磁共振增强特征和ADC值与乳腺癌分子分型相关性的logistic分析

Logistic analysis of correlaiton of enhanced MRI features and ADC value with molecular subtyping of breast cancer

目的:探讨乳腺癌MRI增强特征和ADC值与分子分型之间的相关性。方法:90例共95个乳腺癌病灶纳入本研究,并根据免疫组化结果分为3组:HR组、TN组和Herb-2组。比较各组间病灶的MRI增强特征及ADC值的差异。将有统计学意义的变量引入多分类logistic回归分析,寻找预测肿瘤分子特征的危险因素。结果:95个病灶中HR组61个,TN组13个,Herb-2组24个。单因素分析结果显示,病灶的强化方式及乳腺背景实质强化(BPE)在各组间的差异有统计学意义(P<0.05),ADC值在不同组间差异无统计学意义。多分类logistic回归分析显示,与Herb-2组比较,HR组倾向于肿块样强化(OR=4.594,95%CI:1.368~15.427;P=0.014),BPE为2级时,乳腺癌分子分型为HR的概率增加(OR=5.581,95%CI:1.120~27.814;P=0.036),TN组倾向于肿块样强化(OR=6.984,95%CI:1.208~40.37;P=0.030)。TN组ADC值在3组中最低,但是3组间差异无统计学意义(P>0.05)。结论:依据MRI增强特征鉴别乳腺癌分子亚型是可行的。

Objective:The aim of this study was to investigate the correlation between enhanced MRI features and ADC value with molecular subtyping of breast cancer.Methods:95 lesions of 90 cases were included and divided into 3 groups based on immunohistochemical:HR group,TN group and Herb-2 group.Enhanced MRI features and ADC value were compared between any two groups.Variables with statistical significance were introduced into the multinomial logistic regression equation in search for the risk factors of predicting tumor molecular characteristics.Results:There were 61 cases in HR group,13 in TN group and 24 in Herb-2 group out of 95 lesions.Univariate analysis showed a significant difference(P<0.05)between enhancement pattern and background parenchymal enhancement(BPE),and that there was no significant difference in ADC values between different groups.Multinomial logistic regression analysis indicated that HR group tended to appear as mass-like enhancement[OR=4.594(95%CI:1.368~15.427),P=0.014]compared with Herb-2 group.The probability of molecular subtyping to HR increased with BPE of grading 2(OR=5.581,95%CI:1.120~27.814;P=0.036).TN group tended to appear as mass-like enhancement(OR=6.984,95%CI:1.208~40.37;P=0.030).TN group had the lowest ADC value in three groups but without significant difference(P>0.05).Conclusion:It is feasible to identify the molecular subtypes of breast cancer according to enhanced MRI characteristics.