| Objective:To explore the prognostic independent risk factors affecting the OS and CSS of LACC,and construct a nomogram prediction model to predict the OS and CSS of patients with LACC based on the independent risk factors.To compare the survival outcomes of patients with LACC after different treatment methods.Methods:1870 patients with LACC admitted to the SEER database between 2004and 2014 are selected as subjects.The diagnostic year,age,race,marital status,clinical stage,tumor diameter,pathological type,degree of differentiation,treatment methods,survival status,survival time and cause of death are extracted.The included LACC patients are randomly divided into training set(1309)and validation set(561)according to the principle of 7:3.In the training set,use the Cox to analyze the independent prognostic risk factors of OS and CSS.The nomogram prediction models of OS and CSS at 1,3,5,and 8 years are constructed based on the independent prognostic risk factors.The C-index、AUC、calibration curve and DCA are used to evaluate and verify the calibration degree,fit degree,and clinical benefit of the prediction model in the training set and validation set.Overall data are analyzed for survival with different treatment methods.Kaplan-Meier is used to calculate overall survival and cancer-specific survival,and survival curves are plotted for different subgroups of treatment mothods,and the log-rank test is used for differences in survival.P<0.05 is considered statistically significant.Results:1.Analysis of prognostic factors:Age,tumor diameter,pathological type,degree of differentiation and treatment methods are independent prognostic risk factors for LACC.68～95 years old(OS:HR=2.619,95%CI:1.960～3.498,P<0.001)(CSS:HR=2.165,95%CI:1.539～3.045,P<0.001),tumor diameter≥6.70cm(OS:HR=1.291,95%CI:1.033～1.613,P=0.025)(CSS:HR=1.402,95%CI:1.095～1.794,P=0.007),other pathological type(OS:HR=1.374,95%CI:1.016～1.859,P=0.039),degree of differentiation are poor differentiation(OS:HR=2.107,95%CI:1.282～3.464,P=0.003)(CSS:HR=2.392,95%CI:1.353～4.226,P=0.003)and undifferentiation(OS:HR=3.658,95%CI:1.768～7.569,P<0.001)(CSS:HR=4.236,95%CI:1.886～9.512,P<0.001),treatment methods are surgery combined with chemotherapy(OS:HR=2.179,95%CI:1.034～4.593,P=0.041)(CSS:HR=2.745,95%CI:1.290～5.842,P=0.009),radiotherapy(OS:HR=2.834,95%CI:1.942～4.136,P<0.001)(CSS:HR=2.230,95%CI:1.424～3.493,P<0.001)and untreated(OS:HR=4.093,95%CI:2.479～6.757,P<0.001)(CSS:HR=4.337,95%CI:2.522～7.456,P<0.001)are independent risk factors for LACC patients.2.Construct,evaluate and validate the nomogram:In this study,the nomogram prediction models of OS and CSS in LACC patients are successfully constructed,and the models are evaluated and validated in the training set and validation set.The results show that the C-index of OS and CSS in the training set are 0.735(95%CI:0.676～0.788)and 0.765(95%CI:0.690～0.844),in the validation set are 0.726(95%CI:0.664～0.770)and 0.747(95%CI:0.683～0.828);the AUC of OS and CSS in the training set and validation set are between 0.50～0.70 in the 1-year and 3-year os in training set,and between 0.70～0.90 of others,indicating that the models have a good degree of discrimination.The 1-year,3-year,5-year,and 8-year calibration curves of OS and CSS in the training set and validation set are all close to the standard line,indicating that the models have a good degree of calibration.The 1-year,3-year,5-year and 8-year DCA curves of OS and CSS of the training set and validation set are all above the“None line”and“All line”,indicating that the models have strong clinical practicability.3.Survival analysis of different treatment methods:There are differences in survival analysis among patients with different treatment methods(OS:c~2=161.190,P<0.001)(CSS:c~2=109.744,P<0.001).1-year,3-year,5-year,8-year overall survival and cancer-specific survival:the overall higher are surgery combined with radiotherapy(OS::96.28%,87.00%,82.38%,80.84%)(CSS:96.28%,87.56%,83.44%,81.88%),surgery(OS:95.51%,85.50%,81.17%,78.62%)(CSS:97.07%,88.48%,85.77%,83.39%),surgery combined chemoradiotherapy(OS:96.14%,83.80%,79.68%,74.53%)(CSS:96.14%,85.04%,81.70%,78.76%),and the differences are not statistically significant among the three groups.Followed by radiotherapy combined with chemotherapy(OS:93.31%,77.23%,71.80%,67.16%)(CSS:93.91%,79.61%,74.41%,71.59%).While surgery combined with chemotherapy(OS:95.24%,57.14%,47.14%,47.14%)(CSS:95.24%,57.14%,52.38%,47.14%),chemotherapy(OS:84.62%,53.85%,43.08%,43.08%)(CSS:91.67%,75.00%,65.63%,52.50%),radiotherapy(OS:81.08%,50.09%,41.61%,35.41%)(CSS:91.67%,75.00%,65.63%,52.50%)are overall lower,and the differences are not statistically significant among the three groups.Conclusions:Age,tumor diameter,pathological type,degree of differentiation and treatment methods are independent prognostic risk factors for LACC patients.The prediction model based on the above factors has good discrimination,fit degree,and clinical practicability.Surgery,surgery combined with radiotherapy,surgery combined chemoradiotherapy can significantly improve the survival outcome of patients with LACC,and radiotherapy combined with chemotherapy followed. |
