头颈部滑膜肉瘤临床预测模型的构建和验证:基于SEER数据库

Synovial cell sarcoma of the head and neck clinical features and prognosis:a population-based analysis

目的本研究旨在探索头颈部滑膜肉瘤(synovial cell sarcoma of the head and neck,SCSHN)患者的临床病理特征和生存预后。同时,通过开发列线图来预测SCSHN患者的总生存(overall survival,OS)和疾病特异性生存(diseasespecific survival,DSS),从而为患者提供更准确的风险分层。方法回顾性收集了2004年至2019年期间来自监测、流行病学和最终结果(Surveillance Epidemiology and End Results,SEER)数据库中SCSHN患者的数据。利用Kaplan-Meier进行生存分析,同时利用单变量和多变量Cox分析确定其预后影响因素,并构建OS和DSS列线图。此外,使用C指数、时间相关的受试者操作特征曲线以及校准图验证了列线图的性能。结果共纳入了166例患者,平均年龄为36岁。5年OS率和DSS率分别为71.0%和74.8%。大多数患者接受了手术(84.9%)和放疗(68.1%)。单因素分析结果显示,年龄、肿瘤大小、SEER分期和是否接受放疗是OS和DSS的重要预后因素。在多因素Cox回归分析中,年龄≤50岁、肿瘤直径>5cm和SEER分期为远处病变的患者具有较差的OS和DSS。基于多因素筛选出的预后因素,本研究开发了相应的列线图,并通过C指数、受试者操作特征曲线和曲线下面积以及校准曲线评估和验证列线图的性能,而且与传统的SEER分期系统相比,该列线图预测SCSHN患者的OS和DSS有更高的准确性和更好的性能。此外,基于列线图开发新型风险分层系统,可观察到高风险亚组患者的OS和DSS显著差于中、低风险亚组(P<0.001)。结论本研究构建的列线图方面优于传统的SEER分期,可以对SCSHN患者的长期生存进行个体化预测,并根据风险有效地指导治疗。

Objective Synovial sarcoma of the head and neck(SSHN)is a rare malignancy with few data reported.This study aimed to explore the clinicopathological features and survival prognosis of SCSHN.Besides,we aimed to develop Nomograms to predict the overall survival(OS)and disease-specific survival(DSS)of SCSHN patients,thereby providing more accurate risk stratification for patients.Method We retrospectively collected data from patients with SCSHN in the Surveillance Epidemiology and End Results(SEER)database between 2004 and 2019.Survival analysis was performed by using Kaplan-Meier,univariate and multivariate Cox analyses were used to identify prognostic factors,and OS and DSS Nomograms were constructed.In addition,the performance of the Nomograms were validated using the C-index,timedependent receiver operating characteristic curve,and calibration plot.Result A total of 166 patients were identified,with a mean age of 36 years.The 5-year OS and DSS rates were 71.0%and 74.8%,respectively.Most patients underwent surgery(84.9%)and radiotherapy(68.1%).There were significant differences in OS and DSS between patients with different subgroups of age,tumor size,SEER stage,and radiotherapy.In multivariate Cox regression,patients with age≤50 years,tumor diameter>5cm,and distant disease had poorer OS and DSS.Based on this,this study developed and validated the corresponding Nomograms to reliably estimate OS and DSS in SCSHN patients,with more satisfactory performance compared with the traditional SEER staging system.In addition,stratification systems were developed based on Nomograms.it was found that the OS and DSS of patients in the high-risk subgroup were significantly worse than those in the intermediate and low-risk subgroups(P<0.001).Conclusion The OS-and DSS-nomogram are superior to traditional SEER staging in individualized prediction of long-term survival and help guide risk-adaptive therapy in patients with SCSHN.