首页|期刊导航|昆明医科大学学报|基于超声造影的列线图预测乳腺癌患者铂类化疗敏感性的构建与验证

基于超声造影的列线图预测乳腺癌患者铂类化疗敏感性的构建与验证OA

Construction and Validation of Nomogram Based on Contrast-enhanced Ultrasound for Predicting Platinum Chemotherapy Sensitivity in Breast Cancer Patients

中文摘要英文摘要

目的 探讨基于超声造影(contrast-enhanced ultrasound,CEUS)定量参数构建的列线图模型对乳腺癌患者铂类药物化疗敏感性的预测价值,为临床治疗提供参考.方法 前瞻性选取广元市中心医院2021年1月至2024年12月收治的275例乳腺癌患者为建模队列,按7∶3比例选取2025年1月至2025年7月118例乳腺癌患者为时间验证队列,均行铂类药物化疗,根据化疗后是否达到病理学完全缓解(pathologic complete response,pCR)分为铂类耐药组(未达到pCR,n=94)、铂类敏感组(达到pCR,n=181).多因素Logistic回归分析乳腺癌患者铂类药物化疗敏感的影响因素,构建列线图预测模型;进行模型验证,以受试者工作特征(receiver operating characteristic,ROC)曲线、校准曲线、决策曲线(decision curve analysis,DCA)评估列线图模型的区分度、一致性及临床实用性.结果 建模队列、时间验证队列临床资料比较均无明显差异(P>0.05);建模队列中两组分子分型、ER表达、Ki-67表达、血流信号分级、淋巴结状态、峰值时间(peak time,TTP)、始增时间(arrival time,AT)、上升支斜率(wash-in rate,WiR)比较差异明显(P<0.05);多因素Logistic回归分析结果显示,分子分型、ER表达、Ki-67表达、淋巴结状态、TTP、AT、WiR均是乳腺癌患者铂类药物化疗耐药的独立影响因素(P<0.05);基于Logistic回归分析结果构建列线图预测模型.ROC曲线显示,预测乳腺癌患者铂类药物化疗耐药的列线图模型在建模队列中预测的曲线下面积(area under the curve,AUC)为0.886(95%CI:0.826~0.946),敏感度为92.55%、特异度为88.95%,在时间验证队列中预测的AUC为0.830(95%CI:0.780~0.880),敏感度为91.43%、特异度为86.75%,模型区分度良好;采用Hosmer-Lemeshow拟合优度检验显示P>0.05,在建模队列、时间验证队列中校准曲线的Brier分数分别为0.122、0.141;DCA曲线提示,当阈概率在40%至80%范围内时,使用该列线图模型指导临床决策能提供比"全治"或"全不治"策略更高的净获益,临床实用性较高.结论 整合了分子分型、ER、Ki-67、淋巴结状态及CEUS定量参数(TTP、AT、WiR)的列线图模型,对乳腺癌患者铂类药物化疗耐药风险具有良好的预测效能.

Objective To explore the predictive value of a nomogram model based on quantitative parameters from contrast-enhanced ultrasound(CEUS)for platinum-based chemotherapy sensitivity in breast cancer patients,providing clinical reference for treatment decisions.Methods A prospective study enrolled 275 breast cancer patients treated at Guangyuan Central Hospital from January 2021 to December 2024 as the modeling cohort,with 118 breast cancer patients from January to July 2025 selected at a 7∶3 ratio as the temporal validation cohort.All patients received platinum-based chemotherapy and were stratified based on achievement of pathologic complete response(pCR)following chemotherapy into platinum-resistant group(without pCR,n=94)and platinum-sensitive group(with pCR,n=181).Multivariate logistic regression was used to analyze the influencing factors of platinum-based chemotherapy sensitivity in breast cancer patients,and a nomogram prediction model was constructed.The discrimination,consistency and clinical practicability of the nomogram model were evaluated by receiver operating characteristic(ROC)curve,calibration curve and decision curve(DCA).Results Clinical characteristics showed no significant differences between the modeling and temporal validation cohorts(P>0.05).In the modeling cohort,significant differences were observed between the two groups in molecular subtype,ER expression,Ki-67 expression,blood flow signal grade,lymph node status,peak time(TTP),arrival time(AT),and wash-in rate(WiR)(P<0.05).Multivariate Logistic regression analysis showed that molecular subtype,ER expression,Ki-67 expression,lymph node status,TTP,AT and WiR were all independent influencing factors for platinum chemotherapy resistance in breast cancer patients(P<0.05).Based on the results of Logistic regression analysis,a nomogram prediction model was constructed.ROC curve showed that the nomogram model for predicting platinum-based chemotherapy resistance had an area under the curve(AUC)of 0.886(95%CI:0.826~0.946)in the modeling cohort with sensitivity of 92.55%and specificity of 88.95%,and an AUC of 0.830(95%CI:0.780~0.880)in the temporal validation cohort with sensitivity of 91.43%and specificity of 86.75%,demonstrating good discrimination.The Hosmer-Lemeshow goodness-of-fit test showed P>0.05,with Brier scores of 0.122 and 0.141 for calibration curves in the modeling and temporal validation cohorts,respectively.The DCA curve suggests that when the threshold probability ranges from 40%to 80%,using this nomogram model for clinical decision-making provides higher net benefit compared to"treat all"or"treat none"strategies,demonstrating good clinical utility.Conclusion A nomogram model integrating molecular subtype,ER,Ki-67,lymph node status,and CEUS quantitative parameters(TTP,AT,WiR)demonstrates good predictive efficacy for platinum-based chemotherapy resistance risk in breast cancer patients.

曹爱萍;王玫;李小双

广元市中心医院超声医学科,四川广元 628000广元市中心医院超声医学科,四川广元 628000广元市中心医院超声医学科,四川广元 628000

医药卫生

乳腺癌超声造影列线图模型化疗铂类敏感铂类耐药

Breast cancerContrast-enhanced ultrasoundNomogram modelChemotherapyPlatinum sensitivePlatinum resistance

《昆明医科大学学报》 2026 (8)

90-99,10

四川省卫生健康科研项目(23PJ134)

10.12259/j.issn.2095-610X.S20260809

评论