Cancer Letters, 59 (1991) 171-175 Elsevier Scientific Publishers Ireland
Oncogene
171 Ltd
expression
and prognosis
in cervical cancer
R. Pillai Department
Centre,
of Family and Community
Trioandrum
Medicine.
Unioersity
of Arizona,
Tucson. Arizona (U.S.A.j
and Regional
Cancer
(India)
(Received 17 April 1991) (Revision received 20 June (Accepted 21 June 1991)
1991)
Summary
Introduction
Cancer of the uterine cervix accounts for 80 - 85 93 of all female genital tract malignancies in India and also remains a major problem for oncologists in other parts of the world. A major concern regarding the disease is the lack of specific tumor markers for early detection,
Cancer of the uterine cervix remains a major problem for oncologists both in developed and developing countries. It is estimated that thirteen thousand women in the United States develop invasive cervical cancer annually [ 131. However the figure is much higher in other countries. In India for example, cervical cancer accounts for approximately 20 -50% of all cancers and 80-85% of all gynecologic cancers [6]. Race, socioeconomic status and numerous other risk factors have been recognized, many of them relating to sexual behavior. It is also believed that human papilloma viruses are important etiological factors and their detection in precancerous lesions is a risk factor for malignant transformation [8]. If these viruses do have a role in initiation of cervical cancer, then other co-factors may be needed for malignant transformation. One such factor of concern is the role of oncogenes. Cellular oncogenes are DNA sequences found in the genome of normal and neoplastic tissue [2]. Even though their exact functions are just beginning to be understood, neoplastic transformation may be related to oncogene activation caused by point mutations, chromosomal translocations, deletions and gene amplification [21]. Such abnormalities in oncogene expression have been reported in a high number of cancer cell lines and in many fresh tumor samples [2,21].
for accurate prediction of biological behaviour and for accurate assessment of prognosis. A new and exciting answer to this issue may now be available with the description of specific oncogenes and oncoproteins associated with this malignancy. On a clinical level these genes and their products may allow us to improve our
understanding of disease etiology, and prouide diagnostic, prognostic and more precise individual characterization of therapeutic tumors. This paper discusses the possibilities of using altered expression of oncogenes and their products in neoplastic tissue as markers for the diagnosis and prognosis of cervical cancer. These data support the view that detailed analysis
of
potential
to predict
such
gene
expression
cervical cancer; Keywords: prognosis; tumor marker Correspondence to: 21/842, Nedumcaud,
has
the
tumor behauior.
oncogenes;
Dr. Radhakrishna Pillai, Karthika, TC Karamana. Trivandrum 695 002, Kerala
state, India.
0304-3835/91/$03.50 Published and Printed
0 1991 Elsevier Scientific in Ireland
Publishers
Ireland Ltd
172
Prognostic
features
in
cervical
cancers The biological behavior of invasive cervical cancer is not always predictable. Even when the lesion is localized to the cervix, 15 - 20% of patients have recurrences [11,22]. Many clinical prognostic factors have been associated with high risk of relapse including lymph node involvement, tumor size, histological type and grade of differentiation [21]. Lymph node involvement appears to be a most important factor, since patients with early stages of cervical cancer but who also have para aortic and multiple lymph node metastases are likely to rapidly develop distant metastases [ 11,221. However, in many cases with involved lymph nodes there are no recurrences and conversely the absence of positive nodes does not ensure disease free status. The problem therefore for oncologists is to try and accurately predict those tumors which will relapse in order to modify or modulate treatment protocol. It is thus important to establish biological markers which can be associated with disease course. One such recent advance was the description of the immunological staging system where immunological functions could be correlated well with disease course [ 121. Since oncogenes are associated with proliferation control and cancer [2,21], they offer a new parameter for estimation of prognosis and disease course besides providing insights into carcinogenic mechanisms. Alterations in the c-iia-ras cervical cancer
gene
in
One of the oncogene families most commonly implicated in a variety of human cancers is the ras gene family [ 11. These genes are related structurally and encode a group of closely related 21-kDa proteins termed p21 [l]. There are several reports describing the relationship between the t-as oncogene and its product ~21 and the process of early carcinogenesis in different cancers such as colon, bladder and breast [1,7,20]. Ohuchi et al. [lo]
described the higher grade of patho-histology in breast cancer to correlate with more freexpression of ~21 during carquent cinogenesis. Michelassi et al. [7] have suggested that the levels of the ras oncogene protein increases with the malignant potential of benign human colonic conditions and thus, high risk patients should be monitored and treated before actual malignant transformation occurs. These results suggest that the ras oncogene and its ~21 protein may play an important role in malignant transformation. Current data also support a role for the ras oncogene in cervical cancers. To examine the correlations between ras gene expression and the development of cervical cancer, Sagae et al. [18] studied the reactivity of cervical intraepithelial neoplasia (CIN) and microinvasive lesions of the cervix by using an anti ras p21 monoclonal antibody, rp35. The frequency of positive ~21 staining increased with higher grades of malignancy from 17.9% in CIN-1 to 28.9% in CIN-2 and 53.9% in CIN-3 while it was 50% in microinvasive carcinomas. Furthermore ten cases of lesions that regressed during a l-year follow-up period, 20% were positive for ras ~21 while 50% of the 14 lesions that progressed to invasive and higher grade lesions were positive for ~21. The data thus showed that the frequency of p2 1 increased with higher grades of dysplasia suggesting that ~21 positivity increases during early carcinogenesis from CIN to invasive carcinoma. The same authors in another study [19] using the same anti ras ~21 antibody studied reactivity of 170 squamous carcinomas of the uterine cervix belonging to different histological grades. Overexpression of ~21 was noted in 57.1% of keratinising tumors (K type) and 54.2% of non keratinizing type (LNK type) but only 38.7% of the small cell type (S type). Statistical analysis of the p21 expression and patient prognosis showed the p21 positive K and LNK types to have poorer prognosis whereas the positive S type showed better prognosis in comparison with the negative cases. The authors concluded that the expression of ras ~21 is a prognostic indicator for cervical
173
cancers but the mode of its prognostic correlation was dependent on tumor histology. Another finding of clinical significance was the report showing a correlation between elevation of ~21 expression in cervical carcinomas and incidence of lymph node metastases [4]. This concordant relationship in cervical carcinomas suggests that enhanced production of ~21 occurs late in the oncogenic process and may relate more to the malignant phenotype than to the early process of malignant transformation. However the number of patients in the study was too small to allow any firm conclusion. Alterations in the c-myc oncogene cervical carcinomas
in
Previous studies [14,15,17] of the c-myc oncogene in a limited number of cervical tumors had shown that this oncogene was activated in most evolved cancers, suggesting that the c-myc activation was associated with tumor progression. Further studies on larger number of patients showed the gene to be amplified from 3 to 30 times in 21% of the tumors [16]. Similar results were seen from other studies [9] and no amplification of c-myc was noticed in DNA preparations from normal cervical epithelium or from lymphocytes. Amplification of c-myc was found to be much higher in Stage III and IV tumors indicating that the gene amplification is significantly associated with advanced cervical cancers [16]. A similar finding [9] also suggested that patients whose tumors showed c-myc overexpression at levels 4 - 20 times that of normal tissues had an g-fold greater incidence of early relapse than other patients. The c-myc expression even outweighed nodal status in this study as a prognostic factor. However different findings have also been reported. Quantitation by flow cytometry of the c-myc oncoprotein content in archival biopsies of uterine cervix neoplasia showed normal biopsies to have higher levels than carcinomas [5], a finding difficult to explain in light of most studies showing increased expression of the c-myc gene. Since
the function of the c-myc oncoprotein is not yet understood, the functional significance of this observation is not known. Nevertheless, a number of possibilities exist for such a finding. These include a post-translational protein modification or c-myc gene mutation in malignant cells, increased turn over of the protein which is known to have a short half-life in stimulated cells [3] and a possible increase in the susceptibility of the protein to proteolysis in neoplastic cells during preparation of tissue for the assay. Yet another possibility would be greater accessibility of the c-myc oncoprotein in normal cells and also the possibility that the particular antibody used in the study (MYC l-6ElO) recognizes an epitope of some other nuclear associated protein(s) [5]. Several observations can therefore be drawn from these c-myc studies. A c-myc gene amplification and/or overexpression are significantly associated with most aggressive cervical cancers. It is also possible that the activated cmyc gene could be associated with a cancer cell’s capacity to proliferate. Another interesting observation was that 46% of cervical cancers contained an overexpressed gene with a single gene copy [16], indicating that amplification is not the only mechanism by which the c-myc oncogene is expressed [16]. It was also seen that proportion of tumors with amplified c-myc gene among the tumors with gene overexpression dramatically increases in cases with the worst prognosis (Stages III and IV). This observation indicates that c-myc overexpression is an early event in contrast to gene amplification, which arises in advanced cancers. Such results also strongly suggest that the analysis of c-myc gene amplification as a routine clinical assay will not provide any new information on tumor progression. However c-myc gene expression may be of value as a prognostic indicator in early cervical cancers. In a study of 72 patients with Stage I or II tumors, Riou et al [16] found high levels of c-myc RNA in 25 samples. On estimating an univariate risk of relapse, it was found to be associated with c-myc overexpression, nodal status and geographical origin (40 patients
174
were from France and 32 from Africa). A subsequent multivariate analysis showed only c-myc overexpression and nodal status to be associated with risk of relapse. Here, the major prognostic factor was c-myc overexpression. The 18 month relapse free survival was 49% (95% confidence interval 25- 72%) for patients with c-myc overexpression and 90% (95% confidence interval SO-99%) for patients without c-myc overexpression. Of the 25 patients with high levels of c-myc RNA, 13 relapsed. By contrast of the 47 patients without c-myc overexpression, only 4 relapsed. Thus a significant association between raised levels of c-myc transcripts and risk of relapse was found irrespective of other prognostic factors [16]. Further the overexpression of the c-myc gene outweighed nodal status as a prognostic factor. These results have to be confirmed with larger studies involving more patients. If it is confirmed, then analysis of c-myc RNA in cervical cancers will provide a means of identifying patients at high allowing more risk of early recurrence, individualized approach to management. Conclusions There is no doubt that molecular biology has greatly influenced all aspects of clinical medicine and basic research. The discovery of oncogenes and anti oncogenes has clarified the process of carcinogenesis, opening up a whole new front in tumor markers. New genetic markers can provide biologically more relevant and accurate methods of tumor classification. Such markers can also be tested for their ability to function as tumor markers on the basis of a priori knowledge of their functions. At present, a wide range of discrepancies in oncogene expression and disease course still exist for cervical cancer. It is likely that other factors are involved in the development and progression of cervical cancer including human papilloma viruses [HPV] [8]. Viral DNA sequences have been found to be integrated in the cellular genome of cervical cancer cells and transcribed in the majority of
them [8]. Moreover, when specific types of HPV are detected in cervical preneoplastic lesions, the risk of malignant transformation is increased [8]. Therefore detection of viral sequences along with alterations in ras and cmyc genes could provide a potent test for identifying high risk patients. Though the precise role of viral involvement in cervical carcinogenesis is yet to be fully characterized, experimental models have provided tempting concepts. HPV type 16 DNA has been shown to cooperate with mutated c-Ha-ras gene in transforming primary cells and HPV sequences integrated near the c-myc gene may be involved in the activation of the protooncogene in cervical cancer cell lines [8]. The availability of such new biological markers, can make it possible to detect premalignant lesions capable of transformation or early invasive cancers likely to relapse. Better surveillance of such patients will also result in better forms of management. With the progressive refinement of current molecular biology techniques enabling high sensitivity and specificity while at the same time being simple enough for routine clinical practice, the possibility of oncogenes and their products functioning as true tumor markers is almost fully realized. References Balrnain, multistage
A. (1985) Transforming ras carcinogenesis. Br. J. Cancer,
oncogenes 51. 1- 7.
and
Bishop, J.M. (1987) The molecular genetics of cancer. Science, 235, 305- 311. Hann, S.R., Thompson, C.B. and Eisenman, R.N., (1985) c-myc oncogene protein is independent of the cell cycle in human and avian cells. Nature, 314, 366- 367. Hayshi. Y., Hachisuga, T., Iwasaka, T., Kukuda, K., Okuma, Y., Yokoyama, M. and Sugimori, H. (1991) Expression of ras oncogene product and EGF receptor in cervical squamous cell carcinomas and its relationship to lymph node involvement. Gynecol. Oncol., 40, 147- 151. Hendy-Ibbs, P.. Cox, H., Evan, G.I. and Watson, J.V. (1987) Flow cytometric quantitation of DNA and c-myc oncoprotein in archival biopsies of uterine cervix neoplasia. Br. J. Cancer, 55, 275-282. Jussawaia, D.J. and Gangadharan, P. (1973) Epidemiology of cancer in the Indian sub-continent. Ind. J. Cancer, 10, B- 10.
175 7
8
9
10
11
12
13 14
Michelassi, F., Handcock, M. and Lock, G.E. (1987) Ras oncogene ~21 levels parallel malignant potential of different colonic benign conditions. Arch. Surg., 122, 1414- 1416. Nair, B.S. and Pillai, R. (1991) Oncogenesis of squamous carcinoma of the uterine cervix. Int. J. Gynecol. Pathol., (in press). Ocadiz. R.. Sauceda. R., Cruz. M., Graef. A.M. and Gariglio, P. (1987) High correlations between molecular alterations of the c-myc oncogene and carcinoma of the uterine cervix. Cancer Res., 47, 4173-4177. Ohuchi, N., Thor, A., Page, D.L., Horan Hand, P., Halter, S.A. and Schlom, J. (1986) Expression of the 21000 molecular weight ras protein in a spectrum of benign and malignant human mammary tissues. Cancer Res., 46, 2511- 2519. Pejovic, M.H.. Wolff, J.P., Kramer, A. and Golodfarb, E. (1981) Cure rate estimation and long term prognosis of uterine cervix carcinoma. Cancer, 47, 203 - 206. Pillai. R., Balaram. P., Chidambaram, S., Padmanabha, T.K. and Nair, M.K. (1990) Development of an immunological staging system to prognosticate disease course in malignant cervical neoplasia. Gynecol. 200 - 205. Piver, M.S. (1990) Invasive cervical cancer Semin. Surg. Oncol.. 6, 359-363.
Oncol.,
15
16
Riou, G., Barrois. M., Dutronquay. V. and Orth, G. (1985) Presence of papillomavirus DNA sequences and amplification of c-myc and H-ras oncogenes and enhanced expression of c-myc in carcinomas of the uterine cervix. in: Papillomaviruses: Molecular and Clinical Aspects. Vol. 32, pp. 47-56. Editors: P.M. Howley and T.R. Broker. Alan R. Liss. New York.
M., Tordjman,
I.. Dutronquay,
V. and
17
Riou, G. (1988) Protooncogenes and prognosis in early carcinoma of the uterine cervix. Cancer Surv.. 7. 441- 456. Riou, G.. Le. M.G., Doussal, V.L.. Barrois. M., George
18
and Haie, C. (1987) c-myc protooncogene expression and prognosis in early carcinoma of the uterine cervix. Lancet. ii, 761- 763. Sagas, S., Kudo. R.. Kuzumaki, N., Hisada, T., Mugikura,
19
Y., Nihei. T.. Takeda, T. and Hashimoto, M. (1990) Ras oncogene expression and progression in intraepithelial neoplasia of the uterine cervix. Cancer. 66, 295-301. Sagae, S., Kuzumaki, N., Hisada. T.. Mugikura, Y..
20
Kudo. R. and Hashimoto. M. (1989) Ras oncogene expression and prognosis of invasive squamous cell carcinomas of the uterine cervix. Cancer, 63, 1577 - 1582. Spandidos, D.A. and Ker, I.B. (1984) Elevated expression
21
of the human ras oncogene family in premalignant malignant tumors of the colorectum. Br. J. Cancer, 681- 688. Spandidos, D.A. and Anderson, M.L.M. (1989)
37,
in the 1990s.
Riou, G.. Barrois,
Orth, G. (1984) Presence de genomas de papillomavirus et amplification des oncogenes c-myc et c-Ha ras dans cancers envahissants du col de l’uterus. C. R. Acad. Sci.. 299, 575 - 580.
22
and 49, On-
cogenes and oncosuppressor genes: their involvement in cancer. J. Pathol., 157, l- 10. Van Bommel, P.F.J., Van Lindert, A.C.M., Kock, H.C.L.V., Leers, W.H. and Neijt, J.P. (1987) A review of prognostic factors in early stage carcinoma of the cervix (FIG0 lb and IIA) and implications for treatment strategy. Eur. J. Obstet. Gynecol. Rep. Biol., 26, 69- 84.
Oncogene
171 Ltd
expression
and prognosis
in cervical cancer
R. Pillai Department
Centre,
of Family and Community
Trioandrum
Medicine.
Unioersity
of Arizona,
Tucson. Arizona (U.S.A.j
and Regional
Cancer
(India)
(Received 17 April 1991) (Revision received 20 June (Accepted 21 June 1991)
1991)
Summary
Introduction
Cancer of the uterine cervix accounts for 80 - 85 93 of all female genital tract malignancies in India and also remains a major problem for oncologists in other parts of the world. A major concern regarding the disease is the lack of specific tumor markers for early detection,
Cancer of the uterine cervix remains a major problem for oncologists both in developed and developing countries. It is estimated that thirteen thousand women in the United States develop invasive cervical cancer annually [ 131. However the figure is much higher in other countries. In India for example, cervical cancer accounts for approximately 20 -50% of all cancers and 80-85% of all gynecologic cancers [6]. Race, socioeconomic status and numerous other risk factors have been recognized, many of them relating to sexual behavior. It is also believed that human papilloma viruses are important etiological factors and their detection in precancerous lesions is a risk factor for malignant transformation [8]. If these viruses do have a role in initiation of cervical cancer, then other co-factors may be needed for malignant transformation. One such factor of concern is the role of oncogenes. Cellular oncogenes are DNA sequences found in the genome of normal and neoplastic tissue [2]. Even though their exact functions are just beginning to be understood, neoplastic transformation may be related to oncogene activation caused by point mutations, chromosomal translocations, deletions and gene amplification [21]. Such abnormalities in oncogene expression have been reported in a high number of cancer cell lines and in many fresh tumor samples [2,21].
for accurate prediction of biological behaviour and for accurate assessment of prognosis. A new and exciting answer to this issue may now be available with the description of specific oncogenes and oncoproteins associated with this malignancy. On a clinical level these genes and their products may allow us to improve our
understanding of disease etiology, and prouide diagnostic, prognostic and more precise individual characterization of therapeutic tumors. This paper discusses the possibilities of using altered expression of oncogenes and their products in neoplastic tissue as markers for the diagnosis and prognosis of cervical cancer. These data support the view that detailed analysis
of
potential
to predict
such
gene
expression
cervical cancer; Keywords: prognosis; tumor marker Correspondence to: 21/842, Nedumcaud,
has
the
tumor behauior.
oncogenes;
Dr. Radhakrishna Pillai, Karthika, TC Karamana. Trivandrum 695 002, Kerala
state, India.
0304-3835/91/$03.50 Published and Printed
0 1991 Elsevier Scientific in Ireland
Publishers
Ireland Ltd
172
Prognostic
features
in
cervical
cancers The biological behavior of invasive cervical cancer is not always predictable. Even when the lesion is localized to the cervix, 15 - 20% of patients have recurrences [11,22]. Many clinical prognostic factors have been associated with high risk of relapse including lymph node involvement, tumor size, histological type and grade of differentiation [21]. Lymph node involvement appears to be a most important factor, since patients with early stages of cervical cancer but who also have para aortic and multiple lymph node metastases are likely to rapidly develop distant metastases [ 11,221. However, in many cases with involved lymph nodes there are no recurrences and conversely the absence of positive nodes does not ensure disease free status. The problem therefore for oncologists is to try and accurately predict those tumors which will relapse in order to modify or modulate treatment protocol. It is thus important to establish biological markers which can be associated with disease course. One such recent advance was the description of the immunological staging system where immunological functions could be correlated well with disease course [ 121. Since oncogenes are associated with proliferation control and cancer [2,21], they offer a new parameter for estimation of prognosis and disease course besides providing insights into carcinogenic mechanisms. Alterations in the c-iia-ras cervical cancer
gene
in
One of the oncogene families most commonly implicated in a variety of human cancers is the ras gene family [ 11. These genes are related structurally and encode a group of closely related 21-kDa proteins termed p21 [l]. There are several reports describing the relationship between the t-as oncogene and its product ~21 and the process of early carcinogenesis in different cancers such as colon, bladder and breast [1,7,20]. Ohuchi et al. [lo]
described the higher grade of patho-histology in breast cancer to correlate with more freexpression of ~21 during carquent cinogenesis. Michelassi et al. [7] have suggested that the levels of the ras oncogene protein increases with the malignant potential of benign human colonic conditions and thus, high risk patients should be monitored and treated before actual malignant transformation occurs. These results suggest that the ras oncogene and its ~21 protein may play an important role in malignant transformation. Current data also support a role for the ras oncogene in cervical cancers. To examine the correlations between ras gene expression and the development of cervical cancer, Sagae et al. [18] studied the reactivity of cervical intraepithelial neoplasia (CIN) and microinvasive lesions of the cervix by using an anti ras p21 monoclonal antibody, rp35. The frequency of positive ~21 staining increased with higher grades of malignancy from 17.9% in CIN-1 to 28.9% in CIN-2 and 53.9% in CIN-3 while it was 50% in microinvasive carcinomas. Furthermore ten cases of lesions that regressed during a l-year follow-up period, 20% were positive for ras ~21 while 50% of the 14 lesions that progressed to invasive and higher grade lesions were positive for ~21. The data thus showed that the frequency of p2 1 increased with higher grades of dysplasia suggesting that ~21 positivity increases during early carcinogenesis from CIN to invasive carcinoma. The same authors in another study [19] using the same anti ras ~21 antibody studied reactivity of 170 squamous carcinomas of the uterine cervix belonging to different histological grades. Overexpression of ~21 was noted in 57.1% of keratinising tumors (K type) and 54.2% of non keratinizing type (LNK type) but only 38.7% of the small cell type (S type). Statistical analysis of the p21 expression and patient prognosis showed the p21 positive K and LNK types to have poorer prognosis whereas the positive S type showed better prognosis in comparison with the negative cases. The authors concluded that the expression of ras ~21 is a prognostic indicator for cervical
173
cancers but the mode of its prognostic correlation was dependent on tumor histology. Another finding of clinical significance was the report showing a correlation between elevation of ~21 expression in cervical carcinomas and incidence of lymph node metastases [4]. This concordant relationship in cervical carcinomas suggests that enhanced production of ~21 occurs late in the oncogenic process and may relate more to the malignant phenotype than to the early process of malignant transformation. However the number of patients in the study was too small to allow any firm conclusion. Alterations in the c-myc oncogene cervical carcinomas
in
Previous studies [14,15,17] of the c-myc oncogene in a limited number of cervical tumors had shown that this oncogene was activated in most evolved cancers, suggesting that the c-myc activation was associated with tumor progression. Further studies on larger number of patients showed the gene to be amplified from 3 to 30 times in 21% of the tumors [16]. Similar results were seen from other studies [9] and no amplification of c-myc was noticed in DNA preparations from normal cervical epithelium or from lymphocytes. Amplification of c-myc was found to be much higher in Stage III and IV tumors indicating that the gene amplification is significantly associated with advanced cervical cancers [16]. A similar finding [9] also suggested that patients whose tumors showed c-myc overexpression at levels 4 - 20 times that of normal tissues had an g-fold greater incidence of early relapse than other patients. The c-myc expression even outweighed nodal status in this study as a prognostic factor. However different findings have also been reported. Quantitation by flow cytometry of the c-myc oncoprotein content in archival biopsies of uterine cervix neoplasia showed normal biopsies to have higher levels than carcinomas [5], a finding difficult to explain in light of most studies showing increased expression of the c-myc gene. Since
the function of the c-myc oncoprotein is not yet understood, the functional significance of this observation is not known. Nevertheless, a number of possibilities exist for such a finding. These include a post-translational protein modification or c-myc gene mutation in malignant cells, increased turn over of the protein which is known to have a short half-life in stimulated cells [3] and a possible increase in the susceptibility of the protein to proteolysis in neoplastic cells during preparation of tissue for the assay. Yet another possibility would be greater accessibility of the c-myc oncoprotein in normal cells and also the possibility that the particular antibody used in the study (MYC l-6ElO) recognizes an epitope of some other nuclear associated protein(s) [5]. Several observations can therefore be drawn from these c-myc studies. A c-myc gene amplification and/or overexpression are significantly associated with most aggressive cervical cancers. It is also possible that the activated cmyc gene could be associated with a cancer cell’s capacity to proliferate. Another interesting observation was that 46% of cervical cancers contained an overexpressed gene with a single gene copy [16], indicating that amplification is not the only mechanism by which the c-myc oncogene is expressed [16]. It was also seen that proportion of tumors with amplified c-myc gene among the tumors with gene overexpression dramatically increases in cases with the worst prognosis (Stages III and IV). This observation indicates that c-myc overexpression is an early event in contrast to gene amplification, which arises in advanced cancers. Such results also strongly suggest that the analysis of c-myc gene amplification as a routine clinical assay will not provide any new information on tumor progression. However c-myc gene expression may be of value as a prognostic indicator in early cervical cancers. In a study of 72 patients with Stage I or II tumors, Riou et al [16] found high levels of c-myc RNA in 25 samples. On estimating an univariate risk of relapse, it was found to be associated with c-myc overexpression, nodal status and geographical origin (40 patients
174
were from France and 32 from Africa). A subsequent multivariate analysis showed only c-myc overexpression and nodal status to be associated with risk of relapse. Here, the major prognostic factor was c-myc overexpression. The 18 month relapse free survival was 49% (95% confidence interval 25- 72%) for patients with c-myc overexpression and 90% (95% confidence interval SO-99%) for patients without c-myc overexpression. Of the 25 patients with high levels of c-myc RNA, 13 relapsed. By contrast of the 47 patients without c-myc overexpression, only 4 relapsed. Thus a significant association between raised levels of c-myc transcripts and risk of relapse was found irrespective of other prognostic factors [16]. Further the overexpression of the c-myc gene outweighed nodal status as a prognostic factor. These results have to be confirmed with larger studies involving more patients. If it is confirmed, then analysis of c-myc RNA in cervical cancers will provide a means of identifying patients at high allowing more risk of early recurrence, individualized approach to management. Conclusions There is no doubt that molecular biology has greatly influenced all aspects of clinical medicine and basic research. The discovery of oncogenes and anti oncogenes has clarified the process of carcinogenesis, opening up a whole new front in tumor markers. New genetic markers can provide biologically more relevant and accurate methods of tumor classification. Such markers can also be tested for their ability to function as tumor markers on the basis of a priori knowledge of their functions. At present, a wide range of discrepancies in oncogene expression and disease course still exist for cervical cancer. It is likely that other factors are involved in the development and progression of cervical cancer including human papilloma viruses [HPV] [8]. Viral DNA sequences have been found to be integrated in the cellular genome of cervical cancer cells and transcribed in the majority of
them [8]. Moreover, when specific types of HPV are detected in cervical preneoplastic lesions, the risk of malignant transformation is increased [8]. Therefore detection of viral sequences along with alterations in ras and cmyc genes could provide a potent test for identifying high risk patients. Though the precise role of viral involvement in cervical carcinogenesis is yet to be fully characterized, experimental models have provided tempting concepts. HPV type 16 DNA has been shown to cooperate with mutated c-Ha-ras gene in transforming primary cells and HPV sequences integrated near the c-myc gene may be involved in the activation of the protooncogene in cervical cancer cell lines [8]. The availability of such new biological markers, can make it possible to detect premalignant lesions capable of transformation or early invasive cancers likely to relapse. Better surveillance of such patients will also result in better forms of management. With the progressive refinement of current molecular biology techniques enabling high sensitivity and specificity while at the same time being simple enough for routine clinical practice, the possibility of oncogenes and their products functioning as true tumor markers is almost fully realized. References Balrnain, multistage
A. (1985) Transforming ras carcinogenesis. Br. J. Cancer,
oncogenes 51. 1- 7.
and
Bishop, J.M. (1987) The molecular genetics of cancer. Science, 235, 305- 311. Hann, S.R., Thompson, C.B. and Eisenman, R.N., (1985) c-myc oncogene protein is independent of the cell cycle in human and avian cells. Nature, 314, 366- 367. Hayshi. Y., Hachisuga, T., Iwasaka, T., Kukuda, K., Okuma, Y., Yokoyama, M. and Sugimori, H. (1991) Expression of ras oncogene product and EGF receptor in cervical squamous cell carcinomas and its relationship to lymph node involvement. Gynecol. Oncol., 40, 147- 151. Hendy-Ibbs, P.. Cox, H., Evan, G.I. and Watson, J.V. (1987) Flow cytometric quantitation of DNA and c-myc oncoprotein in archival biopsies of uterine cervix neoplasia. Br. J. Cancer, 55, 275-282. Jussawaia, D.J. and Gangadharan, P. (1973) Epidemiology of cancer in the Indian sub-continent. Ind. J. Cancer, 10, B- 10.
175 7
8
9
10
11
12
13 14
Michelassi, F., Handcock, M. and Lock, G.E. (1987) Ras oncogene ~21 levels parallel malignant potential of different colonic benign conditions. Arch. Surg., 122, 1414- 1416. Nair, B.S. and Pillai, R. (1991) Oncogenesis of squamous carcinoma of the uterine cervix. Int. J. Gynecol. Pathol., (in press). Ocadiz. R.. Sauceda. R., Cruz. M., Graef. A.M. and Gariglio, P. (1987) High correlations between molecular alterations of the c-myc oncogene and carcinoma of the uterine cervix. Cancer Res., 47, 4173-4177. Ohuchi, N., Thor, A., Page, D.L., Horan Hand, P., Halter, S.A. and Schlom, J. (1986) Expression of the 21000 molecular weight ras protein in a spectrum of benign and malignant human mammary tissues. Cancer Res., 46, 2511- 2519. Pejovic, M.H.. Wolff, J.P., Kramer, A. and Golodfarb, E. (1981) Cure rate estimation and long term prognosis of uterine cervix carcinoma. Cancer, 47, 203 - 206. Pillai. R., Balaram. P., Chidambaram, S., Padmanabha, T.K. and Nair, M.K. (1990) Development of an immunological staging system to prognosticate disease course in malignant cervical neoplasia. Gynecol. 200 - 205. Piver, M.S. (1990) Invasive cervical cancer Semin. Surg. Oncol.. 6, 359-363.
Oncol.,
15
16
Riou, G., Barrois. M., Dutronquay. V. and Orth, G. (1985) Presence of papillomavirus DNA sequences and amplification of c-myc and H-ras oncogenes and enhanced expression of c-myc in carcinomas of the uterine cervix. in: Papillomaviruses: Molecular and Clinical Aspects. Vol. 32, pp. 47-56. Editors: P.M. Howley and T.R. Broker. Alan R. Liss. New York.
M., Tordjman,
I.. Dutronquay,
V. and
17
Riou, G. (1988) Protooncogenes and prognosis in early carcinoma of the uterine cervix. Cancer Surv.. 7. 441- 456. Riou, G.. Le. M.G., Doussal, V.L.. Barrois. M., George
18
and Haie, C. (1987) c-myc protooncogene expression and prognosis in early carcinoma of the uterine cervix. Lancet. ii, 761- 763. Sagas, S., Kudo. R.. Kuzumaki, N., Hisada, T., Mugikura,
19
Y., Nihei. T.. Takeda, T. and Hashimoto, M. (1990) Ras oncogene expression and progression in intraepithelial neoplasia of the uterine cervix. Cancer. 66, 295-301. Sagae, S., Kuzumaki, N., Hisada. T.. Mugikura, Y..
20
Kudo. R. and Hashimoto. M. (1989) Ras oncogene expression and prognosis of invasive squamous cell carcinomas of the uterine cervix. Cancer, 63, 1577 - 1582. Spandidos, D.A. and Ker, I.B. (1984) Elevated expression
21
of the human ras oncogene family in premalignant malignant tumors of the colorectum. Br. J. Cancer, 681- 688. Spandidos, D.A. and Anderson, M.L.M. (1989)
37,
in the 1990s.
Riou, G.. Barrois,
Orth, G. (1984) Presence de genomas de papillomavirus et amplification des oncogenes c-myc et c-Ha ras dans cancers envahissants du col de l’uterus. C. R. Acad. Sci.. 299, 575 - 580.
22
and 49, On-
cogenes and oncosuppressor genes: their involvement in cancer. J. Pathol., 157, l- 10. Van Bommel, P.F.J., Van Lindert, A.C.M., Kock, H.C.L.V., Leers, W.H. and Neijt, J.P. (1987) A review of prognostic factors in early stage carcinoma of the cervix (FIG0 lb and IIA) and implications for treatment strategy. Eur. J. Obstet. Gynecol. Rep. Biol., 26, 69- 84.
