分享:
分享到微信朋友圈
X
临床研究
基于DCE-MRI的3D-MIP重建及多参数评估BI-RADS 4类乳腺肿瘤
梁泓冰 宁宁 赵思奇 李远飞 武玥琪 宋清伟 杨洁 高雪 张莫云 张丽娜

Cite this article as LIANG H B, NING N, ZHAO S Q, et al. 3D-MIP reconstruction and multi parameter evaluation of BI-RADS 4 breast tumors based on DCE-MRI[J]. Chin J Magn Reson Imaging, 2024, 15(5): 94-101.本文引用格式梁泓冰, 宁宁, 赵思奇, 等. 基于DCE-MRI的3D-MIP重建及多参数评估BI-RADS 4类乳腺肿瘤[J]. 磁共振成像, 2024, 15(5): 94-101. DOI:10.12015/issn.1674-8034.2024.05.016.


[摘要] 目的 探讨动态对比增强MRI(dynamic contrast-enhancement MRI, DCE-MRI)瘤周血管特征结合瘤内血流动力学参数在乳腺影像报告和数据系统(breast imaging reporting and data system, BI-RADS)4类肿瘤中的鉴别诊断价值。材料与方法 回顾性分析2018年8月至2023年3月于大连医科大学附属第一医院行乳腺MRI检查为BI-RADS 4类且病理结果明确肿瘤的女性病例102例,其中良性组43例,恶性组59例。记录患者年龄、病灶最大径(dmax)、乳腺DCE-MRI基本影像学特征、瘤周血管特征及瘤内血流动力学参数值。通过单因素和多因素logistic回归分析比较两组间多参数的差异,利用受试者工作特征(receiver operating characteristic, ROC)曲线以及曲线下面积(area under the curve, AUC)分析瘤周血管特征指标与瘤内参数值联合应用对BI-RADS 4类乳腺良恶性两组肿瘤鉴别的诊断效能。应用DeLong检验对AUC进行比较。结果 乳腺良性组和恶性组病例在年龄、dmax、背景实质强化(background parenchymal enhancement, BPE)、纤维腺体组织量(fibroglandular tissue, FGT)、瘤周相邻血管征(adjacent vascular sign, AVS)数目、瘤周血管最大径、患侧瘤周与健侧同一象限血管直径差值(∆d)、瘤周血管出现期相以及瘤内容积转移常数(volume transfer constant, Ktrans)、速率常数(flux rate constant, Kep)、最大增强斜率(maximum slope of increase, MSI)和时间-信号强度曲线(time-signal intensity curve, TIC)类型的差异均具有统计学意义(P<0.05),而病变位置、信号增强率(signal enhancement ratio, SER)和血管外细胞外间隙容积比(volume fraction of extravascular extra vascular space, Ve)差异无统计学意义(P>0.05)。通过多因素logistic回归分析结果显示,∆d、dmax、MSI和Ktrans为区分两组间的独立影响因素,其中优势比最大的是MSI值(AUC为0.923)。将瘤周血管特征∆d分别与dmax、MSI和Ktrans进行两者联合模型比较,以∆d与MSI联合模型的诊断效能最高(AUC为0.933,敏感度和特异度分别为93.2%和83.7%),且∆d联合MSI与∆d联合Ktrans比较的差异具有统计学意义(P=0.001);其他联合指标在两两比较时差异无统计学意义(P>0.05),联合模型高于单独MSI模型的诊断效能。结论 瘤周血管特征指标(∆d)联合瘤内半定量(MSI)血流动力学参数对评价BI-RADS 4类乳腺肿瘤具有较好的鉴别诊断价值。
[Abstract] Objective To ascertain whether dynamic contrast-enhancement MRI (DCE-MRI) is a useful diagnostic tool for intratumoral and peritumoral vascular features in breast imaging reporting and data system (BI-RADS) 4 of tumors.Materials and Methods A retrospective collection of 102 female cases with BI-RADS4 breast MRI examination and clear pathological results from August 2018 to March 2023 at the First Affiliated Hospital of Dalian Medical University, 43 cases were benign group and 59 cases were breast malignant group. Record the patient's age, maximum lesion diameter (dmax), and basic imaging features of breast DCE-MRI, as well as peritumor vascular characteristics and intratumor hemodynamic parameters. Differences in several parameters between the two groups were analyzed by univariate and multivariate Logit models. The diagnostic efficacy of combining various peritumor vascular characteristic indexes and intratumor parameter values to differentiate between benign and malignant BI-RADS4 lesions in the breast was analysed using the receiver operating characteristic, (ROC) curves and the area under the curve (AUC). Evaluate AUC with the DeLong test.Results There were statistically significant differences in number of adjacent vascular signs (AVS), maximum diameter (dmax) of peritumoral blood vessels, the difference in diameter of blood vessels around the tumor between the affected and healthy sides (∆d), peritumoral vascular appearance phase, volume transfer constant (Ktrans), flux rate constant (Kep), maximum slope of enhancement (MSI), type of time signal intensity curve (TIC), background parenchymal enhancement (BPE) and fibrous glandular tissue (FGT) between the two groups of benign and malignant breast cases (P<0.05), while there was no statistically significant difference between signal enhancement ratio (SER) and volume fraction of extracellular space (Ve) (P>0.05). ∆d, dmax, MSI and Ktrans were independent factors that affected how the two groups differentiated from one another, according to multiple logistic regression analysis, with MSI values having the largest predominance ratio (AUC of 0.923). Comparing peritumor vascular characteristic indexes (∆d) with dmax, MSI, and Ktrans, the combined model of ∆d and MSI showed the highest diagnostic performance (AUC value of 0.933, sensitivity and specificity of 93.2% and 83.7%, respectively), and the difference between ∆d combined with MSI and ∆d combined with Ktrans was statistically significant (P=0.001); When comparing additional joint indicators pairwise, there was no statistically significant difference (P>0.05). The joint model performed better diagnostically than the individual MSI model.Conclusions The combination of peritumoral vascular characteristic indexes (∆d) and intratumoral semi quantitative parameters (MSI) has high application value in identifying benign and malignant breast lesions in BI-RADS4.
[关键词] 乳腺肿瘤;动态对比增强;瘤周血管;最大密度投影;磁共振成像
[Keywords] breast tumors;dynamic contrast-enhancement;peritumoral blood vessels;maximum intensity projection;magnetic resonance imaging

梁泓冰 1   宁宁 1   赵思奇 1   李远飞 1   武玥琪 1   宋清伟 1   杨洁 2   高雪 3   张莫云 1   张丽娜 1*  

1 大连医科大学附属第一医院放射科,大连 116011

2 大连医科大学公共卫生学院,大连 116044

3 大连医科大学附属第一医院病理科,大连 116011

通信作者:张丽娜,E-mail:zln201045@163.com

作者贡献声明::张丽娜设计本研究的方案,对稿件重要内容进行了修改;梁泓冰起草和撰写稿件,获取、分析和解释本研究的数据并进行了统计分析;宋清伟、杨洁、宁宁、赵思奇、李远飞、武玥琪、张莫云、高雪获取、分析和解释本研究的数据,对稿件重要内容进行了修改;张丽娜获得了2022辽宁省成人教育学会继续教育教学改革研究课题、2022年度大连市医学重点专科“登峰计划”一般项目资助;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。


基金项目: 2022辽宁省成人教育学会继续教育教学改革研究课题项目 LCYJGZXYB22100 2022年度大连市医学重点专科“登峰计划”一般项目 2022DF042
收稿日期:2024-01-22
接受日期:2024-04-23
中图分类号:R445.2  R737.9 
文献标识码:A
DOI: 10.12015/issn.1674-8034.2024.05.016
本文引用格式梁泓冰, 宁宁, 赵思奇, 等. 基于DCE-MRI的3D-MIP重建及多参数评估BI-RADS 4类乳腺肿瘤[J]. 磁共振成像, 2024, 15(5): 94-101. DOI:10.12015/issn.1674-8034.2024.05.016.

0 引言

       乳腺癌是发病率最高的恶性肿瘤,全世界大约六分之一的女性癌症死亡是由乳腺癌引起的,且从绝对病例数和死亡人数来看,中国位居世界第一[1]。MRI在乳腺癌的精准诊疗中发挥越来越重要的作用,其中动态对比增强MRI(dynamic contrast-enhancement MRI, DCE-MRI)具有多方位和多时相成像的成像优点,利用最大信号强度投影(maximum intensity projection, MIP)重建技术,能立体、精准地呈现出病灶周边微血管的主要分布特点,为评估患者术前情况制订合适治疗方案和监测预后奠定了重要的基础。乳腺影像报告和数据系统(breast imaging reporting and data system, BI-RADS)下MRI分类标准表明影像诊断4类肿块的恶性概率为3%~94%,已有学者进一步对4类肿块具体分类,分为4A(>2%,≤10%)、4B(>10%,≤50%)和4C(>50%,≤94%),且随着BI-RADS类别的增加,恶性肿瘤的阳性预测值(positive predictive values, PPV)逐渐增加[2]。研究显示乳腺肿瘤部分影像学表现与恶性程度有关,包括不规则形态、毛刺边缘、肿块边缘强化和节段分布、不均匀增强和瘤周水肿等常与恶性肿瘤相关[3, 4]。然而,BI-RADS 4类良、恶性病灶的影像学表现有相当大的重叠,MRI诊断特异度变化比较大,其亚分类仍存在挑战,尤其对于血供丰富的肿瘤,其中文献报道约50.2%的病灶经病理证实为良性[5, 6],导致患者进行不必要的穿刺或手术治疗,给临床处理造成许多困扰,增加患者心理及经济负担。目前对此类疾病的研究主要集中于肿瘤内MRI、超声造影和血流定量参数以及基于影像组学诊断等方面的研究[7, 8, 9, 10, 11, 12]。而将基于MIP技术的血管指标应用于乳腺BI-RADS 4类肿瘤的鉴别诊断方面尚未见报道。本研究拟通过基于DCE-MRI观察瘤内以及瘤周血管影像特征进一步评估MR BI-RADS4类乳腺良恶性肿瘤的鉴别诊断价值,以期为该类患者的精准诊治提供新的思路。

1 材料与方法

1.1 一般资料

       回顾性分析2018年8月至2023年3月大连医科大学附属第一医院行乳腺MRI检查结果为BI-RADS 4类且病理诊断明确的120例肿瘤病例资料。纳入标准:(1)在接受MRI检查之前,没有接受过任何化疗、放疗或其他治疗方式;(2)具备MRI平扫和DCE-MRI资料;(3)单侧发病,患者双侧乳房均存在。排除标准:(1)多发病灶;(2)既往患有乳腺炎;(3)既往乳腺有手术瘢痕导致局部血管变形或扭曲;(4)患侧存在无可以辨认的瘤周血管或者健侧同一象限内无可以辨认的血管。根据排除标准共排除18例,最终入组102例进行研究,年龄29~70(47.20±10.57)岁;肿瘤最大径10~82 mm。根据MR-BI-RADS 4类病变的病理结果将其分为良性组和恶性组两组。本研究遵守《赫尔辛基宣言》,经大连医科大学附属第一医院医学伦理委员会批准,免除受试者知情同意,批准文号:PJ-KS-KY-2023-302。

1.2 检查方法

       所有患者均使用8通道乳腺相控阵专用线圈进行3.0 T MRI检查(GE signa HDxt 3.0 T MR,GE Healthcare,美国),绝经前期患者应在月经周期7~10天行MRI检查;DCE-MRI采用三维T1WI序列,具体参数包括:TR 3.9 ms,TE 2.1 ms,视野320 mm×320 mm,层厚4.4 mm,翻转角12°,矩阵256×170,扫描时间480 s。包括1期蒙片和30期对比后图像,每期15 s,在获得蒙片图像后,静脉注射钆喷酸葡胺(中国北京北陆药业股份有限公司),剂量为0.1 mmol/kg,使用磁共振高压注射器(SSMR300EP,拜耳医药保健有限公司,德国)以2.5 mL/s的速率注射,注射结束后采用20 mL生理盐水冲管。

1.3 图像分析与处理

1.3.1 基本影像学特征评估

       由两位有乳腺影像诊断经验的医师(观察者1为具有16年工作经验的主任医师,观察者2为5年工作经验的住院医师)描述并记录乳腺MRI基本影像学特征,两位观察者不知晓患者病理资料。MRI基本影像学特征包括:纤维腺体组织量(fibroglandular tissue, FGT)(分类:几乎全部由脂肪组成、散在纤维腺体组成、不均质纤维腺体组成和绝大部分为纤维腺体组成)、乳腺背景实质强化(background parenchymal enhancement, BPE)(分类:几乎不强化、轻微强化、中度强化和明显强化)[13]。若出现观察结果不一致,由两位观察者商议决定。

1.3.2 DCE相关参数获取

       所有MRI原始数据传送到AW 4.2后处理工作站,由Functool软件进行后处理,在病变强化的最显著期相勾画感兴趣区(region of interest, ROI),面积为5.0~10.0 mm2,包括病灶的实性部分,同时避开正常的纤维腺体、坏死或囊变区域,自动生成参数值。瘤内血流动力学特征包括定量参数即容积转移常数(volume transfer constant, Ktrans)、速率常数(flux rate constant, Kep)、血管外细胞外间隙容积比(volume fraction of extravascular extra vascular space, Ve)、半定量参数即最大增强斜率(maximum slope of increase, MSI)和信号增强率(signal enhancement ratio, SER)。选择病灶最大及上下两个层面三个ROI的平均值作为最终参数测量值,见图1

       时间-信号强度曲线(time-signal intensity curve, TIC)分为流入型[病灶的信号增强率(the slope of signal intensity, SIslope)>10%]、平台型(-10%≤SIslope≤10%)和廓清型(SIslope<-10%)[14]。早期信号增强率SI1%的计算公式为:SI1%=100%×(SI1-SI0)/SI0。SI1为增强后第一期图像信号强度,SI0为增强前信号强度,参考ACR 2013版MR BI-RADS的TIC分型标准,通过计算SI1%将上升曲线(增强早期)分为缓升、中等和速升,若SI1%<50%则为缓升,50%≤SI1%≤100%为中等,SI1%>100%则为速升[15]。根据计算结果本研究将平台型区分为缓升平台型和速升平台型两种。通过3D-MIP后处理软件记录瘤周血管特征包括相邻血管征(adjacent vascular sign, AVS)数目、瘤周血管最大径、瘤周血管出现期相和患侧瘤周与健侧同一象限血管直径差值(∆d),见图1。AVS数目测量采用KUL等[16]提出的标准,选择血管长度≥3 cm且直径≥2 mm,或长度<3 cm且直径≥3 mm的血管指定为有意义的血管进行计数,计数时不须对乳腺动脉和静脉做特殊鉴别,对病灶3 cm范围内的肿瘤血管数目进行统计;还需对病灶侧和健侧同一象限的最粗血管内径进行3次测量,取其平均值分别作为瘤周血管最大径和健侧血管最大径以及进行差值计算∆d,见图1。对两位观察者测量结果进行ICC一致性检验。ICC≤0.20为一致性极差,0.20<ICC≤0.40为一致性较差,0.40<ICC≤0.60为一致性一般,0.60<ICC≤0.80为一致性较好,ICC>0.80为一致性好。

图1  女,58岁,右乳腺浸润性癌。DCE-MRI定量和半定量参数测量及3D-MIP相关参数测量。1A:病变定位,选择ROI,dmax=23 mm;1B:Kep=0.982 min-1;1C:Ktrans=0.681 min-1;1D:Ve=0.534;1E:MSI=5.46;1F:SER=1.08;1G:瘤周血管出现期相为4期;1H:3D-MIP(AVS数目为5条,瘤周血管最大径为4.4 mm,健侧血管最大径为1.6 mm,∆d为2.8 mm)。DCE-MRI:动态对比增强MRI;MIP:最大信号强度投影;ROI:感兴趣区;dmax:病灶最大径;Kep:速率常数;Ktrans:容积转移常数;Ve:血管外细胞外间隙容积比;MSI:最大增强斜率;SER:信号增强率;AVS:瘤周血管数目;∆d:患侧瘤周与健侧同一象限血管直径差值。
Fig. 1  A 58-year-old female presented with invasive carcinoma of the right breast. DCE-MRI quantitative and semi quantitative parameter measurements, as well as 3D-MIP related parameter measurements. 1A: Lesion localization, select ROI, dmax=23 mm; 1B: Kep=0.982 min-1; 1C: Ktrans=0.681 min-1; 1D: Ve=0.534; 1E: MSI=5.46; 1F: SER=1.08; 1G: Phase 4; 1H: 3D-MIP (the number of AVS is 5, the maximum diameter of peritumoral vessels is 4.4 mm, the maximum diameter of contralateral vessels is 1.6 mm, and ∆d is 2.8 mm). DCE-MRI: dynamic contrast-enhancement MRI; MIP: maximum intensity projection; ROI: region of interest; dmax: maximum diameter of lesion; Kep: flux rate constant; Ktrans: volume transfer constant; Ve: volume fraction of extravascular extra vascular space; MSI: maximum slope of increase; SER: signal enhancement ratio; P: phases of peritumoral blood vessels: AVS: number of peritumoral blood vessels; ∆d: difference in diameter of blood vessels in the same quadrant on the affected side and on the unaffected side.

1.4 统计学分析

       采用SPSS 26.0(IBM SPSS statistics, USA)软件包进行统计学分析。采用单因素分析,分别分析两组数据正态性,如果数据服从正态分布,则应用独立样本 t检验;如果不服从正态分布,则应用非参数秩和检验进行分析,分类变量选用卡方检验或Fisher's精确检验进行分析,分别比较两组影像学特征及DCE参数的差异,P<0.05表示差异具有统计学意义。将P<0.05的变量纳入多因素logistic回归分析中,计算具有95%置信区间(confidence interval, CI)的优势比,以评估每个独立因素的相对风险,筛选P<0.05的因素进行联合诊断。应用受试者工作特征(receiver operating characteristic, ROC)曲线评估联合诊断效能,计算诊断恶性病变的敏感度、特异度和曲线下面积(area under the curve, AUC),以确定最佳阈值。采用DeLong检验比较各单一瘤周血管特征指标与瘤内血流动力学相关指标联合应用对乳腺BI-RADS 4类乳腺肿瘤鉴别的诊断效能。

2 结果

2.1 临床病理结果及DCE基本影像学特征

       经术后病理诊断证实良性组肿瘤43例、恶性组肿瘤59例。其中良性组肿块型33例,非肿块10例;恶性组肿块型41例,非肿块型18例。43例良性组中,年龄29~61(43.23±9.71)岁,纤维腺瘤26例(60.5%),纤维腺瘤伴硬化性腺病15例(34.9%),导管内乳头状瘤2例(4.7%);59例恶性组中,患者年龄31~70(50.08±10.28)岁,非特殊型浸润性导管癌50例(84.7%),导管内癌6例(10.2%),神经内分泌癌1例(1.7%)、乳头状癌2例(3.4%)。两组间年龄呈正态分布,差异具有统计学意义(P=0.001)。两组在TIC类型、dmax、BPE和FGT差异均具有统计学意义(P均<0.05),而在病变位置上差异无统计学意义(P>0.05),具体结果见表1

表1  乳腺BI-RADS 4类良恶性肿瘤一般特征比较
Tab. 1  Comparison of general characteristics of breast BI-RADS 4 benign and malignant tumors

2.2 DCE瘤内定量及半定量参数值的诊断效能

       两观察者间参数测量一致性均良好(ICC>0.9)。BI-RADS 4类良恶性肿瘤瘤内定量及半定量参数值见表2。恶性组的Ktrans、Kep和MSI均高于乳腺良性组,Ktrans、Kep和MSI在两组间差异均具有统计学意义(P均<0.001)。两组间Ve和SER值的差异均无统计学意义(P均>0.05)。将差异有统计学意义的参数分别进行诊断效能分析,Ktrans值AUC为0.798,敏感度为67.8%,特异度为86.0%,阈值为0.42 min-1;Kep值AUC为0.819,敏感度为88.1%,特异度为72.1%,阈值为0.45 min-1;MSI值AUC为0.923,敏感度为93.2%,特异度为81.4%,阈值为3.38,具体结果见表3

表2  BI-RADS 4类良恶性肿瘤瘤内血流动力学参数及瘤周MR血管特征参数比较
Tab. 2  Comparison of intratumoral and peritumoral MR vascular characteristic parameters in BI-RADS 4 benign and malignant tumors
表3  DCE各参数及联合模型对乳腺BI-RADS 4类良恶性肿瘤鉴别诊断效能
Tab. 3  The diagnostic efficacy of DCE parameters and combined models in distinguishing benign and malignant breast BI-RADS 4 tumors

2.3 DCE瘤周血管特征的诊断性能

       两观察者间参数测量一致性均良好(ICC>0.9)。BI-RADS 4类良恶性肿瘤瘤周血管特征指标见表2。通过MIP进行测量,良恶性两组间AVS数目、瘤周血管出现期相、瘤周血管最大径和∆d的差异均具有统计学意义(P值分别为<0.001、0.002、<0.001和<0.001)具体结果见表2。将差异有统计学意义的参数分别进行诊断效能分析,AVS数目的AUC值为0.782,敏感度50.8%,特异度95.3%,阈值为4条;瘤周血管出现期相的AUC值为0.359,敏感度69.5%,特异度81.4%,阈值为第5期;瘤周血管最大径的AUC值为0.819,敏感度72.9%,特异度81.4%,阈值为3.75 mm;∆d的AUC值为0.727,敏感度59.3%,特异度86.0%,阈值为1.35 mm;具体结果见表3

2.4 BI-RADS4类恶性肿瘤影像学独立风险因素及多参数联合诊断性能比较

       单因素及多因素logistic回归分析见表4,结果显示∆d、dmax、MSI和Ktrans为两组鉴别诊断的影像学独立风险因素(AUC分别=0.727、0.895、0.923和0.798),其鉴别诊断BI-RADS 4类良恶性肿瘤的单独诊断效能见图2。瘤外、瘤内参数两两联合诊断效能见图3表3,其中以∆d与MSI联合模型的诊断效能最高(AUC值为0.933,敏感度和特异度分别为93.2%和83.7%),高于MSI模型(AUC为0.923),差异具有统计学意义(P<0.05)。∆d与dmax联合模型的AUC值为0.908;∆d与Ktrans联合模型的AUC值为0.818,高于Ktrans模型(AUC为0.798)。DeLong检验∆d联合Ktrans与∆d联合MSI的差异具有统计学意义(P=0.001);其他联合指标两两比较差异无统计学意义(P>0.05)。

图2  BI-RADS 4类乳腺癌独立风险因素诊断效能ROC曲线图。∆d、dmax、MSI和Ktrans的AUC分别为0.727、0.895、0.923和0.798。其中MSI值优势比最大,对应AUC为0.923,敏感度和特异度分别为93.2%和81.4%。
图3  BI-RADS 4类乳腺癌的dmax、MSI和Ktrans分别与∆d进行联合诊断效能ROC曲线图。dmax、MSI和Ktrans分别与∆d进行联合诊断的AUC分别为0.908、0.933和0.818。DeLong检验只有∆d+Ktrans与∆d+MSI的差异具有统计学意义(P=0.001),AUC的差异为0.115;其他两两比较差异无统计学意义。ROC:受试者工作特征;dmax:病灶最大径;∆d:患侧瘤周与健侧同一象限血管直径差值;MSI:最大增强斜率;Ktrans:容积转移常数;AUC:曲线下面积。
Fig. 2  ROCs curve of diagnostic efficacy of independent risk factors for BI-RADS 4 breast cancer. The AUC for ∆d, dmax, MSI, and Ktrans are 0.727, 0.895, 0.923, and 0.798, respectively. The MSI value has the highest advantage ratio, corresponding to an AUC of 0.923, with sensitivity and specificity of 93.2% and 81.4%, respectively.
Fig. 3  ROCs of dmax, MSI and Ktrans of BI-RADS 4 breast cancer and ∆d for joint diagnosis efficiency respectively. The AUC for combined diagnosis of dmax, MSI, and Ktrans with ∆d are 0.908, 0.933, and 0.818, respectively. The DeLong test shows that only the difference between ∆d+Ktrans and ∆d+MSI is statistically significant (P=0.001), and the difference in AUC is 0.115. There is no statistically significant difference between the other pairwise comparisons. ROC: receiver operating characteristic; dmax: maximum diameter of lesion; ∆d: difference in diameter of blood vessels in the same quadrant on the affected side and on the unaffected side; MSI: maximum slope of increase; Ktrans: volume transfer constant; AUC: area under the curve.
表4  乳腺BI-RADS 4类良恶性肿瘤鉴别的单因素及多因素logistic回归分析
Tab. 4  Univariate and multivariate logistic regression analysis for the differentiation of benign and malignant breast BI-RADS 4 tumors

3 讨论

       本研究主要针对BI-RADS 4类乳腺良恶性肿瘤进行瘤外和瘤内多参数指标联合分析,采用单因素分析以及二元logistic回归分析建立联合诊断模型,提示∆d、dmax、MSI和Ktrans是BI-RADS 4类乳腺癌的独立风险因素,进一步将∆d分别与瘤内血流动力学参数进行两两联合诊断效能比较,以∆d与MSI联合模型的诊断效能最高。本研究是国内首次提出通过3D-MIP分析比较患侧瘤周与健侧同一象限血管直径差值(∆d)和AVS数目对BI-RADS 4类乳腺良恶性肿瘤进行鉴别诊断,认为通过联合瘤周血管特征及瘤内血流动力学参数,有助于提高BI-RADS 4类乳腺良恶性肿瘤的鉴别诊断效能和特异度。

3.1 基于3D-MIP的瘤周血管特征在鉴别BI-RADS 4类肿瘤中的价值

       BI-RADS主要是针对动态增强的形态学和血流动力学半定量评价,是目前应用最广泛的乳腺报告描述和诊断系统。DCE-MRI可以通过MIP获得血管的三维减影图像,清晰反映血管在乳腺瘤周中的分布情况。本研究通过3D-MIP测量血管的长度和内径,相比于断层图像血管显示更清晰,测量血管的直径更准确,可以重建出不同层厚的冠状和矢状位成像,避免单纯横断面造成的重叠影响[17, 18]

       乳腺是一种富血供的腺体,乳腺癌相对于乳腺良性肿瘤血供更加丰富。近期研究[19, 20, 21]发现乳腺良恶性肿块或非肿块性病变存在AVS常提示恶性肿瘤,且AVS征、血管直径和动脉和静脉可视化之间的时间间隔(time interval between arterial and venous visualization, AVI)等对于良恶性鉴别效果显著,表明结合乳腺血管形态和功能特征的定量表征可能为诊断乳腺癌提供准确的方法。本研究将BI-RADS 4类良恶性肿瘤瘤周血管特征指标纳入分析,其中AVS的具体数目在鉴别BI-RADS 4类乳腺良恶性肿瘤的AUC值为0.782,特异度高达95.3%,恶性组的患侧瘤周血管数量多≥4条。

       有研究[20, 22, 23]提出乳腺癌患者双侧血管存在一定差异,尤其是肿瘤直径越大时,患侧血管数目及直径高于健侧越显著;乳腺良性肿瘤患侧与健侧血管数目及直径接近,但差异无统计学意义。本研究中恶性组中瘤周血管最大径、∆d相比良性组明显增高,而瘤周血管出现期相缩短,且健侧血管最大径在两组之间差异无统计学意义。我们创新性地提出将乳腺双侧同一象限内的血管最大径进行差值即∆d的测量比较,认为其诊断特异度高于瘤周血管最大径,且在瘤周血管特征指标中,仅∆d是良恶性两组鉴别诊断的独立风险因素,故选择∆d作为瘤周血管特征指标进一步在BI-RADS 4类肿瘤良恶性鉴别诊断的联合效能评估中进行研究。

       本研究DCE每期15 s,且恶性组瘤周血管出现期相多为第4期,早于良性组常出现的第5期,但其差异并不显著。本研究中恶性组多为富血供的非特殊类型浸润性癌性病变,其可能为出现血管期相较良性组提前的原因之一。此外,乳腺癌与乳腺良性肿瘤的脉管系统发育的扭曲度和排列有所不同,恶性程度越高,瘤周供血越丰富,其对比剂越早显影。这与研究[24, 25]表明瘤周动脉血管显影早于良性肿瘤一致。分析其差异不显著的原因在于,本研究未进行瘤周动脉和静脉区分,且样本量不足可能会对结果产生一定影响,更加客观的结论需要更大的样本量及快速DCE成像进行论证。

3.2 DCE血管相关定量及半定量动力学参数在鉴别BI-RADS 4类肿瘤中的价值

       DCE-MRI技术可以对瘤内微血管结构和功能进行非侵入性的半定量和定量分析[26, 27]。研究者通过研究定量参数用以区分乳腺良恶性肿瘤,发现恶性组Ktrans、Kep高于良性组(P<0.05)[28]。XIAO等[29]提出半定量DCE-MRI可用于观察乳腺癌亚群的血管生成微环境,SER、MSI可以非侵入性地补充评估乳腺癌的血管生成,能够反映肿瘤血流灌注程度。本研究依据选择信号强度达峰期相所显示的病灶强化最明显的层面(即病灶的实性部分)对应进行ROI的勾画,同时避开正常的纤维腺体、坏死或囊变区域,获得了良好的测量者间结果一致性。本研究得出,BI-RADS 4类瘤内血流动力学参数中,Ktrans、Kep和MSI在两组间差异均具有统计学意义,且恶性组的Ktrans、Kep值均较良性组高,这与ZHANG等[30]学者研究结果类似。对于恶性肿瘤中Ktrans、Kep值较高的确切原因和病理生理学基础目前尚不完全清楚,大部分研究[26, 27, 31, 32]认为乳腺恶性病变表现为微循环血容量容易增加,通透性明显低于正常乳腺实质,从而Ktrans、Kep值较高;而SER和Ve在两组间差异并无统计学意义,分析其原因在于SER和Ve值更易受多种因素影响,包括细胞密度、血管通透性等,SER值更高表明其TIC类型多为流出型[33],本研究中BI-RADS 4类病例的TIC多为平台型,导致其SER值在两组间差异不显著。

       MSI值是反映TIC上升陡峭程度最好的指标,可以有效反映肿瘤内部对比剂的进出速度,肿瘤的血流灌注以及细胞增殖活跃程度[32, 34]。本研究显示MSI值是单独诊断效能最高的指标,且恶性组MSI值显著高于良性组,其截断值为3.38,AUC和敏感度均在90%以上,特异度为81.4%。KIM等[21]研究发现MSI值与乳腺癌患者的组织病理学分级有关,MSI值越高,说明肿瘤组织恶性程度更高。本研究选用的病例均为单发、单侧病灶,健侧的血液循环对于患侧的影响更小,良恶性肿瘤对比的效果更显著和可靠[35],恶性组多为浸润性癌,病理表现具有血流灌注丰富、血管结构紊乱和管径更粗的特征,未来可以扩大样本在研究中进一步研究MSI值与BI-RAD 4类乳腺癌恶性程度的相关性。

3.3 BI-RADS 4类恶性肿瘤独立风险因素及多参数联合诊断价值

       在临床实践中,BI-RADS 4类病变虽常见,但诊疗选择难度也最大。在专家共识观点一致的情况下,仍无法明确恶性肿瘤的风险分层程度。有研究表明,基于瘤内和瘤周早期动态增强的定量参数的组学模型可用于预测BI-RADS 4类肿瘤的良恶性[32, 34]。本研究BI-RADS 4类良恶性两组间年龄、TIC类型、BPE和FGT差异均具有统计学意义,进一步对肿瘤内外影像多参数进行多因素logistic回归分析后结果显示,仅dmax、∆d、MSI和Ktrans为两组鉴别诊断的独立风险因素。

       乳腺癌肿瘤大小与血管的分布、形态、数目以及代谢情况有关,乳腺癌血管生成对于肿瘤生长至关重要。MILOSEVIC等[36]研究等提出肿瘤病灶大小、微血管密度(microvessel density, MVD)水平与瘤周血供及淋巴血管浸润的发生有关。本研究得出肿瘤最大径越大,其瘤周血管的管径越粗,与上述研究一致。进一步将瘤周血管特征指标∆d分别与MSI、Ktrans和dmax进行联合分析,认为∆d与MSI联合模型的诊断效能最高,敏感度高的同时也提升了特异度(83.7%),两者的联合模型均高于单独MSI模型、∆d模型、∆d与dmax或∆d与Ktrans的联合模型。本研究首次探讨了瘤周血管特征指标联合瘤内血流动力学特征在BI-RADS 4类乳腺良恶性肿瘤鉴别中的价值,有助于为今后BI-RADS 4类病变亚分类研究提供新的诊断依据和思路。

3.4 本研究局限性

       本研究存在一定的不足之处:(1)病种局限且样本量相对较小,需要在后续研究中扩大BI-RADS 4类病种和样本量进一步验证我们的结论,以期为BI-RADS 4类病变的细化提供影像依据;(2)尽管我们根据月经周期选用恰当的时机进行MRI扫描,尽量避免BPE对病灶检出产生影响,但本研究良性组中度强化为主,难免会影响瘤周血管特征的观察;(3)目前评估分类仍依靠影像专家阅片经验,暂未有明确的统一化标准,导致测量结果可能存在误差,以后需要继续完善BI-RADS描述术语和对应分类的相关性,结合人工智能等技术,对病灶进行准确、全面的数据测量。

4 结论

       总之,3D-MIP瘤周血管特征指标联合瘤内DCE多参数有助于提高BI-RADS 4类乳腺良恶性肿瘤的诊断效能。dmax、∆d、MSI和Ktrans是预测BI-RADS 4类乳腺癌的独立影响因素,其中∆d与MSI联合模型的诊断效能最高。今后,可以借助影像组学以及多参数联合成像等手段,进一步挖掘3D-MIP血管参数值在BI-RADS 4类乳腺癌精准诊断中的潜力和价值。

[1]
WILKINSON L, GATHANI T. Understanding breast cancer as a global health concern[J/OL]. Br J Radiol, 2022, 95(1130): 20211033 [2024-01-12]. https://pubmed.ncbi.nlm.nih.gov/34905391/. DOI: 10.1259/bjr.20211033.
[2]
HONDA M, KATAOKA M, KAWAGUCHI K, et al. Subcategory classifications of Breast Imaging and Data System (BI-RADS) category 4 lesions on MRI[J]. Jpn J Radiol, 2021, 39(1): 56-65. DOI: 10.1007/s11604-020-01029-w.
[3]
DE ALMEIDA J R, GOMES A B, BARROS T P, et al. Predictive performance of BI-RADS magnetic resonance imaging descriptors in the context of suspicious (category 4) findings[J]. Radiol Bras, 2016, 49(3): 137-143. DOI: 10.1590/0100-3984.2015.0021.
[4]
FUJIWARA K, YAMADA T, KANEMAKI Y, et al. Grading system to categorize breast MRI in BI-RADS 5th edition: a multivariate study of breast mass descriptors in terms of probability of malignancy[J/OL]. AJR Am J Roentgenol, 2018, 210(3): W118-W127 [2024-01-12]. https://pubmed.ncbi.nlm.nih.gov/29381382/. DOI: 10.2214/AJR.17.17926.
[5]
SHIN S R, KO E Y, HAN B K, et al. Benign adenomyoepithelioma of the breast: imaging characteristics[J]. J Korean Soc Radiol, 2023, 84(2): 398-408. DOI: 10.3348/jksr.2022.0021.
[6]
LEITHNER D, WENGERT G, HELBICH T, et al. MRI in the Assessment of BI-RADS® 4 lesions[J]. Top Magn Reson Imaging, 2017, 26(5): 191-199. DOI: 10.1097/RMR.0000000000000138.
[7]
XING B Y, CHEN X Y, WANG Y L, et al. Evaluating breast ultrasound S-detect image analysis for small focal breast lesions[J/OL]. Front Oncol, 2022, 12: 1030624 [2024-01-12]. https://pubmed.ncbi.nlm.nih.gov/36582786/. DOI: 10.3389/fonc.2022.1030624.
[8]
金金, 何文, 于腾飞, 等. S-Detect联合超声造影对乳腺BI-RADS4类病灶的应用价值[J]. 中华超声影像学杂志, 2023, 32(5): 392-398. DOI: 10.3760/cma.j.cn131148-20221020-00709.
JIN J, HE W, YU T F, et al. Application value of S-Detect combined with contrast-enhanced ultrasound in the Breast Imaging Reporting and Data System 4 breast lesions[J]. Chin J Ultrason, 2023, 32(5): 392-398. DOI: 10.3760/cma.j.cn131148-20221020-00709.
[9]
HONDA M, KATAOKA M, ONISHI N, et al. New parameters of ultrafast dynamic contrast-enhanced breast MRI using compressed sensing[J]. J Magn Reson Imaging, 2020, 51(1): 164-174. DOI: 10.1002/jmri.26838.
[10]
VARGHESE B A, LEE S, CEN S, et al. Characterizing breast masses using an integrative framework of machine learning and CEUS-based radiomics[J]. J Ultrasound, 2022, 25(3): 699-708. DOI: 10.1007/s40477-021-00651-2.
[11]
CUI Q, SUN L, ZHANG Y, et al. Value of breast MRI omics features and clinical characteristics in Breast Imaging Reporting and Data System (BI-RADS) category 4 breast lesions: an analysis of radiomics-based diagnosis[J/OL]. Ann Transl Med, 2021, 9(22): 1677. [2024-01-12]. https://pubmed.ncbi.nlm.nih.gov/34988186/. DOI: 10.21037/atm-21-5441.
[12]
DEBBI K, HABERT P, GROB A, et al. Radiomics model to classify mammary masses using breast DCE-MRI compared to the BI-RADS classification performance[J/OL]. Insights Imaging, 2023, 14(1): 64 [2024-01-12]. https://pubmed.ncbi.nlm.nih.gov/37052738/. DOI: 10.1186/s13244-023-01404-x.
[13]
SPAK D A, PLAXCO J S, SANTIAGO L, et al. BI-RADS® fifth edition: a summary of changes[J]. Diagn Interv Imaging, 2017, 98(3): 179-190. DOI: 10.1016/j.diii.2017.01.001.
[14]
吴祺, 王卓, 宁宁, 等. IVIM联合动态增强MRI在非肿块强化腺病与乳腺癌中的鉴别诊断价值[J]. 磁共振成像, 2023, 14(2): 37-43, 49. DOI: 10.12015/issn.1674-8034.2023.02.007.
WU Q, WANG Z, NING N, et al. Differential diagnostic value of IVIM combining with dynamic enhanced MRI in non-mass enhancement adenosis and breast cancer[J]. Chin J Magn Reson Imag, 2023, 14(2): 37-43, 49. DOI: 10.12015/issn.1674-8034.2023.02.007.
[15]
EDWARDS S D, LIPSON J A, IKEDA D M, et al. Updates and revisions to the BI-RADS magnetic resonance imaging lexicon[J]. Magn Reson Imaging Clin N Am, 2013, 21(3): 483-493. DOI: 10.1016/j.mric.2013.02.005.
[16]
KUL S, CANSU A, ALHAN E, et al. Contrast-enhanced MR angiography of the breast: evaluation of ipsilateral increased vascularity and adjacent vessel sign in the characterization of breast lesions[J]. AJR Am J Roentgenol, 2010, 195(5): 1250-1254. DOI: 10.2214/AJR.10.4368.
[17]
SARDANELLI F, IOZZELLI A, FAUSTO A, et al. Gadobenate dimeglumine-enhanced MR imaging breast vascular maps: association between invasive cancer and ipsilateral increased vascularity[J]. Radiology, 2005, 235(3): 791-797. DOI: 10.1148/radiol.2353040733.
[18]
ÇETINKAYA E, YıLDıZ Ş, OTÇU H, et al. The value of adjacent vessel sign in malignant breast tumors[J]. Diagn Interv Radiol, 2022, 28(5): 463-469. DOI: 10.5152/dir.2022.211228.
[19]
LIU D D, BA Z G, GAO Y, et al. Subcategorization of suspicious non-mass-like enhancement lesions(BI-RADS-MRI Category4)[J/OL]. BMC Med Imaging, 2023, 23(1): 182 [2024-01-12]. https://pubmed.ncbi.nlm.nih.gov/37950164/. DOI: 10.1186/s12880-023-01144-w.
[20]
雷鸣峰, 邓宏亮, 金汉葵, 等. 磁共振成像评价乳腺血供在乳腺良恶性疾病鉴别诊断中的价值[J]. 中国CT和MRI杂志, 2020, 18(10): 70-73. DOI: 10.3969/j.issn.1672-5131.2020.10.021.
LEI M F, DENG H L, JIN H K, et al. The value of ipsilateral breast vascularity in differential diagnosis of benign and\r malignant breast lesions by MRI[J]. Chin J CT MRI, 2020, 18(10): 70-73. DOI: 10.3969/j.issn.1672-5131.2020.10.021.
[21]
KIM J J, KIM J Y, HWANGBO L, et al. Ultrafast dynamic contrast-enhanced MRI using compressed sensing: associations of early kinetic parameters with prognostic factors of breast cancer[J]. AJR Am J Roentgenol, 2021, 217(1): 56-63. DOI: 10.2214/AJR.20.23457.
[22]
SCHULZE A K, HOSKIN T L, MOLDOVEANU D, et al. Tumor characteristics of bilateral breast cancer compared with unilateral breast cancer[J]. Ann Surg Oncol, 2024, 31(2): 947-956. DOI: 10.1245/s10434-023-14451-x.
[23]
侯婉青. 以彩色超声评估乳腺癌患者乳房动脉血管差异[J]. 现代医用影像学, 2021, 30(11): 2137-2139. DOI: 10.3969/j.issn.1006-7035.2021.11.048.
HOU W Q. Evaluation of breast arterial vascular differences in patients with breast cancer by color ultrasound[J]. Mod Med Imageology, 2021, 30(11): 2137-2139. DOI: 10.3969/j.issn.1006-7035.2021.11.048.
[24]
KATAOKA M, HONDA M, OHASHI A, et al. Ultrafast dynamic contrast-enhanced MRI of the breast: how is it used?[J]. Magn Reson Med Sci, 2022, 21(1): 83-94. DOI: 10.2463/mrms.rev.2021-0157.
[25]
ONISHI N, KATAOKA M, KANAO S, et al. Ultrafast dynamic contrast-enhanced MRI of the breast using compressed sensing: breast cancer diagnosis based on separate visualization of breast arteries and veins[J]. J Magn Reson Imaging, 2018, 47(1): 97-104. DOI: 10.1002/jmri.25747.
[26]
TSAI W C, CHANG K M, KAO K J. Dynamic contrast enhanced MRI and intravoxel incoherent motion to identify molecular subtypes of breast cancer with different vascular normalization gene expression[J]. Korean J Radiol, 2021, 22(7): 1021-1033. DOI: 10.3348/kjr.2020.0760.
[27]
CHANG L, LAN H. Effect of Neoadjuvant Chemotherapy on Angiogenesis and Cell Proliferation of Breast Cancer Evaluated by Dynamic Enhanced Magnetic Resonance Imaging[J/OL]. Biomed Res Int, 2022, 3156093 [2024-01-12]. https://pubmed.ncbi.nlm.nih.gov/35915805/. DOI: 10.1155/2022/3156093.
[28]
ZHANG Q, SPINCEMAILLE P, DROTMAN M, et al. Quantitative transport mapping (QTM) for differentiating benign and malignant breast lesion: Comparison with traditional kinetics modeling and semi-quantitative enhancement curve characteristics[J/OL]. Magn Reson Imaging, 2022, 86: 86-93 [2024-01-12]. https://pubmed.ncbi.nlm.nih.gov/34748928/. DOI: 10.1016/j.mri.2021.10.039
[29]
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 [2024-01-12]. https://pubmed.ncbi.nlm.nih.gov/33863924/. DOI: 10.1038/s41523-021-00247-3.
[30]
张晖, 王勇. 磁共振扩散结合灌注成像技术对可疑恶性乳腺病变的鉴别诊断价值[J]. 磁共振成像, 2018, 9(4): 265-269. DOI: 10.12015/issn.1674-8034.2018.04.005.
ZHANG H, WANG Y. The value of diagnosis of suspicious malignant breast lesions by magnetic resonance diffusion combined with perfusion imaging technique[J]. Chin J Magn Reson Imag, 2018, 9(4): 265-269. DOI: 10.12015/issn.1674-8034.2018.04.005.
[31]
AO F, YAN Y, ZHANG Z L, et al. The value of dynamic contrast-enhanced magnetic resonance imaging combined with apparent diffusion coefficient in the differentiation of benign and malignant diseases of the breast[J]. Acta Radiol, 2022, 63(7): 891-900. DOI: 10.1177/02841851211024002.
[32]
THAWANI R, GAO L, MOHINANI A, et al. Quantitative DCE-MRI prediction of breast cancer recurrence following neoadjuvant chemotherapy: a preliminary study[J/OL]. BMC Med Imaging, 2022, 22(1):182 [2024-01-12]. https://pubmed.ncbi.nlm.nih.gov/36266631/. DOI: 10.1186/s12880-022-00908-0
[33]
汪军, 江广斌, 谢兴佳. IVIM及DCE-MRI半定量参数鉴别不同管腔上皮类型乳腺癌的价值[J]. 实用放射学杂志, 2022, 38(10): 1617-1620. DOI: 10.3969/j.issn.1002-1671.2022.10.013.
WANG J, JIANG G B, XIE X J. The value of IVIM and DCE-MRI semi-quantitative parameters in differentiating luminal epithelial types of breast cancers[J]. J Pract Radiol, 2022, 38(10): 1617-1620. DOI: 10.3969/j.issn.1002-1671.2022.10.013.
[34]
LIU H L, ZONG M, WEI H, et al. Differentiation between malignant and benign breast masses: combination of semi-quantitative analysis on DCE-MRI and histogram analysis of ADC maps[J]. Clin Radiol, 2018, 73(5): 460-466. DOI: 10.1016/j.crad.2017.11.026.
[35]
常馨, 卢涛, 黄金昶. 电针围刺诱导小鼠乳腺癌微血管正常化的初步研究[J]. 四川大学学报(医学版), 2023, 54(5): 972-977. DOI: 10.12182/20230960401.
CHANG X, LU T, HUANG J C. Preliminary study on microvasculature normalization induced by peritumoral electroacupuncture in mice with breast cancer xenografts[J]. J Sichuan Univ Med Sci, 2023, 54(5): 972-977. DOI: 10.12182/20230960401.
[36]
MILOSEVIC V, EDELMANN R J, WINGE I, et al. Vessel size as a marker of survival in estrogen receptor positive breast cancer[J]. Breast Cancer Res Treat, 2023, 200(2): 293-304. DOI: 10.1007/s10549-023-06974-4.

上一篇 心脏磁共振特征追踪技术对乳癌患者化疗期间心脏功能的评估价值
下一篇 乳腺X线及MRI特征联合临床病理预测乳腺导管原位癌伴微浸润
  
诚聘英才 | 广告合作 | 免责声明 | 版权声明
联系电话:010-67113815
京ICP备19028836号-2