应用“数学模型”预测临床T_(1～2)N_0M_0乳腺癌患者腋窝淋巴结转移情况的研究

Predicting the status of axillary lymph nodes metastases in patients with T_(1～2)N_0M_0 breast cancer using a mathematical model

目的：建立一个数学模型，预测临床T1～2N0M0乳腺癌患者腋窝淋巴结转移情况。方法：对256例行根治术乳腺癌患者进行回顾性研究，按腋窝淋巴结转移情况分为两组，行多因素Logistic回归分析及判别分析。结果：Logistic多元回归分析提示微淋巴管浸润、肿瘤大小、肿瘤部位、癌周浸润、间质浸润等5个指标为腋窝淋巴结转移的高危因素；据高危因素所建立的判别函数，阴性预测值高达88．9％，阳性预测值71．8％，判别准确率83.2％。结论：本判别函数数学模型简单，在理论上可以较准确地判断腋窝淋巴结的转移情况，可能有一定的临床参考价值。

Objective: To establish a mathematical model to predict the status of axillaly lymph nodes metastases in patients with T1 ~ 2N0M0 breast cancer. Methods: 256 cases of T1~ 2N0M0 breast cancer were divided into two groups according to the status of axillary lymph nodes, and were retrospectively analysed using multifactorial logistic regression and discriminant analysis. Results: Logistic regression analysis showed that the microlymphatic invasion, diameter of primary tumor, location of tumor, cireumferential invasion and stromal invasion were the risk factors of axillary lymph nodes metastases. Fisher's linear discriminant function was set up using these 5 factors by discriminant analysis. According to this mathematical model, the negative predictive value, positive predictive value and predictive accuracy of the statue of axilleq lymph nodes metastases were 88. 9%, 71. 8% and 83. 2, respectively. Conclusions: The mathematical model of discriminant function was simple, could predict the status of axillary lymph nodes metastases accurately. It might have some value in clinical application.