基于T2WI影像组学联合临床特征预测浸润性乳腺癌新辅助治疗疗效研究

Prediction of Efficacy of Neoadjuvant Therapy for Invasive Breast Cancer Based on T2WI Radiomics Combined with Clinical Features

目的:探讨T2WI脂肪抑制影像组学联合临床特征对浸润性乳腺癌新辅助治疗(NAT)后病理完全缓解(pCR)的预测价值。方法:回顾性收集本院2021年4月至2022年12月经病理确诊为浸润性乳腺癌并行NAT的104例女性患者资料。依据术后病理评估(Miller/Payne评分系统)将患者分为pCR组(38例)和non-pCR组(66例)。按7∶3比例将入组患者分成训练组(72例)和测试组(32例)。基于磁共振T2WI脂肪抑制序列提取的最佳影像组学特征,单因素分析2组差异有统计学意义的临床特征通过logistic回归构建影像组学模型、临床模型及影像组学和临床特征联合模型。使用受试者工作特征(ROC)曲线等评估模型的预测性能,决策曲线(DCA)对比3种模型的临床效能,DeLong检验比较曲线下面积(AUC)差异。结果:影像组学模型、临床模型及联合模型在训练集中的AUC分别为0.901(95%CI 0.834~0.968)、0.772(95%CI0.661~0.883)和0.955(95%CI 0.913~0.997),在测试集中的AUC分别为0.821(95%CI 0.667~0.975)、0.883(95%CI 0.748~1)和0.925(95%CI 0.845~1)。DeLong检验显示在测试集中,联合模型的预测效果均高于影像组学模型和临床模型(P<0.05)。结论:基于磁共振T2WI脂肪抑制序列影像组学模型,尤其与临床特征构建的联合模型可有效地预测浸润性乳腺癌NAT后疗效,该方法可对临床治疗方案决策提供参考。

Purpose:To investigate the predictive value of T2WI fat suppression radiomics combined with clinical features in the pathological complete remission(pCR)after neoadjuvant therapy(NAT)for invasive breast cancer.Methods:The data of 104 female patients with pathologically diagnosed invasive breast cancer who underwent NAT from April 2021 to December 2022 were retrospectively collected.According to the results of postoperative pathological evaluation(Miller/Payne scoring system),the patients were split into 38 cases in the pCR group and 66 cases in the non-pCR group.They were divided into a training set(n=72)and a test set(n=32)at a ratio of 7∶3.Based on the optimal radiomics characteristics extracted by magnetic resonance T2WI fat suppressed sequence,the statistically significant differences in clinical features between the two groups were analyzed by logistic regression to construct radiomics models,clinical models and joint models of radiomics and clinical features.The prediction capability of the model was evaluated by receiver operating characteristic(ROC)curve,the decision curve analysis(DCA)was used to compare the clinical efficacy of the three models,and the DeLong test was used to compare the area under ROC curve(AUC)difference.Results:In the training set,the AUCs of clinical model,radiomics model and combined model were 0.901(95%CI 0.834-0.968),0.772(95%CI 0.661-0.883)and 0.955(95%CI 0.913-0.997),respectively,while the AUCs in the test set were 0.821(95%CI 0.667-0.975),0.883(95%CI 0.748-1)and 0.925(95%CI 0.845-1),respectively.The DeLong test showed that the prediction effect of the combined model was higher than that of the radiomics model and the clinical model in the test set(P<0.05).Conclusion:The radiomics model based on magnetic resonance T2WI fat suppressed sequence,especially the combined model constructed with clinical features,can be used to effectively predict the efficacy of invasive breast cancer after NAT,and this method can provide a reference for the decision of clinical treatment.