基于炎症相关参数建立根治性膀胱切除术术后的列线图预测模型及验证

目的:探讨接受根治性膀胱切除术的膀胱癌患者相关预测因素，建立并验证列线图预测模型。方法:回顾性收集我院2009年01月至2018年01月期间行根治性膀胱切除术患者的实验室检查和病理结果等临床资料，由术前的血常规结果计算中性粒细胞与淋巴细胞比值（NLR）、血小板与淋巴细胞比值（PLR）和全身免疫炎症指数（SII）。根据约登指数计算SII及NLR的最佳分界值，术后对患者进行随访，使用多因素Cox回归模型分析影响患者术后总生存率的独立危险因素，然后将独立危险因素纳入并构建预测非肌层浸润性膀胱癌（NMIBC）患者3、5年总生存率的列线图，并对模型的预测准确性进行外部验证，通过一致性指数（C指数）和校准曲线来确定列线图的预测精度和一致性。结果:建模组患者中位总生存期（OS）为21个月(1~66个月)，1年、3年和5年的OS率分别为85.2%、68.5%和59.1%。多因素分析显示T分期、N分期、SII和NLR是膀胱癌根治性膀胱切除术后患者的独立危险因素。SII的ROC曲线下面积（AUC）大于NLR，差异有统计学意义，SII预测患者总生存率的准确度更高。我们建立了一个预测根治性膀胱切除术后OS的列线图预测模型，C指数为0.87(95%CI 0.83~0.90)，并对模型进行外部验证，校正曲线显示预测和观察的3、5年生存率之间有很好的一致性。结论:本研究建立的接受根治性膀胱切除术的膀胱癌患者列线图预测模型对膀胱癌患者总生存率具有较高的预测价值，验证相关指标能有效预测患者的预后。

The development and validation of a predictive model for columnar maps after radical cystectomy based on inflammation-related parameters

Objective:To explore the predictive factors associated with bladder cancer patients undergoing total bladder resection,and to develop and validate a columnar graph prediction model.Methods:Clinical data including laboratory tests and pathological findings of patients who underwent total bladder resection in our hospital between January 2009 and January 2018 were retrospectively collected,and neutrophil-to-lymphocyte ratio (NLR),platelet-to-lymphocyte ratio (PLR) and SII values were calculated from the preoperative routine blood results.The optimal cutoff values for SII and NLR were calculated based on the Jorden index.Patients were followed up after surgery and independent risk factors affecting overall survival after surgery were analysed using a multifactorial Cox regression model,and then independent risk factors were incorporated and constructed to predict the overall survival of NMIBC patients at 3 and 5 years,and the predictive accuracy of the model was externally validated by the consistency index (C index) and calibration curves to determine the predictive accuracy and consistency of the column line graphs.Results:The median overall survival (OS) of patients of modeling set was 21 months (1~66 months),with OS rates of 85.2%,68.5% and 59.1% at 1,3 and 5 years,respectively.A multifactorial analysis showed that stage T,N,SII and NLR were independent risk factors for patients after radical cystectomy for bladder cancer.The area under the ROC curve (AUC) for SII was greater than the NLR,a statistically significant difference,and SII was more accurate in predicting overall patient survival.A nomogram prediction plot for predicting OS after radical cystectomy with a C index of 0.87 (95%CI 0.83~0.90) was developed and the model will be externally validated,with calibration curves showing good agreement between predicted and observed 3- and 5-year survival rates.Conclusion:The column line graph prediction model developed in this study for bladder cancer patients undergoing total bladder resection has high predictive value for them and validates that the relevant indicators are effective in predicting patient prognosis.