In cell line studies, BCL-2, BAX, as well as novel MEK1 protein levels have strong influence on ovarian cancer response to cisplatinbased chemotherapy. However, such associations have not been demonstrated clinically. We evaluated prognostic/predictive significance of these proteins with regard to TP53 status. Immunohistochemical analysis was performed on 229 ovarian carcinomas FIGO stage IIB -IV treated with platinum-based chemotherapy; the results were analysed by the Cox and logistic regression models. Clinical parameters (residual tumour size, patient age, FIGO stage) were the only indicators of overall survival (OS) and the strongest predictors of complete remission (CR). On the other hand, BAX expression was the strongest (P ¼ 0.005) or the only (in FIGO IIIC, P ¼ 0.02) prognostic indicator of disease-free survival (DFS) in the TP53(+) group. TP53(+) and TP53(À) ovarian carcinomas differed in clinical and molecular prognostic and predictive factors. Another novel finding is that CR was negatively influenced by high BAX expression in all patients group (P ¼ 0.047) and by BCL2 expression in the TP53(À) group (P ¼ 0.05). High MEK1 expression was associated with endometrioid and clear cell carcinomas (P ¼ 0.049); its loss was found with advancing FIGO stage (P ¼ 0.002). Our results suggest that binomial TP53 status divides ovarian carcinomas into two biologically distinct groups. BAX expression is an important factor of DFS in the TP53(+) group. BCL-2 and BAX, but not MEK1 expressions have predictive value in ovarian cancer patients treated with platinum-based chemotherapy.

We have shown an independent predictive value of P21(WAF1) LI in ovarian carcinoma patients. The prognostic value of P21(WAF1) and P21(WAF1) plus P27(KIP1) LI was determined by TP53 status. A high frequency of C-MYC overexpression in endometrioid and clear cell carcinomas may suggest its role in the development of these tumor types.

Changes in cell survival contribute to tumour development, influence tumour biology and its response to chemotherapy. p53 gene alterations should negatively affect apoptosis by impaired p53-dependent apoptotic response. We looked for associations between spontaneous apoptosis, p53 gene mutation, p53 protein accumulation, growth fraction, bcl-2 expression and histological parameters in 64 ovarian, four tubal and three peritoneal carcinomas. Apoptotic cells were detected with the TUNEL method. p53 gene variants were detected by the single-strand conformation polymorphism and were sequenced directly. P53, Ki-67 and bcl-2 protein expressions were detected immunohistochemically. A weighed multiple logistic regression model was applied. Apoptotic index (Al) ranged 0.02–0.18 (mean 0.11); proliferation index (PI) ranged 3–90% (mean 54%). p53 gene mutations were present in 51, p53 protein accumulation in 46, and diffuse bcl-2 expression in 29 of 71 tumours. The AI was positively associated with the presence of p53 gene mutation ( P = 0.011). However, the PI included into the analysis did positively influence the AI ( P = 0.02) and diminished the association with p53 gene mutation ( P = 0.082). The AI was negatively associated with good histological differentiation ( P = 0.0006), the serous tumour type ( P = 0.002), and diffuse bcl-2 expression ( P = 0.025). Strong bcl-2 expression was associated with endometrioid tumour type ( P = 0.002). FIGO stage and p53 protein accumulation were the only parameters that influenced overall survival time. Thus, our results suggest that histological tumour type and grade are major determinants of spontaneous apoptosis in ovarian carcinomas; p53 alterations do not adversely but rather positively affect spontaneous apoptosis by increasing growth fraction. This, in turn, suggests p53-independency of spontaneous apoptosis in ovarian carcinomas. © 2000 Cancer Research Campaign

No abstract

No abstract