Expression of Tenascin in Human Cervical Cancer—Association of Tenascin Expression with Clinicopathological Parameters

Gynecologic Oncology 73, 415– 421 (1999) Article ID gyno.1999.5405, available online at http://www.idealibrary.com on

Expression of Tenascin in Human Cervical Cancer—Association of Tenascin Expression with Clinicopathological Parameters H. Pilch, M.D.,* U. Scha¨ffer, M.D.,* K. Schlenger, Ph.D.,* A. Lautz, M.D.,* B. Tanner, M.D.,* M. Ho¨ckel, M.D., Ph.D.,† and P. G. Knapstein, M.D.* *Department of Obstetrics and Gynecology, University of Mainz, Mainz, Germany; and †Department of Obstetrics and Gynecology, University of Leipzig, Leipzig, Germany Received November 17, 1998

mediated by immunological mechanisms, especially in terms of the infection with human papilloma virus (HPV) [2]. On the other hand the process of tumor invasion and metastasis requires complex changes in the normal epithelial cell– cell and epithelial cell–mesenchymal stroma interactions [3]. The receptors which mediate these interactions are likely to be important for local tumor progression and metastatic spread. Beside integrins and cellular adhesion molecules, extracellular glycoproteins like fibronection, laminin, or tenascin may be involved in the complex biological cascade of cancer invasion and metastasis [1, 4 – 6]. Tenascin is a glycoprotein component of the extracellular matrix with a six-armed disulfide-bonded macromolecular structure, consisting of three isoforms of the molecule [7, 8]. Tenascin is synthesized by fibroblasts and is a marker of morphogenesis of tendon and muscle in chicks. It is formed during the fetal development in organs such as the gut and kidney [9, 10]. It may have a role in cell adhesion and motility, guidance along cell migration pathways, shedding of epithelial cells from surface, promotion of cell growth, demarcation of tissue boundaries, and tissue modeling [12]. Tenascin is expressed in the mesenchyma at sites of epithelial–mesenchymal interactions during embryogenesis, wound healing in normal adult tissue, and in the stroma of several human tumors [9, 11]. Interactions between tumor cells and extracellular matrix are of importance in tumor invasion and metastasis and the expression of tenascin may suggest an altered cell matrix interaction that may facilitate epithelial tumor cell invasion during carcinogenesis and tumor progression [13]. The amino-acid sequence of tenascin is made up of epidermal growth factor (EGF)-like repetitions which could bind to EGF-receptors of tumor cells, suggesting that tenascin may play a crucial role in tumor invasion and metastasis [15]. Thus, tenascin is believed to be a multifunctional macromolecule that may be a key extracellular matrix component for growth regulation of epithelial and mesenchymal tissues. Different theories existing on the influence of tenascin for the tumor cell–stroma relationship have demonstrated the capability of mammary carcinoma cell

Objective. Tenascin is an extracellular matrix glycoprotein, relevant for embryonal and fetal development, which is reexpressed in the stroma of benign and malignant tumors. Little is known about the molecular interaction of tenascin during neoplastic transformation and tumor progression in cervical cancer. Method. We studied the expression of tenascin in normal tissue of the cervix uteri, cervical carcinoma in situ, and invasive cervical carcinoma in paraffin sections by immunohistochemistry using a monoclonal antibody. Tenascin immunoreactivity was compared with various prognostic parameters. Results. In normal cervical tissue (n 5 5) and in cervical carcinoma in situ (n 5 10) only vessel walls showed a weak tenascin cross-reactivity, whereas tenascin was not expressed in the epithelial layer or the underlying connective tissue. In invasive cervical carcinoma (n 5 89) tenascin expression was markedly increased. In 84% (n 5 75) of the cases examined a strong tenascin immunoreactivity was noted around and within the tumor cell nests. Sixteen percent (n 5 14) of infiltrating cervical carcinomas showed no tenascin immunoreactivity. A definite correlation was found between weak or no tenascin expression and slight desmoplastic mesenchymal reactivity (n 5 42/91%, P < 0.001), lymphatic space invasion (n 5 54/81%, P < 0.001), and lymph node metastases (n 5 30/77%, P < 0.05). Tenascin-positive patients had a significantly better prognosis than tenascin-negative patients (mean survival time of 56.5 6 4.1 months versus 31.9 6 5.6 months, P < 0.05). Conclusion. Based on these findings we discuss that the appearance of tenascin is an indicator of an adequate biological defense in cervical cancer patients. The tenascin staining may therefore be useful for detecting a subgroup of invasive cancer patients missing tenascin reactivity with alterations of stromal defense and a poorer prognosis. © 1999 Academic Press Key Words: tenascin; human cervical cancer; immunohistochemistry.

INTRODUCTION Cervical cancer growth and its invasion and metastatic progression is highly dependent on the local interactions with adjacent cells and tissue [1]. Some of these interactions are 415

0090-8258/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

416

PILCH ET AL.

lines to stimulate tenascin production in normal fibroblasts by secretion of TGF-b, so tenascin expression in tumors may be a consequence of a paracrine effect by which tumor cells force the surrounding stroma cells to organize the so-called tumor stroma [9, 14]. To our knowledge, there have been no reports about tenascin expression in human cervical cancer. In this retrospective study, we assessed the expression of tenascin in normal tissue of the uterine cervix, in cervical carcinoma in situ, and in invasive cervical carcinoma. We also evaluated its relationship to clinicopathological parameters using standard immunohistochemistry. MATERIAL AND METHODS A total of 143 cervical cancer patients who had been surgically treated by radical hysterectomy and pelvic lymphadenectomy in the Department of Obstetrics and Gynecology of the University of Mainz between 1988 and 1993 were reviewed. Follow-up was obtained from tumor registry at the Department of Ob/Gyn of the University of Mainz and supplemented with records from the primary physician. Pathology The histopathology was available for review in 122 cases. Histological classifications were done independently by two pathologists. The following histological types of cervical cancer were identified: squamous cell carcinoma, adenocarcinoma, and adenosquamous carcinoma. Squamous carcinomas are characterized by unique histopathological forms of growth pattern, which are not directly comparable with those of adenocarcinomas or adenosquamous carcinomas. This study is partly aimed at evaluating the growth patterns of squamous cell carcinomas of the uterine cervix and correlating the results with tenascin immunoreactivity, so we did not include patients with adenocarcinomas or adenosquamous carcinomas. All reviewed cases of squamous cell carcinomas (n 5 89) were considered for further immunohistochemical and statistical analysis. All resected specimens were fixed in 10% formalin and routinely processed for paraffin embedding. In this study the tumor was totally paraffin-embedded and histologically examined. Five-micrometer-thick sections made from each block were stained with hematoxylin and eosin. All cervical carcinomas were classified according to WHO classification [16]. All pathological diagnoses and classifications were based on FIGO classification: 28 (31%) were stage IB, 54 (61%) were stage II A 1 B, and 7 (8%) were stage III A [17]. This majority distribution of patients with stage II cervical cancer may be explained by the position of our department as regional center for surgical treatment of advanced cervical cancer. Histopathologic Characteristics The specimens from radical surgery were processed in giant (cross) sections for histological examination to verify tumor

extension and depth of cervical invasion. The presence of tumor cell clusters within an endothelial-lined space was regarded as lymphatic space invasion. The stromal component of the carcinomas examined was qualified as none or weak (intratumorous mesenchyma less than 25%/giant section) and intense or strong (intratumorous mesenchyma more than 25%/ giant section). The stromal component means the desmoplastic fibrocollagenic mesenchyma within the carcinoma is a sign of host reactivity to neoplastic invasion. Especially regarding the peripheral invasion front, the cancerous growth pattern was noted as diffuse (more than 50% disseminated infiltration with single cell dissolution/giant section) and cohesive (more than 50% tumor islands/giant section). Inflammatory reaction was qualified as weak (less than 50 inflammatory cells/HPF) and intense (more than 50 inflammatory cells/HPF). Tenascin Immunohistochemistry Immunostaining was performed with the monoclonal antibody TN2 (Dako-Tenascin). TN2 was raised in immunized mice and binds to activated fibroblasts of the desmoplastic tumor stroma. We used bovine serum containing TN2 at a dilution of 1:50 in phosphate-buffered saline (PBS) for immunostaining. A representative tissue block was selected from each tumor and used for immunohistological examination. Five-micrometer sections were cut from paraffin blocks, mounted on glass slides, and then deparaffinized in xylene and alcohol, according to standard procedures. The sections were then boiled in 0.01 mol/L sodium citrate buffer (pH 6.0) for 15–30 min. Following an incubation with normal goat serum (2–5% in PBS; Vector Laboratories, Burlingame, CA) for 30 min to reduce background staining, the sections were incubated with the primary TN antibody at 4°C overnight. After being thoroughly washed in PBS, the specimens were incubated with biotinylated secondary antibodies to rabbit (Dako, Glostrup, Denmark; diluted 1:200 in PBS) or mouse immunoglobulins (Dako, diluted 1:300 in PBS) at room temperature for 1 h. After being washed in 0.05 mol/L Tris (pH 7.5)– 0.05 mol/L NaCl buffer, streptavidin– biotinylated alkaline phosphatase complex (Dako) was placed on the section for 30 min. The bound complexes were visualized with DAB. The slides were then lightly counterstained in Mayer’s hematoxylin. Sections of breast cancer known to be immunoreactive with anti-tenascin antibodies were used as positive controls. Negative controls were treated in the same way as described above, but the primary anti-tenascin antibody was replaced with PBS. Immunohistochemical Analysis The sections were evaluated without knowledge of clinical outcome. The semiquantitative evaluation of TN expression was performed by means of a score of immunoreactivity (IRS, min 5 0 max 5 12), which was adapted from a method of Remmele and Stegner, who originally established IRS as a measure for receptor density in breast cancers [18]. This im-

417

TENASCIN EXPRESSION IN HUMAN CERVICAL CANCER

TABLE 1 Qualitative and Semiquantitative Assessment of Tenascin Expression in Normal Tissue of the Uterine Cervix, in Carcinoma in Situ, and Invasive Carcinoma Anti-tenascin Mab TN2 neg

Anti-tenascin Mab TN2 pos

IRS 0–4

IRS 6–12

Histology

N

N

%

N

%

N

%

N

%

Normal cervical tissue Carcinoma in situ Invasive cancer

5 10 89

5 10 14

100 100 16

0 0 75

0 0 84

51

57

38

43

Note. IRS, immunoreactivity score.

munoreactivity score was defined as the average number of staining signals in a high-power field (magnification 3200) multiplied by the intensity of the immunostaining reaction (arbitrarily defined 1 5 1; 11 5 2; 111 5 3). The anti-TN reactivity was scored as negative when the cancer surrounding stroma completely lacked immunostaining. The immunoreactivities of each tissue section were reviewed (H.P., A.H.) and similar results were obtained. By this procedure, if tenascin was present it was expressed immunohistochemically in different intense thin bands in dense cancer–mesenchymal junctions of cancerous areas. As positive control a defined specimen of breast cancer was used. The monoclonal antibody TN2 showed some cross-reactivity with vascular muscles, which was used as an additional internal control [19]. Histological sections of normal cervical tissue were used as negative controls. The results of the immunohistochemical analysis were correlated with pathohistomorphological parameters such as tumor size, pathological tumor stage, lymphatic space invasion, nodal status, grade of desmoplastic stromal reaction, and growth pattern. Statistical Analysis Statistical analysis was performed with SAS software, Version 6.04. For comparison, Fisher’s exact test was applied. Adjusted survival distribution was calculated by the productlimit method of Kaplan and Meier. Univariate analysis was performed using the log-rank test. Multivariate analysis was performed using the Cox proportional hazards model. Probability values with P , 0.05 were considered to indicate statistical significance. RESULTS Tumors ranged in size from 1.0 to 6.0 cm (mean, 3.9 cm). Forty-four tumors (49%) were smaller than 4 cm in max diameter, whereas 45 tumors (51%) were larger than 4 cm in max diameter. Lymphatic space invasion was identified in 66 (74%) cases. Pelvic lymph node metastases were observed in 39 (43%) cases. A strong stromal component could be evaluated in 43 (48%) cases. The review of growth pattern revealed

56 (62%) cases with a predominant diffuse tumor cell invasion, whereas 33 (38%) cases were predominantly characterized by cohesive tumor cell complexes. Nuclear grading was not reviewed. In 53 (60%) cases a weak inflammatory reaction was found in contrast to 36 (40%) cases with an intense lymphoid cell infiltration. Results of immunohistochemistry with TN2, an anti-tenascin antibody, are summarized in Table 1. Immunoreactivity was exclusively confined to the intra- and peritumoral mesenchymal stroma with a fine fibrillary staining pattern (Fig. 1). The connective tissue around the tumor was completely negative in all cases examined. In the normal cervical tissue (n 5 5) and in cervical carcinoma in situ (n 5 10) only vessel walls showed a weak tenascin cross-reactivity, whereas tenascin was not expressed in the epithelial layer or the underlying connective tissue. In invasive cervical cancer a positive tenascin staining was observed in 84% (n 5 75) of the primary tumors examined. A semiquantitative assessment of tenascin expression by use of an immunoreactivity score as described under Material and Methods revealed a range from IRS 0 to 12, recording no or weak (IRS 0 – 4) and intense (IRS 6 –12) tenascin immunoreactivity (Table 1). Correlation of Tenascin Immunohistochemistry with Pathoanatomic Parameters The results are summarized in Table 2. Eighty-nine squamous cell carcinomas of the cervix were surgically treated by radical hysterectomy and pelvic lymphatic nodectomy. Lymphatic space invasion was identified in 66 (74%) of the cervical carcinomas reviewed. However, 23 (26%) tumors showed no evidence of lymphatic tumor cell spread. In the majority of the tumors (n 5 54, 81%) with lymphatic space invasion none or only weak tenascin expression (IRS 0 – 4) was observed. In contrast, the majority of the tumors (n 5 18, 78%) without lymphatic space invasion showed a positive immunoreactivity with a strong staining pattern (IRS 6 –12). The correlation between tenascin expression and lymphatic space invasion was statistically significant (P , 0.001, Fisher’s exact test). Thirtynine (43%) of the primary tumors examined showed pelvic

418

PILCH ET AL.

FIG. 1. Squamous cell carcinoma of the uterine cervix with a strong tenascin immunoreactivity (IRS 5 12), 2003 magnification.

lymph node metastases. The majority of these cases (n 5 30, 77%) revealed no or only weak immunoreactivity for the anti-tenascin antibody (IRS 0 – 4), whereas 35 (70%) of the tumors without lymph node metastases could be evaluated with a strong immunohistochemical staining pattern (IRS 6 –12). The correlation between tenascin immunoreactivity and nodal status was also significant (P , 0.05, Fisher’s exact test). Similarly, a significant relationship between the grade of desmoplastic stromal reaction and tenascin immunoreactivity was found. The majority of the carcinomas with no or only weak desmoplastic stromal component (n 5 42, 91%) was characTABLE 2 Correlation of Tenascin Immunohistochemistry with Pathoanatomic Parameters Tenascin expression LSI negative (n 5 23) LSI positive (n 5 66) LNM negative (n 5 50) LNM positive (n 5 39) DSC negative (n 5 46) DSC positive (n 5 43) Cohesive (n 5 33) Diffuse (n 5 56)

IRS 0–4 (n %) 5 54 15 30 42 17 14 45

22 81 30 77 91 39 42 80

IRS 6–12 (n %) 18 12 35 9 4 26 19 11

78 18 70 23 9 61 58 20

P value

terized by no or only weak tenascin expression (IRS 0 – 4), in contrast to the majority of the tumors with a strong desmoplastic stromal component (n 5 26, 61%), which produced a strong immunoreactive staining pattern (IRS 6 –12) (P , 0.001, Fisher’s exact test). In the same way there was a significant correlation between tenascin expression and growth pattern: cervical carcinomas with predominantly cohesive growth pattern showed significantly more often an intense tenascin immunoreactivity (n 5 19, 58%) than the majority of tumors with a diffuse tumor cell spread (n 5 45, 80%), which were identified by no or only weak tenascin expression (IRS 0 – 4) (P , 0.001, Fisher’s exact test). A significant relationship could also be evaluated between the grade of cancerous stromal content and growth pattern, since intense stromal component and cohesive respective compact tumor growth pattern were definitively correlated (P , 0.01, Fisher’s exact test). However, there was no relationship between tenascin immunoreactivity and tumorsize-respective pathological tumor stage and grade of inflammatory reaction (n.s.).

0.001 0.05

Correlation between Tenascin Immunoreactivity and Survival

0.001

All 89 patients were observed with a median follow-up period of 62 months (range 7–92 months). Tenascin-positive cancer patients had a definitely better clinical outcome than tenascin-negative tumor patients (P , 0.05, log-rank test, Fig. 2). Tenascin-positive patients showed a mean survival time of

0.001

Note. LSI, lymphatic space invasion; LNM, lymph node metastases; DSC, desmoplastic stromal component; IRS, immunoreactivity score.

TENASCIN EXPRESSION IN HUMAN CERVICAL CANCER

419

FIG. 2. Overall survival probabilities calculated with the Kaplan–Meier method for 89 cervical cancer patients treated with primary surgery stratified for tenascin expression (P , 0.05, log-rank test).

56.5 6 4.1 months in contrast to tenascin-negative patients with a mean survival time of 31.9 6 5.6 months. Multivariate analysis revealed FIGO stage (P , 0.001), lymphatic space invasion (P , 0.001), and nodal status (P , 0.001) to be the only three independent prognostic factors for overall survival, whereas tenascin expression was not significant as found with univariate analysis (n.s.). DISCUSSION The present study demonstrates that tenascin is expressed in the desmoplastic fibrous stroma of cervical cancer. There is no evidence of tenascin reactivity in benign or malignant epithelium. It has been reported that tenascin is expressed in the stromal tissue of various human malignancies such as breast cancer, lung cancer, gastric cancer, colon cancer, and malignant melanoma [6, 19 –22]. To our knowledge there are no other previous studies concerning tenascin expression in cervical cancer. Other studies suggest that tenascin expression may be important for embryonal morphogenesis, but it may also play a crucial role for epithelial–mesenchymal stroma interaction in adult tissue and cancer [9, 23]. No tenascin expression in epithelial tumor cells could be observed, which is in accordance with the results of previous studies and our own investigation [24 –26]. However, in vitro studies performed by Chiquet have shown that exogenously added tenascin may influence the growth of breast cancer cell lines. Moreover, it is thought that tenascin in tumor tissue is synthesized by stromal fibroblasts, which are induced by the tumor cells [27]. Therefore, tenascin has been discussed as a consequence of a para-

crine effect by which tumor cells force the surrounding stroma cells to organize the desmoplastic fibrous stroma, which may play a crucial role in local tumor propagation [8]. In vitro studies by Probstmeier revealed that the adherent growth of the human colon cancer cell line HAT-29 can be inhibited by a tenascin-containing substrate, supporting the theory of a major fibronectin antagonizing role of tenascin by Chiquet-Ehrisman [29, 30]. In the present study tenascin was observed to be expressed in the stromal tissue of cervical cancer, but not in the normal cervical tissue or in carcinoma in situ and tenascin reactivity correlates significantly with the degree of desmoplastic fibrous stroma. Most of the cervical carcinomas in this study examined exhibited a positive stromal tenascin staining whereas a prominent tenascin staining pattern was found in carcinomas with dense desmoplastic stroma reaction and cohesive tumor growth pattern. Weak or no tenascin staining was observed in carcinomas with loose or no desmoplastic mesenchymal reaction and disseminated tumor growth pattern. The correlation between the amount of desmoplastic stroma formation, tenascin expression, and tumor growth pattern was significant. Furthermore, based on our pathological definition of lymphatic space invasion the tenascin reactivity was correlated with high significant prognostic parameters such as lymph node metastases as well as lymphatic space invasion, knowing well that the detection of lymphatic space invasion in pathological specimens continues to be a controversial issue. Therefore, we suggest that tenascin may not only stimulate tumor growth and progression, but it could also be associated with a defensive role by virtue of the formation of dense desmoplastic fibrous

420

PILCH ET AL.

stroma, which shows prominent tenascin content [15]. Other authors such as Tetsuji et al. and Sugawara et al. reported similar results of tenascin expression in breast and colonic cancer supporting our hypothesis [31, 32]. The missense of such a stromal defense in cancer patients may be explained by a decreased proliferation of probably altered stromal fibroblasts within the cancerous connective tissue. These results indicate that tenascin could be associated with malignant transformation and local tumor spread of cervical cancer, so tenascin staining in surgical specimens might be used to predict the metastatic or aggressive potential of cervical cancer in order to select those patients with invasive cervical cancer who have a poorer prognosis, represented by lymphatic space invasion or lymph node metastases. Similar results are reported by Sugawara, who found that colon carcinomas examined a significant correlation between strong tenascin reactivity and lymph node negative status as well as a better clinical outcome [32]. In our study of 89 cervical cancer patients, Kaplan–Meier survival tables revealed a definitely better clinical outcome for cancer patients with a positive tenascin reactivity. Multivariate analysis, however, revealed only conventional histomorphological parameters as independent prognostic factors, so further studies must clarify the prognostic impact of tenascin expression in cervical cancer. In breast cancer the prognostic value of tenascin staining is controversely discussed [23, 31]. A comparative analysis of the tenascin expression and established prognostic factors in human breast cancer by Moch et al. gave no significant association between tenascin expression and strong prognostic parameters like lymph node metastases, blood vessel invasion, tumor necrosis, or DNA ploidy type [23]. Other authors observed a significantly better prognosis in tenascin-positive breast cancer patients [31]. These discrepancies may be explained by the use of different kinds of monoclonal antibodies and by different sensitivities of the antibodies depending on tissue preparation. It has been suggested that tenascin production is modified by local intratumorous inflammation, speculating that tenascin is responsible for local fine tuning of the immune response and may protect normal or malignant tissues from the effect of T-cells. In breast cancer Moch and co-workers noted a significant relationship between tenascin expression and the degree of inflammation, so that increased tenascin expression in areas of intense inflammatory infiltration indicates a modifying influence of or an interaction between inflammatory cells and tenascin [23]. In the present study there is no evidence of a relationship or interaction between tenascin expression and inflammatory infiltration. This is perhaps due to the situation of HPV infection in most cases of cervical cancer with a special influence of the intratumorous lymphoid response (publication in preparation). Finally, in the current study there was no evidence of tenascin reactivity in the connective tissue of the normal cervix uteri or carcinoma in situ, which is in contrast to invasive cervical cancer, so the application of tenascin staining in surgical specimen might be diagnostically helpful in terms of the histolog-

ical differentiation of squamous intraepithelial lesion and cervical cancer with early stroma invasion. In conclusion we suggest that tenascin could be associated with the malignant transformation and local tumor progress of cervical cancer, in so far that a prominent tenascin reactivity may reflect a strong mesenchymal tumor defense with a lower rate of lymphatic space invasion, and lymph node metastases and a better clinical outcome. The pathophysiological and pathomorphological aspects of tenascin in cervical cancer have not yet been clearly evaluated and should be further investigated. REFERENCES 1. Speiser P, Wanner C, Tempfer C, Mittelbo¨ck M, Hanzal E, Bancer D, Gitsch G, Reitaller A, Kainz C: CD44 is an independent prognostic factor in early-stage cervical cancer. Int J Cancer 74:185–188, 1997 2. Falk S, Falk C, Stutte HJ: Local immunoreactions in CIN and cervical carcinoma—a histological and immunohistological study. Geburts u Frauenheilk 51:39 – 44, 1991 3. Aufderheide E, Chiquet-Ehrismann R, Ekblom P: Epithelial–mesenchymal interactions in the developing kidney lead to expression of tenascin in the mesenchym. J Cell Biol 105:599 – 608, 1987 4. Christensen L, Nielsen M, Anderson J, Clemmensen I: Stromal fibronectin staining pattern and metastasizing ability of human breast carcinoma. Cancer Res 48:6227– 6233, 1988 5. Fujita S, Suzuki H, Kinoshita M, Hirohashi S: Inhibition of cell attachment, invasion and metastasis of human carcinoma cells by anti-integrin b1 subunit antibody. Jpn J Cancer Res 83:219 –229, 1984 6. Mackie EJ, Chiquet-Ehrismann R, Pearson CA, Inaguma Y, Taya K, Kawarada Y, Sakatura T. Tenascin is a stromal marker for epithelial malignancy in the mammary gland. Proc Natl Acad Sci USA 84:4621– 4625, 1987 7. Chiquet-Ehrismann R, Mackie EJ, Pearson CA, Sakatura T: Tenascin: an extracellular matrix protein involved in tissue interactions during fetal development and oncogenesis. Cell 47:131–139, 1986 8. Chiquet-Ehrismann R, Matsuoka Y, Hofer U, Spring J, Bernasconi C, Chiquet M: Tenascin variants: differential binding to fibronectin and distinct distribution in cell cultures and tissues. Cell Regul 2:927–938, 1991 9. Erikson HP, Bourdon MA: Tenascin: an extracellular matrix protein prominent in specialized embryonic tissues and tumours. Annu Rev Cell Biol 5:71–92, 1989 10. Chiquet M, Fambrough DM: Chick myotendinous antigen. A monoclonal antibody as a marker for tendon and muscle morphogenesis. J Cell Biol 98:1926 –1936, 1984 11. Mackie EJ, Halfter W, Liverani D. Induction of tenascin in healing wounds. J Cell Biol 107:2757–2769, 1988 12. Inaguma Y, Kusakabe M, Mackie EJ, Pearson CA, Chiquet-Ehrismann, Sakatura T. Epithelial induction of stromal tenascin in the mouse mammary gland: from embryogenesis to carcinogenesis. Dev Biol 128:245– 255, 1988 13. Lee SK, Park SC, Chi JG, Sakamato F, Shresta P, Mori M: Expression of tenascin in hamster buccal pouch mucosa during experimental carcinogenesis. Oral Oncol Eur J Cancer 31B(3):188 –192, 1995 14. Pearson CA, Pearson D, Shibahara S, Hofsteenge J, Chiquet-Ehrismann R: Tenascin: cDNA cloning and induction by TGF-b. EMBO J 7:2977–2981, 1988 15. Jones FS, Burgoon MP, Hoffmann S, Crossin KL, Cunnigham BA, Edel-

TENASCIN EXPRESSION IN HUMAN CERVICAL CANCER

16.

17. 18. 19.

20.

21.

22.

23.

mann GM: A cDNA clone for cytotactin contains sequences similar to epidermal growth factor-like repeats and segments of fibronectin and fibrinogen. Proc Natl Acad Sci USA 85:2186 –2190, 1988 WHO: Histological Typing of Female Genital Tract Tumours, in Scully et al. (eds): International Histological Classification of Tumours, 2nd ed, 1994 Hermanek P, Sobin LH: TNM Classification of Malignant Tumours. Fifth ed, New York, Springer and International Union Against Cancer, 1997 Remmele W, Stegner HE: Rezeptorbestimmung in Mammakarzi¨ rztebl Ausgabe A 48:3359, 1986 nomgewebe. Dt A Ikeda Y, Mori M, Kajiyama K, Haraguchi Y, Sasaki O, Sugimachi K: Immunohistochemical expression of tenascin in normal stomach tissue, gastric carcinomas and gastric carcinoma in lymph nodes. Br J Cancer 72:189 –192, 1995 Soini Y, Paako P, Nuorva K, Kamel D, Linnala A, Virtanen I, Lehto VP: Tenascin immunoreactivity in lung tumors. Am J Clin Pathol 100:145– 150, 1993 Sakai T, Kawakatsu H, Hirota N, Yokoyama T, Sakakura T, Saito M: Specific expression of tenascin in human colonic neoplasms. Br J Cancer 67:1059 –1064, 1993 Natali PG, Nicotra MR, Bigotti A, Botti C, Castellani P, Risso AM, Zardi L: Expression and production of tenascin in benign and malignant lesions of melanocyte lineage. Int J Cancer 46:586 –590, 1990 Moch H, Torhorst J, Du¨rmu¨ller U, Feichter GE, Sauter G, Gudat F: Comparative analysis of the expression of tenascin and established prognostic factors in human breast cancer. Path Res Pract 189:510 –514, 1993

421

24. Chiquet-Ehrismann R, Kalla P, Pearson CA: Participation of tenascin and transforming growth factor beta in reciprocal epithelial–mesenchymal interactions of MCF cells and fibroblasts. Cancer Res 49:4322– 4325 25. Howeedy AA, Virtanen I, Laitinen L, Gould NS, Koukoulis GK, Gould VE: Differential distribution of tenascin in the normal, hyperplastic and neoplastic breast. Lab Invest 63:798 – 806, 1990 26. Koukoulis GK, Gould VE, Bhattacharyya A, Gould IP, Howeedy AA, Virtanen I: Tenascin in normal, reactive, hyperplastic and neoplastic tissues: biologic and pathologic implications. Human Pathol 22:636 – 643, 1991 27. Ekblom P, Aufderheide E: Stimulation of tenascin expression in mesenchyme by epithelial–mesenchymal interactions. Int J Dev Biol 33:71–79, 1989 28. Engel J: EGF-like domains in extracellular matrix proteins: localized signals for growth and differentiation? FEBS Lett 251:1–7, 1989 29. Probsmeier R, Martini R, Schachter M: Expression of J1/tenascin in the crypt villus unit of adult mouse small intestine: implications for ist role in epithelial cells shedding Development 109:313–321, 1990 30. Chiquet-Ehrismann R, Kalla P, Pearson CA, Beck K, Chiquet M: Tenascin interferes with fibronection action. Cell 53:383–390, 1988 31. Tetsuji S, Tomoo K, Airo T, Yoshinori H, Takehiko H, Koshiro H, Sokokichi M. Tenascin positivity and the survival of patients with invasive breast carcinoma. J Surg Res 55:295–297, 1993 32. Sugawara I: Reduced tenascin expression in colonic carcinoma with lymphogenous metastasis. Invas Metast 11:325–331, 1991