Construction and verification of the prognostic model of ovarian serous adenocarcinoma
10.3781/j.issn.1000-7431.2010.02.008
- VernacularTitle:卵巢浆液性腺癌预后评分模型
- Author:
Jun HU
;
Lirong ZHU
;
Qinping LIAO
;
Yunong GAO
;
Hong ZHENG
- Publication Type:Journal Article
- Keywords:
Ovarian neoplasms;
Adenocarcinoma;
serous;
Prognosis;
Models;
statistical
- From:
Tumor
2010;(2):125-129
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To analyze the prognosis-related factors of ovarian serous adenocarcinoma in order to set up a prognostic model of serous adenocarcinoma and verify the effectiveness of the model as prognostic clinical criteria.Methods:The clinical, patholo-gical and follow-up data from 181 training samples with ovarian serous adenocarcinoma in Peking University First Hospital during January 1995 to December 2003 and another 42 detection samples with ovarian serous adenocarcinoma in Beijing Cancer Hospital during January 1999 to December 2005 were analyzed retrospectively. Kaplan-Meier univariate analysis was used to screen out prognostic factors; COX univariate and multivariate analyses were used to determine the risk coefficient of each factor and different layers in each factor; Pearson rank correlation analysis was used to identify the correlation of each factor. The prognostic model of ovarian serous adenocarcinoma was established on the conversion of risk coefficient to prognostic score and receiver operating characteristic (ROC) analysis was used to determine the cut-off value. The clinical data were collected to verify the sensitivity and specificity of the prognostic model based on the 3-year survival rate and Ki67 value.Results:The survival rate of patients with ovarian serous adenocarcinoma correlated with 6 factors including FIGO stage, histological grade, residual size, metastasis of lymph nodes, general condition after chemotherapy, and serum CA125 levels. The postoperative chemotherapy was an independent factor for prognosis. The prognostic model directly reflected the survival probability of classical COX risk ratio model, with higher score indicating lower survival probability and poorer prognosis. If combining prognostic score with Ki67 the sensitivity and specificity reached 64.7% and 96.0%, respectively.Conclusion:FIGO stage, histological grade, residual size, metastasis of lymph nodes, postoperative chemotherapy and serum CA125 were the prognostic factors of patients with ovarian serous adenocarcinoma. The prognostic model of ovarian serous adenocarcinoma can moderately reflect the actual survival status, and combination with Ki67 will increase the sensitivity and specificity.
