MR脂肪抑制T2WI联合弥散加权成像与增强MRI鉴别乳腺黏液癌与黏液样纤维腺瘤

MR fat suppression-T2WI combined with diffusion weighted imaging and enhanced MRI for differentiating mucinous breast carcinoma and breast myxoid fibroadenoma

目的 对比观察MR脂肪抑制(FS)-T2WI(FS-T2WI)联合弥散加权成像(DWI)与增强MRI鉴别乳腺黏液癌(MBC)与乳腺黏液样纤维腺瘤(MFA)的价值。方法 回顾性分析经术后病理证实的14例MBC(MBC组)与22例乳腺MFA(MFA组),比较组间FS-T2WI信号强度(SI)指数、DWI-表观弥散系数(ADC)、强化特点及时间信号强度曲线(TIC)等MRI表现差异。针对FS-T2WI联合DWI及增强MRI参数以二元logistic回归分析拟合回归方程,绘制受试者工作特征(ROC)曲线,计算曲线下面积(AUC),评估2种回归方程鉴别MBC与乳腺MFA的效能。结果 组间FS-T2WI信号SI指数(Z=3.780,P<0.001)、DWI-ADC(t=4.230,P<0.001)、强化均匀与否(P=0.006)、早期强化方式(P<0.001)、强化填充方式(P<0.001)及TIC类型(P=0.001)差异均有统计学意义,延迟期强化方式差异无统计学意义(P=0.062)。基于FS-T2WI联合DWI参数建立回归方程如下:Logit(P)=-10.434+0.003×ADC+0.748×FS-T2WI SI指数;基于增强MRI参数建立回归方程Logit(P)=31.666+0.287×强化均匀与否-18.319×早期强化方式+19.945×强化填充方式-36.591×延迟期强化方式+20.225×TIC类型。上述2个回归方程鉴别MBC与乳腺MFA的AUC(Z=1.890,P=0.059)、敏感度(χ^(2)=1.050,P=0.305)、特异度(χ^(2)=1.100,P=0.294)和准确率(χ^(2)=0.660,P=0.416)差异均无统计学意义。结论 FS-T2WI联合DWI可鉴别诊断MBC与乳腺MFA,其诊断效能与增强MRI相当。

Objective To compare the value of MR fat suppression-T2 WI(FS-T2 WI) combined with diffusion weighted imaging(DWI) for differentiating mucinous breast carcinoma(MBC) and breast myxoid fibroadenoma(MFA). Methods MRI data of 14 patients with MBC(MBC group) and 22 patients with breast MFA(MFA group) confirmed by postoperative pathology were retrospectively analyzed. MRI manifestations, including the signal intensity(SI) index on FS-T2 WI, DWI-apparent dispersion coefficient(ADC), enhancement characteristics and the time-intensity curve(TIC) pattern were compared between groups. Using parameters of FS-T2 WI combined with DWI and enhanced MRI, regression equations were fitted with binary logistic regression analysis, then receiver operating characteristic(ROC) curves were drawn and the area under the curve(AUC) was calculated to assess the efficacy of the regression equations for identifying MBC and breast MFA. Results SI index on FS-T2 WI(Z=3.780, P<0.001), DWI-ADC(t=4.230, P<0.001), uniform enhancement or not(P=0.006), early enhancement feature(P<0.001), enhancement filling mode(P<0.001) and TIC type(P=0.001) were significantly different, while there was no significant difference of delayed enhancement style(P=0.062) between groups. Based on the parameters of FS-T2 WI combined with DWI, the logistic regression model was established, i.e. Logit(P)=-10.434+0.003×ADC+0.748×SI index on FS-T2 WI;while based on the parameters of enhanced MRI, the established logistic regression model was Logit(P)=31.666+0.287×uniform enhancement or not-18.319×early enhancement feature+19.945×enhancement filling mode-36.591×delayed enhancement style+20.225×TIC type. No significant difference of AUC(Z=1.890, P=0.059), sensitivity(χ^(2)=1.050, P=0.305), specificity(χ^(2)=1.100, P=0.294) nor accuracy(χ^(2)=0.660, P=0.416) was found between the 2 regression equations for identifying MBC and breast MFA. Conclusion FS-T2 WI combined with DWI could be used for differentiating MBC and breast MFA, with diagnostic efficacy comparable to enhanced MRI.