PDF(Pubmed)
Abstract：
UNASSIGNED: Although immunotherapy has revolutionized the treatment landscape of lung cancer and improved the prognosis of this malignancy, many patients with lung cancer still are not able to benefit from it because of many different reasons. The expression of programmed death ligand-1 (PD-L1) in tumor cells has been approved for the prediction of immunotherapy efficacy; however, its clinical application has been limited by the invasiveness of PD-L1 determination and the heterogeneity of tumor cells. As a promising technology, radiomics has made significant progress in the diagnosis and treatment of lung cancer. Thus, we constructed a noninvasive predictive model which based on radiomics to predict the immunotherapy efficacy of lung caner patients.
UNASSIGNED: Data of 82 patients with stage IIIa/IVb NSCLC who received immunotherapy at the First Affiliated Hospital of Soochow University from December 2019 to January 2023 were retrospectively collected. These patients were followed up for durable clinical benefit (DCB), as defined by whether progression-free survival (PFS) reached 12 months. The least absolute shrinkage and selection operator (LASSO) algorithm was used to screen for the radiomic features in the training set, and a radiomics score (Rad-score) was calculated. The clinical baseline data were analyzed, and the peripheral blood inflammation indices were calculated. Univariate and multivariate analyses were performed to identify the applicable indices, which were combined with the Rad-score to create a comprehensive forecasting model (CFM) and nomograms. Internal validation was performed in the validation set.
UNASSIGNED: Up to the last follow-up time, 48 of 82 patients had a PFS of more than 12 months. The area under the receiver operating characteristic (ROC) curve (AUC) of the Rad-score was 0.858 and 0.812, respectively, in the training set and validation set. A systemic immune-inflammation index (SII) score of <500.88 after two cycles of immunotherapy was a protective factor for PFS >12 months [odds ratio (OR) 0.054; P=0.003]. The CFM had an AUC of 0.930 and 0.922, respectively, in the training and validation sets. The calibration curves and decision curve analysis (DCA) demonstrated the reliability and clinical applicability of the model, respectively.
UNASSIGNED: The radiomics model performed well in predicting whether patients with locally advanced or metastatic NSCLC can achieve DCB after receiving immunotherapy. The CFM had good predictive performance and reliability.
UNASSIGNED: Data of 82 patients with stage IIIa/IVb NSCLC who received immunotherapy at the First Affiliated Hospital of Soochow University from December 2019 to January 2023 were retrospectively collected. These patients were followed up for durable clinical benefit (DCB), as defined by whether progression-free survival (PFS) reached 12 months. The least absolute shrinkage and selection operator (LASSO) algorithm was used to screen for the radiomic features in the training set, and a radiomics score (Rad-score) was calculated. The clinical baseline data were analyzed, and the peripheral blood inflammation indices were calculated. Univariate and multivariate analyses were performed to identify the applicable indices, which were combined with the Rad-score to create a comprehensive forecasting model (CFM) and nomograms. Internal validation was performed in the validation set.
UNASSIGNED: Up to the last follow-up time, 48 of 82 patients had a PFS of more than 12 months. The area under the receiver operating characteristic (ROC) curve (AUC) of the Rad-score was 0.858 and 0.812, respectively, in the training set and validation set. A systemic immune-inflammation index (SII) score of <500.88 after two cycles of immunotherapy was a protective factor for PFS >12 months [odds ratio (OR) 0.054; P=0.003]. The CFM had an AUC of 0.930 and 0.922, respectively, in the training and validation sets. The calibration curves and decision curve analysis (DCA) demonstrated the reliability and clinical applicability of the model, respectively.
UNASSIGNED: The radiomics model performed well in predicting whether patients with locally advanced or metastatic NSCLC can achieve DCB after receiving immunotherapy. The CFM had good predictive performance and reliability.
摘要：
■尽管免疫疗法彻底改变了肺癌的治疗领域并改善了这种恶性肿瘤的预后，由于许多不同的原因，许多肺癌患者仍然无法从中受益。程序性死亡配体-1(PD-L1)在肿瘤细胞中的表达已被批准用于预测免疫治疗的疗效;PD-L1测定的侵袭性和肿瘤细胞的异质性限制了其临床应用。作为一项有前途的技术，影像组学在肺癌的诊断和治疗方面取得了重大进展。因此，我们构建了一个基于影像组学的无创性预测模型来预测肺癌患者的免疫治疗效果。
■回顾性收集2019年12月至2023年1月在苏州大学附属第一医院接受免疫治疗的82例IIIa/IVb期NSCLC患者的数据。对这些患者进行了随访，以获得持久的临床益处（DCB），根据无进展生存期（PFS）是否达到12个月定义。最小绝对收缩和选择算子(LASSO)算法用于筛选训练集中的放射学特征,并计算放射组学评分(Rad-score)。对临床基线资料进行分析,计算外周血炎症指标。进行了单变量和多变量分析，以确定适用的指标，将其与Rad评分相结合，创建综合预测模型(CFM)和列线图。在验证集中执行了内部验证。
■直到最后一次随访时间，82例患者中有48例的PFS超过12个月。Rad评分的受试者工作特征(ROC)曲线(AUC)下面积分别为0.858和0.812,在训练集和验证集中。经过两个周期的免疫治疗后，全身免疫炎症指数（SII）评分<500.88是PFS>12个月的保护因素[比值比（OR）0.054;P=0.003]。CFM的AUC分别为0.930和0.922，在训练集和验证集中。校准曲线和决策曲线分析（DCA）证明了模型的可靠性和临床适用性，分别。
■影像组学模型在预测局部晚期或转移性NSCLC患者在接受免疫治疗后是否可以实现DCB方面表现良好。CFM具有良好的预测性能和可靠性。
■回顾性收集2019年12月至2023年1月在苏州大学附属第一医院接受免疫治疗的82例IIIa/IVb期NSCLC患者的数据。对这些患者进行了随访，以获得持久的临床益处（DCB），根据无进展生存期（PFS）是否达到12个月定义。最小绝对收缩和选择算子(LASSO)算法用于筛选训练集中的放射学特征,并计算放射组学评分(Rad-score)。对临床基线资料进行分析,计算外周血炎症指标。进行了单变量和多变量分析，以确定适用的指标，将其与Rad评分相结合，创建综合预测模型(CFM)和列线图。在验证集中执行了内部验证。
■直到最后一次随访时间，82例患者中有48例的PFS超过12个月。Rad评分的受试者工作特征(ROC)曲线(AUC)下面积分别为0.858和0.812,在训练集和验证集中。经过两个周期的免疫治疗后，全身免疫炎症指数（SII）评分<500.88是PFS>12个月的保护因素[比值比（OR）0.054;P=0.003]。CFM的AUC分别为0.930和0.922，在训练集和验证集中。校准曲线和决策曲线分析（DCA）证明了模型的可靠性和临床适用性，分别。
■影像组学模型在预测局部晚期或转移性NSCLC患者在接受免疫治疗后是否可以实现DCB方面表现良好。CFM具有良好的预测性能和可靠性。
