宫颈癌患者预后因素分析及预测模型的初步建立

Analysis of prognosis factor and establishment of prediction model to patients with cervical cancer

目的探讨影响宫颈癌预后的影响因素，在此基础上建立宫颈癌患者生存率的预测模型以指导临床个体化治疗。方法对广东省阳江市妇幼保健院2004年1月-2008年3月宫颈癌患者的临床病理及随访资料进行回顾性分析，利用Kaplan—Meier和Cox—Regression方法进行单因素及多因素的预后分析，并基于此建立预后的数学模型。结果FIGO分期、病理类型、分化程度、肿瘤直径、淋巴结阳性及CA125水平等经单因素分析与宫颈癌患者的预后有关，FIGO分期、病理类型、分化程度经多因素分析是患者独立的预后因素；成功建立了宫颈癌患者3年生存率的预测模型：S（36）=O．984exp（0．653xFIGO分期-0．712x分化程度＋0．484x病理类型代码）。结论FIGO分期、病理类型及分化程度影响宫颈癌患者预后的主要因素，针对上述因素制定个体化的治疗方案结合生存率预测模型可为改善患者预后提供参考。

Objective To assess prognostic factors in patients with cervical carcinoma and establish a prediction model to help the decision of adjuvant therapy. Methods The chnical, pathological and following-up data of patients with cervieal cancer who underwent operation between January 2004 and March 2008 in Guangdong Yangjiang Maternity and Child Health Hos- pital were retrospected and selected. Based on univariate analysis and multivariate analysis of cox-regression model which were taken to identify the prognostic factors for the patients with cervical cancer, a prediction model was established. Results After univariate analysis, FIGO stage, tumor histological grade, cell differentiation, tumor size, lymph node metastasis and serum CA125 values were progonostic factors and were associated with the survival rate of the cervical cancer patients; FIGO stage, tumor histological grade and cell differentiation, according to the multivariate analysis of cox-regression model, were associated with the the survival rate of patients with cervical cancer; the prediction model of S （36） = 0.984exp （0.653× FIGO stage -0.712× cell differentiation ＋0.484× tumor histological grade） was established. Conclusion FIGO stage, tumor histo- logical grade and cell differentiation are independent survival progonostic factors of in patients with cervical cancer. The pre- diction model based on the progonostic factors is useful for clinical therapy.