基于磁共振成像参数、多模态超声构建局部晚期乳腺癌新辅助化疗效果的预测模型
欧兴密 , 郑英 , 左思阳 , 陈蕊 , 闫瑞玲
中国现代医学杂志 ›› 2025, Vol. 35 ›› Issue (16) : 1 -8.
基于磁共振成像参数、多模态超声构建局部晚期乳腺癌新辅助化疗效果的预测模型
Predictive model for neoadjuvant chemotherapy response in locally advanced breast cancer based on MRI parameters and multimodal ultrasound
目的 构建并验证基于磁共振成像(MRI)参数、多模态超声的局部晚期乳腺癌新辅助化疗效果预测模型。 方法 回顾性分析2019年1月—2023年5月中国人民解放军联勤保障部队第九四〇医院收治的238例局部晚期乳腺癌患者的临床资料,按照8∶2随机分为训练集(167例)和内部验证集(71例)。另回顾性分析2023年6月—2024年6月该院收治的66例局部晚期乳腺癌患者的临床资料为外部验证集。患者新辅助化疗后行手术切除并统计病理完全缓解(pCR)情况。新辅助化疗前后行MRI、多模态超声检查,分析影响局部晚期乳腺癌患者新辅助化疗效果的因素,构建并验证基于MRI参数、多模态超声的局部晚期乳腺癌新辅助化疗效果预测模型。 结果 训练集有39例(23.35%)达到pCR,内部验证集与外部验证集分别有16例(22.54%)、16例(24.24%)达到pCR。pCR组肿瘤分期Ⅲ期占比、峰值强度(PI)差值、新辅助化疗后达峰时间(TTP)、TTP差值、新辅助化疗后表观扩散系数(ADC)、ADC差值均高于非pCR组(P <0.05),新辅助化疗后PI低于非pCR组(P <0.05)。多因素逐步Logistic回归分析结果显示:肿瘤分期Ⅲ期[O^R =4.627(95% CI:1.582,13.538)]、ADC差值大[O^R =4.371(95% CI:1.494,12.788)]、PI差值大[O^R =3.785(95% CI:1.294,11.073)]均是影响局部晚期乳腺癌新辅助化疗效果的危险因素(P <0.05)。以影响因素为预测变量,建立列线图预测模型,风险范围0.08~0.56。列线图模型验证结果显示预测局部晚期乳腺癌新辅助化疗效果的校正曲线趋近于理想曲线(P >0.05)。训练集受试者工作特征(ROC)曲线结果显示:列线图模型预测局部晚期乳腺癌新辅助化疗效果的敏感性为79.94%(95% CI:0.673,0.887),特异性为78.66%(95% CI:0.676,0.873),曲线下面积(AUC)为0.823(95% CI:0.751,0.913)。内部和外部验证集ROC曲线结果显示:列线图模型预测局部晚期乳腺癌新辅助化疗效果的敏感性分别为82.67%(95% CI:0.711,0.917)和79.13%(95% CI:0.681,0.878),特异性分别为79.25%(95% CI:0.682,0.879)和83.19%(95% CI:0.715,0.923),AUC分别为0.874(95% CI:0.779,0.983)和0.867(95% CI:0.754,0.962),该模型诊断效能良好。 结论 肿瘤分期、ADC差值、PI差值与局部晚期乳腺癌患者新辅助化疗效果有关,基于此构建局部晚期乳腺癌新辅助化疗效果的预测模型效能良好。
Objective To develop and validate a predictive model for neoadjuvant chemotherapy response in locally advanced breast cancer based on magnetic resonance imaging (MRI) parameters and multimodal ultrasound. Methods A retrospective analysis was conducted on data from 238 patients with locally advanced breast cancer who were treated at the 940 Hospital of the Joint Service Support Force of the Chinese People's Liberation Army from January 2019 to May 2023. They were randomly divided into the training set (167 cases) and the internal validation set (71 cases). Another retrospective analysis was conducted on data from 66 patients with locally advanced breast cancer who were treated at the same hospital from June 2023 to June 2024 for external validation. Patients underwent surgical resections after neoadjuvant chemotherapy, and the pathological complete response (pCR) was analyzed. MRI and multimodal ultrasound examinations were performed before and after neoadjuvant chemotherapy to analyze factors influencing the therapeutic response in patients with locally advanced breast cancer. Based on MRI parameters and multimodal ultrasound features, a predictive model for neoadjuvant chemotherapy efficacy was developed and validated. Results Thirty-nine cases (23.35%) reached pCR in the training set, and 16 (22.54%) and 16 cases (24.24%) reached pCR in the internal validation set and the external validation set, respectively. The proportion of stage III tumors, the change in PI (ΔPI), post-chemotherapy TTP, change in TTP (ΔTTP), post-chemotherapy ADC, and change in ADC (ΔADC) were all significantly higher in the pCR group than in the non-pCR group (P < 0.05), whereas the post-chemotherapy PI was significantly lower in the pCR group (P < 0.05). Multivariable stepwise Logistic regression analysis revealed that stage III tumors [O^R = 4.627 (95% CI: 1.582, 13.538) ], high ΔADC [O^R = 4.371 (95% CI: 1.494, 12.788) ], and high ΔPI [O^R = 3.785 (95% CI: 1.294, 11.073) ] were all risk factors affecting the effect of neoadjuvant chemotherapy in locally advanced breast cancer (P < 0.05). Influencing factors were used as predictor variables to establish a nomogram prediction model, with the predicted risk ranging from 0.08 to 0.56. The calibration curve of the nomogram model for predicting the efficacy of neoadjuvant chemotherapy in patients with locally advanced breast cancer closely approximated the ideal curve (P > 0.05). In the training set, the ROC curve analysis showed that the nomogram had a sensitivity of 79.94% (95% CI: 0.673, 0.887), a specificity of 78.66% (95% CI: 0.676, 0.873), and an area under the curve (AUC) of 0.823 (95% CI: 0.751, 0.913). In the internal and external validation sets, the model demonstrated sensitivities of 82.67% (95% CI: 0.711, 0.917) and 79.13% (95% CI: 0.681, 0.878), specificities of 79.25% (95% CI: 0.682, 0.879) and 83.19% (95% CI: 0.715, 0.923), and AUCs of 0.874 (95% CI: 0.779, 0.983) and 0.867 (95% CI: 0.754, 0.962), respectively, indicating good diagnostic performance of the model. Conclusion Tumor stage, change in ADC, and change in PI are associated with the efficacy of neoadjuvant chemotherapy in patients with locally advanced breast cancer. The predictive model for treatment response constructed based on these factors demonstrates good performance.
| [1] |
SCUDELER M M, MANÓCHIO C, BRAGA PINTO A J, et al. Breast cancer pharmacogenetics: a systematic review[J]. Pharmacogenomics, 2023, 24(2): 107-122. |
| [2] |
GLUZ O, NITZ U, KOLBERG-LIEDTKE C, et al. De-escalated neoadjuvant chemotherapy in early triple-negative breast cancer (TNBC): impact of molecular markers and final survival analysis of the WSG-ADAPT-TN trial[J]. Clin Cancer Res, 2022, 28(22): 4995-5003. |
| [3] |
BLAINE-SAUER S, SAMUELS T L, YAN K, et al. The protease inhibitor amprenavir protects against pepsin-induced esophageal epithelial barrier disruption and cancer-associated changes[J]. Int J Mol Sci, 2023, 24(7): 6765. |
| [4] |
MAO N, SHI Y H, LIAN C, et al. Intratumoral and peritumoral radiomics for preoperative prediction of neoadjuvant chemotherapy effect in breast cancer based on contrast-enhanced spectral mammography[J]. Eur Radiol, 2022, 32(5): 3207-3219. |
| [5] |
胡奇兰, 霍敏, 胡益祺, 术前MRI在预测乳腺癌新辅助化疗后病理反应评估中的价值[J]. 实用放射学杂志, 2023, 39(12): 1962-1966. |
| [6] |
WEKKING D, PORCU M, de SILVA P, et al. Breast MRI: clinical indications, recommendations, and future applications in breast cancer diagnosis[J]. Curr Oncol Rep, 2023, 25(4): 257-267. |
| [7] |
于丹阳, 吴桐, 荆慧, 乳腺癌多模态超声评估新辅助化疗后腋窝淋巴结病理状态的研究[J]. 中华超声影像学杂志, 2022, 31(8): 685-690. |
| [8] |
葛红军, 张子宁, 周菊英, 基于多模态超声参数的列线图预测乳腺癌新辅助化疗后腋窝淋巴结病理完全缓解的价值[J]. 临床超声医学杂志, 2023, 25(12): 996-1000. |
| [9] |
WANG X Y, XIE T S, LUO J R, et al. Radiomics predicts the prognosis of patients with locally advanced breast cancer by reflecting the heterogeneity of tumor cells and the tumor microenvironment[J]. Breast Cancer Res, 2022, 24(1): 20. |
| [10] |
中国抗癌协会乳腺癌专业委员会. 中国抗癌协会乳腺癌诊治指南与规范(2019年版)[J]. 中国癌症杂志, 2019, 29(8): 609-679. |
| [11] |
HAGENAARS S C, de GROOT S, COHEN D, et al. Tumor-stroma ratio is associated with miller-payne score and pathological response to neoadjuvant chemotherapy in HER2-negative early breast cancer[J]. Int J Cancer, 2021, 149(5): 1181-1188. |
| [12] |
莫丹, 陈喜裕, 何婕, 血清microRNA-186-5p、microRNA-328-5p表达和乳腺癌患者临床病理特征与新辅助化疗效果的关系[J]. 中国现代医学杂志, 2023, 33(5): 9-15. |
| [13] |
AEBI S, KARLSSON P, WAPNIR I L. Locally advanced breast cancer[J]. Breast, 2022, 62(Suppl 1): S58-S62. |
| [14] |
KORDE L A, SOMERFIELD M R, CAREY L A, et al. Neoadjuvant chemotherapy, endocrine therapy, and targeted therapy for breast cancer: ASCO guideline[J]. J Clin Oncol, 2021, 39(13): 1485-1505. |
| [15] |
PROVENZANO E. Neoadjuvant chemotherapy for breast cancer: moving beyond pathological complete response in the molecular age[J]. Acta Med Acad, 2021, 50(1): 88-109. |
| [16] |
李阳, 李玉梅, 邓军. 乳腺癌新辅助化疗前后ER、PR、Her-2和Ki-67的变化与化疗疗效的关系分析[J]. 中华全科医学, 2024, 22(9): 1500-1503. |
| [17] |
贺春燕, 张啸飞, 刘兵, 动态增强MRI预测乳腺癌新辅助化疗后病理完全缓解的准确性[J]. 中国临床医学影像杂志, 2022, 33(2): 96-100. |
| [18] |
梁云, 肖运平, 主晓磊, 多模式MRI联合CA125、CA153、CA199预测乳腺癌术后复发转移的临床价值研究[J]. 中国CT和MRI杂志, 2024, 22(2): 92-94. |
| [19] |
WU L, YE W T, LIU Y, et al. An integrated deep learning model for the prediction of pathological complete response to neoadjuvant chemotherapy with serial ultrasonography in breast cancer patients: a multicentre, retrospective study[J]. Breast Cancer Res, 2022, 24(1): 81. |
| [20] |
范晓东, 杨志企, 陈湘光, 多模态MRI影像组学模型预测乳腺癌新辅助化疗疗效不敏感的价值[J]. 实用放射学杂志, 2022, 38(7): 1108-1112. |
| [21] |
LIU Y, WANG Y, WANG Y X, et al. Early prediction of treatment response to neoadjuvant chemotherapy based on longitudinal ultrasound images of HER2-positive breast cancer patients by Siamese multi-task network: a multicentre, retrospective cohort study[J]. EClinicalMedicine, 2022, 52: 101562. |
| [22] |
刘召弟, 杨蔚, 刘开惠, 基于临床、病理及影像特征的列线图预测乳腺癌新辅助治疗后残余小病灶病理完全缓解[J]. 中国医学计算机成像杂志, 2023, 29(1): 32-38. |
| [23] |
GU J H, TONG T, XU D, et al. Deep learning radiomics of ultrasonography for comprehensively predicting tumor and axillary lymph node status after neoadjuvant chemotherapy in breast cancer patients: a multicenter study[J]. Cancer, 2023, 129(3): 356-366. |
| [24] |
LI Y F, CHEN D B, XUAN H J, et al. Construction and validation of prognostic nomogram for metaplastic breast cancer[J]. Bosn J Basic Med Sci, 2022, 22(1): 131-139. |
甘肃省科技计划重点研发项目(23YFFA0035)
/
| 〈 |
|
〉 |