P53 IS A PERSISTENT AND PREDICTIVE MARKER IN ADVANCED OVARIAN CARCINOMAS - MULTIVARIATE-ANALYSIS INCLUDING COMPARISON WITH KI67 IMMUNOREACTIVITY

Citation
P. Rohlke et al., P53 IS A PERSISTENT AND PREDICTIVE MARKER IN ADVANCED OVARIAN CARCINOMAS - MULTIVARIATE-ANALYSIS INCLUDING COMPARISON WITH KI67 IMMUNOREACTIVITY, Journal of cancer research and clinical oncology, 123(9), 1997, pp. 496-501
Citations number
19
Categorie Soggetti
Oncology
ISSN journal
01715216
Volume
123
Issue
9
Year of publication
1997
Pages
496 - 501
Database
ISI
SICI code
0171-5216(1997)123:9<496:PIAPAP>2.0.ZU;2-X
Abstract
p53 mutation and p53 protein overexpression are common findings in ova rian carcinomas. In order to evaluate the prognostic significance of t he p53 status and its role in metastasis, we examined 104 ovarian carc inomas, among them 83 cases with follow-up data, and 40 pairs of prima ry tumors and metastases, by p53 immunohistochemistry and temperature- gradient gel electrophoresis. Comparison of primary tumors and their m etastases revealed identical results in 88%-90% of the cases, indicati ng that, in most cases, mutant p53 occurs prior to metastatic spread a nd remains clonally conserved. With respect to all tumors, moderate/hi gh p53 expression was significantly more prevalent in serous-papillary types, carcinomas with high grade, and high Ki67 scores, but was not associated with age, stage, or hormone receptor status. Kaplan-Meier a nalysis of 83 cases, followed-up for 9-96 months, demonstrated that mo derate/high p53 overexpression in the group of 66 stage T3/M1 tumors w as associated significantly (P = 0.0028 and P = 0.0105) with shorter o verall and recurrence-free survival. Multivariate analysis revealed th at advanced clinical stage and p53 positivity were the only independen t predictive variables. No significance was seen in regard to second-l ook results and outcome of 50 patients receiving platinum-based chemot herapy. These observations show that p52 immunohistochemistry is an in dependent prognostic indicator at the given cut-off level, but does no t reliably predict chemotherapy response.