开发并验证脉络膜黑色素瘤患者总生存和肿瘤特异性生存的预测模型

Develop and validate a predictive model for overall survival and tumor-specific survival in patients with choroidal melanoma

目的利用从监测、流行病学和终末结果研究计划(SEER)数据库中提取的数据构建列线图,预测脉络膜黑色素瘤(CM)患者的总体生存率(OS)和癌症特异性生存率(CSS),评估CM患者的流行病学特征、生存期和预后影响因素。方法从SEER数据库中提取了2010年至2020年间诊断为CM的患者数据,纳入的患者按7:3的比例随机分为训练集(n=1841)和验证集(n=789)。在训练集中,进行单变量Cox回归分析;然后将其纳入多变量Cox比例风险回归模型。在多变量Cox回归模型中筛选独立影响因素,分别构建预测CM的3、5年OS和CSS的列线图。使用决策曲线分析(DCA),通过量化列线图辅助决策的净收益来评估预测模型的临床效用,并且与SEER stage模型比较。根据已建立的列线图获得个体风险评分。结果共有2630例患者被纳入研究。结果显示,性别、年龄、肝转移、手术、放疗和化疗是影响CM患者OS的独立危险因素。年龄、肝转移、手术和化疗是影响CM患者CSS的独立危险因素。CM的3、5年OS和CSS的列线图表明列线图具有较强的区分能力。此外,在OS和CSS的验证集中,DCA提示列线图具有良好的临床潜在价值。Kaplan-Meier(K-M)曲线显示,在训练集和验证集中,高危组患者的OS和CSS率均低于低危组。结论年龄、肝转移、手术和化疗是共同预测CM患者OS和CSS的影响因素。基于SEER数据库建立一个对CM患者相对完善、准确的预后列线图模型,经过校准及后续进一步完善的列线图模型可在临床中应用,在患者治疗及预后中提供指导。

Objective To construct nomograms using data extracted from the surveillance,epidemiology,and end results(SEER)program database to predict overall survival(OS)and cancer-specific survival(CSS)for patients with choroidal melanoma(CM),and to evaluate the epidemiological characteristics,survival periods,and prognostic factors of CM patients.Methods Data on patients diagnosed with CM from 2010 to 2020 were extracted from the SEER database.The included patients were randomly divided into a training set(n=1,841)and a validation set(n=789)at a 7:3 ratio.Univariate Cox regression analysis was conducted in the training set,followed by incorporation into a multivariate Cox proportional hazards regression model.Independent influencing factors were screened in the multivariate Cox regression model to construct nomograms predicting 3-year and 5-year OS and CSS for CM.Decision curve analysis(DCA)was used to assess the clinical utility of the prediction models by quantifying the net benefit of the nomograms in decision-making support,and comparisons were made with the SEER stage model.Individual risk scores were obtained based on the established nomograms.Results A total of 2,630 patients were included in the study.The results indicated that gender,age,liver metastasis,surgery,radiotherapy,and chemotherapy were independent risk factors affecting OS in CM patients.Age,liver metastasis,surgery,and chemotherapy were independent risk factors affecting CSS in CM patients.The nomograms for 3-year and 5-year OS and CSS showed strong discriminative ability.Furthermore,in the validation set for OS and CSS,DCA indicated that the nomograms had good clinical potential.Kaplan-Meier(K-M)curves demonstrated that in both the training and validation sets,patients in the high-risk group had significantly lower OS and CSS rates compared to those in the low-risk group.Conclusion Age,liver metastasis,surgery,and chemotherapy are common predictors of OS and CSS in CM patients.A relatively comprehensive and accurate prognostic nomogram model based on the SEER database has been established.After calibration and further refinement,this nomogram model can be applied clinically to guide the treatment and prognosis of patients.