术前超声联合临床多因素回归模型在评估pN1期乳腺癌患者腋窝淋巴结转移中的价值

发布时间：2018-04-08 16:56

本文选题：乳腺癌　切入点：腋窝　出处：《浙江大学》2015年硕士论文

【摘要】：目的：探讨超声声像图特征与pNl期乳腺癌腋淋巴结转移的相关性,并联合临床、病理信息建立多因素回归模型预测腋淋巴结的pNl期转移。方法：采用飞利浦HDII彩色超声仪,5～12MHz高频线阵探头,对乳腺癌患者的乳腺病灶及腋窝淋巴结进行全面、细致的超声检查。以前哨淋巴结活检或腋窝淋巴结清扫的病理结果作为淋巴结转移的金标准。选取经病理证实的80例pN1期和pN0期乳腺癌患者临床、病理和病灶超声特征等22项参数和腋窝淋巴结超声特征等6项参数与腋窝淋巴结状态进行分析。采用SPSS16.0统计软件对以上28个参数进行单因素及多因素分析,以P0.05作为检验水准,并绘制受试者工作曲线检验回归模型的预测效能。结果： 80例乳腺癌患者和92枚淋巴结分析显示：年龄、病灶位置、病灶血流分级、有无毛刺、后方回声、前方脂肪层回声、乳腺后间隙、组织学分级、皮质厚度、皮髓质分界、皮质均匀性、皮门比等以上12项参数为腋窝淋巴结转移相关因素(P0.05)。多因素分析显示：年龄40岁、病灶血流Ⅲ级、前方脂肪层回声增强、乳腺后间隙消失为腋窝淋巴结转移危险因素,其Odd Ratio (OR)值分别为17.78(95%CI2.50～126.54),2.47(95%CI0.16～32.97),4.57(95%CI1.13～18.53),4.4(95%CI0.87-22.18);皮质非均匀性改变、皮门比1.0为腋窝淋巴结转移危险因素,其Odd Ratio (OR)值分别为5.1(95%CI1.68～15.44),3.53(95%CI1.16～10.75)。绘制前方脂肪层回声参数和多参数组合(年龄、病灶血流分级、前方脂肪层回声、乳腺后间隙)的ROC曲线,2条曲线下面积(Area Under the Curve, AUC)分别为0.679(95%CI0.542～0.816),0.824(95%CI0.709～0.940).绘制皮质均匀性、皮门比等参数以及多参数组合(皮质均匀性、皮门比)的ROC曲线,3条曲线的AUC分别为0.696(95%CI0.569～0.823),0.670(95%CI0.541～0.799),0.757(95%CI0.640～0.873).绘制综合所有指标(年龄、病灶血流分级、前方脂肪层回声、乳腺后间隙、皮质均匀性、皮门比)的ROC曲线,曲线的AUC为0.909(95%C10.825～0.994)。综合所有的指标有较实用的术前预测价值,其灵敏度和特异度分别为72.7%和98.2%。本研究超声评估pN1期腋淋巴结转移的敏感性和特异性分别为91.67%和32.14%,阳性预测值和阴性预测值分别为46.48%和85.71%。结论：病灶超声特征联合临床信息对乳腺癌患者pNl期腋淋巴结转移有一定的预测价值(AUC为0.82),为腋窝淋巴结的扫查提供重要信息；腋窝淋巴结超声皮质非均匀性改变是腋淋巴结早期转移的特征,结合淋巴结皮门比大于1预测pNl期腋淋巴结的转移有一定的准确性(AUC为0.76)；将病灶超声特征、腋淋巴结超声特征和患者年龄等信息结合在预测pNl期腋淋巴结的转移上有较好的准确性(AUC为0.909)。
[Abstract]:Objective:To investigate the correlation between ultrasonographic features and axillary lymph node metastasis in pNl stage breast cancer, and to establish a multivariate regression model for predicting axillary lymph node metastasis in pNl stage.Methods:The breast lesions and axillary lymph nodes of breast cancer patients were examined with Philips HDII color ultrasound instrument with a high frequency and 12 MHz linear array probe.Sentinel lymph node biopsy or axillary lymph node dissection was used as the gold standard for lymph node metastasis.Twenty-two parameters of pathology and ultrasonic characteristics of axillary lymph nodes and the status of axillary lymph nodes were analyzed in 80 patients with pN1 and pN0 stage breast cancer confirmed by pathology.The SPSS16.0 software was used to analyze the above 28 parameters by single factor and multivariate analysis. P05 was taken as the test level, and the predictive effectiveness of regression model was tested by drawing the operating curve of the subjects.Results:Analysis of 80 patients with breast cancer and 92 lymph nodes showed: age, location of lesion, blood flow grade of lesion, no burr, posterior echo, echo of anterior fat layer, posterior space of mammary gland, histological grade, thickness of cortex, dividing of dermatomedullary.The cortical homogeneity and the ratio of skin to hilum were all related to axillary lymph node metastasis (P 0.05).The Odd Ratio ORs were 5.1 / 95 CI = 1.68 ~ 15.44 ~ 3.53 ~ 95 CI 1.16 ~ 10.75g / I, respectively.The ROC curves of all indexes (age, blood flow grade, echo of anterior adipose layer, posterior space of mammary gland, cortex homogeneity, ratio of skin to hilum) were plotted. The AUC of the curve was 0.909 (950.825 卤0.994).The sensitivity and specificity were 72.7% and 98.2%, respectively.The sensitivity and specificity of ultrasound in evaluating axillary lymph node metastasis in pN1 stage were 91.67% and 32.14% respectively. The positive predictive value and negative predictive value were 46.48% and 85.71% respectively.Conclusion:Ultrasound features combined with clinical information have certain predictive value for axillary lymph node metastasis in breast cancer patients at pNl stage. AUC is 0.82g, which provides important information for axillary lymph node scanning.The heterogeneity of axillary lymph nodes was the characteristic of early axillary lymph node metastasis. It was accurate to predict axillary lymph node metastasis in pNl with the ratio of lymph node dermatocutaneous to greater than 1. The ultrasonic features of axillary lymph nodes were analyzed.The combination of ultrasonic features of axillary lymph nodes and age of patients had a good accuracy in predicting axillary lymph node metastasis in pNl stage.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：R737.9

【参考文献】