文献速递肿瘤 PET/CT核素与放射性配体治疗

基于深度学习的[18F]FDG PET/CT无创预测食管癌PD-L1状态及免疫治疗生存分层

Deep learning-based non-invasive prediction of PD-L1 status and immunotherapy survival stratification in esophageal cancer using [18F]FDG PET/CT.
PubMed 文献 · Eur J Nucl Med Mol Imaging · 2025-08-14

本研究回顾性纳入两个中心331例食管癌患者,开发并验证基于治疗前[18F]FDG PET/CT的四种3D ResNet10深度学习模型预测PD-L1状态,并对50例接受免疫治疗的患者随访评估衍生深度学习模型评分(DLS)的生存分层价值。模型预测PD-L1状态准确,CPC和PC模型表现更优,多数模型DLS在合适截断点可有效分层总生存和无进展生存(P<0.05),优于基于PD-L1状态(联合阳性评分≥10)的分层。

要点

  • 基于[18F]FDG PET/CT的深度学习模型可准确预测食管癌PD-L1状态
  • CPC和PC模型预测性能优于PET、CT及临床模型
  • 多数模型DLS可有效分层免疫治疗后总生存和无进展生存

核心结论

  • 基于[18F]FDG PET/CT的深度学习模型可准确预测食管癌患者PD-L1状态。
  • 衍生DLS可有效分层免疫治疗后的生存结局,尤其与临床因素联合时。

研究设计与样本

  • 回顾性研究,纳入两个中心331例食管癌患者,分为训练、内部验证和外部验证队列。
  • 对50例接受免疫治疗的患者进行随访。
  • 开发四种基于3D ResNet10的模型:PET+CT+临床因素(CPC)、PET+CT(PC)、PET(P)和CT(C),并构建纳入临床因素的逻辑回归模型(临床模型)作为比较。
  • 评估DLS作为生存分层的影像学标志物,并构建预测生存的列线图。

主要发现

  • 训练队列预测PD-L1状态的AUC:CPC 0.927、PC 0.904、P 0.886、C 0.934、临床模型0.603。
  • 内部验证队列AUC:CPC 0.882、PC 0.848、P 0.770、C 0.745、临床模型0.524。
  • 外部验证队列AUC:CPC 0.843、PC 0.806、P 0.759、C 0.667、临床模型0.671。
  • CPC和PC模型表现出更优的预测性能。
  • 生存分析显示多数模型的DLS在合适截断点可有效分层总生存和无进展生存(P<0.05),优于基于PD-L1状态(联合阳性评分≥10)的分层。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

PURPOSE: This study aimed to develop and validate deep learning models using [18F]FDG PET/CT to predict PD-L1 status in esophageal cancer (EC) patients. Additionally, we assessed the potential of derived deep learning model scores (DLS) for survival stratification in immunotherapy. METHODS: In this retrospective study, we included 331 EC patients from two centers, dividing them into training, internal validation, and external validation cohorts. Fifty patients who received immunotherapy were followed up. We developed four 3D ResNet10-based models-PET + CT + clinical factors (CPC), PET + CT (PC), PET (P), and CT (C)-using pre-treatment [18F]FDG PET/CT scans. For comparison, we also constructed a logistic model incorporating clinical factors (clinical model). The DLS were evaluated as radiological markers for survival stratification, and nomograms for predicting survival were constructed. RESULTS: The models demonstrated accurate prediction of PD-L1 status. The areas under the curve (AUCs) for predicting PD-L1 status were as follows: CPC (0.927), PC (0.904), P (0.886), C (0.934), and the clinical model (0.603) in the training cohort; CPC (0.882), PC (0.848), P (0.770), C (0.745), and the clinical model (0.524) in the internal validation cohort; and CPC (0.843), PC (0.806), P (0.759), C (0.667), and the clinical model (0.671) in the external validation cohort. The CPC and PC models exhibited superior predictive performance. Survival analysis revealed that the DLS from most models effectively stratified overall survival and progression-free survival at appropriate cut-off points (P < 0.05), outperforming stratification based on PD-L1 status (combined positive score ≥ 10). Furthermore, incorporating model scores with clinical factors in nomograms enhanced the predictive probability of survival after immunotherapy. CONCLUSION: Deep learning models based on [18F]FDG PET/CT can accurately predict PD-L1 status in esophageal cancer patients. The derived DLS can effectively stratify survival outcomes following immunotherapy, particularly when combined with clinical factors.

来源:Eur J Nucl Med Mol Imaging,查看原文。版权归原作者所有。

引用

  1. Xie F, Zhang M, Zheng C, et al. Deep learning-based non-invasive prediction of PD-L1 status and immunotherapy survival stratification in esophageal cancer using [18F]FDG PET/CT. Eur J Nucl Med Mol Imaging. 2025. PMID: 40810801. doi:10.1007/s00259-025-07463-0

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。