing et al. [16] recently reported that, although all of their 14 carcinosarcomas expressed c-KIT, no KIT-activating mutations in exons 11 or 17 of c-kit were found in any of their tumors. They therefore concluded that these tumors are not candidates for treatment with imatinib mesylate since it is effective mainly in tumors with KIT-activating mutations [10,11]. Yet, as pointed out by these authors themselves, other mechanisms for KIT activation may occur, c-kit may contain other mutations, it might be activated by non-mutational mechanisms [10] and other TKIs may be identified. Moreover, imatinib mesylate inhibits not only c-KIT but also several other tyrosine kinases not assayed in the present study. In view of the multiple molecular targets that are activated by imatinib mesylate, it seems questionable to predict its lack of effectiveness in carcinosarcomas based on the present study. Furthermore, negative immunohistochemical staining does not necessarily mean that c-kit expression is truly negative. Other modalities of testing can be applied. Obviously, immunohistochemistry does not address kinase activation or the presence of c-kit mutations. There is also as yet no clearly defined level at which tumors must express cKIT to be subjects for effective TKI treatment.
214
J. Menczer et al. / Gynecologic Oncology 96 (2005) 210–215
Further studies of c-kit expression in larger series of carcinosarcomas in order to elucidate these issues and to settle the controversies with regard to its frequency of expression and its prognostic and therapeutic value are therefore warranted.
References [1] Yarden Y, Kuang WJ, Yang-Feng T, Coussens L, Munemitsu S, Dull TJ, et al. Human proto-oncogene kit, a new cell surface-receptor tyrosine kinase for unidentified ligand. EMBO J 1987;6:3341 – 51. [2] Williams DE, Eisenman J, Baird A, Rauch C, Van Ness K, March CJ, et al. Identification of a ligand for the c-kit proto-oncogene. Cell 1990;63:167 – 74. [3] d’Auriol L, Mattei MG, Andre C, Galibert F. Localization of the human c-kit protooncogene on the q11–q12 region of chromosome 4. Hum Genet 1988;78:374 – 6. [4] Nakata Y, Kimura A, Katoh O, Kawaishi K, Hyodo H, Abe K, et al. C-kit point mutation of extracellular domain in patients with myeloproliferative disorders. Br J Haematol 1995;91:6610 – 30. [5] Rubin BP, Singer S, Tsao C, Duensing A, Lux ML, Ruiz R, et al. KIT activation is a ubiquitous feature of gastrointestinal stromal tumors. Cancer Res 2001;61(22):8118 – 21. [6] Demetri GD. Identification and treatment of chemoresistant inoperable or metastatic GIST: experience with the selective tyrosine kinase inhibitor imatinibe mesylate (STI571). Eur J Cancer 2002;38(Suppl 5):S52–9. [7] Druker BJ, Sawyers CL, Kantarjian H, Resta DJ, Reese SF, Ford JM, et al. Activity of a specific inhibitor of the BCR ABL tyrosine kinase in the blast crisis of chronic myeloid leukemia and acute lymphoblastic leukemia with the Philadelphia chromosome. N Engl J Med 2001;344:1038 – 42. [8] DeMatteo RP. The GIST of targeted cancer therapy: a tumor (gastrointestinal stromal tumor), a mutated gene (c-kit), and a molecular inhibitor (STI571). Ann Surg Oncol 2002;9:831 – 9. [9] Demetri GD. Targeting c-kit mutations in solid tumors: scientific rationale and novel therapeutic options. Semin Oncol 2001;28(Suppl 17):19 – 26. [10] Heinrich MC, Rubin BP, Longley BJ, Fletcher JA. Biology and genetic aspects of gastrointestinal stromal tumors: KIT activation and cytogenetic alterations. Hum Pathol 2002;33:484 – 95. [11] Heinrich MC, Blanke CD, Druker BJ, Corless CL. Inhibition of KIT tyrosine kinase activity: a novel molecular approach to treatment of KIT-positive malignancies. J Clin Oncol 2002;20:1692 – 703. [12] Elmore LW, Domson K, Moore JR, Kornstein M, Burks RT. Expression of c-kit (CD117) in benign and malignant human endometrial epithelium. Arch Pathol Lab Med 2001;125:146 – 51. [13] Scobie JV, Acs G, Bandera CA, Blank SV, Wheeler JE, Pasha TL, et al. C-kit immunoreactivity in endometrial adenocarcinomas and its clinicopathological significance. Int J Gynecol Pathol 2003;22: 149 – 55. [14] Alkushi A, Irving J, Hsu F, Dupuis B, Liu CL, Rijn M, et al. Immunoprofile of cervical and endometrial adenocarcinomas using a tissue microarray. Virchows Arch 2003;442:271 – 7. [15] Winter III WE, Seidman JD, Krivak TC, Chauhan S, Carlson JW, Rose GS, et al. Clinicopathological analysis of c-kit expression in carcinosarcomas and leiomyosarcomas of the uterine corpus. Gynecol Oncol 2003;91:3 – 8. [16] Rushing RS, Shajahan S, Chendil D, Wilder JL, Pulliam J, Lee EY, et al. Uterine sarcomas express KIT protein but lack mutation(s) in exon 11 or 17 of c-KIT. Gynecol Oncol 2003;91:9 – 14. [17] Klein WM, Kurman RJ. Lack of expression of c-kit protein (CD117) in mesenchymal tumors of the uterus and ovary. Int J Gynecol Pathol 2003;22:181 – 4.
[18] Ramondetta LM, Burke TW, Jhingran A, Schmandt R, Bevers MW, Wolf JK, et al. A phase II trial of cisplatin, ifosfamide, and mesna in patients with advanced or recurrent uterine malignant mixed Mullerian tumors with evaluation of potential molecular targets. Gynecol Oncol 2003;90:529 – 6. [19] Huh WK, Conner MG, Straughn Jr JM, Leath CA, Barnes MN, Alvarez RD, et al. Immunohistochemical evaluation of the c Kit protooncogene in sarcomas of the uterus. Gynecol Oncol 2002;84:531 [abstract]. [20] Constantinescu M, Turbat-Herrera EA, Nordberg ML, et al. CD117 and CD34 expression by immunohistochemistry in uterine mixed Mullerian tumors. Lab Invest 2003;83:186A – 7A [abstract]. [21] Sawada M, Tsuda H, Kimura M, Okamoto S, Kita T, Kasamatsu T, et al. Different expression patterns of KIT, EGFR, and HER-2 (c erbB-2) oncoproteins between epithelial and mesenchymal components in uterine carcinosarcoma. Cancer Sci 2003;94:986 – 91. [22] Leath III CA, Straughn Jr JM, Conner MG, Barnes III MN, Alvarez EE, Partridge EE, et al. Immunohistochemical evaluation of the c-kit proto-oncogene in sarcomas of the uterus: a case series. J Reprod Med 2004;49:71 – 5. [23] Gadducci A, Sartori E, Landoni F, Zola P, Maggino T, Cosio S, et al. The prognostic relevance of histological type in uterine sarcomas: a Cooperation Task Force (CTF) multivariate analysis of 249 cases. Eur J Gynecol Oncol 2002;23:295 – 9. [24] Lucas DR, al-Abbadi M, Tabaczka P, Hamre MR, Weaver DW, Mott MJ. c-Kit expression in desmoid fibromatosis. Comparative immunohistochemical evaluation of two commercial antibodies. Am J Clin Pathol 2003;119:339 – 45. [25] Wang L, Felix JC, Lee JL, Tan PY, Tourgeman DE, O’Meara AT, et al. The proto-oncogene c-kit is expressed in leiomyosarcomas of the uterus. Gynecol Oncol 2003;90:402 – 6. [26] Miettinen M, Sobin LH, Sarlomo-Rikala M. Immunohistochemical spectrum of GISTs at different sites and their differential diagnosis with reference to CD117 (KIT). Mod Pathol 2000;13:1134 – 42. [27] Abargel A, Avinoach I, Kravtsov V, Boaz M, Glezerman M, Menczer J. Expression of p27 and p53: comparative analysis of uterine carcinosarcoma and endometrial carcinoma. Int J Gynecol Cancer 2004;14:354 – 9. [28] Bamberger A, Riethdorf L, Milde-Langosch K, Bamberger CM, Thuneke I, Erdmann I, et al. Strongly reduced expression of the cell cycle inhibitor in endometrial neoplasia. Virchows Arch 1999;434: 423 – 8. [29] Raspollini MR, Villanucci A, Amunni G, Paglierani M, Taddei A, Taddei GL. C-kit expression in uterine leiomyosarcomas: an immunocytochemical study of 29 cases of malignant smooth muscle tumors of the uterus. J Chemother 2003;15:81 – 4. [30] McCluggage WG. Uterine carcinosarcomas (malignant mixed Mullerian tumors) are metaplastic carcinomas. Int J Gynecol Cancer 2002;12:687 – 90. [31] Kounelis S, Jones MW, Papadaki H, Bakker A, Swalsky P, Finkelstein SD. Carcinosarcoma (malignant mixed Mullerian tumors) of the genital tract: comparative molecular analysis of epithelial and mesenchymal components. Hum Pathol 1998;29:82 – 7. [32] Guarino M, Giordano F, Pallotti F, Polizzoti G, Tricomi P, Cristofori E. Malignant mixed Mullerian tumor of the uterus. Feature favoring its origin from a common cell clone and an epithelial to mesenchymal transformation of histogenesis. Tumori 1998;84:391 – 7. [33] Szukala SA, Marks JR, Burchette JL, Elbendary JL, Krigman HR. Coexpression of p53 by epithelial and stromal elements in carcinosarcoma of the female genital tract: an immunohistochemical study of 19 cases. In J Gynecol Cancer 1999;9:131 – 6. [34] Arber DA, Tamayo R, Weiss LM. Paraffin section detection of the ckit gene product (CD117) in human tissues: value in the diagnosis of mast cell disorders. Hum Pathol 1998;29:498 – 504. [35] Tonary AM, Macdonald EA, Faught W, Senterman MK, Vanderhyden BC. Lack of expression of c-KIT in ovarian cancers is associated with poor prognosis. Int J Cancer 2000;89:242 – 50.
J. Menczer et al. / Gynecologic Oncology 96 (2005) 210–215 [36] Schmandt RE, Broaddus R, Lu KH, et al. Expression of c-ABL, cKIT, and platelet derived growth factor receptor-beta in ovarian serous carcinoma and normal ovarian surface epithelium. Cancer 2003;98:758 – 64. [37] Taniguchi M, Nishida T, Hirota S, Isozaki K, Ito T, Nomura T, et al.
215
Effect of c-kit mutation on prognosis of gastrointestinal stromal tumors. Cancer Res 1999;59:4297 – 300. [38] Feng F, Liu XH, Xie Q, Liu WQ, Bai CG, Ma DL. Expression and mutation of c-kit gene in gastrointestinal stromal tumors. World J Gastroenterol 2003;9:2548 – 51.