Expression and mutational analysis of c-kit in ovarian surface epithelial tumors

Coexpression of Kit ligand and c-kit has been reported in some gynecologic tumors. To determine whether imatinib mesylate is useful in ovarian epithelial tumors, we performed immunohistochemical and mutational analysis. The cases consisted of 33 cases, which included 13 serous cystadenocarcinomas, 1 borderline serous tumor, 8 mucinous cystadenocarcinomas, 6 borderline mucinous tumors and 5 clear cell carcinomas. Five cases of serous cystadenoma and 5 cases of mucinous cystadenoma were also included. In the immunohistochemical study, 3 cases (3/6, 50%) of borderline mucinous cystic tumor and two cases (2/8, 25%) of mucinous cystadenocarcinoma show positive staining for KIT protein. Only one case (1/13, 7.7%) of serous cystadenocarcinoma had positive staining. On mutational analysis, no mutation was identified at exon 11. However, two cases of borderline mucinous tumors and one case of mucinous cystadenocarcinoma had mutations at exon 17. In these cases, the immunohistochemistry also shows focal positive staining at epithelial component. Although, KIT protein expression showed higher incidence in mucinous tumors than serous tumors, they lack KIT-activating mutations in exon 11. Thus, ovarian surface epithelial tumors are unlikely to respond to imatinib mesylate.     

基质金属蛋白酶-7在卵巢浆液性肿瘤中的表达

目的：探讨MMP－7在卵巢浆液性肿瘤中的表达情况。方法：采用免疫组化SP法对6例正常卵巢、12例卵巢浆液性囊腺瘤、6例交界性囊腺瘤及22例卵巢浆液性囊腺癌MMP－7的表达进行了研究。结果：正常卵巢不表达MMP－7。大部分卵巢浆液性肿瘤的胞浆及间质中都有MMP－7的阳性表达。MMP－7的卵巢良性，恶性及交界性浆液性肿瘤胞浆中的表达无明显差异；而在肿瘤间质中，恶性及交界性卵巢浆液性肿瘤中的表达远高于良性肿瘤（P＜0．05）。在交界性及恶性浆液性卵巢肿瘤中，部分肿瘤细胞的细胞核中也有MMP－7的表达，为国内外首次报道。结论：MMP－7可能在卵巢浆液性肿瘤的进展中发挥重要作用。     

卵巢粘液性及浆液性囊腺瘤癌胚抗原的免疫组化研究

以CEA单克隆抗体,双PAP法研究卵巢粘液性和浆液性囊腺瘤组织内CEA的定位。结果表明:良性瘤23例,CEA全部(-)。交界性肿瘤中粘液性22例,CEA(+)者21例;浆液性15例,CEA(+)者7例。恶性肿瘤中粘液性23例,CEA(+)者18例,浆液性37例,CEA(+)者7例。由于粘液性交界性肿瘤有较高的CEA检出率,可作为诊断参考指标。     

Differential expression of MUC2 and MUC5AC mucin genes in primary ovarian and metastatic colonic carcinoma

Colonic adenocarcinoma, the most common tumor metastatic to the ovary, may closely mimic primary ovarian adenocarcinoma, especially that of mucinous or endometrioid histology. The differential diagnosis is important for therapeutic considerations. Mucin gene expression is relatively organ-specific and may therefore have use in distinguishing between colonic carcinomas metastatic to the ovary and primary ovarian tumors. In this study, we compared the expression of MUC2 and MUC5AC apomucins in 10 colonic adenocarcinomas metastatic with the ovary, 10 ovarian endometrioid carcinomas (4 primary, 6 metastatic), and 32 primary mucinous ovarian tumors (12 cystadenomas, 10 borderline tumors, and 10 cystadenocarcinomas). Monoclonal antibodies CCP58 and 45M1 were used for immunostains of MUC2 and MUC5AC apomucin, respectively. All but 1 of the 10 metastatic colon adenocarcinomas expressed MUC2, whereas none expressed MUC5AC. None of the 10 endometrioid carcinomas expressed MUC2, and only 2 showed weak immunoreactivity with MUC5AC. All 32 primary mucinous ovarian tumors expressed MUC5AC. The percentages of MUC2-positive immunostaining for cystadenomas, borderline tumors, and cystadenocarcinomas were 0% (0/12), 50% (5/10), and 70% (7/10) respectively. These studies show that MUC2 and MUC5AC are useful markers in the distinction between colonic carcinoma metastatic to the ovary and primary ovarian carcinoma.     

Aberrant hypermethylation of RASSF1A promoter in ovarian borderline tumors and carcinomas

The newly identified 3p21.3 tumor suppressor gene RASSF1A is inactivated by hypermethylation in variable solid tumors, including those of the lung, breast, prostate, kidney, and ovary. The purpose of this study was to evaluate the methylation status of RASSF1A in various types and stages of ovarian epithelial tumors. We analyzed the DNA methylation status of ovarian tumors using methylation-specific polymerase chain reaction in 54 frozen ovarian tumor tissues and in 97 cases of archival ovarian serous epithelial tumors using a microdissection procedure. Hypermethylation statuses were examined vs clinicopathologic findings. RASSF1A promoter methylation rates in the various types of fresh ovarian tissues were as follows: serous cystadenoma (1/5), serous tumor of borderline malignancy (2/7), serous adenocarcinoma (4/10), mucinous cystadenoma (0/5), mucinous tumor of borderline malignancy (2/7), mucinous adenocarcinoma (3/6), transitional-cell carcinoma (1/3), clear-cell carcinoma (3/3), and malignant müllerian mixed tumor (3/3). In archived serous tumor tissues, RASSF1A promoter hypermethylation was detected in serous cystadenoma (1/6, 16.6%), serous tumor of borderline malignancy (20/41, 48.8%), and in serous adenocarcinoma (25/50, 50%). The status of RASSF1A hypermethylation in borderline tumors was found to correlate statistically with the presence of microinvasion (p=0.002), peritoneal implant (p<0.001), and bilaterality (p=0.019). The RASSF1A promoter hypermethylation was frequently found in borderline tumors and carcinomas, suggesting that RASSF1A promoter hypermethylation may be a useful molecular marker for the early detection of ovarian tumors.     

Fine-needle aspiration cytology of cystic ovarian lesions

A 4-year study (January 1986 through December 1990) was performed to evaluate the cytomorphologic features of cystic ovarian lesions. Fluid from 103 cases was obtained either during surgical removal of the ovary (48 cases; mean age 50 years) or by fine needle aspiration (55 cases; mean age 32 years). Of the 48 cystic lesions with histologic correlation, 30 (62.5%) were neoplastic and 18 (37.5%) were non-neoplastic. Ten (18%) of the 55 aspirates were unsatisfactory. The remaining 45 cases (82%) were benign, predominantly non-neoplastic entities which included follicle, corpora luteal, and endometriotic cysts. Neoplastic cystic lesions included serous, mucinous, and Brenner tumors, germ cell neoplasms, a sex cord-stromal tumor, and an undifferentiated carcinoma. Follicle and corpora luteal cysts were composed of loose cell clusters of granulosa cells and/or luteinized granulosa cells. Endometriotic cysts contained hemosiderin-laden macro-phages and endometrial cells. Serous and mucinous cystadenomas were composed of cohesive sheets and/or papillary clusters of epithelial cells. A cystic Brenner tumor showed sheets of cells with grooved nuclei, and a benign cystic teratoma contained mature squamous cells admixed with vacuolated cells of presumed sebaceous origin. Although the distinction between benign and malignant entities posed few diagnostic difficulties, borderline tumors could not be distinguished from well-differentiated cystadenocarcinomas. The results of this study indicate that the majority of cystic ovarian lesions can be diagnosed accurately on cytology. Cytologic evaluation of non-neoplastic ovarian cysts is important for women who want to retain their fertility and in the clinical management of women with neoplastic lesions. Diagn Cytopathol 1994; 11:9–14. © 1994 Wiley-Liss, Inc.     

Clinicopathological significance of DEK overexpression in serous ovarian tumors

To investigate the significance of DEK protein expression in ovarian lesions, a total of 113 ovarian serous tumors, including 62 serous cystadenocarcinomas and 19 serous borderline tumors, were studied on immunohistochemistry. For comparison, 32 benign serous tumors, including 12 serous papillary cystadenomas, 10 serous cystadenomas, and 10 serous surface papillomas, were also included. DEK was positive in 93.5% of serous cystadenocarcinomas (58/62), 63.2% of serous borderline tumors (12/19), and weakly positive in 15.6% of benign serous tumors (5/32). The strong positive signal was detected only in serous adenocarcinomas (80.6%, 50/62) and borderline tumors (21.1%, 4/19), but no serous benign tumors were strongly positive (0%, 0/32). Meanwhile, the strong positivity rate of DEK protein was significantly higher in grade 2 and grade 3 than in grade 1 ovarian cancers ( P < 0.05), but there was no significant association between DEK protein expression level and International Federation of Gynecology and Obstetrics (FIGO) stage of serous ovarian adenocarcinoma ( P > 0.05). In summary, DEK plays an important role in the progression of ovarian serous cancers. The detection of DEK protein expression should be useful for the diagnosis and prognosis of ovarian serous cancers, and DEK might be a useful molecular target for ovarian cancer therapy.     

Immunohistochemical localization of survivin in serous tumors of the ovary.

The aim of this study was to determine the immunohistochemical distribution of survivin in benign, borderline, and malignant serous tumors of the ovary. Survivin was localized by an indirect immunoperoxidase method in 42 cases of serous tumors of the ovary (15 cystadenomas, 15 borderline tumors, and 12 cystadenocarcinomas). Nuclear staining and cytoplasmic staining were separately scored. Cytoplasmic staining was detected in 27% of adenomas/borderline tumors and in 58% of carcinomas. Nuclear staining was detected in 87% of adenomas/borderline tumors but in only 42% of carcinomas. Although the differences in the intensity of cytoplasmic staining between adenomas and borderline tumors versus carcinomas were not significant, the differences in the intensity of nuclear staining between low-grade versus malignant tumors were significant. These findings suggest that survivin is widely expressed in benign, borderline, and malignant serous tumors but that nuclear localization of survivin is more common in benign or borderline tumors than in malignant serous tumors of the ovary. The molecular mechanisms that determine the subcellular distribution of this protein may reflect the role of survivin in the regulation of apoptosis during the processes of malignant transformation.     

Mucin histochemistry of ovarian mucinous cystadenomas expressing gastrointestinal characteristics

Fifty-four cases of ovarian mucinous cystadenoma, including twa cases containing ciliated cells representing the müllerian epithelial origin, one case admixed with serous adenoma component and six cases associated with mature teratoma, were examined for the demonstration of gastrointestinal characteristics using periodic acid-Schiff, alcian blue, galac-tose oxidase-Schiff, paradoxycal concanavalin A (ConA), Grimelius and Fontana-Masson stains. Of 41 endocervical-type mucinous cystadenomas not associated with teratoma, 34 cases (83%) showed ConA positivity, expressing gastrointestinal characteristics. As both cases with ciliated cells and the case with serous adenoma component exhibited ConA positivity, the ovarian surface epithelium is supposed to undergo mucinous metaplasia possessing gastrointestinal characteristics. As to the histogenesis of the ovarian mucinous tumors, the metaplasia theory is suggested.     

Prognostic significance of HLA-DR antigen in serous ovarian tumors

Abstract. The antigens encoded by the major histocompability complex (HLA-DR) are cell glycoproteins that play a fundamental role in the regulation of the immune response. The prognosis of ovarian cancer is dependent on the histological type and on the clinical stage at diagnosis. Our study reports the value of HLA-DR antigen as a prognostic marker of ovarian serous adenocarcinoma. We studied 31 cases of serous ovarian cystadenoma, 12 cases of serous ovarian borderline cystadenoma, and 39 cases of well-differentiated cystadenocarcinoma for HLA-DR monoclonal antigen. We also studied the T helper marker (CD4) in the tumor stroma of the relevant cases, given that it is now known that the dependence of immune responsiveness on the class II antigens reflects the central role of these molecules in presenting antigen to T helper cells. HLA-DR was expressed in 20 of 31 cystadenomas (64.5%), 4 of 12 borderline tumors (33.3%), and in 10 of 39 invasive carcinomas (25.6%). CD4 was expressed in 9 of 31 cystadenomas (29%), 5 of 12 borderline tumors (42%), and in 26 of 39 invasive carcinomas (67%). There was a statistically significant difference for the two examined antigens in cystadenomas (p<0.001) and invasive carcinomas (p<0.001), whereas there was no statistical difference in borderline tumors (p<0.5). The results showed decreased expression of HLA-DR and increased expression of CD4 as the lesion progressed to malignancy. The aberrant expression of HLA-DR by epithelial cells of cystadenomas, of borderline tumors, and of invasive adenocarcinomas agrees with the hypothesis of the adenoma/adenocarcinoma sequence. The immune attraction mechanism by low HLADR signaling seems to be of minor importance in the malignant and metastatic potential of serous ovarian tumors.     

卵巢粘液性囊腺瘤恶变和肠上皮化生的关系

Ninety cases of ovarian mucinous tumors were studied histologically and histochemically. Intestinal metaplasia was found in 48.2% (14/29) of benign. 73.6% (14/19) of borderline and 92.9% (39/42) of malignant mucinous cystadenomas. The differences between these three groups are statistically significant (P less than 0.01). Among 67 cases of intestinal metaplastic mucinous tumors, 43 contained argyrophil cells, and 36 contained argentaffin cells. The coexistence of intestinal metaplasic and uterocervical canal type epithelia was observed in 2/3 of borderline and 1/3 of malignant intestinal mucinous cystadenomas. In addition, there were 5 cases of borderline and 3 cases of malignant uterocervical canal type mucinous cystadenomas among the 90 cases. It is evident that the malignant transformation of ovarian mucinous cystadenoma was closely related to intestinal metaplasia. Anyhow, it seems not necessary for malignant transformation of all ovarian mucinous cystadenomas to pass through a stage of intestinal metaplasia: some of the malignant mucinous cystadenomas were considered probably originating from the uterocervical canal type epithelium.     

Overexpression of p21WAF1/CIP1 and MDM2 characterizes serous borderline ovarian tumors

Ovarian epithelial tumors are classically divided into benign, malignant, and borderline or of low malignant potential. It is controversial whether this last group of tumors should be considered benign or malignant. Expression of cell cycle markers has recently been linked to tumor behavior and response to treatment. It has been shown that one of the pathways through which the p53 gene controls the cell cycle is by transactivating p21WAF1/CIP1, a cyclin-dependent kinase (cdk) inhibitor. By inhibiting cdks, p21WAF1/CIP1 blocks the G-1 to S-phase transition in the cell cycle. p53 can be regulated by MDM2 (murine double minute-2) through direct inactivation or promotion of its cytoplasmic degradation. In an attempt to investigate the cell cycle checkpoint mechanisms of these tumors, we studied the expression of p53, Ki-67, MDM2, and p21WAF1/CIP1 by immunohistochemistry. We analyzed the expression of these proteins in 19 cystadenomas (8 serous and 11 mucinous), 40 borderline tumors (31 serous and 9 mucinous), and 18 serous carcinomas of the ovary. p21WAF1/CIP1 was expressed in 7 of 19 (37%) benign cystadenomas, 32 of 40 (80%) borderline tumors (93.5% of serous and 33% of mucinous), and in 9 of 18 (50%) serous carcinomas. Ki-67 was only weakly expressed in 8 of 19 (42%) benign cystadenomas, all borderline tumors showed Ki-67 staining in less than 50% of the cells, and 55% of serous carcinomas stained in more than 50% of tumor cells. p53 was absent in all but 1 of the cystadenomas, was expressed in 9 of 40 (22.5%) borderline tumors (25.8% of serous and 11% of mucinous), and in 10 of 18 (55%) carcinomas. All 11 implants of serous borderline tumors expressed p21WAF1/CIP1. Most serous borderline tumors expressed higher levels of MDM2 compared with the benign cystadenomas and carcinomas. Four of the serous borderline implants (40%) expressed MDM2. Coexpression of p21WAF1/CIP1 and MDM2 characterizes serous borderline tumors of the ovary and their implants, which suggests that these cell cycle control proteins are important in these tumors and may be related to tumor progression. Low expression of p53 protein in serous borderline tumors might be in part mediated by MDM2. This suggests that the p53 pathway is intact in most of these tumors, in contrast with carcinomas, in which high expression of p53 has been related to mutations of this gene.     

Inhibitory role of prohibitin in human ovarian epithelial cancer

Objectives: To characterize the exact individual roles of gonadotropins on ovarian epithelial carcinogenesis, an earlier study showed that prohibitin was significantly up-regulated by luteinizing hormone (LH). To further clarify the role of prohibitin in ovarian carcinogenesis and its association with LH, herein we studied the expression of prohibitin in various ovarian tissues including different developmental stages of ovarian epithelial tumors. Methods: A total of 135 samples were studied by immunohistochemistry. These included benign ovarian cases with follicles, ovarian surface epithelia and ovarian epithelial inclusions (OEI) (n=30), serous cystadenoma (n=14), serous borderline tumor (n=12), serous carcinoma (n=20), mucinous cystadenoma (n=10), mucinous borderline tumor (n=10), mucinous carcinomas (n=10), endometrioid carcinomas (n=12), poorly/undifferentiated carcinomas (n=5), and fallopian tube (n=12). Results: Strong and diffuse staining of prohibitin was detected in luteinized ovarian stromal cells, follicular cells, fallopian tube, and OEI with serous differentiation. A significantly higher prohibitin expression in luteinized stromal cells than in non-luteinized stromal cells was observed (P<.01). Within the ovarian epithelium, the level of prohibitin expression was basically negative in ovarian surface epithelia, but highly expressed in OEI. However, compared to the level of prohibitin expression in OEI, it showed a trend of gradual loss from benign ovarian tumors, to borderline tumors and to carcinomas (P<.0001). Compared to the serous tumors, epithelial tumors with mucinous differentiation showed a significant lower level of prohibitin (P<.0001). An inverse correlation was noted between prohibitin expression and cancer grade. It is interesting to note that a high prohibitin expression level was seen in the fallopian tube, which is similar to OEI. Conclusions: These data further suggest that prohibitin plays a tumor suppressing role, which is probably associated with LH mediated protection role against ovarian epithelial carcinoma. In addition to the tumor suppressive role of prohibitin, it also plays a role in cellular differentiation, which may be helpful to differentiate ovarian mucinous tumors from the tumors with serous differentiation in clinical settings. More importantly, our findings are supportive that the ovarian epithelial cancers, particularly the serous cancers including those precursors with serous differentiation are likely to be derived from fallopian tube instead of ovarian surface epithelia.     

Ovarian tumors of borderline malignancy (tumors of low malignant potential): a critical appraisal.

Before the designations borderline malignancy and low malignant potential were used, the surface epithelial-stromal tumors of the ovary were simply classified into benign and malignant categories. The introduction of the borderline category of tumors has been a great advancement in classification, because it has set apart from the general group of surface epithelial cancers a subgroup with a much better prognosis, stage-for-stage, than that of conventional ovarian carcinomas. Over the last 20 years, pathologists have learned to recognize the distinctive clinicopathologic features of serous borderline tumors as well as the adverse prognostic significance of the associated invasive peritoneal lesions, whether they may represent true implants or independent primary tumors. We have urged our surgical colleagues to search for the peritoneal lesions, and sample them meticulously, and advised our fellow oncologists not to administer adjuvant therapy to patients with noninvasive implants lacking malignant epithelial cells. There is now convincing evidence in the literature that the only fatal cases of serous borderline tumors are those associated with invasive implants, and chemotherapy is indicated only for these rare tumors. It has also been demonstrated that Stage I intestinal mucinous borderline tumors and noninvasive well-differentiated mucinous carcinomas both have an almost equally good prognosis. The current treatment for pseudomyxoma peritonei is still unsatisfactory, but we have recently learned that most of the mucinous ovarian tumors associated with pseudomyxoma peritonei are secondary to similar tumors of the appendix. In the remaining cases, however, the ovarian tumor appears to be responsible for the pseudomyxoma peritonei. Borderline tumors also exist in the endometrioid, clear cell, and Brenner surface epithelial categories, but these tumors have been too rare for clear delineation of their clinical and pathologic features. Recently, some investigators have proposed to abandon the borderline category and to return to the old benign-malignant classification system by dividing unevenly the borderline tumors into a larger group of atypical proliferative epithelial cystadenomas and a smaller category of recently described but still not well-characterized noninvasive carcinomas. In the author's opinion, such a recommendation is misleading because it ignores the possibility of rare but significant behavioral exceptions on each of these two groups of tumors. Furthermore, careful tumor staging is mandatory in both instances regardless of the type of terminology used. It is hoped that by keeping the borderline designation, knowledge on this group of ovarian tumors will continue to expand as it has been until now.     

ThinPrep evaluation of fluid samples aspirated from cystic ovarian masses

The cytological evaluation of ovarian cystic fluid using ThinPrep has not been reported. To determine the diagnostic accuracy of ThinPrep cytology in distinguishing between benign and nonbenign ovarian cystic lesions, we examined 65 fluid samples aspirated during intraoperative consultation with subsequent histologic correlation. One ThinPrep slide was prepared from each sample aspirated from surgically removed ovarian cystic masses and reviewed blindly by a panel of three cytopathologists. The parameters used in cytological evaluation were cellularity, cell types, cellular arrangement, and background. Four samples were acellular and excluded from the study. The consensus cytologic diagnoses were compiled for 61 cases which were assigned to one of the following diagnostic categories: negative for malignant cells (40 cases), atypical cytology (13 cases), and suspicious or positive for malignancy (8 cases). Histologic correlation of the cytological benign/negative cases showed that 26/40 (65%) were histologically benign and 14/40 were false-negative (35%, 5 carcinomas and 9 borderline tumors) with 10 of these cases being mucinous tumors. Most false-negative cytologic samples (11/14 or 79%) did not have an epithelial component. Of the 21 cytological nonbenign diagnoses (atypical/suspicious/positive), 15 (71%) were confirmed on histology (10 carcinomas and 5 borderline tumors). However, a nonbenign cytologic diagnosis was rendered in 6 histologically benign cases, including 2 serous cystadenomas, 1 mucinous cystadenoma, 1 serous cystadenofibroma, 1 endometriosis, and 1 corpus luteal cyst. The diagnostic sensitivity by ThinPrep evaluation of ovarian cystic masses is 81% (26/32) for benign and 52% (15/29) for nonbenign lesions. Our results concluded that ThinPrep examination of ovarian cystic fluid is not accurate in distinguishing benign from malignant cysts, given the significant number of false-negative diagnoses. Major contributing factors include sparse cellularity of the fluid samples and mucinous differentiation of the tumors.     

Alpha-lactalbumin in "common" epithelial tumors of the ovary. An immunohistochemical study

Alpha-lactalbumin (AL) has been widely used as an immunohistochemical marker for mammary carcinoma. The authors have investigated the AL reactivity of 36 unselected ovarian epithelial neoplasms using the avidin-biotin-peroxidase immunoperoxidase technic. In eight serous cystadenocarcinomas, both the primary neoplasm and its metastases were examined. Of the 36 tumors, 7 (19.4%) showed positive staining for AL and 5 were of the serous type. The serous tumors showed moderate to strong staining reaction. Three serous cystadenocarcinomas were AL positive in both primary and metastatic sites, while the remaining five were negative in all sites. One mucinous cystadenoma and one clear cell carcinoma showed weak to moderately positive AL reactivity. There was no good correlation between AL positivity and the presence of malignancy or between AL positivity and tumor grade. In view of the relatively high AL reactivity in ovarian neoplasms, it is advisable to exercise caution in interpreting the presence of AL positivity as specific marker for mammary carcinoma.     

S-100 protein in ovarian tumors. A comparative immunohistochemical study of 135 cases

The distribution of S-100 protein in 135 ovarian tumors, of which 127 were epithelial, was investigated using the immunoperoxidase method. S-100 protein has been demonstrated previously in tumors of various origins. The present study further revealed its characteristic distribution in common epithelial tumors of the ovary. S-100 protein was present in 69% of serous tumors (benign, 50%; borderline, 100%; malignant, 71%), as well as in the serous elements of serous & mucinous mixed tumors (30%). S-100 protein was also demonstrated in 25% of clear cell carcinomas and 29% of endometrioid carcinomas. Interestingly, none of the mucinous tumors were positive for S-100 protein. In addition, the expression of S-100 protein by epithelial ovarian tumors was compared with that of CA 125 and carcinoembryonic antigen (CEA). The distribution of S-100 protein was similar to that of CA 125, since both antigens were frequently present in serous tumors, although their expression patterns were different. On the other hand, S-100 protein-positive cases were almost negative for CEA or vice versa. Our observations indicate that demonstration of S-100 protein in common epithelial tumors of the ovary and comparison of S-100 protein distribution with that of CA 125 and CEA may further clarify the characteristics of common epithelial tumors of the ovary.     

S-100 Protein in Ovarian Tumors

The distribution of S 100 protein in 135 ovarian tumors, of which 127 were epithelial, was investigated using the im-munoperoxidase method. S 100 protein has been demonstrated previously in tumors of various origins. The present study further revealed its characteristic distribution in common epithelial tumors of the ovary. S 100 protein was present in 69% of serous tumors (benign, 50%; borderline, 100%; malignant, 71%), as well as in the serous elements of serous & mucinous mixed tumors (30%). S 100 protein was also demonstrated in 25% of clear cell carcinomas and 29% of endometrioid carcinomas. Interestingly, none of the mucinous tumors were positive for S 100 protein. In addition, the expression of S-100 protein by epithelial ovarian tumors was compared with that of CA 125 and carcinoembryonic antigen (CEA). The distribution of S 100 protein was similar to that of CA 125, since both antigens were frequently present in serous tumors, although their expression patterns were different. On the other hand, S 100 protein positive cases were almost negative for CEA or vice versa. Our observations indicate that demonstration of S 100 protein in common epithelial tumors of the ovary and comparison of S-100 protein distribution with that of CA 125 and CEA may further clarify the characteristics of common epithelial tumors of the ovary.