IDEAL-IQ序列在乳腺肿块良恶性鉴别诊断中的应用价值探究

分享到微信朋友圈

• 特别关注 •

于佳平杜思瑶韩瑞赵睿萌张立娜

Cite this article as: YU J P, DU S Y, HAN R, et al. Application value of IDEAL-IQ sequence in differential diagnosis of benign and malignant breast masses[J]. Chin J Magn Reson Imaging, 2024, 15(1): 14-20, 42.本文引用格式：于佳平, 杜思瑶, 韩瑞, 等. IDEAL-IQ序列在乳腺肿块良恶性鉴别诊断中的应用价值探究[J]. 磁共振成像, 2024, 15(1): 14-20, 42. DOI:10.12015/issn.1674-8034.2024.01.003.

摘要

[摘要] 目的探讨非对称回波最小二乘估算法迭代水脂分离序列（iterative decomposition of water and fat with echo asymmetrical and least-squares estimation quantitation sequence, IDEAL-IQ）来源的R2^*值在乳腺良恶性肿瘤鉴别诊断中的价值，并与传统多回波T2^*梯度回波（gradient recalled echo, GRE）序列来源的R2^*值进行比较。材料与方法 回顾性分析2021年9月至2023年10月在中国医科大学附属第一医院连续收治的42名患者的50个良性肿瘤病灶，在本院影像归档和通信系统（picture archiving and communication systems, PACS）中使用倾向性评分匹配方法匹配肿瘤所在最大层面的最长径，按1∶3的比例纳入150名患者的150个恶性肿瘤病灶。将恶性肿瘤根据预后因子[雌激素受体（estrogen receptor, ER）、孕激素受体（progesterone receptor, PR）以及人表皮生长因子受体2（human epidermal growth factor receptor 2, HER-2）]的阳性/阴性表达情况进行分组。所有患者均接受包含IDEAL-IQ和多回波T2^* GRE序列的多参数MRI，测量以下定量参数：IDEAL-IQ序列R2^*值（R2^* IDEAL）、多回波T2^* GRE序列R2^*值（R2^* GRE）、表观扩散系数（apparent diffusion coefficient, ADC）及肿瘤长径。根据原始资料类型的不同，分别利用单因素分析（独立样本t检验、Mann-Whitney U检验等方法）对比分析各参数的差异。采用Spearman相关性分析R2^* IDEAL与R2^* GRE及二者与ADC的相关性。采用配对样本t检验比较R2^* IDEAL与R2^* GRE的差异。采用logistic回归分析建立联合诊断模型，并使用受试者工作特征（receiver operating characteristic, ROC）曲线及曲线下面积（area under the curve, AUC）分析单独及联合参数鉴别乳腺肿瘤良恶性的效能。结果相关性分析显示乳腺肿瘤患者的R2^* IDEAL与R2^* GRE呈中度相关（r=0.763，P<0.001），二者与ADC值均呈负性弱相关[r=-0.300（R2^* IDEAL），-0.306（R2^* GRE），P<0.001]。良性组与恶性组中，R2^* IDEAL与R2^* GRE均呈中度相关（r=0.745、0.680，P<0.001），二者与ADC均无相关性。两种序列所得的R2^*值差异有统计学意义（P<0.001）。R2^* IDEAL在良恶性组间差异有统计学意义（P<0.001），管腔HER-2阴性型R2^*值最高。对于单一参数，ADC值鉴别良恶性的AUC最高（0.857）；对于联合参数，R2^* IDEAL+ADC鉴别良性组与管腔阴性组的AUC最高（0.927）；差异均有统计学意义（P<0.05）。结论 IDEAL-IQ序列生成的R2^*值可用于区分良恶性乳腺肿块，可能成为除ADC外辅助乳腺肿瘤良恶性鉴别的又一无需对比剂参数。

[Abstract] Objective To investigate the diagnostic significance of R2^* values obtained from iterative decomposition of water and fat with echo asymmetrical and least-squares estimation quantitation sequence (IDEAL-IQ) in distinguishing between benign and malignant breast tumors, and compare these values with those obtained from traditional multiple echo T2^* gradient recalled echo (GRE) series.Materials and Methods A total of 50 cases of benign tumors in 42 patients admitted to the First Hospital of China Medical University from September 2021 to October 2023 were retrospectively analyzed. The propensity score matching was used to match the longest diameter of the largest plane of the tumor in picture archiving and communication systems (PACS), and 150 cases of malignant tumors in 150 patients were included according to the 1∶3 ratio. Malignant tumors were grouped based on the positive/negative expression of prognostic factors such as estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER-2). All patients underwent multi-parameter MRI with IDEAL-IQ and multi-echo T2^* GRE sequences, and the following quantitative parameters were measured: R2^* IDEAL from IDEAL-IQ sequence, R2^* GRE from multi-echo T2^* GRE sequence, apparent diffusion coefficient (ADC), and tumor diameter. The intra-class correlation coefficient (ICC) was used to evaluate the consistency between the researchers. Depending on the type of raw data, the differences of each parameter were compared and analyzed using one-way analysis (independent samples t-test, Mann-Whitney U-test, etc.). Spearman correlation analysis was used to analyze the correlation between R2^* IDEAL and R2^* GRE, as well as their correlation with ADC. The difference between R2^* IDEAL and R2^* GRE was compared by paired sample t-test. A joint diagnostic model was established by using logistic regression analysis, and the receiver operating characteristic (ROC) curve and the area under the curve (AUC) were used to analyze the efficacy of single and combined parameters in differentiating benign and malignant breast tumors.Results Correlation analysis showed that R2^* IDEAL and R2^* GRE in patients with breast tumors were moderately strongly correlated (r=0.763, P<0.001), and both were weakly negatively correlated with ADC values [r=-0.300 (R2^* IDEAL), -0.306 (R2^* GRE), P<0.001]. In benign group and malignant group, R2^* IDEAL and R2^* GRE showed moderate correlation (r=0.745, 0.680, P<0.001), and there was no correlation between them and ADC. The R2^* values obtained by the two sequences were statistically different (P<0.05). There was a significant difference in R2^* IDEAL between benign and malignant groups (P<0.001), and the R2^* value of luminal HER-2 negative group was the highest. For a single parameter, ADC value had the largest AUC (0.857) in differentiating benign and malignant groups. For the combined parameters, R2^* IDEAL+ADC had the largest AUC (0.927) in differentiating benign group from luminal negative group. The differences were statistically significant (P<0.05).Conclusions The R2^* value generated by IDEAL-IQ sequence can be used to distinguish benign and malignant breast tumors, and may be another non-contrast parameter in addition to ADC to assist the differentiation of benign and malignant breast tumors.

[关键词] 乳腺肿瘤；良恶性鉴别；分子分型；非对称回波最小二乘估算法迭代水脂分离序列；扩散加权成像；磁共振成像

[Keywords] breast neoplasms；distinguish between benign and malignant；molecular subtype；iterative decomposition of water and fat with echo asymmetrical and least-squares estimation quantitation sequence；diffusion weighted imaging；magnetic resonance imaging

作者信息

于佳平 ¹ 杜思瑶 ¹ 韩瑞 ² 赵睿萌 ¹ 张立娜 ^1*

¹ 中国医科大学附属第一医院放射科，沈阳　110001

² 中国医科大学第一临床学院，沈阳　110001

通信作者：张立娜，E-mail：zhanglnda@163.com

作者贡献声明：：于佳平、杜思瑶查阅文献并构思本研究框架，获取收集数据及整理分析，起草和撰写稿件；韩瑞、赵睿萌整理分析及解释本研究数据，对稿件重要内容进行修改；张立娜设计本研究的方案，对稿件的重要内容进行修改，获得国家自然科学基金、辽宁省科学技术基金资助。全体作者都同意发表最后的修改稿，同意对本研究的所有方面负责，确保本研究的准确性和诚信。

出版信息

基金项目： 国家自然科学基金项目 81971695 辽宁省科学技术基金项目 2022JH2/101300027

收稿日期：2023-08-30

接受日期：2023-12-29

中图分类号：R445.2 R737.9

文献标识码：A

DOI: 10.12015/issn.1674-8034.2024.01.003

本文引用格式：于佳平, 杜思瑶, 韩瑞, 等. IDEAL-IQ序列在乳腺肿块良恶性鉴别诊断中的应用价值探究[J]. 磁共振成像, 2024, 15(1): 14-20, 42. DOI:10.12015/issn.1674-8034.2024.01.003.

正文

0 引言

乳腺癌是女性最常见的恶性疾病，且发病率持续上升^[¹^]。作为一种复杂的高度异质性疾病^[²^,³^]，乳腺癌不同的分子分型在恶性程度、治疗方案及预后、转归等方面存在显著差异^[⁴^]。现阶段，临床主要通过术前穿刺活检来获得乳腺癌分子分型，但针芯活检并不能反映肿瘤的整体水平，且易造成水肿、感染等增加患者负担^[⁵^]。乳腺MRI有多序列成像的优势，对乳腺癌检测的敏感度较高^[⁶^]，可以弥补活检的局限性，为临床提供更加全面的参数。目前的乳腺MRI主要依赖于动态对比增强MRI（dynamic contrast-enhanced MRI, DCE-MRI）序列和弥散加权成像（diffusion weighted imaging, DWI）。前者可以提供高分辨率的形态学及微循环灌注信息^[⁷^,⁸^]，但是对比剂的使用可能使部分肾功能不全的患者加重肾损害^[⁹^,¹⁰^]，并且对钆对比剂过敏的患者受到很大的限制^[¹¹^,¹²^]；后者可以提供目前唯一无需注射对比剂的功能参数，即表观扩散系数（apparent diffusion coefficient, ADC），用来评估水分子的布朗运动，反映不同组织中细胞密度的差异^[¹³^]，成为区分恶性和良性乳腺肿瘤的有效参数^[¹⁴^,¹⁵^]，其鉴别效能约为0.8^[¹⁶^,¹⁷^,¹⁸^]，仍有待提升，且ADC值易受b值设置差异的影响^[¹⁹^,²⁰^]。因此，如果基于术前影像学参数可以区分乳腺肿瘤的良恶性及良性肿瘤与不同分子分型的恶性肿瘤，对于临床治疗方案的选择具有重要作用。所以，我们需要探索更多无需对比剂的、高效的定量参数。

非对称回波的最小二乘估算法迭代水脂分离方法（iterative decomposition of water and fat with echo asymmetrical and least-squares estimation quantitation sequence, IDEAL-IQ）是一种新型磁共振水脂分离技术，利用优化回波位移和梯度回波成像，能准确量化脂肪分数（fat fraction, FF）和弛豫率（R2^*）^[²¹^]。IDEAL-IQ序列的扫描时间短，其临床应用包括但不限于骨骼、肌肉系统及心肌与肝脏铁含量的测定等方面。已有研究表明R2^*值可用于评估多种恶性肿瘤缺氧^[²²^,²³^,²⁴^,²⁵^]，血管生成紊乱引起的缺氧已被认为是包括乳腺癌在内的多种实体肿瘤的恶性特征之一^[²⁶^]。目前，同样有研究表明R2^*值在乳腺癌^[²⁷^,²⁸^]的缺氧评估中具有重要作用，但尚未发现R2^*值针对乳腺肿瘤良恶性鉴别及良性乳腺肿瘤与乳腺癌不同分子分型鉴别的文献报道。因此本研究旨在探讨IDEAL-IQ序列获得的R2^*值单独及联合鉴别乳腺肿瘤良恶性的价值，为临床高效诊断、确定治疗方案等提供一定的指导。

1 材料与方法

1.1 临床资料

本研究遵守《赫尔辛基宣言》，经中国医科大学附属第一医院伦理委员会批准，免除受试者知情同意，批准文号：2019-33-2。回顾性分析了2021年9月至2023年10月期间在中国医科大学附属第一医院接受乳腺MRI检查的患者资料。利用本院的影像归档和通信系统（picture archiving and communication systems, PACS），首先选取42名患者的50个良性乳腺肿瘤病灶，采用倾向性评分匹配方法，根据肿瘤所在最大层面的最长径按照1∶3的比例匹配了150名患者的150个恶性肿瘤病灶。所有患者均符合统一的纳入和排除标准。纳入标准：（1）乳腺多参数MRI在术前一周内扫描完成；（2）MRI序列完整，同时包括DWI、IDEAL-IQ、多回波T2^* GRE及DCE-MRI序列；（3）DCE-MRI上显示病灶为肿块型强化。排除标准：（1）初次MRI扫描前接受放疗、化疗或放化疗等在内的任何治疗；（2）病灶直径<5 mm；（3）部分序列图像质量未达到分析标准；（4）临床资料不完整或术后病理诊断不明确。

本研究共纳入192名女性患者，其中8名为双病灶或多病灶，共200例病灶纳入研究。其中恶性病灶150例（年龄29~69岁），包括浸润性导管癌137例、浸润性小叶癌8例、实性乳头状癌1例、导管原位癌4例；良性病灶50例（年龄32~64岁），包括乳腺腺病26例、纤维腺瘤13例、导管内乳头状瘤6例、管状腺瘤2例、腺肌上皮瘤样增生1例、纤维病1例、良性叶状肿瘤1例。

1.2 扫描方法

扫描之前留置静脉针，使用美国GE Healthcare SIGNA Pioneer 3.0 T超导型MR仪器，采用8通道相控阵乳腺线圈。协助患者取俯卧位，双乳自然悬垂于线圈内，足先进，两臂前伸，确定乳头与定位中心相齐，对患者的两侧腋窝及乳腺组织等部位进行扫描，嘱患者保持呼吸均匀，避免影响图像质量。增强扫描时，用高压注射器（MEDRAD^® Spectris Solaris EP，德国拜耳医疗）经手背静脉团注钆双胺（GE医药保健有限公司，美国），剂量为0.2 mmol/kg，速率为2.0 mL/s，之后跟注20 mL生理盐水。对比剂注入前扫描1个时相为蒙片，对比剂注入同时开始连续扫描20个时相。扫描序列及相关参数见表1。

点击查看表格

表1 乳腺MRI扫描具体参数
Tab. 1 Specific parameters of breast MRI scan

1.3 定量参数测量

将图像传至GE Advantage Workstation 4.7工作站，由2名乳腺影像诊断经验分别为5年的住院医师和10年的主治医师利用后处理软件独立测量相关参数。IDEAL-IQ序列扫描完成后系统自动迭代重建水像、脂像、R2^*弛豫图像及脂肪分数图像。（1）生成的四种图像中，脂肪分数图像显示病变的边界最为清晰。参考并对照DCE图像，在脂肪分数图上测量病变最长径。（2）浏览DCE图像确定肿瘤最大层面，在IDEAL R2^*图上定位至同一层面，对照DCE图像沿边缘手动勾画感兴趣区（region of interest, ROI），勾画过程中通过参考T2WI和DCE图像从ROI中排除坏死、囊变及出血区域。每个病灶测量三次，取平均值，生成R2^* IDEAL。（3）使用R2 Star软件将多回波信号高度拟合到单指数衰减曲线以导出T2^*：St=ρ0×e(-TEn/T2*)，其中TEn=第n个回波时间，ρ0=原始信号值，并生成T2^*及R2^*衰减图（R2^*＝1/T2^*）。在R2^*衰减图上勾画ROI，勾画层面的选择及勾画方式同上，生成R2^* GRE。（4）ADC图使用4个b值生成：Sb/S0=exp（-b·ADC），其中S0是无扩散梯度的信号强度。选择病灶最大层面在DWI上沿边缘手动勾画ROI，将ROI自动复制到ADC图中，生成ADC。

1.4 免疫组化指标及分子分型

调取患者的病理记录，获得肿瘤的雌激素受体（estrogen receptor, ER）、孕激素受体（progesterone receptor, PR）和人表皮生长因子受体2（human epidermal growth factor receptor 2, HER-2）的表达情况。ER和PR的阳性肿瘤细胞核≥1%为阳性，<1%为阴性。根据美国临床肿瘤学会/美国病理学家学会指南^[²⁹^,³⁰^]对HER-2的判断标准：阴性和1+判定为HER-2阴性，3+为HER-2阳性；对2+者进一步行荧光原位杂交法检测，基因扩增者定义为HER-2阳性，反之为阴性。根据术后病理分析，分子分型分为管腔HER-2阴性型（ER和/或PR阳性和HER-2阴性）、HER-2阳性型（HER-2阳性、与ER及PR状态无关）和三阴性型乳腺癌（即ER、PR和HER-2阴性）三组。

1.5 统计分析

使用Med Calc 20.0软件和SPSS 25.0软件进行统计分析。通过组内相关系数（intra-class correlation coefficient, ICC）评估研究者间的一致性水平，ICC>0.75认为一致性较好，0.40≤ICC≤0.75为一致性一般，ICC<0.40为一致性较差。采用Shapiro-Wilk检验计量资料的正态性，符合正态分布的计量资料以均数±标准差表示，2组间的比较采用独立样本t检验；偏态分布的计量资料以中位数（四分位数间距）表示，2组间的比较采用Mann-Whitney U检验。对良性组及恶性组及不同分子分型乳腺肿瘤临床病理资料的差异性分析采用卡方检验。采用Spearman相关性分析R2^* IDEAL与R2^* GRE及二者与ADC的相关性，|r|≥0.8为强相关，0.5≤|r|<0.8为中度相关，0.3≤|r|<0.5为弱相关，|r|<0.3为相关程度极弱。对于两种序列得到的R2^*值，符合正态分布的采用配对样本t检验，不符合正态分布的采用Wilcoxon符号秩检验。将有统计学意义的参数分别使用logistic回归建立单因素及多因素联合诊断模型，通过绘制受试者工作特征（receiver operating characteristic, ROC）曲线分析单独及联合参数鉴别乳腺肿瘤良恶性的效能，曲线下面积（area under the curve, AUC）的比较采用DeLong检验。P<0.05为差异有统计学意义。

2 结果

2.1 一般资料

乳腺良、恶性肿瘤患者及不同分子分型乳腺癌患者的临床病理特征见表2、表3。乳腺良、恶性肿瘤患者间年龄、肿瘤大小、绝经状态、病变形态及边缘的差异均不具有统计学意义（P>0.05），两组患者的影像示意图如图1、图2所示；不同分子分型乳腺癌患者间的年龄、肿瘤大小的差异不具有统计学意义（P>0.05），淋巴结转移状态的差异具有统计学意义（P<0.001）。

点击查看大图

图1 女，46岁，左乳纤维腺瘤（黄箭）。1A：R2^* IDEAL-IQ图；1B：多回波T2^* GRE序列R2^*图；1C：DCE-MRI图；1D：ADC图。R2^* IDEAL值为7.8 s^-1，R2^* GRE值为26.89 s^-1，ADC值为1.296×10^-3 mm²/s，最长径为14 mm。
图2 女，63岁，右乳浸润性导管癌（黄箭）。2A：R2^* IDEAL-IQ图；2B：多回波T2^* GRE序列R2^*图；2C：DCE-MRI图；2D：ADC图。R2^* IDEAL值为26.4 s^-1，R2^* GRE值为39.04 s^-1，ADC值为0.443×10^-3 mm²/s，最长径为18 mm。IDEAL-IQ：非对称回波最小二乘估算法迭代水脂分离序列；GRE：梯度回波；DCE-MRI：动态对比增强MRI；ADC：表观扩散系数。
Fig. 1 Female, 46 years old, with fibroadenoma of the left breast (yellow arrow). 1A: R2^* IDEAL-IQ image; 1B: Multi echo T2^* GRE sequence R2^* image; 1C: DCE image; 1D: ADC image. The R2^* IDEAL value is 7.8 s^-1, the R2^* GRE value is 26.89 s^-1, the ADC value is 1.296×10^-3 mm²/s, and the maximum diameter is 14 mm.
Fig. 2 Female, 63 years old, with invasive ductal carcinoma of the right breast (yellow arrow). 2A: R2^* IDEAL-IQ image; 2B: Multi echo T2^* GRE sequence R2^* image; 2C: DCE image; 2D: ADC image. The R2^* IDEAL value is 26.4 s^-1, the R2^* GRE value is 39.04 s^-1, the ADC value is 0.433×10^-3 mm²/s, and the maximum diameter is 18 mm. IDEAL-IQ: iterative decomposition of water and fat with echo asymmetrical and least-squares estimation quantitation sequence; GRE: gradient recalled echo; DCE-MRI: dynamic contrast-enhanced MRI; ADC: apparent diffusion coefficient.

点击查看表格

表2 患者基线特征表
Tab. 2 Patients characteristics

点击查看表格

表3 不同分子分型乳腺癌患者基线特征表
Tab. 3 Patients characteristics with different molecular subtypes of breast cancer

2.2 各定量参数组间一致性检验

两名放射科医师测量的肿瘤R2^* IDEAL值、R2^* GRE值、ADC值均显示出较好的一致性，ICC分别为0.899、0.902、0.953。

2.3 R2* IDEAL与R2* GRE及二者与ADC之间的Spearman相关性分析

相关性分析显示乳腺肿瘤患者的R2^* IDEAL与R2^* GRE呈中度相关（r=0.763，P<0.001），二者与ADC值均呈负性弱相关[r=-0.300（R2^* IDEAL），-0.306（R2^* GRE），P均<0.001]。在良性组与恶性组中，R2^* IDEAL与R2^* GRE均呈中度相关（r=0.745、0.680，P<0.001），但配对样本t检验显示，R2^* IDEAL与R2^* GRE差异有统计学意义（P<0.001），二者与ADC均无相关性（表4）。

点击查看表格

表4 R2^* IDEAL与R2^* GRE及二者与ADC之间的相关性分析
Tab. 4 Correlation analysis: R2* IDEAL with R2* GRE and each with ADC

2.4 良性组与恶性组及良性组与各分子分型的参数分析

良性组与恶性组、各分子分型组的定量参数比较见表5。200例乳腺肿瘤中，恶性组的R2^* IDEAL大于良性组（P<0.001）。在各分子分型中，管腔HER-2阴性组的R2^* IDEAL大于良性组（P<0.001）。良性组与HER-2阳性及三阴性组的差异不具有统计学意义（P<0.05）。良性组的ADC均大于恶性组及各个分子分型（P<0.001）。不同组别代表性患者的R2^*测量值见图3。

点击查看大图

图3 不同组别患者的R2^* IDEAL-IQ图。3A：女，44岁，右乳纤维腺瘤，R2^* IDEAL测量值为26.5 s^-1；3B：女，51岁，左乳浸润性导管癌，R2^* IDEAL测量值为12.5 s⁻¹，分子分型为三阴性型；3C：女，61岁，右乳浸润性导管癌，R2^* IDEAL测量值为25.5 s^-1，分子分型为HER-2阳性型；3D：女，46岁，左乳实性乳头状癌，R2^* IDEAL测量值为53.4 s⁻¹，分子分型为管腔HER-2阴性型。IDEAL-IQ：非对称回波最小二乘估算法迭代水脂分离序列；HER-2：人表皮生长因子受体2；ROI：感兴趣区。
Fig. 3 R2^* iterative decomposition of water and fat with echo asymmetrical and least-squares estimation quantitation sequence (IDEAL-IQ) images of patients in different groups. 3A: Female, 44 years old, with fibroadenoma of the right breast, R2^* IDEAL value is 26.5 s^-1. 3B: Female, 51 years old, with invasive ductal carcinoma of the left breast, R2^* IDEAL value is 12.5 s^-1, triple-negative. 3C: Female, 61 years old, with invasive ductal carcinoma of the right breast, R2^* IDEAL value is 25.5 s^-¹, HER-2-positive. 3D: Female, 46 years old, with solid papillary carcinoma of the left breast, R2^* IDEAL value is 53.4 s^-1, luminal HER-2-negative. IDEAL-IQ: iterative decomposition of water and fat with echo asymmetrical and least-squares estimation quantitation sequence; HER-2: human epidermal growth factor receptor 2; ROI: region of interest.

点击查看表格

表5 良性组与恶性组、各分子分型组的定量参数比较
Tab. 5 Quantitative parameter comparison: benign vs. malignant and molecular subtype groups

2.5 各参数单独及联合鉴别乳腺肿瘤良恶性的效能比较

R2^* IDEAL值、ADC值及联合鉴别良性组与恶性组、良性组与管腔HER-2阴性组乳腺肿瘤患者的效能见表6、表7，并绘制ROC曲线，见图4、图5。单参数中，ADC值鉴别恶性组与良性组乳腺肿瘤的AUC最大（AUC=0.857，P<0.001）；联合参数中，R2^* IDEAL+ADC联合鉴别管腔HER-2阴性组与良性组的AUC最大（AUC=0.927，P<0.001）。DeLong检验示，单参数R2 ^*IDEAL与R2^* IDEAL+ADC联合鉴别管腔HER-2阴性组与恶性组的AUC差异有统计学意义（Z=3.021，P=0.003）。

点击查看大图

图4 各参数单独及联合鉴别良性组与恶性组乳腺肿瘤的受试者工作特征曲线。
图5 各参数单独及联合鉴别良性组与管腔HER-2阴性组乳腺肿瘤的受试者工作特征曲线。HER-2：人表皮生长因子受体2；R2^* IDEAL：非对称回波最小二乘估算法迭代水脂分离序列获得的R2^*值；ADC：表观扩散系数。
Fig. 4 Receiver operating characteristic curve of each parameter alone and in combination to differentiate benign and malignant breast tumors.
Fig. 5 Receiver operating characteristic curve of each parameter alone and in combination to differentiate benign and luminal HER-2-negative group. HER-2: human epidermal growth factor receptor 2; R2^* IDEAL: R2^* value from iterative decomposition of water and fat with echo asymmetrical and least-squares estimation quantitation sequence; ADC: apparent diffusion coefficient.

点击查看表格

表6 各参数单独和联合鉴别乳腺肿瘤良性组与恶性组的效能
Tab. 6 The efficacy of individual and combined parameters in distinguishing benign group from malignant group

点击查看表格

表7 各参数单独及联合鉴别良性组与管腔HER-2阴性组的效能
Tab. 7 The efficacy of individual and combined parameters in distinguishing benign group from luminal HER-2 negative group

3 讨论

本研究通过在IDEAL-IQ序列图像上勾画ROI直接提取肿瘤的R2^*值，探讨了其鉴别乳腺良恶性肿瘤的价值。研究表明，恶性乳腺肿瘤包括管腔HER-2阴性型R2^*值高于良性乳腺肿瘤。联合参数R2^* IDEAL+ADC值对鉴别良性组与管腔HER-2阴性组具有最高的诊断效能，可以为乳腺肿瘤患者的治疗策略提供更准确的信息。

3.1 IDEAL-IQ序列的成像优势分析

既往有研究中R2^*值来源于多回波T2^* GRE序列，但是该序列扫描时间长、需要后续软件的计算，并且生成的R2^*衰减图对于肿瘤的显示的效果不佳。本研究使用的IDEAL-IQ序列作为一种新型磁共振水脂分离技术，可实现全自动计算R2^*图像和R2^*校正以后的脂像、水像以及脂肪分数图，全面展现肿瘤成分，并且可以获得准确量化的R2^*值与脂肪分数值^[³¹^,³²^,³³^]，目前已广泛应用于脂质代谢研究领域以及铁含量的测定等方面。此外，本研究IDEAL-IQ序列的扫描仅需49 s，且无需使用对比剂，因此，添加到常规MRI的IDEAL-IQ序列成像易于在短时间内执行，并且不会造成额外的负担，尤其适用于一些肾功能不全及对比剂过敏的患者。在本研究中，IDEAL-IQ序列获得的R2^*值（R2^* IDEAL）与传统的多回波T2^* GRE序列获得的R2^*值（R2^* GRE）具有明显的差异性，但二者呈中度相关，在一定程度上可以相互替代，从而大致判断R2^*值的大小，为鉴别乳腺肿瘤的良恶性提供一定的价值。

3.2 IDEAL-IQ序列R2*值鉴别乳腺肿瘤良恶性分析

WU等^[³⁴^]发现前列腺癌的平均R2^*值显著高于正常前列腺组织，且较高的R2^*值与较高的前列腺癌Gleason评分显著相关，其原因在于恶性肿瘤快速生长，病变区域相对缺氧，改变了肿瘤恶性潜能的基因表达模式，导致更具攻击性的生存特征^[³⁵^,³⁶^]，并引起肿瘤中脱氧血红蛋白浓度的增加，较高的局部脱氧血红蛋白导致T2^*弛豫时间的减少和R2^*的相应增加^[³⁷^]。目前，R2^*值评估肿瘤氧合状态的可行性已在乳腺癌的研究中得到证实^[²⁷^,³⁸^,³⁹^]，但是据我们所知，尚未有研究证实其对于鉴别乳腺良恶性肿瘤的价值。在本研究中，恶性乳腺肿瘤平均R2^*值明显高于良性乳腺肿瘤。因此，R2^*可作为定量成像生物标志物，为乳腺肿瘤鉴别诊断提供额外信息。

3.3 IDEAL-IQ序列R2*值鉴别良性组与管腔HER-2阴性组乳腺肿瘤分析

既往研究发现，不同分子分型的乳腺肿瘤，其恶性程度^[⁴^]与缺氧程度^[⁴⁰^]均有所差异。WANG等^[⁴¹^]的研究表明，高级别膀胱癌的平均R2^*值显著高于低级别膀胱癌；LIU等^[²⁷^]的研究报道了组织学分级为Ⅱ级、Ⅲ级乳腺恶性肿瘤的R2^*值显著高于Ⅰ级。由这些类似的研究可以推测，R2^*与肿瘤恶性程度呈正相关，最具侵袭性的三阴性乳腺癌仅在纤维化中心表现出强烈的缺氧^[⁴⁰^]，相应R2^*值也应较高，而本研究结果显示，三阴性乳腺癌的R2^*值与良性乳腺肿瘤差异无统计学意义，这可能是由于我们匹配的恶性肿瘤的体积较小，肿瘤边缘的丰富血供^[⁴²^]为肿瘤内部提供了较多的氧气，还尚未形成较大区域的纤维化瘢痕，导致其R2^*值较低。另外，HER-2阳性型的恶性肿瘤表面具有更多的血管生成^[¹⁸^{, 42]}，导致较小体积肿瘤的R2^*值较低。管腔HER-2阴性型的侵袭性及缺氧程度均较低^[⁴⁰^]，并且生长缓慢，有时与良性肿瘤难以鉴别。本研究发现，在同样大小的肿瘤中，管腔HER-2阴性型的R2^*值较高，尤其与良性组的差异较大，且R2^*值在鉴别良性肿瘤与管腔HER-2阴性恶性肿瘤的效能高于ADC值，这可能是由于本研究纳入的肿瘤中最多的是管腔HER-2阴性型，此结果存在一定的偶然性，后续需要更加大量的患者数据进行进一步的比较。此外，R2^*值与ADC值的联合进一步提高了鉴别效能，这可能说明R2^*在区分良性肿瘤及低度恶性肿瘤中具有重要价值。

3.4 IDEAL-IQ序列R2*值与ADC值相关性及互补性分析

ADC值可以提供水分子在肿瘤内扩散程度的信息，间接性反应组织微观结构的改变^[⁴³^,⁴⁴^]，在本研究中，ADC值与R2^*值呈负性弱相关，所以我们推测ADC值无法推测乳腺肿瘤的氧合程度，这与MIYATA等^[³⁸^]的研究结果一致，他们认为ADC值大小与HIF-1α（一种应答缺氧应激的关键调节因子）和纤维化病灶（由缺氧引起的乳腺癌中央瘢痕样区域）的存在与否均无显著相关性。本研究获得的R2^*值可以表征肿瘤氧合状态，与ADC值在良恶性鉴别诊断中呈现互补价值，并且二者联合可以进一步提高了乳腺肿瘤良恶性鉴别的诊断效能，有助于临床进一步针对治疗。

3.5 本研究的局限性

本研究可能存在一些局限性：（1）本研究纳入的HER-2阳性及三阴性乳腺癌患者比例相对较少，可能在结果上造成一定的偏差；（2）本研究未纳入非肿块强化病灶，此后在临床实践中，将进一步探讨对非肿块强化病变良恶性鉴别有价值的影像参数；（3）本研究仅选择肿瘤最大层面勾画ROI并测量各定量参数，不能准确反映全肿瘤特征；（4）本研究对于肿瘤的形态、强化特征等形态学未做探讨，未来的研究中将联合形态学进行更加全面地分析。

4 结论

综上所述，初步研究表明，IDEAL-IQ序列获得的R2^*值可以作为鉴别乳腺肿瘤良恶性的辅助手段，尤其对于管腔HER-2阴性型，单参数效能高于ADC值，与ADC值联合时效能最高，为进一步提高MRI鉴别不同分子分型乳腺肿瘤的准确性提供了新思路。

参考文献

[1]

SIEGEL R, MILLER K, WAGLE N S, et al. Cancer statistics, 2023[J/OL]. CA A Cancer J Clin, 2023, 73: 17-48 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/36633525/. DOI: 10.3322/caac.21763.

[2]

BALIU-PIQUÉ M, PANDIELLA A, OCANA A. Breast cancer heterogeneity and response to novel therapeutics[J/OL]. Cancers, 2020, 12(11): 3271 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/33167363/. DOI: 10.3390/cancers12113271.

[3]

CORTAZAR P, ZHANG L J, UNTCH M, et al. Pathological complete response and long-term clinical benefit in breast cancer: the CTNeoBC pooled analysis[J]. Lancet, 2014, 384(9938): 164-172. DOI: 10.1016/S0140-6736(13)62422-8.

[4]

LEITHNER D, HORVAT J V, MARINO M A, et al. Radiomic signatures with contrast-enhanced magnetic resonance imaging for the assessment of breast cancer receptor status and molecular subtypes: initial results[J/OL]. Breast Cancer Res, 2019, 21(1): 106 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/31514736/. DOI: 10.1186/s13058-019-1187-z.

[5]

JEONG Y S, KANG J, LEE J, et al. Analysis of the molecular subtypes of preoperative core needle biopsy and surgical specimens in invasive breast cancer[J]. J Pathol Transl Med, 2020, 54(1): 87-94. DOI: 10.4132/jptm.2019.10.14.

[6]

CONTI A, DUGGENTO A, INDOVINA I, et al. Radiomics in breast cancer classification and prediction[J/OL]. Semin Cancer Biol, 2021, 72: 238-250 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/32371013/. DOI: 10.1016/j.semcancer.2020.04.002.

[7]

XIAO J, RAHBAR H, HIPPE D S, et al. Dynamic contrast-enhanced breast MRI features correlate with invasive breast cancer angiogenesis[J/OL]. NPJ Breast Cancer, 2021, 7(1): 42 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/33863924/. DOI: 10.1038/s41523-021-00247-3.

[8]

SAD L M A E A, DABEES N L, MOHAMED D A E, et al. Assessment of suspected breast lesions in early-stage triple-negative breast cancer during follow-up after breast-conserving surgery using multiparametric MRI[J/OL]. Int J Breast Cancer, 2022, 2022: 4299920 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/35223102/. DOI: 10.1155/2022/4299920.

[9]

BLOMQVIST L, NORDBERG G F, NURCHI V M, et al. Gadolinium in medical imaging-usefulness, toxic reactions and possible countermeasures-a review[J/OL]. Biomolecules, 2022, 12(6): 742 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/35740867/. DOI: 10.3390/biom12060742.

[10]

WEINREB J C, RODBY R A, YEE J, et al. Use of intravenous gadolinium-based contrast media in patients with kidney disease: consensus statements from the American college of radiology and the national kidney foundation[J]. Radiology, 2021, 298(1): 28-35. DOI: 10.1148/radiol.2020202903.

[11]

AHN Y H, KANG D Y, PARK S B, et al. Allergic-like hypersensitivity reactions to gadolinium-based contrast agents: an 8-year cohort study of 154 539 patients[J]. Radiology, 2022, 303(2): 329-336. DOI: 10.1148/radiol.210545.

[12]

LIU X, ZHANG Y F, CUI X, et al. Gadopentetate meglumine activates mast cells to cause IgE-independent allergic reactions both in vitro and in vivo[J/OL]. Int Immunopharmacol, 2022, 106: 108602 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/35131571/. DOI: 10.1016/j.intimp.2022.108602.

[13]

HORVAT J V, BERNARD-DAVILA B, HELBICH T H, et al. Diffusion-weighted imaging (DWI) with apparent diffusion coefficient (ADC) mapping as a quantitative imaging biomarker for prediction of immunohistochemical receptor status, proliferation rate, and molecular subtypes of breast cancer[J]. J Magn Reson Imaging, 2019, 50(3): 836-846. DOI: 10.1002/jmri.26697.

[14]

PARTRIDGE S C, NISSAN N, RAHBAR H, et al. Diffusion-weighted breast MRI: clinical applications and emerging techniques[J]. J Magn Reson Imaging, 2017, 45(2): 337-355. DOI: 10.1002/jmri.25479.

[15]

MENG L S, ZHAO X, GUO J X, et al. Improved differential diagnosis based on BI-RADS descriptors and apparent diffusion coefficient for breast lesions: a multiparametric MRI analysis as compared to Kaiser score[J]. Acad Radiol, 2023, 30(Suppl 2): S93-S103. DOI: 10.1016/j.acra.2023.03.035.

[16]

CHEN X, HE X J, JIN R, et al. Conspicuity of breast lesions at different b values on diffusion-weighted imaging[J/OL]. BMC Cancer, 2012, 12: 334 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/22853049/. DOI: 10.1186/1471-2407-12-334.

[17]

LIN C X, TIAN Y, LI J M, et al. Diagnostic value of multiple b-value diffusion-weighted imaging in discriminating the malignant from benign breast lesions[J/OL]. BMC Med Imaging, 2023, 23(1): 10 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/36631781/. DOI: 10.1186/s12880-022-00950-y.

[18]

ZHAO M, FU K, ZHANG L, et al. Intravoxel incoherent motion magnetic resonance imaging for breast cancer: a comparison with benign lesions and evaluation of heterogeneity in different tumor regions with prognostic factors and molecular classification[J]. Oncol Lett, 2018, 16(4): 5100-5112. DOI: 10.3892/ol.2018.9312.

[19]

DORRIUS M D, DIJKSTRA H, OUDKERK M, et al. Effect of b value and pre-admission of contrast on diagnostic accuracy of 1.5-T breast DWI: a systematic review and meta-analysis[J]. Eur Radiol, 2014, 24(11): 2835-2847. DOI: 10.1007/s00330-014-3338-z.

[20]

OTIKOVS M, NISSAN N, FURMAN-HARAN E, et al. Diffusivity in breast malignancies analyzed for b > 1000 s/mm2 at 1 mm in-plane resolutions: insight from Gaussian and non-Gaussian behaviors[J]. J Magn Reson Imaging, 2021, 53(6): 1913-1925. DOI: 10.1002/jmri.27489.

[21]

JEON K J, LEE C N, CHOI Y J, et al. Assessment of bone marrow fat fractions in the mandibular condyle head using the iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL-IQ) method[J/OL]. PLoS One, 2021, 16(2): e0246596 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/33635882/. DOI: 10.1371/journal.pone.0246596.

[22]

PENG Y, ZOU X L, CHEN G, et al. Chemical shift-encoded sequence (IDEAL-IQ) and amide proton transfer (APT) MRI for prediction of histopathological factors of rectal cancer[J/OL]. Bioengineering, 2023, 10(6): 720 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/37370651/. DOI: 10.3390/bioengineering10060720.

[23]

ZHANG L J, ZHANG Z L, XU J, et al. Carbogen gas-challenge blood oxygen level-dependent magnetic resonance imaging in hepatocellular carcinoma: initial results[J]. Oncol Lett, 2015, 10(4): 2009-2014. DOI: 10.3892/ol.2015.3526.

[24]

SHI G Z, CHEN H, ZENG W K, et al. R2* value derived from multi-echo Dixon technique can aid discrimination between benign and malignant focal liver lesions[J]. World J Gastroenterol, 2021, 27(12): 1182-1193. DOI: 10.3748/wjg.v27.i12.1182.

[25]

KIM Y, PARK J J, KIM C K. Blood oxygenation level-dependent MRI at 3T for differentiating prostate cancer from benign tissue: a preliminary experience[J/OL]. Br J Radiol, 2022, 95(1131): 20210461 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/34235962/. DOI: 10.1259/bjr.20210461.

[26]

ZHU P P, HE F, HOU Y X, et al. A novel hypoxic long noncoding RNA KB-1980E6.3 maintains breast cancer stem cell stemness via interacting with IGF2BP1 to facilitate c-Myc mRNA stability[J]. Oncogene, 2021, 40(9): 1609-1627. DOI: 10.1038/s41388-020-01638-9.

[27]

LIU M, GUO X J, WANG S K, et al. BOLD-MRI of breast invasive ductal carcinoma: correlation of R2* value and the expression of HIF-1α[J]. Eur Radiol, 2013, 23(12): 3221-3227. DOI: 10.1007/s00330-013-2937-4.

[28]

MCPHAIL L D, ROBINSON S P. Intrinsic susceptibility MR imaging of chemically induced rat mammary tumors: relationship to histologic assessment of hypoxia and fibrosis[J]. Radiology, 2010, 254(1): 110-118. DOI: 10.1148/radiol.2541090395.

[29]

HAMMOND M E, HAYES D F, DOWSETT M, et al. American Society of Clinical Oncology/College Of American Pathologists guideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer[J]. J Clin Oncol, 2010, 28(16): 2784-2795. DOI: 10.1200/JCO.2009.25.6529.

[30]

WOLFF A C, HAMMOND M E, HICKS D G, et al. Recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American Society of Clinical Oncology/College of American Pathologists clinical practice guideline update[J]. J Clin Oncol, 2013, 31(31): 3997-4013. DOI: 10.1200/JCO.2013.50.9984.

[31]

GE M M, ZHANG J, WU B, et al. Effect of gadolinium on hepatic fat quantification using multi-echo reconstruction technique with T2* correction and estimation[J]. Eur Radiol, 2016, 26(6): 1913-1920. DOI: 10.1007/s00330-015-3981-z.

[32]

AOKI T, YAMAGUCHI S, KINOSHITA S, et al. Quantification of bone marrow fat content using iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL): reproducibility, site variation and correlation with age and menopause[J/OL]. Br J Radiol, 2016, 89(1065): 20150538 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/27356277/. DOI: 10.1259/bjr.20150538.

[33]

HISANAGA S, AOKI T, SHIMAJIRI S, et al. Peritumoral fat content correlates with histological prognostic factors in breast carcinoma: a study using iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL)[J]. Magn Reson Med Sci, 2021, 20(1): 28-33. DOI: 10.2463/mrms.mp.2019-0201.

[34]

WU L M, CHEN X X, XUAN H Q, et al. Feasibility and preliminary experience of quantitative T2* mapping at 3.0 T for detection and assessment of aggressiveness of prostate cancer[J]. Acad Radiol, 2014, 21(8): 1020-1026. DOI: 10.1016/j.acra.2014.04.007.

[35]

WANG H, JIANG H, VAN DE GUCHT M, et al. Hypoxic radioresistance: can ROS be the key to overcome it?[J/OL]. Cancers, 2019, 11(1): 112 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/30669417/. DOI: 10.3390/cancers11010112.

[36]

SRIVASTAVA N, USMANI S S, SUBBARAYAN R, et al. Hypoxia: syndicating triple negative breast cancer against various therapeutic regimens[J/OL]. Front Oncol, 2023, 13: 1199105 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/37492478/. DOI: 10.3389/fonc.2023.1199105.

[37]

CHAVHAN G B, BABYN P S, THOMAS B, et al. Principles, techniques, and applications of T2*-based MR imaging and its special applications[J]. Radiographics, 2009, 29(5): 1433-1449. DOI: 10.1148/rg.295095034.

[38]

MIYATA M, AOKI T, SHIMAJIRI S, et al. Evaluation of the R2* value in invasive ductal carcinoma with respect to hypoxic-related prognostic factors using iterative decomposition of water and fat with echo asymmetry and least-squares emission (IDEAL)[J]. Eur Radiol, 2017, 27(10): 4316-4323. DOI: 10.1007/s00330-017-4832-x.

[39]

JIANG L, WEATHERALL P T, MCCOLL R W, et al. Blood oxygenation level-dependent (BOLD) contrast magnetic resonance imaging (MRI) for prediction of breast cancer chemotherapy response: a pilot study[J]. J Magn Reson Imaging, 2013, 37(5): 1083-1092. DOI: 10.1002/jmri.23891.

[40]

BENNANI-BAITI B, PINKER K, ZIMMERMANN M, et al. Non-invasive assessment of hypoxia and neovascularization with MRI for identification of aggressive breast cancer[J/OL]. Cancers, 2020, 12(8): 2024 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/32721996/. DOI: 10.3390/cancers12082024.

[41]

WANG Y C, SHEN Y Q, HU X M, et al. Application of R2* and apparent diffusion coefficient in estimating tumor grade and T category of bladder cancer[J]. AJR Am J Roentgenol, 2020, 214(2): 383-389. DOI: 10.2214/AJR.19.21668.

[42]

SYED A K, WHISENANT J G, BARNES S L, et al. Multiparametric Analysis of Longitudinal Quantitative MRI data to Identify Distinct Tumor Habitats in Preclinical Models of Breast Cancer[J/OL]. Cancers, 2020, 12(6): 1682 [2023-08-29]. https://pubmed.ncbi.nlm.nih.gov/32599906/. DOI: 10.3390/cancers12061682.

[43]

BALTZER P, MANN R M, IIMA M, et al. Diffusion-weighted imaging of the breast-a consensus and mission statement from the EUSOBI International Breast Diffusion-Weighted Imaging working group[J]. Eur Radiol, 2020, 30(3): 1436-1450. DOI: 10.1007/s00330-019-06510-3.

[44]

HAAK H E, MAAS M, LAHAYE M J, et al. Selection of patients for organ preservation after chemoradiotherapy: MRI identifies poor responders who can go straight to surgery[J]. Ann Surg Oncol, 2020, 27(8): 2732-2739. DOI: 10.1245/s10434-020-08334-8.

上一篇 MRI及临床病理特征对乳腺癌人表皮生长因子受体2表达状态的鉴别诊断价值

下一篇 MRI特征在BRCA基因突变与非突变型乳腺癌间的差异研究