Zhang Lijuan, Zhang Zhihong, Wang Yiqiao, Zhu Yu, Wang Ziying, Wan Hongwei
Department of Nursing, Shanghai Proton and Heavy Ion Center, Fudan University Cancer Hospital; Shanghai Key Laboratory of Radiation Oncology; and Shanghai Engineering Research Center of Proton and Heavy Ion Radiation Therapy, Shanghai 201315 China.
Columbia University, New York City, NY 10027, United States.
Radiother Oncol. 2025 Mar;204:110712. doi: 10.1016/j.radonc.2025.110712. Epub 2025 Jan 9.
Few studies have examined the factors associated with xerostomia during proton and carbon ion radiotherapy for head and neck cancer (HNC), which are reported to have fewer toxic effects compared to traditional photon-based radiotherapy. This study aims to evaluate the performance of machine learning approaches in predicting grade 2 + xerostomia in adults with HNC receiving proton and carbon ion radiotherapy.
A retrospective study involving 1,769 adults with HNC who completed proton or carbon ion radiotherapy was conducted. Xerostomia was graded using the Radiation Therapy Oncology Group criteria. Eight machine learning models with different combinations sampling methods and class weights were compared to identify the model with the highest balanced accuracy.
The mean age of patients was 47.8 years (range 18-80), with 33.5 % female. The average total radiation dose was 71.0 GyE (SD = 5.7). Grade 1 xerostomia was recorded in 572 patients (32.3 %) and grade 2 in 103 patients (5.8 %). No cases of grade 3 or higher xerostomia were reported. A support vector machine with a linear kernel, a 1:2 positive-to-negative class weight, and SMOTE oversampling achieved the highest balanced accuracy (0.66) and AUC-ROC (0.69) for predicting grade 2 xerostomia, outperforming the logistic regression model (balanced accuracy:0.50, AUC-ROC. 0.67).
The prevalence of grade 2 radiation-induced xerostomia during proton and carbon ion radiotherapy was low in adults with HNC, posing challenges for accurate prediction. Further research is needed to develop improved methods for predicting xerostomia during proton and carbon ion radiotherapy.
很少有研究探讨头颈部癌（HNC）质子和碳离子放射治疗期间口干症的相关因素，据报道，与传统光子放射治疗相比，质子和碳离子放射治疗的毒性作用更少。本研究旨在评估机器学习方法在预测接受质子和碳离子放射治疗的HNC成年患者2级及以上口干症方面的性能。
进行了一项回顾性研究，纳入1769例完成质子或碳离子放射治疗的HNC成年患者。采用放射治疗肿瘤学组标准对口干症进行分级。比较了八种具有不同组合采样方法和类别权重的机器学习模型，以确定平衡准确率最高的模型。
患者的平均年龄为47.8岁（范围18 - 80岁），女性占33.5%。平均总辐射剂量为71.0 GyE（标准差 = 5.7）。572例患者（32.3%）记录为1级口干症，103例患者（5.8%）记录为2级口干症。未报告3级或更高级别口干症病例。对于预测2级口干症，具有线性核、正负类别权重为1:2以及SMOTE过采样的支持向量机实现了最高的平衡准确率（0.66）和AUC-ROC（0.69），优于逻辑回归模型（平衡准确率：0.50，AUC-ROC：0.67）。
HNC成年患者在质子和碳离子放射治疗期间2级放射性口干症的患病率较低，这给准确预测带来了挑战。需要进一步研究以开发改进的方法来预测质子和碳离子放射治疗期间的口干症。
