| 孟帆,袁军辉,方少伯,张孝先,郭兰伟,陈天东,张宏凯,曲金荣,张仁知,陈学军.T1 mapping联合弥散加权成像预测浸润性乳腺癌肿瘤浸润淋巴细胞水平[J].中国医学影像技术,2025,41(1):84~89 |
| T1 mapping联合弥散加权成像预测浸润性乳腺癌肿瘤浸润淋巴细胞水平 |
| Combining T1 mapping and diffusion weighted imaging for predicting tumor-infiltrating lymphocyte level in invasive breast cancer |
| 投稿时间：2024-07-02 修订日期：2024-10-23 |
| DOI：10.13929/j.issn.1003-3289.2025.01.018 |
| 中文关键词: 乳腺肿瘤 淋巴细胞,肿瘤浸润 磁共振成像 |
| 英文关键词:breast neoplasms lymphocytes, tumor-infiltrating magnetic resonance imaging |
| 基金项目:二○二四年度河南省医学科技攻关计划项目(LHGJ20240123)。 |
| 作者 | 单位 | E-mail | | 孟帆 | 郑州大学附属肿瘤医院/河南省肿瘤医院医学影像科, 河南 郑州 450008 | | | 袁军辉 | 郑州大学附属肿瘤医院/河南省肿瘤医院医学影像科, 河南 郑州 450008 | | | 方少伯 | 郑州大学人民医院/河南省人民医院影像科, 河南 郑州 450008 | | | 张孝先 | 郑州大学附属肿瘤医院/河南省肿瘤医院医学影像科, 河南 郑州 450008 | | | 郭兰伟 | 郑州大学附属肿瘤医院/河南省肿瘤医院肿瘤防治研究办公室, 河南 郑州 450008 | | | 陈天东 | 郑州大学附属肿瘤医院/河南省肿瘤医院病理科, 河南 郑州 450008 | | | 张宏凯 | 郑州大学附属肿瘤医院/河南省肿瘤医院医学影像科, 河南 郑州 450008 | | | 曲金荣 | 郑州大学附属肿瘤医院/河南省肿瘤医院医学影像科, 河南 郑州 450008 | | | 张仁知 | 国家癌症中心/国家肿瘤临床医学研究中心/中国医学科学院北京协和医学院肿瘤医院影像诊断科, 北京 100021 | | | 陈学军 | 郑州大学附属肿瘤医院/河南省肿瘤医院医学影像科, 河南 郑州 450008 | chenxuejun1967@163.com |
|
| 摘要点击次数: 392 |
| 全文下载次数: 94 |
| 中文摘要: |
| 目的 观察T1 mapping联合弥散加权成像(DWI)术前无创预测浸润性乳腺癌的肿瘤浸润淋巴细胞(TIL)水平的价值。方法 回顾性纳入143例浸润性乳腺癌并根据术后病理所见示TIL水平分为高水平组(TIL≥10%,n=73)与低水平组(TIL＜10%,n=70)。记录临床-病理信息,观察乳腺癌病灶MRI特征,测量其平均T1值(T1mean)、ADC值(ADCmean)并进行组间比较。以多因素logistic回归筛选TIL水平的独立预测因素,基于回归模型绘制列线图,以受试者工作特征(ROC)曲线及其曲线下面积(AUC)评估其预测价值。结果 相比低水平组,高水平组人表皮生长因子受体2(human epidermal growth factor receptor 2, HER2)阳性占比较高(P＜0.05),病灶呈圆形/椭圆形、边缘光整占比较高(P均＜0.05),而瘤周水肿占比较低(P＜0.05)。组间病灶强化模式差异有统计学意义(P＜0.05)。高水平组T1mean、ADCmean均高于低水平组(P均＜0.05)。强化模式、T1mean及ADCmean均为乳腺癌TIL水平的独立预测因素,以之联合构建预测模型并绘制列线图预测TIL水平的AUC为0.848,高于单一病灶强化模式(AUC=0.569;Z=5.384,P＜0.05)及T1mean(AUC=0.662;Z=3.876,P＜0.05)而与ADCmean(AUC=0.814)差异无统计学意义(Z=1.578,P=0.115)。决策曲线分析显示列线图临床应用价值较好。结论 T1 mapping联合DWI可于术前有效预测乳腺癌TIL水平。 |
| 英文摘要: |
| Objective To observe the value of T1 mapping combining diffusion weighted imaging (DWI) for noninvasive preoperative predicting tumor-infiltrating lymphocyte (TIL) level in invasive breast cancer. Methods Totally 143 patients with invasive breast cancer were retrospectively collected and divided into high group (TIL≥10%, n=73) and low group (TIL＜10%, n=70) according to TIL level by postoperation pathology. Clinicopathological information were collected, MRI features of breast cancer lesions were documented, mean T1 values (T1mean) and mean ADC values (ADCmean) were measured,and then were compared between groups. Multivariate logistic regression analysis was used to identify independent predictive factors of TIL levels, and a nomogram was constructed based on regression model. The receiver operating characteristic (ROC) curve and the area under the curve (AUC) were used to evaluate the predictive value for TIL levels. Results Compared with low group, high group had higher proportion of human epidermal growth factor receptor 2 (HER2) positivity (P＜0.05), and showed more circular/oval shapes and more smooth margins but less peritumoral edema (all P＜0.05). Significant differences of lesions enhancement pattern was found between groups (P＜0.05). T1mean and ADCmean were both higher in high group than those in low group (both P＜0.05). Lesions enhancement pattern, T1mean and ADCmean were all independent predictors of TIL levels in breast cancer. The AUC of nomogram combining the above 3 factors for predicting TIL level was 0.848, significantly higher than that of lesions enhancement pattern (AUC=0.569, Z=5.384, P＜0.05) and T1mean (AUC=0.662, Z=3.876, P＜0.05), but not statistically different with that of ADCmean (AUC=0.814, Z=1.578, P=0.115). Decision curve analysis showed that this nomogram had good clinical application value. Conclusion Combining T1 mapping and DWI could effectively predict level of TIL level in breast cancer before surgery. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
