文献速递乳腺癌放疗放疗物理与技术

乳腺癌放疗副作用预测:临床与基因组数据整合

Prediction of Side Effects from Breast Radiation Therapy-Integration of Clinical and Genomic Data.
PubMed 文献 · Int J Radiat Oncol Biol Phys · 2026-08-18

基于多中心REQUITE队列(n=2067)回顾性分析,比较12种机器学习模型预测乳腺癌放疗后24个月8类副作用的表现。仅用临床数据时随机森林最佳(平均AUC-ROC=0.84,5×重复10折交叉验证);联合临床与单核苷酸多态性数据时逻辑回归整体性能最高(平均AUC-ROC=0.92),集成树分类器次之(0.91)。手臂淋巴水肿预测最佳,其次为乳房水肿和乳头回缩。

要点

  • 联合临床与SNP数据时逻辑回归预测性能最高(平均AUC-ROC=0.92)且透明性更好。
  • 仅用临床数据时随机森林表现最佳(平均AUC-ROC=0.84)。
  • 手臂淋巴水肿是最稳健的预测终点,优于乳房水肿和乳头回缩。

核心结论

  • 乳腺癌放疗副作用预测高度依赖终点,淋巴水肿是特别稳健的结局。
  • 联合临床与单核苷酸多态性数据时,逻辑回归性能与复杂方法相当且更透明。

研究设计与样本

  • 多中心REQUITE队列回顾性分析,纳入2067例接受放疗的乳腺癌患者。
  • 评估12种机器学习模型,采用重复k折交叉验证,以AUC-ROC和PR曲线下面积评价。

主要发现

  • 仅用临床数据时随机森林平均AUC-ROC=0.84(5×重复10折交叉验证)。
  • 联合临床与SNP数据时逻辑回归平均AUC-ROC=0.92(5×重复10折交叉验证)。
  • 集成树分类器联合数据平均AUC-ROC=0.91(5×重复10折交叉验证)。
  • 手臂淋巴水肿预测最佳,其次为乳房水肿和乳头回缩。

关键数值与推荐

  • 副作用评估时间为放疗后24个月,共8个终点。
  • 预测性能以平均AUC-ROC和PR曲线下面积衡量,采用5×重复10折交叉验证。

临床怎么用

  • 可基于临床数据用随机森林初步预测放疗副作用风险。
  • 联合SNP数据时优先考虑逻辑回归,兼顾性能与透明度。
  • 淋巴水肿可作为重点预测和监测的副作用。

局限性

  • 需外部验证和成本效益评估。
  • 为回顾性分析,证据级别受限。

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: To compare the ability of different machine learning models to predict the risk of side effects in patients with breast cancer undergoing radiation therapy. METHODS AND MATERIALS: Data from the multicenter REQUITE cohort (n = 2067) were analyzed retrospectively. Side effects (8 endpoints) were assessed 24 months post radiation therapy. Clinical, treatment, and genetic data were available. Predictive performance of 12 machine learning models was assessed using area under the receiver operating characteristic curve (AUC-ROC) and area under the precision-recall curve, with n-repeated k-fold cross-validation (CV). RESULTS: Random forest achieved the best performance using clinical data alone (mean AUC-ROC = 0.84, 5 × repeated 10-fold CV). Combining clinical and single nucleotide polymorphism data yielded the highest overall performance with logistic regression (mean AUC-ROC = 0.92, 5 × repeated 10-fold CV), as the highest followed by ensemble tree classifiers (mean AUC-ROC = 0.91, 5 × repeated 10-fold CV). The best prediction was obtained for arm lymphedema, followed by breast edema and nipple retraction. CONCLUSIONS: In this large multicenter cohort, radiation therapy side effect prediction was highly endpoint dependent, with lymphedema emerging as a particularly robust outcome. When combining clinical and single nucleotide polymorphism data, logistic regression performed comparably to more complex approaches, while offering greater transparency. External validation and cost-benefit evaluation are required, but our findings provide a comparative benchmark and support the feasibility of targeted prediction of breast radiation therapy side effects.

来源:Int J Radiat Oncol Biol Phys,查看原文。版权归原作者所有。

引用

  1. Janapy AA, Webb AJ, Jandu HK, et al. Prediction of Side Effects from Breast Radiation Therapy-Integration of Clinical and Genomic Data. Int J Radiat Oncol Biol Phys. 2026. PMID: 42612831. doi:10.1016/j.ijrobp.2026.08.026

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。