基于DCE⁃MRI探讨T3、T4直肠癌患者肿瘤形态学与新辅助放化疗后肿瘤退缩相关性  

To Investigate the Correlation between Tumor Morphology and Tumor Regression after Neoadjuvant Chemoradiotherapy in Patients with T3 and T4 Rectal Cancer Based on DCE⁃MRI

作　　者：冯涛 许双燕 刘洋洋 宋雪[1] 曹振东[1] FENG Tao;XU Shuangyan;LIU Yangyang;SONG Xue;CAO Zhendong(Radiology Department,Chengde Medical University Affiliated Hospital,Hebei 067000,China;Department of Ultrasound,Tianjin Hospital,Tianjin 300211,China)

机构地区：[1]承德医学院附属医院放射科,河北067000 [2]天津市天津医院超声科,天津300211

出　　处：《影像科学与光化学》2025年第2期47-54,72,共9页Imaging Science and Photochemistry

基　　金：河北省医学科学研究课题计划(20231368)。

摘　　要：目的:基于动态增强磁共振成像(DCE⁃MRI)探讨T3、T4直肠癌患者肿瘤形态学与新辅助放化疗(neoadjuvant chemoradiotherapy,nCRT)后肿瘤退缩相关性。方法:选取2020年8月至2022年8月我院168例T3、T4直肠癌患者,所有患者均于nCRT前行DCE⁃MRI检查,根据肿瘤退缩分级将患者分为有反应组和无反应组,记录两组DCE⁃MRI肿瘤形态学及定量参数[容积分数(Ve)、容量转移常数(K^(trans))及速率常数(K_(ep))],采用LASSO、Logistic回归方程分析T3、T4直肠癌患者nCRT后肿瘤退缩影响因素,构建列线图预测模型。结果:168例接受nCRT的T3、T4直肠癌患者中,87例纳入有反应组,81例纳入无反应组。两组淋巴结转移、肿瘤分化程度、病灶边缘、病灶形态、病灶内部强化特征、累及周围器官及Ve、K^(trans)、K_(ep)比较差异明显(P<0.05);T3、T4期直肠癌患者病灶边缘、病灶形态、病灶内部强化特征、累及周围器官、K^(trans)、K_(ep)与肿瘤消退呈正相关,Ve与肿瘤消退呈负相关(P<0.05)。选取lambda.lse值为0.015时为最优模型,此时进入模型的变量涉及病灶内部强化特征、累及周围器官、K^(trans)、K_(ep)、V_(e),形态学及定量参数同时纳入Logistic回归分析,发现上述5项变量是T3、T4直肠癌患者nCRT肿瘤退缩影响因素(P<0.05),进一步建立列线图预测模型,该模型AUC为0.939,在训练集和验证集人群中结果与实际观察结果之间相关性良好。结论:T3、T4直肠癌患者nCRT肿瘤退缩受病灶内部强化特征、累及周围器官、K^(trans)、K_(ep)、V_(e)等因素影响,基于DCE⁃MRI检测肿瘤形态学及定量参数对肿瘤退缩预测具有重要价值。Objective:To investigate the correlation between tumor morphology and tumor regression after neoadjuvant chemoradiotherapy(nCRT)in patients with T3 and T4 rectal cancer based on dynamic contrast⁃enhanced magnetic resonance imaging(DCE⁃MRI).Methods:168 patients with T3 and T4 rectal cancer in our hospital were selected from August 2020 to August 2022,all patients underwent DCE⁃MRI before nCRT,and patients were divided into responsive and non⁃responsive groups according to tumor regression grading.The tumor morphology and quantitative parameters[volume fraction(V_(e)),volume metastasis constant(K^(trans))and rate constant(K_(ep))]of the two groups were recorded.LASSO and Logistic regression equations were used to analyze the influencing factors of tumor regression after nCRT in T3 and T4 colorectal cancer patients,and the prediction model of nomogram was constructed.Results:Among 168 T3 and T4 rectal cancer patients who received nCRT,87 cases were included in the response group and 81 cases were included in the non⁃response group.There were significant differences in lymph node metastasis,tumor differentiation degree,lesion margin,lesion morphology,internal enhancement characteristics,involvement of peripheral organs,V_(e),K^(trans)and K_(ep) between the two groups(P<0.05).Lesion margin,lesion morphology,enhancement characteristics,involvement of peripheral organs,K^(trans)and Kep were positively correlated with tumor regression in patients with T3 and T4 rectal cancer,while Ve was negatively correlated with tumor regression(P<0.05).When lambda.lse value was 0.015,the optimal model was selected.At this time,variables entering the model involved internal enhancement features,involvement of surrounding organs,K^(trans),K_(ep),V_(e),morphological and quantitative parameters were also included in Logistic regression analysis.It was found that the above 5 variables were influencing factors of nCRT tumor regression in T3 and T4 colorectal cancer patients(P<0.05).The nomogram prediction model was further establi

关 键 词：直肠癌 动态增强磁共振成像 新辅助放化疗 肿瘤 LOGISTIC 列线图 

分 类 号：R445.2[医药卫生—影像医学与核医学]