基于SEER数据库构建大唾液腺癌患者的远处转移预测模型

doi:10.6040/j.issn.1673-3770.0.2025.199

摘要/Abstract

摘要： 目的明确大唾液腺癌(major salivary gland cancer, MaSGC)远处转移(distant metastasis, DM)的影响因素,并构建MaSGC发生DM的预测模型。方法纳入来自SEER数据库(Surveillance, Epidemiology, and End Results, SEER)的1 729例MaSGC患者和来自青岛大学附属烟台毓璜顶医院的218例MaSGC患者。单因素与多因素logistic回归分析用于识别MaSGC患者发生DM的风险因素。支持向量机(support vector machine, SVM)、逻辑回归(logistic regression, LR)、自适应提升(adaptive boosting, AdaBoost)、决策树(decision tree, DT)、随机森林(random forest, RF)和极端梯度提升(eXtreme Gradient Boosting, XGB)六种机器学习算法用于建立预测模型,并经外部验证。准确率、召回率、受试者操作特征曲线(receiver operating characteristic, ROC)下面积(area under the ROC, AUC)、精确度和F1分数用于验证模型优劣。SHapley可加性解释模型(SHapley Additive exPlanation, SHAP)用于模型可解释性分析。结果研究发现,性别(OR=0.20,P<0.001)、肿瘤原发部位(OR=2.83,P=0.014)、T分期(OR=1.88,P=0.004)、肿瘤组织学类型(P=0.008)、肿瘤大小(OR=1.84,P<0.001)、淋巴结比率(lymph nodes ratio, LNR)(OR=7.67,P<0.001)是MaSGC患者远处转移的显著影响因素。基于以上六个影响因素建立的XGB预测模型明显优于其他预测模型(训练集、验证集和测试集的AUC分别为 0.97、0.94和0.75)。结论性别、肿瘤原发部位、T分期、肿瘤组织学类型、肿瘤大小、LNR是预测MaSGC患者远处转移的重要危险因素,基于上述变量构建的XGB模型可实现对MaSGC患者的DM风险进行个体化预测。

关键词: 大唾液腺癌, 远处转移, 机器学习, 预测模型

Abstract: Objective To identify the factors influencing distant metastasis(DM)in major salivary gland cancer(MaSGC)and to develop a predictive model for the occurrence of DM in MaSGC. Methods The study included 1729 MaSGC patients from the Surveillance, Epidemiology, and End Results(SEER)database,as well as 218 patients from the Affiliated Yantai Yuhuangding Hospital of Qingdao University. Univariate and multivariate logistic regression analyses were used to screen for risk factors for DM in MaSGC patients. Predictive models were built utilizing six machine learning models including support vector machine(SVM), logistic regression(LR), adaptive boosting(AdaBoost), decision tree(DT), random forest(RF), eXtreme Gradient Boosting(XGB)and externally validated. The performance of the model was evaluated using a range of metrics, including accuracy, recall, area under the receiver operating characteristic(ROC)curve(AUC), precision, and the F1 score. The SHapley Additive exPlanation(SHAP)was utilized for the purpose of generating explanations of the model's decisions. Results Gender(OR=0.20, P<0.001), primary tumor site(OR=2.83, P=0.014), T stage(OR=1.88, P=0.004), histological type(P=0.008), tumor size(OR=1.84, P<0.001), and lymph node ratio(LNR)(OR=7.67, P<0.001)were found to be significant influencing factorsfor DM in MaSGC patients. The XGB prediction model built based on the above six factors significantly outperforms the other prediction models(AUCs of 0.97, 0.94, and 0.75 for the internal training cohort, internal validation cohort, and external validation cohort, respectively). Conclusion Gender, primary tumor site, T stage, histological type, tumor size, and LNR were significant risk factorsfor predicting DM in MaSGC patients. The development of the XGB model was predicated on the aforementioned factors, with the objective being the provision of individualised risk predictions for DM in MaSGC patients.

Key words: Major salivary gland cancer, Distant metastasis, Machine learning, Prediction model

中图分类号:

R762

陈曦,姚依松,方钰慧,李东宪,李玉梅,宋西成. 基于SEER数据库构建大唾液腺癌患者的远处转移预测模型[J]. 山东大学耳鼻喉眼学报, 2026, 40(4): 81-89.

CHEN Xi, YAO Yisong, FANG Yuhui, LI Dongxian, LI Yumei, SONG Xicheng. Machine learning models for predicting distant metastasis in patients with major salivary gland cancers based on SEER database[J]. Journal of Otolaryngology and Ophthalmology of Shandong University, 2026, 40(4): 81-89.

参考文献

[1] Ullah A, Khan J, Waheed A, et al. Mucoepidermoid carcinoma of the salivary gland: demographics and comparative analysis in U.S. children and adults with future perspective of management[J]. Cancers, 2023, 15(1): 250. DOI:10.3390/cancers15010250
[2] Del Signore AG, Megwalu UC. The rising incidence of major salivary gland cancer in the United States[J]. Ear Nose Throat J, 2017, 96(3): E13-E16. DOI:10.1177/014556131709600319
[3] Shi X, Dong F, Wei WJ, et al. Prognostic significance and optimal candidates of primary tumor resection in major salivary gland carcinoma patients with distant metastases at initial presentation: a population-based study[J]. Oral Oncol, 2018, 78: 87-93. DOI:10.1016/j.oraloncology.2018.01.009
[4] Nam SJ, Roh JL, Cho KJ, et al. Risk factors and survival associated with distant metastasis in patients with carcinoma of the salivary gland[J]. Ann Surg Oncol, 2016, 23(13): 4376-4383. DOI:10.1245/s10434-016-5356-3
[5] Li JH, Rao YF, Wang XY, et al. Prognostic effects of previous cancer history on patients with major salivary gland cancer[J]. Oral Dis, 2024, 30(2): 492-503. DOI:10.1111/odi.14530
[6] Shi JY, Fan YJ, Long JZ, et al. Development and validation of nomograms to predict risk and prognosis in salivary gland carcinoma patient with distant metastases[J]. Ear Nose Throat J, 2023: 01455613231212060. DOI:10.1177/01455613231212060
[7] Tekke?瘙塂in A(·overI). Artificial intelligence in healthcare: past, present and future[J]. Anatol J Cardiol, 2019, 22(S2): 8-9. DOI:10.14744/anatoljcardiol.2019.28661
[8] Vandenbroucke JP, von Elm E, Altman DG, et al. Strengthening the reporting of observational studies in epidemiology(STROBE): explanation and elaboration[J]. PLoS Med, 2007, 4(10): e297. DOI:10.1371/journal.pmed.0040297
[9] Moons KGM, Altman DG, Reitsma JB, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis(TRIPOD): Explanation and Elaboration[J]. Ann Intern Med, 2015, 162(1): W1-W73. DOI:10.7326/m14-0698
[10] Jiang HL, Mao HF, Lu HM, et al. Machine learning-based models to support decision-making in emergency department triage for patients with suspected cardiovascular disease[J]. Int J Med Inform, 2021, 145: 104326. DOI:10.1016/j.ijmedinf.2020.104326
[11] Mimica X, McGill M, Hay A, et al. Distant metastasis of salivary gland cancer: Incidence, management, and outcomes[J]. Cancer, 2020, 126(10): 2153-2162. DOI:10.1002/cncr.32792
[12] Ali S, Bryant R, Palmer FL, et al. Distant metastases in patients with carcinoma of the major salivary glands[J]. Ann Surg Oncol, 2015, 22(12): 4014-4019. DOI:10.1245/s10434-015-4454-y
[13] 张明君, 姚依松, 陈曦, 等. 唾液腺黏液表皮样癌局部淋巴结转移风险因素分析及列线图模型构建[J]. 中华耳鼻咽喉头颈外科杂志, 2024, 59(6): 614-620. DOI: 10.3760/cma.j.cn115330-20231106-00187
[14] Shi WY, Wu WM, Zhang LY, et al. Prognosis of thyroid carcinoma patients with osseous metastases: an SEER-based study with machine learning[J]. Ann Nucl Med, 2023, 37(5): 289-299. DOI:10.1007/s12149-023-01826-z
[15] Haderlein M, Scherl C, Semrau S, et al. High-grade histology as predictor of early distant metastases and decreased disease-free survival in salivary gland cancer irrespective of tumor subtype[J]. Head Neck, 2016, 38(S1): E2041-8. DOI:10.1002/hed.24375
[16] Costantino A, Canali L, Festa BM, et al. Development of machine learning models to predict lymph node metastases in major salivary gland cancers[J]. Eur J Surg Oncol, 2023, 49(9): 106965. DOI:10.1016/j.ejso.2023.06.017
[17] Ali S, Bryant R, Palmer FL, et al. Distant metastases in patients with carcinoma of the major salivary glands[J]. Ann Surg Oncol, 2015, 22(12): 4014-4019. DOI:10.1245/s10434-015-4454-y
[18] Schwentner I, Obrist P, Thumfart W, et al. Distant metastasis of parotid gland tumors[J]. Acta Oto Laryngol, 2006, 126(4): 340-345. DOI:10.1080/00016480500401035
[19] Wen JM, Wei Y, Jabbour SK, et al. Comprehensive analysis of prognostic value of lymph node staging classifications in patients with head and neck squamous cell carcinoma after cervical lymph node dissection[J]. Eur J Surg Oncol, 2021, 47(7): 1710-1717. DOI:10.1016/j.ejso.2021.01.020
[20] Luo ZY, Hei H, Qin JW, et al. Lymph node ratio as a tool to stratify patients with N1b papillary thyroid cancer[J]. Langenbeck’s Arch Surg, 2023, 408(1): 315. DOI:10.1007/s00423-023-03033-w
[21] Kawada K, Taketo MM. Significance and mechanism of lymph node metastasis in cancer progression[J]. Cancer Res, 2011, 71(4): 1214-1218. DOI:10.1158/0008-5472.can-10-3277
[22] Cho JK, Hyun SH, Choi N, et al. Significance of lymph node metastasis in cancer dissemination of head and neck cancer[J]. Transl Oncol, 2015, 8(2): 119-125. DOI:10.1016/j.tranon.2015.03.001
[23] Ettl T, Gosau M, Brockhoff G, et al. Predictors of cervical lymph node metastasis in salivary gland cancer[J]. Head Neck, 2014, 36(4): 517-523. DOI:10.1002/hed.23332
[24] Shi N, Xia W, Ji KT, et al. Anatomy and nomenclature of tree shrew lymphoid tissues[J]. Exp Anim, 2022, 71(2): 173-183. DOI:10.1538/expanim.21-0150
[25] Meng F, Yuan JH, Zhang X, et al. Influence of parotid lymph node metastasis on distant metastasis in parotid gland cancer[J]. Front Oncol, 2023, 13: 1244194. DOI:10.3389/fonc.2023.1244194
[26] Benchetrit L, Mehra S, Mahajan A, et al. Major salivary gland cancer with distant metastasis upon presentation: patterns, outcomes, and imaging implications[J]. Otolaryngol - head Neck Surg, 2022, 167(2): 305-315. DOI:10.1177/01945998211058354
[27] Xia RH, Zhou RR, Tian Z, et al. High expression of H3K9me3 is a strong predictor of poor survival in patients with salivary adenoid cystic carcinoma[J]. Arch Pathol Lab Med, 2013, 137(12): 1761-1769. DOI:10.5858/arpa.2012-0704-oa
[28] Dai W, Zhou Q, Xu ZF, et al. Expression of TMPRSS4 in patients with salivary adenoid cystic carcinoma: correlation with clinicopathological features and prognosis[J]. Med Oncol, 2013, 30(4): 749. DOI:10.1007/s12032-013-0749-7
[29] Jaehne M, Roeser K, Jaekel T, et al. Clinical and immunohistologic typing of salivary duct carcinoma: a report of 50 cases[J]. Cancer, 2005, 103(12): 2526-2533. DOI:10.1002/cncr.21116

多维度评价

Viewed

Full text

Abstract

Cited

Shared

Discussed