应用COX模型分析影响乳腺浸润性癌预后的生物学因素

An analysis of biological factors influencing prognosisof invasive breast cancer by COX model

目的：探讨增殖细胞核抗原（ＰＣＮＡ）、原癌基因（ＣｅｒｂＢ２）、组织蛋白酶Ｄ（ＣａｔｈＤ）、转移抑制基因（ｎｍ２３Ｈ１）、微血管数（ＭＶＣ）、肥大细胞数（ＭＣ）、雌激素调节基因（ＰＳ２）、前列腺特异抗原（ＰＳＡ）９个生物学因素与乳腺浸润性癌（ｉｎｖａｓｉｖｅｂｒｅａｓｔｃａｎｃｅｒ，ＩＢＣ）预后的关系。方法：单因素分析：ＫａｐｌａｎＭｅｉｒ生存曲线法。列联检验法。多因素分析：ＣＯＸ比例风险模型。结果：经ＣＯＸ模型ＭＰＬＲ方法检验显示出明显影响乳腺浸润性癌预后的４个因素，ＰＳ２、ＭＶＣ、ｎｍ２３Ｈ１、ＣｅｒｂＢ２．用比例风险模型计算出每个患者的预后指数ＰＩ（Ｐｒｏｇｎｏｓｔｉｃｉｎｄｅｘ），根据预后指数大小将８６例乳腺癌术后患者分为２组，分别建立其术后生存率预测模型。结果提示ＰＩ值愈大，预后愈差，反之预后则好。结论：ＰＳ２、ＭＶＣ、ｎｍ２３Ｈ１、ＣｅｒｂＢ２是乳腺浸润性癌术后独立的预后指标。ＰＩ可能是临床评价病人预后、识别ＩＢＣ术后复发的高危险性病人很有实用价值的指标。

Purpose:To detect the relationship between prognosis of invasive breast cancer(IBC)and nine biological factors:proliferating cell nuclear antigen(PCNA),C erbB 2,CathepsinD,nm23 H1,microvessel count(MVC),mast cell(MC),PS2 and prostate specific antigen(PSA).Methods:Univariate analysis:Kaplan Meier survival curve,crosstabs statistics,multaivariate analysis:COX proportional hazard modle.Results:MPLR test whowed the factors wiich influencing prognosisof invasive breast cancer included PS2,MVC,nm23 H1 and C erbB 2.Using proportional hazard model,the PI(Prognostic index)of each cases were calculated.according to their PI,86 cases of breast cancer accepted surgical treatment were divided into 2 groups and the predicting model of survival rates in each group were determined.Result the larger PI,the poorer the prognosis,the smaller PI,the better the prognosis.Conclusions:PS2,MVC,nm23 H1 and C erbB 2 coud be considered as the independent prognostic predictors in invasive breast cancer after resection.The PI may be a useful clinical tool for evaluating the prognosis of patients and identifying the population with“high risk”of recurrence and death in IBC.