MRI快速检查法对致密型乳腺中乳腺癌检出的应用价值

Abbreviated protocol of magnetic resonance imaging for detecting breast cancer in dense breasts

目的探讨MRI快速检查法（AP）在致密型乳腺中对乳腺癌检出的应用价值。方法回顾性收集2014年9月至2015年6月在南京医科大学附属苏州医院就诊的患者资料，对172例x线摄影（MG）检查符合致密型乳腺的患者行MRI动态增强检查[AP和常规检查法（FDP）]及穿刺活组织检查，AP仅仅读取首次增强减影及最大密度投影图像，FDP读取全部MRI动态增强图像，分别计算MG、AP及FDP三种方法在致密型乳腺中对乳腺癌检出的敏感度、特异度、阳性似然比（PLR）和阴性似然比（NLR）。采用受试者工作特征（ROC）曲线比较三者对乳腺癌的诊断能力。结果172例致密型乳腺中共检出56例乳腺癌，病理结果显示，DCIS12例，浸润性导管癌37例，浸润性小叶癌2例，髓样癌3例，黏液癌2例。其中MG检出19例，AP检出53例，FDP检出55例，敏感度分别为33．9％（19／56）、94．6％（53／56）及98．2％（55／56），特异度分别为49．1％（57／116）、79．3％（92／116）及91．4％（106／116），PLR分别为0．67、4．57及11．42，NLR分别为1．35、0．07及0．02。ROC曲线结果显示：MG、AP及FDP曲线下面积分别为0．679±0．053、0．930±0．026及0．948±0．023，95％置信区间分别为0．574～0．784、0．881—0．980及0．900～0．994。Z检验结果显示，AP及FDP曲线下面积较MG差异均有统计学意义（Z=3．031、3．497，P均〈0．050），而AP和FDP之间差异无统计学意义（Z=1．114，P〉0．050）。结论AP及FDP对致密型乳腺中乳腺癌的诊断效能明显高于MG，AP有望成为一种值得推广的乳腺癌筛查模式。

Objective To evaluate the application of an abbreviated protocol （AP） of MRI for detecting breast cancer in patients with dense breast tissue. Methods In this study, we retrospectively analyzed the clinical data of patients in Suzhou Hospital, Nanjing Medical University from September 2014 to June 2015. Totally 172 patients with dense breast tissue screened in mammography （MG） received dynamic contrast-enhanced MRI [AP and full diagnostic protocol （FDP） ] and needle biopsy. AP consisted of the first posteontrast subtraction and the maximum-intensity projection ; FDP consisted of all MRI dynamic enhanced images. The sensitivity, specificity, positive likelihood ratio （PLR）, negative likelihood ratio （NLR） of MG, AP and FDP for breast cancer detection were calculated in dense breasts. The diagnostic capabilities of the three methods were compared using the receiver operating characteristic （ROC） curve. Results Totally 56 cases of breast cancer were detected in 172 patients with dense breasts, including 12 cases of DCIS, 37 invasive ductal carcinoma, 2 invasive lobular carcinoma, 3 medullary carcinoma and 2 colloid carcinoma. Nineteen cases were detected with breast cancer by MG, 53 cases by AP and 55 cases by FDP, with the sensitivity of 33.9% （ 19/56）, 94. 6% （53/56） and 98.2% （55/56）, the specificity of 49. 1% （57/116）, 79. 3% （92/116） and 91.4 % （106/116）, PLR of 0.67,4.57 and 11.42, and NLR of 1.35, 0.07 and O. 02, respectively. ROC curves showed that area under the curve was 0. 679±0. 053 in MG, 0. 930±0. 026 in AP and 0. 948±0. 023 in FDP, with 95% confidence interval of 0. 574-0. 784, 0. 881-0. 980 and 0. 900- 0. 994 respectively. The area under the curve of AP or FDP was significantly by bigger than that of MG （ Z = 3.031, 3.497 ,both P〈0. 050）, while no significant difference was found between AP and FDP （Z= 1.114, P〉0. 050）. Conclusion AP and FDP have higher diagnostic capability for detecting breast cancer in dense breasts compared with MG, and AP may be applied in breast cancer screening.