基于临床及剂量学特征构建模型预测鼻咽癌放疗后颞叶损伤的研究

Research on constructing a model based on clinical and dosimetric characteristics to predict temporal lobe injury after radiotherapy for nasopharyngeal carcinoma

摘要

目的: 基于鼻咽癌患者的临床及剂量学特征构建模型，并评估其对放疗后出现放射性颞叶损伤（temporal lobe injury，TLI）的预测价值。方法: 回顾性分析2018年6月—2023年11月365例（包括TLI组42例和无TLI组323例）鼻咽癌患者的临床资料及剂量学特征，使用单因素及多因素Logistic回归分析筛选出与放射性TLI相关的独立预测因素。利用随机森林机器学习算法构建临床特征模型及临床+剂量学特征联合模型。绘制受试者操作特征（receiver operating characteristic，ROC）曲线，计算曲线下面积（area under curve，AUC）、灵敏度、特异度等参数，绘制校准曲线及决策曲线分析评估模型的预测效能。结果: 通过单因素、多因素Logistic回归分析得出临床特征中的T分期及剂量学特征中的D_{1 cm3}、V70 Gy这3个特征的差异有统计学意义（P<0.05）。联合模型和临床模型的AUC值、灵敏度、特异度和准确率分别为0.853和0.635、66.67%和85.71%、86.38%和39.94%、84.00%和45.00%。DeLong检验显示，联合模型与临床模型的预测效能差异具有统计学意义（P<0.05）。校准曲线和决策曲线分析结果均表明联合模型具有更高的校准度和临床净获益。结论: T分期、D_{1 cm3}、V70 Gy是放射性TLI的独立影响因素，基于三者的联合模型对鼻咽癌放疗后TLI具有较高的预测能力。

Abstract

Objective: To establish a prediction model based on the clinical and dosimetric characteristics of patients with nasopharyngeal carcinoma（NPC），and assess its predictive efficacy for radiation-induced temporal lobe injury（TLI）following radiotherapy. Methods: We performed a retrospective analysis of the clinical data from 365 NPC patients（42 patients in the TLI group and 323 patients in the non-TLI group）who underwent radiotherapy-based comprehensive treatment in the radiotherapy department of our institution from June 2018 to November 2023. Clinical and dosimetric characteristics were collected，and univariate and multivariate logistic regression analyses were employed to identify independent risk factors associated with radiation-induced TLI. The random forest（RF）machine learning algorithm was used to construct a clinical feature model and a combined clinical and dosimetric feature model. Receiver operating characteristic（ROC）curves were plotted，and parameters such as the area under the ROC curve（AUC），sensitivity，and specificity were calculated. Calibration curves and decision curve analysis（DCA）were also conducted to evaluate the predictive performance of the models. Results: Univariate and multivariate logistic regression analyses revealed statistically significant differences in three features：T-stage（clinical characteristic），as well as D _{1 cm 3} and V70Gy（dosimetric characteristics）（all P<0.05）. The AUC values，sensitivity，specificity，and accuracy of the combined model and clinical model were 0.853 and 0.635，66.67% and 85.71%，86.38% and 39.94%，and 84.00% and 45.00%，respectively. The DeLong test indicated a statistically significant difference on predictive performance between combined and clinical models（P<0.05）. Both calibration curves and DCA results demonstrated that the combined model exhibited superior calibration and clinical net benefit. Conclusion: T-stage，D _{1 cm 3}，and V70Gy are independent predictors of radiation-induced temporal lobe injury. The combined model based on these three factors demonstrates high predictive capability for temporal lobe injury following NPC radiotherapy.

关键词

鼻咽癌 / 放射性颞叶损伤 / 机器学习 / 预测模型

Key words

nasopharyngeal carcinoma / radiation-induced temporal lobe injury / machine learning / predictive model

参考文献

[1] SU Z Y, SIAK P Y, LWIN Y Y, et al. Epidemiology of nasopharyngeal carcinoma: current insights and future outlook[J]. Cancer Metastasis Rev, 2024, 43(3):919-939

[2] OUYANG P Y, ZHANG B Y, GUO J G, et al. Deep learning—based precise prediction and early detection of radiation—induced temporal lobe injury for nasopharyngeal carcinoma[J]. EClinicalMedicine, 2023, 58:101930

[3] HOU J, HE Y, LI H, et al. Evolution of radiation—induced temporal lobe injury after intensity—modulated radiation therapy in nasopharyngeal carcinoma: a large cohort retrospective study[J]. Radiat Oncol, 2024, 19(1):9

[4] HE Y Q, WANG T M, YANG D W, et al. A comprehensive predictive model for radiation—induced brain injury in risk stratification and personalized radiotherapy of nasopharyngeal carcinoma[J]. Radiother Oncol, 2024, 190:109974

[5] LI Y, GONG F, GUO Y, et al. Predictive accuracy of machine learning for radiation—induced temporal lobe injury in nasopharyngeal carcinoma patients: a systematic review and meta—analysis[J]. Transl Cancer Res, 2023, 12(9):2361-2370

[6] 李兵, 朱锡旭, 於文雪, 等. 基于轮廓优化技术在鼻咽癌调强放疗计划设计中的应用[J]. 南京医科大学学报（自然科学版）, 2011, 31(6): 898-902

[7] LI B, ZHU X X, YU W X, et al. Application of contour optimization technique in intensity—modulated radiotherapy treatment planning for nasopharyngeal carcinoma[J]. Journal of Nanjing Medical University(Natural Sciences), 2011, 31(6): 898-902

[8] GUAN Y, YU B B, LIU S, et al. Research trends of radiation—induced temporal lobe injury in patients with nasopharyngeal carcinoma from 2000 to 2022: a bibliometric analysis[J]. Radiat Oncol, 2023, 18(1):151

[9] SU S F, HUANG S M, HAN F, et al. Analysis of dosimetric factors associated with temporal lobe necrosis(TLN) in patients with nasopharyngeal carcinoma(NPC) after intensity modulated radiotherapy[J]. Radiat Oncol, 2013, 8:17

[10] BENTZEN S M, CONSTINE L S, DEASY J O, et al. Quantitative Analyses of Normal Tissue Effects in the Clinic(QUANTEC): an introduction to the scientific issues[J]. Int J Radiat Oncol Biol Phys, 2010, 76(3 Suppl):S3-S9

[11] WEN D W, LIN L, MAO Y P, et al. Normal tissue complication probability(NTCP) models for predicting temporal lobe injury after intensity—modulated radiotherapy in nasopharyngeal carcinoma: a large registry—based retrospective study from China[J]. Radiother Oncol, 2021, 157:99-105

[12] BAO D, ZHAO Y, LI L, et al. A MRI—based radiomics model predicting radiation—induced temporal lobe injury in nasopharyngeal carcinoma[J]. Eur Radiol, 2022, 32(10):6910-6921

[13] WANG Y X, KING A D, ZHOU H, et al. Evolution of radiation—induced brain injury: MR imaging—based study[J]. Radiology, 2010, 254(1):210-218

[14] 区晓敏, 杨昭志, 周鑫, 等. 放射性脑损伤的研究进展[J]. 中华放射肿瘤学杂志, 2026, 35(3): 213-222

[15] OU X M, YANG Z Z, ZHOU X, et al. Research progress on radiation—induced brain injury[J]. Chinese Journal of Radiation Oncology, 2026, 35(3): 213-222

[16] HOU J, LI H, ZENG B, et al. MRI—based radiomics nomogram for predicting temporal lobe injury after radiotherapy in nasopharyngeal carcinoma[J]. Eur Radiol, 2022, 32(2):1106-1114

[17] DONG J, NG W T, WONG C H L, et al. Dosimetric parameters predict radiation—induced temporal lobe necrosis in nasopharyngeal carcinoma patients: a systematic review and meta—analysis[J]. Radiat Oncol, 2024, 195:110258

[18] OUYANG H, LIU Y, HE X, et al. A comparative study of different parameter estimation methods for predictive models of normal tissue complication probability(NTCP) of radiation—induced temporal lobe injury following intensity—modulated radiotherapy in nasopharyngeal carcinoma[J]. BMC Cancer, 2025, 25(1):572

[19] GUAN W, XIE K, FAN Y, et al. Development and validation of a nomogram to predict temporal lobe injury in patients with nasopharyngeal carcinoma after intensity—modulated radiotherapy: a propensity score—matched analysis[J]. Front Oncol, 2020, 10:1317

[20] WANG L, QIU T, ZHOU J, et al. A pretreatment multiparametric MRI—based radiomics—clinical machine learning model for predicting radiation—induced temporal lobe injury in patients with nasopharyngeal carcinoma[J]. Head Neck, 2024, 46(9):2132-2144

[21] WU V W C, TAM S Y. Radiation—induced temporal lobe necrosis in nasopharyngeal cancer patients after radical external beam radiotherapy[J]. Radiat Oncol, 2020, 15(1):112

[22] YANG S S, OUYANG P Y, GUO J G, et al. Dosiomics risk model for predicting radiation—induced temporal lobe injury and guiding individual intensity—modulated radiation therapy[J]. Int J Radiat Oncol Biol Phys, 2023, 115(5):1291-1300

[23] HUANG L X, YANG Z X, ZENG Z S, et al. MRI—based radiomics models for the early prediction of radiation—induced temporal lobe injury in nasopharyngeal carcinoma[J]. Front Neurol, 2023, 14:1135978

[24] SUN X, ZHU H, QIAO J, et al. Machine learning—based integration of radiomics and dosiomics for early prediction of radiation—induced temporal lobe injury in nasopharyngeal carcinoma: A multicenter study[J]. J Appl Clin Med Phys, 2026, 27(1):e70420

[25] 刘君懿, 李阳, 王力, 等. 基于剂量组学模型预测鼻咽癌调强放疗后放射性颞叶损伤[J]. 中华放射肿瘤学杂志, 2025, 34(3): 240-248

[26] LIU J Y, LI Y, WANG L, et al. Prediction of radiation—induced temporal lobe injury after intensity—modulated radiotherapy for nasopharyngeal carcinoma based on dosiomics model[J]. Chinese Journal Radiation Oncology, 2025, 34(3): 240-248

[27] 孟慧慧, 余静, 吴飞云, 等. 基于临床及MR影像特征评估直肠癌新辅助放化疗疗效的临床研究[J]. 南京医科大学学报（自然科学版）, 2025, 45(4): 443-452

[28] MENG H H, YU J, WU F Y. Clinical study on evaluating the efficacy of neoadjuvant chemoradiotherapy for rectal cancer based on clinical and MR imaging features[J]. Journal of Nanjing Medical University(Natural Science), 2025, 45(4): 443-452

[29] PAN J, LIANG J, LI Y, et al. Composite quantitative structural magnetic resonance imaging—based risk scoring model for predicting radiation—induced temporal lobe necrosis in nasopharyngeal carcinoma: a novel risk stratification model[J]. Radiat Oncol, 2025, 20(1):160

基金资助

国家重点研发计划(2022YFC2401604)